Acerca de este libro

Esta es una copia digital de un libro que, durante generaciones, se ha conservado en las estanteras de una biblioteca, hasta que Google ha decidido
escanearlo como parte de un proyecto que pretende que sea posible descubrir en lnea libros de todo el mundo.
Ha sobrevivido tantos aos como para que los derechos de autor hayan expirado y el libro pase a ser de dominio pblico. El que un libro sea de
dominio pblico significa que nunca ha estado protegido por derechos de autor, o bien que el perodo legal de estos derechos ya ha expirado. Es
posible que una misma obra sea de dominio pblico en unos pases y, sin embargo, no lo sea en otros. Los libros de dominio pblico son nuestras
puertas hacia el pasado, suponen un patrimonio histrico, cultural y de conocimientos que, a menudo, resulta difcil de descubrir.
Todas las anotaciones, marcas y otras seales en los mrgenes que estn presentes en el volumen original aparecern tambin en este archivo como
testimonio del largo viaje que el libro ha recorrido desde el editor hasta la biblioteca y, finalmente, hasta usted.

Normas de uso

Google se enorgullece de poder colaborar con distintas bibliotecas para digitalizar los materiales de dominio pblico a fin de hacerlos accesibles
a todo el mundo. Los libros de dominio pblico son patrimonio de todos, nosotros somos sus humildes guardianes. No obstante, se trata de un
trabajo caro. Por este motivo, y para poder ofrecer este recurso, hemos tomado medidas para evitar que se produzca un abuso por parte de terceros
con fines comerciales, y hemos incluido restricciones tcnicas sobre las solicitudes automatizadas.
Asimismo, le pedimos que:

+ Haga un uso exclusivamente no comercial de estos archivos Hemos diseado la Bsqueda de libros de Google para el uso de particulares;
como tal, le pedimos que utilice estos archivos con fines personales, y no comerciales.
+ No enve solicitudes automatizadas Por favor, no enve solicitudes automatizadas de ningn tipo al sistema de Google. Si est llevando a
cabo una investigacin sobre traduccin automtica, reconocimiento ptico de caracteres u otros campos para los que resulte til disfrutar
de acceso a una gran cantidad de texto, por favor, envenos un mensaje. Fomentamos el uso de materiales de dominio pblico con estos
propsitos y seguro que podremos ayudarle.
+ Conserve la atribucin La filigrana de Google que ver en todos los archivos es fundamental para informar a los usuarios sobre este proyecto
y ayudarles a encontrar materiales adicionales en la Bsqueda de libros de Google. Por favor, no la elimine.
+ Mantngase siempre dentro de la legalidad Sea cual sea el uso que haga de estos materiales, recuerde que es responsable de asegurarse de
que todo lo que hace es legal. No d por sentado que, por el hecho de que una obra se considere de dominio pblico para los usuarios de
los Estados Unidos, lo ser tambin para los usuarios de otros pases. La legislacin sobre derechos de autor vara de un pas a otro, y no
podemos facilitar informacin sobre si est permitido un uso especfico de algn libro. Por favor, no suponga que la aparicin de un libro en
nuestro programa significa que se puede utilizar de igual manera en todo el mundo. La responsabilidad ante la infraccin de los derechos de
autor puede ser muy grave.

Acerca de la Bsqueda de libros de Google

El objetivo de Google consiste en organizar informacin procedente de todo el mundo y hacerla accesible y til de forma universal. El programa de
Bsqueda de libros de Google ayuda a los lectores a descubrir los libros de todo el mundo a la vez que ayuda a autores y editores a llegar a nuevas
audiencias. Podr realizar bsquedas en el texto completo de este libro en la web, en la pgina http://books.google.com
This is a reproduction of a library book that was digitized
by Google as part of an ongoing effort to preserve the
information in books and make it universally accessible.

https://books.google.com
 LIBRARY
{}{\0 STATE UNIVERSITY
 |&OC

Age and Origin of ** 26 g,

Fluvial Terraces in the

Central Coast Range,
Western Oregon

U.S. GEOLOGICAL SURVEY BULLETIN 2038

gd an e-24-34 V
GEO
GE75B9) v-293, c=& -0.
+++

+ 2 of 3
Age and Origin of
Fluvial Terraces in the
Central Coast Range,
Western Oregon
By STEPHEN F. PERSONIUS

U.S. GEOLOGICAL SURVEY BULLETIN 2038

Radiocarbon and thermoluminescence dating methods

are used to determine the ages and probable
origins of fluvial terraces

UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1993

#. *]; 7+|
01-141-00 se
 U.S. DEPARTMENT OF THE INTERIOR

BRUCE BABBITT, Secretary

U.S. GEOLOGICAL SURVEY

Robert M Hirsch, Acting Director

For sale by
USGS Map Distribution
Box 25286, Building 810
Denver Federal Center
Denver, CO 80225

Any use of trade, product, or firm names in this publication is for descriptive purposes only and
does not imply endorsement by the U.S. Government

Library of Congress Cataloging-in-Publication Data

Personius, Stephen F.
Age and origin of fluvial terraces in the central Coast Range, western Oregon / by
Stephen F. Personius.
cm. (U.S. Geological Survey bulletin ; 2038)
Includes bibliographical references.
Supt. of Docs. no.: I 19.3:B2038
1. Terraces (Gcology)Oregon. 2. Terraces (Geology)Coast
Ranges. I. Title. II. Series.
QE75.B9 no. 2038
[GB595.07]
557.3 sdc20 9223352
[551.4'42] CIP
 ***...... "****..
****...

*** *... * *

..... "....
*** - ...... " " " " - ...
*** *.

****...... "****...
*... . . . *** * *... . . " " " -- e.
***....... "****....
*** *..... """"- e.
*** *..... "" " - e.
***"....
****... *** *.. ****
*** *...
****.... * *
***...... " ...
... . . . * *** * *.
*** *..... """" " ...
*** - ... . . " "*** * *
........ "---
*** *... * *
..... " ...
***...... "****...

.......
. .. . . *****.

***"...... "****....

****...

****...... "**** ....

*** * *
****...
*** * *

****.. ** * * *
***..
****.... *** * *
...... ....... ***...
***.... . . " .... "- ..
*** ...... "**** ...... "****...
*** *..... "" " - ..... """" " ...
****...
*** *...
****....
*** *..
*** *..

**** ...

**** ...

****...
****....
*** ...

****..

**** ....

****.

*** *..

****...
IV CONTENTS

DiscussionContinued
Relations Between the PHT and Other Coast Range Terraces................................... 52
Conclusion............................................................................................................................. 52
Acknowledgments................................................................................................................. 53
References Cited ................................................................................................................... 53

PLATES

[Plates are in pocket]

Generalized map of terraces along the lower Umpqua River, central Coast Range, western Oregon.
Generalized map of terraces along the lower Smith River, central Coast Range, western Oregon.

: Generalized map of terraces along the lower Siuslaw River, central Coast Range, western Oregon.
Generalized map of terraces along the lower Siletz River, central Coast Range, western Oregon.

FIGURES

Map showing regional tectonic setting of the Pacific Northwest................................................................................ 2

; Map showing the location of major rivers in western Oregon and Washington.........................................................
Stratigraphic columns of three fluvial terrace exposures along the Umpqua River where both thermoluminescence
(TL) and radiocarbon samples were collected ............................................................................................................
3

9
Stratigraphic columns of radiocarbon-dated fluvial sediments used to calculate sedimentation rates........................ 13
Longitudinal profile of terraces along the lower reach of the Umpqua River............................................................. 15

: Photograph of PHT (Pleistocene-Holocene transition) terrace on the Umpqua River near McGee Creek.................
Rod-and-level profile of stripped bedrock strath and PHT (Pleistocene-Holocene transition) terrace near
McGee Creek on the Umpqua River...........................................................................................................................
16

17
Stratigraphic column from a PHT (Pleistocene-Holocene transition) terrace exposure near McGee Creek at
station 88U2............................................ 18
Stratigraphic column from a terrace exposure about 2 km downstream from the highway bridge near Scottsburg
at station 87-U14....................................................... 18
10. Stratigraphic column of a low terrace near the Brandy Bar on the Umpqua River at station 87U13....................... 19
11. Diagrammatic topographic profiles of fluvial and marine terraces in the Reedsport area.......................................... 22
12. Stratigraphic column of an intermediate terrace in Reedsport at station 87U21 ....................................................... 23
13. Stratigraphic columns of two exposures of the high terrace underlying the town of Reedsport................................. 24
14. Stratigraphic column of an exposure of a high terrace along Crestview Drive near Reedsport at station 87U19.... 25
15. Stratigraphic column from an exposure of a high terrace along Henderer Road at station 87U1............................. 26
16. Stratigraphic columns of two exposures of an intermediate terrace along Henderer Road......................................... 28
17. Longitudinal profile of terraces along the lower reach of the Smith River................................................................. 29
18. Stratigraphic column of a terrace exposure on Vincent Creek near the confluence with the Smith River at
station 87-US7.. 30
19. Stratigraphic columns of two terrace exposures upstream from Smith River Falls near Bear Creek......................... 31
20. Diagrammatic topographic profile of the present valley bottom in an abandoned meander bend on the
Smith River.. 32
21. Stratigraphic columns of terrace exposures in The Island on the Smith River........................................................... 33
22. Stratigraphic columns of terrace exposures in the abandoned mender bend near Brainard Creek on the
Smith River... 34
23. Stratigraphic column of an exposure of a high terrace on the Smith River at station 87US2................................... 35
24. Topographic profile across the Siuslaw River near river km. 22.5.............................................................................. 37
25. Longitudinal profile of terraces along the lower reach of the Siuslaw River.............................................................. 38
26. Stratigraphic column of an exposure of a possible PHT (Pleistocene-Holocene transition) terrace at station 88Sl
on the Siuslaw River.................................. 39
 CONTENTS V

27. Stratigraphic columns of exposures of several high terraces on the Siuslaw River.................................................... 40
28. Photograph of basal fluvial gravels overlying sandstone bedrock in a high terrace east of Hanson Creek at
station 87-S2............................................................................................................................................................... 41
29. Longitudinal profile of terraces along the lower reach of the Siletz River.................................................................. 42
30. Stratigraphic columns of exposures of Holocene terraces on the Siletz River............................................................ 44
31. Stratigraphic columns of exposures of late Pleistocene terraces on the Siletz River.................................................. 45
32. Examples of stratigraphic columns of exposures of older terraces on the Siletz River............................................... 47
33. Map of lower Coquille River showing locations of the towns of Bandon, Coquille, and Myrtle Point, and the
confluences of the North, Middle, and South Forks of the Coquille River................................................................. 48
34. Map of the flood plain of the Coquille River near Coquille........................................................................................ 49
35. Stratigraphic columns of two auger cores in abandoned channels on the Coquille River flood plain........................ 50

TABLES

List of radiocarbon ages obtained in this study........................................................................................................... 6

! List of thermoluminescence ages obtained in this study............................................................................................. 8
 AGE AND ORIGIN OF FLUVIAL TERRACES IN THE
CENTRAL COAST RANGE, WESTERN OREGON
By Stephen F. Personius
ABSTRACT
Fluvial terraces are generally poorly preserved along
most rivers in the central Coast Range of western Oregon,
but some terraces along the Umpqua, Smith, Siuslaw, and
Siletz Rivers can be correlated and provide information
about their age and origin. Most terraces on inland reaches
of these rivers are underlain by a bedrock strath and have
a sequence of sediments consisting of a 1- to 3-m-thick
basal deposit of sandy pebble channel gravel overlain by a
2- to 7-m-thick deposit of overbank sand and silt. Terraces
generally converge near the coast; along tidal reaches, ter
races consist of fine sand, mud, and peat of fluvial and
estuarine origin. Only a few of the older, higher terrace
remnants near the coast are underlain by exposed bedrock
straths.
The ages of some younger terrace sediments in the
Coast Range were determined by radiocarbon dating of
detrital charcoal and wood. Radiocarbon analyses from the
silty overbank sediments in the lowest continuous terrace
found along most Coast Range rivers provided generally
consistent early Holocene (710 ka) ages. The regional
distribution and consistent age of this terrace (herein infor
mally referred to as the Pleistocene-Holocene transition or
PHT terrace) indicates that climatic change was proba
bly the cause of stream aggradation in latest Pleistocene
and early Holocene time. Aggradation is thought to have
been mostly related to accelerated landsliding brought on
by the warmer, drier early Holocene climate rather than to
base level rise related to rising sea level or an influx of
glacial outwash sediment. These types of climatic ter
races, which form extensive synchronous surfaces, are rel
atively rare in the Coast Range.
Older terraces are present at elevations up to 200 m
or more above present river levels. Most of the strati
graphic sequences exposed in the older terraces are identi
cal to those in the lower terraces, although clay-rich,
deeply oxidized soils on the higher terraces have obscured
some of the original textures and bedding. Ages from
experimental thermoluminescence (TL) dating of older ter
race sediments from the Umpqua and Smith Rivers
generally increase with increasing terrace height and
ranged from about 15 ka to >200 ka. Ages of older fluvial
terraces near the coast were also estimated by tentatively
correlating them to poorly preserved marine terrace depos
its and assumed correlations with the marine oxygen
isotope record. Although poor terrace preservation pre
vented definitive correlation, reasonable projections of
stream gradients similar to the modern rivers support cor
relation of some older fluvial terraces to marine terraces
and middle to late Pleistocene sea-level high stands. The
nearly continuous PHT terraces, which converge to sea
level near the coast, probably are analogs of these older
fluvial terraces.
Numerous isolated terrace remnants are present at
various elevations along Coast Range streams, but most of
these terraces probably were formed as a result of com
plex responses to periodic floods or isolated sedimenta
tion events like localized landsliding and do not represent
long-term, regional changes in stream conditions. Other
isolated terraces are present on the insides of meander
bends and probably formed as a result of minor vertical
incision that accompanies stream meandering (slip-off"
terraces). Most of these isolated remnants were never parts
of extensive terrace surfaces. Even less common in the
Coast Range are tectonic terraces, examples of which
are presently warped over an anticline in Eocene bedrock
along the Siuslaw River.
INTRODUCTION
SCOPE AND PURPOSE
The Coast Range is a belt of coastal mountains of
moderate elevation that lie between the Olympic Moun
tains of northwestern Washington and the Klamath Moun
tains of southern Oregon and northern California; the
range forms part of the forearc of the Cascadia subduction
zone, an active subduction zone that stretches from north
ern California to British Columbia (fig. 1). Relatively little
is known about the Quaternary history of the Coast Range,
2 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

135 130 125 120

I I I I I I T T

British Columbia

A. North
50 L , American
A Plate

2.
* A
-

2:
Washington
a A * *... alve
// Plate | columbia."
A

45 H
; *

: i
* O r 9 O n

Pacific
Plate

40 -

- I l l I 1 l 1 1A

Figure 1. Map showing regional tectonic setting of the Pacific Northwest, locations of major mountain ranges,
and locations of volcanoes in the Cascade Range (triangles); after Rogers (1988) and Spence (1989).

although several recent studies have addressed deforma
tion of Quaternary deposits along the Oregon and Wash
ington coasts (Adams, 1984; Atwater, 1987, 1992; West
and McCrumb, 1988; Darienzo and Peterson,
Kelsey, 1990; McInelly and Kelsey, 1990; Muhs and
others, 1990; Nelson, 1992; Nelson and Personius, 1991).
In this report, I discuss the distribution of terraces along
several rivers in the central Coast Range of western
1990; Oregon, the methods and results of terrace dating and pos
sible controls on terrace formation as the basis for a
 INTRODUCTION

126 124 122

48 O2,
Puget
100 km ** 2C. Sound
25%,

Washington
QD
O)
C
(U
OC

46 H

s Siletz River

Yaquina River
Alsea River

44 H. Siuslaw River
Smith River
Umpqua River

Coos Bay

Coquille River

Sixes River
Elk .."
Cape
Blanco

42
|

Figure 2. Map showing the location of major rivers in western Oregon and Washington. Heavier lines mark
reaches of rivers examined in this study.

separate study of fluvial terraces as recorders of Quater Rivers (fig. 2). The first four rivers were chosen because
nary deformation in this region. they all flow through similar bedrock and have several
The rivers chosen for this study are, from north to ages of fluvial terraces preserved along their lengths; the
south, the Siletz, Siuslaw, Smith, Umpqua, and Coquille Coquille River was included in this study because of its
4 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
location in a region of known Quaternary deformation.
With the exception of the Smith River, which flows into
the Umpqua River about 18 km from the coast, all rivers
examined in this study are trunk streams that ultimately
flow into the Pacific Ocean.
GEOMORPHIC SETTING
The Coast Range in Oregon is a rugged mountain
range of moderate elevation that has been deeply dissected
by fluvial incision. Relief is substantial although maxi
mum elevations only reach about 1,250 m; the highest ele
vations in the drainages examined in this study are less
than 800 m. Trunk streams in the Coast Range are usually
entrenched in deep valleys, which indicates probable
stream antecedence. Valley shapes and variations in river
profiles (Rhea, 1993) and numerous examples of stream
capture in the Coast Range (Niem, 1976; Baldwin, 1981)
are probably indicative of long-term Quaternary uplift.
Steep slopes and abundant rainfall aid in the formation of
accumulations of colluvium on most slopes; periodic land
sliding of thicker colluvial deposits from bedrock hollows
is one of the dominant processes by which sediment is
delivered to stream channels in the Coast Range (Pierson,
1977; Dietrich and Dunne, 1978; Benda and Dunne, 1987;
McDowell, 1987; Benda, 1990). Other Quaternary features
and deposits in the region include marine terraces and
sand dunes along the coast, estuarine and salt marsh
deposits in the tidal reaches of coastal rivers, scattered
landslides, and fluvial terrace deposits along Coast Range
StreamS.
CLIMATE
The geomorphology of the Coast Range is strongly
influenced by the temperate, humid climate of the region.
Rainfall is abundant and strongly influenced by elevation.
Annual amounts are about 150 cm near the mouth of the
Coquille River and 180200 cm near the mouths of the
Umpqua, Siuslaw, and Siletz Rivers; annual amounts are
250300 cm in the Coast Range headwaters of the
Coquille, Umpqua, Smith, and Siuslaw Rivers, and
450500 cm in the headwaters of the Siletz River (Oregon
State Water Resources Board, 1958, 1965; Beaulieu and
Hughes, 1975; Schlicker and others, 1973; Schlicker and
Deacon, 1974). Most precipitation occurs as rain during
the winter months (November to March). Snow occurs
above elevations of about 600 m, but only persists on the
ground for more than a day or two above elevations of
about 1,000 m. Summers are relatively dry and cool; tem
peratures are generally more moderate near the coast and
become warmer inland away from maritime influence and
summer fogs. Near the coast, mean summer temperatures
are about 16C and mean winter temperatures are about
7C; temperatures inland in the Coast Range are also mild,
but show wider variations. Despite the rugged topogra
phy, abundant rainfall and mild temperatures combine to
make the Coast Range one of the most productive and
heavily logged forest regions in the country.
Very little is known about Quaternary paleoclimates
of the region, but the Coast Range was probably not glaci
ated during late Pleistocene glaciations that affected the
higher parts of the Olympic Mountains to the north, the
Klamath Mountains to the south, and the Cascade Moun
tains to the east (Porter and others, 1983). Pollen studies
in southern Washington and northern California indicate
cold, dry conditions during the late Wisconsin glacial
maximum, followed by cool, more humid conditions in
the latest Pleistocene; warmer, drier conditions than today
are thought to have been prevalent during the early and
middle Holocene, which were followed by cooler, wetter
climatic conditions in the late Holocene (Heusser, 1983,
1985; Barnosky, 1985; Adam, 1988). These patterns are
consistent with other late Quaternary paleoclimate studies
in the Western United States (for example, Baker, 1983;
COHMAP, 1988).
BEDROCK GEOLOGY
Most of the central Coast Range is underlain by a
sequence of Eocene and younger sedimentary and volca
nic rocks (Wells and Peck, 1961; Walker and Duncan,
1989) deposited in a forearc basin during an earlier phase
of subduction (Snavely, 1987). Most of the study area is
underlain by the Eocene Tyee Formation, a thick
sequence of arkosic and lithic turbidite sandstone and silt
stone. Minor felsic volcanic and intrusive rocks are found
throughout the Coast Range; many of these igneous out
crops form resistant cores that underlie the highest peaks
in the range. The predominant structures in bedrock are
north- and northeast-trending folds and northeast- and
northwest-trending high-angle faults (Wells and Peck,
1961; Peterson and others, 1986; Snavely, 1987; Walker
and Duncan, 1989).
METHODS
TERRACE DATING METHODS
RADIOCARBON DATING
Because of the heavy forest cover and frequent forest
fires, charcoal is relatively abundant in modern stream and
younger terrace sediments in the Coast Range, so I relied
heavily on conventional gas proportional and accelerator
mass spectrometry (AMS) radiocarbon ages in this study
 METHODS
(table 1). Several problems are commonly associated with
radiocarbon dating of organic matter from fluvial terrace
sediments. I was primarily interested in determining the
age of the initiation of aggradation of terrace deposits, so I
tried to obtain radiocarbon samples from the lowest parts of
the terrace sediments. However, very few samples were
obtained from the lower channel-facies gravel, because
charcoal was either rarely deposited in these sediments or
has been oxidized by subsequent pedogenesis and ground
water flow. Therefore, most of my samples were obtained
from the lower part of the fine-grained overbank facies that
forms the upper part of most terrace exposures.
The detrital origin of fluvial charcoal samples can
cause additional problems, because detrital charcoal will
sometimes have a substantial inherited age before it is
deposited with fluvial sediment. Charcoal enters the flu
vial system either as recently burned logs or twigs washed
off the land surface, or more rarely may be recycled by
erosion of previously deposited charcoal from older ter
race sediments. One study in Australia showed that small
(<2 mm) fragments of modern fluvial detrital charcoal
sampled from the active stream bed had inherited ages of
1,400-1,600 radiocarbon years (Blong and Gillespie,
1978). I did not obtain multiple radiocarbon ages on char
coal from modern stream sediments as a test of this poten
tial problem, but several of my radiocarbon analyses on
charcoal from low, young terraces yielded modern ages or
ages of less than a few hundred years, so inherited age is
probably not a serious problem in most of my samples.
Perhaps the easily eroded sandy sediment characteristic of
the stream studied by Blong and Gillespie (1978) is more
susceptible to charcoal reworking than the silty sediment
that makes up the bulk of fluvial terraces in the Coast
Range. In addition, most fragments of charcoal in Coast
Range fluvial sediments are very small (<5 mm) and the
climate is very conducive to rapid breakup of organic
materials, so charcoal fragments of this size are likely to
disintegrate quickly upon remobilization.
Some of the radiocarbon ages reported here are anom
alously young. The samples may have been contaminated
with small pieces of in situ burned roots, or partially
decomposed roots and other young organic material from
animal burrows, resulting in anomalous ages. Modern and
decomposed roots were abundant in almost all terrace
exposures examined in this study. Although extreme care
was taken to avoid including such materials in the submit
ted samples, most of my charcoal samples were collections
of small fragments of disseminated charcoal, so some con
tamination by younger materials was expected. Another
potential source of anomalously young ages is erroneous
sampling of charcoal from younger, non-fluvial sediment.
Because the silty overbank sediments in the Coast Range
are often devoid of characteristic fluvial bedding features,
they can easily be confused with younger colluvial
sediment that has similar appearance and texture. Extreme
5
care was taken to avoid sampling exposures where fluvial
features such as well-sorted sand beds were not clearly
apparent or where angular colluvial clasts were present, but
some samples from overlying colluvial units may have
been included in the radiocarbon analyses.
A final problem with radiocarbon dating is the non
linear relationship between the radiocarbon and calendric
time scales caused by variable rates of C production in
the atmosphere (Stuiver and Quay, 1980). To overcome
this problem, I used conversion data (Stuiver and Reimer,
1986; Bard and others, 1990; Becker and others, 1991;
Stuiver and others, 1991) to calibrate some of my radio
carbon ages. All calibrated ages mentioned in this report
are listed in calibrated years B.P. to distinguish them
from uncalibrated radiocarbon ages or age ranges, which
are listed in yr B.P. (years before present1950) or
ka (thousands of years ago). Modern (post-1950) sam
ples contain some radiocarbon produced by atomic-bomb
testing in the atmosphere in the 1950's, and therefore are
listed as negative ages (10 + 70 yr B.P.) or as ages with
>100 percent of modern radiocarbon (104% modern).
THERMOLUMINESCENCE DATING
I used experimental thermoluminescence (TL) dating
to estimate the ages of several samples of fluvial sedi
ment. The dates reported herein (table 2) are some of the
first TL dates on fluvial sediment in the Pacific North
west. TL dating is a relatively new technique that shows
great promise for directly dating many types of Quater
nary sediment (Wintle and Huntley, 1982; Berger, 1988;
Forman and others, 1988). The principles behind this
method are that during deposition in certain types of envi
ronments, sediments have their inherited TL signal
reduced (bleached) to some low residual level by exposure
to sunlight. The sediment then slowly acquires a new geo
logical TL signal through exposure to ionizing radiation
from radioactive decay of uranium, thorium, and potas
sium in the surrounding sediment. The accumulated TL
signal in the sample is a measure of cumulative exposure
to ionizing radiation. In the laboratory, the TL signal in a
sample is calibrated to a known radiation level called the
equivalent dose. A TL age is determined by dividing the
equivalent dose by the dose rate, which is a measure of
the environmental radioactivity of the surrounding sedi
ment. Common complications to the technique include the
incomplete bleaching of sediment prior to deposition
(which can give the sample an inherited age), fluctuations
in water content of the sediment, and physical and chemi
cal changes in the sediment due to pedogenesis (which
may tend to give anomalously young ages). All these
problems have probably had an effect on the TL ages of
fluvial sediment in the Coast Range.
 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

Table 1. List of radiocarbon ages obtained in this study.

[Radiocarbon laboratory abbreviations: AAUniversity of Arizona/NSF Accelerator Facility; BETABeta Analytic, Inc., Coral Gables, Florida;
ETH-Eidgenossische Technische Hochschule, Zurich (through Beta Analytic); GXGeochron Laboratories, Cambridge, Massachusetts; PITTUniversity of
Pittsburgh Radiocarbon Laboratory; USGS-U.S. Geological Survey, Menlo Park Radiocarbon Laboratory]

Field and Field station Latitude Longitude Geologic material Reported Remarks
laboratory number and sample weight Radiocarbon age
sample number (g) (yr B.P.)

UMPQUA RIVER

SP88UR1 88-U2 43 31.45' 123 32.54' charcoal 9,120 + 150 Disseminated sample.
BETA-27962 5.0

SP87UR2 87-U4 43 31.44" 123 32.21." charcoal 9,800 + 410 Single fragment exposed in stream cut
BETA-22662 3.0

SP87UR3 87-U6 43 28.98 123 30.88. charcoal 6,545 + 75 Single fragment.

PITT-0126 3.9

SP88UR3 88-U3 43 29.05' 123 30.84' charcoal 6,440 + 60 Disseminated sample

USGS 2722 9.1

SP87UR4 87-U7 43 38,00' 123 34.11" charcoal 8,630 + 100 AMS (accelerator mass spectrometry) age on
AA-2752 0.24 disseminated sample, 400 cm depth.

SP87UR 10 -- ** -- charcoal 730 + 80 Disseminated sample, same location as SP87UR4

USGS 2643 1.0 at 50 cm depth

SP87UR5 87-U8 43 37.90 123 34.02. charcoal 210 + 80 Single fragment; 250 cm depth.
BETA-22663 4.1

SP87UR6 -- -- - wood 95 + 40 Single fragment, 180 cm depth.

USGS 2646 9.5

SP87UR7 -- -- - wood 11.5 + 40 Single fragment; 120 cm depth.

USGS 2645 18.0

SP87UR8 -- -- - charcoal - 10 + 70 Disseminated sample, 50 cm depth.

USGS 2644 1.0

SP87UR 15 87-U14 43 38.86" 123 52.50 charcoal 10.4% modern Disseminated sample, probably deposited onto the
PITT-0127 2.0 terrace in a debris flow from a tributary stream.

SP87UR18 87-U20 43 40.03' 124 05.30' charcoal >26,400 Disseminated sample, 425 cm depth.
BETA-22664 1.8

SP87UR 19 -- -- - charcoal 22,240 + 630 Disseminated sample, 325 cm depth.

PITT-0186 5.8

SP87UR21 87-U22 43 41.67" 124 00.25 peat 1,180 + 130 Bulk sample from auger core; 225-250 cm depth.
BETA-23533 7.4

SMith River

SP87USR2 87-US3 43 44.94' 124 01.06' charcoal 2,685 + 95 AMS age; disseminated sample.
BETA-23531 0.3
ETH-3546

SP87USR3 87-US5 43 48.14' 123 48.44" charcoal >43,600 AMS age; disseminated sample.
AA-2753 0.23

SP87 USR5 87-US7 43 47.45 123 46.39' charcoal 7,950 + 110 AMS age; disseminated sample from terrace on Vincent
BETA-23532 0.55 Creek, 0.25 km upstream from confluence with
ETH-3547 Smith River.

SP87USR6 87-US 10 43 47.20 123 49.50' charcoal 2,860 + 215 Minimum age; disseminated sample from colluvium
PITT-0128 1.4 overlying stripped terrace.

SP87USR7 87-US 11 43 46.95' 123 47.97' charcoal 10,165 + 85 AMS age; disseminated sample from alluvial-fan
AA-2754 0.28 sediment.

SP87USR8 87-US 12 43 47.15' 123 48.65 charcoal 1,020 + 50 Minimum age on single fragment; probably a burned
USGS 26.42 5.6 root.

SP87USR 12 87-US 15 43 44.29' 124 04.31' charcoal, peat, wood 123.7% modern Mixed sample from auger core; 125 cm depth.
PITT-0129 4.4

SP87USR13 -- -- -- charcoal, wood 1,825 + 65 Several fragments from same core as SP87USR12;
PITT-0130 3-5 360 cm depth.
 METHODS 7

Table 1. List of radiocarbon ages obtained in this studyContinued.

Field and Field station Latitude Longitude Geologic material Reported Remarks
laboratory number and sample weight Radiocarbon age
sample number (g) (yr B.P.)

Siuslaw river

sp88SR1 88-S1 44 03.72" 123 52.98. charcoal 7,010 + 90 AMS age; disseminated sample.
betA-27967 0.2
eth-4489

Siletz river

SP88SZR2 88-SZ2 44 42.89' 123 55.92. charcoal 12,110 + 95 Minimum AMS age; disseminated sample probably
BETA-28.357 2.0 contaminated with modern carbon.
ETH-4746

SP88SzR4 88-SZ6 44 44.63" 123 47.76 charcoal 9,030 + 110 AMS age; disseminated sample,
betA-27966 0.3
ETH-4488

SP88SZR5 88-SZ7 44 47.17' 123 47.75' charcoal >36,000 Charcoal split; several fragments.
BETA-27963 5.8

SP88Szr - - wood 41,600 + 1300 Wood split; several fragments from same location as
USGS 2724 19.4 SP88SZR5.

SP88SZR7 88-SZ8 44 46.70' 123 50.07" charcoal 2,230 + 60 Disseminated sample.

USGS 2723 2.6

SP89SZR1 89-SZ2 44 43.79' 123 55.42' charcoal 9,695 + 440 Disseminated sample.
GX-15311 3.0

COQUILLE RIVER

SP89Cr14 89-C3 43 0.9.51' 124 11.88' wood, charcoal 2,860 + 155 Disseminated sample from auger core in northern
GX-15309 6.1 abandoned channel; 350 cm depth.

SP89CR7 89-C4 43 09.38' 124 11.82 wood, charcoal 2,835 + 140 Disseminated sample from auger core in southern
GX-15310 30 abandoned channel; 375-400 cm depth.

A relatively simple TL sampling technique was used
in this study. The terrace exposures were first excavated
about 2050 cm to expose fresh silty overbank sediment
that had not recently been exposed to sunlight. Samples
were taken by screwing or pushing soup-can sized steel
cans into the fresh exposure, and the cans were then
sealed with aluminum foil and plastic tape to preserve
moisture content and prevent exposure to sunlight.
Three TL ages were obtained on samples from radio
carbon-dated exposures on the Umpqua River in order to
establish the accuracy of the technique (fig. 3). TL analy
sis on modern overbank sediment (radiocarbon age on
charcoal of 210 + 80 yr B.P.) yielded a TL age of
7.4 + 2.6 ka (fig. 3A). The sampled sediment is a loosely
consolidated silty sand that contains abundant well
preserved fragments of charcoal and wood. Two other
radiocarbon-dated sites yielded mixed results; the TL age
on sample SP87UT3 (8.9 + 5.9 ka) was only slightly older
than two radiocarbon ages on charcoal (6,440 + 60 yr B.P.
and 6,545 + 75 yr B.P.) from the same location (fig. 3B);
whereas, the TL age on sample SP87UT1 (18.5 + 2.4 ka)
was nearly twice as old as a radiocarbon age on charcoal
of 9,800 + 410 yr B.P. (fig. 3C) from the same location.
The anomalously old Holocene ages indicate that inherited
TL ages of 2.59 kyr, which are probably related to
incomplete bleaching of the sediment during transport,
may be characteristic of fluvial-overbank sediment from
the upper reaches of Coast Range rivers. In contrast to the
apparent incomplete bleaching seen in the Holocene sam
ples, most of my TL analyses on older sediments from the
tidal reaches of the Umpqua and Smith Rivers appeared to
yield younger ages than geologic relations would indicate
(table 2), but the geologic age estimates on these older
sediments are poorly constrained. The TL ages on older
sediments are discussed in more detail in sections on
Umpqua and Smith River terraces.
Some dating studies have found reasonably good con
cordance between TL and other numerical ages of fluvial
sediment (Nanson and Young, 1987; Nanson and others,
1991; Murray and others, 1992) and the effectiveness of
natural zeroing of suspended fluvial sediment (Huntley,
1985; Berger and others, 1990), but other TL studies of
8 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

Table 2. List of thermoluminescence (TL) ages obtained in this study.

[All analyses were performed by Steven L. Forman at the Institute of Arctic and Alpine Research (INSTAAR), University of Colorado, Boulder, Colo.]

Field and Field Station Latitude Longitude Geologic material TL age Geologic age Remarks
laboratory number (ka) estimate

sample number (ka)

UMPQUA RIVER

SP87UT1 87-U4 43 31.44' 123 32.21' silty sand 18.5 + 2.4 9-11 Sample located near SP87UR2 (9,800+
ITL84 410 yr B.P.)

SP87UT3 87-U6 43 28.98' 123 30.88' silty sand 8.9 + 5.9 6-7 Sample located near SP87UR3 (6,545 +
ITL116 75 yr B.P.)

SP87UT5 87-U8 43 37.90' 123 34.02 sandy silt 7.4 + 2.6 (). 1-0.3 Sample located near SP87UR5
ITL85 (210 + 80 yr B.P.)

SP87UT7 87-Ul 43 38.39' 123 36.30 clayey, silty sand 116 + 20 125 Sample located in lower part of Bt soil
ITL118 horizon; may correlate with substage
5e marine terrace.

SP87UT18 87-U18 43 41.82' 124 07:52. 65.2 + 7.3 125 May correlate with substage 5e marine
ITL117 terrace.

SP87UT20 87-U19 43 41.84' 124 0'5.81" clayey, sandy silt >200 >200 At TL saturation: may correlate with
ITL173 stage 7 or older marine terrace.

SP87UT22 87-U21 43 41.99' 124 06.40' sandy silt 14.9 + 3.1 15-30 May correlate with Scholfield Creek
ITL86 fluvial terrace.

SMITH RIVER

SP87UST2 87-US2 43 44.12' 124 02.92' sandy silt 55 + 12 125 May comelate with substage 5e marine
ITL174 terrace.

fluvial sediments have reported less encouraging results
(Berger, 1990). While not without inconsistencies, my TL
results on fluvial sediments are encouraging and, when
used with other data, may be useful in determining rela
tive ages of older fluvial terrace sediments in the Coast
Range. Unfortunately, these analyses are very expensive
and time consuming, so only a limited number of TL anal
yses were performed. I hope that demonstration projects
such as this one help to establish the viability and utility
of TL dating, so that additional studies that address the
problems identified in the present study can be conducted.
OTHER DATING METHODS
Studies of soil development are commonly used in
fluvial terrace studies (for example, Birkeland, 1984;
Harden, 1987; Dethier, 1988), but I found that the steep
climatic gradients (especially rainfall and summer temper
atures) from the coast to inland valleys and poor terrace
preservation reduced the utility of soils as a method for
estimating terrace ages. In addition, most exposures of ter
race sediments that I examined in this study are in
roadcuts that may not be situated in stable geomorphic
positions. For these reasons, I did not describe soils in
detail during my studies of terraces in the Coast Range,
and have not included soil profiles on the stratigraphic
Figure 3 (facing page). Stratigraphic columns of three fluvial
terrace exposures along the Umpqua River where both thermolumi
nescence (TL) and radiocarbon samples were collected. Radiocar
bon ages are listed in years before present (1950; yr B.P.) and TL
ages are listed in thousands of years (ka); field sample numbers (in
parentheses) are listed for reference to table 1. Column thicknesses
are measured from the sediment/bedrock contact. Exposure loca
tions are (A) lower terrace near Elkton at station 87U8, (B) com
posite column from a terrace near river km 130 at stations 87U6
and 88-U3, and (C) terrace near McGee Creek at station 87U4.
Legend shows patterns and labels for all stratigraphic columns in
this report. Columns show complete exposures of terrace sedi
ments, from the bedrock strath to the surface of the terrace, except
where marked with covered contact symbol or where bedrock is not
shown at the base of the column. Soil development is not shown in
the columns. Grain size scale shows breaks between mud (silt and
clay), sand, and gravel. Note differences in vertical scale.
 METHODS

...; . Sand N Bedrock /\/\/ Covered contact

A A Disseminated or bulk
Fluvial |||||| Buried A 14C sample
gravel \! \| A horizon
A Single 14C sample
: Colluvial a 2 .
- *: gravel 23 Pebbly zone
O TL sample

-10+ 70 yr B.P. ..]9,800 +410 yr B.P.

...] (SP87UR8) N] (SP87UR2)
+18.5 +24ka
(SP87UT1)
115 + 40 yr B.P.
(SP87UR7)

95 + 40 yr B.P.
(SP87UR6)

# 210+ 80 yr B.P.
*::::::::..] (SP87UR5)
EA' f:

# 2 -

- *7.4 + 2.6 ka
(SP87UT5)
5 -

;
-

; 3

..: 6,545 + 75 yr B.P.

Exx.jsp&uR3)
::= 8.9 + 5.9 ka
2-%
... ;|(SP87UT3)
1-#### 6,440+60yr B.P. 1-
i (SP88UR3)
P23 S P.32
.

N I

# #

10 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
columns in this report. The intense rainfall induces severe
oxidation and in situ weathering; soils were generally red
der, more oxidized, contained larger amounts of clay, and
contained more weathered gravel clasts (measured by hand
crushing) with advancing age. However, some soil charac
teristics such as distinct horizonation, clay films, and soil
structure were less developed than might be expected in
the older soils, so some of the deeper weathering may be
related to groundwater circulation.
Thicknesses of weathering rinds have been used as a
relative dating technique on fluvial gravel clasts (Colman
and Pierce, 1981; Knuepfer, 1988). Preliminary studies by
Colman and Pierce (1981) indicated that measurement of
weathering rinds on basaltic clasts might have utility as a
dating technique along the Siletz River, but I found the
technique unusable for several reasons. Of the five rivers
examined in this study, only the Siletz River drainage
basin contains enough outcrops of basaltic volcanic rocks
to provide a source of clasts of suitable lithology. Unfortu
nately, gravels are restricted to the lower parts of most
Coast Range terrace sequences, far below the upper parts
of the B horizon where the most consistent weathering
rinds are formed. I examined weathering rinds at several
locations in the Coast Range, but most rinds were widely
variable in thickness and appeared to be related to
intermittent groundwater flow. Rainfall amounts also
affect rind thickness (Colman and Pierce, 1981), so the
steep rainfall gradients between the coast and higher parts
of the Coast Range would also complicate comparisons
between rind thickness and terrace age.
A final dating technique commonly used in studies of
older terraces is the correlation of fluvial terraces with
marine terraces and the prehistoric sea-level record. The
best preserved sequence of marine terraces in the Pacific
Northwest is present between Cape Blanco and Coos Bay
(fig. 2) in southern Oregon. These terraces have been the
subject of numerous studies but only recently have gener
ally consistent uranium series, amino acid, and stable iso
tope age data become available. The youngestWhiskey
Run terracehas been correlated with the 80 ka, oxygen
isotope substage 5a (Shackleton and Opdyke, 1973) sea
level high stand; the next olderPioneer terracehas
been correlated with the 105 ka, substage 5c sea-level high
stand (Kelsey, 1990; Muhs and others, 1990); and the next
oldestSilver Butte terracehas been tentatively corre
lated with the 125 ka, substage 5e sea-level high stand
(Kelsey, 1990). Detailed mapping by Kelsey (1990) indi
cates that all three of these terraces occupy the same
wave-cut platform in the Cape Blanco region, so the lack
of significant differences in altitude between the three
stage-5 terraces precludes the use of elevation spacing as a
precise method of correlating scattered marine terrace
remnants in central Oregon to dated sites in southern Ore
gon. Detailed soils studies at Cape Blanco and Cape
Arago (Bockheim and others, 1992), however, indicate
that soil development can be used as a marine terrace cor
relation tool. My reconnaissance examination of soils on
marine terrace sediments along the central Oregon coast
has indicated that the lowest marine terraces present are
more likely to have formed during one of the stage-5 high
stands, (80 ka, 105 ka, or 125 ka) than during stage 7 (230
ka) or older high stands. More detailed soils studies (H.M.
Kelsey, written commun., 1992) indicate that in most
places along the central Oregon coast, the lowest marine
terrace is correlative with the oxygen-isotope substage 5e
(125 ka) sea-level high stand, and all higher terraces prob
ably correlate with stage 7 (230 ka) or older high stands.
Some tentative correlations of marine terraces with fluvial
terraces near the mouths of the Umpqua, Smith, Siuslaw,
and Siletz Rivers will be discussed.
TERRACE HEIGHTS AND
PROFILE CONSTRUCTION
Fluvial terrace remnants were initially identified on
1:12,000-scale air photos, and most were later checked in
the field. Several methods were used to measure terrace
heights. Some were measured with an altimeter, espe
cially those located within a few minutes walk or drive
from a benchmark where the instrument could be accu
rately calibrated. Other terrace remnants were surveyed
with a hand level and rod from points of known elevation;
the lowest terraces were commonly measured in this fash
ion from the height of the river surface. Some terrace
heights also were measured directly from 1:24,000-scale
topographic maps. However, many high remnants were
inaccessible due to thick vegetation or land ownership
problems, so I employed a photogrammetric method to
measure additional terrace heights. A computer-assisted
photogrammetric mapping system (Dueholm and Pilmore,
1989) was used to construct detailed topographic profiles,
usually measured from the river surface to the terrace sur
face through points of known elevation. These various
methods yielded probable errors of about +5 m for the
higher terraces and +2 m or less for the lowest terraces.
In all cases, height measurements were originally made in
feet and converted to meters above mean sea level.
Longitudinal profiles of the stream and terrace rem
nants were constructed for the Umpqua, Smith, Siuslaw,
and Siletz Rivers. These profiles were constructed with the
terrace height data and stream elevations taken from
1:24,000-scale topographic maps. The elevation data were
plotted against stream length (in km), which was con
verted from river mileage values taken directly from the
topographic maps. Some discrepancies occurred in plotting
stream length of terraces preserved in abandoned meander
bends on the Umpqua and Smith Rivers. In these cases,
terrace remnant lengths were projected downstream from
the location of the upstream end of the remnant. The
profiles clearly show the poor preservation of terrace
 GENERAL CHARACTERISTICS OF COAST RANGE TERRACES
surfaces and the tenuous nature of most terrace correla
tions. Poorly constrained correlations (dotted lines) mostly
were made by mimicking the gradients of the better con
strained correlations (solid and dashed lines).
GENERAL CHARACTERISTICS OF
COAST RANGE TERRACES
TERRACE STRATIGRAPHY
The stratigraphy of fluvial terrace deposits found
along Coast Range rivers is remarkably consistent, regard
less of terrace height above the present river level. Along
the upper reaches above tidal influence, all terraces are
strath terraces consisting of a tread cut into bedrock over
lain by 3 to 10 m of unconsolidated sediment. These sedi
ments consist of a 1- to 3-m-thick lower channel-gravel
facies of sandy pebble and cobble gravel, overlain by a 2
to 7-m-thick overbank facies of sand, sandy silt or silty
sand, and silt; thin, discontinuous beds of sand and sandy
gravel are commonly present in the overbank facies. I
found very few examples of buried soils in the more than
60 exposures of terrace sediments examined in this study.
These buried soils were mostly thin (<2030 cm), poorly
developed A horizons that probably represent hiatuses in
deposition of less than a few tens of years, so in most
cases the sediments in both the channel-gravel and over
bank facies must have been deposited fairly continuously.
The stratigraphy of fluvial terrace sediments near the
coast is more variable, and probably is a reflection of
short- and long-term fluctuations in sea level. The lower
reaches of all rivers examined in this study are tide influ
enced and are flowing in deep, sediment filled, fiord-like
valleys. These valleys were excavated by stream incision
during periods of low sea level associated with full glacial
conditions and now act as traps for fine-grained estuarine
sediment. Most fluvial terrace sequences in these reaches
generally consist of fine sand, mud (silt and clay), and
peat, may be very thick (water wells in the Reedsport area
on the Umpqua River have penetrated more than 43 m of
unconsolidated fluvial sediment (Curtiss and others, 1984),
and usually do not have exposed bedrock straths. Similar
fluvial sequences extend up river to the tide limit of most
Coast Range rivers. A few older, higher terrace remnants
near the coast, however, have thinner stratigraphic
sequences and bedrock straths like their upstream counter
parts, supporting the conclusion that hydrologic conditions
near the mouths of Coast Range rivers have not remained
static through time.
CONTROLS ON TERRACE FORMATION
The modern hydrologic setting of the nontidal
reaches of rivers in western Oregon is one of active stream
11
incision driven by regional uplift of the Coast Range.
Active incision is indicated by the presence of bedrock in
the beds of most Coast Range streams above tide limit and
by the presence of very narrow modern flood plains.
Long-term Quaternary incision is implied by the presence
of high strath terraces and entrenched meanders, confine
ment of the upstream reaches of all rivers in relatively nar
row, bedrock cored, V-shaped valleys, and by numerous
examples of stream capture in the Coast Range (Niem,
1976; Baldwin, 1981). Stream incision has probably domi
nated throughout most of the late Quaternary, and it is a
condition not conducive to extensive deposition of fluvial
sediment. So how do extensive fluvial terraces form in
such an environment? Initial deposition of fluvial sediment
requires hydrologic conditions that allow streams to
migrate laterally. Periods of lateral migration can be
related to short hiatuses in regional uplift or to periods of
aggradation caused by an influx of sediment or a decrease
in stream competence related to a rise in base (sea) level.
Experimental data show that lateral migration will domi
nate stream incision if the sediment load of a stream is
increased beyond the ability of the stream to entrain its
load (see reviews in Schumm and others, 1987), so large
sediment inputs not only provide the material for aggrada
tion and are the source of terrace sediment, but large
influxes of sediment also induce the lateral stream migra
tion that cuts the bedrock straths that underlie the terrace
sediments. Finally, terrace formation requires that the ter
race tread eventually be abandoned by renewed incision
caused by decreased sediment supply or a drop in base
level caused by renewed uplift or falling sea level.
I have roughly grouped fluvial terraces in the Coast
Range into four main types. The smallest, most areally
restricted terraces are formed by vertical incision concur
rent with minor lateral incision that accompanies meander
ing of the river channel. This stream behavior results in
unpaired slip-off terraces (Cotton, 1940) preserved on
the insides of meander bends. Some of these isolated rem
nants may correlate with more extensive terraces, but most
are probably not synchronous with other terrace surfaces.
Isolated slip-off terraces are present at many places
along most Coast Range rivers.
Other aerially restricted terraces form as a result of
short-lived or localized aggradation events. These are
complex response terraces (see reviews in Schumm and
others, 1987, and Bull, 1990) that form in response to
minor stream perturbations; these terraces also do not
form extensive synchronous terrace surfaces. At scattered
locations in the Coast Range, such terrace remnants are
preserved at various levels and have gradients that often
are steeper than the better preserved, more extensive ter
races. Most of these small terraces probably were formed
as a result of transient changes in stream regimesuch as
aggradation related to large floods or localized land
slidingand flood plain abandonment by meander cutoff.
12 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
Another type of terrace does form major, roughly
synchronous terrace surfaces in the Coast Range. These
are terraces produced by regional stream aggradation
events related to major changes in climate (climatic ter
races of Bull, 1990). The best example is an extensive ter
race found on all Coast Range rivers that I examined that
yields consistent radiocarbon ages of 710 ka from the
lower parts of the silty overbank facies. These terraces
formed during a period of regional aggradation probably
related to increased colluviation of hillslope sediment into
small tributaries induced by warmer, drier post-glacial cli
matic conditions, and perhaps influenced to some degree
along lower reaches by decreasing stream gradients
induced by rapidly rising base level (see more extensive
discussion under Terraces along the Umpqua River and
Discussion sections). If such extensive regional aggrada
tion is the typical response of Coast Range rivers to post
glacial climate change, then earlier climatic fluctuations
probably induced the formation of similar regionally
extensive fluvial terraces. TL ages and relations between
scattered fluvial and marine terrace remnants near the
coast indicate that some of the older, higher terraces were
formed under similar climatic and hydrologic conditions.
The last type of fluvial terrace present in the Coast
Range forms by episodic stream incision caused by local
ized tectonic movements; these tectonic terraces differ
from others that are related to regional-uplift-induced
stream incision because they are associated with active
structures. Examples of tectonic terraces were examined
on the Siuslaw River as part of this study; topographic
profiles clearly show that terraces on the lower reach of
the river have been warped over an anticline in the under
lying Eocene bedrock. Although the mode of growth of
this anticline (episodic versus continuous) is unknown, the
presence of multiple terrace surfaces over the anticline
supports the inference that localized tectonic uplift has
influenced terrace formation.
In summary, fluvial terraces in the Coast Range
appear to have a number of origins. At least four different
types of terrace are present, but only those terraces
induced by major changes in climate are likely to be
extensive and synchronous along the lengths of the exam
ined rivers. Other terraces formed by complex response
mechanisms, river meandering, and movement on local
tectonic structures in most cases did not form extensive
synchronous surfaces; most isolated terrace remnants
found along Coast Range rivers were probably formed by
these processes. More details on terrace formation are in
the following discussions.
SEDIMENT ACCUMULATION RATES
Sediment accumulation rates are useful for delineat
ing the time spans over which different fluvial deposits
accumulate and for helping to estimate the ages of terrace
surfaces. For this study I was primarily interested in
determining the age of initiation of terrace aggradation, so
I sampled for radiocarbon from as low in the terrace sedi
ments as possible. Unfortunately, only one radiocarbon
sample was found in the basal gravel facies, so sediment
accumulation rates for these deposits could not be deter
mined. However, I found only one example of a buried
soil in the gravel facies, so the gravels were probably
deposited rapidly and remained entrained in active stream
channels until they were abandoned by lateral stream inci
sion and buried by fine-grained overbank deposits.
Only a few exposures yielded multiple radiocarbon
ages that defined closed time intervals that could be used
to determine finite sediment accumulation rates, and all
the resulting ages were located in the overbank deposits.
Radiocarbon ages were used to calculate long-term sedi
ment accumulation rates from the lowest continuous ter
race on the Umpqua River near Elkton at river km 80 (fig.
4A; station 87U7, plate 1) and from a terrace on Schol
field Creek, a major tributary of the Umpqua about 11.5
km upstream from Reedsport (fig. 4B, station 87U20,
plate 1). Radiocarbon ages of 730 + 80 yr B.P. and 8,630
+ 100 yr B.P. on charcoal collected from 0.5 m and 4.0 m,
respectively, below the surface of the terrace near Elkton
yielded a sediment accumulation rate of about 0.45
mm/yr. A maximum rate of about 0.25-0.3 mm/yr from
the terrace on Scholfield Creek was obtained from ages on
charcoal of 22,240 + 630 yr B.P. and >26,400 yr B.P.
from depths of 3.25 m and 4.25 m, respectively.
Additional sites yielded shorter term sediment accu
mulation rates from low terraces near the coast that are
substantially higher than the rates discussed above. Near
the mouth of the Smith River, data from an auger core in
a low terrace at about river km 4.5 (fig. 4C, station
87US 15, plate 2) yielded a maximum accumulation rate
of about 1.3 mm/yr from radiocarbon ages of 123.7%
modern (post 1950) and 1,825 + 65 yr B.P. at depths of
1.25 and 3.6 m, respectively. An auger core in a similar
low terrace on the Umpqua River at about river km 26
(fig. 4D; station 87U22, plate 1) yielded a minimum
accumulation rate of about 1.9 mm/yr from a single radio
carbon age of 1,180 + 130 yr B.P. on a peat sample from
a depth of 2.25 m. Feiereisen (1981) obtained a similar
rate from a shallow drill hole at about river km. 23.5 on
the Siuslaw River flood plain; a single radiocarbon age of
3,250 + 70 yr B.P. on a sample of peaty wood from a
depth of 5.5 m yielded a minimum sediment accumulation
rate of about 1.7 mm/yr.
These data indicate that accumulation rates calculated
over the life of a terrace on the upper reaches of Coast
Range rivers are on the order of 0.30.5 mm/yr; rates are
about 4 times higher (1.31.9 mm/yr) on lower (tidal)
reaches. No other suitable sites were found for determina
tion of sediment accumulation rates, but these rates may
 C

B.P.
130y
7.180+ Saehdriaocmtakeirngbotrenas-tpdaihetosncd terrace
4.
rFigure
scolumns
calculate
to
used
sediments
tfluvial
column
rates;
section
exposed
of
base
the
from
measured
are (Prleaisntosceint-Hiolocne) of
PHT
(A)
are
locations
core
Exposure
core.
tor
Creek
Scholfield
(B)
87U7,
station
at
Elkton
near
River
Umpqua
on lest
a
from
(D)
and
2),
(pl.
87US15
station
at
Smith
the
lowest
in
core
auger
3.75-m
(C)
w87U20,
m
300
about
River
Umpqua
on
terrace
Creek
Dean
ow

(SP87UR21) ja
a
is
#
3
|
*is
3
3
|
#
|@
@|||
3
|
#

3
3
-

#H
*HS:
|H-

B.P.
yr
>26,400
65
+
"|1,825
B.P.
yr
#AA
u AA
(SP87UR4)
modern
123.7% C/D
oaln"
o-
lspushio
0

=
2(L
ESP.T.
87USR12)
CC

2#
#
TT2
-Fi
-

.
LLI
.
>
AA
(SP87UR18
A.A.'
:

B.P.
yr
630
Al22,240+

(SP87UR19)
*

dif erences
Note
87U22.
station
at
scale.
vertical
in

fss20+yr
2
BP
v
100 B.F.
"
K.T.
-7
H-7
Oc -

B.P.
yr
80
j730+

*|(SP87UR10)
14 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
be characteristic of sedimentation rates on most similar
sized rivers in the region. The variations in the accumula
tion rate data probably do not allow reasonable estimation
of the age of terrace surfaces, but they do suggest that
most of the thick terrace overbank sequences accumulated
over time periods of thousands of years.
TERRACES ALONG THE
UMPQUA RIVER
The Umpqua River is the largest river examined in
this study (11,800 km drainage area; Oregon State Water
Resources Board, 1958), and it is the only river that I
examined whose headwaters are in the Cascade Range of
west-central Oregon. Although fluvial terraces are rela
tively poorly preserved along the Umpqua River (fig. 5;
pl. 1), I have used age and sedimentologic data from
Holocene and older terrace remnants along the lower 200
km of the river to determine some of the processes respon
sible for terrace formation in the Coast Range.
The only terrace that can be correlated with any cer
tainty along the Umpqua River is a nearly continuous ter
race surface that lies 215 m above river level.
Radiocarbon dating of charcoal from this terrace indicates
that sediments near the base of the terrace are latest Pleis
tocene to early Holocene in age, so I informally refer to
this terrace as the Pleistocene-Holocene transition or
PHT terrace. If the geomorphology of this terrace is typ
ical of older fluvial terraces formed along the Umpqua
River (and the similar stratigraphy of the terrace sediments
suggests that it is), then the poor preservation of older ter
races can be easily explained by the morphology of the
PHT terrace. The PHT terrace is commonly unpaired,
remnants are usually narrow (<600 m), and the surface is
cut in places by younger terraces.
With few exceptions, the stratigraphy of older terrace
deposits along the Umpqua River is almost identical to
that associated with nearby Holocene terraces. Soil devel
opment, however, is progressively greater in the older
deposits. Soil properties most consistent with terrace
height and therefore age were degree of reddening,
amount of clay and oxidation products in the B horizon,
and degree of clast weathering where gravel is present in
the B horizon. Soil oxidation also appeared to increase
away from the coast, presumably as a response to hotter,
drier summer weather and fluctuating soil-moisture condi
tions. Prominent terrace exposures are discussed below.
LATEST PLEISTOCENE AND
HOLOCENE TERRACES
I examined the PHT terrace in detail at three sites on
the upper Umpqua River, at stations 87U4 and 88U2
near McGee Creek, and at station 87U7 near Elkton (pl.
1); these sites yielded charcoal for radiocarbon dating
from nearly complete exposures of the terrace sediments.
A fourth site downstream from the highway bridge near
Scottsburg on the lower Umpqua (station 87U14) yielded
a modern age that probably is related to deposition from a
tributary stream.
The relative scarcity of continuous terraces below the
PHT terrace (fig. 5) indicates that the latest period of
extensive terrace formation on the upper Umpqua River
occurred in latest Pleistocene and early Holocene time,
and only a few scattered terraces have developed since
then. I examined lower, younger terraces at three sites on
the upper Umpqua River, at stations 87U6 and 88U3
near the Bullock Bridge and at station 87U8 near Elkton.
I also examined low terraces on the lower Umpqua River
near Reedsport at stations 87U13 and 87U22.
PLEISTOCENE-HOLOCENE TRANSITION
(PHT) TERRACE EXPOSURES NEAR MCGEE CREEK
The PHT terrace near McGee Creek has the longest
continuous exposure of fluvial terrace sediments along the
length of the Umpqua River examined in this study (fig.
6). Part of the bedrock strath has been exhumed and the
sediments have been completely removed on part of the
terrace; the excellent exposures at this location probably
resulted from scour by recent flooding. A topographic sur
vey across the stripped bedrock strath with rod and Abney
level documents the height of the strath beneath the ter
race surface and shows the morphology of the site (fig. 7).
Shallow abandoned channels and other minor surface
irregularities are not shown on the profile, but a 0.51.0
m-high scarp preserved along one of several strike-slip
faults exposed in the bedrock strath is included. This and
other similar scarps are not related to recent surface fault
ing but are the result of differential erosion of softer silt
stone in fault contact with more resistant massive
sandstone. Other examples of exhumed bedrock surfaces
are common along this reach of the Umpqua River.
The stratigraphy in the McGee Creek terrace is typi
cal of that seen along most of the PHT terraces, and the
radiocarbon results illustrate the types of problems associ
ated with dating fluvial sediments (figs. 3C, 7, 8). The two
exposures at stations 87U4 and 88U2 are about 500 m
apart, consist of basal sandy gravel channel deposits over
lain by silty overbank deposits, and yield inverted radio
carbon ages of 9,800 + 410 yr B.P. (station 87U4) and
9,120 + 150 yr B.P. (station 88U2). The charcoal sam
pled at station 87U4 (fig. 3C) was a single fragment
found protruding from the silty overbank deposits. This
sample may have been exposed to growth of fungi or
molds; but the fragment was probably only recently
 C.

Longitudinal
5.
Figure
that
connect
lines
lighter
and
River
elevation
is
line
solid
Heavier
River.
Umpqua
reach
lower
the
along
terraces
of
profile loumcianerscbenocen) estimated
ap roximating
bars
terraces,
marine
heights
The
asterisks.
with
marked
are
error
seriously
be
to
thought
ages
ago);
years
of
(thousands
ka
in
listed
and
circles
vertical
oprerceulataitoievndes;,. ages
Terrace
physically
be
can
probably
that
terraces
connect
sdashed
are
lines
dotted
cand
m
+5
of
errors
vertical
have
may
heights ftahedrmioor
m
+2
terraces
higher
Rfor
terraces.
lower
(the
TL
B.P.,
years
in
listed
and
triangles
solid
with
located
are
250 200 150
-ZA
O-T-
I
-T-S
^

terrace
PHT
Teerrace
m|asurement (ar owheads)
bends
meander
abandoned
Locations
figure.
of
side
left
on
MT"
labelled
errors,
axis.
distance
along
shown
are
(TR")
range
tide
the
and
meander
Abandoned
Z

Radiocarbon
A
age
terrace
Marine
MTL
range
Tide
TL

age

|-

|r,
+ "...]
6,545
7T
B.P.
yr
.5 .....!
...
B.P.
yr
6,440+60
....
--

A TR r.s.
r, MT
5.9|..
+
8.9
ka 50

|--

...
...
410......
+
9,800
B.P.
yr |....."
150
+
9,120
B.P.
yr

(km)
COAST
FROM
DISTANCE

1
...
r
Mi
ggo,
"99-24
tooyee

20ka
...iter
ka
>200
'"
...
...."
MTL. ...r.
*" .... \
...'
100
H'
|
---|.......

|--
--"Pr"
...

RIVER
UMPOUA TE
40
100
200
2O
80
60
180
160
140
120
0
....... .1
P.

7'"
+
*65.2
"
....--"
ka
..3
..
-

'"?!
31
14.9,
16 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

-
-
- *--->

Figure 6. Photograph of PHT (Pleistocene-Holocene transition) terrace at about river km 121 on the Umpqua River near McGee
Creek; station 87U4 is located at right edge of photograph. Note extensive exhumed bedrock strath to the left of the terrace and in
the foreground. Profile shown in figure 7 was measured across the strath near the center of the photograph.

exposed by flooding, and sample pretreatments should
have removed such contaminants. At station 88U2 (fig.
8), numerous small pieces of charcoal were sampled from
a silty interbed in the basal gravel; this sample was not
exposed prior to excavation, but the enclosing sediments
contained abundant modern roots and rootlets, so the sam
ple was carefully picked prior to dating. Although slight
contamination of one or both samples is possible, the ages
nearly overlap within their laboratory error limits. Thus,
the radiocarbon results indicate that the basal McGee
Creek terrace sediments were deposited a minimum of
9,00010,000 radiocarbon years ago. Radiocarbon ages in
this range have been corrected to 9,90011,000 calibrated
years B.P. on the dendrochronological time scale (Becker
and others, 1991; Stuiver and others, 1991).
PHT TERRACE NEAR ELKTON
I examined a similar exposure of the PHT terrace
near Elkton at river km 77.5 (station 87U7, plate 1).
Small fragments of charcoal from the lower part of the
overbank facies yielded an accelerator (AMS) radiocarbon
age of 8,630 + 100 yr B.P., and fragments from the upper
part of the overbank facies yielded a conventional radio
carbon age of 730 + 80 yr B.P. (fig. 4A). The consistent
surface height and early Holocene age of the sample near
the middle of the overbank facies indicate that the PHT
terraces at McGee Creek and Elkton can be correlated
with confidence.
PHT TERRACES NEAR THE COAST
A somewhat unusual stratigraphic sequence was
exposed in a tributary stream cut in the PHT terrace near
river km 42 (fig. 9; station 87U14, plate 1). The terrace
surface is about 7 m above river level at this site and is
continuous with the PHT terrace upstream. The stream cut
exposed a basal unit of coarse (clasts to boulders in size),
angular, matrix-supported gravel, overlain by a pebbly silt
and sand unit that has a weak soil formed on its surface. A
sandy pebble gravel that contains stream-rounded frag
ments of brick is present at the surface of the exposure.
The texture of the coarse basal gravel indicates that it
probably was deposited as a debris flow from the
 S

RPorledai-sntaoscneid-tlHioelovcne)le ages
7.
(Figure
PHT
and
strath
bedrock
stripped
of
tprofile
River.
Umpqua
the
on
Creek
McGee
near
terrace faxdpilor
3
figure
in
legend
reSee
two
locations
projected
Note
patterns.
lithologic
of
gradient
the
8)
and
3C
figs.
(see oacnartbion erosion.
micfault
m
25
about
at
profile
step
The
surface.
strath
fexposed
a
of
location
athe
been
has
that
bedrock
din
by
arks
enrteunatieadl
43.

(SP88UR1)
6
<C
Erosional
- os
-

10-
> *

-

surface
Terrace
-

- - - --H. .-:-.

:
3

150
+
9,120
B.P.
yr
o.
2:;
S
:
I
H.

E.yr
4110
+ 2
9,800
B.P.S
c
4:
5
Scarp
fault
W th
*.....
:
_ ---

:
T
#
(SP87UR2) (m)
DISTANCE

SN
i
HRIVER
(
9)

--~SSN0
- \/

*#
UMPOUA0.033
-

130
O
20
70
60
50
100
90
80
0
40
10-
##.
-:

8-
:
g
18 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
;
9,120 + 150 yr B.P.
(SP88UR1)
0 %3
s 3.
| C/D
|#
a i
Figure 8. Stratigraphic column of a PHT (Pleistocenc
Holocene transition) terrace exposure near McGee Creek
on the Umpqua River at station 88U2. See legend in
figure 3 for explanation of lithologic patterns and
symbols.
unnamed tributary stream flowing northward into the river
through the site. The modern radiocarbon age (104% mod
ern) in the silt indicates that these sediments were depos
ited after 1950, probably during a period of increased
precipitation or flooding. This conclusion is supported by
the presence of brick fragments in the upper gravel. The
unusual stratigraphy exposed at the site and evidence of
extensive modern deposition indicate that the PHT terrace
-- 1.04% modern
T(SP87UR15)
Figure 9. Stratigraphic column from a terrace exposure
about 2 km downstream from the highway bridge near
Scottsburg on the Umpqua River at station 87U14. See
legend in figure 3 for explanation of lithologic patterns
and symbols. Note presence of colluvial gravel at base of
the exposure and fluvial gravel at the top of the exposure.
surface is commonly reoccupied by flood waters along
tidal reaches of the Umpqua River.
Another exposure of the PHT terrace was examined
along the lower Umpqua River near the Brandy Bar at
river km 32.5 (fig. 10; station 87U13, plate 1). The ter
race surface is only about 2.5 m above river level at the
site and consists of about 1.0 m of sandy silt overlying a
minimum of 1.0 m of sandy pebble and cobble gravel. No
radiocarbon ages were obtained from this exposure, but
only minimal oxidation was visible so the terrace sedi
ments are probably very young. The stratigraphy exposed
at the site is significant because this is the last exposure
of an upper reach fluvial sequence (silty overbank sedi
ment overlying basal channel gravel) that I observed in
the tidal reach of the Umpqua River. At all other sites
downstream, exposed sediments in the low terraces con
sist of estuarine sand and mud. However, water wells
drilled near the Brandy Bar site show that thick sequences
of sandy gravel are present beneath the fine overbank sed
iment in this terrace; two wells about 0.5 km downstream
from station 87U13 penetrated about 5.5 m of overbank
sediment and a minimum of 13 and 19 m of sandy gravel,
respectively, without intersecting bedrock (Curtiss and
others, 1984).
 TERRACES ALONG THE UMPQUA RIVER
; #####$g
-
:35%;
Vs
Figure 10. Stratigraphic column of a low terrace near
the Brandy Bar on the Umpqua River at station 87U13.
See legend in figure 3 for explanation of lithologic
patterns and symbols. Nearby water wells have penetrat
ed as much as 19 m of sandy gravel without encountering
bedrock.
LOW TERRACES NEAR THE COAST
As is obvious from the preceding discussion, the
height of the PHT terrace decreases rapidly as the river
nears the coast. By about river km 26, the terrace is only 1
m above river level, and by about river km 19 the terrace
merges with salt marshes in the estuary of the Umpqua
River. The transition from fluvial terrace to salt marsh
marks the culmination of terrace convergence that is evi
dent along the lower reaches of all rivers examined in this
study. While fluvial terraces commonly converge down
stream, convergence on Coast Range rivers may be accen
tuated by some combination of coastal subsidence and
rising sea level in the late Holocene. Some estuaries in
northern Oregon and Washington apparently have been
subject to coseismic subsidence associated with regionally
extensive earthquakes on the Cascadia subduction zone
(Atwater, 1987, 1992; Darienzo and Peterson, 1990, Atwa
ter and Yamaguchi, 1991; Nelson and Personius, 1991),
but regional coseismic subsidence along the central Ore
gon coast is equivocal (Nelson, 1992; Nelson and
Personius, 1991).
Several cores in the lowest terrace and in nearby salt
marshes revealed sequences of interbedded peat, mud, and
minor sand. A 4-m-deep auger core in a terrace about 1 m
above river level near Dean Creek at river km 26 (fig. 4D;
station 87U22, plate 1) exposed the muddy sediments
that typify most low terraces near the coast. A similar
sequence was obtained in a 3.75-m-deep auger core in a
low terrace also about 1 m above river level on the Smith
River (fig. 4C, station 87US15, plate 2) about 4.5 km
19
upstream from its confluence with the Umpqua River.
Nelson (1992) found similar sequences of peat and mud in
several cores in salt marshes along the lower Umpqua
River. The muddy sediments in the lowest terraces are
finer grained, contain peats that are not present in fluvial
sequences upriver, and reflect the estuarine influence on
sedimentation in the lower reaches of the river. The total
thickness of these deposits is unknown, because I could
find no records of water wells in these deposits and
because only the upper 45 m of these sediments could be
observed in natural exposures and cores. The presence of
fine-grained estuarine sediments in the lowest terraces is
an important clue to past climatic conditions, because
most of the older, higher terraces near the coast have simi
lar sequences of thick, fine-grained deposits. Some
(most?) of these older terraces were probably deposited
during periods of high sea level similar to present condi
tions, and therefore correlate with past sea-level high
stands. Thus, the present low terraces are modern analogs
to some of the older fluvial terraces.
OTHER HOLOCENE TERRACES
Although younger fluvial terraces are not well devel
oped below the PHT terrace, lower terraces are present at
a few places along the Umpqua River. The most extensive
younger terrace is about 8 m above river level and is inset
into the 12- to 13-m-high PHT terrace from about river
km 133.5 to km 123 (fig. 5). Two natural exposures of the
lower terrace sediments yielded charcoal for radiocarbon
analysis (fig. 3B; stations 87U6 and 88-U3, plate 1). The
consistent radiocarbon ages from these sites (6,545 + 75 yr
B.P. and 6,440 + 60 yr B.P., respectively, fig. 3B) support
the conclusion that this is a younger terrace surface inset
into the older PHT terrace, rather than a stripped PHT ter
race. Inset terraces of this age are not part of a widespread
middle Holocene terrace surface because relatively few
terraces of this height are present along the Umpqua
River.
One intriguing idea about the origin of this terrace is
the possible association of aggradation of this terrace with
the eruption of Mt. Mazama in the headwaters of the
North Umpqua River (D.R. Sherrod, oral commun., 1991).
Mt. Mazama erupted 6.77 ka (Bacon, 1983; Sarna
Wojcicki and others, 1983), and this eruption must have
dumped large quantities of debris-flow sediment into the
North Umpqua River (Sherrod, 1991). No detailed sedi
mentologic studies were done on the sediments in this ter
race, but the radiocarbon ages that indicate that the terrace
sediments are just a few hundred years younger than the
eruption are at least suggestive of a genetic relationship.
A lower, much younger terrace consisting of basal
gravels and sandy overbank sediment extends for about
1.5 km upstream from Elkton (fig. 3A; station 87U8,
20 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
plate 1). Four radiocarbon ages on charcoal and wood col
lected from the loosely consolidated sandy sediment in
this terrace indicate that the overbank sediment is less than
300 years old.
The radiocarbon ages and textural differences in the
young terrace deposits near Elkton indicate that the over
bank sediments were deposited in three discrete units: (1)
a lower, siltier unit that yielded a radiocarbon age of 210
+ 80 yr B.P. and is capped by a thin buried A horizon, (2)
a middle sandy unit that yielded radiocarbon ages of 95 +
40 yr B.P. and 115 + 40 yr B.P., and (3) an upper sandy
unit that yielded a modern radiocarbon age (10 + 70 yr
B.P.). If each of these overbank units was deposited dur
ing a major flood, then the radiocarbon ages may be use
ful in establishing the frequency of such events in the past
few hundred years. The highest sample yielded a modern
(post 1950) age that indicates recent deposition on the sur
face of this terrace, perhaps from the latest large flood
event in the region which occurred in December, 1964
(Rantz and Moore, 1965). Analysis of all the radiocarbon
ages obtained at this site indicates that major depositional
events have taken place on average about every 105120
radiocarbon years for the last two cycles, with depositional
events at about 210 yr B.P., 105 yr B.P. (average of 95 yr
B.P. and 115 yr B.P.), and post-1950 (1964?).
Unfortunately, when these radiocarbon ages are den
drochronologically calibrated (Stuiver and Reimer, 1986),
the resulting ages are not usable for calculating discrete
recurrence intervals. For example, averaging and calibra
tion of the 95 + 40 yr B.P. and 115 + 40 yr B.P. ages
yields at least 5 calibrated age choices ranging from 0 to
243 calibrated years B.P., and a 1 sigma error range of 0
to 262 calibrated years B.P. Calibration of the 210 + 80 yr
B.P. age yields a single age of 288 calibrated years B.P.,
but this age has a 1 sigma error range of 0 to 311 cali
brated years B.P. So a more conservative approach to cal
culating recurrence intervals from these deposits would
indicate that there have been 3 depositional (flood) events
in about the last 300 years, with an average recurrence
interval of about 150 years.
A comparison of radiocarbon ages from the low ter
race at Elkton with the historic record of large floods in
the Pacific Northwest (Helley and LaMarche, 1973) indi
cates that the last two depositional events at Elkton might
be related to major floods in 186061 and 1964. Studies
of deposits related to the 1964 and earlier floods in north
ern California yielded recurrence intervals estimates of
6080 yr (Kelsey, 1987) to 100200 yr (Helley and
LaMarche, 1973) for large landscape-altering floods.
The intense floods of December 1964 (Rantz and
Moore, 1965) may be a recent analog for the flooding
events responsible for the creation of other small Coast
Range terrace remnants, but the restricted distribution of
the Elkton and similar isolated terraces (fig. 5) indicates
that large floods may be more often recorded in the
stratigraphic record as discrete beds in the overbank depos
its of other terraces. Thus, small, isolated terraces like the
low Elkton terrace are examples of terraces formed as
complex responses to relatively minor changes in river
hydrology, rather than the regional, climatic changes asso
ciated with aggradation of the PHT terrace.
OLDER TERRACES
Although less well preserved than the younger ter
races, numerous exposures of older fluvial terrace sedi
ments are present throughout the Coast Range. Only one
older terrace on a small tributary of the Umpqua River
yielded materials suitable for radiocarbon dating, so exper
imental thermoluminescence (TL) dating was attempted on
sediments from several older terraces. These results for the
individual sites will be discussed later. Correlation of flu
vial terraces to the marine terrace record can be a useful
relative-dating tool, so a discussion of marine terrace rem
nants present near the mouth of the Umpqua River will
precede discussion of older fluvial terraces.
MARINE TERRACES
Marine terrace cover sediments consist primarily of
thick sequences of eolian and minor beach sand and are
present at various elevations at several places near the
mouth of the Umpqua River; I measured surface eleva
tions on some of these deposits but did not map or exam
ine the sediments in detail. In most cases terrace back
edges are not preserved and no exposures of wave-cut
platforms were observed, so the elevations of shoreline
angles could not be determined. Golder Associates (1986,
table 6) list, but do not locate, heights of marine terraces
along the Oregon coast, including an elevation of 18.8 m
for a Whiskey Run-equivalent marine terrace (isotope sub
stage 5a, 80 ka) near Winchester Bay. West and McCrumb
(1988, fig. 6) apparently used this and other elevations in
their calculation of uplift rates along the Oregon coast. I
found no terrace back edges and only one exposure of
possible marine terrace sediments in the Winchester Bay
area, in a roadcut on a dirt road on the north side of Win
chester Creek about 0.5 km from U.S. Highway 101.
Although I found no evidence of a preserved terrace, the
elevation of these sediments is consistent with that given
by Golder Associates (1986).
Marine terrace deposits are more abundant on the
north side of the Umpqua River, where several exposures
of sandy terrace sediments are found along the road that
parallels Threemile Creek. Above the east shore of Three
mile Lake an apparent terrace surface and back edge are
preserved at an elevation of 2428 m. Several meters of
parallel bedded eolian sand are visible in an exposure of
 TERRACES ALONG THE UMPQUA RIVER
this surface near the southern end of the lake. No evidence
of a wave-out platform or lower terraces was observed
below the Threemile Lake surface, but most of the topo
graphically low areas in this region are covered by sand
dunes (Beaulieu and Hughes, 1975) that may have buried
such features. Preliminary soils correlations (H.M. Kelsey,
written commun., 1992) indicate that the lowest preserved
marine terrace along this part of the Oregon coast is no
younger than oxygen-isotope substage 5e (125 ka), so I
have concluded that marine terrace sediments at both Win
chester Bay and Threemile Lake probably are parts of the
same substage 5e, 125 ka terrace.
Sandy marine terrace sediments are exposed at eleva
tions of 4454 m along the Threemile Creek road, but
unequivocal terrace surfaces or wave cut platforms were
not observed. Similar deposits at even higher elevations
are exposed in a borrow pit near the intersection of the
Threemile Creek road and U.S. Highway 101, at an eleva
tion of 8595 m. Some speculative but geologically rea
sonable correlations between marine terrace deposits near
the mouth of the Umpqua River and older fluvial terraces
will be discussed.
INTERMEDIATE TERRACES NEAR REEDSPORT
I found a single exposure of charcoal-bearing sedi
ment in an intermediate-aged fluvial terrace on Scholfield
Creek, a tributary that joins the Umpqua River at Reeds
port (station 87U20, plate 1). This exposure is in a small
terrace remnant whose surface stands about 13.5 m above
Scholfield Creek at an elevation of 16.5 m. Two charcoal
samples were collected for radiocarbon dating from the
upper 4 m of the terrace exposure, which consisted of fine
to medium sand with minor silt and sandy gravel interbeds
(fig. 4B). The resulting radiocarbon ages (22,240 + 630 yr
B.P. and >26,400 yr B.P.) are somewhat problematic
because they correlate the Scholfield Creek terrace with a
period of fluctuating sea level that was not conducive to
fluvial terrace aggradation.
Formation and preservation of a 22 ka fluvial terrace
near the coast is unusual because sea level is thought to
have fallen from an interstadial high stand of about 3540
m below present sea level, between 25 and 30 ka (Bloom
and others, 1974; Chappell and Veeh, 1978; Pinter and
Gardner, 1989) to a late Wisconsin minimum about 1718
ka that may have been 100150 m below present sea level
(Bloom, 1983; Chappell and Shackleton, 1986; Tushing
ham and Peltier, 1991). The apparent age of the Scholfield
Creek terrace does not fit well with these changes in glo
bal sea level because fluvial terrace treads near the coast
that may have aggraded during the interstadial high stand
should have been abandoned by incision brought about by
the steep, post-25 ka drop in sea level to the late
Wisconsin minimum. Examples of stream terraces that
21
may correlate with the 2530 ka interstadial high stand
have been described in central and southern California by
Alexander (1953) and Birkeland (1972), and in northern
California by Merritts and Vincent (1989, fig. 5), but
2025 ka terraces are uncommon and appear to be
restricted to areas undergoing very high rates of uplift,
such as near Ventura in southern California (Rockwell and
others, 1988).
The history and extent of late Pleistocene sea-level
fluctuations are poorly known along the Oregon coast, so I
searched for evidence of a late Wisconsin sea-level mini
mum by examining the bathymetry of the region off the
mouth of the Umpqua River on the U.S. Geological Sur
vey Reedsport, Oregon, 1:100,000 scale topographic
bathymetric map. Numerous minor irregularities are appar
ent on the sea bottom, but a major break in slope that
might be related to a sea-level low stand is present at an
elevation of 104 m (fig. 11), a depth that is consistent
with recent models of sea-level change in the late Quater
nary (Tushingham and Peltier, 1991). However, I found no
evidence of a submarine canyon or an Umpqua River
paleochannel on the continental shelf, so either rapid sedi
mentation has obliterated any traces of such features or the
drop to the late Wisconsin sea-level minimum had only
temporary effects on the lower reaches of the Umpqua
River and its tributaries. Thus, one possible explanation
for the problematic age of the Scholfield Creek terrace is
that the creek was isolated from any incision brought on
by the post-25 ka drop in sea level until 2022 ka, when
entrenchment of the Umpqua River system retreated
upstream to the vicinity of the Scholfield Creek terrace.
The Scholfield Creek terrace ages have also been
affected by the recently documented systematic errors in
the radiocarbon time scale determined from comparisons
with geomagnetic intensities (Stuiver and others, 1991)
and precise uranium-series dating of fossil corals (Bard
and others, 1990). These studies indicate that radiocarbon
ages in the 20- to 30-ka range are 2,5004,000 years too
young. Thus, the two Scholfield Creek radiocarbon ages
yield calibrated ages of 2530 ka, which are more compat
ible with a period of aggradation during an interstadial
sea-level high stand, rather than to the period of rapidly
falling sea level following this high stand.
Terraces are poorly preserved along Scholfield Creek,
so correlation of the dated terrace with surfaces on the
lower part of Scholfield Creek and on the Umpqua River
near the confluence is uncertain. However, the dated terrace
might correlate with a small terrace remnant at an elevation
of about 9 m in the town of Reedsport, just upstream from
the confluence with the Umpqua River (station 87U21,
plate 1). About 5 m of sandy silt and silty sand, oxidized to
7.5YR hues (color terminology of Munsell Color Co.,
1954), were exposed in an old roadcut in this intermediate
terrace (fig. 12). Unfortunately, no charcoal for radiocarbon
dating was found, but a TL age of 14.9 + 3.1 ka was


Dioapgorgarmapahtic part
11.
tFigure
Schol
River,
Umpqua
Ocean,
Pacific
profiles
are
lines
Heavy
area.
Reedsport
the
in
terraces
marine
and
fluvial
of low-gradient fluvial
The
arrows.
with
marked
River
Umpqua
confluence
and
coast
the
Creek;
field
in
are
Creek
Scholfield
of HRhigh
circles:
filled
with
shown
are
text
in
discussed
remnants
of
Locations
range.
tide
Ithe
Reed
at
terrace Ccrhotvelirfdireancleds profiles
Re-indtserpmoedrita,e Creek,
SPsport,
are
lines
Dashed
River.
Smith
the
on
SRhigh
Lake,
Threemile
at
terrace
TLmarine terrace
ctentative
between
Tgradients
text.
the
in
discussed
pterraces
several
readings,
altimeter
from
are
data
horptoregleralmtaipeothnerisdc; errors
d:eoai12prs04oic,uga0rtsea-0imps-ocehanilstcde
(see
and
vt1methods)
mmaps;
greater
reflect
circles)
(filled
symbols
terrace
of
size
in

Scholfield
Creek
SR
*
g
River
Umpqua
Pacific
Ocean

3.
CoCoast
nfluence
t
T

TLC--
-

(km)
COAST
TO
RELATIVE
DISTANCE

-100O
Possible
sea-level

Stand
low
LL]
--

-1501I
------|--
I-I- -30140
O
-10
20
30
200

exag eration.
vertical
Note
terraces.
higher
of
heights

50 O

LLI LLI l
5. LL] C) OH
-50 > H<[ LL] DC E
-

I
 TERRACES ALONG THE UMPQUA RIVER
*Has 3.1ka
...](SP87UT22)
I
# || i
Figure 12. Stratigraphic column of an intermediate ter
race in Reedsport at station 87U21. See legend in figure
3 for explanation of lithologic patterns and symbols.
obtained on sediment from this exposure. The projected
gradient between the intermediate Reedsport terrace and
the terrace on Scholfield Creek (0.0008) is substantially
higher than the tide influenced gradient of modern Schol
field Creek (0.0002) in the same reach (fig. 11); gradients
on the upper part of Scholfield Creek are 0.00270.0036. If
these two terrace remnants are correlative, then the cali
brated Scholfield Creek radiocarbon ages indicate that the
TL age on the intermediate terrace at Reedsport is too
young by 10,000 to 15,000 years.
The Scholfield Creek terrace might also correlate
with other intermediate terrace remnants located near
Providence Creek, 11.5 km north of station 87U18.
These terraces are mapped as two narrow, northwest
trending surfaces with elevations of about 12 m on plate 1.
23
No exposures were found in these terraces. The projected
gradient between the Scholfield Creek terrace and the
intermediate terrace remnants near Providence Creek is
about 0.0004. This gradient is less steep than the projected
gradient to the 9-m-high intermediate Reedsport terrace,
but no age information was obtained from the Providence
Creek terraces, so either the Providence Creek or interme
diate Reedsport terraces might reasonably be correlated
with the Scholfield Creek terrace (fig. 11).
My limited age and gradient data indicate that fluvial
sediments in a late Pleistocene terrace remnant on Schol
field Creek, and possibly in one or more intermediate ter
races on the Umpqua River near Reedsport, may have
aggraded during an interstadial sea-level high stand 2530
ka, and were subsequently abandoned during the fall of
sea level to the late Wisconsin minimum.
HIGH TERRACE AT REEDSPORT
An extensive high terrace surface underlies much of
the western part of the town of Reedsport at about river
km 18 (stations 87U15 and 87U18, plate 1), and occu
pies an abandoned meander bend of the Umpqua River at
an elevation of 1820 m. Water wells in the high Reeds
port terrace about 0.6 km northwest of station 87U18
have penetrated more than 40 m of unconsolidated sedi
ment without encountering bedrock (Curtiss and others,
1984), so a very thick sequence of unconsolidated fluvial
and estuarine sediment must underlie the meander bend.
Two exposures in the high Reedsport terrace revealed sim
ilar stratigraphic sequences of massive silts overlying fine
to medium sand and pebbly sand (fig. 13); these sediments
are similar to most other fluvial and estuarine terrace sedi
ments exposed along the lower reaches of all rivers exam
ined in this study. Charcoal was not found in the two
exposures, but a TL age of 65.2 + 7.3 ka was obtained on
silt from the station 87U18 site (fig. 13A).
The TL age is somewhat problematic, because it indi
cates that the high terrace at Reedsport correlates with a
sea-level low stand at oxygen-isotope stage 4 (Shackleton
and Opdyke, 1973; Chappell and Shackleton, 1986; Mar
tinson and others, 1987). I cannot substantiate the TL age
with other numerical dates, but the presence of marine ter
race remnants near the mouth of the Umpqua River sug
gests correlation of the Reedsport terrace to a different
part of the oxygen-isotope record. The surface of the pos
sible substage 5e (125 ka) marine terrace remnant at
Threemile Lake is 46 m higher than the surface of the
high fluvial terrace at Reedsport, but common variations
of +2 m in the original altitude of the shoreline angle
(Wright, 1970), the presence of an unknown thickness of
marine and eolian sediment on the marine terrace surface,
or minor tilting of folding could easily allow the correla
tion of these terrace surfaces.
 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

-
2 - -

*65.2 + 7.3 ka
|TOE (SP87UT18)

; 1 '...? S.
t..:06:06.

;
"... r I

c
| 3 |
a i #
| 5 |
Figure 13. Stratigraphic columns of two exposures of the high terrace underlying
the town of Reedsport: (A) station 87U18 and (B) station 87U15. See legend in
figure 3 for explanation of lithologic patterns and symbols.

The two possible terrace ages discussed above (65 ka
from the TL age or 125 ka from correlation with the
marine terrace record) associate the high terrace at Reeds
port with very different sea-level positions and geomor
phic settings. A 125 ka age would correlate the fluvial
terrace with a time period in which sea level stood about 6
m above present sea level (Chappell and Shackleton,
1986), so fluvial conditions near the mouth of the Umpqua
River were probably similar to those of today. In contrast,
a 65 ka age would correlate the Reedsport terrace with a
sea-level low stand (oxygen-isotope stage 4) at a time
when the heights and ages of uplifted coral reefs on the
Huon Peninsula in New Guinea indicate that sea level
stood more than 60 m below present sea level (Chappell
and Shackleton, 1986). The lower Umpqua River must
have been flowing in a deeply entrenched canyon under
such conditions. Marine and near-coastal fluvial terraces
formed during sea-level low stands are rarely preserved
above sea level except in areas of very high uplift (Lajoie,
1986). No studies of coastal uplift have shown uplift rates
of more than about 0.4 mm/yr for the central Oregon coast
(Adams, 1984; West and McCrumb, 1988), and more
recent studies (H.M. Kelsey, written commun., 1992)
indicate coastal uplift rates are 30.2 mm/yr in this region,
so high uplift rates are an unlikely cause of preservation of
a 65 ka aged fluvial terrace at Reedsport.
The geomorphology and sedimentology of the high
terrace at Reedsport support the inference that the terrace
formed during a sea-level high stand. The terrace tread
nearly fills its ancestral valley (pl. 1), a pattern which geo
morphically resembles the present valley of the Umpqua
Riverthat is, a relatively flat, steep-walled valley filled
with fine-grained fluvial and estuarine sediment. If the
present morphology of the lower Umpqua is typical of
valleys formed during sea-level high stands, then it seems
highly unlikely that an extensive fluvial terrace formed
during a sea-level low stand could be preserved above the
modern river. Therefore, I conclude that the TL age of
about 65 ka for the Reedsport terrace deposits has under
estimated the age of this terrace by about 60,000 years,
and that the terrace is correlative with possible 125 ka
marine terrace sediments near the mouth of the Umpqua
River (fig. 11).
If the high fluvial terrace at Reedsport is correlative
with a 125 ka marine terrace, then the rapid drop in base
(sea) level that followed the substage 5e high stand is a
 TERRACES ALONG THE UMPQUA RIVER 25

likely mechanism for initiating the incision that cut off the
Reedsport meander and stranded the high Reedsport flu
vial terrace. Drastic drops in base level (on the order of
many tens of meters) may be the most likely mechanism
for vertical incision and abandonment of most terrace
treads located within a few tens of kilometers of the coast.
: -200 ka
HIGH TERRACE ALONG CRESTVIEW DRIVE (SP87UT20)

A very high (90- to 92-m elevation) fluvial terrace is

preserved along Crestview Drive in the southeastern part
of Reedsport near river km. 18.5 (station 87U19, plate 1).
A deep roadcut in this terrace revealed a 6-m-thick basal
channel facies of sandy pebble and cobble gravel with
minor interbeds of fine sand, overlain by a 3-m-thick
overbank facies of sandy silt (fig. 14). The stratigraphy in
this exposure is similar to that seen in terraces along the
upper reaches of the Umpqua River, but the gravel facies
#
is thicker and the clasts are somewhat coarser than in
upstream exposures. Although an unknown amount of the
upper surface of the Crestview terrace may have been
removed by erosion, the Crestview sequence is consider
ably coarser than other terrace sequences near the coast
(compare figs. 13 and 14). The sedimentologic differences
between the Crestview and other Reedsport area terrace
exposures are so striking that depositional environments
near the coast may have changed considerably since the
Crestview terrace was deposited. The sediments in the
high Reedsport terrace described previously appear to
have been deposited in an estuarine-dominated environ
ment similar to the present tidal reach of the Umpqua
River; however, the coarser Crestview sediments are
clearly a product of a higher energy fluvial environment
similar to the present Umpqua River upstream from tidal
influence.

The height of the Crestview terrace indicates that the
age of this terrace is well beyond the range of radiocarbon
dating, so a TL age was obtained on the silty overbank
deposits (fig. 14). The TL analysis determined that the
Crestview sediments are at TL saturation and indicates a
probable age of greater than 200 ka. This minimum age is
consistent with the height of the terrace and the soil devel
opment found in the deposit (oxidation extends through the
entire 10-m-deep exposure, and 2.5YR hues are present in
the B horizon). An age of >200 ka is also consistent with
tentative correlation with the marine terrace record.
Although poorly preserved, marine terrace sediments are
present about 7 km downstream at about the same elevation
as the Crestview terrace. The 8595 m elevation of these
marine deposits indicates that they are substantially older
than the 24- to 28-m-high, isotope-stage 5 terrace at Three
mile Lake, so they most likely correlate with a stage 7 (230
ka) or older sea-level high stand. Some aspects of soil
development on the Crestview terrace, particularly soil
Figure 14. Stratigraphic column of an exposure of a
high terrace along Crestview Drive near Reedsport at sta
tion 87U19. See legend in figure 3 for explanation of
lithologic patterns and symbols.
hues, are consistent with soils on the highest marine terrace
at Cape Blancothe Poverty Ridge terrace(Bockheim
and others, 1992), which is thought to be >500 ka (H.M.
Kelsey, written commun., 1992).
If the Crestview terrace correlates with marine terrace
deposits, then the coarse texture of the fluvial sediments
implies high stream gradients and the absence of a well
established estuary at the mouth of the Umpqua River at
the time of terrace aggradation. Perhaps the mouth of the
Umpqua River during Crestview time was similar to the
mouths of the present day Elk and Sixes Rivers in the
Cape Blanco area of southern Oregon (fig. 2); these rivers
are nontidal, gravel-dominated fluvial systems that empty
26 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

into the Pacific Ocean through relatively narrow (1- to 2 4

km-wide), flat-bottomed valleys. The gradients of the
present-day lower portions of the Elk and Sixes Rivers
(0.0009 and 0.0007, respectively) are much higher than
the gradient of the tidal reach of the Umpqua River
(0.0001). Unfortunately, the gradient between the Crest
view terrace and possibly correlative marine terrace depos
its cannot be calculated because of the poor preservation
of both the marine and fluvial terrace surfaces, but the
coarse texture of the Crestview terrace sediments indicates
a steeper gradient on the ancestral lower Umpqua River.
The inferred higher gradient indicates that either the whole
ancestral Umpqua River had higher gradients or the rate of
uplift of this part of the Coast Range was higher during
Crestview time. High uplift rates certainly have affected
the hydrology of the Elk and Sixes Rivers, because the
Cape Blanco region has the highest uplift rates (0.81.5
mm/yr) determined along the entire Oregon coast (West
and McCrumb, 1988; Kelsey, 1990; Muhs and others,
;
1990). Because stream gradients commonly decrease as
stream discharge increases, a presumably smaller ancestral
Umpqua River should have had a higher gradient at its
mouth than the present river. So both greater uplift and a
smaller discharge may have contributed to higher gradi
ents when the Crestview terrace was formed. Some rivers
in the Pacific Northwest have demonstrated similar pat
terns of increasing gradients with older terraces, such as
the Cowlitz River in western Washington (Dethier, 1988,
fig. 4), while others have essentially parallel older ter
races, such as the Mattole River in northern California
(Merritts and Vincent, 1989, fig. 5). Unfortunately, the
preservation of older fluvial terraces is so poor that longi
|
tudinal profiles of older terrace remnants (fig. 5) must be
considered speculative, and therefore yield relatively little # || #
conclusive information about the gradients of the ancestral
Umpqua River.
Figure 15. Stratigraphic column from an exposure of a
high terrace along Henderer Road about 3 km west of
HIGH TERRACES ALONG HENDERER ROAD Elkton at station 87U1. See legend in figure 3 for expla.
nation of lithologic patterns and symbols.
Several high terraces are relatively well preserved
along the Umpqua River in the area around Elkton
between river km 65 and 95 (fig. 5; pl. 1). About 4 meters hand specimen. Strong oxidation extends through the
of extremely weathered fluvial sediment is exposed in one entire exposure and into several meters of exposed saproli
tic sandstone bedrock.
of the high terraces in a roadcut near the upstream end of
the terrace at river km 75 (fig. 15; station 87U1, plate 1). Soils are better developed on the Henderer Road ter
The exposed sediments differ from the usual fluvial race than on the previously described Crestview terrace,
stratigraphy found in upper Umpqua River terraces in that even though the difference in height of these terraces
the sandy gravel at the base of the section is thinner than above the Umpqua River (41 m versus 90 m, respectively)
basal gravels at most other sites. The thin gravel is over indicates that the Henderer Road terrace is much younger.
lain by sandy silt that contains thin gravelly interbeds and I believe the apparent change in rates of soil development
some gravel clasts floating in the sandy silt matrix. Soils between these two sites is related to differences in climate,
on this terrace are extremely well developed, with 2.5YR in particular to the occurrence of the summer dry, the
hues in the B horizon, and obliteration of most deposi hotter, drier, sunnier weather experienced some distance
tional fabric. In addition, all gravel clasts are crushable in inland from maritime influence along the Oregon coast.
 TERRACES ALONG THE SMITH RIVER
Coastal soils develop under cooler, moister, and foggier
summer conditions than do soils in inland parts of the
Coast Range, so the drier summer soil conditions inland
may be more conducive to increased oxidation. The
increased oxidation is clearly apparent in the Henderer
Road and other terrace exposures on the Umpqua River
that lie more than about 40 km from the coast. However,
this apparent limit of coastal influence on soil develop
ment may not extend as far inland on other parts of the
Oregon coast, because coastal fogs often move farther
inland along major river valleys than they do across
unbroken mountains. Additional variations in coastal influ
ence on soil development through time may exist as a
result of changes in climate related to fluctuating sea lev
els. Such variations in soil development are examples of
the problems inherent in using soils as a tool for correlat
ing terrace remnants along Coast Range rivers.
A TL age of 116 it 20 ka was obtained on silty sedi
ment from one of the higher terrace exposures along Hen
derer Road (fig. 15). As with the other TL sites, I
attempted to sample unweathered sediment for analysis,
but this proved to be impossible because soil development
extended through the entire exposure. The affect of pedo
genesis on the TL signal preserved in these sediments is
unknown, but oxidation, influx of illuviated silt and clay,
and fluctuating groundwater conditions probably have
altered the amounts and types of radioactive sources in the
sediment. However, the TL age obtained on the higher
Henderer Road terrace may be geologically reasonable,
especially if the TL age of 65.2 + 7.3 ka on the high ter
race at Reedsport, to which the Henderer Road terrace
might be correlated (fig. 5), is a minimum age. The TL
age of 116 it 20 ka on the Henderer Road terrace appears
to be accurate within its laboratory error limits if both flu
vial terraces are graded to the Threemile Lake marine ter
race, which in turn is correlative with a substage 5e, 125
ka sea-level high stand. This is a reasonable interpretation,
but the advanced soil development on the high Henderer
Road terrace, lack of physical continuity, and poor preser
vation of both fluvial and marine terraces suggests that
other correlations cannot be ruled out.
Several roadcuts in a lower terrace along Henderer
Road at about river km. 6870 (stations 87U3 and
87U23, plate 1) exposed similar sequences of sandy silts
overlying sandy pebble and cobble gravels (fig. 16). The
elevation of the lower Henderer Road terrace surface near
these exposures is about 43 m, about 26 m above the
Umpqua River. No age information was obtained on this
terrace, but comparable well-developed soils are present
on the deposits at both sites. Strong oxidation extends
through the entire exposures, most gravel clasts are crush
able in hand specimen, and colors reach 5YR hues in the
B horizons. If the higher Henderer Road terrace correlates
with a possible substage 5e (125 ka) marine terrace at
Threemile Lake, then the lower Henderer Road terrace
27
surface might be graded to sea-level high stands at 80 or
105 ka, but marine terraces of this age are probably below
sea level along this part of the Oregon coast (H.M.
Kelsey, written commun., 1992).
TERRACES ALONG THE
SMITH RIVER
The Smith River, one of the largest (900 km drain
age area; Oregon State Water Resources Board, 1958)
tributaries of the Umpqua River, enters the Umpqua River
near Reedsport (fig. 2). Unfortunately, fluvial terraces
along the Smith River are even less well preserved than
along the Umpqua River (fig. 17; pl. 2). Most terraces are
small, but at four locations terraces are better preserved in
abandoned meander bends. As with the Umpqua River,
terrace sediments are found at heights as much as 100 m
above present river level, but lower, younger terraces are
more Common.
The only terrace that can be physically correlated
with any certainty along the Smith River is a mostly con
tinuous terrace, 210 m above river level, that probably is
correlative with the PHT terrace on the Umpqua River.
Unfortunately, no radiocarbon ages were obtained from
upstream exposures of this terrace on the Smith River that
could be compared to dated sites on the Umpqua River.
Several radiocarbon ages were determined on charcoal
from terraces in abandoned meander bends near Brainard
Creek and in The Island near Smith River Falls, and near
the confluence of Vincent Creek and Smith River; a single
TL age was obtained from a high terrace near the mouth
of the river. Most of these ages are not consistent with
other ages on the Smith River or with the better estab
lished chronology on the Umpqua River. The dated sites
and several other terraces are discussed below.
TERRACES UPSTREAM FROM THE ISLAND
Fluvial sediments are exposed in a logging roadcut
along Vincent Creek, just upstream from the confluence of
the creek with the Smith River near river km 53 (station
87US7, plate 2). In the exposure about 1 m of fluvial
sandy silt overlies about 0.75 m of sandy gravel (fig. 18);
an accelerator (AMS) radiocarbon analysis on small frag
ments of charcoal from the sandy silt yielded an age of
7,950 + 110 yr B.P. This age seems anomalously young
because the terrace surface is about 19 m above the eleva
tion of Vincent Creek and about 24 m above the elevation
of Smith River, 250 m downstream. An infinite radiocar
bon age (>43,600 yr B.P.) on a lower terrace near Bear
Creek about 4 km downstream and the presence of abun
dant decomposed roots, worm casts, and animal burrows
 Stratigraphic columns of two exposures of an intermediate terrace along
Henderer Road, 68 km west of Elkton: (A) station 87U3 and (B) station 87U23.
AN VYRZNOSSENYOT.
-}}30:|-
;);
J40.
}'o<>Q Gravel
FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

- -
<>0}
V;?}{};-
\-;&&Sand - -

See legend in figure 3 for explanation of lithologic patterns and symbols.

-
i|$ Mud

}%0|-

SHE LEWN
Gravel


O`Sojc;0\\$3. Os |-

$%$ Sand
- v

* : * *:
..
i.*-$3

.z2.
$$$$$$$
0$$.

$\.\))$ $$$$$$$
..}}. }. }. }}} -}}$$

% |- -
Mud
- -

Figure 16.
I
CN

S

SHE LEW
28

Loaxndpgilor
17.
fFigure
5
figure
see
River;
Smith
reach
lower
the
along
terraces
Reprofile
symbols.
charcoal
on
is
B.P.
yr
85
10,165+
of
age oatcnuadrtibnoanl
........"
" 100

#
80
|.
2-| *
_2
|
|60
|
=
OA
+-
IST-I
2-
~
H-
i

-
Hyr
|9. .- - 5- -
+
2,685
B.P.
-:
-H
terraCe
PHT
| |
4*3,600
B.P.
yr ...
.
*

terrace
Marine
MT
|| | >40
yr . . . .2. . *
1ka
+
155 - 5.
+0.P-*
"

*1,020
B.P.
yr
....
B.P.
->39.500y
. . .....*River
S. MT.
. . ."
falonufvew
a
River
Umpqua
of
mouth
near
projected
are
terraces
Marine
area.
Island
The
in
csediment
the
from
downstream
km liuaeln-cfea.n
*
.....
d
110
+
*7,950
B.P.
yr
| X-
.
Falls
mith

'" F
.....
85
+
10,165
B.P.
yr
...

|.
CONFLUENCE
FROM
DISTANCE
(km)
RIVER
UMPQUA
WITH
Ta;

age
- L

...!...
r
...
...
*-

ro
130
O
20
50
60
0

RadiocSMITH
age
140 RIVER
arb
on|.'"

Teerrace
m|
- asurem..]
ent | #
meander|
Abandoned
2- - |
r
1825-6sy
f......A.
B.P.|l-

Tide
A -
3.
120 range o;
....."
MT
| -
|
(d
[........20
H -

|
'"
...
A- -
||
TR
|

Mr.
|.......
.*| |
-
30 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
*7,950 + 110 yr B.P.
(SP87USR5)
; racy of several Holocene radiocarbon ages on terraces
Figure 18. Stratigraphic column of a terrace exposure
on Vincent Creek near the confluence with the Smith
River at station 87US7; the radiocarbon age is probably
too young (see text). See legend in figure 3 for explana
tion of lithologic patterns and symbols.
in the exposure indicate that the Vincent Creek radiocar
bon sample may have been contaminated with younger
carbon.
Soil development cannot be used with much confi
dence at this site because the upper meter or two of the
terrace was not exposed and the roadcut was located in a
partially eroded part of the terrace surface. The present
soil at the site is weakly developed, with only minor oxi
dation and clast weathering visible in the exposure; moist
colors in the sediments are 10YR 5/4 to 10YR 5/6, so lit
tle reddening of the soil is apparent. A terrace of similar
height along Henderer Road on the Umpqua River is
more oxidized, has 5YR hues in the B horizon and hand
crushable gravel clasts, so the height of the Vincent
Creek terrace may be related to higher incision rates on
Vincent Creek rather than to advanced age. As with sev
eral other sites on the Smith River, the geologic and
radiocarbon data are somewhat contradictory. The height
of the Vincent Creek terrace indicates that the radiocar
bon age from the site is too young, but the weak soil
development in the deposit supports a relatively young
age. The gradient of Vincent Creek is four times as steep
as the Smith River in this area (0.008 versus 0.002), thus,
both sample contamination and higher incision rates may
be the cause of the apparent anomalously young age of
the Vincent Creek terrace.
Several terrace surfaces are preserved along a fairly
straight stretch of the Smith River near Bear Creek, 34
km upstream from Smith River Falls (stations 87US5 and
87US6, plate 2). Exposures in two of these terraces
revealed similar sequences of basal sandy pebble and cob
ble gravel overlain by pebbly sand and pebbly sandy silts
and silty sands (fig. 19). An infinite AMS radiocarbon age
(>43,600 yr B.P.) was obtained on disseminated charcoal
collected from an exposure of the lower terrace that stands
about 15 m above the Smith River (fig. 19A). This radio
carbon age is important because it casts doubt on the accu
presently found at greater heights above the Smith River.
As previously discussed, young organic materials (roots,
filled animal burrows) are abundant in most terrace expo
sures, so some anomalously young radiocarbon ages were
expected. Anomalously old ages, however, are more diffi
cult to explain because sources of dead carbon, such as
fragments of coal or organic-rich shale, are rare in the
underlying bedrock in this part of the Coast Range. Older
fragments of charcoal could be a source, but the very
small fragments preserved in the station 87US5 exposure
could not have been transported very far by the Smith
River without disintegrating. Therefore, I conclude that the
infinite radiocarbon age from the lower terrace is accurate,
and that most of the Holocene ages on nearby terraces,
including Vincent Creek and most of the terraces in The
Island (to be discussed) are probably not accurate esti
mates of terrace age.
An exposure in the higher terrace near Bear Creek,
which stands about 50 m above the Smith River (fig.
19B), revealed a stratigraphic sequence similar to but
somewhat coarser than the nearby lower terrace. Soil
development was better in the upper terrace (higher clay
content and the sand had a firmer consistence in the B
horizon), but in both exposures the soils are only oxidized
to 10YR hues and the gravel clasts in the exposures are
only slightly weathered and not crushable in hand speci
men. This degree of weathering is substantially less than
terraces of similar height on the nearby Umpqua River
(for example, the Henderer Road terraces), which implies
either that incision rates on the Smith River are very high,
or that soil development at this site has been inhibited by
coastal weather patterns. The Bear Creek terraces are
located about 30 km from the coast and the Henderer
Road terraces are located about 45 km from the coast, so
the limit of coastal climatic suppression of soil oxidation
may lie somewhere between 3045 km along major
stream valleys in this part of the Coast Range.
TERRACES IN THE ISLAND
The Island is a large abandoned meander bend
located at about river km 46 on the Smith River (stations
87US10, 11, and 12, plate 2). Smith River Falls is
located about 0.6 km upstream from the meander cutoff
and probably is a retreating knickpoint created by the
 TERRACES ALONG THE SMITH RIVER 31

:* - 43,600 yr B.P.
(SP87USR3)

# #

Figure 19. Stratigraphic columns of two terrace exposures upstream from Smith River
Falls near Bear Creek: (A) exposure of lower terrace at station 87US5, and (B) exposure
of higher terrace at station 87US6. See legend in figure 3 for explanation of lithologic
patterns and symbols.

abandonment of about 6 km of former channel of the
Smith River now isolated in The Island. Several fluvial
terraces are preserved in the area as a result of abandon
ment, but the present geomorphology of the bend has been
complicated by subsequent landsliding, drainage reversals,
and alluvial deposition by small streams. A topographic
profile measured along the present valley bottom of The
Island illustrates some of the pertinent aspects of the geo
morphology of the area (fig. 20): (1) remnants of at least
four pre-abandonment terraces are preserved in the bend;
(2) a substantial amount of colluvial and alluvial sediment
has been deposited in the bend since abandonment; (3) the
gradient of the modern Smith River (0.0029) in the vicin
ity of The Island is only slightly higher than the projected
gradients of the four terraces in the bend (average of four
values=0.0023); and (4) two radiocarbon ages yielded
unexpectedly young ages and probably are only minimum
ages.
The timing of the abandonment of the The Island is
unknown because of inconsistent radiocarbon ages. Char
coal from the lowest abandoned terrace (station 87US12,
plate 2), which probably was the Smith River flood plain
at the time of abandonment (dashed line in fig. 20),
yielded an age of 1,020 + 50 yr B.P., but this age is
almost certainly a minimum value. This sample was taken
from a wedge-shaped zone of reddish-brown silty sand
containing abundant angular wood and charcoal fragments
in the sandy overbank facies (fig. 21A). The shape and
color of this zone and the abundance of apparently non
fluvial organic debris indicate that this zone is probably a
burned tree root or tree-throw cavity filled with sediment
and organic material that post-dates the fluvial terrace sur
face. Therefore, the radiocarbon age yields only a mini
mum age for the abandonment of this terrace.
A roadcut in a fluvial terrace about 21 m above the
lowest terrace (station 87US10, plate 2) exposed an
apparently eroded section of fine sand and sandy silt over
lying bedrock (fig. 21B). This sediment has probably been
reworked by colluvial processes because the exposed sec
tion is very thin, does not include the basal sandy gravel
found in almost all other exposures of fluvial sediments in
the area, and contains some angular pebbles. The exposure
also appears to have been extensively bioturbated by
rodents and contains abundant decomposed roots, so some
decomposed roots may have been included in the radiocar
bon sample. Thus, the radiocarbon age of 2,860 it 215 yr
B.P. also yields only a minimum age of this terrace.
A third radiocarbon age (10,165 + 85 yr B.P.) from a
roadcut exposure at station 87US11 (pl. 2) gives a better
assessment of the antiquity of terraces in The Island. As is
evident from its position on the topographic profile (fig.
20) and on plate 2, the dated sediments lie near the top of
a thick sequence of post-abandonment colluvial and allu
vial sediment. The stratigraphy in the exposure (fig. 21C)


Dioapgorgarmapahtic
20.
tFigure
shown
Also
2).
plate
Island
The
(labelled
River
Smith
on
bend
meander
abandoned
an
in
line)
(solid
bottom
valley
present
the
of
profile ap roximates ages
line)
(dashed
surface
terrace
projected
lowest
The
(0.0029).
reach
this
in
River
Smith
modern
gradient
than
steep
less
slightly
only
are
of
plain
flood
the aobasdtn-iadobacnadmorenbmeotn;t sequence,
meander
time
at
River
pSmith
of
volume
large
the
rnote
youngest
two
The
surface.
this
overlying
area)
(unshaded
sediment rolasdtu-iavobiacnadlor-nbfmeoant km
error;
in
probably
are
asterisks)
with
a(marked
charcoal
on
is
fan
Creek
Buck
from
page
of
top
the
near
sediment
so ocarbon of
maerrace
yields
rit
infinite
An
cutoff.
meander
tof
from
B.P.)
yr
(>43,600
fage
a
Creek,
Bear
near
pre
the
with
correlates
probably
19A)
(fig.
upstream
inimum
diew baebtanerdeosntmieante
a
gives
and
Island
The
in
terrace
fan
Creek
Buck
between
trapped
is
drainage
where
located
marsh
small
Note
meander.
abandoned
age
the
loapnodgsrlaipdheisc;
Tdownstream
Creek.
Little
and
Creek
Buck
between
divide
drainage
modern
the
forms
area
this
photogram
several
readings,
altimeter
from
are
data
River
|Smith
River
Smith
215
+
$2,860
B.P.
yr
OI
T
T
I|
I
|
I
I
T
IT
r-T
I-T
I proer-aebalnadtoinmoenst
several
care
probable
and
lines),
solid
(short
remnants
terrace
0.0023)
(average
surfaces
correlated
the
of
gradients
The
surfaces.
these
between
lines)
(dotted

|
Creek
Buck
<!--
D
Creek
Buck
Little A 50
+
*1,020
B.P.
yr

fan
Creek
Buck
Landslides
Marsh

(m)
DISTANCE

scale-topographic
1:24,000
and
profiles
metric
maps.

e----

- -

O
12000
3000
4000
000
5000
6000
O O O O O O 20 10
8 7 6 5 4 3
 TERRACES ALONG THE SMITH RIVER
2 . .. . 1 ; :
|*1,020 + 50 yr B.P. #.
|(SP87USR8) #:
C/D g n 2,860 + 215 yr B.P.
, f^*A* (Sgusfie,
H. 1 H. 0 N
> >
N
N
>
Cl
# #
Figure 21. Stratigraphic columns of terrace exposures in The Island: (A) exposure of lowest terrace at station 87US12, (B) exposure of
intermediate terrace at station 87US10, and (C) exposure of Buck Creek alluvial-fan sediments at station 87US11. See legend in figure 3
for explanation of lithologic patterns and symbols.
is somewhat unusual because a thin bed of sandy pebble
and cobble gravel is interbedded with pebbly sandy silt
and silt. This stratigraphy and the geomorphology at the
site indicate that the exposed sediments were deposited in
an alluvial or debris fan formed where Buck Creek enters
the abandoned cutoff and are not part of a Smith River
fluvial terrace. This conclusion is supported by the charac
teristic fan shape of the ground surface where Buck Creek
enters the meander, and lack of older fluvial terrace sur
faces in the area (pl. 2). In addition, the roadcut in the
Buck Creek fan is located near the center of the valley,
where preservation of a high Smith River terrace remnant
is unlikely. Projection of the lowest Island terrace (dashed
line in fig. 20) beneath the colluvial and alluvial sediments
now present in the cutoff shows that as much as 38 m of
post-abandonment sediment has been deposited at the apex
of the Buck Creek fan. Such a large amount of sediment
has accumulated in the meander because Buck Creek,
which now occupies most of the former valley of the
Smith River, is underfit and incapable of carrying the
large sediment loads supplied to the valley bottom from
adjacent slopes. The location of the Buck Creek radiocar
bon sample near the surface of the fan sequence may indi
cate that the latest period of significant alluviation in this
part of The Island occurred in latest Pleistocene or early
Holocene time, coincident with aggradation of PHT ter
races along several Coast Range rivers (see Discussion
section).
The radiocarbon age from the previously described
lower terrace near Bear Creek (>43,600 yr B.P.) may be a
more reasonable constraint on the minimum age of mean
der abandonment than the three radiocarbon ages from
33
1
Lll
H.
Lll
> 0
10,165 + 85 yr B.P.
(SP87USR7)
The Island, because the lowest Island terrace and the Bear
Creek terrace are at about the same height above the
Smith River (fig. 17), and because the infinite radiocarbon
age is more consistent with the large amount of post
abandonment sediment present in The Island (fig. 20).
TERRACES BETWEEN THE ISLAND AND
BRAINARD CREEK
Downstream from The Island, fluvial terraces are best
preserved on the insides of meander bends and in aban
doned meanders near Sulphur Springs at about river km
35 and near North Fork at about river km 26 (fig. 17).
Both abandoned meanders contain high, old terraces that
are mostly covered by colluvium. No exposures were
apparent so these terraces were not examined in detail.
However, the terrace spacing at these and other scattered
sites was regular and consistent, so tentative correlations
have been made between the scattered remnants (fig. 17).
The gradients of these correlated surfaces roughly parallel
the better constrained terrace surfaces in The Island but
are somewhat steeper than the modern Smith River and
low terraces along the same reach. Most terraces are not
correlated downstream from the North Fork area because
almost no terraces are preserved in a 15-km-long gap
between North Fork and Brainard Creek (fig. 17).
TERRACES NEAR BRAINARD CREEK
Several terraces are preserved in a large abandoned
meander near Brainard Creek at about river km 9.5
 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

1
| 2,685 + 95 yr B.P.

i (SP87USR2)
#.
N I
>

# Cl |
|
#
Figure 22. Stratigraphic columns of terrace exposures in the abandoned meander
bend near Brainard Creek: (A) exposure of the lowest terrace at station 87US3, and
(B) exposure of the highest terrace at station 87US4. See legend in figure 3 for ex
planation of lithologic patterns and symbols.

(fig. 17). The surface of the lowest terrace in the meander is
at an elevation of about 7 m and is dissected, but enough of
this surface is preserved to show the extent of the pre
abandonment path of the Smith River. A roadcut in this
surface (fig. 22A; station 87US3, plate 2) exposed about
1.25 m of overbank silty sand overlying at least 0.75 m of
silty clay that may be estuarine sediment, quiet-water sedi
ment deposited in an oxbow lake, or lacustrine sediment
deposited behind a landslide dam. No landslide scars large
enough to dam the Smith River are evident downstream
from this exposure, so the muddy sediment was probably
deposited in an oxbow lake or by estuarine processes. An
AMS radiocarbon age of 2,685 + 95 yr B.P. was obtained
on small fragments of charcoal from the base of the sand
and the top of the underlying silt. The exposure contained
numerous decomposed roots and worm burrows that could
be sources of younger organics, but the sample was picked
carefully to remove such materials before submission. The
lowest terrace in the Brainard Creek meander is about 4 m
higher than the lowest terrace along the Smith River, which
has yielded radiocarbon ages of 123.7% of modern at 1.25
m depth and 1,825 + 65 yr B.P. at 3.6 m depth from an
auger core about 5 km downstream at station 87US15 (fig.
4C; pl. 2). The lowest Brainard Creek terrace is also about
2 m lower than an intermediate terrace near the confluence
at Reedsport that yielded a TL age of 14.9 + 3.1 ka. Thus,
the radiocarbon age from the lowest terrace in the Brainard
Creek meander is bracketed in correct stratigraphic order
by radiocarbon and TL ages from nearby terrace sediments
and may be a reasonable maximum age for the age of
meander abandonment.
An eroded exposure of fluvial sediment in the highest
terrace near Brainard Creek was visible in a driveway
roadcut (fig. 22B, station 87US4, plate 2). Only the
lower 1 m of this exposure appeared to be in place, and
the sediment was extremely weathered (all gravel clasts
crushable in hand specimen, extensive oxidation and clay
alteration, 5YR hues) but fluvial bedding was still appar
ent in the sediment. Although the age of this terrace is
unknown, the height (61 m), extensive weathering of the
terrace sediments, and relations to marine terraces at the
mouth of the Umpqua River (fig. 17) indicate that this ter
race is more than 125 ka. Four other terrace surfaces are
preserved in the Brainard Creek meander between the two
terraces discussed here, but no exposures of these surfaces
were found.
TERRACES DOWNSTREAM FROM
BRAINARD CREEK
A roadcut in a 32-m-high terrace on the south side of
the Smith River at about river km 6.5 (station 87US2,
plate 2) exposed about 2 m of sandy silt and pebbly sand
(fig. 23). A TL age obtained on these sediments (55 + 12
ka) is younger than the TL age obtained on fluvial sedi
ment from the high terrace at Reedsport (65.2 + 7.3 ka),
but these ages overlap within their laboratory error limits
(table 2). The high terrace at Reedsport may correlate with
possible 125 ka marine terrace deposits near the mouth of
the Umpqua River, so if these terraces are correlative, then
both TL ages on these terraces are too young.
 TERRACES ALONG THE SMITH RIVER
C/D
C.
Lll
H.
Lll
> - -
*T*55 + 12 ka
- (SP87UST2)
# | #| 3
ti | 5.
Figure 23. Stratigraphic column of an exposure of a
high terrace on the Smith River at station 87US2. See
legend in figure 3 for explanation of lithologic patterns
and symbols.
Comparisons between projected terrace gradients and
the present gradients of the Smith and Umpqua Rivers do
not conclusively indicate whether the high terraces at
Reedsport and on the Smith River are correlative. The
present tide-influenced gradients of the lower Smith and
Umpqua Rivers are both about 0.0001; gradients upstream
from tidal influence are variable, but generally average
about 0.003 for the next 45 km of the Smith River and
about 0.0006 for the next 45 km of the Umpqua River.
The projected gradient between the high Smith River and
Reedsport terraces is 0.002, and the average gradient pro
jected between the Smith River terrace and the marine ter
race surface at Threemile Lake is about 0.0004. The
anomalously high gradient between the Smith River and
Reedsport terraces may indicate that they are not correla
tive, it may be an artifact of vertical errors in measuring
the height of one or both of these terraces, or it could indi
cate that the high Smith River terrace formed at a time
when the lower reach of the Smith River was substantially
steeper than at present, perhaps similar to conditions on
the modern nontidal reach of Scholfield Creek (fig. 11). A
steeper gradient also would be consistent with nontidal
conditions during a period of lower sea level, as indicated
by the TL ages. However, the fine sand and silt in the two
high terraces and the geomorphic setting of the Reedsport
terrace indicate that both terraces were probably formed
under estuarine conditions similar to those of today.
Thus, the heights, projected gradients, TL ages, and
relations to marine terrace deposits indicate at least three
possible scenarios for the age and correlation of the two
high terrace remnants on the lower Smith and Umpqua
Rivers. The terraces may not be correlative, but TL ages
on these sediments overlap within their laboratory error
35
limits. If the terraces are the same age, then they were
aggraded either about 5565 ka during a sea-level low
stand, or about 125 ka during a sea-level high stand. Flu
vial terraces formed during sea-level low stands near the
coast are rarely preserved in the emergent geologic record
(Lajoie, 1986), and the deep exposures of the Reedsport
terrace show no evidence of coarse sediment that should
be associated with steeper stream gradients. So while the
geologic evidence and TL ages are somewhat contradic
tory, I conclude that the Reedsport and Smith River ter
races are probably the same age, and that they correlate
with marine terrace deposits near Threemile Lake that
may be 125 ka. The variations in projected gradients
between these terrace surfaces (fig. 11) are probably
related to errors in the measurement of the heights of ter
race surfaces and the presence of an unknown thickness of
marine and eolian sediment on the marine terrace platform
at Threemile Lake. If these correlations are correct, then
TL ages determined on sediment from the fluvial terraces
are too young by 60,000 years for the Reedsport terrace
and by 70,000 years for the Smith River terrace. The TL
age discrepancies may be related to pedogenic changes in
the fluvial sediments, because at both sample locations,
extensive soil oxidation extended throughout the
exposures.
Charcoal from a fluvial terrace on Hudson Slough, a
tributary that enters the Smith River at about river km 5.5,
has been radiocarbon dated as part of a recent study of
colluvial deposits and erosion rates in the Coast Range
(Reneau, 1988). The terrace surface is about 15 m above
the bed of Hudson Slough at an elevation of 2530 m;
charcoal from about 1 m below the terrace surface at a site
3 km upstream from the Smith River yielded a radiocar
bon age of >39,500 yr B.P. (Reneau, 1988, p. 220). Sev
eral other apparently correlative terrace remnants are at
the same elevation in the Hudson Slough drainage (pl. 2),
apparently indicating that a fairly extensive terrace surface
existed in the slough sometime prior to 40 ka. This terrace
may have been similar to the modern surface of low relief
in Hudson Slough and therefore may have formed during
a sea-level high stand similar to present sea-level condi
tions. The height of this terrace indicates that it might be
correlative with TL-dated terraces on the south side of the
river just upstream (fig. 17) and at Reedsport on the
Umpqua River. All these terraces are probably correlative
with the marine terrace at Threemile Lake, which may be
related to the oxygen-isotope substage 5e (125 ka) sea
level high stand, although Reneau (1988) suggested that
erosion rates calculated in his study indicated that the
Hudson Slough terrace might be related to a Whiskey-Run
equivalent (80 ka) sea-level high stand.
A 3.75-m-deep auger core was obtained from a low
terrace about 4.5 km upstream from the mouth of the
Smith River (station 87US15, plate 2) near the mouth of
Franz Creek. The core contained several layers of peaty
36 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
and silty sediment (fig. 4C). Two radiocarbon ages on
mixtures of peat, detrital charcoal, and wood were
obtained from depths of 1.25 and 3.6 m; as previously dis
cussed, these ages (123.7% modern and 1,825 + 65 yr
B.P., respectively) yielded a maximum sediment accumu
lation rate of about 1.3 mm/yr at this site.
TERRACES ALONG THE
SIUSLAW RIVER
The Siuslaw River has a drainage basin of about
2,000 km (Oregon State Water Resources Board, 1965)
and is located 3035 km north of the Umpqua River (fig.
2). Extensive older fluvial terraces initially were recog
nized along the lower Siuslaw River by Baldwin (1956),
but Schlicker and Deacon (1974) modified Baldwin's map
ping by removing terraces mapped on the meander bend
west of Mapleton. My mapping generally confirms Bald
win's original work in this area (pl. 3). The distribution of
fluvial terraces along the lower Siuslaw River is unusual
because almost all older terraces are preserved on a high
bench on the north side of the river. The present topogra
phy of the valley is U-shaped, with a flat, sediment-filled
floor and steep sides (fig. 24). Because extensive high ter
races are not evident on the south side of the river, the
river appears to have migrated unidirectionally to the
south in Quaternary time. Another unusual feature of the
Siuslaw terraces is the apparent anticlinal warping of older
terraces near river km 25 (fig. 25). This warping coincides
with an anticline in the underlying bedrock (Baldwin,
1956; Schlicker and Deacon, 1974). The anomalously high
gradients of the terraces on the downstream limb of the
anticline were attributed by Schlicker and Deacon (1974)
and Adams (1984) to folding, but neither noted the pres
ence of multiple terraces or the terrace gradient reversals
on the upstream limb of the anticline. The multiple ter
races present above the Siuslaw River anticline appear to
be examples of tectonic terraces formed as a result of
repeated movements on the anticline.
HOLOCENE TERRACE
A single low terrace is found along much of the
lower Siuslaw River at about 210 m above river level; a
radiocarbon determination on charcoal from an exposure
in this terrace near river km 38 (fig. 26; station 88S1,
plate 3) yielded an age of 7,010 + 90 yr B.P. The radio
carbon sample was taken from near the base of sandy
overbank sediment that overlies about a meter of basal
sandy gravel. The small exposure of overbank sediment
was covered at this location by slumped sands that may be
inset into the older terrace surface. At the exposure site,
the terrace surface is at an elevation of 8.5 m and is about
6.5 m above river level. The age and height of this terrace
are similar to those of the PHT terrace on the Umpqua
River, which yielded slightly older radiocarbon ages from
similar positions in the silty overbank deposits. The
younger apparent age of the Siuslaw terrace may be
related to the location of the present tide limit, which is
only 2 km downstream from the sample site. Terrace
aggradation in such locations may have been more
affected by changes in sea level than terraces located far
ther upstream. The Holocene terrace on the Siuslaw River
converges with the river downstream and eventually
merges with tidal marshes near the coast.
In an earlier study, Feiereisen (1981) conducted an
investigation of the geomorphology, sedimentology, and
stratigraphy of the modern flood plain along the lower
Siuslaw River, which included most of the Holocene ter
race described above. He identified cyclical patterns of
fluvial sedimentation and attributed these patterns to forest
fires and large floods in the watershed. Feiereisen obtained
a single radiocarbon age of 3,250 + 70 yr B.P. on peat and
wood from a depth of about 5.5 m in a core in the flood
plain sediments near river km. 23.5.
OLDER TERRACES
Several exposures of older, higher terraces on the
north side of the Siuslaw River revealed extensively
weathered fluvial sequences that were commonly partially
stripped and overlain by colluvium. Because most of the
exposures were clustered in a group of high surfaces, only
subtle differences in weathering characteristics were evi
dent. The most apparent differences in weathering charac
teristics were degree of reddening, oxidation, and
preservation of gravel clasts. I did not obtain any numeri
cal age information from the older Siuslaw River fluvial
terraces, so correlation with marine terraces at the coast
may be the best means of estimating fluvial terrace ages.
A discussion of marine terraces present near the mouth of
the river will precede discussion of the older fluvial
terraCeS.
MARINE TERRACES
Marine terrace deposits are present at several eleva
tions near the mouth of the Siuslaw River. Golder Associ
ates (1986, table 6) listed an elevation of 18.2 m for a
Whiskey Run-equivalent terrace (80 ka) and an elevation
of 48.5 m for a Pioneer(?)-equivalent terrace (105 ka) at
Florence. Golder Associates (1986) also listed an elevation
of 30.3 m for a Pioneer-equivalent terrace at Sutton Lake,
about 10 km north of the Siuslaw River. This terrace may
extend southward to Florence as a relatively flat, sand
 SOUTH NORTH
250

200
150
100
Siuslaw
River
-

OO 5 i *

H-T
-T T T
T- T I-T -T
O 500 1500 2000 2500 3000
(m)
DISTANCE

Topographic
24.
Figure
profile
1:24,000-scale
(from
map)
topographic
River
Siuslaw
the
across
km.
river
near
terraces
fluvial
22.5;
arrows.
with
marked
are
Note
flat-floored
U-shaped,
preservation
and
valley
older
of
only
terraces
fluvial
side
north
the
on
river.

3
 3.

160

|
measurement
Terrace
S|USLAW
RIVER 140

A
Radiocarbon
age
A
range
Tide
T 120

MT
terrace
Marine
<HD-
Anticline 100

|80

|60
MT terrace
PHT

MT

MT

DISTANCE
(km)
COAST
FROM
Longitudinal
25.
Figure
along
profile
Siuslaw
reach
lower
the
5for
figure
see
River;
explanation
Note
symbols.
terraces
higher
of
warping
anticline;
an
over
indicates
symbol
arrow
double
axis
fold
of
location
bedrock.
underlying
the
in
 TERRACES ALONG THE SIUSLAW RIVER
7,010 + 90 yr B.P.
; * S;
223
;;
: o 3.3%d: dG3.
# 0
N
*> z
:
,23;:$:8.
W
S <-->
Figure 26. Stratigraphic column of an exposure of a
possible PHT (Pleistocene-Holocene transition) terrace
at station 88Sl on the Siuslaw River. See legend in fig
ure 3 for explanation of lithologic patterns and symbols.
covered surface at 3035 m elevation. Although marine
terrace deposits are not shown on recent geologic maps
(Schlicker and Deacon, 1974), deposits are present near
the elevations of Golder Associates (1986) in roadcuts
south of Florence. Recent soils work in the region (H.M.
Kelsey, written commun., 1992) indicates that an age of
125 ka is most reasonable for the 18.2 m terrace, so the
higher terraces probably do not correlate with the 105 ka
Pioneer terrace, but they are related to stage 7 (230 ka) or
older sea-level high stands. Older and higher marine ter
race sediments are also present northeast and east of Woa
hink Lake, a few kilometers southeast of Florence. The
heights of the three marine terraces of Golder Associates
(1986) have been plotted in figure 25 for comparison with
fluvial terraces on the Siuslaw River.
OLDER FLUVIAL TERRACES
Several exposures of fluvial sediment were visible in
logging roadcuts at an elevation of about 109 m about 0.5
km east of Hanson Creek near river km. 22.5 (station
87S2, plate 3). A composite stratigraphic column of two
exposures shows about 3 m of sandy silt overlying about 2
m of sandy gravel (fig. 27A). Strong oxidation extends
through the entire profile, soil hues reach 2.5YR in the
upper part of the B horizon, and 30- to 40-cm-deep solu
tion pits are developed in the saprolitic bedrock strath sur
face (fig. 28). All sandstone gravel clasts are easily
crushable in hand specimen, but a few clasts of quartzite
are still relatively unweathered.
39
A similar sequence, with an upper unit of silt, was
revealed in a not as well exposed section of fluvial sedi
ment in a terrace a few kilometers upstream, about 1 km
west-northwest of Beck Station near river km. 24.5 (fig.
27B; station 87S3, plate 3). Soil development, colors,
and depth and degree of oxidation are all similar to that
seen at station 87S2, but this terrace is about 6 m higher
in elevation (115 m) and probably correlates with the
highest terrace surface on the downstream limb of the
anticline.
A lower terrace at about 88 m elevation was exposed
in a roadcut about 1 km northwest of Wendson near river
km 16.5 (fig. 27C; station 87S4, plate 3). The fluvial sed
iments exposed in this roadcut are similar to the sediments
described at stations 87S2 and 87S3, but soil develop
ment is not as extensive as that seen in the higher terraces.
Strong oxidation is present throughout the profile, but soil
reddening only reaches a maximum of 5YR hues in the B
horizon. Most gravel clasts are crushable in hand speci
men but solution pits are not evident in the underlying
bedrock strath surface.
A much lower terrace at about 35 m elevation was
exposed in a roadcut about 1.5 km south of Mapleton near
river km 32 (fig. 27D; station 87S5, plate 3). This terrace
is the lowest in an extensive flight of high terraces pre
served on the inside of a major bend in the Siuslaw River.
The exposure revealed a basal channel facies of sandy
pebble gravel overlain by an overbank facies of pebbly
fine sand and sandy silt. The soil in this exposure is less
well developed than on the higher terraces; soil reddening
reaches 5YR hues in the upper part of the B horizon, but
the sediment has 10YR hues just below the B horizon.
Oxidation extends throughout the profile, but clasts in the
basal gravel are not crushable in hand specimen.
The ages of the older fluvial terraces are unknown.
Adams (1984) recognized a single deformed terrace on the
lower Siuslaw River that he correlated with a hypothetical
100 ka marine terrace at the coast. My studies have delin
eated multiple older terraces, rather than a single terrace,
along the same reach of the Siuslaw River (fig. 25; com
pare with Adams, 1984, fig. 8). The exact relations between
the deformed fluvial terraces and marine terraces at the
coast are unknown. However, unless the terrace gradients
on the downstream limb of the anticline get steeper near the
coast, which is unlikely, these terraces cannot be reason
ably correlated with the lowest marine terrace and therefore
must predate the probable 125 ka age of the lowest marine
terrace. An age of >200 ka is consistent with the degree of
weathering and soil development of the highest fluvial ter
races and with comparisons to a TL age on the 90-m-high
Crestview terrace on the Umpqua River, 3035 km to the
south. The lowest fluvial terrace in the sequence near
Mapleton (station 87S5, plate 3) appears to be relatively
undeformed, and it is at an elevation (35 m) that could
reasonably be projected to the 18.2 m or 30 to 35-m-high
 soquus
pue sumed ooloII Jo uopeutIdxojoj ajnu u pu331 32s gs-/g uopens leuomdeWreou oeluoi qj. Jo ainsodx2 (a) pue sg uopens le uos
-puAreou peum3. Jo ainsodx (o) S-l uopuns le uopens xoog Jeou ooeluan3!!!Jo aunsodx2 (g) zsL8 uonens exl uosueHJouseo 23eun
3q e jo saunsodx elas jo uumoo nsoduloo (y):12ai^eisnis q uo sooeluom qq elaos jo sansodx Jo suunoo ojderpenslz aun?!
Gravel
Gravel

ano

| 2
Mud
Mud

|
O
I
~
FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

* I j .ts

~

CAN

V

()
o
s
6% s o W
&2 Yu S

S,
o .

28.
A
SS

SHE LEW
YN.

SEELEW
,
SV

W
o
-

:
-

SHE LEW
SHELEW
 TERRACES ALONG THE SIUSLAW RIVER
L
-
- -
----:
Figure 28. Photograph of basal fluvial gravels overlying sandstone bedrock in a high terrace east of Hanson Creek at station 87S2; see
fig. 27A. Note probable solution pits in bedrock (arrows) and weathering of fluvial clasts. Pocket knife handle at right center (circled) is
10 cm long.
marine terraces near Florence. If one of these correlations is
accurate, then this fluvial terrace is 125 ka or older.
I do not have enough age information to confirm a
relationship between the multiple high fluvial terraces on
the Siuslaw River and sea-level high stands, but the pres
ervation of at least six older fluvial terraces near Mapleton
indicates that some of these individual terraces may have
formed as a result of episodic tectonic uplift rather than
climatic perturbations and accompanying fluctuations in
sea level.
TERRACES ALONG THE SILETZ RIVER
The lower reaches of the Umpqua, Smith, and Siu
slaw Rivers are located within about 30 km of each other
and flow through similar bedrock. The Siletz River has a
slightly smaller drainage area than the Smith River (800
km2; Oregon State Water Resources Board, 1965) is
located about 80 km north of the Siuslaw River (fig. 2),
and flows through more heterogeneous bedrock than the
other Coast Range rivers examined in this study. Fluvial
terraces and bedrock geology have been mapped along the
41
lower reach of the Siletz River by Schlicker and others
(1973) and Snavely and others (1976a, 1976b); Niem
(1976) discussed additional information on the geologic
history and hydrology of the Siletz River drainage basin.
The lower 86 km of the Siletz River flows through Ter
tiary sandstone and siltstone of the Yaquina, Alsea, Nes
tucca, Yamhill, and Tyee Formations. These rocks are
perhaps a bit finer grained and less massively bedded but
are otherwise similar to the bedrock of the Tyee Forma
tion underlying the Siuslaw, Smith, and Umpqua Rivers.
Above river km 86 (pl. 4), however, the Siletz River
is flowing on basalts and volcaniclastic sedimentary rocks
of the Eocene Siletz River Volcanics, which have been
intruded in places by dikes and sills of granophyric gabbro
(Schlicker and others, 1973; Snavely and others, 1976b).
These rocks appear to be considerably more resistant to
fluvial erosion than the sedimentary rocks downstream.
Extensive fluvial terraces are not present along the river
where it flows on the volcanic rocks, and the contact
between the Siletz River Volcanics and the sedimentary
rocks downstream coincides with the mouth of a narrow
gorge (fig. 29; labelled Lower Gorge" on plate 4).


250 200 O 100

5 O
1 5
- - 100

1300
+
|-fl1,600
B.P.
yr
B.P.
yr
>36,000
90
80

Loxnpglor
29.
fFigure
5
figure
see
River;
Siletz
reach
lower
the
along
terraces
eprofile
symbols.
of iatnuadtinoanl
B.P.
y
+440
9,695 70

(km)
COAST
FROM
DISTANCE
I 60

RIVER
SILETZ
50

measurement
Terrace

Radiocarbon
age
terrace
Marine

range
Tide

| TR MT
A A
 TERRACES ALONG THE SIUSLAW RIVER
Terraces are rare above the mouth of the gorge because
the basalts of the Siletz River Volcanics and younger
intrusive gabbroic rocks are much more resistant to the
lateral stream incision that is necessary for formation of
extensive fluvial terraces. The gradient of the Siletz River
steepens above the mouth of the gorge (from 0.003 to
0.008), so vertical stream incision has also been inhibited
by the more resistant bedrock of the Siletz River Volca
nics. These changes in stream incision are classic exam
ples of bedrock control on stream behavior.
Holocene terraces are well preserved along the Siletz
from near the mouth of Lower Gorge to the coast (pl. 4).
The lowest continuous terrace has yielded two early
Holocene radiocarbon ages (fig. 29) and is equivalent to
the PHT terrace on the Umpqua River. Late Pleistocene
and older terraces are less well preserved, but they are
present at various elevations downstream from Lower
Gorge. Prominent terrace exposures are discussed below.
LATEST PLEISTOCENE AND
HOLOCENE TERRACES
The low, nearly continuous PHT terrace is 2.515 m
above river level and has been sampled for radiocarbon
dating at two sites. A cut in a large gravel pit a few hun
dred meters northwest of Logsden (fig. 30A; station 88
SZ6, plate 4) near river km 78.5 exposed a minimum of 2
m of sandy silt overlying 0.25 m of medium to coarse
sand and 1.75 m of sandy gravel. An AMS radiocarbon
determination on small fragments of charcoal from near
the base of the sandy silt yielded an age of 9,030 + 110 yr
B.P. A second exposure of the PHT terrace was examined
in a natural stream cut near river km. 60, northwest of the
town of Siletz (fig. 30B, station 89SZ2, plate 4). This
stream cut exposed an overbank sequence of fine to
medium sand with thin stringers of sandy silt, overlying
channel facies sandy pebble and cobble gravel; charcoal
from the lower part of the overbank sands also yielded an
early Holocene age (9,695 + 440 yr B.P.). The similar
early Holocene ages support the physical correlation of the
terrace surface between these sites (fig. 29) and also sup
ports the correlation of this terrace to the PHT terrace on
the Umpqua River. The presence of PHT-equivalent ter
races on the Umpqua, Siletz, and probably the Smith and
Siuslaw Rivers strongly implies a regional, probably cli
matic cause for stream aggradation in latest Pleistocene
and early Holocene time.
Scattered younger terrace remnants are present at
lower elevations along much of the Siletz River. Most of
these remnants are on the insides of meander bends and
none form surfaces that are continuous for more than a
few hundred meters. In one low terrace at river km 84.5
near the confluence with Palmer Creek (fig. 30C, station
88SZ8, plate 4), charcoal from an exposure of pebbly,
43
sandy silt and sandy pebble and cobble gravel had a radio
carbon age of 2,230 + 60 yr B.P. Small, isolated terraces
like these probably formed as complex responses to rel
atively minor changes in river hydrology and sediment
supply, rather than to the regional, climatic changes asso
ciated with aggradation of the PHT terrace.
LATE PLEISTOCENE TERRACES
Radiocarbon analyses on charcoal collected from two
higher terrace exposures indicate that stream aggradation
also occurred in the late Pleistocene. A roadcut in a well
preserved terrace surface southwest of Siletz near river km
64 (fig. 31A; station 88SZ2, plate 4) exposed about 2 m
of overbank pebbly fine sand overlying about 0.5 m of
channel facies sandy pebble and cobble gravel. A sample
of disseminated charcoal from near the base of the over
bank sand yielded an AMS radiocarbon age of 12,110 +
95 yr B.P. This age is younger than expected because the
terrace is twice as high above the river (24 m versus 12
m) as the 10 ka PHT terrace at the same location. No
other examples of 12 ka terraces have been found in the
Coast Range, so the anomalously young age may be the
result of inclusion of small fragments of modern and
decayed roots in the radiocarbon sample. Alternatively, if
the radiocarbon age is accurate, then the height and age of
this terrace indicate that rates of stream incision were very
high in latest Pleistocene time, just prior to deposition of
the PHT terrace sediments. My dating results from other
Coast Range rivers indicate that inclusion of younger car
bon in the radiocarbon sample may be the best explanation
for the anomalously young age at this site.
A complex sequence of fluvial and colluvial sedi
ments was examined in a quarry exposure of a 40-m-high
late Pleistocene terrace near river km 88.5 in the Lower
Gorge (fig. 31B, station 88SZ7, plate 4). A typical
sequence of channel-facies sandy pebble and cobble
gravel, overlain by overbank medium sand, is present at
the base of the exposure. The sand is in turn overlain by
0.6 m of organic-rich silt and 0.3 m of sandy, pebbly silt.
The silts are unusual because such fine sediment is not
presently being deposited along this steep reach (gradient
of 0.008) of the Siletz River. The silts are overlain by 2 m
of matrix-supported, sandy, silty, colluvial gravel, which
contains abundant angular pebbles and cobbles of Tyee
Formation siltstone. The presence of Tyee Formation
clasts indicates that this gravel was not fluvially trans
ported, because upstream from the terrace exposure, the
Siletz River has incised more than 200 m below the base
of the overlying Tyee Formation into basalts and volcani
clastic rocks in the underlying Siletz River Volcanics
(Schlicker and others, 1973; Snavely and others, 1976b).
The colluvial gravel is overlain by a 1.5-m-thick sequence
of fluvial sandy pebble gravel, sandy silts, and silts.
44 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

C
Lll
H. -

Lll *::::::::::
> '': '...' O #%
j9,695 + 440 yr B.P. 2222.
... (SP89SZR1) a t
...A. A. 2,230 + 60 yr B.P.
." - HAA. (SP88SZR7)
:Fr 2,030 + 110 yr B.P.
- -

Sx. $. 't
-

*(sPasszR4 *
u
Lll
..
.

CD 1
> 3.3% s
.< 5.3's 3: 23 H.
LL

>: : >

& :

& S g
S ::
: oos : oo s,

2 ' - 0
Y 0


: P:
N
>

Q0
)
G)
-
; :

Q
-
;

: Q
-

Figure 30. Stratigraphic columns of exposures of Holocene terraces on Siletz River: (A) exposure of a PHT (Pleistocene-Holocene tran
sition) terrace in gravel pit near Logsden at station 88SZ6, (B) PHT terrace exposure northwest of Silctz at station 89SZ2, and (C) expo
sure of low terrace near the confluence of Palmer Creek at station 88SZ8. See legend in figure 3 for explanation of lithologic patterns and
symbols.

The complex stratigraphy exposed at this site appears topography in the floor of the drainage in which the site is
to be the result of one or more episodes of landsliding and located (pl. 4) indicate that the colluvial gravel probably
possible damming of the Siletz River in the Lower Gorge. was deposited as a debris flow. It is not clear if this debris
The bowl-shaped ridge at the head and hummocky flow dammed the Siletz River, but the colluvium was
 TERRACES ALONG THE SIUSLAW RIVER

; 3-A* 41,600 +1300 y B.P.

poa (SP88SZR5)

(SP88SZR6)

...O.
1 - S 3-X. o
...S:
*12,110 + 95 yr B.P.
-

# | (SP88SZR2)
SS o
3.32 2.
3
-d S3Ss o
$32,
:
:
o
3.
N
0
: o

N I I
N I
O)
I

# | 3 | | # y

| a
# a.
| a
Figure 31. Stratigraphic columns of exposures of late Pleistocene terraces on Siletz River: (A) terrace
exposure southwest of Siletz at station 88SZ2, and (B) terrace exposure in rock quarry in thc Lower
Gorge at station 88SZ7. Sce legend in figure 3 for explanation of lithologic patterns and symbols.
46 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
subsequently buried by fluvial sediment before the terrace
surface was abandoned. The fine silty sediment that
underlies the colluvium may have been deposited as a
result of an earlier period of landsliding that dammed the
river downstream from the quarry site. The sediment that
formed this dam could have originated as debris flows
either from colluvium-filled hollows located on both sides
of the lower reach of Lower Gorge, or from two small
southward-flowing streams that enter the river a few kilo
meters downstream from the terrace site. The presence of
fluvial sediments at the top of the exposed sequence indi
cates that any landslide dams formed during the aggrada
tion of the quarry terrace were eventually breached and
fluvial deposition was re-established before the terrace
surface was finally abandoned. Such temporary landslide
damming has probably been a common process in the nar
row confines of the Lower Gorge.
Radiocarbon samples were taken from 1- to 2-cm
thick mats of mixed detrital charcoal and wood from the
organic-rich silt unit at the Lower Gorge site (fig. 31B).
The charcoal and wood fragments were separated in the
field and submitted to different radiocarbon laboratories;
the charcoal fraction yielded an age of >36,000 yr B.P.,
and the wood fraction yielded an age of 41,600 + 1,300 yr
B.P. (table 1). Although the wood fraction yielded a finite
age, the presence of decayed roots in the surrounding sedi
ment could have caused contamination of the sample and
as a result yields a minimum age for the terrace sediments.
OLDER TERRACES
Numerous older, higher terrace remnants are pre
served along the Siletz River (fig. 29) but as with other
Coast Range rivers, most remnants are widely scattered so
terrace correlations are tenuous. No numerical age data are
available from any of the older terraces, but some age
information can be inferred from comparisons between the
heights of these older surfaces and the heights of radiocar
bon-dated terraces and from relations to marine terraces at
the coast. A discussion of marine terraces present near the
mouth of the Siletz River will precede discussion of older
fluvial terrace exposures.
MARINE TERRACES
Correlation of Siletz River terraces with the marine
terrace record is very tenuous because few fluvial and
marine terraces are preserved near the coast (fig. 29).
Golder Associates (1986) listed an elevation of 24.2 m for
a Whiskey Run-equivalent marine terrace at Glenenden
Beach, a few kilometers south of the mouth of the Siletz
River. My reconnaissance in the Lincoln City and
Glenenden Beach areas (pl. 4) confirmed that marine ter
race remnants are present at 2428 m in this area.
Kennedy (1978) used amino-acid ratios on shells to estab
lish an oxygen-isotope substage 5a (80 ka) age on a ter
race of similar height at Yaquina Bay, 30 km south of the
Siletz River, but recent soils work (H.M. Kelsey, written
commun., 1992) indicates that several major structures are
present between Yaquina Bay and the Siletz River, and
that the lowest marine terrace at Glenenden Beach is most
reasonably correlated to the substage 5e (125 ka) sea-level
high stand. Marine terrace deposits are also present at
3640 m in Lincoln City and just south of the mouth of
the Siletz River in Glenenden Beach. These deposits are
probably related to stage 7 (230 ka) or older sea-level high
stands. Some scattered fluvial terraces near the coast may
correlate with these marine deposits, but the preservation
of both types of terraces is so poor that such correlations
are very tentative.
OLDER FLUVIAL TERRACES
Older fluvial terraces are well developed in two loca
tions downstream from Lower Gorge, but poor access and
a lack of good exposures hampered examination of most
older fluvial terraces along the Siletz River. The best pre
served terrace flight is located west of Logsden (river km
7080) and has terrace remnants from 40 m to over 100 m
above present river level. The longest terrace remnants
have surface gradients that are somewhat less steep than
lower terraces and the modern river in this reach (fig. 29).
The lower gradients may be an artifact of the methods
used to measure terrace heights, or they could signify reju
venation and steepening of the Siletz River in late Pleis
tocene time. Unfortunately, the poor preservation of older
terraces elsewhere along most of the river prevents terrace
correlation over enough distance to determine whether
rejuvenation has taken place. No good exposures of these
sediments was observed.
Older terraces are also fairly well developed near
Siletz. A deeply weathered exposure of fluvial sediment in
one of these terraces was examined in a logging roadcut
west of Siletz at about river km 63 (fig. 32A; station 88
SZ3, plate 4). The dissected terrace surface is at an eleva
tion of about 93 m and has been covered by colluvium.
Examination of the sediment exposed in the roadcut
revealed about 1 m of sandy, pebbly colluvium overlying
an overbank sequence of pebbly, sandy silt; a well pre
served buried A horizon present in the upper 0.15 m of the
silt probably marks the original terrace surface. The lower
part of the sandy silt and about 1.7 m of clayey silt and
silt were observed in a 3-m-deep core at this site. I was
unable to penetrate the entire thickness of the fluvial sedi
ment with the corer, so the character of the lower part of
 TERRACES ALONG THE SIUSLAW RIVER 47

# # ||
Figure 32. Examples of stratigraphic columns of cxposures of older terraces on the
Siletz River: (A) exposure of high terrace west of Siletz at station 88SZ3; lower 3 m of
section from an auger core, and (B) terrace exposure near Melco Landing at station
88SZ9. Sce legend in figure 3 for explanation of lithologic patterns and symbols.
48 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
12430'
4315'
. .
O 5 km
4300'
Figure 33. Map of lower Coquille River showing locations of the towns of Bandon, Coquille and Myrtle Point, and the conlucnces of the
North, Middle, and South Forks of the Coquille River; shaded area denotes uplands.
the terrace sequence is unknown. The fluvial sediments
underlying the colluvium are oxidized to 5YR hues in the
B horizon and sedimentary and volcanic gravel clasts in
the sandy silt are crushable in hand specimen.
A few very high terrace remnants are present at eleva
tions of 150175 m above river level near the mouth of
Lower Gorge (pl. 4). These terraces have straths formed in
the Tyee Formation, so they predate incision of the river into
the more resistant Siletz River Volcanics and subsequent
formation of the gorge. No good exposures of these older
terrace sediments were observed.
Most older terraces are poorly preserved along the
lower reach of the Siletz River, but two terrace surfaces are
fairly well preserved at elevations of 2325.5 m and 33.535
m near Melco Landing (pl. 4). A landslide scar just west of
Melco Landing near river km 17 (fig. 32B, station 88SZ9,
plate 4) exposed a sequence of fluvial sediments consisting
of 4.5 m of sandy pebble gravel interbedded with thin string
ers of sandy silts, which is overlain by 1 m of silty sand and
1.5 m of fine to medium sand. The terrace sediments are
oxidized to 7.5YR hues and the gravel clasts are not
crushable in hand specimen. Few other exposures of older
12400'
2T
f ITV. ATI
N.1L, -
H Coquille
sl Il
YE
y
terrace sediments were observed along the lower reach of
the river.
TERRACES ALONG THE
COQUILLE RIVER
The Coquille River is located about 60 km south of
the Umpqua River (fig. 2) and enters the Pacific Ocean at
Bandon in an area that has recently been the subject of
several marine terrace studies (Kelsey, 1990; McInelly and
Kelsey, 1990; Muhs and others, 1990). My initial work in
the area consisted of a search for fluvial terraces along the
lower 60 km of the Coquille River, downstream from the
town of Myrtle Point and the confluence of the North and
South Forks of the river (fig. 33). Unfortunately, fluvial
terrace sediments are present in only a few places along
the lower Coquille (Beaulieu and Hughes, 1975), and I
found no evidence of multiple Holocene terraces down
stream from Myrtle Point. Because of the dearth of pre
served terraces, I examined the extensive flood plain in the
vicinity of the town of Coquille (fig. 34), where the flood
 TERRACES ALONG THE COQUILLE RIVER 49

12415'00" 12410'00"
4312'30"

&
&
C
2Y

Station
89- C3

Station
89-C4

Abandoned
Nva channel

O 1 km

4307'30"

Figure 34. Map of the flood plain of the Coquille River near Coquille; shaded areas denote uplands.
Note presence of two extensive abandoned channels on the floodplain southwest of Coquille and locations
of two auger cores (stations 89C3 and 89C4) shown in figure 35 and discussed in the text.

plain is as much as 3 km wide and presently used as pas
ture land. Prior to drainage and diking associated with
agricultural development, the flood plain had extensive
areas of freshwater marsh and was frequently covered
with ponded flood waters. The flood plain and the sur
rounding towns are still flooded during periods of
extended rainfall in the winter months (Beaulieu and
Hughes, 1975).
ancestral Coquille River (fig. 34). These channels are clearly
visible in air photos and are expressed in the field as 25- to
40-m-wide linear depressions that lie less than a meter below
the general elevation of the flood plain and are flanked by
50- to 100-m-wide levees. The southernmost abandoned
river channel extends continuously across the Coquille Val
ley for about 10 km, and represents a major prehistoric shift
in the course of the river. A second abandoned river channel

One of the most interesting features on the flood plain is preserved a couple of hundred meters north of the more
are extensive, well preserved abandoned channels of the extensive abandoned channel. The northern channel
50 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON

A B

3-E

?
LLI
2+--- |- - -

H. *- -

LL - -
| -

H.--
>

;
1 TT

Fl2,860 + 155 yr B.P.

AA (SP89CR4)
0 : , | G) 12,835 + 140 yr B.P.
| # | # (SP89CR7)

#
Figure 35. Stratigraphic columns of two auger cores in abandoned channels on the Coquille River flood
plain: (A) core in southern abandoned channel at station 89C4, and (B) core in northern abandoned chan
nel at station 89C3. Stations are located in figure 34. See legend in figure 3 for explanation of lithologic
patterns and symbols.

apparently was not occupied as long because its central
channel is shallower and it has low levees that have been
breached by numerous small tidal channels.
No exposures of flood plain or channel sediments
were found near Coquille, so I obtained several 4- to 5-m
deep auger cores in both abandoned channels. The simple
stratigraphy found in these cores is illustrated by examples
from each of the channels (stations 89C3 and 89C4, fig.
34). The two cores contained sandy channel deposits that
were overlain by more than 3 m of overbank silt (fig. 35).
Radiocarbon analyses on detrital charcoal and wood from
the upper parts of the channel sands from the two cores
yielded nearly identical ages of 2,860 + 155 yr B.P. and
2,835 + 140 yr B.P.
These radiocarbon ages and the geomorphology of
the Coquille River flood plain can be used to estimate the
relative position of late Holocene sea level. The present
low-water (summer) elevation of the surface of the
nearby Coquille River is less than 1 m above sea level,
and the elevation of the top of the channel deposits in the
two cores (about 2 m below sea level) is about the same
as the present elevation of the channel bottom of the
modern river in the vicinity of Coquille. This relation
implies that little if any change in relative sea level has
occurred in the last 3,000 years in this area. The preser
vation of numerous abandoned tidal and river channels
on the present flood plain and lack of multiple fluvial ter
races also indicate stable relative sea-level conditions.
 DISCUSSION
DISCUSSION
ORIGIN OF LATEST
PLEISTOCENE/EARLY HOLOCENE
AGGRADATION
Radiocarbon ages and the distribution of Pleistocene
Holocene transition (PHT) terraces are remarkably similar
along the Umpqua, Siletz, and probably the Smith and
Siuslaw Rivers. In addition, Grabau (1990) identified what
appears to be a PHT equivalent, 10 ka terrace on Drift
Creek, a tributary of the Alsea River located about equi
distant between the Siuslaw and Siletz Rivers. Under
standing the genesis of the PHT terraces is important
because they are the best preserved terraces in the Coast
Range and may be analogs for some of the older fluvial
terraces in the region. The consistent ages and regional
distribution of PHT terrace sediments indicate a linkage
between the formation of extensive fluvial terraces and
some regional control, such as climatic change.
Changes in stream behavior associated with forma
tion of PHT terraces most likely include rapidly rising
base (sea) level associated with late Wisconsin deglacia
tion and/or increased sediment load in streams. Tidal
reaches of Coast Range streams were certainly affected by
base-level changes, but rises in base level may not have
strongly affected fluvial systems very far inland from the
coast (McDowell, 1987, p. 549). This conclusion is sup
ported by studies of river behavior upstream from recently
flooded reservoirs (Leopold and others, 1964; Leopold and
Bull, 1979), by studies of fluvial response to late Quater
nary climate and base-level change in the Texas Gulf
Coast (Blum, 1990), and by the lack of radiocarbon-dated
terraces in the Coast Range associated with the steep sea
level rise that accompanied worldwide late Wisconsin
deglaciation. A recent modelling study of post-Wisconsin
sea-level changes (Tushingham and Peltier, 1991) indi
cates that about 40 m of sea-level rise occurred between
the late Wisconsin minimum (18 ka) and 11 ka, but no
well-documented examples of extensive fluvial terraces of
this age have been found in the Coast Range. I conclude
that effects of base-level rise were restricted to the lower
(presently tidal) reaches of Coast Range streams, and had
no significant impact on aggradation of the upstream
reaches.
Thus, the aggradation of PHT terrace sediments
appears to be driven by increased sediment inputs into
Coast Range streams. Late Pleistocene stream aggradation
is commonly attributed to sediment inputs from glaciers,
but only one river in this study (the Umpqua) had late
Pleistocene glaciers in its headwaters, and I observed no
sedimentologic differences between terrace sediments on
the Umpqua and other Coast Range rivers. Therefore, the
PHT terraces in the Coast Range are not aggradational
outwash terraces similar to those commonly found down
51
stream from glaciated headwaters in the Western United
States (for example, Moss, 1974; Dethier, 1988) but rather
are the result of stream aggradation related to the input of
large sediment loads from other sources. As suggested by
Grabau (1990), the most likely source of this sediment is
increased frequency of mass movements in the Coast
Range.
The most important source of sediment input into
modern Coast Range streams is mass movements (usually
debris flows) that originate from colluvium-filled hollows
(Pierson, 1977; Dietrich and Dunne, 1978; Benda and
Dunne, 1987; Benda, 1990), but the upper reaches of most
modern Coast Range streams are not rapidly aggrading, so
present rates of sediment input into streams are not suffi
cient to deposit extensive alluvial sediments. Formation of
the PHT and other extensive terraces must have required
much higher rates of sediment input, and the most likely
source of this sediment is increased debris-flow activity
originating from colluvium-filled hollows. Studies of the
ages of basal colluvial deposits from hollows in Washing
ton suggest a period of accelerated evacuation of collu
vium in the early Holocene (Reneau and others, 1989;
Reneau and Dietrich, 1990), and studies from central and
northern California suggest a similar period during the
Pleistocene/Holocene transition (Marron, 1985; Reneau
and others, 1986, 1990; Rypins and others, 1989). The
occurrence of accelerated mass movements in other parts
of the Pacific Northwest at the same time as initial deposi
tion of PHT terrace sediments in the Coast Range points
to climatically induced mass movements as the most likely
source of sediment in these terraces. Colluvium-filled hol
lows are widely dispersed along all drainages in the Coast
Range, so climatically controlled, accelerated evacuation
of these hollows could have provided the regionally dis
tributed sediment supply needed for extensive alluvial
aggradation at the Pleistocene/Holocene transition.
Unfortunately, a recent study of hollows in the Coast
Range (Reneau, 1988; Reneau and Dietrich, 1990, 1991)
did not identify an early Holocene period of landsliding
but rather found basal colluvial ages clustering in the mid
dle Holocene (47.5 ka). However, at least one sample
from this study (Reneau, 1988) and one sample from
another study in the Siuslaw River drainage (Benda and
Dunne, 1987) yielded early Holocene and latest Pleis
tocene basal colluvial ages. I interpret these latter ages,
and the radiocarbon age of about 10 ka from the Buck
Creek alluvial/debris fan on the Smith River, as evidence
of an earlier period of colluvial evacuation and debris fan
formation that produced the large volume of sediment now
present in the PHT terraces. The relative dearth of middle
Holocene fluvial terraces in the Coast Range indicates that
the PHT episode must have evacuated significantly
greater volumes of colluvium than the middle Holocene
event postulated by Reneau (1988). The relative lack of
basal colluvial ages of latest Pleistocene and early
52 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
Holocene age in the Coast Range is probably related to the
small number of radiocarbon ages on basal colluvial
deposits in the region and to the removal of older colu
vium during subsequent landsliding episodes.
The exact causes of the postulated periods of acceler
ated landsliding are not known. In historic time, periods of
accelerated landsliding have usually been caused by short
term, localized perturbations, such as large forest fires,
and by periods of flooding related to intense precipitation
and rapid snowmelt. The late Pleistocene/early Holocene
periods of accelerated landsliding in California and Wash
ington have generally been attributed to longer-term late
Quaternary climatic changes, such as increased storm
intensities in latest Pleistocene time and/or vegetation
changes and increased frequency of forest fires induced by
warmer, drier early Holocene climates (Reneau and others,
1986, 1989, 1990; Reneau, 1988; Rypins and others, 1989;
Reneau and Dietrich, 1990). Such processes, if active for
many hundreds to thousands of years, probably are
responsible for the large inputs of sediment into Coast
Range streams near the Pleistocene/Holocene transition.
RELATIONS BETWEEN THE PHT AND
OTHER COAST RANGE TERRACES
The PHT terraces are excellent examples of
climatic terraces and can be viewed as analogs for some
of the older terraces in the Coast Range, but most isolated
older terrace remnants probably were formed by processes
other than aggradation induced by regional climate and
never formed extensive synchronous terrace surfaces.
Those that did form extensive surfaces probably shared
several characteristics with the PHT terraces: the older
climatic terraces originally formed fairly extensive, nearly
continuous surfaces that converged and correlated with
marine terrace sediments (proxies for high sea level) near
the coast, and the terraces consisted of sediments whose
age ranges may have spanned 10,000 years or more.
Most older terraces are presently very poorly pre
served in the Coast Range, but I have made some very
tentative correlations between some older fluvial terraces
and marine terrace deposits near the coast. The connection
between eustatic sea-level changes and fluvial terraces
near the coast is geologically reasonable, because with the
exception of a single terrace exposure on the Umpqua
River, older fluvial terraces near the coast mostly consist
of fine-grained sediments deposited under fluvial and estu
arine conditions similar to those affecting the lower
reaches of modern rivers. Examples of probable older cli
matic terraces include the two extensive older terraces
along Henderer Road on the Umpqua River (river kms
6875), and the extensive high terraces just downstream
from Logsden at about river km 75 on the Siletz River.
Most other Coast Range terraces of all ages are prob
ably either slip-off terraces, such as the numerous terraces
present on the insides of meander bends along the Smith
River, or formed as complex responses to major floods
or other short-duration sedimentation events. Examples of
the latter are the discontinuous terraces below the PHT
terrace near Elkton and at river km 123133 on the
Umpqua River, and terraces below the PHT terrace at
river kms 4660 and 7884 on the Siletz River. Many
(most?) isolated older terrace remnants on all Coast Range
rivers are probably complex response terraces that never
formed continuous, synchronous surfaces.
Examples of older tectonic terraces are preserved
on the lower reach of the Siuslaw River. Although the
older terraces near Mapleton have been described by
Reneau (1988, p. 219) as slip-off terraces, the anticlinal
warping of these surfaces indicates that tectonic uplift
probably forced the abandonment of at least some terrace
surfaces. The Siuslaw River anticline plunges to the south
(Baldwin, 1956), so perhaps the terraces are localized on
the north side of the river because the river has periodi
cally slipped southward off the anticline in the direction
of fold plunge. Some of the older Siuslaw terraces proba
bly correlate with marine terraces near the coast, but uplift
has masked any climatic signal in this terrace sequence.
CONCLUSION
Fluvial terraces are mostly discontinuous and small in
the Coast Range of western Oregon, but I have made
some inferences about terrace origin based on the distribu
tion and age of terrace deposits on the Siletz, Siuslaw,
Smith, Umpqua, and Coquille Rivers. Widespread, nearly
continuous Pleistocene/Holocene transition (PHT) terraces
are 215 m above the present level of most Coast Range
rivers and began to aggrade in nontidal reaches in latest
Pleistocene and early Holocene time; radiocarbon analysis
of charcoal from terrace sediments on the Siletz, Siuslaw,
and Umpqua Rivers yielded ages of 710 ka. The ubiqui
tous distribution of the PHT terraces implies a regional,
probably climatic cause of initial terrace formation. The
most likely cause of terrace aggradation at the Pleis
tocene/Holocene transition is accelerated landsliding
associated with regional climatic change and its effects on
vegetation and sediment yield, rather than rising base level
or an influx of glacial outwash sediment. Similar periods
of accelerated colluviation in Washington and California
have been attributed to increased storm intensities in latest
Pleistocene time and vegetation changes and increased
intensity of forest fires induced by warmer, drier early
Holocene climates, but the exact cause of accelerated
landsliding in Oregon is unknown.
 ACKNOWLEDGMENTS
Most of the late Holocene, aerially restricted fluvial
terraces in the Coast Range probably formed as complex
responses to major floods or localized landsliding. These
terraces represent short-duration interruptions in what
appears to be a long term trend of stream incision since
the PHT terrace sediments were aggraded.
Older terrace remnants are scattered along the lengths
of Coast Range rivers up to 200 m or more above present
river levels, but they are less frequent as the rivers
approach the coast. Tentative correlations of these ter
races with poorly preserved marine terrace deposits indi
cate that some of the older fluvial terraces were formed
during sea-level high stands in the middle and late Pleis
toCene.
With the exception of the Smith River, I obtained
generally consistent results from radiocarbon dating of
detrital charcoal and wood from the lower terrace sedi
ments. Anomalously young ages are thought to be related
to contamination of the samples with young carbon from
the inclusion of root material or from sampling in younger
colluvial sediment. Experimental TL dating of older ter
race sediments yielded mixed results; the TL dates on
Holocene deposits generally yielded maximum ages and
most of the TL dates on older deposits appeared to yield
minimum ages. However, the TL ages were in strati
graphic order and may be useful as relative age indicators.
Problems with TL ages are poorly understood, but they
may be related to some combination of incomplete light
bleaching during deposition and to post-depositional
hydrologic or pedogenic changes. Examination of soils
proved to be less useful as a relative-dating technique
because of apparent changes in rates of soil development
inland from the coast. Apparent differences in oxidation
in soils on terraces of similar age indicate that soils which
formed under the cooler, foggier summer weather near the
coast developed more slowly than inland soils, where hot
ter, drier conditions prevail during the summer months.
ACKNOWLEDGMENTS
The author wishes to thank the numerous landowners
who allowed free access to their property in the Coast
Range, Steve McDuffy for assistance in the field, Danelle
Alloway and especially Lee-Ann Bradley for drafting
assistance, Kathy Haller and Sherry Agard for manuscript
reviews, Steve Reneau for comments on erosion rates in
the Coast Range, and Harvey Kelsey for comments on
marine terraces along the central Oregon coast. This work
was supported by the National Earthquake Hazards
Reduction Program of the U.S. Geological Survey and the
U.S. Nuclear Regulatory Commission.
53
REFERENCES CITED
Adam, D.P., 1988, Pollen zonation and proposed informal climatic
units for Clear Lake, California, cores CL734 and CL737,
in Sims, J.D., ed., Late Quaternary climate, tectonism, and
sedimentation in Clear Lake, northern California Coast
Ranges: Geological Society of America Special Paper 214,
p. 6380.
Adams, John, 1984, Active deformation of the Pacific Northwest
continental margin: Tectonics, v. 3, no. 4, p. 449472.
Alexander, C.S., 1953, The marine and stream terraces of the
Capitola-Watsonville area: University of California
Publications in Geography, v. 10, no. 1, p. 144.
Atwater, B.F., 1987, Evidence for great Holocene earthquakes
along the outer coast of Washington State: Science, v. 36,
p.942944.
1992, Geologic evidence for earthquakes during the past
2000 years along the Copalis River, southern coastal
Washington: Journal of Geophysical Research, v. 97,
p. 19011919.
Atwater, B.F., and Yamaguchi, D.K., 1991, Sudden, probably
coseismic submergence of Holocene trees and grass in coastal
Washington State: Geology, v. 19, p. 706709.
Bacon, C.R., 1983, Eruptive history of Mount Mazama and Crater
Lake caldera, Cascade Range, U.S.A.: Journal of Volcanology
and Geothermal Research, v. 18, p. 57115.
Baker, R.G., 1983, Holocene vegetation history of the western
United States, in Wright, H.E., Jr., ed., The Holocene, Vol. 2 of
Wright, H.E., Jr., ed., Late Quaternary environments of the
United States: Minneapolis, University of Minnesota Press,
p. 109127.
Baldwin, E.M., 1956, Geologic map of the lower Siuslaw River
area, Oregon: U.S. Geological Survey Oil and Gas
Investigations Series Map OM-186, scale 1:62,500.
1981, Geology of Oregon (third edition): Dubuque, Iowa,
Kendall/Hunt, 170 p.
Bard, Edouard, Hamelin, Bruno, Fairbanks, R.G., and Zindler,
Alan, 1990, Calibration of the C timescale over the past
30,000 years using mass spectrometric U-Th ages from
Barbados corals: Nature, v. 345, no. 6274, p. 405410.
Barnosky, C.W., 1985, Late Quaternary vegetation near Battle
Ground Lake, southern Puget Trough, Washington: Geological
Society of America Bulletin, v. 96, p. 263271.
Beaulieu, J.D., and Hughes, P.W., 1975, Environmental geology of
western Coos and Douglas Counties, Oregon: Oregon
Department of Geology and Mineral Industries Bullctin 87,
148 p.
Becker, Bernd, Kromer, Bernd, and Trimhorn, Peter, 1991, A
stable-isotope tree-ring timescale of the Late Glacial/Holoccne
boundary: Nature, v. 353, p. 647649.
Benda, Lee, 1990, The influcmce of debris flows on channcls and
valley floors in the Oregon Coast Range, U.S.A.: Earth Surface
Processes and Landforms, v. 15, p. 457466.
Benda, Lee, and Dunne, Thomas, 1987, Sediment routing by debris
flow, in Beschta, R.L., Blinn, T., Grant, G.E., Swanson, F.J.,
and Ice, G.G., eds., Erosion and sedimentation in the Pacific
Rim: International Association of Hydrological Sciences
Publication no. 165, p. 213223.
54 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
Berger, G.W., 1988, Dating Quaternary sediments by
luminescence, in Easterbrook, D.J., ed., Dating Quaternary
Sediments: Geological Society of America Special Paper 227,
p. 1350.
1990, Effectiveness of natural zeroing of the
thermoluminescence in sediments: Journal of Geophysical
Research, v. 95, no. B8, p. 12,37512,397.
Berger, G.W., Luternauer, J.L., and Clague, J.J., 1990, Zeroing
tests and application of thermoluminescence dating to Fraser
River Delta sediments: Canadian Journal of Earth Sciences, v.
27, p. 17371745.
Birkeland, P.W., 1972, Late Quaternary eustatic sea-level changes
along the Malibu coast, Los Angeles County, California:
Journal of Geology, v. 80, p. 432448.
1984, Soils and geomorphology: New York, Oxford
University Press, 372 p.
Blong, R.J., and Gillespie, R., 1978, Fluvially transported charcoal
gives erroneous C ages for recent deposits: Nature, v. 271,
p. 739741.
Bloom, A.L., 1983, Sea level and coastal morphology of the United
States through the late Wisconsin glacial maximum, in Porter,
S.C., ed.,The Late Pleistocene, Vol. 1 of Wright, H.E., Jr., ed.,
Late Quaternary environments of the United States:
Minneapolis, University of Minnesota Press, p. 215229.
Bloom, A.L., Broecker, W.S., Chappell, J.M.A., Matthews, R.K.,
and Mesolella, K.J., 1974, Quaternary sea level fluctuations on
a tectonic coastNew *Th/*U dates from the Huon
Peninsula, New Guinea: Quaternary Research, v. 4,
p. 185205.
Blum, M.D., 1990, Climatic and eustatic controls on Gulf Coastal
Plain fluvial sedimentationAn example from the Late
Quaternary of the Colorado River, Texas: Gulf Coast Section,
Society of Economic Paleontologists Foundation Eleventh
Annual Research Conference, Programs and Abstracts,
p. 7183.
Bockheim, J.G., Kelsey, H.M., and Marshall, J.G. III, 1992, Soil
development, relative dating, and correlation of late
Quaternary marine terraces in southwestern Oregon:
Quaternary Research, v. 37, p. 6074.
Bull, W.B., 1990, Stream-terrace genesisImplications for soil
development, in Knuepfer, P.L.K. and McFadden, L.D., eds.,
Soils and landscape evolution: Geomorphology, v. 3,
p. 351367.
Chappell, J., and Shackleton, N.J., 1986, Oxygen isotopes and sea
level: Nature, v. 324, p. 137140.
Chappell, J., and Veeh, H.H., 1978, *Th"U age support of an
interstadial sea level of -40 m at 30,000 yr B.P.: Nature, v. 276,
p. 602603.
COHMAP, 1988, Climatic changes of the last 18,000
yearsObservations and model simulations: Science, v. 241,
p. 10431052.
Colman, S.M., and Pierce, K.L., 1981, Weathering rinds on
andesitic and basaltic stones as a Quaternary age indicator,
western United States: U.S. Geological Survey Professional
Paper 1210, 56 p.
Cotton, C.A., 1940, Classification and correlation of river terraces:
Journal of Geomorphology, v. 3, p. 2737.
Curtiss, D.A., Collins, C.A., and Oster, E.A., 1984, Water resources
of western Douglas County, Oregon: U.S. Geological Survey
Water-Resources Investigations Report 834017, 81 p.
Darienzo, M.E., and Peterson, C.D., 1990, Episodic tectonic
subsidence of late Holocene salt marshes, northern Oregon
coast, central Cascadia margin, U.S.A.: Tectonics, v. 9,
p. 122.
Dethier, D.P., 1988, The soil chronosequence along the Cowlitz
River, Washington: U.S. Geological Survey Bulletin
1590Fr, p. F1F47.
Dietrich, W.E., and Dunne. Thomas, 1978, Sediment budget for a
small catchment in mountainous terrain: Zeitzschrift fr
Geomorphologie, Supplementband 29, p. 191206.
Dueholm, K.S., and Pilmore, C.L., 1989, Computer-assisted
geologic photogrammetry: Photogrammetric Engineering and
Remote Sensing, v. 55, no. 8, p. 11911196.
Feiereisen, J.J., 1981, Geomorphology, alluvial stratigraphy, and
sedimentsLower Siuslaw and Alsea River valleys, Oregon:
Eugene, University of Oregon, Ph.D. dissertation, 281 p.
Forman, S.L., Jackson, M.E., McCalpin, J., and Maat, P., 1988, The
potential of using thermoluminescence to date buried soils
developed on colluvial and fluvial sediments from Utah and
Colorado, U.S.A.Preliminary results: Quaternary Science
Reviews, v. 7, p. 287293.
Golder Associates, 1986, WNP-3 geologic support services, coastal
terrace study: Richland, Washington, unpublished report
prepared for Washington Public Power Supply System, 106 p.
Grabau, P.C., 1990, Floodplain aggradation in an intermontane area
of the Oregon Coast Range as a geomorphic response to the
Pleistocene-Holocene transition: Bellingham, Western
Washington University, M.S. thesis, 45 p.
Harden, J.W., 1987, Soils developed in granitic alluvium near
Merced, California: U.S. Geological Survey Bulletin 1590A,
65 p.
Helley, E.J., and LaMarche, V.C., Jr., 1973, Historic flood
information for northern California streams from geological
and botanical evidence: U.S. Geological Survey Professional
Paper 485E, p. E1E16.
Heusser, C.J., 1983, Vegetation history of the northwestern United
States including Alaska, in Porter, S.C., ed., The Late
Pleistocene, Vol. 1 of Wright, H.E., Jr., ed., Late Quaternary
environments of the United States: Minneapolis, University of
Minnesota Press, p. 239258.
1985, Quaternary pollen records from the Pacific Northwest
CoastAleutians to the Oregon-California boundary, in
Bryant, V.M., Jr., and Holloway, R.G., eds., Pollen records of
Late-Quaternary North American sediments: Dallas, American
Association of Stratigraphic Palynologists, p. 141165.
Huntley, D.J., 1985, On the zeroing of the thermoluminescence of
sediments: Physics and Chemistry of Minerals, v. 12,
p. 122127.
Kelsey, H.M., 1987, Geomorphic processes in the recently uplifted
coast ranges of northern California, in Graf, W.L., ed.,
Geomorphic systems of North America, part of Chapter 13,
Pacific Coast and Mountain System: Geological Society of
America Centennial Special Volume 2, p. 550560.
1990, Late Quaternary deformation of marine terraces on
the Cascadia subduction zone near Cape Blanco, Oregon:
Tectonics, v. 9, p. 9831014.
Kennedy, G.L., 1978, Pleistocene paleoecology, zoogeography and
geochronology of marine invertebrate faunas of the Pacific
Northwest coast (San Francisco to Puget Sound): Davis,
University of California, Ph.D. thesis, 824 p.
 REFERENCES CITED
Knuepfer, P.L.K., 1988, Estimating ages of late Quaternary stream
terraces from analysis of weathering rinds and soils:
Geological Society of America Bulletin, v. 100, p. 12241236.
Lajoie, K.R., 1986, Coastal tectonics, in Wallace, R.E., panel
chairman, Active Tectonics: Washington, D.C., National
Research Council, Studies in Geophysics, National Academy
Press, p. 95125.
Leopold, L.B., and Bull, W.B., 1979, Base level, aggradation, and
grade: Proceedings of the American Philosophical Society, v.
123, p. 168202.
Leopold, L.B., Wolman, M.G., and Miller, J.P., 1964, Fluvial
processes in geomorphology: San Francisco, W.H. Freeman
and Company, 522 p.
Marron, D.C., 1985, Colluvium in bedrock hollows on steep slopes,
Redwood Creek drainage basin, northwestern California, in
Jungerius, P.D., ed., Soils and Geomorphology: Catena
Supplement 6, p. 5968.
Martinson, D.C., Pisias, N.G., Hays, J.D., Imbrie, John, Moore,
T.C., Jr., and Shackleton, N.J., 1987, Age dating and the orbital
theory of the ice agesDevelopment of a high-resolution 0 to
300,000-year chronostratigraphy: Quaternary Research, v. 27,
p. 129.
McDowell, P.F., 1987, Geomorphic processes in the Pacific coast
and mountain system of Oregon and Washington, in Graf,
W.L., ed., Geomorphic systems of North America, part of
Chapter 13, Pacific Coast and Mountain System: Geological
Society of America Centennial Special Volume 2, p. 539549.
McInelly, G.W., and Kelsey, H.M., 1990, Late Quaternary
deformation in the Cape Arago-Bandon region of coastal
Oregon as deduced from wave-cut platforms: Journal of
Geophysical Research, v. 95, no. B5, p. 6699-6713.
Merritts, Dorothy, and Vincent, K.R., 1989, Geomorphic response
of coastal streams to low, intermediate, and high rates of uplift,
Mendocino triple junction region, northern California:
Geological Society of America Bulletin, v. 101, p. 1373-1388.
Moss, J.H., 1974, The relation of river terrace formation to
glaciation in the Shoshone River basin, western Wyoming, in
Coates, D.R., ed., Glacial Geomorphology: Binghamton, State
University of New York, Publications in Geomorphology,
p. 293314.
Muhs, D.R., Kelsey, H.M., Miller, G.H., Kennedy, G.L., Whelan,
J.F., and McInelly, G.W., 1990, Age estimates and uplift rates
for Late Pleistocene marine terracesSouthern Oregon
portion of the Cascadia forearc: Journal of Geophysical
Research, v.95, no. B5, p. 66856698.
Munsell Color Company, 1954, Munsell soil color chart:
Baltimore, Maryland, Munsell Color Company, Inc.
Murray, Andrew, Wohl, Ellen, and East, Jon, 1992,
Thermoluminescence and excess *Ra dating of late
Quaternary fluvial sands, East Alligator River, Australia:
Quaternary Research, v. 37, p. 2941.
Nanson, G.C., Price, D.M., Short, S.A., Young, R.W., and Jones,
B.G., 1991, Comparative Uranium-Thorium and
thermoluminescence dating of weathered Quaternary alluvium
in the tropics of northern Australia: Quaternary Research, v.
35, p. 347366.
Nanson, G.C., and Young. R.W., 1987, Comparison of
thermoluminescence and radiocarbon age determinations from
Late-Pleistocene alluvial deposits near Sydney, Australia:
Quaternary Research, v. 27, p. 263269.
55
Nelson, A.R., 1992, Holocene tidal-marsh stratigraphy in south
central OregonEvidence for localized sudden submergence
in the Cascadia subduction zone, in Fletcher, C.P., and
Wehmiller, J.F., eds., Quaternary coasts of the United
StatesMarine and lacustrine systems: SEPM Special
Publication 48, p. 287301.
Nelson, A.R., and Personius, S.F., 1991, The potential for great
earthquakes in Oregon and WashingtonAn overview of
recent coastal geologic studies and their bearing on
segmentation of Holocene ruptures, central Cascadia
subduction zone, in Rogers, A.M., Walsh, T.J., Kockelman,
W.J., and Priest, G.R. compilers, Earthquake hazards in the
Pacific Northwest of the United States : U.S. Geological
Survey Open-File report 91441-A, 29 p.
Niem, W.A., 1976, Drainage basin morphology in the central Coast
Range of Oregon: Corvallis, Oregon State University, M.S.
thesis, 99 p.
Oregon State Water Resources Board, 1958, Umpqua River basin:
Salem, Oregon State Water Resources Board, 200 p.
1965, Mid-coast basin: Salem, Oregon State Water
Resources Board, 122 p.
Peterson, C.P., Kulm, L.D., and Gray, J.J., 1986, Geologic map of
the ocean floor off Oregon and the adjacent continental
margin: Oregon Department of Geology and Mineral
Industries Geologic Map Series map GMS-42, scale
1:500,000.
Pierson, T.C., 1977, Factors controlling debris-low initiation on
forested hillslopes in the Oregon Coast Range: Seattle,
University of Washington, Ph.D. thesis, 166 p.
Pinter, Nicholas, and Gardner, T.W., 1989, Construction of a
polynomial model of glacio-eustatic fluctuationestimating
paleo-sea levels continuously through time: Geology, v. 17,
p. 295298.
Porter, S.C., Pierce, K.L., and Hamilton, T.D., 1983, Late
Wisconsin mountain glaciation in the western United States, in
Porter, S.C., ed., The Late Pleistocene, Vol. 1 of Wright, H.E.,
Jr., ed., Late Quaternary environments of the United States:
Minneapolis, University of Minnesota Press, p. 71111.
Rantz, S.E., and Moore, A.M., 1965, Floods of December 1964 in
the far western United States: U.S. Geological Survey Open
File Report, 205 p.
Reneau, S.L., 1988, Depositional and erosional history of
hollowsApplication to landslide location and frequency,
long-term erosion rates, and the effects of climatic change:
Berkeley, University of California, Ph.D. dissertation, 327 p.
Reneau, S.L., and Dietrich, W.E., 1990, Depositional history of
hollows on steep hillslopes, coastal Oregon and Washington:
National Geographic Research, v. 6, p. 220230.
1991, Erosion rates in the southern Oregon Coast
RangeEvidence for an equilibrium between hillslope
erosion and sediment yield: Earth Surface Processes and
Landforms, v. 16, p. 307-322.
Reneau, S.L., Dietrich, W.E., Donahue, D.J., Jull, A.J.T., and
Rubin, Meyer, 1990, Late Quaternary history of colluvial
deposition and erosion in hollows, central California Coast
Ranges: Geological Society of America Bulletin, v. 102,
p.969982.
Reneau, S.L., Dictrich, W.E., Dorn, R.I., Berger, C.R., and Rubin,
Meyer, 1986, Geomorphic and paleoclimatic implications of
56 FLUVIAL TERRACES, CENTRAL COAST RANGE, WESTERN OREGON
latest Pleistocene radiocarbon dates from colluvium-mantled
hollows, California: Geology, v. 14, p. 655658.
Reneau, S.L., Dietrich, W.E., Rubin, Meyer, Donahue, D.J., and
Jull, A.J.T., 1989, Analysis of hillslope erosion rates using
dated colluvial deposits: Journal of Geology, v. 97, p. 4563.
Rhea, Susan, 1993, Geomorphic observations of rivers in the
Oregon Coast Range from a regional reconnaissance
perspective: Geomorphology, v. 6, p. 135150.
Rockwell, T.K., Keller, E.A., and Dembroff, G.R., 1988,
Quaternary rate of folding of the Ventura Avenue anticline,
western Transverse Ranges, southern California: Geological
Society of America Bulletin, v. 100, p. 850858.
Rogers, G.C., 1988, An assessment of the megathrust earthquake
potential of the Cascadia subduction zone: Canadian Journal of
Earth Science, v. 25, p. 844852.
Rypins, Steven, Reneau, S.L., Byrne, Roger, and Montgomery,
D.R., 1989, Palynologic and geomorphic evidence for
environmental change during the Pleistocene-Holocene
transition at Point Reyes Peninsula, central coastal California:
Quaternary Research, v. 32, p. 7287.
Sarna-Wojcicki, A.M., Champion, D.C., and Davis, J.O., 1983,
Holocene volcanism in the conterminous United States and the
role of silicic volcanic ash layers in correlation of latest
Pleistocene and Holocene deposits, in Wright, H.E., Jr., ed.,
The Holocene, Vol. 2 of Wright, H.E., Jr., ed., Late Quaternary
environments of the United States: Minneapolis, University of
Minnesota Press, p. 5277.
Schlicker, H.G., and Deacon, R.J., 1974, Environmental geology of
coastal Lane County Oregon: Oregon Department of Geology
and Mineral Industries Bulletin 85, 116 p.
Schlicker, H.G., Deacon, R.J., Olcot, G.W., and Beaulieu, J.D.,
1973, Environmental geology of Lincoln County, Oregon:
Oregon Department of Geology and Mineral Industries
Bulletin 81, 171 p.
Schumm, S.A., Mosley, M.P., and Weaver, W.E., 1987,
Experimental fluvial geomorphology: New York, John Wiley
and Sons, 413 p.
Shackleton N.J., and Opdyke, N.D., 1973, Oxygen isotope and
paleomagnetic stratigraphy of Equatorial Pacific core V28
238Oxygen isotope temperatures and ice volumes on a 105
and 106 year scale: Quaternary Research. v. 3, p. 3955.
Sherrod, D.R., 1991, Geologic map of a part of the Cascade Range
between latitudes 4344, central Oregon: U.S. Geological
Survey Miscellaneous Investigations Series Map I1891, scale
1:125,000.
Snavely, P.D., Jr., 1987, Tertiary geologic framework,
neotectonics, and petroleum potential of the Oregon
Washington continental margin, in Scholl, D.W., Grantz, A.,
Published in the Central Region, Denver, Colorado
Manuscript approved for publication June 11, 1992
Edited by Tom Kohnen
Graphics by Stephen F. Personius and Roger Highland
Photocomposition by Shelly A. Fields
and Vedder, J.G., eds., Geology and resource potential of the
continental margin of western North America and adjacent
ocean basinsBeaufort Sea to Baja California: Circum
Pacific Council for Energy and Mineral Resources, Earth
Science Series, v. 6, p. 305335.
Snavely, P.D., Jr., MacLeod, N.S., Wagner, H.C., and Rau, W.W.,
1976a, Geologic map of the Yaquina and Toledo quadrangles,
Lincoln County, Oregon: U.S. Geological Survey
Miscellaneous Investigations Series Map I-867, scale
1:62,500.
1976b, Geologic map of the Cape Foulweather and Euchre
Mountain quadrangles, Lincoln County, Oregon: U.S.
Geological Survey Miscellaneous Investigations Series Map
I868, Scale 1:62,500.
Spence, William, 1989, Stress origins and earthquake potentials in
Cascadia: Journal of Geophysical Research, v. 94, no. B3,
p. 30763088.
Stuiver, Minze, Braziunas T.F., Becker, Bernd, and Kromer, Bernd, f
t *
1991, Climatic, solar, oceanic, and geomagnetic inluences on
Late-glacial and Holocene atmospheric 14C/12C change:
Quaternary Research, v. 35, p. 124.
Stuiver, Minze, and Quay, P.D., 1980, Changes in atmospheric
carbon-14 attributed to a variable sun: Scicnce, v. 207,
p. 1119.
Stuiver, Minze, and Reimer, P.J., 1986. A computer program for
radiocarbon age calibration: Radiocarbon. v. 28,
p. 10221030.
Tushingham, A.M., and Pelticr. W.R., 1991, Ice-3GA new
global model of Late Pleistocene deglaciation based upon
geophysical predictions of post-glacial relative sea level
change: Journal of Geophysical Research, v.96, p. 44974523.
Walker, G.W., and Duncan, R.A., 1989, Geologic map of thc Salem
1 by 2 quadrangle, western Oregon: U.S. Geological Survey
Miscellaneous Investigations Series Map I-1893, scale
s
1:250,000.
Wells, F.G., and Peck, D.L., 1961, Geologic map of Oregon west
of the 121st meridian: U.S. Geological Survey Miscellaneous
Investigations Series Map I325, scale 1:500,000.
West, D.O., and McCrumb, D.R., 1988, Coastline uplift in Oregon
and Washington and the nature of Cascadia subduction-zone
tectonics: Geology, v. 16, p. 169172.
Wintle, A.G., and Huntley. D.J., 1982. Thermoluminescence dating
of scdiments: Quaternary Science Reviews, v. 1, p. 3153.
Wright, L.W., 1970, Variation in the level of the cliff/shore
platform junction along the south coast of Great Britain:
Marine Geology, v. 9, p. 347353.
* U.S. GovePNMENT PRINTING OFFICE: 1993-573-038186000
 w *
Ll /9%)(Z_

-- \UG gr
Advances Related to United States and
| International Mineral Resources: ...
Developing Frameworks and f *
Exploration Technologies "

| T.E.9, 2 s UNIVERSITY tnes

fld on: 07-29-93

EEZEB9, U-203, c >7-2

COVER.Part of a color shaded relief aeromagnetic map of north-central Minnesota.
Color changes represent broad changes in magnetic field strength. The warm colors (red,
orange, and yellow) represent areas of magnetic field highs; the cool colors (green, blue,
and purple) represent magnetic field lows.
Advances Related to United States and
International Mineral Resources:
Developing Frameworks and
Exploration Technologies

Edited by RICHARD W. SCOTT, JR., PAMELA S. DETRA, and BYRON R. BERGER

U.S. GEOLOGICAL SURVEY BULLETIN 2039

The Bulletin reports some accomplishments of the

U.S. Geological Survey in research on
mineral resources in 1992

UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1993

 U.S. DEPARTMENT OF THE INTERIOR

BRUCE BABBITT, Secretary

U.S. GEOLOGICAL SURVEY

Dallas L. Peck, Director

For sale by
USGS Map Distribution
Box 25286, Building 810
Denver Federal Center
Denver, CO 80225

Any use of trade, product, or firm names in this publication is for descriptive purposes only and
does not imply endorsement by the U.S. Government

Library of Congress Cataloging-in-Publication Data

Advances related to United States and international mineral resources : developing frameworks and exploration
technologies / edited by Richard W. Scott, Jr., Pamela S. Detra, and Byron R. Berger.
p. cm. (U.S. Geological Survey bulletin; 2039)
Includes bibliographical references.
1. ProspectingEquipment and supplies. 2. Mines and mineral resourcesUnited States. I. Detra,
Pamela S. II. Berger, Byron R. III. Series.
QE75.B9 no. 2039
[TN270)
557.3 sdc20 9243213
[622'.1'028] CIP
 CONTENTS
[Letters designate chapters]

Introduction USGS-Meeting our Nation's need for mineral resources

through applied, basic, and technology research
By Byron R. Berger and Pamela S. Detra............................................
Part 1. A Framework for Understanding United States Mineral Resources
A. Tertiary calderas and regional extension of the east-central part of the
TinticDeep Creek mineral belt, eastern Great Basin, Utah
By Douglas B. Stoeser..........................................................................
Basement structure in the Railroad ValleyGrant Range region, east
central Nevada, from interpretation of potential-field anomalies
By H. Richard Blank............................................................................ 25
Magnetic interpretation of a mineralized shear zone in the southern
part of the Idaho batholith
By M. Dean Kleinkopf......................................................................... 31
Role of lithospheric flexure and plate convergence in the genesis
of some Mississippi-Valley-type zinc deposits in the Appalachians
By Dwight C. Bradley.......................................................................... 35
Rb-Sr isotopic redistribution and character of hydrothermal fluids
in the Catheart Mountain Cu-Mo deposit, Maine
By Robert A. Ayuso and Nora K. Foley.............................................. 45
F. Structure and origin of the Ely copper deposit, east-central Vermont
By Terry W. Offield, John F. Slack, and Sarah A. Wittinbrink........... 60

Part 2. A Framework for Understanding International Mineral Resources

G. Gold, silver, and base metal epithermal mineral deposits around the
Gulf of California, Mexico: Relationship between mineralization
and major structures
By John-Mark G. Staude...................................................................... 69
Cenozoic geology and mineral deposits of the Berenguela district,
northwestern Bolivia
By Alan R. Wallace, Richard F. Hardyman, Richard M. Tosdal,
Nestor Jimenez, Jos Lus Lizeca, and Fernando Murillo................... 79
K-Arages of Bolivian Tertiary polymetallic vein deposits
By Steve Ludington, Edwin H. McKee, and Nora Shew..................... 87
Gold and silver in acid-sulfate alteration and quartz-sericite-pyrite
stockworks, La Espaola prospect, northwestern Altiplano, Bolivia
By Albert Hofstra, Richard F. Hardyman, Luis Barrera, and
Orlando Sanjines.................................................................................. 95
Mineralogy and chemistry of gold-associated skarn from Nambija,
Zamora Province, Ecuador: A reconnaissance study
By Jane M. Hammarstrom.................................................................... 107
Geochronology and geochemistry of the Ladolam gold deposit, Lihir
Island, and gold deposits and volcanoes of Tabar and Tatau,
Papua New Guinea
By James J. Rytuba, Edwin H. McKee, and Dennis P. Cox................ 119

Part 3. Developing Technologies for Mineral Resource Assessment and Exploration

M. Volatiles in clay minerals from sedimentary and hydrothermal
environments: A potential petrologic and minerals-assessment
tool?
By Gene Whitney and Gary P. Landis................................................. 127
III
IV CONTENTS

Mapping minerals with imaging spectroscopy

By Roger N. Clark, Gregg A. Swayze, and Andrea Gallagher............ 141
Satellite image processing for enhanced spectral discrimination
and interpretability
By Daniel H. Knepper, Jr..................................................................... 151
40Ar/39Ar studies of fluid inclusions in vein quartz from
Battle Mountain, Nevada
By Edwin H. McKee, James E. Conrad, Brent D. Turrin,
and Ted G. Theodore........................................................................... 155
Heavy minerals at the Fall ZoneA theoretical model of grain size,
density, and gradient
By Curtis E. Larsen.............................................................................. 167

Part 4. Testing Technologies for Mineral Resource Assessment and Exploration

R. Use of geochemical surveys in Ti-Hf-REE-Th-U placer exploration
A Mid-Atlantic-States example
By Andrew E. Grosz............................................................................. 181
The effects of hydrothermally altered bedrock on natural forest
vegetation in the Snow CampSaxapahaw area, North Carolina,
and the resulting expressions in Landsat TM imagery
By Alba Pays, Robert G. Schmidt, and Andreu Bonet....................... 189
Aeromagnetic survey of north-central Minnesota
By Robert J. Horton, Warren C. Day, and Robert E. Bracken............. 201
Mineralogic and fluid-inclusion studies of the Pea Ridge iron
rare-earth-element deposit, southeast Missouri
By Gary B. Sidder, Warren C. Day, Laurence M. Nuelle,
Cheryl M. Seeger, and Eva B. Kisvarsanyi......................................... 205
Mineral resource potential of the White River National Forest and the
Dillon Ranger District of the Arapaho National Forest, Colorado
By Margo I. Toth, Anna B. Wilson, Theresa M. Cookro, Viki Bankey,
Greg K. Lee, and John S. Dersch......................................................... 217
Geophysical studies of the White River National Forest, northwestern
and central Colorado
By Viki Bankey......................................................................... 233
Locating buried conductive material along the Getchell trend, Osgood
Mountains, Nevada: Implications for gold exploration and the
carbon-gold association(?)
By V.J.S. Grauch and Donald B. Hoover............................................. 237
Gold in Devonian carbonate rocks at Cedar Peak, southern Snake
Mountains, northeastern Nevada
By Charles H. Thorman and William E. Brooks.................................. 245
Occurrences of native gold containing silver, mercury, copper, and
palladium in Lemhi County, east-central Idaho
By George A. Desborough, William H. Raymond, and James Nishi. 263
Part 5. Gaps in the Information Base
AA. The importance of source rocks in formation of metallic-sulfide
ore deposits
By Bruce R. Doe................................................................................... 275
 INTRODUCTION

USGS-MEETING OUR NATIONS NEED FOR

MINERAL RESOURCES THROUGH APPLIED,
BASIC, AND TECHNOLOGY RESEARCH
By Byron R. Berger' and Pamela S. Detra'

Mineral resources are an essential component of the
foundation of a strong society and healthy economy. Pro
viding information for the public good about the origin,
quality, and quantity of mineral resources world-wide has
been an important mission of the U.S. Geological Survey
(USGS) since its creation in 1879. This information con
tributes to (1) the national security by assuring a supply of
raw materials; (2) the economic well-being of the United
States by maintaining a healthy minerals industry as well
as other industries, such as aerospace and steel, that
depend upon minerals; (3) public health by providing
understanding of the occurrence of metals in the natural
environment and the behavior of metals in potentially haz
ardous situations; (4) land-use decisions and land-use clas
sification by public and private interests; and (5)
government policy making.
To effectively accomplish a USGS goal of providing
mineral resource information for the public good, the
Office of Mineral Resources, Geologic Division of the
USGS, seeks to quickly and effectively disseminate
research results to government, commercial, and general
public interests. This Bulletin, Advances related to United
States and international mineral resources: Developing
frameworks and technologies, reports progress on some of
the current activities of the Office of Mineral Resources
and reflects the necessity for an international perspective
regarding mineral resources and resource technology.
The organization of this Bulletin is based on the phi
losophy that the needs of mineral-resource-information
users are best met by (1) establishing a geologic frame
work for mineral deposits nationally and internationally,
(2) developing technologies that will aid in the exploration
and evaluation of mineral resources and mineral resource
'U.S. Geological Survey, Mail Stop 973, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
potential, and (3) providing tests of new technologies to
provide measures of their utility. For many years, the
USGS has used this programmatic approach to fulfill its
mission of earth science in the public interest.
ESTABLISHING A GEOLOGIC
FRAMEWORK FOR
MINERAL DEPOSITS
The science of economic geology matured as a disci
pline within the earth sciences in the latter half of the 19th
century. Whereas earlier engineers and geologists tended
to classify ore deposits on the basis of form and (or) min
eralogy, genetic concepts became the preferred underpin
ning of the field as it evolved from a practiced art to a
science. Many USGS geologists contributed to the evolu
tion of thought in this science in its infancy (e.g., Samuel
F. Emmons, George Becker, Waldemar Lindgren, Raphael
Pumpelly, Frederick Ransome, and Josiah Spurr). Each of
these geologists and their colleagues believed that the gen
esis and occurrence of mineral deposits were best under
stood in the context of their geologic framework. The
presumption that different deposit types reflect differing
pressure and temperature environments has formed the
intellectual framework of economic geology for most of
the 20th century; this framework was established by the
early years of the century.
During the past three decades, a new perceptual
frameworkplate tectonics, in conjunction with new
research technologieshas revolutionized our understand
ing about the origin and occurrence of mineral deposits.
But, as in the early years of this century, it is still a frame
work based on sound information about regional geology
and detailed understanding about individual deposits.
2 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

30
Magmatic
segregation
24
CD
Cr
H High-T Mesozonal
; epizonal
18
C
SZ Dynamo
2 hydrothermal
+ 12
5.
Lll Low-T
O epizonal
6

ice

200 400 600 800

TEMPERATURE, IN DEGREES CENTIGRADE

Figure 1. Classification of mineral deposits in terms of pressures and temperatures that exist in different tectonic settings where specific
deposit types form (inferred from stable alteration mineral assemblages and geothermal gradients). The mineral-deposit styles are as fol
lows: (1) magmatic segregation, (2) mesozonal, (3) high-temperature epizonal, (4) dynamo-hydrothermal, and (5) low-temperature
epizonal. Selected, typical, mineral-deposit styles and examples for each of the deposit types are given in table 1.

With the advances in technology, scientists have
achieved new insights into the pressure and temperature
conditions under which mineral deposits are formed
(fig. 1) and which mineral-deposit styles fit into the pres
sure-temperature types (table 1). As a result, scientists also
now recognize the important relation between tectonic
environment and deposit type. The USGS is actively par
ticipating in this revolution in technology and understand
ing. Several papers in parts 1 and 2 of this volume reflect
our continuing attempt to understand the all-important
geologic framework within which deposits occur. Contri
butions include Stoeser (Chapter A); Blank (Chapter B);
Kleinkopf (Chapter C); Bradley (Chapter D); and Luding
ton and others (Chapter I). The remaining papers in parts
1 and 2 deal more specifically with mineral deposits or
mining districts (see Ayuso and Foley (Chapter E); Offield
and others (Chapter F); Staude (Chapter G); Wallace and
others (Chapter H); Hofstra and others (Chapter J); Ham
marstom (Chapter K); and Rytuba and others (Chapter L)).
DEVELOPING AND TESTING
TECHNOLOGIES FOR MINERAL
RESOURCE ASSESSMENT AND
EXPLORATION
The development and utilization of appropriate and
state-of-the-art technologies is integral to mineral
resource assessments and exploration. Regional targeting
of permissive mineral-resource terranes contributes to
both of these ends. Larsen (Chapter Q) is developing a
geologic model to target heavy mineral placers in the
southeastern United States. New and improved technolo
gies in remote sensing have resulted in major advances in
the utility of this approach to regional analysis, particu
larly in arid and semiarid regions. Clark and others
(Chapter N) and Knepper (Chapter O) describe important
advances in technology development in this field.
Once a regional assessment or exploration analysis
has been completed, follow-up technologies are useful for
outlining specific mineralized areas. Whitney and Landis
(Chapter M) are investigating volatiles in clay minerals as
a tool for outlining permissive or prospective areas within
sedimentary sequences. McKee and others (Chapter P) are
investigating the utility of the thermochronology of fluid
inclusions in mining-district-scale studies.
When technologies have been developed that appear
to be useful, performance testing under normal field con
ditions is valuable both to document the conditions under
which a technology is useful and to transfer that technol
ogy to other users. The papers in part 4 all take developed
technologies and test them with such objectives in mind.
Stream-sediment geochemistry is applied by Grosz (Chap
ter R) and lithogeochemistry by Thorman and Brooks
(Chapter Y), and both are components in the multidisci
plinary study described by Toth and others (Chapter V).
Stressed vegetation is tested by Pays and others (Chapter
S). Geophysical techniques are tested by Horton and
others (Chapter T) and Grauch and Hoover (Chapter X).
 USGS-MEETING OUR NATION'S NEED FOR MINERAL RESOURCES 3

Table 1. Selected, typical, mineral-deposit styles and examples Mineralogy is utilized by Sidder and others (Chapter U) to
for deposit types shown in figure 1. provide genetic insights into the Pea Ridge deposit, and
mineral chemistry is applied to gold placers by
Mineral-deposit Examples Desborough and others (Chapter Z).
style (locations)
Magmatic deposits
Layered mafic intrusions..................... Bushveld
Stillwater
GAPS IN THE INFORMATION BASE
Anorthosite titanium ........................... Allard Lake
Kimberlite diamonds........................... Kimberley Viewing mineral deposits from a source to sink
Carbonatite rare-earth ......................... Mountain Pass perspective has long been the subject of academic science.
Mesozonal deposits However, detailed studies of natural abundances of trace
Pegmatites, lithium ............................. Kings Mountain elements in all varieties of rocks have failed thus far to
Manono-Kitole elucidate which rocks unequivocally have served as metal
Rhyolite tin......................................... Black Mountains sources for most types of mineral deposits. Therefore,
High-heat-flow granites....................... Cornwall
source regimes have had little application in mineral
High-temperature epizonal deposits resource assessment and exploration activities. Neverthe
Porphyry copper.................................. Bingham Canyon less, as Doe (Chapter AA) asserts, the investigation of
El Salvador
characterizing source rocks should continue.
Porphyry molybdenum
(Climax style)................................. Climax
Epithermal gold-silver......................... Comstock lode
Pachuca
Polymetallic veins, replacements......... Bingham Canyon
Bolivia
Skarn-related copper........................... Bingham Canyon
Christmas
Cyprus massive................................... Cyprus
Turner-Albright
Kuroko massive sulfide....................... Kuroko
Kidd Creek
Dynamo-hydrothermal deposits
Polymetallic veins............................... Coeur d'Alene
Low-sulfide quartz veins..................... Mother Lode
Ballarat
Muruntau
Low-temperature epizonal deposits
Sedimentary iron................................. Lake Superior
Clinton
Mississippi Valley Pb-Zn.................... Viburnum
Cracoc-Silesia
Roll-front uranium .............................. Grants
Powder River
 CHAPTER A
TERTIARY CALDERAS AND REGIONAL
EXTENSION OF THE EAST-CENTRAL PART OF
THE TINTICDEEP CREEK MINERAL BELT,
EASTERN GREAT BASIN, UTAH
By DOUGLAS B. STOESER!
ABSTRACT
Caldera-related Oligocene (=34 Ma) rhyolitic tuff and
volcanic megabreccia occur at three localities in the east
central part of the TinticDeep Creek mineral belt in the
Basin and Range of western Utah. The volcanic rocks at
these localities (Maple Peak (West Tintic Mountains),
south Cherry Creek, and Desert Mountain) are
8,00010,000 ft, 4,200 ft, and > 5,000 ft thick, respec
tively. Megabreccia at Maple Peak and Desert Mountain
contain blocks > 1,000 ft long, whereas, at south Cherry
Creek, blocks are as much as 100 ft long. The tuff and
megabreccia at the Maple Peak and Desert Mountain
localities are interpreted to be intracaldera in origin and
belong to the Maple Peak and Desert Mountain calderas.
The megabreccia of the Maple Peak area had previously
been interpreted to be large breccia pipes but here is con
sidered to be intracaldera collapse breccia. The Maple
Peak caldera had been previously unrecognized. The south
Cherry Creek volcanic rocks may be either intracaldera or
caldera proximal and are interpreted to be associated with
the Maple Peak caldera. The relationship between the
intracaldera Maple Peak section and volcanic rocks of the
East Tintic Mountains, where a caldera (the Tintic caldera)
has been proposed, remains unresolved, but the volcanic
rock sections between the two ranges do not appear to
correlate.
The three sections of tuff and megabreccia dip an
average of 4050 ENE. This structural orientation is
interpreted as being caused by rotation of blocks above a
westward-dipping detachment, named here the Tintic
'U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
Valley detachment. Extensional deformation above the
detachment dismembered at least two caldera complexes,
the Maple Peak and Desert Mountain, segments of which
were tectonically transported west on the detachment.
Overall structural relationships are obscured by extensive
alluvial cover in the study area.
INTRODUCTION
An east-west-trending belt of Tertiary igneous rocks
and mineral deposits (the TinticDeep Creek mineral belt)
extends from the Tintic district west to the Deep Creek
Range in the northern part of the Delta 1x2 quadrangle
(figs. 1, 2) (Hilpert and Roberts, 1964; Shawe and Stew
art, 1976). The TinticDeep Creek belt contains two suites
of igneous rocks, a late Eocene to mid-Oligocene (4234
Ma) suite of intermediate to rhyolitic volcanic and intru
sive rocks (Lindsey and others, 1975; Morris and Lover
ing, 1979; Lindsey, 1982; Shubat and Snee, in press) and
an early Miocene to Holocene suite (=21 Ma) of basalts
and highly evolved, high-silica and topaz rhyolite and
granite (Turley and Nash, 1980; Lindsey, 1982; Chris
tiansen and others, 1986). The Eocene-Oligocene igneous
rocks align with the belt (the eastern part of which com
prises the Thomas RangeTintic volcanic field of Stokes,
1986), whereas the Miocene to Holocene igneous rocks
are distributed throughout the Sevier Desert region
roughly along a north-south axis. The Eocene-Oligocene
volcanic rocks have traditionally been viewed as having
formed prior to Basin-and-Range extension, whereas the
younger group of igneous rocks is clearly related to it.
Gans and others (1989), however, have interpreted vol
canic rocks of mid-Oligocene age in east-central Nevada
6 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
114 112
42 | |
- 110
T of the West Tintic Mountains shows the Maple Peak area
TNTic-DEEP CREEK
MNERALBELT .
40H
STUDY :
AREA l! :
DELTA QUAD. f rocks as being comprised chiefly of purplish-gray,
A. COLORADO
A. PLATEAU
BASIN AND A
RANGE A
38 H. ...'
A | | |
Figure 1. Index map of Utah showing the location of the study
area and the Delta 1X2 quadrangle.
(at the west end of the TinticDeep Creek belt) as being
coeval with early Basin-and-Range extension.
This paper presents preliminary evidence that the
igneous rocks and mineral deposits of the east-central part
of the TinticDeep Creek belt belong to several calderas
that have been disrupted and extended by large-scale
extensional faulting. These conclusions are based on study
of mid-Tertiary volcanic rocks at three localities: Maple
Peak, south Cherry Creek, and Desert Mountain (fig. 3),
where thick sections of tuff and megabreccia have been
previously reported (Morris and Kopf, 1970a, 1970b; Kat
telman, 1968; Walsh, 1987). This paper presents new
mapping and stratigraphy for these rocks and interprets
them as chiefly intracaldera in origin. Based on their
structural setting, it is also proposed that these rocks lie
within a highly extended terrane that is related to an
underlying detachment fault. Geochronologic and petro
logic studies of the igneous rocks from the study area are
in progress, and cooperative mapping of the Maple Peak
area is being conducted with the Utah Geologic Survey.
This report is a product of the Delta CUSMAP (Contermi
nous United States Mineral Assessment Program) project
of the Office of Mineral Resources of the U.S. Geological
Survey. This paper benefited from reviews and critical
comments by Daniel R. Shawe, David A. Lindsey, and
Thomas A. Steven.
MAPLE PEAK
Maple Peak includes the highest point in the West
Tintic Mountains and forms the southernmost mountain
within that range (figs. 2, 4). Previous geologic mapping
as being underlain by Precambrian quartzite, Paleozoic
limestone, a thick section of Tertiary undifferentiated vol
canic rocks, Tertiary megabreccia, and Tertiary monzo
- nitic intrusions, all of which are overlain by the Salt Lake
Series of clastic sediments of late Tertiary age (Groff,
1959; Morris and Kopf, 1970a; Pampeyan, 1989). Morris
and Kopf (1970a) described the undifferentiated volcanic
medium-grained, porphyritic latite and quartz latite lava
containing phenocrysts of andesine, sanidine, biotite, horn
blende, and quartz in a fine-grained crystalline matrix.
They also stated that the volcanic rocks locally include a
basal unit of white, fine-grained tuffs and buff unsorted
-
tuff breccias. They estimated the total thickness of vol
canic rocks to exceed 1,000 ft.
Morris and Kopf (1967) described three large breccia
pipes in the southern part of the West Tintic Mountains,
which they named the Maple Peak, Cherry Creek, and
Hassel breccia pipes (fig. 4). These pipes were thought to
consist of elongate bodies of coarse breccia cutting the
Tertiary volcanic rocks as well as older sedimentary rocks.
The megabreccias were described as consisting of a mas
sive chaotic breccia with 65 percent silicified, calcitized,
and chloritized felsic volcanic matrix and 35 percent
inclusions, including blocks as much as 1,200 ft long.
Inclusions in the breccia consist of quartzite, phyllite,
argillite, conglomerate, limestone, shale, chert, latite,
quartz latite, monzonite, quartz monzonite, and aplitic
granite. The breccia matrix was described as containing
unsorted angular and broken crystals of quartz, orthoclase,
andesine, biotite, and lithic fragments, in a very fine
grained groundmass.
Reexamination of the volcanic rocks and breccia
along a northeast traverse across Maple Peak revealed a
thick section of intercalated rhyolitic welded tuffs and
coarse volcanic megabreccia capped by a lacustrine(?)
sedimentary unit (fig. 4, table 1). The volcanic rocks at
Maple Peak dip 2065 NE., with an average dip for the
entire section of approximately 4045.
The lower part of the volcanic-rock section at Maple
Peak is approximately 6,500 ft thick and consists chiefly of
altered rhyolitic welded tuffs (units tits, fig. 4, table 1)
(part of the lower section was not seen along the traverse
where it is intruded by monzonite porphyry, fig. 4). The
tuffs all contain phenocrysts of plagioclase, sanidine,
quartz, and biotite and range from relatively crystal rich to
crystal poor with abundant pumice fragments and few lithic
clasts. Two massive latite units (fl and f2, fig. 4) were
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH 7

Lake CRICKET
MTNS

* >
22 |

25 MILES

Figure 2. Index map to geographic features of the Delta 1X2 quadrangle and to figure 3. Solid lines are paved highways and roads;
dashed lines are improved dirt roads. Light-gray areas are ranges and areas of outcrop. White areas are intermontaine alluvium. Dark-gray
areas within the Sevier Desert are Pliocene and Quaternary basaltic volcanic rocks that lie on basin fill. Areas with diagonal lines are lakes.

encountered on the traverse that may be either flows or
flow domes intercalated in the section, allochthonous slabs
of older volcanic rock incorporated within the t2 tuff unit,
or hypabyssal intrusives. The hypabyssal-intrusion interpre
tation is based on the observation that monzonite and latite
intrude mid-Tertiary volcanic rocks in the East and West
Tintic Mountains and at Desert Mountain (see below).
The lower part of the traversed section contains sev
eral areas of dense phenocryst-poor andesite breccia (unit
Tab, fig. 4). Morris and Kopf (1970a) described this brec
cia as consisting of fragments of fine-grained, dark-red
dish-brown andesite; monzonite porphyry; quartzite; and
carbonate rocks in a sparse matrix of hematite-stained, pul
verized andesite. Where seen by the author, this unit con
sists of monomict breccia with fine-grained andesitic clasts
in a fine-grained andesitic matrix. These rocks appear to
have formed by brecciation of a massive andesite and may
be flow breccia or agglomerate, or they may have been
brecciated subsequent to cooling. The andesite breccia was
interpreted by Morris and Kopf (1970a) to be breccia
intruded into the volcanic rocks, but, to me, they appear to
be inclusions of andesite breccia incorporated into the
welded tuffs. I also found several 6- to 8-ft-diameter clasts
of the andesite breccia in the lower Maple Peak megabrec
cia. These andesitic breccia bodies occur north and west
of, but spatially near, a large mass of andesite at the south
eastern part of Maple Peak (fig. 4)these breccia bodies
may represent older volcanic rock that was dismembered
during formation of the tuffs and megabreccias.
The upper part of the Maple Peak section consists of
rhyolitic welded tuffs and coarse megabreccia (fig. 4, table
1). Although Morris and Kopf (1967) indicated that the
megabreccia of Maple Peak intrudes the volcanic rocks, I
reinterpret the megabreccia as being volcanic megabreccia
that is intercalated in the volcanic section and that overlies
the lower tuff section. This reinterpretation is, in part,
based on the observation that the central part of the Maple
Peak megabreccia contains a clast-poor zone (fig. 4). The
clast-poor zone is poorly exposed, but, where examined, it
is a eutaxitic welded tuff. Furthermore, this tuff dips
approximately 40 N.E., indicating that this unit has a
strike and dip conformable with the volcanic section
above and below the megabreccia. Morris and Kopf
(1970a) showed several 85 N.E. dips for the megabreccia
in the traverse area, but I was unable to replicate these
measurements when I visited these locations, and they are
8 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
11300'
4000'
3930'
H O 5 10 MILES
H-3,
EXPLANATION
Tertiary and Quaternary alluvium
Pliocene and Quaternary basaltic volcanic rocks
Miocene and Pliocene rhyolite (including topaz rhyolite)
Oligocene and Miocene granite and monzonite
Oligocene volcanic megabreccia
Eocene and Oligocene intermediate and silicic
volcanic rocks
Cretaceous and Paleocene sedimentary rocks
Figure 3. Simplified geologic map of the northeastern part of the Delta 1X2 quadrangle and index map to figures 4, 5, and 6. Also
shown are known and proposed calderas and caldrons as discussed in text: A, Thomas caldera; B, Picture Rock caldera; C, Keg cal
dron; D, Dead Ox caldron; E, Desert caldera; and F, Tintic caldera.
not shown on figure 4. Morris and Kopf (1967) were well
aware of the clast-poor zone within the Maple Peak mega
breccia stating on page C69 of their paper that The cen
tral area of this pipe is dominantly composed of tuffaceous
material and rare boulder-sized clasts of sedimentary
rocks. They ascribed the clast-poor central area to zona
tion within the breccia pipe. It is also clear from their text
that I am not the first to propose that the central zone of
the megabreccia is a welded tuff: Morris and Kopf also
note on the same page that The central zone of the brec
cia pipes closely resembles ashfall arkosic tuffs and com
monly has been mistaken for a pyroclastic deposit. They
11.230 11.200'
H Paleozoic sedimentary rocks (mainly limestone)
Precambrian and Cambrian clastic sedimentary rocks
- Contact
-A-A-A- Thrust fault -- Saw teeth on upper plate
- Fault
1 Caldera margin, dashed where approximate
- - - - - - - -
Caldera or caldron margin, inferred (see text for explanation)
also stated in several places within their paper that the
welded tuffs of the volcanic section superficially resem
ble the tuffaceous and fragmental matrix of the pipes, but
they did not state in what way the breccia-pipe matrix dif
fers from the tuffs. If, indeed, the central part of the Maple
Peak megabreccia is a welded tuff that is conformable
with the rest of the section, it is difficult to retain the
interpretation of Morris and Kopf (1967) that the mega
breccia represents a breccia pipe that intrudes the volcanic
rocks. The megabreccia of Maple Peak is here reinter
preted to be two volcanic megabreccia units that are inter
calated in the welded tuff section. The lower megabreccia
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH 9

Table 1. Description of stratigraphic units along traverse shown in figure 4.

[--, part of section not seen in interval now occupied by monzonite intrusion (Tm)]

Map unit Thickness (feet) Rock type Description

sl >59() sedimentary rocks Calcareous shale, pumiceous(?) siltstone.
t7 1,360 rhyolitic welded tuff Altered, tan to light-brown, pumice rich,
lithic-clast poor, good eutaxitic texture,
~25% phenocrysts of quartz, feldspar,
biotite and hornblende.
mb2 94() megabreccia Altered crystal-rich rhyolitic matrix with
phenocrysts of quartz, feldspar, and biotite.
t 34() rhyolitic welded tuff Light-gray, densely to poorly welded, weak
layering with good eutaxitic texture, -3-6%
phenocrysts of sanidine, quartz, plagioclase
and biotite.
mbl 63() megabreccia Cream-colored rhyolitic matrix, with < 1.5%
phenocrysts of sanidine, plagioclase, quartz,
and biotite.
tS 0-170 rhyolitic welded tuff Light-gray, massive, -10-12% phenocrysts
(sanidine, plagioclase, quartz, trace biotite),
~5% lithic clasts, wedges out to east.
t4 61() rhyolitic welded tuff Medium-brown, very altered, flow banded
with distinct eutaxitic texture (with greatly
flattened pumice (fiamme), * 10% pheno
crysts of quartz, feldspar, and biotite(?),
sparse lithic clasts.
t3 1,260 rhyolitic welded tuff Medium-gray, altered, massive, struc
tureless, pumice rich, crystal poor (~10%,
quartz, feldspar, biotite), -35% lithic clasts.
f2 250 porphyritic latite(?) Massive, with ~35% phenocrysts
(plagioclase, hornblende, biotite, clino
pyroxene) in devitrified groundmass, lower
part of unit light-gray, upper part altered,
mottled green and light-purple.
t2 1,540 rhyolitic welded tuff Mostly massive, structureless, altered,
crystal poor, cream to buff colored, ~15%
phenocrysts plagioclase, quartz, sanidine,
biotite, sparse small lithic clasts, dense,
well-laminated zone near top with good
eutaxitic texture.
f] 215 porphyritic latite(?) Approximately 5% plagioclase phenocrysts
in very fine grained, dense, light-gray
groundmass.
-- 2,020 -- Part of section not seen in interval now
occupied by monzonite intrusion (Tim).
tl 420 rhyolitic welded tuff Dark brown, 40% phenocrysts (plagioclase,
sanidine, quartz, biotite), -30% pumice
fragments, sparse lithic fragments.
Total 10,345

along the traverse is approximately 630 ft thick; the upper
megabreccia is approximately 940 ft thick; and the inter
vening tuff is 340 ft thick (table 1). Preliminary 39Ar/40Ar
biotite and sanidine ages for the tuff yield 34.6+0.1 and
34.4+0.05 Ma respectively (L.W. Snee, unpub. data).
Even without the evidence of a welded tuff unit
within the Maple Peak megabreccia, it is difficult to
accept the megabreccias of the study area as breccia pipes
because of their size. Their approximate dimensions are:
1.5 by 3.5 mi for the Maple Peak megabreccia, 1 by 3 mi
for the Cherry Creek megabreccia, and 0.5 by 1 mi for the
Hassel megabreccia. Lipman (1984) noted that, although
diatremes (breccia pipes) produce structures containing
coarse breccias similar to the megabreccias of ash-flow
10 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

11222'30" 11220'00" 11217'30"

| * E I
TC TV7

3950'00"

3950'00"

TV

Trn

Figure 4 (explanation is on facing page). Geologic map of Maple Peak area, including detailed stratigraphy along a traverse
across Maple Peak. Traverse location bounded by heavy lines in east-central part of map. Labeled rock units within traverse
area are described in table 1. Map is modified from Morris and Kopf (1970a). Megabreccias are shown with names assigned
by Morris and Kopf (1967).

calderas, the maximum known diameter of diatremes is
less than 1 mi. Another feature of these megabreccias is
their spatial association with the welded tuff section,
which would easily be understood if they were layered
units within that section. If the megabreccias are layered
units within the section, the map pattern of the Cherry
Creek and Maple Peak megabreccias shown on figure 4
suggests that they are continuous, and not separate, mega
breccias.
A notable feature of the Hassel breccia pipe, as
mapped by Morris and Kopf (1970a), is the large number
of strike and dip measurements within the unit (only some
of which are shown on fig. 4), most of which are 4585
NE. A well-defined, northeast-dipping layering is an odd
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH
EXPLANATION
Quaternary-Tertiary alluvium
Conglomerate
[T] Jasperoid
Monzonite porphyry
Intrusive latite
t Fault
Q)
ETH Basalt H.
Megabreccia
Volcanic rocks, undifferentiated
Tab Andesite breccia
feature for a breccia pipe, but it is consistent if the Hassel
unit is a layered unit within the volcanic section. I have
examined only the southeasternmost part of the area
mapped as the Hassel pipe and found it to be a massive,
clast-poor, welded tuff.
Overlying the upper Maple Peak megabreccia is
another thick (1,360 ft) rhyolitic tuff, which, in turn, is
overlain by sedimentary rocks that are a minimum of 590
ft thick (table 1). These sedimentary rocks were not exam
ined closely but, where observed on the traverse, consist
of calcareous shale and siltstone. This unit was not recog
nized by Morris and Kopf (1970a) in their mapping. The
relationship of the sedimentary rocks to the tuff is not
known, but the attitude of the sedimentary rocks is
approximately conformable with the tuff. Morris and Kopf
(1970a) mapped several basalt units to the north of the
traverse (fig. 4), and, based on stratigraphic relationships
observed at the south Cherry Creek study area (discussed
below), these may be basalt flows intercalated with the
sedimentary rocks.
The occurrence of thousands of feet of very thick
welded tuffs intercalated with very coarse volcanic mega
breccia is distinctive of an intracaldera ash-flow sequence
(Lipman, 1976, 1984; Best and others, 1989), and I, there
fore, conclude that the tuff and megabreccia of the Maple
Peak area were deposited within an ash-flow caldera.
Given the fairly consistent dip to the northeast and lack of
evidence for a repeat in the tuff-megabreccia section, the
entire intracaldera sequence appears to be exposed in the
Maple Peak area. The approximate thickness of 10,000 ft
for the Maple Peak section suggests that the caldera was
large.
The overall geometry of the proposed caldera is
unclear, and resolution of its features must await further
mapping; however, a few features of the caldera can be
11
HESH Paleozoic sedimentary rocks (mainly limestone)
Precambrian quartzite
H45 Strike and dip of beds
- Contact
AA Thrust fault - Saw teeth on upper plate
noted here. Best and others (1989) observed that, in other
calderas, lenses of landslide breccia within the intra
caldera tuff thin away from the caldera wall from which
they were derived, but can extend many kilometers into
the center of the depression. The shape of the Maple
PeakCherry Creek megabreccia fits this description (very
thick to the northwest and wedging out to the southeast)
and suggests that the collapsed-wall source of the mega
breccia inclusions was to the northwest. Another notable
feature of the Maple Peak sequence as mapped by Morris
and Kopf (1967, 1970a) is the large mass of Paleozoic
limestone that overlies the Maple PeakCherry Creek
megabreccia and, in turn, appears to be overlain by tuff
(fig. 4). This limestone may be a very large, intact block
of the wall that slumped in as the caldera wall collapsed.
The intact nature of the Maple Peak section suggests that
the floor of the caldera is exposed to the southwest of the
basal volcanic units and that the West Tintic mining dis
trictimmediately to the west of the Hassel megabreccia
unitand the intrusives contained in that district (Gard
ner, 1954; Stein and others, 1988) are located in the floor
of the caldera. The sedimentary rocks that overlie the
Maple Peak volcanic rocks may be late caldera fill or they
may be part of a regional series of mid- to late-Tertiary
sediments that are found throughout the region (Heylmun,
1965; McDonald, 1976).
SOUTH CHERRY CREEK
The south Cherry Creek study site (fig. 3) consists of
a series of low hills south of the Sheeprock Mountains
where Cherry Creek empties southwards into the Sevier
Desert. The area was first mapped by Morris and Kopf
(1970b), who recognized four stratigraphic units (from
12 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
bottom to top): Tertiary welded tuff and breccia, volcanic
sandstone and limestone, gravel, and a basalt flow. Morris
(1987, p. 4) later described the tuff and breccia unit as a
massive to crudely layered rhyolitic tuff breccia that
erupted from breccia pipes located in the southern part of
the West Tintic and Sheeprock Mountains. He also pre
sented a K-Ar biotite age for the tuff of 34.5 Ma (although
Morris did not cite a source for this age determination,
possibly it is derived from the 34.6+1.7-Ma K-Ar biotite
age of Marvin and Dobson (1979) for a quartz-rich rhyo
lite tuff at the south Cherry Creek locality). Sanidine from
one of the Cherry Creek tuffs yielded a preliminary 39Ar/
40Ar age of 34.7+0.1 Ma (excess 40Ar) (L.W. Snee,
unpub. data).
In examining the tuff, I found vitrophyric layers
within the tuff, some of which could be traced for more
than one-half mile. Mapping these vitrophyres led to
defining an internal stratigraphy for the tuff based on
alternating vitrophyric and non-vitrophyric tuff zones (fig.
5). In addition, two megabreccia zones within the tuff
section were also mapped (Tml and Tm2, fig. 5). The
tuffs are pumice rich and lithic poor but highly variable in
degree of induration and phenocryst content, and the non
vitrophyric tuffs are pervasively altered such that the
mafic minerals have rarely survived. Dips within the tuff
section vary from 20 to 65 NE. Based on an average
dip of about 40 NE., the exposed tuff section is roughly
4,200 ft thick. Of particular note is the welded tuff above
the upper megabreccia unit (Tm2), which is an altered,
tannish-brown, pumice-rich, phenocryst-poor tuff that is as
much as 2,000 ft thick.
Vitrophyres occur throughout the tuff section and, in
general, are relatively thin (about 1.5 to 5 ft) and uniform
in thickness. Notable, however, is the vitrophyre at the top
of the tuff section that, although its thickness appears to
vary significantly along strike, is locally as much as 100 ft
thick. Most of the vitrophyres can be traced for hundreds
of feet or more, but some occur as thin lenses of only a
few tens of feet in length. One angular discordancy is
present in the lower part of the tuff section, where the
lowest vitrophyre is sharply truncated by the overlying vit
rophyre and tuff (fig. 5). The vitrophyres contain few to
abundant lithic clasts and phenocrysts of sanidine, plagio
clase, quartz, biotite, sparse hornblende, and trace clinopy
roxene in a glassy matrix. In hand specimen, the
vitrophyres are a dull black (perlitic) color, suggesting that
the glass is hydrated.
The lower megabreccia (Tim1), stratigraphically the
lowest unit observed at south Cherry Creek, consists
chiefly of large blocks of quartzite and limestone as much
as 60100 ft in diameter. The eastern two-thirds of the
southernmost outcrop of the lower megabreccia consists of
a very altered intermediate volcanic rock that I interpret to
be a large block within the megabreccia. The matrix of the
megabreccia was not observed, and thus, it is not known if
this unit is a volcanic megabreccia. The upper megabrec
cia (Tim2) forms a distinct layer within the tuff section.
The clast content of the upper megabreccia changes along
strike such that, in the southeast, it is a clast-rich breccia
with a tuff matrix and blocks as much as about 40 ft in
diameter, and, to the northwest, it passes into a tuff with
sparse clasts (< 5 percent) that are mostly less than 6 ft in
diameter. The breccia is polymict, with quartzite being the
dominant clast type, but clasts of limestone and various
types of volcanic rocks are also present, including crystal
rich rhyolite clasts similar to some of the underlying tuffs.
The upper megabreccia is less than 100 ft thick in the
southeast and thickens to the northwest. It includes a cap
zone of purplish, clast-poor, crystal-rich, densely welded,
agglomeratic tuff that grades downward into the mega
breccia.
The tuff sequence is overlain by a poorly exposed
section of sedimentary rocks composed of volcaniclastic
sandstone, water-laid tuff(?), and minor limestone. A oli
vine-basalt flow overlies the sedimentary rocks. Both units
dip to the northeast and appear to have about the same
structural attitude; both are roughly conformable with the
underlying tuff. The basalt may be intercalated with the
sedimentary rocks, but no rock crops out above the basalt.
Based on a dip of 3040 NE. for the sedimentary rocks,
they are 6001,000 ft thick. Morris and Kopf (1970b)
mapped tuff above the sediments and basalt, but field
examination failed to locate any tuff outcropsthe areas
mapped as tuff by them is here remapped as part of the
Tertiary pediment alluvium unit of figure 6.
The south Cherry Creek section bears an overall sim
ilarity to the Tertiary section found in the Maple Peak
area. Both are composed of rhyolitic tuffs that include
lower and upper megabreccias and are capped by a thick
section of sedimentary rocks and associated basalt.
DESERT MOUNTAIN
Desert Mountain is a small, isolated range that con
sists of two main geological units: a rhyolitic tuff with
megabreccia along the eastern margin and a granitic com
plex that forms the bulk of the range (fig. 6). Precambrian
quartzite and diamictite of the Sheeprock Group also crop
out at the northern end of the range (fig. 6). Desert Moun
tain has been mapped previously by Kattelman (1968),
Shawe (1972), Rees and others (1973), and part of Desert
Mountain along the Callao-Jericho road was mapped by
Walsh (1987). Calkins (1972) also did a magnetic and
gravity survey of Desert Mountain. The present report is
based, in part, on a new map of Desert Mountain being
prepared by D.B. Stoeser and D.A. Lindsey.
The tuff of Desert Mountain, as named by Walsh
(1987), occurs along the entire eastern flank of Desert
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH 13

11227'00"

3947'30"

EXPLANATION

Quaternary alluvium

Pediment alluvium

% Basalt

Tuffaceous Sandstone
shale, and conglomerate
|
Vitrophyre
|

T welded tuff
Megabreccia: Tm1, lower,
Tm2, upper -

Strike and dip of layering

3945'00"

Figure 5. Geologic map of south Cherry Creek area. Based in part on mapping by Morris and Kopf (1970b).

Mountain. As described by Walsh, it consists of two
facies: massive rhyolitic tuff and megabreccia with a rhy-
olitic matrix (the facies are gradational and simply reflect
that the tuff ranges from clast rich to clast poor). The tuff
is crystal rich, with phenocrysts of plagioclase, sanidine,
quartz, biotite, and sparse hornblende in an aphanitic
matrix that contains sparse, small, compacted pumice frag
ments and sparse to numerous lithic fragments. The tuff
has been dated by Lindsey and others (1981) at 34.5+1.3
Ma (zircon fission track).
14 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

11237'30" 11235'00" 11252'30"

;
EXPLANATION

Quatemary alluvium Qal

3950'00" H
Tuff of Desert Mountain, with
3. inclusion (see caption) Coyote Knoll
+ + - prospect area
To . Desert Mountain grantic complex ; ;
H ||

Quartz diorite

|ped
ped Diamictite
- --
E
5 | } -
-
-

~ E

|pe.-
Quartzite
Cl

3947'30" JERICHO ROAD

..... ===

3945'00"

3942'30" | | |

Figure 6. Geologic map of Desert Mountain. Main units described in explanation box within figure; other labeled units on
map refer to either inclusions within the tuff of Desert Mountain or dikes within the granitic complex of Desert Mountain.
Labeled inclusions are: a, Tertiary andesite; d. Precambrian diamictite; q, Precambrian quartzite; r, rhyolite; and v, andesite to
dacitic volcanic rocks, lahar, and volcanogenic sedimentary rocks. Labeled dikes are: f. felsite; m, monzonite; p, lamprophyre;
and t, latite. Also shown are areas of jasperoid (j) formed by replacement of tuff. Other symbols are the same as on figure 4.

Most of the megabreccia occurs south of the Callao zones are dominantly quartzite, with lesser amounts of
Jericho road (figs. 3, 6). The megabreccia is chaotic, diamictite, argillite, volcanic rocks, and sparse, small (<
matrix rich (> 80 percent), and contains a few blocks up to 10 cm) plutonic rocks. The volcanic-rock inclusions are
about 1,500 ft in length. Inclusions in the megabreccia comprised of andesite, latite, dacite, and sparse rhyolite.
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH
The large volcanic blocks located at the southern end of
the tuff outcrop (fig. 6) consist of intact sections of dacite,
latite, lahar(?), and volcaniclastic sedimentary rock.
The structural attitude of the tuff is problematical
because the tuff is massive and generally without well
defined layering so that determining attitudes in the field
is very difficult. Based on local, weak layering and orien
tation of pumice lappilli, it appears that the tuff has an
overall attitude of dipping steeply to the east (fig. 6).
Shawe (1972) and Walsh (1987) also indicate that the tuff
and megabreccia dip steeply east and is intruded by the
granite. Based on dips shown in figure 6, the exposed sec
tion of tuff and megabreccia is approximately 5,000 ft
thick.
The tuff of Desert Mountain is largely altered such
that the mafic minerals are replaced by aggregates of
opaque minerals, and hematite and feldspar are sericitized.
In some places, the tuff is very dense and hard, as if silici
fied. Of approximately 50 samples of the tuff that I exam
ined in thin section, primary mafic minerals were found
only in three samples. In addition to the regional low
grade alteration of the tuff, most of the tuff north of the
Callao-Jericho road has been argillically altered and
locally silicified to form jasperoid (fig. 6). The most
intensely altered area of tuff, the Coyote Knoll prospect
area (fig. 6), was explored for gold by Freeport-McMoran
during 19891990. The granite of Desert Mountain, north
of the road, also has been weakly to strongly argillically
altered, and, therefore, the alteration postdates the main
igneous units of Desert Mountain.
The granitic complex of Desert Mountain consists of
multiple intrusive units of biotite monzogranite (not shown
on fig. 6) and includes large roof pendants of quartz dior
ite. Granite intrudes the tuff of Desert Mountain and is
therefore younger. Although the granite has been dated
several times, it is not clear what units were dated because
previous workers recognized only a single granite at
Desert Mountain. Armstrong (1970) dated the granite by
the K-Ar method at 28.5+0.6 Ma (biotite) and 26.5+0.5
Ma (orthoclase). Lindsey and others (1975) using the fis
sion-track method obtained ages of 28.2 Ma (sphene) and
32.4 Ma (zircon). Walsh (1987) also cites an unpublished
Sr-Rb isochron age for the granite by E.H. Christiansen of
34.8+0.8 Ma. The granitic complex is intruded by dikes of
lamprophyre, mafic-rich porphyritic monzonite and quartz
monzonite, latite, and granitic felsite (fig. 6).
Both Kattleman (1968) and Rees and others (1973)
interpreted the tuff to be a shallow intrusive; Rees and
others, in particular, considered it to be a diatreme. Shawe
(1972) and Walsh (1987) interpreted the tuff to be extru
sive and intracaldera in origin. Given the great thickness
of the tuff-megabreccia section and the large size of some
inclusions in the megabreccia, I concur with the intra
caldera interpretation.
15
DISCUSSION
CALDERAS
A number of calderas have been proposed to lie
within the eastern half of the TinticDeep Creek belt (fig.
3). Of these, only the 3836-Ma Thomas caldera is clearly
defined and well accepted (Shawe, 1972; Lindsey, 1982).
Other proposed calderas and caldrons include the Tintic
caldera of the East Tintic Mountains (Morris, 1975), the
Desert caldera (Shawe, 1972), the Dead Ox and Keg cal
drons of the Keg Mountains (or Keg caldera of Shawe,
1972; Shubat and Snee, in press), and Picture Rock
caldera, which is immediately south of the Keg Mountains
(ARCO, unpub. report, Anaconda Geological Document
Collection, University of Wyoming, Laramie; Shubat and
Snee, in press). Only the Tintic caldera, the Desert
caldera, and the Maple Peak caldera (a newly proposed
caldera) are located within the study area and are dis
cussed here.
TINTIC AND MAPLE PEAK CALDERAS
Morris (1975) first proposed the Tintic caldera
primarily on the basis of the existence of a circular mag
netic anomaly that coincided with a thick section of inter
mediate to rhyolitic volcanic rocks in the southern part of
the East Tintic Mountains. The inferred caldera was
approximately 8 mi in diameter. Magmatic rocks of the
East Tintic Mountains were divided into two age groups
by Morris and Lovering (1979); an older group of about
3532 Ma and a younger group of about 18 Ma. The pro
posed Tintic caldera assemblage belongs to the older
group and was thought to have included two major out
flow units: the Fernow and Packard Quartz Latites. The
existence of a caldera in the southern East Tintic Moun
tains has been supported by Hannah and Macbeth (1990),
Hannah and others (1990, 1991), Macbeth (1990), and
Keith and others (1989, 1991)these workers felt that
caldera formation was related to rocks assigned to the
Copperopolis latite formation of Morris (1975) and Morris
and Lovering (1979) rather than to the Fernow or Packard
Quartz Latites.
Hannah and Macbeth (1990) and Macbeth (1990)
mapped a section of volcanic tuffs and flows, approxi
mately 3,200 ft thick, that they subdivided into lower and
upper pyroclastic sequences. The lower pyroclastic
sequence consists of andesitic tuff and tuff breccia that is
overlain by andesitic flows. The lower tuffs include mega
breccia with clasts exceeding 325 ft in diameter. The upper
pyroclastic sequence contains a lower rhyolitic lapilli tuff
overlain by lacustrine sediments and andesitic flows and
agglomerate. The two pyroclastic sequences are overlain
and underlain by thick sections of mafic (dominantly
16 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
andesitic) flow rocks. They propose two phases of caldera
collapse based on the tuffs and tuff breccias in the lower
part of each sequence. Available age dating of these rocks
cited by Hannah and Macbeth (1990) indicate they were
deposited at approximately 3537 Ma. Although Hannah
and Macbeth (1990) and Macbeth (1990) support a caldera
interpretation for these rocks, they admit that most of the
overt features of a caldera are absent due to complications
induced by post-caldera cover and structural events, includ
ing the caldera being truncated and downdropped on the
west by the Tintic Valley fault (Mabey and Morris, 1967).
They also reduced the size of the proposed caldera to a
diameter of approximately 6.5 mi.
Keith and others (1989, 1991) studied the upper pyro
clastic sequence, which they described as members of the
Copperopolis latite of Morris (1975) and Morris and Lov
ering (1979). They published 39Ar/40Ar geochronologic
data that support an age of 3537 Ma for these rocks.
They also showed that overlying andesitic and rhyolitic
rocks, as well as monzonite intrusions, were implaced
between 34.5 and 33.6 Ma. Based on their work they con
cluded that a large caldera related to the eruption of the
younger Packard Quartz Latite and Fernow Quartz Latite
is not present in this area. They continued on to conclude
that a smaller caldera related to the eruption of the tuff
member of the Copperopolis latite may be present.
In the Tintic area, two caldera-related volcanic-rock
sequences are now recognized: those previously assigned
to the Tintic caldera in the southern East Tintic Mountains
(Hannah and Macbeth, 1990; Keith and others 1989, 1991;
Macbeth, 1990; and Morris, 1975) and those of the West
Tintic Mountains as proposed in this paper. The two
sequences are separated by the approximately 10-mi-wide
Tintic Valley, and it appears that the rocks of both
sequences extend beneath the cover of the valley. Based
on the lack of correlation of the stratigraphy of the two
sections (i.e., 3,200 ft of mostly andesitic rocks in the East
Tintic Mountains versus 9,600 ft of rhyolitic tuff and
megabreccia in the West Tintic Mountains), it seems
unlikely that they represent the same calderas. In addition,
the East Tintic caldera-related rocks appear to be some
what older that those of the West Tintic Mountains (3537
Ma as opposed to 34.5 Ma), although certainly more work
is needed to resolve the age of formation of these rocks.
They are close in age, however, and it seems likely that
multiple caldera-forming events have occurred in the Tin
tic area during the early Oligocene. Because the caldera
related rocks of the West Tintic Mountains do not appear
to correlate with those in the East Tintic Mountains, I pro
pose recognizing a separate ash-flow caldera, the Maple
Peak caldera, named after the location where the best and
most complete exposures of the intracaldera rocks are
found. A relationship between East Tintic volcanic rocks
may exist based on: (a) the occurrence of magmatic events
equivalent in age to the Maple Peak intracaldera rocks in
the East Tintic Mountains, and (b), as proposed below,
Tintic Valley may have formed by extension, and the two
volcanic sequences may have initially been more in prox
imity than they are now. One interesting possibility is that
the Fernow Quartz Latite may be outflow from the Maple
Peak caldera. The Fernow is a large, welded, ash flow that
occurs throughout the southern part of the East Tintic
Mountains and for at least another 20 mi to the south. In
spite of the quartz latite name, the Fernow is a rhyolite
with 6870 percent silica and contains abundant pheno
crysts of plagioclase, quartz, biotite, and minor sanidine
and hornblende (Morris, 1975; de Vries, 1990). The more
recent works by Hannah and Macbeth (1990) and Keith
and others (1989, 1991) have failed to identify a source
for the Fernow in the East Tintic Mountains, but its lithol
ogy fits well with the rhyolitic tuffs of the Maple Peak
intracaldra sequence (table 1).
DESERT MOUNTAIN CALDERA
The Desert caldera was first proposed by Shawe
(1972) based on the presence of rhyolitic to quartz-latitic
ash-flow tuff units along the eastern margin of the range
and on a circular magnetic high over the western part of
the range. Shawe showed the caldera as including all of
Desert Mountain and centered on the western part of the
range (fig. 3). Both Shawe (1972) and Walsh (1987) inter
preted the granite to be resurgent into the tuff, which is
supported by available dating as shown by Walsh (1987)
and by recent mapping. Although the Desert Mountain
igneous rock assemblage has been assigned to the Desert
caldera by Shawe, I recommend changing the name to the
Desert Mountain caldera in order to make its name fit
more appropriately with its type locality.
Structural complications induced by regional exten
sion, as discussed below, indicate that calderas in the east
central part of the TinticDeep Creek belt have been dis
membered and that the exposed parts of these calderas
may not correspond with the locations of their root zones
(that is, the upper parts of the calderas have been trans
ported laterally away from their underlying intrusive
zones). Another significant problem is that alluvial cover
obscures most of the study area. As already noted, the
great thicknesses of tuff and large size of the inclusions in
the megabreccias at Maple Peak and Desert Mountain
indicate that these rocks must be intracaldera in origin.
Although a thick section of tuff and megabreccia occurs at
south Cherry Creek, the clast size of the megabreccia is
significantly less than for megabreccia at the other two
localities, and the south Cherry Creek volcanic rock
assemblage may be either intracaldera or very proximal in
origin. If proximal rocks lie between the two intracaldera
assemblages of Maple Peak and Desert Mountain, it is
unlikely that the Desert Mountain assemblage belongs to
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH 17

*r-
{4} MAPLE
~ ric
-

r & PEAK s

DESERT
MOUNTAIN
CALDERA

SEVIER
DESERT
COCORP LINE 1

EXTENSIONAL
** DOMAIN

39

Figure 7. Map of the Delta 1X2 quadrangle showing location of proposed calderas in the study area and the Sevier Desert extensional
domain and related features. Tintic caldera as shown by Hannah and Macbeth (1990). White arrows show direction of proposed extension
related to deformation of caldera-related volcanic rocks. Medium- and dark-gray areas are ranges and basaltic volcanic fields from figure
2. AA marks location of cross section shown on figure 8. BB' marks location of cross section shown on figure 10. Dotted line labeled
G7 represents a seismic line described by Planke and Smith (1991).

the same caldera as that of Maple Peak. If it did, then the
rocks of Desert Mountain were transported over 13 mi to
the west from the site of the Tintic calderathis requires
that a deep part of that caldera was transported the far
thest, a relation opposite to that expected. Thus, I con
clude that the intracaldera rocks of Desert Mountain
belong to a separate caldera from those of Maple Peak,
and I support the caldera interpretation of Shawe (1972)
for Desert Mountain (fig. 7).
REGIONAL EXTENSION
The study area lies north of the Sevier Desert Basin,
which underlies the central part of the Sevier Desert (figs.
2, 8). The basin is located between the Canyon Range on
the east and the Cricket and Drum Mountains on the west
and is more than 10,000 ft deep (Planke and Smith, 1991).
The Sevier Desert Basin is one of the major extensional
features of the eastern Basin and Range and has received
considerable study (McDonald, 1976; Lindsey, and others,
1981; Von Tish and others, 1985; Mitchell and McDonald,
1986; Planke and Smith, 1991), including COCORP seis
mic reflection profile line 1 (fig. 7) (Allmendinger and
others, 1983, 1985). The basin was formed by extension
on the Sevier Desert detachment, a low-angle normal fault
that heads along the west flank of the Canyon Range, dips
westward at about 1012, and extends at least as far
west as the House Range (Allmendinger and others, 1983;
Smith and Bruhn, 1984). An interpretative cross section of
the COCORP Seismic data across the Sevier Desert exten
sional domain by Von Tish and others (1985) is shown on
figure 8. Sharp (1984) has estimated that total displace
ment on the Sevier Desert detachment is between 28 and
38 km (1724 mi). Extension on the Sevier Desert detach
ment occurred primarily during two episodes (Von Tish
and others, 1985): the first episode began no later than the
middle Oligocene (before 28 Ma) and continued into the
middle Miocene, and the second episode began in the
middle Pliocene and continues to the present.
The Sevier Desert extensional domain and the under
lying Sevier Desert detachment have been mapped as far
north as the latitude of Leamington (fig. 7) (Von Tish and
others, 1985; Gans, 1987; Planke and Smith, 1991) or to
18 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
O A. -
W Sevier
Co
[+7-77
/ .
7 } ~!:-----------------------......
- ""-----
O s- f / -
3
O
1 - 7 '
; 2 t PAieczotc.And PRECAMERLAN
2 x of CRICKET. MuntalN:Bxck
LL - HOUSE RANGE
= 3 DETACHMENT
>
_TEsea
~~
l
g
# 4 -
_
gene
5
Figure 8. Interpretative cross section of Sevier Desert Basin along COCORP seismic reflection line 1, between A and A on figure 7
(2X vertical exaggeration) (from Von Tish and others, 1985).
about 12 mi south of Desert Mountain. Defining regional
extension to the north is problematical. The presence of
regional extension in the study area is suggested by atti
tudes of the Oligocene volcanic rocks within the eastern
part of the TinticDeep Creek belt. The volcanic rocks of
the West Tintic Mountains, south Cherry Creek, and
Desert Mountain all have a consistent, moderately steep
dip to the northeast, whereas volcanic rocks of similar age
in the East Tintic Mountains, to the east of the study area,
and those of the Keg Mountains, Thomas Range, and
Drum Mountains to the west, are nearly flat-lying. That is,
the Oligocene volcanic rocks that are north of and on the
trend of the known Sevier Desert extensional domain are
tilted and those to the flanks are not. The indicated area of
regional extension, however, does continue eastward from
the Sevier Desert extensional domain to Tintic Valley
north of the Leamington fault (fig. 7). It should be noted
that the eastern portion of the Sevier Desert detachment
has been removed by erosion due to uplift in the area of
the Canyon Range (Sharp, 1984). In addition to the struc
tural evidence, Planke and Smith (1991) describe results
from a seismic line that crosses the study area (dotted line
labeled G7 on fig. 7), that indicates a weak, discontinu
ous, westward-dipping reflector, whose projection inter
sects the surface just west of Jerichothey propose that
this may be the Sevier Desert detachment.
The occurrence of limestone xenoliths within the
Desert Mountain intrusive complex may provide direct
evidence for the Sevier Desert detachment beneath Desert
Mountain. Late, mafic-rich, monzonite porphyry dikes in
the northwestern part of the intrusive complex (fig. 6)
contain blocks of limestone and marble as much as 50 ft
in diameter (as well as large blocks of granite and small
inclusions of various types of mafic intrusive rocks).
Because it seems unlikely that these carbonate blocks
could have been derived from above the emplacement
level of the dikes, they must have come from below. This
==
Desert Basin - E
-
--- t-ft--.
O
K
Q
\
S SA
""------
---
* | 2 |
&\\
->
Pliocene and Quaternary
MIOCENE &V s | alluvium
& O ------ Pl basalt
2. -
: :*
imentary r
Pakeozoic and Precambrian
sedimentary rocks
5 MILES - *
relationship indicates that there is a bottom or floor to the
intrusive complex. The Desert Mountain intrusive complex
may be laccolithic in form, or alternatively, a detachment
may underlie the complex and the monzonite dikes post
date initiation of the detachment and sampled limestone
from beneath the detachment.
Regional extension in the Sevier Desert region can be
further defined by the gravity gradient map of the Delta
1x2 quadrangle (fig. 9) (Bankey and Cook, 1989). This
map shows the horizontal rate of change of the gravity
field. Where the gravity field changes sharply over a
short distance, it indicates boundaries between crustal
domains with significant differences in density. Typically,
these zones correlate with high-angle faults that bound
basins; these zones are due to the sharp gravity contrast
between alluvial basin fill and bedrock and, in most cases,
locate faults formed during regional extension. In figure
9, lines have been drawn along these zones for emphasis.
The western margin of the Sevier Desert extensional
domain, along the west flank of the Sevier Desert, is
well defined by a steep gravity-gradient zone that runs
west of the Cricket Mountains and along the east flank
of the Drum Mountains (fig. 9). Based on figure 9, I
have extended the western boundary of regional exten
sion northward between the Thomas Range and Keg
Mountain along an apparent, direct continuation of this
zone. The eastern margin of the domain in the study area
is drawn along the east margin of Tintic Valley where
Mabey and Morris (1967) and Morris (1975, 1987) locate
a major range-front fault. Tintic Valley exhibits a major
gravity low that Mabey and Morris (1967) and Cook and
others (1991, p. 18) have interpreted as indicating
approximately 6,000 to 7,000 ft of fill, although some
contribution to the gravity anomaly might be due to low
density caldera material beneath the valley. As previously
stated, Planke and Smith (1991) indicate that a detach
ment heads at the location of Jericho, which lies on the
 auperre
2
figure
boundaries
Range
(1989).
Cook
and
Bankey
data
original
from
adapted
quadrangle
1X2
Delta
the
of
map
gradient
gravity
Simplified
9.
sFigure
imposed 3uepresent
>
dark
1.53.5,
gray
Qmedium
Pliocene
are
pattern
diagonal
with
rAreas
D
and
C,
B,
A,
2.
figure
on
shown
fields
volcanic
east-west
a.5.
ternary

gxorpalidanzaoetniotnas)l.:
hover
magnitude
represent
eShades
for
1989,
Cook,
and
Bankey
(see
field
gravity
the
of
gradient
0.71.5,
gray
light
0.00.7,
white

Lu ; ~ O> :o

MILES
25

|
RANGE
MIDDLE

text.
the
in
discussed
are
that
lineaments

40 39
20 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

SW NE

DESERT MTN. S. CHERRY CK. MAPLE PEAK E. T.INTIC MTNS.

TiNTIC VALLEY

TINTIC VALLEY DETACHMENT

l APPRoximatel-Y 10 MILEs

Quaternary alluvium and # Tertiary sedimentary rocks Tertiary volcanic Tertiary andesitic flows,
Tertiary sedimentary megabreccia agglomerate, and tuff
rocks

Tertiary monzonite s Tertiary granite, with quartz Tertiary welded tuff = Paleozoic and Precambrian
diorite roof pendants F sedimentary rocks

Figure 10. Generalized interpretive cross section of study area (along BB' on figure 7). Geology of the East Tintic Mountains adapted
from Hannah and Macbeth (1990) and Macbeth (1990), Tintic Valley from a gravity interpretation by Cook and others (1991), Maple Peak
from Morris and Kopf (1970a) and this paper, south Cherry Creek from this paper, and Desert Mountain from Walsh (1987) and this paper.
Blank areas of the figure indicate parts for which the geology in unknown due to alluvial cover. No vertical exaggeration.

East Tintic range-front fault (fig. 8, 9). Tintic Valley is
interpreted to be underlain by a major listric fault that
forms a half-graben and that soles westward to form a
detachment that underlies the study region (fig. 10).
The gravity gradient data for the Delta 1x2 quad
rangle also exhibits four east-west lineaments (labeled A,
B, C, and D). The southern three lineaments, A, B, and
C, all mark disruptions in the north-south grain of the
data. These north-south features usually indicate exten
sional faults, either as range-front faults or intrabasinal
extensional faults. Lineaments A and B simply mark such
disruptions and do not have notable surface expression.
Lineament C is discussed separately below. Lineament D
is distinct from the other three in that it is a gravity
gradient lineament, which suggests that it may be a major
fault. Lineament D appears to extend almost the complete
length of the Delta 1x2 quadrangle. It is disrupted by
north-south-trending gravity-gradient features, indicating
that it is older than regional extension.
Lineament C is a major feature within the north
central part of the quadrangle. Its main characteristics
west of the Sevier Desert extensional domain are:
1. Disruption of north-south-trending gravity-gradient
features and, therefore, presumably disruption of
faults related to post mid-Tertiary extensional faulting,
2. Aligns with east-west strike-slip faults at the southern
end of the Deep Creek range (Smith and others, 1991),
3. Aligns with topographic highs and anomalous range
alignments as shown by the east-west-trending Middle
Range and smaller positive areas west of the Drum
Mountains; also aligns with truncations of the Deep
Creek, Fish Springs, and House Ranges,
4. Corresponds to an inversion in dip of Paleozoic sedi
mentary rocks between the Fish Springs and House
Ranges,
5. Aligns with intrusions in the area of Sand Pass, which
is located at the join between the Fish Springs and
House Ranges, and
6. Aligns with the southern margin of the Thomas caldera
and the flaring of the topography of the northern part
of the Drum Mountains.
East of the Drum Mountains, the projection of the lin
eament is largely under alluvial cover, but on the gravity
gradient map (fig. 9), there is a change from north-south
trending gravity-gradient features to the south of the linea
ment to east-west-trending ones to the north. Because of
the features defining the lineament at Sand Pass at the join
between the Fish Springs and House Ranges, the lineament
is informally named here the Sand Pass lineament. A major
feature of the Sand Pass lineament is that it forms the
southern boundary for the Oligocene magmatism of the
TinticDeep Creek belt. With few exceptions, all intru
sions and vents of this age are near or north of the Sand
Pass lineament. Altogether, the features associated with the
Sand Pass lineament indicate that crust to the north of the
lineament reacted differently to regional extension com
pared to crust to the south. East-west lineaments with simi
lar features have also been reported in southwestern Utah
and southern Nevada (Ekren and others, 1976; Rowley and
others, 1978). The nature of the Sand Pass lineament is
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH
unclear: it may mark a major crustal discontinuity in the
basement between Proterozoic terranes or between
Archean and Proterozoic terranes, or more simply, it may
represent a transition zone between two extensional
domains. Based on the evidence for regional extension and
a detachment beneath the study area and because features
of the Sevier Desert detachment are truncated at the Sand
Pass lineament, I conclude that a separate and probably
coeval detachment, named here the Tintic Valley detach
ment, and a related extensional domain underlie the east
central part of the TinticDeep Creek belt (fig. 10).
REFERENCES CITED
Allmendinger, R.W., Sharp, J.W., Von Tish, D., Oliver, J., and
Kaufman, S., 1985, COCORP profile across the Cordilleran
hingeline, west-central Utah, in Gries, R.R. and Dyer, R.C.,
eds., Seismic Exploration of the Rocky Mountain Region:
Rocky Mountain Association of Geologists, p. 2330.
Allmendinger, R.W., Sharp, J.W., Von Tish, D., Serpa, L., Brown,
W., Kaufman, S., Oliver, J.E., and Smith, R.B., 1983, Cenozo
ic and Mesozoic structure of the eastern Basin and Range Prov
ince, Utah, from COCORP seismic relection data: Geology, v.
11, p. 532536.
Armstrong, R.L., 1970, Geochronology of Tertiary igneous rocks,
eastern Basin and Range Province, western Utah, eastern Ne
vada, and vicinity, U.S.A.: Geochimica et Cosmochimica
Atca, v. 34, p. 202232.
Bankey, Viki, and Cook, K.L., 1989, Complete Bouguer gravity
map and related geophysical maps of the Delta 1X2 quadran
gle, Utah: U.S. Geological Survey Miscellaneous Field Stud
ies Map MF2081A, scale 1:250,000.
Best, M.G., Christiansen, E.H., Deino, A.L., Gromme, C.S.,
McKee E.H., and Noble, D.C., 1989, Excursion 3A
Eocene through Miocene volcanism in the Great Basin of
the western United States, in Chapin, C.E. and Zidek, Jiri,
eds., Field Excursions to Volcanic Terranes in the West
ern United States, Volume II: Cascades and Intermountain
West: New Mexico Bureau of Mines and Mineral
Resources Memoir 47, p. 91133.
Calkins, W.G., 1972, Magnetic and gravity study of Desert Moun
tain, Juab County, Utah: Utah Geological and Mineralogical
Survey Bulletin 95, 21 p., 1 plate.
Christiansen, E.H., Sheridan, M.F., and Burt, D.M., 1986, The ge
ology and geochemistry of Cenozoic topaz rhyolites from the
western United States: Geological Society of America Special
Paper 205, 82 p.
Cook, K.L., Mabey, D.R., and Bankey, Viki, 1991, The new
gravity map of UtahAn overview of its usefulness for geo
logic studies: Utah Geological Survey, Survey Notes, v. 24,
no. 4, p. 1221.
Ekren, E.B., Bucknam, R.C., Carr, W.J., Dixon, G.L., and Quinliv
an, W.D., 1976, East-trending structural lineaments in central
Nevada: U.S. Geological Survey Professional Paper 986, 16 p.
Gans, P.B., 1987, An open-system, two-layer crustal stretching
model for the eastern Great Basin: Tectonics, v. 6, p. 112.
21
Gans, P.B., Mahood, G.A., and Schermer, Elizabeth, 1989, Synex
tensional magmatism in the Basin and Range Province; a case
study from the eastern Great Basin: Geological Society of
America Special Paper 233, 53 p.
Gardner, W.C., 1954, Geology of the West Tintic mining district
and vicinity, Juab County, Utah: Salt Lake City, University of
Utah, unpub. M.S. thesis, 43 p.
Groff, S.L., 1959, Geology of the West Tintic Range and vicinity,
Tooele and Juab Counties: Salt Lake City, University of Utah,
unpub. Ph.D. thesis, 183 p., map scale 1:20,439.
Hannah, J.L., and Macbeth, Alec, 1990, Magmatic history of the
East Tintic Mountains, Utah: U.S. Geological Survey Open
File Report 900095, 24 p., 1 plate, scale 1:24,000.
Hannah, J.L., Stein, H.J., and Macbeth, Alec, 1990, A caldera in the
East Tintic Mountains, UtahField relations between Tertiary
magmatism and Tintic-type ore deposits [abs.), in Geology and
Ore Deposits of the Great Basin, Reno/Sparks, Nevada: Geo
logical Society of Nevada and U.S. Geological Survey, 1990
Program with Abstracts, p. 63.
Hannah, J.L., Stein, H.J., and Macbeth, A.P., 1991, Field relations
between Tertiary magmatism and Tintic-type ore deposits,
East Tintic Mountains, Utah, in Raines, G.L., Lisle, R.E.,
Schafter, R.W., and Wilkinston, W.H., eds., Geology and Ore
Deposits of the Great Basin: Geological Society of Nevada,
Symposium Proceedings, v. 1, p. 485489.
Hilpert, L.S., and Roberts, R.J., 1964, Economic Geology, in Min
eral and water resources of Utah: Washington, D.C., 88th U.S.
Congress, 2nd. Session, Committee Print, p. 2834.
Heylmun, E.B., 1965, Reconnaissance of the Tertiary sedimentary
rocks in western Utah: Utah Geological and Mineral Survey
Bulletin 75, 38 p.
Kattelman, D.F., 1968, Geology of the Desert Mountain intrusives,
Juab County, Utah: Brigham Young University Geology Stud
ies, v. 15, pt. 1, p. 85107.
Keith, J.D., Dallmeyer, R.D., Kim, Choon-Sik, and Kowallis, B.J.,
1989, A reevaluation of the volcanic history and mineral poten
tial of the central East Tintic Mountains, Utah: Utah Geologi
cal and Mineral Survey Open-File Report 166, 74 p.
Keith, J.D., Dallmeyer, R.D., Kim, Choon-Sik, and Kowallis, B.J.,
1991, The volcanic history and magmatic sulfide mineralogy
of latites of the central East Tintic Mountains, Utah, in Raines,
G.L., Lisle, R.E., Schafter, R.W., and Wilkinston, W.H., eds.,
Geology and Ore Deposits of the Great Basin: Geological So
ciety of Nevada, Symposium Proceedings, v. 1, p. 461483.
Lindsey, D.A., 1982, Tertiary volcanic rocks and uranium in the
Thomas Range and northern Drum Mountains, Juab Coun
ty, Utah: U.S. Geological Survey Professional Paper 1221,
p. 157.
Lindsey, D.A., Glanzman, R.K., Naeser, C.W., and Nichols, D.J.,
1981, Upper Oligocene evaporites in basin fill of Sevier Desert
region, western Utah: Bulletin of the American Association of
Petroleum Geologists, v. 65, p. 251260.
Lindsey, D.A., Naeser, C.W., and Shawe, D.R., 1975. Age of vol
canism, intrusion, and mineralization in the Thomas Range,
Keg Mountain, and Desert Mountain, western Utah: U.S. Geo
logical Survey Journal of Research, v. 3, no. 5, p. 597604.
Lipman, P.W., 1976, Caldera-collapse breccias in the western San
Juan Mountains, Colorado: Geological Society of America
Bulletin, v. 87, p. 13971410.
22 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Lipman, P.W., 1984, The roots of ash flow calderas in western
North AmericaWindows into the tops of granitic batholiths:
Journal of Geophysical Research, v. 89, p. 88018841.
Mabey, D.R., and Morris, H.T., 1967, Geologic interpretation of
gravity and aeromagnetic maps of Tintic Valley and adjacent
areas, Tooele and Juab Counties, Utah: U.S. Geological Survey
Professional Paper 516D, 10 p.
Macbeth, A.P., 1990, Geology of the southwestern East Tintic
Mountains, central Utah: University of Vermont, unpub. M.S.
thesis, 164 p.
Maciulaitis, P.E., 1991a, Jericho property, Juab County, Utah:
Boulder, Colorado, Euro-Nevada Mining Corporation, Inc.,
unpub. report, 28 p.
1991b, Rockwell/Sahara property, Juab County, Utah:
Boulder, Colorado, Euro-Nevada Mining Corporation, Inc.,
unpub. report, 24 p.
Marvin, R.F., and Dobson, S.W., 1979, Radiometric ages: Com
pilation B, U.S. Geological Survey: Isochron/West, no. 26,
p. 332.
McDonald, R.E., 1976, Tertiary tectonic and sedimentary rocks
along the transitionBasin and Range province to plateau and
thrust belt province, Utah, in Newman, G.W. and Goode, H.D.,
eds., Basin and Range Symposium and Great Basin Field Con
ference: Rocky Mountain Association Geologist Guidebook,
p. 503514.
Mitchell, G.C., and McDonald, R.E., 1986, History of Cenozoic ex
tension in central Sevier Desert, west-central Utah, from CO
CORP seismic reflection dataDiscussion: Bulletin of the
American Association of Petroleum Geologists, p. 10151021.
Morris, H.T., 1975, Geologic map and sections of the Tintic Moun
tain quadrangle and adjacent part of the McIntyre quadrange,
Juab and Utah Counties, Utah: U.S. Geological Survey Miscel
laneous Investigations Series Map I-883, scale 1:24,000.
Morris, H.T., 1982, Regional relations of Wah Wah-Frisco thrust
fault in western Utah: U.S. Geological Survey Professional Pa
per 1150, 64 p.
1983, Interrelations of thrust and transcurrent faults in the
central Sevier orogenic belt near Leamington, Utah, in Miller,
D.M., Tood, V.R., and Howard, K.A., eds., Tectonic and
Stratigraphic Studies in the Eastern Great Basin: Geological
Society of America Memoir 157, p. 7581.
1987, Preliminary geologic map of the Delta 2 quadrangle
Tooele, Juab, Millard, and Utah Counties, Utah: U.S. Geolog
ical Survey Open-File Report 87185, 18 p., scale 1:250,000.
Morris, H.T., and Kopf, R.W., 1967, Breccia pipes in the West Tin
tic and Sheeprock Mountains, Utah: U.S. Geological Survey
Professional Paper 575C, p. C66C71.
1970a, Preliminary geologic map and cross sections of the
Maple Peak quadrangle and adjacent part of the Sabie Moun
tain quadrangle, Juab County, Utah: U.S. Geological Survey
Open-File Report 70234, scale 1:24,000.
1970b, Preliminary geologic map and cross sections of the
Cherry Creek quadrangle and adjacent part of the Dutch Peak
quadrangle, Juab County, Utah: U.S. Geological Survey
Open-File Report 70233, scale 1:24,000.
Morris, H.T., and Lovering, T.S., 1979, General geology and mines
of the East Tintic mining district, Utah and Juab counties,
Utah: U.S. Geological Survey Professional Paper 1024, 203 p.
Pampeyan, E.H., 1989, Geologic map of the Lynndyl 30- by 60
minute quadrangle, west-central Utah: U.S. Geological Survey
Miscellanceous Investigations Series Map I-1830, scale
1:100,000.
Planke, Sverre, and Smith, R.B., 1991, Cenozoic extension
and evolution of the Sevier Desert Basin, Utah, from Seis
mic reflection, gravity, and well log data: Tectonics, v.
10, p. 345366.
Rees, D.C., Erickson, M.P., and Whelan, J.A., 1973, Geology and
diatremes of Desert Mountain, Utah: Utah Geological and
Mineralogical Survey, Special Studies 42, 12 p., 1 plate.
Rowley, P.D., Lipman, P.W., Mehnert, H.H. Lindsey, D.A., and
Anderson, J.J., 1978, Blue Ribbon lineament, an east-trending
structural zone within the Pioche mineral belt of soutwestern
Utah and eastern Nevada: U.S. Geological Survey Journal of
Research, v. 6, no. 2, p. 175192.
Sharp, J.W., 1984, West-central Utah: palinspastically restored sec
tions constrained by COCORP seismic reflection data: Ithaca,
New York, Cornell University, unpub. M.S. thesis, 60 p.
Shawe, D.R., 1972, Reconnaissance geology and mineral potential
of Thomas, Keg, and Desert calderas, central Juab County,
Utah, in Geological Survey Research 1972: U.S. Geological
Survey Professional Paper 800B, p. B67B77.
Shawe, D.R., and Stewart, J.H., 1976, Ore deposits as related to tec
tonics and magmatism, Nevada and Utah: American Institute
of Mining Engineers, Transactions, 1977 Annual Meeting, Las
Vegas, Nevada, v. 260, p. 225232.
Shubat, M.A., and Snee, L.W., in press, High-precision 40Ar/39Ar
geochronology, volcanic stratigraphy, and mineral deposits of
Keg Mountain, west-central Utah: U.S. Geological Survey
Bulletin 2012, Chapter G.
Smith, D.L., Gans, P.B., and Miller, E.L., 1991, Palinspastic resto
ration of Cenozoic extension in the central and eastern Basin
and Range Province at latitude 3940N., in Raines, G.L.,
Lisle, R.E., Schafter, R.W., and Wilkinston, W.H., eds., Geol
ogy and Ore Deposits of the Great Basin: Geological Society
of Nevada, Symposium Proceedings, v. 1, p. 7586.
Smith, R.B., and Bruhn, R.L., 1984. Intraplate extensional tectonics
of the eastern Basin-RangeInferences on structural style
from seismic reflection data, regional tectonics, and thermal
mechanical models of brittle-ductile deformation: Journal of
Geophysical Research, v. 89, p. 57335762.
Spenser, J.E., and Welty, J.W., 1986, Possible controls of base- and
precious-metal mineralization associated with Tertiary detach
ment faults in the lower Colorado River trough, Arizona and
California: Geology, v. 14, p. 195198.
Stein, H.J., Kelley, D.L., Kaminsky, J.F., and Gordon, I.R.,
1988, Field trip guide for the West Tinitic mining district,
western Utah: U.S. Geological Survey Open-File Report
880558, 12 p.
Stokes, W.E., 1986, Geology of Utah: Salt Lake City, Utah Geo
logical and Mineral Survey, 280 p.
Turley, C.H., and Nash, W.P., 1980, Petrology of late Tertiary and
Quaternary volcanism in western Juab and Millard Counties,
Utah: Utah Geological and Mineral Survey Special Study 52,
p. 133.
Von Tish, D.B., Allmendinger, R.W., and Sharp, J.W., 1985, His
tory of Cenozoic extension in central Sevier Desert, west-cen
tral Utah, from COCORP Seismic reflection data: Bulletin of
the American Association of Petroleum Geologists, v. 69, no.
7, p. 10771087.
 TERTIARY CALDERAS AND REGIONAL EXTENSION, EASTERN GREAT BASIN, UTAH 23

Vries, R.D. de, 1990, Tales of Tertiary tuffs in central Utah: De
Kalb, Northern Illinois University, unpub. M.S. thesis, 160 p.
Walsh, D.V., 1987, Geological and geochemical relationships of
the Desert Mountain igneous complex rocks: Salt Lake City,
University of Utah, unpub. M.S. thesis, 64 p.
Wilkins, J., Jr., Beane, R.E., and Heidrick, T.L., 1986, Mineraliza
tion related to detachments faultsA model, in Beatty, B., and
Wilkinson, P.A.K., eds., Frontiers in Geology and Ore Depos
its of Arizona and the Southwest: Arizona Geological Society
Digest, v. 16, p. 108117.

BASEMENT STRUCTURE IN THE
RAILROAD VALLEYGRANT RANGE REGION,
EAST-CENTRAL NEVADA,
FROM INTERPRETATION OF
POTENTIAL-FIELD ANOMALIES
By H. RICHARD BLANK!
Regional gravity and aeromagnetic anomaly maps are
widely used to delineate first-order structural features
expressed by marked density and magnetization contrasts.
In east-central Nevada and elsewhere in the miogeoclinal
domain of the Basin and Range extensional province, the
gravity method is especially effective for distinguishing
areas where the bedrock is veneered by pediment gravels
(or by Cenozoic volcanic and volcaniclastic rocks) from
areas underlain by relatively thick, low-density valley fill.
The density basement is generally rock of the miogeo
clinal carbonate succession or Mesozoic and Cenozoic
intrusive rock. The aeromagnetic method, on the other
hand, detects magnetic crystalline rock beneath the non
magnetic carbonates. This basement may be either
Phanerozoic intrusives or the Precambrian igneous-meta
morphic complex. Signatures of strongly magnetic vol
canic rocks of the Cenozoic section are typically
distinguished by their shorter spatial wavelengths, which
result from greater proximity to the detector and lesser
depth extent.
The original objective of recent U.S. Geological Sur
vey (USGS) gravity and aeromagnetic investigations in
east-central Nevada was to aid in mineral
assessment of Bureau of Land Management Wilderness
Study Areas. Later, the scope of investigation was
expanded to include a study of the structural framework
of sedimentary basins with known or potential hydrocar
bon deposits. Studies of basement structure in the Rail
road ValleyGrant Range region (fig. 1) focus on both
"U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
CHAPTER B
objectives and involve integration of potential-field, seis
mic, drilling, and geologic-map data. Railroad Valley
has a strike length of nearly 175 km, with an average
north-northeasterly trend. Its maximum width of about
25 km occurs where it is flanked on the east by the
Grant Range. This central sector is the locus of oil pro
duction in the valley. The nearby Troy district, in the
southern Grant Range, achieved prominence in the past
as a producer of gold, silver, and tungsten.
Figure 2 shows the complete-Bouguer gravity anom
aly field of the Railroad ValleyGrant Range region. The
color interval on the map is 2 milligals (mGal), and the
absolute anomaly level at Currant is -210 mgal. Primary
gravity data were obtained from files of the National Cen
ter for Geophysical and Solar-Terrestrial Data (Boulder,
Colorado 80303), supplemented by several dozen stations
recorded by the USGS in conjunction with current pro
grams. A reduction density of 2.67 g/cm3 was employed,
and all stations have been terrain-corrected out to a radius
of 167 km (see Cordell and others, 1982, for a description
of standard USGS gravity-reduction procedures). Strati
form geologic units shown on the figure are: Paleozoic
resource
sedimentary rocks, chiefly miogeoclinal carbonates (P.),
Tertiary volcanic and sedimentary rocks (T), and Quater
nary surficial deposits (Qa). Granitic intrusive rocks of
Cretaceous and Tertiary age (KTi) are represented by a
diagonal pattern; the largest such body exposed is the
Troy pluton. Also shown is the Shell No. 1 discovery hole
in Railroad Valley, which intersected quartz monzonite at
a depth of 10,330 ft (3.15 km). Quartz monzonite has been
intersected in several other drill holes at shallower depths
in the valley.
26 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
118
I others, 1985; Effimoff and Pinezich, 1986). The eastern
41
NEVADA
Area of study
Figure 1. Index map of Great Basin showing Railroad
ValleyGrant Range region, east-central Nevada.
The Bouguer anomaly map shows a north-northeast
trending anomaly depression in Railroad Valley that is
narrower than the full width of the valley and that contains
two collinear axial lows separated by a weak rise. All six
of the producing oil fields are situated on the margins of
the northern low, which has a residual amplitude of 43
mCal. Exploration drilling has established that the thick
ness of valley fill (Tertiary and Quaternary deposits,
including Tertiary volcanic rocks) in this area exceeds 4.2
km (see Garside and others, 1988, for a recent compen
dium of drilling data). At such depths, the prevailing geo
thermal gradient may be sufficient to convert organic
material to oil (Hulen and others, 1991), and the presence
of high-angle faults on the margins of the deep could have
facilitated upward migration to suitable traps and reser
voirs. The mean density contrast corresponding to a 43
mGal anomaly produced by 45 km of valley fill is much
less than the 0.400.45 g/cm3 predicted for fill relative to
Paleozoic carbonates (Guion and Pearson, 1979) or to
igneous intrusive rock. This discrepancy is probably attrib
utable to an abundance of high-density carbonate materials
in the Tertiary and Quaternary valley fill, as suggested by
down-hole sonic and lithologic logs. Megaclasts of Paleo
zoic carbonate rocks are locally present and, in some
places, have been detected by detailed, high-precision,
commercial gravity surveys.
The 43-mCal anomaly low has horizontal gradients
along its eastern flank of as much as 11 mGal/km,
whereas those on the western flank are about half as
strong. Accordingly, the bedrock surface beneath Railroad
Valley can be modeled as an asymmetric graben or half
graben with its floor tilted down to the eastthis is in
agreement with seismic-reflection results (Bortz and
Murray, 1979; Vreeland and Berrong, 1979; Foster and
marginal gradient is a continuous, well-constrained gravity
feature that follows the trend of the range front for a dis
tance of 65 km, from near Currant to the southwest corner
of the map area (fig. 2). The locus of maximum gradients
in this belt is a line 210 km west (basinward) of the near
est bedrock outcrops. The line of maxima roughly coin
cides with a line of springs and groundwater seepage and
is believed to mark the approximate trace of the main
boundary fault (range-front fault) on the west side of the
Grant Range (Foster, 1979). Opposite the two gravity
lows, where the gradient is steepest, the fault or fault zone
is probably a high-angle structure, but in the intervening
segment (in the vicinity of the Troy pluton), dips as low as
20 have been interpreted from seismic data (Potter and
others, 1991). Either flatter dip or deeper burial of a spe
cific density discontinuity would weaken the gradient.
Down-hole logs show the rocks immediately east of the
gradient maxima, just north of the Troy pluton, to be pre
dominantly high-velocity, high-density, Paleozoic carbon
ate sand/gravel and breccia (identified as valley fill) below
the uppermost 400 m or so of section. A transverse seis
mic profile in this area reveals east-dipping reflectors
abutting an underlying, gently west-dipping reflector that
is interpreted as a low-angle normal fault. Thus, where the
gravity gradients are weaker, the range-front fault appar
ently truncates sections that include detached and rotated
allochthonous Paleozoic strata (either megaclasts in val
ley fill or strata that form the hanging wall of a low
angle normal fault).
Residual aeromagnetic total-field intensity for the
Railroad ValleyGrant Range region is shown on figure 3.
This map was prepared in the USGS Branch of Geophys
ics by merging data from several different surveys (after
removing the International Geomagnetic Reference Field
corresponding to the date of each survey, draping to an
elevation of 1,000 ft (305 m) above terrain, and reducing
to the pole). The generalized geologic units on this figure
are the same as those of figure 2, and, in addition, the
locus of maximum horizontal gradients of the Bouguer
anomaly field on the east side of Railroad Valley (range
front fault; see above) is indicated by a heavy dashed line.
The color interval on the map is 10 nanoteslas (nT), and
the absolute residual field intensity over the Shell No. 1
discovery hole is about +12 nT.
The main aeromagnetic feature of interest is a broad
positive anomaly with three principal crests that encom
passes the northern half of Railroad Valley (most of the
sector shown on fig. 3) and most of the Grant Range. The
axis of this anomaly tends to parallel that of the negative
gravity anomaly associated with Railroad Valley: both
extend N.-S. across the area of figure 3 and are convex
eastward. No offset of the aeromagnetic feature is
observed where it is transected by the line of maximum
 BASEMENT STRUCTURE IN EAST-CENTRAL NEVADA 27

11545' 11530' 115 15'

- f --: WHITE PINE 3.

- " oak/

O 5 10 MILES
L I I

O 5 10 KILOMETERS
||

Figure 2. Complete-Bouguer gravity anomaly map of Railroad ValleyGrant Range region, east-central Nevada. Color interval 2
mCal. P., Paleozoic sedimentary rocks; KTi, Cretaceous and Tertiary granitic intrusive rocks; T, Tertiary volcanic and sedimentary
rocks; Qa, Quaternary surficial deposits. Quartz monzonite (qm) was penetrated at 10,330 ft in Shell No. 1 discovery hole.
28 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

11545' 11530' 115 15'

O ; 10 MILES
O 5 10 KILOMETERS
l

Figure 3. Residual aeromagnetic total-field intensity map of Railroad ValleyGrant Range region, east-central Nevada. Color
interval 10 nT. Symbols are the same as for figure 2.
 BASEMENT STRUCTURE IN EAST-CENTRAL NEVADA
gravity gradients, which nearly bisects it. Therefore, the
range-front fault must be listric with respect to, or termi
nate at, a low-angle structure at or near the surface of the
magnetic body. This structure may be the low-angle nor
mal fault north of the Troy pluton that is interpreted from
seismic data. Mississippian Chainman Shale, a source rock
for the hydrocarbons (Barker and Peterson, 1991), could
locally have served as the zone of detachment. The ques
tion of whether the low-angle fault passes beneath the
Grant Range at depth or connects with a low-angle fault
exposed on the west side of the Grant Range (as inferred
by Lund and others, 1991, and Potter and others, 1991) is
not likely to be resolved from the potential-field data.
Because no strongly magnetic Paleozoic or Late Prot
erozoic strata are known from geologic mapping in the
region, possible sources of the broad aeromagnetic anom
aly are limited to the Precambrian igneous-metamorphic
crystalline basement complex and Phanerozoic intrusive
bodies. The most likely source is considered to be a Phan
erozoic granitic batholith. Granitic plutons are exposed in
the southwestern White Pine Range, in the southern Grant
Range, and in the northern Quinn Canyon Range, and, as
noted above, quartz monzonite has been penetrated in sev
eral drill holes in Railroad Valley. All of these occur
rences are spatially associated with the broad anomaly,
although typical rock of the Troy pluton (two-mica gran
ite) is known to be only very weakly magnetic (M.R. Hud
son, oral commun., 1992). Potassium-argon ages for
biotite of the plutons range from 23 to 36 Ma, according
to Kleinhampl and Ziony (1985), who cite Schilling
(1965) and Armstrong (1970). Yet a 70-Ma minimum age
of crystallization has been determined for the Troy pluton
by Fryxell (1988) on the basis of whole-rock Rb/Sr analy
sis. These apparently contradictory results can be recon
ciled by postulating that the aeromagnetic anomaly source
is a composite body consisting of weakly magnetic granite
that was emplaced during Late Cretaceous time and
strongly magnetic monzonitic rock emplaced during an
episode of renewed heating and magmatism in the
Tertiary.
The batholith, if present, must have had a major role
in the tectonic evolution of the Railroad ValleyGrant
Range region. Initial emplacement would have occurred
during a late stage of synkinematic regional metamor
phism that is recorded in lower Paleozoic rocks of the
southern Grant Range (Kleinhampl and Ziony, 1985).
Renewed magmatism in Oligocene-Miocene time may
have contributed to arching of the range, as well as to high
heat flow in Railroad Valley. Also, a structural parallelism
indicates that the batholith may have influenced the con
figuration of youthful extensional structures, such as the
Railroad Valley graben.
Some implications of the batholith for mineral
resource exploration have been discussed in detail else
where (Lund and others, 1987, 1988). Of special interest
29
are the prominent anomaly crestal areas north of the Troy
pluton and north of the Shell No. 1 discovery hole (fig. 3).
Such highs superimposed on the broad positive anomaly
could be expressions of structural relief, or locally
enhanced magnetization, or both. (A third prominent high
on the broad anomaly, centered 15-km west of Currant at
the north edge of the map, is much more intense and
sharply bounded than the other two and could be due to a
buried mafic volcanic complex. Unlike the others, it has
no apophysis beneath Paleozoic rocks.) If these highs are
produced by intrusive rock, they flag potential loci of
hydrothermal alteration and mineralization. Moreover, the
gravity data strongly suggest that much of the area
beneath the two aeromagnetic highs is pedimented bed
rock and is therefore possibly accessible for exploitation,
should mineral deposits be proved.
REFERENCES CITED
Armstrong, R.L., 1970, Geochronology of Tertiary igneous rocks,
eastern Basin and Range Province, western Utah, eastern Ne
vada, and vicinity: Geochimica et Cosmochimica Acta, v. 34,
p. 203232.
Barker, C.E., and Peterson, J.A., 1991, Burial history of the Missis
sippian Chainman Shale and the Eocene Sheep Pass Forma
tion, Railroad and White River Valleys, eastern Nevada, in
Flanigan, D.M.H., Hansen, M., and Flanigan, T.E., eds., Geol
ogy of White River Valley, the Grant Range, Eastern Railroad
Valley and Western Egan Range, Nevada: Reno, Nevada
Petroleum Society, 1991 Field Trip Guidebook, p. 3746.
Bortz, L.C., and Murray, D.K., 1979, Eagle Springs oil field, Nye
County, Nevada, in Newman, G.W., and Goode, H.D., eds.,
1979 Basin and Range Symposium: Denver and Salt Lake
City, Rocky Mountain Association of Geologists and Utah
Geological Association, p. 441454.
Cordell, L.E., Keller, G.R., and Hildenbrand, T.G., 1982, Bouguer
gravity map of the Rio Grande rift, Colorado, New Mexico,
and Texas: U.S. Geological Survey Geophysical Investigations
Map GP-949, scale 1:1,000,000.
Effimoff, I., and Pinezich, A.R., 1986, Tertiary structural develop
ment of selected basinsBasin and Range Province, north
eastern Nevada, in Mayer, L., ed., Extensional Tectonics of the
Southwestern United StatesA perspective on Processes and
Kinematics: Geological Society of America Special Paper 208,
p. 3142.
Foster, N.H., 1979, Geomorphic exploration used in the discov
ery of the Trap Spring field, Nye County, Nevada, in New
man, G.W., and Goode, H.D., eds., 1979 Basin and Range
Symposium: Denver and Salt Lake City, Rocky Mountain
Association of Geologists and Utah Geological Association,
p. 477488.
Foster, N.H., Vreeland, J.H., and Dolly, E.D., 1985, Basin and
Range seismic profiles in Railroad Valley, Nye County, Ne
vada, in Gries, R.R., and Dyer, R.C., eds., Seismic Explora
tion of the Rocky Mountain Region: Denver, Rocky
Mountain Association of Geologists and Denver Geophysi
cal Society, p. 283288.
30 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Fryxell, J.E., 1988, Geologic map and descriptions of stratigraphy
and structure of the west-central Grant Range, Nye County,
Nevada: Geological Society of America and Chart Series MCH
064, scale 1:24,000, 16 p.
Garside, L.J., Hess, R.H., Fleming, R.C, and Weimer, B.S., 1988,
Oil and gas developments in Nevada: Nevada Bureau of Mines
and Geology Bulletin 104, 136 p.
Guion, D.J., and Pearson, W.C., 1979, Gravity exploration for
petroleum in Railroad Valley, Nevada, in Newman, G.W., and
Goode, H.D., eds., 1979 Basin and Range Symposium: Denver
and Salt Lake City, Rocky Mountain Association of Geologists
and Utah Geological Association, p. 544556.
Hulen, J.B., Bortz, L.C., and Bereskin, S.R., 1991, Geothermal pro
cesses in evolution of the Grant Canyon and Bacon Flat oil res
ervoirs, Railroad Valley, Nye County, Nevada, in Flanigan,
D.M.H., Hansen, M., and Flanigan, T.E., eds., Geology of
White River Valley, the Grant Range, Eastern Railroad Valley
and Western Egan Range, Nevada: Reno, Nevada Petroleum
Society, 1991 Field Trip Guidebook, p. 4754.
Kleinhampl. F.J., and Ziony, J.I., 1985, Geology of northern Nye
County, Nevada: Nevada Bureau of Mines and Geology Bulle
tin 99A, [including geologic map, scale 1:250,000), 172 p.
Lund, Karen, Beard, L.S., Blank, H.R., Jr., Hofstra, A.H., and
Hamilton, M.M., 1988, Mineral resources of the Riordens Well
Wilderness Study Area, Nye County, Nevada: U.S. Geological
Survey Bulletin 1731H, 16 p.
Lund, Karen, Beard, L.S., and Perry, W.J., Jr., 1991, Structures of
the northern Grant Range and Railroad Valley, Nye County,
NevadaImplications for oil occurrences, in Flanigan,
D.M.H., Hansen, M., and Flanigan, T.E., eds. Geology of
White River Valley, the Grant Range, Eastern Railroad Valley
and Western Egan Range, Nevada: Reno, Nevada Petroleum
Society, 1991 Field Trip Guidebook, p. 16.
Lund, Karen, Nash, J.T., Beard, L.S., Blank, H.R., Jr., and Tutin,
S.E., 1987, Mineral resources of the Blue Eagle Wilderness
Study Area, Nye County, Nevada: U.S. Geological Survey
Bulletin 1731D, 19 p.
Potter, C.J., Grow, J.A., Lund, K., Perry, W.J., Jr., Miller, J.J., and
Lee, M.W., 1991, Comparison of basin geometry and faulting
styles along a regional seismic-reflection profile from Railroad
Valley to Lake Valley, Nevada, in Flanigan, D.M.H., Hansen,
M., and Flanigan, T.E., eds., Geology of White River Valley,
the Grant Range, Eastern Railroad Valley and Western Egan
Range, Nevada: Reno, Nevada Petroleum Society, 1991 Field
Trip Guidebook, Reno, p. 5960.
Schilling, J.H., 1965, Isotopic age determinations of Nevada rocks:
Nevada Bureau of Mines and Geology Report 10.
Vreeland, J.H., and Berrong, B.H., 1979, Seismic exploration in
Railroad Valley, Nevada, in Newman, G.W., and Goode, H.D.,
eds., 1979 Basin and Range Symposium: Denver and Salt Lake
City, Rocky Mountain Association of Geologists and Utah
Geological Association, p. 557572.
 CHAPTER C
MAGNETIC INTERPRETATION OF A
MINERALIZED SHEAR ZONE IN THE
SOUTHERN PART OF THE
IDAHO BATHOLITH
By M. DEAN KLEINKOPF'
In the northwestern part of the Hailey 1x2 quadran
gle, located in the southern part of the Idaho batholith,
magnetic-anomaly data that is displayed in shaded relief
reveal a conspicuous northeast-trending linear belt of steep
magnetic gradient (fig. 1). This belt is interpreted as a
buried shear zone that may have influenced emplacement
of mineral deposits in this area.
Analyses of magnetic-anomaly data were made as
part of a multidisciplinary team effort under CUSMAP
(Conterminous United States Mineral Assessment Pro
gram) to assess the mineral resource potential of the
Hailey and western Idaho Idaho Falls 1x2 quadrangles.
The aeromagnetic-anomaly data are applicable for direct
detection of mineralized features and are applied in con
junction with other geophysical, geological, and geochem
ical investigations to delineate major fault and shear
zones, map buried intrusive complexes, and identify local
ities that might be hydrothermally altered or mineralized.
The bulk of the rocks of the Cretaceous Idaho
batholith are biotite granodiorites or two-mica granites
(Kiilsgaard and Lewis, 1985; Lewis and Kiilsgaard, 1991).
The batholithic terrane exhibits exceptionally uniformly
low magnetization and shows few distinctive anomalies,
except for those associated with minor occurrences of
diorite intrusions and border facies of tonalite, which are
more strongly magnetized than the main mass of the
batholith. The Tertiary granitic batholiths and stocks are
considerably more magnetic than the granitic rocks of the
Cretaceous Idaho batholith. The highest magnetic values
on the map (fig. 1) correlate spatially with high-elevation,
"U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
v,
Tertiary, epizonal plutons such as the granitic intrusions of
the Sawtooth batholith (S); the stock at Steele Mountain
(SM); and the Sheep Mountain batholith (SHM). The
magnetic-anomaly data indicate that these composite intru
sive complexes are more extensive in the subsurface than
in outcrop. The belt of steep magnetic gradient is nearly
continuous along the southeastern side of the Sawtooth,
Steele Mountain, and Sheep Mountain intrusions.
The southern part of the Idaho batholith is broken
into a series of rhombic blocks (Bennett and Lewis, 1989)
resulting from intersections of northeast- and northwest
trending faults. Northeast-trending faults parallel the trans
Challis fault system (TCFS), which is related to Eocene
extension. Northwest-trending faults relate to basin-and
range extension in the Miocene. Many of the faults are
reflected in the magnetic-anomaly data as contour align
ments and deflections of contours across the magnetic
grain. In particular, the belts of steep magnetic gradient
often are seen to correlate with faults or postulated faults
and shear zones. The prominent example in the study area
(fig. 1) is the northeast-trending belt of steep magnetic
gradient that occurs near Atlanta and is interpreted to
reflect a buried shear zone, probably part of the trans
Challis fault system (TCFS). Other evidence supports the
interpretation that the belt of steep magnetic gradient may
be a tectonic feature. Two regional basin-and-range faults,
the Deer Park (DPF) and Montezuma (MF) faults and cor
relative magnetic trends, systematically change strike from
south-southeast north of the belt of steep magnetic gradi
ent to southeast on the south side of the belt (fig. 1).
Northeast of Atlanta, a number of gold deposits are
clustered along the belt of high magnetic gradient (fig. 1).
Additional undiscovered mineral deposits may occur along
the belt to the northeast and southwest of Atlanta. If the


noetXrna2otle-qmhriuawpngetdrlnseiatzsneigrtldcnye of
Map
1.
1Figure
Idaho,
Hailey,
the
of
tpart
residual,
gashowing
relief;
shaded
in
ianomalies
geology; zone;
pshear
metals
aprecious
sources
For
ddeposits.
geology
(1989);
others
and
Kleinkopf
see
data
anomaly
min
ieosrltoyrmiaebgtunatelionc

MILES
15
O
10
5
KLOMETERS
O
15
5
10
AREA
STUDY
of
Location
Porecious
po
or
metal
deposit
lymetal ic Total-iO
shown
is
field
anomaly
magnetictensity'
n3
batholith
Mountain
SHM-Sheep Trans-Chal is
TCFS
system
fault
Mountain
Steele
at
SM-Stock >
(
terrane
batholith
Idaho
Cretaceous
|
Miocene
.
terrane
rock
volcanic
+Eocene
* .Eocene
terrane
rock
volcanic terrane batholith
rock
intrusive
-

SSawtooth terrane
shale
black
Paleozoic
Sss
DPFDeep fault
fault
Park MFMontezuma

Interpreted
==
zone
shear
intervals)
(nanoTesla
50-nt
in
ATION
Explan

Major
-
fault

--

116
114

(1989).
Johnson
and
Worl
from
modified
deposits
eral

11600'

44 43
 MAGNETIC INTERPRETATION OF SHEAR ZONE, SOUTHERN IDAHO BATHOLITH
trend is part of the TCFS, it may hold potential for
extensive epithermal gold mineralization, which is charac
teristic of the TCFS (Bennett, 1984; Kiilsgaard and others,
1986). In order to study spatial relationships between the
geophysical anomalies and the occurrence of precious
metals and polymetallic deposits, a generalized summary
of plots taken from Worl and Johnson (1989) was pre
pared that combines distribution of precious-metal and
polymetallic-vein, stockwork, and skarn deposits. The
deposit types correspond to mineral deposit models, or
modifications of models, found in Cox and Singer (1986).
In summary, new potential areas for undiscovered
mineral resources may be located northeast and southwest
of Atlanta along the belt of steep magnetic gradient. For
detailed mineral resource exploration that might be consid
ered in the future, closely spaced magnetic data collected
by aircraft or on the ground may be used to detect seg
ments of fault zones or granitic intrusions that are hydro
thermally altered and possibly mineralized. Detection is
based on the assumption that the magnetic susceptibilities
of the hydrothermally altered areas would be perceptibly
reduced relative to the unaltered parts of the features.
REFERENCES CITED
Bennett, E.H., 1984, The trans-Challis fault zone: A major crustal
discontinuity in central Idaho (abs.]: Geological Society of
America Abstracts with Programs v. 16, no. 6, p. 442.
33
Bennett, E.H., and Lewis, R.S., 1989, The geology and ore depos
its of the Idaho batholith in the Hailey 1X2 quadrangle, Ida
ho, in Winkler, G.R., and others, eds., Geology and Mineral
Deposits of the Hailey and Western Idaho Falls 1X2 Quad
rangles, Idaho: U.S. Geological Survey Open-File Report
89639, p. 2632.
Cox, D.P, and Singer, D.A., eds., 1986, Mineral deposit models:
U.S. Geological Survey Bulletin 1693,379 p.
Kiilsgaard, T.H., and Lewis, R.S., 1985, Plutonic rocks of Creta
ceous age and faults of the Atlanta Lobe of the Idaho batholith,
in McIntyre, D.H., ed., Symposium on the Geology and Min
eral Deposits of the Challis Quadrangle, Idaho: U.S. Geologi
cal Survey Bulletin 1658B, p 2942.
Kiilsgaard, T.H., Fisher, F.S., and Bennett, E.H., 1986, The trans
Challis fault system and associated precious metal deposits,
Idaho: Economic Geology, v.81, p. 721724.
Kleinkopf, M.D., Bankey, Viki, and McCafferty, A.E., 1989, Grav
ity and magnetic anomaly patterns applied to mineral resource
exploration, Hailey 1X2 quadrangle, Idaho, in Winkler,
G.R., and others, eds., Geology and Mineral Deposits of the
Hailey and Western Idaho Falls 1x2 Quadrangles, Idaho:
U.S. Geological Survey Open-File Report 89639, p. 2632.
Lewis, R.S., and Kiilsgaard, T.H., 1991, Eocene plutonic rocks in
south-central Idaho: Journal of Geophysical Research, v. 96,
no. B8, p. 1329513311.
Worl, R.G., and Johnson, K.M., 1989, Geologic terranes and min
eral deposit types in the Hailey and western Idaho Falls 1X2
quadrangles, Idaho, in Winkler, G.R., and others, eds.,
Geology and Mineral Deposits of the Hailey and Western Ida
ho Falls 1X2 Quadrangles, Idaho: U.S. Geological Survey
Open-File Report 89639, p. 316.
 CHAPTER D
ROLE OF LITHOSPHERIC FLEXURE AND
PLATE CONVERGENCE IN THE GENESIS
OF SOME MISSISSIPPI-VALLEY-TYPE
ZINC DEPOSITS IN THE APPALACHIANS
By DWIGHT C. BRADLEY!
INTRODUCTION
Ordovician-hosted Mississippi-Valley-type (MVT)
zinc deposits in the Appalachian fold-thrust belt and fore
land basin (fig. 1) generally are regarded as having formed
from basinal brines that migrated cratonward from the
orogen or its foreland basin (Oliver, 1986; Bethke and
Marshak, 1990; Kesler and Van der Pluijm, 1990). The
age of mineralization, however, has been subject to much
debate, and as a consequence, mineralization has been
linked only in very general terms to Appalachian plate
tectonic history. Various researchers have argued that min
eralization occurred during Middle Ordovician (Hoagland
and others, 1965; Callahan, 1968; Bradley, 1989a), Late
Ordovician to Devonian (Kesler and Van der Pluijm,
1990), and late Paleozoic (Hearn and others, 1987).
Depending on the age of mineralization, any connections
between metallogenesis and plate-tectonic events would be
sought in either the Alleghanian, Acadian, or Taconic
orogenies.
One aspect of ore deposit geology that can be linked in
detail to Appalachian plate tectonics is the geologic history
of host strata. In this paper, I discuss geologic effects of the
Taconic arc-passive margin collision (Bradley and Kusky,
1986; Bradley, 1989b) that now control, in part, the distri
bution of many MVT deposits. Two related processes are
of particular interest. Lithospheric flexure caused forebulge
uplift and normal faulting; both were instrumental in
ground preparation. Flexure also drove subsidence of a
foredeep basin that probably expelled mineralizing fluids.
"U.S. Geological Survey, Branch of Alaskan Geology, 4200 Univer
sity Drive, Anchorage, AK 99508.
Plate convergence drove the Taconic thrust belt over the
foredeep and forebulge and caused these belts to migrate
toward the craton. This paper documents an association
noted by Mitchell (1985) and Kaiser and Ohmoto (1988)
between economic MVT deposits and flexure-induced fea
tures related to an arc-passive margin collision. Explora
tion for MVT deposits in some frontier regions may benefit
from similar plate-tectonic considerations.
THE TACONIC OROGENY AND
LATER TECTONIC EVENTS
The early Paleozoic evolution of the Appalachians
was a Wilson cycle of opening, then partial closing, of the
Iapetus ocean basin. Late Proterozoic rifting of Protero
zoic basement was followed by Cambrian and Early
Ordovician subsidence of a passive margin with two com
ponents: (1) a miogeoclinal wedge of mainly platformal
carbonate rocks up to 3.5 km thick, which now host the
MVT deposits, and (2) farther east, a much thinner apron
of slope and rise facies, now preserved in far-traveled
thrust sheets.
This passive margin endured until its collision with an
arc complex during the Ordovician Taconic orogeny (Row
ley and Kidd, 1981; Bradley, 1989b) (fig. 2). The earliest
stratigraphic record of collision is an Early to Middle
Ordovician influx of orogen-derived flysch, which con
formably overlies older slope and rise deposits. The arrival
of flysch from the east was synchronous with gentle uplift
of the platform. Uplift, therefore, can be viewed as a distal
effect of collision. In the carbonate rocks that now host the
MVT deposits, this uplift is recorded by an unconformity:
36 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
ORDOVICIAN-HOSTED MVT DEPOSITS
IN THE APPALACHIAN OROGEN
cENTRA- TENNESEE
g *s.
Figure 1. Map of Appalachian orogen showing main MVT deposits or districts (black dots). Taconic foredeep (foreland basin) formerly
occupied stippled area but has been partially eroded away. Late Paleozoic Alleghanian deformation front (open barbs) lies cratonward of
Taconic thrust front (solid barbs) in central and southern Appalachians.
the post-Knox unconformity in the southern Appalachians
(Mussman and Read, 1986), the post-Beekmantown uncon
formity in Pennsylvania and New York (Lash, 1989), and
the St. George unconformity in Newfoundland (Knight and
others, 1991). Mussman and Read (1986) reported ero
sional relief in Virginia up to 140 m; they also reported
sinkholes and caves extending down to 65 m below the
unconformity. Knight and others (1991) reported erosional
relief in Newfoundland up to 50 m and karst features to
depths of 120 m. Comparable erosional relief and depth of
karst features have been reported from all of the MVT dis
tricts discussed below. The unconformity has been widely
attributed to passage of a forebulge (Jacobi, 1981; Rowley
and Kidd, 1981; Bradley and Kusky, 1986; Lash, 1989;
Knight and others, 1991) because uplift was synchronous
with flysch sedimentation farther east and was followed
immediately by collision-induced drowning of the platform
and progradation of flysch. The unconformity and subse
quent events vary in age along the orogen (Bradley,
1989b), indicating that eustatic fall in sea level cannot have
been the sole cause of emergence.
Overlying the unconformity is an upward-deepening,
diachronous succession of carbonates, black shales, and
turbidites. The shales and turbidites comprise the synoro
genic clastic fill of the Taconic foredeep, a basin that
migrated cratonward through time across the former conti
nental shelf (Rowley and Kidd, 1981; Bradley, 1989b).
Normal faulting, induced by lithospheric flexure, accom
panied foredeep subsidence (Bradley and Kidd, 1991).
Paleogeographic relations across the Taconic foredeep are
summarized in figure 3. This cross section depicts a quasi
steady-state configuration that migrated cratonward
through time; it is appropriate for a given transect until
thrusting ceased.
Two younger collisionsthe Acadian in the Devo
nian (Osberg and others, 1989) and the Alleghanian in the
Mississippian to Permian (Hatcher and others, 1989)had
markedly different effects on Taconic paleogeographic
elements in the northern versus the southern parts of the
&
c
O 200 kiLor ETERs
-
s
N "*Reog
A
DANIELS 22
vo.
orogen. In the northern Appalachians, the Taconic thrust
front is the cratonward limit of significant shortening in
the northern Appalachians (fig. 1); Ordovician foredeep
deposits west of the Taconic thrusts are subhorizontal. In
the northern Appalachians, the Taconic front is accurately
located by fossiliferous wildflysch deposits (olistostromes
and tectonic melange) that immediately flank far-traveled
allochthons. In the southern and central Appalachians, late
Paleozoic thrusts and folds deform a broad region to the
west of the Taconic thrust front; this region includes
Ordovician foredeep strata as well as MVT deposits in
older carbonate rocks. In the central and southern Appala
chians, the position of the Taconic thrust front is not
known precisely, but it can be estimated by assuming that
Ordovician facies belts in the foredeep (black shales, tur
bidites) were the same width and distance from the thrust
front as in the northern Appalachians (Bradley, 1989b).
ORDOVICIAN-HOSTED MVT
DEPOSITS
Ordovician-hosted MVT deposits in the Appalachian
fold-thrust belt and (or) foreland basin (Clark, 1987) are
hosted by strata that were laid down only a few million
years before the Taconic orogeny. Mineralized areas dis
cussed below include the Elmwood and nearby deposits of
the Central Tennessee district, the MascotJefferson City
and Copper Ridge deposits of the East Tennessee district,
the Timberville deposit in Virginia, the Friedensville
deposit in Pennsylvania, and the Daniel's Harbour deposit
in Newfoundland. The deposits can be divided into three
groups on the basis of the state and age of deformation of
the host strata: (1) flat-lying strata, (2) deformed strata
that were flat-lying until the late Paleozoic Alleghanian
orogeny, and (3) deformed strata that were overthrust dur
ing the Taconic orogeny.
 ROLE OF FLEXURE AND CONVERGENCE IN APPALACHIANZINC DEPOSITS
marrow/ng
Ayass/ve- OC62O/7
marg/n
North America
... --- * * * normal
fau/fs
fore deep ....
Figure 2. Schematic model of Taconic orogeny as an arc-passive margin collision. Plate convergence shown in top part of figure inevi
tably led to collision illustrated in bottom part of figure. Rise of Taconic forebulge and subsidence of Taconic foredeep were secondary
effects of collision.
DEPOSITS IN FLAT-LYING STRATA
MVT deposits of this group are exemplified by the
Central Tennessee district, on the Nashville dome (fig. 1).
Lower Ordovician carbonate rocks of the Knox Group
host the mineralization, which was controlled by solution
collapse breccias formed during karst formation along the
post-Knox unconformity (Gaylord and Briskey, 1983).
The Nashville dome and Cincinnati arch together comprise
part of an elongate uplift that flanks the southern and cen
tral Appalachian thrust belt (Quinlan and Beaumont,
1984). During Ordovician time, the axis of the Nashville
dome was located about 65 km east of the Devonian axis
(Stearns and Reesman, 1986). The position next to the
orogenic belt, the magnitude of uplift, and the cratonward
shift of the arch between Ordovician and Devonian times
together suggest that the Nashville domeCincinnati arch
began as a flexural effect of Taconic loading (Quinlan and
Beaumont, 1984).
Figure 3 shows the reconstructed position of the Cen
tral Tennessee district as projected onto the schematic
cross section through the Taconic foredeep at the time
when Ordovician thrusting ceased. This distance was esti
mated by summing (1) the present distance from the Elm
wood deposit on the Nashville dome to the Chattanooga
thrust, (2) the pre-late-Paleozoic palinspastic distance from
the Chattanooga to the Knoxville thrust (Woodward and
37
//orma/
accref/onary
fau/fs
prism
O
|-l-l
1 OO K|LONME TERS
Gray, 1985), and (3) the palinspastic distance (Walker and
others, 1983) from the Knoxville thrust to the eastern mar
gin of the Sevier Basin (a local depocenter of the Taconic
foredeep) (Shanmugam and Walker, 1980). At the close of
the Taconic orogeny, the Central Tennessee MVT district
lay about 275 km cratonward of the thrust front.
DEPOSITS IN STRATA DEFORMED DURING
THE LATE PALEOZOIC
MVT deposits of this group include MascotJefferson
City, Copper Ridge, and Timberville. These deposits are
hosted in folded and thrust-faulted Ordovician carbonate
rocks within the late Paleozoic southern Appalachian fold
thrust belt. As in the Central Tennessee district, mineral
ization is localized by karst-related solution-collapse brec
cias that lie immediately below the post-Knox
unconformity. In the East Tennessee district, detrital
sphalerite sands in solution-collapse deposits, described by
Kendall (1960) and Hoagland and others (1965), have dips
comparable to the bedding dips of the host strata, proving
that mineralization preceded Alleghanian deformation.
When thrusting ended in the Ordovician, the host
strata of the MascotJefferson City, Copper Ridge, and
Timberville deposits were located on the undeformed cra
tonic flank of the Taconic foredeep (fig. 3). The positions


Graedlonevigcreiaongl-rihaozpshetidc buried
3.
OFigure
of
Positions
foredeep.
Taconic
through
pcross
to
relative
section
onto
projected
are
deposits
MVT
when
elements cFeorcnitveodnerigscevainlcley
(plate
occur
now
deposits
MVT
where
locations
Two
tceased.
were
Harbour)
Daniel's
and
the
during
sheets
thrust
Taconic
beneath
KILOMETERS
100
-
* - * *.
> *
*
Overriding
c..
Tennessee
ss *
*
*,
*,
* -
* *,
*
*
plate *,*,
-
-
-
-
-
THRUST *,
"Fri-edensvil e
-
-
-
-
-
LOAD * *, -
-
* * * *, -
-
-
w"
- -
* * -
* - -
-
~ ^ -
-
ORE
F
C
GENI
RO
NO
SY
POSITS
DE
DEEP
* f
*".*.*.bul
*.*.*.*.*.*.*.*
a
- -
-
- ~ -
-
^ - *
-
-

*
w
orebulge
Harbour
<^^^^^^ -- -
-
-
50

-
-

>|< -

-
-

-
-
-
-
* -

level
sea -

-
- turbidites0O
Graywacke 2
-
-
*
-

-
-
[H]
rocks
Carbonate
-
migration
-
*-
*-
-

FOREDEEP
FOREBULGE -

-
EXPLANATION
=-| -
Shale
mineralization
7
window -

MVT
possible -

- - - -
* * - - - -
* - - - - - -
- * * * - - - -
- - - - - -
- ~ * * - - - -
- - - - - -
- * * * - - - -
~...~...~- - - - - - -
- a
- - * * * - - - -
-- - - - - - -
- - * * * - - - -
~ * - - - - - -
- * * * - - - -
- - - - - -
- * * * - - - -
- - - - - -
* * * - - - -
>|< - - - - -
* * * - - -
- - - - -
-
-
- platform
margin
passive
of
rocks
* * * - - -
- - - - -
* * * - - -
migration
-
-
*~
*-
- -
- - - - - -
* * * - - -
|< - - - - -
* * * - - -
- clastic
minor
and
Carbonate
- - - - - -
* * * * - - -
- - - - - -
KS
O
R
C
GENI
RO
PRE-O
- * * * * - - -
- - - - - -
- * * * * - - -
- - - - - -
- - * * * * - - - -
- - - - - -
* * * * - - - -
- - - - - -
* * * * - - - -
- - - - - -
* * * * - - - -
- - - - - -
* * * * - - - -
- - - - - -
* * * * - - - -
- - - - - -
- * * * * - - - -
- - - - - -
- * * * * - - -
- - - - -
-
-
Precambrian
- * * * * - - - -

- *
- - - - -
* * * - - - -
- basement
- - - - - - -
- * * * * - - - -
- - - - - - -
- * * * * - - - -
- - - - - - -
^^^^^^*. - * * * * - - - -
- - - - - - -
EMERGING - * * * * - - -
- - - - -
* * * - - -
-
-
- * -
SHELF - *
- - - - -
* * * - - - -
-
- - - - - -
- * * * * - - - -
- - - - - -
* * * * - - - -
- - - - - -
* * * * * - - - - Ordovician.
*
*
- - - - -
* * * - - -
- - - - -
* * * - - -
-

*
-
- EEE
- - - - - -
* * - * - - - -
*. *. *. *. *. *. *. *. * - - - - - -
 ROLE OF FLEXURE AND CONVERGENCE IN APPALACHIANZINC DEPOSITS 39

DANIEL'S HARBOUR TECT ONIC FRIEDENS VILLE

SYSTEM SERIES STAGE STRATIGRAPHY INTER PRETATION STRATIGRAPHY

Caradocian final
4.E= Martinsburg Formation
emplacement of
thrusts ====El Jacksonburg
Limestone

--- foredeep
Llandeilian subsidence

Middle
Mainland forebulge D *% IT
- Sandstone unconformity
Ordovician Llanvirnian T l
Table Head h

Group

*~ MVT
Arenigian DE POSIT
I

HE St. George
HELEH Group
L lt. Beekmantown
OWer Tremadocian * passive Group
margin * I

Undivided subsidence : : :
I I

platform ITI

sequence Undivided
Cambrian T: platform
- - - - - - sequence

M.
~~
* * * *
*,
* * * *
S,,, G ill 'J',,, G ille
w" r f f renville * r f r f renvi
Precambrian (r basement
M. * * ^^^^l basement
kssss "S,,,
z r f * f z. z .

Figure 4. Cambrian and Ordovician stratigraphy at Daniel's Harbour and Friedensville deposits. Forebulge unconfor
mity, foredeep deposits, and Taconic thrusting differ in age between the two sections, but the relative order of events is
the same.

of the MascotJefferson City and Copper Ridge deposits
were estimated from a palinspastic cross section of the
Sevier Basin (Walker and others, 1983, p. 705, upper sec
tion). The position of Timberville was estimated from a
palinspastic cross section through northern Virginia (Read,
1980, p. 1582, section A). Estimated palinspastic distances
to the Taconic thrust front are 55 km for the MascotJef
ferson City deposit, 85 km for the Timberville deposit,
and 115 km for the Copper Ridge deposit.
DEPOSITS INSTRATA TECTONICALLY
BURIED DURING THE ORDOVICIAN
The Daniel's Harbour and Friedensville deposits (fig.
1) have the most intriguing connection to Taconic paleo
geography. The Daniel's Harbour deposit in western New
foundland is hosted by Lower Ordovician platform
carbonates of the St. George Group, and the ore bodies
occur immediately beneath a major karsted unconformity
at the top of the St. George Group (fig. 4). Ore postdates,
but is spatially associated with, dolomitic breccia bodies
and Ordovician sinkholes formed by karst processes
(Lane, 1984). These breccia bodies are elongate parallel to
orogenic strike, and they appear to have nucleated along
orogen-parallel normal faults that were active during and
immediately after forebulge uplift (Knight and others,
1991). Lane (1989) reported rotated geopetal sediments
in ore-stage cavities and displaced ore bodies along faults
as evidence that mineralization preceded contractional
deformation. Host rocks at the Daniel's Harbour deposit
are part of the gently dipping, parautochthonous platfor
mal sequence that was tectonically buried beneath the
Humber Arm allochthon during the Middle Ordovician
(Cawood and Williams, 1988). When thrusting ceased, the
deposit lay at least 10 km to the east of (behind) the thrust
front. In figure 3, the position of the Daniel's Harbour
deposit was estimated by extrapolation 20 km along strike
to the south, where the Humber Arm allochthon has not
been eroded away and extends at least as far west as the
coast. According to Cawood and Williams (1988), the
parautochthonous ore-hosting strata at the Daniel's Har
bour deposit were not penetratively deformed during
emplacement of the allochthon; gentle bedding dips and
40 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

=s: 75
Hamburg ##S- --

----------- *F. Taconic T

L-L-L-L LI -
foreland-bash
- - - - - - - - - - -

i HTTTTTTTTTH I " -- -R THERTHER - - - - - - - deposits - - -

|-|--|--|--|--|--|--|--|--|--|--T-r-->r-H-I-I-I-I-I-R-E
I I I I I |--|--|--|--|--|--|--|--|--|--|--|--|--|--|--RP-e

N. M
- - --T-I-T-I-T-I-T-I-T-I-T-T-L-T-I-T-I-T-I-T-T- E 41 o
*... - - -
76 2 *41-r-
-
-
-
-
- - - - - -

-
- -

- - - - - - - - *
- - - -

-
*
-
-
-
-
-
z *
*
z
*
-
*
-
- - - -
-
-
-
-
-
-
-
-
-
- - - - - - -
-

- - - - - - - - - - -
- - - ~ *
- - - - - '
- - - - - -
7.

- -
*
Friedensville -
* - ^^*. * ~ *-- ~~ - - - - -
- -

MVT deposit
~~~ *
- a z -
*.*.*.*.*.*, O 25 50 KILOMETERS
- *.*.*.*.*.
- - - - -
1 l
- - - -
- - - - - -

75
EXPLANATION

Middle Ordovician
Mesozoic rift facies carbonate rocks

Ordovician and Cambrian

. . . . .. . . .. . . . Carboniferous molasse facies slope and rise facies

Lower Ordovician and Cambrian

Devonian and Silurian, carbonate rocks
mainly molasse facies

Middle Ordovician flysch 'ssss's] Precambrian basement

Figure 5. Generalized geologic map of Friedensville, Pennsylvania, and vicinity. Friedensville was tectonically buried during the
Taconic orogeny by the Hamburg allochthon, which advanced toward N. 18 W. (large arrow; Lash and Drake, 1984). Present
erosional front of Hamburg klippe lies about 20 km cratonward of Friedensville.

high-angle reverse faults in the Daniel's Harbour area date and faulted along with enclosing strata, and hence, miner
from the Devonian Acadian orogeny. alization predated deformation (Callahan, 1968, p. 103).
The MVT deposit at Friedensville, Pennsylvania (fig. As at the Daniels Harbour deposit, host rocks at the
5) is hosted by platformal carbonate rocks of the Lower Friedensville deposit are part of the parautochthonous plat
and Middle Ordovician Beekmantown Group. The ore formal sequence that was tectonically buried beneath the
occurs in tabular, solution-collapse breccias that lie just Hamburg allochthon during the Middle Ordovician (Lash
below the post-Beekmantown unconformity (fig. 4). Stra and Drake, 1984). When thrusting ceased, the Friedens
tabound ore bodies at the Friedensville deposit are folded ville deposit lay about 20 km to the east of (behind) the
 ROLE OF FLEXURE AND CONVERGENCE IN APPALACHIANZINC DEPOSITS
thrust front. In figure 3, the position of Friedensville was
estimated by extrapolation 20 km along strike to the south,
where the Ordovician thrust front is marked by a belt of
Ordovician wildflysch immediately west of the leading
edge of the Hamburg klippe (Root and MacLachlan,
1978). Ore-hosting strata at Friedensville are folded and
cut by several generations of faults. Deformation is either
Taconic (Lash and Drake, 1984, p. 32) or Alleghanian.
DISCUSSION
5.
Events leading up to and including the Taconic orog
eny in the Appalachians provided many of the conditions
necessary for MVT mineralization. During the Cambrian
and Early Ordovician in the Appalachians, carbonate host
rocks were deposited along a thermally subsiding passive
margin in a seaward-thickening miogeoclinal wedge. Dur
ing the earliest stages of collision, uplift of the platform at
a forebulge resulted in widespread ground preparation, and
much ground was prepared in this fashion because plate
convergence caused the forebulge to migrate. Passive-mar
gin subsidence and collision at low latitudes is a key fac
tor because forebulge uplift exposed carbonate rocks to
karst formation. In contrast, karsted forebulge unconformi
ties are unlikely to form during high-latitude collisions,
such as the Brooks Range orogeny, involving a clastic
rather than a carbonate platform. Collision drove the sub
sidence of a shale- and turbidite-filled foreland basin and
also caused the basin to migrate cratonward. The migra
tion of facies belts capped the permeable carbonate plat
form with an impermeable shale aquitard. Evidence from
the Daniels Harbour deposit suggests that flexure-induced
normal faults dating from early collision played a role in
ground preparation. Normal faults also appear to have
helped to focus mineralizing fluids in the Northern Arkan
sas MVT district in the Ouachita foredeep (Bradley and
Kidd, 1991) and the Taconic foredeep of New York (Bra
dley and Kidd, 1991). Mitchell (1985) suggested a similar
fluid-focusing mechanism for the Irish base-metal district
in the Variscan foreland. Flexure-induced normal faults in
other foredeeps, therefore, are promising exploration
targets.
.
When Taconic thrusting ceased, six MVT localities
were distributed across a 300-km-wide belt that stretched
from forebulge to thrust belt (fig. 3): the Central Tennes
mit see district on the forebulge; the Copper Ridge, Timber
|||}| ville, and MascotJefferson City deposits on the cratonic
flank of the foredeep; and the Daniel's Harbour and
Friedensville deposits beneath thrust sheets. A simple
explanation for this distribution is that mineralization
occurred during the Middle Ordovician by foreland basin
dewatering, as discussed below. But if so, then there were
at least two episodes of MVT mineralization in the
41
Appalachian foreland, a second event being necessary to
explain numerous MVT occurrences and small deposits in
the Appalachian foreland in rocks as young as Silurian,
Devonian, and even Pennsylvanian (Clark, 1987).
The case for Middle Ordovician mineralization is
summarized here. The prevailing view of Appalachian
MVT mineralization as a product of foreland basin dewa
tering (Kesler and others, 1988) can be reconciled with the
distribution of deposits in figure 3 only if mineralization
occurred during the Middle Ordovician. The last time the
Daniel's Harbour or Friedensville deposits could have
been connected in any hydrologically simple way to a
foreland basin was immediately before tectonic burial dur
ing the Middle Ordovician. Younger (Acadian and
Alleghanian) foreland basins may have existed to the east
of the Daniel's Harbour and Friedensville deposits, but
any such basins would have been superimposed on the
eroded roots of the Taconic thrust belt and not on the
more deeply buried parautochthon within which the ore
bodies occur. Younger foreland basins certainly existed
west of the Daniel's Harbour and Friedensville deposits
(fig. 5), but they are unlikely to have expelled large quan
tities of fluid toward and beneath the frontal thrusts. Fore
land basin brines could only have migrated through the
Daniels Harbour and Friedensville areas during the brief
(510 m.y.) Middle Ordovician interval between the time
of karst-related ground preparation at the forebulge and
the time of tectonic burial by thrusts. If this scenario is
valid, the 300-km width of the belt of MVT localities
records the end of plate convergencethis halted an ore
deposit production line that had MVT deposits at vari
ous places along it (fig. 3). Each deposit would have
formed on the forebulge or cratonic flank of the foredeep;
post-mineralization plate convergence would have con
veyed the earlier formed deposits away from the relative
position at which they formed, toward or beneath the oro
gen. This orogenic-foreland setting for mineralization dif
fers slightly from the setting proposed for MVT deposits
of the Ozarks in the foreland of the Ouachita orogen
(Leach and Rowan, 1986) and those in the foreland of the
Canadian Rockies (Garven, 1985). In the latter two
regions, mineralizing fluids are believed to have been
driven by gravity across a sediment-filled foreland basin
that had a surface that sloped down toward the craton. In
contrast, Middle Ordovician mineralizing fluids in the
Appalachians would necessarily have migrated across a
foredeep that sloped up toward the craton and had a deep
axial depression (as in fig. 3).
Mineralization of the Ordovician-hosted deposits
could be younger than Middle Ordovician; but if so, two
very different genetic models must apply to six similar
deposits: one mechanism for the Daniels Harbour and
Friedensville deposits, and another for the Central Tennes
see district and the Copper Ridge, MascotJefferson City,
and Timberville deposits. These last four could have been
42 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
formed from Acadian or early Alleghanian foreland-basin
brines sometime after the Middle Ordovician. Post-Mid
dle-Ordovician mineralization at the Daniels Harbour and
Friedensville deposits, however, could only have taken
place beneath the Taconic thrust belt, and the mineralizing
fluids would have been something other than foreland
basin brines.
The distribution of deposits in figure 3, therefore,
highlights an unanswered question in Appalachian metal
logenesis that future studies should address. One possibil
ity is that there were two or more mineralizing events: an
earlier one during the Middle Ordovician and a later one
during the late Paleozoic. The other possibility is that
there was a single late Paleozoic mineralizing event but
two very different environments of mineralization: sub
thrust and orogenic foreland. Future research should indi
cate whether the Daniels Harbour and Friedensville
deposits have textural, isotopic, fluid-inclusion, or other
properties that set them apart from the other Ordovician
hosted MVT deposits.
REFERENCES CITED
Bethke, C.M., and Marshak, S. 1990, Brine migrations across North
AmericaThe plate tectonics of groundwater: Annual Re
views of Earth and Planetary Science, v. 18, p. 228315.
Bradley, D.C., 1989a, Mississippi Valley-type Pb-Zn mineraliza
tion during Ordovician arc-continent collision in the Appala
chians: Geological Society of America, Abstracts with
Programs, v. 21, p. A8.
1989b, Taconic plate kinematics as revealed by foredeep
stratigraphy, Appalachian orogen: Tectonics, v. 8,
p. 10371049.
Bradley, D.C., and Kidd, W.S.F., 1991, Flexural extension of the
upper continental crust in collisional foredeeps: Geological So
ciety of America Bulletin, v. 103, p. 14161438.
Bradley, D.C., and Kusky, T.M., 1986, Geologic evidence for rate
of plate convergence during the Taconic arc-continent colli
sion: Journal of Geology, v. 94, p. 667681.
Callahan, W.H., 1968, Geology of the Friedensville zinc mine, Le
high County, Pennsylvania, in Ridge, J.D., ed., Ore Deposits of
the United States, 19331967 [The Graton-Sales Volume, v.
1]: American Institute of Mining, Metallurgy, and Petroleum
Engineers, p. 95107.
Cawood, P.A., and Williams, H.S., 1988, Acadian basement thrust
ing, crustal delamination, and structural styles in and around
the Humber Arm allochthon, western Newfoundland:
Geology, v. 16, p. 370373.
Clark, S.H.B., 1987, Metallogenic map of zinc, lead, and barium
deposits and occurrences in Paleozoic sedimentary rocks, east
central United States: U.S. Geological Survey Miscellaneous
Investigations Series Map I-773, scale 1:2,500,000.
Garven, G., 1985, The role of regional fluid flow in the genesis of
the Pine Point deposit, western Canada sedimentary basin:
Economic Geology, v. 80, p. 307-324.
Gaylord, W.B., and Briskey, J.A., 1983, Geology of the Elmwood
and Gordonsville mines, Central Tennessee zinc district, in
Tennessee Zinc Deposits Field Trip Guide Book: Blacksburg,
Virginia, Virginia Tech Department of Geological Sciences,
Guide Book Number 9, p. 116151.
Hatcher, R.D., Jr., Thomas, W.A., Geiser, P.A., Snoke, A.W.,
Mosher, S., and Wiltschko, D.V., 1989, Alleghanian Orogen,
in Hatcher, R.D., Jr., Thomas, W.A., and Viele, G.W., eds.,
The Appalachian-Ouachita Orogen in the United States: Boul
der, Colorado, Geological Society of America, The Geology of
North America, v. F-2, p. 233318.
Hearn, P.P., Sutter, J.F., and Belkin, H.E., 1987, Evidence for late
Paleozoic brine migration in Cambrian carbonate rocks of the
central and southern Appalachians: Implications for Mississip
pi Valley-type mineralization: Geochimica et Cosmochimica
Acta, v. 51, p. 13231334.
Hoagland, A.D., Hill, W.T., and Fulweiler, R.E., 1965, Genesis of
the Ordovician zinc deposits in East Tennessee: Economic Ge
ology, v. 60, p. 693714.
Jacobi, R.D., 1981, Peripheral bulgeA causal mechanism for the
Lower/Middle Ordovician unconformity along the western
margin of the northern Appalachians: Earth and Planetary Sci
ence Letters. v. 56, p. 245251.
Kaiser, C.J. and Ohmoto, H., 1988, Ore-controlling structures of
Mississippi valley-type mineralization on the North American
midcontinent as products of late Paleozoic convergent plate
tectonism, in Kisvarsanyi, G., and Grant, S.K., eds., North
American Conference on the Tectonic Control of Ore Deposits
and the Vertical and Horizontal Extent of Ore Systems, Pro
ceedings Volume: Rolla, Missouri, University of Missouri
Rolla, p. 424430.
Kendall, D.L., 1960, Ore deposits and sedimentary features, Jef
ferson City mine, Tennessee: Economic Geology, v. 55.
p. 9851003.
Kesler, S.E., Jones, L.M., and Ruiz, J., 1988, Strontium-isotopic
geochemistry of Mississippi Valley-type deposits, east Ten
nessee: Implications for age and source of mineralizing brines:
Geological Society of America Bulletin, v. 100, p. 13001307.
Kesler, S.E., and Van der Pluijm, B.A., 1990, Timing of Mississip
pi Valley-type mineralization: Relation to Appalachian oro
genic events: Geology, v. 18, p. 11151118.
Knight, I., James, N.P., and Lane, T.E., 1991, The Ordovician St.
George unconformity, northern Appalachians: The relation
ship of plate convergence at the St. Lawrence Promontory to
the Sauk-Tippecanoe sequence boundary: Geological Society
of America Bulletin, v. 103, p. 12001225.
Lane, T.E., 1984, Preliminary classification of carbonate breccias,
Newfoundland Zinc mines, Daniel's Harbour, Newfoundland:
Geological Survey of Canada Paper 841A, p. 505512.
Lane, T.E., 1989, Sphalerite/dolomite stratigraphy and the tectonic
origin of an MVT deposit, Daniel's Harbour, western New
foundland, Canada: Geological Society of America, Abstracts
with Programs, v. 21, p. A8.
Lash, G.G., 1989, Middle and Late Ordovician shelf activation and
foredeep evolution, central Appalachian orogen, in Keith,
B.D., ed., The Trenton Group (Upper Ordovician Series) of
Eastern North America: Deposition, Diagenesis, and Petro
leum: American Association of Petroleum Geologists, Studies
in Geology 29, p. 3753.
Lash, G.G., and Drake, A.A., Jr., 1984, The Richmond and Green
wich slices of the Hamburg klippe in eastern Pennsylvania
Stratigraphy, sedimentology, structure, and plate tectonic
 ROLE OF FLEXURE AND CONVERGENCE IN APPALACHIANZINC DEPOSITS
implications: U.S. Geological Survey Professional Paper 1312,
40 p.
Leach, D.L., and Rowan, E.L., 1986, Genetic link between Ouach
ita foldbelt tectonism and the Mississippi Valley-type lead
zinc deposits of the Ozarks: Geology, v. 14, p. 931935.
Mitchell, A.H.G., 1985, Mineral deposits related to tectonic events
accompanying arc-continent collision: Institute for Mining
and Metallurgy Transactions, Section B, v. 94, p. B115B127.
Mussman, W.J., and Read, J.F., 1986, Sedimentology and develop
ment of a passive- to convergent-margin unconformity: Middle
Ordovician Knox unconformity, Virginia Appalachians: Geo
logical Society of America Bulletin, v.97, p. 282295.
Oliver, J., 1986, Fluids expelled tectonically from orogenic belts:
Their role in hydrocarbon migration and other geologic phe
nomena: Geology, v. 14, p. 99102.
Osberg, P.H., Tull, J.F., Robinson, P., Hon, R., and Butler, J.R.,
1989, The Acadian Orogen, in Hatcher, R.D., Jr., Thomas,
W.A., and Viele, G.W., eds., The Appalachian-Ouachita
Orogen in the United States: Boulder, Colorado, Geological
Society of America, The Geology of North America, v.
F2, p. 179232.
Quinlan, G.M., and Beaumont, C., 1984, Appalachian thrusting,
lithospheric flexure, and the Paleozoic stratigraphy of the east
ern interior of North America: Canadian Journal of Earth Sci
ence, v. 21, p. 973996.
43
Read, J.F., 1980, Carbonate ramp-to-basin transitions and foreland
basin evolution, Middle Ordovician, Virginia Appalachians:
American Association of Petroleum Geologists Bulletin, v. 64,
p. 15751612.
Root, S.I, and MacLachlan, D.B., 1978, Western limit of Taconic
allochthons in Pennsylvania: Geological Society of America
Bulletin, v. 89, p. 15151528.
Rowley, D.B., and Kidd, W.S.F., 1981, Stratigraphic relationships
and detrital composition of the medial Ordovician flysch of
western New England: Implications for the tectonic evolution
of the Taconic orogeny: Journal of Geology, v. 89, p. 199218.
Shanmugam, G., and Walker, K.R., 1980, Sedimentation, subsid
ence, and evolution of a foredeep basin in the Middle Ordovi
cian, southern Appalachians: American Journal of Science, v.
280, p. 479496.
Stearns, R.G., and Reesman, A.L., 1986, Cambrian to Holocene
structural and burial history of Nashville dome: American As
sociation of Petroleum Geologists Bulletin, v. 70, p. 143154.
Walker, K.R., Shanmugam, G., and Ruppel, S.C., 1983, A model
for carbonate to terregenous clastic sequences: Geological So
ciety of America Bulletin, v. 94, p. 700712.
Woodward, N.B., and Gray, D.R., 1985, Southwest Virginia, Ten
nessee, and northern Georgia sections, in Woodward, N.B.,
ed., Valley and Ridge Thrust Belt: Balanced Structural Sec
tions, Pennsylvania to Alabama: Knoxville, University of Ten
nessee, Studies in Geology, v. 12.
 CHAPTER E
Rb-Sr ISOTOPIC REDISTRIBUTION AND
CHARACTER OF HYDROTHERMAL FLUIDS
IN THE CATHEART MOUNTAIN
Cu-Mo DEPOSIT, MAINE
By ROBERT A. AYUSO! and NORA. K. FOLEY
ABSTRACT
The Catheart Mountain porphyry Cu-Mo deposit is
composed of hydrothermally altered equigranular grano
diorite and porphyritic dikes that contain as much as 4
weight percent sulfides (pyrite, molybdenite, chalcopyrite,
galena, and sphalerite). Propylitic to potassic alteration
produced a succession of mineral assemblages: biotite
was replaced by chlorite and then by muscovite; plagio
clase by albite + epidote; chlorite by muscovite; K-feld
spar by muscovite; and sphene by rutile. All rocks in the
hydrothermal system can be systematically arranged in
terms of a relative measure of alteration intensity (alter
ation index) that depends on the sum of the total abun
dances of K-feldspar, muscovite, and quartz. Up to 6.9
weight percent K2O, 3,900 ppm Cu, and 320 ppm Mo are
found in rocks with high alteration indices; these rocks
also have lower contents of CaO and Na2O, higher values
of Fe"/Fe and K/Na, and higher abundances of Cu and
Mo relative to rocks with lower alteration indices. Sr con
tents decreased (from about 900 to 50 ppm), and Rb con
tents generally increased to as much as 250 ppm in rocks
with high alteration indices; this resulted in generally
higher values of Rb/Sr (as much as 3.5). Rb-Sr whole
rock isotopic analyses of nine Catheart Mountain samples
from the equigranular granodiorite hosts and porphyritic
dikes that vary widely in their indices of alteration (from
40 to 74) yield an apparent isochron age of 422+15 Ma
'U.S. Geological Survey, Mail Stop 954, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092.
*U.S. Geological Survey, Mail Stop 959, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092.
and Sri (strontium initial ratio) = 0.70603+0.00019 (2 oy.
The bulk of the geochemical data indicate that the Rb-Sr
isotopic system was reset as a result of alteration, that the
isochron age is not the crystallization age of the stock or
of the mineralization, and that the Sri of the fluids
involved during alteration modified the magmatic Sri
value. Fluid-inclusion studies indicate that late
mineralizing fluids responsible for Cu mineralization were
pervasive throughout the system; they were of moderate
temperatures (< 300C); and they were characterized by
generally low salinities (up to 5 equivalent weight percent
NaCl) and moderate CO2 contents (515 mole percent).
Widely varying CO2/H2O values of inclusions in quartz
veins suggest that CO2 effervescence may have contrib
uted to deposition of the second stage of mineralization
that produced chalcopyrite + sphalerite + galena + quartz
+ pyrite + calcite.
INTRODUCTION
The Catheart Mountain stock is associated with one
of the best examples of porphyry-style Cu-Mo mineraliza
tion in the New England Appalachian Mountains.
Together with the nearby molybdenite occurrence at Sally
Mountain, they belong to the northern Appalachian por
phyry Cu-Mo province (Hollister and others, 1974). The
Catheart Mountain stock and granitic rocks at Sally Moun
tain intruded the Attean pluton in the Chain Lakes massif
and are part of an allochthonous block that rests on
Proterozoic Grenville basement (fig. 1) (Stewart and
others, 1985). The purpose of this study is to investigate
the response of the Rb-Sr isotopic system to the hydrothe
mal alteration at Catheart Mountain and to present a
 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

\
\
\
\
\
\
\
\
\
\
\
\
O e \
N
132. .*
_/ Attean \
\
_2^ pluton \

T ~~ Hog Island
T ~~~ pluton
T-- ~ Sally Mtn
T ~~~ pluton

EXPLANATION Chain
|
Lakes * ~
pluton

Alteration zones "*" ":

pluton
Potassic

[T] Quartz-sericitic
[T] Propylitic

O 500 1000 METERS

1

Figure 1. Map showing the generalized hydrothermal alteration areas and location of the drill core samples used at Catheart Mountain
(modified after Schmidt, 1974). Line AB represents the southernmost extension of the North American basement (Grenville) proposed by
Zen (1983). Numbered dots identify drill hole locations; numbered X's show locations of outcrops.

preliminary characterization of the fluid inclusions associ
ated with the alteration.
Schmidt (1974) determined the extent of the regional
alteration zones at Catheart Mountain. He proposed that
the stock was associated with a porphyry Cu-Mo system
similar to that described in the classical porphyry model
(Lowell and Guilbert, 1970). Atkinson (1977) identified
breccia pipes and proposed that the alteration also
included a silicic, fluorite-rich phase. Ayuso (1987)
established compositional variations in white micas from
all the alteration zones and attempted to relate white-mica
chemistry to the Cu and Mo content of the bulk rocks;
Ayuso and others (1988) determined the range in O and
Pb isotopic compositions, and Ayuso (1989) presented
results of a study of the geochemistry of the Catheart
Mountain stock. Recently, in an effort to systematically
characterize the chemical and mineralogical changes dur
ing hydrothermal alteration, an index of alteration was
developed for samples at varying stages of alteration at
Catheart Mountain (Molling and Ayuso, 1990).
ANALYTICAL PROCEDURES
One hundred and sixteen holes were drilled on Cat
heart Mountain by the Scott Paper Company, and 45 sam
 CATHEART MOUNTAIN Cu-Mo DEPOSIT, MAINE
from these holes were obtained for geochemical studies.
Modal abundances were estimated by point counting and
used a computational technique that employed whole-rock
data and microprobe compositions of minerals of each rock
(Molling and Ayuso, 1990) in a scheme first proposed by
Ferry (1985). A suite of samples containing abundant fluid
inclusions in quartz phenocrysts and in narrow quartz +
muscovite + sulfide veins was selected from 19 core loca
tions within and adjacent to the two mineralized centers at
Catheart Mountain (fig. 1). Fluid inclusions in samples of
the Attean pluton and of nearby plutons, including Sally
Mountain, Hog Island, and Chain of Ponds, were also
examined for comparison. Petrographic studies of fluid
inclusions were made on over 35 thin sections of material
from the equigranular and porphyritic parts of the stock.
The thermometry data were collected using equipment mar
keted by Fluid, Inc. of Denver, Colorado. The stage was
calibrated using the melting point of ice, synthetic fluid
inclusions prepared by Fluid, Inc., and various other stan
dards of known melting point. A small number of samples
were crushed in oil, with a crushing stage constructed by
Chaixmeca, Ltd. (France), to confirm the presence of non
compressible gas in some inclusions. The distribution of all
types of inclusions in phenocrystic quartz and vein quartz
and calcite was characterized, and heating and freezing
studies were conducted on representative inclusions; initial
results are summarized in table 1 and compared with data
for other granite-related mineral deposits of the northern
Appalachian Mountains and with porphyry Cu deposits in
the Western United States and Canada.
Whole-rock splits of nine samples from Catheart
Mountain were analyzed by isotope dilution for Rb and Sr
using a 6-inch, 60-degree-sector, NBS-type mass spec
trometer at the U.S. Geological Survey in Reston, Vir
ginia. A special effort was made to obtain the least altered
samples available from drill core in an attempt to estimate
the original isotopic composition of the Catheart Mountain
stock. Five of the samples from Catheart Mountain are
from the porphyritic dikes, and four are from equigranular
host rocks. Three outcrop samples from the nearby Sally
Mountain Mo occurrence were also analyzed for compari
son with the Catheart Mountain stock. The ratio of *Sr/
*Sr of 0.1194 was used for normalizing the Sr isotopic
measurements. Sr blanks are less than 2 ng and those for
Rb are 0.2 ng or less. Replicate analyses of NBS SRM
987 give an averge *'Sr.Sr of 0.71016+0.00006 (95 per
cent confidence level). The precision of individual sam
ples at the 95 percent confidence level is 0.022 percent for
*'Sr*Sr, 1.2 percent for Rb, and 2.0 percent for Sr. The
isotopic and decay constants recommended by Steiger and
Jager (1977) were used for all the isotopic calculations.
The regression and uncertainty calculation of York (1966)
was used to determine ages and their uncertainty at the 2
sigma level.
47
GEOLOGIC SETTING
The Catheart Mountain stock lies within the Attean
pluton on the southeast flank of the Boundary Mountains
anticlinorium (Albee and Boudette, 1972) in the Chain
Lakes massif, an allochthonous crustal block consisting
for the most part of a thick sequence of diamictite,
metasedimentary rocks (quartzites, feldspathic metasand
stone), and sparse mafic metavolcanic rocks (Boone and
others, 1970). The Chain Lakes massif was regionally
metamorphosed to sillimanite grade and retrograded to
chlorite facies in the middle Paleozoic. Granitic rocks
exposed at Sally Mountain also contain molybdenite min
eralization (Albee and Boudette, 1972), but the extent of
mineralization is minor relative to Catheart Mountain, and
the regional alteration zones are not well developed.
CATHEART MOUNTAIN PETROLOGY
The Catheart Mountain stock consists of equigranular,
fine- to medium-grained granitic to granodioritic rocks
(referred to as quartz monzonites by Schmidt, 1974) and
irregular dikes of porphyritic granodiorite and granite
(quartz porphyries of Schmidt, 1974). Sulfide mineraliza
tion is found in both the dikes and host rocks, in contrast to
the Attean pluton, which is generally sulfide free. Schmidt
(1974) delineated two main centers of hydrothermal alter
ation and five major gradational regional alteration zones.
In relative order of increasing intensity of alteration, they
consist of the propylitic, outer phyllic, phyllic, quartzose
potassic, and potassic alteration zones (fig. 1 shows a sum
mary of the alteration zones). Sulfide minerals are found
within siliceous veinlets (stockwork) and are disseminated
in the granodiorites and dikes; these veinlets are oriented
generally east-west, especially within the potassic alteration
zone. Some of the veinlets in the propylitically altered zone
consist of a central zone, containing ore minerals and
quartz-sericite (phyllic) alteration, surrounded by a K-feld
spar-rich alteration zone (potassic) (Ayuso, 1987); this is in
agreement with the zonal arrangement in many porphyry
Cu deposits (e.g., Titley, 1975; Titley, 1982; Gustafson and
Hunt, 1975; Brimhall, 1977).
Granitic rocks at Catheart Mountain, which have pre
served textural remnants of probable magmatic origin,
consist of alkali feldspar and plagioclase, quartz, biotite,
and accessory minerals (sphene, zircon, allanite, apatite,
and magnetite) (Ayuso, 1987). Hornblende may have been
originally present; virtually all rocks contain large
amounts of secondary minerals, predominantly secondary
K-feldspar, quartz and muscovite; epidote and chlorite are
also ubiquitous, as are calcite, pyrite, and hematite. Trace
amounts of gypsum were also found. Together with the
paragenetic relations established by textural studies on the
&

Canyon
Copper
1973)
Nash,
Theodore
1971,
Theodore,
and
(Nash
L(~5
vol.)
X
150
500
Cu
(Reynolds
1981)
Titley,
Beane
1985,
Beane,
and
Rita
Santa
Cu
V
X
22()
+800

Gasp
Madeleine
(Cu
1989)
others,
and
LW(~590);
X il iams-Jones
120
420 Aipncrlouxsimoantse
V*
and
IV
types
for
contents
percent.
volume
vol,
noted,
unless
percent
mole
in
given
all
vapor,
V,
liquid,
L,
only,
1985)
Spooner,
and
(Thomas
TV327
Tanco
X
265
aariable
-Sn-Li-Cs Bagdad
C(X
Cuand
1974)
unNash
L ingham,
(Preece
1982)
Beane,
and
Sierrita
X
120
440
Cu (ranWilson
1980)
others,
and
+1,000
GCu
200
X isle/Bel
1976)
Nash,
study,
(This
L(~20250)
Catheart
X
199
>300
Cu-Mo
George
Lake
1987)
others,
and
(Seal
L(~1710 )
Sb-W-MO
X
142
405

Mountain
Sally
(93
Xstudy)
This
Cu-Mo
~200
L
content?
CO2
types!
Inclusion
Deposit
late
initial
Th,
Metals
References
Gasp
Mines
1983)
(Shelton,
L(~515
vol.)
5()6
Cu-Mo
X
11() (Samson,
1990)
W-Mo-Sn-F
Pleasant
Mount
X
100
49()
City
Park
(John,
1989)
X
185
+600
Cu-Au

Crhanriatce-reilsatiecds
1.
gTable
selected
Canada.
States
United
Western
and
Eastern
the
of
deposits
mineral

Ianclusion
h!
III,
H2O-rich,
I,
(1976).
Nash
from
modified
types
CO2.
L-V
V,
H2O-CO2,
mixed
IV,
lite-bearing,

Canada
and
States
United
Eastern Canada
and
States
United
Western
I-mole
((C)
I90)
location
V
IV
IIII

Combined
*
CH4.
equivalent
as
expressed
CO2+CH4+N2

hoemopgenriaztauiroen
t[Th,
C)
in 2emper,ature
t
unknown.

.
Columbia
British
Brunswick.
New
Mexico.
New
Manitoba.
Maine.
Quebec: Arizona. Nevada.
Utah.
 CATHEART MOUNTAIN Cu-Mo DEPOSIT, MAINE
basis of fracture density and alteration intensity (Ayuso,
1987), the alteration index (AI) defined by Molling and
Ayuso (1990), which consists of the sum of the abun
dances of K-feldspar + muscovite + quartz (weight per
cent), can be used to monitor the relative degree of
alteration in these rocks. This alteration index uses miner
als that are normally produced during potassic and quartz
sericite alteration in classical porphyry Cu systemsthese
minerals are usually abundant and are normally associated
with rocks containing variable amounts of alteration, up to
the highest grades of mineralization. Values for the alter
ation index are thus thought to provide useful, systematic
estimates of the relative alteration intensity in this hydro
thermal system because virtually all rocks can be
expressed in terms of different combinations of K-feld
spar, muscovite, and quartz.
Most mineralized samples are characterized by a high
density of quartz veins and a high abundance of sulfide
minerals and especially by the occurrence of abundant,
turbid K-feldspars (Molling and Ayuso, 1990); the turbid
ity is thought to represent very small fluid inclusions and
(or) a network of grain boundaries produced during exso
lution (Ferry, 1985). In fact, tiny fluid inclusions have
been found in some hydrothermal K-feldspars that
exchanged their oxygen isotopic composition during alter
ation (O'Neil and Taylor, 1967). In general, turbid K-feld
pars at Catheart Mountain are particularly abundant in
rocks with the highest values of AI and have much lower
contents of Na and Ba relative to non-turbid K-feldspars
this is consistent with the suggestion that these elements
are removed from igneous rocks as a result of hydrother
mal alteration (Molling and Ayuso, 1990). Plagioclase is
found as sparse phenocrysts in the granodiorite, but it is
most abundant in the matrix where it has been replaced by
muscovite, epidote, and calcite; plagioclase was replaced
by hydrothermal K-feldspar in the strongly altered sam
ples. Reaction of feldspars and fluids has also produced
albitic rims surrounding plagioclase and turbid K-feldspar;
plagioclase compositions are increasingly more potassic as
a function of alteration, again indicating wholesale
removal of Ca and Na and influx of K during alteration
and mineralization of the system. Biotite was first altered
to chlorite and sphene and, subsequently, to complexly
intergrown epidote and chlorite. Clusters of rutile in
biotite are thought to represent exsolution of Ti during
recrystallization. Epidote compositions depend almost
exclusively on the composition of the mineral undergoing
replacement. In fact, epidote compositional variability at
Catheart Mountain may indicate availability of Fe" as
controlled by the prevailing O fugacity. Chlorite ranges
from chlinochlore to pycnochlorite and is strongly depen
dent on the composition of the precursor mineral (Molling
and Ayuso, 1990).
Two generations of muscovite that differ in size,
abundance, and sulfide mineral association were produced
49
as a result of the hydrothermal event (Molling and Ayuso,
1990). In addition, deuteric, fine- to medium-grained mus
covite, as found in most evolved granitic rocks, is found
as elongated blades oriented along cleavage planes replac
ing plagioclase. Growth of deuteric muscovite was not
accompanied by significant addition of sulfide minerals.
Infiltration of hydrothermal fluids at the onset of the
hydrothermal event reequilibrated the deuteric muscovite
and produced coarse-grained muscovite during the first
stage of mineralization (pyrite + molybdenite). This first
influx of sulfides was overprinted in the porphyritic dike
rocks by an assemblage that included fine-grained musco
vite and quartzthis is broadly equivalent to quartz-seri
cite alteration (sericitic) in classical porphyry-type Cu
deposits. Fine-grained muscovite thus replaced all other
minerals, and in particular, fine-grained muscovite was
produced as a result of recrystallization and reequilibration
of coarser grained muscovite contemporaneously with the
second influx of sulfide minerals into the veins (chalcopy
rite + sphalerite + galena + quartz + pyrite it calcite).
Muscovites have a wide range in octahedral cation con
tents, and average compositions of coarse-grained musco
vites are more aluminous adjacent to molybdenite than the
finer grained muscovites in contact with pyrite and chal
copyrite (Ayuso, 1987; Molling and Ayuso, 1990). This
trend may reflect temperature of formation because more
aluminous muscovites reflect higher temperatures (Velde,
1965; McDowell and Elders, 1980), and because, in gen
eral, in many porphyry systems, molybdenite is associated
with higher temperatures and earlier episodes of mineral
ization and chalcopyrite is associated with lower tempera
tures and later episodes of mineralization (e.g., Brimhall,
1977; Gustafson and Hunt, 1975).
Preliminary temperature estimates obtained by a two
feldspar geothermometer (Haselton and others, 1983), a
plagioclase-muscovite geothermometer (Green and Usdan
sky, 1986), and chlorite geothermometers (e.g., Cathelineau,
1988) indicate temperatures in the range of about 300+50C.
Molling and Ayuso (1990) suggested that mineral composi
tions recorded progressive reequilibration during hydrother
mal activity at Catheart Mountain, consistent with a system
initially dominated by magmatic water with temperatures of
approximately 400C. Subsequent alteration may have
resulted from influx of meteoric water at temperatures as
low as 200250C (generally in the temperature range found
in many other porphyry Cu systems: 200350C at Butte
(Brimhall, 1979); less than 350C in El Salvador (Gustafson
and Hunt, 1975)). In most porphyry Cu deposits, tempera
tures of later stages of mineralization reported from fluid
inclusion studies are near 200C (e.g., table 1), and salinities
are low. P-T ranges typical of later mineralizing stages of
many Cu porphyry deposits (less than 400C and less than
4 kb) are shown on figure 2; fluids responsible for most Cu
porphyry deposits are CO2 poor and highly saline, at least
initially (table 1) (e.g., Hollister, 1978; Nash, 1976).
50 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
4
- Salinity = 6 wt % NaCl
3 Catheart Mountain porphyry (fig. 1) fall into groups based
XCO2 = 0.04 XCO2 = 0.17
1S T and liquid-vapor-solid (L-V-S) inclusions predominantly
* |- - - - -\- - - - - - -
Q contain a variable number of daughter minerals. Halite,
2 Fluid
Q mainly in small inclusions in quartz phenocrysts; these
Cl
Tl - - - - - - A - - - - - - - - OnSct of
CO2 boiling
1
Liquid + Vapor
O O I T
200 I I
400 T so
Temperature (C)
Figure 2. Phase relations in the system H2O-CO2-NaCl, from
Bowers and Helgeson (1983). P-T diagram of CO2 in 6 percent
NaCl solutions at mole fraction CO2 of 0.04 and 0.17. Modified
from Seal and others (1987). Probable P-T range for late mineral
ization of Catheart Cu-Mo porphyry based on geologic constraints
is shown by dashed lines and shaded area.
Virtually all elements at Catheart Mountain have
been strongly affected by metasomatic changes, but Al, Ti,
and P are among those that appear to be relatively immo
bile (Ayuso, 1989). Rocks containing a progressively
higher AI are relatively depleted in CaO, Na2O, Fe0, and
Sr among others and have relatively higher contents for
K2O, Fe2O3, H2O, CO2, Rb, S, etc. The least altered rocks
have Cu contents as low as 10 ppm and Mo as low as 0.2
ppm, with increasing alteration, Cu content is up to 3,900
ppm, and Mo is up to 320 ppm. According to Molling and
Ayuso (1990), samples with AI < 50 represent the stage at
which coarse-grained muscovite and K-feldspar had devel
opedthis stage predates the production of fine-grained
muscovite. Samples with AI - 50 invariably contain fine
grained muscovite in the alteration assemblage; in these
samples, the Ti-bearing mineral is rutile instead of sphene.
Moreover, with increasing AI, plagioclase content
declines, as does the content of epidote, calcite, and pyrite.
FLUID INCLUSIONS AND
THE NATURE OF THE
HYDROTHERMAL FLUIDS
Fluid inclusions found in quartz and calcite from the
on textural relationships and bulk fluid chemistry. Primary
vapor-rich (V) inclusions coexist with liquid-vapor (L-V),
occur in vein quartz (fig. 3A, 3B). The L-V-S inclusions
and possibly sylvite, occur as rare daughter minerals,
may have formed during the early Mo-bearing stage of
mineralization because they are not found in vein quartz
associated with chalcopyrite. Small flakes of muscovite
are a common solid inclusion occurring both in host
quartz and in fluid inclusions (fig. 3C). When muscovite
occurs in fluid inclusions, the mica appears to have acted
as a trap for inclusion fluids during crystal growth and sig
nifies the primary nature of the inclusions. In addition, one
or two minute unidentified birefringent solids also occur,
as do small opaque solids that may be chalcopyrite (fig.
3D). Calcite also contains some apparently primary liquid
vapor inclusions (fig. 3E) that are large relative to the
secondary inclusions and aligned generally parallel to
growth zones. These inclusions also have regular liquid
to-vapor ratios and lack trapped or daughter solids.
Numerous inclusions containing two liquids and a gas
(L1-L2-V) at room temperature are present in both vein
and phenocrystic quartz. The second liquid, identified as
CO2 on the basis of crushing and thermometry studies, is
readily visible in most larger inclusions. L1-L2-V
Figure 3 (facing page). Photomicrographs of fluid-inclusion
types that characterize phenocrystic and vein quartz of the Catheart
Mountain stock. A, Primary vapor-rich inclusions (v) coexist with
liquid-vapor (21+v, 14-v) and liquid-vapor-solid inclusions in quartz.
Field of view equals 150 p.m. B. Predominant solids in inclusions
contained in quartz are tentatively identified as muscovite, halite,
and sylvite. Field of view equals 250 pum. C. Muscovite flakes (m)
occur in many inclusions in quartz. Field of view is 150 pum. D,
Opaque solids, possibly chalcopyrite (c2), are also found with other
solids in inclusions in quartz. Field of view = 150 pm. E, Primary
and secondary liquid-vapor inclusions in calcite. The larger prima
ry inclusions parallel a crystal face, whereas secondary inclusions
are aligned along a healed microfracture at an angle to the growth
surface. Field of view equals 1.5 mm. F, Inclusions containing two
liquids and a gas (21+v) coexisting with vapor-rich inclusions (v)
are present in both vein and phenocrystic quartz. Field of view
equals 150 p.m. G, CO2-bearing inclusions that have variable ratios
and that contain muscovite flakes (m) are found in quartz. Field of
view equals 200 p.m. H, Li+L2+V inclusions that have regular
H2O/CO2 values and that contain solids (21+v4-s) occur in quartz
grains encased in chalcopyrite. Field of view equals 150 pum.
 CATHEART MOUNTAIN Cu-Mo DEPOSIT, MAINE 51

- - 2|+vrs
-
- -

- \
*

D H
52 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
inclusions with highly variable volumetric proportions pre
dominate in the coarse vein quartz (fig. 3F). Vapor-rich
inclusions also occur with L1-L2-V inclusions; however,
they are less common. The three-phase inclusions have
visually estimated C2O2/(H2O+CO2) ratios that range
from 5 to 47 percent; gas-rich inclusions are also present.
The majority of these inclusions have CO2/(H2O+CO2) of
3045 percent. Some L1-L2-V inclusions also contain one
or two solids (fig. 3G); these inclusions are found in the
narrower sulfide-bearing quartz veins that cut the core of
the stock. Some of these liquid CO2-bearing inclusions
are attached to muscovite blades partially imbedded in
quartz; hence, they are probably primary in origin (fig.
3H). The volumetric proportions of CO2 and H2O in
these inclusions are more regular than those found in the
coarse quartz veins, volume percent CO2 ranges from 5 to
15 for the larger, more readily visible inclusions.
Secondary L-V inclusions and inclusions of an
ambiguous nature are abundant in quartz phenocrysts and
in vein quartz and calcite. These inclusions occur in
numerous curved and linear arrays that cut across aggre
gates of the host mineral. The inclusions are generally
small and have regular liquid-to-vapor ratios; they rarely
contain trapped solids.
Preliminary optical study of selected samples of
molybdenite-bearing rocks in the Sally Mountain stock
indicates that L-V, V, and L-V-S are the predominant
types of inclusions. Only one single inclusion has been
found containing two liquids plus vapor (L1-L2-V ) in
samples from Sally Mountain. Quartz phenocrysts in the
Attean pluton also contain L-V, V, and L-V-S inclusions,
and it apparently contains no two-liquid-plus-vapor inclu
sions. Quartz phenocrysts in samples of the nearby Devo
nian Hog Island and Chain of Ponds plutons (Albee and
Boudette, 1972; Heizler and Lux, 1984) also contain
mainly inclusions of L-V, V, and L-V-S types. A single
inclusion in quartz from the Chain of Ponds pluton is pos
sibly of the L1-L2-V type. Further studies are in progress
to place more quantitative limits on the temperature distri
bution, salinities, gas contents, and daughter minerals or
trapped solids present in primary and secondary inclusions
occurring in these plutons.
At the present time, the data suggest that the liquid,
CO2-bearing inclusions are restricted predominantly to the
granitic rocks associated with mineralization at Catheart
Mountain. The generally uniform spatial distribution of
inclusion types in vein quartz and in quartz phenocrysts,
and the large number of L1-L2-V inclusions compared to
other plutonic rocks in the region, suggests that the ore
fluids were localized in and around the Catheart Mountain
stock. The regular distribution of inclusion types (all types
are generally found at all depths in all drill core) also sug
gests that fractures and veinlets were open to most fluids,
at least during the later stages of the evolution of the
hydrothermal system, and that the fluid flow in the
hydrothermal system generally was pervasive, occurring
throughout the pluton.
Homogenization temperatures of primary and second
ary inclusions in quartz from Catheart Mountain range
from 199C to 305C. Visual estimates of low liquid-to
vapor ratios for some other inclusions (e.g., vapor-rich
inclusions in fig. 3A, 3B) suggest that the actual tempera
ture range may be even higher. Many CO2-bearing inclu
sions decrepitated prior to homogenization. CO2
effervescence is suggested by inclusions in coarse vein
quartz that have highly variable H2O/CO2 values and
coexisting L-V, L1-L2-V, and V-rich inclusions. Esti
mates of the salinity of measured inclusions based on last
melting temperatures of solids (ice or clathrate-hydrate)
range from about 0 to 5 weight percent NaCl according to
data summarized by Collins (1979). Methane is assumed
to be virtually absent based on melting temperatures of
solid CO2 to liquid CO2 in the presence of vapor of about
56+1C.
Most of the inclusion data presented here are repre
sentative of the later hydrothermal fluid that resulted in
the second pervasive influx of sulfide minerals into the
system consisting of veins of chalcopyrite + sphalerite +
galena + quartz + pyrite + calcite and fine-grained musco
vite. The preliminary results suggest a model whereby a
hydrothermal fluid of moderate temperature and salinity
(1), gradually became saturated in CO2 with decreasing
temperature and (2), began to effervesce and deposited
quartz, chalcopyrite, and a second generation of muscovite
in veins. This effervescence due to CO2 followed an initial
phase of mineralization that deposited quartz, molybden
ite, and the first generation of muscovite in veins at Cathe
art Mountain. Magmatic feldspars, the primary sinks for
Rb and Sr in the granodioritic rocks, were systematically
destroyed during this alteration, and, in their place, hydro
thermal albite and K-feldspar were formed. Figure 2
shows the probable P-T range for formation of the second
stage of sulfide mineralization; an onset of CO2 efferves
cence occurring at about 300340C would yield pressures
in the range of about 1.42.3 kb, consistent with geologic
estimates of a shallow depth of emplacement for the Cat
heart Mountain stock (35 km). CO2 is not typical of Cu
Mo porphyry systems, although it has been identified in
some deposits (table 1). For deposits where CO2 is
present, amounts of CO2 may range from as low as about
3 mole percent CO2 (limits of optical detection) to more
than 30 mole percent in CO2-rich inclusions; some inclu
sions may have no optically detectable CO2 but may have
large CO2 pressures at room temperature (Nash, 1976).
Liquid, CO2-bearing inclusions are thought to be associ
ated with relatively low-temperature porphyry-type depos
its or the low-temperature stages of such deposits (Nash,
1976). Unlike H2O boiling of a brine, CO2 effervescence
generally results in an inital drop in total salinity due to
loss of CO2 to the vapor phase; this is followed by a
 CATHEART MOUNTAIN Cu-Mo DEPOSIT, MAINE 53

Table 2. Rb-Sr isotopic data and alteration indices of the Catheart Mountain porphyry Cu-Mo
deposit and Sally Mountain Cu-Mo occurrence, Maine.
[--, data not available]

Sample! Rb Sr. Rb/Sr. 87Rb86Sr 87Sr/86Sr AI 2

CATHEART MOUNTAIN
Porphyritic dikes
1 14-494 94.() 861 (). 1092 ().3164 ().70750 49
12370 136.0 498 ().2731 ().7915 ().71079 56
5-680 127.0 379 ().3351 ().9695 ().7l 191 65
5-74() 143.0 249 ().5743 1.6674 0.71685 73
31-282 21().() 168 1.2500 3.6300 ().72674 74
CATHEART MOUNTAIN
Equigranular granodiorite host
1 14-442 96.9 680 (). 1425 0.4121 0.70854 40
31-210 126.0 621 0.2029 ().5868 0.7096] 46
31-52 87.4 736 (). 1 188 0.3435 0.70816 48
18-190 | | ().() 486 ().2263 () 6530 (), 71003 50
SALLY MOUNTAIN

AT-05 147.4 1()4 | 4 || 73 4. 1095 ().73286 -

AT-06 174.1 69 2.5232 7.325|| ().74871 -

AT-07 213.9 5] 4, 1615 12. 1247 ().77885 -

| Sample numbers correspond to drill hole number and depth.

* Alteration index, (AI) consists of the sum of the abundances (in weight percent) of K
feldspar, muscovite, and quartz (Molling and Ayuso, 1990).

gradual increase in salinity with continued boiling as salts
are concentrated in the fluid. The low and variable salini
ties of the late hydrothermal fluid may reflect this process
in part. In the Eastern United States and Canada, CO2 is a
more commonly reported constituent of inclusions in
quartz from granite-related mineral deposits compared to
deposits from the Western United States and Canada
(table 1). However, compared to such granite-related
deposits, those from the East have a distinct metallogenic
association, are possibly more deeply eroded, are generally
older, and have penetrated fundamentally different base
ment rocks. Other explanations are also possible, and fur
ther interpretation must await a more comprehensive
database.
RUBIDIUM AND STRONTIUM
GEOCHEMISTRY
Isotopic data for Rb and Sr are given in table 2 and are
plotted in figure 4. Whole-rock data for nine samples from
both petrographic units of Catheart Mountain plot on or
near the regression line within analytical uncertainty and
indicate an age of 422+15 Ma (fig. 4). The initial *'Sr.Sr
is 0.70603+0.00019. Two samples of the porphyritic dikes
and one from the equigranular host fall off the regression
line, suggesting that they were geochemically disturbed in a
different manner than the rest of the samples. If these three
samples were eliminated from the age calculation, the
remaining six samples would indicate an age of 420+4 Ma
for the Catheart Mountain stock; the initial *'SrSr ratio
for these six samples is 0.7060910.00003these results are
very similar to the entire set of data.
In a general way, measured Sr isotopic ratios are
higher in rocks with higher alteration indicesthis is con
sistent with the idea that the fluids associated with alter
ation were enriched in radiogenic Sr. The fluids could
have acquired their isotopic composition by leaching of
older minerals with high Rb/Sr values, such as biotite. The
resulting isotopic composition, thus, would have been
enriched in radiogenic *'Sr, and this mechanism would
have produced enrichment of radiogenic Sr in newly pre
cipitated albite rims and epidote involved in the destruc
tion of magmatic plagioclase.
Rb and Sr abundances in the Catheart Mountain stock
are directly related to the extent of alteration of the rock
(fig. 5). With increasing alteration, as indicated by AI val
ues, Rb contents are relatively enhanced and Sr contents
are relatively lowered (fig. 5A, 5B). The coherent changes
in Rb and Sr contents reflect the progressive destruction of
plagioclase by hydrothermal fluids. As a result of this
alteration, relatively calcic magmatic plagioclase eventu
ally disappeared, and this also resulted in a net removal of
Sr from the whole rocks. Although rims of albite and epi
dote were also formed as a result of the alteration, these
minerals are Sr poor (e.g., Dickin and Jones, 1983; Andre
54 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

.7350
250
Catheart Mountain, Maine | O
.7300 H.
I O [H D
,7250 H. 200 [T] L
b
on
7200 H. E 1504 [] O
s
on Cl + [] O
r
Co 7150 -
Cl. | D O
.C. -

.71 OO H. r 100+. O

.7050 H.
Age = 422 +15 m.y.
it; 21 87c, 186 -

Initial 8/Sr/ 86Sr = .70603+.00019

.7OOO 1 l l _l l l l
50- []. Equigranular host
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1 A O Porphyritic dike
87Rb/86 Sr
04------------
40 50 60 70 80 90 100
Figure 4. Rb-Sr isochron diagram for the Catheart Mountain
pluton. Size of individual data points reflects the 95 percent con
fidence level. 900 O
800
and Deutsch, 1986). 87Rb/86Sr isotopic values are gener
*H H
[I] Equigranular host
O Porphyritic dike
ally low at Catheart Mountain (fig. 4), from about 0.3 to 600
3.5, and broadly increase with higher AI in both the I
equigranular host rocks and porphyritic dikes, consistent
with the suggestion that the low values of 87.Rb/80Sr are
500- H
best interpreted as the result of hydrothermal leaching. Rb/ 3 400:
400- O]
Sr values increased up to about 3.5 in the most altered , 300
rocks (fig. 5). 200 H O
The three samples analyzed for their isotopic compo
sitions from Sally Mountain do not plot on the regression
line for the Catheart Mountain stock but broadly indicate
100-
j B
H. CDC

an age of 393+29 Ma and initial *'Sr/"Sr value of 0HHHHT-----T

0.70949 (table 2). Values of the "Rb/*Sr ratio range 40 50 60 70 80 90 100
from 4.1 to 12.1, and Sr content is as low as 51.4 ppm in
4
these samples. Thus, the isotopic data do not indicate that
Sally Mountain and Catheart Mountain belong to a com D Equigranular host
mon isotopic systemthis is consistent with the sugges Porphyritic dike s
3 P

tion that the two hydrothermal systems did not share the
same type of mineralizing fluids.
| e.
2 e
SUMMARY AND DISCUSSION O

The whole-rock Rb-Sr isochron age and initial 87Sr/

O[...] [T]
1 |

*Sr of the porphyritic dikes and equigranular rocks at C *,

Catheart Mountain can be interpreted in at least two ways: |
O
(1) The isotopic compositions survived the hydrothermal O 111. 1
event, and they are magmatic. In this interpretation, the 40 50 60 70 80 90 100
whole-rock Rb/Sr age of about 422 Ma and the Sr initial Alteration index
value of about 0.70603 are primary; or (2) the isotopic
compositions in the Catheart Mountain pluton formed Figure 5. Abundances of A. Rb; B, Sr.; and C, values of Rb/Sr
under the influence of the hydrothermal event. In this plotted as a function of hydrothermal alteration as represented by
interpretation, the isochron age is not the crystallization the alteration index, Al= K-feldspar + muscovite + quartz (in
age but reflects an open isotopic system and probably weight percent). Figure shows all samples analyzed in this study
mixing of different isotopic reservoirs (e.g., a magmatic and samples analyzed by Ayuso (1989).
 CATHEART MOUNTAIN Cu-Mo DEPOSIT, MAINE
component plus an assimilation component). These two
alternatives are explored in more detail below.
In the event that new Rb-Sr heterogeneities were
established by the remobilization of Rb and Sr isotopes
and a common initial Sr ratio developed at the onset of
alteration (perhaps by mechanisms similar to those found
in other altered systems elsewhere (Zeck and Wallin,
1980)), it is conceivable that whole-rock Rb-Sr ages
reported in this study approximate the age of crystalliza
tion (and mineralization) of the Catheart Mountain stock.
In this case, the whole-rock Rb-Sr isochron would provide
a minimum estimate of the crystallization agean esti
mate that would be aproximately similar to the mineraliza
tion age because mineralization usually occurs shortly
after crystallization of the pluton, in a way similar to that
of other porphyry Cu systems (Norton, 1978).
This interpretation of the Rb-Sr isotopic data, how
ever, does not appear to be viable because the hydrother
mal system was open to migration of Rb (resulting in a
net gain in Rb content) and Sr (net loss of Sr) as both the
original fabric and mineralogy of the pluton were drasti
cally modified. A net loss of Sr in the system implies that
radiogenic Sr was also lost. Moreover, a magmatic inter
pretation of the whole-rock Rb-Sr age is in conflict with
previously reported age determinations on muscovite from
molybdenite veins from Catheart Mountain that yielded
somewhat older ages (K-Ar = 441+8 Ma; Rb-Sr model
age = 447+10 Ma) (Lyons and others, 1986); the analyti
cal uncertainties of these determinations, however, indi
cate that all of these ages are roughly equivalent. Lyons
and others (1986) indicated that the Rb-Sr model age
(447+10 Ma) obtained on Catheart Mountain was similar
to the zircon age of 443+4 Ma of the Attean pluton; they
also suggested that the Attean and the Catheart Mountain
stock were comagmaticwe do not favor this interpreta
tion on the basis of geochemical contrasts between the
Attean and the Catheart Mountain stock (Ayuso, 1989). In
fact, the K-Ar muscovite age and Rb-Sr model age (Lyons
and others, 1986) contrast with preliminary results of our
Ongoing Ar-Ar dating study, indicating that muscovites in
textural equilibrium with molybdenite and chalcopyrite
have Ar-Ar plateau ages of about 435+2 Ma (Ayuso and
Kunk, unpub. data). If this Ar-Ar muscovite age is mean
ingful, the implication is that neither our whole-rock Rb
Sr results (422+15 Ma) nor the previously determined Rb
Sr model ages and K-Ar ages (447+10 Ma and 441+8 Ma,
respectively; Lyons and others, 1986) give a precise esti
mate of the age of the hydrothermal event. However, one
important and unexpected result of this study is that,
despite the strong alteration effects at Catheart Mountain,
the whole-rock Rb-Sr age differs by less than 3 percent
from our preliminary Ar-Ar age.
There is no direct way to recover the original 87Sr/
86Sr value, unless contamination throughout the life of the
hydrothermal event was produced by magmatic fluids con
taining the same "Sr/"Sr ratio as the rocks. The samples
55
would have lost Sr (higher Rb/Sr values only), but the ini
tial 87Sr/86Sr signature would have been preserved, and
the estimated age would be a general approximation of the
age of the hydrothermal event.
Age-corrected Sr initial ratios at 434 Ma, which is
the suggested Ar-Ar age of muscovite associated with the
mineralization at Catheart Mountain (Ayuso and Kunk,
unpub. data), range from about 0.7055 to about 0.7066.
The only striking exception is from a highly altered sam
ple (31282; AI = 74) that does not contain epidote and
that has a value of initial Sr of about 0.7044. In contrast
to hydrothermal alteration systems elsewhere (Andre and
Deutsch, 1986), epidote-poor rocks (higher alteration
index) at Catheart Mountain are the most intensely
altered, and we think that they are unlikely to contain pri
mary isotopic features. If, as we think, the hydrothermal
fluids differed from the magmatic fluids and were proba
bly more radiogenic than the rocks, the resulting Sr initial
ratio represents a mixture of the magmatic Sr initial ratio
and the radiogenic ratio of the contaminating fluid,
although the end-member compositions cannot be estab
lished at this time. In fact, the plot of age-corrected initial
Sr versus 1/Sr (not shown) demonstrates that the porphy
ritic dikes and the equigranular host rocks can be sepa
rated into two general fields and that, at least in the
porphyritic dikes, there is a hint that the higher initial Sr
values are in the Sr-poor rocks (high AI).
At the height of hydrothermal activity at Catheart
Mountain, the fluid was probably moderately hot (at least
as hot as 300C), oxidizing (high values of Fe"/Fe"),
and saline (occurrence of halite, table 1). The hydrother
mal fluid progressively became saturated in CO2 with
decreasing temperature, effervesced, and deposited quartz,
chalcopyrite, and fine-grained muscovite in veins. Effer
vescence may indicate a period of second boiling as a
result of CO2a period subsequent to the event that
deposited quartz, molybdenite, and coarser muscovite.
Destruction of biotite accounts for the loss of Mg in the
whole rock. Igneous plagioclase released Ca and Na (and
Sr) to the fluid, and reactions with the fluid ultimately pre
cipitated hydrothermal K-feldspar and produced an enrich
ment in K and Rb; these reactions also produced high
values of K/Na and Fe"/Fe in hydrothermally altered
rocks by base-exchange reactions (Hemley and Jones,
1964)this also lowered the KCl/HCl ratio in the aqueous
phase (Burnham, 1979). The fluid was thus probably
enriched in radiogenic Sr as a result of these reactions.
Hydrothermally produced K-feldspar and muscovite
account for the bulk of the K and H2O influx into the
system; albitic plagioclase accounts for the Na and at least
some of the Sr retained in the rock after alteration. Epi
dote and chlorite formed during alteration also partially
account for the Ca, Fe, Mg, and Sr in the rock, and the
sulfide minerals account for the influx of Cu, Mo, and S.
Toward the end of the alteration events, hydrolysis reac
tions increased values of the KCl/HCl ratio and formed
muscovite from K-feldspar (Burnham and Ohmoto, 1980).
56 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
ACKNOWLEDGMENTS
This study benefitted from the generous help of R.G.
Schmidt during selection of core samples and through dis
cussions on the nature of the mineralization and alteration
in the Catheart Mountain area. We thank J.G. Arth for his
critical help and advice in the isotope laboratory and for
discussions on the geochemistry of granitic rocks. The
paper benefitted greatly from reviews by B.R. Doe and R.
Seal.
REFERENCES CITED
Albee, A., and Boudette, E.L., 1972, Geology of the Attean quad
rangle, Somerset County, Maine: U.S. Geological Survey Bul
letin 1297, 110 p.
Andre, L. and Deutsch, S., 1986, Magmatic *'Sr/*Sr relicts in hy
drothermally altered quartz diorites (Brabant Massif, Belgium)
and the role of epidote as a Sr filter: Contributions to Mineral
ogy and Petrology, v. 92, p. 104112.
Atkinson, D., 1977, Catheart Mountain, an Ordovician porphyry
copper molybdenum occurrence in northern Appalachia: Lon
don, Ontario, University of Western Ontario, Ph.D. disserta
tion, 192 p.
Ayuso, R.A., 1987, Mineralized rocks of the Catheart Mountain
Cu-Mo porphyry, Maine: Aluminous white micas as indices of
mineralization: U.S. Geological Survey Bulletin 1803, 16 p.
Ayuso, R.A., 1989, Geochemistry of the Catheart Mountain por
phyry copper deposit, Maine: Maine Geologcial Survey, Stud
ies in Maine Geology, Jackson Volume, p. 139162.
Ayuso, R.A., Horan, M., and Criss, R.E., 1988, Pb and O isotopic
geochemistry of granitic plutons from northern Maine: Ameri
can Journal of Science, v. 288, p. 421460
Beane, R.E., and Titley, S.R., 1981, Porphyry copper deposits, Part
II. Hydrothermal alteration and mineralization: Economic Ge
ology, 75th Anniversary Volume, p. 214269.
Boone, G.M., Boudette, E.L., and Moench, R.H., 1970, Bedrock
geology of the Rangeley LakesDead River basin region, west
ern Maine, in New England Intercollegiate Geologic Confer
ence, 62nd Annual Meeting, Rangeley Maine: Guidebooks for
Field Trips in the Rangeley LakesDead River Basin, Western
Maine, p. 124.
Bowers, T.S., and Helgeson, H.C., 1983, Calculation of the thermo
dynamic and geochemical consequences of nonideal mixing in
the system H2O-CO2-NaCl on phase relations in geologic sys
temsEquation of state for H2O-CO2-NaCl fluids at high
pressures and temperatures: Geochimica et Cosmochimica
Acta, v. 47, p. 12471275.
Brimhall, G.H., 1979, Lithologic determination of mass transfer
mechanisms of multiple stage porphyry copper mineralization
at Butte, Montana: Vein formation of hypogene leaching and
enrichment of potassium-silicate proto-ore: Economic Geolo
gy, v. 74, p. 556589.
Brimhall, G.H., 1977, Early fracture-controlled disseminated
mineralization at Butte, MT: Economic Geology, v. 72,
p. 3759.
Burnham, C.W., 1979, Magmas and hydrothermal fluids, in
Barnes, H.L., ed., Geochemistry of Hydrothermal Ore Depos
its, 2nd edition: John Wiley and Sons, p. 71136.
Burnham, C.W., and Ohmoto, H., 1980, Late-stage processes
of felsic magmatism: Mining Geology Special Issue No. 8.
p. 111.
Cathelineau, M., 1988, Cation site occupancy in chlorites and illites
as a function of temperature: Clay minerals, v. 23, p. 471485.
Collins, P.L.F., 1979, Gas hydrates in CO2-bearing fluid inclusions
and the use of freezing data for estimation of salinity: Eco
nomic Geology, v. 74, p. 14351444.
Dickin, A.P., and Jones, N.W., 1983, Relative element mobility
during hydrothermal alteration of a basic sill. Isle of Skye.
N.W. Scotland: Contributions to Mineralogy and Petrology, v.
82, p. 147153.
Ferry, J., 1985, Hydrothermal alteration of Tertiary igneous rocks
from the isle of Skye, northwest Scotland: II. Granites: Con
tributions to Mineralogy and Petrology, v. 91, p. 283304.
Green, N.L., and Usdansky, S.I., 1986, Ternary feldspar mixing
relations and thermobarometry: American Mineralogist, v.
71, p. 11001108.
Gustafson, L.B., and Hunt, J.P., 1975, The porphyry copper
deposits at El Salvador, Chile: Economic Geology, v. 70.
p. 857912.
Haselton, H.T., Jr., Hovis, G.L., Hemingway, B.S., and Robie,
R.A., 1983, Calorimetric investigation of the excess entropy of
mixing in analbite-sanidine solid solutionslack of evidence
for Na, K, short range order and implications for two-feldspar
geothermometry: American Mineralogist, v. 68, p. 398413.
Heizler, M.T., and Lux, D.R., 1984 "ArAr incremental release
geochronology on hornblende-biotite pairs for the Chain of
Ponds pluton, west-central Maine [abs.]: Geological Society of
America, Abstract with Programs, v. 16, p. 2223.
Hemley, J.J., and Jones, W.R., 1964, Chemical aspects of hydro
thermal alteration with emphasis on hydrogen metsomatism:
Economic Geology, v. 59, p. 538569.
Hollister, L.S., 1978, Geology of the Porphyry Copper Deposits of
the Western Hemisphere: New York, American Institute of
Mining, Metallurgical, and Petroleum Engineers, 219 p.
Hollister, V.F., Potter, R.R., and Barker, A.L., 1974, Porphyry-type
deposits of the Appalachian orogen: Economic Geology, v. 69,
p. 618630.
John, D.A., 1989, Evolution of hydrothermal fluids in the Park Pre
mier stock, central Wasatch Mountains, Utah: Economic Ge
ology, v. 84, p. 879902.
Lowell, J.D., and Guilbert, J.M., 1970, Lateral and vertical alter
ation-mineralization zoning in porphyry ore deposits: Eco
nomic Geology, v. 63, p. 645654.
Lyons, J.B., Aleinikoff, J.N., and Zartman, R.E., 1986, Urani
um-thorium-lead ages of the Highlandcroft plutonic suite.
northern New England: American Journal of Science, v.
286, p. 489509.
McDowell, S.D., and Elders, W.A., 1980, Layer silicate mineral
ogy in borehole Elmore 1, Salton Sea geothermal field, Cali
fornia, USA: Contributions to Mineralogy and Petrology, v.
74, p. 293310.
Molling. P. and Ayuso, R.A., 1990, Mineral paragenesis and hydro
thermal alteration in the Catheart Mountain porphyry Cu-Mo
deposit [abs.]: Geological Society of America, Abstracts with
Programs, v. 22, p. 5758.
 CATHEART MOUNTAIN Cu-Mo DEPOSIT, MAINE
Nash, J.T., 1976, Fluid inclusion petrologyData from porphyry
copper deposits and applications to exploration: U.S. Geolog
ical Survey Professional Paper 907D, 16 p.
Nash, J.T., and Cunningham, C.G., Jr., 1974, Fluid-inclusion stud
ies of the porphyry copper deposit at Bagdad, Arizona: U.S.
Geological Survey Journal Research, v. 2, p. 3134.
Nash, J.T., and Theodore, T.G., 1971, Ore fluids in a porphyry cop
per deposit at Copper Canyon, Nevada: Economic Geology, v.
66, p. 385399.
Norton, D., 1978, Source lines, source regions, and path lines for
fluids in hydrothermal systems related to cooling plutons:
Economic Geology, v. 73, p. 2128.
O'Neil, J.R. and Taylor, H.P., 1967, The oxygen isotope and cation
exchange chemistry of feldspars: American Mineralogist, v.
52, p. 14141437.
Preece, R.K., III, and Beane, R.E., 1982, Contrasting evolutions of
hydrothermal alteration in quartz monzonite and quartz diorite
wall rocks at the Sierrita porphyry copper deposit, Arizona:
Economic Geology, v. 77, p. 16211641.
Reynolds, T.J., and Beane, R.E., 1985, Evolution of hydrothermal
fluid characteristics at the Santa Rita, New Mexico, porphyry
copper deposit: Economic Geology, v. 80, p. 13281347.
Samson, I.M., 1990, Fluid evolution and mineralization in a sub
volcanic granitic stock: The Mount Pleasant W-Mo-Sn de
posits, New Brunswick, Canada: Economic Geology, v. 85,
p. 145163.
Schmidt, R.G., 1974, Preliminary study of rock alteration in the
Catheart Mountain molybdenum-copper deposit, Maine: U.S.
Geological Survey Journal Research, v. 2, p. 189194.
Seal, R.R., II, Clark, A.H., and Morrissy, C.J., 1987, Stockwork
tungsten (scheelite)-molybdenum mineralization, Lake
George, southwestern New Brunswick: Economic Geology,
v. 82, p. 12591282.
Shelton, K.L., 1983, Composition and origin of ore-forming fluids
in a carbonate-hosted porphyry copper and skarn deposit: A
fluid inclusions and stable isotope study of mines, Gasp, Que
bec: Economic Geology, v. 78, p. 387421.
Steiger, R.H., and Jager, E., 1977, Convention on the use of decay
constants in geo- and cosmochronology: Earth and Planetary
Science Letters, v. 36, p. 359362.
57
Stewart, D. B., Luetgert, J.H., Unger, J.D., and Phillips, J.D., 1985,
Characterization of the crust of Maine by seismic reflection
and refraction [abs.]: Geological Society of America, Abstracts
with Programs, v. 17, p. 727.
Theodore, T.G., and Nash, J.T., 1973, Geochemical and fluid zona
tion at Copper Canyon, Lander County, Nevada: Economic
Geology, v. 68, p. 565570.
Thomas, A.V., and Spooner, E.T.C., 1985, Occurrence, petrology,
and fluid inclusion characteristics of tantalum mineralization
in the Tanco granitic pegmatite, S.E. Manitoba, in Taylor, R.P.,
and Strong, D.F., eds., Granite-Related Mineral Deposits, ex
tended abstracts of papers presented at the CIM conference on
granite-related mineral deposits): Halifax, Canada, CIM Geol
ogy Division, p. 274278.
Titley, S.R., 1982, The style and progress of mineralization and al
teration in porphyry copper systems, in Titley, S.R., ed., Ad
vances in Geology of the Porphyry Copper Deposits,
Southwestern North America: Tucson, Arizona, University of
Arizona, p. 93116.
Titley, S. R., 1975, Geological characteristics and environment of
some porphyry copper occurrences in the southwestern Pacif
ic: Economic Geology, v. 70, p. 499514.
Velde, B., 1965, Phengite micas: Synthesis, stability, and natural
occurrence: American Journal of Science, v. 263, p. 886913.
Williams-Jones, A.E., Samson, I.M., and Linnen, R.L., 1989, Fluid
evolution and its role in the genesis of the granite-related
Madeleine copper deposit, Gasp, Quebec: Economic Geolo
gy, v. 84, p. 15151524.
Wilson, J.W.J., Kessler, S.E., Cloke, P.L., and Kelly, W.C., 1980,
Fluid inclusion geochemistry of the Granisle and Bell porphyry
copper deposits, British Columbia: Economic Geology, v. 75,
p. 45661.
York, D., 1966, Least squares fitting of a straight line: Canadian
Journal of Physics, v. 44, p. 10791086.
Zeck, H.P., and Wallin, B., 1980, A 1,220+60 m.y. Rb-Sr isochron
age representing a Taylor-convection caused recrystallization
event in a granitic rocks suite: Contributions to Mineralogy and
Petrology, v. 74, p. 4553.
Zen, E-an, 1983, Exotic terranes in the New England Appala
chiansLimits, candidates, and ages: A speculative essay:
Geological Society of America Memoir 158, p. 5581.
 CHAPTER F
STRUCTURE AND ORIGIN OF THE
ELY COPPER DEPOSIT,
EAST-CENTRAL VERMONT
By TERRY W. OFFIELD, JOHN F. SLACK," and SARAH A. WITTINBRINK?
ABSTRACT
The Ely copper deposit, in the Orange County copper
district, east-central Vermont, occurs within pelitic and
psammitic schist of the Middle Silurian(?) to Lower Devo
nian Gile Mountain Formation in a stratigraphic zone con
taining relatively abundant amphibolite. Stratabound
layers and lenses of massive sulfide are concentrated near
the crest of a prominent regional fold and are distributed
en echelon down the plunge of the fold. The sulfide is
interpreted to be of exhalative volcanogenic origin, formed
in association with sea-floor basalts. Subsequent to depo
sition on or near the sea floor, layer-parallel foliation
developed during intense isoclinal folding (F1). This was
followed by open to isoclinal folding (F2) that produced
reclined folds; the latest deformational episode consisted
of regional arching and local structural doming (F3). Each
tectonic episode was accompanied by regional metamor
phism. Sulfide layers were tectonically thickened and
thinned, folded, and repeated by ductile thrust faulting.
The present distribution of sulfide largely is controlled by
the geometry of F2 parasitic folds. Folding and shearing
were intensified by ductility contrast and mobilization of
the sulfide during deformation and metamorphism.
INTRODUCTION
The Ely mine, also known as the Copperfield or Ver
shire mine, is located about 5 km southeast of the town of
'U.S. Geological Survey, Mail Stop 954, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092.
*Department of Geological Sciences, Northwestern University, Evan
ston, IL 60208.
Vershire Center in the Orange County copper district of
east-central Vermont (fig. 1). Ely was discovered in 1820,
and for a time during the 1850's, was the largest copper
producer in the United States (Stevens, 1911, p. 1772).
The mine closed in 1892 after producing approximately
35,000,000 pounds of copper from about 500,000 tons of
ore that had an average grade of 3.5 percent Cu (White
and Eric, 1944). Attempts at production around the turn of
the century had little success, and the mine was aban
doned in 1903.
Geologic studies of the mine area are limited. An
unpublished thesis by Smith (1905) included observations
made while most of the mine workings were accessible as
well as descriptions of the ore by mine supervisors. The
history and production of the mine and descriptions of the
orebodies were presented by Wheeler (1883), Smyth and
Smith (1904), and Weed (1911); Buerger (1935) described
the ore and gangue mineralogy. White and Eric (1944)
provided cross sections of the workings accessible above
flooding level, a map of foliations and lineations measured
at the surface around the mine, and a brief description of
the local geology. The regional geology was outlined by
Hitchcock (1912) and Doll (1944). Hermance and others
(1949) reported on the findings from nine holes drilled in
the mine area by the Bureau of Mines.
The present study is part of work in progress to deter
mine the regional and local structural setting and deforma
tional history of the sulfide deposits of the district. Results
so far differ from those of prior studies in two important
regards: (1) With one exception, earlier workers recognized
only two episodes of deformation; we find good evidence
of three deformational events in which the regional folds of
the study area formed in an episode between the two defor
mational events recognized earlier; and (2) prior interpreta
tions involved syn- to post-metamorphic introduction of
sulfide into structural loci, following the conventional epi
genetic model of the times, whereas we interpret the origin
60 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

7230'
4400' 7215 largely of hornblende-plagioclase amphibolite, with a local
EXPLANATION tholeiitic chemistry (Hepburn, 1984; J.F. Slack, unpub.
Gile Mountain Formation data) and generally has been interpreted as a series of
(Middle Silurian to tally Devonian)
basalt flows transgressing a regional pelite-carbonate
Standing Pond Volcanic member
SP (Middle Silurian)
GM facies boundary (see Annis and others, 1983). Near
P, Waits River Formation
Springfield, Vermont, 70 km south of Ely, a felsic dike
(Middle Silurian) Vershire
O Center that cuts the Standing Pond has been dated as Middle Sil
\; 3
?
\ *
Ely
inine
urian (423+4 Ma) by Aleinikoff and Karabinos (1990)
\ z WR
this requires that the underlying Waits River Formation be
i; 3: no younger than Middle Silurian in age. The Gile Moun
West
| A Fairlee o tain Formation is here given an age of Middle Silurian(?)
A walker Gove to Early Devonian. The Connecticut Valley trough
../ mine 2 mine
sequence, therefore, is apparently Silurian and Devonian
in age, deposited shortly before the Devonian Acadian
GM orogeny.
SP
Most structural studies of the region identified two
A South stages of deformation that were assumed to be products of
, Strafford Elizabeth
i O finine the Acadian orogeny (e.g., White and Jahns, 1950; Lyons,
O : 1955; Chang and others, 1965; Fisher and Karabinos,
E
SP
Li i 1980). These and other workers considered the deforma
LL :
< |
CC :
tion to have involved pervasive isoclinal folding followed
F- | by an episode of regional arching that included the forma
C/D :
! tion of local structural domes. This arching broadly
WR l o 1 2 3 MILES
warped the older foliation about an axis that trends
/
- 2 3 kilomet ERs approximately north-south along the west side of the cop
4345'
per district. Large folds (the zigzag folds of Doll, 1944)
outlined by the Standing Pond Volcanic Member (fig. 1)
Figure 1. Map showing structural pattern and Ely and Elizabeth were believed to have formed in the arching episode and
deposits in the Orange County copper district. Geology after Howard were interpreted by White and Jahns (1950) as reverse
(1969). drags that cascaded off the rising Strafford dome (fig. 1).
The earlier studies also concluded that the two deforma
tions each were accompanied by episodes of metamor
of the sulfide syngenetic (volcanogenic) with deposition of
phism to garnet grade or higher.
sulfide prior to deformation and metamorphism.
Our recent geologic studies indicate, instead, that at
least three episodes of deformation occurred and that the
regional arching and structural doming warped rocks
GEOLOGIC SETTING
already deformed by two episodes of intense folding. This
more complicated deformational history was also recog
The rock units in the study area make up the Con nized by Woodland (1977) in an area about 18 km south
necticut Valley trough sequence, which extends nearly 400 west of the Ely mine, and evidence documenting it has
km from western Massachusetts to southeastern Quebec been found consistently throughout the Connecticut Valley
(Hatch, 1988). The Gile Mountain Formation, stratigraphi trough sequence of east-central Vermont by the authors
cally at the top of the sequence, consists predominantly of (Offield and Slack, 1990). The metamorphic history asso
pelitic schist, with minor quartzite and amphibolite in ciated with this series of deformations has been docu
places. It is considered to be, in part, Early Devonian on mented and described by Menard (1991) and Menard and
the basis of plant fossils found in rocks in Quebec that are Spear (1991). D1, evidenced by abundant, small-scale
believed, but not proven, to be equivalent (Hueber and isoclines and layer-parallel axial-surface foliation, was
others, 1990). The Gile Mountain overlies calcareous accompanied by biotite-grade metamorphism. D2 formed
metapelite and minor quartzose marble of the Waits River open to isoclinal folds with highly variable development
Formation in apparent lithologic transition, suggesting a of axial-surface foliation, depending on rock type and
transgressive facies boundary (e.g., Lyons, 1955). The structural locus. D2 was associated with garnet-grade
Standing Pond Volcanic Member of the Waits River metamorphism during a significant increase in pressure.
occurs mainly along the mapped Waits RiverGile Moun D3, the regional arching event, was associated with the
tain contact but departs slightly from it to the north of the development of metamorphic grades up to sillimanite in
Strafford dome (fig. 1). The Standing Pond consists the Ely area, decreasing in grade to the east and west.
 ELY COPPER DEPOSIT, EAST CENTRAL VERMONT
The zigzag folds are present only on the east side of
the Connecticut Valley trough and only in the copper dis
trict. These folds, which are reclined folds of the D2/F2
generation, are increasingly tighter and steeper in plunge
the closer they are to the Strafford dome. They occur
along the east flank of the D3 regional arch and were tilted
perhaps 10 to 30 degrees by a combination of the arching
and formation of the Strafford dome (fig. 1). Our work
indicates that the zigzag folds predate the doming and
were deformed by it. The zigzag folds are of interest in
the context of mineral resources because both the Ely and
Elizabeth orebodies occur near the crests of anticlines
(interpreted facing direction) in this group of folds and
because the F2 folding was important in shaping the sul
fide deposit at Ely.
GEOLOGY OF THE ELY MINE AREA
The Ely mine exploited stratabound, massive sulfide
deposits within the lower part of the Gile Mountain
Formation, about 800 m stratigraphically (apparent thick
ness) above the Standing Pond Volcanic Member.
Throughout eastern Vermont and the copper district, the
Gile Mountain Formation is composed predominantly of
pelitic schist. In the Ely mine area, however, a large pro
portion of the exposed rocks is psammitic (quartzose)
schist. Wheeler (1883) first recognized this relation and
noted the contrast with the setting of the Elizabeth ore
body, which is mainly within pelitic schist and altered
amphibolite (Howard, 1969; Annis and others, 1983). Drill
cores from the Ely mine area examined by the authors
(Offield and Slack) show the psammitic schist in grada
tional contact and interleaved with pelitic schist and lesser
amphibolite. The psammitic schist is mainly composed of
quartz, with minor plagioclase, mica, and garnet; some
samples display a mylonitic fabric. The pelitic schist con
tains abundant muscovite and biotite, minor quartz and
plagioclase, and local concentrations of staurolite and kya
nite. Garnets in the pelitic schist commonly show inclu
sion-rich, anhedral, rolled cores and clear euhedral rims
that reflect the complex deformation and metamorphism in
the area (see also Menard, 1991).
Schistose amphibolite layers are not generally
exposed at the surface but are seen in drill cores. The
amphibolites are common in a 300-m-thick zone (apparent
stratigraphic thickness) that hosts the Ely deposit. This
zone maintains the same stratigraphic/structural position
above the Waits RiverGile Mountain contact for at least
12 km to the south, where it contains the larger massive
sulfide deposits of the Elizabeth mine (fig. 1). The
amphibolite layers occur in both the pelitic and psammitic
schists and, like the Standing Pond Volcanic Member, are
believed to represent basaltic flows emplaced across
61
Figure 2. Block of psammitic schist showing asymmetric F2
Z-folds and almost complete lack of S2 foliation.
locally transgressive facies of shaly and sandy sediments.
In at least two places at Ely, the amphibolites are closely
associated with coticules, which are also found at Eliza
beth (Annis and others, 1983). The coticules consist of
fine-grained, Mn-rich garnet and quartz in multiple layers
less than 1 cm thick, which in places display complex,
tight to isoclinal folds.
Throughout the area, mesoscopic folds are common
in most outcrops. These folds are especially well dis
played in the psammitic schist, where they range in form
from attenuated isoclines to V shapes with one limb
longer than the other and with wavelengths up to 40 cm
(fig. 2). These are F2 folds that deform a prominent folia
tion parallel to compositional layering (S1=0). Tight to
isoclinal F1 folds only millimeters in wavelength, and to
which the dominant S1 foliation is axial planar, are com
mon on F2 fold limbs and crests, as can be seen by close
inspection of structures in drill cores (fig. 3A). S2 axial
surface foliation tends to be poorly developed in most F2
folds; it consists of alignments of biotite and quartz that
form surfaces nearly parallel to the long limbs of the
folds, rather than in a bisecting position. In some more
schistose rocks, S2 foliation is defined by a well-devel
oped crenulation cleavage. The same geometry of folds
and foliation is seen in pelitic schist and amphibolite,
except that fold crests in those lithologies tend to be
marked by a series of small, rounded plications of highly
variable shape (fig. 3B). Mineral lineations consisting of
weak alignments of quartz and biotite grains commonly
are present, oriented parallel to the axes of F2 folds. In
places, similar lineations parallel the axes of small Fl
folds and wrap over the crests of F2 folds.
White and Eric (1944) did not specify what foliation
is portrayed on their map, but because S2 foliation is
uncommon and the foliation varies in strike and dip (sug
gesting that it has been deformed), their measured folia
tions are likely to be mainly, if not entirely, S1. A
62 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Figure 3. Core from drillhole E-9, showing style of F2 folds out
lined by S1=0. A, Amphibolite displays what are probably plica
tions on the crest of a larger F2 fold. The folded foliation is S1,
axial planar to very small, tight to isoclinal folds. No S2 foliation
is developed. B, Mixed pelitic-psammitic schist shows irregular F2
folds in S1=0, with partial, local development of S2 axial-planar
foliation. Diameter of core is 3 cm.
stereographic plot of the White and Eric (1944) map data
(fig. 4) shows the foliation distribution to define a -axis of
N. 37 E., 25. Axes of small F2 folds and cleavage-bed
ding intersection lineations (White and Eric, 1944, p. 6)
on the plot are scattered, with a mean orientation of N.
35 E. 24, very close to the 3-axis. These lineations
should all relate to F2 axes; the meaning of the scatter in
orientation is not entirely clear. Although it is possible
that some of the fold-axis lineations plotted on figure 4 are
of F1 rather than F2 generation, it is unlikely that this
Figure 4. Stereographic plot of S1=0 (contoured; n=717) and F2
fold axes and cleavage-bedding intersection lineations (dots; n=50)
of White and Eric (1944, pl. 2). X, 3-axis N. 37 E. 25; o, mean
lineation N. 35 E.24. Contours of foliation density 0.1, 5, 10, 25
percent (maximum density is 42 percent).
explains the scatter. Fl folds are relatively scarce, and also
the scatter of data for 13 folds we confidently identified as
F2 is as great as that shown by the lineations plotted on
figure 4. The event that produced the scatter did not reori
ent the foliation as much as it did the fold lineations, sug
gesting post-F2 rotational slip on foliation surfaces around
axes steeply inclined to those surfaces. The scatter of the
lineations is approximately on a great circle, suggesting
dispersal by conical slip folding. The geometry is consis
tent with reshaping of F2 structures during the D3 forma
tion of the Strafford dome.
Small-scale F2 folds also show systematic changes in
the sense of asymmetry, indicating the presence and
geometry of larger scale folds in the mine area. Most
workings are clustered about halfway up the ridge that
contains the mine (fig. 5), in a zone of mostly pelitic
schist with minor layers of psammitic schist and amphibo
lite. Mesoscopic folds in surface outcrops in this zone
either are symmetric (M) or alternate in asymmetry (Z or
S) on a scale of centimeters to meters, indicative of the
axial zone of a larger fold. From 15 m above the main
workings almost to the ridge crest, 70 m higher, is a zone
of well-developed Z-folds in psammitic schist that
includes a few interbeds of amphibolite. Just below the
crest, the asymmetry reverses (S-folds), indicating the
axial surface of a larger fold.
White and Eric (1944) noted the location of the Ely
deposit on or near the axial plane of one of the regional
zigzag folds but stressed the importance of local cleavage
rolls (which they believed postdated the folds) as the main
features controlling the occurrence of ore. Smith (1905)
referred to such structures as stepdowns and believed
 ELY COPPER DEPOSIT, EAST CENTRAL VERMONT 63

4355'42"

EXPLANATION

Fault, dashed where infered

30 Strike and dip of cleavage

2." Trend and plunge of

linear element

L' shaft

Underground workings

H Drill hole

E E. Y. Y'
1 Mine sections, Fig. 6

A A" - -

L-J Interpretive cross-section,

shown on Fig. 7

Mesoscopic F, told shape

--
0 500 FEET

Figure 5. Map showing mine-area topography, mine workings, and structural elements in the area of the Ely deposit (after White and
Eric, 1944, pl. 2). AA' is section line for face of block diagram shown on figure 7. E-E' and Y-Y are mine sections shown on figure 6.
Contour interval is 50 ft.

they had localized epigenetically introduced sulfides. The
structures are not specifically delineated on the map of
White and Eric (1944) and are not readily discernible in
the pattern of 717 foliation readings shown on their map.
White and Eric (1944) may have interpreted bends in ore
layers and curved ore connections from one stope to
another, shown on their mine sections, as indicating these
cleavage rolls. Because warps of foliation would be
expected from both inhomogeneous F2 folding and the
effects of D3/F3 deformation, we cannot assign the cleav
age rolls or stepdowns to one or the other of these
eVentS.
All three phases of deformation involved shear that
resulted in the rotation or offset of marker layers and any
earlier structures. Discrete thrust faults are rarely
exposed, however. One excellent example of an exposed
thrust is in a small adit halfway between the two main
mine openings. The contact zone between psammitic
schist in the roof of the adit and pelitic schist containing
considerable sulfide below is marked by an intense shear
fabric and structural truncations. This thrust contact
appears to represent a structure associated with F1 isocli
nal folding and foliation development because it is
deformed by tight F2 folds on the east side of the adit. In
an adit about 25 m below this locality, White and Eric
(1944) described a thrust fault that bounded sulfide ore.
Their map shows the thrust connecting both adits (see fig.
5), but this connection seems doubtful unless there is
structural repetition of the thrust by F2 folding.
Mining was done primarily in tunnels and stopes off
a main shaft (actually an inclined adit that followed the
ore down plunge in the crestal area of a structurally com
plex system of folds). It was driven about N. 40 E. more
than 1,000 m down an incline of about 25, approximately
on the trend and plunge of the mesoscopic F2 folds (fig.
5). Figure 6 presents mine sections drawn by White and
Eric (1944) of workings off the upper part of the main
shaft. They show that ore occurs in thin layers, commonly
64 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Y'
_
_-T
Figure 6. Mine section drawings from White and Eric (1944, pl.
6). Heavy black is massive sulfide including schist with abundant
disseminated sulfide"; light dashed lines are foliation. Displace
ment sense shown on small vertical fault. Adits, stopes, mine rubble
indicated. E-E", Y-Y" located on figure 5. Numbers on vertical
axes indicate elevation.
in stacks or sets of two to as many as six. In several parts
of the mine workings shown, two or three of the layer sets
occur, separated by up to 10 m of largely barren country
rock. Although the mine sections reveal truncation and
convergence of ore-layer sets and of layers within sets,
they are especially important in showing that the ore lay
ers are nearly everywhere parallel to schistosity (noted
earlier by Wheeler, 1883). The few folds portrayed on the
sections deform both ore and foliation. The sections also
show disjunctive relations with one ore layer truncating
another, apparently along small thrust faults.
Wheeler (1883, p. 221) provides a detailed descrip
tion of the occurrence of ore in the mine:
The vein really consists of two parallel ore sheets or strata of
pyritic matter which are separated by an intervening layer of
schist from ten to twenty-five feet thick. The lower vein or
ore sheet is from one to eight feet in thickness, averaging
about four feet, and on the whole is richer and more regular
than the upper vein. The upper vein fluctuates very widely
both in size and richness, as it will widen out from a one-foot
vein of mundic [magnetic pyrite], assaying say one per cent
[sic] in copper, to a twenty-five foot vein that will average
over ten per cent [sic]. This great variation in the strength of
the lode, more particularly the upper vein, is due to its highly
lenticular character, as it really consists of a series of lenses
or ore bodies extending both laterally and vertically, that are
united by connecting sheets of ore. These connecting sheets
will range from one to four feet in thickness, while the bo
nanzas or ore bodies will range from six to thirty feet in
thickness. In addition to the great increase in strength at the
ore bodies, these lenses are usually composed of high grade
ore, so that they are veritable bonanzas, especially as they are
apt to be nearly solid ore, with few or no bands of quartz or
country rock. The connecting sheets are usually either mun
dic having very little copper pyrites mixed through it, or else
are small mixed veins of nearly pure copper pyrites, inter
mixed through sheets of quartz. There are frequently one or
more sheets of schist running through the ore, and these,
when of large size, have frequently been mistaken for the
hanging- and foot-walls of the vein, as no true or defined
walls exist, though the ore is usually encased between quite
even and regular surfaces.
This description from mine exposures is especially
valuable, because many of the ore faces were not available
for study even when Smith worked there in 1905. Smith's
observations, and those of the miners he interviewed, led
him to emphasize the distribution of ore in lenses. These
lenses were described as being strung out en echelon
down the inclined main shaft. Lenses to the north occurred
7 to 10 m below the lenses to their south, generally with
overlaps of about 17 m but with one example of about 65
m of overlap. Individual lenses tended to be about 100 m
long down plunge, about 17 to 80 m wide, and up to 10 m
thick (average 3 m). Smith (1905) described the ore lenses
as ending variously in sharp terminations associated with
crushed quartz; or grading into disseminated, uneconomic
ore; or fraying into unminable stringers. The ores consist
of pyrrhotite, chalcopyrite, minor pyrite, and sphalerite in
a gangue of quartz, mica, carbonate, feldspar, and minor
tourmaline (Wheeler, 1883; Weed, 1911; Buerger, 1935).
Reference by Wheeler (1883, p. 222) to quartz schist that
is highly garnetiferous, on the hanging side of the vein
may indicate the presence of coticules in the wallrocks of
the sulfide ores.
Rare vertical faults that displace ore as much as 3 m
typically are filled by carbonate and local barite (Smith,
1905; White and Eric, 1944). These faults apparently post
date all of the deformation described above and probably
are a product of the Mesozoic deformation in the region.
Similar late faults at the Elizabeth mine are filled by dia
base and lamprophyre dikes of presumed Mesozoic age
(Howard, 1969; Annis and others, 1983).
DISCUSSION AND INTERPRETATION
Previous workers considered the Ely deposit to have
formed by epigenetic replacement during or following
 ELY COPPER DEPOSIT, EAST-CENTRAL VERMONT
regional metamorphism (Wheeler, 1883; Smith, 1905;
Weed, 1911; White and Eric, 1944). Their interpretations
were influenced greatly by theories, popular at the time, of
epigenetic ore genesis and were supported by the fact that
the sulfide bodies at Ely in places crosscut previously
deformed silicate wallrocks (Smyth and Smith, 1904, p.
678; Smith, 1905, p. 127). Mine sections of White and Eric
(1944) also show sulfide layers that appear to lie in the
axial planes of F2 folds, indicating emplacement of some
sulfide after D2 deformation. We consider such occurrences
to be misleading, however, and believe that they represent
minor tectonic repositioning or mobilization of sulfides,
like those recognized in many other deformed and meta
morphosed massive sulfide deposits throughout the world
(e.g., Vokes, 1971; Gilligan and Marshall, 1987).
We suggest that the massive sulfide deposits at Ely
are largely of exhalative volcanogenic origin and that the
majority of the sulfide lenses originally formed more or
less synchronously with deposition of the Gile Mountain
Formation. Evidence for an early volcanogenic origin of
the sulfides at Ely includes their overall stratabound
geometry, their association with mafic metavolcanic rocks,
and their deformation by two periods of folding; a photo
graph of folded, alternating layers of pyrrhotite and chal
copyrite shown by Weed (1911, pl. IV) clearly documents
a predeformational age for the sulfides. Similar arguments
can be made with respect to the other massive sulfides of
the Orange County copper district, which also are believed
to be of submarine volcanogenic origin. The largely sedi
ment-hosted copper deposits of the district are considered
excellent examples of the Besshi class of massive sulfide
deposits (Slack, in press).
Because of the intense deformation in the mine area,
we cannot distinguish between a purely exhalative origin
for the sulfides and an origin involving replacement of
permeable sediments during sedimentation or diagenesis.
Although sub-sea-floor replacement, in part or whole, can
not be ruled out, we prefer a largely exhalative origin, par
ticularly because of the layered association of sulfide
deposits with amphibolite (metabasalt) and coticule. Coti
cules are believed to represent metamorphosed mixtures of
detrital, clastic sediments and exhalative metalliferous pre
cipitates enriched in iron and manganese (e.g., Spry, 1990)
that formed directly on the sea floor.
Diagenetic mineralization seems less likely, based on
the massive character of most of the ores and their similar
ity in composition and setting to the massive sulfide depos
its that are actively forming on sediment-covered ridge
crests. Modern analogs in such settings are at Guaymas
Basin in the Gulf of California (Koski and others, 1985;
Peter and Scott, 1988) and Middle Valley and Escanaba
Trough, off the coasts of Washington and California,
respectively (Goodfellow and Blaise, 1988; Koski and oth
ers, in press).
65
In modern submarine vent areas, sulfides can be irreg
ularly distributed in layers, lenses, and talus flows on the
sea floor. Sulfides may also occur irregularly below the sea
floor, a product of the movement of hydrothermal fluids
though permeable sediments. In the Orange County copper
district, such inferred complexities have been compounded
by intense deformation and metamorphism. Most of the
small-scale structures evident in the mine sections por
trayed by White and Eric (1944) involve folding and fault
ing of ore layers and S1=0 foliation; they thus must be part
of the F2 folding event. Much of the detail of individual ore
layers, however, and perhaps even the occurrence of ore in
lenses, may be due to F1 folding or to the original configu
ration of syngenetic (or synsedimentary) sulfides.
From Wheeler's (1883) description given above, it
seems likely that the two main ore layers are stratigraphi
cally separate and distinct and not a single layer repeated
by folding. The abundance of tiny F1 isoclines distributed
throughout the host rocks, however, virtually requires the
sulfide layers to have been deformed and reconfigured by
F1. Because such tight isoclinal folding seldom occurs
without shearing and thrust faulting, the ore lenses
described as having internal septa of schist may be isocli
nal F1 fold forms, sheared or offset by layer-parallel thrust
faults so as to make fold crests difficult to recognize. On
the other hand, the shaping of ore in lenses by a combina
tion of original sea floor accumulation and F1 deformation
does not readily account for the en echelon distribution of
those lenses down the plunge of the first-order F2 fold
structure. We believe that this geometry is a result of the
configuration of F2 parasitic folds, perhaps with later mod
ification of trends and plunges by rotation and shear
during D3.
Concentration of sulfide in semi-massive to massive
sulfide bodies would offer a conspicuous ductility contrast
with silicate wallrocks and promote mobility during defor
mation (e.g., Vokes, 1971; Gilligan and Marshall, 1987).
This ductility contrast locally would influence and accen
tuate deformation, and these effects probably would be
heightened by the accompanying regional metamorphism.
Both F1 and F2 folding would produce or exaggerate the
lenticular shape of ore layers on fold limbs and result in
prominent lensing of sulfide in fold crests. Ore layers
would readily serve as slip surfaces in thrust dislocations,
and sulfides under high P-T metamorpic conditions would
be subject to remobilization, including injection as vein
like bodies into adjacent rock. Such injected sulfide was
observed in the mine (e.g., Wheeler, 1883; Weed, 1911)
and was considered evidence of ore localization in open
space and for the structurally controlled, epigenetic intro
duction of sulfides.
Using surface structural observations and lithologic
information from a fence of nine drill holes, we interpret
the setting of the Ely deposit as shown in figure 7. The ore
occurs principally in a stratigraphic section of pelitic schist
 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

EXPLANATION

25 Foliation strike and dip

* Plunge of fold-axis lineation

Figure 7. Block diagram showing interpreted position of ore zone on F2 fold structure, location of main shaft, drill-hole
information (drill holes shown by circled numbers), small-scale fold shapes observed in outcrop and drill holes (M, Z, S),
and representative surface measurements of foliations and F2 axial lineations. The drill holes lie off the section drawn on
the front face of the block by as much as 200 m (in front and behind) and are presented in assumed simple updip and downdip
projections to the face. Brown, dominantly pelitic schist with subordinate psammitic schist and amphibolite. Yellow, dom
inantly psammitic schist with subordinate pelitic schist and amphibolite. Thick black portions of drill holes, significant am
phibolite layers. xx, sulfide in stopes and adits; x, significant sulfide occurrence in drill hole. AS, axial surface of large F2
fold. Front face is section AA on figure 5.

and amphibolite between sections of mixed psammitic
schist and amphibolite. Ore lenses are localized in small
folds on the flank and crest of a much larger fold (F2). The
drill-hole information indicates little likelihood for ore lay
ers to occur on the upper flank of the large fold. If sulfide
layers originally extended laterally beyond the small area
explored by mining, however, our interpretation suggests
that the most favorable structural locus for thickening and
repetition of sulfide layers would be in the crestal zone of
the large fold. Projection of drill holes 4 and 5 downdip
 ELY COPPER DEPOSIT, EAST-CENTRAL VERMONT
puts sulfide occurrences exactly in that crestal zone on the
cross section illustrated. Assuming continuity of the origi
nal sea-floor distribution of sulfides, lenses of ore may
occur far down the plunge of the main F2 fold.
A more general conclusion regarding mineral deposits
in the district is that massive sulfide bodies might reason
ably be expected anywhere within amphibolite-rich strati
graphic intervals in the Gile Mountain Formation. Within
those intervals, the axial zones of the reclined folds would
be sites where sulfide layers, if present, could have been
thickened and perhaps enhanced by the polyphase deforma
tion recognized in the area (Offield and Slack, 1990). These
stratigraphic and structural considerations, together with
recent geochemical data (Watts, 1990; Slack and others,
1990), provide guidelines in the exploration for new mas
sive sulfide deposits in the district.
REFERENCES CITED
Aleinikoff J.N., and Karabinos, Paul, 1990, Zircon U-Pb data for
the Moretown and Barnard Volcanic Members of the Missis
quoi Formation and a dike cutting the Standing Pond Volca
nics, Southeastern Vermont, in Slack, J.F., ed., Summary
Results of the Glens Falls CUSMAP Project, New York, Ver
mont, and New Hampshire: U.S. Geological Survey Bulletin
1887, p. D1D10.
Annis, M.P., Slack, J.F., and Rolph, A.L., 1983, Stratabound mas
sive sulphide deposits of the Elizabeth mine, Orange County,
Vermont, in Sangster, D.F., ed., Field Trip Guidebook to Stra
tabound Sulphide Deposits, Bathurst Area, N.B., Canada and
West-Central New England, U.S.A.: Geological Survey of
Canada Miscellaneous Report 36, p. 4151.
Buerger, N.W., 1935, The copper ores of Orange County, Vermont:
Economic Geology, v. 30, p. 434443.
Chang, P.H., Ern, E.H., Jr., and Thompson, J.B., Jr., 1965, Bedrock
geology of the Woodstock quadrangle, Vermont: Vermont
Geological Survey Bulletin 29, 65 p.
Doll, C.G., 1944, A preliminary report on the geology of the Straf
ford quadrangle, Vermont: Vermont Geological Survey, Re
port of the State Geologist, 24th, 19431944, p. 1428.
Fisher, G.W., and Karabinos, Paul, 1980, Stratigraphic sequence
of the Gile Mountain and Waits River Formations near Roy
alton, Vermont: Geological Society of America Bulletin, v.
91, p. 282286.
Gilligan, L.B., and Marshall, Brian, 1987, Textural evidence for re
mobilization in metamorphic environments: Ore Geology Re
views, v. 2, p. 205229.
Goodfellow, W.D., and Blaise, Bertrand, 1988, Sulfide formation
and hydrothermal alteration of hemipelagic sediment in Mid
dle Valley, northern Juan de Fuca Ridge: Canadian Mineralo
gist, v. 26, p. 675696.
Hatch, N.L., Jr., 1988, Some revisions to the stratigraphy and struc
ture of the Connecticut Valley trough, eastern Vermont:
American Journal of Science, v. 288, p. 10411059.
67
Hepburn, J.C., 1984, Geochemical evidence for the origin of the
Standing Pond Volcanics, eastern Vermont [abs.]: Geologi
cal Society of America, Abstracts with Programs, v. 16, no.
1, p. 23.
Hermance, H.P., Neumann, G.L., and Mosier, McHenry, 1949, In
vestigation of Ely mine copper deposit, Orange County, Vt.:
U.S. Bureau of Mines Report of Investigations 4395, 11 p.
Hitchcock, C.H., 1912, The Strafford quadrangle: Vermont Geo
logical Survey, Report of the State Geologist, 8th, 19111912,
p. 10045.
Howard, P.F., 1969, The geology of the Elizabeth mine, Vermont:
Vermont Geological Survey, Economic Geology No. 5, 73 p.
Hueber, F.M., Bothner, W.A., Hatch, N.L., Jr., Finney, S.C., and
Aleinikoff, J.N., 1990, Devonian plants from southern Que
bec and northern New Hampshire and the age of the Con
necticut Valley trough: American Journal of Science, v.
290, p. 360395.
Koski, R.A., Benninger, L.M., Zierenberg, R.A., and Jonasson,
I.R., in press, Composition and growth history of hydrothermal
deposits in Escanaba Trough, southern Gorda Ridge, in Mor
ton, J.L., Zierenberg, R.A., and Reiss, C.A., eds., Geologic,
Hydrothermal, and Biologic Studies at Escanaba Trough, Gor
da Ridge, Offshore Northern California: U.S. Geological Sur
vey Bulletin 2022.
Koski, R.A., Lonsdale, P.F., Shanks, W.C., Berndt, M.E., and
Howe, S.S., 1985, Mineralogy and geochemistry of a sedi
ment-hosted hydrothermal sulfide deposit from the southern
trough of Guaymas Basin, Gulf of California: Journal of Geo
physical Research, v.90, p. 66956707.
Lyons, J.B., 1955, Geology of the Hanover quadrangle, New
HampshireVermont: Geological Society of America Bulletin,
v. 66, p. 105145.
Menard, Thomas, 1991, Metamorphism of calcitic pelitic schists,
Strafford dome, Vermont: Rensselaer Polytechnic Institute,
unpub. Ph.D. thesis, 295 p.
Menard, Thomas, and Spear, F.S., 1991, Metamorphic history of
the Strafford dome, Vermont: An update [abs.]: Geological So
ciety of America, Abstracts with Programs, v. 23, no.1, p. 103.
Offield, T.W., and Slack, J.F., 1990, Polyphase folding and thrust
faulting in the Vermont copper belt [abs.]: Geological Society
of America, Abstracts with Programs, v. 22, no. 2, p. 61.
Peter, J.M., and Scott, S.D., 1988, Mineralogy, composition, and
fluid-inclusion microthermometry of seafloor hydrothermal
deposits in the southern trough of Guaymas Basin, Gulf of Cal
ifornia: Canadian Mineralogist, v. 26, p. 567587.
Slack, J.F., in press, Descriptive and grade-tonnage models for
Besshi-type massive sulphide deposits, in Kirkham, R.V.,
Sinclair, W.D., Thorpe, R.I., and Duke, J.M., eds., Ore De
posit Models: Geological Association of Canada Special
Publication.
Slack, J.F., Atelsek, P.J., and Whitlow, J.W., 1990, Geochemistry
of stream sediments and heavy-mineral concentrates from the
Orange County copper district, east-central Vermont, in Slack,
J.F., ed., Summary Results of the Glens Falls CUSMAP
Project, New York, Vermont, and New Hampshire: U.S. Geo
logical Survey Bulletin 1887, p. Q1Q21.
Smith, P.S., 1905, Orange County copper district, Vermont: Har
vard University, unpub. Ph.D. thesis, 177 p.
68 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Smyth, H.L., and Smith, P.S., 1904, The copper deposits of
Orange County, Vermont: Engineering and Mining Journal,
v. 77, p. 667678.
Spry, P.G., 1990, Geochemistry and origin of coticules (spessar
tine-quartz rocks) associated with metamorphosed massive
sulfide deposits, in Spry, P.G., and Bryndzia, L.T., eds., Re
gional Metamorphism of Ore Deposits and Genetic Implica
tions: Ultrecht, The Netherlands, VSP Publishers, p. 4975.
Stevens, H.J., 1911, The copper handbook [19101911): Hough
ton, Michigan, H.J. Stevens, Publisher, v. 10, 1902 p.
Vokes, F.M., 1971, Some aspects of the regional metamorphic mo
bilization of pre-existing sulphide deposits: Mineralium
Deposita, v. 6, p. 122129.
Watts, K.C., Jr., 1990, Regional exploration geochemistry of the
Glens Falls 1x2 quadrangle, New York, Vermont, and New
Hampshire, in Slack, J.F., ed., Summary Results of the Glens
Falls CUSMAP Project, New York, Vermont, and New Hamp
shire: U.S. Geological Survey Bulletin 1887, p. F1F70.
Weed, W.H., 1911, Copper deposits of the Appalachian States:
U.S. Geological Survey Bulletin 455, 166 p.
Wheeler, H.A., 1883, The copper deposits of Vermont:
School of Mines Quarterly, Columbia College, New York,
v.4, p. 219224.
White, W.S., and Eric, J.H., 1944, Preliminary report, Geology of
the Orange County copper district, Vermont: U.S. Geological
Survey Open-File Report, 37p.
White, W.S., and Jahns, R.H., 1950, Structure of central and east
central Vermont; Journal of Geology, v. 58, p. 179220.
Woodland, B.G., 1977, Structural analysis of the Silurian
Devonian rocks of the Royalton area, Vermont: Geological So
ciety of America Bulletin, v. 88, p. 11111123.
 CHAPTER G
GOLD, SILVER, AND BASE METAL EPITHERMAL
MINERAL DEPOSITS AROUND THE
GULF OF CALIFORNIA, MEXICO:
RELATIONSHIP BETWEEN MINERALIZATION
AND MAJOR STRUCTURES
By JOHN-MARK G. STAUDE'
ABSTRACT
A compilation of epithermal precious-metal occur
rences and major tectonic structures around the Gulf of
California indicates that there is a regional correspondence
and control of mineral deposits. This analysis is a regional
evaluation rather than a mine- or district-specific study,
and it permits better assessment of the overall correlation
between locations and types of structures and their relative
mineral endowment.
Compilation of 133 lode-gold, base-metal (Ag + Pb +
Zn), and copper occurrences from western and northwest
ern Mexico shows that most of the deposits are structur
ally controlled. The structural controls vary in style and
size as do the types of deposits hosted throughout the Gulf
extensional province. Systematic trends in mineralization
types, fault directions, and fault offsets indicate that the
correspondence is not uniform throughout western Mexico
and that different regions have varying types of deposits
and degrees of correspondence to major tectonic features.
The percentage of deposits occurring along major
faults is highest in Sonora and North Baja and lowest in
western mainland Mexico. The map compilation indicates
a control of deposit type by different basement regions
and by proximity to rift centers. Deposits close to the rift
axis (i.e., northeastern Baja) are more gold rich and are of
the hot-spring type, whereas those around the periphery of
the Gulf are fissure veins and quartz-adularia-sericite
'U.S. Geological Survey, Center for InterAmerican Mineral
Resources Investigations, Tucson, AZ 85705.
veins. The compilation also shows that deposits are more
gold dominated in the northwest part of the Gulf of Cali
fornia and more base metal and silver rich in the southeast
portion of the Gulf of California region. These composi
tional differences may indicate differences in hydrother
mal fluid chemistries, different source-rock compositions,
or different degrees of magmatic gas input into the epi
thermal system. The deposits are diachronous around the
Gulf: the hot-spring-type deposits in Baja formed more
recently than the Late Cretaceousearly Tertiary (Lara
mide) and middle Tertiary (Oligocene-Miocene) epither
mal vein systems.
Regional mineral exploration in transform-fault-tec
tonic settings should not only focus on following the main
tectonic suture but also upon sampling, mapping, and
determining the associated fault splays off the main suture.
Few known deposits in the area studied are located
directly on the major transform (i.e., San Andreas);
instead, most structurally associated metal occurrences are
along splays associated with smaller extensional and
strike-slip faults.
INTRODUCTION
Structural control of ore deposits is by no means a new
topic. Agricola (1556) recognized the importance of frac
tures for permitting metal-bearing fluids to infiltrate the
host rock and to deposit metals in open spaces. Early in this
century Spurr (1916), Hulin (1925), and Lindgren (1933)
suggested that structural features were keys to ore deposi
tion. Recently, the merging of structural interpretation with
70 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
mineral deposits has become significant: many new ore
deposit discoveries have been made from structural analy
ses (e.g., Mesquite mine, McLaughlin mine, etc.). Most
recent studies concentrate on theoretical models (Sibson,
1981, 1986) or individual occurrences, including such work
as the investigation of the Henderson mine by Guha and
others (1983) and the Sigma mine study by Robert and
Brown (1986). Regional studies of Proterozoic ductile
shear in the Canadian Abitibi Belt (Poulsen and Robert,
1989; Robert, 1990) and Archean ductile domains such as
the Norseman-Wiluna Belt of Western Australia (Barley
and Groves, 1988; Groves and Barley, 1988) have contrib
uted to an overall understanding of the regional association
of mineralization in ductile environments. However, com
paratively little analysis of regional data with precise mine
and structure locations has been done for brittle environ
ments; the relation between plate margins and structurally
controlled epithermal ore deposits has, likewise, received
preliminary analysis. In the Gulf of California region, tec
tonic exposures are mostly shallow and expose cataclastic
shear zones, making this area ideal for a regional study of
structures and mineralization.
The Gulf of California region is a distinguishing fea
ture of the North American Cordilleran orogenic system
because of its unique tectonic history and complex geo
logic relations. The Gulf extends over 1,700 km south
from the U.S.-Mexico border to Jalisco, Mexico (JA on
fig. 1). A simplified compilation of the geologic structures
and mineral occurrences is presented on plate 1 (plate 1 is
in the pocket on cover 3 of this volume). The Gulf is long
and narrow, forming an actively extending graben between
the 5,000-ft-high peaks of the Sierra Pedro Martir (Baja
California batholith) and the Basin and Range province of
western Sonora (Stock and Hodges, 1989). Proposed mod
els that attempt to explain the tectonic relations and impli
cations of gulf formation and rifting are varied in their
interpretations of age, mechanisms, and rate of opening.
In 1980, Homestake Mining Company discovered the
McLaughlin orebody in Calistoga, California, and in 1984,
Gold Fields Mining Company found a large orebody that
they developed into the Mesquite mine near Yuma, Ari
zona. Both of these large gold deposits occur along faults
associated with the San Andreas fault system. Other pre
cious metal discoveries at the Picacho and Temco mines,
near Mesquite, have led to increased interest in mineral
deposits associated with transtensional faulting and rifting,
especially around the mouth of the Gulf of California
(plate 1). Recently, structural geologists and economic
geologists have joined forces to work out kinematic-con
trol and fluid-flow models for structurally associated ore
deposits; these models have direct application to regional
structural analyses (Sibson, 1987).
The abundance of gold deposits in southeastern Cali
fornia that parallel the San Andreas fault suggests the like
lihood of undiscovered deposits in northwest Mexico. The
San Andreas structural system continues southward from
California into North Baja, California and Sonora, eventu
ally becoming obscured by the Colorado River delta and
the sedimentary basins of the Gulf. The main transform
goes into the Gulf north of Puerto Penasco (PP on plate
1), but tensional graben faults and transtensional shear
zones can be traced on land for over 1,200 km down both
sides of the Gulf. Splays like the Imperial Valley fault and
Laguna Salada fault host active hydrothermal centers like
those at Cerro Prieto (CP on plate 1).
The continuation of the transform fault system in the
submarine environment has been studied by means of
bathymetric mapping, seismic reflection, and direct sub
marine observations (e.g., Von Damm and others, 1985;
Lonsdale, 1989; and many others). Few geologic maps
compare this detailed submarine information with surface
mapping on both sides of the Gulf. The relations between
submarine and subaerial tectonics need to be integrated to
look at the structures of the entire Gulf extensional region.
Analysis of tectonic features in the Gulf gives insight into
the principle stresses, styles of faulting, and relative tim
ing of offset. The results of this analysis can then be used
to assess and predict ore deposit occurrences around the
Gulf region. The map (plate 1) differs from previously
published maps of the Gulf of California because it com
bines mineral deposit locations with detailed structures.
By compiling structures and ore deposits on the same
map, it becomes possible to determine which deposits
occur along faultsthis compilation is useful for regional
geologic studies.
This paper (1) investigates the dip orientation and
types of faults most commonly hosting mineralization
around the Gulf of California, (2) quantifies the percent
age of the known occurrences that are associated with dif
ferent types of faults, (3) discusses how plate tectonics
and mineral-deposit studies together can be used to deter
mine the geologic history of the Gulf, and (4) presents
data showing that submarine rift-graben faults and strike
slip splays from transform faults can be matched with doc
umented structures on land.
PRINCIPAL DATA SOURCES AND
STUDY-AREA BOUNDARIES
Compilation of faults from bathymetric and seismic
reflection studies across the Gulf of California provides
excellent data to compile the portion of the map covering
the floor of the Gulf. In general, the Gulf is shallow, and
tectonic features are rapidly blanketed by sediment, mak
ing it difficult to map the exact location of faults (Larson,
1972). Abundant submarine pull-apart basins like the
Guaymas and Huatabampo Basins (GU and HU on plate
1) define zones of extension along the transform margin.
Submarine dives, discussed by Lonsdale and others
 BASE METAL MINERAL DEPOSITS, GULF OF CALIFORNIA, MEXICO 71

EXPLANATION

Area of plate 1

l1111 Region of Tertiary

extension

- - - - - - State boundary

---- - - International boundary

USA

*--
EXICO

CH
Area of report

Cocos plate

Figure 1. Gulf of California extended region and area covered in plate 1. The solid line outlines the limits of the structural
and mineral deposit compilation. Abbreviations are Arizona (AZ), California (CA), Chihuahua (CH), Colima (CO),
Durango (DU), Jalisco (JA), Nayarit (NA), Michoacan (MI), North Baja California (NBC), Sierra Pedro Martir fault
(SPMF), Sonora (SN), and South Baja California (SBC).
72 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
(1980), aid in defining submarine topography and hydro
thermal-fluid characteristics. Deep sea drilling results and
maps published in deep sea drilling reports add significant
tectonic-structural data useful in compiling maps (Curray
and Moore, 1982). Heat flow studies by Lawyer and Will
iams (1979) provide data that can be used to verify asym
metric extension and general structural regimes.
Published tectonic maps from eastern Baja California
by Rangin (1986), Hamilton (1971), Stock and Hodges
(1989), Lonsdale (1989), de Cserna (1989), Dokka and
Merriam (1982), Gastil and others (1975), Lopez Ramos
(1982), and Sociedad Programacin and Presupuesto (a
Mexican governmental agency) were combined with maps
showing mineral deposits from Wisser (1954), de la
Fuente Moises (1985), and Leonard (1989) to generate
plate 1. Additional field mapping and visits to mineral
deposit sites by the author where conducted from 1987
through 1992.
The map compilation for the western Mexican main
land combines data from Gastil and Krummenacher
(1977), Henry and Fredrikson (1987), Wahl (1973), and
Swanson and McDowell (1984) in addition to information
from the Baja California maps listed above. Mineral
resource maps by Leonard (1989), Gonzalez Reyna
(1956), Wisser (1966), Salas (1975), Consejo de Recursos
Minerales (1985, 1987), Consejo de Recursos Minerales
and State of Durango (1980), Universidad Autonima de
Mexico (197477), Anaconda Mining Company (unpub.
data), as well as files from the USGS Center for Mineral
Resource Investigations and mines shown on the geotec
tonic maps listed above were used to construct plate 1.
The boundaries of the Gulf extensional province are
difficult to define (see Stock and Hodges, 1989, for dis
cussion) but are important for identifying the scope of this
study. The western limit is defined by the eastern range
front fault of the Peninsular batholith (SPMF, fig. 1). West
of the major range-front normal fault, there is considerably
less faulting, and the batholith acts as a coherent block.
The range-front fault system continues south along the
east coast of the Baja Peninsula for nearly 800 km. In cen
tral Baja, the boundary becomes more difficult to define
because, in many places, normal and strike-slip faults are
covered by young volcanics, and where exposed, fault
scarps often become diffuse (e.g., Bahia de Los Angeles
(BL, plate 1), Santa Rosalia (SR, plate 1)).
In southern Baja, the escarpment follows the coast,
and, in places, the sea cliffs themselves mark the escarp
ment. Faulting at the southern tip of Baja likely occurred
before the early Miocene strike-slip motion that is associ
ated with oblique subduction west of the Baja Peninsula
(Spencer and Normack, 1979) and was not part of the late
Miocene opening of the Gulf. Faults extending from La
Paz to Cabo San Lucas are older than the rift faults but
may have been reactivated and further mineralized during
the rift-extensional events.
The eastern limit of extensional structures is also dif
ficult to define. On the east, the transition grades into the
Basin and Range province, and there is no single escarp
ment or main tectonic feature separating the Gulf from the
entire Basin and Range. Geologically, the Gulf with its
transtensional faulting is a continuation of basin-and-range
extension. The Sierra Madre Occidental is a comparatively
stable block and, in the southern part of the study area
(Sinaloa and points further south), it is the eastern limit of
the Gulf extensional province. In the northern part of the
study area, the boundary is limited to within 200 km east
of the eastern edge of the Gulf of California. This defini
tion is necessary to confine the study area to the Gulf
region and not the entire Basin and Range province.
DATA ANALYSIS
By comparing the tectonic features, hot springs, and
mineralization around the Gulf on the map compilation,
certain spatial and temporal correlations become apparent.
The mineral occurrences can be quantified by counting the
number of mineral occurrences that occur along major
faults versus those not on regionally mapped faults. For
the occurrences on faults, the direction and style (either
normal or strike-slip) can be tabulated. For all occur
rences, the principal commodities (gold, base metals, or
copper) are distinguished for later comparison among
deposit types.
The circum-Gulf region can be divided into 5 regions
based on geographic position and State in order to facili
tate comparison of structural and other data. North Baja
California, South Baja California, and Sonora are each
treated independently. Sinaloa and Durango comprise the
fourth region, and the southern mainland States of Nayarit,
Jalisco, and Colima are grouped as the fifth region. A tab
ulation of the number of mineral occurrences associated
with strike-slip faults, normal faults, and those not associ
ated with any fault using the five geographic groups is
presented in table 1. Histograms based on this data table
showing comparative features of structural control and
mineral deposits around the Gulf are presented in figures
that follow.
Comparing the various types of regional information
highlights some broad correlations. Firstly, the percentage
of mineral occurrences located on mapped faults differs
throughout the Gulf of California region, as shown on fig
ure 2. In North Baja and Sonora, the number of mineral
occurrences on faults greatly exceeds those not coincident
with structural features. However, in the southern part of
the study area (South Baja, Sonora, and Sinaloa-Durango),
the correspondence is only 50 percent, and in the southern
mainland region, only about 30 percent of the mineral
occurrences are on faults. The difference in mineralization
and structural control around the Gulf indicates that there
 BASE METAL MINERAL DEPOSITS, GULF OF CALIFORNIA, MEXICO 73

Table 1. Number of mineral deposits associated with each type of Mineral occurrences
structure in the Gulf of California extended region.
on faults
[The southern mainland Mexico group (S. mainland Mex.) includes the States of
Nayarit, Jalisco, and Colima) Mineral occurrences
not on faults

Strike-slip Normal Not on Total

fault fault fault number

Au-Ag-base-metal deposits
N. Baja 2 6 1 9
S. Baja l 3 2 6
Sonora 7 2 4 13
Sinaloa-Durango 7 4 8 19
S. mainland Mex. 6 () 19 25

Gold-dominated deposits
N. Baja 9 8 4 21
S. Baja 2 l () 3
o co to - to
Sonora 6 2 3 |1 ~ *- -> C
CO CO 2 O ~
Sinaloa-Durango 2 () () 2 - C
2 C/D C/D 5
S. mainland Mex. () () () () v) E
Copper deposits Regions v.
N. Baja 2 3 l 6
S. Baja () l 5 6 Figure 2. Comparison of mineral occurrences on faults to
Sonora 3 () 7 10 those not on faults in each of five regions: North Baja (N.
Sinaloa-Durango () () 1 l Baja); South Baja (S. Baja); Sonora; Sinaloa and Durango (Sina
S. mainland Mex. () l () l Dur); and the southern mainland group of Nayarit, Jalisco, and
Colima, (S. mainland).
Totals: 47 31 55 133

in the northern Gulf region, especially in North Baja (fig.

is a variation in the importance of structural control and a
corresponding difference in deposit type. In North Baja
and Sonora, the higher correspondence suggests that min
eralization has a first-order correspondence to faulting
such that fluids may be channeled through structures,
whereas south of North Baja and Sonora, the occurrences
do not correspond with major fault structures and, instead,
are found throughout the area. Part of this disparity could
be due to the varying quality of the data but, because
maps were used of comparable scales and from similar
sources, the observations are clearly not simply an artifact
of the source mapping.
Combining all 133 occurrences and looking at the
overall picture, 58 percent of the mineral occurrences are
associated with some structure; 35 percent occur on nor
mal faults; and 23 percent occur on strike-slip faults. The
Not on fault column in table 1 indicates the number of
mineral occurrences not coincident with a mapped fault.
These percentages are important because they show that,
over the entire Gulf region, a large percentage of the pre
cious- and base-metal deposits are associated with major
mapped structures and that the association between struc
tures and deposits is particularly high in North Baja and
3). The relation between mineral-deposit composition and
region can be compared with the earlier observation that
structural association with mineral occurrence is highest in
North Baja California and Sonora. From the compilation,
it appears that the gold deposits lie along major strike-slip
structures more often than do the Ag-Pb-Zn veins.
The direction of dip of the faults that host mineral
occurrences is independent of the side of the Gulf. In the
simple extensional model often proposed for rift areas, the
structures are drawn dipping toward the center of the rift.
But in the Gulf of California, mineral occurrences are
present in approximately equal numbers on east- and west
dipping structures on both sides of the Gulf. Figure 4 is a
plot of number of mineral occurrences on east-dipping
normal faults plotted next to the number of occurrences on
west-dipping faults. The number of occurrences on east
dipping normal faults and west-dipping normal faults is
nearly identical for each region. An interesting idea that
may be drawn from this is that mineralizing fluids may
not be derived from near the center of the rift and flow
outward along faults, and instead, they may be regionally
dispersed and localized near individual hydrothermal
CenterS.

Sonora. Most of the hot springs in the Gulf extensional prov

A comparison of different deposits by region shows ince occur along the western margin of the Gulf and
that base-metal occurrences are highly concentrated on the within the Gulf, along the rift transform and associated
southern mainland and that gold deposits are concentrated extensional grabens. Hot springs on the eastern coast of
 74 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Au-dominated deposits

Base-metal-dominated N
deposits

Cu-dominated deposits

i
!

\
,
( CH
} MEXICO

; :
Regions |
Figure 3. Comparison of mineralization types (showing the ma
jor metallic element) in each of five regions.
O 400 KILOMETERS
|

East-dipping
normal fault

Figure 5. Generalized locations of active and recently active

20 - West-dipping
gn normal fault
hot springs around the Gulf of California that are shown on plate
8 18 1. Hot springs are marked by star; abbreviations are: North Baja
E 16 -
(NB), South Baja (SB), Sonora (SN), Chihuahua (CH), Sinaloa
(SI). Durango (DU), Nayarit (NA), Jalisco (JA), Colima (CO),
* 14 Michoacan (MI).
3
O
12
o 10 -
'o 8
5 Eight of the l l hot-spring areas that occur on land around
3 6 -
the Gulf are along faults. The other three possibly connect
5 4 with structures at depth, but no fault is identified on the
* 2 | -- surface. The springs are important because they could be a
0 [
.0, to co - to
modern analog to the formation of epithermal mineral
co 3 5 deposits and they are usually associated with a major
CO CO 5 & E
fault.
2
- -

C/D C/D
C
*7, :
--

w
Regions
Figure 4. Diagram of dip directions of mineralized normal faults
around the Gulf of California.
Baja occur along faults that can be traced back to active
transforms in the Gulf. Cerro Prieto (CP) and other hot
spring areas have mineralized areas at the surface and
extensive alteration along local structures (Elders and oth
ers, 1981). A simplified hot-spring location map is pre
sented to show the location of geothermal fields (fig. 5).
TECTONIC ANALYSIS
The geology of the Gulf of California has been stud
ied in detail, and the references cited in the Principal Data
Sources and Study-Area Boundaries section of this report
provide many of the basic observations used in outlining
the geologic history of the Gulf. However, from the map
and data presented in this compilation, several newly
observed features become apparent. To put these in the
tectonic context, we will first briefly summarize the salient
tectonic events.
 BASE METAL MINERAL DEPOSITS, GULF OF CALIFORNIA, MEXICO 75

Baja California was originally part of mainland Mex EXPLANATION

ico, with subduction extending along the eastern Pacific
Ocean margin for over 5,000 km from northwestern Normal fault-bounding
major extension
Alaska to Central America. The continental crust sur
6 million year distance
rounding the Gulf lies near the edge of the North Ameri l of movement - relative
---, to North American craton
can craton and is inherently weak due to its long and
complex history of continental-margin sedimentation, tec
tonic accretion, and magmatism. These inherited weak
nesses became future zones of failure and eventually gave
rise to the Gulf and the development of a transtensional
continental margin. During the Miocene, much of the
future site of the Gulf was occupied by andesitic arc vol
canoes overlying lithosphere that was subducted along the
western margin of Baja California. To the east, the south
ern Cordillera was extending and low-angle detachments
extended the Gulf region, forming a graben known as the
protogulf. Differing accounts of the events leading to the
opening of the protogulf are proposed by Karig and Jen
sky (1972), and Moore (1973). More recently, Gastil and
others (1975, 1981), Spencer and Normack (1979), and
Atwater (1989) brought new insights into the timing and
order of events that led to the opening of the protogulf,
and these events continue to open the Gulf today. Origi
O 500 KILOMETERS
nally, the protogulf concept was proposed to explain the 1
anomalously old oceanic crust in the mouth of the Gulf
adjacent to the Mexican margin (Moore and Buffington,
1968). More recently, this concept has been expanded to Figure 6. Reconstruction of protoGulf of California, showing
include late Miocene extensional faulting, magmatism, and width of gulf at 6 Ma and direction and distance moved in the last
marine sedimentation in north and central Gulf areas 6 million years relative to the North American plate. Diagram
shows extent of study area within the extensional regime of north
(McLean and others, 1987).
west Mexico. The solid line outlines the geographic shape of the
Subduction along the western Baja California conti Gulf at 6 Ma, and the dotted line outlines the present-day geograph
nental margin ceased 1210 Ma, and east-west extension ic position. 6SD, location of San Diego 6 million years ago; 0SD,
marked by basin-and-range faulting and alkali-basalt vol present location of San Diego. Tucson (TS) and Hermosillo (HE)
are reference cities.
canism increased. The structurally weak zone, which was
already developing inland of the continental margin, sub
sided even more, and the principal plate boundary shifted faults that host many of the high-angle, fault-related epith
250 km eastward into the extended continental crust at 6 ermal deposits in Sonora and Baja California.
Ma. Over the past 6 m.y., the Gulf has opened approxi Represented on the map of structures (plate 1) are
mately 300 km. Figure 6 is a reconstruction of the position varying stages of superimposed extension and strike-slip
of Baja relative to North America at 6 Ma. The recon faults. In the Laguna Salada Fault zone (LSF on plate 1),
struction closes some of the extension and is tied by simi strike-slip faults cut the normal faults in some areas,
lar conglomerate and volcanic sequences across the now whereas in other areas normal faults displace strike-slip
open Gulf. The Pacific and North American plates were shear zones. The compilation also shows that splays off
moving 1020 obliquely to the early San Andreas trans the main submarine transform can be traced onto land
formthis caused opening of small intra-Gulf pull-apart along the eastern edge of the Baja Peninsula. Finally, fur
basins linked by en echelon faults (Atwater, 1989). These ther study of the volcanic sequences and age determina
faults not only opened basins but also propagated on land, tions on the volcanosedimentary stratigraphy can be used
offsetting upper Miocene sedimentary and volcanic rocks. with this compilation of structures to constrain the com
According to Atwater (1989), by 6 to 4 m.y. ago, the mar plex paleo-stress history.
gin strengthened, probably cooled and became more rigid,
and a weaker inland zone began to accommodate shearing
between the two plates. Elevated heat flow and hydrother DISCUSSION AND CONCLUSIONS
mal waters were liberated due to newly formed crustal
scale transform shears. It is these transform faults, associ This paper presents a structural and mineral-occur
ated pull-apart graben faults, and reactivation of normal rence compilation that is useful in connecting submarine
76 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
and subaerial geologic mapping. From this structural com
parison, it is apparent that fault splays and en echelon
transform faults extend significant distances away from
the main transform plate boundary (e.g., Puertocitos, Isla
La Guardia, etc.) and act as conduit structures for mineral
izing fluids. The Gulf region plate boundary is a zone of
rifting and transtensional shear through which heat is con
vected and hydrothermal fluids evolve. These fluids may
transport metals, carrying them into and along fault struc
tures in over 50 percent of the known epithermal mineral
occurrences. Present-day geothermal activity in the Gulf
seems to be continuing this production and convection of
enriched fluids in the Gulf province. Compiled mineral
occurrences show that structural control is a significant
factor for epithermal mineral deposits, particularly in the
gold-dominated deposits and especially in Sonora and
northern Baja California.
Oxide and sulfide copper deposits, such as those at
Boleo and southwest of Caborca, may be related at depth
to large structures, but these structures are not mapped at
the surface. Perhaps additional geophysics and satellite
image interpretation will be useful in not only locating
alteration areas but also major faults. Copper porphyries
are older than the Gulf extension and are not included in
this project. Detailed maps for El Arco and porphyries in
Sonora and Sinaloa show that major structures play a criti
cal role in magma emplacement and hydrothermal fluid
flow.
The greater abundances of hot springs in the western
Gulf rather than in the eastern Gulf indicates that an asym
metry exists across the Gulf. This suggests the possibility
of hotter crust and more active tectonics to the west. Heat
flow studies by Lawyer and Williams (1979) support this
idea by showing increased heat flow in the western Gulf.
Another observation of apparent asymmetry is that fault
ing on the western side of the Gulf is confined to a narrow
strip, within a few tens of kilometers of the coastline.
Faulting in western mainland Mexico, on the other hand,
extends from the Gulf extensional region hundreds of kilo
meters eastward into the southern Basin and Range exten
sional regime (plate 1). The Gulf of California is also
asymmetric in that it has a thinner crust in the west com
pared to that in the east (Gastil and others, 1975). The
thinner crust has higher heat flow, which may drive the
geothermal anomalies and which may concentrate the hot
spring-type gold mineralization along the western side of
the Gulf, particularly in North Baja, where the transform
comes on land (i.e., Imperial fault, Laguna Salada fault,
etc.).
Another significant observation based on the map
compilation (plate 1) is that faults originating at or near
the transform extend long distances away from transforms
and can be matched with faults mapped on land, thus
tying the submarine-seafloor maps with circum-Gulf land
maps. For example, in the Baha Los Angeles area (BL on
plate 1), strike-slip faults come on land and host small
gold concentrations in Pliocene sandstones. And at Puer
tocitos (PU on plate 1), the Puertocitos fault (a splay off
the main transform) comes on land and hosts the hot
spring-type Au occurrence. In Puertocitos, active hot
springs in Pliocene and Quaternary volcanics are localized
along this same splay. This has important implications for
mineral exploration in the Gulf of California and other
similar regions where submarine faults could be traced on
land to discover new deposits.
Other transtensional settings, like the Red Sea, also
show regional structural variations, and these appear to be
controlled by preexisting structurally weak zones. In the
Red Sea, the rift is superimposed upon deformed Precam
brian basement. The preexisting weaknesses control the
location of later structural failure. The Gulf of California
inhabits a structurally weak trough, and the asymmetries
can be explained from preexisting structures. Along the
Atacama fault in western Chile, brittle transtensional shear
structures host many gold deposits, yet, unlike the Gulf of
California, the Atacama fault does not have hot-spring
type features (Brown and others, 1991).
Around the Gulf of California, the epithermal metal
systems are gold dominated in the north and base metal
dominated in the souththis could be caused by vertical
zoning, crustal contamination, fluid chemistries, or timing
and spatial relations to fluid and heat sources. The chemi
cal variations might be due to preferential metal complex
ing and carrying of different metals or fluid dynamics and
water-rock interactions. The fluid chemistry could likely
be constrained by fluid-inclusion studies and detailed pet
rographic analyses of deposits throughout the region; how
ever, this information is not available. Other options, such
as differences in basement composition or magmatic
sources could explain the variation. In North Baja and
Sonora, mineralization is commonly very shallow, with
aprons of hot-spring cinter, open-space cavities, and reed
and plant detritus that is silicified in auriferous hematitic
silica in the Tertiary volcanic and sediment-hosted occur
rences. In the eastern Baja Peninsula, mesothermal quartz
veins are common in metamorphosed Paleozoic vol
canosedimentary rocks and granitic intrusions. These
veins are probably related to mid to late Mesozoic
batholithic plutonism and not to the waning stages of Ter
tiary subduction and the onset of extension in the Gulf of
California. The mesothermal veins are not included in
this study because they likely predate Gulf extension. The
vein deposits on the mainland are of a higher temperature
type than those that commonly contain tourmaline, quartz,
pyrite and minor sericite. These veins are, in general,
breccia zones with a quartz + carbonate matrix, and their
orientation varies as does their age. In this preliminary
analysis, the veins and ore breccias are simply compared
 BASE METAL MINERAL DEPOSITS, GULF OF CALIFORNIA, MEXICO
in terms of their metal association. Further age dating will
help constrain the relative timing of different mineraliza
tion events.
In conclusion, the map compilation (plate 1) provides
a timely summary of existing tectonic data and correlates
this data with a new mineral deposit map for western
Mexico. Many of the epithermal mineral deposits are
structurally controlled, particularly the gold-rich deposits
in North Baja and Sonora. The type of mineralization
changes from gold rich in the north to base metal rich in
the south. Finally, gold deposits are most often associated
with strike-slip faults, whereas the base-metal deposits
occur in roughly equal numbers on both normal and
strike-slip structures.
ACKNOWLEDGMENTS
This paper has benefited from careful review by
Geoff Plumlee, Robert Kamilli, and Floyd Gray. Earlier
drafts were reviewed by Peter Coney and Spencer Titley
of the University of Arizona. Acknowledgments are
extended to Mark Barton, Joaquin Ruiz, and Norm Page,
who have contributed to discussions of the data and inter
pretations. Felipe Rendon, Porfirio Padilla, Dave Johnson,
and Ryan Mouser accompanied me in the field and helped
obtain hard-to-find Mexican geologic data. Thanks are due
to Perry Durning of Fisher-Watt Gold Company and to the
many other property owners who allowed access to their
properties over the past five years while this study was
completed.
REFERENCES CITED
Agricola, G., 1556, De Re Metallica, [English translation by H.C.
Hoover and L.H. Hoover, 1950), New York, Dover Publica
tions, 130 p.
Atwater, T., 1989, Plate tectonic history of the northeast Pacific and
western North America: Geological Society of America,
Decade of North American Geology, Geology of North Amer
ica, Volume N, p. 2172.
Barley, M.E., and Groves, D.I., 1988, Geological setting of gold
mineralization in the Norseman-Wiluna Belt, Eastern Gold
fields Province, Western Australia: Geological Society of Aus
tralia, Bicentennial Gold 88, Excursion no. 6, p. 1846.
Brown, M., Daz, F., and Grocott, J., 1991, Atacama Fault system;
history of displacement and tectonic significance for the Meso
zoic-Recent evolution of northern Chile: Santiago, Chile, Con
greso Geologico Chileno, Resumenes Ampliados, p. 129132.
Consejo de Recursos Minerales, 1985, Geologic and mineral dis
trict map of Durango: Consejo Recursos Minerales Publica
tion, scale 1:500,000, 2 sheets.
Consejo de Recursos Minerales, 1987, Plano de recursos mineros
activos y potenciales del Estado de Nayarit: Consejo Recursos
Minerales unpub. map, scale 1:400,000.
77
Consejo de Recursos Minerales and State of Durango, 1980, Topo
graphic and mineral commodities map of Durango, [1st ed.]:
Consejo Recursos Minerales Publication, scale 1:500,000.
Cserna, Z. de, 1989, An outline of the geology of Mexico: Geolog
ical Society of America, Decade of North American Geology,
Geology of North America, Volume A., p. 233264.
Curray, J.R., and Moore, D.G., eds., 1982, Reports of the Deep Sea
Drilling Project, v. 64, Parts I and II: Washington, U.S. Gov
ernment Printing Office, 260p.
Dokka, R.K., and Merriam, R.H., 1982, Late Cenozoic extension of
northeastern Baja California, Mexico: Geological Society of
America Bulletin, v. 93, p. 371378.
Elders, W.A., Hoagland, J.R., and Williams, A.E., 1981, Distri
bution of hydrothermal mineral zones in the Cerro Prieto
geothermal field of Baja California, Mexico: Geothermics, v.
10, p. 245253.
Fuente Moises, M. de la, 1985, Analisis metalogenetico regional de
la Peninsula de Baja California: Mexico City, Universidad Au
tonima de Mexico, unpub. thesis, 156 p.
Gastil, R.G., and Krummenacher, D., 1977, Reconnaissance geolo
gy of coastal Sonora between Puerto Lobos and Bahia Kino:
Geological Society of America Bulletin, v. 88, p. 189198.
Gastil, R.G., Morgan, G.J., and Krummenacher, D., 1981, The
tectonic history of peninsular California and adjacent Mexi
co, in Ernst, W.G., ed., The Geotectonic Development of
California, Rubey Volume 1: Englewood Cliffs, N.J., Pren
tice-Hall, p. 284306.
Gastil, R.G., Phillips, R.P., and Allison, E.C., 1975, Reconnais
sance geology of the State of Baja California: Geological Soci
ety of America Memoir 140, 170 p.
Gonzalez Reyna, J., 1956 [updated in 1966], Los criadores de Au,
Ag, Pb, Zn, Cuen la Republica Mexicana: Mexico City, Twen
tieth International Geologic Congress, scale 1:2,000,000.
Groves, D.I., and Barley, M.E., 1988, Part 2: Gold mineralization
in the Norseman-Wiluna Belt, Eastern Goldfields Province,
Western Australia: Geological Society of Australia, Bicenten
nial Gold 88, Excursion no. 6, p. 4866.
Guha, J., Archambault, G., and Leroy, J., 1983, A correlation
between the evolution of mineralizing fluids and the geo
mechanical development of a shear zone as illustrated by
the Henderson 2 mine, Quebec: Economic Geology, v. 78,
p. 16051618.
Hamilton, W., 1971, Recognition on space photographs of struc
tural elements of Baja California: U.S. Geological Survey Pro
fessional Paper 718, 26 p.
Henry, C.D., and Fredrikson, G., 1987, Geology of part of southern
Sinaloa, Mexico adjacent to the Gulf of California: Geological
Society of America, Map and Chart Series MCH063, [text and
map), scale 1:250,000.
Hulin, C.D., 1925, Structural control of ore deposition: Economic
Geology, v.24, p. 1549.
Karig, D.E., and Jensky, W., 1972, The protoGulf of California:
Earth and Planetary Science Letters, v. 17, p. 169174.
Larson, R.L., 1972, Bathymetry, magnetic anomalies, and plate tec
tonic history of the mouth of the Gulf of California: Geologi
cal Society of America Bulletin, v. 83, p. 33453360.
Lawyer, L.A., and Williams, D.L., 1979, Heat flow in the cen
tral Gulf of California: Journal of Geophysical Research, v.
84, p. 34653478.
78 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Leonard, K.R., 1989, Preliminary deposit-type map of northwest
ern Mexico: U.S. Geological Survey Open-File Report
89158,330 p.
Lindgren, W., 1933, Mineral Deposits: New York, McGraw-Hill,
930 p.
Lonsdale, P.F., 1989, Geology and tectonic history of the Gulf of
California: Geological Society of America, Decade of North
American Geology, The Geology of North America, Volume
N, p. 499521.
Lonsdale, P.F., Bischoff, J.L., Burns, V.M., Kastner, M., and
Sweeney, R.E., 1980, A high-temperature hydrothermal de
posit on the seabed at a Gulf of California spreading center:
Earth and Planetary Science Letters, v. 49, p. 820.
Lopez Ramos, E., 1982, Geologia de Mexico, [3rd ed.]: Mexico
City, Universidad Autonima de Mexico, 454 p.
McLean, H., Hausback, B.P., and Knapp. J.H., 1987, The geology
of west-central Baja California Sur, Mexico: U.S. Geological
Survey Bulletin 1579, 16 p.
Moore, D.G., 1973, Plate-edge deformation and crustal growth,
Gulf of California structural province: Geological Society of
America Bulletin, v. 84, p. 18831906.
Moore, D.G., and Buffington, E.C., 1968, Transform faulting and
growth of the Gulf of California since the late Pliocene: Sci
ence, v. 161, p. 12381241.
Poulsen, K.H., and Robert, F., 1989, Shear zones and goldprac
tical examples from the southern Canadian Shield, in Bursnall,
J.T., ed., Mineralization and Shear Zones: Geological Associ
ation of Canada, Short Course Notes 6, p. 239266.
Rangin, C., 1986, Contribution a L'etude geologique du systeme
Cordillerain Mesozoidue du nord-ouest du Mexique; Une
coupe de la Basse Californie centrale a la Sierra Madre occi
dentale en Sonora: Paris, Memoires de la Societe Geologique
de France, Memoire no. 148, 133 p.
Robert, F., 1990, Structural setting and control of gold-quartz veins
of the Val D'Or area, southeastern Abitibi Subprovince, in Ho,
S.E., Robert, F., Groves, D.I., eds., Gold and Base-Metal Min
eralization in the Abitibi Subprovince, Canada, with Emphasis
on the Quebec Segment: The University of Western Australia,
Short Course, p. 167209.
Robert, F., and Brown, A.C., 1986, Archean gold-bearing quartz
veins at the Sigma Mine, Abitibi greenstone belt, Quebec; Part
1. Geologic relations and formation of the vein system: Eco
nomic Geology, v. 81, p. 578592.
Salas, G.P., 1975, Carta y provincias metalogeneticas de la Repub
lica Mexicana: Consejo de Recursos Minerales, Publication 21
E., Scale 1:2,000,000.
Sibson, R.H., 1981, Fluid flow accompanying faulting; field evi
dence and models, in Simpson, D.W., Richards, P.G., eds.,
Earthquake Prediction, an International Review: American
Geophysical Union, Maurice Ewing Series, v.4, p. 593603.
Sibson, R.H., 1986, Brecciation processes in fault zones: Pure and
Applied Geophysics, v. 124, p. 159175.
Sibson, R.H., 1987, Earthquake rupturing as a mineralizing agent in
hydrothermal systems: Geology, v. 15, p. 701704.
Spencer, J.E., and Normack, W.R., 1979, Tosco-Abreojos fault
zone; a Neogene transform plate boundary within the Pacif
ic margin of southern Baja California, Mexico: Geology, v.
7, p. 554557.
Spurr, J.E., 1916. The relation of ore-deposition to faulting: Eco
nomic Geology, v. 11, p. 601622.
Stock, J.M., and Hodges, K.V., 1989, Pre-Pliocene extension
around the Gulf of California and the transfer of Baja Califor
nia to the Pacific plate: Tectonics. v. 8, no. 1, p. 991 15.
Swanson, E.R., and McDowell, F.W., 1984, Calderas of the Sierra
Madre Occidental volcanic field, western Mexico: Journal of
Geophysical Research, v. 89, p. 87878799.
Universidad Autonima de Mexico, 19741977, Carta Geologica de
Sinaloa, Mexico: U.N.A.M. Instituto de Geologia, scale
1:100,000, [series of 32 maps).
Von Damm, K.L., Edmond, J.M., Measures, C.I., and Grant, B.,
1985, Chemistry of submarine hydrothermal solutions at
Guaymas Basin, Gulf of California: Geochimica et Cosmo
chimica Acta, v. 49, p. 22212237.
Wahl, D.E., 1973, Geology of the El Salto strip, Durango, Mexico:
Austin, University of Texas, unpub. M.S. thesis, 112 p.
Wisser, E., 1954, Geology and ore deposits of Baja California,
Mexico: Economic Geology, v. 49, no. 1, p. 4476.
Wisser, E., 1966, The epithermal precious-metal province of north
west Mexico: Nevada Bureau Mines Geological Report No.
13, p. 6392.
 CHAPTER H
CENOZOIC GEOLOGY AND MINERAL DEPOSITS
OF THE BERENGUELA DISTRICT,
NORTHWESTERN BOLIVIA
By ALAN R. WALLACE,' RICHARD F HARDYMAN, RICHARD M. TOSDAL. NESTOR
JIMENEZ, JOSE LUIS LIZECA, and FERNANDO MURILLO"
ABSTRACT
The Berenguela Ag-Pb-Zn-Cd district of the north
western Bolivian Altiplano was the site of episodic mid to
late Tertiary deformation during basin sedimentation and
explosive volcanism. Oligocene continental sedimentation
produced the Berenguela Formation and the lower member
of the Mauri Formation, both of which were buried by
ignimbrites and volcaniclastic sediments of the Miocene
upper member of the Mauri Formation. Dacitic to rhyolitic
domes, lahars, and ash-flow tuffs were emplaced during
middle Miocene to Pliocene time. Episodic tectonism and
associated broad folding and high-angle faulting produced
major unconformities in the sedimentary record.
The district contains two types of base- and precious
metal deposits. Late Oligoceneearly Miocene(?) cad
mium-rich veins formed along fractures in the Berenguela
and lower member of the Mauri Formations, and middle
Miocene or younger cadmium-poor veins formed in
Miocene volcaniclastic and intrusive rocks. The different
chemistries and mineralogies suggest different ages and
metal sources for the two vein systems. Distinct thoro
genic Pb isotopic compositions for galena in the two types
of veins are consistent with this interpretation.
'U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, University of Nevada, Mackay Building,
Reno, NV 89557.
*U.S. Geological Survey, Mail Stop901,345 Middlefield Road, Menlo
Park, CA 94025.
Servicio Geolgico de Bolivia, Calle Frederico Zuazo No. 1673,
Casilla 2729, La Paz, Bolivia.
INTRODUCTION
The Berenguela district, approximately 160 km south
west of La Paz, Bolivia (fig. 1), consists of a network of
mineralized faults that have been mined intermittently since
Spanish colonial times, in the 1500's. The ores are hosted
by Cenozoic continental sedimentary rocks, inviting com
parison with the Corocoro district to the east, and by
Miocene volcanic and intrusive rocks. Although the Beren
guela district is of only minor economic importance, geo
logic relations provide many clues to the Cenozoic history
of the northwestern Altiplano. Recent mapping by U.S.
Geological Survey (USGS) and the Servicio Geolgico de
Bolivia (GEOBOL) geologists redefined the stratigraphic
and tectonic history of the region and placed the mineral
deposits in a more regional geologic framework.
REGIONAL GEOLOGIC SETTING
The Berenguela district, which surrounds the small
town of the same name, is in the western Altiplano (fig.
1), the middle of three Andean morphotectonic elements
that formed in response to subduction of the Nazca plate
beneath the South American craton along the western mar
gin of the continent (Baranzagi and Isacks, 1976). To the
west, the Cordillera Occidental is underlain by Proterozoic
to Tertiary metamorphic, igneous, and sedimentary rocks
and is capped by a chain of Oligocene to Holocene strato
volcanoes. To the east, the Cordillera Oriental is domi
nated by thin-skinned thrust belts (Jordan and others,
1983; Sheffels, 1990). The Altiplano is underlain by an
easterly thickening section of Tertiary sedimentary rocks
that overlie, at varying depths, Proterozoic to Mesozoic
sedimentary and crystalline rocks.
80 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

72- 64-
-
- -
68"
- -
-

SEDIMENTARY AND
B RA ZI L IGNEOUS ROCK UNITS
\ 27 -----

c- ! 2^ l
N /* The rocks of the Berenguela district record Oligocene
Ys--S w, i |
and younger clastic sedimentation and plutonic and volca
RSS \ | nic igneous activity. The oldest rocks exposed in the area
A N & ^
are continental redbeds of the Oligocene Berenguela For
| PERU N *... -->.S.
N (q, **~ mation (fig. 2), which includes more than 1,000 m of red
ttzo N. ! *. \ *4. S. ~
quartz arenite, siltstone, and mudstone. Thin conglomerate
14 H
N.
\
/
, N
&
,
^*
|
:

beds are present locally, and they contain subrounded
S ( c. N *, S o / clasts of granite, gneiss, chert, and quartzite (110 cm in
\ N.) N N \
diameter), which were presumably derived from either a
*Nsv
A.
Aar
AREQUIPA
S :
** N
*
*>
&
N
N
\.
". pre-Tertiary source or reworked from older Tertiary con
N. Y/ v N- ~o N glomerate layers. Bedding is generally parallel, although
N Fs{
: o,
e
Coc HABAMBA | 4,
e sania ("Ruz
Y~
* channelling is locally common beneath massive sandstone
18- o oruro \ -

.* N p . layers. The uppermost unit of the Berenguela exposed in

* \\ , \ . | the district is a massive sandstone. The red coloration is a
\\
2 < 2
o
(s.
N ". | ...A ----------
product of diagenetic to post-diagenetic alteration of what
*r \ . * | / ever mafic minerals (biotite, hornblende, and magnetite)
\, \ }, \ | may have been present in the sediments. The Berenguela
\g\\o | IARUA | /:
7 Formation has been correlated with redbeds of the Tihua
- V \l : '4'
22 \\Q_/\ |
, !-- S
j \/
-- ". e. . ".
nacu Formation north of La Paz, the upper part of which
CHI L E \; (Z 21- y
N. 7
Y z
-

\ A has an age of 29 Ma (Swanson and others, 1987), and the

e ANTityraasia
Y ||}

>

/
A R G E N T IN A
I
.

N.
Corocoro Formation, which has a late Oligocene to early
Miocene age (1725 Ma; data from A.H. Clark, cited in
- o too 200 300 MILES

H Flint, 1990). Therefore, the Berenguela Formation is prob

o 100
Figure 1. Location of the Berenguela, Bolivia, area with respect
to major morphotectonic features of the central Andes. Heavy
dashed lines separate morphotectonic elements.
Arc-related igneous activity related to the Andean
cycle has persisted since Mesozoic time. Jurassic, Late Cre
taceous, and Paleocene magmas were emplaced along the
western cratonic margin, west of the modern chain of vol
canoes. After a period of igneous and tectonic quiescence,
magmatic activity was renewed in the late Oligocene, pro
ducing the central volcanic zone, which extends from
southern Peru into Chile (Thorpe, 1984). The western part
of the zone is a chain of Miocene to Holocene andesitic
stratovolcanoes along the crest of the Cordillera Occidental.
Many of the identified structures in the Altiplano
region are products of late Oligocene and younger con
traction and uplift related to subduction and formation of
the Andean orocline. The region was tectonically inactive
from the late Eocene until the late Oligocene, and conti
nental clastic sediments were shed from the eroding Cor
dillera Occidental to produce a subdued landscape (Tosdal
and others, 1984; Jordan and Alonso, 1987). Renewed epi
sodic tectonism produced folds and thrusts in the eastern
Altiplano (Jordan and others, 1983) and rapid uplift and
multiple stages of pedimentation in the western Altiplano
(Tosdal and others, 1984; McKee and Noble, 1989).
200 300 kilomeTERs
ably of Oligocene age as well.
The Mauri Formation overlies the Berenguela Forma
tion, and it reflects a progressive change from continental
clastic sedimentation with a distal crystalline source
(lower member) to volcaniclastic sedimentation with a
nearby volcanic-arc source for the sediments and interca
lated ignimbrites (upper member). The lower member
(Mauri 1 and 3 of Sirvas C., 1964) is arkosic sandstone to
siltstone, with a biotite- (and, near the base, plagioclase-)
rich mineralogy that is distinct from that of the Berenguela
Formation. Rare conglomerate units near the base of the
Mauri contain subrounded pebbles to boulders of gneiss
and granite (some with Late Proterozoic U-Pb-isotopic
ages; R. Tosdal, unpub. data, 1991). Higher in the section,
conglomerates are common and contain cobbles of distinc
tive basaltic volcanic rocks and minor sandstones in addi
tion to clasts of gneiss and granite. The volcanic rocks are
petrologically similar to those in the Abaroa Formation,
which crops out some 50 km to the west and southwest.
The conglomerate is coarser to the southwest of the town
of Berenguela and must reflect source-region uplift to the
west and south. Andesitic dikes and sills occur within the
lower member (Mauri 2 and 4 of Sirvas C., 1964); these
have been dated at 25.2+ 1.0 Ma (lava flow in Mauri 2;
Lavenu and others, 1989) and 25.6 Ma (lava flow in
Mauri 4: Evernden and others, 1977).
The upper member of the Mauri Formation is com
posed of a heterogeneous sequence of principally volcani
clastic tuffaceous sandstone and conglomerate and
GEOLOGY AND MINERAL DEPOSITS, BERENGUELA DISTRICT, NORTHWESTERN BOLIVIA 81

6930' 6915'
1700'
Z \ |

e Santiago de Machaca
Q

Co. Huaricunca
1715'

Trnu

- is p
- Q Q
- Ta
|
1730'
0 5 10 MILES
| |
T I
0 5 10 KILOMETERS

EXPLANATION

Q Sedimentary rocks and alluvium, Ta Abaroa Formation (upper Oligocene

undivided (Quaternary) to lower Miocene)

Tp Perez Ignimbrite (Pliocene-Miocene) * andesitic section

Lavas and pyroclastic rocks, Cerke and Mauri Formation, lower member
Tcy ... -- Trn!
Huaricunca volcanoes (upper Miocene) (Oligocene)

-
Fi Intrusive
middlerocks, undivided (upper to
Miocene) Tb Berenguela Formation (Oligocene)
-

Tmu Mauri Formation,

to middle upper member (upper
Miocene) Fault

Figure 2. Generalized geologic map of the Berenguela region. CT, Cerro Tatitocollo prospect; DA,
Dos Amigos prospect; X, mine or group of mines. Based on unpublished mapping (1990) by the au
thors of this paper.
82 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
ignimbrite. As such, the upper member represents a pro
found change in type and source of sedimentation from
the rocks of the lower member. The units in the upper
member, previously included as Mauri 5 And 6 on the east
side of the area and the Cerke Formation to the west (Sir
vas C., 1964), represent dynamic and rapid basin filling
east of an active, explosive volcanic source. Units on the
east side of the Berenguela district include interbedded
sandstone, conglomerate, reworked tuffaceous sediments,
and pyroclastic airfall deposits. To the west, the section is
composed of predominantly pyroclastic deposits derived
from a chain of volcanoes that include the modern Cerros
Cerke and Huaricunca. Basalt-rich conglomerates in the
Mauri east of Berenguela probably were derived from the
late Oligocene to middle Miocene Abaroa Formation to
the south. To the east, the younger rocks disconformably
overlie the upper part of the lower member of the Mauri,
and to the west, they unconformably overlie the Beren
guela and lower part of the lower member of the Mauri
Formations. In the central part of the Berenguela district,
isolated remnants of a locally derived, trough-filling sedi
mentary breccia overlie the Berenguela and the lowermost
part of the Mauri Formations. A tuff in the upper member
of the Mauri, northeast of Berenguela, has been dated at
about 10.5 Ma (Evernden and others, 1977). Construc
tional dacitic to andesitic volcanic activity along the
Cerke-Huaricunca chain buried the older, more felsic
pyroclastic material with flows and extensive mudflow
breccias. These andesites and dacites were erupted at
5.7+0.5 Ma (Lavenu and others, 1989).
Subvolcanic intrusive bodies of rhyolite to dacite were
emplaced along a north-northwest-trending belt in the west
ern part of the study area. The stocks intrude the Beren
guela Formation and both the lower and upper members of
the Mauri Formation, but they are overlain by mudflow
breccias of Cerro Cerke. Therefore, intrusive ages probably
range from about 15 to 5 Ma, based upon existing geochro
nologic data and new stratigraphic relations.
The youngest unit in the area is the Perez ignimbrite,
a moderately welded, quartz-phenocrystic tuff that was
erupted at 2.23.3 Ma (Evernden and others, 1977;
Lavenu and others, 1989). The tuff sheet extends for per
haps 100 km south of Berenguela and west to near the
borders with Peru and Chile; it represents at least 250 km3
of erupted magma.
STRUCTURAL GEOLOGY
Major unconformities and changes in sediment grain
size and composition record regional tectonic events.
Major unconformities occur between the Berenguela and
lower member of the Mauri, between the lower and upper
members of the Mauri, and between the upper member of
the Mauri and the 5-Ma dacitic volcanic rocks of Cerro
Cerke. The contact between the Berenguela and lower
member of the Mauri is essentially parallel in the center of
the district, but south of Berenguela, it is an angular
unconformity of as much as 10; also, the thick, prominent
sandstone bed at the top of the Berenguela Formation was
locally eroded prior to Mauri deposition. This indicates
broad folding and erosion of the Berenguela prior to
Mauri time, and the distinct change in lithology between
the two formations indicates a profound change in source
terranes. Regional stratigraphic correlations and limited
geochronologic data indicate that this unconformity proba
bly developed in the late Oligocene, perhaps between
about 30 and 26 Ma. This time is regionally correlative
with the onset of Cordilleran uplift and formation of the
Bolivian orocline (Sempere and others, 1990; Isacks,
1988).
The second major disconformity occurs below the
upper member of the Mauri Formation and related pyro
clastic rocks. These sediments and tuffs were deposited on
various eroded levels of the lower member of the Mauri
and, in places, the Berenguela. The disconformity sepa
rates the dominantly continental clastic sedimentary rocks
of the lower member of the Mauri and Berenguela Forma
tions from the dominantly volcanic and volcaniclastic
rocks of the upper member of the Mauri Formation. The
age of the disconformity can only generally be placed
between 26 Ma (the age of basalt flows in the lower mem
ber of the Mauri) and 10 Ma (the age of a tuff in the
upper member of the Mauri). This major unconformity
likely correlates with a regionally extensive period of
Neogene deformation, perhaps broadly equivalent to the
Quechua period of deformation (McKee and Noble, 1989).
This episode, recognized in Peru, peaked at about 19 Ma.
The third unconformity is between the upper member
of the Mauri/Cerke sequence and the dacitic volcanic
rocks and mudflows derived from Cerros Cerke and Huar
icunca. The older rocks, including the 10-Ma tuff, are
broadly folded and faulted, and some erosion occurred
prior to deposition of the dacitic material at about 5 Ma.
This event reflects deformation throughout the region
this event is referred to as the late Miocene Quechua II
deformational event in Peru (McKee and Noble, 1982,
1989). The flat-lying 2-Ma Perez ignimbrite indicates that
the region has been tectonically quiet since that time.
Deformational structures within the Berenguela dis
trict include broad, open folds and high-angle faults.
Polyphase brittle structures in the Berenguela Formation
formed during four distinct deformation events. Open
folds with westerly trending fold axes (N. 90 W.N. 70
W.) that deform the Berenguela Formation but not the
overlying Mauri Formation formed during the oldest
event. This event records mild north-south-oriented con
traction and was responsible for the observed unconfor
mity between the Berenguela and Mauri Formations.
 GEOLOGY AND MINERAL DEPOSITS, BERENGUELA DISTRICT, NORTHWESTERN BOLIVIA
The second event is recorded by high-angle faults that
are best exposed north of the town of Berenguela. Here,
faults cut the sedimentary rocks and basalt intrusions of the
Berenguela and lower Mauri Formations and the previously
open folds. Two major trends of steeply dipping (60-90)
faults are evident. One set is dominated by conjugate faults
that strike generally east-west and dip to the north and
south; less common faults trend northwesterly, N. 2050
W., and commonly link longer strike segments of the more
westerly striking faults. A second fault set strikes northeast
erly, N. 2050 E., and cuts and is cut by the westerly
striking faults; this mutually crosscutting relationship indi
cates that the faults are contemporaneous. Slickensides on
exhumed fault planes indicate multiple stages of movement
on the faults, and most faults show evidence of normal-slip,
oblique-slip, and (or) locally strike-slip movements, with
both dextral and sinistral sense of shear. Fracture and
microfault geometries in the sandstones of the Berenguela
Formation are identical to the map-scale fault geometry,
and these structures are commonly filled with silica and
control the bleaching alteration of the red sandstones.
Moreover, the cadmium-rich veins are localized along these
faults and fractures, and they probably formed during this
second faulting event.
The conjugate, westerly striking faults of the older
fault set bound grabens in which the contact between the
lower member of the Mauri Formation and the underlying
Berenguela Formation has been downdropped with respect
to the bounding mesas. The evidence for oblique slip on
the faults indicates that the grabens formed during limited
horizontal shearing, and the geometry of the grabens and
the slickslides indicates that the regional tectonic regime
was dextral shear. In this interpretation, the westerly strik
ing faults are analogous to Riedel shears, which form
early in the development of a strike-slip fault system (Tch
alenko, 1970; Wilcox and others, 1973), and the north
westerly striking faults, veins, and fractures formed as
normal or extensional structures. Shortening at this time
would have been oriented approximately northwest-south
east, and extension would have been oriented northeast
southwest.
A third event is recorded by broad, open folds with
north-trending (N. 0 E.-N. 15 E.) axes that deform the
upper member of the Mauri Formation and subjacent
rocks. The folds record mild west-east-oriented contrac
tion. Near the town of Berenguela, the conjugate westerly
striking faults that formed during the previous event were
reactivated; the cadmium-rich veins were extensively brec
ciated during this deformation. Sinistral oblique movement
on the conjugate faults is suggested by the slickslides
along the exhumed fault planes. This sense of movement
is consistent with the inferred direction of contraction. The
superposition of folds and faults that developed during the
first three deformations formed the broad structural dome
in the Berenguela region.
83
A fourth episode of deformation is recorded by north
erly striking, high-angle faults that cut Mauri and Beren
guela Formation. Movement on faults is consistently
normal. Faults cutting the Pliocene volcanic rocks west of
Berenguela (Sirvas C., 1964) perhaps formed at this time.
MINERAL DEPOSITS
The Berenguela district includes two similar, but nev
ertheless distinct, types of silver-bearing mineral deposits:
cadmium-rich, quartz-poor, base-metal veins in the Beren
guela and lower member of the Mauri Formations, and
cadmium-poor, quartz-rich veins associated with Miocene
volcanic rocks. These were previously described by
Schneider-Scherbina (1962) and Sirvas C. (1964).
DEPOSITS HOSTED BY
SEDIMENTARY ROCKS
Silver- and base-metal-rich veins occur in steeply
dipping fractures in the upper part of the Berenguela For
mation, in the lower member of the Mauri sedimentary
rocks, and occasionally in the andesites in the lower mem
ber of the Mauri (fig. 2). Many of the veins were exposed
at the surface; concealed or buried veins were exploited in
Colonial times by piques (steep, narrow, stone-lined
declines that lead to drifts along the veins).
The veins formed along east-northeast- and north
west-trending faults in the sedimentary rocks, and major
ore intercepts occur at the intersections of these two fault
sets. Slickensides along fault surfaces cut the ore, indicat
ing that fault movement continued after mineralization.
The depth of oxidation is shallow (1020 m), and most of
the dumps from the shallow workings contain minor
amounts of sulfides that could not be processed during
Colonial times. Therefore, the veins likely continue to
greater depths than those reached by historical mining.
Topographic relations indicate that veins extend at least
200 m above and at least 100 m below the Berenguela
Mauri contact.
As exposed in surface workings, the veins are gener
ally less than 10 cm wide, although millimeter-wide, silica
filled fractures cut altered wallrocks on either side of some
veins. Rocks for a few centimeters to a meter on either side
of the vein are bleached; in some areas of greater fracture
density, the entire mass of host rock has been bleached.
Some beds, especially massive sandstones bounded by
mudstones, are preferentially bleached as much a 20 meters
from the vein. Petrographic observations indicate that the
major apparent chemical change is wholesale removal of
iron during alteration. In some zones, barite has completely
replaced the original quartz cement of the sandstones, and
secondary copper carbonates and sulfates indicate that
84 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
copper (and other elements?) was deposited in the altered
rocks. Evidence of silicification and carbonatization is
notably absent in the alteration halo.
Vein minerals, as identified in samples collected from
dumps, include galena, greenockite, barite, and sphalerite,
with minor amounts of chalcopyrite, chalcocite, and rare
marcasite. Paragenetic relations indicate that bleaching and
minor dissemination of sulfides into the wallrock was the
earliest paragenetic event. Barite was then deposited as a
cement in the wallrocks, and the rock was then cut locally
by thin chalcedonic veinlets. Bladed barite was the first
major mineral to be deposited in the veins, followed by
minor marcasite. Subsequently, sphalerite, galena, and
greenockite were deposited in the vein. Galena and
sphalerite are intergrown with delicately banded, yellow
ish, colloform minerals; in part, these might be secondary
anglesite derived from the galena, but textures indicate
that many of these bands are primary. The ores, both pri
mary and secondary, are texturally and paragenetically
similar throughout the district, and no systematic zoning
was noted.
Much of the silver-rich ore mined from the oxidized
parts of the veins was selectively hand sorted by Colonial
miners and is therefore largely unavailable for study. Sec
ondary iron oxides are primarily orangish goethite, with
considerably less hematite and even less jarosite. Second
ary copper carbonates and sulfates, including malachite,
azurite, and possibly antlerite, form ubiquitous coatings on
fractures. These secondary iron and copper minerals sug
gest that chalcocite was the primary copper sulfide in the
veins and that pyrite, chalcopyrite, and the total volume of
iron and copper sulfides were relatively minor (Anderson,
1982).
A number of prospects occur along a fault that is in
the upper part of the lower member of the Mauri and that
continues into the upper member of the Mauri, east of
Berenguela. These deposits are characterized by abundant
malachite, tenorite, milky quartz, and chalcedony, and tex
tures are very different from those in the Berenguela dis
trict. This striking difference may indicate that these
deposits are products of a mineralizing event unrelated to
and possibly younger than the typical Berenguela ores.
Conversely, the veins may represent a distal silica- and
copper-rich zonation of the Berenguela mineralizing
system.
DEPOSITS ASSOCIATED WITH
VOLCANIC ROCKS
The mineral deposits that are associated with volcanic
rocks occur in Miocene volcaniclastic and volcanic rocks
at the Dos Amigos and Cerro Tatitocollo prospects. In
contrast to the sedimentary-rock-hosted deposits, these
deposits contain abundant silicified rock and, as indicated
by preliminary geochemical data, very little cadmium (a
few tens of parts per million versus several percent; J.L.
Lizeca, unpub. data, 1991). Dos Amigos, a vein 2 km
southwest of Berenguela, is along a fault that trends N.
70 W. in the Berenguela Formation along which a pre
mineralization rhyolite dike was emplaced; slickensides in
ore indicate subsequent fault movement. All rocks are
altered to a quartz-sericite assemblage, and quartz is the
dominant gangue mineral. Galena, pyrite, and sphalerite
were observed on the dumps; Sirvas C. (1964) reported
tennantite and pearceite. Both fault and hydrothermal
breccias are common along the vein. Recent carbonate
bearing, low-temperature waters have formed a line of
travertine aprons along the fault zone.
At Cerro Tatitocollo, 1 km south of Berenguela,
rocks of the lowermost part of the lower member of the
Mauri are unconformably overlain by volcaniclastic rocks
of the upper member of the Mauri. The dominant host
rock is an epiclastic breccia to pebble conglomerate of the
upper member of the Mauri that is composed largely of
subangular rhyolite fragments in a coarse-sand matrix.
These rocks were intruded by flow-banded rhyolite plugs.
Mineralization was concentrated along north-trending frac
tures; hydrothermal fumarolic breccias were localized
along the fractures, and silicification increased towards the
breccias. Silver, copper, arsenic, zinc, lead, and antimony
have been reported from this zone, which has a maximum
dimension of no more than a few hundred meters; gold
occurs in a small area southwest of the silver and base
metal zone.
AGES AND GENESIS OF THE DEPOSITS
The distinct differences in cadmium and silica in the
two types of ore deposits suggest two ages of mineraliza
tion and different sources of fluids and metals. The rela
tively more radiogenic thorogenic Pb isotopic
compositions in the cadmium-rich veins (208Pb/204Pb =
38.7338.84; 4 deposits, 8 samples) relative to the volca
nic rock and associated veins (208Pb/204Pb = 38.36; 1
deposit, 1 sample) furthermore suggest that the two fluids
interacted, or were derived from, two reservoirs of slightly
different Pb isotopic compositions. Veins in the Beren
guela and lower member of the Mauri Formations formed
after deposition of at least the lower member of the Mauri
and emplacement of the basaltthese events have been
dated at 25 Ma. Sparse field relations suggest that veins in
the Berenguela are unconformably overlain by unmineral
ized volcaniclastic rocks of the upper member of the
Mauri Formation, thereby placing an upper age limit on
the mineralization. In contrast, the volcanic-associated,
cadmium-poor deposits are no older than middle Miocene,
the inferred age of the igneous host rocks. The copper
silica deposits, for which no cadmium data are available,
 GEOLOGY AND MINERAL DEPOSITS, BERENGUELA DISTRICT, NORTHWESTERN BOLIVIA
in the upper member of the Mauri east of Berenguela, are
likely less than 10 m.y. old, the same age as the tuff unit
in the upper member.
Mineralizing fluids and contained metals could have
been derived from more than one source. Unlike ores in
the Corocoro district to the east, which formed in host
rocks during diagenesis and were later remobilized into
fractures (Flint, 1990), the ores at Berenguela are epige
netic. As a result, these deposits may not have formed in a
manner like those at Corocoro, where the metals were
remobilized by basin fluids from copper-rich basalts or
copper deposits in the Cordillera Occidental (Flint, 1990).
However, cadmium-rich ores are typically derived from
basin fluids; fluids of demonstrable igneous derivation
rarely contain abundant cadmium. Therefore, the abun
dance of cadmium in the Berenguela-hosted ores strongly
suggests a basin fluid-origin for some or all of the
metalsthe difference between the deposits at Berenguela
and Corocoro might be the presence or absence of fracture
permeability at the time of initial fluid migration. If
indeed the deposits formed before deposition of upper
member of the Mauri volcaniclastic sediments, then min
eralization might have occurred during the early to middle
Miocene deformational event.
In contrast, the fluids involved in the formation of the
igneous-hosted deposits formed during or after late Mauri
sedimentation and related igneous activity. Metals may
have been derived from an igneous source during Miocene
volcanic activity; as with many volcanic-hosted epithermal
systems, much of the mineralizing fluid may have been
heated meteoric water. Remobilization of some of the older
mineral deposits cannot be discounted; however, the rela
tive absence of cadmium in the volcanic-hosted veins rela
tive to silver and zinc, which have very similar mobilities
(Rose and others, 1979), argues against significant remobi
lization of the metals in the older Zn-Cd-Ag-Pb veins.
ACKNOWLEDGMENTS
The field work for this study was part of a joint
project, funded by the Interamerican Development Bank,
between the USGS and GEOBOL to study precious-metal
deposits in the Altiplano. C.G. Cunningham of the USGS
served as overall project coordinator; Marcello Claure and
Franz Tavera of GEOBOL were invaluable for their coor
dinating efforts in Bolivia. We thank Compaa Minera
del Sur (COMSUR) and Rio Tinto Zinc for the use of
their facilities in Berenguela.
REFERENCES CITED
Anderson, J.A., 1982, Characteristics of leached capping and tech
niques of appraisal, in Titley, S.R., ed., Advances in Geology
85
of the Porphyry Copper Deposits, Southwestern North Ameri
ca: Tucson, University of Arizona Press, p. 275295.
Barazangi, M., and Isacks, B.L., 1976, Spatial distribution of earth
quakes and subduction of the Nazca plate beneath South Amer
ica: Geology, v.4, p. 686692.
Evernden, J.F., Kriz, S.J., and Cherroni M.C., 1977, Potassium
argon ages of some Bolivian rocks: Economic Geology, v.
72, p. 10421061.
Flint, S.S., 1990, Sediment-hosted stratabound copper deposits of
the central Andes, in Boyle, R.W., Brown, A.C., Jefferson,
C.W., Jowett, E.C., and Kirkham, R.V., eds., Sediment-Hosted
Stratiform Copper Deposits: St. John's Geological Association
of Canada, p. 371398.
Isacks, B.L., 1988, Uplift of the central Andean plateau and bend
ing of the Bolivian orocline: Journal of Geophysical Research,
v. 93, p. 32113231.
Jordan, T.E., and Alonso, R.N., 1987, Cenozoic stratigraphy and
basin tectonics of the Andes Mountains, 2028 south lati
tude: American Association of Petroleum Geologists Bulletin,
v. 71, p. 4964.
Jordan, T.E., Isacks, B.L., Allmendinger, R.W., Brewer, J.A., Ra
mos, V.A., and Ando, C.J., 1983, Andean tectonics related to
geometry of subducted Nazca plate: Geological Society of
America Bulletin, v. 94, p. 341361.
Lavenu, A., Bonhomme, M.G., Vatin-Perignon, N., and De Pacht
ere, P., 1989, Neogene magmatism in the Bolivian Andes be
tween 16 S. and 18 S.: Stratigraphy and K/Ar
geochronology: Journal of Latin American Geosciences, v. 2,
p. 3547.
McKee, E.H., and Noble, D.C., 1982, Miocene volcanism and
deformation in the western Cordillera and high plateaus of
south-central Peru: Geological Society of America Bulletin,
v. 93, p. 657662.
McKee, E.H., and Noble, D.C., 1989, Cenozoic tectonic events,
magmatic pulses, and base- and precious-metal mineralization
in the central Andes, in Ericksen, G.E., Canas Pinochet, M.T.,
and Reinemund, J.A., eds., Geology of the Andes and its Rela
tion to Hydrocarbon and Mineral Resources: Circum-Pacific
Council for Energy and Mineral Resources Earth Science Se
ries, v. 11, p. 189194.
Rose, A.W., Hawkes, H.E., and Webb, J.S., 1979, Geochemistry in
Mineral Exploration, [2nd ed.]: New York, Academic Press,
657 p.
Schneider-Scherbina, A., 1962, Los yacimientos polimetalicos de
Berenguela: Departmento Nacionl de Geologia, Informe No.
39 de la Misin Geologica Alemana en Bolivia, 54 p.
Sempere, T., Herail, G., Oller, J., and Bonhomme, M.G., 1990, Late
Oligoceneearly Miocene major tectonic crisis and related ba
sins in Bolivia: Geology, v. 18, p. 946949.
Sheffels, Barbara M., 1990, Lower bound on the amount of crustal
shortening in the central Andes: Geology, v. 18, p. 812815.
Sirvas C., J.F., 1964, Estudio geolgico de la regin Tambo Mauri
Berenguela, Provincia Pacajes del Departamento de La Paz,
Republica de Bolivia: unpub. thesis, Universidad Mayor de
San Andreas.
Swanson, K.E., Noble, D.C., McKee, E.H., Sempere, T., Martinez,
C., and Cirbian, M., 1987, Major revisions in the age of rock
units and tectonic events in the northern Altiplano basin of Bo
livia (abs.]: Geological Society of America Abstracts Pro
grams. v. 19, no. 6, p. 456.
86 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Tchalenko, J.S., 1970, Similarities between shear zones of differ
ent magnitudes: Geological Society of America Bulletin, v.
81, p. 1625640.
Thorpe, R.S., 1984, The tectonic setting of active Andean magma
tism, in Harmon, R.S., and Barreiro, B.A., eds., Andean Mag
matism: Chemical and Isotopic Constraints: Nantwhich,
United Kingdom, Shiva Publishing Limited, p. 48.
Tosdal, R.M., Clark, A.H., and Farrar, E., 1984, Cenozoic
polyphase landscape and tectonic evolution of the Cordillera
Occidental, southernmost Peru: Geological Society of America
Bulletin, v. 95, p. 13181332.
Wilcox, R.E., Harding, T.P., and Seely, D.R., 1973, Basic wrench
tectonics: American Association of Petroleum Geologists Bul
letin, v. 57, p. 7496.
 CHAPTER I

K-Ar AGES OF BOLIVIAN TERTIARY

POLYMETALLIC VEIN DEPOSITS

By STEVE LUDINGTON, EDWIN H. MCKEE, and NORASHEW3

INTRODUCTION

Silver and tin mining has been a mainstay of the

Bolivian economy for over four centuries. In the Andes,
south of the capital city of La Paz, dozens of large mines
and hundreds of small mines and prospects occur in the
Cordilleras Oriental and Occidental and on the Altiplano
(fig. 1). Production has been mainly from the Cordillera
Oriental.
Many of the mines in the Bolivian Andes are poly
metallic vein deposits that are closely related to subvolca
nic dacitic intrusions, and, as far as is known, most of
these Tertiary intrusions and their associated polymetallic
vein deposits are Miocene in age. The intrusions are
emplaced in two distinct host-rock types: (1) Paleozoic
marine sedimentary and metasedimentary rocks, primarily
in the Cordillera Oriental, and (2) Tertiary continental sed
imentary and volcanic rocks in the Cordillera Occidental
and on the Altiplano. I

The polymetallic veins consist primarily of sulfide

minerals, chiefly pyrite, marcasite, and pyrrhotite. Lesser Figure 1. Index map showing location of physiographic
amounts of sphalerite, galena, cassiterite, arsenopyrite, provinces of Bolivia referred to in text.
chalcopyrite, stibnite, tetrahedrite, wolframite, bismuthin
ite, argentite, native bismuth, native gold, and complex
sulfosalt minerals occur in various combinations. There is This report summarizes the results of geochronologic
a notable scarcity of nonsulfide gangue, although quartz, studies of samples collected during 1990 and 1991 during a
barite, or manganese carbonate occur locally in the veins, general study of the mineral deposits of the Altiplano of
primarily in the upper parts of the hydrothermal systems. Bolivia.
Sericite, or white mica, is the most common alteration
mineral, along with kaolinite. Alunite also occurs in a few
deposits.
K-Ar AGE DETERMINATIONS

'U.S. Geological Survey, Mail Stop984, 345 Middlefield Road, Men
lo Park, CA 94025.
*U.S. Geological Survey, Mail Stop 901, 345 Middlefield Road, Men
lo Park, CA 94025.
*U.S. Geological Survey, Branch of Alaskan Geology, 4200 Univer
sity Drive, Anchorage, AK 99508.
Twelve new conventional K-Ar age determinations,
from ten mineral deposits in the Cordillera Occidental and
on the Altiplano of Bolivia, were performed during this
study. The ages and analytical data are presented in table
1. Five of the ages directly date mineralization and (or)
88 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 1. K-Arages and analytical data for samples from the Altiplano and Cordillera Occidental, Bolivia.
[rad, radiogenic. Constants used were *K/K=1.167x10" mol/mol; l-4.962x10"/yr. 1.-1.-0.581z10"/yr. Samples 90BSLO14, 109, 110,90BDR010, 90BBR004a, 10, 14, and
90BBG024 were analyzed by E.H. McKee, Samples 90BDR030, 35, 53, and 55 were analyzed in duplicate for argon by Nora Shew, yielding a single age. Potassium measurements
by S.T. Pribble. Mass spectrometry by J.Y. Saburomaru, J.C. Von Essen]

Sample Locality Type of Mineral K2O *Arrad *Arrad Age

ino. rock analyzed (weight (mol/g (percent of (Ma)
percent) x10-11) total)
Direct dates on mineral deposits
90BDR()] () San Cristobal (Toldos) Dacite in pit Biotite 747 9. 136 54 8.5+0.3
90BBG024 Carangas Rhyolite flow Biotite 8, 12 18.03 59 15.4+0.5
9()BSL()14 Buena Vista Vein wallrock Sericite 4.57 7.306 55 | 1, 14-0.3
90BSL1().9 Kiska Late vein Alunite 8, 19 16.08 36 13.6+0.5
9()BSLl 1() Kori Kollo Late vein Alunite 9.59 6,555 50 4.7+().2
Dates on intrusions, lavas, and ash-flow tuffs that constrain mineral deposit ages
90BBR004a Todos Santos Dacite Biotite 5.42 4.739 17 6, 14-0.2
9()BBR01() Negrillos Ash-flow tuff Biotite 8.46 26.6 | 75 21.7+().7
90BBR014 Mara Elena Andesite dike Hornblende 783 709] 17 6.3+().2
9()BDR()3() San Cristobal Dacite stock Biotite 8.9() | ().32 49 8.0+0.1
8.9 | 1(). 34 56
90BDR035 Eskapa (Cuidado) Dacite Biotite 8.46 7,696 37 6.3+0.1
8.5 | 7 727 44
9()BDR()53 Mina Escala Dacite debris low Biotite 8.08 21.43 60 18.2+0.3
8.09 21.07 S3
9()BDR()55 Mina Escala Rhyolite stock Biotite 9.02 23.43 6() 18.0+().2
9 (). 23.62 73

alteration events. The other seven ages are on intrusions,
tuffs, or lava flows at five additional localities and indi
rectly aid in constraining the time of mineralization. The
mineral deposits dated during this study were selected
because they are broadly representative of Bolivian min
eral deposits in the central Andes and little was known of
their geochronology. Samples from the deposits dated
directly yielded alunite, biotite, or sericite for dating;
biotite and hornblende from the intrusions, ash-flow tuffs,
and lava flows were used for age determination (table 1).
K-Ar analyses were performed using standard iso
tope-dilution techniques similar to those described by Dal
rymple and Lanphere (1969). The mineral concentrates
were made by heavy-liquid, magnetic, electrostatic, and
hand-picking procedures. Potassium analyses were by lith
ium-metaborate-flux-fusion flame photometry, the lithium
serving as an internal standard (Ingamells, 1970). Argon
analyses were done using a five-collector mass spectrome
ter (Stacy and others, 1981). The precision of the data,
shown as the value in table 1, is the estimated analytical
uncertainty at one standard deviation (Cox and Dalrymple,
1967). It represents uncertainties in the measurement of
radiogenic 40Ar and K2O based on experience with hun
dreds of replicated analyses in the Menlo Park laboratories
of the U.S. Geological Survey. Mass discrimination of the
spectrometer is routinely determined on the basis of multi
ple analyses of purified air. The constants used in age
determination are those from the Subcommission on Geo
chronology (Steiger and Jaeger, 1977).
The new ages, along with brief descriptions of the
mineral deposits, are given below. More detailed deposit
descriptions can be found in a report on the mineral
resources of the Bolivian Cordillera Occidental and Alti
plano (U.S. Geological Survey and Servicio Geolgico
de Bolivia, 1991). Locations of the studied deposits
together with all known major polymetallic vein deposits
are shown on figure 2, which includes the new K-Ar
ages reported here and all known previously reported
ages (Ludington, Orris, and others, 1991) for polymetal
lic vein deposits.
KORI KOLLO DEPOSIT
The La Joya district includes several polymetallic
vein deposits. Principal among them is the one at Cerro
Kori Kollo, which contains about 150 tonnes of gold and
1,000 tonnes of silver and is the first Bolivian polymetal
lic vein deposit to produce substantial amounts of gold
(Long and others, 1991). Cerro Kori Kollo is one of four
hills in the district that are apparently cupolas on a large
dacitic intrusive body that intrudes siltstone and sandstone
of Silurian age. The ore consists of sulfide and oxide veins
that cut a sericitized dacite porphyry.
The age of the La Joya dacite intrusion, based on a
biotite K-Ar date, is 14.3+0.4 Ma, whereas an age deter
mined on sericite from the orebody at Kori Kollo was
15.7+0.5 Ma (Redwood, 1987). These dates, although
 K-Ar AGES OF BOLIVIAN TERTIARY POLYMETALLIC VEIN DEPOSITS

69 68 67 66
FIL i I I
Z y
* uticaca
S 13 Ma LA PAZ
,--" OUIMSA CHATA 24 Ma
SJ S VILOCO
A
17 H.
3%;
( &
" . canacoles
23 Ma
( } -

J
2-
BERENGUELA CHICOTE
O
S O 23 Ma 79
| - COLOUIRI o
& y 14 Ma Y."
3. N\ * -
La JA
s O

I *::: s:Orur c COLCHA

is H. (") s W A le f ":"
- S * *** MoRococal &vows
O P v : 79
73 W P e MON-e 21 Ma
* SERRAT
AVICAYA .
21 Ma COLOUECHACA
i \ v
21 Ma
* \ ..neanutos <21.7 Ma y Lago
r & ..canawaas su- 2 *
Poop
e MALMISA
* st - THEREsA

Ye -
HUARI
.colav
20 Ma.
caRGUAlcoLLo \ ^* 9ERRo Rico
Z. DE POTosi
13 Ma

PORCO
O
2,
~~
--

SANTA Rosa
DE KASIRI -

EL ASIENTO y -

PULACAYO
*
*.

-> UBINA
- O
Uyuni TASNA
16 Ma

CHOROLOUE
CHOCAYA o 16 Ma
13 Ma -
0.5 Ma
O TATAs.
O O ISCA ISCA
SAN cRESTOBAL.
e SAN
16 MaviceNTE O CHOROMA

Escala
comuniDAD e < 18 Ma. SANTA ISABEL
- - - - - - - - - - e "-.
sAN e sueNA w87A 11 Me
. ANTONIO e!ESMORACA
. *NYPole B
; ...?& " " " `MoRoKF ---t-via
~" AR
GFAr
N- -

<15 / Ma. roveous ENIN.

23 20 o 20 40 Go so 100 kilometers
|- Ececem-e

I I I 1 I

Figure 2. Map showing locations and ages of mineral deposits discussed in this study and additional polymetallic
vein deposits in western Bolivia. Solid circles are polymetallic vein deposits; X's are other types of deposits. Names
and dates in bold italics indicate deposits specifically studied. Shaded area indicates the young volcanic rocks of the
Cordillera Occidental, which are post-mineral cover to the polymetallic vein deposits.
90 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
discordant in detail, are consistent with field evidence that
indicates that intrusion of the dacite bodies and mineral
ization are closely related. We dated a quartz-alunite vein
that crosscuts mineralized structures in the Kori Kollo
mine and obtained a K-Ar age of 4.7+0.2 Ma (table 1). If
all the dates are accurate, they indicate that the hydrother
mal system was extremely long lived. Additional studies
are needed to eliminate the possibility that the alunite age
documents either a later hydrothermal event that is unre
lated to the Miocene intrusion or supergene alteration.
NEGRILLOS AREA
In the Negrillos area, silver-bearing polymetallic
veins are emplaced in a series of andesitic lava flows and
volcanic breccias (Gamble and others, 1991). These rock
units appear to be stratigraphically above an ash-flow tuff
dated at 21.7+0.7 Ma (table 1), suggesting that the veins at
Negrillos are younger than 21.7 Ma.
CARANGAS AREA
In the Carangas area, silver-bearing polymetallic
veins are emplaced in altered tuffs of the Tertiary Caran
gas Formation (Gamble and others, 1991). The most sig
nificant deposits appear to be on Cerro Espritu Santo,
where zones of intense hydrothermal alteration are
present. A distinctive rhyolite porphyry is found both as
dikes and small plutons that intrude the hydrothermally
altered rocks on the hill and as clasts within hydrother
mal(?) mineralized breccias. Analysis of biotite from a
dike in this rhyolite porphyry gave an age of 15.4+0.5 Ma
(table 1). Although the field relations that constrain the
meaning of this date are equivocal, the age of mineraliza
tion may be about 15 Ma.
TODOS SANTOS DEPOSIT
The deposit at the Todo Santos mine is related to a
rhyolite dome complex that intrudes rock of the Tertiary
Carangas Formation (Gamble and others, 1991). The
main Todos Santos dome consists of flow-banded rhyo
lite that intrudes its own carapace of tuffs and breccias.
The associated mineral deposit consists of a group of
veins and disseminations in and near a fault that cuts the
dome and its carapace of pyroclastic rocks. The prospect
is of interest primarily for its precious-metal values, but
base-metal values are also appreciable. Biotite from a
coarse block-and-ash flow of pumiceous rhyodacite
yielded a K-Ar age of 6.1+0.2 Ma (table 1); the block
and-ash flow is intruded by a small rhyolite dome, simi
lar in appearance to the main dome at Todos Santos. If
the two domes are temporally equivalent, the mineraliza
tion at Todos Santos may have formed less than 6.1 m.y.
ago.
SAN CRISTOBAL DISTRICT (TOLDOS MINE)
Mineral deposits in the San Cristobal district occur
near the center of a deeply eroded late Tertiary volcanic
complex that consists of a number of subvolcanic stocks
that intrude siltstone, sandstone, and conglomerate of
probable Eocene and Oligocene age and younger volcanic
and volcaniclastic rocks of probable Miocene age (Richter,
Brooks, Ludington, and others, 1991). Currently, the most
important mineral deposit in the district is being exploited
at the Toldos mine, where an open-pit, heap-leach opera
tion recovers silver from an orebody of approximately
3,000,000 tonnes at about 120 grams/tonne (g/t) silver.
The district is typical of many Altiplano polymetallic
vein districts in that tin and gold are found in very low
concentrations. In the Toldos deposit, alteration envelopes
around silver-bearing pyrite veins are primarily composed
of sericite. The dacite porphyry that hosts the deposit and
others in the district have been affected by a form of potas
sic alteration that has resulted in whole-rock K2O contents
greater than 8 weight percent, even though petrographic
studies have been unable to document and clarify the nature
of the alteration. Fresh biotite phenocrysts are a distinctive
characteristic of these altered rocks, and K-Arages on these
biotites yielded ages of 8.0+0.1 Ma (90BDR030) and
8.5+0.3 Ma (90BDR010) (table 1). Sample 90BDR010 is
from the open pit at the Toldos mine, whereas sample
90BDR030 is from an outcrop of a separate intrusion, sev
eral kilometers to the north. Thus, 8.5 Ma would appear to
be the maximum age of mineralization.
ESCALA DEPOSIT
The Escala deposit consists of a series of silver-bear
ing polymetallic veins emplaced in a subvolcanic rhyolite
porphyry stock (Richter, Brooks, Cox, and others, 1991).
The rhyolite stock intrudes a series of dacitic ash-flow
tuffs and debris flows. Near the deposit, a date of 18.2+0.3
Ma was obtained from biotite in a dacitic debris flow in
the upper part of this volcanic series (table 1). Biotite
from fresh intrusive rhyolite porphyry a few kilometers
north of the Escala deposit yielded an age of 18.0+0.2 Ma.
These dates constrain the age of the Escala deposit to be
less than 18 Ma (table 1).
BUENA VISTA DEPOSIT
In the Buena Vista area, several polymetallic veins are
emplaced in Miocene dacitic ash-flow tuffs, lava flows, and
volcanic breccias (Ludington, Czamanske, and others,
1991). Unlike most of the Bolivian polymetallic vein
 K-Ar AGES OF BOLIVIAN TERTIARY POLYMETALLIC VEIN DEPOSITS
deposits, they have no obvious association with dacitic
intrusive bodies. The veins contain lead, zinc, silver, anti
mony, and gold; the principal ore minerals are stibnite,
galena, sphalerite, chalcopyrite, argentite, and various sil
ver sulfosalts. Alteration sericite that is part of a quartz
sericite-pyrite envelope around a vein yields a K-Ar age of
11.1+0.3 Ma (table 1). This confirms an earlier determina
tion of 11.3+0.6 Ma (Instituto de Geologa Econmica,
1985). This age is considered to be the age of mineraliza
tion at Buena Vista; it is considerably younger than the
enclosing pyroclastic rocks, which appear to have been
erupted between about 24 Ma and 15 Ma (Kussmaul and
others, 1975).
MARA ELENA PROSPECT
Anomalous values of gold and silver are reported from
a small, rhyolitic intrusive body near the top of Cerro Cul
ebra, a composite stratovolcano about 6.5 km south of
Todos Santos (Ratt and others, 1991). This intrusive is
extensively silicified as are lava flows surrounding and
overlying it. Base metals are generally absent from this
mineralized system, which is probably an epithermal vein
deposit rather than a polymetallic vein. The altered rocks
at Cerro Culebra are cut by an unaltered hornblende-bear
ing andesite dike that yields a K-Ar age of 6.3+0.2 Ma
(table 1), which provides a minimum age for the deposit.
CERRO ESKAPA AREA (CUIDADO MINE)
Cerro Eskapa is part of a large field of late Tertiary
and Quaternary stratovolcanoes in the Cordillera Occiden
tal and is about 125 km southwest of Uyuni. Two mineral
deposits are hosted in the lavas of this volcano. One, the
Cuidado mine, is near the summit of the volcano and con
sists of an area of conspicuous epithermal, chiefly argillic,
hydrothermally altered dacitic breccias that contains
anomalous values of gold, silver, arsenic, antimony, and
tellurium. The other is on the flanks of the volcano and
consists of copper veins in a massive dacitic lava dome
flow that contains phenocrysts of sanidine, plagioclase,
biotite, and some quartz (Richter, Brooks, Shew, and oth
ers, 1991). Biotite from this rock yielded a K-Ar age of
6.3+0.1 Ma (table 1)this serves as a maximum age of
mineralization for the copper deposit.
KISKA PROSPECT
The Kiska gold-silver prospect is on an isolated hill
southwest of the Kori Kollo deposit in the La Joya district
(Ludington, du Bray, and others, 1991). The entire hill,
which is composed primarily of Paleozoic sedimentary
91
Table 2. Ages of mineral deposits on the Altiplano and in the
Cordillera Occidental, Bolivia.
[Deposits followed by an asterisk (*) are not polymetallic vein deposits; Mara Elena
and the Cuidado mine on Cerro Eskapa are probably epithermal precious-metal depos
its; Kiska is not assignable at present]
Mineral deposit Age
Kori Kollo >5 Ma
Negrillos <21.7 Ma
Carangas 15 Ma
Todos Santos <6 Ma
San Cristobal (Toldos mine) 8.5 Ma
Mina Escala <18 Ma
Buena Vista 11 Ma
Mara Elena >6 Ma
Eskapa (Cuidado mine)* <6 Ma
Kiska" > 14 Ma
rocks, has been hydrothermally altered, primarily by
strong silicification, and intruded by a dacite porphyry.
The mineralized rock is a stockwork that consists of at
least two generations of veins that contain fine silica,
jarosite, hematite, and alunite . Alunite from the youngest
and apparently unmineralized set of veins yielded an age
of 13.6+0.5 Ma (table 1), providing a minimum age for
the deposit.
DISCUSSION
The available radiometric ages of minerals associated
with Tertiary polymetallic vein deposits in the Bolivian
Andes south of La Paz are latest Oligocene or Miocene.
This includes 14 ages taken from published sources (Lud
ington, Orris, and others, 1991) and the ages determined
or constrained during this study (table 2, fig. 2). The ages
range from about 24 Ma to about 5 Ma and span the entire
Miocene with no apparent temporal maxima in mineraliza
tion. Three other deposits (Mara Elena, Eskapa, Kiska),
which are not polymetallic vein deposits, also have com
parable ages.
There do seem, however, to be spatial patterns in time
of-mineralization in this region. Deposits in the Cordillera
Oriental are older than in the Altiplano and Cordillera
Occidental, and deposits become progressively younger
southward in both Cordilleras and in the Altiplano.
In the Cordillera Oriental, over a distance of about
500 km, the deposits decrease in age from about 24 Ma, at
Viloco in the north, to about 16 Ma, at San Vicente in the
south (Cerro Rico de Potos appears to be an exception at
13 Ma).
A similar but less well-defined pattern of westward
younging of polymetallic vein deposits is seen in the
92 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Cordillera Occidental and on the Altiplano of Bolivia. The
ages of deposits in this belt are significantly younger than
in the Cordillera Oriental to the east: from about 15 Ma at
Kori Kollo to less than 6 Ma at Todos Santos.
RELATIONSHIP OF MINERAL
DEPOSITS TO ANDEAN VOLCANISM
The Andean Cordillera of western South America is
commonly referred to as the classic example of a conver
gent continental plate margin (Dewey and Bird, 1970).
The area of this report, southwestern Bolivia, lies in the
central Andes, astride a major oroclinal bend in the Cor
dillera. Differences in the tectonic and igneous history of
different parts of the Andes are the result of different rates
and (or) angles of subduction between the Pacific Nazca
plate and the continental South American plate, as well as
to varying crustal thickness. Offshore from Bolivia and
northern Chile, a segment of the subducting Nazca plate is
bounded on the north by the Easter fault zone (subparallel
to the Nazca Ridge) and on the south by the Challenger
fault zone (Schweller and others, 1981). This Nazca plate
segment is directly related to tectonism and volcanism in
the Bolivian portion of the Andes.
In Bolivia, the Cordillera consists of three contiguous
morphotectonic provinces, which are, from east to west,
the Cordillera Oriental, Altiplano, and the Cordillera Occi
dental (figs. 1 and 2). Built across these provinces is the
central volcanic zone (Thorpe and others, 1982), the
largest of the three active volcanic chains in the Andean
Cordillera. Here, volcanism has also been voluminous dur
ing most of the latter half of the Cenozoic, especially dur
ing the last 27 million years. The sites of former
volcanoes define two arcs that merge at the southern end
of the Altiplano. The eastern arc lies primarily in the Cor
dillera Oriental and consists of eroded calc-alkaline volca
noes that are late Oligocene and early to middle Miocene
in age (27 to 11 Ma). The western arc (primarily in the
Cordillera Occidental) forms the high crest of the Andes
and is comprised of late Miocene to Holocene andesitic
stratovolcanoes and voluminous ignimbrite sheets (Francis
and others, 1983).
Calc-alkaline volcanic and intrusive rocks in the
eastern arc (Cordillera Oriental) are the host for most of
the Cenozoic polymetallic vein deposits in Bolivia and
are closely associated with the hydrothermal activity that
formed the veins. The volcanic and intrusive rocks range
in age from about 27 to 11 Ma. Mineralization dates
coincide with this time interval, with an apparent concen
tration between about 25 Ma and 15 Ma. In most miner
alized areas, the age of mineralization is only slightly
younger than intrusive and extrusive ages, confirming the
observation that hydrothermal activity is fundamentally a
late-stage part of the igneous system (McKee and Noble,
1990).
Bolivian polymetallic vein deposits appear to be
smaller and less common on the Altiplano and in the Cor
dillera Occidental than in the Cordillera Oriental. This
may be more apparent than real and could be the result of
the relative lack of erosion of the younger volcanoes in the
west or burial of older, mineralized hydrothermal systems
by young, unaltered lava flows and ash-flow tuffs.
Because it is much less deeply eroded, the western arc
(Cordillera Occidental) appears to be more likely to host
epithermal precious-metal deposits than polymetallic vein
deposits. Because of the generally younger age of the
igneous rocks, mineralization there is also younger.
REFERENCES CITED
Cox, Allan, and Dalrymple, G.B., 1967, Statistical analysis of
geomagnetic reversal data and the precision of potassium-ar
gon dating: Journal of Geophysical Research, v. 72, no. 10,
p. 26032614.
Dalrymple, G.B., and Lanphere, M.A., 1969, Potassium-argon dat
ingPrinciples, techniques, and applications to geochronolo
gy: San Francisco, W.H. Freeman, 258 p.
Dewey, J.F., and Bird, J.M., 1970, Mountain belts and new
global tectonics: Journal of Gcophsyical Research, v. 75,
p.26252647.
Francis, P.W., Halls, C., and Baker, M.C.W., 1983, Relationships
between mineralization and silicic volcanism in the central
Andes: Journal of Volcanology and Geothermal Research, v.
18, p. 165190.
Gamble, B.M., Ratt, J.C., Carrasco, Raul, Soria-Escalante, Edu
ardo, and McKee, E.H., 1991, Todos Santos district, in U.S.
Geological Survey and Servicio Geolgico de Bolivia, Geolo
gy and mineral resources of the Altiplano and Cordillera Occi
dental, Bolivia, with a section on Application of economic
evaluations to mineral deposit models, by D.I. Bleiwas and
R.G. Christiansen, U.S. Bureau of Mines: U.S. Geological Sur
vey Bulletin 1975, p. 136143.
Ingamells, C.O., 1970, Lithium metaborate flux in silicate analysis:
Analytica Chimica Acta, v. 52, p. 323334.
Instituto de Geologa Econmica, 1985, K-Arages of mineraliza
tion at the Caracoles, Siglo XX, Colquechaca, Colavi, Huari
Huari, Unificada, Tasna, Inocentes, and Buena Vista mines
in Bolivia: La Paz, Universidad Mayor de San Andres de
Bolivia, Instituto de Geologa Econmica, Publicacin Espe
cial No. 2, p. 270281.
Kussmaul, S., Jordan, L., and Ploskonka, E., 1975. Isotopic ages of
Tertiary volcanic rocks of SW. Bolivia: Bundesanstalt fr
Bodenforschung und geologischen Landesmter, Geologische
Jahrbuch, part B. v. 14, p. 111120.
Long, K.R., Ludington, Steve, du Bray, E.A., Andr-Ramos, Or
lando, and McKee, E.H., 1991, La Joya district, in U.S.
Geological Survey and Servicio Geolgico de Bolivia, Geol
ogy and mineral resources of the Altiplano and Cordillera
Occidental, Bolivia, with a section on Application of eco
nomic evaluations to mineral deposit models, by D.I.
 K-Ar AGES OF BOLIVIAN TERTIARY POLYMETALLIC VEIN DEPOSITS
Bleiwas and R.G. Christiansen, U.S. Bureau of Mines: U.S.
Geological Survey Bulletin 1975, p. 131136
Ludington, Steve, Czamanske, G.K., Carrasco, Raul, Hinojosa-Ve
lasco, Alberto, and McKee, E.H., 1991, Buena Vista area, in
U.S. Geological Survey and Servicio Geolgico de Bolivia,
Geology and mineral resources of the Altiplano and Cordillera
Occidental, Bolivia, with a section on Application of economic
evaluations to mineral deposit models, by D.I. Bleiwas and
R.G. Christiansen, U.S. Bureau of Mines: U.S. Geological Sur
vey Bulletin 1975, p. 164174.
Ludington, Steve, du Bray, E.A., and McKee, E.H., 1991, Kiska
prospect, in U.S. Geological Survey and Servicio Geolgico de
Bolivia, Geology and mineral resources of the Altiplano and
Cordillera Occidental, Bolivia, with a section on Application
of economic evaluations to mineral deposit models, by D.I.
Bleiwas and R.G. Christiansen, U.S. Bureau of Mines: U.S.
Geological Survey Bulletin 1975, p. 191192.
Ludington, Steve, Orris, G.J., Cox, D.P., and Asher-Bolinder, Si
grid, 1991, Mineral deposit models, in U.S. Geological Survey
and Servicio Geolgico de Bolivia, Geology and mineral re
sources of the Altiplano and Cordillera Occidental, Bolivia,
with a section on Application of economic evaluations to min
eral deposit models, by D.I. Bleiwas and R.G. Christiansen,
U.S. Bureau of Mines: U.S. Geological Survey Bulletin 1975,
p. 6490.
McKee, E.H., and Noble, D.C., 1990, Cenozoic tectonic events,
magmatic pulses, and base- and precious-metal mineralization
in the central Andes, in Ericksen, G.E., Pinochet, Maria Teresa
Caaas, and Reinemund, J.A., eds., Geology of the Andes and
its relation to hydrocarbon and mineral resources: Circum-Pa
cific Council for Energy and Mineral Resources Earth Science
Series, v. 11, p. 189194.
Ratt, J.C., Gamble, B.M., McKee, E.H., Soria-Escalante, Eduardo,
and Carrasco, Raul, 1991, Mara Elena prospect (Cerro Cule
bra), in U.S. Geological Survey and Servicio Geolgico de Bo
livia, Geology and mineral resources of the Altiplano and
Cordillera Occidental, Bolivia, with a section on Application
of economic evaluations to mineral deposit models, by D.I.
Bleiwas and R.G. Christiansen, U.S. Bureau of Mines: U.S.
Geological Survey Bulletin 1975, p. 110112.
Redwood, S.D., 1987, The Soledad Caldera, BoliviaA Mi
ocene caldera with associated epithermal Au-Ag-Cu-Pb-Zn
mineralization: Geological Society of America Bulletin, v.
99, p. 395404.
Richter, D.H., Brooks, W.E., Cox, D.P., Shew, Nora, Bailey, Eliz
abeth, Hinojosa-Velasco, Alberto, and Escobar-Diaz, Angel,
93
1991, Escala district, in U.S. Geological Survey and Servicio
Geolgico de Bolivia, Geology and mineral resources of the
Altiplano and Cordillera Occidental, Bolivia, with a section on
Application of economic evaluations to mineral deposit mod
els, by D.I. Bleiwas and R.G. Christiansen, U.S. Bureau of
Mines: U.S. Geological Survey Bulletin 1975, p. 159161.
Richter, D.H., Brooks, W.E., Ludington, Steve, Hinojosa-Velasco,
Alberto, Escobar-Diaz, Angel, McKee, E.H., and Shew, Nora,
1991, San Cristobal district, in U.S. Geological Survey and
Servicio Geolgico de Bolivia, Geology and mineral resources
of the Altiplano and Cordillera Occidental, Bolivia, with a sec
tion on Application of economic evaluations to mineral deposit
models, by D.I. Bleiwas and R.G. Christiansen, U.S. Bureau of
Mines: U.S. Geological Survey Bulletin 1975, p. 153156.
Richter, D.H., Brooks, W.E., Shew, Nora, Hinojosa-Velasco, Al
berto, and Escobar-Diaz, Angel, 1991, Cerro Eskapa, in U.S.
Geological Survey and Servicio Geolgico de Bolivia, Geolo
gy and mineral resources of the Altiplano and Cordillera Occi
dental, Bolivia, with a section on Application of economic
evaluations to mineral deposit models, by D.I. Bleiwas and
R.G. Christiansen, U.S. Bureau of Mines: U.S. Geological Sur
vey Bulletin 1975, p. 116118.
Schweller, W.J., Kulm, L.D., and Prince, R.A., 1981, Tectonics,
structure, and sedimentary framework of the Peru-Chile
trench, in Kulm, L.D., Dymend, E.J., Dasch, E.J., and Hus
song, D.M., eds., Nazca plateCrustal formation and An
dean convergence: Geological Society of America Memoir
154, p. 323349.
Stacey, J.S., Sherrill, N.D., Dalrymple, G.B., Lanphere, M.A., and
Carpenter, N.V., 1981, A five-collector system for the simulta
neous measurement of argon isotopic ratios in a static mass
spectrometer: International Journal of Mass Spectrometry and
Ion Physics, v. 39, p. 167180.
Steiger, R.H. and Jaeger, E., 1977, Subcommission on Geochronol
ogyConvention on the use of decay constants in geo- and
cosmochronology: Earth and Planetary Science Letters, v. 36,
p. 359362.
Thorpe, R.S., Francis, P.W., Hammill, M., and Baker, M.C.W.,
1982, The Andes, in Thorpe, R.S., ed., Andesites: New York,
John Wiley and Sons, p. 187205.
U.S. Geological Survey and Servicio Geolgico de Bolivia, 1991,
Geology and mineral resources of the Altiplano and Cordillera
Occidental, Bolivia, with a section on Application of economic
evaluations to mineral deposit models, by D.I. Bleiwas and
R.G. Christiansen, U.S. Bureau of Mines: U.S. Geological
Survey Bulletin 1975, 365 p.
 CHAPTER J
GOLD AND SILVER IN ACID-SULFATE
ALTERATION AND QUARTZ-SERICITE-PYRITE
STOCKWORKS, LA ESPANOLA PROSPECT,
NORTHWESTERN ALTIPLANO, BOLIVIA
By ALBERT HOFSTRA, RICHARD F HARDYMAN, LUIS BARRERA,
and ORLANDO SANJINES
ABSTRACT
La Espaola prospect is located in the Bolivian por
tion of the Neogene to Quaternary magmatic arc of the
central Andes. La Espaola was selected for study
because it is the most intense color anomaly recognized on
remote sensing imagery in the region. Field mapping and
preliminary laboratory studies were conducted by geolo
gists from GEOBOL (Servicio Geolgico de Bolivia) and
the USGS (U.S. Geological Survey) as part of a training
program in volcanic-hosted precious metal deposits that
was funded by the Inter-American Development Bank.
The hydrothermal system is centered on a 1.5 by 3
km composite dacite porphyry stock that intrudes upper
Oligocene to middle Miocene andesitic to dacitic flows,
lahars, and volcaniclastic sedimentary rocks. The stock
consists of at least four intrusive phases that are sericiti
cally altered, including one distinguished by an abundance
of miarolitic cavities. Several small unaltered dacite por
phyry dikes and plugs intrude altered dacite porphyry. K
Ar determinations on hornblende in pre-mineralization
dacite flows (11.2+0.7 Ma) and on biotite from a pre-min
eralization intrusive phase in the composite dacite por
phyry stock (11.8+0.8 Ma) have similar ages. A K-Ar
determination on alunite yielded an age of 10.3+0.3 Ma.
'U.S. Geological Survey, Mail Stop 973, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, University of Nevada, Makay Building,
RMSM 101A, Reno, NV 89557.
*Servicio Geolgico de Bolivia, Calle Frederico Zuazo No. 1673,
Casilla 2729, La Paz, Bolivia.
The hydrothermal system is therefore about 1 Ma younger
than the host composite dacite porphyry stock.
La Espaola prospect contains three types of mineral
occurrences: (1) epithermal acid-sulfate-type mineraliza
tion characterized by an early stage of vuggy silica and
quartz-alunite alteration and a second stage of silicification
with pyrite-barite + enargite + galena it sphalerite + gold +
silver; (2) gold- and silver-bearing quartz-sericite-pyrite
stockwork veinlets in the sericitic zone of a dacite por
phyry stock; and (3) epithermal adularia-sericite-type
quartz-pyrite-chalcopyrite-sphalerite veins on the margins
of the stock.
Gold potential is believed to lie primarily in low
grade stockwork zones within the stock and in epithermal
acid-sulfate mineralization along structures within and
adjacent to the stock. Bulk samples from stockwork zones
yield gold values of 50 to 1,700 ppb, and values of 2,000
ppb have been measured in hand samples of acid-sulfate
mineralization. Potential for undiscovered stockwork and
(or) acid-sulfate mineralization exists at depth within the
system.
INTRODUCTION
La Espaola prospect is located at 1714' south lati
tude and 6931' west longitude in the Bolivian Andes, 187
km southwest of the capitol city of La Paz (fig. 1). The
prospect is on the northwest margin of the Altiplano
where it meets the eastern flank of the Cordillera Occiden
tal. The elevation of the valley floor in this part of the
Altiplano is about 4,000 m, and many of the volcanic
summits in the area rise over 5,000 m. The prospect is
96 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

76 72 68
12 I I
sensing, volcanology, and hydrothermal systems, followed
Lima
by field and laboratory studies of several anomalous areas.
PERU Processing of the remote-sensing imagery during the
first year lead to the discovery of a strong color anomaly
3- C)
at the La Espaola prospect that is indicative of iron
BOLIVIA oxides and sericite (Jimenez and others, 1991). This is the
largest and most intense anomaly recognized in the region.
|6 ||- Agy oLa Paz -
During the second year, the La Espaola prospect was the
La Espaola site of preliminary geologic mapping at scales of 1:50,000
N D. Oruro
and 1:10,000 by geologists at GEOBOL. Much of the
()
-

Arica mapping was conducted by Luis Barrera of GEOBOL as

part of his senior thesis at the University of San Andres in
La Paz. Since 1988, the potential for bulk minable pre
20 PACIFIC t 8 cious-metal deposits at La Espaola has been under inves
OCEAN |2 tigation by Expromin S.R.L. of La Paz.
This report is based on three days of field work spent
reviewing the geology, alteration, and mineralization at La
Antofagasta
CHILE a. -
Espaola and builds on the geologic framework studies of
24 r
u
geologists at GEOBOL and Expromin. Two days were
Salta
also spent collecting samples for geochemical studies that
will be the subject of another report. Several thin sections
* Tucuman
and X-ray diffraction spectra were also studied in the lab
u
Copiapo Y
oratory. The purpose of this preliminary report is to pro
28 - vide brief descriptions of some of the different types of
0. 200 400 km ARGENTINA alteration and mineralization present at La Espaola and to
La Serena
evaluate the significance of these features within the con
text of modern ore deposit models. Future field and labo
1 I
ratory studies are planned to test the validity of these first
impressions.
Figure 1. Location map showing the position of La Espaola
prospect within the Neogene to Quaternary volcanic arc of the cen
tral Andes (stippled pattern). Adapted from Ericksen and others
(1986). LOCAL GEOLOGY

All of the rocks exposed in the La Espaola area (fig.

within the Neogene to Quaternary volcanic arc of the cen
tral Andes (fig. 1). Much of the exploration effort in the
region has focused on the numerous prominent alteration
zones that are spatially associated with eruptive centers,
especially stratovolcanoes, caldera complexes, and dome
fields. Ore types present in this setting include Ag-Au
veins and disseminations; polymetallic Agbase metal
veins; polymetallic Sn veins; porphyry Au-Cu, Ag-Cu, and
Sn deposits; and numerous geothermal springs (Ericksen
and others, 1986; Cunningham and others, 1991). Recent
discoveries of large, low-grade precious metal deposits
include La Joya (Bolivia) and La Coipa (Chile), which are
mined by open-pit methods.
This report is an outgrowth of a two-year training
project funded by the Inter-American Development Bank.
The project is directed at training geologists from the
national geological institutions of Bolivia, Chile, and Peru
in modern techniques for studying and evaluating pre
cious-metal deposits in volcanic rocks of the central
Andes. The project included initial training in remote
2) are late Oligocene or younger in age and were depos
ited in a volcanic-chain depositional environment (Wallace
and others, 1991). The geology consists of an upper Oli
gocene to middle Miocene sequence of andesitic to dacitic
volcanic flows, volcaniclastic sedimentary rocks, and
lahars that were intruded by a composite dacite porphyry
stock at about 11 Ma. Alteration and mineralization are
centered on the dacite porphyry intrusive complex and
appears to be about the same age. Figure 3 is a view
looking southwest of the La Espaola prospect. The two
light-colored hills in the foreground, San Geronimo and
Jichu Cunca, are composed of the altered dacite porphyry
intrusive complex. The major lithologic units are
described below.
MAURI FORMATION
Rocks of the upper Oligocene to middle Miocene
Mauri Formation crop out along the east and south sides
 GOLD AND SILVER, LA ESPANOLA PROSPECT, NORTHWESTERN ALTIPLANO, BOLIVIA 97

8098
11.2 + 0.7 Ma e -

8096

11.8 + 0.8 Ma e Jichuc

NG. 4521 m. 8094

Trn Tik mine

370 Trn

A 10S,

Tik Tik

| 1 km |
8092
441 443 445 447

Explanation
C Pervasive quartz-sericite-pyrite alteration Qal Quaternary alluvium
SN Tid 1-4 Post-mineralization
& Gold-bearing quartz stockworks
dacite porphyry intrusives
G D Acid-sulfate alteration Tid Pre-mineralization dacite porphyry
Z- - -
intrusives, Tid" has miarolitic cavities
g Intrusive breccia Ttk Volcanics at Thola Kkollu
X Prospect pit Trm Mauri and Abaroa Formations

X- Adit O K-Ar sample site and date

Figure 2. Simplified geologic map of La Espaola prospect, modified from mapping by L. Barrera and Expromin geologists. AA' is the
line of cross section for figure 11. Coordinates in thousands of meters are from zone 19 of the transverse Mercator projection, international
spheroid.
98 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Jichu Cunca
- -
- -
. -
*** - -
Figure 3. Photograph of La Espaola prospect looking southwest. San Geromino hill (4,584 m) and Jichu Cunca hill (4,521 m) are
comprised of the altered dacite porphyry intrusive complex. Thola Kkollu hill (4,842 m) is visible in the background, and the crest of
the Cordillera Occidental is in the far background.
of La Espaola and strike approximately N. 30 W. and
dip to the west at about 40 (fig. 2). The steep dips are
probably due to block faulting associated with an episode
of middle Miocene uplift that affected this part of the Alti
plano (Wallace and others, 1991). The base of the Mauri
Formation is not exposed at La Espaola, although, about
20 km to the east, it lies unconformably on red arkosic
sandstones of the middle Oligocene Berenguela Formation
(Jimenez and others, 1991; Wallace and others, 1991). In
the Berenguela area, the Mauri Formation has been
divided into upper and lower parts. The upper part is not
present in the La Espaola area. The lower part consists
of purple and green volcaniclastic conglomerate, sand
stone, siltstone, and mudstones deposited in an alluvial-fan
environment. These rock units are overlain by the Abaroa
Formation, which consists of volcaniclastic breccias and
lahars that contain blocks of pyroxene andesite in a volca
niclastic matrix. Small pyroxene andesite flows and dikes
are also present. Recent mapping by GEOBOL and
USGS geologists indicate that the Abaroa Formation is the
time-equivalent western facies of the lower Mauri in the
Berenguela area. The Mauri and Abaroa Formations have
Thola Kkollu
San Geronimo
therefore been combined on figure 2 (area labeled Tm).
The upper age limit of the lower Mauri is constrained by a
K-Ar age on biotite from a tuff near Berenguela that
yielded an age of 13.6+1.1 Ma (Carlos Perez de Arce R.,
Laboratorio De Geocronologia, Servicio Nacional De Geo
logia Y MineriaChile, unpub. data). Evernden and oth
ers (1977) and Lavenu and others (1989) report ages for
the Abaroa Formation that range from 21.6 to 13.5 Ma.
VOLCANICS AT THOLA KKOLLU
At La Espaola, the Mauri Formation is unconform
ably overlain by a sequence of nearly flat-lying lahars and
dacite flows informally named here the volcanics at Thola
Kkollu (Ttk on fig. 2). Some of the flows appear to be
derived from a volcanic center at Thola Kkollu hill (fig.
3), 1 km southwest of the map area on figure 2, although
this remains to be confirmed. Dips measured in these
units are less than 12, suggesting that there has been little
or no tilting since deposition. The age of the rocks is
constrained by two K-Ar determinations. Hornblende
 GOLD AND SILVER, LA ESPANOLA PROSPECT, NORTHWESTERN ALTIPLANO, BOLIVIA
from a dacite dike that crosscuts one of the lahars yielded
an age of 11.2+24 Ma (Carlos Perez de Arce R. at the
Laboratorio De Geocronologia, Servicio Nacional De Geo
logia Y MineriaChile, unpub. data). Another horn
blende from a dacite flow from the same area yielded and
age of 11.2+0.7 Ma (M. Macintyre, Scottish Universities
Research Reactor Center, SURRC, U.K., unpub, data).
DACITE INTRUSIVE COMPLEX
A composite dacite porphyry stock was emplaced
into the Mauri Formation and the volcanics at Thola
Kkollu prior to mineralization (Tid and Tid* on fig. 2).
From existing exposures it was impossible to determine
whether the intrusive reached the surface, although if it
did, all remnants of the overlying volcanic edifice have
been removed. The stock measures about 1.5 km by 3 km
in plan, is sericitically altered, and locally hosts some of
the gold mineralization. Intrusive breccias are present at a
few localities within the complex (fig. 2). Five or more
intrusive phases are recognized, based on differences in
phenocryst assemblages or abundance of miarolitic cavi
ties. The contacts between all of the different intrusive
phases have not yet been mapped. The intrusive dacite
porphyry, labeled Tid on figure 2, is thought to consist of
four or more intrusive phases. Most of the composite
intrusive (Tid) is composed of two intrusive phases; one
that contains greater than 20 percent phenocrysts and
another that contains less than 5 percent phenocrysts. In
both, seriate zoned plagioclase (17 mm) and hornblende
(0.54 mm) are the most prominent phenocrysts. Biotite
(< 1 mm) is usually visible in thin section. Another phe
nocryst-rich intrusive phase on the southwest side of Jichu
Cunca hill can be distinguished by the presence of
embayed (~ 1 mm) phenocrysts of quartz. The dacite por
phyry intrusion mapped on top of Jichu Cunca hill (Tid*
on fig. 2) is a phenocryst-rich plagioclase-hornblende dac
ite that is distinguished by the presence of abundant miar
olitic cavities, 5 mm to 3 cm in size, lined with drusy
quartz.
On the west side of the complex is another dacite
intrusive phase that can be distinguished by the presence
of 3 to 6 mm phenocrysts of biotite. This intrusive is
sericitically altered except near its westernmost contact
where it is nearly fresh. Biotite from this locality (fig. 2)
yielded a K-Ar age of 11.8+0.8 Ma (M. Macintyre, Scot
tish Universities Research Reactor Center, SURRC, U.K.,
unpub. data).
Several near-vertical dacite porphyry dikes emanate
from the south side of the main intrusive body and appear
to occupy a radial fracture system (fig. 2). Some are phe
nocryst rich, and some are phenocryst poor. Two of the
dikes dip toward the main intrusive body at about 45 and
may have been emplaced along ring fractures or, more
99
Figure 4. Photomicrograph of unaltered plagioclase-hornblende
dacite porphyry from the top of San Geronimo hill (Tidl on fig. 2).
Note the seriate zoned plagioclase, twinned hornblende, and apha
nitic groundmass. Field of view is 16 mm wide, crossed polars.
likely, along the steeply dipping bedding of the Mauri
Formation. The contact between the dacite porphyry and
the Mauri Formation is well exposed in this area and is
knife sharp.
The early-altered dacites (Tid and Tid*) are intruded
by several small dacite porphyry dikes and plugs that are
fresh or exhibit weak propylitic alteration. A small pla
gioclase-hornblende dacite porphyry plug is present (Tid1
on fig. 2) on top of San Geronimo hill that is mineralogi
cally and texturally very similar to the pre-mineralization
phenocryst-rich dacite except that it is unaltered and lacks
miarolitic cavities. The close similarity of this plug to the
early-altered dacites (Tid and Tid*) suggests that both
were derived from the same parent magma. Figure 4 is a
photomicrograph of a sample from this intrusive (Tid*)
showing seriate zoned plagioclase and twinned hornblende
in an aphanitic groundmass. This plug crosscuts quartz
sericite-pyrite (QSP) and acid-sulfate-altered dacite por
phyry (fig. 2). Also present are green- and brown-colored
plagioclase-hornblende dikes (Tid2 and Tid3 on fig. 2) that
contain about 10 percent phenocrysts and have much
smaller phenocrysts (< 3 mm) than the other dacite intru
sions. Some of these dikes cut QSP alteration and gold
bearing stockwork zones. Another unaltered phenocryst
rich plagioclase-hornblende dacite porphyry projects
through alluvial cover in the northeast part of the map area
(Tid4 on fig. 2). Miarolitic cavities were not observed in
any of the post-mineralization intrusives.
The K-Ar ages suggest that the volcanics at Thola
Kkollu and the pre-mineralization dacite porphyry stock
are about the same age, 11 Ma. Mineralization is younger
than the volcanics at Thola Kkollu (11.2+0.7 Ma) that are
altered and mineralized and younger than the biotite-rich
intrusive phase of the composite dacite porphyry stock
that is also altered and mineralized (11.8+0.8 Ma).
100 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

ALTERATION AND MINERALIZATION

Several types of alteration and mineralization are

present in and around the intrusive complex. Some
exhibit characteristics of porphyry-style mineralization and
others exhibit characteristics of epithermal acid-sulfate
and adularia-sericite-type mineralization. Descriptions of
these rocks and the genetic implications of some of the
features present are discussed below.

DRUSY QUARTZ VUGS

Vugs lined with drusy quartz that are up to 3 cm in

diameter (e.g., fig. 5A) occur sporadically throughout the
intrusive complex but are most common in the dacite por
phyry exposed near the summit of Jichu Cunca hill (Tid*
on fig. 2). At some localities, pyrite and chalcopyrite are
associated with quartz. Quartz crystals in these cavities
contain secondary vapor-dominant inclusions and aqueous . . * T

inclusions that contain halite daughter minerals (figs. 5B

and 5C). The fluid inclusions are characteristic of the por b. . . . * : . C.
phyry environment and provide evidence of boiling. In
----------:
some cases, the quartz also contains microscopic inclu --:
sions of rutile (fig. 5D), another feature commonly found
in porphyry systems (Rose and Burt, 1979; Cox, 1986).
Vugs lined with drusy quartz occur regardless of alteration
zone and appear in vuggy silica, quartz-alunite, and QSP
altered rocks. Alunite encrustations on top of the drusy
quartz clearly show that acid-sulfate alteration postdates
the drusy quartz. The vugs of drusy quartz are thought to
be miarolitic cavities or segregations that formed from a
residual, immiscible, low-density fluid as the dacite por
phyry intrusions cooled. The presence of miarolitic cavi
ties in the pre-mineralization intrusives attests to the high
volatile content of the magmas. -
-

: -

Figure 5. A, Photograph of miarolitic cavities in dacite porphyry

PERVASIVE ALTERATION
The dacite porphyry intrusive complex and portions
of the surrounding volcanic host rocks are pervasively
altered to an assemblage of quartz, sericite, and pyrite
(QSP) (fig. 2). Thin sections show that, in the least
altered samples, mafic minerals are destroyed and plagio
clase is partially altered to very fine grained (< 20
microns) sericite. In the most intensely altered samples,
all of the rock-forming minerals have been converted to an
assemblage of quartz, sericite, and pyrite, and the sericite
is coarser, ranging up to 1 mm (fig. 6). In a few samples,
crystals of potassium feldspar are present that locally are
partially to completely replaced by sericite (fig. 6). X-ray
diffraction studies of samples from Jichu Cunca hill and
San Geronimo hill show that QSP-altered dacite often
contains gypsum and occasionally kaolinite or potassium
from Jichu Cunca hill (Tid*). The scale is in centimeters. B, Pho
tomicrograph of secondary gas-dominant inclusion with halite
daughter mineral. C. Photomicrograph of secondary aqueous inclu
sion with halite daughter mineral. The fluid inclusions are in quartz
from a miarolitic cavity. The black bar is 12 microns long. D, Pho
tomicrograph of miarolitic cavity filled with quartz containing in
clusions of rutile. Field of view is 1 mm wide, plane polarized light.
feldspar in addition to sericite. The gypsum is thought to
be the secondary hydration product of primary anhydrite.
The kaolinite may be hydrothermal in origin or, more
likely, is a supergene alteration product related to the oxi
dation of pyrite.
QSP alteration is largely restricted to the dacite por
phyry intrusive complex, although locally, the surrounding
host rocks are also altered (fig. 2), especially along
 GOLD AND SILVER, LA ESPANOLA PROSPECT, NORTHWESTERN ALTIPLANO, BOLIVIA
Figure 6. Photomicrograph of fibrous sericite, rhomb-shaped po
tassium feldspar, and opaque pyrite in a sample from the southwest
side of Jichu Cunca hill. Fluid inclusions are visible in the potassi
um feldspar. Field of view is 4 mm wide, crossed polars.
fracture zones. For example, the dacite porphyry dikes
emanating from the south side of the intrusive complex
exhibit weak QSP alteration even though some of the
dikes are less than 1 m in width. The host upper Mauri
Formation at this locality is only propylitically altered
even when in direct contact with QSP-altered dacite por
phyry. However, on the north and west sides of the main
intrusion, the volcanic host rocks exhibit pervasive QSP
alteration that grades outward into propylitic alteration.
Within 0.5 to 1 km of the intrusive complex, the
Mauri Formation and volcanics at Thola Kkollu are propy
litically altered and contain chlorite, variable amounts of
disseminated pyrite, calcite + quartz veinlets, disseminated
calcite, and disseminated epidote. On the northeast side of
the intrusive, in the vicinity of the Tarutani veins, dissemi
nated hematite in the Mauri Formation is present that may
be hydrothermal in origin.
The presence of relict potassium feldspar and gypsum
(after anhydrite) is important because K-spar and anhy
drite are characteristic of the high-temperature, potassic
zone of porphyry systems (Rose and Burt, 1979). The
pervasive QSP alteration zone apparently collapsed
inward, overprinting the earlier potassic alteration at La
Espaolathis is common in many porphyry systems.
STOCKWORK MINERALIZATION
Previous to our field work, Expromin geologists iden
tified three areas of stockwork mineralization that contain
gold. Each of the stockwork zones is located near the
margin of the dacite porphyry intrusive complex (fig. 2).
Figure 7A is a photograph of stockwork mineralization out
cropping on the east side of San Geronimo hill. Figure 7B
is a closer view of the stockworks that occur throughout
101
--
-
- -
--
ls N
Figure 7. A, Photograph of gold-bearing stockwork zone on the
east side of San Geronimo hill. B, Outcrop of stockwork zone that
contains 1.2 ppm gold. Each veinlet contains quartz, sericite, and
iron oxides after pyrite.
this zone. The stockworks consist of several generations of
quartz veinlets, typically less than 2 mm in width, that also
contain sericite and pyrite. Greater concentrations of dis
seminated pyrite are observed in and around the stockwork
zones. The fracture density is around 10 fractures per
square meter, and the fractures lack a strong preferred ori
entation. The intensity of gold mineralization in the stock
work zones correlates with the amount of quartz
introduction; i.e., the more prominent the quartz veinlets,
the higher the gold values (R. Terrazas, Expromin, oral
commun., 1991). Fracture density is less important.
Chemical analyses of channel samples (~ 0.1 m wide by 3
m long) and of bulk samples from prospect pits in stock
work zones yield gold values of 50 to 1,700 ppb (R. Terra
zas, Expromin, oral commun., 1991). High gold values
correlate spatially with high copper values. Silver values
range between about 0.5 and 20 ppm, and copper values of
up to 0.16 percent were encountered. The Ag/Au ratios in
all three areas are generally less than 10. Arsenic, anti
mony, silver, lead, and zinc concentrations are sporadically
102 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Figure 8. Photograph of hand specimen showing vuggy silica al

teration developed in brecciated dacite porphyry from the south
west side of San Geronimo hill. The dark material is geothite and
hematite after pyrite. Scale is in centimeters.

anomalous, but mercury is not (R. Terrazas, Expromin, oral

commun., 1991). Drainage sediments below stockwork
mineralization on the northeast side of San Geronimo con
tain gold flakes up to 3 mm in maximum dimension, with 1
mm flakes being relatively common. This suggests that the
gold in the stockwork zones is fairly coarse grained. The
geologic setting, mineralogy, and geochemical signature of
the stockwork mineralization at La Espaola is similar to
the gold-rich porphyry systems of the Maricunga belt of
Chile (Vila and Sillitoe, 1991) and to other gold-rich por
phyry systems on the Pacific Rim (Rytuba and Cox, 1991).

ACID-SULFATE ALTERATION ON
SANGEROMINO HILL

Several zones of acid-sulfate alteration are present on

top of San Geronimo hill (fig. 2). These rocks form
prominent blocky outcrops resistant to erosion. Two types
of acid-sulfate alteration are present that are superimposed
on the earlier, pervasive QSP alteration. Vuggy silica
alteration consists of the residual silica left over after
extreme acid leaching of the host rock and variable
amounts of sugary quartz introduced by the hydrothermal
fluids (fig. 8). Quartz-alunite alteration consists of a fine
grained intergrowth of quartz, alunite, and pyrite that
locally contains clots of bladed alunite up to 2 mm in size
(e.g., fig. 9B). The vuggy silica and quartz-alunite-altered
rocks usually contain abundant red and brown iron oxides
after pyrite. Occasionally vuggy silica can be observed to
grade outward into quartz-alunite alteration, but more
often it grades abruptly into QSP alteration. No barite or
Figure 9. A, View looking east of vuggy silica and quartz
alunite altered dacite flow at the Kollota mine. Vuggy silica rub
ble on top of San Geronimo hill is visible in the upper right. B.
Photomicrograph of quartz-alunite alteration. Lath-shaped crys
tals are alunite; ground mass is predominantly quartz. Field of
view is 5 mm wide, crossed polars. C, Barite-enargite veinlet in
quartz-alunite-altered dacite flow from the Kollota mine dump.
sulfides other than pyrite were observed in these outcrops.
X-ray diffraction studies of samples from this locality
show that kaolinite and pyrophyllite are also present.
Some of the vuggy silica zones form dike-like bodies with
a strong vertical aspect. Float boulders of vuggy silica
 GOLD AND SILVER, LA ESPANOLA PROSPECT, NORTHWESTERN ALTIPLANO, BOLIVIA
were also observed on Jichu Cunca hill, suggesting that
acid-sulfate alteration is present in the southern part of the
intrusive complex.
The geologic setting, presence of vuggy silica with
quartz-alunite alteration, and the dike-like morphology of
the altered zones are characteristic of magmatic hydrother
mal acid-sulfate systems (Rye and others, 1991). This
type of acid-sulfate alteration typically develops at shallow
levels some distance above the mineralized portions of
porphyry systems (Vila and Sillitoe, 1991). These rocks
are also often the site of subsequent precious-metal miner
alization (Heald and others, 1987; Hedenquist, 1987). The
textural characteristics of the acid-sulfate alteration pre
served on San Geronimo suggest that it formed at some
distance below the surface and that there has been signifi
cant erosion since its formation. Opaline silica and chal
cedonic quartz are absent; quartz and alunite are relatively
coarsely crystalline; pyrophyllite is present, suggesting
temperatures of 260C or more (White and Hedenquist,
1990); and no flat-lying paleosurface or paleo-water-table
features are preserved.
A sample of coarsely crystalline (12 mm) alunite
from the acid-sulfate-altered zone on top of San Geronimo
hill (fig. 2) was dated by the K-Ar method and yielded an
age of 10.3+0.3 Ma (E.H. McKee, unpub. data). This is
about 1 m.y. younger than the age (11.8+0.8 Ma) of the
biotite-rich phase of the host intrusive complex.
KOLLOTA MINE
The Kollota mine is located on the west side of the
intrusive complex where several prospect pits and an adit
are developed along an east-west-striking fracture zone
(fig. 2). The volcanic dacite flows along the structure are
pervasively altered in a 2- to 4-m-wide zone to vuggy sil
ica, quartz-alunite, or quartz-kaolinite that are resistant to
erosion (fig. 9A). Figure 9B is a photomicrograph of
quartz-alunite alteration from this outcrop. The acid-sulfate
alteration is surrounded by a 10- to 25-m-wide envelope of
quartz-sericite-pyrite alteration that abruptly grades into the
propylitically altered volcanics. Locally, the fracture zone
is brecciated and partially cemented with sugary, fine
grained quartz and pyrite. Unoxidized samples of vuggy
silica often contain as much as 10 percent disseminated
pyrite (< 1 mm size), and barite crystals up to 2 cm long are
locally present in open spaces. Some of the acid-sulfate
altered dump samples are also crosscut by barite-enargite
veinlets. Figure 9C is a photo of a hand sample that split
along a barite-enargite veinlet. A rock sample containing
barite-enargite veinlets contained 2 ppm gold (R. Terrazas,
Expromin, oral commun., 1991). X-ray diffraction studies
of samples from the Kollota mine dumps confirmed the
field identification of alunite, barite, kaolinite, and sericite.
103
A small, dark reddish-brown gossan about 1 by 3 m in size
was also observed at this locality, suggesting that larger
concentrations of sulfides are present locally within this
ZOne.
The geologic features of the Kollota mine are charac
teristic of epithermal acid-sulfate or high-sulfidation sys
tems (Heald and others, 1987; Hedenquist, 1987). The
geologic setting, mineral assemblages, and two-stage
paragenesis characterized by early acid-sulfate alteration
followed by later quartz-pyrite-barite-enargite-gold precip
itation is typical of these systems.
SANTA ROSA MINE
On the south side of Jichu Cunca hill are several
prospect pits and adits of the Santa Rosa mine that date
from Spanish Colonial times. The workings are along a
pyritic, silicified fracture zone that contains high silver
values. The structure strikes approximately N. 45 E.
within QSP-altered dacite porphyry (fig. 2). Many of the
rocks on the dump are breccias that contain fragments of
vuggy-silica-altered dacite porphyry that are cemented by
sugary, fine-grained quartz and pyrite. However, no
alunite or kaolinite alteration was recognized in the field.
Some of the breccias are cemented by a later stage of fine
grained quartz containing pyrite, barite, sphalerite, galena,
and a relatively abundant, unidentified, dark-gray sulfosalt
mineral. The sulfides and sulfosalt minerals are generally
less than 2 mm in size, whereas barite crystals range up to
1 cm. A later stage of silicification consists of gray to
white chalcedonic quartz that locally fills the remaining
porosity in the breccia. Mineralization and silicification at
Santa Rosa appears to be younger than the pervasive QSP
alteration in the host dacite porphyry.
The mineralization at Santa Rosa has many similari
ties to the epithermal acid-sulfate mineralization at Kol
lota, except that alunite and kaolinite alteration is poorly
developed, and greater amounts of silica and base-metals
were introduced. The breccia fragments of vuggy silica,
intense silification, abundance of pyrite and barite, and the
relatively small amounts of open-space filling are charac
teristic of epithermal acid-sulfate systems.
TARUTANIVEINS
On the northeast flank of the main intrusive body, the
lower Mauri Formation is host to Cu-Zn-bearing quartz
veins of the adularia-sericite type (figs. 2 and 10). All of
the Tarutani veins fill open spaces along fractures with lit
tle or no silicification of the wallrock. Pyrite, chalcopy
rite, and sphalerite are the main sulfides present. To date,
no geochemistry has been done to determine whether the
veins carry significant gold or silver values. Figure 2
104 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

shows two prospect pits that were developed on two veins,

1 to 3 cm in width, each enclosed within a narrow 13 cm
envelope of QSP alteration. Mine adits were developed
on the two largest veins. The northern adit exploited a 6
to 10-cm-wide vein that strikes N. 62 E. and dips 70 to
the north (fig. 10A). It is composed of several generations
of 1- to 3-mm-size quartz that contains 1 to 10 volume
percent sulfides. The vein is locally brecciated. The sul
fides in the vein are generally less than 5 mm in size.
Some of the vein material on the dump consists of quartz
that contains about 5 volume percent chalcopyrite. This
vein is surrounded by a symmetrical 2- to 4-cm envelope
of QSP alteration that abruptly grades outward into propy
litic alteration (fig. 10B). The southern adit exploited a 1
m-wide vein that strikes N. 70 E. and dips 85 to the
north. This vein is mineralogically similar to the others
except that it is brecciated and contains fragments of wall
rock. The QSP alteration envelope around this vein is 4 to
10 cm wide. Although adularia was not identified in any
of the Tarutani veins or alteration, the mineralization is
very similar to adularia-sericite-type veins commonly
found on the margins of porphyry systems (Margolis and
others, 1991).

SUPERGENE ALTERATION

Supergene processes resulted in oxidation and leach

ing of sulfides to depths of generally less than 35 m.
Recent prospect pits and trenches often reveal fresh rock
with unoxidized sulfides a few meters below the surface.
Geothite, hematite, and locally jarosite are the primary
oxidation products of pyrite. Sulfuric acid produced by
this process locally altered silicates to form kaolinite.
Hydration of anhydrite produced gypsum. Supergene cop
per minerals including digenite, chalcocite, malachite, and
azurite are present in dump material from the Tarutani
veins. The lack of intense supergene oxidation is proba
bly due to the arid climatic conditions that existed in the
region since the middle Miocene (Vila and Sillitoe, 1991).
The lack of deep oxidation is important because the pyrite
and other sulfides present in unoxidized rock can have a
deleterious affect on the recovery of gold by heap-leach
ing methods.
DISCUSSION
A preliminary hypothesis that is consistent with our
observations is that La Espaola represents the juxtaposi
tion of a magmatic-hydrothermal, acid-sulfate system on an
earlier, gold-rich porphyry system with development of
adularia-sericite-type veins on the periphery. Although we
do not yet have the data to fully support these speculations,
Figure 10. A, View along strike of the northern Tarutani vein. B,
Photograph showing the narrow QSP alteration halo around the
quartz-pyrite-chalcopyrite-sphalerite vein.
they provide a starting point for further field and laboratory
studies. We envision a sequence of igneous and hydrother
mal events as follows.
After deposition of the Mauri and Abaroa Forma
tions, during late Oligocene to middle Miocene time, and
after a period of uplift and erosion in the middle Miocene,
the La Espaola area became the site of renewed volcan
ism. Eruption of dacite flows and fragmental volcanics at
about 11 Ma, from a volcanic center at Thola Kkollu and
from other volcanic centers nearby, resulted in the forma
tion of the volcanics at Thola Kkollu. During the latter
part of this episode, igneous activity commenced at La
Espaola and produced the dacite porphyry intrusive com
plex. The geologic relations suggest that La Espaola was
the site of repeated intrusive activity before, during, and
after mineralization. As some of the more volatile-rich
dacite porphyry intrusions cooled and crystallized, a low
density fluid or vapor separated to form miarolitic cavities.
As other volatile-rich intrusions crystallized, dense mag
matic brines separated, moved into surrounding fracture
systems, and boiled to produce potassic alteration contain
ing pyrite, chalcopyrite, it gold. As the generation of
 GOLD AND SILVER, LA ESPANOLA PROSPECT, NORTHWESTERN ALTIPLANO, BOLIVIA 105

San Geronimo hill

A
Acid-sulfate
! A"
alteration
METERS

soo-N-Tholakkollu hill
4400
| mineralization

Tarutani
-300 veins

4200

--00
Magmatic vapor plume
--~~~

Tim (?)
3900 -----

Possible stockwork\G
3700

~~ \- water

3500
Inferred porphyry intrusion
3400

Figure 11. Schematic cross section along the line of section AA on figure 2 showing the major features of La Espaola prospect. Two
phases of the dacite porphyry stock (Tid) are shown, although the position of the contact between the two intrusive phases is only approxi
mate. One or the other of these intrusive phases is thought to be responsible for the stockwork zone observed at the surface. Also shown
is an inferred porphyry intrusion that is thought to be the source of the magmatic volatiles that produced the acid-sulfate alteration on San
Geronimo hill and at the Kollota mine. The inferred porphyry intrusion may also have produced stockwork mineralization similar to that
observed at the surface. The curved arrows represent possible flow paths of convecting meteoric water. See figure 2 for explanation of map
symbols. See the Discussion section of the text for more details.

magmatic fluids waned, convectively circulating cells of
meteoric water collapsed into the earlier formed zone of
potassic alteration to produce pervasive quartz-sericite
pyrite alteration. The QSP stockwork zones may have
formed outboard of the potassic zone as magmatic brines
cooled and mixed with meteoric water. This sequence of
intrusion, magmatic fluid generation and escape, and col
lapse of meteoric fluids into the system may have
occurred several times. The three stockwork zones dis
covered at La Espaola may have been produced by sepa
rate intrusive events.
Magmatic hydrothermal acid-sulfate alteration formed
at shallower levels in the system. The vuggy silica and
quartz-alunite alteration observed on San Geronimo hill, at
the Kollota mine, and at the Santa Rosa mine typically
forms at the top of the magmatic vapor plume (fig. 11)
where volatiles such as HCl, HF, and SO2 condense and
where SO2 disproportionates to H2SO4 and H2S (Rye and
others, 1991). In the zone of condensation above the
vapor plume, small volumes of extremely acidic fluid are
generated that are progressively neutralized by reaction
with the host rocks and by dilution with meteoric water.
The acid-sulfate-altered rocks produced by this process are
usually permeable and often are the site of subsequent
quartz-pyrite-barite + enargite + gold + silver + base-metal
mineralization such as that observed at the Kollota and
d Santa Rosa mines. Ore deposition in acid-sulfate systems
has been attributed to cooling, to pH increases associated
with progressive mixing between magmatic and meteoric
waters, and to boiling (White and Hedenquist, 1990).
Acid-sulfate alteration and mineralization typically
forms well above the level of potassic alteration in por
phyry systems (often 500700 m). The presence of acid
sulfate alteration within rocks formerly subjected to potas
sic alteration at La Espaola suggests that the hydrology
of the system changed significantly. A drop in the water
table or erosional unroofing during the life span of the
igneous-hydrothermal system has been called upon to
explain the lowering of magmatic-hydrothermal acid-sul
fate alteration into the sericitic and potassic zones of por
phyry systems (White and Hedenquist, 1990; Vila and
Sillitoe, 1991). Further, based on the magmatic hydrother
mal acid-sulfate model (Rye and others, 1991), it is possi
ble to infer the presence of an unexposed intrusion at
depth (e.g., fig. 11) that was the source of the magmatic
volatiles necessary to produce the acid-sulfate alteration
observed at the surface. The inferred intrusion may also
have produced stockwork mineralization similar to that
observed at the surface (fig. 11).
The dacite porphyry intrusive complex provided the
thermal drive for convecting cells of meteoric water (fig.
11) that produced widespread propylitic alteration in the
surrounding host rocks. Locally, these fluids were
focused along high-angle faults and fractures in the
106 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
upwelling part of the cell to produce adularia-sericite-type
veins, such as the Tarutani veins (fig. 11). Ore deposition
in these veins may have occurred by mixing of meteoric
water and fluids emanating from the intrusive or by boil
ing. The predominance of base-metal sulfides and lack of
adularia in the veins suggests that fluid mixing was more
important.
After the early dacite porphyry intrusions cooled and
the hydrothermal system collapsed, La Espaola continued
to be the site of intrusive activity. Numerous small plugs
and dikes were emplaced into rocks that had previously
been the site of hydrothermal activity. The later dacite
porphyry intrusions probably remained unaltered because
they contained lower concentrations of volatiles or were
too small to drive convecting cells of meteoric water.
Uplift and erosion over the last 10 million years
exposed the altered and mineralized rocks and produced
the topographic relief observed today. Gold-bearing rock,
which likely was present in the shallower parts of the sys
tem, was removed by erosion. Potential for placer gold
may therefore exist in alluvial valley fill on the flanks of
the La Espaola prospect.
ACKNOWLEDGMENTS
The authors would like to acknowledge our colleges,
F. Tavera, N. Jimenez, J.L. Lizeca, F. Murillo, O. Flores,
and R. Learned for stimulating discussions in the field that
improved our understanding of La Espaola. We thank
our drivers W. Mallea, W. Nina, J. Valdez, and R. Tapia
for their help and support in camp. Publication was
encouraged by F. Tavera and authorized by GEOBOL.
Special thanks are offered to R. Terrazas of Expromin
S.R.L. for allowing us to examine geologic maps and
geochemical data collected during the course of explora
tion at La Espaola and for discussions regarding the ori
gin of the mineralization. We are grateful to S. Sutley and
T. Botinelly for help with the identification of minerals
from X-ray diffraction spectra. And lastly, we thank the
Inter-American Development Bank for funding.
REFERENCES CITED
Cox, D.P., 1986, Descriptive model of porphyry Cu-Au, in Cox,
D.P., and Singer, D.A., eds., Mineral Deposit Models: U.S.
Geological Survey Bulletin 1693, p. 110114.
Cunningham, C.G., McNamee, J., Vasquez, J.P., and Ericksen,
G.E., 1991. A model of volcanic dome-hosted precious metal
deposits in Bolivia: Economic Geology, v. 86, p. 415421.
Ericksen, G.E., Eyzaguirre, V.R., Urquidi, F., and Salas, R., 1986,
Neogene-Quaternary volcanism and mineralization in the cen
tral Andes, in Horn, M.K., ed., Transactions of the Fourth Cir
cum-Pacific Energy and Mineral Resources Conference,
Singapore, p. 537550.
Evernden, J.F., Kriz, J.S., and Cherroni, C., 1977, Potassium-ar
gon ages of some Bolivian rocks: Economic Geology, v. 72,
p. 10421061.
Heald, P.W., Foley, N.K., and Hayba, D.O., 1987, Comparative
anatomy of volcanic-hosted epithermal deposits: Acid-sulfate
and adularia-sericite types: Economic Geology, v. 82, p. 125.
Hedenquist, J.W., 1987, Mineralization associated with volca
nic-related hydrothermal systems in the circum-Pacific ba
sin, in Horn, M.K., ed., Transactions of the Fourth Circum
Pacific Energy and Mineral Resources Conference, Sin
gapore, p. 513524.
Jimenez, N., Lizeca, J.L., Murillo, F., Sanjines, O., Barrera, L., and
Flores, O., 1991, Aplicacion de metodos de percepcion remota
en el estudio geolgico del area Berenguela-Charana, Bolivia
[WorkshopRemote sensing applied to geological investiga
tions]. Sevicio Geolgico de Bolivia, Proyecto BID - USGS -
GEOBOL,1221 de Agosto de 1991, 8 p.
Lavenu, A., Bonhomme, M.G., Vatin-Perignon, N, and de Pacht
ere, P., 1989, Neogene magmatism in the Bolivian Andes be
tween 16 S. and 18 S. Stratigraphy and K/Ar
geochronology: Journal of South American Earth Sciences, v.
2, p. 3547.
Margolis, J., Reed, M.H., and Albino, G.V., 1991, A process orient
ed classification of epithermal systems: Magmatic volatile
rich versus volatile-poor fluid paths [abs.]: Geological Society
of America, Abstracts with Programs, v. 23, no. 5, p. A230.
Rose, A.W., and Burt, D.M., 1979, Hydrothermal alteration, in Bar
nes, H.L., ed., Geochemistry of Hydrothermal Ore Deposits,
2nd ed., New York, John Wiley and Sons, p. 173235.
Rye, R.O., Bethke, P.M., and Wasserman, M.D., 1991, The stable
isotope geochemistry of acid-sulfate alteration and vein form
ing alunite: U.S. Geological Survey Open-File Report 91257,
58 p.
Rytuba, J.J., and Cox, D.P., 1991, Porphyry gold: A supplement to
U.S. Geological Survey Bulletin 1693: U.S. Geological Sur
vey Open-File Report 91116. 7 p.
Vila, T., and Sillitoe, R.H., 1991, Gold-rich porphyry systems in
the Maricunga belt, northern Chile: Economic Geology, v.
86, p. 12381260.
Wallace, A.R., Hardyman, R.F., Tosqal, R.M., Jimenez, N., Lizeca.
J.L., and Murillo, F., 1991, Episodic midlate Tertiary defor
mation during basin sedimentation and explosive volcanism.
northwestern Altiplano, Bolivia {abs.]: Geological Society of
America, Abstracts with Programs, v. 23, no. 5, p. A 132.
White, N.C., and Hedenquist, J.W., 1990, Epithermal environments
and styles of mineralization: Variations and their causes, and
guidelines for exploration: Journal of Geochemical Explora
tion, v. 36, p. 445474.
 CHAPTER K
MINERALOGY AND CHEMISTRY OF
GOLD-ASSOCIATED SKARN FROM NAMBIJA,
ZAMORA PROVINCE, ECUADOR:
A RECONNAISSANCE STUDY
By JANE M. HAMMARSTROM"
ABSTRACT
Gold mineralization is associated with skarn formed
in a heterogeneous xenolith or roof pendant within the
Jurassic Zamora batholith in southeastern Ecuador. Petro
graphic descriptions, electron microprobe data for miner
als, and geochemical data for rock samples are presented
for a suite of 16 samples collected as part of a reconnais
sance study of the Nambija gold deposit (McKelvey,
1991). Hydrothermal alteration affected all the lithologies
present at Nambija, including syenite porphyry, volcanic
rocks and tuffs, and breccias. Zoned, anisotropic grossu
lar-andradite garnet partially replaces phenocrysts and
groundmass in breccia and is the dominant mineral in
massive skarn; isotropic, pure andradite is locally devel
oped. Other minerals present in skarn include diopsidic
pyroxene, epidote, K-feldspar, actinolite, chlorite, quartz,
apatite, calcite, pyrite, and traces of chalcopyrite. Free
gold is present as electrum in quartz within massive garnet
skarn and in garnet-rich breccia; gold is not associated
with sulfide minerals in the sample suite studied. Temper
ature estimates from chlorite geothermometry are compati
ble with a < 300C retrograde skarn stage. The Nambija
deposit has some characteristics of relatively oxidized
gold-bearing skarn deposits elsewhere but is unusually
gold rich and copper poor.
'U.S. Geological Survey, Mail Stop 959, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092.
INTRODUCTION
Gold has been mined in the Zamora Province of
southeastern Ecuador since the 16th century. In the early
1980's, the gold potential of the area was rediscovered,
initiating a modern gold rush centered around the town of
Nambija, which is along the eastern edge of the Cordillera
de Nanguipa (Barragn, 1989). Gold deposits at Nambija,
and nearby at Campana and Campanilla (fig. 1), form a
north-south belt of structurally controlled mineralization
hosted by a polylithologic xenolith or roof pendant within
the Jurassic Zamora batholith. The xenolith rocks are ten
tatively correlated with the Jurassic Santiago and Macuma
Formations (Salazar, 1988). McKelvey (1991) described
the general geology of the Nambija deposits and showed
that the host rocks at Nambija include limestone, shale,
sandstone, and volcanic tuffs and flows, all of which are
variably altered to skarn. Mineralization appears to be
concentrated in a structurally controlled zone that extends
for 1.5 km along strike, especially where breccias, faults,
and dikes cut the skarn assemblages. The extremely high
gold grade of the Nambija ore, purported to be in excess
of 50 g/t (grams per tonne), has previously attracted inten
sive, small-scale, unorganized mining activities and,
recently, the attention of the international mining industry.
The coarse-grained nature of the gold at Nambija and also
at Campanilla makes it difficult, due to nugget effects, to
assess ore grades based on reconnaissance sampling. No
reliable grade or tonnage data are available, but geologists
visiting the area in the late 1980's speculated that the
Nambija deposit has a potential ore tonnage of 25,000,000
short tons (23,000,000 metric tonnes) or more and that
grade estimates for the active mining at the time of their
visit were probably on the order of 0.5 to 1 oz Au/short
108 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

79 deposit are well above the median values of 213,000 met

i
ric tonnes and 8.6 g/t Au, respectively, calculated for 39
gold-bearing skarns by Theodore and others (1991). If
Nambija is a deposit of this type, it is certainly one of the
4 *
-
f L -

oja
Pl
aCes
s
(Cmpana & T
most significant gold-bearing skarns in the world.
*a.
Zamoras frambla : METHODS OF STUDY
, , ., ,. . ." S, s i

Paleozoic * \ , , , X."
Petrographic descriptions reported here are based on
metamorphic rocks * , , , , s examination of hand specimens and thin sections for 16
(Zamora Group) . . . . . ;-) samples collected by McKelvey (1991). Polished thin sec
Zamora (; tions were prepared for 11 samples; several sections were
tholith ,:
N. S - - -- prepared for some samples to include variations in texture
\ *v - 1> .
within individual specimens. Minerals were analyzed by
*~ , , .
the author using an ARL-SEMQ electron microprobe at
t . .
t * . S- S
\ * .. . $ the U.S. Geological Survey in Reston, Va. Natural and
ECUADOR ... .. synthetic mineral standards were used for silicate miner
\w
>, --
-
als; synthetic electrum standards were used for precious
\ > - s metal analyses. The number of analyses determined for a
PERU is ...' given mineral reflects the modal abundances of that min
eral in the sample suite. A scanning electron microscope
N. 0 20
was used for qualitative analyses and identification of sub
X------- -
5* * . KILOMETERS -
- I microscopic minerals. Fifteen 2-kg bulk-rock samples (fig.
2, NAM-1 through NAM-15) were analyzed for a suite of
Figure 1. Location map for the Nambija, Campanilla, Campana, 44 elements in U.S. Geological Survey analytical laborato
and placer gold deposits, Zamora Province, Ecuador. Geology from ries in Denver.
Mapa Geolgico Nacional de la Repblica del Ecuador (1982).

ton (14 to 484 g/t) (Mineral Resources Data System, 1991,
record number TC00085).
Campanilla, a gold deposit recently developed by Pla
teau Mining Plc., reportedly has a significant tonnage of
ore with gold grades of 0.5 oz/ton or more (Mining Maga
zine, 1990). The deposit at Campanilla is hosted by andes
itic volcanic tuff. The tuff is locally altered to skarn; gold is
associated with late quartz-carbonate veins that cut the
skarn. Campanilla is classified as a vein-type gold deposit
(Mining Magazine, 1990) that postdates skarn formation.
Nambija has been described (Salazar, 1988) as a complex
skarn deposit that cannot be readily classified using the
criteria for different classes of calcic skarns defined by Ein
audi and others (1981). Among the unusual features of the
skarn at Nambija, Salazar (1988) noted that gold is present
in a variety of textures and lithologies, that scheelite is
present with sulfides, and that gold is deposited late in the
paragenesis and may have been remobilized.
Mineral assemblages are one of the most useful
guides for classifying skarn deposits and for understanding
the environment of ore deposition. In this study, new data
on mineral chemistry are presented for samples collected
by McKelvey (1991) from Nambija, and the data are eval
uated in terms of recent models for gold skarns (Meinert,
1989; Theodore and others, 1991; Ettlinger and Ray,
1989). The apparent size and grade of the Nambija
SAMPLE DESCRIPTIONS
Lithologies sampled include altered syenite porphyry,
tuff, breccia, chlorite skarn, and garnet skarn. Sample
locations and the distribution of lithologies are shown in
figure 2. McKelvey's (1991) sketch map of the district
shows the distribution of outcrops; it also shows a central
zone of garnet skarn cut by dikes and breccias that is
flanked by an outer zone of chlorite skarn. Mining activity
is concentrated along the central zone. The distribution of
rock types on figure 2 is generalized; the distribution of
garnet-rich versus chlorite-rich skarn in this sample suite
is much more erratic. This relation is not surprising in
light of the heterogeneous skarn protolith and the extent of
shearing and brecciation in the immediate area of the
deposit.
Rock types (field names assigned by McKelvey,
1991) and mineral assemblages for samples studied by
electron microprobe are summarized in table 1. A number
of minerals reported from Nambija were not observed in
the samples examined for this study, including scheelite,
arsenopyrite, rhodochrosite, and brochantite.
SYENITE PORPHYRY
The predominant rock type in the Zamora batholith
is a quartz-plagioclase-K-feldspar-hornblende tonalite
 MINERALOGY AND CHEMISTRY OF SKARN, ZAMORA PROVINCE, ECUADOR
EXPLANATION
Syenite(?) porphyry
Volcanic rocks
[ ] Breccia
[ ] Garnet skarn
[T] Chlorite skarn
Sample location (NAM-)
KILOMETERS
Figure 2. Generalized geologic map of the Nambija area show
ing sample locations. Modified from McKelvey (1991).
(Salazar, 1988) or granodiorite. Granite occurs locally.
McKelvey (1991) described sample NAM-13 as a syen
ite porphyry, based on the absence of quartz, the abun
dance of K-feldspar phenocrysts, and the presence of
relict mafic silicate minerals. The relation of this rock,
and the two syenite porphyry dikes (fig. 2) that are
associated with the main mineralized zone, to the main
body of the batholith is unknown. The syenite porphyry
is extensively hydrothermally altered throughout the
mapped area. In hand sample, syenite-porphyry sample
NAM-13 is a medium-grained, even-textured rock com
posed of chalky white 12-mm-long phenocrysts set in a
buff-colored groundmass. Rusty clots mark the locus of
oxidized, hydrated iron-oxide minerals. Phenocrysts are
plagioclase (now pure albite) or orthoclase, clouded by
alteration to sericite and clays. Shreddy, brown biotite is
present in the groundmass. No calc-silicate minerals or
sulfides are present.
TUFF
Sample NAM-10 is a distinctive black rock,
described as a tuff (McKelvey, 1991), from an outcrop in
the garnet skarn zone on figure 1. Based on thin-section
study, the rock is composed of a dense swarm of plagio
clase laths (0.5 mm wide by 2 mm long) altered to clays.
109
Irregular patches of chlorite, typically 2 mm across,
enclose blocky, 0.55 mm and smaller, disseminated mag
netite grains that are partly oxidized to hematite; chlorite
also fills in between plagioclase laths. These chlorite
patches replace most of the igneous mafic silicate minerals
in the rock; some biotite (table 2, analysis 9) is preserved.
Epidote is the only calc-silicate mineral in this rock; pyrite
and chalcopyrite are observed in chlorite. This rock repre
sents endoskarn formed in an igneous protolith that was
more mafic than the syenite porphyry.
BRECCIA AND SKARN
McKelvey (1991) noted that mineralization at Nam
bija appears to be concentrated in a linear, structurally
controlled zone of multi-clastic intrusion and tectonic
breccias that crosscut skarn assemblages. The distinction
between breccia and skarn is based on outcrop texture; in
thin section the distinction is less clear. Breccia samples
(table 1) all contain altered feldspar phenocrysts, igneous
clasts having porphyritic textures, and fine-grained volca
nic(?) clasts. The groundmass is replaced by skarn min
erals to varying degrees. Garnet partially replaces altered
feldspar phenocrysts and groundmass in breccia. Epidote
is locally developed in a few samples, and fibrous actino
lite is present in veins, in groundmass, and as pseudomor
phs that replace phenocrysts. Skarn samples NAM-3, 8,
12, and 14 are predominantly igneous rocks partly veined
by (or locally altered to) skarn, whereas samples NAM-7,
9, and 15 are massive garnetites. In the few samples that
contain both pyroxene and garnet, garnet is coarser, more
abundant, and appears to be paragenetically later than
pyroxene. Green and brownish chlorite pods, interstitial
to massive garnet, are cut by blebby veinlets (fig. 3A) that
are opaque in transmitted light. These blebs are fine
grained mixtures of quartz and iron oxides and form
chains that appear as veinlets, or they line the walls of late
calcite veins. In some samples, a brown, biotite-like
material forms a spidery network of veinlets and patches
throughout the rock. Microprobe analyses show that,
although this material can be chemically homogeneous on
the scale of tens of microns, it generally is heterogeneous
across a single grain or along a veinlet. This material
probably represents a mixture of several mineral phases,
but in some cases, it has a composition intermediate
between a biotite and a chloritei.e., low totals that sug
gest a hydrous mineral, significant but variable K2O con
tent (0.5 to 5 weight percent), and a highly variable silica
COIntent.
K-feldspar is ubiquitous in breccia and in skarn. In
some samples, K-feldspar grew along or selectively
replaced zones in garnet, imparting a bullseye texture to
the garnet. K-feldspar forms monomineralic veinlets, or
veinlets cored by calcite, and some K-feldspar appears to
E

heden
%Hd,
andradite;
percent
mole
%'Ad,
observed;
not
---,
amount
trace
Tr,
present;
mineral
X,
numbers.
sample
are
porphyry)
tuff,
breccia,
(skarn,
headings
rock-type
under
(Numbers Porphyry
Tuff
Breccia
Skarn

NAM-13
NMineral
]
89TT
NAM-11
NAM-5
NAM-l
NAM-15
NAM-14
NAM-12
NAM-9
NAM-8
NAM-7
NAM-3
AM-10 K-feldspar. . . . . .
X -
----
--
---
--
-- X
Chlorite. . . . . . .
---
--
---
--- Magneti e. . . . . .
X ---
--
---
---
-
-
--
-
---

Plagioclase. . . . .
X ---
--- X
Bioti e. . . . . . . .
---
---
---
--
--- Bioti e/chlorite. . . .
X -
---
---
--- Quartz. . . . . . . .
X --
---
-- TPyrite. . r. . . .
X ---
---
---
---

X
.........
Electrum
-
---
-
--
--
---
--- (%P6-36
Epidote
s). . . . . . .
22430
X ---
---
-
--- X
Calcite. . . . . . .
---
---
---
--- Sphen . . . . . . .
X ---
----
--
---
-
---
--

.X
Amphibole
.......
---
--
---
--
---
--
---
---

aiscreompbrloabge.s
mineral
of
Summary
1.
mTable
electron
by
studied
were
that
Nambija
from
samples
13
in
present

21-63
98
34-55,
(%Ad). . . . .
Garnet
X
58-65
3272
44-69
33-58
2945
4862--
--
---
---

Apati e. . . . . . .
X ---
---
---
--
-
--- Chalcopyrite. . . . .
X --
---
--
---
---

pOther
s!
galena,
zircon,
includes
barite.
and
halerite,

pistacite]
percent
mole
Ps,
k
bergite,
.X
Other!
.......
---
---
---
-
---
---

(%Hd). . .
Pyroxene
16-17
22
28
2134
24 ---
-
--
----
--
 MINERALOGY AND CHEMISTRY OF SKARN, ZAMORA PROVINCE, ECUADOR 111

Table 2. Microprobe data for chlorite and biotite.

[Total iron reported as FeC), water computed assuming full hydroxyl site occupancy, -, no data. Explanation of analysis num
bers:" 1 and 2, sample NAM-1, green chlorite, 2 different grains; 3, sample NAM-7, average of 5 points taken along a 0.8-mm
long clot of pale-green chlorite in rosettes, interstitial to gamet;4, sample NAM-7, brown chlorite, average of 2; 5, sample NAM
8, green chlorite, average of 2; 6, sample NAM-9, average of 3 points on pale, brownish-green chlorite in K-feldspar, interstitial
to garnet; 7 and 8, sample NAM-10, averages of 3 points on each of 2 grains of green chlorite adjacent to epidote;9, sample NAM
10, average of 4 points from a coarse-grained (1 mm across) diffuse patch of biotite (patches are mixtures of biotite- and green
chlorite-rich areas; they are riddled with magnetite inclusions and appear to represent alteration of relict mafic silicate minerals in
the volcanic protolith)]

Analysis number
l 2 3 4 5 6 7 8 9

(Results shown in weight percent)

SiO2............ 29.26 31.64 28.95 32.62 28. 19 33.33 26.96 26.25 34.76
TiO2............... - - - 0.03 0.02 0.00 0.04 0.06 1.14

Al2O3 ......... 17.05 15.58 17.43 16.81 17.67 15.01 2008 20.73 17.70
FeC)............. 14.84 14.27 17.26 15, 19 14.99 13.49 17.54 17.41 15.66
MnO.............. 1.23 1.16 3.23 1.07 3.61 ().71 (),63 ().7] 0.41
MgO............ 23.31 22.59 19.71 21.58 21.38 24.20 21.07 21.75 14.23
CaO............... 0.25 (),66 (). 14 ().28 0.05 0.41 ().06 0.03 0.01

Na2O.............. - () ()] 0.02 0.02 0.02 0.06 ().01 ().00 0.09

K2O.............. ().03 (). 10 (). 14 () 47 () ()] (). 17 0.61 0.08 9.63
H2O............ 12.02 12.1() | 1.74 12,03 11.81 12.06 11.85 11.87 3.95
F - - ().09 (). 12 ().()5 0.25 - ().00 0.08
Cl.. - - - - - 0.01 0.01 ().00 0.05
=F................ - - ().()4 ().05 ().02 (). 11 - 0.00 0.03
O=Cl.............. - 0.00 ().00 ().00 0.01
Total......... 98.0 98.1 98.7 I ()().() 98.5 99.8 98.9 98.9 97.6
(Results shown as number of cations14 O equivalents for chlorite; 22 O equivalents for biotite)
Si IV.............. 2.99 3.21 2.99 3.23 2.91 3.29 2.76 2.68 5.33
Al IV............. | ()| ().79 | ()1 0.77 1.09 ().7] 1.24 1.32 2.67
T site............. 4.00 4.()() 4.()() 4.00 4.00 4.00 4.00 4.00 8.00
Al VI............. 1.04 1.07 1.11 1, 19 1.06 1.04 1.19 1.18 0.53
Ti... - - - (), ()() 0.00 0.00 ().00 0.00 (). 13
Fe2+............... 1.27 1.21 1.49 1.26 1.35 1.11 1.50 1.49 2.01
Mn................ (). 11 (). 1() ().28 ().09 ().32 0.06 ().05 0.06 0.05
Mg................ 3.55 3.41 3.04 3.19 3.29 3.56 3.22 3.32 3.25
O site!........... 5.96 5.79 5.92 5.72 6.01 5,77 5.97 6.05 5.97
Ca....... ().()3 ().07 ().02 ().03 ().()] ().04 () ()] () ()() ().00
Na.................. - () ()() () ()() () ()() () ()() 0.01 ().00 ().00 ().03
K.. () ()() () () | ().02 0.06 () ()() 0.02 ().08 ().01 1.88
OH ................ 8.00 8.00 7.97 7.96 7.99 7.92 8.00 8.00 3.95
F - - ().03 0.04 ().()] 0.08 - () ()() 0.04
Cl... - ().00 0.00 () ()() ().01
Mg/(Mg+Fe). 0.74 0.73 ().67 0.72 ().71 ().76 (),68 0.69 0.62
Fe/(Fe+Mg)... 0.26 ().27 ().33 ().28 ().29 0.24 ().32 0.31 0.38
T ("C)2...... 26] 192 264 186 288 165 337 362 -

1 Octahedral site sum exclusive of Ti for chlorite, includes Ti for biotite.

2 Temperature estimates computed from Jowett's (1991) equation for chlorite
geothemometry, T (C)=319(AllW) 69, where AllW =All V40.1(Fe/Fe+Mg).

have grown unimpeded into open space before infilling by
calcite (fig. 3B). Calcite veinlets, generally on the order of
0.2 mm in width, crosscut all other assemblages and
appear to record the latest mineral deposition within these
rocks.
Pyrite is sparsely disseminated throughout most of the
samples; however, only one sample, NAM-3, can be
described as sulfide rich where pyrite forms a network of
massive and mottled grains that enclose rounded crystals of
garnet and other silicates. This mottling results from
incomplete replacement of preexisting silicate minerals.
Chalcopyrite is the second most abundant sulfide present,
and micron-size blebs of galena, pyrrhotite, and barite were
identified in pyrite by scanning electron miscroscopy.
112 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Figure 3. Photomicrographs showing skarn mineral textures in
sample NAM-9, plane polarized light, field of view 1 mm across.
A, Blebby, opaque quartz vein (Q) cutting chlorite (chl) interstitial
to massive garnet (grt). B, Potassium feldspar (kfs) and calcite (cal)
vein in garnet (grt).
Pyrite grains are brecciated and veined by late silicate min
erals and calcite.
GOLD MINERALIZATION
Salazar (1988) described six types of gold occur
rences at Nambija: (1) an interstitial, late phase between
garnet crystals in assemblages with quartz, K-feldspar, cal
cite, and plagioclase; (2) with disseminated pyrite in calc
silicate minerals and volcanic breccias; (3) on K-feldspar
grain boundaries; (4) in siliceous sedimentary rocks; (5) in
quartz veins and veinlets; and (6) in fractures in shear
zones. He concluded that gold was deposited after garnet
and perhaps after K-feldspar and probably had a pro
tracted history of deposition and remobilization that
spanned skarn alteration, brecciation, and quartz veining in
the district. Geochemical data for 15 samples representing
various lithologies at Nambija show the highest gold val
ues (as much as 47 ppm) are associated with quartz in
garnet-rich skarn (table 3).
Macroscopic free gold is present as electrum in
quartz within garnet-skarn sample NAM-15 and in sample
89TT1, a brecciated skarn from a locality just north of the
area shown in figure 1. Both samples contain abundant
garnet and neither show evidence for extensive retrograde
alteration or sulfide deposition. Relatively rare, two-phase
(liquid + vapor) fluid inclusions are present in quartz sur
rounding electrum and more commonly in calcite through
out the sample suite, but the fluid inclusions in the
samples available are too small (< 10 microns) and too
sparse for conventional geothermometry.
Electrum forms discrete grains and parts of discontin
uous veinlets up to 0.6 mm long (fig. 4). Within each sam
ple, electrum is compositionally homogeneous, but
electrum in the breccia is enriched in gold (Aug2 Ags) rel
ative to electrum in garnet skarn (Au?? Ag23). Given the
limited number of samples available for study, it is impos
sible to determine how typical these values are for the
entire district; the electrum compositions may reflect sec
ondary gold enrichment through remobilization of pre
cious metals, late in the life of the hydrothermal system.
MINERAL CHEMISTRY
GARNET
Most of the garnet at Nambija is yellowish green to
green in hand sample. In thin section, the garnet is color
less or locally yellow in plane polarized light. Garnet var
ies in texture from patchy, to veins, to massive and is
sector twinned, mostly anisotropic, and shows delicate,
oscillatory growth zoning. Microprobe data for garnet
were recalculated in terms of ideal garnet stoichiometry (8
cations per 12 oxygens and assuming no hydrogarnet com
ponent) for the estimation of ferric and ferrous iron con
tents and the calculation of molecular proportions of end
members. Most Nambija garnets are intermediate in com
position between grossular and andradite, and the sums of
the pyrope, almandine, and spessartine components are
less than or equal to 10 mole percent (fig. 5A). Garnet
rims tend to be cleaner (inclusion-free) and enriched in
iron (i.e., normally zoned) relative to cores, despite fine
scale oscillations in composition (see McKelvey and Ham
marstrom, 1991, fig. 1). In gold-bearing garnet skarn sam
ple NAM-15, however, cores, zones, and irregular patches
of yellow, isotropic, nearly pure andradite are locally
developed. Andradite is rimmed by garnet of intermediate
composition, typically in coarse-grained pods. Hammar
strom (in Theodore and others, 1992) noted similar
 E

analyzed
Flonaredstcumytca-rteoalistevcomelpinyct
pi2
scoupled
emission
Sb,
Pb,
Mo,
Cu,
Cd,
Bi,
As,
Ag
for
analysis
partial
with
1983)
Crock
1987,
others,
and
(Lichte
package
analytical
NAM-7
NAM-6
NAM-5
NAM-4
NAM-3
NAM-2
NAM-1
NAM-15
NAM-14
NAM-13
NAM-12
NAM-11
NAM-10
NAM-9
NAM-8 TpbAesrtrace
d|
(elements
methods
following
by
Hg,
1990).
visible
W,
method,
vapor
ascold
furnace
graphite
Au,
atomic
Te, cobetropgmstaicosnpoteny,d
0.35
0.15
0.10
2.80
0.30
Te
0.05
0.50
N0.05
<0.05
E|||
BfCCCIdTuff
Porphyry
Skarn
Breccia

3
Yb
1
2
4
l

Ag
2.9
0.30
0.051
0.55
0.25
0.58
0.28
N0.045
1.1

0.5
1.0
W
6.0
N0.5
11.0
2.0
1.0

0.002
0.90
NO.002
7.30
0.024
1.20
0.008
<0.002
0.012
0.028
0.034
47.0
1.65
Au
million)
per
parts
(in
analysis"
element
Trace spectroscopy?
Optical
million)
per
parts
(in
spectroscopy?
Optical
percent)
weight
(in

eadings
tuff,
skam,
(breccia,
copnecenitfriateido]n
the
at
detected
not
N,
numbers.
sample
are
sporphyry)

Be
1
<
2
<<l
<1
1l

Hg
0.18
0.20
0.12
0.04
0.02
0.36
0.08
N0.02
0.28

1pm0
<
Bi
1988).
(Motooka,
pZn
2
&
Cd
and
ppm
analyst.
Fey,
D.L.
samples.
all
for

1.6
12
N0.60
2.5
8.6
4.1
3.6
9.4
7.3
48
20
0.68
As

absorption.
analysts.
Roumer,
T.A.
and
Roushey,
B.H.
Hageman,
P.L.
forsamples
"data
"ock
Nambija.
from

head;
WPe

halytic
-

errock-l

Table
3 A
und
(Numbers

1.
-

14
8.
Al
12
6.
9.53
5.56
7.94
9.95
8.65
8.04
10.0
8.26
6.84 15.3
10.7
5.93
5.43
10.3 8.57
Ca 3.65
6.89
0.49
1.93
10.1
1.74
5.60
21.0
1.29
12
2.
0.61
11.0
0.11 6.68
12.3
3.47
3.26
Fe
7.18
2.43
1.78
6.03
4.32
7.91
3.73
7.40
14
1.
9.74 Mg
0.56
2.47
0.45 2.31
2.79
0.48
3.62
3.18
0.27
0.62
4.85
0.70
0.91
0.10
0.06
0.49 0.85
4.40
3.56
0.02
0.23
2.92
Na
0.89
4.88
3.17
4.78
4.66
3.33
3.54
0.04 Ti
0.34
0.49
0.31
0.53
0.62
0.21
0.28
0.54
0.42
().38
0.26
0.35
0.32 2,070
1,880
57
602
625
120
853
772
Ba
103
196
200
16
685
597
14 17
18
56
31
<48 18
8
10
Ce
227
7
14
15
11
63
21 118
6
Co
7
29
18
2
4
31
708 33
122
329
36
34
77
38
15
13
10
3Cr
475 203
54
2
|Cu
1
|
9
6 | 23
250
117
239
745
662
50
46 18 40
20
24
32
26
17
19
27
16
Ga 914384 Li
La
5
210
7
48
13
6
1
| 5
6
764176 5,910
310
17
12
8
11
4 8,320
4,480
4,320
2,630
6,770
2,660
4,440
7,550
8,630
2,570
4,860
1,130
135
Mn
1,350 <2
<Mo
3
8
42 5
Nb
8
<6
<4
4 33
9
Nd
17 427 Ni
1<14
5
11
221
8
35
12
16 <2
421
9
27
5
15
1101420 13
1<4
5
Pb
16
14
12
73
8
4
24
22
8 8
<5
Sb
<
5
15
7
9 0 16
12
18
Sc
19
20
9
323
8
30
100 45
133 0 7
1178
1
92
261
4
31
65
Sr
166
61
509
195
39
18
62 <Th04
6
19
5
<4
4 13
35
15
19
2Zn
8
171
83
26
239
125
223
69
36
10
33.32
0.75
6.77
3.48
1.33
0.87
6.64
1.23
13
().
3.49
K
5.79
1.79
0.14
0.80
.97 00.58
P
0.03
0.04
0.14
0.07
0.06
0.09
0.11
0.41
.06 147
1154
138
133
123
80
200
97
V
84
172
59
239
302
298
28 28
26
1Y7
6
10
21
15
22
38
40
16
11
114 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Py + Alm + Sp

o NAM 1
NAM 3
NAM 7
NAM8
NAM9
NAM 12
NAM 15
89|| 1

Gr Ad
Figure 4. Photomicrograph showing coarse electrum (E) in
quartz (Q) with zoned garnet (grt). Sample NAM-15, plane polar A
ized light, field of view 2 mm across.
2.00
n lt n 1 O 32
occurrences of multiple generations of garnet, including a s lony- 5 (0.26 o

a ! Iqt 7 O 28
break in composition between isotropic and anisotropic | Jorn 8 0 21
* ! or Q () 4 o $3
varieties, in skarn associated with gold mineralization in I lm 12 5 * n

the Battle Mountain mining district of north-central

Nevada. Meinert (1989) stated that pure andradite is rela
#
Q1)
1.50 | : ," " " - -
. . .
O o o c
k- ~ on
tively uncommon in gold skarns compared to intermediate Q1) * - c, 0.
o

*a

grossular-andradite compositions. In five of the Nevada C- n s "

deposits, however, a stage of nearly pure andradite garnet + a an
C) 1.00 - - -o

growth appears to correlate with gold mineralization in

skarn (Theodore and others, 1991, fig. 10) and may repre o . . ;
sent a stage of iron influx or remobilization. This influx of .9-
-
...'.
o -

iron could be related to sulfide breakdown during late ret

rograde skarn stages within the hydrothermal system. It is 9N 0.50 " .. ..
Q - o o a * *
conceivable that gold, originally deposited with sulfide H. . . . ."
minerals, may have been remobilized and concentrated as * ' ' * -

part of such a process. Therefore, samples that only con
tain garnet populations of a uniform composition may be
less prospective than samples with more complex garnet
populations.
TiO2 contents of the pure andradites are negligible.
Garnets in the two breccia samples (open symbols in fig
ure 5B) are more titaniferous than those in massive skarn,
which is typical of skarn developed in an igneous rather
than a carbonate-rock protolith. Titanium content of garnet
does not reflect the whole-rock titanium concentrations
(table 3); the least titaniferous garnets are in pyroxene
and sphene-bearing samples.
PYROXENE
Pyroxene is much less common than garnet in the
samples studied. Where present, pyroxene is finer grained
and less abundant than coexisting garnet. The pyroxene is
0.00 l-i--------, 1 * * *a fl-* . . . . . ------------
O 20 40 60 80 1 OO
Andradite, in mole percent
B
Figure 5. A, Ternary diagram showing Nambija garnet composi
tions plotted in terms of pyrope (Py) + almandine (Alm) + spessar
tine (Sp) components, grossular (Gr) component, and andradite
(Ad) component. B. Variation of TiO2 contents of Nambija garnets
as a function of andradite content. Numbers in parentheses refer to
whole-rock Ti content reported by McKelvey (1991).
diopsidic (fig. 6), with a persistent, but relatively minor,
manganese component (MnO ranges from 1.5 to 3.0
weight percent). Titanium contents are negligible, and alu
mina varies from below detection limits to 0.5 weight per
cent Al2O3. Pyroxene compositions for Nambija samples
overlap pyroxenes from other mineralized skarns but lack
the enhanced Al2O3 and TiO2 contents noted by Ettlinger
 MINERALOGY AND CHEMISTRY OF SKARN, ZAMORA PROVINCE, ECUADOR 115

5.0
intergrown with chlorite or chlorite replacing a preexisting
/ *
, ITT N- N n=22
layer silicate may invalidate the use of the method. Chlo
H. Y , Gold and silver rite geothermometry was used in a study of a massive sul
f | s/ fide deposit by Krandidiotis and MacLean (1987), and
Jowett (1991) reevaluated their conclusion that a correc
Lll ~- | \ tion is needed to account for variations in chlorite Fe/Mg
h- S ] \ ratios. Temperatures computed for Nambija chlorites
H. X
5 3.0 r ||N \ (table 2) from Jowett's (1991) equation range from 165C
to 362C. Relatively high concentrations of CaO and K2O
uj |||}| \ Base and
in analyses 2, 4, 6, and 7 (table 2) suggest that intergrown
> | \ ferrous
2: 2.0 H. | \ metal
smectite or mica may be present. Temperatures computed
for analyses containing these impurities are low (< 200C
for analyses 2, 4, and 6; table 2) relative to temperatures
+ 1.0 H. | N-l-l-l-l- / computed for other analyses. For two analyses from tuff
sample NAM-10, a lower temperature is obtained for K
g f y - |/ bearing analysis 7 (337C; table 2) than for K-free analy
sis 8 (362C; table 2); both temperatures are high, beyond
the range of 150325C for which Jowett's (1991) equa
O 10 20 30
tion is applicable, and textures suggest that chlorite
Feo, IN WEIGHT PERCENT replaces biotite or other mafic silicate minerals in this
EXPLANATION rock. Similarly, the temperature computed for brown, K
D Sample 3 x Sample 12 and Ca-bearing chlorite in skarn sample NAM-7 (table 2,
Sample 8 e Sample 14 analysis 4) is lower than the temperature computed for
purer green chlorite from the same rock. The Nambija
Figure 6. Pyroxene compositions, plotted in terms of weight per chlorites are much more manganiferous than the chlorites
cent Al2O3 + TiO2 vs Fe0. Fields for pyroxenes associated with from geothermal systems on which the geothermometers
gold and silver skarns and base- and ferrous-metal skarns from Et are based, and the assumption of low ferric iron content
tlinger and Ray (1989). may be invalid. Chlorite in skarn samples such as the
interstitial rosettes, represented by analysis 3 and analysis
5, which yield temperatures of 264C and 288C, respec
tively, are the most likely chlorites to record meaningful
and Ray (1989) for pyroxenes from precious-metal
enriched skarns elsewhere.
temperatures in the data set because textures are compati
ble with growth from a fluid, and Ca and K contents are
negligible. Although it would be presumptuous to attach
any significance to the individual absolute temperatures,
CHLORITE
the range of temperatures for chlorite overlaps the temper
Chlorite is a relatively minor mineral in most of the ature range (220320C) of fluid inclusions in chlorite
calc-silicate-bearing samples studied, despite the extensive dominant skarn at the Tomboy-Minnie gold skarn deposit
(Theodore and others, 1986).
area mapped as chlorite skarn on figure 2. Chlorite clearly
postdates garnet and tends to form clots of pale-green to Johnson (1991) applied chlorite geothermometry to
brownish rosettes interstitial to garnet in samples of skarn. hydrothermally altered sedimentary rocks, including skarn,
Chlorite compositions (table 2) are all relatively magne in the New World gold-copper district, Montana, to exam
sian (Mg/(Mg+Fe) = 0.68 to 0.76). Cathelineau and Nieva ine district-wide temperature gradients and fluid-flow
(1985) described a chlorite solid-solution geothermometer paths. His study, with caveats on the application of such
based on an empirical correlation between observed authi methods to different hydrothermal systems, showed a gen
genic chlorite compositions and temperatures measured in eral pattern of temperature decrease away from postulated
active geothermal systems (e.g., Salton Sea geothermal fluid conduits such as sills and dikes, fault zones, and
field, McDowell and Elders, 1980). Cathelineau (1988) lithologic contacts. Most of the calculated temperatures
refined this geothermometer and suggested that the fell in the range 226321C. Nambija may be an excel
method could be applied with caution to hydrothermal and lent mining district for similar studies, especially if cou
metamorphic chlorites because the temperature-sensitive pled with fluid-inclusion work. Unfortunately, the
parameter, the tetrahedral aluminum content (AllW), available data set for Nambija is as yet too limited in dis
appears to be independent of lithology or fluid composi tribution and number of samples to rigorously evaluate
tion. However, he noted that the presence of smectite trends in chlorite temperatures.
116 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

OTHER MINERALS 4.
| --~~
I

- | ----. Fe
Feldspars in all rock types are nearly pure albite or f
st I
. . . CU
Cu * -----
orthoclase. Intermediate plagioclase compositions (Ab2) t | - e , .. .. . t

are only observed in phenocryst cores in altered tuff sam O) f | Au r t Ag-Au

ple NAM-10; rims are albite. Partial analyses of apatite <
5 2EF |
I
--------'

indicate a fluorine-rich variety, containing less than 0.2 O E e |

O) 1; t - - - - - - - - - - - - - - - - - - - - - - - - - - -

weight percent chlorine. Barite, micron-size blebs of o

galena, pyrrhotite, sphalerite, and an unidentified telluride Explan ATION

mineral were detected in coarse-grained pyrite in sample O

O
e Skarn
A Breccia
NAM-3. Calcite fluoresces blue or red under the micro
- Porphyry
probe beam and appears to be relatively pure based on
energy dispersive analyzer (EDS) spectra.
log Cu/Au

Figure 7. Metal ratios for Nambija rock samples. Fields for gold,
GEOCHEMISTRY silver-gold, copper, and iron skarns from Ettlinger and Ray (1989).

Analytical data (table 3) for samples representative of

the various lithologies at Nambija show that gold contents
range from below detection limits (0.002 ppm) to 47 ppm
and that the highest values are present in skarn. AuAg
ratios are greater than 1 for most samples. No particular
elemental correlations are apparent in the overall data set,
nor does the spatial distribution of elemental abundances
indicate any zoning pattern. Tellurium ranges from below
detection levels (0.05 ppm) to 2.80 ppm. Bismuth was
sought, but not detected, at the 10 ppm level. Arsenic is
present in most samples (< 48 ppm), and both tungsten
and arsenic are relatively enriched in skarn sample NAM
8. Copper ranges from 2 to 745 ppm; molybdenum values
are all 8 ppm, and most values are below detection lim
its (< 2 ppm). Fe/Mg ratios are greater than 1 for all sam
ples; Na/K ratios are highly variable.
In their discussion of problems in classifying pre
cious-metal-enriched skarns, Ettlinger and Ray (1989) note
that such skarns are commonly present in geological set
tings that host iron and copper skarns and copper porphyry
deposits. They proposed a skarn classification diagram
using Cu/Au and Cu/Ag ratios that could be used with
assay data as well as production data. Ratios for the Nam
bija data set are plotted on such a diagram in figure 7.
Points are plotted for the eight samples for which gold,
silver, and copper are all present above detection levels.
Two of the Nambija skarn samples plot outside of the
fields defined by Ettlinger and Ray (1989) on the basis of
a worldwide skarn data set; these extremely high gold val
ues undoubtedly reflect nugget effects in sampling. Three
Nambija skarn samples plot in or near the gold skarn field;
one plots in the silver-gold skarn field; and none plot in
the copper or iron skarn fields. The scatter in the data
points reflects the heterogeneity of the Nambija deposit
and the unusually gold-rich nature of the mineralization.
Neither the arsenic, tellurium, and (or) bismuth signature
that is commonly associated with gold skarn deposits (as
well as with other intrusion-related gold deposits (Sillitoe,
1991)) nor the antimony-arsenic-mercury chemical signa
ture that is typical of some epithermal gold deposits are
present in the samples studied here.
DISCUSSION
Meinert (1989), Theodore and others (1991), and
Ettlinger and Ray (1989) compiled characteristics of a
number of gold-bearing skarn deposits and summarized
features that may be useful in exploration (table 4). Distinc
tive mineralogic characteristics of gold skarns noted by
Meinert (1989) include garnet of intermediate grossular
andradite composition, high modal pyroxene/garnet ratios,
relatively iron-rich (> H50) pyroxene compositions,
enhanced aluminum contents of pyroxene, and sulfide
assemblages dominated by arsenopyrite, pyrrhotite, and
marcasite. The presence of these minerals reflects the over
all reducing environment that characterizes distal, gold-rich
skarns (e.g., Fortitude, Nevada) that form the most attrac
tive targets for exploration. Many proximal, copper-rich
skarns commonly associated with porphyry copper systems
also have significant quantities of gold, but these are char
acterized by a more oxidized, garnet-rich prograde assem
blage and a dominant pyrite-chalcopyrite sulfide
assemblage (Theodore and others, 1991; Meinert, 1989).
The place of gold deposition in the complex spatial and
temporal evolution of most skarn deposits from initial ther
mal metamorphism through prograde to retrograde metaso
matism is not well understood. In many such deposits, gold
appears to be deposited with sulfides late in the retrograde
stages that develop best along structures permeable to
hydrothermal fluids. In other cases, however, gold is asso
ciated with either paragenetically late quartz in skarn where
prograde minerals such as garnet and pyroxene are
 MINERALOGY AND CHEMISTRY OF SKARN, ZAMORA PROVINCE, ECUADOR
preserved or with a late quartz-pyrite overprint. The avail
able data for the Nambija deposits (McKelvey, 1991;
Salazar, 1988; this study) show that Nambija exhibits some
of the characteristics of a relatively oxidized gold skarn, as
indicated by garnet and pyroxene compositions, the pres
ence of epidote, and the nature of the sulfide minerals
(pyrite > pyrrhotite). More reduced mineral assemblages
may lie at depth, and the gold observed in quartz within
preserved prograde skarn assemblages may represent gold
that was deposited or remobilized late in the history of the
system. The protolith, albeit a diverse assemblage, appears
to be dominated by volcaniclastic material rather than a
relatively pure carbonate protolith. Neither copper, bis
muth, arsenic, nor tellurium appear to be present in anoma
lous concentrations in the reconnaissance sample suite
(table 4), although a Te-bearing mineral has been identified
in pyrite. Gold/silver ratios are high, and copper/gold ratios
for the Nambija samples appear to be extremely low com
pared to most gold skarn systems (McKelvey and Hammar
strom, 1991). Mineralization appears to be concentrated
along major structures; however, the extensive alteration
present in all of the samples studied from the area (notably
alkali metasomatism of intrusive porphyry, igneous clasts
in breccia, and skarn and extensive quartz veining) docu
ment the effects of a large magmatic-hydrothermal system,
perhaps associated with an unidentified, possibly buried,
porphyry (gold-copper) system.
ACKNOWLEDGMENTS
Greg McKelvey initiated this study, provided sam
ples, base maps, and geologic data, and graciously encour
aged me to complete the study when he left the project.
Paul Schruben assisted in map preparation. John Slack and
Ted Theodore provided extremely helpful critical reviews.
REFERENCES CITED
Arbogast, Belinda, F., ed., 1990, Quality assurance manual for the
Branch of Geochemistry, U.S. Geological Survey: U.S. Geo
logical Survey Open-File Report 90668, 184 p.
Barragn, Jorge, 1989, El potencial aurfero del Ecuador: Geomi
met, v. 16, no. 158, p. 8990.
Cathelineau, Michel, 1988, Cation site occupancy in chlorites an
dillites as a function of temperature: Clay Minerals, v. 23,
p. 471485.
Cathelineau, Michel, and Nieva, David, 1985, A chlorite solid
solution geothermometer, the Los Azufres (Mexico) geother
mal system: Contributions to Mineralogy and Petrology, v.
91, p. 235244.
Crock, J.G., Lichte, F.E., and Briggs, P.H., 1983, Determination
of elements in NBS geological reference materials SRM268
obsidian and SRM688 basalt by inductively coupled
117
Table 4. Characteristics of gold skarn deposits.
[Based on Theodore and others, 1991; Meinert, 1989; and Ettlinger and Ray, 1989)
Clastic or volcaniclastic rock component in skarn protolith
Relatively reduced, mafic calcalkaline/subalkalic plutons
Anomalous contents of As, Bi, and (or) Te
Gold enriched in distal skarn
Gold deposited in retrograde skarn, together with sulfides
Dominant skarn minerals:
[Reduced type]
Pyroxene (hedenbergitic, Al rich) > garnet (andradite
grossular)
Narrow, Al-rich zones in pyroxene and garnet
[Oxidized type]
Garnet (andradite-grossular) > pyroxene (diopside rich)
Reverse zoning and multiple generations of garnet
Dominant sulfides:
[Reduced type]
Pyrrhotite, arsenopyrite, marcasite
[Oxidized type]
Chalcopyrite, pyrite
Other gangue minerals:
K-feldspar, biotite, prehnite, idocrase, Scapolite, apatite,
sphene, cuspidine, epidote (replaces garnet), actinolite or
hornblende (replaces pyroxene), chlorite, quartz, calcite
plasmaatomic emission spectroscopy: Geostandards News
letter, v. 7, no. 2, p. 335340.
Einaudi, M.T., Meinert, L.D, and Newberry, R.J., 1981, Skarn de
posits, in Skinner, B.J., ed., Seventy-Fifth Anniversary Vol
ume, 19051980, Economic Geology: New Haven, Conn.,
Economic Geology Publishing Company, p. 317301.
Ettlinger, A.D., and Ray, G.E., 1989, Precious metal enriched
skarns in British Columbia: An overview and geological study:
British Columbia Ministry of Energy, Mines, and Petroleum
Resources Paper 19893, 128 p.
Johnson, T.W., 1991, Geology, hydrothermal alteration, and Cu
Ag-Au skarn and replacement mineralization in the northern
part of the New World district, Park County, Montana: Pull
man, Washington State University, unpub. M.S. thesis, 326 p.
Jowett, E.C., 1991, Fitting iron and magnesium into the hydrother
mal chlorite geothermometer [abs.]: Geological Association of
CanadaMineralogical Association of CanadaSociety of
Economic Geologists, Program with Abstracts, v. 16, p. A62.
Krandidiotis, P., and MacLean, W.H., 1987, Systematics of chlorite
alteration at the Phelps Dodge massive sulfide deposit, Mata
gami, Quebec: Economic Geology, v. 82, p. 18981911.
Lichte, F.E, Golightly, D.W., and Lamothe, P.J., 1987, Inductively
coupled plasma-atomic emission spectrometry: U.S. Geologi
cal Survey Bulletin 1770, p. B1-B10.
Mapa Geolgico Nacional de la Repblica del Ecuador, 1982, scale
1:1,000,000.
McDowell, S.D., and Elders, W.A., 1980, Authigenic layer silicate
minerals in borehole Elmore 1, Salton Sea geothermal field,
118 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
California, USA: Contributions to Mineralogy and Petrology,
v. 74, p. 293310.
McKelvey, G.E., 1991, Interest shown in Nambija gold deposits,
Zamora Province, Ecuador: Mining Engineering, v.43, no. 12,
p. 14121414.
McKelvey, G.E., and Hammarstrom, J.M., 1991, A reconnaissance
study of gold mineralization associated with garnet skarn at
Nambija, Zamora Province, Ecuador [abs.], in Good, E.E.,
Slack, J.F., and Kotra, R.K., eds., USGS Research on Mineral
Resources1991, Program and Abstracts: U.S. Geological
Survey Circular 1062, p.55.
Meinert, L.D., 1989, Gold skarn depositsGeology and explora
tion criteria, in Keays, R.R., Ramsay, W.R.H., and Groves,
D.I., eds., The Geology of Gold Deposits: The Perspective in
1988: Economic Geology Monograph 6, p. 537552.
Mining Magazine, November 1990, Campanilla gold mine, p.
322323.
Mineral Resources Data System (MRDS) [computer database],
1991, U.S. Geological Survey, Branch of Resource Analysis,
Record number TC00085 for the Nambija district, Ecuador.
[Available from USGS, BORA, MS 920, 12201 Sunrise Valley
Drive, Reston, VA 22092].
Motooka, J.M., 1988, An exploration geochemical technique for
the determination of preconcentrated organometallic halides
by ICP-AES: Applied Spectroscopy, v.42, p. 12931296.
Salazar, Edgar, 1988, Nambija: Conacimiento geologico y mineral
ogico hasta la presente: INEMIN, Quito, 11 p.
Sillitoe, R.H., 1991, Intrusion-related gold deposits, in Foster, R.P.,
ed., Gold Metallogeny and Exploration: Glasgow and London,
Blackie, p. 165209.
Theodore, T.G., Blake, D.W., Loucks, T.A., and Johnson, C.A.,
1992, Geology of the Buckingham stockwork molybdenum
deposit and surrounding area, Lander County, Nevada, with a
section on Potassium-argon and 40Ar/39Ar geochronology of
selected plutons in the Buckingham area, by E.H. McKee, and
a section on Economic geology, by T.A. Loucks and C.A.
Johnson, and a section on Supergene copper deposits at Cop
per Basin by D.W. Blake, and a section on Mineral chemistry
of Late Cretaceous and Tertiary skarns, by J.M. Hammarstrom:
U.S. Geological Survey Professional Paper 798D, 307 p.
Theodore, T.G., Howe, S.S., Blake, D.W., and Wotruba, P.R.,
1986, Geochemical and fluid zonation in the skarn environ
ment at the Tomboy-Minnie gold deposits, Lander County, Ne
vada, in Nichols, E.E., ed., Exploration for Ore Deposits of the
North American Cordillera; Selected Papers of the Symposium
of the Association of Exploration Geochemists held in Reno,
Nevada, March 2528, 1984: Journal of Geochemical Explora
tion, v. 25, nos. 1, 2, p. 99128.
Theodore, T.G., Orris, G.J., Hammarstrom, J.M., and Bliss, J.D.,
1991, Gold-bearing skarns: U.S. Geological Survey Bulletin
1930, 61 p.
 CHAPTER L

GEOCHRONOLOGY AND GEOCHEMISTRY

OF THE LADOLAM GOLD DEPOSIT,
LIHIRISLAND, AND GOLD DEPOSITS AND
VOLCANOES OF TABAR AND TATAU,
PAPUA NEW GUINEA

By JAMES J. RYTUBA, EDWIN H. MCKEE, and DENNIS P. COX?

INTRODUCTION 152* 153

I I 1 ----------------

The Tabar-Feni chain of alkaline volcanoes extends 2* -

~ O

for a distance of over 250 km off the east coast of New -- - PNG >
Pigiput deposit w
Ireland, Papua New Guinea (fig. 1). Several gold deposits Simberi TABAR islands - So *
- r
S.

and prospects are present within the volcanic chain, which o \!-------
T I
extends southeastward toward Bougainville Island, the site 3 -> LIHIR 154* 155
of the Panguna porphyry coppergold deposit. The world ISLANDS

class Ladolam deposit, on the island of Lihir, was discov C

ered in 1982 (Davies and Ballantyne, 1987) and has

adolm T TANGA. ISLANDS
deposit
proven reserves of 29.7 million oz (923.7 t) of gold at a
cutoff grade of 1 g/t (0.029 oz per short ton) and a geolog O | FENIISLANDS

ical reserve of 43 million oz (1,330 t) (Getz and Ward,

1990). The deposit formed very recently and still has b] GREEN ISLANDS

active hot-spring and solfatara activity within parts of the

orebody. This paper presents new age and geochemical
data for the Ladolam deposit and for volcanic units and Solomon Sea
gold deposits in the northern part of the Tabar-Feni volca
BOUGAINVILLE
nic chain on the islands of Simberi and Tatau.
1 Uss.
o 50 1oo 150 KILOMETERS

REGIONAL GEOLOGY
Volcanoes within the Tabar-Feni chain have been
regarded to range in age from Miocene to Holocene
based primarily on their morphology and a few isotopic
age determinations. The five K-Ar isotopic ages reported
'U.S. Geological Survey, Mail Stop 901, 345 Middlefield Road,
Menlo Park, CA 94025.
*U.S. Geological Survey, Mail Stop 984, 345 Middlefield Road,
Menlo Park, CA 94025.
Figure 1. Location of Lihir and Tabar Islands and the Ladolam,
Pigiput, Simberi, and Tatau gold deposits and prospects. PNG, Pap
ua New Guinea.
previously for the chain range from 1.5 to 0.18 Ma with
no systematic age progression along the chain (Wallace
and others, 1983). The volcanic chain is characterized
by shoshonitic alkaline volcanism, and petrochemical
studies of volcanic suites from the volcanic chain by
Wallace and others (1983) demonstrate a broad range of
120 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 1. Data for samples collected from islands of Lihir, Simberi, and Tatau, Papua New Guinea.

(Constants used were K/K-1.167x10"mol/mol. Ap=496210'yr, B.B.EO 581x10"/yr.

Sample Locality Sample Mineral K2O *Ariad *Ariad Age

no. type (mol/gx10-12) (percent of total) (Ma)
DDHL53427.7m .....Lihir Vein Biotite 8.72 4.21927 9.01 0.34+0.03
DDHL63-424.8 ........Lihir Porphyry Biotite 9.39 4.894.91 3.52 0.36+0.03
853 1 ........................ Simberi Trachyte Whole rock 1.31 6.63084 43.9 3.5+0.2
8532 ........................ Simberi Trachyte Whole rock 1. 14 6.007] 30.5 3.7+0.2
8533 ........................ Simberi Basalt Whole rock ().65 2, 16529 5.9 2.3+0.4
8534 ........................Simberi Basalt Whole rock 0.44 1.98.842 12.19 3.1+0.3
8537........................Tatau Trachybasalt Whole rock 1, 1] 4.44735 22.1 2.8+0.2
8539 ........................Tatau Trachybasalt Whole rock 3.97 1.1 1036 22.0 1.91+0.1
852.7 ........................ Simberi Vein Alunite ().28 5.2.1741 0.46 1.2+2.0
8528 ........................ Simberi Vein Alunite ().88 3.57.103 5.65 2.8+.6

nepheline-normative volcanic and intrusive rocks. The
geochemical data, including radiogenic isotopic values,
suggest similarities to an intraoceanic arc; however, the
absence of andesitic volcanic rocks and the distinctly
alkaline character of the volcanism makes the origin of
the volcanism problematic (Plimer and others, 1988;
Wallace and others, 1983; and Johnson,
Johnson's (1979) hypothesis that the volcanism results
from partial melting of mantle lithosphere previously
modified by subduction is favored by recent studies.
K-Ar DATING
for one sample (Stacey and others, 1981) and using a
15.2-cm-radius, Neir-type mass spectrometer for five sam
ples. The precision of the data, shown as the t value
(table 1), is the estimated analytical uncertainty at one
standard deviation (Cox and Dalrymple, 1967). It repre
sents uncertainties in the measurement of radiogenic 40Ar
1979). and K2O based on experience with hundreds of replicated
analyses in the Menlo Park laboratories of the U.S. Geo
logical Survey. Mass discrimination of the spectrometer is
routinely determined on the basis of multiple analyses of
purified air. The constants used in age determination are
those recommended by the Subcommission on Geochro
nology (Steiger and Jaeger, 1977).

Six volcanic units and two mineral deposits on the

islands of Simberi and Tatau and two samples from the
Ladolam deposit on the island of Lihir were dated by K
Ar methods. The samples of volcanic units on Simberi and
Tatau are representative of the volcanic edifices and also
provide maximum-age constraints for the gold deposits
and prospects on the two islands. Alunite samples from
two gold deposits on Simberi directly date the mineraliza
tion (table 1). The two samples from the Ladolam deposit
include biotite phenocrysts from the porphyry intrusion
that is genetically related to the gold deposit and hydro
thermal biotite from the mineralized stockwork part of the
deposit (table 1).
METHODS
K-Ar analyses were performed using the standard iso
tope-dilution techniques similar to those described by Dal
rymple and Lanphere (1969). Mineral concentrates were
made by heavy-liquid, magnetic, electrostatic, and hand
picking procedures. Potassium analyses were by lithium
metaborate-flux-fusion flame photometry, the lithium serv
ing as an internal standard (Ingamells, 1970). Argon anal
yses were done using a five-collector mass spectrometer
GEOLOGY OF THE
LADOLAM DEPOSIT
The three stratovolcanoes that comprise the island of
Lihir are Pleistocene in age and rest on a basement of
Pliocene volcanic rocks (Plimer and others, 1988). The
Ladolam gold deposit occurs within the Luise caldera, a 6
by-4-km summit caldera developed on a stratovolcano
located on the east side of the island. The Ladolam deposit
consists of several zones of gold mineralization within the
floor of the Luise caldera (Davies and Ballantyne, 1987),
and the deposit contains a geological reserve of 43 million
oz Au (Getz and Ward, 1990).
Gold is hosted in a variety of breccias and is closely
associated with a subvolcanic monzonite porphyry that
also contains low-grade stockwork gold mineralization.
The breccias are a mixture of collapse breccias derived
from the oversteepened caldera walls, primary volcanic
and intrusive breccias, and hydrothermal breccias that
crosscut other breccia types. The highest gold grades are
present within a subhorizontal breccia unit composed of
volcanic and intrusive clasts, originally termed the boil
ing zone (Davies and Ballantyne, 1987). This breccia is
 LADOLAM GOLD DEPOSIT, LIHIR ISLAND, PAPUA NEW GUINEA
_
--
Im. |||||
O 1 2 3 4 5
Figure 2. Core sample of high-grade gold ore from the Lienetz
zone, Ladolam gold deposit. Open-space breccia (locally termed
the boiling zone") consists of monzonite clasts characterized by
biotite alteration. Clasts are coated with fine-grained pyrite. Matrix
has been leached from the rock.
characterized by an open-space texture in which the
matrix has been leached (fig. 2). Clasts rich in secondary
biotite are coated by fine-grained pyrite, giving the breccia
a dark-black color. Gold forms grains that are : 20 pum in
diameter within pyrite in the sulfide zone (Moyle and oth
ers, 1990).
Samples from the Leinetz ore zone, within the cen
tral part of the Ladolam deposit, were examined petro
graphically and with the scanning electron microscope
(SEM). Breccia textures most common within
deposit consist of crackle breccias at depth, grading
upward to clast-supported and finally matrix-supported
breccias in the upper part (fig. 3). In general, rocks that
appear as clasts in the breccias show a gradual transition
in texture upward from fine- to medium-grained
equigranular rocks beneath the boiling zone to
hypabyssal porphyries with microaplitic groundmass and,
121
CALDERA IN HIGH-K
BASALTIC LAVAS ACTIVE HOT SPRING WITH LOCALIZED
ALUNITE-SULFUR-PYRITE SINTER
TEXTURES
ALTERATION
BREccia witH
VOLCANIC CLASTS
OPEN TExTURED
\
| ALUNITED, wancaste
CLAY WITH SPOTTY Au
BREccia MINEPALIZATION
500m
BREccla witH
CLASIS of K BKOTITE-ANHYDRITE
syenite X MAGNETITE-(PYRITE)
| ]/, , V,
| Stockwork
SYENITE ; , \ UNALTERED
MOnzoni TE
- -
Figure 3. Idealized cross section through the Ladolam gold de
posit, Lihir Island, showing distribution of breccia and textures and
zoning of ore and alteration.
finally, to volcanic porphyries with flow-aligned ground
mass feldspars in the upper alunite alteration zones.
Weakly altered intrusive rocks are of two types. The
first type is a monzonite (DDH L63, 425 m) consisting of
12-mm-wide, radiating grains of plagioclase (Ango) and
interstitial K-feldspar. Biotite forms large euhedral grains
that poikilitically enclose plagioclase, and clinopyroxene is
rare. Magnetite is abundant as small rounded grains and in
clusters with biotite. The second type is a syenite (DDH
L44, 344 m) composed of interlocking anhedral grains of
K-feldspar, euhedral clinopyroxene, rare biotite, and abun
dant magnetite.
Grading upward from these weakly altered rocks into
6
the ore zone (fig. 3), biotite alteration becomes prominent,
but, except for anhydrite-calcite veins, stockwork veinlets
typical of most porphyry systems are rare. However, nar
row, widely spaced veinlets of clinopyroxene-magnetite
albite were noted in the syenite. Anhydrite-calcite veins
are very abundant beneath the ore zone. They contain bar
ite and irregularly distributed small grains of sphalerite,
galena, chalcopyrite, and pyrite. Silver telluride was noted
in one sample as a small inclusion in pyrite. A mineral
rich in rare earth elements and phosphorous was detected
by the SEM in several samples but was not found in thin
section. Vapor-rich fluid inclusions are abundant in the
veinlets, and many inclusions contain the daughter miner
als celestite and barite. Chloride minerals (such as halite
and sylvite), characteristic of fluid inclusions in porphyry
systems, were not detected.
In the ore breccia, clasts of monzonite altered to
the biotite and white mica are surrounded by a matrix com
posed of biotite, anhydrite, and pyrite. The anhydrite is
partially replaced by gypsum. Pyrite is abundant and is
partially replaced by chalcopyrite. In addition to these
minerals, PbS, ZnS, HgS, and the rare-earth phosphate
were noted with the SEM. Bismuthian stibnite, similar to
that present in the Paradise Peak deposit, Nevada (John
and others, 1991), is also present.
122 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Samples in which primary minerals were totally
replaced by alunite retained their original texture so that
breccia clasts and igneous rock textures could be
observed. Pyrite is the most abundant sulfide, and marca
site is subordinate in abundance. Pyrite forms inclusion
filled grains surrounded by overgrowths of inclusion-free
pyrite. It also occurs as fine wisps, atolls, and framboids.
Magnetite was observed as inclusions in pyrite. Rarely,
chalcopyrite replaces both pyrite and magnetite. Other
inclusions in pyrite grains were identified by SEM meth
ods as lead sulfide, molybdenum sulfide, and silver-gold
telluride.
AGE AND GEOCHEMISTRY OF THE
LADOLAM DEPOSIT
Biotite from phenocrysts in the fine-grained phase of
the porphyritic monzonite at a depth of 424.8 m below the
surface has an age of 360+30 ka (table 1). Veinlets that
comprise the stockwork in the porphyry are characterized
by an early phase of hydrothermal biotite intergrown with
magnetite and overgrown by later anhydrite. Coarse
grained biotite from a vein containing 0.8 ppm Au at a
depth of 427.7 m has an age of 340+30 ka (table 1).
These ages are the same within analytical uncertainty and
support a close genetic relationship between the porphyry
and gold mineralization. However, the data are permis
sive of mineralization following emplacement of the por
phyry by as much as 80,000 years.
Geochemical analyses of drill-hole and surface sam
ples from the Lienetz zone of the Ladolam deposit (table
2) represent both porphyry and epithermal environments.
Factor analysis was used to define elemental associations.
The R-mode principal components analysis (Davis, 1986),
using an orthogonal transformation solution, defines three
factors with positive loading values of greater than 0.8.
The elements in each factor are listed in order of decreas
ing loading and are summarized in table 3. Factor 1 con
sists of Au, Ba, Cs, As, Ag, and Rb. This elemental suite
reflects the epithermal environment of precious metal min
eralization. Factor 2 consists of W, U, Mo, Cu, Pb, and Cr
and reflects mineralization associated with the porphyry
environment. Factor 3 consists of Fe, Co, V, and Ce and
reflects the original magmatic variation of these elements
in the volcanic and plutonic host rocks.
GEOCHEMISTRY AND AGE OF GOLD
DEPOSITS AND HOST ROCKS,
SIMBERI AND TATAU
Simberi and Tatau form the northern two volcanoes
of the Tabar group and occur at the northernmost part of
the volcanic chain. The four volcanic units dated from
Simberi represent the major geologic units on the island
and span a short interval from 3.5 to 2.3 Ma. Based on the
analytical uncertainty in the sample ages, the volcanic
event could have lasted as little as 0.8 m.y. or as long as
2.0 m.y. (table 1). The alunite ages, 1.2+2.0 and 2.8+0.6
Ma, from the mineralized zones on Simberi are younger
than the volcanic rocks, but because of the large analytical
uncertainty in the ages, alunite could have formed during
mineralization that occurred either during the last stage of
volcanic activity or just after volcanism ended, at about
2.2 Ma.
Two volcanic units on the island of Tatau have ages
of 2.8+0.2 and 1.94:0.1 Ma (table 1). The volcanic units
provide a maximum age for the high-grade gold-quartz
adularia and base-metal veins present on Tatau. Volcanic
activity on Tatau and Simberi overlap in part, and the sim
ilar ages for the end of volcanism suggest that the mineral
izing event on Tatau occurred at the end of volcanic
activity and at about the same time as on Simberi.
The Pigiput deposit on Simberi is hosted by volcanic
breccias, flows, and associated plugs. Although air-photo
analysis suggests circular structures on Simberi, the flat
dips of the pyroclastic units observed in drilling suggest
that the island is an erosional remnant of a tuff sheet
rather than a remnant of a volcanic center (R. Rogerson,
Papua New Guinea Geological Survey, oral commun.,
1986).
Geochemical data of surface samples from the
Pigiput deposit (table 2) were evaluated through correla
tion-coefficient and factor analysis. Gold has a strong cor
relation (> 0.8) with Ag, As, Fe, Mo, Sb, and V. Principal
components factor analysis defines three significant ele
mental associations. The R-mode principal components
analysis, using an orthogonal transformation solution,
defines three factors with positive loading values of
greater than 0.8 (table 3). Factor 1 consists of Fe, Mo, Sb,
Au, As, V, and Ag; this elemental suite reflects precious
metal mineralization. Factor 2 consists of Na, Cs, and Co
and reflects the intense argillic alteration of the volcanic
host rocks. Factor 3 consists of Sm, La, and Ce and
reflects the magmatic fluid enrichment of light rare-earth
elements (LREE) superimposed on the original LREE
variation of these elements in the volcanic rocks (see Dis
cussion section).
DISCUSSION
Geochronologic data presented in this paper and
previously published ages (Wallace and others, 1983)
indicate that the Tabar-Feni volcanic chain is Pliocene to
Holocene in age, and there is no isotopic evidence for
Miocene volcanism. Volcanic units from the Tabar

hornblende,
Hb,
feldspar,
Feld,
Symbols:
Ag.
Pb,
V,
Cu,
for
absorption
atomic
INNA
by
Analysis
percent.
weight
sodium
and
iron
ppb);
(in
gold
except
ppm
in
values
[All
Bio,
breccia;
Bx,
porphyry:
Por,
pyrite;
Py,
Sample
Pb
Na
Mo
La
Fe
Co
Cu
Cr
Cs
Ce
Ba
As
Ag
Au
Location
V
U
Th
Sc
Sm
Sb
Rb
W

H160
Makopena
3Bx
57ydr10o6thermal
910
8535
50
970
16
6,700
19
12
110
0.1
11
6.9
413
6,660
l
0.5
41

ore
stwk
monz
Bio
1L63-424.8
2
220
1,680
Lihir
zone,
2Lienetz
l
42
1,900
989
3
18
20
6.9
23
105
1
6.1
17
54
l 00
0.44

flow
in
ore
Oxide
8
S3
1
93
260
18]
0.2
140
Simberi
0Samat,
535
3.6
10
120
ll
6.8
37
.30
40
145
3.1
2.5
4

Tatau
Makapena,
veins
py-q
Feldpor,
32
1.7
1.6
3.4
8
2
33
1,100
2
M3
100
0.9
298
18
2.1
10
28
7
5
4.l
1.493
.3

Geioncehreamliszterdy
2.
mTable
rocks
plutonic
volcanic
of
Guinea.
New
Papua
Islands,
Tabar
and
Simberi
Tatau,
Lihir,
from
samples

flow
por
Basalt
LC2
Lihir
wall,
W.
caldera
Luise
1.3
28
4.8
2ll
77
38
060
l
23
1
29
5
5
90
0.4
196
0.5
.2
80
]
ore
Sulfide
L54-1945
Lihir
zone,
ore
Lienetz
210
3.6
1.9
16
8.4
8210
6
100
1,460
78
130,000
l
5.2
0.2
246
14
137
162
150
4200

ore
Oxide
L16
Lihir
zone,
ore
2Lienetz
8
72
2.2
2.5
l
110
1,200
60
0.6
504
4
l
10
26
1
5
0.2
84
8.6486 Simberi
Pigiput,
FeC)
and
Mn
Vein,
9
240
5.4
5.7
15
360
540
4,100
0.2
19,020
P3
8l
1
19
90
67
162
60
37
0.1
43
60
6

flow
por
Hb
Lihir
wall,
caldera
Luise
2018
l
133
104
0.5
1.2
34
5.8
0.2
48
LCl
2
3.3
80
90
.6
13
7.7
51
144 tuff
Mafic
Lihir
wall,
W.
caldera
Luise
59LC8
l
40
15
078
2
380
1.1
48
5.1
0.2
100
.2
2.1
4
108
1
|
10
11 Lihir
zone,
Lienetz
Monzonite
L63-568.2
7120
l
32
340
220
6
13
15
588
5
12
4.9
1.8
2
147
0.8
0.9
0.2 40
250
14
1.2
13040 Monzonite
.843 Sulfide
Lihir
zone,
ore
Lienetz
zone
ore
4L71-177
l
5,800
655
5.8
4,050
4,130
120
7.3
27
3,600
10
8.2
235
1,250
0.4 por
Lihir
zone,
ore
64Lienetz
l
30
310
0.7
80
2
122
1.5
1.9
6.9
8074
.4
L87-411
9
0.2
22
27
1
4
6
53
4.4 Tatau
Makapena,
por
Trachyte
54
48
144
2.3
1.5
24
5.8
M2
1.9
22
3.8
15
265
160
l.
40
970
36
0.1
130
10
4 flow
in
ore
Oxide
Simberi
1Samat,
l
72
290
29
5.5
10
1,700
215
5.3
40
8.5
13
01,230
2
2.9
418
0.8
4
8
S420
.2
80 Tabar
Big
flow
mafic
Por
10
8.3
2
275
0.7
0.9
27
4.3
1.3
140
1TBl
l
16
420
24
0.2
46
157
6
4
5 60 ore
grade
high
Vein,
Tabar
Big
3 34
170
2
305
1.5
1.3
20
6.3
l
45
410
746
0.1
11,500
22
TB2
5
6
8
2.2
15
8.80 2.2
7 Simberi
Sorowar,
vein
Alunite
3
2
300
l
68
270
1,200
0.5
2,590
03
16
8.5
31
28
8528
8
10
50
1.3
37 2 2
.2
8 flow
mafic
Por
28531
l
52
170
87
0.1
1,370
Simberi
Manambu,
1Mt.
19
8.7
39
80
7.2
21
20
4
7306
.3
220
0.8
1.3 flow
mafic
Por
Simberi
coast
NE.
Asllawa,
2
172
1.2
1.3
32
6.7
78533
l
54
100
0.1
1330
62
120
1.9
83
21804 Por
7.4
4
5 .9 Simberi
coast
NE.
Asllawa,
flow
mafic
2
8534
3480
l
53
170
34
10.2
3
19
7.3
39
1156
5
35
6.2
1.8
004
.8
188
0.9
1.6

Lihir
wall,
W.
caldera
Luise
LC4
3.9
2
1.2
l
0.1
515
1
019
34
67
0.3
14
0089 Flow
Flow
2
3
10
290
0.6 Lihir
wall,
W.
LC5 1 Basalt
caldera
Luise
1301.1
43
155
21
39.3
28
9.1
0.3
2
250
0.5
6
l.3
10
40
44
5.4 flow
por
Creek
Kapit
LC6
2
30.2
9l
32
1
53082
||
7.8
37
140
205
0.5
1.1
27
4.8
0.2

Simberi
Samat,
ore
vein
1Oxidized
1
220
510
130
44
10
265
0.3
57
2.7
l
1.5
30,000
2 ,800
64590
6.2.
47
S5
221
31
572
Simberi
Pigiput,
Tuff
3
7.7
9.2
1P2
l
O.
0100
2
5,110
390
5
1 .240
4.6
10
191
7
8
1.1
112
406

flow
por
Basalt
LC7
Lihir
wall,
E.
caldera
Luise
2
304
13
35
17
7.6
32
2162
5
20
5.9
5.32
00
30
.9
210
1.3
1.7
0.3 76
200
Lihir
wall,
W.
caldera
Luise
tuff
Mafic
2]
27
31
2
235
1.8
2.6
43
6.6
LCl2
0.2
2
1
4.
050
10
1.3
.3
8

flow
Mafic
2
Simberi
Manambu,
2Mt.
50
1310
3
22
6.6
331
3.9
10
17
1.9
8606
.6
175
1.7
44
0.1
140
70
8532

sotnzcoknwiotrek];
Stwk,
mbiotite;
Monz,
quartz;
Q, Tatau
Makapena,
7Ml
47
1540
14
64
0.5
Lava
2
163
93
3
9 49
20.9
4
19
5.9
24
.2
.6 Simberi
Pigiput,
tuff
Pumice
20
58
4.4
7.5
9.5
1,660
0.1
4,970
6
1
36
100
4
8
8.9
35
1.4
113
10
48
0.2
Pl2
1 Simberi
Pigput,
15
179
3.5
3.7
8.8
1
88
280
1,850
0.1
2,630
6
020
2 5 Por
intrusive
Por
P4
6
310
77
250
17.28 Simberi
Pigiput,
4.7
intrusive
9P54 zone
3
71
100
2,470
0.2
3,070
150
12
0.2
17
26
8.2
15
132
19
11
68
27
260
4.3
prospect
Tiro
0Mt.
lava
Altered
5
255
1.4
18538
4
62
3,100
1,310
26
180
10
140
192
29
5.3
35
150
23
6
8.2
10 2
1.9 prospect
Tiro
Mt.
lava
Por
45
1670
17
6.3
32
157
230
1.2
1.4
20
6.5
5.7
10
1.1
25
.2
20
8539
71
31
Argillic
Simberi
Samat,
1S1A
8
58
640
0.2
267
5.7
24
3108
392202
.2
168
2.6
1.9
26
8.2
1.3

Makopena
2
flow
mafic
Por
16
228.3
6
215
0.7
18537
1
31
940
53
0.4
51
.1
508
88
4
10
39
60
1.4

type
no.
124 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 3. Array of orthogonal varimax loadings for three factors determined for the Ladolam and
Pigiput deposits.
[Bold numbers indicate elements defining each factor. See text for definition of factors)

Ladolam deposit, Lihir Pigiput deposit, Simberi

Factor 1 Factor 2 Factor 3 Factor 1 Factor 2 Factor 3

Log (Au).............. 0.989 () ()38 () ()()] 0.951 0.074 ().274

Log (Ag).............. 0.848 ().507 (). 103 0.888 0.078 (),033
Log (As)..............0.954 0.262 ().093 0.937 0.108 0.294
Log (Ba).............. 0.989 (),013 0.001 0.313 0.676 0.552
Log (Ce).............. 0.344 (),696 0.233 0.328 0.014 0.940
Log (Cs).............. 0.979 0. 145 0.064 0.123 0.939 0.164
Log (Cr).............. ().088 0.853 (). 171 0.836 0.291 0.108
Log (Cu)............ 0.017 0.956 ().281 0.017 0.243 0. 135
Log (Co)............ 0.053 ().548 0.827 0.201 0.847 0.383
Log (Fe).............. 0.209 ().295 0.883 0.985 0.015 (). 158
Log (La).............. (),652 ().719 ().04() ().499 0.003 0.861
Log (Mo)............. ().012 0.966 ().25() 0.972 0.045 (). 167
Log (Na)............ 0.455 ().551 0.291 0,069 0.966 0.177
Log (Pb).............. 0.277 0.917 ().283 (),099 0.310 0.356
Log (Rb).............. 0.808 ().453 0.21() 0.267 0.475 0.020
Log (Sb).............. 0.572 0.785 ().230 0.949 0.097 0.252
Log (Sm)............. ().480 ().536 0.087 0.239 0.025 0.953
Log (Sc).............. 0.020 0. 176 0.341 ().722 (). 165 0.001
Log (Th)..............0, 175 0.014 0.892 0.241 0.404 0.070
Log (U) ...............0. 130 0.975 (). 182 0.641 0.396 0.439
Log (V) ............... 0.418 ().399 0.804 0.923 ().007 0.248
Log (W)............... (),067 0.982 (). 155 0.248 0.338 0.058

group in the northernmost part of the chain range in age
from about 3.5 to 1.9 Ma. Within the Tabar group, vol
canic units on Simberi and Tabar Island overlap in age.
A previously reported age of 0.99+0.08 Ma (Wallace
and others, 1983), for a quartz trachyte on Tabar Island,
is considerably younger and may reflect an episode of
younger volcanic activity separate from the mid-Pliocene
eVent.
The Ladolam deposit on Lihir Island represents a
unique gold deposit in many aspects. The deposit is the
single largest gold orebody known and contains a geo
logical reserve of 43 million oz (1,330 t) of gold (Getz
and Ward, 1990). The close relationship of the gold
deposit to a subvolcanic porphyry that intruded to very
shallow levels of the volcanic edifice (within 400 m of
the floor of the summit caldera) (fig. 3) demonstrates
the transition from porphyry to hot-spring environment
over a very telescoped interval. The age of the por
phyry, about 360 ka, and the associated stockwork vein
ing, about 340 ka, indicate a genetic relationship
between the porphyry and the gold deposit. Biotite dated
from coarse-grained biotite-magnetite-gold stockwork
represents the older, high-density, hydrothermal fluid
associated with the subvolcanic-porphyry style of miner
alization. Mixing of the older, high-density fluid with
meteoric water resulted in deposition of gold in the
open-space breccia at depths of 50 to 200 m below the
paleosurface. Near-surface argillic, advanced argillic, and
phyllic alteration developed in the shallow hot-spring
environment, and alunite from the advanced argillic zone
has an age of 151#15 ka (Davies and Ballantyne, 1987).
Detailed geochemical studies of REE demonstrates that
boiling of the LREE-enriched magmatic fluids occurred
at depth and resulted in condensation of vapor phases
into the meteoric system, and this developed an acid-sul
fate fluid in the near surface (Lottermoser, 1990). The
presence of active hot-spring and solfatara activity
within the deposit is indicative of this waning stage of
acid-sulfate hydrothermal activity. The episode of por
phyry intrusion, porphyry-style alteration and mineraliza
tion, and hot-spring gold deposition occurred over a
remarkably short interval of about 360,000 years and
demonstrates that long-lived hydrothermal systems are
not necessary for the formation of world-class gold
deposits. As Henley (1990) has noted, the Ladolam
deposit demonstrates a clear continuum from magmatic
hydrothermal to the meteoric-water-dominated hydrother
mal ore deposition.
Similarities and differences between Ladolam and
typical porphyry coppergold deposits are listed in table
4. These differences imply that the Ladolam deposit
formed at a much higher level in the crust than is typical
 LADOLAM GOLD DEPOSIT, LIHIR ISLAND, PAPUA NEW GUINEA 125

Table 4. Comparison of Ladolam deposit to geologic features common to porphyry coppergold

deposits.

Characteristic Ladolam Porphyry Cu-Au

Intrusive rock texture....... Equigranular, minor porphyry Porphyry texture is predominant.
Alteration........................ Biotitic, argillic, and acid-sulfate Porphyry texture is predominant.
Ore assemblage Marcasite, pyrite, magnetite, Chalcopyrite, magnetite, (pyrite),
gold. gold.
Stockwork vein types Anhydrite veinlets only Biotite-K-feldspar, Quartz
magnetite-chalcopyrite.
Temperature.................... 140150C and 200220C 400700C.
(Moyle and others, 1990).
Fluid salinity...................3.8 weight percent NaCl 30-75 weight percent NaCl
equivalent (Moyle and others, equivalent.
1990).
Fluid inclusion daughter
minerals.......................... Sr and Ba sulfates in anhydrite Na-K-Fe chlorides.
Breccia types...................Phreatic and primary volcanic Hydrofracturing.

of porphyry systems. The pressure quench that produces
a porphyry texture did not occur during the late mag
matic phase of intrusion nor did the evolution of a high
pressure fluid capable of hydrofracturing the intrusion to
produce stockwork veins. Breccia, more characteristic of
phreatic explosive activity, became the important ore
trapping structure. The anhydrite stockwork, as pointed
out by Moyle and others (1990), probably formed during
invasion by marine water. This conclusion is supported
by the presence of strontium sulfate as the important
daughter mineral in the fluid inclusions.
Ladolam differs from epithermal deposits mainly in
its high content of hydrothermal biotite. It differs from
epithermal, alunite-rich deposits in the low abundance of
enargite or luzonite. It differs from gold telluride deposits
associated with alkalic igneous rocks (Cripple Creek, Col
orado, and Emperor, Fiji) in its low abundance of gold or
silver tellurides. Although vanadium content is high at
Ladolam, the green vanadium mica typical of gold tellu
ride deposits has not been observed.
In conclusion, it is difficult to find similarities
between Ladolam and any known type of deposit. More
over, it does not resemble any of the known prospects in
the Tabar-Feni island chain. Until other similar deposits
are discovered, Ladolam must be considered unique.
ACKNOWLEDGMENTS
We wish to express out gratitude to Kennecott Explo
ration (Australia, Ltd.) for providing access to the deposits
on Lihir, Tatau, and Simberi. In particular we wish to
thank Rod Davies, Peter Morrissey, and John Wier for
providing information and ideas as well as samples from
each of the deposits. Our appreciation goes to Rick Roger
son of the Papua New Guinea Geological Survey for
logistical help and discussions during our field work.
REFERENCES CITED
Cox, Alan, and Dalrymple, G.B., 1967, Statistical analysis of
geomagnetic reversal data in the precision of potassium
argon dating: Journal of Geophysical Research, v. 72, no.
110, p. 26032614.
Dalrymple, G.B., and Lanphere, M.A., 1969, Potassium-Argon
Dating: San Francisco, W.H. Freeman, 258 p.
Davies, R.G., and Ballantyne, G., 1987, Geology of the Ladolam
gold deposit, Lihir Island, Papua New Guinea: Proceedings,
Pacific Rim Congress, 1987, Gold Coast, Queensland, Austra
lia Institute of Mining and Metallurgy, p. 943949.
Davis, J.C., 1986, Statistics and Data Analysis in Geology: New
York, John Wiley, 646 p.
Getz, A.J., and Ward, R.J., 1990, Western Pacific precious metals
projects updated: Mining Engineering, v. 42, no. 9, p.
10671070.
Henley, R.W., 1991, Epithermal gold deposits in volcanic ter
ranes, in Foster, R.P., ed., Gold Metallogeny and Exploration:
Glasgow and London, Blackie, p. 133164.
Ingamells, C.O., 1970, Lithium metaborate flux in silicate analysis:
Analytica Chimica Acta, v. 52, p. 323334.
John, D.A., Nash, Thomas, Clark, C.W., and Wolftange, W.H.,
1991, Geology, hydrothermal alteration, and mineralization at
the Paradise Peak gold-silver deposit, Nye County, Nevada, in
Geology and Ore Deposits of the Great Basin: Reno, Nevada,
Geological Society of Nevada, p. 10201049.
Johnson, R.W., 1979, Geotectonics and volcanism in Papua New
GuineaA review of the late Cainozoic: Bureau Mineral Re
sources, Journal of Australian Geology and Geophysics, no. 4,
p 18.1207.
126 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Lottermoser, B.G., 1990, Rare-earth element and heavy-metal be
havior associated with the epithermal gold deposit on Lihir Is
land, Papua New Guinea: Journal of Volcanology and
Geothermal Research, v. 40, p. 269289.
Moyle, A.J., Doyle, B.J., Hoogvliet, H. and Ware, A.R., 1990, Lad
olam gold deposit, Lihir Island, in Hughes, F.E., ed., Geology
of the Mineral Deposits of Australia and Papua New Guinea:
Australia Institute Mining and Metallurgy Monograph No. 14,
v. 2, p. 17931805.
Plimer, I.R., Andrew, A.S., Jenkins, R., Lottermoser, B.G., 1988,
The geology and geochemistry of the Lihir gold deposit, Papua
New Guinea: Melbourne, Geological Society of Australia,
Extended Abstracts and Programs, Bicentennial Gold 88, p.
139143.
Stacey, J.S., Sherrill, N.D., Dalrymple, G.B., Lanphere, M.A., and
Carpenter, N.V., 1981, A five-collector system for the simulta
neous measurement of argon isotopic ratios in a static mass
spectrometer: International Journal of Mass Spectrometry and
Ion Physics, v. 39, p. 167180.
Steiger, R.H. and Jaeger, E., 1977, Subcommission on geochronol
ogyConvention on the use of decay constants in geo- and
cosmochronology: Earth and Planetary Science Letters, v. 36,
p. 359362.
Wallace, D.A., Johnson, R.W., Chappell, B.W., Arculus, R.J., Per
fit, M.R., and Crick, I.H., 1983, Cainozoic volcanism of the
Tabar, Lihir, Tanga, and Feni Islands, Papua New Guinea; ge
ology, whole rock analyses and rock forming mineral compo
sition: Bureau Mineral Resources, Geology and Geophysic
Australia Report, 243 p.
 CHAPTER M

VOLATILES IN CLAY MINERALS FROM

SEDIMENTARY AND HYDROTHERMAL
ENVIRONMENTS: A POTENTIAL PETROLOGIC
AND MINERALS-ASSESSMENT TOOL2

By GENE WHITNEY! and GARY P. LANDIS

ABSTRACT released from the clay samples. Preliminary results show

Clay samples from two different geochemical envi
ronments were examined using quadrupole mass spectro
metric analysis of gases evolved during heating to
determine whether systematic differences in gas composi
tions can be used as petrologic-geochemical indicators of
geologic environment. An illitic, interstratified illite/smec
tite was collected from the Shannon Sandstone Member of
the Upper Cretaceous Steele Shale in the Powder River
Basin, Wyoming. This rock currently serves as a petro
leum reservoir rock and contains oil. A suite of clay sam
ples also was collected from a fossil acid-sulfate
geothermal system at Alum Mountain, New Mexico, for
comparison. These samples include illitic clays, nearly
pure smectite, nearly pure kaolinite, and a mixture of
kaolinite, smectite, and illitethe samples typify clay
alterations in hydrothermal ore systems. Comparison of
the volatile species liberated and the temperatures at
which they are released provide information about the
abundance and residence of the gases in the clays. This
information may be used to infer compositions of fluids in
which the clays reacted.
All samples were heated from 60C to 1,060C at
5C/min while 20 mass/electric charge (m/e) channels
were analyzed by mass spectrometer. This technique per
mits the evaluation of volatile composition and the specifi
cation of the temperature at which various volatiles were
dramatic differences in volatile compositions of clays from
the two geochemical environments. The petroleum-reser
voir clay bears a distinctive organic signature, even after
particle size separation and dispersal, whereas the geother
mal clays show more distinctive characteristics related to
sulfide and sulfate volatiles. Mineralogically similar sam
ples from the two environments show roughly similar
water-evolution profiles but dramatically different behav
ior for other gases. For example, the profiles for organic
species, CO2, and sulfur gases are entirely different for
illitic clays from the two environments. On the other hand,
samples with different mineral compositions from the
same geochemical environment also showed distinctly dif
ferent gas-evolution profiles: kaolinite and smectite from
the fossil geothermal system exhibit entirely different pro
files for HF, HCl, and sulfur gases. These data suggest
that volatile composition and evolution profiles for clay
minerals may be useful for fingerprinting geochemical flu
ids in which the clays formed; volatile composition and
evolution profiles may be quantifiable with additional
basic research on siting and partitioning of volatiles in
clays and clay-rich rocks. Site-specific characterization of
gas species in clays may prove a valuable tool in mineral
resource assessment studies by quantifying the degree to
which ore-forming fluids have interacted with host rocks.
INTRODUCTION

"U.S. Geological Survey, Mail Stop 904, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
Clay minerals may form and react in a variety of
aqueous fluids but are usually too fine grained to contain
useful fluid inclusions. Do these fine-grained minerals
retain traces of the fluids (including volatile components)
128 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
in which they reacted? Volatiles released from heated clay
minerals have been identified and measured by mass spec
trometry and by analysis of the condensed gases (Keller,
1986; Heller-Kallai and others, 1988; Gibson and Johnson,
1972; Gibson, 1973; Wicks and Ramik, 1990; Morgan and
others, 1988; Mller-Vonmoos and others, 1977; Heller
Kallai and others, 1989). Important volatile species identi
fied include H2O, CO2, O2, NH3, Ar, He, H2, nitrates, sul
fates, and a variety of additional organic and inorganic
compounds. Although these volatiles have been identified,
there has been little attempt to relate the volatile content
of clay minerals to details of crystal chemistry or to the
geochemistry of the environment in which the clays
formed, despite the fact that the composition of selected
evolved volatiles has been successfully used as a petro
logic tool in igneous and metamorphic environments (Mat
son and others, 1984, 1986; Garcia and others, 1980). The
goal of the present research is to measure the variations in
abundance and composition of clay volatiles as a function
of clay mineralogy and crystal chemistry and to relate the
volatile chemistry of clays to their environment of forma
tion. The work presented here is preliminary but provides
some measure of the usefulness of this technique. It also
points to the need for further research, especially in spe
cific ore-forming environments.
VOLATILES AND THE CLAY
MINERAL STRUCTURE
Volatiles in clay minerals are potentially important
indicators of geologic processes because clays form in a
wide variety of diagenetic and hydrothermal environments,
and clay minerals themselves exhibit a range of composi
tion and structure. Volatiles derived from clay minerals
during heating may be released from different sites within
the clay sample. These sites fall into three general catego
ries: surface sites, interlayer sites, and structural sites (fig.
1). Surface sites are external sites on the clay particle onto
which neutral or charged species may be adsorbed. Sur
face sites in clays do not differ, in a qualitative sense,
from surface sites on other minerals. The strength of
attraction between the volatile species and the mineral sur
face may vary greatly, depending upon the charge and
polarity (if any) of the volatiles, the nature of the mineral
surface, and the medium in which attachment occurs. The
strength of attachment ranges from very weak (simple
adsorption of a neutral species on a neutral surface) to
very strong (chemisorption of a charged species on a
charged surface). Clays differ from other minerals prima
rily in their unusually fine grain size and associated large
surface area. Thus, clays may retain volatiles in large
quantities simply by virtue of available surface area. In
addition, certain clays bear a layer charge, which, if
SURFACE
ADSORPTION
SITES
W STRUCTURAL SITES
->) A N
I_LI
N Zs.
| | SITES
/ N
Figure 1. Schematic diagram of clay mineral structure showing
possible sites of volatile species in clays.
unsatisfied on surfaces or edges, may facilitate pro
nounced adsorption-chemisorption phenomena (Yariv and
Cross, 1979). Because of high surface area and structural
layer charge, clays often contain substantial amounts of
adsorbed or chemisorbed atmospheric gases that are driven
off by heating at low to moderate temperatures.
The second site potentially occupied by volatile com
ponents is the interlayer space of clay minerals. Clays are
comprised of stacks of silicate sheets. Clay minerals with
a net negative charge on the silicate layers (smectites, ver
miculite, mica) contain cations between layers to balance
that charge (fig. 1). In the expandable clays (smectite and
vermiculite), these interlayer cations are exchangeable and
frequently hydrated. Thus, expandable clays usually con
tain large amounts of free molecular water. This expand
able characteristic also permits the interchange of other
molecules and ions between the surrounding medium and
the interlayer space. For example, polar organic molecules
may be associated with interlayer cations in expandable
clays (Lagaly, 1984). Mica is a nonexpandable structure,
and the interlayer cation is not exchangeable. Therefore,
there is little interchange of ions or molecules after crys
tallization between the interlayer space of mica and the
surrounding medium. However, volatile species that
occupy the interlayer space of mica (such as Ar and NH3)
may be released during heating. Ordinarily, mica must be
heated to a temperature sufficient to disrupt the silicate
sheet structure before these interlayer species are liberated.
The third site for volatiles in clay minerals is in
structural cation and anion sites in the silicate layers.
Substitution of volatiles for structural cations is relatively
unusual in the clay structure, but anion substitution is
common. The most abundant structural volatile in clay
minerals is hydroxyl. When clays are heated (in vacuum)
to a temperature sufficient to dehydroxylate the silicate
layers, some free hydrogen and hydroxyl is released and
combines to form molecular water, leaving a dehydrated
oxide structure. Substitution of F or Cl for structural
hydroxyl is common in small amounts, and when the
mineral reaches its dehydroxylation temperature, much of
the F and Cl also may be released. The structural sites
 VOLATILES IN CLAY MINERALS
are the most restrictive because the silicate layer structure
will accommodate only a few species that do not disrupt
the crystal structure because of charge and ionic-radius
constraints. Therefore, nitrates, sulfates, carbonates, and
organic species must reside in the interlayer space or on
surface sites: they cannot be accommodated in the silicate
layer structure.
The range of species that may occupy the three dif
ferent sites in clay minerals, and the temperatures at which
they are likely to be liberated, varies greatly for the three
types of sites. Surface sites may be occupied by any vola
tile species that exists in the environment that will attach
to the mineral. These volatiles are generally loosely bound
and are released at relatively low temperatures (~ 100C),
though chemisorption may cause some species to be
retained until heating to 300C or above (Adamson, 1982).
Interlayer species are somewhat more restricted and are
likely to be released at temperatures of 100C to 300C
(Wicks and Ramik, 1990). Structural volatiles are limited
almost exclusively to anions that may occupy hydroxyl
sites. These species are released at temperatures from
450C to 1,200C, depending upon the specific clay and
its structure (Mackenzie, 1970, 1972). Adsorbed molecular
water and structural hydroxyl are consistently the most
abundant volatiles released from clays upon heating. Nev
ertheless, it may be minor volatile species that reveal the
composition of formation fluids or the presence of hydro
carbons or other distinctive volatiles in a particular geo
logic environment.
SAMPLES AND
EXPERIMENTAL METHODS
In order to systematically investigate the ability of
clay minerals to retain volatiles from their environment of
formation, two suites of well-characterized samples were
selected for preliminary analysis (table 1). The first set
consists of authigenic clay minerals from core samples of
petroleum reservoir sandstones in the Powder River Basin,
Wyoming (Hansley and Whitney, 1991). The second set
of samples is from outcrop samples of a fossil acid-sulfate
geothermal system at Alum Mountain, New Mexico
(Northrop and Whitney, 1987). The intent of sample selec
tion was to check the variation in volatile composition of
different clay minerals from the same environment, and
the same (or similar) minerals from different environ
ments, in order to assess how well the volatiles reflected
the environment of formation and whether particular clay
minerals tend to collect particular volatile species. Only
selected samples from each suite are described here.
The laboratory preparation for all samples was the
same. Each whole-rock sample was disaggregated ultra
sonically in distilled water and the fine-size fractions were
129
Table 1. Description of samples selected for this study.
Sample number Mineralogy and description
PRB3 .................. Authigenic
interstratified illite/smectite (15
percent expandable) from Upper Cretaceous
Shannon Sandstone Member of the Steele
Shale, Powder River Basin, Wyoming.
Sample contains < 10 percent chlorite as
the only identifiable impurity. Particle size
< 0.2 plm.
B2408G................. Illite from the Alum Mountain fossil
geothermal system. Contains trace amounts
of kaolinite, K-feldspar and calcite(?).
Particle size ~ 1.0 pum.
GW-86-53 ............. Illitic clay from Alum Mountain fossil
geothermal system. Contains < 15 percent
kaolinite and a trace of jarosite. Particle
size ~ 1.0 p.m.
GW-8644 .............Kaolinite from Alum Mountain fossil
geothermal system. Contains a trace of
natroalunite or jarosite. Particle size ~ 1.0
|Im.
GW-8640 ............. Smectite from Alum Mountain fossil
geothermal system. Contains a trace of
kaolinite. Particle size ~ 1.0 pum.
GW-86-33 ............. Mixture of smectite, kaolinite, and illite
from Alum Mountain fossil geothermal
system. May contain a trace of sulfate
mineral. Particle size ~ 1.0 p.m.
separated by centrifugation. Size separates were oriented
on glass slides using the technique of Drever (1973), and
clay mineral composition was determined by X-ray dif
fraction. Air-dried, ethylene-glycol-saturated, and heated
(550C) mounts of each sample were X-rayed for mineral
identification. Separate aliquots of each sample were dried
in Teflon beakers at 60C for subsequent thermal and vol
atile analysis.
Each sample was subjected to thermogravimetric
analysis using a Perkin-Elmer TGS-2 programmable ana
lyzer. Ten milligrams of each sample were heated at
10C/min from ambient to 1,000C in a flowing-nitrogen
atmosphere at approximately 1 bar (105 Pa) pressure.
Weight loss and first derivative of weight loss were
recorded with a computer-aided data-acquisition system
consisting of MacADIOS interface hardware and Lab
VIEW II data-collection and analysis software on an
Apple Macintosh computer.
Analysis of volatile species was performed by placing
a separate 20 mg aliquot of each sample in a quartz tube
(7 mm inside diameter) located in a muffle furnace and
connected to the mass spectrometer. The quartz sample
tube was evacuated, and the sample was heated stepwise
from 60C to 1,060C at 5C/min under high vacuum (10
8 torr). Samples were routinely baked out at 60C or
100C prior to analysis. Evolved volatiles were analyzed
130 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
with a Balzers 511 quadrupole mass spectrometer auto
mated with a DEC PDP-11/83 data-collection system
(Landis and Hofstra, 1991). Twenty masses (m/e) were
measured every 12 seconds during the temperature ramp
and recorded digitally.
RESULTS: BACKGROUND AND
REPRODUCIBILITY TESTS
We performed a set of analyses designed to test the
sample-preparation and analytical techniques and to estab
lish background and reproducibility for a variety of vola
tiles. Of particular concern is the adsorption of
atmospheric gases by fine-grained clay samples. It is
important to distinguish geologically important gases from
those adsorbed from the atmosphere during the sample
preparation process. In addition, we wanted to detect any
gases being given off by the quartz sample tubes (or
leaked gases) during heating.
The first test measured the amount of gas released by
the quartz sample tubes, attributable to either adsorption of
atmospheric gases or the presence of gas bubbles within the
quartz that makes up the tubes. Figure 2 shows a direct
comparison of the empty quartz sample tubes heated after
exposure to the atmosphere and then reheated without
exposure to the atmosphere. This test determines what
gases are from the atmosphere and what gases are from the
quartz. Detectable amounts of atmospheric gases are
clearly present on the opened tubes, whereas the quartz
tubes kept under vacuum show virtually no extraneous
gases. The rise in background of both opened and
unopened tubes must represent virtual leakage into the
mass spectrometer at the higher temperatures from vac
uum-system surfaces. It is important to note that the gas
release events exhibited by the opened tube actually con
stitute very low level background compared to the amounts
of gas released by samples (fig. 3, for comparison).
The second test utilized quartz powder (clear, euhe
dral crystals ground and elutriated to a 20- to 54-pum grain
size and acid treated to remove hyperfine particles) to
determine amounts of volatiles released by surface pro
cesses alone. (In order to measure the effects of surface
adsorption for an inert solid, the solid must release only
surface species and no volatiles from its own structure.)
For the quartz powder used in this test, CO2 and H2O
exhibit sudden release events near 200C, 290C and
520C, suggesting that the quartz may contain microinclu
sions that burst at those temperatures (fig. 3A). Neverthe
less, the amounts of gases released are roughly two orders
of magnitude smaller than the amounts of these gases
released during a typical clay analysis (fig. 3B). Aside
from these small events from the quartz, no significant
amounts of volatiles are released, suggesting that simple
surface adsorption of atmospheric gases will not play a
significant role in the analysis of clay volatiles.
The third test measured the reproducibility for multi
ple analyses of a single sample. Figure 4 shows duplicate
runs of sample GW-86-44 for three species. The tempera
tures at which volatiles are released are reproducible
within the precision of the temperature-measurement tech
nique. The absolute amounts of volatiles released varied
somewhat from run to run, but these changes may be
attributable to sample size and different surface areas, and
to sample inhomogeneity.
RESULTS AND PRELIMINARY
INTERPRETATION:
VOLATILE COMPOSITIONS OF
NATURAL CLAYS
1. PETROLEUM RESERVOIR CLAYS
The first clay sample analyzed was an authigenic
interstratified illite/smectite (I/S) from the Upper Creta
ceous Shannon Sandstone Member of the Steele Shale in
the Powder River Basin, Wyoming. This sandstone has
served as a petroleum reservoir and still contains petro
leum (Hansley and Whitney, 1991). Thermogravimetric
(TGA) analysis (in N2 at 1 atmosphere pressure) of the
<0.2-pum fraction of this reservoir clay reveals the loss of
adsorbed volatiles at 100C and dehydroxylation and
structural modification of the clay at about 570C (fig. 5).
The evolution of water during heating (under vacuum) in
the furnace attached to the quadrupole mass spectrometer
produced a peak at around 450500C (fig. 6A). In both
the TGA and the mass spectrometer pyrolysis analyses,
the water-loss events are composed of multiple peaks, but
it is not certain what each of these peaks indicates. The
lower temperature at which dehydroxylation occurs in the
mass spectrometer is attributable to the vacuum because
gas-release reactions are accelerated under high vacuum.
Virtually all of the adsorbed water on the clay had been
lost during the 100C bakeout prior to the analysis.
In addition to the release of structural water, several
other species were released from this petroleum-reservoir
clay in significant amounts during the water-release event.
For example, HF (m/e = 20) and HCl (m/e = 36) mimic
the water-release event exactly, as would be expected for
anion species that substitute for structural hydroxyl. Other
species that cannot substitute for structural anions, such as
H2S (m/e = 34) and Ar (m/e = 40), also exhibit events at
the dehydroxylation temperature (fig. 6B, 6C), suggesting
that they occupy interlayer spaces of the clay and are only
released during the structural breakdown of the mineral.
On the other hand, several volatiles were released that
show no correlation whatsoever with the dehydroxylation
event of the clay. These include several organic species or
 VOLATILES IN CLAY MINERALS 131

1.2x10"
4.0.10" H- +------

3.0x10* H

2.010* |

1.0x10* H

Pretired tube
l l l

100 300 500 700 900 1,100 100 300 500 700 900 1100

1.5x10" I I i r 8.0x10 I I I I I

1.2x10" F
6.0x10" | | | | -

9.0x10" F
4.0x10" H
6.0x10" H

20x10' HAL W -

3.0x10" |
Prefred tube *~
I I

B 100 300 soo 100 300 500 700 900 1100

8.0x1 0.11 - T I I T T
3.6x10

2.710
6.0x10"

1.8x10
4.0x10"

9.0x10"
2.0x10"

- - - : - Pietired tube
0 i i 1 i i i i 11
C 100 300 500 700 900 1,100 100 300 500 700 900 1100

Temperature (C) Temperature (C)

Figure 2. Background test. Comparison of the gas evolution pro
files (for m/e = 28, 34,64) for a quartz sample tube exposed to the at
mosphere (opened) compared to the same quartz sample tube
reanalyzed without exposure to the atmosphere (prefired). Volatiles
released are clearly attributable to adsorption of atmospheric compo
nents. A, m/e = 28, attributable to N2 and the CO ion fragment from
CO2; B, m/e = 64, attributable to SO2: C, m/e = 34, attributable to
H2S.
Figure 3. Background test for the sample tube and an inert sub
stance (in this case quartz powder) illustrating the relative amounts
of water (m/e = 18) adsorbed and then released from exposure to at
mosphere versus that of sample (smectite clay). A, release from
quartz powder in opened vacuum furnace tube; B, release from typ
ical clay sample, smectite; C, release from prefired vacuum furnace
tube. Although the quartz powder exhibits some evolution of water,
the concentrations are at least two orders of magnitude lower than
during a clay analysis and occurs at lower temperatures.
 132 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

4.0x10 i I T T I T I I T I _ 100 -

3 98 T-I- | - -

2. H,o = 0. wn |
# 3.0x10
9
94 - - -
3.
: 92 | | | | |
> 90 /~ - -

# 2.010 4 - * \- -

E -- t - - - - -

2 1st derivative
|
- -

s 1.0x10
=
l -

0 100 200 300 400 500 600 700 800 900

0 1 1

Temperature (C)
* 100 900 1 100

Figure 5. Thermogravimetric analysis of sample PRB3, showing

major dehydroxylation peak at 570C under 1 atm, N2 gas flow. Up
5.0x10 I
per trace is gravimetric response and lower trace is first derivative.

3. 4.010* ||
E 2. FOSSIL ACID-SULFATE GEOTHERMAL
3.
> 3.0x10" H SYSTEM CLAYS
7,
E
S
Analysis of several samples with different clay-min
5 2.010* H
c eral compositions from a fossil acid-sulfate geothermal
r
quo system permits evaluation of how volatile retention varies
= 1.0x10" H as a function of mineralogy and permits qualitative analy
sis of gas species present. Five samples from the Alum
0 Mountain fossil geothermal system were selected (table 1).
B 100 These samples come from different alteration zones within
the geothermal system (Northrop and Whitney, 1987), so
each mineral or assemblage may have formed in contact
8.0x10 I T T T-I- T I I T I
with fluids with slightly (but systematically) different
compositions. Such variations of fluid composition within
2. N2. co, CO2 : - ! -

a geothermal system is an additional variable to consider

# 6.010* |
c;
:
-
;
:
:
-
when examining differences among samples.
^- i\;
>, - 1 - Some variation in thermal behavior as a function of
: i i ! .. mineralogy is seen in the m/e = 18 profiles. The main
# 40x10" | | || "...'. A event for each sample is the loss of structural water (dehy
droxylation). For kaolinite (fig. 8A), dehydroxylation
Run 2 /.
occurs as a simple, symmetric event near 420C under
E 2.010* | *J
-

*\^ /*N
:-- Run 1 high vacuum. There is little evidence for adsorbed water
or for any other minerals with different breakdown tem
0 I l l l | l l l
peratures in this sample. The illitic samples (fig. 8B, 8C)
C
100 300 500 700 900 1100 exhibit more complex dehydroxylation events, but both
Temperature (C) occur near 430450C. The smectite (fig. 8D) dehydrox
ylates at 600C but shows water evolution almost continu
Figure 4. Duplicate analyses of the same material, analyzed two
weeks apart. A, m/e = 18, the major peak of water; B, m/e = 20, a mi ally from 100C to 600C. In fact, both the smectite
nor peak of water and the major peak of HF: C, m/e = 28, the major sample and the illitic samples (which contain a small pro
peak of N2 and CO and ion-fragment peak of CO2. portion of smectitic layers) exhibit low-to-moderate-tem
perature gas evolution. This effect probably reflects
retention of water in the interlayer space of these clays
fragments with m/e = 27, 28, and 29 (fig. 7A, B, C). These that is greater than simple adsorption (Sposito, 1984;
events probably represent thermal pyrolysis of organic Sposito and Prost, 1982). As would be expected, a sample
matter that is associated with the clay but is not contained consisting of mixtures of these minerals exhibits volatile
structurally in the clay. evolution curves that are composites of the individual
 VOLATILES IN CLAY MINERALS 133

3.6x10' 7.2x10 T T T i I T I T T i

2.7x10' 5.4x10 - -

1.8x10" 3.6x10 - -

9.0x10" 1.8x10 - -

0 i l l I I i i i i i

A 100 300 500 700 900 A 100 300 500 700 900 1100

2.010 1.2x10' I T i I I I

1.5x10 9.0x10"

1.0x10 6.0x10"

5.0x10" 3.0x10"

B 100 300 500 700 900 B 100 300 500 700 900 1100

1.6x10 I I T I T-I- I T I T
6.0x10

1.210 H -
-
4.5x10

| 8.0x10" H

4.0x10" H
3.0x10

: 1.5x10

C 100 300 500 700 900 1100

C 100 300 500 700 900 1100
Temperature (C)
Temperature (C)

Figure 6. Gas-evolution profiles for sample PRB3. Both structur Figure 7. Gas-evolution profiles for organic species or ion frag
al and nonstructural volatile species are evolved over a narrow tem ments in sample PRB3 (m/e = 27, 28, 29). The m/e = 28 peak (part
perature range. A, water, m/e = 18; B, hydrogen sulfide, m/e = 34; B) saturates the electrometer at 1x10' amps from the major nitro
C, argon, m/e = 40. gen peak at the same m/e as organic-species ion fragments.
134 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

2.010 I T-I- I I-1 I I 7.2x10' T I I I I T T- T i

A Kaolinite 3. 7 Smectite
1.5x10" H * - 5 5.4x10' -

>

1.0x10 3.610' -

E
oc

5.0x10 a 1.8x10" -

1.5x10 I-I-I-I-I-I-I-T-T

,- - Mineral mixture
Illite 1
3. Illite + kaolinite + smectite
5 - -

... 1.010
'7,

S
E

* 5.010'

E

E 100 300 500 700 900 1100

Temperature (C)

Figure 8. Water-release profiles (m/e = 18) for several samples

Illite 2
with different mineral compositions from the Alum Mountain fossil
geothermal systemthis demonstrates mineralogical differences
for a single volatile species from a single geologic environment. A,
kaolinite, <1-pm size; B, illitic clay, 0.5- to 5-pum size; C, illitic
clay, 31-pum size; D, smectite; E, mixed illite-kaolinitesmectite
clay, 31-pum size.

sample (fig. 9A). All of the samples release H2S above

C 100 300 500 700 900
Temperature (C)
components, in proportion to the amount of each mineral
present (fig. 8E).
Because this suite of samples formed in a sulfur-bear
ing system, the behavior of sulfide and sulfate volatiles is
of interest. For m/e = 34 (representing H2S), there are sev
eral distinctive features of the gas-release curves. All of
the samples exhibit some H2S release between 400C and
500C that is roughly correlative with clay water-release
events (fig. 9). The kaolinite sample exhibits an H2S dou
blet at 200C and 250C that is not observed in any other
700C, but no two samples exhibit the same temperatures
or patterns of gas release. It is likely that these high-tem
perature events are not related to structural changes in the
1,100
clays but represent calcining of sulfide minerals present in
amounts too small to detect with X-ray diffraction (as
much as a few percent). They may also represent intra
crystalline H2S that is released when some thermal thresh
old is reached. This hypothesis is supported by the
behavior of SO2, which also shows patterns of liberation
that seem to be unrelated to dehydroxylation or other ther
mally induced changes in the clay minerals (fig. 10). For
example, both the kaolinite sample and one of the illitic
samples exhibit a sharp SO2 peak at 850C. Such a simi
larity cannot be easily explained in terms of clay crystal
chemistry because the two minerals are structurally and
chemically different.
 VOLATILES IN CLAY MINERALS 135

1.610 2.0x10" T I I i I I I I I

1.210 1.5x10"

8.010" 1.0x10"

4.0x10" 5.0x10"

D 100 300 500 700 900 1,100

Mineral mixture
Illite + kaolinite + smectite
2.4x10 9.0x10" H

1.610

8.0x10"
| 6.010"

3.0x10"
:
500 700 900 1,100
Temperature (C)

Figure 9. Hydrogen sulfide evolution profiles (m/e = 34) for min

eralogically diverse samples from the Alum Mountain fossil geo
1.2x10 H thermal system. A through E are the same as in figure 8.

8.0x10" its dehydroxylation temperature. These differences in

anion compositions for different minerals from the same
geochemical environment suggest that certain volatiles
4.0x10" may be partitioned between clay minerals, and such parti
tioning may be useful as indicators of temperature and
fluid composition. Alternatively, the different clay miner
als may have formed from fluids with different composi
C 100 300 500 700 900 1100 tions. In this case, the differences in volatile composition
Temperature (C) between clays may reflect an evolution in fluid composi
tion with time.

Gas-evolution patterns for several other species are

potentially useful for crystal chemical and geochemical
3. COMPARISON OF PETROLEUM
interpretations. For example, the release pattern of HF
(m/e = 20) follows that of water for both the smectite RESERVOIR AND GEOTHERMAL CLAYS

and kaolinite samples (fig. 11), reflecting normal anionic

substitution of F for OH in the clay structure. However,
despite the fact that C1 also may substitute for OH', the
smectite sample displays no Cl (measured as HCl, m/e =
36) release at the dehydroxylation temperature (fig. 11D),
whereas the kaolinite sample shows a strong Cl- peak at
Besides the variation in volatile content for different
clays from the same environment, it is instructive to exam
ine the differences and similarities in the general volatile
composition of clays from the two different environments.
Because various clay minerals incorporate several volatiles
136 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

1.2x10" I 3.6x10"

9.0x10 2.7x10"

| 6.0x10

3.0x10
| 1.8x10"

9.0x10
: :
D 100 300 500 700 900 1100

Mineral mixture i

Illite + kaolinite + smectite

2.7x10" H

1.8x10" H
:
9.0x10

-|
900 1,100 E 100 300 500 700
Temperature (C)
900 1,100

Figure 10. Sulfur dioxide evolution profiles (m/e = 64) for geother
mal clays from Alum Mountain showing great diversity of release
1.0x10"
for different minerals from the same geochemical environment. A
through E are the same as in figure 8.

5.0x10" system is skewed toward lower temperature (fig. 12). The

C 100 300 500 700 900 1100
Temperature (C)
in different ways related to their structure, it is perhaps
most important to compare structurally similar minerals
from the two environments. The clay from the petroleum
reservoir rock is a highly illitic, interstratified, illite/smec
tite and is compared with an illitic clay from the geother
mal system. As expected for two mineralogically similar
samples formed in aqueous environments, the dehydroxy
lation behavior for the two samples is similar, but the
water-release event for the illite from the geothermal
illitic clay from the petroleum reservoir displays several
gas-release events that are probably related to the presence
of organic matter in the sample. For m/e channels repre
senting organic compounds and their fragments (e.g., m/e
= 26, 27, 29), the petroleum-related illite produced events
that are missing in the geothermal illite runs, particularly at
about 430C (fig. 13). In addition, the petroleum reservoir
clay released carbon dioxide and related fragments (m/e =
12, 28, 44) at 590C, an event that is not seen in the geo
thermal illite (fig. 14). Similarly, H2S (m/e = 34) and SO2
(m/e = 64) were released from the petroleum clay at
500550C and, again, there were no similar release
events from the geothermal illite (fig. 15). Conversely, the
geothermal illite released several of these species between
750C and 850C, with no parallel events in the petro
leum-related illitic material. It is not yet clear what these
high-temperature events represent, but they must represent
accessory minerals (e.g., carbonates and sulfides) present
 VOLATILES IN CLAY MINERALS 137

1.2x10" I I T I I T T I I T 2.8x10 I I I I I I I I T i

? HF HF
2.110 Smectite
5 9.010 Kaolinite .1x10' -

5
# 6.0x10
c
r

# 3.0x10

A
0

100 300 500 700 900

l

1,100
-

| 1.4x10

7.0x10"

C
0
-

3.6x10" T T T T T T I | T I 1.5x10" I I T-I i I T I T I

? HCI HCI
2.7x10" H Kaolinite - Smectite
- ... I X - : -

> 1.0x10" -

= 1.8x10" |
\
er,

s
E 9.0x10-11

B
0 l

500
1 l

700
l

900
l l 1

1,100
-

| 5.0x10"

D
0
100 300 500 700 900
|
1100
-

Temperature (C) Temperature (C)

Figure 11. Comparison of HF (m/e = 20) and HCl (m/e = 36) liberation behavior for smectite and kaolinite from the Alum Mountain
fossil geothermal system.A, kaolinite HF, B, kaolinite HCl; C, smectite HF: D, smectite HCl.

in very small amounts or volatiles trapped within clay par 1.2x10 II --

ticles as intracrystalline molecular species. In either case, it
is clear that these mineralogically similar samples from
Petroleum I/S
2. i i ;
two diverse geochemical environments produce distinctly # 9.0107 F
different gas spectra during thermal ramping. S.
>
# 6.0.10" i i ; ; ;
3 * Hydrothermal illite
DISCUSSION AND FUTURE WORK
2 - - -

s: 3.0x10
E
The purpose of the present study was to determine
whether clay minerals formed in different geochemical 0 l 1 l-l l l

environments might retain some memory of the compo 100 300 500 700 900 1,100
sition of formation fluids as volatiles adsorbed on their
Temperature (C)
surfaces, trapped in interlayer spaces of clays, or incorpo
rated within structural sites in the silicate layers of the Figure 12. Dehydroxylation profiles of mineralogically similar
clay. Although a great deal of systematic study is still illitic samples from both the petroleum reservoir (PRB3) and geo
required, it is apparent that the gas chemistry of mineral thermal (B2408) systems. Although the geothermal sample (illite)
ogically similar clays from a petroleum environment and a has more low-temperature activity than the petroleum-reservoir
geothermal-hydrothermal environment give strikingly sample (I/S), comparable dehydroxylation behavior is achieved
different gas-evolution profiles when heated. Further, from the same basic mineralogy.
138 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

1.0x10 I-I-I-I-I-I-I-I-H 1.0x10 T I I I I I I I T I

Hydrothermal illite
- - Hydrothermal
.9 || Petroleum, illite Petroleum I/S
#
,-
8.0x10 Is N i
3 8.010" |
- -

S. ; : I 5
* 6.0x10 H
wn : :
= 6.010
v

5.
5 4.0x10 H N.
;
:
:
:
S
5 4.0x10
\-
r
s s
E 2.010 H E 2.010

0 Id l l I I l l l l 1
0
A 100 300 500 700 900 1,100 A 100

1.0x10 T I
1.610

2.
3. 8.010 F
E # 1.210" |
3. 3.
> 6.0x10 H : ; >
*2 ydrothermal
5. illite # 8.0x10" |
5 4.0x10 -
E
t
en
s 4.0x10" H
s
B 2.010 : E Q
Hydrothermal illite
0 I l 1 1 I l l l l I I I I l I
0
B 100 300 500 700 900 1,100 B 100 300 500 700 900 1,100

1.610 T-I 4.0x10

Petroleum I's
2. 2.
Hydrothermal
= 1.210* | # 3.0.10" | illite
S. S.
> >
.9
# 2.010* |
5 8.0x10 F
E
*
| E
| *r
3. *r

s 4.0x10 F l s, 1.010* ||
E * \ E

: Hydroth *
i i
l i illite l l i l l i l
0
1 1 1 l l 1 l

C 100 300 500 700 900 1,100 C 100 300 500 700 900 1.100
Temperature (C) Temperature (C)

Figure 13. Comparison of the organic-species and ion-fragment
(m/e = 26, 27, 29) evolution for illitic samples from both environ
ments. Despite their mineralogic similarity, there is almost no cor
relation between the evolution of these species. The geothermal
system (illite) and petroleum reservoir (I/S) are labeled on the dia
grams for A, m/e = 26; B, m/e = 27; and C, m/e = 29.
Figure 14. Carbon dioxide gas release. Comparison of the evolu
tion of C+ (m/e = 12), CO+ (m/e = 28), and CO2+ (m/e = 44) ion
fragments for the illitic clay sample from the petroleum reservoir (L/
S) and for the illitic clay sample from the geothermal system (illite).
A, m/e = 12; B, m/e = 28; C, m/e = 44.
 VOLATILES IN CLAY MINERALS 139

2.010 H--- -- I samples with organic solvents may alter the compositions
: : i : Petroleum I/S :
drastically. Are any adsorbed species retained during sam
Hydrothermal i Ni 1 :
2. i illite ple preparation or are they replaced by different species
# 1.510 ; : from preparation solutions or from the atmosphere? A sys
&
tematic analysis of samples during the course of sample
P
separation and preparation would help answer this ques
# 1.010 tion. A corollary to this question is the effect of particle
E size. Small particles have a higher surface area per unit
r
er, 10
mass than larger particles and would therefore retain a
s 5.0x10"
B higher concentration of adsorbed species. On the other
hand, small particles have less volume for the retention of
0 i interlayer or intracrystalline species and would contain a
lower concentration of these volatiles. Thus, a detailed
A 100 300 500 700 900 1,100
analysis of particle-size effects for a uniform gas-mineral
1.0x10 I I I I I I I I combination would be beneficial.
Petroleum I's Il y rm. Third, and most critically, the quantitative incorpora
3 8.0x10 [. \ . tion of volatiles into a particular mineral, and their exact
E location within the mineral sample, must be determined.
3.
For example, if kaolinite evolves significant amounts of
* 6.0x10 H Cl during dehydroxylation but smectite does not, does

S that imply that Cl occupies basal hydroxyl sites in the

5 4.0x10 H
r
\c
quo
= 2.010 H
0 sition. On the other hand, if the Cl is present simply as an
B 100 300 500 700 900
Temperature (C)
Figure 15. Sulfur gas release. A, The H2S (m/e = 34); and B, SO2
(m/e = 64) profiles for the illitic samples from the petroleum reser
voir (I/S) and the geothermal system (illite).
these different gas compositions are consistent with the
contrasting geochemical environments of the clays.
Based on these preliminary data, several specific areas
of research are now required to permit a more rigorous
interpretation of these differences in volatile chemistry.
First, natural samples must be analyzed chemically
and mineralogically in order to detect very small amounts
of secondary phases. Evolution of CO2, H2S, and SO2
from accessory minerals undetectable by X-ray diffraction
must be distinguishable from these gases held on clay
mineral surfaces or within the interlayer space of expand
able minerals. This work must be performed in conjunc
tion with the characterization of the volatiles evolved by
pure samples of the candidate phases.
Second, if certain volatiles are present as surface spe
cies adsorbed or chemisorbed on mineral surfaces, then
the effects of sample-preparation technique must be
assessed. For example, if organic compounds are present
in the petroleum reservoir rock as grain coatings or inter
layer species, then ultrasonic disaggregation or washing
kaolinite (sites that do not exist in smectite)? If anion par
titioning occurs between different structural sites within
clay minerals, then there is a potential for quantifying
such partitioning in terms of temperature and fluid compo
1,100 intracrystalline inclusion, then its presence simply reflects
crystallographic contiguity and will tell us nothing about
specific temperatures. This problem is probably best
approached using experimental syntheses of clay minerals
in solutions of precisely known compositions.
In conclusion, this preliminary study has shown that
there are significant differences in the composition of
retained volatiles in clay minerals formed in different
geochemical environments. Volatiles for mineralogically
similar samples from a petroleum reservoir rock and from
a fossil geothermal system are distinctly different in com
position and thermal evolution profiles. In addition, miner
alogically different samples from the same geothermal
system exhibit dramatically different gas-evolution pro
files. By systematically determining the abundance and
sites of volatiles in clay minerals and how these volatiles
are incorporated into the clays during formation, we may
be able to use composition and thermal evolution of vola
tiles from clays as petrologic tools to interpret tempera
tures and fluid chemistry of the clay-forming environment.
ACKNOWLEDGMENTS
This research was funded by the U.S. Geological
Survey Development of Assessment Techniques (DAT)
Program.
140 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
REFERENCES CITED
Adamson, A.W., 1982, Physical Chemistry of Surfaces: New York,
John Wiley & Sons, 664 p.
Drever, J.I., 1973, The preparation of oriented clay mineral speci
mens for X-ray diffraction analysis by a filter-membrane peel
technique: American Mineralogist, v. 58, p. 553554.
Garcia, M.O., Muenow, D.W., and Liu, N.W.K., 1980, Volatiles in
Ti-rich amphibole megacrysts, southwest U.S.A.: American
Mineralogist, v. 65, p. 306312.
Gibson, E.K., Jr., 1973, Thermal analysismass spectrometer com
puter system and its application to the evolved gas analysis of
Green River shale and Lunar soil samples: Thermochimica
Acta, v. 5, p. 243255.
Gibson, E.K., Jr., and Johnson, S.M., 1972, Thermogravimetric
quadrupole mass spectrometric analysis of geochemical sam
ples: Thermochimica Acta, v.4, p. 4956.
Hansley, P.L., and Whitney, C.G., 1991, Petrology, diagenesis, and
sedimentology of the Upper Cretaceous Shannon Sandstone
Member of the Steele Shale in the Powder River Basin, Wyo
ming: U.S. Geological Survey Bulletin 1917B, 33 p.
Heller-Kallai, L., Miloslavski, I., and Aizenshtat, Z., 1989, Reac
tions of clay volatiles with n-alkanes: Clays and Clay Minerals,
v. 37, p. 446450.
Heller-Kallai, L., Miloslavski, I., Aizenshtat, Z., and Halicz, L.,
1988, Chemical and mass spectrometric analysis of volatiles
derived from clays: American Mineralogist, v. 73, p. 376382.
Keller, W.D., 1986, Compositions of condensates from heated
clay minerals and shales: American Mineralogist, v. 71,
p. 14201425.
Lagaly, G., 1984, Clay-organic interactions: Philosophical Trans
actions of the Royal Society of London, v. A311, p. 315332.
Landis, G.P., and Hofstra, A.H., in press, Fluid inclusion gas chem
istry as a potential minerals exploration toolCase studies
from Creede, CO, Jerritt Canyon, NV, Coeur d'Alene district,
ID and MT, southern Alaska mesothermal veins, and mid-con
tinent MVT's: Journal of Geochemical Exploration, v.42.
Mackenzie, R.C., 1970, Differential Thermal Analysis, Volume
lFundamental Aspects: London, Academic Press, 775 p.
Mackenzie, R.C., 1972, Differential Thermal Analysis, Volume
2Applications: London, Academic Press, 607 p.
Matson, D.W., Muenow, D.W., and Garcia, M.O., 1984, Volatiles
in amphiboles from xenoliths, Vulcan's Throne, Grand Can
yon, Arizona, U.S.A.: Geochimica et Cosmochimica Acta, v.
48, p. 16291636.
Matson, D.W., Muenow, D.W., and Garcia, M.O., 1986, Volatile
contents of phlogopite micas from South Africa kimberlite:
Contributions to Mineralogy and Petrology, v. 93, p. 399408.
Morgan, D.J., Warrington, S.B., and Warne, S.St.J., 1988, Earth
sciences applications of evolved gas analysisA review:
Thermochimica Acta, v. 135, p. 207212.
Mller-Vonmoos, M., Kahr, G., and Rub, A., 1977, DTA-TG-MS
in the investigation of claysQuantitative determination of
H2O, CO and CO2 by evolved gas analysis with a mass spec
trometer: Thermochimica Acta, v. 20, p. 387393.
Northrop, H.R. and Whitney, G., 1987, Clays and associated min
erals in the Alum Mountain fossil geothermal system near Gila
Hot Springs, Grant County, New Mexico, in An Excursion to
Selected Zeolite and Clay Deposits in Southwestern New Mex
ico and Eastern Arizona: Brockport, New York, International
Committee on Natural Zeolites, p. 112.
Sposito, G., 1984, The Surface Chemistry of Soils: New York, Ox
ford University Press, 234 p.
Sposito, G., and Prost, R., 1982, Structure of water adsorbed on
smectites: Chemical Reviews, v. 82, p. 553573.
Wicks, F.J., and Ramik, R.A., 1990, Vacuum thermogravimetric
analysis and evolved gas analysis by mass spectroscopy, in
J.W. Stucki, D.L. Bish, and F.A. Mumpton, eds., Thermal
Analysis in Clay Science: Boulder, Colorado, The Clay Miner
als Society Workshop Lectures, v. 3, p. 160189.
Yariv, S., and Cross, H., 1979, Geochemistry of Colloid Systems:
New York, Springer-Verlag, 450 p.
 CHAPTER N
MAPPING MINERALS WITH
IMAGING SPECTROSCOPY
By ROGER N. CLARK,' GREGG A. SWAYZE,' and ANDREA GALLAGHER"
Imaging spectroscopy is a new mapping tool and rep
resents the next generation in remote-sensing technology.
The narrow spectral channels of an imaging spectrometer
form a continuous reflectance spectrum of the Earth's sur
facethis contrasts with the 4 to 7 channels of the previ
ous generation of imaging instruments, like the Landsat
Thematic Mapper (TM) and Multispectral Scanner (MSS)
instruments. Although systems like Landsat can distin
guish general brightness and slope differences in the
reflectance spectrum of the surface, imaging spectroscopy
not only does that but also resolves absorption bands in
the spectrum that can be used to identify specific species.
Spectroscopic analysis of imaging-spectroscopy data
allows any material (mineral, vegetation, man-made,
water, snow, etc.) with unique absorption features in the
measured spectral region to be mapped.
NASA is now flying the Advanced Visual and Infra
Red Imaging Spectrometer (AVIRIS) instrument.
AVIRIS acquires data in the spectral range from 0.4 to
2.45 m in 224 spectral channels. The instrument is flown
in an ER-2 aircraft (a modified U-2 spy plane) at an alti
tude of 19,800 m (65,000 ft). The ground resolution is 20
m; the swath width is about 11 km (614 pixels); and the
swath length can be up to about 1,000 km. After initial
poor performance in 1987 and 1989, the AVIRIS instru
ment now produces superb signal-to-noise data.
In 1989, we developed a new analysis algorithm that
uses a digital spectral library of known materials and a
fast, modified-least-squares method of determining if a
single spectral feature for a given material is present
(Clark and others, 1990). We have made a major advance
in the mapping algorithm: now multiple minerals using
multiple spectral features are mapped simultaneously.
This is done by a modified-least-squares fit of spectral
"U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
features from data in our digital spectral library to corre
sponding spectral features in the image data. The algo
rithm does not force a detection as do many other
algorithms in use. For example, many algorithms take a
set of curves and best fit them to the observed data, often
requiring a set of parameters (like mineral fraction) to sum
to one. Our algorithm only produces values indicating the
presence of those minerals we choose to map. If the min
erals do not exist in that area, the algorithm produces
zeros, indicating they are not detected.
Our mapping algorithm produces, for each pixel in
the image, a spectral feature depth (correlated to abun
dance) and a fit number (least-squares correlation coeffi
cient) to the reference spectra (giving a measure of
confidence in the result) for each mineral mapped. The
depth values for each pixel and each mineral form a set of
images of the minerals correlating to abundance. The fit
values form a set of images corresponding to the confi
dence level of the identification. We combine single-min
eral fit and depth images from several minerals, assigning
a color to each mineral map. In this way, we produce
multi-mineral maps. For example, red might be assigned
to hematite; shades of red (from bright red to dark red)
show relative abundance of hematite, brighter red indicat
ing a stronger spectral signature. In these maps of miner
als, black indicates none of the given set of minerals were
detected at that location.
We have used the algorithm on AVIRIS data of
Cuprite, Nevada, to illustrate some of the mapping possi
bilities with the new generation of sensors. The geologic
and alteration maps are shown in figures 1 and 2. A false
color image of Cuprite (like one that might be produced
by broad-band remote-sensing instruments) is shown in
image A (all image figures are shown at the end of this
chapter). An example minerals map is shown for iron
bearing minerals (image B). A color mineral map of clays
and sulfates is shown in image C.
 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

QAL
T
ALA uvial f Arus And
STREAM CHANNEL of Posits

OP

GRAVELS AND
Ou DER ALLuvial FANS

TF
[ ]
POPPH Y RiT C FEL si TE Dix ES

Tr

1 kilometer
|
;. T HiRST Y CAN YOn TuFF
(SPEARHEAD AND TRAIL RIDGE MEMBERS)

TB
[T]
: ; TS
BAS ALT

[T]
sie BERT TUFF

; RHY OLTE OR QuART 2 LA Tit E

As H-FLOw TUFF

ce
i#

EMIGRANT FORMATION

CMS

i :i Mule spring LiMEST on E

cH f
N, : HARKLE'ss Formation

2-N.
Map Location
|
-

\
NEVADA

Figure 1. Geologic map of the Cuprite, Nevada mining district. The map was produced by conventional field work combined with remote
sensing TM data. From Abrams and Ashley (1980) and R.P. Ashley and R.C. Evarts (unpub. data, 1976) as modified by Hook (1990).

We can detect very subtle spectral differences, like
degrees of kaolinite crystallinity (see image D), the differ
ence between Na-montmorillonite versus Ca-montmorillo
nite (see image C), and individual members of the Na-K
alunite solid-solution series (images E1 and E2). At
Cuprite, maps of these subtle differences depict the alter
ation zone remarkably well. For example, poorly crystal
line kaolinite or halloysite seems to be a general
weathering product that occurs throughout the image, but
as one moves closer to alteration zones, the kaolinite
becomes progressively more crystalline. The highly crys
talline kaolinite occurs just outside of the alunite zones.
Alunite zones can be subdivided spectrally into areas
where K-alunite occurs at the center and is surrounded by
Na-alunite (see images E1 and E2).
These mineral maps have the potential to be extraor
dinary tools for mineral exploration because they show
variations in mineral chemistry and, hence, pressure, tem
perature, and chemcal gradients in areal detail never seen
before. Coupled with further geochemical research, detec
tion of spectral variations in mineral solid solution series
may provide a means to map temperature gradients on a
large scale in a matter of hours. This information can also
be used to locate areas where critical relationships need
investigation. The applications seem boundless.
We have developed our methodology of calibration
and mapping analysis to be a near-routine method. Imag
ing spectroscopy could now be used in U.S. Geological
Survey (USGS) projects, and we believe the results illus
trated here show that imaging spectroscopy could greatly
enhance mineral mapping by the USGS. We also feel the
method could be used for environmental problems because
the spectral mapping algorithm will work on any material
having diagnostic spectral absorption or emission features.
Imaging spectroscopy could also be used in laboratory
analysis of hand samples or in the field for investigating
small areas.
 MAPPING MINERALS WITH IMAGING SPECTROSCOPY 143

37030'

ALTERATION MAP OF
CUPRITE, NEVADA
SILICIFIED

& OPALIZED
E ARGILLIZED
[ ] UNALTERED
MILES

117 o | COTTONTAIL o KILOMETER

t c- r-- n
15' 2. RANCH
NEVADA

Figure 2. Alteration map of the Cuprite, Nevada mining district. The map was produced by conventional field work combined with
remote-sensing TM data. From Abrams and Ashley (1980) and R.P. Ashley and R.C. Evarts (unpub. data, 1976) as modified by Hook
(1990).
144 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
REFERENCES CITED
Abrams, M.J., and Ashley, R.P., 1980, Alteration mapping using
multispectral imagesCuprite Mining district, Esmeralda
County, Nevada: U.S. Geological Survey Open-File Report
80367, 17 p.
Clark, R.N., Gallagher, A.J., and Swayze, G.A., 1990, Material ab-
sorption band depth mapping of imaging spectrometer data
using a complete band shape least-squares fit with library ref
erence spectra: Proceedings of the 2nd Airborne Visible Infra
red Imaging Spectrometer (AVIRIS) Workshop, Jet
Propulsion Laboratory, Publication 9054, p. 176186.
Hook, S.J., 1990, The combined use of multispectral remotely
sensed data from the short wave infrared (SWIR) and thermal
infrared (TIR) for lithological mapping and mineral explora
tion: Fifth Australasian Remote Sensing Conference, Proceed
ings, Oct., 1990, v. 1, p. 371-380.
 MAPPING MINERALS WITH IMAGING SPECTROSCOPY 145

1170 10

37023

F
2 KILOMETERS

Image A. A color infrared image of Cuprite, Nevada from AVIRIS data. North is up in this 11-km-wide by 14-km-long scene. Ground
resolution is 20 m. The linear feature running north-south to the right of center in the image is Highway 95. The area to the east of the road
is the well-studied Cuprite mining district. This area consists of hydrothermally altered volcanic rocks and contains an intensely altered
central silica cap surrounded by less altered zones of opalized and argillized rock. The area west of the highway consists of altered volcanic
rocks and Cambrian siltstones and limestones. It also contains silicified, opalized, and argillized zones. Altered siltstones comprise most
of the altered rocks in the northern part of the Cuprite Hills.
146 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

1170 10

37023

|--
2 Kll OMETERS

Image B. A color mineral map of iron-bearing minerals for Cuprite, Nevada. Hematite is red and is restricted to the opalized and argillized
volcanic rocks on both sides of the highway. Goethite is green and is probably a weathering product (it is a common soil component).
Jarosite is blue and occurs in the opalized zones on both sides of the highway; the largest occurrences are west of the highway.
 MAPPING MINERALS WITH IMAGING SPECTROSCOPY 147

1170 10

Image C. A color mineral map of clays and sulfates for Cuprite, Nevada. Alunite is red and occurs in opalized and argillic zones on both
sides of the highway. Dickite is orange and is closely associated with kaolinite in the altered zones. Kaolinite (well crystallized) is yellow.
Kaolinite (medium crystallized) is yellow green. Kaolinite (poorly crystallized) is green. (Halloysite and poorly crystallized kaolinite are
spectrally indistinguishable at 2.2 p.m.) Ca-montmorillonite is light blue and occurs in the northeastern portion of the scene. Na-montmoril
lonite is blue and occurs in rock units and as loess accumulations on alluvial fans and in playas. Some muscovites have a similar spectral
signature and are also mapped as montmorillonite (with further research, these minerals may be separated). Buddingtonite is purple and is
located only in a few pixels east of the highway. Buddingtonite can be better seen in white on the expanded image E2. Paragonite is magenta
and occurs mostly in the lower left center of the image. Chlorite occurs as an intimate mixture with the paragonite. Opalized tuff is white
and occurs in the lower left corner as well as in the central region of the Cuprite alteration zone.
148 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

1170 10.

3723'

Image D. A color mineral map of kaolinite crystallinity for Cuprite, Nevada. Well crystallized kaolinite is blue and occurs near the outer
margins of the alunite areas (which are black in this image). Medium crystallized kaolinite is green and occurs between the highly crys
tallized and poorly crystallized kaolinite. Poorly crystallized kaolinite is red and occurs farthest from the alunite areas. Poorly crystallized
kaolinite or halloysite may be due to weathering processes. (Halloysite and poorly crystallized kaolinite are spectrally indistinguishable
at 2.2 pum.)
 MAPPING MINERALS WITH IMAGING SPECTROSCOPY 149

3723'

Image E1. A color mineral map of the alunite-natroalunite solid solution series, ammonium minerals, and two clays for Cuprite, Nevada.
Potassium alunite is green and occurs in opalized and argillic zones on both sides of the highway. Natroalunite (Na mole fraction 0.65) is
magenta and occurs next to potassium alunite, mostly to the west of the highway. Natroalunite (Na mole fraction 0.80) is blue and occurs
with other natroalunites, mostly to the west of the highway. Buddingtonite is white and is located in a few pixels east of the highway and in
a few small areas west of the highway. Ammonium illite-smectite is yellow and occurs in the alteration zones near alunite and buddingtonite.
Kaolinite (well crystallized) is red and occurs near alunites in the alteration zones. Na-montmorillonite is cyan and occurs in rock units and
loess accumulations on alluvial fans and in playas.
 150 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
1170 10

3705.
23 ER.

Image E2. An enlarged color mineral map of the alteration zones. This is an enlargement of the east-central part of image El and includes
the alunite-natroalunite solid solution series, ammonium minerals, and two clays for Cuprite, Nevada. Potassium alunite is green and occurs
in opalized and argillic zones on both sides of the highway. Natroalunite (Na mole fraction 0.65) is magenta and occurs next to potassium
alunite, mostly to the west of the highway. Note how this natroalunite surrounds many of the potassium alunite zones (in green). Natroalunite
(Na mole fraction 0.80) is blue and occurs with other natroalunites, mostly to the west of the highway. Buddingtonite is white and is located
in a few pixels east of the highway and in a few small areas west of the highway. Note that buddingtonite is surrounded by ammonium illite
smectite. A few buddingtonite localities occur west of the highway. Ammonium illite-smectite is yellow and occurs in the alteration zones
near alunite and buddingtonite. Note how the ammonium illite-smectite occurs at the edges of potassium alunite (green) zones. Kaolinite
(well crystallized) is red and occurs near alunites in the alteration zones. Na-montmorillonite is cyan and occurs in rock units and loess ac
cumulations on alluvial fans and in playas.
 CHAPTER O
SATELLITE IMAGE PROCESSING FOR
ENHANCED SPECTRAL DISCRIMINATION
AND INTERPRETABILITY
By DANIEL H. KNEPPER, JR.'
BACKGROUND
Landsat Thematic Mapper (TM) satellite digital
image data have been successfully applied to mapping
hydrothermally altered rocks, lithologic units, and geo
logic structures in numerous CUSMAP, preassessment,
and wilderness studies; the data have also been used by
several U.S. Geological Survey (USGS) cooperative, inter
national, mineral resource assessment projects. For the
immediate future, at least until the end of this decade, TM
data will be the only digital satellite data with worldwide
coverage and spectral bands for detecting minerals with
absorption features in the 2.22.3-lim region (mostly
hydroxyl-bearing clay minerals and carbonates). Conse
quently, it is in the interest of the USGS to explore new
methods of processing TM digital data to optimize the
capability for extracting geologic information and to
increase the interpretability of the digital-image products.
Color-ratio composite (CRC) images prepared from
ratios of TM bands are commonly used to map areas of
hydrothermally altered rocks and to discriminate rock
units based on the presence of clay, carbonate, and iron
oxide minerals (Knepper, 1989; see also selected bibliog
raphy in Knepper, 1989). Two new image-processing tech
niques for preparing more useful CRC images were tested
in the Geophysics Environmental and Minerals Demon
stration Area (Watson and others, 1989), centered around
Canon City in south-central Colorado. The two techniques
are used in combination with standard processing proce
dures to increase the contrast between spectrally dissimilar
materials and to display the TM band ratios in a more
interpretable form. Spectral contrast is enhanced by a
'U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
process called iterative ratioing (Crippen, in press a).
Improved image presentation is obtained by combining
three spectral components (band ratios) with a topographic
component (average image) in a four-component display
(Crippen, in press b) of the CRC image data. Encouraging
results from this study prompted the application of the
methods to additional test areas in the Altiplano of
Bolivia, the Wind River Basin, Wyoming, and southern
California with equally successful results.
ITERATIVE RATIOING
The signal reaching the TM sensor is composed of
light that is reflected from the ground surface and light
that is scattered by the atmosphere. Iterative ratioing is a
visually guided technique for removing the atmospheric
scattering component from the signal measured by the
TM sensor in each band. When band ratios are prepared,
the iterative ratio correction provides a better measure of
the actual spectral contrast between different materials
and is based on the TM data alone without the need to
apply complicated atmospheric models or to collect
ground measurements.
Conceptually, iterative ratioing is a process of testing
estimated atmospheric correction factors by visually evalu
ating the degree of topographic expression that remains in
a test-band-ratio image after applying an estimated correc
tion factor. Spectral-band-ratio values of a surface should
be independent of the slope of the surface relative to solar
illumination direction. However, in low illumination con
ditions, such as shadows, the dominant signal detected by
the TM sensor is an additive term due to atmospheric scat
tering. Consequently, ratios of uncorrected bands in these
shadowed areas are not the same as in full illumination,
and they express an aspect of topography. The degree of
152 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

topography remaining in a test ratio is a measure of how

well the estimated atmospheric correction factor removes
the scattered-light component from the band data. If topo
graphic expression remains, then the correction constant is
too small; if the topographic expression is reversed (shad
ows are white), then the correction factor is too large.
Operationally, the test-ratio image is formed using
one TM band for which the atmospheric correction factor
is known (numerator) and one band for which the correc
tion factor is to be determined (denominator). TM band
7 is the longest wavelength band of the TM data
(2.082.35 pum) and is the least affected by atmospheric
scattering. Removing an atmospheric correction factor of
2 data numbers (DN) from TM band 7 provides good
correction for most TM scenes. Correction factors for
any of the remaining TM bands can be determined by:
(1) applying an estimated correction factor to the band,
(2) forming a ratio image between the corrected TM band
7 (numerator) and the band with the estimated correction
factor applied (denominator), (3) evaluating the degree of
topography remaining in the test-ratio image, and (4)
adjusting the correction factor and repeating steps (2), Figure 1. Standard Landsat TM color-ratio composite image of
(3), and (4) until the amount of topography in the test the Canon City area, Colorado, made without atmospherically cor
recting the band-image data. Image has been contrast stretched for
ratio image is minimized.
display. Image area is approximately 10 miles across.

FOUR-COMPONENT DISPLAY mathematical relationship that allows the mean adjust

Atmospherically corrected TM band-ratio images dis
play a minimum of topographic information, and color
ratio composite (CRC) images prepared from three atmo
spherically corrected band ratios are often difficult to
interpret in the appropriate geologic and geomorphic con
text. In addition, potentially important spectral anomalies
are difficult or impossible to accurately locate on maps.
Selected amounts of topographic information can be
merged into the CRC image by multiplying each of the
band-ratio images by an image containing suitable topo
graphic expression. An average image composed of the
average of two or more single bands works well. The
average image can be edge enhanced to increase the
sharpness of tonal boundaries reflecting shadowing before
merging with the band ratios. Alternatively, the spectral
information from the TM band ratios can be merged with
higher spatial resolution images, such as SPOT or digital
radar image data, after resampling the TM band ratios to
the same pixel size as the higher resolution data and regis
tering the two data sets.
By carefully adjusting the means of the band ratios
and the average image before merging, a CRC image can
be prepared that provides a topographic reference for the
spectral information in the band ratios without signifi
cantly affecting the hues of the colors depicting the spec
tral information. Crippen (in press b) provides a simple
ments to be made in a straightforward and consistent man
ner. The resulting CRC image can be visually interpreted
and confidently used for photogeologic mapping.
DISCUSSION
Figure 1 is a standard TM CRC image centered on
the Canon City, Colorado, areathis area lies in the cen
ter of the south-plunging (top to bottom) Oil Creek syn
cline, which is outlined by prominent hogbacks of
Cretaceous sedimentary strata. Some topographic expres
sion is apparent in the image because no atmospheric
corrections were applied to the bands of TM data, but
there is an overall subduing of most topographic features.
The colors on the image indicate compositional variations
of surface materials: red represents vegetation, magenta
represents materials containing clays and (or) carbonate
minerals, green represents minerals containing reduced
(ferrous) iron minerals, and blue represents iron-oxide
(ferric) minerals. Figure 2 is a four-component TM CRC
image in which the individual TM bands have been
atmospherically corrected using the iterative ratioing pro
cess. Color hues on the image are basically the same as
on figure 1, although some differences are apparent
because the atmospheric corrections provide a better (and
different) measure of the actual spectral differences
 SATELLITE IMAGE PROCESSING
Figure 2. Four-component Landsat TM color-ratio composite
image of the Canon City area, Colorado, prepared from atmospher
ically corrected bands using iterative ratioing and merged with an
edge-enhanced average image prepared from TM bands 3,4, 5, and
7. Image has been contrast stretched for display. Image area is ap
proximately 10 miles across.
between materials. The addition of the topographic com
ponent to the image clearly enhances the interpretability
of TM CRC data.
The software necessary to produce atmospherically
corrected, four-component Landsat TM CRC images is
relatively simple and should be available in most image
153
processing packages. The functions that are necessary are:
(1) adding, subtracting, multiplying, and dividing images
and constants to images, (2) contrast stretching, (3) image
statistics (mean and standard deviation), and (4) edge
enhancement (although this can be accomplished with a
simple boxfilter program and the arithmetic functions
listed in (1) above). Programs contained in REMAPP-PC
(Livo and Gallagher, 1991) were used to prepare figures 1
and 2. Details of the image-processing procedures are con
tained in Crippen (in press a, in press b).
REFERENCES CITED
Crippen, R.E., in press a, The iterative ratioing method of determin
ing atmospheric corrections for scenes with rugged terrain:
Photogrammetric Engineering and Remote Sensing.
in press b, Image display of four components of spectral
data: Remote Sensing of Environment.
Knepper, Daniel H., Jr., 1989, Mapping hydrothermal alteration
with Landsat Thematic Mapper data, in Lee, Keenan, ed., Re
mote Sensing in Exploration Geology: 28th International Geo
logical Congress Field Trip Guidebook T-182, p. 1321.
Livo, K.E., and Gallagher, A.J., 1991, REMAPP-PCRemote
sensing image processing software for MS-DOS personal com
puters, Version 2.00: U.S. Geological Survey Open-File Re
port 91449, 81 p.
Watson, Kenneth, Knepper, D.H., Jr., Clark, R.C., Pitkin, J.A.,
Labson, V.F., Campbell, D.L., Rowan, L.C., Kruse, F.A.,
Lee, Keenan, and Livo, K.E., 1989, Plans for integrated air
borne geophysical study of the Geophysics Environmental
and Minerals Demonstration Area, south-central Colorado
[abs.]: USGS Research on Mineral Resources1989, 5th An
nual V.E. McKelvey Forum on Mineral and Energy Resourc
es, Program with Abstracts, U.S. Geological Survey Circular
1035, p. 7778.
 CHAPTER P
40Ar/39Ar STUDIES OF FLUID INCLUSIONS IN
VEIN QUARTZ FROM BATTLE MOUNTAIN,
NEVADA
By EDWIN H. MCKEE," JAMES E. CONRAD,' BRENTD. TURRIN,"
and TED G. THEODORE!
ABSTRACT
Fluid inclusions in vein quartz from the Buckingham
stockwork molybdenum deposit and from leucogranites
north of the Buckingham deposit yield 40Ar/39Ar isochron
ages that agree well with K-Ar and 40Ar/39Ar radiometric
ages of phenocrysts from these rocks. (The Buckingham
deposit is Late Cretaceous in age, and the leucogranites
north of the Buckingham deposit are late Eocene to early
Oligocene in age.) The younger period of igneous activity
(late Eocene to early Oligocene) is clearly defined by
three isochrons at about 38 Ma. The Late Cretaceous age
is recognized by three isochrons, one well defined at
75.5+1.1 Ma, one that is 55.1+3.7 Ma, and one that is
51.2+16.3 Ma. All samples had inherited excess 40Ar in
amounts that equal about 30 percent of the total argon.
Because of the excess argon, individual age determina
tions are highly variable and cannot be interpretediso
chrons based on several analyses are necessary to yield a
meaningful age. Samples with small amounts of potas
sium proved unusable; sufficient potassium must be
present to produce enough radiogenic 40Ar to mask the
excess 40Ar.
INTRODUCTION
Igneous activity and associated mineralization in the
Battle Mountain district, Nevada (fig. 1), occurred during
the Late Cretaceous and again in late Eocene to early
"U.S. Geological Survey, Mail Stop 901, 345 Middlefield Road,
Menlo Park, CA 94025
Oligocene time. These age designations are based on
many K-Ar and 40Ar/39Ar age determinations (Theodore
and others, 1973; McKee, 1992). The purpose of this
study was to determine the radiometric ages of quartz
veins from the Cretaceous Buckingham stockwork molyb
denum deposit and nearby Eocene and Oligocene quartz
veins in this district using single-grain laser-fusion 40Ar/
39Ar methods. The fluid-inclusion-derived ages from this
well-dated district are compared with existing K-Ar and
40Ar/39Ar dates to test the feasibility of obtaining usable
geochronology from fluid inclusions. Vein quartz can be
dated using argon produced by radioactive decay of potas
sium-bearing phases in fluid inclusions. All argon from
quartz should be derived from fluid inclusions because
there is no other site possible for the large K atom in the
SiO2 structure of quartz. Argon is present as a component
of paleohydrothermal fluids trapped in fluid inclusions. As
the data were gathered, however, it became evident that
excess argon is present within the vein quartz. Only by
constructing isochrons from a number of individual quartz
grains is it possible to evaluate the amount of excess argon
present. The ages determined by this technique confirm
those derived by K-Ar and 40Ar/39Ar total fusion and
incremental heating of potassium-bearing minerals from
the Buckingham deposit (McKee, 1992).
The presence of argon in fluid inclusions in quartz
was first demonstrated by Wahler (1956). Subsequently,
excess Ar and He in beryl, cordierite, and tourmaline,
which were assumed to come from fluid inclusions, was
reported by Damon and Kulp (1958). The presence of
Ar, Kr, and Xe in gas from gas inclusions in sylvite from
the Bereznikovsk mine in Russia was reported by Nes
melova (1959). Excess 40Ar in several quartz and fluo
rite samples was documented by Rama and others (1965),
and they pointed out the difficulties that this presents for
156 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

11730' 1179 for a period of 60 h. Irradiation flux was monitored by

41
use of the MMhb-1 hornblende standard, which has an age
HUMBOLDT COUNTY
of 513.9 Ma. Sample handling techniques and corrections
for Ca- and K-derived isotopes used in the Menlo Park
laboratory are as described by Dalrymple and Lanphere
BATTLE
MOUNTAIN
| (1971, 1974). Samples 1471760, 1471-173, 79C77, and
81TT121 were irradiated in the Berkeley TRIGA reactor
DISTRICT . (University of California, Berkeley) for a period of 10 h.
Trenton Canyon
Mo-Cu prospect
Moru pros
| BATTLE
: MOUNTAIN
Irradiation flux was monitored by use of standards MMhb
1 (hornblende), GHC 305 (biotite, with an age of 103.76
O
Ma), and P-207 (muscovite, with an age of 82.1 Ma).
Copper Basin Potassium and calcium corrections for the Berkeley
Cu-Au deposits
TRIGA reactor were determined using optical grade CaF2
Buckingham Mo deposit
and a laboratory potassium glass.
Galena Pb-Zn-Ag deposits
The 40Ar/39Ar analyses were conducted at the U.S.
/ -
Copper Canyon
Cu-Au deposits Geological Survey, Menlo Park, California, and at the
Berkeley Geochronology Center using similar laser-fusion
microextraction systems to analyze single grains of vein
quartz that were approximately 0.51.0 mm in size. An
on-line, ultrasensitive mass spectrometer was used for
argon analyses.

LANDER COUNTY
40 l
Figure 1. Map of north-central Nevada, showing Battle
Mountain district and mineral deposits within it.
K-Ar dating. Little work has been published on K-Ar or
40Ar/39Ar dating of fluid inclusions since these studies.
Recently, Kelly and others (1986) examined, by
incremental heating 40Ar/39Ar techniques, the Ar isotopes
of fluid inclusions in quartz veins associated with granite
hosted tungsten mineralization in southwestern and north
ern England. In this study, the authors noted that argon in
the fluid inclusions is released during incremental heating
in amounts that correlate with the potassium and chlorine
content of fluid inclusions. The argon in the potassium
correlated component is derived primarily from in situ
decay of potassium in solid phases in the inclusions.
Argon from the chlorine-correlated component is associ
ated with brines that constitute the liquid phase of the
inclusions. Inherited argon from the original magmatic
fluid was not noted.
METHODS
Samples 79C103 and 79C104 were irradiated in the
U.S. Geological Survey TRIGA reactor (Denver, Colo.)
GEOLOGIC SUMMARY
The Buckingham stockwork molybdenum deposit in
north-central Nevada, about 10 km southwest of the town
of Battle Mountain (fig. 1), contains one of the largest
known resources of molybdenum in the United States.
However, under present (1991) economic conditions, the
relatively low molybdenum grades of the deposit make it
uneconomic. The Buckingham deposit is a calc-alkaline,
stockwork-molybdenum-bearing system located on the
east flank of the Battle Mountain mining district (fig. 1).
The deposit contains an estimated 1,000 million tons of
mineralized rock, averaging approximately 0.10 percent
molybdenum (as MoS2). The deposit also contains small
amounts of silver, tungsten, copper, and gold (Theodore
and others, 1992). Molybdenum mineralization is related
to emplacement of a Late Cretaceous composite porphyry
system that intruded surrounding Paleozoic rocks and
altered much of the country rock to hornfels.
Two periods of plutonism have been recognized in
the Battle Mountain area (Theodore and others, 1992).
The older period, represented by monzonites at Bucking
ham and the granodiorite of Trenton Canyon (on the west
edge of the mining district and about 9.5 km due west of
the Buckingham system), has been dated at about 86 Ma
(McKee, 1992) (table 1, fig. 2). The younger period
includes at least six Tertiary intrusive bodies in the
Buckingham vicinity, ranging in age from 41.1 to 37.3 Ma
(Theodore and others, 1973) (table 1, fig. 2).
 *ArAr STUDIES, BATTLE MOUNTAIN, NEVADA 157

Table 1. K-Ar and 40Ar/39Ar ages of Buckingham stockwork umbrella-shaped shells of molybdenite mineralization that
system and nearby intrusive rocks. drape over the stocks (Theodore and others, 1992).
(o, standard deviation. Modified from McKee (1992)]
Approximately one-half of the Buckingham deposit is
hosted in metamorphosed and intensely veined rocks
Rock Mineral Apparent age belonging to the Upper Cambrian Harmony Formation.
type dated (Mato) Secondary, partially chloritized biotite is pervasive
Samples from the Buckingham stockwork system throughout the Buckingham porphyry and yields a K-Ar
Monzogranite porphyry Muscovite 61.3 + 1.5
age significantly younger (by 10 to 15 percent) than the
Monzogranite porphyry Muscovite 61.7+ 1.5 86-Ma age that is considered to be the age of initial cool
Monzogranite porphyry Biotite and chlorite 65.1 + 1.6 ing of the pluton. This younger age is interpreted to result
Monzogranite porphyry Muscovite 68.6 + 1.5 from partial loss of argon from the 86-Ma porphyry
Aplite Whole rock 70.3 + 1.7
(McKee, 1992).
Monzogranite porphyry Muscovite 75.7+ 1.6
Monzogranite porphyry Muscovite 77.4 + 1.6 The argon release pattern from the incremental-heat
Monzogranite porphyry Muscovite 85.5 + 1.9 ing 40Ar/39Ar experiment support the partial-argon-loss
Monzogranite porphyry Biotite 85.7+0.4! model, indicating a complex thermal history. The second
Monzogranite porphyry Muscovite 86.1 + 2.0 ary thermal event, although not pervasive enough to com
Monzogranite porphyry Muscovite 88.0+2.0
pletely reset the K-Ar clock, is clearly recognized in the
Sampies of intrusive rocks north and southwest of
release pattern (fig. 3). Heat and hydrothermal activity
the Buckingham stockwork system
Rhyolite Biotite 37.3 + 1.1
generated by the late Eocene and early Oligocene igneous
Leucotonalite Hornblende 37.7+ 1.4 events show varying amounts of intensity at different
Monzogranite Biotite 38.8 + 1.1 places throughout the Buckingham system. Almost all
Monzogranite Hornblende 39.0 + 1.1 samples from the western part of the system yield ages
Monzogranite Hornblende 39.3+ 1.0 that are variably reset, although none of them are com
Rhyolite Biotite 39.1 + 1.0
pletely reset to the approximate 38-Ma age of the younger,
Sample of intrusive rock northeast of and cutting nearby plutons.
the Buckingham stockwork system
Granodiorite porphyry Hornblende 35.4 + 1.1

1 Age from the two greatest release steps from an incremental TERTIARY GRANITIC AND RHYOLITIC
heating 40Ar/39Ar experiment.
INTRUSIVE ROCKS

The six small monzogranite, granodiorite porphyry,

CRETACEOUS MONZOGRANITES OF THE
BUCKINGHAM AREA
The age of emplacement of the Buckingham system
is based on evaluation of one incremental-heating 40Ar/
39Ar determination and 10 K-Ar age determinations
(McKee, 1992) (table 1). The ages range from 88.0+2.0
to 61.3+1.5 Ma; four of the K-Ar ages and the major
argon release steps of the incremental heating experiment
cluster around 86 Ma. This age (86 Ma) is considered to
represent the initial time of cooling of the porphyry
system.
Eight major intrusive phases identified in the Buck
ingham system were intruded sequentially from south to
north across a distance of several kilometers. The present
configuration of the intrusive center is an east-west align
ment of two stocks (each a composite of several porphy
ries) and several outlying intrusive bodies (fig. 2). All
eight intrusive phases underwent molybdenum mineraliza
tion. The two oldest intrusions form border phases, and
the youngest, an outlying intrusion, produced relatively lit
tle molybdenum mineralization. The main Buckingham
molybdenum deposit, formed by five igneous phases asso
ciated with the main part of the two stocks, developed
and rhyolite stocks and dikes in the vicinity of the Bucking
ham deposit (fig. 2) yield K-Ar ages on biotite and horn
blende of about 38 Ma (table 1). These bodies are mostly
satellites of the main Cretaceous monzogranite porphyry of
the Buckingham deposit. A system of granodiorite por
phyry dikes, dated at about 35 Ma, cut the Buckingham
porphyry (table 1, fig. 2). Hydrothermal alteration and min
eralization related to late phases of this pervasive middle
Tertiary period of igneous activity are recognized south of
the Buckingham deposit, at Copper Canyon, and at many
other places in the Battle Mountain mining district. The
thermal effect of late Eocene and early Oligocene intrusive
events manifests itself in the Cretaceous Buckingham sys
tem as partially reset K-Ar ages.
STRUCTURE
Deformation of the Buckingham system during the
middle Tertiary included significant brittle-type faulting
that is manifested primarily along three major, low-angle,
normal faults. This Tertiary extensional tectonism pro
gressed from east to west during Oligocene and possibly
158 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

81TT121
(38 Ma)

| Q 6.5
tt-d
+ Kit CMETER

EXPLANATION

[T] Quaternary alluvium

Tertiary igneous rocks
Cretaceous Buckingham stockwork systemincludes monzogranite (black)

[...] Paleozoic sedimentary and metasedimentary rocks

38.8 Ma
K-Ar or *ArAr sample localityshowing age determinations from McKee (1992)
75.5 Ma
[] *ArAr quartz vein, fluid-inclusion sample site and age
Faultdotted where concealed

Figure 2. Geologic map of the Buckingham stockwork molybdenum deposit and surrounding area. Sample sites for this report
are represented by squares; K-Ar sample sites and one 40Ar/39Ar incremental-heating-age sample site represented by circles.
 *ArAr STUDIES, BATTLE MOUNTAIN, NEVADA 159

120

100HH-85.7+0.4 Ma
FUSE

1015C

60 H.

] 525

40
O 10 20 30 40 50 60 70 80 90 100

39Ar RELEASED, IN CUMULATIVE PERCENT

Figure 3. 40Ar/39Ar age spectra for a biotite from the Buckingham stockwork system (monzogranite porphyry). The six heating steps
(and thus temperature in C) are shown as horizontal bars. This release pattern is similar to patterns produced from minerals with disturbed
thermal histories. The most likely age calculated using the two oldest steps (which produced more than 50 percent of the Ar) is 85.7+0.4
Ma (from McKee, 1992).

Miocene time (Theodore and others, 1992). Imbricate,
down-to-the-east, listric-normal faults merge into a single,
flat-lying dislocation east of the Buckingham porphyry
system. In addition, major reverse dislocations in the
hanging wall of the major fault probably reflect develop
ment of antithetic reverse faults in the toe region of a
rigid-body glide block (Theodore and others, 1992).
FLUID INCLUSIONS
Hydrothermal fluids in the Buckingham system show
complex histories that reflect successive influxes of fluids
from newly injected magma into loci of earlier emplaced
igneous phases (Theodore and others, 1992). In distal
parts of the system, fluid inclusions in andradite-diopside
suggest temperatures of about 300C at extremely low
mole fractions of CO2. In proximal parts of the system,
where the bulk of the quartz stockworks and molybdenite
was deposited, fluids were relatively low-density, low
salinity NaCl brines that included variable amounts of
CO2. These early-stage fluids established an ascending
plume of fluid that was considerably less saline than flu
ids associated with most porphyry copper systems in the
southwestern United States. The magmas at the Bucking
ham deposit did not evolve large volumes of very dense,
NaCl-rich fluids. High in the system, very sparse halite
bearing fluid inclusions formed through condensation
from an earlier, higher temperature saline fluid. In the
western part of the system, the high-temperature and
high-saline halite-bearing fluid inclusions, having salini
ties as much as 65 weight percent NaCl equivalent and
homogenization temperatures as high as 500C, probably
reflect ponded magmatic fluids from a deep source. Over
all distribution of veins associated with such fluids
appears to be restricted. Although a fluid-immiscibility
model is probably applicable to some porphyry molybde
num systems where the exposed parts of the system are
above a buried concentration of molybdenum, the geome
try of the molybdenite-enriched shells and their proximity
to the igneous source at the Buckingham deposit argue
against this model. These shells are tightly constrained to
the flanks and apical parts of the two mineralized stocks,
and the shells show a genetic relationship to a quartzpo
tassium feldspar porphyry phase in the core of the stocks.
Furthermore, an approximately 1-km-long composite ver
tical section through the Buckingham system is available
for study in the three tectonically offset blocks, and
nowhere is there widespread occurrence of hypersaline
fluids. The bulk of the moderately saline, H2O-NaCl-CO2
mineralizing fluids associated with this deposit is related
to quartzpotassium feldspar porphyry and slightly deeper
160 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 2. Temperature and salinity data for vein quartz samples used for fluid inclusion studies.

[Locations of samples used for *ArAr analyses shown on fig. 2; salinity ranges given in weight percent NaCl equivalent, --, not determined]

Homogenization Fluid
Sample no. Host rock" temperature inclusion Salinity range? Comments
range (C)2 types
Outcrop samples
79C77 Km -- I,II, IV - Quartz veins on Vail Ridge associated
with Buckingham system.
81TT121 Km 229324 I, IV 011.6 Quartz veins in upper copper-rich part of
Buckingham system.
78C103 Ch -- I, II, III, IV -- Quartz stockworks near Tertiary
leucogranite.
78C104 Ch -- I, II, III -- Quartz stockworks in pendant within
Tertiary leucogranite.
87TT215A Chg 175550 I, II, III 1.348.5 Quartz stockworks near Tertiary
leucogranite.
87TT215B Chg 250600 I, II, III 3.062 Quartz stockworks near Tertiary
leucogranite.
Samples from diamond drill core
1453259.8 Km 261328 I, II 7.810.8 Drill hole collared in east stock of
Buckingham system.
1471-173 Km -- I, II, IV -- Drill hole collared in Vail block part of
Buckingham system.
1471760 Km 292345 I, II, IV - Drill hole collared in Vail block part of
Buckingham System.
1 Km, Cretaceous monzogranite; Ch, Upper Cambrian Harmony Formation hornfels; Chg, garnet skarn developed in Upper Cambrian
Harmony Formation and related genetically to 38-Ma leucogranite (see text).
* From Theodore and others (1992); Theodore and Hammarstrom (1991).
* At 22+2C: I, low vapor volume, mostly liquid H2O. II, high vapor volume, mostly liquid H2O, III, halite bearing, commonly with other
daughter minerals; IV, mostly CO2 vapor with some liquid CO2, mostly H2O liquid (see text).

equivalent igneous phases. A summary of fluid-inclusion
types recognized in the Buckingham stockwork system is
shown in figure 4, and temperature and salinity data for
the fluid inclusions is in table 2.
In this study, 40Ar/39Ar values from fluid inclusions
were measured on 29 approximately 1-mm grains from
four samples of vein quartz from three localities (includ
ing a drill core) that are associated genetically with the
Cretaceous Buckingham stockwork molybdenum system
(table 2). These four samples were not found to include
any of the type-III fluid inclusions (fig. 4), i.e., the daugh
ter-mineral-bearing variety; nonetheless, it is possible that
some sylvite and (or) other potassium-bearing minerals
may have been present in the quartz fragments selected for
argon analysis because other samples from the system
contain some sylvite-bearing fluid inclusions. In addition,
40Ar/39Ar measurements were made from fragments of
vein quartz associated with the 38-Ma leucogranite. These
samples are from two localities (fig. 2): one vein
(780.104) intrudes the Harmony Formation about 500 m
west of a relatively large body of leucogranite, and the
other (78C103) intrudes the leucogranite. In contrast with
the samples from the Buckingham system, the vein quartz
associated with the 38-Ma leucogranite contains abundant
concentrations of type III, daughter-mineral-bearing fluid
inclusions (fig. 4). Many of these type-III fluid inclusions
contain sylvite (Theodore and others, 1992; Theodore and
Hammarstrom, 1991).
40Ar/39Ar ANALYSIS OF FLUID
INCLUSIONS
40Ar/39Ar analyses on fluid inclusions in quartz from
the Battle Mountain district show a wide range of ages.
Apparent ages of individual quartz grains are from about 40
Ma to about 1,747 Ma (table 3, no. 101-6). These ages are
interpreted to represent the product of inherited excess 40Ar
from magmatic sources trapped in fluid inclusions plus the
component of radiogenic 40Ar derived from the decay of
potassium-bearing fluids and solids trapped in fluid inclu
sions. This interpretation is supported by the association of
the younger ages with a relatively high release of 39Ar and
a corresponding increase in age with lower 39Ar releases.
Fluid inclusions containing essentially no potassium
 *ArAr STUDIES, BATTLE MOUNTAIN, NEVADA 161

LH2O (+ Liquid CO2 at temperatures < 20C) This is similar to the range of K-Ar and 40Ar/39Ar incre
mental heating ages from biotite, muscovite, and whole
Type I -

rock samples from the Buckingham stockwork system

+ Nonmagnetic opaque material(s)
(McKee, 1992) (table 1).
Mostly H2O vapor
Sample 81TT121, which was expected to give a Late
/N Liquid H,o (t Liquid CO2 at Cretaceous age (60 to 76 Ma), yielded a well-constrained
I temperatures < 20C)
I Type II isochron of 37.2+5.3 Ma (fig. 8). This late Eocene or
I + Nonmagnetic opaque material(s)
I
early Oligocene age is the same as many of the small
Mostly H2O vapor igneous bodies in the Battle Mountain district (McKee,
BROADLY 2
CONTINUOUS
WITH ONE Type IV Liquid H2O
- -
1992) (table 1). We assume that this age represents either
ANOTHER Mostly almost complete replacement of the bulk of the fluid in
I CO2 vapor older Cretaceous quartz by middle Tertiary fluid or that
Liquid CO2 the quartz vein is a late Eocene to early Oligocene vein
I Vapor CO2 that is not recognized as such in the Cretaceous Bucking
I
I Liquid H2O ham stockwork system. The former alternative is more
N/ Liquid CO2 (+CH4) likely. The fluid-inclusion population hosted by the quartz
vein in sample 81TT121 differs in many aspects from that
Type III Liquid H2O ostly H2O vapor near the fringes of the system. Firstly, the overwhelming
NaCl bulk of the fluid inclusions, mostly type I (fig. 4), are sec
ondary relative to vein quartz, and they are concentrated in
Cl dense swarms along annealed microcracks oriented at high
Mostly H2O vapor angles to the walls of the veins (Theodore and others,
Pyrite (and (or) arsenopyrite, 1992). These fluid inclusions must have been repeatedly
chalcopyrite, iron oxide)
trapped mostly during circulation of fluids associated with
NaCl
a widespread and locally intense, late-stage, argillic over
+ CaCl i Potassium feldspar or KCl print that has affected these rocks. Nonetheless, almost all
these veins still show relict potassic-alteration assemblages
Figure 4. Types of fluid inclusions recognized in the Bucking
along their selvages that suggest that initial emplacement
ham stockwork molybdenum system (modified from Theodore and of veins occurred during potassic-alteration stages of the
others, 1992). system. Although the overwhelming bulk of the fluid
inclusions in the veined porphyry are the type I variety,
these veins also seem to show a slightly increased abun
therefore give anomalously old 40Ar/39Ar ages, whereas dance of halite-bearing, type-III-variety fluid inclusions
fluid inclusions with high potassium content can produce relative to veins at the fringes of the system. Furthermore,
enough radiogenic 40Ar to almost entirely mask the pres liquid-vapor homogenization tests of fluid inclusions from
ence of the excess 40Ar component. these quartz veins in the monzogranite porphyry yield a
The ages determined from individual grains of sam plot significantly different from that derived from quartz
ples of the Buckingham quartz stockwork are extremely veins that are minimally affected by postpotassic alter
variable and give no indication of the age of the fluid ation. This difference results in a 100C difference
inclusions, assuming that they are Cretaceous as is indi between medians for the two temperature-of-homogeniza
cated by previous studies (McKee, 1992). As noted, sam tion plots (380C for the potassic-alteration assemblage,
ples with very little potassium give anomalously old ages compared with 280C for veins showing intense develop
with analytical precision values as great or greater than ment of the intermediate argillic overprint). Thus, the
the age. For instance, two grains from sample 79C77 (fig. apparent preferred concentration of fluid inclusions that
2) that contain very little potassium yield ages of homogenized to vapor at the high-temperature end of the
1,748+1,306 Ma and 1,314+660 Ma (table 3). Grains with distribution plot for samples of veined porphyry in the
greater amounts of potassium, however, provide points on general area of sample 81TT121 suggests that these high
a 40Ar/36Ar versus 39Ar?6Ar diagram from which an iso temperature fluid inclusions may be largely relict from
chron can be drawn that defines the age of the fluid inclu early potassic stages (Theodore and others, 1992). In
sions in the total sample (broken to produce the individual addition, circulation of fluids during the middle Tertiary in
grains). The isochron also gives a measure of the amount this part of the Cretaceous porphyry system is probably
of excess argon in the sample (intercept on the ordinate related to small bodies of Tertiary intrusive rocks not
with a value greater than 295.5). Samples 1471-760, shown on figure 2.
1471-173, and 79C77 define isochrons (figs. 5, 6, and 7) Of the samples that produced Cretaceous isochron
that are interpreted to be between about 60 and 76 Ma. ages, number 1471-760 (760 ft depth in core hole shown
162 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 3. Summary of single-grain laser-fusion 40Ar/39Ar analyses of quartz from the Battle Mountain district.

[Ages calculated using Ae=0.581x1019 yr", Ap=496210"yr", "K/Kotal=1.167x10"moles/mole. Emors are estimates of the standard deviation (o) of analytical precision]

Sample Experiment 40Ar/ 38A/ 37A/ 36A/ 40Ar rad 40Ar rad/ 40Ar rad Apparent age
no. no. 39Ar 39Ar 39Ar 39Ar (x 1014 moles) 39Ar (percent) (+10) (Ma)
Quartz veins from the Buckingham stockwork system
1471-760 100-1 143.36 (). 17742 1,0008 ().29231 3.1992 57,085 39.81 268.4 + 46.2
(fig. 5) 100-2 542.86 ().65356 0.93087 1.631.4 6.8823 60.881 11.21 284.9 + 18.1
1003 17.762 0.021477 0.0501.99 0.008106 50.853 15.363 86.49 76.24 + 0.59
100-4 184.32 0.54344 0.54672 (). 19733 0.78051 126,07 68.39 547.1 + 113.0
100-5 17,213 ().019658 0.000707 ().005415 21.353 15.604 90.65 77.42 + 0.73
100-6 874.42 1.6889 0.48291 1.4799 0.87615 437.22 49.99 1,445 + 1,022
1007 1183.2 1.5772 0.38085 2.7903 ().91207 358.75 30.32 1,257+ 1,343

1471-1731 114-1 17569 ().27240 0.025 152 0.31226 4.5563 83,411 47.48 379.9 + 60.0
(fig. 6) 114-2 29,938 ().047228 0.35130 ().03736 27,980 18,922 63.20 93.46 + 2.77
114-3 33.953 0.057976 ().38035 (),05322 9.6808 18.253 53.75 90.23 + 6.63
114-4 183.46 0.46293 5.2656 ().21 180 1,5298 121.52 66.14 530.0+ 312.0

79C771 101-1 168.64 0.28784 0.058985 0.369.45 12.585 59.465 35.26 278.8 + 28.7
(fig. 7) 101-2 470.22 ().83678 0.29027 1.1120 19,681 141.64 30. 12 604.4 + 65.0
101-3 167.31 ().39749 ().022914 ().23509 2.7926 97.837 58.47 438.2 + 75.9
101-4 1077.0 3.5915 (). 19807 2.3541 1.870.9 381.39 35.41 1,314 + 659
101-5 22.366 0.024696 ().003383 ().02698 19,424 14.387 64.32 71.50 + 1.92
101-6 1272.3 3.4560 0.32611 2.3367 1.7346 581.86 45.73 1,747+ 1,306

81TT1211 99-1 11.777 ().021847 () () 12340 0.012292 25,827 8.1376 69.10 40.79 + 5.57
(fig. 8) 99-3 21.739 0.025867 0.000779 0.036843 14.269 10.844 49.88 54.15 + 2.42
99-4 17.261 0.024.701 0.003130 0.010519 54,165 14.145 81.94 70.31 + 0.89
99-5 21.742 0.013739 ().000925 0.026,141 0.97.164 14,009 64.43 69.65 + 22.58
996 238.79 1,0214 0.019467 ().55847 ().36695 73.750 30.89 339.8 + 276.8
99-7 44,022 (). 11787 ().001111 ().091029 1.4932 17.115 38.88 84.74 + 16.84
99-8 321.18 1.38.10 1,1708 0.62517 0.89.309 136.58 42.51 586.0 + 287.1
99-9 16.615 ().02.1366 ().000136 () ()20051 7.6325 10.682 64.29 53.35 + 2.24

Quartz veins associated with late Eocene to early Oligocene igneous event
79C1032 8970.245 5.0156 (). 12399 0.22852 () ()()31478 9.4073 4.0745 81.22 45.51 + 0.28
(fig. 9) 8970246 8,5362 0.38236 ().067216 ().011091 1.4048 5.2356 61.33 58.27+ 0.48
897.0247 10.825 0.47548 1.5676 0.015082 0.674.11 6.4626 59.64 71.66 + 0.85
897.0248 7.5420 ().36763 1.1612 ().010515 ().70169 4.4961 59.57 50.15 + 0.58
897.0249 5.1823 (). 16034 1.2324 0.004525 1,0897 3.9112 75.41 43.71 + 0.38
897.0250 3,8967 0.037980 ().35426 ().0011497 2.8827 3,5556 91.22 39.78 + 0.25

79C1043 8970.240 4,6007 0.20468 0.008.856 0.0028211 7.4874 3.7393 81.27 41.68 + 0.25
(ig. 10) 897.0242 8.3961 1,5301 (),029.108 0.012850 0.49378 4.5726 54.46 50.83 + 0.76
897.0243 7.2386 ().7508 2.1875 ().010203 0.41811 4.3639 60.20 48.54 + 0.82
897.0244 3.9638 ().075151 0.055961 ().0013574 10.673 3.5385 89.27 39.46 + 0.24
897.0251 3.7181 ().031822 ().50.783 ().0005269 0.47926 3.5728 96.06 39.84 + 0.59
897.0252 5.2010 ().54699 ().3.1335 () ()()38603 0.23390 4,0559 7797 45.16 + 1.27

| Reactor corrections: *"Ark/*Ark =0.0086, *Arca/?'Arca = 0.0003, and *Arca/'Arca = 0.0003. J = 0.00281.
* Reactor corrections "Ark/3Arg = 0.0285, *Arca/?'Arca =0.000671, and *Arca/37Arca = 0.000251. J = 0.00627.
* Reactor corrections: *"Ark/*Ark = 0.0285, *Arca/37Arc, = 0.000671, and *Arca/?'Arca = 0.00025 1. J = 0.00625.
 *ArAr STUDIES, BATTLE MOUNTAIN, NEVADA 163

4000 1000

T + not used in fit

- O used in fit

;3OOO 800

*-
600
3.
2000 k
<
S.
400

1OOO
200 k
- + not used in fit age=51.1+3.7 Ma
age=75.5+1.1 Ma L O used in fit 40Ar/36Ar-420.1+11.7
40Ar/36Ar-324.6+6.6

I I L 1 O |- 1 I 1 1 I I I

O 50 1OO 150 200 O 10 20 30 40

39AT/36Ar. 39Ar/36Ar.

Figure 5, 40Ar/39Ar versus 39Ar?6Ar diagram with isochron plot Figure 7. 40Ar/39Ar versus 39Ar?6Ar diagram with isochron plot
for sample 1471-760. Isochron age is 75.5+1.1 Ma. for sample 79C77. Isochron age is 55.1+3.7 Ma.

1OOO 2OOO

T + not used in fit

O used in fit

8OO M. +
1500

600 3.

Q 1000
<
< | c
<r

400 H.

5OO
2OO }
|- + not used in fit age=51.2+16.3 Ma |- age=37.2+5.3 Ma
|- O used in fit 40Ar/36Ar:-507.2+76.2 - 40Ar/36Ar-437+45.9

O - l l I I 1 o I I I l 1111 l 1 11111111 1

O 10 2O 30 O 25 50 75 100

39AI/36Ar. 39Ar/36Ar.

Figure 6. 40Ar/39Ar versus 39Ar?6Ar diagram with isochron plot Figure 8, 40Ar/39Ar versus 39Ar?6Ar diagram with isochron plot
for sample 1471-173. Isochron age is 51.2+16.3 Ma. for sample 81TT121. Isochron age is 37.2+5.3 Ma.
164 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
4OOO
|
|- + not used in fit
- O used in fit
3OOOH
3. |
H
-
&
$ 2000!
1 OOOH
age=38.5+0.3 Ma
40Ar/36Ar-4OO.2+6.1
O I 1 1 11111 I 1 L I 1 I I I
O 250 500 750
39Ar/36Ar.
Figure 9. 40Ar/36Ar versus 39Ar/36Ar diagram with isochron plot
for sample 78C103. Isochron age is 38.5+0.3 Ma.
on fig. 2) gives the best data for drawing an isochron
because two analyzed grains had relatively large amounts
of potassium (fig. 5). The influence on the isochron by
samples with a high value of 39Ar6Ar (an indication of
the amount of potassium) is a significant factor because
the points plot far from the ordinate on the 40Ar/36Ar ver
sus 39Ar6Ar diagram and have relatively more leverage
in drawing the best line fit between the plotted points. Six
grains define an isochron that is 75.1+1.1 Ma (fig. 5).
A second sample from the drill hole, 1471-173, also
yields an isochron that can be interpreted as Late Creta
ceous in age (fig. 6). It is much less well constrained,
however, and has a much larger uncertainty due to the
small amount of potassium in the six grains analyzed.
The isochron age of 51.2+16.3 Ma, based on three points,
is of little value in determining the age of the quartz fluid
inclusions.
Sample 79C77 presented the same problems as sam
ple 1471-173 (see above). Because of very small amounts
of potassium and consequently small amounts of 39Ar, the
analysis is heavily biased by the excess 40Ar in the fluid
of the inclusion. The isochron drawn through four points
defines an age of 55.1+3.7 Ma (fig. 7). The slope of the
line is strongly influenced by one analysis with the highest
40Ar/39Ar value (101-5, table 3); the other three analyses
cluster near the ordinate and near the 425 ratio, indicating
about 30 percent excess argon.
Two samples (78C103 and 78C104, fig. 2) of vein
quartz associated with a 38-Ma leucogranite north of the
10000
T O used in fit
7500
; 5O00
2500
age=38.3+0.6 Ma
40Ar/36Ar-395+20
1 I l
O L-111111111111111111111111.
1OOO O 500 1000 1500 2000 2500
39AI/36Ar.
Figure 10. 40Ar/36Ar versus 39Ar/36Ar diagram with isochron
plot for sample 78C 104. Isochron age is 38.3+0.6 Ma.
Buckingham quartz stockwork system produce well
defined isochron ages of 38.5+0.3 Ma and 38.3+0.6 Ma,
respectively (figs. 9 and 10). The leucogranite bodies
have K-Ar ages on hornblende and biotite of 37.7+1.4 Ma,
38.8+1.1 Ma, and 39.0+1.1 Ma (table 1); this is identical
to the fluid-inclusion ages. The trapping of fluid in the
quartz must have taken place at the time of cooling of
these igneous bodies, and there has been no subsequent
modification of the igneous system.
A summary 40Ar/36Ar versus 39Ar/36Ar diagram (fig.
11) shows the data points from all the analyses used from
the four samples of quartz from the Buckingham quartz
stockwork. The age is shown above or below each line.
This figure emphasizes the fact that samples with low
potassium (and, hence, a low 39Ar36Ar. ratio) cluster near
the ordinate and limit the construction of an isochron if
grains that have a higher potassium content are not avail
able. Ideally, a wide variation of potassium content would
produce the best isochrons. Samples 78C103 and 78C104
(figs. 9 and 10) are examples in which isochrons are con
trolled by a wide spread of 39Ar36Ar values. They are
not plotted on figure l l because they have points an order
of magnitude higher on the 39Ar/36Ar axis than the other
samples.
CONCLUSIONS
40Ar/39Ar dating of fluid inclusions in igneous miner
als has promise as a means of dating the fluid and, hence,
 *ArAr STUDIES, BATTLE MOUNTAIN, NEVADA 165

35OO early Oligocene) proved datable with reasonable con

A fidence. The Late Cretaceous event was evident but
3OOOH 2 obscure. The mechanism for partial argon loss from
S ^ fluid inclusions and for multiple periods of fluid en
.3% 33% trapment in minerals is poorly understood but has a
2500 H. 9
great effect on dating of mixed populations of fluid in
*- A f clusions by the 40Ar/39Ar technique.
$ 2000!- 21
3.
*-
^ 243 Q
<
1500k REFERENCES CITED

1OOO!- * Dalrymple, G. B., and Lanphere, M.A., 1971, ArAr technique

e 1471-760
D 81 TT121
5OO O79C77
n 1471-173
O 1 l I
O 40 8O 12O 160
39Ar/36Ar
Figure 11. 40Ar/36Ar versus 39Ar/36Ar diagram with isochron
plots from all Buckingham stockwork samples from the Battle
Mountain district.
the age of mineralization (assuming that the two are
related and nearly synchronous). Several limiting factors
are apparent from the quartz samples from the Battle
Mountain mining district, Nevada.
1. Excess inherited 40Ar may be present, and should be
expected, in the trapped fluid. Because of this, a num
ber of analyses are needed to provide points on a 40Ar/
36Ar versus 39Ar6Ar diagram from which an iso
chron can be drawn. The slope of the isochron gives
the age, and the intercept on the ordinate (40Ar/36Ar.
axis) indicates the amount of excess argon (if the value
is above 295.5 or the 40Ar/36Ar ratio of air).
2. Samples with small amounts of potassium that yield
proportionally small amounts of 39Ar on irradiation
are difficult to analyze with good precision. These
samples produce points on a 40Ar/36Ar versus 39Ar/
36Ar diagram that cluster near the 40Ar/36Ar axis and
are of little value in describing an isochron. Samples
containing small amounts of potassium proved nearly
unusable in this study.
3. Areas with several periods of igneous activity, such as
Battle Mountain (Late Cretaceous and late Eocene to
early Oligocene), probably produce several popula
tions of fluid inclusions. At Battle Mountain, the
youngest period of igneous activity (late Eocene to
of K-Ar dating: A comparison with the conventional tech
nique: Earth and Planetary Science Letters, v. 12, p. 300308.
1974, 40Ar/39Ar age spectra of some undisturbed terrestrial
samples: Geochimica et Cosmochimica Acta, v. 38, p.
715738.
2OO
Damon, P.E., and Kulp, J.L., 1958, Excess helium and argon
in beryl and other minerals: American Mineralogist, v. 43,
p. 433459.
Kelley, S., Turner, G., Butterfield, A.W., and Shepherd, T.J., 1986,
The source and significance of argon isotopes in fluid inclu
sions from areas of mineralization: Earth and Planetary Sci
ence Letters, v. 79, p. 303318.
McKee, E.H., 1992, Potassium-argon and *ArArgeochronolo
gy of selected plutons in the Buckingham area, in Theodore,
T.G. and others, 1992, Geology of the Buckingham Stock
work Molybdenum Deposit and Surrounding Area, Lander
County, Nevada: U.S. Geological Survey Professional Paper
798D, p. D36D98.
Nesmelova, Z.N., 1959, Gases in potassium salts of the Bereznik
ovsk mine: Trudy Vsesoyuznogo Nauchno-issledovatel'skogo
Instituta Galurgii, no.35, p. 206243.
Rama, S.N.I., Hart, S.R., and Roedder, Edwin, 1965, Excess radio
genic argon in fluid inclusions: Journal of Geophysical Re
search, v. 70, p. 509511.
Theodore, T.G., Silberman, M.L., and Blake, D.W., 1973,
Geochemistry and potassium-argon ages of plutonic rocks in
the Battle Mountain mining district, Lander County, Nevada:
U.S. Geological Survey Professional Paper 798A, 24 p.
Theodore, T.G., and Hammarstrom, J.M., 1991, Petrochemistry
and fluid-inclusion study of skarns from the northern Battle
Mountain mining district, Nevada, in Aksyuk, A.M., ed.,
SkarnsTheir Genesis and Metallogeny: Athens, Theophras
tus Publications, S.A., p. 497554.
Theodore, T.G., Blake, D.W., Loucks, T.A., and Johnson, C.A.,
1992, Geology of the Buckingham stockwork molybdenum
deposit and surrounding area, Lander County, Nevada: U.S.
Geological Survey Professional Paper 798D, 318 p.
Wahler, William, 1956, ber die in Kristallen eingeschlossenen
Flssigkeiten und Gose: Geochimica et Cosmochimica Acta,
v. 9, p. 105135.
 CHAPTER Q
HEAVY MINERALS AT THE FALL ZONE
A THEORETICAL MODEL OF GRAIN SIZE,
DENSITY, AND GRADIENT
By CURTIS E. LARSEN'
ABSTRACT
Placers are concentrations of heavy minerals sorted
by density and shape through fluvial, littoral, and (or)
eolian processes. A minerals specific gravity and
hydraulic equivalence relative to quartz determines the
type of sedimentary deposit in which it is found. For
example, a grain of gold or a titanium mineral such as
ilmenite or rutile in the very fine sand range (0.060.12
mm) tends to concentrate in dissimilarly sized deposits.
Gold of this size occurs with very coarse quartz sand
(1.02.0 mm) or gravel. Very fine grained titanium min
erals, as well as zircon and monazite, are deposited with
fine-grained quartz sand (0.130.25 mm).
The size of sediment carried and subsequently
depostied by a river depends on its discharge, slope,
depth, and velocity. Sediments with limiting grain sizes
that host placers are deposited in predictable reaches of
the system where slope and velocity decrease. In the
southeastern U.S., rivers draining the eastern Appalachians
commonly decrease in gradient and velocity until they
reach the outer edge of the Piedmont. Gradients of streams
flowing across the Piedmont steepen where crossing struc
tural or lithologic boundaries and then flatten abruptly at
the Fall Zone, where rivers enter the Coastal Plain. The
abrupt change in river dynamics preferentially deposits
coarse-grained sediments near the Fall Zone and medium
and fine-grained sand at the inner edge of the Coastal
Plain. Depending on upstream source terranes, the Fall
Zone may be targeted for exploration for placer gold,
whereas the inner edge of the Coastal Plain has potential
"U.S. Geological Survey, Mail Stop 954, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092
as the site for monazite, zircon, rutile, and ilmenite accu
mulations.
Upper Cretaceous fluvial sediments near the Fall
Zone in Georgia and South Carolina contain small concen
trations of monazite, as well as zircon, rutile, and ilmenite.
Fine-grained Cretaceous sands and silts of the upper delta
and delta facies concentrate less dense, but hydraulically
equivalent, heavy mineral suites that are characterized by
staurolite, kyanite, amphiboles, pyroxenes, epidote, and
rutile. A similar gradation of heavy minerals is also
present within the Cenozoic fluvial facies of the Coastal
Plain.
Although a qualitative relationship between certain
heavy mineral suites and the inner Coastal Plain fluvial
deposits has been known for some time, the role of stream
gradient has not been examined critically. Detailed geo
morphological and stratigraphic research in the Fall Zone
region will allow development of more effective predictive
models of placer mineral occurrence. Drainage basin
analysis utilizing stream discharge, gradient, and channel
width and depth, can be linked with geologic studies of
favorable source terranes to identify for exploration spe
cific river reaches that contain sediment sizes that are con
ductive to specific heavy mineral suites. Such studies
allow the extensive database of stream flow characteristics
in the southeastern U.S. to be coupled with economic geo
logic knowledge to refine heavy mineral prospecting tech
niques in the Fall Zone region.
INTRODUCTION
The inner edge of the Atlantic Coastal Plain has been
successfully targeted for heavy mineral exploration at vari
ous sites in the southeastern U.S. The availability of
favorable source terranes clearly influences the placement
168 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
of heavy mineral deposits, but differential sorting by
facies in fluvial systems is another important determinant
of their location and is an effective guide to exploration.
This paper examines heavy mineral occurrences near the
inner edge of the Coastal Plain of southeastern U.S. and
their relationship to stream gradients and fluvial facies. It
is focused on the Deep RiverCape Fear River system of
North Carolina to model downstream changes in heavy
mineral location.
River systems draining the eastern flank of the Appa
lachian Mountains commonly decrease in gradient until
they reach the outer edge of the Piedmont Upland. Gradi
ents are commonly abruptly steeper at structural or litho
logic boundaries in crystalline rocks and then decrease at
the inner edge of the Coastal Plain, where rivers cross
Mesozoic and Cenozoic sediments. Coastal Plain rivers
are graded to sea level. Zones of rapids near the contact
between the crystalline rocks of the Piedmont and the sed
iments of the Coastal Plain have historically been termed
the Fall Line or, more accurately, the Fall Zone (Fenne
man, 1938).
A change in river dynamics takes place near the Fall
Zone. Rivers flowing through constricted channels in the
crystalline terranes of the Piedmont carry the bulk of dis
charge from their drainage basins along with the sus
pended and entrained sediment load. Upon entering the
more erodable sediments of the Coastal Plain, channels
widen, causing a related decrease in stream velocity. As a
result, river systems tend to deposit coarse-grained sedi
ments near the Fall Zone, and they tend to deposit
medium- and fine-grained sand in the fluvial-deltaic facies
downstream.
Heavy mineral species are closely allied with the
grain size of their host sediments. Partially as the result of
hydraulic equivalence, grain shape, and bed roughness,
heavy minerals supplied to the fluvial system are differen
tially sorted by decreasing specific gravity. Although
source terrane is critical, differentiation by sorting is also
a critical control on the location of deposits. Stratigraphic
studies by Sohl and Owens (1991), Owens (1989), and
Owens and Gohn (1985) have indicated qualitative rela
tionships between heavy minerals and sedimentary facies
in the Cretaceous sediments of the Coastal Plain, but the
concept of linking facies with grain size to evaluate the
heavy mineral potential of the Coastal Plain has not previ
ously been attempted.
The Cape Fear River system of North Carolina is an
example of discharge and gradient changes conductive to
preferred deposition of heavy mineral suites. The Fall
Zone, marking the approximate inland boundary between
the Atlantic Coastal Plain and Piedmont Uplands, suggests
a representative model for preferential heavy mineral dep
osition at gradient changes along rivers draining the Caro
lina slate belt. This river system provides a model for
gradient and grain-size changes along former river courses
draining the same source terranes, and it may have local
ized undiscovered paleoplacers.
HEAVY MINERAL EQUIVALENCE
The formation of heavy mineral placers is a function
of weathering of primary source terranes, transport of
detrital material, and deposition in suitable environments.
Concentration of heavy minerals in placers results from
differential sorting by specific gravity and shape through
fluvial, littoral, and (or) eolian processes. The relationship
is complex but is controlled by density differences
between heavy minerals and other minerals (quartz, feld
spar) and their host deposit.
The weight equivalence and hydraulic-equivalent set
tling velocity relative to quartz (Rubey, 1933; Ritten
house, 1953; Tourtelot, 1968; Pettijohn and others, 1973)
bear directly on the type of sedimentary deposit in which
it is found. For example, a 1-mm diameter sphere of
quartz (specific gravity 2.65) is weight equivalent to a
gold sphere (specific gravity 18.0) of 0.528-mm diameter,
less than one phi-size interval smaller than the quartz. On
the basis of settling velocities in water, the hydraulically
equivalent size variation between gold and quartz is dif
ferent. The same 1-mm diameter sphere of quartz is
hydraulically equivalent to a gold sphere of 0.15-mm
diameter, approximately 2.5 phi sizes smaller than the
quartz equivalent. Less dense heavy minerals such as
ilmenite (specific gravity 4.75) settle at velocities nearer
to that of quartz. In terms of size, ilmenite spheres tend to
concentrate with equivalent-size quartz that are from 0.75
to 1.0 phi size larger.
Particle shape further complicates simple weight
equivalence. Flakes settle far less rapidly than spheres. A
flake with a diameter 10 times its thickness tends to settle
at only about one quarter the rate of a sphere of similar
diameter (Tourtelot, 1968). This translates to less than a
one-phi size difference.
Where sediments enter the fluvial system, laminar
and turbulent flow in channels dominate heavy mineral
sorting and deposition. Laminar flow at low discharge
periods relegates sediment to entrained transport move
ment of grains along stream beds by saltation. As velocity
increases, turbulent flow captures saltating grains and car
ries them as suspended load to where velocity decreases.
Differential-density settling velocity returns grains to the
sediment entrained in the bed load (Friedman and Sanders,
1978). The heavy minerals are the first to be returned to
the bed load, where they begin to be concentrated. Sling
erland (1977, 1984) studied the relative movement and
concentration of magnetite as entrained grains in bed load
in laboratory experiments. He demonstrated that heavy
mineral concentration within a sandy bed load is less
dependent on the hydraulic equivalency of grains than on
 HEAVY MINERALS AT THE FALL ZONE 169

3 ro 'S no S 3. - - RS 3 op
c g c c c g 5 c c g c c
O ! 1 I l l l l l l l l 1 l l

- & -

s
5
- v -

q)
- c -

- 3 -

7. -

i e Coase Sand to # I
10- ------> t
f
H. 5
z t
LL - C -

C 3.
5 - Percent Heavy Minerals -> M.

Cl

H - |

5
LL
:
-

|

3. - I
10 O- very Fine Sand 3-4c -

- - - - - - - - - - - - - - - - - N.

1 D-
-
sT__
Fine Sand 2-30
_ _ _ _ C.
-

1001 I I I T I I I I I i I T

DEPTH BELOW TERRACE SURFACE, IN CENTIMETERS

Figure 1. Grain size and heavy mineral distributions in stratified Holocene alluvial fill, Richard B. Russell Reservoir, Savannah River,
Georgia.

the coarseness of the bed load itself. Increasing bed rough
ness corresponds to a greater heavy mineral concentration.
Heavy minerals tend to nest in the interstices of larger
grains.
Studies of heavy mineral deposition in glaciolacus
trine deltas by Force and Stone (1990) examined grain
size differences and heavy mineral species in natural set
tings. They noted that the grain size of the host bed was
the only variable correlating well with the percentage of
heavy minerals. Each species showed maximum concen
tration at a particular grain size of the host bed. In general,
the more equant-shaped heavy mineral grains studied,
ranging in composition and density from magnetite (spe
cific gravity 5.18) to tourmaline (specific gravity 3.25),
showed median grain sizes from 0.5 phi to 1 phi size finer
than the total sample mode.
Similar relationships between grain size of the heavy
mineral species and the host are evident in alluvial depos
its as well. Figure 1 shows an example of a stratigraphic
analysis of Holocene, overbank, floodplain deposits along
the Savannah River in the Richard B. Russell Reservoir.
The weight percent of the total heavy mineral content of
each sample is plotted with the percentage of the individ
ual sand fractions. Weight percent is plotted logarithmi
cally, but depth is shown arithmetically on the x-axis.
This plotting technique identifies potential ratios among
the variables shown by similarity between curves. Figure
1 shows greatest similarity between the total heavy min
eral suite and the medium-sand fraction of each sample.
- Total heavy minerals are in ratio with the medium-grained
sand. These curves are emphasized with solid lines. The
comparison indicates that the heavy minerals present in
each sample are dominantly found within this size frac
tion. Heavy mineral species for the same samples are
shown on figure 2. Epidote and the amphiboles, the least
dense of the suite (specific gravity 3.03.5), are the most
abundant species. Epidote and, to a lesser extent, the
amphiboles account for the pattern of the total heavy min
eral curve on figure 1. As mineral species of greater
density are considered, the similarity with the medium
grained fraction becomes less distinct, and higher density
minerals tend to track with the fine- and very fine grained
fractions. Differentiation by grain size and density is
apparent.
The relationship between density and grain size of
the host deposit is clearly illustrated on figure 3, which
plots Rittenhouse's (1943) data on heavy mineral suites
from fluvial samples as adapted by Pettijohn and others
(1973). Heavy mineral species are plotted as the log of
specific gravity versus the difference in phi size between
the size of heavy minerals and the size of the host quartz
grains in the deposit. Species with specific gravities
170 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

; f ; ; ; ; ; ; ; ; ; ; f

001

Zircon

O :
-
H.

H. -

z
LL -

O - -

CC - -

Lll - -

Q- ! 0 F

E
(2 AmpnboleS

LL |

3.

100- very Fine Sand 3.4c 2 N T

-- - - - - - - - - - - - *- ~ - 25 9.25 mn ~ N

*-------- - - - / ---e. >

-

Medium Sand 129 _ _ "

- F--Rs. 8.
N/ S very
Fine sand 349_s^
- -
%,
So;
- T - -

H - - - -il-t---------__ . Y-- F--------- 2 * S$,

| * T T Erie Sand 23G - Fine Sand -- `s, -

C St. T
I E
OC) | I | | T | | |

DEPTH BELOW TERRACE SURFACE, IN CENTIMETERS

Figure 2. Distribution of heavy mineral species and grain size in stratified Holocene alluvial fill, Richard B. Russell Reservoir, Savannah
River, Georgia.

between 2.8 and 5.0 tend to correlate with host quartz
sands that are as much as 1.2 phi units coarser. The origi
nal work (Pettijohn and others, 1973) has been adapted
here by extending the curve to include higher density min
erals, such as cassiterite, silver, the platinum-group ele
ments, and gold. Whereas the relationship shown between
heavy mineral grains and their quartz equivalents on fig
ure 3 is simplified and applies to equant-shaped grains, it
is useful for assessing the heavy mineral potential of sand
bodies on the basis of grain size.
Weight equivalency and hydraulically equivalent set
tling velocities as described by Rubey (1933) and Tourte
lot (1968) are based on ideal shapes and conditions.
Fluvial experiments by Slingerland (1977, 1984), on the
other hand, rely on mathmatical expressions to explain
complex relationships between heavy minerals and the
grain size of entrained deposits in bed load. Observations
by Force and Stone (1990) are useful for understanding
heavy mineral concentrations in a single natural setting.
None of these techniques alone, however, specifically
addresses heavy mineral resource potential. Figure 3 pre
sents a testable procedure for identifying and assessing the
heavy mineral resource of fluvial sand bodies on the basis
of grain size.
STREAM GRADIENT AND
SEDIMENT SIZE
The relationship between the slope of a stream chan
nel and the size of its sediment load is complex. As early
as 1877, Gilbert (1877) tried to relate decrease in stream
gradient with the quantity of stream discharge. Sternberg,
 HEAVY MINERALS AT THE FALL ZONE 171

100.0 -
-

>- o T
O I Gold
E 3: I
3. O
-C
-- --
N
CC O Platinum Group Elements (PGE's)
(D s * - .
O 10.0 - * * O- Silver

L - g *; ;
# 5 3- 555
C
LL - 92 a 5 * : D- Cassite rite
0
Q) |s# s; 53:# S= *|
#3
- = e g; oq |
; : I Lee
J.O
3 | |
| Garnet

Amphiboles & Pyroxenes

1.0 I I I I I

O 1 2 3 4 5
PHI UNITS SMALLER THAN OUARTZ EOU!VALENT

Figure 3. Relationship of heavy mineral grains to hydraulically equivalent quartz grains in a fluvial environment (adapted from Pettijohn
and others, 1973; Rittenhouse, 1943).

on the other hand, (cf. Woodford, 1951, p. 813) believed
that the slope of the stream channel is directly related to
the size of the bed load. Rubey (1933) analyzed stream
slopes using data from Gilbert's (1914) classic flume stud
ies. He concluded that, if the depth-to-width ratio of the
channel was constant, the graded slope decreased with a
decrease in sediment load or particle size or with an
increase in discharge. Hence, a stream might be adjusted
to an increase in particle size by an increase in depth (rel
ative to width) or an increase of slope.
Hack (1957) attempted to clarify this relationship.
He noted that stream discharge is important in controlling
slope, but it was difficult to measure as a quantity. On the
other hand, stream discharge relates to its drainage area.
He was able to show a direct relationship with the slope
of a stream and the mean diameter of the bed material
when drainage area was taken into account. For example,
a clear relationship was present between slope and the
ratio of mean sediment size to drainage area, and sediment
size clearly decreased with a downstream decrease in
172 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
slope. In general, perennial streams show concave pro
files, with coarse debris associated with steep slopes and
fine debris with gentle slopes.
The complexity of the relationship between sediment
size and slope in stream channels was pointed out by
Leopold and others (1964). They showed that slope,
channel roughness (size of bed load), depth, and velocity
are interrelated through stream discharge. As discharge
increases downstream by the entrance of tributaries, a
channel may respond by an increase in width, depth, or
velocity. With perennial streams, the response is normally
an adjustment through increased width and depth of the
downstream channel; concomitantly, velocity decreases.
Expressed somewhat differently, the rate of change in
depth increases with a decrease in the rate of change in
slope. The rate of change in velocity decreases with the
increase in depth and decrease in slope. Roughness
decreases with slope and velocity as well (Leopold and
others, 1964, p. 257). All the variables adjust to overall
changes in discharge, such as those caused by increased
precipitation. The processes of differential sorting of sed
iments by shape, size, and density takes place within this
fluvial framework. Stream velocity and channel form are
key factors in sediment transport and deposition. Turbu
lent and laminar flow within reaches of river systems
influence the size and concentration of sediment carried
by saltation as bed load, or by suspension. Changes in
slope and channel form, however, determine where such
sediments are deposited. This is of major importance in
the study of heavy mineral placers because it serves as an
aid in the location of specific economic deposits.
Slope or stream gradient, with respect to drainage
area, has been shown by Hack, (1957, fig. 21) to have a
predictable relationship with the median size of bed mate
rial. Hack showed the interrelationship with the equation:
M 0.6
s- (*)
where
S is the channel slope in ft/mi,
M is the median size of bed material in mm,
A is the drainage area in mi2, and
18 is a constant determined for the units of
measurement.
Tourtelot's (1968) presentation of hydraulic-equivalent
grain sizes for heavy minerals can be adapted to Hack's
geomorphic presentation. This is useful because the drain
age basin is already used as a common unit to define
stream-sediment sampling sites in mineral resource assess
ment studies.
Channel form displays similar controls on the grain
size of deposits and the associated heavy minerals. Stream
velocity, together with turbulent and laminar flow
conditions, is a key factor in sediment transport and sort
ing and is related to channel depth and width. Particles of
a certain size are transported at a given velocity and set of
flow conditions. They will be deposited where the flow
conditions are no longer able to carry them as saltating or
suspended grains. Particles transported along confined
channels are commonly not deposited along river reaches
until the channels are no longer restricted and channel
width can increase. As the channel width varies, a con
comitant decrease in velocity results. Particles carried in
suspension in the higher velocity flow of the confined
channels are subsequently deposited as overbank deposits
where channel constriction is removed. These are left as
alluvial deposits of predictable grain size. Examples might
include local areas of abrupt lithologic variation between
resistant and less resistant rock types that allow locally
wider channels due to differential erosion. Similarly,
down-faulted or subsiding basins may exert geologic con
trol on channel width as well as gradient. Using hydraulic
equivalence as a guide and adjacent source terranes as a
requirement, anticipated heavy mineral suites may be
sought in such areas on the basis of the predicted grain
size of deposits.
Figure 4 presents a longitudinal profile of the Deep
RiverCape Fear River system superimposed on the
underlying geologic controls of the Carolina slate belt,
Deep River Mesozoic Basin, and the Atlantic Coastal
Plain. The profile of the Haw River, the major tributary of
the Cape Fear River is also shown. Gaging stations along
the profile are plotted as well. The profile shows a strong
geologic control with steep gradients associated with felsic
volcanic rocks and gentler gradients corresponding with
the metamudstone, argillite, siltstone, and shale. The low
est gradients occur with the sediments of the Coastal
Plain. Table 1 presents the calculated median grain size
for this river system on the basis of drainage area and
slope from Hacks equation above. Downstream from the
confluence of the Deep River and Haw River, grain size
of the Cape Fear River bed load decreases with increasing
discharge and decreasing slope. The most pronounced
change in grain size is between the Fall Line, near Lilling
ton, N.C. (Md = 129 mm), and Huske Lock (Md = 6.65
mm). Although the abrupt decrease in grain size down
stream from the Fall Line is partially attributable to the
grain size of the underlying Cretaceous sediments, the
change in slope also contributes to this variation.
Abrupt gradient changes along stream systems also
offer controls on the preferential deposition of sediment
and associated heavy minerals. Although influenced by
channel depth and width, locally or regionally steepened
gradients coincide with increased stream velocities and
larger diameter particles as bed or suspended loads.
Changes to less steep slopes have the opposite effect,
with finer sized grains deposited rapidly along the channel
as an adjustment to a decrease in stream velocity.
 300

Carolina
1
69
belt
Slate
Q)
~\
9. .9-
2C.
o
9
5

> 9O
200 2 Cn
C/D
3=

co
*
=
C)
-- -
=5
> L
LLi-
O
deposits
Quaternary
#s"
Coastal -

- s

O Q)
A.Q)
so.
#Plain
|OZone
Fall
3.S
~
2 -
w -
3
#*2\
f N Confluence
o
ov)
Riv 5 -.
dQ
wn
Rivers
Haw
and
Deep
Ne
##: un
-
S

> s% S/-
-
%:SSChatham
(Upper
Group
*-
-$Triassic)
3
7
69
do I
CQ
RpExPLANATION
;:3.Rp -
Formation
Pekin
~
-

c
M &3C9
9
`s
-
-;
C
Format
k
urnation
o
Fear
Cape
River
$Sozo
*::: nnock r
(Tertiary,
Formation
Duplin
Td
-
`-yr
Pliocene) -
e
wan-
+--
terrace IC PPg
Pennsylvanian)
(Permian
rocks
Granitic
and
P
i*rac
~ -
>
-
Seea
(Upper
Formation
Peedee
Kp
basin -
Cretaceous)
C9.
granitic
Metamorphosed
(Cambrian)
rocks
Kp
Lve
>
*
undett
Krn/Kct etentiated
c
(Upper
Group
Creek
Black
Ctv
Cretaceous)
rocks
metavolcanic
Felsic
(Cambrian)
`s~
Ko
t~ *...Kb K
-
cmv
rocks
metavolcanic
Mafic
`s Formation
Creek
Donoho
Krn
(Cambrian)
Ko
*-
s
Formation
Bladen
Cmd
Ko
metaargillite
and
Metamudstone
(Cambrian)
~

>
*
Metavolcanic-epielastic
Cve
Formation
Heel
Tar
Kt
rocks
(Cambrian)
~
99
stations
Gaging
Fear
Middendorf-Cape
Kmy
Formation,
>
(Upper
undivided
Kc
Cretaceous)
*~
T
-
-
T
I-T

500 200
400
300
O
100
DISTANCE
(KILoMETERs)

4.
Figure
profile
Longitudinal
Fear
Cape
the
of
North
system,
River
Carolina.

174 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 1. Calculated median grain size for the Cape Fear River formed a barrier to navigation (McGee, 1888). Fenneman
system bed load as a function of cumulative drainage area and (1938) broadened the definition to represent a zone, 510
slope. mi wide, that denoted a transition from the crystalline
rocks of the Piedmont to unconsolidated sediments of the
[Gaging station numbers are shown on figs. 4 and 7]
Coastal Plain. The term Fall Line was used instead to
Gaging station Cumulative Slope Median grain mark the point where Piedmont rocks passed beneath the
drainage area (mi') (ft mi) size (mm) sediments of the Coastal Plain. In this respect, the term
1. Deep River near became a boundary between physiographic provinces
High Point, N.C. ............ 46.9 |3 | 27.6 rather than a fluvial attribute. Hack (1982) returned to a
2. Deep River near fluvial definition. He showed that major breaks in modern
Randelman, N.C........... 125.() 4.5 12.4 stream gradients occurred within the Fall Zone and that
3. Deep River near these were often found within the Piedmont rather than the
Ramseur, N.C. ............. 349 () 1().5 142.1 Coastal Plain. Thus, there is ambiguity in the use of the
4. Haw River at
term. In this paper, Fall Line denotes the contact
Haw River, N.C. .......... 596.() 5.() 70.5 between Piedmont rocks and Coastal Plain sediments
5, Haw River near
Pittsboro, N.C. .......... 1,310.0 7. 9 332.()
exposed along modern stream beds. Fall Zone, on the
6. Deep River at
other hand, incorporates the steepened reaches upstream
Moncure, N.C.'......... 1,434.0 3.7 102.7
from the Fall Line to the inner edge of Coastal Plain sedi
7. Cape Fear River at ments. In this respect, the changing position of the contact
Lillington, N.C.......... 3,464.0 2.5 129 ()
between the Piedmont and Coastal Plain can be taken into
----------------------------- Fall Line ---------------------------------- account as major stream systems were incised or alluvi
8. Cape Fear River at ated during the post-Mesozoic history of tectonically and
Huske Lock.............. 4,810.0 0.34 6.7 eustatically induced base-level changes. Fall Zone, in this
9. Cape Fear River at view, incorporates the innermost deposits of the Coastal
Lock No. 1 ................ 5,2200 (). 12 1.3
Plain but recognizes the potential proximity of former
!Confluence of Deep River and Haw River. Head of Cape Fall Line positions with the contemporaneous fluvial
Fear River. facies of inner Coastal Plain deposits.
On the basis of past and present stream gradient
Hydraulic equivalency of heavy minerals in natural changes, similar in pattern to those of the modern system,
settings is coupled with the composite array of variables the inner Coastal Plain is preferentially suited for heavy
that influence stream velocity and particle size in fluvial mineral deposition and presents its own placer potential.
systems. The relationship is complex and resists simple Here again, the gradient change of streams draining the
explanation. In a general sense, however, specific reaches Piedmont played the major role.
along river channels may be highlighted as having the The Late Mesozoic and Cenozoic rocks of the
greatest potential for differential deposition of specific Coastal Plain form a complex wedge of fluvial and marine
heavy mineral suites as a response to settling velocities sediments that rises in altitude inland until pinching out
and stream flow. Two clear examples are geologic against the crystalline rocks of the Piedmont. Upper Creta
boundaries that (1) allow for abrupt widening of channels ceous strata rest unconformably on the weathered surface
in areas downstream from confined channels, and (2) of the crystalline rocks. This relationship is common for
abrupt changes to less steep gradients. Lithologic the Atlantic Coastal Plain from New Jersey to Georgia
changes from resistant to less resistant rock types, either (Frazier, 1979; Gohn and others, 1979; Owens and Gohn,
locally or regionally, can contribute to both types of pre 1985; Sohl and Owens, 1991). The basal Cretaceous units
ferred depositional settings. are sometimes weathered, both in outcrop and in subsur
face, and consist of very coarse to fine-grained sand. At
their basal contact they often incorporate coarse clasts of
LATE MESOZOIC AND CENOZOIC crystalline rocks. The basal Cretaceous formations include
the Tuscaloosa Formation in Georgia, the Cape Fear and
PLACER POTENTIAL Middendorf Formations in the Carolinas, and the Potomac
AT THE FALL ZONE Group in the Middle Atlantic area. All are fluvial sedi
ments related to ancestral river systems that once drained
the Piedmont. They are upper delta plain sediments correl
The Fall Line is loosely associated with the inner ative with contemporaneous shoreline positions that lie
edge of the Coastal Plain in the southeastern U.S., where east of the present coastline (Sohl and Owens, 1991). The
it originally implied an imaginary line drawn through the Lower Cretaceous sediments are overlain by Upper Creta
downstream location of falls on major river systems that ceous deltaic and shelf deposits.
 HEAVY MINERALS AT THE FALL ZONE 175

35 81 o 80 79 73: 76

i -

3. s
3. . &
$2. *\}}

&

*...
S e.

Pamlico
Sound
Lw

34,- \
-
-

jockinghai, a
A.
NQRTHCARQLINA.Z.
SOUTHCAROLINA DELTA PLAIN
zirconvmonazite

2: #
flix V, II PRODELTA
FRONTs *. staurolite
II * s <art.
DELTA PLAIN
*F - O
rence

zirconmonazite QNT
33%- P -

\)

W
IV
PRODELTA SHELF
& &V EXPLANATION
ee * ITar Heel Formation
II-Bladen Formation
&. IIIDonoho Creek Formation
2
(? O 10 20 30 40 50 MILES IVPeedee Formation

- 25 & 6 rs
32 1 S) I I l

Figure 5. Generalized paleoenvironmental facies analysis of the Black River Group and lower Peedee Formation between Columbia,
South Carolina, and Greenville, North Carolina.

The crystalline basement is an eastward subsurface
extension of rocks of the Piedmont Province. From the
inner edge of the Coastal Plain, the surface of the base
ment slopes eastward at about 3.43.8 m/km to where the
surface is about 450 m below sea level at the present
coastline. In the southeastern U.S., the surface is tectoni
cally deformed by the Cape Fear arch in the Carolinas and
by the Georgia uplift in Georgia and Florida. The Salis
bury Embayment lies northeast of the Cape Fear arch and
the Southeast Georgia Embayment lies to the southeast.
Both are basins with thick sedimentary sequences that
show Cretaceous and Cenozoic deformation.
Along the axis of the Cape Fear arch, the slope of the
basement surface beneath the inner Coastal Plain is 2.0 m/
km for 100 km downstream from the Deep River Meso
zoic Basin. Farther east, however, the slope of the base
ment surface is comparable to steeper areas both north and
south of the arch (3.8 m/km). The slope of the basement
below the Coastal Plain is similar to that of the present
Piedmont surface and suggests continuity of this surface
through time, although river systems have changed. As an
example, the modern gradient of the Deep River in the
eastern Piedmont is approximately 1.7 m/km (fig. 5). This
compares with the slope of the basement below the adja
cent inner Coastal Plain of 2.0 m/km. By contrast, the
modern gradient of the Cape Fear River across the Coastal
Plain is approximately 0.28 m/km (Soller, 1988). Thus,
Cenozoic deposition, complicated by tectonism and sea
level changes, has built up the Coastal Plain surface and
decreased stream gradients relative to the eastern edge of
the Piedmont. The change in gradient at the inner edge of
the Coastal Plain represents a hydraulic break conducive
to preferential heavy mineral deposition.
The PiedmontCoastal Plain transition zone was
identified by Mertie (1958, 1975) as a favorable location
for accumulations of detrital monazite and zircon. He
found these minerals within the heavy mineral suites of
basal Cretaceous sediments deposited on Piedmont crystal
line rocks in the Carolinas and Georgia. Mertie attributed
the monazite-zircon concentrations in Cretaceous rocks of
the inner Coastal Plain to simple heavy mineral dispersal
from source terranes in granitic rocks at the eastern edge
176 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
of the Piedmont. Detrital heavy mineral concentrations
decreased rapidly downstream from the source deposits
because they were diluted by mineral suites supplied by
nearby sources (Pettijohn, 1957, p. 566567). Grain-size,
hydraulic-equivlance, and sedimentary-facies relationships
received less coverage in Mertie's reports. Nonetheless, he
identified reworked sediments of Cretaceous and Tertiary
age at the Fall Zone as potential sites for placers of mona
zite and zircon. Following a dispersal model, he sought
out enriched fluvial placers for monazite and zircon along
specific secondary stream drainages that drained basal
Cretaceous formations.
HYDRAULIC SORTING WITHIN
CRETACEOUS FLUVIAL-DELTAIC DEPOSITS
The dispersal model was limited by the outcrop area
of basal Cretaceous rocks. There was little knowledge of
the heavy mineral suites of subsurface rocksnot much
was known of their grain size, stratigraphy, or depositional
environments. Recent work by Sohl and Owens (1991) on
the Cretaceous stratigraphy of the Carolina Coastal Plain
provides insights into the heavy mineral suites. Here, the
basal Cretaceous is represented by the Cape Fear Forma
tion, which rests unconformably on the basement. The
Cape Fear Formation, consists of poorly sorted, fine- to
coarse-grained fluvial sand. Sohl and Owens point out that
the heavy mineral suite in the outcrop area along the Cape
Fear River consists of ilmenite, zircon, tourmaline, rutile,
and significant monazite. To the northeast, along the Tar
River, monazite is subordinate to staurolite and kyanite. In
subsurface deposits, the heavy minerals at Charleston,
S.C., are more varied but contain zircon, rutile, and mona
zite. The coarse- to fine-grained fluvial sediments of the
Cape Fear Formation represent deposits of the upper delta
plain of a drainage system with a shoreline position near
the present coastline.
The overlying Middendorf Formation displays differ
ent lithofacies and mineralological relationships from
those of the Cape Fear. Both are of fluvial-deltaic origin
however, and the lithofacies of each changes from upper
delta plain to delta plain and delta front deposits from
southwest to northeast along the strike of the outcrop area.
The terrestrial facies grade into marine facies between the
Pee Dee and Cape Fear Rivers. Prodelta and shelf deposits
of the Middendorf are present in the subsurface near
Charleston. A decrease in grain size accompanies the
facies transition, as does an apparent hydraulically equiva
lent change in the heavy mineral suites. Zircon, rutile, and
monazite are common in the upper delta plain deposits
(Owens and Grosz, 1989), but are replaced by epidote,
garnet, and andalusite along the delta front and prodelta.
Hydraulic sorting by grain size, density, and lithofa
cies is well defined, where appropriate grain-size data are
available, in the overlying Black Creek Group. The Upper
Cretaceous of the Carolinas was formerly thought to rep
resent a single transgressive event with the Middendorf
Formation, Black Creek Group, and Peedee Formation as
lithofacies of the same unit that increased in altitude west
ward across the basement during transgression (Swift
and Heron, 1969). Sohl and Owens (1991) more reliable
dating of lithofacies along the Cape Fear valley shows a
series of transgressive and regressive events, each charac
terized by separate fluvial-deltaic facies.
As redefined by Owens (1989), the Black Creek
Group includes the Tar Heel Formation, the Bladen For
mation, and the Donoho Creek Formation. The Tar Heel
Formation incorporates fluvial sediments once identified
as Middendorf along the Cape Fear River. The group is
overlain by shelf deposits of the Peedee Formation. Figure
5 shows the outcrop pattern and facies relationships of the
Black Creek Group and Peedee Formation (Owens, 1989;
Sohl and Owens, 1991). Changes from delta plain, and
delta front, prodelta, to shelf are apparent from southwest
to northeast and across the Cape Fear arch.
Heavy mineral suites in these formations reflect the
facies changes. The coarse-grained Tar Heel Formation is
a representative example. Along the Peedee River, the
upper delta plain facies of the Tar Heel Formation con
tains abundant zircon and monazite. The delta front facies
exposed along the Cape Fear valley has little monazite and
less zircon, but garnet is abundant. In the prodelta facies
that crops out along the Neuse River, garnet decreases
with an increase of staurolite and kyanite. The shelf
deposits contain chloritoid as the dominant heavy mineral
species. The Bladen and Donoho Creek Formations are
progressively finer grained than the Tar Heel but show
similar heavy mineral suites and similar density sorting in
the identified deltaic facies.
The theoretical relationship between heavy mineral
suites, grain size, and deltaic facies is shown on figure 6,
which illustrates grain-size changes of the Upper Creta
ceous formations exposed along the Cape Fear valley from
the Fall Line to the coast. This figure shows first percen
tile and median grain sizes analyzed and plotted by Swift
and Heron (1969) to illustrate a downstream decrease
grain size in a fluvial-marine depositonal system. Their
data are adapted here to reflect the redefined facies
mapped on figure 5. It should be noted that Sohl and
Owens (1991) show downstream lithofacies in each for
mation of the Black Creek Group to progess from south
west to northeast along the strike of the outcrop rather
than along the Cape Fear River. Nonetheless, sections
exposed along the Cape Fear River show the same pro
gression between individual facies, although, in this case,
each facies represents a different formation. The grain-size
data show a consistent decrease and are used here to
model downstream heavy mineral distribution at the inner
edge of the Coastal Plain or Fall Zone.
 HEAVY MINERALS AT THE FALL ZONE 177

ExPL ANATION
O Cass tet ite

0 Monazite O Heavy mineral populations

with mean 2 Phi
O Magnet,te
O liment e O First per centile
O Zircon o Median diameter

O autue
O-Stauroitte
O Garnet
O Epidote
2 O Amphiboles
J O Chior it ord
* O Kyante
- Deep River - Outer Li- O I out malim
line

.3 r-Piedmont Mesozoic Basin Piedmon

-
- I~
~l.
~I s---
N. 1-f Inner Coastal Plain
- - - - - -
i
~ S. S Cape Fear Fm Donoho Creek Fm fine pebbles
N ~
-
- - -

2 `S. Tar Heel Fm Biaden Fm. | Pee dee F m. -

U) ~
= ~ ~ very fine pebbles
-> - 1 H. > C. -

- Ad
r ~
d- ven y coarse sand
LL O H. -

N
coarse sand

2: 1 -

<
C.
O medium sand

2 H. -

fine sand

3- Meta- Meta- -

Sediments Mesozoic sediments granites

*-i-e >|< ** very line sand
Upper Cretaceous Sediments
4 1 I l l l
0 50 100 150 200 250 300
DISTANCE ALONG CAPE FEAR RIVER (KILOMETERS)

Figure 6. Grain size of Upper Cretaceous Formations exposed along the Cape Fear River (adapted from Swift and Heron, 1969). Median
and first percentile grain size is plotted downstream from the Fall Line. Shown above are the theoretical ranges of heavy mineral species
calculated from median grain size of the Cape Fear Formation and Black River Group along the Cape Fear River. Heavy mineral popula
tions are calculated from an assumed mean grain size of 2 phi + 1 phi unit. The ranges of heavy mineral species are from figure 3.

Grain size progressively decreases from delta plain
facies through delta front facies in the Tar Heel to
prodelta facies in the Bladen and finally shelf facies in
the Donoho Creek. Anomalous, coarse grain sizes in the
overlying Peedee Formation, a shelf deposit, reflects a
basal concentration of reworked phosphatic clasts consist
ing of bored, subrounded cobbles, internal molds of mol
lusks, and bone and teeth material. Plotted above the
grain size data on figure 6 are the calculated equivalent
ranges for fine-grained heavy mineral species (+ one phi
size) shown on figure 3 and keyed to the median grain
size data of Swift and Heron (1969). The heavy mineral
suite characterized by monazite, ilmenite, zircon, and
rutile plot within the delta plain and delta front facies
suggested by Sohl and Owens (1991). The less dense
heavy minerals staurolite, garnet, epidote, and the
amphiboles plot farther downstream in the delta front and
prodelta facies. The theoretical and observed heavy min
eral identifications point to downstream sorting by
density as a key factor in the location of specific heavy
mineral suites.
The Cretaceous fluvial-marine distribution of heavy
mineral suites provides a conceptual model for understand
ing the depositional context of heavy mineral deposits on
the Coastal Plain. Figure 7 shows the heavy mineral ranges
of figure 6 superimposed on the longitudinal profile of the
Cape Fear River system (fig. 4) and plotted relative to the
Fall Line. The monazite-ilmenite-zircon-rutile zone, char
acterized by the grain size of the delta plain and delta front
facies of the Tar Heel Formation and the Cape Fear Forma
tion, lies immediately downstream from the modern Fall
Line in the vicinity of the Orangeburg Scarp and the inner
edge of the Duplin Formation. The less dense staurolite
epidote-garnet-amphibole zone within the Cretaceous is
anticipated farther downstream and beneath the Duplin sur
face and Quaternary alluvial deposits. Above the Fall Line
the coarser grained upper delta plain sediments would be
more conducive to gold and PGE deposition.
:

IC
P_T-1
and
(Perman
rocks
Grantic
PPg
sea
en sylvaKp n)==
(Upperlve
Kp
Formation
Peedee
nis
aO &Si
MCert~
(C9.
rocks
grantic
eamaobcrepoihuoas)ne)d
~

I
r-
T
I
I
I 200
O30
100
400
500
9
l2.00
CD
O G
O)
CnE
- O)f
E
a
#
25. luence n
F
5
O
Rivers
Haw
and
Deep
100
Cov)
ofnful :- C
- O
~
rr Rp
XChw
Pekin
Formation
EOP*LAq_ler
t, ace
C~ NATI)ON
`-
`s
.
>
CertaemtvboarlcieKb
(Upper
Group
Creek
mBlack
Felsic
(Chy
rocks Kd
~ Kt aoniu)cs)
`s
ss
mDonoho
Matic
(cmv
rocks
eCKo
Krn tamvborlciani)c
Formation
Creek
~

#
an
deposit
Quaternary
E
Triassic)
(Upper
Group
Chatham
O
*-7,
*<7
`s`srocks
MKt
Cve
(Cambrian)
Formation
Heel
Tar
etavolcanic-epielastic
~

Coastal
f I-
q2
q)
C
-
O#
Plain
*i
Zone
Fall
u

w
cheerew
c$Sozo
$:
T Be
Pliocene)
(Tertiary.
Formation
Duplin
Towance
--
Cumnock
Formation errace
-~letrace
-

5E:
#
3
= s.
s
2
;
:
Rp 69
CD

j-Fs.
&$
69
3
C`
f
$c
sq-

~
M In
>
Kndrnikct
-
u` if erentiated
-

,~
,
er,
s,
MBladen
mCmd
and
(Cambrian)
eta*mrugd~isltointe
Formation
Kb.

o
--C-
=>O

Threotraecteiocuasl
7.
CFigure
Upper
6.
figure
of
basis
on
River
Fear
Cape
the
along
zonation
mineral
heavy

Q1)9.
C5 3.
=
L1- *-
s
&
#
as
s TERs)
(KILowe
Distance
G
-

belt
Slate
Carolina >
stations
Gaging
99
Miodremnadotrif-oCnap,e
FKmy
Fear

~ Cretaceous)
(Upper
undivided
Kc

5&
300 Cl-

35.
#
3
s */
;
:
*E
 HEAVY MINERALS AT THE FALL ZONE
CENOZOIC REWORKING OF THE INNER
COASTAL PLAIN OF THE CAROLINAS
The surficial deposits of the Atlantic Coastal Plain
are dominantly Cenozoic in age. Although lower Tertiary
(Paleocene and Eocene) sediments are preserved at depth
within the subsiding Southeast Georgia and Salisbury
Embayments, they are commonly absent from the adjacent
arches (Gohn and others, 1979; Owens, 1989). Instead,
Miocene and lower Pliocene marine and coastal margin
facies transgress the innermost Coastal Plain deposits from
the Carolinas (Owens, 1989) to Virginia (Mixon and oth
ers, 1989). The transgressive events reworked the underly
ing Upper Cretaceous formations, as demonstrated by the
Duplin Formation on figure 7, and incorporated equivalent
heavy mineral suites within younger fluvial-deltaic and
marine facies. Quaternary transgressive deposits occur at
progressively lower altitudes eastward across the Coastal
Plain to the present coast and are contemporaneous with
alluvial fills along the major river systems.
The Cenozoic deposits of the Coastal Plain continue
the fluvial-deltaic style of heavy mineral sorting demon
strated for the Cretaceous. Stream gradients across the
Coastal Plain, however, decreased progressively through
time relative to the Piedmont and the Fall Zone while a
thickening wedge of sediment was deposited across the
Mesozoic Coastal Plain. In effect, a rising base level
adjusted to eustatic sea level and local tectonic and isos
tatic changes gave rise to lower gradient rivers across the
Coastal Plain and emphasized the abrupt change to a
lower gradient at the Fall Zone. As a consequence, finer
grained fluvial facies were deposited across the Coastal
Plain together with a suite of hydraulically equivalent
lighter heavy minerals dominated by staurolite, kyanite,
amphiboles, pyroxenes, and epidote. In contrast, the some
what coarser grained facies nearer the Fall Zone produced
concentrations of a high-density mineral suite: monazite,
magnetite, ilmenite, zircon, and rutile. Thus using stream
gradient and hydraulic equivalence as the dominant vari
ables, the Coastal Plain Cenozoic deposits near the Fall
Zone, as well as their Upper Cretaceous counterparts, tend
to be the preferred setting for monazite, ilmenite, rutile,
and zircon (Owens and Grosz, 1989). Subsequent rework
ing by coastal processes at the higher interglacial sea level
stands further concentrated the inner Coastal Plain fluvial
placers into barrier island beach and dune facies. These, in
turn, resemble the classic economic coastal placer settings
of southeastern Georgia and northern Florida.
CONCLUSION
The Fall Zone marks the transition between the crys
talline terranes of the Piedmont and the sediments of the
Atlantic Coastal Plain. Modern river systems like the Cape
179
Fear drain the Piedmont and enter the Coastal Plain at the
Fall Line, where stream gradients become gentle and the
grain size of sediments often begins a pronounced down
stream decrease. Although the position of the Fall Line
has changed since the Mesozoic, the Coastal Plain sedi
ments preserve a characteristic facies transition from upper
delta plain, to delta front, prodelta, and shelf that begins at
the edge of the Piedmont. A downstream decrease in grain
size is accompanied by a related sorting of heavy mineral
species by density and size. Coarser delta plain sediments
near the Fall Line receive more dense heavy minerals like
monazite, ilmenite, and zircon. Farther downstream the
assemblage is more often comprised of staurolite, garnet,
amphiboles, and epidote.
Low-gradient coastal plains offer a resource opportu
nity for concentrations of heavy minerals in suites gov
erned by grain size and density that are dependent on
upstream source terranes. With the Atlantic Coastal Plain
of the southeastern U.S. as an example, and given appro
priate source terranes, medium- to coarse-grained sand of
the upper delta plain facies can be expected to concentrate
a fine-grained heavy mineral suite characterized by mag
netite, monazite, ilmenite, zircon, and subordinate rutile.
The fine- to medium-grained sand of the delta front, on
the other hand, has an anticipated heavy mineral suite
dominated by very fine grained amphiboles, pyroxenes,
staurolite, kyanite, epidote, and tourmaline with subordi
nate and even finer grained ilmenite, zircon, and rutile.
In short, coarse clastics are deposited in proximity to
the Fall Zone, with grain size decreasing with distance
downstream from the edge of the Piedmont. The con
tained heavy mineral suites are sorted accordingly by
grain size, governed by hydraulic equivalence. A close
relationship between heavy minerals and fluvial facies
suggests that economic exploration should more closely
investigate extant stratigraphic facies reconstructions for
future economic studies.
REFERENCES CITED
Conley, J.F., 1962, Geology and mineral resources of Moore Coun
ty, North Carolina: North Carolina Department of Conserva
tion and Development Bulletin 76, 40 p.
Fenneman, N.M., 1938, Physiography of the Eastern United States:
New York, McGraw-Hill, 714 p.
Force, E.R., and Stone, B.D., 1990, Heavy mineral dispersal and
deposition in sandy deltas of glacial Lake Quinebaug, Con
necticut: U.S. Geological Survey Bulletin 1874, 21 p.
Friedman, G.M., and Sanders, J.E., 1978, Principles of Sedimentol
ogy: New York, John Wiley & Sons, 792 p.
Frazier, W.J., 1979, Sedimentology and paleoenvironmental analy
sis of the Upper Cretaceous Tuscaloosa and Eutaw Formations
in western Georgia, in Arden, D.D., Beck, B.F., and Morrow,
E., eds., Proceedings, Second Symposium on the Geology of
180 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
the Southeastern Coastal Plain: Georgia Geologic Survey In
formation Circular 53, p. 3952.
Gallagher, P.E. and Hoffman, C.W., 1990, Geology and heavy
mineral exploration of sedimentary deposits along the Fall
Zone of Northampton, Halifax, Nash, and Wilson Counties,
North Carolina: North Carolina Geological Survey Open-File
Report 904, 16p.
Gilbert, G.K., 1877, Report of the geology of the Henry Mountains:
U.S. Geographic and Geological Survey, Rocky Mountain
Region, 160 p.
Gilbert, G.K., 1914, The transportation of debris by running water:
U.S. Geological Survey Professional Paper 86, 259 p.
Gohn, G.S., Bybell, L.M., Christopher, R.A., Owens, J.P., and
Smith, C.C., 1979, A stratigraphic framework for Cretaceous
and Paleogene margins along the South Carolina and Georgia
coastal Sediments, in Arden, D.D., Beck, B.F., and Morrow, E.,
eds., Proceedings, Second Symposium on the Geology of the
Southeastern Coastal Plain: Georgia Geologic Survey Infor
mation Circular 53, p. 6474.
Hack, J.T., 1957, Studies of longitudinal stream profiles in Virginia
and Maryland: U.S. Geological Survey Professional Paper
294B, p. 5363.
Hack, J.T., 1982, Physiographic divisions and differential uplift in
the Piedmont and Blue Ridge: U.S. Geological Survey Profes
sional Paper 1265, 49 p.
Leopold, L.B., Wolman, M.G., and Miller, J.P., 1964, Fluvial
Processes in Geomorphology: San Francisco, W.H. Freeman,
522 p.
McGee, W.J., 1888, The geology of the head of the Chesapeake
Bay, U.S. Geological Survey, 7th Annual Report (18851886),
p. 537646.
Mertie, J.B., Jr., 1958, Zirconium and hafnium in the Southeastern
Atlantic States: U.S. Geological Survey Bulletin 1082A, 25 p.
1975, Monazite placers in the Southeastern Atlantic States:
U.S. Geological Survey Bulletin 1390, 41 p.
Mixon, R.B., Berquist, C.R., Jr., Newell, W.L., and Johnson, G.H.,
1989, Geologic map and generalized cross sections of the
Coastal Plain and adjacent parts of the Piedmont, Virginia:
U.S. Geological Survey Miscellaneous Investigations Series
Map I2033, 2 sheets.
Owens, J.P., 1989, Geologic Map of the Cape Fear region, Florence
1X2 quadrangle and northern half of the Georgetown 1%x2
quadrangle, North Carolina and South Carolina: U.S. Geolog
ical Survey Miscellaneous Investigations Series Map
I1948A, Scale 1:250,000, 2 sheets.
Owens, J.P., and Gohn, G.S., 1985, Depositional history of the Cre
taceous Series in the U.S. Atlantic Coastal Plain: Stratigraphy,
paleoenvironments, and tectonic controls of sedimentation, in
Poag, C.W., ed., Geological Evolution of the United States At
lantic Margin: New York, Van Nostrand Reinhold, p. 2586.
Owens, J.P., and Grosz, A.E., 1989, Rare-earth resource poten
tialMonazite distribution in the Fall Zone Coastal Plain
sediments of western South Carolina: U.S. Geological Sur
vey Research on Mineral Resources1989 Program with
Abstracts, Fifth Annual V.E. McKelvey Forum on Mineral
and Energy Resources: U.S. Geological Survey Circular
1035, p. 51.
Pettijohn, F.J., 1957, Sedimentary Rocks: New York, Harper &
Row.
Pettijohn, F.J., Potter, P.E., and Siever, Raymond, 1973, Sand and
Sandstone: New York, Springer Verlag, 618 p.
Rittenhouse, Gordon, 1953, Transportation and deposition of
heavy minerals: Geological Society of America Bulletin, v.
54, p. 17251780.
Rubey, W.W., 1933, Settling velocities of gravel, sand, and silt
particles: American Journal of Science, 5th series, v. 25,
p. 325338.
Slingerland, R.L., 1977, The effects of entrainment on the hydraulic
eqivalence relationships of light and heavy minerals in sands:
Journal of Sedimentary Petrology, v. 47, p. 753770.
1984, Role of hydraulic sorting in the orgin of fluvial plac
ers: Journal of Sedimentary Petrology, v. 54, p. 137150.
Sohl, N.F., and Owens, J.P., 1991, Cretaceous stratigraphy of the
Carolina Coastal Plain, in Horton, J.W., Jr., and Zullo, V.A.,
eds., Geology of the Carolinas: University of Tennessee Press,
p. 191220.
Soller, D.R., 1988, Geology and tectonic history of the lower Cape
Fear River valley, southeastern North Carolina: U.S. Geologi
cal Survey Professional Paper 1466A, 60 p.
Swift, D.J.P., and Heron, S.D., Jr., 1969, Stratigraphy of the Caro
lina Cretaceous: Southeastern Geology, v. 10, p. 201245.
Tourtelot, H.A., 1968, Hydraulic equivalence of grains of quartz
and heavier minerals, and implications for the study of placers:
U.S. Geological Survey Professional Paper 594F, p. FlF13.
Wilson, W.F., Carpenter, P.A., III, Burt, E.R., McDaniel, R.D.,
Coffey, J.C., and McKensie, B.J., 1978, Geology of the Ra
leigh 192 quadrangle, North Carolina: U.S. Department of
Energy, Grand Junction, Colo., Report no. GJBX51.
Woodford, A.O., 1951, Stream gradients and Monterey Sea Valley:
Geological Society of America Bulletin, v. 62, p. 799852.
 CHAPTER R
USE OF GEOCHEMICAL SURVEY'S IN
Ti-Hf-REE-Th-U PLACER EXPLORATION.
A MID-ATLANTIC-STATES EXAMPLE
By ANDREW E. GROSZ!
ABSTRACT
Selected elements from the hydrogeochemical and
stream sediment reconnaissance database of the National
Uranium Resource Evaluation Program (NURE) are diag
nostic and may predict the presence of commercially
important heavy minerals. Titanium (Ti), hafnium (Hf),
rare-earth elements (REE: Ce, Dy, Eu, La, Lu, Sm, Yb,
Y), and thorium (Th) are among the elements analyzed
and reported in the NURE database. Concentrations of
these elements in samples from the Atlantic Fall Zone and
Coastal Plain Province can predict the presence of com
mercially important heavy minerals such as ilmenite
(FeTiO3), rutile (TiO2), zircon ((Zr, Hf,U)SiO2), monazite
((Ce, a,Y,Th,U)PO4), and xenotime (YPO4). Ilmenite and
rutile are the most commonly occurring Ti-bearing miner
als in Fall Zone and Coastal Plain sediments. Zirconium
and hafnium possess extraordinary geochemical affinity;
thus, the presence of Hf (zirconium (Zr) was not analyzed)
usually indicates the presence of zircon.
Known placer deposits of heavy minerals in (1) the
North CarolinaSouth Carolina State-line region along the
Fall Zone, (2) coastal South Carolina, and (3) at the base
of the Fall Zone near the South CarolinaGeorgia State
line are outlined by the anomalies. The characteristic Ti
Hf-REE-Th-U signature is identified over unconsolidated
Upper Cretaceous and Tertiary sediments along the Fall
Zone and over North Carolina's Coastal Plain Province,
particularly over Pliocene and Pleistocene fossil shoreline
deposits. The geochemical data, therefore, may be used to
locate other deposits.
'U.S. Geological Survey, Mail Stop 954, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092
A large area in North Carolina's Coastal Plain Prov
ince is outlined for Ti-Hf-REE-Th-U-bearing heavy min
erals exploration. The Cape Fear arch is shown to have
had a pronounced effect on the distribution of placer
minerals and, therefore, on potential placer deposits in
Coastal Plain sediments in North Carolina. Upper Creta
ceous sediments in South Carolina and in North Carolina
south of the Cape Fear arch constitute a REE-Hf-Th-U
superprovince.
Other parts of the conterminous United States, partic
ularly in the largely unexplored Central Plains, appear
suitable for similar study. The potential for Ti-Hf-REE
Th-U-bearing deposits of heavy-mineral suites in the
United States may be much greater than previously
thought.
INTRODUCTION
Deposits of heavy minerals in unconsolidated sedi
ments, both of the fluvial and beach-complex type, furnish
most of the world's ilmenite and rutile (ores of titanium),
zircon (ore of zirconium and hafnium), and monazite (ore
of rare earths and thorium). Garnet, staurolite, tourmaline,
and aluminosilicates (kyanite and sillimanite) are common
byproducts of placer mining operations and are used as
abrasives or minerals for refractories.
Exploration for deposits of heavy minerals has tradi
tionally relied upon methods guided largely by geomrpho
logic and geologic rationale. Exploration commonly
involves preliminary field sampling of surficial sediments
followed by drilling on geomorphologically defined tar
gets. Most commonly these are Quaternary(?) beach ridges
and associated foredunes or their fossil equivalents in
coastal plain sediments.
182 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
More recently, airborne gamma-ray radiation and
high-resolution magnetic surveys have been shown to be
useful in outlining targets for sampling (Robson and Sam
path, 1977; Force and others, 1982; Grosz, 1983; Grosz
and others, 1989). Induced polarization is another promis
ing technique for onshore and offshore placer exploration
(Wynn and others, 1990).
This report details principles and techniques of placer
deposits exploration by use of stream-sediment geochemi
cal data and provides examples from the Carolinas that are
applicable to the conterminous States and Alaska. The
principles and techniques are applicable to a host of other
deposit types, metallic and nonmetallic, in hard-rock and
sedimentary settings. They may also prove useful in
regional geologic mapping.
THE NURE-HSSR PROGRAM
The National Uranium Resource Evaluation (NURE)
Program was established by the U.S. Department of
Energy to gather data with which to evaluate uranium
resources and to identify areas in the United States poten
tially favorable for uranium deposits (Averett, 1984). The
hydrogeochemical and stream sediment reconnaissance
(HSSR) was one element of the NURE program. The
HSSR effort involved collection and analysis of samples
of stream sediment, ground water, and surface water from
the conterminous 48 States and Alaska to determine con
centrations of uranium and other selected elements
(Arendt and others, 1980). The data were expected to help
outline geochemical provinces and to suggest favorable
areas for more detailed studies. Tabulations of areas sur
veyed during the HSSR and the information that is avail
able for those areas are given by Averett (1984).
Sampling and reporting in the HSSR were done on a
quadrangle basis using the 1X2 quadrangles of the
National Topographic Map Series (NTMS) of the U.S.
Geological Survey. HSSR data reports were issued for
330 (70.5 percent) of the 468 quadrangles in the 48 con
terminous States and in 104 (68 percent) of the 153 quad
rangles in Alaska. About 70 percent of the Nation has
geochemical coverage. All samples were analyzed for
uranium, and nearly all were also analyzed for up to 59
elements (Averett, 1984).
METHODOLOGY
The approach used in this study is based on the
hypothesis that elements such as titanium (Ti) and
hafnium (Hf) in stream-sediment samples reflect on the
presence of common economic heavy minerals such as
ilmenite (FeTiO3) and rutile (TiO2) in the case of Ti, and
zircon ((Zr, Hf)SiO2) in the case of Hf (Zr was not
analyzed in HSSR). The rare-earth elements (REE) cerium
(Ce), dysprosium (Dy), europium (Eu), lanthanum (La),
lutetium (Lu), samarium (Sm), and ytterbium (Yb) are
commonly associated with yttrium (Y), thorium (Th), and
uranium (U) in heavy minerals such as monazite
((Ce,La,Y,Th)PO4), xenotime (YPO4), and allanite (epi
dote group: (Ca,Ce,Y)2(Al,Fe)3(SiO4)3(OH))other REE
were not analyzed in HSSR. Because zirconium and
hafnium possess extraordinary geochemical affinity, Hf
usually indicates the presence of zircon (Vlasov, 1966).
Although other commonly occurring heavy and light min
eral species can carry these elements as well (for example
Ti in amphiboles; Hf in pyroxenes; REE, U, and Th in
apatite), they typically contain only trace abundances of
these elements.
Mertie (1975), OverStreet (1967), Force and others
(1982), and Owens and others (1989) show that monazite
and zircon are present in the area of this report and that
the sediments have undergone extensive weathering that
removed much of the labile mineral component (amphib
oles and pyroxenes, for example). Force and others (1982)
and Owens and others (1989) also show that monazite
(with its Th and U content) and zircon (with its U and Th
content) control the aeroradiometric signature of surfi
cially exposed heavy-mineral concentrations.
The HSSR database for each quadrangle contains all
the sample types, their location coordinates, various
descriptive fields, and their analytical data. The database
for each of the quadrangles discussed in this report was
searched for stream-sediment samples (sample type 50 in
the database; 100 mesh (149 micron) sieve fraction)
having data for Be, Ce, Cr, Dy, Eu, Hf, La, Lu, Nb, P,
Sm, Sn, Th, Ti, U, W, Y, and Yb by use of a query-by
example relational database program.
For the 10 1x2 quadrangles used in this study,
10,093 stream-sediment samples met the search criteria.
For any given sample, elemental data were: (1) a positive
number (expressed as parts per billion (ppb)), (2) zero, (3)
a negative number (signifying less than a given level of
concentration; i.e., -2,000 ppb), or (4) no value (analytical
data not available). In this study, only positive numbers
were utilized for statistical analyses and graphic displays;
data are discussed in percent format.
First, cumulative frequency plots for individual ele
ments were constructed to identify anomalous outliers. As
an example, figure 1 shows REE4-Y for 8,174 samples. A
maximum flexure in the curve near 7,600 on the abscissa
indicates that the upper 9.3 percentile group (758 samples)
has > 0.05 percent REE+Y. This group was plotted on a
generalized geologic base map for North and South Caro
lina on which the Upper Cretaceous and Tertiary sedi
ments (excluding those on the lower portions of the
Coastal Plain) are shown as a single unit (fig. 2). The Fall
Zone, at an average altitude of 240 feet, is generally
coincident with the western edge of this unit. This
 USE OF GEOCHEMICAL SURVEYS IN PLACER EXPLORATION 183

1.0 Figure 3 displays linear relationships between the

F
REE+Y and Th for the 758 samples having > 0.05 percent
2 +
REE+Y. There appear to be at least two populations, dis
5
$2
ul
1.5 tinguished by their Th to REE+Y ratios, that have not
a

2 +
been intermixed to any great extent. The correlation of Th
>
and REE4Y for both populations is expected for monazite;
20 f this is confirmed below. However, xenotime can be shown
E # to be important locally by plots of the Y-subgroup (heavy
cc
|-- REE) and the Ce-subgroup (light REE). Variability in the
C 0.5
Th content of monazite (Overstreet, 1967) may also
; account for the two populations.
O Zircon is a commonly occurring U-bearing mineral
O l 2 3 4 5 6 7 8 9 in terrigenous clastic assemblages. A plot of Hf versus U
NUMBER OF SAMPLES, IN THOUSANDS (fig. 4) for the group of Hf-bearing samples (60 percent
of the total sample population) suggests that at least two
Figure 1. Cumulative frequency plot of 8,174 samples having incompletely mixed groups (end members) differing in
REE+Y > 0.0 percent. Maximum flexure in the curve near 7,600
their U/Hf ratios exist in the sample suite. Because the
on the abscissa indicates that the upper 9.3 percentile group (758
samples) has > 0.05 percent REE4-Y. samples used for this plot span a broad spectrum of geo
logic units, compositional differences in zircons (for
example, perhaps normal zircons and a metamict zircon
approach was followed for Hf, U+Th, and Ti. Other population with low concentrations of U) may account
placer-mineral-indicating elements mentioned above were for the scatter. However, other mineral species (for exam
also plotted and showed patterns of distribution paralleling ple Hf-rich pyroxenes) contributing to the scatter may be
those shown in the figures. represented as well. Moreover, a plot of Th versus U for

81O 77o
I I

IVIRGINIA
- - - +-----: F---ey
360
No data
Nort N CAROLINA+
+
7.A dj
-

N alliico
Sound

#; + -'g
*CARQLINA i:

Tertiary and Cretaceous

rocks, undivided

Figure 2. Locations of samples having > 0.05 percent REE+Y.

 184 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

0.25 samples having > 0.05 percent REE4-Y (fig. 5) suggests

that Th-rich monazite is an important source for both
radioelements.
0.20

0.15
DISCUSSION

: 0.05
Samples from the locations shown on figure 2 con
tain more than 0.05 percent REE+Y and represent 758
(7.5 percent) of the total sample population. The greatest
concentration of anomalous values follows the western

0.50 1.00 1.50

SUM OF REE + Y, IN PERCENT
Figure 3. Plot of the sum of REE-Y against thorium for 758 samples
having > 0.05 percent REE+Y (plotted on fig. 2).
boundary of the undifferentiated Upper Cretaceous and
Tertiary sediments (the Fall Zone) with remarkable accu
racy and persistence from the South CarolinaGeorgia
State line north to about central North Carolina. North of
this central region, the anomalous REE4-Y pattern contin
ues approximately linearly onto Mississippian and Penn
sylvanian intrusive and interlayered metasedimentary and
metavolcanic rocks of Late Proterozoic and (or) Cambrian
age in the Piedmont, lying to the west of the Cretaceous
and Tertiary sediments.
An anomalous REE--Y site on the lower portion of
the Coastal Plain near Pamlico Sound is approximately
coincident with known REE4-Y-bearing sedimentary phos
phorite occurrences. To the south of the approximate axis
of the Cape Fear arch, anomalous REE4-Y locations

. +
extend to the lower reaches of the Coastal Plain, particu
larly along the Cape Fear and Cooper Rivers. Sedimentary
phosphorite is also known to occur in this region (Force
and others, 1982). Shoreward dispersal of monazite-bear
ing sediments by rivers seems to be an important mecha
O. 10 0.15 0.20 0.25 nism only to the south of the Cape Fear arch. Anomalous
Hf, IN PERCENT REE+Y concentrations are also evident in the region of
the Horse Creek deposit, located at the Fall Zone near the
Figure 4. Plot of Hf against U for the Hf-bearing sample population
(N = 6,102).
GeorgiaSouth Carolina State line (Mertie, 1975). Staatz
and Armbrustmacher (1979) defined a rare-earth province
coincident with the Cretaceous and Tertiary trend because
of the high monazite content of the sediments.
0.04
The locations shown on figure 6 contain more than
0.15 percent Ti and represent 671 (6.6 percent) of the total
0.03 sample population and 11.4 percent of the Ti-bearing sam
ples. The distribution pattern is significantly different
from that of REE4-Y, although no less remarkable in its
0.02 persistence and localization over Fall Zone sediments in

. 0.01
north-central North Carolina.

To the east, and parallel to the Fall Zone sediments,

anomalous Ti concentrations in Coastal Plain sediments
extend south to the vicinity of the northeastern flank of the
Cape Fear arch. Many of these anomalous concentrations
0.10 0.15 0.20
are approximately coincident with the late Tertiary to
Th, IN PERCENT
early Quaternary Surry Scarp (a major geomorphic fea
ture), as shown by Blackwelder and Cronin (1981). South
Figure 5. Plot showing the relationship between Th and U in sam of the Cape Fear arch, anomalous Ti concentrations are
ples having > 0.05 percent REE+Y. much less common in Coastal Plain sediments than to the
 USE OF GEOCHEMICAL SURVEYS IN PLACER EXPLORATION
810
I
Rl
O Cata
NS r;
f F:
- - ++
EoRGIA%
Figure 6. Locations of samples having > 0.15 percent Ti (N = 671).
north; the locations of samples containing anomalous con
centrations of Ti shift westward onto crystalline rocks
above the Fall Zone. In southern North Carolina, anoma
lous Ti values persist over the western edge of Upper Cre
taceous sediments. In the North Carolina Piedmont,
anomalous Ti concentrations occur in distinctive north
east-southwest-trending belts over Cambrian and Late Pro
terozoic metasedimentary, metavolcanic, and igneous
rocks. The few anomalies over beach-complex sediments
in coastal South Carolina are approximately coincident
with heavy-mineral concentrations found by previous aer
oradiometric surveys (Force and others, 1982).
Samples from the locations shown on figure 7 con
tain more than 0.015 percent Hf and represent 523 (5.2
percent) of the total sample population and 8.6 percent of
the Hf-bearing samples. The distribution pattern shows
more scatter than do the REE--Y and Ti distributions
(compare fig. 7 with figs. 2 and 6). Over Fall Zone sedi
ments in northern North Carolina, the Hf distribution
closely resembles that of Ti, and over Coastal Plain sedi
ments in South Carolina, the Hf pattern closely resembles
that of Ti and REE+Y. The high cutoff value used to
delimit anomalous Hf concealed subtle Hf anomalies that
coincide with the Ti anomalies in coastal North Carolina,
185
Tertiary and Cretaceous
rocks, undivided
discussed above. South of the Cape Fear arch, anomalous
Hf is remarkably persistent along the western boundary of
Upper Cretaceous sediments. Crystalline rocks, specifi
cally Silurian to Pennsylvanian intrusives appear to be
important Hf (and REE+Y) sources in South Carolina.
Figure 8 shows the geographic distribution pattern of
631 (6.3 percent) of the total sample population and 8.8
percent of the samples that contain more than 0.01 percent
U and Th combined. As predicted by figures 2, 3, and 4,
the distribution pattern shown by these radioactive ele
ments is remarkably consistent with that for REE+Y and
much less so with that for Hf. This suggests that monazite
may be the major control on both the U and Th distribu
tion. Although less well defined because of large flight
line spacing, NURE spectral aeroradiometric surveys
(Duval and others, 1989, 1990) also show anomalous U
and Th signatures over many of these same areas. More
closely spaced total-count aeroradiometric data are also in
agreement (Force and others, 1982; Owens and others,
1989).
In contrast, the NURE-HSSR chemical data clearly
outline the area of heavy-mineral deposits at the North
CarolinaVirginia State line where gamma-ray radiation
surveys do not. Thus, the NURE data may be used to find
186 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
810
T |
+ VIRGINI
+---------.
O N AROLINA
*
h **'. - ;
Figure 7. Locations of samples having > 0.015 percent Hf (N = 523).
Ti-bearing placer deposits that do not contain appreciable
quantities of radioactive heavy minerals (i.e., monazite +
zircon). The congruence of a number of elements indica
tive of several placer minerals, therefore, is interpreted as
an unambiguous indicator of the presence, and relative
abundance, of commercially important heavy minerals.
Thus, in conjunction with geologic, topographic, and
airborne geophysical information, the HSSR stream-sedi
ment data can be used as a powerful exploration tool for
placer deposits.
CONCLUSIONS
NURE-HSSR geochemical data is well suited for out
lining small- to large-scale areas within the conterminous
U.S. and Alaska permissive of Ti-Hf-REE-U-Th-bearing
deposits of placer heavy minerals. The geochemical signa
ture of known deposits of heavy minerals is identified
over a large portion of North Carolinas Coastal Plain and
Fall Zone regions. The Fall Zone region (generally coinci
dent with exposures of Cretaceous and Tertiary sediments)
in central Georgia, south-central Alabama, northeastern
Mississippi, and western Tennessee shows a continuation
770
-
**** T rS.
Tertiary and Cretaceous
rocks, undivided
of this geochemical pattern. Examination of only the most
anomalous samples in the total sample population for any
given element, as done in this paper, excludes much valu
able information on potential large-volume/low-grade
deposits of heavy minerals, particularly in coastal plain
sediments where their geochemical expression is subdued.
Examination of the 70 to 90 percentile group of the popu
lations reveals persistent anomalies in Atlantic Coastal
Plain sediments. The 50 to 90 percentile group over some
midcontinent and western areas show similar geochemical
signatures. Large areas in northeastern Kansas and eastern
Nebraska, for example, may also have exploration poten
tial for this mineral suite. The potential for Ti-Hf-REE-U
Th-bearing deposits of placer heavy minerals in surficially
exposed sediments in the U.S. may be much greater than
previously thought.
ACKNOWLEDGMENTS
Appreciation is extended to John C. Reed, Jr. for pro
viding the digital geographic and geologic base map and
to Terry L. Klein for reformatting the data. E. Force, J.
 USE OF GEOCHEMICAL SURVEYS IN PLACER EXPLORATION 187

810 77o

VIRGINIA
F---ty
\
f + -

NORTR CAROLINA+". +

W -

; : ...

Tertiary and Cretaceous

rocks, undivided

Figure 8. Locations of samples having > 0.01 percent U and Th combined (N = 631).

Philpotts, and K. Schulz provided thoughtful and construc
tive technical reviews.
REFERENCES CITED
Grosz, A.E., 1983, Application of total-count aeroradiometric maps
Arendt, J.W., Butz, T.R., Cagle, G.W., Kane, V.E., and Nichols,
C.E., 1980, Hydrogeochemical and stream sediment recon
naissance procedures of the uranium resource evaluation
project: U.S. Department of Energy, GJBX32(80), 53 p.
Averett, W.R., 1984, National uranium resource evaluation
Grosz, A.E., Cathcart, J.B., Macke, D.L., Knapp, M.S., Schmidt,
guide to data reports of the hydrogeochemical and stream
sediment reconnaissance: U.S. Department of Energy,
GJBX5(84), 119 p.
Blackwelder, B.W., and Cronin, T.M., 1981, Atlantic Coastal Plain
geomorphology illustrated by computer-generated block dia
grams: U.S. Geological Survey Miscellaneous Field Studies
Map MF-1242.
Duval, J.S., Jones, W.J., Riggle, F.R., and Pitkin, J.A., 1989, Equiv
alent uranium map of the conterminous United States: U.S.
Geological Survey Open-File Report 89-478, 10 p., scale
1:2,500,000.
Duval, J.S., Jones, W.J., Riggle, F.R., and Pitkin, J.A., 1990, Potas
sium and thorium maps of the conterminous United States:
U.S. Geological Survey Open-File Report 90-338, 16 p., scale
1:2,500,000.
Force, E.R., Grosz, A.E., Loferski. P.J., and Maybin, A.H., 1982,
Aeroradioactivity maps in heavy-mineral exploration
Charleston, South Carolina, area: U.S. Geological Survey Pro
fessional Paper 1218, 19 p.
to the exploration for heavy-mineral deposits in the coastal
plain of Virginia, with a section on Field-spectrometer-data re
duction, by K.L. Kosanke: U.S. Geological Survey Profession
al Paper 1263, 20 p.
Walter, and Scott, T.M., 1989, Geologic interpretation of the
gamma-ray aeroradiometric maps of central and northern Flor
ida: U.S. Geological Survey Professional Paper 1461, 48 p.
Mertie, J.B., Jr., 1975, Monazite placers in the southeastern Atlan
tic States: U.S. Geological Survey Bulletin 1390, 41 p.
Overstreet, W.C., 1967, The geologic occurrence of monazite: U.S.
Geological Survey Professional Paper 530, 327 p.
Owens, J.P., Grosz, A.E., and Fisher, J.C., 1989, Aeroradiometric
map and geologic interpretation of part of the Florence and
Georgetown 1%2 quadrangles, South Carolina: U.S. Geolog
ical Survey Miscellaneous Investigations Series Map
I1948B.
188 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Robson, D.F., and Sampath, N., 1977, Geophysical response of
heavy-mineral sand deposits at Jerusalem Creek, New South
Wales: Bureau of Mineral Resources Journal of Australian Ge
ology and Geophysics, v. 2, p. 149154.
Staatz, M.H., and Armbrustmacher, T.J., 1979, Preliminary map of
rare-earth provinces in the conterminous United States: U.S.
Geological Survey Open-File Report 79576-T.
Vlasov, K.A., ed., 1966, Geochemistry and mineralogy of rare
elements and genetic types of their deposits [translated
from Russian by Z. Lerman and Y. Brenner, Israel Pro
gram for Scientific Translations): Jerusalem, Israel, 1633 p.
Wynn, J.C., Grosz, A.E., and Carlson-Fosca, V.L., 1990, In
duced polarization response of some titanium-bearing plac
er deposits in the southeastern United States, in Fink, J.B.,
ed., Induced PolarizationApplication and Case Histories:
Tulsa, Oklahoma, Society of Exploration Geophysicists, In
vestigations in Geophysics No. 4, p. 280303.
 CHAPTERS
THE EFFECTS OF HYDROTHERMALLY ALTERED
BEDROCK ON NATURAL FOREST VEGETATION IN
THE SNOW CAMP-SAXAPAHAWAREA,
NORTH CAROLINA, AND THE RESULTING
EXPRESSIONS IN LANDSAT TM IMAGERY
By ALBA PAYAS, ROBERT G. SCHMIDT, and ANDREU BONET3
ABSTRACT
Areas of intensely hydrothermally altered rocks in the
Carolina slate belt are partially marked by anomalous
patches of forest that are characterized by the forest com
munity present and by the general condition of the canopy
and understory. Parts of many of these hydrothermally
altered areas are recognizable on processed satellite
images and thus provide a useful exploration tool for
locating unknown altered areas. The causative factors for
the unusual forests and for their recognition on satellite
images have been the subject of much speculation, but
controlled data that may aid in identifying the factors have
been difficult to obtain.
The Snow CampSaxapahaw area within the Carolina
slate belt includes large masses of volcanic rocks that have
been intensely altered by hydrothermal processes of the
high-sulfidation type. Taxonomic and physiological stud
ies of relevant forests and comparisons of soil analyses
indicate that three factors are probably most important in
influencing the distribution of satellite-detectable anoma
lous forest vegetation: (1) low amounts of available nutri
ents, especially Ca and Mg, in soils formed on the areas of
most intense hydrothermal alteration, (2) rocky, relatively
thin soils, and (3) south- or southeast-facing slopes. Both
'Servei Geolgic de Catalunya, Barcelona, Spain
*U.S. Geological Survey, Mail Stop 954, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092
*Botany Department, University of Barcelona, Barcelona, Spain
the second and third factors can be assumed to augment
water stress during dry periods.
INTRODUCTION
The Snow CampSaxapahaw area chosen for this
study is located within the Carolina slate belt, 20 km south
of Burlington and 30 km west of Chapel Hill, in Ala
mance and Chatham Counties, North Carolina (fig. 1).
Most of the study area is currently being mapped at a
scale of 1:24,000. The 268-km area is underlain by volca
nic and volcaniclastic rocks of a wide variety of textures
and compositions, from basaltic to rhyolitic. As in many
other parts of the slate belt, these volcanic rocks are
intruded by many plutons and small stocks of shallow
intrusive rocks (also of diverse compositions, from gab
broic to granitic), many of which have associated strong
hydrothermal alteration. Mineral potential of the area
includes high-grade pyrophyllite, which has been quarried,
and gold worked in various prospects and small-scale
mines. Outcrops are locally abundant in the Snow Camp
Saxapahaw area, though for the most part, they are widely
scattered, and in some of the area, they are rare.
Forest vegetation of the ridges and resistant hills in
the Piedmont area have been described as more like that
of the mountains farther west (Braun, 1950) and unlike the
oak-pine association of the surrounding Piedmont (Milton
and others, 1983). Because many of the intensely altered
core zones form erosion-resistant ridges, there is a ten
dency for the forest communities on them to be different
due to the topography alone; the effects of rock alteration
are superimposed on that predictable variation.
190 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
.
KENTUCKY 2 V | R G|N|A
-
- ~~
|- - - - - - - - -
TENNESSEE
Figure 1. Index map showing the location of the Snow CampSaxapahaw study area (rectangle with diagonal line pattern) in the Carolina
slate belt, North Carolina.
Identification of areas of hydrothermally altered rocks
using processed Landsat images has been tested in the
region with some success (Schmidt and Koslow, 1986;
Schmidt and others, 1987).
Objectives.The major objective of this study was to
more precisely define the factors enabling discrimination
on satellite images of forest areas that are present over
bedrocks that have undergone intense hydrothermal alter
ation. We wanted to compare forest characteristics at
approximately the same date, November 9, as that of our
two image sets from previous years.
Field survey method.Surveys were conducted
November 110, 1989, to determine species present, size,
tree density, canopy density, and foliage condition in areas
with different slopes. Transects were run in areas of
altered rocks and also on unaltered background areas;
these transects were restricted to locations where we
believed the geology to be well known. Transects were
DAN RIVER BASIN
//
%hapel;5||
As%
ft;
elegy's
&\s
O 50 100 KILOMETERS
laid out using magnetic compass, and observation sites on
the lines were spaced 20 m apart with a calibrated nylon
rope. At each field observation point, the immediately sur
rounding area was divided into four equiangular quadrants
measured from the compass bearing of the transect line. In
each quadrant, the closest canopy tree (tree whose top is
not shaded by another) was selected, the species and diam
eter 1 m above the base determined, and the distance from
the observation point measured. Observation-site-to-tree
distances were measured by tape and tree diameters by
caliper. Identification of species was made in situ, and for
a few less common trees such as Carya, only genus was
determined. Four visual approximations of leaf abundance
and greenness were made per 20 m interval by each of
two observers, and the results were averaged. Relative
greenness is a value obtained by estimating percent of
chlorophyll retention (G values of 1 to 4 indicate: 1,
125 percent of leaves present are green; 2, 2550 percent
 LANDSAT TM IMAGERY, SNOW CAMP-SAXAPAHAW AREA, NORTH CAROLINA
are green; 3, 5075 percent are green; and 4, 75100 per
cent are green), and percent of leaves remaining on tree
(L values of 1 to 4 indicate: 1, 025 percent of leaves
remain on tree; 2, 2550 percent remain; 3, 5075 percent
remain; and 4, 75100 percent remain). The relative
greenness was then derived by multiplying G times L,
yielding index values of 116, which were subsequently
normalized to 100. Percent canopy cover was determined
by observers who used sighting frames (camera view find
ers) to estimate the amount of sky covered by tree canopy.
Two observers each made four observations per site, and
the results were averaged. From the data collected in this
manner, graphs were prepared showing the abundance of
Quercus prinus (chestnut oak), relative greenness, canopy
density, and trees per hectare.
Soil samples for nutrient analysis were collected from
points on the transects in March 1990 in the manner
described in Milton and others (1983).
Time of data collection.Our study was restricted to
Landsat TM data sets already in hand, and these data sets
were obtained during the normal time of leaf coloration
and leaf fall (November 9, 1982, and November 9, 1988).
These data sets gave good results in our study, but we do
not know if a different time of year would have been
better. Forest foliage changes rapidly from day to day
during leaf senescence, and one assumes that conditions
vary somewhat on the same calendar date from year to
year. However, in applying Landsat data to forest studies,
we are comparing one part of the forest to another in
essentially an instant of time rather than comparing
against any fixed standard. Certainly some of the foliage
characteristics that we found significant may not be sig
nificant at other times of the year. The 1982 TM data
gave much better discrimination of quartz-rich soils in
tilled fields, perhaps as a result of dryer soil on the data
collection date.
Acknowledgments.-Our research was undertaken
cooperatively by the U.S. Geological Survey, Branch of
Eastern Mineral Resources, and the Servei Geolgic de
Catalunya, Barcelona, Spain. It has built upon earlier work
in the area by the U.S. Geological Survey in cooperation
with Instituto Tecnolgico GeoMinero de Espaa, Madrid.
Transportation and field expenses in North Carolina
for Pays and Bonet were supported by NATO Collabora
tive Project No. 530/88. Images for field use were pro
cessed by the Institut Cartografic de Catalunya, Barcelona,
Spain, which also transcribed Landsat TM data sets to
floppy disks for further processing in the United States.
We greatly appreciate discussions of methods and inter
pretations as well as help in the field by L.C. Rowan,
N.M. Milton, C.M. Ager, M.S. Power, M.H. Podwysocki,
Richard Brooks, Karl Shaffer, and E.M. Haskin. We are
especially grateful to the North Carolina Department of
Agriculture, Agronomic Division, Soil Testing Laboratory,
Dr. D.W. Eaddy, Director, and M.R. Tucker, Chief, Soil
191
Testing Section, for analyses of available nutrients in soils
we collected from the study area.
GEOLOGY
The study area, in central North Carolina, is under
lain by rocks that are generally typical of the Carolina
slate belt in central and northern North Carolina. The slate
belt consists of mostly arc volcanics of a variety of com
positions, and they have been intruded by several genera
tions of subvolcanic and plutonic rocks. These rocks and
the hydrothermal alteration that has affected many of them
have been discussed in greater detail by Schmidt and oth
ers (1990).
Volcanic and volcaniclastic rocks.-Roughly four
fifths of the area is underlain by volcanic and volcaniclas
tic rocks that are provisionally divided into two volcanic
rock groups (fig. 2). The oldest group is a complex of
small heterogeneous volcanic units within which only a
few units could be mapped separately, and the younger is
an extensive crystal-rich rhyodacite and dacite unit that
rests unconformably upon the older group (Schmidt and
others, 1990). Both groups are tentatively interpreted to
correspond to 600560-Ma rocks that have been dated in
other parts of the slate belt. The basaltic to rhyolitic older
group is mostly andesitic and dacitic but includes a flow
banded siliceous rhyodacite-rhyolite unit shown separately
in figure 2.
Intrusive rocks.Plutonic and subvolcanic rocks of
several ages are very common in parts of the Snow
CampSaxapahaw area. The largest plutonic bodies are
granite, quartz diorite, or tonalite, in part forming intrusive
complexes with masses and screens of hornfels and
metavolcanic rocks. Fine-grained granophyric granite and
quartz monzonite form plutons in the Cane Creek Moun
tains that probably connect to form one body at depth.
Formed at shallow depth, these plutons are surrounded by
somewhat silicified zones, their extent indicated by soils
type. All of these pre-metamorphic intrusives may have
been formed during a long span of time, but before about
320 Ma. The younger tonalite and quartz diorite are proba
bly close to 300 Ma, the approximate age of several, dated
young plutons of the slate belt. Areas mapped as little
metamorphosed tonalite and quartz diorite also include
considerable pre-metamorphism granitic rock into which
masses of the younger rock have intruded. The youngest
intrusive rocks are mafic dikes of Mesozoic age that are
not shown on figure 2.
Metamorphism.All of the volcanic rocks and per
haps three-fourths of the intrusive rocks in the area have
undergone regional metamorphism, mostly to mid-green
schist facies, with local development of new actinolite or
hornblende at a few sites. Open fractures and vugs have
192 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

7925 79 20
I
Q FG |
------------
I
- I | CAT W
U.e. GR T==
\. *
S5 i
S. --

\-. Saxapahaw
FR `----" "
N&S
KN
|FB/ | GB
\
V N DAT
GB

\ }| /
35
55

359
50'

O 3 KILOMETERS

Figure 2 (above and facing page). Geologic map of the Snow CampSaxapahaw study area showing the distribution of hydrothermal al
teration, former gold mines, and gold anomalies as determined by panning of stream sediments. Field mapping, 19851989, by Robert G.
Schmidt, Elizabeth H. Hughes, Pablo Gumiel, Alba Pays, and Carmen Anton-Pacheco. Geologic units in southeast corner of map adapted
from Wilkinson (1978).

been eliminated in the strongly altered rocks, thus chang
ing their appearance substantially from that of unmetamor
phosed equivalents as observed in other regions. The
youngest plutonic and subvolcanic granitoid intrusives are
essentially unmetamorphosed.
Structural geology.Volcanic rocks of all of the
Snow CampSaxapahaw area have open to tight folding,
but recognizable layering is so rare that folding is not
mappable. Straight, northeast-trending, linear features sug
gest faults and shear zones, and many of those checked in
the field were found to be zones of sheared or brecciated
rocks. A few north- and northwest-trending faults are sug
gested by offset alteration zones and rock units, but none
of these were specifically identified in the field.
Hydrothermal alteration.Extensive hydrothermal
alteration has taken place in the volcanic and intrusive
 LANDSAT TM IMAGERY, SNOW CAMP-SAXAPAHAWAREA, NORTH CAROLINA
EXPLANATION
VOLCANIC ROCKS
REGIONAL METAMORPHISM-----------------------------------
[RD) Crystal-rich rhyodacite/dacite with many
small inclusions, very uniform texture
and composition throughout. Thin cob
ble-rich fragmental layer at base not
shown separately here
------------------------ UNCONFORMITY------------------------>
DAT Heterogeneous mostly dacitic and andesitic
FR tuffs, part fragment-rich; includes
basaltic to rhyolitic units, includes dis
tinctive lenses of flow-banded rhyo
dacite/rhyolite (FR) that are interpreted
to thicken in northwest corner of area
(Cane Creek Mountains)
HYDROTHERMALLY
ALTERED ROCKS
Quartz-sericite-pyrite and quartz-paragonite
rocks; also includes potassic and epidote
rich alteration near plutons. Protolith tex
tures generally destroyed in central parts
of zones
Intensely altered very siliceous central core
zones and associated pods of pyrophyl
lite-andalusite-pyrite rock
OTHER AREAS
Areas unmapped in this study, mostly inter
mediate volcanic rocks
rocks, and it is most significant in the oldest units where
alteration is of the high-sulfidation type as defined by
White and Hedenquist (1990). The very large Snow
CampSaxapahaw alteration systems make up a northeast
trending zone about 13 km long and up to nearly 4 km
wide (Hughes, 1987). They share many characteristics
with over 40 similar areas in the Carolina slate belt
(Schmidt, 1985; Schmidt and Klein, 1985).
The most intensely altered parts of the systems, pre
sumably innermost as the systems were formed, consist of
large volumes of almost pure, fine-grained quartz rock.
Near the edges of these cores are several smaller, in part
minable, masses of pyrophyllite-quartz, including various
amounts of pyrite and andalusite. These core zones are
193
INTRUSIVE ROCKS
Tonalite and quartz-diorite, little metamor
phosed. Mapped areas include Some
areas of pre-metamorphism plutonic
rock as well
Quartz diorite/hornfels/volcanic rock injec
tion complexes
[HD] Hornblende granodiorite
Granite, quartz diorite and tonalite, plagio
clase oscillatory zoning destroyed by
regional metamorphism
Fine-grained granophyric granite and
quartz-monzonite
Gabbro, porphyritic gabbro, and horn
blende gneiss
~~~ Geologic contact, dashed where inferred
Inferred fault
MINES AND PROSPECTS
O Pyrophyllite or sericite mines and prospects
N Gold prospects and Small mines
X Panned stream-sediment concentrate with
visible gold
now mostly surrounded by extensive envelopes of
quartzwhite mica (sericite and paragonite) altered volca
nic and subvolcanic rocks. Propylitic alteration probably
formed a zone exterior to the quartzwhite mica envelope,
but we were not able to map it separately from the sur
rounding volcanic rocks that have undergone greenschist
facies metamorphism.
Metallic mineralization and pyrophyllite deposit.
Presently recognized gold and silver mineralizations are
mostly in areas of strong quartzwhite mica and probably
propylitic alteration. Such mineralization is also found in
some narrow zones along northeast-trending shears and is
generally accompanied by quartz-sericite alteration (fig. 2)
(DAgostino and Schmidt, 1986; Schmidt and others,
194 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 1. Nutrient elements available to plants in composite sam composition and condition can be recognized in pro
ples of A-horizon soils. cessed Landsat TM imagery. At the same time, this
[Analyses by Soil Testing Section, Agronomic Division, North Carolina Department
application has greatly increased the need for better
of Agriculture. The soil test indices and factors for conversion of indices to parts per understanding of the processes involved.
million are used as defined in the Soil Testing Section, Standard Soil Test Report]
The most common soil types in the area are the Hern
don, Goldston, and Goldsboro silt loams. Although certain
Substrate
Unaltered Probably Strongly
mapped soil types correlate well with specific mapped
rocks altered rocks altered rocks bedrock, the quartz-rich soil types associated with highly
(2 samples) (2 samples) (16 samples) altered bedrock are also commonly present on other rock
substrates.
Cation exchange capacity
(CEC, in meg/100 cm)..... 3.3, 4.4 3.3, 3.5 2.34.3 Bedrock-controlled soil characteristics that strongly
avg 3.3 influence forest condition in this area include available
Base saturation nutrients, textures, and trace metals, as well as drainage,
(percent of CEC) .............. 36, 40 21, 28 720 rockiness, soil depth, and slope direction.
avg | 1.3 We suspected intuitively that the core-zone soils
pH.... 4.5, 5 44, 4.4 3.84.6 lacked available nutrients because of the siliceous nature
avg 4.2 of the altered bedrock and the slow growth of trees on
Phosphorus index x these sites, but standard total chemical analyses were inad
conversion factor (ppm)...... 1, 5 l, 4 15
equate to measure the soil nutrients available to growing
Potassium index, converted to
plants. Consequently, 16 soil samples, each a composite of
available K (ppm)............. 63,35 70, 78 3163
avg 44 more than 10 subsamples taken within a 10 m circle, were
Calcium (percent of CEC) obtained from the A horizon, below the organic layer, in
converted to areas of strongly hydrothermally altered bedrock, and 4
available Ca (ppm)......... 159, 18] 60, 100 2080 soil samples were obtained from areas of relatively weak
avg 28 alteration or unaltered rock. These were analyzed for a
Magnesium (percent of CEC) variety of constituents (table 1), and several distinct and
converted to consistent variations were found between areas of unal
available Mg (ppm).......... 62, 53 29, 29 1020 tered and highly altered bedrock; however, we recognize
avg 15 the limitations of our small data set for unaltered rock.
Manganese index
Soils and vegetation, areas of unaltered bedrock.
(converted to ppm).......... 98, -100 30, 20 145
Most areas of volcanic and plutonic rocks in the Carolina
avg 5
Copper index
slate belt are overlain by soils that support healthy natural
(converted to ppm)........... 0.6, 0.4 0.2,0.4 0.21.0
forests with many species of deciduous and evergreen
avg ().4 trees and understory trees and shrubs. We recorded 1015
species on single transects in normal forests, and a total of
24 species were recorded in all transects on unaltered
rocks, not including shrubs and understory trees. The can
1987). Several spotty occurrences of anomalous gold con opy is generally closed, but when openings are present,
centrations have been found in high-silica core-zone rocks they permit a dense second growth of canopy-tree saplings
as well. Pyrophyllite ore was extracted from one major and understory species.
quarry and two smaller pits in the area. In five forest transects (252 trees observed) in areas
of unaltered rocks, no species exceeds 19 percent (table
2). The seven most abundant species are: Acer rubrum,
BEDROCK, SOILS, AND FOREST 19.0 percent; Quercus alba, 17.5 percent; Q. stellata, 10.0
percent; Q. velutina, 9.0 percent; Carya sp., 6.0 percent;
VEGETATION ASSOCIATIONS
Liquidambar styraciflua, 4.8 percent; and Liriodendron
tulipifera, 4.0 percent. Typical understory trees and shrubs
General correlations of bedrock type, especially include Ilex, Vaccinium, Gaylussacia, Azalea, Rhododen
hydrothermally altered rocks, and forest tree communities dron, Hamamelis, and Oxydendrum.
have been understood in a general way and have been Nutrient analyses in the normal forest, especially for
used in prospecting for pyrophyllite for many decades. calcium and magnesium, tend to be relatively low (table
By using Landsat remote-sensing techniques during geo 1) but seem adequate to permit healthy forest growth. In
logic mapping in the region, we have been able to greatly normal forests (areas of unaltered rock) that were not on
expand the known areas of hydrothermal alteration any transect, metals analyses ranged from below detection
because, to an important degree, forests of unique up to (in ppm), Cu = 120, Mo = 27, Pb = 39, and Zn = 73
 LANDSAT TM IMAGERY, SNOW CAMPSAXAPAHAW AREA, NORTH CAROLINA 195

Table 2. Tree species abundance related to bedrock type. on highly altered bedrock, the spectral reflectances in TM
bands 4 and 5 are reduced there about 30 percent; band 3
[All numbers, except total number of trees in sample, in percent. Rk, rock; sl, slopes)
is reduced about 20 percent; but band 2 is reduced only
Underlying bedrock type and slope about 4 percent (fig. 4). Our estimates of greenness,
Species Unaltered rk Strongly altered rock admittedly somewhat subjective, indicate more chlorophyll
All All South, North, remaining in the canopy on level and sun-facing slopes,
slopes slopes SE. Sl N.W. sl
which one would expect to reduce spectral reflectance
Quercus prinus 1.2 58.5 53.9 55.0 there (fig. 4, transects L and D-1).
alba 17.5 10.3 14.4 10.6
In soil analyses made to determine available nutri
falcata 3.6 2.4 1.1 3.3
ents, base-exchangeable Ca and Mg in hydrothermally
velutina 9.1 5.8 5.6 6.1
marilandica 1.2 1.2 1.1 1.7
altered areas were found to be only fractions (one-sixth
stellata 10.0 1.2 0.6 1.1 and one-fourth) of the amounts present in forest soils
formed on unaltered volcanic rock substrates (table 1),
phellos 3.2
Fraxinus sp. 3.6 which are already relatively poor in Ca and Mg, thus sug
Prunus serotina ().8 gesting a significant deficiency in these important nutri
Acer rubrum 19 () 6.0 8.3 5.6 ents. The relative amounts of other available nutrients
Pinus virginiana ().8 3.4 4.4 3.3 determined do not seem significantly different from those
echinata 2.8 5.8 5.() 5.6 in nearby forest soils over unaltered bedrock.
taeda 0.4
Maximum metal values within areas of strong hydro
Liriodendron tulipifera 4.0 0.4 (),6
thermal alteration and anomalous forest vegetation ranged
Liquidambar styraciflua 4.8
from below detection to the following (in ppm): Cu = 49,
Ulmus sp. 1.2
Pb = 30, Zn = 20, and Mo = 7.3. These contrast with the
Juniperus virginiana 2.4
Cornus florida 3.6 greater range (from below detection to greater amounts for
Oxydendrum arboreum 3.2 2.0 1.7 2.8 each metal) in areas without anomalous vegetation. Thus,
Nyssa sp. 0.4 1.2 1.1 2.2 the distribution of these metals in the Snow CampSaxap
Carya sp. 6.0 1.6 2.8 1.7 ahaw area does not appear to be a factor in the growth of
Fagus grandifolia 1.2 0.2 0.6 the anomalous forests.
Total 100.0 100.0 100.0 100.2
Although there seems to be no significant difference
in tree species present on north and northwest slopes as
Total trees in sample 252 5()4 18() 180
compared with sun-facing slopes (table 2), visual exami
nation of the forests indicates considerable difference in
the condition of the trees in these two topographic situa
(all total element present) without observable effects on tions. While the forest canopy in both tends to be rela
the condition of trees.
tively open and the understory vegetation very sparse
Soils and vegetation, areas of hydrothermally altered despite the favorable light conditions, trees on the south
bedrock.The forest community common on areas of facing slopes are more spreading, slower growing, and
intensely hydrothermally altered bedrock is noticeably less some have many dead or weakened limbs. In contrast,
diverse than the surrounding forest (table 2, fig. 3). Some trees on the north-facing slopes are of normal height and
what fewer total species are present411 on each of robustness. It would appear, therefore, that differences in
seven transectsand of these, a small number generally the spectral response between the two slope environments
dominate. Only six species, Quercus prinus, Q. alba, Q. recorded in Landsat data depend on forest conditionthis
velutina, Acer rubrum, Pinus virginiana, and P. echinata may include tree structure that permits more sunlight on
make up 89.8 percent of the 504 trees sampled. However, the sunny-slope forest floor, although this effect would be
the most noticeable difference is in the abundance of
partly counteracted by more green leaf retention.
Quercus prinus, the chestnut oak. This species makes up
most of the canopy forest over extensive areas at these
sites and averages 58 percent of the forest for all of the
DATA ANALYSIS
altered area sampled (table 2). Furthermore, forests at
these sites are characterized by a relatively open canopy
and lack of a significant understory. Except on transect B, Geobotanical data collected in the field were com
fewer species were present on the northwest-facing slope pared with bedrock geology and with spectral reflectance
than on the sun-facing slope. The north slope of transect D data collected by satellite (Landsat Thematic Mapper). Of
had only four species (fig. 3). Because species composi 12 transects surveyed November 110, 1989, six were
tion of the forests on north and northwest slopes is much selected for final analysis on the basis of completeness of
like that on the sunny slope and remains typical of forests data and the representative nature of the forests.
 R
T
C
E
S
N
A
LD1
D2
level)
(relatively
(south
slope)
(north
Substrate
Unaltered
Strongly
Altered
Substrate
No
prinus
Quercus
present
density
Tree
and
abundance
600
600
600

e
hectare
per
Trees
soo
Le-Trees
per
hectare
hectare
per
eTrees
H.
500
500
prinus,
E
Q.
E
of
number
out
trees
4psite
ofer
400
-
H.
400
400

k
300 H.
300
300
g00
.S
S.200

H.
200
200
-4
-4
G
+3
-
100
100
H.
100
H2
$
oII
L
l
o
EO1
ll
L
I11
209
L208
213
212
21.1
210
217
216
215
214
89
90
D91
84
85
86
87
88
->
81
82
83
95
94
93
D92
OO
99
98
97
96
10
2.
Site
number

SPECIES
(Canopy
PRESENT
AT
trees)
OBSERVATION
SITES
,(7.5%)
(2.3%)
(6.7%)
(11.4%)
(2.5%)
sp.
Carya
Q.
ego
marilandica
P.
echinata
(1.7%)
P.
echinata
Liriodendron
P.
virginiana
tulipifera
(4.5%)
..
styraciflua
P.
virginiana
(3.3%)
J.
virginiana
(2.3%)
(5.0%) Nyssa
sp.
(1.7%)
falcata
Q. Prunus
serotina

(23.2%)
(79.5%)
stellata
Q.
prinus
Q.

(14.9%)
sp.
Fraxinus
(1.7%)
(1.7%)
marilandica
Q.
phellos
Q.

Trees
3.
Figure
Quercus
and
hectare
per
distribution
prinus
species
figures),
(top
three
for
abundance
conditions.
forest
typical
L epresents
rorest
afTransect
level
on
ground
unaltered
represents
D-1
sbedrock;
aftransect
non outh-facing
highly
with
slope
and
altered
corest
is
D-2
ontinuation
D-1,
of
bedrock,
similar
orth-facing
genera
Some
slope.
as
abbreviated
are
J,
Cornus;
C,
follows:
Liquidambar;
L.
Juniperus;
Quercus.
Q,
Pinus;
P:
 3
Go
go
co H
E 4-
$ -
2c.
-
400 ||
20
20
W*--:
*. 3Tw-e
o0
30
g
III o2
>
2OO S.
1
canopy
of
Percent
e
-e-Relative
greenness 330
-
C.
0 28
-g
25
5 2-0O
go --
0. eolorest
Site
4.
rFigure
cTM
forest
typical
three
for
cover
canopy
of
percent
greenness,
relative
fvalues,
a
represents
L.
Transect
and
ground
level
onnfedlvieatcioansc.e, of
fornuorest
represents
D-1
csunaltered
is
D-2
transect
and
altered
highly
nslope
a
on
bedrock,
similar
with
D-1,
thi-nfuactinog
230
- 8
as
3so
+35H
H
canopy
of
Percent
---
s5o
; 2
O 5
1
o

greenness-
Relative
-e- 440
70
* 0
77o

||
TM-2
-- 0TM-4
Elevation
-e-
--TM-4
70 -

5
TM-2
---
TM-5
e-
TM-5
Q_

numbers
Site
slope)
(north numbers
Site

D2
Substrate
Altered
Strongly

211
210
209
L2O8
217
216
215
214
213
212
88
89
82
83
64
85
97
96
90
D91
at
94
as
87
93
D92
95
101
100
99
98

80
90
Bo
250
80
250
250 TM-3190
-- 1*-*
0-1o
O -11-1-1-1-1 80
Bo
45
A8O
45 P.
70
7o -

Reflectance
Spectral
Sites
of
Elevation
and
15
*1"
tts-t-t-t-t-t-t-ta-.
E,
l
1210; -

greenness -
Relative
e- canopy
of
Percent
I
l
1
I
L
l

C
E
S
N
A
R
T Canopy
and
Greenness

*T*T-
*-*---e.
w-w-w
"H

--
Elevation
-e-
TM-4
10
-

numbers
Site
slope)
(south -
TM-2
---
TM-5
e-200

D1

81
82
83
84
85
86
87
88
89
90
D91

s_eeo
:so
230
so
3
|

Substrate
Unaltered

(relatively
level)
numbers
Site

5
Elevation
-e-
M-4
||
-*-
240
-2-
-24040
L

55F
2
50
30
30 20 slope.
>
co co
* 9
40
o
40 > 85
c
198 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

(351). WLZ
PA

1 KILOMETER

Figure 5. Detail of a digitally classified image of the Major Hill area superimposed on a stan
dard topographic map. Green, ordinary forest; red, forest on high-silica or altered rock substrate;
yellow, ordinary bare earth or stubble field or green, cropped field; dark blue, water. Geologic
units are coded as follows: RD, crystal-rich rhyodacite/dacite; DAT, dacitic and andesitic tuffs;
FR, flow-banded rhyodacite/rhyolite; QS, quartz-sericite-pyrite altered rock; QC, quartz-rich core
zone. In our trials, a supervised classification using a maximum likelihood method resulted in im
ages that most accurately represented forested areas above altered bedrock or highly siliceous bed
rock and general cultural features. Some, if not most, of the areas that seem to be falsely classified
as forest on altered or high-silica bedrock are mixed pixels of forest and field or grassed areas.
These generally include pasture with scattered trees (probably areas P, P', P) or unusually open
forest (F). Falsely classified forests marked F are not explained. C, D, K, and I are botanical
transects. The image was generated on a 286-type personal computer using the Dragon Image Pro
cessing System and a maximum likelihood classifier. The circles show typical areas used to obtain
characteristic TM reflectance values for hydrothermally altered rock (circles 1, 2, and 3) and nor
mal forest (circles 4, 5, and 6). Statistical characteristics for the supervised classification of field
and water areas were extracted in a similar manner.

Digital values for TM bands 2, 3, 4, and 5 along the
transects were extracted by generating an enlarged portion
of a classified image on a CRT screen and superimposing
a topographic map on which the transect had been plotted
in the field. Some inaccuracy in transect location is
assumed because the 1:24,000 topographic maps are
imperfect representations of minor land-surface details,
and some areas lacked features that could be used as exact
reference points. Digital numbers were extracted for each
TM band for each pixel on the transect line and for one
pixel on each side. Reflectance values that were finally
used were moving averages on the center points of nine
point squares. Spectra for three typical transects are
shown in figure 4.
Influence of slope direction on forest characteris
tics.-Digital classification of Landsat data can be used to
identify forests growing on strongly altered rock substrates
on level ground and sun-facing slopes but is much less
reliable on north-facing slopes. Five transects were laid
out on areas of highly altered rocks, each one including
both south or southeast slopes and north- or northwest
facing slopes, but full data for greenness index and canopy
were not collected for three of the transects. These slopes
range from about 1:4.5 to 1:8. Transect D, the one judged
to have the most typical forests throughout, is shown
graphically in figures 3 and 4 and on a detailed segment
of a classified image with topography superimposed (fig.
5). The abundance of Quercus prinus on the north slope is
similar to that on the south slope (fig. 3); and the percent
age of forest canopy is about the same, but there are sig
nificant differences in the relative greenness (table 3).
Reflectance spectra of the north slope more closely resem
ble those of transect L and of other areas with unaltered
bedrock than they do those of the south slope of transect
D (fig. 4). We believe that the anomalous forests can be
distinguished by their TM reflectance spectra where the
 LANDSAT TM IMAGERY, SNOW CAMPSAXAPAHAW AREA, NORTH CAROLINA 199

Table 3. Relative greenness of canopy forests on contrasting Mg in the soil, (2) a thin and (or) very rocky soil that
slope environments. accentuates moisture stress during dry periods, and (3)
position of the forest on top of a dry ridge or on a south
Date of Mean
or southeast-facing slope, causing additional moisture
Group Transect observation relative
stress. The many ridge crests and south- or southeast
(Nov. 1989) greenness
facing slope areas identified as anomalous by digital
Hydrothermally altered substrate
SE. slopes D-1 5 44.7
classification had all three factors present. Other anoma
K-2 8 26.1 lous ridges, generally with Quercus prinus forests, are
Relatively level F 6 35.1 underlain by siliceous rocks that are either unaltered or not
I 7 25.0 more than moderately altered.
NW, slopes K-1 8 29.1 Metals present in the hydrothermally altered rocks do
D-2 5 22.7 not appear to be a stress-causing factor in this study area.
Unaltered substrate Several areas have been identified where the species
E 5 24.6 present reliably reflect the strongly altered rock substrate
J 7 21.1 but where, due to a location on north- or northwest-facing
L 9 20.0 slopes, water stress is presumably less and the forest floor
G 6 5.3 less illuminated in mid-morning (the time of satellite data
collection). These sites have reflectance spectra closer to
those of a normal forest.
The Landsat data sets available to us were collected
relatively open canopy combined with flat or sun-facing
slope permit good illumination of the ground surface, irre during periods of partial leaf coloration and leaf fall. All
other experiments with which we are familiar using digital
spective of the measured greater greenness there. We also
classification of images of southeastern Piedmont forests
believe that more bare ground is exposed on the sun
have used autumn or winter data sets. Obviously, data sets
facing slopes. Not surprisingly, spectra characteristic of
obtained earlier in the growing season need to be studied
anomalous forests are also obtained in open, forested pas
before we will know what part of the year gives the best
tureland and where individual pixels are made up of part
forest and part field. images for these geologic purposes.
Digital classification.Detailed results of a maxi
mum likelihood digital classification of a small part of REFERENCES CITED
the area are shown in figure 5. Identification of anoma
lous forests growing on highly altered substrates is quite Braun, E.L., 1950, Deciduous Forests of Eastern North America:
reliable on ridge crests and sun-facing slopes, but the sim Philadelphia, Blakiston, 596 p.
ilar forests on north and northwest slopes are generally not D'Agostino, J.P., and Schmidt, R.G., 1986, Gold in panned concen
differentiated. Anomalous bedrock classifications in trates from southern Alamance County, central North Caroli
unaltered areas are believed to be related to very siliceous na: U.S. Geological Survey Open-File Report 86453, 5 p.
bedrocks, and some appear to be edge effects caused by Hughes, E.H., 1987, The geology and alteration centers of the
pixels including more than one surface type. A few unex Snow Camp mineMajor Hill area, central Carolina slate belt,
Alamance and Chatham Counties, North Carolina: U.S. Geo
plained anomalous areas remain to be field checked.
logical Survey Open-File Report 87180, 29 p.
Milton, N.M., Collins, William, Chang, Sheng-Huei, and Schmidt,
R.G., 1983, Remote detection of metal anomalies on Pilot
CONCLUSIONS Mountain, Randolph County, North Carolina: Economic Geol
ogy, v. 78, p. 605617.
Schmidt, R.G., 1985, High-alumina alteration systems in volcanic
Many areas of intensely hydrothermally altered rocks rocks and their significance to mineral prospecting in the Caro
of the Carolina slate belt can be identified on processed lina slate belt: U.S. Geological Survey Bulletin 1562, 59 p.
Landsat TM imagery, thus providing a useful aid to min Schmidt, R.G., and Klein, T.L., 1985, High-alumina alteration sys
eral exploration. Our data confirm that forests in areas of tems and their role in mineral exploration, southeastern United
intensely altered rocks are characterized by a diminished States, in Krafft, Kathleen, ed., USGS Research on Mineral
number of species and by dominance of Quercus prinus, Resources1985, Program and Abstracts: U.S. Geological
Survey Circular 949, p. 4748.
the chestnut oak, as well as by differences in tree
Structure. Schmidt, R.G., and Koslow, M.H., 1986, Preliminary map showing
the occurrence of siliceous rocks in central North Carolina de
We believe that three main factors contribute to the
rived from enhanced principal component Landsat images:
forest conditions that can be recognized in Landsat TM U.S. Geological Survey Miscellaneous Field Studies Map
data. These are: (1) seriously diminished available Ca and MF1817.
200 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Schmidt, R.G., Pays, Alba, Gumiel, Pablo, and D'Agostino,
J.P., 1987, The Saxapahaw, North Carolina, gold occurren-
cesSearch for a deposit model and new remote sensing
techniques, in Sachs, J.S., ed., USGS Research on Mineral
Resources1987, Program and Abstracts: U.S. Geological
Survey Circular 995, p. 6263.
Schmidt, R.G., Gumiel, Pablo, and Pays, Alba, 1990, Provisional
geologic map of the Snow CampSaxapahaw area, North
Carolina: U.S. Geological Survey Open-File Report 90417.
White, N.C., and Hedenquist, J.W., 1990, Epithermal environments
and styles of mineralization: Variations and their causes, and
guidelines for exploration: Journal of Geochemical Explora
tion, v. 36, p. 445474.
Wilkinson, S.E., 1978, The geology of the northeast quarter of the
Silk Hope quadrangle, Carolina slate belt, North Carolina:
Chapel Hill, University of North Carolina, unpub. M.S. thesis,
56 p.
 CHAPTER T
AEROMAGNETIC SURVEY OF
NORTH-CENTRAL MINNESOTA
By ROBERT J. HORTON, WARREN C. DAY, and ROBERT E. BRACKEN"
INTRODUCTION
As part of the International Falls and Roseau Conter
minous United States Mineral Assessment Program (CUS
MAP) projects, the U.S. Geological Survey (USGS)
conducted a detailed aeromagnetic survey over an exten
sive area in north-central Minnesota (fig. 1). Greater than
95 percent of the bedrock in the survey area is covered by
Quaternary glacial deposits. Information from available
drill holes (Klein and Day, 1989) and sparse outcrop data
are inadequate for geologic mapping and resource assess
ment. The aeromagnetic data, therefore, played a crucial
role in the interpretation of the bedrock geology for these
CUSMAP projects.
AEROMAGNETIC SURVEY
The 60,000 line-kilometer survey was conducted
from 1984 through 1986. The aeromagnetic data were col
lected using a Fairchild Heli-Porter research aircraft, oper
ated by the USGS, Branch of Geophysics. For the north
central Minnesota aeromagnetic survey, the primary flight
line direction was north-south, with a flight-line spacing of
about 380 meters. The survey altitude was radar controlled
at about 91 meters above ground level. The magnetic field
was measured using a Geometrics G-813 proton preces
sion magnetometer, with a cycle time of 0.5 seconds and a
sensitivity of 0.5 nanoTeslas. The average speed of the
aircraft was 90 nautical miles per hour. The flight path
recovery was accomplished using visual and electronic
(LORAN and Mini-Ranger) methods.
"U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225
*U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225
The color shaded relief aeromagnetic map (fig. 2)
was generated from a gridded data set (Bracken, 1991)
that was produced from the flight-line data (Bracken and
Petrafeso, 1991). The apparent relief is produced by com
puter-generated illumination from the southwest. The
color-shading technique emphasizes subtle level changes
in anomaly amplitude and enhances linear features such as
faults and dikes that are aligned at oblique angles to the
illumination direction.
GEOLOGY
The survey area is located in the Superior province of
the Canadian Shield. Westward-thickening Quaternary
deposits, ranging from less than 1 to greater than 60
meters thick, overlie most of the Precambrian bedrock
within the survey area (Horton and others, 1989). Geo
logic mapping by Day and others (1990a, 1990b) has
shown that this Late Archean terrane contains rocks of
three subprovinces: the Wabigoon in the north, the Quet
ico in the central part, and the Wawa-Shebandowan in the
south (fig. 1).
The Wabigoon subprovince is a granite-greenstone
terrane composed of volcanic and coeval plutonic rocks
and volcaniclastic, epiclastic, and chemical sedimentary
rocks. The Quetico subprovince is composed of metasedi
mentary rocks, gneiss, and potassic granite. The Wawa
Shebandowan subprovince is a granite-greenstone terrane,
composed of metavolcanic rocks, interlayered metasedi
mentary rocks, and granitoid intrusions. An Early Protero
zoic dike swarm strikes northwest across the three older
subprovinces (Southwick and Day, 1983).
MAGNETIC CHARACTERISTICS
The rocks of the Wabigoon subprovince have a wide
range of magnetic characteristics. The granitic rocks of the
202 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

9315'

Minn

STUDY AREA
-**
* -->

Sl
E:
AND of
-

. ..
15 3O KM

Figure 1. Diagram showing the location of the north-central Minnesota aeromagnetic survey and major lithotectonic terranes and
StructureS.

Wabigoon generally produce broad, low-amplitude mag
netic anomalies. The Wabigoon greenstone belts produce
the relatively long, narrow, high-amplitude anomalies that
are visible across the eastern third of the survey's northern
border and along the west-central boundary of the survey
area. Within these greenstone belts, iron formations pro
duce the thin, highly magnetic anomalies. Mafic intrusive
bodies produce two irregularly shaped, moderately high
amplitude anomalies seen in the north-central portion of
the survey area.
The metasedimentary rocks of the Quetico subprov
ince produce the low-gradient, low-amplitude magnetic
terrane across the central portion of the survey area.
Within this metasedimentary terrane, an east-west-striking
magnetic high is produced by the migmatites of the Late
Archean Vermilion Granitic Complex (Southwick and
Sims, 1980), which are more magnetic than the metasedi
mentary units. Two prominent west-northwest-trending
Archean dikes, located in the western portion of the Quet
ico subprovince, are offset by younger, northwest-trending
Proterozoic dikes.
The rocks of the Wawa-Shebandowan subprovince
also have a wide range of magnetic characteristics. As in
the Wabigoon subprovince, the granitic rocks generally
produce broad, low-amplitude magnetic anomalies,
whereas the greenstone belts produce the relatively long,
 AEROMAGNETIC SURVEY OF NORTH-CENTRAL MINNESOTA 203

93 15'

4730' +

Figure 2. Color shaded relief aeromagnetic map of the north-central Minnesota survey. The color changes represent broad changes in
magnetic field strength. The warm colors (red, orange, and yellow) represent areas of magnetic field highs, the cool colors (green, blue,
and purple) represent magnetic field lows. The apparent relief is computer generated by illumination from the southwest. The illumination
emphasizes subtle level changes in magnetic amplitude and enhances features that are aligned oblique to the illumination direction.

narrow, high-amplitude anomalies. Iron formations pro
duce thin, high-amplitude magnetic anomalies. The large,
broad feature in the southwestern corner of the survey area
is produce by the Bemidji batholith. Numerous low-ampli
tude, subrounded anomalies are produced by granitoid
intrusions. Many of these granitoid intrusions are partially,
or wholly, encircled by narrow magnetic highs. These
narrow, encircling magnetic features are usually caused by
magnetic units within the enclosing greenstone terrain.
However, in the southeastern corner of the subprovince,
the magnetic ring around the elliptical anomaly (which is
hosted within the northeast-trending magnetic high) is
thought to be produced by a magnetic border phase within
the pluton. In the extreme southeastern corner of the
mapped area, the irregularly shaped, high-amplitude
anomaly is produced by an iron formation deformed by
faulting and intense folding.
Several prominent structural features are readily seen
in the aeromagnetic map (figs. 1 and 2). The Vermilion
fault strikes west-northwestward across the center of the
survey area. Faults that separate the rocks of the Wabi
goon and Quetico subprovinces (the Rainy LakeSeine
River and Four Town faults) lie within the high, northeast
tending gradients produced by the magnetic greenstone
204 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
terrane in contact with metasediments. The Early Protero
zoic mafic dikes are readily apparent and can be seen cut
ting across major structural features and the three
subprovince terranesthese dikes produce narrow, north
west-tending, linear anomalies.
SUMMARY
The north-central Minnesota airborne survey covered
an extensive area where bedrock is buried by thick glacial
deposits. Reliable geologic maps could not be produced
based on the available outcrop and drill hole information;
however, aeromagnetic data provided a means of mapping
buried bedrock features, such as contacts and structures,
by their magnetic characteristics. The high resolution of
the survey generated data that is detailed enough to pro
duce 1:24,000-scale aeromagnetic maps that greatly aid in
geologic mapping. This survey demonstrates that aero
magnetic data, when used in conjunction with drill-core
information, geochemical studies, and other geophysical
methods, can be used to interpret bedrock geology and aid
in mineral assessment in terranes with little or no bedrock
exposure.
REFERENCES CITED
Bracken, Robert, 1991, Description of magnetic tapes contain
ing gridded aeromagnetic anomaly data from north-central
Minnesota: Sioux Falls, South Dakota, USGS-EROS Data
Center, Data release serial no. D91-0597.
Bracken, Robert, and Petrafeso, Frank, 1991, Description of mag
netic tapes containing north-central Minnesota aeromagnetic
flight line data: Souix Falls, South Dakota, USGS-EROS Data
Center, Data release serial no. D91-0439.
Day, W.C., Klein, T.L., and Schulz, K.J., 1990a, Preliminary bed
rock geologic map of the Roseau 192 quadrangle, Minneso
ta, U.S.A., and Ontario, Canada: U.S. Geological Survey
Open-File Report 900544A and 900544B, scale 1:250,000.
Day, W.C., Southwick, D.L., Schulz, K.J., and Klein, T.L., 1990b,
Bedrock geologic map of the International Falls 1x2 quad
rangle, Minnesota, U.S.A., and Ontario, Canada: U.S. Geolog
ical Survey Miscellaneous Investigations Map Series
I1965B, scale 1:250,000.
Horton, R.J., Meyer, G.N., and Bajc, A.F., 1989, Reconnaissance
Quaternary geology map of the International Falls 1x2quad
rangle: U.S. Geological Survey Open-File Report 89654,
Scale 1:250,000.
Klein, T.L., and Day, W.C., 1989, Tabular summary of lithologic
logs and geologic characteristics from diamond drill holes in
the western International Falls and Roseau 192 quadrangles,
northern Minnesota: U.S. Geological Survey Open-File
Report 89346.
Southwick, D.L., and Day, W.C., 1983, Geology and petrology of
Proterozoic mafic dikes, north-central Minnesota: Canadian
Journal of Earth Sciences, v. 20, no. 4, p. 622638.
Southwick, D.L., and Sims, P.K., 1980, The Vermilion Granitic
ComplexA new name for old rocks in northern Minnesota:
U.S. Geological Survey Professional Paper 1124A, p.
AlA11.
 CHAPTER U
MINERALOGIC AND FLUID-INCLUSION
STUDIES OF THE PEA RIDGE
IRON-RARE-EARTH-ELEMENT DEPOSIT,
SOUTHEAST MISSOURI
By GARY B. SIDDER, WARREN C. DAY, LAURENCE M. NUELLE,
CHERYL M. SEEGER, and EVA. B. KISVARSANYI?
ABSTRACT
Iron ore and reserves of rare-earth-element ore at the
Pea Ridge mine occur in what is thought to be an eroded
volcano-plutonic complex in the St. Francois Mountains
of southeast Missouri. Middle Proterozoic rhyolitic tuffs
host the iron ore. Rare-earth-element-bearing breccia
pipes cut the iron ore and altered wallrocks. Comagmatic
aplite dikes cut the deposit, which suggests that the min
eralization was contemporaneous with the igneous activ
ity in the St. Francois terrane.
The Pea Ridge deposit consists of four mappable
zones: the amphib ole-quartz, magnetite, hematite, and
silicified zones. Fluid-inclusion studies indicate that the
deposit formed from highly saline, high-temperature,
magmatically derived hydrothermal fluids. The mineral
izing fluids are possibly related to rocks of an iron-rich
trachyte suite, which is present regionally as ring intru
sions associated with the volcano-plutonic complex.
Halite homogenization in three-phase fluid inclusions in
the amphibole-quartz, magnetite, hematite, and silicified
zones indicate that temperatures varied from greater than
530C to about 400C during deposition and alteration
associated with the magnetite ore. The rare-earth
element-bearing breccia pipes formed as late-stage fluids,
'U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225
*Doe Run Co., P.O. Box 500, Viburnum, MO 65566
*Division of Geology and Land Survey, Missouri Department of
Natural Resources, P.O. Box 250, Rolla, MO 65401
derived from the magnetite ore system, that boiled. The
breccia pipes were rapidly and explosively emplaced at a
temperature of about 300C.
INTRODUCTION
This study represents a contribution from the ongoing
Midcontinent Strategic and Critical Minerals Program's
Olympic Dam Project, which is a cooperative effort
between the U.S. Geological Survey and the Missouri
Geological Survey. The project is funded through the
USGS Office of Mineral Resources. Research on the Pea
Ridge deposit was an outgrowth of work by Sims and oth
ers (1987), who recognized that the Middle Proterozoic
granite-rhyolite terranes in the Midcontinent may host
Olympic-Dam-type copper-uranium-gold-silver deposits
(Reeve and others, 1990).
A critical facet of the study of the Pea Ridge iron
rare-earth-element deposit is to establish the type, compo
sition, and paragenesis of ore and alteration minerals.
Detailed underground mapping combined with petro
graphic and electron microbeam analyses have docu
mented the paragenesis in the Pea Ridge deposit and the
nature of the ore and alteration minerals. This paper
highlights new ore-petrography, mineral-chemistry, and
fluid-inclusion data in an attempt to assess the conditions
of ore formation. The reader is referred to Nuelle and
others (1992) for a detailed description of the ore deposit,
mineral paragenesis, and ideas on formation of the rare
earth-element-bearing breccia pipes.
206 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
38H
OBonne Terre
O 20 KILOMETERS
EXPLANATION
Z% Granite central pluton Active underground mine
N Trachyte ring intrusion O Inactive mine
s.ss. Granite ring intrusion A Undeveloped deposit
Subvolcanic granite
. . . . . . . . . . Rhyolite
Figure 1. Geologic map of the Middle Proterozoic St. Francois
terrane, southeast Missouri (from Kisvarsanyi, 1980).
GEOLOGIC SETTING
Middle Proterozoic rocks, 1,4801,450 Ma (Bickford,
1988), of the St. Francois granite-rhyolite terrane in south
east Missouri host 32 known magnetite and hematite
deposits and prospects that together form the Missouri
iron metallogenic province (Kisvarsanyi and Proctor,
1967). The St. Francois terrane consists predominantly of
rhyolite, which was intruded by coeval subvolcanic granite
massifs (fig. 1). Later, granite porphyry and an iron-rich
trachyte suite were emplaced as ring-dike intrusions
around the margins of caldera complexes. Bodies of two
mica granite, which are enriched in fluorine, uranium, and
thorium in addition to large-ion-lithophile elements
(LILE), such as Sn, W, Nb, Y, Be, Li, Rb, and Ba, are
inferred to form resurgent plutons within the central part
of caldera complexes (Kisvarsanyi, 1980, 1981; Kisvarsa
nyi and Kisvarsanyi, 1989a).
The Missouri iron deposits belong to the family of
Olympic-Dam- and Kiruna-type Fe, Cu, REE, and Au
deposits (Oreskes and Einaudi, 1990; Reeve and others,
1990). The Missouri ores are predominantly magnetite and
(or) hematite, with variable amounts of pyrite, chalcopy
rite, and bornite. They are enriched in the LILE, and some
are enriched in phosphorous. As is typical of this family
of deposits, they are distinct from deposits associated with
layered mafic intrusions in both mode of occurrence and
by the fact that they have low titanium contents.
The Missouri iron deposits predominantly occur as
intrusive or replacement bodies within volcanic rocks of
the St. Francois terrane. Late-stage aplite or granitic
dikes cut some of the deposits, which indicates that the
mineralization was synchronous with igneous activity in
the terrane (Day and others, 1989; Kisvarsanyi and Kis
varsanyi, 1989b).
THE PEA RIDGE DEPOSIT
The host rocks in the Pea Ridge deposit consist of
crystal-rich, subalkaline, rhyolitic ash-flow tuffs that
strike N. 80 W. and dip about 75 NE. Both the deposit
and host rocks are interpreted to form a large roof pen
dant on top of a circular, 20-km-wide, central granite
pluton (Nuelle and others, 1992) (fig. 1). This pluton was
interpreted as the core of a buried caldera complex (Kis
varsanyi, 1981; Nuelle and others, 1991) on the basis of
regional compilations of aeromagnetic and gravity data.
The margin of the caldera complex is delineated by an
iron-rich trachyte ring intrusion (fig. 1). The Pea Ridge
ore system may have evolved as immiscible fluids from
the source or a derivative magma of the iron-rich trachyte
suite (Kisvarsanyi and Kisvarsanyi, 1981; Day and
others, 1991).
The tabular Pea Ridge iron orebody discordantly cuts
its host rhyolitic rocks (fig. 2). The orebody strikes N.
55-60 E. and dips 75 SE. to nearly vertical (Husman,
1989). Replacement and open-space fill textures character
ize the deposit (Nuelle, Seeger, and others, 1989), and
cross-cutting, replacement, and alteration relations define
four mappable zones (Nuelle and others, 1992). They are,
from earliest to latest: (1) the amphibole-quartz zone; (2)
the magnetite ore zone, with massive magnetite and mag
netite-cemented breccia ore; (3) the hematite zone; and (4)
the silicified zone (fig. 2). Quartz veins, mafic dikes, and
aplite dikes cut the four main zones. REE-bearing breccia
pipes cut rocks of the footwall and the eastern margin of
the deposit (Marikos, Nuelle, and Seeger, 1989; Nuelle
and others, 1992).
 PEA RIDGE IRON-RARE-EARTH-ELEMENT DEPOSIT, SOUTHEAST MISSOURI 207

MINE
LEVEL
(feet)

1375

1475

1675

1825

1975

2125

2275
237O
2475

REE-rich breccia Magnetite zone

Silicified zone Amphibole-quartz-magnetite zone
Hematite zone Rhyolite

Figure 2. Isometric view of the Pea Ridge deposit. View looking northwest.

AMPHIBOLE-QUARTZ ZONE
The amphibole-quartz zone occurs on both the hang
ing wall and footwall of the deposit. This zone thins from
about 40 m wide on the eastern edge of the hanging wall
westward into host rock with fractures that are filled with
amphibole and have silicified walls (Emery, 1968). The
footwall amphibole-quartz zone is less brecciated and con
tains fewer quartz pods than that of the hanging wall.
Contacts between the amphibole-quartz zone and host rhy
olite wallrock are generally gradational; some wallrock
exhibits only incipient replacement by actinolite (Nuelle
and others, 1992).
The amphibole-quartz zone consists of massive,
coarse-grained actinolite with interstitial quartz, apatite,
magnetite, pyrite, chalcopyrite, and calcite (fig. 3). Actin
olite forms crystals as much as 5 cm long with seams and
disseminations of magnetite and sulfides. Fluorine content
in actinolite ranges from about 0.20 to 0.90 weight per
cent, whereas chlorine concentration is less than 0.05
weight percent. Quartz is present as both interstitial
grains and as massive pods 1 to 50 cm in diameter.
The amphibole-quartz zone represents a metasomatic
alteration front that preceded emplacement of the magne
tite orebody and formed a skarn in the host rhyolitic rocks.
Relict textures of rhyolite in the amphibole-quartz zone of
the hanging wall and incipient replacement of rhyolite by
actinolite confirm that rhyolite is the protolith of the
amphibole-quartz zone (Nuelle, Marikos, and others, 1989;
Nuelle, Seeger, and others, 1989; Nuelle and others,
1992).
208 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Figure 3. Representative sample of the amphibole-quartz zone.
Actinolite forms crystals as much as 5 cm long. Calcite fills inter
stices. Magnetite is disseminated in actinolite.
MAGNETITE ZONE
The magnetite ore zone consists of massive magnetite
and, on its outer margins, completely assimilated rhyolite
and magnetite-cemented breccias, known locally as
pseudobreccia and heterolithic breccia (Nuelle and others,
1992). Magnetite comprises about 60 to 90 modal percent
of the massive ore, with associated gangue minerals such
as apatite, quartz, phlogopite, chlorite after phlogopite,
pyrite, chalcopyrite, calcite, fluorite, barite, and potassium
feldspar in decreasing abundance accounting for about 12
to 35 modal percent. Martite varies between less than 1
and about 13 modal percent throughout the magnetite ore
zone. The texture of magnetite in the ore zone is generally
massive, coarse to fine grained, and granular, with subcon
choidal fracture. The gangue minerals form interstitial
grains, net-textured veinlets, and pods within the massive
magnetite.
Magnetite-cemented breccias contain variably altered
fragments of rhyolite in a matrix of magnetite + actinolite.
Alteration of rhyolite varies towards the orebody from rel
atively minor in pseudobreccia to intense in heterolithic
breccia. Locally, the contact between pseudobreccia and
rhyolite is sharp. Between heterolithic breccia or pseudo
breccia and massive magnetite is an ill-defined zone of rhy
olite and (or) amphibole-quartz-zone rock that is massively
replaced and assimilated by magnetite without any breccia
texture or with only faint ghosts of magnetite-replaced frag
ments. This zone also contains a higher proportion of
gangue minerals such as quartz and trace to minor amounts
of apatite, actinolite, chlorite, calcite, pyrite, and chalcopy
rite. These gangue minerals form seams, pods, and dissem
inations in magnetite. The pseudobreccia contains rhyolite
fragments that are neither rotated nor displaced, and the
breccia may be either matrix or clast supported. Heteroli
thic breccia is well developed along the hanging wall. It is
predominantly a matrix-supported breccia that contains
Figure 4. Photomicrograph of martite after magnetite in the mag
netite zone. Relict magnetite in cores of grains is replaced by hema
tite. Some hematite formed as overgrowths on magnetite and as
plates interstitial to magnetite. Field of view about 1.8 mm wide.
fragments of relatively unaltered rhyolite, chloritized rhyo
lite, and amphibole-quartz-zone rock in a magnetite matrix.
Fragments in pseudobreccia are commonly angular to sub
angular, whereas those in heterolithic breccia are subangu
lar to rounded due to resorption on the margins of
fragments.
Magnetite throughout the ore zone is compositionally
homogeneous and essentially pure Fe3O4. The titanium
content in magnetite from the ore zone, as analyzed on the
electron microprobe, ranges from 0.01 to about 0.02
weight percent TiO2, and the x'usp (apparent mole fraction
ulvospinel, Fe2TiO4 (Stormer, 1983)) for 20 grains is 3
0.001. Other oxides of elements such as nickel, magne
sium, chromium, manganese, silicon, aluminum, and zinc
are present in total amounts < 0.40 weight percent.
Hematite (martite) in the massive magnetite zone is
common as grains pseudomorphic after magnetite. The
amount of replacement is variable from minor alteration
on the rims of magnetite grains to complete replacement
(fig. 4). The hematite zone on the top and footwall of the
magnetite orebody (fig. 2) represents a zone of nearly
complete replacement of magnetite by martite. Martite in
the magnetite zone is virtually pure hematite, with x'ilm
(apparent mole fraction ilmenite, FeTiO3 (Stormer, 1983))
< 0.001 and the sum of oxides of the trace elements
< 0.25 weight percent.
Apatite and quartz are the most abundant gangue
minerals in the magnetite zone. Both are present in
amounts between about 1 and 10 modal percent. They
occur most commonly as interstitial grains to magnetite as
well as in pods and lining vugs. Semiquantitative analyses
on the scanning electron microscope (SEM) and the elec
tron microprobe indicate that apatite contains as much as 5
weight percent fluorine and less than 0.5 weight percent
 PEA RIDGE IRON-RARE-EARTH-ELEMENT DEPOSIT, SOUTHEAST MISSOURI
-
: ---
---
t
-
- (Mg+Fe) > 0.88), such as penninite or clinochlore, to iron
Figure 5. Photomicrograph of apatite with solid inclusions of
monazite. Equidimensional inclusions of monazite are randomly
oriented, whereas elongated crystals are crystallographically ori
ented. Field of view about 0.5 mm wide.
rare earth elements (REE). However, solid inclusions of
REE-bearing minerals such as monazite, xenotime, and
possibly britholite are abundant in apatite (Sidder, unpub.
data; Whitten and Cornell, 1991). In contrast, quartz is
distinctly free of solid inclusions. This is a key character
istic in distinguishing between quartz and apatite in pol
ished thick sections. The mineral inclusions in apatite vary
from equidimensional grains to elongated rods as much as
50 m in length. Although the former are oriented ran
domly, the long prismatic grains appear to be oriented
crystallographically in the apatite (fig. 5). Moreover, the
mineral inclusions in apatite appear to be zoned vertically
through the deposit, with abundant inclusions in apatite at
depth and inclusion-free apatite with associated, intersti
tial, discrete crystals of monazite at higher levels in the
mine. These features suggest that the REE-bearing miner
als included in apatite formed by exsolution. mli (1975)
and McKeown and Klemic (1956) described similar inclu
sions in apatite in the Gloserheia granite pegmatite, Nor
way, and in the magnetite deposits in the Mineville
district, New York, respectively. mli (1975) concluded
that the REE-bearing minerals in apatite formed by exso
lution, possibly accompanied by metasomatism.
Phlogopite or phlogopitic biotite is generally present
in amounts of 3 0.20 to 2.70 modal percent and rarely to
as much as 12 modal percent. Phlogopite forms grains that
are commonly intergrown with apatite interstitial to mag
netite and as thin veinlets in magnetite. The mica varies in
color from moderate reddish brown phlogopitic biotite to
pale reddish brown and colorless phlogopite. This color
variation appears to be related to the amount of TiO2.
Those grains that have recognizable color contain greater
than 1.50 weight percent TiO2, whereas those grains that
are faintly colored to colorless have less than 0.50 weight
209
percent TiO2. The ratio Mg/(Mg+Fe) (assuming total iron
aS Fe2+) also varies with color:darker grains have a ratio
of about 0.80, and nearly colorless grains have a ratio
greater than 0.80 and as much as 0.87. Fluorine abundance
in phlogopite and phlogopitic biotite varies between about
0.7 and 1.8 weight percent, whereas chlorine is < 0.2
weight percent.
Chlorite commonly replaces phlogopite. Its abun
dance varies from 0.50 to about 1 modal percent and
rarely to 5 modal percent. Chlorite is typically pleochroic
from light yellow to bluish or olive green, and it has low,
first-order gray to anomalous Prussian blue birefringence.
Its composition ranges from magnesian-rich chlorite (Mg/
magnesium-rich chlorite (Mg/(Mg+Fe) between 0.55 and
0.76), such as pycnochlorite or prochlorite (ripidolite).
This compositional variation is probably due to the com
positional differences in phlogopite and phlogopitic biotite
that it replaced.
Pyrite and chalcopyrite are the main sulfide minerals
in the magnetite zone. Pyrite is present in amounts that
vary from about 0.10 to as much as about 14 modal per
cent, whereas chalcopyrite varies from 0.10 to about
1.50 modal percent. Pyrite occurs as disseminations in
magnetite, as well as interstitial grains, seams, pods, and
stringer veinlets in the magnetite ore. Chalcopyrite is
sparsely disseminated in magnetite and is present as grains
associated with pyrite interstitial to magnetite and as blebs
in pyrite associated with pyrrhotite and possibly cubanite.
Some pyrrhotite is nickeliferous, with about 4.0 weight
percent nickel. Sulfide distribution is not zoned vertically
or horizontally within the orebody. The composition of
chalcopyrite approximates ideal stoichiometry without any
trace elements. Pyrite also is stoichiometric; however,
cobalt is a common trace metal in pyrite; and nickel,
arsenic, silver, copper, manganese, and zinc are rarely
present in amounts less than about 0.70 weight percent
total. The content of cobalt in pyrite in the magnetite zone
is generally zoned from as much as 5.00 weight percent in
the core of grains to about 0.50 weight percent on the rim.
This is in contrast to later pyrite that is associated with
blades of hematite in the magnetite and hematite zones
and contains < 0.50 weight percent cobalt.
Melonite (NiTe2), coloradoite (HgTe), and cassiterite
are present in trace amounts in the magnetite zone. They
form inclusions in pyrite and chalcopyrite (fig. 6). Melo
nite forms euhedral grains as much as 85 pum in diameter
and is disseminated in pyrite and associated with chal
copyrite that is interstitial to magnetite and apatite. A pea
nut-shaped grain of coloradoite about 18 pum long has
been recognized in cassiterite within pyrite (fig. 6E).
Other telluride minerals have not been distinguished in the
magnetite ore. Altaite (PbTe) is the predominant telluride
mineral in the REE-bearing breccia pipes. One sample of
magnetite ore with both melonite and coloradoite is the
210 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

only one that contains detectable amounts of gold (400

ppb).
Calcite, fluorite, barite, calcium sulfate, and potas
sium feldspar are rare (< 1 modal percent) gangue miner
als in the magnetite zone. Calcite, fluorite, and barite
occur mostly in pods and vugs associated with apatite and
quartz within massive magnetite. Electron microprobe
and SEM analyses have detected calcium sulfate (gypsum
or anhydrite) interstitial to apatite and chalcopyrite in
magnetite ore (figs. 6A, C).

HEMATITE ZONE

The hematite zone irregularly mantles the margins of

the magnetite orebody (fig. 2). The hematite zone sepa
rates the silicified zone from the magnetite orebody along -----
-
-
-

the footwall and host rhyolite from the magnetite zone

along the eastern edge of the deposit. Contacts are gener
-
-

J
ally gradational, with irregularly distributed patches of C- - r. -

magnetite in the hematite zone and hematite in the magne

tite zone. Hematite is also localized along faults and frac
tures in the magnetite zone (Emery, 1968; Nuelle and
others, 1992).
Hematite varies in abundance throughout the deposit
from 3 1.0 modal percent in the magnetite zone to about
80 modal percent in the hematite zone. In general, hema
tite decreases in abundance with depth. Most hematite is
an alteration product of magnetite and, where massive
replacement is present, forms fine-grained massive mar
tite. However, some hematite in the magnetite and hema
tite zones is platy or bladed (fig. 7), and some hematite
forms overgrowths on magnetite (fig. 4) as well as occur
ring in veins and pods in the magnetite zone. Some of the
martite and hematite may have formed at high temperature
in response to an increased fugacity of oxygen in the min
eralizing fluids. As noted previously, martite after magne
tite is relatively pure Fe2O3. Platy hematite is slightly
more variable in composition, with x'ilm varying between
< 0.001 and 0.031 and containing as much as 1.58 weight
percent TiO2 and a total of 0.20 weight percent zinc, sili
con, aluminum, manganese, and chromium oxides.
Quartz, apatite, pyrite, and calcium sulfate are the
most common gangue minerals associated with hematite.
Pyrite interstitial to and included within plates of hematite
(fig. 7) is notably low in cobalt (< 0.60 weight percent) in
contrast to early pyrite in the magnetite zone. Calcium
sulfate (gypsum or anhydrite) is typically included within
the plates of hematite (fig. 7).
SILICIFIED ZONE
Massively silicified rhyolite with extensive potassium
feldspar alteration is developed in part of the footwall.
This silicified zone is present between the hematite zone
:
Figure 6A. Enhanced SEM image of telluride-bearing magnetite
ore. Mt, magnetite; py, pyrite; cp, chalcopyrite; ap, apatite; anhy,
calcium sulfate; cas, cassiterite; mel, melonite.
and host rock on the north side of the deposit and between
magnetite ore and host rock on the west (Husman, 1989;
Seeger and others, 1989). Massive white to light-gray
quartz that variably replaces rhyolite and fills open spaces
characterizes this zone (fig. 8). Potassium feldspar flood
ing, or potassic alteration, has converted the gray to red
dish-brown host volcanic rocks to moderate reddish
orange rock (Nuelle and others, 1992). Locally within the
silicified zone, quartz and orthoclase form pegmatitic pods
and veins. Sericite, muscovite, biotite, fluorite, barite,
tourmaline, epidote, calcite, pyrite, chalcopyrite, hematite,
rutile, apatite, monazite, zircon, calcium sulfate, and topaz
are accessory minerals in the silicified zone (Husman,
1989; Nuelle and others, 1992). Sericite, fluorite, and tour
maline typically coat fractures, joints, and fragment sur
faces in the silicified zone (Seeger and others, 1989;
Nuelle and others, 1992), and calcite is common in vein
lets, whereas the other accessory minerals occur as dis
seminations in quartz and orthoclase. Quartz and (or)
orthoclase veins that cut other zones of the orebody were
in part contemporaneous with massive silicification. These
veins also contain minor amounts of barite, calcite, fluo
rite, chalcopyrite, monazite, xenotime, and rutile (fig. 8)
(Nuelle, Seeger, and others, 1989). A uranium-lead date of
1.46 Ga on xenotime from a quartz vein (W.R. Van
Schmus, University of Kansas, written commun., 1988)
confirms the age of the Pea Ridge deposit to be contempo
raneous with regional igneous activity.
 PEA RIDGE IRON-RARE-EARTH-ELEMENT DEPOSIT, SOUTHEAST MISSOURI 211

|
t:
299.52 MICRONS H
Figure 6B. Element maps of sample in 6A. Elements are, left to right, Sn, Fe, Hg, and Te on top, and S, Ni, Cu, and P on bottom.

- -
---

1UPM 2CKU CU J.
(

Figure 6C. SEM photomicrograph of calcium sulfate shown

in 6A.

Figure 6D. Expanded view of area in 6A that contains cassiterite.

Mt, magnetite; py, pyrite; cp, chalcopyrite; ap, apatite; cas, cassiter
ite; mel, melonite; col, coloradoite.
212 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
S
77.87 MICRONS H-4
Figure 6B. Element maps of sample in 6D. Elements are, left to right, Sn, Fe, Hg, and Te on top; S, Ni, Cu, and Mg on bottom.
Figure 7. Photomicrograph of martite and hematite with intersti
tial pyrite and quartz in the hematite zone. Magnetite has been com
pletely replaced by martite. Hematite plates enclose pyrite and
quartz. Field of view about 4.8 mm wide.
Mapping of underground workings indicates that
silicification postdated the emplacement of magnetite ore
(Nuelle and others, 1992). For example, underground
workings show silicified magnetite ore and pseudobreccia
where both the rhyolite fragments and magnetite cement
are silicified (fig. 8). However, in contrast, some drill core
displays pseudobreccia with fragments of silicified rhyo
lite and quartz in unsilicified magnetite cement, which
suggests that some silicification occurred before emplace
ment of the magnetite ore. Alternatively, magnetite
emplacement and silicification may have overlapped in
space and time. The diverse assemblage of minerals in the
silicified zone and its ambiguous paragenetic relation with
the magnetite zone suggest that the conditions of deposi
tion, such as temperature, pH, and the fugacities of oxy
gen, sulfur, and carbon dioxide, changed through time.
Work is in progress to document more closely mineral
paragenesis in the silicified zone and to distinguish
between relict textures and minerals in altered and
replaced rhyolite versus secondary minerals introduced
with quartz and potassium feldspar.
RARE-EARTH-ELEMENT-BEARING
BRECCLA PIPES
At least four REE-bearing breccia pipes cut the iron
ore and altered rhyolite wallrocks in the Pea Ridge deposit
at or near the contacts between the different zones of the
deposit. For example, one pipe is along the contact
 PEA RIDGE IRON-RARE-EARTH-ELEMENT DEPOSIT, SOUTHEAST MISSOURI
Figure 8. Photographs, taken underground, of the silicified zone.
Top, Zone of massive silicification. Bottom, Silicified pseudo
breccia.
between the silicified and magnetite zones, and another is
between the silicified and hematite zones. Locally, apo
physes of breccia pipes protrude into adjacent rocks
(Nuelle and others, 1992). Three pipes are relatively soft
and friable, whereas one is hard and indurated.
The temporal relation between the silicified zone and
the REE-bearing breccia pipes is apparently ambiguous.
Fragments of silicified wallrock and massive quartz from
the silicified zone in the friable, soft, REE-bearing breccia
pipes clearly indicate that silicification preceded emplace
ment of the breccia pipes, whereas the hard pipe is appar
ently silicified (Marikos, Seeger, and Nuelle, 1989).
However, the hardness of the latter is probably related to a
greater abundance of primary quartz in the pipe and is not
an overprint due to secondary quartz introduced into the
breccia pipe after its emplacement. The matrix of the hard
breccia pipe is not silicified.
213
The forms of the hard and soft breccia pipes are simi
lar. The pipes dip > 60 and are elongate to ovoid in plan
view (Nuelle and others, 1992). They range in length from
several meters to as much as 60 m and in width from 3 1
to about 15 m. Their maximum vertical extent is not
known, but at least one pipe is exposed over a distance of
about 120 m.
Fragments of rhyolite, magnetite- and hematite-rich
ore, and silicified-zone rocks in the breccia pipes range in
diameter from 3 1 mm to several meters and document the
late-stage emplacement of the pipes. Some minerals such
as sanidine, quartz, and barite form megacrysts or possibly
phenocrysts (Nuelle and others, 1991). These breccias are
both matrix and clast supported (Nuelle and others, 1992).
The matrix of the breccias consists predominantly of rock
flour, potassium feldspar (predominantly orthoclase), chlo
rite, barite, monazite, xenotime, quartz, apatite, allanite,
and calcite (Marikos, Seeger, and Nuelle, 1989). Other
accessory minerals present in trace quantities include other
REE-bearing minerals, altaite, chalcopyrite, pyrite, hema
tite, galena, thorite, and cassiterite.
Monazite is the predominant REE-bearing mineral in
the breccia pipes. Lesser amounts of xenotime and uncom
mon bastnaesite ((Ce,La)CO3(FOH)), britholite ((Ca,Ce);
(SiO4,PO4)3(OH,F)), allanite, and possibly tengerite
(Cay3(CO3)4(OH)33H2O) and synchisite ((Y,Ce)Ca
(CO3)F2) have also been observed (Husman, 1989;
Whitten and Cornell, 1991; Sidder, unpub. data, 1991).
These minerals generally occur as small crystals that are
< 4.0 mm in diameter in the matrix, as fill in cracks in
barite and sanidine megacrysts that are > 2 cm long, and
as disseminations in rock fragments. Allanite forms euhe
dral tablets in acicular groups within the matrix of the
hard breccia pipe. Total REE-oxide (REO) concentrations
from grab samples range from about 5 to 38 weight per
cent, with an average of 20 weight percent. Grades from
bulk samples range from about 7 to 25 weight percent
REO and average about 12 weight percent (Nuelle and
others, 1992). Ore reserves of about 600,000 tons at 12
weight percent REO have been estimated for the hard
breccia pipe at Pea Ridge (Whitten and Yancey, 1990).
Thorium and uranium are present in minor quantities
in the breccia pipes. The thorium concentration in grab
samples averages about 3,320 ppm, and the uranium con
tent averages about 190 ppm (Nuelle and others, 1992).
Thorite has been identified as grains disseminated in the
REE-bearing breccia pipes associated with altaite (PbTe),
monazite, and other REE-bearing minerals (Husman,
1989; Sidder unpub. data, 1991). Radioactivity imparted
by thorium and uranium serves as an exploration guide for
the breccia pipes.
Precious metals are distributed erratically in the brec
cia pipes. Gold and silver are present in trace amounts,
with 58 grab samples of breccia pipe material averaging
about 0.9 ppm gold and 1.2 ppm silver (L.J. Tucker, Pea
214 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Figure 9. Photomicrographs of fluid inclusions in quartz (field of
view about 0.5 mm wide). A, Three-phase fluid inclusions in quartz
in the rare-earth-element-bearing breccia pipes. B, Two-phase fluid
inclusions in the rare-earth-element-bearing breccia pipes. Note the
different proportion of vapor and liquid in the inclusions.
Ridge Iron Ore Co., oral commun., 1989). The hard brec
cia pipe is apparently enriched in precious metals relative
to the soft pipes (Husman, 1989). Precious-metal-bearing
minerals tentatively identified include hessite (Ag3Te),
sylvanite (AuAg ea), and electrum (Husman, 1989).
Altaite is the most common telluride mineral in the
breccia pipes.
FLUID-INCLUSION STUDIES
Quartz in all zones of the Pea Ridge deposit is the
only fluid-inclusion-bearing mineral identified thus far.
Fluorite and barite do not contain fluid inclusions, and
apatite contains only mineral inclusions, as mentioned pre
viously. Primary fluid inclusions in quartz in the amphi
bole-quartz, magnetite, hematite, and silicified zones
consist of three or more phases, with liquid, vapor, halite,
and one or more other daughter minerals (fig. 9A). Those
on the same plane within a given sample have similar pro
portions of the three phases. The inclusions range in size
from less than 5 pum to as much as 16 pum. Fluid inclu
FLUID INCLUSION HOMOGENIZATION
TEMPERATURES AT THE PEA RIDGE MINE
Silicified Zone = m n ml
Hematite Zone
f Magnetite Zone
Amphibole-Quartz Zone
m, m-m
|--|--|--|--|--|
300 350 400 450 500 550
Temperature (C)
Figure 10. Histogram of homogenization temperatures of three
phase (L + V + H) fluid inclusions in the Pea Ridge deposit. All
inclusions are in quartz and homogenize by halite disappearance.
sions in the REE-bearing breccia pipes contain only liquid
and vapor (fig. 9B). However, rock and mineral fragments
from other zones in the breccia pipes make study of the
fluid inclusions in the breccia pipes more difficult. Sec
ondary fluid inclusions in all zones are recognized readily
by their alignment along fractures. They are generally
smaller (< 7 pum in diameter) than the primary inclusions,
consist predominantly of liquid and vapor only, and
homogenize at temperatures less than 250C.
All three-phase inclusions homogenize by halite dis
solution after vapor disappearance. Homogenization tem
peratures of three-phase liquid + vapor + halite fluid
inclusions in quartz within the amphibole-quartz zone
indicate that these inclusions were trapped at temperatures
as great as or greater than 530C (fig. 10). Halite in some
inclusions had shrunk in size and become rounded during
heating, but had not dissolved by 530C, at which point
the heating run was terminated. Salinities calculated using
the method of Potter and others (1977) for the inclusions
in the amphibole-quartz zone range from about 52 to > 56
equivalent weight percent NaCl. Preliminary oxygen iso
tope data for quartz-magnetite pairs suggest that this zone
may have formed at temperatures as high as 680C (Sid
der and others, 1991).
Gangue quartz interstitial to magnetite in the ore zone
contains characteristic, but relatively rare, three-phase
liquid + vapor + halite fluid inclusions. These inclusions
generally homogenize at temperatures between about
480C and 530C; however, halite in some inclusions did
not disappear by 530C (fig. 10). Salinities in the magne
tite zone range from about 54 to > 60 equivalent weight
percent NaCl. A quartz-magnetite pair yields a calculated
temperature from oxygen-isotope data of 480C (Sidder
and others, 1991).
Three-phase liquid + vapor + halite fluid inclusions
are rare in quartz interstitial to hematite plates in the
hematite zone (fig. 7). Preliminary heating studies indicate
that these inclusions homogenize by halite dissolution at
 PEA RIDGE IRON-RARE-EARTH-ELEMENT DEPOSIT, SOUTHEAST MISSOURI
about 410C (fig. 10), and a quartz-hematite pair yields a
calculated temperature of 395C (Sidder and others, 1991).
The calculated salinity in the hematite zone is about 47
equivalent weight percent NaCl.
Multi-phase fluid inclusions are abundant in the silic
ified zone. Liquid + vapor + halit inclusions are most
common, but some inclusions contain additional daughter
minerals. Some of these daughter minerals include hema
tite, recognized by its orangish color; an opaque mineral;
and tabular or rounded translucent crystals, some of which
are birefringent. These inclusions all homogenize by halite
dissolution, although some of the other daughter minerals
do not disappear before halite. The temperature of homog
enization in the silicified zone ranges from about 320C to
430C, and the salinity ranges from about 39 to 49 equiva
lent weight percent NaCl.
Two-phase liquid-and-vapor fluid inclusions with
variable liquid-to-vapor proportions coexist in quartz
within the matrix of the breccia pipes. Such coexisting
populations of vapor- and liquid-rich fluid inclusions are
commonly indicative of entrapment of fluids that under
went boiling (Cunningham, 1978). Because of the variable
proportion of liquid and vapor in these inclusions, temper
atures of disappearance of the vapor or liquid are not nec
essarily representative of the trapping temperature. The
temperature calculated from the sulfur isotope composition
of barite and pyrite in the breccia pipes yielded 305C
(Sidder and others, 1991).
CONCLUSIONS
The Pea Ridge iron and rare-earth-element orebody is
a high-temperature, magmatic-hydrothermal deposit. The
ore was emplaced within what is thought to be an eroded
caldera complex at a relatively shallow level (< 2 km) of
the crust. The deposit consists of four mappable zones, the
amphibole-quartz, magnetite, hematite, and silicified
zones, which formed paragenetically in succession as tem
perature decreased from greater than 530C to about
300C. Ambiguous relations between the silicified and
magnetite zones indicate that their period(s) of deposition
may have overlapped in part of the deposit. Fluids intersti
tial to magnetite in the iron orebody may have escaped,
boiled, and caused brecciation during rapid and explosive
emplacement of the REE-bearing breccia pipes. The mag
netite deposit (and the REE-bearing breccia pipes by
direct association and geologic relations such as the pres
ence of sanidine as a phenocryst phase) formed from
highly saline, high-temperature, magmatically derived
hydrothermal fluids. It is possible that the mineralizing
fluids are genetically associated with rocks of the iron-rich
trachyte suite and that they formed as immiscible fluids.
215
REFERENCES CITED
Amli, Reidar, 1975, Mineralogy and rare earth geochemistry of
apatite and xenotime from the Gloserheia granite pegma
tite, Froland, southern Norway: American Mineralogist, v.
60, p. 607620.
Bickford, M.E., 1988, The formation of continental crust: Part 1.
A review of some principles; Part 2. An application to the Pro
terozoic evolution of southern North America: Geological So
ciety of America Bulletin, v. 100, p. 13751391.
Cunningham, C.G., 1978, Pressure gradients and boiling as mech
anisms for localizing ore in porphyry systems: U.S. Geologi
cal Survey Journal of Research, v. 6, no. 6, p. 745754.
Day, W.C., Kisvarsanyi, E.B., Nuelle, L.M., Marikos, M.A., and
Seeger, C.M., 1989, New data on the origin of the Pea Ridge
iron-apatite deposit, southeast MissouriImplications for
Olympic Dam-type deposits [abs.]: Geological Society of
America, Abstracts with Programs, v. 21, no. 6, p. A 132.
Day, W.C., Sidder, G.B., Rye, R.O., Nuelle, L.M., and Kisvarsanyi,
E.B., 1991, The Middle Proterozoic rhyolite-hosted Pea Ridge
iron and rare-earth element depositA magmatic source for
Olympic Dam-type deposits in the Midcontinent region of the
United States [abs.], in Chappell, B.W., ed., Second Hutton
Symposium on Granites and Related Rocks: Canberra, Austra
lia, Sept. 1991, Abstracts with Program, p. 30.
Emery, J.A., 1968, Geology of the Pea Ridge iron orebody, in
Ridge, J.D., ed., Ore Deposits of the United States,
19331967: New York, American Institute of Mining, Met
allurgical, and Petroleum Engineers, Inc., The Graton-Sales
Volume, v. 1, p. 359369.
Husman, J.R., 1989, Gold, rare earth element, and other potential
by-products of the Pea Ridge iron ore mine: Missouri Depart
ment of Natural Resources, Division of Geology and Land Sur
vey, Contributions to Precambrian Geology No. 21, Open File
Report OFR-89-78-MR, 18 p.
Kisvarsanyi, E.B., 1980, Granitic ring complexes and hotspot activ
ity in the St. Francois terrane, Midcontinent region, United
States: Geology, v. 8, p. 4347.
Kisvarsanyi, E.B., 1981, Geology of the Precambrian St. Francois
terrane, southeastern Missouri: Missouri Department of Natu
ral Resources, Division of Geology and Land Survey, Report
of Investigations No. 64, 58 p.
Kisvarsanyi, E.B., and Kisvarsanyi, Geza, 1989a, Alkaline granite
ring complexes and metallogeny in the Middle Proterozoic St.
Francois terrane, southeastern Missouri, U.S.A.: Geological
Association of Canada Special Paper 38, 14 p.
Kisvarsanyi, Geza, and Kisvarsanyi, E.B., 1981, Genetic relation
ship of Kiruna-type apatitic iron ores to magnetite trachyte and
syenite in the St. Francois terrane, Missouri [abs.]: Geological
Society of America, Abstracts with Programs, v. 13, p. 488.
1989b, Precambrian geology and ore deposits of the south
east Missouri iron metallogenic province, in Brown, V.M.,
Kisvarsanyi, E.B., and Hagni, R.D., eds., Olympic Dam
Type Deposits and Geology of Middle Proterozoic Rocks in
the St. Francois Mountains Terrane, Missouri: Society of Eco
nomic Geologists Guidebook Series, v.4, p. 140.
Kisvarsanyi, Geza, and Proctor, P.D., 1967, Trace element content
of magnetites and hematites, southeast Missouri iron metallo
genic province, U.S.A.: Economic Geology, v. 62, p. 449471.
216 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Marikos, M.A., Nuelle, L.M., and Seeger, C.M., 1989, Geology of
the Pea Ridge Mine, in Brown, V.M., Kisvarsanyi, E.B., and
Hagni, R.D., eds., Olympic Dam-Type Deposits and Geolo
gy of Middle Proterozoic Rocks in the St. Francois Mountains
Terrane, Missouri: Society of Economic Geologists Guide
book Series, v.4, p. 4154.
Marikos, M.A., Seeger, C.M., and Nuelle, L.M., 1989, REE-miner
alized breccia pipes in the Pea Ridge deposit of southeast Mis
souri [abs.]: Geological Society of America, Abstracts with
Programs, v. 21, no. 6, p. A34.
McKeown, F.A., and Klemic, Harry, 1956, Rare-earth-bearing ap
atite at Mineville, Essex County, New York: U.S. Geological
Survey Bulletin 1046B, p. 923.
Nuelle, L.M., Day, W.C., Sidder, G.B., and Seeger, C.M., 1992,
Geology and mineral paragenesis of the Pea Ridge iron ore
mine, Washington County, MissouriOrigin of the rare-earth
element- and gold-bearing breccia pipes, in Day, W.C., and
Lane, D.E., Strategic and Critical Minerals in the Midcontinent
Region, United States: U.S. Geological Survey Bulletin 1989,
Chapter A, p. A1-A11.
Nuelle, L.M., Kisvarsanyi, E.B., Seeger, C.M., Day, W.C., and Sid
der, G.B., 1991, Structural setting and control of the Pea Ridge
magnetite deposit, Middle Proterozoic St. Francois terrane,
Missouri [abs.]: Geological Society of America, Abstracts with
Programs, v. 23, no. 5, p. A292.
Nuelle, L.M., Marikos, M.A., Seeger, C.M., and Day, W.C., 1989,
Hydrothermal aspects of amphibole and magnetite emplace
ment in the Pea Ridge Fe-REE deposit, Washington County,
Missouri [abs.]: Geological Society of America, Abstracts with
Programs, v. 21, no. 6, p. A248.
Nuelle, L.M., Seeger, C.M., Marikos, M.A., and Day, W.C., 1989,
Mineral assemblages of the Pea Ridge Fe-REE deposit: Impli
cations of an Olympic Dam-type variant in southeast Missouri
[abs.]: Geological Society of America, Abstracts with Pro
grams, v. 21, no. 6, p. A34.
Oreskes, Naomi, and Einaudi, M.T., 1990, Origin of rare earth ele
ment-enriched hematite breccias at the Olympic Dam Cu-U
Au-Agdeposit, Roxby Downs, South Australia: Economic Ge
ology, v. 85, p. 128.
Potter, R.W., II, Babcock, R.S., and Brown, D.L., 1977, A new
method for determining the solubility of salts in aqueous solu
tions at elevated temperatures: U.S. Geological Survey Journal
of Research, v. 5, no. 3, p. 389395.
Reeve, J.S., Cross, K.C., Smith, R.N., and Oreskes, Naomi, 1990,
Olympic Dam copper-uranium-gold-silver deposit, in Hughes,
F.E., ed., Geology of the Mineral Deposits of Australia and
Papua New Guinea: Australasian Institute of Mining and Met
allurgy, Monograph no. 14, p. 10091035.
Seeger, C.M., Nuelle, L.M., and Marikos, M.A., 1989, Massive
silicification and late stage quartz veining in the Pea Ridge Fe
REE deposit, southeast Missouri [abs.]: Geological Society of
America, Abstracts with Programs, v. 21, no. 6, p. A34.
Sidder, G.B., Nuelle, L.M., Day, W.C., Rye, R.O., Seeger, C.M.,
and Kisvarsanyi, E.B., 1991, Paragenesis and conditions of
formation of the Pea Ridge iron and rare-earth element deposit,
Missouri [abs.]: Geological Society of America, Abstracts with
Programs, v. 23, no. 5, p. A292.
Sims, P.K., Kisvarsanyi, E.B., and Morey, G.B., 1987, Geology and
metallogeny of Archean and Proterozoic basement terranes in
the northern midcontinent, U.S.A.An overview: U.S. Geo
logical Survey Bulletin 1815, 51 p.
Stormer, J.C., Jr., 1983, The effects of recalculation on esti
mates of temperature and oxygen fugacity from analyses of
multicomponent iron-titanium oxides: American Mineralo
gist, v. 68, p. 586594.
Whitten, C.W., and Cornell, W.L., 1991, Rare earth occurrences
in the Pea Ridge tailings: American Institute of Mining En
gineers Preprint, 20 p.
Whitten, C.W., and Yancey, R.J., 1990, Characterization of the rare
earth mineralogy at the Pea Ridge deposit: U.S. Bureau of
Mines, Report of Investigations 9331, 15 p.
 CHAPTER V
MINERAL RESOURCE POTENTIAL OF THE
WHITE RIVER NATIONAL FOREST AND THE
DILLON RANGER DISTRICT OF THE
ARAPAHO NATIONAL FOREST, COLORADO
By MARGO I. TOTH, ANNA B. WILSON. THERESA M. COOKRO.
VIKI BANKEY." GREG K. LEE, and JOHN S. DERSCH%
ABSTRACT
The mineral resource potential of the White River
National Forest and Dillon Ranger District of the Arapaho
National Forest (referred to as the Forest in this report)
was evaluated by the U.S. Geological Survey (USGS).
Geological, geochemical, and geophysical data were com
piled and compared to characteristics of known and
inferred deposits within or close to the Forest. Models for
14 locatable commodities and five leasable commodities
were identified and evaluated. For locatable commodities,
the Forest contains 751 mi2 of high resource potential, 535
mi2 of moderate resource potential, 83 mi2 of low
resource potential, and 6 mi2 of unknown resource poten
tial. For leasable commodities, the Forest contains 433
mi2 of high resource potential and 85 mi2 of low resource
potential.
"U.S. Geological Survey, Mail Stop 922, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 937, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 973, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
U.S. Forest Service, U.S. Department of Agriculture, 11177 W. 8th
Avenue, P.O. Box 25127, Lakewood, CO 80225.
INTRODUCTION
The assessment of the mineral resource potential of
the White River National Forest and Dillon Ranger Dis
trict of the Arapaho National Forest, Colorado (fig. 1),
was made to assist the U.S. Forest Service in fulfilling the
requirements of Title 36, Chapter 2, part 219.22, Code of
Federal Regulations and to supply resource information
and interpretations so that the mineral resources of the
Forest could be considered, along with other resources, in
land-use planning. The Dillon Ranger District of the Arap
aho National Forest was included in the report on the
White River National Forest because the two areas are
administered by the staff of the White River Forest and
are included in a single planning document. For simplic
ity, only figure 1 distinguishes or outlines the Dillon
Ranger District; in all other figures the district is included
within the Forest boundary. The identified, or known,
resources of the Forest were studied by the U.S. Bureau of
Mines (Brown, 1990); the potential for undiscovered min
eral and energy resources was studied by the U.S. Geolog
ical Survey (Toth and others, in press). This report is an
abbreviated summary of Toth and others (in press).
The Forest encompasses about 2.4 million acres in
central and northwestern Colorado in two separate parcels:
a southern parcel that is elongate in an east-west direction
and a northern, square-shaped parcel. Eight distinct moun
tain ranges and parts or all of ten counties are within the
Forest. The entire Forest is on the western side of the
Continental Divide, which forms the eastern boundary of
parts of the Forest in many places. Except for some water
diversion projects, the Forest watershed drains into the
 108 1079 1069
:
|

APPROXIMATE
OF
BOUNDARY
THE
WHITE
FOREST
NATIONAL
RIVER

4.
%.
iver Ri
3. Colorado
Granby
D
-
Meeker
Kremmling
C
c
f APPROXIMATE
BOUNDARY
THE
OF
[7
40
H.
Buford
JP DISTRICT
RANGER
D|LLON
THE
OF
<& NATIONAL
ARAPAHO
FOREST
&
&
Red
G)
River A.
Mountain
-
-
White
l Plateau
2
or

Ay
99
6
Castle
New
C
[]
Rifl ic
li
Glenwood
Springs
Battlement
st'
**
&
&

Elk
Mountains

Marble
D
Castle
Peak
c 14,265

39H.

pCrested
Butle

O
10
20 3.MILES
lI |

National
Colorado.
Forest,
Index
1.
Figure
location
the
showing
map
National
River
White
of
Ranger
Dillon
and
Forest
Arapaho
District
 MINERAL RESOURCE POTENTIAL, WHITE RIVER N.F., COLORADO
Colorado River, which flows southwestward through cen
tral and western Colorado. Interstate Highway 70 traverses
east-west through the central part of the Forest, and
numerous other State and county highways extend through
or to the Forest. Improved and unimproved roads and jeep
trails provide access throughout the Forest. The major
communities within and near the Forest are Aspen, Breck
enridge, Dillon, Frisco, Vail, and Glenwood Springs.
GEOLOGIC HISTORY
Discussions of the geologic and tectonic history of
central and northwestern Colorado are summarized in
Tweto (1980), Wallace (1990), Mutschler and others
(1987), and Toth and others (in press). The geologic his
tory recorded in the Forest spans more than 1.8 billion
years. The major structural and rock-forming events are as
follows: Early Proterozoic accretion and metamorphism of
volcanic and sedimentary rocks, and intrusion of granitic
plutons; early to middle Paleozoic deposition of marine
and nonmarine sedimentary rocks during repeated epeiro
genic movements; late Paleozoic and early Mesozoic ero
sion and deposition related to the formation of two
highlands, separated by a narrow basin; Late Cretaceous
through early Eocene uplift during the Laramide orogeny,
along with volcanism and emplacement of granitic plu
tons; middle Tertiary tectonic quiescence, accompanied by
volcanism and emplacement of granitic plutons; late Ceno
zoic magmatism, block faulting in an extensional regime,
formation of a bimodal suite of basalt and rhyolite and
emplacement of related plutons; and Pleistocene to
Holocene glaciation and erosion. The simplified geology
and a few of the major structural features in the Forest are
shown on figure 2.
Within the Forest, Proterozoic crystalline rocks are
exposed in mountain ranges in the eastern part of the
Forest and in some of the deeper canyons in the north
western part of the Forest. Paleozoic sedimentary rocks
flank the Proterozoic rocks on the western, northern, and
eastern sides and are also exposed in the northwestern
part of the Forest. Mesozoic sedimentary rocks are
located principally in the northeastern part of the Forest
and along the western boundary of the Forest, including
exposures along the Grand Hogback monocline. To the
west of the monocline, Tertiary sedimentary rocks are
exposed in the Piceance Basin. Tertiary basalt flows are
mostly in the northwestern part of the Forest, where they
overlie Paleozoic sedimentary rocks. Laramide and
younger plutons are concentrated along a northeast-trend
ing zone through the southern and eastern parts of the
Forest. Mineral deposits are also concentrated along this
zone, known as the Colorado Mineral Belt (Tweto and
Sims, 1963) (fig. 3).
219
GEOCHEMISTRY
The geochemical contributions to the mineral
resource assessment of the White River National Forest
were derived from data that include analyses for 2,661
rock, 637 stream-sediment, 820 heavy-mineral-concen
trate, and 1,856 National Uranium Resource Evaluation
(NURE) stream-sediment samples from the U.S. Geologi
cal Survey Branch of Geochemistry databasesaltogether
this represents 5,513 sample locations. The NURE data
were combined from digital copies of analyses listed for
the Leadville (Planner and others, 1981), Grand Junction
(Langfeldt and others, 1981), Montrose (Broxton and oth
ers, 1979), Craig (Bolivar and others, 1979), and Denver
(Shettel and others, 1981) 1%x2 quadrangles. Included in
the geochemical information produced by the USGS were
published and unpublished data previously generated for
mineral resource appraisals of Wildernesses, Wilderness
Study Areas, Primitive Areas, and Roadless Areas (Wal
lace and others, 1989; Van Loenen and others, 1989; Free
man and others, 1985; Ludington and Ellis, 1981; Tweto
and others, 1970; Mallory and others, 1966; Soulliere and
others, 1985a, 1985b, 1986; Theobald and others, 1983).
Analyses for 39 elements were included in the data
and, of these, 15 were selected for primary use in the
geochemical contribution to the assessment of resource
potential in the Forest. These elements are silver, arsenic,
gold, cobalt, chromium, copper, molybdenum, nickel, lead,
antimony, tin, uranium, vanadium, tungsten, and zinc.
The distributions of boron, barium, bismuth, beryllium,
chromium, iron, manganese, and thorium were also exam
ined but were determined to be less significant than the 15
elements mentioned above for use in the evaluation.
To combine the data from different sample media and
analytical methods, a series of mathematical operations
was performed on the data sets. Standardized data sets
were created by performing various mathematical opera
tions, resulting in numeric values that represented the
number of standard deviations that the original value devi
ated from mean. Values between 2 and 3 in the trans
formed data were considered to represent slightly
anomalous concentrations; values of 3 to 5 were treated as
moderately anomalous; and values above 5 were consid
ered highly anomalous. Interpretations were derived from
graphical displays that could simultaneously include all
available data.
These data were spatially interpolated, and contour
plots were produced to provide areal representations of
anomalous concentrations of various elements in the study
area. Point plots of element suites were also generated to
give a detailed accounting of the sample-site locations for
the anomalous areas as well as a view of associations of
elements that are thought to be likely constituents of various
mineral-deposit types. In the Forest, samples containing
220 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

108 107 106

-I | |
APPROxIMATE BOUNDARY OF THE
WHITE RIVERNATIONAL FOREST AND
DILLON RANGER DISTRICT OF THE
ARAPAHO. NATIONAL FOREST

3.
3.

40- -

. . . p. N illiams Range
:r. 38 rust fault

&

o 10 20 30 MILES

- |

EXPLANATION

- ZZZZ Middle and Early Proterozoic

Tertiary basalt % crystalline rocks
III
, Tertiary sedimentary rocks # -

Syncline

Ki. Tertiary and Cretaceous intrusive rocks # Monocline

NNN
Mesozoic sedimentary rocks t Anticline

Pis . Paleozoic sedimentary rocks -AA-A- Thrust fault

Figure 2. Map showing simplified geology and some of the major structural elements within or close to the White River National Forest
and Dillon Ranger District of the Arapaho National Forest, Colorado.

anomalous concentrations of base and precious metals and GEOPHYSICS

related elements occur mostly in the proximity of known
mining districts and mineral-deposit occurrences (Toth and
others, in press). However, many areas of isolated anoma-
lies of base and (or) precious metals were not found to be
related to any known occurrences.
Three sets of geophysical data (gravity, aeromagnetic,
and radiometric) were compiled and interpreted for the
Forest (Toth and others, in press; Bankey, this volume).
Sources of data include studies of six Forest Service and
 MINERAL RESOURCE POTENTIAL, WHITE RIVER N.F., COLORADO
108 107 1069
i APPROxIMATE BOUNDARY OF THE
WHITE RIVER NATIONAL FOREST AND
DILLON RANGER DISTRICT OF THE
ARAPAHO NATIONAL FOREST
40
w!
Colorado Mineral Belt
0 10 20 30 MILES
Figure 3. Map of the Colorado Mineral Belt. Modified from Tweto
and Sims (1963).
BLM Wilderness Study Areas within the Forest (Camp
bell, 1981, 1985; Campbell and Wallace, 1986; Van
Loenen and others, 1989; Tweto and others, 1970; and
Moss and Abrams, 1985). Geophysical studies to support
mineral resource studies in the Colorado Mineral Belt
were made by Case and Tweto (Case, 1965, 1966, 1967;
Tweto and Case, 1972). Other related geophysical studies
include ones by Behrendt and Bajwa (1974) and Isaacson
and Smithson (1976).
A complete Bouguer gravity anomaly map was pro
duced using edited gravity data from 6,500 stations col
lected over the past several decades (Toth and others, in
press). A derivative gravity map was calculated from the
Bouguer gravity grid using the program FFTFIL (Hilden
brand, 1983) to remove or filter anomaly wavelengths
longer than 30 mi. This derivative map emphasizes anom
alies produced by shallow sources and suppresses the
longer wavelength anomalies (> 30 mi) related to deep
sources, such as regional gradients in the Colorado Pla
teau, in the White River uplift, and in the Colorado Min
eral Belt gravity low.
A magnetic anomaly map was compiled showing
regional magnetic anomalies (Toth and others, in press).
The horizontal gradient of the pseudogravity field was cal
culated using the method of Cordell and Grauch (1985).
This method calculates pseudo gravity anomalies caused
by assumed density contrasts of bodies causing magnetic
anomalies, and it is used because gravity anomalies are
simpler to analyze than magnetic anomalies. Maximum
gradient trends of the pseudogravity field were plotted on
a contour map as long, partially discontinuous lines.
Gamma-ray-spectrometer data for the Forest are from
airborne NURE surveys. Measurements detected near-sur
face (shallower than 20 inches) abundances of the radioele
ments uranium, potassium, and thorium. The contractor
221
interpreted amounts of K (in percent), equivalent ura
nium (eu, in ppm), and equivalent thorium (eTh, in ppm)
were contoured and analyzed (Toth and others, in press).
MINING HISTORY
The White River National Forest and Dillon Ranger
District of the Arapaho National Forest have a wealth of
Area gvel
:
() LOr Aid O
-
mineral resources. Mining and mineral exploration have
played a central role in the history of this area and are still
important. The Forest contains 26 mining districts and
mineralized areas (fig. 4).
Most of the mining districts and major orebodies in
the Forest had been discovered by the late 1800's. In
1859, gold placers in the Breckenridge district were the
first deposits to be worked, and gold placers were discov
ered a year later a few miles south of the Forest, at Lead
ville. These two areas include some of the richest placer
deposits in Colorado.
Base- and precious-metal lode deposits were discov
ered a few years after the gold placers. The first silver
vein was discovered in the Forest in the Montezuma min
ing district in 1864. During the next decade, lode deposits
were discovered and exploited throughout the Colorado
Mineral Belt (Tweto and Sims, 1963). In 1878, bonanza
silver deposits were discovered in the Kokomo-Tenmile
district; silver-lead ore was found at Gilman and in the
Aspen district the next year.
The Gilman district, known for its copper and zinc, is
the State's largest producer of metals. The Eagle mine was
the major producer, and it consolidated many older mines
and workings into a single operating unit in 1918. Next in
overall production in the Forest are the Aspen, Brecken
ridge, and Kokomo-Tenmile districts, respectively. Other
districts and mines in the Forest had less significance in
terms of total production.
The Eagle mine in the Gilman district ceased produc
tion in 1981. No large-scale mining has taken place in the
Forest since that time; current mineral activity consists of
annual assessment work on unpatented mining claims,
prospecting, and small-scale mining, mostly in the Breck
enridge area.
MINERAL RESOURCE POTENTIAL
Mineral resource potential information is given in
terms of mineral-deposit types. The deposit types are based
on geological, geochemical, and geophysical characteristics
of known and inferred deposits within or close to the Forest
and similarities to deposits described in Cox and Singer
(1986). Most of the deposit types are represented by type
localities at a mine or within mining districts.
222 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

108 1079 106

T- | I
APPROXIMATE BOUNDARY OF THE
WHITE RIVER NATIONAL FOREST AND
DILLON RANGER DISTRICT OF THE
ARAPAHO NATIONAL FOREST

2.
* e 26 - colorap Riv

40H.
3. a Meeker
B *\
Kremmlinge
-

LDotsero to
25 24

Glenwood
Springs
s
O
s}
SV
C

D
Leadville

*2.
&
3e. 2.
39H. S. -

| | |
O 10 2O 30 MILES

Figure 4. Index map showing mining districts within or close to White River National Forest and Dillon Ranger District of the Arapaho
National Forest, Colorado. Numbers on map are keyed to mining districts as follows: 1, Green Mountain; 2, North Rock Creek; 3, Frisco;
4, Montezuma; 5, Breckenridge; 6, Upper Blue River; 7, Kokomo-Tenmile; 8, Tennessee Pass; 9, Eagle; 10, Gilman; 11, Fulford; 12,
New York Lake; 13, East Lake Creek; 14, East Lake CreekCross Creek; 15, Middle Mountain; 16, Holy Cross CityCross Creek; 17,
Independence; 18, Lincoln Gulch; 19, Aspen; 20, Columbia-Ashcroft; 21, Taylor Peak, 22, Crystal River; 23, Spring Butte; 24, Carbon
ate; 25, Rifle Creek; 26, Uranium Peak. Map modified from Brown (1990).

All available information was assembled and ana LOCATABLE COMMODITIES

lyzed according to procedures outlined by Shawe (1981)
and Taylor and Steven (1983). The assessment of mineral
resource potential in this summary is divided into three
parts: locatable, leasable, and salable commodities. Table
1 summarizes the total size of areas of high, moderate,
low, or unknown resource potential for locatable and leas
able minerals, and figures 510 show areas of resource
potential for all commodities except salable minerals. For
simplicity, only the areas of high and moderate resource
potential are shown on the figures, although areas of low
or unknown potential may be mentioned in the text.
Where areas of differing resource potential overlap, only
the highest degree is shown (i.e., high instead of moder
ate). The reader is referred to Toth and others (in press)
for more detail.
A. Stockwork molybdenum.Formed in the carapaces
of highly differentiated granitic plutons and in the adjacent
country rocks; valuable mainly for molybdenum but also
contain tungsten, tin, and bismuth. Plutons occur in the
southern part of the Forest. Three small areas have high
resource potential, and three small areas have moderate
resource potential (fig. 5).
B. Stockwork copper-molybdenum.Formed in gran
ite bodies; valuable mainly for copper and molybdenum,
but can also contain gold, tungsten, tin, silver, lead, zinc,
and bismuth. One small area in the southern part of the
Forest has high resource potential (fig. 5); a small area just
outside the southeastern part of the Forest has high resource
potential; and two small areas have low resource potential.
 MINERAL RESOURCE POTENTIAL, WHITE RIVER N.F., COLORADO 223

Table 1. Resource potential of lands in the White River National Forest and Dillon Ranger District
of the Arapaho National Forest classified according to type of deposit.
[The White River National Forest and Dillon Ranger District of the Arapaho National Forest contain a total of 3,928 mi. Figures
in columns under each category of resource potential in mi and rounded to the nearest mi. Letters in parentheses preceding the
deposit type indicate the deposit type as discussed below. --, no area of resource potential for that category of resource potential)

Type of deposit Mineral resource potential

High Moderate Low Unknown

Locatable resources

(A) Stockwork Mo.................... l 14 - -

(B) Stockwork Cu-Mo.............. 11 - - -

(C) Skarn ................................ | l 27 --

(D) Polymetallic replacement ... 82 1 || l --

(E) Sherman-type Pb-Zn-Ag..... 7 37 - -

(F) Polymetallic vein................ 382 245 - -

(G) Vein U............................... l l l -

(H) Vein W.............................. - 18 - -

(I) Sandstone Cu..................... 2 - - --

(J) Sandstone U-V................... 54 - 44 -

(K) Placer gold......................... 33 6 8 -

(L) Stratabound sulfides........... -- -- 2 6

(M) Limestone.......................... 175 17 - -

(N) Gypsum ............................. 2 85 -- -

Total locatable resources................. 75 l 535 83 6

Leasable resources

(O) Oil and gas......................... 332 -- - -

(P) Coal................................... 38 33 - -

(Q) Methane............................. 38 33 - -

(R) Oil shale............................ -- 52 -- -

(S) Geothermal energy............. 30 - - -

Total leasable resources.................. 438 118 - -

C. Polymetallic skarn.Formed at the contact
between intrusive igneous rocks and chemically reactive
host rocks, such as limestone; valuable mainly for copper,
lead, zinc, silver, tungsten, manganese, and iron, but may
also contain trace amounts of gold. An area of high
resource potential occurs south of Aspen, where an iron
skarn was previously mined. Two small areas of moderate
resource potential (fig. 5) and two small areas of low
resource potential are in the southeastern part of the Forest.
D. Polymetallic replacement.Formed by hydrother
mal solutions traversing limestone and dolomite; contain
copper, lead, zinc, silver, and manganese, but many also
contain tungsten and gold. The principal area of high
resource potential is along the flanks of the Sawatch
Range where the Leadville Limestone is present (fig. 6A).
Three other small areas have high resource potential, and
two have moderate resource potential; these areas are usu
ally associated with polymetallic vein deposits.
E. Sherman-type Pb-Zn-Ag.Formed by the infilling
of solution-collapse structures in the Leadville Limestone;
contain lead, zinc, and silver. The Leadville Limestone
has a high resource potential in a small area along the
southwestern flank of the Sawatch Range and in a small
area along the eastern flank. The remaining outcrops of
Leadville Limestone have moderate resource potential
(fig. 6B).
F. Polymetallic veins.Related to Laramide and Ter
tiary igneous intrusions; contain lead, zinc, copper, silver,
gold, and manganese; likely byproducts are tin, antimony,
and arsenic. Principal areas of high resource potential are
in the Carbonate, Crystal River, Independence and Lincoln
Gulch, Fulford, New York Lake, East Lake Creek, Cross
Creek, Middle Mountain, Holy Cross City, Gilman, Breck
enridge, Montezuma, North Rock Creek, and Green
Mountain mining districts (figs. 4, 7); three large areas of
moderate resource potential are along the southern and
southeastern parts of the Forest.
G. Vein uranium.Formed from hydrothermal solu
tions associated with igneous intrusions; containing ura
nium and trace amounts of gold, silver, antimony, lead,
zinc, and molybdenum. One small area, northeast of
Aspen, has high resource potential (fig. 7); two small
areas have moderate resource potential.
H. Vein tungsten.Formed from hydrothermal solu
tions associated with igneous intrusions; contain tungsten,
gold, silver, and zinc. Two small areas in the eastern part
of the Forest, near Frisco, have moderate resource poten
tial (fig. 7).
224 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

108 107 106

I | |
APPROXIMATE BOUNDARY OF THE
WHITE RIVER NATIONAL FOREST AND
DILLON RANGER DISTRICT OF THE
ARAPAHO NATIONAL FOREST

Kremmling
- Meeker c

40+ -

[...]

Glenwood
Springs

Leadville
->
3.
z.
wn
EXPLANATION (r,
2.
r,
39- High potential

Moderate potential

20 30 MILES

Figure 5. Mineral resource potential map for stockwork molybdenum (deposit type A), stockwork copper-molybdenum (deposit type
B), and polymetallic skarn deposits (deposit type C) in the White River National Forest and Dillon Ranger District of the Arapaho Na
tional Forest, Colorado.

I. Sandstone copperFormed where copper-rich L. Stratabound sulfide deposits in Proterozoic

solutions encountered reducing environments in sandstone
layers; contain copper and uranium, with byproducts silver
and vanadium. One small area in the Forest, southwest of
Vail, has high resource potential, and another small area,
just west of Aspen, has high resource potential (fig. 8).
J. Sandstone uranium-vanadium.Formed during
diagenesis in locally reducing environments in sandstone;
contain uranium, vanadium, and trace amounts of copper.
Two small areas in eastern part the Forest and one large
area in the northwestern part of the Forest have high
resource potential (fig. 8).
K. Placer gold.Deposited by streams that traversed
and eroded gold-bearing rock; contain gold and minor
quantities of silver and bismuth. One large area near
Breckenridge, along the upper Blue River, has high
resource potential. Seven small areas of moderate resource
potential (fig. 8), in the southeastern part of the Forest, are
downstream from known mining districts.
rocks.Formed where exhalatives from mafic volcanic
rocks interacted with sea water to form brines; deposited
in volcanic and sedimentary rocks in a submarine envi
ronment in Proterozoic time; contain copper, lead, zinc,
gold, and silver. Three small areas in the eastern part of
the Forest have low resource potential; several outcrops
in the west-central part of the Forest and two small areas
in the eastern part of the Forest have unknown resource
potential.
M. High-calcium limestone.Formed in shallow
water marine environment in Mississippian time. Areas
on the flanks of the Sawatch Range, in the southeastern
part of the Forest, and south of the Flat Tops, in the north
western part of the Forest, that are underlain by the Lead
ville Limestone have high resource potential (fig. 9).
Areas with moderate resource potential in the southeastern
part of the Forest are underlain by Leadville Limestone
that has been mineralized.
MINERAL RESOURCE POTENTIAL, WHITE RIVER N.F., COLORADO 225

1079 106

40 I I

EXPLANATION

APPROXIMATE BOUNDARY OF THE

high potential WHITE RIVER NATIONAL FOREST
AND DILLON DISTRICT OF THE
moderate potential ARAPAHO NATIONAL FOREST

39

O 10 2O 30 MILES

107 1069
40

EXPLANATION
APPROXIMATE BOUNDARY OF THE
high potential WHITE RIVER NATIONAL FOREST
AND DILLON DISTRICT OF THE
moderate potential ARAPAHO NATIONAL FOREST

39

Figure 6. A, Mineral resource potential map for polymetallic replacement deposits

(deposit type D), and B, mineral resource potential map for Sherman-type Pb-Zn-Ag
(deposit type E) in the White River National Forest and Dillon Ranger District of the
Arapaho National Forest, Colorado. Where areas of differing potential overlap, only
the area of high potential is shown.


polymetal ic River
for
map
potential
resource
Mineral
7.
Figure
White
the
in
deposits
H)
tungsten
and
G),
uranium
F),
type
(deposit
vein shown;
is
high
area
overlap,
potential
areas
Where
Colorado.
Forest,
Arapaho
the
of
District
Ranger
Dillon
and
Forest
National
(mod)
G
F
s
(mod)
G

(mod)
H
108
1069
1079

Leadville

O

oLoRAP
*-
-

D
3.
Kremmlinga
C Meeker
- D
*

MILES
30
1200
O

AND
FOREST
NATIONAL
RIVER
WHITE
AP ROXIMATE
THE
OF
BOUNDARY THE
OF
DISTRICT
RANGER
DILLON
FOREST
NATIONAL
ARAPAHO

|
|
I

I
L
|

H.
40
L -
39H. potential
Moderate
potential
High
EXPLANATION (mod).
by
indicated
is
potential
moderate
of

3. C
4.


uranium-vanadium in
I),
copper
sandstone
for
map
potential
resource
Mineral
8.
Figure
deposits
K)
gold
placer
and
J),
type
(deposit high
area
only
overlap,
potential
different
areas
Where
Colorado.
Forest,
Arapaho
of
District
Ranger
Dillon
and
Forest
National
River
White
the

108
1069
1079
&
ER |
IV Leadville
R
O
e

coLoRAP
*-
-

[]

'y.v

%
*

KremmlingMeeker[T]
-

3.
MILES

AND
FOREST
NATIONAL
RIVER
WHITE
AP ROXIMATE
THE
OF
BOUNDARY THE
OF
DISTRICT
RANGER
DILLON
FOREST
NATIONAL
ARAPAHO

DotseroD 10
O
2O

l
|

potential
Moderate
&
potential
High
E&
XPLANATION

I
| |

shown.
is
potential

s89
C
H.
40 H.
399
73
 108
1070
1069

I|
APPROXIMATE
BOUNDARY
THE
OF
RIVER
WHITE
FOREST
NATIONAL
AND
RANGER
DILLON
DISTRICT
THE
OF
ARAPAHO
FOREST
NATIONAL

3.
&.

40H

s
89
C

EXPLANATION

potential
High
H.
399
potential
Moderate

_| l
20
1
MILES
30
O0
I

Figure
9.
resource
Mineral
for
map
potential
high-calcium
(deposit
limestone
and
M)
type
evaporite
in
gypsum
River
White
deposits
National
N)
the
Dillon
and
Forest
District
Ranger
Arapaho
the
of
Forest,
Colorado.
National
different
areas
Where
potential
only
overlap,
high
area
shown.
is
 MINERAL RESOURCE POTENTIAL, WHITE RIVER N.F., COLORADO 229

108 107 106

I | I
O APPROXIMATE BOUNDARY OF THE
Q (mod.) WHITE RIVERNATIONAL FOREST AND
DILLON RANGER DISTRICT OF THE
ARAPAHO NATIONAL FOREST

&
e

3. a Meeker
40 H. -

PQ (mod.)

s
89
C

EXPLANATION

High potential
39 H. -

Moderate potential

l l |
O 10 20 30 MILES

Figure 10. Mineral resource potential map for oil and gas (deposit type O), coal (deposit type P), coalbed methane (deposit type Q),
and oil shale (deposit type R), and geothermal resources (deposit type S) in the White River National Forest and Dillon Ranger District
of the Arapaho National Forest, Colorado. Where areas of different potential overlap, the area of high potential is shown; the area of
moderate potential is indicated by (mod).

N. Gypsum in evaporite deposits.-Formed in shal
low-water marine environment. Resource potential is high
north of Aspen, where thick and relatively pure beds of
evaporite are present; resource potential is moderate in
several areas in the northwestern and central parts of the
Forest (fig. 9).
LEASABLE COMMODITIES
O. Oil and gas.-Formed in a near-shore and
coastal-plain environments. Two large areas of high
resource potential are along the western part of the Forest
(fig. 10).
P. CoalFormed in a deltaic system from the
decomposition and alteration of organic remains. One
area in the southwestern part of the Forest has high
resource potential; one area in the northwestern part of
the Forest and two areas in the southwestern part of the
Forest have moderate resource potential (fig. 10).
Q. Coalbed methane.Generated in the maturation
process of coal. One area in the southwestern part of the
Forest has high resource potential; one area in the north
western part of the Forest and two areas in the southwest
ern part of the Forest have moderate resource potential
(fig. 10).
R. Oil shale.Formed in a large, shallow lake and
on broad playa fringes. One large area in the southwest
ern part of the Forest has moderate resource potential
(fig. 10).
S. Geothermal energy.Resource areas consist of
thermal springs located along fault zones. One large area
near Glenwood Springs has high resource potential and
two smaller areas in the south-central part of the Forest
have high resource potential (fig. 10).
230 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 2. Quantitative estimates of undiscovered resources of stockwork molybdenum, polyme

tallic vein, polymetallic replacement, and placer gold deposits in the White River National Forest
and Dillon Ranger District of the Arapaho National Forest.
[Values, except number of deposits, in metric tonnes unless indicated as troy ounces (oz). Percentages in column headings
refer to conidence levels for estimates of undiscovered deposits occurring in the Forest]

Metal 90o 5000 1096 Mean

Stockwork molybdenum
Mo.. () 230,000 1,200,000 430,000
Total mineralized rock ......... O 130,000,000 620,000,000 220,000,000
Number of deposits............... () | l
Polymetallic veins
Pb. 57 4,600 55,000 20,400
Zn. () 1,300 39,000 14,000
Cu...... () 53 370 330
Ag 16,400 oz 1,029,000 oz 14,468,000 oz 8,037,500 oz
Au.... () 612 oz 41,800 oz 22,800 oz
Total mineralized rock.......... 350 56,000 840,000 280,000
Number of deposits............... () 3 5
Polymetallic replacement
Pb. () 29,000 510,000 220,000
Zn...... () 14,000 620,000 250,000
Cu.. () () 21,000 9,800
Ag () 1,896,900 oz 67,515,400 oz. 26,041,500 oz
Au... () () 147,900 oz 99,600 oz
Total mineralized rock.......... () 6,200,000 12,000,000 4,100,000
Number of deposits............... () l l
Placer gold
Au.. 180 oz 3,100 oz 23,000 oz 7,000 oz
Total mineralized rock.......... 1,500 130,000 1,000,000 330,000
Number of deposits............... () l 2

SALABLE COMMODITIES ASSESSMENT OF METAL

Sand and gravel.Numerous deposits of sand and
gravel are located within major drainages in the Forest and
are exploited for concrete and structural aggregate materi
als, road fill, mortar, and other uses. Fifty-three sites that
have produced sand and gravel are on Forest land.
Dimension stone.Sources in the Forest include mar
ble and moss rock. White marble has been mined south of
Marble, near Yule Creek, and black marble has been
mined south of Aspen, in Conundrum Creek. Numerous
moss-rock sites are in the Forest, and the potential
resources are essentially unlimited.
Crushed and lightweight aggregate.Crushed aggre
gates usable for roadway materials and general construc
tion are derived from limestone, basalt, and granite
sources. Rock quarries are scattered around the Forest
where topography does not inhibit excavation. Light
weight aggregates include volcanic ash, pumice, and sco
ria. These are scattered around the Forest in limited
quantities and contain varying amounts of impurities.
Clay.Clay resources can be found in the Dakota
Sandstone, south of Dillon Reservoir. Where clays have
been sampled, they are not of refractory-brick quality.
ENDOWMENT USING GEOLOGIC AND
GRADE-TONNAGE MODELS
At the request of the U.S. Forest Service, the USGS
provided a quantitative assessment of the undiscovered
mineral resources that might exist in the Forest. Based on
the geology, geophysics, and geochemistry of known
deposits in the Forest, deposit types were defined and
compared to other similar deposits worldwide. Using a
computer program entitled MARK3, tonnages for undis
covered deposits in the Forest were estimated from known
tonnages and grades of deposits worldwide. The results
are presented in table 2. Note that these estimates do not
imply that the resources would be economic to produce.
Such a determination would require a complete analysis of
both the costs of discovery and engineering feasibility, and
would require an economic evaluation of production.
Only four deposit types have sufficient grade and
tonnage information to assess them using the MARK3
program: stockwork molybdenum, polymetallic replace
ment, polymetallic vein, and placer gold deposits. All
 MINERAL RESOURCE POTENTIAL, WHITE RIVER N.F., COLORADO
remaining deposits that are known or predicted for the
Forest lack sufficient data, or the deposit type is too
poorly defined for quantitative assessment.
REFERENCES CITED
Behrendt, J.E., and Bajwa, L.Y., 1974, Bouguer gravity map of
Colorado: U.S. Geological Survey Geophysical Investigations
Map GP-895, scale 1:500,000.
Bolivar, S.L., Hill, D.E., Bunker, M.E., Cheadle III, J., Minor,
M.M., Sandoval, W., Talcott, C.L., Trujillo, L., and Water
bury, G.R., 1979, Uranium hydrogeochemical and stream sed
iment reconnaissance data release for the Craig NTMS
quadrangle, Colorado, including concentrations of forty-three
additional elements: U.S. Department of Energy Open-File Re
port GJBX76(79), 238 p., 5 plates.
Broxton, D.E., Morris, W.A., Bolivar, S.L., Apel, C.T., Gallimore,
D.L., George, W.E., Hensley, W.K., McInteer, C., Minor,
M.M., and Zelezny, W.F., 1979, Uranium hydrogeochemical
and stream sediment reconnaissance data release for the Mon
trose NTMS quadrangle, Colorado, including concentrations
of forty-three additional elements: U.S. Department of Energy
Open-File Report GJBX125(79), 255 p., 5 plates.
Brown, S.D., 1990, Mineral appraisal of the White River National
Forest, Colorado: U.S. Bureau of Mines Open-File Report
MLA 990, 378 p.
Campbell, D.L., 1981, Aeromagnetic and complete Bouguer grav
ity anomaly maps of the Hunter-Fryingpan Wilderness area,
Pitkin County, Colorado: U.S. Geological Survey Miscella
neous Field Studies Map MF-1236C, scale 1:50,000.
Campbell, D.L., 1985, Gravity and aeromagnetic maps of the Ma
roon BellsSnowmass Wilderness and additions, Pitkin and
Gunnison Counties, Colorado: U.S. Geological Survey Mis
cellaneous Field Studies Map MF-1647B, scale 1:100,000.
Campbell, D.L., and Wallace, A.R., 1986, Aeromagnetic map of
the Holy Cross Wilderness area, Eagle, Lake, and Pitkin Coun
ties, Colorado: U.S. Geological Survey Miscellaneous Field
Studies Map MF1841B, scale 1:100,000.
Case, J.E., 1965, Gravitational evidence for a batholithic mass of
low density along a segment of the Colorado Mineral Belt
[abs.]: Geological Society of America Special Paper 82, p. 26.
1966, Geophysical investigations over Precambrian rocks,
northwestern Uncompahgre Plateau, Utah and Colorado:
American Association of Petroleum Geologists Bulletin, v. 50,
p. 14231443.
1967, Geophysical ore guides along the Colorado Mineral
Belt: U.S. Geological Survey Open-File Report 67039, 13 p.
Cordell, Lindrith, and Grauch, V.J.S., 1985, Mapping basement
magnetization zones from aeromagnetic data in the San Juan
Basin, New Mexico, in Hinze, W.J., ed., The Utility of Region
al Gravity and Magnetic Anomaly Maps: Society of Explora
tion Geophysics, p. 181197.
Cox, D.P., and Singer, D.A., 1986, Mineral deposit models: U.S.
Geological Survey Bulletin 1693,379 p.
Freeman, V.L., Campbell, D.L., King, H.D., Weisner, R.C., and
Bieniewski, C.L., 1985, Mineral resource potential of the Ma
roon BellsSnowmass Wilderness and additions, Gunnison
231
and Pitkin Counties, Colorado: U.S. Geological Survey Mis
cellaneous Field Studies Map MF-1647A, scale 1:100,000.
Hildenbrand, T.G., 1983, FFTFILA filtering program based on
two-dimensional Fourier analysis on geophysical data: U.S.
Geological Survey Open-File Report 83237, 61 p.
Isaacson, L.B., and Smithson, S.B., 1976, Gravity anomalies and
granite emplacement in west-central Colorado: Geological So
ciety of America Bulletin, v. 87, p. 2228.
Langfeldt, S.L., Youngquist, C.A., D'Andrea, Jr., R.F., Zinkl, R.J.,
Shettel, Jr., D.L., Broxton, D.E., Hansel, J.N., McInteer, C.,
and Minor, M.M., 1981, Uranium hydrogeochemical and
stream sediment reconnaissance data release for the Grand
Junction NTMS Quadrangle, Colorado/Utah: U.S. Department
of Energy Open-File Report GJBX264(81), 142 p., 7 plates.
Ludington, Steve, and Ellis, C.E., 1981, Mineral resource potential
of the Hunter-Fryingpan Wilderness area and the Porphyry
Mountain Wilderness Study Area, Pitkin County, Colorado:
U.S. Geological Survey Miscellaneous Field Studies Map
MF-1236D, scale 1:50,000.
Mallory, W.W., Post, E.V., Ruane, P.J., Lehmbeck, W.L., and Sto
telmeyer, R.B., 1966, Mineral resources of the Flat Tops Prim
itive Area, Colorado: U.S. Geological Survey Bulletin
1230C, 30 p.
Moss, C.K., and Abrams, G.A., 1985, Geophysical maps of the
Vasquez Peak Wilderness Study Area and the Williams Fork
and St. Louis Peak Roadless Areas, Clear Creek, Grand, and
Summit Counties, Colorado: U.S. Geological Survey Miscella
neous Field Studies Map MF-1588D, scale 1:50,000.
Mutschler, F.E., Larson, E.E., and Bruce, R.M., 1987, Laramide
and younger magmatism in ColoradoNew petrologic and
tectonic variations on old themes, in Drexler, J.W., and Larson,
E.E., eds., Cenozoic Volcanism in the Southern Rocky Moun
tains Revisited: A Tribute to Rudy C. EpisPart 1: Colorado
School of Mines Quarterly, v. 82, no. 4, p. 147.
Planner, H.N., Apel, C.T., Fuka, M.A., George, W.E., Hansel, J.M.,
Hensley, W.K., and Pirtle, J., 1981, Uranium hydrogeochemi
cal and stream sediment reconnaissance data release for the
Leadville NTMS quadrangle, Colorado, including concentra
tions of forty-two additional elements: U.S. Department of En
ergy Open-File Report GJBX13(81), 185 p., 1 plate.
Shawe, D.R., 1981, U.S. Geological Survey workshop on non-fuel
mineral-resource appraisal of Wilderness and CUSMAP areas:
U.S. Geological Survey Circular 845, 18 p.
Shettel, Jr., D.L., Langfeldt, S.L., Youngquist, C.A., D'Andrea, Jr.,
R.F., Zinkl, R.J., Bolivar, S.L., Garcia, S.R., George, W.E., and
Hanks, D., 1981, Uranium hydrogeochemical and stream sedi
ment reconnaissance data release for the Denver NTMS Quad
rangle, Colorado: U.S. Department of Energy Open-File
Report GJBX263(81), 153 p., 17 plates.
Soulliere, S.J., Arnold, M.A., Kluender, S.E., and Zelten, J.E.,
1985a, Mineral resources of the Eagle Mountain Wilderness
Study Area, Pitkin County, Colorado: U.S. Geological Survey
Bulletin 1717B, 9 p.
Soulliere, S.J., Arnold, M.A., Kluender, S.E., 1985b, Mineral re
sources of the Hack Lake Wilderness Study Area, Garfield
County, Colorado: U.S. Geological Survey Bulletin
1717A, 5 p.
Soulliere, S.J., Arnold, M.A., Hassemer, J.R., Martin, R.A., Klu
ender, S.E., and Zelten, J.E., 1986, Mineral resources of the
232 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Bull Gulch Wilderness Study Area, Eagle County, Colorado:
U.S. Geological Survey Bulletin 1717C, 12 p.
Taylor, R.B., and Steven, T.A., 1983, Definition of mineral
resource potential: Economic Geology, v. 78, no. 6, p.
12681270.
Theobald, P.K., Bielski, A.M., Eppinger, R.G., Moss, C.K., Kre
idler, T.J., and Barton, H.N., 1983, Mineral resource potential
map of the Vasquez Peak Wilderness Study Area, and the Wil
liams Fork and St. Louis Peak Roadless Areas, Clear Creek,
Grand, and Summit Counties, Colorado: U.S. Geological Sur
vey Miscellaneous Field Studies Map MF 1588A, scale
1:50,000.
Toth, M.I., Wilson, A.B., Cookro, T.M., Bankey, Viki, Lee, G.K.,
and Case, J.E., in press, Mineral resource potential and geology
of the White River National Forest and the Dillon Ranger Dis
trict of the Arapaho National Forest, Colorado, with a section
on Salable commodities, by J.S. Dersch: U.S. Geological Sur
vey Bulletin 2035.
Tweto, Ogden, 1980, Tectonic history of Colorado, in Kent, H.C.,
and Porter, K.W., eds. Colorado Geology: Denver, Rocky
Mountain Association of Geologists, p. 59.
Tweto, Ogden, and Case, J.E., 1972, Gravity and magnetic features
as related to geology in the Leadville 30-minute quadrangle,
Colorado: U.S. Geological Survey Professional Paper 726C,
31 p.
Tweto, Ogden, and Sims P.C., 1963, Precambrian ancestry of the
Colorado Mineral Belt: Geological Society of America Bulle
tin, v. 74, p. 9911014.
Tweto, Ogden, Bryant, Bruce, and Williams, F.E., 1970, Mineral
resources of the Gore RangeEagle's Nest Primitive Area and
vicinity, Summit and Eagle Counties, Colorado: U.S. Geolog
ical Survey Bulletin 1319C, 127 p.
Van Loenen, R.E., Lee, G.K., Campbell, D.L., and Thompson, J.R.,
1989, Mineral resource appraisal of the Mount Massive Wil
derness, Lake County, Colorado: U.S. Geological Survey Bul
letin 1636, 18 p.
Wallace, A.R., 1990, Regional geologic and tectonic setting of the
central Colorado Mineral Belt, in Beaty, D.W., Landis, G.P.,
and Thompson, T.B., eds., Carbonate-Hosted Sulfide Deposits
of the Central Colorado Mineral Belt: Economic Geology
Monograph No. 7, p. 1928.
Wallace, A.R., Lee, G.K., Campbell, D.L., Lundby, William, and
Brown, S.D., 1989, Mineral resources of the Holy Cross Wil
derness Area, Eagle, Pitkin, and Lake Counties, Colorado: U.S.
Geological Survey Bulletin 1879, 22 p.
 CHAPTER W
GEOPHYSICAL STUDIES OF THE WHITE RIVER
NATIONAL FOREST, NORTHWESTERN AND
CENTRAL COLORADO
By VIKI BANKEY!
INTRODUCTION
The U.S. Geological Survey (USGS) is undertaking a
multidisciplinary study to assess the mineral resource
potential of the White River National Forest and part of
the Arapaho National Forest (referred to in this report as
the Forest; fig. 1). The results of this study are summa
rized elsewhere in this volume (Toth and others). Twenty
six mining districts or mineralized areas are found within
or adjacent to the Forest boundaries and include some of
the largest concentrations of metallic mineral deposits in
the Rocky Mountain region, notably in the Gilman, Lead
ville, and Climax districts. Most of these deposits formed
within a zone of hydrothermal mineralization that charac
terizes the Colorado Mineral Belt in the southeastern part
of the Forest (figs. 1 and 2). The deposits of the Colorado
Mineral Belt are associated with Late Cretaceous to late
Tertiary felsic to intermediate intrusive and volcanic rocks
that were localized by numerous northeast-trending base
ment shear faults (Toth and others, this volume). A promi
nent 3050 milli(Sal (mGal) gravity low is centered over
the Colorado Mineral Belt and is produced by a low-den
sity, silicic, batholithic mass that underlies a large part of
the area (Case, 1965; Tweto and Case, 1972; Isaacson and
Smithson, 1976). The outline of this gravity low is shown
on figures 1 and 2. Many of the mapped intrusions and
related mineralized areas probably formed as apophyses
from the top of this massive batholith.
Gravity and aeromagnetic anomaly maps were com
piled and interpreted for the White River National Forest
area as part of the assessment study. The present work
builds on previous geophysical studies made for the Lead
"U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
ville 1x2 quadrangle (Wallace and others, 1988) and for
six Wilderness Study Areas (Campbell, 1981, 1985;
Campbell and Wallace, 1986; Moss and Abrams, 1985;
and Van Loenen and others, 1989). Geophysical studies to
support mineral resource evaluations of the Colorado Min
eral Belt were made by Case and Tweto (Case, 1965,
1966, 1967; Tweto and Case, 1972), Behrendt and Bajwa
(1974), and Isaacson and Smithson (1976).
Gravity and magnetic data contribute significantly to
the understanding of geologic features associated with min
eral deposits. Some of these features include Proterozoic
faults and shear zones, sometimes characterized by offsets
in gravity or magnetic anomalies or both, that provided
conduits for Phanerozoic intrusive magmas and mineraliz
ing fluids that formed the metallic mineral deposits of the
Colorado Mineral Belt. Gravity anomalies may distinguish
low-density porous sandstones (which may host copper, sil
ver, uranium, or vanadium deposits in the Forest) or Paleo
zoic evaporites (host to gypsum, halite, and potassium salts,
and oil and gas reservoirs) or high-density limestones
(which are host to many of the significant metallic ore
deposits in the area (Toth and others, in press)).
GRAVITY STUDIES
Gravity anomalies occur from the juxtaposition of
rocks having measurable density contrasts, which result
from structural or geologic features such as faults, folds,
downwarps, intrusions, basin fill, lithologic contacts, or
facies changes. The number and quality of gravity stations
limits the accuracy of anomaly definition, especially in
mountainous areas where station spacing is often sparse.
However, on a regional scale, gravity mapping is a useful
tool for locating structural breaks and folds and for delin
eating shallow, buried pediments.
234 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

108 1079 106

10.

9.

8.

7.
6.

5.

- 4.

| 2.

! .

0.0

-1 .

- 2.

3.

- 4.

5.

6.

7.

8.

o i 9.
3 || -10.
E - 1 1 .

5 L -30.

Figure 1. High-pass filtered residual gravity map of the White River National Forest. The map shows the Bouguer gravity field with
broad, presumably deep, features removed and shallower features emphasized. Bouguer gravity mapping (Behrendt and Bajwa, 1974)
shows a broad gravity low (hachured outlines) over the low-density batholith underlying the Colorado Mineral Belt.

In the Colorado Mineral Belt, gravity mapping has AEROMAGNETIC STUDIES

defined a broad area that is underlain by a low-density
batholith that is the presumed source of mineralization
(Behrendt and Bajwa, 1974). A shallow intracrustal origin
for the low, having an apex within a few thousand feet of
the surface, a depth extending 40,000 ft below sea level,
and a width averaging 1520 mi, has been demonstrated
by gravity models (Case, 1966; Tweto and Case, 1972;
Isaacson and Smithson, 1976). The broad, deep gravity
low resulting from the batholith masks gravity anomalies
resulting from smaller, shallower, and more mineralogi
cally significant intrusions and hydrothermally altered
areas. To minimize the gravity effect of the deep batholith,
a mathematical filter was applied to a 1.2 mi (2 km) grid
ded data set of complete-Bouguer gravity data that was
produced using gravity data from 6,500 stations. This fil
ter removes wavelengths longer than about 30 mi from the
Bouguer gravity grid (Hildenbrand, 1983). The resulting
high-pass residual gravity map emphasizes anomalies due
to shallow sources by suppressing long-wavelength anom
alies caused by broad, often deep, regional features, such
as the Colorado Mineral Belt, Colorado Plateau, and the
White River uplift.
Aeromagnetic anomalies are caused by rocks that
contain significant amounts of magnetic minerals; these
anomalies reflect variations in the amount and type of
magnetic material and the shape of the body. In general,
igneous and metamorphic rocks contain enough magnetic
minerals to generate magnetic anomalies, whereas sedi
mentary rocks are commonly virtually nonmagnetic.
Aeromagnetic anomaly maps are important tools in
mapping surficial and buried igneous rocks. Aeromagnetic
data can be used to locate and estimate depths to igneous
intrusions that may be related to possible mineral deposits.
For example, magnetic highs are caused by magnetic plu
tons in the Elk Mountains. High-intensity magnetic lows
may indicate igneous rocks that acquired their magnetic
properties during a period of magnetic field reversal, such
as the magnetic lows associated with some outcrops of
basaltic rocks in the White River uplift. Rings of magnetic
highs with central or reentrant lows may indicate porphyry
systems in which hydrothermal alteration has destroyed
preexisting magnetic minerals. Local magnetic highs may
exist where hydrothermal alteration has created secondary
 GEOPHYSICAL STUDIES, WHITE RIVER N.F., COLORADO
108 1079
Aproximate Forest
40
|
* Batt.
Figure 2. Aeromagnetic anomaly map of the White River National Forest and vicinity. The aeromagnetic maps of the area were digitally
merged to generate this aeromagnetic compilation map.
magnetic minerals; for example, in a magnetite-bearing
skarn. Small magnetic lows may occur where less mag
netic Tertiary stocks have intruded more magnetic Protero
zoic rocks.
Nine aeromagnetic surveys were flown over separate
parts of the Forest, with flight-line spacing ranging from
0.5 to 3 mi apart. These aeromagnetic data sets were sepa
rately gridded and digitally merged to a common datum to
produce the map in figure 2.
GEOPHYSICAL INTERPRETATIONS
The Forest has been divided into three geophysical
provinces for this report: the Colorado Mineral Belt and
Gore Range in the eastern and southern Forest; the White
River uplift in the northern Forest; and the Battlement
Mesa area in the western Forest.
In the Colorado Mineral Belt, residual gravity lows
may indicate the presence and subsurface extent of shal
low apophyses of the low-density, silicic, batholithic mass
of Cretaceous to Tertiary age that is presumed to underlie
a large part of the belt. One such string of gravity lows
235
1069
i
occurs on a northeast trend from the stocks at Climax to
Montezuma to Red Mountain (fig. 1), suggesting that
these features are related. Because buried batholiths play
such an important role as a source of mineralizing fluids
and heat in this area, this trend of gravity lows may be
significant.
Among aeromagnetic anomalies of interest to mineral
assessments are the high-amplitude, high-frequency anom
alies associated with the mid-Tertiary outcrops of the Mt.
Sopris, Snowmass, and White Rock plutons that are in the
Elk Mountains in the south-central part of the Forest.
These three plutons are connected at depth, as suggested
by the gravity high that encompasses the area and demon
strated in models by Campbell (1985). However, these
plutons are separated from the Ragged Mountain pluton to
the southwest by the geophysically distinct granite of
Treasure Mountain. The gravity and magnetic anomalies
show the subsurface extent of these mineralized plutons
and indicate some occurrences of magnetic alteration.
In the Sawatch and Gore Ranges, Cretaceous and Ter
tiary intrusive rocks are commonly, but not always, associ
ated with gravity and magnetic lows, whereas the
Proterozoic schists, granites, and gneisses are characterized
236 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
by magnetic highs. Magnetic gradients, caused by contrast
ing lithologies or locally increased magnetization, mark
Proterozoic shear zones and metamorphic layering and
northwest-trending, uplift-related, high-angle faults of
Laramide and younger age. These numerous zones of
crustal weakness are often the location of mineral deposits
(Toth and others, in press).
Between the Sawatch Range and the Gore Range lies
an interesting east-northeast-trending zone of magnetic
lows that corresponds to a residual gravity low. This
zone, which extends from west of Vail to Dillon and
crosses many north-northwest-trending topographic and
structural features, may reflect a larger volume of Phaner
ozoic intrusive rocks.
The magnetic field in the area of the White River
uplift is characterized by high-frequency, small, positive
and negative anomalies over basalt outcrops, some of
which formed during periods of reversed magnetization.
Low residual gravity values occur where basalt becomes
the major rock type at the surface. Because the basalt
magmas intruded through evaporites, the potential exists
for Norilsk-type copper-nickel-platinum deposits (Toth
and others, in press). The southern and southwestern mar
gins of the uplift are flanked by four magnetic highs that
are all similar in appearance, magnitude, extent, and
depth-to-source. A broad, positive gravity anomaly is
associated with these magnetic highs. Because of the loca
tion of these anomalies along the edge of the uplift, they
are presumed to be structurally controlled and may be
caused by locally exposed Proterozoic rocks that are found
in the bottom of some canyons in the southern White
River uplift.
Gravity lows east of the White River uplift are asso
ciated with evaporite facies rocks within the Eagle Valley
Evaporite. These rocks are a source of gypsum, halite, and
potassium salts and may also host oil and gas deposits
(Toth and others, this volume).
The Battlement Mesa area, in the westernmost part of
the Forest, is associated with a residual gravity low and a
broad magnetic high that reflect a presumably deep
source. However, the data are of poor resolution so that
smaller, shallower sources are not represented well.
CONCLUSIONS
Analysis of geophysical features provide new insights
into structural history and subsurface configurations and
can help geologists target new areas for mineral explora
tion. Recognizable patterns of geophysical anomalies cor
respond to geological targets for mineral potential, such as
Proterozoic shear zones and Phanerozoic faults that pro
vided conduits for mineralizing fluids, Cretaceous and
Tertiary intrusive rocks that provided the fluids and heat
for hydrothermal systems, Paleozoic sedimentary rocks
that serve as host rocks for metallic minerals, and evapor
ite-sequence rocks that are both host rocks for oil and gas
deposits and source rocks for salts.
REFERENCES CITED
Behrendt, J.E., and Bajwa, L.Y., 1974, Bouguer gravity map of
Colorado: U.S. Geological Survey Geophysical Investigations
Map GP-895, scale 1:500,000.
Campbell, D.L., 1981, Aeromagnetic and complete Bouguer grav
ity anomaly maps of the Hunter-Fryingpan Wilderness area,
Pitkin County, Colorado: U.S. Geological Survey Miscella
neous Field Studies Map MF-1236C, scale 1:50,000.
1985, Gravity and aeromagnetic maps of the Maroon
BellsSnowmass Wilderness and additions, Pitkin and Gunni
son Counties, Colorado: U.S. Geological Survey Miscella
neous Field Studies Map MF-1647B, scale 1:100,000.
Campbell, D.L., and Wallace, A.R., 1986, Aeromagnetic map of
the Holy Cross Wilderness Area, Eagle, Lake, and Pitkin
Counties, Colorado: U.S. Geological Survey Miscellaneous
Field Studies Map MF-1841B, scale 1:100,000.
Case, J.E., 1965, Gravitational evidence for a batholithic mass of
low density along a segment of the Colorado Mineral Belt
[abs.): Geological Society of America Special Paper 82, p. 26.
1966, Geophysical investigations over Precambrian rocks,
northwestern Uncompahgre Plateau, Utah and Colorado:
American Association of Petroleum Geologists Bulletin, v. 50,
p. 14231443.
1967, Geophysical ore guides along the Colorado Mineral
Belt: U.S. Geological Survey Open-File Report 67039, 13 p.
Hildenbrand, T.G., 1983, FFTFILA filtering program based on
two-dimensional Fourier analysis on geophysical data: U.S.
Geological Survey Open-File Report 83237, 61 p.
Isaacson, L.B., and Smithson, S.B., 1976, Gravity anomalies and
granite emplacement in west-central Colorado: Geological So
ciety of America Bulletin, v. 87, p. 2228.
Moss, C.K., and Abrams, G.A., 1985, Geophysical maps of the
Vasquez Peak Wilderness Study Area and the Williams Fork
and St. Louis Peak Roadless Areas, Clear Creek, Grand, and
Summit Counties, Colorado: U.S. Geological Survey Miscella
neous Field Studies Map MF-1588D, scale 1:50,000.
Toth, M.I., Wilson, A.B., Cookro, T.M., Bankey, Viki, Lee, G.K.,
and Case, J.E., in press, Mineral resource potential and geology
of the White River National Forest and the Dillon Ranger Dis
trict of the Arapaho National Forest, Colorado, with a section
on Salable commodities, by J.S. Dersch: U.S. Geological Sur
vey Bulletin 2035.
Tweto, Ogden, and Case, J.E., 1972, Gravity and magnetic features
as related to geology in the Leadville 30-minute quadrangle,
Colorado: U.S. Geological Survey Professional Paper 726C.
31 p.
Van Loenen, R.E., Lee, G.K., Campbell, D.L., and Thompson, J.R.,
1989, Mineral resource appraisal of the Mount Massive Wil
derness, Lake County, Colorado: U.S. Geological Survey Bul
letin 1636, 18 p.
Wallace, A.R., Ludington, Steve, Lovering, T.G., Campbell, D.L.,
Case, J.E., Grauch, V.J.S., and Knepper, Dan, 1988, A mineral
preassessment of the Leadville 1x2 quadrangle, Colorado:
U.S. Geological Survey Open-File Report 88074, 56 p.
 CHAPTER X
LOCATING BURIED CONDUCTIVE MATERIAL
ALONG THE GETCHELL TREND,
OSGOOD MOUNTAINS, NEVADA:
IMPLICATIONS FOR GOLD EXPLORATION AND
THE CARBON-GOLD ASSOCIATION(?)
By V.J.S. GRAUCH' and DONALD B. HOOVER"
INTRODUCTION
The Getchell trend is an alignment of six sedimen
tary-hosted, disseminated gold deposits along the eastern
side of the Osgood Mountains in north-central Nevada
(fig. 1). In 1988, the U.S. Geological Survey conducted a
multi-sensor airborne geophysical program in this area in
order to demonstrate the utility of integrated airborne geo
physical surveying for exploration or mineral assessment
in covered terranes. The surveys included an airborne
electromagnetic (EM) survey, flown using three different
frequencies. Apparent resistivity maps were constructed
for each frequency; these maps are standardly used to
interpret lateral variations in subsurface resistivities. (The
word apparent is used because there are some simplify
ing assumptions made about the Earth in order to make
the resistivity computation from measured EM parame
ters.) In a preliminary assessment of the apparent resistiv
ity maps for the Getchell-trend area, Hoover and others
(1991) observed that all the exposed gold deposits are
associated with linear conductive zones. Because EM sur
veys can map subsurface resistivities, apparent resistivity
maps, thus, can be an important guide to locating buried
conductive zones, which in turn may be associated with
buried gold deposits.
Unfortunately, the geophysical signature of buried
conductive material is commonly masked by the effect of
overlying or neighboring rocks, making identification on
"U.S. Geological Survey, Mail Stop 964, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
apparent resistivity maps difficult. A step toward isolating
this signature is to take the ratio of apparent resistivities
calculated for two different frequencies (Pierce and
Hoover, 1991). Analogous to a derivative, the ratio of
apparent resistivity at a higher frequency divided by that
for a lower frequency is a measure of how fast resistivity
is changing with depth. Buried conductive material, which
we define as rocks with very low resistivity underlying
rocks with moderate to high resistivity, will then be repre
sented by areas where resistivity is decreasing with depth.
However, areas where highly resistive rocks overlie mod
erately resistive rocks also express decreasing resistivity
and, yet, are areas of no interest. Thus, a better approach
enhances areas where resistivity is decreasing with depth
and the apparent resistivity is low. We call such an opera
tion a buried-conductor enhancement.
The purpose of this report is two-fold: (1) to intro
duce a buried-conductor enhancement of resistivity data
that pinpoints buried conductive material and thus can bet
ter isolate subsurface areas of interest; and (2) to argue
that the primary conductive material reflected in the
enhanced map is the carbonaceous matter that has a com
mon association with deposits of this type. In addition,
perhaps the conductive properties of the carbonaceous
matter have some sort of indirect or direct, but as yet
unknown, relationship to the gold mineralization.
GEOLOGIC SETTING
The Getchell trend lies along the eastern side of the
Osgood Mountains, a structurally complicated range
238 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
a location
Figure 1. Location of the six sedimentary-hosted gold deposits
that make up the Getchell trend along the east side of the Osgood
Mountains. Stippling represents topographic ranges.
composed of Paleozoic sedimentary rocks, Cretaceous gra
nodiorite, and Tertiary extrusive rocks (fig. 2). The
Osgood Mountains host deposits of silver, tungsten, barite,
and manganese as well as gold. The gold deposits are sim
ilar to the well-known Carlin deposit; disseminated, invisi
ble gold is hosted in Paleozoic, thin-bedded,
carbonaceous, calcareous siltstones and silty carbonates.
Mineralization is controlled locally by the Getchell fault
system at the Getchell, Pinson, and possibly Preble depos
its (Bagby and Berger, 1985). Similar range-front faulting
controls mineralization at the Chimney Creek (Osterberg
and Guilbert, 1991) and Mag (Foster and Kretschmer,
1991) deposits. The Rabbit Creek deposit is more strati
form than fault controlled (Bloomstein and others, 1991).
Carbonaceous matter is abundant at all the Getchell
trend gold deposits; this association is typical of sediment
hosted, disseminated gold deposits in general, although no
direct correlation between gold and organic carbon has
been documented within the deposits themselves (Percival
and others, 1988; Bagby and Berger, 1985). The carbon
aceous matter may have originated as hydrocarbons that
were partially remobilized by hydrothermal fluids along
fractures and fault zones (Hausen and Park, 1986; Broili
and others, 1988) and then heated to anthracite (Hausen
and Park, 1986) or higher (Leventhal and Hofstra, 1990)
coal rank. Studies at the Jerritt Canyon deposit in north
eastern Nevada demonstrate that graphitization of the car
bonaceous matter occurred prior to gold mineralization so
that carbon was unavailable for gold transport (Leventhal
and others, 1987). Only a few gold grains are spatially
related to carbonaceous matter (Leventhal and others,
1987; Hausen and Park, 1986). In any case, the genetic
relationship between the carbonaceous matter and gold
mineralization, if any, remains unclear.
APPARENT RESISTIVITY DATA
Resistivity (or conductivity, its inverse) is primarily a
function of the interconnected porosity of the rock, the
amount and quality of contained water, and degree of satu
ration (Keller and Frischknect, 1966). Concentrations of
electronically conductive minerals, such as metallic sul
fides and graphite, can significantly lower resistivities
locally but generally contribute little to bulk resistivities of
rock units. Thus, bulk-earth resistivity is normally deter
mined by ionic conduction through fluids contained in the
pore spaces of rocks. As a rule of thumb, rocks with low
porosity, such as igneous rocks, will have higher resistivi
ties than rocks with high porosity, such as some clastic
sedimentary rocks. On the other hand, a rock unit with
moderate porosity in which the pores are filled with saline
water instead of fresh water has an extremely low resistiv
ity. In addition, large amounts of clay minerals decrease
resistivities dramatically because of their ability to absorb
water (Palacky, 1986).
The depth of exploration of an EM survey is depen
dent on the frequency of the EM measurements and the
resistivity of the subsurface. Low-frequency measurements
have greater exploration depths than high-frequency mea
surements, and the depth of exploration is greater in resis
tive terranes than in conductive ones. For example, at a
given location, depth of exploration using 900 Hz is
greater than for 7,200 or 56,000 Hz. Between locations,
the 900-Hz depth of exploration is greater where subsur
face resistivities are higher.
Apparent resistivity data for the Getchell-trend area
were computed from data collected during an airborne EM
survey flown in 1988 using a DIGHEM IV system. The
Figure 2 (facing page). Generalized geology of the Osgood
Mountains and location of precious metal occurrences. Rock units
adapted primarily from Willden (1964). Additional geologic
names from Hotz and Willden (1964); ages from Erickson and
Marsh (1974) and Silberman and others (1974). CC, Chimney
Creek central pit; CN, Chimney Creek north pit; CS, Chimney
Creek south pit; GC, Getchell central pit; GN, Getchell north pit;
GS, Getchell south pit; HC, Hansen Creek mine; MA, Mag mine;
PI, Pinson mine: PR, Preble mine; RC, Rabbit Creek pit (future
outline); SCA, Summer Camp gold mine; SCO, Silver Coin silver
mine; TR, Turquoise Ridge gold mineralization; 13, section 13
gold mineralization.
 BURIED CONDUCTIVE MATERIAL, GETCHELL TREND, NEVADA 239

117 15'
I
EXPLANATION

Ouaternary alluvium -
Contact

Selected normal fault,

T Pliocene basalt ball on down-thrown side
A Selected thrust fault,
Tba Miocene basalt and basaltic andesite sawteeth on side of upper plate

Pit outline of gold deposits

Cretaceous granodiorite(about 90 Ma)
- Known gold mineralization
Paleozoic sedimentary rods, undivided
A. Silver deposit
Ordovician Comus Formation

Ordovician Valmy Formation

Cambrian Preble Formation

41

15'

41 |_
OO'

o 5 MILES

H
O 5 KILOMETERS
240 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
survey was flown 30 m (100 ft) above ground along north
west- and southeast-trending lines spaced 402 m (1/4
mile) apart (except over the gold deposits, where they
were spaced 201 m (1/8 mile) apart). Data were acquired
using frequencies of 56,000, 7,200, and 900 Hz, and
apparent resistivities were calculated for each of these fre
quencies. Details of the data-acquisition system and the
computation of apparent resistivity are described in Pierce
and Hoover (1991).
The alluvial cover in the Getchell-trend area is typi
cally fairly conductive, so that the depth of exploration
using 900 Hz in those areas ranges from about 10 m (33
ft) to 100 m (330 ft). In areas of exposed rock, exploration
depths are generally greater and, for most lithologies,
range from about 100 m (330 ft) to 1,000 m (3,300 ft).
Pierce and Hoover (1991) discuss the apparent resistivity
results for the Getchell-trend area.
BURIED-CONDUCTOR
ENHANCEMENT
As shown by Pierce and Hoover (1991), the loga
rithm of the ratio between apparent resistivities calculated
for 7,200-Hz and 900-Hz frequencies, respectively, gives
information about how resistivity changes with depth.
Roughly speaking, where the ratio is greater than one (the
logarithm is positive), resisitivity is decreasing with depth.
Where it is less than one (the logarithm is negative), resis
tivity is increasing with depth. For the buried-conductor
enhancement, this ratio is then multiplied by the logarithm
of the 900-Hz apparent conductivity (the inverse of appar
ent resisitivity) after adding a constant to ensure only posi
tive logarithms for the conductivity. The conductivity for
the 900-Hz frequency is used because it represents the
deepest looking frequency. In equation form, the results of
the enhancement (E) can be written for each grid point as:
1 p
F -es (...)") (...)
900
log
900
(1)
where
p000 is the 900-Hz apparent resistivity (ohm-meters),
p7,200 is the 7,200-Hz apparent resistivity (ohm
meters), and
k is a constant that ensures the left-hand side of the
product will be positive for all grid points.
Note that k is usually chosen by inspection of the
minimum value of log [...] or all grid points.
Because the negative values of E represent the unin
teresting case where the resisitivities are increasing with
depth, these values are neglected. Thus, the final plot of E
has high values where the product of the 900-Hz apparent
conductivity and the amount of decrease in resistivity with
depth is great. The highest E values represent places
where there is a large, probably abrupt, decrease in resis
tivity at depth. Moreover, the change in resistivity must
occur at fairly shallow depths because the depth of explo
ration is diminished by the presence of the low resistivities
in the subsurface.
RESULTS FOR THE
GETCHELL-TREND AREA
The result of applying the above buried-conductor
enhancement to apparent resisitivity data for the Getchell
trend area is shown on figure 3. The highlighted features
of the enhanced map are present on both the 900-Hz
apparent resistivity and ratio maps of Pierce and Hoover
(1991) as expected from equation (1), but there are many
fewer targets on the enhanced map. The smaller number
greatly facilitates follow-up investigations.
NATURE OF THE CONDUCTIVE MATERIAL
We surmise that the broad regions of slightly higher
values on figure 3 are due to buried lithologies that are
less resistive than the overlying lithology, whereas the
high enhanced values that occur as spots, lines, or string
ers are due to local concentrations of buried conductive
material. In the case of the spot of high enhanced values
near lat 4102'N., long 11717'W., the conductive material
likely is a buried warm spring. In most other cases, we
deduce that graphite primarily comprises the conductive
material from the following observations: (1) Most of the
highest enhanced spots where subsurface information is
available correlate with rocks that contain large concentra
tions of carbonaceous matter, moderate to large amounts
of clay or other argillaceous rock, and small amounts of
Figure 3 (facing page). Results of the buried-conductor en
hancement applied to apparent resistivity data from an airborne sur
vey over the Getchell-trend area. The dimensionless values
increase as a function of the product of the decrease in resistivity
with depth and the 900-Hz apparent conductivity (the inverse of re
sistivity). The enhancement is presented in color shaded relief, with
false illumination from the east. Negative values represent the un
interesting case where resistivity increases with depth. Broad re
gions of slightly higher values are probably due to buried
lithologies that are less resistive than the overlying lithology.
whereas the high values that occur as spots, lines, or stringers are
most likely caused by concentrations of buried conductive material
that is composed partially or wholly of graphitized carbonaceous
matter. Geologic contacts, pit outlines, and occurrence locations are
from figure 2. Faults are not shown for clarity. Refer to figure 2 for
nameS.
BURIED CONDUCTIVE MATERIAL, GETCHELL TREND, NEVADA 241

11722'30
4115' +

| | | | | | | | I_I_
+
-

0 1 2 3 4 5 KM
1
242 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
sulfides; (2) occurrences of argillaceous rock without car
bonaceous matter do not correspond to low resistivities on
the apparent resistivity maps; (3) the amount of sulfides is
not sufficient to produce the low resistivities measured;
and (4) graphite can lower resistivies of rocks quite dra
matically, even when present in small quantities (Nelson
and others, 1982; Grant and West, 1965). Moreover, labo
ratory analyses of carbonaceous matter sampled from sedi
mentary-hosted disseminated gold deposits elsewhere in
northern Nevada indicate cryptocrystalline graphite (Lev
enthal and Hofstra, 1990) with an anthracite or higher coal
rank (Leventhal and Hofstra, 1990; Hausen and Park,
1986). Both anthracite coal and graphite are excellent
electrical conductors. Moreover, if the cryptocrystalline
graphite is localized along bedding or fault planes, as is
carbonaceous matter at the Carlin deposit (Hausen and
Park, 1986), the interconnectivity of the crystals would
further increase conduction.
From these observations we infer that many, if not
most, of the moderate to high values on the enhanced map
reflect carbonaceous matter that is at least partially graphi
tized. Thus, the enhanced map can help determine the
location and nature of buried carbonaceous matter in the
Getchell-trend area.
Where known, the areas of highest enhanced values
have no associated gold mineralization, although they
commonly occur near or on line with precious-metal
deposits (the two known silver deposits included). The
high values to the west of the Osgood Mountains pluton
are exceptions: no known gold mineralization occurs
nearby. On the other hand, all but two of the gold occur
rences that were detectable by the EM survey correspond
to moderately enhanced values, which generally occur as
lines or stringers. The two exceptions are at the central
and south pits of the Getchell deposit, where the lack of
enhancement may be explained by resisitivities that
increase rather than decrease with depth (Pierce and
Hoover, 1991). Here, the deposit and its associated con
ductive material are entirely exposed at the surface. The
gold mineralization in section 13, the Chimney Creek
south pit, and the Rabbit Creek deposit are deeper than the
EM depths of exploration. The inadequate depth of explo
ration near the Rabbit Creek deposit implies that the large
enhanced area to the south must reflect widespread, buried
conductive material within the overburden, which may or
may not be related to the deposit below (Pierce and
Hoover, 1991).
STRUCTURE
Many faults are manifested on the enhanced map (fig.
3), either as lines or stringers, or as abrupt changes in
value. For example, the northeasterly trending enhanced
line about 3 km west of the central Chimney Creek pit
coincides with part of a mapped thrust fault (Willden,
1964); carbonaceous rocks and some sulfide mineraliza
tion are present at depth (B. Berger, written commun.,
1977). At the Pinson and Mag deposits, linear enhanced
areas trend north and northeast, corresponding to Tertiary
structure (E. Kretschmer, written commun., 1991). The
north-south-trending linear enhanced area east of the Pre
ble deposit is associated with the west edge of a graben
like feature that is interpreted from geophysical data
(Hoover and others, 1991; Grauch and others, 1991) and
corraborated by drilling (E. Kretschmer, oral commun.,
1990). To the east, a more subtle linear area of low values
on the enhanced map that trends northeast may represent
the east edge of the graben. A northeasterly trend is sug
gested by connecting linear enhanced areas to the north
east and southwest of the Preble deposit. The trend may
be a major fault related to the northerly trending fault
mapped in the pit (Kretschmer, 1984).
Faults are also expressed by abrupt changes in char
acter on the enhanced map, such as the two northeast
trending faults along the southeastern and northeastern
edges of the study area. These faults are corraborated by
other geophysical evidence (Hoover and others, 1991).
CONCLUSIONS
The buried-conductor enhancement applied to appar
ent resistivity maps can be a valuable tool for gold explo
ration in the Getchell-trend area. The high enhanced
values are most likely explained by subsurface concentra
tions of graphite. Precious-metal mineralization may occur
adjacent to areas that have the highest enhanced values,
but it does not occur directly over these areas. Mineralized
areas correspond better to low and moderately enhanced
values that commonly follow faults. Possible explanations
for these observations are: (1) Carbonaceous matter is
moderately but not highly conductive in association with
gold mineralization, implying that highly conductive car
bonaceous matter and gold mineralization are mutually
exclusive for a given area; (2) conductive cabonaceous
matter is present in association with gold mineralization,
but rocks overlying mineralized areas are less resistive
than in nonmineralized areas; or (3) conductive carbon
aceous matter is present in association with gold mineral
ization, but the resistive cover has been stripped off during
mining. The last possibility does not explain the moder
ately enhanced values that occur near mines in areas
where the cover has not been stripped off. The remaining
two explanations both suggest that the resistivities of the
rocks are somehow related to gold mineralization. Perhaps
the gold mineralization event destroyed some of the con
ductive properties of the carbonaceous matter or increased
the porosity of the overlying rocks. In situ electrical mea
surements near the Pinson mine suggest that carbonaceous
 BURIED CONDUCTIVE MATERIAL, GETCHELL TREND, NEVADA
ore exhibits higher resistivities compared to carbonaceous
non-ore (Hoover and others, 1987). This observation also
suggests an interrelation between the conductive properties
of carbonaceous matter and associated gold mineralization.
We predict that abundant carbonaceous material (par
tially or wholly graphitized) will be found at depth at the
spots of highly enhanced values under the basalt south of
the Silver Coin mine (fig. 2); under Paleozoic sedimentary
rocks just to the northeast of Turquoise Ridge mineraliza
tion and near lat 417'30"N., long 11722'30"W.; and
under alluvium near the northern tip of the study area. In
addition, faul ts containing conductive material that may
have controlled hydrothermal fluid flow in the past occur
along the southeastern edge of basalt northeast of the Sil
ver Coin mine, along the contact between Preble Forma
tion and alluvium 2 km west of Lone Butte, under
alluvium about 1 and 2 km southeast of the Pinson mine,
and possibly within the Paleozoic rocks west of the sec
tion 13 mineralization. Even if no gold is found in these
areas, the buried-conductor enhancement does well in iso
lating conductive fault zones that may be related to miner
alization. This technique may become extremely useful in
studying the role of carbonaceous matter in sedimentary
hosted, disseminated gold deposits.
REFERENCES CITED
Bagby, W.C. and Berger, B.R., 1985, Geologic characteristics of
sediment-hosted disseminated precious-metal deposits in the
Western United States, in Berger, B.R., and Bethke, P.M., eds.,
Reviews in Economic Geology, Volume 2Geology and
Geochemistry of Epithermal Systems: El Paso, Society of Eco
nomic Geologists, p. 169232.
Bloomstein, E.I., Massingill, G.L., Parratt, R.L., and Peltonen,
D.R., 1991, Discovery, geology, and mineralization of the
Rabbit Creek gold deposit, Humboldt County, Nevada, in
Raines, G.L., Lisle, R.E., Schafer, R.W., and Wilkinson,
W.H., eds., Geology and Ore Deposits of the Great Basin,
Symposium Proceedings: Reno, Geological Society of Neva
da, p. 821845.
Broili, C., French, G. McN., Saddrick, D.R., and Weaver, R.R.,
1988, Geology and gold mineralization of the Gold Bar depos
it, Eureka County, Nevada, in Schafer, R.W., Cooper, J.J., and
Vikre, P.G., eds., Bulk Mineable Precious Metal Deposits of
the Western United States Symposium Proceedings: Reno,
Geological Society of Nevada, p. 5772.
Erickson, R.L., and Marsh, S.P., 1974, Geologic map of the Gol
conda quadrangle, Humboldt County, Nevada: U.S. Geologi
cal Survey Geologic Quadrangle Map GQ1174, scale
1:24,000.
Foster, J.M., and Kretschmer, E.L., 1991, Geology of the Mag de
posit, Pinson mine, Humboldt County, Nevada, in Raines,
G.L., Lisle, R.E., Schafer, R.W., and Wilkinson, W.H., eds.,
Geology and Ore Deposits of the Great Basin, Symposium Pro
ceedings: Reno, Geological Society of Nevada, p. 845856.
243
Grant, F.S., and West, G.F., 1965, Interpretation theory in applied
geophysics: New York, McGraw-Hill, 584 p.
Grauch, V.J.S., Hoover, D.B., and Wojniak, W.S., 1991, Subsur
face structure and lithology near the Getchell gold trend, Os
good Mountains, NevadaGeophysical insights: U.S.
Geological Survey Circular 1062, 34 p.
Hausen, D.M. and Park, W.C., 1986, Observations on the associa
tion of gold mineralization with organic matter in Carlin-type
ores, in Dean, W.C., ed., Proceedings of the Denver Region
Exploration Geologists Society Symposium on Organics and
Ore Deposits: Denver Region Exploration Geologists Society,
p. 119136.
Hoover, D.B., Grauch, V.J.S., Pitkin, J.A., Krohn, M.D., and
Pierce, H.A., 1991, An integrated airborne geophysical study
along the Getchell trend of gold deposits, north-central Neva
da, in Raines, G.L., Lisle, R.E., Schafer, R.W., and Wilkinson,
W.H., eds., Geology and Ore Deposits of the Great Basin,
Symposium Proceedings: Reno, Geological Society of Neva
da, p. 739758.
Hoover, D.B., Smith, B.D., Grauch, V.J.S., and Podwysocki, M.H.,
1987, Geophysical studies in the vicinity of the Getchell fault
system, Humboldt County, Nevada: U.S. Geological Survey
Circular 995, p. 3132.
Hotz, P.E., and Willden, Ronald, 1964, Geology and mineral de
posits of the Osgood Mountains quadrangle, Humboldt Coun
ty, Nevada: U.S. Geological Survey Professional Paper 431,
128 p.
Keller, G.V., and Frischknect, F.C., 1966, Electrical methods in
geophysical prospecting: New York, Pergamon Press, 523 p.
Kretschmer, E.L., 1984, Geology of the Pinson and Preble gold de
posits, Humboldt County, Nevada: Arizona Geological Soci
ety Digest, v. 15, p. 5966.
Leventhal, Joel, and Hofstra, Albert, 1990, Characterization of car
bon in sediment-hosted disseminated gold deposits, north cen
tral Nevada, in Hausen, D.M., Petersen, E.U., and Tafuri, W.J.,
eds., Gold '90, Symposium Proceedings: Littleton, Colorado,
Society for Mining, Metallurgy, and Exploration, p. 365368.
Leventhal, J.S., Hofstra, A.H., Vuletich, A.K., and Mancuso, T.B.,
1987, Sediment-hosted disseminated gold mineralization at
Jerritt Canyon, Nevada. IIIRole of organic carbon [abs.]:
Geological Society of America, Abstracts with Programs, v.
19, no. 7, p. 745.
Nelson, P.H., Hansen, W.H., and Sweeney, M.J., 1982, Induced
polarization response of zeolitic conglomerate and carbon
aceous siltstone: Geophysics, v. 47, no. 1, p. 7188.
Osterberg, M.W., and Guilbert, J.M., 1991, Geology, wall-rock
alteration, and new exploration techniques at the Chimney
Creek sediment-hosted gold deposit, Humboldt County, Ne
vada, in Raines, G.L., Lisle, R.E., Schafer, R.W., and
Wilkinson, W.H., e ds., Geology and Ore Deposits of the
Great Basin, Symposium Proceedings: Reno, Geological So
ciety of Nevada, p. 805820.
Palacky, G.J., 1986, Geological background to resistivity mapping:
Geological Survey of Canada Paper 8622, p. 1927.
Percival, T.J., Bagby, W.C., and Radtke, A.S., 1988, Physical and
chemical features of precous metal deposits hosted by sedi
mentary rocks in the Western United States, in Schafer, R.W.,
Cooper, J.J., and Vikre, P.G., eds., Bulk Mineable Precious
244 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Metal Deposits of the Western United States Symposium Pro
ceedings: Reno, Geological Society of Nevada, p. 1134.
Pierce, H.A., and Hoover, D.B., 1991, Airborne electromagnetic
applicationsMapping structure and electrical boundaries be
neath cover along the Getchell trend, Nevada, in Raines, G.L.,
Lisle, R.E., Schafer, R.W., and Wilkinson, W.H., eds., Geolo
gy and Ore Deposits of the Great Basin, Symposium Proceed
ings: Reno, Geological Society of Nevada, p. 771780.
Silberman, M.L., Berger, B.R., and Koski, R.A., 1974, K-Ar age
relations of granodiorite emplacement and tungsten and
gold mineralization near the Getchell mine, Humboldt Coun
ty, Nevada: Economic Geology, v. 69, p. 646656.
Willden, Ronald, 1964, Geology and mineral deposits of Hum
boldt County, Nevada: Nevada Bureau of Mines and Geolo
gy Bulletin 59, 154 p.
 CHAPTER Y
GOLD IN DEVONIAN CARBONATE ROCKSAT
CEDAR PEAK, SOUTHERN SNAKE MOUNTAINS,
NORTHEASTERN NEWADA
By CHARLES H. THORMAN" and WILLIAM E. BROOKS!
ABSTRACT
A bedrock geochemical sampling program of the
Oxley Peak and adjacent quadrangles shows anomalous
gold concentrations in the Cedar Peak area, located 8 miles
north of Wells, Nevada. Chemical analyses were obtained
for 54 samples of unaltered bedrock and silicified and unsi
licified fault zones throughout the study area and from
zebra-banded Devonian dolomite and interbedded lime
stone at Cedar Peak. The Devonian zebra-banded dolomite
and interbedded limestone occur within a west-northwest
trending structural block (Cedar Peak block). The Cedar
Peak block is bounded on the north and south by west
northwest-trending high-angle faults and on the west by a
west-dipping normal fault. Sixteen samples of dolomite and
limestone contained from less than 0.002 to 0.150 parts per
million gold; seven samples did not have detectable gold.
Gold was not detected in the 31 samples collected from
other locations in the area, including breccia from faults
bounding the Cedar Peak block.
INTRODUCTION
Geologic mapping and geochemical sampling of fault
zones and bedrock in the southern Snake Mountains were
undertaken to study the history of deformation and mineral
ization of the range and to try to determine whether or not
various deformational events and structures had different,
and possibly characteristic, mineralization signatures and
ages. Anomalous geochemical values, especially for gold,
were found in the Cedar Peak study area in the Oxley Peak
quadrangle (fig. 1), in the southern Snake Mountains. Two
"U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
samples (88T04 and 88T06) were collected in 1988 from
the Simonson Dolomite and Guilmette Formation during
field mapping because of the unusual nature of the rocks.
The presence of detectable gold prompted additional sam
pling of the section at Cedar Peak. The presence of gold
and other metals in Devonian zebra-banded dolomites indi
cates that mineralizing fluids passed through these rocks,
possibly several times.
GEOLOGIC SETTING
Northeastern Nevada is in the northern Basin and
Range province and is underlain by rocks ranging in age
from Archean orthogneiss to Holocene volcanic and basin
fill deposits. The stratigraphic column is comprised of three
major sequences, separated by hiatuses representing the
main orogenic events of the region. The oldest event was
rifting of the Precambrian sialic(?) crystalline basement.
Upon the rifted continental margin a second sequence was
deposited, comprised of Proterozoic to Middle Jurassic
strata. A major subdivision of the pre-Mississippian rocks
in Nevada results from the recognition of two distinct sedi
mentary successions: an eastern or carbonate suite of mio
geoclinal rocks deposited adjacent to the North American
craton and a western or siliciclastic suite of eugeoclinal
rocks deposited farther to the west. The transition from the
eastern carbonate facies to the western siliciclastic facies
rocks had a northerly trend and was located in central
Nevada.
In western and central Nevada, Several middle Paleo
zoic to early Mesozoic orogenic events have been recog
nized: the Late Devonian to Early Pennsylvanian Antler,
Late Pennsylvanian Humboldt, and Late Permian to Early
Triassic Sonoma orogenies. These events are mainly
recorded in northeastern Nevada in the stratigraphic record
rather than by structures. From Late Jurassic to early
ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

1150 114052'30"
41022'30" I ;: :

SUMMER CAMP
Plate 4 QUADRANGLE

Plate 5

41015' H -

OXLEY PEAK
QUADRANGLE

Plate 5

O 1 2 3 4
L- I l l I

MILES

41007'30' H -
O Wells

Figure 1. Map showing location of Cedar Peak study area (black rectangle) in the southern Snake
Mountains and the principal structural plates (northern half of map modified after Smith and others,
1990).
 GOLD IN CARBONATE ROCKS, CEDAR PEAK, NEVADA
Plate 1 Plate 2 Plate 3 Plate 4 Plate 5
Mm Ma
Mtp -*
Dg ri Dw Dsh
D.si Ds.
DSOw
Yrk Sw
Ow
* units not in map area. See Smith and others (1990).
Figure 2. Distribution of pre-Tertiary rocks in thrust plates in the Oxley Peak quadrangle and adjacent quadrangles. Arrow between
units Ma and Dw indicate subordinate thrust within plate 4. See appendix for description of map units.
Tertiary time, the area underwent two major compres
sional orogenic events, the Middle to Late Jurassic Elko
orogeny (Thorman and others, 1991) and the Early Creta
ceous to early Tertiary Sevier orogeny. These two oroge
nies resulted in regional metamorphism of miogeoclinal
rocks, the formation of large-scale southeast- and east
directed structures, some extensional structures, plutonism,
and regional uplift resulting in the shedding of enormous
quantities of detritus into basins to the east. Sporadically
preserved intermontane basinal deposits accumulated dur
ing this time (Vandervoort and Schmitt, 1990). The third
major rock sequence was deposited after the Sevier orog
eny; its age ranges from Eocene to Holocene. Rocks of
this sequence accumulated during an extensional orogenic
period that is still going on.
The southern Snake Mountains are underlain by a
sequence of Paleozoic, Triassic, and Cenozoic sedimentary
and volcanic rocks (Thorman and Brooks, 1988; Thorman
and others, 1991; Smith and others, 1990). Direct and indi
rect evidence indicates that these rocks were deformed dur
ing the Antler, Humboldt, Sonoma, Elko, and Sevier
orogenies and by Oligocene to Holocene basin-and-range
extensional faulting. Compression and extension were ori
ented in the same general directions during these events
and caused the reactivation of many faults, especially west
northwest-trending, high-angle faults. Pre-Tertiary rocks in
the southern Snake Mountains range in age from Ordovi
cian through Triassic and occur in a series of imbricate
thrust plates (Thorman and Brooks, 1988; Thorman and
others, 1991). The strata within the thrust plates strike sub
parallel to the bounding faults, which are subparallel to
each other. Figure 2 shows the various units in these thrust
plates; lithologic descriptions of the units are listed in the
Appendix to this chapter. At Cedar Peak, facies of eastern,
247
Plate 6 Plate 7 Plate 8
Tid | Triassic
Pm Pm Pm | Permian
| Lower Mississippian
| Devonian
H Silurian
Ordovician
carbonate (miogeoclinal) rocks and western, siliciclastic
(eugeoclinal) rocks of Ordovician through Devonian age
are present in separate thrust plates that are bounded by
both bedding-parallel and high-angle faults. The altered
rocks discussed below are part of the eastern facies
sequence. Tertiary rocks range in age from late Eocene
(= 41 Ma) to Miocene and Pliocene and have undergone
extensional deformation only.
STRUCTURES IN THE CEDAR PEAK
BLOCK AREA
The Cedar Peak block (fig. 3) consists of east-dipping
Middle Devonian Simonson Dolomite, Upper Devonian
Guilmette Formation, Lower Mississippian Tripon Pass
Limestone, and the Mississippian sandstone of Melandco
(Thorman and Brooks, 1988). These rocks comprise the
lowest thrust sheet, herein referred to as the Bishop Creek
plate (plate 1 of Smith and others, 1990) in the Oxley Peak
quadranglethese rocks belong to the eastern (miogeocli
nal) facies sequence of the Paleozoic miogeocline. The
Finks Canyon plate (plate 5 of Smith and others, 1990)
tectonically overlies the Bishop Creek plate and is com
prised of Devonian limestone, shale, chert, and argillite,
herein interpreted to be transitional between eastern (mio
geoclinal) and western (eugeoclinal) facies strata. The
Oxley Peak plate tectonically overlies both of these plates
and consists of limestone, dolomite, chert, and siltstone of
the Permian Murdock Mountain Formation.
Two types of faults dominate the structure at and near
Cedar Peak: high-angle, west-northwest-trending faults
(which probably have both lateral and vertical displace
ment) and faults that are subparallel or parallel to bedding
248 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

; its
t - - - - -

114 52' 30" 114 50'

30"

Summer Camp Melandco

quadrangle quadrangle

Bishop Creek
Reservoir spillway
1 \2^ 15 ee 14

41 O +
15 - e16 O 13

O
*** 17

Cedar Peak 21 10
6 of 1

| - _y e3O
Figures 3B & 3C
Oxley Peak e12

quadrangle e 18

e 29 Wells Peak

28 quadrangle
Oxley Peak * 22
- O e e 24
27 23

8 .
25. 26
7

O 1 2 Miles

41
07 - -
|
30"

Figure 3A. Map showing samples sites 1 through 30 (see table 1) and location of areas shown in figures 3B and 3C.
 GOLD IN CARBONATE ROCKS, CEDAR PEAK, NEVADA 249

41 114 o 55'58.6 1" 1 14 55'0.605"

13
42.99
11

Mm
6
- ~ *-
Mtp 21 1O
Mm

/
/
7

Dg

CEDAR PEAK
A \ .

Hso O

34

35
Ds i 65, 36 s \ .N. o ... .

43 - 4.4 N

DS h - -* 2 k 20 Mtp

\ ` - . . . . " 46 __ -/*
. T T ~ .*
4 10 Y - - - - - ~~ Dg
13' ~~!

13.75" 0 500 1000 1500 FEET

1 I l |

Figure 3B. Geologic map of the Cedar Peak block showing location of samples collected from the Simonson Dolomite, Guilmette For
mation, and from fault zones. See tables 1 and 2. Topographic base digitized from Oxley Peak 7"h-minute quadrangle. Location of figure
shown on figure 3A. Dotted lines are contours; heavy solid lines are high-angle faults (dashed where inferred); heavy solid lines with teeth
are low-angle to bedding-parallel faults with teeth on upper plate; light dashed lines are jeep trails. See Appendix for description of map
units Dsi, Dg, Mtp, Mm, and Dsh.

(herein referred to as bedding-parallel faults). The bed
ding-parallel faults are of two types: those that bound
major imbricate thrust sheets (both younger-over-older and
older-over-younger relationships) and subordinate bedding
parallel faults within the thrust sheets that occur at bound
aries between contrasting lithologies. In addition, northerly
trending, down-to-the-west normal faults occur in the west
ern portion of the rangethe largest of these is at the west
ern foot of the range.
The Cedar Peak block, a west-northwest-trending
block about 0.7 miles wide comprised of middle Paleozoic
miogeoclinal units, is bounded by high-angle faults that
transect the Snake Mountains. The high-angle faults cut
Paleozoic, Mesozoic, and Tertiary rocks. At least two peri
ods of movement are interpreted to have taken place along
these faults: one during late Mesozoic and the other during
middle to late Tertiary. We consider the faults to be part of
the Wells fault system, a Late Mesozoic west-northwest
trending right-slip fault system that extends across north
eastern Nevada (Thorman and Ketner, 1979). Initial dis
placement is believed to have occurred in the Mesozoic,
during either the Middle to Late Jurassic Elko or during the
Late Cretaceousearly Tertiary Sevier orogenies. No data
exist that may suggest (1) whether fault movement began
during the Elko orogeny and was subsequently reactivated
during the Sevier orogeny, or (2) whether fault movement
occurred during only one orogeny or the other. At the east
ern foot of the Snake Mountains, the high-angle faults jux
tapose middle Miocene rhyolite flows against sandstone of
Melandco in the Bishop Creek platethis dates some of
the faulting as middle Miocene or younger.
The northern bounding fault has left separation of
about 0.7 miles and juxtaposes rocks of the Bishop Creek
plate against higher plates on the eastern side of the
range. The southern bounding fault has apparent down-to
the-south movement and juxtaposes the structurally
higher Finks Canyon and Oxley Peak plates, on the
south, against the Bishop Creek plate. It is not possible,
at this time, to determine the amount and direction of the
slip along these faults.
The bedding-parallel faults are interpreted to be pre
late-Eocene in age because 39-Ma to 40-Ma ash-flow tuffs
250 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

41 114 55'58.61" 1 14 55'0.605."

13' -

42.99" Mm /

Dg

r CEDAR PEAK
A -

40
O. O.02
N
. .
*...
\
s * 0.150 oo,
<0.002 . -
O. 10 <0.002 D
.0. 100 / " * , 9

DS . N'
, :
e

- --O /
<0.002 /

O. O.O4s. _<0.002 ^
e * D / ^ - -
0.002 0.002 9 ^

''... 2 - N Mt p
D sh * ~ - O ~ T
\ > -- - - N -/*
41
13'o 1 --.
.
- - ~~
9
13.75" O 500 1000 1500 FEET
L | 1 |

Figure 3C. Geologic map of Cedar Peak block showing gold values for sites shown in figure 3B. See Appendix for description of
map units.

overlie these structures on the eastern flank of the range
(Thorman and others, 1991). At Cedar Peak, west-north
west-trending, high-angle faults cut both types of bedding
parallel faults (see fig. 3).
STRATIGRAPHY OF GOLD-BEARING
ROCKS AT CEDAR PEAK
The Simonson Dolomite in this part of northeastern
Nevada typically is comprised of dolomite, locally with
subordinate limestone. In contrast, the lower Guilmette is
comprised of a lower unit of argillaceous, platy limestone
approximately 100 ft thick that is overlain by several hun
dred feet of thin- to medium-bedded limestone with vari
able amounts of dolomite beds (generally less than 50
percent). Both the dolomite and limestone beds are charac
terized by laminated, medium-grained, medium- to dark
gray rocks (fig. 4). At Cedar Peak, the upper 400 ft of the
Simonson Dolomite contains approximately 4080 percent
zebra-banded dolomite of an irregular, stratabound nature
and forms a light-colored band on the hillsidethis is in
marked contrast to the otherwise dark-colored rocks in this
portion of the stratigraphic section (fig. 5).
Two lithologies form the zebra-banded rock: a light
colored, bleached-looking, coarse-grained, massive dolo
mite and a darker, coarse-grained, weakly laminated dolo
mite (fig. 6). The darker rock is similar to the typical
laminated Simonson Dolomite lithology but is much
coarser grained. Discontinuous, flat, tabular fragments of
the darker rock occur in a matrix of the light-colored dolo
mite, with the amount of light-colored dolomite ranging
from zero to more than 95 percent. Some of the zebra
banded rock structure has the appearance of crossbedding
(fig. 6); some has the appearance of edgewise conglomer
ate; and, in some beds, the dark fragments are randomly
oriented. In the rocks that appear to be crossbedded or
edgewise conglomerates, the light-colored dolomite crys
tals typically are oriented normal to the edges of the dark
fragments, forming comb structure, commonly with open
spaces. The upper Simonson Dolomite contains numerous
irregular-shaped bodies of zebra-banded rock that are
grossly stratabound, but in detail they cut across bedding
and die out laterally. Small-scale collapse features within
coarse-grained veinlets cut across bedding (fig. 7).
 GOLD IN CARBONATE ROCKS, CEDAR PEAK, NEVADA
Figure 4. Typical exposure of weathered, laminated beds in the
Simonson Dolomite.
Figure 5. Looking northeast at Cedar Peak. Dashed line is con
tact between Simonson Dolomite (Dsi) and Guilmette Formation
(Dg). Grassy slope with Dg label is basal unit of Guilmette. Light
colored beds below contact are zebra-banded dolomite beds in up
per Simonson Dolomite.
The origin of zebra-banded dolomite, or zebra rock,
has been discussed by many workers. Two different pro
cesses considered to be essential to the formation of the
basic zebra structure are: the development of stromatactis
structure synchronous with carbonate sedimentation during
which episodic submarine lithification of the carbonate
muds, which forms a series of crusts between which cavi
ties (stripes) are eroded, cemented, and filled with rhombs,
occurs near the sediment-seawater interface (Landis and
Tschauder, 1990, p. 340); or the development of zebra
banded dolomite is best developed adjacent to karst chan
nelways, and the zebra spar precipitated into voids created
by karst dissolution (Horton and De Voto, 1990, p. 86).
Both models describe early-stage, low-temperature, light
colored (zebra spar) dolomite overgrown by late-stage
251
Figure 6. Zebra-banded dolomite beds in upper Simonson Dolo
mite. Note cross bedded appearance and variation in amount of
light and dark lithology.
Figure 7. Possible collapse structures in zebra-banded dolomite
filled with light-colored dolomite matrix.
dolomite of a higher temperature that is associated with ore
deposition. We have not studied the small-scale features of
the Cedar Peak zebra rock with regard to the origin of the
zebra structure or with regard to how many generations of
dolomite may be present.
NATURE OF MINERALIZED ROCKS
During mapping of the Oxley Peak quadrangle and
adjacent areas, samples were routinely collected for
geochemical analyses from altered rocks and from several
black-shale-bearing western-facies units to determine their
metal content. Samples were also collected from fault
252 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 1. Samples collected in 1988 from the Oxley Peak, northwestern Wells Peak, south
western Melandco, and southeastern Summer Camp quadrangles, northeast Nevada.
[See figure 3A for sample locations. Most of the samples were collected from fault zones; the other samples were from
undeformed rocks to determine background geochemical characteristics. Abbreviations for formation names are presented
in the Appendix; lower limits of detection for gold and mercury are 0.05 ppm and 0.02 ppm, respectively; n, not detected
at limit of determination; H, greater than 36 ppm, V, high-angle fault, B, bedding-parallel fault, E-W, east-west-trending
fault; N-S, north-south-trending fault; X, feature observed in sample, -, feature not observed in sample; /, separates symbols
for map units in fault contact; ?, indicates uncertainty regarding nature of sample]

Map Field Formation Gold Mercury Fault Brecciated Silicified Vein Iron
number number nannes (ppm) (ppm) stained
l 88T()] Mtp n 0.18 V-B X - - X
2 88T02 Mtp n 0.42 V(E-W) X - - X
3 88T03 Mtp n 0.68 V(E-W) X - - X
4 88T04 Dg 0.05 H - - - X X
5 88T06 Ds. 0.10 n Rock samplezebra-banded dolomite

6 88T07 Dg'Mm n () 02 V X X X
7 88E07 Ttm n 0.18 V(N-S) X X - X
8 88E08 Ttm n n V(N-S) X X - X
9 88T10 Trau n n - - X - -

10 88E12 Mtp/Mm n 0 24 V(E-W) X X - X

|1 88E13 Mm n () ()8 V(E-W) X X - X

12 88T 14 Prm/Ds] n 0.02 B X X X
13 88T1 7 DOw n 0 18 Rock sampleblack shale
14 88T18 Sw n 0.28 Rock sampleblack shale
15 88T19 Ow n 0.02 Rock sample basaltic amygdaloidal pillow lava

16 88T20 DOw n 0.20 Rock samplesiltstone

17 88T2] Msh n 0.06 Rock sampleblack shale
18 88T26a Dsh n 0.34 Rock sampleblack argillite
18 88T26b Dsh n 0.22 Rock sampleblack argillite
19 88T27a Dg/Mtp n 0.04 B o X - X

19 88T27b Dg/Mtp n () 12 B o X - X
20 88T28 Dg/Mtp n 0.12 B o X X
2] 88T29 Dg/Mlp n 0.04 V(E-W) X X X
22 88T30 Prm/Mm n 0 32 B X X X
23 88T3 . Prm/Dsh n n B X o -

24 88T32 Prm n 0 16 B o X - X
24 88T33 Pm n 0.04 B X - X
25 88T34 Ttm n n B(N-S) X X - X
26 88T35 Ttm n () 32 B(N-S) X X X
27 SST52 Dsh n 0.40 o - - - X

28 88T57 Pm n 0.80 B X X - X
29 88T58 Dsh n 0.40 Rock sample dark siltstone with black chert and
siliceous nodules
30 88T80 Trp n n V(E-W) X o - -

zones and veins of various types. Figure 3A shows the loca
tions of samples collected in 1988. Figure 3B shows the
locations of samples collected in 1989 and figure 3C shows
the gold content detected from each sample in 1989. Tables
1 and 2 list the 54 samples collected in 1988 and 1989,
respectively, and indicate formation name, gold and mer
cury content, structural setting, and nature of each sample
(breccia, silicified, iron stained, high-angle or bedding-par
allel fault). Tables 3, 4, and 5 list the samples according to
the formation or structure sampled; table 3 lists samples
from the Simonson Dolomite and Guilmette Formation;
table 4 lists samples from fault zones; and table 5 lists
samples collected for geochemical background data from
black shales. Table 6 lists the latitude and longitude of the
samples shown on figure 3A and listed in table 1.
The rocks of primary concern in this report are zebra
banded dolomites in the upper 440 ft of the Simonson
Dolomite and lower 300 ft of the Guilmette Formation
because they are the only rocks that contained detectable
gold. However, various fault zones contain elevated con
centrations of some metals, suggesting that some of the
mineralizing fluids that deposited the gold may have passed
through the fault zones.
Gold values in the zebra-banded rocks range from
0.02 to 0.150 ppm (parts per million); in some cases gold
was detected but below the limit of detection (tables 1, 2, 3,
and 4). Twenty-three samples were collected from these
formations (tables 1 and 2), of which 14 were from the
Simonson, 8 were from the Guilmette, and 1 was from an
east-west-trending fault zone along which the Simonson
 GOLD IN CARBONATE ROCKS, CEDAR PEAK, NEVADA

Table 2. Samples collected from the Simonson Dolomite (Dsi) and Guilmette Formation (Dg) at Cedar Peak in 1989
in the Oxley Peak quadrangle, northeast Nevada.
[See figure 3A for sample locations. Lower limits of detection for gold and mercury are 0.002 ppm and 0.02 ppm, respectively; n, not detected at limit of
determination; c, detected but below limit of determination shown; /, separates map units in fault contact)

Map Field Formation Gold Mercury Sample description

number number nannes (ppm) (ppm)
31 89T24 Dg 0.002 il Fine-grained, zebra-banded dolomite.
31 89T25 Dg 0.150 n Coarse-grained, zebra-banded dolomite.
32 89T26 Dg 0.002 0.04 Atrypa-bearing, thin-bedded, argillaceous limestone.

33 89T27 Dsi 0.002 n Medium-grained, mottled dolomite.

33 89T28 Dsi n n Medium- to coarse-grained, zebra-banded dolomite.
33 89T29 Dsi n ().04 Dolomite breccia.

34 89T30 Dsi <0.002 in Medium-grained, zebra-banded dolomite.

34 89T31 Dsi in in Medium-grained, zebra-banded dolomite.
35 89T32 Dsi 0.100 0.36 Medium-grained dolomite with intraformational breccia,
some thin dolomite veins.

35 89T33 Dsi In 0.04 Medium-grained dolomite with intraformational breccia.

36 89T34 Dsi n n Medium- to coarse-grained, zebra-banded dolomite.
37 89T35 DSi <0.002 In Medium- to coarse-grained, zebra-banded dolomite.

38 89T36 Dsi 0.002 0.20 Medium- to coarse-grained dolomite with fossil hash.
39 89T37 Dsi <0.002 0.04 Medium- to coarse-grained, zebra-banded dolomite.
40 89T38 DSi n n Medium- to coarse-grained, zebra-banded dolomite.

41 89T39 Dsi <0.002 in Medium-grained, mottled dolomite.

42 89T40 Dsi/Dg 0.004 0.16 Silicified, brecciated, and iron-stained carbonate rock
along E.-W. fault.

43 89T41 Dg 0.002 ().08 Brecciated and iron-stained limestone and dolomite.

44 89T.42 Dg 0.002 0.08 Brecciated and iron-stained dolomite with calcite veinlets.
45 89T43 Dg <0.002 (). 16 Brecciated and iron-stained limestone and dolomite with
calcite veinlets.

46 89T.44 Dg in in Brecciated and iron-stained limestone with calcite

veinlets.

and Guilmette are juxtaposed. Eight samples from the
Simonson, seven from the Guilmette, and the fault-zone
sample contained detectable gold. We can see no difference
in hand specimen between the gold-bearing and barren
samples.
Distribution of several other metals in the Simonson
and Guilmette show the following relationships: mercury
generally occurs with gold but was not detected in two of
the four samples with the highest gold values (88T06 and
88T25); antimony, arsenic, cadmium, copper, molybde
num, and zinc have distribution patterns similar to that of
mercury. Silver was detected in only one sample at Cedar
Peak (89T26), which also had anomalous amounts of zinc,
antimony, gold, copper, molybdenum, and mercury.
In general, the following elements have higher con
centrations in the faults than in the Simonson and Guil
mette: zinc, mercury, lead, molybdenum, copper, antimony,
cadmium, arsenic, zirconium, yttrium, vanadium, nickel,
gallium, chromium, cobalt, beryllium, barium, boron, sil
ver, titanium, phosphorous, and iron.
OBSERVATIONS
Conclusions regarding the zebra-banded dolomites
and the alteration and distribution of gold and other metals
are:
1. Gold appears to have been deposited primarily in the
Cedar Peak block;
2. The zebra-banded beds occur primarily in the Cedar
Peak block; Simonson Dolomite and Guilmette
;
[eAtnearmliyntaitcioanl,
method
analyzed,
element
show
headings
column
The
percent.
which
Ti,
P,
Fe,Mg,Na,
Ca,
for
except
million,
per
parts
in
given
are
dresults
of
limit
lower
the
and rebtserpomrvciptnaitvieolny;s
analyses,
furnace
graphite
aag,
vapor
cold
aac,
analyses;
atomic
aaaf,
Other
analysis.
dflame
of
limit
lower
the
at
detected
not
n,
are:
detected,
*,
pena0tdr-ceutrcnmlotigenhrmavetpsihnleotysnc
d(shown
of
Methods
percent).
are
which
Ti,
and
P,
Mg,Na,
Fe,
Ca,
for
except
million,
per
parts
sin
emission
s,
1are:
5-
i-10,
or
i-5
ianalyses;
plasma
coupled Mg
Sample
Cd
Bi
Mn
LaAg
TMo
P
Na
Fe
Ca
BAs
B
Be
Ge
Ga
Cu
Cr
Co
Au
ia
number
s (5)(50)
(10)
(1)
(20)
(200)
(0.5)
(0.002)
(0.2)
(0.02)
(0.05) 5<
n5
2n
1089T24

(),002
<20
i.05
00
()0n In
150 ()0n n
().002
5
<
2n
>
189T25

0().05
|
20
I.2
in 57
150
&
5
< n
I0n 3n
().7
20
89T26
0.05
200
00
in
n 150
&2.002
>
1(89T27

(()5
0<
|
00
).2
i()
n0)n 70
in 5
<
20
(89T28
|()
).2
Il n
n 25
2n
-
(89T29
<
1()
()7
|
3()
15()
&).2
1()
000 n
5
2< .20n n
2<2()
189T3()
00>
&
l
.002
.()
i00
nn 1<20
2n
>
||
89T3
0
<
|
i.05
00
()0n n
.2
n 51()
2n
>
189T32

15
().
0.005
30
00
i0n n
n 5()
2n
> 0 1n
089T33
1<

()()3
.05
00
in 2in

89T34
0>
0.003
1
<
<20
i.()
n00
.20n 1n
220
089T35
<
-
()()3
()
.05
i50
n()
0n n
in ()0 3n
13()
().07
5
<
2n

89T36
7
0.02
&00 ().05
5
<
2in
-
(89T37
1()
0.002
1.5
200
).2
000 3n
n 5
<
2n
1089T38
(
0.003
l
100
00
.05
).2
n0 n 31()
2n
- ()0ln n
189T.39

().05
().005
20iI00
n |
<
5,000
10
25
7
|
20
89T40
0.02
00
In <
n
in
n 5
1300
n(),005
2).05
>
||
(89T4
().5
i50
n 0n In
150
2200
>
89T.42
().5
().07
0.005
5
<i0n n
n n,000
20.005
20
89T43
7
5
<
1
().
i00
nnn 50
1().05
&
n
5
<2150

89T44
().7
().01
1n
n0

n2,000
2().2
15
50>().05
5()
3 n
3,000
33()n10
88T()4
2000
0 <3().002
n
520
88T()6
().2
1n
1
().00
()n
iIl

quadrangle,
Peak
Oxley
in
Formation
Guilmette
and
Dolomite
Simonson
Devonian
the
from
collected
samples
of
Analyses
3.
Table
Nevada.
northeast

of
limit
lower
below
dbut
>,
shown]
value
the
than
greater
be
to
eteremrinmaitnioend;

for
except
million,
per
parts
in
given
are
dresults
of
limit
lower
the
and
(shown dpenatdrceutrcnmotignihravetpsihleoysnc *,
[eAtnearmliyntaitcioanl, coupled
method
analyzed,
element
show
headings
column
The
percent.
which
Ti,
P,
Mg,Na,
Fe,
Ca, of
Methods
percent).
are
which
Ti,
and
P,
Fe,Mg,Na,
Ca,
for
except
million,
per
parts
sin
emission
s,
iare:
10-element
5-
i-10,
or
i-5
analyses;
respec
analyses,
plasma vapor
cold
aac,
analyses;
furnace
graphite
aag,
atively;
Other
analysis.
flame
absorption
atomic
daaf,
of
limit
at
detected
not
n,
are:
lower
the
below
but
detected,
ebtseroemrvipnatiton;s
Ni
Nb
Sample
Hg
Ag
Th
Sr
W
Zn
Sb
Cd
Bi
Pb
Mo
Cu
Snn i-10
Sc
ZV
Y
As
Zr
Au inumber
s -10
aag
s
aac (5)
(20)
(0.045)
(0.03)
S (0.09)
(0.002)
(0.02)
(0.06)
(0.6)
(200)
(100)
(10)
().70
0.85
0.22
0.24
n
89T28
In
in
I
n n
l
1
<
89T29
n
m
15
1.0
0.04
6.6
0.46
2.7
0.96
i0
In
n n 125
<
l
089T3
n
.030
0.74
0.20
.26
I00
n0n
in
In 10.04
<
89T33
n
51
7.2
0.23
0.85
0n
.11
i0n 0.29
10.60
<
89T34
n
n.
0.82
0.17
0.80
0n
in
i
n 0<
5n
m
0.97
1.7
1.043
89T38
5n.1
0 1n
2.8
0.26
1.2
<
(1289T44
0.60
),049
n.3
000
5
<
n
89T37
1.6
0.094
01.3

0.04
1.4
1.002
i0
n n 5
<
n
89T40
10
1
(&
40
4.8
07.6

0.16
4.5
3.0
3.2
),067
.004
i0
in
n n 01.9

0.16
1<
n
17
1.8
5.9
1.3
89T43
1.002
.068
2.0
ni0
n

qeuvadarCaontginlued,.
Peak
Oxley
in
Formation
Guilmette
and
Dolomite
Simonson
Devonian
the
from
collected
samples
of
Analyses
3.
NTable
northeast
0.34
1.3
().002
0.14
189T24
<
n
1.0
i0
n
I
inn ().
2.4
0.26
0.58
150
189T25
<
n
1.4
n0n
In
i 0.54
1.5
0.33
0.40
0.002
5
<
89T27
n
10
in
i
n11
n In
7
0.33
1.2
0.44
19
(().
<
30
2189T
).002
I00
nn0n 00.96
<n
89T35
0.19
1.6
1.002
1.0
n0n
in
In
i 02.2
<(189T39
n
2.1
6.5
2.0
6.3
15
.002
).033
n 05

Zn aaf
Ni
Nb
Sample
Hg
Zn
Sb
Cd
W
V
Th
Sr
Sn
Sc
As
Zr
Au i-5
number
aac (100)
(5)
(200)
(10)
s (0.1)
(20)
(0.05)
(0.02)
(2)

096
H
160
1.05
n 000 ().
n88T()4
15
300
10
50
2200
2
<
1300
()
.9
n 10
288T()6
&
n
01

5
2
<00
.1
n0

12.1
89T32
n
380
0.100
0.36
81
1.3
3.9
3n
10
I00
.5l 89T36
(15
n
1.8
2.5
0.002
0.20
5.5
0.57
2.2
).93
in
in
n 0.002
0.08
1.7
0.37
0.87
02<
1.1
n l
1189T4
.033
ni.4
000n 0.002
189T.42
0<
n
4.8
11
0.72
0.08
4.3
0.18
4.0
.071
i0
n n

n
15
0.97
().002
0.04
9.7
70
3.5
0.063
14.5

5
89T26
0&
3.4
i.11
in
n0n

of
dlimit
>,
shown]
value
the
than
greater
be
to
eteremrinmaitnioend;


[eAtnearmliyntaitcioanl, coupled
method
analyzed,
element
show
headings
column
The
percent.
which
Ti,
P,
Fe,Mg,Na,
Ca,
for
except
million,
per
parts
in
given
are
dresults
of
limit
lower
the
and penatdrceutrcnmotignihravetpsihleoysnc
d(shown
of
Methods
percent).
are
which
Ti,
and
P,
Fe,Mg,Na,
Ca,
for
except
million,
per
parts
sin
emission
s,
iare:
10-element
5-
i-10,
or
i-5
analyses;
plasma
Mg
Sample
Cd
Bl
Be
Ag
Fe
Ca
TCo
P
Na
B
B
Mo
Mn
La
Au
As
Ge.
Ga
Cu
Cri
a number
s
s
(002)
(0.05)
(02)
(50)
(1)
(20)
(200)
(0.5)
(0.002)
(5)
(10) (
|
0
14 ) 5
0.01
0.3
15
5
088T
<
3
20
50
30
.02
n (200
<
15
0l
0.03
)20
088T27a
20
|
700
1.2
n 50
00.3
<
5
O
70.2
88T33
|
700
7.07
07
n 030
<
1.5
0.2
S()
2
0l,000
n.2
88T02
100
1
10
50.5 0l
0 70<
2
0
0.15
5
20
30
n
n.20 70
88T03
0.2
700
50
1 010
(0
5
05
|
0.02
n)7
500
<
88E12
n 088T29
0.03
|
<
030
l0.2
500
300
1n
10 20
.5 l
0.02
5,000
10
7
20
89T40
200
|
<
5n

rebtserpoemrvciptnaitviteolny;s
analyses,
vapor
cold
aac,
analyses;
furnace
graphite
aaag,
Other
analysis.
flame
absorption
atomic
daaf,
of
limit
lower
the
at
detected
not
n,
are:
*,

50
0l
0
0<
0.2
1200
5 .1
.2
88E07
100
( 0
0.07
10
3n
n5) 188T34
2
0
15
5
500
2n
0.1
l
0
<
0.7
7
30n520 (300
15
3
15.5)20 02
88T27b
0
()2
()
<
|
1
7 n 015
7 150
70
088T28
05
0.05
300
5
<
200
|
10
.2
n 70
02
0
0.5
<
50
5
| 5 50
700
1
30
88T3]
l
0.02
n0 5
<
3
0O
07
()
10.22
88T32
0
|
500
70
7.1
n (88TS7
<
105
50
20
5
7
0 2
0.5
0.07
2,000
.5)
300
|
100
200
n 0200
0
1.5
|
50
30
n52
088E13
0.1
0.2
500
10
5
<
20.5

050
2
0
<
l3
200
20
20
88E08
0100
0.5
5
10
1n
.1 50
70
055
<
0.2
()5
1.5
88T35
05
()300
1
0
2.2
.505
n 010
)20
5
()
7 205
88T30
1
0
02
|
500
100
5
<n 0100
088T80
2
0.15
1.5
700
5.07
0.2
10
70
30
3 .20
n

quadrangles,
adjacent
and
quadrangle
Peak
Oxley
the
in
faults
from
collected
samples
of
Analyses
4.
Table
Nevada.
northeast

Bed ing-paral el
faults
North-south
faults
faults
East-west

of
limit
lower
below
but
ddetected,
>,
shown]
value
the
than
greater
be
to
eteremrinmaitnioend;

for
except
million,
per
parts
in
given
are
dresults
of
limit
lower
the
and
(shown dpena0tdr-ceutrcnmlotigenhrmavetpsihnleotysnc *,
[eAtnearmliyntaitcioanl, coupled
method
analyzed,
element
show
headings
column
The
percent.
which
Ti,
P,
Fe,Mg,Na,
Ca, of
Methods
percent).
are
which
Ti,
and
P,
Fe,Mg,Na,
Ca,
for
except
million,
per
parts
sin
emission
s,
1are:
5-
i-10,
or
i-5
ianalyses;
respec
analyses,
plasma vapor
cold
aac,
analyses;
furnace
graphite
aag,
atively;
Other
analysis.
flame
absorption
atomic
daaf,
of
limit
at
detected
not
n,
are:
lower
the
below
but
detected,
ebtseroemrvipnatiton;s
Hg
Ag
Sample
Sb
Cd
Bi
Mo
Cu
ZAs
Y
Th
Sr
Sn
Sc
Zr
Au
Pb
Zn
W
VPb
Ni
Nbn i-10
number
aag
aac
s
(0.002)
(0.02)
(0.06)
(0.09)
(0.045)
(0.03)
(0.6)
(200)
(100)
(10)
(5)
(20)

3.0
3.2
n .067
07.6

0.16
4.5
110
<
40
4.8
89T.40
5.004
n0

Sample
Hg
Pb
Ni
Nb
Sb
Cd
Bi
Zn
As
Y
W
V
Th
Sr.
Sn
Sc
Zr
Au
Zn s
(0.05)
(0.02)
(2) (15

<
0100
301
20
7
2
5 ).5 0100
88E07
1.18
n 15
2
<
5
70
44.2
88E08
10
1()
3
1n
&
n
4 00
n0 <30
010
25
100
1
2
3n 0 30
88T34
5 .1
4 20.32
188T35
5
<
l0100
215
280l
.1
n 5
20110
2
< .170 <
0.02
14
88T
n 5
2
01 0 88T27b
88T27a
7()
(1
10
.04
n). 2
<
07
50
01 0 0.12
.12
n.2 20
2
<
2
1
5
10()
01()
13
50
88T28
n.2
0 2 05
0<.10 20
200
150
1
70
88T30
.32
n 050
2
1<
|
3 07 0.1
88T3]
300
200
44
n <1100
5
050
6
2
< .16
20 0.04
88T32
7n 0100
319
2
< .40 <
88T33
5
200
30
20
n 5
1470
9 .2
60
0.80
88T57
30
100
200
500
l()
2
170
n 200
08
<
6 .1
7 <
0.42
88T02
20
50
25
&
15
2
23
100
00
n 5
0100
2
33
4
8 .1
0.68
88T03
20
30
150
21
n00 010
212
2
<
88E12
20
.160 0.08
40 0.24
1n 220
2
<88E13
5
010
97 <
100
.1
n 20.04
230
(1
88T29
15
5
10
5040 <
).2
n
n 2
5
20
030
11
2588TSO
200
4 .180
n
n

Peak
Oxley
the
in
faults
from
collected
samples
of
Analyses
4.
qTable
adjacent
Nand
northeast
euvadadrCaongtnilgneulsde,.

number
s
(0.1)
(2)
(5)
(100)
(10)
(20)
(200)

Bed ing-paral el
faults
North-south
faults
faults
East-west

of
dlimit
>,
shown]
value
the
than
greater
be
to
eteremrinmaitnioend;


[eAtnearmliyntaitcioanl, coupled
method
analyzed,
element
show
headings
column
The
percent.
which
Ti,
P,
Fe,Mg,Na,
Ca,
for
except
million,
per
parts
in
given
are
dresults
of
limit
lower
the
and pena0tdr-ceutrcnmlotigenhrmavetpsihnleotysnc
d(shown
of
Methods
percent).
are
which
Ti,
and
P,
Na,
Mg,
Fe,
Ca,
for
except
million,
per
parts
sin
emission
s,
1are:
5-
i-10,
or
i-5
ianalyses,
plasma

rebtserpomrvciptnaitvieolny;s,
furnace
graphite
aag,
vapor
cold
aac,
analyses,
atomic
aaaf,
Other
analysis.
dflame
of
limit
lower
the
at
detected
not
n,
are:
<, Mg
Sample
Co
Cd
Ag
Na
Fe
Ca
Mo
Mn
La
Ge
Ga
Cu
B
B
B1
Be
Cri number
TAs
PAu
a s
s
(200)
(0.5)
(0.002)
(02)
(0.05)
(5)
(10)
(0.02)
(50)
(1)
(20) 70
5
1
<
01,000
150
05
88T18
0|
)
()
.2
n
n0 50
(15
5 1005,000
n 50
200
2
88T19
0.7
().5
150
S()
|
<).5
n 5
1100
2,000
500
<
5)
(88T20
()5
()
0
1
0 ()
n
n) 300
2
5 30
l()
l
7
().5
()
88T2]
().2
700
100
nn
50

88T26a
().5
10
300
3
5
(
(<
2
15
500
200
).
n
n) 150
7n
88T26b
l.000
200
1,500
05
5
(
2
()
15
n()
)

30
150
20
7
< 2,000
0|
0
88T17
(
&
007
2
().5
()
()
n
n.2
5
) Hg
Sample
Cd
Nb
Au
Zn
Sb
Z
Y
w
V
Th
Zr
H1
Sr
Sn
Sc
Pb
N1
As s (002)
n number
s (0.05)
(0.1)
(2)
(100)
(10)
(5)
(200) 9
(20) 120
015
100
4
|
2
<
14
|88T1
700
()
n
n 305
8 () 2
<
28
70
5020 20
88T18
1
<
20
.8
n
n 2
<188T19
(100
6
5
0.02
500
150)0 ()
1(X)
50
n
n4 1200
l
2
|
0
<5
20 0) 0.06
l()
6
88T20
(
20
n
n (150
<
2 ) 0.34
150
15
6
88T2]
70
10
00
n
n 05
2
<
7 .370 ()
200
1
70
88T26a
15
50
nn 2
<
950
22
200
01
5
88T26b
1()
5n.4
0

gueaodcrhaenmgilceasl,
for
collected
samples
of
Analyses
5.
qTable
adjacent
and
quadrangle
Peak
Oxley
the
from
background
Nevada.
northeast

of
limit
lower
below
but
ddetected,
>,
shown]
value
the
than
greater
be
to
eteremrinmaitnioend;
 GOLD IN CARBONATE ROCKS, CEDAR PEAK, NEVADA 259

Table 6. Latitude and longitude of samples collected in 1988 range-front fault (which is currently active). This
from the Oxley Peak, northwestern Wells Peak, southwestern Me might explain the higher values for molybdenum and
landco, and southeastern Summer Camp quadrangles, northeast silver obtained from samples along these faults.
Nevada.
Geochemical anomalies have been noted from Middle
(See figure 3A for approximate location of sites: *, degrees;', minutes;", seconds] and Upper Devonian rocks in the northern Pequop Moun
tains (approximately 20 miles east of Wells, Nevada) and in
Map Latitude (N.) Longitude (W.) the Kings Canyon area, in the southern Confusion Range in
number o - -- O - --

western Utah. Zebra-banded dolomite in the Guilmette For

1 41 15 3.615 114 56 23.894 mation, exposed along Interstate Highway 80 just west of
2 41 14 17.839 114 56 46.319
Pequop Pass, contains anomalous concentrations of gold,
3 41 14 19.503 114 56 42.497
mercury, lead, zinc, and barium (Erickson and others,
4 41 14 24.923 114 56 35.122
1966). These beds appear to be at or slightly above the
5 41 13 27.386 114 55 4 1,468
6 41 13 39.200 114 55 55.500
stratigraphic position of the gold-bearing Guilmette beds at
7 41 9 39.371 1 14 58 50.592 Cedar Peak. In the Kings Canyon area, samples from the
Simonson and Guilmette contain anomalous concentrations
8 41 9 57.495 | 14 58 45.659
9 41 9 58.143 1 14 57 10.107 of gold, silver, mercury, arsenic, antimony, fluorine, bar
1() 41 13 41,661 | 14 55 11.577 ium, copper, lead, and zinc (Zimbleman and others, 1990).
|1 41 13 38. 195 114 54 42.945 These samples included jasperoid and breccia; bedrock
12 41 12 40.665 | 1.4 52 49.871 samples for background concentrations were also collected.
13 41 14 57.050 | 1.4 52 9.674
Zebra-banded dolomites are not uncommon in Middle
14 41 15 28.296 114 52 15,430
15 41 15 27.854 | 1.4 52 22.300
and Upper Devonian units in the eastern Great Basin
16 41 14 49. 131 | 14 53 2.724 according to F.G. Poole (oral commun., 1991). The impor
17 41 14 40.360 114 53 10,935 tance of zebra-banded dolomite in minerals exploration is
18 41 12 19.840 | 14 56 41,922 not clear, but the occurrences of anomalous concentrations
19 41 13 23.209 l 14 55 14, 119 of metal in them records the movement of metal-bearing
20 41 13 17,039 114 55 1 1,087 fluids. A closer look at these types of rocks is needed to
21 41 13 40.482 114 55 21,010 understand the origin of the zebra-banded structure, their
22 41 ll 20,089 | 1.4 55 51.012 contained metals, and the timing of fluid movement.
23 41 11 7.838 114 55 15.456
24 41 ll 9,631 114 55 4.826
25 41 9 43,009 | 14 59 5.427
26 41 9 50.053 114 58 47.307 REFERENCES CITED
27 41 ll 8, 160 | 14 56 9,024
28 41 ll 26, 136 114 55 57,187
Erickson, R.L., Marranzino, A.P., Oda, Uteana, and Janes, W.W.,
29 41 11 58. 150 | 14 56 1.897
30 41 13 5.368 114 54 27,235 1966, Geochemical reconnaissance in the Pequop Mountains
and Wood Hills, Elko County, Nevada: U.S. Geological Sur
vey Bulletin 1198E, 20 p.
Horton, H.A., Jr., and De Voto, R.H., 1990, Dolomitization and
Formation are exposed to the north of the block but
contain few zebra-banded beds; diagenesis of the Leadville Limestone (Mississippian), cen
tral Colorado, in Beaty, D.W., Landis, G.P., and Thompson,
3. The distribution patterns for most metals in faults, both T.B., eds., 1990, Carbonate-Hosted Sulfide Deposits of the
high and low angle, are similar, except that gold was Central Colorado Mineral Belt: Economic Geology Mono
not detected in any fault zone except one at Cedar Peak graph 7, p. 86107.
where Simonson and Guilmette are juxtaposed;
Landis, G.P., and Tschauder, R.J., 1990, Late Mississippian karst
4. The siliciclastic units juxtaposed against the Simonson
caves and Ba-Ag-Pb-Zn mineralization in central Color
Dolomite and Guilmette Formation do not appear to be adoPart II. Fluid inclusion, stable isotope, and rock
altered; geochemistry data and model of ore deposition, in Beaty,
5. The youngest mineralizing event took place after D.W., Landis, G.P., and Thompson, T.B., eds., 1990, Carbon
about 14 Ma, the age of a rhyolite flow cut by east ate-Hosted Sulfide Deposits of the Central Colorado Mineral
west-trending, high-angle faults at the east end of the Belt: Economic Geology Monograph 7, p. 339366.
Cedar Peak block; Smith, J.F., Jr., Ketner, K., Hernandez, G.X., Harris, A.G., Stamm,
6. Several mineralizing events, separated in both time R.G., and Smith, M.C., 1990, Geologic map of the Summer
and space, may have occurred. The north-south-trend Camp quadrangle and part of the Black Butte quadrangle, Elko
ing faults along the foot of the range are highly silici County, Nevada: U.S. Geological Survey Miscellaneous In
fied, probably as the result of hot-spring activity at the vestigations Series Map I-2097, scale 1:24,000.
260 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
Thorman, C.H., and Brooks, W.E., 1988, Preliminary geologic map
of the Oxley Peak quadrangle, Elko County, Nevada: U.S.
Geological Survey Open-File Report 88755, scale 1:24,000.
Thorman, C.H., and Ketner, K.B., 1979, West-northwest strike-slip
faults and other structures in allochthonous rocks in central and
eastern Nevada and western Utah, in Newman, G.W., and
Goode, H.D., eds., 1979 Basin and Range Symposium: Rocky
Mountain Association of Geologists, p. 621636.
Thorman, C.H., Ketner, K.B., Brooks, W.E., Snee, L.W., and Zim
mermann, R.A., 1991, Late MesozoicCenozoic tectonics in
northeastern Nevada, in Raines, G.L., Lisle, R.E., Schafer,
R.W., and Wilkinson, W.H., eds., Geology and Ore Deposits of
the Great Basin, Symposium Proceedings: Geological Society
of Nevada, v. I, p. 2545.
Vandervoort, D.S., and Schmitt, J.G., 1990, Cretaceous to early
Tertiary paleogeography in the hinterland of the Sevier thrust
best, east-central Nevada: Geology, v. 18, p. 567570.
Zimbleman, D.R., Nutt, C.J., Hageman, Phil, Hill, R.H., Fey, D.L.,
Arbogast, B.F., and Bullock, J.H., Jr., 1990, Selected analytical
results, sample locality map, and discussion of trace-element
anomalies for rock samples near Kings Canyon, Confusion
Range, Millard County, Nevada: U.S. Geological Survey
Open-File Report 90317, 16 p.
als
r
w
APPENDIX TO CHAPTER Y
R.
l.
::
... S DESCRIPTION OF MAP UNITS SHOWN ON FIGURE 2
EASTERN-FACIES UNITS
Td Dinwoody(?) Formation (Triassic)Shale and
limestone. Interbedded greenish, gray, and choco
late brown thin-bedded to fissile shale with subor
dinate brownish, thin-bedded to platy, fine-grained
limestone to shaley limestone.
Pm Murdock Mountain Formation (Permian)Lime
stone, chert, and siltstone. Gray, medium- to thick
bedded, medium- to coarse-grained bioclastic lime
stone to dolomitic limestone; brachiopod valves and
spines are common.
Sandstone of Melandco (Lower Mississippian)
Sandstone, conglomerate, and shale. Brown to red
dish-brown, medium- to thick-bedded, medium- to
coarse-grained to gritty chert-quartz sandstone.
Mtp Tripon Pass Limestone (Lower Mississippian)
Gray, medium- to thin-bedded, fine- to medium
grained limestone with graded bedding. Locally be
comes sandy, the grains being quartz and chert.
Guilmette Formation (Upper Devonian)Gray,
medium- to thick-bedded, medium-grained lime
stone. Contains medium- to thin-bedded dolomite as
interbeds in lower 3 to 6 m. Basal part of unit con
sists of thin-bedded, argillaceous, Atrypa-bearing
limestone that forms a smooth slope. Formation is
approximately 300 m thick.
Simonson Dolomite (Middle Devonian)Gray,
medium- to thick-bedded, fine- to medium-grained,
laminated dolomite with minor interbeds of lime
stone. Base is not exposed. Thickness probably ex
ceeds 450 m.
WESTERN-FACIES UNITS
Argillite, shale, and chert (Mississippian)Black
siliceous argillite with chert-quartz arenite beds.
Black shale and siltstone with scarce, bedded, black
chert and bedded barite. Fine-grained sandy beds.
Upper and lower contacts are low-angle faults.
Thickness probably exceeds several hundred feet.
Dsh Shale, limestone, and chert (Devonian)Dominant
lithology is black to dark brownish gray, fissile to
thin-bedded, locally weakly phyllitic shale. Con
tains interbedded, medium-gray, thin- to medium
bedded, medium-grained, in part bioclastic, lime
stone; locally beds contain well-rounded, frosted,
medium-grained quartz. Black, thin-bedded chert is
present locally as thin interbeds. Thickness of unit
exceeds 250 m. Upper and lower contacts are low
angle faults. Contains four early Late Devonian con
odont assemblages (B. Wardlaw, written commun.,
1988). Equivalent to unit Dt of Smith and others
(1990) in the Summer Camp quadrangle, where it
ranges from Early to Late Devonian in age.
Shale, limestone, and chert, undivided (Devo
nian)Contains lithologies found in both units Dsh
and Dsl.
DSl Shale, limestone, and chert (Upper Devonian)
Contains more or less equal amounts of fissile, black
shale; yellowish thin-bedded to platy, argillaceous,
siliceous, fine-grained limestone; and black, thin
bedded, platy chert. Contains sparse bedded barite.
Thickness is unknown, but probably exceeds 150 m.
The lower and upper contacts are low-angle faults.
Two conodont assemblages are middle Middle De
vonian and Devonian to Early Mississippian in age
(B. Wardlaw, written commun., 1988). Equivalent
to unit DSOw of Smith and others (1983), which
they determined to range in age from Ordovician to
Devonian.
Sw Siltstone, chert, and argillite (Silurian)Graptolite
bearing, yellow-weathering, greenish-gray, mica
ceous siltstone, chert, and argillite underlain by
black, white, red, and green chert.
Shale and chert (Middle to Upper Ordovician)
Black bedded chert containing sparse graptolites,
underlain conformably by thick, graptolite-rich,
black shale.
DSOw Shale, siltstone, chert, limestone, and argillite, undi
vided (Devonian, Silurian, and Ordovician)Un
differentiated western-facies rocks, including units
mapped separately as Dsl, Dw, Sw, and Ow. Thick
ness probably exceeds 200 m.
 CHAPTER Z
OCCURRENCES OF NATIVE GOLD
CONTAINING SILVER, MERCURY, COPPER,
AND PALLADIUM IN LEMHI COUNTY,
EAST-CENTRAL IDAHO
By GEORGE A. DESBOROUGH, WILLIAM H. RAYMOND, and JAMES NISHI?
ABSTRACT
Native gold grains collected from streams in Lemhi
County, Idaho, range in chemical composition from pure
gold to varieties containing up to 36.3 weight percent mer
cury, 30 weight percent silver, 4 weight percent copper,
and 2 weight percent palladiumall of which are in solid
solution with gold. Very fine grained (< 150 pum) native
gold grains were recovered from stream-sediment samples
and studied by scanning-electron-microscope (SEM) meth
ods to determine their size, morphology, and chemical
composition. Analyses of individual native gold grains
show significant variations in chemical composition
among grains in a single stream-sediment sample and also
show distinct variations in compositions between samples
from different localities. The chemical compositions of
Some of these grains are unique with regard to palladium
or copper content when compared with published data for
samples from other localities in the United States.
The chemical diversity in the compositions of the
grains, locally and regionally, is significantly greater than
compositional variations reported for native gold for most
other regions of North America.
Most of the samples studied are from streams that
drain areas of Middle Proterozoic strata; most samples
from drainages underlain chiefly by Paleozoic sedimentary
or Cenozoic volcanic rocks do not contain native gold
grains. However, gold was found in samples from streams
"U.S. Geological Survey, Mail Stop 905, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 921, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
*U.S. Geological Survey, Mail Stop 917, P.O. Box 25046, Denver
Federal Center, Denver, CO 80225.
near large faults that juxtapose Cenozoic volcanics and
Middle Proterozoic rocks.
The geochemical signatures of native gold reported
here are sufficiently unique to warrant further studies with
a higher density of samples in areas of the Lemhi Range
that are currently being geologically mapped. This should
allow delineation of potential lode sources. Additionally,
regional studies east of the study area should aid in delin
eating the region that contains copper- and palladium
bearing gold. Because of the association, known from the
literature, of copper- and palladium-bearing gold with
mafic rocks, such sources in or beneath the Middle Prot
erozoic rocks should be considered.
INTRODUCTION AND
METHODS OF STUDY
Most of the study area is composed of siltite and
fine-grained feldspathic quartzite of Middle Proterozoic
age that may be as thick as 19,000 m (J. Connor, written
commun., 1991); metamorphism of these strata is in the
greenschist facies. Proterozoic-age granite occurs in the
Leesburg area (fig. 1). Minor amounts of Ordovician syen
ite are present in the northwestern part of the study area,
and small localized areas of lower Paleozoic strata are
present in the central and southern parts. Remnants of
Cenozoic extrusive and lacustrine rocks are present in
many areas, mainly in valleys and on mountain flanks, but
are a small proportion of the exposed rocks.
The morphology, size, and chemical composition of
native gold grains collected from streams in Lemhi
County, Idaho, have been analyzed using scanning
264 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

1 * 4
-

|
<)
North Fork
-y

SCALE
O 5 10 m i le s
|--|--|-1

O 5 10 k i ! ome ters Wick gm

| ode

*_-3 -

Diamond Creek
e

& Word Gu ch

:z Smith
Gu ch

CO
Leesburg ~

&2o
*2,.

*.
3.

Cobo t ->

45 -

Hayne's Creek

Tendoy

McDev i t t Creek

L
( )
2.
10 <
Po is on C.
Creek Lemhi

T
X &ear *zz. /

| DAHO _*& H

****/ c
>
* Boise A lv
Poco te | | o

Figure 1. Drainages and sample localities (solid circles) in Lemhi County, Idaho.

electron-microscope (SEM) techniques coupled with and to establish local and regional variations in the
simultaneous semiquantitative energy dispersive analysis. geochemistry of native gold.
The results of these analyses are used to interpret the ori Previous studies of gold in stream sediments from
gin of the gold grains and included or attached minerals Lemhi County indicated that native gold grains are
 NATIVE GOLD, LEMHI COUNTY. IDAHO
abundant in areas that have little or no history of mining
or prospecting (Desborough and others, 1990). These areas
had been previously considered to be unfavorable for the
occurrence of gold deposits. Native gold was concentrated
in the laboratory from bulk stream-sediment samples.
The gold grains were examined with a binocular
microscope (40 or 80), and the sizes and masses of gold
grains were estimated. Based on this examination, the
number of gold grains per kilogram of bulk stream
sediment sample was reported, and estimated gold concen
trations in parts per million (ppm) were reported.
For the present study, bulk samples collected in the
field ranged from about 1 to 5 kg; the number of gold
grains in these ranged from about 10 to 50 gold grains per
kg. A large sample was selected in order to collect a mini
mum of about 20 gold grains for each locality. However,
this was not achieved for several samples. Locations of
samples studied and major drainages in the area are shown
on figure 1.
Native gold grains in stream sediments were concen
trated and isolated by laboratory mechanical-panning tech
niques (English and others, 1987). The size of the native
gold grains range from about 20 to 800 pum; most grains
were in the size range of 20 to 100 pum. Individual gold
grains were hand picked with a single bristle on an artists
paint brush and mounted on sticky tape attached to SEM
mounts that were then coated with carbon. Because of
their small size and the difficulty of handling of each
grain, about 10 to 20 percent of the grains were lost in
transfer from the pan to the SEM mount.
Prior to chemical analysis (for Au, Ag, Cu, Pd, and
Hg), the morphology of the native gold grains was exam
ined. Most of the gold grains were relatively smooth;
some were slightly porous and had amorphous or irregular
shapes. Subhedral grains were sparse or rare, and euhedral
shapes were virtually absent.
For this SEM study, detection limits of palladium and
silver are about 0.1 weight percent. The detection limits
for copper and mercury are estimated to be about 0.2 and
about 1.0 weight percent, respectively. The high detection
limit for mercury is due to interference by gold. These
detection limits are about 100 times higher than those
from the semiquantitative emission spectroscopy method
and about 10 times higher than those from the electron
microprobe, which requires a flat, polished surface.
RESULTS
A summary of the compositional data for grains from
each locality studied is given in table 1.
Allison Creek.-Seven native gold grains from this
locality range from about 40 to 80 pum in their largest
dimensions. Pure gold, silver-bearing, mercury-bearing,
265
and palladium-bearing grains were found. Copper-bearing
grains were not found. The highest amount of palladium
(2 weight percent) was found at this locality. The
geochemical signature for this locality is similar to that of
Poison Creek; both of these localities are along or near a
large north-south-trending fault that juxtaposes Middle
Proterozoic and Cenozoic rocks.
Arnett Creek.This locality is upstream from the
Haidee lode gold mine and is in the headwaters of Arnett
Creek. Only 12 native gold grains, ranging up to 200 pum
in average diameter, were analyzed. Grains from this
locality contain all five elements for which samples were
analyzed (table 1) and also contain the highest proportion
of copper-bearing grains. The geochemical signature is
similar to those from the four other sample localities in
group 3 (figs. 1 and 2). Only traces of pyrite are present
in the concentrate.
Bear Valley Creek-This locality, near the mouth of
the creek, has the largest native gold grain (800 pum in
diameter) recovered and also has grains of the highest
average purity. Two SEM mounts were made. Of 20
grains analyzed, 7 are pure gold. A SEM photo of 14
grains is shown on figure 3A. Palladium, copper and mer
cury were not detected. One gold grain has a pyrite inclu
sion; another has both quartz and a potassium-, iron-,
magnesium-, aluminum-silicate inclusion with micaceous
morphology that is presumed to be biotite. Both pyrite
and chalcopyrite are present in the heavy-mineral concen
trate. The geochemical signature is simple, but it is unlike
the other samples studied (fig. 2).
Blackbird Creek.-This sample was collected less
than 1 km downstream from the Blackbird cobalt mine.
The largest of 23 grains studied is about 200 pum in diame
ter. Neither mercury nor copper was detected, and thus,
this sample is similar to those from Allison Creek and Poi
son Creek (fig. 2).
Fourth of July Creek. Seventeen grains ranging
up to 250 pum in diameter were analyzed from the mouth
of the creek. A SEM photo of these grains is shown on
figure 3B. Palladium and copper are present, but mercury
is not; this sample has the highest mean silver content of
all samples (table 1). It is also from the mouth of a very
large drainage basin (about 50 km2), as is the Sandy Creek
sample, which has a similar geochemical signature (fig. 2).
Freeman Creek-Twenty-six grains from near the
head of Freeman Creek were analyzed. Two SEM
mounts, one of 21 grains and one of 19 grains were made.
A SEM photo of the 19-grain mount is shown on figure
3C. In addition to gold, the concentrate contains abundant
pyrite and minor galena and chalcopyrite. The locality is
downstream from the Ore Cash gold mine, which ceased
operations around 1900. A single grain was pure gold.
The largest grain is about 200 pum in the largest dimen
sion, but most grains are smaller than 100 pm (fig. 3C).
The high proportion of mercury-bearing grains is thought
266 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 1. Number of grains and percentage of native gold grains containing only gold, silver, palla
dium, copper, or mercury for localities in Lemhi County, Idaho.
[Number in parentheses is the percentage of grains containing that metal. Number in brackets is the range or amount of metal
concentration(s). Number in bold print is mean silver content of silver-bearing grains)

Locality 100% Silver- Palladium- Copper- Mercury

(no. of grains) Au bearing bearing bearing bearing
grains grains grains grains
1 6 1 () 1
Allison Creek (7) (14) (86) (14) (14)
(0.8-15.1] [20] [57]
10.4
2 8 1 5 2
Arnett Creek (12) (17) (66) (8) (42) (16)
(0.5-16.4) 1.7 (0.5-0.9) [5.4-142]
5.1
7 13 () () 0
Bear Valley (35) (65)
Creek (20) (0.2-1.8]
0.6
8 13 5 () 0
Blackbird (35) (56) (22)
Bear Valley (0.3-300) [0.5-0.9)
4.25
1 16 3 6 ()
Fourth of July (6) (94) (18) (35)
Creek (17) (0.5-29.2] [04-06) (0.4-1.3]
12.1
1. 25 7 1 19
Freeman (4) (96) (27) (4) (83)
Creek (26) (0.2-10.3] (0.4-1.0] [4.0] [4.7-25.9)
5.5
() 29 6 10 4
Lake Creek (29) (100) (21) (34) (14)
[12-74) (0.40.9) (0.4-1.3] (4.816.4]
4.8
14 23 12 () 4
Poison Creek (44) (32) (52) (27) (9)
(0.2-182] (0.5-1.6] [30-23.2]
3.8
1 15 6 7 ()
Sandy Creek (20) (5) (75) (30) (35)
(0.3-174) (0.3-0.6] (0.5-1.0)
6.4
0 12 3 8 ()
Smith Gulch (13) (92) (23) (62)
[24-13.0] |0.6-0.9) (0.4-1.2]
7.6
() 19 3 2 21
Sourdough Gulch (21) (90) (14) (10) (100)
(0.3-118] (0.407) (0.5-0.6] (8.6-36.3]
5.2
5 20 4 7 1
Ward Gulch (26) (19) (77) (15) (27) (4)
(0.2-14.6] (0.5-1.0] (0.5-18] [13.5]
7.0
3 11 3 13 0
Wickam lode (21) (14) (52) (14) (62)
(0.3-3.0] [07-09] (0.4-1.3]
0.9
 NATIVE GOLD, LEMHI COUNTY. IDAHO
GROUPS
1 | 2 1 3
80
7O
6O
50
40
30
2O
1O
O i ||
i| ; ! :|
Figure 2. Percentages of gold grains that are silver bearing, copper bearing, palladium bearing, and pure gold for each locality stud
ied. Five geochemical groups are tentatively recognized. Percentages were obtained by dividing the number of grains in each category
by the total number of grains at each locality. These four parameters were normalized to 100 percent.
to be an artifact, and thus, the geochemical signature is
placed with that of the Sourdough Gulch sample (fig. 2).
Hayden Creek.-A sample from above the Bear Val
ley Creek confluence contained two grains of gold-copper
minerals. One grain has Au-70.5, Cu=22.0, Ag=5.0, and
Zn=2.6 (all weight percent), corresponding to Auo.425
Cu0.438Ag0.0582no,051, which is the mineral tetraauricu
pride (AuCu). The second grain has Au-58.6, Cu-37.8,
and Ag+3.6 (all weight percent), corresponding to
Cu3Au/Ag0.125, which is essentially the composition of th
e mineral auricupride (Cu3Au). Both grains are 40 to 50
pum in their largest dimension. No native gold was recov
ered from this sample site.
Lake Creek.-This sample was taken about 1.5 km
upstream from Williams Lake. Twenty-nine grains were
studied (fig. 3D). Pyrite is present in the concentrate. In
addition to the grains shown on figure 3D, three coin
shaped grains of gold (mounted separately) appear to be of
bacterial-replacement origin based on the ornate morphol
ogy observed at high magnification in SEM studies (J.R.
Watterson, oral commun., 1991; Brooks and Watterson, in
press). These three grains range from about 100 to 300
267
Ag bearing
Cubearing
- Pd bearing
[] 100% Au
| :; ii
pum in diameter. A SEM photomicrograph of one of these
grains (fig. 4) shows a gold-replaced bacterial colony at
low and high magnifications. Only silver and gold were
detected, and the silver content is 3.6 weight percent.
Poison Creek.-These gold grains range in size from
about 20 to 250 pum; only five are 100 pum or larger (fig.
3E). Forty-three grains were analyzed (table 1). One grain
is a complex intergrowth of gold and pyrite and an uni
dentified mercury mineral that could not be analyzed;
another grain (not analyzed) is a finely intergrown mixture
of biotite and gold. Pyrite is present in the concentrate as
grains that are mostly smaller than 100 pum. A very small
spherical grain (a on figure 3E) is an artifact (discussed
later).
Sandy CreekTwenty grains analyzed from this
locality range up to about 200 pum in maximum diameter
(table 1). Mercury was not detected. Minor pyrite is
present in the concentrate. This sample is from a large
drainage area (about 30 km2), and thus, the gold may be
from many sources (fig. 1).
Smith Gulch.-This locality lies on the south side of
the Beartrack gold deposit of Meridian Gold, Inc., near
268 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

400nM Flw

4001N 2UAV

Figure 3. Scanning electron microscope photographs of native gold grains from six localities in Lemhi County, Idaho. Bar
Scales on each photo are in micrometers. Whitest grains are gold; gray grains present in some photos are other heavy minerals.
Samples are from: A, Bear Valley Creek (14 gold grains); B, Fourth of July Creek (17 gold grains); C, Freeman Creek (19 gold
grains); D, Lake Creek (29 gold grains); E, Poison Creek (44 gold grains); a is spherical artifact with 13 weight percent Cu
and 5 weight percent Hg; F. Ward Gulch (43 gold grains).
 NATIVE GOLD, LEMHI COUNTY. IDAHO
Figure 4. Scanning electron microscope photographs of a gold
replaced bacterial colony from Lake Creek. Bar scales in pum; black
line in lower right photograph represents 40 p.m. Photograph cour
tesy of John R. Watterson.
Leesburg (fig. 1). Only 13 grains were analyzed (table 1).
The largest gold grain is about 200 pum in the largest
dimension. Mercury was not detected, but copper and pal
ladium are present, and thus, the geochemical signature is
grouped with four other localities with similar composi
tions (fig. 2). Trace amounts of pyrite are present.
Sourdough Gulch.-This locality is about 3.5 km
north of the abandoned mining camp of Gilmore. It is just
southeast of the area shown on figure 1. The drainage
basin is about 3 km long and contains Paleozoic sedimen
tary and Tertiary intrusive rocks (Ruppel and Lopez,
1988). Galena, pyrite, and thorite are abundant in the con
centrate. Fine-grained gold (as much as 150 pum in diame
ter) is very abundant, amounting to about 50 gold grains
per kg (0.050.1 ppm). Mercury is present on all grains
analyzed (table 1) but is suspected to be an artifact
because adits with wastepiles are present upstream from
the sample site.
Ward Gulch.-This locality is near Smith Gulch, and
this short drainage is on the south side of the Beartrack
269
gold deposit. A SEM photograph of 43 gold grains from
this locality is shown on figure 3F, and analytical results
are given in table 1. The geochemical signature of the
sample is similar to that of Smith Gulch, which is the
adjacent drainage, 0.5 km to the east (figs. 1 and 2).
Minor pyrite is present in the concentrate.
Wickam lode.This sample is from gossan in an adit
about 1 km north of Diamond Creek; the host rock is
Proterozoic strata. It contains abundant native gold that
was associated with sulfide prior to oxidation. Twenty
one grains, all smaller than 100 pum, were analyzed (table
1). This sample contains the highest proportion of copper
bearing grains, and mercury is absent (table 1). The
geochemical signature is similar to four other localities
that are west of the Salmon River (figs. 1 and 2).
GEOCHEMICAL SIGNATURES OF
NATIVE GOLDPRIOR STUDIES
The most extensive recent study of the geochemistry
of native gold was done on Alaskan lode and placer
deposits by emission spectroscopy. One hundred and
ninety-one analyses for 28 lode occurrences and 2,438
analyses for 295 placer localities have been reported by
several investigators (table 2). Maximum amounts of pal
ladium, copper, and mercury are shown for lode deposits,
and maximum amounts of palladium and copper for placer
deposits are listed in table 2.
Desborough (1970), using electron microprobe meth
ods, performed 3,000 quantitative analyses on about 300
individual gold grains from 24 placer localities in 6 West
ern States and Alaska; maximum copper concentrations
(0.1-2.0 weight percent) were found in 18 grains from 3
localities. Desborough and others (1970), in a study of
placer gold in the northeastern part of the Colorado Mineral
Belt, analyzed 991 grains from 17 localities and found
detectable copper (0.10.25 weight percent) in 44 grains.
Of these 44 grains with detectable copper, 37 were from 4
large drainages near Boulder, Colorado. All of the copper
bearing grains analyzed in these studies contained 1 or
more weight percent of silver. Desborough and others
(1971), using the same method, analyzed 788 grains from 5
placer localities and 1 lode deposit in 5 Western States. The
results show that 138 grains from Ninemile Creek, Mon
tana, contained between 0.1 and 0.6 weight percent copper.
Thus, of 47 studied localities in the Western U.S.
and Alaska, only 12 have native gold with more than 0.1
weight percent copper. In summary, prior geochemical
studies of native gold in the Western United States and
Alaska show that only 5 of 370 localities contain more
than 1 weight percent copper in native gold; only 3 of
370 localities contain 210.2 weight percent copper in
native gold.
270 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992

Table 2. Summary of reported emission-spectrographic analytical data for copper, palladium, and
mercury in native gold for Alaskan lode deposits and summary of copper and palladium in native gold
for Alaskan placer deposits.
[Maximum palladium, copper, and mercury concentrations are shown in weight percent; conc., concentration. Mercury values re
ported for Alaskan placer gold are not given here because of possible introduction of mercury during commercial gold-recovery
efforts. Detection limits are: Cu=0.0001 weight percent, Hg=0.002 weight percent, Pd=0.0005 weight percent (Mosier, 1975)]

No. of No. of Maximum Maximum Maximum

Location analyses localities palladium copper mercury
conc. Conc. conc.

Lode gold
Fairbanks mining district
plus others'...... 191 28 0.002 0.24 6.2

Placer gold
Circlequadrangle ------------------------- 476 51 0.002 0.85

Livengood quadrangle" ----------------- 247 21 ().002 0.2

Various locations'......................... 160 33 < 0.0005 ().2

Fairbanks mining district............ 199 37 < 0.0005 0.1

Bonnifield mining district"........... 196 27 ().08 2.1
Tolovana and Rampart
mining districts........................ 48 5 ().04 ().3
Fortymile mining district 248 30 ().01 ().09
Four districts in Tanana and
Livengood quadrangles.......... 404 45 0.01 10.2
Koyukuk-Chandalar
mining districts"... 460 46 ().002 1.2
11 grains: 110.2%.
Cathrall and others (1989a, 1988a, 1990, 1987).
Cathrall and others (1988b).
Cathrall and others (1988a).
Cathrall and others (1990).
Cathrall and others (1989b).
Cathrall and others (1989c).
Cathrall and others (1987).
Cathrall and others (1989a).
McDanal and others (1988).
10
Mosier and Lewis (1986); Mosier and others (1989).

An electron microprobe study of native gold and gold
rich alloys in the gold-bearing horizon of the Skaergaard
intrusion, East Greenland (Bird and others, 1991, table 1, p.
1090), reports copper values of 1.5 and 11 weight percent,
respectively, in 2 native gold grains (analyses 1 and 2); 3
other grains (analyses 3, 4, and 5) are the mineral tetraauri
cupride, which contains 23.225.5 weight percent copper,
although the authors refer to all 5 analyses as representing
gold-rich alloys. In addition, the two copper-bearing
native gold grains contain 0.8 and 5.8 weight percent palla
dium; the three tetraauricupride grains contain 0.15, 3.0,
and 5.5 weight percent palladium and from 0.25 to 1.6
weight percent platinum (Bird and others, 1991, table 1).
Electron microprobe studies of precious-metal miner
als in the Jijal layered ultramafic-mafic complex,
Pakastani Himalayas (Miller and others, 1991, table 2, p.
1100), have revealed three gold- and copper-bearing
minerals. Two grains contain 24 and 22.7 weight percent
copper, respectively, and were identified as tetraauricu
pride. Another grain contains 5.3 weight percent copper
and is a gold-silver alloy referred to as electrum. The
two tetraauricupride grains contain 6.1 and 6.3 weight per
cent palladium. The copper-bearing electrum grain con
tains 0.6 weight percent palladium.
CONTAMINATION, ARTIFACT, AND
BACTERIAL GOLD
DiLabio and others (1988) conducted an interesting
study of the spherical form of gold, which investigated
many aspects of the chemical signature of both natural and
artifactual (e.g., jewelry-scrap gold) gold alloys. Laboratory
 NATIVE GOLD, LEMHI COUNTY. IDAHO
and sample contamination was shown to be significant in
some cases. In our Poison Creek sample, 1 of 44 native
gold grains analyzed was spherical and about 20 pum in
diameter; it contained 13 weight percent copper, 5 weight
percent mercury, and the balance is mostly gold and minor
amounts of silver (grain a, fig. 3E). This grain has a
microcrystalline structure (incipient crystal growth or
quenched crystal growth) on the surface like that shown by
DiLabio and others (1988, fig. 1E), which is typical of
quenched-liquid metals. We consider this grain to be a con
taminant (artifact) and do not report the results as repre
senting native gold. This aspect is discussed to illustrate our
awareness of possible metal contamination problems.
Mercury-bearing gold grains present significant prob
lems to interpretation because of the potential that mer
cury is an artifact. For some localities that have a history
of mining and prospecting (e.g., Freeman Creek and Sour
dough Gulch), abundant mercurian gold was expected
(table 1); 77 percent of the mercury-bearing grains are
from these 2 localities. However, localities at Allison
Creek, Lake Creek, and Poison Creek have no evidence or
history of mining or prospecting, yet mercury is present in
native gold at these localities. We expect that some mer
curian gold grains from the study area is of natural occur
rence, but we cannot prove it. Mercury is present in 13
percent, or 40 of the 308 grains analyzed. Naturally occur
ring mercurian gold is known from many lode deposits
and is reported in table 2 for Alaskan lode gold.
Brooks and Watterson (in press) and Watterson
(1992) have only recently demonstrated that gold may be
precipitated in, on, or by bacteria and (or) bacterial colo
nies and that native gold preserves the original morpholo
gies of the organisms. The gold pseudomorphs after
bacteria apparently form in near-surface environments
such as streams, soils, or deeply weathered bedrock. This
gold is of high purity and apparently contains less than
about 5 weight percent silver. Gold concentrated by bacte
ria in this form has not yet been shown to be of economic
significance. However, gold of bacterial-replacement ori
gin is easily recognized by the delicate and microscopi
cally ornate morphology identifiable in SEM studies
(Brooks and Watterson, in press; Watterson, 1992). Only
three disc-shaped native gold grains that we found in
Lemhi County may have a bacteria-replacement origin
(fig. 4). These are from a locality in Lake Creek (fig. 1) in
which fine-grained native gold is very abundant (> 20
gold grains per kg).
PERSPECTIVE AND DISCUSSION
Our purpose is to try to understand the possible
geochemical significance and the potential value to explo
ration that gold geochemical signatures may provide.
271
Pd (11)
Ag Cu
(93) (9)
Figure 5. Ternary diagram of Ag-Pd-Cu showing the relative
amounts of silver, palladium, and copper (weight percent) in mercu
ry-free native gold grains from Lemhi County, Idaho. Numbers at
apices indicate number of grains with only silver (93), only Pd(11),
and only Cu (9).
Figure 5 shows a ternary plot of copper, silver, and palla
dium for 182 mercury-free grains. Only nine grains (5 per
cent) contain detectable amounts of all three elements.
Ninety-three silver-bearing grains (51 percent) contain no
copper or palladium; 11 palladium-bearing grains (6 per
cent) contain neither silver nor copper; and 9 copper-bear
ing grains (5 percent) contain neither silver nor palladium.
Of these 182 grains, 32 (18 percent) contain silver and
copper (without palladium), and 23 (13 percent) contain
silver and palladium (without copper). These data are
interpreted to indicate a general incompatibility of exten
sive solid solution of all three elements (Ag, Pd, and Cu)
simultaneously in native gold under the conditions of dep
osition. With the exception of the nine gold grains con
taining all three elements, and the occurrence of
auricupride and tetraauricupride in Hayden Creek, most
gold grains are not similar in composition to native gold
found in mafic and (or) ultramafic rocks (Bird and others,
1991; Miller and others, 1991).
The occurrence of minor amounts of copper and
minor amounts of palladium in native gold in the study
area is significant. If we include all gold grains (including
those containing mercury): (1) Palladium occurs in 12 of
13 localities studied (92 percent), and palladium occurs in
54 of 308 gold grains analyzed (17 percent); (2) copper
occurs in 9 of 13 localities studied (69 percent), and cop
per occurs in 59 of 308 gold grains analyzed (19 percent).
When compared to the Alaskan native-gold data shown in
table 2, these occurrences of both palladium and copper in
native gold in Lemhi County, Idaho, must be considered
272 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
unusual. Differences and similarities in the populations of
gold grains from each locality studied are shown on figure
2. The number of silver-bearing grains, copper-bearing
grains, palladium-bearing grains, and pure gold grains was
determined for each sample, and the relative percentage of
each of these parameters was normalized to 100. On this
basis, five compositional groups of native gold signatures
are tentatively identified on figure 2. The first group has
no copper-bearing grains and has a high proportion of
pure gold grains. The second group of only one locality
has neither copper- nor palladium-bearing grains but has a
high proportion of pure gold grains. The third group has a
high proportion of copper-bearing grains and also has pal
ladium-bearing grains; all of these localities are west of
the Salmon River (fig. 1). The fourth group has a very
high proportion of silver-bearing grains and a low propor
tion of copper-bearing grains in comparison to the propor
tion of palladium-bearing grains. The fifth group has a
modest percentage of copper- and palladium-bearing
grains; this group might be better considered as part of the
third group. However, the average silver content of the
third group is 5.3 weight percent, whereas the average sil
ver content of the fifth group is 9 weight percent. We
hope that additional studies will determine whether or not
these tentative groups of gold geochemical signatures will
be distinct or not.
Based on the occurrences of native gold along post
Eocene faults, we infer that some of the gold was depos
ited during Cenozoic time. The extensive occurrence of
gold in areas underlain chiefly by Proterozoic metasedi
mentary rocks (lower greenschist facies) and locally by
Proterozoic igneous granitic rocks or thin mafic sills and
dikes is inferred to indicate the strong possibility that
some of the gold may have been deposited during Protero
zoic time. We tentatively infer that some of the unusual
compositions of the gold in this area may be related to
Proterozoic mineralization. Another possibility is that
some of the gold was deposited in Proterozoic rocks but
was remobilized by later mineralizing events that may not
have destroyed the Proterozoic geochemical signatures.
Further research on the native gold in the region
should reveal whether or not the copper-bearing and pal
ladium-bearing gold is geologically and (or) geographi
cally limited to certain terranes. Focus will be placed on
greater sample density in areas currently being geologi
cally mapped.
REFERENCES CITED
Bird, D.K., Brooks, C.K., Gannicott, R.A., and Turner, P.A., 1991,
A gold-bearing horizon in the Skaergaard intrusion, East
Greenland: Economic Geology, v. 86, p. 10831092.
Brooks, R.R. and Watterson, J.R., in press, On the microbial bio
geochemistry of noble metals, in Brooks, R.R., ed, Biological
Systems and Noble Metals, Chapter 6: West Palm Beach, Flor
ida, CRC Press.
Cathrall, J.B., Albanese, M., VanTrump, G., Mosier, E.L., and
Lueck, L., 1989, Geochemical signatures, analytical results,
mineralogical data, and sample locality map of placer and lode
gold, and heavy-mineral concentrates from the Fortymile min
ing district, Eagle quadrangle, Alaska: U.S. Geological Survey
Open-File Report 89451, 32 p.
Cathrall, J.B., Antweiler, J.C., and Mosier, E.L., 1987, Occurrence
of platinum in gold samples from the Tolovana and Rampart
mining districts, Livengood quadrangle, Alaska: U.S. Geolog
ical Survey Open-File Report 87300, 12 p.
Cathrall, J.B., Antweiler, J.C., VanTrump, G., and Mosier,
E.L., 1989a, Gold analytical results and gold signatures from
the Fairbanks mining district, Fairbanks and Livengood quad
rangles, Alaska: U.S. Geological Survey Open-File Report
89490, 23 p.
1989b, Gold, platinum, and silver analytical results and
gold signatures from the Bonnifield mining district, Fairbanks
and Healy quadrangles, Alaska: U.S. Geological Survey Open
File Report 89461, 23 p.
1990, Gold analytical results and gold signatures from the
Anchorage, Charley River, Healy, Iditarod, Juneau, Mt. Hayes,
Mt. McKinley, Ophir, Ruby, and Talkeetna quadrangles, Alas
ka: U.S. Geological Survey Open-File Report 90210, 25 p.
Cathrall, J.B., McDanal, S.K., VanTrump. G., Mosier, E.L., and
Tripp, R.B., 1988, Analytical results, geochemical signatures,
mineralogical data, and sample locality map of lode gold, plac
er gold, and heavy mineral concentrates from the Tolovana
mining district, Livengood quadrangle, Alaska: U.S. Geologi
cal Survey Open-File Report 88578, 32 p.
Cathrall, J.B., Tripp, R.B., McDanal, S.K., Mosier, E.L., and Van
Trump, G., 1988, Analytical results, geochemical signatures.
mineralogical data, and sample locality map of placer gold and
heavy mineral concentrates from the Circle mining district,
Circle quadrangle, Alaska; U.S. Geological Survey Open-File
Report 88676, 48 p.
Desborough, G.A., 1970, Silver depletion indicated by microanal
ysis of gold from placer occurrences, Western United States:
Economic Geology, v. 65, p. 304311.
Desborough, G.A., Raymond, W.H., and Iagmin, P.J., 1970, Distri
bution of silver and copper in placer gold derived from the
northeastern part of the Colorado Mineral Belt: Economic Ge
ology, v. 65, p. 937944.
Desborough, G.A., Heidel, R.H., Raymond, W.H., and Tripp, J.,
1971, Primary distribution of silver and copper in native gold
from six deposits in the Western United States: Mineralium
Deposita; v. 6, p. 321334.
Desborough, G.A., Raymond, W.H., and Evans, K.V., 1990,
Fine-grained gold in stream sediments of east-central Lemhi
County, Idaho: U.S. Geological Survey Open-File Report
90332, 9 p.
DiLabio, R.N.W., Newsome, J.W., McIvor, D.F., and Lowenstein.
P.L., 1988. The spherical form or gold: Man-made or second
ary: Economic Geology, v. 83, p. 153162.
English, B.L., Desborough, G.A., and Raymond, W.H., 1987. A
mechanical panning technique for separation of fine-grained
gold and other heavy minerals: U. S. Geological Survey Open
File Report 87364.9 p.
 NATIVE GOLD, LEMHI COUNTY, IDAHO
McDanal, S.K., Cathrall, J.B., Mosier, E.L., Antweiler, J.C.,and
Tripp, R.B., 1988, Analytical reseults, geochemical signatures,
mineralogical data, and sample locality map of placer gold and
heavy-mineral concentrates from the Manley Hot Springs,
Tofty, Eureka, and Rampart mining districts, Tanana and Liv
engood quadrangles, Alaska: U.S. Geological Survey Open
File Report 88443, 54 p.
Miller, D.J., Loucks, R.R., and Ashraf, M., 1991, Platinum-group
element mineralization in the Jijal layered ultramafic-mafic
complex, Pakastani Himalayas: Economic Geology, v. 86,
p. 10931102.
Mosier, E.L., 1975, Use of emission spectroscopy for the semi
quantitative analysis of trace elements and silver in native gold,
in Ward, F.N., ed., New and Refined Methods of Trace Analy
sis Useful in Geochemical Exploration: U.S. Geological Sur
vey Bulletin 1408, p. 97105.
273
Mosier, E.L., Cathrall, J.B., Antweiler, J.C., and Tripp, R.B.,
1989, Geochemistry of placer gold, Koyukuk-Chandalar
mining district, Alaska: Journal of Geochemical Explora
tion, v. 31, p. 97-115.
Mosier, E.L. and Lewis, J.S., 1986, Analytical results, geochemical
signatures, and sample locality map of lode gold, placer gold,
and heavy-mineral concentrates from the Koyukuk-Chandalar
mining district, Alaska: U.S. Geological Survey Open-File Re
port 86345, 172 p.
Ruppel, E.T. and Lopez, D.A., 1988, Regional geology and mineral
deposits in and near the central part of the Lemhi Range, Lemhi
County, Idaho: U. S. Geological Survey Professional Paper
1480, 121 p.
Watterson, J.R., 1992, Preliminary evidence for the involvement of
budding bacteria in the origin of Alaskan placer gold: Geology,
v. 20, p. 315318.
 CHAPTER AA
THE IMPORTANCE OF SOURCE ROCKS IN
FORMATION OF METALLIC-SULFIDE
ORE DEPOSITS
By BRUCE R. DOE'
ABSTRACT
Knowledge of the source rocks for carbon is impor
tant in finding economic accumulations of oil and natural
gas. Source rocks are also seriously considered for some
other commodities like uranium, but they are yet to enter
the mainstream of considerations in evaluations of metal
lic-sulfide ore deposits. Five factors need to be consid
ered about source rocks: the abundance of metals in
source rocks; the site of the metals (accessibility); the
kinds of anions available (sulfur, chlorine, ammonia, car
bon dioxide, methane, fluorine) and organic compounds
present; accompanying metals involved (e.g., iron); and
complex sources where metals, anions, and organic com
pounds may come from several geologic units. An appeal
is made for information relating to the importance of
source rocks in the formation of metallic-sulfide ore
deposits.
INTRODUCTION
Although source rocks long have been considered
vital to the formation of petroleum and natural gas accu
mulations, source rocks are generally considered to be
unimportant to the formation of metallic mineral deposits
with certain notable exceptions (e.g., Titley, 1987). In oil
and gas evolution (Tissot and Welte, 1984), source rocks
must contain more than a certain level of carbon in order
to produce commercial crude oil (> 0.5 percent if the car
bon is in the form of algal matter, the most favorable
material). If the organic matter is composed of land plants,
"U.S. Geological Survey, Mail Stop 923, 12201 Sunrise Valley Drive,
National Center, Reston, VA 22092.
the prime product will tend to be natural gas; if marine
plants, then oil most likely will be generated. At low tem
peratures, only biogenic gas will be produced. The tem
perature must reach about 50C for crude oil to be
produced, but, if it is greater than about 150200C, the
oil will break down into thermogenic gas. The temperature
interval favorable for generation of crude oil is called the
oil window. Much can be learned about the economic
potential of a sedimentary basin for oil or gas even from a
dry well.
Initial efforts on various non-sulfidic ores (e.g., sedi
mentary uranium, phosphate, manganese, iron, etc.) give
every indication that source rocks are important to the
development of other natural resources as well. Develop
ment of a theory for the importance of source rocks in
formation of metallic sulfide ore deposits is now timely.
The goals of developing such a theory emphasize the bear
ing of source rocks on formation of super-large, super-rich
metallic sulfide ore deposits and what might be learned
from surface geology, geophysics, and dry holes.
DISCUSSION
A number of aspects need to be considered in evalu
ating the importance of source rocks in the formation of
metallic sulfide ore deposits:
1. Metal-enriched sources.The prime targets in oil
exploration are the largest deposits or supergiant fields.
Thus, shales are the prime source rock of interest because
they can be carbon rich. Limestones may produce valuable
quantities of oil, but shale should yield the largest deposits.
For metallic sulfides, we can expect a similar relationship.
Especially for ore deposits related to igneous processes (the
magmatothermal case, where fluid amountseither mag
matic fluids or circulating hydrothermal cellsand
276 ADVANCES RELATED TO U.S. AND INTERNATIONAL MINERAL RESOURCES, 1992
duration of hydrothermal systems are limited), formation of
supergiant ore deposits should be favored by enriched
sources. The reasoning is simple. At 1 ppm of an element,
there are only 10,000 tons of an element in a cubic mile of
rock. Therefore, 100 mi of rock that contains this element
at a concentration of 1 ppm would have to be extracted to
get a deposit yielding 1 million tons of metal. But the
largest deposits would require a considerably greater rock
volume. Broken Hill, Australia, will produce more than 20
million tons (20 Mt) of zinc, which would require more
than 2,000 mi with only 1 ppm extraction. It stands to
reason that formation of a supergiant ore deposit would
be easier at the 10 or 100 ppm level than at the 1 ppm level.
More subtle concentration fluctuations may also be para
mount. For example, a parent magma that starts with twice
the copper content of a normal magmatic system may
yield a porphyry copper deposit of 0.45 percent rather than
0.23 percent. The former may well be economic to mine,
the latter will not. Thus, we may have to look for small
enrichments in concentrations.
In my experience, most people who feel that source
rocks are not important in the generation of major ore
deposits feel that way because they imagine that the
source being considered is just a lower grade mineraliza
tion, i.e., a regeneration of older deposits (these people
oppose the geostill hypothesis, where ore fluids are
either directly derived from subducted, metal-rich, exhala
tive sediments or indirectly from magmas in which such
sediments are a component). There is no doubt that nature
does form some ores through a succession of enrichment
stages (a bootstrap operation)supergene enrichment of copper, a compatible element, becomes too depleted in the
porphyry copper may be among the least controversial
examples. Indeed, supergene enrichment is often more
important for putting the metal into a form in which the
ore may be economically processed (e.g., by heap leaching
of oxides and carbonates rather than sulfides) than it is for
enriching concentrations; however, element enrichment is
also important (e.g., gossan gold). The importance of
regeneration certainly needs to be considered.
2. Site of the metals.In addition to the abundance
of an element in the source rock, the location of the ele
menti.e., the accessibility of the elementwithin the
rock can also be important. This factor is recognized as
most important in the generation of uranium ore deposits.
Uranium in a major uranium-bearing phase like zircon is
relatively inaccessible to leaching, whereas much uranium
is often in cryptic, easily-mobilized phases. In sediments
containing normal concentrations of uranium, volcanic
glass shards, from which uranium is readily mobilized,
have been proposed as the source of uranium in some
major sandstone uranium deposits (Zielinski, 1983). Thus,
unusual concentrations of an element may not be required
if the element is readily available to the ore fluid.
3. Anions.There is little doubt that anions can play
a predominant role in ore formation. For metallic-sulfide
ore deposits, a source of sulfur is clearly needed. Indeed,
the reason that a major deposit has not been found in the
Salton Basin that was deposited from the hot, metal-rich,
chloride brines owes to a lack of sulfur. But other anions
are also important (e.g., chlorine, carbon dioxide, ammo
nia, fluorine), and organic compounds are often mentioned
as being important for ore formation. Chlorine can be of
paramount importance in complexing metals to more solu
ble form. For example, hot, chloride-rich brines (e.g., Sal
ton-Sea brine) can carry metals that are enriched in the
1001,000 ppm range. These brines also are highly reac
tive: They will acquire the necessary concentrations of
heavy metals even from metal-poor source rocks if the
high solubility of these metals can be maintained either
through remaining hot or through decreasing pressure
(Hemley and others, 1992) over a long flow path. The
source of chlorine in such brines seems to be some combi
nation of connate water and evaporites because the sulfur
isotopic composition (634S) is usually heavy (as expected
for sulfur derived from anhydrite in evolved sea water and
evaporites). Thus a dry hole in a sedimentary basin that
shows the presence of evaporites might still be encourag
ing for the presence of mineral deposits.
Philip Candela and colleagues at the University of
Maryland have stressed the importance of both water and
chlorine in the source in order to produce certain kinds of
ore deposits (e.g., Candela and Holland, 1986). They feel
that a porphyry copper results from incorporating a
hydrous source in the melt so that a chloride-rich fluid
phase separates early in magmatic differentiation before
melt. Paucity of water in a source, such that a fluid phase
might form late in magmatic differentiation, might pro
duce a different kind of mineralization for an incompatible
element such as molybdenum (i.e., molybdenum por
phyry). Thus, one might deduce that amphibolite in the
lower crust may favor copper porphyry, and granulite may
favor molybdenum porphyry. Firming up such deductions
may not be beyond our abilities. In some cases, the nature
of the lower crust may be deduced from surface geology,
in other cases from seismology or isotopes.
4. Accompanying metals.-The metals accompany
ing those elements of economic interest may also play an
important role in ore formation. The importance of the
cationic elements not of economic interest (e.g., iron) is
well known. Although copper is as abundant as zinc in
most mafic source-rock candidates, the richest porphyry
copper ores rarely reach 3.5 percent, whereas zinc deposits
have supergiant representatives that average 17 percent
(e.g., Broken Hill, Australia, and Red Dog, Alaska).
These deposits also tend to contain greater ore grades for
lead (about 5 percent) than the best copper ores, even
though lead is almost certainly much less abundant in the
source rocks. A part of this discrepancy may relate to
discussions by Hemley and others (1992), who point out
 FORMATION OF METALLIC-SULFIDE ORE DEPOSITS
that iron is relatively common in ore fluids. Iron readily
combines with copper to form only a moderately rich cop
per mineral, chalcopyrite (CuFeS2), rather than a richer
equivalent, such a bornite (Cu5FeS4). Thus, an iron-poor
source might favor richer copper deposits, such as the cop
per-rich Colorado Pipe in Mexico, although relations of
sulfur and oxygen in the ore fluid can be important also.
5. Complex sources.Although there are many anal
ogies between petroleum and metallic-sulfide deposits,
there are also clear differences. Metals may be acquired
from several sources and from over the total path length of
the ore fluid; yet, petroleum may have mixed sources too.
Both require permeability and some sort of trap (structural
control). Both may involve a large range of temperatures
of-generation (e.g., biogenic gas and thermal gas; red-bed
copper deposits and Mississippi Valley Zn-Pb deposits; or
porphyry copper deposits and massive sulfide Cu-Zn-Pb
deposits). Both can be spoiled by gangue, which for
crude oil is the presence of hydrogen sulfide or carbon
dioxide. Petroleum formation, however, has nothing simi
lar to important controls on ore deposition, such as: con
centration by solution cooling or neutralization; the need
for two or more components such as metal and sulfur,
which can be derived from separate sources; and perhaps
some complexing agent such as chlorine that might come
from yet a third source. The magmatic differentiation
escalator, which can be important in enriching metals, is
not involved in crude-oil generation. To what extent all
these aspects can be sorted out remains to be seen.
There are no illusions about the difficulty of what is
being undertaken. Many metallic-sulfide ore deposits, for
example, are polymetallic without any economic abun
dances in any one metal taken by itself. Also, hot, chlo
Published in the Central Region, Denver, Colorado
Manuscript approved for publication October 26, 1992
Edited by Richard W. Scott, Jr., Branch of Central Technical Reports
Graphics prepared by Dick Walker and M.A. Marty Simmons,
Branch of Geochemistry; some author drafting used
Assistance with manuscript preparation by Margo Johnson and
Barbara Ramsey, Branch of Geochemistry
Type composed by Shelly A. Fields, Branch of Central Technical Reports
Cover design by Art Isom, Branch of Central Technical Reports
r U.S. GOVERNMENT PRINTING OFFICE: 1992774-049/66,044. REGION NO 8
277
ride-rich brines will acquire metals from anything through
which they pass, so the problem in forming a major ore
deposit may be more one of acquiring the chloride and
keeping the solubility up over long flow paths than the
need for metal-rich sources. The author would like to hear
from you concerning interesting references and any ideas
you might wish to relate concerning the importance of
source rocks in the formation of metallic-sulfide ore
deposits. In return, he will send you a copy of his paper
(Doe, 1991) upon request.
REFERENCES CITED
Candela, P.A., and Holland, H.D., 1986, A mass transfer model for
copper and molybdenum in magmatic hydrothermal systems
The origin of porphyry-type ore deposits: Economic Geology,
v. 81, p. 117.
Doe, B.R., 1991, Source rocks and the genesis of metallic mineral
deposits: Global Tectonics and Metallogeny, v. 3, p. 1319.
Hemley, J.J., Cygan, G.L., Fein, J.B., Robinson, G.R., and d'An
gelo, W.M., 1992, Hydrothermal ore-forming processes in
the light of studies in rock-buffered systemsI. Iron-copper
zinc-lead sulfide solubility relations: Economic Geology, v.
87, p. 122.
Tissot, B.P., and Welte, D.H., 1984, Petroleum formation and oc
currence (2nd ed.): Berlin, Springer-Verlag, 699 p.
Titley, S.R., 1987, The crustal heritage of silver and gold ratios in
Arizona ores: Geological Society of America Bulletin, v.99, p.
814826.
Zielinski, R.A., 1983, Tuffaceous sediments as source rocks for
uraniumA case study of the White River Formation, Wyo
ming: Journal of Geochemical Exploration, v. 18, p. 285306.
 Cr t, "....
n 3. ls
L f
TH E OH!
H!C) STATE Ui, VE::** -> ! Y.
* --- ** ...A. 292.9
* 55 & CA A. DR"JF.
53. pate Du
 The Oh Uni

III.iii.|
3 2435 056588

CHECK FOR 5 PART

H-o-o-s-ATE unive-s-L-L-BRARIEs