Vous êtes sur la page 1sur 22

Bull Volcanol (2009) 71:859–880

DOI 10.1007/s00445-009-0271-0

RESEARCH ARTICLE

Eruptive style of the young high-Mg basaltic-andesite


Pelagatos scoria cone, southeast of México City
Marie-Noëlle Guilbaud & Claus Siebe &
Javier Agustín-Flores

Received: 5 March 2008 / Accepted: 2 March 2009 / Published online: 31 March 2009
# Springer-Verlag 2009

Abstract The eruption of the Pelagatos scoria cone in the elevated crystal contents at T < 1,100°C, and blocky
Sierra Chichinautzin monogenetic field near the southern surfaces. Later, the closure of the fissure by cooling dikes
suburbs of Mexico City occurred less than 14,000 years focused the magma flow at a narrow section of the fissure.
ago. The eruption initiated at a fissure with an effusive This led to an increased magma ascent velocity. Rapid and
phase that formed a 7-km-long lava flow, and continued shallow degassing (<3 km deep) triggered ~40 vol.%
with a phase of alternating and/or simultaneous explosive microlite crystallization. Limited times for gas-escape and
and effusive activity that built a 50-m-high scoria cone on higher magma viscosity (6×105–4×106 Pa s) drove strong
the western end of the fissure and formed a compound lava explosions of highly (60–80 vol.%) and finely vesicular
flow-field near the vent. The eruption ended with the magma. Coarse clasts broke on landing, which implies
emplacement of a short lava flow that breached the cone brittle behavior due to complete solidification. This requires
and was accompanied by weak explosions at the crater. sufficient time to cool and in turn implies ejection heights
Products consist of a microlite-rich high-Mg basaltic of over 1 km, which is much higher than “normal”
andesite. Samples were analyzed to determine the magma’s Strombolian activity. Hence, magma viscosity significantly
initial properties as well as the effects of degassing-induced impacts eruption style at monogenetic volcanoes because
crystallization on eruptive style. Although distal ash fallout it affects the kinetics of shallow degassing. The long-
deposits from this eruption are not preserved, a recent lasting eruptions of Jorullo and Paricutin, which produced
quarry exposes a large section of the scoria cone. Detailed similar magmas in western México, were more explosive.
study of exposed layers allows us to elucidate the mode of This can be related to higher magma fluxes and total
cone-building activity. Petrological and textural data, erupted volumes. Implications of this study are important
combined with models calibrated by experimental work because basaltic andesites are commonly erupted to form
and melt-inclusion analyses of similar magmas elsewhere, monogenetic scoria cones of the Trans-Mexican Volcanic
indicate that the magma was initially hot (>1,200°C), gas- Belt.
rich (up to 5 wt.% H2O), crystal-poor (~10 vol.% Fo90
olivine phenocrysts) and thus poorly viscous (40–80 Pa s). Keywords Monogenetic volcanism . Basaltic andesite .
During the early phase, low magma ascent velocity at the Sierra Chichinautzin . Strombolian activity . Scoria cone .
fissure vent allowed low-viscosity magma to degas and Pelagatos . Trans-Mexican Volcanic Belt
crystallize during ascent, producing lava flows with

Introduction
Editorial responsibility: J. White
M.-N. Guilbaud (*) : C. Siebe : J. Agustín-Flores Because of their small size, simple morphology (e.g., Porter
Departamento de Vulcanología, Instituto de Geofísica, 1972; Wood 1980), and mafic to intermediate composition,
Universidad Nacional Autónoma de México,
it is often not recognized that scoria cones can display a
Ciudad Universitaria,
Coyoacán, México D.F., México wide variety of eruptive styles. These styles can range from
e-mail: m.guilbaud@geofisica.unam.mx relatively harmless Hawaiian lava fountaining (Head and
860 Bull Volcanol (2009) 71:859–880

Wilson 1989) to short-reaching ballistic Strombolian bursts cones of such composition are common in the Trans-
(McGetchin et al. 1974), up to violent-Strombolian or sub- Mexican Volcanic Belt.
Plinian explosions with high eruptive columns and consid-
erable ash-and-gas emissions (Valentine et al. 2005;
Valentine et al. 2007; Valentine and Gregg 2008; Pioli et Background and hazards
al. 2008). Such variations are thought to arise from the
sharp changes that rapidly-decompressed gas-rich magmas The Trans-Mexican Volcanic Belt (TMVB) is a volcanic
experience at shallow levels. Of these, degassing-induced arc that is related to the subduction of the Cocos Plate
crystallization is believed to strongly affect magma viscos- underneath the North American Plate (insert in Fig. 1a).
ity, which may ultimately control the intensity of magma Besides stratovolcanoes, calderas and domes, the TMVB
fragmentation (e.g., Taddeucci et al. 2004; Polacci et al. includes more than 3,000 scoria cones of Quaternary age.
2006; Sable et al. 2006; Lautze and Houghton 2007; Pioli Scoria cones are not distributed evenly in the TMVB but
et al. 2008). Other factors that intervene in the eruptive form clusters or “fields”. The most prominent of these
style are the inevitable evolution of the feeding system and fields is the ~40,000 km2 Michoacán-Guanajuato volcanic
vent geometry during the eruption. For example, in long- field that consists of ~900 Quaternary edifices including
lasting eruptions, a conduit may develop within the cone the historical Paricutin (1943–1952) and Jorullo (1759–
(e.g. Keating et al. 2008), promoting efficient magma–gas 1774) cones for which eyewitnesses accounts exist (e.g.,
segregation (Pioli et al. 2008), or a shallow reservoir may Foshag and Gonzalez 1956; Hasenaka and Carmichael
form underneath it (Valentine and Krogh 2006), favoring 1985; Luhr and Simkin 1993). The Sierra Chichinautzin
some degree of crustal contamination and enhanced Volcanic Field (SCVF) borders the south of México City,
shallow degassing (McBirney et al. 1987; Johnson et al. one of the largest cities in the world and the economic and
2008). Many variables are involved in the system, which political capital of Mexico. It consists of more than 220
makes these processes still not well understood. monogenetic vents, most of which are scoria cones
This paper focuses on a recent (<14,000 years B.P.) (Bloomfield 1975; Martin Del Pozzo 1982; Márquez et
monogenetic volcano that formed within the potentially al. 1999). Radiocarbon dating of young volcanic products
active Sierra Chichinautzin volcanic field (SCVF) border- indicates that a minimum of eight scoria cone eruptions
ing Mexico City. The eruption featured a main, fissure-fed occurred within the SCVF during the last 10,000 years
effusive phase and a later explosive phase that formed a (Siebe et al. 2005). The last eruption occurred at Xitle
small scoria cone at one end of the fissure. Products consist ~1,670 years B.P. (Siebe 2000). This indicates that the
of a phenocryst-poor and microlite-rich basaltic andesite, SCVF should be considered potentially active. Future
which allows study of the effects of magma viscosity and monogenetic activity in this area would pose important
shallow crystallization on the eruptive style of a small hazards (Siebe et al. 2004a, 2005). The compositional
monogenetic volcano. Although distal ash fallout deposits range of eruptive products and the distribution of volcanic
from this eruption are not preserved, a recent quarry centers within the SCVF are roughly known (see: Mooser
exposes a large stratigraphic section of the scoria cone. et al. 1974; Bloomfield 1975; Martin Del Pozzo 1982;
Easy access to the interior beds of the cone allowed study Márquez et al. 1999; Wallace and Carmichael 1999;
of the mode of cone-building activity. Cervantes and Wallace 2003; Siebe et al. 2004a, 2004b,
After a brief review of potential hazards from monoge- 2005; Schaaf et al. 2005). Yet, much more needs to be
netic activity in the Trans-Mexican Volcanic Belt, the understood about the age of the volcanoes and their
products from the Pelagatos are described, and field, eruptive behavior. To partly fill this gap, this study
grain-size, grain-shape, and density data of clasts in scoria analyzed in detail the well-preserved and well-exposed
beds exposed throughout the main cone are presented. In products of the young Pelagatos scoria cone eruption in
addition, compositional, petrographical, and textural data the eastern part of the SCVF. Products of the Pelagatos are
are considered in order to reconstruct the eruption chronol- among the most primitive of the SCVF (see sample CHI-
ogy, interpret the mode of cone-building activity, and 70 of Márquez and de Ignacio 2002, sample 96365 of
discuss the changes that the magma experienced during Schaaf et al. 2005; this study). They share the singular
the course of the eruption and how these changes characteristics of high-Mg andesites, which are high
influenced eruptive style. This eruption is then compared contents of compatible elements (MgO, Ni, Cr, low FeO/
with the historically documented eruptions of Jorullo and MgO) and the occurrence of Mg-rich olivine phenocrysts
Paricutin in western México, both of which involved with Cr-spinel inclusions (e.g. Tatsumi and Ishizaka
similar magma types. Peculiarities of the eruptive style of 1982). Results of this study shed light on the processes
magmas of intermediate composition (basaltic andesite) are that govern monogenetic eruptions of phenocryst-poor
highlighted. This is of particular importance because scoria basaltic andesite magmas.
Bull Volcanol (2009) 71:859–880 861

Location, distribution, morphology, and age basal width) of 0.14–0.17 was calculated for the cone. This
of Pelagatos products is close to the typical ratio of 0.18 reported by Porter (1972)
for young cones. A section through the south-western flank
Pelagatos is located 30 km west-northwest from Popocatépetl of the cone is exposed by a quarry at location N19°08′33.0″
stratovolcano and 20 km southeast from Mexico City W98°58′17.1″. A blocky lava flow issued from a high level
(Fig. 1a). The area includes several young cones and on the cone and now covers part of the eastern cone flank.
associated lavas that are forested by pine trees, and older It may be a crater overflow or a rootless flow formed by
cones and lavas that are overlain by cultivated soils coalescence of pyroclasts landing hot on the cone flank
developed on tephra layers (Fig. 2). The Pelagatos eruption (e.g. Head and Wilson 1989). Direct evidence for the latter
formed a main scoria cone, two scoria ridges, and an is however missing and welding is a rare feature within the
elongated lava flow-field which filled a depression between deposits. The easternmost scoria ridge (600 m from the
cones and lavas pre-dating the eruption (Figs. 1b and 2). main cone) is ~180 m long and <40 m high, and is curved
The main cone and the two scoria ridges are aligned on the eastern end, displaying a half-moon shape. Scoria
roughly in an E–W direction (Fig. 1b). The main cone is mounds <1 m in height and several meters in length were
slightly asymmetric with an irregular crater rim. It is ~50 m found at the base of the eastern margin of the ridge on the
high, 300–350 m in diameter at its base, and its south side downflow side. These mounds are aligned parallel to flow
is breached. With this data a geometric ratio (cone height/ direction. The scoria ridge located 400 m east from the

Fig. 1 a Location of Pelagatos


(<14,000 years B.P.) and Xitle North American Plate MEXICO CITY
(~1670 years B.P.) scoria cones
in the SCVF. Note the growth of
Mexico City to the south, in- TMVB
Xitle 19˚15'
vading the volcanic field. Inset:
map showing the location of the
Sierra Chichinauzin Volcanic
Field (SCVF) in the subduction- SCVF Pelagatos Popocatépetl
related Trans-Mexican Volcanic Cocos
Plate 19˚00'
Belt (TMVB). Major stratovol-
500 km
canoes are shown by filled tri-
angular symbols in the inset. b
Detailed map of the Pelagatos SCVF
cone, scoria ridges, and lava
flow-field. Dots indicate lava 18˚45'
sampling locations. The location 30 km
and name of older cones that
diverted the Pelagatos lava flow a 99˚20' 99˚00' 98˚40'
are also indicated
exic lta
ity

inferred fissure
oC
To M ilpa A

2900 m altitude above sea level


To M

m
00
29
main PG18 m
scoria 00
cone 27
hitepe
c
2800 m

ridges to Juc
quarry shea
P67 r zon PG16
PG10 EARLY es
P59
LATER FLOW
PG17
FLOWS
P61 PG15
PG13
Volcán
ec

Cohuazalo
tep
Oax

Volcán
Ahuazatepetl

Volcán Huehuel Volcán Huehuelcon 1 km


b
862 Bull Volcanol (2009) 71:859–880

Cerro Del Agua cone Cilcuayo cone Tláloc cone


rich in plagioclase phenocrysts and therefore quite different
from the olivine-bearing Pelagatos products. It seems that
Pelagatos main cone ash fallout deposits from Pelagatos were relatively thin (few
Cilcuayo lava shield cm?) and easy to erode, given that older ash deposits
belonging to other nearby cones are tens of cm thick and
well-preserved.
The Tutti Frutti plinian pumice from Popocatépetl dated
Pelagatos at ~14,000 years B.P. (Siebe et al. 1999) forms an important
lava field
marker horizon in the area. Its absence over the products
from Pelagatos indicates that the Pelagatos eruption is
younger than 14,000 years B.P. The base of the Pelagatos
lava and its contact with the underlying paleosol is not
Volcán
Ahuazatepetl Volcán exposed. This makes it difficult to obtain a conclusive
Cohuazalo radiocarbon age for the eruption. Carbon found in reworked
products ponding against the margin of the lava flows
(sample P60, Fig. 1b) was dated at 2,520±105 years B.P. at
the Radiocarbon Laboratory, Department of Geosciences,
Volcán Tlapexcua Tucson, Arizona. This date represents only a minimum age
for this eruption. Hence, a satisfactory radiocarbon age is
Fig. 2 Photograph towards the west taken by C.S. during a helicopter still pending.
flight on Dec. 29, 1994. The forested Pelagatos lava flow-field and
adjacent older farmed scoria cones are clearly discernible

Physical characteristics of scoria cone deposits


main cone is ~60 m long, ~10 m high, and slightly curved.
The top of the ridge is made of locally exposed layers of A recent quarry cuts an oblique section through the SW
partly-welded reddish scoria. flank of the main cone, exposing its uppermost layers
The lava flow field is ~7 km long and 0.5 to 1 km wide (Fig. 3). The NW–SE oriented face of the quarry displays
(Figs. 1b and 2). It comprises a main flow with lateral shear the outer-wall facies of the cone and a small outer part of
zones and lobate margins. The proximal 1.5 km of the flow the crater fill overlying the crater wall unconformity
is covered by a superposition of smaller flows that form a (Fig. 3; the terms used here are defined in figure 5 of
distinct platform banking against the scoria ridges to the Houghton and Schmincke 1989). Note that it does not cut
north (“later flows” in Fig. 1b). Spatial relationships down to the base of the cone which lies about 20 m farther
suggest that the main flow came early from a fissure vent down. The crater fill consists of coarse beds (clasts >1 cm
stretching from the main cone to the easternmost scoria in diameter, little ash) of (1) small (<10 cm), loose, dark-
ridge. Late flows were apparently emitted from the western grey to bluish clasts and (2) larger, viscously-deformed,
end of the fissure, and post-dated the formation of the often weakly welded (partly sintered), pink oxidized clasts
scoria ridges. Some may pre-date the formation of the up to 30 cm in length. By contrast, the outer-wall facies is
scoria cone. The last flow emerges from the cone’s breach entirely composed of loose scoria clasts which are fresh,
and its source is covered by bombs. dark, glassy, and often have an iridescent sheen, except in
The area covered by the flows was measured at 4.9 km2 the upper 2 m of the section where clasts are yellow, due to
(estimated by importing a contour of the flow field drawn surface weathering. Locally, they are partly covered by a
on the computer into Scion Image processing software), white coating (mostly amorphous silica) that was apparent-
and 5.2 km2, using a digital planimeter and a 1/50,000 ly precipitated by fumarolic steam from rainwater percolat-
topographic map. Flows are 7.5 m-thick on average, giving ing through the still hot cone. Some clasts (<1 vol.%)
a bulk lava volume of 3.7–3.9×107 m3 and 3.1–3.5×107 m3 contain white, round to rectangular quartz aggregates of
DRE (average vesicularity of 10 vol.%). The bulk volume metamorphic origin, which often display a thin reaction rim
of the main cone is 1.7×106 m3 (Márquez et al. 1999) and of vesicular glass and clinopyroxene microlites.
0.68×106 m3 DRE (average vesicularity of 60 vol.%). The The following paragraphs describe the outerwall facies.
cone scoria thus represents only ~0.02 vol.% of the total Terminologies used are those by White and Houghton
erupted DRE volume (neglecting the scoria ridges). It is (2006) and Houghton and Wilson (1989) and describe the
worth mentioning that no ash blanket related to this ranges in grain-size and vesicularity, respectively.
eruption was found extending beyond the vent area. A Fine-grained layers (coarse to very coarse ash) form
>1 m-thick ash layer cropping out at nearby exposures is distinct horizons in the cone section. Most of these layers
Bull Volcanol (2009) 71:859–880 863

define the base of the inversely-graded part of distinct beds collapses contribute to the rapid growth of an important
(see below). They were traced on a mosaic of photographs basal talus (Fig. 3).
of the quarry wall to image the structure of the cone. The
resulting representation (Fig. 3) reveals the slight asymme-
Clast morphologies and textures
try of the cone expressed by the location of the crater
unconformity 1–2 m below and a few meters to the east of
Clast morphologies and textures can reveal diagnostic
the point of highest elevation. Also, south-dipping beds are
features indicative of eruptive conditions. Microscopic
generally thicker than north-dipping beds and show clear
characteristics were observed under the scanning electron
downslope thickening. These two features can be explained
microscope (SEM) operated at the Laboratorio Universi-
by the construction of the cone on the older lava field of
tario de Petrología at UNAM, México.
nearby Cilcuayo volcano that slopes to the east (Fig. 2).
The different clast types found in the deposit are the
This situation promoted instability and avalanching of
following:
scoria forming beds deposited on the south side of the
cone. The paucity of faults deforming the beds seems to be 1. Large bombs (>10 cm in diameter; largest bomb: 80 cm
typical of deposits from outer-wall cone facies (Houghton long and 30 cm thick). These bombs are often broken,
and Schmincke 1989) and can be attributed to the granular exposing interiors with a coarsely-vesicular microcrystal-
nature of the material. Note that here the loose nature of the line core cut by curvi-planar fracture surfaces covered by
deposit makes the quarry face highly unstable. Frequent a rough, irregular, 0.2 to 0.5 cm-thick finely-vesicular

crater wall
NW unconformity crater fill
coarse deposit consisting of blue SE
highest point
to reddish, loose to partly-welded,
angular to contorded scoria clasts layers are continuous
obscured by rainwashed soil across vegetated area
Fig. 9b
weathered top

faults
fault
~7 m thick high concentration Fig. 9c
of bombs

led
mp
10 m

sa
nd
coarse layers
TALUS

da
thickening downslope

ge
log
on
5m
cti
se
Fig. 9a
Fig. 8

Fig. 3 Internal structure of the main Pelagatos cone as exposed in a on the sketch. Field observations indicate that they correspond to the
large quarry. The quarry wall runs at ca. N140° (~NW–SE) and cuts base of each bed. Areas contoured with a broken line denote zones
across the SW flank of the cone. It does not expose the interior of the where the quarry wall is hidden behind bushes and small trees, or
central part of the cone nor its base but its outer and uppermost layers. coated with mud. Note that distances are not to scale but distorted
The sketch was constructed by outlining preferentially eroded fine- because the drawing is based on a photo mosaic and photos were
grained horizons on a mosaic of high-resolution photographs of the taken from a point located closer to the right side of the quarry
quarry wall shown below. These horizons are shown as fine dark lines
864 Bull Volcanol (2009) 71:859–880

Fig. 4 Photographs of different


clast types. a Large bomb in
coarse bed. Note broken interior
and coarsely-vesicular core. b
Composite bomb in finer-
grained bed. Bomb consists of
distinct clasts welded together.
Outer crust is thin, convoluted,
and coated by lapilli-size scoria.
c Angular vesicular clast with
thin crust on two sides and
coarsely-vesicular core bounded
by curvi-planar surfaces. d An-
gular vesicular clast with rough, a b
glassy, highly-stretched light-
brown outer crust on one side.
Note vesicle coarsening from
crust to core. e Rough vesicular
clast with ragged, highly-
stretched outer crust surround-
ing the clast. f Dense broken
clast with rounded smoothed
shape, and finely-vesicular core.
Note small size (<1 mm) and
deformation of vesicles, in
comparison to highly-vesicular
clasts

c d

5 cm
e f

glassy crust (Fig. 4a). Some are composite, i.e., 70 vol.%) clasts that have a coarsely-vesicular
composed of distinct clasts welded together (Fig. 4b). core bounded on one to three sides by a ragged
Large, single bombs have dense cores that were viscously-stretched crust (Figs. 4c and d). They
completely degassed before solidification. Most bombs are similar to broken-up parts of bombs (i.e.,
are elongate and parallel to bedding surfaces, but a few Fig. 4a). In thin-sections of glassy crust, vesicles
are in an oblique position. Their abundance and size occur in a bimodal distribution. The first popula-
decrease laterally from the cone summit downslope, tion is <40 µm-across and consists of small
which can be related to their progressive break-up upon isolated vesicles. The second consists of larger
rolling down, and the shorter flight distances of large interconnected coalesced vesicles (Fig. 5a). The
bombs. Young cones sometimes display a basal ring proportion of each population ranges from 40 to
made of large bombs that rolled down the slope without 60% on back-scattered electron (BSE) micro-
breaking (e.g. McGetchin et al. 1974; Heiken 1978), a scope images. Vesicles are notably coarser in the
feature not observable at Pelagatos. core, where they form a tight network of round to
slightly deformed mm-sized vesicles with larger
2. Coarse clasts (16 mm–10 cm):
cm-sized coalesced vesicles (Fig. 5b).
2.1. Coarse, angular vesicular clasts (~80–90 vol.% of 2.2. Rough vesicular clasts (~10–20%) are irregularly-
coarse clasts) are moderately-vesicular (60– shaped vesicular clasts (70–80 vol.% vesicle
Bull Volcanol (2009) 71:859–880 865

glassy crust core

500 mm 1 mm

Fig. 5 Textures of vesicular angular clasts. a Back-scattered electron lites are abundant in groundmass. b Scanning electron microscope
(BSE) microscope image of thin-section of fragment of glassy outer (SEM) image of fragment of coarsely-vesicular scoria core (sample
crust (sample P13). Note long coalesced vesicles and smaller, isolated P01). Note larger size of vesicles compared to crust samples. Large
vesicles. Vesicles are round to slightly elongated. Plagioclase micro- coalesced vesicles are surrounded by smaller, isolated vesicles

content) with extremely rough-jagged surficial to abundant <10-µm-size crystallites exposed by


glassy crusts that are highly-stretched at a mm-to- fracture surfaces (Fig. 7f). Note that there is a
cm scale and micro-vesicular (Fig. 4e). continuous gradation between sideromelane and
2.3. Angular dense clasts (<5%) are randomly- tachylite types (e.g. Heiken 1978).
distributed, poorly-vesicular (5–20 vol.%) clasts 3.3. Olivine phenocrysts with thin glass veneers (1–3%).
with smooth, rounded surfaces sometimes cut by
5–10 cm-spaced linear fractures. They are often
Clasts between 1 and 16 mm have characteristics
broken into very angular fragments that form
intermediate between coarse clasts and fines. They consist
jigsaw pieces of fusiform bombs or are randomly
of (1) angular crystalline tachylite-type clasts (Fig. 7d), (2)
dispersed in the beds. Cores often display <1 mm
irregular, rough, glassy sideromelane-type clasts (Fig. 7a),
vesicles stretched parallel to the outer clast
and (3) composites with both groundmass types. A few
surface (Fig. 4f).
dense clasts observed in the −4Φ fraction of sample P46
3. Fines (<1 mm):
display finely-vesicular mm-scale cauliflower-shaped sur-
3.1. Sideromelane clasts (65–95% of fines, see below) faces. Only two samples of platy clasts with fine-ash
are light-brown, delicate, irregularly shaped, coatings were found. These were 5 cm-long, 0.5 cm-thick,
translucent and frothy (70–90 vol.%). In thin-
section, they display a glassy matrix and dense,
small, round to elongated smoothly-shaped
vesicles, some of which are coalesced (A in
Fig. 6). On SEM images, they display smooth B
curved glass surfaces (Fig. 7b), tubular textures Ol Ol
(Fig. 7c), thin (<10 µm-across) highly stretched
Ol
glass filaments, and vesicles up to 2 mm in
diameter exposed by fracture surfaces (Fig. 7b).
C
3.2. Tachylite clasts (5–35%) comprise dark-brown to
black, angular, opaque clasts generally less vesicu-
lar than sideromelane (60–70 vol.%). In thin-
section, they show a devitrified to cryptocrystalline A
500 µm
matrix, and fewer but larger, coalesced vesicles than
in sideromelane clasts (B and C in Figs. 6, and 7e). Fig. 6 Photograph of ash fragments under the polarizing microscope.
On SEM images, they display slightly-irregular a Glassy sideromelane type with delicate outer surface and microlites.
b, c Tachylite fragments showing devitrified glass texture, large
glass surfaces with sparse to abundant µm-to-mm coalesced vesicles, and cluster of olivine microphenocrysts and laths
scale fractures arranged in parallel sets, and sparse of plagioclase microlites
866 Bull Volcanol (2009) 71:859–880

Fig. 7 Scanning electron microscope images of scoria particles. Side- degree of coalescence of vesicles compared to (a). e Moderately-
romelane type: a Highly-vesicular glassy clast 8 mm-across; note vesicular droplet broken on one side; fracture surface marked by µm-
irregular, delicate surface and small size and dense packing of vesicles. size striae; note low amount and size of vesicles compared to (b). f
b Highly-vesicular 1 mm-across droplet broken on one side. c Close-up Close-up of microcrystalline surface of clast shown in (d); angular
of viscously-stretched flexible glassy skin formed on contact with equant crystals are probably olivines, and tiny laths are plagioclases;
atmosphere (Heiken 1972). Tachylite type: d Coarsely-vesicular clast, surface draped by flexible glassy skin
8 mm-across; note angular outer surface and larger size and higher

finely-vesicular, and had surface fractures resembling mud processed separately (sample pairs: P33 and P34, P31 and
cracks. P32, and P47 and P48).
It is worth noting that all clasts were solid when The sequence may be subdivided into three main units
deposited. Only one bomb showed signs of having on the basis of bed thickness and coarseness (Fig. 8).
deformed plastically upon impact. This bomb had a pink Transitions between units are gradual.
(Fe-oxidized) crust on which small clasts were adhered by Unit A consists of laterally-discontinuous, often inter-
sintering. nally heterogeneous, thin to moderately thick (<70 cm),
poorly-sorted (σΦ 1.75–1.93), medium lapilli (Md −2.2 to
Grain-size, vesicularity, and componentry variations −5.1Φ) beds (Figs. 8 and 9a) that often have a lenticular
shape with lateral pinching edges and undulating bound-
The sequence of beds forming the SE side of the quarry aries (Fig. 9a). Coarse clasts are concentrated in the middle
(Fig. 3) was studied in detail in the field. Samples from the part of the beds, while the base and top of the beds are
lower side of the section were accessible by standing on the distinctly finer-grained (Fig. 9a).
upper part of the talus and analyzed in the laboratory Beds of Unit B are generally thicker (60–300 cm) and
(Fig. 8, analytical methods applied and results are included have better-sorted (σΦ 0.9–1.7), coarser interiors (Md −4.8
in Appendix 1). An arbitrary distinction was made between to −5.5Φ) (Figs. 8 and 9b). They are laterally continuous
thick (>60 cm) and thin (<60 cm) beds. Individual beds over the entire cone’s slope, with the exception of the
present a distinct inversely-graded fine-grained base that uppermost bed of the unit, which transforms laterally into a
grades from coarse ash to fine-to-medium lapilli upwards, sequence of discontinuous thin beds (Fig. 3). Beds have
and is overlain by the main part of the bed (medium to bases consisting of thin (<10 cm) reversely-graded fine-to-
coarse lapilli and few bombs). The main part of the beds medium lapilli (samples P47, P31, and P33) and main parts
often displays normal grading at the top (see below). The composed of coarse lapilli sometimes fining to medium
fine base is best developed at thick beds. In three cases, the lapilli toward the top (Fig. 9c). Internal heterogeneities
main part and fine base of the bed were sampled and include fine-lapilli horizons that pinch out laterally, and
Bull Volcanol (2009) 71:859–880 867

compositional
samples for

analysis
KEY

Weathered top
1800
medium to coarse lapilli scoria coarse interior of thick bed
fine lapilli scoria fine-grained base of thick bed
truncated lapilli horizons
thin bed (<60 cm)

1600 30 P36
40
P36
P23 30
20
20
10

crude
internal
C 0

30 P35
10

0
50 55 60 65 70 75 80 85 90
P36 bedding P36 20
10
0
P21

average of all
individual 50 P51
P35 40
1400 P35 thin bed 30
20
10
P19 crude P51 0
P51 internal

25 vol.%
40 P50
bedding 30
P50 P50 20
10
0 40
50 P34 P34
30
40
P17 30
1200 P34 P34 20
20
10 10
P42 P33 0
P33 P33 0
20 50 55 60 65 70 75 80 85 90

10

0
P15 P49 P49 50 P49
Level in section (cm from base)

40

LAVA
highly vesicular
30
20 40

coarse lapilli

Medium lapilli
10 P32
1000 0 30
well sorted

40 P32 20
poorly sorted

30
20 10
P13 10 0
0 50 55 60 65 70 75 80 85 90
30 P31
20

P32
B P32
10
0
40

P48
800 50 P48
30

40 20
30

Fine lapilli
20 10

P31 10
0 0
50 55 60 65 70 75 80 85 90
P31 30
P47
20
10
30
0
P48 P30
600 P11 P48 40
30
P30 20

20
10
P47 10
0
0
P47 30
P29 50 55 60 65 70 75 80 85 90

P09 P30 P30 20


10 30
0
P29 P29 P46
30 P46 20

400 P07 P46 P46 20


10
10

0 0
P45 P45
20 P45 50 55 60 65 70 75 80 85 90

P05 P44 P44 10

0
P03 50
P44
20 40 P27
faint lamination
200 P27 A P27
10

0
30
20
10
30 P27
0
P26 P26 20
50 55 60 65 70 75 80 85 90
10 60
0 50 P26
30
P26 40
30
20
20
10
10
P01 0
0 -6 -5 -4 -3 -2 -1 0 1 2 3 4 >4 0
50 55 60 65 70 75 80 85 90

1 2 3 -6 -5 -4 -3 -2 -1 60 70 80 0 10 20 30 40 220 240 260 280

Stratigraphic log Units Sorting Median Average vesicularity Tachylite/ Bulk Ni (ppm) Proportion (wt.%) Proportion of clasts (%)
(see text) coefficient diameter (φ) (vol.%) sideromelane vs clast size range (φ) vs vesicularity range
(%) (vol.%)*

Fig. 8 Section studied at main cone quarry (see also Fig. 3). From left ity of clasts 16–32 mm, tachylite/sideromelane proportions in the 1Φ
to right: Stratigraphic log showing internal grading of beds and size fraction, and bulk Ni contents in samples across section. Grain-
location of samples studied by bulk analysis (left side of log) and size and vesicularity distribution histograms organized according to
sieving (right side). Section subdivided in three units (A, B and C). stratigraphic height. Vesicularity values recalculated from density
Variation of sorting, coarseness (median diameter), average vesicular- using DRE=2.6 g/cm3. See text for description

trains, lenses, or clusters of well-sorted, clast-supported, with a peak at grain-sizes −4 or −5Φ and a tail at grain-sizes
angular, coarse lapilli. smaller than −3Φ. The distributions of thin beds are less
Beds in Unit C are thinner than in Unit B, finer-grained, skewed, have a main mode at lower grain-sizes (−3 to
less-well sorted and laterally discontinuous (Figs. 8 and −2Φ), broader peaks, and higher relative proportions of
9b). Beds show crude vertical internal stratification over clasts smaller than −3Φ. The inversely-graded fine-grained
10–20 cm. The uppermost beds are strongly altered. bases of thick beds (samples P47, P31, and P33) have
The average vesicularity of clasts in the 16–32 mm poorly skewed to symmetrical distributions with broad
(−4Φ) size-range changes slightly, but not significantly peaks at small grain sizes (1 to −3Φ). Samples from the
through the sequence (Fig. 8). Clasts are highly vesicular latter are better sorted than samples from thin beds with
(average vesicularity=73±6 vol.%, average density: 0.76± similar Md. Note that the fine-grained base (P47) and
0.16, DRE=2.6 g/cm3). coarse interior (P48) of a thick bed spread the whole range
Histograms of grain-size and vesicularity distributions in coarseness and sorting measured in the samples.
provide additional insight into the nature of the deposits Vesicularity distributions of clasts in the 16–32 mm size
(Fig. 8). Grain-size distributions are different for (1) thick range are essentially unimodal (Fig. 8), and their shape or
beds, (2) thin beds, and (3) the fine base of thick beds. range does not vary with bed coarseness. They have a main
Thick beds have strongly coarsely-skewed distributions peak at 65–80 vol.%, with few clasts in the range 85–
868 Bull Volcanol (2009) 71:859–880

broken bomb

Unit B
1m

c c
b

Fig. 9 Photographs of bed sequences indicated on Fig. 3. a Base of light-colored horizons marking the base of beds. Rectangular outline
sequence (unit A). Broken bomb is the same as in Fig. 4b. Note lateral on lower right marks left side of (c). c Close-up of a single coarse bed.
discontinuity of fine-grained (coarse ash) horizons (broken white Note fine-grained material of variable thickness at the base and top of
lines). b Top of sequence (units B and C). From bottom to top: each bed. At base of bed, on left side, fine-grained material
laterally-continuous thick beds interleaved with thinner beds that corresponds to pinching edge of thin lenticular bed (see b) whose
disappear laterally, changing into thinner, laterally discontinuous beds. upper limit is delineated by broken white line
Trains of well-sorted coarse material are observable. Note fine-grained

90 vol.% (one sample) and in the range 50–60 vol.% (three has tachylite proportions that correlate with the coarseness
samples). One sample (P46) covers the entire Pelagatos of the main part of the corresponding beds (samples P32
vesicularity range (51–87 vol.%, 0.34–1.29 g/cm3). The and P48). This suggests that the fine base is the product of
densest fragments in this sample (50–60 vol.% vesicles) the segregation of particles from the main part of the bed. In
have cauliflower-type surface textures (see above). other words, particles forming the fine base are the same as
The proportions of sideromelane and tachylite clasts small particles in the main part (see ‘Evidence for
were determined in the 1Φ size fraction of nine samples avalanching processes’). This can also explain the better
covering the total range in bed coarseness. The plot of sorting of fine bases in comparison to individual beds with
tachylite/sideromelane versus bed coarseness reveals a similar Md. The data collected do not support a linear
positive relationship (Fig. 10): tachylite proportions in- variation of tachylite proportions in the samples with
crease with increasing proportions of coarse clasts (or increasing stratigraphic height (Fig. 8). Samples from units
decreasing proportions of fines) in the deposit. In this B and C tend to be richer in tachylite (>25 vol.%, Fig. 8)
graph, the fine base of thick beds (samples P31 and P47) because they were obtained from coarser beds.
Bull Volcanol (2009) 71:859–880 869

35
intermediate position. Lavas define a smooth trend of
Proportion of tachylite (%)
30 decreasing SiO2 content with increasing distance from the
25 vent (Fig. 11d). Proximal samples from the “later-flows”
field (see Fig. 1b) are distinctly more evolved than more
20
distal lavas erupted earlier (“early flow” field on Fig. 1b).
15 This trend and the more evolved composition of scoria
P35
10 P29 P26 samples from the late-formed cone suggest an increasing
P27 P45
P32 P47 degree of evolution of the magma produced during the
5
P36 P31A eruptive activity. The distinctly less evolved composition of
0 the vesicular bomb and dense angular clast samples, in
-6 -5 -4 -3 -2 -1
comparison to vesicular scoria samples (Fig. 11), is
Median diameter (φ)
intriguing, given that they were also deposited during cone
Fig. 10 Proportion of tachylite over sideromelane clasts versus growth. Their composition is also not matched by any of
median diameter. Results were obtained by separating and counting the lava samples. They may possibly be: (1) recycled pieces
particles (>300 per sample) using a binocular microscope. Symbols
of a magma which was originally erupted at a different
inside squares indicate samples from the fine-grained base of thick
coarse beds. Such a linear correlation is interpreted to be the result of stage of the eruption and is not included in the sample set,
magma ascent rates that control the degree of magma cooling and e.g., from early flows that are now covered, (2) eruptive
fragmentation (see text) products of a magma batch that has undergone a different
ascent path from that of the main magma body, or (3)
fragments of older lavas which were incorporated by the
Composition and petrography of lava and scoria ascending magma. All vesicular scoria samples cover a
narrow range in bulk composition (Table 1, Fig. 11a–c) and
Fresh scoria, bomb, and lava samples were collected for most elements display only minor unsystematic variations
chemical and petrographical analyses in order to constrain throughout the entire sequence. However, Ni contents in
the cooling, degassing, and crystallization of the magma scoria show large variations with the highest values (234–
before and during eruption. Scoria samples from different 269 ppm) in basal units and lowest values (226–248 ppm)
levels of the quarry section were studied to track variations in the upper units. This strong variation contrasts with the
during cone-growth (see Fig. 8). All samples are angular narrow range of Ni in lava samples (246–251 ppm) (Figs. 8
vesicular clasts belonging to the dominant type 2.1 (i.e. and 11e). However, it is interesting to note that the average
‘Clast morphologies and textures’), with the exception of a Ni-content in scoria samples (246±13 ppm) is similar to the
dense and angular clast that belongs to type 2.3 (sample average Ni-content in lava samples (249±2 ppm). The three
P42 on Fig. 8). Quenched glassy parts of the outer crust of analyzed scoria samples have a similar glass composition
three vesicular samples from the base, middle, and top of (Table 2), that is andesitic (~59 wt.% SiO2) with higher
the section were analyzed for glass and crystal composi- FeO, TiO2, and P2O5, and lower CaO and MgO contents
tions. They preserve the texture and composition of the than bulk rock scoria (Table 2, Fig. 11a–c).
liquid magma when erupted explosively at the vent. For Samples contain olivine phenocrysts±plagioclase and
comparison, the dense interior of a vesicular bomb sampled quartz xenocrysts (<3 vol.%), and plagioclase, olivine,
from the base of the talus was also analyzed. Fresh vesicle- clinopyroxene, ±Fe-oxides as groundmass crystals. Modal
poor lava samples were collected at proximal, medial, and crystal abundances and compositional ranges are reported
distal locations of the flow-field (Fig. 1b), to detect in Tables 3 and 4. Olivine phenocrysts are 50–600 µm in
variations in the composition and texture of lava emitted size and occur in clusters. They commonly have Fo-and-Ni
during the eruption. rich euhedral cores and Fo-and-Ni poor skeletal rims
Whole rock analyses are reported in Table 1 and results (Table 4), and contain euhedral <50 µm Cr-rich spinel
are plotted with the average scoria glass composition in inclusions. They may enclose irregular, generally <20 µm-
Fig. 11. All samples cover a relatively small range in bulk sized, glass inclusions, and rarely display arrow-like
rock compositions (52.94–54.81 SiO2 wt.%; 9.27–10.24 hopper-shapes. Xenocrysts form >100 µm-sized plagio-
MgO wt.%, compositions normalized to 100%). MgO, clases with resorbed cores intergrown with olivines and
CaO, and FeO decrease with increasing SiO2 wt.% clinopyroxenes. Groundmass crystals are distinctly more
(Fig. 11a–c). Low amounts of olivine±clinopyroxene± abundant (Table 3) and more evolved (Table 4) in the lava
plagioclase fractionation can explain such a trend. The than in the scoria. Plagioclase is the dominant groundmass
vesicular bomb and the dense angular clast plot at the less phase (34–39 vol.% in lava, 33 vol.% in scoria). Crystals
evolved end of the trend (Fig. 11a–c), vesicular scoria are distributed with equal density in the two sample types,
samples at the most evolved end, and lavas in an having slightly larger sizes in the lava. They are lath-
870 Bull Volcanol (2009) 71:859–880

Table 1 Whole rock chemical analyses of Pelagatos lava, bomb, and scoria samples

Sample Scoriaa Scoriaa Scoriab Vesicular Dense Medial Distal Distal Medial Prox. Cone
type bomb clast lava lava lava lava Lava overflow
Sample Average Stdev P53 PG0510 P42 PG0515 PG0516 PG0517 PG0518 P61-1 P67-1
no.
Long. 19°05′ 19°05′ 19°05′ 19°05′ 19°05′ 19°05′ 19°07′ 19°06′ 19°06′ 19°05′ 19°05′
30.06″ 30.06″ 33.32″ 34.60″ 30.06″ 59.41″ 23.42″ 47.77″ 29.08″ 06.7″ 35.6″
Lat. 98°57′ 98°57′ 98°57′ 98°57′ 98°57′ 98°55′ 98°54′ 98°54′ 98°56′ 98°57′ 98°57′
44.65″ 44.65″ 47.73″ 44.94″ 44.65″ 53.64″ 30.92″ 26.84″ 23.2″ 34.7″ 35.5″
Alt. (m) 2,973 2,973 2,980 2,998 2,973 2,819 2,688 2,671 2,838 2,950 2,982

wt.%
SiO2 54.21 0.17 54.23 52.53 52.51 53.40 53.12 53.19 53.26 53.62 54.68
TiO2 0.82 0.01 0.79 0.82 0.87 0.82 0.79 0.79 0.85 0.83 0.81
Al2O3 15.51 0.11 15.20 15.64 16.31 15.57 15.21 15.28 15.43 15.47 15.50
Fe2O3(t) 7.76 0.04 7.88 8.20 8.15 7.83 7.95 7.91 7.86 7.60 7.50
FeO(t)c 6.98 0.04 7.09 7.38 7.33 7.05 7.15 7.12 7.07 6.84 6.75
MnO 0.12 0.00 0.13 0.13 0.13 0.12 0.12 0.12 0.12 0.12 0.12
MgO 9.49 0.12 9.65 10.16 9.67 9.84 10.09 9.98 9.78 9.50 9.79
CaO 7.66 0.10 7.77 8.09 7.55 7.92 7.95 7.98 7.64 7.55 7.70
Na2O 3.44 0.05 3.22 3.41 3.52 3.45 3.32 3.34 3.41 3.42 3.53
K2O 0.92 0.03 0.93 0.91 1.05 0.89 0.91 0.86 0.92 0.94 1.02
P2O5 0.17 0.00 0.17 0.16 0.18 0.16 0.15 0.15 0.17 0.17 0.17
LOI <0.01 0.00 < 0.01 0.00 0.03 0.14 0.03 < 0.01 0.00 0.00
Total 99.32 0.32 99.18 99.22 99.12 99.22 98.82 98.81 98.66 98.46 100.07
Ni (ppm) 226–269 236 241 222 250 250 250 251 247 246

For analytical procedures see Appendix 2


a
Average and standard deviation (Stdev) of bulk rock scoria analyses (samples P01 to P23, see Fig. 8)
b
Sample from medium distance of section to NE side of quarry. Long. longitude, Lat. latitude, Alt. altitude (GPS data). Prox. proximal. Fe2O3(t)=
total Fe as Fe2O3
c
FeO(t)=total Fe recalculated as FeO (=0.8998×Fe2O3(t)); Total calculated using FeO(t)

shaped, often in parallel-branching arrangements, and with average composition was used. Olivines with Fo87–91
skeletal outgrowths. Groundmass olivines and clinopyrox- cores are the only phenocrysts in the Pelagatos products.
enes are rare and small in scoria but more numerous and The model “PELE” developed by Boudreau (1999) esti-
larger in lava (Table 3). Olivines are oval to rectangular in mates that ~Fo90 olivine is the first phase to crystallize in a
shape and most often isolated. Clinopyroxenes form magma with the bulk composition of the Pelagatos products
<10 µm-size isolated laths in scoria, but aggregates of at 1,260–1,270°C. This estimated temperature range also
granular crystals up to 50 µm across in bomb and lava agrees with temperatures calculated using several theoret-
samples. Fe-oxides (titanomagnetite) are only present in ical and empirical models (Beattie 1993; figure 2 in
lava and bomb samples. They are surprisingly abundant in Putirka et al. 2007; Montierth et al. 1995; Sugawara
the former (4–11 vol.%) where they form granular crystals. 2000) and experimental data on high-Mg andesites similar
in composition to Pelagatos (Tatsumi 1982; Pichavant and
MacDonald 2007). The occurrence of chromium spinels as
Discussion inclusions in the olivine crystals implies that they formed
slightly before or during olivine crystallization.
Cooling, degassing and crystallization history of the magma Melt inclusion data from compositionally similar mag-
mas can help constrain the pressure of crystallization of the
Models and experimental data from the literature were olivines (and Cr-spinels) and the melt water content.
compared with results obtained during this study in order to Primitive olivines in other high-Mg magmas from the
constrain the conditions of the evolving magma prior to and SCVF (Xitle, Jumiltepec, Las Tetillas, Tuxtepec, Tepetlapa)
during the Pelagatos eruption. The bulk magma composi- contain glass inclusions that represent captured melt with
tion was considered homogeneous in the modeling, and an 1.3–5.2 wt.% H2O at 100–600 MPa (Cervantes and Wallace
Bull Volcanol (2009) 71:859–880 871

Fig. 11 Compositional varia- 12 55.0


tion in bulk rock samples and
scoria glasses. Symbol key on 10
a
54.5
(a) applies to all plots. See
8 Bulk rock

MgO (wt.%)
Table 1 for chemical data and

SiO2 wt.%
sample locations. Broken line= Vs bomb 54.0
6 Later
Dense clast
trends in plots. In symbol key: flows
Lava
Vs=vesicular, Vs scoria1 = 4 53.5
Vs scoria1 Early flow
average composition and
2 Vs scoria2 Scoria
standard deviation of vesicular 53.0
Lava (Schaaf et al. 2005) glass
scoria samples from section 0 d
shown in Fig. 8, Vs scoria2 = 50 55 60 52.5
composition of vesicular scoria 0 2 4 6 8
sample from opposite side of 8.5
approx. distance from source (km)
quarry (all vesicular scoria sam- b
ples have similar composition). 8.0 280
FeO (wt.%)

For plots (d) and (e), only bulk Scoria


270
rock data is plotted. For plot (e), 7.5 260 Unit A
two additional symbol types are

Ni (ppm)
used to distinguish between 250
scoria samples from basal layers 7.0
240 Units
(Unit A) and upper layers B&C
(Units B and C) 230
6.5
50 55 60 220 e
210
8.5 50 52 54 56 58
8.0 SiO2 wt.%
CaO (wt.%)

7.5

7.0

6.5 c
6.0
50 55 60
SiO2 wt.%

2003). Olivines in lavas erupted at Jorullo and Paricutin mid-crustal levels (100–600 MPa) from a melt that
trapped inclusions under a similar range of pressure and contained ≤5 wt.% H2O. Such pressures are in agreement
melt water content (Johnson et al. 2008; Pioli et al. 2008). with the narrow bulk compositional range of the Pelagatos
Thus, olivines in the Pelagatos magma probably formed at magma. The low abundance of phenocrysts implies that the
magma was not stored at depth and that olivines mainly
Table 2 Scoria glass composi- crystallized during rapid ascent.
tions obtained by electron mi- Wt.% Average Stdev
A melt with 5 wt.% H2O reaches saturation levels at
croprobe analysis
SiO2 58.63 0.315 ~6 km depth or 200 MPa (Pineau et al. 1998). It is probable
TiO2 1.26 0.045 that the microlites formed above this depth, under con-
Al2O3 15.49 0.218 ditions of melt undercooling induced by gas-loss during
FeO 7.73 0.298 rapid magma rise. This effect has been widely observed and
MnO 0.13 0.037 is well expressed by the high proportion, dense distribution,
MgO 3.92 0.130 small size, and skeletal shape of plagioclases, which is the
CaO 7.01 0.165 phase that is most affected by undercooling (e.g. Hammer
Na2O 3.82 0.207 and Rutherford 2002). The maximum temperature of the
For analytical procedures see
Appendix 2 K2O 1.40 0.072
erupted magma can be estimated at 1,100°C taking into
Average and standard deviation P2O5 0.26 0.037
account evidence for co-crystallization of olivine, clinopyr-
(Stdev) of 59 electron microprobe NiO 0.02 0.025
oxene, and plagioclase microlites in the products (PELE
point analyses (26 analyses model). Scoria clasts contain minor amounts of olivine and
of sample P13, 24 analyses of Cr2O3 0.00 0.005
clinopyroxene microlites, suggesting magma temperatures
sample P21, 9 analyses of sample Total 99.67 0.494
P01); Mg#=MgO/(MgO+FeO), close to 1,100°C, which is confirmed by models applied to
Mg# 33.64 0.011
FeO total microlite and scoria–glass compositional pairs (Beattie
872 Bull Volcanol (2009) 71:859–880

Table 3 Modal mineralogical abundances in representative lava, bomb, and scoria samples

Oliv pheno Gd oliv Gd plag Gd cpx Gd Fe-ox Gd glass Total cryst.

Bomb and lavaa


PG0510A Bomb 10.1 3.4 32.2 14.2 0.9 39.2 60.8
PG0513 proximal lava 7.0 2.3 51.6 14.9 3.9 20.4 79.6
PG0518 medial lava 12.9 1.3 34.2 20.6 10.9 20.1 79.9
PG0515 medial lava 11.0 2.4 41.8 22.2 6.2 16.4 83.6
PG0516 distal lava 10.8 1.4 39.1 25.8 8.6 14.3 85.7
PG0517 distal lava 12.8 1.0 39.0 25.5 6.3 15.5 84.5
Scoriab 12.7 33.3 3.2 0 50.8 49.2

Oliv olivine, pheno phenocryst, Gd groundmass, cpx clinopyroxene, plag plagioclase, Fe-ox iron-oxides, cryst. crystallinity
a
Samples analyzed using point counter (total counts ~800, vesicles subtracted from total count to calculate modal %). Modal abundance of olivine
phenocrysts includes <0.3 vol.% of Cr-spinel enclosed in the crystals. Samples are classified in table according to their distance from source. Note
absence of correlation. Most variations are thought to correlate with rate of post-emplacement cooling
b
Calculated using average scoria matrix glass and bulk rock compositions, end-member crystal compositions, and program using least-square
method (MINSQ, Herrmann and Berry 2002). Fits are very good (sum of squared residuals=0.03), showing that assumption of in-situ crystal
fractionation is valid. Note that most (in terms of vol.) olivine crystals in scoria are phenocrysts. Values were recalculated to 100%

1993; Putirka 2005) and scoria–glass composition alone supersaturated (<5 wt.% H2O) melt, possibly over a wide
(Sugawara 2000; Montierth et al. 1995). The coarser lava range of pressure (depth <10 km). Shallow degassing
texture implies larger amounts of cooling. Such cooling (<6 km depth) triggered intense microlite crystallization.
may have taken place either within the conduit, because the During explosive events at the vent, the magma ascended
degassed nature of the lava suggests relatively low magma rapidly and was quenched as scoria at ~1,110°C. During
ascent rates, or during lava flowage, because the blocky effusive phases, the magma cooled more slowly and to
surface of the lava provides poor insulation to the interior larger degrees before final solidification.
of the flow. The former process might have been less
important judging by evidence that supports strong under- Chronology of the eruption
cooling during ascent (high density of plagioclases) which
can be best explained by rapid magma rise. The abundance The alignment of the main cone with the spatter ridges
of Fe-oxides in the lava implies temperatures well below strongly suggests that the eruption was structurally con-
1,040°C (PELE model). However, most of these crystals trolled by a fissure. The fissure is oriented E–W, parallel to
may have formed during final solidification within the normal faults that affect the local basement, and other
interior of the flow (lava solidifies at ~700°C). The lava extensional tectonic lineaments within the volcanic field
emplacement temperature could thus have been similar to (Márquez et al. 1999; Siebe et al. 2004b). This suggests that
that measured at Paricutin (1,020–1,100°C, Luhr and the feeder dike was either “captured” by a pre-existing
Simkin 1993, and references therein). normal fault (e.g. Valentine et al. 2006; Gaffney et al. 2007;
In summary, the P–T path followed by the Pelagatos Valentine and Keating 2007), or rose parallel to the
must have been similar to that of the early Jorullo magma direction of least principal stress determined by the active
as calculated by Johnson et al. (2008). Olivine and Cr- regional extensional regime (e.g. Nakamura 1977; Márquez
spinels grew from a high-temperature (T>1,200°C) H2O- et al. 1999).

Table 4 Crystal compositional ranges in representative scoria and lava samples obtained by electron microprobe analysis

Samples Oliv pheno Cr-spinels Gd oliv Gd plag Gd cpx

Scoria (P01, P13, P21) Cores: Fo 87–91 NiO 0.141–0.492, Cr# 57–68 Fe2# 35.2–56.2 =pheno rims An54–69 Wo40–42
Rims: Fo 83–87 NiO<0.214 En48–50a
Lava (P59–1) =scoria =scoria Fo60–80 An48–65 Wo38–46
En46–52

For analytical procedures see Appendix 2. Same abbreviations as in Table 3. Fo=100×Mg/(Mg+Fe); Cr#=Cr/(Cr+Al), Fe2#=Fe2 /(Fe2 +Mg);
An=100×Ca/(Ca+Na+K); Wo=Ca/(Ca+Fe+Mg); En=Mg/(Ca+Fe+Mg). NiO in wt.%
a
Two microlites only. A plagioclase xenocryst analyzed was distinctly more evolved (An64–70)
Bull Volcanol (2009) 71:859–880 873

Early eruptive products include the main lava flow and observed within the flow-field. They may have been
the scoria ridges exposed at the eastern end of the fissure (it dispersed by the lava flow, or if they exist, be hidden by
is likely that scoria ridges also formed on the western part vegetation.
of the fissure and were later covered by flows from the Note that the eruption of Pelagatos is considered
main cone). The first eruptive phase was thus dominantly monogenetic because of the lack of evidence for a
effusive. Flow morphology suggests that the lava issued significant time interval between the two eruptive phases.
from the entire length of the fissure. The magma discharge Its duration can be crudely estimated using data collected at
rate was probably relatively high and sufficiently continu- active volcanoes. The small size of the cone (50 m) implies
ous to enable the formation of a single flow, and not a that it formed in less than a day to about a month (Fig. 7 in
sequence of overlapping flows. The length and narrowness Wood 1980). The total volume of lava erupted (~4×107 m3)
of the flow were promoted by the flow of the lava into a would have required from 5 days to 15 months to
depression delineated by the margins of flows from older accumulate at average effusion rates of 1 to 100 m3/s
nearby cones. The presence of truncated scoria ridges on (range for a final lava flow length of 7 km, see figure 4 in
the eastern end of the fissure indicates that some mildly- Walker 1973). Accordingly, the eruption may have lasted
explosive activity was taking place while the lavas were from 6 days to 16 months.
emitted. Such activity may have corresponded to a
Hawaiian-style lava fountaining episode that created Mode of growth of the main scoria cone
simultaneously rheomorphic flows and a scoria cone (e.g.,
Head and Wilson 1989). Such a scenario is however Evidence for avalanching processes
considered unlikely in this case because diagnostic features
indicative of fountaining (spatter, agglutinates, and fluidal When a growing scoria cone built of loose material exceeds
clasts) are not observed at Pelagatos. This might be due to a given size and its slopes reach the angle of repose for
the relatively high magma viscosity (see below). Instead, loose scoria (32°), deposits are redistributed by avalanching
explosive and effusive activity might have alternated. (McGetchin et al. 1974). Evidence for avalanching is
Evidence for this is the scoria mounds observed at the base widespread at the Pelagatos quarry and includes: the
of the easternmost scoria ridge. These mounds represent lenticular shape of thin beds, the downslope thickening of
pyroclastic material deposited on the down-flow side of the coarse beds, and the irregular shape of the base of thick
fissure during a pause in lava effusion, that were later rafted beds. The systematic basal reverse grading of the beds is
by the flows when effusive activity resumed. Such unsteady best accounted for by self-sieving of the very loose material
eruptive behavior may have coincided with closing stages upon sliding (Fisher and Schmincke 1984). This process
of activity at that section of the fissure. can still be observed to occur at present day, since small
The second phase focused on the western end of the avalanches detach frequently from the unstable base of the
fissure. It featured more sporadic lava output, forming a quarry face and add material to talus. Sedimentary
compound lava field (“later flows” in Fig. 1b), and more structures originated by avalanching are particularly well
intense explosive activity that built a 50-m-high scoria developed at Pelagatos due to the loose and friable nature
cone. The transition from phase 1 to phase 2 may have been of the deposits, and the asymmetry of the cone built on a
gradual: the main scoria cone may have started to form as sloping lava field.
the main lava flow was being emplaced. Final activity was
concentrated at the main cone, featuring weak explosions Evidence for eruptive style: a case for strong Strombolian
and a final small lava flow. There is evidence for at least activity
one cone-breaching event, although there might have been
several, based on the unstable nature of the cone that is The geometry of fallout deposits is generally used to assign
small, consists of loose material, and was built on an older a particular style to an eruption because it reflects the
sloping lava flow fan. The last flow apparently erupted characteristics of the eruptive column. However, medial and
from the crater of the cone, although earlier flows (or some distal fallout deposits are not preserved at Pelagatos:
of them) originated from the base of the cone. All effusive pyroclastic deposits are only observed on the steep-sided
phases may have been associated with cone collapse, cone, where they were partly redistributed by avalanching.
whether lava was issuing from the crater or the base of Yet, clast morphologies and clast vesicularity, and certain
the cone. Cone breaches might have healed rapidly, leaving sedimentary features of the proximal cone deposits provide
little evidence for their occurrence, as observed at active constraints on the eruptive style.
cones (e.g., Foshag and González 1956; Luhr and Simkin In a broad sense, the deposits are comparable to products
1993). Remnants of rafted parts of the scoria cone (e.g. from magmatic (=dry) Strombolian activity. They are well-
Riggs and Duffield 2008; Valentine et al. 2006) were not bedded, loose, have low proportions of lithics and poorly-
874 Bull Volcanol (2009) 71:859–880

vesicular dense clasts, and lack the fluidal shape of clasts so The products also show sedimentary features that are
typical for Hawaiian-style eruptions (e.g., Valentine and diagnostic of deposition from elevated and sustained
Gregg 2008). Furthermore, the deposits are coarse, poorly eruptive columns. For instance, the normal grading dis-
sorted, and have strongly skewed grain-size distributions played at the upper part of many beds is best explained by
similar to other studied Strombolian deposits (Houghton segregation processes of particles falling from an eruptive
and Hackett 1984; Houghton and Schmincke 1989). cloud according to their size. Some of the laterally-
The large amount of angular vesicular clasts derived discontinuous fine lapilli horizons that occur within single
from solidified bombs 5–10 cm across (−5 to −7Φ) in the beds may also result from weak density currents associated
Pelagatos deposits is however somewhat atypical for with heavy ash fallout, as observed at Etna (e.g., Taddeucci
Strombolian products. Indeed, during regular activity at et al. 2004). Finally, many beds are internally stratified,
Stromboli, flight times of bombs of this size (5–10 cm) are which is another feature typical of fallout deposits (e.g.,
not long enough (~10 s to slightly longer, McGetchin et al. Valentine et al. 2005, 2007; Valentine and Keating 2007).
1974) for the clasts to cool completely and degas Ejection heights estimated for the explosions at Pelagatos
significantly (to produce the coarsely-vesicular cores) prior are high compared to those regularly observed at Stromboli
to landing. Low magma temperatures could allow in-flight (e.g., scoria ejection heights <300 m, Lautze and Houghton
solidification, but this does not apply to Pelagatos where 2007), but they are not uncommon. Cone-forming activity
the magma was erupted at ~1,110°C. during the eruption of Heimaey (Iceland) in 1973 involved
One process that increases total clast transport times and significantly stronger explosions than those characterizing
thus time for cooling is clast recycling at the vent. This process Stromboli (Blackburn et al. 1976). At Heimaey, bombs
has been observed directly in scoria cone eruptions (e.g., were ejected up to 500 m above the crater rim, fragments
Foshag and González 1956; Blackburn et al. 1976). Failing to 2–10 mm in size fell from up to 1 km high, and little spatter
reach the upper cone rim, large bombs fall back into the vent, was produced (Self et al. 1974). Curiously, this type of
are re-ejected into the air, and so on, until they finally land on activity seems to be more common than that observed at
the cone’s outer slope. This process generally produces clasts Stromboli (Blackburn et al. 1976). However, this latter is
covering a wide range in vesicularity, including dense round more commonly used as a reference. Even stronger activity
clasts (Fig. 4f), and composite bombs (Fig. 4b) formed by was observed at Paricutin where eruptive columns reached
welding of recycled clasts rolling back into the crater. The 6 km high (Ordoñez 1947), and such eruptive style was
proportion of such clasts is however small in the Pelagatos defined as “violent Strombolian” (e.g., Pioli et al. 2008,
deposits. Hence, the recycling process must have played a Valentine and Gregg 2008). Violent Strombolian activity,
subordinate role at Pelagatos, and thus cannot be held however, typically produces thick ash fallout deposits (e.g.,
responsible for the formation of the bulk of the deposits. Pioli et al. 2008, Valentine and Gregg 2008) which are
On the other hand, the minor occurrence of recycled deposits absent at Pelagatos. Because the explosive activity at
implies that the vent must have stayed open (un-chocked by Pelagatos was of lower magnitude, the term “strong
debris) during the largest part of the activity at the cone. Strombolian” is preferred.
Long flight times and bomb breakup upon final impact
(in contrast to breakup by the recycling process) seems to Origin of “strong Strombolian” activity at Pelagatos
be the only other process that can account for the high
abundance of angular, but vesicular clasts in the Pelagatos Many factors can promote “dry” highly-explosive activity
products. Such clasts mainly cool during fall (e.g. Thomas at scoria cones (e.g., Pioli et al. 2008, Valentine and Gregg
and Sparks 1992). Fall heights required for cooling the 2008). At Pelagatos, it seems that a matching combination
largest clasts present in the Pelagatos scoria beds were of relatively high magma viscosities and high magma
estimated using figure 5a from Thomas and Sparks (1992). ascent rates caused the “strong Strombolian” eruptive style.
Accordingly, fall heights of 1 km would be sufficient to The viscosity of the Pelagatos magma was calculated using
cool 32–64 mm-size clasts (−5Φ) from 1,110 to 700°C and procedures proposed by Giordano et al. (2008) and Vetere
thus allow for solidification before landing. For the same et al. (2006) for deriving melt viscosity, and the Einstein–
temperature range, complete cooling of larger clasts (64– Roscoe equation to take crystal content into account
128 mm in diameter; −6Φ) would require heights of fall (Pinkerton and Stevenson 1992). Magma viscosity was
between 2 and 3 km. When falling from a height of only initially low (40–80 Pa s) but increased sharply during the
1 km, the cores of clasts of this size would only reach latest stages of ascent (6×105–4×106 Pa s) due to cooling,
~800°C upon landing. But this might be cool enough for intense degassing, and associated microlite crystallization.
them to break. In conclusion, angular vesicular clasts in the Accordingly, magma yield-and-tensile strength increased,
Pelagatos deposits could have formed by strong explosions promoting efficient magma fragmentation and high ejection
ejecting bombs to heights >1 km. velocities, as well as brittle behavior of the clasts upon
Bull Volcanol (2009) 71:859–880 875

landing. The vesicular texture of the scoria glasses at decreasing magma ascent rates. The thinner and more
Pelagatos implies intense degassing (high vesicularity) but discontinuous, yet relatively coarse deposits at the top of
limited bubble coalescence (large amounts of isolated small the sequence (Unit C) may represent conditions of
vesicles). High magma ascent rates combined with rela- sharply fluctuating magma ascent rates towards the end
tively high magma viscosity can cause such features. Low of the cone building phase. It should be noted, however,
magma permeability can in turn lead to rising gas that the eruption sequence may have been more complex
overpressure, which can be sustained by rapid crystalliza- and hours to days may have elapsed between individual
tion enriching the melt in volatiles (Taddeucci et al. 2004). explosions.
In this context, it is worth mentioning that at Heimaey (see
above), the fragmentation of some bombs during flight was Conduit processes and bulk Ni variations
related to rapid vesiculation of an apparently-viscous
magma (Self et al. 1974). Such a process may also have The wide range in bulk Ni contents of scoria and bomb
occurred at Pelagatos. samples is in strong contrast with the narrow range
displayed by lava samples (see Table 1), and suggests a
Variations of eruptive intensity during cone growth process that was segregating olivine phenocrysts during
the explosive activity. The total range of variation detected
Changes in bed thickness and coarseness through the upper (47 ppm) can be accounted for by only <3 vol.% crystals,
cone section (Fig. 8) express variations in eruptive intensity because Ni partitions strongly into olivine phenocryst
during the final stages of cone growth. Causes for such cores (Table 4). Although part of the variation might be
variations can be several. A change in bulk magma viscosity an artifact of insufficiently large sample sizes for the
can be discarded in this case due to the lack of variation in vesicular scoria samples submitted for analysis, this
bulk rock composition, or in crystal or vesicle contents effect cannot be the only explanation because the dense
through the sequence. There is, however, a systematic bomb sample (P42) has a low Ni content (Fig. 11). At
relation between bed coarseness and the proportion of shallow levels, rapid magma rise, high bubble content,
degassed and cooled ash particles (tachylite) relative to the and elevated melt viscosity might create heterogeneities
amount of highly-vesicular glassy particles (sideromelane) within the flow that segregate dense olivine phenocryst
(Fig. 10). Different processes have been invoked to explain clusters. This segregation process seems to be weak
variations in the abundance of such fragments within during effusive activity but important during explosive
Strombolian fallout deposits (Heiken 1978; Taddeucci et activity, perhaps promoted by high magma ascent rates.
al. 2004; Valentine et al. 2005; Polacci et al. 2006; Sable et It is not clear why basal scoria layers in the cone tend
al. 2006; Lautze and Houghton 2007). In this case, such to be relatively enriched and upper layers relatively
variations are best explained by a correlation with depleted in olivine phenocrysts, but this might be linked
magma ascent rates. High magma ascent rates will limit to evolving conditions of magma ascent during the
magma residence times in the conduit and thus the extent course of the activity. In particular, the magma may
of magma cooling and degassing prior to quenching (low have resided for a short time in shallow small sills
tachylite proportions), and promote high degrees of branching off the main conduit (e.g. Valentine and
fragmentation due to high rates of magma vesiculation. Krogh 2006), and crystals may have settled during this
Resulting deposits will be less coarse. Conversely, low time. The drawing of these magma batches into the main
magma ascent rates will produce coarser deposits rich in conduit at late stages of the eruption could explain the
tachylite clasts. Alternatively, coarser beds at a given variations observed.
location might be related to higher eruption intensity
(higher eruption column) because high amounts of The shallow plumbing system and the evolution
tachylite suggest high magma viscosity and yield strength of the activity during the eruption
(e.g., Taddeucci et al. 2004). However, tachylite-rich
deposits produced by strong explosions at Etna are fine- The Pelagatos activity changed from effusive to explosive
grained and associated with large amounts of dense bombs during the course of the eruption. Lava fountaining during
(Taddeucci et al. 2004), a feature not observed at Pelagatos. the first phase seems unlikely (see ‘Chronology of the
Hence an inverse correlation between bed coarseness and eruption’). Assuming that the first phase featured the
tachylite content, and magma ascent velocities and explo- effusion of degassed magma, then the transition to
sion intensity is preferred for the case of the Pelagatos explosive activity could express (1) increasing gas contents,
eruption. The sequence observed through Unit A and B of (2) increasing magma ascent velocity, or (3) increasing
the section (i.e., thickening and coarsening of beds, Fig. 8) magma viscosity. The second and third hypotheses seem
could therefore be interpreted as the result of overall more likely, although more data would be needed to test the
876 Bull Volcanol (2009) 71:859–880

first. Closure of the fissure to form a single conduit through the crust (figure 16 in Carmichael 2002). This
provides a process by which magma ascent rates can allows such magmas to reach shallow levels with elevated
increase, leading to higher explosivity, because of insuffi- temperature, high H2O contents, and low crystallinity.
cient time for the gas to escape passively from the magma Their eruptive style will therefore be controlled by
(Klug and Cashman 1996). Moreover, the ~40 vol.% shallow degassing processes that increase dramatically
microlites formed in the rapidly-decompressed magma, as the magma’s viscosity, reduce the time for passive magma
well as its higher bulk SiO2 contents (Fig. 11) compared to degassing, and thus induce more highly explosive activity
magma delivered during phase 1, would have increased its than is generally expected at scoria cones. The Jorullo and
viscosity, thus promoting higher explosivity. Fluctuations in Paricutin eruptions in western Mexico, which were fed by
the intensity of the explosions during phase 2 and magmas similar to the Pelagatos eruption, differed
alternating and/or simultaneous effusive phases indicate markedly from the latter by their long-lasting highly-
that magma-and-gas fluxes in the conduit were variable. explosive phases that produced considerable amounts of
These variations can be partly explained by the formation ash. At Paricutin, this has been related to high magma-
of shallow sills and dikes branching off from the conduit and-gas flux rates and efficient magma–gas segregation
walls where parts of the magma could have fractionated and within the large cone formed early during the eruption
gas segregated (e.g. Valentine and Krogh 2006; Keating et (Pioli et al. 2008). Ongoing work at Jorullo also suggests
al. 2008). The rise of degassed magma batches within the that the highly-explosive dry phase coincided with the
cone, and/or their extrusion at the base of the cone, would construction of the main cone. The localization of magma
have fed the lava flows. flow, and concomitant increase in magma ascent velocity,
and subsequent construction of a cone permitted by the
long duration of the eruption and/or the tectonic setting (at
Summary and implications Paricutin), may then be the main factors determining the
higher explosivity of those magmas. At Pelagatos, the
The Pelagatos eruption was mainly effusive, with a late explosive phase of the eruption was relatively short,
short-lived (<1 month) explosive phase forming a small building only a small cone, but we propose that it would
cone. The magma ascended relatively rapidly through the have developed into a more explosive eruption forming a
crust (low phenocryst contents, primitive bulk composi- large cone and thick distal fallout deposits if magma
tion), facilitated by high temperatures (>1,200°C), elevated supplies had continued.
volatile contents (<5 wt.% H2O), and pre-existing fractures.
Depressurization triggered degassing at shallow levels
(<6 km) and microlite crystallization. During the first Conclusions and hazards
phase, low magma ascent rates allowed magma to degas
at shallow levels, producing fissure-fed lava flows with Recent radiocarbon dating of scoria cones in the TMVB
elevated crystal contents, T<1,100°C, and blocky surfaces. (e.g. Siebe et al. 2005; Siebert and Carrasco-Núñez 2002;
Later, increasing closure of the dike focused the magma Sieron and Siebe 2008) have revealed that monogenetic
flow and increased magma ascent velocity. The combina- eruptions seem to be more frequent in Mexico than
tion of sharply increasing magma viscosity due to ~40 vol. generally recognized before. For this reason the evalua-
% microlite crystallization driven by rapid gas exsolution, tion of risks for surrounding populations and infrastruc-
higher bulk SiO2 content, and limited time for gas escape, ture from this type of eruption has been neglected until
resulted in strong explosions (>1 km high eruptive recently. Many of the young monogenetic vents of the
columns) of relatively hot (1,100°C) and highly vesicular TMVB, including those from historic eruptions of Jorullo
magma. Unsteady magma-and-gas flow rates modulated the and Paricutin, are composed of olivine-bearing basaltic
intensity of the explosions. As a result, a scoria cone andesite (Jorullo) to andesite (e.g. Paricutin) magmas.
formed, while the rise of degassed magma batches within This study provides insights into the processes that
the cone and their extrusion at the base of the cone fed control the explosive intensity and eruptive style of such
occasional lava flows. These flows breached the unstable magmas, which may be of use for making hazard
cone and formed a proximal compound flow-field. assessments and designing geophysical monitoring strat-
High-Mg phenocryst-poor basaltic andesite magmas, egies. In particular, the localization of the magma flow
such as the Pelagatos magma, are thought to originate during the course of a fissure eruption and resulting
from partial melting of a peridotite enriched in slab- increase in magma ascent rates can lead to explosive
derived volatiles (Tatsumi and Ishizaka 1982; Carmichael eruptive phases producing high amounts of vesicular,
2002; Cervantes and Wallace 2003). Their high initial microlite-rich scoria and ash. The duration and explosiv-
water contents and thus low density facilitate rapid ascent ity of such phases will be dictated by magma flux rates
Bull Volcanol (2009) 71:859–880 877

Tachyl.
and the total erupted magma volume. In the case of a

22.7

33.1

27.4

20.0
25.4
11.0
6.0
new eruption of this type in the future, close petrological

Crystal-types (%)
monitoring of the emitted products would be recommen-

Siderom.
ded to trace shallow degassing and crystallization

77.3
93.3
89.0

66.9

72.6

80.0
74.6
processes, and possibly detect precursors of dangerous
explosive phases.

1132
435
758

709

756

456
611
N
Acknowledgements Institutional and logistic support was provided

Stdev
Vesicularity
by Instituto de Geofísica, Universidad Nacional Autónoma de México.

3.8
4.5

6.1

5.0

5.6

6.6
The first author wishes to thank Agnes Mazot for providing pleasant

(vol.%)
and useful assistance in the field, Carlos Linares López and Adela

74.8
75.1

71.3

72.5

72.3

75.0
Avg
Margarita Reyes Salas for technical assistance with the microprobe
and the SEM, Julie Roberge for insightful discussions, and Lilia Arana
and Renato Castro-Govea for their general help. The present work was

0.51
0.35
0.32
0.52
0.17
0.14
0.38
0.27
0.33
0.37
0.50
0.33
αΦ
carried out with the aid of a postdoctoral fellowship from Coordina-
ción de la Investigación Científica (UNAM) to M.-N. Guilbaud and a

2.33
2.08
1.93
1.98
1.45
1.28
1.53
1.93
1.05
2.23
2.40
2.30
stipend from UNAM-Posgrado to J.A. Flores. Analytical costs were

σΦ
Statistics
defrayed by projects CONACyT-50677-F and DGAPA-UNAM-

−3.75
−2.25
−2.25
−5.15
−3.30
−3.40
−5.30
−3.00
−5.25
−3.10
−4.45
−3.35
IN101006-3 assigned to Claus Siebe. Two anonymous journal

MdΦ
reviewers, Greg Valentine, and editor James White made useful
suggestions for improving this article.

0.9
1.4
1.2
0.4
1.3
0.9
0.4
1.1
0.3
0.9
0.6
1.4
>4

2.1
2.6
2.2
0.9
1.6
1.3
1.1
1.6
0.7
2.0
1.4
2.5
4
Appendixes

3.3
3.9
3.4
1.7
2.2
1.9
1.8
2.5
1.4
3.5
2.5
4.2
3

4.1
5.2
4.4
2.5
2.7
2.4
2.4
3.4
2.1
4.4
3.5
5.8
Appendix 1: Analysis of cone scoria

4.6
7.4
8.0
3.0
2.8
2.4
2.5
4.5
2.2
5.1
4.4
8.0
Representative samples of scoria were taken from nearly 1
each bed in the sequence. Sample size was chosen in

13.7
4.9
9.0

3.0
2.9
2.6
2.2
6.3
1.9
6.4
4.8
9.2
accordance with the coarseness of the sampled bed.
0

Samples were first processed in the field for large clast- 14.0
10.9

12.3
11.4
% Weight per grain-size fraction in Φ

6.6

3.9
6.0
4.9
2.1

1.9
9.3
5.4
−1

sizes (−5 and −6Φ), using large specially-built sieves. The


remaining sample (clasts smaller than −5Φ) was divided
30.0
29.1

21.4
20.0

20.0

17.2

22.5
9.9

6.4

3.0

2.4

8.1
−2

into smaller fractions in the laboratory using a mechan-


ical divider, and a sub-sample was then processed. All
19.8
17.6
21.0

34.0
37.1

25.1

25.4

22.1
11.4
6.1

6.8

4.4
−3

samples were hand-sieved for the grain-size fractions −4


to −2Φ. An automatic machine was used for a few
33.4

17.0
18.3
19.7
17.8
20.2
22.2
20.3
23.1
11.9
7.0
5.1
−4

minutes to sieve smaller grain-size fractions. Two sub-


10.3

39.6

40.5

51.2

31.2
samples of sample P31 (sub-samples P31A and P32B)
1.9
1.0

6.7
6.7

3.8

5.6

0.0
−5

were processed separately, in order to check the


15.6

19.5

representativity of the samples and the reproducibility


0.0
0.0
0.0

0.0
0.0

0.0
9.3
0.0
3.7
0.0
−6

of the measurements. Results obtained are similar (see


Lab sieved

Fig. 8 and Appendix 1). The median diameter (Md),


Total sample weight (g)

(Φ ≥−4)

sorting coefficient (σΦ) and skewness (αΦ) were calcu-


738
628
425
785
477
509
760
591
797
366
632
403

lated applying the formulas of Inman (1952) using the


Sequential Fragmentation/Transport software (©KWARE,
Field sieved
(Φ −6, −5)

University of California 2000).


10125

For vesicularity measurements, 25–26 clasts of the


3630
1335
2035
7960
2250
2250

1315
8105
1425
7735

sieved −4Φ size-fraction for samples from distinct


0

stratigraphic levels were analyzed using the method of


Bed-type
Appendix 1

Houghton and Wilson (1989). To measure water displace-


ment, clasts were coated with a paraffin wax film and
A1

A1
A2
A2
A1
A2
A1

A1
B
B

loaded with a steel ballast. No direct relation between


Sample

particle weight (≈size) and vesicularity was found. This


P31A
P31B
P26
P27
P29
P30

P32
P33
P34
P35
P36
P44

verifies the representativity of the sample size (Houghton


878 Bull Volcanol (2009) 71:859–880

Tachyl.

Grain size, vesicularity, and componentry data collected on scoria beds from the main cone. Grain-sizes and statistical parameters are reported in Φ units. Bed type: A= thick (>60 cm-
thick) beds. A1 =sample from main part. A2 =sample from fine-grained base; B= thin (<60 cm-thick) individual beds (sample from interior). MdΦ= median diameter, σΦ = sorting coefficient,
and Wilson 1989). The density of three dense bomb

17.6

28.4
Crystal-types (%) samples was measured with the same method and used as
a reference to calculate the vesicularity of scoria clasts

Siderom.
(density DRE=2.6 g/cm3). The vesicularity of four large

82.4

71.6
angular clasts ranging in size from 6 to 8 cm (−6Φ) was
also measured for comparison.
494

292
N

Appendix 2: Chemical analyses


Stdev
Vesicularity

αΦ=skewness (calculated using Inman 1952). Avg average, Stdev standard deviation, N number of particles counted, Siderom. sideromelane, Tachyl. tachylite
8.6

6.3
(vol.%)

Whole rock compositional data was measured using


68.7

70.3
Avg

Inductively Coupled Plasma Emission Spectrometry


(ICPES) and Instrumental Neutron Activation Analysis
−0.02

(INAA) at Activation Laboratories, Ancaster, Canada.


0.29
0.46

0.17
0.50
0.51
0.56
αΦ

Crystals and scoria glasses were analyzed by a JEOL


2.45
1.75
1.88
0.90
1.70
2.07
1.58

JXA-8900R electron microprobe at the Laboratorio


σΦ
Statistics

Universitario de Petrología (LUP), Instituto de Geo-


−2.45
−4.60
−1.04
−5.50
−4.85
−4.56
−5.10

física, UNAM, México City. Measuring conditions were


MdΦ

a beam current of 10 nA, an accelerating potential of


20 kV, and a beam diameter of 15 µm for glass and 1–
1.3
0.6
1.5
0.2
0.4
0.4
0.3
>4

5 µm for crystal analysis. During glass analysis, Na and


2.5
1.1
2.4
0.5
0.8
1.0
0.6

K were analyzed first and using short counting times


4

(10 s) to avoid loss by volatilization. Counting times


4.2
1.8
3.9
0.8
1.6
2.0
1.3
3

were 60 s for minor elements (P, Ti, Cr, Ni) and 40 s


5.9
2.4
5.9
1.2
2.4
2.9
1.9

for the other elements. Only analyses with totals


2

>99 wt.% were considered, assuming that the glass


11.5
8.0
2.9

1.3
3.1
3.9
2.6

contained minor amounts of H2O. Basaltic glass standard


1

USNM 111240 VG-2 was mounted on a block and used


24.1
9.3
3.2

1.1
3.2
4.0
3.1

to calibrate Na, Al, Ti, and Fe before each run. This


0

standard was also analyzed at the beginning and end of


12.3

18.4
% Weight per grain-size fraction in Φ

4.1

1.2
3.5
4.1
3.4
−1

each run to monitor the stability of the instrument during


a given run and between different runs, and to check the
18.4

17.1
6.6

1.8
4.5
5.4
5.8
−2

consistency of the data with respect to accepted values. In


19.9
10.8
12.2

general, the standard compositional data obtained was


3.2
7.4
9.5
7.9
−3

consistent with published data (Jarosewich et al. 1980),


18.0
32.5

16.4
30.5
29.8
18.9

although minor elements showed relatively large uncer-


3.0
−4

tainties (average of all analyses with standard deviations:


33.8

51.4
42.4
36.8
54.3

50.41± 0.26 SiO2, 1.87± 0.06 TiO2, 14.08± 0.12 Al2O3,


0.0

0.0
−5

12.28 ± 0.15 FeO, 0.20 ± 0.05 MnO, 6.70 ± 0.16 MgO,


20.9

10.78 ± 0.11 CaO, 2.67 ± 0.10 Na2O, 0.18 ± 0.02 K2O,


0.0
0.0
0.0

0.0
0.0
0.0
−6

0.22± 0.04 P2O5, 99.42 ±0.54 total wt.%). Crystals were


Lab sieved

analyzed using adequate mineral standards. For analyzing


Total sample weight (g)

(Φ ≥−4)

chromium spinel, a chromium oxide (100% Cr2O3) was


438
556
289
680
976
830
405

used.
Field sieved
(Φ −6, −5)

4300

5910
8075
3705
3240

References
0

0
Bed-type
Table (continued)

Beattie P (1993) Olivine–melt and orthopyroxene–melt equilibria.


Contrib Mineral Petrol 115:103–111. doi:10.1007/BF00712982
A1
A2
A1
A1

A1
B

Blackburn EA, Wilson L, Sparks RSJ (1976) Mechanisms and


dynamics of Strombolian activity. J Geol Soc Lond 132:429–440
Sample

Bloomfield K (1975) A late-Quaternary monogenetic volcano field in


P45
P46
P47
P48
P49
P50
P51

central Mexico. Geol Rundsch 64:476–497


Bull Volcanol (2009) 71:859–880 879

Boudreau AE (1999) PELE; a version of the MELTS software Lautze N, Houghton BF (2007) Linking variable explosion style and
program for the PC platform. Comput Geosci 25:201–203. magma textures during 2002 at Stromboli volcano, Italy. Bull
doi:10.1016/S0098-3004(98)00117-4 Volcanol 69:445–460. doi:10.1007/s00445-006-0086-1
Carmichael IAE (2002) The andesite aqueduct: perspectives on the Luhr JF, Simkin T (1993) Paricutin, the volcano born in a cornfield.
evolution of intermediate magmatism in west-central (105–99°W) Geoscience Press, Phoenix
Mexico. Contrib Mineral Petrol 143:641–663. doi:10.1007/ Márquez A, De Ignacio C (2002) Mineralogical and geochemical
s00410-002-0370-9 constraints for the origin and evolution of magmas in Sierra
Cervantes P, Wallace PJ (2003) Magma degassing and basaltic Chichinautzin, Central Mexican Volcanic Belt. Lithos 62:35–62.
eruption styles: a case study of the 2000 yr B.P. eruption of doi:10.1016/S0024-4937(02)00069-5
Xitle Volcano, central Mexico. J Volcanol Geotherm Res Márquez A, Verma SP, Anguita F, Oyarzun R, Brandle JL (1999)
120:249–270. doi:10.1016/S0377-0273(02)00401-8 Tectonics and volcanism of Sierra Chichinautzin: extension at the
Fisher RV, Schmincke HU (1984) Pyroclastic rocks. Springer, New front of the central Trans-Mexican Volcanic Belt. J Volcanol
York Geotherm Res 93:125–150. doi:10.1016/S0377-0273(99)00085-2
Foshag WF, González J (1956) Birth and development of Paricutin Martin Del Pozzo AL (1982) Monogenetic volcanism in Sierra
Volcano Mexico. US Geol Surv Bull 965D:355–489 Chichinautzin, Mexico. Bull Volcanol 45:9–24
Gaffney ES, Damjanac B, Valentine GA (2007) Localization of McBirney AR, Taylor HP, Armstrong RL (1987) Paricutin re-
volcanic activity. Earth Planet Sci Lett 263:323–338. doi:10. examined: a classic example of crustal assimilation in calc-
1016/j.epsl.2007.09.002 alkaline magma. Contrib Mineral Petrol 95:4–20
Giordano D, Russell JK, Dingwell DB (2008) Viscosity of magmatic McGetchin TM, Settle M, Chouet BA (1974) Cinder cone growth
liquids: A model. Earth Planet. Sci Lett 271:123–134. modeled after Northeast Crater, Mount Etna, Sicily. J Geophys
doi:10.1016/j.epsl.2008.03.038 Res 79:3257–3272
Hammer JE, Rutherford MJ (2002) An experimental study of the Montierth C, Johnston AD, Cashman KV (1995) An empirical glass-
kinetics of decompression-induced crystallization in silicic melt. composition-based geothermometer for Mauna Loa lavas. An
J Geophys Res 107:ECV 8-1. doi:10.1029/2001JB000281 Geophys Union Geophys Mon 92:207–217
Hasenaka T, Carmichael ISE (1985) The cinder cones of Michoacán- Mooser F, Nairn AEM, Negendank JFW (1974) Palaeomagnetic
Guanajuato, central Mexico: their age, volume and distribution, investigations of the Tertiary and Quaternary igneous rocks
and magma discharge rate. J Volcanol Geotherm Res 25:105– (VIII): a paleomagnetic and petrologic study of volcanics of the
124. doi:10.1016/0377-0273(85)90007-1 Valley of Mexico. Geol Rundsch 63:451–483
Head J, Wilson L (1989) Basaltic pyroclastic eruptions: influence of Nakamura K (1977) Volcanoes as possible indicators of tectonic stress
gas-release patterns and volume fluxes on fountain structure, and orientation—principle and proposal. J Volcanol Geotherm Res
the formation of cinder cones, spatter cones, rootless flows, lava 2:1–16
ponds and lava flows. J Volcanol Geotherm Res 37:261–271. Ordoñez ME (1947) El volcán de Paricutin. Fantasía, Mexico
doi:10.1016/0377-0273(89)90083-8 Pichavant M, MacDonald R (2007) Crystallization of primitive
Heiken G (1972) Morphology and petrography of volcanic ashes. basaltic magmas at crustal pressures and genesis of the calc-
Geol Soc Am Bull 83:1961–1988 alkaline igneous suite: experimental evidence from St Vincent,
Heiken G (1978) Characteristics of tephra form Cinder Cone, Lassen Lesser Antilles arc. Contrib Mineral Petrol 154:535–558.
Volcanic National Park, California. Bull Volcanol 41:119–130 doi:0.1007/s00410-007-0208-6
Herrmann W, Berry RF (2002) MINSQ—a least squares spreadsheet Pineau F, Shilobreeva S, Kadik A, Javoy M (1998) Water solubility
method for calculating mineral proportions from whole rock and D/H fractionation in the system basaltic andesite–H2O at
major element analyses. Geochem Explor Environ Anal 2:61– 1250°C and between 0.5 and 3 kbars. Chem Geol 147:173–184.
368. doi:10.1144/1467-787302-010 doi:10.1016/S0009-2541(97)00180-0
Houghton BF, Hackett WR (1984) Strombolian and phreatomagmatic Pinkerton H, Stevenson RJ (1992) Methods of determining the
deposits of Oakune Craters, Ruapehu, New Zealand: a complex rheological properties of magmas at sub-liquidus temperatures.
interaction between external water and rising basaltic magma. J J Volcanol Geotherm Res 53:47–66. doi:10.1016/0377-0273(92)
Volcanol Geotherm Res 21:207–231. doi:10.1016/0377-0273(84) 90073-M
90023-4 Pioli L, Erlunda E, Johnson ER, Cashman KV, Wallace PJ, Rosi M,
Houghton BF, Schmincke HU (1989) Rothernberg scoria cone, East Delgado H (2008) Explosive dynamics of violent Strombolian
Eifel: a complex Strombolian and phreatomagmatic volcano. Bull eruptions: the eruption of Paricutin Volcano 1943–1952 (Mexico).
Volcanol 52:28–48. doi:10.1007/BF00641385 Earth Planet Sci Lett 271:359–368. doi:10.1016/j.epsl.2008.04.026
Houghton BF, Wilson CJN (1989) A vesicularity index for pyroclastic Polacci M, Corsaro RA, Andronico D (2006) Coupled textural and
deposits. Bull Volcanol 51:451–462. doi:10.1007/BF01078811 compositional characterization of basaltic scoria: insights into the
Inman DL (1952) Measures for describing the size distribution of transition from Strombolian to fire-fountain activity at Mt Etna,
sediments. J Sed Petrol 22:125–145 Italy. Geology 3:201–204. doi:10.1130/G22318.1
Jarosewich E, Nelen JA, Norberg JA (1980) Reference samples for Porter SC (1972) Distribution, morphology, and size frequency of
electron microprobe analysis. Geostand Newslett 4:43–47 cinder cones on Mauna Kea Volcano, Hawaii. Geol Soc Am Bull
Johnson ER, Wallace PJ, Cashman KV, Delgado Granados H, Kent A 83:3607–3612
(2008) Magmatic volatile contents and degassing-induced crys- Putirka KD (2005) Igneous thermometers and barometers based on
tallization at Volcán Jorullo, Mexico: implications for melt plagioclase+liquid equilibria; tests of some existing models and
evolution and the plumbing systems at monogenetic volcanoes. new calibrations. Am Mineral 90:336–346. doi:10.2138/
Earth Planet Sci Lett 269:478–487. doi:10.1016/j.epsl.2008. 03.004 am.2005.1449
Keating GN, Valentine GA, Krier DJ, Perry FV (2008) Shallow Putirka KD, Perfit M, Ryerson FJ, Jackson MG (2007) Ambient and
plumbing systems for small-volume basaltic volcanoes. Bull excess mantle temperatures, olivine thermometry, and active vs.
Volcanol 70:563–582. doi:10.1007/s00445-007-0154-1 passive upwelling. Chem Geol 241:177–206. doi:10.1016/j.
Klug C, Cashman KV (1996) Permeability development in vesiculat- chemgeo.2007.01.014
ing magmas: implications for fragmentation. Bull Volcanol Riggs NR, Duffield WA (2008) Record of complex scoria cone
58:87–100. doi:10.1007/s004450050128 eruptive activity at Red Mountain, Arizona, USA, and implica-
880 Bull Volcanol (2009) 71:859–880

tions for monogenetic mafic volcanoes. J Volcanol Geotherm Res Tatsumi Y (1982) Origin of high-magnesian andesites in the Setouchi
178:763–776. doi:10.1016/j.volgeores.2008.09.004 volcanic belt, southwest Japan, II. Melting relations at high
Sable J, Houghton BF, Del Carlo P, Coltelli M (2006) Changing pressures. Earth Planet Sci Lett 60:305–317. doi:10.1016/0012-
conditions of magma ascent and fragmentation during the Etna 821X(82)90009-7
122 BC basaltic Plinian eruption: evidence from clast micro- Tatsumi Y, Ishizaka K (1982) Origin of high-magnesian andesites in
textures. J Volcanol Geotherm Res 158:333–354. doi:10.1016/j. the Setouchi volcanic belt, southwest Japan, I. Petrological and
jvolgeores.2006.07.006 chemical characteristics. Earth Planet Sci Lett 60:293–304.
Schaaf P, Stimac J, Siebe C, Macías JC (2005) Geochemical evidence doi:10.1016/0012-821X(82)90008-5
for mantle origin and crustal processes in volcanic rocks from Thomas RME, Sparks RSJ (1992) Cooling of tephra during fallout
Popocatépetl and surrounding monogenetic volcanoes, Central from eruption columns. Bull Volcanol 54:542–553. doi:10.1007/
Mexico. J Petrol 46:1243–1282. doi:10.1093/petrology/egi015 BF00569939
Self S, Sparks RSJ, Booth B, Walker GPL (1974) The 1973 Heimaey Valentine GA, Krogh KEC (2006) Emplacement of small dikes
strombolian scoria deposit, Iceland. Geol Mag 111:539–548 and sills beneath a small basaltic volcanic center—the role of
Siebe C (2000) Age and archaeological implications of Xitle volcano, pre-existing structure (Paiute Ridge, southern Nevada, USA).
southwestern Basin of Mexico City. J Volcanol Geotherm Res Earth Planet Sci Lett 246:217–230. doi:10.1016/j.epsl.2006.
104:45–64. doi:10.1016/S0377-0273(00)00199-2 04.031
Siebe C, Schaaf P, Urrutia-Fucugauchi J (1999) Mammoth bones Valentine GA, Keating GN (2007) Eruptive styles and inferences
embedded in a late Pleistocene lahar from Popocatépetl Volcano, about plumbing systems at Hidden Cone and Little Black Peak
near Tocuila, central Mexico. Geol Soc Am Bull 111:1550–1562 scoria cone volcanoes (Nevada, U.S.A.). Bull Volcanol 170:105–
Siebe C, Rodríguez-Lara V, Schaaf P, Abrams M (2004a) Radiocarbon 113. doi:10.10007/s00445-007-0123-8
ages of Holocene Pelado, Guespalapa, and Chichinautzin scoria Valentine GA, Gregg TKP (2008) Continental basaltic volcanoes—
cones, south of Mexico City: implications for archeology and future processes and problems. J Volcanol Geotherm Res 177:857–873.
hazards. Bull Volcanol 66:203–225. doi:10.1007/s00445-003-0304-z doi:10.1016/j.jvolgeores.2008.01.050
Siebe C, Rodríguez-Lara V, Schaaf P, Abrams M (2004b) Geochemistry, Valentine GA, Krier D, Perry FV, Heiken G (2005) Scoria cone
Sr–Nd isotope composition, and tectonic setting of Holocene construction mechanisms, Lathrop Wells volcano, southern
Pelado, Guespalapa, and Chichinautzin scoria cones, south of Nevada, USA. Geology 33:629–632. doi:10.1130/G21459.1
Mexico City. J Volcanol Geotherm Res 130:197–226. doi:10.1016/ Valentine GA, Perry FV, Krier D, Keating GN, Kelley RE, Cogbill
S0377-0273(03)00289-0 AH (2006) Small-volume basaltic volcanoes: eruptive products
Siebe C, Arana-Salinas L, Abrams M (2005) Geology and radiocarbon and processes, and posteruptive geomorphic evolution in Crater
ages of Tláloc, Tlacotenco, Cuauhtzin, Hijo del Cuauhtzin, Flat (Pleistocene), southern Nevada. Geol Soc Am Bull
Teuhtli, and Ocusacayo monogenetic volcanoes in the central part 118:1313–1330. doi:10.1130/B25956.1
of the Sierra del Chichinautzin, México. J Volcanol Geotherm Res Valentine GA, Krier DJ, Perry FV, Heiken G (2007) Eruptive and
141:225–243. doi:10.1016/j.jvolgeores.2004.10.009 geomorphic processes at the Lathrop Wells scoria cone volcano. J
Siebert L, Carrasco-Núñez G (2002) Late-Pleistocene to precolombian Volcanol Geotherm Res 161:57–80. doi:10.1016/j.jvolgeores.
behind-the-arc mafic volcanism in the eastern Mexican Volcanic 2006.11.003
Belt: implications for future hazards. J Volcanol Geotherm Res Vetere F, Behrens H, Holtz F, Neuville DR (2006) Viscosity of
115:179–205. doi:10.1016/S0377-0273(01)00316-X andesitic melts—new experimental data and a revised calculation
Sieron K, Siebe C (2008) Revised stratigraphy and eruption rates of model. Chem Geol 228:233–245. doi:10.1016/j.chemgeo.2005.
Ceboruco stratovolcano and surrounding monogenetic vents 10.009
(Nayarit, México) from historical documents and new radiocar- Walker GPL (1973) Lengths of lava flows. Phil Trans R Soc Lond A
bon dates. J Volcanol Geotherm Res 176:241–264. doi:10.1016/j. 274:107–118
jvolgeores.2008.04.006 Wallace PJ, Carmichael ISE (1999) Quaternary volcanism near the
Sugawara T (2000) Thermodynamic analysis of Fe and Mg partition- Valley of Mexico: implications for subduction zone magmatism
ing between plagioclase and silicate liquid. Contrib Mineral and the effects of crustal thickness variations on primitive magma
Petrol 138:101–113. doi:10.1007/s004100050011 compositions. Contrib Mineral Petrol 135:291–314. doi:10.1007/
Taddeucci J, Pompilio M, Scarlato P (2004) Conduit processes during s004100050513
the July–August 2001 explosive activity of Mt Etna (Italy): White J, Houghton B (2006) Primary volcaniclastic rocks. Geology
inferences from glass chemistry and crystal size distribution of 34:377–680. doi:10.1130/G22346.1
ash particles. J Volcanol Geotherm Res 137:33–54. doi:10.1016/j. Wood CA (1980) Morphometric evolution of cinder cones. J Volcanol
jvolgeores.2004.05.011 Geotherm Res 7:387–413

Vous aimerez peut-être aussi