Benchmarks and sediment source(s) of the 1755 Lisbon tsunami

deposit at Boca do Rio Estuary

Eric Font
a,
, Cristina Veiga-Pires
b
, Manuel Pozo
c
, Slvia Nave
d
, Susana Costas
d
, Francisco Ruiz Muoz
e
,
Manuel Abad
e
, Nuno Simes
b
, Slvia Duarte
a
, Joaqun Rodrguez-Vidal
e
a
IDL-FCGUL, Faculdade de Cincias de Lisboa, Ed. C3.3.22, Campo Grande, 1749-016, Portugal
b
CIMA-FCT, Universidade do Algarve, Campus de Gambelas, 8005-139 Faro, Portugal
c
Departamento de Geologa y Geoqumica, Universidad Autnoma de Madrid, 28049-Madrid, Spain
d
Laboratrio Nacional de Energia e Geologia (LNEG), Apartado 7586, Alfragide, 2721-866, Portugal
e
Departamento de Geodinmica y Paleontologa, Facultad de Ciencias Experimentales, Campus de El Carmen, Avda. Tres de Marzo, 21071-Huelva, Spain
a b s t r a c t a r t i c l e i n f o
Article history:
Received 17 October 2012
Received in revised form 7 June 2013
Accepted 8 June 2013
Available online 20 June 2013
Communicated by J.T. Wells
Keywords:
tsunami deposit
estuary
geochemical proxies
mineralogy
hydrodynamics
Standardizing the signature of tsunami deposits has been identied as a major limitation for the identica-
tion of paleo-tsunami deposits. This limitation mostly arises from the strongly source-dependent nature of
these deposits, which in turn determines their composition and depositional architecture, and from the effect
of the local morphology of the corresponding depositional environment. Here, we provide new high-
resolution mineralogical, geochemical and micro/macrofauna data of the 1755 tsunami layer of Boca do Rio
estuary (Algarve, Portugal) with the aim of unraveling the signatures of estuarine tsunami deposits and
linking them to possible sediment sources. We also apply for the rst time diffuse reectance spectropho-
tometry (DRS) analysis. Our results show that the 1755 tsunami deposit of the Boca do Rio estuary is featured
by an enhancement in Sr and Ca, which are linked to the input of biogenic and detrital carbonates (shell frag-
ments and limestone clasts) from the beach foreshore and a strong depletion in most terrestrial- and
marine-sensitive indicators. The latter is interpreted as resulting from the reworking of the estuarine clays
and subsequent dilution within a huge volume of sand eroded from the coastal barrier. It conrms that in
the case of the Boca do Rio estuary, the sediment source is essentially proximal and coastal. Textural and min-
eralogical features between the base and the top of the tsunami layer suggest the imprint of run-up and back-
wash currents derived from a unique wave. Micro and macrofauna analysis and DRS data of the siliciclastic
fraction show slight but signicant environmental changes occurring just after the tsunami, which could be
provoked by an eventual closure of the estuary mouth.
2013 The Authors. Published by Elsevier B.V. All rights reserved.
1. Introduction
In the last decades, numerous investigations have focused on re-
solving tsunami deposit benchmarks in order to identify sediment
sources and to evaluate their impact as natural hazards. However,
if recent tsunamis are well described on the basis of historical records
(e.g. Delange and Healy, 1986; Baptista et al., 1998; Papadopoulos,
2003; Dominey-Howes, 2007; Baptista and Miranda, 2009; Ambraseys
and Synolakis, 2010), the identication of paleotsunami deposits is in
counterpart more controversial since it reposes on geological evidences
that can be shared by other high-energy events such as storm-induced
deposits (e.g. Cundy et al., 2000; Pratt, 2002; Goff et al., 2004a;
Kortekaas and Dawson, 2007; Morton et al., 2007; Tappin, 2007;
Barbano et al., 2010; Chague-Goff, 2010; Goff et al., 2012; Ramirez-
Herrera et al., 2012). To attempt to solve this problem, geologists have
accessed a wide range of geological, geophysical and geochemical prox-
ies to distinguish between storm- and tsunami-induced deposits,
improving the so-called tsunami proxy toolkit (see review in
Chague-Goff et al., 2011). Sedimentological (i.e. grain size), macro-
and micropaleontological and geomorphological features are the most
conventional proxies (see review in Morton et al., 2007; Chague-Goff
et al., 2011; Goff et al., 2012). Geochemical approaches are less explored
despite the fact that chemical composition of interstitial waters and
sediment fromcoastal and lagoonal environments has beensuccessfully
used as an indicator of tsunami inundation (Andrade et al., 2003;
Chague-Goff, 2010; Chague-Goff et al., in press). Also, magnetic ap-
proaches (i.e. bulk rock magnetic properties and anisotropy of magnetic
susceptibility) were recently applied to modern tsunami deposits
Marine Geology 343 (2013) 114
This is an open-access article distributed under the terms of the Creative Commons
Attribution-NonCommercial-No Derivative Works License, which permits non-commercial
use, distribution, and reproduction in any medium, provided the original author and source
are credited.
Corresponding author at: IDL-UL, Universidade de Lisboa, Edifcio C8-8.3.22, Campo
Grande, 1749-016 Lisboa, Portugal. Tel.: +351 217500811.
E-mail address: font_eric@hotmail.com (E. Font).
0025-3227/$ see front matter 2013 The Authors. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.margeo.2013.06.008
Contents lists available at SciVerse ScienceDirect
Marine Geology
j our nal homepage: www. el sevi er . com/ l ocat e/ mar geo
(Andrade et al., 2003; Font et al., 2010; Wassmer et al., 2010) but still
need to be expanded and calibrated in other case studies worldwide.
Although a signicant database of tsunami benchmarks is now
available, the identication of tsunami deposits is still ambiguous.
This is because their signatures depend essentially on the composi-
tion and texture of the sedimentary source. It is therefore crucial to
connect tsunami deposit benchmarks with their environmental and
geomorphological settings and to link them to sediment source(s).
Here, we provide new geochemical and mineralogical data of the
1755 Lisbon earthquake tsunami deposit of the Boca do Rio estuary
(Algarve, Portugal) in order to i) improve the tsunami toolkit; ii) iden-
tify sediment source(s); iii) connect the sub-units of the deposit to
distinct hydrodynamics regime and iv) study environmental changes
of the estuary after the 1755 tsunami. The 1755 Lisbon tsunami has
been extensively described in the historical records and the deposit is
well studied in terms of sedimentological, paleoecological, geochemical
and magnetic properties in the Boca do Rio estuary (Dawson et al.,
1995; Hindson and Andrade, 1999; Oliveira et al., 2009; Font et al.,
2010; Costa et al., 2012).
Our new mineralogical and geochemical data match previous results
and provide new clues to identify sediment source(s) of the 1755 tsuna-
mi deposit andits geological signature. Inaddition, grainsize andtextural
data yield new hydrodynamic constraints that help in distinguishing
overwash from backwash currents within the sub-unit of the tsunami
layer. Finally, changes in composition and grain size between pre- and
post-tsunami estuarine clays witness the impact of the tsunami wave
on the geomorphology and environmental conditions. Accordingly, we
proposed an environmental model for other bar-built estuaries and dis-
cuss the use of geochemical proxies as indicator of estuarine tsunami
deposit.
2. Geological settings and sampling
The tsunami deposit induced by the 1755 Lisbon earthquake was
found in the region of Algarve, Southern Portugal, in the estuary
of Boca do Rio (Dawson et al., 1995; Hindson and Andrade, 1999;
Hindson et al., 1999; Andrade et al., 2003; Costa et al., 2011, 2012)
(Fig. 1). The estuary corresponds to a supratidal ood plain separated
fromthe sea by a rocky and sandy barrier richin boulders and shell frag-
ments. Regarding the lithostratigraphy of the estuary, we followed the
lithofacies proposed by Hindson and Andrade (1999), namely: a grayish
marine sand unit with shell fragments at the base (unit D) and a tsuna-
mi deposit (sand and gravel, unit B) intercalated between reddish silty
clay (units A and C; Fig. 1). The tsunamigenic origin of unit B is con-
rmed by i) presence of coarse sandand gravels indicating anabrupt in-
terruption in the low-energy sedimentation of the estuary, ii) presence
of clay balls ripped from the over washed surface of the estuary
mud-ats, and iii) occurrence of different marine species contrasting
with the estuarine setting of the depositional environment and pres-
ence of limestone clasts (Hindson and Andrade, 1999; Hindson et al.,
1999; Costa et al., 2011). The tsunami deposit was dated between
1730 60 and 1800 80 yr AD by thermoluminescence and optical-
ly stimulated luminescence (Dawson et al., 1995). More recently, Cunha
et al. (2010) presented quartz optical dating and
14
C dating of the Boca
do Rio and Martinhal estuaries. They showed that the data were re-
markably reproducible within the same site but overestimated in the
1755 tsunami deposit by up to 120230 yr. They attributed the dispar-
ity due to fast erosion and deposition of older sediments carried from
near offshore.
Lateral facies and texture variations of the 1755 tsunami deposit at
Boca do Rio have been extensively studied by Hindson and Andrade
(1999) who globally divided it into sub-unit B1 corresponding to a
level rich in limestone clasts and mud balls and sub-unit B2 without
limestone clasts. The authors showed that despite the decreasing
thickness of the tsunami from downstream to upstream, the succes-
sion of B1 at the base and B2 at the top is almost systematic. In
addition, we present mass specic magnetic susceptibility data
(reecting variations in clay and iron oxide content) of four holes
along the NS transect in the estuary and show that results are repeat-
able in all proles (Fig. 1C). It means that any high-resolution log-
prole from a hole is expected to be repeatable in several others sites
from the estuary. In order to complement what has been done at Boca
do Rio, we thus opt to study a single hole but compensated by providing
multidisciplinary and high-resolution datasets and by testing new
methods such as diffuse reectance spectrophotometry. Samples were
collected from the trench T4 using 3 gutters of 5 cm in width.
3. Methods
Mass specic magnetic susceptibility data presentedat Fig. 1 were ac-
quiredusing a MFK1 (AGICO) (see Methodin Font et al., 2010). X-ray dif-
fraction (XRD) was carried out in 11 samples (bulk sample and clay
fraction b2 m) using a SIEMENS D-5000. Powdered (b63 m) bulk
samples were scanned from2 to 65 2. Clay fraction samples were pre-
pared from suspensions oriented on glass slides air-dried, with ethylene
glycol solvation and after heating at 550 C. Quantitative estimation was
carried out using the intensity factors calculated by Schultz (1964) and
van der Marel (1966).
Geochemical data were obtained on sediment samples by three
techniques: ICP (AES and MS), AAS and XRF. Chemical analysis of
Si, Al, Mg, Fe, K, Mn, Ti, P and Ca was performed by means of
the MagiX X-ray uorescence (XRF) spectrometer of PANalytical. A
Varian 220-FSQU-106 atomic absorption spectrometer (AAS) was
used for the determination of Na. Samples were analyzed after glass
disk elaboration from molten mixture with lithium tetraborate. Loss
on ignition (LOI) was determined at 950 C. Trace element concentra-
tions were determined by inductively coupled plasma spectrometry
in both ICP-AES and ICP-MS instruments. Samples were analyzed
after standard procedure by acid digestion (HF, HNO3 and HCl).
Carbon, hydrogen, nitrogen and sulphur of 11 samples were analyzed
by means of LECO CHNS-932.
Atotal of 22 semi-quantitative XRF analyses (cps) were performedto
identify trends at higher-resolution (Croudace et al., 2006; Rothwell and
Rack, 2006; Martin-Puertas et al., 2011). Bulk samples were freeze dried
overnight and hand crushed in an opal mortar. Powdered samples
(~200 mg) were irradiated using an automated wavelength dispersive
spectrometer Philips PW 1400 equipped with a rhodium tube to collect
ve successive counting reads (30 s xed counting-time period).
Diffuse reectance spectrophotometry (DRS) is a rapid and non-
destructive technique based on the percent reectance of a sample
relative to white light. It provides qualitative information about min-
eralogical and grain size variations (Adkins et al., 1997; Ji et al., 2001).
DRS approaches have been applied to a wide range of studies (Clark et
al., 1990; Adkins et al., 1997; Ji et al., 2004) but has never been tested
in the case of tsunami-induced deposits. Reectance spectra data are
conventionally represented in the CIE (Commission Internationale de
l'clairage) 1976 L*a*b* space (Pauli, 1976) where L* is the axis
representing lightness/darkness, a* is the axis of redness/greenness
and b* is the axis of yellowness/blueness. Here, and every 2 cm on
the sediment core, three measurements of color data were obtained
using a X-Rite Colourtron spectrophotometer.
Grainsize was measured using a MalvernMastersizer Microlaser an-
alyzer counting particles ranging from0.5 to 300 m(11 to 1.75 ) and
on a Coulter LS230 laser particle analyzer accounting for the size frac-
tion between 0.04 and 2000 m (~15 to 1 ), i.e. from very ne
clay to very coarse sand size fraction (Blott and Pye, 2012). Bulk sedi-
ments analyzed with the Malvern method were rst processed after
adding powdered sodiumhexametaphosphate (1 g/dm
3
) and mechan-
ical agitation for deocculation. The mineral or inorganic fraction,
analyzed with both the Malvern and the Coulter counter methods,
was then analyzed after organic matter removal through peroxide oxi-
dation. Finally, the siliciclastic fraction, also only analyzed with the
2 E. Font et al. / Marine Geology 343 (2013) 114
Malvern method, was analyzed after removing the carbonated fraction
by successive 10% HCl acidic attacks. This allows discussing the grain
size of the ne terrestrial fraction (estuarine clays) by avoiding the de-
trital and biogenic carbonate fraction from the dune and surrounding
cliff.
Micro and macrofauna analysis were performed on 10 samples
(25 g). Bulk samples were washed through a 1-mm sieve. Bivalves
and gastropods were identied to species level and counted for
semi-quantitative analysis. The abundance of a species per sample
was described as follows: very rare (one specimen), rare (25 speci-
mens), frequent (510 specimens) or very frequent (10 specimens
or more). For microfossil analysis, 10 samples (10 g) were washed
through a 63-m sieve to remove the mud fraction before drying.
Both the abundance of each species and the number of individuals
per gram in each sample were calculated.
4. Results
4.1. Grain size analysis
Grain size analysis from ~0.02 to 2000 m of the Boca do Rio tsuna-
mi deposit has been already conducted and published by Hindson et al.
(1996) and Hindson and Andrade (1999). In order to correlate our re-
sults with those aforementioned we conducted grain size analysis in
Fig. 1. A) Geomorphological map of the Boca do Rio estuary; B) photograph of one of the gutter sampled from trench 4. C) Mass specic magnetic susceptibility proles of several
boreholes from Font et al. (2010) (T1 to T3) and this study (T4) showing that the same trend in bulk mineralogical composition is observable along the whole estuary.
3 E. Font et al. / Marine Geology 343 (2013) 114
the range b2000 m using Coulter granulometer. However, our main
goal is to increase the resolution of the b300 m fraction, which corre-
sponds to the range responsible for the bulk magnetic signal, i.e. clays
and iron oxides (Font et al., 2010; Hateld et al., 2010), by using a
Malvern granulometer. Results are illustrated at Fig. 2 (see also Tables
1 and 2 at Supplementary data).
Results of the b2000 m fraction in sub-unit B1 show an inversely
graded layer, that is an upward increase in mean grain size due to en-
hancement in medium sand (350 m, Fig. 2A). The increase in medi-
um sand is coincident with a reduction in silt (Fig. 2) whereas coarse
and very coarse sands are less abundant showing a more uniform
presence (Fig. 2). This lower unit also shows a signicant difference
in grain size distribution in the ne sediment analysis after the
removal of the carbonated fraction (Fig. 2A). The upper sub-unit
(B2) corresponds instead to an upward ning sequence dominated
by a gradual increase in very ne sands (100 m) and silts (Fig. 2) ob-
servable in both the b300 and b2000 m fractions. This gradual in-
crease in ner sediments is interrupted to the top by the entrance
of coarser sediments (medium and coarse sand) at the top end of
sub-unit B2, causing a net increase of the mean grain size in both
the b300 and b2000 m fractions (Fig. 2). The transition from B1 to
B2 is represented by the increase of ne and very ne sands and the
absence of medium and coarse sands which translates into a reduc-
tion of the mean grain size from 400 to 100 m (Fig. 2B).
Coulter counter (b2000 m) and Malvern (b300 m) sediment
grain size are presented in Fig. 2B. The tsunami layer presents a
Fig. 2. Grain size, sorting and mean (in m). Upper panel shows the grain size distribution for the entire trench, while lower panel shows the distribution of the different grain size
fractions within the tsunami deposit. Grain sizes of ne inorganic and siliciclastic fractions are obtained after organic matter removal through peroxide oxidation and after removing
the carbonated fraction by HCl, respectively.
4 E. Font et al. / Marine Geology 343 (2013) 114
sharp contact with underlying unit C and exhibits very poor and var-
iable sorting values, ranging between 3 and 12 mmostly due to mul-
timodal distribution (Fig. 2). Note that low values in mean grain size,
sorting and sand content observed at 74 cm in the b300 m analyses
are due to the presence of a mud ball in the middle of unit B. This in-
terpretation is consistent with visual observation of the core. On the
other hand, surrounding units C and A can be characterized as medi-
um silt after removal of the organic content, with the upper 40 cm of
unit A becoming ner with a mean grain size of ne silt (Fig. 3). Fur-
thermore, although these two units are both classied as medium silt,
their siliciclastic mean grain size and sorting, as well as their grain
size distribution, are statistically different (n = 52, p b 0.001). Sum-
marizing, the grain size population of the tsunami deposit suggests
the occurrence of three modes within the sand domain; a fourth
mode would be represented by the limestone clasts and gravels at
the base of the deposit, and nally a fth mode represented by
muddy sediments. This would correspond to the ve sub-units al-
ready described in Hindson and Andrade (1999).
4.2. Mineralogy
4.2.1. Bulk mineralogy
Mineralogical results are obtained using X-ray diffraction and are
shown in Fig. 3 (see also Table 3 at Supplementary data). The bulk
sample mineralogy shows two different assemblages: Assemblage 1
is comprised of phyllosilicatesquartz(feldspars, dolomite, calcite).
Assemblage 2 consists of phyllosilicates, quartz, calcite, dolomite
(aragonitefeldspars). Assemblage 1 is observed in silty clays from
units A and C. Phyllosilicates (6679%) and quartz (1930%) with
subordinate feldspars (K-feldspar > plagioclase) and local traces of
calcite and dolomite are noted. Assemblage 2 occurs in unit B (i.e. tsu-
nami deposit) and is made-up of a lower content in phyllosilicates
(4061%) with quartz (3237%), subordinated calcite (213%) and
dolomite (411%). Traces of feldspars (K-feldspar > plagioclase)
and aragonite are also identied.
4.2.2. Clay mineralogy
The study of the clay fraction (b2 m) revealed the existence of
three main minerals, which, after the appropriate treatments, led us
to infer the existence of a predominant swelling phase and two highly
subordinated non-swelling phases (Fig. 3; Table 3 at Supplementary
data). Indeed, the study of air-dried oriented aggregates revealed
d-spacing with a very intense reection at 1415 and two smaller
reections at 10.01 and 7.157.17 , respectively. Treatment with
ethylene glycol caused the reection at 1415 to shift to 16.80 ,
which revealed a 2:1 swelling phyllosilicate. Treatment at 550 C re-
vealed a shift in reection to 9.9 , which was consistent with a struc-
tural collapse process in a mineral with the interlayer space occupied
by hydrated exchangeable cations. All the identied features pointed
to smectite as the swelling mineral of the sample. The 10 clay min-
erals showed no changes in its basal d-spacing after application of the
above described treatments, remaining around 9.9 with ethylene gly-
col and after thermal treatment, being identied as illite. The third clay
mineral around 7 showed no changes in its basal d-spacing after
treatment with ethylene glycol but disappears after thermal treatment
identifying kaolinite. In all the samples traces of irregular illitesmectite
mixed layers are recognized.
In the two differentiated assemblages, the clay mineralogy associ-
ation is composed of illite (6572%), smectite (1628%) and kaolinite
(up-to 10%). No special trends have been detected but on the basis of
the phyllosilicate content these clay minerals accumulate mostly in
Assemblage 1. Indeed in the bulk sample of this assemblage the calcu-
lated content varies between 45 and 58% for illite, 12 and 21% for
smectite and 4 and 8% for kaolinite. Whereas in Assemblage 2 a rela-
tively lower content in illite (2942%), smectite (815%) and kaolinite
(34%) is observed.
Froma genetic point of viewthe mineralogical assemblages are con-
sidered inherited from the adjacent clay units whereas the presence of
calcite and aragonite is chiey associated to bioclastic remains.
4.3. Geochemistry
4.3.1. Normalization
For a short discussion of the use of normalization of geochemical
elements, we refer here to the work of Chague-Goff (2010). For the
present study, we represent original and Al
2
O
3
normalized data
(Figs. 45; Tables 4, 5 and 6 at Supplementary data). Original data
better show the connection of the tsunami deposit with the underly-
ing clayish units, whereas Al
2
O
3
normalized data rather inform on
material from an additional and distinct source.
4.3.2. Major elements
Major elements contained in the tsunami deposit show a consider-
able enhancement in Ca content in relation to the adjacent clayish
units. SiO
2
content slightly increases (45%) in the tsunami deposit
while other elements occur inmuchlower concentrations (Fig. 4). Com-
positions in geochemical elements globally followed a couplet of two
gradual trends observed within each sub-unit: i) an increase in major
and trace elements from the base until the top of sub-unit B1, followed
by ii) a decrease fromthe base to the top of sub-unit B2. Ca is essentially
found as calcite and/or aragonite (detrital and biogenic) while Si origi-
nates from detrital siliciclastic input (sand and clays). Na and Cl are
slightly depleted in the tsunami deposit but show a single positive
peak at the base of the tsunami deposit (Fig. 4). Depletion in Fe and Ti,
essentially found as rutile (TiO2) and titanomagnetite (Fe
2-x
Ti
x
O
3
),
matches the low values of magnetic susceptibility observed in the tsu-
nami deposit (Font et al., 2010). Concerning organic matter distribution
the elemental analysis (CHNS) suggests a depletion of organic carbon
but enrichment in inorganic carbon in the tsunamigenic deposit sam-
ples (Fig. 4). This interpretation is supported by the decrease of N and
H despite the fact that the higher content in C is mostly due to carbon-
ates (bioclasts and rock fragments). S is poorly concentrated in all units
(Fig. 4).
These results suggest that the main geochemical signatures of the
tsunami deposit are originated from inherited phyllosilicates contained
in the adjacent clayish units and diluted within a huge volume of
non-magnetic minerals (i.e. calcite and silicon). Quantitative data nor-
malized by Al
2
O
3
conrm the connection between the signature of the
tsunami deposit and those of the mudat. These results suggest an
external source of SiO
2
and CaO (shell fragment and detrital calcite
and silicate). This is also demonstrated by bivariate analysis that Fig. 3. Mineralogical composition of bulk sample and clay fraction (wt.%).
5 E. Font et al. / Marine Geology 343 (2013) 114
displays an excellent positive correlation (R
2
> 0.95) between Al
2
O
3
and major elements commonly associated with inherited minerals
(TiO
2
, K
2
O, Na
2
O, Fe
2
O
3
) (Fig. 5). Conversely, SiO
2
andCaOdisplay a neg-
ative correlation with Al
2
O
3
reaching a correlation coefcient of 0.95. In
addition to inherited phyllosilicates fromthe clays and biogenic/detrital
carbonates from the beach, a slight but notable increase in other ele-
ments (K
2
O, Na
2
O and MgO) is observed (Fig. 5). MgO is mostly associ-
ated to dolomite while Na
2
O and K
2
O can be related to micaceous
minerals and feldspars. However the role played by salts forming from
seawater is not discarded.
4.3.3. Trace elements
Trace element analyses match major element and mineralogical
interpretations by showing a relative depletion in some marine- and
continental-sensitive elements (Br, Ba, Zr in cps) in the tsunami
deposit (Fig. 4). Conversely, Sr is enhanced in the tsunami deposit.
This is consistent with the Ca trend, which suggests a biogenic origin
(shell fragments). Relative depletioninBa and Br, whichare usually con-
sidered reliable indicators of marine seawater invasion (Chague-Goff,
2010), is better explained by reworking of the estuarine clays and subse-
quent dilution within the huge volume of sand and carbonates fromthe
beach. The same is valid for Zr that mimics Fe, Ti and Al trends as well as
ne sand and silt contents.
Quantitative data (in g/g) conrmthat the signature of the tsunami
deposit is mostly associated with inherited minerals including TTEs
(transition trace elements), REEs (rare earth elements and Y), Rb and
Ba (Fig. 4). Al-normalized values show little variation (except for Sr),
being interpreted as associated to clay fraction (phyllosilicates) and/or
feldspars. In counterpart, the tsunami deposit is enhanced in Sr
with slight increases in U, Th, As and Pb (Fig. 4). Increase in Sr con-
tent is easily explained by the input of biogenic carbonate whereas
increase in the others elements can result from the input of a distinct
terrestrial source, anthropogenic inuence and/or post-depositional
mobility.
Fig. 4. Geochemical data. Major elements (above): quantitative data (wt.%) and trends in marine seawater (Ca, Na, Cl) and terrestrial/high-energy (Al, Fe, Ti, Zr) indicators (in cps).
Elemental analysis (C, N, H, S) is also shown. Trace elements (below): trends in marine seawater (Sr, Ba, Br) and high-energy (Zr) indicator (in cps) and quantitative Al
2
O
3
-
normalized trace elements.
6 E. Font et al. / Marine Geology 343 (2013) 114
4.4. Diffuse reectance spectrophotometry
DRS results for the bulk sediment and the siliciclastic fraction are
presented in Fig. 7 and show the mean values of the three measure-
ments with their associated standard deviations. Results from the bulk
sediment samples show very small differences in color between the
three sedimentary units. The tsunami layer seems to be slightly clearer,
exhibiting a small increase in L* values, which is in accordance with the
decrease in organic matter content and increase in calcite or carbonated
elements (see Section 4.3). This layer also exhibits much more variance
in the value for the three color components, expressed by the large
standard deviations. This characteristic agrees with the sedimentary
heterogeneity, in grain size and origin of particles, of these specic
high-energy sedimentary deposits.
Conversely, the DRS results of the siliciclastic fraction present sig-
nicant differences (n = 52, p b 0.001) between the three different
sedimentary units (Figs. 67). Units B and A also present differences
between the lower and upper parts. From a more specic point of
view, unit C is the darker, redder and yellower unit, which seem to
correspond also to the characteristics of the lower part of the tsunami
unit (unit B). In contrast the upper part of this unit seems to be a tran-
sition layer to the lower part of unit A, becoming progressively ligh-
ter, less red and less yellow (Fig. 6). The upper part of unit A shows
color stabilization along the top ~30 cm. It is important here to
make two points. Firstly, the standard deviations observed for the
siliciclastic fraction are very small and consistent all along the core,
even in the tsunami layer. This suggests that this fraction is pure
or homogeneous in each sample, in contrast to what was previously
observed for bulk sediment samples. Secondly, the siliciclastic frac-
tion present in the tsunami layer seems to correspond to a mixture
between two end-members that correspond to the siliciclastic frac-
tions of the underlying and overlying sedimentary units (Fig. 7).
4.5. Macro and micropaleontology
Macro and microfaunal analysis of trench T4 identied three main
units from the base to the top (Fig. 8):
Unit C (just below the tsunami deposit) reveals scarce foraminifera
specimens (Ammonia beccarii, Quinqueloculina seminulum, Elphidium)
and ostracods (Loxoconcha rhomboidea). Some fragments of the bi-
valve Mactra stultorum have been found in the upper part of these
clays. This foraminifera assemblage is similar to those found in
other estuaries from Southwestern Spain, in marine areas close to
river mouths (salinity: 3436) (Gonzalez-Regalado et al., 2001).
In these zones, the ostracod L. rhomboidea lives in the channel
margins or the bottom of very shallow tributary channels connected
with the main channels of estuaries (Ruiz et al., 2000).
These environments present usually a clayey/silty substrate such as
the one observed here (Fig. 2).
Unit B (tsunami deposit) can be divided into two intervals. The
base (B1) of the deposit includes large specimens (17 cm long) of
disarticulated bivalves (Mytilus edulis, Mactra stultorum, Glycymeris
glycymeris, Ostrea edulis, Acanthocardia tuberculata) and gastropods
(Hinia reticulata, Calyptraea chinensis, Rissoa sp.). Most of them
are partially fragmented or present evidences of reworking, with
rounded margins or bioerosional marks in the valves or carapaces.
Foraminifera (Elphidium macellum, Heterolepa bellincionii, Pullenia
bulloides) and ostracods (adults of Loxoconcha elliptica) are rare,
whereas some thick and ne spines of echinoderms are also present.
The large size of bioclasts and the presence of reworked species
indicate transport under turbulent conditions. The bivalve record
indicates a marine origin although most of themhave been probably
transported from adjacent beaches or dunes, as attested by bio-
erosional structures and rounded margins that imply a previous, in-
tense reworking or a feedback of valves by the perforating organism
after death. Similarly, presence of the estuarine brackish ostracod
L. elliptica also indicates reworking of the underlying estuarine clays.
The overlain interval (B2) is characterized by juvenile shells (b2 cm
in diameter) of bivalves (Macoma melo, Mactra stultorum, Mactra
glauca), some rare gastropods (Calliostoma zyzyphinum), a more di-
verse foraminiferal assemblage (Elphidium spp., Lobatula lobatula,
Quinqueloculina seminulum, Triloculina trigonula), adults and juve-
nile instars of ostracods (Loxoconcha elliptica) and numerous frag-
ments (plates, thick and ne spines) of echinoderms (Echinocorys?
Paracentrotus?). This layer may originate from erosion of nearby
sandy deposits (beaches or dunes) and small, internal estuarine
channels with populations of L. elliptica. The presence of juvenile
instars of this specie implies a suspension transport. The relative de-
crease in grain size (including smaller bioclasts of molluscs) relative
to the B1 suggests a decrease in the ow velocity. Note the absence
of the infralittoral ostracod assemblages that inhabit the adjacent
marine areas of Southern Portugal and Spain (Ruiz et al., 1997;
Cabral et al., 2006) in the whole tsunamigenic deposits, which
could indicate a nearby source for these materials in this case. This
paleontological pattern (lower interval: scarce micropaleontological
record; upper interval: diverse micropaleontological assemblage)
and the decrease in mean grain size have been detected in other
tsunamigenic layers deposited by the great Chilean earthquake of
February 2010 (Horton et al., 2011).
Unit A(overlaying the tsunami deposit) canbe dividedinto three sec-
tions. The lowermost interval (A1) is characterized by the presence of
Fig. 5. Bivariate plots of major elements versus Al
2
O
3
(%) showing a positive correlation of terrestrial inherited elements between the tsunami deposit and the estuarine clays (on
the left) and a negative correlation of CaO and SiO
2
(on the right) indicating a distinct origin (i.e. dune).
7 E. Font et al. / Marine Geology 343 (2013) 114
few individuals of ostracods (Cyprideis torosa) and some fragmented
and reworked carapaces of foraminifera, whereas microfauna are ab-
sent. The specimen C. torosa is a worldwide species that lives in
ponds, lagoons, saltworks or marshes and that prefers quieter, more
restricted environments in comparison with Loxoconcha elliptica or
Loxoconcha rhomboidea. The punctuated ornamentation is related to
ne substrates and implies a drop in salinity in relation to unit C
(Cabral et al., 2011). A2 is characterized by the absence of any of the
groups studied, probably due to a transition to supratidal environ-
ments. Finally, the uppermost interval (A3) includes the most diverse
freshwater ostracod assemblage of the trench (Herpetocypris reptans,
Cyprinotus salinus, Cyclocypris laevis, Ilyocypris lacustris), while a
distinctive record of characeans (Chara hispida, Chara vulgaris) sug-
gests a freshwater and very shallow pond or marsh similar to the
present-day conditions (Mischke et al., 2003).
Our paleontological data match previous studies and conrm that
the 1755 tsunami deposit contains reworked foraminifera and ostra-
cods which indicate marine conditions as opposed to the brackish
to freshwater assemblages of the estuarine clays (Hindson et al.,
1999). These faunal data suggest that the 1755 tsunami caused partial
isolation of this area in relation to marine inputs, giving an explana-
tion for the decrease in salinity (from C to A) and the presence of a
calmer environment.
5. Discussion
Because tsunami deposit benchmarks are highly dependent on the
sediment source(s), it is crucial to establish reliable benchmarks of
tsunami deposits in function of their depositional and environmental
settings. Here we focused on bar-built estuarine settings such as
those located along the Southern Portuguese coast, Gulf of Cadiz,
where several tsunami deposits have been identied (Hindson and
Andrade, 1999; Hindson et al., 1999; Kortekaas and Dawson, 2007;
Font et al., 2010; Costa et al., 2011, 2012). First, we use sedimentolog-
ical, mineralogical and geochemical data of Boca do Rio estuary to dis-
cuss the tsunami deposit signature and sediment source(s). We then
discuss the use of geochemical benchmarks for tsunami deposit iden-
tication in these estuarine settings. Finally, we interpret grain-size
and compositional data in terms of hydrodynamics constraints to at-
tempt to discriminate between eventual successive waves or run-up/
backwash currents.
Fig. 6. Visible diffuse reectance spectrophotometry data represented in the CIE L*a*b*c system for the bulk sediment and the siliciclastic fraction. The siliciclastic fraction was
obtained after removing carbonates by HCl.
Fig. 7. Bivariate plots of b* versus mean grain size of the siliciclastic fraction (AS) show-
ing that the base of the tsunami is better correlated with the siliciclastic fraction of unit
C whereas the top rather correlates with unit A. Samples located in the middle of the
tsunami layers show intermediate positions between unit C and unit A.
8 E. Font et al. / Marine Geology 343 (2013) 114
5.1. The 1755 tsunami deposit at Boca do Rio: Signature and sediment
source(s)
During the last decades numerous studies dealt withtsunami deposit
identication, although few of them attempted to couple geochemical
data to sedimentological, mineralogical and paleontological approaches
(Minoura and Nakaya, 1991; Minoura et al., 1994; Goff and Chague-Goff,
1999; Brookeld et al., 2006; Schlichting and Peterson, 2006; Nichol et
al., 2007; Chague-Goff, 2010; Pozo et al., 2010; Chague-Goff et al.,
2012; Jayawardana et al., 2012; Ramirez-Herrera et al., 2012; Roy et
al., 2012; Chague-Goff et al., in press). However, such multidisciplinary
approach can provide crucial information about tsunami deposit signa-
tures and sediment origin, as is our goal here.
From a mineralogical point of view and considering the deposit as a
whole (without discriminating lithological sub-units), the 1755 tsuna-
mi deposit of the Boca do Rio estuary is featured by an enhancement
in quartz, calcite and dolomite (Fig. 3) and a depletion in phyllosilicates
(Fig. 3) and iron oxides (Font et al., 2010) relative to the enclosing estu-
arine clay units. Clay assemblages (kaolinite, smectite, illite) are identi-
cal in the tsunami deposit and in the clay units (Fig. 3) and corroborate
presence of unit C clay-balls in the tsunami deposit (Hindson and
Andrade, 1999). These mineralogical signatures are associated to char-
acteristic geochemical trends, namely a relative increase in Sr and Ca
and a systematic decrease in marine (Ba, Br) and high energy indicators
(Fe, Ti, Zr) as well as in most trace elements (Fig. 4).
The identication of three modes (900 m, 400 m and 100 m)
within the sand fraction suggests different sediment sources. These
fractions are compared to the present day beach sediment analyzed
by Hindson and Andrade (1999) suggesting a high correspondence
between the medium and coarser sands of unit B and the sediments
in the beach foreshore and dune. However, the sediments from the
coastal barrier cannot explain the mode of very ne sands identied
within the upper subunit as their presence is very rare or anecdotic
in the present littoral barrier with (very low concentration in the
dune sediments). In order to understand the origin of the very ne
sands we have compared their variability with the geochemical sig-
nature nding a high correspondence with the increase of Ti and Fe.
Macro and microfauna assemblages of the tsunami deposit include
reworked species (foraminifera and ostracods) from the estuarine
clays and admixture of shell fragments (bivalves, gastropods) from
the pit of the coastal barrier (Figs. 89). We thus interpret depletions
in phyllosilicates, heavy minerals (Ti, Fe, Zr) and common marine-
seawater indicators (Na, Cl, Ba, Br and other trace elements) as the re-
sult of estuarine clay reworking and subsequent dilution within huge
volumes of siliciclastic and carbonate material (calcite and quartz)
from the coastal barrier.
Indeed, depletion in Fe and Ti contents in the tsunami deposit
(Fig. 5) well matches mass specic (i.e. normalized by density) magnet-
ic susceptibility and isothermal remanent magnetization data (Font et
al., 2010), conrming that the concentration of iron oxides (originated
from the reworking of underlying mudats) is diluted within huge vol-
umes of non-magnetic materials (i.e. calcite and silicate). This is also
comforted by enhancement in Sr and Ca in the tsunami deposit, which
is linked to the input of shell fragments (foraminifera, ostracods,
bivalve, gastropods) and detrital calcite/dolomite (Fig. 9). Detrital
calcite/dolomite is found in the reliefs and substratumof the area (Cre-
taceous limestones and Jurassic dolostones; Fig. 1) as well as inthe form
of boulders accumulated at the pit (Fig. 9). Both shell fragments and
boulders are deposited today at the upper beach during high energy
events such as storms (Fig. 9). Therefore a huge volume of calcite and
silicate (quartz) is interpreted to originate from the erosion of the
pre-1755 coastal barrier of the Boca do Rio estuary. Because coastal bar-
riers are highly dynamic environments and present high porosities,
their sediment deposits are organic matter depleted (which can x Ba
for example). In addition, mobile elements typical of marine environ-
ments (Ba, Br, Na, Cl) are washed out within beach sediments
explaining their dilution in the tsunami deposit. In addition, the vari-
ability of the Fe and Ti in the prole presents a high correspondence
Fig. 8. Micro- and macropaleontological assemblages.
9 E. Font et al. / Marine Geology 343 (2013) 114
Fig. 9. Environmental and depositional model of the 1755 tsunami deposit of the Boca do Rio estuary.
10 E. Font et al. / Marine Geology 343 (2013) 114
withthe presence of very ne sands, suggesting that the latter is derived
from terrestrial sediments by contrast with the medium and coarse
sands that were sourced by the coastal barrier during the marine
inundation.
These results suggest that most of the material brought by the tsu-
nami wave has a proximal and coastal (estuary and adjacent beach
and dune) origin, in agreement with previous studies from the Algar-
ve region (Hindson and Andrade, 1999; Costa et al., 2012). The geo-
morphological mapping of this estuary (Fig. 1A) and the eld
observations show a great erosion of the Holocene dune barrier dur-
ing the 1755 tsunami that generated a present sandy cliff-face and as-
sociated boulder eld with terrace morphology at the mouth. Some
isolated marine boulders (20 cm) are located perched 6 m a.s.l. in
the rock cliff. This scenario is also compatible with the Sendai plain
case where diatom assemblages of the 2011 Tohoku-oki tsunami
were derived from erosion of the beach and the soil on land (Goto
et al., 2012).
5.2. Geochemical benchmarks of estuarine tsunami deposits
Recently, Chague-Goff (2010) has brought to discussion the chem-
ical imprints as a valuable tool for paleo-tsunami identication. Even
though their use is not straightforward, their combination with other
geological and mineralogical indicators can help in identifying tsuna-
mi induced deposits.
A number of elements reect a marine inuence due to their
higher concentration in seawater than in freshwater. Sulfur (S), chlo-
rine (Cl), sodium (Na), strontium (Sr), bromine (Br), barium (Ba) and
calcium (Ca) are among the elements that have been successfully
used as proxies for tsunami inundation (Andrade et al., 2003; Cundy
et al., 2006; Nichol et al., 2007; Chague-Goff, 2010; Chague-Goff et
al., 2012; Ramirez-Herrera et al., 2012; Chague-Goff et al., in press).
Heavy elements such as Fe, Ti and Zr are usually related to siliciclastic
components, especially clay minerals and iron oxides, varying directly
with the terrigenous fraction of the sediment but also strongly de-
pend on the composition of the source. They thus reect a terrestrial
origin, being in lower concentrations in seawater than freshwater
(Wedepohl, 1971). Additionally, Ti has also been used as an indicator
of high-energy deposition causing iron sand enrichment during inun-
dation (Goff et al., 2004b).
The reliability of these geochemical proxies is limited by the large
variety of depositional and compositional settings making themstrong-
ly source-dependent (e.g. Chague-Goff et al., in press). Moreover, in the
case of estuarine environments, geochemical cycles of these elements
are peculiar and complex while few data exist in the literature to
allow systematic interpretations (Malcolm and Price, 1984; Bruchert
and Pratt, 1996; Chen et al., 1997; Coffey et al., 1997; Gerritse, 1999;
Lopez-Buendia et al., 1999; Mayer et al., 2007). This is the case of the
Boca do Rio estuary, where geochemical data strongly contrast with
some study cases described in the literature (Ramirez-Herrera et al.,
2012; Chague-Goff et al., in press) and brought into discussion the reli-
ability of these elements as estuarine tsunami deposit indicators.
Concerning Sr and Ba as indicators of marine seawater invasion,
Ramirez-Herrera et al. (2012) found two sandy units interpreted as
tsunami deposits in the Ixtapa estuary (Mexico) that retain higher
concentrations of Ba and Sr in comparison to the over- and underly-
ing marshy layers. Relative maxima in Sr and Ba were also found, to-
gether with abundant marine and brackish diatoms, in the 1826 AD
tsunami deposit of the Okarito Lagoon (South Island, New Zealand)
(Nichol et al., 2007). At the opposite, Chague-Goff et al. (in press) re-
cently found higher concentrations of Sr but lower concentrations of
Ba in the 2011 Tohoku-oki and 869 AD Jogan tsunami deposits of
the Sendai plain, in Japan. In the case of the Boca do Rio estuary, our
data are similar to the tsunami deposits of the Sendai plain but
contrasting to those of the Ixtapa estuary (Ramirez-Herrera et al.,
2012). These paradoxical observations witness the complexity of
estuarine settings and the caution that must be taken when consider-
ing Ba as a paleosalinity indicator of tsunami deposits. Indeed Ba is
usually higher in estuarine mudats than in marine sediments be-
cause estuaries act as a source of dissolved barium, from fresh river-
ine suspended particulate matter or clay-borne Ba exchanging with
Mg and Ca from ocean water (Chen et al., 1997; Coffey et al., 1997;
Stecher and Kogut, 1999). Depending on the biological productivity
in the estuary, i.e. sulphate and nitrate availability, dissolved barium
can precipitate as barite, resulting in Ba depleted near-shore marine
sediments. It has been shown that the death of the diatom bloom
recorded in the spring of 1996 in the estuary of the Delaware
Bay, USA, was associated with a dramatic removal of dissolved Ba
interpreted as resulting from barite precipitation (Stecher and Kogut,
1999). However, in the case of the 1755 tsunami deposit of the Boca
do Rio estuary, neither diatoms or barite crystals were observed under
a scanning electron microscope (Font et al., 2010) while no signicant
correlation between dissolved Ba and barite is observed in the case of
the Sendai tsunami deposits suggesting that barite is not the dominant
form of Ba occurrence here (Chague-Goff et al., in press).
Br is also of particular interest because it is thought to be found in
marine water but not in terrestrial sources and has been used in estu-
arine settings to distinguish marine from terrigenous organic matter
(Fuge, 1988; Mayer et al., 2007). Schlichting and Peterson (2006) pro-
posed that a sand layer can be considered to be of marine origin if
it displayed marine markers, such as beach sand, marine diatoms,
and/or elevated bromine concentration. However, and similarly to
Ba, a relative enhancement of Br contents in tsunami deposit is
not systematically observed. On the contrary, the tsunami deposits
(sandy units) of the Pololu wetland of Hawaii showed depleted Br
content in comparison to intermediate peat layers (Chague-Goff et
al., 2012). The same is valid for the 1755 tsunami deposit of the
Boca do Rio estuary that shows a dramatic decrease in Ba and Br
contents in comparison to the clayish layers (Fig. 5). Contrarily to
the Pololu tsunami deposits, where Br mimics Na and Cl contents
(Chague-Goff et al., 2012), our data do not show signicant correla-
tions of Br and Na, and Cl.
The use of Fe, Ti and Zr as indicators of high-energy deposition is also
litigious in the case of the Boca do Rio tsunami deposit. Indeed Fe and Ti
are essentially found in all units as titanomagnetite and rutile phase
(Font et al., 2010). The linear trend of the Fe/Ti ratio along the vertical
prole indicates that no additional Fe or Ti was brought in the tsunami
deposit but rather reects a common (terrestrial) origin. Lower values
of magnetic susceptibility and FeTi content (in cps; Fig. 4) in the tsuna-
mi deposit argue for the dilution of reworked estuarine clays within a
huge volume of non-magnetic aeolian sand from the beach. Indeed,
aeolian sands are generally depleted in iron oxides whereas the latter
is accumulated, by density, in lowland area (i.e. placers, Luther et al.,
1982; Burdige, 2011). These ndings demonstrate the dependence of
tsunami deposit features with respect to the source and witness the
caution that must be taken when considering geochemical proxies as
indicators of estuarine tsunami deposits.
5.3. Hydrodynamics: Two waves or one wave with run-up and backwash?
Because our work focuses on benchmark identication through
the high-resolution analysis of a unique trench, a detailed interpreta-
tion of our whole data in terms of hydrodynamic constraints during
tsunami ooding is beyond the scope of our study. However, our nd-
ings yield new clues to improve the model previously proposed by
Hindson and Andrade (1999).
Detailed textural analysis showed abrupt changes in particle sizes
along the tsunami deposit: the base is constituted by a chaotic mix-
ture of grain size of gravel and clasts suggesting a highly turbulent
ow while the top is predominantly constituted by graded sand, silt
and clay particles indicative of a lower-energy ow (Hindson and
Andrade, 1999). These features led Hindson and Andrade (1999) to
11 E. Font et al. / Marine Geology 343 (2013) 114
suggest the occurrence of two successive waves or a backwash origin
for the uppermost layer of the tsunami deposit.
Most case studies evocate a unique couplet of sand as representing
an entire tsunami incursion (Benson et al., 1997): a sand layer at the
base deposited during the run-up and a muddy layer at the top de-
posited during backwash. Thus, the presence of successive tsunami
layers with ning-upward sequences of the b2000 m fraction is ev-
idence of the repetition of tsunami incursions. Therefore, our results
are more compatible with the hypothesis of a unique inundation
wave while the 2-wave scenario becomes less feasible. Indeed, the
relative upward increase of less dense minerals and calcium in
sub-unit B1 (Fig. 4) suggests that the base becomes less dense up-
wards, which, together with the inverse grading, is compatible with
traction carpet sedimentation at the base of the tsunami (Moore et
al., 2011). The transition between B1 and B2 is marked by a maximum
in the percentage of medium and coarse sands and a remarkable im-
provement of the sorting (Figs. 23). The latter could eventually mark
a break on the energy of the ood allowing a more selective and nor-
mal grading sedimentation. Such decrease in ow velocity between
B1 and B2 is also supported by micro- and macrofauna grain size
(see Section 4.5).
In addition we hypothesized a backwash origin for sub-unit B2,
which is majorly composed of the input of upstream terrestrial sedi-
ments as illustrated in Fig. 9. Indeed, the increase in very ne sands
(Fig. 2), together with Ti and Fe contents (Fig. 4) in the upper layer sug-
gests that, in addition to low ow energy, a change in sediment source
occurred in sub-unit B2 as suggested in Section 5.1. Such change is also
demonstrated by a strong correlation between siliciclastic fractions
from unit A and those fromsub-unit B2, whereas the base of the tsuna-
mi layer (B1) better correlates with estuarine sediments from unit C
(Fig. 7). Finally, the upper part of B2 shows an increase of the mean
grain size due to the entrance of medium and coarse sands in a very
low percentage (Fig. 2). This is interpreted as the local reworking of
B1 sediments by the backwash current.
Although the upper section of unit C is lithologically indistinguish-
able from unit A (Hindson and Andrade, 1999), slight variations in
grain size, color and paleontological data were identied here and
suggest that differences within unit A and unit C rely on lateral facies
variation such as those illustrated in Fig. 9. Indeed unit A shows ner
siliciclastic grain size associated to fresh water (uvial) ostracod as-
semblage, whereas the siliciclastic fraction of unit C shows coarser
grain size associated to brackish foraminifera and ostracods (Figs. 3
and 8). We thus conclude that the difference in unit A and unit C re-
sides in the distance to shoreline: unit C represents the estuarine
inlling facies proximal to shoreline, frequently inuenced by saline
seawater invasion which in turn leads to increase in siliciclastic ne
fraction by aggregation/occulation (i.e. processes by which clay min-
erals aggregate together and to organic matter to form coarser parti-
cles; e.g. Goldberg et al., 1991; Thill et al., 2001); whereas unit A is
located upstream, in ood plains or uvial networks, where fresh
water limits occulation processes. Such scenario implies that units
A and C co-exist during the tsunami event, and that the backwash
current was able to transport and deposit sediments from a landward
source.
The shift on the type of facies sedimentation within the estuary
from C to A, following the catastrophic event, suggests the gradual
inlling of the estuary and the advance of the upstream facies sea-
ward (Fig. 9). This trend could also be compatible with a more re-
stricted estuary, a more stable and less permeable coastal barrier
that breached less frequently.
6. Conclusions
Our multidisciplinary approach demonstrates that the material
constituting the 1755 tsunami deposit of the Boca do Rio estuary
has a proximal origin inherited from the reworking of estuarine
clays mixed with the input of sand, detrital calcite and dolomite,
shell fragments and limestone clasts from the beach.
The geochemical and mineralogical signatures of the 1755 tsunami
deposit are featured by a depletion in marine sensitive proxies (Na, Cl,
Ba, Br) and heavy minerals (Ti, Fe and Zr) resulting from the reworking
of estuarine clays and subsequent dilutionwithina huge volume of sand
from the coastal barrier. An additional source of very ne sand distinct
fromthose found at the beach and dune has been also identied. Strong
correlations between this very ne population and content in Ti and Fe
suggest an upstream continental origin.
Diffuse reectance spectrophotometry (DRS) provides informa-
tion about the color of the sediment and has been applied here, for
the rst time, in a tsunami deposit. The 1755 tsunami deposit at
Boca do Rio is characterized by large color variations interpreted to
result fromthe heterogeneity of the material constituting the tsunami
deposit. Considering that DRS analyses are rapid and non-destructive,
it represents a new high-resolution tool to identify high-energy
deposits composed by reworked materials.
The abrupt shift in hydrodynamic conditions from traction carpet
(B1) to free falling (B2) under turbulent hydrodynamic conditions in
the tsunami layer suggests the action of a single wave recording
run-up and backwash phases, respectively. This is supported by the
comparison of diffuse reectance spectrophotometry data of the
siliciclastic fraction versus mean grain size, which shows a strong cor-
relation between underlying unit C (downstream) and the base of the
tsunami deposit (B1), and overlying unit A (upstream) and the top of
the tsunami deposit (B2).
Finally, micro and macrofauna analysis and DRS data of the
siliciclastic fraction show slight but signicant environmental changes
occurring just after the tsunami.
Supplementary data to this article can be found online at http://
dx.doi.org/10.1016/j.margeo.2013.06.008.
Acknowledgments
This work was funded by the FCT (Fundao para a Cincia e
Tecnologia; ref. PTDC/CTEGIX/110205/2009), the Institute Dom
Luz (Pest-OE/CTE/LA0019/2011-IDL), the Spanish MICINN-FEDER
(CGL2010-15810/BTE) and the Research Group (RNM-238) of the
PAIDI (Andalousia Board, Spain). The authors thank Luis Rebelo and
Rachid Omira for their help in the eld; and Celia Lee and Ana Sousa
for the administrative supply. The X-ray laboratory group from URMG
at LNEG is thanked for the XRF measurements. We thank John
Triantalis for the language corrections. We thank the Editor John T.
Wells and anonymous referees for their constructive review.
References
Adkins, J.F., Boyle, E.A., Keigwin, L., Cortijo, E., 1997. Variability of the North Atlantic
thermohaline circulation during the last interglacial period. Nature 390, 154156.
Ambraseys, N., Synolakis, C., 2010. Tsunami catalogs for the Eastern Mediterranean,
revisited. Journal of Earthquake Engineering 14, 309330.
Andrade, C., Freitas, C., Miranda, J.M., Baptista, M.A., Cacho, M., Silva, P., Munh, J.,
2003. Recognizing possible tsunami sediments in the ultradissipative environment
of the Tagus estuary (Portugal). Coastal Sediments '03 The fth International Sym-
posium on Coastal Engineering and Science of Coastal Sediment Processes, 1823
May, Clearwater Beach, Fl., ed. CD-ROM, 14 pp.
Baptista, M.A., Miranda, J.M., 2009. Revision of the Portuguese catalog of tsunamis.
Natural Hazards and Earth System Sciences 9, 2542.
Baptista, M.A., Miranda, P.M.A., Miranda, J.M., Victor, L.M., 1998. Constrains on the
source of the 1755 Lisbon tsunami inferred from numerical modelling of historical
data on the source of the 1755 Lisbon tsunami. Journal of Geodynamics 25,
159174.
Barbano, M.S., Pirrotta, C., Gerardi, F., 2010. Large boulders along the South-Eastern
Ionian coast of Sicily: storm or tsunami deposits? Marine Geology 275, 140154.
Benson, B.E., Grimm, K.A., Clague, J.J., 1997. Tsunami deposits beneath tidal marshes on
Northwestern Vancouver Island, British Columbia. Quaternary Research 48, 192204.
Blott, S.J., Pye, K., 2012. Particle size scales and classication of sediment types based on
particle size distributions: review and recommended procedures. Sedimentology
59, 20712096.
12 E. Font et al. / Marine Geology 343 (2013) 114
Brookeld, M.E., Blechschmidt, I., Hannigan, R., Coniglio, M., Simonson, B., Wilson, G.,
2006. Sedimentology and geochemistry of extensive very coarse deepwater sub-
marine fan sediments in the Middle Jurassic of Oman, emplaced by giant tsunami
triggered by submarine mass ows. Sedimentary Geology 192, 7598.
Bruchert, V., Pratt, L.M., 1996. Contemporaneous early diagenetic formation of organic
and inorganic sulfur in estuarine sediments from St. Andrew Bay, Florida, USA.
Geochimica Et Cosmochimica Acta 60, 23252332.
Burdige, D.J., 2011. Estuarine and coastal sediments coupled biogeochemical cycling.
In: Wolanski, Eric, D. M. B. T.-T. (Eds.), Treatise on Estuarine and Coastal Science, 5,
pp. 279316 (on E. and C. Science).
Cabral, M.C., Freitas, M.C., Andrade, C., Cruces, A., 2006. Coastal evolution and Holocene
ostracods in Melides lagoon (SW Portugal). Marine Micropaleontology 60, 181204.
Cabral, M.C., Freitas, C., Andrade, C., Moreira, S., Cruces, A., 2011. Holocene ostracods of
Pederneira (Nazar, Portugal), a structurally-segmented inlled lagoon. Joannea
Geologie und Palontologie 11, 3638.
Chague-Goff, C., 2010. Chemical signatures of palaeotsunamis: a forgotten proxy? Ma-
rine Geology 271, 6771.
Chague-Goff, C., Schneider, J.L., Goff, J.R., Dominey-Howes, D., Strotz, L., 2011.
Expanding the proxy toolkit to help identify past events lessons from the 2004
Indian Ocean tsunami and the 2009 South Pacic tsunami. Earth-Science Reviews
107, 107122.
Chague-Goff, C., Goff, J., Nichol, S.L., Dudley, W., Zawadzki, A., Bennett, J.W., Mooney, S.D.,
Fierro, D., Heijnis, H., Dominey-Howes, D., Courtney, C., 2012. Multi-proxy evidence
for trans-Pacic tsunamis in the Hawai'ian Islands. Marine Geology 299, 7789.
Chague-Goff, C., Andrew, A., Szczucinski, W., Goff, J., Nishimura, Y., 2012. Geochemical
signatures up to the maximum inundation of the 2011 Tohoku-oki tsunami - Im-
plications for the 869 AD Jogan and other palaeotsunamis. Sedimentary Geology
282, 6577.
Chen, Z.Y., Chen, Z.L., Zhang, W.G., 1997. Quaternary stratigraphy and trace-element in-
dices of the Yangtze delta, Eastern China, with special reference to marine trans-
gressions. Quaternary Research 47, 181191.
Clark, R.N., King, T.V.V., Klejwa, M., Swayze, G.A., Vergo, N., 1990. High spectral resolu-
tion reectance spectroscopy of minerals. Journal of Geophysical Research Solid
Earth and Planets 95, 1265312680.
Coffey, M., Dehairs, F., Collette, O., Luther, G., Church, T., Jickells, T., 1997. The behaviour
of dissolved barium in estuaries. Estuarine, Coastal and Shelf Science 45, 113121.
Costa, P.J.M., Andrade, C., Freitas, M.C., Oliveira, M.A., da Silva, C.M., Omira, R., Taborda,
R., Baptista, M.A., Dawson, A.G., 2011. Boulder deposition during major tsunami
events. Earth Surface Processes and Landforms 36, 20542068.
Costa, P.J.M., Andrade, C., Freitas, M.C., Oliveira, M.A., Lopes, V., Dawson, A.G., Moreno,
J., Fatela, F., Jouanneau, J.M., 2012. A tsunami record in the sedimentary archive of
the Central Algarve coast, Portugal: characterizing sediment, reconstructing
sources and inundation paths. The Holocene 22, 899914.
Croudace, I.W., Rindby, A., Rothwell, R.G., 2006. ITRAX: Description and Evaluation of a
New Multi-Function X-Ray Core Scanner. The Geological Society of London, Special
Publications, London 5163.
Cundy, A.B., Kortekaas, S., Dewez, T., Stewart, I.S., Collins, P.E.F., Croudace, I.W., Maroukian, H.,
Papanastassiou, D., Gaki-Papanastassiou, P., Pavlopoulos, K., Dawson, A., 2000. Coastal
wetlands as recorders of earthquake subsidence in the Aegean: a case study of the
1894 Gulf of Atalantis earthquakes, Central Greece. Marine Geology 170, 326.
Cundy, A.B., Sprague, D., Hopkinson, L., Maroukian, H., Gaki-Papanastassiou, K.,
Papanastassiou, D., Frogley, M.R., 2006. Geochemical and stratigraphic indicators
of late Holocene coastal evolution in the Gythio area, Southern Peloponnese,
Greece. Marine Geology 230, 161177.
Cunha, P.P., Buylaert, J.P., Murray, A.S., Andrade, C., Freitas, M.C., Fatela, F., Munha, J.M.,
Martins, A.A., Sugisaki, S., 2010. Optical dating of clastic deposits generated by an
extreme marine coastal ood: the 1755 tsunami deposits in the Algarve (Portugal).
Quaternary Geochronology 5, 329335.
Dawson, A.G., Hindson, R., Andrade, C., Freitas, C., Parish, R., Bateman, M., 1995. Tsunami
sedimentation associated with the Lisbon earthquake of 1 November AD 1755
Boca-do-Rio, Algarve, Portugal. The Holocene 5, 209215.
Delange, W.P., Healy, T.R., 1986. New-Zealand tsunamis 18401982. NewZealand Journal
of Geology and Geophysics 29, 115134.
Dominey-Howes, D., 2007. Geological and historical records of tsunami in Australia.
Marine Geology 239, 99123.
Font, E., Nascimento, C., Omira, R., Baptista, M.A., Silva, P.F., 2010. Identication of
tsunami-induced deposits using numerical modeling and rock magnetism tech-
niques: a study case of the 1755 Lisbon tsunami in Algarve, Portugal. Physics of
the Earth and Planetary Interiors 182, 187198.
Fuge, R., 1988. Sources of Halogens in the environment, inuences on human and
animal health. Environmental Geochemistry and Health 10, 5161.
Gerritse, R.G., 1999. Sulphur, organic carbon and iron relationships in estuarine and fresh-
water sediments: effects of sedimentation rate. Applied Geochemistry 14, 4152.
Goff, J.R., Chague-Goff, C., 1999. A late Holocene record of environmental changes from
coastal wetlands: Abel Tasman National Park, New Zealand. Quaternary Interna-
tional 56, 3951.
Goff, J., McFadgen, B.G., Chague-Goff, C., 2004a. Sedimentary differences between the
2002 Easter storm and the 15th-century Okoropunga tsunami, Southeastern
North Island, New Zealand. Marine Geology 204, 235250.
Goff, J.R., Wells, A., Chagu-Goff, C., Nichol, S.L., Devoy, R.J.N., 2004b. The elusive AD
1826 tsunami, South Westland, New Zealand. New Zealand Geographer 60, 2839.
Goff, J., Chague-Goff, C., Nichol, S., Jaffe, B., Dominey-Howes, D., 2012. Progress in
palaeotsunami research. Sedimentary Geology 243, 7088.
Goldberg, S., Forster, H.S., Heick, E.L., 1991. Flocculation of illite kaolinite and illite
montmorillonite mixtures as affected by sodium adsorption ratio and pH. Clays
and Clay Minerals 39, 375380.
Gonzalez-Regalado, M.L., Ruiz, F., Baceta, J.I., Gonzalez-Regalado, E., Munoz, J.M., 2001.
Total benthic foraminifera assemblages in the Southwestern Spanish estuaries.
Geobios 34, 3951.
Goto, K., Sugawara, D., Abe, T., Haraguchi, T., Fujino, S., 2012. Liquefaction as an impor-
tant source of the A.D. 2011 Tohoku-oki tsunami deposits at Sendai Plain, Japan.
Geology 40, 887890.
Hateld, R.G., Cioppa, M.T., Trenhaile, A.S., 2010. Sediment sorting and beach erosion
along a coastal foreland magnetic measurements in Point Pelee National Park, On-
tario, Canada. Sedimentary Geology 231, 6373.
Hindson, R.A., Andrade, C., 1999. Sedimentation and hydrodynamic processes associated
with the tsunami generated by the 1755 Lisbon earthquake. Quaternary International
56, 2738.
Hindson, R., Andrade, C., Parish, R., 1999. A microfaunal and sedimentary record of envi-
ronmental change within the late Holocene sediments of Boca do Rio (Algarve,
Portugal). Geologie En Mijnbouw Netherlands Journal of Geosciences 77, 311321.
Hindson, R.A., Andrade, C., Dawson, A.G., 1996. Sedimentary Processes Associated with
the Tsunami Generated by the 1755 Lisbon Earthquake on the Algarve Coast. Por-
tugal. Phys. Chem. Earth 21, 5763.
Horton, B.P., Sawai, Y., Hawkes, A.D., Witter, R.C., 2011. Sedimentology and paleontol-
ogy of a tsunami deposit accompanying the great Chilean earthquake of February
2010. Marine Micropaleontology 79, 132138.
Jayawardana, D.T., Ishiga, H., Pitawala, H.M.T.G.A., 2012. Geochemistry of surface sedi-
ments in tsunami-affected Sri Lankan lagoons regarding environmental implica-
tions. International Journal of Environmental Science and Technology 9, 4155.
Ji, J.F., Balsam, W., Chen, J., 2001. Mineralogic and climatic interpretations of the
Luochuan loess section (China) based on diffuse reectance spectrophotometry.
Quaternary Research 56, 2330.
Ji, J.F., Chen, J., Balsam, W., Lu, H.Y., Sun, Y.B., Xu, H.F., 2004. High resolution hematite/
goethite records from Chinese loess sequences for the last glacialinterglacial
cycle: rapid climatic response of the East Asian monsoon to the tropical Pacic.
Geophysical Research Letters 31.
Kortekaas, S., Dawson, A.G., 2007. Distinguishing tsunami and storm deposits: an ex-
ample from Martinhal, SW Portugal. Sedimentary Geology 200, 208221.
Lopez-Buendia, A.M., Bastida, J., Querol, X., Whateley, M.K.G., 1999. Geochemical data
as indicators of palaeosalinity in coastal organic-rich sediments. Chemical Geology
157, 235254.
Luther, G.W., Giblin, A., Howarth, R.W., Ryans, R.A., 1982. Pyrite and oxidized iron min-
eral phases formed from pyrite oxidation in salt-marsh and estuarine sediments.
Geochimica Et Cosmochimica Acta 46, 26652669.
Malcolm, S.J., Price, N.B., 1984. The behavior of iodine and bromine in estuarine surface
sediments. Marine Chemistry 15, 263271.
Martin-Puertas, C., Valero-Garces, B.L., Mata, M.P., Moreno, A., Giralt, S., Martinez-Ruiz,
F., Jimenez-Espejo, F., 2011. Geochemical processes in a Mediterranean Lake: a
high-resolution study of the last 4,000 years in Zoar Lake, Southern Spain. Journal
of Paleolimnology 46, 405421.
Mayer, L.M., Schick, L.L., Allison, M.A., Ruttenberg, K.C., Bentley, S.J., 2007. Marine vs.
terrigenous organic matter in Louisiana coastal sediments: the uses of bromine:
organic carbon ratios. Marine Chemistry 107, 244254.
Minoura, K., Nakaya, S., 1991. Traces of tsunami preserved in inter-tidal lacustrine and marsh
deposits some examples from Northeast Japan. Journal of Geology 99, 265287.
Minoura, K., Nakaya, S., Uchida, M., 1994. Tsunami deposits in a lacustrine sequence of
the Sanriku Coast, Northeast Japan. Sedimentary Geology 89, 2531.
Mischke, S., Herzschuh, U., Kurschner, H., Fuchs, D., Zhang, J.W., Fei, M., Sun, Z.C., 2003.
Sub-recent ostracoda from Qilian Mountains (NW China) and their ecological sig-
nicance. Limnologica 33, 280292.
Moore, A., Goff, J., McAdoo, B.G., Fritz, H.M., Gusman, A., Kalligeris, N., Kalsum, K.,
Susanto, A., Suteja, D., Synolakis, C.E., 2011. Sedimentary deposits from the 17
July 2006 Western Java tsunami, Indonesia: use of grain size analyses to assess
tsunami ow depth, speed, and traction carpet characteristics. Pure and Applied
Geophysics 168, 19511961.
Morton, R.A., Gelfenbaum, G., Jaffe, B.E., 2007. Physical criteria for distinguishing sandy
tsunami and storm deposits using modem examples. Sedimentary Geology 200,
184207.
Nichol, S.L., Goff, J.R., Devoy, R.J.N., Chague-Goff, C., Hayward, B., James, I., 2007. Lagoon
subsidence and tsunami on the west coast of New Zealand. Sedimentary Geology
200, 248262.
Oliveira, M.A., Andrade, C., Freitas, M.C., Costa, P.J., 2009. Modeling volume transfer
between beachforedune and the backshore by the 1755 Lisbon tsunami at Boca
do Rio lowland, Algarve (Portugal). Journal of Coastal Research 15471551.
Papadopoulos, G.A., 2003. Tsunami hazard in the Eastern Mediterranean: Strong earth-
quakes and tsunamis in the Corinth Gulf, Central Greece. Natural Hazards 29, 437464.
Pauli, H., 1976. Proposed extension of CIE recommendation on uniform colour spaces,
colour difference equations, and metric colour terms. Journal of the Optical Society
of America 66, 866867.
Pozo, M., Ruiz, F., Carretero, M.I., Vidal, J.R., Caceres, L.M., Abad, M., Gonzalez-Regalado,
M.L., 2010. Mineralogical assemblages, geochemistry and fossil associations of
PleistoceneHolocene complex siliciclastic deposits from the Southwestern Donana
National Park (SW Spain): a palaeoenvironmental approach. Sedimentary Geology
225, 118.
Pratt, B.R., 2002. Storms versus tsunamis: dynamic interplay of sedimentary, diagenetic,
and tectonic processes in the Cambrian of Montana. Geology 30, 423426.
Ramirez-Herrera, M.T., Lagos, M., Hutchinson, I., Kostoglodov, V., Machain, M.L., Caballero,
M., Goguitchaichvili, A., Aguilar, B., Chague-Goff, C., Goff, J., Ruiz-Fernandez, A.C.,
Ortiz, M., Nava, H., Bautista, F., Lopez, G.I., Quintana, P., 2012. Extreme wave deposits
on the Pacic coast of Mexico: tsunamis or storms? A multi-proxy approach.
Geomorphology 139, 360371.
13 E. Font et al. / Marine Geology 343 (2013) 114
Rothwell, R.G., Rack, F.R., 2006. New techniques in sediment core analysis: an introduc-
tion. The Geological Society of London, Special Publications, London, pp. 129.
Roy, P.D., Jonathan, M.P., Macias, M.C., Sanchez, J.L., Lozano, R., Srinivasalu, S., 2012.
Geological characteristics of 2011 Japan tsunami sediments deposited along the
coast of Southwestern Mexico. Chemie Der Erde Geochemistry 72, 9195.
Ruiz, F., GonzalezRegalado, M.L., Munoz, J.M., 1997. Multivariate analysis applied to
total and living fauna: seasonal ecology of recent benthic Ostracoda off the North
Cadiz Gulf coast (Southwestern Spain). Marine Micropaleontology 31, 183203.
Ruiz, F., Gonzalez-Regalado, M.L., Baceta, J.I., Munoz, J.M., 2000. Comparative ecological
analysis of the ostracod faunas from low- and high-polluted Southwestern Spanish
estuaries: a multivariate approach. Marine Micropaleontology 40, 345376.
Schlichting, R.B., Peterson, C.D., 2006. Mapped overland distance of paleotsunami high-
velocity inundation in back-barrier wetlands of the Central Cascadia Margin, USA.
Journal of Geology 114, 577592.
Schultz, L.G., 1964. Quantitative interpretation of mineralogical composition from X ray
and chemical data for the Pierre Shale. Geological Survey Professional Paper 391-C
(31 pp.).
Stecher, H.A., Kogut, M.B., 1999. Rapid barium removal in the Delaware estuary.
Geochimica Et Cosmochimica Acta 63, 10031012.
Tappin, D.R., 2007. Sedimentary features of tsunami deposits their origin, recognition
and discrimination: an introduction. Sedimentary Geology 200, 151154.
Thill, A., Moustier, S., Garnier, J.M., Estournel, C., Naudin, J.J., Bottero, J.Y., 2001. Evolu-
tion of particle size and concentration in the Rhone river mixing zone: inuence
of salt occulation. Continental Shelf Research 21, 21272140.
van der Marel, H.W., 1966. Quantitative analysis of clay minerals and their admixtures.
Contributions to Mineralogy and Petrology 12, 96138.
Wassmer, P., Schneider, J.L., Fonfrege, A.V., Lavigne, F., Paris, R., Gomez, C., 2010. Use of
anisotropy of magnetic susceptibility (AMS) in the study of tsunami deposits: ap-
plication to the 2004 deposits on the eastern coast of Banda Aceh, North Sumatra,
Indonesia. Marine Geology 275, 255272.
Wedepohl, K.H., 1971. Environmental inuences on the chemical composition of shales
and clays. Physics and Chemistry of the Earth 8.
14 E. Font et al. / Marine Geology 343 (2013) 114