Lamb QSR 2009 Ip

ARTICLE IN PRESS
Quaternary Science Reviews xxx (2009) 1–17
Contents lists available at ScienceDirect
Quaternary Science Reviews

journal homepage: www.elsevier.com/locate/quascirev
Tepexpan Palaeoindian site, Basin of Mexico: multi-proxy evidence for

environmental change during the late Pleistocene–late Holocene
Angela L. Lamb a, Silvia Gonzalez b, *, David Huddart b, Sarah E. Metcalfe c, Christopher H. Vane d,
Alistair W.G. Pike e
a
NERC Isotope Geosciences Laboratory, British Geological Survey, Keyworth, Nottingham NG12 5GG, UK
b
School of Natural Sciences and Psychology, Liverpool John Moores University, Liverpool L3 3AF, UK
c
School of Geography, University of Nottingham, Nottingham NG7 2RD, UK
d
British Geological Survey, Keyworth, Nottingham NG12 5GG, UK
e
Department of Archaeology and Anthropology, University of Bristol, Bristol BS8 1UU, UK
a r t i c l e i n f o a b s t r a c t
Article history: The Tepexpan Palaeoindian skeleton was discovered in 1947 close to the former Lake Texcoco margin, in
Received 11 November 2008 the Basin of Mexico. The find has been the object of considerable interest and discussion over the last
Received in revised form 60 years regarding its real age and archaeological interpretation. Here we report new AMS radiocarbon
1 April 2009
dates associated with the sedimentary succession at Tepexpan with ages between 19,110 90 and
Accepted 5 April 2009
612 22 14C years BP and a new uranium-series date for the skeleton with an age of 4700 200 years BP
that indicates a mid Holocene age. The sedimentary succession was studied in detail using: stable
isotopes, diatoms, organic geochemistry and tephrochronology. The multi-proxy evidence suggests large
changes around the margins of Lake Texcoco in terms of the balance between aquatic and terrestrial
plants, C3 and C4 plants, saline, alkaline and freshwater conditions, volcanic activity, marginal reworking
of lake sediments and input from the drainage basin through the late Pleistocene–late Holocene. These
changes had large impacts on the prehistoric human populations living by the lake shores since the late
Pleistocene in the Basin of Mexico.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction much discussion about the real age of the human skeleton and its
archaeological interpretation, as it marks the beginning of the
The aim of this paper is to understand environmental change in study of early Americans in Mexico and the use of science based
the Basin of Mexico over the last 20,000 years at the site of the archaeology methods (geophysics, radiocarbon dating and forensic
discovery of the Tepexpan Man (Hombre de Tepexpan) Palae- reconstruction) in the country. Unless stated otherwise, all dates
oindian remains. This near complete human skeleton was found in referred to in the text are uncalibrated radiocarbon years.
the eastern margin of the former Lake Texcoco, in marginal lake The separation of climatic change from volcanic and human-
sands, silts and clays, with diatom and volcanic ash layers (De Terra, induced changes has been a problem in Mexican palae-
1947, 1949, 1951). Tepexpan is an adult male with an age-at death of oclimatology(Sears, 1951; Sears and Clisby, 1955; Metcalfe et al.,
25–30 years. The cranium is meso-to brachicephalic, with a cranial 1991; Caballero-Miranda, 1995 Lozano Garcia and Ortega Guerrero,
index of 79.4 (Gonzalez et al., 2003). The skeleton was found under 1997; Huddart and Gonzalez, 2006). There has also been much
a layer of caliche in sediments associated with several mammoth discussion regarding the climatic evidence based on various types
finds nearby (De Terra et al., 1949). At the time of discovery, the of proxy (Brown, 1985; Markgraf, 1993; Buckler et al., 1998; Lou-
human skeleton (assumed to be at least 10,000 years old because it nejeva Baturina et al., 2006; Solleiro Rebolledo et al., 2006) and
was thought to be in coexistence with the mammoth remains) was from various altitudes and locations within Mexico (Metcalfe, 1997,
regarded as an important Palaeoindian skeleton. Here, a sedimen- 2006; Metcalfe et al., 2000). There has also been a debate related to
tary succession has been studied 80 m to the east of the original site Mexico in terms of the timing, route and origin of the first coloni-
where the skeleton was found. Since its discovery there has been sation of the Americas (Gonzalez et al., 2006a,b) and at least 27
Palaeoindian remains have been identified in Central Mexico, from
excavations and chance finds during the last 60 years. This collec-
* Corresponding author. Tel.: þ44 (0)151 231 2213; fax: þ44 (0)151 207 3224. tion known as the Preceramic Human collection is kept at the
E-mail address: [email protected] (S. Gonzalez). National Museum of Anthropology in Mexico City and has been
0277-3791/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.quascirev.2009.04.001
Please cite this article in press as: Lamb, A.L., et al., Tepexpan Palaeoindian site, Basin of Mexico: multi-proxy evidence for environmental...,
Quaternary Science Reviews (2009), doi:10.1016/j.quascirev.2009.04.001
ARTICLE IN PRESS
2 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17
radiocarbon dated to late Pleistocene to mid Holocene (Gonzalez deltaic and lacustrine sedimentary environments (Fig. 1). Hence at
et al., 2001, 2003). Recently, suggested human footprints from the Tepexpan, lake environments would have been brackish and
Valsequillo Basin (east of the Basin of Mexico, south of Puebla) have marshy, fed by springs emerging from the surrounding volcanic
been dated to at least 40,000 years old (Gonzalez et al., 2006b), lavas, rather than like the saline central lake. Bradbury (1989) found
although this has resulted in much controversy (Renne et al., 2005; that the system is currently sodium- and bicarbonate-rich, with the
Gonzalez et al., 2006a). Nevertheless, there are increasing sugges- central lake being alkaline (pH >9), compared to the higher
tions that Mexico played a central role in the origin and dispersal of elevation freshwater basins (pH 7.6–8.4).
early Americans since the late Pleistocene, with ages of at least Basin climate is currently subtropical, with monsoonal rainfall
10,755 75 years BP (OxA-10,112) for the earliest directly radio- from the western part of the Bermuda high falling predominantly in
carbon dated humans found in the Basin of Mexico (Gonzalez et al., the warm, summer months between July and October (500–
2003). Consequently, understanding the role of past climate, 1000 mm). However, the mountains provide a wide range of
volcanism and environmental change in the development of early precipitation and temperature gradients and as a result the basin
human groups and their associations with migration and settle- seems highly sensitive to climatic change (see Metcalfe et al., 2000
ment patterns, is particularly significant in this area. for a review). The Tepexpan site is drier (<600 mm) than the
southern basin (Sanders, 1976). Solleiro Rebolledo et al., (2006)
2. Regional setting suggest that most of the Teotihuacán valley just to the north of
Tepexpan, between 2250 and 2800 m a.s.l., is semi-arid with
2.1. Characteristics of the Basin of Mexico and Lake Texcoco a mean annual temperature of 14.9 C and annual precipitation of
563.3 mm. It is not clear, however, when either the current
The Basin of Mexico is a high altitude (ca 2240 m a.s.l.), flat- monsoonal rainfall patterns, or the differences in rainfall across the
floored basin, surrounded by volcanic mountains (Fig. 1) and has basin evolved (Lozano Garcia and Ortega Guerrero, 1998). The
been a closed hydrographic system since ca 700 ka BP (Vazquez modern vegetation at Tepexpan is highly disturbed, but along the
Sanchez and Jaimes Palomera, 1989). There are a series of inter- altitudinal gradient, the natural vegetation comprises: Juniper
connected sub-basins which formerly contained lakes. Prior to forests, xerophytic scrubland, pine-oak forests, Abies forests,
artificial drainage, these lakes formed a united lake approximately grassland (Hillaria cenchroides) and subalpine grassland (Rze-
1000 km2 in size (Sanders, 1976), with Lake Texcoco the lowest and dowski and Rzedowski, 1979, 1985). The pre-drained environment
most saline of these basins acting as the terminal lake. According to would have been one that supported emergent and submergent
Bradbury (1989), a natural sill at the north end of the basin would vegetation. Climatic changes over the last 50,000 years BP have
have limited the lake’s maximum depth to 24 m in the recent seen fluctuations in the balance between grasslands around the
geological past. This shallowness, combined with thermal spring lakes and the mixed pine and oak forests of the upper basin (Lozano
inflow and carbonate-rich groundwater, explain Texcoco’s high Garcia and Vazquez Selem, 2005).
salinity. After draining in the 1900s (Bradbury, 1971), the lake now Human influence in the basin may go back at least as far as
occupies only a small area to the NE of Mexico City surrounded by 22,000 BP from the evidence suggested at the Tlapacoya Paleo-
salt marshes. Tepexpan lies in a marginal lake position on the indian site to the SE of the basin (Fig. 1; Lorenzo and Mirambell,
northern side of Texcoco town, at an altitude of 2241 m. The Rı́o San 1986), although most of this very early archaeological evidence is
Juan flows into Lake Texcoco just to the south of the site forming also disputed (Waters, 1985; Huddart and Gonzalez, 2006). There is
Fig. 1. Map of the Basin of Mexico showing the former extent of Lake Texcoco and other major water bodies. The location of the Tepexpan Man skeleton and the studied sedi-
mentary succession are marked with a black solid circle.
ARTICLE IN PRESS
A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 3
however, firm evidence for human presence in Tlapacoya back to due to low microfossil preservation and intense human activity
10,200 65 years BP (OxA-10,225) by direct radiocarbon dating of from w3500 BP onwards (Metcalfe et al., 1991). Adding to these
a human skull (Gonzalez et al., 2003). problems, many sedimentary successions suffer from poor chro-
nological control, largely due to a lack of available organic material
3. Previous palaeoenvironmental work in the Basin of Mexico to be dated.
Despite these problems, some patterns of basin palaeoclimate
Lacustrine studies in the basin have focussed predominantly on are emerging. Most lacustrine records point to the Last Glacial
Lake Chalco to the south of Texcoco (Sears, 1952; Bradbury, 1971, Maximum (LGM) (ca 23–15 ka BP) being cool and relatively dry;
1989; Watts and Bradbury, 1982; Lozano Garcia et al., 1993; Lozano involving the expansion of grasslands around the lake basins
Garcia and Ortega Guerrero, 1994; Caballero Miranda, M.E. 1997; (Lozano Garcia et al., 1993; Lozano Garcia and Ortega Guerrero,
Caballero Miranda and Ortega Guerrero, 1998; Ortega Guerrero et 1994; Ortega Guerrero et al., 2000). However, in the highlands,
al., 2000) and Tecocomulco in the north of the basin (Caballero open forest communities, demonstrated by high arboreal pollen
Miranda, et al., 1999). There has also been a pollen study of a core levels (e.g. Lozano Garcia and Ortega Guerrero, 1998), suggest more
from the centre of Lake Texcoco (Lozano Garcia and Ortega variable conditions than first envisaged (Bradbury, 1997; Metcalfe
Guerrero, 1998) and diatom and pollen work on sections from et al., 2000; Sedov et al., 2001) and a more recent study of palae-
across the basin, including several sites around Texcoco (see osols from the Teotihuacán Valley suggest more humid conditions
Bradbury, 1971, 1989). Here diatoms were used to reconstruct past at the LGM (Solleiro Rebolledo et al., 2006). The Lateglacial period is
environments over the last 30,000 years BP, although diatom less well understood, partly due to frequent volcanic events, which
preservation and dating control were poor (Table 1). Texcoco varied have widely obscured the lacustrine, sedimentary record (Ortega
between a saline lake, saline marsh, alkaline lake and alkaline Guerrero et al., 2000). There is increasing evidence for dry, possibly
marsh, depending on effective moisture in the basin. During drier very dry, conditions in the basin at this time, but local conditions
periods, the level of Texcoco would drop, forming algal-rich, are variable with arguments both for a relatively moist period
concentrated pools and creating large areas of river and spring-fed (15–10 ka BP) (Lozano Garcia et al., 1993; Caballero Miranda and
freshwater marshes around the margins of the lake. This is because Ortega Guerrero, 1998) and increased aridity (Bradbury, 1989;
there is very little change in altitude between the base of Texcoco Caballero Miranda et al., 1999), with Lake Texcoco becoming saline
and the upper basins. During humid periods, and possibly on and the marsh areas expanding. In the south of the basin increased
a seasonal basis, Texcoco would rise and flood the surrounding freshwater input from the catchment may have been caused by
freshwater marshes with saline water (Bradbury, 1971). reduced catchment vegetation due to the intense volcanism
The lakes in the Basin of Mexico are some of the most intensely (Caballero Miranda and Ortega Guerrero, 1998). The combined
studied in the Neotropics yet an array of problems have hindered results of drying and volcanism have been used to explain why
a consensus on general patterns of environmental change (Lozano many records suggest a hiatus in sedimentation around Texcoco
Garcia et al., 1993; 1993; Metcalfe, 1997; Metcalfe et al., 2000; from ca 14,500–6000 BP (Bradbury, 1989; Lozano Garcia and Ortega
Sedov et al., 2001; Huddart and Gonzalez, 2006; see Table 1). Guerrero, 1998; Sedov et al., 2001).
Finding continuous sedimentary successions, particularly back to Most studies suggest an arid period from 9000 to 6000 BP in the
20,000 years is difficult because of numerous late Pleistocene basin (e.g., Lozano Garcia et al., 1993; Caballero Miranda et al.,
volcanic events that have led to sediment discontinuities and 1999). The main lake was saline and may have occasionally dried
multiple ash layers that render the stratigraphy difficult to analyse out (Bradbury, 1989; Lozano Garcia and Ortega Guerrero, 1998). The
and interpret. The Holocene period, which is better represented in pollen record from Texcoco however, suggests that the aridity was
many sedimentary successions, has also been difficult to interpret not as intense as the Lateglacial. In the nearby Lerma Basin initially
Table 1
Existinginterpretations of the climate of Lake Texcoco and the Basin of Mexico during the last 20,000 radiocarbon years BP compared to new findings from this study.
Chronology (ka) Lake Texcoco Lake Texcoco Lake Texcoco General basin
(this study) 1, 2 3, 4* climate 5, 6, 7, 8, 9, 10,
11, 12, 13
0–5 Low P/E, spring and river Small increase in Warmer and wetter Some brief humid periods
input dominate. Lake drying out precipitation but disturbed
5–10 Alkaline, shallow lake; arid conditions Dry, reduced *Dry warm conditions Dry conditions
infiltration, warm 6.1 ¼ Eruption hiatus Low lake levels
10–16 Volcanism. Initially periods of low P/E Lake very shallow, Dry & cold (23–10 ka) Humid but variable (12–9 ka)
then more humid. P/E declines towards saline, marginal lakes
the Holocene with some freshwater nourished by freshwater
activity evident springs
14 Hiatus (14.4–6.1 ka) – Hiatus – lake drying Volcanic activity after 15 ka
lake dry at margins causes fluctuating lake acidity.
and eroding Some humid events
14–18 Increasingly dry but with periods of *Cool, moist conditions Volcanic disturbances
increased river input to the lake Climate unclear – Chalco
(19–16.5 ka) warmer/drier
18–23 Shallow & saline lake Dry & cold (23–10 ka) – Cool & dry
Cooler/drier – some some increase in winter New records ¼ variability
increase in winter precipitation precipitation
23–25 Lakes fell, marginal sites Large eruption of Drier (rise in temp after 23.5 ka)
became fresh and shallow Popocatépetl Volcano
(23.6–22 ka)
Numbers in first row refer to the following references: 1. Bradbury, 1971; 2. Bradbury, 1989; 3. Lozano Garcia and Ortega Guerrero, 1998; 4. Gonzalez Quintero and Fuentes
Mata, 1980; 5. González Quintero, 1986; 6. Lozano Garcia et al., 1993; 7. Lozano Garcia and Ortega Guerrero, 1994; 8. Lozano Garcia and Ortega Guerrero, 1998; 9. Caballero
Miranda, 1997; 10. Caballero Miranda et al., 1999 (northern basin); 11. Ortega Guerrero et al., 2000; 12. Metcalfe et al., 2000; 13. Sedov et al., 2001.
ARTICLE IN PRESS
the Holocene was cool, indicated by glacial advances (10,000–

9000 BP), followed by climatic amelioration, which was strongest
from 8000 to 3700 BP and correlates to the expansion of Abies
pollen in the Basin of Mexico. Brief wet periods are evident at 6800
and 5100 BP, followed by widespread deforestation from 3500 to
1200 BP. Many inferences about the nature and timing of arid
intervals since the late Pleistocene are at variance from the rest of
Central Mexico (Metcalfe et al., 2000; Table 1). This variance may be
due to the complications referred to, but also the fact that Mexico
lies between Tropical and Temperate climate regimes.
d18O applications have been shown to provide good evidence for
past moisture availability in Mexico (Davies, 1995; Bridgwater et al.,
1999), but the number of studies has been restricted by a lack of
carbonates in volcanic terrain lakes. Here, we present a range of
geochemical analyses of a late Pleistocene sedimentary succession
excavated at Tepexpan that has the potential to provide detailed
records of palaeoclimate. This will allow further understanding of
the Central Mexican palaeoenvironment and its influence on
human populations and migration patterns. This information,
combined with an improved geochronological framework for key
early human sites will aid an understanding of Mexican Palae-
oindian geoarchaeology during the late Pleistocene–early
Holocene.
4. Methods, results and interpretations
4.1. Lithology
The new sedimentary succession reported here is situated close

to the eastern margin of the former Texcoco lake (Fig. 1) and is
located at 19 360 5200 N, 98 560 4700 W at an altitude of 2255 m a.s.l.,
Fig. 2. Lithology and physical properties (by loss-on-ignition) of the Tepexpan sedi-
approximately 80 m from the discovery of Tepexpan Man. Organic mentary succession. The star symbol represents the position of Tepexpan Man in the
and carbonate contents of the sediment were estimated from sedimentary succession.
weight loss-on-ignition at 550 and 950 C respectively, along with
% water content. XRD analysis on 18 samples spaced through the
section was carried out using a Philips model PW 1730 X-ray with these previous studies. The lowest part of our Unit 4 they
generator, with a PW 1716 diffractometer and PW 1050/25 detector consider to be a syn-sedimentary formed Fluvisol which has no
in order to check carbonate species present. clear pedogenic horizons but has silica infillings of root channels.
We consider this to be largely a tephra fall into the lake and not of
4.2. Lithology results pedogenic origin. Our Units 5 and 6 they consider to be dry-land
palaeosols separated by pyroclastic sediment (Lounejeva Baturina
The studied sedimentary succession comprises 4 m of fine et al., 2006), partly redeposited by fluvial processes. The lowest of
sands, silts and clays with volcanic ash, calcrete and ostracod layers the three upper palaeosols is classified as a Fluvial Cambisol, the
(Fig. 2). The lithology can be split into six distinct units. Unit 1 (400– middle has a large, hard, carbonate-rich ACk horizon and dark
360 cm) is variable consisting of 12–20 cm interbeds of brown silty brown–black root infillings, whilst the upper palaeosol is thinner
clay and approximately 3–8 cm fine sand units, with included with dark-brown Ah horizons and abundant powdery, neoformed
ostracods, diatoms and ash. Unit 2 (360–310 cm) is more uniform, calcite. The upper two palaeosols are classified as Calcaric Fluvisols.
consisting predominantly of ash-rich, brown, laminated fine sand, We also noted a set of deep, narrow fissures beginning in their
with a band of silt to fine sand at 321–326 cm. There is a colour middle paleosol into the lacustrine sedimentary succession below.
change to a pale brown fine sand in Unit 3 (310–235 cm), with
occasional calcareous and diatom-rich laminations and ripples. Unit 4.3. Stable isotopes
4 (235–138 cm) is volcanic ash dominated and calcareous. The
dominant grain size is silty-sand and the upper part of the unit 4.3.1. Modern samples
(138–180 cm) is fine sand, with fine ash laminations. In Unit 5 (138– Water samples were collected for 18O/16O analysis from the
124 cm) there is a band of dark brown silt with carbonate nodules. remnant lake and surrounding basin springs to characterise the
Unit 6 has an indurated, calcareous, brown silt (124–30 cm) and modern hydrology and were analysed using the equilibration
a caliche (calcrete) layer at 62–62.5 cm. Top soil forms the top method for oxygen (Epstein and Mayeda, 1953). Analytical preci-
30 cm of the section. The % water content of the sediments varies sion is typically 0.05& for d18O (1s). Isotope values are reported
widely from 5 to 60% (Fig. 2). % carbonate and % organic matter as per mil (&) deviations of the isotopic ratio (18O/16O) from
follow similar trends; they are lowest in the ash-rich, Unit 4 and standard mean oceanic water (V-SMOW).
high in Units 1–3 (Fig. 2). XRD analyses indicated the samples
contained calcite, except one sample of aragonite at 256 cm. 4.3.2. Modern isotope hydrology results and interpretation
The sedimentary succession described here was also studied by The present, highly evaporated, Lake Texcoco has a d18O value of
Lounejeva Baturina et al. (2006) and Solleiro Rebolledo et al. (2006) 3.3& (in 2003). Spring water has a mean d18O value of 11.7
and it is pertinent to discuss the differences in the current work 0.2& (n ¼ 3) and is likely to represent the d18O composition of
ARTICLE IN PRESS
precipitation in the basin. As described above, the hydrology of Lake (cellulose, Sigma Chemical prod. no. C-6413) calibrated against
Texcoco is unusual as during drier periods freshwater may become NBS-19 and NBS-22. Replicate analysis gave a precision of <0.1&
more influential giving rise to low marginal lake d18O values, (1s).
approaching 12& or lower. During humid periods the lake level
rises introducing saline, evaporated and thus isotopically heavy 4.3.5. Carbonate d18O and d13C results and interpretation
water to the surrounding freshwater marshes (Bradbury, 1971). The carbonate d18O and d13C show good correspondence sug-
gesting that they are responding to changes in climatically-induced,
4.3.3. Carbonate d18O and d13C methods moisture variations (variations in P–E) (see Fig. 3). Units 1–3 show
Sediment samples were treated with 5% sodium hypochlorite similar trends where the values start relatively low and increase
solution for 24 h to oxidise reactive organic material, washed three through the unit. d18Ocalcite in Units 1 and 2 increases from 3 to 4
times in distilled water and sieved at 85 m. The <85 m fraction was to þ7& and similarly in Unit 3, increases from 2 to over þ7&,
filtered through quartz microfibre filter paper, dried at 40 C and before progressively decreasing at the top of the unit to 2&.
ground in agate. The isolated material was reacted with anhydrous d13Ccalcite shows similar patterns in Units 1–2, increasing from
phosphoric acid in vacuo overnight at a constant 25 C. The CO2 around 0 to þ6& and similarly in Unit 3 it increases from 2 to
thus liberated was separated from water vapour and collected for þ6& but unlike d18Ocalcite, does not decrease at the top of the unit.
analysis on a VG Optima mass spectrometer. Overall analytical In the ash-rich Unit 4, values for carbon and oxygen are variable,
reproducibility for this type of sample is normally around 0.2& for beginning low (<10 and <6& respectively) and then stabilising
13
C and 18O (2s). Isotope values (13C and 18O) are reported as per mil for d13Ccalcite at around 1 to þ1&; but remaining variable for
(&) deviations of the isotopic ratios (13C/12C and 18O/16O) from the d18Ocalcite at 7 to þ2&. d18O and d13C values are relatively stable in
standard V-PDB. Units 5 and 6, around 5 to 3& and 0 to þ2& respectively. The
top three values have lower d18O and d13C values coinciding with
4.3.4. d13Corganics and C/N methods lower C/N values and higher % water content (Fig. 3).
Percentage carbon and nitrogen, used to calculate C/N, were
measured on sediments treated with 5% HCl to remove carbonates 4.3.6. d13Corganics and C/N results and interpretation
using a Carlo Erba elemental analyser, calibrated through an d13Corganics and C/N ratios are usually measured together as
internal acetanilide standard. Replicate analysis of samples gave d13Corganics can be difficult to interpret without knowing the source
a precision of þ<0.1 (1s). 13C/12C analyses were performed by of the organic material reaching the lake (Figs. 2 and 4). C/N ratios
combustion using a Carlo Erba 1500 on-line to a VG Triple Trap and can distinguish between aquatic vegetation (C/N ratios of 4–10) and
Optima dual-inlet mass spectrometer. 13Corganic values were calcu- terrestrial, vascular plants (ratios >20) (Talbot and Lærdal, 2000).
lated to the V-PDB scale using a within-run laboratory standard Normally, lacustrine sedimentary C/N ratios of between 10 and 20
Fig. 3. Stratigraphy, dating and isotope data for the Tepexpan section. The star symbol represents the position of Tepexpan Man in the sedimentary succession.
ARTICLE IN PRESS
Fig. 4. d13Corganic and C/N ratios for sediments in the Tepexpan section arranged by unit. Typical ranges of various plant types are shown (see text for references).
represent a mixture of aquatic and higher plant material (Meyers, plants dominate. Low % TOC values in Unit 4 mean that C/N inter-
1994). Plankton has average C/N ratios of between 5 and 8 (Tyson, pretations are unreliable and are thus not shown or discussed. C/N
1995) but freshwater macrophytes are more varied and record ratios begin high at the section base (ca 20) and fall progressively
ratios of between 12 and 30. Phytoplankton and macrophytes will through Units 1 and 2 (to <5) before returning to high values (15–
utilise dissolved CO2 in the lake in the same way as C3 plants and 20) in Unit 3. Values are again low in Unit 5 (ca 5) and then increase
thus will have a similar range of d13C values (17 to 9&; Degens to around 15 in the upper unit.
et al., 1968). If dissolved CO2 becomes depleted in a lake due to d13Corganics is relatively stable in Units 1–3 ranging from 18 to
elevated pH, or increased productivity, HCO 3 will be utilised 21&. There is a peak of heavier values around 18& at 331–
instead, which has a higher d13C ratio, thus increasing aquatic plant 345 cm, coinciding with C/N values of <5. In Units 5 and 6, d13Cor-
d13C values to C4 ranges (Laws et al., 1995). If C/N ratios indicate ganics values are considerably higher, ranging from 16.4 to 14.7&
terrestrial sources of plant material, C4 grasses (17 to 9&; and are approaching C4 values (Figs. 3 and 4).
Deines, 1980) can be distinguished from C3 plants (33 to 23&)
O’Leary et al. (1992). Common C4 plants, with high d13C values, 4.4. Organic geochemistry
include Poaceae, Chenopodiaceae (Cerling et al., 1988) and Cyper-
aceae (sedge family; Tyson, 1995), whereas most other plants 4.4.1. Methods
(temperate grasses, shrubs, trees and some forbs) have lower d13C For lipid extraction, four samples characteristic of different
values and are C3 category. There is an intermediary group of plants lithologies from the Tepexpan section were freeze-dried and
(CAM), including most succulents, that have intermediary d13C powdered to pass a 63 mm sieve. Accurately weighed sediment
ratios of 11 to 28& (Schleser, 1995). portions were spiked with a known amount of hexatriacontane (n–
%C closely follows trends in % water content, with peaks in both C36 alkane) dissolved in dichloromethane (DCM). The sediments
occurring concurrently in Units 1 and 2 at 389, 365 and 323 cm. In were mixed with an equal amount of anhydrous sodium sulphate
Unit 3, %C begins at around 4% but drops off rapidly through the and extracted with DCM:methanol (9:1 v/v) in an accelerated
unit to <1% that is maintained through Unit 4, before it increases solvent extraction system ASE 200 (Dionex). The solvent was
gradually to a few % through the upper 2 units (Fig. 3). Neither %C, reduced under a stream of dry nitrogen gas to a volume of 5 ml. The
%N, or C/N decrease systematically down the section, suggesting total lipid extracts were split into acid and neutral fractions by solid
that diagenetic processes have not severely affected the sediments phase extraction using normal phase C18 cartridges (Aminopropyl
(Fig. 3). %C and %N positively correlate (r2 ¼ 0.91, n ¼ 151) sug- Bond Elute) pre-washed with DCM:isopropanol, 2:1 v/v. The
gesting that they are both organically bound (cf. Talbot and neutral fraction was recovered with 10 ml DCM:isopropanol, 2:1 v/
Johannessen, 1992). C/N ratios vary between units (Figs. 3 and 4); in v and separated into hydrocarbon, alcohol and polar portions using
Units 1 and 3, values are predominantly greater than 15, suggesting thin layer chromatography (TLC) (silica 60 G, 0.25 mm thick). The
a dominance of terrestrial or vascular plants. In Unit 6 C/N is plates were developed using 1% acetic acid in hexane:ethylacetate
intermediate, around 10–15 and suggests a mixed environment. In 7:2 v/v) and the aliphatic and alcohol fractions collected. Combined
Units 2 and 5, C/N values range from 3 to 13 indicating that aquatic gas chromatography-mass spectrometry (GC-MS) was performed
ARTICLE IN PRESS
using a Fisons 8000 GC directly coupled to a Fisons 800 MD single- Table 2

quadrupole mass spectrometer. Sample application (1 ml) was by Distribution parameters from n-alkanes from selected samples at Tepexpan.
on-column injection. The GC oven was temperature-programmed Sample Cnoa range Cnob max OEP TAR Paq
from 40 C (1 min isothermal) to 310 C at 4 C/min and held Tx-8 13–33 33 1.7 2.8 0.4
isothermally at 310 C for 5 min. Peak areas of the products and Tx-49 14–33 20 0.7 0.4 0.7
internal standard were measured using the total ion current (TIC) Tx-103 13–33 31 4.4 6.6 0.4
Tx-146 13–33 31 4.2 6.9 0.4
on the GC-MS. For Elemental Analysis and Rock–Eval Pyrolysis
concentrations of total organic carbon (TOC) were determined on Terrigenous/aquatic ratio (TAR) ¼ (nC27 þ nC29 þ nC31)/(nC15 þ nC17 þ nC19). Odd
over even preference indices OEP ¼ (nC25 þ 6nC27 þ nC29)/(4nC26 þ 4nC28). n-
the carbonate free samples using an Elementar C, N, S analyser
alkane proxy of submerged/floating freshwater macrophytes (Paq) ¼ (C23 þ C25)/
operated in C mode. Kerogen typing was conducted using a Rock– (C23 þ C25 C29 C31).
Eval II instrument (precision of measurement for HI, OI; 5%. a
Carbon number range.
b
Hydrogen indices (HI ¼ S2 100/TOC) and oxygen indices Chain length of the most abundant homologue.
(OI ¼ S3 100/TOC) were measured using TOC values.
4.4.2. Results and Interpretation submerged/floating species. Evidence that this section of the lake
The n-alkane distributions of the sediment samples from represents a changed palaeoenvironment from that above (118 cm,
selected palaeoenvironments (Units 1, 3, 4 and 6) are shown in Unit 6), or below (389 cm, Unit 1 and 304 cm, Unit 3) is supported
Fig. 5. Similar n-alkane distributions in the range of nC13–nC33 and by the high relative abundance of low molecular weight homo-
maximal n-alkane at nC31 were observed in samples from the logues, even n-alkane preponderance, as well as TAR value of 0.4
lowest part of the section (389 cm, Unit 1 and 304 cm, Unit 3) (Table which suggest increased sources of aquatic sources of n-alkanes
2). The range and relative abundance of n-alkanes are consistent (Table 2). The moderate OEP value of 0.7 could be caused by
with a dominant input from land plant, epicuticular waxes (Meyers, thermal maturation due to volcanism since carbon number pref-
1997, 2003). Freshwater microalgae, such as Botryococcus braunii, erence is known to diminish with thermal maturity/marine input.
also yield odd numbered long chain n-alkanes in the range nC25–35; The uppermost section (118 cm, Unit 6) probably represents
however, sediments rich in B. braunii also contain significant a mixture of different terrestrial and aquatic sources as evidenced
concentrations of macrocyclic-alkanes in the range C15–C24 (Audino by the high abundance of odd carbon numbered high molecular
et al., 2001). The absence of macrocyclic-alkanes here confirmed weight homologues (nC31–nC33). However, a significant contribu-
that the odd carbon numbered high molecular weight n-alkanes tion from lower molecular weight n-alkanes, as well as Paq of 0.4
are mainly sourced from terrestrial plant waxes. This notion is (emergent macrophytes), and moderate TAR of 0.4, possibly indi-
supported in part by the high terrigenous/aquatic ratio (TAR) values cates mainly terrestrial watershed sources (Meyers, 1997). The
of 6.6 and 6.9 which indicates elevated terrestrial sources of lipids input of C3 and C4 plants into agricultural soils can be delineated on
relative to aquatic sources, such as algae (Meyers, 1997) (Table 2). the basis that the former are slightly depleted in nC33 as compared
The odd to even carbon preference values of w4 for both samples to their C4 vegetation (Wiesenberg et al., 2004). The high propor-
show a strong odd n-alkane preponderance which is consistent tion of nC33 relative to nC29 and nC31 in sample 118 cm may
with terrigenous input from either grasses, or deciduous trees, or therefore indicate an increase in C4 plant input (Fig. 6).
both, and low thermal maturity (Scalan and Smith, 1970) (Table 2).
Ficken et al. (2000) have shown in four surface lake sediments
and 23 plant species from Kenya that submerged and floating
leaved aquatic plants yield enhanced abundances of C23 and C25 n-
alkanes as compared to emergent and fully terrestrial plants which
are dominated by C29, C31 and C33 homologues. This was formulated
into the proxy, Paq for submerged/floating aquatic plants versus
emergent terrestrial plant sources. At Tepexpan samples at 304 and
389 cm give intermediate values of 0.4 corresponding to emergent
macrophytes, or possibly a mixture of terrestrial, emergent and
submerged vegetation (Table 2). The sample at 199 cm (Unit 4) with
Paq value of 0.7 corresponds with n-alkanes derived mainly from
Fig. 6. Ternary diagram showing the relative composition of the three dominant
Fig. 5. Histograms of the distribution of odd and even numbered n-alkanes from n-alkanes (C29, C31 and C33) from selected Tepexpan sediments. The sample at 118 cm
selected Tepexpan sediments. (Unit 6) plots in the region indicative of C4 plants (Wiesenberg et al., 2004).
ARTICLE IN PRESS
The Rock–Eval pyrolysis data presented in Fig. 7 can be used to 4.5. Diatom ecology
characterise bulk organic matter; the hydrogen and oxygen indices
provide proxy information for algal or land plant derived organic 4.5.1. Methods, results and interpretation
matter (Holtvoeth et al., 2001; Meyers, 2003). Samples 118, 304 and Here we present preliminary results from 17 samples through
389 cm all have HI values above 50 mg (HC/g TOC) but lower than the section between 401 and 40 cm in depth. Samples were
200 mg (HC/g TOC). With variable OI values, this suggests that the prepared using a standard method to remove carbonate and
hydrocarbons are probably derived from a Type III kerogen, which organic matter (Battarbee, 1986). Diatoms were identified under
is predominantly composed of organic matter sourced from land 1000 magnification using standard floras and interpreted using
plants. In contrast, Type II kerogens originate from bacterial, both published information and personal data on the ecology of
phytoplankton and or zooplankton sources, or from Type II/III modern Mexican diatoms (e.g. Davies et al., 2002). Diatom abun-
mixed marine/terrigenous inputs, which yield HI values of 300–600 dance and preservation through the section is variable from few to
and 200–300 mg (HC/g TOC) respectively. Type IV kerogens are relatively abundant. The flora indicates shallow water, but of
generally classified as having HI values of <50 mg. However, varying degrees of salinity/alkalinity, with some signs of changes
previous studies of HI indices from late Quaternary sediments from between NaCO3 and NaCl chemistry (Fig. 8). In Unit 1, the diatoms
Lakes Victoria and Rukwa have suggested that HI values of <200 indicate variable conditions. Shallow water is indicated with small
possibly indicate either a mixed source of organic matter, or Nitzschia spp. or Fragilaria spp. dominating. The species suggest
a significant coal or charcoal component as well as possible mostly freshwater conditions with occasional increases in alka-
reworking and oxidation (Talbot and Livingstone, 1989). linity and water chemistry varying between HCO 3 , CO3 and Cl.
The HI/OI indices of sample 199 cm (Unit 4) cannot be readily Diatoms are abundant and well preserved and indicate vegetation
explained in terms of kerogen type, or sources of organic matter nearby in some places (388–390 cm), but are sparse and poorly
since HI values of >1000 have not been previously described even preserved, with signs of mechanical damage and corrosion in
in algal Type I source rocks, which are derived in the main from others (378–380 cm). At 378–380 cm very alkaline conditions are
reworked algal debris deposited in lacustrine or marine settings. suggested by thinly silicified Nitzchia spp. and the lake appears
One plausible explanation for this spurious data point maybe that shallower and more evaporated than the base of the unit. Between
the very low TOC value is not reliable, or that the mineral matrix 364–366 cm more abundant diatoms indicate slightly fresher
contributes to CO2 production, thus elevating S3 (mg CO2/g rock) conditions again and a shallow, vegetated, fairly alkaline lake. In
and OI indices. Unit 2 the diatoms change to being predominantly thinly silicified
and generally preservation is poor. This is likely to be due to
increasing alkalinity through this unit. Where species can be
identified confidently, they comprise epiphytic and benthic species,
indicating moderately alkaline and shallow, HCO 3 /CO3 water (338–
334 cm), apart from at 323–324 cm where slightly deeper and
fresher conditions are indicated. Unit 3 also has poor preservation
to begin with, improving up through the unit, with the species
present suggesting fluctuating conditions, shallow water and
indications of NaCl chemistry (C. clypeus). The presence of vegeta-
tion is still evident. Unit 4 is rich in silica debris but has few dia-
toms. Those which are present indicate shallow, NaCl and CO3
waters, at the edge of a lake, with some vegetation. Again there is
very poor preservation in Unit 5 but there are indications of a mixed
alkaline (NaCl–CO3), shallow water flora and there is less sign of
epiphytic taxa. At the top of the section (40 cm) only very small
fragments of diatom can be seen again, possibly because the lake
was too alkaline to preserve the frustules well. Generally the dia-
toms indicate shallow water, with moderate to high alkalinity, with
only the sample from 388 to 390 cm suggesting fresher water.
There is more evidence for NaCl waters in the middle part of the
core (e.g. 300–250 cm), with increasing loss of preservation in the
upper section. This suggests that an already shallow system became
unsuitable for diatom preservation and no longer supported many
aquatic macrophytes.
4.6. Tephra
4.6.1. Methods
Selected samples from the Tepexpan sedimentary succession
were analysed for major element geochemistry of the tephras
(volcanic glass) using a Cameca SX100 Electron Probe Micro-
analyser at 10 kV beam voltage (Appendix A). Total iron is
expressed as FeO and analyses of pure glass with totals above 95%
were normally present. Carbon coated slide samples were analysed
using the wavelength dispersive method, with an accelerating
voltage of 10 kV. Homogeneous Lipari glass and an andradite
Fig. 7. Van Krevelen plot for selected Tepexpan samples. The organic matter of three of (garnet) were analysed at regular intervals to establish the probe
the samples is classified as Type III–IV kerogen. stability.
ARTICLE IN PRESS
Fig. 8. Diatom species changes from selected samples from the new Tepexpan section.
4.6.2. Tephrachronology results and interpretation piedmont fringe sequences of pyroclastic deposits, lava flows and
The results presented are the first study of the tephra glass fluvial sediments were considered to be mainly Pleistocene in age
geochemistry at this archaeological site, although Gonzalez and (Mooser et al., 1996).
Huddart (2007) presented data from the nearby Tocuila Mammoths (c) In the succession between 340 and 370 cm there is a mixed
site, whilst Ortega Guerrero and Newton (1998) studied tephra tephra component with rhyolitic (73–75% SiO2) and basaltic–
geochemistry from mainly the southern part of the Basin of Mexico. andesitic (64–65% SiO2) grains. This suggests that the reworking of
Prior work in the basin by Mooser (1967), Lambert (1986) and the rhyolitic ash from the upper basin catchment still continued,
Bradbury (1989) used tephras as stratigraphic markers using along with the complete reworking of the Pomez con Andesita
mainly subjective criteria (colour and grain size of the layers). The (Pumice-with-Andesite ash) (ca 14,500–14,000 BP). This ash has
tephra geochemistry is given in the Appendix A. A summary of the not been recognised in situ at Tepexpan, although it has been noted
results is as follows: at many other locations throughout the basin, such as Tlapacoya,
(a) There are a range of basaltic, basaltic–andesite and rhyolitic Tocuila and pockets of this ash have been found in lacustrine clay as
ash contributions to the sediments, with strong evidence of ‘ash balls’ at the nearby Santa Isabel Iztapan II Mammoth site
reworking. The only real tephra layer in situ is the 20 cm basaltic– (Huddart and Gonzalez, 2006). Bradbury (1971) proposed that
andesite tephra in samples Tx-59 and Tx-55. This means that there Tepexpan Man was as old as the Pumice-with-Andesite ash,
are several characteristic tephra markers from the Basin of Mexico because he found pumice clasts from this ash in the sediments
that are missing in the Tepexpan sedimentary succession, based on immediately below the caliche layer, where the skeleton was found
the 14C dates obtained for this section (see the results reported in (Fig. 2). This illustrates that there has been significant reworking of
Mooser (1967) and Lozano garcia and Ortega Guerrero (1998) and this ash at Tepexpan and many other sites in the basin, such as at
the discussion in Gonzalez and Huddart (2007)). Tlapacoya from the slopes above the hill (Huddart and Gonzalez,
(b) In the lowest part of the succession between 370 and 2006) and at Tocuila from the flanks of lahar channels (Gonzalez
404 cm, dated to between 16,730 75 and 19,110 90 BP, there is and Huddart, 2007). From these examples it is clear that the
a component of rhyolitic tephra (72–75% SiO2) in the silty fine sand, reworking process is very common and must be recognised in the
although there are a few grains at 64% SiO2 (Tx-151, Tx-142). sections, to avoid confusion and to be able to construct a correct
However there is no obvious rhyolitic tephra marker reported and accurate tephrachronology.
previously in the basin at this time period (which could have been (d) The only tephra found in situ in the section is a 20 cm
reworked), except for the rhyolitic component in the pyroclastic basaltic–andesitic to andesitic tephra found between ca 200 and
flows and ash on the piedmont fringe to the East. It has been sug- 220 cm, with SiO2 contents between 54 and 59% (Appendix A). It
gested by Huddart and Gonzalez (2006) that these pyroclastic flows cannot be the Gran Ceniza Basaltica because of its age and it may be
and associated rhyolitic ash were active as late as ca 37,000 BP from the Chimalpa tephra dated to 14,015 130 BP (Ortega-Guerrero
one of the rhyolitic domes found around the Quetzaltepec Sierra and Newton, 1998), and although this had similar % of SiO2 (54–
(Volcan Telapon). This is much later than had formerly been sug- 58%) it was only 5 cm thick in the southern Chalco Basin. The likely
gested, as these volcanoes were always thought to be Tertiary in age source of this tephra may be one of the monogenetic cinder cones
(Mooser, 1975). It is clear that the relationships between these close to Tepexpan such as Cerro Chiconautla, Cerro Tlahuilco or
deposits, and their dating, is poorly understood because the dacitic Cerro Santiago.
lavas and pyroclasts with large quantities of pumice that together (e) Between 200 and 130 cm the buff fine sand has a major
form the Tlaloc Formation were thought by Vazquez Sanchez and component of volcanic ash, although with no obvious tephra
Jaimes Palomera (1989) to be between 0.6 and 0.7 Ma, while the horizons in situ. For example, in samples Tx-29, Tx-14 and Tx-6
ARTICLE IN PRESS
there is a mixed tephra population including basaltic, basaltic–

andesite and rhyolitic grains. There is no evidence of the 10,500 BP,
Upper Toluca Pumice in situ which produced a medium to fine sand
pumice layer up to 50 cm thick in the Basin of Mexico (Arce et al.,
2003), which is present at the Tocuila Mammoths site (Gonzalez
et al., 2001; Gonzalez and Huddart, 2007). The rhyolitic component
from the upper drainage basin may also have contributions from
the rhyolitic tephras (76–78% SiO2) dated to around 13,450 40 BP
from Tequexquinahuac in the upper part of the piedmont slope to
the south-east of Tepexpan (Huddart and Gonzalez, 2006). This
could be equivalent to the San Martin tephra (67–73% SiO2),
described by Ortega Guerrero and Newton (1998) from a location in
the southern Lake Chalco Basin, which was dated to
13,990 100 BP However, it seems more likely that there are two
different rhyolitic eruptions in this period. At the Tocuila Fig. 9. Uranium and U-series dates across the bone section from the outer to the inner
Mammoths site there is a beige, sandy silt, with pumice clasts just surface. Dates rejected as contaminated with detrital Th are shown in grey and
above a white rhyolitic ash (73–74% Si). It included gastropods excluded from the calculation of the mean age of 4700 200 years.
which have been dated by AMS to 10,016 39 BP (OxA-15,840).

Whilst it is possible that this ash is a phase of the Upper Toluca 5.1. Uranium-series dating of the Tepexpan human skeleton
Pumice, the silica content from that tephra is between 69 and 73%
(Newton and Metcalfe, 1999) and hence lower than this ash found Direct dating of bone by U-series presents a particular challenge
in Tocuila. The other components to this part of the Tepexpan because bone remains an open system for uranium, and any
sedimentary succession must come from the basaltic and basaltic- attempt to calculate a date from U-series measurements requires
andesite ash, lower in the sedimentary succession, the Pumice- a model of uranium uptake. Here we employ the method of Pike
with-Andesite and Upper Toluca Pumice, reworked from the lake et al. (2002) who have shown that the U uptake regime for a bone is
margins and from all these sediments from the drainage basin to reflected in the distribution of U and U-series isotopes across a bone
the north-east. This reworking can be also seen in sample Tx-6 section. Analyses were made on a section of bone from the
where there are incorporated mud balls with volcanic ash grains in Tepexpan human skeleton, from the outer (periosteal) to inner
the centre. (endosteal) surfaces using laser ablation plasma mass spectrometry
(f) Below the caliche layer, the pale white, CaCO3-rich silt has according to the method of Eggins et al. (2005). The U distribution
a mixed tephra population with components from the PWA, UTP is W-shaped while the dates are relatively uniform towards the
and a basaltic ash (Tx-4) at 70 cm and again evidence for reworking outer surface getting older towards the centre (Fig. 9, Table 3).
with the presence of muddy balls. Gastropods from this layer at However, the low 230Th/232Th indicates significant contamination
88.5 cm have been dated to 6333 30 BP whilst sample Tx-2 at by detrital 232Th, presumable from sediment, for many of the
50 cm above the caliche again shows a mixed population and samples. Detritus brings with it230Th, and in most cases will lead to
muddy balls and has been dated to 612 22 BP However, there is older apparent ages. In this case, all of the older dates towards the
another rhyolitic tephra from core E at Chalco (the Huitzilzingo centre of the bone are heavily contaminated. Rejecting any dates
tephra) dated to 2645 55 BP (Ortega Guerrero and Newton 1998) with 230Th/232Th >10 leaves a relatively uniform distribution of
but this is not apparent at Tepexpan. Whilst there is no obvious dates with a mean of 4700 200 years. A uniform distribution of U-
source for the rhyolitic tephras in the Basin of Mexico apart from series dates, with a W-shaped uranium concentration profile is
the Tlaloc–Telapon massif volcanic system, this has implications for indicative of the early cessation of uranium uptake by the bone
the region as the associated dating suggests that perhaps these (Pike et al., 2002), so we have taken the mean of the closed system
rhyolitic, explosive eruptions have been periodic throughout the dates to be out best estimate of the age of the bone. There could, of
time period from 37,000 to 2645 BP in the Basin of Mexico. course, have been a delay before the onset of uranium uptake, so
this should be taken as a minimum age and therefore not incon-
5. Tepexpan chronology sistent with the radiocarbon date in humus in the layer where
Tepexpan Man was found, of 5600 40 years BP reported by
One of the main problems for the study of the Tepexpan Solleiro Rebolledo et al. (2006).
Palaeoindian site has been the lack of a well constrained chro-
nology. This is because the only directly dated material was the 5.2. Radiocarbon dating
Tepexpan human skeleton itself, with lots of problems of contam-
ination with younger material added to the specimen at the time of Two gastropod shell samples for AMS radiocarbon dating were
discovery during the 1940’s, to try to consolidate it. This situation extracted from a section adjacent to the Tepexpan Man section
gave as a result an erroneous young (contaminated) radiocarbon described here and correlated to the section whilst in the field
date of around 2000 years BP (Lorenzo, 1989; Stafford et al., 1991; (equivalent depths of 46 and 88.5 cm respectively). The whole and
Gonzalez et al., 2003). We report in here a new Uranium-series date broken shells were pre-treated using an acid wash procedure to
of 4700 200 years BP (Fig. 9) that has been obtained directly from remove the external shell layer in case of any post-depositional
the skeleton, which is more or less in agreement with an AMS change and crushed before acid hydrolysis treatment, CO2 collec-
radiocarbon date of 5600 40 years BP (Beta 205,077) obtained by tion, graphitisation and AMS dating. Two additional samples for
Solleiro Rebolledo et al. (2006) for humus associated with the layer dating were taken from ostracod-rich layers at 365 and 400 cm.
where the skeleton was found. Also, an independent chronology for Ostracods were separated by treating with 5% sodium hypochlorite,
the new sediment section was obtained here using a combination sieved at >180 mm, then screened under a microscope where
of AMS radiocarbon dating on ostracods, gastropods and detailed juveniles, broken shells, any with carbonate overgrowths were
tephrochronology studies. The different methods used are removed, along with other contaminants. Between 10 and 20 mg of
explained below. ostracods were selected and checked for the presence of aragonite
ARTICLE IN PRESS
Table 3 Pleistocene–early Holocene related to volcanic events and subse-

U-series measurements on the Tepexpan human skeleton. quent reworking as the ones already found in the lahar channels
Analysis 230
Th/232Th 234
U/238U 230
Th/238U Closed and discontinuities in the Tocuila Mammoths sedimentary
system succession (Gonzalez and Huddart, 2007) and those proposed from
date (ky) ca 14,000 years BP onwards (Bradbury, 1989; Lozano-Garcı́a and
1 23.2 1.382 0.032 0.0488 0.005 3.903 0.457 Ortega-Guerrero, 1998). The significant shift in the geochemical
2 32.4 1.402 0.033 0.0643 0.008 5.099 0.671
proxies occurring at 310 cm at the lithological boundary between
3 8.77 1.388 0.033 0.0578 0.007 4.621 0.640 Reject
4 45.7 1.370 0.032 0.0574 0.007 4.649 0.615 the ash-rich fine sand of Unit 2 to the more calcareous, paler fine
5 45.1 1.390 0.033 0.0593 0.007 4.738 0.623 sand of Unit 3 (see below) might indicate a discontinuity. Bradbury
6 46.8 1.385 0.032 0.0613 0.007 4.918 0.625 (1989) suggested that when the margins of Lake Texcoco were
7 9.50 1.419 0.035 0.0512 0.006 3.994 0.540 Reject exposed, or covered by shallow water, the sediments would accu-
8 65.7 1.415 0.034 0.0640 0.008 5.025 0.697
9 7.37 1.365 0.033 0.0566 0.007 4.599 0.664 Reject
mulate very slowly and thus creating a reliable detailed chronology
10 36.4 1.401 0.033 0.0560 0.007 4.431 0.634 at Tepexpan maybe difficult. Hence the chronology in here is only
11 11.9 1.372 0.034 0.0421 0.006 3.387 0.525 a guide to the timing of the inferred climate changes.
12 15.4 1.422 0.034 0.0717 0.010 5.612 0.860
13 4.65 1.410 0.035 0.0631 0.007 4.975 0.643 Reject
6. Overall interpretation of lake conditions
14 7.99 1.393 0.033 0.0749 0.010 6.000 0.850 Reject
15 12.3 1.412 0.034 0.0616 0.008 4.844 0.661
16 6.24 1.400 0.033 0.0471 0.006 3.719 0.525 Reject Bradbury’s. (1989) diatom ecology reflects the fluctuating
17 3.01 1.385 0.032 0.0805 0.008 6.495 0.745 Reject influence of the Rio San Juan that feeds into the Texcoco Lake via
18 5.78 1.249 0.028 0.0819 0.008 7.360 0.827 Reject the Teotihuacan Valley. Deeper water periods associated with clays
Low 230Th/232Th indicates contamination with detrital thorium, and dates with and planktonic species also indicate saline water; whereas fresh-
230
Th/232Th <10 have been rejected. The mean of the uncontaminated dates is water and alkaline water diatoms occur during the regression of
4700 200 years.
Lake Texcoco, when the river water influence becomes relatively
strong. The multi-proxy study discussed here also suggests large
with Fiegl’s solution. As the staining indicated only calcite, ostra- changes in the nature of Lake Texcoco at its margins in terms of the
cods were lightly crushed and rinsed in water, before being AMS balance between aquatic and terrestrial plants, C3 and C4 plants,
dated using direct combustion. The dates are reported as uncali- saline, alkaline and freshwater conditions and volcanic activity. The
brated radiocarbon years BP after correction for isotopic fraction- overall profile represents a volcano-lacustrine, regressive sequence.
ation (Table 4). Unit 1 The presence of clays and silts suggest relatively deeper
water. The water content of the sediment is high, and Bradbury
5.3. Sediment sequence dates (1989) suggested this may be due to high clay content in the
sediment caused by in-wash of weathered ash from the catchment.
The four AMS radiocarbon dates range from 19,110 90 years at Other proxies also suggest high in-wash from the catchment, the
the base of the section to 612 22 years at 46 cm from the surface main peaks in water content corresponds to peaks in %C which may
(Table 4). Assuming continuous sedimentation, the rate of accu- indicate increased nutrient and plant inflow to the margins. The
mulation ranges from w0.15 mm year1 (19,110 90 to organic geochemistry sample at 389 cm suggests organic material
16,730 75), w0.27 mm year1 (16,730 75 to 6334 30) and dominated by terrestrial plant waxes and that these are likely to be
w0.07 mm year1 (6334 30 to 612 22). Rates of 0.15 and a mix of terrestrial plants from the catchment and emergent
0.27 mm year1 are fairly slow, but not uncommon for lake envi- macrophytes. This is also suggested by the C/N and d13Corganic data
ronments, however given the marginal location of the Tepexpan which points to terrestrial plant material of C3 type. The fact that
site in relation to the lake basin, with a river nearby, some sediment carbonate d18O and d13C values are high to begin with (around 0&)
erosion events are to be expected. The very slow accumulation rate and increase through the unit, suggest that the lake water was fairly
from the period between 6334 and 612 years BP almost certainly saline and was progressively evaporating through the unit, possibly
does indicate a discontinuity. There are suggestions that the sedi- as climate became drier. The variability in the unit is emphasised by
mentary record of the Holocene in Mexico is commonly incomplete the changing diatom ecology, which fluctuates between periods of
due to human disturbance of the environment and also due to freshwater and very alkaline conditions in a marginal lake setting.
intense drying in some locations and this may be reflected here too. Thus overall the proxies point to high river inflow into the margins
In the late Pleistocene the large accumulation of ash in Unit 4 was of a saline lake, indicating a variable climate, possibly becoming
deposited quickly as a series of instantaneous ashfalls and hence drier through the unit.
the accumulation rate between 16,730 75 and 6334 30 BP may Unit 2 by contrast is less variable and has a generally coarser
overestimate the true lake sedimentation rate. grain size, suggesting shallower water. Carbonate d18O and d13C
Based on the tephra studies it seems likely that there are data suggest further evaporation and concentration of lake water
important gaps in the sedimentary succession during the late indicating the continued aridification of climate and lower lake
levels. C/N and d13C begin and end with higher values typical of
terrestrial plants, but in the centre of the unit (at 331–345 cm) fall
Table 4
New AMS radiocarbon ages on marginal lake shells and ostracods from Lake Texcoco
to below 5 and this is concurrent with higher d13C values that
at the Tepexpan Man sedimentary sequence. approach 18&. This would suggest a period of increased phyto-
plankton numbers, indicating algal-rich pools utilising bicarbonate
Section Sample description Laboratory d13C Radiocarbon age
depth (cm) number PDB & 14
C years BP 1s rather than CO2. The diatom preservation is generally poor in this
unit and is likely to be due to increasing alkalinity which is sup-
46 Mixed marginal lake OxA-15,842 8.8 612 22
gastropod shells ported by some epiphytic and benthic species which indicate
88.5 Mixed marginal lake OxA-15,841 2.9 6334 30 moderately alkaline and shallow, HCO 3 /CO3 water (338–334 cm). A
gastropod shells brief wet phase indicated by a band of silt to fine sand at 321–
365 Ostracod shells OxA-15,842 0.6 16730 75
326 cm and diatoms at 323–324 cm indicate slightly deeper and
400 Ostracod shells OxA-15,877 2.8 19110 90
fresher conditions. Thus the lake margins at this point appear to
ARTICLE IN PRESS
have become relatively shallow and stagnant, with less river inflow diatom preservation, increase in C4 vegetation and lower d18O and
and the development of alkaline pools, although conditions remain d13C carbonates in the upper section suggest the lake was drying
variable with interruptions from brief wetter periods. out and the spring and river water was feeding more marginal
Unit 3 is much like Unit 1 and the organic geochemistry suggests parts.
a change back to the dominance of terrestrial or emergent plants
rather than aquatics. C/N ratios >15 concur with this indicating 7. Discussion
a higher plant dominance. Very high %C at the beginning of the unit,
followed by a rapid but progressive decline, suggests in-wash is Although the chronology of the sedimentary succession is
high and climate is humid at the start, then drying through the unit limited, general comparisons with existing palaeoclimate recon-
supported by high and increasing d18O and d13C carbonates. Diatom structions in the basin can be made as follows:
preservation improves through the unit and suggests fluctuating
shallow water, including indications of NaCl chemistry, with the 7.1. Last Glacial Maximum (19,000–16,500 years BP)
presence of vegetation.
Unit 4 The dominance of volcanic ash makes interpretations of Unit 1 captures the later part of the LGM (ca 19–16.5 ka BP)
the palaeoclimate difficult as little organic material is preserved. although there maybe some hiatuses present. The geochemical
What is there suggests a submerged or floating aquatics source for and diatom interpretation here points to high sediment in-wash
organics from the organic geochemistry and limited C/N data. from the catchment into a relatively saline lake. This suggests that
d18O and d13C from preserved carbonates are lower and variable the climate may have been generally dry in the basin (increasingly
particularly to begin with and initially d18O and d13C carbonates so), tying in with widespread evidence for the expansion of
fall suggesting an increase in freshwater inflow. The few diatoms grasslands at this time (Lozano Garcia et al., 1993; Lozano Garcia
preserved in this environment suggest shallow, NaCl and CO3 and Ortega Guerrero, 1994; Ortega Guerrero et al., 2000). The
waters at the edge of a lake with some vegetation. Thus the interruption of this aridity, with evidence for high river inflow
limited evidence points to shallow, freshwater re-established in from the catchment suggests a more variable picture though,
a highly variable, volcanically-disturbed horizon. Solleiro-Rebol- supporting other evidence for some increased humidity (Brad-
ledo et al. (2006) suggest that their syn-sedimentary Fluvisol bury, 1997; Lozano Garcia and Ortega Guerrero, 1998; Metcalfe
developed in a transitional environment at the periphery of et al., 2000; Sedov et al., 2001; Solleiro Rebolledo et al., 2006).
a drying lake with sedimentation hampering the soil profile, Indeed Bradbury (1989) infers extended open water environments
although they recognised weak humus accumulation, silica neo- at 19–18 ka BP and increases in winter precipitation, however the
formation and a restricted and late calcite precipitation (poten- chronology is uncertain. There is some evidence in northern
tially indicating a hiatus?). Mexico for the cessation of the summer monsoon regime, wetter
Unit 5 Following this volcanically-disturbed period, the lake conditions (Metcalfe et al., 2002; Castiglia and Fawcett, 2006) and
changes character dramatically to one of initially aquatic vegeta- for the southward displacement of the mid-latitude Westerlies
tion in concentrated alkaline pools (low C/N ratios and high (Brunner, 1982) due to the expansion of the Laurentide ice sheet
d13Corganic), with increased evaporation, suggested by the to the north and the effects may have reached as far south as
carbonate isotopes and diatom species indicative of increasingly central Mexico (González Quintero and Fuentes Mata, 1980; Van
alkaline conditions At the top of the section (Unit 6), the organic Devender and Burgess, 1985; Bradbury, 1997). Metcalfe et al.
geochemistry suggests a mix of aquatic and terrestrial plants and (2000) suggest that the LGM was probably cool and humid due to
higher d13Corganics for the terrestrial plants may relate to an decreases in evaporation rather than increased precipitation. An
increase in C4 Cyperaceae and thus possibly spring flow. However, alternative explanation for enhanced periods of river inflow is
increasing human influence in the region, in the form of cereal increases in meltwater flow from mountain glaciers in the basin.
growing, is a more likely reason for the increase in C4 plants. The Lozano Garcı́a et al. (1993) postulate that Popocatépetl volcanism
consistently high d13Corganic values also suggest that the aquatic may have increased meltwater runoff to Lake Chalco in the LGM,
portion of the organics were utilising HCO 3 , thus lake conditions however this was most prevalent from 22.5 to 19 ka BP (Bradbury,
are likely to have been very alkaline. The diatom flora suggest less 1989; Caballero Miranda and Ortega Guerrero, 1998). Fluctuations
aquatic vegetation and a mixed alkaline (NaCl–CO3), shallow between acid and neutral water at Chalco between 19 and 15 ka
water flora. Solleiro Rebolledo et al. (2006) recognised palaeosols BP may also have been due to volcanism affecting the mountain
in this part of the succession which had characteristics related to ice fields above the lake, or the direct consequence of ashfall on
hydromorphic conditions which resulted from high groundwater lake chemistry (Caballero Miranda, 1997). The Teotihuacán Valley
levels indicated by Fe–Mn (with humus) hypercoatings in the could have also received increased meltwater associated with
voids. They also had characteristics related to a dry environment, enhanced volcanism at this time which found its way into the
such as deep fissures, indicating intense desiccation and river input at Tepexpan. Alternatively, the enhanced inflow into
secondary carbonate precipitation, the latter especially in the Lake Chalco is thought to have been high enough to allow it to
uppermost palaeosol and modern soil. There are also impure clay overflow into Lake Texcoco, increasing the freshwater input
coatings in the strongly calcified palaeosols and the presence of further north (Caballero Miranda, 1997).
lithogenic calcite suggests that they were never carbonate-free.
They attribute this particle translocation to infiltration of flood 7.2. Lateglacial (15,000–10,000 years BP)
waters, saturated with suspended sediment and so conditions
suitable for illuvial pedofeatures were possible even in the The lack of dates from this part of the sedimentary succession
carbonate-rich soils of an arid environment. makes pinpointing the Lateglacial problematic, although the
Generally, the evidence indicates periods of high inflow at the tephras indicate that it is likely to fall across Units 2–4. There are
base of the sedimentary succession that fluctuated with periods of certainly volcanically-disturbed and reworked sediments and the
marginal flooding by saline water and periods of enhanced evap- suggestion of hiatuses in the Tepexpan sedimentary succession at
oration. In the middle of the sedimentary succession, volcanic this point indicate it may fall during this time period.
events disturb the proxy evidence but lower lake levels and some Many studies from Lake Chalco suggest a moist Lateglacial
freshwater input is suggested. Following this, increasing loss of (Lozano Garcı́a et al., 1993; Caballero Miranda and Ortega
ARTICLE IN PRESS
Guerrero, 1998). However, much inter-basin variability is based on some evidence related to periodic flooding. In the Lerma
apparent and a pollen study from Lake Texcoco and Lake Chalco Basin, to the west of this basin, the period from 8000 to 3700 BP is
comparing the period sees evidence for drying conditions in the warm and dry with some brief wet periods at 6800 and 5100 BP
Lateglacial, with Lake Texcoco becoming saline and the fresh- (Caballero et al., 2002; Lozano-Garcı́a et al., 2005). There is no
water marsh areas expanding, possibly due to reduced catchment evidence for such wet periods from this sedimentary succession,
vegetation due to the intense volcanism (Caballero Miranda and apart from the hydromorphic conditions and flooding evidence
Ortega Guerrero, 1998). In the north of the basin, deposition provided by Solleiro-Robolledo et al. (2006).
apparently ceased at Tecocomulco indicating very dry conditions
(Caballero Miranda et al., 1999). The conditions in the central part 7.4. Mid–late Holocene (5000 years BP–present)
of the Tepexpan sedimentary succession are certainly variable
with the lake initially being shallow and quite stagnant, with There is almost certainly a hiatus between the two upper
reduced inflow (Unit 2), changing to more humid conditions and radiocarbon dates from the sedimentary succession in the late
deeper water at the beginning of Unit 3, before indications of P/E Holocene, caused by periods when the lake dried out resulting
declining, with some suggestions of brief periods of spring or in erosion at the site. From 6300 BP, the organic geochemistry
river inflow. Deposition of volcanic ash in Unit 4 obscures many and C/N data suggest an increase in terrestrial plants and
of the proxy indicators, however the carbonate d18O and d13C expansion of C4 plants. Increases in C4 Cyperaceae, and thus
suggest P/E remained low during the volcanism. This period in possibly spring flow, may be one part of this expansion, but arid
the basin is unlikely to be fully resolved until sedimentary C4 grasses and the expansion of agriculture are likely. The
successions less affected by volcanism can be located, although diatom flora also indicate less aquatic vegetation and a mixed
this may be impossible because of the major ash thicknesses for alkaline (NaCl–CO3), shallow water lake. The indurated nature of
the Pumice-with-Andesite and Upper Toluca Pumice that fell in the sediment additionally signifies intense drying through the
this basin (Huddart and Gonzalez, 2006). In their paleosol study late Holocene.
at Tepexpan, Solleiro Rebolledo et al. (2006) measured a rela- Many Mexican records are highly disturbed in the late Holocene
tively low d13C (soil organic matter) value on humus (25&), resulting in poor understanding of climatic fluctuations. There is,
equivalent to the lower part of Unit 4 in this study, suggesting however, widespread evidence for drying in the mid Holocene,
a lowering of lake level and the development of a swamp which seems to have increased in severity from south to north
ecosystem, although still with relatively humid climatic across present day Mexico and up in the USA. Typical of many
conditions. Mexican records, widespread deforestation and human disturbance
is evident from 3500 to 1200 BP. Buckler et al. (1998) suggest drier
7.3. Early–mid Holocene (10,000–6000 years BP) conditions beginning at 6500 BP and by 5000 BP the region was
more xeric than today. Solleiro Robolledo et al. (2006) argue that
As the period of intense volcanism is dated as ending around the later cultural and agricultural development of the Teotihuacan
the onset of the Holocene, Unit 5 and the bottom part of Unit 6 civilisation appeared to develop in an arid environment, backed up
are likely to represent the early–mid Holocene period. Although by the archaeological evidence for water management and irriga-
difficult to age, Unit 5 indicates alkaline, very shallow conditions tion (Nichols et al., 1991; McClung de Tapia, 2000). For the last
in the lake and stagnant pools of water with aquatic plants. The 2000 years though the climate appears not to have changed dras-
carbonate data suggest that these pools were spring or river fed tically (Lounejeva Baturina et al., 2006), although evidence for this
and that the lake was reduced in size at this point, indicating arid period from the Basin of Mexico is extremely sparse due to
conditions. The base of Unit 6, where Tepexpan Man Was found shrinkage and deflation of the most recent sediments following
also suggests aridity in terms of the carbonate data, but higher C/ drainage of the basin.
N ratios suggest more terrestrial plants, perhaps from the
expansion of C4 plants and/or agriculture. Despite some parts of 8. Conclusions
central Mexico being wet in the early Holocene (Metcalfe, 2006),
the Basin of Mexico has several reconstructions that point to very There are several general conclusions which can be made with
arid conditions (e.g., Bradbury, 1989; Lozano Garcı́a et al., 1993; regard to the Tepexpan Palaeoindian site specifically and the
Lozano Garcı́a and Xehuantzi Lopez, 1997; Caballero Miranda general palaeoclimatic reconstructions for the Basin of Mexico
et al., 1999; Solleiro Robolledo et al., 2006). Bradbury (1989) during the last 20,000 years BP:
describes a shallow, fairly saline lake with periods of desiccation
and Lozano Garcı́a and Ortega Guerrero (1998) find increased (1) The multi-proxy evidence suggests large changes around the
numbers of Ruppia maritima which indicates increased salinity. margins of Lake Texcoco in terms of the balance between
Lozano Garcı́a and Ortega Guerrero (1998) argue that the pollen aquatic and terrestrial plants, C3 and C4 plants, saline, alkaline
record from Texcoco shows arid conditions, but less so than the and freshwater conditions, volcanic activity and marginal
Lateglacial period, with more water in the basin in the early reworking of lake sediments and input from the drainage basin
Holocene. Lake Chalco had low lake levels and alkaline/saline through Units 1–6 (late Pleistocene–late Holocene). These
waters at this time (Caballero Miranda, 1995). However, the changes will have had a large impact on the prehistoric pop-
glaciations of Mexican volcanoes indicate greater moisture in the ulations living by the lake shore.
period 10,000–8200 BP (Heine, 1984, 1988) as ice advances (2) Overall correlations between the multi-proxy evidence and
occurred in this period and between 8000 and 6000 BP (Vázquez climate conditions is thought to be good for the last 20,000 BP
Selem and Heine, 2004). Buckler et al. (1998) and Lozano Garcı́a from the Tepexpan site despite problems caused by persistent
and Vázquez Selem. (2005) suggest generally wetter conditions volcanic activity. However, a more detailed multi-proxy anal-
with occasional arid intervals (6000–5000 BP). Solleiro Robolledo ysis is needed in a deeper water location in Texcoco Lake to
et al. (2006) suggest that the sharp increase of pedogenic d13C unravel climatic changes that were complicated in the
(SOM) values accompanied by carbonate precipitation in their marginal lake environment of this site.
upper palaeosols marks a change towards drier environmental (3) Many of the main marker tephras found in the Basin of Mexico
conditions. However, they noted that this was a gradual transition, (Pumice-with-Andesite, Upper Toluca Pumice, etc.) are eroded
ARTICLE IN PRESS
and reworked in the marginal environments of Tepexpan, archaeologists interested in the study of the early peopling of
where river, deltaic and laharic (volcanic mudflow) events the Americas.
were common. Only one tephra was found in situ (between 200
and 220 cm depth) and that is likely to have had a local source
from a monogenetic cinder cone. It may well be equivalent to Acknowledgements
the Chimalpa tephra of Ortega Guerrero and Newton (1998). It
is suggested also that rhyolitic tephras have been an important We are very grateful to the Instituto Nacional de Antropologia e
input into the lake from 37,000 to 2645 BP which is much Historia (INAH) in Mexico for granting access to the Tepexpan
younger than previously suggested. Paleoindian Site, also José Concepción Jimenez López and José
(4) The Tepexpan Man Palaeoindian skeleton has been assigned Antonio Pompa y Padilla, from the Dirección de Antropologı́a Fı́sica
a minimum date of 4700 200 years BP (mid Holocene) by U- (DAF), INAH for help during the sampling of the Tepexpan skeleton.
series dating in this study and 5700–5900 years BP from an This work was supported by the following NERC grants: NER/T/S/
AMS date on humus in the layer where the skeleton was found 2002/00467 and NE/C519446/1 and a NERC Isotope Geosciences
(Solleiro Robolledo et al. (2006)). Younger 14C dates of around Laboratory grant (IP/814/0504). The following colleagues are
2000 years reported before for the skeleton (Lorenzo, 1989; thanked for their isotope advice: Melanie Leng and Carol Arrow-
Stafford et al., 1991; Gonzalez et al., 2003) are interpreted here smith at NIGL, Nottingham, UK. Tom Higham, ORAU, Oxford
as being contaminated by younger preservatives used on the University is thanked for advice and comments on radiocarbon
skeleton during conservation, according to the values of C/N dating of ostracods and gastropods. The NERC Electron Microprobe
obtained for the skeleton, and should be discarded from Unit at Edinburgh University, UK was used for the tephra analysis
interpretations. Despite the mid Holocene age obtained for the and Anthony Newton and David Steel are thanked for their advice
Tepexpan skeleton and associated sediments it remains a very on tephra analysis. The manuscript was improved by comments
important piece of evidence for Palaeoindian populations in from Steven Driese and an anonymous reviewer. NIGL publication
the Basin of Mexico that requires consideration by the number 886.
Appendix A
Geochemical analysis from tephra shards from selected samples at Tepexpan.

Sample SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O P2O5 Total %
Tx-2a 56.361 7.819 13.352 10.266 0.029 0.769 2.048 4 2.722 0 97.365
Tx-2b 66.144 0.574 14.899 2.793 0.08 1.465 2.789 4.94 2.536 0.165 96.385
Tx-2c 66.484 0.773 14.391 2.521 0 0.571 2.929 4.837 3.072 0.094 95.672
Tx-2d 65.881 0.594 14.191 3.017 0.045 1.11 2.75 4.731 2.611 0.213 95.144
Tx-2e 66.207 0.64 14.43 2.901 0.055 1.326 2.77 4.467 2.561 0.087 95.444
Tx-2f 65.845 0.568 14.568 2.829 0.087 1.422 2.831 4.649 2.672 0.071 95.543
Tx-2g 66.14 0.829 14.465 4.262 0.125 1.519 3.891 4.057 3.029 0.086 98.402
Tx-2h 65.819 0.803 14.356 4.094 0.077 1.405 3.599 3.877 3.392 0.109 97.532
Tx-2i 62.36 1.375 13.916 5.792 0.035 1.185 3.903 5.526 2.196 0.217 96.506
Sample Tx-2, depth: 50 cm, mix population of tephra compositions but mainly between 62 and 66% SiO2, very small tephras, presence of ‘‘muddy balls’’.
Tx-4a 73.559 0.245 11.534 1.026 0.068 0.386 0.924 3.368 4.351 0 95.6
Tx-4b 58.527 1.477 16.267 7.324 0.163 2.641 6.029 4.748 2.185 0.431 99.793
Tx-4c 74.332 0.077 12.093 0.01 0.055 0.404 3.786 4.714 0 96.47
Tx-4d 58.615 1.593 15.439 7.663 0.103 3.334 5.896 3.631 2.325 0.396 98.995
Tx-4e 68.021 0.185 15.896 1.609 0.02 0.364 3.258 4.566 2.224 0.055 96.199
Tx-4f 69.835 0.708 14.439 2.797 0.042 0.58 2.106 4.247 3.167 0.055 97.976
Tx-4g 69.358 0.751 11.935 3.017 0.062 0.469 2.478 4.074 3.14 0.687 95.97
Tx-4h 73.577 0.025 12.613 0.539 0.124 0.062 0.319 3.952 4.008 0.024 95.244
Tx-4I 72.155 0.235 14.174 0.618 0.019 0.183 1.693 4.268 3.687 0.063 97.096
Sample Tx-4, depth: 70 cm, very mix population of tephras, reworked sediments.
Tx-6c 62.554 0.768 18.022 3.374 0 0.928 5.912 4.619 1.794 0.156 8.127
Tx-6e 63.908 1.146 15.632 4.635 0.096 1.845 4.211 4.682 2.166 0.357 8.679
Tx-6f 54.727 0.571 20.279 3.156 0.051 2.687 8.225 4.485 1.096 0.154 5.431
Tx-6h 57.848 1.433 16.476 6.558 0.137 2.656 6.756 4.099 2.105 0.734 8.801
Tx-6i 58.158 1.375 16.74 5.211 0.067 2.073 5.92 4.018 2.827 0.766 7.156
Tx-6j 67.185 0.752 14.047 2.659 0.013 0.648 3 3.753 4.195 0.109 6.363
Tx-6k 70.155 0.765 14.256 2.46 0.074 0.316 2.293 4.831 3.054 0.126 8.332
Tx-6m 62.827 1.461 15.523 5.475 0.102 1.53 4.491 4.72 2.347 0.41 8.884
ARTICLE IN PRESS
(continued )
Tx-6o 71.824 0.093 12.049 2.583 0.083 0.789 0.501 3.208 4.38 0 5.511
Tx-6p 73.341 0.016 12.084 0.621 0.165 0.135 0.395 3.121 5.10 0.032 5.011
Tx-6r 69.173 0.266 13.47 2.054 0.064 0.46 1.939 4.388 2.594 0.064 4.474
Sample Tx-6, depth: 110 cm, mix population of tephras, presence of ‘‘muddy balls’’ with ash grains in the centre, lots of organics, very small tephra grains.
Tx-29a 73.787 0.033 11.822 0.844 0.033 0.025 0.4 3.728 4.156 0 94.2
Tx-29b 72.167 0.019 12.073 0.629 0.075 0.034 0.313 3.909 4.258 0 93.475
Tx-29c 64.667 0 21.008 0.065 0 0.026 2.245 9.696 1.074 0.055 98.837
Tx-29d 74.282 0.115 13.25 0.194 0.036 0.031 0.673 2.756 6.473 0 97.809
Tx-29e 59.833 1.635 14.753 7.527 0.216 2.255 5.283 4.595 1.963 0.634 98.695
Tx-29f 77.9 0.075 11.536 0.467 0.052 0.038 0.829 3.927 3.078 0 97.902
Tx-29g 70.002 0.258 14.524 0.76 0.052 0.143 2.042 3.962 4.608 0 96.35
Tx-29h 76.794 0.05 12.344 0.662 0.033 0.244 1.549 4.034 3.076 0 98.786
Tx-29i 55.21 1.124 20.151 4.69 0.126 1.887 8.448 4.073 1.482 0.438 97.629
Tx-29j 72.693 0.043 12.374 0.677 0.091 0.044 0.323 3.808 4.358 0 94.411
Tx-29k 73.464 0.16 12.803 1.082 0.013 0.218 1.187 3.588 3.651 0.063 96.229
Sample Tx-29, depth: 159–161 cm, laminated white sand, ash þ diatoms, SiO2 mainly 72–74%, but a few grains with w55–59%.
Tx-55a 57.456 1.637 16.569 5.473 0.124 2.248 7.294 4.728 1.689 0.839 98.059
Tx-55b 56.515 1.363 14.115 6.996 0.108 4.838 7.519 3.685 1.905 0.713 97.756
Tx-55c 57.973 1.689 13.649 7.315 0.117 2.455 4.905 4.227 2.939 0.875 96.145
Tx-55d 56.79 1.202 19.231 4.785 0.035 1.249 7.035 4.974 1.552 0.446 97.3
Tx-55e 56.583 1.507 12.683 8.237 0.263 5.101 4.956 3.055 1.999 1.053 95.437
Tx-55f 59.289 1.682 13.976 6.917 0.194 2.637 4.832 4.724 2.766 1.107 98.123
Tx-55g 56.477 1.104 16.071 6.54 0.141 4.596 6.568 4.222 1.328 0.484 97.53
Tx-55h 54.53 0.927 17.918 5.311 0.105 2.876 7.175 4.256 1.35 0.635 95.083
Tx-55i 57.705 1.673 15.007 7.687 0.172 2.615 5.992 4.028 2.16 0.838 97.877
Tx-55j 57.637 1.127 18.828 4.472 0.133 1.154 6.578 4.764 1.989 0.611 97.293
Tx-55k 56.56 1.496 15.192 6.833 0.185 2.859 5.835 4.163 2.349 0.93 96.402
Tx-55l 55.514 1.237 17.748 6.007 0.086 1.859 6.947 4.346 2.174 0.775 96.694
Tx-55m 56.906 0.812 21.14 3.836 0 0.81 8.091 4.883 1.263 0.436 98.177
Tx-55m 55.882 1.455 14.125 6.928 0.152 3.972 8.01 4.284 1.357 0.703 96.87
Tx-55o 57.685 1.423 15.783 5.513 0.092 2.386 6.36 4.507 2.034 0.709 96.491
Tx-55p 55.32 1.532 13.601 7.113 0.201 3.765 7.284 3.738 2.396 0.606 95.556
Tx-55q 54.907 0.479 22.121 3.554 0.17 2.773 10.154 4.153 0.658 0.306 99.276
Tx-55r 55.529 1.049 17.358 5.644 0.137 2.802 7.52 4.191 2.017 0.624 96.872
Tx-55s 55.784 1.535 14.266 7.476 0.219 2.661 5.692 3.897 1.936 0.709 94.175
Tx-55t 56.926 1.703 13.212 7.049 0.098 2.575 4.803 4.215 2.469 0.958 94.008
Sample Tx-55, depth: 210–212 cm, dark brown silty volcanic ash, SiO2 ¼ 54–59%, tephra in situ.
Tx-59a 57.467 1.289 16.389 6.223 0.161 2.686 7.056 4.277 1.963 0.729 98.241
Tx-59b 56.501 1.606 14.029 8.647 0.206 3.406 5.821 3.991 2.24 0.975 97.422
Tx-59c 55.851 1.074 18.281 5.354 0.084 2.072 7.072 4.576 1.596 0.7 96.659
Tx-59d 57.941 1.5 14.935 7.151 0.145 2.723 5.744 4.274 2.458 0.901 97.77
Tx-59e 56.027 1.535 15.255 7.786 0.177 2.91 6.5 3.995 2.188 1.107 97.48
Tx-59f 57.071 1.308 18.235 6.251 0.058 1.779 6.923 4.848 1.542 0.639 98.655
Tx-59g 56.629 1.58 14.175 8.137 0.215 2.998 5.607 4.211 2.144 0.893 96.592
Tx-59h 56.338 0.963 17.675 5.881 0.149 3.287 7.671 4.009 1.45 0.623 98.046
Tx-59i 56.53 1.234 16.553 6.615 0.074 3.094 6.904 4.185 1.84 0.722 97.751
Tx-59j 56.611 1.541 14.387 8.279 0.151 4.152 6.815 3.916 1.983 0.828 98.663
Tx-59k 56.978 1.632 14.4 8.154 0.122 3.03 6.007 4.048 2.394 0.875 97.641
Sample Tx-59, depth: 218–220 cm, dark brown silty volcanic ash, SiO2 ¼ 56–57%, tephra in situ.
Tx-75a 66.501 0.489 15.054 2.45 0.058 0.504 1.723 5.25 3.949 0.086 96.064
Tx-75b 75.0 0.049 11.746 0.986 0.013 0.024 0.435 3.807 4.153 0 96.212
Tx-75c 73.853 0.082 11.858 0.608 0.121 0.07 0.609 3.68 3.863 0 94.743
Tx-75d 73.236 0.011 12.678 0.468 0.065 0.064 0.386 3.795 4.257 0.063 95.024
Tx-75e 70.929 0.028 15.923 0.259 0.068 0.066 3.225 5.723 0.34 0 96.56
Tx-75f 64.564 0.887 15.504 3.041 0.055 1.684 3.892 4.356 2.807 0.225 97.015
Tx-75g 73.286 0.074 11.684 0.982 0.046 0.04 0.371 3.646 4.95 0.024 95.101
Tx-75h 64.361 1.128 13.902 3.933 0.12 1.478 2.73 4.43 3.203 0.256 95.541
(continued on next page)
ARTICLE IN PRESS
(continued )
Tx-75i 65.51 1.152 13.848 4.007 0.078 1.43 2.775 3.867 3.122 0.295 96.084
Tx-75j 74.296 0.109 11.907 1.057 0.052 0.089 0.4 2.829 4.949 0 95.688
Tx-75k 73.487 0.002 12.059 0.785 0.029 0.041 0.424 3.355 4.68 0 94.862
Tx-75l 73.242 0.026 12.225 0.453 0.039 0.063 0.367 3.962 4.064 0.016 94.457
Tx-75m 73.641 0.002 12.749 0.662 0.062 0.08 0.389 3.351 4.371 0 95.308
Sample Tx-75, depth: 250 cm, laminated fine sand with volcanic ash, diatomites and ostracods, SiO2 ¼ 73–75% mainly with a few grains 64–65%.
Tx-123a 74.137 0 12.679 0.605 0.153 0.061 0.31 3.988 4.063 0 95.997
Tx-123b 74.546 0.127 12.159 0.527 0.114 0.212 0.828 3.64 4.288 0 96.442
Tx-123c 74.793 0.095 11.907 0.542 0.082 0.097 0.813 3.679 3.694 0 95.701
Tx-123d 72.164 0.067 11.525 1.108 0.016 0.061 0.37 3.378 4.733 0 93.422
Tx-123e 74.504 0.207 12.126 0.981 0.013 0.216 1.066 3.193 4.192 0.039 96.538
Tx-123f 74.119 0.025 13.089 0.64 0.094 0.043 0.367 4.394 4.015 0 96.788
Tx-123g 74.082 0 12.932 0.502 0.098 0.103 0.406 4.114 4.254 0.008 96.498
Tx-123h 64.703 1.079 15.352 4.718 0.048 1.894 4.36 4.92 2.418 0.385 99.877
Tx-123i 77.393 0 12.358 1.344 0.244 0.069 0.168 2.958 4.971 0.032 99.536
Tx-123j 72.908 0 12.101 0.536 0.081 0.046 0.337 2.553 6.001 0.016 94.579
Sample Tx-123, depth: 342–344 cm, laminated fine sand with volcanic ash, diatomites and ostracods, SiO2 ¼ 72–74% mainly.
Tx-131a 74.355 0.005 12.728 0.56 0.163 0.035 0.337 4.012 4.257 0.008 96.459
Tx-131b 73.572 0.037 12.496 0.662 0.17 0.06 0.417 3.956 4.275 0.024 95.669
Tx-131c 73.042 0.015 12.524 0.585 0.059 0.044 0.387 3.624 4.487 0.04 94.806
Tx-131d 73.95 0 12.647 0.598 0.202 0.01 0.338 3.966 4.349 0.032 96.091
Tx-131e 74.26 0.021 12.842 0.586 0.133 0.049 0.324 3.997 3.918 0 96.13
Tx-131f 72.473 0 13.226 0.511 0.114 0.031 0.44 3.898 3.979 0.079 94.751
Tx-131g 73.263 0.03 12.441 0.672 0.062 0.328 0.571 3.956 4.007 0.095 95.425
Tx-131h 72.643 0.037 12.613 0.55 0.143 0.046 0.417 3.714 4.501 0.055 94.72
Tx-131i 74.531 0 13.064 0.609 0.033 0.058 0.351 4.171 4.36 0 97.176
Tx-131j 68.293 1.009 12.893 3.105 0 0.337 2.055 4.125 3.01 0.196 95.022
Sample Tx-131, depth: 358–360 cm, silty clay, start of sequence of white laminated fine sands with diatoms and ostracods, SiO2 ¼ 73–74% mainly.
References implications for the late Holocene human occupation in the Upper Lerma Basin,
Central Mexico. Palaeogeography, Palaeoclimatology. Palaeoecology 186,
217–235.
Arce, J.L., Macias, J.L., Vazquez Selem, L., 2003. The 10.5 ka Plinian eruption of
Castiglia, P., Fawcett, P.J., 2006. Large Holocene lakes and climate change in the
Nevado de Toluca, Mexico, stratigraphical and hazard implications. Geological
Chihuahan desert. Geology 34, 113–116.
Society of America Bulletin 115 (2), 230–248.
Cerling, T.E., Ehleringer, J.R., Harris, J.M., 1988. Carbon dioxide starvation, the
Audino, M., Grice, K., Alexander, R., Kagi, R.I., 2001. Macrocyclic-alkanes: a new class
development of C-4 ecosystems, and mammalian evolution. Philosophical
of biomarker. Organic Geochemistry 32, 759–763.
Transactions of the Royal Society of London. Series B-Biological Sciences 353,
Battarbee, R.W., 1986. Diatom analysis. In: Berglund, B.E. (Ed.), Handbook of Holo-
159–170.
cene Palaeoecology and Palaeohydrology. Wiley, Chichester, pp. 527–570.
Davies, H.L., 1995. Quaternary palaeolimnology of a Mexican crater lake. Ph.D.
Bradbury, J.P., 1971. Paleolimnology of Lake Texcoco, México. Evidence from
Thesis, Kingston University, London, UK.
Diatoms. Limnology and Oceanography 16, 180–200.
Davies, S.J., Metcalfe, S.E., Caballero, M.E., Juggins, S., 2002. Developing diatom-
Bradbury, J.P., 1989. Late Quaternary lacustrine paleoenvironments in the Cuenca de
based transfer functions for Central Mexican lakes. Hydrobiologia 467, 199–213.
México. Quaternary Science Reviews 8, 75–100.
Degens, E.T., Gullard, R.R., Sackett, W.M., Hellebust, J.A., 1968. Metabolic fraction-
Bridgwater, N.D., Heaton, T.H.E., O’Hara, S.L., 1999. A late Holocene palae-
ation of carbon isotopes in marine plankton. 1. Temperature and respiration
olimnological record from central México, based on faunal and stable-isotope
experiments. Deep-Sea Research 15, 1–9.
analysis of ostracod shells. Journal of Paleolimnology 22, 383–397.
Deines, P., 1980. The isotopic composition of reduced organic carbon. In: Fritz, P.,
Brown, R.B., 1985. A summary of late-Quaternary pollen records from Mexico west
Fontes, J.C. (Eds.), Handbook of Environmental Isotope Geochemistry. The
of the Isthmus of Tehuantepec. In: Bryant, V.M., Holloway, R.G. (Eds.), Pollen
Terrestrial Environment, vol. 1. A. Elsevier, Amsterdam, pp. 329–406.
Records of Late-Quaternary North American Sediments. American Association
De Terra, H., 1947. Preliminary note on the discovery of fossil man at Tepexpan in
of Stratigraphic Palynologists Foundation, Dallas, pp. 71–94.
the Valley of Mexico. American Antiquity 13 (1), 40–44.
Brunner, C., 1982. Paleoceanography of surface waters in the Gulf of Mexico during
De Terra, H., Romero, J., Stewart, T.D., 1949. Early man in Mexico. In: Tepexpan Man.
the late Quaternary. Quaternary Research 17, 105–119.
Viking Foundation. Publications in Anthropology 11, 13–86.
Buckler, E.S.I.V., Pearsall, L.M., Holtsford, T.P., 1998. Climate, plant ecology, and
De Terra, H., 1951. Comments on radiocarbon dates from Mexico. In: Homenaje
central Mexican archaic subsistence. Current Anthropology 39, 152–164.
a Alfonso Caso. Socieded Mexicana de Antropologia, Mexico, pp. 377-388.
Caballero-Miranda, M.E., 1995. Late Quaternary palaeolimnology of Lake Chalco, The
Eggins, S.M., Grun, R., McCulloch, M.T., Pike, A.W.G., Chappell, J., Kinsley, L.,
Basin of Mexico: new evidence for palaeoenvironmental and palaeoclimatic
Mortimer, G., Shelley, M., Murray-Wallace, C.V., Spotl, C., Taylor, L., 2005. In situ
change in central Mexico during the last 45,000 years. Ph.D. thesis, University of
U-series dating by laser ablation multicollector ICPMS: new prospects for
Hull, Hull, U.K.
Quaternary geochronology. Quaternary Science Reviews 24 (23-24), 2523–2538.
Caballero Miranda, M.E., 1997. The last glacial maximum in the Basin of Mexico: the
Ficken, K.J., Li, B., Swain, D.L., Eglinton, G., 2000. An n-alkane proxy for the sedi-
diatom record between 34,000 and 15,000 years BP from Lake Chalco.
mentary input of submerged/floating freshwater aquatic macrophytes. Organic
Quaternary International 43/44, 125–136.
Geochemistry 31, 745–749.
Caballero Miranda, M.E., Ortega Guerrero, B.,1998. Lake levels since about 40,000 years
Gonzalez, S., Huddart, D., 2007. Paleoindians and megafaunal extinction in the Basin
ago at Lake Chalco, near México City. Quaternary Research 50, 69–79.
of Mexico: the role of the 10.5 K Upper Toluca Pumice eruption. In: Grattan, J.,
Caballero Miranda, M.E., Lozano Garcia, S., Ortega Guerrero, B., Urrutia
Torrence, R. (Eds.), Living Under the Shadow: The Cultural Impacts of Volcanic
Fucugauchi, J., Macias, J.L., 1999. Environmental characteristics of Lake Teco-
Eruptions. One World Archaeology Series 53. Left Coast Press, California, pp.
comulco, northern Basin of Mexico, for the last 50,000 years. Journal of Pale-
90–106.
olimnology 22, 399–411.
Gonzalez, S., Huddart, D., Morett-Alatorre, L., Arroyo-Cabrales, J., Polaco, O.J. 2001.
Caballero Miranda, M.E., Ortega Guerrero, B., Valadez, F., Metcalfe, S., Macias, J.L.,
Mammoths, Volcanism and Early humans in the Basin of Mexico during the late
Sugiura, Y., 2002. Sta. Cruz Atizapán: a 22-ka lake level record and climatic
ARTICLE IN PRESS
Pleistocene–Holocene. In: Cavarreta, G., Gioia, P., Mussi, M., Palombo, M.R. Meyers, P.A., 2003. Applications of organic geochemistry to paleolimnological
(Eds.), The World of Elephants. Proceedings of the 1st International Congress, reconstructions: a summary of examples from the Laurentian Great Lakes.
Roma, Italy, pp. 704–706. Organic Geochemistry 34, 261–289.
Gonzalez, S., Jiménez-López, C., Hedges, R., Huddart, D., Ohman, J., Turner, A., Mooser, F. 1967. Tefracronologı́a de la Cuenca de México para los últimos 30,000
Pompa, y, Padilla, J.A., 2003. Earliest humans in America: new evidence from años. Boletin del INAH, No. 30, Mexico, D.F, INAH, pp. 12-15.
Mexico. Journal of Human Evolution 44, 379–387. Mooser, F. 1975. Historia geologica de la Cuenca deMexico, In: Departamento del
Gonzalez, S., Huddart, D., Bennett, M., 2006a. Valsequillo Pleistocene archaeology Distrito Federal, Memoria de las obras del sistema de drenaje profundo del
and dating: ongoing controversy in Central Mexico. World Archaeology 38 (4), Distrito Federal, Mexico, No. 1, pp. 7–38.
611–627. Mooser, F., Montiel, A., Zúniga, A. 1996. Nuevo mapa geologico de la Cuenca de
Gonzalez, S., Huddart, D., Bennett, M.R., González-Huesca, A., 2006b. Human foot- Mexico, Toluca y Puebla, Mexico, Comision Federal de Electricidad, Mexico, D.F.
prints in Central Mexico older than 40,000 years. Quaternary Science Reviews Nichols, D.L., Spence, M.W., Borland, M.D., 1991. Watering the fields of Teotihuacan:
25, 201–222. early irrigation in the ancient city. Ancient Mesoameica 2, 119–219.
González Quintero, L., 1986. Análisis polı́nicos de los sedimentos. In: Lorenzo, J.L., O’Leary, M.H., Madhavan, S., Paneth, P., 1992. Physical and chemical basis of carbon
Mirambell, L. (Eds.), Tlapacoya: 35,000 Años de Historia del Lago de Chalco. isotope fractionation in plants. Plant. Cell and Environment 15, 1099–1104.
Colección Cientifica, Serie Prehistoria, Instituto Nacional De Antropologı́a e Ortega Guerrero, B., Thompson, R., Urrutia Fucugauchi, J., 2000. Magnetic properties
Historia, pp. 157–166. of Lake Chalco, central Mexico and their paleoenvironmental implications.
González Quintero, L., Fuentes Mata, M., 1980. El Holoceno de la porción central de Journal of Quaternary Science 15, 127–140.
la Cuenca de México. Memorias Instituto Nacional de Antropologı́a e Historia, III Pike, A.W.G., Hedges, R.E.M., Van Calsteren, P., 2002. U-series dating of bone using the
Coloquio sobre Paleobotánica y Palinologı́a (Mexico City, 1977) 86, 113–132. diffusion-adsorption model. Geochimica et Cosmochimica Acta 66, 4273–4286.
Heine, K., 1984. The classical Late Weichselian climatic fluctuations in Mexico. In: Renne, P., Feinbeck, J.M., Waters, M.R., Arroyo Cabrales, J., Ochoa-Castillo, P., Perez-
Mörner, N.A., Karlén, W. (Eds.), Climatic Changes on a Yearly to Millenial Basis. Campa, M., Knight, K.B., 2005. Age of Mexican Ash with alleged footprints.
Reidel, Boston, pp. 95–115. Nature 438, E7–E8.
Heine, K., 1988. Late Quaternary glacial chronology of the Mexican volcanoes. Die Rzedowski, J., Rzedowski, C.G., 1979. Flora fanerogámica del Valle de México.
Geowissenschaften 7, 197–205. C.E.C.S.A. México Vol. I, 403 pp.
Holtvoeth, T., Wagner, T., Horsfield, B., Schubert, C.J., Wand, U., 2001. Late-Quater- Rzedowski, J., Rzedowski, C.G., 1985. Flora fanerogámica del Valle de México.
nary supply of terrigenous organic matter to the Congo deep-sea fan (ODP site Escuela Nac. Cienc. Biol. IPN. Inst. Ecol Vol. II, 74 pp.
1075): implications for equatorial African paleoclimate. Geo-Marine Letters 21, Sanders, W.T., 1976. The natural environment of the Basin of Mexico. In: Wolf, E.R.
23–33. (Ed.), The Valley of Mexico, Studies in Pre-Hispanic Ecology and Society. School
Huddart, D., Gonzalez, S., 2006. A review of environmental change in the Basin of of American research University of New Mexico Press, Albuquerque, 337 pp.
Mexico (40,000–10,000 BP): implications for early humans. In: Jiménez Scalan, R.S., Smith, J.E., 1970. An improved measure of the odd-to-even predomi-
López, J.C., Gonzalez, S., Pompa y Padilla, J.A., Ortiz Pedraza, F. (Eds.), El hombre nance in the normal alkanes of sediment extracts and petroleum. Geochimica et
temprano en América y sus implicaciones en el poblamiento de la Cuenca de Cosmochimica Acta 34, 611–620.
México: Primer Simposio Internacional. Colección Cientı́fica, Serie Antropolo- Schleser, G.H., 1995. Parameters determining carbon isotope ratios in plants. In:
gı́a, 500. Instituto Nacional de Antropologia e Historia, Mexico, D.F., pp. 77–106. Frenzel, B., Stauffer, B., Weib, M. (Eds.), Problems of Stable Isotopes in Tree-
Laws, E.A., Popp, B.N., Bidigare, R.R., Kennicutt, M.C., Macko, S.A., 1995. Dependence Rings, Lake Sediments and Peat Bogs as Climate Evidence for the Holocene.
of phytoplankton carbon isotopic composition on growth rate and (CO2)aq: Fischer Verlag, Stuttgart, pp. 71–96.
theoretical considerations and experimental results. Geochimica et Cosmochi- Sears, P.B., 1951. Pollen profiles and cultural horizons in the Basin of Mexico. In:
mica Acta 59, 1131–1138. 29th International Congress of Americanistas, vol. 1. University of Chicago Press,
Lorenzo, J.L., 1989. Fechamiento de la Mujer de Tepexpan. Boletin del Instituto Chicago, pp. 57-61.
Nacional de Antropologia e Historias. Nueva Epoca 28, 2–15. Sears, P.B., 1952. Palynology in southern North America: Part 1. Archaeological
Lounejeva Baturina, E., Morales Puente, P., Cabadas Báez, H.V., Cienfuegos horizon in the Basin of Mexico. Geological Society of America Bulletin 63,
Alvarado, E., Vallejo Gomez, E., Solleiro Rebolledo, E., 2006. Late Pleistocene to 241–254.
Holocene environmental changes from d13C determinations in soils at Teoti- Sears, P.B., Clisby, K.H., 1955. Palynology in North America, 4: Pleistocene climate in
huacan, Mexico. Geofisica Internacional 45, 85–98. Mexico. Bulletin of the Geological Society of America 66, 521–530.
Lozano Garcia, M.S., Ortega-Guerrero, B., 1994. Palynological and magnetic Sedov, S., Solleiro Rebolledo, E., Gama Castro, J.E., Vallejo Gomez, E., Gonzalez
susceptibility records of Lake Chalco, central Mexico. Palaeogeography, Palae- Velazquez, A., 2001. Buried paleosols of Nevado de Toluca: an alterative record
oclimatology. Palaeoecology 109, 177–191. of Late Quaternary environmental change in Central Mexico. Journal of
Lozano Garcia, M.S., Ortega Guerrero, B., 1997. Late Quaternary environmental Quaternary Science 16, 375–389.
changes of the central part of the Basin of Mexico. Correlation between Texcoco Solleiro Rebolledo, E., Sedov, S., McClung de Tapia, E., Cabadas, H., Gama Castro, J.,
and Chalco basins. Review of Palaeobotany and Palynology 99, 77–93. Vallejo Gomez, E., 2006. Spatial variability of environmental change in the
Lozano Garcia, M.S., Vázquez Selem, L., 2005. A high elevation Holocene pollen Teotihuacan Valley during the late Quaternary: paleopedological inferences.
record from Iztaccihuatl volcano, central Mexico. The Holocene 15, 329–338. Quaternary International 156-157, 13–31.
Lozano Garcia, M.S., Ortega Guerrero, B., Caballero Miranda, M., Urrutia- Stafford, T.W., Hare, P.E., Currie, L., Jull, A.J.T., Donahue, D.J., 1991. Accelerator
Fucugauchi, J., 1993. Late Pleistocene and Holocene palaeoenvironments of radiocarbon dating at the molecular level. Journal of Archaeological Science 18,
Chalco Lake. Quaternary Research 40, 332–342. 35–72.
Lozano Garcia, M.S., Sosa Najera, S., Sugiura, Y., Caballero Miranda, M., 2005. Talbot, M.R., Johannessen, T., 1992. A high resolution palaeoclimatic record for the
23,000 yr of vegetation history of the Upper Lerma, a tropical high-altitude last 27 500 years in tropical West Africa from the carbon and nitrogen isotopic
basin in Central Mexico. Quaternary Research 64, 70–82. composition of lacustrine organic matter. Earth and Planetary Science Letters
Markgraf, V., 1993. Climatic history of Central and South America since 18,000 yr 110, 23–37.
BP: comparison of pollen records and model simulations. In: Wright, H.E.J., Talbot, M.R., Lærdal, T., 2000. The late Pleistocene–Holocene palaeolimnology of
Kutzbach, J.E., Webb, T.I., Ruddiman, W.F., Street-Perrott, F.A., Bartlein, P.J. (Eds.), Lake Victoria, East Africa, based upon elemental and isotopic analyses of sedi-
Glacial Climates since the Last Glacial Maximum. University of Minneapolis mentary organic matter. Journal of Paleolimnology 23, 141–164.
Press, Minneapolis, pp. 357–385. Talbot, M.R., Livingstone, D.A., 1989. Hydrogen index and carbon isotopes of
McClung de Tapia, E., 2000. Prehispanic agricultural systems in the Basin of Mexico. lacustrine organic matter as lake level indicators. Palaeogeography, Palae-
In: Lentz, D.L. (Ed.), Imperfect Balance: Landscape Transformations in the Pre- oclimatology. Palaeoecology 70, 121–137.
Columbian Americas. Columbia University Press, New York, pp. 121–146. Tyson, R.V., 1995. Sedimentary Organic Matter: Organic Facies and Palynofacies.
Metcalfe, S.E., 1997. Palaeolimnological records of climate change in Mexico-frus- Chapman and Hall, London.
trating past, promising future? Quaternary International 43/44, 111–116. Van Devender, T.R., Burgess, T.L., 1985. Late Pleistocene woodlands in the Bolson de
Metcalfe, S.E., 2006. Late Quaternary environments of the Northern Deserts and Mapimi: a refugium for the Chihuahuan desert? Quaternary Research 24, 346–353.
Central Transvolcanic Belt of Mexico. Annals of the Missouri Botanical Garden Vázquez Selem, L., Heine, K., 2004. Late Quaternary glaciation in Mexico. In: Ehlers, J.,
93, 258–273. Gibbard, P.L. (Eds.), Quaternary Glaciations- Extent and Chronology, Part 111, South
Metcalfe, S.E., Street-Perrott, F.A., Perrott, R.A., Harkness, D.D., 1991. Palae- America. Asia, Africa, Australia, Antarctica. Elsevier, Amsterdam, pp. 233–242.
olimnology of the Upper Lerma Basin, Central México. Journal of Paleolimnol- Vázquez Sanchez, E., Jaimes Palomera, R., 1989. Geologia de la Cuenca de Mexico.
ogy 5, 197–218. Geofisica International 28, 133–190.
Metcalfe, S.E., O’Hara, S.L., Caballero, M., Davies, S.J., 2000. Records of Late Pleis- Waters, M.R., 1985. Early man in the New World: an evaluation of the radiocarbon
tocene–Holocene climatic change in México. Quaternary Science Reviews 19, dated Pre-Clovis Sites in the Americas. In: Mead, J.I., Melzer (Eds.), Environ-
699–721. ments and Extinctions: Man in the Late Glacial North America. Center for the
Metcalfe, S.E., Say, A., Black, S., McCulloch, R., O’Hara, S., 2002. Wet conditions Study of Early Man. University of Maine at Orono, pp. 125–143.
during the last glaciation in the Chihuahuan desert, Alta Babicora Basin, Mexico. Watts, W.A., Bradbury, J.P., 1982. Paleoecological studies at Lake Pátzcuaro on the
Quaternary Research 57, 91–101. west-central Mexican plateau and at Chalco in the Basin of Mexico. Quaternary
Meyers, P.A., 1994. Preservation of elemental and isotopic source identification of Research 17, 56–70.
sedimentary organic matter. Chemical Geology 114, 289–302. Wiesenberg, G.L.B., Schwarzbauer, J., Schmidt, M.W.I., Schwark, L., 2004. Source and
Meyers, P.A., 1997. Organic geochemical proxies of paleoceanographic, paleo- turnover of organic matter in agricultural soils derived from n-alkane/ n-
limnologic, and paleoclimatic processes. Organic Geochemistry 27, carboxylic acid compositions and C-isotope signatures. Organic Geochemistry
2163–2250. 35, 1371–1393.

Lamb QSR 2009 Ip

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lamb QSR 2009 Ip

Uploaded by

Copyright:

Available Formats

ARTICLE IN PRESS

Quaternary Science Reviews xxx (2009) 1–17

Contents lists available at ScienceDirect

Quaternary Science Reviews

Tepexpan Palaeoindian site, Basin of Mexico: multi-proxy evidence for

2 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 3

4 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

the Holocene was cool, indicated by glacial advances (10,000–

4. Methods, results and interpretations

The new sedimentary succession reported here is situated close

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 5

6 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 7

using a Fisons 8000 GC directly coupled to a Fisons 800 MD single- Table 2

8 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 9

10 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

there is a mixed tephra population including basaltic, basaltic–

which have been dated by AMS to 10,016 39 BP (OxA-15,840).

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 11

Table 3 Pleistocene–early Holocene related to volcanic events and subse-

12 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 13

14 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

Geochemical analysis from tephra shards from selected samples at Tepexpan.

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 15

16 A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17

A.L. Lamb et al. / Quaternary Science Reviews xxx (2009) 1–17 17

You might also like