Reconstruction of late Pleistocene climate in the Valsequillo Basin (Central Mexico) through isotopic analysis of terrestrial and freshwater snails more

Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 Contents lists available at SciVerse ScienceDirect Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo Reconstruction of late Pleistocene climate in the Valsequillo Basin (Central Mexico) through isotopic analysis of terrestrial and freshwater snails Rhiannon E. Stevens a,⁎, Sarah E. Metcalfe b, Melanie J. Leng c, d, Angela L. Lamb c, Hilary J. Sloane c, Edna Naranjo e, Silvia González f a McDonald Institute for Archaeological Research, University of Cambridge, Downing Street, Cambridge, CB2 3ER, UK School of Geography, University of Nottingham, University Park, Nottingham, NG7 2RD, UK NERC Isotope Geosciences Laboratory, British Geological Survey, Keyworth, Nottingham, NG12 5GG, UK d Department of Geology, University of Leicester, Leicester LE1 7RH, UK e Instituto de Biología Departamento de Zoología, Universidad Nacional Autonoma de Mexico, Apartado Postal 70-153, Mexico D.F. 04510, Mexico f School of Natural Sciences and Psychology, Liverpool John Moores University, Byrom Street, Liverpool, L3 3AF, UK b c a r t i c l e i n f o a b s t r a c t We aim to reconstruct the climatic and environmental conditions in the Valsequillo Basin during the deposition of the Valsequillo gravels between c. 40,000 and 8000 years ago, when large mega-fauna and potentially humans occupied the basin. Fossil freshwater (Fossaria sp. and Sphaeriidae (Family)) and terrestrial (Polygyra couloni, Holospira sp. and Cerionidae (Family)) snail shells from sections within the Barranca Caulapan were collected for oxygen and carbon stable isotope analysis. Oxygen and carbon isotopes in terrestrial and freshwater snail shells relate to local climatic parameters and environmental conditions prevailing during the lifetime of the snail. Whole shell isotope analysis showed that c. 35,000 years ago climate in the Valsequillo Basin was similar to the present day. Between c. 35,000 and 20,000 BP conditions became increasingly dry, after which conditions became wetter again, although this record is truncated. Intra-shell isotopic analyses show that the amount of precipitation varied seasonally during the late Pleistocene. If people did reach this part of the Americas in the late Pleistocene they would have experienced changing long-term and seasonal climatic conditions and would have had to adapt their life strategies accordingly. © 2012 Elsevier B.V. All rights reserved. Article history: Received 19 April 2011 Received in revised form 24 November 2011 Accepted 12 December 2011 Available online 9 January 2012 Keywords: Mollusc Shell Palaeoclimate Pleistocene Seasonality Barranca Caulapan Human occupation Isotope 1. Introduction 1.1. Background to study Late Pleistocene deposits are present within the Valsequillo basin, located in the volcanic highlands of central Mexico (N 18° 56, W 98° 07), south of Puebla (Fig. 1A and B). Palaeoclimate records from Mexico for the last glacial period and early Holocene are rather scarce and sometimes contradictory, thus the presence of molluscs within the deposits within the Valsequillo Basin provide an opportunity to obtain much needed additional data. Since the 1960s archaeologists have suggested that archaeological artefacts found in the basin provide evidence for very early human colonization of the Americas. Here we briefly describe the evidence for human occupation in the basin and present new isotope evidence for climatic conditions in the Valsequillo Basin in the Late Pleistocene. The first geological map of the Valsequillo Basin was produced by H. Malde in 1968. He showed that the underlying bedrock, the Balsas group, is a coarse Cretaceous limestone conglomerate cemented ⁎ Corresponding author. E-mail address: res57@cam.ac.uk (R.E. Stevens). 0031-0182/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2011.12.012 together by a matrix of red mudstone. The Pleistocene Basin deposits were thought to be not more than 70 m thick and to consist of four major units: The Amomoloc lake beds (lower), the Xalnene tuff/ash, the Atoyatenco lake beds (upper), and the Batan Lahar. Malde described the Amomoloc lake beds as rich in limestone, indicative of a closed basin, and the Atoyatenco lake beds as noncalcareous, indicative of external drainage. The intervening Xalnene ash/tuff was thought to represent a very explosive subaqueous basaltic eruption. In the late Pleistocene the Rio Atoyac and its tributaries incised the basin deposits, and this valley system was filled by alluvium known as the Valsequillo gravels, which are presently up to 30 m thick. The lower part of the gravels include many channel deposits of chert and limestone gravels with pebbles up to 0.5 m long interbedded with lenticular finer-grained alluvium and in the upper part fairly regular beds of sand, silt and clay several centimetres thick. Thin layers of volcanic ash, rhyolitic and basaltic, together with pumice lapilli are also present within the gravels sequence (Malde, 1968). From the Valsequillo gravels archaeologists have recovered both humanly modified bone (butchered and engraved) and stone artefacts in association with megafaunal remains (Irwin-Williams, 1962). Since the 1940s a manmade reservoir has filled part of the basin. Meteorological data from Puebla City indicate the present day climate is marked R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 17 Fig. 1. A. Location of the Valsequillo Basin. B. Location of sections studied in the Valsequillo Basin, Central Mexico. by highly seasonal precipitation. The region typically receives around 800 mm a year of precipitation which mostly falls between May and October (summer) in response to the northward migration of the ITCZ and the onset of the North American Monsoon. Almost no rain falls between December and February. Annual mean temperatures are around 16 °C, with winter mean temperature around 13 °C and summer mean temperature around 18 °C. In January, the coolest month, temperatures typically range from 8 °C to 17 °C and in April, the warmest month, temperatures typically range from 13 °C to 18 °C. In winter months there can be occasional ground frost (http://www.weatherbase.com). Fieldwork undertaken within the basin through the 1960s and in 1978, mostly focused on the archaeological site at Hueyatlaco (Fig. 1B), which Irwin-Williams interpreted as being a ‘kill site’ (Irwin-Williams, 1978). Dating of the Valsequillo gravels at Hueyatlaco and the artefacts within was problematic (see Szabo et al., 1969; VanLandingham, 2004, and González et al., 2006a,b). Subsequent work at Hueyatlaco suggests that the sediments are unlikely to be in situ and that the early uranium series and fission track dates are erroneous (González et al., 2006a,b). González et al. (2006a) confirmed Szabo et al.'s (1969) results that the Valsequillo gravels in the Barranca Caulapan, a small river gully found on the northeast side of the Valsequillo basin (Fig. 1B) are in situ and date from 38,900 to 9150 14C BP (uncalibrated). One shell radiocarbon dated to c. 20,000 14C BP and was found in close proximity to a lithic scraper, suggesting that humans may have been present in the Valsequillo Basin at this time (Szabo et al., 1969). The partial adult human skeleton found in Texcal Cave in the Valsequillo basin provides incontrovertible evidence for the presence of humans by the early Holocene with a radiocarbon date of 7480 ± 55 14C BP for the skeletal remains (González et al., 2003). Markings within the Xalnene ash from Cerro Toluquilla (Fig. 1B), initially interpreted as human footprints relating to a pre-40,000 BP human migration phase, have been re-interpreted as quarry marks (Renne et al., 2005; González et al., 2006a; Feinberg et al., 2009; Mark et al., 2010; Morse et al., 2010). Although uncertainties remain concerning when humans were first present in the Valsequillo basin and the overall chronology of the basin, the age of the Valsequillo gravels themselves is well constrained (as described above). For this reason, along with the extensive presence of megafaunal remains within this stratigraphic unit and the possible presence of humans within the basin at the time this unit was deposited, we focus our palaeoenvironmental investigations on these gravels. This paper aims to reconstruct the climatic and environmental conditions in the Valsequillo Basin during the deposition of the Valsequillo gravels, when large mega-fauna and potentially humans occupied the basin. Within the basin we have studied two sections through the Valsequillo gravels that are visible in exposed sections within the Barranca Caulapan (Fig. 1B). Today the Barranca contains an ephemeral stream, fed by a small number of springs, which cuts through the Valsequillo gravels. The Quaternary gravels are observed to overlie the Tertiary Balsas conglomerate, with an unconformity between the two. Samples were collected from two exposed sections one located upstream (Upper Barranca Caulapan, Log A, Figs. 1B and 2) (N 18° 56′ 53.8, W 98° 07′ 48.7, elevation 2091 m), and one downstream (Lower Barranca Caulapan, Log B, Figs. 1B and 3) (N 18° 56′ 36.1, W 98° 07′ 32.2, elevation 2064 m at base of sequence) of the road which intersects the Barranca. The sedimentology and stratigraphy for these two sections are shown in Figs. 2 and 3. Fossil shells of both freshwater and terrestrial molluscs were collected directly from the exposed section face for radiocarbon dating and stable isotope analysis. Modern shells of terrestrial and freshwater molluscs were also collected from leaf litter and rock pools to compare directly with the isotopic composition of the fossil shells and aid with the environmental reconstruction. 1.2. Ecology of the molluscs 1.2.1. Terrestrial snails Modern and fossil land snail shells collected from the Barranca Caulapan include a Cerionidae, Holospira sp., Polygyra couloni (Shuttleworth, 1852), Hawaiia minuscula (Binney, 1840) and Rotadiscus hermanni hermanni (Pfeiffer, 1866). Of these, an unknown genus of the family Cerionidae and a Holospira sp. were first found in the Barranca Caulapan as fossil specimens during this research. They will be described and named by Naranjo elsewhere. The three terrestrial mollusc types which were analysed for carbon and oxygen isotopic composition were modern and fossil Cerionidae and Holospira sp. and fossil P. couloni. Not all specimens of the Cerionidae and the Holospira were identified to species level, therefore isotopic data for the Cerionidae and Holospira sp. are considered collectively and hereafter are referred to as Cerionidae/Holospira sp. Most of the known species of the family Cerionidae live on exposed shores in the Florida Keys, the Bahamas, the Greater Antilles (except Jamaica) to the Cayman Islands and Curaçao (Pilsbry, 1946). The American mainland distribution of the family in the past was possibly very extensive, since fossil material from the Uppermost Cretaceous Hell Creek Formation, Montana and early Miocene Tampa Limestone of Florida (Roth and Hartman, 1998) had been assigned to Cerionidae. The presence of this cerionid in central Mexico extends its distribution south, along with new species of extant cerionids from Queretaro state, recently described by Fred G. Thompson (manuscript in revision). Living cerionids can be found in isolated patches of pine 18 R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 Fig. 2. Sedimentary Log A of the Upper Barranca Caulapan sequence. Whole shell carbonate δ18O and whole shell carbonate δ13C. Fig. 3. Sedimentary Log B of the Lower Barranca Caulapan section. Whole shell carbonate δ18O and whole shell carbonate δ13C. Range in modern Cerionidae/Holospira sp. and Fossaria sp. δ13C is not shown as values are off the scale, see Fig. 4. R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 19 or oak forest near the Valsequillo dam (F.G. Thompson, personal communication). Roth and Hartman (1998) noticed that various Western North America Late Cretaceous and Paleogene terrestrial molluscs now live in tropical lands (Roth and Hartman, 1998). Living specimens of the family Cerionidae have not been found around the Barranca Caulapan. The Holospira genus belongs to the family Urocoptidae, which is considered by Pilsbry (1946) as strictly calcicolous. The distribution of Holospira is from southern Arizona, central Texas to south Mexico. They are very abundant in central Mexico (Mexican Plateau) at altitudes of 1200 to 2000 m, and are always associated with limestone and usually with very hot and dry habitats (Bequaert and Miller, 1973). A few species of Holospira however live in more humid environments in association with tropical deciduous forests (Thompson, 1971). The distributions of most species are very local (Pilsbry, 1946). Holospira are usually gregarious; 44 species have been found within the southern states of Mexico (Bequaert and Miller, 1973). Modern Holospira sp. were found to be living in the Barranca Caulapan. The modern specimens were discovered adhering to the side of the tributary walls, typically over 30 cm above ground level, and were often found in association with decomposing leaf litter on which most terrestrial snails feed. The Polygyra genus belongs to the family Polygyridae, whose current distribution is from Alaska and eastern Canada, south to tropical America (Burch, 1962). Polygyra are found in the southeastern United States, through Central America from Mexico to Belize and Guatemala (Thompson, 2008). Snails in the Polygyridae family are of medium to large size (approx. 5 to 40 mm), their lip is reflected and some species have teeth in the aperture. The Polygyra genus is distributed in the southern United States (close to the 100th meridian), Mexico — where it is less abundant over the plateau, Cuba, the Bahamas and Bermuda (Pilsbry, 1940). Polygyra often inhabit forests and can be commonly found under rocks, humus and tree trunks, and primarily consume fungi. As with the majority of terrestrial molluscs, Polygyra are active during humid nights (Cheatum and Fullington, 1971). Some species prefer open spaces close to thick shrubs and grass (Pilsbry, 1940). Polygyra couloni (Shuttleworth, 1852) is known from the states of Guerrero, Jalisco, Distrito Federal, Morelos, Veracruz and now Puebla. A few modern P. couloni were found living in open grassland in the Barranca Caulapan. Isotopic analysis was only carried out on fossil P. couloni specimens. The life span of terrestrial molluscs varies; small species typically live one or two years, and larger species (greater than 2 cm) can live 4 years or more in captivity (Miller and Naranjo, personal observation). Land snails have a preference for moist habitats with high relative humidity (Van der Schalie and Getz, 1961, 1963) as this reduces water loss due to respiration, evaporation, and slime secretions (Howes and Wells, 1934; Prior, 1985; Barnhart, 1986; Balakrishnan and Yapp, 2004). Most land snails feed at night and rest during the day (Edelstam and Palmer, 1950; Newell, 1966; Gelperina, 1974; Cook, 1979; Balakrishnan and Yapp, 2004). During periods when environmental conditions are particularly dry or when temperatures exceed 27 °C, land snails aestivate (Cowie, 1984; Thompson and Cheny, 1996; Balakrishnan and Yapp, 2004), and shell growth ceases (Cowie, 1984). Shell growth occurs during the activity phase, when conditions are moist (Balakrishnan and Yapp, 2004). 1.2.2. Freshwater snails Fossil freshwater molluscs found in the Barranca Caulapan include Fossaria cf. obrussa (Say, 1825), Fossaria cf. cockerelli (Pilsbry and Ferriss, 1906), Drepanotrema lucidum (Pfeiffer, 1839), Sphaeriidae (Family), and Valvata humeralis (Say, 1829). The aquatic pulmonates have a terrestrial origin and have retained a high tolerance of prolonged aerial exposure and are adapted to shallow water, eutrophic microenvironments where exposure to air, desiccation, and hypoxic conditions can be frequent (McMahon, 1983). Thus unlike most freshwater species Fossaria cf. obrussa and F. cf. cockerelli are able to survive in intermittent streams and ponds by settling into the sediment on the bottom and aestivating in otherwise dry conditions (Natureserve, 2008). Valvata humeralis live in permanent water which is quiet, or with a slight current, either in the shallow margins of a lake or in protected situations at the edge of a stream (Taylor, 1957). Drepanotrema lucidum live in streams, shallow pools, ponds and marshes (Pointier and Augustin, 1999). Sphaeriidae (Family) live in freshwater lakes, streams, rivers, ponds, and ephemeral habitats. Modern Fossaria cf. obrussa, F. cf. cockerelli and D. lucidum were found to be living in rock pools in the Barranca Caulapan. Although the rate of shell growth can increase or decrease, shell growth is generally continuous throughout life. The freshwater molluscs which were analysed for carbon and oxygen isotope composition were modern and fossil Fossaria cf. obrussa and F. cf. cockerelli and fossil Sphaeriidae (Family). Not all specimens of Fossaria analysed for isotopes were identified to species level, therefore isotopic data for Fossaria cf. obrussa and Fossaria cf. cockerelli are considered collectively and hereafter are referred to as Fossaria spp. 1.3. Carbon and oxygen isotopes in terrestrial and freshwater snail shells Carbon and oxygen isotope compositions of terrestrial and freshwater snail shells have been used to reconstruct past environmental conditions (Abell, 1985; Goodfriend and Magaritz, 1987; Goodfriend and Ellis, 2000; Balakrishnan and Yapp, 2004). The oxygen isotope (δ 18O) composition of terrestrial snail shell carbonate is related to local climatic parameters and represents the combined effect of temperature, relative humidity, and the oxygen isotope signature of ambient meteoric water (Goodfriend et al., 1989). The carbon isotope (δ 13C) composition of terrestrial snail shell carbonate is primarily related to the carbon isotope signatures of the diet, and thus that of the local vegetation (e.g. C3, C4, and CAM plants) (Goodfriend and Magaritz, 1987; Stott, 2002; Metref et al., 2003; Balakrishnan and Yapp, 2004; Balakrishnan et al., 2005), however ontogenic factors can affect the snail shell δ 13C (Goodfriend et al., 1989; Leng et al., 1998). In freshwater snails the oxygen and carbon isotope values of shell carbonate reflect the temperature and isotopic composition of water in which they lived (Fritz and Poplawski, 1974; Leng et al., 1999). For both terrestrial and freshwater snails, average shell isotope signatures can be obtained from homogenized whole shells, whereas seasonal variation in isotope signatures can be investigated through analysing sub-samples from incremental growth layers and have been used to reconstruct palaeoenvironmental conditions (cf. Leng et al., 1998, 1999). Snail shells can be composed of aragonite or calcite, an aragonitic shell is often used as a means to confirm original preservation of the shells composition. Conversion to calcite (since aragonite is metastable) will effectively ‘reset’ the isotope signal (Leng and Marshall, 2004) so here we used X-ray diffraction to establish the carbonate type. 2. Materials and methods 2.1. Carbon and oxygen isotope analysis of shell carbonate Fossil shells were collected by hand picking directly from the exposed stratigraphic sections. Bulk sediment samples were initially sieved for shells, but since the shells are extremely fragile this caused the shells to fragment. The vertical height of each shell relative to the Tertiary Balsas unit in the Upper Barranca Caulapan (Log A) and the top of the massive sands layer in the Lower Barranca Caulapan (Log B) was measured prior to collection and each individual shell was numbered. In the laboratory, shells were bleached overnight in 5% sodium hypochlorite solution in order to remove clays and any organic residue. Those samples selected for whole shell analysis were then 20 R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 crushed in agate and the powder collected for isotope analysis and some also underwent XRD analysis (see Section 2.2). For intra-shell isotope analysis, up to 42 sequential holes were drilled in the shells from aperture to apex, the powder was collected for isotope analysis alone since there was insufficient for XRD. This form of analysis was only carried out on the Upper Barranca Caulapan section. Larger samples (>5 mg, mainly the whole shell powders) were analysed using a classical vacuum method, whereas smaller samples (b5 mg, sequential shell analysis) were analysed via an automated method. Samples analysed using the classical method were reacted with anhydrous phosphoric acid in vacuo overnight at a constant 25 °C. The CO2 liberated was separated from water vapour under vacuum and collected for analysis. Measurements were made on a VG Optima mass spectrometer. The smaller samples were analysed directly using an automated common acid bath VG Isocarb + Optima mass spectrometer. Isotope values (δ 13C, δ 18O) are reported as per mil (‰) deviations of the isotopic ratios ( 13C/ 12C, 18O/ 16O) calculated to the V-PDB scale using a within-run laboratory standard calibrated against NBS standards. Analytical reproducibility is typically b0.1‰ for both δ 13C and δ 18O (based on similarly sized laboratory standards). 2.2. X-ray diffraction X-ray diffraction (XRD) was carried out on a selection of whole shells in order to determine the mineralogy, and importantly to check for conversion of aragonite to calcite. Aliquots of shell powder were reground under acetone for 5 min in a tema-mill and dried at 55 °C before being back-loaded into standard aluminium sample holders. XRD analysis was carried out using a Phillips PW1700 series diffractometer and Co-Kα radiation operating at 45 kV and 40 mA. Xray diffraction results from a selection of the fossil terrestrial shells show that the aragonite in the shells was pristine and had not converted to calcite. 2.3. Radiocarbon Three terrestrial mollusc shells (Cerionidae/Holospira sp.) were selected from the Upper Barranca Caulapan, Log A (Fig. 2) for accelerator mass spectrometry radiocarbon dating. The shell VS 17 was first stained with Feigl's solution to check the mineralogy and hence test for any alteration. This method was used as the staining can be removed from the shell prior to radiocarbon dating. 3. Results and interpretation 3.1. Assessing the relationship between modern shell δ 18O and climate 3.1.1. Terrestrial molluscs Whole shell δ 18O analysis was carried out on modern Cerionidae/ Holospira sp. (n = 14) and results ranged from approximately −6.9‰ to −2.6‰ (V-PDB) (Fig. 4). In order to understand the local relationship between terrestrial shell δ 18O and climate in the Valsequillo Basin it was necessary to evaluate which parameters were influencing the δ 18O signatures of the local precipitation and local water. As the isotopic composition of precipitation is not routinely measured in the Valsequillo region, we used the Online Isotope in Precipitation Calculator (OIPC Version 2.2: http://www.waterisotopes.org/) to obtain a model-generated weighted-annual mean and monthly mean δ 18O for the Valsequillo precipitation. The online isotope in precipitation calculator models data from the Global Network for Isotopes in Precipitation database (IAEA/WMO, 2006) and uses an algorithm that incorporates the main parameters controlling the isotopic signature of precipitation such as Rayleigh distillation, latitude, altitude, and vapour-transport pathways (Bowen and Wilkinson, 2002; Bowen and Revenaugh, 2003; Baldini et al., 2007). No clear relationship exists between temperature measured at Puebla City (the closest meteorological station to Valsequillo) between 2001 and 2007 and the model-generated mean monthly precipitation δ 18O using the OIPC. By contrast a correlation (R 2 = 0.3993) exists between modelgenerated mean monthly precipitation δ 18O and the mean monthly amount of precipitation measured at Puebla City over the same time period (Fig. 5). An inverse relationship between precipitation δ 18O and the amount of precipitation (the amount effect) has been observed in regions which are subject to significant precipitation and/ or high humidity (Dansgaard, 1964; Yapp, 1982; Rozanski et al., 1993). This method has previously been used to calculate local precipitation δ 18O for comparison with land snail shell oxygen isotope signatures (Baldini et al., 2007). The weighted-annual mean δ 18O for the Valsequillo precipitation generated by the model was −7.9‰ (V-SMOW) ± 0.9‰ (95% confidence interval). This calculated weighted-annual mean precipitation δ 18O is very similar to the isotopic composition of water samples collected in January 2005 from a spring feeding the Barranca Caulapan stream, from the Barranca Caulapan stream and from the Valsequillo reservoir, which ranged from −8.0‰ to −7.7‰ (V-SMOW). The Fig. 4. δ18O and δ13C of modern and fossil terrestrial shells from the Barranca Caulapan. R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 21 Fig. 5. Model generated mean monthly precipitation δ18O versus measured mean monthly precipitation in mm. Fig. 6. Calculated seasonal variation in shell aragonite δ18O (calculated using Eqs. (1) and (2), the OIPC generated mean monthly precipitation δ18O for the Valsequillo basin, and seasonal variation in temperature measured at the Puebla meteorological station). isotopic composition of shell carbonate that would typically form from the precipitation/water in the Valsequillo region, at a given temperature, can be calculated using a palaeotemperature equation such as that of Kim and O'Neil (1997), here presented in the re-expression by Leng and Marshall (2004): 2 TBC ¼ 13:8–4:58ðδc–δwÞ þ 0:08ðδc–δwÞ : high fossil shell δ 18O values indicate relatively dry conditions. The calculated annual pattern (Fig. 6) is consistent with the seasonal modern climate, with its strong summer wet season. 3.1.2. Freshwater molluscs The δ 18O of the modern Fossaria sp. (n = 17) ranged from approximately −9.0‰ to − 6.1‰ (V-PDB) (Fig. 4), with a mean δ 18O of −7‰ ± 0.7‰. Their δ 18O values are consistent with these freshwater specimens precipitating their shell aragonite in isotopic equilibrium with the stream water in which they lived. As the calculated precipitation δ 18O is very similar to the measured isotopic composition of the water samples (see above), the freshwater shells δ 18O values appear to be closely linked to the precipitation δ 18O and in turn to the amount of precipitation due to the amount effect. 3.2. Upper Barranca Caulapan (Log A) The 4.5 metre section of Valsequillo gravels in the Upper Barranca Caulapan (Log A) consists of several fine gravel and sandy units. The molluscs extracted for isotopic analysis from the Upper Barranca Caulapan section included Cerionidae/Holospira sp. and Polygyra couloni, all of which are terrestrial species. Three new radiocarbon dates on terrestrial shells from section A show that the sequence covers the time period approximately 35,980 ± 260 to 18,875 ± 75 uncalibrated 14 C yr BP (Fig. 2, Table 1). Sedimentation rates must have varied, and we do not know the age of the upper 65 cm of sediment. More recent sediments from the Upper Barranca Caulapan sequence are assumed to have been eroded away. 3.2.1. Whole shell oxygen isotopes Whole shell carbonate δ 18O analysis was carried out on fossil and modern Cerionidae/Holospira sp. (n = 53) and fossil Polygyra couloni (n = 17) from the Upper Barranca Caulapan and results are presented in Figs. 2 and 4. The δ 18O of the fossil and modern Cerionidae/Holospira sp. range from −6.9‰ to − 2.6‰ and from − 5.8‰ to −0.2‰ (V-PDB) respectively (Fig. 4) and their mean δ 18O values are significantly different (fossil: −3.0‰ 1σ = 1.4‰; modern: −4.7‰ 1σ = 1.1‰, p ≤ 0.001). These results show that under the current climatic regime substantial variation is observed in whole shell δ 18O values. Tightly clustered δ 18O values indicate suppressed seasonality, whereas variable δ 18O values indicate a highly seasonal climatic regime (Yanes et al., 2008) (see also Figs. 5 and 6 for the effects of modern seasonality). There is substantial overlap between the modern and fossil Cerionidae/Holospira sp. δ 18O values indicating that overall, climatic conditions during the late Pleistocene were similar to those of the current day (Fig. 4). The slightly higher mean, maximum and minimum fossil Cerionidae/Holospira sp. δ 18O values indicate that conditions were slightly drier in the past than today. ð1Þ where: T°C = temperature degrees Celsius δc = δ 18O of calcite (V-PDB) δw = δ 18O of water (V-SMOW) As this calculation only provides the composition of calcite precipitated from water at a certain temperature, conversion to aragonite (Grossman, 1982; Abell and Williams, 1989) can be done using: δa ¼ δc þ 0:6 where: δa = δ 18O of aragonite Using mean annual temperature data recorded between 2001 and 2007 at Puebla (+16.7 °C), the shell aragonite precipitated from average rainwater would have had a mean δ 18O value in the region of −7.8‰ (V-PDB) (−9.1‰ to − 7.3‰ = 95% confidence interval). The δ 18O values of the modern Cerionidae/Holospira sp. range from −6.9‰ to − 2.6‰ (V-PDB) (Fig. 4) with a mean of −4.7‰ ± 1.1‰. The offset between the calculated δ 18O of aragonite typically precipitated under the mean climatic conditions and δ 18O of the modern Cerionidae/Holospira sp. shell aragonite may be because not all of the oxygen taken in by land snails comes from rain water (i.e. some may come from evaporated water in pools or in the soil) and water may only be ingested at certain times of the year. Terrestrial snails also ingest oxygen through eating plants, which often have higher δ 18O than precipitation due to evapotranspiration, and soil and rock carbonate which tend to have higher δ 18O values. Using the OIPC model-generated mean monthly precipitation δ 18O for the Valsequillo region, and the Puebla temperature data, we calculated (using Eqs. (1) and (2)) the expected oxygen isotope composition of shell aragonite precipitated each month by land snails living in the Valsequillo region. The calculated shell aragonite δ 18O values ranged from a minimum of −10.8‰ (V-PDB) to a maximum of − 5.4‰ (V-PDB) (Fig. 6). Although the relationship between model calculated monthly precipitation δ 18O and precipitation amount is not strong, it does suggest that there is a relationship between our shell δ 18O and climate; low shell δ 18O values indicate relatively wet conditions and ð2Þ 22 R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 Table 1 AMS radiocarbon results from Upper Barranca Caulapan (Log A). 14 C code Sample code VS50 VS45 VS17 Material Shell Shell Shell Species Cerionidae/Holospira sp. Cerionidae/Holospira sp. Cerionidae/Holospira sp. Distance above Balsas 3.85 m 2.52 m 0.77 m δ13C (‰) − 6.8 − 8.3 − 6.6 14 C date yr BP OxA-15970 OxA-15969 OxA-15997 18,875 ± 75 20,290 ± 140 35,980 ± 260 Although there is substantial variation in the fossil Cerionidae/ Holospira sp. δ 18O values broad trends can be seen through the sequence (Fig. 2). At the base of the section the range of fossil Cerionidae/Holospira sp. δ 18O values overlap with, but are slightly higher than, those of the modern specimens, suggesting conditions were a little drier than today. Although highly variable, δ 18O values overall become gradually higher from the base of the sequence to approximately 2 m above the Balsas Conglomerate, indicating that climatic conditions may have become drier. They then gradually become lower between approximately 2 and 2.7 m indicating wetter conditions. The few data points available above 2.7 m provide an indication of the direction of the δ 18O trend in the upper part of the sequence. The trend in fossil Polygyra couloni δ 18O values through the upper Barranca Caulapan series is not as clear as those of the fossil Cerionidae/Holospira sp. Overall, however, the fossil P. couloni δ 18O values broadly appear to increase and decrease concurrently with those of the fossil Cerionidae/Holospira sp. but an offset is visible between δ 18O signatures of the two. Such an offset has been previously observed and is likely to be due to the different ecotopes (e.g., aestivation time, length of growth season, and ingested food) of the two species (Zanchetta et al., 2005). Unfortunately no isotopic data for modern Polygyra were available to compare with Cerionidae/Holospira sp. and test the nature of this relationship today. 3.2.2. Whole shell carbon isotopes δ 13C data is obtained with the δ 18O data (Section 3.2.1) and results range from approximately −9.0‰ to −5.1‰ and from − 7.7‰ to −5.5‰ from the fossil and modern Cerionidae/Holospira sp. respectively (Fig. 4). These results show that under current environmental conditions substantial variation is observed in modern Cerionidae/ Holospira sp. δ 13C values. There is some overlap between fossil and modern Cerionidae/Holospira sp. values (Fig. 4) although their mean δ 13C values (fossil: −7.7‰ 1σ = 0.8‰; modern: −6.4‰ 1σ = 0.6‰) are significantly different (p ≤ 0.001). An offset between modern and fossil shell δ 13C values is expected as the burning of fossil fuels has affected the carbon pool and has contributed a significant amount of isotopically light carbon ( 12C) to the atmosphere which has resulted in a 1‰ to 1.5‰ depletion from pre-industrial values (Friedli et al., 1986). The modern Cerionidae/Holospira sp. mean δ 13C is however 13C enriched relative to that of the fossil specimens. Shell aragonite δ 13C values primarily reflect the diet of the snail however shell δ 13C is typically around 14‰ heavier than diet, probably because respiratory gas exchange discards CO2 and retains the isoto− pically heavier HCO3 (McConnaughey and Gillikin, 2008). Shell δ 13C values therefore can be used as a proxy for the δ 13C signatures of the plants and organic matter consumed. The δ 13C values for C3 and C4 plants typically range from −22‰ to − 35‰ and − 9‰ to −16‰ respectively. Thus the shell carbonate δ 13C results suggest that during the formation of the Upper Barranca Caulapan sequence a temperate environment dominated by C3 plants prevailed within this area of the Valsequillo Basin. The fossil Polygyra couloni δ 13C values very closely track those of the fossil Cerionidae/Holospira sp. throughout the sequence (Fig. 2) with the exception of between 2 m and 2.5 m (c. 20,000 14C BP) where the fossil P. couloni δ 13C rise prior to the fossil Cerionidae/Holospira sp. δ 13C values. The increase δ 13C of the two species broadly corresponds to the increase in shell δ 18O. Together these signatures may indicate the amount of C4 plants slightly increased towards 20 ka BP. Expansion in C4 plants during the last Glacial Maximum in Central Mexico has been previously attributed to low CO2 partial pressure and increased aridity (Huang et al., 2001). 3.2.3. Intra shell oxygen isotopic analysis Intra-shell carbon and oxygen isotope analysis was conducted on one modern (Fig. 7A) and three fossil Cerionidae/Holospira sp. (Fig. 7B–D), and two fossil Polygyra couloni (Fig. 8A–B) shells. The measured intra-shell δ 18O of the modern Cerionidae/Holospira sp. range from −6.2‰ to − 0.6‰ (V-PDB). Thus the amplitude of the calculated oxygen isotope variation (5.4‰) and the measured intra-shell isotopic variation (5.6‰) are very similar and can be compared with our modelled values (Fig. 6). As for the whole shell oxygen isotope analysis, the offset between the calculated intra-shell δ 18O and the measured modern Cerionidae/Holospira sp. intra-shell δ 18O may be due to intake of oxygen through consumption of plants and carbonate and through exchange with the atmosphere. The modern Cerionidae/ Holospira sp. intra-shell δ 18O values are relatively low towards the apex of the shell, suggesting that the snail began growth during the wettest part of the year. Just over two sinusoidal cycles can be seen in the shell δ 18O values (Fig. 7A). We presume these sinusoidal cycles track the seasonal variation in precipitation δ 18O (see also Figs. 5 and 6) and indicate that the snail lived for just over two years. The three fossil Cerionidae/Holospira sp. specimens were sampled from near the bottom (specimen VS-10 at 0.85 m above the Balsas conglomerate) (Fig. 7B), middle (specimen VS-44 at 2.5 m above the Balsas conglomerate) (Fig. 7C), and top (specimen VS-51 at 4.16 m above the Balsas conglomerate) (Fig. 7D) of the Upper Barranca Caulapan sequence. The variation in amplitude of intra-shell δ 18O for the modern and fossil Cerionidae/Holospira sp. specimen is relatively small (2.3‰) considering the amount of δ 18O variation observed within North America shell populations. Potentially the slightly higher amplitude variation in intra-shell δ 18O seen in specimens VS-44 and VS-51 may indicate slightly greater seasonal climatic variations than those that occur today, however the data are limited and the potential difference small. As for the modern Cerionidae/Holospira sp. we presume the sinusoidal cycles observed in the fossil Cerionidae/Holospira sp. δ 18O track the seasonal variation in precipitation δ 18O. On this basis specimens VS-44 and VS-51 are calculated to have lived for 2– 3 years. For specimen VS-10 the duration of life is less clear as a complete sinusoidal pattern is not observed in its δ 18O signatures. Intra-shell isotopic analysis was not conducted on a modern Polygyra couloni specimen, but was undertaken on two fossil P. couloni specimens. The amplitude of the δ 18O variation measured in specimen VS-53 (found 3.59 m above the Balsas Conglomerate, Fig. 8A) was similar to that observed in the Cerionidae/Holospira sp. specimens, but was much greater in the P. couloni specimen VS-61 (found 1.55 m above the Balsas Conglomerate, Fig. 8B). Without intra-shell δ 18O results from a modern P. couloni it is not possible to determine whether the difference in the amount of intra-shell variation for this species is typical or atypical. Only one sinusoidal cycle was observed in the P. couloni δ 18O results indicating that both specimens probably lived for around a year. 3.2.4. Intra shell carbon isotopic analysis Very little variation was observed in the modern Cerionidae/ Holospira sp. intra-shell δ 13C values (Fig. 7A), suggesting the diet of the modern snail was relatively homogenous throughout life, and ontogenic factors do not appear to have influenced its δ 13C values. R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 23 Fig. 8. Intra-shell δ18O and δ13C isotope analysis. A. Fossil Polygyra couloni VS — 53, 1.55 m above the Balsas Conglomerate. B. Fossil Polygyra couloni VS — 61, 3.59 m above the Balsas Conglomerate. δ18O = closed symbols, δ13C = open symbols. the number of cycles does not correspond to the number of presumed annual cycles observed in the δ 18O values of this specimen. The absolute δ 13C values of all of the fossil Cerionidae/Holospira sp. specimens are relatively similar to that measured for the modern Cerionidae/ Holospira sp. Intra-shell isotopic analysis was not conducted on a modern Polygyra couloni, but the intra-shell δ 13C values for the fossil P. couloni VS-53 (Fig. 8A) lacked substantial variation and were similar to that observed in the Cerionidae/Holospira sp. specimens. Greater intra-shell δ 13C variation was observed for P. couloni specimen VS-61 (Fig. 8B). As previously discussed, substantial variation was observed in this specimen's intra-shell δ 18O values, potentially indicating greater seasonal variation during the life of specimen VS-61. Specimen VS-61 was found 3.65 m above the Balsas conglomerate. Although very limited bulk shell δ18O and δ13C results are available from this part of the sedimentary sequence, a sizeable shift is seen in the δ18O and δ13C values of both the fossil Cerionidae/Holospira sp. and P. couloni (Fig. 2). Thus the bulk and intra-shell isotope changes may be recording different aspects of the same shift in climatic conditions. Fig. 7. Intra-shell δ18O and δ13C isotope analysis. A. Modern Cerionidae/Holospira sp. B. Fossil Cerionidae/Holospira sp. VS — 10, 0.85 m above the Balsas Conglomerate. C. Fossil Cerionidae/Holospira sp. VS — 44, 2.5 m above the Balsas Conglomerate. D. Fossil Cerionidae/Holospira sp. VS — 51, 4.16 m above the Balsas Conglomerate. δ18O = closed symbols, δ13C = open symbols. 3.3. Lower Barranca Caulapan (Log B) The Lower Barranca Caulapan section (Figs. 1B and 3) consists of several interbedded gravel and sand layers and was in very close proximity to the section recorded and described in detail by González et al. (2006a). The Valsequillo gravels in the lower Barranca Caulapan have previously been dated by Szabo et al. (1969) and González et al. (2006a). The Upper and Lower Barranca Caulapan sections can be broadly correlated as the Lower Barranca Caulapan section represents in part an expanded version of the Upper Barranca Caulapan section. Shells were collected from a thick sandy layer overlain by a matrix supported gravel layer (Fig. 3). The boundary between the two was positioned approximately 10.5 m above the Balsas Conglomerate (Fig. 3). Although the shells were not taken from the same section recorded by González et al. (2006a), it has The intra-shell δ 13C values for the fossil Cerionidae/Holospira sp. VS10 also lacked substantial variation (Fig. 7B) and were similar in isotopic composition to that of the modern Cerionidae/Holospira sp. Almost the same can be said for specimen VS-44 (Fig. 7C), however a limited amount of δ 13C variation was observed towards the apex of the shell indicating that ontogenic factors may have influenced the δ 13C composition of this specimen during early life. Relatively limited variation (approximately 1.5‰) occurs in the δ 13C values of specimen VS-51 (Fig. 7D), however, the variation does appear to be cyclical, yet 24 R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 been possible (through comparison of Gonzalez et al.'s stratigraphic log and photographs taken in the field) to pinpoint the top of the sandy layer in both sequences and therefore link the chronology of the González et al. (2006a) sequence to the section investigated in this study. An electron spin resonance date on a Mammoth molar found approximately 1.25 m below the top of the sandy layer gave an age of 27,800 ± 3800 BP and a radiocarbon date on a mollusc found approximately 1.85 m below the top of the sandy layer gave a date of 27,880 ± 240 14C yr BP (González et al., 2006a; Fig. 3). Mollusc specimens found in the Lower Barranca Caulapan section include Fossaria cf. obrussa, Fossaria cf. cockerelli, Drepanotrema lucidum, Sphaeriidae (Family), Valvata humeralis, which are freshwater species, and a Cerionid, Cerionidae/Holospira sp., Polygyra couloni, Hawaiia minuscula, Rotadiscus hermanni hermanni, all of which are terrestrial species. As previously discussed (see Section 1.2) the majority of freshwater species found in the Lower Barranca Caulapan section generally live in small pools or streams that may only be present seasonally. The presence of these freshwater species in the same layer as the terrestrial species indicates that the Lower Barranca Caulapan sequence was deposited under mostly terrestrial conditions, although ephemeral pools or stream pools may have formed locally for part of the year. Whole shell isotopic analysis was only undertaken on the three most abundant of these fossil species: terrestrial Cerionidae/Holospira sp. (n = 15) and freshwater Fossaria sp. (n = 19) and Sphaeriidae (n = 7). Intra-shell isotopic analysis was not carried out on specimens from the Lower Barranca Caulapan sequence. slightly lower than the fossil Fossaria sp. (Fig. 3). The offset between the two types of molluscs could be due to the different ecotopes. 3.3.2. Whole shell carbon isotopes δ 13C analysis was carried out on the same fossil specimens from the Lower Barranca Caulapan sequence (Figs. 3 and 4). The range in δ 13C measured for fossil Cerionidae/Holospira sp. from this layer (−6.7‰ to −0.9‰ V-PDB) was substantially higher than that observed for the same fossil species from the Upper Barranca Caulapan sequence (−9.0‰ to −5.1‰) and than that observed for the modern specimens (−7.7‰ to − 5.5‰) (Fig. 4). Moreover the mean δ 13C of the fossil Cerionidae/Holospira sp. from the Lower Barranca Caulapan (−3.1‰, 1σ = 1.3‰) is statistically different (p ≤ 0.001) to that of fossil specimens from the Upper Barranca Caulapan (− 7.7‰, 1σ = 0.8‰) and that of the modern specimens (− 6.4‰, 1σ = 0.6‰). The latter two groups are also statistically different from each other (p ≤ 0.001). The disparity between the δ 13C values of fossil Cerionidae/Holospira sp. from two contemporary sequences only a few kilometres apart is thought to relate to the topography. The Lower Barranca Caulapan valley is much wider and open and therefore the snails are more likely to have consumed C4 vegetation such as grasses. The Upper Barranca Caulapan is an incised valley which currently has trees growing on its steep sides. The fossil Cerionidae/Holospira sp. from the Upper Barranca Caulapan are more likely to have consumed leaf litter which has lower δ 13C values as trees are C3 plants. The difference between the mean δ 13C of the fossil Cerionidae/Holospira sp. from the Lower Barranca Caulapan and that of the modern specimens is also likely to relate to the sampling location of the modern Cerionidae/Holospira sp. which were collected from leaf litter in the Upper Barranca Caulapan. However, as mentioned in Section 3.2.2, some but not all of the difference in δ 13C values between the fossil Cerionidae/Holospira sp. from the Lower Barranca Caulapan and the modern specimens, may be related to changes in the δ 13C of the atmosphere due to the burning of fossil fuels. The range in δ 13C measured for fossil Fossaria sp. from the Lower Barranca Caulapan (−5.4‰ to −0.6‰ V-PDB) was substantially higher than that observed for the modern specimens (−14.5‰ to −8.9‰) (Fig. 4). Moreover the mean δ 13C of the fossil Fossaria sp. (−3.1‰, 1σ = 1.3‰) was significantly different (p ≤ 0.001) to that of the modern specimens (−11.2‰, 1σ = 1.3‰). These results suggest that climatic conditions were drier and/or warmer during the formation of the lower Barranca Caulapan sequence than during the present day. Isotopic analysis was not conducted on modern Sphaeriidae, however the fossil Sphaeriidae δ 13C values are similar to those of the fossil Fossaria sp. (Fig. 3). As all the fossil Cerionidae/Holospira sp. and Fossaria sp. were extracted from essentially the same 30 centimetre sandy layer, we have assumed that this is an essentially synchronous unit and do not discuss any trends in the data (Fig. 3). 3.4. Summary of Barranca Caulapan palaeoclimate and comparisons with surrounding regions Whole shell δ 18O and δ 13C values indicate that before approximately 36,000 14C BP conditions in the Upper Barranca Caulapan were similar to today although perhaps a little drier, with C3 vegetation prevailing. An increase in the δ 18O and δ 13C values of terrestrial shells from the Upper Barranca Caulapan after that date suggests conditions became drier. The relatively high δ 18O and δ 13C values from terrestrial and freshwater shells in the Lower Barranca Caulapan dating to approximately 28,000 14C BP confirm this trend. Terrestrial shell δ 18O and δ 13C values from the Upper Barranca Caulapan indicate that some time after 28,000 14C BP, but before approximately 20,000 14 C BP, conditions became gradually cooler and/or wetter. Although limited data is available, terrestrial shell δ 18O and δ 13C values from the Upper Barranca Caulapan suggest conditions continued to 3.3.1. Whole shell oxygen isotopes The results of the whole shell isotopic analysis of specimens from the Lower Barranca Caulapan are presented in Figs. 3 and 4. The mean δ 18O (−2.6‰, 1σ = 1.4‰) measured for fossil Cerionidae/Holospira sp. from the Lower Barranca Caulapan was not statistically different (p = 1.00) from that of the same species from the Upper Barranca Caulapan (−2.6‰, 1σ = 1.4‰). The range in δ 18O for fossil Cerionidae/Holospira sp. from the Lower Barranca Caulapan (−5.5‰ to −0.8‰ V-PDB) was also similar to that observed for fossil Cerionidae/Holospira sp. from the Upper Barranca Caulapan (−5.8‰ to −0.2‰) (Fig. 4). The mean δ 18O of the fossil specimens from both the Upper and Lower Barranca Caulapan was statistically different (p ≤ 0.001 for both) to that of the modern Cerionidae/Holospira sp. (−4.7‰, 1σ = 1.1‰). In addition, the range in δ 18O for the modern Cerionidae/Holospira sp. (−6.9‰ to − 2.6‰) was slightly lower than that of the fossil specimen, although there is substantial overlap (Fig. 4). These results suggest that the climate conditions recorded in both the Lower and Upper Barranca Caulapan were not too dissimilar to present day climatic conditions, although are likely to have been a little drier. The δ 18O of the fossil Fossaria sp. from the Lower Barranca Caulapan range from − 6.8‰ to −3.7‰ (V-PDB) whereas the range in δ 18O of the modern Fossaria sp. is much lower, from −9.0‰ to − 6.1‰ (Fig. 4). Moreover, the mean δ 18O of the fossil specimens (−5.2‰ 1σ = 0.9‰) is statistically different (p ≤ 0.001) to the mean of the modern specimens (−7.0‰, 1σ = 0.7‰). This suggests that the fossil Fossaria sp. from the Lower Barranca Caulapan lived in water that had higher δ 18O values than that of the modern spring/stream/reservoir water. Precipitation δ 18O which is controlled by the amount affect and the source of moisture, is the primary control on the δ 18O of these water sources. The higher fossil shell δ 18O values could indicate either that the amount of precipitation around 28,000 years ago was lower than that of today, or that the source of the moisture was different to the present day. The latter of these two scenarios is unlikely as if the source of the moisture had switched from the Atlantic (Gulf of Mexico/Caribbean sea) to the North Pacific Ocean, the fossil shell δ 18O values should be lower. Isotopic analysis was not conducted on modern Sphaeriidae, however the fossil Sphaeriidae δ 18O values are R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 25 become cooler and/or wetter between approximately 20,000 and 18,0000 14C BP and post 18,0000 14C BP. As the majority of the palaeoclimate studies from central Mexico have reported radiocarbon dates in uncalibrated 14C years BP, we do so for ease of comparison. Unfortunately other palaeoclimate records from Mexico for the period covered by the Barranca Caulapan sections are rather scarce and sometimes contradictory. The largest number of records comes from the Basin of Mexico (west of Puebla, Fig. 1), but the hydrological complexity of the basin, the impacts of glacial meltwater (largely in the south), tephra deposition and apparently contradictory interpretations of different proxies have hindered an agreed palaeoclimatic reconstruction (Bradbury, 1989; Caballero et al., 1999; Metcalfe et al., 2000; Solleiro-Rebolledo et al., 2006). There is a broad consensus however that the early part of the last glacial was wetter than present (certainly until 30,000 BP and perhaps as late as 25,000 BP), although the overall trajectory was towards drier conditions (González-Quintero, 1986; Bradbury, 1989; LozanoGarcía and Ortega-Guerrero, 1998; Caballero et al., 1999). The interpretation of conditions around the LGM is far more problematic. Pollen extracted from a core from the centre of Lake Texcoco was interpreted as indicating that between c. 26,000 and 24,000 BP the climate was cold and dry, then became more moist between 24,000 and 21,000 BP (González-Quintero and Fuentes-Mata, 1980; Brown, 1985; Bradbury, 1989). Between 21,000 and 19,000 BP an increase in Pinus indicated that conditions became drier and more temperate. Changes in the relative proportion of different tree pollen (Quercus, Pinus, Alnus) after 19,000 BP were thought to indicate cooler more moist conditions (González-Quintero and Fuentes-Mata, 1980; Brown, 1985; Bradbury, 1989). A later study of a different core from the same basin reported that conditions post 23,000 BP were both dry and cold (Lozano-García and Ortega-Guerrero, 1998). At the Tepexpan man site NE of Texcoco, Lamb et al. (2009) report high river inflow into a drying lake between about 19,000 and 16,500, while results from the Teotihuacan valley (slightly further NE) suggest a cool and humid late Pleistocene (Solleiro-Rebolledo et al., 2006). Similar discrepancies in interpretations of conditions around 22,000 have come from Chalco in the south of the Basin of Mexico (Bradbury, 1989; Caballero and Ortega Guerrero, 1998) but it seems likely that both pollen and diatom records may have been affected by the last major eruption of Popocatepetl around 22,000 BP. In the north of the basin, the record from Lake Tecocomulco (Caballero et al., 1999) indicates very low lake levels after 25,700 terminating in desiccation of the lake around 15,000 BP. Other sites in the central highlands with reasonable chronologies for the mid to late glacial all lie to the west of Valsequillo and the Basin of Mexico. The Upper Lerma basin seems to have been quite wet from about 22,000, with a shallow water phase between 19,000 and 16,000 (Caballero et al., 2002). Palaeosol records from the Nevado de Toluca (adjacent to the Lerma basin) also indicate humid, forest environments in the mid to late glacial (Sedov et al., 2001). Further west, data from Lake Cuitzeo show conditions wetter than present from 35,000 to 22,000, followed by shallowing (Velázquez Duran et al., 2001; Israde et al., 2002). The severity of this drying is unclear as there is a major hiatus in one core, whilst isotopic data from another indicates quite wet conditions round the LGM. The Pátzcuaro Basin, also to the west has records extending over the last 44,000 years (Bradbury, 2000; Metcalfe et al., 2007) and shows deep water conditions through the LGM and into the early Holocene. Pollen from a fluvial sequence in the Barranca Rancho Viejo, just north of Pátzcuaro lake also suggests cool moist conditions around 26,000 BP (Robles-Camacho et al., 2010). In summary, the results from the Barranca Caulapan are broadly consistent with other data from the wider region. The Pátzcuaro record is unequivocal about the persistence of high lake levels through the LGM and has been interpreted in terms of more effective winter precipitation. For basins to the east (including Valsequillo) things are less clear cut and it does appear that conditions were drying from about 30,000 BP (Caballero et al., 2010). Perhaps winter rainfall did not increase significantly in these more easterly parts of Mexico. It is increasingly clear that there was significant temperature depression in this area around the LGM (Lozano-García and Vázquez Selem, 2005) and this would certainly have had an effect on water balance. The various records of wetter conditions between about 22,000 and 18,000 may, therefore, also be related to reduced evaporation. The intra-shell isotope analyses from Barranca Caulapan indicate that there was significant seasonal variation in the amount of precipitation in the Valsequillo basin during the late Pleistocene. Central to the modern climate regime in Mexico are the monsoonal rains which occur in summer driven by seasonal changes in the position of the inter-tropical convergence zone and the development of low pressure over NW Mexico/SW USA (Poore et al., 2005). Although the modern climatic regime seems to have dominated the northern hemisphere neotropics since about 9000 BP, prior to that the winter rains are thought to have been enhanced due to the southwards displacement of the mid-latitude westerlies by the Laurentide ice sheet, however the timing and nature of this transition is a focus of debate (Metcalfe et al., 2000). The intra-shell isotope analyses from the Valsequillo basin suggest highly seasonal precipitation between 35,000 and 18,000 BP although it is not possible from the shell data to determine whether the increase rainfall occurred in the summer or the winter. Our palaeoclimate investigations in the Valsequillo Basin, in summary, indicate the following palaeoenvironmental record: • c. 35,000 BP: climate was similar to the present day. • Between c. 35,000 and 20,000 BP: conditions became increasingly dry. • Post 20,000 BP: conditions became increasingly humid again, although this record is truncated. • Precipitation varied seasonally. • The timing of wet season is not clear. Humans living in or passing through the Valsequillo Basin would have experienced changing long-term and seasonal climatic conditions and would have had to adapt their life strategies accordingly. It is clear, however, that conditions in the basin would never have been so extreme as to discourage human occupation if people did reach this part of the Americas in the late Pleistocene. Acknowledgements RES was in receipt of an Anne McLaren Fellowship at the University of Nottingham. We would like to thank Richard Jefferies, Jennifer Tripp, Sarah Davies, and Ben Aston for help with sample collection. We are grateful to Carol Arrowsmith and Joanne Green for technical assistance with isotope analyses. Teresa Needham and Graham Morris are thanked for assistance with laboratory equipment. Tom Higham, Jean-Luc Schwenninger and Alistair Pike are thanked for their assistance with chronology and radiocarbon dating of shells. This project was also funded by NERC research grants (NE/C519446/ 1), NIGFSC grant (IP/847/0505), and a Royal Geographical Society expedition grant (GFG 18/05). References Abell, P.I., 1985. Oxygen isotope ratios in modern African gastropod shells — a database for paleoclimatology. Chemical Geology 58, 183–193. Abell, P.I., Williams, M.A.J., 1989. Oxygen and carbon isotope ratios in gastropod shells as indicators of paleoenvironments in the Afar region of Ethiopia. Palaeogeography, Palaeoclimatology 74, 265–278. Balakrishnan, M., Yapp, C.J., 2004. Flux balance models for the oxygen and carbon isotope compositions of land snail shells. Geochimica et Cosmochimica Acta 68, 2007–2024. 26 R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 Howes, N.H., Wells, G.P., 1934. The water relations of snails and slugs: 1. Weight rhythms in Helix pomatia. Journal of Experimental Biology 11, 327–343. Huang, Y., Street-Perrott, F.A., Metcalfe, S.E., Brenner, M., Moreland, M., Freeman, K.H., 2001. Climate change as the dominant control on glacial–interglacial variation on C3 and C4 plant abundance. Science 293, 1647–1651. IAEA/WMO, 2006. Global Network of Isotopes in Precipitation. The GNIP Database. http://isohis.iaea.org. Irwin-Williams, C., 1962. Preliminary report on investigations in the region of Valsequillo Reservoir. Unpublished report submitted to Departamento de Prehistoria Instituto Nacional de Antropologia e Historia, México D.F. pp. 1–32. Irwin-Williams, C., 1978. Summary of archaeological evidence from the Valsequillo region, Puebla, Mexico. In: Browman, D.L. (Ed.), Cultural Continuity in Mesoamerica. Mouton, London, pp. 7–22. Israde, Alcantara, I., Garduño Monroy, V., Ortega, R., 2002. Paleoambiente lacustre del cuaternario tardío en el centro del lago de Cuitzeo. Hidrobiológica 2, 61–78. Kim, S.T., O'Neil, J.R., 1997. Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. Geochimica et Cosmochimica Acta 61, 3461–3475. Lamb, A.L., Gonzalez, S., Huddart, D., Metcalfe, S.E., Vane, C.H., Pike, A.W.G., 2009. Tepexpan Palaeoindian site, Basin of Mexico: multi-proxy evidence for environmental change during the late Pleistocene–late Holocene. Quaternary Science Reviews 28, 2000–2016. Leng, M.J., Marshall, J.D., 2004. Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quaternary Science Reviews 23, 811–831. Leng, M.J., Heaton, T.H.E., Lamb, H.F., Naggs, F., 1998. Carbon and oxygen isotope variations within the shell of an African land snail (Limicolaria kambeul chudeaui Germanin): a high-resolution record of climate seasonality? The Holocene 8, 407–412. Leng, M.J., Lamb, A.L., Lamb, H.F., Telford, R.J., 1999. Palaeoclimatic implications of isotopic data from modern and early Holocene shells of the freshwater snail Melanoides tuberculata, from lakes in the Ethiopian Rift Valley. Journal of Paleolimnology 21, 97–106. Lozano-García, M.S., Ortega-Guerrero, B., 1998. Late Quaternary environmental changes of the central part of the Basin of Mexico. Correlation between Texcoco and Chalco basins. Review of Palaeobotany and Palynology 99, 77–93. Lozano-García, S., Vázquez Selem, L., 2005. A high elevation Holocene pollen record from Iztaccihuatl volcano, central Mexico. The Holocene 15, 329–338. Malde, H.E., 1968. Preliminary draft on the stratigraphy of Valsequillo region (except valsequillo gravel, volcanic ash stratigraphy, etc.). Submitted to the U.S. Geological survey records library, Denver, Colorado, pp. 1–157. Mark, D.F., Gonzalez, S., Huddart, D., Bohnel, H., 2010. Dating of the Valsequillo volcanic deposits: resolution of an ongoing archaeological controversy in Central Mexico. Journal of Human Evolution 58, 441–445. McConnaughey, T.A., Gillikin, D.P., 2008. Carbon isotopes in mollusk shell carbonates. Geo-Marine Letters 28, 287–299. McMahon, R.F., 1983. Physiological ecology of freshwater pulmonates: volume 6, ecology. In: Russell-Hunter, W.D. (Ed.), The Mollusca. : Ecology., Volume 6. Academic Press, Orlando, Florida, pp. 359–461. Metcalfe, S.E., O'Hara, S.L., Caballero, M., Davies, S.J., 2000. Records of Late Pleistocene– Holocene climatic change in Mexico — a review. Quaternary Science Reviews 19, 699–721. Metcalfe, S.E., Davies, S.J., Braisby, J.D., Leng, M.J., Newton, A.J., Terrett, N.L., O'Hara, S.L., 2007. Long and short-term change in the Pátzcuaro Basin, central Mexico. Palaeogeography, Palaeoclimatology, Palaeoecology 247, 272–295. Metref, S., Rousseau, D.D., Bentaleb, I., Labonne, M., Vianey-Liaud, M., 2003. Study of the diet effect on delta C-13 of shell carbonate of the land snail Helix aspersa in experimental conditions. Earth and Planetary Science Letters 211, 381–393. Morse, S.A., Bennett, M.R., Gonzalez, S., Huddart, D., 2010. Techniques for verifying human footprints: reappraisal of pre-Clovis footprints in Central Mexico. Quaternary Science Reviews 29, 2571–2578. NatureServe, 2008. NatureServe Explorer: An Online Encyclopedia of Life. Newell, P.F., 1966. The nocturnal behaviour of slugs. Medical & Biological Illustration 16, 146–159. Pfeiffer, L., 1839. Bericht über die Ergebnisse meiner Reise nach Cuba im Winter 1838– 1839. Archiv für Naturgeshichte 5, 346–358. Pfeiffer, L., 1866. Beschreibung neuer Landschnecken. Malakozoologische Blätter 13, 76–91. Pilsbry, H.A., 1940. Land mollusca of North America (North of Mexico), volume 1 part 2. Academy of Natural Sciences of Philadelphia Monographs 3, 575–994. Pilsbry, H.A., 1946. Land mollusca of North America (North of Mexico), volume 2 part 1. Academy of Natural Sciences of Philadelphia Monograph 3, 1–520. Pilsbry, H.A., Ferriss, J.H., 1906. Mollusca of the Southwestern States II. Proceeding of Academy of Natural Sciences of Philadelphia 58, 123–175. Pointier, J.P., Augustin, D., 1999. Biological control and invading freshwater snails: a case study. Comptes Rendus de l'Académie des Sciences de Paris 322, 1093–1098. Poore, R.Z., Pavich, M.J., Grissino-Mayer, H.D., 2005. Record of the North American southwest monsoon from Gulf of Mexico sediment cores. Geology 33, 209–212. Prior, D.J., 1985. Water-regulatory behavior in terrestrial gastropods. Biological Reviews of the Cambridge Philosophical Society 60, 403–424. Renne, P.R., Feinberg, J.M., Waters, M.R., Arroyo-Cabrales, J., Ochoa-Castillo, P., PerezCampa, M., Knight, K.B., 2005. Age of Mexican ash with alleged ‘footprints’. Nature 438, E7–E8. Robles-Camacho, J., Corona-Chávez, P., Morales-Gámez, M., Guzmán, A.F., Polaco, O.J., Domínguez-Vázquez, G., Israde-Alcántara, I., Oliveros-Morales, A., 2010. Estratigrafía y paleoambiente asociados a un Gomphoteriidae (Cuvieronius hyodon) en Tzintzuntzan, Michoacán, México. Revista Mexicana de Ciencias Geologicas 27, 530–544. Balakrishnan, M., Yapp, C.J., Theler, J.L., Carter, B.J., Wyckoff, D.G., 2005. Environmental significance of 13C/12C and 18O/16O ratios of modern land-snail shells from the southern great plains of North America. Quaternary Research 63, 15–30. Baldini, L.M., Walker, S.E., Railsback, L.B., Baldini, J.U.L., Crowe, D.E., 2007. Isotopic ecology of the modern land snail Cerion, San Salvador, Bahamas: preliminary advances toward establishing a low-latitude island paleoenvironmental proxy. Palaios 22, 174–187. Barnhart, M.C., 1986. Respiratory gas tensions and gas exchange in active and dormant land snails, Otala lactea. Physiological Zoology 59, 733–745. Bequaert, J.C., Miller, W.B., 1973. The Mollusks of the Arid Southwest, with an Arizona Check List. University of Arizona Press, Tucson, p. 271. Binney, A., 1840. A monograph of the helices inhabiting the United States. Boston Journal of Natural History 3, 421–438. Bowen, G.J., Revenaugh, J., 2003. Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39, 1299–1311. Bowen, G.J., Wilkinson, B., 2002. Spatial distribution of delta O-18 in meteoric precipitation. Geology 30, 315–318. Bradbury, J.P., 1989. Late Quaternary lacustrine paleoenvironments in the Cuenca de Mexico. Quaternary Science Reviews 8, 75–100. Bradbury, J.P., 2000. Limnologic history of Lago de Patzcuaro, Michoacan, Mexico for the past 48,000 years: impacts of climate and man. Palaeogeography, Palaeoclimatology, Palaeoecology 163, 69–95. Brown, R.B., 1985. A summary of late-Quaternary pollen records from Mexico west of the Isthmus of Tehuantepec. In: Bryant, V.M., Holloway, R.G. (Eds.), Pollen Records of Late Quaternary North American Sediments. American association of stratigraphic palynologists, Dallas, pp. 71–93. Burch, J.B., 1962. How to Know the Eastern Land Snails. W.M.C. Brown Co., Dubuque, Iowa, p. 214. Caballero, M., Ortega Guerrero, B., 1998. Lake levels since about 40,000 years ago at Lake Chalco, Mexico City. Quaternary Research 50, 69–79. Caballero, M., Lozano, S., Ortega, B., Urrutia, J., Macias, J.-L., 1999. Environmental characteristics of Lake Tecocomulco northern Basin of Mexico, for the last 50,000 years. Journal of Paleolimnology 22, 399–411. Caballero, M., Ortega, B., Valadez, F., Metcalfe, S.E., Macias, J.-L., Sugiura, Y., 2002. Sta. Cruz Atizapan: a 22-ka lake level record and climatic implications for the late Holocene human occupation in the Upper Lerma Basin, central Mexico. Palaeogeography, Palaeoclimatology, Palaeoecology 186, 217–235. Caballero, M., Lozano-Garcia, S., Vazquez-Selem, L., Ortega, B., 2010. Evidencias de cambio climático y ambiental en registros glaciales y en cuencas lacustres del centro de México durante el último máximo glacial. Boletín de la Sociedad Geológica Mexicana 62, 359–377. Cheatum, E.P., Fullington, R.W., 1971. The aquatic and land Mollusca of Texas. Part one. The Dallas Museum of Natural History Bulletin, 1, p. 77. Cook, A., 1979. Homing in the Gastropoda. Malacologia 18, 315–318. Cowie, R.H., 1984. The life-cycle and productivity of the land snail Theba-pisana (Mollusca, Helicidae). Journal of Animal Ecology 53, 311–325. Dansgaard, W., 1964. Stable isotopes in precipitation. Tellus 16, 436–468. Edelstam, C., Palmer, C., 1950. Homing behaviour in gastropodes. Oikos 2, 259–270. Feinberg, J.M., Renne, P.R., Arroyo-Cabrales, J., Waters, M.R., Ochoa-Castillo, P., PerezCampa, M., 2009. Age constraints on alleged ‘footprints’ preserved in the Xalnene Tuff near Puebla, Mexico. Geology 37, 267–270. Friedli, H., Lötscher, H., Oeschger, H., Siegenthaler, J., Stauffer, B., 1986. Ice core record of the 13C/12C ratio of atmospheric CO2 in the past two centuries. Nature 324, 237–238. Fritz, P., Poplawski, S., 1974. 18O and 13C in the shells of freshwater molluscs and their environments. Earth and Planetary Science Letters 24, 91–98. Gelperina, A., 1974. Olfactory basis of homing behaviour in the giant garden slug, Limax mimus. Proceedings of the National Academy of Sciences of the United States of America 71, 966–970. González, S., Jimenez-Lopez, J.C., Hedges, R., Huddart, D., Ohman, J.C., Turner, A., Padilla, J., 2003. Earliest humans in the Americas: new evidence from Mexico. Journal of Human Evolution 44, 379–387. González, S., Huddart, D., Bennett, M.R., González-Huesca, A., 2006a. Human footprints in Central Mexico older than 40,000 years. Quaternary Science Reviews 25, 201–222. González, S., Huddart, D., Bennett, M.R., 2006b. Valsequillo Pleistocene archaeology and dating: ongoing controversy in Central Mexico. World Archaeology 38, 611–627. González-Quintero, L., 1986. Analisis polinicos de los sedimentos. In: Lorenzo, J.L., Mirambell, L. (Eds.), Tlapacoya: 35,000 Años de historia del lago de Chalco. Colección cientifica, Serie prehistoria, Instituto Nacional de Antropologia e Historia, pp. 157–166. González-Quintero, L., Fuentes-Mata, M., 1980. El Holoceno de la porción central de la Cuenca de México. In: Sánchez, F. (Ed.), Memoria III colóquio sobre paleobotánica y palinología: Colección científica Instituto de Antropología e Historia, 86, pp. 113–132. Goodfriend, C.A., Ellis, G.L., 2000. Stable carbon isotope record of middle to late Holocene climate changes from land snail shells at Hinds Cave, Texas. Quaternary International 67, 47–60. Goodfriend, G.A., Magaritz, M., 1987. Carbon and oxygen isotope composition of shell carbonate of desert land snails. Earth and Planetary Science Letters 86, 377–388. Goodfriend, G.A., Magaritz, M., Gat, J.R., 1989. Stable isotope composition of land snail body-water and its relation to environmental waters and shell carbonate. Geochimica et Cosmochimica Acta 53, 3215–3221. Grossman, E.L., 1982. Stable isotopes in live benthic foraminifera from the southern California Borderland, Thesis, University of Southern California, La Jolla, California, USA. R.E. Stevens et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 319-320 (2012) 16–27 Roth, B., Hartman, J.H., 1998. A probable Cerion (Gastropoda: Pulmonata) from Uppermost Cretaceous Hell Creel Formation, Garfield County, Montana. Paleobios 18 (2&3), 16–20. Rozanski, K., Araguas-Araguas, L., Gonfiantini, R., 1993. Isotopic patterns on modern global precipitation. In: Swart, P.K., Lohmann, K.C., McKenzie, J., Savin, S. (Eds.), Climate Change in Continental Isotopic Records. American Geophysical Union, pp. 1–36. Say, T., 1825. Descriptions of some new species of fresh water and land snails of the United States. Journal of the Academy of Natural Sciences of Philadelphia 5, 119–131. Say, T., 1829. Descriptions of some new terrestrial and fluviatile shells of North America. The new harmony disseminator of useful knowledge. . (3, 260. Aug. 26). Sedov, S., Solleiro-Rebolledo, E., Gama-Castro, J.E., Vallejo-Gomez, E., Gonzalez-Velazquez, A., 2001. Buried palaeosols of the Nevado de Toluca: an alternative record of Late Quaternary environmental change in central Mexico. Journal of Quaternary Science 16, 375–389. Shuttleworth, R.J., 1852. Diagnosen neuer Mollusken. Mittheilungen der Naturforschenden Gesellschaft in Bern, 260/261, pp. 289–304. Solleiro-Rebolledo, E., Sedov, S., de Tapia, E.M., Cabadas, H., Gama-Castro, J., VallejoGomez, E., 2006. Spatial variability of environment change in the Teotihuacan Valley during the Late Quaternary: paleopedological inferences. Quaternary International 156, 13–31. Stott, L.D., 2002. The influence of diet on the delta C-13 of shell carbon in the pulmonate snail Helix aspersa. Earth and Planetary Science Letters 195, 249–259. Szabo, B.J., Malde, H.E., Irwin-Williams, C., 1969. Dilemma posed by uranium-series dates on archaeologically significant bones from Valsequillo, Puebla, Mexico. Earth and Planetary Science Letters 6, 237. Taylor, D.W., 1957. Pliocene fresh-water mollusks from Navajo County, Arizona. Journal of Paleontology 31, 654–661. 27 Thompson, F.G., 1971. Some Mexican land snails of the genera Coelostemma and Metastoma (Urocoptidae). Bulletin of the Florida State Museum 15, 267–302. Thompson, F.G., 2008. An Annotated Checklist and Bibliography of the Land and Freshwater Snails of Mexico and Central America(On-line publication) http://www. flmnh.ufl.edu/malacology/mexico-central_america_snail_checklist/. Thompson, R., Cheny, S., 1996. Raising Snails. National Agriculture Library Special Reference Briefs. (NAL SRB 96-05). Van der Schalie, A., Getz, L.L., 1961. Comparison of adult and young Pomatiopsis cincinnatiensis (Lea) in respect to moisture requirements. Transactions of the American Microscopical Society 80, 211–220. Van der Schalie, A., Getz, L.L., 1963. Comparison of temperature and moisture responses of the snail genera Pomatiopsis and Oncomelania. Ecology 44, 73–83. VanLandingham, S.L., 2004. Corroboration of Sangamonian age of artifacts from the Valsequillo region, Puebla, Mexico by means of diatom biostratigraphy. Micropaleontology 50, 313–342. Velázquez Duran, R., Israde Alcantare, I., Lozano García, S., 2001. Significado ecológico de los palinomorfos depositados en el lago de Cuitzeo, Michoacán, durante los últimos 35,000 años. Ciencia Nicolaita 29, 45–61. Yanes, Y., Delgado, A., Castillo, C., Alonso, M.R., Ibanez, M., De La Nuez, J., Kowalewski, M., 2008. Stable isotope (delta O-18, delta C-13 and delta D) signatures of recent terrestrial communities from a low-latitude, oceanic setting: endemic land snails, plants, rain, and carbonate sediments from the eastern Canary Islands. Chemical Geology 249, 377–392. Yapp, C.J., 1982. A model for the relationships between precipitation D/H ratios and precipitation intensity. Journal of Geophysical Research — Oceans and Atmosphere 87, 9614–9620. Zanchetta, G., Leone, T.G., Fallick, A.E., Bonadonna, F.P., 2005. Oxygen isotope composition of living land snail shells: data from Italy. Palaeogeography, Palaeoclimatology, Palaeoecology 223, 20–33.