Quaternary Geochronology 10 (2012) 430e435 Contents lists available at SciVerse ScienceDirect Quaternary Geochronology journal homepage: www.elsevier.com/locate/quageo Research paper Challenges in constraining pluvial events and hominin activity: Examples of ESR dating molluscs from the Western Desert, Egypt B.A.B. Blackwell a, b, *, A.R. Skinner a, F. Mashriqi b, A.E. Deely b, R.A. Long b, J.J.J. Gong b, M.R. Kleindienst c, J.R. Smith d a Dept. of Chemistry, Williams College, Williamstown, MA 01267-2692, USA RFK Science Research Institute, Glenwood Landing, NY 11547-0866, USA Dept. of Anthropology, University of Toronto at Mississauga, Mississauga, ON L5L 1C6, Canada d Dept. of Earth & Planetary Sciences, Washington University in St. Louis, St. Louis, MO 63130-4862, USA b c a r t i c l e i n f o a b s t r a c t Article history: Received 15 October 2011 Received in revised form 10 January 2012 Accepted 11 January 2012 Available online 20 January 2012 Receiving <0.1 mm/y of precipitation, Egypt’s hyperarid Western Desert, today lacks naturally occurring surface water. Artesian spring deposits, tufa deposited by springs and carbonate-rich silty lacustrine sediment attest that oases in the Western Desert had surface water during the Pleistocene. Paleolithic artefacts, fossil ungulate teeth, and snails occurring within the Pleistocene deposits and dotting the surface record times when higher rainfall and/or groundwater tables during pluvial events allowed surface water to exist in wetlands, small ponds and lakes, enabling hominin habitation. Archaeological finds ranging from Early to Later Stone Age (ESAeLSA) occur in gravel lags, within sedimentary deposits, and on the older geomorphic surfaces. Near Kharga, large tufa deposits ranging from a few hectares to more than 10 km2 in area, such as Matana and Medauwara, dot the edge of the Libyan Plateau. Molluscs were dated using standard ESR protocols. To test for reworked fossils, multiple samples from a single sample were dated independently. In some units at Medauwara, multiple gastropod populations from different times have been preserved, while others appear to only preserve a single population. To see the effects of the cosmic dose rate on ESR ages, ages were calculated using zero cosmic dose rate, the full modern cosmic dose rate, and time-averaged cosmic and sedimentary dose rates. For gastropods from Matana, no significant difference in ESR ages resulted from different cosmic dose rate assumptions. Therefore, at Matana 2, the shells dated at 27.7  1.9 assuming time-averaged external dose rates, while at Matana 3, they averaged 65.1  4.1 ka, suggesting that water was present for hominin use at times during OIS 2 and 4. Ó 2012 Published by Elsevier B.V. Keywords: ESR (electron spin resonance) dating Matana Kharga oasis Egypt Medauwara Kharga oasis Egypt Cosmic dose rate modeling External dose rate modeling Reworking 1. Introduction Today, the hyperarid Western Desert, Egypt, receives w0.01e0.07 mm/y precipitation, but suffers >2 m/y of evaporation, ensuring that area lacks naturally occurring standing water (Smith et al., 2007). In past pluvial events, however, this area may have received as much as 30 cm/y (Hawkins et al., 2001). Hominins inhabited the Western Desert at sites around Kharga Oasis (Fig. 1), and left behind their Earlier (ESA), Middle (MSA), and Later Stone Age (LSA) artefacts. At Kharga, fossil freshwater molluscs that sit in lacustrine silt and mud beds record times when surface water did * Corresponding author. Dept of Chemistry, Williams College, Williamstown, MA 01267-2692, USA. Tel./fax: þ1 516 759 6092. E-mail address: bonnie.a.b.blackwell@williams.edu (B.A.B. Blackwell). 1871-1014/$ e see front matter Ó 2012 Published by Elsevier B.V. doi:10.1016/j.quageo.2012.01.005 exist for hominins and their game to drink, probably thanks to higher rainfall during pluvial events (Adelsberger and Smith, 2010). Dating these events, however, has proved to be more challenging than finding the artefacts that document the hominin presence. Rarely do the artefacts occur within any sedimentary context that can be correlated to any datable materials, while intermittent fluvial and the pervasive aeolian erosion have removed most paleoenvironmental indicators. Thus, dating proxy materials becomes imperative. Pleistocene lake sediment, artesian spring and tufa deposits, all show that Kharga Oasis had surface water repeatedly during the Pleistocene. Tufa can only form when spring water brings dissolved salts to the surface and evaporates to deposit the carbonate. Where tufa dams block water flow, small basins form within the tufa deposits. In order for basins to trap, and hence, preserve silt or mud deposits, those basins must contain standing water. Freshwater 431 B.A.B. Blackwell et al. / Quaternary Geochronology 10 (2012) 430e435 Table 1 Samples in the study. Fig. 1. The Kharga Oasis Depression. Near Kharga, large tufa deposits dot the escarpment edge between the Libyan Plateau and the Kharga Oasis Depression. Lacustrine silts containing freshwater snails were sampled at sites within Matana (also spelled as Metana in some publications) and Medauwara. snails, like Melanoides tuberculata, Gyraulus sp., Lymnaea stagnalis, and Plantries planorbis, can only survive and breed when abundant surface water fills lakes. Hominids, and many ungulate species, require relatively fresh surface water to drink daily. At times during the Pleistocene, water was sufficiently abundant to maintain a large, relatively diverse ecosystem attractive to hominins, who left numerous Paleolithic artefacts scattered across the landscape ranging from Oldowan choppers to LSA settlements, and even Neolithic hearths and beads to Roman pottery. For dating the archaeological artefacts, datable materials are rare. Although 230Th/234U can sometimes date tufas younger than w400 ka, detrital 230Th contaminates the tufa complicating or preventing successful 230Th/234U dating (Blackwell and Schwarcz, 1995). Moreover, the porous travertines and lake carbonates can contain multiple generations of calcite deposition that preclude accurate dates with any method. Some tufa units from Kharga have been dated successfully, mainly from Medauwara (Kleindienst et al., 2008; Smith et al., 2004a,b). ESR (electron spin resonance) can directly date freshwater molluscs from 5 ka to 2 Ma in age (Blackwell, 2006). No tufa units at Kharga have yielded large molluscs, but Matana and Medauwara (Fig. 1) have produced freshwater gastropods for ESR dating. Freshwater gastropods from Matana (also spelled Metana) and Medauwara were analyzed to assess their suitability for ESR dating (Table 1). Potential effects from variable cosmic dose rates, sedimentary dose rates, and evidence for reworked fossils within deposits were examined. 2. Geology and archaeology at Kharga Oasis Below the Libyan Plateau in Egypt’s Western Desert, roughly 650 km southwest of Cairo at 25 N 30 E and 100 m asl, the Kharga Oasis depression extends w200 km north-south, and w75 km westeast (Fig.1). The oasis formed as a depression ablated by the wind into Cretaceous-Tertiary marine limestone, shale and sandstone units. Eocene Thebes Group limestone now rims the depression (Smith et al., 2004a,b; Nicoll et al., 1999). In 1930e1933, Caton-Thompson and Gardner studied the archaeology of Kharga Oasis (Caton- Number Location ESR Outcrop GPS coordinates Species Catalogue Name North East RM47 KAR17 Big Snail Gully Medauwara N2760628 24 56.9290 E0307807 31 05.9050 Melanoides tuberculata RM51-RM54 KAR17a-d Big Snail Gully Medauwara N2760628 E0307807 Melanoides tuberculata FM77-FM80 KAR102a-f Parking Lot Medauwara 24 570 49.600 31 050 15.800 Melanoides tuberculata FM81-FM84 KAR105a-d Railway 3 Medauwara 24 570 16.500 31 050 02.100 Melanoides tuberculata FM85 KAR105e Railway 3 Medauwara 24 570 16.500 31 050 02.100 Gyraulus sp. AM08a,b KAR106a,d Matana 2 25 010 59.100 30 570 17.800 Melanoides tuberculata AM07a-b KAR107a-b Matana 3 25 010 59.000 30 570 18.300 Melanoides tuberculata Thompson, 1952). Kleindienst et al. (2008) and Smith et al. (2007) relocated some of their localities on the escarpment at Naqb el Refûf, Bulaq, and Matana. The Dakhleh Oasis and Kharga Oasis Prehistory Projects have investigated Medauwara area since 1996. Along the Libyan Plateau that rims the depression, springs have deposited several large tufa accumulations, such as Medauwara and Matana (Figs. 1 and 2), which mark nickpoints for the local groundwaters at slope breaks. Locally, tufa dammed ponds up to a few square metres and even small lakes up to 1e5 km2. These ponds and lakes accumulated lacustrine silts, which often comprise large amounts of carbonates mixed with small inputs of clastic sediment. Both the lacustrine silts, and the tufa preserves freshwater gastropods, including M. tuberculata and Gyraulus sp. (Smith et al., 2004a). Since these gastropods can only live in fresh or brackish water, dating their shells indicates when hominins could inhabit the area (Adelsberger and Smith, 2010). Since tufa resists aeolian erosion better than does the lacustrine and playa sediment, the tufa itself and the lake deposits rimmed or overlain by tufa often preserve the only evidence for human occupation in the Western Desert (Hawkins et al., 2001). Because the springs that result from aquifer storage deposit tufa from water at ambient temperature, tufa also records when an ample water supply existed (Nicoll et al., 1999). Freshwater gastropods also record wetter climatic conditions, because they can only exist when ample water is present in ponds or lakes. A region hosting substantial tufa deposits >12 km2 in area near an ancient caravan route (Nicoll et al., 1999), Naqb (Wadi) el Medauwara contains Earlier Stone Age, Upper Acheulean, Older and Younger MSA, Aterian, Khargan, Epipaleolithic, and Neolithic occurrences. Chert nodules which may have served as the raw material for lithic production occur in the gravel lags within some basins. Although Smith et al. (2004a,b) recognized three tufa units, the Plateau Tufa, and Wadi Tufas 1 and 2 (Fig. 2), more recent analyses suggest a much more complex stratigraphy. Complicating the problem is the limited extent for many small lacustrine basins, coupled with the extensive erosion that has removed intervening mapable units. Deposited by direct precipitation near spring orifices, these tufas occur at different altitudes on the escarpment. Tufa dammed small ponds with areas of 1e5 m2 that trapped silts. Small lakes with areas up to 1e5 km2 also formed in scattered locations, including at Railway Outcrop and Big Snail Gully. Unfortunately, subsequent erosion and some anthropogenic surface modification has left isolated outcrops containing molluscs 432 B.A.B. Blackwell et al. / Quaternary Geochronology 10 (2012) 430e435 Fig. 2. The spring tufa deposits at Medauwara. Medauwara contains several generations of spring tufas. The tufas dammed ponds and small lakes which collected carbonate lacustrine silt and freshwater gastropods, such as Melanoides tuberculata. Gastropods were collected from three localities. that are difficult to relate to one another (Kieniewicz and Smith, 2007). At Kharga, however, few sections containing more than a few stratigraphic units occur anywhere. Since mapped tufa units do not occur as one outcrop, but are discontinuously distributed laterally across the area, interbasinal correlation has been difficult, and correlation between scattered tufa deposits even more challenging. Even where two outcrops occur only few tens of metres apart, correlating the units is often impossible. Hence, chronostratigraphic dating will help to map their stratigraphy and archaeological associations. Thus, ages for the samples dated here will be used to build a more detailed chronostratigraphic map of the area. During the Oxygen Isotope Stage (OIS) 6/5e pluvial event, the same precipitation source fed all the regions and the aquifer sources, likely deriving from air masses originating over the Atlantic Ocean. This rainfall produced low d18O ratios in the tufa, as well as deep and shallow groundwater in Nubian and Libyan aquifers, and maintained a small recurring paleolake, which provided a stable water source at Medauwara. In the early Holocene pluvial period before 5 ka, relatively abundant C4 species, such as grasses, carpeted the area near the springs (Smith et al., 2004a; Kieniewicz, 2007). A smaller tufa accumulation than Medauwara, Matana clings to the Libyan Plateau in a region not accessed by roads, which requires a lengthy climb over dunes and small cliffs up the escarpment edge. Few studies have occurred at Matana due to the difficulty in accessing the site without camels. Nonetheless, Gardner reported several MSA localities there (Caton-Thompson, 1952). Although Kharga tufas are well preserved, colluvial and eolian silt and sand can contaminate tufas, thereby complicating 230 Th/234U dating efforts. Tufas overlying lake silts were dated at 126  4 ka by 230Th/234U at Medauwara (Smith et al., 2004a), 103  14 ka at KOPP Matana 02 (Smith et al., 2007), 240  5 ka and 125 þ 1.6 ka at El Refûf (Kleindienst et al., 2008). 3. ESR Dating At Medauwara, molluscs from three areas were tested to determine the prevalence of reworked fossils (Table 1, Figs. 1 and 2). All samples were collected from within the snail-bearing horizons in areas where snails were found eroding from the underlying sandy silts. While the snail-bearing horizons at all three localities may or may not represent penecontemporaneous lake(s) deposits given their altitudes and locations within the tufa deposit, the presence of visibly different taphonomic characteristics in at least two of them makes it necessary to examine the collections for possible reworked samples. Detailed stratigraphic relationships between the outcrops are impossible to verify except through chronological analyses. At Matana, two mollusc samples were collected for ESR dating. Both came from silty sand units that were capped by dirty discontinuously eroded tufa units, that overlie the two sampling locations (Table 1, Fig. 1). Approximately 5 m apart vertically, the two sampling locations were separated by some 30 m laterally, from which a significant portion of the capping travertine had been eroded or had never existed. Intervening fine sediment had also 433 B.A.B. Blackwell et al. / Quaternary Geochronology 10 (2012) 430e435 been winnowed away, making it impossible to relate these two units stratigraphically. ESR can provide absolute dates for aragonitic molluscs over 5 kae2 Ma using the stable signal at g ¼ 2.007 (Skinner, 1988; Molod’kov, 1993; Blackwell, 2001). Ages are found using Eq. (1): A S ¼ Z t 0 ðDint ðtÞ þ Dsed ðtÞ þ Dcos ðtÞÞdt (1) where A S ¼ the total accumulated dose in the sample, Dint(t) ¼ the dose rate from internal sources: U, its daughters, and any other radioisotopes, Dsed(t) ¼ the dose rate from external sources: sedimentary U, Th, and K, Dcos(t) ¼ the dose rate from external sources: cosmic radiation, t ¼ the sample’s age (Deely et al., 2011). Due to the climatic variability in the Sahara during the Pleistocene, successful ESR dating must consider how reworking and deflation affect the calculation for the sedimentary and cosmic dose rates. Molluscs were prepared using standard ESR dating procedures (Skinner, 1988; Skinner and Shawl, 1994). Using a 60CO g source, all but one aliquot were irradiated at a dose rate of 56e132 Gy/s with added doses ranged from 8 to 1300 Gy. Molluscs were annealed for 1.0 h at 90  C to remove unstable interference. At 25  C, aliquots were scanned in a JEOL RE1X ESR spectrometer at 2 mW power under a 100 kHz field modulation of 5 mT, using a 0.3 s time constant. The spectra were scanned over 50 mT, centered at 330 mT for molluscs with an 8.0 min sweep time. Gains were set to maximize signal intensity. To measure Dint(t), U concentrations within the fossils were analyzed geochemically with NAA (see Blackwell, 1989 for analytical details) and were modelled using the standard early uptake (EU, p ¼ 1), linear uptake (LU; p ¼ 0), and recent uptake (RU, p ¼ 10) models using Rosy (Brennan et al., 1997). To determine Dext(t), volumetrically averaged sedimentary geochemical analysis was used. For sedimentary geochemistry, sediment samples from within 30 cm of the sample were geochemically analyzed by NAA, for radioactive elements and then volumetrically averaged for each layer around the fossil(s). Due to low sedimentary radioactivity, Dsed(t) values for the molluscs constitute w20e50% of DS(t), the total dose rate. Thus, as the source for the remaining 20e35% of DS(t), Dcos(t) may present a significant uncertainty in the calculated ages. To examine such uncertainties for each dating sample, several time-averaged cosmic dose rates, Dcos ðtÞ, were modelled as noted above (Deely et al., 2011). To test for reworked molluscs, where sample sizes permitted, several random grab samples from each mollusc-bearing bed were tested for age coherence. Mean ages were calculated by inversely weighting the ages by their associated standard deviations. 4. Results Dating multiple samples from in situ stratigraphic deposits associated with lacustrine or fluvial units will often demonstrate if a deposit contains fossils of different ages that were mixed during reworking. If their ages, U concentrations, and accumulated doses agree statistically, then the possibility for reworked fossils becomes less likely (e.g., Table 2; Blackwell, 1994). To test for reworking at three outcrops, several samples were analyzed. For the reworking tests, only the LU ages will be discussed. The trends in the EU or RU ages give similar results. For KAR17 from Big Snail Gully, the shells were chosen randomly from within the whole sample. U concentrations in these subsamples show little variation, all falling in the range 0.66e0.81  0.02 ppm U (Table 2a). Although the accumulated Table 2 Tests for reworked snails from Medauwara, Kharga Oasis. Accumulated dose, A S (Gray) Agea,b LU (ka) 0.66  0.02 67.8  5.3 104.1  9.2 grab sample 0.76  0.02 61.0  2.2 94.1  5.2 grab sample 0.75  0.02 58.6  2.9 90.1  5.9 grab sample 0.67  0.02 62.3  3.3 95.9  6.5 grab sample 0.81  0.02 68.7  4.0 105.5  7.6 3 unweathered 7.46  0.02 29.5  1.6 24.0  3.0 25.97  0.02 56.5  3.1 31.8  2.8 15.34  0.02 42.8  4.2 34.4  5.1 26.09  0.02 75.2  3.4 40.1  5.3 11.52  0.02 57.3  3.3 50.2  5.8 1.16  0.02 40.5  4.0 63.6  8.2 0.93  0.02 54.7  4.7 87.2  10.5 1.06  0.02 42.4  2.1 67.1  6.5 1.26  0.02 54.8  4.6 84.3  10.0 Sample Description a. Big Snail RM47 KAR17 RM51 KAR17c RM54 KAR17b RM53 KAR17a RM52 KAR17d Gully grab sample b. Railway FM85 KAR105e FM84 KAR105d FM82 KAR105b FM83 KAR105c FM81 KAR105a light, rugose unweathered dark, rugose unweathered light, smooth weathered dark, smooth weathered c. Parking Lot Basin FM77 small, light, rugose KAR102a unweathered FM80 medium, red-stained, KAR102f rugose, unweathered FM79 medium size, gray, KAR102e smooth; weathered FM78 medium, light beige, KAR102d smooth; weathered [Umol]a (ppm) a Abbreviations: LU ¼ assuming linear (continuous) U uptake, p ¼ 0, [Umol] ¼ U concentration in the molluscs. b Ages calculated with parameters listed in Table 4 and cosmic dose rate, Dcos ðtÞ ¼ 0.293  0.010 mGy/y. doses show somewhat more variation, the ages do not differ significantly given their 2 s errors (95% confidence limit). Hence, it appears that KAR17 does not contain multiple populations of shells. For KAR105, however, the samples had visually identifiable subpopulations within the sample. For each, these were first separated into the subpopulations and then analyzed as separate splits (Table 2b). Among the M. tuberculata splits, the U concentrations ranged from 12 to 26 ppm, while the Gyraulus had only 7.5 ppm. Their accumulated doses showed similar variation. For KAR105, unweathered shells that retained their rugose outer surface gave younger ages, ranged from 24.0  3.0 to 34.4  5.1 ka. In the strongly polished shells, the ages ranged from 40.1  5.3 to 50.2  5.8 ka. Although the weathered shells do not differ statistically from the oldest of the unweathered shells, they are statistically different from the other two unweathered shell populations. Therefore, at Railway 3, at least two populations appear to be present, an older weathered population, and a younger unweathered population. At Parking Lot Basin, KAR102 again visually identifiable subpopulations were separated before analysis (Table 2c). With U concentrations ranging from 0.9 to 1.3 ppm, these shells appear similar. Their accumulated doses and ages, however, hint that two populations may be mixed here, one with accumulated doses averaging 42.0  1.9 Gy and ages at 65.8  5.1 ka and another with accumulated doses at 54.8  3.3 Gy and ages of 85.7  7.3 ka. Based on their accumulated doses, these two groups are statistically different at the 95% confidence limit. In this instance, therefore, neither the gastropod size nor the presence of weathering uniquely distinguished the two subpopulations. 434 B.A.B. Blackwell et al. / Quaternary Geochronology 10 (2012) 430e435 ages. Nonetheless, modelling both Dsed ðtÞ, and Dcos ðtÞ allows us to constrain the fossils’ ages. To model Dcos ðtÞ, species environmental preferences, geological and geomorphological indicators are used to estimate past water or sediment depths and, thus, the minimum and maximum cover thicknesses, as well as the amount of erosion and its relative timing. With ramped box models with these constraints, minimum and maximum Dsed ðtÞ and Dcos ðtÞ; minimum and maximum ages can then be derived (for a detailed discussion, see Deely et al., 2011). Fig. 3 illustrates such an analysis for snails from Matana. At both sites, DBG ðtÞ ranged from 0.40  0.04 to 0.48  0.05 mGy/y, and sed:b DBG ðtÞ from 0.461 to 0.496  0.03 mGy/y (Table 3). Despite the sedg Fig. 3. AM07b’s age vs. external dose rate, Dext(t). As Dext(t) rises, the ages drop and the differences between the different model ages decrease. Considering the 2 s errors (95% confidence limit), the ages calculated with a cosmic dose rate, Dcos(t) ¼ 0 mGy/y, do not differ significantly from those calculated using the time-averaged external dose rate, Dext ðtÞ, which was modeled for changes in both Dcos(t) and Dsed(t) with time. The ages calculated using the modern Dcos(t), 292.9 mGy/y, do not significantly differ from those ages calculated with Dext ðtÞ. Possibly, reworking or deflation produced these two sets of mixed shell populations. In a blowout, as the wind winnows away the fine sediment and the blowout deepens, the heavier grains, such as hominin artefacts, gastropod shells, mammalian teeth and bones, drop onto the sediment surface. All traces for some units may disappear. Thus, deflation can mix coarser grains removed from any higher eroded units, making it more challenging to date the artefacts or fossils. Consequently, before final ages can be calculated, the region must be mapped more extensively to attempt to determine the original units from which these shells derived, and to estimate the time-averaged cosmic dose rate, Dcos ðtÞ, and time-averaged sedimentary dose rate, Dsed ðtÞ, for the shells in Railway 3 and Parking Lot Basin. In open air desert sites, determining Dcos ðtÞ, is essential. Since deflation reworks fossils, changing both their Dsed(t) and their Dcos(t), this requires modelling in order to calculate reliable ESR age difference between the two sites, the similarity in the sedimentary geochemical data suggests that both Matana 2 and 3 sediment derived from the same source, whether that source was aeolian sediment, authigenic carbonate deposited from lacustrine/ pond/spring water at the time the snails lived or later. At Matana 3, potential sediment cover thicknesses fell in the range of w1e2 m, while the water depths, although likely of limited duration, varied from possibly 10 cm to not more than 1 m. The presence of a significant tufa unit above the snail-bearing silts does suggest that the water ponded existed here for more than a few years. Using Dsed ðtÞ and Dcos ðtÞ for AM07b led to a mean Dext ðtÞ ¼ 1036  80 mGy/y, yielding an LU age of 69.2  6.9 ka (Fig. 3). This age does not differ significantly from the age, 62.8  5.9 ka calculated assuming continuous exposure to the modern Dcos(t) (Table 4). AM07b’s age with time-averaged dose rates also does not differ significantly, 82.5  7.7 ka, calculated assuming Dcos(t) ¼ 0 mGy/y. While this is clearly an impossibly low Dcos ðtÞ; given that the snails now sit under w1.0e1.5 m of sediment, it does set the absolute minimum age. At Matana 2, the mean timeaveraged Dext(t) at 1106  80 mGy/y gave a mean LU age of 27.7  1.9 ka. Again, these ages did not differ significantly from those using the modern Dcos(t) ¼ 292.9 mGy/y or Dcos(t) ¼ 0 mGy/y (Table 4). The reality lies somewhere between these extrema, but does not yield any significant change in the calculated ages for these snails. Since two samples at both Matana 2 and 3 constituted the same species, these samples can be averaged to give a mean for each site. At Matana 2, the U in the molluscs, [Umol], averaged 0.73  0.09 ppm, but [Umol] ¼ 0.55  0.06 ppm at Matana 3 (Table 4). At Matana 2, A S agree well with a mean at 31.6  1.5 Gy, while at Matana 3, they agree well, averaging 69.0  2.0 Gy. At Matana 2, using the time-averaged Dext(t) of 1106  80 mGy/y, gave a mean LU age of 27.7  1.9 ka. At Matana 3, the time-averaged Dext(t) ¼ 1036  80 mGy/y, yielding a mean LU age of 65.1  4.1 ka. Therefore, Matana 2 and 3 were deposited during the Late Pleistocene, in OIS 2 and 4 respectively, indicating that freshwater was available then for hominin use. Table 3 Sedimentary radioactivity at Matana. Dose ratesa Sample Concentrations NAA/type Field U (ppm) Th (ppm) K (wt%) DBG ðtÞb sed;b (mGy/y) DBG ðtÞc sed;g (mGy/y) AM08a-b sed KAR106b AM07a-b sed KAR107d 1.60  0.02 4.68  0.28 0.54  0.02 0.476  0.050 0.496  0.033 1.76  0.02 4.12  0.27 0.44  0.01 0.396  0.043 0.461  0.032 a Abbreviations: DBG ðtÞ ¼ sedimentary dose rate from b sourcesDBG ðtÞ ¼ sedimentary dose rate from g sourcesCalculated with aragonite density, rmol ¼ 2.95  0.01 g/cm3, sed;g sed;b carbonate sediment density, rcal ¼ 2.96  0.01 g/cm3, quartz sediment density, rqtz ¼ 2.66  0.01 g/cm3, sedimentary water concentration, Wsed ¼ 10.0  5.0 wt%, All errors are 1 s. b Uses thicknesses from the nearest shell. c Uses cosmic dose rate, Dcos ðtÞ ¼ 0.0  0.0 mGy/y. 435 B.A.B. Blackwell et al. / Quaternary Geochronology 10 (2012) 430e435 Table 4 ESR ages for Matana molluscs. Sample AM07a KAR107a AM07b KAR107b Matana 3 Mean (n ¼ 2) AM08a KAR106a AM08b KAR106d Matana 2 Mean (n ¼ 2) [Umol] (ppm) Accumulated dose, AS (Gray) ESR Agesa,b ESR Agesa,c Dext(t) (mGy/y) EU (ka) LU (ka) RU (ka) Dext(t) (mGy/y) EU (ka) LU (ka) RU (ka) 0.65  0.02 67.30  2.22 1149.9  80.0 54.1  1.9 56.5  1.9 58.1  1.9 1036.2  80.0 59.3  4.7 62.4  5.1 64.4  5.4 0.82  0.02 75.63  5.00 1149.9  80.0 59.3  5.5 62.8  5.9 65.2  6.2 1036.2  80.0 65.9  6.3 69.2  6.9 72.2  7.3 0.74  0.09 68.96  2.04 1149.9  80.0 54.8  1.8 57.2  1.8 58.9  1.8 1036.2  80.0 62.0  3.8 65.1  4.1 67.6  4.4 0.61  0.02 31.90  1.95 1264.9  80.0 24.0  2.0 24.6  2.2 25.1  2.3 1106.1  80.0 27.2  2.5 28.0  2.6 28.7  2.7 0.49  0.02 31.05  2.26 1264.9  80.0 25.7  2.6 26.3  2.6 26.9  2.7 1106.1  80.0 26.7  2.7 27.4  2.8 27.9  2.9 0.55  0.06 31.55  1.48 1264.9  80.0 24.7  1.6 26.4  1.7 25.9  1.8 1106.1  80.0 27.0  1.8 27.7  1.9 28.3  2.0 a Abbreviations: EU ¼ assuming early U uptake, p ¼ 1, LU ¼ assuming linear U uptake, p ¼ 0, RU ¼ assuming recent U uptake, p ¼ 10, Ages calculated using a/g factor, ka ¼ 0.10  0.01, initial U activity ratio, (234U/238U)0 ¼ 1.2  0.20, aragonite density, rmol ¼ 2.95  0.01 g/cm3, radon loss from the shells, Rnmol ¼ 0.0  0.0 vol%, carbonate sediment density, rcal ¼ 2.96  0.01 g/cm3, quartz sediment density, rqtz ¼ 2.66  0.01 g/cm3, sedimentary water concentration, Wsed ¼ 10.0  5.0 wt%, All errors are 1 s. b Uses cosmic dose rate, Dcos ðtÞ ¼ 0.293  0.020 mGy/y. c Uses Dsed ðtÞ and Dcos ðtÞ. 5. Conclusions Testing for reworked gastropod assemblages at Medauwara revealed that two of three did likely contain more than one population. For these assemblages, further geological sampling is needed to reveal the sources for the two populations before accurate ages can be calculated. At Matana, two different units were dated by building box models to calculate the time-averaged cosmic and sedimentary dose rates. The gastropods from Matana 2 averaged 27.7  1.9 ka, while those from Matana 3, had a mean LU age of 65.1  4.1 ka. Thus, during OIS 2 and 4, water was available at Matana for hominin uses. Therefore, the Khargan Complex occupations that occur atop the Libyan Plateau and on the escarpment face above Kharga Oasis depression would have had sufficient water. Acknowledgements A.C. Montoya and S.M. Baboumian assisted with some sample preparation. Jean Johnson and Alice Pedruczny, McMaster University Nuclear Reactor, did the NAA. We thank the National Science Foundation (EAR-0447357 to JRS, and ILI 9151111 to ARS), Williams College, RFK Science Research Institute, McMaster University Nuclear Reactor for logistical and financial support. Editorial handling by: R. Grun References Adelsberger, K.A., Smith, J.R., 2010. Paleolandscape and paleoenviornmental interpretations of spring deposited sediments (sic) in Dakhleh Oasis, Western Desert, Egypt. Catena 83, 7e22. Blackwell, B.A., 1989. Laboratory Procedures for ESR Dating of Tooth Enamel. McMaster University Department of Geology Technical Memo, 89.2, 234 pp. Blackwell, B.A., 1994. Problems associated with reworked teeth in electron spin resonance (ESR) dating. Quaternary Geochronology (Quaternary Science Reviews) 13, 651e660. Blackwell, B.A., 2001. Electron spin resonance (ESR) dating in lacustrine environments. In: Last, W.M., Smol, J.P. (Eds.), Tracking Environmental Change Using Lake Sediments. Basin Analysis, Coring, and Chronological Techniques, vol. 1. Kluwer, Dordrecht, pp. 283e369. Blackwell, B.A.B., 2006. Electron spin resonance (ESR) dating in karst environments. Acta Carsologica 35, 123e153. Blackwell, B.A., Schwarcz, H.P., 1995. The Uranium Series Disequilibrium Dating Methods. In: Dating Methods for Quaternary Deposits. Geological Association of Canada, St. John’s. GeoText 2: 167e208. Brennan, B.J., Rink, W.J., McGuirl, E.L., Schwarcz, H.P., 1997. Beta doses in tooth enamel by “one-group” theory and the Rosy ESR dating software. Radiation Measurements 27, 307e314. Caton-Thompson, G., 1952. Kharga Oasis in Prehistory. Ahtlone Press, London, p. 213. Deely, A.E., Blackwell, B.A.B., Mylroie, J.E., Carew, J.L., Blickstein, J.I.B., Skinner, A.R., 2011. Testing cosmic dose rate models for ESR: dating corals and molluscs on San Salvador, Bahamas. Radiation Measurements 30, 1e7. Hawkins, A.L., Smith, J.R., Giegengack, R., McDonald, M.M.A., Kleindienst, M.R., Schwarcz, H.P., Churcher, C.S., Wiseman, M.F., Nicoll, K., 2001. New research on the prehistory of the escarpment in Kharga Oasis, Egypt. Nyame Akuma 55, 8e14. Kieniewicz, J.M., 2007. Pleistocene Pluvial Lakes of the Western Desert of Egypt: Paleoclimate, Paleohydrology, and Paleolandscape Reconstruction. Ph.D. thesis, University of Washington in St. Louis, St. Louis, 271 pp. Kieniewicz, J.M., Smith, J.R., 2007. Hydrologic and climatic implications of stable isotope and minor element analyses of authigenic calcite silts and gastropod shells from a mid-Pleistocene pluvial lake, Western Desert, Egypt. Quaternary Research 68, 341e344. Kleindienst, M.R., Schwarcz, H.P., Nicoll, K., Churcher, C.S., Frizano, J., Gigengack, R.W., Wiseman, M.F., 2008. Water in the Desert: First Report on Uranium-series Dating of Caton-Thompson’s and Gardner’s "classic" Pleistocene Sequence at Refûf Pass, Kharga Oasis. In: Wiseman, M.F. (Ed.), The Oasis Papers 2: Proceedings of the Second International Conference of the Dakhleh Oasis Project. Oxbow, Oxford, pp. 25e54. Molod’kov, A., 1993. ESR dating of non-marine mollusk (sic) shells. Applied Radiation & Isotopes 44, 145e147. Nicoll, K., Giegengack, R., Kleindienst, M., 1999. Petrogenesis of artifact-bearing fossilspring tufa deposits from Kharga Oasis, Egypt. Geoarchaeology 14, 849e863. Skinner, A.F., 1988. Dating of marine aragonite by electron spin resonance. Quaternary Science Reviews 7, 461e464. Skinner, A.R., Shawl, C.E., 1994. ESR dating of terrestrial Quaternary shells. Quaternary Geochronology (Quaternary Science Reviews) 13, 679e684. Smith, J.R., Giegengack, R., Schwarcz, H.P., 2004a. Constraints on Pleistocene pluvial climates through stable isotope analysis of fossil spring tufas and associated gastropods, Kharga Oasis, Egypt. Palaeogeography, Palaeoclimatology, Palaeoecology 206, 157e175. Smith, J.R., Giegengack, R., Schwarcz, H.P., McDonald, M.M.A., Kleindienst, M.R., Hawkins, A.L., Churcher, C.S., 2004b. A reconstruction of Quaternary pluvial environments and human occupations using stratigraphy and geochronology of fossil-spring tufas, Kharga Oasis, Egypt. Geoarchaeology 19, 1e34. Smith, J.R., Hawkins, A.L., Asmerom, Y., Polyak, V., Giegengack, R., 2007. New age constraints on MSA occupations of Kharga oasis, Western desert, Egypt. Journal of Human Evolution 52, 690e701.