OSL Chronology for Sediments and MSA Artefacts at the Toteng Quarry, Kalahari Desert, Botswana

South African Archaeological Society

OSL Chronology for Sediments and MSA Artefacts at the Toteng Quarry, Kalahari Desert,BotswanaAuthor(s): George A. Brook, Pradeep Srivastava, Fong Z. Brook, Lawrence H. Robbins, Alec C.Campbell and Michael L. MurphySource: The South African Archaeological Bulletin, Vol. 63, No. 188 (Dec., 2008), pp. 151-158Published by: South African Archaeological SocietyStable URL: http://www.jstor.org/stable/20475010 .

Accessed: 16/08/2013 14:01

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

South African Archaeological Society is collaborating with JSTOR to digitize, preserve and extend access toThe South African Archaeological Bulletin.

http://www.jstor.org

This content downloaded from 152.3.102.242 on Fri, 16 Aug 2013 14:01:14 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=saas

http://www.jstor.org/stable/20475010?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


South African Archaeological Bulletin 63 (188): 151-158, 2008 151

Field and Technical Report

OSL CHRONOLOGY FOR SEDIMENTS AND MSA ARTEFACTS AT THE TOTENG QUARRY, KALAHARI DESERT, BOTSWANA

GEORGE A. BROOK1, PRADEEP SRIVASTAVA2, FONG Z. BROOK1, LAWRENCE H. ROBBINS3, ALEC C. CAMPBELL4 & MICHAEL L. MURPHY5

'Department of Geography, University of Georgia, Athens, GA 30602, USA E-mail: [email protected]

2Sedimentology Group, Wadia Institute of Himalayan Geology, 33 GMS Road, Dehradun 248001, India

3Department of Anthropology, Michigan State University, East Lansing, MI 48824, USA

4PO. Box 306, Crocodile Pools, Gaborone, Botswana

5Kalamazoo Valley Community College, 6767 West 'O'Avenue, P O. Box 4070, Kalamazoo, MI 49003-4070, USA

(Received April 2008. Revised October 2008)

INTRODUCTION Dating the Middle Stone Age (MSA) is a key problem in

southern African archaeology because this is the critical period in human evolution when fossil and genetic evidence suggests that the first anatomically modern humans appeared in Africa (McBrearty & Brooks 2000; Henshilwood & Marean 2003;

Mellars 2006; Grine et al. 2007). Currently, MSA dates range from earlier than 200 ka to as recent as c. 20 ka. Within this considerable range, stratified cave deposits show that most of the MSA is earlier than the period dateable by radiocarbon (beyond 50 ka). For this reason, a variety of other methods, including Optically Stimulated Luminescence (OSL), Thermo luminescence (TL), Electron Spin Resonance (ESR), Uranium series (U-series), and Amino Acid Racemization (AAR), have been used to date this important period of human development.

However, despite these efforts, there are still comparatively few absolute dates available for occurrences in the interior of southern Africa and only two sites with MSA dates in the Kalahari of Botswana (Robbins & Murphy 1998; Robbins et al. 2000). These sites include the pan margin site of Gi, and the Tsodilo Hills White Paintings Rock Shelter, both of which are located in northwestern Botswana (Fig. 1). This paper reports on new dates from the Toteng area in the Kalahari, even though comparatively few artefacts were recovered. We will also briefly compare the dates from Toteng with the findings at Gi and with new OSL ages for White Paintings Rock Shelter.

Toteng is at the eastern end of Lake Ngami southwest of the Okavango Delta, near the confluence of the Kunyere and Nchabe rivers (Figs 1 & 2). The Toteng area is best known for archaeological sites containing Bambata pottery and the earliest evidence of domesticated livestock in Botswana (Campbell 1992; Huffman 1994; Robbins et al. 2005). There are also Later Stone Age deposits as well as surface sites that contain earlier materials most likely belonging to the MSA and ESA. For example, in 1988, one of us (Campbell) noted ESA handaxes and MSA artefacts just to the east of where the new bridge was going to be built across the Nchabe River at Toteng. Well-made Acheulean hand axes have also been found in the bed of the Boteti River (Robbins & Murphy 1998).

Toteng 3 is an important site that is located on the upper edge of a modern quarry situated immediately behind the Toteng Primary School (Campbell 1992). The site overlooks the Nchabe River valley, which is one to two minutes' walk from the site itself. Excavations at this site, at the edge of the quarry face, have yielded Bambata pottery, LSA lithic materials, small ostrich eggshell beads, bones of fish and bones of other animals

including domestic livestock (Campbell 1992). These archaeo logical remains are found buried in up to a metre of sand that caps a well-consolidated layer of pedogenic calcrete. The underlying calcrete deposits are exposed in the wall of the quarry and any artefacts that occur either within or below the calcrete would, on stratigraphic grounds, be earlier than the Bambata and LSA materials.

During the course of our work at Toteng 3 we searched for artefacts embedded in the wall of the quarry in an attempt to extend the time span of archaeological record in the area. The quarry face can be described as a cliff exposing about 5.5 m of deposits. In July 2003, G.A. Brook noticed several stone artefacts embedded in the wall of the cliff about 0.7 m above the base and about 4.8 m from the top (Fig. 3a). This location is designated Toteng 3A because of its proximity to site 3. A close inspection revealed that there were only six widely dispersed artefacts embedded in the wall. For this reason, we decided to measure their depth and relative position and carefully remove them, rather than conduct an overall excavation through the

massive amount of cemented overburden. The six artefacts were dispersed over a horizontal distance

of 221 cm and the total vertical thickness of the artefact dispersal zone over this area was 8 cm. We did not find any soil horizon, or other distinctive feature that was associated with the artefacts, nor were there any other embedded artefacts visible at other locations along the cliff face. These few objects could have been part of a low-density artefact scatter at the edge of the ancient Nchabe River valley. However 26 additional artefacts were collected from the sloping erosion surface immediately below the vertical sediment face containing the embedded MSA artefacts (Fig. 3b). We believe that the

majority, and possibly all, of the surface finds were eroded from the cliff face and for reasons stated below, are most likely associ ated with the MSA rather than the LSA.

THE QUARRY (TOTENG 3A) SEDIMENTS The 5.5m thick sediment sequence exposed in the quarry

on the right bank of the Nchabe River (20?21.40'S, 22?57.30'E) can be divided into three main units (Fig. 4). Unit 1 at the base is bioturbated sandy silt, Unit 2 calcretized bioturbated fine sand, and Unit 3, at the top, gray bioturbated sandy silt.

Unit 1 is 1.95 m thick, yellowish in colour, and highly bioturbated with dispersed calcrete nodules up to 2 cm in diameter. Animal burrows are filled by faecal pellets and vary from inclined to sub-horizontal. The upper 75 cm is less bioturbated and laminations are clearly visible. The unit



152 South African Archaeological Bulletin 63 (188): 151-158, 2008

Tsod_o gavango\v \ Namibia Botswana

GI Maun wazilan

< Toteng /Toteng-3/| thaz Ma,kgoapelwg/adi

fPans;>o/ z Maximumextensionof Gaborone le in 20th century |

FIG. 1. Location of the Okavango River Delta, Lake Ngami, Toteng, and the Tsodilo Hills. Sites along the Nchabe River that are mentioned in the text are shown

and the extent of Lake Ngami when its surface is 920 m and 923.5 m asl is indicated.

contains dispersed stone artefacts believed to be of MSA affinity. The lithology and physical characteristics of the unit indicate sedimentation in a river floodplain environment.

Unit 2 is 200 cm of calcretized, bioturbated fine sand with a transitional contact with Unit 1. There is evidence of bioturbation by animal as well as root burrows. This unit may represent a bioturbated river channel bar.

There is a sharp contact between Units 2 and 3. Unit 3 is 155 cm of fine sandy silt with three sub units with gradational contacts. Sub-Unit 1 at the base is pinkish white, bioturbated sandy silt with dispersed calcrete nodules.

Sub-Unit 2 is grayish black fine sand, and Sub-Unit 3 is bioturbated black silt. Unit 3 does not appear to have a fluvial (river) origin and was probably deposited in an upland setting by both sheet flows and aeolian activity.

Units 1 and 2 in the Toteng Quarry face appear to represent ancient Nchabe River channel and floodplain deposits at eleva tions between 934 and 936 m asl that are overlain by Unit 3 fine silts washed onto the fluvial deposits by a combination of over land flow and aeolian action. As the fluvial deposits lie several

metres above the present channel of the Nchabe they imply higher levels in nearby Lake Ngami at the time of their deposi tion. Lake Ngami is fed by flow along the Okavango River,

which responds to rainfall in the uplands of southeastern Angola. After reaching its delta, the Okavango River may spill into the Ngami basin via the Thaoge and/or Kunyere river systems. In fact Lake Ngami filled to 923.5 m in July and August 2004, and again in the winters of 2005 and 2006, after being

totally dry from 1982-2003. In all three years, flow was along the Kunyere River while the Thaoge River remained dry. In 2005, the flow of the Kunyere River into the Lake Ngami basin was sufficient to cause flooding of the lower section of the Nchabe channel below the Toteng Quarry. In the past, the Lake Ngami basin may have filled until its elevation exceeded that of the c. 936 m threshold between the Nchabe and Boteti rivers. This would have reversed flow in the Nchabe valley directing it to the east-southeast in the direction of the Boteti River and

Makgadikgadi Pans (Shaw 1988; Thomas & Shaw 1991; Shaw et al. 2003). Substantial flow along the Kunyere and Nchabe valleys, when the quarry sediments were deposited, implies much wetter conditions than occur at the site today.

OSL DATING

METHODS OSL dating is based on the assumption that sediments are

zeroed or bleached of luminescence at the time of deposition and burial (see Aitken 1985, 1998). Bleaching of aeolian sands occurs through exposure to sunlight during transport and several recent studies of fluvial sediments have shown that they are bleached sufficiently at deposition to be dated by OSL (e.g. Colls et al. 2001; Stokes et al. 2001; Wallinga et al. 2001; Eitel et al. 2002; Heine & Heine 2002; Bourke et al. 2003; Rittenour et al. 2003; Leigh et al. 2004; Srivastava et al. 2004, 2005, 2006; Brook et al. 2006). It has been argued that some river transported sediments contain a proportion of unbleached




!*4 s :

:@t0X:-WF:~~~~~~~~-A

r~~~~~4; FI. . epemer6,204 atllteimgeshwig heKuyee ndNcab rves,an te octinsofth Tteg rcaeloicl its.MS ateacs erfonda

Toen3Aner hebaeoa lifi te otngQurr ().Thposilextntofhe93 mlaea 5.5+ .2kninreatontoth Ttegste i sow i ()

grains (e.g. Roberts et al. 1998, 1999; Wallinga et al. 2001; Wallinga 2002; Olley et al. 2004), in which case minimum OSL ages are a more appropriate estimate of true sediment age (see also Srivastava et al. 2006; Brook et al. 2006). However, one danger in using minimum ages is that they could represent more recent periods of sediment bioturbation rather than the true period of deposition. The absence of bimodality in our aliquot age distributions, the lack of prominent tails either old or young, and the lack of evidence for significant bioturbation in the study area (e.g. Robbins et al., 2008), suggests that

weighted mean ages for the Toteng (and also White Paintings) samples are the best estimates of the time of sediment deposition and burial.

To provide a chronology for the Toteng Quarry sediment sequence, three samples were collected for OSL dating by hammering an opaque plastic pipe horizontally into the vertical face. Samples were obtained from Sub-Unit 2 of Unit 3 (NCH-5), the upper section of Unit 2 (NCH-4) and the basal section of Unit l(NCH-1). In dating, a 4 mm-diameter aliquot size was used and the central aliquot age was taken as the best

estimate of the true age of the sample. Sample preparation and handling for OSL dating was carried out in the laboratory under controlled red-light conditions. Five centimetres of sediment were removed from each end of the PVC sample tubes for dose rate estimation. Luminescence measurements

were made on the central section of the sediment cylinder that was least likely to have been exposed to sunlight during sam pling.

Core sediments for OSL analysis were washed with water and then treated with 10% HCl and 30% H202 to remove carbonates and organic material. Sieving isolated the 120-150 ,m-size fraction and then density separation using Na-Polytungstate (p = 2.58 g/cm3) was used to separate quartz from feldspar minerals. The quartz fraction was etched with 48% HF for 80 minutes followed by 36% HCl for 40 minutes to remove the alpha skin. Quartz grains were mounted on stainless steel discs with the help of Silkospray?. Light stimula tion of quartz mineral extracts was undertaken with a RIS0 array of combined blue LEDs centered at 470 nm. Detection optics comprised two Hoya 2.5 mm thick U340 filters and a




FIG. 3. MSA artefacts from Toteng 3A. (A) Units 1 and 2 at the base of the Toteng Quarry face and the MSA artefact layer in Unit 1 (indicated by the arrow). (B) Sample of the MSA artefacts recovered. Artefacts in the upper row were in situ in the quarry wall. Artefacts in the middle and lower rows were surfacefindsfrom the area near the base of the quarry wall and a little downslope of it; the last specimen on the right in the lower row is a bonefragment.

3 mm thick Schott GG420 filter coupled to an EMI 9635 QA photomultiplier tube. Measurements were taken with a RISO TL-DA-15 reader. A 25-mCi 90Sr/90Y built-in source was used for sample irradiation. A thick source Daybreak alpha counting system was used to estimate U and Th for dose rate calculation. Potassium was measured by ICP90, with a detection limit of 0.01%, using the sodium peroxide fusion technique at the SGS Laboratory in Toronto, Canada. The cosmic-ray dose was estimated for each sample as a function of depth and altitude following Prescott & Hutton (1994).

The SAR protocol (Murray & Wintle 2000) was used to

determine palaeodose. Fifteen 4 mm diameter aliquots from each sample were analysed. A five-point regenerative dose strategy was adopted with three dose points to bracket the palaeodose, a fourth zero dose to test for recuperation effects, and a fifth repeat dose, usually of the smallest regenerative dose. The OSL response to the repeat dose was measured to monitor whether the sensitivity change correction incorporated in the SAR protocol was successful. All measurements were

made at 125?C for 100 seconds after a pre-heat to 220?C for 60 seconds. For all aliquots the recycling ratio between the first and fifth point ranged within 0.95-1.05. Sediment water

u NCH2-5 (4.7?O.9ka) . . ... ...... .;

2 -D NCH2-4 (26.5?5.1ka)

~~~~~~~~~~1)~~~~~~~~~~~~~~~1

_s4 D -v

. '3 v 4 a - U a

b C ~ ~ ~ ~ ~ ~ ~~LZZ nt1:Boubae anysl

Base o sectionx G . m t a t e r (

x x x x x x x x x x x x x x x x x x ......

5- , x x x x x x x x NC21(1.i.k).xUnt2Caceidbourtdfnesd

.x x x x x x x x x x x x x x x x x x Unit 3: GryBioturbated sandy silt

Base of section

FIG. 4. Sediment section at the Toteng Quarry (Toteng 3A) in the Nchabe River valley, Botswana, showing OSL ages and the location of MSA artefacts.




TABLE 1. Central OSL ages and dosimetry data for Toteng Quarry (NCH) and White Paintings rock shelter (WPS) samples.

Sample Depth (m) U (ppm) Th (ppm) K (%) Cosmic ray Dose rate Paleodose Age (ka) dose (Gy/ka) (Gy/ka) (Gy)

NCH2-5 0.25 1.95 + 0.3 2.44 ? 0.9 0.26 0.2756 1.1 ? 0.1 5.3 ? 0.9 4.7 ? 0.9

NCH2-4 1.7 1.54 ? 0.3 2.41 + 0.9 0.28 0.2032 1.0 ? 0.1 25.9 ? 4.3 26.5 ? 5.1

NCH2-1 5.0 2.53 ? 0.3 3.19 ? 1.3 0.28 0.1338 1.2 ? 0.1 61.0 ? 6.0 51.5 ? 7.2

WPSOSL-7 4.15 5.37 ? 0.8 4.71 ? 2.7 0.13 0.1518 1.8 ? 0.3 93.7 ? 7.0 51.9 ? 8.5

WPSOSL-8 4.5 6.13 ? 0.9 5.52 ? 3.0 0.12 0.1454 2.0 ? 0.3 112.6 ? 13.0 55.8 ? 10.5

WPSOSL-9 5.1 4.95 ? 0.7 3.49 ? 2.5 0.18 0.1353 1.7 ? 0.2 96.7 ? 9.1 58.4 ? 10.2

content was assumed to be 5 + 2%. Data were analysed using the ANALYST program of Duller (1999).

RESULTS Sample age and dosimetry data are presented in Table 1.

From the youngest sediment in the sequence to the oldest, OSL ages were 4.7 ? 0.9, 26.5 ? 5.1, and 51.5 ? 7.2 ka. The ages for the fluvial sediments of Units 1 and 2 indicate deposition in the period 51-26 ka and place the age of the MSA artefacts at c. 51.5 ? 7.2 ka. A more detailed chronology will be needed to determine if sedimentation was continuous between these dates or if they represent two distinct periods of river activity separated by one or more periods of drier conditions. The late Holocene age for the upper part of Unit 3 suggests increased rainfall and sheet wash at a time when the river had abandoned its high-level floodplain. Declining Lake Ngami water levels are implied. Abandonment of the floodplain level may have been due to channel avulsion, a drier climate, or diversion of Okavango Delta flow into other distributaries.

Our other studies of relict deltaic sediments at the eastern end of Lake Ngami provide additional evidence of high water levels in the lake during the late Pleistocene. At several sites fluvial sediments immediately underlie diatomites that are taken to record maximum lake depths. The contact between the fluvial sands and the overlying diatomites is usually well defined and there is no apparent mixing of the diatomite and sands as a result of bioturbation. As a result, we believe that OSL minimum ages for the underlying fluvial sands are better estimates of the ages of the diatomites than are central ages, even if these ages reflect bioturbation of the sands before they were buried by diatoms. At Sites 6 and 11 (Fig. 1) fluvial sands with minimum OSL ages of 33.5 ? 3.6 ka and 25.0 + 2.3 ka, respectively, underlie lake diatomites, suggesting maximum lake levels shortly after these dates. At Site 15, in a small tributary valley south of the Nchabe River, a diatomite overlies fluvial sands that have a minimum OSL age of 54.3 ? 7.7 ka. Lacustrine clays and then fluvial sands that have provided a central OSL age of 43.3 ? 7.3 ka overlie the diatomite. The diatomite thus records a period of high lake level soon after 54.3 ? 7.7 and the overlying fluvial sediments show that Lake Ngami no longer extended as far as Site 15 by 43.3 ? 7.3 ka. Also, at Toteng 7, fluvial sands beneath a diatomite have a minimum OSL age of 45.1 ? 5.1 ka indicating a high lake level some time after this date. Additional evidence of past lake conditions comes from Mogapelwa (Fig. 1) where an augur hole revealed a diatomite underlain by fluvial sands with a minimum OSL age of 37.7 ? 4.4 ka and overlain by sands with a central OSL age of 28.7 ? 2.7 ka. Thus, the diatomite records high lake levels in the period c. 38-29 ka. Together, the data from these five sites at the eastern end of Lake Ngami suggest

wetter climates and high lake levels during two periods from 54-43 ka and from 38-25 ka. Evidence from several other sites along the Nchabe River indicates increased flow into the basin

and higher lake levels in the period 6.1-3.0 ka and at c. 1.7 ka (Brook, unpublished data).

Data from a sediment core taken from the centre of Lake Ngami at 922 m, very close to its sump at 921 m, agree well with our record from deposits at the eastern end of the lake basin. A diatomite at the base of the core (4.6 m) indicates a deeper lake at 40.5 ? 3.2 ka (TL age) followed by low lake levels at 35.6 + 3.8 ka (TL age), indicated by fluvial sands (Huntsman-Mapila et al. 2006). High lake levels are also recorded in the Makgadik gadi sub-basin by a TL age of 43.4 ? 4.4 ka (Ringrose et al. 2005). Above the sands in the Lake Ngami core is a diatomite that is overlain by organic-rich sediment with a calibrated radiocar bon age of 19.3 ? 0.7 ka indicating a shallow wetland near the lowest point in the lake basin around the time of the Last Glacial

Maximum (LGM). The underlying diatomite records a deeper lake before the LGM and an overlying diatomite a deeper lake shortly after the LGM. The upper diatomite is followed by dia tom-rich silt indicating declining lake levels by 16.6 ? 0.6 ka continuing to the early Holocene. A younger diatomite records another lacustrine episode at c. 5 ka (Huntsman-Mapila et al. 2006).

Thus, data from the centre and eastern end of the Lake Ngami basin, indicating an expanded and deeper Lake Ngami at 54-43, 38-25, 6-3, and c. 2 ka, support evidence from the Toteng Quarry of active stream flow into a deeper lake at 51.5 + 7.2 ka and 26.5 ? 5.1 ka. Deep incision of the fluvial sediments after c. 25 ka implies periodic streamflow into Lake Ngami when its elevation was less than 934 m. Clearly an extensive Lake Ngami at 54-43 ka, with all of the varied resources it would provide, would have been a magnet for humans, providing an explanation for the presence of MSA artefacts at Toteng Quarry at 51.5 ? 7.2 ka.

ARTEFACTS Table 2 presents a list of the recovered artefacts and Fig

ure 3b shows a selection of the artefacts recovered from the Toteng 3A site. The first six specimens in the table (above the line) were embedded in the deposits and the rest were the surface finds. While no retouched tools were found embedded in the quarry wall, the multiplatform core is identical to MSA cores found elsewhere in Botswana. The surface finds, which we consider as a representative collection, display several MSA features including another prepared core (number 7 in Table 2), a discoidal core (8), a large flake with a facetted striking plat form (20) and the large flake blade (23). We might also add the pointed scraper (28) and the pointed flake with denticulate retouch (32).

In addition to these few MSA typological characteristics, the artefacts from the quarry wall as well as the surface finds, are significantly larger than the typical LSA microlithic materials that were excavated in the sand deposits that cap the calcrete. For example at Toteng 3, Square 3 (a one by one metre unit), 105 cm of sand above the calcrete was excavated in 5 cm




TABLE 2. Artefacts from Toteng 3A.

1. MSA multi platform core S 4.1 x 3.6 x 2.1

2. Flake, no platform (early reduction stage) S 3.4 x 2.5 x 1.2

3. Flake S 1.4 x 0.9 x 0.2

4. Flake, no platform S 2.4 x 1.1 x 0.6

5. Flake fragment S 1.4 x 1.5 x 0.5

6. Flakefragment Q 2.5 x 2.4 x 0.75 7. MSA multi platform, prepared core S 4.4 x 4.4 x 4.2

8. Discoidal core made on flake S 4.4 x 4.8 x 1.5

9. Flat core S 3.4 x 3.1 x 1.3

10. Broken core/core scraper S 4.4 x 2.8 x 2.6

11. Highly reduced core S 2.8 x 2.1 x 2.1

12. Core/scraper C 3.4 x 3.1 x 1.3

13. Core, weathered C 3.5 x 2.0 x 1.6

14. Broken core Q 2.5 x 2.0 x 1.2

15. Chunk Q 3.0 x 1.9 x 1.5

16. Chunk Q 2.8 x 2.6 x 0.8

17. Chunk Q 3.65 x 2.9 x 1.3

18. Chunk Q 2.2 x 0.8 x 0.7

19. Broken cobble with 2 grinding facets S 4.4 x 3.3 x 2.7

20. Flake with facetted striking platform S 3.2 x 2.8 x 1.25

21. Flake, weathered C 1.7 x 1.7 x 0.5

22. Flake, battered (early removal from core) S 3.0 x 3.3 x 1.3

23. Flake-blade with cortex S 5.8 x 2.6 x 1.0

24. Flake fragment 0 2.3 x 2.6 x 0.9

25. Flake fragment S 3.0 x 1.3 x 0.7

26. Flake fragment C 2.1 x 0.7 x 0.7

27. Bladelet removed from edge of core C 2.1 x 1.0 x 0.6

28. Pointed side scraper S 3.4 x 1.9 x 1.1

29. End Scraper S 3.0 x 4.2 x 0.7

30. Flake with possible weathered retouch C 2.8 x 1.5 x 0.6

31. Retouched flake-Awl projection on end S 2.8 x 2.3 x 0.3

32. Pointed flake with denticulate retouch C 3.7 x 2.0 x 0.35

Key: the raw material symbols are: S = silcrete (grey, brown, green/yellow and white), C = chert (brown, red, honey-coloured, multi-coloured), Q = quartz, 0 = other rock. Measurements are in cm. On flakes, the length was measured at a right angle from the striking platform to the end of the flake.

arbitrary levels. The mean length of 61 flakes, bladelets and tools found in these deposits (surface to 105 cm) is 1.5 cm compared to a mean 3.2 cm for 32 artefacts from site 3A. Since cores were not found in the Toteng 3, Square 3 assemblage and chunks were not measured, we can eliminate them from the 3A sample to produce a more meaningful comparison. This reduces the mean length for 3A to 2.9 cm (n = 19) but this reduc

tion still reveals a clear difference in size. In addition to the distinction in size, there is also a difference in the percentage of raw material types used. The measured sample from Square 3 (in the sand that caps the calcrete) consisted of 31% silcrete, 67% chert and 2% quartz, whereas the 3A sample consists of 56% silcrete, 22% chert, 19% quartz and 1% other. While the artefact sample from 3A is small and one cannot rule out bias in the raw material sample, the increased use of silcrete and the larger size of artefacts at site 3A are consistent with findings in the MSA levels of White Paintings Shelter, as well as at the MSA surface locality of Kudiakam Pan (Robbins 1987).

DISCUSSION Chronologies have been developed for MSA deposits at

several sites in South Africa, Namibia, and Botswana; some of the most important are listed in Table 3. At Wonderwerk Cave in the Northern Cape of South Africa U-series ages range from > 220 to 70 ka (Beaumont & Vogel 2006) while at Border Cave in KwaZulu-Natal, South Africa, a revised ESR chronology indi cates ages from 227-41 ka (Grun & Beaumont 2001). AAR ages for ostrich eggshell, as well as conventional radiocarbon ages for the Border Cave sediments, show a transition from MSA to early LSA at about 47-40 ka (Beaumont et al. 1978; Beaumont 1980; Miller et al. 1999). Burnt lithics in MSA deposits at Blombos Cave, 300 kilometres east of Cape Town, have pro vided TL ages of 105-67 ka (Tribolo et al. 2006), while OSL ages on MSA sediments indicate deposition from 143 to c. 70 ka (Jacobs et al. 2006). MSA levels at Klasies River main site on the south coast of South Africa indicate occupation from 115-50 ka (Feathers 2002). TL ages on burnt lithics from MSA layers at Rose Cottage Cave near Ladybrand, in the eastern Free State of South Africa, range from 76-42 ka while OSL ages for sediment range from 86-33 ka (Valladas et al. 2005). At Apollo 11 Cave in southern Namibia, MSA deposits date to ?83 ka and the transi tion from MSA to early LSA occurred at 41 ka based on radio carbon dating of ostrich eggshell, although radiocarbon ages in

Wendt (1976) show this transition between 26-20 ka (Miller et al. 1999). MSA deposits at Die Kelders cave on the South African coast have yielded OSL ages of 70-60 ka (Feathers & Bush 2000). At Sibudu Cave about 40 km north of Durban post-Howieson's Poort, late and final MSA deposits have OSL

TABLE 3. Ages for MSA sediments at selected sites in Southern Africa.

Site* * Dating method Material dated Age range (ka) Reference

Wonderwerk Cave (SA) U-series Speleothem 220-70 1

Sibudu Cave (SA) OSL Sediment c. 59-37 2

Rose Cottage Cave (SA) TL Burnt Lithic 76-42 3 OSL Sediment 86-33

Blombos Cave (SA) TL Burnt Lithic 105-67 4 OSL Sediment 143-70

Die Kelders Cave (SA) OSL Sediment 70-60 5

Border Cave (SA) ESR Tooth Enamel 227-41 6

Klasies River Cave (SA) OSL Sediment 115-50 7

Boomplaas Cave (SA) 14C OES*/Organic >49-18 8 AAR OES* 65-18

Apollo 11 Cave (NAM) 14C OES*/Organic >49-26 9 AAR OES* ?83-41

Toteng Quarry (BOT) OSL Sediment c. 51 10

White Paintings Shelter (BOT) OSL Sediment 61-52 11

Gi (BOT) AAR OES* 85-65 ka 12

1 = Beaumont & Vogel (2006); 2 = Jacobs et al. (2008); 3 = Valladas et al. (2005); 4 = Jacobs et al. (2003a,b; 2006), Tribolo et al. (2006); 5 = Feathers & Bush (2000); 6 = Grun & Beaumont (2001); 7 = Feathers (2002); 8 = Miller et al. (1999); 9 = Wendt (1976), Miller et al. (1999); 10 = this paper; 11 = Brook (unpublished data); 12 = Brooks et al. (1990).

*OES = Ostrich eggshell **SA = South Africa; NAM = Namibia; BOT = Botswana.




ages of 58.5 ? 1.4, 47.7 ? 1.4 and 38.6 ? 1.9 ka, respectively (Jacobs et al. 2008). The MSA to LSA transition at Boomplaas Cave, near the southern tip of South Africa, occurs at 18 ka based on radiocarbon dates for ostrich eggshell and other materials with AAR ages on ostrich eggshell indicating MSA occupation back to at least 65 ka (Deacon et al. 1984; Miller et al. 1999; Vogel 2001). Another site to mention is Pinnacle Point Cave (PP13B) located near Mossel Bay, South Africa with OSL ages of 164 ? 12 ka (Lower MSA), 132 ? 12 ka (Middle MSA) and 120 ? 7 ka (Upper MSA) (Marean et al. 2007).

Sediments exposed at the Toteng Quarry were deposited in a fluvial floodplain environment at a time when flow through the Okavango River system and Delta was greater than today. As the quarry sediments are several metres above the present river channel, and appear to reach 934-936 m asl, they imply a higher level of water in Lake Ngami and more regular discharge along the Kunyere and Nchabe rivers at 51.5 ? 7.2 and 26.5 ? 5.1 ka. The elevation of the floodplain sediments suggests an extensive lake at 936 m at least seasonally, and flow to the northeast along the Nchabe River towards the Boteti rather than southwest towards Lake Ngami, which occurs at lake elevations of less than 936 m (Shaw 1988).

The MSA artefacts were found near the base of Unit 1 of the quarry sediments, which provided an OSL age of 51.5 ? 7.2. This chronology fits well with data for the MSA from White Paintings Shelter in the Tsodilo Hills (Murphy 1999; Robbins et al. 2000). At White Paintings MSA artefacts were encountered first at 4.1 to 4.2 m depth and thereafter to about 7 m depth.

New OSL ages for White Paintings' sediments at 4.15, 4.5, and 5.1 m suggest deposition of sands coeval with MSA artefacts at 51.9 ? 8.5,55.8 ? 10.5, and 58.4 ? 10.2 ka, respectively (Tables 1 and 3).

Beginning at c. 3.0 m at White Paintings, there is a noticeable shift from a microlithic technology to the use of large blades struck from prepared cores (Robbins et al. 2000). The only comparable blade industry in the Kalahari that has been dated is the intermediate industry 2C at Gi Pan (Fig. 1), which dates to about 34 ka (Brooks et al. 1990). Using protein diagenesis of ostrich eggshell, Brooks et al. (1990) estimate their Unit 4, the MSA unit at Gi, to be 65-85 ka.

As Table 3 shows, the MSA at major sites in southern Africa dates in the range 227-18 ka with terminal ages for MSA deposits varying from 70-18 ka. At Wonderwerk Cave, just beyond the southern margin of the Kalahari Desert, the upper boundary of the MSA is 70 ka; within the Kalahari proper it is 65 ka at Gi and 52 ka at White Paintings rock shelter. Therefore, the 51 ka age for the MSA artefacts at Toteng Quarry is close to the youngest ages for other sites within the Kalahari or near its

margin, and in good agreement with MSA ages for sites else where on the subcontinent. However, so far, the upper ages for the MSA in the Kalahari (70-51 ka) are much older than the youngest MSA deposits around the South African coast (c. 20 ka) begging the question as to how long the MSA per sisted in this region. Clearly, more research will be needed to shed light on this important issue.

CONCLUSION The Kalahari has figured prominently in the discussion of

the origin of modern humans in Africa, yet there has been little dating evidence available, especially in regard to the ending, or latter part of the MSA. OSL dates for sediments at the Toteng Quarry, and at White Paintings Rock Shelter suggest that the MSA lasted until at least 51 ka in the Kalahari Desert of Botswana. This estimate is not inconsistent with the transition from MSA to an LSA blade industry at Gi dated to c. 34 ka and a youngest

age for MSA Unit 4 of 65 ka (Brooks et al. 1990). At White Paint ings Shelter the MSA is also followed by a large blade industry (Murphy 1999; Robbins et al. 2000). Further work may refine the chronology of the blade industry and enable us to confirm whether or not the transition to blades is, in fact, part of a broad pattern in the area. The oldest ages for MSA strata at White Paintings and Gi indicate MSA peoples in the Kalahari definitely by 61 ka and probably much earlier given the ostrich eggshell protein diagenesis age of 85 ka for the MSA at Gi and the TL age of c. 94 ka for 6.05 m at White Paintings Shelter (Brooks et al. 1990; Robbins et al. 2000). The ages of floodplain sediments in the Ngami basin at Toteng Quarry and at several other sites in the centre and at the eastern end of the lake basin, indicate that a large lake (at least seasonally) with its shoreline at 936 m asl and a maximum depth of 19 m, filled the Ngami basin at 54-43 ka and again at 38-25 ka. The existence of such a large lake that almost certainly offered an abundance of hunt ing, gathering and possibly fishing opportunities at 51.5 ? 7.2 ka

may explain why MSA peoples were in the area at this time.

ACKNOWLEDGEMENTS This research was funded by National Science Foundation

grant BCS-0313826. The Office of the President in Botswana granted research clearance. We are grateful to our colleagues from the Botswana National Museum and the University of Botswana who participated in the fieldwork at Toteng. We are especially grateful to Troy Shick and we also thank Alison Brooks for collecting OSL samples for us from White Paintings Rock Shelter in 2004.

REFERENCES Aitken,M.J. 1985. Thermoluminescence Dating. London: Academic Press.

Aitken, M.J. 1998. An Introduction to Optical Dating. Oxford: Oxford

University Press.

Beaumont, PB. 1980. On the age of Border Cave hominids 1-5.

Palaeontologia africana 23: 21-33.

Beaumont, PB., de Villiers, H. & Vogel, J.C. 1978. Modern man in

sub-Saharan Africa prior to 49,000 B.P: a review and evaluation with

particular reference to Border Cave. South African Journal of Science 74: 409-419.

Beaumont, PB. & Vogel, J.C. 2006. On a timescale for the past million

years of human history in central South Africa. South African Journal of Science 102: 217-228.

Bourke, M.C, Child, A. & Stokes, S. 2003. Optical age estimates for

hyper-arid fluvial deposits at Homeb, Namibia. Quaternary Science Reviews 22:1099-1103.

Brook, G.A., Srivastava, P. & Marais, E. 2006. Characteristics and OSL minimum ages of relict fluvial deposits near Sossus Vlei, Tsauchab

River, Namibia, and a regional climate record for the last 30 ka. Journal

of Quaternary Science 21: 347-362.

Brooks, A.S., Hare, P.E., Kokis, J.E., Miller, G.H., Ernst, R.D. & Wendorf, F. 1990. Dating archaeological sites by protein diagenesis in ostrich

eggshell. Science 248: 60-64.

Campbell, A.C. 1992. Southern Okavango Integrated Water Development Study: Archaeological Survey of Proposed Maun Reservoir. Gaborone:

Department of Water Affairs.

Colls, A.E., Stokes, S., Blum, M.D. & Straffin, E. 2001. Age limits on the

late Quaternary evolution of the upper Loire River. Quaternary Science Reviews 20: 743-750.

Deacon, H.J., Deacon, J., Scholtz, A., Thackeray, J.E, Brink, J.S. & Vogel, J.C. 1984. Correlation of palaeoenvironmental data from the Late

Pleistocene and Holocene deposits at Boomplaas Cave, southern

Cape. In: Vogel, J.C. (ed.) Late Cainozoic Palaeoclimates of the Southern

Hemisphere: 339-351. Rotterdam: Balkema.

Duller, G.A.T 1999. Luminescence Analyst Computer Program.

Aberystwyth: University of Wales.

Eitel, B., Blumel, WD. & Huser, K. 2002. Environmental transition between 22 ka and 8 ka in monsoonally influenced Namibia - a

preliminary chronology. Zeitschrift f?r Geomorphologie 126: 31-57.




Feathers, J.K. 2002. Luminescence dating in less than ideal conditions: case studies from Klasies River main site and Duinefontein, South Africa. Journal of Archaeological Science 29:177-194.

Feathers, J.K. & Bush, D.A. 2000. Luminescence dating of Middle Stone

Age deposits at Die Kelders. Journal of Human Evolution 38: 91-119.

Grine, F.E., Bailey, R.M., Harvati, K., Nathan, R.P, Morris, A.G., Henderson, G.M., Ribot, I. & Pike, A.W.G. 2007. Late Pleistocene human skull from Hofmeyr, South Africa, and modern human

origins. Science 315: 226-229.

Gr?n, R. & Beaumont, P. 2001. Border Cave revisited: a revised ESR

chronology. Journal of Human Evolution 40: 467-482.

Heine, K. & Heine, J.T. 2002. A paleohydrologic reinterpretation of

Homeb Silts, Kuiseb River, central Namib Desert (Namibia) and

paleoclimatic implications. Catena 48:107-130.

Henshilwood, CS. & Marean, GW. 2003. The origin of modern human

behavior: A review and critique of models and test implications. Current Anthropology 44: 627-651.

Huffman, T.N. 1994. Toteng pottery and the origins of Bambata. South ern African Field Archaeology 3: 3-9.

Huntsman-Mapila, P., Ringrose, S., Mackay, A.W, Downey, WS., Modisi, M., Coetze, S.H., Tiercelin, J-J., Kampunzu, A.B. &

Vanderpost, C. 2006. Use of the geochemical and biological sedimen

tary record in establishing palaeo-environments and climate change in the Lake Ngami basin, NW Botswana. Quaternary International 148:

51-64.

Jacobs, Z., Wintle, A.G. & Duller, G.A.T 2003a. Optical dating of dune sand from Biombos Cave, South Africa : I - multiple grain data.

Journal of Human Evolution 44: 519-612.

Jacobs, Z., Duller, G.A.T. & Wintle, A.G. 2003b. Optical dating of dune

sand from Biombos Cave, South Africa: II - single grain data. Journal

of Human Evolution 44: 613-625.

Jacobs, Z., Duller, G.A.T, Wintle, A.G. & Henshilwood, CS. 2006.

Extending the chronology of deposits at Biombos Cave, South Africa, back to 140 ka using optical dating of single and multiple grains of

quartz. Journal of Human Evolution 51: 255-273.

Jacobs, Z., Wintle, A.G., Duller, G.A.T, Roberts, R.G. & Wadley, L. 2008.

New ages for the post-Howiesons Poort, late and final Middle Stone

Age at Sibudu, South Africa. Journal of Archaeological Science 35:

1790-1807.

Leigh, D.S., Srivastava, P. & Brook, G.A. 2004. Late Pleistocene braided

rivers on the Atlantic Coastal Plain, USA. Quaternary Science Reviews

23: 65-84.

Marean, CW, Bar-Matthews, M., Bernatchez, J., Fisher, E., Goldberg, P.,

Herries, A.I.R., Jacobs, Z., Jerardino, A., Karkanas, P., Minichillo, T,

N?ssen, P.J., Thompson, E., Watts, I & Williams, H.M. 2007. Early human use of marine resources and pigment in South Africa during the Middle Pleistocene. Nature 449: 905-908.

McBrearty, S. & Brooks, A.S. 2000. The revolution that wasn't: a new

interpretation of the origin of modern human behavior. Journal of Human Evolution 39: 453-563.

Mellars, P. 2006. Why did modern human populations disperse from Africa ca. 60 000 years ago? A new model. Proceedings of the National

Academy of Sciences 103: 9381-9386.

Miller, G.H., Beaumont, PB., Deacon, H.J., Brooks, A.S., Hare, P.E. & Jull, A.J.T. 1999. Earliest modern humans in southern Africa dated by isoleucine epimerization in ostrich eggshell. Quaternary Science Reviews 18: 1537-1548.

Murray, A.S. & Wintle, A.G. 2000. Luminescence dating of quartz using an improved single aliquot regenerative-dose protocol. Radiation

Measurement 32: 57-73.

Murphy, M.L 1999. Changing human behavior: the contribution of the

White Paintings Rock Shelter to an understanding of changing lithic

reduction, raw material exchange, and hunter-gatherer mobility in

the interior regions of southern Africa during the Middle and Early Late Stone Age. Unpublished PhD dissertation. East Lansing: Michi

gan State University.

Olley, J.M., Pietsch, T & Roberts, R.G. 2004. Optical dating of Holocene sediments from a variety of geomorphic settings using single grains of quartz. Geomorphology 60: 337-358.

Prescott, J.R. & Hutton, T 1994. Cosmic ray contributions to dose rates

for luminescence and ESR dating: large depths and long-term time

variations. Radiation Measurements 23: 497-500.

Ringrose, S., Huntsman-Mapila, P., Kampunzu, A.B., Downey, WS.,

Coetze, S., Vink, B., Matheson, W & Vanderpost, G 2005. Sedimento

logical and geochemical evidence for palaeo-environmental change

in the Makgadikgadi subbasin, in relation to the MOZ rift depres sion, Botswana. Palaeogeography, Palaeoclimatology, Palaeoecology 217:

265-287.

Rittenour, T.M., Goble, R.J. & Blum, M.D. 2003. An optical age chronology of Late Pleistocene fluvial deposits in the northern lower Mississippi valley. Quaternary Science Reviews 22:1105-1110.

Robbins, L.H. 1987. Stone age archaeology in the northern Kalahari, Botswana, Savuti and Kudiakam Pan. Current Anthropology 28: 567-569.

Robbins, L.H. & Murphy, M.L. 1998. The Early and Middle Stone Age. In: Lane, P., Segobye, A. & Reid, D.A.M. (eds) Ditswa Mmung: the

Archaeology of Botswana: 50-64. Gaborone: Botswana Society & Pula Press.

Robbins, L.H., Murphy, M.L., Brook, G.A., Ivester, A.H., Campbell, A.C., Klein, R.G., Milo, R.G., Stewart, K.M., Downey, WS. & Stevens, N.J. 2000. Archaeology, paleoenvironment, and chronology of the Tsodilo Hills White Paintings rock shelter, northwest Kalahari Desert, Botswana. Journal of Archaeological Science 27:1085-1113.

Robbins, L.H., Campbell, A.C., Murphy, M.L., Brook, G.A., Srivastava, P. & Badenhorst, S. 2005. The advent of herding in southern Africa:

early AMS dates on domestic livestock from the Kalahari Desert, Botswana. Current Anthropology 46: 671-677.

Robbins, L.H., Campbell, A.C., Murphy, M.L., Brook, G.A., Liang, E,

Skaggs, S.A., Srivastava, P., Mabuse, A.A., Badenhorst, S. 2008. Recent archaeological and paleoenvironmental research at Toteng, Botswana: early domesticated livestock in the Kalahari. Journal of African Archaeology 6(1): 131-149.

Roberts, R., Bird, M., Olley, J., Galbraith, R., Lawson, E., Laslett, G., Yoshida, H., Jones, R., Fullagar, R., Jacobsen, G. & Hua, Q. 1998.

Optical and radiocarbon dating at Jinmium rock shelter in northern Australia. Nature 393: 358-362.

Roberts, R.G., Galbraith, R.F., Olley, J.M., Yoshida, H. & Laslett, G.M. 1999. Optical dating of single and multiple grains of quartz from

Jinmium rock shelter, northern, Australia: Part II, results and implica tions. Archaeometry 41: 365-395.

Shaw, P.A. 1988. After the flood: the fluvio-lacustrine landforms of

northern Botswana. Earth Science Reviews 25: 449^456.

Shaw, P.A., Bateman, M.D., Thomas, D.S.G. & Davies, F. 2003. Holocene

fluctuations of Lake Ngami, Middle Kalahari: chronology and

responses to climatic change. Quaternary International 111: 23-35.

Srivastava, P., Brook, G.A. & Marais, E. 2004. A record of fluvial

aggradation in the northern Namib Desert during the Late Quater

nary. Zeitschrift f?r Geomorphologie N.F., Suppl., Vol. 133:1-18.

Srivastava, P., Brook, G.A. & Marais, E. 2005. Depositional environment and luminescence chronology of the Hoarusib River Clay Castles

sediments, northern Namib Desert, Namibia. Catena 59:187-204.

Srivastava, P., Brook, G.A., Marais, E., Morthekai, P. & Singhvi, A.K.

2006. Depositional environment and OSL chronology of the Homeb Silt Deposits, Kuiseb River, Namibia. Quaternary Research 65:478^91.

Stokes, S., Bray, H.E. & Blum, M.D. 2001. Optical resetting in large drainage basins: tests of zeroing assumptions using single-aliquot procedures. Quaternary Science Reviews 20: 879-885.

Thomas, D.S.G. & Shaw, PA. 1991. The Kalahari Environment. Cambridge: Cambridge University Press.

Tribolo, C, Mercier, N., Selo, M., Valladas, H., Joron, J-L., Reyss, J-L., Henshilwood, C, Sealy, J. & Yates, R. 2006. TL dating of burnt lithics from Biombos Cave (South Africa): further evidence for the antiquity of modern human behaviour. Archaeometry 48: 341-357.

Valladas, H., Wadley, L., Mercier, N., Froget, L, Tribolo, C, Reyss, J-L. &

Joron, J-L. 2005. Thermoluminescence dating on burnt lithics from

Middle Stone Age layers at Rose Cottage Cave. South African Journal of Science 101:169-174.

Vogel, J.C. 2001. Radiometrie dates for the Middle Stone Age in South

Africa. In: Tobias, P.V, Raith, M.A., Moggi-Cecchi, J. & Doyle, G.A.

(eds) Humanity from African Naissance to coming Millennia: 261-268.

Florence: Firenza University Press.

Wallinga, J. 2002. Optically stimulated luminescence dating of fluvial

deposits: a review. Boreas 31: 303-322.

Wallinga, J., Duller, G.A.T., Murray, A.S. & T?rnqvist, TE. 2001. Testing optically stimulated luminescence dating of sand-sized quartz and

feldspar from fluvial deposits. Earth and Planetary Science Letters 193: 617-630.

Wendt, WE. 1976. Art mobilier' from the Apollo 11 cave, South West Africa: Africa's oldest dated works of art. South African Archaeological Bulletin 31: 5-11.



Documents

OSL Chronology for Sediments and MSA Artefacts at the Toteng Quarry, Kalahari Desert, Botswana