Research Article Olduvai fossil isotopic ecology Page 1 of 14

Isotopic ecology of fossil fauna from Olduvai Gorge AUTHOR: at ca 1.8 Ma, compared with modern fauna Nikolaas J. van der Merwe'

AFFILIATION: Light stable isotope ratios (ô'^C and 5"0) of tooth enamel have been widely used to defermine the diets 'Department ot Archaeology, and water sources of fossil fauna. The carbon isotope ratios indicate whether the plants at the base of the University ot Cape Town, Cape food web used Cj or C^ photosynthetic pathways, while the oxygen isotope ratios indicate the composition Town, South Atrica ot the local rainfall and whether the drank water or obtained it from plants. The contrasting diets of two early hominin species - Homo habilis and Paranthropus boisei - of ca 1.8 Ma (million years ago) in CORRESPONDENCE TO: were defermined by means of stable carbon isotope analysis of their tooth enamel in a previous Nikolaas van der Merwe study. The diets of two specimens ot R boisei, from Olduvai and Peninj, proved to be particularly unusual, EMAIL: because 80% of their carbon was derived trom C^ plants. It was suggested that their diet consisted primarily nikolaas.vandermerwe(g>uct. ot plants, with particular emphasis on papyrus, a C^ sedge. The dominance ot C^ plants in the diet of P. boisei ac.za is a finding supported in another study of 22 specimens from . The isotopic ecology and diets of fossil fauna that were present at the same time as the two fossil hominin species are described here, in order to POSTAL ADORESS: provide a fuller understanding of their contrasting diets and of the moisture sources of their water intake. Department ot Archaeology, This informafion was then compared with the isotopic composition of modern fauna from the same region University ot Cape Town, of Tanzania. The carbon isotope ratios tor both fossil and modern specimens show that the habitats in which Private Bag, Rondebosch 7701, South Atrica these faunal populations lived were quite similar - grassland or wooded grassland. They had enough bushes and frees to support a few species of browsers, but most of the animals were grazers or mixed feeders. DATES: The oxygen isotope ratios of the fossil and modern tauna were, however, very different, suggesting strongly Received: 12 Apr 2013 that the source of moisture for the rain in the Oiduvai region has changed during the past 1,8 million years. Revised: 24 Jul. 2013 Accepted: 22 Aug. 2013 Introduction KEYWORDS: The purpose of this article is to provide isotopic information on the ecology and diets of fossil tauna trom GIduvai carbon isotopes: oxygen Gorge Beds I and II at ca 1.8 Ma, in order to provide an ecological context for the diets of two species of early isotopes: Oiduvai tossils: hominins. Recent isotopic analyses of the tooth enamel ot Homo habilis and Paranthropus boisei (formerly moisture sources: Zinjanthropus boisei, also Australopithecus boisei) showed that the two species had distinctly different diets.'-^ vertebrate ecology Three specimens of Homo habilis from Olduvai had, respectively, 23%, 27% and 49% of carbon derived from C^ HOW TO CITE: plants, much like early hominins trom ": their diets probably included grass-eating animals and/or Van der Merwe NJ. Isotopic insects. However, two specimens ot P. boisei trom Olduvai and Peninj had 77% and 81 % of carbon derived from C^ ecology ot tossil tauna trom plants. Because modem are limited to about 20-50% protein-rich toods tor their energy requirements,' it Olduvai Gorge at ca 1.8 Ma, was suggested that the diet o\P boisei included a large component ot C^ plants. As grasses, especially edible seeds, compared with modern tauna. are highly seasonal at latitudes applicable to Olduvai, it was also suggested that the C, sedge Cyperus papyrus, S AtrJ Sei. 2013:109(11/12), which was presumably available in the treshwater swamps at Lake Olduvai and Lake Natron (Peninj), may have Art. #2013-0105,14 pages. been a major component in the diet ot P boisei. The very high contribution ot C, plants to the diet ot this hominin http://dx.doi.org/10.1590/ has been confirmed by tbe isotopic analysis ot 22 individuals trom sites in Kenya that stretch over 700 km of the sajs.2013/20130105 Ritt Valley.^ The authors of the latter study suggested that P boisei had a diet that comprised mostly C, plants, without specitying whether these were grasses or sedges. An online comment on their article by Lee-Tborp^ supports the opinion that sedges were predominant in the diet. Recent publications provide turtber evidence to put these early hominin diets in context.^'" An assessment of the isotopic ecology at Peninj and Olduvai during the presence of P boisei cannot settle the argument about grasses and sedges in P boisei's diet, but it can illuminate whether the environment was dominated by C^ plants and their consumers. This assessment is ot particular importance at Olduvai around 1.8 Ma, for which the isotope values of the two hominins are available. In East Africa, C^ plants appeared in the Late Miocene and continued through the Pliocene,"'^ but did not become dominant until ca 1.8 Ma."' At Olduvai, the carbon isotopes in palaeosol carbonates indicate that C^ plants made up about 40-60% of tbe biomass during the time of Beds I and II," but a preliminary study ot the carbon isotopes in tooth enamel ot fossil tauna indicate a much higher C^ component.'^"' An assessment ot the isotopic ecology at Peninj was undertaken by measuring the carbon and oxygen isotopes in the tooth enamel ot 40 specimens ot fossil fauna from the Maritinane Type Section.^ These fossils were all grazers and the carbon isotopes were closely similar to those ot modern fauna of related species from the modern Serengeti. The environment was essentially open grassland with very tew trees. However, the fossil specimens were some 1.3 million years old, so the results do not provide information about the ecology of Olduvai at ca 1.8 Ma. It was noted that the S'^O values ot the Peninj tossils were distinctly negative, relative to the Vienna Pee Dee Belemnite (VPDB) standard, while those ot the modern fauna were all positive. It was suggested that the moisture source ot rain in this region had presumably changed during the past 1.3 million years: this suggestion is addressed in detail here. For this purpose, tooth enamel samples were obtained from 145 specimens of tossil fauna from Olduvai Gorge Middle Bed I (ca 1.785-1.83 Ma) and Lowermost Bed II (ca 1.75-1.83 Ma) and their 5'^C and 5"0 values were ©2013. The Authors. measured. These values were then compared with those of 77 modern animals from six wildlife reserves, primarily Published under a Creative Serengeti and Maswa, but also Lobo, Lukwati, Ugalla and Selous (Figure 1 ). Their carbon and oxygen isotope ratios Commons Attribution Licence. were measured to illuminate their diets and water sources.

Soutii African Journal of Science Volume 109 \ Number 11/12 http://www.sajs.co.za 64 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 2 of 14

orongoro Conservation area Olduvai-Gorge

Figure 1 : Map of Tanzania showing the six wildlife reserves from which modem faunal specimens were coiiected, as well as Olduvai and Peninj.

Tooth enamel and stable light isotopes earliest measurements on bone apatite were done on tossil bone,^" ^^ but The assessment of prehistoric diets and environments by means ot tooth enamel has proved to be the most dependable material tor isotopic isotopic analysis of bone has been developed over the past 30 years and analysis.^f^" Tooth enamel is highly crystalline and resists aiteration by is wideiy used in archaeoiogy"*-^' (for review see Van der Merwe^'). in carbonates in groundwater Contaminating carbonates may precipitate in the early development ot this method, the major interest involved "C/"C cracks in the tooth enamel, but this materiai is much more soluble than ratios (5"C values) in bone collagen, which provide a measure ot the tooth enamel and is readily removed with dilute acetic acid. proportion ot C3 and C^ plants at the base ot the tood web. However, collagen in bone has a iimited litetime, especially in hot and humid Tooth enamel is a biologicai apatite with about 3% carbonate. With environments where organic materiais deteriorate rapidly. The oldest appropriate pretreatment, the carbon and oxygen isotope ratios ot hominin collagen specimens that have been anaiysed isotopically were the carbonate can be measured with confidence to provide dietary those of Neanderthals trom cold, dry .^^'^^ The tossiiised taunal intormation. The carbon isotope ratio is an average ot the plants at the specimens trom Oiduvai contain no collagen, but carbon and oxygen base ot the tood web, acquired by herbivores trom eating the plants and isotopes can be measured in the mineral phase of their skeletons. The passed along the tood chain to omnivores and carnivores, with some

South African Journai of Science Volume 109 ¡ Number 11/12 http://www.sajs.co.za 65 November/December 2013 Research Article Ofduvai fossil isotopic ecology Page 3 of 14 digestive differences between ditterent species.^' The oxygen isotope drink surface water that is closest to meteoric water and feed at night, ratio is a measure of the body water of an , acquired from plant when humidity increases and plant 5"0 decreases."^ water or from drinking water and is altered by the thermophysiology of the animal. The oxygen isotope ratio can contribute to dietary and Materials and methods environmentai reconstructions^ ^'Î^" and is particuiarly indicative ot Samples of tooth enamel used for isotopic analysis in this study were changes in humidity and aridity.^^'^^ removed from the tooth specimens at their location of storage. In the To reach dietary conclusions based on carbon isotope ratios, it is case of Olduvai Beds I and II, this sampling was done at the National necessary to determine the C3 and C^ end members for a given time Museum ot Natural History in Arusha and at the Olduvai field station and place. This determination is most readily achieved by measuring the (Mary Leakey's old field camp, which has been expanded). Most of tooth enamel 5"C values of dedicated browsers and grazers. Girattes, the teeth from Olduvai Beds I and II had been excavated by the Olduvai for example, rarely eat anything but C3 leaves of trees and shrubs, whiie Landscape Palaeoanthropology Project (OLAPP) - an international alcelaphine antelope like the wildebeest concentrate on C^ grasses. At this group of researchers who have been working at Olduvai during the mid- point in time, dedicated browsing herbivores otthe South Atrican interior year field season since 1989.* Specimens from the OLAPP collections (whose tooth enamel the Cape Town laboratory has often analysed) have can be identified in Tables 1 and 2 on the basis ot their catalogue mean 5"C values of -14.5%o, while dedicated grazers have mean values numbers, which are in the format 95/43;156 (excavated 1995, trench of -0.5%o. These values can be altered by changes in ciimate and in the 43, number 156). Some specimens trom Mary Leakey's collection, atmosphere. The 'industrial ettect' ot the past 200 years, tor example, which is stored at the Olduvai field station, were also sampled. These substantially increased the 00^ content of the atmosphere, as a result are from Bed II (Table 2), excavation sites FLK N and HWKE. In total, 145 of fossil fuel burning, resulting in the 5"C value of the atmosphere (and faunal specimens were sampled from the collections of Olduvai Middle that of ail plants) becoming more negative by 1.5%o. The 5"C values Bed I (ca 1.785-1.83 Ma) and Lowermost Bed II (ca 1.75-1.78 Ma). of all modern specimens must consequently be corrected by adding The modern samples from the Serengetl National Park, Maswa Game 1.5%o to the measured number. Climatic changes that alter 5"C values Reserve (southern Serengeti), Lobo (northeast Serengeti) and Selous of plants include increased humidity, which may make the 5"C values of Game Reserve (southeast Tanzania) were obtained from collections at C3 plants (but not C^) more negative by 2%o.3' Tropical torests provide an Seronera, the headquarters of Serengeti National Park, and from some extreme exampie in this regard -their C3 plants (they have no C^ plants) of the other ranger stations in the park. A few specimens from Lukwati have 5"C values as much as 10%o more negative than those of plants and Ugalla Game Reserves were sampled in Arusha at a taxidermy shop growing in the open; this difference results from high humidity, low light and at the Rock Art Centre. A collection of 77 modern bone specimens, and, especially, from recycled CO^ produced by rotting leaf litter on the primarily from Serengeti National Park which adjoins Olduvai, was also forest floor.^^ In contrast, the 5"C values may become more positive as a assembled (120 measurements). result of aridity and bright sunlight,^'^^' while C^ plants may respond with slightly more negative values, as a result of the increased occurrence of Samples were removed from tooth specimens by using a variable speed enzymatic C^ subtypes that are adapted to such conditions. drill with a dental diamond drill tip of 0.5-mm diameter. A sample of 3 mg enamel powder is sutticient for at least two isotopic analyses. The approximate ratio of C3 and C^ plants in the diet of herbivores can This amount is the equivalent of about two sugar grains; if it is removed be arrived at by interpoiating between the C3end member (100% C3 diet) from a broken enamel edge, the sampling scar is usually invisibie. By and C^ end member (100% C^ diet) of the ecosystem under study, moving the drill tip along an enamel edge, broken along the length of a although there are some slight ditterences between different species that tooth, seasonal variations (particularly in "*0) are averaged. The enamel have the same diets. powder was gathered on smooth weighing paper and poured into a small centrifuge vial with snap lid, in which all the subsequent chemical Oxygen isotope ratios in tooth enamel can help to understand certain pretreatment was carried out at the Stable Light Isotope Facility ot the elements of a local habitat. The primary source of oxygen in biological University ot Cape Town. The powder was pretreated with 1.5-2.0% apatite is from water (drinking water or water in food) and from oxygen sodium hypochlorite (to remove organic materials and humic acids), bound in food. The light ^'^0 isotope evaporates more readiiy than the rinsed with water, and then reacted for 15 min with 0.1 M acetic acid (to heavy '^0 isotope, while the latter precipitates more readily. These remove readily dissolved carbonates). characteristics have produced such localised ettects as the water in modern East Atrican lakes being enriched by 5-10%o compared to the After washing and drying, 1 mg of the enamel was weighed into an waters flowing into them." Uitimately, the ^"0 content of water in an area individual reaction vessel of a Kiel II autocarbonate device (Thermo depends on precipitation - or meteoric water (5^*0 J - which decreases Electron Corporation, Bremen, ). The powder sample was with distance from the source of moisture (an ocean or lake), increasing reacted at 70°C with 100% phosphoric acid, which produces CO^ from latitude, increasing altitude and decreasing temperature.*^^ An example the carbonate in the enamel. This sample gas was cryogenically distilled ot the ditterent moisture sources can be seen in East Africa, where S^^O and its isotope ratios were measured in a Finnegan MAT 252 ratio mass values of waters in Kenya and Ethiopia ditter by about 2-3%o.''3 The b^'^O spectrometer (Thermo Electron Corporation, Bremen, Germany). The value of leaf water derived from the same water source also varies with 6^^C and S'^O were calibrated against the international VPDB carbonate aridity, because of evapotranspiration. standard by using a calibration curve established from National Bureau The integrity of 5"C and 5"0 values in tooth enamel for a given time of Standards standards 18 and 19, as well as by inserting internal and place can best be judged by observing the pattern of results for all laboratory standards of 'Lincoln limestone' and 'Carrara 2 marble' at the animal species in the biome. A considerable database for the Plio- regular intervals in the autocarbonate device. The precision of replicate Pleistocene of Africa is available by now and can be used to judge the analyses was better than 0.1 %o. results ot a given study. Browsing animals (C3 plant feeders) will have distinctly more negative 5"C vaiues than grazers (C^ plant teeders), with Results the most dedicated browsers and grazers differing by about 14%o. There The 5"C and S^^O values for fossil fauna from Olduvai West Middle Bed I is also a distinct pattern between ditterent species of grazers, depending (Table 1) and Olduvai East Lowermost Bed II (Table 2) are combined on the different amounts of C3 plants like forbs that are included in their in Figure 2 to illustrate the faunal community in the Olduvai region at diets. Wherever they have been compared, specimens of Damaliscus ca 1.8 Ma. These vaiues can be compared with the isotope ratios of spp. (e.g. the tsessebe and ) have more positive 5"C values than modern fauna from the adjoining Serengeti National Park, plus some Connochaetes spp. (wildebeest), which, in turn, have more positive specimens from Maswa, Lobo, Ugalla, Lukwati and Selous wildlife values than Equus spp. (zebra). For oxygen Isotopes, S^'O values of reserves (Table 3, Figure 3). The 5"C and 5^*0 values of the fossil and hippos are usually more negative than those of equids, because they modern fauna are compared in Figures 4 and 5, respectively.

South African Journal of Science Volume 109 ¡Number 11/12 http;//www.sais.co.za 66 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 4 of 14

Table 1: Carbon and oxygen stable isotope ratios in tooth enamel of fossil fauna from Olduvai West Middle Bed I. [Stratigraphy: above Tuff 1B, 1.83 Ma: below 1.785 Ma].*

UCTno. TZno. OLAPP no. (year, Irench, no.) Tooth 5"C 5»0 ORDER: CARNÍVORA FAMILY:CA'NIDAE Species: Canis mesomeias (black-backed Jackal) UCT7059 2 TZ150 96/57:862 -7.3 -4.4 ARTIOOACTYU SUIDAE 1 1 Phacochoerus modestus (cf. warthog) UCT7200 M, 0.1 -1.4 UCT7186 W -0.3 0.1 Mean (s.d.)1, n=2 -0.12(0.29) -0.65(0.01) 3CVI0AE Sub-family: Reduncinae Kobus sp. UCT7056 " jri T2147 95/57:536 -2.0 UCT7057 TZ148_ 95/57:626 -5.0 -7.3 UCT7058 TZ149 95/57:621 -3.8 -3.8 Mean (s.(l.).n=3 -3.6(1.54) -4.17(2.95) [Tfiese three specimens could be from one animal] Hippotraginae Hippotragus gigas UCT7055 ___3 T2146 96/65:443 RM, -0.3 -3.0 Anteloplnae Gaze//a//leAl«;erosJcf^la£ge morpb Grant's gazelle/ impala) -1.54 j TZ144 0.1 -4.4 UCT7054 ' TZ145 -2.0 -1.2 Mean (s.d.), n=2 -2.79(2.25) Gazeiia/ Antidorcas (cf. Tbomson's gazej[e/ sprin^jO UCT7049 '!" ' ^ZZrn 96/65:287 -3.5 UCT7050 TZ141 97/88:204 0.7 TZ142 96/65:481 RM, -4.1 -4.6 UCT7052 TZ143 95/57:863 RM,,. -1.9 Mean (s.d.), n=4 -2.22(1.85) -3.21(0.97) Megaiofragus sp. UCT7037 TZ128 96/65:540 LM'« -1.6 -3.7 Repeat -0.9 -3.5 UCT7038 TZ129 96/57A:18 M« -0.9 -4.8 UCT7039 TZ130 96/65:42 LM'« 1.7 -4.0 UCT7040 TZ131 96/65:480 M" 0.2 -5.4 Mean (s.d.), n=4 -0.29(1.31) -4.28(0.79) BeatragusI Connochaetes (cf. wildebeest) UCT7045 TZ136 96/65:148 RM« 0.3 -5.7 UCT7046 TZ137 96/65:438 M« -1.1 -3.4 Repeat -1.1 -4.4 Repeat -1.1 -3.9 UCT7043 TZ132 96/88:67VI/ M« 0.2 -6.9 UCT7044 TZ133 96/88:73Vi; RMi" 0.2 -4.0 Mean (s.d.),/7=4 -0.11(0.66) -5.01(1.6) Parmularius sp. UCT7041 17132 96/65:286 LM« 1.26 -2.50 UCT7042 TZ133 96/65:349 M" 1.09 -1.63 Mean (s.d.), n=2 +1.17(0.12) -2.06(0.61) Mean (s.d.), n=8 (all alcelaphines) +0.14(1.06) -4.21(1.44) SQUAMATA GROCODYtlDAE Crocodyius niioticus UCT7032 TZ123 96/57A:312 -2.7 -5.1 UCT7033 TZ124 96/57A:313 -3.1 -5.0 UCT7034 TZ125 96/57A187 -2.2 -5.0 UCT7C36 TZ127 96/57A:75 -3.3 -6.3 Mean (s.d.),n=4 -2.82(0.48) -5.36(0.64) [These four specimens could be from one animal] UCT7035 TZ126 96/57:491-2 -4.4 -7.6 Mean (s.d.), n=5 (all crocodiles) -3.14(0.82) -5.81(1.15) Mean and standard deviation of 5"0 for 27 animals -4.95(-1.83)

28 measurements trom Olduvai West + 2 from Oiduvai East are inciuded in fhe table. 5"C and 5"0 vaiues for individual specimens have been rounded off to 0.1 but mean and s.d. are based on originai measurements. In the column headings, the UCT numbers refer fo the database of the Sfabie Light Isotope Facility in the Archaeology Department at the University of Cape Town; TZ numbers are from the author's field coiiecfion ot specimens from Tanzania; and OLAPP numbers are from the excavations ot the Oiduvai Landscape Paiaeoanthropology Project.

South African Journal of Science Volume 109 {Number 11/12 http://www.sajs.co.za 67 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 5 of 14

Tabie 2: Carbon and oxygen stable isotope ratios in tooth enamel ot fossil fauna from Olduvai East Lowermost Bed II (plus three hippos trom Lower Bed II). [Stratigraphy: overlying Tutf 1F: 1.75-1.78 Ma].

UCT no. TZno. OLAPP no. (year, trench, no.) Tooth 6"0 ORDER: CARNÍVORA Family: CANIDAE Hyaenid UCT7016 TZ107 95/43:157 LP, -1.3 -6.7 UCT7017 TZ108 89/3:106 RPS -2.5 -5.8 Mean (s.d.), n=2 -1.9(0.86) -5.73(0.03) Acinonyx sp. (ct. cheetah) UCT7018 J TZ109 95/44:168 LI, -2.0 1 -4.8 Lycaon sp. (cf. wild hunting dog) UCT7019 _^ TZ110 LP, -2.4 -7.88 UCT7020 TZ111 LM, -1.9 -6.66 Mean (s.d.), n=5 -2.02(0.48) -6.168(1.15) PRÜBÜSCIDEA ELEPHANTIDAE Elephas recki (grazing elephant) UCT9936 TZ179 HWKE/104.4L1 Juvenile? 2.3 -1.8 GRÍP3A UCT9937 TZ180 HWKE/104.4L1 Deciduous 1.3 -2.0 UCT9937 TZ180F HWKE/104.4L1 Fragments of TZ180 0.4 -2.7 Mean (s.d.), n=3 -1-1.3(1.0) -2.2(0.5) Deinotherium bozasi r — - - ÜCT9938 TZ181 HWKE/104.4L1 Fist size tooth -11.1 -4.4 Grid 3A:35 UCT9938 Tziai Repeat measurement -11.0 -4.3 -11.1 -4.3 UCT9939 TZ182 HWKE/104.4L1:36 Prababiy same animai as TZ181 -10.2 -5.8 UCT9939 TZ182 Repeat measurement -10.4 -5.7 Mean (s.d.), n=4 -10.68(0.42) -5.04(0.83) Mean (s.d.), n=2 animals -10.76(0.4) -4.9(0.78) PERISSODACTYU EOUiDAE Equus olduwayensis UCT7561 TZ59 94/34:474 P/M, 0.6 -2.0 UCT7563 TZ61 94/27:47 M" 0.2 -2.8 ÜCT7564 TZ62 94/27:47 M" -0.6 -2.3 TZ61 and TZ62 are the same tooth; mean: -0.4 -2.6 UCT7565 1 T763 94/34:222 M, -0.2 -2.6 UCT7566 "1 TZ64 94/34:780 M, 1.4 -1.9 UCT7568 TZ65 94/53:86 1.0 -1.8 UCT7569 T266 96/44:1508 M., -0.8 -3.7 UCT7573 TZ70 94/2t:229 M* 0.7 -3.3 UCT7575 T772 94/34:230 -1.7 -3.0 Mean (s.d.), n=8 +0.01(0.9) -2.69(0.6) Equus oiduwayensis? UCT9940 TZ184 HWKE/104.2L8 2.7 -3.1 Equid (no species identification) UCT994t TZ185 HWKE/104.4L1:7 3.0 1.1 Hipparion sp. UCT7572 1769 94/39:4 M, 1.0 -5.3 UCT7001 TZ92 89/3:105 LPyM, -1.8 -4.6 Mean (s.d.), n=2 -0.4(2.0) -4.9(0.6) Ambiguous (smaii morph) UCT7559 TZ57 96/43:496 M" -1.2 -2.8 UCT7560 TZ58 96/43:644 fJC 0.9 -2.1 UCT7562 TZ60 94/21:302 1 -0.4 -3.8 UCT7570 TZ67 95/43:655 M' 1.9 -1.7 UCT7571 T268 95/43:? 1.2 -4.0 UCT7574 T771 95/43:646 M" -0.2 -2.8 Mean (s.d.), n=6 -HO.37(1.16) -2.9(0.9) Aiiequids,n=18 0.36(1.32) -2.76(1.30) ARTIODACTYLA HiPPOPOTAMiDAE Hippopofamus gorgops UCT7180 TZ01 94/21:321-26 ? 1.8 -6.2 UCT7t99 TZ02 95/43:650 pm 1.4 -6.0 UCT7239 TZ03 94/34:5 1 0.9 -4.3 UCT7185 TZ04 95/45:283 P/M -0.8 -5.4 UCT7225 TZ05 94/34:18 ? 1.7 -3.8 UCT7423 TZ06 94/21:220 7 0.6 -2.4 UCT7181 TZ07 94/34:420 Tusi( -0.8 -4.2 Continued on next page

South African Journal of Science Volume 109 \ Number 11/12 http;//vtfww.sajs.co.za 68 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 6 of 14

UCT no. TZno. OUPP no, (year, trench, no.) Tooth 8"C 6"0

UCT6982 TZ73 97/98:36 0.5 -5.1 Mean (s.d.), n=8 -1-0.66(1.0) -4.7(t.2) LOWER BED II Hippopotamus gorgops UCT9952 TZ196 HWkE/104.4 Canine 2.2 -7.2 GRID1C, L1:25 UCT9953 TZ197 L_ HWkE/tO4.4 Mj 0.6 -5.7 GRI01A, L1:37 UCT9954 T2198 HWkE/104.4 M,? -0.9 -3.6 GRID1B,L1:11 UCT9954 Repeat -0.8 -3.8 Average UCT9954 -0.85 -3.7 Mean (s.d.), n=3 t.64(1.5) -5.53(1.7) SUIOAE Ptiacoctioerus modestus UCT7196 94/27:321 UCT6987" 94/35:Ï6 ' -2.9 95/44:331-2 '-4.7 94/27:4^ RC -111 -2.4 94/27:48 LC -1.5 -3.1 94/27:48 0.0 -1.8 [Jijeas are trom the same animal; means -0.9, -2.4] Mean (s.d.), n=4 Koipociioerus limnetes UCT7191 95/45:327 UCT7428 95/45:328 -0.9 -2.8 UCT6990' 95/45:202 JL. -3.1 Mean (s.d.), n=3 -10(0.4) [These three teeth from trench 45 could be frqrn the same animal] koipoctioerus afarensis ÜCT7230 95/44:1509 UCT7240 95/43:824^ UCT6992 "95/44:659^ UCT6993 95/44:t5t6 UCT6994 UCT6995 UCT6996' UCT6997'

[Teeth troni Jrenj;h_44^ould be frorn the sarne animal] "7 Reduncinae Kobus sp. (ct. waterbuck; UCT7015 TZ106 95/43:662 LM'" -0.6 -4.4 Tragelaphinae Trageiaphusjtrepsiceros (greater Mu)_ UCT6983 ^ ' I " TZ74 "" 89/11:45 -2.4 -0.3 UCT6984 I TZ75 -5.7 -0.1 Trageiaphus smpius (bushbuck) UCT6985_ „ T^r "Zl? LM, -5.7 0.6 Trageiaphus oryx (eland) UCT7237 ^ TZ21 94/24:12 -4.2 -2.2 Trageiaptius sp. (no species identification) UCT7201_" TZ20 94/21:104 ID? M? -2.9 -3.8 UCT76986 TZ77 94/23:73 -9.2 0.3 Mean (s.d.), n=6 Alcelaphlnae Connodiaetes / Beatragus (cf. wildebeest)

Parmuiarius sp. UCT6981 ]2] TZ33 94/34:79 LM,,. 2.0 -2.3 UCT7198 " 95/50:11 2.1 -1.3 UCT7228 ^ JRL. 2.5 -4.2 UCT7238 95/44:334 TZ38 3.8 -1.1 Continued on next page

Soufii African Journal of Science Volume 109 \ Number 11/12 http://www.sajs.co.za 69 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 7 ot 14

UCT no. TZ no. OLAPP no. (year, trench, no.) Tooth 8»C 6"0 UCT7184 TZ40 95/21-.101-102 1 M,,j 0.9 -1.6 ÜCT7545 TZ44 95/50:10 M,„ 1.3 -1.3 UCT6999 TZ90 95/44:21 RM,,, 1.1 -1.1 UCT7000 TZ91 95/43:142 RM,,, 1.7 -4.5 Mean (s.d.), n=8 -H .9(1.0) -2.2(1.4) Megalotragus?_ ÍÍCT9942 ^ Tr104.2L8C 2.7 -3.1 Hippotraginae Hippotragus gigas (cf. roan, sable antelope) UCT7002 JZ93 95/43:48^^ UCT7003 95/43:1141' M" UCT9943"" "Tr104.2L8 ' 0.5 0.2 ÖCT9944" 1.3 UCT9946" >r104.2L8 0.2 -0.1 Repeat 0.2 IJCT9946 0.2 mmf Tr104.2L8 0.6 JTI92_ Tr104.2L8C 1.1 ÜCT9949 J7193 ÏÏCT9950 J TZÍ94 0.8 O.f Mean (s.d.), n=9 -h0.54(0.58) -0.33(1.51) Antelopinae Gazetia sp.[large morph. ct. Grant's gazelle//lepyceros (impaia)] UCT7007 TZ98 95/44:1278 _0.2_ -e.2. UCT7008 TZ99 • Î 95/44:904 -7.7 UCT7006 TZ97 "1 95/44:1325^ -9.0 -7.5 j Mean (s.d.), n=3_ -5.5(5.0) [Compare moderjjnipalaJrornMaswa,5"C =1.8] Gazelia sp. [small morph, ct. Thomson's gazelle//Iní/tíoreas (springbok)] UC17009 TZ100 1.2 TZ103 95/3:107 RM'« 0.3 -4.7 TZ105 95/43:35 RM"' 0.0 UCT70J0 95/43:147 -5.2 -3.4 IJCT70J1 95/43:163 -5.6 11 UCT70T3" TZ103 95/44:1314 RM,,, -9.1 -5J Mean (s.d.), n=6 .^1 :M -2.85(2.4) [Compare modern Thornson's gazelle trom Masvya, 5"C = -5^-6.:^: springbok sjecie^ have beerijrazers (A toncff) an^rowsers (A.jrjajsupialis)] GIRAFFIDAE Girafta sp. UCT7Ï87"

Sivatherium sp. (giant grazing giraffe) ÏÏcf7429 I ^fÏ UCJ7546 UCT6998

SQUAMATA CROCODYLIDAE_ _ Crocodylus niloticus

120 measurements are included in ttie table. 5"C and&"0 values for individual specimens have been rounded off to 0.1 but mean and s.d. are based on original measurements. In the column tieadings. ttie UCT numbers refer to the database of the Stable Ligfit isotope Faciiity in the Arcfiaeoiogy Department at the University of Cape Town; 7Z numbers are from ttie author's field coilection of specimens from Tanzania: and OLAPP numbers are from the excavations of tfje Olduvai Landscape Paiaeoanffiropology Project.

South African Journai of Science Volume 109 ¡Number 11/12 http://www.sais.co.za 70 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 8 of 14

Table 3: Carbon and oxygen stable isotope ratios in tooth enamel of modern fauna from Serengeti National Park and Maswa Game Reserve (southwest Serengeti), Lobo (northeast Serengeti), Selous Game Reserve (southeast Tanzania), Ugalla and Lukwati (southwest Tanzania)

UCT no. TZno. Collection locus Tooth 5"C 6"0

Order: CARNÍVORA Farrily: FELIDAE Panthera ßardus (leopard) UCT7076 TZ167(1) Lui

Mean (s.d.), n=3 HYAENIDAE Crocuta crocuta (spotted hyaena) UCT9994 TZ240 Serengeti N/A oniy organic UCT10002 TZ248 Serengeti -5.2 -0.7 UCT10018 TZ264 Serengeti -3.1 -2.8 Mean (s.d.),n=3 -4.18(1.5) -1.76(1.5) Lycaonpictus (wiid hunting dog) UCT10019 j TZ265 Serengeti -4.2 0.3 PROBOSCIDEA ELEPHANTiOAE Lomäonta africana UCT10014 TZ260 Serengeti -1.6 PERISSODACTYLA EQUiOAE Equus burcheiti UCT9983 TZ229 Serengeti 7 0.6 2.8 UCT9984 TZ230 Serengeti 9 0.5 1.7 UCT9985 TZ231 Serengeti ? 0.6 3.4 UCT9986 TZ232 Serengeti ? 1.2 2.3 UCT9987 TZ233 Serengeti 7 1.1 3.3 UCT9988 TZ234 Serengeti ? 0.8 1.5 Mean (s.d.), n=6 0.77(0.3) 2.49(0.81) UCT7079 TZ169(1) Lukwati RM' -0.7 -1.2 REPEAT TZ169(1) Luiiwati RM' -1.4 -2.6 UCT7079 TZ169(2) Lui

South Atrican Journal of Science Volume 109 \ Number 11/12 http;//www.sajs.co.za 71 November/December 2013 Research Article OlduvaÍ fossil isotopic ecology Page 9 of 14

UCTno. TZno. Collection locus Tooth 6"C 8"0

UCT10060 TZ306 Serengeti Í -12.6 3.3 Mean (s.d.),n=3 -12.71(0.14) 3.59(0.66) BOVIDAE Syncerus caffer (buffalo) UCT7071 TZ162 Maswa RM' 0.7 -O.B UCT9967 TZ213 Serengeti 2.0 3.3 UCT9968 TZ214 Serengeti 1.9 2.6 UCT9969 TZ215 Serengeti 2.3 3.6 UCT9970 TZ216 Serengeti -0.4 4.5 UCT9971 TZ217 Serengeti -0.2 3.2 UCT9975 TZ221 Serengeti 0.2 2.1 UCT10000 TZ246 Serengeti 1.4 3.3 UCT10001 TZ247 Serengeti 0.5 2.3 Mean (s.d.), n=9 0.93(1.0) 2.69(1.48) Tragelaphus strepsiceros (greater kudu) UCT7062 TZ153 Maswa LM' -14.1 2.0 UCT7063 TZ154 Maswa RM' -16.6 -3.5 Mean (s.d.),n=2 -15.35(1.7) -0.78(3.9) Tragelaphus imberbes (lesser kudu) UCT7061 TZ152 Maswa RM^ -12.6 0.5 Trageiaphus scriptus (bushbuck) UCT7069 TZ160 Maswa RM' -14.8 1.5 UCT7070 TZ161 Maswa RM' -15.4 -1.8 Mean (s.d.),n=2 -15.11(0.4) -0.13(2.3) Redunca redunca (Bohor reedbuck) JCT1067 TZ15B Maswa RM' 0.4 Kobus eflipsiprymnus (common waterbuck) 1 UCT706B TZ159 Maswa LM' 0.5 [ :a7 - - - Hippotragus niger (sable) UCT 7060 [ TZ151 Maswa RM' 0.2 -0.1 Hippofragus eguinus (roan) UCT7C7B TZ168(1) Ugalla 1.6 -0.2 UCT7078 TZ168(2) Ugalla 2.1 0.5 Mean (s.d.), n=1 (same animai) 1.87(0.33) 0.67(0.45) Connochaetes taurinus (blue wiidebeest) UCT7065 TZ156 Selous RM' -1.2 Connochaetes gnu (black wildebeest) UCT9965 I TZ211 Serengeti 2.5 2.5 UCT9966 JlPZ TZ212 Serengeti 1.2 3.9 JCT9976 TZ222 Serengeti 1.4 3.1 UCT9977 ! TZ223 Serengeti 0.6 3.0 UCT9978 T~ TZ224 Serengeti 1.3 4.8 UCT9978 I TZ225 Serengeti 2.66 -1.24 Repeat 2.69 -0.71 Average 2.68 -0.98 UCT99B0 TZ226 Serengeti 1.5 2.B UCT99B1 TZ227 Serengeti 2.7 2.9 UCT99B2 TZ228 Serengeti 1.6 4.9 UCT9996 TZ242 Serengeti 1.9 5.6 Mean (S.d.), n=10 1.74(0.71) 2.54(2.3) Aiceiaphus buseiaphus (Kongoni, Coke's ) UCT7064 TZ155 Maswa LM' 0.5 -0.7 UCT10023 TZ269 Serengeti -1.0 1.7 UCT10024 TZ270 Serengeti 2.2 1.4 Mean (s.d.),n=3 0.55(1.63) 0.78(1.31) Continueil on next page

South African Journal of Science Volume 109 \ Number 11/12 http://www.sajs.co.za 72 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 10 of 14

tJCT no. TZno. Coilection iocus Tooth 8"C 6"0

Alcelaphus iichtensfeini (Lichtenstein's tiartebeest)

UCT7076 TZt66 Lobo LMä 2.1 0.4

Repeat 2.4 -0.6

Average 2.25 -0.27 UCT7076 TZ166(1) 1.6 3.7 UCT7076 TZ166X 1.9 3.6 Repeat TZ166X 1.9 1.1

Mean (s.d.), n=i + 1.79(0.21) 2.80(1.51)

Mean (s.d.), n=2 2.08(0.22) 0.39(0.9) Damaiiscus lunatus (topi)

UCT10005

UCT10005

Average

UCT9990

UCT7066

Aepyceros melampus (impaia^

UCT7074 T TZt66

UCT9992 T TZ238

UCT9999 1 TZ246

^ Mean (s.d.

Gazelia thomsonii (Thompson's gazeiie)

UCTt0016 TZ262 Serengeti -1.3 4.8

UCT10016 -1.2 6.1

UCTt0016 Average n=^ -1.2 4.9 UCT10006 TZ252 Serenget -0.9 6.2

UCT10007 TZ263 Serengeti -2.4 6.0

UCT10008 TZ264 -2.8 4.8

UCTt0013 TZ259 Serengeti -3.0 4.9

UCT7072 TZ163 Maswa -6.7 -0.9

UCT7073 TZ164 Maswa -6.7 -0.8

Mean (s.d.), n=7 -3.26(2.19) 3.29(2.85) Gazeiia granti

UCT10020

UCT10020

UCT10020

UCT10021

UCTt0021

UCT10021

UCTt0022

UCTt0022

UCTtOO22

Ostrich eggsheii UCT10003 TZ249 Serengeti -10.2 6.7 Repeat -10.0 6.7 Repeat -9.8 6.8

Mean (s.d.), n=3 -9.99(0.19) 6.71(0.06) 5"0 mean for Serengeti animais (tooth enamei) only, n=50: +1.78 (2.66)

78 tooth enamei samples (72 animais) and 1 ostrich eggsheii are inciuded ih the table. Most of these specimens were sampled from the collections in Serehgeti Nationai Park; the specimens from Ugaiia and Lukwati were sampled at a taxidermist in Arusha. b"Candb"O vaiues for individuai specimens have been rounded off fo 0.1 but mean and s. d. are based on originai measurements. To compare modern 5"C vaiues with those of fossii specimens, 7.5% shcuid be added to correct for the industriai effect offhepasf 200 years. in ttie coiumn headings, the UCT numbers refer to the database of fhe Stabie Light Isotope Facility in the Archaeoiogy Department af the University of Cape Town; 7Z numbers are from tiie author's fieid coiiection of specimens from Tanzania.

South African Journal of Science Volume 109 \ Number 11/12 http://www.sajs.co.za 73 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 11 of 14

Bed I: A Bed«: 2

1 Giraftidae: Sivatherium 0 Bovidae: Trageiaphus -1 Suidae: Phacoohoerus

' -2 Giraffidae: Giraffa Elephantidae: Elephas Equidae:, Bovidae: Gazella -3 Equus & (2) ovidae: Gazella Hipparion (1) -4 Suidae: Phacoohoerus & Canidae: Canis Kolpochoerus -5 Crooodylidae: r . Elephantidae: Loxodonta Crocodylus I Hippopotamidae: -6 Crocodylidae: Crocodylus Hippopotamus Canidae: Hyaenid, Àcinonyx, Lycaon -7 (1) (2) Bovidae: Kobus, Cohnochaetes, Parmuiarius, Megaiotragus & Hippotragus

-12 -11 -10 -9 -8 -7 -5 -3 -2 -1

Figure 2: S'^C and S'^O values obtained trom tooth enamel of fossil fauna from Olduvai Beds I and II.

Serengeti: • Maswa: * Selous: x Ugaiia: » Liil

Bovidae: Syncerus, Connochaetes, Aiceiaphus, Oamaliscus, Aepyceros

Equidae: Equus

Bovidae: Hippopotamidae: Bovidae: Redunca, Kcbus, Aiceiaphus Hippotragus, Aiceiaphus, Hippopotamus Oamaiiscus, Aepyceros & Bovidae4: Syncerus Hippotragus Feiidae: Panthera Bcvidae: Gazeila X Bovidae: Hyaenidae: Connochaetes Elephantidae: Loxodohta Crocuta & Lycacn Equidae: Equus

-16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4-3-2-10123

Figure 3: 5^^C and 5^^0 values obtained from tooth enamei of modern fauna from Serengeti, Maswa, Lobo, Ugalla, Lukwati and Selous wiidife reserves.

Diceros ùkxmls G¡mm sp Legend: Modern: black •; Bed I: grey .4; Bed I!; grey m GirBiïa camelopardalis Loxoäor}la africana Deinoíhsrmm bûsssi

Panthern feo Cemivcrss ¡H-i^snKi. A Crocuta &ocúía Car)¡$ mesomsias Kobus sp. hipparion sp. Gazolia sp.

Megaiotragus sp. .A A .4 4 Equus burchelli I 1 -i B'~ Equus oSdowayansis 1—•—i -111 Hippopotamus sp, Hippotragus gigas -J , 10 Bephas sp. Connochaeies gnu MMlaphinss

-14 -13 -12 -U -10 -9 -7 -6 -S -4 -3 -2 -1 0 1 2 Î 4 5

+ 7.5%ii Aas been added to the modern values to correct for the industrial effect Figure 4: 6'^C trom tooth enamel ot modern tauna trom the Serengeti (in black) and trom tossil tauna trom Olduvai Beds I and II at ca t .8 Ma (in grey).

South African Journal of Science Volume 109 ¡Number 11/12 http://www.sajs.co.za 74 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 12 of 14

IXcemsblcomls I 2 Legend: Modem: black •; Bed I: grey i; Bed II: grey • GIraffa sp • • 3 Giraría camelopardalls •• «3 • 1 Loxodonta africana Deinotherium bozasi Suids Phacochoerus modestus Panlhera leo C-:i'" I": • '• ^enid, Acinonyx) Crola rocta

Kobus sp. » 1 A » « 2

Magalotragus sp. Equus burcttelll Equus oläowayönsis Hippopotamus sp. Hippotragus gigas Crocodylus nlloticus -H10 Bephas sp. Connochaates gnu Alcelaphines iConnochaataslBaatragus & Parmularius sp.)

Figure 5: 5"0 from tooth enamel of modern fauna from the Serengeti (in biack) and from fossii fauna from Oiduvai Beds i and II at ca 1.8 Ma (in grey).

In general, the carbon isotope ratios ot the same or related species have it shouid be noted that the Bed II specimens are clustered toward the stayed remarkabiy simiiar over 1.8 miilion years. The avaiiable plant grazing end ot the spectrum with tew browser representatives. This toods have obviousiy stayed much the same. The oxygen isotope ratios, pattern is especially evident when the Bed II assemblage is compared in contrast, have become enriched by some 6%o. One obvious ditterence with that ot the modern Serengeti, which has a similar scarcity of can be tound in the case of hippos: the 11 specimens ot Hippopotamus browsers. It is noteworthy that the tragelaphines (kudu, bushbuck gorgops trom Beds i and II were dedicated grazers, whiie the single and eland) trom Bed II have a mean 5"C value ot -5.0%o. They were modern specimen of Hippopotamus amphibious trom the Serengeti was evidentiy mixed teeders, in contrast to the modern kudu and bushbuck a mixed teeder Hippos are generally regarded as dedicated grazers, but trom Maswa, which are dedicated browsers with a mean 5"C value this assumption is not correct"" - they will eat C3 plants it grasses are ot-14.7%0. not avaiiable nearby. The 5"C values tor modern tauna (Table 3) have not been corrected tor the industrial ettect. In order to compare them with the 6"C values tor Carbon isotopes Olduvai Beds I and II, it is necessary to add 1.5%o. Figure 4 shows the corrected vaiues. The collection trom Oiduvai West Bed I is somewhat limited: the 27 specimens comprise t black-backed jackal, 6 crocodiies and 20 grazing The 6"C values ot modern animais (Table 3) represent several different animals (suids and bovids). No browsers are included, which makes it biomes in Tanzania. This muiti-representation is immediateiy obvious impossible to determine the C3 end member ot this taunai community. when one compares the 5"C values tor Equus burcheiii (plains zebra) At the C„ end of the collection, the most positive 5"C value is that ot trom Serengeti Nationai Park (-1-O.8) with those trom Lukwati (-2.0). -Hi.t%o tor two specimens ot Parmuiarius (an alcelaphine). The 5"C Lukwati is located near Lake Rukwa, between Lakes Tanganyika and values trom the Bed I specimens are very similar to those trom the larger Nyasa, and its grazers clearly have more C3 plants in their diet. Simiiarly, collection ot 118 specimens trom Bed II, which inciudes canids, suids, the browsers trom the Serengeti inciude the modern giratte {Giratfa hippotragines, anteiopines, alcelophines and crocodyiids. The vaiues tor cameiopardis, 5"C = -12.7), elephant {Loxodonta africana, 5"C = Beds i and II have been added together in the tigures to represent the -11.9) and black rhino (0/cerosô/com/s, 5''C = -13.9). Maswa Game tossii tauna. Reserve, which borders Serengeti on the southwest side but is much more wooded, is represented among the browsers by three tragelaphines: Olduvai Bed II is represented by 118 specimens, which provide a the greater kudu {Trageiaphus strepsiceros, 5"C = -15.4), iesser comprehensive view ot the community at ca 1.8 Ma. The browsing kudu (7: imberbes, 5"C = -12.6) and bushbuck {I scriptus, 5"C = end ot the spectrum is represented by two specimens of Deinotherium -15.1). The grazing end ot the Serengeti spectrum is represented by bozasi (-11.05: -10.32), a giant browsing elephant with tusks growing Hippotragus equinus (roan antelope, 5"C = -1-1.9), Connochaetes trom its lower jaw and bent downwards towards the ground. A specimen gnu (black wiidebeest, 5"C = -H1.7) and Alcelaphus iichtensteini ot giratte, Giratta sp. (probabiy graciiis), has a less negative 5"C value (Lichtenstein's hartebeest, 5"C = -1-2.0). in calculating the C3 and C,, of -9.1 %o. It is of interest that Eiephas recki, an elephant species that end members, the Serengeti specimens have been emphasised, yielding occurred widely trom East to southern Atrica at this time, was clearly approximately -12.5 and -(-1.5. When corrected tor the industrial ettect, a grazer (5"C = -1-1.3), while Sivatherium sp. (also Libytherium sp.), the end members are about -11.0 and -1-3.0. This result is similar to that a short-necked grazing giratte (5'^C = -1.5), was also present during trom Olduvai Bed li: however, similarity does not mean that the piant Bed II times. These species ot elephant and giraffe were later replaced communities ot the two biomes were the same, only that the avaiiability by the modern dedicated browsing elephant, Loxodonta africana, and ot C3 and C,, plants tor browsers and grazers were similar giratte, Giraffa camelopardls. The grazing end otthe spectrum at Olduvai In Figure 1, a selection ot tooth enamel 5"C values trom Olduvai Bed II in Bed II times was represented by aiceiaphines otthe generaßeafragus/ and the modern Serengeti are compared: the modern vaiues have been Connochaetes (5"C = -1-2.1), Parmularius (-H1.9) and Megaiotragus corrected tor the industriai ettect. The results are extraordinarily simiiar, (-1-2.7). Based on these values and that ot Deinotherium bozasi, one suggesting two similar landscapes ot wooded grassland. The exception can estimate the 5"C values tor the C3 and C,, end members tor Bed II at is a singie 5"C value tor a modern hippo trom the Serengeti, which is about -11.5%(j and -i-2.5%o, respectiveiy. evidently a mixed teeder.

South African Journai of Science Voiume 109 ¡ Number 11/12 http://www.sajs.co.za 75 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 13 of 14

Oxygen isotopes An alternative explanation for the change in enamel 6^'*0 ratios is that the From the ô^^O values reported in Tables 1, 2 and 3, it is clear that primary source ot meteoric water changed at some point during the past oxygen isotope ratios vary substantially with time and place. Among 1.8 iVla. It is suggested that this change took place with the introduction 27 measurements for Olduvai West Bed I (27 animals. Table 1) there of the Indian Ocean monsoon to this part of East Africa. Currently, the is only a single positive value for 5''0. Among 124 measurements for major source of rain is the Indian Ocean monsoon which brings the 'long Olduvai East Bed II (111 animals. Table 2) there are only 10 positive 5^*0 rains' at mid-year, while the Atlantic Ocean provides the 'short rains' values. Among 77 measurements tor ô'^O in the tooth enamel of modern at the end of the year Given the long distance from the Atlantic Ocean, animals (74 animals. Table 3), 56 values are positive. This contrast is across Angola and Tanzania, the heavy isotopes of hydrogen and oxygen well illustrated in Figure 5, which shows that modern animals have 5"0 are substantially rained out along the way. This phenomenon has been values that are more positive than those of the fossil specimens by about documented for the rainfall at the research station at Seronera in the 6%o. To be precise, the 50 animals trom the Serengeti are 6.24%o more Serengeti National Park. A research project, as yet unfinished, has been positive than 27 animals from Bed I and 5.56 more positive than those launched by the author in collaboration with Cassian Mumbi of TAWIRI from Bed II (112 animals). Of the 21 negative 5"0 values for modern (Tanzania Wildlife Research Institute) and statt members at Seronera, animals, 9 are for animals from the Serengeti (out of 53 animals, exact to measure the oxygen isotopes in samples from the rain gauge over location unknown), 8 are for animals from Maswa (out of 13 animais) nearly 2 years. The oxygen isotope values ot rain from the two oceanic and 2 are for animals from Lukwati (out of 4 animals). moisture sources differ by as much as 10%o. It is also possible that nearby Lake Victoria, which was only formed around 50 ka, could be As the body water of animals is largely controlled by the rain in the adding '^0 to the local rainfall. area, it is suggested that the source(s) of rain in the Serengeti, Maswa and Lukwati are different. This hypothesis is under investigation, with It is noteworthy that the same isotopic phenomenon (no change in 5^^C, the aim of estabiishing when the change in the rain source in the but substantial enrichment in 5"0) can be observed at Lake Natron Olduvai/ Serengeti area occurred. between ca 1.5 Ma and modern times.^ As more isotopic analyses are done on fossil tooth enamel from northwest Tanzania, it is likely that the Conclusions timing of this change will be established. The stable carbon and oxygen isotope ratios of faunal tooth enamel at Acknowledgements Olduvai and in the adjoining Serengeti National Park in Tanzania provide valuable information about the environment in that region at ca 1.8 Ma The fossil specimens from Olduvai were obtained from the collections and in modern times, in general, the carbon isotope ratios indicate of the Olduvai Landscape Archaeology and Palaeoanthropology Project that the environment of Oiduvai Bed I and II was very similar to that (OLAPP), curated at the National History Museum in Arusha and at the of the modern Serengeti - that is, 'savannah grassland with scrub and Olduvai Field Camp. Rutgers graduate student Amy Cushing identified bush' (Gentry and Gentry, quoted by Cerling and Hay"). In terms otthe the fossil specimens and helped with sampling. The modern specimens UNESCO definition,'"' it was a wooded grassland with 10-40% woody from Serengeti National Park were obtained from the Research Station at plant cover or a grassland with less than 10% woody plants. When Seronera. Emily Kisamo, the acting chiet warden of Serengeti National the carbon isotope ratios of modern fauna have been corrected for the Park at the time, was instrumental in providing sampling access. The industrial effect in the modern atmosphere, the values of the fossil fauna isotopic ratios of all the specimens were measured at the Stable Light are essentially the same as those of their modern counterparts. There is, isotope Facility ot the Archaeology Department at the University of Cape however, a contrast between the fossil and modern tragelaphines. The Town by Ian Newton and John Lanham. Cape Town graduate students woody piants growing at Oiduvai at ca 1.8 Ma were evidently insufficient helped with the production ot the manuscript; Mariagrazia Galimberti for these animals to be dedicated browsers. produced the figures, while Kerryn Warren typed the manuscript and checked the statistical calculations. I thank all of these individuals most A major change in the environment at Olduvai obviously occurred when heartily. Funds for the research were provided by the National Research Lake Olduvai dried up when the Gorge was formed (later than Bed II). This Foundation of South Atrica, the University of Cape Town, and the Landon change removed a number of water-related plant species and freshwater T Clay Fund of Harvard University. fauna from the environment, but did not result in major changes in the diets of browsing and grazing animals, or in their relative prevalence in References the landscape. However, If the major component of the diet of R boisei 1. Van der Merwe NJ, IVlasao FT Bamford iVlK. isotopic evidence for contrasting was papyrus, then their staple tood disappeared. diets ot early hominins Homo habiiis and Austraiopitiiecus boisei ot Tanzania. In contrast to the lack of difference in carbon isotope signatures in the SAfrJSoi. 2008;104:153-156. fauna at Olduvai between fossil and modern specimens, their oxygen 2. Van der Merwe NJ. isotopio eooiogy and diets of fossil fauna trom the T-1 isotope signatures are distinctly different. The 5^^0 ratios of the modern (Type Seotion, iVlaritanane) paieosurfaoe. in: Dominguez-Rodrigo iVl, Aioaia L, faunal community are about 6%o more positive than those of the fossil Luque L, editors. Peninj, a research project on origins (1995-2005). fauna from Beds 1 and II. A change of this magnitude has been observed Oxford: Oxbow Books; 2009. p. 109-114. in palaeosol carbonates in Olduvai Gorge by Cerling and Hay", with the 3. Ceriing TE, iVIbua E, Kirera FiVI, iVlanthi FK, Grine FE, Leakey MG, et ai. Diet of major change taking place at about 0.5 Ma. Paranthropus boisei in the eariy Pieistocene ot East Africa. Proc Nati Acad Sei USA. 2011:108:9337-9341. http://dx.doi.0rg/10.1073/pnas.1104627108 A ready explanation for such a change in 5"0 ratios is that there was a major change in temperature and humidity at some point during the 4. Lee-Thorp JA, Thackeray JF, Van der iVIerwe NJ. The hunters and the hunted past 1.8 Ma. This environmental change would have caused a major revisited. J Hum Evoi. 2000;39:565-576. http://dx.doi.org/10.1002/oa.659 change in the rate of leaf water evaporation and would have increased 5. Lee-Thorp JA, Sponheimer iVI, Van der iVlerwe NJ. What do stabie isotopes the difference in enamel S^^O ratios between water-independent teii us about the hominid dietary and eooiogicai niches in the Piiocene? int J (evaporation-sensitive) Girattidae and water-dependent (evaporation- Osteoarchaeoi. 2003:13:104-113. http://dx.doi.org/10.1002/oa.659 insensitive) Hippopotimidae. Such an increase is not observable in this case. The 5'°0 ratios of Giraffa sp. (-2.2) and Hippopotamus gorgops 6. Van der iVlerwe NJ, Thackeray JF, Lee-Thorp JA, Luyt J. The carbon isotope ecoiogy and diet ot Austraiopithecus africanus at Sterktontein, South (-5.5) from Lower Bed II differ by 3.3. Among the modern fauna, the S^^O Africa. J Hum Evoi. 2003;44:581-597. http://dx.dol.org/10.1016/S0047- ratios ot Giraffa cameiopardis (-3.6) and Hippopotamus amphibious 2484(03)00050-2 (-0.2) ditter by 3.4. Therefore the difference did not change, indicating that the humidity and temperature did not change. The relatively small 7. Noii D, Avery G. Protein poisoning and coastai subsistence. J Archaeoi Sei. difference between the 5"0 ratios of water-dependent and water- 1988:15:395-401. http://dx.d0i.0rg/i0.1016/0305-4403(88)90037-4 independent animals aiso indicates that the environment in each case 8. Lee-Thorp JA. The demise of 'Nutcracker iVlan'. Proc Nati Acad Sei USA. was not very dry and that water was permanently available. 2011 ;108:9319-9320. http://dx.d0i.0rg/i0.1073/pnas.1105808108

South Atrican Journal of Science Volume 109 \ Number 11/12 http://www.sajs.co.za 76 November/December 2013 Research Article Olduvai fossil isotopic ecology Page 14 of 14

9. Ceding TE, Manthi FK, Mbua EN, Leakey LN, Leakey MG, Leakey RE, et al. 28. Passey BH, Robinson TF, Ayliffe LK, Cerling TE, Sponnheimer M, Dearing Stable isotope-based diet reconstructions of Turkana Basin hominids. Proc MD, et at Carbon isofope fractionation befween diet, breath, and bioapatite Nati Acad Sei USA. 2013:110(26):10501-10506. http://dx.d0i.0rg/i0.1073/ in different . J Archaeol Scl. 2005:32:1459-1470. http://dx.doi. pnas.1222568110 org/1 C.I 016/j.jas.2005.03.015

10. Sponnheimer M, Alamseged Z, Ceding TE, Grine FE, Kimbel WH, Leakey 29. Quade J, Cerling TE, Barry JC, Morgan ME, Pilbeam D, Chivas AR, et al. A MG, ef at Isotopic evidence of eady hominin diefs. Proc Nati Acad Sei USA. 16- Ma reeord of paleodief using carbon and oxygen isotopes in fossil teeth 2013:110:10513-10518. http://dx.d0i.0rg/i0.1 C73/pnas.1222579110 from . Chem Geol (Isotope Geosc). 1992:94:183-192. http://dx.doi. 11. Cerling TE, Bovi/man JC, O'Neill JR. An isofopic study of a fluvial-lacustrine org/10.1 CI 6/C168-9622(92)90011 -X sequence: The Pllo-Pleistocene Koobi for a sequence. East Afdca. 30. Sponnheimer M, Lee-Thorp JA. Oxygen isotope ratios in enamel carbonate Palaeogeogr Palaeoclimatol Palaeoecol. 1988:63:335-356. http://dx.doi. and their ecological significance. J Archaeol Sei. 1999:26:723-728. http:// org/10.1016/0031-0182(88)90104-6 dx.doi.org/10.1C06/iasc.1998.C388 12. Cerling TE. Developmenf of grasslands and savannas in East Africa during the 31. Helliker BR, Ehleringer JR. Differential ^"0 enrichment of leaf cellulose in C3 Neogene. Palaeogeogr Palaeoclimatol Palaeoecol. 1992:97:241-247. http:// versus C, grasses. Functional Plant Biol. 2C02:29:435-442. http://dx.doi. dx.doi.org/10.1016/0031 -0182(92)90211 -M org/10.1071/PP01122 13. Kingston JD, Marina BD, Hill A. Isotopic evidence for Neogene hominid 32. Levin NE, Cerling TE, Passey BH, Harris JM, Ehleringer JR. A stable paleoenvironmenfs in fhe Kenya Rift Valley Science. 1994:264:955-959. isotope aridity index for terrestrial environments. Proc Nati Acad Sei. http://dx.doi.org/1C.1126/science.264.5161.955 2006:1 C3:11201-11205. http://dx.d0i.0rg/i 0.1073/pnas.O6C47191C3 14. Wolde-Gabriel G, Haile-Selassie Y Rennie PR, Hart WK, Ambrose SH, Asfavii B, et al. Geology and palaeontology of the Late Miocene Middle 33. White TD, Ambrose SH, Suwa G, Su DF, DeGusta D, Bernot RL, et.al. ^ms^] Valley, Afar Rift, Ethiopia. Nature. 2001:412:175-178. http://dx.doi. Macrovertebrate paleontology and the Pliocene habitat of Ardipithecus org/10.1038/35084058 ramidus. Science. 2009:326:67-93. http://dx.doi.org/10.1126/science. 1175822 15. Wynn JG. Influence of Plio-Pleistocene aridification on human evolution: Evidence from paleosols of the Turkana Basin, Kenya. Am J Phys Anthropol. 34. TIeszen LL. Natural variation in the carbon isotope values of plants: 2004:123:106-118. http://dx.doi.ürg/10.1002/ajpa.10317 Implications for archaeology, ecology and palaeoeeology J Archaeol Sei. 1991:18:227-248. http://dx.d0i.0rg/i 0.1016/0305-4403(91 )9OO63-U 16. Lee-Thorp JA, Sponnheimer M. Contdbution of sfabie light isotopes to paleoenvironmental reconstruction. In: Henke W, Tattersall TH, editors. 35. Van der Merwe NJ, Medina E. Photosynthesis and the "C/'^C ratios in Handbook of paleoanthropology vol.1. Bremen: Springer Verlag: 2007. p. Amazonian rain forests. Geochim Cosmochim Acfa. 1989:53:1091-1094. 289-310. http://dx.d0i.0rg/i0.1007/978-3-540-33761-4_9 http://dx.d0i.0rg/i C.1016/0016-7037(89)90213-5

17. Cerling TE, Hay RL. An isotopic study of paleosol carbonates from Olduvai 36. Ehleringer JR, Field CB, Lin ZF, Kuo CY Leaf carbon isotope and mineral Gorge. Quartemary Res. 1986:25:63-78. http://dx.doi.org/1C.1016/0033- composition in subtropical plants along an irradianee eline. Oecologia. 5894(86)9CC44-X 1986:70:520-526. http://dx.doi.org/10.1C07/BF00379898

18. Van der Merwe NJ, Cushing AE, Blumenschine RJ. Stable isofope ratios of 37. Ehleringer JR, Cooper TA. Correlations between carbon isotope ratio and fauna and the environment of palaeolake Olduvai. J Hum Evol. 1999:34:A24- microhabitat in desert plants. Oecologia. 1988:76:562-566. A25. http://dx.doi.org/10.1038/276815a0 38. Bowen GJ, Wilkinson B. Spatial distribution of 8"0 in meteoric precipitation. 19. VogelJC, Van der Merwe NJ. Isotopic evidencefor early maize cultivation in New Geology 2002:30:315-318. http://dx.d0i.0rg/i 0.1130/0091 -7613(2002) York State. Am Antiq. 1977:42:238-242. http://dx.doi.org/1C.2307/278984 C3C<0315:SDOOIM>2.0.CO:2

20. Van der Merwe NJ, Vogel JC. '^C content of human collagen as a measure 39. Dansgaard W. Stable isotopes in precipitation. Tellus. 1964:16:436-438. of prehistoric diet in woodland North America. Nature. 1978:276:815-816. http://dx.d0l.0rg/i 0.1111/J.2153-3490.1964 .tbOCI 81 .x

21. Van der Merwe NJ. Carbon isotopes, photosynthesis, and archaeology Am 40. Epstein S, Mayeda T Variation of '"0 content of wafers from natural sources. Sei. 1982:70:596-606. Geochim Cosmochim Acta. 1953:4:213-224. http://dx.doi.org/1C.1C16/ 22. Bocherens H, Billiou D, Mariotti A, Pathou-Mathi M, Otte M, Bojean D, 0016-7037(53)90051-9 et at Palaeoenvironmental and paiaeodietary implications of isotopic 41. Gat JR. Oxygen and hydrogen isofopes in the hydrologie cycle. Annu Rev Earth biogeochemistry of last interglacial Neanderthal and bones in PI Sei. 1996:24:225-262. http://dx.doi.Org/10.1146/annurevearfh.24.1.225 (). J Archaeol Sei. 1999:26:599-607. http://dx.doi. org/10.10C6/iasc.1998.0377 42. Koch PL. Isotopic reconstruction of past continental environments. Annu Rev Earth PI Sei. 1998:26:573-613. http://dx.doi.org/10.1146/annurevearth. 23. Richards MR Pettitt PB, Trinkhaus E, Smith FH, Paunovic M, Karavanic I. Neanderthal diet at Vindija and Neanderthal prédation: The evidence from 26.1.573 sfabie isotopes. Proc Nati Acad Sei. 2000:97:7663-7666. http://dx.doi. 43. Levin N, ZIpser E, Cerling TE. Isotopic composition of wafers from Ethiopia org/10.1073/pnas.120178997 and Kenya: Insights into moisture sources from Easfern Africa. J Geophys 24. Sullivan CH, Krueger HW. Carbon isotope analysis of separafe phases Res. 2009:114:D233C6. http://dx.doi.org/1C.1029/2009JDC12166 in modern and fossil bone. Nature. 1981:292:333-335. http://dx.doi. 44. Cerling TE, Harris JM, Hart JA, Kaleme R Klingel H, Leakey MG, etal. Stable org/10.1038/292333a0 isotope eeology of the common hippopotamus. J Zool. 2C08:276:204-212. 25. Sullivan CH, Krueger \W. Carbon isotope ratios of bone apatite and animal diet http://dx.d0i.0rg/i 0.111 i/j.1469-7998.2C08.0045C.X reconstruction. Nature. 1983:301:177. http://dx.doi.org/1C.1C38/3C1177aO 45. Blumenschine RJ, Peters CR, Masao FT Clarke RJ, Deino AL, Hay RL, et 26. Lee-Thorp JA, Van der Merwe NJ. Carbon isotope analysis of fossil bone al. Late Pliocene Homo and hominid land use from western Olduvai Gorge, apatite. S Afr J Sei. 1987:83:712-715. Tanzania. Scienee. 2003:299:1217-1221. http://dx.d0i.0rg/i 0.1126/seienee. 1075374 27. Lee-Thorp JA, Van der Merwe NJ. Aspects of the chemistry of fossil and modern biological apatites. J Archaeol Sei. 1991:18:343-354. http://dx.doi. 46. White F The vegetation of Africa. Nafural Resources Research. Vol. 2C. Paris: org/10.1016/0305-4403(91 )9007C-6 UNESCO: 1983.

South African Journal of Science Volume 109 \ Number 11/12 http://www.sajs.oo.za 77 November/December 2013 Copyright of South African Journal of Science is the property of Academy of Science of South Africa and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.