Neogene biomarker record of vegetation change in eastern Africa

The evolution of C4 grassland ecosystems in eastern Africa has been intensely studied because of the potential influence of vegetation on mammalian evolution, including that of our own lineage, hominins. Although a handful of sparse vegetation records exists from middle and early Miocene terrestrial fossil sites, there is no comprehensive record of vegetation through the Neogene. Here we present a vegetation record spanning the Neogene and Quaternary Periods that documents the appearance and subsequent expansion of C4 grasslands in eastern Africa. Carbon isotope ratios from terrestrial plant wax biomarkers deposited in marine sediments indicate constant C3 vegetation from ∼24 Ma to 10 Ma, when C4 grasses first appeared. From this time forward, C4 vegetation increases monotonically to present, with a coherent signal between marine core sites located in the Somali Basin and the Red Sea. The response of mammalian herbivores to the appearance of C4 grasses at 10 Ma is immediate, as evidenced from existing records of mammalian diets from isotopic analyses of tooth enamel. The expansion of C4 vegetation in eastern Africa is broadly mirrored by increasing proportions of C4-based foods in hominin diets, beginning at 3.8 Ma in Australopithecus and, slightly later, Kenyanthropus. This continues into the late Pleistocene in Paranthropus, whereas Homo maintains a flexible diet. The biomarker vegetation record suggests the increase in open, C4 grassland ecosystems over the last 10 Ma may have operated as a selection pressure for traits and behaviors in Homo such as bipedalism, flexible diets, and complex social structure.

stratigraphically continuous latest Miocene to Pleistocene sites in eastern Africa.
Pollen and phytolith records from floodplain sediments, paleosols, or lakes (21)(22)(23) reflect local vegetation, whereas distal records, such as those from marine cores, represent regional vegetation (9,24) and are subject to greater dispersal bias (25). In eastern Africa, Raymonde Bonnefille's perseverance in extracting pollen from seemingly unpromising terrestrial sediments has yielded vegetation reconstructions at numerous Plio-Pleistocene hominin sites (22,23,26,27). The pollen data illustrate that, within the shift toward more arid conditions over the past ∼5 My, vegetation at individual sites remained highly variable through time and, at any given time interval, there was heterogeneity between sites.
Phytoliths are microscopic opaline silica bodies with taxonomically diagnostic morphologies (28). Their presence in the geologic record is powerful because of their greater resistance to degradation than pollen under oxidative conditions and, critically, because of their ability to distinguish C 3 and C 4 grasses apart from one another (29). Increasingly, phytoliths are being used to reconstruct Miocene to Pleistocene ecosystems in Africa (5,14,29,30).
Another perspective on hominin paleoenvironments comes from carbon isotope analysis. Plants using the Calvin cycle, or C 3 photosynthetic pathway, including trees, shrubs, forbs, and cool season grasses, have more negative carbon isotope values, whereas plants using the Hatch−Slack cycle, or C 4 pathway, including warm season grasses and sedges, have more positive carbon isotope values (31). In eastern Africa today, the carbon isotope ratio of plants can be used to discriminate between grassy and woody vegetation because nearly all grasses below 3,000 m use the C 4 pathway, whereas nearly all woody vegetation uses the C 3 pathway (32,33). Direct reconstruction of vegetation in eastern Africa using carbon isotopes comes almost exclusively from pedogenic carbonate, which, when formed in equilibrium with soil CO 2 , reflects the proportion of C 3 to C 4 vegetation in past landscapes (34,35). Carbon isotope data from pedogenic carbonate have contributed substantially to our understanding of vegetation at Miocene to Pleistocene hominid fossil sites (ref. 4

and references therein).
Carbon isotope data from fossil tooth enamel have been fundamental to understanding dietary changes in large mammals in eastern Africa (e.g., refs. 36 and 37) and, more broadly, paleoenvironments at eastern African fossil sites. The carbon isotope ratio of fossil enamel reflects the proportion of C 3 and C 4 vegetation in the diets of mammals (38)(39)(40). Large isotope data sets representative of mammalian fauna provide important information about the relative contributions of C 3 and C 4 biomass on the landscape (e.g., refs. [41][42][43], although the dietary information does not directly reflect quantitative vegetation distributions. Another carbon isotope archive for reconstructing eastern African paleovegetation that has emerged as a powerful tool in the past decade is terrestrial plant waxes from sediments (7,8,44). Plant wax biomarkers derive primarily from epicuticular leaf waxes that protect the leaf tissue from abrasion by dust; attack from insects, microbes, and fungi; and water loss from the leaf surface. The waxes are primarily long chain, linear or normal (n-)alkyl lipids made up of carbon and hydrogen atoms (45). Terrestrial plant waxes are commonly transported via fluvial and aeolian processes and are preserved in terrestrial and marine sediments (46)(47)(48). Marine sediments typically accumulate slowly and continuously and can be well dated. Thus, terrestrial plant waxes from marine sediments serve as useful archives of regional vegetation change (49), making them an excellent complement to terrestrial sedimentary deposits. Paleovegetation records from terrestrial sediments are biased toward lowland vegetation simply because sediment accumulation occurs in basins. Thus, carbon isotopes in plant waxes from terrestrial sites provide a local vegetation signal (44), whereas waxes from marine sediments provide a regionally integrated signal.
Beyond elucidating the paleoenvironments and diets of eastern African fauna, carbon isotope data from pedogenic carbonates, fossil enamel, and plant waxes, were critical in the discovery of the global expansion of C 4 grasslands in the late Miocene (see review in ref. 50). This includes pedogenic carbonate records from the Siwaliks (51)(52)(53), North America (54), and Africa (52,55). Tooth enamel data were also used to document the rise of grasslands between 8 Ma and 5 Ma in these same regions (56)(57)(58). More recently, carbon isotope records from plant waxes in the Siwaliks (48), North America (49), and eastern Africa (7) have further characterized this global ecological transition. Several of these isotopic records indicate minor amounts of C 4 vegetation in the middle Miocene of North America (49,54) and possibly in eastern Africa (52). Plant wax (7) and more recent tooth enamel data (36,59) suggest the rise of C 4 grasses in eastern Africa was earlier than in other parts of the world, yet there is no continuous vegetation record from the Neogene, and data before 10 Ma are sparse.
Characterizing the timing and nature of the C 4 grassland expansion and evaluating its potential role in the evolution of mammals and humans in eastern Africa requires a regional, full Neogene record of vegetation. The patchy and discontinuous distribution of early and middle Miocene fossil sites in eastern Africa precludes terrestrial sediments from providing such a record. Marine sediments, on the other hand, offer the opportunity to produce a long, continuous, regionally representative vegetation record from the Neogene to present. A series of Deep Sea Drilling Project (DSDP) cores proximal to eastern Africa were recovered from the Somali Basin (western Indian Ocean), the Red Sea, and the Gulf of Aden in the early 1970s. One of these cores, DSDP Site 231 from the Gulf of Aden, has provided vegetation data going back 11.3 My (7).
Here, we present an ∼24-My vegetation record from carbon isotopes of n-alkane plant waxes from four marine cores that collectively span the Neogene to present. Volcanic ashes in the cores correlated to those deposited in the Turkana and Awash Basins provide evidence that plant waxes are sourced from eastern Africa (60). Vegetation data from the cores provide new insight into the timing, nature, and spatial patterns of C 4 grassland expansion in eastern Africa by significantly expanding the spatial and temporal range of existing data. Early and middle Miocene vegetation data from the Somali Basin address whether C 4 vegetation was present in Africa before the late Miocene. We use the carbon isotope data to evaluate the role of vegetation change in mammalian evolution, and we evaluate C 4 grassland expansion at the macroevolutionary scale in eastern Africa and in a global context.

Site Selection and Interpretive Framework
The four cores, from DSDP Sites 228, 232, 235, and 241, range in latitude from 19.1°N to 2.4°S ( Fig. 1) and have basal ages of 3.9 Ma, 5.5 Ma, 13.5 Ma, and 23.4 Ma, respectively, from the sampled intervals. Age−depth models are shown in Fig. S1, data are given in Dataset S1, and methods are described in SI Materials and Methods. Sample ages, presented in Dataset S1, were interpolated using the age−depth models for each site based on calcareous nannofossil biostratigraphy updated to the most recent biohorizon ages (61,62), and from radiometrically dated volcanic ashes in the cores (60). Each site was nominally sampled at ∼0.5to 1.4-My resolution to capture long-term, Neogene paleovegetation trends and provide a broad temporal context to shorter but higher-resolution records from terrestrial and marine archives. DSDP site information for each core is given in Dataset S2 and sample lithologies are summarized in Dataset S3. We interpret regional vegetation using carbon isotope data from the n-C 31 alkane because it best reflects the actual distribution of vegetation on the landscape. Garcin et al. (63) found that, among n-alkane homologs commonly measured for carbon isotopes, the concentration of the n-C 31 alkane is nearly identical in C 3 and C 4 plants in Africa, effectively removing plant wax production biases related to plant functional type. However, the significant differences in concentrations of other n-alkane homologs between C 3 plants and C 4 grasses also contain powerful information for reconstructing vegetation. For example, C 4 grasses have higher concentrations of longer chain n-alkanes (C 33 and C 35 ) than C 3 plants. Conversely, C 3 plants have higher concentrations shorter chain lengths (e.g., C 27 or C 29 ) (63)(64)(65). Furthermore, C 35 is absent in modern rainforest taxa, extremely low (∼3% of total n-alkanes) in savanna trees, shrubs, and herbs, but relatively abundant (up to 20% of total n-alkanes) in C 4 grasses (64,65). As a result, the δ 13 C value of the n-C 35 alkanes (δ 13 C C35 ) is extremely responsive to the presence of C 4 vegetation on the landscape. This allows us to use the δ 13 C value of the n-C 35 alkane as a highly sensitive indicator of C 4 vegetation on the landscape. Additionally, the absence of the n-C 35 alkane and dominance of the n-C 29 alkane in plant wax molecular distributions are characteristic of modern rainforest plants (64). We suggest that similar patterns in n-alkane distributions from ancient sediments coupled with carbon isotope values representative of rainforest taxa would be indicative of rainforest vegetation in the past.

Results and Discussion
Concentration and carbon isotope data for the n-C 29 , n-C 31 , n-C 33 , and n-C 35 alkanes extracted from marine core sediments (n = 46) are presented in Datasets S4 and S5, respectively. Molecular distributions indicate long chain n-alkanes are derived from terrestrial plants (SI Materials and Methods and Fig. S2). The average chain length (ACL) and carbon preference index (CPI) are coarse metrics commonly used to evaluate plant wax distributions (Dataset S4). The former reflects the relative amounts of long and shorter n-alkanes, whereas the latter measures the strong biosynthetic preference for odd-numbered n-alkanes in plants versus the approximately equal preference in petroleum (see also SI Materials and Methods). The ACL and CPI for all samples are similar to modern African plant values (64), indicating n-alkanes are derived from terrestrial plants with no indications of significant diagenesis or addition of exogenous long chain n-alkanes (e.g., microbial sources or petrogenesis). The oldest sample analyzed, from Site 241, is characterized by a large unresolved complex mixture (UCM) in the range of short-chain to midchain lengths that does not appear to affect the δ 13 C values of longchain plant-derived n-alkanes (Fig. S2). Other samples exhibit small UCMs in the short-chain and midchain length range (Fig. S2).
Carbon Isotope Values and Molecular Distributions from 24 Ma to 10 Ma. The δ 13 C C31 values from 24 Ma to 10 Ma come from the Site 235 and 241 cores in the Somali Basin and are very depleted and nearly invariant, ranging from −33.9‰ to −32.2‰ (Fig. 2), indicative of C 3 ecosystems. The range of δ 13 C C35 values before 10 Ma is similarly small, from −33.3‰ to −31.3‰ ( Fig. S4 and Dataset S5). The other n-alkane homologs, C 29 and C 33 , have a pooled range of −34.0‰ to −30.8‰. Given the similarity in δ 13 C values of the four homologs through time and the relative geographic proximity of Sites 235 and 241, we combine data from both sites into a single Somali Basin record (Fig. S4). Combining the two sites into a single record is further supported by clay mineralogy of the Somali Basin sediments, which indicates a similar sediment provenance of the Horn of Africa and possibly Central Africa for both sites (67). All values from Somali Basin samples are indicative of C 3 ecosystems in eastern Africa before 10 Ma. The limited range within each homolog from 20 Ma to 10 Ma suggests that the δ 13 C values likely define endmember values for C 3 ecosystems in eastern Africa. The persistently low δ 13 C values of the n-C 35 homolog are particularly diagnostic, indicating no isotopic evidence for C 4 vegetation in east Africa from 24 Ma to 10 Ma.
The average molecular distribution of n-alkanes for pre-10-Ma samples has a n-C 31 maximum, followed in order of abundance by n-C 29 , n-C 33 , n-C 27 , and n-C 35 (Fig. S5). Although the data do not preclude minor amounts of rainforest vegetation, the n-alkane distribution does not support widespread rainforests in eastern Africa, nor do measured δ 13 C C31 values, which do not reach the low values characteristic of rainforest vegetation (see figure 6a in ref. 64 for reference n-alkane distributions). Rather, the molecular distributions and δ 13 C C31 values reflect a combination of nonrainforest trees, shrubs, herbs, and C 3 grasses. Relative contributions of these C 3 plant types cannot be ascertained from the molecular distribution data. Two possible exceptions to this are the molecular distributions from samples from 19.18 Ma (Site 241, sample KU357) and 1.15 Ma (Site 228, sample KU315), which have distributions similar to modern rainforest plants (Fig. S6). However, the δ 13 C values are −32.4‰ and −25.2‰, respectively, neither of which reflect modern rainforest plants values of −35.6 ± 2.8‰ (corrected to preindustrial atmospheric δ 13 C values) (64).

Comparing Biomarker Data with Existing Vegetation Records.
The absence of C 4 vegetation at the regional scale before 10 Ma based on n-alkane carbon isotope and molecular distribution data is supported by vegetation reconstructions from paleobotanical records of predominantly forest or wooded ecosystems in Ethiopia, Kenya, and Uganda (15-17, 19, 27).
A single pedogenic carbonate value from the Hiwegi Formation (17.8 Ma) on Rusinga Island in Lake Victoria has a δ 13 C PC value of −5.7‰ (68). However, the corresponding δ 13 C value from coexisting soil organic matter is −24.3‰, consistent with water-stressed C 3 vegetation or minor amounts (ca. 15%) of C 4 vegetation. There is an enrichment in 13 C between soil organic matter and pedogenic carbonate that is a function of kinetic processes (diffusion of CO 2 in soils) and equilibrium isotopic exchange during carbonate precipitation that results in a theoretical enrichment of +14‰ to +17‰ (35). The observed enrichment factor of +18.6‰ for the Hiwegi Formation sample falls slightly outside this theoretical value. The other δ 13 C PC values that suggest C 4 vegetation existed before 10 Ma come from the Tugen Hills (Baringo Basin) in the central Kenyan Rift Valley (52,69). Five pedogenic carbonates from the ∼15.5-Ma sediments in the Muruyur Formation have δ 13 C PC values ranging from −7.4‰ to −5.2‰. Only one of the five samples (MU 453, δ 13 C PC = −7.4‰) has a corresponding organic matter value with an enrichment value within the theoretical range of +14-17‰, making it the most reliable value from this period (70). Two δ 13 C PC values around −6.5‰ from the ∼10.5-Ma section of the Ngorora Formation, also in the Tugen Hills, indicate minor amounts of C 4 vegetation. A corresponding organic matter δ 13 C value from one of the samples is −24.3‰, which falls within the range of water-stressed C 3 vegetation. Even if all pre-10-Ma pedogenic carbonate values >−8‰ are accepted and considered with other paleovegetation data from Rusinga and the Tugen Hills, the picture that emerges from both sites can be summarized as wooded to forested environments with minor amounts of C 4 vegetation present. The early presence of minor amounts of C 4 vegetation in eastern African is entirely consistent with an emergence of the C 4 photosynthetic pathway in the late Eocene or early Oligocene (71,72). Despite an early emergence, C 4 grasses remained minor constituents of global ecosystems until their late Miocene expansion (73).
Fort Ternan (14 Ma) is the final middle Miocene site that warrants discussion, because well-preserved grass macrofossils were found there (18) and 54% of total pollen comes from grasses (27). Pedogenic carbonate values range from −14.9‰ to −9.8‰, leading to the conclusion that C 3 grasses were present at the site (74). Fort Ternan illustrates that, without plant fossils, whether from pollen, phytoliths, or macrofossils, C 3 grasses are invisible when viewed solely through an isotopic lens. The well-documented lack of pollen and the rarity of plant macrofossils in the Neogene of eastern Africa demonstrate the need to find other proxies that will enable us to investigate the Cenozoic history of grasses. Phytoliths could play a pivotal role; they have proven to be robust indicators of grasslands in the Cenozoic in other parts of the world (75). Other proxies for detecting C 3 grasses, where stable isotope methods are silent on the matter, include grass cuticles (76) or, possibly, n-alkane distributions (64,65).

Onset of C 4 Grassland Expansion at 10 Ma and Effects on Fauna.
The onset of C 4 grassland expansion in eastern Africa occurs at ∼10 Ma. It is marked by a sharp 3.1‰ increase in the δ 13 C value of the n-C 35 alkane that occurs between samples dated at 10.08 Ma and 9.65 Ma (Fig. 3). This is the first occurrence, to our knowledge, of a δ 13 C value greater than −30‰ in the n-alkane record. Enrichment in 13 C occurs across all other alkane homologs with deceasing magnitude from C 33 to C 29 (Fig. S4), which follows the predicted pattern based on chemotaxonomic differences in n-alkane production observed in modern C 3 and C 4 plants in Africa (63) and unequivocally dates the onset of expansion of C 4 grasses in eastern Africa.
A remarkable consequence of the appearance of C 4 grass in eastern Africa is the immediate and, in some cases, dramatic dietary response of large herbivores (36). The n-alkane δ 13 C values reveal a synchroneity between the onset of expansion of C 4 grasses and dietary shifts in proboscideans (Fig. 3) and other large herbivore lineages (36). Most striking is the shift from C 3 -dominated to C 4 -grazing diets in equids documented in Kenya between 9.9 Ma and 9.3 Ma. Other large mammals that followed suit by 9 Ma include bovids and rhinocerotids. Equids only arrived in Africa around 10.5 Ma, so there is no dietary record in eastern Africa for them in the earlier Neogene (59). Proboscideans, which include modern elephants and their ancestors, are an Afrotherian order for which dietary records exist from the early Miocene to present. Fig. 3 illustrates the dietary evolution of proboscideans from eastern and central (i.e., Chad) Africa over the past 17 My based upon carbon isotope data from their teeth (Dataset S7, compiled primarily from ref. 77). Small amounts of C 4 vegetation are present in proboscidean diets beginning at 9.9 Ma. The dominant early and middle Miocene proboscideans, Deinotheriidae and Gomphotheriidae, had brachydont molars with transverse lophs for shearing vegetation (78,79). Although the inclusion of some C 3 grass in their diets cannot be ruled out, their craniodental morphology indicates they were browsers. If C 3 grasses were abundant at this time, proboscideans did not capitalize on this dietary resource and instead waited until the appearance of C 4 grasses at ∼10 Ma to begin grazing. In fact, as one of the dominant early Neogene herbivores, it is possible that browsing proboscideans maintained or even increased grasslands, as their extant browsing relatives, Loxodonta africana, do today (80).
Carbon Isotope Data from 10 Ma to Present. The nature of grassland expansion after 10 Ma. The regional signal from the Somali Basin cores shows that C 4 vegetation increases linearly from 10 Ma to present ( Fig. 2 and Fig. S7). Post-10-Ma δ 13 C C31 values from the Somali Basin Sites 235 and 241 have nearly identical slopes, with R 2 values of 0.87 and 0.97, respectively (Fig. S7A). When combined, the Somali Basin samples yield a slope that indicates a δ 13 C increase of 1.1‰ per million years (R 2 C31 = 0.92) or 8.0% C 4 vegetation per million years (Fig. S7B). Adding the 10 samples from Site 228 (0.2-3.9 Ma) yields a similar trend, with an increase of +1.0‰ per million years (R 2 C31 = 0.88) or 7.3% C 4 vegetation per million years. From 9.7 Ma to 5.1 Ma, Somali Basin δ 13 C C31 values increase by 3.8‰, whereas δ 13 C C33 and δ 13 C C35 increase by 5.3‰ and 5.0‰, respectively, indicating significant increases in C 4 biomass during the late Miocene.
Molecular distributions suggest the increase in C 4 biomass detected from plant wax isotopes reflects increases in grasses rather than C 4 shrubs, such as arid-adapted shrubs found in the families Chenopodiaceae and Amaranthaceae. This is because the marked increase in n-C 33 and n-C 35 alkane relative abundances in post-10-Ma samples (Fig. S5) resembles n-alkane distributions in extant African C 4 grasses (65), whereas limited data of n-alkane distributions from C 4 shrubs (Amaranthaceae spp.) show they produce little to no n-C 33 and n-C 35 alkanes (81). We therefore interpret the increase in δ 13 C values and relative abundances of n-C 33 and n-C 35 alkanes after 10 Ma to primarily reflect an increase in C 4 grasses. Interestingly, pollen from Chenopodiaceae and Amaranthaceae occurs in variable and occasionally high proportions (20-80%) in the late Miocene to Pliocene pollen record from DSDP 231 (9). Crassulacean acid metabolism (CAM) plants can have δ 13 C values more positive than C 3 plants; however, today, their overall contribution to biomass in the region is small. Thus, CAM plants likely contributed minor or insignificant amounts of n-alkanes to the sediments in the Somali Basin cores (Sites 235 and 241).
Carbon isotope data from Sites 228 and 232 encompass the latter part of the C 4 expansion period and lie geographically much farther north than Sites 235 and 241 (Fig. 1). Unlike the deep water Somali Basin sites, sediments from Site 228 in the Red Sea include fluvial sediments from the (paleo) Sudanese Delta (82). Site 228 samples have a higher CPI (generally >8) than samples from the other three sites (Dataset S4), probably related to differences in plant wax source and fluvial transport and depositional processes. Despite the vast distance separating the Somali Basin from the Red Sea, carbon isotope values from the two regions follow the same trend (Fig. 2).
Site 232 is located at the eastern entrance to the Gulf of Aden close to the Arabian Peninsula (Fig. 1). The δ 13 C C31 values from this site, with an estimated age of 5.54-0.39 Ma, range from −28.5‰ to −25.3‰. Although the core contains at least four ash layers correlated to the Turkana Basin (60), it is also highly likely that there is plant wax input from the Arabian Peninsula and, to a lesser extent, Southern Asia. This is based on the proximity to those areas and regional wind patterns (e.g., figure 1 in ref. 83). Provenance studies from nearby marine cores support this interpretation (84,85). The Arabian Peninsula and Southern Asia (i.e., the Siwaliks) both had significant amounts of C 4 vegetation by 6 Ma, and therefore it is impossible to disentangle contributions from the three source areas (51,86). Because of this potential complication, we refrain from making any interpretation of eastern Africa vegetation based on data from Site 232.   Comparison with existing paleoenvironmental data from 10 Ma to present. An extensive carbon isotope record from plant wax biomarkers exists for another Gulf of Aden core, Site 231, which is situated closer to the Horn of Africa than Site 232 (7). Site 231 biomarker data come from n-C 30 alkanoic acids, also derived from terrestrial plants, but with different isotope systematics and endmember values than n-C 31 alkanes (87). Despite these differences, general comparisons can be drawn between the two biomarker data sets. Both show an overall trend toward more positive δ 13 C values through time from ∼10 Ma to present (Fig. S4). The Somali Basin data exhibit a larger range (9.5‰ in the n-C 31 alkane, from −33.9‰ to −24.4‰) than Site 231 data (7.9‰ in the n-C 30 acid, from −30.2‰ to −22.3‰), despite the former having a smaller sample size (n = 26) than the latter (n = 372). The absence of 24-to 12-Ma data from Site 231, a period characterized by exclusively C 3 vegetation, may explain the lack of more negative values characteristic of the n-alkane data in the early Neogene ( Fig. S4 and Dataset S5). Site 231 δ 13 C values are generally more positive than the δ 13 C C31 values, probably, in part, due to differences in the lipid types (n-C 31 alkane vs. n-C 30 alkanoic acids). It may also be a result of different plant wax source areas. Site 231 δ 13 C values exhibit high variability over short timescales (∼10-40 ky) associated with orbitally driven vegetation change in eastern Africa.
One curious feature of our records is a notable lack of variability that might be anticipated to result from undersampling (aliasing) of orbital-scale climate variability. Where sampling resolution was adequate to resolve such higher-frequency changes, plant wax biomarker records from Gulf of Aden Site 231 document large (±2-3‰) orbital-scale cycles over the last 4.5 Ma. Orbitally driven vegetation change has also been documented in lacustrine sediments from Olduvai Gorge (8). However, our low-resolution (∼0.5-to 1.4-My sampling) records consistently document gradual, smooth increases in C 4 vegetation from 10 Ma to present (Fig. 2). Given that large, orbital-scale variability is observed in marine and terrestrial archives when sedimentation rates are high and sequences are adequately sampled (7,8,44), we conclude that some other factor may be smoothing or obscuring this orbital signal in our records. All of our sites were rotary drilled and, in some cases, spot cored, and coringrelated disturbance was observed in places. The Somali Basin average sedimentation rates are low (2−3 cm/ky), so bioturbation may have blurred orbital-scale variability. These sites are also quite distal from the coast, and so the plant wax catchment areas are likely quite large, implicating large spatial averaging as well. What is remarkable, however, is the uniformity of the C 4 expansion timing and the slopes of the C 4 expansion trends observed across all sites (Fig. S7).
There are numerous isotopic records from pedogenic carbonate from the Plio-Pleistocene of eastern Africa and several from the late Miocene available for comparison with the Somali Basin n-alkane data. We use the δ 13 C PC data set from eastern Africa recently compiled by Levin (88) and include new data added from the Karonga Basin in northern Malawi, a fossil locality that critically links eastern and southern African hominin sites (89) (Fig. 3C). We include the Malawi data because of its proximity to Site 241 and therefore likelihood of being part of the plant wax source area.
Data from the Tugen Hills (9.4-5 Ma) range from −9.3‰ to −4.2‰ and have a mean of −7.5 ± 1.3‰ (1σ). These values indicate C 3 -dominated to mixed C 3 −C 4 ecosystems (52,55). At Lothagam, a site located in the Turkana Basin, δ 13 C PC values come from a slightly narrower age window of 8.4-5.7 Ma. They range from −9.0‰ to −1.3‰ with a mean of −5.0 ± 2.1‰ (90). At Lothagam, the landscape was more open than the Tugen Hills, with values reflecting mostly mixed C 3 −C 4 to C 4 -dominated ecosystems. Major increases in the number of species and the abundance of grazing mammals in the Turkana, Baringo, and Awash Basins during this period provide supporting evidence for the increase C 4 biomass and it's increasing significance as a dietary resource for African mammals (36,41,43).
Pedogenic carbonate records from 5 Ma to 1 Ma are abundant from hominin sites in eastern Africa, particularly from the Turkana and Awash Basins (4). The main trend is a shift toward increased C 4 vegetation from ∼4 Ma onward (Fig. 3C). Initially, the range of environments mainly comprises C 3 -dominated to mixed C 3 −C 4 ecosystems until about ∼2 Ma, when C 4 -dominated ecosystems become more prevalent and C 3 -dominated ecosystems essentially disappear from the record, with the exception of the Karonga Basin in northern Malawi (Fig. 3C). The diets of large herbivores in the Turkana Basin reflect this transition. From 4.1 Ma to 2.35 Ma, the fossil record is made up of taxa primarily with mixed C 3 −C 4 diets and transitions to one comprised of mostly C 4 grazers from 2.35 Ma to 1 Ma (37). Both diet and vegetation reconstructions from terrestrial sites are in agreement with the Somali Basin n-alkane data, which indicate continued increases in C 4 vegetation through the Plio-Pleistocene.
Vegetation Change and Human Evolution. The earliest purported hominins, Sahelanthropus tchadensis and Orrorin tugenensis, appeared in the late Miocene (7-6 Ma) (Fig. 3D). By that time (6.5 Ma), the Somali Basin n-alkane data show an increase to 19% C 4 vegetation on the landscape, or a 2.6‰ increase in δ 13 C C31 values above average early Neogene C 3 values based on the linear regression equation for all data (Fig. S7C). An even greater increase of 4.6‰ occurs in δ 13 C C35 values by this time. These two early possible hominins emerged amid a period of major change in faunal communities and diets driven by changes in ecosystem structure associated with the expansion of C 4 grasses. The dietary shift toward C 4 grazing was well underway in many mammalian lineages (36,43), but there are not yet any isotopic analyses of teeth to ascertain the diets of these early hominids.
However, extensive isotopic data from Plio-Pleistocene hominin teeth provide a clear picture of diets in different lineages, temporal trends, and dietary niche partitioning between Pleistocene genera (91)(92)(93)(94)(95)(96). Two early species (4.4-4 Ma), Ardipithecus ramidus and Australopithecus anamensis, are anomalous for their C 3 -dominated diets (−11.6‰ to −8.5‰). After 3.8 Ma, Australopithecines and Kenyanthropus platyops show a dramatic increase in isotopic range and, presumably, dietary breadth, with δ 13 C values ranging from −13‰ to −2.7‰. Australopithecus bahrelghazali from Toros-Menalla (Chad, central Africa) shows the earliest evidence, to date (3.6 Ma), of a hominin with a consistent, mostly C 4 -based diet (range −4.4‰ to −0.8‰). It is possible that the dietary evolution of central African hominins differed from those in eastern Africa, as is the case with the southern African hominins (97). In eastern Africa, the diet of Paranthropus evolves to become dependent on C 4 -based foods from 2.7 Ma to 1.5 Ma (92). Overall, the increase in C 4 -based foods in eastern African hominin diets mirrors the increasing prevalence of C 4 biomass on the landscape. Our genus, Homo, does not follow this trend.
The isotopic range of Homo diet remains variable and broad (−9.9‰ to −2.2‰, mean = 6.1‰), indicating dietary flexibility from ∼2.4 Ma to 1.3 Ma (95). Microwear studies suggest dietary differences between Homo species based on tooth wear, supporting the broad isotopic range (98). An important observation, then, is that Homo diet remained flexible during a time of rising environmental inflexibility imposed by increasing aridity and C 4 vegetation on the landscape. One critical element of hominin diet that remains unknown, and is especially important for Homo, is when meat became an important food source. Carbon isotopes do not provide information on trophic level; thus isotope-based reconstructions of hominin diet reflect some combination of plant foods and of higher trophic level foods such as insects, mammals, or aquatic fauna.
The plant wax data from the Somali Basin, along with pedogenic carbonate and Gulf of Aden (Site 231) plant wax data, show that continued increases in the proportion of C 4 grasses on the landscape ultimately influenced hominin diet, possibly from the inception of the lineage. The shift to more open environments presented hominins with new challenges as to how they acquired their food, avoided predators, and managed higher heat loads. The challenges posed by the changing environmental conditions were met with changes to diet (C 4 -based foods and, eventually, meat), the innovation of tools, and, ultimately, increased social complexity such as the cooperative acquisition of resources through foraging, scavenging, or hunting.
Global Context of C 4 Grassland Expansion in Eastern Africa. The Neogene plant wax record provides the opportunity to place the expansion of C 4 grasslands in eastern Africa into a global context. C 4 grassland expansion has been investigated most intensively in the Siwaliks of South Asia (48,51,53), the Americas (49,54,(99)(100)(101), southern Africa (102,103), and, to a lesser degree, China (104,105), the Mediterranean (106), and the Middle East (85,86). The onset of expansion at 10 Ma in eastern Africa precedes by at least 2 My that of well-dated terrestrial records from the Siwaliks (∼8 Ma), North and South America (∼8-6 Ma), and South Africa (∼8-7 Ma).
An important feature of the eastern Africa biomarker record is that it captures the transition from isotopically invariant C 3 -dominated ecosystems during the first part of the Neogene to the period of dynamic change after 10 Ma (Fig. 2). The expansion of C 4 grasslands in eastern Africa differs markedly from other regions around the world in its earlier onset and the monotonic increase in C 4 grasses at the regional scale. The mode of expansion differs from the step-like change that occurred in the Siwaliks, arguably the most well documented case of grassland expansion (51). It also differs from the Great Plains of North America, which shows possible minor amounts of C 4 vegetation as early as ∼23 Ma but an expansion beginning around 7 Ma, based on the pedogenic carbonate record (54). A North American vegetation reconstruction from n-C 31 alkane data from a Gulf of Mexico core (DSDP Site 94) also suggests minor amounts of C4 vegetation by 9.1 Ma or possibly earlier (49). Two marine sediments cores (Ocean Drilling Program Sites 1081 and 1085) spanning ∼10 degrees of latitude from southern Africa both indicate the onset of expansion at ∼7.5 Ma based on n-C 31 alkane data, lagging eastern Africa by ∼2.5 My (102,103). A Neogene vegetation record from western Africa could provide information on mechanisms behind the expansion in Africa because it would show whether or not expansion occurred synchronously across eastern and western Africa. Clearly, there is the need for continued efforts to produce well-dated records of C 4 expansion to identify additional regional, continental, or global patterns. This includes characterizing the vegetation and climatic conditions before the expansion, which were not uniform globally. Improved spatial and chronological resolution will greatly enhance our understanding of the driving mechanisms of grassland expansion.
Commonly invoked causes for C 4 grassland expansion include a late Miocene decrease in atmospheric CO 2 concentrations (pCO 2 ) or increases in aridity, seasonality of precipitation, fire, and herbivory (56,(107)(108)(109). These factors are not mutually exclusive; therefore, debate continues about the relative importance of each variable (110). Most are difficult to quantify in the geologic record, and all continue to benefit from new or improved proxy methods for reconstruction in deep time. This includes characterizing fire frequency and intensity, identifying C 3 grass abundance, quantifying seasonality and aridity, and refining pCO 2 estimates.

Conclusions
The Neogene n-alkane carbon isotope record from the Somali Basin shows no evidence for C 4 vegetation at the regional scale during the early or middle Miocene of eastern Africa, which is consistent with terrestrial isotopic, faunal, and floral records. The first evidence of C 4 vegetation is indicated by a 3.1‰ increase in the n-C 35 alkane homolog at 10 Ma. Molecular distributions suggest the increase in C 4 biomass was from grasses rather than from arid-adapted C 4 shrubs. From 10 Ma to present, C 4 grasses increase monotonically on the landscape at the regional scale, as evidenced in the agreement between δ 13 C records in cores stretching from the Somali Basin to the Red Sea. Despite the strong secular trend in the n-alkane marine core records, terrestrial records document persistent vegetation heterogeneity at the local scale, and the Site 231 biomarker record demonstrates dramatic vegetation changes driven by orbital cyclicity. Collectively, the n-alkane vegetation record and dietary reconstructions show that vegetation change beginning at 10 Ma influenced the evolution of mammalian diets and, later, hominin diets. The Somali Basin record demonstrates that C 4 grassland expansion in eastern Africa was different in timing and mode from other regions in the world. Although this does not preclude common, global phenomena (e.g., pCO 2 ) as important drivers or preconditions to the global expansion of C 4 grasslands, it further underscores regional variables as critical determinants for the timing and mode of late Neogene vegetation change.

Materials and Methods
All isotopic data are presented in delta notation, where δ 13 C = ([ 13 C/ 12 C sample ]/ [ 13 C/ 12 C standard ] -1) and the standard is VPDB. Plant waxes were extracted from cleaned and crushed marine sediments with organic solvents. The n-alkanes were isolated by silica gel column chromatography, identified by gas chromatography mass spectrometry, and quantified by comparison with an internal standard and standard mixtures of C 8 −C 40 n-alkanes. Carbon isotope ratios were measured using a gas chromatography−combustion interface coupled to an isotope ratio mass spectrometer. Carbon isotopes values were corrected based on values of isotopic standards using the methods of Polissar and D'Andrea (111). SI Materials and Methods contains a more detailed description of organic geochemical and stable isotope methods.