The results show a major difference in both habitat structure and species composition between the earliest and latest time periods. Habitats are more open at 2.0-1.6 Ma as evinced by greater percentages of grazing fauna and the presence of few arboreal and frugivorous taxa. This is not supported by directed change from closed and wet to dry and open in any of the time successive members of the basins examined here,
Fig. 14.8 Correspondence analysis of the adaptations with the modern localities and the late time period fossil sites. Rainforests and grasslands are at opposite ends of the spectrum and outlined in solid lines. The fossil sites are distributed near open, highly seasonal modern habitats, i.e., grasslands. Although only the Upper Burgi and KBS collection areas fall within modern secondary grassland ranges. The Olduvai sites have extremely large numbers of grazers compared with all extant localities. It is perhaps expected due to depositional environments that some of the Turkana Basin sites are more like the Kafue Flats with abundant wetland grasslands and flood plains.
Fig. 14.9 Tree diagram of a cluster analysis based on the Dice/Sorenson Index of species similarity among fossil sites used in this study. The major difference among the sites is geographical rather than chronological (South Africa vs. East Africa). However, the mid-to late Turkana Basin sites from which Homo erectus/ergaster has been recovered form a smaller cluster away from the Kalochoro Member and all other localities. The early Turkana Basin sites represent Australopithecus and Paranthropus sites.
although overall it is apparent that environments became more seasonal and dry. This correlates to other findings (e.g., Behrensmeyer et al., 1997; Spencer, 1997; Reed, 1997, 1998; Bobe and Behrensmeyer, 2004; Bobe and Leakey, 2009) of increased numbers of taxa adapted for grazing and as an extension suggests fewer trees in more open grasslands. However, the increase in aquatic and fresh grass grazing fauna, particularly at West Turkana and Hadar in the later time period also indicates that while habitats were relatively treeless, they were wet, in contrast to other areas of Africa that were drying out. These data support findings based on the analysis of lacustrine sediments indicating the presence of a large lake in the Turkana basin fed by the proto-Omo River between 1.9 and 1.7 Ma and a sizeable lake in the Olduvai basin between 1.92 and 1.7 Ma (Potts, 1998a; Trauth et al., 2005). Other slight differences in adaptive patterns between formations demonstrate the importance of accounting for local environmental influences on the structure and composition of faunal assemblages and not assuming that all East African habitats were changing at the same rates or in the same directions.
While the results of this study apply to a relatively small range of time, it can be concluded that the effects of climate change as indicated by deMenocal (1995) at 2.8 Ma can be observed in fauna before and after this time period in a variety of sites. However, this did not occur at the same time at each locality. Southern African localities, as one might expect from temperate sites, appear to be more seasonal and perhaps drier than penecontemporaneous East African fossil sites. This does not mean the South African sites were accumulated in more open environments, however. While the East African sites are fluvial or lacustrine deposits, the South African deposits have been accumulated by carnivores in caves. Despite this, the South African sites have aquatic taxa represented and the difference in habitats seems to be more a result of latitude rather than taphonomy. Additionally, a different faunal community in the two regions is indicated by the primates alone in that there are more colobine taxa in East Africa traditionally thought to inhabit more closed environments.
Based on our results it seems likely that habitats in both east and south Africa were gradually changing but, perhaps more importantly, fluctuating, over long periods of time, resulting in different species compositions in mammal communities dependant on geographic region. Wynn (2004) has suggested that mosaic environments consisting of drier and more open habitats on floodplains, in contrast to wetter and more closed refuge habitats near rivers and lakes across this time period resulted in an increased number of localized ecological niches for species to fill. This premise could also account for species similarities being higher among sites and regions in close proximity (e.g., Olduvai, early Hadar, and the Shungura Formation).
There appear to be only slight environmental changes that were associated with the emergence of species of the genus Homo, and indeed also with species of Paranthropus that appeared at roughly the same time. Habitats tended to become slightly more xeric and open, but not at the same time in different areas. More open habitats are also an indication of greater seasonality in rainfall as higher levels of annual rainfall in less seasonal environments support greater woody vegetation. Rainforests have perhaps a 3 month "dry" season whereas open woodlands can have dry seasons lasting up to 8 months in either a unimodal or bimodal pattern. There are increases in grazing mammals at 2.35 Ma (Kalochoro), 2.0 Ma (Upper Burgi), 2.33 Ma (Hadar), and 2.4 Ma (Shungura F). As Olduvai Bed I is the first in that sequence, there is nothing with which to compare to what came before in that region, but the grazing adaptations as represented by the fossil assemblage in Olduvai Bed I times exceeds 40%. Sterkfontein 4 has high numbers of grazing animals, but lianas have also been recovered from the site (Bamford, 1999). Thus, the mammal community gives evidence of drier climate, but a probable ecotone habitat in which grasslands were contiguous with forests. Although the presence of Homo from Sterkfontein 4 is debated, there is an increase in mammals indicating more arid habitats compared with Makapansgat 3.
We may have evidence of a new hominin taxon appearing in the fossil record as early as 2.6 Ma with the manufacture of stone tools. Thus, habitats in the middle time period that are associated with Homo and other hominins are likely more seasonal and more open than earlier sites from which Australopithecus afarensis has been recovered in East Africa. The key to understanding climatic factors that are associated with the emergence of Homo and other taxa such as A. garhi and P. aethiopicus, both recovered from sediments of 2.5 Ma, is to comprehend the causes of the extinction of earlier aus-tralopithecines. This is a complex issue but with more research in individual basins, an understanding of the patterns of localized tectonic influenced changes as well as climate changes may shed light on this issue. We know that during the period from ~3.0 to 2.0 Ma, fossil and other evidence reveal habitat changes and species turnovers. This does not appear to be a pan-African event where species changed in all regions at once. For example, there is evidence at Hadar of a mammal species turnover at ~3.0 Ma that did not involve A. afarensis (Reed, 2008). On the other hand, it has been reported that there are morphological differences in the mandibles of A. afarensis before and after this faunal shift (Lockwood et al., 2000).
While there were no stark divergences in habitat between members in which H. erectus fossils are found (upper Burgi, KBS, Kaito, Olduvai Bed II, Okote, Natoo) compared to those from which H. erectus has not been recovered (Kalochoro, Member G, Olduvai Bed I), it is clear that the composition of mammalian assemblages differ, both with respect to time and location. While the Olduvai Beds group together in the cluster analysis, they are not similar to one another. Furthermore, there was no abrupt change in fauna (in terms of speciation, extinction or migration events) around the time in which H. erectus first appears (~1.9Ma). It is interesting however, that H. erectus has been recovered from all members of the middle and late Turkana Basin cluster (sensulato) in Fig. 14.8 except for the Kalochoro Member, which is the most distant site in the cluster.
The results of this study emphasize prolonged turnover of fauna in the late Pliocene to early Pleistocene, correlated with gradual and fluctuating shifts in paleoclimate and thus habitat representation. It is clear that while environments across Africa were generally becoming drier and more open, large mammals (including hominins) existed in a variety of habitats. And as the Hadar record shows, individual hominin species were able to live in a variety of environments (Reed, 2008).
Was this article helpful?