Chapter 9
Savannas did not feature during the dawn of the age of mammals following the demise of the dinosaurs 66 Ma. The first grasses had appeared by then, but remained a miniscule constituent of the vegetation. Through the Palaeocene and Eocene, forests and closed woodlands blanketed the Earth during the extremely hot and wet conditions that prevailed from the equator to the poles.1 Most of the modern tree families were represented by then, including the legumes. Grasses became prominent first in South America during the Oligocene around 30 Ma.2 By the mid-Miocene 12 Ma, grasslands had become sufficiently widespread in North America to support a herbivore fauna with dentition honed for grazing, dominated by horses. Grasses came to the fore in Africa during the late Miocene around 10 Ma, including species showing the C4 photosynthetic pathway, as recorded by plant wax biomarkers in marine sediments.3,4 After 8 Ma, C4 grasslands had expanded throughout the tropics and subtropics, under climates that became cooler and drier, especially in Africa.5 By the mid-Pliocene 3.5 Ma, savannas dominated by C4 grasses were widely established through most of Africa.
However, the record of past vegetation in Africa is fragmentary in time and space, dependent on where sediments retain plant remains. Studies have been focused on rift valley settings in eastern Africa, where sediments can be dated quite precisely from volcanic layers. However, gaps exist in the sediment record. Marine sediments deposited offshore provide a broad regional record of past vegetation in catchments of rivers based on the pollen and other biomarkers they contain, although these can get shifted around by ocean currents. Plant remains from the limestone caves in interior southern Africa where fossils are preserved are more difficult to date. Sites documenting vegetation changes from pollen and other residues become more widely available from late in the Pleistocene.
Grasses can be identified as a group from their pollen, but are not readily distinguished at tribal or generic level. Silica bodies (or phytoliths) contained in grass leaves support only tribal distinctions. However, C4 grasses leave a characteristic signature in the form of carbon isotope ratios, recorded in soil carbonates, organic matter and bones and teeth of the animals feeding on them (see Chapter 7). Woody plant families or genera can be distinguished from their pollen. Trees that are wind-pollinated produce profuse pollen, but those pollinated by insects, like the acacias, are more sparing of pollen and thus under-represented. Further records of past vegetation are provided by fossilised leaves or trunks (Figure 9.1) as well as from biomarkers like leaf waxes.
Figure 9.1
Fossil wood in Luangwa Valley, Zambia.
Savanna Expansion
Carbon isotope ratios in soils document the expansion of C4 grasses through the Pliocene into the early Pleistocene, from 5 to 2 Ma. This increase was probably partly at the expense of grasses with the C3 pathway.6 A further indicator of grassland expansion through this period comes from wear patterns on the teeth of herbivores supported by carbon isotope ratios in dental enamel. I will describe the evidence separately for eastern and southern regions of Africa because of the differences in the sources used.
Eastern Africa
During the late Oligocene around 28 Ma, broad-leaved deciduous woodland resembling miombo was present both in Ethiopia and Tanzania, and presumably in between.7,8,9 Acacias were also present by that time in Tanzania.10 Wet forest occurred near Lake Victoria ~18 Ma, but by 12 Ma had become replaced by a heterogeneous mix of forest, woodland and even savanna.9,10 However, the grass pollen at that time came solely from C3 grasses. By the late Pliocene ~3 Ma, C4 grasses had achieved overwhelming dominance.3,4,9,11 Tectonic movements related to rift valley formation contributed to shifts in local climates during the Pliocene, causing the eastern African interior to be drier than projected from global temperature conditions,12 supported by increased evidence of fires.13 This brought about a reduction in forest cover over eastern Africa.9 Spines and other structural defences against large herbivores became a feature of several plant lineages during this time.14 This infers increased representation of woody plants typical of dry/eutrophic savannas with structural rather than chemical defences against herbivores. The projected woody plant cover, estimated from stable carbon isotope ratios in soils, had fallen below 40 percent at Awash in Ethiopia, and perhaps around half of this in the Turkana region of northern Kenya, by 6 Ma.15 By the mid-Pliocene ~4 Ma, most of the modern plant families could be recognised from fossil wood at Laetoli in central Tanzania, including acacias.16 Genera present today that seem missing locally at that time include Balanites (desert date), Combretum (bushwillows), Croton and Euclea (gwarri).
The onset of Pleistocene glaciations after 2.7 Ma was associated with a sharp drop in tree pollen and corresponding rise in C4 grass pollen in marine deposits off eastern Africa.9 Chloridoid grasses indicative of dry conditions became prominent between 2.7 and 2.5 Ma. Acacias and wild myrrhs, typical of dry savanna, were notable in the woody vegetation around Olduvai, along with pollen from forest trees such as yellowwoods and junipers, probably present in nearby highlands.12,17 Within the Omo valley, grass pollen increased substantially in its representation between 2.8 and 2.35 Ma.18 Tree species representative of dry savanna woodland had become established there by 2 Ma. Vegetation to the east of Lake Turkana retained a local predominance of riparian woodland species ~1.5 Ma interspersed mosaically with more grassy formations.19 Underground trees or geoxyles, adapted to withstand frequent fires, became prominent constituents of the vegetation during the course of the early Pleistocene.20
Southern Africa
During the late Miocene and Pliocene, pollen from marine deposits off the coasts of Angola and Namibia show that both broad-leaved and fine-leaved savanna trees were present in the adjacent interior,21 but the white rhinos found in the early Pliocene fauna at Langebaanweg in the south-western Cape around 5 Ma still grazed only C3 grasses.22 The savanna bushveld around Makapansgat in South Africa’s northern Limpopo Province also contained only C3 plants at that time.23 The first indication of C4 grasses at Makapansgat comes from stable isotopes in small mammal fossils dated around 3.4 Ma, although ostrich eggshells indicate that C4 grasses had been present in Namibia a little earlier.24 C4 grasses expanded greatly in their representation at Makapansgat only after 2 Ma, somewhat later than in eastern Africa.25 In the vicinity of the dolomite caves in the South African Highveld, an open savanna of protea bushes along with riverine forest patches existed in the late Pliocene/early Pleistocene, similar to the modern-day grassland–forest mosaic prevailing in the Drakensberg foothills (Figure 9.2).21,26 Montane grasslands with protea shrubs and heathlands on rocky slopes are distributed further north on high-lying remnants of the African Erosion Surface into eastern Africa.27
Figure 9.2
Protea shrub savanna, a relict highland grassland formation resistant to fire. (A) Drakensberg foothills, South Africa; (B) Nyika Plateau, Malawi, post-fire.
Pleistocene Oscillations
The onset of glacial advances and retreats initiating the Pleistocene after 2.6 Ma was associated with oscillations between extremes of aridity and wetness over much of Africa. Initially, these were driven by the 41 kyr periodicity in the Earth’s angle of rotation, but the 21 kyr period generated by precession was particularly influential for rainfall in the tropics.
Eastern Africa
Over eastern Africa, cooler glacial periods were associated with expansions of grassy savanna, while tree cover increased during warmer interludes.28 In the Serengeti region, rainfall projected from plant wax biomarkers apparently fluctuated between extremes of 400–900 mm, and thus from semi-arid to mesic, during the 200 kyr interval between 1.9 and 1.7 Ma when climatic variation became accentuated.29 Despite rain-shadow effects intensified by the rise of Ngorongoro and other large volcanoes in the east, a lake became established at Olduvai due to blocked drainage.12 C4 grasses or sedges were prominent in the vegetation there in lowlands, while Afromontane forest was present on nearby highlands. The mixture of dry savanna, bushland and wooded grassland around Olduvai after 1.7 Ma closely resembled the heterogeneous mosaic shown by the modern dry savanna.12,30 Lake Olduvai dried up following rift movements around 1.3 Ma but became re-established intermittently later.
Further north in Kenya at Turkana West, the C4 grass contribution apparently fluctuated between extremes as wide as 5 percent to 100 percent, as inferred from wax biomarkers in soils, between 2.3 and 1.7 Ma.31 In East Turkana the woody vegetation around 1.95 Ma, as revealed by silicified wood, comprised species typical of riverine woodland.32 Thereafter, grassy savanna expanded generally.9,15 Conditions became temporarily moister in eastern Africa between 1.1 and 0.9 Ma, but the consequences for vegetation changes have not been elucidated. Savanna grasslands expanded at the expense of rainforests through western and central Africa, even into the Congo basin, during and following the coldest period of the last glaciation around 20 ka.33
Southern and South-Central Africa
Sediments deposited in the Indian Ocean by the Zambezi, Limpopo and other rivers draining southern parts of Africa record the vegetation in their catchments during the Middle and Late Pleistocene, based on fossil pollen and plant leaf waxes.34,35,36 In the Limpopo catchment, montane forest including yellowwood trees gave way to open heathland resembling Cape fynbos during extreme glacial conditions. Savanna woodland expanded during the brief interglacial interludes. Pollen from C4 grasses and sedges was more constantly present. Sediments deposited by the Zambezi River show a closely similar pattern, including the presence of ericaceous heathland as well as montane forest during glacial periods, but with greater representation of mopane during transitional periods. Miombo woodland does not feature, perhaps simply because its trees are poor pollen producers (this applies also to acacias). On the western side of the subcontinent, sediments off the Angolan coast indicate the presence of open grassy or shrubby vegetation during glacial periods and an expansion of miombo woodland during interglacials.21
Sediments contained within Lake Malawi reflect abrupt vegetation changes over quite short periods after 135 ka, when low lake levels reflected mega-drought conditions.37 Yellowwood pollen gave way to grass pollen after 105 ka, while pollen from miombo woodland increased in prominence between 75 and 65 ka. This was followed by an expansion by yellowwood forest after 60 ka when global temperatures were descending towards the LGM. Following the LGM, miombo woodland increased at the expense of montane forest.38
In the Cradle region of South Africa’s Highveld, shrubby grassland with riverine woodland flanking rivers persisted through the early Pleistocene from 1.8 to 1.2 Ma.39,40 In the vicinity of Wonderwerk Cave located in the Northern Cape, close to the current transitions between the arid savanna, grassland and Nama-Karoo biomes, a mixture of C4 and C3 grasses persisted from 1.95 until 0.99 Ma, as shown by pollen trapped in stalagmites and hyrax dung deposits.41,42 Thereafter, C4 grasses increased in prominence as aridity intensified.
Within interior South Africa, the vegetation near Florisbad after 300 ka, shortly preceding the earliest modern humans, fluctuated between being grassy with karoo shrubs during warm periods and fynbos shrubbery when cooler conditions ensued.43 Two sites located in the northern bushveld region of South Africa, Wonderkrater spring and Tswaing impact crater, provide an exceptionally detailed record of vegetation shifts in their vicinity after 200 ka, derived from fossil pollen.44,45,46 At Tswaing, the local vegetation shifted from dry savanna woodland with bushwillow prominent to yellowwood forest or fynbos as glacial conditions set in after 190 Ma. Vegetation near Wonderkrater was evidently a grassy savanna prior to 45 ka, approaching the LGM, with both acacias and bushwillows prominent along with camphor bush (Tarchonanthus camphoratus).47 By 35 ka, with conditions becoming cooler but wetter, C3 grasslands and ericaceous shrubs took over and C4 grasses were absent. Pollen from yellowwood (Podocarpus spp.), wild olive (Olea africana) and boer-bean (Schotia brachypetala) trees indicated the nearby presence of evergreen forest. There was a hiatus in pollen through the LGM around 20 ka when temperatures were perhaps 5°C cooler than at present and conditions became very dry. After 14 ka, savanna woodland with acacia trees became established once more as conditions warmed into the current interglacial period. On the high Lesotho plateau, low shrubland including plant genera currently found much further south prevailed during and following the LGM 23–17 ka, indicating a northward shift by winter rainfall conditions. In the environs of Sibudu Cave, located close to the KwaZulu-Natal coast, yellowwood forest was more prominent prior to 60 ka than it is today, indicating cooler and wetter conditions.
Overview
The record of past vegetation in Africa, records three noteworthy features: (1) a progressive trend towards greater grass cover through the Pliocene and Pleistocene; (2) wide oscillations in the local presence of different vegetation formations, especially during the late Pleistocene; and (3) greater representation of grassland–forest interspersion along with ericaceous shrubs during the cooler glacial periods that predominated during the greater portion of the Pleistocene.
Savanna formations first appeared within Africa during the Miocene, before C4 grasses became overwhelmingly dominant. Spinescent acacias, characteristic of dry/eutrophic savannas, were prominent in eastern Africa by the mid-Pliocene. Nevertheless, riparian woodlands prevailed in the Omo and Turkana basins throughout the early Pleistocene, adjoining the lakes that formed and rivers that flowed. C4 grasses became dominant somewhat later in subtropical southern Africa than in tropical eastern regions. During the Middle and Late Pleistocene, ericaceous heathland (fynbos) became prominent in fossil pollen as far north as the Zambezi River catchment during glacial extremes, along with widened representation of evergreen forests. Nevertheless, grass pollen remained abundant, while pollen from miombo woodland trees was surprisingly uncommon. Montane grassland–forest mosaics were more widely prevalent during the cooler glacial conditions that predominated during the Pleistocene than they are at present over southern and south-central Africa, replacing broad-leaved savanna woodlands. Fine-leaved, acacia-dominated savanna had become established in parts of eastern Africa by this time and was well represented in the south-west within present-day Namibia.
Wide fluctuations in the local vegetation within quite short periods are strikingly evident in the finer temporal resolution provided by sedimentary accumulations in South Africa and Malawi through the late Pleistocene. They are probably responses to the 21 kyr oscillation in the Earth’s rotational wobble. Whatever the composition, grassy vegetation remained predominant throughout eastern and southern Africa from the late Pliocene onward. An additional factor beside climate contributing to a more open tree cover through most of the Pleistocene was the prevalently low atmospheric concentration of carbon dioxide.48 This restricts the growth rate of woody plants and hence their chances of growing beyond the fire trap.
SUGGESTED FURTHER READING
Bamford, MK, et al. (2016) Pollen, charcoal and plant macrofossil evidence of Neogene and Quaternary environments in southern Africa. In Knight, J; Grab, SW (eds) Quaternary Environmental Change in Southern Africa. Cambridge University Press, Cambridge, pp. 306–323.
Barboni, D. (2014) Vegetation of northern Tanzania during the Plio-Pleistocene: a synthesis of paleobotanical evidences from Laetoli, Olduvai and Peninj hominin sites. Quaternary International 322–323:264–276.
Bonneville, R. (2010) Cenozoic vegetation, climate changes and hominid evolution in tropical Africa. Global Planetary Change 72:390–412.
Jacobs, BF, et al. (2010) A review of the Cenozoic vegetation history of Africa. In Werderlin, L; Sanders, WJ (eds) Cenozoic Mammals of Africa. University of California Press, Berkeley, pp. 57–72.
REFERENCES
1.Jacobs, BF, et al. (1999) The origin of grass-dominated ecosystems. Annals of the Missouri Botanical Garden 86:590–643.
2.Strömberg, CAE. (2011) Evolution of grasses and grassland ecosystems. Annual Review of Earth and Planetary Sciences 39:517–544.
3.Uno, KT, et al. (2016) Neogene biomarker record of vegetation change in eastern Africa. Proceedings of the National Academy of Sciences of the United States of America 113:6355–6363.
4.Polissar, PJ, et al. (2019) Synchronous rise of African C4 ecosystems 10 million years ago in the absence of aridification. Nature Geoscience 12:657–660.
5.Cerling, TE, et al. (1997) Global vegetation change through the Miocene/Pliocene boundary. Nature 389:153–158.
6.Feakins, SJ, et al. (2013) Northeast African vegetation change over 12 my. Geology 41:295–298.
7.Jacobs, BF. (2004) Palaeobotanical studies from tropical Africa: relevance to the evolution of forest, woodland and savannah biomes. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 359:1573–1583.
8.Jacobs, BF, et al. (2010) A review of the Cenozoic vegetation history of Africa. In Werdelin, L; Sanders, WJ (eds) Cenozoic Mammals of Africa. University of California Press, Berkeley, pp. 57–72.
9.Bonnefille, R. (2010) Cenozoic vegetation, climate changes and hominid evolution in tropical Africa. Global and Planetary Change 72:390–411.
10.Bobe, R. (2006) The evolution of arid ecosystems in eastern Africa. Journal of Arid Environments 66:564–584.
11.Edwards, EJ, et al. (2010) The origins of C4 grasslands: integrating evolutionary and ecosystem science. Science 328:587–591.
12.Barboni, D. (2014) Vegetation of Northern Tanzania during the Plio–Pleistocene: a synthesis of the paleobotanical evidences from Laetoli, Olduvai, and Peninj hominin sites. Quaternary International 322:264–276.
13.Keeley, JE; Rundel, PW. (2005) Fire and the Miocene expansion of C4 grasslands. Ecology Letters 8:683–690.
14.Charles-Dominique, T, et al. (2016) Spiny plants, mammal browsers, and the origin of African savannas. Proceedings of the National Academy of Sciences of the United States of America 113:E5572–E5579.
15.Cerling, TE, et al. (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51–56.
16.Bamford, MK. (2011) Fossil woods. In Harrison, T (ed.) Paleontology and Geology of Laetoli: Human Evolution in Context. Springer, Dordrecht, pp. 217–233.
17.Andrews, P; Bamford, M. (2008) Past and present vegetation ecology of Laetoli, Tanzania. Journal of Human Evolution 54:78–98.
18.Bobe, R; Eck, GG. (2001) Responses of African bovids to Pliocene climatic change. Paleobiology 27:1–47.
19.Bamford, MK. (2017) Pleistocene fossil woods from the Okote Member, site FwJj 14 in the Ileret region, Koobi Fora Formation, northern Kenya. Journal of Human Evolution 112:134–147.
20.Maurin, O, et al. (2014) Savanna fire and the origins of the ‘underground forests’ of Africa. New Phytologist 204:201–214.
21.Neumann, FH; Bamford, MK. (2015) Shaping of modern southern African biomes: Neogene vegetation and climate changes. Transactions of the Royal Society of South Africa 70:195–212.
22.Franz-Odendaal, TA, et al. (2002) New evidence for the lack of C4 grassland expansions during the early Pliocene at Langebaanweg, South Africa. Paleobiology 28:378–388.
23.Hopley, PJ, et al. (2007) Orbital forcing and the spread of C4 grasses in the late Neogene: stable isotope evidence from South African speleothems. Journal of Human Evolution 53:620–634.
24.Ségalen, L, et al. (2007) Timing of C4 grass expansion across sub-Saharan Africa. Journal of Human Evolution 53:549–559.
25.Hopley, PJ, et al. (2007) High- and low-latitude orbital forcing of early hominin habitats in South Africa. Earth and Planetary Science Letters 256:419–432.
26.Luyt, CJ; Lee-Thorp, JA. (2003) Carbon isotope ratios of Sterkfontein fossils indicate a marked shift to open environments c. 1.7 Myr ago. South African Journal of Science 99:271–273.
27.Meadows, ME; Linder, HP. (1993) Special paper: a palaeoecological perspective on the origin of afromontane grasslands. Journal of Biogeography 20:345–355.
28.Dupont, LM, et al. (2011) Glacial–interglacial vegetation dynamics in South Eastern Africa coupled to sea surface temperature variations in the Western Indian Ocean. Climate of the Past 7:1209.
29.Magill, CR, et al. (2013) Ecosystem variability and early human habitats in eastern Africa. Proceedings of the National Academy of Sciences of the United States of America 110:1167–1174.
30.Albert, RM, et al. (2015) Vegetation landscape at DK locality, Olduvai Gorge, Tanzania. Palaeogeography, Palaeoclimatology, Palaeoecology 426:34–45.
31.Uno, KT, et al. (2016) A Pleistocene palaeovegetation record from plant wax biomarkers from the Nachukui Formation, West Turkana, Kenya. Philosophical Transactions of the Royal Society B: Biological Sciences 371:20150235.
32.Bamford, MK. (2011) Late Pliocene woody vegetation of Area 41, Koobi Fora, East Turkana Basin, Kenya. Review of Palaeobotany and Palynology 164:191–210.
33.Malhi, Y, et al. (2013) African rainforests: past, present and future. Philosophical Transactions of the Royal Society B: Biological Sciences 368:20120312.
34.Dupont, L. (2011) Orbital scale vegetation change in Africa. Quaternary Science Reviews 30:3589–3602.
35.Castañeda, IS, et al. (2016) Middle to Late Pleistocene vegetation and climate change in subtropical southern East Africa. Earth and Planetary Science Letters 450:306–316.
36.Dupont, LM; Kuhlmann, H. (2017) Glacial–interglacial vegetation change in the Zambezi catchment. Quaternary Science Reviews 155:127–135.
37.Beuning, KRM, et al. (2011) Vegetation response to glacial–interglacial climate variability near Lake Malawi in the southern African tropics. Palaeogeography, Palaeoclimatology, Palaeoecology 303:81–92.
38.Ivory, SJ, et al. (2012) Effect of aridity and rainfall seasonality on vegetation in the southern tropics of East Africa during the Pleistocene/Holocene transition. Quaternary Research 77:77–86.
39.Reed, KE. (1997) Early hominid evolution and ecological change through the African Plio–Pleistocene. Journal of Human Evolution 32:289–322.
40.Bamford, MK, et al. (2016) Pollen, charcoal and plant macrofossil evidence of Neogene and Quaternary environments in southern Africa. In Knight, J; Grab, SW (eds) Quaternary Environmental Change in Southern Africa. Cambridge University Press, Cambridge, pp. 306–323.
41.Ecker, M, et al. (2018) The palaeoecological context of the Oldowan–Acheulean in southern Africa. Nature Ecology & Evolution 2:1080–1086.
42.Brook, GA, et al. (2010) A 35 ka pollen and isotope record of environmental change along the southern margin of the Kalahari from a stalagmite and animal dung deposits in Wonderwerk Cave, South Africa. Journal of Arid Environments 74:870–884.
43.Scott, L; Neumann, FH. (2018) Pollen-interpreted palaeoenvironments associated with the Middle and Late Pleistocene peopling of Southern Africa. Quaternary International 495:169–184.
44.Scott, L. (1999) Vegetation history and climate in the Savanna biome South Africa since 190,000 ka: a comparison of pollen data from the Tswaing Crater (the Pretoria Saltpan) and Wonderkrater. Quaternary International 57:215–223.
45.Scott, L, et al. (2003) Age interpretation of the Wonderkrater spring sediments and vegetation change in the Savanna Biome, Limpopo province, South Africa. South African Journal of Science 99:484–488.
46.Scott, L. (2016) Fluctuations of vegetation and climate over the last 75 000 years in the Savanna Biome, South Africa: Tswaing Crater and Wonderkrater pollen sequences reviewed. Quaternary Science Reviews 145:117–133.
47.Backwell, LR, et al. (2014) Multiproxy record of late Quaternary climate change and Middle Stone Age human occupation at Wonderkrater, South Africa. Quaternary Science Reviews 99:42–59.
48.Bond, WJ, et al. (2003) The importance of low atmospheric CO2 and fire in promoting the spread of grasslands and savannas. Global Change Biology 9:973–982.