East and Southern African Neolithic: Geography and Overview
- 122 Downloads
State of Knowledge and Current Debates
The Neolithic period of eastern and southern Africa was a form of niche fulfillment when viewed within the context of its origins and distribution. In Africa, Neolithic cultures include those groups who herd domesticated animals, use stone tools and ceramics as part of their subsistence package. Climatic and demographic pressures after 6000 BP pushed Neolithic pastoralists living in the Nile River Valley southward, and the savanna habitats of eastern and southern Africa that these migrants encountered were attractive ecosystems rife for herding. The early Neolithic pastoralists in eastern Africa tended to be less mobile and had a broad-based subsistence compared to later pastoral groups, who colonized southern Africa after 2000 BP. When placed within the larger paleoenvironmental and cultural context, pastoralists adapted their subsistence to specific ecosystems based on the potential of the land to provide adequate sustenance for themselves and their livestock, but there are constant demographic pressures facing pastoralists who have low productive yields per unit of land compared to sedentary farmers.
In this entry, the Neolithic period in eastern and southern Africa will be reviewed from the perspective of how animal herders occupy distinct niches on a landscape. The current state of archaeological evidence for early pastoral cultures in the subcontinent will be examined to evaluate how prehistoric people may have opportunistically exploited specific ecological niches in the landscape. The present-day mosaic of genetic and linguistic groups across the subcontinent is used to fill in the gaps of knowledge where archaeological data is insufficient to chronicle the demographic and subsistence shifts throughout the late Holocene. The arrival of Iron Age agropastoralists in southern Africa will also be explored in order to attempt a clarification what the Neolithic entailed in this specific context. African Neolithic contexts are not a uniform culture or subsistence adaptation, and people have remained flexible in the ways they procure food, water, and shelter from the land into the modern era.
The Neolithic in Eastern and Southern Africa
The roots of animal herding cultures of eastern and southern Africa are found in the now desiccated landscape of the eastern Sahara. Genetic data supports the notion that a rapid demographic expansion of the genetic forebears of the Niger-Kordofanian (Bantu), Nilo-Saharan (Nilotic), and Afroasiatic (Cushitic) language families (broadly falling in the L2 mtDNA haplotype) occurred at the end of the Younger Dryas between 10,000 and 12,000 years BP (Trombetta et al. 2015). It was during this period that experimentations with domesticating cattle (Bos taurus) are thought to have occurred in the eastern Sahara (Jórdeczka et al. 2013), although this view is not universally accepted (Linseele et al. 2014; Stock and Gifford-Gonzalez 2013).
At 10,000 years BP, the Sahara was a very different ecosystem than it was during the previous stadial period or what it is like today. The zenith of the so-called African Humid Period (AHP) saw the Sahara transform into a vast network of lakes and swamps that hosted hippos, crocodiles, a wide array of bovids and communities of hunter-gatherers who lived adjacent to the water bodies and had broad-based subsistence patterns. Pan-African distributions of “wavy line” and “dotted wavy line” ceramics in conjunction with bone harpoons and evidence for high predation on fish and riparian resources is evidence for a transcontinental movement of people during the early Holocene, occurring in tandem with population growth (Manning and Timpson 2014).
As climate conditions began to dry out, productive ecological biomes contracted around permanent sources of water. At archaeological sites adjacent to Nabta Playa in the southwestern Egyptian Sahara and further north in the Fayum region, there is evidence that animals and people were tethered to the same resources. It is during this period that the earliest domesticates appear to have been introduced into northern African from southwestern Asia (Barich 2016). Leaving contentious claims of cattle in Nabta Playa, the earliest, unambiguous domesticated animals date to “ca. 5600 cal BC” (7550 cal. years BP, if taken sensu stricto) in the form of goats (Capra hircus) from the site of Fayum (Linseele et al. 2016) and within 200 years at Kerma in Upper Nubia (Honegger and Williams 2015). Sheep (Ovis aries) were present in the Nile Delta, eastern Sahara, and Red Sea region by 7000 years BP (Muigai and Hanotte 2013). Goat and/or sheep remains have been recovered in contexts dating to ca. 7250 years BP in Sodmein Cave in the Red Sea Hills (Vermeersch et al. 2015). Current scholarship indicates that donkeys (Equus asinus) were domesticated in the Nile Valley between 7000 and 6000 years BP (Kimura et al. 2013). However, domesticates remain scarce in the archaeozoological record relative to wild animals (and fish) for the next 1500 years. During this period, the relationship between humans and these animals transformed into one of co-dependency. As lakes were drying up, humans needed a dependable food source and cattle had been bred to the point where their natural foraging instincts had disappeared.
Early domesticates were probably part of an integrated exchange system spanning the Sahara and Nile River Basin into southwestern Asia in which complex foragers/pastoralists maintained close technological and cultural networks over vast territories. These networks involved seasonal mobility in which riparian areas were utilized extensively during the dry season and grasslands were exploited in the rainy season. As riparian woodlands turned into grasslands by 6000 years BP and grasslands turned into scrubland and sand by 4000 years BP, settlement in the eastern Sahara was abandoned and the extensive exchange network collapsed (Hoelzmann et al. 2001).
Prior to the introduction of domesticated animals, the primary (archaeologically visible) subsistence strategy in the eastern Sahara and Nile Basin was fishing using barbed bone points (a.k.a. “bone harpoons”). Archaeological sites in the Nile Basin have increasing numbers of domesticated animals represented in their faunal assemblages between 6000 and 4000 years BP indicating the importance livestock was gaining as a resource that was depended upon (Sadig 2010). Settlements with evidence for uses of aquatic, wild terrestrial, and domesticated resources also appear along the Gash River at the western foot of the Ethiopian Highlands around 5000 years BP (Gautier and Van Neer 2006), trickling into the Ethiopian Highlands after 3000 years BP (Lesur et al. 2014). Cultivation of sorghum along the banks of the Nile River also intensified by 5000 years BP (Haaland 1995) and may have further enticed people to accelerate settlement within riparian regions. Cultivated food was only one component of a very broad subsistence strategy practiced in the riverine regions of northeastern Africa, and long-distance exchange networks were maintained with foragers living in the Ethiopian Highlands (Fernández et al. 2007) and Turkana Basin (Wright et al. 2015).
East Africa and the Horn
Shortly after 5000 years BP, Neolithic pastoralists reached the Turkana Basin (Grillo and Hildebrand 2013). In the Koobi Fora region on the northeastern shore of Lake Turkana, harpoon fishing and riparian foraging was replaced with livestock husbandry and intensive riparian foraging. Stylistic commonalities between Nderit tradition (Turkana) and Shaheinab tradition (Sudanese Nile) ceramics (Barthelme 1985), similar subsistence patterns (Marshall et al. 1984), and linguistic reconstructions (Ehret 1998) have been used to argue that there was a middle Holocene migration of Sudanese (Afroasiatic/Cushitic) pastoralists into the Lake Turkana region. This marks the beginning of the “Pastoral Neolithic” (PN) period in eastern Africa, which is defined as animal herding cultures who used ceramics and Later Stone Age (LSA) tools (Bower et al. 1977).
However, the arrows of pastoral diffusion are difficult to untangle during the middle Holocene. Teeth of domesticated cattle have been unearthed from Quiha in northeastern Ethiopia with a stratigraphic date of ~5000–6000 years BP (Barnett 1999). Domesticated animals are found at other sites such as Asa Koma in Djibouti (Lesur 2004), Danei Kawlos in northeastern Ethiopia (Marshall and Negash 2002), Gobedra Rock Shelter in northern Ethiopia (Phillipson 1977), and Laga Oda (Clark and Prince 1978) and Lake Besaka (Brandt 1984) in the northeastern Rift Valley. At the Gogoshiis Qabe Rockshelter in southern Somalia, domesticated goats and cattle were recovered from pre-3500 year BP contexts (Brandt 1986). There is virtually no evidence to suggest a stylistic replacement of indigenous ceramic or lithic traditions in the Horn of Africa such as what is presumed to have occurred in the Turkana Basin (Lesur 2004).
For approximately 1000 years, occurrences of livestock in eastern Africa remain restricted primarily to the Turkana Basin and the Horn. Analyses of obsidians from the Turkana Basin suggest that once pastoralism was established in the region, exchange networks were intense but localized (Nash et al. 2011). There were only small “trickles” of pastoralists south of 3°N prior to 3000 years BP (Bower 1991). There are four published archaeological site locations with domesticated animals south of Lake Turkana prior to 3000 years BP: Enkapune ya Muto in the central Rift Valley (Ambrose 1998), Gogo Falls (Karega-Mũnene 2002) and Usenge 3 (Lane et al. 2007) near Lake Victoria, and Kahinju in the Kenyan Coastal Plains (Wright 2007). All of these sites have evidence for minimal reliance on domesticated animals, but instead show foraging-based adaptations in which domesticated animals became one (relatively minor) subsistence component.
It is likely that the recorded drops in lake levels and increases in xeric plant species documented across equatorial Africa between 4000 and 3000 years BP impeded the spread of transhumant pastoralists beyond predictable resource bases. The environments in which domesticated animals are found during this time period are areas with high topographic relief or are close to permanent water bodies. Furthermore, domesticated animals are found in very low numbers relative to wild taxa during this period from sites outside the Turkana Basin and Horn. It is a distinct possibility that livestock in these settings were rustled from established herds elsewhere and not kept for extensive periods prior to slaughtering (Marshall 2000). In that case, the presence of domesticated animals in equatorial Africa prior to 3000 years BP represented an additional resource within the broad-based foraging complex of those who possessed them and was neither culturally transformative nor even necessarily permanent.
Eburran 5 sites are those where there is evidence for a limited acquisition of domesticated stock, used ceramics, but maintained the same general stone tool traditions (long, narrow blades and microliths, narrow endscrapers, sidescrapers, and microburins) and a broad-spectrum foraging patterns from preceding periods (Wilshaw 2016). These sites are located on savanna and forest ecotones and are mostly known from the Central Rift Valley.
Most Kansyore sites are restricted to the areas adjacent to fluvial systems and Lake Victoria and are delayed-return hunter-fishers-gatherers who adopted domesticated animals by as early as 4500 years BP (Prendergast 2010b). The Kansyore ceramic tradition dates to as early as 6000 years BP and includes hemispherical bowls that are highly decorated with circular bands of rocker stamp and punctate impressions (Dale and Ashley 2010). The lithic traditions are not uniform, which reflects the LSA traditions of each locale rather than being a widely distributed cultural entity (Seitsonen 2010). Domesticated animals were incorporated into an indigenous foraging economy that involved seasonal residential mobility to exploit aquatic resources and access wild terrestrial game available at different times of the year.
Elmenteitan sites were first described by Louis Leakey (1931) and are found in more closed habitat locations along the western Rift Valley above 1900 m. These sites have a higher percentage of ovicaprids compared to cattle and have a distinct lithic industry comprised of large, backed blades, long endscrapers, and large sidescrapers (Ambrose 1984). Elementeitan sites are also characterized as having a distinct pottery style (globular vessels with vertical lugged handles) and using stone bowls as mortuary artifacts.
Savanna Pastoral Neolithic (SPN) sites are distributed throughout eastern Africa and were predominantly cattle pastoralists who supplemented their diets with plains game or, in the case of Lake Turkana, fish (Ambrose 1984). This archaeological entity has also been called “Wilton” (Leakey 1931) referring to the short, broad blades and convex endscrapers that characterize the lithic industry. However, there is considerable diversity in the lithic technologies used, which have stylistic continuity from LSA traditions predating pastoralism. Four named ceramic traditions called Nderit, Narosura, Akira, and Maringishu have been described from SPN sites as well as a distinct herringbone-motif decorated vessel. Stone bowls and other ground stones are commonly found on these sites.
The period from 3000 to 2000 years BP is known as the so-called splash of pastoralism into equatorial eastern Africa (Bower 1991). This has also been conceptualized as the period when pastoralism incubated in the whole of eastern Africa, as opposed to the preceding frontier phase when pioneers ventured into uncharted territories to search for new pastures or escape demographic pressures in their ancestral lands (Lane 2004). During “consolidation phases” of human frontier expansion, intensification and specialization of agricultural production techniques occurs (Alexander 1980; Lane 2004). Indeed, many archaeological sites with high proportions (>90%) of domesticated animals in faunal assemblages are found in the central Rift Valley and portions of the savannas of Kenya and northern Tanzania from this time period. However, many contemporaneous archaeological sites across eastern Africa have evidence that people were nonspecialists, maintaining small herds of animals that supplemented their diets which primarily revolved around foraging of which Eburran and Kansyore forager-(fisher)-pastoralists are good examples. In the case of the inhabitants along the Galana River in the Kenyan Coastal Plains, there is little evidence of significant changes in settlement or technology from 6000 to 1300 years BP (Wright 2007).
The precipitation regime of eastern Africa is bimodal with rainy seasons dictated by the movement of the Intertropical Convergence Zone (ITCZ). However, there is significant spatial and temporal variability in the distribution of rainfall based on global and regional differences in the distribution of solar heat and sea surface temperatures. Generally, eastern equatorial Africa has been wetter in the last 4000 years than during the middle Holocene (Verschuren et al. 2009), while the western portion of the region was driest around 2000 years ago with gradually increasing precipitation thereafter (Nash et al. 2016). In general, rainfall patterns in the late Holocene have been unpredictable on the decadal scale relative to the earlier Holocene (Plisnier et al. 2000), which appears to have pushed people toward exploiting a more predictable resource base. The opening up of a grassy corridor by 2000 years BP in the Lake Victoria region (Chritz et al. 2015) and in south-central Africa (Castañeda et al. 2009; Robinson and Rowan 2017) enabled pastoralists to move into the region without fear of encountering swarms of tsetse flies (Glossina sp.), which carry the disease trypanosomiasis (Gifford-Gonzalez 2017).
The use of Y-chromosome and mtDNA genetic markers is used in southern Africa to disentangle the history of population movements over the last several 1000 years (see discussion below). However, using these tools is a much more difficult task in eastern Africa where genetic admixture has been more thorough despite the retention of four language families and over 200 self-identified ethnic groups. One example of this is the Luo ethnic group from western Kenya who speak a Nilo-Saharan language but have shown predominantly Kordofanian/Bantu ancestry (Scheinfeldt et al. 2010). A classic ethnographic analysis tracing the movements of material culture items and marriage partners across ethnic boundaries in the Lake Baringo region of Kenya illustrates that ethnically diverse landscapes necessitate the presence of strong cooperative networks in order to function well (Hodder 1977). There is also evidence of admixture of genes from southwest Asian populations introduced into East Africa between 3300 and 2700 years ago (Pickrell et al. 2014). Cohesive social networks built on complex social values, kinship, and exchange patterns are crucial components of niche fulfillment in precarious ecological settings and have deep roots in African prehistory.
Demic Routes of the Neolithic to Southern Africa
The most commonly accepted demic route of the first livestock into southern Africa holds that a small group of migrants left the Great Lakes region of Uganda around 2000 years BP and settled in Okavango River Basin of what is now northern Botswana and the Caprivi Strip of Namibia. From there, the agro-pastoralists came into contact with Khoe-Kwadi speaking hunter-gatherers, some of whom rapidly adopted a pastoral lifestyle themselves, and went on to colonize vast tracts of southern Africa by 1800 years BP. (The linguistic distinction of the Khoe-Kwadi language family(-ies) is the subject of some debate among linguists (Ehret 2008; Güldemann 2008; Mitchell 2010). One common heuristic device, though not linguistically or biologically accurate, is to refer to pastoral communities of southern Africa as “Khoe” and the hunter-gatherers as “San” (Scheinfeldt et al. 2010; Soodyall et al. 2008).)
Separate migrations of Bantu-speaking agropastoralists occurred down the Atlantic and Indian Ocean coasts between 1800 and 1600 years BP (de Filippo et al. 2012), but there are many potential memes that remain undefined in this explanation of livestock diffusion. It is clear that the spread of domesticated animals did not occur as a coherent population expansion of one culture group from eastern to southern Africa. There has been considerable debate regarding the degree to which livestock herding techniques were adopted by indigenous foragers or whether pastoral people who entered southern Africa settled alongside foragers and there was a gradual melding of subsistence and culture systems via intermarriage and intercommunity trade. The picture is further complicated by the fact that Bantu-speaking agropastoralists enter southern Africa within 200 years of the non-Bantu-speaking populations (“Khoekhoen”). In the end, there are three primary lines of evidence that can be used to retrace the spread of pastoralism out of eastern Africa: archaeology, genetics, and linguistics. In this entry, these data sets (that are not always in total agreement with one another) will be reviewed to evaluate the current state of knowledge surrounding the origins of Neolithic pastoral economies in southern Africa.
The lack of archaeological evidence for a southward expansion of Neolithic pastoralists from 3000 to 2000 years BP into in the areas that now include southern Tanzania, northern Malawi, northern Mozambique, and eastern Zambia may be as a result of the dearth of well-dated archaeological assemblages from this time period. The majority of archaeological finds in the Rufiji River Delta and surrounding areas (Chami and Kweksason 2003) most likely date to after 3000 years BP rather than before, as the authors suggest. The artifacts, including numerous Narosura ceramic tradition sherds, broad blades, and groundstones, are all suggestive of a well-developed SPN demic expansion from 3000 to 2000 years BP. Ancient DNA analyses from Figira and Hora rockshelters in Malawi show complete genetic replacement of indigenous populations by modern-day Bantu-speaking populations, while genetic succession was less complete among populations in the modern day countries of Kenya and Tanzania (Skoglund et al. 2017).
Early Iron Age settlement of the region adjacent to the Zambezi and Machili Rivers in southwestern Zambia has been reported during the time period when Neolithic pastoralists had likely reached points south of the Zambezi. The sites of Kunyengenya, Salumano, Lusu, and Situmpa have Iron Age ceramic assemblages, evidence for domesticated sheep and radiocarbon ages clustering around 2200 years BP (Katanekwa 1978; Phillipson 1989). The so-called Situmpa Ware is documented across Zambia and Zimbabwe and has radiocarbon ages dating to around 2000 years BP (Clark and Fagan 1965). In the Namib Desert, the site of Snake Rock has ceramics and sheep bones dating to 2000 years BP (Kinahan 2016). Despite intensive investigations, evidence of Neolithic pastoralists have yet to be located in this region from this time period, so other forms of data must be explored to identify the routes by which the Neolithic moved from eastern into southern Africa.
Y-chromosomal genetic evidence shows that pastoralists from eastern Africa en route to southern Africa passed through Tanzania estimated at ca. 2000 years BP (Henn et al. 2008) and shortly thereafter made their way down a corridor along the Zambezi River (Russell et al. 2014). Direct dating of archaeological materials from this region is scant and DNA evidence has a large range of associated error in its clock, so there is much room for speculation regarding who the first pastoralists were, the routes they took to pass into southern Africa, and when it occurred.
Linguistic analyses provide an interpretive framework for understanding the migration of pastoralists and farmers into southern Africa. Early Neolithic groups from southern Africa spoke Khoe-Kwadi languages, which originated from Okavango River Basin. However, Khoe-Kwadi languages also have eastern African links to root words like “year,” which have similar cognates in Cushitic and eastern Bantu languages (Güldemann 2008). More tellingly, the overlap of cognates between East African languages associated with Sandawe hunter-gatherers (northern Khoe family) and Khoe-Kwadi are numerous (Ehret 2008; Güldemann 2008). The genetic divergence dates of click-speaking populations in Tanzania (e.g., Sandawe, Hadzabe) from southern Khoe-San populations are estimated to be older than 35,000 years (Tishkoff et al. 2007), so one should expect that proto-Khoe-Kwadi speaking migrants to southern Africa were conversant with animal husbandry as well as foraging subsistence prior to their exodus. When the immigrant northern Khoe speakers reached southern Africa, both genetic and linguistic analyses indicate that the proto-Khoe-Kwadi language family diversified enormously as intercommunity exchange networks developed (Güldemann 2008; Salas et al. 2002; Soodyall et al. 2008). As will be demonstrated in the following section, the archaeological evidence reflects the diversity of subsistence adaptations that accompanied the movement of people and technological ideas after 2000 years BP.
Southern African Neolithic?
Insofar as it is understood, the southern African Neolithic consists of two facies: herders who hunted and gathered and hunter-gatherers who kept small amounts of livestock (Sadr 2003). As during the pioneer phases of agricultural expansion in many other places in Africa, pastoralism entered South Africa after 2000 years BP as an experimental concept that indigenous hunter-gatherers integrated into their subsistence economies. In the years that succeeded the introduction of Neolithic and Iron Age lifeways into southern Africa, the ecology of the landscape was impacted by the demographics of settlement and nonanthropogenic changes to the climate, which pushed people into narrower subsistence niches.
The earliest securely dated archaeological finds of domesticated animals in southern Africa are comprised of sheep remains at Spoegrivier Cave in Namaqualand dated to 2105 ± 65 14C years BP (1990–2300 cal. years BP; Sealy and Yates 1994). In archaeological sites studied across Namaqualand, there is no discernable material culture distinction between archaeological assemblages prior to and immediately following 2000 years BP, suggesting that pastoralism in this region was not introduced by a separate, immigrant pastoral group (Webley 2007). Following the introduction of ceramic technology into the region after 1900 years BP, incremental changes to lithic assemblages and ostrich eggshell bead production could represent the evolution of a “herder-forager” culture, distinct from those groups who did not adopt domesticated animals (Sadr 2003).
Sheep and cattle are found at Toteng in northern Botswana directly dated with radiocarbon ages of 2020 ± 40 and 2070 ± 40 14C years BP (1900–2040 cal. Years BP and 1990–2110 cal. Years BP, respectively; Robbins et al. 2005). In both cases, the vast majority of the faunal remains were from wild animals, with domesticated stock comprising <5% of the total diagnostic NISP assemblage (Robbins et al. 2008). Settlement of the Toteng sites correlates to higher lake levels of Lake Ngami and archaeological materials recovered reflect that subsistence had a distinctly riparian resource predation strategy (Robbins et al. 2008). Occupation of the site throughout the late Holocene conforms to the hunters-who-kept-livestock model advanced by Sadr (2003).
The first livestock economies to enter southern Africa are argued to have preferred sheep as opposed to cattle (Smith 1992). However, Thomas Huffman (2001, 2007) argues that archaeofaunal assemblages from the southern African Iron Age underrepresent the numbers of cattle people actually kept because African pastoralists usually use cattle byproducts (milk, blood) and rarely eat their cattle. He calls this the “Central Cattle Pattern,” which is documented in numerous ethnographic examples from across the continent. Subsequent analysis of Early Iron Age phytolith assemblages from dung deposits in southern Africa reflect taxa consumed by small stock and not cattle confirming the hypothesis that early Neolithic pastoralism was centered on tending small stock (Badenhorst 2009). Middle and Late Iron Age agropastoralists were more reliant on cattle and the faunal assemblages from across the region reflect this trend.
Using linguistics, the earliest cattle herding cultures of southern Africa are thought to have been Khoe-Kwadi speakers who settled in the Limpopo Valley approximately 2000 years ago (Ehret 2008; Güldemann 2008). Tom Güldemann (2008) sets the homelands for the Khoe-Kwadi language family in the northeastern Kalahari along what is today the northern borders of Namibia and Botswana. However, the archaeological association between Khoe-speakers and pastoralism between 2000 and 1000 years BP is shaky (Sadr 2008b). Khoekhoen people were known as highly mobile, lugged-ceramic using pastoralists living in the western Cape of South Africa (Sadr 2003). However, none of the antecedent material culture features associated with ethnographic accounts of Khoekhoen pastoralists (Bollong et al. 1997) are apparent from archaeological sites prior to 1000 years BP. When lugged ceramics first appear in the archaeological record at Kasteelberg in the southern Cape after 1000 years BP, residue analysis from the vessels indicates that they were used to process seal blubber, not milk as would be expected from people who were dependent on livestock (Patrick et al. 1985).
Neolithic pastoralists arrived in the western Cape approximately 2000 years BP and the eastern Cape and western coast regions of South Africa by 1800 years BP and shared the low-altitude plains with hunter-gatherers through the late Holocene (Lewis 2002; Sadr 2015). The site of Kasteelberg has been interpreted as supporting two types of populations: mobile, herder-foragers who subsisted on inland resources and another population of semi-sedentary herder-foragers who exploited coastal ecosystems (Sadr et al. 2003). Small-stock herding at the site is dated to prior to ca. 2000 years BP and shows continuity in the overall archaeological assemblage with previous nonlivestock tending occupations of the site (Sadr et al. 2003).
After the arrival of pastoralism in the Elands Bay region, hunter-gatherers appear to have continued to practice a highly mobile, flexible, opportunistic foraging strategy, which suggests noncompetitive interactions between groups with different subsistence practices (Jerardino et al. 2009). Stable isotope analysis performed on human skeletal materials from the southwestern Cape indicates that the diets of coastal inhabitants had δ13C values that strongly indicate a very limited mobility range and focus on marine resources (Sealy and Merwe 1988). The persistence of hunter-gatherers into the late Holocene at sites in the southern Natal (Cable 1984) and the northern Cape (Parsons 2003) further confirms variability in the settlement patterns along ecological gradients between highland, fluvial riparian, and coastal lowland settings.
Bone chemistry of two contemporaneous burial sites located within 13 km of one another in the Robberg Peninsula along the southern Cape coast show that there is strong differentiation of dietary patterns between the two populations (Sealy 2006). Individuals buried at Plettenberg Bay show consumption of high trophic marine animals, whereas individuals from Matjes River had a much more broad-based diet with lower trophic marine and terrestrial resources (Sealy 2006). Combined with other archaeological data, the implication of this finding is that there was clear differentiation of subsistence strategies, territories, and cultural practices among late Holocene hunter-gatherers of southern Africa. Therefore, as Neolithic pastoralism spread along the coasts and hinterlands of southern Africa, the perceived usefulness of the resource appears to have varied among the indigenous inhabitants of the region.
The timing and degree of exogenous genetic introgression into indigenous southern African pre-agricultural populations would be helped by detailed genetic studies of living populations across the southern Cone. Unfortunately, most genetic analyses of hunter-gatherer and agricultural populations of southern Africa has been restricted to the geographic outskirts of the region. Y-chromosome and mtDNA analysis of modern Khoe-San populations show deep ancestral links to the L0 haplotype (which is the base of the African population tree according to Gonder et al. 2007) due to intermarriage of the Khoekhoen with indigenous hunter-gatherer populations after 2000 years BP, but there is no genetic discrimination between ethnic groups presently designated as “Khoe” (Khoe-Kwadi language family) and “San” (Ju-ǂHõa language family) (Schlebusch et al. 2012; Soodyall et al. 2008). In contrast, Bantu-speaking populations bear distinct mutations that separate their lineages from the Khoe-San clade and haplotypes (Gonder et al. 2007; Henn et al. 2008; Pickrell et al. 2014; Tishkoff et al. 2009), but these are generalizations to which one cannot ascribe a tight chronology. Linguistically, the same basic relationships bear out unevenly as there is significant linguistic borrowing between the language families although distinct language families can be discriminated between Khoe, Bantu, and other non-Khoe speakers (Güldemann 2008; Hammond-Tooke 2004).
The combined archaeological, genetic, and linguistic evidence from southern Africa supports the notion that exploitation of ecological niches diversified following the introduction of domesticated stock. There is continuation of the LSA Wilton lithic traditions from the early Holocene into the historic period (100 years BP) with only modest changes in tool frequencies and morphology alongside communities heavily dependent on using copper, tin, and iron (Duncan 2002; Mitchell 2005). The evolution of the Smithfield LSA large endstruck scrapers and distinct pottery tradition of the eastern Karoo after 800 years BP shows evolution of formalized stone tool production techniques despite the presence of iron throughout the region (Mitchell 2005). Due to the high amount of regional subsistence diversity, the separation of Neolithic from Iron Age settlements is not straightforward in southern Africa. The overlap of timing and material cultures of the two archaeologically distinguished groups is the result of similar demographic and ecological pressures, but differences in the migration and settlement patterns are worth noting to contextualize modes of early food production in Africa.
Iron Age Pastoralism. Previous models advanced in linguistics supported the notion of a “bow-wave” expansion of Bantu-speaking, iron-using agropastoralists originating from the West Africa via western Lake Victoria who territorially replaced indigenous hunting (Khoekhoen) and pastoral (Nilo-Saharan and Afro-Asiatic) populations (e.g., Vansina 1995). Although the extent to which there was replacement of indigenous hunter-gatherers varies by geography, this theory has been revised with a nuanced view of incremental movements and intermixing of members of the Kordofanian/Bantu language family as they migrated across other portions of the subcontinent (Eggert 2005). Studies of the distribution of African genetic markers clearly demonstrate the migration of two separate, Bantu clades into southern Africa approximately 2000 years ago, but the lineages are hybridized, as is the archaeological evidence of stone versus iron using and hunting versus pastoral versus farming subsistence patterns (Mitchell 2010).
Gross analyses of ceramic assemblages from southern Africa offer a complex view of culture change in the South African Neolithic period. After 2300 years BP, a few thick-walled ceramics (e.g., Situmpa, mentioned above) and thin-walled variants appear within LSA sites associated with mixed foraging-herding economies for about 600 years (Sadr 2008a). The archaeological site of Bambata and toponymically derived ceramic tradition known from western Zimbabwe is argued to have the earliest occurrence of cattle on the Zimbabwe Plateau and is argued to have predated Iron Age settlement of the area (Walker 1983). However, the clear association between domesticated animals and the 2100 years BP radiocarbon date has been questioned because an older radiocarbon date comes from above the Bambata lens (Huffman 2005). Thomas Huffman (2005) also places thin-walled Bambata pottery within the Kalundu ceramic tradition, which is a subvariant of the Chifumbaze Complex, and associated with Bantu-speaking populations. The other subvariant of the Chifumbaze Complex is recognized as Urewe ceramic tradition, which is an East African Early Iron Age ware from western Lake Victoria occurring after 2500 years BP and spread down the Indian Ocean coast after 2000 years BP in a modified form called “Kwale Ware” or “Early Iron Ware.” In Huffman’s (2005) analysis, Bambata pottery was first produced by Iron Age farmers in Angola and transported by (Khoe-speaking) foragers (who probably kept some domestic stock) into southeastern Africa around 1750 years BP. The style was later adopted by Bantu-speaking agropastoralists of eastern Botswana and western Zimbabwe around 1600 years BP (Huffman 2005). By 2000 years BP, Bambata ceramics and a lithic scraper tradition spread along the Atlantic seaboard to the southern tip of Africa, but disappeared from the archaeological record in the Zambezi and Limpopo watersheds by 1500 years BP (Sadr 2015).
Settlement data indicates that village-dwelling Iron Age agropastoralists who farmed yams and millet entered southern Africa between 1800 and 1400 years BP (Vogel and Fuls 1999). Phytoliths recovered from one of the earliest occupied of such sites, Broederstroom, include wetland species of sedges and herbaceous grasses indicating wetter than present conditions (Huffman 1996). The immigrating Iron Age people were clearly attracted to southern Africa because of its agricultural potential. Pulses of expansion of Iron Age agropastoralists into new niches in southern Africa generally correlates with wetter and warmer conditions than are presently found (Huffman 1996). The presence of the newcomers would have presented a source of resource competition for pastoral populations. However, the linguistic evidence indicates that Khoekhoe-speaking pastoralists remained in northern and northeastern South Africa for at least 1000 years after the arrival of Bantu-speaking populations (Ehret 2008).
Early Iron Age farmers who settled in southern Africa after 1800 years BP appear to have kept sheep, but there is only limited evidence for cattle and goat husbandry. Throughout the Iron Age, farming communities on the Zimbabwe Plateau grew increasingly invested in participating in the trans-Indian Ocean exchange networks, and their settlements were located near stone-walled enclosures that grew into loosely organized states. Between 1400 and 1000 years BP, the distribution of Bambata Ware and other thin-walled ceramics becomes increasingly restricted and thick-walled vessels associated with crop-raising/iron-producing communities is widely distributed throughout southern Africa (Sadr 2008a). Butchering cattle became increasingly important to these farming communities by the Late Iron Age (ca. 700–200 years BP), which has been attributed to the arrival of Nguni and Sotho-Tswana speakers from northeastern South Africa (Badenhorst 2010).
Successive waves of Bantu-speaking migrants into southern Africa during the Middle and Late Iron Ages (Hammond-Tooke 2004) as well as indigenous changes in niche fulfillment strategies resulted in a complex cultural mosaic at the time of European colonization. For at least the last 270,000 years, rainfall in eastern and southern Africa has been antiphasing, meaning that as rainfall decreased in eastern Africa, it was simultaneously increasing in southern Africa, and vice versa (Simon et al. 2015). Such climatic factors have catalyzed interregional migrations of agropastoral people throughout the late Holocene seeking good pasture and farmland. With the immigration of new populations into southern Africa, subsistence and settlement patterns of indigenous Khoekhoen changed – in some cases moving toward a foraging-intensive exploitation pattern (Wilmsen 1994) or, in other cases, they moved to a livestock-intensive exploitation pattern (Kinahan 1991).
Discussion of the Evidence
Neolithic pastoralists entering East Africa entered into new niches at the same time ecological conditions in their homelands were desiccating. It is likely that pastoralists entered the Ethiopian Highlands for the same reasons at approximately the same time, but they are less archaeologically visible than in Lake Turkana. Pastoralism took shallow roots south of Lake Turkana as a seemingly minor compliment to indigenous foraging systems. Marshall (2000) has suggested that raiding of herds may have been the mechanism in which livestock appear in new niches. In any event, the evidence is relatively clear that early pastoralism had little effect on indigenous settlement and subsistence practices. Livestock were incorporated as just one more resource within a broad-base subsistence regime.
Given the evidence for antiphasing climates between eastern and southern Africa in the late Holocene, livestock husbandry in southern Africa may have arisen from small populations of environmental refugees leaving East Africa around 2000 years ago and settling in the greener pastures of southern Africa (Scott et al. 2012). The low archaeological visibility of these migrants is typical of itinerant pastoralists, who tend to leave few traces on the landscape due to the fact that having copious amounts of material culture is antithetical to a mobile lifestyle (Shahack-Gross et al. 2003). As in the pioneer phases of livestock introgression in eastern Africa, some southern African foragers recognized the newcomers as bringing an innovation that could be useful if resources became scarce.
Changing environments and cultural practices in eastern and southern Africa during the late Holocene incubated livestock husbandry until it began to assume an ever-more important role in dictating settlement and subsistence practices. In the savanna plains and Rift Valley, there is strong evidence for the development of a “cattle-first” economy after 3000 BP. However, there remained many foraging-first communities across eastern and southern Africa interacting with their agricultural and pastoral neighbors into modern times. From a cultural ecology perspective, the degree to which people relied on domesticated and wild resources will vary by geography and time, depending on the ability of different communities to feed themselves from the land or subsist through trade with their neighbors.
In East Africa, late Holocene economies of the Highlands and Coastal Plains continue to practice the broad-based subsistence regimes inherited from middle Holocene foraging-herders. Elmenteitan communities in the Highlands were vertically transhumant, but had a broad subsistence base. In the Coastal Plains and Lake Victoria region, pastoralists were tethered to resource-rich riparian areas, but little is known of how they lived in the immediate coastal hinterlands. Evidence from Galana River in southeastern Kenya and the Lower Rufiji River in Tanzania show that there are cultural connections to livestock-dependent pastoralists living in the interior, but the indigenous subsistence economy and lithic traditions remain constant throughout the Later Stone Age Neolithic period. A similar pattern is followed among Kansyore sites in the Lake Victoria region, albeit with varying degrees of integration of PN subsistence and tool production techniques (Prendergast 2010a). In these lacustrine and fluvial settings, domesticated animals were integrated into riparian foraging strategies, which included hunting, fishing, and collecting edible plants and invertebrates that grew in aggrading floodplains.
Immigration of Bantu-speaking populations who raised yams and millet and introduced iron production represented a new subsistence niche, which may have clashed with transhumant pastoralism. Until recently, most scholars believed that approximately at 2000 years BP, a unified Bantu cultural complex immigrated to the northern Lake Victoria region from western Africa bringing Urewe tradition ceramics, iron working and sedentary farming with them (Vansina 1995). However, well-dated archaeological data sets from Lake Victoria (Lane et al. 2007), archaeogenetics (Pickrell et al. 2014; Skoglund et al. 2017) and revised linguistic analyses (Ehret 1998) indicate that the introduction of these cultural features were asynchronous. Regardless, the overall effect of these new cultural features realigned subsistence practices in the region with some pastoral groups eschewing cultivation altogether, while others (particularly in the Lake Victoria Basin, Highlands and Coastal Plains of the Indian Ocean) adopted plant cultivation as yet another resource that could be utilized as needed.
The impact of agricultural/grazing practices and iron production significantly altered the distribution of woody and grassy taxa across eastern Africa after 5000 years BP. Burning and tree coppicing were intentional landscape management tools undertaken during the Neolithic across the African continent (Boivin et al. 2016). After 2000 years BP, significant shifts in the distribution of grasslands at the expense of primary-growth forests found in lakes all across the western Rift Valley are evidence of increasing proprietorship of land. Intensive management of patches of land would result in circumscription of territories that was not necessarily beneficial to unrestricted movement of transhumant pastoralists. In this case, the indigenous populations were forced to adopt aspects of the new economy, move to a new area, or attempt to maintain a transhumant pastoral lifestyle in an increasingly territorially-circumscribed landscape.
Some groups clearly remained rooted in the Neolithic pattern practicing transhumant pastoralism as evidenced by sites dating to after 3000 years BP on the Mara Plains and Laikipia Plateau of the Central Rift Valley and in the East African Highlands. It is likely that these pastoralists traded animal products and byproducts with their farmer/forager neighbors, but there were also likely clashes over access to land in which different cultural groups competed for resources. Such clashes were commonly documented in the ethnographic record beginning in the nineteenth century. Maa-speaking pastoralists frequently fought with Kikuyu and Luo groups in an attempt to gain pasture in the fertile Central Highlands and Lake Victoria Basin, respectively (Galaty 1993). During the nineteenth century, the highly fatal cattle disease rinderpest had devastated Maasai herds and severely stressed intergroup relations in the form of rampant livestock raiding (Sobania 1993). A further climatic stress is also possible occurring during the end of the Little Ice Age (AD 1300–1850) in which wetter precipitation regimes gave way to much more arid conditions than had been witnessed in several 100 years (Verschuren 2001). A cultural antipathy developed amongst the Maa for their neighbors, and vice versa in which neither group viewed the others as humans (Berntsen 1979). Such a cultural proclivity is common amongst groups who compete for a common resource pool. For the purposes of the present discussion, it is significant to the extent that niche fulfillment under mutually symbiotic conditions can quickly be thrown off-balance when resource bases become strained.
The move of pastoralism across the Zambian and Zimbabwe plateaus into southeastern Africa was rapid and features of the Bantu expansion can be interpreted as a having immigrated as a “package.” When viewed through the lens of John Alexander’s (1980) and Paul Lane’s (2004) analyses of Neolithic frontiers, the colonization of this broad area occurred so rapidly that the “settling in” phase of frontier expansion had little time to occur. In this sense, the moving frontiers of incipient East African pastoralism (ca. 4000–3000 years BP), eastern equatorial Africa (3000–2000 years BP) and southern African pastoralism (2000–1500 years BP) shared some common aspects (Sadr 2003). The initial spread of domesticated animals into eastern equatorial Africa between 4000 and 3000 years BP have many of the characteristics of Sadr’s (2003) “foragers who herd.” There is no evidence for radical changes in indigenous tool technologies, and hunting and gathering seems to persist in much the same way that it did before livestock were introduced (Dusseldorp 2016). After a period of incubation in the Turkana Basin, pastoral occupation of eastern equatorial Africa after 3000 years BP diversifies into a range of niche fulfillment specialists, which adapted further after the arrival of Bantu agro-pastoralists after 2000 years BP. Likewise, after incubating in eastern equatorial Africa from 3000 to 2000 years BP, the spread of pastoralism into southern Africa was initially limited, but diversified over time as niches filled and complex exchange networks evolved.
New demographic pressures faced after sedentary farmers circumscribed access to pasture lands encouraged some of these pastoral communities to look for new pastures elsewhere. The expansion southward represented the only possible route for expansion given the dense stands of rainforest located to the west, established and entrenched pastoral communities to the north and ocean to the east. By 1600 years BP, pastoralism had reached the southern tips of Africa, but the migrations left little obvious archaeological footprint on the landscape that would separate them from the indigenous economies. In this regard, despite genetic and linguistic evidence for a demic expansion of Khoekhoen at the same approximate time that small domestic stock begin appearing at a few sites in southern Africa, the overall subsistence economy remained weighted toward hunting and gathering for several 100 years until the arrival of the Bantu speaking agro-pastoralists. As in the East African case, the abundance of wild resources in the wetter-than-present southern Africa at around 2000 years BP provided many subsistence options, and keeping and culling numerous quantities of domesticated animals does not appear to have been culturally preferred.
Within 200 years of the initial appearance of livestock in southern Africa, farming communities expanded into the region. This migration further filled the ecological niches on the African subcontinent, so new relationships between pastoral, farming and foraging communities were forged. As Andrew Smith (1992) states, there were three choices left for indigenous hunter-gatherers who were in southern Africa as pastoralists filled their niches: (1) adopt a pastoral lifestyle themselves, (2) enter into a subservient patronage relationship with food producers, or (3) move into marginal environments and exploit lower-ranked resources. Based on the present-day archaeological evidence, it seems that there was a combination of all three factors amongst the populations of eastern and southern Africa between 2000 and 1600 years BP.
In many other regions of the world, the Neolithic was a transformative event in which agriculture permanently so radically changed the landscape that foraging was no longer possible. However, this was not the scenario in eastern and southern Africa. The subsistence practices of individual communities were and remain fluid to this day. Many! Kung of the Central Namib Desert have shifted between foraging-and pastoral-dependent economic strategies for hundreds of years (Kinahan 1991). Domesticated animals in eastern and southern Africa represented just one more subsistence choice in a quiver holding many arrows.
The eastern and southern African Neolithic is diverse and has evolved throughout the Holocene to incorporate many different subsistence and settlement practices. Early pastoral communities in eastern Africa from 5000 to 3000 BP had broad-based subsistence strategies and remained tethered to localized resource bases. After 3000 BP, pastoralists on the grassland plateaus and Rift Valley were seasonally mobile and appear to have restricted their diets to meat, animal byproducts and probably other foods, which do not leave traces in the archaeological record. Pastoralists living in the East African Highlands and Coastal Plains of the Indian Ocean continued to subsist on a wide array of wild products, and continued to use domesticated animals as a secondary food source.
Experimentation with small-scale horticulture is possible prior to the emergence of intensive yam and millet cultivation after 2000 BP. It is likely that demographic pressures placed restricted access on transhumance after sedentary farming took hold in the fertile regions of East Africa. Mobile pastoralists chose to migrate south into the Zimbabwe Plateau, Miombo Woodlands, Mozambiquan Coastal Plains and southern African Highveld, filling most of the viable pastoral niches within 400 years subsequent to leaving East Africa. The archaeological invisibility of the Neolithic (Khoekhoen) herders suggests that the pastoralists who entered southern Africa were highly mobile—carrying only the bare essentials, which would have left a small footprint on the landscape. The southern African Neolithic manifested in the form of foragers who kept some domesticated animals, and, as Bantu Iron Age agropastoralists entered the region, limited plant cultivation was practiced as well.
The beginning of the Iron Age in southern Africa after 2000 years BP reconfigured niche fulfillments and subsistence choices, but the ethnic legacies of these migrations endure genetically and linguistically into the present day. The persistence of foraging and herding subsistence strategies across eastern and southern Africa into modern times as well as the continued use of stone tools despite the presence of metal technologies attests to the success of these adaptations in these settings. To this extent, the end of the Neolithic is diffuse and cannot be ascribed on the basis of the beginning of the Iron Age. Iron Age and Neolithic lifeways were operating in tandem throughout the late Holocene with people shifting subsistence strategies regularly, entering into new exchange relationships or fighting for land with other communities based on what they needed to do to survive that particular year. This legacy persists to the present day as flexible modes of subsistence remain critical aspects of modern farming systems in Africa.
- Alexander, J. 1980. The spread and development of iron-using in Europe and Africa. In Proceedings of the 8th Panafrican congress of prehistory and quaternary studies, ed. R.E. Leakey and B.A. Ogot, 327–330. Nairobi: The International Louis Leakey Memorial Institute for African Prehistory.Google Scholar
- Ambrose, S.H. 1984. The introduction of pastoral adaptations to the highlands of East Africa. In From hunters to farmers: The consequences of food production in Africa, ed. J.D. Clark and S.A. Brandt, 212–239. Berkeley: University of California Press.Google Scholar
- Badenhorst, S. 2009. Phytoliths and livestock dung at Early Iron Age sites in southern. Africa South African Archaeological Bulletin 64: 45–50.Google Scholar
- Barnett, T. 1999. The emergence of food production in Ethiopia, Cambridge monographs in African archaeology 45, BAR international series S763. Oxford: Archaeopress.Google Scholar
- Barthelme, J.W. 1985. Fisher-hunters and Neolithic pastoralists in east Turkana, Kenya, BAR international series 254. Oxford: British Archaeological Reports.Google Scholar
- Berntsen, J.L. 1979. Pastoralism, raiding, and prophets: Maasailand in the nineteenth century. Ph.D., University of Wisconsin-Madison.Google Scholar
- Brandt, S.A. 1984. New perspectives on the origins of food production in Ethiopia. In From hunters to farmers: The causes and consequences of food production in Africa, ed. J.D. Clark and S.A. Brandt, 173–190. Berkeley: University of California Press.Google Scholar
- Cable, C. 1984. Economy and technology in the Late Stone Age of southern Natal, Cambridge monographs in African archaeology 9. British archaeological research international series 201. Oxford: British Archaeological Reports.Google Scholar
- Castañeda, I.S., J.P. Werne, T.C. Johnson, and T.R. Filley. 2009. Late quaternary vegetation history of southeast Africa: The molecular isotopic record from Lake Malawi. Palaeogeography, Palaeoclimatology, Palaeoecology 275: 100–112. https://doi.org/10.1016/j.palaeo.2009.02.008.CrossRefGoogle Scholar
- Chritz, K.L., F.B. Marshall, M.E. Zagal, F. Kirera, and T.E. Cerling. 2015. Environments and trypanosomiasis risks for early herders in the later Holocene of the Lake Victoria basin, Kenya. Proceedings of the National Academy of Sciences 112: 3674–3679. https://doi.org/10.1073/pnas.1423953112.Google Scholar
- Eggert, M.K.H. 2005. The bantu problem and African archaeology. In African archaeology. Blackwell studies in global archaeology, ed. A.B. Stahl, 301–326. Malden: Blackwell.Google Scholar
- Ehret, C. 1998. An African classical age: Eastern and southern Africa in world prehistory history, 1000 BC to AD 400. Charlottesville: University Press of Virginia.Google Scholar
- Ehret, C. 2008. The early livestock-raisers of southern Africa. Southern African Humanities 20: 7–35.Google Scholar
- Galaty, J.G. 1993. Maasai expansion and the new east African pastoralism. In Being Maasai: Ethnicity and identity in East Africa, ed. T. Spear and R. Waller, 61–86. London: James Currey.Google Scholar
- Güldemann, T. 2008. A linguist’s view: Khoe-Kwadi speakers as the earliest food-producers of southern Africa. Southern African Humanities 20: 93–132.Google Scholar
- Haaland, R. 1995. Sedentism, cultivation, and plant domestication in the Holocene Middle Nile Region. Journal of Field Archaeology 22: 157–174.Google Scholar
- Hammond-Tooke, W.D. 2004. Southern Bantu origins: Light from kinship terminology. Southern African Humanities 16: 71–78.Google Scholar
- Huffman, T.N. 2001. The Central Cattle Pattern and interpreting the past. Southern African Humanities 13: 19–35.Google Scholar
- Huffman, T.N. 2005. The stylistic origin of Bambata and the spread of mixed farming in southern Africa. Southern African Humanities 17: 57–79.Google Scholar
- Huffman, T.N. 2007. Handbook to the Iron Age: The archaeology of precolonial farming societies in southern Africa. Pietermaritzburg: University of Kwa-Zulu Natal Press.Google Scholar
- Karega-Mũnene. 2002. Holocene foragers, fishers and herders of western Kenya, BAR international series 1037, Cambridge monographs in African archaeology. Oxford: Archaeopress.Google Scholar
- Kinahan, J. 1991. Pastoral nomads of the central Namib Desert. Windhoek: New Namibia Books.Google Scholar
- Leakey, L.S.B. 1931. Stone age cultures of Kenya colony. Cambridge: Cambridge University Press.Google Scholar
- Lesur, J. 2004. Exploitation de la diversité faunique et début du pastoralisme dans la corne de l'afrique à l’Holocène. Paris: Université de Paris I.Google Scholar
- Linseele, V., W. Van Neer, S. Thys, R. Phillipps, R. Cappers, W. Wendrich, and S. Holdaway. 2014. New archaeozoological data from the Fayum “Neolithic” with a critical assessment of the evidence for early stock keeping in Egypt. PLoS One 9: e108517. https://doi.org/10.1371/journal.pone.0108517.CrossRefGoogle Scholar
- Linseele, V., S.J. Holdaway, and W. Wendrich. 2016. The earliest phase of introduction of southwest Asian domesticated animals into Africa. New evidence from the Fayum Oasis in Egypt and its implications. Quaternary International 412: 11–21. https://doi.org/10.1016/j.quaint.2015.12.028.CrossRefGoogle Scholar
- Marshall, F. 2000. The origins and spread of domestic animals in East Africa. In The origins and development of African livestock: Archaeology, genetics, linguistics and ethnography, ed. K.C. MacDonald and R.M. Blench, 191–221. London: UCL Press.Google Scholar
- Marshall, F., and A. Negash. 2002. Early hunters and herders of northern Ethiopia: The fauna from Danei Kawlos and Baati Ataro rockshelters. In: Society of Africanists archaeologists 16th Biennial conference, University of Arizona, Tucson, 18–21 May 2002.Google Scholar
- Mitchell, P. 2005. Modeling Later Stone Age societies in southern Africa. In African archaeology. Blackwell studies in global archaeology, ed. A.B. Stahl, 150–173. Malden: Blackwell.Google Scholar
- Mitchell, P. 2010. Genetics and southern African prehistory: An archaeological view. Journal of Anthropological Sciences 88: 73–92.Google Scholar
- Out, W.A., P. Ryan, J.J. García-Granero, J. Barastegui, L. Maritan, M. Madella, and D. Usai. 2016. Plant exploitation in Neolithic Sudan: A review in the light of new data from the cemeteries R12 and Ghaba. Quaternary International 412: 36–53. https://doi.org/10.1016/j.quaint.2015.12.066.CrossRefGoogle Scholar
- Phillipson, D.W. 1989. The first south African pastoralists and the Early Iron Age. Nsi 6: 127–134.Google Scholar
- Sadig, A.M. 2010. The Neolithic of the Middle Nile Region: An archaeology of Central Sudan and Nubia. Kampala: Fountain Publishers.Google Scholar
- Sadr, K. 2008b. Invisible herders? The archaeology of Khoekhoe pastoralists. Southern African Humanities 20: 179–203.Google Scholar
- Simon, M.H., Ziegler, M., Bosmans, J., Barker, S., Reason, C.J.C., and Hall, I.R. 2015. Eastern South African hydroclimate over the past 270,000 years. Scientific Reports 5. https://doi.org/10.1038/srep18153.
- Skoglund, P. et al. 2017. Reconstructing prehistoric African population structure. Cell 171(1): 59–71. https://doi.org/10.1016/j.cell.2017.08.049
- Smith, A.B.. 1992. Pastoralism in Africa: Origins and development ecology. London: Hurst & Company.Google Scholar
- Sobania, N. 1993. Defeat and dispersal: The Laikipiak and their neighbours at the end of the nineteenth century. In Being Maasai: Ethnicity and identity in East Africa, ed. T. Spear and R. Waller, 61–86. London: James Currey.Google Scholar
- Soodyall, H., H. Makkan, P. Haycock, and T. Naidoo. 2008. The genetic prehistory of the Khoe and San. Southern African Humanities 20: 37–48.Google Scholar
- Trombetta, B., et al. 2015. Phylogeographic refinement and large scale genotyping of human Y chromosome Haplogroup E provide new insights into the dispersal of early pastoralists in the African continent. Genome Biology and Evolution 7: 1940–1950. https://doi.org/10.1093/gbe/evv118.CrossRefGoogle Scholar
- Verschuren, D., J.S. Sinninghe Damste, J. Moernaut, I. Kristen, M. Blaauw, M. Fagot, and G.H. Haug. 2009. Half-precessional dynamics of monsoon rainfall near the East African Equator. Nature 462: 637–641. http://www.nature.com/nature/journal/v462/n7273/suppinfo/nature08520_S1.html.CrossRefGoogle Scholar
- Wilmsen, E.N. 1994. The creation of subsistence foraging in the colonial era. In Pastoralists at the periphery: Herders in a capitalist world, ed. C. Chang and H. Koster, 140–158. Tucson: University of Arizona Press.Google Scholar
- Crowther, A., Prendergast, M.E., Fuller, D.Q., and Boivin, N. 2017. Subsistence mosaics, forager-farmer interactions, and the transition to food production in eastern Africa. Quaternary International. https://doi.org/10.1016/j.quaint.2017.01.014.
- Ehret, C. 2013. Sub-Saharan Africa: Linguistics. In The global prehistory of human migration, ed. P. Bellwood, 96–106. Malden: Wiley.Google Scholar
- Heyer, E., and J. Rocha. 2013. 14 Sub-Saharan Africa: Human genetics. In The encyclopedia of global human migration, ed I. Ness. Oxford: Blackwell. doi: https://doi.org/10.1002/9781444351071.wbeghm814.
- Lane, P.J. 2013. Trajectories of pastoralism in northern and central Kenya: An overview of the archaeological and environmental evidence. In Pastoralism in Africa: Past, present and future, ed. M. Bollig, M. Schnegg, and H.-P. Wotzka, 104–144. Oxford: Berghahn Books.Google Scholar
- Mitchell, P.M. 2016. Later Stone Age hunter-gatherers and herders. In Quaternary environmental change in southern Africa: Physical and human dimensions, ed. J. Knight and S.W. Grab, 385–396. Cambridge: Cambridge University Press.Google Scholar
- Sadr, K. 2013. A short history of early herding in southern Africa. In Pastoralism in Africa: Past, present and future, ed. M. Bollig, M. Schnegg, and H.-P. Wotzka, 171–197. Oxford: Berghahn Books.Google Scholar
- Vogelsang, R., and B. Keding. 2013. Climate, culture, and change: From hunters to herders in northeastern and southwestern Africa. In Comparative archaeology and paleoclimatology. Socio-cultural responses to a changing world, ed. M.O. Baldia, T.K. Perttula, and D.S. Frink, 43–62. Oxford: Archaeopress.Google Scholar