Several decades ago, Harlan et al. (1976) suggested that Africa, outside of the Nile River Valley, might be the most useful setting for developing a fuller understanding of plant domestication and agricultural origins (Harlan et al. , 5). It seems that in Africa the earliest indigenous plant domestication occurred relatively late (ca. 2000 BC) compared to most other regions of the world (Harlan et al, 7-8).
Whether this was due to a method of harvesting that was not artificially selective, such as beating versus cutting with stone or iron sickles, a lack of intentional re-sowing of harvested grains, or reliance in some cases on non-domesticable plants remains unknown, but it seems clear that wild grain collection was part of a variety of adaptive strategies until at least about 2000 BC. Unlike the Near East, most of Africa’s native domestic plants appear to have different temporal and geographic origins.
In other words, crop domestication in Africa did not arise in a single region, but developed from diverse vegetative zones (Harlan et al, 12). From the critical and historical perspectives, it is important to understand and analyze the development of agricultural patterns in any historio-geographical region, African Sahara in this particular case, because it is from there that the first evidence emerges of village-based communities, pastoralism and intensive use of wild grains.
Over the past 75 years, theories of the origins and spread of agriculture have been numerous and diverse. Explanations have ranged from cultural progress, climate change, diffusion of agriculture from single hearths, to population pressure, status enhancement, feasting, and to simply viewing the variety of agricultural approaches around the globe as increasingly extractive adaptations of foraging behavior. Increasingly, however, it appears that multiple factors led to the development of agriculture and that the processes may have been different in each region of the world.
Archaeological evidence from centers of independent domestication provides numerous opportunities to explain the process, but from the critical viewpoint, it gives little insight into what might have been the ultimate stimulus for such a broad shift. Today, the Egyptian Western Desert (also known as the Eastern Sahara or the Libyan Desert) is extremely inhospitable with little or no rainfall, high daily temperatures, relentless sandstorms, and life that can be supported only near the occasional well or oasis (Wendorf and Schild, 1984, 1-5).
Increased rainfall around 9000 BC led to the formation of seasonal ponds around Bir Kiseiba and Nabta Playa (Wendorf and Schild, 1984, 2). Although the Eastern Sahara remained unpredictable, peoples migrating west from the Nile Valley or from the desert to the south began to temporarily inhabit its better-watered areas (Close and Wendorf, 64).
No structures, storage pits, or wells were recovered from the earliest sites, and pottery was rare (Wendorf and Schild, 1984, 5). Grinding stones were present in the oldest levels, and the plant remains suggest reliance on wild grasses (Wendorf and Schild, 1998, 99). Wild animals such as hare and gazelle comprised the majority of faunal remains, and domesticated cattle were possibly included in the subsistence regime (Wendorf and Schild, 1998, 103).
By 8000 to 7000 BC, the area around Nabta was scattered with desert lakes and dotted with the trees of Tamarix, Acacia, and probably Ziziphus, swampy plants (sedges), and wild grasses (Close and Wendorf, 68). Occupation of the Western Desert was still likely seasonal, with abandonment during the summer monsoons. The sites were larger than those of the previous period, and the remains of small and large huts, bell-shaped storage pits, and deep wells suggest intensified habitation (Close and Wendorf, 69).
Lithics, bone points, grinding stones, and pottery were present (though pottery was still somewhat rare), and the fauna continued to consist mainly of hare, gazelle, and possibly domesticated cattle (Wendorf and Schild 1998, 107). The evidence for domesticated cattle in these earliest levels is debated. Bones, tentatively identified as such, mainly teeth and foot remains, are morphologically similar to both modern domesticated and wild cattle (Bos primigenius f. taurus and B. rimigenius, respectively), but not to other large bovids in the area. Gautier argues for the presence of domesticated cattle rather than wild cattle because the latter probably could not survive on their own in an arid climate without the aid of humans to guide them to known water sources (qtd in Close and Wendorf 1984, 61-62). Support for domesticated cattle comes also from the lack of bones from medium-sized bovids that typically roam with wild cattle (Wendorf and Schild 1998, 108).
Cattle bones are present but not common in the assemblages, which is used to argue for cattle-keeping (for milk and blood) rather than for cattle-eating (Close and Wendorf, 66). Interestingly, Close and Wendorf suggest that it was this expansion into the Sahara that might have pushed cattle-herders towards cattle-keeping and not slaughter, as during the same time in the Nile Valley, cattle apparently were being killed for consumption and not maintained for their products (Close and Wendorf, 68).
In addition to hunting, and cattle milk and blood, the collection of wild plants also provided food. The best studied plant remains come from the site of E-75-6 at Nabta Playa, dating to around 6000 BC (Wasylikowa, 128). Wendorf and Schild interpret the sites of Nabta Playa as representing an important transition in prehistory, that of incipient domestication (Wendorf and Schild, 1998, 105). The intensive use of wild grains by pastoralist-hunters suggests a broad-spectrum approach to subsistence, but one that also incorporates semi-sedentism and delayed use of resources.
Although the pastoralists at Nabta Playa apparently revisited the same locations on a seasonal basis, they probably were forced to remain mobile due to their reliance on cattle and the need for abundant grass cover. Archeologists and historians suggest that groups migrating from the west introduced domesticated African grains to the upper Middle Niger Delta (MND) is has been supported by material remains through various archeological sites (McIntosh, 56).
For instance, ceramics and bone harpoon-type points with affinities to sites in the Mema and Dhar Tichitt suggest that there was some early interaction or occupation at Dia by fisher-forager and agro-pastoralist groups from these more western areas. Evidence from Dhar Tichitt suggests that domesticated millet was introduced prior to 1900 BC, and that millet farming and herding existed well before 600 BC (McIntosh, 71). Ceramics from Mema sites indicate that indigenous fisher-foragers first inhabited the Mema area, but by 1300-800 BC, pastoralist immigration into the region had begun.
It has been proposed by Mcintosh that these groups of herders and fishers might have assimilated to some degree in the Mema, and then perhaps fissioned into proto-Bozo and Nono groups upon entering the modern MND sometime between 800 and 400 BC (McIntosh, 79). Movement into the deeper channels of the upper MND was likely one response to increasing desiccation of the ancient floodplain margins and encroachment of the Sahara during the early first millennium BC. The human-plant relationship at MND appears from the earliest times to be based on rice farming and collection of wild plant resources.
This trend continues throughout the occupation of the sites, even during periods of seasonal habitation or partial abandonment (Horizons II and III of Dia). Early in the second millennium however, several species (pearl millet, bread wheat, and cotton) occur that suggest the development of new or intensified relationships between Dia and the outside world. The increased presence of pearl millet noted especially on Mara probably signals enhanced trade or exchange with other communities, or perhaps the movement of new peoples into the area.
Mcintosh writes of oscillating drying trends during this time that might have allowed cultivation of pearl millet in areas previously too wet, perhaps at Dia or at outlying hamlets (Mcintosh, 83). This important cereal was likely domesticated somewhere between the Sahara and the Sahel of West Africa. The earliest evidence of domesticated pearl millet comes from Tichitt, dating 1900-1500 BC, and from Birimi in northern Ghana, where two grains were directly dated to 1740 BC and 1130-1250 BC (McIntosh, 93).
Pearl millet occurs frequently at later sites and is a common and important cereal across much of West Africa. The four bread wheat grains found on both Shoma and Mara, one grain directly radiocarbon dated to AD 779-1157, may also signal trade, or more likely, visitors from abroad. Native to west Asia and introduced into North Africa by way of Egypt, these wheat grains probably made their way to MND via one of the major Saharan trade towns such as Sijilmasa, where to according to medieval Arabic travelers and traders, wheat was cultivated (McIntosh, 99-100).
In sum, it increasingly appears that there was an independent domestication of cattle in the eastern Sahara around 8000 BC, well before the introduction of cattle, goat, and sheep from the Near East around 5000 BC. Practicing a broad-spectrum approach to food getting, these early herders spread west and south across the Sahara, eventually entering West Africa around 2000 BC.
The earliest domesticated grass (pearl millet) occurs around this time in a broad band across the southern Sahara and Sahel beginning earliest at Dhar Tichitt (Mauritania) and moving rapidly eastward to Lake Chad (northeastern Nigeria) by about 1200 BC (Wendorf and Schild, 1998, 122). These sites are invariably associated with the remains of domesticated cattle, suggesting that Saharan pastoralists introduced domesticated grasses into sub-Saharan Africa and played a pivotal role in the development of other African regions.