The Oxford Research Encyclopedia of African History will be available via subscription on April 26. Visit About to learn more, meet the editorial board, learn about subscriber services.

Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, AFRICAN HISTORY ( (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 21 April 2019

Development of Plant Food Production in the West African Savannas: Archaeobotanical Perspectives

Summary and Keywords

The West African savannas are a major area of independent plant domestication, with pearl millet, African rice, fonio, several legumes, and vegetable crops originating there. For understanding the origins of West African plant-food-producing traditions, it is useful to have a look at their precursors in the Sahara during the “African humid period” between 10,500 and 4,500 years ago. The Early and Middle Holocene Saharan foragers and pastoralists intensively used wild grasses for food but did not intentionally cultivate. Due to increasing aridity in the late 3rd millennium bce, the pastoralists migrated southward into the savanna zone. In this context pearl millet was domesticated and spread rapidly in West Africa during the 2nd millennium bce. It was first cultivated by agro-pastoral communities, predominantly on a small scale. The 1st millennium bce was a transitional phase: most of the early agricultural societies disappeared, but it was also a time of numerous economic and social innovations. Due to increasing aridity, the floodplains around Lake Chad and the valleys of the rivers Senegal and Niger became accessible to farming populations after 1000 bce. In the 1st millennium ce, agriculture intensified, with mixed cultivation of cereals and legumes and the integration of new African domesticates, such as sorghum, fonio, roselle, and okra. Pearl millet remained the major crop in most areas, while sorghum dominated in northern Cameroon. Imported wheat, date palm, and cotton appeared in the first half of the 2nd millennium ce. The combined exploitation of cultivated cereals, legumes, and wild fruit trees (e.g., shea butter tree) in agroforesty systems eventually resulted in a cultural landscape as it is still visible in West Africa today.

Keywords: cultivation, domestication, Neolithic, Iron Age, West Africa, Sahara, pearl millet, sorghum, African rice, cultural landscape

Theoretical Background

The development of food production is one of the most important turning points in human evolution. For more than two million years, early humans had been relying on the exploitation of wild plants and animals for food. Only later did these plants and animals become more and more under human control. Although management of wild plants and animals to increase their potential as food resources may have begun already in the Late Pleistocene, more than 20,000 years ago, it is only in the Holocene after 9,600 bce, with its warmer and more stable climate, that domesticated plants and animals appear in many areas of the world. A plant is considered to be domesticated when genetic and morphological changes become apparent as a result of conscious or unconscious selection through human activities, and the species become more dependent on humans for reproduction. Cultivation, by contrast, comprises the entirety of human activities designed to increase the yield of economically important plants, be they wild or domesticated. At least in the primary centers of plant domestication, there is a long transition phase from initial cultivation to full farming, eventually relying completely on domesticated plants.

Plant cultivation and domestication are usually connected to distinct sociocultural and economic changes, such as demographic growth, technological innovation, increasing sedentism, and intensification of resource exploitation. Depending on geographical and sociocultural factors, the timing and the trajectories to plant-food production show a high degree of variability in different regions of the world. The pathway to plant food production in Africa is unique in two major aspects: (1) In contrast to the other continents where plant domestication precedes or is contemporary with animal domestication, in Africa the reverse pattern can be observed: mobile pastoralism with domesticated cattle and small livestock developed several thousand years before plants were domesticated (“cattle before crops”).1 (2) As a consequence, the transition to plant-food production happened quite late, in comparison to other primary centers of plant domestication, such as China, the Near East, or Mesoamerica.2 The phenomenon is related to specific geographical conditions in Africa, notably the paleo-environmental history of the “Green Sahara,” which offered a favorable environment for mobile foragers and pastoralists between 9600 and 2500 bce and may have negated the need for more extensive cultivation leading to domestication.

Since Gordon Childe first described the “Neolithic Revolution,” food production has been viewed as a central element for the definition of the Neolithic in the Near East and Europe.3 But the different geographical areas in Africa have highly diverse trajectories to food production which are difficult to subsume in a common framework. The absence of contemporaneity between initial animal and plant food production, but also of sedentism, storage technology and ceramics, traditionally regarded as markers for the Neolithic, precludes the use of the term for Africa.4 Similar difficulties arise in the definition of the Iron Age.5 Therefore, this paper tries to avoid “Ages and Stages” terms as much as possible and uses instead absolute chronological designations. For pragmatic reasons, the term “Iron Age” is used according to the definition of de Corse & Spiers to describe societies with knowledge of iron production from c. 500 bce until 800 ce.6

History of Archaeobotanical Research

Archaeological work on the history of food production started quite late in Africa. While in the Near East and Central America the first excavations with the aim to find evidence for plant cultivation were already conducted shortly after World War II, it was not before the 1980s that systematic archaeological exploration and excavations with archaeobotanical participation started in the Sahara and in sub-Saharan Africa. The delayed interest in the history of African crops can be traced back to the dominant influence of Vavilov’s theory on the centers of plant domestication, developed in the 1920s.7 Vavilov saw Africa only as a passive recipient of introduced crops, except for the montane hinterlands of the Mediterranean coast and the Ethiopian highlands. Vavilov’s centers were refuted in a seminal publication by plant geneticist Jack Harlan, who brought the indigenous African crops into the focus of attention.8 By that time, knowledge about the history of African crops was mainly based on hypothetical assumptions and modern ethnographic and botanical evidence, apart from a handful of archaeological studies with minor archaeobotanical contributions. Harlan emphasized a new paradigm for the archaeological research on the origins of African plant domestication. He used the distribution of the potential wild ancestors of the crops for the definition of a domestication center in the eastern African highlands and a vast “non-center” in the savanna and forest belt south of the Sahara, with the majority of potential domestication areas situated between Senegal and Lake Chad.

Some West African multidisciplinary archaeological research projects, including systematic sampling for plant remains, started in the 1990s. But in a wider perspective, the intensity of archaeobotanical exploration in West Africa remains heterogeneous, and research activities have been and still are limited and regionally restricted. Detailed archaeobotanical data have been published from Mauritania, Senegal, Mali, Ghana, Burkina Faso, and Nigeria; archaeobotanical studies for Benin are in progress. But for some countries, such as Togo, Côte d’Ivoire, or Guinea, virtually no information is available. The dots as well as the gaps in the West African archaeobotanical landscape represent the absence of evidence and should not be construed as evidence of absence. In the early 21st century, the West African savanna regions witness today serious political and civil unrest; some of the areas where intensive archaeological research was conducted from the 1990s onward are no longer accessible. Prominent examples are northeast and central Nigeria, Mali, northern Burkina Faso, and Niger, which have become insecure through Islamist terrorism and crime including kidnapping and robbery. It seems, unfortunately, that the archaeobotanical data for some areas, as presented in this compilation, will remain the state of the art for the coming years.

This article mainly refers to plant remains from excavated sites as the major source of information and interprets them in their archaeological context. However, the full potential of interpretative approaches for the reconstruction of ancient plant use can only be tapped in a wider discourse with other disciplines, including ethnography, archaeozoology, archaeochemistry, genetics, linguistics, and palaeoenvironmental research.

Modern Climate, Vegetation, and Agricultural Systems

The term “savanna,” as it is used here, describes an ecosystem with coexisting grass and tree species, in a climate with seasonal rainfall. The West African savannas include a broad range of different vegetation types, from almost pure grasslands with a very sparse tree cover to dense woodlands with large trees in dense stands and some grass in the herb layer. Precipitation is restricted to the summer months and is largely determined by the changing position of the Intertropical Convergence Zone (ITCZ) over the course of the year.

The vegetation zones are arranged in roughly parallel belts from the Sahara to the evergreen rainforest, in relation to the amount of annual rainfall (Figure 1). The climate of the Sahel, bordering the Sahara to the south, is characterized by high temperatures all year round and annual precipitation of 100–600 mm. Rainfall is mainly restricted to the months between July and September a few months in summer, while the rest of the year remains completely dry. The distribution and amount of rainfall are very irregular in terms of time and space, and drought is common. The vegetation consists mostly of low annual grasses with very few trees, dominated by acacias. The most sustainable economic activity is transhumant pastoralism based on cattle, small livestock, and camels. Around Lake Chad and in the large river valleys, specialized fishing is practiced.

Development of Plant Food Production in the West African Savannas: Archaeobotanical PerspectivesClick to view larger

Figure 1. West Africa’s vegetation zones and annual precipitation.

Map drawn by Monika Heckner.

The Sudan zone experiences annual average precipitation between 600 and 1000 mm. Its natural vegetation is a tree-and-shrub savanna with high perennial grasses that are burned annually. Pearl millet, sorghum, and maize are the main crops cultivated during the rainy season, while fonio and rice (today mostly Asian rice, Oryza sativa) are regionally dominant. The traditional agricultural system is shifting cultivation and agroforestry. The natural savanna is cleared, used for cultivation for a few years, and finally left for a fallow period of five to thirty years to allow soil regeneration. During clearing, wild tree saplings with valuable fruits are protected. In this way, the typical West African park savanna develops, which is a patchwork of fields with large trees as well as fallows in different stages of succession. Regular bush fires, wood collecting, and grazing by domesticated cattle and small livestock also have a significant influence on the vegetation. However, the regeneration capacity of the trees is high, and completely denuded areas are rare in the Sudan zone. In recent decades, cultivation of cash crops, such as peanuts and cotton, has become more and more important, and large permanent fields have replaced the traditional park savannas. The savannas of the Guinea zone with annual precipitation of 1000–1600 mm have a denser tree cover than those of the Sudan zone. They are mainly secondary formations derived from different forest types that are regularly cleared. Shifting cultivation based on sorghum, maize, and root crops such as yams, manioc, and cocoyam is the dominant subsistence type.

Early and Middle Holocene Foragers (9600–2500 bce)

Foraging in the West African Savannas and Forests

For much of the Holocene, hunter-gatherers with microlithic Later Stone Age (LSA) technologies, sometimes with ceramics, inhabited the vast landscapes of the West African savannas and forests, which offered rich wild food resources: honey, edible fruits from trees, tubers such as wild yams, plenty of game, and aquatic animals.9 But very little is known about these foragers, due to the scarcity of well-dated archaeological sites. The rock shelters Bosumpra (Ghana), Shum Laka (Cameroon), and Péntènga (Burkina Faso) offer a glimpse into the life of Early and Middle Holocene African hunter-gatherers.10 All three sites have rich archaeozoological inventories, charcoal, and remains from edible tree fruits. In Bosumpra and Shum Laka, oil palm (Elaeis guineensis) and Canarium schweinfurthii dominate; both species have fruits rich in oil with high nutritional value. In Péntènga edible fruits from fifteen tree species were found; most of them are still collected today and sometimes sold in the local markets. Apart from a few unidentified grass caryopses in Bosumpra, there is no evidence for the use of wild grasses by the inhabitants of the rock shelter.11 This could be the result of excavation techniques because floating and sieving with fine mesh widths was conducted at neither Shum Laka nor Péntènga. But it is also conceivable that the perennial grass species of the humid savannas were not attractive to humans because their seed production is much lower than that of annual grasses distributed farther north under more arid conditions. The development of wild-grass exploitation in Africa, eventually resulting in domestication, did not place in the West African savannas, but in the Sahara.

The “Green Sahara”: A Garden of Eden?

During the hyper-arid Late Pleistocene, the Sahara Desert had extended 400 kilometers farther south than at present, presenting an extremely hostile environment for plants, people, and animals. This changed around c. 13,500 bce and most dramatically after 9500 bce when lakes and swamps began to form in the central and southern Sahara as a result of increasing monsoon rains during the “African humid period.”12 After 8500 bce—when lakes reached their maximum extension—tropical grasslands and savanna-like habitats had spread as far north as 24°N, and the “Green Sahara” was connected to the extensive savanna ecosystems covering the largest part of the African continent.13 The richness in game, aquatic resources, and wild grasses made the Sahara attractive for foragers. But the “Green Sahara” was not a Garden of Eden. Tree species with edible fruits were still scarce. Rainfall was low, and several dry spells occurred even during the more humid phases of the Early and Middle Holocene. In the face of unpredictable availability of water, plant, and animal resources, hunter-gatherers had to employ strategies of high mobility and flexibility in seasonal movements.

Wild annual grasses, mainly from the grass tribe Paniceae, played an important role in the subsistence of Saharan foragers. The invention of pottery was likely an adaptive strategy for the effective exploitation of wild grasses and other plant resources. Grinding equipment and pottery occur at Early Holocene sites in the northern Sahel and the central and eastern Sahara shortly after the beginning of the Holocene around 9000 bce.14 Because the grains of wild grasses need processing before consumption, including chaff removal and cooking, the frequency of grinding stones and pottery can be interpreted as an indirect marker for the role of these resources in the diet.

The best-studied areas with detailed archaeological sequences and ample data on plant and animal exploitation are the eastern Sahara and southwestern Libya. The eastern Sahara was repopulated when the monsoon rains suddenly resumed around 8500 bce. One of the earliest signs of reoccupation comes from the Egyptian Great Sand Sea, which is today one of the most arid and barren areas of the Sahara.15 Exploited plants of the archaeological site of Nabta Playa display a high diversity and include Paniceae grasses, wild sorghum, various small seeded plants, tree fruits, and (though rarely found) tuber plants such as water lily (Nymphaea) and cattail (Typha).16

In the Tadrart Acacus of southwestern Libya, several caves and rock shelters have excellent preservation of faunal and floral remains, in some cases allowing DNA analyses of desiccated plants and residue analysis of potsherds.17 Among the numerous plant species exploited during the Early Holocene, grasses from the tribe Paniceae are especially common, which today are known as important food resources in the southern Sahara and the Sahel, among them Urochloa, Echinochloa, Brachiaria, Setaria, and Pennisetum.18 Glumes of wild Paniceae adhere to some of the well-preserved basketry fragments from the cave sites, demonstrating that sophisticated baskets were produced for collection and storage of grass grains.19

Domesticated animals were integrated into mobile subsistence practices in both the central and eastern Sahara at around 6000 bce.20 While the issue of independent cattle domestication in the eastern Sahara remains controversial, the introduction of sheep and goats from the Near East is undisputed. The early pastoralists left traces in the form of rock art with depictions of domestic animals, especially in the central Sahara. In the 5th millennium bce, there is evidence for consumption and processing of dairy products.21 The exploitation of wild grasses continued, and in combination with pastoralism and some hunting and fishing, this appears to have been the best-adapted strategy for survival in a semi-arid environment with large-scale temporal and spatial variation of pastures and edible plant resources. Even today some populations in the transition zone between the Sahel and the Sahara practice this mixed subsistence and collect wild grasses during their seasonal movements with cattle herds (Figure 2). The combination of pastoralism with the exploitation of wild grasses remained successful until the Sahara finally desiccated completely. In the eastern Sahara, population shifts to the south began around 5300 bce, and mobile pastoralism eventually ended in the southernmost areas of the eastern Sahara by around 1500 bce. In the central and western Sahara, pastoralists moved to the south in the 3rd millennium bce in response to increasing aridity. It is in the context of these migrations that the first domesticated plants appear.

Development of Plant Food Production in the West African Savannas: Archaeobotanical PerspectivesClick to view larger

Figure 2. Traditional basket from Niger for harvesting grains of wild grasses.

Photo by Jennifer Markwirth.

The Beginnings of Plant Cultivation (2500–1000 bce)

During the “African humid period,” the savannas south of the Sahara had not been favorable for pastoralists because of the presence of tsetse-fly–borne trypanosomias, and other diseases harmful to cattle and humans. With increasing aridity, the northern border of tsetse distribution shifted to the south, making the Sahel attractive for immigrants from the Sahara. Similarities in ceramic styles between sites in the southern Sahara and the Sahel, and occasional presence of cattle bones in archaeological sites, support the idea of an immigration of pastoralists from the north.22

The first crop to be domesticated was pearl millet (Pennisetum glaucum). The earliest remains of domesticated pearl millet have been found in the Tilemsi Valley of Mali, dated around 2500 bce.23 The Tilemsi Valley, currently a desert environment, was one of the last refuges for the pastoralists in the desiccating southern Sahara, offering water, rich tree growth, and aquatic animals. The domestication process of pearl millet is still not well understood. Wild pearl millet had not been one of the most preferred wild cereals in the Sahara, and it is unclear why pastoralists selected this species for cultivation and not one of the other wild Paniceae grasses that had already been collected for some thousand years. The pearl millet remains of the Lower Tilemsi Valley are fully domesticated, indicated by two distinct domestication features: club-shaped grains and chaff remains with a firmly attached rachis. Probably the initial domestication of pearl millet took place elsewhere in the Sahara prior to 2500 bce because only a few specimens demonstrate features of the wild cereal. Genetic studies suggest a potential domestication area between eastern Mali through western Niger.24 This hypothesis cannot be currently tested archaeologically because the region is inaccessible to researchers.

After 2000 bce, the knowledge of pearl-millet cultivation spread rapidly in West Africa. In Mauritania, Mali, Ghana, Burkina Faso, and Nigeria, where intensive archaeological research has been conducted, domesticated pearl millet is present in sites of the 2nd millennium bce, albeit in very diverse cultural contexts. With the exception of the Kintampo tradition in Ghana, pearl millet is the only crop in these 2nd-millennium sites. Plant cultivation in the 2nd millennium bce was part of a risk-management strategy with highly diverse subsistence including fishing, hunting, herding, and collecting of wild-plant foods. Pearl-millet cultivation was in most cases only practiced on a small scale, in a system of “low-level food production.”25 Three well-defined cultural traditions in West Africa have seen in-depth research on early plant cultivation and its socioeconomical context: Tichitt in Mauritania, Gajiganna in northeast Nigeria, and Kintampo in Ghana. Although the three traditions are distributed in ecologically quite different areas, they all offer evidence for at least a partial sedentary way of life.

First Agro-Pastoralists of the Gajiganna Complex (Northeast Nigeria)

The Chad Basin of northeast Nigeria is an environment characterized by extensive seasonally flooded clay plains with intermingling sand dunes. Due to high water levels of the Mega Lake Chad, which existed during the African humid period, the area was inaccessible before 2000 bce. When the waters of Lake Chad receded, cattle pastoralists originating from the central Sahara started to settle in small campsites in the Bama Deltaic Complex, which is at the outer reaches of former Mega Lake Chad. The Gajiganna Complex, named after a village in the area where its remains were first detected, has been extensively studied by a team from Goethe University (Frankfurt, Germany).26 Archaeozoological data confirm the importance of cattle and small livestock, but hunting and fishing were also practiced.27 Plant impressions in chaff-tempered ceramics illustrate distinct changes in the use of food plants throughout the Gajiganna Complex between 1800 and 800 bce.28 In the first phase (Gajiganna I) from 1800 to 1500 bce, the people were pastoralists and tempered their pottery with inorganic materials. Domesticated pearl millet appears in the temper between 1500 and 1200 bce (Gajigana IIa), but only in small numbers. During phase Gajiganna IIb (1200–1000 bce), the amount of pearl millet chaff increased significantly, indicating that cultivation became an important activity, with large amounts of waste from threshing and winnowing, available for pottery production. At the same time, there is evidence for a more sedentary way of life, resulting in the formation of settlement mounds that represent permanent hamlets and small villages on the dunes. A continuous element in the subsistence of the Gajiganna people was the use of wild grasses, which were growing in abundance on the seasonally inundated clay plains. Among these grasses are also sorghum and wild rice, which, albeit so common in the Chad Basin, were not domesticated there.

Between Savanna and Forest: The Kintampo Tradition (Ghana)

Sites of the well-studied Kintampo tradition of Ghana, dated between 2100 and 1400 bce, are distributed from the drier wooded savanna in the north of the country to the humid rainforest in the south. In spite of its broad distribution throughout different ecological zones, Kintampo material culture is remarkably homogenous, including pottery, celts, grinding stones, bifacial projectile point, rasps, and decorative objects.29 The Kintampo settlement pattern comprises semi-sedentary villages with daub architecture, as well as upland rock shelters. Plant remains have been studied from Birimi, the rock-shelter B-sites, and Bosumpra Cave.30 The Kintampo people had a mixed subsistence, including the early exploitation of oil palm (Elaeis guineensis), incense tree (Canarium schweinfurthii), and other wild plants, as well as cultivation of pearl millet and cowpea (both directly dated as early as the 2nd millennium bce), hunting, fishing, and herding of cattle and small livestock.31 Corresponding to the large ecological gradient in the distribution of Kintampo sites, subsistence patterns show some variation. The most conspicuous is the dominance of oil palm and incense tree in the southern sites of the forest-savanna ecotone, and their absence in the northernmost site Birimi.

Although unambiguous evidence points to an introduction of domesticated animals and pearl millet from the north, the origin of the Kintampo tradition is still debated. Finds of oil palm from around 10,000 bce in Bosumpra and incense tree dated to the mid-8th millennium bce point to a continuity in the management of wild resources, eventually resulting in early agroforestry practices. But the Kintampo tradition is also strongly influenced by a migrant population coming from the north and adapting to the special ecological conditions in the savanna-forest ecotone.32

Both the pearl-millet caryopses from Birimi and the cowpea seeds from the B-sites have morphological characteristics that clearly attribute them to domesticated varieties. However, it is difficult to assess the role of their cultivation in the overall economy. At Birimi, the almost complete absence of wild grasses is remarkable as opposed to the good representation of pearl millet. As in the earlier LSA sites Shum Laka and Péntènga, the perennial wild grasses of the Sudanian savannas apparently were favored by neither hunter-gatherers nor early food producers, probably due to their low seed production in comparison to annual grasses of the Sahara and the Sahel.

Pearl-Millet Cultivation in the Tichitt Tradition (Mauritania)

In Mauritania, numerous sites of the Tichitt tradition, dated to 2000–200 bce, are concentrated along the escarpments and on the plateau of the Dhars Tichitt, Oualata, and Néma. Sites on the plateau have stone structures and are up to 80 hectares large, thus representing the oldest villages in West Africa.33 Some of the stone structures are interpreted as granaries, suggesting that plant cultivation was an important activity. Today the Tichitt-Oualata area is a hostile desert landscape, but during the 2nd millennium bce, it was suitable for herding, cultivation of pearl millet, and even a sedentary way of life thanks to the presence of springs and lakes at the edge of the escarpment. Pearl-millet finds consist of directly AMS-dated impressions in chaff-tempered ceramics of the Classic Tichitt phases, beginning around 1600 bce. The massive abundance of the chaff implies that the crop was processed near pottery-producing areas.34 Because the pearl-millet remains display distinct domestication markers, it is highly probable that the plant was not domesticated in the Tichitt area, but was introduced from elsewhere already in a domesticated form. Apart from the hackberry (Celtis cf. integrifolia), it is unknown whether other plants were used in the Tichitt area during the 2nd millennium bce because no macrobotanical remains other than the chaff in pottery have been recovered.

Diverse Subsistence Patterns in Mali and Burkina Faso

Some 2nd-millennium bce sites in the Mema on the northwestern banks of the Inland Niger Delta in Mali (southeast of the Tichitt-Oualata area) show cultural affinities with the Classic Tichitt tradition. However, these people were not agriculturalists, but cattle herders who tempered their pottery with bone and wild fonio chaff.35 Another population, the Kobadi folk, whose subsistence was largely based on hunting, fishing, and gathering, also lived in the Méma at the same time. This is an example of multicultural and multi-economic contemporaneity within a single region, which was probably very common during the early phase of plant cultivation.36

Southeast of the Inland Niger Delta in Mali, 2nd-millennium bce pearl millet has been reported from the Windé Koroji Complex and Ounjougou. At Windé Koroji, domesticated pearl millet is present in small numbers between 2100 and 1100 bce.37 After 1600 bce, it is associated with domesticated cattle. Remains of wild fruit trees are also attested, such as Vitex doniana and Lannea microcarpa, which were collected in a gallery forest nearby. Windé Korodji is interpreted as being part of a wider geographical network of mobile populations moving between the Niger floodplain and regions farther south, with variable subsistence patterns, including fishing, hunting, gathering of wild plants, and some seasonal herding.

At Ounjougou on the edge of the Dogon Plateau, domesticated pearl millet, dated around 1800 bce, was detected in a complex series of fluvial sediments. Ounjougou, today situated at the boundary between the Sahel and the Sudanian zone, is not far from the Lower Tilemsi Valley, where pearl-millet cultivation has been attested dating to about 700 years earlier. With its lush vegetation and permanent water bodies, Ounjougou might have provided an alternative settlement area for the Saharan agro-pastoralists pushed to the south with increasing aridity after 2000 bce. The 2nd-millennium pearl-millet finds of Ounjougou, which were not found in a settlement context, indicate the presence of the crop, but do not allow yet for an assessment of its role in subsistence. However, contemporary finds of abundant charcoal and phytoliths are interpreted as signs of increasing aridity and initial disturbance of the landscape, including clearance and bush fires.38

In northern Burkina Faso, there are also a few sites with evidence for pearl-millet cultivation, dated to the end of the 2nd millennium bce.39 These seem to be remnants of mobile populations that left only scarce material remains. At Tin-Akof, some charred pearl-millet grains were found in shallow pits, but the small size of the site rather speaks for a seasonal occupation of hunter-gatherers who practiced small-scale cultivation.40

Development of Plant Food Production in the West African Savannas: Archaeobotanical PerspectivesClick to view larger

Figure 3. Important cultural traditions, excavation areas, and archaeological sites mentioned in the text.

Map drawn by Monika Heckner.

Collapse and Innovation in the 1st Millennium bce

The 1st millennium bce was a period of dramatic changes in West Africa. All early food-producing societies ended, albeit not synchronically, and most of them had no immediate successors in their former core areas. Some researchers have suggested that increasing aridity after 800 bce is responsible for the disappearance of the mobile pearl-millet cultivators in Burkina Faso, and for the collapse of the Gajiganna Complex and the Tichitt Tradition.41 However, palaeoclimatic data are sparse, and interpretation often not unequivocal; the exact temporal correlation between palaeoclimate proxy data and archaeological evidence is in most cases not possible, much less a causal relationship. It is also conceivable that internal economic and social difficulties among the first sedentary communities were responsible for their collapse, and that their modes of food production were not sustainable under the challenging conditions of an environment with strong fluctuations in seasonal rainfall.

Potential reactions of a society facing an ecological or economic crisis are migration to another region with more favorable living conditions, a shift of subsistence practices, application of new technologies, and/or changes in social organization. In Burkina Faso, higher mobility of pastoralist or hunting communities might explain the almost complete absence (or invisibility) of archaeological sites in the 1st millennium bce. Although the conditions under which the first food-producing societies collapsed are not well understood, it is remarkable that contemporaneous with their disappearance, new sedentary traditions emerged, some of them in other ecologically different zones of West Africa.

In terms of technology, the most important innovation of the 1st millennium bce in West Africa was iron metallurgy.42 A few iron objects, slag, and remains of furnaces for iron reduction date between 800 and 400 bce in several West African countries, with chronological insecurity because of a plateau in the radiocarbon calibration curve. There is still no final agreement on the question if iron metallurgy was independently developed in Africa, or if it was introduced from the Mediterranean or the Near East.43 In any case, the knowledge of iron metallurgy does not seem to have changed the modes of agricultural production or the exploitation of wild-plant resources in the 1st millennium bce. Iron remained initially a rare commodity. Just over than a thousand years later, in the early 2nd millennium ce, mass production of iron in West Africa fueled the manufacture of iron tools and would thus contribute to intensified agricultural activities.

Between 400 and 200 bce, a decline of settlement activities is noted in many areas of the West African Sahelian and Sudanian zone, such as in Senegal, northeastern and central Nigeria, and Mali.44 This is contemporary with major population movements from the savanna zone into the central African rainforest.45 Increasing aridity with dry spells and stronger seasonality of rainfall are suggested as being responsible for these dramatic changes, disrupting settlement and subsistence patterns, and eventually leading to their reorganization after 200/300 ce.46

Organizational Complexity and Surplus Production: Zilum and Maibe (Northeast Nigeria)

In the Gajiganna area in northeastern Nigeria, the settlement mounds of the “classic” Gajiganna Complex disappeared after 800 bce and were replaced by sites extending into the plain. Most of these sites are small, with diameters less than 100 m, and are not situated on mounds. But Zilum and Maibe, two sites dated around 500 bce, each extends over more than 10 hectares; their spatial extension is recognizable by accumulations of millions of potsherds. For Zilum the number of inhabitants is estimated to have been between 1750 and 2500, while Maibe might have hosted up to 6,000 persons. These large villages were part of a hierarchically organized settlement system and show signs of organizational complexity that had been unknown before.47 In the geomagnetic plan of Zilum, a ditch is visible, completely surrounding the settlement and interpreted by the excavators as a defensive measure.48 Some archaeological features are related to specialized craft activities, such as tanning, pottery production, and manufacturing of bone points or stone beads.

Increasing social complexity in Zilum is accompanied by distinct economic changes. Hundreds of storage pits illustrate the importance of cultivated crops and indicate surplus agricultural production. Besides pearl millet, which was the basis of their daily meal, the people of Zilum also cultivated cowpea (Vigna unguiculata), a legume with high protein content. The combined cultivation of cereals and legumes, often together in the same field, is an effective method of increasing yields because legumes enrich the soil with additional nutrients provided by nitrogen-fixing bacteria. In many regions all over the world, including the ancient civilizations of the Near East, the combination of cereals and legumes was the basis for the establishment of large sedentary communities.

In comparison with the sites of the Gajiganna tradition, the concentration of animal bones in Zilum is much lower. Probably the additional protein from the cowpea replaced to a certain extent the animal component in the diet, thus contributing to a more sedentary way of life and allowing demographic growth and specialization. However, in spite of their higher productivity, Zilum and Maibe existed for little more than a century and ended around 400 bce, coinciding with a period of strong erosion in the Chad Basin due to increasing aridity and seasonality of rainfall.49

Conservative Pearl-Millet Farmers: Nok Culture (Central Nigeria)

Nok Culture of Central Nigeria, located near the modern Nigerian capital Abuja, has been known since the middle of the 20th century, when elaborate terracotta figurines were detected in alluvial deposits of open-air tin mines near the village of Nok. The terracotta figurines are made of burned clay, and, although each figurine is unique, they show a high degree of stylistic uniformity.

Recently, a long-term multidisciplinary research project of Goethe University shed light on the history of Nok culture by revealing the archaeological context and chronology of the figurines.50 Subsistence, settlement system, and environment, as well as technologies of pottery, terracotta, and iron production of Nok culture were studied in detail. Although the exact origin of the Nok people is unknown, pearl-millet finds of the earliest sites with Nok pottery clearly indicate that they came from the north around 1500 bce. Thus, the early Nok sites fall chronologically into the first wave of expansion of pearl millet from the arid southern Sahara to the more humid savanna zones.51 The fabrication of terracotta figurines started around 1000 bce and continued until circa 400 bce. During this time, the number of sites increased significantly, but no hierarchical settlement system developed. The Nok people lived in small, dispersed settlements, and there are no signs of larger villages or urban centers.

Nok subsistence was based on pearl millet and cowpea cultivated in a mixed farming system. Due to the strongly acidic soils, no animal remains have been preserved in Nok sites, and the contribution of animal products to the diet remains unknown. However, preliminary residue analysis on Nok potsherds suggests that plants and plant products provided the majority of food resources for the Nok people.52 It is improbable that Nok people had cattle because the area lies within the tsetse belt, which would only permit the existence of small livestock. A key component of the diet was the oil plant Canarium schweinfurthii, which might have served as a substitute for animal fats. Although the Nok people started iron metallurgy between 800 and 400 bce, there is no evidence that this new technology changed their mode of agricultural production.

After 400 bce, Nok culture declined, clearly indicated by the abrupt decrease of archaeological sites with the characteristic pottery and terracotta fragments, and it completely disappeared from the area around the beginning of the Common Era.53 It seems that Nok society was conservative, producing highly standardized terracotta figurines over several centuries, and continuing a cultivation system with pearl millet and cowpea even when it was no longer sustainable after 400 bce. From several sites in West Africa, there is evidence for stronger climatic seasonality around 400 bce, which might have led to heavy erosion in the Nok area, and thus to a deterioration of the conditions for cultivation.54

A convincing argument for the conservative nature of the Nok people is that they did not adopt the oil palm, one of the most valuable natural resources of their home territory. Elaeis guineensis is common in the river valleys of central Nigeria, and although there are no available archaeobotanical data for earlier foragers in the area, the oil palm finds of the LSA and Kintampo sites in Ghana clearly illustrate the long and continuous tradition of oil-palm exploitation in West Africa.55 By contrast, the Nok people ignored oil palm, although it was growing in the direct vicinity of their settlements. By around 200 ce, a new population with a different pottery style arrived in the Nok area. They appreciated oil palm and also cultivated new crops, most notably fonio (Digitaria exilis), which is adapted to poor and degraded soils and lower rainfall.56

Settling the Floodplains

In the humid conditions of the Early and Middle Holocene, three areas in West Africa were inundated and therefore not accessible to humans: the borders of Mega Lake Chad in northeast Nigeria, the Middle Senegal Valley (MSV), and the Inland Niger Delta (IND) in Mali. In the Chad Basin, the Bama Deltaic Complex had already desiccated after 2000 bce, enabling immigration of the Gajiganna people. But the inner reach of the inundation plain, the Chad Lagoonal Complex, remained under water until 1000 bce. With increasing aridity after 1000 bce and retreating waters of rivers and lakes, the MSV, IND, and Chad Lagoonal Complex became attractive for human seasonal or permanent occupation. They enabled a diversified subsistence with wild and domesticated plants and animal resources, buffering the risks of living in an unpredictable Sahelian environment.

Chad Lagoonal Complex (Northeast Nigeria)

The Chad Lagoonal Complex consists of extensive lacustrine clay deposits with only a few interspersed sand dunes. After 1000 bce, settlement mounds were constructed as an adaptation to seasonal flooding. Still today, the settlement mounds stand out like small islands in an ocean during the rainy season, surrounded by the inundated clay plains, locally called firki. Today the firki are extensively farmed with sorghum in the dry season, while rain-fed pearl millet is grown on the sandy islands. From two deeply stratified settlement mounds, Kursakata and Mege, which were excavated in the 1990s, detailed archaeobotanical data have been reported. Animal exploitation included fishing and cattle keeping.57 Pearl millet is present in both sites from the beginning of the occupation, but foraging of wild resources was also very important, as evidenced by the remains of Paniceae grasses, wild rice, water lily, and tree fruits. In Kursakata, which was settled from 1000 bce, pearl-millet production intensified after 800 bce, when inundation periods became shorter and less intense and larger sandy areas were available for farming, favoring permanent occupation. In 1st-millennium bce Mege, by contrast, the wild-plant components of the diet are still dominant. Mege remained a small island in the vast firki clay plains and was seasonally visited by herding, fishing, and foraging people who exploited the wild wetlands.

Middle Senegal Valley

The Middle Senegal Valley (MSV) is an inland floodplain along the River Senegal extending over 400 km along the border of Senegal and Mauritania. Although situated in a zone with low annual rainfall of 300–400 mm, the MSV is one of the most productive agricultural areas in West Africa. It supports rain-fed farming on the higher grounds, and flood-recession cultivation (décrue), mainly with sorghum and rice, on the seasonally inundated plains. Since 1990, a joint archaeological project of Rice University (Houston, TX) and the IFAN (Institut Fondamentale d’Afrique Noire at Dakar, Senegal) excavated several deep-stratified sites, dated between 800 bce and 1000 ce, among them Walaldé, Cubalel, Sincu Bara, and Siwré, which all were sampled for plant remains. For the earliest period, from 800 to 200 bce, the site of Walaldé was studied in detail, including its archaeobotanical remains.58 The people who arrived at Walaldé around 800 bce were agro-pastoralists who brought cattle, small livestock, and domesticated pearl millet. Fishing was a minor component of their subsistence, according to the small number of fish bones in the site. Large concentrations of potsherds in some levels suggest intensive occupation, but the deposits also show multiple signs of abandonment. In the archaeobotanical assemblage, pearl millet is dominant, accompanied by fruit remains of Ziziphus and other wild trees, caryopses of wild grasses, and seeds of other herbs. The Walaldé people practiced rain-fed cultivation of pearl millet on the levees and surrounding dunes, but they also used the wild resources from a range of environments. Agricultural production was probably not very intensive. Cowpea, sorghum, and African rice are all absent, which indicates that neither mixed farming with cereals and legumes nor décrue agriculture on the clay plains was a focus of subsistence. Iron metallurgy was known, but obviously no iron tools were used for soil tilling; this became only later necessary when the heavy clay soils were taken in cultivation.

The Inland Niger Delta (Mali)

The Inland Niger Delta (IND) is a vast wetland in the middle ranges of the River Niger and its tributary, the Bani, extending over circa 400 kilometers. The delta consists of braided streams, marshes, and lakes, which are extremely productive for fishing and agriculture, but present difficulties for permanent occupation due to seasonal flooding. As in the MSV, occupation along the River Niger started when the inundations became less intensive after 1000 bce. From 800 bce, agro-pastoralists concentrated in favorable places, resulting in large settlement mounds. The Dia site consists of three settlement mounds of over 100 hectares in total, of which two, Shoma and Mara, were excavated. Most of Shoma’s radiocarbon dates fall within the range of 800 to 400 bce, contemporary with Walaldé in the Middle Senegal Valley. But Dia Shoma is different in terms of its archaeobotanical inventory. From the earliest occupation levels, rice caryopses were present, which could be securely placed into the stratigraphy through direct AMS dating. To determine if they represented wild or domesticated African rice, Shawn S. Murray undertook a metrical study (length:width ratios) of the archaeological rice caryopses and compared them with modern domesticated rice, Oryza glaberrima, and the two African wild rice species, O. barthii and O. longistaminata.59 The length:width ratio of the ancient rice caryopses clearly resemble those of the domesticated O. glaberrima. This provides clear evidence for the domesticated status of rice, although the caryopses were considerably shorter than the ones from the modern crop. Caryopses size does not change very much throughout all levels of occupation. It is therefore likely that initial domestication took place elsewhere, and that the crop was imported into the IND already in its domesticated form.

Intensive Farming in the 1st and 2nd Millennia ce

The 1st and early 2nd millennia ce are a time of demographic growth, intensification of agricultural production, establishment of urban centers and long-distance trade networks, and eventually the rise of the West African empires. A typical archaeological feature are groups of large settlement mounds indicating permanent occupation over longer periods. After 800 ce, a general elaboration of trends noted during the preceding period can be observed, partly associated with the introduction of Islam.60 Archaeobotanical information for stratified mound sites is available from Senegal, Mali, Burkina Faso, Nigeria, and Cameroon, and for sites in Benin, archaeobotanical studies are in progress. Some urban centers, such as Jenne-Jeno, Gao, and Essouk-Tadmakka (all in Mali) also yielded archaeobotanical data.61 The archaeobotanical information is very heterogeneous, ranging from detailed quantitative studies to single observations that are hardly comparable except for the mere presence/absence of useful plants in the record.

New Crops and Production Systems

In the 1st millennium ce, new crops appear in the archaeobotanical assemblages of savanna sites. Among them are roselle (Hibiscus sabdariffa), Bambara groundnut (Vigna subterranea), bottle gourd (Lagenaria siceraria), and okra (Abelmoschus esculentus).62 These were cultivated in advanced cropping systems together with cereals and cowpea. Mixed cropping of cereals, legumes, and vegetables enabled higher yields and greater resilience during periods of climatic stress and were sufficiently productive to support large sedentary communities. The emerging urban centers were dependent upon surplus production in rural areas. Increasingly sophisticated iron metallurgy eventually resulted in mass production of iron after 1000 ce, and it is quite probable that a larger part of the available iron was used in the manufacturing of agricultural tools. With expanding trade networks in the first half of the 2nd millennium ce, exotic crops arrived in West Africa, notably wheat, date palm, watermelon, and cotton.

Despite the higher diversity of domesticates, cereals remained the nutritional basis of most populations in the Sahel and the Sudanian zone. Specialized pastoralism based on cattle emerged, and exchange systems between grain-crop farmers and herders developed.63 Wild plants continued to be used, but with a few exceptions, they were of minor significance in comparison to cereals. Pearl millet continued to be the staple in most areas, while rice was dominant in the Inland Niger Delta. In addition, two other cereals entered the scene: sorghum and fonio. For the West African history of these four crops, already cited by Harlan in his seminal paper, archaeobotanical data are available, albeit still in small quantities.64

Ecological and Cultural Diversity of Cereal-Based Agriculture

Pearl Millet (Pennisetum glaucum)

Pearl millet was the major carbohydrate plant cultivated in the 1st and 2nd millennia ce all over West Africa—a continuous tradition that can be traced back to its first appearance in archaeological sites around 2500 bce. Pearl millet, today the staple food for over 100 million people in tropical Africa and India, is perfectly adapted to the highly seasonal climate with low rainfall that is characteristic of the Sahelian zone. Some drought-resistant cultivars thrive under as little as 250 mm of rainfall and can mature within fifty-five to sixty-five days. The crop grows well on light and acid soils with low nutrient status.65 These characteristics make pearl millet the ideal crop for the Sahel with its extensive dune landscapes (Figure 4), and this is reflected in the archaeobotanical assemblages of the 1st-millennium ce sites. In Cubalel and Sincu Bara (MSV), pearl millet is dominant, and no other crops are recorded. It seems that the pattern of plant use in the MSV did not change very much in comparison to the 1st millennium bce, when the area was first settled.66 In Burkina Faso, thousands of pearl-millet grains and chaff remains have been found in Oursi and other 1st-millennium ce mound sites, while wild grasses were insignificant. This indicates that agriculture with pearl millet and other domesticates had become the dominant mode of subsistence.67 Persistent and intensive cultivation resulted in increasing grain size over time, as shown in the detailed quantitative study by Kahlheber.68

Development of Plant Food Production in the West African Savannas: Archaeobotanical PerspectivesClick to view larger

Figure 4. Pearl millet diversity in a village of the Sahel in Burkina Faso.

Photo by Katharina Neumann.

Sorghum (Sorghum bicolor)

Although sorghum is today widely grown in West Africa, its archaeobotanical evidence in the 1st and early 2nd millennia ce is patchy. The domestication area of sorghum is suggested to have been located between Lake Chad, the Nile Valley, and the Red Sea, but because of political instability in the recent decades, very few data are available.69 The oldest fully domesticated sorghum is reported from the Sudanese Nile Valley in the 1st millennium bce. Recent studies on impressions of wild and domesticated chaff remains in pottery 2nd millennium bce suggest cultivation of sorghum from 3500 bce in the Kasala region of eastern Sudan.70 Harlan included the Chad Basin in the potential domestication area, because wild sorghum is very common in the inundated clay plains. In spite of careful archaeobotanical studies in the Chad Basin, no domesticated sorghum older than 400 ce has been recovered. West of the Chad Basin, occasional finds of sorghum demonstrate that the crop was known in the 1st millennium ce, but it never played a significant role.71

Sorghum can be cultivated on a wide range of soils. Some cultivars thrive especially well on heavy clay soils, which are inundated in the rainy season. In these areas, sorghum is grown in the dry season after the waters recede. Dry-season cropping of sorghum enables a second harvest in addition to rainy-season pearl millet. However, tilling the heavy clay soils requires the use of iron hoes. In a seminal paper, John Sutton has highlighted the relationship between the spread of sorghum and iron technology throughout Africa from the early 1st millennium ce onward.72 It is astonishing that sorghum is completely absent, for example, in the MSV in the 1st millennium ce, although clay soils suitable for décrue agriculture are widely available, and it is only sporadically present in other West African sites. By contrast, farther east in the contemporary settlement mounds of Cameroon, sorghum was the most important crop. In the site of Salak, sorghum is attested from 450 ce with two different cultivars, while pearl millet is absent.73 In the Nigerian Chad Basin, dry-season cultivation on clay soils probably started around 700 ce.74

The reasons for this clear difference in crop inventories between the Cameroonian and northeastern Nigerian sites, on the one hand, and the areas farther west, on the other hand, might be cultural or ecological or both. Ecological factors determining the suitability of a region for sorghum cultivation are soils and annual rainfall. In the Nigerian Chad Basin, sandy soils for pearl-millet cultivation are rare, which might have favored the adoption of sorghum as a staple. The 1st-millennium ce Cameroonian sites all lie in the Sudanian zone with 800–900 mm annual rainfall, where sorghum thrives well in the rainy season. By contrast, the excavated contemporary sites farther to the west mostly are situated in the Sahel with lower annual rainfall between 250 and 600 mm, which are ecologically better suitable for pearl millet. With additional sites excavated in the Sudanian zone, the pattern may change in the future. However, even at the 1st-millennium ce site Arondo in the Senegal Valley, which received higher precipitation of 600–800 mm, sorghum was only a minor crop in comparison to pearl millet.75 In Jenne-Jeno (Mali), sorghum was present, but it was much less important than rice.76 The general picture of Iron Age sorghum in West Africa is that it was known, but remained a minor crop, in spite of iron technology and suitable soils for its cultivation. It is also possible that the area around Lake Chad constituted a cultural border between an eastern and western Iron Age spheres with culturally determined preferences for different cereals as staple food. This hypothesis can only be tested when the area between Lake Chad and the Nile Valley, which has been inaccessible during the last decades due to political conflicts, will be open again for systematic research.

African Rice (Oryza glaberrima)

Cultural preferences can also be assumed for the uneven Iron Age distribution of domesticated African rice in the plains of the large West African rivers and Lake Chad. At the sites of Kursakata and Mege in the Chad Lagoonal Complex in northeastern Nigeria, wild rice is regularly present throughout the local sequence from 1000 bce to the 19th century ce, but it is not dominant and is considered to be a component of the larger group of wild grasses collected for food.77 In the 1st-millennium ce sites of Cubalel and Sincu Bara in the MSV, rice is absent, although the local floodplain environments would have been highly suited for its cultivation. By contrast, rice is very common in the IND sites of the 1st millennium ce, notably Jenne-Jeno and Dia. In Dia-Mara, dated around 500–1900 ce, the domesticated status of rice has been confirmed by metrical studies, as was already completed for the 1st-millennium bce deposits of Dia Shoma. The rice remains of Jenne-Jeno, dated between 400 bce and 1400 ce, have never been studied in detail to determine their domestication status.78 However, the short distance of only 70 km between the two site complexes, their similar ecological environment, and the continuous presence of rice throughout the archaeological deposits of Jenne-Jenno make it very likely that domesticated rice was a staple. The different presence/absence and domestication status of rice in the three floodplain areas during the 1st millennium ce suggests that regional cultural traditions were responsible for this diversity.

Fonio (Digitaria exilis, D. iburua)

Digitaria exilis (fonio) is a small-grained grass species widely cultivated in West Africa. The similar D. iburua (black fonio) has a more reduced cultivation area in Benin, Togo, and Nigeria. Digitaria exilis requires low inputs for cultivation, can grow on poor soils without fertilization, and has excellent nutritional properties. A further advantage of fonio is its short growing cycle, so that it can be harvested before the other crops ripen. The wild progenitor of Digitaria exilis could be Digitaria longiflora distributed in the Inland Niger Delta.79 No detailed study on the feature characteristic of wild and domesticated Digitaria species has been conducted, and therefore most fonio identifications are preliminary. The earliest reported fonio finds, dated to the first half of the 1st millennium ce, come from the Nigerian site Janruwa C in the Nok area.80 Later on it is present in the IND.81 In the MSV site Cubalel, dated from 0 to 900 ce, Digitaria, has also been found in the earliest levels, but it is not clear if it was a wild or domestic species.82 The appearance of fonio in the Early Iron Age is interpreted as a sign of diversification in the context of demographic growth, increasing social complexity, and the emergence of urban centers.

The Development of a Cultural Landscape

Today the West African savannas are a cultural landscape deeply influenced by agricultural practices, bush fires, firewood exploitation, and pastoralism. How far back in time can human impacts be traced? Pollen analysis, which is an excellent tool for detecting human-induced changes in Neolithic Europe, unfortunately is not effective in the West African savannas because pollen of cultivated crops and plants from disturbed areas are not distinguishable from those of the species in the open savanna.83 Although bush fires can be traced throughout the Holocene in sediment sequences from lakes, their anthropogenic nature is difficult to demonstrate.84 In the sediment sequence of Ounjougou (Mali), the presence of charcoal in annually laminated sediments is an indicator for regular burning of the savanna from circa 2000 bce onward, which fits well with the first signs of pearl-millet cultivation.85 However, the influence of humans on the landscape during the 1st and 2nd millennia bce was limited.

This changed in the Iron Age and the following period of the West African empires, due to intensive agriculture, increasing sedentism, and growing populations. Charcoal from archaeological sites is an excellent source of information for the reconstruction of agricultural practices and their influence on vegetation. Charcoal assemblages from stratified settlement mounds in Burkina Faso illustrate the intensification of land use between 0 and 1400 ce.86 During this time park savannas with useful trees developed, with the shea-butter tree, Vitellaria paradoxa, as the most important species.87 Increasing percentages of fallow species after 1000 ce indicate that more land came under cultivation and fallow cycles became shorter. The presence of Faidherbia albida, which is a valuable fodder tree, points to the growing importance of livestock keeping.88 This tree species is characteristic of landscapes in the Sahel where pearl millet cultivation and cattle pastoralism coexist.

It has been suggested that intensive iron production in the first half of the 2nd millennium ce had a distinct impact on the natural vegetation. But was it really devastating, was it a process of deforestation denuding the landscape? Detailed studies on the charcoal remains from slag mounds in Mali show a much more differentiated picture.89 Tree species with wood of a high calorific value, such as Prosopis africana, declined, indicating over-exploitation, but at the same time other species with high regeneration capacity gained importance. Although the intensity of land use was quite high in the 1st and early 2nd millennia ce, only the species composition of the woody vegetation changed, but there are no signs of desertification on the same scale as that observed in the 20th century.90 The most distinct anthropogenic changes of the natural vegetation seem to be more recent. Numerous tree species that were still present in the first half of the 2nd millennium ce can no longer be found today, especially in the Sahel zone.91 The disappearance of these species probably took place only in the last few hundred years, in a combination of overexploitation and periods of drought.

Discussion of the Literature

As a starting point, it might be useful to go back to the first comprehensive publication on African plant domestication, resulting from a Wenner-Gren symposium on Burg Wartenstein in 1972.92 Many papers in this book still provide valuable considerations on the history of West African crops. The progress in African archaeobotany since then is illustrated in the proceedings of the International Workshop of African Archaeobotany (IWAA). The first IWAA was organized in Mogilany near Krakov (Poland) in 1994, and since then regularly every three years: 1997 in Leicester (England), 2000 in Frankfurt (Germany), 2003 in Groningen (Netherlands), 2006 in London (England), 2009 in Cairo (Egypt), 2012 in Vienna (Austria), and 2015 in Modena (Italy). The proceedings of the Modena IWAA are in preparation.93 The 2018 IWAA is announced for Gran Canaria (Canary Islands, Spain).

In earlier volumes on the origins of agriculture in a global perspective, Africa has been mostly underrepresented due to the scarcity of archaeobotanical evidence. In more recent overviews, Africa has gained more importance, for example in the linguistically orientated work of Bellwood.94 The compilations of Barker and Barker and Goucher offer the chance to look at African food production in comparison to primary and secondary centers of early food production on other continents.95 For an introduction to the archaeological background, Africa, which is volume 1 of the Encyclopedia of Prehistory, as well as Ann Stahl’s critical introduction to African archaeology (especially the chapter on plant cultivation) and the recently published Oxford Handbook of African Archaeology, are recommended.96

Detailed information about West African crops and wild plants can be found in the sixteen volumes of the series Plant Resources of Tropical Africa and in the six volumes of Burkill’s Useful Plants of West Tropical Africa.97

Further Reading

Barker, Graeme, and Candice Goucher, eds. The Cambridge World History, volume 2: A World with Agriculture, 12000 BCE—500 CE. Cambridge, U.K.: Cambridge University Press, 2015.Find this resource:

Breunig, Peter. “Pathways to Food Production in the Sahel.” In The Oxford Handbook of African Archaeology. Edited by Peter Michell and Paul Lane, 555–570. Oxford: Oxford University Press, 2013.Find this resource:

Fahmy, Ahmed G., Stefanie Kahlheber, and A. Catherine d’Andrea, eds. Windows on the African Past. Frankfurt am Main: Africa Magna Verlag, 2011.Find this resource:

Fuller, Dorian Q., and Elisabeth Hildebrand. “Domesticating Plants in Africa.” In The Oxford Handbook of African Archaeology. Edited by Peter Michell and Paul Lane, 507–525. Oxford: Oxford University Press, 2013.Find this resource:

Harlan, Jack R. “Agricultural Origins: Centers and Noncenters.” Science 174 (1971): 468–474.Find this resource:

Höhn, Alexa. “Archaeobotanical Remains.” In Field Manual for African Archaeology. Edited by Alexandre Livingstone Smith, Els Cornelissen, Olivier P. Gosselain, and Scott McEachern, 206–209. Tervuren: Royal Museum for Central Africa, 2017.Find this resource:

Höhn, Alexa, and Katharina Neumann. “Shifting Cultivation and the Development of a Cultural Landscape during the Iron Age (0–1500 AD) in the Northern Sahel of Burkina Faso, West Africa: Insights from Archaeological Charcoal.” Quaternary International 249 (2012): 72–83.Find this resource:

Kahlheber, Stefanie, and Katharina Neumann. “The Development of Plant Cultivation in Semi-Arid West Africa.” In Rethinking Agriculture. Edited by Tim Denham, José Iriarte, and Luc Vrydaghs, 320–346. Walnut Creek: Left Coast Press, 2007.Find this resource:

Kay, Andrea U., and Jed O. Kaplan. “Human Subsistence and Land Use in Sub-Saharan Africa, 1000 BC to AD 1500: A Review, Quantification, and Classification.” Anthropocene 9 (2015): 14–32.Find this resource:

Linseele, Veerle. “Did Specialized Pastoralism Develop Differently in Africa than in the Near East? An Example from the West African Sahel.” Journal of World Prehistory 23 (2010): 43–77.Find this resource:

Marshall, Fiona, and Elisabeth Hildebrand. “Cattle before Crops: The Beginnings of Food Production in Africa.” Journal of World Prehistory 16.2 (2002): 99–143.Find this resource:

Mitchell, Peter, and Paul Lane, editors. The Oxford Handbook of African Archaeology. Oxford: Oxford University Press, 2013.Find this resource:

Neumann, Katharina. “The Romance of Farming: Plant Cultivation and Domestication in Africa.” In African Archaeology: A Critical Introduction. Edited by Ann Brower Stahl, 249–275. Malden, MA: Blackwell, 2005.Find this resource:

Peregrine, Peter N., and Melvin Ember, eds. Encyclopedia of Prehistory, volume 1: Africa. New York: Kluwer, 2001.Find this resource:

Stahl, Ann Brower, ed. African Archaeology: A Critical Introduction. Malden, MA: Blackwell, 2005.Find this resource:

Stevens, Chris, Sam Nixon, Mary Ann Murray, and Dorian Q. Fuller, eds. Archaeology of African Plant Use. Walnut Creek, CA: Left Coast Press, 2014.Find this resource:

Sutton, John E. G. “Africa, Agriculture and Iron.” In Combining the Past and the Present. Edited by Terje Oestergaard, Nils Anfinset, and Tore Saetersdal, 107–118. BAR International Series 1210. Oxford: Archaeopress, 2004.Find this resource:


(1.) Fiona Marshall and Elisabeth Hildebrand, “Cattle before Crops: The Beginnings of Food Production in Africa,” Journal of World Prehistory 16.2 (2002): 99–143.

(2.) See ORE article by Freda Nkirole M-Mbogori, “Farming and Herding in East Africa: Archaeological and Historical Perspectives.”

(3.) Gordon Childe, Man Makes Himself (Nottingham, Great Britain: Spokesman, 2003).

(4.) Melinda A. “The Neolithic Macro-(R)evolution: Macroevolutionary Theory and the Study of Culture Change,” Journal of Archaeological Research 17 (2009): 1–63.

(5.) John E. G. Sutton, “Africa, Agriculture and Iron,” in Combining the Past and the Present, eds. Terje Oestergaard, Nils Anfinset, and Tore Saetersdal, BAR International Series 1210 (Oxford: Archaeopress, 2004), 107–118.

(6.) Christopher de Corse and Sam Spiers, “West African Iron Age,” in Encyclopedia of Prehistory, vol. 1: Africa, eds. Peter N. Peregrine and Melvin Ember (New York: Kluwer, 2001), 313–318.

(7.) Nikolaj I. Vavilov, Origin and Geography of Cultivated Plants (Cambridge, U.K.: Cambridge University Press, 1992).

(8.) Jack R. Harlan, “Agricultural Origins: Centers and Noncenters,” Science 174 (1971): 468–474.

(9.) Susan K. McIntosh, “West African Late Stone Age,” in Encyclopedia of Prehistory, vol. 1: Africa, eds. Peter N. Peregrine and Melvin Ember (New York: Kluwer, 2001), 319–322; and Joanna Casey, “The Stone to Metal Age in West Africa,” in Oxford Handbook of African Archaeology, eds. Peter Mitchell and Paul Lane (Oxford: Oxford University Press, 2013), 603–614.

(10.) Sarah E. Oas, A. Catherina d’Andrea, and Derek J. Watson, “10,000 Year History of Plant Use at Bosumpra Cave, Ghana,” Vegetation History and Archaeobotany 24 (2015): 635–653; Philippe Lavachery, “The Holocene Archaeological Sequence of Shum Laka Rockshelter (Grasslands, Western Cameroon),” African Archaeological Review 18 (2001): 213–247; Thomas Frank et al., “The Chaine de Gobnangou, SE Burkina Faso: Archaeological, Archaeobotanical, Archaezoological and Geomorphological Studies,” Beiträge zur Allgemeinen und Vergleichenden Archäologie 21 (2001): 127–190; and Stephen A. Dueppen and Daphne Gallagher, “Adopting Agriculture in the West African Savanna: Exploring Socio-economic Choices in First Millennium CE Southeastern Burkina Faso,” Journal of Anthropological Archaeology 32 (2013): 433–448.

(11.) Caryopsis (pl. caryopses) is the scientific term for the grains of grasses (Poaceae). The caryopsis is not a seed, as usually assumed, but a special type of fruit with the seed coat closely adhering to the pericarp.

(12.) Anne-Marie Lézine et al., “Sahara and Sahel Vulnerability to Climate Changes, Lessons from Holocene Hydrological Data,” Quaternary Science Reviews 30 (2011): 3001–3012; and Peter deMenocal et al., “Abrupt Onset and Termination of the African Humid Period. Rapid Climate Responses to Gradual Insolation Forcing,” Quaternary Science Reviews 19 (2000): 347–361.

(13.) Christelle Hély, Anne-Marie Lézine, and APD Contributors, “Holocene Changes in African Vegetation: Tradeoff Between Climate and Water Availability,” Climate of the Past 10 (2014): 681–686.

(14.) Eric Huysecom et al., “The Emergence of Pottery in Africa during the Tenth Millennium Cal BC: New Evidence from Ounjougou (Mali),” Antiquity 83 (2009): 905–917; and Friederike Jesse, “Early Pottery in Northern Africa—An Overview,” Journal of African Archaeology 8.2 (2010): 219–238.

(15.) Rudolph Kuper and Stefan Kröpelin, “Climate-Controlled Holocene Occupation in the Sahara: Motor of Africa’s Evolution,” Science 313 (2006): 803–807.

(16.) Krystyna Wasylikowa et al., “Site E–6: Vegetation and Subsistence of the Early Neolithic at Nabta Playa, Egypt, Reconstructed from Charred Plant Remains,” in Holocene Settlement of the Eygyptian Sahara, vol. 1: The Archaeology of Nabta Playa, eds. Fred Wendorf, Romuald Schild, and Associates (New York: Kluwer, 2001), 544–606.

(17.) Linda Olmi et al., “Morphological and Genetic Analyses of Early and Mid Holocene Wild Cereals from the Takarkori Rockshelter (Central Sahara, Libya),” in Windows on the African Past: Current Approaches to Archaeobotany, eds. Ahmed G. Fahmy, Stefanie Kahlheber, and A. Catherine d’Andrea, Reports in African Archaeology 3 (Frankfurt: Africa Magna Verlag, 2011), 175–187; and Julie Dunne et al., “Earliest Direct Evidence of Plant Processing in Prehistoric Saharan pottery,” Nature Plants 3 (2016): 161–194.

(18.) Anna Maria Mercuri, “Human Influence, Plant Landscape Evolution and Climate Inferences from the Archaeobotanical Records of the Wadi Teshuinat Area (Libyan Sahara),” Journal of Arid Environments 72 (2008): 1950–1967.

(19.) Savino di Lernia, Isabella Massamba N’siala, and Anna M. Mercuri, “Saharan Prehistoric Basketry: Archaeological and Archaeobotanical Analysis of the Early-Middle Holocene Assemblage from Takarkori (Acacus Mts., SW Libya),” Journal of Archaeological Science 39 (2012): 1837–1853.

(20.) Diane Gifford-Gonzalez and Olivier Hanotte, “Domesticating Animals in Africa: Implications of Genetic and Archaeological Findings,” Journal of World Prehistory 24 (2011): 1–23.

(21.) Julie Dunne et al., “First Dairying in Green Saharan Africa in the fifth millennium BC,” Nature 486 (2012): 390–394.

(22.) Sylvain Ozainne et al., “Developing a Chronology Integrating Archaeological and Environmental Data from Different Contexts: The Late Holocene Sequence of Ounjougou (Mali),” Radiocarbon 51.2 (2009): 457–470.

(23.) Katie Manning et al., “4500-Year Old Domesticated Pearl Millet (Pennisetum glaucum) from the Tilemsi Valley, Mali: New Insights into an Alternative Cereal Domestication Pathway,” Journal of Archaeological Science 38 (2011): 312–322.

(24.) Ibrahima Oumar et al., “Phylogeny and Origin of Pearl millet (Pennisetum glaucum [L.] R. Br) as Revealed by Microsatellite Loci,” Theoretical and Applied Genetics 117 (2008): 489–497.

(25.) Bruce D. Smith, “Low-Level Food Production,” Journal of Archaeological Research 9 (2001): 1–43.

(26.) Peter Breunig, “Pathways to Food Production in the Sahel,” in The Oxford Handbook of African Archaeology, eds. Peter Michell and Paul Lane (Oxford: Oxford University Press, 2013), 555–570.

(27.) Veerle Linseele, Archaeofaunal Remains from the Past 4000 Years in Sahelian West Africa: Domestic Livestock, Subsistence Strategies and Environmental Changes (Oxford: Archaeopress, 2007).

(28.) Marlies Klee, Barbara Zach, and Hans-Peter Stika, “Four Thousand Years of Plant Exploitation in the Lake Chad Basin (Nigeria), part III: Plant Impression in Potsherd from the Final Stone Age Gajiganna Culture,” Vegetation History and Archaeobotany 13 (2004): 131–142.

(29.) Derek J. Watson, “Within Savanna and Forest: A Review of the Late Stone Age Kintampo Tradition, Ghana,” Azania 45 (2010): 141–174.

(30.) A. Catherine d’Andrea and Joanna Casey, “Pearl Millet and Kintampo Subsistence,” African Archaeological Review 19.3 (2002): 147–173; A. Catherine D’Andrea, Amanda Logan, and Derek J. Watson, “Oil Palm and Prehistoric Subsistence in Tropical West Africa,” Journal of African Archaeology 4.2 (2006): 195–222; A. Catherine D’Andrea et al., “Early Domesticated Cowpea (Vigna unguiculata) from Central Ghana,” Antiquity 81 (2007): 686–698; and Oas et al., “Bosumpra.”

(31.) A. Catherine d’Andrea, Amanda L. Logan, and Derek J. Watson, “Oil Palm,” 195; and Achille Gautier and Wim Van Neer, “The Continuous Exploitation of Wild Animal Resources in the Archaeozoological Record of Ghana,” Journal of African Archaeology 3.2 (2005): 195–212. See ORE article by A. Catherine d’Andrea and Richard T. Chiah, “Food Production in the Forest Zone of West Africa. Archaeological and Historical Perspectives.”

(32.) Derek J. Watson, “Under the Rocks. Reconsidering the Origin of the Kintampo Tradition and the Development of Food Production in the Savanna-Forest/Forest of West Africa,” Journal of African Archaeology 3.1 (2005): 3–55.

(33.) Kevin MacDonald, “The Tichitt Tradition in the West African Sahel,” in The Cambridge World History, vol. 2: A World with Agriculture, 12,000 BCE-500 CE, eds. Graeme Barker and Candice Goucher (Cambridge, U.K.: Cambridge University Press, 2015), 499–513.

(34.) Dorian Fuller, Kevin MacDonald, and Robert Vernet, “Early Domesticated Pearl Millet in Dhar Nema (Mauritania): Evidence of Crop Processing Waste as Ceramic Temper,” in Fields of Change: Progress in African Archaeobotany, ed. René Cappers (Groningen, The Netherlands: Barkhuis, 2007), 71–76.

(35.) Kevin MacDonald, “Betwixt Tichitt and the IND: the pottery of the Faita Facies, Tichitt Tradition,” Azania 46.1 (2011): 62.

(36.) Kevin MacDonald, “Tichitt-Walata and the Middle Niger: Evidence for Cultural Contact in the Second Millennium BC,” in Aspects of African Archaeology, Papers from the 10th Congress of the Pan African Association for Prehistory and Related Studies, eds. Gilbert Pwiti and Robert Soper (Harare: University of Zimbabwe Publications, 1996), 429–440.

(37.) Kevin MacDonald, Louis Champion, and Katie Manning, “Windé Koroji Ouest (Mali, Third and Second Millennia BCE): The Environmental and Subsistence Evidence,” in Winds of Change, eds. Nicole Rupp et al. (Bonn, Germany: Habelt, 2017), 165–180.

(38.) Barbara Eichhorn and Katharina Neumann, “Holocene Vegetation Change and Land Use at Ounjougou, Mali,” in Archaeology of African Plant Use, eds. Chris J. Stevens et al. (Walnut Creek, CA: Left Coast Press, 2014), 83–96; and Aline Garnier et al., “Phytolith Taphonomy in the Middle- to Late-Holocene Fluvial Sediments of Ounjougou (Mali, West Africa),” The Holocene 23.3 (2012): 416–431.

(39.) Stefanie Kahlheber and Katharina Neumann, “The Development of Plant Cultivation in Semi-Arid West Africa,” in Rethinking Agriculture, eds. Tim Denham, José Iriarte, and Luc Vrydaghs (Walnut Creek, CA: Left Coast Press, 2007), 320–346.

(40.) Stefanie Kahlheber, “Perlhirse und Baobab - Archäobotanische Untersuchungen im Norden Burkina Fasos” (PhD diss., University of Frankfurt, 2004).

(41.) Breunig, “Pathways,” 562.

(42.) See ORE article by Louise Iles, “African Iron Production and Iron-Working Technologies.”

(43.) David Killick, “What Do We Know about African Iron Working?” Journal of African Archaeology 2.1 (2004): 97–112.

(44.) Breunig, “Pathways,” 562; Gabriele Franke, “A Chronology of the Central Nigerian Nok Culture: 1500 BC to the Beginning of the Common Era,” Journal of African Archaeology 14.3 (2016): 257–290; and Laurent Lespez et al., “High-Resolution Fluvial Records of Holocene Environmental Changes in the Sahel: the Yamé River at Ounjougou (Mali, West Africa),” Quaternary Science Reviews 30 (2011): 737–756.

(45.) Katharina Neumann et al., “First Farmers in the Central African Rainforest: A View from Southern Cameroon,” Quaternary International 249 (2012): 53–62.

(46.) Jean Maley and Robert Vernet, “Populations and Climatic Evolution in North Tropical Africa from the End of the Neolithic to the Dawn of the Modern Era,” African Archaeological Review 32 (2015): 198.

(47.) Breunig, “Pathways,” 563.

(48.) Carlos Magnavita et al., “Zilum: A Mid-First Millennium BC Fortified Settlement Near Lake Chad,” Journal of African Archaeology 4.1 (2006): 153–170.

(49.) Maley and Vernet, “Climatic Evolution,” 189.

(50.) Peter Breunig and Nicole Rupp, “An Outline of Recent Studies on the Nigerian Nok Culture,” Journal of African Archaeology 14.3 (2016): 237–256.

(51.) Sylvain Ozainne et al., “A Question of Timing: Spatio-temporal Structure and Mechanisms of Early Agriculture Expansion in West Africa,” Journal of Archaeological Science 50 (2014): 359–368.

(52.) Julie Dunne, e-mail message to author, September 08, 2016.

(53.) Franke, “Chronology Nok,” 273.

(54.) Alexa Höhn and Katharina Neumann, “The Palaeovegetation of Janruwa (Nigeria) and Its Implications for the Decline of the Nok Culture,” Journal of African Archaeology 17.3 (2016): 331–353.

(55.) d’Andrea, Logan and Watson, “Oil Palm,” 214.

(56.) Höhn and Neumann, “Janruwa,” 350.

(57.) Marlies Klee, Barbara Zach, and Katharina Neumann, “Four Thousand Years of Plant Exploitation in the Chad Basin of Northeast Nigeria I: The Archaebotany of Kursakata,” Vegetation History and Archaeobotany 9 (2000): 223–237; Barbara Zach and Marlies Klee, “Four Thousand Years of Plant Exploitation in the Chad Basin of Northeast Nigeria II: Discussion on the Morphology of Caryopses of Domesticated Pennisetum and Complete Catalogue of Seeds and Fruits,” Vegetation History and Archaeobotany 12 (2003): 187–204; Gerlinde Bigga and Stefanie Kahlheber, “From Gathering to Agricultural Intensification: Archaeobotanical Remains from Mege, Chad Basin, NE Nigeria,” in Windows on the African Past. Current Approaches to Archaeobotany, eds. Ahmed G. Fahmy, Stefanie Kahlheber, and A. Catherine d’Andrea, Reports in African Archaeology 3 (Frankfurt: Africa Magna Verlag, 2011), 19–66; and Linseele, “Archaeofaunal Remains,” 159.

(58.) Alioune Déme and Susan K. McIntosh, “Excavations at Walaldé: New Light on the Settlement of the Middle Senegal Valley by Iron-Using Peoples,” Journal of African Archaeology 4.2 (2006): 317–348; and Shawn S. Murray and Alioune Déme, “Early Agro-Pastoralism in the Middle Senegal Valley: The Botanical Remains from Walaldé,” in Archaeology of African Plant Use, eds. Chris J. Stevens et al. (Walnut Creek: Left Coast Press, 2014), 97–102.

(59.) Shawn S. Murray, “Identifying African Rice Domestication in the Middle Niger Delta (Mali),” in Fields of Change: Progress in African Archaeobotany, ed. René Cappers (Groningen, The Netherlands: Barkhuis, 2007), 53–62.

(60.) Christopher de Corse and Sam Spiers, “West African Regional Development,” in Encyclopedia of Prehistory, vol. 1: Africa, eds. Peter N. Peregrine and Melvin Ember (New York: Kluwer, 2001), 339–345.

(61.) Susan K. McIntosh, ed., Excavations at Jenné-Jeno, Hambarketolo, and Kaniana (Inland Niger Delta, Mali), the 1981 Season (Berkeley: University of California Press, 1995), 348–351; Dorian Q. Fuller, “The Botanical Remains,” in Urbanism, Archaeology and Trade, ed. Timothy Insoll (Oxford: Archaeopress, 2000), 28–35; and Sam Nixon, Mary A. Murray, and Dorian Q. Fuller, “Plant Use at an Islamic Merchant Town in the West African Sahel: the Archaeobotany of Essouk-Tadmakka (Mali),” Vegetation History and Archaeobotany 20 (2011): 223–239.

(62.) Kahlheber and Neumann, “Plant Cultivation,” 331.

(63.) Veerle Linseele, “Did Specialized Pastoralism Develop Differently in Africa than in the Near East? An Example from the West African Sahel,” Journal of World Prehistory 23 (2010): 43–77.

(64.) Harlan, “Centers,” 471.

(65.) D. J. Andrews and K. A. Kumar, “Pennisetum glaucum (L.) R. Br.,” in Plant Resources of Tropical Africa, vol. 1: Cereals and Pulses, eds. M. Brink and G. Belay (Wageningen: PROTA Foundation/Backhuys, 2006), 128–133.

(66.) Shawn S. Murray, “A Report on the Charred Botanical Remains from Sincu Bara, A Mid-first Millennium AD Middle Senegal Valley Site,” Nyame Akuma 69 (2008): 56–63; Mary A. Murray, Dorian Q. Fuller, and Cecilia Capezza, “Crop Production on the Senegal River in the Early First Millennium AD: Preliminary Archaeobotanical results from Cubalel,” in Fields of Change: Progress in African Archaeobotany, ed. René Cappers (Groningen, The Netherlands: Barkhuis, 2007), 63–70; and Daphne Gallagher and Shawn S. Murray, “Palaeoethnobotanical Analysis: Seeds and Fruits from Cubalel, Unit C-3A,” in The Search for Takrur: Archaeological Excavations and Reconnaisssance along the Middle Senegal Valley, eds. Roderick McIntosh, Susan K. McIntosh, and Hamady Bocoum (New Haven, CT: Yale University Press, 2013), 299–310.

(67.) Kahlheber and Neumann, “Plant Cultivation,” 331.

(68.) Kahlheber, “Perlhirse und Baobab,” 221.

(69.) Harlan, “Centers,” 471.

(70.) Dorian Q. Fuller, “Early Kushitic Agriculture. Archaeobotanical Evidence from Kawa,” Sudan and Nubia 8 (2004): 70–74; Alemseged Beldados, “Sorghum Exploitation at Kassala and Its Environs, North Eastern Sudan in the Second and First Millennia BC,” Nyame Akuma 75 (2011): 33–39; and Frank Wichell, Chris J. Stevens, Charlene Murphy, Louis Champion, and Dorian Q. Fuller, “Evidence for Sorghum Domestication in Fourth Millennium BC Eastern Sudan,” Current Anthropology 58.5 (2017): 673–683.

(71.) Kahlheber and Neumann, “Plant Cultivation,” 332; and Nixon, Murray, and Fuller, “Essouk-Tadmakka,” 237.

(72.) Sutton, “Africa, Agriculture and Iron,” 111.

(73.) Thierry Otto and Michèle Delneuf, “Evolution des ressources alimentaires et de paysages au nord du Cameroun: Apport de l’archéologie,” in Plantes et paysages d’Afrique, ed. Monique Chastenet (Paris: Karthala, 1998), 491–514.

(74.) Bigga and Kahlheber, “Mege,” 41.

(75.) Daphne Gallagher, “Analysis of Seeds and Fruits from the Sites of Arondo and Ft. Senebudu, Senegal” (unpublished senior honors thesis, Rice University, Houston, 1999).

(76.) McIntosh, “Jenné-Jeno,” 349–350.

(77.) Klee et al., “Kursakata,” 232; and Bigga and Kahlheber, “Mege,” 40.

(78.) S. S. Murray, “African Rice,” 60; and McIntosh, “Jenné-Jeno,” 349–350.

(79.) S. R. Vodouhè and E. G. Achigan Dako, “Digitaria exilis (Kippist) Stapf,” in Plant Resources of Tropical Africa, vol. 1: Cereals and Pulses, eds. M. Brink and G. Belay (Wageningen, the Netherlands: PROTA Foundation/Backhuys, 2006), 54–57.

(80.) Höhn and Neumann, “Janruwa,” 350.

(81.) Kahlheber and Neumann, “Plant Cultivation,” 327.

(82.) M. A. Murray, Fuller, and Capezza, “Crop Production,” 67; and Gallagher and Murray, “Cubalel,” 305.

(83.) Ulrich Salzmann and Martyn Waller, “The Holocene Vegetational History of the Nigerian Sahel Based on Multiple Pollen Profiles,” Review of Palaeobotany and Palynology 100 (1998): 39–72.

(84.) Ulrich Salzmann, Philipp Hoelzmann, and Irena Morczinek, “Late Quaternary Climate and Vegetation of the Sudanian Zone of NE-Nigeria Deduced from Pollen, Diatoms and Sedimentary Geochemistry,” Quaternary Research 58 (2002): 73–83.

(85.) Yann Le Drézen et al. “Hydrosedimentary Records and Holocene Environmental Dynamics in the Yamé Valley (Mali, Sudano-Sahelian West Africa),” Comptes rendus géoscience 342.3 (2010): 244–252.

(86.) Alexa Höhn and Katharina Neumann, “Shifting Cultivation and the Development of a Cultural Landscape during the Iron Age (0–1500 ad) in the Northern Sahel of Burkina Faso, West Africa: Insights from Archaeological Charcoal,” Quaternary International 249 (2012): 72–83.

(87.) Daphne E. Gallagher, Stephen A. Dueppen, and Rory Walsh, “The Archaeology of Shea Butter (Vitellaria paradoxa) in Burkina Faso, West Africa,” Journal of Ethnobiology 36.1 (2016): 150–171.

(88.) Höhn and Neumann, “Shifting Cultivation,” 79.

(89.) Barbara Eichhorn et al., “Iron Metallurgy in the Dogon Country, (Mali, West Africa)—‘Deforestation’ or Sustainable Use?” in Proceedings of the Fourth International Meeting of Anthracology, ed. Freddy Damblon (Oxford: Archaeopress, 2013), 57–70.

(90.) Andrea U. Kay and Jed O. Kaplan, “Human Subsistence and Land Use in Sub-Saharan Africa, 1000 BC to AD 1500: A Review, Quantification, and Classification,” Anthropocene 9 (2015): 14–32.

(91.) Alexa Höhn, “Where Did All the Trees Go?” in Fields of Change: Progress in African Archaeobotany, ed. René Cappers (Groningen, The Netherlands: Barkhuis, 2007), 35–42.

(92.) Jack R. Harlan, Jan M. J. de Wet and Ann B. L. Stemler, ed., Origins of African Plant Domestication (The Hague, The Netherlands: Mouton, 1976).

(93.) Leon Stuchlik and Krystyna Wasylikowa, eds., Acta Palaeobotanica 35(1) (Kraków, Poland: Polish Academy of Sciences, 1995); Marike Van der Veen, ed., The Exploitation of Plant Resources in Ancient Africa (New York: Kluwer/Plenum, 1999); Katharina Neumann, Ann Butler, and Stefanie Kahlheber, eds., Food, Fuel and Fields (Cologne: Heinrich-Barth Institut, 2003); René Cappers, ed., Fields of Change: Progress in African Archaeobotany (Groningen, The Netherlands: Barkhuis, 2007); Chris Stevens et al., eds., Archaeology of African Plant Use (Walnut Creek, CA: Left Coast Press, 2014); Ahmed G. Fahmy, Stefanie Kahlheber, and A. Catherine d’Andrea, eds., Windows on the African Past (Frankfurt: Africa Magna Verlag, 2011); and Ursula Thanheiser, ed., News from the Past (Groningen, The Netherlands: Barkhuys, 2016).

(94.) Peter Bellwood, First Farmers: The Origins of Agricultural Societies (Malden, MA: Blackwell, 2007).

(95.) Graeme Barker, The Agricultural Revolution in Prehistory: Why Did Foragers Become Farmers? (Oxford: Oxford University Press, 2006); and Graeme Barker and Candice Goucher, eds., The Cambridge World History, vol. 2: A World with Agriculture, 12000 BCE—500 CE (Cambridge, U.K.: Cambridge University Press, 2015).

(96.) Peter N. Peregrine and Melvin Ember, eds., Encyclopedia of Prehistory, vol. 1: Africa (New York: Kluwer/Plenum, 2001); Ann Brower Stahl, ed., African Archaeology: A Critical Introduction (Malden, MA: Blackwell, 2005); Katharina Neumann, The Romance of Farming: Plant Cultivation and Dometication in Africa, in African Archaeology: A Critical Introduction, ed. Ann Brower Stahl (Malden, MA: Blackwell, 2005), 249–275; and Peter Mitchell and Paul Lane, eds., The Oxford Handbook of African Archaeology (Oxford: Oxford University Press, 2013).

(97.) Plant Resources of Tropical Africa, 16 vols. (Wageningen, The Netherlands: PROTA Foundation/Backhuys, 2006–2012); and H. M. Burkill, ed., The Useful Plants of West Tropical Africa, vols. 1–6, 2d ed. (Kew, Great Britain: Royal Botanic Gardens, 1985–2004).