Beyond intensification: landscapes and livelihoods in Mali’s Guinea Savannah

For more than a decade, sub-Saharan Africa has been the focus of calls for a new Green Revolution. Like its predecessor, the African Green Revolution aims to increase the productivity of smallholder farmers, improving their own food security and income as well as that of the continent as a whole. This is to be done with minimum environmental damage, through “sustainable intensification.” While sustainable intensification has shown potential in places where high population density precludes cropland expansion, evidence of its effectiveness in land-abundant, labor-limited areas is limited. One such land-abundant, labor-limited area is the Guinea Savannah region of West Africa, which the World Bank called a “Sleeping Giant” where agricultural development could drive economic growth both locally and at the national level. Within the Guinea Savannah region, we use southern Mali’s Bougouni district as a case study to explore potential futures for smallholder agriculture in the area. We explored the history of the area’s agriculture using a panel data set for three villages, as well as remote sensing analysis and census data. Over the period of the panel data (1994-2012), agricultural change was minor. Cultivated area per household was highly correlated with household size and the number of draft animals a household owned. This relationship remained constant over the full period, suggesting little change in labor productivity. Yields of major crops remained stagnant, even as fertilizer input increased. Cropland expansion occurred in parallel with population growth, but up to the present, over half the arable land in the study villages was not cultivated. Because uncultivated rangeland made up such a large percentage of the land, we characterized the productivity, management and use of these rangelands (Chapter 3). In two villages, we assessed biomass quantity and species composition at 2-month intervals, tracked a sample of village herds, and used remote sensing combined with regression analysis to map the productivity of herbaceous biomass in a woody savannah landscape. We found that rangelands produced a seasonal peak of 2-2.5 t/ha of herbaceous biomass, from a diverse mix of annual and perennial species, notably Andropogon gayanus and A. pseudapricus. Herds covered distances of 10-18 km each day, with distance and location variable based on the season. During most of the year, the forage supply far exceeded the demands of grazing herds, but in the late dry season forage becomes scarce and herders supplement grazing with cut tree fodders, or send herds on transhumance to the south. While rangelands are exploited for a variety of uses, local management has thus far maintained high levels of productivity and biodiversity. In order to evaluate the potential of sustainable intensification to meet its goals of reducing poverty and improving food security, we explored the solution space of possible gains from intensification for farm households in three villages. With yields equivalent to the best farmers yields in the area, over 90% of households can achieve food self-sufficiency, and most can raise income levels above the threshold for extreme poverty. Reaching attainable yield levels, equal to those obtained in on-station trials, improved the picture further. However, agriculture must compete with other income generating options, which can be considerably more profitable. The average annual income for a gold miner in the area was $1225. Even at attainable yield levels, only 25% of households in the study villages could earn higher per capita incomes from their current cropped area. If we consider options beyond intensification, we find that expanding cultivated land area can increase this fraction to 59%, while also timing crop sales to correspond to peak price points allows over 90% of households to earn more from agriculture than the average gold mining income. Dairy production has potential to provide high income to a few households with large herds, but would require large investments in infrastructure and improved market access. Production of small ruminants for meat, particularly rams sold at peak holiday prices, could raise incomes for a larger number of households, because initial costs are modest. While small ruminant production does not require the complex infrastructure of dairy marketing, current production potential is limited by a lack of veterinary services and limited market access. Because of the limited gains from intensification, new options are needed for land-abundant, labor-constrained farming systems like those in southern Mali. We worked with local farmers in two villages to develop and analyze future scenarios. Scenarios were based on key drivers farmers identified: tractor availability and increased cashew production. These were explored further by developing a game in which the board represented the village territory, and players with varying initial assets could make decisions about planting trees, purchasing or renting tractors, and clearing new land. The agent-based model Mali-sene (Multi-agent land-use and intensification socio-ecological niche exploration) simulated behavior seen in the land use game, and was used to explore a wider range of scenarios, with different rates of tree planting as well as access to tractor rental and purchase. Scenarios with extensive tree planting resulted in high rates of land conversion, with the majority of cultivated land in tree plantations, and resulted in incomes of up to $1600 per capita. Scenarios where tractor rental was available but tree planting was minimal resulted in somewhat lower rates of land conversion, but converted land was planted to annual staple crops, while tractor rental without the introduction of cashew increased annual incomes to $400 per capita, still twice the initial value. It seems clear that cropland expansion is highly likely to occur in this area, and preventing expansion comes at a real cost to local farmers. Tractor availability and cashew planting both led to land conversion, but the environmental impacts of cashew, as a perennial tree crop, are likely to be lower than the impacts of annual staples. A holistic evaluation of sustainability that considers farmer livelihoods might therefore conclude that expansion is as sustainable as intensification. The process of developing agricultural technology innovations in sub-Saharan Africa is generally led by scientists, but has many commonalities with engineering and product design methodologies. Increased attention to the steps in this process, from problem definition to developing design specifications to testing possible solutions, could help research for development projects develop more relevant technical solutions for farmers