Will Trees Alleviate Hunger in Africa?

The challenges in food production and crop yields are strikingly different across the globe. Here in Africa, the continent with the highest proportion of people in poverty, yields of many staple food crops are far below the potential of even currently available crop germplasm. For example, good stands of maize grow throughout eastern and southern Africa, although in some locations, their occurrence is confined to a few well-managed farms.

The major biophysical challenges for boosting yields are improving soil and water management. Recent developments in what are called fertilizer tree systems are showing robust promise on the first of these challenges. Although there is still much to be done and although improvements in water management will be essential, the progress is raising hopes that expanding food production in Africa is within reach.

Although soil management in many parts of the world consists of soil conservation and the use of mineral fertilizer, these have historically not been sufficiently applied on African soils. Soil health restoration in Africa once relied significantly on the practice of fallowing (periodically leaving fields unplanted), but because of ever-increasing rural populations, the practice has all but disappeared. Farmers have responded by increasing the intensity of soil-conservation and nutrient-management practices—for example, the application of animal manure and a return of crop residues to the soil. But these practices have been inadequate in general, and soil quality has degraded. Constraints on soil health differ across the landscape, even at small scales, so the identification of best practices is context specific.

Efforts to promote an increased use of fertilizer are under way, and this is widely advocated. But after 50 years of efforts to promote fertilizer, it is clear that smallholder farmers still use very little of it—only 9 kilograms per hectare, about the same amount as in the 1970s. Its high cost is the major deterrent. Yet recent studies suggest that the very low use of fertilizer on staple foods in Kenya and Uganda is actually close to the economically optimal amount. The implication is that availing more credit may have little effect and that the price ratios of output (i.e., crops) to fertilizer need to be fundamentally increased in order to encourage higher usage of fertilizer. Accordingly, some countries have recently created or expanded fertilizer-subsidy programs.

This means that three strategies for advancing soil-fertility management are now being pursued in Africa: (1) Fertilizer subsidies should be expanded, because African farmers are disfavored in terms of the relative prices of fertilizer and crops. (2) Fertilizer subsidies (and therefore, the continent's dependency on imported fertilizers) may be scrapped, and instead, the focus could be shifted to increasing the use of locally sourced nutrients, such as animal and green manures. (3) More attention may be focused on integrated soil-fertility management systems. The third approach acknowledges the constraints that farmers must contend with in acquiring nutrients from all sources while recognizing that there may be local synergies that affect the optimum strategy—for example, where the use of animal manure may increase the efficiency of fertilizer use and may thus make it more attractive even at unsubsidized costs.

Agroforestry systems—in particular, those aimed at addressing soil-health problems, such as fertilizer tree systems—have been tested as components of integrated soil-fertility management systems. They have also been investigated as stand-alone options.

Some important examples are well known and have been practiced traditionally. These include the parkland system of the Sahel, in which mature trees are integrated into fields, including millet and sorghum fields. Other systems have been developed more recently. These include rotational or relay fallows (a dense planting of trees in rotation with crops), intercropping systems (including natural regeneration, such as in the Sahel, and more intensive and dense planting of trees with crops), and biomass-transfer systems (where trees are grown away from crop fields and biomass is transferred in). Other agroforestry practices, such as windbreaks and contour plantings that prevent soil erosion, are also important for soil health, but the latest developments in fertilizer tree systems justify special attention.

How do fertilizer trees benefit crop yields? One way is through the addition of nutrients to the soil. The most important fertilizer tree species are nitrogen fixing: They accumulate nitrogen in their roots and leaves. Agroforestry systems with high densities of trees can generate more than 100 kilograms of fixed nitrogen per hectare. Trees can also recycle nutrients from lower soil depths—from deep roots to the leaves—and can make them available to crops through leaf litter fall. Modest amounts of phosphorus, potassium, and other nutrients can be recycled this way, but agroforestry alone will not be able to significantly replenish soils that are depleted of these nutrients.

A second way fertilizer trees can benefit crop yields is through physical and biological effects. Recent studies have shown that under agroforestry systems, the amount of carbon in the soil increases and higher densities of key soil macrofauna and microfauna are found (Barrios et al. 2012). In addition, other research has shown that the stabilizing effect of trees' roots and the cover provided by their canopies reduce soil erosion, break hardpans, and improve water infiltration and moisture levels. All of these can help improve overall soil health and can also enhance the efficiency of mineral fertilizer and water use.

A third form of benefits comes from microclimate effects. The shading effects of fertilizer trees can be beneficial during the dry season because they reduce the number of weeds and soil temperatures in advance of planting. Shade can also reduce evapotranspiration during the rainy season.

Through these effects, fertilizer trees can significantly increase crop yields. Such systems have been tested fairly widely in Africa, on research plots and farmers' fields, alone and in combination with other soil-fertility practices. They have been tested in all regions, although largely with maize as the crop. In a recent publication, Sileshi and colleagues (2008) synthesized a large set of studies on the effects of fertilizer tree systems on maize in Africa. Sileshi and colleagues (2008) concluded that in about two-thirds of the cases, the systems alone have doubled maize yields. When combined with microdosing of mineral fertilizer, the yield results are almost always superior. There is some understanding of the conditions under which they do not perform as well, but more research is required to help sharpen our understanding of which options are likely to work in different environments.

Other studies have shown that agroforestry systems are profitable in terms of the returns to land and labor relative to the current low-input systems and natural fallow systems. They also compete relatively well with mineral fertilizer options where phosphorus is not a limiting factor and fertilizer is not subsidized.

So, are the systems being adopted by farmers? The evidence is mixed, although for some, it is still early. It is first important to point out some of the resource requirements of the systems. The rotational agroforestry systems require land to be taken out of production and are therefore more suitable to farms where reasonably large areas can be managed. Conversely, households with labor shortages may find them attractive. Intercrop systems that enable continuous planting of the crops are more suited to farms with scarce land resources. All agroforestry systems can be established with minimal cash outlays, requiring only the purchase of germplasm.

Traditional parkland systems are widespread and have been in use for centuries. There is also evidence to suggest that farmers have been recently motivated to rejuvenate them in Niger (more than 5 million hectares) and elsewhere in the Sahel. The Conservation Farming Unit in Zambia has widely extended a key parkland species (Faidherbia albida) in its recent conservation-farming dissemination program. For the more recently developed systems, dissemination efforts have been ongoing only since the late 1990s. The evidence shows some success, as well as the need for improved targeting. Improved fallow systems were extended into a high-population area of Kenya in the late 1990s. Although farmers were able to implement them, and studies showed that the majority performed well, many farmers abandoned the practice. On the other hand, a recent study in eastern Zambia, where there are larger farms, showed that the majority of farmers who tested improved fallows some 10 years earlier were still engaged in the practice. A newly developed intercrop system using Gliricidia sepium has been recently disseminated widely in Malawi, but it is too early to assess its adoption.

Some of these fertilizer tree systems represent a new way of farming, and this could well be an impediment to their diffusion. There are also other constraints to diffusion, such as a lack of tree germplasm and the lack of immediate benefits. However, it is important to note that in rain-fed smallholder agriculture in Africa, cereal yields are stubbornly low, precisely because the usage of all soil-fertility management practices is very low. In the face of poor and risky incentives, the integrated soil-fertility management approach seems the best way forward, largely because it promotes the use of a range of nutrient sources. Fertilizer tree systems are an important option, given their low cost and proven ability to generate significant changes in soil health.

Author notes

Frank M. Place (Skype: frankicraf) is an economist and head of impact assessment at the World Agroforestry Centre, Nairobi, Kenya (www.worldagroforestry.org).