Maximal rates of BNF recorded in the tropics reach an astonishing 5 kg N ha-1 day-1 with the green manure Sesbania rostrata (Giller, 2001). We have measured more than 250 kg N ha-1 of fixed N2 in soybean in southern Africa on smallholder farms with associated grain yields of more than 3.5 t ha-1. This demonstrates that the potential rates of BNF in legumes are not limited by the efficiency of BNF in the legume-rhizobial symbiosis. But sadly, less than 8 kg N ha-1 year-1 is often fixed by grain legumes when expressed across the whole area of an African smallholder farm due to a lack of substantial area (<15% of the farm area) planted with legumes and environmental stresses that prevent effective BNF (<35 kg N ha-1 fixed) (Mapfumo et al., 2001). Successful BNF by legumes in the field depends on the interaction:
(Legume genotype × Rhizobium strain) × Environment × Management
where environment encompasses climate (temperature, rainfall, day length etc to encompass length of growing season) and soils (acidity, aluminum toxicity, limiting nutrients etc). Management includes aspects of agronomic management (use of mineral fertilizers, sowing dates, plant density, weeding). Incidence of diseases and pests are also a function of (GL × GR) × E × M. Thus establishment of effective BNF depends on optimizing all of these components together. Legumes are often women’s crops, grown for home consumption and they are thus often grown in poorer soils with little application of manure. They are also allocated less attention in terms of labor for crop management. This means that E and M often override the potential of the legume/rhizobium symbiosis for BNF.
A step-wise approach will be used to enhancing BNF by grain and forage legumes. As shown in Figure 1, substantial increases in BNF can be made based on existing legume and inoculant technologies (for GL × GR × E × M) using current knowledge. Further substantial increases can be gained by increasing the area under legumes and retaining the legume residues in the field, which will be achievable within 4 years in many regions. Selection of rhizobial strains with enhanced BNF efficiency can be selected (including field testing) within 3-5 years. Moving to a longer time scale breeding of legumes for enhanced BNF takes longer, probably 10-15 years before varieties incorporating newly- identified traits can be moved to release. Adaptation to environmental stress in the legume/rhizobial symbiosis is poorly understood and there is a strong need for detailed plant physiology research in this area to support such breeding efforts to enhance BNF. Finally, genetic engineering approaches to enhance BNF hold promise, but may lead to enhanced BNF in farmers’ fields only in 30-50 years.
Enhancing BNF and scaling-up of legume integration in smallholder farming systems will be achieved in this project in the short-term (0-4 years) by introducing high quality rhizobial inoculants, selecting legume and rhizobial genotypes with enhanced potential for BNF and targeting these legume/rhizobium combinations within different agroecologies, with best-practice agronomy. This will lead to a major expansion of the area under legumes that fix substantial amounts of nitrogen that will also help to improve soil fertility for other crops. Our confidence that we can achieve substantial impacts in increasing amounts of N2 fixed in smallholder agriculture in sub-Saharan Africa arises from experience in a number of ‘success stories’ that have already achieved such impact with thousands of farmers. We introduce the technical background underpinning this approach in the next sections.
Figure 1: Potential returns to investment on research and development investment through different approaches to enhance the inputs from BNF (modified from Giller and Cadisch, 1995).