Sustainable intensification and diversification options with grain legumes for smallholder farming systems in the Guinea savanna of Ghana

Food security is a critical issue in the Guinea savanna of Ghana where about 60% of the rural population, mostly smallholder farmers are food insecure. Food insecurity results from poor crop yields due to low soil fertility compounded by erratic unimodal rainfall and the inability of households to purchase required supplemental food. Rapid population growth means that the numbers of food insecure people are likely to increase, necessitating sustainable intensification and diversification to increase crop production per unit area of land. This thesis focused on testing spatial and temporal intensification and diversification options suitable for the variable biophysical and socio-economic conditions of smallholder farming systems in the Guinea savanna to increase productivity, mitigate the risk of crop failure, and thus to increase food self-sufficiency. One site in the southern Guinea savanna and one in the northern Guinea savanna were selected which differed in biophysical and socio-economic resources. In each site, field experiments were conducted on three fields differing in soil fertility (fertile, medium fertile, poorly fertile) to quantify: N2-fixation and N contribution to soil fertility by grain legumes in sole and intercropping; impact of replacement intercropping on increasing resource use efficiency and crop productivity; and productivity of relay (additive) intercropping and rotation of grain legumes with maize. Scenario analysis was performed with data from the N2Africa Ghana project supplemented with data from the on-farm experiments and literature to test the impacts of intensification and diversification options on household food self-sufficiency. Sole legumes fixed larger amounts of N2 than under intercropping. The soil N balance was generally positive and similar between intercrops and sole crops suggesting that both systems could be sustainable intensification and diversification options. Poor fields stimulated grain legumes to rely on atmospheric N2 for growth leading to more positive soil N balances than in fertile fields. Consequently, legumes in poor fields were more competitive with maize and led to greater intercrop yield advantage than in fertile fields. Across all fields and sites, intercropping enhanced the efficiency in resource use resulting in a 26% to 46% yield advantage over sole cropping. Intercrops were more efficient and productive in the drier northern Guinea savanna than in the wetter southern Guinea savanna. Yet the absolute larger grain yields achieved in fertile fields and in the southern Guinea savanna with more favourable soil fertility and rainfall resulted in greater net benefits. This suggests that intercropping is beneficial both in poorly fertile and fertile fields though the benefits take different dimensions. Legume-cereal rotation was superior in increasing the yield of maize without N fertiliser ranging from 0.38 t ha-1 in NGS to 1.01 t ha-1 in SGS due to residual N and non-N benefits compared with continuous maize cropping. Sowing cowpea first and relaying maize decreased maize grain yield substantially from 0.29 t ha-1 (14%) in SGS to 0.82 t ha-1 (83%) in NGS, representing 14% and 83% grain yield reductions relative to maize sown at the beginning of the season. These grain yield reductions were due to inadequate rainfall received by the relay maize. When maize was sown from the onset of the season and the cowpea relayed, the cowpea grain yield reduction was relatively smaller compared with that of maize. Such cowpea grain yield decline was similar between the SGS and NGS and ranged from 28% (0.18 t ha-1) to 47% (0.26 t ha-1) relative to the cowpea sown from the onset of the season. The cumulative grain yield of this relay system over two seasons was similar to that of the legume-cereal rotations even with cowpea failing to yield in the first season. The scenario analysis revealed a high incidence of food insufficiency among smallholder farm households in the Guinea savanna of Ghana. This ranged from 56% in the Northern region with relatively favourable rainfall, soil fertility and larger land area cropped per farm to 45% in the Upper East and Upper West regions with comparatively less rainfall, poor soils and smaller land area cropped. In addition, 21% of households in the Northern region and 37% in the Upper East and Upper West regions could only survive on their own food production for six months or less. However, the scenario analysis suggested that through intensification and diversification with grain legumes, the proportion of food self-sufficient households in the Guinea savanna could increase by 25 – 43% and those self-sufficient for a maximum of half a year decreased to 3 – 15%. Households could also generate substantial marketable surpluses to earn income. However, the total size of land cropped by a farm household matters, and improved access to markets and credit are needed to acquire the relevant inputs. Also, multi-year analysis using modelling would be relevant in providing insights on long-term nutrient balances, especially of N and soil organic matter to understand the long-term sustainability of the various options.