How do cowpea and groundnut improve soil N fertility and yields of succeeding sorghum crop in the Guinean savannah zone of Burkina Faso (West Africa)?

Field experiments were carried out in a weakly acid (pH H2O 6.5) Ultisol at Farako-Ba (4 deg 20' West, 11 deg 6' North and 405 m altitude) in the Guinean savannah zone of Burkina Faso to study the beneficial effects of cowpea (Vigna unguiculata) and groundnut (Arachis hypogaea) on grain yield of succeeding sorghum. The first experiment was a factorial of three crop rotations (cowpea-sorghum, groundnut-sorghum and sorghum-sorghum) as first factor and five fertilizer treatments (NPK fertilizer, NPK+ dolomite, NPK+ manure, NK+ phosphate rock and a control) as second factor in a split plot design with a randomised block arrangement and four replications. Soil mineral N, nematode infection and fertilizer N recoveries were studied during three years (2000 to 2002). Biological nitrogen fixation (BNF) in cowpea and groundnut was measured in the second year (2001) with 15N isotopic dilution method using a non-fixing cowpea as reference crop. Groundnut fixed 8 to 23 kg N ha-1 and the percent of N derived from the atmosphere (% Ndfa) varied from 27 to 34%. Cowpea fixed 50 to 115 kg N ha-1 and the Ndfa varied from 52 to 56%. Compared to mineral NPK fertilizer alone, legumes fixed more nitrogen from the atmosphere when NPK fertilizer was combined with dolomite or manure. Compared to water-soluble phosphate, phosphate rock increased BNF by cowpea. Highest responses to N fertilizer applications were found when sorghum was rotated with legumes and the lowest in mono cropping of cereals. Compared to mono cropping of cereals (maize-sorghum or sorghum-sorghum), sorghum produced 2.9 and 3.1 times more grain yields when it was rotated with groundnut or cowpea respectively. In a second experiment, the response of a succeeding sorghum to N fertilizer, nitrogen fertilizer equivalencies (NFE) of the two legumes and nematode infection in the soil and roots of the succeeding sorghum were studied. The NFE of groundnut (35 kg N ha-1) was higher than that of cowpea (25 kg N ha-1). However, soil and succeeding sorghum roots were most infected by nematodes in cowpea-sorghum rotation while groundnut in the rotation decreased nematode infections. The high NFE of groundnut compared to cowpea indicated that positive effect of groundnut on nematode suppression was accounted in NFE as N effect. A better use of N fertilizer was observed in legume-sorghum rotations. In continuous sorghum, fertilizer N use efficiency (NUE) was 20% whereas NUEs in cowpea-sorghum and groundnut-sorghum rotations were 28 and 37% respectively. The soils of legume-sorghum rotations contained more mineral N (40-50% more than the monocropping control), thus increasing soil N supply to succeeding sorghum compared to mono cropping of sorghum. The highest total N uptake by sorghum was found in legume-sorghum rotations. It was estimated from the response curves that 51 and 58 kg N ha-1 should be applied to sorghum when it was rotated with groundnut or cowpea respectively to achieve optimum grain yields of 2,000 kg ha-1. Recommended NPK fertilizer should be applied to the two legumes. Best results were obtained when legumes received NPK fertilizer plus dolomite or manure. But the local phosphate rock can be used as source of P on groundnut. (author)