An on-farm study to evaluate three methods of applying legume residues in maize-based cropping systems in the central Kenya highlands was initiated in 1997 and conducted for three years. The legumes involved in the study were velvet bean (Mucuna pruriens (L.) DC) and crotalaria (Crotalaria ochroleuca G. Don), which were planted between maize (Zea mays L.) rows, two weeks after planting maize. The legumes were left growing in the field after maize harvest, until land preparation for the next maize crop. The legume forage was harvested and was either incorporated into the soil, left on the surface as mulch, or removed from the field before planting maize. The first cropping season when green manure legumes were establishing, maize grain yields were depressed by an average of 33% compared to the non legume control treatment. The legume treatments increased maize grain yields from a mean of 1.0 t ha⁻¹ in the controls to a mean of 1.6 t ha⁻¹ during the 1998 and 1999 long rains cropping seasons. Over the two cropping seasons, incorporation of legume residue into the soil gave a higher maize yield (2.1t ha⁻¹) compared to leaving the legume biomass on the surface as mulch (1.4t ha⁻¹). The effects of below-ground biomass on maize yields were minimal. The legume treatments did not significantly affect soil N, P, K, and organic carbon. During the study, farmers were able to identify characteristics of suitable green manure legumes and also identify possible niches for green manure legume technologies.

KEYWORDS

• Legume residue management,
• maize-based cropping systems,
• on-farm studies,
• smallholder farming