Use of15N for fertilizer N recovery and N mineralization studies in semi-arid Kenya

Maize and beans were grown on a ferralsol at Kiboko, Kenya, with up to 120 kg N ha^–1. Within the 10 kg N ha^–1 plots,¹⁵N labelled fertilizer was applied in microplots. There was no significant response in yield to fertilizer N and labelled N recovery was low, being 7.5% or less in one season and 17.7% or less in the second season. Samples of Kiboko soil at four different water contents were incubated and the rate of gross N mineralization over 7 days was calculated, utilizing¹⁵N labelling of the mineral N. Gross N mineralization increased greatly with soil moisture and a fitted relationship between gross N mineralization rate and soil water content was obtained. Using measurements of soil water content at the field site, daily values of the soil N supply by gross mineralization were calculated. On average, modelled gross soil N mineralized could supply much (> 69%) of the N removed from the plots. It is suggested that the lack of response to fertilizer N may be explained by the coincidence of a high rate of N mineralization, and increased crop demand, caused by the onset of rain.