Main content area

Strong spatial-temporal patterns in maize yield response to nutrient additions in African smallholder farms

Njoroge, Samuel, Schut, Antonius G.T., Giller, Ken E., Zingore, Shamie
Field crops research 2017 v.214 pp. 321-330
NPK fertilizers, corn, exchangeable potassium, fertilizer application, grain yield, nitrogen, phosphorus, potassium, potassium fertilizers, small-scale farming, soil, soil analysis, soil fertility, soil organic matter, Kenya
Large variability in crop responses to macronutrient application at various spatial scales present challenges for developing effective fertilizer recommendations for crop production in smallholder farming systems of sub-Saharan Africa. We assessed maize yield responses to nitrogen (N), phosphorus (P) and potassium (K) application and evaluated relationships between crop responses to N, P and K application and soil analysis data. Nutrient omission trials were conducted on 23 farms located in Sidindi, Western Kenya, selected to be representative of the main soil and management factors in maize based systems in Siaya County. Treatments included a control and PK, NK, NP and NPK applications. The trials ran for six consecutive cropping seasons, without changing treatments or plot location, covering the period 2013–2015. Strong spatial-temporal patterns in maize yield responses to N, P and K applications were observed. Average maize yields in the control, PK, NK, NP and NPK treatments were 2.8, 3.2, 5.1, 5.1 and 5.5tha−1 at 88% dry matter respectively in the first cropping season, and 1.1, 1.4, 2.9, 3.6 and 5.3tha−1 at 88% dry matter respectively in the sixth cropping season. In all seasons, variability in maize yield between fields was greatest in the control treatment followed by the NK treatment and least in the NPK treatment. Mean relative yield was 0.6, 0.92 and 0.93 for N, P and K respectively, in the first cropping season, and 0.25, 0.52 and 0.68, respectively, in the sixth cropping season. Six main maize yield response categories were identified that differed in the maize grain yield responses to recursive N, P and K applications. Maize yield responses to N, P and K were not fully accounted for by soil organic matter, soil available P and exchangeable K respectively. Our results indicate that current methods for soil analysis do not adequately predict the response to application of N, P and K fertilizer under the highly variable soil fertility conditions encountered in smallholder farming systems. The strong spatial-temporal patterns observed present major challenges for the development of effective site-specific fertilizer recommendations. Potential avenues for future research and options for more effective intensification strategies are discussed.