Assessing the Biophysical Determinants of Variability in Yields and Response to Fertilizers in Smallscale Farming in Southwethern Uganda.

There are a growing number of initiatives that aim to increase fertilizer use in many parts of sub-Saharan Africa in support of the African Green Revolution. Numerous reports indicate a high variability in yields within and between farms; this could affect the crop’s response to fertilizers and therefore their efficiency.  Such variability has been attributed to existence of a soil fertility gradient, resulting from differences in resource allocation associated with management strategies.  Other biophysical characteristics beyond management, such as soil physical and chemical properties, topography and other environmental attributes could play a major role in this variability.   Understanding the determinants of this variability could be a step forward in predicting response to fertilizers and establishing guidelines on efficient use of fertilizers tailored to specific locations.  

Multi-location nutrient omission trials were conducted in 16 fields distributed with a 10km x 10km grid in southwestern Uganda to determine the importance of individual major nutrients and of micronutrients for maize production as a way to identify appropriate fertilizers adapted to the area. A total of 10 treatments were run per field: a Control treatment with no nutrients added, a treatment with the application of the 3 major nutrients N, P, and K (NPK treatment), 3 treatments in which N, P, and K were omitted from NPK treatment one at a time, a treatment with S added to NPK,  a treatment  with multinutrients (Ca, Mg, S, micronutrients) in addition to NPK, 2 treatments in which K is not applied but S or S and Zn were added, and a treatment in which lime is added to NPK.

Maize yields in the control treatment were highly variable between the various fields, ranging from below 1Mg ha^-1 to 6 Mg ha^-1.   The response to the combination of N,P and K, representing common fertilizers used in Africa,  varied between fields, with 7 fields showing little or no response. We link the soil parameters and environmental attributes to yield data to understand the major determinants of variability in control yields and in the maize crop’s response of fertilizers. Major reasons for the observed non-responsiveness to fertilizers are discussed.