HighlightsIn upto 25% of fields, maize is non-responsive to fertilizer and amendments.Multiple factors that vary by site explain poor crop response to fertilizers.Low Mn, Cu and B contents are the most striking differences between the poor non-responsive cluster and others.Site specific management recommendations are needed to improve the efficiency of fertilizer application.

24Sub-Saharan Africa (SSA) is plagued by low productivity and little research is available 25 on the attainable responses and profitability to applied nutrients under variable 26 environments. The objective of this study was to determine the attainable maize grain 27 response to and potential of profitability of N, P and K application in SSA using Both response to fertilizer and value cost ratio (VCR) are highly variable and no more 32 than 61% cases for N, 43% for P and 25% for K attain VCR of 2 or more. Also, based 33 on the recent AfSIS data, VCR exceeds 1 in just 67% (N), 57% (P) and 40% (K) of the 34 cases, even when best management practices are applied on a research farm, and 35 interest rates are zero. Chances for profitability are highest when soil organic carbon is 36 1 -2% and control maize grain yield is 1 -3 t ha -1 but also depends on relatively static 37 soil properties (primarily texture and mineralogy) that are not under farmer control. We 38 conclude that return on investment of macronutrient fertilizer is highly variable and can 39 be substantially increased by helping farmers decide where to apply the fertilizers. 40Consequently, farmers need access to information on factors influencing economic 41 returns of fertilizer use in order to make the right decisions.

No abstract

16This paper reports on a Participatory Learning and Action Research (PLAR) process that 17 was initiated in three villages in Eastern Uganda in September 1999 to enable small-scale farm-18 ers to reverse nutrient depletion of their soils profitably by increasing their capacity to develop, 19 adapt and use integrated natural resource management strategies. The PLAR process was also 20 used to improve the participatory skills and tools of research and extension personnel to sup-21 port this process. The farming systems of the area were characterised for socio-economic and 22 biophysical conditions that included social organisations, wealth categories, gender, crop, soil, 23 agro forestry and livestock production. Farmers identified soil fertility constraints, their indi-24 cators, and causes of soil fertility decline, and suggested strategies to address the problem of 25 soil fertility decline. Soil fertility management diversity among households indicated that most 26 farmers were not carrying out any improved soil fertility management practices, despite pre-27 vious research and dissemination in the area. Following the diagnosis stage and exposure visits 0308-521X/$ -see front matter Ó

No abstract