Nutrient balances and expected effects of alternative practices in farmingsystems of Uganda

Nutrient balances were estimated for small-scale farming systems at four su b-humid, medium altitude locations in eastern and central Uganda. Data from several sources were used to estimate nutrient fluxes, including farmer in terviews, observations of the fanning systems, soil analyses, and the outpu t of simulation models. Results from the three locations were combined beca use of similarity of farming systems, with the Palissa location treated sep arately. Nutrient balances were determined to be negative for all crops exc ept for N and P in the banana-based land use type (LUT) which benefited fro m the transfer of nutrients from other land use types in organic materials. The annual crops-LUT occupied more land than other LUTs and accounted for more nutrient loss than all other LUTs combined. Crop harvest and soil eros ion were the major causes of nutrient losses at the crop and LUT levels. N and K balances were positive and negative, respectively, for the fallow and pasture LUTs. Harvests and livestock dung and urine were brought to the ho usehold LUT, and the losses in this relatively small LUT appeared to be hig h from burning, volatilization and erosion. No farmer used fertilizer and n utrients in any form were applied to banana, only. Marketing and purchasing of commodities resulted in farm-level nutrient deficits for Palissa but ha d little effect on the balances at other locations. The cumulative effect o f several low-input management practices was estimated to give nutrient bal ances of near zero for N and P, but K losses at the held-level continued to be high. (C) 1998 Elsevier Science B.V. All rights reserved.