CALCULATING SOIL NUTRIENT BALANCES IN AFRICA AT DIFFERENT SCALES .2. DISTRICT SCALE

In a recent study on the NPK balance of land use systems in sub-Sahara n Africa, it was found that scale-inherent simplifications were inevit able (Stoorvogel et al., 1993). This article reports on a similar exer cise in a well-inventorized smaller area (Kisii District, Southwestern Kenya). Land use types and land/water classes (combinations of rainfa ll zones and soil units) were combined into geographically well-define d land use systems with NPK inputs by mineral fertilizers, manure, wet and dry deposition, and biological N fixation, and outputs by abovegr ound crop parts, leaching, denitrification, and erosion. Primary data were available on applied mineral fertilizers and manure, crop yields, nutrient contents, residue removal and erosion. Deposition, leaching and denitrification were estimated using rainfall, clay, N and K conte nt, and fertilizer input. Erosion was estimated along the lines of the Universal Soil Loss Equation. The aggregated nutrient balance for the Kisii District was -112 kg N, -3 kg P, and -70 kg K ha-1 yr-1. For al l nutrients, removal of harvested product was the strongest negative c ontributor, followed by erosion. In terms of land use, nutrient deplet ion was highest under pyrethrum and lowest under tea. Sensitivity anal ysis revealed that changing mineralization rate and soil N content had an important impact on the N balance. Varying slope gradient and leng th, soil erodibility, land cover and the enrichment factor for eroded material affected all nutrients. Examples are given of possible ways t o improve the NPK balance in the Kisii District by manipulating inputs and outputs. The methodology can prove valuable in any area where the farming community is receptive to integrated nutrient management syst ems.