Application of computerized land evaluation systems in Tanzania: a case study in Kilosa District, Morogoro Region

Abstract

Automated Land Evaluation System (ALES) and Land Evaluation Computer System (LECS) were applied in Kilosa District, Tanzania to test their applicability and adaptability in the area within the context of low-input rainfed maize farming. The study comprised physical land suitability classification of dominant soils identified through reconnaissance soil and land resources survey and comparison of predicted yields with field data. ALES predicted that four soil units had moderate potential with 80% yield, while two soil units had poor potential with 40% yield. Comparison of farmers' reported yields with those of ALES indicates significant correlation (R>0.9, P<0.001). LECS predicted that two soil units had low potential with yields of 30% while other soils were not suitable with zero yield and poor correlation (R<0.5, P<0.5). LECS results after improvement with Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS) model show that LECS predictions are still poor (R<0.06, P<1) and two soil units classify as not suitable with zero yield. Comparisons of yields predicted by the systems with those reported by farmers show that ALES was readily adapted to the physical conditions in the area. LECS, a fairly rigid system, requires improvement before it could be applied in the area. Given the strength of ALES, it could considerably facilitate land evaluation in Tanzania.

UNISWA Jnl of Agric Vol 10 2001: pp 40-50