| Chance Constrained Programming (CCP) and compound Poisson (negative binomial) models are used respectively, to study the profitability and diffusion of erosion-control technologies in two Extension Planning Areas (EPAs) in the Ntcheu Rural Development Project in Malawi. Two conservation technologies (agroforestry/vetiver grass combination (AV) and bare contour bunds (BB)) and two erosive technologies (wrong spacing (WS) and down slope and no conservation (NC) were studied on 200 randomly selected farms.; The results showed that capital and labor are constraints to production among WS and NC farmers. The AV technology, which was the most effective in erosion control, had also the highest per hectare net farm income. When the AV technology is effective, the average soil loss converges to the soil loss tolerance level of the farm. Furthermore, when the soil loss is greater than the soil loss tolerance level of the farm, the optimal land allocated to crops declines implying that some land should be left as fallow to regenerate or be allocated to more profitable activities. The negative binomial regression model demonstrated that the major factors in the diffusion of erosion-control technologies are extension, farmer experience, and land quality.; The analysis suggests that the following are needed: (1) joint research efforts between farmers and agricultural researchers, particularly on indigenous erosion-control technologies; (2) an increased number of extension workers in the field to work with farmers; (3) the introduction of appropriate technologies for planting activities; (4) targeted subsidized credit; and (5) ecological diversity in the production of crops most suited to the Ntcheu Rural Development Project. |
