Chakula bila kulima? : trade-offs concerning soil and water concervation in heterogeneous smallholder

Soil and water conservation practices need to be tailored to suit the diverse local conditions in smallholder farms.Using a combination of survey methods, field experimentation over several seasons and farm scale analysis, this research explored the targeting of recommended options to field and farm types. Smallholder farmers’ in Mbeere and Meru South Districts of Central Kenya acknowledged the occurrence of soil erosionin their farms and understood the water erosion process. Trash lines were common in the low potential Mbeere area for the control of erosion, except for farmers with high resource endowment who instead preferred fanya juu and vegetation barriers. In Meru South, contour farming was popular for different farmers although the preference was for vegetative barriers with multiple benefits. Three field types on a relative scale of soil fertility were identified by the farmers: good, medium and poor. Physical and vegetative measures were more common and well maintained in good fields but rare and neglected in poor fields. Farming on sloping arable fields with no vegetative barriers lead to soil degradation and establishment of vegetative barriers curbed soil erosion. Napier grass barriers were efficient in conserving soil and water but competed with crops for available water. This competition was especially strong with minimum tillage even when the Napier was intensely harvested. Leucaena barriers had a complementary water use pattern with crops across tillage practices but were less efficient for soil and water conservation. Considering economic returns and the soil conserved, leucaena barriers had attractive and less risky economic returns across tillage practices but conserved less soil. Napier barriers with regular tillage presented a win-win scenario for farmers and environmental impacts because of simultaneous attractive economic returns and efficient soil conservation. Cumulative maize grain yields in the good fields were above 15 Mg ha^-1 across cropping seasons and were not influenced by tillage and crop residue retention. The cumulative grain yields in the medium fields were above 10 Mg ha^-1 across cropping seasons and were greater with crop residue retention. In the poor fields, cumulative grain yield was less than 10 Mg ha^-1 across seasons and minimum tillage resulted in yield decrease while crop residue retention did not affect yields. For the poor fields, emphasis should be placed on the rehabilitation of soil physical and chemical attributes. At farm level, retention of crop residues was not viable due to use of crop residues for livestock feed. Minimum tillage was of interest to well-endowed farmers who had labour constraints. Poor farmers were interested but would not afford herbicides and had no access to sprayer pumps. Long term studies and farm scale modelling are necessary to unravel further the complexity in heterogeneous smallholder farming system for better fitting of recommended soil and water conservation options.