Water conservation and crop production under two agroforestry systems: a Laikipia district case study

Abstract

The influence of Plectranthus barbatus live fence and GreviIlea robusta trees on soil moisture and crop production was explored for two growing seasons (October - January 1990/91 and March - August, 1991) at Matanya, Laikipia District, Kenya. Two experiments were set up separately for the live fence and trees. There were two management practices; root pruning and no pruning (30 cm deep and 40 cm away from the live fence and trees). Soil moisture was monitored at three distances away (94 cm, 188 cm and 375 cm) from the live fence/trees using a neutron probe upto 170 cm deep. Grop performance and product ion for maize and beans intercropping were evaluated row-wise in the live fence and at four distances radially from the trees. In the unpruned live fence experiment, in the first season maize biomass production at the third and fourth rows was significantly different (at 5 % level) from the first and second rows. The comparison of the pruned and unpruned live fence plots showed that maize biomass production in the the first season was 147 %, 59 % and 5 % higher in the first three rows of the pruned plot from the first three rows of the unpruned plot. During rainy periods available soil water within 45 cm depth was higher at 375 cm away from the live fence by up to 37 % and 100 % than at 188 cm and 94 cm, respectively in the unpruned plot. Furthermore up to 65 % (at 94 % cm) and 20 % (at 188 cm) higher available soil water was monitored in the pruned plot . than the unpruned plot. However during dry periods probably evaporation (at 375 cm) compensated root competition for moisture (at 94 cm) and thus there was no difference between treatments. In general it could be said that competition for moisture could reach up to the second row of maize (188 cm). The investigation in the tree experiment has shown that crop performance and product ion were not significantly different between the four distances and between the two management practices. Pruning has not proved to be beneficial in terms of increasing yield by reducing root competition for moisture. Comparison of the tree system with the local and mulching methods proved that the system used soil water that would otherwise have been lost as deep percolation, without reducing the crop production. Thus the overall productivity of the system was better than the local and mulching methods.