Productivity of hedgerow shrubs and maize under alleycropping and block planting systems in semiarid Kenya

A field experiment was conducted for eight cropping seasons from 1988 to 1991 in semiarid Machakos, Kenya, to compare the productivity of shrubs and crops in intercropping (alleycropping) versus block (sole) planting systems. The study, conducted in a split-plot experiment with three replications, consisted of two tree species (Leucaena leucocephala andSenna siamea syn.Cassia siamea) in the main-plots and combinations of two planting systems (alleycropping and block planting of the shrubs and maize) with five different ratios of land allocation for the shrub and crop (1000, 1585, 2080, 2575, and 0100) in sub-plots. Thus, a sole maize and a sole tree were included in the subplot treatments. The spacing between hedgerows of shrubs in intercropping was 6.7, 5, and 4 m, respectively, corresponding to 15, 20, and 25% land allocation to shrubs. The trees were pruned to 0.5 m height four times a year. Intercropped hedgerows of senna and leucaena produced 10% and 24% more biomass than their respective block planting systems. On average, leucaena produced more biomass than senna. Maize alleycropped with leucaena yielded 16% less grain than sole-crop maize, whereas senna intercropping caused hardly any maize-yield reduction. Compared with the respective sole-crop systems, leucaena intercropping did not affect land equivalent ratio (LER), whereas LER increased by 28% with senna intercropping. The different tree:crop land occupancy ratios did not affect the production of either the component species or of the total system, except that LER declined with incrreased spacing between hedgerows. It is concluded that in semiarid highlands of Kenya, leucaena and crops should better be grown in sole blocks, not in alleycropping. In the case of slow-growing species such as senna, intercropping is worthwhile to consider only if the additional labour needed does not pose a serious problem for management, and the species has fodder value.