Competition for growth resources between woody and crop species is sai
d to be the main reason for failure of hedgerow intercropping (alleycr
opping) in semiarid tropics, but the mechanisms of competition are not
clearly understood. Tn this study, conducted in the semiarid highland
s of Kenya, soil-water changes and root dynamics were monitored during
two rainy seasons from a long-term, replicated, alleycropping experim
ent. The treatments were (i) maize (Zea mays L.) grown alone, without
fertilizer; (ii) maize, without fertilizer, intercropped between hedge
rows of leucaena [Leucaena leucocephala (Lam.) de Wit], with hedgerow
prunings returned to the alleys; and (iii) maize grown alone, with fer
tilizer at 40 kg N and 18 kg P ha(-1). Available water in 1.25 m of so
il depth under alleycropping was lower than with maize alone. Depletio
n of soil water by hedgerows continued after maize harvest and carried
water deficits in alleycropping from one season to the next. Leucaena
provided 1.45 Mg ha(-1) leaf biomass, which contained 41.6 kg N and 2
.5 kg P ha(-1); moreover, it added 0.51 Mg ha(-1) season(-1) of root b
iomass to the soil, equivalent to 7 kg N and 0.2 kg P ha(-1). Therefor
e, N contributed through the alleycropping system was equivalent to th
e recommended fertilizer level; the system, however, did not meet the
P requirements of the crop. Maize did not respond to fertilizer in bot
h seasons, and the alleycropped maize yielded lower than the unfertili
zed maize. Our study indicates that under water-limiting conditions, a
lleycropping is detrimental to crop yields because competition of tree
s with crops for water outweighs the likely soil-fertility benefits.