Soil water dynamics in cropping systems containing <Emphasis Type="Italic">Gliricidia sepium</Emphasis>, pigeonpea and maize in southern Malawi

The water dynamics of cropping systems containing mixtures of Gliricidia sepium (Jacq.) Walp trees with maize (Zea mays L.) and/or pigeonpea (Cajanus cajan L.) were examined during three consecutive cropping seasons. The trees were pruned before and during each cropping season, but were left unpruned after harvesting the maize; prunings were returned to the cropping area in all agroforestry systems to provide green leaf manure. The hypothesis was that regular severe pruning of the trees would minimise competition with crops for soil moisture and enhance their growth by providing additional nutrients. Neutron probe measurements were used to determine spatial and temporal changes in soil moisture content during the 1997/98, 1998/99 and 1999/00 cropping seasons for various cropping systems. These included gliricidia intercropped with maize, with and without pigeonpea, a maize + pigeonpea intercrop, sole maize, sole pigeonpea and sole gliricidia. Soil water content was measured to a depth of 150 cm in all treatments at 4–6 week intervals during the main cropping season and less frequently at other times. Competition for water was apparently not a critical factor in determining crop performance as rainfall exceeded potential evaporation during the cropping season in all years. The distribution of water in the soil profile was generally comparable in all cropping systems, implying there was no spatial complementarity in water abstraction by tree and crop roots. However, available soil water content at the beginning of the cropping season was generally lower in the tree-based systems, suggesting that the trees continued to deplete available soil water during the dry season. The results show that, under rainfall conditions typical of southern Malawi, the soil profile contains sufficient stored water during the dry season (ca. 75–125 mm) to support the growth of gliricidia and pigeonpea, and that gliricidia trees pruned before and during the cropping season did not deleteriously compete for water with associated crops. Water use efficiency also appeared to be higher in the tree-based systems than in the sole maize and maize + pigeonpea treatments, subject to the proviso that the calculations were based on changes in soil water content rather than absolute measurements of water uptake by the trees and crops.