Plant and Soil 264: 129–139, 2004. © 2004 Kluwer Academic Publishers. Printed in the Netherlands. 129 Competition in tree row agroforestry systems. 3. Soil water distribution and dynamics S.J. Livesley 1,4 , P.J. Gregory 2 & R.J. Buresh 3 1 Forest Science Centre (DSE), Water Street, Creswick, Victoria 3363, Australia. 2 Soil Science Department, The University of Reading, Whiteknights, P.O. Box 233, Reading, RG6 6DW, U.K. 3 IRRI, DAPO Box 777, Metro Manila, The Philippines. 4 Corresponding author ∗ Received 24 September 1999. Accepted in revised form 10 December 2003 Key words: agroforestry, Grevillea robusta, maize, Senna spectabilis, soil water content, water balance Abstract The purpose of this study was to test the hypothesis that soil water content would vary spatially with distance from a tree row and that the effect would differ according to tree species. A field study was conducted on a kaolinitic Oxisol in the sub-humid highlands of western Kenya to compare soil water distribution and dynamics in a maize monoculture with that under maize (Zea mays L.) intercropped with a 3-year-old tree row of Grevillea robusta A. Cunn. Ex R. Br. (grevillea) and hedgerow of Senna spectabilis DC. (senna). Soil water content was measured at weekly intervals during one cropping season using a neutron probe. Measurements were made from 20 cm to a depth of 225 cm at distances of 75, 150, 300 and 525 cm from the tree rows. The amount of water stored was greater under the sole maize crop than the agroforestry systems, especially the grevillea-maize system. Stored soil water in the grevillea-maize system increased with increasing distance from the tree row but in the senna-maize system, it decreased between 75 and 300 cm from the hedgerow. Soil water content increased least and more slowly early in the season in the grevillea-maize system, and drying was also evident as the frequency of rain declined. Soil water content at the end of the cropping season was similar to that at the start of the season in the grevillea-maize system, but about 50 and 80 mm greater in the senna-maize and sole maize systems, respectively. The seasonal water balance showed there was 140 mm of drainage from the sole maize system. A similar amount was lost from the agroforestry systems (about 160 mm in the grevillea-maize system and 145 mm in the senna-maize system) through drainage or tree uptake. The possible benefits of reduced soil evaporation and crop transpiration close to a tree row were not evident in the grevillea-maize system, but appeared to greatly compensate for water uptake losses in the senna-maize system. Grevillea, managed as a tree row, reduced stored soil water to a greater extent than senna, managed as a hedgerow. Introduction Annual crops often exploit only a small fraction of the available rainfall and stored soil water reserves. The integration of perennial trees within a farming system can increase the amount of water transpired and increase overall biomass productivity (Ong et al., 1992; Wallace et al., 1995). This may be achieved directly when trees exploit the rainfall and stored wa- ter reserves outside the cropping seasons and/or when a greater proportion of the rainfall within a cropping ∗ FAX No: +61-3-5321-4166. E-mail: stephen.livesley@dse.vic.gov.au season is transpired rather than evaporated, run-off or drained to below the rooting zone (Ong et al., 1992). It may also be achieved indirectly when modification of microclimatic conditions by trees increases the tran- spiration efficiency of the crop, the unit production of biomass per unit water transpired (Brenner, 1996). In semi-arid and dry sub-humid areas, evapora- tion from the soil surface can account for 30–60% of the annual rainfall (Cooper et al., 1983; Wallace, 1991). Trees can reduce these losses and conserve soil water by providing shade, reducing wind speed and increasing infiltration with mulch layers and improved soil structure (Torquebiau and Kwesiga, 1996; Young,