Tree effects on crop growth on a phosphorus-fixing Ferralsol

In agroforestry systems trees affect crops. This study aimed at understanding how trees grown in lines for pole production affect maize on an unfertilized, P-fixing Ferralsol, in the agricultural area of Western Kenya. I hypothesized that, because maize production was primarily limited by P-deficiency, trees would affect maize growth by effects on P-uptake of maize, as determined by P-availability of the soil, soil water contents and root-length densities.Field-based experiments were used to investigate 1) rhizosphere modifications by trees and maize roots and their effects on P-availability, 2) tree water-use and extraction from the soil profile, and 3) root distributions of trees and maize. The tree-crop interaction model WaNuLCAS was adjusted and used to compute the effects on maize production of soil drying and rhizosphere modifications as caused by the trees.Soil drying occurs if water-use of trees is larger than water-inputs by rainfall. High water-use of trees in tree lines was mainly related to high leaf-area and low leaf-efficiency. Meteorological variables influenced water-use of tree lines less. The importance of leaf-area as determinant of water-use of tree lines was due to the large range over which water-use of tree-lines responds linearly to increases in leaf-area.Water extraction from the soil with distance to the tree line (horizontal dimension) was highest in the zones near the tree line, and was not clearly related to relative root length density. This was due to the decreasing water-potential gradient, between root and soil, with increasing distance to the stem-base. Vertically, root length densities decreased more strongly than water-potential gradient, and determined the water-extraction profile.Intermediate to high organic anion contents and pH increases in rhizosphere soil of Cassia spectabilis (syn. Senna spectabilis ) and Grevillea robusta mobilized soil P, which was taken up by the trees themselves. Residual P-availability was not increased. The relatively long duration of pH-increases and organic anion exudation in the rhizosphere of Cassia , combined with the relatively high synlocation of Cassia and maize roots, suggests that maize can profit from P mobilized by Cassia roots.Acid phosphatase activity was highest in the rhizosphere soil of Grevillea and probably caused the observed shift from organic to inorganic P-fractions in the soil. However, maize showing low acid phosphatase activity in the rhizosphere, depleted organic-P fractions in the rhizosphere soil similarly as Grevillea . Therefore, it is likely that uptake of organic P from soil is more limited by its availability in the rhizosphere than by hydrolysis catalyzed by phosphatase activity.Model simulations revealed that a decrease in soil water content of 2.5 % (at high pF) decreased maize production by 30- 40 %, due to 1) decreased P-diffusion due to soil drying, and 2) the cumulative effect of decreased P-diffusion on decreases in maize and maize-root growth. Maize yield increases of 12 % as measured near Cassia tree lines could only be simulated when organic anion and pH effects on P-availability in the rhizosphere of Cassia were included in the model. Without rhizosphere modifications, Cassia would compete with maize, due to its high demand for P. However, positive rhizosphere effects as exerted by Cassia were minimal if soil-drying induced P-deficiency would occur simultaneously.