OBSERVATIONS OF SOME TREE ROOT SYSTEMS IN AGROFORESTRY INTERCROP SITUATIONS, AND THEIR GRAPHICAL REPRESENTATION

Root systems of unpruned Cassia siamea and Leucaena kecocephala trees and regularly pruned hedges of C. siamea, L. leucocephala and Gliricidia sepium were examined in a semi-arid area in Kenya. Trenches were dug across rows of trees or hedges, soil profiles washed, and different sized roots counted in small 5 x 5 cm grids fully covering both faces of the trench. The roots were displayed using two computer software programmes written for this purpose: a spline technique that displays smoothed root densities in different shades; and a random dot method which displays roots as dots, using any specified scale. The fine roots of four-and-a-half-year-old C. siamea penetrated the soil to meet with rocks at depths between 0.60 and 2.0 m, and spread laterally to 9 m, traversing adjacent crop plots. Roots of eight-year-old L. leucocephala covered an extensive soil volume and included a greater density of fine roots than C. siamea at comparable distances. Above-ground pruning restricted the roots of hedges to a depth of 1.5 m, although roots of different hedges spaced 5 m apart intermingled freely in the alley space. L. leucocephala roots filled the alley very densely while those of G. sepium were sparse, particularly in the middle of the alley. Rooting density of C. siamea was intermediate. The dot method is more precise when rooting densities are low but when densities are high the splining method appears to be more visually helpful. These graphical methods greatly facilitate overall comparisons of sections of tree root systems, but quantification of root densities at different distances away from trees is still needed to understand more precisely the competitive effects of trees and associated crops.