Effect of combined deficit irrigation and grass competition at plantation on peach tree root distribution

Agroforestry systems success is predicated on the assumption of root spatial separation between the tree and the crop species. Such a feature, where tree roots prospect subsoil horizons whereas intercrop roots prospect top soil horizon is thought to eventually happen as the system ages and trees get older. However, since roots are very plastic, we can hypothesise that it is possible to shape the tree root system when trees are young, i.e. after plantation. In this experiment, we tested the hypothesis that combining a moderate water deficit to change the carbon allocation pattern in favour of root over shoot growth combined with the top soil competition brought about by the intercrop will force tree roots to grow deeper, thus leading to the vertical separation of the trees and the intercrop root systems. To test this hypothesis, a peach tree orchard was established in 2014 with three water regime treatments: (i) a fully irrigated control without intercrop (C), (ii) a moderate water deficit treatment without intercrop (RDI) and (iii) a moderate water deficit treatment intercropped with a grass + legume mixture (RDI + G). Roots were manually excavated at the end of the first and the second growing seasons and root length and biomass per soil horizon and distance to the tree trunk were measured. The juvenile tree root system in all treatments was mainly plagiotropic, reaching 1.5 m from the tree trunk (middle of the inter row) horizontally as opposed to 0.7 m vertically, without difference between treatments. The combination of water deficit and intercrop competition reduced tree root biomass fourfold in 2014 and by 3 in 2015. Tree roots in RDI + G were also excluded from the topsoil horizon (0-10 cm) due to grass + legume competition, but because of the strong reduction in their total biomass, they did not grow in deeper soil horizons than the control tree roots (C). Secondary vertical tap roots were only starting to grow by the end of the second year, suggesting that root growth at depth could not take place the year after plantation especially in grafted scions.