Effect of spatial variation of tree root characteristics on slope stability. A case study on Black Locust (Robinia pseudoacacia) and Arborvitae (Platycladus orientalis) stands on the Loess Plateau, China

ACL-12-02International audienceVegetation is widely used for controlling shallow landslides. The mechanisms by which roots increase apparent soil cohesion is well documented and many values of root additional cohesion are available in the literature for different plant species. However, less information is given about the spatial variation of soil reinforcement by roots at a slope scale and its influence on slope stability, in particular in forest areas. The goal of this paper is to describe the spatial variability of root additional cohesion on two monospecific 17-y-old stands of Robinia pseudoacacia and Platycladus orientalis grown on slopes in the semiarid Loess Plateau of China, and to analyze numerically the effect of this variability on slopes stability. For this purpose, vertical trenches were dug at different distance and directions around trees situated at three different slope locations, i.e. up-, mid- and down-slope. Grids with a 10×10 cm mesh were placed on vertical walls. Roots were counted within each grid cell and split according to their diameter class. Root area ratio (RAR) was estimated and compared among different positions around the trees and at three different locations along the slope. Roots tensile strength was determined with laboratory mechanical tests. RAR and root tensile strength were used as inputs in six different root reinforcement models to calculate root additional cohesion. A 2D finite element model of slope stability was developed and used to calculate the increase in factor of safety (FoS) due to root additional cohesion on rectilinear and terraced slopes. Results showed that both root tensile strength and Young's modulus of R. pseudoacacia was about two times higher than tensile strength of P. orientalis. RAR distribution had a strong relationship with local soil moisture content measured in July during the raining season, and was significantly different with regards to tree location on the slope. The six theoretical models used to estimate the root additional cohesion gave different vertical profiles of root reinforcement distribution according to the underlying hypothesis on how forces are transferred to the roots. Theoretical analyses of slope stability showed that terraced slopes were 20% more stable than rectilinear slopes, disregarding the differences in hydrological regimes between these two terrain morphologies. Numerical sensitivity analyses also showed that the FoS reached an asymptotic value when increasing root additional cohesion. Actual additional cohesions of the two studied sites corresponded to FoS that were already close to this asymptotic values. Consequently variations of these actual root cohesions would not much affect slope stability. However it was showed that more attention should be given to the reinforcement of the bottom part of the actual slopes, where roots have a larger positive impact on the Fo