New approaches for stabilizing slopes along a river valley in China: investigating the potential off local species for controlling landslides through the reinforcing effects off their root systems

National audienceThe foothills of Himalayan mountain ranges are areas where plant biodiversity is extremely rich due to the semi-tropical climate and the wide variety of soil substrates. Landslides and erosion are frequent because of natural causes e.g. mountain orogeny and earthquakes, but are largely due to anthropological causes e.g. deforestation by logging and agriculture, road and dam construction. The aim of this study is to propose a new approach for slope stabilisation, by focussing on the careful management of degradation hotspots as well as biodiversity hotspots, and understanding better the processes leading to the formation of each type. We will study how plant roots reinforce soil, with an emphasis on rooting strategies of plants growing under strong ecological and mechanical constraints e.g. landslides and erosion. Our field site is in the Salween River valley, Yunnan province, southern China. We are currently examining root and shoot structure as well as mechanics of local herbaceous species, shrubs and creeping plants to determine which might have a mechanically reinforcing effect on a slope, as well as their ethno botanical advantages. Thus we will identify those which can be considered as “tools for eco-engineering” in this area. Using these data, we will then explore an emerging concept of eco-engineering: the management of erosion and biodiversity hotspots with a view to vegetating the entire degraded zone. Our research aims at studying in detail revegatation strategies with regard to ‘degradation’ hotspots (not all hotspots we will examine are due to erosion processes), which are our target zones of management. The use of mechanistic models based on finite elements methods (Kokutse et al, 2006, 2009) will enable us to simulate and calculate the Factor of Safety (FoS) of slopes for different management scenarios. This model allows individual plants to be positioned on a slope, therefore we can determine the effect of links between hotspots, as well as vegetation type, over the entire slop