干旱区生态环境知识资源中心

The effects of vegetation root distribution on near-surface water partitioning can be two-fold. On the one hand, the roots facilitate deep percolation by root-induced macropore flow; on the other hand, they reduce the potential for deep percolation by root-water-uptake processes. Whether the roots impede or facilitate deep percolation depends on various conditions, including climate, soil, and vegetation characteristics. This paper examines the effects of root distribution on deep percolation into the underlying permeable bedrock for a given soil profile and climate condition using HYDRUS modelling. The simulations were based on previously field experiments on a semiarid ponderosa pine (Pinus ponderosa) hillslope. An equivalent single continuum model for simulating root macropore flow on hillslopes is presented, with root macropore hydraulic parameterization estimated based on observed root distribution. The sensitivity analysis results indicate that the root macropore effect dominates saturated soil water flow in low conductivity soils (K-matrix below 10(-7) m/s), while it is insignificant in soils with a K-matrix larger than 10(-5) m/s, consistent with observations in this and other studies. At the ponderosa pine site, the model with simple root-macropore parameterization reasonably well reproduces soil moisture distribution and some major runoff events. The results indicate that the clay-rich soil layer without root-induced macropores acts as an impeding layer for potential groundwater recharge. This impeding layer results in a bedrock percolation of less than 1% of the annual precipitation. Without this impeding layer, percolation into the underlying permeable bedrock could be as much as 20% of the annual precipitation. This suggests that at a surface with low-permeability soil overlying permeable bedrock, the root penetration depth in the soil is critical condition for whether or not significant percolation occurs. Copyright (C) 2010 John Wiley & Sons, Ltd.