Spatial vegetation structure and its effect on wind erosion of Alxa dryland ecosystem

In wind erosion models, previous parameters related to vegetation morphology and density are limited in describing the spatial distribution of vegetation that influences surface heterogeneity. Thus, it is not fully understood how spatial vegetation patterns affect wind erosion on a field-scale. Based on an investigation of 36 plots of vegetation in Alxa Plateau, northwestern China, we established a multivariate linear model for temporally and spatially averaged aerodynamic roughness length ( Z _0 ) incorporating the height, roughness density, regularity of vegetation patches (curvature) and spacing between patches (connectivity). The curvature positively interacted with the connectivity in affecting the mean Z _0 , while it was the most important factor affecting the standard deviation of Z _0 . The connectivity modulated the roughness density in affecting the standard deviation of Z _0 . The spatial-related terms contributed 37% and 62% to the model variance of the mean and standard deviation of Z _0 , respectively. Our results validate the importance of spatial vegetation structure in the vegetation-airflow interactions, with a suggestion of estimating the heterogeneity of surface erodibility by intuitive spatial parameters. Based on that spatial vegetation patterns reflect the ecosystem states, a strengthened linkage between wind erosion and vegetation stability may be useful in erosion regulation in drylands.