Shrub-soil interaction and its feedbacks to vegetation patterns in Alxa plateau in NW China.

Aims: Interactions among plants are the main biotic driver for the development of spatial vegetation patterns in dryland ecosystems. The modification of soil properties by perennial plants is an important mechanism. We aimed to examine how spatial vegetation patterns were affected by the local changes in soil physiochemical properties beneath shrubs in different communities. Methods: We investigated 36 plots containing shrub vegetation on the Alxa Plateau, northwestern China. Patch sizes of all shrubs were measured. The median and curvature of patch size distribution (obtained by quadratic regression) were selected as indicators for spatial patterns. We determined soil physiochemical properties beneath the canopy of dominant and other shrub species and in adjacent intershrub spaces. A multi-model inference procedure was employed to analyze the standardized effect of the changes in soil properties and their interactions on the spatial vegetation patterns. Results: Local changes in soil physiochemical properties associated with shrub-soil feedbacks affected spatial vegetation patterns in different aspects. Salt aggregation was positively correlated with deviations from the power-law distribution, while nutrient concentration had no significant effect. Interaction between nutrient and salinity concentration positively affected the median patch size, maintaining a stable Turing-like pattern. Heterogeneity analysis showed that the dominant shrub species are representative of plant-soil interactions for the entire community. Conclusions: Our study suggests that soil salinity accumulated under the canopy of shrubs is the key soil property driving the spatial pattern of vegetation patches on the Alta Plateau. [ABSTRACT FROM AUTHOR]