Diverse response of vegetation growth to multi-time-scale drought under different soil textures in China's pastoral areas

Abstract The pastoral areas of China are mainly located in ecologically fragile regions, where its ecosystems are highly sensitive to drought trends. Although numerous studies have been carried out on the response of vegetation to droughts, it is not entirely clear whether soil properties can influence this relationship. Using the Normalized Difference Vegetation Index (NDVI) and the Standardized Precipitation Evapotranspiration Index (SPEI), covering the period 1982 to 2015, we carefully analyzed drought impacts on vegetation in China's pastoral areas, to determine the effects of vegetation communities and soil types on vegetation response to multi-time-scale drought. Significantly positive correlations between NDVI and SPEI were observed in most regions, properly indicating that vegetation was largely influenced by drought, particularly the pastures in Inner Mongolia. Generally, forest was sensitive to longer time-scales of droughts, while grassland and cropland showed a close relationship with shorter or median drought time-scales. However, noticeable differences were found on the Tibetan Plateau, mainly because drought was not the main factor affecting vegetation growth in the region. The NDVI-SPEI correlations and the corresponding SPEI time-scales of each soil texture differed considerably, even in areas of the same land cover type, revealing that soil properties, here mainly refer to soil texture (classified by fractions of each separate soil, i.e., sand, silt, and clay), can assuredly affect the resistance and resilience of vegetation to drought stress. The underlying mechanism is the difference in particle size and permeability which can alter the storage and position of available soil water, thus affecting water absorption by the root system. Our results highlight the considerable importance of properly integrating edaphic factors when exploring the impact of likely climate change on ecosystems.