Bacterial community structure in bulk soil and rhizosphere of alpine plants exhibits opposite longitudinal patterns

• Alpine plants could shape their specific bacterial communities in their rhizosphere that was distinct from those in surrounding bulk soil. • The taxonomic diversity (observed richness of generalist or total OTUs, and Shannon index) of bacterial communities in the rhizosphere versus bulk soil exhibited opposite longitudinal patterns, partly due to their differential responses to soil properties (e.g. soil pH, ammonium nitrogen, and available potassium). • The bacterial community dissimilarity between paired bulk soil or rhizosphere samples was positively related with difference in longitude (i.e. distance-decay hypothesis), which could be predicted by abiotic (climate and soil properties) and biotic factors (plant phylogeny). Soil bacteria play a key role in determining vegetation dynamics and ecosystem functions in natural grasslands, and can change along biogeographic gradients due to the sensitivity to abiotic and biotic factors. However, it remains unclear whether there are differences in longitudinal patterns and main drivers of bacterial community structure between bulk soil and rhizosphere, especially in alpine grassland ecosystems. In this study, we characterized bacterial communities in rhizosphere of alpine plants (28 species) and bulk soil at 13 sites of three alpine grassland types along the longitudinal gradient (from 82.5°E to 93.8°E) in China. Our results showed that climate and soil properties (e.g. mean annual precipitation, soil pH, and organic matter content), as well as bacterial taxonomic diversity in these two soil origins, changed along the longitudinal gradient. Total OTUs richness, Shannon index, and generalist richness of bacterial communities in the bulk soil increased along longitude, apparently due to the higher organic matter content and lower available potassium content. On the contrary, Shannon index in the rhizosphere decreased along longitude and was not related with any environmental factor. The dissimilarity in bacterial communities between paired samples was positively related with the differences in longitude and environmental factors (mean annual precipitation and soil pH in particular) for the two soil origins. There was a positive relationship between bacterial community dissimilarity among rhizosphere samples and the phylogenetic distance of co-occurring plant species in the individual sites of different alpine grassland types, suggesting phylogenetic conversation in plant-bacteria interactions under field conditions. Our findings suggest that the differential responses of bacterial communities in the rhizosphere versus bulk soil to abiotic and biotic factors may underlie their opposite longitudinal patterns, which highlights the necessity of integrating biogeographic and phylogenetic approaches in exploring plant-soil bacterial associations.