Long-term warming and nitrogen addition drive interface-specific bacterial assembly and enhance soil-leaf microbial connectivity

Abstract Plant–microbe interactions are strongly influenced by global environmental change, but bacterial responses in diversity, assembly, and cross-interface connectivity across multiple plant–soil interfaces remain poorly understood. Here, we investigated bacterial communities associated with the dominant grass species Stipa breviflora in leaf epiphytes, leaf endophytes, rhizosphere soil, and bulk soil, using an 18-year field experiment with infrared warming and N addition in a temperate desert steppe. We evaluated interface-specific responses in bacterial community structure, assembly processes, and the contributions of different interfaces to leaf endophytic bacterial communities. We found that interface identity plays the dominant role in shaping bacterial phylogenetic α- and β-diversity, with soil communities exhibiting higher diversity than phyllosphere communities. Nitrogen addition significantly reduced phylogenetic α-diversity and resulted in a more stochastic phylogenetic structure in bulk soil, whereas warming enhanced phylogenetic clustering in rhizosphere soil. Notably, warming increased phylogenetic dispersion and β-diversity of leaf epiphytic bacteria but decreased β-diversity in rhizosphere soil, indicating contrasting above- and belowground responses. Community assembly across interfaces was dominated by homogeneous selection, with stochastic processes contributing in an interface-dependent manner. Leaf endophytic communities remained remarkably stable, suggesting strong host filtering. Microbial source tracking revealed that warming and N addition increased soil-derived contributions to leaf endophytes, with over 70% of the community originating from soil under global change treatments. Overall, our results demonstrate that long-term warming and N addition drive interface-specific bacterial assembly and enhance soil to leaf microbial connectivity, highlighting the importance of a multi-interface perspective in understanding microbial responses to global change.