Background: Disentangling the biogeographic patterns of rare and abundant microbial sub-communities is essential in order to understand the generation and maintenance of microbial diversity with respect to the functions they provide. However, little is known about ecological assembly processes and environmental adaptation of rare and abundant microbial sub-communities across large spatial-scale wetlands. Using Illumina sequencing, we investigated the taxonomic and phylogenetic β-diversity of rare and abundant bacterial and fungal sub-communities in Qinghai-Tibet Plateau wetland soils. Additionally, we determined environmental breadths and phylogenetic signals of ecological preferences of rare and abundant microbial sub-communities, and investigated community assembly processes of microbial taxa. Results: We found that both taxonomic and phylogenetic similarities of rare and abundant microbial sub-communities attenuated with geographical distance. Based on threshold indicator taxa analysis and Blomberg’s K statistic, abundant microbial taxa exhibited broader environmental thresholds and stronger phylogenetic signals for ecological traits than rare microbial taxa. The strong correlations between community compositional dissimilarity and phylogenetic distance of rare microbial sub-communities also revealed that rare taxa may be more sensitive to environmental changes. In addition, the rare microbial sub-communities exhibited closer phylogenetic clustering compared with abundant microbial sub-communities. The null model analysis revealed that dispersal limitation belonging to stochastic process dominated ecological assembly of abundant bacterial sub-community, and rare and abundant fungal sub-communities; variable selection belonging to deterministic process governed community assembly of rare bacterial taxa. Neutral model analysis and variation partitioning analysis further confirmed that abundant microbial sub-communities were less environmentally constrained. Soil ammonia nitrogen was the crucial factor in mediating the balance between stochasticity and determinism of both rare and abundant microbial sub-communities, as reflected by distinct differences in stochastic process with higher ammonia nitrogen content.Conclusions: Abundant microbial sub-communities may have better environmental adaptation potential and are less dispersed by environmental changes compared with rare microbial sub-communities. Our findings extend knowledge of the adaptation of rare and abundant microbial taxa to ongoing environmental change and could facilitate prediction of biodiversity loss caused by global climate change and increasing human activity in wetlands of the Qinghai-Tibet Plateau.