Located on the southwest slope and plain areas of the Altai Mountains in China, this study aims to explore bird composition variation (beta diversity) in mountain landscape (metacommunity M), riparian landscape (metacommunity R), desert landscape (metacommunity D) and across the three landscapes (metacommunity A), and to assess how patch connectivity with environmental and spatial factors influence species distributional patterns across multiple metacommunities. In 78 transect lines over the study area, 9724 detections of 139 bird species were detected. We calculated the beta diversity, its turnover and nestedness components in four metacommunities. We used the variation partitioning method to investigate the relative importance between the environment, spatial variation and locality connectivity in driving bird community composition variation. We found high beta diversities with a small contribution of nestedness components in all four metacommunities. When only a single set of predictors is contained in the model, the predictor that best explains the variation of bird community composition is connectivity in metacommunity M, R and D and spatial predictor in metacommunity A. In all three sets of predictors, 73.8~85.4% of variations of community composition can be explained in the four metacommunities, and connectivity always contributed the most. High beta diversity and a high turnover component imply that regional-scale conservation efforts should be thought of as preserving overall biodiversity. A conservation strategy is to keep stepping-stone habitats with good connectivity in the middle of the riparian landscape. Along with the Altai-Sayan biodiversity ecoregion, the desert and riparian environments are essential for birds residing in the mountainous terrain. Furthermore, they should be regarded as integral parts of the ecoregion.
Keywords: beta diversity; bird; landscape; locality connectivity; metacommunity.