The Qaidam Basin, a typical alpine arid inland basin on the northern Qinghai-Xizang Plateau, China, hosts wetland ecosystems that are strongly constrained by topography and extreme climate. These ecosystems exhibit pronounced spatiotemporal heterogeneity and fragmented distribution patterns, rendering them highly sensitive to environmental change. This study integrated Sentinel-2 remote sensing imagery with the SedInConnect model to delineate wetland patch distributions and calculate the Index of Connectivity (IC) values across the basin. Based on IC values, we stratified field sampling sites into high-, moderate-, and low- connectivity gradient groups to analyze the relationships among plant community characteristics, vegetation spatial patterns, and wetland connectivity in the Qaidam Basin. Partial Least Squares Path Modeling (PLS-PM) was further employed to quantify the driving mechanisms underlying wetland vegetation characteristics. The results revealed that wetland connectivity across the basin was generally low, with IC values up to 1. 32 and displaying a west-to-east decreasing gradient. The west and northwest were characterized by relatively continuous high-connectivity wetland networks, while fragmented and low-connectivity wetlands predominated in the east and southeast. Connectivity regulated wetland vegetation patterns primarily by affecting patch size, fragmentation, and internal adjacency. High-connectivity areas had higher class area (CA), largest patch index (LPI), and area-weighted mean patch size (AREAAM) than low-connectivity areas. Connectivity had the strongest effect on vegetation coverage, which declined sharply from 87. 577% in high-connectivity areas to 12. 152% in low-connectivity areas. Meanwhile, species diversity showed a moderately negative response to connectivity changes, whereas species evenness remained relatively unaffected. PLS-PM explained 78. 300% and 67. 500% of the variance in vegetation community and vegetation pattern, respectively. Climate played a dominant role in shaping vegetation characteristics, with significant negative effects on both vegetation community and pattern. Topography influenced vegetation indirectly through climate, and connectivity was influenced by both drivers and exerted positive effects on vegetation community and pattern. This study reveals the multi-pathway driving mechanisms underlying vegetation pattern formation in alpine wetlands, providing a theoretical foundation and decision-support framework for the scientific conservation and adaptive management of wetlands in the Qaidam Basin.
Yifan et al. (Wed,) studied this question.