Extreme environments foster phylogenetically diverse microorganisms and unique community assembly patterns. Plateau saline marsh lakes represent understudied extreme habitats characterized by dual stressors of high salinity and low temperature. Here, we analyzed the soil bacterial and archaeal diversity in three salt marshes of the Qaidam Basin on the Qinghai-Tibetan Plateau, China. While the bacterial and archaeal alpha diversity showed no significant differences among the three salt marshes, the community composition varied significantly. Notably, soil salinity (indicated by electric conductivity, EC) exerted opposing effects on microbial diversity—suppressing bacterial while promoting archaeal communities. Stochastic processes were the predominant mechanism for both bacterial and archaeal community assembly, where the weights were, in descending order, drift, homogeneous selection, and dispersal limitation. Network analysis revealed predominantly positive co-occurrence patterns within both bacterial and archaeal communities. We did not find a direct relationship between any bacterial or archaeal co-occurrence network properties and soil EC, but there was a significant correlation of network complexity to microbial diversity, which was influenced by EC. Our findings indicate distinct responses of bacterial and archaeal diversity to varying salinity levels, while the underlying assembly processes appear to be conserved in driving shifts in community diversity in plateau salt marsh wetlands.

Urban wetlands, encompassing both natural and constructed ecosystems, are vital for urban resilience. Understanding how plant functional traits adapt to these distinct habitats is crucial for ecological management. This study investigates the intraspecific variation and trait covariation patterns of the common reed (Phragmites australis) in two contrasting urban wetland types in Ganzhou City: a stagnant, engineered constructed wetland and a dynamic, natural riverine wetland. This contrast represents a key gradient in hydrological regime and anthropogenic influence. We measured 22 morphological and chemical traits to assess trait differences, variability (coefficient of variation), and correlation patterns. Volcano plot analysis revealed significant habitat effects: reed in natural wetlands exhibited higher levels of Cu, P, N, and leaf moisture content (LMC), whereas those in constructed wetlands had higher Ca content. Traits such as Na, Mn, and Al showed high intraspecific variability. Correlation analyses revealed significant trade-offs and integrations among traits, such as positive correlations between LMC and nutrients (K, Cu), and negative correlations between Ca and key leaf morphological traits. Principal component analysis (PCA) further confirmed a significant separation along PC1, driven primarily by nutrient elements (Cu, P, K) and LMC, with natural wetlands scoring higher. In contrast, PC2, associated with leaf morphological traits (e.g., leaf area, leaf width), showed no significant inter-habitat difference. Our findings demonstrate that P. australis employs distinct ecological strategies by adjusting its functional traits and resource allocation in response to different urban wetland environments. This highlights the critical role of intraspecific trait variation in plant adaptation and has important implications for wetland restoration and the design of constructed ecosystems.