Knowledge about the ecological patterns of abundant and rare bacteria in regulated river ecosystems, especially with respect to their community traits, is an important but poorly investigated subject. In this study, we examined the community assembly of abundant and rare bacteria, as well as their environmental adaptation, across complex environmental gradients in sediments of the Yarlung Tsangpo River on the Tibetan Plateau. Results showed that abundant taxa exhibited broader environmental thresholds and stronger phylogenetic signals for ecological traits than rare taxa. In contrast, rare taxa were more sensitive to environmental changes and showed stronger phylogenetic clustering. Although both subcommunities exhibited significant distance-decay patterns, the abundant subcommunity was governed primarily by dispersal limitation, while the rare subcommunity was strongly driven by heterogeneous selection. The similar distribution patterns but contrasting assembly mechanisms affecting abundant and rare subcommunities resulted from the differences in environmental adaptation. Forest area and total nitrogen were key factors in determining the stochastic and deterministic assembly for abundant and rare subcommunities, respectively. Additionally, rare taxa might play potential roles in maintaining network stability, although they were less connected and located more peripherally within the network. Collectively, our study provides a new perspective for the ecological significance of abundant and rare bacteria in fluvial sediments, and facilitates the prediction of microbial responses to ongoing environmental changes in the Yarlung Tsangpo River.

Microbes play a critical role in regulating the size, composition, and turnover of dissolved organic matter (DOM), which is one of the largest pools of carbon in aquatic ecosystems. Global change may alter DOM-microbe associations with implications for biogeochemical cycles, although disentangling these complex interactions remains a major challenge. Here we develop a framework called Energy-Diversity-Trait integrative Analysis (EDTiA) to examine the associations between DOM and bacteria along temperature and nutrient gradients in a manipulative field experiment on mountainsides in contrasting subarctic and subtropical climates. In both study regions, the chemical composition of DOM correlated with bacterial communities, and was primarily controlled by nutrients and to a lesser degree by temperature. At a molecular-level, DOM-bacteria associations depended strongly on the molecular traits of DOM, with negative associations indicative of decomposition as molecules are more biolabile. Using bipartite networks, we further demonstrated that negative associations were more specialized than positive associations indicative of DOM production. Nutrient enrichment promoted specialization of positive associations, but decreased specialization of negative associations particularly at warmer temperatures in subtropical climate. These global change drivers influenced specialization of negative associations most strongly via molecular traits, while both molecular traits and bacterial diversity similarly affected positive associations. Together, our framework provides a quantitative approach to understand DOM-microbe associations and wider carbon cycling across scales under global change.