Water eutrophication is a global ecological issue, and thermal stratification of water bodies can enable eutrophication. We examined bacterial communities in the stratified water columns and sediments in two different trophic reservoirs along the Wujiang River using quantitative real-time PCR and high-throughput sequencing. Bacterial 16S rRNA gene copies varied from 3.70 × 10⁷to 5.27 × 10⁸copies/L in the water column of Hongjiadu (HJD) Reservoir (60 m water depth) with slightly stratified variation; while in Wujiangdu (WJD) Reservoir (70 m water depth), bacterial abundance decreased markedly from the surface to the bottom(1.74 × 10⁹to 2.38 × 10⁷copies/L). The vertical distribution patterns of bacteria in both reservoirs resembled those of water Chlorophylla(Chla) concentrations. The abundance was negatively correlated with water depth (D), total nitrogen (TN), nitrate (NO₃^–-N), and positively correlated with water temperature (T) and dissolved oxygen (DO) level. In contrast, the alpha diversity of bacteria showed the opposite trend in the vertical water column.Proteobacteria,Actinobacteria, andBacteroideteswere the predominant phyla in the water column of both reservoirs. Compared to WJD Reservoir, HJD Reservoir displayed marked vertical spatial difference in bacterial community structure during thermal stratification. In particular,Pseudomonaswas frequently detected at the bottom of the HJD Reservoir. These results were consistent with predictive metagenomic profiling that revealed different vertical functional variation patterns of the bacterial communities in the two reservoirs. The bacterial community structure of HJD Reservoir was associated with water D, ammonium (NH₄⁺-N), nitrite (NO₂^–-N), and total phosphorus (TP). The community structure of WJD Reservoir was related to water T, Chla, NO₃^–-N, and TN. The findings highlighted the important roles played by thermal stratification and nutrients in shaping the water bacterial community structure. Additionally, the absolute abundance of water nitrifiers (AOBgene copies) and denitrifiers (narG,nirS,norB, andnosZgene copies) displayed significant vertical differences in the water columns of both reservoirs. Gene copies involved in denitrification were significantly higher than those involved in nitrification. Water phosphorus and nitrogen contents were important variables influencing the absolute abundance of ammonia oxidizers and denitrifying bacteria, respectively. Our study revealed that the emergence of thermal stratification was responsible for the vertical stratification of bacteria in water and affected the bacterial community structure together with nutrients.

Sedimentary microorganisms play crucial roles in maintaining the functional stability of aquatic ecosystems. However, their taxonomic composition and assembly processes are not well known in estuarine–coastal margins because of their complex environment. We investigated microbial communities, co-occurrence relationships, and underlying mechanisms in 33 surface sediment samples collected in the Jiulong River Estuary and the Taiwan Strait to reveal their composition dynamics. The abundance, diversity, and composition of microorganisms demonstrated obvious spatial variables. Methanobacterium and Methanosarcina, as well as Candidatus_Nitrosopumilus and Nitrososphaeraceae were the main methanogenic and ammonia-oxidizing archaea, with an average abundance of more than 5.91% and 4.27%, respectively. Along with a salinity gradient increase, the relative abundance of methanogenic archaea (from 42.9% to 16.6%) contrasted with the trend of ammonia-oxidizing archaea (from 6.04% to 18.7%). The number of methanogenic archaea gradually decreased with increasing geographic distance (p < 0.05), whereas ammonia-oxidizing archaea showed no significant change (p > 0.05). In co-occurrence patterns, closer inter-taxa connections were observed among archaea–archaea and bacteria–bacteria than in archaea–bacteria, which indicated that coexistence within the same kingdom was greater than interaction between different kingdoms in shaping the community structure along the salinity gradient. Furthermore, null model analyses of the microbial community showed that undominated was the most prominent process, explaining over 44.9% of community variation, followed by heterogeneous selection and dispersal limitation, which contributed to 27.7% and 16.3%, respectively. We demonstrated that stochasticity, rather than determinism, regulates community assembly. These results further highlight that intra-kingdom co-occurrence and stochastic processes shape the structure and assembly of microbial communities in estuarine–coastal margins.