Flooding variations affect soil bacterial communities at the spatial and inter-annual scales.

Hydrological variations have substantial effects on the diversity and composition of soil bacterial communities in wetlands. At the spatial scale, the responses of soil bacterial diversity and composition to hydrological variations in wetlands have been extensively investigated. However, at the temporal scale, especially at the inter-annual scale, the corresponding bacterial responses are rarely reported. Therefore, we explored the effects of flooding variations on the diversity and composition of soil bacterial communities at a lakeshore wetland in two hydrological contrasting years. Three flooding variables, i.e. flooding duration (FD), total duration of the growing season (TGD), and exposure duration of the growing season (EGD), were used to characterize flooding regime. Soil bacterial communities were determined using 16S rRNA gene sequencing method. We found a very high soil bacterial diversity at the lakeshore wetland. The Shannon's indexes of soil bacterial communities varied from 5.61 to 7.11 in two years. Soil bacterial α-diversity followed a unimodal curve along the elevation gradient, and was significantly lower in the flooding year than in the drought year. Principal coordinate analysis demonstrated that the compositions of soil bacterial communities were separated in order of elevation and year along the first and second axes, respectively. The apparent habitat preferences of soil bacterial families were closely connected with their respiratory traits, and this trend was stronger at the inter-annual scale than at the spatial scale. Soil bacterial compositions were predominantly determined by the direct (by changing respiratory traits) and indirect (by changing soil pH) effects of TGD at the spatial scale, while they were simultaneously regulated by the direct effects of three flooding variables at the inter-annual scale. Our results enhance the understanding of soil microbial communities in wetlands and have large implications for developing general theories to predicting soil microbial functions.