Sediment microbial communities play critical roles in the health of fish and the biogeochemical cycling of elements in aquaculture ecosystems. However, the response of microbial communities to temporal and spatial variations in interconnected aquaculture pond and ditch systems remains unclear. In this study, 61 sediment bacterial samples were collected over one year from 11 sites (including five ponds and six ditches) in a 30-year-old fish aquaculture farm. The 16S rRNA approach was used to determine the relative abundances of microbial communities in the sediment samples. The relationships among nutrients, heavy metals, and abundant microorganisms were analyzed. Our results showed that Proteobacteria, Bacteroides and Chloroflexi were the predominant phyla in the sediments of aquaculture pond, with average abundances of 36.33%, 18.60%, and 14.58%, respectively. The microbial diversity in aquaculture sediments was negatively correlated (P < 0.05) with the concentrations of total nitrogen and total phosphorus in sediments, indicating that the microbial diversity is highly associated with the remediation of nutrients in sediments. The sediment samples with high similarities were discovered by the t-distributed stochastic neighbor embedding (t-SNE) method. The site-specific correlations between specific microorganisms and heavy metals were explored. The network analysis revealed that the microbial diversities in aquaculture ponds were more stable than that in aquaculture ditches. The network analysis also illustrated that the microbial genera with low relative abundances may become key groups of microbial communities in sediment ecosystems. Our work deepens the understanding of the relationships between microbial communities and the spatiotemporal characteristics of surface water and sediments in aquaculture farms.
Keywords: Aquaculture pond; Microbial community; Sediment; Spatiotemporal variations; T-distributed stochastic neighbor embedding.
Copyright © 2021 Elsevier B.V. All rights reserved.