Abstract

Summary Activated sludge ( AS ), which has been in use for 100 years, has been the most popular biological process in various wastewater treatment plants ( WWTPs ), in which bacteria plays central roles in pollutant removal. However, the potential relationship between bacteria taxa and the niches occupied by specific functional bacteria in AS are largely unknown. Here, correlation‐based network analysis was applied to a 16S rRNA gene pyrosequencing dataset containing > 760 000 sequences of 50 AS samples from globally distributed full‐scale WWTPs . The results showed that (i) bacterial assembly in AS was nonrandomly arranged by taxonomic relatedness and (ii) intra‐ and inter‐phylum/class co‐occurrence higher than expected by chance was induced by multiple deterministic processes, such as habitat filtering and competition. Moreover, based on bacterial occupancy, a prevalent core set of cosmopolitan functional bacteria (e.g. multiple nitrogen‐cycling‐related bacteria) was widely distributed in the AS of different WWTPs , showing strong ecological associations among them. Additionally, the AS network has statistical and structural characteristics similar to those of previously reported ecological networks, such as power‐law connectivity distribution and nonrandomly connected properties. Overall, this work provides novel insights into the bacterial associations within AS and sheds light on the ecological rules guiding bacterial assembly in WWTPs .