Influence of seasonality, wastewater treatment plant process, geographical location and environmental parameters on bacterial community selection in activated sludge wastewater treatment plants treating municipal sewage in South Africa.

ABSTRACT

This is the first comprehensive study that focusses on the correlation between the bacterial community composition and a range of previously identified selective criteria in activated sludge wastewater treatment plants on the African continent. Multivariate statistical analyses were used to determine the relative significance of the geographical location (factor site), wastewater treatment plant process (factor configuration), seasonality (factor season), and environmental parameters on the bacterial communities in nine wastewater treatments plants from two sites in South Africa using terminal restriction fragment length polymorphism as a screening tool to rationalize the number of samples (to 50 samples) for high throughput (Illumina MiSeq) sequencing. Site was the most significant factor (Global ANOSIM R value = 0.91, p = 0.001), and it was established that the inter-site differences were not climatic in origin but related to differences in the composition of the influent and activated sludge. Previous studies that have reported associations between microbial community structure and environmental parameters have measured influent chemistry, and this is the first time, to our knowledge, that the comprehensive chemical character of activated sludge itself has been included in this type of study. It was found using BEST analysis that the activated sludge ammonia, activated sludge total phosphate and influent chemical oxygen demand were the most significant (p < 0.001) drivers for inter-site bacterial community selection (ANOSIM Global R values of 0.862, 0.782 and 0.428, respectively). This link would not have been established with only influent chemical analyses as there was no significant difference (t-test, p > 0.05) in the average influent phosphate concentrations between the 2 sites, but there was a highly significant difference (p < 0.001, t (15.5)>t-crit (2.01)) in the activated sludge total phosphate concentrations (20.8 ± 17.0 and 127.8 ± 40.2 mg/L). This is notable for all future studies on a global level aimed at identifying factors for selection of microbial communities in activated sludge.