Wastewater treatment plant upgrade drives microbial succession and niche expansion of nitrogen-cycling bacteria in a receiving river: Insights from interpretable machine learning

Wastewater treatment plant (WWTP) upgrades improve effluent quality, yet their long-term effects on receiving-water microbiomes remain unclear. We conducted a long-term pre- and post-upgrade monitoring at the Gaobeidian WWTP in Beijing and its receiving river (five sites along the upstream-outfall-downstream gradient). Integrating water-quality parameters, 16S rRNA gene sequencing, and metagenomic profiles, we applied interpretable AutoML and community assembly modeling to investigate the ecological responses to the upgrade. The upgrade shifted microbial communities from heterotrophic degradation dominance toward a methylotrophy-associated partial denitrification regime. Increased dissolved oxygen and oxidation-reduction potential after ozonation coincided with pronounced enrichment of Methylotenera at the effluent outfall and downstream sites, particularly in winter, whereas Dokdonella , Dechloromonas , and Bacillus declined. The expansion of methylotrophic bacteria may enhance nitrogen transformation, while nitrogen-cycling gene profiles suggest that incomplete denitrification may also increase the potential risk of N 2 O accumulation. AutoML identified high oxidation-reduction potential, low nutrient availability, and low temperature as key factors associated with Methylotenera niche expansion. Community assembly also showed reduced consistency with neutral expectations and stronger signals of environmental filtering. Overall, this study links WWTP-upgrade-induced water-quality reshaping with microbial succession, nitrogen-cycling pathway shifts, and community assembly transitions, providing a microbial ecological basis for assessing advanced treatment impacts and managing watershed-scale water environmental risks.