在难降解废水处理系统中的长期适应过程中,细菌组装和关键分类单元对于优化系统性能和改进管理策略至关重要。因此,这项研究采用了高通量测序,再加上机器学习模型和统计分析方法,全面阐明农药废水处理系统中细菌群落和组装过程的关键特征。为期9个月的监测显示,两个相互连接的生物处理单元(称为BA和PA)之间的细菌群落多样性和组成存在很大差异。两个单位的细菌群落动态相似。此外,水质在调节活性污泥的细菌群落结构中起着至关重要的作用,主要是由确定性模式驱动的。同质选择贡献了BA和PA样品的62.85%和64.43%的变异,分别。此外,网络分析显示,两组的细菌组成具有显着的模块化。线性回归分析确定了与代谢和降解功能相关的主要细菌模块。值得注意的是,PA样品中的模块2与核苷酸代谢相关的功能具有显著正相关,氨基酸,和外源性物质,以及苯甲酸盐降解。此外,随机森林模型确定的生态模块中的关键分类单元,如假单胞菌,鞘氨醇,和PHOS-HE28是具有代谢和降解功能的优势种群。特别是,鞘氨醇,似乎是一种潜在的多功能降解细菌,与氨基酸和异源生物代谢有关,以及脂肪酸,缬氨酸,亮氨酸,异亮氨酸,氟苯甲酸酯,和氨基苯甲酸酯降解。这些发现对于制定运行策略以在难降解废水处理过程中保持稳定的系统性能非常重要。
Bacterial assembly and key taxa during long-term
acclimatization in refractory wastewater treatment systems is of paramount importance for optimizing system performance and improving management strategies. Therefore, this study employed high-throughput sequencing, coupled with machine learning models and statistical analysis approaches, to comprehensively elucidate key features of bacterial communities and assembly processes in pesticide wastewater treatment systems. A nine-month monitoring showed substantial variation in diversity and composition of bacterial community between two interconnected biological treatment units (designated as BA and PA). Dynamics of bacterial communities in both units were similar. Moreover, water quality played crucial roles in regulating the bacterial community structure of activated sludge, which were primarily driven by deterministic patterns. Homogeneous selection contributed to 62.85 % and 64.43 % of the variations in BA and PA samples, respectively. Additionally, network analysis revealed significant modularity in bacterial compositions in both groups. Linear regression analysis identified major bacterial modules associated with metabolism and degradation functions. Notably, Module 2 in PA samples has significant positive correlations with functions related to metabolism of nucleotide, amino acid, and xenobiotics, as well as benzoate degradation. Furthermore, key taxa in ecological modules identified by Random Forest model, such as Pseudomonas, Sphingobium, and PHOS-HE28, were dominant populations with metabolism and degradation functions. Particularly, Sphingobium, appeared to be a potential multifunctional degrading bacterium, related to amino acid and xenobiotics metabolism, as well as fatty acid, valine, leucine, isoleucine, fluorobenzoate, and aminobenzoate degradation. These findings are important for developing operating strategies to maintain stable system performance during refractory wastewater treatment.