Despite the rapid development of research on aquatic environment microbiota, limited attention has been paid to exploring the complex interactions between microbial communities and aquatic environments. Particularly, the mechanisms underlying fish diseases based on such dynamic interactions remain unknown. This study aimed to address the gap by conducting microbiome and co-occurrence network analyses on the typical freshwater aquaculture systems. High-throughput 16S rRNA gene sequencing results revealed significant differences in the microbiota between the disease and healthy groups. Notably, disease mortality varied consistently with the gradient of relative abundance of Proteobacteria (intestine, R2 = 0.46, p < 0.05) and Cyanobacteria (gill, R2 = 0.67, p < 0.01), indicating their potential use as diagnostic criteria. Furthermore, the elevated hepatosomatic index, NO3-N, COD and TC (sediment) were directly correlated with diseases (r > 0.54, p < 0.01). Mean concentrations of NO3-N, COD and TC were elevated by 78.87%, 25.63% and 44.2%, respectively, in ponds where diseases occurred. Quantitative analysis (qPCR) revealed that Aeromonas sobria infected hosts through a potential pathway of “sediment (4.4 × 105 copy number/g)-water (1.1 × 103 copy number/mL)-intestine (1.2 × 106 copy number/g)”. Similarly, the potential route for Aeromonas veronii was sediment (4.9 × 106 copy number/g) to gill (5.1 × 105 copy number/g). Additionally, the complexity of microbial networks in the intestine, water, and sediment was significantly lower in the disease group, although no similar phenomenon was observed in the gill microbial network. In summary, these findings reveal that elevated concentrations of crucial environmental factors disrupt the linkages within microbiota, fostering the growth of opportunistic bacteria capable of colonizing fish gut or gills. This offers new insights into potential mechanisms by which environmental factors cause disease in fish.