Spatial Variation in Stream-Water Bacterial Communities Across Sites Differing in Wild Chinese Giant Salamander...

Wildlife monitoring and environmental microbiome profiling provide complementary perspectives for assessing habitats of protected aquatic species. This study integrated long-term computer vision monitoring, 16S rRNA amplicon sequencing, and water-quality measurements to examine stream sites differing in the camera-confirmed detection frequency of wild Chinese giant salamanders (Andrias davidianus). Monitoring from July 2024 to January 2026 recorded 47 independent detection events, of which 35 occurred at S1. S1 was therefore designated as the high-frequency detection site, while S2-S6 served as comparison sites. Stream-water bacterial communities were characterized from one composite sample per site, together with measurements of water temperature, pH, dissolved oxygen, conductivity, ammonia nitrogen, and nitrate nitrogen. S1 differed from several comparison sites in bacterial taxonomic composition, alpha-diversity estimates, and site-specific ASVs. Comamonadaceae, Rhodoferax, and Flavobacterium showed relatively high abundances at S1. However, RDA, envfit, Mantel, partial RDA, and variation-partitioning analyses found no significant independent relationship between the measured water-quality variables and bacterial-community variation. The observed spatial patterns likely reflect interactions among water quality, hydrology, substrate, riparian shading, refuge availability, organic-matter inputs, and anthropogenic disturbance. These results establish an exploratory spatial baseline and support the use of stream-water bacterial communities as a complementary component of Chinese giant salamander habitat monitoring.