Abstract: To investigate the characteristics and seasonal variations of the aquatic food web in the Heishui River Basin, fishery resources were collected from different river sections in July (wet season) and November (dry season) of 2024. Based on the carbon and nitrogen stable isotope values of the main consumers and prey organisms, the SIBER model was used to analyze the seasonal differences in fish trophic levels and community trophic structure. The δ¹³C values of organisms ranged from -20.45‰ to -5.27‰ in the wet season and and from -29.40‰ to -1.12‰ in the dry season; δ¹⁵N values ranged from 0.88‰ to 14.68‰ in the wet season and from 0.87‰ to 12.66‰ in the dry season. The trophic level of fish ranged from 1.817 to 3.918, with no significant overall difference between the wet and dry seasons (P>0.05). However, significant seasonal variations occurred in certain species: the average trophic level of Abbottina obtusirostris and Zacco platypus was significantly higher in the wet season (P<0.001), whereas that of Schistura fasciolatawas was significantly higher in the dry season (P<0.01). These findings reflect species-specific adaptations to seasonal hydrology and food resource changes. Trophic structure indices (NR, CR, TA, CD, MNND, and SDNND) were generally higher in the dry season, indicating a more diverse basal food source, more distinct trophic niche differentiation, higher community diversity, weaker interspecific trophic competition, and lower trophic redundancy during the dry season. These results provide a case reference for food webs research in mountain river ecosystems and offer a theoretical basis for ecosystem health assessment and targeted conservation management of the Heishui River.