Fish isotopic niches associated with environmental indicators and human disturbance along a disturbed large subtropical river in China

Abstract

Stable isotopes are increasingly used to detect and understand the impacts of environmental changes on riverine ecological properties. The δ13C and δ15N signatures of fish with different feeding habits were measured in a large subtropical river to evaluate how fish isotopic niches respond to environmental gradients and human disturbance. From basal resources to fish consumers, the high values of epilithic periphyton (biofilm) δ13C and suspended particulate organic matter δ15N concurrently determined the niche ranges and space (e.g., convex hull area) of fish communities. Along a longitudinal gradient (except in the industrial zone), the number of fish trophic guilds identified by Bayesian ellipses continuously increased; meanwhile, higher trophic diversity and less redundancy were observed near the lower reaches and estuary. Variance inflation factors were estimated to detect the multicollinearity of 40 environmental variables, 14 of which were selected as indicators. Relative importance (RI) analysis was used to evaluate the explanatory power of these indicators for the spatial variation in isotopic niche metrics; the results showed that riffle habitat area, water nitrate concentration, gravel-cobble substrate, and riparian buffer width were the 4 key environmental indicators (average RI > 12%) that determined the longitudinal pattern of fish isotopic niches. These findings suggested that community-level δ13C signatures are more responsive to changes in habitats (e.g., riffle) and substrates (e.g., gravel-cobble) supporting the productivity of autochthonous diatoms while δ15N signatures respond to water quality altered by nitrogen pollution from manure-fertilized farming and poultry livestock effluent. Furthermore, δ15N may be more robust and interpretable than δ13C as an isotopic indicator of ecosystem change in rivers exposed to multiple or complex anthropogenic stressors.