Impacts of environmental factors on the food web structure, energy flows, and system attributes along a subtropical urban river in southern China.

Tropical and subtropical rivers are being subjected to multiple stressors from human disturbance (e.g., water pollution and habitat degradation). Understanding the relationship between environmental conditions and the river ecosystem is important for improving river management. We built 14 Ecopath models composed of 28 functional groups (trophic levels [TLs] of 1.0-3.8) along a subtropical urban river to explore the influence of environmental changes on system attributes. From headwaters to downstream, the model outputs showed that the transfer efficiency (TE), energy flow parameters, and ecosystem theory indices exhibited significant (P < 0.05) differences across a longitudinal gradient of disturbance, indicating heterogeneous attributes of local river segments. The high TE values of TLs I, II, and III separated the upper, middle, and lower reaches, respectively, which could be attributed to the shift in dominant consumption flows from upstream 'periphyton - aquatic insects - insectivorous fish' to midstream 'detritus - shrimp - crustaceavorous fish' and to downstream 'phytoplankton - filter-feeding invertebrates/fish'. Structural equation modelling was used to test the causal relationships among environmental variables and demonstrated that abiotic factors directly influenced biomass composition and indirectly influenced trophic networks. Water quality, including dissolved oxygen and flow velocity; habitat characteristics, such as riffles, cobble-gravel substrate, and seasonal floodplain; and biological indicators, including the relative contributions (%) of decapods, insectivorous fish, and insect scrapers to biomass composition, had significant (P < 0.05) positive impacts on system maturity (evaluated by omnivory, connectance, and cycling indices). In the future, it will be possible to evaluate the health of river ecosystems by monitoring representative environmental factors, which could be a cost-effective approach to system-level improvement.