River flow regulations have great impact on the downstream aquatic ecosystem. It is important to investigate the response of target species to the changes in hydrological regimes so that we can explore possible remediation measures. Ecohydraulics models that integrate hydrodynamics processes and ecological processes have been shown to be efficient in achieving these objectives.This study developed an integrated model which combined a two-dimensional hydrodynamic module with a vegetation evolution module and a fish habitat module. Owing to the ability to represent spatial heterogeneity and local interactions, the vegetation module used an unstructured cellular automata (UCA) approach. To describe the ambiguous relations between the physical conditions and habitat suitability, a fuzzy inference method was applied to the fish habitat module. The developed model was applied to a compound channel of the Lijiang River in southwest China, which has been greatly affected by the flow regulations of the Qingshitan Reservoir for navigation purposes. Through scenario simulations, the effects of flow regulation on riparian vegetation and fish habitat were analyzed. According to the results, water releases in the dry season imposed negative effects on the downstream semi-aquatic plant Rumex maritimus (R. maritimus) and Polygonum hydropiper (P. hydropiper) and favored the upland species Leonurus heterophyllus (L. heterophyllus). Regarding to the effects on fish Spinibarbus hollandi (S. hollandi) the results showed that water releases increased the suitability of the spawning conditions, especially during wet and dry years, but had little impact on the overwintering conditions. Key words 9 9 9 9 cellular automata, fuzzy inference, habitat modelling, river ecosystem, river regulation
ABBREVIATION AND NOTATION