摘要:Physical Habitat Simulation (PHABSIM) is used to predict the impact of flow regime alterations on habitat availability of fish. Species-specific predictions of habitat availability for fish are often made by applying generic habitat suitability curves (HSC), thereby implicitly assuming that HSCs are fixed among rivers. We tested the hypothesis that HSCs for a widely-distributed North American minnow, the longnose dace (Rhinichthys cataractae), were similar in two rivers in western Canada. Utilization curves differed between rivers for all three parameters, but the mechanism producing these differences was different. Although transferability tests provide an objective method of assessing whether an existing curve is applicable to a new river system, our findings indicated the tests often failed because the test statistic was not applicable to the raw data. R. cataractae populations displayed large variations in HSC for water depth, water velocity, and substratum composition between the study rivers and previously published data. We conclude that the ability of PHABSIM to predict available habitat for a forage fish such as R. cataractae will be low unless the HSC are generated in the river systems where the model is to be applied.