The use of Habitat Suitability Index (HSI) models for fisheries impact assessment has not been field validated for most fish species. This research sought to test several of the fundamental assumptions inherent in applications of Suitability Index (SI) curves with the Habitat Evaluation Procedures (HEP) of the u.s. Fish and Wildlife Service. Using draft SI's for rock bass (Ambloplites rupestris), tests for positive correlations between standing stocks and measures of fish habitat were conducted in 1982-1983. Multiple regression modeling and analysis of fish production were used to further evaluate the efficacy of habitat assessments for rock bass based on HSI models.

Highly significant positive correlations (p<.01, R²=.68) between Habitat Units (HU's) and biomass were observed in Back Creek for November 1982. However, significant positive correlations were not observed consistently for other months, models, or streams evaluated. Factors which apparently accounted for low correlations included seasonal fish movements, fish sampling biases, habitat homogeneity, and the omission of potentially important variables, stitch as depth and cover, from HSI models. Fish production was essentially the same in Little Walker and Back Creeks and averaged 1.04 gm⁻²yr⁻¹. The factors influencing correlations between fish abundance and habitat measurements prevented meaningful comparisons of rock bass production and the related physical habitat.

Results from this study suggest that summer low flow periods may not always be most limiting to rock bass habitat in Virginia streams. Furthermore, seasonal movements of fish and sampling biases, if not considered, may seriously confound attempts to relate indices of carrying capacity and habitat model outputs. Incorporating the annual home range of the evaluation species into future study designs is recommended to expedite subsequent attempts to validate HSI models.