Fish Species of Greatest Conservation Need in Wadeable Iowa Streams: Current Status and Effectiveness of Aquatic Gap Program Distribution Models

Effective conservation of fish species of greatest conservation need (SGCN) requires an understanding of species– habitat relationships and distributional trends. Thus, modeling the distribution of fish species across large spatial scales may be a valuable tool for conservation planning. Our goals were to evaluate the status of 10 fish SGCN in wadeable Iowa streams and to test the effectiveness of IowaAquatic Gap Analysis Project (IAGAP) species distribution models. We sampled fish assemblages from 86 wadeable stream segments in the Mississippi River drainage of Iowa during 2009 and 2010 to provide contemporary, independent fish species presence–absence data. The frequencies of occurrence in stream segments where species were historically documented varied from 0.0% for redfin shiner Lythrurus umbratilis to 100.0% for American brook lamprey Lampetra appendix, with a mean of 53.0%, suggesting that the status of Iowa fish SGCN is highly variable. Cohen’s kappa values and other model performance measures were calculated by comparing field-collected presence–absence data with IAGAP model–predicted presences and absences for 12 fish SGCN. Kappa values varied from 0.00 to 0.50, with a mean of 0.15. The models only predicted the occurrences of banded darter Etheostoma zonale, southern redbelly dace Phoxinus erythrogaster, and longnose dace Rhinichthys cataractae more accurately than would be expected by chance. Overall, the accuracy of the twelve models was low, with a mean correct classification rate of 58.3%. Poor model performance probably reflects the difficulties associated with modeling the distribution of rare species and the inability of the large-scale habitat variables used in IAGAP models to explain the variation in fish species occurrences. Our results highlight the importance of quantifying the confidence in species distribution model predictions with an independent data set and the need for long-term monitoring to better understand the distributional trends and habitat associations of fish SGCN.