Although the American eel Anguilla rostrata occurs in a variety of habitats over large geographic areas, little is known regarding the specific habitat relations that regulate eel distribution and abundance in freshwater streams. We evaluated the importance of 17 physical habitat, chemical, and biological variables in predicting eel density in five major river basins in Maryland. Because artificial structures impede eel migration in all five basins, only sites determined to be on unblocked streams or downstream of structures that significantly restrict eel passage were used in the analysis. Stepwise regression identified a model consisting of four variables—velocity–depth diversity, the log‐transformed distance (km) to the Chesapeake Bay, the log‐transformed density of noneel fishes, and the distance to a semipassable or impassable structure—as the best predictor of eel densities. When applied to a random subset of data not used for model development, the model correctly predicted 44.4% of eel densities and had a high mean prediction error (1.86 eels/75 m2) compared with the mean eel density for all sites included in model development (3.33 eels/75 m2). However, the Pearson correlation coefficient between the observed and predicted eel densities (r = 0.64; P < 0.05) indicated that the model was generally successful in predicting trends in observed densities. The study results are consistent with previous research suggesting the lack of significant eel–stream habitat associations, although the presence of a variety of velocity–depth regimes may support higher eel densities. Higher eel densities near artificial structures known to affect upstream movement suggest that eels are accumulating downstream of these structures because of impeded migration.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. The University of Notre Dame is collaborating with JSTOR to digitize, preserve and extend access to American Midland Naturalist.ABSTRACT: Discriminant function analysis was used to quantify both microhabitat (within-cave) and macrohabitat (among-cave) locations selected by five species of cave-dwelling bats in the limestone caves and mines of Maryland, Pennsylvania and West Virginia. Microhabitat parameters reflected microclimate, physical structure and behavioral interactions at the roosting site. Macrohabitat parameters described cave macroclimate, structure, size and external habitat features. Three discriminant functions accounting for more than 95 % of the variance were derived in the microhabitat analysis. Function I was correlated with microclimate and the tendency to cluster, function II with the exposure of the site, and function III with the draftiness of the location. There was a large degree of overlap along many of the univariate parameters, particularly temperature. Multivariate classification of the five species was 63.5% correct. Parameters that were important on a macrohabitat scale differed among species. Zones of suitable microhabitat available within a cave or mine were important factors, but presence or absence of a species at a cave or mine could not be accounted for solely by differences at this level. Cave size is shown to be important to several species in this guild. Correct multivariate classifications varied by species from 82.0 to 96.0%.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.