| Abstract: | ABSTRACT: Forty‐six independent stream reaches in southeastern Pennsylvania were surveyed to assess the relationships between geomorphic and habitat variables and watershed total impervious area (TIA) and to test the ability of the impervious cover model (ICM) to predict the impervious category based on stream reach variables. Ten variables were analyzed using simple and multivariate statistical techniques including scatter‐plots, Spearman's Rank correlations, principal components analysis (PCA), and discriminant analysis (DA). Graphical analysis suggested differences in the response to TIA between the stream reaches with less than 13 percent TIA and those with greater than 24 percent TIA. Spearman's Rank correlations showed significant relationships for large woody debris and sinuosity when analyzing the entire dataset and for depth diversity and the standard deviation of maximum pool depths when analyzing stream reaches with greater than 24 percent TIA. Classification into the ICM using DA was 49 percent accurate; however, the stream reaches did support the ICM in other ways. These results indicate that stream reach response to urbanization may not be consistent across geographical regions and that local conditions (specifically riparian buffer vegetation) may significantly affect channel response; and the ICM, used in the appropriate context, can aid in the management of stream reaches and watersheds. |
