Degree Name

Master of Science (MS)

Semester of Degree Completion

1997

Thesis Director

Robert U. "Bud" Fischer

Abstract

In agricultural watersheds, channelization and the conversion of native vegetation into agricultural crops contribute to a loss or simplification of near-stream vegetation and a reduction in the complexity of the physical stream environment. Since changes to near-stream vegetation can have a significant effect on the stream biological community, this study investigated the relationship between a habitat quality index and stream fish assessment indices used to evaluate stream quality. Fish data and stream habitat quality data were collected from 13 sampling localities associated with the Embarras River basin. Habitat quality was measured using the Stream Habitat Assessment Procedure (SHAP), while stream fish quality was determined by species richness and the Index of Biotic Integrity (IBI). A positive linear correlation was observed between SHAP and IBI (r2 = 0.409, n = 13, p < 0.05) for the 13 sites sampled. However no significant correlation was observed between SHAP and species richness (r2 = 0.195, n = 13, p = 0.13). Multiple stepwise forward regression analysis was performed to determine which SHAP variables would be the best predictors of stream biotic integrity as measured by IBI and species richness. Two significant variables (pool quality and bank vegetation) were used in creating a model for predicting IBI in which 76% of the variance in IBI was explained by these SHAP variables. In addition, a model which explained approximately 84% of the variance in species richness among sites was created using the four significant SHAP variables of pool quality, pool variability, deposition and canopy cover. The SHAP variables included in these models are related to features such as riparian zone vegetation and channel morphology, and play an important role in creating and maintaining the key qualities of the stream ecosystem. Models such as those suggested in this paper demonstrate the potential for predicting the biotic integrity of a stream fish community from commonly collected, and often readily available, habitat data. Thus, the models may be an important management tool which will allow for the rapid prediction of the biotic integrity of a stream, and thus permit intensive management practices to be focused on critical sites within a stream basin.

Download

Share

COinS