Date of Award

January 2014

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Biology

First Advisor

Steven Kelsch

Abstract

The relationships between fish species guilds, riparian cover, and vegetation disturbances in the surrounding landscape were examined across the 11 western tributaries of the Red River of the North. Archival stream sampling data, collected from 1993-2011 by North Dakota state agencies, were analyzed relative to temporally-appropriate land-cover predictors generated from National Land Cover Database and National Agricultural Imagery Program products.

The 0-30 m riparian cover width was the most influential landscape predictor influencing fish structure. The 0-30 m riparian cover displayed interactive effects with 30-50 m riparian cover width and watershed land-cover disturbance. These riparian scales were identified by a PCA of intact riparian area, determined from digitized 1m remotely sensed images. Tolerant and omnivorous species guilds had higher percent compositions where riparian cover in the 0-30 m scale was degraded. Conversely, insectivorous and benthic insectivorous species guilds had higher percent compositions where the 0-30 m riparian cover was more intact. Although suspended sediment loading resulting from riparian disturbance is suspected as a potential mechanism for the riparian effect, the limits of the 0-30 m riparian scale are recognized. The 0-30 m riparian scale is presently a proxy variable, as the results identify a structural relationship with the landscape and assumes mechanisms.

The investigation of riparian scaling also has implications for the incorporation of riparian effects into fisheries landscape analysis. Relationships between fish communities and riparian integrity or riparian composition have been reported at a variety of arbitrarily selected scales. To test the effects of generalizing riparian scale, a 0-50 m riparian scale was used rather than the 0-30 m scale determined to be the most important. The more general scale displayed slightly different relationships than were shown to exist. Caution should therefore be exercised if arbitrarily selecting riparian scale widths for fisheries landscape analysis.

Download

Share

COinS