Publication Date


Degree Program

Department of Biology

Degree Type

Master of Science


The composition of lotic assemblages is influenced by many landscape factors. Foremost among these are basin geomorphology, fluvial and substrate characteristics, stream size, and land use. The primary purpose of this investigation was to determine if fish assemblages within Kentucky's Green River and Tradewater River Basins were associated with local scale environmental gradients and, if so, which gradients were the most important in shaping fish assemblages and which fish species were most closely associated with these gradients. A secondary purpose was to determine which regionalization or characterization strategy best segregated streams within the study area according to fish assemblages. Eighty-eight sites within the Green River and Tradewater River Basins were sampled with seine and backpack electroshocker between August 2001 and March 2002. Sampling resulted in the capture of 83 species. A suite of environmental characteristics was estimated or measured at each site. Fish and environmental data were tested for significant relationships with canonical correlation analysis (CCA). Further analyses for the effect of environmental gradients on individual species were done with Pearson correlations. Species data were exposed to detrended correspondence analysis (DCA) to analyze the effect of geographic and local scale classification strategies. CCA analysis indicated there were significant relationships between species and environmental data for both CCA Axis 1 and 2. Further analysis revealed that stream size and substrate composition had the strongest effect in structuring fish assemblages within the study area. Pearson correlations suggested species with special habitat needs were those mainly influenced by local environmental gradients, and thus, were most influential in the structuring offish assemblages. Habitat quality was higher in the eastern portion of the study area where high gradient streams were most common. This may be a reflection of expected upstream to downstream processes or of differing land use due to topography. Larger streams were found to generally support a greater diversity of fish species than smaller streams. DCA analysis of classification strategies indicated fish assemblages best grouped by combinations of hydrologic units (sub-drainages) within the study area. These groupings followed a general east to west trend. Further investigation is needed to separate the effects of natural riverine processes from that of anthropogenic land use.


Medical Sciences