Document Type

Dissertation - Open Access

Award Date


Degree Name

Doctor of Philosophy (PhD)


Wildlife and Fisheries Science

First Advisor

Charles R. Berry


Assessments of the health of rivers and streams in prairie environments would improve if the role of systemic patterns and processes among geologic-climatic settings in controlling physical habitat and fish communities were better defined. My research approach was based on a premise that assessments of the health of rivers and streams in prairie environments would benefit from studies that 1) examine the moderating effects of systemic patterns and processes by comparing physical habitat continua and fish communities among geologic-climatic settings, 2) determine the relative influence of locally interacting variables (e .g., channel shape and riparian vegetation) and systemic processes in structuring physical habitat and fish communities among a range of streams sizes, and 3) test the effects of biotic and abiotic controls on habitat partitioning by fish during critical periods of low flows common in prairie streams. My research had three complementary parts: two field studies and a laboratory study. The first field study tested the hypothesis that systemic processes moderate physical riverine environments, and thus biological communities, in distinct ways between a semi-arid region and a subhumid region. The second field study tested the hypothesis that in a subhumid region the interactions of local variables have greater influence on physical habitat and fish communities in small streams than in rivers. The laboratory study tested the influence of predators, large woody debris, and turbidity on habitat use by common minnow species under simulated low flow conditions. In the first study, longitudinal trends in channel morphometry, bankfull dimensions, stream bank and riparian traits, substrate, and fish community attributes were compared between the Bad River in a semi-arid region of western South Dakota and the Big Sioux River in the sub-humid region in eastern South Dakota. Findings suggest that along the Big Sioux River, the longitudinal changes in physical characteristics are gradual and continuous. Bankfull dimensions, channel bottom width, sand substrate, and streambank as deposition increased longitudinally while silt substrate, and percent of bank with vegetation decreased. In contrast, in the Bad River, longitudinal changes in these physical characteristics showed either a random pattern or a pattern of no change. In the Bad River, vertical channel dimensions (i.e., bank length and bank height) did increase with watershed size but not uniformly as they did in the Big Sioux River. Relationships of fish community attributes with watershed size were similar to the physical patterns. For example, in the Big Sioux River, most fish community attributes showed a continuous pattern of change either upward or downward with watershed size. In contrast, in the Bad River, most attributes show no discrete changes with watershed. My findings show that while prairie streams in sub-humid regions exhibit a truncation of the river continuum concept (RCC), physically and biologically, in semiarid regions, further truncation of the upper part of the RCC occurs. While both rivers had similarities in pioneering species in the upper parts of each river, only the Big Sioux River had a headwater component in the upper part. Also, a general randomness or lack of pattern in the physical and biological structure in the Bad River can be conceptualized as a longitudinal stretching of a reach of river into a longer segment of river. This would account for the seemingly lack of pattern in fishes. In the second study, findings indicate that variation in channel morphometry, physical habitat, and riparian-r1elated habitat decreases with watershed size in a subhumid prairie stream. Variation in channel morphometry, physical habitat, and riparian related habitat in tributaries could not be explained by local riparian conditions or adjacent land use. In fact, land use was or had been pasture, which limited comparison with other adjacent land use types (e.g., cropland). Furthermore, within pastures the level of animal vegetation use could not account for variation in riparian-related cover types among tributary sites. In this study the range of physical conditions among tributaries was coupled with a range of biological attributes. However, very few significant correlations were found between the biological and physical attributes. This suggests that the systemic controls even among small watersheds do have an influence on site-specific physical habitat and biological attributes. Several phenomena are probably responsible for this variation: hydrologic alterations caused by upland conversion to agriculture, cumulative loss of riparian buffering capacity, subtle differences in sub-watershed hydrology and geology, and flow fluctuations. In the laboratory study, low flow conditions were simulated in three experimental streams. Two suites of trials were performed: low and high turbidity trials.

Library of Congress Subject Headings

Fish communities -- South Dakota
Fishes -- Habitat -- South Dakota
Stream ecology


Includes bibliographical references (page 161-171)



Number of Pages



South Dakota State University


Copyright © 2001 Craig Lee Milewski. All rights reserved.