Redband trout density was examined in three forested streams in central Oregon at two scales, pool channel unit and microhabitat. Two streams were in roadless areas and one was in a "highly managed" watershed. At the larger spatial scale, trout displayed a seasonal shift in habitat use from early to late summer. There was a positive correlation between trout density and pool structural complexity during summer base flow. The association was intensified throughout the summer as stream flow continued to drop. The structural complexity of each pool was quantified using an index integrating structural variability and depth. Twenty-two pools were divided into microhabitats, or pool subunits with similar characteristics using a qualitative classification scheme describing different structural elements comprising the pool habitat. Microhabitat was partitioned between fry and older trout throughout the summer: Fry generally used stream margins, backwaters, and shallow areas; whereas, trout one year and older used deep areas (depth greater that 0.5 m) and cover associated with substrate and wood. The use of cover by trout one year and older doubled from June to August. This change in use was coupled with the increased association with structural complexity at the channel unit scale. The large substrate and wood that provided cover also increased the structural complexity of the pool. As trout increased their use of cover, their densities increased in more complex pools. Man-made log weir pools in a simplified stream were evaluated for their structural complexity and compared to the shallow natural pools in the same reach. Log weir pools had greater average depth, but were less complex than natural pools, and could maintain a similar density of trout. The structural complexity in natural pools appears to compensate for their shallow depth. Log weir pools enhanced with placement of a rootwad supported higher densities of trout. Addition of a rootwad provided microhabitats associated with cover that were lacking in weir pools without other structures added. When using instream construction to create pool habitat, complex structure that provides microhabitats associated with cover is more effective at holding higher densities of fish under late summer low flow conditions. / Graduation date: 1996

Habitats for juvenile chinook salmon (Oncorhynchus tshawytscha) were observed at a high flow during the spring of 1989 and at a low flow during the spring of 1991 in the lower American River of California. Parameters of microhabitats in eight macrohabitats were measured. Mean column water velocity and total water depth variables were used to statistically compare juvenile chinook salmon microhabitat distributions between and among macrohabitat type/reach and flow combinations. Using mean column water velocity and total water depth as independent variables, each macrohabitat type/reach appeared to be unique. Mean column water velocity seemed to be a better descriptor of juvenile chinook salmon habitat selection than was total water depth. Riffles appeared to be the preferred macrohabitats for large non-schooling chinook salmon at both flow levels. Root wads, woody debris, and submerged terrestrial vegetation were utilized extensively and provided a significant cover for schooling chinook salmon during the high flow period. Habitat utilization by chinook salmon appears to be influenced by nonhydraulic variables. Application of models in current instream flow studies need to include factors such as water temperature and aquatic insect production. / Graduation date: 1993

Copies of author's previously published articles inserted. / Bibliography: leaves 317-359. / xxii, 360 leaves : ill. (some col.), maps (some col.) ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Examines the generality of the Flood Pulse Concept as a model for the role of flow in large floodplain rivers of all climatic zones and reviews the literature on hydrology-biology relations, with particular reference to the ecology of riverine fish. Using techniques of ordination, clustering and analysis of similarities (ANOSIM), examines the relations between flow variability, econology and climate in large rivers worldwide. Using a fie-year database for Cooper Creek in Central Australia, relates structures of fish, macroinvertebrate and zooplankton assemblages and indices of fish health and behaviour to hydrology at several spatial and temporal scales. These relations are established using multivariate techniques, univariate correlation and regression, and neural networks modelling. / Thesis (Ph.D.)--University of Adelaide, Dept. of Zoology, 1999

Copies of author's previously published articles inserted. / Bibliography: leaves 317-359. / xxii, 360 leaves : ill. (some col.), maps (some col.) ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Examines the generality of the Flood Pulse Concept as a model for the role of flow in large floodplain rivers of all climatic zones and reviews the literature on hydrology-biology relations, with particular reference to the ecology of riverine fish. Using techniques of ordination, clustering and analysis of similarities (ANOSIM), examines the relations between flow variability, econology and climate in large rivers worldwide. Using a fie-year database for Cooper Creek in Central Australia, relates structures of fish, macroinvertebrate and zooplankton assemblages and indices of fish health and behaviour to hydrology at several spatial and temporal scales. These relations are established using multivariate techniques, univariate correlation and regression, and neural networks modelling. / Thesis (Ph.D.)--University of Adelaide, Dept. of Zoology, 1999