Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Ecology and Evolutionary Biology

Major Professor

James A. Drake

Committee Members

Kenneth A. Rose, Richard J. Strange, Louis J. Gross


Riverine fishes evolved to life in a highly variable, flow-driven environment. During the two past centuries, large rivers have been substantially altered by human activities. This has resulted in declines of fish populations that depend on the large river environment.

The research described here uses models to evaluate the effects of human activities on the viability of fish populations in rivers. I focused on five modifications of the river environment associated with impoundment: (1) seasonal allocation of river flow; (2) diversion of river flow; (3) fragmentation of the river habitat by dams; (4) conversion of free-flowing river to reservoir habitat; and (5) alteration of migration patterns.

To understand the role of flow regulation on chinook salmon (Oncorhynchus tshawytscha) recruitment, I developed an individual-based model to predict recruitment as a function of seasonal flow patterns in the Tuolumne River, California. I used simulated annealing to find flow patterns that maximize chinook recruitment under wet and dry hydrologic conditions. As water availability increased, I found that the optimal flow pattern shifted from allocating low flows uniformly across seasons to a pattern with high spring flows. When I considered a new objective: maximizing the variance of spawning times among recruits, the optimal flow regime called for a winter pulse in flow just before the peak spawning date for the minority (late-fall) run.

To evaluate the recovery options for chinook salmon in the Tuolumne River, Ideveloped an age-based model to conduct a population viability analysis (PVA). I developed a flow-dependent spawner-recruitment relationship from the recruitment model. Its shape depended on the flow regime, suggesting that such relationships are not fixed properties of species, but depend on environmental conditions. The PVA model suggested that recovery, in the absence of straying, would be enhanced most by significantly reducing ocean harvest, followed by reduced diversion of water from the river.

For white sturgeon (Acipenser transmontanus) populations in the Snake River, Idaho a main concern is habitat fragmentation by dams resulting in smaller, isolated populations. Simulation experiments to evaluate the effects of fragmentation suggested that population viability was higher when dams were spaced widely enough apart to retain free-flowing habitat. A simulation experiment to evaluate the effects of altered migration patterns associated with impoundment showed that both the likelihood of persistence and the genetic diversity among white sturgeon populations were enhanced by balanced upstream and downstream migration rates.

Models that simulate the responses of fish populations to modified river habitat do not consider the potential for an evolutionary response. I designed a PVA model simulating the genetic basis of age at maturity for individual fish. Simulated individual variation in this trait lead to increased population viability only when the variation was heritable and subjected to an altered selective regime. The results support the idea that predicting population viability depends on estimating the potential for evolution in fitness-related traits for populations exposed to anthropogenic changes in the environment that impose strong, directional selective forces.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."