Modeling Fisheries Impacts Can Be Reliable at Basic Levels, but Predictions Become Tenuous as Complexity Increases

Developing models to assess fisheries impacts caused by the operation of Hungry Horse and Libby Dams in northwestern Montana provided lessons for restoring ecosystem functions at other dams.  Component models allow assessment of model reliability at each step, from physical processes through trophic levels.  Empirical data were used model development, and new data were used for model verification.  Simulations were used to develop dam operating strategies that balance fisheries needs with flood risk management and power generation.  Several fish population models failed to produce reliable results, so were replaced with simplified fish growth models.  River models were designed to share data with reservoir models, so that operations could be optimized for native fish species and critical life stages, upstream and downstream of the dams.  Streamlined versions were applied to other headwater storage projects to assess various operating scenarios for the Columbia River Treaty review.  Occasionally, dam operations must be assessed rapidly and inexpensively, as illustrated by a simple benthic biomass model that was designed for use at any location where wetted perimeter has been calibrated to stream discharge.  Alternative dam operation strategies can be reliably compared and ranked across a wide range of seasonal water supplies from drought to flood.