Stream restoration is often targeted to improve fish habitat. However, predictions or evaluations of habitat or fish population responses to these actions are rarely assessed or more often assumed. Model predictions that describe both physical and biological effects from different types of restorations can help test hypotheses to improve our understanding of fish habitat requirements, especially if restoration is treated as an experiment. In addition, models can help synthesize data to determine project effectiveness to link fish responses to physical habitat responses, promoting a mechanistic understanding that is transferrable to other systems. We are applying a mechanistic model to estimate net energy intake (NEI) potential for juvenile steelhead throughout a stream reach. NEI modeling approaches attempt to describe the quantity of food a fish can ingest at a particular site (as a function of food availability, temperature, and fish size) and the energetic costs of living there (as a function of flow patterns, temperature, and fish size). The collection of NEI estimates can be used to help understand the site's overall energetic profitability for fish and to generate an estimate of carrying capacity. We collected topographic information to create digital elevation models (DEMs) of a stream channel along with drifting invertebrates, temperature, discharge, and substrate type, before and after the addition of large wood in the Asotin watershed. This information was used to create hydraulic models to estimate water velocities and flow direction which were then used to predict delivery rates of drift, swimming costs to fish, and ultimately NEI potential throughout the reach. The model was used to predict restoration benefits based on expected geomorphic responses to large wood. In addition, the model was used to translate modest observed changes in topography from the restoration to increases in NEI and carrying capacity of the stream reach. We believe the NEI modeling approach will be useful for restoration planning and evaluation; however, efforts to simplify the model will likely be necessary to gain broader use of this tool.
Available at: http://works.bepress.com/joseph_wheaton/71/