Salmonid Fry Stranding Mortality Associated with Daily Water Level Fluctuations in Trail Bridge Reservoir, Oregon ETHAN BELL,* SHARON KRAMER, 1 DAVID ZAJANC, AND JENNIFER ASPITTLE Stillwater Sciences, 850 G Street, Arcata, California 95521, USA Abstract.—Little information exists on how reservoir fluctuations affect stranding risk for early life stages of salmonids. This study focuses on the effects of hydroelectric-related water level fluctuations in Trail Bridge Reservoir, Oregon, where salmonids, including the bull trout Salvelinus confluentus and spring Chinook salmon Oncorhynchus tshawytscha, which are listed under the U.S. Endangered Species Act, commonly occur. A distance-from-line sampling design was employed using permanently established transects to estimate the magnitude of stranding of juvenile salmonids during 30 surveys over 3 months in spring 2006. All stranded fish observed during field surveys were mapped onto spatially rectified, low-elevation aerial photographs to assess the patterns in stranding. Most fish were stranded in habitats with a slope of less than 6%, typically in interstitial spaces among cobbles, and in ‘‘potholes.’’ Fish were stranded in similar numbers following small or large fluctuations, and no relationship was apparent between the range in fluctuation and the number of stranded fish or between the average rate of water surface decline and the number of stranded fish. Based on extrapolation, we estimated that 808 spring Chinook salmon fry and 444 brook trout S. fontinalis fry were stranded in Trail Bridge Reservoir during spring 2006. One dead bull trout was observed, but based on the abrasions, open wounds, and signs of infection that we observed, it died prior to the decline in reservoir elevation. Our findings suggest that stranding in this reservoir could be reduced (while retaining the hydroelectric function of the reservoir) by restricting fluctuations to specific elevations during vulnerable fish migration periods, increasing the slope of areas identified as having a high stranding risk, or both. Water levels in rivers and lakes fluctuate naturally according to the regional hydrologic cycle; however, hydroelectric operations that generate power during peaks in demand for electricity can cause water level fluctuations beyond those that occur naturally (Hunter 1992). These effects can be especially acute in reregulation reservoirs, which are designed for large fluctuations in water volume to reduce fluctuations in downstream reaches. These fluctuations can ‘‘strand’’ various life stages of fish and other aquatic species (Bauersfeld 1978; Fraley and Graham 1982; Hunter 1992; Hoffman et al. 2000; Saltveit et al. 2001). Stranding is defined here as the separation of fish from their primary water body (river or reservoir), leading to injury or mortality. The early life stages of salmonids may be particularly susceptible to stranding mortality owing to their reduced swimming ability and prefer- ence for shallow-water habitat with cover (Healey 1991; Baxter and McPhail 1996). Although the effects of water surface elevation fluctuations on salmonids have been well documented for river habitats (as reviewed in Cushman 1985; Hunter 1992), the effects of fluctuations in reservoir habitats have not been closely examined. Furthermore, there are many chal- lenges to studying the effects of stranding, including difficulties in detecting fish (especially fry), which can often be partially or completely buried in mud, interstices of gravel, or vegetation (Becker et al. 1981; Saltveit et al. 2001); accounting for predation of stranded fish prior to detection; and controlling for changes in susceptibility of fish to stranding as they grow and move between habitats of varying stranding risk. As a result, many studies are not able to provide accurate estimates of the magnitude of stranding, nor can they collect enough detailed information on spatial patterns to recommend appropriate changes to opera- tions, or other management actions (Hunter 1992). This study was designed to address the central challenges in estimating the number of salmonid fry that become stranded during the period of greatest vulnerability (downstream fry migration) in a hydroelectric re- regulation reservoir. In addition, we evaluated the spatial and temporal patterns of stranding to determine whether particular substrate types, elevations, project operations, or other factors influenced the incidence of stranding. The results of this study informed recom- mendations for specific management actions to de- crease stranding risk. Methods Study site.—Trail Bridge Reservoir is located on the upper McKenzie River approximately 113 km east of * Corresponding author: ethan@stillwatersci.com 1 Present address: H. T. Harvey and Associates, 1125 16th Street, Suite 209, Arcata, California 95521, USA. Received February 12, 2007; accepted February 8, 2008 Published online October 13, 2008 1515 North American Journal of Fisheries Management 28:1515–1528, 2008 Ó Copyright by the American Fisheries Society 2008 DOI: 10.1577/M07-026.1 [Article]