Effects of water storage and delivery on spawning activity of endangered Lost River Suckers in Upper Klamath Lake, Oregon

ABSTRACT Objective Lake-spawning Lost River Suckers Deltistes luxatus in Upper Klamath Lake spawn in shallow shoreline locations where springs discharge into the lake. Because of the shallow depths of the spawning habitat, water management that changes lake surface elevation during the spawning season could cause behavioral changes in spawning fish due to reduced available habitat and decreasing depths. Our goal was to examine the association between lake surface elevation, including rapid declines in lake surface elevation, and the interannual use of specific spawning sites, fidelity to these sites, daily spawning activity, and spawning duration. Methods We quantified the number of springs visited annually by individual Lost River Suckers and investigated variation in use of specific springs during different environmental conditions. We constructed generalized additive models to describe patterns in the daily number of unique sucker detections at the shoreline springs during the spawning season relative to environmental covariates. Lastly, we used extended Cox models with time-varying covariates to test hypotheses about the effects of lake surface elevation on the amount of time from the first to last annual detections of suckers at the shoreline springs. Results Low surface elevations in Upper Klamath Lake reduced the number of springs visited by lake-spawning Lost River Suckers and affected spring usage. Site fidelity to specific springs during typical years was evident, but spawning activity primarily took place at two of the four springs during years with low lake surface elevation. Lake surface elevation and rapid declines in surface elevation were not related to total daily unique sucker detections at the springs, but rapid declines were limited in our data. Low lake surface elevations resulted in a reduced duration between the first and last detections at the springs for both females and males. Conclusions Low Upper Klamath Lake surface elevations were associated with Lost River Suckers visiting a reduced number of spawning sites, which spawning sites were primarily visited, and shorter durations between the first and last detections. Although rapid declines in lake surface elevation did not have an effect in any of our models, this result is limited to the magnitude of declines in lake surface elevation documented in our data, which was small. Larger declines may influence spawning sucker behavior.