Laserfiche WebLink
1 <br />1 <br /> <br />need to be made cautiously, especially with regard to relationships with environmental variables <br />such as stream flows. <br />One possible inference from light-trap data for both the middle and lower Green River <br /> <br />ii <br /> <br /> <br />1 <br />L <br /> <br /> <br />1 <br /> <br /> <br />1 <br /> <br /> <br /> <br /> <br />reaches is that the number of larvae captured since the early to mid-1990's appears to be declining, <br />in terms of absolute numbers and CPUE data. An obvious reason for reduced reproduction is the <br />decline of the presumed largest population of wild fish in the middle Green River. Other reasons <br />for reduced reproduction are not apparent. Initially, we thought light-trap data from a site or sites <br />that has a known history of captures and from sites present in all years regardless of flow level, <br />may provide some index of reach-wide reproductive success over time. Unfortunately, capture <br />data were too sparse in many years to allow for such an analysis. <br />Use of light-trap data to reliably estimate annual reproductive success of razorback suckers <br />in the Green River may be possible, but additional information is needed. For example, no <br />information exists to describe the probability larvae will be detected in a suitable habitat if light- <br />trap sampling is conducted. Minimally, the relationship between fish density, backwater size, <br />sampling effort, and detection probabilities should be quantified. At a reach scale, an aggregate of <br />samples gathered to estimate a CPUE index may give an adequate picture of reproductive success <br />only if capture probabilities are equal or known across sites and years. One way to estimate this <br />would be to mark and recapture larvae over the sampling season that were released upstream of <br />capture areas. The eventual loss of remaining wild fish and their reproductive output seemingly <br />makes it easy to detect reproductive success by stocked fish. Light-trap sampling may give some <br />reliable measure of reproductive success of stocked fish, especially if numbers of larvae captured <br />increase by an order of magnitude (e.g., thousands instead of < than 100 captured per year). <br />26 <br />