Laserfiche WebLink
<br />Population estimates and capture-recapture simulations. Population estimates were made <br /> <br /> <br />using two methods. First, the computer program CAPTURE (Otis et al. 1978, White et al. 1982) <br /> <br /> <br />was used to make population estimates from the capture-recapture data. However, the first two <br /> <br /> <br />years of the study we had no recaptures; only in 2000 did we obtain recaptures so that CAPTURE <br /> <br />could make a population estimate. For the second method, we assumed the population was <br /> <br /> <br />stable between 1998 and 2000 (i.e., mortality equaled recruitment and therefore a constant N), <br /> <br /> <br />and we used the binomial probability density function to calculate likelihoods of obtaining the <br /> <br /> <br />recapture data for the second and third passes each year, and since the data sets are independent <br /> <br /> <br />we multiplied them together to obtain the likelihood for the entire data set (Haddon 2001). We <br /> <br />then found N that maximized the joint likelihood. Since we did not know how many tagged fish <br /> <br /> <br />survived from year to year, we ignored any recaptures from previous years (see Appendix B for <br /> <br />calculation details). Ninety-five percentile confidence limits were estimated using the likelihood <br /> <br /> <br />profile. The population estimates apply to fish> 150 mm TL (the smallest fish tagged). <br /> <br /> <br />The simulation module in CAPTURE was used to examine how our sampling effort <br /> <br /> <br />could be improved. We used time varying probabilities of capture (generated by beta <br /> <br /> <br />distributions with alpha = 1.5 and beta = 2.5 for all runs) and the Darroch population estimator. <br /> <br /> <br />Relative bias was defined as (average Nhat-N)/N and coefficient of variation (CV) as SE/Nhat. <br /> <br /> <br />Each trial consisted of 100 replications. <br /> <br />Catch per unit effort (CPUE) and simulations. Catch of humpback chub> 150 mm TL <br /> <br /> <br />per hour of raft electrofishing was recorded for two sampling passes through Yampa Canyon for <br /> <br />1998, 1999, and 2000. Effort was measured by the timer on the Smith-Root GPP, which counted <br /> <br />6 <br />