Laserfiche WebLink
estimator which allowed estimation of Colorado pikeminnow abundance for every sampling year <br />and potentially, estimation of survival between years of sampling. The Huggins model allowed <br />use of the covariate TL, which was important because fish size affected capture probabilities and <br />survival rates of Colorado pikeminnow. Incorporation of size effects into survival and capture <br />probabilities allowed for more efficient and realistic population modeling. Effects of TL on <br />survival rates also revealed empirical information about the demographics of Colorado <br />pikeminnow in the Green River Basin. <br />The multi-strata aspect of the estimating model was important because of the sampling <br />design, and because differences in size-structure of Colorado pikeminnow among reaches (strata) <br />affected capture probabilities and ultimately abundance estimates. Differences in capture <br />probabilities across strata might also be expected because of differences in geomorphic <br />conditions (canyon vs. valley), sampling crews, and the type of sampling craft (raft vs. boat) used <br />in each reach. This was important because sampling crews noted differences in capture <br />efficiency among sampling trips within a year and across years because of differences in <br />turbidity, flow level, or availability of off-channel habitat. Finally, this model offered the further <br />flexibility of estimating effects of fish moving from one reach to another (0) among years. Thus, <br />the manner in which the data were collected and the biology of the subject animal logically led to <br />use of this estimating model. <br />Fulfilling the assumptions of the underlying model is a critical first step in obtaining <br />reliable abundance estimates. Aspects of the experimental design employed in this study ensured <br />that most assumptions of closed-model abundance estimators were met. The assumption of <br />demographic closure was met, in part, because within-year sampling was limited to a relatively <br />short time period in spring prior to Colorado pikeminnow migration to spawning areas. This <br />40