Laserfiche WebLink
<br />1978). These models give maximum likelihood estimates of population size and probability of <br />capture for each sampling occasion. Population estimates were transformed to densities <br />(capture-recapture estimate / study-reach length) to allow comparison of estimates made during <br />the first year of this project (l6-km reach) with subsequent years (32-km reaches). <br /> <br />Overwinter survival probability (S) was calculated by dividing each spring population <br /> <br />estimate (N2) by the estimate for the previous autumn (N1) : <br /> <br />^ N2 <br />S=-;:-. <br />N1 <br /> <br />Variance of S was calculated by the delta method for propagation of errors (Robson and Spangler <br />1978): <br /> <br />v (S ) =( ~J' V, (N,) +( ::)' V, (N,) . <br /> <br />Comparison of CPUE and Capture-Recapture Estimates <br />The same capture data used to calculate capture-recapture estimates were also used to <br />estimate CPUE. Catch per unit effort for a backwater was defined as the number of Colorado <br />squawfish captured per seine haul. To explore the effect of effort on CPUE estimates, CPUE <br />was calculated for a single pass and for combined sampling passes. Catch per unit effort for a <br />single pass was estimated by averaging CPUE of each backwater for that pass. Combined CPUE <br />was calculated by averaging CPUE of each backwater from all three passes. <br />Estimates of abundance and overwinter survival from CPUE and capture-recapture were <br />compared by inspection and by testing for an association between the measures. Pearson <br />correlation coefficient was used to evaluate the association between CPUE and population <br />density (Zar 1984). <br /> <br />6 <br />