8105

• UCREFRP Catalog Number: 8105
• Author: Haines, G. B., D. W. Beyers and T. Modde.
• Title: Estimation of Winter Survival, Movement and Dispersal of Young Colorado Squawfish in the Green River, Utah.
• USFW Year: 1998.
• USFW - Doc Type: Recovery Program Project 36,
• Copyright Material: NO

Overwinter survival estimates derived by dividing CPUE in the spring by CPUE in the
autumn assumes that the fish are equally vulnerable to capture in the spring and autumn. To test
this relationship, we calculated seining catchability coefficients. Seining catchability coefficient
for a sampling pass was defined as the probability that a Colorado squawfish would be caught
per seine haul (Ricker 1975). Seining catchability coefficient was calculated by dividing the
probability of capture for a sampling pass (derived from CAPTURE) by the number of seine
hauls in that pass. This measure accounts for the effect of variable effort so that catchability
coefficients from different sites and occasions can be compared. The difference between autumn
and spring catchability coefficients were compared by calculating a (-statistic and comparing it to
a two-tailed Student's critical value. Pearson correlation coefficient with Bonferroni adjustment
(Zar 1984) was used to evaluate the association between catchability coefficients and three
measures of water temperature: main channel, backwater, and the difference between backwater
and main channel temperature.
To compare variability of overwinter survival estimates from CPUE and
capture-recapture methods, the coefficient of variation (CV) was computed. The CV is defined
as (standard error of the estimate · 100)/mean). Coefficients of variation allow equitable
comparison of precision among estimates with differing means because variance is expressed as
a percent of the mean (Zar, 1984). Statistical analyses were conducted with SYSTAT software
(Wilkinson 1990).

Movement
Marked Colorado squawfish were also used to study movement pattems. The 32-km
study reaches were divided into 8-km sections. By varying mark location and color, fish were
given a unique mark that identified which study section they were captured in. When a fish was
recaptured, the marks allowed identification of the section where that individual fish was
originally captured and the number of days at large.
To determine how annual spring flooding redistributed young Colorado squawfish,
surveys were conducted in late July or early August to search for marked fish. Surveys began at
the upstream boundary of each study reach and extended downstream 76 Ian for the Canyonlands

7