9435

• UCREFRP Catalog Number: 9435
• Author: Haines, B. and T. Modde.
• Title: Humpback Chub Monitoring in Yampa Canyon, 1998-2000.
• USFW Year: 2002.
• USFW - Doc Type: Project Number 22a4,
• Copyright Material: NO

profile likelihood confidence limits (Table 5). While noting that the data consisted of only 44

fish captured and 2 recaptures, examination of Table 5 suggests that population size might be

somewhere between a few hundred and a few thousand fish> 150 mm TL. We note that the

point estimates (Nhat) in Table 5 are probably biased too low. In 1998 and 1999 two passes

failed to recapture a single fish that was marked in that year; if these data were considered a

larger estimate would be generated by CAPTURE. And, as we will see below, simulation runs

showed that the low capture probabilities also contribute to under-estimates.

When all capture-recapture data were combined into a likelihood model that assumed the

population was stable in 1998-2000, we estimated that the population was 391 with 95% CI
180-2750 (Appendix B).

We used the simulation module in CAPTURE to examine how the capture-recapture

study could be modified to produce a satisfactory estimate of the Yampa Canyon humpback chub

population. Simulations were run for three probabilities of capture and three population sizes

(Table 6). The probability of capture for the Yampa Canyon humpback chub was about 0.03

(based on an average of 12 captures per pass from a population of 400). In future studies the

probabilities might be doubled, but this would still result in serious bias (10-30 percent too low)

and imprecise (CV range 0.3-0.5) estimates. The probability of capture would need to increase

to about 0.1 in order to get satisfactory bias (< 10 percent) and precision (CV < 0.2), provided the

population size is > 200.

Catch per unit effort. Electrofishing catch per hour for humpback chub> 150 mm TL

ranged from 0.72 to 1.54, and the CV ranged from 0.07 to 0.67 with a mean of 0.31 (Table 7).

11