9588

• UCREFRP Catalog Number: 9588
• Author: Bestgen, K. R. and e. al.
• Title: Population Status of Colorado Pikeminnow in the Green River Basin, Utah and Colorado.
• USFW Year: 2005.
• USFW - Doc Type: Fort Collins, CO.
• Copyright Material: NO

inherent in most estimating models in program CAPTURE. Because capture probabilities of
individuals often vary spatially or over time (this study), it is useful to adjust probabilities of
capture by a covariate (here TL) that is known to affect rates of capture of the animals in the
sample. Use of a continuous covariate such as fish length results in different probabilities of
capture for fish of different lengths, and each then contributes differently to the abundance
estimate. For example, for every 500-mm-TL Colorado pikeminnow captured that has a
hypothetical and relatively high probability of capture of 0.20, there must be an additional four
individuals of that size in the population that were not captured (1 / 0.2 = 5 total pikeminnow of
that size in the population). Alternatively, if a relatively larger 800-mm-TL Colorado
pikeminnow was captured that has a lower probability of capture of 0.05, that means there may
be an additional 19 individuals in the population that were not captured (1 / 0.05 = 20 total
pikeminnow of that size). The abundance estimate for the population results from summing the
total number of fish captured and not captured. It is important to remember that probabilities of
capture are themselves estimated quantities with associated variances and their accuracy is
fundamental to the accuracy and precision of abundance estimates.
As was apparent from model selection, probabilities of capture were relatively low and

varied widely among sampling occasions, reaches, and years (Table 8). Maximum capture
probabilities (usually for a Colorado pikeminnow about 580-mm TL) were relatively low and
ranged from about 0.01 to 0.20. Average probabilities of capture were slightly higher for the
White River and the lower Green River reach than for the other reaches. Average annual values
of probabilities of capture were lower for most river reaches in 2002 and 2003, drought years
characterized by a short run-off period and a low maximum discharge level. Average
33