7835

• UCREFRP Catalog Number: 7835
• Author: Osmundson, D. B., et al.
• Title: Studies Of Colorado Squawfish In The Upper Colorado River, Final Reports.
• USFW Year: 1997.
• USFW - Doc Type: Recovery Implementation Program, Project No. 14,
• Copyright Material: NO

Kaeding 199 1). If the lower basin was once the source of the upper basin's large fish, as we
suggest, subsequent blockage of upstream movement by the construction of mainstem dams
and the eventual extirpation of lower basin populations might explain the disappearance of
such large individuals in recent times.
Survival
The range of suitable (P < 0.05) survival estimates of 0.83-0.87 (1991 through 1994 data)
was higher than that calculated by Gilpin (1993) for the Green River population (0.81) and
may reflect differences in environmental factors between the two rivers. However, had
Gilpin's (1993) methodology provided a range of estimates rather than a point estimate, his
range may have overlapped ours, suggesting no differences in survival rates between the two
rivers. Also, deviations from a stable population size in either river would affect the
comparability of these estimates since both were based on the assumption of stable population
size. Other differences in methodologies or assumptions may explain some of the difference in
estimated rates.
The method used to determine survival rate in this study was not designed to replace currently
available capture-recapture models for open populations such as Jolly-Seber (Jolly 1965,
Seber 1965) and variations thereof. However, it can be applied when consistent capture effort
and long capture histories necessary for the Jolly-Seber model (see Pollock et al. 1990) are
not available. Chief limitations of the method used here are the need for population structure
information over several years (to determine if age distribution is stable), and capture methods
that sample various-sized fish at rates representative of the actual population. Tests indicated
that population size structure of Colorado squawfish z 550 mm TL in the Colorado River did
not change significantly from 1990 through 1995 and was similar to that in 1982. We were
not able to test for capture bias in trammel nets, but disproportionate numbers of large fish
caught with electrofishing gear was consistent with known biases of this gear type (Reynolds
1983). When fish captured by electrofishing were included in the analysis, estimated survival
rates increased as expected (not shown).
Though historic survival rates are unknown, there are some new sources of mortality that may
have lowered rates during recent times, while other sources have probably changed little.
Accidents during high water events (abrasions, strandings, etc.), stress-related factors during
spawning, and avian predation by great blue herons (Ardea herodias) and bald eagles
(Halweetus leucocephalus) have likely always been primary sources of adult mortality.
Significant change in mortality rates from these factors is unlikely except perhaps that the
frequency of strandings may have increased in recent times because of the excavation of
numerous gravel-pit ponds along the floodplain of the upper reach. Colorado squawfish are
attracted to these habitats when ponds flood during spring runoff and strandings have been
documented (FWS unpublished data) when declining flows later re-isolate the ponds.
Substantial angling and netting activities of settlers and residents from the 1890's through the
A-18