7055

• UCREFRP Catalog Number: 7055
• Author: Archer, D. L., L. R. Kaeding and B. D. Burdick.
• Title: A Study of the Endangered Fishes of the Upper Colorado River.
• USFW Year: 1985.
• USFW - Doc Type: Grand Junction, Colorado.
• Copyright Material: NO

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

size classes for each comparison -- < 500 mm vs ) 500 mm, < 550 mm vs ) 550
mm. and ( 600 mm \IS J\ 500 mm.

The larval and post-larval, age-O catch data contained such high
proportions of zero catch-per-unit-effort (CPUE) values that comparisons of
catch rates among years or river reaches using normal statistical tests (e.g.
ANOVA) were precluded. The high zero catch rates resulted in highly skewed
distributions that could not be transformed to approximate a normal
distribution necessary to meet the assumptions of ANOVA. Relative abundance
of larval and post-larval, age-Q Colorado squawfish was therefore compared
using a chi-square test of the observed and expected number of samples that
contained at least one squawfish and those that contained none (Hamley et al.
1983). Accordingly, the frequency of zero CPUE (rather than mean CPUE) was
used as the index of relative abundance for each sampling period (see Bannerot
and Austin 1983 for a discussion of this technique).

Results

Radiotelemetry

Sixty-one adult-size Colorado squawfish were equipped with radio
transmitters during the study: 22 in 1982; 21 in 1983; and 18 in 1984 (Tables
1-3). Eleven fish were tagged with 540-day transmitters in early-fall 1982;
the remainder was tagged with 180 or 300-day transmitters in early spring
(1982-1984) prior to the onset of spawning. Most fish were tagged in the
upper reach (27 fish), whereas 16 were tagged in upper lake Powell (Reach 1)
and 16 in reaches 2 and 3 (Figure 2). Contact was lost with 16% of the fish
immediately after release, whereas we had periodic contact with 72% of the
fish for at least 30 days and with 5~ for more than 100 days (Tables 1-3:
Appendix Tables 25-27). The longest period of intermittent contact was 410 days
(Fish No. 23).

Only one radio-tagged fish (No.2) moved among adjacent river reaches
(Figure 2). It was in lake Powell during early and mid summer, 1982, but had
moved 199 mi upstream to the Grand Valley by late summer. Although it might be
speculated that this fish was one that had been tagged near Grand Junction but
was later mistakenly thought to be from lake Powell, such confusion could not
have occurred. Small changes in the frequency of some radio transmitters have
been noted, but in 1982 all fish from lake Powell were given radios with
frequencies that differed markedly from those of fish elsewhere in the river.
A frequency change of the magnitude necessary to confuse a lake Powell fish
with fish from elsewhere would have been unprecedented. The longest distance
moved by any of the other radio-tagged fish was 40 m; (No. 20).

No difference in movement patterns was noted among the three years of the
study or between the various sizes classes of fish. With the exception of the
fish that moved from lake Powell to Palisade, fish generally moved short
distances up and downstream within their respective reaches and did not move
into adjoining reaches. Most movement occurred during the presumed spawning
period, but movement of radio-equipped fish to one or two relatively short
river reaches during the spawning season (as has been reported in the Green
River basin) was not observed. Shortly after tagging, fish in upper Lake
Powell tended to move downstream into the main body of the reservoir where

10