8161 - Laserfiche WebLink

Home Browse Search

GROWTH AND SURVIVAL OF COLORADO SQUAWFISH 689 Westwater Canyon (rk 181-200) served as the de- marcation zone (Figure 1) and was not sampled. Fish Capture and Marking During late April to mid-June 1991-1994 efforts were made to capture Colorado squawfish through- out the study area. Previous radiotelemetry studies indicated high seasonal use of low-velocity, off- channel habitats as main-stem flows increased from spring runoff (Osmundson and Kaeding 1989). Trammel nets were used to quickly block mouths of backwaters, and a powerboat "drove" fish from the backwaters toward the channel, there- by ensnaring fish, an active rather than passive capture method. Each year, three passes were made through the upper study area and two were made through the lower study area. Every backwater suspected of sheltering adult Colorado squawfish was sampled on each pass (91-104 different sites annually). In some reaches, where backwaters were rare, shore- lines were electrofished with a 4.9-m johnboat equipped with a Coffelt VVP-15 (Coffelt Manu- facturing, Flagstaff, Arizona) that produced pulsed DC. Captured fish were scanned and tagged if a PIT tag was not detected. Fish were first anesthetized with tricane methanesulfonate. Tags were implant- ed in the body cavity with a hypodermic needle that was inserted 2-5 mm posterior to the base of the left pelvic fin. Five to eight scales were re- moved between the lateral line and dorsal fin in- sertion. Maximum total length (Anderson and Gu- treuter 1983) was measured and fish were released after recovery from the anesthetic. Our growth data were supplemented by addi- tional records obtained from various sources, in- cluding a pilot exercise conducted in the upper reach in 1990. The Colorado Division of Wildlife (CDOW) contributed data from 1991 through 1995, and some recapture data were provided by the Utah Division of Wildlife Resources (UDWR); both agencies conduct annual spring electrofishing surveys. Records from a U.S. Fish and Wildlife Service (FWS) 1994-1996 survey of the lower 3.5 km of the Gunnison River and a 1995-l996 survey near rk 262 were also used. Length data collected by FWS in 1982 were used in tests for a stable age distribution. Calculation of Growth A combination of techniques was applied to as- certain TL at various ages. Our goal was to use measured changes of individuals as much as pos- sible and minimize reliance on scales. Recaptures of juvenile and subadult Colorado squaw fish were few, however, so measurements of fish aged by various means, including scales (ages 2-7), were used for young fish. Ages 0-7.-Colorado squawfish average 7.7 mm long at hatching (Snyder 1981). For l-year-old Colorado squawfish, mean TL was calculated from measurements of 73 fish seined on 28 June 1989 (about a year after hatching) from backwaters near rk 87, for which a unimodal si'~e distribution in- dicated all were of one age-grcup. Beginning with 2-year-old fish, annual incre2.ses represent growth from one spring to the next, rather than between hatching periods or between winters. Mean TL at each age for Colorado squawfish 2-7 years old was derived from specimens aged by scales captured between 28 April and 21 June 1991-1994, pri- marily from the lower reach. Because hatching generally occurs in July or August, many of these fish were 1-J months younger than the reported age. Scale annuli were counted; the first annulus, as- sumed missing (e.g., Seethaler 1978; Hawkins 1992), was added to estimate age. Four to eight scales were examined from each fish and the num- ber of annuli most frequently identified was used. We used scales from known-age pond-reared Col- orado squawfish (2-4 years old) for developing our technique and later testing our accuracy. Ages 8 and older.-For fish 8 years and older, sufficient recapture data were available for deter- mining mean annual growth increments. Data were typically from fish with capture-recapture inter- vals of 1 year. However, because of lower sample sizes for fish 600 mm and larger, annual increments were also calculated for fish with recapture inter- vals of 2 or 3 years. Annual increments were av- eraged by size-class of fish (50-mm categories) at initial capture. Analysis of variance (ANOV A) was used to test for differences in rates among size- classes. Mean TL at each age was first calculated by adding the appropriate mean growth increment (based on length-class) to mean TL of the preced- ing age. Average TL of 7-year-old fish as deter- mined from scales was the starting point. The pro- cess was continued to 900 mm, roughly the upper size limit of Colorado squawfish captured in recent years. Monte Carlo simulation based on mean and vari- ance of growth increments for each size-class was also used to estimate mean TL at each age. Twenty simulations were conducted with different ran-