8163 - Laserfiche WebLink

Home Browse Search

958 OSMUNDSON AND BURNHAM identifying demographic trends via temporal changes in age or size structure, and (4) examining historical accounts to reveal clues to former abun- dance. Systematic, riverwide sampling and mark- ing of adult and sub adult Colorado squawfish was conducted during a 4-year period to address anal- yses 1-3. In addition, we evaluated age-O abun- dance, monitored during a 9-year period, and size- frequency data, collected by other researchers over 20 years. We then discuss the potential for pop- ulation persistence based on these analyses and the present trends in habitat alteration. Methods Study area.-The study area included the entire portion of the Colorado River occupied by Colo- rado squawfish upstream of the Green River con- fluence and the lower 3.5 km of the Gunnison Riv- er upstream to the base of the Redlands diversion dam. Colorado River locations are described in river kilometers (rkm) converted from river miles as mapped by Belknap and Belknap (1974). The study area extended from the Green River conflu- ence (rkm 0.0) upstream to rkm 298.1 at Palisade, Colorado, where further upstream movement of fish is blocked by two diversion dams. Based on the distribution pattern of adults and juveniles, we partitioned the study area into two major reaches, upper and lower (described in Os- mundson et al. 1998, this issue), and excluded from study the intervening 19-km Westwater Can- yon. To examine distribution of adults within the upper reach, three subreaches were identified based on discharge, average gradient, and land- form type. The upper subreach (rkm 298.1-275.1), in an alluvial valley, had mean discharge of 110 m3/s and average gradient of 1.70 m/km. The mid- dle subreach (rkm 275.1-245.5), in the same al- luvial valley, had mean discharge of 175m 3 Is and average gradient of 1.27 m/km. The lower sub- reach (rkm 245.5-200.0), largely canyon bound, had mean discharge of 175 m3/s and average gra- dient of 0.91 m/km. These lower, middle, and up- per subreaches correspond to strata 5, 6, and 7, respectively, described in Osmundson et al. (1998). Subadult and adult capture efforts.-Subadult (250-500 mm long) and adult (450-900 mm long) Colorado squawfish were captured from late April to mid-June 1991-1994 in backwaters throughout the entire study area. Subadults and adults con- gregate in these low-velocity habitats during spring runoff when main-channel flows increase dramatically (Osmundson and Kaeding 1989). Fish were actively entrapped in nets by a method we dubbed "scare and snare." Using a 4.3-m-long johnboat, we first blocked the open end of each backwater with a trammel net; the boat then en- tered the backwater by passing over the net with the motor raised. To scare fish toward the net, the boat was vigorously motored back and forth be- ginning at the far end of the backwater and work- ing toward the mouth. Nets were pulled as soon as sufficient "scare" effort was expended (5-15 min, depending on backwater size). In very large backwaters or flooded ponds, additional nets were set once the mouth was blocked. Trammel nets were 1.8 m deep with a 2.5-cm-bar-mesh inner panel and a 25-cm-bar-mesh outer wall. These nets captured Colorado squawfish as small as 259 mm (all fish lengths reported as total length) but caused gill damage in some small individuals. After the first year, finer-mesh netting (l.3-cm-bar inner mesh and 18-cm-bar-mesh outer wall) was used in the lower reach to prevent gill damage. Ensnared Colorado squawfish were placed in a live well until all fish were removed from the nets. Fish were anesthetized, measured for maximum total length (Anderson and Gutreuter 1983), and electronically scanned for the presence of a passive integrated transponder (PIT) tag (Biomark, Inc., Boise, Idaho). If a PIT tag was not found, one was implanted in the body cavity by using a hypoder- mic needle inserted 2-5 mm posterior to the base of the left pelvic fin. Fish were released after re- covery from the anesthetic. Three passes through the upper study reach were made each spring, and every backwater that might hold Colorado squawfish was netted in each pass. Each pass generally took 7-9 d to complete. In the lower reach, two passes were made each spring (except in 1991 when only one pass was made). In some portions of both reaches where backwater habitats were rare, both shorelines were electro- fished with a 4.9-m-long johnboat equipped with a Coffelt VVP-l5 electrofisher (Coffelt Manufac- turing, Flagstaff, Arizona) that produced pulsed DC. Capture data for portions of some passes were also supplemented with fish electrofished by the Colorado Division of Wildlife (lower subreach of the upper study reach) and U.S. Fish and Wildlife Service (lower 3.5 km of Gunnison River). Survival rate.-Capture-recapture data from the multiple passes were used for estimating survival rates in the upper and lower reaches by using Cor- mack-Jolly-Seber (CJS) models (see Lebreton et al. 1992) because no assumptions were needed concerning abundance, recruitment, or trends in ,