8163 - Laserfiche WebLink

Home Browse Search

STATUS OF ENDANGERED COLORADO SQUAWFISH 965 study period, though one yearling-sized individual (105 mm) was captured there in 1986. Larvae were collected from the middle subreach (1-8 fish an- nually) during 6 of the 9 years. All were collected downstream ofrkm 263.5. In 1986, one larva was also collected from the lower 3.5 km of the Gun- nison River. The highest number of larvae col- lected from the middle subreach was in 1986; how- ever, differences in mean CPUE among years (Fig- ure 5) were not significant (X2 = 9.24, P = 0.32, df = 8, K-W ANOV A). Age-O Colorado squawfish were captured (1-29 fish annually) from the mid- dle subreach also during 6 of the 9 years; all were captured downstream of rkm 262.7. Mean CPUE was significantly different among years (X2 = 27.08, P < 0.0007, df = 8) with 1986 being the only year significantly higher than others (P < 0.05). Though CPUE in 1987 appeared higher than in subsequent years, the difference was not sig- nificant (P > 0.05). These results for age-O fish were similar to those of McAda et al. (1994) for the lower reach, where mean CPUE for 1986 was double that of any other year (1986-1992). Earlier, riverwide surveys of age-O Colorado squawfish by the USFWS, extending back to 1982, indicated an even higher CPUE in 1985 than in 1986 (McAda and Kaeding 1989). Discussion Demographics The best point estimate of abundance of adult Colorado squawfish in the upper reach during 1991-1994 was about 250 fish; in the lower reach, the best point estimate was about 350 fish. Al- though sizable variability was associated with these estimates, different estimators produced sim- ilar point estimates. Many individuals in the lower reach subpopulation appeared to be part of 1-3 strong year-classes spawned in the mid-1980s, and an increasing trend in the upper-reach subpopu- lation during 1991-1994 resulted from upstream movements of these cohorts (see Osmundson et al. 1998). At the beginning of this study, very few adult-sized fish were captured in the lower reach. Thus, the 250 adult fish in the upper reach might have represented the bulk of the breeding popu- lation of the entire river until this pulse of new fish recruited to adulthood. This census number indicates an effective population size (Ne; usually considerably less than the census or actual popu- lation size) well below thresholds suggested for maintaining adaptive genetic variance for long- term population viability: 5,000 suggested by Cil 0.12 58 C. E 0.10 co Je .<= 0.08 III !E- O> 0.06 ~ .<= J,l 0.04 B t: 0.02 co 0> ::;; 0.00 2.5 Larval fish 54 52 72 ~tttn 72 f 14 0 2.0 f " 0 - .<= III 1.5 !E- $ e! 1.0 .<= 0 i3 i- t: 0.5 co 14 0> ::;; Age-O fish 16 16 0.0 16 16 1~1m1~1~1~1~1~1~1_ FIGURE 5.-Annual catch per unit effort (::'::SE) of larval (top) and age-O (bottom) Colorado squawfish in the middle subreach from 1986 through 1994. Catch rates are geo- metric means; sample size is reported above SE bars. Lande (1995) or 500 by Franklin (1980) and Soule (1980). The best point estimate of annual survival of adult Colorado squawfish in the upper reach (0.86) was very similar to the survival rate for Colorado squawfish greater than 550 mm (0.85) estimated by Osmundson et al. (1997) using a different ap- proach. Though both methods used the same cap- ture records, estimates here were based on cap- ture-recapture histories whereas Osmundson et al. (1997) used a modified Chapman-Robson (Seber 1982) model where survival is based on declining numbers of increasingly older individuals. Though estimates in both studies were higher than one re- ported by Gilpin (1993) for the Green River pop- ulation (0.81), actual differences between popu- lations may be small: Gilpin's sensitivity analysis indicated that an overestimate in his growth cal- culations would have resulted in underestimating adult survival rate. In other words, if adult growth averaged 10 mm in length per year (as previously estimated by Tyus 1988) instead of the 15 mm estimated by Gilpin (1993), his survival rate es- timate would increase to 0.87. In both rivers, adult survival is relatively high and therefore probably not a major constraint on population increase. Length frequencies of adult Colorado squawfish differed between the upper and lower reaches, with larger (>500 mm) fish predominant in the upper reach. Length frequencies in the lower reach, how-