8175 - Laserfiche WebLink

Home Browse Search

DD at Jensen was 37% less overall than that at Green River. Based on former studies (Black and Bulkley 1985; Kaeding and Osmundson 1988), we indicated potential growth indices on the graph of average annual temperatures (Figure 10). Optimal growth temperatures are not attained near Jensen but are reached for an average period of approximately 2 months near Green River during July and August. Also, spawning temperatures above 190C in upper Green River were reached 2 weeks to 1 month after that which occurs near the Green River gage, which may add length to the growing season in lower Green River reaches if spawning occurs sooner at this site. A comparison of summer seasons before and winter seasons after collection of age-O Colorado pikeminnow in fall at the two study sites showed that DD remained relatively constant from year to year and that the lower Green River was consistently warmer than the middle Green River. WhenDD was calculated as degrees above zero, winter accumulation is approximately half of summer; however, when DD was calculated as degrees above 130C multiplied by the potential growth indices, winter DD was less than one-tenth that of summer (Figure 11a and b). The latter more closely reflected available growing season length in the respective seasons, because growth potential was incorporated. Stepwise multiple regressions demonstrated that average summer flows (July through September) predicted 52% of variation in mean size ofage-O fish in fall (p = 0.008) (Figure 12); the model excluded peak discharge, date of peak discharge and summer DD. Degree-day accumulation did appear to be loosely associated with fish size in the lower Green River reach, however this relationship was not significant (p = 0.17). No relationship between DD and size of age-O fish in fall was detected in the middle Green River reach (Figure 13). No model of the physical variables examined significantly explained relative abundance (measured as geometric mean CPE) of age-O fish in fall. Appendix 1 summarizes variables used in this analysis to determine associations among physical variables and year class strength. For age-l fish in spring, 67% of variation in length was predicted by length in fall (p = 0.009), relative abundance in fall (p = 0.075), and winter DD (November through March) (p = 0.114). Winter flows were excluded from this model. Ninety-seven percent of variation in relative abundance in spring was explained by relative abundance of age-O fish in fall (p < 0.000); mean length in fall, winter DD and wint~r flows were excluded from this model. Figures 14 and 15 show the relationship between length offish in fall versus spring and number offish in fall versus spring. DISCUSSION Scale circuli counts of 9 or fewer preceding the first growth check on 75% of adult Colorado pikeminnow suggest that these fish were large enough to form a growth check in their first year. However, approximately 50% of age-O fish captured and measured in fall of 1991 were probably too small to form a first-year growth check, and 45 to 90% of age-O fish collected in other years were too small. We suggest this apparent discrepancy is a result of size-dependent overwinter mortality and growth of age-O Colorado pikeminnow, which was supported by examination of empirical QQ plots for seven seasons. These findings are significant for two reasons. First, the discovery of a first year growth check on most adults conflicts with 10