Laserfiche WebLink
increased to a range from 13% to 25%. However, error rates for classification <br />backwaters alone were almost always lower than for Colorado squawfish backwate <br />of up to 80% were observed for Colorado squawfish backwaters, with most greate <br />Backwaters not used by Colorado squawfish were incorrectly classified 12% of t <br />Some examples of discriminant modeling are presented in Appendix A. <br />Fish Community and Catch-per-unit-effort <br />Catch-per-unit effort (CPUE; fish/100m3) for Colorado squawfish was gene <br />in summer and lowest in spring (Table 7). Suppositions that catch rate was in <br />run-off flows could not be substantiated. Although a negative relationship ex <br />CPUE and volume of peak flow, date of peak flow and duration of peak flow; the <br />were not very strong (R2<0.61) and were not significant (p>0.15) (Figure 5). <br />correlation analyses indicated no relationships (p>0.15) between flow paramete <br />Colorado squawfish (Table 8). <br />Twelve species of fish were caught in backwaters in addition to Colorado <br />native and eight non-native (Table 9). Red shiner (Notropis Iutrensis) was th <br />species. This fish was found in 91% (n=945) of all samples. CPUE for this sp <br />surpassed CPUE for any other species. The second most common species was fath <br />(Pimephales promelus). This species was present in 70% of all sites. Sand sh <br />Stramineus) was present in 51% (n=535) of all backwaters. During the period 1 <br />shiners were common in samples, but they increased in abundance and distributi <br />