Laserfiche WebLink
these seasons, 33s and 32~ of backwaters, respectively, held Colorado squawfis <br />comparison, only 22% of backwaters sampled in the spring yielded this species <br />Preference for use of habitat classes was tested through contingency tab <br />for the 668 habitats for which type classifications had been recorded.. Observ <br />classes was significantly different (p<0.05) from expected use for spring fa <br />combined (Table 4). Observed vs expected use for summer samples was significa <br />p<0.10. Contingency table analysis of backwater distribution by class for Col <br />backwaters vs the class distribution of backwaters not used by Colorado squawf <br />significantly different (Fisher's Exact 2-tailed p<0.05; Table 5). The diffe <br />distributions was driven by a greater than expected use of secondary (scour) c <br />less than expected use of migrating sand waves. <br />The Wilcoxon rank-sum test detected significant differences (p< 0.05) in <br />variables between Colorado squawfish backwaters and unused backwaters. Nurser <br />used by this species were larger (length and width), deeper, more turbid, had <br />but lower difference from the main channel temperature, and had less cover ava <br />devoid of Colorado squawfish (Table 6). These differences were most easily re <br />comparisons of mean or computed values (area and volume). However, these calc <br />were all derived from individual observations also showing significant differe <br />area and volume of Colorado squawfish habitats, both derived from significant <br />variables, were considerable greater than in other habitats; often several ord <br />Some variability did exist, though. In 1995, many of these variables showed v <br />the general trend. <br />