Laserfiche WebLink
<br />substrate size and habitat type were weighted according to the total number of <br />fish in the sample and were entered into. the analysis concurrently. <br />Histograms for each of these four habitat variables were also developed for <br />comparisons among species. <br />Habitat overlap (B) between age-0 Colorado squawfish and each of the <br />sympatric species was estimated using the Schoenen Index (Schoenen 1970): <br />n <br />B=1-0.5(1_llPxi-Pyi~) <br />where n = number of habitat categories, PXi = the proportion of species x <br />collected from habitat category i, and P~ = the proportion of species ~ <br />collected from habitat category i. For these analyses habitat categories <br />consisted of the 17 intervals that occurred between canonical variates along <br />Discriminant Function 1 (Matthews and Hill 1980). <br />For the habitats from which Colorado squawfish were collected,. <br />correlations between catch per effort (C/E; number of fish/m2 seined) for age- <br />0 Colorado squawfish and C/E for each of the six sympatric species were <br />estimated using the Pearson Correlation Coefficient. All analyses were done <br />using SXSTAT (Wilkinson 1988), a statistical package for use on <br />microcomputers. <br />Results <br />Analyses showed little discrimination between habitats occupied by age-0 <br />Colorado squawfish and those used by the other species. Wilks's lambda, a <br />measure of the discrimination between the habitat use of Colorado squawfish <br />and that of a sympatric species ranged from 0.556 for Gila spp. in the Green <br />River to 0.999 for red shiner in the Colorado River (Table 1, Figures 1 and <br /> <br />