8166 - Laserfiche WebLink

Home Browse Search

1 0.9 and three cases between 0.7 and 0.8. Gila spp. ranged from 0.587 to 0.938, only one value exceeded 0.9. Most values were between 0.8 and 0.9. Speckled dace showed the most discrimination in habitat use with Colorado I squawfish. Lambda ranged from 0.2 to 0.770, with two cases greater than 0.7, two cases between 0.6 and 0.7, and four cases less than 0.6. Habitat ,and diet overlap indices are often considered to be biologically significant at values exceeding 0.6 (e.g., Wallace 1981). Although technically not an overlap index, lambda behaves in much the same way by 1 measuring the difference in habitat use between the species. Thus the assumption that values of lambda greater than 0.6 are biologically meaningful ' may be reasonable. Certainly values greater than 0.6 reflect great similarity in habitat use between species. The correlations between the discriminating variables and the dependent canonical factors indicate which of the measured variables (depth, velocity or substrate) had the greatest effect on separation in habitat use between species (Appendix Tables 5 and 6). Depth was the most important contributor to discrimination between Gila spp. and Colorado squawfish habitat, whereas velocity and substrate size were the most important variables to discriminate 1 between age-0 Colorado squawfish habitat and that utilized by channel catfish and speckled dace (Appendix Tables 5 and 6). Because no distinction existed ' between habitat used by Colorado squawfish and that occupied by red shiner, sand shiner or fathead minnow, the relative contribution of the measured habitat variables is meaningless. In the Green River, C/E of fathead minnow was significantly correlated with C/E of age-0 Colorado squawfish in backwaters in five of the six cases L examined. However, the correlation coefficients (r) were low, ranging ' 10 t