Laserfiche WebLink
<br />considered an adult. For large-bodied fish (Catostomus <br /> <br />discobolus and Catostomus latipinnisl, sub-adults were age-l <br /> <br />fish, as determined by length-frequency histograms. <br /> <br />For each species and age-class, a weighted mean and standard <br /> <br />deviation for each habitat variable was calculated. All habitat <br /> <br />measures were weighted by abundance (number of fish in a <br /> <br />habitat) <br /> <br />Discriminant function analysis was used to classify <br /> <br />species (and age-classes) by habitat variables (SSPS 1996). <br /> <br />Initially, a stepwise discriminate analysis was used to determine <br /> <br />those variables that accounted for the greatest amount of <br /> <br />variation among species and age-classes. Only variables that <br /> <br />were significant (P < 0.15) were included in the discriminant <br /> <br />analysis. The entire data set was randomized and one half was <br /> <br />chosen to create the d~scriminant function and the other half was <br /> <br />used to test how well the function predicted the presence of a <br /> <br />particular species or age-class. <br /> <br />Habitat association was only examined between native and <br /> <br />nonnative species. We did not consider association within groups <br /> <br />of native species or nonnative species. Overlap in habitat use <br /> <br />between the species was calculated as: <br /> <br />IfN+I <br />%overlap=---- x 100 <br />HN <br /> <br />Where HN+I = the number of seine hauls where the chosen native and <br /> <br />nonnative species/age-class occurred together and HN = the total <br /> <br />number of seine hauls where the native species/age-class <br /> <br />8 <br />