8035 - Laserfiche WebLink

Home Browse Search

- , photographs of 741 live animals taken in the wild at several si tes in the Upper Colorado River drainage. The data were analyzed using principal component analysis, a multivariate method for sorting collections of individuals with discrimination of diagnostic characters. Discriminating measures were found about the mid-body and head; caudal peduncle length, was also significant, but not the least depth of the caudal peduncle as in earlier studies. Fishes greater than 200 cm SL were sortable into distinct groups, even without direct consideration of the occipital hump. Several hypotheses were suggested: (a) G. robusta robusta is a distinct taxon; (b) intergrades or previ- ously-unidentifiable fish are predominately ~. cypha (i.e., G. cypha is more variable than previously thought); (c) hybridiza- tion seems rare when evaluated solely on a morphological basis; (d) Green River fishes collected by the Utah Division of Wildlife Resources in 1985 and 1986 were mostly ~. cypha; (e) ~. robusta also were collected in the Green River (e.g., along with~. cypha at Three Canyon, but alone at Whirlpool Canyon); (e) Westwater and Black Rocks on the mainstem Colorado River are occupied by ~. cypha and G. robusta, wi th morphological similarities between localities suggesting that they may belong to common gene pools; (f) ~. cypha and ~. elegans are morphologically so similar that their current status as two species might be questioned (however, the abili ty to distinguish between these species may have been confounded by evaluation of hatchery-reared ~. elegans which may have had a morphology different from wild ones). Overall, 2