8235 - Laserfiche WebLink

Home Browse Search

168 DOUGLAS ET AL. FIGURE 2.-A three-dimensional plot depicting ar- rangement of 167 adult Gila robusta, G. cypha, and G. elegans in canonical variate space. Three groups are ap- parent: G. elegans (left rear), G. cypha (right rear), and G. robusta (foreground). The white spheres represent three G. elegans classified as G. cypha. The single stip- pled sphere represents a G. cypha classified as G. ele- gans. instead to G. cypha, and the 12 misclassified G. cypha were assigned to G. robusta. The score for one additional (correctly identified) G. cypha was virtually identical to that for G. robusta. The two misclassified G. elegans juveniles were assigned to G. robusta, and two other (correctly identified) G. elegans had scores virtually identical to G. cy- pha and G. cypha/G. robusta, respectively. Among the putative hybrids, G. robusta X G. cypha was assigned 60% of the time (6 of 10) to G. robusta and 40% (4 of 10) to G. cypha. Putative G. cypha X G. elegans was assigned 50% of the time (2 of 4) to G. cypha, and 50% (2 of 4) to G. elegans. Gila seminuda was categorized as G. ro- busta in 66.7% of cases (2 of 3) and as G. cypha in 33.3% (1 of 3). Discussion Discrimination of Species Holden (1968) and Holden and Stalnaker (1970) noted that G. robusta and G. elegans shared little morphometric similarity (81 % of the former and 89% of the latter could be segregated). Each phe- notype thus exhibited a high degree of within- group homogeneity, in spite of the fact that sam- ples were taken from throughout the Colorado Riv- er basin. The five characters which best discrim- inated between the two species were: dorsal and anal rays, head depth, length of upper jaw, and least depth of caudal peduncle. Our results support those of Holden and Stal- naker (1970). Of all species-pairs under evalua- tion, G. robusta/G. elegans was most easily sep- arated, and by all three character sets. However, our results are more parsimonious than those of Holden and Stalnaker, in that three characters (two meristic and one morphometric) were found to be maximally discriminating. Holden (1968) and Holden and Stalnaker (1970) noted that, although morphological relationships between G. cypha and G. elegans are unclear, hump and snout characteristics (Figure 1) are critical to their discrimination (G. elegans has a smooth hump and no overhanging snout; "typical" G. cy- pha has an abrupt hump and long, overhanging snout). Results from our study (Table 1) concur; G. cypha/G. elegans are indeed the most difficult of the three species-pairs to separate. Gila cypha, with its exaggerated nuchal morphology (Figure 1), had long been considered by fishery managers as a male G. elegans (reviewed in Douglas 1993). Holden and Stalnaker (1970) argued that al- though G. cypha and G. elegans are separate en- tities, they are completely bridged by a third group that lacks the internal cohesiveness or homoge- neity of the first two. Some members of the third group are more similar to G. cypha, and others resemble G. elegans. Holden and Stalnaker (1970) offered three hypotheses to account for this situ- ation: (1) incomplete speciation with subsequent introgressive hybridization; (2) one morphologi- cally variable species; or (3) two species, with in- tergrades being variants of one or the other. Smith et al. (1979) argued that the majority of intergrades in the Holden and Stalnaker (1970) study were actually G. cypha. The discrepancy was due to re- duced precision in the original analyses when con- tinuous data were collapsed to categorical. In this study, meristic data (Table 1) best seg- regated G. cypha and G. elegans. However, meris- tic and morphometric subsets used characters that were not useful for field analysis. The field subset (in spite of losing three important characters) still delineated individuals 95.3%. Trophic characters were apparent in all three data subsets; this at least offers the suggestion that some form of resource partitioning may occur between these species. Depth of caudal peduncle is also a discriminating character, suggesting that G. cypha (at least with regards to its peduncle) is less of a "bony tail" ~