|
<br />DOUGlAS ET AL.-GlLA GEOMETRIC MORPHOMETRICS
<br />
<br />395
<br />
<br />DISCUSSION
<br />
<br />
<br />Methodology.-With 0: = 1 (i.e., large-scale pat-
<br />terns emphasized), the RWA still depicted the
<br />nuchal hump as an overriding factor in shape
<br />discrimination among populations of G. cypha
<br />and between it and G. robusta. This would be
<br />expected when only the two species were com-
<br />pared, because nuchal characteristics are para-
<br />mount in discriminating among them (Douglas
<br />et aI., 1998). However, subtle discrimination
<br />among populations of G. cypha was surprising
<br />and attests to the power of the geometric ap-
<br />proach. Variation is also apparent in other body
<br />regions (Appendices 1-2), and these also con-
<br />tributed to discrimination.
<br />
<br />
<br />
<br />Previous research.,..-Numerous studies have been
<br />undertaken to clarify the taxonomic confusion
<br />surrounding G. robusta, G. elegans, and G. cypha.
<br />Most relied upon pre-1960s Colorado River sur-
<br />veys (i.e., preimpoundment surveys; reviewed by
<br />Holden, 1991). Holden (1968) evaluated 309
<br />individuals over 19 (of 35) characters. Speci-
<br />mens were: 130 G. robusta; 99 G. elegans; 16 G.
<br />cypha; 55 G. elegans X G. cypha; and three G. ro-
<br />busta X G. elegans. Holden concluded that G.
<br />ro/msta and G. elegans should be considered valid
<br />species. However, the relationship between G.
<br />cypha and G. elegans is clouded by the presence
<br />of what appeared to be intergrade forms. Spec-
<br />imens varied from typical cypha (i.e., abrupt
<br />hump and long snout) to typical elegans (i.e.,
<br />smooth hump and no snout). Holden (1968)
<br />also concluded that collections to that time
<br />were not sufficiently large to clearly demon-
<br />strate morphological differences between the
<br />two forms. Apparent intergrades suggested ei-
<br />ther introgressive hybridization or phenotypic
<br />variability associated with environmental condi-
<br />tions.
<br />Holden and Stalnaker (1975) also reported
<br />intergrades in the upper basin, whereas R. R.
<br />Miller argued these were actually G. cyPha (T.
<br />W. Joseph, J. A. Shinning, R. J. Behnke, and P.
<br />B. Holden, U.S. EPAjU.S.F.W.S., 1977, un~
<br />publ.). Other important studies were reviewed
<br />by Douglas et al. (1989), Douglas (1993), and
<br />McElroy and Douglas (1995). Douglas et al.
<br />(1998) again demonstrated (as did Holden,
<br />1968) that G. robusta was most easily separated
<br />from G. cypha and G. elegans.
<br />
<br />Present study.-Differences with regard to overall
<br />body size are apparent within both G. cypha and
<br />G. robusta. In the former, the Little Colorado
<br />population has the largest individuals. Those
<br />from Desolation, Westwater, and Black Rocks
<br />
<br />Canyons are intermediate in size, whereas fish
<br />from Yampa River and Cataract Canyon are
<br />smallest. For G. robusta, largest individuals are
<br />found in. Rifle and Debeque Canyons, whereas
<br />smallest are again at Cataract Canyon. Remain-
<br />ing populations are intermediate in size.
<br />Size-free shape coordinates tell a similar sto-
<br />ry: G. cypha from the Little Colorado River is
<br />most distinct, displaying classic humpback chub
<br />morphology (i.e., pronounced nuchal hump;
<br />small, concave head; thin caudal peduncle).
<br />However, Desolation and Cataract Canyon G. cy-
<br />pha show robusta-like morphology, with larger
<br />heads and reduced nuchal humps. Desolation
<br />Canyon G. cypha have longer and thicker pe-
<br />duncles irrespective of body size, whereas those
<br />from Cataract Canyon have more truncated pe-
<br />duncles. Both Desolation and Cataract Canyon
<br />G. cypha group closely with .G. robusta from the
<br />same location when both species were included
<br />in the same analysis. However, Desolation and
<br />Cataract Canyon G. cypha also reflect differenc-
<br />es in morphology when both were contrasted
<br />one against the other.
<br />The unique morphology of Cataract Canyon
<br />G. cyPha has been previously noted. R. A. Valdez
<br />(U.S. Bureau of Reclamation, 1988, unpubl.)
<br />observed that Cataract Canyon G. cypha were
<br />generally smaller fish as adults (i.e., < 200 mm
<br />TL), with blunt heads and overhung snouts,
<br />shallow nuchal humps, and scaleless napes and
<br />breasts (many of these characteristics were not-
<br />ed herein). Dorsal and anal ray counts of Cat-
<br />aract Canyon G. cypha were inconsistently 9 and
<br />10, respectively (which are typical counts for the
<br />species). Valdez argued that genetic studies
<br />were needed to ascertain whether these fish
<br />were indeed G. cypha, because they certainly dif-
<br />fered from the typical G. cypha phenotype. Our
<br />results (and those of McElroy and Douglas,
<br />1995) suggest these fish (and those from Deso-
<br />lation Canyon) are indeed phenotypically dif-
<br />ferent, appearing more robusta-like in their mor-
<br />phologies than do other G. cypha populations.
<br />Another difficulty with the Cataract Canyon
<br />G. cyPha population was noted by Karp and Tyus
<br />(1990), who stated that all upper basin G. cypha
<br />are presumed native except those in Cataract
<br />Canyon. The latter may be derived (at least in
<br />part) from a 1981 stocking of juvenile fish from
<br />Black Rocks. However, G. cypha from Cataract
<br />Canyon are phenotypically quite distinct from
<br />those at Black Rocks, particularly with regard to
<br />relative length and depth of head and relative
<br />length of caudal peduncle. This may indicate
<br />those G. cypha stocked in 1981 from Black Rocks
<br />did not survive to reproduce.
<br />Failure of indidgenous fishes to establish
<br />
<br />
<br />
|