|
<br />A.{/vF!f '" ' I{ JIIf/ I ~
<br />:/' /'./u '^J a.S J
<br />
<br />
<br />
<br />,1
<br />
<br />i"
<br />
<br />Of al
<br />
<br />Copeia, 2001(2), pp. 389-400
<br />
<br />Use of Geometric Morphometries to Differentiate Gila (Cyprinidae)
<br />within the Upper Colorado River BasinI
<br />
<br />MICHAEL E. DOUGlAS, MAllIS R. DOUGlAS, JOHN M. LYNCH, AND DOUGlAS M. McELROY
<br />
<br />Video images of 215 adult Gila robusta and 148 endangered Gila cypha were col-
<br />lected from May 1991-0ctober 1992 at eight Colorado River basin localities (seven
<br />upper basins and one lower basin). The two species were sympatric at five of these
<br />locations; G. robusta was absent at one site, whereas G. cypha was missing at two
<br />others. Saggital views of each individual were videotaped and 25 morphological
<br />points (15 anatomical landmarks and 10 helping points) identified. Bookstein shape
<br />coordinates were calculated from Cartesian coordinates of these landmarks and
<br />points, whereas centroid size was used as a measure of body size. Shape differences
<br />were evaluated among populations of each species using MANOVA and canonical
<br />variates analysis. In G. cypha, variation encompassed three aspects: nuchal hump
<br />(most pronounced in Grand Canyon forms), relative head size (larger in Cataract
<br />Canyon fonns), and caudal peduncle dimensions (shorter with a tapering depth in
<br />Cataract Canyon forms but longer and uniformly deeper in those from Desolation
<br />Canyon). Nuchal development in G. robusta is slight, hence only head and peduncle
<br />dimensions distinguished populations. Those individuals from Cataract Canyon had
<br />relatively shorter peduncles that (again) tapered in depth from anterior to posterior,
<br />whereas G. robusta from Desolation Canyon possessed peduncles thatwere much
<br />longer and of uniform depth. Specimens from Debeque and Rifle Canyons had
<br />proportionally smaller heads. Variation among all 13 populations (i.e., both species
<br />together) was evaluated using relative warp analysis, with G. cypha and G. robusta
<br />clearly separated at all sympatric locations except those from Desolation and Car-
<br />atact Canyons. Here, body shapes of the two species converged. Overall, shape
<br />variation in both species is c1inal. Although results from our geometric morpho-
<br />metric analysis were statistically similar to those based on distances derived from a
<br />truss analysis, the geometric approach visually demonstrated phenotypic differences
<br />among populations and species and this, in turn, has management implications.
<br />
<br />As a discipline, multivariate morphometries
<br />. (sensu stricto) is just four decades old (Jol-
<br />icoeur and Mosimann, 1960; Blackith and Rey-
<br />ment, 1971). Arguably, it has had three revivals
<br />during this period. The first was an improve-
<br />ment in shape quantification, with nonortho-
<br />gonal distances being combined with horizon-
<br />tal/vertical ones to form a superimposed truss
<br />(Le., a series of abutting triangles) over the
<br />shape in question (Strauss and Bookstein,
<br />1982). The second was an analytical refinement
<br />in the quantification of body size and shape,
<br />which permitted each to be analyzed separately
<br />(Humphries et a!., 1981; Rohlf and Bookstein,
<br />1987; Bookstein, 1989a). The final was, appro-
<br />priately enough, a return to an older, more in-
<br />tuitive but (until recently) computationally in-
<br />tractable methodology espoused by D'Arcy
<br />Wentworth Thompson (1917). This new ap-
<br />proach now emphasizes three aspects: the exact
<br />definition of anatomical homologies (or land-
<br />
<br />1 This paper is dedicated to F. James Rohlf on the
<br />occasion of his 65th birthday.
<br />
<br />marks) among forms; the quantification of
<br />these landmarks in shape space; and (as in
<br />Thompson's case) the use of deformation grids
<br />among forms to actually visualize landmark-by-
<br />landmark transformations (i.e., to visualize and
<br />to quantify shape change as one form is super-
<br />imposed onto another). This new synthesis, apt-
<br />ly termed geometric morphometries, was for-
<br />mulated initially by Bookstein et a!. (1985) and
<br />continued to evolve through collaboration with
<br />other researchers, particularly during a series of
<br />workshops (e.g., Rohlf and Bookstein, 1990;
<br />Marcus et aI., 1993, 1996). Its uses now extend
<br />into ecological applications (Adams and Rohlf,
<br />2000). The mathematical tools for this reedifi-
<br />cation are provided in Bookstein (1991), where-
<br />as a brief, readable synopsis is given by Rohlf
<br />and Marcus (1993). The historical development
<br />of geometric morphometries is provided by
<br />Marcus and Corti (1996).
<br />As a technique, geometric morphometries is
<br />characterized by the following (as per Rohlf and
<br />Marcus, 1993): Capture of 2-dimensional (Le.,
<br />2-D) or 3-D coordinates from previously defined
<br />
<br />@ 200! by the American Society or Ichthyologists and Herpetologists
<br />
<br />L-/;J0 ~.
<br />
<br />
<br />
<br />
<br />
<br />
|