|
<br />DOUGLAS-SEXUAL DIMORPHISM IN GILA CYPHA
<br />
<br />341
<br />
<br />modestly different characters out of 53 (i.e.,
<br />3.8%) is certainly not evidence for pronounced
<br />sexual dimorphism. If, in fact, a conservative
<br />Bonferroni approach (Harris, 1975:96-101) is
<br />used to compensate for the evaluation of mul-
<br />tiple tests, the probability value for acceptance
<br />of a significant ANCOV A would be P > 0.001
<br />(0.05/53). Hence, no characters would be
<br />judged significant in the ANCOV A.
<br />Besides, one could argue that sexual dimor-
<br />phism is an evolutionary phenomenon that is
<br />generally pervasive when it occurs, being ex-
<br />hibited in numerous phenotypic characters (see
<br />for example, Schnell et a1. 1985: Douglas et aI.,
<br />1986). The two allegedly dimorphic characters
<br />in this analysis displayed little apparent function
<br />in sexual display or sexual attraction among
<br />cyprinids, which further challenges the idea that
<br />variation in these characters was sexuallv based.
<br />Results of the discriminant analysis '(Fig. 4)
<br />further discourage the idea that sexual dimor-
<br />phism is pervasive in G. cypha. Whereas the dis-
<br />criminant function was composed of three char-
<br />acters [i.e., characters (22), (33), and (48): Table
<br />1 ], only 60% of all individuals could be correctly
<br />segregated to sex through its application. Thi's
<br />was scant improvement over the original hy-
<br />pothesis that group membership for a given in-
<br />dividual was of equal (e.g.. :')0%) probability.
<br />Morphological shape differences among in-
<br />dividuals in this study are apparent, regardless
<br />of sex. These differences are primarily due to
<br />extent and development of the nuchal hump
<br />and to concomitant changes in concavity and
<br />length of the head, which accompanies this de-
<br />velopment (PC2 and H2, Tallie I). For exam-
<br />ple, distance from pupil to top of head [char-
<br />acter (3)] decreases as distance from pupil to
<br />bottom of head [character (4)] increases. These
<br />modifications point to the development of the
<br />scalloped cranium characteristic of mature G.
<br />cypha (Fig. 1 B). In addition, as scalloping pro-
<br />ceeds, there is a reduction in four distance mea-
<br />stires that extend from pectoral origin to vari-
<br />ous landmarks along the dorsal edge of the head
<br />and nape [characters (9)-(12), Table 1]. These
<br />reductions suggest a tendency for the head (from
<br />snout through approximately the nape) to
<br />shorten or become more compact as the con-
<br />cavity ofthe skull increases. Finally, those linear
<br />distances from nape through vertical pectoral
<br />[characters (24)-(25)] decrease or shorten as
<br />distances from vertical pectoral through mid-
<br />point between vertical pectoral and dorsalori-
<br />gin [characters (26)-(27)] increase. These trans-
<br />formations essentially describe the vertical
<br />development of the nuchal hump: they indicate
<br />that distances immediately preceeding the hump
<br />
<br />decrease or shorten to form a ramp, whereas
<br />distances involved directly with the hump itself
<br />lengthen or increase.
<br />Size-related shape changes in the head and
<br />nuchal region of G. cypha are complex. Com-
<br />parisons of the results from PCA and sheared
<br />PCA (PC2 and H2, Table I) suggested that the
<br />anterodorsal hump and its associated structures
<br />vary complexly as growth proceeds, with some
<br />characters increasing in magnitude while others
<br />decrease. How ontogeny influences the onset
<br />and development of this hump is still to be elu-
<br />cidated. However, sexual differences in the de-
<br />gree of nuchal development can clearly be ruled
<br />out as one hypothesis pertaining to the origin
<br />and/or maintenance of this unusual structure.
<br />
<br />ACKNOWLEDGMENTS
<br />
<br />Special thanks to D. A. Hendrickson (Uni-
<br />versity of Texas, Austin-UT) and C. O. Minck-
<br />ley (Northern Arizona University-N AU, Flag-
<br />staff) who organized and executed the Arizona
<br />Game and Fish Monitoring Program for hump-
<br />back chub (1987-90) and who invited my par-
<br />ticipation in the project. During this period,
<br />many individuals provided me with assistance
<br />ill the canyon: B. Bagley, C. Lutz, D. Papoulias,
<br />D. Valenciano, R. Van Haverbeke (all of AZGF),
<br />and P. Ryan and J. Nystadt (Navajo Natural
<br />Heritage Program). Grand Canyon National
<br />Park kindly allowed research within the park's
<br />boundaries. GCNP and AZGF provided permits
<br />to collect fish at the confluence, and Navajo
<br />Game and Fish supplied a collecting permit for
<br />the upstream Little Colorado River. The Unit-
<br />ed States Fish and Wildlife Service allowed me
<br />to engage in endangered fish research as a sub-
<br />permitee under Federal Permit 676811. The
<br />cooperation and assistance of all of these or-
<br />ganizations and agencies is appreciated. The bi-
<br />ology of Gila in general, and humpback chub
<br />in particular, was discussed earnestly and often
<br />with W. L. Minckley, P. C. Marsh, T. E. Dow-
<br />ling, and B. D. DeMarais (Arizona State Uni-
<br />versity), C. O. Minckley (NAU), D. A. Hen-
<br />drickson (UT), and D. Kubly (AZGF).
<br />Manuscript reviews were graciously provided
<br />by W. L. Minckley, P. C. Marsh, C. O. Minckley,
<br />T. E. Dowling, K. M. Somers, W. J. Rainboth,
<br />and three reviewers. This research was sup-
<br />ported by NSF-BSR-87-14552. Manuscript
<br />preparation was subsizided by BOR-I-FC-90-
<br />10490 and BOR-I-SP-40-0970A. Donations by
<br />Intel Corporation (Hillsboro, Oregon) to the
<br />endangered species program of MED are also
<br />acknowledged.
<br />
|