9327

• UCREFRP Catalog Number: 9327
• Author: Starnes, W. C.
• USFW Year: 1995.
• Copyright Material: NO

nigrescens-like form and another Colorado Basin form, such as G.
elegans. Hybrid origin has been hypothesized for another taxon
in the genus as will be discussed later. Though discussed, none
of the major studies has proposed agenetic ecophenotypy as
explanation of the extent of variation observable in Colorado
Basin Gila forms.
Testing of the various hypotheses with data presented
herein does not unambiguously support any particular one but may
be sufficient to falsify others. Niether patterns of allozyme
variation described herein, nor those previously described by
DeMarais (1992), provide evidence for the existence of
reproductively isolated taxonomic units corresponding to
morphological entities discussed above. However, clearly, the
constrasting morphologies among various forms of Gila must have
genetic bases. Several have been propagated in hatcheries (e.g.,
Hamman, 1981) under similar conditions and develop true to
parental morphology. While allozymes may not be expressed by
totally neutral genes and are thus subject to some selection
(e.g., Selander, 1970; Mueller et al., 1985), they probably are
not subject to the same regimes or degree of selection as those
which govern morphology, effectively decoupling them. Thus,
while allozymes may corroborate reproductive isolation with fixed
differences, the converse is not true; lack of such differences
does not necessarily indicate ongoing gene exchange. Mostly
allopatric taxa which have come in contact from time to time and
exchanged genetic material through introgression may
morphologically "sort themselves back out" by backcrossing
selection for those hybrids which are morphologically most
similar to the predominant resident form (e.g., an intermedia
population that has experienced a small incursion of robusta
genes or vice versa) but are not necessarily allozymically more
similar. Thus, alleles for allozyme loci exchanged in this
manner may remain in the population virtually indefinitely or be
selected against much more gradually by different processes.
It is questionable how a "superspecies" concept (sensu Smith
et al., 1979) could be tested or just what the genetic
configuration of such an entity would be; presumably, however,
consistent differences would be expected among its components
which is not the case with lower basin Gila. Rather, Gila
populations under study have no demonstrable reproductive
isolation based on allozyme studies, and thus fit the criteria of
being a single "biological species" rather than a complex of
separate ones. Yet the extent and patterns (lack of homogeneity),
of variation among them are more indicative of contributions from'.
two or more groups with largely separate histories. These groups
each comprise populations of common origin which have shared
genetic histories and have thus primarily responded to selection
collectively (thus not fitting the criteria for parallel
individually selected "ecotypic" populations either). While the
lack of fixed differences at allozyme loci among Gila populations
19