9327

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

the Gila Basin and Bill Williams system have been regarded as
three taxa. The greatly attenuated bonytail chub, Gila elegans
Baird and Girard (1853), formerly occurred in larger streams,
such as the lower Gila River and mainstem Colorado, but has been
extirpated for decades except in Lake Mohave. The more
generalized roundtail chub, Gila robusta Baird and Girard (1853),
inhabited rivers and larger creeks of the region and, while
extirpated from much of its former range, fragments of what were
probably formerly more continuous populations remain in several
larger creeks of the Gila Basin and Bill Williams system. The
third species, the thick-bodied Gila chub, G. intermedia (Girard,
1856), typically inhabited smaller tributary habitats, including
springs, and, again, while many populations have been extirpated,
several persist at isolated localities. Samples from many
populations are clearly separable into these three entities on
shape and meristic characters (Rinne, 1976; DeMarais, 1986;
Douglas, herewith). However, some problematic populations,
showing various degrees of morphological intermediacy between
robusta and intermedia, have also been identified in streams in
the Gila Basin, mostly in tributaries to the Verde River and in
the upper Gila proper. Rinne (1976) applied the epithet grahami
Baird and Girard (1853) to these populations, treating them as a
subspecies of G. robusta. The extant populations of Gila forms
are broadly sympatric in the Gila Basin but are generally not
known to be syntopic (Rinse, 1976). Contacts between the various
forms may therefore be relatively rare.
Alternate hypotheses have been advanced as explanations for
the array of Gila forms occurring in the Gila Basin. Miller
(1946) theorized that intermedia and other forms (exclusive of
elegans) were "ecological subspecies" (i.e., genetically
determined) of G. robusta. Rinne (1976) hypothesized that
robusta, intermedia, and and the morphologically intermediate
"grahami'l constituted independant lineages with "grahamill closest
related to robusta. Smith et al. (1979) regarded these forms and
all others in the Colorado Basin (e.g., bonytail, humpback chubs)
as members of a single "superspecies" complex, a term applicable
to a group of closely related species (Mayr, 1969) and thus one
of questionable utility beyond species complex or even genus, and
one which might equate to several definitions of "species flock"
posed in papers compended in Echelle and Kornfield (1984).
DeMarais (1986), based on morphological and geological evidence,
held that robusta and intermedia constituted distinct lineages
but that phenotypically intermediate populations had arisen
through multiple hybridization events. These occurred when those
taxes were brought into occasional contact by various events in
the Pliocene which integrated formerly separate basins into the
present configuration of the lower Colorado and Gila basins.
DeMarais (1992) again advanced this scenario as the most viable
hypothesis based on studies of allozyme variation. One
hypothesis that has not been advanced to date is that the robusta
form of Gila may be of hybrid origin between an intermedia-
18