9327

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

the eastern upper Yaqui drainage with samples from populations to
the north. Alleles shared by one or both samples of cf. pulchra
and samples from the Rio Grande-Pecos, Guzman, and Gila basins
but not Mexican populations here referred to as menacae occur at
AK-1, EST-2, and IDH-2 (Figs. 2,6,8). Alleles are shared by the
Los Gavilones cf. pulchra sample, but not the La Junta sample,
with samples from the Rio Grande-Pecos but not the Guzman Basin
at PK-1 (Fig. 12) and it, in fact, groups more closely with Rio
Grande samples than the cf. pulchra sample from La Junta.
However, an allele at IDH-2 shared by both cf. pulchra samples
but not the samples from basins to the north may serve to
genotypically distinguish among them. These differences bear
further investigation because the Guzman Basin sample, derived
solely from the Rio Mimbres of New Mexico, is very small (n=3).
The Gila population in the San Bernadino Creek portion of
the Yaqui in Arizona is currently known as Gila purpurea (Girard,
1856), the Yaqui chub, and this name was later applied to
populations inhabiting the rios Sonora and Mdtape to the south
(e.g., Miller, 1959). DeMarais (1991) later described samples
from these rivers, as well as one site in the Rio Moctezuma
portion of the Yaqui drainage as a separate taxon, Gila eremica,
the desert chub. Genetic differences cited above (e.g., at PK-1)
certainly support recognition of the San Bernadino and Matape
samples as taxa distinct from one another and from all others
examined. However, samples from the rios Sonora and Moctezuma
were not available to include in the analysis.
Miller (1945) and DeMarais (1991), based on morphology,
regarded purpurea and eremica as members of the Gila subgenus
Temeculina Cockerell (19019) along with G. orcutti (Eigenmann and
Eigenmann, 1890), which is endemic to the Los Angeles Basin,
?lifornia, and G. ditaenia Miller (1945) of the Rio Concepci6n
inage, Mexico and Arizona. However, while purpurea and
.mica were grouped as a pair by the UPGMA analysis (Fig. 13),
among loci examined herein, no alleles are exclusively shared by
that pair that might serve as markers for such a subgeneric
grouping. Further, DeMarais (1992), based on allozyme studies,
concluded that Temeculina was paraphyletic with regard to other
Gila spp., thus not forming a natural group. It might be borne
in mind that past introgression between members of this putative
grouping and, say., G. menacae, may have altered former patterns
of genetic variation by exchanging alleles which were derived
within the respective lineages of the hybridizing taxa (e.g.,
Smith, 1992); such a scenario would confound attempts to
reconstruct relationships via cladistic analyses and thereby
might incorrectly reject an hypothesis of monophyly for
Temeculina.
6