9327

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

samples taken to the north, enters the area only as a low
frequency occurrence in the Sycamore Creek sample from the Bill
Williams system.
Several loci treated herein were probably also studied by
DeMarais (1992) in lower Colorado Basin Gila populations with
overall consistent findings. His results at the "AK-A" locus are
highly congruent with those at AK-1 herein (Fig. 2) as were those
at CK-A (Fig. 5) and CBP-1 (Fig. 4), including, in the latter,
the minor polymorphisms in Burro (=Francis) Creek portion of the
Bill Williams system and Bonita Creek in the upper Gila sytem
with some additional polymorphism in his lower Eagle Creek
sample. DeMarais also found a similar pattern at LDH-B (Fig. 9)
with a slower allele present in upper Gila Basin samples with the
Cienega Creek sample homozygous for that allele. He found
minimal variation at "sIDHP-A" and "mIDHP-A", one of which may be
analogous to IDHP-2 (Fig. 8) herein but his Williamson Valley
sample did not score for the high incidence of the slow allele
scored herein and disagreements in scorings elsewhere (e.g., San
Bernadino) suggest that different loci are involved. DeMarais'
"PGM-A" may be analogous to PGM herein (Fig. 11) but, while
results were similar for most samples, there were contradictory
scorings (opposite alleles) for the West Clear Creek and
Pahranagat Valley samples and he did not report polymorphism in
the Grand Canyon sample reported herein.
Heterozygosities (Appendix, Table 5) are relatively low (as
compared to other cyprinid species, e.g., Avise, 1977) to
nonexistent in several Gila samples from the Gila Basin while
those from the Bill Williams system were moderate (e.g., .042 in
Francis Creek). In the Gila Basin, extreme cases are the total
lack of heterozygosity in the Fossil and Cienega creeks samples;
relatively low levels (<.030) occur in several population
samples, including those from Blue River, Williamson Valley, East
Verde River, and Cherry Creek and moderate levels (.030 - .050)
occur in several others.
Sixteen samples from the Gila Basin were at variance with
Hardy-Weinberg expectations at one or more loci (Appendix, Table
6). Twelve deviated significantly at the EST-2 locus, which may
have problematic results in some cases as discussed below, but
only five of these deviated at other loci. Samples from Tonto
and Cienega creeks, Harden Cienega, and Black River deviated at
one other locus and the upper Verde River sample (VDP) at two
because of the heterozygote difficiencies caused by the
occurrence of single homozygotes for an uncommon allele with no
heterozygotes present; in three cases, this was at MPI-2.
Reviews of scoring for these samples indicate that these
occurrences should be real, though the presence of such rare
homozygotes in the absence of heterozygotes in samples of these
sizes (usually 20) exceeds the probability of random occurrence
as indicated by the exact probabilities analyses in Table 6.
13