9327

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

population. However, DeMarais (1992) examined a sample of 14
from this population and found results at AK-1 and and possibly
IDHP-2 consistent with those herein; he did not examine EST-2 or
PK-1 nor samples from Mexican populations.
All samples from the Culiacan, Sinaloa, Fuerte, and Mayo
drainages appear to conform to Hardy-Weinberg expectations at all
polymorphic loci (Appendix, Table 6). The AK-1 locus has
possible occurrences of rare alleles and a deficiency of expected
heterozygotes in the Mayo sample resulting in a significant
deviation (P<.05) from expected values in the chi-square analysis
but not the exact probabilites analysis; the latter indicates it
may not deviate significantly from random possibilities. The San
Bernadino sample is two small (n=4) for meaningful tests of
conformance; DeMarais (1992) reported four polymorphic loci in
his sample of 14 from this population but did not report results
of Hardy-Weinberg tests.
Genetic variability (Appendix, Table 5) in several samples
from Mexican Pacific drainages is relatively low, including the
combined Culiacan-Sinaloa-Fuerte sample (mean heterozygosity
.029), and Mayo (.017), lower Yaqui (.000), Tomochic (.006) and
Mdtape (.007) samples as compared to the mean calculated for
fishes in general (.051, Nevo et al., 1984) or that calculated
for several species of North American cyprinids (0.038 ± .008,
Avise, 1977) with only the first approaching that level.
Heterozygosities were not computed separately for the three
component samples combined in the CUL+ sample but examination of
scores clearly indicates that this attribute was increased by
this combination through cumulative diversity of heterozygotes at
GPI-1 and bGAL, though GPI-1 was totally homozygous in the SIN
sample and bGAL was so in the FUE sample. However, even taken
singly, these component samples exceeded the number of observed
heterozygotes in the Mayo sample and the others.
Echelle (1991) has discussed the relatively low variability
found in desert dwelling fishes, several having heterozygosties
in the range reported above for Gila, and hypothesized that
repeated bottlenecking (sensu Motro and Thomson, 1982) and
relative isolation are responsible for this characteristic. The
hydrographic histories of the respective Gila sampling sites are
not well enough known to corroborate or cast doubt on the
applicability of such a theory to Mexican populations. However,
populations in such large rivers as the Yaqui questionably would
be subjected to such phenomena. It should be noted that samples
from the upper Papigochic (JUN and GAV) and the San Bernadino
sample (though small), from relatively small habitats, exhibited
much higher heterozygosties (.040 - .071) than the above samples.
Taxonomic Implications
4