DH -Texas Memorial Museum, Texas Natural History Collections, University of Texas, Austin, TX 78712-1100 USA; JK -Dept. Zoology, University of Texas, Austin, TX 78712
Biogeography, ecology and conservation status of the Mexican blindcats, genus Prietella (Ictaluridae)
Recent explorations have revealed that the two species of the blind, troglobitic Ictalurid genus Prietella are more widely distributed, and probably more abundant, than formerly realized. The Mexican blindcat, P. phreatophila, has now been collected at many sites scattered throughout northern Coahuila, ranging from the type locality near Múzquiz to 160 km N, very near the Rio Grande. Additionally, numerous local residents widely scattered around Coahuila report periodically seeing blind catfishes in well buckets used at hand-dug wells which intersect natural fractures, or have seem them in surface flows from such wells following storms, but our repeated attempts to collect specimens at these sites have remained unproductive. Four specimens taken at a site about 50 km north of the type locality for P. lundbergi, and 600 km SE of the nearest known locality of P. phreatophila, are believed to represent P. lundbergi, previously known from only a single specimen. Judging by its invertebrate fauna, and water chemistry, the isolated sump from which the P. cf. lundbergi were collected has long been isolated from large, adjacent spring discharges. While the genus lives below surface habitats ranging from moist sub-tropical forests to arid Chihuahuan desert, sub-surface habitats so far examined share a relatively narrow temperature range, an absence of other fishes, and floods following local or nearby rains, thus indicating nearby recharge areas. These range, however, from large and permanent streams with continual surface discharges, through isolated 60 m deep sumps, to shallow, intermittent streams. In some sites without flow, physico-chemical stratification was observed, and fish distributions appeared to indicate pronounced preferences. In general, the caves in which these fishes occur are typically fracture-based systems, and those sampled so far are not either extremely large or deep (70 m). Though these data indicate that habitats are closely tied to local meteoric water sources, genetic data indicate little geographic variation within P. phreatophila, but considerable differentiation between the two species. This seems to indicate no subterranean aquatic connections between Coahuila and Tamaulipas, but probable interconnectedness of aquifers across much of Coahuila, and relatively high rates of gene flow over substantial distances in that state. It seems likely that most sites we have been able to sample are peripheral, and larger populations likely exist in deeper, inaccessible parts of the aquifer. It is estimated that the range of P. phreatophila may cover about 13,000 km2 in N. Coahuila, and it would not be surprising if populations were to be found in Texas in the area between Big Bend and Eagle Pass. Populations continue to be maintained in the lab, but varied attempts to induce reproduction in captivity, including hormone injections, have failed.
García de León,F.*;Hendrickson,D.A.;Hillis,D.M.
FG -Instituto Tecnológico de Ciudad Victoria, Cd. Victoria, Tamaulipas, México; DAH -Texas Memorial Museum, Texas Natural History Collections, University of Texas, Austin, TX 78712-1100 USA; DMH -Zoology Department, University of Texas, Austin, TX 78712
Molecular phylogeny of Ictaluridae with emphasis on the Mexican blindcats, genus Prietella
Sequences were obtained for mitochondrial cytochrome-b (648 bases)
and 12S rRNA (390 bases) genes of representative species of Ictaluridae
(Ameiurus natalis, Ictalurus punctatus, Pylodictus olivaris,
Noturus insignis, N. gyrinus, N. flavus, Prietella
cf. lundbergi, P. phreatophila) and a single representative
of the family Bagridae (Mystus sp.), considered on the basis of
morphological characters to be the sister group of the Ictaluridae (Pinna
1993). Phylogenetic relationships were investigated in PAUP version 4.0.0d56
(D. Swofford, Smithsonian Institution), using the combined gene sequences
with branch and bound and the maximum parsimony optimality criterion. A
total of 194 characters were found to be informative. The single best tree
(((((Ameiurus(Ictalurus(Prietella(Pylodictus(Noturus)
had a score of 682, and consistency index of 0.6716. The next best tree
was 2 steps longer. Prietella was monophyletic in the optimal tree,
but constraining the tree to include the traditional Prietella +
Noturus sister-group relationship (Taylor 1969; Amemiya et al.
1986; Lundberg 1982, 1992) required an additional 9 steps (C.I. = 0.6628),
and 100 of the possible trees were more parsimonious than any tree which
included this traditional relationship. A likelihood ratio test indicated
that the traditional Prietella + Noturus relationship is
a significantly poorer fit to the data compared to the optimal tree, although
some other alternatives are not significantly different from the placement
of Prietella as the sister-group to Pylodictus + Noturus..
Divergence between the two species of Prietella is substantial,
but the P. phreatophila analyzed (22 specimens from 8 sites) were
nearly invariant for both genes, with only single base substitutions in
each gene in the northernmost population. It thus appears that Prietella
lundbergi in Tamaulipas has long been isolated from congeners in Coahuila,
but that subterranean interconnections among aquifers of Coahuila have
allowed for relatively recent or continuing gene flow among populations
of P. phreatophila.