Development of the hybrid swarm between Pecos pupfish (Cyprinodontidae: Cyprinodon pecosensis) and sheepshead minnow (Cyprinodon variegatus): a perspective from allozymes and mtDNA.
Echelle and Connor (1989) invoked extremely rapid selection to explain the observed genetic changes because genetic swamping as a direct result of the magnitude of the introduction is difficult to envision for Pecos pupfish. The species is a small omnivore, its diet being primarily diatoms and detritus (Davis 1981), and, for a similar pupfish, Naiman (1976) documented extremely high productivity. Correspondingly, C. pecosensis is abundant in waters that, like most areas occupied by the hybrids, are too saline to support a diverse freshwater fish fauna (Echelle and Echelle 1978). This, together with the likelihood that C. variegatus was introduced as an incidental release of a small (but unknown) number of individuals, makes genetic swamping unlikely under normal circumstances.
For added information on genetic structure, we include both mitochondrial DNA (mtDNA) and the previously examined allozymes in a survey of variation among hybrid populations. Frequency comparisons of species-specific cytoplasmic (mtDNA) and nuclear (allozymes) markers can give considerable insight into the dynamics of genetic introgression (Lansman et al. 1981; Dowling et al. 1989; Rand and Harrison 1989). In our analysis, we reasoned that equivalent, within-sample levels of introgression across a variety of markers would be compatible with genetic swamping of C. pecosensis by large numbers of C. variegatus. Divergent frequencies would suggest alternative explanations (Ferris et al. 1983; Powell 1983; Dowling and Hoeh 1991; Dowling and Childs 1992). From our results and repeated reports of massive fish kills in the study area (Rhodes and Hubbs 1992), the most plausible explanation is genetic swamping during a period of catastrophically low population densities for the endemic species.
Pupfish were sampled from 10 localities [ILLUSTRATION FOR FIGURE 1 OMITTED]. For each site in the following list, our site designation is followed by parentheses showing first the site number used by Echelle and Connor (1989) and then the one used by Wilde and Echelle (1992) - hyphens indicate that the site was not sampled: P1 (-, P7); A (-, P3); B (3, near P6); C (7, 2); D (8, 3); E (14, 8); F (18, 11); G (-, 14); V1, Lake Balmorhea, 3 km S Balmorhea, Reeves Co.; V2, Edinburg Water Supply Canal, 1 km N Edinburg, Hidalgo Co. Dates of collection were June 1991 for sites P1 and B-F, April 1992 for V2, and August 1992 for V1, A, and G.
Standard methods of horizontal starch-gel electrophoresis (Murphy et al. 1990) were used to examine allozyme products of four gene loci exhibiting fixed or nearly fixed differences (Echelle and Connor 1989) between C. pecosensis and C. variegatus: alcohol dehydrogenase-1 (ADH-1, EC 188.8.131.52), esterase-1 (EST-1, EC 184.108.40.206), glucose-6-phosphate isomerase-A (GPI-A, EC 220.127.116.11), proline dipeptidase-1 (PEPD-1, EC 3.1.1.-). Tissues and buffer systems were as follows: ADH-1 and PEPD-1, liver, tris-citrate buffer, pH 8.0 (Shaw and Prasad 1970); EST-1, eye and brain combined, Tris-borate-EDTA buffer system (Turner 1983); GPI-A, eye and brain combined, discontinuous lithium hydroxide buffer system (Selander et al. 1971).
Genomic DNA was extracted, with minor modifications, following Hillis et al. (1990) and digested with two restriction enzymes (HindIII, XhoI). Fragments were separated in 1% agarose gels, transferred to nylon membranes (Southern 1975), hybridized to probe mtDNA labeled with [32P]dCTP (random priming kit, U.S. Biochemical), and visualized by autoradiography. Hybridization conditions and washes followed Davis (1986). Probe DNA consisted of a cocktail of pUC-19 plasmid clones of two fragments of XbaI-digested mtDNA from Cyprinodon macularius. The probe DNAs represent the 4.5- and 4.7-kilobase XbaI fragments mapped by Echelle and Dowling (1992).
Allozyme data were analyzed using BIOSYS-1 (Swofford and Selander 1981). The exact-probability test was used to find deviations from Hardy-Weinberg genotypic proportions. Fixation-index values (F) were calculated, and tests for allele frequency differences between samples taken in different years were performed using the heterogeneity chi-square option. Burrows's composite measure of gametic disequilibrium (D; Weir 1990) was calculated from genotypic frequencies for each pairwise combination of allozyme encoding genes. To test the hypothesis that D equals zero, we used the [[Chi].sup.2]-distributed statistic Q (Weir 1990). We report the standardized measure of linkage disequilibrium (D[prime] = D/[D.sub.max]; [D.sub.max] is the maximum possible value for the sample; Hedrick 1983). Tests of cytonuclear (allozyme vs mtDNA) gametic disequilibrium followed Asmussen et al. (1987). The data used to assess gametic disequilibrium are available elsewhere (Childs 1993). In reporting statistical significance for multiple, within-site tests, we use the sequential Bonferonni method (Rice 1989) to ensure a maximum Type I error rate of 0.05.
The reference samples of Cyprinodon pecosensis and C. variegatus were effectively fixed for different alleles at the four allozyme-encoding gene loci, although there was a low level ([less than] 0.05) of allele sharing at ADH-1, EST-1, and PEPD-1 (Table 1). All samples from areas previously exhibiting genetic introgression had high frequencies of introduced (C. variegatus) allozymes except the sample from lower Salt Creek (site B, [ILLUSTRATION FOR FIGURE 1 OMITTED]) where a low level of introgression ([less than] 0.03) occurred in samples taken in 1988 (Wilde and Echelle 1992). We again detected low frequencies ([less than]0.03) of introduced alleles at this site in 1991 (Table 1). This site of low introgression is excluded in further references to "hybrid populations."
We resolved an introduced GPI-A allele (GPI-[A.sup.c]) that previously had not been detected in the hybrid populations. The product of this allele migrated slightly faster than the product [TABULAR DATA FOR TABLE 1 OMITTED] typical of reference samples of C. pecosensis; the two products comigrated in previous assays with a different buffer system (Echelle and Connor, 1989; Wilde and Echelle 1992). The GPI-[A.sup.c] allele occurred at a frequency of 0.15 and 0.07, respectively, in our reference samples of C. variegatus from V1 and V2. It was absent in our reference sample of C. pecosensis and, in the six hybrid populations, its frequency varied directly with the mean frequency of introduced alleles across the other three loci (Pearson's r = 0.999, P [less than] 0.001). Thus, GPI-[A.sup.c] apparently was introduced into the Pecos River with C. variegatus.
In all samples from hybrid populations, GPI-[A.sup.c] formed a much higher percentage (29-48%) of the introduced GPI-A alleles than would be predicted from its apparent frequencies in C. variegatus. A slow electromorph presumably representing GPI-[A.sup.c] has been reported in only 8 of 31 samples of C. variegatus (Darling 1976; Echelle and Connor 1989; Wilde and Echelle 1992), including Darling's (1976) analysis of samples from 28 coastal Atlantic and Gulf of Mexico localities; in eight native populations of the species from Texas, the maximum recorded frequency is 0.07. The highest frequency so far detected in C. variegatus is the value of 0.15 for our sample from the introduced population in Lake Balmorhea (site VI).
A high GPI-[A.sup.c] frequency seems characteristic of all hybrid populations, including those documented earlier but not sampled in this study. Earlier studies reported a consistently low estimate of introgression for GPI-A relative to other allozyme loci in all of the more than 25 sites where hybrids have been collected (Echelle and Connor 1989; Wilde and Echelle 1992; Wilde 1994). Failure to resolve GPI-[A.sup.c] explains this observation.
HindIII and XhoI provided mtDNA restriction fragment patterns diagnostic for the two species (Table 1). A single mtDNA haplotype (type A) was detected in both reference samples of C. pecosensis from Salt Creek. Two different haplotypes were found in the reference samples of C. variegatus - haplotype B in the sample from Lake Balmorhea and haplotypes B and D in the sample from Edinburg (Table 1). Haplotype D is not discussed further because it was not detected in hybrid populations. Haplotype B was the common introduced haplotype in hybrid populations. The observed haplotype B fragments [ILLUSTRATION FOR FIGURE 2 OMITTED] conform with expectations from published (Echelle and Dowling 1992) restriction-site maps for C. variegatus and the source species for our 9.2-kb mtDNA probe; no such map exists for C. pecosensis.
Individuals from hybrid populations typically exhibited mtDNA haplotype A of C. pecosensis or haplotype B of C. variegatus (Table 1, [ILLUSTRATION FOR FIGURE 2 OMITTED]). The only exception was a unique haplotype found in Red Bluff Reservoir. This third haplotype (haplotype C, [ILLUSTRATION FOR FIGURE 2 OMITTED]) had a HindIII fragment pattern identical to haplotype B except for the addition of a 2.3-kb fragment. Accordingly, the fragments produced by XhoI suggested that haplotype C was approximately 18.7 kb, about 2 kb larger than reported for C. variegatus by Echelle and Dowling (1992). This haplotype was likely derived from haplotype B through a large length-mutation in a HindIII fragment that overlaps our probe and was not detected in haplotype B because of its small size (ca. 300 bp). Such a fragment occurs in C. variegatus and eight other species of pupfish (Echelle and Dowling 1992).
With the sequential Bonferroni correction for four within-sample tests, we detected only one significant deviation from Hardy-Weinberg proportions for allozyme genotypes in the six samples from hybrid populations. This was a PEPD-1 heterozygote deficiency (P = 0.005) at Horsehead Crossing. Fixation indices (9 positive, 15 negative) indicated no consistent pattern of heterozygote deficiencies.
Temporal and Spatial Variation
Geographic variation in allozyme and mtDNA frequencies in hybrid populations in 1991-1992 [ILLUSTRATION FOR FIGURE 3 OMITTED] generally conformed with the clinal patterns of introduced allozymes in 1985 and 1986 (Echelle and Connor 1989; Wilde and Echelle 1992). However, a change in amplitude of the cline is indicated. At Pecos, the site where frequencies of introduced alleles were highest in 1985 and 1986, the frequencies for all four loci were lower in 1991-1992 [ILLUSTRATION FOR FIGURE 4 OMITTED]; this difference was statistically significant for the EST-1 locus [ILLUSTRATION FOR FIGURE 4 OMITTED]. All other sites (A, C, E-G) showed the opposite pattern. With one exception (ADH-1 at Pandale), frequencies of introduced alleles at these sites were higher in 1991-1992 than in both 1985 and 1986. Also at these sites, five of the 32 tests comparing frequencies of introduced alleles in 1991-1992 with those in 1985 or 1986 indicated statistically significant increases in 1991-1992, none indicated a decrease [ILLUSTRATION FOR FIGURE 4 OMITTED]. These results parallel those for mean, within-sample allozyme frequencies [ILLUSTRATION FOR FIGURE 5 OMITTED].
To assess whether the observed changes were stochastic in nature, we performed concordance tests on the rank abundances of introduced alleles for the four allozyme loci. Concordance was significant among samples in 1985 (sites C-E Kendall's W = 0.75, P [less than] 0.05; data from Echelle and Connor, 1989) and 1986 (sites A, C-G, W = 0.58, P [less than] 0.05; data from Wilde and Echelle, 1992) and nearly significant in 1991-1992 (W = 0.40; 0.05 [less than] P [less than] 0.10). Significant concordance also occurred across all sites and years (Kendall's W = 0.57; P [less than] 0.001; GPI-A not included because of the previously mentioned scoring problem). These results are consistent with directional change mediated, as discussed later, by gene flow in neutral clines.
We found little evidence for within-site interlocus differences in frequencies of diagnostic allozymes. The only significant differences occurred at Horsehead Crossing (2 x 4 heterogeneity [[Chi].sup.2] = 10.9, 3 df, P = 0.012). There also was little within-sample difference between observed numbers of introduced mtDNAs and the numbers predicted from the frequencies of introduced allozymes (averaged across the four allozyme loci). The observed number was greater than predicted at sites A and G, equal at site C, and less at sites D, E, and F. The only significant difference was in the sample from Pandale (G = 4.5, P [less than] 0.05), where the observed frequency of introduced mtDNA (0.93) was greater than predicted (0.75).
Associations between Loci
The only statistically significant linkage disequilibrium value was the one for the GPI-A/EST-1 combination at site G (Table 2). GPI-A and EST-1 appear to be physically linked in pupfishes (Echelle and Connor 1989; Wilde and Echelle 1992), a conclusion that remains valid despite the previously mentioned GPI-A scoring difficulty in earlier studies. Those studies reported much larger D[prime] values for GPI-A/EST-1 than for all other locus pairs (including GPI-A/ADH-1 and GPI-A/PEPD-1), a bias not explained by the scoring problem for GPI-A.
Sample sizes were too small and, at some sites, allele frequencies were too extreme for assessments of the statistical significance of individual disequilibrium values for unlinked gene loci (Brown 1975). Overall, however, disequilibrium has declined since the earlier studies (Echelle and Connor 1989; Wilde and Echelle 1992) of the hybrid populations. At equilibrium, D[prime] values should be distributed symmetrically around zero (Forbes and Allendorf 1991). In our samples there were five values of 0.000 and 15 positive and 16 negative values. In contrast, hybrid populations exhibited excess positive values in 1985 (24 positive, 6 negative; [[Chi].sup.2] = 10.8, P [less than] 0.005) and 1986 (34 positive, 14 negative, [[Chi].sup.2] = 8.3, P [less than] 0.005).
The cytonuclear (mtDNA/allozyme) disequilibrium values are available elsewhere (Childs 1993). Within-sample tests revealed no sexual differences in frequencies of mtDNA haplotypes or allozyme-encoding alleles characteristic of the two species (2 x 2 G-tests; P [greater than] 0.10 in all tests). Thus, the sexes were not analyzed separately. None of the disequilibrium values was statistically significant and there was no excess of positive or negative values (12 positive, 12 negative).
System Approaching Equilibrium
Our allozyme results indicated relatively little change in the hybrid pupfish populations in the Pecos River since the earlier surveys in 1985 and 1986. A survey encompassing more localities (Wilde 1994) supports this conclusion and indicates that the observed allele frequency changes are generally attributable to the reduction in between-site heterogeneity expected (Endler 1977) as a result of gene flow in neutral clines. These observations contrast markedly with the apparently rapid changes that occurred over more than 500 river-kilometers in the early history of the hybrid populations. Within five or fewer years, beginning sometime between 1980 and 1984, 20% to 80% of the endemic genome at different localities had been replaced by the introduced genetic material of C. variegatus (Echelle and Connor 1989).
The genetic patterns in this system were similar to those observed in a variety of hybrid zones (recently reviewed by Harrison 1993). Individual samples exhibited random association among nuclear (allozymes) and cytoplasmic (mtDNA) markers, and, among samples, there were concordant allele frequency shifts among loci. In situations involving massive introductions (e.g., western trouts - Gyllensten et al. 1985; Forbes and Allendorf 1991), such observations could be explained as a result of genetic swamping by large numbers of introduced fish, but, as previously discussed, this is difficult to envision under normal population densities for the Pecos pupfish.
Early History of Introgression
Fish Kills as a Factor
One possible explanation for the Pecos pupfish situation is that the introduction of C. variegatus may have occurred at a time when the native pupfish had been catastrophically depleted, and was therefore susceptible to genetic swamping. This would explain the rapid spread of introduced genetic material in the Pecos River and the observed rank concordance of allele frequencies among sample localities.
Pecos River fish kills in the study area have been reported by landowners since the 1950s (Rhodes and Hubbs 1992). Traditionally, these kills were attributed to high salinities or pollution from oil field activities, but, more recently, toxins released during blooms of a chrysophyte alga (Prymnesium parvum) have been implicated (James and De La Cruz 1989; Rhodes and Hubbs 1992). Massive kills extending over more than 100 river-kilometers occurred in November-December of 1985, 1986, 1988, and 1989 (Rhodes and Hubbs 1992; J. Brooks, pers. comm.). The 1988 kill extended from upstream of Red Bluff Reservoir downstream about 420 km and virtually all of our study area has experienced massive fish mortalities since 1985 (Rhodes and Hubbs 1992). No massive kill was reported for the period when the hybrid pupfish populations became established (sometime between 1980 and 1984), but the area is remote and such a kill might have gone unreported.
The relative lack of introduced markers in the Salt Creek population of C. pecosensis provides support for the role of a fish kill in explaining the genetic structure of hybrids in the Pecos River. Salt Creek supports a dense population of C. pecosensis, the bulk of which may have escaped the fish kills. Observations following the 1989 kill indicated that Salt Creek fishes were unaffected, whereas those in Red Bluff Reservoir and downstream in the Pecos River were severely depleted (J. Brooks, pers. comm.). Movement of Pecos River hybrids into Salt Creek is impeded by a series of small waterfalls. However, in 1987, 1988, and 1989, the spillway for Red Bluff Reservoir overflowed into Salt Creek, perhaps accounting for the low level of introgression ([less than]3%) that now occurs in lower Salt Creek (Wilde and Echelle 1992). Probably because of the dense population of C. pecosensis, there has been no explosion of hybrids like the one that occurred in the Pecos River in the early 1980s.
The anomalously high frequency of the introduced GPI-[A.sup.c] allele in the hybrids suggests that a single sample of C. variegatus was involved in the origin of the hybrid populations. The highest frequency of GPI-[A.sup.c] so far detected in C. variegatus was in the Lake Balmorhea population (0.15 versus 0.00-0.07 in other samples from Texas), which was established by the 1960s, apparently as a result of bait transport by humans (Stevenson and Buchanan 1973). Founder effect in a sample from Lake Balmorhea would explain the high frequencies of GPI-[A.sup.c] in the hybrids (29-48% of the introduced GPI-A alleles).
Echelle and Connor (1989) suggested that the initial introduction of C. variegatus occurred in the vicinity of Pecos where there was a 4:1 ratio of introduced to native allozymes. This now seems untenable because of the discovery of introgression [TABULAR DATA FOR TABLE 2 OMITTED] upstream of the high ([greater than] 50 m) dam forming Red Bluff Reservoir, where pupfish exhibit the high frequency of GPI-[A.sup.c] found in all other hybrid populations. Echelle and Connor (1989) sampled upstream of the dam, but at a site approximately 30 km beyond the reservoir where, as of this writing, there was no evidence of genetic introgression (A. A. Echelle, pers. obs.).
It seems likely that, if only a single introduction of C. variegatus was involved in the evolution of the hybrid populations, it would have occurred in Red Bluff Reservoir. Existing frequencies of introduced markers in Red Bluff Reservoir would reflect the relative abundances of the two species at the time of reproductive fusion, an event that might have been delayed in a large reservoir with patchy habitat from which the native species had been virtually eliminated. Dispersal of C. variegatus from Red Bluff Reservoir before reproductive fusion with C. pecosensis could explain the presently higher frequencies of introduced markers in some downstream areas relative to Red Bluff Reservoir. This presumes that C. variegatus or hybrids expanded in population size and dispersed throughout the study area before C. pecosensis was able to return to normal population densities, possibly as a result of differential selection during a period of expanding population size.
Absence in downstream populations of one of the two introduced mtDNAs (haplotype C) in Red Bluff Reservoir may be attributable to founder effect. The effective population size for mtDNA in vertebrates is about four times smaller than that for nuclear genes, heightening its susceptibility to genetic drift (Birky et al. 1989). Correspondingly, the variation we observed in mtDNA frequencies is greater than that in the nuclear markers [ILLUSTRATION FOR FIGURE 3 OMITTED]. Haplotype C was absent in the 78 specimens of C. variegatus we examined and in 36 additional specimens we have assayed from three coastal localities on the Gulf of Mexico. This mtDNA variant may represent a length mutation that occurred in Red Bluff Reservoir and then attained a moderately high frequency (15%) at a time of low population density.
In summary, the repeating nature of massive fish kills over large areas of the Pecos River has made genetic swamping at a time of low population density plausible as an explanation of the genetic structure of the hybrid populations. The minimal amounts of change detected in frequencies of introduced markers since the early 1980s seems generally attributable to gene flow in neutral clines. Our analysis does not exclude the possibility of rapid natural selection in the establishment of the hybrid populations. Differential selection during a period of increasing numbers of pupfish might explain how C. variegatus or hybrids were able to expand in population size before the endemic species was able to return to high population densities.
We thank J. Derr and S. Davis for cloning the probe DNAs, L. Richardson and J. Gold for guidance on Southern blotting, R. Edwards for specimens of C. variegatus, N. Ashbaugh and A. F. Echelle for assistance, J. Brooks and G. Garrett for helpful information, and S. Fox and K. McBee for reviewing an early draft of the manuscript. Financial support provided by the National Science Foundation (DEB 9208264).
ASMUSSEN, M. A., J. ARNOLD, AND J. C. AVISE. 1987. Definition and properties of disequilibrium statistics for associations between nuclear and cytoplasmic genotypes. Genetics 115:755-768.
BIRKY, C. W., JR., P. FUERST, AND T. MARUYAMA. 1989. Organelle gene diversity under migration, mutation, and drift: Equilibrium expectations, approaches to equilibrium, effects of heteroplasmic cells, and comparison to nuclear genes. Genetics 121:613-627.
BROWN, A. D. H. 1975. Sample sizes required to detect linkage disequilibrium between two or three loci. Theoret. Pop. Biol. 8:184-201.
CHILDS, M. R. 1993. Dynamics of introgressive hybridization between Pecos pupfish (Cyprinodon pecosensis) and sheepshead minnow (C. variegatus) in the Pecos River, Texas. Master's thesis, Oklahoma State Univ., Stillwater.
DARLING, J. 1976. Electrophoretic variation in Cyprinodon variegatus and systematics of some fishes of the subfamily Cyprinodontidae. Ph.D. diss. Yale Univ., New Haven, CT.
DAVIS, J. R. 1981. Diet of the Pecos River pupfish, Cyprinodon pecosensis (Cyprinodontidae). Southwest. Nat. 25:535-540.
DAVIS, S. K. 1986. Population structure and patterns of speciation in Geomys: An analysis using mitochondrial and ribosomal DNA. Ph.D. diss. Washington Univ., St. Louis, MO.
DOWLING, T. E. AND M. R. CHILDS. 1992. Impact of hybridization on a threatened trout of the southwestern United States. Conserv. Biol. 6:355-364.
DOWLING, T. E., AND W. R. HOEH. 1991. The extent of introgression outside the contact zone between Notropis cornutus and Notropis chrysocephalus (Teleostei: Cyprinidae). Evolution 45:944-956.
DOWLING, T. E., G. R. SMITH, AND W. M. BROWN. 1989. Reproductive isolation and introgression between Notropis cornutus and Notropis chrysocephalus (Family Cyprinidae): Comparison of morphology, allozymes, and mitochondrial DNA. Evolution 43:620-634.
ECHELLE, A. A., AND P. J. CONNOR. 1989. Rapid, geographically extensive genetic introgression after secondary contact between two pupfish species (Cyprinodon, Cyprinodontidae). Evolution 43:717-727.
ECHELLE, A. A., AND T. E. DOWLING. 1992. Mitochondrial DNA variation and evolution of the Death Valley pupfishes (Cyprinodon, Cyprinodontidae). Evolution 46:193-206.
ECHELLE, A. A., AND A. F. ECHELLE. 1978. The Pecos River pupfish, C. pecosensis n. sp. (Cyprinodontidae), with comments on its evolutionary origin. Copeia 1978:569-582.
ENDLER, J. A. 1977. Geographic variation, speciation and clines. Princeton Univ. Press, Princeton, NJ.
FERRIS, S. D., R. D. SAGE, C-M. HUANG, J. N. NIELSON, U. RITTE, AND A. C. WILSON. 1983. Flow of mitochondrial DNA across a species boundary. Proc. Nat. Acad. Sci. 80:2290-2294.
FORBES, S. H., AND F. W. ALLENDORF. 1991. Associations between mitochondrial and nuclear genotypes in cutthroat trout hybrid swarms. Evolution 45:1132-1349.
GYLLENSTEN, U., R. F. LEARY, F. W. ALLENDORF, AND A. C. WILSON. 1985. Introgression between two cutthroat trout subspecies with substantial karyotypic, nuclear, and mitochondrial genomic divergence. Genetics 111:905-915.
HARRISON, R. G. 1993. Hybrid zones and the evolutionary process. Oxford Univ. Press, New York.
HEDRICK, P. W. 1983. Genetics of populations. Science Books, Boston, MA.
HILLIS, D. M., A. LARSON, S. K. DAVIS, AND E. A. ZIMMER. 1990. Nucleic acids III: Sequencing. Pp. 318-372 in D. M. Hillis and C. Moritz, eds. Molecular systematics. Sinauer, Sunderland, MA.
JAMES, T. L., AND A. DE LA CRUZ. 1989. Prymnesium parvum Carter as a suspect of mass mortalities of fish and shellfish communities in western Texas. Texas J. Sci. 41:429-430.
LANSMAN, R. A., R. O. SHADE, J. F. SHAPIRA, AND J. C. AVISE. 1981. The use of restriction endonucleases to measure mitochondrial DNA sequence relatedness in natural populations. J. Mol. Evol. 17:214-226.
MURPHY, R. W., J. W. SITES, JR., D. G. BUTH, AND C. H. HAUFLER. 1990. Proteins I: Isozyme electrophoresis. Pp. 45-126 in D. M. Hillis and C. Moritz, eds. Molecular systematics. Sinauer, Sunderland, MA.
NAIMAN, R. J. 1976. Productivity of a herbivorous pupfish population (Cyprinodon nevadensis) in a warm desert stream. J. Fish Biol. 9:125-137.
POWELL, J. R. 1983. Interspecific cytoplasmic gene flow in the absence of nuclear gene flow: Evidence from Drosophila. Proc. Nat. Acad. Sci. 80:492-495.
RAND, D. M., AND R. G. HARRISON. 1989. Ecological genetics of a mouse hybrid zone: Mitochondrial, nuclear, and reproductive differentiation of crickets by soil type. Evolution 43:432-449.
RHODES, K., AND C. HUBBS. 1992. Recovery of Pecos River fishes from a red tide fish kill. Southwest. Nat. 37:178-187.
RICE, W. R. 1989. Analyzing tables of statistical tests. Evolution 43:223-225.
SELANDER, R. K., M. H. SMITH, S. Y. YANG, W. E. JOHNSON, AND J. B. GENTRY. 1971. Biochemical polymorphism and systematics in the genus Peromyscus. I. Variation in the old-field mouse (Peromyscus polionotus). Studies in genetics. Univ. Tx. Publ. 7103:49-90.
SHAW, C. R., AND R. PRASAD. 1970. Starch gel electrophoresis of enzymes - A compilation of recipes. Biochem. Genet. 4:297-320.
SOUTHERN, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517.
STEVENSON, M. M., AND T. M. BUCHANAN. 1973. An analysis of hybridization between the cyprinodont fishes Cyprinodon variegatus and C. elegans. Copeia 1973:682-692.
SWOFFORD, D. L., AND R. B. SELANDER. 1981. BIOSYS-1: A FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. J. Hered. 72:281-283.
TURNER, B. J. 1983. Genic variation and differentiation of remnant natural populations of the desert pupfish, Cyprinodon macularius. Evolution 37:690-700.
WEIR, B. S. 1990. Genetic data analysis. Sinauer, Sunderland, MA.
WILDE, G. R. 1994. Spatial and temporal patterns in the covariance of genetic and morphologic characters in a pupfish hybrid swarm. Ph.D. diss. Oklahoma State Univ., Stillwater.
WILDE, G. R., AND A. A. ECHELLE. 1992. Genetic status of Pecos pupfish populations after establishment of a hybrid swarm involving an introduced congener. Trans. Am. Fish. Soc. 121:277-286.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||mitochondrial deoxyribonucleic acid|
|Author:||Childs, Michael R.; Echelle, Anthony A.; Dowling, Thomas E.|
|Date:||Oct 1, 1996|
|Previous Article:||Severe inbreeding depression and rapid fitness rebound in the butterfly Bicyclus anynana (Satyridae).|
|Next Article:||Life-history variation and comparative phylogeography of darters (Pisces: percidae) from the North American central highlands.|