2010 SOUTHEASTERN NATURALIST 9(2):373–384 Multiple Nuclear Gene Analysis of the Divergence Between Populations of the Tallapoosa Darter, Etheostoma tallapoosae Leos G. Kral* Abstract - Populations of Etheostoma tallapoosae (Tallapoosa Darter) have previously been shown to be genetically divergent for mitochondrial DNA. In this study, PCR primers were developed to amplify portions of six nuclear genes, and sequences of these genes were assessed in three of the most divergent Tallapoosa Darter populations. This analysis shows that these populations are also highly divergent for nuclear gene sequences and thus any adaptive or potentially adaptive variation is likely to be partitioned among the populations. Sequences of the six nuclear genes have also been determined in the closely related species E. coosae and E. brevirostrum, and the utility of these nuclear gene sequences to the elucidation of darter phylogeny is discussed. Introduction Etheostoma tallapoosae Suttkus and Etnier (Tallapoosa Darter) is endemic to the Tallapoosa River system above the fall line (Suttkus and Etnier 1991). Analysis of mitochondrial DNA in a previous study (Brogdon et al. 2003) has shown that this species is subdivided into at least four genetically divergent reproductively isolated populations that were designated as management units (MU) according to the definition of Moritz (1994a, 1994b). These MUs do not share any mitochondrial haplotypes. The upper Tallapoosa River MU and the Little Tallapoosa MU are the most divergent, with a 1.6% mitochondrial sequence divergence. The mitochondrial sequence divergence between the Enitachopco Creek MU and the upper Tallapoosa River and Little Tallpapoosa River MUs is 1.04% and 1.10%, respectively. The Jay Bird Creek MU is very closely related to the upper Tallapoosa River MU, with a mitochondrial sequence divergence of 0.26%, and was not included in this study. The maximum mitochondrial sequence divergence of 1.6% between the most genetically diverged populations is fairly low, and no obvious morphological differences have been noted between those populations (Suttkus and Etnier 1991). Furthermore, since mitochondrial genomes sort faster than nuclear genomes, it is not certain that these populations have significantly different allelic frequencies of nuclear genes such that actual or potential adaptive diversity associated with the nuclear genome is partitioned among the various populations. To help ascertain the degree of nuclear gene divergence in these populations, PCR primers were developed to amplify portions of six singlecopy protein-coding nuclear genes. These primers are complementary to *Department of Biology, University of West Georgia, Carrollton, GA 30118; lkral@ westga.edu. 374 Southeastern Naturalist Vol. 9, No. 2 conserved exon sequences and also amplify the targeted gene segments from other darter species (both Etheostoma and Percina). As such, these primers can be used to assess the nuclear genetic heterogeneity of other darter species and may have utility in the estimation of darter phylogeny. Current estimation of darter phylogeny is mainly based on morphological and mitochondrial DNA sequence analysis (e.g., Near 2002, Porter et al. 2002), although some recent studies utilize both mitochondrial sequences and intron 1 of the nuclear S7 ribosomal protein gene (e.g., Keck and Near 2008). The problem with estimating phylogeny with mitochondrial sequences alone is that incorrect phylogenies may be inferred if hybridization has resulted in mitochondrial introgression with subsequent replacement by drift or selective sweep in one of the species (Alves et al. 2003, Fredsted et al. 2006, Ray et al. 2008). Basing phylogenies on single nuclear genes is also less accurate than basing phylogenies on multiple nuclear genes (Edwards et al. 2007, Gadagkar et al. 2005, Gontcharov et al. 2004, Liu et al. 2008). In this study, the nuclear sequence heterogeneity of six genes was determined in three Tallapoosa Darter populations, previously designated as the most divergent MUs, to ascertain if nuclear gene divergence exists in these populations. Using the six nuclear genes, relationships of these populations were also assessed to the closely related species Etheostoma coosae Fowler (Coosa Darter) and Etheostoma brevirostrum Suttkus and Etnier (Holiday Darter) to determine the congruence of nuclear gene phylogenies and previously determined mitochondrial phylogenies (Brogdon et al. 2003, Porter et al. 2002). Materials and Methods PCR primers were designed to amplify portions of six protein-coding genes for a total of about 10,900 base pairs of nuclear DNA, of which 3900 base pairs represent protein-coding exon sequences. The following gene segments were amplified: Glyceraldehyde-3-phosphate dehydrogenase, Ribosomal Protein S6, Lysophospholipase II, Phospholipase C-gamma-1, Kelch Repeat and BTB (POZ) Domain Containing Protein, and an open reading frame of about 1050 base pairs corresponding to an exon within a gene of unknown function (designated the uORF gene). The PCR primers are anchored within conserved exon sequences and amplify across distantly related darter species (data not shown). All but one of the gene segments were initially characterized from randomly selected cDNA clones of Tallapoosa Darter muscle/skin mRNA. Orthologous genes and likely intron/exon structure of those genes were identified by alignments of the cDNA sequences with genomes of Denio rerio Hamilton (Zebra Danio), Takifugu rubipres Temminck and Schlegel (Japanese Pufferfish), and Tetraodon nigroviridis de Proce (Green Puffer Fish) on the Ensembl Blast server. Exon-anchored PCR primers were then designed with Primer Premier (Biosoft) to amplify across each intron, and sizes of the Tallapoosa Darter introns of these gene segments were determined. Final 2010 L.G. Kral 375 exon-anchored PCR primer pairs were designed that amplified about 1000 base pair segments, except in those cases where a single intron exceeded that length. Intron/exon structures of gene segments and amplification products of these gene segments are diagramed in Figure 1. PCR primers that amplify the six genes in twelve fragments are listed in Table 1. Figure 1. Intron/exon structure of gene fragments utilized in this study. Boxes represent exons and connecting lines represent introns. Gene fragments greater than 1000 nucleotides are amplified in segments, and these are indicated below each gene fragment. Exons and portions of exons are labeled A, B, C, etc. in the downstream direction and relative to the gene fragment characterized. Sizes of gene fragments are specific for the Tallapoosa Darter population in the Tallapoosa River. 376 Southeastern Naturalist Vol. 9, No. 2 The uORF gene was not identified from a cDNA clone, but rather was identified from a clone of RAPD-PCR amplified genomic DNA. This sequence is homologous to an exon of a gene present in T. rubripes and T. nigroviridis, but it is possible that this uORF gene fragment is an amplification of a portion of a pseudo gene. However, since all indel mutations of this uORF gene segment in the Tallapoosa Darter and the Coosa Darter maintain an open reading frame (data not shown), it is likely that this gene is functional in the darters. Some of the genomic DNA samples previously isolated for analysis of Tallapoosa Darter population structure (Brogdon et al. 2003) were utilized in this study. Specifically, DNA was utilized from the three most genetically divergent sets of populations previously identified as MUs by mitochondrial DNA sequence analysis: MU 1—the population in the upper Tallapoosa River system (8 individuals), MU 2—the population in the Little Tallapoosa River system (10 individuals) and the population in the portion of the Tallapoosa River system just south of Harris reservoir (8 individuals), and MU 3—the population in Enitachopco Creek (5 individuals). To minimize costs of identifying nucleotide differences fixed in each MU sample, PCR amplifications were performed on combined genomic DNA from multiple individuals. Equal amounts of genomic DNA from individuals Table 1. Primers that amplify gene segments diagramed in Figure 1. Letters that follow gene designation in the primer name correspond to exons as labeled in Figure 1. Gene Primer name Primer sequence 5’ → 3’ uORF uorf-sense CTCTTCTATTTTGCAGCACC uorf-antisense TCCAGGTCCATGATTAGCT Kelch kelch-sense GCAAGAAACCAGGCTAAACA kelch-antisesense GCATGAAATGCCGACTGT GAPDH gapdh-sense GAGTCCACAGGTGTATTCACA gapdh-antisense GTCAGGTCAACCACGGACA LPL lplAC-sense CCCGATTCCCGTCACTCT lplAC-antisense GAGAAGCCACCGAGCATTAT lplCE-sense ACCGTATAATGCTCGGTGG lplCE-antisense TGAGGGTTGACTATGGATTTGAG lplDF-sense AGTTGGCTGGCATTGTGGC lplDF-antisense TTTGGGGCAAATACTTCTC PLC plcAC-sense CAGCAGCCAAGAAGAACTCA plcAC-antisense GGGCAGAGGGTCGTAGTT plcCE-sense GTCATCCTTCCCTGAGACCA plcCE-antisense TATGCCGCGTCCGTGCTT plcEF-sense ACGGACGCGGCATAGTTT plcEF-antisense CCACAAAGCGCAAGAAGG S6 s6AB-sense CAAGGGTCACTCCTGCTACCG s6AB-antisense CAACATACTGTCTAACGTCATCCTC s6BC-sense CGGGCTTACCGACAGCAC s6BC-antisense CAGCAGCTTGGCATACTCAGA s6CD-sense GGCAAGAAGCCCAGAACTAA s6CD-antisense CGCCTCTTGGCAATCTGTTC 2010 L.G. Kral 377 of each population were combined to form four population samples, and each sample was amplified in 50-ul reactions containing 100 ng of the combined DNA, 25 ul of Qiagen HotStarTaq Master Mix, and 0.5 uM concentration of a sense primer and 0.5 uM concentration of the matching antisense primer. The PCR conditions were identical for all twelve primer pair reactions: 95 °C for 15 minutes to activate the Taq polymerase, followed by 35 cycles of 30 seconds at 94 °C, 1 minute at 55 °C, and 2 minutes at 72 °C. The final cycle was followed by a 10-minute incubation at 72 °C. Amplified PCR products were gel purified using the Qiagen Qiaquick Gel Extraction kit or the Zymogen Zymoclean Gel DNA Recovery kit. The purified PCR products were then sequenced directly utilizing both the sense and antisense PCR primers as sequencing primers. The purified PCR products were also cloned into pSMART GC HK plasmids utilizing the Lucigen GC Cloning and Amplification Kit according to manufacturers instructions, and clones of the PCR products were sequenced utilizing the flanking SL1 and SR2 primers. Sequencing was carried out by Functional Biosciences, Inc., Madison, WI (www.functionalbio.com). Initial contig assembly from sequence alignments of direct PCR product reads as well as two to three clones of each gene fragment were carried out in GeneJockey II DNA analysis software (Biosoft). All variable site differences that were fixed in the population samples were identified by aligning gene contigs of the four population samples with the direct PCR product sequence traces in Geneious Pro bioinformatics software (Biomatters, Ltd.), and chromatogram peaks at each variable site were examined to ascertain lack of nucleotide heterogeneity within each population. Eight independent clones of each of the twelve PCR-amplified gene fragments from the combined Enitachopco Creek genomic DNA samples were also sequenced, and sequence heterogeneity observed among the cloned sequences was compared to that observed in sequence traces obtained from a population mixture of PCRamplified DNA. Sequences of these six genes were also determined from individual Coosa Darter and Holiday Darter samples by sequencing of both the direct PCR products as well as two to three clones of those PCR products as described above. Phylogenetic analyses were conducted using MEGA 4.0 software (Kumar et al. 2008, Tamura et al. 2007). Exact test of population differentiation was carried out with ARLEQUIN population genetics software version 3.1 (Excoffier et al. 2005, Raymond and Rousset 1995). Results A total of about 10,900 base pairs of single-copy genomic DNA was sequenced in each of three previously defined Tallapoosa Darter MUs, and 95 sites that are variable among these MUs were identified (Table 2). Of these 95 sites, 75 are single-nucleotide polymorphisms (SNPs), 18 are indels, and 2 are GT dinucleotide simple tandem repeats (STRs). Of the 93 sites that are 378 Southeastern Naturalist Vol. 9, No. 2 either SNPs or indels, 89 of the variants have been verified as being fixed in each of the MU samples by examination of sequence traces. Four SNP sites, identified as variable among the populations from sequences of cloned fragments, could not be verified as being fixed among the population samples because the sequences of that portion of the gene fragment in combined population samples were not of sufficient quality. Verification that heterogeneous sites would be detectable in sequence traces obtained from combined DNA samples was obtained by a comparison of the sequence data from combined Enitachopco Creek DNA samples and eight individually sequenced clones of those DNA samples. The sites that were heterogeneous within the direct PCR product sequence chromatograms of the combined population sample proved to also be variable among the individual cloned sequences. The approximately 10,900 base pairs of DNA sequenced in this study are composed of about 7000 base pairs of intron sequence and about 3900 base pairs of exon sequence. The two STR loci, 17 of the 18 indels, and all but 13 of the 75 SNPs are located in the intron regions. All but one of the 13 SNPs in the exon regions result in synonymous substitutions. The one SNP that resulted in a non-synonymous substitution results in an arginine to a serine substitution in the uORF gene. The one indel located in an exon region is a deletion of two amino acids from the uORF sequence present in the upper Tallapoosa River and Enitachopco Creek populations that maintains the same Table 2. Nuclear gene sequence variation among populations. Only variable sites are shown at their respective base numbers. Abbreviations within reference sequence: I = indel (+ indicates sequence present and – indicates sequence absent), S = STR (L, M, and S indicate relative length). Abbreviations of populations: TR = Tallapoosa River, EC = Enitachopco Creek, and LT = identical sequence in both Little Tallapoosa River and south of Harris Reservoir populations. Population uOR FK GAPDH LPL PLC 0000000 00 0000000 00000000000 000000000000000000 0000001 11 1222222 23333333444 555666666666667777 3466790 14 9112677 90223359036 148113466788991345 8223322 83 5385606 30691458103 090586857925066407 1003813 50 0715139 24717666398 921378107141218060 GAITTAT TA TGTTTTC ACCACACGIST GTIASCAGAATTCAIIGG TR ..-.... .. ....... ........+M. ..+TL..C....A.-+.A EC C.-.... A. .T..... ........+L. ..+.S........G--.. LT .C+CCTC .G C.ACCCT TTTGTCTT-SC AG-.MTG.TGCG..++A. S6 00000000000000000000000000000000000111111111111111 77888888888888888999999999999999999000000000000000 89012234556677889000122246678888899002334555666789 52670612474908029125328882286666707682780069148540 85562561603164835628720728971789683772996952157115 TCICTCTGAICATATIIICGIIGAAAACCTTGAAACIGIITITGTAIAIG TR .T-.....T-.....--+..-+...GG..AGT....+A--.-....-.-. EC ..-.AG...+.....-+-A.-+T.....T...G.G.+.++.+....+.-. LT C.+A..GA.-TTCTC+--.T+-.GG..T.....G.A-.++C+CAAG+G+T 2010 L.G. Kral 379 reading frame as the sequence present in the Little Tallapoosa River population as well as in the Coosa Darter and the Holiday Darter (data not shown). Variants of 89 SNP/indel sites appear to be fixed in each of the three MU samples. These 89 variable sites define three ancestral alleles. At least another 20 variable sites are not fixed among the population samples (data not shown) and represent variations in alleles derived from the three ancestral alleles. Thus, the three ancestral alleles and their derivatives are fixed among the Tallapoosa Darter population samples of the three separate MUs. No attempt has been made to reconstruct the derived alleles since the intra population variable sites were primarily identified in sequence traces of mixed-population PCR-amplified gene fragments and three of the complete gene sequences were reconstructed from multiple fragments. Statistical tests of population divergence can not be performed directly because population sequence data was obtained from combined DNA samples and not from individuals. A single ancestral allele type was detected in the combined sequence data from 5 to 10 individuals of a population (potentially 10 to 20 different alleles), but it is possible that a few copies of the other ancestral alleles may be present in the sample and these escaped detection. In an attempt to determine if the combined data of this small sample size could support nuclear genetic divergence among populations, an exact test of population differentiation (Raymond and Rousset 1995) was performed utilizing worst-case scenario where each population sample contained only 80% of the detected allele and 10% of each of the other two alleles. The populations previously defined as MUs show significant differences under these simulated conditions (Table 3). Note that the Little Tallapoosa River and south of Harris Reservoir populations are part of the same MU and show no significant differentiation. Maximum parsimony (MP), neighbor joining, minimum evolution, and UPGMA analyses of individual genes and concatenated sequences of all genes were performed for all three Tallapoosa Darter MUs as well as the Holiday Darter and the Coosa Darter. All analyses of concatenated sequences resulted in the same tree topologies as the MP tree shown in Figure 2. Twenty-one of the 327 variable sites in the concatenated gene sequences among the three species were parsimony-informative. Topologies of trees obtained for individual genes were essentially concordant with the concatenated dataset except for the uORF gene, where the Holiday Darter Table 3. Nuclear genetic differentiation of Tallapoosa Darter populations under simulated condition of 80% population specific allele frequency. Significance level (P-value) of nondifferentiation test. Abbreviations: NS = not significant, TR = Tallapoosa River, LT = Little Tallapoosa River, HR = south of Harris Reservoir, and EC = Enitachopco Creek. TR LT HR TR LT 0.00319 ± 0.0005 HR 0.00558 ± 0.0006 1.0 (NS) EC 0.02853 ± 0.0014 0.01790 ± 0.0013 0.02555 ± 0.0016 380 Southeastern Naturalist Vol. 9, No. 2 partitions within the Tallapoosa Darter clade (data not shown). The uncorrected pairwise sequence divergence of the concatenated genes between the Tallapoosa Darter populations ranged from 0.17% to 0.55% (Table 4), and the average distance between Tallapoosa Darter populations and the Holiday Darter averaged 0.84% (0.80% to 0.88%, Table 4). Discussion Analysis of nuclear DNA sequences showed that the Tallapoosa Darter populations previously identified as being genetically diverged for mitochondrial haplotypes (Brogdon et al. 2003) are also genetically diverged for single-copy nuclear protein-coding genes. Of the approximately 10,900 nucleotides sequenced, 89 variable sites have been confirmed as fixed among the population samples analyzed. These fixed differences show that the MU samples are each homogeneous for one major ancestral allele type at each of the six loci examined, and heterogeneity within each MU sample is comprised of variants of those ancestral allele types. The sample size is sufficient to determine that significant nuclear gene divergence is partitioned among the MUs (Table 3), but the total number of individuals examined is not great enough to conclude that these allele-frequency differences are fixed in the actual populations from which the samples were analyzed. The variation observed at the six gene fragments examined in these MU samples is, with the possible exception of the two codon deletion and nonsynonymous substitution in the uORF gene, most likely neutral variation. Table 4. Uncorrected sequence divergence of combined nuclear gene sequences. Abbreviations: TR = Tallapoosa River, EC = Enitachopco Creek, LT = Little Tallapoosa River, BR = Holiday Darter, and CS = Coosa Darter. TR EC LT BR TR EC 0.17% LT 0.55% 0.51% BR 0.85% 0.80% 0.88% CS 2.55% 2.51% 2.61% 2.78% Figure 2. Maximum parsimony phylogram of combined nuclear gene sequences. Bootstrap values (1000 replicates) are given at branches. Abbreviations: TR = Tallapoosa River, EC = Enitachopco Creek, LT = Little Tallapoosa River, BR = Holiday Darter, and CS = Coosa Darter. 2010 L.G. Kral 381 The high level of allelic divergence present among these MUs suggests that any non-neutral or potentially non-neutral mutations present in the genome are likely not homogeneously distributed among the populations. Since no obvious morphological differences are evident in these genetically divergent populations, such potentially non-neutral variation would represent cryptic variation that may be adaptive (potential adaptive variation) under new environmental conditions or change in genetic background such as future hybridization between populations or accumulation of new mutations (Le Rouzic and Carlborg 2007). Brogdon et al. (2003) designated populations of the Tallapoosa Darter as MUs based on mitochondrial sequence divergence, and this study characterized these MUs as showing a high degree of nuclear gene variation. These results strengthen the need to monitor these populations as MUs since it is likely that potential adaptive variation is sequestered among these populations. Maintenance of such variation enhances the evolutionary potential of the species. Fixation of advantageous traits by drift or selection is more likely and occurs faster if advantageous alleles are already present with multiple copies in a population than if such alleles have to arise de novo by mutation (Barrett and Schluter 2008). A number of studies have demonstrated that clades defined by mitochondrial haplotype variation within a darter species are distributed among river drainages (Krabbenhoft et al. 2008, Piller et al. 2008, Powers et al. 2004, Ray et al. 2006), as well as allopatrically distributed among tributatries within major drainages (Hollingsworth and Near 2009, Krabbenhoft et al. 2008). Etheostoma blenniodes Rafinesque (Geenside Darter) clades defined by intron 1 of the S7 ribosomal protein gene also are distributed among river drainages (Piller et al. 2008), and populations of Etheostoma mariae Fowler (Pinewood Darter) showed significant genetic structuring based on AMOVA of S7 intron 1 sequence analysis (Krabbenhoft et al. 2008). Segregation of nuclear gene variation is, therefore, likely sequestered among populations of other darter species as well. It may well be worthwhile to apply the multigenic analysis presented in this study to populations of other darter species to more fully define the segregation of nuclear genetic variation among their populations. As new low-cost, high-throughput sequencing methodologies are becoming available, a more thorough quantitative study of such variation should become practical in the near future. The elucidation of darter evolution will be more complete if an analysis of a significant sampling of nuclear gene variation complements analysis of mitochondrial and microsatellite markers as well as morphological variation. Along with the studies of Krabbenhoft et al. (2008) and Hollingsworth and Near (2009), this study shows that variation within a darter species does not only occur among major drainages, but also can occur among tributaries within a drainage. This result shows that allopatric darter evolution occurs at much smaller geographic scales than previously thought. This phenomenon is common across three major eastern US drainages, the Atlantic slope 382 Southeastern Naturalist Vol. 9, No. 2 (Krabbenhoft et al. 2008), the Cumberland (Hollingsworth and Near 2009), and the Mobile (this study). The relatively low-cost method utilized in this study is effective in identifying nuclear genomic variation at these small geographic scales. It has also been demonstrated that multigenic analysis based on the six genes utilized in this study is suitable for analysis of darter phylogeny. The cladograms obtained from multigenic analysis (Fig. 2) are concordant with cladograms of the same species obtained by mitochondrial sequence analysis (Brogdon et al. 2003, Porter et al. 2002). At the population level, analysis of the Tallapoosa Darter mtDNA haplotype data from Brogdon et al. (2003) shows that the Tallapoosa River MU and the Little Tallapoosa River MU are evolutionarily about equidistant from the Enitachopco Creek MU. However, multigenic analysis shows that the Tallapoosa River MU and the Enitachopco Creek MU are more closely related to each other than either is to the Little Tallapoosa MU (Fig. 2, Table 4). This discrepancy is likely due to faster lineage sorting of the haploid unisexually transmitted mitochondrial genome, which would be more susceptible to genetic drift than the nuclear genome. Furthermore, the multigenic analysis is based on a larger number of presumably independently assorting genes, whereas mitochondrial haplotypes represent essentially a single locus. Thus, it is likely that multigenic analysis may provide a better understanding of the evolutionary relatedness of populations than mitochondrial sequence analysis in situations where populations are divergent at nuclear loci. The major problem with reconstruction of phylogenies with nuclear genes is that sorting of gene lineages may be independent of speciation events (Maddison 1997, Moore 1995, Nichols 2001). That is, alleles of genes that have arisen prior to any speciation events may be stochastically sorted during subsequent speciation events. Under these conditions, phylogeny reconstruction using single genes may be incongruent. An example of this is seen in the phylogeny of the Tallapoosa Darter based only on the uORF gene, where Tallapoosa Darters and the Holiday Darter are not resolved as separate lineages (data not shown). Utilization of multiple genes increases accuracy. Utilizing a dataset of 106 orthologous genes in eight species of yeast, Rokas et al. (2003) found that concatenations of 20 or more genes provided accurate trees with >95% bootstrap values at each branch. Bootstrap support of >70% was obtained with as few as three concatenated genes. Using computer simulations, Gadagkar et al. (2005) have found that concatenations of 10 genes provide >95% accuracy. The phylogenetic trees were generated in this study from concatenated gene sequences. It should be noted that under some conditions concatenated data do not produce accurate species trees (Kubatko and Degnan 2007). A number of statistical methodologies are being developed to estimate species trees from multiple nuclear gene and allele DNA sequence data (Brumfield et al. 2008, Edwards et al. 2007, Liu et al. 2008). Regardless of methodology, it will probably be necessary to develop primers for more single-copy nuclear genes than the six genes utilized in this study to increase accuracy of estimated darter species relationships. 2010 L.G. Kral 383 Acknowledgments I thank Ian Stansbury, Catherine Singleton, Levy Trusty, Gregory Ayuk, and Ivey Holland, students in the department of biology at the University of West Georgia, for laboratory assistance during various phases of this project. I also thank Thomas Near for DNA samples of E. coosae and E. brevirostrum. I further thank all the reviewers of this manuscript for many helpful comments. This study was supported by Faculty Research Grant and Faculty Research Enhancement Award funds obtained from the University of West Georgia. Literature Cited Alves, P.C., N. Ferrand, F. Suchentrunk, and D.J. Harris. 2003. Ancient introgression of Lepus timidus mtDNA into L. granatensis and L. europaeus in the Iberian Peninsula. Molecular Phylogenetics and Evolution 27:70–80. Barrett, R.D.H., and D. Schluter. 2008. Adaptation from standing genetic variation. Trends in Ecology and Evolution 23:38–44. Brogdon, S.M., C.R. Tabit, and L.G. Kral. 2003. Population structure of the Tallapoosa Darter (Etheostoma tallapoosae). Southeastern Naturalist 2:487–498. Brumfield, R.T., L. Liu, D.E. Lum, and S.V. Edwards. 2008. Comparison of species tree methods for reconstructing the phylogeny of Bearded Manakins (Aves: Pipridae: Manacus) from multilocus sequence data. Systematic Biology 57:719–731. Edwards, S.V., L. Liu, and D.K. Pearl. 2007. High-resolution species trees without concatenation. Proceedings of the National Academy of Sciences of the United States of America 104:5936–5941. Excoffier, L., G. Laval, and S. Schneider. 2005. Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1:47–50. Fredsted, T., T. Wincentz, and P. Villesen. 2006. Introgression of Mountain Hare (Lepus timidus) mitochondrial DNA into Brown Hares (Lepus europaeus) in Denmark. BMC Ecology 6:17–22. Gadagkar, S.R., M.S. Rosenberg, and S. Kumar. 2005. Inferring species phylogenies from multiple genes: Concatenated sequence tree versus consensus gene tree. Journal of Experimental Zoology (Molecular and Developmental Evolution) 304B:64–74. Gontcharov, A.A., B. Marin, and M. Melkonian. 2004. Are combined analyses better than single gene phylogenies? A case study using SSU rDNA and rbcL sequence comparisons in the Zygnematophyceae (Streptophyta). Molecular Biology and Evolution 21:612–624. Hollingsworth, P.R., and T.J. Near. 2009. Temporal patterns of diversification and microendemism in eastern highland Barcheek Darters (Percidae: Etheostomatinae). Evolution 63:228–243. Keck, B.P., and T.J. Near. 2008. Assessing phylogenetic resolution among mitochondrial, nuclear, and morphological datasets in Nothonotus darters (Teleostei: Percidae). Molecular Phylogenetics and Evolution 46:708–720. Krabbenhoft, T.J, F.C. Rohde, A.N. Leibman, and J.M. Quattro. 2008. Concordant mitochondrial and nuclear DNA partitions define evolutionarily significant units in the imperiled Pinewoods Darter, Etheostoma mariae (Pisces: Percidae). Copeia 2008:909–915. Kubatko, L.S., and J.H. Degnan. 2007. Inconsistency of phylogenetic estimates from concatenated data under coalescence. Systematic Biology 56:17–24. 384 Southeastern Naturalist Vol. 9, No. 2 Kumar, S., M. Nei, J. Dudley, and K. Tamura. 2008. MEGA: A biologist-centric software for evolutionary analysis of DNA and protein sequences. Briefings in Bioinformatics 9:299–306. Le Rouzic, A., and O. Carlborg. 2007. Evolutionary potential of hidden genetic variation. Trends in Ecology and Evolution 23:33–36. Liu, L., D.K. Pearl, R.T. Brumfield, and S.V. Edwards. 2008. Estimating species trees using multiple-allele DNA sequence data. Evolution 2080–2091. Maddison, W.P. 1997. Gene trees in species trees. Systematic Biology 46:523–536. Moore, W.S. 1995. Inferring phylogenies from mtDNA variation: Mitochondrial gene trees versus nuclear gene trees. Evolution 49:718–726. Moritz, C. 1994a. Defining “evolutionary significant units” for conservation. Trends in Ecology and Evolution 9:373–375. Moritz, C. 1994b. Applications of mitochondrial DNA analysis in conservation: A critical review. Molecular Ecology 3:401–411. Near, T.J. 2002. Phylogenetic relationships of Percina (Percidae: Etheostomatinae). Copeia 2002:1–14. Nichols, R. 2001. Gene trees and species trees are not the same. Trends in Ecology and Evolution 16:358–364. Piller, K.R., H.L. Bart, and D.L. Hurley. 2008. Phylogeography of the Greenside Darter complex, Etheostoma blennioides (Teleostomi: Percidae): A wide-ranging polytypic taxon. Molecular Phylogenetics and Evolution 46:974–985. Porter, B.A., T.M. Cavender, and P.A. Fuerst. 2002. Molecular phylogeny of the snubnose darters, subgenus Ulocentra (Genus Etheostoma, Family Percidae). Molecular Phylogenetics and Evolution 22:364–374. Powers, S.L., R.L. Mayden, and D.A. Etnier. 2004. Conservation genetics of the Ashy Darter, Etheostoma cinereum (Percidae: subgenus Allohistium), in the Cumberland and Tennessee Rivers of the southeastern United States. Copeia 2004:632–637. Ray, J.M., R.M. Wood, and A.M. Simons. 2006. Phylogeography and post-glacial colonization patterns of the Rainbow Darter, Etheostoma caeruleum (Teleostei: Percidae). Journal of Biogeography 33:1550–1558. Ray, J.M., N.J. Lang, R.M. Wood, and R.L. Mayden. 2008. History repeated: Recent and historical mitochondrial introgression between the Current Darter Etheostoma uniporum and Rainbow Darter Etheostoma caeruleum (Teleostei: Percidae). Journal of Fish Biology 72:418–438. Raymond, M., and F. Rosset. 1995. An exact test of population differentiation. Evolution 49:1280–1283. Rokas, A., B.L. Williams, N. King, and S.B. Carroll. 2003. Genome-scale approaches to resolving incongruence in molecular phylogenies. Nature 425:798–804. Suttkus, R.D., and D.A. Etnier. 1991. Etheostoma tallapoosae and E. brevirostrum, two new darters, subgenus Ulocentra, from the Alabama River drainage. Tulane Studies in Zoology and Botany 28:1–24. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24:1596–1599.