University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange

Masters Theses Graduate School

12-2006

Insights into the Etheostoma spectabile species complex: Incongruence between mitochondrial and nuclear gene sequence data

Christen M. Bossu University of Tennessee - Knoxville

Follow this and additional works at: https://trace.tennessee.edu/utk_gradthes

Part of the Ecology and Evolutionary Biology Commons

Recommended Citation Bossu, Christen M., "Insights into the Etheostoma spectabile species complex: Incongruence between mitochondrial and nuclear gene sequence data. " Master's Thesis, University of Tennessee, 2006. https://trace.tennessee.edu/utk_gradthes/1509

This Thesis is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Masters Theses by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council:

I am submitting herewith a thesis written by Christen M. Bossu entitled "Insights into the Etheostoma spectabile species complex: Incongruence between mitochondrial and nuclear gene sequence data." I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the equirr ements for the degree of Master of Science, with a major in Ecology and Evolutionary Biology.

Thomas J. Near, Major Professor

We have read this thesis and recommend its acceptance:

James A. Fordyce, Christine R. Boake

Accepted for the Council: Carolyn R. Hodges

Vice Provost and Dean of the Graduate School

(Original signatures are on file with official studentecor r ds.) To the Graduate Council:

I am submitting herewith a thesis written by Christen M. Bossu entitled “Insights into the Etheostoma spectabile species complex: Incongruence between mitochondrial and nuclear gene sequence data.” I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Master of Science, with a major in Ecology and Evolutionary Biology.

______Thomas J. Near______Thomas Near, Major Professor

We have read this thesis and recommend its acceptance:

___ James A. Fordyce______

______Christine R. Boake______

Accepted for the Council:

__ ___Anne Mayhew ______Vice Chancellor and Dean Of Graduate Studies

(Original signatures are on file with official student records.) Insights into the Etheostoma spectabile species complex: Incongruence between the mitochondrial and nuclear gene sequence data.

A Thesis Presented for the Master of Science Degree The University of Tennessee, Knoxville

Christen M. Bossu December 2006 Dedication

I would like to dedicate my thesis to my parents, Frank P. Bossu and Nancy

Uridil. I experienced the passion they would exude in every endeavor, and that dedication and passion I inherited made this research possible.

ii Acknowledgements

I would like to express my sincere appreciation to my faculty advisor, Dr. Thomas

J. Near, for his guidance and support throughout the course of this thesis effort, making it clear that distance was never an issue. The insight and experience was certainly appreciated. I would, also, like to thank my other committee members Dr. Christine

Boake and Dr. James Fordyce for both the guidance and conversations that helped focus and shape this thesis.

I am also indebted to my labmates whose presence in the field and the molecular lab made my thesis a joy to do. So to Ben Keck, Phillip Hollingsworth, Jacob Kendrick and Richard Harrington, I thank you for the conversations around the fire, your enthusiasm for the field and the fish and the constant laughter.

I would also like to express my sincere thanks to the Ecology and Evolutionary department at the University of Tennessee, Knoxville. From the summer research funding that made my field collections and lab work possible to the afternoon patio time conversations that provided insight and spirits to my research and to my life.

iii Abstract

Hybridization is recognized as an evolutionary process that can provide a significant source of genetic variation and whose genetic consequences have been investigated across a wide taxonomic range of plants and animals. Darters (Percidae:

Etheostomatinae) are a clade with documented interspecific hybridization and many species with a recent evolutionary origin, yet most molecular phylogenetic analyses of darters to date have relied primarily on mitochondrial DNA (mtDNA) sequences.

Inferring relationships within and between closely related species using a single locus gene tree is potentially confounded by introgression as well as retention of ancestral polymorphisms. This can lead to incongruence between the gene tree and the species tree, and confound interpretations of phylogeography and species relationships.

Considering these limitations, I utilized both mtDNA and six nuclear genes to reconstruct the phylogeny of the E. spectabile species complex, a hypothesized reciprocally monophyletic group with known instances of intergradation and hybridization. My objectives were twofold: 1) to determine if the molecular evidence supported the recent species delimitations based on meristics and breeding male coloration and 2) to determine the phylogenetic utility and congruence of mtDNA and nuclear DNA data to address possible hybridization in the species complex. I found concordance between distinct genetic signals, meristics and geographic distributions that supported many, but not all of the recognized species. I also found that introgression is prevalent throughout the history of the E. spectabile species complex, and confounds the monophyly of the complex, specifically with E. fragi and E. uniporum mtDNA haplotypes grouping outside of the

iv complex. Understanding the prevalence of introgression is crucial for future investigation of the evolution of these fishes.

v Table of Contents Section Page

1. Introduction...... 1

2. Materials and Methods ...... 8

2.1 Specimen sampling and DNA sequencing...... 8

2.2 Phylogenetic analysi...... 21

2.3 Tests of incongruence ...... 23

3. Results...... 25

3.1 Mitochondrial and nuclear DNA sequencing

of the Etheostoma spectabile complex ...... 25

3.2 Nuclear phylogenetic analyses...... 32

3.3 Incongruence among the mitochondrial and nuclear genes of

the Etheostoma spectabile complex ...... 41

3.4 Total combined data analysi ...... 43

3.5 Repeatability of clade ...... 46

4. Discussion ...... 51

4.1 Phylogeography of the Etheostoma spectabile complex...... 51

4.1.1 Species relationships...... 51

4.1.2 Status of species...... 54

4.2 Hybridization and introgression...... 55

Literature Cited...... 60

Vita ...... 72

vi List of Tables

Table Page

1. Specimen examined, tissue voucher number with clone names,

and collection localities ...... 10-17

2. Nuclear Primers used for amplification and sequencing ...... 19

3. Models of evolution selected by AIC in Modeltest 3.06 for each

character partition of each gene ...... 23

4. Summary of the overall resolution power of individual nuclear

genes and the presence and absence of particular nodes for each individual

nuclear analysis ...... 39

5. Shimodaira-Hasegawa (SH) test of incongruence between the mtDNA

and nuclear DNA topologies of the E. spectabile species complex...... 43

6. Shimodaira-Hasegawa (SH) test of incongruence between the Bayesian

individual nuclear gene and mtDNA tree topologies ...... 44

7. Shimodaira-Hasegawa (SH) test of incongruence between the Bayesian

tree topologies of the total combined data and the nuclear DNA and mtDNA

data partitions...... 47

8. Repeatability of clades (A-O) in the total combined data set looking at

independent individual data sets ...... 48

vii List of Figures

Figure Page

1. Etheostoma spectabile species complex range map...... 9

2. A portion of the 95% majority rule tree resulting from a parsimony

ratchet analysis of the combined mtDNA dataset, Section 1...... 26

3. A portion of the 95% majority rule tree resulting from a parsimony

ratchet analysis of the combined mtDNA dataset, Section 2...... 27

4. A portion of the 95% majority rule tree resulting from a parsimony

ratchet analysis of the combined mtDNA dataset, Section 3...... 28

5. A 95% majority rule phylogeny resulting from the Bayesian analysis

of the S7 intron 1 nuclear gene of the E. spectabile species complex ...... 29

6. Tree resulting from Bayesian analysis of the S7 ribosomal subunit

intron 1...... 33

7. Tree resulting from Bayesian analysis of the Kelch repeat nuclear

gene...... 34

8. Tree resulting from Bayesian analysis of the Mixed lineage leukemia

nuclear gene ...... 35

9. Tree resulting from the Bayesian analysis of the Myoglobin nuclear

gene...... 36

10. Tree resulting from Bayesian analysis of the TMO-4C4 nuclear gene .....37

11. Tree resulting from the Bayesian analysis of the Rag1 Exon 3 nuclear

gene...... 38

viii 12. Tree resulting from the Bayesian analysis of the combined nuclear genes...... 42

13. Tree resulting from the total combined data of the nuclear DNA and mtDNA data partitions (6550 bp) Bayesian analysis ...... 45

14. Phylogeny resulting from the Bayesian analysis of the total combined data with the nodes used in the repeatability analysis labeled...... 49

ix 1. Introduction

Until recently, it was thought that one of the main consequences of hybridization was the reinforcement of reproductive isolation (Arnold 1997). Hybridization was regarded as a detrimental process that could produce unfit or inviable hybrid offspring.

Therefore prezygotic mating barriers evolved to prevent this exchange as selection against hybridization (Howard 1993) making hybridization rare in species. This is no longer the case. Hybridization is now known to be widespread taxonomically and the genetic consequences have been investigated in both plants and animals (Arnold 1992;

Grant and Grant 1992; Dowling and DeMarais 1993; Dowling and Secor 1997, Rieseberg

1997). Hybridization is also recognized as an evolutionary process that can provide a significant source of genetic variation (Arnold 1992; Harrison 1993).

Among vertebrates, introgression or gene movement between species mediated by hybridization and backcrossing is most common in fishes (Hubbs 1955; Verspoor and

Hammer 1991, Avise 2004). Examples of extensive hybridization occur in many fish species and the genetic consequences vary greatly. Hybridization can introduce new genes and novel phenotypes (Lewontin and Birch 1966; Seehausen 2004), it can create new species via immediate polyploid speciation events (Schultz 1969; Alves et al. 2001;

Gromicho et al. 2006) or give rise to new species without an increase in ploidy, called homoploid speciation or hybrid speciation (Smith et al. 1979; DeMarais et al. 1992;

Salzburger et al. 2002; Smith et al. 2003; Meyer et al. 2006). Hybridization can also lead to lineage extinction (Rhymer and Simberloff 1996; Dowling and Childs 1992;

Childs et al. 1996; Rosenfield et al. 2004; Taylor et al. 2006). Past studies of homoploid hybridization on the origin of a new species includes hybridization involving the cyprinid 1 species Gila robusta and G. elegans as the proposed origin of G. seminuda in the Virgin

River, of Nevada, Arizona, and Utah (Smith et al. 1979; DeMarais et al. 1992). Another instance of a species originating via hybridization is the swordtail Xiphophorus clemenciae that involved hybridization of a swordless southern platyfish species female and a male from a sworded southern species (Meyer et al. 2006).

Many of the extinctions through introgression that occur in fishes are caused by introductions for sport or commercial fishing, biological control or through accidental introduction of bait species (Rhymer and Simberloff 1996). For instance, extensive introductions of the rainbow trout, Oncorhynchus mykiss, and the cutthroat trout, O. clarki, in western watersheds has led to the loss of diversity of native species such as the threatened Apache trout, O. apache, and the endangered Gila trout, O. gilae (Dowling and Childs 1992). Another instance of extinction via hybridization is illustrated under the guise of hybrid vigor, or heterosis, in which the hybrid is more fit than the parental species. The hybrids of the Pecos pupfish (Cyprinidon pecosensis) and the closely related sheepshead minnow (C. variegates) show hybrid vigor in swimming endurance.

The ecological superiority of the hybrids permitted its rapid spread throughout the range of C. pecosensis and the eventual genetic loss of the parental lineage (Childs et al. 1996;

Rosenfield et al. 2004).

Studies of genetic variation within species-rich lineages can provide invaluable insight into the patterns and processes involved in speciation and hybridization

(Barraclough et al. 1999; Bhagabati et al. 2004). Darters (Percidae: Etheostomatinae) are an ideal animal system to research the resulting phylogenetic signal of hybridization as well as the genetic consequences of hybridization. Darters constitute a species-rich clade 2 comprising approximately 225 species endemic to Eastern North America with known instances of hybridization. Darters are also an extensively researched clade with systematic studies conducted at different levels of evolutionary divergence. For instance, a large-scale phylogeny of percids (Sloss et al. 2004) investigated the relationship of darters within the more inclusive Percidae. On a more narrow evolutionary scale, species-level phylogenies (Ammocrypta, Near et al. 2000; Barcheek darters, Page et al.

2003; Percina, Near 2002; Ulocentra, Porter et al. 2002; Nothonotus, Near and Keck

2005) as well as intraspecific analyses (Percina evides, Near et al. 2001; Etheostoma cinereum, Powers et al. 2004; Crystallaria asprella, Morrison et al. 2006; Etheostoma caeruleum, Ray et al. In press) have also been investigated.

A phylogenetic analysis of a darter species complex with known hybridization, the Etheostoma spectabile species complex, has not yet been investigated and would focus the question on the extent and role of hybridization in the evolution of this species complex. Investigation of the variation within the Etheostoma spectabile (Agassiz) complex began with Distler’s (1968) initial systematic examination of populations west of the Mississippi River. In his review, five subspecies were identified and described based on meristics and geographic locality (Etheostoma s. spectabile, E. s. pulchellum, E. s. squamosum, E. s. fragi, E. s. uniporum), as well as multiple zones of intergradation between the designated subspecies. An electrophoretic analysis of the lactate dehydrogenase locus followed shortly after (Wiseman et al. 1978), suggesting that populations of E. s. pulchellum and E. s. squamosum were fixed for a shared allele, while notably E. s. uniporum populations were fixed for a unique allele and populations of E. s. fragi were fixed for an allele with a unique staining pattern. Investigations of the 3 subspecies within the E. spectabile complex have relied heavily on meristics and male breeding coloration. In a study of the E. spectabile complex, Ceas and Page (1997) described four previously undescribed species (E. bison, E. burri, E. kantuckeense and E. tecumsehi) and elevated the former subspecies E. s. fragi and E. s. uniporum to species.

Ceas and Burr (2002) followed with the description of E. lawrencei from the

Cumberland, Green and Salt River systems, bringing the number of species in the E. spectabile complex to eight, with plans for the description of an additional two species

(Sheltowee and Ihiyo darters). All three systematic studies (Distler, 1968; Ceas and Page

1997; Ceas and Burr, 2002) assume that the E. spectabile species complex is monophyletic.

In contrast, the subgenus Oligocephalus, of which E. spectabile is a member, is thought to be polyphyletic (Page 1981). Oligocephalus, diagnosed by Bailey and Gosline

(1955), is described as an artificial assemblage of species that includes the E. spectabile complex, E. caeruleum, E. radiosum, E. whipplei, E. luteovinctum and ten other species

(Page 1981, 1983). These fish are characterized by a blue marginal band in the first dorsal fin above a well defined red band. Phenetic and phylogenetic analysis using morphological characters showed Oligocephalus to be polyphyletic (Page 1981).

Members of Oligocephalus commonly co-occur in the same streams, and hybrid zones have been identified between E. s. pulchellum and E. radiosum (Linder 1955, 1958;

Distler 1968; Branson and Campbell 1969; Echelle et al. 1974; Echelle et al. 1975).

Therefore the E. spectabile species complex presents an opportunity to investigate phylogenetic relationships as well as the possible hybridization between other species classified in Oligocephalus. 4 Analysis of mitochondrial DNA (mtDNA) is widely used to study phylogenetic relationships and population genetics (Nielsen et al. 1994; Brown et al. 2002). In fact, a majority of the darter investigations previously mentioned have focused on the patterns revealed by the mtDNA cytochrome b (cytb) gene. There are advantages to using mitochondrial genes when investigating species with recent diversification. The mitochondrial genome is maternally inherited with a lack of recombination, a rapid rate of mutation and reduced effective population size allowing for a better chance of tracking the species tree (Moore 1995). However, due to the rapid mutation rates, information is lost over evolutionary time (Smith 1992), so the utility of mtDNA to investigate deeper nodes is questionable. For instance, a phylogeny of the Percidae that included the darters described the inter-relationships of the clade with an uninformative polytomy (Sloss et al.

2004). Also, within the Percina phylogeny (Near 2002) there was strong bootstrap support for the most recent species divergences, but most internal nodes lacked strong bootstrap support.

The use of a single mtDNA gene to infer a species tree has limitations beyond the lack of resolution at deeper nodes. Incomplete lineage sorting of ancestral polymorphisms could result in incongruence between the gene tree and species tree. Speciation results in initial polyphyly of gene trees and eventually, given enough time and a reproductive barrier, the alleles will sort and will result in reciprocally monophyletic lineages and a more accurate species tree (Avise 1989, Moore 1995, Funk and Omland 2003). Several mechanisms may slow down coalescence time to where the pattern of reciprocal monophyly is will not be seen. First, the demography of the species or a large effective population size reduces drift. Second, if the isolation is a recent event the genes could 5 reflect the initial polyphyly due to incomplete lineage sorting. Third, balancing natural selection acting on a species, fixing ancient polymorphisms will result in incongruent gene and species trees (Maddison 1997; Funk and Omland 2003).

An additional limitation of mtDNA analysis is that incongruence between the gene tree and a species tree could be a product of extensive hybridization and introgression. Introgression produces polyphyly by introducing phylogenetically divergent mitochondrial lineages across species boundaries (Shaw 2002). Therefore the analysis of mtDNA would reflect the heterospecific origin of its mitochondrial genome

(Smith 1992), thereby incorrectly inferring a species tree from a gene tree.

Detecting hybridization and introgression can be problematic and to this effect, the ability to use different molecular markers to becomes valuable. In the past, intermediate morphology was the indicator used to assess hybridization (Branson and

Campbell 1969), but morphology is far too unconstrained genetically and plastic (Smith

1992). One method to detect hybridization involves the evaluation of a mitochondrial gene tree against a nuclear background that identifies species participating in the hybridization (Funk and Omland 2003). If used in concert with nuclear genes, mtDNA is valuable when investigating hybridization because its long persistence allows the determination of historical mitochondrial lineages (Gyllensten et al. 1985). Using multiple molecular markers, both mitochondrial DNA and nuclear DNA to detect hybridization provides the best hope of discriminating hybrid individuals and overcoming the ambiguities of morphology. In fact non-concordant evolutionary histories of nuclear and mitochondrial genes have been documented within various animal groups (Allendorf et al. 2001; Shaw 2002). 6 The goal of this research is to use phylogenies inferred from mitochondrial and multiple nuclear markers to investigate the evolutionary history of the E. spectabile species complex and to determine whether species designations based on meristics and male color characteristics are supported by the monophyly of the two mitochondrial genes. In addition, I am investigating the pattern of hybridization in the E. spectabile complex. Unique to this investigation, the utilization of six nuclear genes of varying length and regions (exon vs. intron) are the nuclear background on which I’ll compare the mtDNA phylogeny to identify potential hybrids within Oligocephalus. If there is incongruence between the mtDNA and nuclear DNA, introgression may have played a substantial role in the diversification of the E. spectabile species complex.

7 2. Materials and Methods

2.1 Specimen sampling and DNA sequencing.

I sampled all species and subspecies of the Etheostoma spectabile species complex. I also included representatives of all major lineages within the darter clade. I specifically chose three outgroup clades in order to explore the relationship within the E. spectabile complex with respect to its placement within the entire darter clade. First I chose a closely related outgroup consisting of E. microperca, E. proelaire and E. fonticola. I then included other members of Oligocephalus that sometimes co-occur with species of the E. spectabile complex (E. caeruleum, E. radiosum, E. whilpplei and E. luteovinctum) to explore potential hybridization. I sequenced many individuals of E. caeruleum that were found in sympatry with members of the E. spectabile species complex to determine if putative hybrids within the E. spectabile species complex shared haplotypes with the E. caeruleum individuals of that geographic locality. Finally, I bracketed the whole darter clade with non-darter percids including Sander vitreum and

Perca flavescens. I collected members of E. spectabile throughout their range (Figure 1) between 2004 and 2006 using standard seining techniques; the collection localities are identified in Table 1. Additional specimens of E. s. spectabile and E. s. pulchellum were supplied by Dr. Pat Ceas (St. Olaf College) and added to the analyses.

I clipped tissue from the right pectoral fin and fixed it in absolute ethanol for later

DNA isolation and sequencing and identified it with a University of Tennessee Tissue

Collection (UTTC) voucher number. Each individual fish was then tagged with the corresponding UTTC number and placed in 10% formalin solution for later deposition into the University of Tennessee Research Collection of Fishes. 8 E. lawrencei Sheltowee darter E. kantuckeense E. burri E. tecumsehi E. s. squamosum E. uniporum Mamequit darter

Kansas Kentucky Ok lahoma Missouri Tennessee

Arkansas Ihiyo darter Ozark darter E. fragi E. bison

Figure 1. Etheostoma spectabile species complex range map. E. s. pulchellum has an extensive range in the Great Plains and western Central lowlands of the United States. E. s. spectabile also has an extensive range, occurring in drainages east of the Mississippi River and is found in the river systems adjacent to the Ozark species, E. burri, E. uniporum, E. fragi and the Ozark darter. E. s. squamosum occurs on the Springfield Plateau in the lower Neosho-Spring River System of Oklahoma, Kansas, Arkansas and Missouri. The Mamequit, Sheltowee and Ihiyo darters are recognized as undescribed species.

9 Table 1. Specimen examined, tissue voucher number with clone names, and collection localities. Species Voucher number (clone name) Collection locality data Etheostoma bison UTTC 2127 (EbisB) UTTC 5376, 5377 (EbisD, EbisE) Beaverdam Creek @ W. Beaverdam Rd., Hickman Co., TN UTTC 5380, 5381 5382 (EbisH, Cane Branch of Cane Creek @ Lower Cane Creek Rd, EbisI EbisJ) Hickman Co., TN UTTC 5386, 5387, 5388 (EbisM, Lick Creek @ TN 438, Perry Co., TN EbisN, EbisO) UTTC 5395, 5396 (EbisS, EbisT) Trib to Hurricane Creek @ Spann Rd. & main stem, Humphreys Co., TN Etheostoma burri UTTC 2252 (EburB) Mill Spring, NE edge of the town Mill Spring, Wayne Co. Missouri (Type locality) UTTC 8034, 8035, 8035 (EburC, East Prong Indian Creek @ C.R. 534, Butler Co. MO EburD, EburE) UTTC 8044, 8045 8048 (EburF, Trib to Black River @ C.R. 523, Butler Co., MO EburG, EburH) UTTC 8055, 8059 (EburI, EburJ) Logan Creek @ C.R. 422, Reynolds Co., MO Etheostoma cf. spectabile UTTC 2211 (EmamB) Unnamed trib. to Half Pone Creek @ Thomasville Rd., (Mamequit darter) Cheatham Co., TN UTTC 5400, 5401, 5402 (EmamC, Wells Creek @ Hwy 13 & Campground Rd., Houston Co., EmamD*, EmamE) TN UTTC 6253, 6254 (EmamJ, South Fork of Little River @ Little River Church Rd., EmamK*) Christian Co., KY Etheostoma fragi UTTC 2253, 2254, 2255 (EfraA, Mill Cr @ Hwy 115, Sharp Co., AR EfraB, EfraF) UTTC 2765, 2766 (EfraD, EfraE) Mill Creek @ AR 117, Sharp Co., AR UTTC 8194, 8195 (EfraG, EfraH) Bray Branch @ C.R. 27, Fulton Co., AR UTTC 8196, 8197, 8198, 8205 Hackney Creek @ Unnamed Rd., ~2.5 km E of AR St. HWY (EfraI, EfraJ, EfraK, EfraL) 289 & US 62/412 Intersection, Fulton Co., AR

10 Table 1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma cf. spectabile UTTC 2216 (EhatA) Smith Fork @ Statesville TN Hwy 267, Wilson Co. TN (Ihiyo darter) UTTC 4849, 4852, 4853 (EhatC, Dixon Creek k@ jct Dixon Creek Rd & Golden Hollow EhatF, EhatG) Rd., Trousdale Co., TN UTTC 8310, 8312, 8313 (EhatT, Walker Creek @ Walker Creek Rd., Dekalb Co., TN EhatU, EhatV) Etheostoma kantuckeense UTTC 2228 (EkntA) Glover Creek @ Wilborn Rd., Barren Co. KY UTTC 6353, 6354, 6355 (EkntZ, Trammel Creek @ Concord Church & St.Rt. 2106, Allen EkntAA, EkntAB*) Co., KY UTTC 6357, 6358 (EkntAD, EkntAE) Long Creek @ Dotty Branch & St. Rt. 1578, Allen Co., KY UTTC 4870, 4873 (EkntD, EkntG) Long Creek @ Dotson Rd., Macon Co., TN UTTC 5367, 5368, 5369 (EkntI, West Fork Drake Creek @ N. Hunter Rd., Sumner Co., EkntJ, EkntK) TN UTTC 6347, 6348, 6349 (EkntT, Unnamed trib to Middle Fork of Drake's Creek @ New EkntU, EkntV) Roe, Allen Co., KY Etheostoma lawrencei UTTC 2220 (ElawA) West Blackburn Fork @ Plunk Whiston Rd., Putnam Co. TN UTTC 2226 (ElawB) Koger Creek along Hwy 415 @ confluence of McIver Creek, Clinton Co. Kentucky UTTC 2241 (ElawC) Trib of Frey Creek @ Frey Creek Church, Casey Co. Kentucky UTTC 6378, 6384, 6385 (ElawZ, Trib to East Fork Little Barren River @ St. Rt. 80, ElawAA*, ElawAB) Metcalfe Co., KY UTTC 6460, 6461, 6462 (ElawAJ, Goose Creek @ Pattieville Rd., Russel Co., KY ElawAK, ElawAL)

11 Table 1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma lawrencei UTTC 5244, 5246 (ElawAS, ElawAU) Jims Creek Spring @ Will Patton Rd., Fentress Co., TN UTTC 5231 (ElawAW) Pogue Creek off William Creek Rd., Fentress Co., TN UTTC 8266, 8278 (ElawAZ, ElawBC) Mill Creek @ Oil Hollow Rd, Clay Co., TN UTTC 4837, 4839 (ElawD, ElawF) Unnamed trib of Cumberland River @ Hwy80, Smith Co., TN UTTC 5209, 5210 (ElawK, ElawL) Beechlog Creek @ Beech Log Rd., Wilson Co., TN UTTC 5222, 5225 (ElawQ, ElawT) Shaw Branch of Martin Creek @ jct. 96 & Martin Creek Rd., Putnam Co., TN UTTC 5226, 5227, 5228 (ElawU, ElawV, Maxwell Brach of Mine Lick Creek @ I-40 exit, ElawW) Putnam Co., TN Etheostoma spectabile UTTC 3793, 3792, 3794 Spring Creek @ Hwy 51, Cooke Co., TX pulchellum (EpulA,EpulBE*,EpulBG) UTTC 4153, 4154, 4155, (EpulN*, Spring Creek, 4 mi N Danville, Yell Co., AR EpulO*, EpulB) UTTC 3788, 3789, 3790 (EpulAB, Mill Branch of Bull Creek off Hwy 31, White Co., AR EpulZ, EpulAC) UTTC 4182, 4184 (EpulAH*, EpulAJ) Deer Creek, 1.5 mi W Stull, Douglas Co., KS UTTC 4160, 4162 (EpulX, EpulAM) Jack Creek @ Hwy 177, Chase Co., KS UTTC 4143, 4150 (EpulAN, EpulAU) Big Skin Bayou @ E 1070 Rd., Sequoyah Co., OK UTTC 4164, 4166 (EpulAX, EpulC) Little Chetopa Creek, 5.5mi E Buffville, Wilson Co., KS UTTC 4178, 4179, 4181 (EpulAZ, Pennington Creek @ Mill Creek Rd., Johnston Co., EpulBB*, EpulBD*) OK UTTC 4186 (EpulBH) Wells Creek @ Sanders Rd., Le Flore Co., OK UTTC 5002 (EpulBQ) Guadalupe River @ Hope Crossing, Kerr Co., TX

12 Table 1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma spectabile UTTC 5123, 5126, 5128 (EpulCL, Sansaba River, 9 mi SE Menaro, Menaro Co., TX pulchellum EpulCO*, EpulCQ) UTTC 3800, 3801, 3802 (EpulD, Crooked River @ Hwy FF, Ray Co., MO EpulE, EpulF) UTTC 3867, 3868 (EpulG, EpulH) Trib of Sallisaw Creek, 2 mi SW Marble City, Sequoyah Co., OK UTTC 3869-3872 (EpulI, EpulJ, North Fork Verdigris River, 8 mi E Matfield Green, Lyon EpulK, EpulL*) Co., KS UTTC 3795, 3796, 3797 (EpulP, Trib of E. Branch Mill Creek, Lake Wabawnsee spillway EpulQ*, EpulR) @ Hwy 99, Waubaunsee Co., KS UTTC 4156, 4157 (EpulU, EpulV) Cadron Creek @ Hwy 124, Van Buren Co., AR UTTC 2133, 2266 (EpulCX, Spring River @ Business 60, Lawrence Co. MO EpulCY) Etheostoma cf. spectabile UTTC 2221 (EsheF) Unnamed trib to Boone Creek @Boone Creek Rd & (Sheltowee darter) Pleasant Valley Church, Garrard Co., KY UTTC 6540, 6542, 6543 (EsheA, Boone Creek @ Boone Creek Rd, Garrard Co., KY EsheB, EsheC) UTTC 7000, 7001 (EsheG, EsheH) Dix River @ Rigsby Rd., Lincoln Co., KY UTTC 7012, 7013 (EsheJ, EsheK) Hanging Fork Creek @ HWY 300, Lincoln Co., KY Etheostoma spectabile UTTC 2268 (EspeA) Marmaton River @ unnamed county road off old KS 54, spectabile Bourbon Co. KS UTTC 4169, 4170, 4171 (EspeAC. Banister Branch @ Hwy B, St. Francois Co., MO EspeAD, EspeAE) UTTC 4137, 4139 (EspeC, Pawnee Creek @ HWY 39, Bourbon Co., KS EspeAH) UTTC 3787 (EspeAJ) Petite Saline Creek @ Hwy E, Cooper Co., MO

13 Table 1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma spectabile UTTC 4175, 4176, 4177 (EspeAL*, EspeF, Middle Fork Saline River @ C. R. 100N, spectabile EspeAM*) Hamilton Co., IL UTTC 8133, 8134, 8137 (EspeAN, EspeAO, Maries River at C.R. 521, Osage Co., MO EspeAP) UTTC 3865, 3866, 3874 (EspeG, EspeH, Chicken Creek @ Hwy 370, Lyon Co., KS EspeE) UTTC 3778, 3779, 3780 (EspeL, EspeM, Green Branch @ Hwy BB, Saline Co., MO EspeN) UTTC 4172, 4173, 4174 (EspeP, EspeQ, Indian Creek, IL EspeR) UTTC 3806, 3807, 3808 (EspeU, EspeV*, West Fork Middle Fork of Big Creek, .2mi W EspeW) Buckhorn, Madison Co., MO UTTC 3875, 3876 (EspeX, EspeY) Doe Run, 2 mi W town of Doe Run, St. Francois Co., MO UTTC 4190, 4192 (EspeAQ, EspeAR) Cedar Creek, 2 mi N Simpson, Johnson Co., IL Etheostoma cf. spectabile UTTC 2289 (EozdA Thomas Creek @ the Thomas Creek B&B, (Ozark darter) Newton Co. AR UTTC 8170, 8171, 8172 (EozdB, EozdC, N. Fork White River @North Fork Rec Area, off EozdD) ST. HWY CC, Ozark Co.,MO UTTC 8189, 8193 (EozdE, EozdF?) Big Creek @ St. Rte. 14, Searcy Co., AR Etheostoma spectabile UTTC 4141, 4142 (EsqmB, EsqmC) Big Creek @ California Rd, Allen Co., KS squamosum UTTC 3782, 3783, 3784 (EsqmF, EsqmD, Illinois R @ US 62, Cherokee Co., KS EsqmE) Etheostoma tecumsehi UTTC 2206 (EtecM) Buck Fork of Pond River, Todd Co., KY UTTC 6281, 6294, 6295, 6298 (EtecA, EtecD, Buck Fork and trib, @Flat Rock Rd Todd Co., EtecE*, EtecH*) KY

14 Table 1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma UTTC 2257, 2258 (EuniC, EuniD) Mulberry Creek @ Hwy 21, Ripley Co. MO uniporum UTTC 8071, 8073, 8074, 8075, 8076, 8077 (EuniG, Pine Valley Creek @ C.R. D, Van Buren, Carter Co., EuniH, EuniI, EuniJ, EuniK, EuniL) MO UTTC 8096 (EuniM) Tenmile Creek @ C.R.419, ~7miSE of Ellsinore, Carter Co., MO UTTC 8101, 8102, 8105 (EuniO, EuniP, EuniQ*) Trib to South Prong of Little Black River @ confluence & St. HWY K, Ripley Co., MO UTTC 8116, 8117, 8118 (EuniR, EuniS*, EuniT) Hurricane Creek @ St. Rte U, Oregon Co., MO Etheostoma UTTC 3803, 3804, 3805 (EwhiA, EwhiB, EwhiC) Cove Creek @ Hwy 336, Van Buren Co., AR whipplei Etheostoma UTTC 1961 (EcaeA) Piney Fork Fort Campbell Military Reservation on caeruleum HWY 345, Montgomery Co., TN UTTC 1450 (EcaeB) Middle Fk. Of the Vermillion R. @ 900E Rd. Bridge, Vermillion Co., IL UTTC 2768 (EcaeE) Mill Creek at AR 117, Sharp Co., AR UTTC 4867, 4868 (EcaeG, EcaeH) Long Creek @ Dotson Rd, Macon Co., TN UTTC 6319 (EcaeO) Unnamed trib to Middle Fork of Drake's Creek @ New Roe, Allen Co., KY UTTC 6388 (EcaeS) trib to East Fork Little Barren River @ St. Rt. 80, Metcalf Co., KY UTTC 8056, 8057 (EcaeT, EcaeU) Logan Creek @ C.R. 422, Reynolds Co., MO UTTC 8078, 8079 (EcaeV, EcaeX) Pine Valley Creek @ C.R. D, Carter Co., MO UTTC 8104 (EcaeY) Trib to South Prong of Little Black R @ confluence & St. HWY K, Ripley Co., MO UTTC 8120 (EcaeZ) Hurricane Creek @ St. Rte U, Oregon Co., MO

15 Table1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma caeruleum UTTC 8173 (EcaeAA) N. Fork White River @ North Fork Rec Area off ST. HWY CC, Ozark Co.,MO UTTC 8190, 8192 (EcaeAB, Big Creek @ St. Rte. 14, Searcy Co., AR EcaeAC) UTTC 8204 (EcaeAD) Hackney Creek @ Unnamed Rd., ~2.5 km E of AR St. HWY 289 & US 62/412 Intersection, Fulton Co., AR Sander vitreum UTTC 312 (SvitA) Mississippi R. Rock Island Co. IL Perca flavescens UTTC 261 (PflaA) Lake Andrusia (south east shore) 5 mi WNW Cass Lake, Beltrami Co., MN Percina aurantiaca UTTC 88 (PauaA) Emory River, 4 mi SW Wartburg, Morgan Co., TN Percina roanoka UTTC 76 (ProaA) Blackwater River, 3 mi N Rocky Mount, Rt. 220 bridge, Franklin Co., VA Crystallaria asprella UTTC 686 (CaspB) Cahaba River at Centreville, Bibb Co., AL Ammocrypta pellicuda UTTC 104 (ApelA) Embarrass River, St. Hwy 121 bridge at Greenup Cumberland Co., IL Etheostoma cinereum UTTC 689 (EcinA) Rockcastle River, Rockcastle Co., KY Nothonotus camurum UTTC 6492 (NcamZ) Buck Creek, @ Homer Ping (Goochtown Rd.), ~ 2.7 km SE of Goochtown, Pulaski Co., KY Etheostoma blennioides UTTC 756 (EbleA) trib. West Fork Pond River 4 mi NW Fruit Hill, Johnson Mill Rd., Christian Co., KY Etheostoma flabellare UTTC 1438 (EflaB) Middle Fork of the Vermillion River, 2 mi. ENE Collison at 900E Rd bridge, Vermilion Co., IL Etheostoma vitreum UTTC 976 (EvitA) Blackwater River, 3 mi N Rocky Mount, Rt. 220 bridge, Franklin Co., VA Etheostoma radiosum UTTC 863 (EradA) Gap Creek, ~0.9 mi SE Caddo Gap, St. Hwy. 8, Montgomery Co., AR

16 Table 1. Continued. Species Voucher number (clone name) Collection locality data Etheostoma whipplei UTTC 3805 (EwhiC) Cove Cr., 1 mi W Green Tree @ Hwy 336, Van Buren Co. AR Etheostoma luteovinctum UTTC 2456 (ElutD) Unnammed trib. to North Prong Barren Fork, at Haneger Rd., ca. 11.3 km ESE of McMinnville, Warren Co., TN Etheostoma punctulatum UTTC 465 (EpunA) Osage Creek, Carroll Co., AR Etheostoma fonticola UTTC 2105 (EfonA) National Fish Hatchery, San Marcos, TX Etheostoma proelaire UTTC 953 (EproA) Max Creek, Johnson Co., IL Etheostoma microperca UTTC 1191 (EmicA) East Fork Raccoon Creek, 4 mi W Beloit, Spring Creek Rd., Rock Co., WI * indicates an individual removed from analysis because its haplotype was identical to that of another individual in the same population

17 DNA was isolated with standard phenol-chloroform extraction and ethanol precipitation methods or with a Qiagen DNAeasy Tissue Extraction Kit following the manufacturer’s protocol. For all individuals, mitochondrial genes cytochrome b (cytb),

NADH subunit 2 (ND2) and the nuclear S7 ribosomal first intron were amplified using the polymerase chain reaction (PCR) protocols found in Near (2002), Kocher et al.

(1995) and Chow and Hazama (1998), respectively.

From the 20 nuclear genes screened, I chose five additional nuclear genes based on ease of amplification and sequencing, rather than variable character sites that are sequenced for a subset of the total taxa sampled (Table 2). To investigate the ancient introgression and the more recent hybridization, I focused on populations of E. spectabile species complex whose mtDNA haplotypes were grouping outside of the species complex. The subset includes approximately 27 species with multiple individuals and populations within the species E. fragi, E. uniporum, E. caeruleum, and the populations of E. burri, E. s. spectabile and E. s. pulchellum with putative hybrids for a total of 46 individuals. If there was past introgression, I’d expect individuals with heterospecific mtDNA haplotypes to have nuclear haplotypes that are more closely related to haplotypes of the members of their own species group. I included multiple individuals of E. caeruleum to determine if there was a geographic signal in nuclear introgression as well as the hypothesized mtDNA introgression or whether the nuclear genes supported the monophyly of the species complex.

18 Table 2. Nuclear primers used for amplification and sequencing. Gene Primer Sequence Length Reference S7 Intron 1 S7RPEX1F 5’-TGGCCTCTTCCTTGGCCGTC-3’ 543 bp Chow & Hazama 1998 S7RPEX2R 5’-AACTCGTCTGGCTTTTCGCC-3’ Rag1Exon3 Rag1F1A 5’-CTGAGCTGCAGTCAGTACCATAAAGATGT-3’ 1377 bp Lopez et al. 2004 R1-4090R A 5’-CTGAGTCCTTGTGAGCTTCCATRAAYTT-3’ Rag1F2 S 5’-GATCTTTCAGCCCCTGCAYNCCC-3’ Designed Rag1R2 S 5’-GCTCAAAGGCCTTTGACTGRC-3’ Mixed Lineage MLL25AF 5’-GCNCGNTCNAAYATGTTYTTYGG-3’ 807 bp Dettai and Leukemia Lecointre 2005 MLL25AR 5’-ATRTTNCCRCARTCRTCRCTRTT-3’ MLL-F2 5’-TTGGCCTCACRCCTTTCTACGGGG-3’ Designed MLL-R2 5’-CTGAGSGACAGTTGGGCCTC-3’ TMO-4C4 TMO4C4-F2 5’-GAKTGTTTGAAAATGACTCGCTA-3’ 473 bp Streelman et al. 1998 TMO4C4-R2 5’-AAACATCYAAMGATATGATCATGC-3’ Myoglobin MVO-F2b 5’-AGCTGTTCCCCAAGTTTGCCG-3’ 409 bp MVO-R2 5’-GCATCGAGTCCTGCCTTCTC-3’ KELCH Kelch-F 5’-GCAAGAAACCAGGCTAAACA-3’ 754 bp Leos Kral Kelch-R 5’-GCATGAAATGCCGACTGT-3’ Unless otherwise indicated by an A (amplification primer only) or S (sequencing primer only), the primer for amplification was also used for sequencing

19 I sequenced the more slowly evolving 3’ segment of the Rag1 gene, Rag1 Exon 3

(Rag1Ex3) that is approximately 1400 bp. I used the forward primer Rag1-F1 and the reverse primer F1-4090R (Lopez et al. 2004) for amplification. I developed new sequencing primers, Rag1-F2 and Rag1-R2 (Table 2) based on Etheostoma caeruleum

(AY308768) and Perca flavescens (AY430226) sequences on Genbank and a complete E. caeruleum sequence from my sampled taxa. The resulting fragment is approximately

1377 bp, slightly shorter than the sequences based on the original amplification primers.

Mixed lineage leukemia (MLL) contains an intron spacer surrounded by exon 25 and exon 26. I designed MLL-F2 and MLL-R2 based on the Percina roanoka, E. microperca and E. flabellare sequences that I had complete using MLL25AF and

MLL25AR primers (Dettai and Lecointre 2005). The MLL region spans approximately

800 bp, although the fragment length differs between the outgroup taxa and members of the E. spectabile species complex. Myoglobin (Mb) contains 201 bp of an exon region followed by a 208 bp intron region. The primers we designed for the Mb gene were

MVOF2-b and MVOR2 and were used in both amplification and sequencing.

Thermal cycling conditions for the cytb and ND2 included an initial denaturation step of 94ºC for 3 min followed by 30 cycles of denaturation at 94ºC (30 sec) primer annealing at 50ºC (30 sec), and extension at 72ºC (1.5 min) and a final incubation of 72ºC for 5 min was added at the end of the cycles. Thermal cycling conditions were generally the same for the nuclear genes, except the optimal primer annealing temperature for

20 Rag1Ex3 and MLL was 57.4ºC, and 55ºC for the S7, Mb, TMO-4C4, and KELCH nuclear genes.

The amplifications of all genes were purified using enzymatic purification with

Exonuclease1 and Shrimp Alkaline Phosphatase (SAP), cycled for 15 minutes at 37ºC, followed by 15 minutes at 80ºC to inactivate the enzymes, or Qiaquick PCR Purification kits before sequencing. Cleaned PCR products were used as templates for Big Dye

(Applied Biosystems) cycle sequencing and sequencing reactions were read using an ABI

3100 automated sequencer at multiple sequencing labs. Individual reaction files were edited and contigs were built with the edited sequence files in Sequencher (GeneCodes,

Ann Arbor, MI). The cytb and ND2 sequences did not require gaps and were aligned by eye in PAUP*4.01 (Swofford 2001). All other nuclear fragments that varied in length and spanned both exon and intron regions were aligned in ClustalX 1.8 (Thomson et al.

1997) followed by minor adjustment by eye in PAUP. The alignment was checked in

MacClade 4.0 (Maddison and Maddison 2000) using the alignment of the amino acids from the coding sequences as confirmation.

2.2 Phylogenetic analyses.

Maximum parsimony (MP; Fitch 1971) and Bayesian maximum likelihood

(BML; Larget and Simon 1999) analyses were used to generate phylogenetic hypotheses from the mtDNA and nuclear DNA. MP analyses, excluding missing and ambiguous data were run on 4 different partitions: cytb and ND2 sequences combined for the E. spectabile species complex, each nuclear gene separately for the taxa subset, a

21 combination of all nuclear genes and an analysis with the combination of both mitochondrial and nuclear sequences. The computer program PAUP* 4.01 (Swofford

2001) was used for all MP analyses to find the most parsimonious tree. The program utilized a full heuristic search with 10 addition sequence replicates and TBR branch swapping. The robustness of the nodes was assessed using non-parametric bootstrap analysis with 100 pseudoreplicates. Due to the extensive dataset used for the combined mtDNA analysis of the E. spectabile species complex, we used a parsimony ratchet

(Nixon 1999) method to analyze the data. The parsimony ratchet analysis randomly selects a subset of current characters, performs TBR branch swapping and keeps one or a few trees for 200 iterations. The ratchet samples many tree islands with fewer trees from each island, so it provides much more accurate estimates of the “true” consensus than collecting many trees from few islands under traditional parsimony methods (Nixon

1999).

Character partitions utilized in the Bayesian analyses were assigned based on region (intron vs. exon) and codon positions in the coding region for each gene. The S7

Intron 1 nuclear gene had only one partition corresponding to the intron region. The mitochondrial genes (cytb and ND2), the protein coding TMO-4C4 and the Rag1 Exon 3 nuclear genes were assigned three character partitions, one for each of the three codon positions. The remaining genes (Myoglobin, Kelch and MLL) had four character positions, one for the 1st, 2nd and 3rd codon positions and one partition for the intron region. The models of evolution for each character position utilized in the Bayesian analyses and reported in Table 3 were selected in Modeltest 3.06 (Posada and Crandall

22 Table 3. Models of evolution selected by AIC in Modeltest 3.06 for each character partition of each gene. Gene 1st codon 2nd codon 3rd codon intron position position position Cytb TIMef+I+G GTR+I+G GTR+I+G N/A ND2 TrN a +I+G GTR+I+G GTR+I+G N/A S7 Intron 1 N/A N/A N/A TVM+I TMO-4C4 F81 b +I F81 b SYM d +G N/A Myoglobin TrNef TVM HKY c TVM Kelch K81 f uf+I TrN a +I F81 b +I HKY c Mixed Lineage TrN a TrN a +I K80 e +G TIM Leukemia Rag1 Exon 3 TVM+I TVM+I TVM+G N/A a Tamura and Nei 1993, b Felsenstein 1981, c Hasegawa et al. 1985, d Zharkikh 1994, e Kimura 1980, f Kimura 1981 GTR refers to the General Time Reversible Model, TVM refers to the Transversion Model (variable transversions, transitions equal) and TIM refers a Transition model (variable transitions, transversions equal). G indicates there is a gamma distribution, I indicates that the proportion of invariant sites is greater than 0, ef indicates equal base frequency, and uf indicates unequal base frequency.

1998) using the Akaike Information Criterion (AIC). The Bayesian analysis was run in

MrBayes (Huelsenbeck and Ronquist 2001) for 2,000,000 generations to allow the

MCMC algorithm to run for an appropriate number of generations to allow convergence in the estimation of the parameters. Four chains were run simultaneously in each analysis. The burn in period of the MCMC analysis was 1 X 106 generations, and the parameters and trees saved before the burn in period were discarded.

2.3 Tests of incongruence

The Shimodaira-Hasegawa (SH) test was executed in PAUP*4.01 (Swofford

2001) to test the null hypothesis that all trees in the set of probable tree topologies resulting from the mitochondrial and nuclear gene sequences were equally good 23 explanations for the data. The SH test using Rell Bootstrap with 1000 replicates was run on several tree partitions: 1) the BML of the concatenated mtDNA and the S7 gene for the large E. spectabile complex dataset after removing identical haplotypes (108 individuals), 2) for each BML topology of the 6 nuclear genes for the subset of taxa, and

3) between the BML of the mtDNA tree and the concatenated nuclear tree.

The Shimodaira-Hasegawa test is a conservative test, and one difference in a node could result in the rejection of congruence, therefore we also performed a node by node reliability test. If the same clade is repeated despite the possibility that different genes resolve different parts of the phylogeny, that clade is considered to be reliable (Chen et al. 2003). Therefore relationships found in independent data sets are robust, whether or not there is strong statistical support or robust bootstrap values. I created a single resolved node constraint tree for each internal node in the total data phylogeny, and then used that constraint tree to search the posterior tree space from each single gene analysis.

The percent of trees that have that particular node was used to assess significance.

24 3. Results

3.1 Mitochondrial and nuclear DNA sequencing of the Etheostoma spectabile complex.

Mitochondrial DNA sequencing was performed on 212 individuals of the E. spectabile complex including outgroup taxa for both cytb and ND2. The concatenated mtDNA parsimony ratchet phylogeny (Figure 2, 3 and 4) provides evidence of a geographic signal as well as ancient and possible current hybridization.

The S7 intron 1 nuclear fragment was sequenced for all 212 individuals, including outgroup taxa. There were a number of identical or very closely related haplotypes in the mitochondrial analysis; therefore, I removed identical haplotypes to simplify the tree.

The resulting phylogeny (Figure 5) is from a Bayesian analysis of 108 individuals.

The parsimony ratchet analysis of 212 individuals indicates a distinct geographic split with the E. s. pulchellum and E. s. squamosum clade that are found in the Great

Plains and western Central lowlands and the Springfield Plateau respectively, and an

Eastern clade with species in the Ozark Region and east of the Mississippi. Concordant with the allozyme study that indicated that these two subspecies were fixed for the same

LDH allele (Wiseman et al. 1978), the mtDNA sequencing also indicates that members of

E. s. squamosum and E. s. pulchellum share haplotypes (Figure 3).

Also concordant with the previous LDH electrophoretic analysis (Wiseman et al.

1978), my mtDNA data show that E. uniporum and E. fragi form two distinct clades using mitochondrial sequencing, but in the phylogeny their haplotypes group outside of the E. spectabile complex (Figure 2).

25 Sander vitreum A Perca flavescens A Nothonotus camurum Z Percina aurantiaca A Percina roanoka A Crystallaria asprella B E. cinereum A Ammocrypta pellicuda A E. vitreum A E. flabellare B E. blennioides A E. fragi E E. fragi F E. fragi G E. fragi H E. fragi K E. fragi A E. fragi J E. fragi B E. fragi D E. fragi I E. fragi L E. luteovinctum D E. radiosum A E. whipplei B E. whipplei A E. whipplei C E. spectabile pulchellum Z E. spectabile pulchellum AB E. spectabile pulchellum AC E. spectabile spectabile A E. caeruleum AA E. caeruleum AE E. caeruleum AB E. caeruleum AC E. uniporum R E. uniporum O E. uniporum P E. uniporum H E. uniporum C E. uniporum D E. uniporum I E. uniporum J E. uniporum L E. uniporum G E. uniporum K E. caeruleum X E. uniporum M E. caeruleum A E. caeruleum B 10 changes E. caeruleum G E. caeruleum S E. caeruleum H E. caeruleum O 10 E. caeruleum E E. caeruleum AD cha E. caeruleum Z nge E. uniporum T E. caeruleum V s E. caeruleum Y E. burri H E. caeruleum T E. caeruleum U E. spectabile spectabile Y E. spectabile spectabile E E. spectabile spectabile X Figure 2. A portion of the 95% majority rule tree resulting from a parsimony ratchet analysis of the combined mtDNA dataset, Section 1. The red highlighted taxa are putative hybrids.

26 E. fonticola A E. proelaire A E. punctulatum A E. microperca A E. spectabile pulchellum T E. spectabile pulchellum E E. spectabile squamosum B E. spectabile pulchellum AJ E. spectabile pulchellum Q E. spectabile pulchellum D E. spectabile pulchellum F E. spectabile squamosum D E. spectabile squamosum F E. spectabile pulchellum AH E. spectabile pulchellum P E. spectabile pulchellum R E. spectabile pulchellum AM E. spectabile pulchellum X E. spectabile squamosum C E. spectabile pulchellum AN E. spectabile pulchellum AU E. spectabile pulchellum BH E. spectabile squamosum E E. spectabile pulchellum B E. spectabile pulchellum U E. spectabile pulchellum V E. spectabile pulchellum G E. spectabile pulchellum H E. spectabile pulchellum C E. spectabile pulchellum AX E. spectabile pulchellum I E. spectabile pulchellum J E. spectabile pulchellum K E. spectabile pulchellum CZ E. spectabile pulchellum CX E. spectabile pulchellum CY 10 changes E. spectabile pulchellum AZ E. spectabile pulchellum A E. spectabile pulchellum BG E. spectabile pulchellum BQ E. spectabile pulchellum CL E. spectabile pulchellum CQ

10 chan ges

Figure 3. A portion of the 95% majority rule tree resulting from a parsimony ratchet analysis of the combined mtDNA dataset, Section 2. Includes E. spectabile pulchellum and E. s. squamosum. E. s. squamosum haplotypes are highlighted in red to illustrate the admixture of haplotypes.

27 E. cf. spectabile Ozark darter A E. cf. spectabile Ozark darter B E. cf. spectabile Ozark darter C E. cf. spectabile Ozark darter D E. cf. spectabile Ozark darter E E. cf. spectabile Sheltowee darter C E. cf. spectabile Sheltowee darter G E. cf. spectabile Sheltowee darter H E. cf. spectabile Sheltowee darter A E. cf. spectabile Sheltowee darter J E. cf. spectabile Sheltowee darter B E. cf. spectabile Sheltowee darter F E. cf. spectabile Sheltowee darter K E. spectabile spectabile AD E. spectabile spectabile AE E. spectabile spectabile AC E. spectabile spectabile P E. spectabile spectabile Q E. spectabile spectabile R E. spectabile spectabile AR E. spectabile spectabile F E. spectabile spectabile AO E. spectabile spectabile AH E. spectabile spectabile AN E. spectabile spectabile H E. spectabile spectabile AP E. spectabile spectabile C E. spectabile spectabile G E. spectabile spectabile U E. spectabile spectabile W E. spectabile spectabile L E. spectabile spectabile M E. spectabile spectabile AJ E. spectabile spectabile N E. burri B E. burri C E. burri D E. burri E E. burri G E. burri J E. burri F E. burri I E. cf. spectabile Mamequit darter J E. cf. spectabile Mamequit darter A E. cf. spectabile Mamequit darter B E. cf. spectabile Mamequit darter C E. cf. spectabile Mamequit darter E E. bison B E. bison O E. bison M E. bison N E. bison I E. bison H E. bison J E. bison D E. bison E E. bison S E. bison T E. kantuckeense A E. kantuckeense G E. kantuckeense D E. kantuckeense AD E. kantuckeense AE E. tecumsehi A E. tecumsehi D E. tecumsehi M E. kantuckeense AA E. kantuckeense Z E. kantuckeense T E. kantuckeense U E. kantuckeense V E. kantuckeense I E. kantuckeense J E. kantuckeense K E. lawrencei A E. cf. spectabile Ihiyo darter A E. cf. spectabile Ihiyo darter T E. cf. spectabile Ihiyo darter U E. cf. spectabile Ihiyo darter V E. lawrencei C E. lawrencei AJ E. lawrencei AK E. lawrencei AL E. lawrencei AB 10 changes E. lawrencei Z E. lawrencei V E. cf. spectabile Ihiyo darter F E. cf. spectabile Ihiyo darter G E. lawrencei Q E. lawrencei B E. lawrencei F E. lawrencei AW E. lawrencei AS E. lawrencei AU E. cf. spectabile Ihiyo darter C E. lawrencei D E. lawrencei P E. lawrencei T E. lawrencei U E. lawrencei K E. lawrencei L E. lawrencei W E. lawrencei AZ E. lawrencei BC

Figure 4. A portion of the 95% majority rule tree resulting from a parsimony ratchet analysis of the combined mtDNA, Section 3. The haplotypes highlighted illustrate those species that are not monophyletic.

28 Sander vitreum A Perca flavescens A Percina aurantiaca A Percina roanoka A E. cinereum A E. blennioides A Crystallaria asprella B Ammocrypta pellicuda A Notnonotus camrurum Z E. microperca A E. fonticola A E. proelaire A E. punctulatum E. vitreum A AE. flabellare B E. luteovinctum D

E. caeruleum (5)

E. whipplei A E. uniporum C E. radiosum A E. caeruleum G E. caeruleum (2) E. uniporum M E. uniporum O E. fragi (4) E. uniporum R E. uniporum T E. uniporum (2)

E. spectabile spectabile (13)

E. cf. spectabile Ihiyo darter (2) E. cf. spectabile Mamequit darter

E. bison (5)

E. lawrencei (11)

E. kantuckeense (4) E. cf. spectabile Ihiyo darter (2) E. cf. spectabile Sheltowee darter (3) E. tecumsehi E. cf. spectabile Mamequit darter E. tecumsehi (2) E. cf. spectabile Ozark darter E. burri (3) E. burri H E. spectabile spectabile (2) E. spectabile pulchellum (2)

E. spectabile pulchellum (13)

E. spectabile squamosum (3) 5 changes

Figure 5. A 95% majority rule phylogeny resulting from the Bayesian analysis of the S7 intron 1 nuclear gene of the E. spectabile species complex. Includes 108 individuals.

29 Etheostoma uniporum haplotypes form a clade that is completely nested within the sampled E. caeruleum haplotypes. Etheostoma fragi haplotypes are reciprocally monophyletic, but the pattern of haplotype evolution is not similar to E. uniporum. That is, although the E. uniporum clade is completely bracketed by the E. caeruleum haplotypes, E. fragi’s haplotypes are evolutionary divergent from any other

Oligocephalus species. In fact there is no haplotype within the phylogenetic analysis that

E. fragi appears closely related to.

Figure 4 shows genetic structure between the Ozark and eastern clade members.

The Ozark darter found in the North Fork of the White River in Missouri is a reciprocally monophyletic and basal to the rest of the eastern clade.

Haplotypes from E. s. spectabile populations in the Castor River, Lamine River and Petite Saline Creek of the Missouri and Mississippi River drainages in Missouri are sister to the reciprocally monophyletic E. burri, endemic to the Black River in Missouri.

Thus my data show E. s. spectabile is paraphyletic. There is an isolated Osage River,

Missouri E. s. spectabile clade as well as a clade that includes E. s. spectabile found in

Illinois in the Ohio River drainage. There also appears to be an isolated clade in the

Meramec River, Arkansas that is basal to the E. s. spectabile of Missouri and Illinois.

The undescribed Sheltowee darter found in the southern Kentucky River is a sister taxon to the E. s. spectabile and E. burri clade.

Etheostoma kantuckeense of the Barren River system in Kentucky is not monophyletic, because E. tecumsehi, which is endemic to the Pond River system in

Kentucky, is more closely related to the E. kantuckeense found west of Barren River

30 Lake as compared to the E. kantuckeense populations found in tributaries south and east of Barren River Lake.

Etheostoma lawrencei was described as a species found in three disjunct drainages

(Ceas and Burr 2002) and geographically adjacent to the undescribed Ihiyo darter. My data illustrate an admixture of both species’ haplotypes indicating that E. lawrencei is not monophyletic which could be a result of some introgression or incomplete lineage sorting. Although there is haplotype sharing between the Ihiyo darter and E. lawrencei,

E. lawrencei haplotypes show a geographic signal. There are three distinct genetic groups, distinct haplotype is from an E. lawrencei individual in the Roaring River,

Tennessee, another clade is formed from haplotypes in the Green River, Kentucky and E. lawrencei populations in the Cumberland River, Tennessee drainage form the third distinct clade.

There is evidence for both ancient and more recent hybridization among the species in the E. spectabile complex when the mtDNA analysis is compared to the S7 nuclear analysis. Evidence of the major ancient introgression was previously mentioned in terms of the very divergent and monophyletic E. fragi and E. uniporum haplotypes.

All of the hypothesized hybridization events occurred in species west of the Mississippi drainage such as E. s. spectabile, E. spectabile pulchellum, E. burri and E. uniporum. In at least two of the sampled populations all individuals sequenced are fixed for heterospecific haplotypes, supporting the hypothesis of fixation after an ancient hybridization event. The mtDNA haplotypes in the E. s. spectabile Doe Run population in St. Francis County, Missouri are more closely related to those of E. caeruleum found in neighboring Reynolds County, Missouri. A population of E. s. pulchellum of Mill 31 Branch of Bull Creek in White County, Arkansas, is fixed for haplotypes closely related to those of E. whipplei.

There is also evidence of recent hybridization events in which only a one of several individuals that were sequenced had a heterospecific mtDNA haplotype. In each case, the occurrence of putative hybridization occurs in populations west of the

Mississippi River. For instance, E. uniporum (EuniM, EuniT), E. s. spectabile (EspeE) and E. burri (EburH) mtDNA haplotypes are more closely related to E. caeruleum mtDNA haplotypes found in the same geographic region than those of their own species.

For all these putative hybrids, fin ray counts and infraorbital canal diagnostics confirmed the identity of the individuals as members of the E. spectabile species complex. In order to avoid ambiguities in identifying putative hybrids based on morphology, I created a nuclear phylogeny using S7 ribosomal subunit 1 for all individuals (Figure 5) and then an additional subset of nuclear genes for a subset of the darters, focusing on the putative hybrids for further investigation.

3.2 Nuclear phylogenetic analyses.

The Bayesian analysis of each independent nuclear gene is similar to the maximum parsimony analysis, but since the Bayesian analysis incorporates a model of molecular evolution and illustrates greater node support, those results will be reported here. Consistent with utility of nuclear genes at deeper nodes, all nuclear genes support the monophyly of the darters, but there is limited resolution of the more recent clades, and variable resolution among the different genes for different parts of the phylogeny

(Figure 6-11; summarized in Table 4).

32 Sander vitreum A Perca flavescens A * Percina aurantiaca A Percina roanoka A Ammocrypta pellicudaA * Crystallaria asprella B * E. cinereum A E. blennioides A Nothonotus camurum Z * E. microperca A E. fonticola A E. proelaire A E. punctulatum A E. flabellare B E. vitreum A * E. caeruleum AA E. caeruleum E E. caeruleum O E. luteovinctum D E. uniporum C E. caeruleum G * E. whipplei C * E. radiosum A E. uniporum M E. uniporum O E. uniporum R * * E. uniporum D E. uniporum G E. fragi B * * E. fragi D E. fragi G E. fragi J E. lawrencei AB Mamequit darter B E. burri B * E. burri C E. burri F * >95% posterior probability E. burri H * E. s. spectabile H 5 changes * E. s. spectabile X E. s. spectabile Y E. s. pulchellum AC E. s. pulchellum AB * E. s. pulchellum Z * E. s. pulchellum AZ E. s. squamosum E Figure 6. Tree resulting from Bayesian analysis of the S7 ribosomal subunit intron 1. The red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are the putative hybrids.

33 Sander vitreum A Perca flavescens A Nothonotus camurum Z Crystallaria asprella B E. cinereum A E. vitreum A Percina aurantiaca A * Percina roanoka A Ammocrypta pellicuda A * E. blennioides A * E. flabellare B * E. punctulatum A E. proelaire A * E. fonticola A E. microperca A E. luteovinctum D E. radiosum A * E. whipplei C * E. caeruleum E E. caeruleum G * E. caeruleum O E. caeruleum AA E. burri B E. burri C * E. burri F E. burri H E. fragi B E. fragi D E. fragi G E. fragi J E. lawrencei AB Mamequit darter B * E. s. pulchellum AZ E. s. spectabile H E. s. squamosum E E. uniporum M * >95% posteriror E. uniporum O probabilities E. pulchellum AC * >90% posterior E. pulchellum AB probabilities E. pulchellum Z 5 changes E. s. spectabile X * E. s. spectabile Y E. uniporum R * E. uniporum C E. uniporumD * E. uniporum G Figure 7. Tree resulting from Bayesian analysis of the Kelch repeat nuclear gene. The red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are the putative hybrids.

34 Sander vitreum A Perca flavescens A Ammocrypta pellicuda A Crystallaria asprella B * Percina aurantiaca A Percina roanoka A * E. cinereum A E. blennioides A E. vitreum A Nothonotus camurum Z * E. flabellare B * E. punctulatum A * E. microperca A * E. fonticola A E. proelaire A * E. whipplei C E. radiosum A E. luteovinctum D E. caeruleum AA E. caeruleum E * E. caeruleum G * E. caeruleum O E. burri B E. burri C * E. burri F E. burri H E. lawrencei AB Mamequit darter B E. s. spectabile H E. uniporum D * E. s. spectabile X E. s. spectabile Y * E. fragi B * E. fragi D * posterior probabilities E. fragi G >95% E. fragi J .5 change E. s. pulchellum AZ E. s. squamosum E * E. s. pulchellum AC E. s. pulchellum AB E. s. pulchellum Z E. uniporum M * E. uniporum O E. uniporum R * E. uniporum C E. uniporum G Figure 8. Tree resulting from Bayesian analysis of the Mixed lineage leukemia nuclear gene. The red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are the putative hybrids.

35 Sander vitreum A Perca flavescens A Ammocrypta pellicuda A Crystallaria asprellaB E. luteovinctum D * Percina aurantiaca A Percina roanoka A * E. cinereum A Nothonotus camurum Z * E. caeruleum E E. caeruleum AA * E. caeruleum G * E. caeruleum O * E. whipplei C E. radiosum A E. blennioides A E. vitreum A * * E. flabellare B E. punctulatum A * E. microperca A * E. fonticola A E. proelaire A E. burri B E. burri C E. burri F E. burri H E. lawrencei AB Mamequit darter B E. s. pulchellum AZ E. s. spectabile H E. squamosum E E. uniporum C * E. uniporum D E. uniporum G E. uniporum R E. s. pulchellum AC E. s. pulchellum AB E. s. pulchellum Z * E. uniporum O E. uniporum M * E. s. spectabile X E. s. spectabile Y * posterior probabilities >95% E. fragi B 1 change * E. fragi D E. fragi G E. fragi J Figure 9. Tree resulting from Bayesian analysis of the Myoglobin nuclear gene. The red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are the putative hybrids

36 Sander vitreum A Perca flavescensA Percina aurantiaca A Percina roanoka A E. cinereum A Nothonotus camurum Z E. flabellare B E. blennioides A E. vitreum A * Ammocrypta pellicuda A Crystallaria asprella B E. caeruelum E E. caeruelum G E. caeruelum O * E. caeruelum AA E. luteovinctum D E. punctulatum A E. burri B E. burri C E. burri F E. burri H E. s. pulchellum AZ E. s. spectabile H E. s. squamosum E E. s. spectabile X E. s. spectabile Y * * E. whipplei C E. radiosum A * E. lawrencei AB Mamequit darter B * E. microperca A * E. fonticola A E. proelaire A E. s. pulchellum AC * E. s. pulchellum AB E. s. pulchellum Z * >95% posterior probability E. fragi B E. fragi D * >90% posterior * probability E. fragi G 5 changes E. fragi J E. uniporum C E. uniporum D * E. uniporum G E. uniporum M E. uniporum O E. uniporum R Figure 10. Tree resulting from Bayesian analysis of the TMO-4C4 nuclear gene. The red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are the putative hybrids

37 Sander vitreum A Perca flavescens A Percina aurantiaca A * Percina roanoka A * Ammocrypta pellicuda A Crystallaria asprella B * * E. cinereum A E. blennioides A Nothonotus camurum Z E. flabellare B * * E. vitreum A E. punctulatum A E. microperca A * * * E. fonticola A E. proelaire A E. radiosum A * * E. whipplei C E. uniporum M E. caeruleum G * * E. caeruleum O * E. caeruleum E E. caeruleum AA * E. luteovinctum D E. fragi B * E. fragi D E. fragi G E. fragi J * * E. uniporum C E. uniporum D * E. uniporum G E. uniporum O E. uniporum R * E. s. pulchellum AZ * E. s. squamosum E E. s. pulchellum AC * E. s. pulchellum AB E. s. pulchellum Z * E. burri B E. burri C * > 90% posterior E. burri F probability * E. burri H * >95% posterior E. lawrencei AB probability Mamequit darter B E. s. spectabile H 5 changes * E. s. spectabile X E. s. spectabile Y Figure 11. Tree resulting from Bayesian analysis of the Rag 1 Exon 3 nuclear gene. The red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are the putative hybrids

38 Table 4. Summary of the overall resolution power of individual nuclear genes and the presence and absence of particular nodes for each individual nuclear analysis. Nuclear % Darter Percina Microperca Oligocephalus E. Ammocrypta E. Intraspecific gene Resolved clade clade clade clade spectabile and blennioides clade complex Crystallaria and E. support clade cinereum S7 36% yes yes no yes yes yes yes E. uniporum, E. s. pulchellum Kelch 26% yes yes yes yes yes no no E. uniporum MLL 45% yes yes yes yes yes no yes E. caeruleum Myoglobin 33% yes yes yes no yes no no no TMO-4C4 24% no no yes no no yes no no Rag1Ex3 50% yes yes yes yes yes yes yes E. caeruleum

39 The S7 ribosomal gene significantly resolved 36% of nodes of the Percidae phylogeny (Figure 6). The monophyly of both the Oligocephalus and the E. spectabile complex are supported with a greater than a 95% posterior probability. The Percina clade is also well supported with posterior probability support. There is significant support for a clade that includes Ammocrypta and Crystallaria, and also a clade that includes E. cinereum and E. blennioides.

The Kelch repeat phylogeny (Figure 7) is similar to that of S7 in that it shows significant posterior probability support of the Oligocephalus, E. spectabile complex and the Percina clades. There is also strong support for an E. flabellare and E. punctulatum relationship, and the Microperca clade that includes E. fonticola, E. proelaire and E. microperca.

The Mixed Lineage Leukemia (MLL) gene illustrated that in addition to the clades mentioned previously that were well supported by S7 and Kelch, a greater proportion of nodes (45%; Figure 8) are significantly resolved. There is a greater resolution of shallow nodes among the intraspecific clades. A clade including E. spectabile pulchellum and E. spectabile squamosum is monophyletic as well as three distinct clades of E. uniporum, E. fragi and E. caeruleum.

The Myoglobin phylogeny only has 33% of the nodes significantly resolved as illustrated in Figure 9. Unlike the other nuclear genes, Oligocephalus is not resolved as monophyletic, whereas significant support for the E. spectabile complex clade does exist.

The TMO-4C4 protein-coding region has the lowest resolution, with only 24% of the nodes significantly resolved (Figure 10). There is significant posterior support for an

Ammocrypta and Crystallaria clade as well as the E. fonticola, E. proelaire and E. 40 microperca clade, but there is no support for either the monophyly of the E. spectabile complex, nor the Oligocephalus clades.

Rag 1 Exon 3 is the largest nuclear fragment with 1377 bp and also has the highest percentage of the nodes that are significantly resolved (50%; Figure 11). Similar resolution was found using MLL, but Rag1 Exon 3 also showed significant resolution of an eastern clade of darters that includes E. burri, E. s. spectabile, E. lawrencei and the

Mamequit darter.

When the nuclear genes are added for a combined analysis (Figure 12) that includes 4403 bp, the resolution increases enormously. Eighty-three percent of the nodes are significantly resolved and many of the intraspecific nodes show high posterior probabilities as well as the deeper nodes. The Ammocrypta and Crystallaria clade retains a well-supported node, as does the clade that includes E. cinereum and E. blennioides.

Etheostoma uniporum and E. fragi that had mtDNA haplotypes removed from the E. spectabile complex are sister species, still distantly related to the remaining E. spectabile complex, but the E. spectabile complex is significantly monophyletic as is the

Oligocephalus subgenus.

3.3 Incongruence among the mitochondrial and nuclear genes of the Etheostoma spectabile complex.

For the parsimony ratchet analysis of the 212 individuals for the E. spectabile species complex, a 95% majority rule phylogeny was identified for both the S7 nuclear gene and the combined mtDNA. The SH test rejects the congruence of the combined mtDNA and S7 nuclear tree topologies (Table 5).

41 Sander vitreum A Perca flavescens A * Percina aurantiaca A Percina roanoka A * Ammocrypta pellicudaA * Crystallaria asprella B * Etheostoma cinereum A Etheostoma blennioides A * Nothonotus camurum Z Etheostoma vitreum A * * * Etheostoma flabellare B Etheostoma punctulatum A Etheostoma microperca A * * * Etheostoma fonticola A Etheostoma proelaire A Etheostoma luteovinctum D * * Etheostoma whipplei C Etheostoma radiosum A * * Etheostoma caeruleum G * * Etheostoma caeruleum O * Etheostoma caeruleum E Etheostoma caeruleum AA Etheostoma fragi B * Etheostoma fragi D * Etheostoma fragi G * Etheostoma fragi J Etheostoma uniporum C * * Etheostoma uniporum M * Etheostoma uniporum O * Etheostoma uniporum R * Etheostoma uniporum D * Etheostoma uniporum G * Etheostoma pulchellum AZ Etheostoma squamosum E * Etheostoma pulchellum AC * Etheostoma pulchellum AB Etheostoma pulchellum Z * Etheostoma spectabile H * Etheostoma lawrencei AB Mamequit darter B * * Etheostoma spectabile X Etheostoma spectabile Y Etheostoma burri B * >95% posterior probability * Etheostoma burri C Etheostoma burri F * > 90% posterior probability Etheostoma burri H 10 changes

Figure 12. Tree resulting from Bayesian analysis of the combined nuclear genes. The analysis includes 4363 bp and the red stars indicate nodes with a >95% posterior probability, and the red highlighted taxa are putative hybrids

42 Table 5. Shimodaira-Hasegawa (SH) test of incongruence between the mtDNA and nuclear DNA topologies of the E. spectabile species complex. Tree -ln L of -ln L of tree Difference in -ln P Tree1 2 L Bayes_S7spectabile (best) Bayes_mtDNA 3834.05 4792.09 958.05 <0.0001* spectabile All significant p-values are followed by an asterisk.

To determine whether the pattern of incongruence was present in other nuclear genes, the SH test was run for the resulting tree topologies of the additional five individual nuclear gene analyses using a subset of the taxa (Table 6). Similar to the SH analysis of the total dataset the mitochondrial data using the subset of taxa reject the congruence with the S7 tree topology as well as the rest of the nuclear gene topologies.

In addition, incongruence prevails between the nuclear tree topologies. The S7 and Rag1

Exon 3 nuclear data reject congruence between all the nuclear tree topologies as well as the mtDNA tree topology. In fact, TMO-4C4 is the only nuclear gene whose topology is not significantly different than the mtDNA topology nor any of the other nuclear gene topologies.

3.4 Total combined data analysis

The total combined dataset that includes both the mtDNA and nuclear DNA data partitions shows a high proportion of resolution (81%; Figure 13), comparable to the concatenated nuclear analysis resolution (83%). Interestingly, there was reduced support for three interior nodes that I hypothesize are due to incongruence between gene tress

43 Table 6. Shimodaira-Hasegawa (SH) test of incongruence between the Bayesian individual nuclear genes and mtDNA tree topologies. Tree Bayes_mtDNA Bayes_S7 Bayes_TMO4C4 Bayes_Mb Baues_Kelch Bayes_MLL Bayes_Rag1Ex3 Bayes_ mtDNA <0.0001* 0.131 <0.0001* 0.001* <0.0001* <0.0001* Bayes_S7 <0.0001* 0.390 <0.0001* <0.0001* <0.0001* <0.0001* Bayes_TMO4c4 <0.0001* <0.0001* 0.001* <0.0001* <0.0001* <0.0001* Bayes_Mb <0.0001* <0.0001* 0.431 0.010* <0.0001* <0.0001* Bayes_Kelch <0.0001* <0.0001* 0.052 0.022* 0.001* <0.0001* Bayes_MLL <0.0001* <0.0001* 0.083 <0.0001* 0.054 0.001* Bayes_Rag1Ex3 <0.0001* <0.0001* 0.822 0.003* 0.083 0.074 All significant p-values are followed by an asterisk.

44 Sander vitreum A Perca flavescens A * P. aurantiaca A P. roanoka A A. pellicuda A * C. asprella B E. cinereum A * * N. camurum Z * E. blennioides A E. flabellare B * * E. vitreum A E. punctulatum A E. microperca A * E. fonticola A * * E. proelaire A * E. s. pulchellum AZ * E. s. squamosum E * E. lawrencei AB * * Mamequit darter B * E. s. spectabileH E. burri B * * * E. burri C E. burri F E. fragi G * E. fragi J * E. fragi B E. fragi D * E. luteovinctum D E. radiosum A * * E. whipplei C * * E. s. pulchellum AB E. s. pulchellum AC E. s. pulchellum Z * >90% posterior E. caeruleum AA probabilities E. caeruleum E * >95% posteriror * * E. caeruleum G probabilities * E. caeruleum O * 50 changes E. burri H * E. s. spectabileX * * E. s. spectabileY E. uniporum M * E. uniporum O * E. uniporum R * E. uniporum C * E. uniporum D E. uniporum G

Figure 13. Tree resulting from the total combined data of the nuclear DNA and mtDNA data partitions (6550 bp) Bayesian analysis. The branches are scaled to the number of substitutions. The red stars illustrate >95% posterior probability, the green starts indicate >90% posterior probability and the putative hybrids are highlighted in red text.

45 inferred from the nuclear DNA and mtDNA. The SH test that compared the nuclear and mitochondrial tree topologies to that of the total combined analysis (Table 7) indicates that the total data phylogeny is significantly different than the one obtained in the combined nuclear analysis, but not significantly different from the mtDNA gene tree analysis. In the total combined analysis, there is 6550 bp, with 2187 belonging to the mtDNA partition, and 4363 bp from the sampled nuclear genes. Despite the fact that the number of sampled nucleotide sites from the nuclear genes is double that of the sites sampled from the mtDNA genome, there are more parsimony informative sites within the mtDNA that potentially swamp out the phylogenetic signal of the variation of the nuclear

DNA.

3.5 Repeatability of clades

Table 8 summarizes the repeatability of nodes (A-O, Figure 14) from the total combined phylogeny in the individual nuclear gene analyses. The nodes that are supported in both the nuclear and mitochondrial analyses are highlighted in dark grey (A,

B and O). These include the presence of the dater clade, the Percina clade and the sister relationship between E. fonticola and E. proelaire. The medium grey shading indicates that the nodes found in the combined data phylogeny are not present in any of the individual analyses (G, H, I and L). These nodes could indicate cryptic congruence, in which the individual analysis did not have significant support for the nodes so they were collapsed; yet when all the data were combined, strong bootstrap support existed.

Finally, the light shading highlights clades that have clear nuclear and mitochondrial

46 Table 7. Shimodaira-Hasegawa (SH) test of incongruence between the Bayesian tree topologies of the total combined data and the nuclear DNA and mtDNA data partitions. Tree -ln L Difference -ln L P Bayes_Total 43912.11362 (best) Bayes_mtDNA 44150.17779 238.06417 0.100 Bayes_Nuclear DNA 46819.77904 2907.66542 <0.0001 *

All significant p-values are followed by an asterisk.

47 Table 8. Repeatability of clades (A-O) in the total combined data set looking at independent individual data sets. The numbers represent the percentage of trees in posterior tree space in which the node is present. The dark shading highlights nodes that are always present in posterior tree space in both nuclear and mitochondrial analyses, the medium shading highlights nodes that are present in the total combined analysis, but not present in any of the independent analyses and the light highlights the nodes that are found in the mtDNA analysis, but not in any of the nuclear DNA analyses. A B C D E F G H I J K L M N O

Kelch 1.00 1.00 0 0 0.52 0 0 0 0 0 0 0 0 1.00 0.88

Mb 1.00 0.99 0.10 0.42 1.00 0 0 0 0 0 0 0 0 0.99 1.00

MLL 1.00 1.00 0 0.70 0 0 0 0 0 0 0 0 0 1.00 1.00

RAG1 1.00 1.00 0 0.99 0 0.16 0 0 0 0 0 0 0 1.00 1.00

S7 1.00 1.00 0 1.00 0 1.00 0 0 0 0 0 0 0 0.79 1.00

TMO-4C4 1.00 0.89 0 0.97 0.97 0 0 0 0.52 0.97 0 0 0 0.96 1.00 mtDNA 1.00 1.00 0.99 0.52 0.06 0.76 0 0 0 0.79 1.00 0.13 1.00 0 1.00

48 Sander vitreum A Perca flavescens A B Percina aurantiaca A Percina roanoka A D Ammocrypta pellicuda A Crystallaria asprella B C A E E. cinereum A Nothonotus camurum Z G E. blennioides A F E. flabellare B E. vitreum A E. punctulatum A N E. microperca A M O E. fonticola A E. proelaire A H E. s. pulchellumAZ E. s. squamosum E E. lawrencei AB Mamequit darter B E. s. spectabile H I E. burri B E. burri C E. burri F E. fragi G E. fragi J E. fragi B E. fragi D J E. luteovinctum D E. radiosum A E. whipplei C K E. s. pulchellum AB E. s. pulchellum AC E. s. pulchellum Z E. caeruleum AA L E. caeruleum E E. caeruleum G E. caeruleum O E. burri H 50 changes E. s. spectabile X E. s. spectabile Y E. uniporum M E. uniporum O E. uniporum R E. uniporum C E. uniporum D E. uniporum G

Figure 14. Phylogeny resulting from the Bayesian analysis of the total combined data with the nodes used in the repeatability analysis labeled.

49 conflict (C, K and M). In these cases, the nodes in the total combined data are present in the mtDNA but not in the nuclear DNA, indicating a swamping effect of mtDNA at these nodes. For instance, node C represents the combined relationship of Ammocrypta,

Crystallaria, E. cinereum and Nothonotus. This node was found within the mtDNA, though a polytomy exists explaining the interrelationships. On the other hand, the nuclear data provide two relationships: 1) Ammocrypta and Crystallaria and 2) E. cinereum and Nothonotus, but did not combine them into one single node C. The remaining nodes show variable support depending on the individual gene analysis used, illustrating the differing evolutionary histories among the nuclear and mitochondrial gene.

50 4. Discussion

4.1 Phylogeography of the Etheostoma spectabile complex

Defining species based on molecular or morphological characters is an ongoing debate and depends largely on the species concept that is used. Concordant patterns of divergence observed across multiple, independent genetically based characters have been considered strong evidence of phylogenetic distinction (Avise 2004). For my thesis, I present the molecular analysis of multiple independent data sets (two mitochondrial and six nuclear genes) and utilize the phylogenetic species concept in which species are recognizably monophyletic to determine whether there is phylogenetic support for the recent species delimitations based on meristics and breeding coloration within the E. spectabile complex. In addition, I compared mitochondrial and nuclear data and found that hybridization is a process that has implications for both the ancient and recent evolutionary histories of the E. spectabile species complex.

4.1.1 Species relationships

I can now use molecular phylogenies to address several assertions made about the interrelationships of the species complex based on morphological characters. The mtDNA data was most valuable for phylogeographic patterns of the E. spectabile species complex, and the combined nuclear analysis was most valuable for detecting the evolutionary history of hybridization and introgression within the complex.

First, the proposal that E. tecumsehi is the putative sister taxon to E. lawrencei that had evolved as a peripheral isolate population of E. lawrencei (Ceas and Burr 2002) is not supported using the mitochondrial genes. Instead, the mtDNA suggests that E. tecumsehi is the sister taxon to E. kantuckeense. The Sheltowee darter was another 51 undescribed species with a distribution adjacent to E. lawrencei, yet is also not closely related. Surprisingly, the Sheltowee darter of Kentucky is closely related to E. s. spectabile populations west of the Mississippi River and E. burri rather than to the species east of the Mississippi in the Tennessee and Kentucky drainages.

My mitochondrial analysis supports the conclusion Ceas and Page (1997) that E. s. spectabile is actually a composite of several diagnosable species based on male breeding coloration. Indeed, distinct genetic clusters sort between multiple drainages.

Distinct genetic clades are found in 1) the Osage River system in Missouri, 2) Illinois

River System in Illinois, 3) the Missouri/Mississippi River system in Missouri, 4) the St.

Francis River System in Missouri and Arkansas and 5) the Meramac River system in

Missouri.

Etheostoma lawrencei is described as having among the most widespread and complicated distribution in the E. spectabile complex (Ceas and Burr 2002). Found in three disjunct river systems in Tennessee and Kentucky, E. lawrencei is recognized as one typological species. According to the mitochondrial haplotypes, three distinct clades are congruent with the disjunct distribution. One individual from the Roaring River,

Tennessee has a haplotype that is basal to a Cumberland River, Tennessee clade and a

Green River, Kentucky clade. In addition to the geographic structure found within the haplotypes, E. lawrencei is not reciprocally monophyletic. Haplotypes of the Ihiyo darter, found in the Cumberland River and adjacent Caney Fork River of Tennessee are interspersed throughout the distinct clades, indicating possible introgression, incomplete lineage sorting of two very closely related species, or the over splitting of these fishes.

52 Distler’s systematic analysis (1968) and the subsequent lactate dehydrogenase

(LDH) analysis showed that the greatest differentiation in the E. spectabile complex occur in streams draining the Ozark Uplands (Wiseman et al. 1978) thus supporting the origin of the species complex in the Ozark Uplands. A parallel pattern was found in the phylogeography of E. caeruleum (Ray et al. In press). My mitochondrial analysis provides additional support for the hypothesis of a Ozark origin of the E. spectabile complex. The Ozark darter, found in the North Fork of the White River, Arkansas, is basal to the remaining species in the Ozarks and those species that are found in the streams east of the Mississippi River. In addition to the basal Ozark darter, there are 5 distinct haplotype clusters in adjacent streams that likely evolved in isolation in Missouri and Arkansas (E. fragi, E. uniporum, E. burri, and three clades within the E. s. spectabile) supporting the hypothesis of high diversity in the Ozarks, followed by an eastern dispersal into Tennessee and Kentucky.

In terms of geography, E. fragi is restricted to the Strawberry River of the Black

River system, Arkansas, while E. uniporum is abundant in the upland tributaries of the

Black River System, Missouri and Arkansas (Ceas and Burr 2002). The close geography and morphology of these species indicates a likely sister relationship. Yet the mitochondrial haplotypes of these species and between these species and the species of the E. spectabile complex are extremely divergent. The combined nuclear analysis highlighted the discordant pattern of the mitochondrial analysis and indicates that E. uniporum and E. fragi are sister species.

53 4.1.2 Status of species

Of the ten described species (E. s. spectabile, E. s. pulchellum, E. s. squamosum,

E. fragi, E. uniporum, E. bison, E. burri, E. tecumsehi, E. kantuckeense and E. lawrencei) and the four undescribed species (Ozark darter, Mamequit darter, Ihiyo darter and

Sheltowee darter) only 8 are reciprocally monophyletic in mtDNA gene trees.

The mitochondrial data do not support the monophyly of E. s. spectabile. In fact, the monophyletic E. burri is closely related to E. s. spectabile of the Petite Saline River,

Missouri, producing a paraphyletic species.

The parapatric E. lawrencei and the Ihiyo darter are also not monophyletic. The

Ihiyo darter, found in adjacent Cumberland streams, has haplotypes shared amongst the

E. lawrencei haplotypes of the Cumberland and Green drainages, resulting in two polyphyletic species. The same is true for the polyphyly of E. s. pulchellum and E. s. squamosum through the admixture of haplotypes.

The final species is not monophyletic based on my genetic data is E. kantuckeense. Endemic to the Barren River System, E. kantuckeense populations east of the Barren River Lake are more closely related to the monophyletic E. tecumsehi, a species endemic to Pond River of the Green River system, than to E. kantuckeense populations of the Barren River system, but found in tributaries south and east of the

Barren River Lake.

With only eight out of fourteen monophyletic species, we can ask whether the members of the Etheostoma spectabile complex were over-split. But the answer is not simple. Two species pairs share haplotypes, E. s. squamosum and E. s. pulchellum, and

E. lawrencei and the Ihiyo darter, respectively, which could indicate a possible case of 54 over-splitting. On the other hand, there appear to be many more distinct clusters of haplotypes than species designations. The greater number of haplotype clusters that correspond to distinct river drainages support a possible conservative nature of diagnosing species, and a number of undescribed species. In any case, there is a need for further investigation.

4.2 Hybridization and introgression

I used the nuclear data to develop two main conclusions regarding hybridization.

First, introgressive hybridization has occurred at multiple times throughout the evolutionary history of the E. spectabile species complex. The nuclear genes support the monophyly of E. spectabile complex, which is incongruent with the mtDNA gene tree.

This incongruence, specifically the divergent mitochondrial haplotypes of E. fragi and E. uniporum, is evidence of ancient introgression. A pattern of more recent hybridization is also observed in which there is heterospecific mtDNA haplotype introgression in only a single individual in a population, although this pattern is seen only among populations west of the Mississippi River. My second conclusion is that there is an asymmetrical pattern of mtDNA introgression between Oligocephalus species and species in the E. spectabile complex.

Etheostoma fragi’s mitochondrial haplotypes are extremely divergent from the E. spectabile complex haplotypes, up to 20.2% uncorrected sequence divergence. Yet the combined nuclear phylogeny illustrates that E. fragi is a member of a monophyletic E. spectabile group. So the question arises, where did the mitochondrial genome of E. fragi come from? In order to resolve this, it will be necessary to develop a phylogeny using complete taxon sampling in order to discern the mitochondrial donator. Currently, E. 55 fragi haplotypes are related to none of Oligocephalus haplotypes nor to any of the other outgroup taxa haplotypes I studied. The isolated and divergent nature of E. fragi’s haplotypes may illustrate a possible mtDNA fossil, in that whatever species donated the mtDNA has subsequently become extinct.

In addition to the haplotypes of E. fragi, the placement of E. uniporum haplotypes indicates ancient introgression. Etheostoma uniporum haplotypes form a clade that is completely nested within E. caeruleum haplotypes. Ancient hybridization with E. caeruleum and fixation of that E. caeruleum haplotype within E. uniporum is the best explanation of the observed pattern.

There are several populations of the E. spectabile species complex that have heterospecific mtDNA genomes introgression, but only in a few individuals of the population. Etheostoma uniporum is one of the lineages that shows evidence of current hybridization with E. caeruleum. Single individuals from two separate populations

(EuniM, EuniT) have heterospecific mitochondrial haplotypes that are not seen in other individuals from the same population. The same is true of the introgressed haplotypes of the individuals of E. s. spectabile (EspeX, EspeY) and E. burri (EburH). The population of E. s. pulchellum (EpulAB, EpulAC, EpulZ) also showed introgressed mitochondrial

DNA related to E. whipplei. In all of these cases the nuclear and meristic data indicate that these individuals are good members of the E. spectabile complex.

The ancient and more recent hybridization events also indicate an asymmetry in the mtDNA haplotype transmission. Often, the exchange of genes between species is unequal (Bacilieri et al 1996; Mendelson 2003) and the tendency is for DNA to flow

56 from common to rare species (Dowling et al. 1989, Taylor and Herbert 1993). In this case, the E. caeruleum or E. whipplei mtDNA haplotypes were found in E. spectabile individuals, indicating that E. spectabile complex males mated more frequently with E. caeruleum or E. whipplei females than vice versa.

The mechanism of the unequal introgression could result from differential viability of the hybrid offspring or prezygotic asymmetry in the mating behavior.

Branson and Campbell (1969) documented a hybrid population of E. spectabile and E. radiosum cyanorum in the Thomas Bricken Spring, Oklahoma. Hybrids were diagnosed based on meristic and color characteristics, and the authors noted that the hybrids tended towards E. spectabile characteristics, meaning that they had more E. spectabile diagnosable characteristics than E. radiosum cyanorum characteristics. In laboratory crosses Hubbs (1967) noted that eggs of E. spectabile inhibit the sperm of E. caeruleum while the eggs of E. caeruleum invigorate the sperm of E. spectabile. The explanation could also be applied my observation that the only hybrids detected had E. caeruleum mtDNA transferred into E. spectabile complex members. In contrast to Hubbs’ later results (1967), he had earlier found little significant difference between egg viability that could account for asymmetrical mtDNA transfer (Hubbs 1959). He noted that the viability of the eggs resulting from a cross with an E. spectabile female with a male of E. spectabile was 0.62, and not significantly different from the viability of an E. spectabile female and E. caeruleum male, 0.66. Unfortunately the reciprocal cross was not performed, but the intraspecific and interspecific (in the wrong mtDNA transmission

57 direction) egg viability was not significantly different indicating that low viability of eggs may not be the mechanism of asymmetrical E. caeruleum mtDNA transmission.

Asymmetry in mating probabilities could also result in the unidirectional haplotype transfer E. caeruleum mtDNA into species of the E. spectabile complex.

Mendelson (2003) illustrated that there was asymmetrical behavior between E. spectabile pulchellum and E. radiosum paludosum that could result in unidirectional mtDNA transfer. Her data indicated that E. s. pulchellum males mated with E. r. paludosum females significantly more frequently than vice versa. Although the E. r. paludosum male was larger and aggressively defended both species females, he failed to mount and spawn with the E. s. pulchellum female, resulting in the majority of E. s. pulchellum females mating with E. s. pulchellum males that sneaked past the territorial E. r. paludosum male. Whether the female failed to send the correct signal, or the male was not stimulated by the female’s signal is unclear. What is clear is that I have evidence for asymmetric transfer of E. caeruleum mtDNA haplotypes into individuals of the E. spectabile complex in my comparison of mtDNA and nuclear DNA.

My results show that introgression can cloud the phylogeography of a species complex. Previous morphological studies and my mitochondrial data presented here are consistent with the reciprocally monophyletic nature of half the species in the Etheostoma spectabile complex. The mtDNA analysis also reveals previously undescribed geographic structure and cryptic species in E. s. spectabile and E. kantuckeense. But, the mtDNA analysis also indicates that E. uniporum and E. fragi are more distantly related to the E. spectabile species complex than suggested by morphology and geography. In

58 contrast, the nuclear DNA analyses show that E. uniporum and E. fragi are members of the complex. Furthermore, the incongruence between the mtDNA and nuclear DNA results indicate that introgression is an evolutionary process that can be seen throughout the history of the E. spectabile species complex. The molecular evidence of recent hybridization events invokes questions about the extent of current reproductive isolation.

Future behavioral investigations are required to study the processes or mechanisms of the recent hybridization events. Understanding the prevalence of introgression is crucial for future investigation of the evolution of these fishes.

59 Literature Cited.

60 Literature Cited

Allendorf, F. W., R. F. Leary, P. Spruell, and J. K. Wenburg. 2001. The problems with

hybrids: setting conservation guidelines. Trends in Ecology & Evolution 16:613-

622.

Alves, M. J., M. M. Coelho, and M. J. Collares-Pereira. 2001. Evolution in action

through hybridisation and polyploidy in an Iberian freshwater fish: a genetic

review. Genetica 111:375-385.

Arnold, M. L. 1992. Natural hybridization as an evolutionary process. Annual Review of

Ecology and Systematics 23:237-261.

Arnold, M. L. 1997. Natural hybridization and evolution. Oxford University Press,

Oxford.

Avise, J. C. 1989. Gene trees and organismal histories - A phylogenetic approach to

population biology. Evolution 43:1192-1208.

Avise, J. C. 2004. Molecular markers, natural history and evolution. 2nd edition. Sinauer

Associates, Sunderland, MA.

Bacilieri, R., A. Ducousso, R. J. Petit, and A. Kremer. 1996. Mating system and

asymmetric hybridization in a mixed stand of European oaks. Evolution 50:900-

908.

Bailey, R. M. and W. A. Gosline. 1955. Variation and systematic significance of

vertebral counts in the American fishes of the family Percidae. Miscellaneous

Publications. Museum of Zoology, University of Michigan 93:1-44.

Barraclough, T. G., J. E. Hogan, and A. P. Vogler. 1999. Testing whether ecological

factors promote cladogenesis in a group of tiger beetles (Coleoptera : 61 Cicindelidae). Proceedings of the Royal Society of London Series B-Biological

Sciences 266:1061-1067.

Bhagabati, N. K., J. L. Brown, and B. S. Bowen. 2004. Geographic variation in Mexican

jays (Aphelocoma ultramarina): local differentiation, polyphyly or hybridization?

Molecular Ecology 13:2721-2734.

Branson, B. A., and J. B. Campbell. 1969. Hybridization in darters Etheostoma spectabile

and Etheostoma radiosum cyanorum. Copeia:70-&.

Brown, R. P., N. M. Suarez, and J. Pestano. 2002. The Atlas mountains as a

biogeographical divide in North-West Africa: evidence from mtDNA evolution in

the Agamid lizard Agama impalearis. Molecular Phylogenetics and Evolution

24:324-332.

Ceas, P. A., and Burr, B. M. 2002. Etheostoma lawrencei, a new species of darter in the

E. spectabile species complex (Percidae: subgenus Oligocephalus), from

Kentucky and Tennessee. Ichthylo. Explor. Freshwat 13:203-216.

Ceas, P. A., and L. M. Page. 1997. Systematic studies of the Etheostoma spectabile

complex (Percidae; subgenus Oligocephalus), with descriptions of four new

species. Copeia:496-522.

Chen, W. J., C. Bonillo, and G. Lecointre. 2003. Repeatability of clades as a criterion of

reliability: a case study for molecular phylogeny of Acanthomorpha (Teleostei)

with larger number of taxa. Molecular Phylogenetics and Evolution 26:262-288.

Childs, M. R., A. A. Echelle, and T. E. Dowling. 1996. Development of the hybrid swarm

between pecos pupfish (Cyprinodontidae: Cyprinodon pecosensis) and

62 sheepshead minnow (Cyprinodon variegatus): A perspective from allozymes and

mtDNA. Evolution 50:2014-2022.

Chow, S., and K. Hazama. 1998. Universal PCR primers for S7 ribosomal protein gene

introns in fish. Molecular Ecology 7:1255-1256.

Demarais, B. D., T. E. Dowling, M. E. Douglas, W. L. Minckley, and P. C. Marsh. 1992.

Origin of Gila-seminuda (Teleostei, Cyprinidae) through introgressive

hybridization - Implications for evolution and conservation. Proceedings of the

National Academy of Sciences of the United States of America 89:2747-2751.

Dettai A. and Lecointre, G. 2005. Further support for the clades obtained by multiple

molecular phylogenies in the acanthomorph bush. C R Biol. 328:674-689.

Distler, D. A. 1968. Distribution and variation of Etheostoma spectabile (Agassiz)

(Percidae, Teleostei). Univ. Kans. Sci. Bull. 48:143-208.

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.

Dowling, T. E., and M. R. Childs. 1992. Impact of hybridization on a threatened trout of

the southwestern United-States. Conservation Biology 6:355-364.

Dowling, T. E., and B. D. Demarais. 1993. Evolutionary significance of introgressive

hybridization in cyprinid fishes. Nature 362:444-446.

Dowling, T. E., and C. L. Secor. 1997. The role of hybridization and introgression in the

diversification of animals. Annual Review of Ecology and Systematics 28:593-

619. 63 Echelle, A. A., A. F. Echelle, M. H. Smith, and L. G. Hill. 1975. Analysis of genic

continuity in a headwater fish, Etheostoma radiosum (Percidae). Copeia:197-204.

Echelle, A. A., J. R. Schenk, and L. G. Hill. 1974. Etheostoma spectabile-E. radiosum

hybridization in Blue River, Oklahoma. American Midland Naturalist 91:182-

194.

Felsenstein, J. 1981. Evolutionary trees from gene-frequencies and quantitative characters

- finding maximum-likelihood estimates. Evolution 35:1229-1242.

Fitch, W. M. 1971. Toward defining course of evolution - minimum change for a specific

tree topology. Systematic Zoology 20:406-&.

Funk, D. J., and K. E. Omland. 2003. Species-level paraphyly and polyphyly: Frequency,

causes, and consequences, with insights from animal mitochondrial DNA. Annual

Review of Ecology Evolution and Systematics 34:397-423.

Grant, P. R., and B. R. Grant. 1992. Hybridization of bird species. Science 256:193-197.

Gromicho, M., J. P. Coutanceau, C. Ozouf-Costaz, and M. J. Collares-Pereira. 2006.

Contrast between extensive variation of 28S rDNA and stability of 5S rDNA and

telomeric repeats in the diploid-polyploid Squalius alburnoides complex and in its

maternal ancestor Squalius pyrenaicus (Teleostei, Cyprinidae). Chromosome

Research 14:297-306.

Gyllensten, U., R. F. Leary, F. W. Allendorf, and A. C. Wilson. 1985. Introgression

between 2 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 University

Press, New York. 64 Hasegawa, M., H. Kishino, and T. A. Yano. 1985. Dating of the human ape splitting by a

molecular clock of mitochondrial-DNA. Journal of Molecular Evolution 22:160-

174.

Howard, D. J. 1993. Reinforcement: Origin, dynamics, and fate of an evolutionary

hypothesis. Pp. 46-69 in R. G. Harrison, ed. Hybrid zones and the evolutionary

process. Oxford University Press, New York.

Hubbs, C. 1959. Laboratory hybrid combinations among Etheostomatine fishes. Texas

Journal of Science 11:49-56.

Hubbs, C. 1967. Geographic variations in survival of hybrids between Etheostomatine

fishes. Bull. Texas Mem. Mus. 13:1-72.

Hubbs, C. L. 1955. Hybridization between fish species in nature. Systematic Zoology

4:1-20.

Huelsenbeck, J. P., and F. Ronquist. 2001. MRBAYES: Bayesian inference of

phylogenetic trees. Bioinformatics 17:754-755.

Kimura, M. 1980. A simple method for estimating evolutionary rates of base

substitutions through comparative studies of nucleotide-sequences. Journal of

Molecular Evolution 16:111-120.

Kimura, M. 1981. Estimation of evolutionary distances between homologous nucleotide-

sequences. Proceedings of the National Academy of Sciences of the United States

of America-Biological Sciences 78:454-458.

Kocher, T. D., J. A. Conroy, K. R. McKaye, J. R. Stauffer, and S. F. Lockwood. 1995.

Evolution of NADH dehydrogenase subunit 2 in east African cichlid fish.

Molecular Phylogenetics and Evolution 4:420-432. 65 Larget, B., and D. L. Simon. 1999. Markov chain Monte Carlo algorithms for the

Bayesian analysis of phylogenetic trees. Molecular Biology and Evolution

16:750-759.

Lewontin, R. C., and L. C. Birch. 1966. Hybridization as a source of variation for

adaptation to new environments. Evolution 20:315-336.

Linder, A. D. 1955. The fishes of the Blue River in Oklahoma with descriptions of two

new percid hybrid combinations. American Midland Naturalist 54:173-191.

Linder, A. D. 1958. Behavior and hybridization of two species of Etheostoma (Percidae).

Trans. Ky. Acad. Sci. 61:195-212.

Lopez, J. A., W. J. Chen, and G. Orti. 2004. Esociform phylogeny. Copeia:449-464.

Maddison, D. R. and W. P. Maddison. 2000. MacClade 4: Analysis of phylogeny and

character evolution. Sinauer Associates, Sunderland, MA.

Maddison, W. P. 1997. Gene trees in species trees. Systematic Biology 46:523-536.

Mendelson, T. C. 2003. Evidence of intermediate and asymmetrical behavioral isolation

between orangethroat and orangebelly darters (Teleostei : Percidae). American

Midland Naturalist 150:343-347.

Meyer, A., W. Salzburger, and M. Schartl. 2006. Hybrid origin of a swordtail species

(Teleostei : Xiphophorus clemenciae) driven by sexual selection. Molecular

Ecology 15:721-730.

Moore, W. S. 1995. Inferring phylogenies from mtdna variation - mitochondrial-gene

trees versus nuclear-gene trees. Evolution 49:718-726.

66 Morrison, C. L., D. P. Lemarie, R. M. Wood, and T. L. King. 2006. Phylogeographic

analyses suggest multiple lineages of Crystallaria asprella (Percidae :

Etheostominae). Conservation Genetics 7:129-147.

Near, T. J. 2002. Phylogenetic relationships of Percina (Percidae : Etheostomatinae).

Copeia:1-14.

Near, T. J., and B. P. Keck. 2005. Dispersal, vicariance, and timing of diversification in

Nothonotus darters. Molecular Ecology 14:3485-3496.

Near, T. J., L. M. Page, and R. L. Mayden. 2001. Intraspecific phylogeography of

Percina evides (Percidae : Etheostomatinae): an additional test of the Central

Highlands pre-Pleistocene vicariance hypothesis. Molecular Ecology 10:2235-

2240.

Near, T. J., J. C. Porterfield, and L. M. Page. 2000. Evolution of cytochrome b and the

molecular systematics of Ammocrypta (Percidae : Etheostomatinae). Copeia:701-

711.

Nielsen, J. L., C. Gan, and W. K. Thomas. 1994. Differences in genetic diversity for

mitochondrial-DNA between hatchery and wild populations of Oncorhynchus.

Canadian Journal of Fisheries and Aquatic Sciences 51:290-297.

Nixon, K. C. 1999. The Parsimony Ratchet, a new method for rapid parsimony analysis.

Cladistics-the International Journal of the Willi Hennig Society 15:407-414.

Page, L. M. 1981. The genera and subgenera of darters (Percidae, Etheostomatini).

Occasional Papers of the Museum of Natural History 90:1-69.

Page, L. M. 1983. Handbook of darters. TFH Publications, Neptune City. NJ.

67 Page, L. M., M. Hardman, and T. J. Near. 2003. Phylogenetic relationships of barcheek

darters (Percidae : Etheostoma, Subgenus Catonotus) with descriptions of two

new species. Copeia:512-530.

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.

Posada, D., and K. A. Crandall. 1998. MODELTEST: testing the model of DNA

substitution. Bioinformatics 14:817-818

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:632-

637.

Ray, J. M., Wood, R. M., Simons, A. M. 2006. Phylogeography and post-glacial

colonization patterns of the rainbow darer, Etheostoma caeruleum (Teleostei:

Percidae). Journal of Biogeography:1-9.

Rhymer, J. M., and D. Simberloff. 1996. Extinction by hybridization and introgression.

Annual Review of Ecology and Systematics 27:83-109.

Rieseberg, L. H. 1997. Hybrid origins of plant species. Annual Review of Ecology and

Systematics 28:359-389.

Rosenfield, J. A., S. Nolasco, S. Lindauer, C. Sandoval, and A. Kodric-Brown. 2004. The

role of hybrid vigor in the replacement of Pecos pupfish by its hybrids with

sheepshead minnow. Conservation Biology 18:1589-1598.

68 Salzburger, W., S. Baric, and C. Sturmbauer. 2002. Speciation via introgressive

hybridization in East African cichlids? Molecular Ecology 11:619-625.

Schultz, R. J. 1969. Hybridization, unisexuality, and polyploidy in Teleost Poeciliopsis

(Poeciliidae) and other vertebrates. American Naturalist 103:605-619.

Seehausen, O. 2004. Hybridization and adaptive radiation. Trends in Ecology &

Evolution 19:198-207.

Shaw, K. L. 2002. Conflict between nuclear and mitochondrial DNA phylogenies of a

recent species radiation: What mtDNA reveals and conceals about modes of

speciation in Hawaiian crickets. Proceedings of the National Academy of

Sciences of the United States of America 99:16122-16127.

Sloss, B. L., N. Billington, and B. M. Burr. 2004. A molecular phylogeny of the Percidae

(Teleostei, Perciformes) based on mitochondrial DNA sequence. Molecular

Phylogenetics and Evolution 32:545-562.

Smith G. R., R. R. Miller and W. D. Sable. 1979. Species relationships among fishes of

the genus Gila in the upper Colorado River drainage. Proc. 1st Conf. Sci. Res.

Nat. Parks. 1:613–623.

Smith, G. R. 1992. Introgression in fishes - Significance for paleontology, cladistics, and

evolutionary rates. Systematic Biology 41:41-57.

Smith, P. F., A. Konings, and I. Kornfield. 2003. Hybrid origin of a cichlid population in

Lake Malawi: implications for genetic variation and species diversity. Molecular

Ecology 12:2497-2504.

69 Streelman, J. T., R. Zardoya, A. Meyer, and S. A. Karl. 1998. Multilocus phylogeny of

cichlid fishes (Pisces : Perciformes): Evolutionary comparison of microsatellite

and single-copy nuclear loci. Molecular Biology and Evolution 15:798-808.

Swofford, D. L. 2001. PAUP. Version 4.01. Smithsonian Institution.

Tamura, K., and M. Nei. 1993. Estimation of the number of nucleotide substitutions in

the control region of mitochondrial-DNA in humans and chimpanzees. Molecular

Biology and Evolution 10:512-526.

Taylor, D. J., and P. D. N. Hebert. 1993. Habitat-dependent hybrid parentage and

differential introgression between neighboringly sympatric daphnia species.

Proceedings of the National Academy of Sciences of the United States of America

90:7079-7083.

Taylor, E. B., Boughman, J. W., Groenenboom, M., Sniatynski, M., Schluter, D. and

Gow, J. L. 2006. Speciation in reverse: morphological and genetic evidence of the

collapse of a three-spined stickleback (Gasterosteus aculeatus) species pair.

Molecular Ecology 15:343-355.

Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997.

The CLUSTAL_X windows interface: flexible strategies for multiple sequence

alignment aided by quality analysis tools. Nucleic Acids Research 25:4876-4882.

Verspoor, E., and J. Hammar. 1991. Introgressive hybridization in fishes - the

Biochemical-Evidence. Journal of Fish Biology 39:309-334.

Wiseman, E. D., A. A. Echelle, and A. F. Echelle. 1978. Electrophoretic evidence for

subspecific differentiation and intergradation in Etheostoma spectabile (Teleostei

Percidae). Copeia:320-327. 70 Zharkikh, A. 1994. Estimation of evolutionary distances between nucleotide-sequences.

Journal of Molecular Evolution 39:315-329.

71 Vita

Christen Marie Bossu was born in Cincinnati, Ohio, on November 12, 1979. She attended elementary school in the Lakota School District and junior high school in the

Sycamore School District in Cincinnati, Ohio. For her freshman and sophomore year of high school she attended Ursuline Academy in Cincinnati, Ohio, followed by junior and senior year at Ursuline Academy in Dallas, Texas where she graduated with honors in

1998. The following August she went to Purdue University and graduated with honors with a Bachelor of Science specializing in Ecology and Evolutionary Biology in 2002.

She pursued postgraduate research in New South Wales, Australia, and Paramaribo,

Suriname for a year before entering the University of Tennessee in 2003, where she received a M.S. in Ecology and Evolutionary Biology in 2006. Christen will begin pursuing a Ph.D. in Ecology and Evolutionary Biology at Yale University, New Haven,

Connecticut in August 2006.

72