1 COMPARATIVE MOR PHOLOGY AND EVOLUTION OF GENITAL PAPILLA E IN A GENUS OF DARTERS (PERCIDAE: ETHEOSTOMA ) By ZACHARY PIERCE MARTIN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 201 3
2 201 3 Zachary Pierce Martin
3 To all the researchers who have included notes about genital pa pillae in the darter literature
4 ACKNOWLEDGMENTS I thank Dr. Larry Page for his continued guidance, encouragement, and constructive criticism on my research. I t hank Drs. Max Nickerson and Jimmy Liao for their const ructive criticism and comments. Also, I thank Drs. Gordon Burleigh and Rebecca Kimball who provided s uggestions on methodology and editorial notes. A special thanks goes to Dr. David Reed at FLMNH for allowing access to his photograph ic equipment. I thank Lunide Orleus for her help with taking photographs. For specimen loans I thank Dr. Jon athan Ambruste r (AUM), Dr. Chris Taylor (INHS), Dr. Wayne Starnes (NCSM), Dr. Brooks Burr (SIU), and Mr. Robert Robins (UF)
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LI ST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRA CT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 14 Recent Phylogenetic Hypotheses for Etheostoma ................................ .................. 14 Taxonomic Classification of Darters ................................ ................................ ....... 16 Trait Diversity in Etheostoma ................................ ................................ .................. 16 Genital Papilla Morphology ................................ ................................ ............... 17 Spawning Behavior ................................ ................................ .......................... 19 Reproductive Traits in an Evolutionary Context ................................ ...................... 20 St udy Objectives ................................ ................................ ................................ ..... 21 2 METHODS ................................ ................................ ................................ .............. 23 Comparative Morphology ................................ ................................ ........................ 23 Overall Shape (0 = Mound, 1 = Tube, 2 = Rosette, 3 = Mound Tube Intermediate) ................................ ................................ ................................ 24 Basal Platform (0 = Absent, 1 = Present) ................................ ......................... 25 Spatulate (0 = Not spatulate, 1 = Spatulate) ................................ ..................... 25 Terminal Concavity (0 = Absent, 1 = Present) ................................ .................. 25 Genital Pore Orientation on Tube Papillae, GPO (0 = Downward, 1 = Anteriad, 2 = Posteriad) ................................ ................................ ................ 26 Genital Pore Position on Tube Papillae, GPP (0 = Genital Pore is Not at Distal Most Point of Genital Papilla, 1 = Genital Pore Positioned at Distal Tip of Genital Papilla) ................................ ................................ .................... 26 Villi on Tube Papillae (0 = Absent, 1 = Present) ................................ ............... 26 Bifurcated Tube Papillae (0 = No Bifurcat ion, 1 = Bifurcation) .......................... 27 Villi on Rosette Papillae (0 = Absent, 1 = Present) ................................ ........... 27 Bifurcated Papillar Shield (0 = Absent, 1 = Pres ent) ................................ ........ 27 Bifurcated Papillar Platform (0 = Absent, 1 = Present) ................................ ..... 27 Pleated Papillar Platform (0 = Absent, 1 = Present) ................................ ......... 28 Genital Pore Position on Mound tube Papillae (0 = Displaced from Distal Most Point of Papilla, 1 = Positioned at Distal Most Point of Papilla) ............ 28 Villi on Mound tube Papillae (0 = Absent, 1 = Present) ................................ .... 28
6 Phylogenetic Data ................................ ................................ ................................ ... 28 Ancestral State Reconstruction ................................ ................................ ............... 29 Materials Examined ................................ ................................ ................................ 31 Ammocrypta ................................ ................................ ................................ ..... 31 Crystallaria ................................ ................................ ................................ ....... 31 Etheostoma ................................ ................................ ................................ ...... 32 Percina ................................ ................................ ................................ ............. 36 3 RESULTS ................................ ................................ ................................ ............... 37 Genital Papilla Morphology ................................ ................................ ..................... 37 Morphological Character Evolution ................................ ................................ ......... 37 Overall Shape ................................ ................................ ................................ ... 38 AFLP phylogeny ................................ ................................ ......................... 38 Nuclear DNA phylogeny ................................ ................................ ............. 39 3TE phylogeny ................................ ................................ ........................... 40 5TE phylogeny ................................ ................................ ........................... 41 Variation In Rosette Papillae ................................ ................................ ............ 42 AFLP phylogeny ................................ ................................ ......................... 42 Nuclear DNA phylogeny ................................ ................................ ............. 43 3TE phylogeny ................................ ................................ ........................... 43 5TE phylogeny ................................ ................................ ........................... 44 Basal Platform ................................ ................................ ................................ .. 44 3TE phylogeny ................................ ................................ ........................... 45 5TE phylogeny ................................ ................................ ........................... 45 AFLP and nDNA phylogenies ................................ ................................ .... 46 Spawning Behavior Guilds ................................ ................................ ...................... 46 Egg Burying Guild ................................ ................................ ............................ 4 7 Surface burying ................................ ................................ .......................... 47 Diving ................................ ................................ ................................ ......... 47 Egg Attaching Guild ................................ ................................ ......................... 48 Rock attachers ................................ ................................ ........................... 48 Macrophyte/Plant material attachers ................................ ......................... 49 Algae attachers ................................ ................................ .......................... 49 Egg Clumping Guild ................................ ................................ ......................... 49 Egg Clustering Guild ................................ ................................ ........................ 50 Alpha clustering ................................ ................................ ......................... 50 Beta clustering ................................ ................................ ........................... 51 Gamma clustering ................................ ................................ ...................... 51 4 DISCUSSION ................................ ................................ ................................ ......... 79 Comparative Genital Papillae Morphology ................................ .............................. 79 Character Evolution and Phylogenetic Implications ................................ ................ 84 Genital Papilla Morphology and Egg Deposition ................................ ..................... 88 The Role of Genital Papillae in the Diversification of Etheostoma .......................... 90
7 APPENDIX: GLOSSARY ................................ ................................ ............................ 97 LIST OF REFERENCES ................................ ................................ ............................... 98 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 105
8 LIST OF TABLES Table page 3 1 Character coding for the 14 characters describing variation in genital papillae morphology in Etheostoma and two outgroup species, Ammocrypta vivax a character as inapplicable to a species due to the overall shape of its genital papilla. ................................ ........ 52 3 2 Retention indices yielded when applying each character to each of the four phylogenies tested. T he scale for retention index is 0 to 1 and an index approaching 1 represents a character that fits extremely well on a phylogeny, displaying little homoplasy across the tree. ................................ ........................ 56 3 3 A behavior al guild classification of 152 darters in Etheostoma along with Ammocrypta vivax and Percina maculata based on mode of egg deposition ..... 57 4 1 The overall shape states, coarse behavioral guilds, and fine behavioral guilds of each species examined in this study grouped by distinct morphologies. ........ 92
9 LIST OF FIGURES Figure page 3 1 V entral views of eight species having a mound morphology. All papillae scored as mound had a posteriorly facing genital pore surrounded by several small villi. ................................ ................................ ................................ ............ 62 3 2 Lateral close ups of eight sp ecies score d as having tube papillae. .................... 63 3 3 Lateral and ventral view photographs of species scored as having a rosette papillae ................................ ................................ ................................ ............... 64 3 4 Ventr al photographs of six species scored as having intermediate mound tube papillae. ................................ ................................ ................................ ..... 65 3 5 Lateral close ups of genital papillae that have a basal platform ......................... 66 3 6 Ventral photographs of tube papillae scored as spatulate and lateral and ventral photographs of tube papillae scored as having a terminal concavity ..... 67 3 7 Latera l close ups of genital papillae representative of posteriad downward, and anteri ad facing genital pores. ................................ ................................ ....... 68 3 8 Lateral and ventral photographs of tube papillae representative of bifurcate (A D), displaced genital pore position (A G), distally positioned genital pore (H), and villi present (A D) character states. ................................ ....................... 69 3 9 Ventral close ups of six species with tube papillae and villi. .............................. 70 3 10 The first row includes ventral photographs of a rosette papilla with a bifurcate papillar shield (PS), and a genital papilla with no bifurcation (fused). The second row displays intraspecific variation in the bifurcat ion of the papillar shield. ................................ ................................ ................................ ................. 71 3 11 Ventral photographs of four species scored as having a bifurcate papillar platform ................................ ................................ ................................ .............. 71 3 12 Max imum likelihood ancestral state reconstruction of the overall shape character using the AFLP tree (Tracy Smith, pers. comm. ). Clades composed of more than two taxa with a completely conserved overall shape state were collapsed. Branches below nodes reconst ructed below the decision threshold are colored black; otherwise they are the color of their associated character state. ................................ ................................ ................................ ................... 72 3 13 Maximum likelihood ancestral state reconstruction of the overall shape character using the two gene nuclear tree (Nick Lang, pers. comm. ). ................ 73
10 3 14 Maximum likelihood ancestral state reconstruction of the overall shape character using the three gene total ev idence tree (Near et al 2011) ................ 74 3 15 Maximum likelihood ancestral state reconstruction of the overall shape character using the five gene total evidence tree (Nick Lang, pers. comm. ).. .... 75 3 16 Ancestral state reconstructions of variation in rosette shaped genital papillae using the AFLP (Tracy Smith, pers. comm.), nDNA (Nick Lang, pers. comm.), 3TE (Near et al ., 2011), and 5TE (Nic k Lang, pers. comm.) phylogenies.. ......... 76 3 17 Ancestral state reconstruction of the presence of the basal platform structure using the three gene total evidence (3TE; Near et al ., 2011) phylogeny. ........... 77 3 18 Retention indices for the whole 14 character matrix, each character individually, and each character state of the overall shape character ................. 78
11 LIST OF ABBREVIATIONS 3TE Three gene (cytb + S7 + RAG1) total evidence phylogeny (Near et al ., 2012) 5TE Five gene (ND2 + cytb + 16S + S7 + EGR2B) total evidence phylogeny ( Nick Lang pers. comm. ) AFLP A MPLIFIED F RAGMENT L ENGTH P OLYMORPHISM PHYLOGEN Y ( SMITH ET AL 2011) AP Anal Pore BP Basal Platform GP Genital Pore GPO Genital Pore Orientation GPP Genital Pore Position MPR Most P arsimonious R econstruction MRCA Most R ecent C ommon A ncestor nDNA Nuclear gene (S7 + EGR2B) phylogeny (Nick Lang, pers. comm. ) PL Propor tional L ikelihood PP Papillar Platform PS Papillar Shield RI Retention Index
12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Master of Science COMPAR A TIVE MOR PHOLOGY AND EVOLUTION OF GENITAL PAPILLA E IN A GENUS OF DARTERS (PERCIDAE: ETHEOSTOMA ) By Zachary Pierce Martin Ma y 2013 Chair: Larry Page Major: Zoology R ecent taxa rich molecular phylogenies for percids have made robust studies on the evolu tion and diversification of darters possible Darters are well known for their diversity in reproductive traits but female genital papillae are rarely the focus of this attention. A comparative morphology study of female genital papillae in Etheostoma was conducted, including an examination of the evolution of these structures using four phylogenetic hypotheses. Fourteen characters were used to describe the variation in papilla morphology throughout Etheostoma Examining papillae, developing descriptive te rminology, and reconstructing ancestral character states led to the identification of 13 distinct papilla morphologies Based on maximum likelihood reconstructions, the basal morphology of the genital papilla in females of Etheostoma was reconstructed as a simple tub e with a distally positioned and posteriorly oriented pore. A mong the significant phylogenetic results were support for a monophyletic Catonotus and a sister group relationship between Boleosoma and Ioa Synapomorphies for Catontous were a roset te papilla and a pleated papillar platform. T he Boleosoma + Ioa clade was supported by a bifurcated papillar platform. C onspicuous villi and a pleated papillar platform separate Ioa from Boleosoma R elationship s between papilla morphology and
13 spawning stra tegies also were examined. Among the s ignificant findings were that all species that attach eggs to a rocky substrate possess a basal platform posterior to the papilla ; this structure may be required to successfully exploit this substrate for oviposition S ome morphologies are specific to oviposition strategies, and the evolution of papilla morphology may have played a major role in the diversification of Etheostoma
14 CHAPTER 1 INTRODUCTION Recent Phylogenetic Hypotheses for Etheostoma Etheostoma a genus of percid fishes with 148 species represents the largest and one of the most diverse groups of fishes in the freshwaters of North America ( Page and Burr 2011 ) The genus belongs to a well studied su bfamily, Etheostomatinae commonly known as darters The evolutionary relationships within percid fishes have been a topic of interest for ichthyologists for over a centur y. Within that time the list of recognized taxa within Etheostomatinae has grown dra matically making classification ever more difficult Further, instances of morphological convergence, hybridization and introgression have plague d the elucidation of relationships among darters (Page and Swofford 1984; Sloss et al ., 2004; Carlson & Wainwr ight 2010 ; Near et al ., 2011) Given these issues numerous phylogenetic hypotheses have been proposed for Etheostoma based on both morphological and molecular data ( Page, 1981 ; Wood & Mayden, 1997 ; Song et al ., 1998 ; Sloss et al ., 2004; Ayache and Near, 2009; Smith et al ., 2011; Near et al ., 2011; Nick Lang, pers. comm. ) Four recent hypotheses on the evolutionary relationships of darters were developed using molecular data from the genome scale (AFLP s ) down to single gene s using nuclear and mitochondrial data to build phylogenetic hypotheses ( Near et al ., 2011; Nick Lang pers. comm. ; Tracy Smith, pers. comm. ) Three of the four hypotheses approach ed a complete taxon omic sample of the subfamily ( Nick Lang pers. comm Near et al 2011). The hypothesis ba sd on AFLP data did not include as many taxa but recent efforts have increase d their sample size ( Tracy Smith pers. comm. ).
15 The nuclear, mitochondrial, and AFLP ( amplified fragment le ngth polymorphism) datasets were analyzed using neighbor joining (Smith et al 2011), maximum parsimony ( Nick Lang pers. comm. ), and Bayesian (Near et al ., 2011 ; Nick Lang, pers. comm ; Tracy Smith pers. comm. ) methods. Lang analyzed three datasets including 228 species, including 1 52 species of Etheostoma : a combined mitoch ondrial dataset (ND2 + cytb + 16S), a combined nuclear dataset (S7 + EGR2B), and a concatenated total evidence (TE) dataset ((ND2 + cytb + 16S) + (S7 + EGR2B)) ( pers. comm. ). However, he only reported the results from the combined nuclear and the TE datase ts because these trees broadly represent ed the major trends in all analyses. A second study used four datasets in their analyses sampling 245 darter taxa (Near et al. 2011) The species missing from their analyses include d Etheostoma lugoi E pottsi and E sellare They included a single gene mitochondrial dataset (cytb), two single gene nuclear datasets (S7 and RAG1), and a concatenated dataset (cytb + S7 + RAG1). Finally, phylogenetic hypotheses based on an AFLP dataset that included 69 darter species with representatives from all genera, 48 of which were species of Etheostoma have been presented (Smith et al ., 2011) All three of these studies consistently recover ed numerous subgeneric clades including Boleosoma Doration Etheostoma Fuscatelum, Lito cara Microperca Nothonotus 1 Poecilichthys Psychromaster and Vaillantia Even though these clades were consistently recovered, Fuscatelum Microperca and Vaillantia have not always been recognized in taxonomic classifications T hree other traditionall y recognized subgener a were recovered in at least one of the studies Catonotus composed of the 1 Nothonotus is a clade consistently recovered in all studies, however, it is recognized at the generic level insome studies (Near et al ., 2011)
16 barcheek, fantail, and spottail darters was recovered as monophyletic (Smith et al ., 2011; Tracy Smith, pers. comm. ) and polyphyletic (Near et al ., 2011 ; Nick Lang, pers. comm ). When the group was polyphyletic the spottail darter group was separate from barcheek and fantail darters and was given the name Stigmacerca (Near et al ., 2011). Secondly, Hololepis was composed of darters distributed through the Southe ast U.S. This clade was consistently unresolved and included members from other subgenera such as Fuscatelum the monotypic Villora and the sometimes recognized Belophlox Finally, t he large group known as Oligocephalus was recovered as monophyletic ( Near et al ., 2011; Smith et al ., 2011; Nick Lang, pers. comm. ) and polyphyletic ( Nick Lang, pers. comm. ). The E. spectabile species group and E. fragi were found outside the larger Oligocephalus clade in the Lang five gene total evidence tree ( pers. comm. ) T axonomic Classification of Darters Numerous taxonomic classifications of darters have been proposed ( Page, 1981; Bailey and Etnier, 1988; Near et al ., 2011). Two recently published classifications were particularly taxa rich including a 175 species taxono my (Page, 2000) which recogniz ed four genera and a 248 darter taxonomy (Near et al ., 2011) which recogniz ed five genera T axonomists have recognized three to five gener a when all previ ou s classification s are considered. At most, Ammocrypta Crystallaria Etheostoma Nothonotus and Percina have been recognized (Near et al ., 2011). Some studies have recognized Crystallaria as a subgenus within Ammocrypta ( Nick Lang, pers. comm. ) and Nothonotus as a subgenus within Etheostoma ( Page, 2000 ; Smith et al ., 201 1). Trait Diversity in Etheostoma Darters are small bodied, benthic fishes living in a range of hydrologic conditions from high flow riffles to lentic systems. Etheostoma exhibits an impressive diversity of
17 color patterns shape and behavior. A wide range of jaw morphologies from short and blunt to long and pointed are indicative of the diversity in feeding strategies that exist in Etheostoma ( Page 1983; Page and Swofford, 1984; Carlson and Wainwright, 2010) D iversity in body shape pore counts, and fin r ay counts were also hypothesized to correlate with ecological and behavioral traits (Page and Swofford, 1984; Guill et al ., 2003) For instance, species of Etheostoma that exploit ed high current environments had deep caudal peduncles and robust bod ies whil e species that exploit ed slower currents ha d narrow caudal pe duncles and were more fusiform. A suite of interspecifically variable traits that Etheostoma has been well known for are reprodu ctive morphology and behavior Male s may develop breeding tubercles on their ventral surface, a variety of bright nuptial color s on body and fins, and egg mimics on pelvic and dorsal fins during the spawning season (Wiley & Collette, 1970; Page, 1983; Page and Swofford, 1984; Gumm et al ., 2011). Female genital papilla e us ed for ovoposition swell in size during peak spawning season (Page, 1983) and are known to vary greatly Spawning behavior varies in the mode of ovoposition employed and the level of investment in male parental care (Page 1985; Kelly et al ., 2012). G enit al papillae and t he mode of ovoposition are of particular interest here. Genital Papilla Morphology The interspecific variation in genital papilla e has received less attention than the variation in spawning behavior. G enital papilla e are fleshy protuberan ces surrounding an apertu re through which eggs and sperm are discharged and are positioned between the anus and the anal fin In Etheostoma they are particularly variable and range from simple tub es to complex, pleated and multi lobed structures. T hey hav e been described and illustrated in systematic and life history studies (Collette, 1961; Collette and Yerger,
18 1962; Page and Burr, 1976; Page and Mayden, 1981; Page, 1983). I n several instances darter papillae have also been put into a phylogenetic contex t and general statements about shared morphologies among reproductive guilds have been made (Page, 1983; Page and Swofford, 1984). Page (1983, F igs 36 38) provided illustrations of the variation in female genital papilla e among 22 species with representati ves from three behavioral guilds Along with th ese figure s Page (1983) provide d a brief description of the structure s and briefly discussed the ir putative functional sig nificance. Papillae of e gg burying species were described as conical or tubular shape s which served to inject eggs i nto substrate. Papillae of e gg cluster ers were said to be wide and flat which serve d to push the eggs against the flat undersides of stones. Egg attaching species had papillae that were described as morphologically diverse, but most were conical or tubu lar. No indication of how these morphologies aide d in attaching eggs to substrates was given. In a later study, papillae of egg clumping papillae were described as flat much like those of egg clustering species of Etheostoma ( Page, 1985) V ariation in papillae of darter s has been used in taxonomic classification s N umerous villi in the anal urogenital area were used as a diagnostic character for the glassy darter, E ( Ioa ). vitreum (Jenkins and Burkhead 1993, p. 763). The mon o typic subgenus Villora has been diagnosed based on unique genital papilla morphology ( Hubbs and Cannon, 1935; Collette and Yerger, 1962 ). The name Villora even reflects the unique morphology with the latin villus and or meaning (Hubbs and Cannon 1935). A s suggested by past research genital papilla morphology
19 is clearly a taxonomically valuable and potentially phylogenetically informative characteristic of Etheostoma Spawning Behavior Based on field and laboratory observations, f our spawning behaviors are recognized within Etheostoma : egg burying, egg attaching, egg clumping, and egg clustering (Page 1983, Page 1985). When describing th ese behaviors, Page (1985) also hypothesized the order in which each of these behaviors evolved based on ovoposition mode and level of parental care. Egg burying was recognized in all darter genera and was suggested to be the most primitive behavior which evolved from the ancestral broadcast spawning (Page, 1985) It was characterized as the female wiggling into the substrate with her genital papilla buried followed by the male mounting her back and eggs are released just below the substrate surface whi le t he male releases sperm. N o post spawning parental care was described in egg buriers The substrates typically utilized in this mode of behavior are loose gravel, sand, a nd mixed gravel and sand (Winn 1958a,b; Page, 1983; Simon and Wallus, 2006) V ariation has been reported in the extent to which the female buries her body, ranging from burying just at the substrate surface to completely bur ying (Page and Simon, 1988; Simon and Wallus, 2006) Egg attaching was hypothesized to have evolved from egg burying a nd was characterized as the female selecting a site of egg deposition and elevating to the site along with the male (Page, 1985) The two vibrate and the female attaches several adhesive eggs to the substrate while the male releases sperm to fertilize the eggs. Substrates used are boulders, macrophytes, live and dead leaves, logs, twigs, fibrous root material, filamentous algae, moss, sides of aquariums, and artificial turf ( see Table
20 3 3 for references ). There was no post spawning parental care described for egg attaching Egg clumping is also thought to have evolved from egg burying (Page, 1985) Egg clumping males select a cavity beneath a rock where the spawning will take place and the male will guard the eggs. The female enters, is courted, and positi ons herself at the junction of the rock and gravel where she deposits her eggs in a grape like mass while the male fertilizes them. The male remains with the eggs and guards them until hatching. Lastly, egg clustering is thought to be the most derived beha vior and evolved from egg attaching (Page, 1985) Egg clustering males select a cavity under a large, flat stone where spawning will take place and the male will guard the eggs. The male clears the area of silt and debris and then the female enters the ca vity and is courted by the male They both invert to deposit and fertilize the eggs on the underside of the stone in a single layer. The substrate utilized is typically stone; however, some species are known to use logs. Interspecific variation exists with in this category, and egg clustering can be subdivided into three groupings based on the amount of time the male and the female remain inverted (Page 1985): both the male and female may briefly invert (alpha clustering) only the female may have prolonged inversion (beta clustering) or both will have prolonged inversions (gamma clustering) Reproductive Traits in a n Evolutionary Context These reproductive traits have received some attention in a phylogenetic context. The spawning behaviors and genital pa pilla morphologies were first mapped on to a phylogeny by Page (1985). Within this study, he hypothesized that egg clumping was derived within the Nothonotus clade, two forms of egg clustering emerge d in the
21 Catonotus clade, a third form of egg clustering w as derived in the Boleosoma clade, and egg attaching was the derived behavior in Vaillantia and Ioa Notably, this hypothesis suggests that egg clustering evolved multiple times in independent lineages. He also states that a spatulate female genital papill a is the derived morphology in Vaillantia and a broad, flat nonbifurcate female genital papilla is the derived morphology in Catonotus T he evolution of spawning behavior s in darters has been further investigated using a recently hypothesized phylogeny an d three behavior categories: egg burying, e gg attaching and egg guarding ( Kelly et al ., 2012). The egg guarding category combines egg clumping and egg clustering to simplify the analysis. They hypothesized that the historically recognized egg clustering p art of their egg guarding guild, evolved only once and was the ancestral state in the Goneaperca ( Catonotus + Boleosoma + Psychromaster + Stigmac erca ) clade. T his character state is secondarily lost in the entire Psychromaster clade and once in Boleosoma. No hypotheses on the evolution of reproductive morphology were presented Study Objectives The amount of variation in external reproductive morphology within Etheostoma is impressive. That variation has yet to be fully described for genital papillae in a s tandardized format. The first objective of this thesis was to conduct a comparative study of genital papilla morphology for species of Etheostoma Descriptive terminology for genital papillae was developed and phylogenetically informative characters were i dentified, both of which should be useful products in future research on evolutionary and reproductive biology of darters. G iven the abundance of existing taxon rich darter phylogenies, a second objective was to examine the evolutionary history of genital
22 papilla morphologies. T his research investigated whether the dist ribution of morphologies agree d with molecular phylogenetic hypotheses by providing evidence that corroborates consistently supported phylogenetic relationships and resolves particularly tr oublesome relationships. Phylogenetic problems of particular interest are the placement of Etheostoma ( Allohistium ) cinereum and the monophyly of Catonotus A third objective was to examine how genital papillae morphology relates to spawning behavior (i.e ., mode of ovopo sition ) in Etheostoma For instance, do all egg clusterer s share a genital papilla morphology? Is genital papilla morphology conserved in any of the four described spawning behavior guild s ? Also d oes a more precise classification of spawni ng guilds help to elucidate pattern in the morphological variation ? For instance, papilla morphology within the egg attaching guild seems to be extremely variable and breaking this guild down even further (i.e., rock attaching, macrophyte attaching, algae attaching) could reveal pattern in this variation that would not otherwise be seen.
23 CHAPTER 2 METHODS Comparative Morphology Specimens were obtained from several institutions including Auburn University Natural History Museum (AUM), Florida Museum of Natur al History (UF), Illinois Nat ural History Survey (INHS) and North Carolina State Museum of Natural Sciences (NCSM) When choosing specimens, visibly gravid females collected during peak spawning months were targeted Months of interest ranged from Februa ry to June, with the focus on March to May when most species are known to spawn (Page 1983 ; Hubbs, 1985 ). Using these constraints 12 8 species of Etheostoma along with two outgroup species, Ammocrypta vivax and Percina maculat a were examined Photograp hs were taken of genital papillae of preserved female specimens in peak spawning condition using Canon 40D and 5D cameras fitted with an Infinity Optics K2 long distance microscope on a Visionary Digital TM P 51 CamLift (Palmyra, Virginia) at UF FLMNH. The structure and variation in genital papilla morphology of 130 species, including the two outgroup species, were described using 14 discrete characters. Twelve of the characters were binary and two were multistate. Character d efiniti ons are provided througho ut the methods and in the Appendix C haracter states were unordered ; the number assigned to each state did not imply anything about whether it was an ancestral or derived state. The characters were scored in a hierarchical framework. Overall shape and basa l platform applied t o all species. The remaining twelve character states scored for each species described variation within its respective overall shape. Characters describing variation of an overall shape that a species did not possess were scored as An assumption made here was that the presence/absence of a n
24 overall shape or basal platform precluded the presence/absence of other descriptive characters These data were co mpiled in a matrix for use in phylogenetic analyses such as anc estral state reconstructions and calculating retention indices. Overall Shape (0 = Mound, 1 = Tube, 2 = Rosette, 3 = Mound Tube Intermediate) Overall shape described the broader shape profile of the genital papillae. Mound described a particularly squat papilla that was held close to the ventral surface of the body, even at peak development ( Fig. 3 1 ). Only 10 species were scored as mound, and the group did not exhibit enough variation to merit additional character coding. Notably, all mound papillae pos sessed villi associated with a posteriorly facing genital pore. Etheostoma sellare was scored as mound for lack of a better descriptor; alternatively, E. sellare could have been coded as rosette. Eighty seven species possessed genital papillae that were s cored as tube papillae. Species scored as tube had an elongate, cylindrical to conical papilla that projected from the ventral surface of the body (Fig. 3 2). Variation in tube papillae was described using six characters: spatulate, terminal concavity (TC) genital pore orientation (GPO), genital pore position (GPP), villi, and bifurcation. Substantial variation in tube length was noted but was too subjective to score as a qualitative character. Twenty six species possessed rosette papillae. Rosette papill ae (Fig. 3 3) had a suite of five defining characteristics: 1) papillar shield present, 2) papillar platform present, 3) basal platform present, 4) genital pore faced downward, 5) genital pore displaced from a distal position anteriorly towards the anus. V ariation in rosette papillae was described using four characters: villi, papillar shield bifurcation, papillar platform bifurcation, and pleats.
25 Seven species were scored as having a mound tube papilla that was intermediate between mound and tube: E. acut iceps, E. chlorobranchium E chuckwachatte E davisoni E. jordani, E rufilineatum and E tippecanoe (Fig. 3 4). Variation in this papilla shape was described using two characters: genital pore position, villi. Basal Platform (0 = Absent, 1 = Present) The presence or absence of the basal platform was scored as a binary character that applied to papillae of each overall shape (Fig. 3 5). There was variation in the size and connection of the basal platform with other papillar structures. Seventy five spe cies were scored as not having a basal platform, while 56 species were scored as having the structure. The basal platform extended laterally around the base of the papillar tissue in thirty seven species and was posterior to all rosette papillae and some t ube papillae. Spatulate (0 = Not spatulate, 1 = Spatulate) Thirty one species with a tube papilla possessed a spatulate tube. Spatulate tubes were dorsoventrally flattened, rather than cylindrical, and commonly had a large genital pore opening (Fig. 3 2A, C; Fig. 3 6A D). Spatulate was a character only applicable to species that were scored as tube in the overall shape character. Terminal Concavity (0 = Absent, 1 = Present) Thirty three species had a terminal concavity, a cup like indention at the distal en d of a tube papilla (Fig. 3 2B, D; Fig. 3 6E H). Interestingly, terminal concavity was mutually exclusive of the spatulate character. Terminal concavity was a character only applicable to species that were scored as tube in the overall shape character.
26 Ge nital Pore Orientation on Tube Papillae, GPO (0 = Downward, 1 = Anteriad, 2 = Posteriad) The direction in which the genital pore faced was given one of three scores (Fig. 3 7). Intraspecific variation was observed, particularly between the downward and pos terior facing positions, potentially as a product of preservation and date of capture. Posteriorly facing genital pores were most common (n = 74 spp), then downward facing genital pores (n = 9 spp), and least common were anteriorly facing genital pores (n = 4 spp). GPO was a character specific to tube papillae. Genital Pore Position on Tube Papillae, GPP (0 = Genital Pore is Not at Distal Most Point of Genital Papilla, 1 = Genital Pore Positioned at Distal Tip of Genital Papilla) The position of the genital pore relative to the distal tip of the papilla was scored as a binary trait. The majority of species (n = 70) had a genital pore located at the distal tip of a tube papilla. The remaining 17 species with tubular papillae had genital pores that were displa ced towards the anus or anal fin (Fig. 3 8). Most pores were displaced anteriorly. Villi on Tube Papillae (0 = Absent, 1 = Present) Eleven of 85 species of Etheostoma with tube papillae possessed villi (Fig. 3 8A C; Fig. 3 9A D). Both outgroup species poss essed tubes with villi. Villi were present in a range of sizes, positions, and numbers; from the few, small villi of E. serrifer that were present at the base of a bifurcation on the ventral side of the tube to the approximately 10 large villi of E. okaloo sae that developed from pleats and surrounded the genital pore. Five species (four Etheostoma and Percina maculata ) had pleated tissue running lengthwise along the tube that developed into villi that surrounded the genital pore.
27 Bifurcated Tube Papillae (0 = No Bifurcation, 1 = Bifurcation) Four species of Etheostoma had tube papillae with a bifurcated ventral surface of the tube or a bifurcated visor (Fig. 3 8A D). The tubes of E. fonticola E. proeliare and E. trisella were fully bifurcated distally form ing two lobes. On E. serrifer the bifurcation started at the distal tip and ran halfway down the length of the tube on the ventral surface rather than being split all the way through the tube. Villi on Rosette Papillae (0 = Absent, 1 = Present) All but one species, E. olmstedi with rosette papillae possessed villi (Fig. 3 3E H). The position and number of villi varied among species. For E. vitreum villi seemed to be the most distinctive feature of the papilla, and for the other species the villi were more of a secondary feature. Bifurcated Papillar Shield (0 = Absent, 1 = Present) The papillar shield was medially bifurcated in specimens of 15 species and was a character specific to rosette papillae (Fig. 3 10). In some instances, the papillar shield was we ll separated into two lobes (e.g. E. nigripinne ), and in others the bifurcation was less developed or partially fused (e.g. E. squamiceps ). Intraspecific variation in this character was also noted. For instance, specimens of E. crossopterum were examined w ith both well developed bifurcation of the papillar shield tissue and completely fused papillar shield tissue. Bifurcated Papillar Platform (0 = Absent, 1 = Present) The papillar platform was medially bifurcated in six species of Etheostoma and was a chara cter specific to rosette papillae (Fig. 3 3A, B; Fig. 3 11). The bifurcation in the papillar platform of E. nigrum E. perlongum and E. vitreum formed a distinctly
28 bilobed structure. Etheostoma longimanum E. olmstedi and E. podostemone had a third lobe of tissue positioned dorsally to the two lobes formed by the bifurcation. Pleated Papillar Platform (0 = Absent, 1 = Present) The papillar platform was radially pleated in 21 species of Etheostoma (Fig. 3 3C F). This character was specific to rosette papi llae. Similar to tube papillae, pleats were always present in combination with villi. The pleats were present at the base of the papillar platform and developed into villi toward the center of the papillar platform. almost completely exclusive of scoring E. vitreum which was Genital Pore Position on Mound tube Papillae (0 = Displaced from Distal Most Point of Papilla, 1 = Po sitioned at Distal Most Point of Papilla) All but one species with mound tube papillae, E. davisoni (Fig. 3 4), had a distally positioned genital pore. In E davisoni the genital pore was displaced anteriorly toward the anus. Villi on Mound tube Papillae (0 = Absent, 1 = Present) Five species with mound tube papillae were scored as having villi, E. acuticeps E. chlorobranchium E. davisoni E. rufilineatum and E. tippecanoe When present, the villi were oriented around the genital pore (Fig. 3 4A, D F). The three latter species had pleated papillar tissue that developed into villi surrounding the genital pore. Phylogenetic Data Four recently hypothesized phylogenies based on molecular data were obtained directly from researchers ( Nick Lang, pers. comm. ; Tom Near, pers. comm. ; Tracy Smith, pers. comm. ) and from supplementary data provided online (Near et al 2011,
29 http://sysbio.oxfordjournals.org ). These trees were chosen because of the large number of dart er taxa sampled within their analyses and the molecular data used to build them. Subgenera given by Page (2000) are represented in each phylogeny and three of the four phylogenies approach full taxon sampling of Etheostoma The Bayesian majority rules con sensus phylogeny built using an AFLP dataset by Tracy Smith ( pers. comm ) is referred to as the AFLP tree. The Bayesian majority rules consensus phyloge ny built with the S7 + EGR2B nuclear genes by Nick Lang ( pers. comm. ) is referred to as the nDNA tree. T he Bayesian majority rules consensus phylogeny built with the combined five gene nDNA and mtDNA dataset ((ND2 + cytb + 16S) + (S7 + EGR2B)) is the 5TE tree (five gene total evidence). The Bayesian majority rules consensus phylogeny built with the combined three gene nDNA and mtDNA dataset ((cytb) + (S7 + RAG1)) is the 3TE tree (three gene total evidence ; Near et al. 2011 ). E ach phylogeny was pruned to include only members of the genus Etheostoma that were included in this study and the two outgroup species Ammocrypta vivax and Percina maculat a, using Mesquite 2. 75 (Maddison and Maddison, 2011). Ancestral State Reconstruction The above phylogenies were each used to map the evolutionary history of the genital papilla morphologies by performing ancestral stat e re constructions in Mesquite 2. 75 An unordered maximum likelihood reconstruction of the ancestral character states was performed for each of the genital papilla characters using the character matrix (Maddison and Maddison, 2006; Maddison and Maddison, 20 11) The Mk1 model of evolution was assumed, meaning that the transition from any one character state to another was equally probable. The likelihood reconstructions yielded negative Log likehood values, which were presented as proportional likelihood valu es. A
30 recommended decision threshold of a 2.0 difference in the log likelihood values between character states was followed for determining which states existed a t each ancestral node. P arsimony reconstruction s were performed in instances where the decisio n threshold was not met to provide additional evidence on the ancestral state of the character in question. By performing the reconstructions, hypothese s about how genital papilla morphology has evolved were formed. In addition, mapping the genital papilla data onto the phylogenies provided tests of the hypothesized relationships Likewise, retention indices were calculated for the character matrix as a whole, for each of the 14 characters individually, and for each of the character states of the overall sh ape character individually Retention indices ranged from 0 to 1, and an index close to or equal to one was indicative of a character that fit s the phylogeny well and displayed little or no homoplasy. Spawning Guild Classifiation Data on commonly used depo sition substrate, position of male during spawning, position of female duri ng spawning were collected from the literature These data along with the narrative description of the spawning behavior were used to assign species to a coarse and a fine spawnin g guild when possible. Coarse spawning guilds reflect the four classically described behavioral guilds: egg burying, egg attaching, egg clumping, and egg clustering. In cases where multiple forms of spawning behavior were described for one species, the mor e derived behavior, as identified by Page (1985), was assigned to that species. S pecies were assigned to one of nine fine behavioral guilds based on more fine grained de scriptions of spawning behavior when possible. Surface burying and diving
31 were subsets of the egg burying guild. Surface burying was characterized by the female wiggling as much as half of her abdomen into the substrate with her genital papilla buried. Diving was characterized by the female choosing a site of ovoposition in sand, gravel, or cobble and plunging herself into the substrate where she was completely buried or ha d only her head, pectoral fins, caudal fin, or dorsal surface exposed. Rock attaching, macrophyte/plant attaching, and algae attaching were subsets of the egg attaching gu ild. Rock attaching was characterized as females depositing eggs on the top, sides, or crevices of a rock or boulder. Macrophyte/plant attaching was characterized as females depositing eggs on any aquatic and terrestrial plant material, both alive and dead Algae attaching was characterized as females depositing eggs in or on filamentous algae which in some instances was attached to rocks. Alpha clustering, beta clustering, and gamma clustering were subsets of the egg clustering guild. Alpha clustering was characterized as both the female and male inverting briefly (seconds) to attach eggs on the underside of a nest rock. Beta clustering was characterized as the female remaining inverted for minutes to hours during oviposition and the male inverting briefly periodically. Gamma clustering was characterized as both the female and male remaining inverted for minutes to hours during oviposition. Materials Examined Ammocrypta Ammocry pta pellucida INHS 96652,1 April 2003 (1); Ammocrypta vivax INHS 74946, 25 27 April 1948 (3). Crystallaria Crystallaria asprella INHS 86995, 22 June 1980 (3).
32 Etheostoma Etheostoma acuticeps INHS 75149, 17 June 1976 (3); Etheostoma aquali INHS 79391,14 April 1978 (2), INHS 87359 5 May 1981 (3); Etheostoma artesiae, UF 147964, 4 March 2005 (1); Etheostoma asprigene INHS 26854, 28 February 1976 (1); Etheostoma atripinne, UF 210678, 26 June 1962 ( 1 ); Etheostoma australe INHS 84082, 24 February 1979 (3); Etheostoma baileyi, INHS 37943, 14 April 1996 (2); Etheostoma barbouri, INHS 8 6917, 20 May 1980 (3); Etheostoma barrenense INHS 29495, 1 April 1993 (3), INHS 82653, 21 April 1973 (2); Etheostoma basilare INHS 37908, 12 April 1996 (3), INHS 58366, 19 April 1990 (3); Etheostoma bellum INHS 64019, 3 April 1995 (4), INHS 79286, 13 Ap ril 1978 (3); Etheostoma blennioides INHS 86489, 6 June 1901 (3), INHS 93390, 26 April 1989 (3); Etheostoma blennius INHS 77006, 14 March 1976 (2), INHS 79390, 14 April 1978 (3); Etheostoma boschungi AUM 39522, 9 March 2002 ( 6 ), INHS 79437, 16 May 1978 (3); Etheostoma brevirostrum UF 100297, 6 April 1995 ( 5 ); Etheostoma burri INHS 37973, 22 April 1996 (3); Etheostoma caeruleum INHS 38687, 15 March 1996 (3), INHS 45466, 23 April 1998 (1), INHS 100920, 21 May 1960 (3); Etheostoma camurum INHS 46514, 22 May 1998 (3), INHS 87449, 28 May 1981 (2); Etheostoma chermocki INHS 37978, 13 March 1996 (2); Etheostoma chienense INHS 60818, 18 April 1991 (3), INHS 63920, 7 April 1988 (3); Etheostoma chlorobranchium INHS 86948, 22 May 1980 (2), INHS 88207, 9 June 1983 (2); Etheostoma chlorosoma INHS 7046, 27 June 1967 (1), INHS 75917, 8 May 1964 (3); Etheostoma chuckwachatte INHS 38630 10 March 1996 (3); Etheostoma cinereum SIUC 4141, 13 March 1982 ( 5 ), UF 43964, 15 May 1970 (1); Etheostoma collettei INHS 7570 5, 18 February 1977 (3), INHS 90234, 20 March 2001 (3); Etheostoma collis INHS 27238, 2 May 1979 (3); Etheostoma colorosum INHS 62992,
33 21 May 1987 (2), INHS 78750, 27 February 1978 (2); Etheostoma coosae INHS 37962, 9 March 1996 (3), INHS 83079, March 1 965 (3); Etheostoma corona INHS 33989, 17 April 1995 (2), INHS 36294, 17 April 1995 (4); Etheostoma cragini INHS 38080, 20 February 1996 (3); Etheostoma crossopterum INHS 58154, 19 March 1990 (3), INHS 62782, 11 April 1987 (3); Etheostoma davisoni INHS 78740 27 February 1978 (1); Etheostoma derivativum INHS 91973, 21 April 2001 (4), INHS 91983, 19 April 2001 (4); Etheostoma ditrema INHS 37964 14 March 1996 (4); Etheostoma douglasi INHS 55758, 26 May 2000 (3); Etheostoma duryi INHS 64074, 8 April 1988 (1); Etheostoma etnieri INHS 82555, 29 April 1967 (3); Etheostoma etowahae UF 97360, 17 June 1994 (5); Etheostoma euzonum INHS 90433, 22 March 2001 (3); Etheostoma exile INHS 4167, 1 June 1964 (3), INHS 60848, 6 May 1991(3); Etheostoma flabellare INHS 3170, 21 May 1960 (2), INHS 62882, 13 May 1987 (3); Etheostoma flavum INHS 36035, 4 April 1995 (4), INHS 80898, 15 April 1978 (4); Etheostoma fonticola, INHS 75668, 29 May 1973 (4); Etheostoma forbesi INHS 27837, 3 April 1992 (3); Etheostoma fragi INHS 36052, 10 April 1995 (3); Etheostoma fusiforme, INHS 75703, 18 February 1977 (3); Etheostoma gracile INHS 9358, 29 March 1963 (3); Etheostoma grahami INHS 84096, 20 February 1979 (3); Etheostoma gutselli UF 27888, 9 April 1980 (1); Etheostoma hist rio INHS 37979, 16 March 1996 (2), INHS 38539, 17 February 1996 (2); Etheostoma inscriptum INHS 47070, 14 May 1998 (3); Etheostoma jessiae INHS 83904, 7 April 1978 (2); Etheostoma jordani INHS 57870 12 March 1989 (3); Etheostoma juliae INHS 79429, 2 5 April 1978 (3); Etheostoma kanawhae INHS 27279, 5 May 1979 (3); Etheostoma kantuckeense INHS 32407, 27 March 1994 (3); Etheostoma kennicotti INHS 40478, 16 April 1973 (3); Etheostoma lachneri INHS
34 64601, 16 March 1989 (1), INHS 76261, 17 April 1977 ( 3); Etheostoma lawrencei AUM 31020, 2 April 1999 ( 5 ), INHS 37837, 3 April 1996 (3); Etheostoma lepidum INHS 30202, 27 May 1993 (3), INHS 68408, 2 May 1985 (4); Etheostoma longimanum INHS 27262, 29 April 1979 (3); Etheostoma luteovinctum INHS 27829, 3 April 1992 (3), INHS 87219, 26 April 1980 (3); Etheostoma lynceum INHS 90121, 19 March 2001 (3); Etheostoma maculatum INHS 62872, 26 April 1987 (1); Etheostoma mariae INHS 46222, 2 April 1997 (3), INHS 58983, 20 June 1990 (3); Etheostoma marmorpinnum I NHS 82442, 16 March 1976 (3); Etheostoma microlepidum INHS 87338, 6 May 1981 (3); Etheostoma microperca INHS 26932, 23 May 1976 (3), INHS 75828, 18 March 1977 (3); Etheostoma moorei INHS 81085, 25 May 1978 (3); Etheostoma neopterum INHS 35693, 1 April 1995 (4); Etheostoma nianguae INHS 74395, 21 June 1965 (2); Etheostoma nigripinne INHS 61719, 23 April 1986 (3), INHS 79364, 14 April 1978 (4); Etheostoma nigrum INHS 45776, 28 April 1998 (3); Etheostoma nuchale UF 43443, 19 April 1968 ( 10 ); Etheostoma obeyense INHS 55590, 17 April 2000 (3); Etheostoma okaloosae UF 43420, 22 March 1968 ( 13 ); Etheostoma olivaceum INHS 79329, 16 April 1978 (3); Etheostoma olmstedi INHS 74147, 23 May 1977 (4), INHS 77003, 1975 or 1976 (3); Etheostoma oophylax INHS 408 62, 7 April 1997 (4); Etheostoma osburni INHS 27104, 28 April 1979 (3); Etheostoma pallididorusm INHS 80745, 28 May 1978 (3); Etheostoma parvipinne INHS 45833, 24 March 1998 (3), INHS 79334, 18 March 1978 (1); Etheostoma percnurum UF 30679, 10 April 19 81 (1); Etheostoma perlongum INHS 90221, 15 March 2001 (3); Etheostoma planasaxatile UF 176721, 13 May 2009 ( 16 ); Etheostoma podostemone INHS 27136, 30 April 1979 (3); Etheostoma pottsi INHS 75300, 14 July 1975 (3); Etheostoma proeliare INHS uncatalog ed (3), INHS
35 40835, 20 May 1976 (3); Etheostoma pyrrhogaster UF 174221, 18 April 2009 ( 6 ); Etheostoma radiosum UF 144178, 27 February 2004 (2); Etheostoma rafinesquei UF 148753, 9 March 2005 ( 3 ); Etheostoma rubrum UF 43934, 16 March 1970 (7); Etheostom a rufilineatum UF 167699, 20 April 2007 ( 22 ); Etheostoma ruprestre UF 43542, 2 May 1968 (1); Etheostoma sagitta UF 22214, 30 June 1976 (2); Etheostoma saludae UF 47151, 10 April 1986 (5); Etheostoma scotti UF 84850, 4 April 1990 (3); Etheostoma sellar e INHS 76085, 10 November 1965 (1); Etheostoma serrifer INHS 27249, 1 May 1979 (3), INHS 29679, 11 May 1993 (2); Etheostoma simoterum INHS 87281, 21 April 1977 (3); Etheostoma smithi INHS 75029, 14 March 1976 (3); Etheostoma spectabile AUM 10130, 3 Ju ne 1965 (5); Etheostoma spilotum UF 43650, 1 June 1968 (2); Etheostoma squamiceps INHS 87287, 31 March 1981 (4), INHS 91945, 18 April 2001 (4); Etheostoma striatulum INHS 79365, 14 April 1978 (3); Etheostoma swaini INHS 78623, 3 March 1978 (3); Etheost oma swannanoa INHS 63834, 24 March 1986 (3), INHS 79111, 20 March 1978 (2); Etheostoma tallapoosae INHS 87751, 6 April 1982 (2); Etheostoma tecumsehi INHS 60827, 20 April 1991 (3); Etheostoma tennesseense, UF 172008, 12 April 2008 (1); Etheostoma tetraz onum INHS 76072, 7 May 1976 (3); Etheostoma thalassinum INHS 88323, 26 June 1983 (3); Etheostoma t ippecanoe INHS 64017, 23 May 1988 (4); Etheostoma trisella INHS 78466, 24 February 1978 (1); Etheostoma tuscumbia INHS 88585, 15 June 1984 (4); Etheostom a uniporum INHS 32595, 12 April 1994 (3), INHS 36221, 9 April 1995 (3); Etheostoma variatum INHS 27152, 28 April 1979 (3), INHS 79216, 22 March 1978 (3); Etheostoma virgatum INHS 55610, 28 April 2000 (3), INHS 165368, 18 April 2006 ( 4 ); Etheostoma vitre um INHS 32566, 17 April 1994 (3), INHS 74162, 23 May 1977 (3);
36 Etheostoma vulneratum UF 50935, 17 July 1971 ( 5 ); Etheostoma whipplei INHS 87649, 29 March 1982 (3); Etheostoma zonale INHS 103276, 17 May 2010 (3); Etheostoma zonistium INHS 38652, 16 Apr il 1996 (3), INHS 82856, 28 April 1978 (3). Percina PERCINA MACULAT A UF 43517, 27 APRIL 1968 (11), UF 43523, 27 APRIL 1968 (1).
37 CHAPTER 3 RESULTS Genital Papilla Morphology The structure and variation in genital papilla morphology of 130 species, includin g the two outgroup species, were described using 14 characters. All characters were discrete, and both binary and multistate characters were included. Character definitions are provided in the appendix. The first character recorded for each species was a s imple description of the overall shape of the genital papillae. Subsequently, each species was scored for characters that describe the variation within its overall shape. Characters specific to an overall shape that a species did not possess were scored as 1. Morphological Character Evolution Likelihood ancestral character state reconstructions were performed on the four phylogenies (Figs. 3 12 through 3 17) Proportional likelihood values (PL) f or transitional nodes are presented in parentheses with the supported character state except when transitions were made terminally ( PL of 1.000 ) Character states that were reconstructed with confidence past the recommended decision threshold of 2.0 ( Mesqu ite manual, Maddison and Maddison 2011 ) are indicated by asterisks. When character states were reconstructed with PL support below the decision threshold, the most parsimonious character state was given as well. C haracter state transitions predicted by the reconstructions provide d additional support for clades hypothesized by molecular phylogenies in which the morphology was conserved but not necessarily for phylogenetic hypotheses in which the calculated r etention indices (RI) were indicative of homoplasy (Table 3 2) Retention indices also identif ied several characters of interest.
38 The ancestor at the root of each of the trees was reconstructed as having a tube papilla (PL( tube ) = 0.994 0.999*). The MRCA to Etheostoma was also reconstructed as having a tu be papilla (PL( tube ) = 0.999 1.000*). At both positions the tube was reconstructed as having a distally positioned genital pore that was oriented posteriad with all other tube characteristics absent. This simple tube morphology was the basal morphology to Etheostoma The highest retention index and best fitting morphological character matrix to the molecular phylogenies was calculated with the three gene total evidence phylogeny (RI = 0.7 84 ). When retention indices were calculated for individual character s they ranged from 0 to 1 (Fig 3 18 ) Five of 14 characters had retention indices 0.800 Also, when the overall shape states were reconstructed individually tube and rosette yielded retention indices 0.800. F ive characters had retention indices 0.900: basal platform, rosette, rosette: villi, rosette: papillar platform bifurcation, r osette: pleats. Along with overall shape these five least homoplastic characters are discussed further Overall Shape The multistate overall shape character was reconstructed on the four phylogenies with a range of state changes from nine to 13 ( Fig. 3 12 AFLP = 9 ; Fig. 3 13, nDNA = 13 ; Fig. 3 14, 3TE = 12 ; Fig. 3 15, 5TE = 12). The retention index was high est using the 3TE (RI = 0.769) Transitions from tube papillae to all other shapes were reconstructed as likely Character state transitions from mound to rosette (0 >2), mound tube to rosette (3 >2), and rosette to mound tube (2 >3) were never reconstructed on any phylogeny. AFLP phylogeny (Figure 3 12)
39 T wo major transitions from tube to rosette were reconstructed at the MRCAs to E. barbouri E. squamicep s (node 39; PL( rosette ) = 0.937*) and E. longimanum E. vitreum (node 79; PL( rosette ) = 0.969*) clades that included representatives from Catonotus and Boleosoma respectively (Fig. 3 12) The rosette morphology was thereafter conserved. An unresolved tran sition away from a tube papillae was reconstructed at the MRCA to E. bellum E. juliae (node 121; PL( mound ) = 0.405, PL( tube ) = 0.400, PL( mound tube ) = 0.192) a clade that included representatives from Nothonotus Within this clade, t he ancestor to E. aqua li + E. sanguifluum (node 130; PL( mound ) = 0.980*) was clearly resolved as having a mound papilla while six terminal transitions were also reconstructed at E. bellum (node 126; mound ), E. camurum (node 127; tube ), E. rufilineatum (node 128; mound tube), E. tippecanoe (node 133; mound tube), E. jordani (node 134; mound tube), and E. juliae (node 135; mound ) The remaining two transitions reconstructed on the AFLP tree were made terminally at E. tuscumbia (node 38; mound ) and E. davisoni (node 60; mound tube ) N uclear DNA phylogeny (Figure 3 13) Three major transitions away from tube to rosette papillae were reconstructed at the MRCAs to E. barbouri E. flabellare (node 136; PL(rosette) = 0.991*), E. chienense E. olivaceum (node 169; PL(rosette) = 0.999*), and E. longimanum E. olmstedi (node 222; PL(rosette) = 0.999*) including representatives from Catonotus Catonotus and Boleosoma + Ioa respectively A transition from tube to mound was reconstructed at the MRCA to E. pallididorsum E. tuscumbia (node 160; P L(mound) = 0.848*). Also, the PL support for a tube p apilla weakened at the MRCA to Nothonotus (node 5), and within this clade transitions to mound tube and mound were supported at
40 the MRCAs to E. acuticeps E. chuckwachatte (node 7; PL( mound tube ) = 0.983* ) and E. aquali E. maculatum (node 17; PL( mound ) = 0.999*) respectively Other transitions within the Nothonotus clade were reconstructed as terminal at E. camurum (node 12; tube), E. moorei (node 22; tube), E. rubrum (node 27; tube), E. tippecanoe (node 29; mound tube), E. jordani (node 32; mound tube), E. etowahae (node 34; tube), E. juliae (node 35; mound) due to uncertainty at deeper nodes. In addition to these terminal transitions, a transition from tube to mound tube was reconstructed at E. davisoni (node 187 ). 3TE p hylogeny (Figure 3 14) A major transition away from tube to rosette papillae was reconstructed at the MRCA to E. derivativum E. squamiceps (node 70; PL( rosette ) = 0.883*), a clade which included representatives from Boleosoma Catonotus Ioa and Psychromaster Rosette papillae were conserved in all clades except Psychromaster which was composed of species with tube ( E. boschungi and E. cragini ) and mound ( E. pallididorsum and E. tuscumbia ) papillae. The likelihood reconstruction did not resolve the character state of Psychromaster only occurred terminal ly An additional major transition away from the ancestral tube to mound tube was reconstructed at the MCRA to E. acuticeps E. tippecanoe (node 221; PL(m ound tube ) = 0.9 91*), essentially Nothonotus wit h out E. juliae The mound tube papilla was conserved in six of the and transition s to mound and tube were reconstructed and the morphologies subsequently conserved at the MRCA to E.
41 aqual i E. vulneratum (node 232; PL( mound ) = 0.993*) and at the MRCA to E. moorei E. rubrum (node 247; PL(tube) = 0.997*) respectively Other transitions in the E. acuticeps E. tippecanoe clade occurred terminally at E. etowahae (node 230; tube), E. bellum (nod e 244; mound), E. camurum (node 245; tube ), E. chlorobranchium (node 246; mound tube ), E. juliae (node 252; mound ) A final terminal transition to mound tube occurred in a clade of species with tube papillae at E. davisoni (node 135 ) 5TE p hylogeny (Figur e 3 15) In the likelihood re construction using the 5TE, a major transition away from tube papillae was reconstructed at the ancestor to the E. acuticeps E. tippecanoe clade (node 6; PL(mound tube intermediate ) = 0.9 57 *) which included representatives from Nothonotus A major transition from mound tube to mound occurred within the E. acuticeps E. tippecacnoe clade at the ancestor to the E. aquali E. vulneratum clade (node 16; PL(mound) = 0.993*) and the morphology was subsequently conserved Additional tran sitions from mound tube to mound within the E. acuticeps E. tippecanoe clade occurred terminally at E. bellum (node 29) and E. juliae (node 37) and from mound tube back to tube at the ancestor to E. moorei + E. rubrum (node 33, PL(tube) = 0.993*) and termi nally at E. etowahae (node 14) and E. camurum (node 30). A nother major transition in the 5TE tree occurred at the ancestor to the E. artesiae E. squamiceps clade (node 43, PL(rosette) = 0.667) which included subgenera Boleosoma Catonotus Ioa Oligocepha lus (in part), and Psychromaster The support for the transition to rosette at node 43 was weak, however, this transition provides a more parsimonious scenario than three independent transitions to rosette at
42 the ancestors to E. barbouri E. kennicotti (nod e 85 PL(rosette ) = 1.000*) E. longimanum E. perlongum (node 103 PL(rosette ) = 0.999*) and E. chienense E. squamiceps (node 114 PL(rosette) = 0.992* ). The rosette shape was then lost at the ancestor to the E. artesiae E. cragini clade (node 46, PL(tube ) = 0.992*), which is predominately a clade of species with tube papillae. There are two exceptions within this clade; E. pallididorsum (node 82) and E. tuscumbia (node 83) are both scored as mound and consequently terminal transitions are reconstructed at those branches. For the most part, the tube papillae w ere reconstructed as conserved throughout the remainder of the 5TE tree. The only other exception is a terminal t ransition that occurred at E. davisoni which was scored as having a mound tube papilla e in a clade of tube bearing species V ariation I n Rosette Papillae Three characters used to describe the variation in rosette papillae fit well onto the four phylogenies (Fig 3 1 6 ) and had retention indices >0.900: villi (RI = 0.913 0.958) bifurcated p apillar platform (RI = 0.957 1.000) and pleated papillar platform (RI = 0.913 0.958) The presence or absence of villi associated with rosette papillae transitioned from two to four times (AFLP = 4, 3TE = 2, 5TE = 2, nDNA = 4). The bifurcated papillar pla tform character also transitioned from two to four times (AFLP = 2, 3TE = 3, 5TE = 4, nDNA = 3). The pleated papillar platform character was consistently reconstructed as transitioning four times in all the phylogenies used. AFLP p hylogeny In the likeliho od analyses using the AFLP tree, rosette papillae with a pleated papillar platform and villi emerge d at the ancestor of the E. barbouri E. squamiceps clade (node 39; PL(villi ) = 1.000* PL(pleats ) = 0.995). R osette papillae with a bifurcated
43 papillar platf orm and villi was reconstructed at the ancestor to the E. longimanum E. vitreum clade (node 79 ; PL(villi ) = 0.999* PL(bifurcation ) = 0.957*). Two terminal character state transitions were reconstructed in the E. longimanum E. vitreum clade as E. olmstedi (node 86) lack ed villi and E. vitreum (node 87) h ad a pleated papillar platform. Nuclear DNA p hylogeny In the likelihood analyse s using the nDNA tree, rosette papillae with a pleated papillar platform and villi emerged at the ancestor s to E. barbouri E. f labellare (node 136; PL(villi) = 1.000*, PL(pleats) = 0.994*) and E. chienense E. olivaceum (node 169; PL(villi) = 1.000*, PL(pleats) = 1.000*). Rosette papillae with a bifurcated papillar platform and villi were reconstructed at the MRCA to E. longimanum E. olmstedi (node 222; PL(villi) = 1.000*, PL(bifurcation) = 1.000*). Two terminal character state transitions occurred in the E. longimanum E. olmstedi clade as E. vitreum (node 230) ha d a bifurcated papillar platform and E. olmstedi (node 231) lack ed vil li. 3TE p hylogeny In the likelihood analysis using the 3TE tree, rosette papillae with pleated papillar platform and villi were reconstructed at the ancestor to the E. derivativum E. squamiceps clade (node 70, PL(villi ) = 1.000* PL(pleats ) = 0.995*). Wit hin this clade, a bifurcated papillar platform was reconstructed at the ancestor to the E. longimanum E. vitreum (node 101; PL(bifurcation) = 0.999*) and the pleats associated with the papillar platform were lost at the ancestor to E. longimanum E. olmsted i (node 102; PL(no pleats) = 0.990*). The two terminal transitions occurred in the E. longimanum E. olmstedi clade as E. olmstedi (node 110) l acked villi and E. vitreum (node 111) gain ed a
44 pleated papillar platform. Interestingly, the rosette papilla was l ost all together to a clade of species with tube and mound shaped papillae, E. boschungi E. cragini 5TE p hylogeny It was noted above that in the likelihood reconstruction using the 5TE tree rosette papillae may have emerged once (most parsimonious, weak proportional likelihood) or three times (less parsimonious, strong proportional likelihood). When the most parsimonious scenario was assumed, rosette papillae emerged accompanied with a pleated papillar platform and villi at the MRCA to E. artesiae E. squa miceps This morphology was conserved in the E. barbouri E. kennicotti (node 85) and E. chienense E. squamiceps (node 114) clades. Rosette papillae w ere reconstructed at the MRCA to E. longimanum E. perlongum as having a bifurcated papillar platform (PL (bi furcation) = 0.996*) keeping the villi, and losing the pleats (PL(no pleats) = 0.934*). Further in the E. longimanum E. perlongum clade, two final terminal transitions were made when E. olmstedi (node 107) lost villi and E. vitreum (node 108) gained a ple ated platform. Basal P latform The basal platform character was reconstructed as a derived character in the genus Etheostoma late in the phylogeny. In all of the likelihood reconstructions, the transition from absence to presence of a basal platform took p lace from one to four times (AFLP = 4; nDNA = 4; 3TE = 1; 5TE = 2). The retention indices were relatively high when compared to the indices for other characters, and it was highest using the 3TE tree ( Figure 3 1 7 ; RI = 0.964). Major transitions to the pres ence of the basal platform were made at deep nodes in both of the total evidence trees and on the AFLP and nDNA trees the structure was reconstructed as a more of derived character in more
45 homoplastic scenarios. Losing the basal platform was rarer than gai ning the structure but losses of the basal platform were reconstructed in both the total evidence trees. 3TE p hylogeny In the 3TE tree, the basal platform emerged only once at the MRCA of E. cinereum E. squamiceps (node 8; PL(basal platform) = 0.989), a cl ade that included species from subgenera Allohistium Boleosoma Catonotus 2 Etheostoma Ioa and Psychromaster The structure was thereafter conserved in almost all taxa. The exceptions were the Psychromaster clade (node 94; PL(no basal platform) = 0.990* ) and E. swannanoa (node 24; PL(0) = 1.000) which transitioned back lacking a basal platform. 5TE p hylogeny The reconstruction using the 5TE tree suggested that the basal platform emerged twice. The basal platform was reconstructed once at that MRCA to E artesiae E flavum (node 42; PL(1) = 0.993), a clade that included species from subgenera Boleosoma Catonotus 3 Etheostoma Ioa Oligocephalus (in part) 4 and Psychromaster The structure was then reconstructed as absent starting at the MRCA to E. artesia e E. cragini (node 46; PL(0) = 0.998) which included the members of Oligocephalus (in part) and Psychromaster With the exception of a terminal loss of the structure at E. swannanoa (node 57), the basal platform was conserved in the remainder of the E art esiae E flavum clade. The second emergence of the basal platform occurred in a 2 Catonotus is paraphyletic in the three gene t otal evidence tree ( Near et al 2011). 3 Catonotus is polyphyletic in the five gene total evidence tree ( Lang et al ., 2011). 4 Oligocephalus is polyphyletic in the fi ve gene total evidence tree ( Lang et al 2011).
46 distantly related clade at E. cinereum that was purportedly related to Nothonotus and Ammocrypta in a trichotomy. AFLP and nDNA p hylogenies Using the AFLP and the nDNA phyloge nies resulted in a more homoplastic reconstruction of the evolutionary history of the basal platform where the structure emerge d in four distinct events. On the AFLP tree, the major transitions were made at the MRCAs to E. baileyi E. zonale (node 10; PL(1) = 0.998), E. barbouri E. squamiceps (node 39; PL(1) = 0.957), and E. longimanum E. vitreum (node 79, PL(1) = 0.963). A final terminal transition to possessing a basal platform was made at E. cinereum (node 136). On the nDNA tree, the major transitions wer e made at E. atripinne E. baileyi (node 78; PL(1) = 0.977), E. barbouri E. flabellare (node 136; PL(1) = 0.991), E. chienense E. olivaceum (node 169; PL(1) = 1.000), and E. longimanum E. olmstedi (node 222; PL(1) = 1.000). A single, terminal loss of the ba sal platform occurred in the nDNA tree at E. swannanoa (node 107). No loss of the basal platform occurred on the AFLP tree, however, E. swannanoa was not included in the AFLP tree (Smith et al ., 2011 Tracy Smith, pers. comm. ). Spawning Behavior Guilds A provisional spawning guild classification was built based on the previously established classification (Page, 2000) and on the variation noted in the o bservations on the spawning habits of 115 darter species The intent was to account for any consistently reported within guild variation that exists within the four classically recognized guilds It was possible to place 98 species into fine grained spawning guilds. Variation in oviposition substrate preference, female position during spawning, and male posit ion during spawning helped to define the fine grain guilds.
47 Egg B urying Guild Thirty one darters, including both outgroup species, are known to bury their eggs. The species included in the egg burying guild can be broken into two fine guilds (e.g., surfac e buriers and divers) based on how deep the female buries herself during egg deposition. Observations were insufficient for seven egg buriers to merit fine scale classification ( E. artesiae E. luteovinctum E. moorei E. osburni E. paludosum E. pulchell um and E. tetrazonum ). Surface burying Surface burying is characterized by the female wiggling as much as half of her abdomen into the substrate with her genital papilla buried. The substrates typically utilized in this mode of behavior are loose grav els, sand, and mixed gravel and sand Ammocrypta vivax Percina maculata and 14 species of Etheostoma from six subgenera were surface buriers (Layman, 1993; Muller, 2008; Page and Simon, 1988; Ruple et al ., 1984; Simmons and Layzer, 2004; Simon, 1997; Sim on and Wallus, 2006; Winn 1958a, Winn 1958b). E theostoma caeruleum is categorized as a surface burier based on the a ccount given in Winn (1958a, b); however, other observations suggest it could be a diver (Pflieger, 1978) N o species of Nothonotus were cla ssified as surface buriers. Diving Diving is an egg burying behavior in which the female choose s a n oviposition site in sand, gravel, or cobble and plunge herself into the substrate where she is completely buried or only her head, pectoral fins, caudal fi n, or dorsal surface is exposed. Ten species from four subgenera ( Litocara Nothonotus Oligocephalus and Psychromaster ) are divers ( Distler, 1972 ; Etnier and Starnes, 1993; Fisher, 1990; James and Taber,
48 1986; Mattingly et al ., 2003; Mount, 1959; Orr and Ramsey, 1990; Page and Simon, 1988; Ross and Wilkins, 1993; Scalet, 1973; Warren et al ., 1986; Widlak and Neves, 1985). Egg A ttaching Guild Fo rty six species are in the egg attaching guild including representatives from 11 subgenera : A llohistium Belophlo x Boleosoma Etheostoma Fuscatelum Hololepis Litocara Microperca Oligocephalus Psychromaster Vaillantia and Villora Egg attaching is characterized by the female selecting a n oviposition site sometimes poking at the site with her snout and then e levating to the site horizontally or inverted at which time the male will follow Fine scale egg attaching guilds are based on the commonly used site of egg deposition. Three major guilds were recognized : rock attachers, macrophyte/plant material attachers and algae attachers. It was not possible to place four of the egg attachers into fine grained guilds based on the available literature. Rock attachers Fifteen species were assigned to the rock attaching guild. With the exception of E cinereum and E vit reum all of the rock attachers are members of the subgenus Etheostoma Furthermore, they are all members of the subgenus Ulocentra recognized in some taxonomic classification s (Near et al ., 2011) The placement of E. vitreum into the rock attaching guild is debatable because the spawning event (Winn and Picciolo, 1960) observed for the glassy darter was in an unnatural setting, just below a concrete structure.
49 Macrophyte/Plant mate rial attachers Twenty darters from eight subgenera were assigned as macrophyte/plant material attachers. Members of Etheostoma Fuscatelum Hololepis Microperca Oligocephalus Psychromaster Vaillantia and Villora were included. Commonly reported ovipos ition substrates included aquatic and terrestrial plant material both dead and alive. The structural dissimilarity of some of these substrates has been noted. Attaching eggs to fibrous root material (e.g., E exile ) could be different from attaching eggs to rush ( Juncus spp) stalks (e.g., E. boschungi ). Algae attachers Seven members of Belophlox Etheostoma Microperca and Oligocephalus were classified as algae attachers. All of the algae attachers from subgenus Etheostoma ( E. blennioides E. histrio and E. zonale ) deposit eggs on algae attached to rocks and the former two species are known to attach eggs on moss too (Pflieger 1978; Steinberg et al 2000; Walters 1994; Winn 1958a; Winn 1958b). Egg Clumping Guild Egg clumping was reported in eight species: E aquali E. camurum E. chlorobranchium, E maculatum E microlepidum E sanguifluum E vulneratum and E wapiti Two additional species that have been noted as possible egg clumpers are E camurum and E chlorobranchium ( Mount, 1959; Page and Simon 1988; Simon and Wallus 2006). However, these observations do not fit what was initially described as egg clumping (Page, 1985). Specifically, it is unclear whether the males of these two species establish territories and guard eggs afterwards. The eggs buried by E camurum like t (1959) and Page and Simon (1988), so at the least this species may
50 display an intermediate behavior from egg burying to egg clumping. In the pres ent study, E. camurum and E. chlorobranchium are scored as egg clumpers Egg Clustering Guild Twenty nine species are known to fall within the egg clustering guild. Page (1985, 2000) classified 22 of the known egg clusterers into three fine guilds based on the amount of time the female and male remain inverted during spawning events. The se were referred to as alpha clustering, beta clustering, and gamma clustering (Page, 1985). Five of the egg clusterering darters ( E. derivativum E. lemniscatum E. longima num E. podostemone E. sitikuense and E. susanae ) were not placed into a fine guild in the present study because the account provided in the literature was too vague to substantiate classification. E. longimanum and E. podostemone are classified as gamma clusterers ( Page, 2000); however, the accounts given as references for these species (Jenkins, 1980; Page et al 1981) only note that single layered cluster nests were found on the underside of rocks guarded by males. Alpha c lustering Eleven species were classified as alpha clusterers. The assignment of E. perlongum is debatable, but it has been clas sified as a gamma clusterer previously (Page, 2000) E. perlongum was observed spawning in a laboratory four tim es ( Lindquist et al 1981). Time of inversion for the female ranged from 0.5 51 seconds and for the male ranged from 0.5 14 seconds. Other alpha clusterers stay within the seconds range too, whereas, gamma clusterers will stay inverted fo r minutes to hours at a time.
51 Beta c lustering Eleven specie s are confirmed beta clusterers. Beta clusterer ers are species in which the female invert s to deposit eggs and stay s inverted for long periods of time (minutes to hours) while the male only briefly inverts periodically to fe rtilize eggs. Gamma c lustering Just two species, Etheostoma nigrum and E. olmstedi were confirmed in the present study as gamma clusterers. As mention ed above, t wo other darters were classified as gamma clusterers (Page, 2000) E. longimanum and E. podostemone based on information in previous publications (Jenkins, 1980; Page et al ., 1981). These authors determined both species were egg clusterers by finding eggs clustered on the underside of nest rocks guarded by males. Information on the act of egg deposition was unavailable so these two species remain unclassified at the fine guild level.
52 Table 3 1. Character coding for the 14 characters describing variation in genital papillae morphology in Etheostoma and two outgroup species, Ammocrypta vivax and Percina maculata es a character as inapplicable to a species due to the overall shape of its genital papilla. Tube Rosette Mound tube Species Overall Shape Basal platform Spatulate TC GPO GPP Villi Bifurcate Villi PS bifurcate PP bifurcate PP pleats GPP Villi A. vivax 1 0 0 0 2 1 1 0 x x x x x x E. acuticeps 3 0 x x x x x x x x x x 1 1 E. aquali 0 0 x x x x x x x x x x x x E. artesiae 1 0 1 0 2 0 0 0 x x x x x x E. asprigene 1 0 1 0 2 1 0 0 x x x x x x E. atripinne 1 1 0 1 2 1 0 0 x x x x x x E. australe 1 0 0 0 2 1 1 0 x x x x x x E. baileyi 1 1 0 1 2 1 0 0 x x x x x x E. barbouri 2 1 x x x x x x 1 1 0 1 x x E. barrenense 1 1 0 1 2 1 0 0 x x x x x x E. basilare 2 1 x x x x x x 1 1 0 1 x x E. bellator 1 1 0 1 2 1 0 0 x x x x x x E. bellum 0 0 x x x x x x x x x x x x E. bison 1 0 1 0 2 1 0 0 x x x x x x E. blennioides 1 1 0 1 2 1 0 0 x x x x x x E. blennius 1 1 0 1 2 1 0 0 x x x x x x E. boschungi 1 0 1 0 2 1 0 0 x x x x x x E. brevirostrum 1 1 0 1 2 1 0 0 x x x x x x E. burri 1 0 1 0 2 1 0 0 x x x x x x E. caeruleum 1 0 1 0 2 1 0 0 x x x x x x E. camurum 1 0 0 0 2 1 1 0 x x x x x x E. chermocki 1 1 0 1 2 1 0 0 x x x x x x E. chienense 2 1 x x x x x x 1 1 0 1 x x E. chlorobranchium 3 0 x x x x x x x x x x 1 1 E. chlorosoma 1 0 0 0 1 0 1 0 x x x x x x E. chuckwachatte 3 0 x x x x x x x x x x 1 0 E. cinereum 1 1 0 0 2 1 0 0 x x x x x x E. collettei 1 0 1 0 2 1 0 0 x x x x x x E. collis 1 0 0 0 0 0 0 0 x x x x x x E. colorosum 1 1 0 1 2 0 0 0 x x x x x x E. coosae 1 1 0 1 2 1 0 0 x x x x x x E. c orona 2 1 x x x x x x 1 1 0 1 x x E. cragini 1 0 1 0 2 1 0 0 x x x x x x E. crossopterum 2 1 x x x x x x 1 1 0 1 x x
53 Table 3 1. Continued Tube Rosette Mound tube Species Overall Shape Basal platform Spatulate TC GPO GPP Villi Bifurcate Villi PS bifu rcate PP bifurcate PP pleats GPP Villi E. davisoni 3 0 x x x x x x x x x x 0 1 E. derivativum 2 1 x x x x x x 1 1 0 1 x x E. ditrema 1 0 1 0 2 0 0 0 x x x x x x E. duryi 1 1 0 1 2 1 0 0 x x x x x x E. edwini 1 0 0 0 1 0 1 0 x x x x x x E. etnieri 1 1 0 1 2 1 0 0 x x x x x x E. etowahae 1 0 1 0 2 1 0 0 x x x x x x E. euzonum 1 0 0 0 2 1 0 0 x x x x x x E. exile 1 0 1 0 2 1 0 0 x x x x x x E. flabellare 2 1 x x x x x x 1 0 0 1 x x E. flavum 1 1 0 1 2 1 0 0 x x x x x x E. fonticola 1 0 0 0 2 0 1 1 x x x x x x E. forbesi 2 1 x x x x x x 1 1 0 1 x x E. fragi 1 0 1 0 2 1 0 0 x x x x x x E. fricksium 1 0 1 0 2 1 0 0 x x x x x x E. fusiforme 1 0 0 0 0 0 0 0 x x x x x x E. gracile 1 0 0 0 1 0 0 0 x x x x x x E. grahami 1 0 0 0 2 1 1 0 x x x x x x E gutselli 1 1 0 1 2 1 0 0 x x x x x x E. histrio 1 1 1 0 2 0 0 0 x x x x x x E. hopkinsi 1 0 1 0 2 1 0 0 x x x x x x E. inscriptum 1 1 0 1 2 1 0 0 x x x x x x E. jessiae 1 0 0 1 2 1 0 0 x x x x x x E. jordani 3 0 x x x x x x x x x x 1 0 E. juliae 0 0 x x x x x x x x x x x x E. kanawhae 1 0 1 0 2 1 0 0 x x x x x x E. kantuckeense 1 0 1 0 2 1 0 0 x x x x x x E. kennicotti 2 1 x x x x x x 1 1 0 1 x x E. lachneri 1 1 0 1 2 1 0 0 x x x x x x E. lawrencei 1 0 1 0 2 1 0 0 x x x x x x E. lepidum 1 0 1 0 2 1 0 0 x x x x x x E. longimanum 2 1 x x x x x x 1 0 1 0 x x E. luteovinctum 1 0 1 0 2 1 0 0 x x x x x x E. lynceum 1 1 0 1 2 0 0 0 x x x x x x E. maculatum 0 0 x x x x x x x x x x x x E. mariae 1 0 0 1 2 1 0 0 x x x x x x E. marmorpinnum 2 1 x x x x x x 1 1 0 1 x x E. microlepidum 0 0 x x x x x x x x x x x x E. microperca 1 0 0 0 0 1 1 0 x x x x x x
54 Table 3 1. Continued Tube Rosette Mound tube Species Overall Shape Basal platform Spatulate TC GPO GPP Villi Bifurcate Villi PS bifurcate PP b ifurcate PP pleats GPP Villi E. moorei 1 0 0 0 2 1 1 0 x x x x x x E. neopterum 2 1 x x x x x x 1 0 0 1 x x E. nigripinne 2 1 x x x x x x 1 1 0 1 x x E. nigrum 2 1 x x x x x x 1 0 1 0 x x E. nuchale 1 0 1 0 2 0 0 0 x x x x x x E. obeyense 2 1 x x x x x x 1 1 0 1 x x E. occidentale 1 1 1 0 2 1 0 0 x x x x x x E. okaloosae 1 0 0 0 0 1 1 0 x x x x x x E. olivaceum 2 1 x x x x x x 1 0 0 1 x x E. olmstedi 2 1 x x x x x x 0 0 1 0 x x E. oophylax 2 0 x x x x x x 1 0 0 1 x x E. osburni 1 0 0 0 2 1 0 0 x x x x x x E. pallididorsum 0 0 x x x x x x x x x x x x E. parvipinne 1 0 0 1 0 1 0 0 x x x x x x E. percnurum 2 1 x x x x x x 1 0 0 1 x x E. perlongum 2 1 x x x x x x 1 0 1 0 x x E. planasaxatile 1 1 0 1 2 1 0 0 x x x x x x E. pod 0 st e mone 2 1 x x x x x x 1 0 1 0 x x E. pottsi 1 0 1 0 2 1 0 0 x x x x x x E. proeliare 1 0 0 0 2 0 1 1 x x x x x x E. pyrrhogaster 1 1 0 1 2 1 0 0 x x x x x x E. radiosum 1 0 1 0 2 1 0 0 x x x x x x E. rafinesquei 1 1 0 1 2 1 0 0 x x x x x x E. rubrum 1 0 0 0 2 1 0 0 x x x x x x E. rufilineatum 3 0 x x x x x x x x x x 1 1 E. rupestre 1 1 0 1 2 0 0 0 x x x x x x E. sagitta 1 0 0 0 2 1 0 0 x x x x x x E. saludae 1 0 0 0 1 0 0 0 x x x x x x E. sanguifluum 0 0 x x x x x x x x x x x x E. scotti 1 1 0 1 0 1 0 0 x x x x x x E. sellare 0 1 x x x x x x 1 0 0 0 x x E. serrifer 1 0 1 0 0 0 1 1 x x x x x x E. smithi 2 1 x x x x x x 1 1 0 1 x x E. spectabile 1 0 1 0 2 1 0 0 x x x x x x E. spilotum 1 0 0 1 2 1 0 0 x x x x x x E. squamiceps 2 1 x x x x x x 1 1 0 1 x x E. stigmaeum 1 0 0 1 2 1 0 0 x x x x x x E. striatulum 2 1 x x x x x x 1 1 0 1 x x
55 Table 3 1. Continued Tube Rosette Mound tube Species Overall Shape Basal platform Spatulate TC GPO GPP Villi Bifurcate Villi PS bifurcate PP bifurcate PP pleats GPP Vill i E. swaini 1 0 1 0 2 1 0 0 x x x x x x E. swannanoa 1 0 0 1 2 1 0 0 x x x x x x E. tallapoosae 1 1 0 1 2 1 0 0 x x x x x x E. tecumsehi 1 0 1 0 2 1 0 0 x x x x x x E. tennesseense 1 1 0 1 2 1 0 0 x x x X x x E. tetrazonum 1 0 0 0 2 1 0 0 x x x X x x E. thalassinum 1 1 0 1 2 1 0 0 x x x X x x E. tippecanoe 3 0 x x x x x x x x x X 1 1 E. trisella 1 0 1 0 0 0 0 1 x x x X x x E. tuscumbia 0 0 x x x x x x x x x X x x E. uniporum 1 0 1 0 2 1 0 0 x x x X x x E. variatum 1 0 0 0 2 1 0 0 x x x X x x E. virgatum 2 1 x x x x x x 1 1 0 1 x x E. vitreum 2 1 x x x x x x 1 0 1 1 x x E. vulneratum 0 0 x x x x x x x x x X x x E. whipplei 1 0 1 0 2 1 0 0 x x x X x x E. zonale 1 1 0 1 2 1 0 0 x x x X x x E. zonistium 1 1 0 1 0 1 0 0 x x x X x x Percina macu lata 1 0 0 0 2 1 1 0 x x x X x x
56 Table 3 2. Retention indices obtained when applying each character to the four phylogenies. The scale for retention index is 0 to 1 and an index approaching 1 represents a character that fits extremely well on a phylogen y, displaying little homoplasy across the tree. Characters Retention Indices on Molecular Phylogenies AFLP 3TE 5TE nDNA Whole Matrix 0.735 0.784 0.758 0.744 Overall Shape 0.714 0.769 0.763 0.737 Basal Platform 0.893 0.964 0.944 0.926 Mound 0.250 0.5 00 0.500 0.375 Mound tube 0.000 0.333 0.167 0.333 GPP 0.000 0.400 0.200 0.400 Villi 0.000 0.200 0.200 0.400 Rosette 0.929 0.960 0.958 0.917 Villi 0.923 0.958 0.957 0.913 PS Bifurcation 0.769 0.750 0.783 0.739 PP Bifurcation 1.000 0.95 8 0.957 0.957 PP Pleats 0.923 0.958 0.913 0.913 Tube 0.783 0.829 0.800 0.775 Spatulate 0.750 0.786 0.757 0.743 Terminal Concavity 0.763 0.808 0.806 0.778 GPD 0.645 0.712 0.667 0.647 GPP 0.613 0.667 0.643 0.607 Villi 0.656 0.717 0.6 73 0.692 Bifurcation 0.750 0.795 0.767 0.744
57 Table 3 3 A behavioral guild classification of 152 taxa of Etheostoma along with Ammocrypta vivax and Percina maculata based on mode of egg deposition Spawning Behavior Guild Reference Egg buriers (n = 34) Unspecified Fine Guild E. ( Nothonotus ) moorei Orr and Ramsey, 1990 E ( Oligocephalus ) artesiae Kelly et al ., 2012* E ( Oligocephalus ) luteovinctum Johnston and Johnson, 2000 E. ( Oligocephalus ) spectabile paludosum M endelson 2003; Kelly et al ., 2012 E. ( Oligocephalus ) spectabile pulchellum Mendelson, 2003; Kelly et al ., 2012 E ( Poecilichthys ) osburni Katula, 1991 E. ( Poecilichthys ) tetrazonum Pflieger, 1978; Page, 1983 Surface Buriers Ammocrypta vivax Simon and Wallus, 2006; Muller, 2008 E. ( Belophlox ) fricksium Layman, 1993 E ( Doration ) akatulo Simmons and Layzer, 2004 E. ( Doration ) jessiae Simon, 1997 E. ( Doration ) stigmaeum Simon, 1997 E. ( Litocara ) sag itta Page and Simon, 1988 E. ( Oligocephalus ) caeruleum Winn, 1958a, b** E. ( Oligocephalus ) collettei Katula, 1991 E ( Oligocephalus ) spectabile Winn, 1958a, b** E ( Oligocephalus ) swaini Ruple et al ., 1984*** E ( Poecilichthys ) variatum Simon and Wallus, 2006 E. ( Psychromaster ) punctulatum Page and Simon, 1988 E. ( Psychromaster ) tuscumbia Simon and Wallus, 2006 Percina ( Alvordius ) maculata Divers E. ( Litocara ) nianguae Mattingly et al ., 2003 E. ( Nothonotus ) acuticeps Etnier and Starnes, 1993 E. ( Nothonotus ) bellum Fisher, 1990 E. ( Nothonotus ) jordani Orr and Ramsey, 1990 E. ( Nothonotus ) juliae James and Taber, 1986 E. ( Nothonotus ) rubrum Ross and Wilkins, 1993 E. ( Nothonotus ) rufilineatum Page, 1983; Widlak and Neves, 1985** E. ( Nothonotus ) tippecanoe Warren et al ., 1986 E. ( Oligocephalus ) radiosum cyanorum Scalet, 1973 E. ( Psychromaster ) cragini Distler, 1972
58 Table 3 3. Continued Spawning Be havior Guild Reference Egg Attaching Behaviors (n = 46) Unspecified Fine Guild E. ( Etheostoma ) bellator Kelly et al ., 2012* E. ( Fuscatelum ) phytophilum Kelly et al ., 2012* E. ( Hololepis ) serrifer Kelly et al ., 2012* E. ( Litoc ara ) spilotum Kelly et al ., 2012* Rock Attachers E. ( Allohistium ) cinereum Jenkins and Burkhead, 1993 E. ( Boleosoma/Ioa ) vitreum Winn and Picciolo, 1960 E. ( Etheostoma ) atripinne Page, 1983 E. ( Etheostoma ) baileyi Porterfie ld, 1998; Steinberg et al ., 2000 E. ( Etheostoma ) barrenense Winn 1958a, b; Steinberg et al ., 2000 E. ( Etheostoma ) brevirostrum Johnston and Shute, 1997; Steinberg et al ., 2000 E. ( Etheostoma ) colorosum Johnston et al ., 1999 E. ( Ethe ostoma ) coosae Johnston and Shute, 1997; Steinberg et al ., 2000 E. ( Etheostoma ) duryi Page et al ., 1982; Porterfield, 1997 E. ( Etheostoma ) etnieri Porterfield, 1998 E. ( Etheostoma ) flavum Keevin et al ., 1989; Porterfield, 1997 E. ( E theostoma ) rafinesquei Carney and Burr, 1989; Porterfield, 1997 E. ( Etheostoma ) scotti Bauer et al ., 1995 E. ( Etheostoma ) simoterum Page and Mayden, 1981; Porterfield, 1997 E. ( Etheostoma ) tallapoosae Johnston et al ., 1999 Macr ophy te/Plant Material Attachers E ( Etheostoma ) pyrrhogaster Carney and Burr, 1989 E. ( Etheostoma ) raneyi Johnston and Haag, 1996 E ( Etheostoma ) swannanoa Simon and Wallus, 2006** E ( Etheostoma ) zonistium Carney and Burr, 1989 E. ( Fuscatelum ) parvipinne Johnston, 1994 E. ( Hololepis ) collis Burkhead and Jenkins, 1991; Kelly et al ., 2012 E ( Hololepis ) fusiforme Fletcher, 1976 E ( Hololepis ) gracile Braasch and Smith, 1967 E. ( Microperca ) microperca Winn, 1 958a, b E. ( Oligocephalus ) asprigene Page et al ., 1982; Cummings et al ., 1984 E. ( Oligocephalus ) ditrema Page et al ., 1982 E ( Oligocephalus ) exile Winn, 1958a, b
59 Table 3 3. Continued Spawning Behavior Guild Reference Egg Attaching Beh aviors Macrophyte/Plant Material Attachers (continued) E. ( Oligocephalus ) grahami Strawn, 1955 E. ( Oligocephalus ) nuchale Duncan et al ., 2010 E ( Psychromaster ) boschungi Boschung, 1986 E. ( Psychromaster ) pallididorsum Johnsto n, 1995 E ( Psychromaster ) trisella Page, 1985 E ( Vaillantia ) chlorosomum Page et al ., 1982 E ( Vaillantia ) davisoni Bart, Jr., 1992 E ( Villora ) edwini Page, 1983; Page, 1985 Algae Attachers E ( Belophlox ) okaloosae Collette and Yerger, 1962 E. ( Etheostoma ) blennioides Fahy, 1954; Winn, 1958a, b; Pflieger, 1978 E ( Etheostoma ) histrio Steinberg et al ., 2000 E ( Etheostoma ) zonale Pflieger, 1978; Walters, 1994 E ( Microperca ) fonticola Strawn, 1 956 E ( Microperca ) proeliare Page and Burr, 1978 E ( Oligocephalus ) lepidum Strawn, 1956 Egg Clumpers (n = 7) E ( Nothonotus ) aquali Page et al ., 1982 E. ( Nothonotus ) camurum Mount, 1959; Page and Simon, 1988 E. ( Noth onotus ) chlorobranchium Simon and Wallus, 2006 E ( Nothonotus ) maculatum Raney and Lachner, 1939; Winn, 1958a, b E ( Nothonotus ) microlepidum Page et al ., 1982 E ( Nothonotus ) sanguifluum Rakes et al ., 1999 E ( Nothonotus ) vulneratu m Etnier and Starnes, 1993 E. ( Nothonotus ) wapiti Rakes et al ., 1999 Egg Clustering Behaviors (n = 29) Unspecified Fine Guild E ( Boleosoma ) longimanum Page et al ., 1981; Jenkins and Burkhead, 1993 E ( Boleosoma ) podestomone E. ( Boleosoma ) susanae Kelly et al 2012* E. ( Catonotus ) lemniscatum Eisenhour and Burr, 2000 E. ( Catonotus ) sitikuense Kelly et al ., 2012*
60 Table 3 3. Continued Spawning Behavior Guild Reference Egg Clustering Behaviors (continued) Alpha Clusterer E ( Boleosoma ) perlongum Lindquist et al ., 1981 E ( Catonotus ) chienense Piller et al ., 1999 E ( Catonotus ) corona Page et al ., 1992; Page, 2000 E ( Catonotus ) crossopterum Page et al ., 1992; P age, 2000 E ( Catonotus ) forbesi Page et al ., 1992; Page, 2000 E ( Catonotus ) neopterum Page and Mayden, 1979; Page et al ., 1992; Page, 2000 E ( Catonotus ) nigripinne Page et al ., 1992; Page, 2000 E ( Catonotus ) olivaceum Page, 198 0 E ( Catonotus ) oophylax Page et al ., 1992; Page, 2000 E. ( Catonotus ) pseudovulatum Page et al ., 1992; Page, 2000 E ( Catonotus ) squamiceps Page, 1974 Beta Clusterer E ( Catonotus ) barbouri Page et al al 1992 E ( Catonotus ) basilare Page, 1983; Page et al ., 1992; Page, 2000 E ( Catonotus ) derivativum Page, 1983; Page et al ., 1992; Page, 2000 E ( Catonotus ) flabellare Lake, 1936 E ( Catonotus ) kennicotti Page, 1975; Page, 1976 E ( C atonotus ) marmorpinnum Layman, 1984; Layman, 1991 E ( Catonotus ) obeyense Page et al ., 1981; Page et al., 1992 E ( Catonotus ) percnurum Layman, 1984; Eisenhour and Burr, 2000 E ( Catonotus ) smithi Page and Burr, 1976 E ( Catonotus ) striatulum Page, 1980; Page et al., 1992; Page, 2000 E ( Catonotus ) vigratum Page et al ., 1992 Gamma Clusterer E ( Boleosoma ) nigrum Winn, 1958a, b E ( Boleosoma ) olmstedi Constantz, 1979 No Behavior Documented (n = 42) E. ( Belophlox ) mariae E. ( Catonotus ) brevispinum E. ( Catonotus ) humerale E ( Etheostoma ) blennius E ( Etheostoma ) cervus E ( Etheostoma ) chermocki E ( Etheostoma ) gutselli E ( Etheostoma ) inscriptum
61 Table 3 3. Continued Spawning Behavior Guild Reference No Behavior Documented (continued) E ( Etheostoma ) lachneri E ( Etheostoma ) lynceum E ( Etheostoma ) occidentale E ( Etheostoma ) orientale E ( Etheostoma ) plan asaxatile E. ( Etheostoma ) ramseyi E ( Etheostoma ) rupestre E ( Etheostoma ) tennesseense E ( Etheostoma ) thalassinum E. ( Hololepis ) saludae E. ( Hololepis ) zonifer E. ( Nothonotus ) chuckwachatte E. ( Nothonotus ) denoncourti E. ( Nothonotus ) douglasi E. ( Nothonotus ) etowahae E. ( Oligocephalus ) australe E. ( Oligocephalus ) bison E. ( Oligocephalus ) burri E. ( Oligocephalus ) fragi E. ( Oligocephalus ) hopkinsi E. ( Oligocephalus ) kantuckeense E. ( Oligocephalus ) lawrencei E. ( Oligocephalus ) lugoi E. ( Oligocephalus ) pottsi E. ( Oligocephalus ) segrex E. ( Oligocephalus ) tecumsehi E. ( Oligocephalus ) uni porum E. ( Oligocephalus ) whipplei E ( Poecilichthys ) erythrozonum E ( Poecilichthys ) euzonum E ( Poecilichthys ) kanawhae E. ( Psychromaster ) autumnale E. ( Psychromaster ) mihileze Etheostoma sellare = Kelly et al (2012) source was Pat Rakes, personal communication **= Another source suggests alternative strategy (Pflieger, 1978 for E. caeruleum and E. spectabile being divers; Page and Simon, 1988 for E. camurum being a surface burier; Simon and Wallu s, 2006 for E. rufilineatum being a clumper; Jenkins and Burkhead, 1993 for E. swannanoa being a burier) ***= Description of spawning was close to diving but not definite, scored as surface burier ngle layer of eggs and guarding male
62 Fig ure 3 1 V entral views of e ight species with a mound papilla : A) E. tuscumbia B) E. sanguifluum C) E. pallididorsum D) E. juliae E) E. maculatum F) E. microlepidum G) E. aquali H) E. bellum
63 Figure 3 2 La teral views of eight species with tub e papillae : A) E. burri B) E. coosae C) E. inscriptum D), E. mariae E), E. microperca F) E. osburni G) E. saludae and H). E. swannanoa
64 Fig ure 3 3 Lateral and ventral view s of s pecies with rosette papillae : A and B ) Etheostoma olmstedi C and D ) E. flabellare E and F ) E. crossopterum and G and H ) E. vitreum. A nal pore (AP), papillar shield (PS), genital pore (GP), papillar platform (PP), and basal platform (BP) are labeled on the ventral view s
65 Fig ure 3 4 Ven tral views of six species with intermediate mound tube papillae : A) E. acuticeps, B) E. tippecanoe, C) E. jordani, D) E. davisoni E) E. chlorobranchium, F) E. rufilineatum.
66 Fig ure 3 5. Lateral views of species with a basal platform : A) Etheostoma cher mocki B) E. cinereum C) E. flavum D) E. lachneri E) E. nigripinne F) E. striatulum
67 Fig ure 3 6 Ventral views of tub e papillae scored as spatulate : A) E. artesiae B) E. bison C) E. etowahae D) E. collettei L ateral and ventral views of tube papil lae with a terminal concavity : E) E. flavum F) E. brevirostrum G) E. etnieri, and H) E. zonistium.
68 Fig ure 3 7 Lateral close E. blennius E. scotti ), E. g racile ) facing genital pores.
69 Fig ure 3 8 Lateral and ventral views of tube papillae representative of the character states bifurcate ( A D ) displaced genital pore position (A G), distally positioned genital pore (H), and villi present (A D) A) Etheosto ma fonticola B) E. proeliare C) E. serrifer D) E. trisella E) E. collis F) E. fusiforme G) E. histrio and H) E. jessiae G enital pore (GP) is labeled in each photograph.
70 Fig ure 3 9 Ventral views of six species with tube papillae and villi : A) E. australe B) E. grahami C) E. microperca D) E. edwini E) P. maculata and F) E. okaloosae.
71 Fig ure 3 10 G enital papilla with a bifurcate papillar shield (PS) : A) Etheostoma basilare and a genital papilla with no bifurcation (fused) : B) E. oophylax I ntraspecific variation in the bifurcation of the papillar shield in E. crossopterum (C and D). Fig ure 3 1 1 Ventral views of four species with a bifurcate papillar platform : A) E. longimanum and B) E. perlongum C) E. podostemone and D) E. vitr eum
72 F igure 3 12 Maximum likelihood ancestral state reconstruction of the overall shape character using the AFLP tree Clades composed of three or more taxa with a completely conserved overall shape state were collapsed. Branches below nodes reconstructed below the decision threshold are colored black; otherwise they are the color of their associated character state.
73 Figure 3 13. Maximum likelihood ancestral state reconstruction of the overall shape character using the two gene nuclear tree. Clades composed of three or more taxa with a completely conserved overall shape were collapsed. Branches below nodes reconstructed below the decision threshold are colored black; otherwise they are the color of their associated character state.
74 Figure 3 14 Maximum like lihood ancestral state reconstruction of the overall shape character using the three gene total evidence tree Clades composed of three or more taxa with a completely conserved overall shape state were collapsed. Branches below nodes reconstructed below th e decision threshold are colored black; otherwise they are the color of their associated character state.
75 Figure 3 15. Maximum likelihood ancestral state reconstruction of the overall shape character using the five gene total evidence tree Clades compos ed of three or more taxa with a completely conserved overall shape state were collapsed. Branches below nodes reconstructed below the decision threshold are colored black; otherwise they are the color of their associated character state.
76 Figure 3 16 An cestral state reconstructions of variation in rosette genital papillae using the AFLP, nDNA, 3TE, and 5TE phylogenies. Branches below nodes reconstructed below the decision threshold are colored black; otherwise they are the color of their associated chara cter state. Variation in the rosette papilla is indic ated with the orange, green and violet bars along branches and with symbols inside nodes Darters in Catonotus clade 1 are the barcheek + fantail darters and in darters in Catonotus clade 2 are the spott ail darters.
77 Figure 3 17 Ancestral state reconstruction of the presence of the basal platform using the three gene total evidence (3TE; Near et al ., 2011) phylogeny.
78 Figure 3 18. Retention indices for the whole 14 character matrix, each character indi vidually, and each character state of the overall shape characte r
79 CHAPTER 4 DISCUSSION Fourteen characters were used to describe the variation in morphology of genital papillae of females throughout Etheostoma Examining genital papill ae, developing descri ptive terminology and reconstructing ancestral character states revealed 13 distinct papillae morphologies and three synapomorphies. Based on maximum likelihood reconstructions, the basal morphology of the genital papilla in females of Etheostoma was reco nstructed as a simple tube with a distally positioned and posteriorly oriented pore. Among the significant phylogenetic results were support for a monophyletic Catonotus and a sister group relationship between Boleosoma and Ioa Synapomorphies for Catontou s were a rosette papilla and a pleated papillar platform. The Boleosoma + Ioa clade was supported by a bifurcated papillar platform. Conspicuous villi and a pleated papillar platform separate Ioa from Boleosoma Also, the r econstruction analyses suggest ed that the diversification of oviposition site choice has been accompanied by adaptations to genital papilla morphology. Comparative Genital Papillae Morphology Genital papilla morphologies in Etheostoma were compared by developing qualitative terminology that described their general shape, the detail in tissue such as any folding and projections and the position and orientation of the genital pore. An overall shape character describe d the genital papillae at the most coarse level using four character stat es: mound, mound tube intermediate, tube, and rosette. Characters specific to each overall shape further described the variation within those four coarse character state s Tube papillae were determined to dominate the genus but vary interspecifically in ma ny respects. In fact, t he papillae that required the most secondary
80 descript ion were the tube papillae which were described using s ix additional characters. Rosette papillae were described using four character s. Mound tube intermediate papillae were descri bed using t wo characters. All mound papillae were similar enough to not require additional characters to describe variation. The fourteenth character, basal platform, was a structure found in association wi th both tube and rosette papillae and was scored a s if applicable to all overall shapes. Gathering data on the 14 characters led to the discovery at least 13 distinct morphologies Out of these 13 morphologies, nine of them were variations on the tube morphology, three of them were v ariations on the roset te morphology, and the final on e is the mound morphology. It was unclear whether the intermediate mound tube m orphology was distinct or whether the specimens examined that received this overall shape score were collected outside of peak spawning, or altern atively, poorly pres erved. The variation in tube genital papillae was impressive, meriting at least nine distinct categories. Some of these variant morphologies were common being found in up to 30 taxa, whereas others were extremely unique being noted in only a single taxa. The most common variant of tube papillae were cylindrical tubes accompanied by basal platforms which were noted in 30 taxa 29 from subgenus Etheostoma and the other one species was E ( Allohistium ) cinereum Further variation within t his group was noted in whether the genital papilla possessed a terminal concavity (27 spp) or the distal ventral edg e of the tube extended past the dorsal edge and was dorsoventrally flattened ( E. cinereum E. histrio E. occidentale ). In addition, there w as variation noted in the lengt h of these tube papillae but the presence of the basal platform was ultimately important enough to
81 bring this group together. The ne xt most common tube variant was the spatulate tube found in 29 taxa from subgenera Hololepis (n = 3), Nothonotus (n = 1), Oligocephalus (n = 22) Poecilichthys (n = 1) and Psychromaster (n = 2) The representative from Poecilichthys ( E. kanawhae ) could alternatively be placed in the next group of distinct morphologies. Also, there was variation n oted in the shape of the lateral edges of these spatulate tubes. Some appeared crenulate, and others were a smooth, uniform lateral edge. The remainder of the distinct morphologies were shared by far fewer taxa. For example, a girth y, slightly conical tub e papilla was noted in six species from two subgenera, Litocara and Poecilicthys The most extreme case of a conical tube was noted in E. tetrazonum Another distinct morphology was quite similar to the most commonly exhibited morphology in its cylindrical nature and the presence of a terminal concavity, however, the difference was it lack ed a basal platform Four species possessed this morphology: E. ( Doration ) jessiae E. ( Belophlox / Hololepis ) mariae E. ( Doration ) stigmaeum E. ( Etheostoma ) swannanoa The next distinct morphology was a tube with a slit shaped genital pore opening that ran anteroposterioly and was displaced from a distal position and was oriented anteriad This morphology was shared by three taxa from subgenus Hololepis : E. fusiforme E. gr acile and E. saludae An incredibly unique morphology was shared by the sister taxa from subgenus Microperca E. fonticola and E. proeliare They possess ed tubes with dorsally bilobed tissue that extend ed beyond the genital pore. For this reason, they wer e scored as having a displaced genital pore position. The ventral half of the tube had anteroposterioly oriented pleats that developed into villi towards the genital pore. The other member of
82 Microperca E. microperca possesse d the only single species mor phology. It look ed similar to the other genital papillae within the subgenus except it lack ed the dorsally bilobed extension of tissue. The next tube variant was one that was noted in previous studies for E ( Villora ) edwini (Collette and Yerger, 1962 ), a short, stout, heavily villiate d tube and it was also shared by E ( Hololepis ) okaloosae T h e final distinct tube morphology was noted in specimens of E. chlorosoma This was an almost rectangular cuboid tube with a genital pore displaced toward s the anus t hat faced anteriad The genital pore was also found along a medially posi tioned anteroposteriorly oriented indention. The sister species to E chlrosoma E. davisoni possibly possess ed a less robust form of this morphology. E theostoma davisoni was scored as having an intermediate mound tube papil la but more specimens of gravid females should be examined to determine whether they truly share d this morphology ; only one specimen of E. davisoni was examined from a collection made in Alabama in late February T here were three distinct rosette morphologies recognized. These morphologies received more attention in the results portion of this thesis than the tube variants The first variants of the rosette papilla were the 20 taxa with a pleated papillar platform which included all the representatives of subgenus Catonotus included in this study. These pleats were generally arranged radially and developed into villi towards the center of the papillar platform. The second distinct rosette papilla morphology was shar ed by the five representatives from Boleosoma included in this study. This rosette papilla morphology was defined by a bifurcated papillar platform that lacked pleats. The bifurcated papillar platform took a bilobed form in E. nigrum and E. perlongum while the other three taxa ( E. longimanum E. olmstedi and E. podostemone ) actually had a third
83 lobe positioned dorsomedially to the two lateral lobes. The final distinct rosette papilla morphology belonged to E ( Ioa ) vitreum and had been used in past studies as diagnostic in an identification key (Jenkins and Burkhead 1993). The papilla was scored as having a pleated papillar platform as well as a bifurcated papillar platform and villi. The pleats were not obvious ly radial ly arrange d they were just noticeab le at the base of the papillar platform below where the tissue developed into villi. The villi were perhaps the most distinguishing feature as they were extremely numerous and large. The bifurcated papillar platform was present, but not in such a robust fo rm as what existed throughout Boleosoma The final distinct morphology was the m ound papilla which belong ed to 10 taxa representing Nothonotus Psychromaster and the extinct E. sellare ( insertae cedis ). The assignment of E. sellare to this group was quest ionable as noted in the results. The score was based on a single specimen collected in November of 1965, well outside of the preferred month constraints stated in the methods. It was included in the study, however, because any insight to its possible phylo genetic position within Etheostoma and evidence of the reproductive strategy it employed would be an exciting product of this study. Outside of these 13 distinct morphologies, the morphology of 13 additional species remained uncertain and consequently the se species were left unassigned. These thirteen species represent ed subgenera Fuscatelum Hololepis Nothonotus and Oligocephalus Included in this group were all species that were scored as having an intermediate mound tube papilla. This could serve as a n additional, fourteenth distinct group, but as of now this designation may be unjustified as many of the specimens
84 examined from this group did not look to be in peak spawning condition, especially the representatives from E. chuckwatchatte and E. tippeca noe The same could be said about the specimen of E. rubrum that was examined which was scored as having a tub e papilla Ultimately, more specimens of all of the thirteen species left unassigned would need to be examined to make accurate statements about t heir genital papilla morphologies. Character Evolution and Phylogenetic Implications It was determined that the diversity in papilla morphology throughout Etheostoma was derived from a simple tub e papilla. reconstr ucted as fairly complex on each phylogeny, particularly in Nothnotus a clade in which several transitions were made to shorter tubes, mound tubes, and mounds. Elsewhere, t he tube morphology was reconstructed as transitioning to rosette from one to three t imes independently. Subsequent t ransitions from rosette to mound and tube were also found, but only with support on the 3TE phylogeny. Tube transitioned to mound tube at E. davisoni on all phylogenies. Lastly, tube transitioned to mound at E. ( Psychromaste r ) pallididorsum and E. ( Psychromaster ) tuscumbia on the AFLP, nDNA, and 5TE phylogenies. All other taxa conserved the tube shape although the simple Genital papilla morphology was extremely valuable in testing phylog enetic relationships among Etheostoma There were examples of this at both the intersubgeneric level and the interspecific level. C haracters that were most informative were overall shape (e.g. mound and rosette character states) characters descriptive of rosette papillae and the presence /absence of a basal platform. These characters were
85 informative for relationships among darters in Allohistium Boleosoma Catonotus Etheostoma Ioa Nothonotus and Psychromaster The basal platform was informative in r esolving deeper, intersubgeneric relationships. The subgenera that had this structure in common were Allohistium Boleosoma Catonotus Etheostoma and Ioa The three gene combined evidence tree presented a clade in which the basal platform was almost comp letely conserved (Near et al 2011) The exception was the inclusion of the subgenus Psychromaster in the clade as all of its constituents lack a basal platform. Either this structure was los t or the reconstruction indicate d a misplacement of Psychromaster in the phylogeny. The position of Allohistium in Etheostoma has been uncertain for years. Some p hylogenetic hypotheses suggest ed that E ( Allohistium ) cinereum was sister to Nothonotus or even more closely related to Ammocrypta or Percina (Song et al 1 998 ; Smith et al ., 2011 ) However, E. cinereum was found nested within the subgenus Etheostoma with strong Bayesian posterior probability support ( 1.00 and 0.94, respectively ) in other analyses (Near et al ., 2011; Nick Lang, pers. comm. ) Similarity in pap illa morphology supported the latter hypothesis, as E. cinereum and 29 o f the 30 representatives from the subgenus Etheostoma had basal platforms in combination with tube papillae. This morphology was unique to the combined Allohistium and Etheostoma clade The one species in this clade scored as not having a basal platform was E. swannanoa which had a loss or reduction in this structure Alternatively, E. swannanoa could be scored incorrectly if the five specimens examined were not representative of the s realized character state. Otherwise, the tube and basal platform morphology was w ell conserved and diagnostic for this clade. T his result also
86 corroborate d the results from a previous phylogenetic analysis that used a m orphological data set (Ayache and Near, 2009). Alternative hypotheses that placed Allohistium sister to Nothonotus or to all other members of Etheostoma present ed a less parsimonious, less desirable scenario of morphological homoplasy R osette papilla e were found to be specific to thr ee subgenera : Boleosoma Catonotus and Ioa These subgenera have been hypothesized to be closely related ; however, their monophyly has rarely been supported in molecular and morphological phylogenetic analyses (Song et al ., 1998) T he only analysis that r ecovered them as monophyletic was weakly supported and the analysis did not include representatives from the S pottail D arter clade (Song et al 1998) The next closest result to monophyly recover ed a clade composed of Boleosoma Catonotus and Ioa with Ps ychromaster sister to Boleosoma + Ioa (Near et al ., 2011; Smith et al ., 2011). The clade compositions were the same between the two studies, but the topologies differ ed The inclusion of P sy chromaster a clade of species with tube and mound papillae, negat e d the use of the rosette papilla as a synapomorphy for Boleosoma Catonotus and Ioa Support for this clade suggest ed that a rosette papilla emerged only once in Etheostoma and was lost at the MRCA to Psychromaster T he likelihood reconstructions using t he five gene combined evidence phylogeny ( Nick Lang, pers comm ) offered an equally parsimonious scenario but included an additional suite of taxa from Oligocephalus in the clade. In this hypothesis, Psychromaster was not recovered as sister to any of the rosette papillae clades. V ariation within rosette papillae was also phylogenetically informative. The t hree distinct morphologi es within the r osette papillae,1) rosette with villi and a pleated
87 papillar platform 2) rosette with bifurcated papillar platfo rm 3) rosette with villi and both a bifurcated a nd a pleated papillar platform, each corresponded to a subgenus, Catonotus Boleosoma and Ioa respectively. The rosette morphology specific to all of the Catonotus species was very suggestive of the s monophyly Barcheek, f antail, and s pottail d arters, the clades that have historically been recognized in Catonotus (e.g., Page 2000) have not been consistently recovered as monophyletic (Smith et al ., 2011) Rather, b archeek and f antail darters often a re recovered as sister clades while spottail darters are recovered as sister to large clades that include the barcheek and fantail darters along with subgenera Psychromaster and Oligocephalus (Near et al ., 2011 Nick Lang, pers comm ) The lat t er pattern was consistent enough that a new clade name was recently proposed for the spottail darters, Stigmacerca and Catonotus was applied to the B archeek + F antail D arter clade (Near et al ., 2011). The rosette morphology specific to Boleosoma and Ioa corroborate d their sister group relationship that has been consistently hypothesized (Wood and Mayden, 1997; Song et al ., 1998; Sloss et al ., 2004; Near et al ., 2011; Smith et al 2011; Nick Lang, pers. comm. ) Both groups have rosette papillae with bifurcated papilla r platforms. While the bifurcated papillar platform was not as conspicuous on the E. ( Ioa ) vitreum examined this morphology still serve d as a synapomorphy for the consistently recovered clade Aside from the shared bifurcated papillar platform, Ioa was se t apart from Boleosoma by the presence of pleats on the papillar platform and the more numerous and conspicuous villi than were present on the papillae of Boleosoma
88 species The se morphological distinctions between the two subgenera agree d with previous f indings (Bailey and Etnier 1988; Jenkins and Burkhead 1993) Mound papillae were found to be homoplastic within the genus because they emerg ed independently in Nothonotus and Psychromaster The reconstruction of e ach of these emergence events provide d su pport for phylogenetic relationships. Firstly, the mound morphology of E. ( Psychromaster ) pallididorsum and E. ( Psychromaster ) tuscumbia supported the ir sister relationship hypothesized in nDNA tree (Nick Lang, pers. comm. ), a topology not consistently rec overed. The other representatives of Psychromaster had tube papillae. The mound morphology also occurred in Nothonotus as many as three times independently. Mound papillae were conserved in the Spotted Darter clade and offer additional morphological suppor t to results consistently recovered in molecular analyses (Near et al 2011; Smith et al 2011; Nick Lang, pers. comm. ; Tracy Smith, pers. comm. ) If mound emerged multiple times in Nothonotus the other two transitions took place terminally at E. bellum a nd E. juliae Based on findings presented here, future research efforts on the phylogenetic relationships within Etheostoma should be focused on the placement of Psychromaster If the constituents of Psychromaster are truly nested within a clade of darters that have rosette papillae and a basal platform then they have secondarily transitioned out of these character states to more primitive character states Genital Papilla Morphology and Egg Deposition The most obvious association between spawning behaviors and papilla morphology was between rosette papillae and egg clustering. All egg clustering darters but few others, were found to have rosette papillae. The exception was E. vitreum which was described as a communal spawner that attached eggs to rocks (W inn and
89 Picciolo 1960). However, this description was based on observation s made in unnatural stream conditions directly downstream of a concrete spillway and may not accurately describe the normal spawning behavior (Larry Page, pers comm ). The mound mor phology belonged to egg buriers, egg attachers, and egg clumpers. S pecies employing the former two behaviors possess ed a diversity of morphologies, but almost all of the egg clumpers possess ed mounds. The one exception was E. chlorobranchium which was ini tially scored as mound tube and has now been determined to be uncertain. Tube papillae belong ed to egg burying species and egg attaching species. It seemed the former behavior did not require a tube papilla because egg buriers were also scored as having m ound and mound tube papilla shapes. D iving buriers most commonly maintain ed the smaller, stouter morphologies with the exception of E. radiosum and to a lesser extent E. cragini In constrast, surface buriers commonly had genital papillae that were simple elongate tubes O ccasionally the tubes were spatulate; e.g. E. collettei E. fricksium E. spectabile and E. swaini The difference in morphology between the two burying behaviors may not be that surprising. Surface buriers may need an elongate genital papilla that can inject eggs into gravel or sand (Page 1983). Presumably, an elongate d genital papilla lost its utility in diving buriers because they accomplish ed with their bodies what the ovipositor did otherwise. Egg attaching species possess ed almos t exclusi vely tube papillae. One possible exception was again E. vitreum with a rosette papilla The other exception was E. davisoni which was initially scored as mound tube but the single specimen examined may have been a poor representation of a female at peak ripeness. Attaching eggs to
90 rocks seemed to require a tube papilla accompanied with a basal platform. This distinct morphology did seem to be versatile though ; some darters that possess ed it attach eggs to macrophytes and algae. T he other behavior al guild that was found to commonly use a rock substrate for oviposition were the egg clusterers which also possess ed basal platforms posterior to their genital papillae. F ive of the most derived distinct tube morphologies belong ed to species that attach e ggs to macrophytes or algae. These points suggest that t he diversification of oviposition site choice was accompanied by adaptations to genital papilla morphology. The Role of Genital Papillae in the Diversification of Etheostoma The interaction between a genital papilla and an oviposition substrate may determine the survivability of an egg. If eggs are not transferred successfully to an appropriate spot they may succumb to predation, infection, or physical damage. Genital papilla morphology may be indicati ve of how effectively a darter deposits eggs on a specific substrate. For example, the basal platform was present in all species that deposit eggs on rocks regardless of where on the rock the egg was placed. When the papilla is extremely swollen during th e spawning season the basal platform may be particularly helpful in propping a papilla downward, away from the body T he basal platform may contribute to precise oviposition on a rock y substrate. Even more specific ally all egg clusterers utilize the und erside s of rocks for ovipos i tion and they all possess ed rosette papillae with a basal platform. R osette papillae may be particularly suited to precisely and successfully attach eggs singly to the underside s of rocks whereas d arters with tube papillae and basal platforms may not have the ability to successfully exploit the underside of a rock as a n ovipositon site. The ability of darters to exploit new spaces and substrates for oviposition surely played a
91 role in their diversification as suggested by the pr ogression from darters simply burying eggs in gravel to more derived behaviors of attaching eggs across a stream landscape o n the sides of rocks, underside of rocks, or on vegetation.
92 Table 4 1. The overall shape states, coarse behavioral guilds, and fi ne behavioral guilds of each species examined in this study g rouped by distinct morphologies Distinct Genital Papillae Morphologies Overall Shape Score Egg Deposition Strategy Fine Behavioral Guild Tubes w/ basal platform E. ( Allohistium ) cinereum Tub e Attaching Rock Attaching E. ( Etheostoma ) atripinne Tube Attaching Rock Attaching E. ( Etheostoma ) baileyi Tube Attaching Rock Attaching E. ( Etheostoma ) barrenense Tube Attaching Rock Attaching E. ( Etheostoma ) bellator Tube Attaching ? E. ( Etheostoma ) blennioides Tube Attaching Algae Attaching E. ( Etheostoma ) blennius Tube ? ? E. ( Etheostoma ) brevirostrum Tube Attaching Rock Attaching E. ( Etheostoma ) chermocki Tube ? ? E. ( Etheostoma ) colorosum Tube Attaching Rock Attaching E. ( Etheostoma ) coosae Tube Attaching Rock Attaching E. ( Etheostoma ) duryi Tube Attaching Rock Attaching E. ( Etheostoma ) etnieri Tube Attaching Rock Attaching E. ( Etheostoma ) flavum Tube Attaching Rock Attaching E. ( Etheostoma ) gutselli Tube ? ? E. ( Etheostoma ) histrio Tube Attaching Algae Attaching E. ( Etheostoma ) inscriptum Tube ? ? E. ( Etheostoma ) lachneri Tube ? ? E. ( Etheostoma ) lynceum Tube ? ? E. ( Etheostoma ) occidentale Tube ? ? E. ( Etheostoma ) planasaxatile Tube ? ? E. ( Etheostoma ) pyrrhogaster Tube Attaching Macrophyte Attaching E. ( Etheostoma ) rafinesquei Tube Attaching Rock Attaching E. ( Etheostoma ) rupestre Tube ? ? E. ( Etheostoma ) scotti Tube Attaching Rock Attaching E. ( Etheostoma ) tallapoosae Tube Attaching Rock Attaching E. ( Etheostoma ) tennesseens e Tube ? ? E. ( Etheostoma ) thalassinum Tube ? ? E. ( Etheostoma ) zonale Tube Attaching Algae Attaching E. ( Etheostoma ) zonistium Tube Attaching Macrophyte Attaching
93 Table 4 1. Continued Distinct Genital Papillae Morphologies Overall Shape Score Egg Deposition Strategy Fine Behavioral Guild Spatulate tubes E. ( Hololepis ) fricksium Tube Burying Surface burying E. ( Hololepis ) hopkinsi Tube ? ? E. ( Hololepis ) serrifer Tube Attaching ? E. ( Nothonotus ) etowahae Tube ? ? E. ( Olig ocephalus ) artesiae Tube Burying ? E. ( Oligocephalus ) asprigene Tube Attaching Macrophyte Attaching E. ( Oligocephalus ) bison Tube ? ? E. ( Oligocephalus ) burri Tube ? ? E. ( Oligocephalus ) caeruleum Tube Burying Surface burying E. ( Oligoc ephalus ) collettei Tube Burying Surface burying E. ( Oligocephalus ) ditrema Tube Attaching Macrophyte Attaching E. ( Oligocephalus ) exile Tube Attaching Macrophyte Attaching E. ( Oligocephalus ) fragi Tube ? ? E. ( Oligocephalus ) kantuckeense Tu be ? ? E. ( Oligocephalus ) lawrencei Tube ? ? E. ( Oligocephalus ) lepidum Tube Attaching Algae Attaching E. ( Oligocephalus ) luteovinctum Tube Burying ? E. ( Oligocephalus ) nuchale Tube Attaching Macrophyte Attaching E. ( Oligocephalus ) rad iosum Tube Burying Diving E. ( Oligocephalus ) spectabile Tube Burying Surface burying E. ( Oligocephalus ) swaini Tube Burying Surface burying E. ( Oligocephalus ) tecumsehi Tube ? ? E. trisella Tube Attaching Macrophyte Attaching E. ( Oligoc ephalus ) uniporum Tube ? ? E. ( Oligocephalus ) whipplei Tube ? ? E. ( Poecilichthys ) kanawhae Tube ? ? E. ( Psychromaster ) boschungi Tube Attaching Macrophyte Attaching E. ( Psychromaster ) cragini Tube Burying Diving
94 Table 4 1. Continue d Distinct Genital Papillae Morphologies Overall Shape Score Egg Deposition Strategy Fine Behavioral Guild Girthy, conical tubes E. ( Litocara ) sagitta Tube Burying Surface burying E. ( Litocara ) spilotum Tube Attaching ? E. ( Poecilichthys ) e uzonum Tube ? ? E. ( Poecilichthys ) osburni Tube Burying ? E. ( Poecilichthys ) tetrazonum Tube Burying ? E. ( Poecilichthys ) variatum Tube Burying Surface burying Cylindrical tubes w/o basal platform E. ( Doration ) jessiae Tube Burying Surface burying E. ( Doration ) stigmaeum Tube Burying Surface burying E. ( Etheostoma ) swannanoa Tube Attaching Macrophyte Attaching E. ( Hololepis ) mariae Tube ? ? Tubes w/ anteriorly oriented genital pore E. ( Hololepis ) fusiforme Tube Attaching Macrophyte Attaching E. ( Hololepis ) gracile Tube Attaching Macrophyte Attaching E. ( Hololepis ) saludae Tube ? ? Rectangular cuboid tubes E. ( Vaillantia ) chlorosoma Tube Attaching Macrophyte Attaching E. ( Vaillantia ) daviso ni Mound tube Attaching Macrophyte Attaching Heavily villiate tubes E. ( Hololepis ) okaloosae Tube Attaching Algae Attaching E. ( Villora ) edwini Tube Attaching Macrophyte Attaching Tubes w/ bilobed distal tissue E. ( Microperca ) f onticola Tube Attaching Algae Attaching E. ( Microperca ) proeliare Tube Attaching Algae Attaching Microperca type villiate tube E. ( Microperca ) microperca Tube Attaching Macrophyte Attaching
95 Table 4 1. Continued Distinct Genital Papillae M orphologies Overall Shape Score Egg Deposition Strategy Fine Behavioral Guild Rosette papillae w/ radially pleated papillar platform E. ( Catonotus ) barbouri Rosette Clustering Beta clustering E. ( Catonotus ) basilare Rosette Clustering Beta clust ering E. ( Catonotus ) chienense Rosette Clustering Alpha clustering E. ( Catonotus ) corona Rosette Clustering Alpha clustering E. ( Catonotus ) crossopterum Rosette Clustering Alpha clustering E. ( Catonotus ) derivativum Rosette Clustering Beta clustering E. ( Catonotus ) flabellare Rosette Clustering Beta clustering E. ( Catonotus ) forbesi Rosette Clustering Alpha clustering E. ( Catonotus ) kennicotti Rosette Clustering Beta clustering E. ( Catonotus ) marmorpinnum Rosette Clustering B eta clustering E. ( Catonotus ) neopterum Rosette Clustering Alpha clustering E. ( Catonotus ) nigripinne Rosette Clustering Alpha clustering E. ( Catonotus ) obeyense Rosette Clustering Beta clustering E. ( Catonotus ) olivaceum Rosette Clustering Alpha clustering E. ( Catonotus ) oophylax Rosette Clustering Alpha clustering E. ( Catonotus ) percnurum Rosette Clustering Beta clustering E. ( Catonotus ) smithi Rosette Clustering Beta clustering E. ( Catonotus ) squamiceps Rosette Clustering Alpha clustering E. ( Catonotus ) striatulum Rosette Clustering Beta clustering E. ( Catonotus ) virgatum Rosette Clustering Beta clustering Rosette w/ bifurcate papillar platform E. ( Boleosoma ) longimanum Rosette Clustering ? E. ( Boleos oma ) nigrum Rosette Clustering Gamma clustering E. ( Boleosoma ) olmstedi Rosette Clustering Gamma clustering E. ( Boleosoma ) perlongum Rosette Clustering Alpha clustering E. ( Boleosoma ) podostemone Rosette Clustering ? Rosette w/ dense vill i and bifurcate papillar platform E. ( Ioa ) vitreum Rosette Attaching Rock Attaching
96 Table 4 1. Continued Distinct Genital Papillae Morphologies Overall Shape Score Egg Deposition Strategy Fine Behavioral Guild Mound E. ( Nothonot us ) aquali Mound Clumping Clumping E. ( Nothonotus ) bellum Mound Burying Diving E. ( Nothonotus ) juliae Mound Burying Diving E. ( Nothonotus ) maculatum Mound Clumping Clumping E. ( Nothonotus ) microlepidum Mound Clumping Clumping E. ( Nothon otus ) sanguifluum Mound Clumping Clumping E. ( Nothonotus ) vulneratum Mound Clumping Clumping E. ( Psychromaster ) pallididorsum Mound Attaching Macrophyte Attaching E. ( Psychromaster ) tuscumbia Mound Burying Surface burying Uncertain morpho logies A. vivax Tube Burying Surface burying E. ( Fuscatelum ) parvipinne Tube Attaching Macrophyte Attaching E. ( Hololepis ) collis Tube Attaching Macrophyte Attaching E. ( Nothonotus ) acuticeps Mound tube Burying Diving E. ( Nothonotus ) camurum Tube Burying Diving E. ( Nothonotus ) chlorobranchium Mound tube Burying Diving E. ( Nothonotus ) chuckwachatte Mound tube ? ? E. ( Nothonotus ) jordani Mound tube Burying Diving E. ( Nothonotus ) moorei Tube Burying ? E. ( Nothonotus ) r ubrum Tube Burying Diving E. ( Nothonotus ) rufilineatum Mound tube Burying Diving E. ( Nothonotus ) tippecanoe Mound tube Burying Diving E. ( Oligocephalus ) australe Tube ? ? E. ( Oligocephalus ) grahami Tube Attaching Macrophyte Attaching P. ( Alvordius ) maculata Tube Burying Surface burying
97 APPENDIX GLOSSARY Basal platform hypertrophied tissue immediately posterior to the base of the genital papilla, between the papilla and the anal fin insertion. This tissue is free of tubular papilla and may be fused to (dependent upon the species and time of year) or free of a papillar platform. Crenulate possessing wavy or notched edges on genital papilla, a character that is most obviously seen in some species with spatulate tubular genital papi lla (e.g. Etheostoma caeruleum ) Medial groove an indentation that runs along the length of the genital papilla Mound a squat, bulbous form of genital papilla that remains close to the ventral surface of the body even during peak spawning season when t he structure is most swollen. (also, see true mound ) Papillar platform tissue immediately posterior of the genital pore present in species of the egg clustering guild. Papillar platforms exist in two major forms, radially pleated and bifurcate (forming two to three lobes). Papillar shield Dorsoventrally flattened tissue covering the genital pore in species of the egg clustering guild. Papillar shields exist in a fused, bilobed, and multi lobed form. Pleats folds in tissue, sometimes associated w ith villi (e.g. flower like genital papillae) Rosette papillae that possess a papillar shield, papillar platform, a basal platform, and a downward oriented genital pore that is displaced from a distal position. Spatulate descriptive term for tubular pa pillae that are dorsoventral flattened, particularly in the distal portion of the tube. Visor posterior tissue (tissue facing the ventral surface of the fish) at the distal portion of a tubular or mound to stout tube papilla that extends beyond the genit al pore (e.g. Etheostoma fonticola ).
98 LIST OF REFERENCES Ayache, N. C. and T. J. Near 2009. The utility of morphological data in resolving phylogenetic relationships of darters as exemplified with Etheostoma (Teleostei: Percidae). Bul letin of the Peabody Museum of Natural History 50:327 346. Bailey, R. M. and D. A. Etnier. 1988. Comments on the subgenera of darters (Percidae) with descriptions of two new species of Etheostoma ( Ulocentra ) from s outheastern United States. Occasional Pape rs of the Museum of Zoology, University of Michigan 175:1 48. Bart, J., H. L. 1992. Spawning b ehavior of Etheostoma davisoni Hay. Copeia 1992:537 539. Bauer, B. H., D. A. Etnier, and N. M. Burkhead. 1995. Etheostoma ( Ulocentra ) scotti (Osteichthys: Percida e), a new darter from the Etowah River system in Georgia. Bulletin of the Alabama Museum of Natural History 17:1 17. Boschung, H. T. 1986. Biology and conservation of the slackwater darter, Etheostoma boschungi (Pisces: Percidae). Southeastern Fishes Counc il Proceedings 4:1 4. Braasch, M. E. and P. W. Smith. 1967. The life history of the slough darter, Etheostoma gracile (Pisces, Percidae). Illinois Natural History Survey Biological Notes 58:1 12. Carlson, R. L. and P. C. Wainwright. 2010. The ecological mo rphology of darter fishes (Percidae: Etheostomatinae). Biological Journal of the Linnean Society 100:30 45. Carney, D. A. and B. M. Burr. 1989. Life histories of the bandfin darter Etheostoma zonistium and the firebelly darter Etheostoma pyrrhogaster i n w estern Kentucky. Illinois Natural History Survey Biological Notes 134:1 16. Collette, B. B. 1961. The systematics and biology of the darters of the subgenera Hololepis and Villora (Pisces, Percidae). Dissertations Abstracts 21. Collette, B. B. and R. W. Yerger. 1962. The american percid fishes of the subgenus Villora Tulane Studies in Zoology 9:213 230. Constantz, G. D. 1979. Social dynamics and parental care in the tessellated darter (Pisces: Percidae). Proceedings of the Academy of Natural Sciences of Philidelphia 131:131 138. Cummings, K. S., J. M. Grady, and B. M. Burr. 1984. The life history of the mud darter, Etheostoma asprigene in Lake Creek, Illinois. Illinois Natural History Survey Biological Notes 122:1 16. Distler, D. A. 1972. Observations o n the reproductive habits of captive Etheostoma cragini Gilbert. Southwestern Naturalist 16:439 441.
99 Duncan, R. S., C. P. Elliot, B. L. Fluker, and B. R. Kuhajda. 2010. Habitat use of the watercress darter ( Etheostoma nuchale ): An endangered fish in an urb an landscape The American Midland Naturalist 164:9 21. Eisenhour, D. J. and B. M. Burr. 2000. Conservation status and nesting biology of the endangered duskytail darter Etheostoma percnurum in the Big South Fork of the Cumberland River, Kentucky. Journa l of the Kentucky Academy of Science 61:67 76. Etnier, D. A. and W. C. Starnes. 1993. The Fishes of Tennessee. University of Tennessee Press, Knoxville, TN. Fahy, W. E. 1954. The life history of the northern greenside darter, Etheostoma blennioides blennio i d es Rafinesque. Proceedings of the Elisha Mitchell Scientific Study 70:139 205. Fisher, W. L. 1990. Life history and ecology of the orangefin darter Etheostoma bellum (Pisces: Percidae). American Midland Naturalist 123:268 281. Fletcher, A. M. 1976. A rar e darter spawning. American Currents 4:20 22. Guill, J. M., C. S. Hood, and D. C. Heins. 2003. Body shape variation within and among three species of darters (Perciformes: Percidae). Ecology of Freshwater Fish 12:134 140. Gumm, J. M., K. D. Feller, and T. C. Mendelson. 2011. Spectral characteristics of male nuptial coloration in darters ( Etheostoma ). Copeia 2011:319 326. Hubbs, C. 1985. Darter reproductive seasons. Copeia 1985:56 68. Hubbs, C. L. and M. D. Cannon. 1935. The darters of the genera Hololepis a nd Villora University of Michigan Museum of Zoology Miscellaneous Publication 30:93. James, P. W. and C. A. Taber. 1986. Reproductive biology and age and growth of the yoke darter Etheostoma juliae Copeia 2:536 540. Jenkins, R. E. 1980. Etheostoma podos temone .in D. S. Lee, C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister, and J. R. Stauffer, editors. Atlas of North American Freshwater Fishes. North Carlina State Museum, Raleigh. Jenkins, R. E. and N. M. Burkhead. 1993. Freshwater f ishes of Vi rginia. American Fisheries Society, Bethesda, Maryland. Johnston, C. E. 1994. Spawning behavior of the goldstripe darter ( Etheostoma parvipinne Gilbert and Swain) (Percidae). Copeia 1994:823 825. Johnston, C. E. 1995. Spawning behavior of the paleback dart er Etheostoma pallididorsum (Percidae). The Southwestern Naturalist 40:422 425.
100 Johnston, C. E., N. A. Farnax, H. L. Bart, and A. K. Howard. 1999. Laboratory observations of spawning behavior in two species of snubnose darters, Etheostoma colorosum and E tallapoosae Southeastern Fishes Council Proceedings 38:5 7. Johnston, C. E. and W. R. Haag. 1996. The life history of the yazoo darter (Percidae: Etheostoma raneyi ), a species endemic to north central Mississippi. Tulane Studies in Zoology and Botany 30 :47 60. Johnston, C. E. and D. L. Johnson. 2000. Sound production during the spawning season in cavity nesting darters of the subgenus Catonotus (Percidae: Etheostoma ). Copeia 2000:475 481. Johnston, C. E. and J. R. Shute. 1997. Observational notes on the spawning behavior of the blue shine r ( Cyprinella caerulea ) and the holiday darter ( Etheostoma brevirostrum ), two rare fishes of the Conasauga River, Georgia and Tennessee. Southeastern Fishes Council Proceedings 35:1 2. Keevin, T. M., L. M. Page, and C. E. Johnston. 1989. The spawning behavior of the Saffron darter ( Etheostoma flavum ). Transactions of the Kentucky Academy of Science 50:55 58. depositi on behaviours and the evolution of male parental care i n darters (Teleostei: Percidae: Etheostomatinae). Journal of Evolutionary Biology 25:836 846. Lake, C. T. 1936. The life hi story of the fan tailed darter Catonotus flabellaris flabellaris (Rafinesque). The American Midland Naturalist 17:816 830. Layman, S. R. 1984. The duskytail darter Etheostoma ( Catonotus ) sp., confirmed as an egg clusterer Copeia 1984:992 994. Layman, S. R. 1991. Life history of the relict, duskytail darter Etheostoma ( Catonotus ) sp., in Little River, Tennessee. Copeia 1991:471 485. Layman, S. R. 1993. Life history of the Savannah d arter, Etheostoma fricksium in the Savannah River d rainage, South Caroli na. Copeia 1993:959 968. Lindquist, D. G., J. R. Shute, and P. W. Shute. 1981. Spawning and nesting behavior of the Waccamaw Darter, Etheostoma perlongum Environmental Biology of Fishes 6:177 191. Maddison, W.P. & D.R. Maddison. 2006. StochChar: A package of Mesquite modules for stochastic models of character evolution. Version 1.1. Maddiso n, W. P. and D. R. Maddison. 2011 Mesquite: a modular system for evolutionary analysis. Version 2.75. http://mesquiteproject.org.
101 Mattingly, H. T., J. Hamilton, and D. L. Galat. 2003. Reproductive ecology and captive breeding of the threatened Niangua Darter Etheostoma nianguae The American Midland Naturalist 149:375 383. Mendelson, T. C. 2003. Evidence of intermediate and asymmetrical behavioral isolation between o rang ethroat and o rangebelly darters (Teleostei: Percidae). American Midland Naturalist 150:343 347. Mount, D. I. 1959. Spawning behavior of the bluebreast d arter, Etheostoma camurum (Cope). Copeia 1959:240 243. Muller, B. 2008. Scaly sand darter ( Ammocrypta vi vax ): observations and captive spawning American Currents 34:1 2. Near, T. J., C. M. Bossu, G. S. Bradburd, R. L. Carlson, R. C. Harrington, J. P. R. Hollingsworth, B. P. Keck, and D. A. Etnier. 2011. Phylogeny and temporal diversification of darters (Per cidae: Etheostomatinae). Journal of Systematic Biology 60:1 31. Orr, J. W. and J. S. Ramsey. 1990. Reproduction in the greenbreast darter Etheostoma jordani (Teleostei: Percidae). Copeia 1990:100 107. Page, L. M. 1974. The life history of the spottail dar ter Etheostoma squamiceps in Big Creek, Illinois, and Ferguson Creek, Kentucky. Illinois Natural History Survey Biological Notes 89:1 20. Page, L. M. 1975. Relations among the darters of the subgenus Catonotus of Etheostoma Copeia 1975:782 784. Page, L. M. 1976. The life history of the stripetail darter Etheostoma kennicotti in Big Creek, Illinois. Illinois Natural History Survey Biological Notes 93:1 15. Page, L. M. 1980. The life histories of Etheostoma olivaceum and Etheostoma striatulum two specie s of darters in central Tennessee. Illinois Natural History Survey Biological Notes 113:1 14. Page, L. M. 1981. The genera and subgenera of darters (Percidae, Etheostomatini). Occasional Papers of the Museum of Natural History, The University of Kansas 90: 1 69. Page, L. M. 1983. The handbook of d arters. TFH Publications, Neptune City, NJ. Page, L. M. 1985. Evolution of reproductive behaviors in percid fishes Bulletin of the Illinois Natural History Survey 33:275 295. Page, L. M. 2000. Etheostomatinae. Page s 225 253 in J. F. Craig, editor. Percid Fishes: Systematics, Ecology, and Exploitation. Blackwell Science, Oxford.
102 Page, L. M. and B. M. Burr. 1976. The life history of the slabrock darter Etheostoma smithi in Ferguson Creek, Kentucky. Illinois Natural History Survey Biological Notes 99:1 12. Page, L. M. and B. M. Burr. 2011. Peterson field guide to freshwater fishes of North America n orth of Mexico. Pages 508 602. Houghton Mifflin Harcourt Publishing Company, New York, NY. Page, L. M., P. A. Ceas, D. L. Swofford, and D. G. Buth. 1992. Evolutionary relationships within the Etheostoma squamiceps complex (Percidae: s ubgenus Catonotus ) with descriptions of five new species Copeia 1992:615 646. Page, L. M., W. L. Keller, and L. E. Cordes. 1981. Etheostoma ( B oleosoma ) longimanum and E ( Catonotus ) obeyense two more darters confirmed as egg clusterers Transactions of the Kentucky Academy of Science 42:35 36. Page, L. M. and R. L. Mayden. 1981. The life history of the Tennessee s nubnose d arter, Etheostoma simo terum in Brush Creek, Tennessee. Illinois Natural History Survey Biological Notes 117:1 11. Page, L. M., M. E. Retzer, and R. A. Stiles. 1982. Spawning behavior in seven species of darters (Pisces: Percidae). Brimleyana 8:135 143. Page, L. M. and T. P. Sim on. 1988. Observations on t he reproductive behavior and eggs of four species of darters, with comments on Etheostoma tippecanoe and E. camurum Transactions of the Illinois State Academy of Science 81:205 210. Page, L. M. and D. L. Swofford. 1984. Morpholo gical correlates of ecological specialization in darters. Environmental Biology of Fishes 11:139 159. Pflieger, W. L. 1978. Distribution, status, and life history of the Niangua darter, Etheostoma nianguae Missouri Department of Conservation Aquatic Serie s 16: 1 25. Piller, K. P. and B. M. Burr. 1999. Reproductive biology and spawning habitat supplementation of the relict darter, Etheostoma chienense a federally endangered species. Environmental Biology of Fishes 55. Porterfield, J. C. 1997. Separation of spawning habitat in the sympatric snubnose darters Etheostoma flavum and E. simoterum (Teleostei, Percidae). Transactions of the Kentucky Academy of Science 58:4 8. Porterfield, J. C. 1998. Spawning behavior of snubnose darters (Percidae) in natural and l aboratory environments Environmental Biology of Fishes 53:413 419. Rakes, P. L., J. R. Shute, and P. W. Shute. 1999. Reproductive behavior, captive breeding, and restoration ecology of endangered fishes. Environmental Biology of Fishes 55:31 42.
103 Raney, E. C. and E. A. Lachner. 1939. Observations on the life history of the s potted d arter, Poecilichthys maculatus (Kirtland). Copeia 1939:157 165. Ross, S. T. and S. D. Wilkins. 1993. Reproductive behavior and larval characteristics of the threatened bayou dart er ( Etheostoma rubrum ) in Mississippi. Copeia 1993:1127 1132. Ruple, D. L., J. R. H. McMichael, and J. A. Baker. 1984. Life history of the g ulf d arter, Etheostoma swaini (Pisces: Percidae). Environmental Biology of Fishes 11:121 130. Scalet, C. G. 1973. R eproduction of the orangebelly darter Etheostoma radiosum cyanorum (Osteichthyes:Percidae). American Midland Naturalist 89:156 165. Simmons, J. W. and J. B. Layzer. 2004. Spawning behavior and habitat of the endangered bluemask darter Etheostoma ( Doratio n ) sp. Copeia 2004:412 417. Simon, T. P. 1997. Ontogeny of the darter subgenus Doration with comments on intrasubgeneric relationships Copeia 1997:60 69. Simon, T. P. and R. Wallus. 2006. Percidae -Perch, Pikeperch, and Darters. CRC Press, Boca Raton, Flo rida, USA. 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, T. A., T. C. Mendelson, and L. M. Pag e. 2011. AFLPs support deep relationships among darte rs (Percidae: Etheostomatinae) consistent with morphological hypotheses. Heredity 107:579 588. Song, C. B., T. J. Near, and L. M. Page. 1998. Phylogenetic relations among percid fishes as inferred from m itochondrial cytochrome b DNA sequence data. Molecular Phylogenetics and Evolution 10:343 353. Steinberg, R., L. M. Page, and J. C. Porterfield. 2000. The spawning behavior of the harlequin darter, Etheostoma histrio (Osteichthyes: Percidae). Ichthyologica l Exploration of Freshwaters 11:141 148. Strawn, K. 1955. A Method of Breeding and Raising Three Texas Darters. The Aquarium Journal 26:12 17, 31, 408 412. Walters, J. P. 1994. Spawning b ehavior of Etheostoma zonale (Pisces: Percidae). Copeia 1994:818 821. Warren Jr., M. L., B. M. Burr, and B. R. Kuhajda. 1986. Aspects of the reproductive biology of Etheostoma tippecanoe with comments on egg burying behavior. American Midland Naturalist 116:215 218.
104 Widlak, J. C. and R. J. Neves. 1985. Age growth, food hab its, and reproduction of the redline darter Etheostoma rufilineatum (Cope) (Perciformes: Percidae) in Virginia. Brimleyana 11:69 80. Wiley, M. L. and B. B. Collette. 1970. Breeding tubercles and contact organs in fishes: their occurrence, structure, and si gnificance. Bulletin of the American Museum of Natural History 143:197 201. Winn, H. E. 1958. Comparative reproductive behavior and ecology of fourteen species of darters (Pisces Percidae). Ecological Monographs 28:155 191. Winn, H. E. 1958. Observation on the reproductive habits of darters (Pisces Percidae). The American Midland Naturalist 59:190 212. Winn, H. E. and A. R. Picciolo. 1960. Communal spawning of the glassy darter Etheostoma vitreum (Cope). Copeia 1960:186 192. Wood, R. M. and R. L. Mayden. 19 97. Phylogenetic relationships among selected darter subgenera (Teleostei: Percidae) as inferred from analysis of allozymes. Copeia 1997:265 274.
105 BIOGRAPHICAL SKETCH Zachary P. Martin was born in Brunswick, Maine. He was raised one of four children. His f ather was in the U.S. Navy and his mother was an educator. Their family traveled throughout the southeast United Sta tes during his childhood Zachary spent most of his teen years in Chesapeake, Virginia While attending high school Zach played soccer, ran track, and served as his high schools beloved mascot, the Great Bridge Wildcat. After graduating high school he earned his Bachelors of Science in Marine Biology at the University of North Carolina Wilmington. During this time he s Fisheries Ecology Lab studying the life history of estuarine fishes. Following his undergraduate career, Zach moved to Gainesville, Florida to work in the Florida Rivers Lab as a laboratory technician. Zachary was a student at the University of Florida in the Department of Biology and pursued his Master of Science in Zoology.