Snelson_48(4)_Cover BULLETIN UNIVERSITY OF FLORIDA GAINESVILLE ELASSOMA GILBERTI, A NEW SPECIES OF PYGMY SUNFISH (ELASSOMATIDAE) FROM FLORIDA AND GEORGIA Franklin F. Snelson, Jr., Trevor J. Krabbenhoft, and Joseph M. Quattro Vol. 48, No. 4, pp. 119-144 2009 ™ SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 145 The FLORIDA MUSEUM OF NATURAL HISTORY is Florida’s state museum of natural history, dedicated to understanding, preserving, and interpreting biological diversity and cultural heritage. The BULLETIN OF THE FLORIDA MUSEUM OF NATURAL HISTORY is a peer-reviewed publication that publishes the results of original research in zoology, botany, paleontology, and archaeology. Address all inquiries to the Managing Editor of the Bulletin. Numbers of the Bulletin are published at irregular intervals. Specific volumes are not necessarily completed in any one year. The end of a volume will be noted at the foot of the first page of the last issue in that volume. Richard Franz, Managing Editor Cathleen Bester, Production Bulletin Committee Richard Franz, Chairperson Ann Cordell Sarah Fazenbaker Richard Hulbert William Marquardt Larry Page Irvy R. Quitmyer David W. Steadman, Ex officio Member ISSN: 0071-6154 Publication Date: September 15, 2009 Send communications concerning purchase or exchange of the publication and manuscript queries to: Managing Editor of the BULLETIN Florida Museum of Natural History University of Florida PO Box 117800 Gainesville, FL 32611-7800 U.S.A. Phone: 352-392-1721 Fax: 352-846-0287 e-mail: lrfranz08@gmail.com SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 119 1 Florida Museum of Natural History, University of Florida, Gainesville, FL 32611 2 Marine Sciences Program and Department of Biological Sciences, University of South Carolina, Columbia, SC 29208 Current addresses: 3 1701 SW 58th Avenue, Gainesville, FL 32608. 4 Department of Biology and Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM 87131. Snelson, F.F. Jr., T.J. Krabbenhoft, and J.M. Quattro. 2009. Elassoma gilberti, a New Species of Pygmy Sunfish (Elassomatidae) from Florida and Georgia. Bull. Florida Museum Nat. Hist. 48(4): 119-144. ELASSOMA GILBERTI, A NEW SPECIES OF PYGMY SUNFISH (ELASSOMATIDAE) FROM FLORIDA AND GEORGIA Franklin F. Snelson, Jr.1,3, Trevor J. Krabbenhoft 2,4, and Joseph M. Quattro2 ABSTRACT A new species of pygmy sunfish, Elassoma gilberti (Elassomatidae), is described from northwestern Florida and extreme southwestern Georgia. It previously has been confused with its sister species, Elassoma okefenokee Böhlke 1956. The two are very similar morphologically, but differ in the number of preopercular canal pores (four in E. gilberti, three in E. okefenokee), in average number of anal fin rays (usually seven in E. gilberti, usually eight in E. okefenokee), and in more subtle differences in coloration, body depth, and dorsal and anal fin size. The distinction of the two species is supported by eight fixed differences at the mitochondrial 16S rRNA locus and 12 fixed differences at the nuclear S7 locus. Phylogenetic analyses using these molecular characters supported monophyletic clades that contained haplotypes and alleles found uniquely in the two taxa. Elassoma gilberti is found in stream systems draining into the Gulf of Mexico from Choctawhatchee Bay in the Florida panhandle south to the Withlacoochee and Homosassa drainages in west-central Florida. Both species occur in the Suwannee River drainage, E. gilberti in the lower and middle sections and E. okefenokee in the middle and upper sections. They remain genetically distinct where sampled in this drainage but have not been found syntopically. The history and nomenclatural status of the name Elassoma evergladei orlandicum Lönnberg 1894 is discussed and a lectotype is designated based on the earlier findings of R. M. Bailey and J. E. Böhlke. Lectotype designation relegates the name to the synonymy of Elassoma evergladei Jordan 1884. TABLE OF CONTENTS Introduction........................................................................................................120 Methods............................................................................................................. 120 Morphology................................................................................................. 120 Molecular Techniques.................................................................................. 120 Phylogenetic Analysis..................................................................................121 Elassoma gilberti n. sp.............................................................................. 121 Morphological Comparison of Elassoma gilberti and E. okefenokee..................128 Molecular Analysis............................................................................................ 130 Results......................................................................................................... 130 Phylogenetic Analysis..................................................................................132 Distribution........................................................................................................ 132 Status of “Elassoma evergladei orlandicum” Lönnberg ................................... 139 Acknowledgements.............................................................................................141 Literature Cited.................................................................................................. 141 Appendix 1.........................................................................................................142 Appendix 2.........................................................................................................144 Key Words: pygmy sunfish, Elassoma, Florida, Georgia, new species, nomenclature. 120 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) INTRODUCTION The pygmy sunfishes, family Elassomatidae, are endemic to the southeastern United States. The family consists of six described species, all in the genus Elassoma. The characters distinguishing the species are summarized by Mayden (1993). Although the phylogenetic affinities and classification of the family have been the subject of con- siderable controversy and speculation (see summary in Nelson 2006), the monophyly of the six Elassoma spe- cies is supported by genetic analysis (Jones & Quattro 1999; Quattro et al. 2001; Roe et al. 2002) and by a number of shared apomorphic morphological traits (Branson & Moore 1962; Johnson 1984). All species of Elassoma are small, averaging between 25-35 mm stan- dard length as adults, and prefer springs, swamps, ditches, or slow moving streams with abundant submerged veg- etation. In this paper we describe a new species of pygmy sunfish that is sister and closely related, both morpho- logically and genetically, to Elassoma okefenokee Böhlke 1956. In the early 1990’s, FFS noticed that E. okefenokee collected in central Florida usually had three pores in the preopercular (PO) branch of the cephalic lateral-line canal, whereas Elassoma evergladei Jor- dan and Elassoma zonatum Jordan collected in the same region usually had four PO pores. Further study of material housed at the Florida State Museum of Natural History revealed that all species of Elassoma normally have four PO pores, including populations of E. okefenokee from the Florida panhandle. Additional study, now supported by genetic analysis and distribu- tional data, reveals that the four-pored form of “okefenokee” is specifically distinct from the true E. okefenokee, and that E. okefenokee is unique in the genus in having three PO pores throughout its range. Herein we describe this new species and detail its distribution in relation to Elassoma okefenokee, espe- cially in the Suwannee River system of Florida and Geor- gia, where both species occur but remain separated geo- graphically. We also present analysis based on both mitochondrial and nuclear gene sequence data that is entirely consistent with the morphological analysis and supports reciprocal monophyly of the two sister taxa. Finally, we re-examine the nomenclatural status of “Elassoma evergladei orlandicum” Lönnberg 1894 and designate a lectotype based on earlier analysis by R. M. Bailey and J. E. Böhlke. METHODS MORPHOLOGY Methods used in making counts and measurements follow Hubbs and Lagler (1974) and Rohde and Arndt (1987). All measurements were made to the nearest 0.1 mm with dial calipers under a dissecting microscope. Head length was measured to the fleshy end of the oper- cular flap and head depth was measured at the occiput. Body depth was measured vertically at the origin of the dorsal fin. Dorsal and anal fin lengths were measured from the base of the first spine to the tip of the longest ray. Paired fin measurements were made on the right side of the body if the left side was damaged or ap- peared abnormal. Pectoral fin length was measured from the structural base to the tip of the longest ray near the middle of the fin, with the fin pressed flat against the body. Pelvic fin length was measured from the base of the spine to the tip of the longest ray. The dorsal and anal fin ray counts include all elements separately, with the last two elements counted independently when they were separated to the body. Fin ray elements were counted under a dissecting microscope with transmitted light. Scale and fin ray counts were made under a dis- secting microscope using a jet of compressed air to aid in scale definition. Scales are small and their patterns are often irregular, making it necessary to repeat some counts several times until a consensus count was reached. The naming of the cephalic lateral line canals fol- lows Branson and Moore (1962) except that their preoperculomandibular canal (POM) is called the preopercular canal (PO) since the mandibular portion of the canal is absent in Elassoma. Preopercular pore counts presented in the format 3-3 mean that there are three pores in the PO on the left side of the head and three on the right. Counts presented in the format 4- 2+2 mean that there are four pores on the left side and, on the right side, the canal is divided into two indepen- dent segments, each containing two pores. Life color descriptions were made with magnifi- cation from live specimens immobilized in ice water, from material preserved in 10% buffered formalin for less than 30 minutes, and from color photographs of live and fresh specimens. MOLECULAR TECHNIQUES Total genomic DNA was obtained from caudal fin clippings using the QIAGEN DNeasy® Blood and Tis- sue Kit following the manufacturers’ protocol. Pres- ence of total genomic DNA was confirmed visually by ethidium bromide-stained 1.5% agarose gel electrophore- sis. A 553 base pair fragment of 16S rRNA (16S, mtDNA) and the entire 583 base pair S7 Intron 1 (S7, nuclear DNA) were amplified via the polymerase chain reaction (PCR). The 16S mtDNA locus was amplified using the 16Sa and 16Sb primer pair described in Palumbi (1996), while the nuclear S7 locus was amplified with SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 121 S7RPEX1F and S7RPEX2R primers described in Chow and Hazama (1998). Reaction conditions consisted of an initial 94 °C disassociation phase for 4 minutes, fol- lowed by 40 cycles of 94 °C for 1 minute, 48 °C for 1 minute and 72 °C for 1 minute. A final 7-minute 72 °C extension phase was added to the end of each 40 cycle reaction profile. Presence of amplicons was confirmed visually by ethidium bromide-stained 1.5% agarose gel electrophoresis. PCR products were precipitated with a 20% poly- ethylene glycol/2.5 M NaCl mixture and the precipitates washed twice with 70% ethanol (Applied Biosystems 1994). The forward and reverse PCR primers were used as forward and reverse sequencing primers in sepa- rate reactions using the ABI BigDye® Terminator ver- sion 3.1 Cycle Sequencing Kit. Sequencing reactions were then read on an ABI 377 automated sequencer. Sequence files were exported into Sequencher™ (Gene Codes Corporation, Ann Arbor, Michigan) and contigs made of forward and reverse sequences from each in- dividual. The accuracy of all base calls for all contigs was checked by eye. Contigs were exported from Sequencher™ as text files for further genetic analyses. Sequences were aligned using ClustalW (Thompson et al. 1994) using the default parameters. Minor modifica- tions to the initial alignment were made manually. Both data matrices in nexus format are available from JMQ upon request. Since sequence variation at the S7 locus involved very few polymorphic sites within species, het- erozygotes could be inferred unambiguously from chro- matograms as overlapping, equally intense bands at single base positions on the trace files; these positions were consistent on both sequenced strands. Nucleotide se- quences of unique mtDNA haplotypes and nuclear DNA alleles have been submitted to Genbank (accession num- bers GQ477414-GQ477438). PHYLOGENETIC ANALYSIS Parsimony and maximum likelihood trees were con- structed individually on unique 16S haplotypes and S7 alleles using PAUP* (Swofford 1998). Sequences ob- tained from Elassoma zonatum were used as an outgroups taxon (sensu Quattro et al. 2001), and two individuals representing the breadth of sequence varia- tion in E. evergladei (TJK unpublished data) were in- cluded to assess the sister group relationship between the two “okefenokee” PO pore type groups. Phyloge- netic analyses were viewed as independent tests of spe- cies-level divergence between the four-pored and three- pored populations of “okefenokee”. Given minor vari- ability in pore counts within populations, we considered whether individuals with four PO pores, sampled from populations where four pores predominate, form mono- phyletic groups in two independent gene trees to the exclusion of those haplotypes or alleles sampled from individuals with three pores taken from populations where three pores predominate. Similarly, we ask whether genetic distinctions between these groups are consis- tent across their respective geographic ranges, espe- cially where they are found in proximity. Bayesian analyses were performed on both data matrices separately, but in all cases the Bayesian results were entirely consistent with both the parsimony and likelihood analyses except for the placement of various terminal nodes that have no bearing on relationships be- tween the three- and four-pored forms. For simplicity, only the parsimony and likelihood results are reported here. For likelihood trees, models of DNA evolution were selected using likelihood ratio tests as implemented in MODELTEST (Posada & Crandall 1998, Version 3.7). Bootstrapping (Felsenstein 1985; 1,000 pseudoreplicates) was used to gauge support for nodes of interest, in particular monophyly of the three- and four- pored forms in both gene trees and a sister group relationship between the two types of “okefenokee”. Elassoma gilberti n. sp. (Fig. 1) Diagnosis.– Elassoma gilberti is distinguished from its close relative E. okefenokee by possessing four pores in the preopercular (PO) canal on each side of the head and usually seven anal fin rays. Elassoma okefenokee has three PO pores and usually eight anal fin rays. Elassoma gilberti has slightly less deep body and slightly smaller dorsal and anal fins than E. okefenokee. Breed- ing females of E. gilberti often have blue dashes below and behind the eye, which are lacking in female E. okefenokee. Otherwise, the two species are almost identical or broadly overlapping in meristic, morphomet- ric, and color features. The distinction of the two spe- cies is supported by molecular data. Eight fixed differ- ences (of 553 bp assayed) at the mitochondrial 16S rRNA locus and 12 fixed differences (of 583 bp assayed) found at the nuclear S7 locus differentiated the two species. Phylogenetic analyses using these molecular characters supported monophyletic clades that contained haplotypes and alleles found uniquely in E. gilberti and E. okefenokee, respectively. Type Material. – All type material is located at the Florida Museum of Natural History (UF). Holotype: UF 173591 (FFS 07-51, tag # 1M); adult male 25.2 mm SL; Econfina drainage, Florida, Taylor County, Econfina River at US Hwy. 27&19 bridge 12.0 road miles NW of jct. with US Hwy. 98W in Perry; 30° 15.09’ N, 83° 42.06’ 122 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) W; 8 April 2007; F. F. Snelson, Jr. Allotype: UF 173592 (FFS 07-51, tag # 14F); adult female 23.1 mm SL; col- lected with the holotype. Paratypes: UF 173593, 29 specimens collected with the holotype. UF 173595 (FFS 06-34); 17 speci- mens (3 of which were fixed and stored in 95% ETOH and were used for DNA tissue samples); collected at the type locality; 25 April 2006; F. F. Snelson, Jr. UF 173606 (FFS 05-25); 13 specimens (5 of which were fixed and stored in 95% ETOH and were used for DNA tissue samples); collected at the type locality; 24 March 2005; F. F. Snelson, Jr. UF 173594 (FFS 92-02); 15 specimens collected at the type locality; 8 April 1992; F. F. Snelson, Jr. Etymology. – The new species in named in honor of Dr. Carter R. Gilbert, Curator of Fishes at the Florida Museum of Natural History from 1961-1998 and now Curator Emeritus. This name will stand in recognition of the many contributions Dr. Gilbert has made to the study of North American fishes and as special thanks from FFS for serving as a guide and mentor for many years. The suggested common name is Gulf Coast Pygmy Sunfish, in view of its distribution only in drain- ages that empty into the Gulf of Mexico. Description. – Average adult body size 22.1 mm SL for males, 22.0 mm SL for females. The largest male measured was 25.5 mm SL, the largest female 26.3 mm SL. The general body shape and appearance are shown in Figure 1. Body laterally compressed, with greatest depth at dorsal fin origin. The head is moder- ately compressed. The anterior profile is narrowly rounded; the mouth is terminal, with the lips projecting slightly beyond the snout tip. All fins are broadly rounded in posterior profile except the pelvics, which are pointed. Proportional measurements are presented in Table 1. Scale and fin-ray counts are presented in Tables 2 and 3. Lateral scale rows 27-32, usually 28-31. No pored lateral line scales. Transverse scale rows 13-19, usually 14-17. Caudal peduncle scale rows 14-20, usu- ally 16-19. Body fully clad in thin, partially embedded cycloid scales. Top of head anterior to nape naked. Dorsal fin spines 3-5, the first spine often short and par- tially embedded. Dorsal fin rays 9-13, usually 10-12. Anal fin spines always 3. Anal fin rays 6-9, usually 7 or Figure 1. Elassoma gilberti. (A) Breeding male 25.2 mm SL. (B) Breeding female 22.7 mm SL. UF 173607. Suwannee River drainage, FL, Dixie County, backwater off Suwannee River at Fanning Springs; 25 April 2006. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 123 Elassoma gilberti Elassoma okefenokee Holotype Allotype Paratypes Male Female Males Females Males Females UF173591 UF173592 n=11 n=10 n=10 n=10 Standard Length 25.2 23.1 21.0 21.2 22.5 21.5 19.7-23.2 18.0-26.3 20.2-23.8 19.6-23.6 Body Depth 298 277 293 294 318 314 269-310 261-325 312-327 306-327 Predorsal Length 444 441 441 442 428 429 427-461 422-455 412-447 407-453 Prepelvic Length 353 355 364 364 372 365 347-379 345-375 346-395 352-379 Preanal Length 579 597 577 605 565 594 564-599 589-626 547-580 576-608 Dorsal Fin Length 496 420 477 436 506 456 451-504 408-453 481-521 439-470 Anal Fin Length 329 273 329 277 348 288 304-357 266-298 335-376 276-304 Pectoral Fin Length 151 147 154 142 157 144 129-167 134-151 146-168 137-152 Pelvic Fin Length 266 225 263 233 270 240 234-287 217-249 254-282 225-260 Caudal Peduncle 258 251 260 245 256 242 Length 244-276 222-277 243-282 208-262 Caudal Peduncle 135 125 135 129 135 130 Depth 123-147 118-140 125-146 124-138 Head Length 333 351 341 341 350 346 326-364 327-362 325-371 332-364 Head Depth 214 190 216 198 214 202 209-226 185-214 200-229 186-217 Snout Length 95 95 87 84 85 83 78-96 78-95 74-98 78-88 Eye Diameter 99 87 98 100 99 101 91-106 94-109 94-104 93-105 Upper Jaw Length 99 95 100 93 105 102 91-110 89-100 98-112 97-111 Table 1. Proportional measurements of Elassoma gilberti and E. okefenokee. Standard length (SL) is in millime- ters. All other measurements are thousandths of SL with the mean given above and the range below. 124 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) 8. Pectoral fin rays 14-18, usually 15 or 16. Pelvic fin always with 1 spine and 5 rays. Branched caudal fin rays 10-13, usually 11 or 12. The canals and pores of the cephalic lateralis sys- tem are similar to the pattern described for Elassoma zonatum by Branson and Moore (1962). The anterior nasal pore of the supraorbital canal (SO) is just anterior and medial to the anterior nare opening. The posterior nasal pore is just medial and posterior to the posterior nare opening, positioned on the rim of or slightly inside the narial depression. Over the center of the eye, the SO canal gives off a short, medially directed canal that terminates in a single pore. This represents the su- praorbital commissure of Branson and Moore (1962), but it is not a true commissure since it does not meet the same canal on the opposite side of the head. The SO canal then curves downward to meet the postocular commissure (POC) at a large pore just behind the upper quadrant of the eye. Posteriorly there is a pore at the point where the POC meets the posttemporal (PT) ca- nal. At this point, the short supratemporal canal branches off dorsally and medially with a single pore at its termi- nus. Further posteriorly, the PT ends in a pore just ante- rior to the upper corner of the opercular opening. The preopercular (PO) canal (POM of Branson & Moore 1962) does not join the POC. It usually has four pores, one at the upper end of the preopercle, one at or just dorsal to the broad curvature of the preopercle where the opercle and subopercle bones meet, one just below the angle of the preopercle where the curvature begins to straighten, and one at the anterior terminus of the preopercle, almost directly below the middle of the eye (Fig. 2). Geographic variation in PO pore number is shown in Table 4 and is discussed later. The PO canal does not extend onto the mandible. The infraorbital ca- nal is greatly reduced, limited to a short tube in the lach- rymal region with two pores, one pore posterior and lat- eral to the anterior nare, the other just below and ante- rior to the anterior edge of the eye near the border of the upper lip. Lateral Scale Rows 26 27 28 29 30 31 32 33 N Mean E. gilberti West 3 7 14 16 12 4 56 29.7 South 3 13 23 14 5 3 61 29.2 E. okefenokee 1 3 15 11 14 9 3 4 60 29.6 Transverse Scale Rows 13 14 15 16 17 18 19 N Mean E. gilberti West 10 17 14 13 2 1 57 15.7 South 3 11 24 19 6 63 15.2 E. okefenokee 2 10 31 19 7 1 70 15.3 Caudal Peduncle Scales 14 15 16 17 18 19 20 N Mean E. gilberti West 1 13 15 19 5 3 56 17.4 South 1 1 10 19 20 9 3 63 17.5 E. okefenokee 2 20 20 16 9 3 70 17.3 Table 2. Scale counts of Elassoma gilberti and E. okefenokee. For E. gilberti, the count of the holotype is bolded, the count of the allotype is underlined. For E. gilberti, west is Choctawhatchee Bay east through the Wakulla drainage; south is the Econfina south through the lower Suwannee drainage. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 125 Ta bl e 3. F in r ay c ou nt s of E la ss om a gi lb er ti an d E . o ke fe no ke e. Fo r E . g il be rt i, th e co un t o f th e ho lo ty pe is b ol de d, th e co un t o f th e al lo ty pe is u nd er lin ed . Fo r E . g il be rt i, w es t i s C ho ct aw ha tc he e B ay e as t t hr ou gh th e W ak ul la d ra in ag e; s ou th is th e E co nf in a so ut h th ro ug h th e lo w er S uw an ne e d ra in ag e. D or sa l S pi ne s D or sa l R ay s A na l R ay s 3 4 5 N M ea n 9 10 11 12 13 N M ea n 6 7 8 9 N M ea n E . gi lb er ti W es t 4 48 5 57 4. 0 1 4 41 9 1 56 11 .1 3 36 17 1 57 7. 3 So ut h 8 5 2 8 68 4. 0 10 3 8 18 1 67 11 .1 3 4 4 20 67 7. 3 E . ok ef en ok ee 1 54 15 70 4. 2 1 32 32 4 69 11 .6 22 42 6 70 7. 8 P ec to ra l R ay s B ra nc he d C au da l R ay s 14 15 16 17 18 N M ea n 9 10 11 12 13 N M ea n E . gi lb er ti W es t 5 31 17 3 56 15 .3 7 22 25 54 11 .3 So ut h 13 30 1 9 4 1 67 15 .3 3 2 8 31 1 63 11 .5 E . ok ef en ok ee 5 35 22 3 1 66 15 .4 1 6 29 32 68 11 .4 126 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) Breeding Colors, Males. – In breeding males, the background color of the body ranges from dark gray to sooty black. The sides of the body are marked with 5-8 narrow iridescent blue bars. The first bar is anterior to the caudal base at about the midpoint of the caudal pe- duncle; bars extend forward a varying distance to mid- body. The blue bars are about half the width of the intervening black spaces. The blue bars are best devel- oped on the caudal peduncle, where they are slightly oblique and often extend nearly from the dorsal to the ventral midline. Anteriorly, the bars are less well devel- oped, confined mostly to the flanks, and usually broken into separate blue spots or dashes. The head is marked by two bright iridescent blue dashes or crescents on the border of the orbit; one lies behind the eye, centered slightly below a horizontal through the middle of the eye; the other lies below the eye, centered slightly in front of a vertical through the center of the eye. There is a narrow gap between the two blue eye dashes. The light band or “racing stripe” (see below) across the snout tip and the lips is usually obliterated by the overall dark pig- mentation of the head. The distal third of the dorsal and anal fins have an iridescent powder-blue wash or band that is bordered by a narrower black band. The background color in the basal half of these fins is sooty black with 1-2 (anal fin) or 2-3 (dorsal fin) irregular rows of bright blue spots, which are best defined in the posterior part of the fins. The caudal fin is dark blue to black basally with bright iridescent blue wash or band distally. It is usually bor- dered terminally by a narrow black border, but this may be lacking in some specimens. Two unpigmented win- dow-like or vertically elongated spots over the caudal fin base range from beige to pale blue. The spine and the first 2-3 rays and intervening membranes of the pel- vic fins are bright iridescent blue; the remainder of the fin is sooty gray or black. The pectoral fins are clear. The blue colors on the body and fins fade rapidly after death. Breeding Colors, Females. – The body colors of breeding females are much as described below for pre- served material, being a combination of tans and browns. The overall intensity of the pigmentation can vary from pallid to dark depending on the color of the water and the nature of the habitat from which they are taken. There is no blue color on the body or fins. The dashes behind and below the eye, described for males, are of- ten present in females, but are less striking than in males and are iridescent blue-green, rather than powder blue in color. Their presence/absence may be a function of readiness to spawn. A concentration of black pigment around the vent is usually conspicuous, and a light band across the snout tip and the dark spotting in the dorsal and anal fins is usually evident on close inspection. Color in Preservation, Males. – After preserva- tion, life colors fade quickly and reveal the general pat- tern shown in Figure 1A. Two unpigmented windows, vertically-elongated or semicircular in shape, are present at the caudal base, one above, one below the midline. A darker area separates the two windows at the horizon- tal midline and they are bordered posteriorly by black pigment. These windows range from very obvious to rather inconspicuous. Anterior to the basicaudal win- dows, there is usually a dark rectangular or crescent- shaped bar or blotch, bordered anteriorly by a lightly pigmented bar or blotch. The dark blotch may not ex- tend to the dorsal and ventral midline. The light bar typically does extend to the dorsal and ventral midline, encircling the caudal peduncle. Often this light area expands back as far as the upper and lower procurrent Figure 2. Patterns of preopercular (PO) pores in (A) Elassoma gilberti with four pores and (B) E. okefenokee with three pores. Pattern C, with 2+2 pores, is found is a small percentage of E. okefenokee, primarily in the St. Johns drainage. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 127 3-3 3-4 4-3 4-4 2+2-3 3-2+2 2+2-2+2 5-4 4-2 N E. gilberti Choctawhatchee 2 2 140 - - - 1 145 Apalachicola 1 - 1 85 - - - 1 88 New 14 14 Ochlockonee 1 - - 78 79 St. Marks 1 1 104 106 Aucilla 1 - 1 87 89 Econfina 6 4 93 103 Fenholloway 16 16 Spring Warrior 10 10 Steinhatchee 1 - 46 47 California Creek 3 3 Suwannee 1 4 3 208 216 Waccasassa 1 12 - - - 1 14 Homosassa 1 1 Subtotal 4 14 13 897 - - - 3 931 E. okefenokee Altamaha 1 1 Satilla 30 1 31 St. Marys 66 2 68 St. Johns 307 7 6 1 5 7 2 - 1 336 Suwannee 241 6 4 3 2 256 Withlacoochee 32 - 1 - - 2 1 36 Hillsborough 9 9 Kissimmee 23 - - - 2 25 Subtotal 708 16 11 4 9 9 4 - 1 762 Grand Total 1693 Table 4. Preopercular (PO) pore counts by drainage system for Elassoma gilberti and E. okefenokee. The counts of the holotype and allotype are bolded. PO Pores 128 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) caudal rays, resulting in a distinctly light spot in the area, especially ventrally. Anterior to the caudal peduncle, the alternating pattern of dark and light bars or blotches becomes variable from specimen to specimen. In some, only one additional band pair is evident. In others, band- ing extends as far forward as mid-body, with up to 5-6 additional band/blotch pairs. Banding becomes progres- sively more obscure anteriorly, often breaking up into a mottled pattern. Pigment is sparser on the venter, breast, and lower half of the head. There is a dark ring of pig- ment surrounding the vent, most conspicuous in speci- mens with overall reduced body pigmentation. The tip of the urogenital papilla is blackened. The median and pelvic fins range from dusky to black, depending on the reproductive condition of speci- mens at the time of collection. In the dorsal fin, pigment is uniformly distributed on the membranes and the spines/ rays. In slightly faded specimens the distal portion of the last 2-4 dorsal rays are darker than the remainder of the fin, resulting in a dark blotch in this area. In faded specimens or in individuals with less intensely pigmented dorsal fins, a series of 2-3 rows of black spots is present on the basal third of the fin, the spots centered over the spines/rays. The tips of the dorsal spines are slightly depigmented compared to the reminder of the fin. The pectoral fin rays are outlined with melanophores but the membranes are clear. A few dark spots, as in the dorsal fin, may also be present in the basal half of the anal fin. The preorbital region, including the area around the nares and the lateral aspects of the upper and lower lips, is more heavily pigmented than the midline of the head, snout, lips, and chin tip. This results in a broad, pale mid- sagittal “racing stripe” across the top of the snout and lips (see Boschung & Mayden 2004:613 for an illustra- tion). This pigment feature may range from conspicuous to faint. It is generally more pronounced in specimens that were not in breeding condition at the time of cap- ture. There may be a faintly defined postorbital stripe to the edge of the operculum. There is no suborbital bar. Breeding Colors, Females. – Females have overall pigmentation much reduced compared to males (Fig. 1B). The unpigmented basicaudal windows are usually less conspicuous in females than in males owing to the less heavily pigmented caudal fin and paler body pigmenta- tion. The alternating light and dark bars or blotches on the posterior trunk and caudal peduncle are usually more prominent in females than in males. From mid-body for- ward, the body pigmentation is usually a mottled or marbled pattern. Often the dark bars/blotches on the caudal third of the body are intensified ventrally, result- ing in a series of dark spots or sub-rectangular blotches along the base of the anal fin. Less frequently, similar spots may also be present along the base of the soft dorsal fin. The undersides of the head, the flanks, and the venter are less heavily pigmented than dorsal as- pects of the body, but there is always an intensification of black pigment around the vent. The mid-sagittal “racing stripe” pattern across the snout tip is typically more conspicuous in females than in males. As in males, there may be a faintly defined post-orbital stripe but there is no suborbital bar. The soft dorsal and anal fins are marked with 2-3 rows of black spots centered over the rays. The inter- radial membranes are lightly stippled with melanophores, sometimes becoming slightly darker distally. The cau- dal fin rays are outlined with melanophores but pigment is faint and scattered over the membranes; occasionally the basal third of the fin may be faintly spotted. The pelvic fins may be clear or may have a few scattered melanophores laterally in the basal third of the fin. The pectoral fins have the rays narrowly outlined with black, but the membranes are immaculate. Sexual Dimorphism. – Sexual differences in col- oration and pigmentation are described above. There are no sexual differences in meristic characters, but di- morphism in fin size is conspicuous. Males have much longer dorsal, anal, and pelvic fins than females, with little or no overlap in proportional size (Table 1). The differences in fin size are most pronounced during the breeding season (March through early May), but are evident in adults at all times of the year. Females also have proportionately greater preanal length than males (Table 1). MORPHOLOGICAL COMPARISON OF ELASSOMA GILBERTI AND E. OKEFENOKEE The characters distinguishing Elassoma okefenokee from all other species of Elassoma known at that time are presented by Mayden (1993). Those same features will also serve to distinguish E. gilberti from all other species in the genus except for E. okefenokee. The colors of breeding males in these two species show no consistent differences. Variation within a spe- cies is evident depending on site-to-site habitat condi- tions such as water clarity and color. Individual varia- tion within a collection seems to be related to reproduc- tive readiness of individual males. The only color differ- ence noted was in breeding females. Blue-green dashes behind and below the eye (see Description) were usu- ally present in heavily gravid female E. gilberti. It ap- peared that specimens in which the dashes were indis- tinct or absent were not in peak reproductive condition. These dashes were never observed in breeding females of E. okefenokee. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 129 To the naked eye, specimens of E. okefenokee appear to be slightly deeper bodied than E. gilberti. This appearance is confirmed by measurements showing that both males and females of E. okefenokee have greater body depth (Table 1). Further, both males and females of E. okefenokee have more expansive dorsal and anal fins (Table 1), which contribute to the appearance of greater body depth. Otherwise, the two species are remarkably similar in their general appearance. Scale counts of the two species are broadly over- lapping and there are no consistent geographic trends within E. gilberti (Table 2). Likewise, fin ray counts are very similar and broadly overlapping (Table 3). For dorsal fin rays, 64% (79 of 123) of the E. gilberti speci- mens examined had a count of 11, but E. okefenokee was equally likely to have 11 or 12. Anal fin ray counts showed the most differentiation: 65% (80 of 124) of E. gilberti counted had seven rays whereas 60% (42 of 70) of E. okefenokee had eight rays (Table 3). The PO pore count is the only reliable morphologi- cal character that distinguishes E. gilberti from E. okefenokee. Because there is only a one pore differ- ence between the two species, absolute identification of single specimens may not be possible on this character alone. However, if several specimens from a site are available, preferably five or more, positive identification of the series is always possible based on the average PO count for the series. Excluding the Suwannee drainage, where both spe- cies occur, only 3 of 715 (0.4%) specimens of E. gilberti examined had a PO count of 3-3, which is the charac- teristic count of E. okefenokee (Table 4). In the Apalachicola drainage, 96.6% of the specimens exam- ined had a count of 4-4 (Table 4). However, a single specimen in a lot of one (UF 105552) has a count of 3- 3. In the Aucilla drainage material, a single specimen in a lot of one (UF 75283) had count of 3-3. Overall, 97.7% of specimens examined from the Aucilla drainage had a count of 4-4. Among the material from the Ochlockonee drainage, a single specimen in a lot of 34 (UF 50309) has the odd count of 3-3; overall, 98.7% of the speci- mens examined from the Ochlockonee had the typical count of 4-4. Likewise, with the exclusion of the Suwannee ma- terial, only 1 of 506 specimens of E. okefenokee has a PO count of 4-4, characteristic of E. gilberti (Table 4). That specimen, from the St. Johns drainage, was in a lot of 13 (UF 40); the other specimens in the lot had PO counts of 4-3 (in 1) and 4-4 (in 11). With the exclusion of the Suwannee drainage ma- terial, 37 of 1221 specimens (3.0%) examined had a bilaterally asymmetrical PO pore count of 4-3 or 3-4 (Table 4). By themselves, such specimens would be equivocal, but in most cases they were from lots with multiple specimens where the predominant count (> 90%) was the typical count for the species. In only seven small series of E. gilberti did the number of specimens with asymmetrical PO counts exceed 10% of the counts for the series: UF 5851 (1 specimen in a lot of 1), WTLC BA150-97 (1 of 1), GMNH 349 (1 of 1), UF 91808 (1 of 2), UF 95958 (2 of 8), UF 95959 (1 of 5), and UF 173606 (4 of 13). Among the series of E. okefenokee exam- ined, specimens with 4-3 or 3-4 PO pores were found to exceed 10% in only eight lots: GMNH 1474 (1 of 4 specimens), UF 56514 (1 of 5), GMNH 1149 (1 of 5), UF 2496 (1 of 5), UF 5852 (1 of 5), UF 5870 (1 of 7), UF 22905 (2 of 5), and UF 173638 (2 of 13). The Econfina drainage material had the highest percentage of 3-4 or 4-3 PO counts among all drainages (9.7%, Table 4). Several other PO pore count patterns were exhib- ited by rare specimens of both species (Table 4). The only consistent pattern was a count of 2+2 on one or both sides of the head. In these cases, the PO canal was interrupted at the angle of the preopercle leaving short tubes in the vertical and horizontal limbs of the bone, each with a pore at either end (Fig. 2C). This pattern was never observed in E. gilberti, but was found in 20 of 506 (4.0%) specimens of E. okefenokee from outside the Suwannee basin, predominately in the St. Johns River system (Table 4). The 2+2 pattern was usually found on only one side of the head. It was bilat- erally symmetrical in only 4 of 762 (0.5%) of E. okefenokee specimens examined. Because both species occur in the Suwannee drain- age, their PO pore count pattern is examined in more detail (Table 5). In the lower Suwannee basin, a single specimen in a lot of 23 (UF 173607) has a count of 3-3. Specimens with asymmetrical pore counts were found in two lots: UF 173608 (2 of 20 specimens) and UF 90972 (5 of 123). All remaining specimens from the lower Suwannee (194 of 202, 96.0%) had the PO count of 4-4 typical for E. gilberti elsewhere in its range. In contrast, the upper Suwannee River drainage in Geor- gia is occupied exclusively by E. okefenokee; 83 of 85 specimens examined have the typical PO pore count of 3-3. Only two specimens had asymmetrical counts (GMNH 1312a, 1 of 11 specimens; and GMNH 2049, 1 of 12). The Santa Fe system in Florida, a tributary to the Suwannee River, is clearly occupied by E. okefenokee as indicated by the predominance of the PO count of 3- 3 (Table 5). There are only five asymmetric 3-4 or 4-3 counts among the specimens examined (UF 7645, 3 of 130 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) 28 specimens; UF 25535, 1 of 28; and UF 101509, 1 of 10). Out of 127 specimens examined from the Santa Fe, only two have a count of 4-4, which is the typical count for E. gilberti (UF 7645, 1 of 28 specimens, and UMMZ 210073, 1 of 1). Only seven specimens from six lots are available from the Ichetucknee Spring and spring run. This small tributary is of interest because it is the most downstream population in the Santa Fe River system before the lat- ter joins the Suwannee River proper. The majority of these specimens (4 of 7) have the count of 3-3, typical of E. okefenokee in the remainder of the Santa Fe sys- tem. Two specimens, each in singleton lots, have the asymmetric count of 3-4 (UF 123566 and UF 126467). One specimen in a singleton lot (UF 4984) has a count of 4-4, which is the typical count for E. gilberti. Only three samples are available from the middle portion of the Suwannee drainage between the mouth of the Santa Fe and the state border to the north. The most downstream site is Allen Mill Pond Spring (UF 30238; Fig. 7, site 4). This series is identified as E. gilberti based on PO counts of 4-4 in nine specimens (Table 5). The next most upstream of the three samples is a small series from White Springs in Hamilton County (UF 4646; Fig. 7, site 5). All five specimens have PO counts of 4-4, consistent with E. gilberti. The third and most upstream of the three middle Suwannee sites is Robinson Branch, a direct tributary to the Suwannee River (UF 173634; Fig. 7, site 6). This sample has PO counts (Table 5) and DNA sequences consistent with E. okefenokee. Only two specimens in the lot of 37 from Robinson Branch have an asymmetrical pore count. We have assumed in all cases that specimens with an odd PO count from a large lot, especially in cases where there is a large sample size from the drainage, are simply atypical individuals of the “expected” spe- cies, not the alternate species. In the case of singletons or small lots, it would be impossible to confidently assign odd specimens without matching DNA sequence data. The situation in the Ichetucknee Spring system, with few specimens in small lots, is problematic in this regard. The co-occurrence there of both species could not be ruled out with the available data. MOLECULAR ANALYSIS The locations from which specimens were sampled for DNA sequence data are shown in Figure 3 and specific collection data are given in DNA Materials Examined. RESULTS 16S. – Sixty-two individuals of “okefenokee” were sequenced for 16S. Thirty specimens had 4-4 PO pores PO Pore Count N 3-3 3-4 4-3 4-4 2+2-3 Lower Suwannee E. gilberti 1 4 3 194 202 Santa Fe E. okefenokee 119 3 2 2 1 127 Ichnetucknee E. okefenokee 4 2 - 1 7 Middle Suwannee E. gilberti 14 14 E. okefenokee 35 - 1 - 1 37 Upper Suwannee E. okefenokee 83 1 1 85 Table 5. Preopercular (PO) pore counts for Elassoma gilberti and E. okefenokee in the Suwannee River drainage of Florida and Georgia. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 131 and were from four-pore populations (E. gilberti) and 32 had 3-3 PO pores and were from three-pore popula- tions (E. okefenokee) (Table 6). In addition, we se- quenced two E. evergladei from divergent locations (western and eastern Florida) and one E. zonatum from Florida. Twenty polymorphic sites (3.75% of 553 as- sayed bases) were found within E. gilberti and E. okefenokee, and these variable positions defined 14 16S haplotypes (Table 7). Diversity was evenly distributed between these two taxa; six polymorphic sites (five tran- sitions, one transversion) defined seven haplotypes in E. okefenokee, while six polymorphic sites (all transitions) defined seven haplotypes in E. gilberti. Eight sites (six transitions, two transversions; 1.50% uncorrected se- quence divergence) exhibited fixed differences between the two species. S7. – Forty-three individuals of “okefenokee” were sequenced for the nuclear S7 locus, 20 specimens with 4-4 PO pores from four-pore populations (E. gilberti) and 23 with 3-3 PO pores from three-pore populations (E. okefenokee) (Table 6), plus two E. evergladei from two divergent locations and one E. zonatum. This re- sulted in 86 sampled alleles excluding the outgroup (Table 8). Of 583 base positions surveyed, 15 polymorphic sites defined five S7 alleles in E. gilberti and E. okefenokee. One polymorphic site (one transversion) defined two S7 alleles in E. okefenokee, while two polymorphic sites (one transition, one transversion) defined three haplotypes in E. gilberti. Twelve sites (three transitions, nine transversions; 2.06% uncorrected sequence divergence) exhibited fixed differences between the two species. Heterozygotes were evident in four populations, one population of E. okefenokee (Santa Fe) and three populations of E. gilberti (Econfina, St. Marks, Ochlockonee). Small sample sizes preclude reasonably powerful tests of Hardy-Weinberg equilibrium, although the distribution of variation within individual populations did not suggest any systematic bias in the frequency of Figure 3. The distribution of samples used in the DNA analysis of Elassoma gilberti (Eg1 – Eg6) and E. okefenokee (Eo1 – Eo10). Specific collection information can be found in the DNA Material Examined section. 132 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) Drainage Abbrev. 16S S7 St. Johns StJ 13 8 Hillsborough Hil 4 2 Santa Fe SaF 9 7 Suwannee Suw 9 9 Steinhatchee Ste 3 1 Econfina Eco 6 4 Aucilla Auc 3 1 St. Marks StM 9 5 Ochlockonee Och 6 6 Total 62 43 genotypes within any individual population. Importantly, although two divergent alleles were sampled, no het- erozygotes composed of these two divergent alleles were detected at the S7 locus in the Suwannee River. PHYLOGENETIC ANALYSIS 16S. – Parsimony analyses on the 14 haplotypes surveyed from E. gilberti and E. okefenokee, two haplotypes from E. evergladei and the outgroup E. zonatum recovered a single tree (length = 66, consis- tency index (CI) = 0.939, retention index (RI) = 0.956) containing three reciprocally monophyletic clades. One of these clades included haplotypes sampled in E. evergladei, while the remaining two clades contained seven haplotypes surveyed from E. gilberti and seven from E. okefenokee (Fig. 4). A sister group relation- ship between E. gilberti and E. okefenokee was sup- ported strongly by bootstrap analysis, as was a sister group relationship between the E. gilberti + E. okefenokee clade and haplotypes surveyed in E. evergladei. Maximum likelihood analyses (best-fit model = K80 from MODELTEST) on these same data recovered a single topology (-Ln likelihood = 1174.646) that was entirely consistent with the shortest tree re- covered from the parsimony analysis. Bootstrap sup- port for these relationships was likewise very strong. S7. – Parsimony analysis on the eight (including E. evergladei and E. zonatum) unique S7 alleles re- covered a single tree (length = 85, CI = 0.988, RI = 0.976) that contained three reciprocally monophyletic clades comprising two alleles sampled in E. gilberti, three alleles sampled in E. okefenokee, and two alleles in E. evergladei (Fig. 5). As in the 16S trees, E. evergladei is recovered as the sister group to the E. gilberti + E. okefenokee clade. Maximum likelihood analyses (best-fit model = K81 from MODELTEST) on these same data recovered a single topology (-Ln likelihood = 1254.102) that was entirely consistent with the shortest tree recovered from the parsimony analy- sis. Monophyly of the three clades was strongly sup- ported by bootstrap analysis, as was the sister group relationship between alleles sampled from E. gilberti and E. okefenokee. DISTRIBUTION Elassoma gilberti is found in stream systems draining into the Gulf of Mexico from the panhandle of Florida and extreme southwestern Georgia south through the western portion of the north-central Florida peninsula (Fig. 6). The western-most drainages occupied in the panhandle are the Choctawhatchee and several smaller stream systems that empty into Choctawhatchee Bay. The species is common in the Florida portion of the Choctawhatchee system but is not yet known from Ala- bama (Boschung and Mayden 2004). In the Apalachicola drainage, the species is found in tributaries to both the Chipola and Apalachicola systems, with a few records from extreme southwestern Georgia. The species is found in the small Whiskey Creek and New drainages to the east of the Apalachicola and is common in the Florida portions of the Ochlockonee drainage further east. Böhlke and Rohde (1980) plotted a record from the Georgia portion of the Ochlockonee drainage, but supporting specimens have not been located. The spe- cies was not reported from the Georgia portion of the Ochlockonee drainage by Swift et al. (1977). East and south of the Ochlockonee, the species is found in all major and some minor drainages south through the Suwannee (discussed in more detail below). The next Gulf drainage south of the Suwannee is the Waccasassa; although material is very limited from this drainage, the PO count of 4-4 in 12 of 14 specimens is consistent with E. gilberti (Table 4). The next Gulf drainage south of the Waccasassa is the Withlacoochee; PO counts from this drainage clearly identify this population as E. okefenokee (Table 4). A single specimen is known from Homosassa Springs Run in Citrus County, south Table 6. Number of individuals of Elassoma gilberti and E. okefenokee sampled for 16S and S7. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 133 Nucleotide Position 1 1 2 2 2 2 2 2 2 2 2 3 3 3 3 3 4 4 Drainage 8 9 1 6 0 1 3 3 5 6 6 6 7 4 5 7 7 9 1 9 8 2 5 7 2 8 4 8 6 2 6 9 1 9 2 7 8 8 5 2 StJ Hil SaF Suw Ste Eco Auc StM Och Eo16S1 GAGGAT C C AG GT T T C C GAT A 3 7 6 Eo16S2 . . . A. . . . . . . . . . . . . . . . 1 Eo16S3 . . . . . . . . . . A. . . . . . . . . 1 Eo16S4 . . . . . . . . . A AA. . . . . . . . 2 Eo16S5 . . . . . . . . . A . A. . . . . . . . 3 4 Eo16S6 . . A . . . . . A . A. . . . . . . . 1 Eo16S7 . . . . . . . T . A . A. . . . . . . . 4 Eg16S1 . . . . GC A. . A . . A. T . AGC G 3 2 3 3 7 Eg16S2 A. . . GC A. . A . . A. T . AGC G 1 Eg16S3 . . . . GC A. . A . . A. T . AGC . 6 Eg16S4 . . . . GC A. . A . . A. T T AGC G 1 Eg16S5 . . . . GC A. GA . . A. T . AGC G 1 Eg16S6 . G. . GC A. . A . . A. T . AGC G 2 Eg16S7 . G. . GC A. . A . . AC T . AGC G 1 Table 7. Nucleotide diversity at the 16S mtDNA locus uncovered in Elassoma okefenokee (Eo) and E. gilberti (Eg). Twenty variable positions define 14 unique haplotypes. Counts are the number of haplotypes (number of individuals) surveyed per population. Periods within the sequence matrix indicate identity to the first haplotype. Numbers refer to specific nucleotide positions in the alignment. Drainage abbreviations can be found in the Table 6. Nucleotide Position 1 3 3 3 3 3 4 4 4 4 5 5 Drainage 4 5 9 6 0 1 4 7 9 3 4 7 7 1 2 6 4 1 2 0 9 9 4 8 0 2 5 9 1 1 StJ Hil SaF Suw Ste Eco Auc StM Och EoS71 C A A C A A G T A T G C A A G 16 4 13 12 EoS72 A . . . . . . . . . . . . . . 1 EgS71 . C G . C T T A C G A A C G T 6 6 2 7 EgS72 . C G . C T T A C G A . C G T 2 2 3 2 EgS73 . C G T C T T A C G A . C G T 10 Table 8. Nucleotide diversity at the S7 nuclear locus uncovered in Elassoma okefenokee (Eo) and E. gilberti (Eg). Fourteen variable positions define five unique haplotypes. Counts are the number of alleles (twice the number of individuals surveyed when added across columns) surveyed per population. Periods within the sequence matrix indicate identity to the first haplotype. Numbers refer to specific nucleotide positions in the alignment. Drainage abbreviations are given in Table 6. 134 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) Figure 4. Phylogenetic relationships among 16S mtDNA haplotypes observed in Elassoma gilberti, E. okefenokee, and E. evergladei. Shown is the shortest topology recovered under the parsimony criterion. Bootstrap support (parsimony on top, maximum likelihood on bottom) is indicated for individual branches. For clarity, only values greater than 50% are shown. Although rooted with a sequence from E. zonatum, this taxon is not shown, but its position is implied from the root as drawn. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 135 of the Withlacoochee drainage. The PO count is 4-4, consistent with E. gilberti. Further south, the last Gulf drainage occupied by either species is the Hillsborough, and it is occupied by E. okefenokee (Table 4). Elassoma okefenokee is distributed from south- eastern Georgia south through much of the Florida pen- insula (Fig. 6). The occurrence of the species in the Altamaha drainage in Georgia is based on a single speci- men (GMNH 1073). The only other record of E. okefenokee from the Altamaha drainage, mapped by Böhlke and Rohde (1980), is based on a misidentified series of 19 E. evergladei (GMNH 1023). South of the Altamaha basin, the species is found in the Satilla drainage in Georgia and the St. Marys and Nassau drain- ages in southeastern Georgia and northeastern Florida. The species also occupies the upper portions of the Suwannee drainage in Georgia and its major tributary, the Santa Fe system, in Florida (details below). Elassoma okefenokee occupies the interior lake basins in north-central Florida and populations are scattered in tributaries of the St. Johns River as far south as Or- lando. There are a few recent records from the Kissimmee River basin in south-central Florida, extend- ing from near Kissimmee to the north shore of Lake Okeechobee. These records suggest that E. okefenokee once may have been more widespread in the southern peninsula. As noted above, the species is also found in three drainages, the Suwannee, Withlacoochee, and Hillsborough, that empty into the Gulf of Mexico. The distribution of the two species in the Suwannee drainage is shown in Figure 7 and the supporting PO count data are given in Table 5. All material from the lower portion of the Suwannee drainage south of the confluence with the Santa Fe (Fig. 7, from site 2 south to site 1) is clearly consistent with E. gilberti both in PO counts and in the DNA analysis (Tables 7, 8). The Ichetucknee Spring system (Fig. 7, site 3) is the lower- most tributary to the Santa Fe River before the latter joins the Suwannee River. As noted earlier, only a few specimens are available from this small tributary. The spread of pore count data (Table 5) suggests there may be, or may have been, co-occurrence or hybridization of the two species in this system. However, based on the predominant PO count of 3-3, we assign this popu- lation to E. okefenokee. The remainder of the Santa Fe system upstream from Ichetucknee Springs is occu- pied by E. okefenokee, as indicated both by PO counts and the DNA analysis. The upper Suwannee River drainage basin above the Florida-Georgia border is oc- cupied exclusively by E. okefenokee, a conclusion based on both PO counts and DNA sequence analysis. In the Figure 5. Phylogenetic relationships among S7 nuclear alleles observed in Elassoma gilberti, E. okefenokee, and E. evergladei. Shown is the shortest topology re- covered under the parsimony criterion. Bootstrap sup- port (parsimony on top, maximum likelihood on bottom, 1000 pseudoreplicates) is indicated for individual branches. For clarity, only values greater than 50% are shown. Although rooted with a sequence from E. zonatum, this taxon is not shown, but its position is im- plied from the root as drawn. 136 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) middle Suwannee basin, the three lots available clearly separate into E. gilberti at the two more downstream locations, Allen Mill Pond (Fig. 7, site 4) and White Springs (Fig. 7, site 5), and into E. okefenokee at the more upstream location, Robinson Branch (Fig. 7, site 6). The Robinson Branch and White Springs localities are about six miles apart. In summary, the known distribution of the two spe- cies in the middle Suwannee drainage is as follows. Elassoma okefenokee extends down the Suwannee drainage proper to Robinson Branch. Elassoma gilberti is then found from White Springs south in the main Suwannee basin, with records scattered sporadically in small tributaries and springs as far south as the mouth of Gopher River (Fig. 7, site 1). The Santa Fe system is occupied by E. okefenokee both above and below the Figure 6. The overall distribution of Elassoma gilberti (dark circles) and E. okefenokee (dark squares) in Florida and southern Georgia based on material examined. The type locality (TL) for E. gilberti is indicated by a black triangle. SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 137 Figure 7. The distribution of Elassoma gilberti (dark circles) and E. okefenokee (dark squares) in the Suwannee River drainage in Florida and southern Georgia based on material examined. Specific sites referenced in the text are as follows: (1) mouth of Gopher River, (2) Guaranto Springs, (3) Ichetucknee Springs, (4) Allen Mill Pond, (5) White Springs, (6) Robinson Branch, and (7) Suwannee River at Fargo. 138 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) Santa Fe Sink and as far downstream as Ichetucknee Springs near the terminus of the Santa Fe. Elassoma gilberti occupies springs that are direct tributaries to the Suwannee River both above and below the mouth of the Santa Fe. Thus the ranges of the two species are in close proximity at two places in the middle Suwannee basin, around the Suwannee-Santa Fe confluence and in the Suwannee basin proper between White Springs and Robinson Branch. With the possible exception of Ichetucknee Spring, there is no indication that their ranges overlap or that they occur sympatrically. Additional material from the middle portions of the Suwannee basin and the lowermost portions of the Santa Fe system, preferably with supporting DNA sequence data, will be needed to refine the distributional pattern of these two species in the middle Suwannee basin. Un- fortunately, recent drought, land use changes, and modi- fications and dewatering of springs has rendered it im- possible to duplicate collections at many of the historic sites and difficult to locate new sites where the species might still occur. The only other sister species that exhibit a similar distribution pattern in Florida are Fundulus cingulatus and F. rubrifrons. Fundulus cingulatus is found pri- marily in the panhandle of Florida and in extreme south- ern Alabama and southwestern Georgia, west of the St. Marks drainage. In contrast, F. rubriforns is found in southeastern Georgia and northeastern Florida and at widely scattered sites south throughout the Florida pen- insula (Gilbert et al. 1992). Both species occur in widely separated areas of the Suwannee River basin. The only area where their populations are in proximity is in the Santa Fe River branch. Like E. gilberti and E. okefenokee, they have not been taken sympatrically. Other taxa, although not differentiated to the de- gree of morphologically diagnosable species, show a simi- lar genetic break across this same geographic area (e.g., Bermingham and Avise 1986). Our inclusion of two disparate samples representing sequence variation within E. evergladei likewise suggests substantial divergence across the Suwannee in other pygmy sunfishes. Given these common patterns in diverse taxa, it was expected that genetic divergence across the Suwannee should occur in “okefenokee” as well. However, we are un- aware of any substantial morphological differentiation that attends genetic differentiation across this common phylogeographic boundary in species other than Fun- dulus cingulatus/rubrifrons and Elassoma gilberti/ okefenokee. Indeed, the complete disequilibrium be- tween nuclear DNA, mitochondrial DNA, and morpho- logical characters within samples of E. okefenokee and E. gilberti from the Suwannee and the absence of het- erozygous individuals for diagnostic nuclear DNA alle- les are compelling support for assigning species-level status to E. gilberti. STATUS OF “ELASSOMA EVERGLADEI ORLANDICUM” LÖNNBERG Einer Lönnberg described this nominal form in 1894 un- der the text heading Elassoma evergladei Jordan. It was based on material collected from several localities in Orange, Osceola, and De Soto counties, Florida. The status of this name was first investigated by Dr. Reeve M. Bailey and Dr. James E. Böhlke (deceased). They examined the known syntypic material and produced a manuscript in 1978 dealing with the name and related nomenclatural issues. That manuscript was never pub- lished. The status of the name was commented on by Gilbert (1998) but its assignment was left unresolved because no lectotype was designated. Gilbert later (2004) considered E. evergladei orlandicum a nomen oblitum under provisions of Article 23.9.1 of the Inter- national Code of Zoological Nomenclature (1999), stat- ing that the name had not been used in the primary lit- erature since 1899. He failed to note that Barney and Anson used the name in 1920 (p. 242): “Dr. Einer Lönnberg has published notes on an Elassoma found at Orlando, Florida, and named provisionally by him E. orlandicum.” We feel that to avoid confusion, the sta- tus of this name is best resolved by the selection of a lectotype. With permission (R. M. Bailey, in litt. to FFS, June 2007), we paraphrase or quote directly from the draft Bailey and Böhlke manuscript where a lectotype was proposed. Material in brackets embedded within quotes is our addition. Lönnberg (1894:122-123) wrote: “This little fish seems to be extremely variable. When I obtained my first specimens in Ferncreek [Orlando, Orange County, FL] I surely believed that I had found a new species. I was led to that opinion by the number of spines and soft rays in the vertical fins. Jordan [1884:323] describes Elassoma evergladei with four spines and 9 or 10 soft rays in the dorsal and three spines and 5 soft rays in the anal. On my specimens I counted five spines (in one only 4) and 11 or 12 soft rays and the formula of the anal was III, 7. There was thus one spine and 1 or 2 soft rays in the dorsal and 2 soft rays in the anal more than in the typical E. evergladei. I therefore believed just to establish a new subspecies with the name ‘orlandicum’ the more as also the color etc. was different.” Bailey and Böhlke wrote: “Lönnberg’s … discus- sion emphasized the variability of evergladei, mentioned sexual differences, and added to the description, includ- ing a detailed account of life colors of both sexes. These SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 139 were carefully drawn and accurately described Elassoma okefenokee Böhlke (1956). The only sub- sequent use of the name orlandicum of which we are aware is by Barney and Anson (1920:242). [They] re- ferred to an Elassoma found at Orlando, Florida, named provisionally by Lönnberg as E. orlandicum.” We add here that Jordan and Evermann (1896) quoted Lönnberg’s entire account as a footnote to their species account of E. evergladei, with the preface “Dr. Einer Lönnberg gives the following account of the specimens observed by him about Orlando, Florida, and provisionally named “Elassoma orlandicum” (ibid:984). Bailey and Böhlke: “In order to investigate the status of orlandicum, we have had the privilege, through the courtesy of Dr. Å. Holm, to examine the 19 syntypes from Upsala Universitets Zoologiska Museum. These are labeled “Elassoma evergladei var orlandicum Lönnb., collected in Orlando, Ferncreek, Orange Co., Florida, Jan. 1893, E. Lönnberg”. Examination immedi- ately provided explanation for Lönnberg’s impression of variability: the syntypic series is complex, consisting of nine specimens of Elassoma evergladei Jordan and ten of E. okefenokee Böhlke.” “Lönnberg’s fin-ray counts agree better with okefenokee than with evergladei …, but we are un- able to explain his recording of five dorsal spines. Only one of the 19 syntypes (an example of okefenokee) has five spines. It seems evident that although Lönnberg had a mixed sample the description was drafted mostly from specimens of okefenokee.” At this point, the Bailey and Böhlke manuscript includes two tables, one typed, one handwritten, that give sex, standard length, counts, proportional measurements, and pigmentation traits of the nine E. evergladei and ten E. okefenokee specimens in the lot. There is no need to duplicate those tables here but we will summa- rize the major distinctions they noted between the two species that were the basis for their sort. Proportional measurements are given as thousandths of standard length and the range is followed by the mean. Counts are given as range followed by mean. In each case, the first set of values is for the nine E. evergladei speci- mens, the second set is for the ten E. okefenokee. Predorsal length: 452-534, 483; 415-450, 430. Caudal peduncle depth: 141-161, 152; 125-144, 132. Longest pectoral ray: 172-214, 182; 130-162, 147. Dorsal fin rays: 10, 10.0; 10-12, 11.0. Anal fin rays: 5-6, 5.6; 6-9, 7.1. Neural spines anterior to first dorsal pterygiophore: 5-6, 5.6; 3-4, 3.9. Top of head: fully scaled versus na- ked. Bailey and Böhlke continue: “We elect to pre- serve current nomenclature for these two species by choosing from the Ferncreek syntypes as lectotype of Elassoma evergladei orlandicum Lönnberg, 1894, a male, 19.7 mm in standard length. This specimen agrees with E. evergladei Jordan (1884), and the name orlandicum should therefore be listed in the synonymy of that species. The lectotype [ZMUU 344a] has counts as follows: dorsal IV, 10; anal III, 6; pectoral 14-14; sum of softray counts of these fins 44 (see Böhlke, 1956:9); body scales 31; vertebrae 28; neural spines anterior to first dorsal pterygiophore 6. The top of the head has a dense investiture of exposed, imbricate scales. The fol- lowing body proportions are in thousandths of the stan- dard length: predorsal length 487; caudal peduncle depth 157; length of dorsal-fin base 350; length of anal-fin base 188; longest pectoral ray 173; length of pelvic fin 274; diameter of eye 102; snout length 76. The pelvic fins are dusky; the dorsal and anal fins are lightly dusted with melanophores and both fins are notably darkened poste- riorly; there are two pale spots at the base of the caudal fin.” “Decision for the selection above was based on the following considerations. (1) E. e. orlandicum, [al- though] the older name, has remained almost unnoticed for nearly 100 years whereas okefenokee is established and has received general acceptance. (2) Lönnberg’s description was somewhat equivocal: “I surely believed I had found a new species…. I therefore believed just to establish a new subspecies …. The variability of the E. evergladei becomes the more evident.” [We add, furthermore, that the name “orlandicum” appears only once in Lönnberg’s paper. It is embedded in the text under the species account of E. evergladei (p. 123), where it is enclosed in quotes and is not italicized, unlike all other scientific names appearing in Lönnberg’s pa- per]. “(3) In view of the above, orlandicum might be interpreted as first published as a synonym, therefore unavailable in nomenclature [International Code of Zoo- logical Nomenclature, 1999:article 11.6.] In light of our lectotype selection uncertainty on this question is obvi- ated.” Dr. Bailey remains convinced that this lectotype designation is the correct course of action (in litt. to FFS, June 2007). The remaining eight syntypes of E. evergladei (ZMUU 344b-i; 16.1-22.1 mm SL) become paralectotypes of Elassoma evergladei orlandicum Lönnberg 1894. The other ten syntypic specimens (ZMUU 344j-s; 15.9-20.7 mm SL) are re-identified as Elassoma okefenokee Böhlke 1956. FFS has re-ex- amined these 19 specimens and confirms and agrees with Bailey and Böhlke’s results, conclusions, and lecto- type designation. Since the Bailey and Böhlke manuscript was writ- 140 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) ten, additional original Lönnberg material has been dis- covered. This material, reported by Gilbert (1998) and also examined by FFS, is as follows: NHRM 14105 (1 specimen, 17.0 mm SL) from Ferncreek, January 1893, re-identified as E. okefenokee; NHRM 14106 (5, 17.0- 20.0 mm SL) from Ferncreek, January 1893, re-identi- fied as E. okefenokee. These two lots appear to repre- sent legitimate syntypes. The lot NHRM 9170 contains three specimens (14.0-17.5 mm SL)) from Bagdad, 27 April, 1893; two are re-identified as E. okefenokee, one is dried and unidentifiable. Gilbert (1998) opined that the specimens from Bagdad are questionable syntypes because Bagdad is not mentioned in the original text as a place where Lönnberg collected pygmy sunfish. How- ever, Gilbert (2004) lists NRM 9170 as syntypic mate- rial and states that “…all extant types of E. evergladei orlandicum are from Fern Creek”. We note that the current town of Bagdad is in Santa Rosa County, Florida, in the Blackwater River drainage basin. Neither E. okefenokee nor E. gilberti are known from that drain- age system. ACKNOWLEDGEMENTS Mrs. Eugenia B. Böhlke (deceased) re-examined the holotype and several specimens from each lot of the paratypes of Elassoma okefenokee Böhlke, located at ANSP, and confirmed that all had three preopercular pores on each side of the head (in litt. to FFS, Septem- ber 1, 1992.) The drawings of the head pores (Fig. 2) are by Jason Bourque. We thank Cathy Bester for help with figure preparation and Rob Robins for handling the cataloging of recent material. Fresh material for DNA extraction was provided by Allen Boatman, Casper Cox, Fritz Rohde, Michael Sandel, and Klaus Schmidt. The helpful comments of two outside reviewers greatly im- proved the manuscript. FFS would like to thank the administration and staff of the Florida Museum of Natu- ral History, especially George Burgess, Larry Page, Rob Robins, and Dave Steadman, for providing office and lab space and other logistic support during the course of this study. LITERATURE CITED Applied Biosystems. 1994. 373 DNA Sequencing Sys- tem User’s Manual. Part number 903204, Revi- sion A. Perkin Elmer, Foster City, CA. Barney, R.L., and Anson, B.J. 1920. Life history and ecology of the pigmy sunfish, Elassoma zonatum. Ecology, 1:241-256. Bermingham, E., and Avise, J.C. 1986. Molecular zoo- geography of freshwater fishes in the southeastern United States. Genetics, 113:939-965. Böhlke, J.E. 1956. A new pygmy sunfish from south- ern Georgia. Notulae Naturae, 294:1-11. Böhlke, J.E., and Rohde, F.C. 1980. Elassoma okefenokee Böhlke, Okefenokee pygmy sunfish. P. 585 in D.S. Lee et al., eds. Atlas of North American Freshwater Fishes. Publication 1980- 12, North Carolina Biological Survey, North Caro- lina State Museum of Natural History, Raleigh, NC. Boschung, H.T., Jr., and Mayden, R.L. 2004. Fishes of Alabama. Smithsonian Books, Washington, DC. Branson, B.A., and Moore, G.A. 1962. The lateralis components of the acoustico-lateralis system in the sunfish family Centrarchidae. Copeia, 1962:1-108. Chow, S., and Hazama, K. 1998. Universal PCR prim- ers for S7 ribosomal protein gene introns in fish. Molecular Ecology, 7:1247-1263. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39:783- 791 Gilbert, C.R. 1998. Type catalog of Recent and fossil North American freshwater fishes: families Cyp- rinidae, Catostomidae, Ictaluridae, Centrarchidae and Elassomatidae. Florida Museum of Natural History, Special Publication No. 1. Gilbert, C.R. 2004. Family Elassomatidae Jordan 1877 – pygmy sunfishes. California Academy of Sci- ences Annotated Checklists of Fishes, No. 33. Gilbert, C.R., Cashner, R.C., and Wiley, E.O. 1992. Taxonomic and nomenclatural status of the banded topminnow, Fundulus cingulatus (Cyprino- dontiformes: Cyprinodontidae). Copeia, 1992:747- 759. Hubbs, C.L., and Lagler, K.F. 1974. Fishes of the Great Lakes region. The University of Michigan Press, Ann Arbor, MI. International Commission on Zoological Nomenclature. 1999. International Code of Zoological Nomencla- ture. The International Trust for Zoological No- menclature, London. Available at: http://iczn.org/ iczn (accessed February 2009). Johnson, D.G. 1984. Percoidei: development and rela- tionships. Pp. 464-498 in H.G. Moser et al., eds. Ontogeny and Systematics of Fishes. American Society of Ichthyologists and Herpetologists, Spe- cial Publication 1. Allen Press, Lawrence, KS. Jones, W.J., and Quattro, J.M. 1999. Phylogenetic af- finities of the pygmy sunfishes (Elassoma) inferred from mitochondrial DNA sequences. Copeia, 1999:470-474. Jordan, D.S. 1884. List of fishes collected in Lake Jessup and Indian River, Florida, by R. E. Earll, with descriptions of two new species. Proceed- SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 141 ings of the U. S. National Museum, 7:322-324. Jordan, D.S., and Evermann, B.W. 1896. The fishes of North and Middle America. Bulletin of the U. S. National Museum, 47(Pt. 1):1-1240. Lönnberg, E. 1894. List of fishes observed and col- lected in South-Florida. Ofversigt af Kongl. Vetenskaps-Akademicus Förhandlingar, No. 3:109- 131. Mayden, R.L. 1993. Elassoma alabamae, a new spe- cies of pygmy sunfish endemic to the Tennessee River drainage of Alabama (Teleostei: Elassomatidae). Bulletin of the Alabama Museum of Natural History, 16:1-14. Nelson, J.S. 2006. Fishes of the World, 4th ed. John Wiley & Sons, Hoboken, NJ. Palumbi, S.R. 1996. Nucleic acids II: The polymerase chain reaction. Pp. 205-247 in D.M. Hillis, C. Moritz, and B.K. Mable, eds. Molecular System- atics. Sinauer, Sunderland, MA. Posada, D., and Crandall, K.A. 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics, 14:817-818. Quattro, J.M., Jones, W.J., Grady, J.M., and Rohde, F.C. 2001. Gene-gene concordance and the phyloge- netic relationships among rare and widespread pygmy sunfishes (genus Elassoma). Molecular Phylogenetics and Evolution, 18:217-226. Roe, K.J., Harris, P.M., and Mayden, R.L. 2002. Phy- logenetic relationships of the genera of North American sunfishes and basses (Percoidi: Centrarchidae) as evidenced by the mitochondrial cytochrome b gene. Copeia, 2002:897-905. Rohde, F.C., and Arndt, R.G. 1987. Two new species of pygmy sunfishes (Elassomatidae, Elassoma) from the Carolinas. Proceedings of the Academy of Natural Sciences, Philadelphia, 139:65-85. Swift, C., Yerger, R.W., and Parrish, P.R. 1977. Distri- bution and natural history of the fresh and brackish water fishes of the Ochlockonee River, Florida and Georgia. Bulletin of the Tall Timbers Research Station, No. 20, 111 pp. Swofford, D.L. 1998. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Version 4.0b10. Sinauer, Sunderland, MA. Thompson, J.D., Higgins, D.G., and Gibson, T.J. 1994. CLUSTAL W: improving the sensitivity of progres- sive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Research, 22:4673-4680. 142 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) APPENDIX 1 MATERIAL EXAMINED Institutional Abbreviations for material examined are as follows: ANSP (Academy of Natural Sciences of Philadel- phia), UMMZ (University of Michigan Museum of Zoology), AUM (Auburn University Museum), GMNH (Georgia Museum of Natural History), TU (Tulane University), UF (Florida Museum of Natural History), and WTLC (Walton Taxonomy Laboratory Collection, Georgia Wildlife Resources Conservation Center). We thank the curators and staff of these institutions for specimen loans. The museum number is followed in parentheses by the number of specimens in the lot. Collection numbers followed by an asterisk contain some specimens field fixed and maintained in 95% ETOH and tissue samples are available for molecular study. Elassoma gilberti Apalachicola Dr. – Florida. Franklin Co.: UF 4594 (15), UF 53382 (3), UF 53386 (6). Jackson Co.: TU 39598 (28), UF 4987 (1), UF 5858 (5), UF 52709 (4), UF 52720 (5), UF 52827 (5), UF 52949 (3), UF 53383 (2), UF 58966 (17), UF 59678 (3), UF 60162 (1), UF 60178 (36), UF 130169 (1). Liberty Co.: UF 117051 (5), UF 120168 (8), UF 144312 (2), UF 144337 (1). Georgia. Decatur Co.: WTLC BA05-041 (1), UF 1772 (5). Dougherty Co.: UF 105553 (1). Miller Co.: UF 105442 (1). Mitchell Co.: GMNH 302 (3). Seminole Co.: UF 4863 (81). Aucilla Dr. – Florida. Jefferson Co.: UF 1263 (1), UF 5867 (23), UF 50945 (4), UF 53376 (4), UF 63321 (11), UF 73687 (4), UF 73799 (1), UF 74004 (12), UF 74287 (4), UF 74468 (2), UF 74495 (1), UF 74940 (8), UF 74956 (4), UF 144619 (4), UF 145860 (5), UF 173613* (24). Jefferson-Madison Co.: UF 74037 (15), UF 74539 (39). Madison Co.: UF 53380 (30), UF 63339 (18), UF 73964 (1). Georgia. Thomas Co.: UF 66710 (10), UF 75009 (1), UF 75026 (1), UF 75283 (1), UF 75284 (16), UF 75289 (2). California Creek Dr. – Florida. Dixie Co.: UF 63857 (3). Choctawhatchee Dr. – Florida. Holmes Co.: UF 5851 (3). Holmes-Jackson Co.: UF 54162 (1), UF 72411 (2), UF 72548 (8). Jackson Co.: UF 54255 (67). Okaloosa Co.: UF 51936 (7), UF 55605 (2), UF 156374 (13). Walton Co.: TU 111445 (25), TU 124327 (34), UF 50182 (1), UF 50369 (1), UF 130058 (1), UF 144903 (1), UF 145082 (1), UF 145092 (7), UF 145309 (1), UF 145435 (6). Washington Co.: UF 53387 (2), UF 54262 (15), UF 54265 (2), UF 55035 (3). Econfina Dr. – Florida. Taylor Co.: UF 73769 (11), UF 74020 (15), UF 74045 (1), UF 74453 (1), UF 74505 (3), UF 74888 (22), UF 91808 (2), UF 95939 (16), UF 95947 (1), UF 95948 (1), UF 95958 (9), UF 95959 (6), UF 96761 (1), UF 104458 (16). Fenholloway Dr. – Florida. Taylor Co.: UF 74308 (7), UF 74378 (7), UF 74845 (4), TU 36125 (4). Homosassa Dr. – Florida. Citrus Co.: UF 120466 (1). New Dr. – Florida. Liberty Co.: UF 71753 (1), UF 71879 (8), UF 71899 (12), UF 71921 (3). Ochlockonee Dr. – Florida. Gadsden Co.: UF 50093 (10), UF 50309 (34), UF 53388 (4), UF 54266 (10), UF 69793 (2), UF 70149 (11). Leon Co.: UF 53199 (2), UF 53377 (3), UF 53378 (11), UF 53379 (9), UF 61109 (1), UF 71843 (2), UF 75258 (2). Liberty Co.: UF 5859 (14), UF 50168 (1), UF 50232 (31), UF 50248 (2), UF 52264 (16), UF 53375 (1), UF 53383 (1), UF 54261 (14), UF 69794 (2), UF 70076 (8). Wakulla Co.: UF 69776 (5), UF 69926 (7), UF 69997 (21), UF 71779 (7), UF 73291 (1). Spring Warrior Dr. – Florida. Taylor Co.: UF 38823 (25), UF 74817 (1). St. Marks-Wakulla Dr. – Florida. Jefferson Co.: UF 5856 (11), UF 53044 (9), UF 77051 (3). Leon Co.: GMNH SNELSON, KRABBENHOFT, and QUATTRO : New Species of Pygmy Sunfish From Florida and Georgia 143 77074 (1), UF 79529 (1), UF 101462 (10), UF 130745 (2), UF 131077 (3), UF 173610* (19). Steinhatchee Dr. – Florida. Dixie Co.: UF 116466 (24). Dixie-Taylor Co.: UF 58463 (11), UF 74198 (21), UF 173612 (2). Lafayette Co.: UF 38124 (2), UF 74324 (20), UF 74703 (1), UF 74809 (1), UF 75058 (4), UF 75085 (4), UF 173611 (8), UF 173614 (4), UF 173609 (77). Taylor Co.: UF 58465 (23), UF 74216 (1). Suwannee Dr. – Florida. Dixie Co.: UF 2490 (9), UF 92238 (11), UF 173607 (26), UF 173608 (36). Gilchrist Co.: UF 58210 (4), UF 58246 (26), UF 173615* (3). Hamilton Co.: UF 4646 (5). Lafayette Co: UF 30238 (9). Levy Co.: UF 90972 (128), UF 110792 (1), UF 120345 (3), UF 120352 (6), UF 120359 (3). Waccasassa Dr. – Florida. Levy Co.: ANSP 151940 (8), GMNH 372 (1), UF 5860 (3), UF 63175 (1), UF 63382 (1). Elassoma okefenokee Altamaha Dr. – Georgia. Wheeler Co.: GMNH 1073 (1). Hillsborough Dr. – Florida. Hillsborough Co.: UF 173640 (5), UF 173643 (3). Pasco Co.: TU 135521 (1). Kissimmee Dr. – Florida. Glades Co.: UF 118701 (1). Highlands Co.: UF 96451 (1), UF 104843 (1). Okeechobee Co.: UF 2487 (4), UF 96452 (1). Orange Co.: UF 173619 (30), UF 173623 (7), UF 173626 (2), UF 173630 (1). Satilla Dr. – Georgia. Bacon Co.: GMNH 791 (10), GMNH 1474 (4). Brantley Co.: UF 23731 (3), UF 23740 (1). Charlton Co.: AUM 11402 (2). Coffee Co.: GMNH 1573 (1). Wayne Co.: GMNH 1022 (16). St. Johns Dr. – Florida. Alachua Co.: UF 40 (13), UF 2496 (5), UF 2498 (8), UF 2500 (4), UF 5852 (11), UF 5854 (14), UF 5855 (14), UF 5857 (3), UF 5862 (1), UF 5870 (8), UF 9670 (2), UF 17259 (22), UF 25528 (23), UF 25529 (4), UF 32854 (6), UF 43736 (10), UF 45077 (14), UF 81232 (4), UF 90700 (32), UF 97363 (27), UF 146398 (2), UF 146919 (10), UF 173632* (20), UF 173625 (23), UF 173637* (7), UF 173644 (9), UF 173624 (20). Clay Co.: UF 22825 (12), UF 96101 (5). Flagler Co.: AUM 33914 (3). Lake Co.: UF 7652 (11), UF 35235 (2), UF 43315 (2), UF 47209 (5), UF 79514 (1), UF 96179 (3), UF 173633* (17). Lake-Seminole Co.: UF 21519 (2), UF 81257 (2), UF 173620 (7), UF 173617 (19), UF173618 (17), UF 173639 (4), UF 173631 (31), UF 173628 (25), UF 173629 (6), UF 173638* (20). Marion Co.: UF 4428 (1), UF 8753 (1), UF 22905 (6), UF 23175 (5), UF 26267 (1), UF 26355 (1), UF 101463 (1), UF 121949 (4), UF 121950 (4). Marion-Putnam Co.: UF 22927 (3), UF 125369 (1), UF 173636 (17). Orange Co.: UF 173622 (29). Putnam Co.: UF 19 (1), UF 1907 (5), UF 23155 (8), UF 35966 (18), UF 41953 (3), UF 42208 (5), UF 43296 (7), UF 45509 (7), UF 47443 (2). Seminole Co.: UF 173627 (4). Volusia Co.: UF 1308 (4), UF 5871 (79). St. Marys-Nassau Dr. – Florida. Baker Co.: UF 26320 (2), UF 34151 (8), UF 56485 (1). Nassau Co.: UF 56476 (1), UF 56514 (5), UF 145734 (3). Georgia. Charlton Co.: AUM 11238 (18), GMNH 1150 (9), GMNH 1149 (9), TU 213320 (94), UF 173635* (23). Ware Co.: GMNH 1476a (17). Suwannee Dr. – Florida. Alachua Co.: UF 9622 (33), UF 25535 (28), UF 34021 (2), UF 38427 (1). Alachua- Bradford Co.: UF 173642 (8). Alachua-Columbia Co.: UF 173641 (3). Bradford Co.: UF 34037 (2), UF 34061 (21). Columbia Co.: UF 25526 (1), UF 101461 (1), UF 101509 (10), UF 173634* (37). Columbia-Suwannee Co.: UF 43728 (1), UF 123566 (1), UF 126467 (1), UF 173621 (2). Columbia-Union Co.: UMMZ 210073 (1). Gilchrist Co.: UF 5865 (1), UF 7645 (28). Suwannee Co.: UF 4984 (1). Union Co.: UF 25536 (1), UF 34140 (10). Georgia. Berrien Co.: AUM 10334 (24). Clinch Co.: AUM 4986 (9), GMNH 1312a (22), UF 173616* (50). Echols Co.: GMNH 2049 (24). Lanier Co.: UF 4001 (1). Lowndes Co.: GMNH 1063 (26), UF 50576 (8). Ware Co.: GMNH 1439 (30). Withlacoochee Dr. – Florida. Citrus Co.: UMMZ 176248 (1), UMMZ 176261 (1), UF 7650 (27), TU 12565 (1). Marion Co.: UF 85364 (9). Sumter Co.: GMNH 843 (3). 144 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 48(4) APPENDIX 2 DNA MATERIAL EXAMINED Each sample entry for Elassoma gilberti and E. okefenokee begins with a map code (Eg1, etc.) that corresponds to those in Figure 3. Elassoma gilberti Eg1: UF 173656, Ochlockonee drainage, FL. Eg2: UF 173610, UF 173657, St. Marks drainage, FL. Eg3: UF 173613, Aucilla drainage, FL. Eg4: UF 173595, UF 173606, Econfina drainage, FL. Eg5: UF173611, Steinhatchee drainage, FL. Eg6: UF 173615, Suwannee drainage, FL. Elassoma okefenokee Eo1: UF 173616, Suwannee drainage, GA. Eo2: UF 173634, Suwannee drainage, FL. Eo3: UF 173641, Suwannee (Santa Fe) drainage, FL. Eo4: UF 173642, Suwannee (Santa Fe) drainage, FL. Eo5: UF 173632, St. Johns drainage, FL. Eo6: UF 173637, St. Johns drainage, FL. Eo7: UF 173636, St. Johns drainage, FL. Eo8: UF 173633, St. Johns drainage, FL. Eo9: UF 173638, St. Johns drainage, FL. Eo10: UF 173643, Hillsborough drainage, FL. Elassoma evergladei UF 173757, Apalachicola drainage, FL. (west). UF 173756, Ochlockonee drainage, FL. (west). UF 173759, St. Johns drainage, FL. (east). Elassoma zonatum UF 173758, Waccasassa drainage, FL. Snelson_48(4)_Cover.pdf vol48no4.pdf