2003 SOUTHEASTERN NATURALIST 2(2):159-178 A NEW SPECIES OF EURYCEA (CAUDATA: PLETHODONTIDAE) FROM NORTH AND SOUTH CAROLINA JULIANR. HARRISON,111'* AND SHELDONI. GUTTMAN2 ABSTRACT - We describe a new species of hemidactyliine salamanderin the genus Eurycea based on a study of variation in morphometric, osteologic, reproductive, and larval characteristics and allozyme-based allele frequencies and fixed differences in samples of two color pattern morphs known to exist in the E. quadridigitata complex in North and South Carolina. The new species differs from its presumptive relative Eurycea quadridigitata (Holbrook) in having a bright yellow venter, a smaller size, relatively longer limbs, fewer paravomerine teeth, a greater number of yolk-laden ovarian eggs, and dorsally spotted hatchling larvae. Osteologic differences include a frontal with a welldeveloped triangular postero-lateral process, a prevomer that lacks a posteromedial bony shelf, a prevomerine tooth series that is not sharply arched, a tetraradiateos triangulare, and fewer trunk vertebrae. The new species exhibits between four and seven fixed allozyme differences from E. quadridigitata. The two species are largely allopatric in distribution in the Carolinas. INTRODUCTION Our study of geographical variation in Eurycea quadridigitata (Holbrook) within North and South Carolina has shown that two color pattern morphs known to exist in this area actually correspond to two separate, largely allopatric species. The new species differs from E. quadridigitata in both morphological and genetic characteristics. In this paper we first discuss the literature concerning color pattern morphs and then describe as a new species certain populations occupying the Piedmont and upper Coastal Plain of South Carolina, together with others that occupy the Piedmont and portions of the Coastal Plain in North Carolina. Bishop (1943) first noted the existence of two color pattern morphs in Eurycea quadridigitata (Holbrook). One of the most striking differences between these two morphs is a completely yellow venter lacking dark markings in what Bishop referred to as the "light form" as contrasted with a uniformly pigmented venter with small dark brown flecks in his "dark form." The dark form or morph also possesses dark brown sides with narrow, longitudinal or oblique yellow or white lines as opposed to simply brownish upper sides in the light form or yellow Department of Biology, The College of Charleston, Charleston, SC, 29424. 2 Department of Zoology, Miami University, Oxford Ohio 45058. * Corresponding author - [email protected]. 160 SoutheasternNaturalist Vol. 2, No. 2 morph. Bishop provided additional details for both morphs, but did not comment on the geographic distribution of either form. It is likely that Bishop (1943) based the description of his "light form" on specimens sent to him by C.S. Brimley from the Raleigh area of the lower Piedmont in North Carolina. Brimley's (1944) characterizationof Eurycea quadridigitata fits Bishop's "light form" only, and his experience with this species was limited primarily to the Raleigh area. Dunn's (1926) description of Eurycea quadridigitata is a composite of both color pattern morphs. The male described, collected in Florida, is a "dark form" specimen, whereas the female, collected at Raleigh by Brimley, is a "light form" specimen. Although Mittleman (1947) studied geographic variation in Eurycea quadridigitata throughout its then-known range, his study was based largely on preserved material. He stated that he took notes on color pattern and pointed out that "... various phases are found with equal frequency in all parts of the range ..." and that he was unable "... to distinguish specimens from one locality or another on the basis of color and/or pattern ..." However, Mittleman did not characterize any of the color patternshe investigated. Only 15 of the 246 specimens Mittleman examined were collected in North or South Carolina, and 11 of the 15 were from Wake Co., NC, an area now known to harbor only populations of the yellow morph. Mittleman (1947) described two new subspecies of Eurycea quadridigitata (Holbrook) based primarily on quantitative differences involving number of prevomerine teeth, number of costal grooves, and number of costal grooves between adpressed toes. Later, and without comment, Mittleman (1967) relegated each of the subspecies he described to the synonymy of Manculus quadridigitatus (Holbrook) and regarded this species as monotypic. These allocations were presumably based in part upon the studies of Cagle (1952) and Neill (1949, 1954) that provided no support for subspecific differentiation within the species. There is no furthermention of color or color patternsin the literature concerning Eurycea quadridigitata until Folkerts (1971) pointed out the existence of three populations of this species from Anderson and Pickens counties in the upper Piedmont of South Carolina. All of the specimens he examined from these three populations were referable to Bishop's "light form." Folkerts characterizedthe Piedmont animals and compared them with 76 "darkform" specimens from the Coastal Plain of South Carolina. In addition to the color pattern differences, he pointed out that the Piedmont specimens were smaller on average, but stated there were no differences in costal groove or prevomerine tooth counts from those given by Mittleman (1967). Holbrook's (1842:331) description and plate of Salamandra quadridigitata are clearly referable to Bishop's (1943) "dark form." 2003 Julian R. Harrison, III and S.I. Guttman 161 The plate (21) shows a pigmented venter and a dark-sided tail, both characteristics of the dark morph. In his description the venter ("abdomen") is said to be bluish silvery-white, a striking feature of living dark morph individuals. Also, the tail is described as "... compressed only towards its tip," a characteristic of the dark morph. In yellow morph specimens the tail is round with little or no indication of lateral compression or a dorsal keel. Holbrook, however, provided no information concerning number of costal grooves or number of prevomerine teeth. The provenance of Holbrook's type (ANSP 490 from South Carolina) is unknown. The type description was based on specimens from Georgia and Florida, as well as South Carolina. Although the type is in rather poor shape (Mittleman 1967), two x-ray photographs have revealed the number of trunk vertebrae in this specimen to be 19 (Richard Highton, pers. comm.). As discussed later in this paper, most specimens of the yellow morph examined in the present study had 17 trunk vertebrae, whereas most specimens of the dark morph had 19 trunk vertebrae; these are estimates based on the number of costal grooves (Highton 1957) and direct counts from the study of stained and cleared material. Schmidt (1953) restricted the type locality of E. quadridigitata to the vicinity of Charleston, SC, an area known to harbor only the dark morph. Cope's (1871) description of Manculus remifer also fits rather well Bishop's "darkform," as he stated that "... color is black above, and dark brown below." On the basis of external morphology, color pattern, and number of trunk vertebrae, we conclude that Holbrook's (1842) description of Salamandra quadridigitata is clearly referable to the dark morph, and that the yellow morph is an undescribed species. One of us (JRH) first became interested in the status of Bishop's two color pattern morphs during a study (Harrison 1973) of the life history of Eurycea quadridigitata in the lower Coastal Plain of South Carolina. All of the more than 600 specimens examined at that time were referable to Bishop's "dark form." Later, in 1974, Harrison collected three specimens from a locality in the Sandhills region of the upper Coastal Plain in South Carolina that proved referable to Bishop's "light form" and distinctly different from other Coastal Plain material. A study of this material and additional specimens from the Clemson area in the upper Piedmont of South Carolina, the Coastal Plain of South Carolina, and the Piedmont and Coastal Plain of North Carolina has shown that Bishop's two color pattern forms in this region are separate, largely allopatric species. In addition, Guttman examined protein variation among samples from Pickens and Barnwell counties, SC, Scotland Co., North Carolina, Osceola Parish, Louisiana, and Beauregard Parish, Louisiana and found fixed genetic differences as well. The new species of Eurycea is described below. 162 Vol. 2, No. 2 SoutheasternNaturalist METHODS Measurements of adults were made with Vernier calipers to the nearest 1.0 mm, larvae to the nearest 0.5 mm with a metric rule and a dissecting microscope. Abbreviations used are SVL (snout to anterior end of vent), T (tail length from anterior end of vent to tip of tail), HL (tip of snout to posterior edge of gular fold). HW (head width), ICF (number of intercostal folds between adpressed limbs), VT (number of prevomerine teeth counted with a dissecting microscope after drying roof of mouth with compressed air), HL/SVL (relative head length), HW/SVL (relative head width), and T/SVL (relative tail length). Specimens for osteological studies were fixed in formalin, measured, sex determined, and cleared in potassium hydroxide, stained with Alizarin Red-S, and stored in glycerine. Ovarian egg counts in gravid females included all obviously enlarged, yolk-laden eggs. Specimens for genetic studies were available from the Clemson area, Pickens Co., SC (yellow morph) Barnwell Co, SC (dark morph), Scotland County, NC (dark morph), Osceola Parish, LA (dark morph) and Beauregard Parish, LA (yellow morph); these were sent to Guttman at Miami University for electrophoretic analysis of protein variation. An organ homogenate was prepared from the kidneys, heart, liver, tongue and rinsed digestive tract of each animal. The tissues were homogenized together in an equal volume of chilled 2% 2-phenoxyethanol and centrifuged at 25,000 g at 4 'C for 45 min to obtain an aqueous protein extract. The supernatant of water-soluble proteins was decanted and frozen at -70 TCuntil used, a maximum of 72 h following preparation. Extracts were analyzed by standard techniques of horizontal starch gel electrophoresis (Selander et al. 1971, Ayala et al. 1972). Seventeen loci were consistently resolvable (see Table 1. Enzymes and buffer systems used in this study. Enzyme system Number Buffer a Aspartate aminotransferase(AAT) Albumin (ALB) Fumaratehydratase (FH) Glucose-6-phosphate isomerase (GPI) Glutamate dehydrogenase (GLUDH) Isocitrate dehydrogenase (IDHP) Lactate dehydrogenase (LDH) Malate dehydrogenase (MDH) Malic enzyme (MEP) Peptidase (1-alanyl-1-leucyl; PEP) Phosphogluconate dehydrogenase (PGDH) Phosphoglucomutase (PGM) Protein-2 (PT) Superoxide dismutase (SOD) 2. 6. 1. 1 LiOH LiOH TC8 LiOH TEB TC8 TEB TC8 TC8 LiOH TC8 TC8 LiOH TC8 4. 2. 1. 2 5. 3. 1. 9 1. 4. 1. 1. 1. 1. 42 1. 1. 1. 27 1. 1. 1. 37 1. 1. 1. 40 3. 4. -. 1. 1. 1. 44 5. 4. 2. 2 -.-.-.- 1. 15. 1. 1 a LiOH, Lithium hydroxide, pH 8.1 and TC8, Tris-citrate, pH 8.0 (Selander, et al. 1971; TEB, Tris- borate-EDTA, pH 9.1 (Ayala, et al. 1972). 2003 Julian R. Harrison, III and S.I. Guttman 163 Table 1 for a list of enzyme loci, abbreviations and numbers following Shaklee et al. 1990). All gels were 12.5% starch (Electrostarch lots #301 [LiOH] and #307 [Tris-citrate and Tris-borate-EDTA], Otto Hiller, Madison, Wisconsin). Analysis of single individual genotypic data was performed using the Biosys I computer program developed by Swofford and Selander (1981). SYSTEMATICS EURYCEA CHAMBERLAINI SP. NOV. CHAMBERLAIN'S DWARF SALAMANDER Holotype. - U.S. National Museum of Natural History (USNM) 547846, an adult female collected from Sesquicentennial State Park, Richland County, South Carolina, on 3 August 1974, by Julian R. Harrison (JRH 2700) and Albert E. Sanders. Paratypes. - Fifteen specimens from South Carolina as follows: Charleston Museum (ChM) CA4566-69 (4) [JRH 3286-89], Barnwell Co., SavannahRiver Site, Rainbow Bay; ChM CA4570 (1) [JRH 3177], Pickens Co., near Clemson; USNM 547847-48 (2) [JRH 2701-02], Richland Co., Sesquicentennial State Park (topotypes, one stained and cleared); NCSM 36021 (8), Sumter Co., N edge of Sumter. Thirty-five specimens from North Carolina in the collection of the North Carolina State Museum (NCSM) as follows: NCSM 14947 (5) and 17103 (3), ChathamCo., 1.5 mi SSW Wilsonville; NCSM 36024 (1), CravenCo., ca. 3 mi E Tuscarora;NCSM 15788 (1), Johnston Co., 2 mi NE Four Oaks; NCSM 19159 (1), Lenoir Co., 2 mi NW Deep Run; NCSM 17470 (2), Moore Co., 8.25 mi SW Robbins; NCSM 36023 (2), Pamlico Co., 7 mi WSW Oriental; NCSM 17556 (2), Sampson Co., 4.8 mi NNE Delway; NCSM 14805 (3), Vance Co., 4 mi E Henderson; ChM CA4571-77 (7) [JRH 3178-84], NCSM 17335 (1), 17579 (4), and 19783 (1), Wake Co., 4.5 mi S Cary;NCSM 36022 (1), Wake Co., 2 mi E Holly Springs. Description of Holotype. - After preservation, total length 81 mm, tip of snout to anterior corner of vent 27 mm, to posterior corner of vent 29 mm; tail length from anterior corner of vent 54 mm, body width and depth (at mid-trunk) 4.5 mm and 2.9 mm respectively, axilla to groin length 17.2 mm, tail height and width just posterior to vent both 2.9 mm; head length from tip of snout to gular fold 6.5 mm, head width at widest portion 3.7 mm, head depth at angle of jaw 2.4 mm; eye diameter 1.8 mm, eye width 1.5 mm, interocular width (shortest distance between eyes) 1.2 mm; internarial distance 1.0 mm, tip of snout to orbit 1.2 mm. Number of costal grooves 16-16, number of intercostal folds between adpressed limbs 4, number of costal 164 SoutheasternNaturalist Vol. 2, No. 2 grooves between adpressed limbs 5. Digits of the forelimbs 1-4-3-2 in order of length from the shortest (first digit); digits of the hindlimbs 1-2-4-3. Number of prevomerine teeth 9-7, these widely separated from those of the parasphenoid series. Dorsum of head, trunk, and tail in life brownish-bronze; in preservative dorsal surfaces light brown, a consequence of abundant tiny punctate melanophores. Dorsum of trunk between dorsolateral stripes with a medial series of 15 dark spots about equally distributed from gular fold to a point just posterior to hindlimbs. Dorsum of head with a short dark occipital bar about as wide as the eyelids and a small dark dot behind the bar. Eyelids more heavily pigmented than remainder of head, except for a darker spot at the posteromedial corner of each eye and a medial darker spot just posterior to the eyes. Dorsum of forelimbs light brown with scattered pigment-free areas. Dorsum of femur with larger, mostly melanophore-free areas forming a fairly conspicuous horizontal band. A dark dorsolateral stripe, approximately 1.0 mm in width extends from the posterior corner of the eye to the tip of the tail. This merges fairly abruptly on the trunk with moderately pigmented sides with scattered pigment-free areas, which in turn are sharply demarcatedfrom the venter by an abruptborderextending from the base of the forelimb to the base of the hindlimb. Sides of tail moderately heavily pigmented with relatively few pigment-free areas. Venter of trunk in life bright yellow, venter of tail dusky-yellow; in preservative venter of trunk and tail pale buff with yellowish tinge, without dark pigment except scattered, very tiny punctate melanophores. Punctate melanophores somewhat more abundanton tail. Etymology. - The name honors Edward Burnham Chamberlain, former Curatorof VertebrateZoology, the Charleston Museum, in special recognition of his contributions to the study of amphibians and reptiles in South Carolina. Diagnosis. - A hemidactyliine plethodontid, probablyclosely related to Eurycea quadridigitata, and sharing with that species the possession of four toes on both fore and hind limbs, but differing in size, color pattern, prevomerine and costal groove count. Eurycea chamberlaini is smaller in size than E. quadridigitata, has a color pattern somewhat resembling that of E. cirrigera, has relatively longer limbs than E. quadridigitata, fewer costal grooves, and fewer prevomerine teeth. It differs also from the nominal species in the configuration of three skeletal elements, the prevomer, frontal and basibranchial II (os triangulare) of the hyobranchial apparatus, in the arrangement of prevomerine teeth, and by having fewer trunk vertebrae. 2003 Julian R. Harrison, III and S.I. Guttman 165 The new species is characterized in life by a brownish-bronze dorsum with distinct dorsolateral stripes and a bright yellow venter; snoutvent lengths of adults (SVL; measured to anteriorcorner of vent) 22-30 mm (g = 25.8 mm); modal numbers of 16 costal grooves and 17 trunk vertebrae; relatively long limbs, 3-6 costal folds between adpressed limbs (g = 4.2); a relatively long tail, tail to snout-vent length ratio 1.08-2.00 (g = 1.62); prevomerine teeth 5-16 (g = 10.0). The prevomer of Eurycea quadridigitata has a posterolateral bony shelf and the prevomerine tooth series is sharply arched (inverted "L" shape), an arrangement that is characteristic of hemidactyliine salamanders (Wake 1966). However, the prevomer of E. chamberlaini lacks the posterolateral bony shelf and the prevomerine tooth series is not sharply arched, an arrangement("plethodonine pattern")similar to that in salamandersof the tribe Plethodontini (Wake 1966) The frontal bone typically bears a short triangular process at its posterolateral corner. Basibranchial II of the hyobranchial apparatusis ossified and is typically tetraradiate,i.e., 4-pointed. The distinctive brownish-bronze dorsum and bright yellow venter is lost in preservative, but such specimens in the Carolinas can be identified through a combination of characteristics including especially the plethodonine prevomerine tooth pattern, narrow dark dorsolateral stripes, and an unpigmented venter, including the venter of the head and tail. Outside the Carolinas, number of costal grooves will not serve as a diagnostic characteristic, as some populations of E. quadridigitata are characterized by 16 costal grooves. The venter of the trunk may only be lightly pigmented in some E. quadridigitata, but the venter of the head and tail are usually moderately to heavily pigmented. Examination of the dorsal oral cavity, particularly if airdried, will reveal the plethodonine arrangement of the vomerine tooth series, as this appears to be a fairly consistent characteristic. Color Pattern. - Notes on color pattern in life are available for twelve specimens (10 females, 2 males) collected in Pickens County, SC (Fig. 1). These specimens displayed a color patternvery close to that of the "light form" described by Bishop (1943). No sexual dimorphism in this character was readily apparent, although only two males were represented. The dorsum of the head and trunk was bronzy-buff or brown in nine specimens and brownish-buff or buff in three others. The pattern of the darker markings on the dorsum was rather variable, ranging from none to variable patterns including faint herring bone-like arrangements,few or numerous flecks, and a mid-dorsal line or irregular row of dots. A deep layer of guanophores impartedthe bronzy aspect to the dorsum; these were not evident laterally. The dorsum of the tail was yellowish-buff or a light bronzy-buff with an orange tinge in eight 166 SoutheasternNaturalist Vol. 2, No. 2 specimens and brownish-buff, orange-buff, pale buff, or light bronzybrown in four others; darker markings present ranged from few to numerous tiny dark flecks. The venter of the head and limbs was flesh-colored with a suffusion of yellow on the posterior portion of the head in some specimens. The venter of the trunk was bright yellow in eleven specimens and yellowish-buff in one. The venter of the tail was orange-yellow or orange in eleven specimens and yellow in one. Seven specimens had a distinct or relatively distinct dorsolateral dark stripe while in four the dorsolateral and lateral areas were uniformly dark in color, sharply demarcatedfrom the yellow of the venter. In one specimen the dorsolateral stripe was replaced by a row of dots. Laterally and ventrolaterally, eight specimens had scattered to moderately numerous white flecks or streaks but four others lacked these. Variation. - Eurycea chamberlaini is, on average, slightly smaller in snout-vent length than E. quadridigitata (Table 2). Forty-seven measured adults had a mean snout-vent length of 25.8 mm in comparison to 28.0 mm for 48 adults of E. quadridigitata. These results are comparable to those obtained by Folkerts (1971) for a small sample of 11 specimens of the yellow morph from Pickens and Anderson counties, SC, and 76 specimens of the dark morphfrom the Coastal Plain of South Carolina. The new species is also relatively longer-limbed having a mean of 4.2 intercostal folds between adpressed limbs as opposed to a Figure 1. Male specimen of Eurycea chamberlaini from Pickens Co., SC, illustrating characteristic bright yellow venter. 2003 Julian R. Harrison, III and S.I. Guttman 167 mean of 5.5 in E. quadridigitata (Table 2). It also averages slightly fewer prevomerine teeth, having a mean of 10.0 in comparison to a mean of 11.8 in E. quadridigitata (Table 2). The means for each of these three characters(snout-vent length, relative limb length, and number of prevomerine teeth) were significantly different (Student's T) at P = <0.01. Little sexual dimorphismin these characteristicswas noted in the samples of either species. Males averaged slightly larger in relative head length, head width and tail length, whereas females averaged slightly more prevomerine teeth and had a slightly greater average relative limb length, but these differences were non-significant. The two species differ appreciably in number of costal grooves. These were counted using the "maximum count," i.e., including a groove in both the axilla and the groin but not a groove lying immediately above the hind limb (Highton 1957). Eighty-five percent of the 48 E. quadridigitata examined had 18-18 costal grooves, and 15% had 1717 or 18-17 (one specimen). By contrast, 83% of the 47 E. chamberlaini examined had 16-16 costal grooves and 17% had 15-15 or 15-16 (one specimen). In contrast, Folkerts (1971) found no differences in costal groove count between specimens of the yellow morph from the South Carolina Piedmont and the values given by Mittleman (1967) for E. quadridigitata. However, this may simply reflect the fact that Mittleman's data were taken from specimens collected over a broad geographic area from North Carolina to Mississippi, and very few of these were from the Carolinas. Ten specimens, six Eurycea quadridigitata and four E. chamberlaini from the Rainbow Bay complex in the Savannah River Site, Barnwell Co., SC, illustrate differences between the two species in one of the two known areas of sympatry. All of the E. chamberlaini have unpigmented venters, pale buff in preservative, whereas four of the E. quadridigitata have venters lightly to moderately pigmented with punctate melanophores. The venters of the remaining two E. quadridigitata have very few, widely scattered punctate melanophores in one and numerous punctate melanophores in the other. Also, the venters of the Rainbow Bay E. quadridigitata are not as heavily pigmented as those of specimens collected elsewhere within the Carolinas. In all of the E. Table 2. Morphometic characteristics of the measured samples of Euycea quadridigitata and E. chamberlaini. See Methods and Materials for identification of abbreviations. Eurycea chamberlaini Eurycea quadridigitata Character N Mean SD SVL T/SVL HL/SVL HW/SVL 28 1.47 0.238 0.136 2.44 0.15 0.013 0.011 48 37 48 48 Range N Mean SD Range 22-33 1.14-1.72 0.213-0.259 0.121-0.184 47 30 47 47 25.8 1.62 0.247 0.143 2.16 0.23 0.019 0.014 22-30 1.08-2.00 0.214-0.291 0.127-0.189 ICF 48 5.5 0.99 3-8 47 4.2 0.87 3-6 VT 48 11.8 2.15 8-16 47 10.0 2.99 5-16 168 Vol. 2, No. 2 SoutheasternNaturalist chamberlaini the lateral aspect of the tail is lightly pigmented, contrasting distinctly with the dorsolateral stripe. The lateral aspect of the tail is dark-sided in all of the E. quadridigitata, not contrasting distinctly with the dorsolateralstripe. Costal groove counts were 15 (1) or 16 (3) for the specimens of E. chamberlaini as contrasted with 16 (1), 17 (4), or 18 (1) for E. quadridigitata. Osteologic Differences. - Twenty-two stained and cleared skeletons were available for study, 10 Eurycea chamberlaini and 12 E. quadridigitata. Some of these preparations,however, are poorly stained and/or in poor condition, making it difficult or impossible to assess the state of one or more of the charactersexamined. The two species differ with respect to the structure and shape of the frontal, the prevomerine tooth series, and basibranchialII of the hyoid apparatus.They also differ in number of trunk vertebrae. Eurycea is one of several plethodontid genera that possess poorly to very well developed posterolateral frontal processes that overlap the parietals (Wake 1966). This character is variable in both the new and the nominal species but in E. chamberlaini the processes are somewhat more strongly developed and typically triangular or nearly triangular in shape, whereas they tend to be short and rounded in E. quadridigitata (Fig. 2). Eight of the 10 specimens of E. chamberlaini have triangular or nearly triangular processes but in the two others they are indiscernible or short with ctenate margins, respectively. In six specimens of E. quadridigitata in which the state of this character could be determined, the processes are relatively short and somewhat rounded in three specimens, relatively short and somewhat rounded A B Figure 2. Conformation of the frontal processes in (A) Eurycea quadridigitata and (B) E. chamberlaini. 2003 Julian R. Harrison, III and S.I. Guttman 169 but with ctenate margins in two specimens, and triangular in one specimen. The vomer of E. quadridigitata typically bears a sharply arched prevomerine tooth series with a bony shelf posterolateral to the series (Fig. 3A). This is representative of Wake's (1966) hemidactyliine pattern. No variation was noted in eight of 12 E. quadridigitata in which the state of this character could be determined. The prevomerine tooth series of E. chamberlaini is not sharply arched and the vomer lacks a posterolateral bony shelf (Fig. 3B), a pattern more characteristic of plethodonine salamanders.A similar condition was present in a stained and cleared specimen of E. cirrigera from York Co., SC. The preorbital process of the vomer is only slightly developed in both species. In six of the 10 specimens of E. chamberlaini examined the tooth series is clearly not arched, whereas it is slightly arched in the remaining four, but not to the degree observed in specimens of E. quadridigitata. The hyobranchial apparatus is similar in the two species, except that basibranchial II (os triangulare) is usually tetraradiate in E. chamberlaini and triradiatein E. quadridigitata (Fig. 4). No variation is apparentin eight of the 10 specimens of E. chamberlaini in which this character could be determined. Five of the 7 specimens of E. quadridigitata in which this character could be determined have the triradiatecondition, whereas two show the tetraradialstate. Based upon the information presented herein, Hilton's (1945) description of the skeleton of Manculus quadridigitatus is applicable to the new species, E. chamberlaini, not the nominal species. Similarly, this is also the case with regard to his (1947) illustration of the hyobranchial apparatusin the genus Manculus. The two species also differ in number of trunk vertebrae. According to Highton (1957), if a costal groove located posterior to the gular fold A Figure 3. Conformation of the prevomerine teeth in (A) Eurycea quadridigitata and (B) E. chamberlaini. B 'A A B Figure 4. Conformation of basibranchial II (os triangulare)in (A) Eurycea quadridigitata and (B) E. chamberlaini. 170 Vol. 2, No. 2 Southeastern Naturalist but anterior to the front limb is counted, then the number of costal grooves in E. quadridigitata is equal to the number of trunk vertebrae. As noted previously, seven of 47 specimens of E. chamberlaini examined have a costal groove count of 15-15, one has a count of 15-16, and Table 3. Allele quadridigitata. frequencies at variable loci in five populations of purported Euycea Missing letters represent alleles found in other species of Eurycea. Population LOCUS (N) AAT-1 E F ALB A C FH A B C GPI-1 B C GLUDH A B IDHP- 1 A B C D LDH-1 C D E LDH-2 A B MDH-1 C D MDH-2 A B PEP-1 B C D PGDH A B D PGM A B C D Yellow morph Pickens Beauregard Parish, LA Co., SC Osceola Parish, LA Eurycea quadridigitata Scotland Co., NC Barnwell Co., SC 11 12 9 7 19 0.05 0.95 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.06 0.67 0.27 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.94 0.06 1.00 1.00 0.06 0.94 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.44 0.56 1.00 0.76 0.24 1.00 1.00 1.00 0.06 0.94 1.00 1.00 1.00 1.00 0.95 0.05 1.00 0.05 0.95 1.00 1.00 1.00 0.14 0.86 0.83 0.17 0.36 1.00 0.64 0.05 0.95 0.38 0.62 1.00 0.09 0.91 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.58 0.42 0.23 0.18 0.59 0.04 0.96 2003 171 JulianR. Harrison,III andS.I. Guttman 40 have a count of 16-16. Six of 48 specimens of E. quadridigitata examined have a count of 17-17, one has a count of 18-17, and 41 have a count of 18-18. Thus, using Highton's method, most of these E. chamberlaini (83%) have 17 trunk vertebrae, the remainder 16 trunk vertebrae.In contrast, most of the E. quadridigitata (85%) have 19 trunk vertebrae, the remainder 18 trunk vertebrae. These results correspond fairly closely with observations on 19 stained and cleared specimens in which this character could be determined. Eight of the skeletons of E. chamberlaini had 17 trunkvertebraeand two had 18 trunkvertebrae. Six of the E. quadridigitata skeletons had 19 trunkvertebrae and three have 18 trunkvertebrae. Genetic Differences. - Thirteen of the 17 protein-producing loci surveyed were either polymorphic within populations or variable across populations (Table 3). LAP, ME, PT2 and SOD showed no detectable variation across all samples. Allele frequencies (Table 3) demonstrate that extensive differentiation occurs among these five samples. Fixed differences exist at 07 loci between population pairs (Table 4). The Pickens Co., SC sample (yellow morph) is the most strongly differentiated with seven fixed differences between it and both the Scotland Co., NC and Barnwell Co., SC populations (dark morphs). The Pickens Co., SC population also maintains four fixed differences from the two Louisiana samples [Osceola Parish (dark morph) and Beauregard Parish (yellow morph)]. In contrast, the Scotland Co., NC and Barnwell Co., SC populations are extremely similar with no fixed differences between them; the Osceola Parish, LA sample has one fixed difference (MDH-1) when compared to the Scotland Co., NC and Barnwell Co., SC populations. Genetic distances (unbiased, Nei 1978) between populations (Table 4), calculated utilizing all loci, range from 0.003 (Barnwell Co, SC and Scotland Co., NC) to 0.752 (Barnwell Co., SC and Pickens Co., SC). When genetic distances are averaged by color morph (Table 5), the mean genetic distance between the dark and yellow morphs is 0.599. Genetic differentiation among the three dark morph samples is slight (mean D = 0.053); however, the two yellow Table 4. Number of loci showing fixed differences between the five populations examined (above diagonal) and matrix of Nei (1978) unbiased genetic distances (below diagonal). Populations 1 and 2 are the yellow morph; populations 3 and 5 are E. quadridigitata. 4 5 1 2 3 Population 1. Beauregard Parish,LA. ***** 4 4 5 4 2. Pickens Co., SC 3. Osceola Co., SC 0.450 0.428 ***** 0.611 4 ***** 7 1 7 0 4. ScotlandCo., NC 5. BarnwellCo., SC 0.553 0.517 0.733 0.752 0.080 0.078 ***** 0.003 0 ***** 172 Vol. 2, No. 2 SoutheasternNaturalist morph samples are genetically quite different from each other (D = 0.450). The electrophoreticdata indicate that the yellow and darkmorphs are extremely different from one another and are genetically isolated. However, the magnitude of the differentiation between the populations from Beauregard Parish, LA and Pickens Co., SC also suggests that the samples of Eurycea from these two sites representdistinct entities that are not interbreeding;thus, the yellow morph genetically analyzed in this study appears to be composed of two species. The genetic data indicate that the populations studied belong in three, genetically distinct, groups; there is no evidence of gene flow between the groups. Therefore, we propose that they be recognized as three species. CarlaHass and Richard Highton (1998; pers. comm.) are currentlyconducting analyses of genetic variation of other populations throughoutthe range of the nominal species, and D. Bruce Means (pers. comm.) and John B. Jensen our investigating the statusof some Alabama,Florida,and Georgiapopulations. Distribution and Habitat. - Eurycea chamberlaini is known currently from portions of the Piedmont in both North and South Carolina, the upper Coastal Plain in South Carolina, and portions of the central Coastal Plain in North Carolina (Fig. 5). It is sympatric with E. quadridigitata in at least two localities, the Rainbow Bay complex (Rainbow Bay, Bullfrog Pond, and Sun Bay) in the SavannahRiver Site, Barnwell Co., SC (R.D. Semlitsch and J.H. Pechmann, pers. comm.), and a site ca. 3 mi WSW Allendale, Allendale Co., SC (specimens in the collection of the CharlestonMuseum). On the SavannahRiver Site, only E. quadridigitata was found at Flamingo Bay and Karen's Pond (R.D. Semlitsch, pers. comm.). Elsewhere in South Carolina the range of E. quadridigitata appears to be parapatric with that of E. chamberlaini along the lower edge of the sandhills region of that state. The former species reaches into North Carolina primarily in those counties bordering the South Carolina state line from the southern edge of the sandhills region in Scotland County eastward to Brunswick County. The pattern of distribution of E. chamberlaini, a range encompassing the Piedmont of both states but extending into the Coastal Plain primarily in North Carolina, is not unique. It is duplicated in part by at least two other species of amphibians, Hemidactylium scutatum and Bufo woodhousei fowleri (Martof et al. 1980, Conant and Collins 1998). Table 5. Matrix of Nei (1978) unbiased genetic distances averaged by color morph. Color Morph No. of populations 1 1. Dark Morph 3 0.053 (0.003-0.080) 2. Yellow Morph 2 0.599 (0.428-0.752) 2 0.450 (0.450-0.450) 2003 JulianR. Harrison,IIIandS.I. Guttman 173 In the Carolinas Eurycea quadridigitata occurs in a variety of habitats, but is especially associated with pine litter or other debris at the margins of pine flatwood or pine savanna ponds (Harrison 1973). It also occurs in the vicinity of Carolinabays, in tupelo-cypress swamps, and in bottomland hardwood forests. Eurycea chamberlaini also occurs in a variety of habitats but seems to prefer seepage areas near streams or ponds, particularly in Piedmont and upper Coastal Plain areas. Thirtythree records of the new species in files of the NCSM provide fairly specific habitat data. Twenty-three (ca. 70%) of these collections were associated with streams and/or seepages from springs, and ten (ca. 30%) were associated with floodplain sites or a pond. At the Savannah River Site in Barnwell Co., SC, both species have been caught in pitfall traps at Carolina bays, presumably entering these habitats to breed. However, E. chamberlaini tends to migrate to Rainbow Bay later in the fall than E. quadridigitata and there are seepage or wet areas in the vicinity of Rainbow Bay and Bullfrog Pond (J. H. Pechmann, pers. comm.). Whether or not E. chamberlaini oviposits in these sites and/or in the seepage or wet areas remains undetermined. Differences in Fecundity. - Data relative to the fecundity of Eurycea chamberlaini are available from 18 gravid females in two samples, one from Wake Co., NC, the other from Barnwell Co., SC. 000o0 0o 0,'D0 0~ o o oD o ro 0 0o 00 o 0 0 00 o 0 00k 50 km 0 Figure 5. Distribution of Eurycea chamberlaini (filled circles) and E. quadridigitata (open circles) in the Carolinas. The type locality of E. chamberlaini is indicated by a filled triangle. Known sites occupied by both species are shown by half-filled circles. Map is based on selected records in the ChM and NCSM collections; it includes only those records of Eurycea chamberlainifromNorthCarolinathathavebeenverified. 174 SoutheasternNaturalist Vol. 2, No. 2 Data for E. quadridigitata are available from 14 gravid females in the collections of the Charleston Museum (ChM) and the senior author (JRH) from the South Carolina Coastal Plain. The number of enlarged, yolk-laden ovarian eggs was rather variable in both species but in E. chamberlaini (gt SVL = 25.6 mm) the mean was 45.2 (35-64), whereas the mean in E. quadridigitata was 39.8 (23-62). Although the female E. quadridigitata were larger (gi SVL = 30.8 mm), they bore on average the smallest number of enlarged yolk-laden ovarian eggs. In contrast, Semlitsch and McMillan (1980) found that females (dark morph; R. D. Semlitsch, pers. comm.) from Rainbow Bay and Ellenton Bay on the Savannah River Site differed significantly in number of ovarian eggs (.t = 32.7 and 21.4 eggs, respectively), and that the larger females from Rainbow Bay produced more eggs. The mean number of eggs in 12 clutches oviposited by E. quadridigitata in the Charleston, SC, area was 21.9 (13-36) (Harrison 1973). Most of these were clutches deposited in the laboratory, so it is possible that some females did not deposit all of their eggs. There are only two other reports in the literature concerning the fecundity and/or field clutches of females from populations herein assigned to E. chamberlaini and only one of these gives the number of eggs oviposited. The mode of egg deposition is similar in the two species. Bishop (1943) states that a female sent to him from the Raleigh area by Brimley deposited 48 eggs in the laboratory. These were deposited in groups of 5-8 on the undersurfaces of damp leaves. Brimley (1923) reports finding field-deposited eggs in leafy trickles attached to leaves and debris singly or in groups of 3-6. These results are comparable to those reported by Harrison (1973) for E. quadridigitata. He observed laboratory-deposited eggs attached singly and ratherloosely to pine needles. Larval Differences. - Larvae (n = 102) of the new species were available for examination. These were compared with Goin's (1951) description of early Eurycea quadridigitata larvae, 21 specimens of hatchling larvae of the latter species from Berkeley Co., SC, and 50 Figure 6. Hatchling larva of Eurycea chamberlaini from Vance Co., NC. Scale represents1.0 mm. 2003 Julian R. Harrison, III and S.I. Guttman 175 older larvae from Bladen Co., NC. The Berkeley County larvae agree closely with the description provided by Goin (1951). As he notes, one of the most striking features of the hatchling larvae is the presence of a type of dorsal fin normally associated with ponds. A similar fin is present in hatchling larvae of the new species (Fig. 6). Hatchling E. quadridigitata are more or less uniformly grayish-brown dorsally but have pigment-free venters. There is a tendency for lateral rows of small rounded, pigment-free areas, especially ventrolaterally. There are no dorsal spots. Older larvae have lateral pigment-free areas in the form of longitudinal dashes or streaks, a characteristic absent in larvae of the new species. In hatchlings the ventral portion of the caudal fin is wider than that of the new species, especially near the vent. In older larvae of E. quadridigitata dark head markings of the postorbital and preorbital areas are more sharply delineated than those in larvae of the new species. Also, in older larvae of that species, the venter of the tail is moderately pigmented whereas in larvae of the new species the venter of the tail is only lightly pigmented or unpigmented. An average of 74% of the larvae of E. chamberlaini have symmetrically or asymmetrically paired dorsal spots (Fig. 4). The number of specimens with spots declines over the larval period. When spots are present these average 4.5 (1-7) on the right and 4.4 (1-8) on the left. The total number of spots averaged 8.9 (2-15). Spots are symmetrically arranged as pairs in 45% of the sample, asymmetrically in 55%. Hatchling larvae of E. quadridigitata are slightly larger than those of the new species averaging 6.8 mm SVL (6.5-7.0) for 10 measured specimens as compared to 6.3 mm SVL (5.5-7.0) for 10 measured specimens of E. chamberlaini. Gill rami of E. quadridigitata are longer and stockier, with the second and third pairs reaching the elbow. Those of E. chamberlaini reach just beyond the brachial mid-point. E. quadridigitata larvae are darker, more heavily pigmented. There are dense patches of melanophores with more or less uniformly dispersed pigment-free areas that come into contact but do not anastomose. Patch edges are relatively discrete. In E. chamberlaini there are anastomosing patches of melanophores enclosing pigment-free areas; patch edges are more dendritic. Concluding Remarks. - In Barnwell County, SC at the Savannah River Site, both Eurycea chamberlaini and E. quadridigitata appear to occur in similar habitats but apparently breed at somewhat different times, with the former species entering bucket traps at drift fences later in the fall. This pattern merits furtherinvestigation, including the question of whether or not different egg deposition sites are chosen. The geographic range and habitat requirements of E. chamberlaini in the Carolinas are incompletely known, but the hiatus between the geo- 176 Naturalist Southeastern Vol.2, No. 2 graphically isolated populations of E. chamberlaini in Pickens Co., SC and populations of this species in the upper Coastal Plain is perhaps an artifact of collecting as the intervening Piedmont is the least explored region of this state. With the exception of papers by Brimley (1923, 1944) the life history and ecology of E. chamberlaini is relatively unknown and should prove to be fruitful areas of research, especially at sites occupied by both the new species and E. quadridigitata. The electrophoretic data indicate that the yellow and dark morphs are extremely different from one another and are genetically isolated. However, the magnitude of the differentiation between the populations from Beauregard Parish, LA and Pickens Co., SC also suggests that the samples of Eurycea from these two sites represent distinct entities that are not interbreeding; thus, the yellow morph genetically analyzed in this study appears to be composed of two species. It is important to conduct genetic analyses of sympatric populations of the two morphs, e.g., Barnwell Co., SC. In addition, genetic studies of additional yellow morph populations are needed to determine the number and range of taxa included in this morph. ACKNOWLEDGMENTS Permission to collect specimens in South Carolina State Parks was granted to Harrison by the South Carolina Departmentof Parks and Tourism. We thank Alvin L. Braswell (North Carolina State Museum of Natural Sciences - NCSM), Albert E. Sanders (The Charleston Museum - ChM) and the late Joseph R. Bailey (Duke University - DU) for permission to study specimens housed in their respective institutions. Specimens in the DU collection have been transferred to NCSM. Specimens for genetic studies from the upper Piedmont of South Carolina were supplied by Richard R. Montanucci (Clemson University), from the Savannah River Site by R.D. Semlitsch, S. J. Morreale, and J.H. Pechmann (Savannah River Ecology Laboratory), and from North Carolina by Alvin L. Braswell (NCSM). We appreciate also the effort of others in collecting the Louisiana specimens used in genetic analyses. Alvin L. Braswell (NCSM) furnished information concerning the larvae, habitats, and distribution of the two species in North Carolina, and otherwise assisted our study in significant ways. R. D. Semlitsch (University of Missouri), J.H. Pechmann (University of New Orleans) and J. W. Gibbons (Savannah River Ecology Laboratory) provided information concerning the distribution and ecology of the two species on the SavannahRiver Site. RichardHighton (University of Maryland)reviewed an earlier version of this paper and has provided valuable information and encouragement to the senior author over the years. We also thank Paul Chippindale (University of Texas at Arlington) and Raymond D. Semlitsch (University of Missouri) for providing many helpful comments and suggestions that significantly improved the manuscript. This is contribution number 223 of the Grice Marine Biological Laboratory,College of Charleston. 2003 JulianR. Harrison,IIIandS.I. Guttman 177 LITERATURE CITED Ayala, F.J., Powell, J.R., M.L. Tracy, C.A. Mourao, and S. Perez-Salas. 1972. Enzyme variability in the Drosophila willistoni group. IV. Genic variation in naturalpopulations of Drosophila willistoni. Genetics 70:113-139. Bishop, S.C. 1943. Handbook of Salamanders: The salamanders of the United States, of Canada, and of Lower California. Comstock Publishing Company., Ithaca, NY. xiv + 555 pp. Brimley, C.S. 1923. The dwarf salamanderat Raleigh, N.C. Copeia 120:81-83. Brimley, C.S. 1944. Amphibians and Reptiles of North Carolina. Carolina Biological Supply Company, Elon College, NC (a combined version of titles from Carolina Tips, 1939-1943). Cagle, F.R. 1952. A Key to the Amphibians and Reptiles of Louisiana. Tulane University Book Store, New Orleans, LA. 42 pp. Conant, R., and J.T. Collins. 1998. A Field Guide to Reptiles and Amphibians of Eastern and Central North America, 3rd Ed., Exp. Houghton Mifflin, New York, NY. 616 pp. Cope, E.D. 1871. Catalogue of Reptilia and Batrachiaobtained by C.J. Maynard in Florida. Annual Reports of the Peabody Academy Sciences 1869:82-85. Dunn, E.R. 1926. The salamandersof the family Plethodontidae. Smith College 50th Anniversary Publ. Northampton,MA. viii + 441 pp. Folkerts, G.W. 1971. Notes on South Carolina salamanders. Journal of the Elisha Mitchell Scientific Society 87(4):206-208. Gibbons, J.W., and R.D. Semlitsch. 1991. Guide to the Reptiles and Amphibians of the SavannahRiver Site. University of Georgia Press, Athens, GA. 131 pp. Goin, C.J. 1951. Notes on the eggs and early larvae of three more Florida salamanders. Annals of the Carnegie Museum 32:253-262, plate 18. Harrison, J.R. 1973. Observations on the life history and ecology of Eurycea quadridigitata (Holbrook). HISS News-Journal 1(2):57-58 (abstract). Hass, C.A., and R. Highton. 1998. How many species are lurking in Eurycea quadridigitata (Plethodontidae: Hemidactyliini)? Abstracts for Contributed Papers to the Fourth Conference on the Biology of Plethodontid Salamanders, Highlands, NC. P. 18. Unpublished. Highton, R. 1957. Correlating costal grooves with trunk vertebrae in salamanders. Copeia 1957(2):107-109. Hilton, W.A. 1945. The Skeleton of Manculus quadridigitatus. Journal of Entomology and Zoology 37(2):59-60. Hilton, W.A. 1947. The hyobranchial skeleton of Plethodontidae. Herpetologica 3(6):191-194. Holbrook, J.E. 1842. North American Herpetology. Vol. 5, vi + 118 pp. J. Dobson and Son, Philadelphia, PA. Martof, B.S., W.M. Palmer, J.R. Bailey, and J.R. Harrison, III. 1980. Amphibians and Reptiles of Virginia and the Carolinas. University of North Carolina Press, Chapel Hill, NC. 264 pp. Mittleman, M.B. 1947. American Caudata.I. Geographic variation in Manculus quadridigitatus. Herpetologica 3(6):209-224. Mittleman, M.B. 1967. Manculus and M. quadridigitatus. Catalogue of American Amphibians and Reptiles 44.1-2. American Society of Ichthyologists and Herpetologists, Kensington, MD. 178 Southeastern Naturalist Vol. 2, No. 2 Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590. Neill, W.T. 1949. A series of Manculus from Georgia. Herpetologica 5(2):2930. Neill, W.T. 1954. Ranges and taxonomic allocations of amphibians and reptiles in the southeastern United States. Publications of the Research Division, Ross Allen's Reptile Institute 1(7):75-96. Selander, R.K., M.H. Smith, S.Y.Yang, W.E. Johnson, and J.B. Gentry. 1971. Biochemical polymorphism and systmatics in the genus Peromyscus. I. Variation in the old-field mouse (Peromyscus polionotus). Stud. Genet. 7103:49-90. Semlitsch, R.D., and M.A. McMillan. 1980. Breeding Migration, population Size structure, and reproduction of the Dwarf Salamander, Eurycea quadridigitata, in South Carolina. Brimleyana 3:97-105. Shaklee, J.B., F.W. Allendorf, D.C. Morizot, and G.S. Whitt 1990. Gene nomenclature for protein-coding loci in fish. Transactions of the American Fisheries Society 119:2-15. Swofford, D.L., and R. Selander. 1981. BIOSYS-1: A FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72:281-283. Wake, D.B. 1966. Comparative osteology and evolution of the lungless salamanders, family Plethodontidae. Memoirs of the Southern California Academy of Science 4:1-111. Appendix I. Specimens examined. In addition to the holotype and paratypes of Eurycea chamberlaini, the study included measurements taken from 37 specimens of E. quadridigitata from North Carolina as follows: Scotland County, NCSM 15401 (4), 1.25 mi SSE Wagram, NCSM 17017 (1), 2 mi SSE Wagram, NCSM 17334 (1), 3.25 mi NW Wagram, NCSM 17016 (1), 3.5 mi S Wagram; NCSM 16984 (13), 5.4 mi NE Laurinburg,NCSM 19872 (1), 5.5 mi NE Laurinburg, NCSM 18615 (2), 6 mi NE Laurinburg,NCSM 16526 (1), NCSM 19230 (2), and NCSM 19873 (1), 6.5 mi NE Laurinburg, NCSM 17015 (1), 7 mi ENE Laurinburg;Robeson County, NCSM 35995 (1) [DU A1103], Shoeheel Creek on Route 501, NCSM 35978 (2) [DU Al 104], 2 mi NE Fairmont; Bladen County, NCSM 35976 (4) [DU A2131], N.C. State Game Club, 0.25 mi toward Council from club house, NCSM 36000 (1) [DU A2172], 3 mi N Bladinboro on NC Highway 131; Columbus County, NCSM 35996 (1) [DU A8566], Cerro Gordo, and 11 specimens from South Carolina as follows: Clarendon County, JRH 3163-3166 (4) and 3193-99 (7), Santee National Wildlife Refuge, Dingle Pond.
© Copyright 2026 Paperzz