Genus Desmognathus) Population in Southern Illinois, USA

Home , Desmognathus

Herpetological Conservation and Biology 11:434–450. Submitted: 23 March 2016; Accepted: 4 October 2016; Published: 16 December 2016.

Molecular Phylogeographic Methods Reveal the Identity and Origin of a Dusky Salamander (Genus Desmognathus) Population in Southern Illinois, USA

Donald B. Shepard1,4, Nicholus Ledbetter2, Amber L. Anderson2, and Andrew R. Kuhns3

1School of Biological Sciences, Louisiana Tech University, P.O. Box 3179, Ruston, Louisiana 71272, USA 2Department of Biology, University of Central Arkansas, 201 Donaghey Avenue, LSC 180, Conway, Arkansas 72035, USA 3Illinois Natural History Survey, Prairie Research Institute, University of Illinois, 1816 South Oak Street, Champaign, Illinois 61820, USA 4Corresponding author, e-mail: [email protected]

Abstract.—Introduced species can negatively affect natural communities and ecosystems through interactions with native species. Dusky salamanders (genus Desmognathus) are commonly collected from the wild and used as fishing bait, which can result in release outside their native population or beyond the limits of the range of the species. Desmognathus conanti is the only species of the genus native to Illinois, where it occurs in Pulaski County in the extreme southern tip of the state. In 1986, a population of Desmognathus was discovered at Jug Spring, Johnson County, about 32 km north of previously known Illinois populations. We generated mitochondrial DNA sequence data for Jug Spring Desmognathus and D. conanti from Pulaski County, and combined them with DNA sequences from GenBank to determine the species identity and geographic origin of Jug Spring Desmognathus. Our analyses confirmed the species identity of Pulaski County D. conanti and showed that Jug Spring Desmognathus are D. fuscus, a species that ranges throughout the eastern U.S. but is not previously known from Illinois. Jug Spring Desmognathus were most closely related to haplotypes of D. fuscus from the Cumberland Plateau of Tennessee, pointing to this region as the likely source of the Jug Spring population. The impacts of the introduced D. fuscus on the Jug Spring ecosystem are unknown, but their presence may negatively affect invertebrates and other salamanders occupying the spring and adjacent habitats. We recommend the population be monitored and that surveys be conducted to determine if this introduced species is expanding its range. Key Words.—amphibian conservation; bait-bucket introduction; Batrachochytrium salamandrivorans (Bsal); forensic herpetology; introduced species; spring lizard

Introduction Dusky salamanders (genus Desmognathus) are com- mon inhabitants of streams and riparian habitats in the Introduced species can have profound impacts on eastern United States (U.S.), reaching their peak diver- the integrity of natural communities and ecosystems sity in the southern Appalachian Mountains (Petranka through competitive and/or predatory interactions with 1998; Bonett et al. 2007). In the southeastern U.S., Des- native species (Parker et al. 1999; Mooney and Cleland mognathus and other salamanders collectively referred 2001; Vilà et al. 2011). An important first step in man- to as spring lizards are commonly (and legally) collected aging introduced species is accurately identifying them from the wild then used or sold as fishing bait (Martof as non-native. This is relatively straightforward when 1953; Jensen and Waters 1999; Copeland et al. 2009). introduced species are easy to discriminate from native Salamanders collected as fishing bait may be transport- species and the introduced species is clearly outside its ed outside their native population or beyond the limits native range. However, it can be problematic when in- of the range of the species (Martof 1953). This human- troduced species are difficult to distinguish morphologi- mediated dispersal can result in gene flow among geo- cally from native species and/or may represent a natural graphically distant populations or the founding of a new disjunct population (Bonett et al. 2007). In such cases, extralimital population when individuals escape or are phylogeographic analysis of DNA sequences is an effec- discarded from bait buckets (Martof 1953). For ex- tive method for confirming suspected species introduc- ample, extralimital introductions of Western Tiger Sala- tions and identifying their geographic origins (Scheffer manders (Ambystoma mavortium) in California and Seal and Lewis 2001; Hebert and Cristescu 2002; McDowall Salamanders (Desmognathus monticola) in Arkansas 2008; Fitzpatrick et al. 2012). are thought to have occurred via the bait industry and

Figure 1. Map of southern Illinois showing locations of native and introduced populations of Desmognathus with inset of eastern United States showing study area (outlined in box) and Putnam County, Tennessee. bait buckets of fisherman (Bonett et al. 2007; Johnson et Fig. 1). Dutchman Lake is a popular fishing locale and al. 2011). Using molecular phylogeographic methods, Desmognathus there occur in spring-fed streams along researchers identified the Great Plains region (Kansas, Dutchman Creek near a parking lot at the end of a road New Mexico, Oklahoma, Texas) as the source of the aptly named, Fishing Hole Lane. California introductions (Johnson et al. 2011) and north- Many species of Desmognathus are undifferentiated ern Georgia as the source of the Arkansas introduction morphologically and can only be identified positively (Bonett et al. 2007). using genetic data (Tilley and Mahoney 1996; Bonett The genus Desmognathus currently comprises 21 2002; Beamer and Lamb 2008). Thus, it was difficult recognized species; however, several of these are com- to determine the species identity of Jug Spring Desmog- posed of multiple morphologically cryptic but geneti- nathus and ascertain whether they were native or intro- cally divergent lineages that likely warrant recognition duced. Using allozyme gel electrophoresis, Moeller as distinct species (Kozak et al. 2005; Beamer and Lamb (1994) compared Jug Spring Desmognathus to 17 popu- 2008; AmphibiaWeb. 2015. AmphibiaWeb: Informa- lations of D. conanti and D. fuscus from Illinois, Indi- tion on amphibian biology and conservation. Available ana, Kentucky, Ohio, and Tennessee, USA. He found from http://AmphibiaWeb.org/ [Accessed 15 December that Jug Spring Desmognathus were more similar genet- 2015]). Only one species, D. conanti (Spotted Dusky ically to D. conanti than to D. fuscus, but they were not Salamander), is native to Illinois, USA, where it is closely related to any of the populations of D. conanti known from Pulaski County in the extreme southern tip that he sampled, including those from Pulaski County, of the state along the Ohio River (Rossman 1958; Smith Illinois (Moeller 1994). The retention of dorsal spots 1961; Brandon and Huheey 1979; Phillips et al. 1999; after metamorphosis and the presence of an orange post- Fig. 1). In 1986, a population of Desmognathus was ocular stripe may help to distinguish D. conanti from D. discovered at Jug Spring, near Dutchman Lake, John- fuscus, but these characters are not consistently diagnos- son County, approximately 32 km (about 20 mi) north tic and vary ontogenetically and geographically (Ross- of the nearest Pulaski County locality (Moeller 1994; man 1958; Bonett 2002).

435 Herpetological Conservation and Biology

At the time of the study on Jug Spring Desmognathus gene. For ND2, we used the PCR primers METf.6 (Moeller 1994), D. conanti and D. fuscus were thought to (AAGCTTTCGGGCCCATACC) and COIr.8 (GCTAT- be sister taxa; both were formerly considered subspecies GTCTGGGGCTCCAATTAT) and followed the ther- of D. fuscus (Rossman 1958). However, we now have mocycler protocol from Kozak et al. (2005). For se- strong evidence from mitochondrial DNA (mtDNA) quencing ND2, we used METf.6 and a newly designed that these species are not sister taxa and that each spe- internal primer (Desmog_ND2_391F: TCAACAT- cies is actually composed of multiple, genetically diver- GACAAAARCTTGCAC). For COX1, we used the gent, phylogeographic lineages (Titus and Larson 1996; primers cox1F (CGGCCACTTTACCYRTGATAATY- Kozak et al. 2005; Beamer and Lamb 2008). Because ACTCG) and cox1R (GTATTAAGATTTCGGTCTGT- Moeller (1994) only compared Jug Spring Desmogna- TAGAAGTAT) for PCR and sequencing and followed thus to populations of D. conanti and D. fuscus, and he the thermocycler protocol from Beamer and Lamb found that Jug Spring salamanders were not closely re- (2008). We verified PCR products using gel electropho- lated to any of the populations of D. conanti that he sam- resis and cleaned them using ExoSap-IT (Affymetrix, pled, it is possible that Jug Spring Desmognathus are Santa Clara, California, USA). Sequencing reactions neither D. conanti nor D. fuscus, but are instead a dif- and automated sequencing were performed by Eurofins ferent species of Desmognathus. Thus, Moeller (1994) Genomics (Louisville, Kentucky, USA) using BigDye provided evidence that Jug Spring Desmognathus might Terminator chemistry and an ABI3730xl 96-capillary be introduced, but he was unable to resolve the species sequencer (Applied Biosystems, Foster City, California, identity and geographic origin of the population. Here, USA). We edited chromatograms by eye to verify base we use phylogeographic analysis of mtDNA sequences calls and assembled contigs from forward and reverse to determine the species identity and geographic origin sequences using Geneious v.7.1.5 (Kearse et al. 2012). of Jug Spring Desmognathus. Our results are important To determine the species identity and geographic not only for documenting the biodiversity of Illinois, but origin of Jug Spring Desmognathus, we sequenced ND2 also for the management of native species, especially D. for one individual and COX1 for all 10 individuals. We conanti, which is listed as Endangered in Illinois (Illi- also sequenced ND2 and COX1 for our sample of D. nois Endangered Species Protection Board 2015). conanti from Pulaski County, Illinois. We combined our ND2 sequences with a data set of ND2 sequences for Materials and Methods Desmognathus from Kozak et al. (2005) and Martin et al. (2016). Together these data comprised 1038 base-pairs Data collection.—We collected tissue samples for (bp) for 106 terminal taxa and included representatives DNA analysis from 10 Jug Spring Desmognathus on 13 of all currently recognized species as well as genetically March 2013. We searched for salamanders by turning divergent phylogeographic lineages within some spe- over rocks in and around streams and caught salaman- cies that may represent morphologically cryptic species ders by hand. For each individual, we clipped the last (Appendix 1). We combined our COX1 sequences with approximately 5 mm of the tail tip and placed it in a a data set of COX1 sequences for Desmognathus from uniquely labeled 1.5 ml tube of 95% ethanol. We re- Beamer and Lamb (2008) and Hibbitts et al. (2015). To- corded GPS coordinates (37.4889N, 88.9145W) and gether these data comprised 551 bp for 102 terminal taxa released salamanders where they were captured. We and included representatives of most recognized species obtained a tissue sample of D. conanti from Pulaski (13 of 21) as well as extensive geographic sampling of County, Illinois, USA (37.1313N, 89.2043W; 18 July genetic variation within several wide-ranging species 2003), from the Illinois Natural History Survey Am- (e.g., D. conanti, D. fuscus) in the southeastern U.S. phibian and Reptile Frozen Tissue Collection (INHS (Appendix 2). We used both ND2 and COX1 because Tissue 1532). Data for these samples are included in the ND2 data set has better taxonomic coverage and bet- Appendices 1 and 2. ter geographic sampling west of the Appalachians in the In the lab, we extracted genomic DNA from each Ohio River basin whereas COX1 has better geographic sample using the DNeasy Blood & Tissue Kit (Qiagen, sampling in the Atlantic and Gulf coastal plains of the Valencia, California, USA) following the protocol of USA. Because the vertebrate mitochondrial genome is the manufacturer. We quantified DNA yield using a inherited as a single haploid unit with no recombination, NanoDrop 2000 spectrophotometer (Thermo Scientific, mitochondrial genes are completely linked. Thus, ND2 Waltham, Massachusetts, USA) and standardized DNA and COX1 will share the same history. concentrations to 10–25 ng/μl. We used polymerase chain reaction (PCR) to amplify the mitochondrially Data analysis.—For both ND2 and COX1, we encoded NADH dehydrogenase subunit 2 (ND2) gene aligned sequences in Geneious using the MUSCLE al- and a portion of the cytochrome c oxidase I (COX1) gorithm and visually inspected alignments to verify an

436 Shepard et al.—Identity and origin of introduced dusky salamanders. open reading frame. We included ND2 sequence of Pha- Desmognathus and Pulaski County D. conanti to gauge eognathus hubrichti (AY728233) and COX1 sequence the relative rates of ND2 and COX1. of D. quadramaculatus (EU311649) from GenBank to Finally, we constructed a phylogenetic network in use as outgroups in phylogenetic analyses (Mueller et SplitsTree4 (Huson and Bryant 2006) using all ND2 al. 2004; Kozak et al. 2009). We used PartitionFinder sequences of the species identified as conspecific with (Lanfear et al. 2012) to determine the best partitioning Jug Spring Desmognathus. We used uncorrected (p) scheme and substitution models for each alignment. genetic distances and the NeighborNet method (Bryant Based on the Bayesian Information Criterion, the best and Moulton 2004; Huson and Bryant 2006). Phylo- partitioning scheme for the ND2 alignment was by co- genetic networks may provide additional insight about don position with the best substitution models being relationships among haplotypes that are not revealed by HKY + Γ + I for first and second positions and GTR + a phylogenetic tree, and may be especially informative Γ for third positions. The best partitioning scheme for when trying to determine the geographic origin of an the COX1 alignment was also by codon position with introduced population (Huson and Bryant 2006; Bonett the best substitution models being K80 + Γ + I for first et al. 2007; Johnson et al. 2011). With a single locus, positions, F81 for second positions, and GTR + Γ for as in our case, a split network allows us to focus on re- third positions. lationships of haplotypes within the species in question We inferred phylogenetic relationships using Bayes- while also conveying the uncertainty in those relation- ian Inference in MrBayes v.3.2 (Ronquist et al. 2012). ships (Huson and Bryant 2006). For both ND2 and COX1, we conducted two independent searches in MrBayes v.3.2 consisting of three Results heated and one cold chain for five million generations, retaining every 1,000th sample. We applied the appropri- Both ND2 and COX1 phylogenies identify Jug ate substitution model to each codon position, unlinked Spring Desmognathus as D. fuscus with high support substitution model parameters across partitions, linked (Bpp = 1.0 for ND2 and 0.99 for COX1; Figs. 2 and 3). branch lengths of partitions and scaled them relative to The 551 bp-region of COX1 we sequenced was identical each other, and used default priors. We assessed burn-in for all 10 Jug Spring individuals (Fig. 3). As expected, within runs and convergence between runs by viewing both phylogenies place Pulaski County D. conanti in plots of likelihood and parameter values in Tracer v.1.6 a clade with other D. conanti and most closely related (Rambaut et al. 2014). We also ensured that Effective to populations from western Tennessee (Bpp = 1.0 for Sample Size (ESS) for all parameters was > 200 (Drum- ND2; Fig. 2) and western Kentucky (Bpp = 0.90 for mond and Bouckaert 2015). We discarded trees sampled COX1; Fig. 3). before burn-in (20%), combined post-burn-in trees from In the ND2 phylogeny, Jug Spring Desmognathus the two independent runs, and generated a 50% majority are most closely related to D. fuscus from the Cumber- rule consensus phylogram. We determined the species land Plateau of Tennessee (Bpp = 1.0; Fig. 2). In the identity of Jug Spring Desmognathus to be the species COX1 phylogeny, Jug Spring Desmognathus are sister (or clade) with which it shares a sister-taxon relationship to a widely distributed clade of D. fuscus comprising in the ND2 and COX1 phylogenies. Relationships were populations from North Carolina to Massachusetts and considered to have high support when the Bayesian pos- westward to Indiana and Kentucky (Bpp = 0.99; Fig. terior probability (Bpp) of nodes was ≥ 0.95. 3). Sequence divergence in ND2 between Jug Spring For both ND2 and COX1, we used MEGA v.5 Desmognathus and D. fuscus from the Cumberland Pla- (Tamura et al. 2011) to calculate uncorrected pairwise teau of Tennessee (Putnam County) is 0.87% whereas sequence divergence (p-distance) between Jug Spring sequence divergence in COX1 between Jug Spring Des- Desmognathus and its sister taxon as well as between mognathus and populations of D. fuscus from its sister Jug Spring Desmognathus and Pulaski County D. conan- clade (North Carolina, Virginia, West Virginia, Massa- ti. We inferred the most likely geographic origin of Jug chusetts, Indiana, Kentucky) averages 6.13%. In com- Spring Desmognathus to be the area from which its sis- parison, sequence divergence between Jug Spring Des- ter taxon was sampled. Because geographic sampling mognathus and Pulaski County D. conanti is 10.67% for differed between ND2 and COX1 phylogenies, we used ND2 and 8.35% for COX1, indicating a slightly higher the level of sequence divergence to identify which of rate of evolution in ND2 (i.e., more substitutions per the two hypothesized geographic origins is more likely. unit time). The ND2 phylogenetic network shows Jug We recognize that ND2 and COX1 may have different Spring Desmognathus and D. fuscus from the Cumber- substitution rates (Mueller 2006) so levels of sequence land Plateau of Tennessee (Putnam and Morgan coun- divergence may not be directly comparable. However, ties) clustered together at the end of a long edge, far we can use the sequence divergence between Jug Spring separated from other haplotypes (Fig. 4).

437 Herpetological Conservation and Biology

Figure 2. Bayesian 50% consensus phylogram for species of Desmognathus based on analysis of ND2 mitochondrial DNA sequences in MrBayes. Taxa are labeled by species, county, and state. Nodes are labeled with Bayesian posterior probabilities (Bpp); asterisks (*) indicate Bpp = 100%. Jug Spring D. sp. and Illinois D. conanti are shown in bold.

438 Shepard et al.—Identity and origin of introduced dusky salamanders.

Figure 3. Bayesian 50% consensus phylogram for species of Desmognathus based on analysis of COX1 mitochondrial DNA sequences in MrBayes. Taxa are labeled by species, county, and state. Nodes are labeled with Bayesian posterior probabilities (Bpp); asterisks (*) indicates Bpp = 100%. Jug Spring D. sp. and Illinois D. conanti are shown in bold.

439 Herpetological Conservation and Biology

Figure 4. Phylogenetic network based on uncorrected (p) genetic distances between ND2 haplotypes of Desmognathus fuscus. Haplotypes are labeled by their geographic origin (County, State) and Jug Spring D. sp. is shown in bold.

Discussion Arkansas were invariable in COX1 (515 bp for seven individuals). Natural populations of Desmognathus are We found that Jug Spring Desmognathus are D. fus- often composed of multiple closely related mtDNA hap- cus, a species that ranges throughout much of the east- lotypes (Tilley et al. 2008; Wooten et al. 2010; Hibbitts ern U.S., but has not been recorded previously from Il- et al. 2015). linois (Phillips et al. 1999). Desmognathus conanti, a The phylogenetic placement of Jug Spring D. fus- species broadly distributed across the southern U.S., is cus and the low level of sequence divergence (0.87% in native to Illinois and our sample from Pulaski County ND2) from D. fuscus of the Cumberland Plateau of Ten- confirmed the identity of this population. The western nessee point to this geographic region as the likely source range limit of D. fuscus occurs in southeastern Indiana, of the Jug Spring introduction. Our results for COX1 central Kentucky, and north-central Tennessee, although were inconclusive concerning the geographic origin of the species extends along the Cumberland River into Jug Spring D. fuscus, but this data set lacked samples western Kentucky where it forms a contact zone with from the Cumberland Plateau (Beamer and Lamb 2008). D. conanti (Bonett 2002; AmphibiaWeb. 2015. op. cit.). Sequence divergence in COX1 between Jug Spring D. Jug Spring D. fuscus are thus outside the known native fuscus and other populations of D. fuscus was consider- distribution of D. fuscus. If Jug Spring D. fuscus rep- ably higher (mean 6.13%) than sequence divergence in resent an isolated remnant population, as opposed to an ND2 between Jug Spring D. fuscus and D. fuscus from extralimital introduction, then we would predict signifi- the Cumberland Plateau (0.87%). Because ND2 had a cant genetic divergence from other populations of D. higher rate of evolution than COX1, a difference in rates fuscus. This appeared to be the case in analyses with cannot explain the disparity in sequence divergence ob- COX1, but analyses with ND2, which had better sam- served between Jug Spring D. fuscus and their sister tax- pling throughout the Ohio River basin, showed that Jug on in each phylogeny. Instead, this disparity is primarily Spring D. fuscus are closely related to and only slightly due to variation in geographic sampling and the absence divergent from populations of D. fuscus from the Cum- of closely related haplotypes in the COX1 data set. The berland Plateau of Tennessee. The portion of COX1 level of ND2 sequence divergence between Jug Spring that we sequenced was also invariable among the 10 Jug D. fuscus and D. fuscus from Putnam County, Tennessee Spring Desmognathus we examined, which would be (0.87%) is less than the sequence divergence between predicted by a founder effect and serves as additional Cumberland Plateau populations of D. fuscus from Put- support that the population is introduced (Fitzpatrick nam and Morgan counties (1.65%), which are separated et al. 2012). Similarly, Bonett et al. (2007) found that by about 64 km (Ron Bonett and Ken Kozak, unpubl. D. monticola introduced to the Ozarks of northwestern data). Species of Desmognathus often show a strong

440 Shepard et al.—Identity and origin of introduced dusky salamanders. pattern of Isolation-by-Distance, meaning that genetic populations largely consists of agricultural and wetland distance and geographic distance between populations habitats, which would impede dispersal and gene flow. are positively correlated (Tilley 2016). Based on this, In the 30 y since the discovery of the Jug Spring Des- we estimate that the source population of Jug Spring D. mognathus, no additional localities along Dutchman fuscus is fairly close (< 64 km) to the Putnam County, Creek have been reported, but we are unaware of any Tennessee locality. targeted surveys. Therefore, we recommend that sur- Moeller (1994) provided evidence that Jug Spring veys be conducted along Dutchman Creek to ensure that Desmognathus were introduced, but his ability to posi- the introduced population of D. fuscus is not expanding tively identify the species and geographic origin was its range from Jug Spring. The population should also hampered by limited sampling (11 populations of D. be monitored periodically and potential impacts on na- conanti and six populations of D. fuscus). He found that tive species should be assessed. Jug Spring Desmognathus were not closely related to Using salamanders as fishing bait is legal in many any of the populations that he sampled, but they were states, including Tennessee, the identified source of Jug more similar genetically to D. conanti than D. fuscus Spring D. fuscus. Other than limits on the species and (Moeller 1994). However, Moeller (1994) sampled only number of individuals that can be collected, most states six populations of D. fuscus from southern Ohio, south- have few regulations on this practice (Nanjappa and Con- ern Indiana, and northern Kentucky. Our phylogenetic rad 2011). Transportation of salamanders outside their and network analyses with ND2 showed that Cumber- native population may facilitate the spread of pathogens, land Plateau populations of D. fuscus form a genetically yet most states do not require disease/pathogen testing divergent lineage apart from other D. fuscus, including for exported or imported animals or those used as bait those from Kentucky and Ohio. Bonett et al. (2007) (Picco and Collins 2008; Nanjappa and Conrad 2011). emphasized the importance of adequate sampling of ge- In response to threats posed by the fungal pathogen, Ba- netic diversity across the range of a species when using trachochytrium salamandrivorans (Bsal), the U.S. Fish molecular-based methods to determine whether a popu- and Wildlife Service (2016) recently listed 201 species lation is introduced and whence it came. Moeller (1994) of salamanders (67 species native to the U.S.) as Injuri- lacked samples from the Cumberland Plateau of Tennes- ous Wildlife under the Lacey Act, which prohibits im- see, which proved critical for determining the origin of portation and interstate transport. Species of the genera Jug Spring D. fuscus in our study. The importance of Desmognathus and Eurycea are commonly used as bait, sampling is further exemplified by our more ambiguous but remain unregulated by this ruling due to a lack of results based on COX1 compared to ND2, of which only data (Copeland et al. 2009; U.S. Fish and Wildlife Ser- the latter included samples from the Cumberland Pla- vice 2016). If species of these genera are eventually teau of Tennessee. The distribution of this lineage of D. found to be susceptible to Bsal or to be potential carriers fuscus (fuscus A clade of Kozak et al. (2005)) is not well of the fungus, then interstate transport would likewise established and more sampling in the region would help be prohibited under federal law with individual states to bolster our conclusions. regulating transport within their own borders. Regard- The impacts of the introduced D. fuscus on the Jug less of whether this happens, the practice of using sala- Spring ecosystem are unknown. Species of Desmogna- manders as bait warrants increased regulation given the thus are predators of invertebrates and other salaman- multitude of potentially negative impacts. To minimize ders, and their presence may negatively affect naïve these impacts without impinging greatly on individual native fauna inhabiting the spring and cave. Peck and fishermen, we recommend that the use and sale of sala- Lewis (1978) did not document any threatened or en- manders as bait be restricted to the drainage basin (HUC dangered invertebrates in Jug Spring Cave during their 4 or 6) from which the salamanders were collected. surveys of subterranean invertebrate fauna of Illinois, Many species of salamanders commonly used as bait in but Cave Salamanders (Eurycea lucifuga) occur within the eastern U.S. are associated with streams and genetic the cave system (Matt Niemiller, pers. comm.). West- data indicate that population connectivity in these spe- ern Tiger Salamanders (Ambystoma mavortium) intro- cies is largely determined by drainage patterns (Jones duced into California have been shown to have deleteri- et al. 2006; Kozak et al. 2006; Kuchta et al. 2016). Be- ous effects on native California Tiger Salamanders (A. cause salamander populations within the same drainage californiense) through displacement and hybridization basin are more likely to be linked by dispersal naturally, (Riley et al. 2003; Ryan et al. 2009). Desmognathus fus- any human-mediated dispersal within a basin would cus and D. conanti hybridize where their ranges contact likely have little to no impact. in western Kentucky (Bonett 2002), but Jug Spring D. Genetic data allow identification of species when fuscus are about 32 km from the nearest locality for D. morphology is uninformative and provide a means to conanti in Illinois. Furthermore, the area between these determine the geographic origin of introduced species

441 Herpetological Conservation and Biology as well as trace their invasion route (Liu et al. 2006; Brandon, R.A., and J.E. Huheey. 1979. Distribution Baker 2008; McDowall 2008; Guillemaud et al. 2010; of the Dusky Salamander, Desmognathus fuscus Huffman and Wallace 2012). In addition to generating (Green) in Illinois. Chicago Academy of Sciences genetic data from populations under study, research- Natural History Miscellanea 205:1–7. ers also frequently need reference genetic data from Bryant D., and V. Moulton. 2004. NeighborNet: an other species or populations for comparison (McDowall agglomerative method for the construction of 2008). The availability of reference genetic data and the phylogenetic networks. Molecular Biology and geographic scale of sampling can limit the success of Evolution 21:255–265. these approaches (Bonett et al. 2007). For these reasons, Copeland, J.E., G.L. Mears, and R.S. Caldwell. 2009. publicly available genetic databases (e.g., GenBank) are Salamanders as fishing bait in the Blue Ridge invaluable resources. However, for the utility of these physiographic province of east Tennessee. Journal of databases to be maximized, researchers not only need to the Tennessee Academy of Science 84:52–54. submit their data, but it is imperative that they provide Drummond, A.J., and R.R. Bouckaert. 2015. Bayesian locality information or make it easy to connect genetic Evolutionary Analysis with BEAST. Cambridge and locality data in their publications (Pope et al. 2015). University Press, Cambridge, UK. The fact that our study would not have been possible Fitzpatrick, B.M., J.A. Fordyce, M.L. Niemiller, and without the genetic and locality data made available by R.G. Reynolds. 2012. What can DNA tell us about other researchers underscores this point. biological invasions? Biological Invasions 14:245– 253. Acknowledgments.—We thank Ethan Kessler and Guillemaud, T., M.A. Beaumont, M. Ciosi, J.M. Cody Roden for help with fieldwork and we thank Chris Cornuet, and A. Estoup. 2010. Inferring introduction Phillips (INHS) for loaning a tissue sample of Desmog- routes of invasive species using approximate nathus conanti from Pulaski County, Illinois. We also Bayesian computation on microsatellite data. thank Scott Ballard for alerting us to Moeller’s Mas- Heredity 104:88–99. ter’s thesis at SIU-Carbondale and Matt Niemiller for Hebert, P.D., and M.E. Cristescu. 2002. Genetic information on the fauna of Jug Spring Cave. We thank perspectives on invasions: the case of the Cladocera. Toby Hibbitts and Gary Voelker for providing DNA Canadian Journal of Fisheries and Aquatic Sciences sequences of Texas Desmognathus conanti from their 59:1229–1234. study. Finally, we thank Ron Bonett, Ken Kozak, Ben Hibbitts, T.J., S.A. Wahlberg, and G. Voelker. 2015. Lowe, and Steve Tilley for providing additional tissues, Resolving the identity of Texas Desmognathus. DNA sequences, and/or locality data for some species of Southeastern Naturalist 14:213–220. Desmognathus. Field collections were made under per- Huffman, J.E., and J.R. Wallace. 2012. Wildlife mit from the Illinois Department of Natural Resources Forensics: Methods and Applications. John Wiley & and under University of Illinois IACUC protocol 14000. Sons, Hoboken, New Jersey, USA. Funding was provided by the University of Central Ar- Huson, D.H., and D. Bryant. 2006. Application of kansas. phylogenetic networks in evolutionary studies. Molecular Biology and Evolution 23:254–267. Literature Cited Illinois Endangered Species Protection Board. 2015. Checklist of Endangered and Threatened Baker, B.W. 2008. A brief overview of forensic Animals and Plants of Illinois. Illinois Endangered herpetology. Applied Herpetology 5:307–318. Species Protection Board, Springfield, Illinois. Beamer, D.A., and T. Lamb. 2008. Dusky salamanders 18 p. Available at: http://www.dnr.illinois.gov/ (Desmognathus, Plethodontidae) from the Coastal ESPB/Documents/2015_ChecklistFINAL_for_ Plain: multiple independent lineages and their webpage_051915.pdf. bearing on the molecular phylogeny of the genus. Jensen, J.B., and C. Waters. 1999. The ‘spring lizard’ bait Molecular Phylogenetics and Evolution 47:143–153. industry in the state of Georgia, USA. Herpetological Bonett, R.M. 2002. Analysis of the contact zone between Review 30:20–21. the dusky salamanders Desmognathus fuscus Johnson, J.R., R.C. Thomson, S.J. Micheletti, and H.B. fuscus and Desmognathus fuscus conanti (Caudata: Schaffer. 2011. The origin of Tiger Salamander Plethodontidae). Copeia 2002:344–355. (Ambystoma tigrinum) populations in California, Bonett, R.M., K.H. Kozak, D.R. Vieites, A. Bare, J.A. Oregon, and Nevada: introductions or relics? Wooten, and S.E. Trauth. 2007. The importance Conservation Genetics 12:355–370. of comparative phylogeography in diagnosing Jones, M.T., S.R. Voss, M.B. Ptacek, D.W. Weisrock, introduced species: a lesson from the seal salamander, and D.W. Tonkyn. 2006. River drainages and Desmognathus monticola. BMC Ecology 7:7. phylogeography: an evolutionary significant

442 Shepard et al.—Identity and origin of introduced dusky salamanders.

lineage of shovel-nosed salamander (Desmognathus Mooney, H.A., and C.E. Cleland. 2001. The evolutionary marmoratus) in the southern Appalachians. impact of invasive species. Proceedings of the Molecular Phylogenetics and Evolution 38:280–287. National Academy of Sciences USA 98:5446–5451. Kearse, M., R. Moir, A. Wilson, S. Stones-Havas, M. Mueller, R.L. 2006. Evolutionary rates, divergence Cheung, S. Sturrock, S. Buxton, A. Cooper, S. dates, and the performance of mitochondrial genes in Markowitz, C. Duran, et al. 2012. Geneious Basic: an Bayesian phylogenetic analysis. Systematic Biology integrated and extendable desktop software platform 55:289–300. for the organization and analysis of sequence data. Mueller, R.L., J.R. Macey, M. Jaekel, D.B. Wake, and Bioinformatics 28:1647–1649. J.L. Boore. 2004. Morphological homoplasy, life Kozak, K.H., R.A. Blaine, and A. Larson. 2006. Gene history evolution, and historical biogeography of lineages and eastern North American palaeodrainage plethodontid salamanders inferred from complete basins: phylogeography and speciation in mitochondrial genomes. Proceedings of the National salamanders of the Eurycea bislineata species Academy of Sciences USA 101:13820–13825. complex. Molecular Ecology 15:191–207. Nanjappa, P., and P.M. Conrad. 2011. State of the Kozak, K.H., A. Larson, R.M. Bonett, and L.J. Harmon. Union: Legal Authority Over the Use of Native 2005. Phylogenetic analysis of ecomorphological Amphibians and Reptiles in the United States. divergence, community structure, and diversification Version 1.03. Association of Fish and Wildlife rates in dusky salamanders (Plethodontidae: Agencies, Washington, D.C., USA. Desmognathus). Evolution 59:2000–2016. Parker, I.M., D. Simberloff, W.M. Lonsdale, K. Goodell, Kozak, K.H., R.W. Mendyk, and J.J. Wiens. 2009. Can M. Wonham, P.M. Kareiva, M.H Williamson, B. Von parallel diversification occur in sympatry? Repeated Holle, P.B. Moyle, J.E. Byers, and L. Goldwasser. patterns of body-size evolution in coexisting clades 1999. Impact: toward a framework for understanding of North American salamanders. Evolution 67:1769– the ecological effects of invaders. Biological 1784. Invasions 1:3–19. Kuchta, S.R., M. Haughey, A.H. Wynn, J.F. Jacobs, Peck, S.J., and J.J. Lewis. 1978. Zoogeography and and R. Highton. 2016. Ancient river systems evolution of the subterranean invertebrate faunas and phylogeographical structure in the spring of Illinois and southeastern Missouri. National salamander, Gyrinophilus porphyriticus. Journal of Speleological Bulletin 40:1–39. Biogeography 43:639–652. Petranka, J.W. 1998. Salamanders of the United Lanfear, R., B. Calcott, S.Y.W. Ho, and S. Guindon. States and Canada. Smithsonian Institution Press, 2012. PartitionFinder: combined selection of Washington, D.C., USA. partitioning schemes and substitution models for Phillips, C.A., R.A. Brandon, and E.O. Moll. 1999. phylogenetic analysis. Molecular Biology and Field guide to amphibians and reptiles in Illinois. Evolution 29:1695–1701. Illinois Natural History Survey Manual 8:1–300. Liu, J., M. Dong, S.L. Miao, Z.Y. Li, M.H. Song, and Picco, A.M., and J.P. Collins. 2008. Amphibian R.Q. Wang. 2006. Invasive alien plants in China: commerce as a likely source of pathogen pollution. role of clonality and geographical origin. Biological Conservation Biology 22:1582–1589. Invasions 8:1461–1470. Pope, L.C., L. Liggins, J. Keyse, S.B. Carvalho, and C. Martin, S.D., D.B. Shepard, M.A. Steffen, J.G. Riginos. 2015. Not the time or the place: the missing Phillips, and R.M. Bonett. 2016. Biogeography and spatio-temporal link in publicly available genetic colonization history of plethodontid salamanders data. Molecular Ecology 24:3802–3809. from the Interior Highlands of eastern North Riley, S.P.D., H.B. Shaffer, S.R. Voss, and B.J. America. Journal of Biogeography 43:410–422. Fitzpatrick. 2003. Hybridization between a rare Martof, B.S. 1953. The “spring lizard” industry: a factor native tiger salamander (Ambystoma californiense) in salamander distribution and genetics. Ecology and its introduced congener. Ecological Applications 34:436–437. 13:1263–1275. McDowall, I.L. 2008. DNA technology and its Rambaut A., M.A. Suchard, D. Xie, and A.J. Drummond. applications in herpetological research and forensic 2014. Tracer v.1.6, Available from http://beast.bio. investigations involving reptiles and amphibians. ed.ac.uk/Tracer Applied Herpetology 5:371–385. Ronquist, F., M. Teslenko, P. van der Mark, D.L. Ayres, Moeller, H.C. 1994. Status of a Johnson County, A. Darling, S. Höhna, B. Larget, L. Liu, M.A. Illinois, population of dusky salamanders, genus Suchard, and J.B. Huelsenbeck. 2012. MrBayes Desmognathus (Amphibia: Plethodontidae). M.Sc. 3.2: efficient Bayesian phylogenetic inference and Thesis, Southern Illinois University, Carbondale, model choice across a large model space. Systematic Illinois, USA. 47 p. Biology 61:539–542.

443 Herpetological Conservation and Biology

Rossman, D.A. 1958. A new race of Desmognathus ochrophaeus complex (Amphibia: Plethodontidae). fuscus from the south-central United States. Herpetological Monographs 10:1–42. Herpetologica 11:158–160. Tilley, S.G., R.L. Eriksen, and L.A. Katz. 2008. Ryan, M.E., J.R. Johnson, and B.M. Fitzpatrick. 2009. Systematics of dusky salamanders, Desmognathus Invasive hybrid Tiger Salamander genotypes impact (Caudata: Plethodontidae), in the mountain and native amphibians. Proceedings of the National Piedmont regions of Virginia and North Carolina, Academy of Sciences USA 106:11166–11171. USA. Zoological Journal of the Linnean Society Scheffer, S.J., and M.L. Lewis. 2001. Two nuclear 152:115–130. genes confirm mitochondrial evidence of cryptic Titus, T.A., and A. Larson. 1996. Molecular species within Liriomyza huidobrensis (Diptera: phylogenetics of the desmognathine salamanders Agromyzidae). Annals of the Entomological Society (Caudata: Plethodontidae): a reevaluation of of America 94:648–653. evolution in ecology, life history, and morphology. Smith, P.W. 1961. The amphibians and reptiles of Systematic Biology 45:451–472. Illinois. Illinois Natural History Survey Bulletin U.S. Fish and Wildlife Service. 2016. Injurious wildlife 28:1–298. species; listing salamanders due to risk of salamander Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. chytrid fungus. Federal Register 81:1534–1556. Nei, and S. Kumar. 2011. MEGA5: Molecular Vilà, M., J.L. Espinar, M. Hejda, P.E. Hulme, V. Jarošík, evolutionary genetics analysis using maximum J.L. Maron, J. Pergl, U. Schaffner, Y. Sun, and P. likelihood, evolutionary distance, and maximum Pyšek. 2011. Ecological impacts of invasive alien parsimony methods. Molecular Biology and plants: a meta-analysis of their effects on species, Evolution 28:2731–2739. communities and ecosystems. Ecology Letters Tilley, S.G. 2016. Patterns of genetic differentiation in 14:702–708. woodland and dusky salamanders. Copeia 104:8–20. Wooten, J.A., C.D. Camp, and L.J. Rissler. 2010. Genetic Tilley, S.G., and M.J. Mahoney. 1996. Patterns of genetic diversity in a narrowly endemic, recently described differentiation in salamanders of the Desmognathus dusky salamander, Desmognathus folkertsi, from the southern Appalachian Mountains. Conservation Genetics 11:835–854.

Donald B. Shepard is an Assistant Professor in the School of Biological Sciences at Louisiana Tech University, USA. His research employs molecular methods and geospatial tools to examine patterns of genetic variation within species, identify cryptic diversity, and understand the processes that drive ecological and evolutionary diversification of amphibians and reptiles. (Photographed by Alex Pyron).

Nicholus Ledbetter is currently a Ph.D. student at the University of Tulsa in Tulsa, Oklahoma, USA. He earned his B.S. at the University of Central Arkansas and was a participant in the NSF summer REU program at Kansas State University. His dissertation research focuses on population genetics of the Oklahoma Salamander (Eurycea tynerensis). (Photographed by Trevor Burton).

Amber L. Anderson is currently a M.S. student at the University of Central Arkansas in Conway, Arkansas, USA. She earned her B.S. from the University of Arkansas at Little Rock. Her thesis research is a multiyear mark-recapture study of the Ouachita Dusky Salamander (Desmognathus brimleyorum) that aims to identify the factors that influence dispersal and determine how dispersal contributes to population dynamics of semi-aquatic salamanders. (Photographed by Chris Robin- son).

Andrew R. Kuhns is the Herpetologist for the Biotic Survey and Assessment Program at the Illinois Natural History Survey within the Prairie Research Institute at the University of Illinois Urbana-Champaign, USA. He conducts surveys and habitat assessments for threatened and endangered amphibians and reptiles in areas scheduled for transportation improvements. Thus, he spends more time than the average person looking for and identifying road kill. As time and funding allows, he pursues independent research pertaining to the distribution, ecology, and conservation of amphibians and reptiles. (Photographed by Deb Maurer).

444 Shepard et al.—Identity and origin of introduced dusky salamanders.

Appendix 1. Specimen and locality information for the ND2 data set with GenBank number or source for each sequence. Species Voucher#/Museum# County/Parish, State GenBank Accession# D. abditus JFBM17798 Cumberland, Tennessee KR732330 D. aeneus KHK51 Gilmer, Georgia AY612342 D. aeneus BTL237 Macon, North Carolina KR732331 D. apalachicolae BTL238 Liberty, Florida KX764600 D. apalachicolae KHK157 Liberty, Florida AY612373 D. auriculatus BTL239 Wakulla, Florida KR732333 D. brimleyorum RMB2173 Garland, Arkansa AY612420 D. brimleyorum FC11578 LeFlore, Oklahoma AY612419 D. brimleyorum RMB2518 Montgomery, Arkansas AY612423 D. brimleyorum RMB2327 Nevada, Arkansas AY612422 D. brimleyorum RMB2201 Polk, Arkansas AY612421 D. brimleyorum MVZS14353 Scott, Arkansas AY612418 D. carolinensis KHK72 Buncombe, North Carolina AY612368 D. carolinensis KHK80 Buncombe, North Carolina AY612369 D. carolinensis KHK118 McDowell, North Carolina AY612372 D. carolinensis KHK103 Yancey, North Carolina AY612371 D. carolinensis KHK85 Yancey, North Carolina AY612370 D. conanti KHK8.399 Abbeville, South Carolina AY612381 D. conanti ASU23228 Anderson, South Carolina AY612382 D. conanti KHK163 Ballard, Kentucky AY612387 D. conanti KHK662 Hardeman, Tennessee AY612386 D. conanti KHK227 Henderson, Tennessee AY612388 D. conanti ASU23805 Jasper, Mississippi AY612415 D. conanti ASU23806 Jasper, Mississippi Kozak et al. 2005 D. conanti RMB251 Lewis, Tennessee AY612389 D. conanti RMB221 Limestone, Alabama AY612384 D. conanti FLN1532 Pulaski, Illinois KX764602 D. conanti TJR2470 Richmond, Georgia KX764601 D. conanti RMB225 Tallapoosa, Alabama AY612383 D. conanti RMB275 Tishomingo, Mississippi AY612385 D. conanti RMB239 Washington, Louisiana AY612390 D. folkertsi KHK340 Union, Georgia AY612351 D. fuscus KHK503 Alleghany, North Carolina AY612408 D. fuscus RMB630 Barbour, West Virginia AY612400 D. fuscus KHK531 Bath, West Virginia AY612403 D. fuscus RMB595 Belmont, Ohio AY612399 D. fuscus RMB831 Bradford, Pennsylvania AY612396 D. fuscus RMB720 Campbell, Virginia AY612402 D. fuscus FC13580 Duplin, North Carolina AY612414 D. fuscus KHK141 Fairfield, South Carolina AY612392 D. fuscus KHK142 Fairfield, South Carolina AY612393 D. fuscus MVZS12956 Franklin, Massachusetts AY612395

445 Herpetological Conservation and Biology

Appendix 1 (continued). Specimen and locality information for the ND2 data set with GenBank number or source for each sequence.

Species Voucher#/Museum# County/Parish, State GenBank Accession# D. fuscus KHK314 Grayson, Virginia AY612409 D. fuscus KHK570 Greene, Virginia KR826999 D. fuscus RMB696 Guliford, North Carolina AY612394 D. fuscus FLN8224 Johnson, Illinois KX764603 D. fuscus KHK468 Lee, North Carolina AY612407 D. fuscus KHK423 Menifee, Kentucky AY612398 D. fuscus KHK429 Morgan, Tennessee AY612411 D. fuscus RMB339 Oldham, Kentucky AY612397 D. fuscus RMB716 Pittsylvania, Virginia AY612401 D. fuscus RMB513 Putnam, Tennessee AY612410 D. fuscus KHK557 Rockingham, Virginia AY612404 D. fuscus KHK434 Rutherford, North Carolina AY612406 D. fuscus KHK435 Rutherford, North Carolina Kozak et al. 2005 D. fuscus RMB2331 Surry, North Carolina AY612412 D. fuscus RMB2332 Surry, North Carolina AY612413 D. fuscus RMB743 Watauga, North Carolina AY612405 D. imitator WRK Jackson, North Carolina KX764604 D. imitator KHK05 Sevier, Tennessee AY612343 D. marmoratus KHK366 Caldwell, North Carolina AY612345 D. marmoratus KHK18 Graham, North Carolina AY612344 D. marmoratus CC44 Rabun, Georgia KR827000 D. marmoratus KHK90 Yancey, North Carolina AY612346 D. monticola KHK65 Buncombe, North Carolina AY612376 D. monticola KHK782 Butler, Alabama AY612379 D. monticola JFBM17794 Graham, North Carolina KX764605 D. monticola KHK16 Graham, North Carolina AY612375 D. monticola KHK60 Transylvania, North Carolina AY612377 D. monticola KHK134 Unicoi, Tennessee AY612374 D. monticola S13225 Westmoreland, Pennsylvania AY612378 D. ochrophaeus RMB1224 Overton, Tennessee AY612367 D. ochrophaeus WKS05 Tompkins, New York AY612366 D. ocoee KHK31 Clay, North Carolina AY612353 D. ocoee KHK44 Clay, North Carolina AY612354 D. ocoee KHK154 Cocke, Tennessee AY612356 D. ocoee KHK22 Graham, North Carolina AY612352 D. ocoee RMB268 Jackson, Alabama AY612362 D. ocoee KHK56 Jackson, North Carolina AY612361 D. ocoee KHK266 Marion, Tennessee AY612360 D. ocoee KHK53 Rabun, Georgia AY612358 D. ocoee RMB2335 Rabun, Georgia AY612359 D. ocoee KHK01 Sevier, Tennessee AY612355

446 Shepard et al.—Identity and origin of introduced dusky salamanders.

Appendix 1 (continued). Specimen and locality information for the ND2 data set with GenBank number or source for each sequence.

Species Voucher#/Museum# County/Parish, State GenBank Accession# D. ocoee KHK62 Transylvania, North Carolina AY612357 D. orestes KHK129 Avery, North Carolina AY612363 D. orestes KHK305 Grayson, Virginia Kozak et al. 2005 D. orestes KHK306 Grayson, Virginia AY612365 D. orestes KHK140 Unicoi, Tennessee AY612364 D. organi KHK73 Buncombe, North Carolina AY612341 D. organi KHK310 Grayson, Virginia KR827001 D. planiceps ST11008 Patrick, Virginia KR732337 D. planiceps ST10865 Pittsylvania, Virginia KX764606 D. quadramaculatus KHK593 Buncombe, North Carolina KR827003 D. quadramaculatus KHK369 Caldwell, North Carolina AY612347 D. quadramaculatus KHK52 Rabun, Georgia AY612350 D. quadramaculatus RMB2329 Surry, North Carolina AY612349 D. quadramaculatus KHK135 Unicoi, Tennessee AY612348 D. quadramaculatus KHK499 Wilkes, North Carolina KR827002 D. quadramaculatus CC14 KX764607 D. santeetlah JFBM18737 Jackson, North Carolina KR732338 D. santeetlah MVZS11775 Jackson, North Carolina AY612391 D. welteri FC14355 Letcher, Kentucky AY612416 D. welteri KHK414 Powell, Kentucky AY612417 D. wrighti JFBM17195 Clay, North Carolina KX764608 D. wrighti JFBM16100 Swain, North Carolina KR732339 Phaeognathus hubrichti FC13612 Butler, Alabama AY728233

Appendix 2. Specimen and locality information for the COX1 data set with GenBank number or source for each sequence. Species Voucher#/Museum# County/Parish, State GenBank Accession# D. apalachicolae DAB861 Leon, Florida EU311708 D. apalachicolae BTL238 Liberty, Florida KX764609 D. apalachicolae DAB218 Liberty, Florida EU311666 D. auriculatus DAB349 Baker, Florida EU311681 D. auriculatus DAB1385 Clinch, Georgia EU311650 D. auriculatus DAB348 Liberty, Georgia EU311680 D. auriculatus BTL239 Wakulla, Florida KX764610 D. brimleyorum FC11578 LeFlore, Oklahoma KX764611 D. brimleyorum RMB2327 Nevada, Arkansas KX764612 D. brimleyorum RMB2201 Polk, Arkansas KX764613 D. carolinensis DAB1105 Buncombe, North Caolina EU311642 D. carolinensis DAB946 Yancey, North Carolina EU311713 D. conanti DAB324 Amite, Mississippi EU311674 D. conanti DAB346 Baldwin, Alabama EU311678 D. conanti DAB252 Barnwell, South Carolina EU311668

447 Herpetological Conservation and Biology

Appendix 2 (continued). Specimen and locality information for the COX1 data set with GenBank number or source for each sequence.

Species Voucher#/Museum# County/Parish, State GenBank Accession# D. conanti DAB345 Butler, Alabama EU311677 D. conanti DAB1387 Effingham, Georgia EU311651 D. conanti DAB647 Grant, Louisiana EU311699 D. conanti DAB646 Henderson, North Carolina EU311698 D. conanti ASU23806 Jasper, Mississippi KX764614 D. conanti DAB322 Jasper, Mississippi EU311672 D. conanti DAB438 Jasper, Mississippi EU311685 D. conanti DAB922 Lawrence, Alabama EU311712 D. conanti DAB222 Livingston, Kentucky EU311667 D. conanti TJH2756 Newton, Texas Hibbitts et al. 2015 D. conanti TJH2757 Newton, Texas Hibbitts et al. 2015 D. conanti TJH3262 Newton, Texas Hibbitts et al. 2015 D. conanti FLN1532 Pulaski, Illinois KX764615 D. conanti TJR2470 Richmond, Georgia KX764616 D. conanti TJH3266 Sabine, Texas Hibbitts et al. 2015 D. conanti TJH3269 Sabine, Texas Hibbitts et al. 2015 D. conanti TJH3270 Sabine, Texas Hibbitts et al. 2015 D. conanti DAB347 Santa Rosa, Florida EU311679 D. conanti TCWC94726 Tyler, Texas Hibbitts et al. 2015 D. conanti TJH2696 Tyler, Texas Hibbitts et al. 2015 D. conanti DAB435 Washington, Florida EU311684 D. conanti DAB323 Washington, Louisiana EU311673 D. conanti DAB867 Wayne, Georgia EU311709 D. conanti DAB868 Wayne, Georgia EU311710 D. conanti DAB321 West Feliciana, Louisiana EU311671 D. fuscus DAB881 Bamberg, South Carolina EU311711 D. fuscus DAB1039 Bath, Kentucky EU311640 D. fuscus DAB201 Beaufort, North Carolina EU311664 D. fuscus DAB637 Berkeley, South Carolina EU311695 D. fuscus DAB1485 Bladen, North Carolina EU311655 D. fuscus DAB508 Bladen, North Carolina EU311687 D. fuscus DAB1042 Bland, Virginia EU311641 D. fuscus DAB755 Burke, North Carolina EU311705 D. fuscus DAB715 Caldwell, North Carolina EU311702 D. fuscus DAB265 Calhoun, South Carolina EU311669 D. fuscus DAB1487 Carteret, North Carolina EU311656 D. fuscus DAB501 Colleton, South Carolina EU311686 D. fuscus DAB209 Craven, North Carolina EU311665 D. fuscus DAB414 Craven, North Carolina EU311682 D. fuscus DAB1505 Davidson, North Carolina EU311659 D. fuscus DAB1484 Davie, North Carolina EU311654

448 Shepard et al.—Identity and origin of introduced dusky salamanders.

Appendix 2 (continued). Specimen and locality information for the COX1 data set with GenBank number or source for each sequence.

Species Voucher#/Museum# County/Parish, State GenBank Accession# D. fuscus DAB1478 Duplin, North Carolina EU311653 D. fuscus FC13580 Duplin, North Carolina KX764617 D. fuscus DAB434 Edgecombe, North Carolina EU311683 D. fuscus DAB806 Florence, South Carolina EU311707 D. fuscus MVZ224931 Franklin, Massachusetts AY728227 D. fuscus DAB1506 Iredell, North Carolina EU311660 D. fuscus DAB1036 Jefferson, Indiana EU311639 D. fuscus FLN8223 Johnson, Illinois KX764619 D. fuscus FLN8224 Johnson, Illinois KX764621 D. fuscus FLN8225 Johnson, Illinois KX764622 D. fuscus FLN8226 Johnson, Illinois KX764620 D. fuscus FLN8227 Johnson, Illinois KX764618 D. fuscus FLN8228 Johnson, Illinois KX764623 D. fuscus FLN8229 Johnson, Illinois KX764624 D. fuscus FLN8230 Johnson, Illinois KX764625 D. fuscus FLN8231 Johnson, Illinois KX764626 D. fuscus FLN8232 Johnson, Illinois KX764627 D. fuscus DAB1488 Montgomery, North Carolina EU311657 D. fuscus DAB1496 Montgomery, North Carolina EU311658 D. fuscus DAB1545 New Hanover, North Carolina EU311663 D. fuscus DAB290 Pitt, North Carolina EU311670 D. fuscus DAB972 Pitt, North Carolina EU311719 D. fuscus DAB603 Randolph, West Virginia EU311694 D. fuscus DAB596 Rockingham, Virginia EU311692 D. fuscus DAB638 Scotland, North Carolina EU311696 D. fuscus DAB782 Scotland, North Carolina EU311706 D. fuscus DAB526 Watauga, North Carolina EU311689 D. fuscus DAB1517 Wilkes, North Carolina EU311662 D. monticola DAB642 Graham, North Carolina EU311697 D. monticola DAB1346 Lumpkin, Georgia EU311646 D. monticola — Monroe, Alabama AY549708 D. monticola — Monroe, Alabama AY549717 D. monticola DAB571 Monroe, Tennessee EU311690 D. monticola DAB954 Transylvania, North Carolina EU311717 D. monticola DAB1256 Union, Georgia EU311644 D. monticola DAB524 Watauga, North Carolina EU311688 D. ochrophaeus DAB602 Randolph, West Virginia EU311693 D. ocoee DAB1406 Douglas, Georgia EU311652 D. ocoee DAB1352 Lumpkin, Georgia EU311647 D. ocoee DAB1122 Macon, North Carolina EU311643 D. ocoee DAB951 Union, Georgia EU311715

449 Herpetological Conservation and Biology

Appendix 2 (continued). Specimen and locality information for the COX1 data set with GenBank number or source for each sequence. Species Voucher#/Museum# County/Parish, State GenBank Accession# D. orestes DAB719 Burke, North Carolina EU311703 D. orestes DAB739 Caldwell, North Carolina EU311704 D. quadramaculatus DAB1356 Madison, North Carolina EU311649 D. santeetlah DAB327 Graham, North Carolina EU311676 D. welteri DAB326 Harlan, Kentucky EU311675

450