2011 SOUTHEASTERN NATURALIST 10(2):251–266 Cave-obligate Biodiversity on the Campus of Sewanee: The University of the South, Franklin County, Tennessee Groves B. Dixon1 and Kirk S. Zigler1,* Abstract - The southern Cumberland Plateau in Tennessee and Alabama has the greatest diversity of cave-obligate animals in the United States. The University of the South in Franklin County, TN is one of the largest private landholders on the southern Cumberland Plateau. Its 13,000-acre campus has more than 30 caves and is underlain by more than 14 km of horizontal passageways. We examined the biodiversity of cave animals on the campus at the species level and at the genetic level. Through a survey of seven caves on the campus, we identified 24 cave-obligate species, including two new county records. This total accounts for half of the cave-obligate species reported for Franklin County. For our genetic analysis, we selected six diverse taxa (two millipedes, a beetle, a fly, an aquatic isopod, and a spider) that were collected from multiple caves, and compared their mitochondrial cytochrome oxidase I gene sequences. Across the six taxa we found: (1) low genetic diversity within caves (mean nucleotide diversity within caves across all taxa: 0.25%), (2) high genetic divergence between caves (divergence between caves within taxa ranged from 2.5%–10.9%, with two exceptions), and (3) little evidence for gene flow between caves (FST between caves within taxa > 0.57, with one exception). Thus, the campus supports tremendous species diversity, and even more remarkable genetic diversity within those species on a small geographic scale (no studied caves were >7 km apart). The divergence between cave populations and lack of gene flow between them that we observed across a range of taxa highlight the importance of cave conservation on a regional scale. Introduction Obligate cave-dwelling animals, both terrestrial (troglobites) and aquatic (stygobites), represent a variety of taxa that have drawn the attention of scientists for centuries. They typically share a suite of convergent morphological characteristics, including eyelessness, pigment loss, appendage elongation, and increased extravisual senses, collectively termed “troglomorphy” (reviewed in Culver and Pipan 2009). Cave-obligate organisms generally have small ranges (Culver et al. 2000), and many are known from only a single cave (e.g., 45% of eastern United States cave species are single-cave endemics (Christman et al. 2005)). In addition to their small range sizes and high rates of endemism, cave-obligate animals are restricted to karst (cave-bearing) regions, with the result that cave-obligate biodiversity is concentrated within relatively few geographical areas. Across nine major karst regions of the United States, the Appalachians and the Interior Low Plateaus have the most cave 1Department of Biology, Sewanee: The University of the South, 735 University Avenue, Sewanee, TN 37383. *Corresponding author - kzigler@sewanee.edu. 252 Southeastern Naturalist Vol. 10, No. 2 biodiversity (Culver et al. 2003). The southern edges of these regions meet at the southern Cumberland Plateau in Tennessee and Alabama. Figure 1. Extent of the southern Cumberland Plateau in Tennessee and Alabama. County borders and names are indicated, and the campus of The University of the South in Franklin County, TN is outlined in grey. Higher altitudes are designated by darker shades. 2011 G.B. Dixon and K.S. Zigler 253 The southern Cumberland Plateau supports the greatest richness of caveobligate species in North America (Culver et al. 2000), is a center of endemism for cave species (Christman et al. 2005), and is a biodiversity hotspot for cave fauna on a global scale (Culver et al. 2006). It is a region of high cave density and high temperate productivity, both of which likely contribute to high levels of cave biodiversity (Culver et al. 2006). The campus of The University of the South (known as “Sewanee”) is located in the northeast corner of Franklin County, TN, and covers 13,000 acres of this region (Fig. 1). As the privately held property of an educational institution, the Sewanee campus offers a unique opportunity for the study and conservation of cave biodiversity. Despite this, there has been no systematic survey of Sewanee’s cave-obligate biodiversity. Most studies of cave biodiversity have focused solely on species diversity. On the southern Cumberland Plateau, a series of surveys (Lewis 2005; Peck 1989, 1995) and taxonomic revisions (Buhay and Crandall 2008, Lewis 2009, Shear 2010) have provided a well-developed picture of the species diversity of cave-obligate taxa for the region. Recently, several studies have examined the genetic diversity of specific cave-obligate taxa in this region. These include two studies on cave crayfish (Buhay and Crandall 2005, Buhay et al. 2007), one on Tennessee cave salamanders (Niemiller et al. 2008), and one on a cave spider (Snowman et al. 2010). These studies provide insights into the population genetics and evolutionary history of these cave-obligate taxa. Buhay and Crandall (2005) and Buhay et al. (2007) emphasized the large population sizes and extensive gene flow observed for the cave crayfish, Niemiller et al. (2008) highlighted evidence for speciation of the Tennessee cave salamanders in the presence of gene flow with surface salamander species, and Snowman et al. (2010) found significant divergence and little evidence for gene flow between cave spider populations. Undoubtedly, studies of genetic diversity are an important complement to studies of species diversity in caves (reviewed in Porter 2007). In this study, we examined the biodiversity of cave-obligate animals on the Sewanee campus at the species level and at the genetic level. We first surveyed seven caves on the campus. We then selected six diverse taxa (two millipedes, a beetle, a fly, an aquatic isopod, and a spider) that were collected from multiple caves, and compared their mitochondrial cytochrome oxidase I gene sequences. Although Sewanee has a large campus, all the caves we surveyed were within 7 km of one another, allowing us to examine gene flow and population divergence on a small scale. Previous genetic analyses on a troglobitic spider from Sewanee’s campus found that cave populations were isolated and genetically divergent (Snowman et al. 2010). By studying diverse taxa, we tested whether cave populations are generally isolated from one another, and whether high levels of intraspecific diversity between caves are general characteristics of cave-obligate species, or if they are a unique trait of a single cave-obligate lifestyle, morphology, or taxon. 254 Southeastern Naturalist Vol. 10, No. 2 Materials and Methods Sample sites We surveyed seven caves on the campus of The University of the South. The caves were selected on the basis of location, size, and accessibility. The sampled caves represent the six largest horizontal caves found within the campus, with a total of ≈6 km of horizontal passageways. In addition, we surveyed TD Cave, a small cave located near Solomon’s Temple Cave. We also gathered previously reported data on Dry Cave and Lost Cove Cave, which are located just outside the northern and southern borders of the campus, respectively, and Wet Cave, which runs beneath the campus, although the entrance is located outside of it. With the exception of TD Cave, which is not listed by the Tennessee Cave Survey (TCS), we refer to all caves by their TCS names. To protect sensitive cave habitats, cave locations are not disclosed in this paper. Field methods Sampling was conducted between 30 January 2009 and 2 February 2010. We visited each cave 4 to 6 times. Visits generally lasted 1 to 2 hours. Our sampling methods included hand collecting, pitfall traps, and Berlese extraction from litter samples. Hand collected samples were stored temporarily in reagent alcohol (90.25% ethanol, 4.75% methanol). For pitfall traps, we used 50- and 250-mL plastic cups. We filled the bottoms of these with reagent alcohol, and covered the openings with 1.3-cm wire mesh to prevent disruption by larger fauna, or their accidental capture. Pitfall traps were usually baited with Limburger cheese, smeared around the inner edge of the cup, though in several cases, cow dung proved an effective substitute. Traps were dug into banks, or wedged between rocks as to allow crawling access to the lip. They were left for two to five days before recovery on a return visit. Litter samples were transferred to the lab in plastic bags, and placed in Berlese funnels for invertebrate extraction. Cave-obligate salamanders and crayfish were observed and recorded, but were not collected. Sample identification In the lab, specimens were sorted and stored in 95% ethanol at -20 °C. We identified as many species as possible using available literature. If we could not identify a species confidently, we deferred to recognized experts (see Acknowledgments). Scientific nomenclature including authorities for all species collected in this study are listed in Table 2. DNA extraction, amplification, and sequencing We selected six cave-obligate taxa for genetic analysis: Pseudotremia minos (Russell Cave Millipede), Pseudotremia barri (Barr’s Cave Millipede), Ptomaphagus hatchi (Hatch’s Cave Fungus Beetle), Spelobia tenebrarum (Cave Dung Fly), Caecidotea bicrenata (Two-toothed Cave Isopod), and Nesticus barri (Barr’s Cave Spider). These were chosen for their abundance in our sample sites, their responsiveness to our sequencing methods, and their taxonomic and 2011 G.B. Dixon and K.S. Zigler 255 morphological variety. For Barr’s Cave Spider, we supplemented sequences already collected from two caves on the campus by Snowman et al. (2010). For DNA extractions, we used the DNeasy Blood and Tissue kit (Qiagen #69506) according to the manufacturer’s protocol. DNA extractions were examined by electrophoresis to ensure success before amplification was attempted. We amplified fragments of the mitochondrial cytochrome oxidase I (COI) gene using the polymerase chain reaction (PCR). Our PCRs were 30 μL in volume: 15 μL AmpliTaq Gold PCR Master Mix (Applied Biosystems #4316753), 12 μL distilled water, 1 μL each of the selected 5’ and 3’ primers at 10 μm concentration, and 1 μL of extracted DNA. All amplifications were originally conducted using the universal COI primers HCOI-2198 and LCOI-1490 (Folmer et al. 1994). After a successful sequence was acquired for a species, we developed species-specific primers to increase the efficiency of our amplification and sequencing reactions (Table 1). Our PCRs ran as follows: 5 min at 95 °C; then 35 cycles of 15 sec at 95 °C, 15 sec at 45 °C, and 1 min at 72 °C; with a final 7 min at 72 °C. When using species-specific primers, we increased the annealing temperature to 50 °C. PCRs were examined by electrophoresis. Successful PCRs were purified (Qiagen #28106) and sequenced on both strands. All sequences have been submitted to GenBank (Accession #HM245857–HM245919). Sequence analysis COI fragments were aligned and edited using Sequencher 4.9 (Gene Codes Corporation, Ann Arbor, MI). Variable sites were verified manually from chromatograms, and primers were excluded from all alignments. No indels were present. To assess the level of genetic diversity of populations within individual Table 1. Focal taxa, their common names, and the primer combinations which gave the most consistent results for each. All sequences were first amplified using the universal primers HC01-2198 and LC01-1490 (Folmer et al. 1994) before development of species-specific primers. The Nesticus barri primers are from Snowman et al. (2010). Species Primer Sequence (5'-3') Pseudotremia barri HC01-tremia GTTGATATAAAATTGGGTCCCCTCC (Barr's Cave Millipede) LC01-1490 GGTCAACAAATCATAAAGATATTG Pseudotremia minos HC01-tremia GTTGATATAAAATTGGGTCCCCTCC (Russell Cave Millipede) LC01-1490 GGTCAACAAATCATAAAGATATTG Ptomaphagus hatchi HC01-ptom GGGACATCCTTAAGACTTTTAATTC (Hatch’s Cave Beetle) LC01-ptom GCTGGTAAAGAATTGGATCCCC Spelobia tenebrarum HC01-ten GAACTTTATATTTTATATTTGGGGC (Cave Dung Fly) LC01-ten GTCTCCTCCACCAGCAGGGTC Caecidotea bicrenata HC01-caec AGGGTCCCTCCCCCCTGGGG (Two-toothed Cave Isopod) LC01-caec GGGGCTTGAGCCGGAAGAGTCGG Nesticus barri HC01-2198 TAAACTTCAGGGTGACCAAAAAATCA (Barr’s Cave Spider) LC01-barri GGACTTTGTATTTTATTCTTGGGTC 256 Southeastern Naturalist Vol. 10, No. 2 Table 2. Cave-obligate species found in and just outside Sewanee’s campus and the caves they inhabit. X designates observations made in this study, L designates records from Lewis (2005), and S designates records from Shear (2010). Unconfirmed records are indicated with a ‘?’. Caves are listed from south to north and abbreviated as follows: LC = Lost Cove Cave, GV = Grapeville, BB = Buckets of Blood, BH = Sewanee Blowhole, MC = Miller Cave, TD = TD Cave, ST = Solomon’s Temple, WC = Wet Cave, WS = Walker Springs, DC = Dry Cave. Cave Class Order Species LC GV BB BH MC TD ST WC WS DC Arachnida Araneae Anthrobia monmouthia Tellkampf X Liocranoides archeri Platnick X X X X L Nesticus barri Gertsch L X X X X X X Phanetta subterranea (Emerton) X X X X X L L Pseudoscorpionida Hesperochernes mirabilis (Banks) X X X Kleptochthonius magnus Muchmore L Kleptochthonius tantalus Muchmore L Opiliones Tolus appalachius Goodnight & Goodnight L X X Diplopoda Callipodida Tetracion jonesi Hoffman X X X Chordeumatida Pseudotremia barri Lewis X X L X L Pseudotremia minos Shear X X Scoterpes stewartpecki Shear S X X ? Scoterpes ventus Shear S S S Entognatha Collembola Pseudosinella pecki Christiansen & Bellinger L Pseudosinella hirsuta (Delamare) X X X X X X Pseudosinella spinosa (Delamare) X X 2011 G.B. Dixon and K.S. Zigler 257 Table 2, continued. Cave Class Order Species LC GV BB BH MC TD ST WC WS DC Diplura Litocampa cookei (Packard) L Litocampa valentinei (Conde) X X X X L Insecta Coleoptera Pseudanopthalmus humeralis Valentine L Pseudanopthalmus intermedius Valentine X L Ptomaphagus hatchi Jeannel X X X X X X L X L Subterrochus ferus (Park) L Subterrochus steevesi Park X Diptera Spelobia tenebrarum (Aldrich) X X X L X L Malacostraca Isopoda Caecidotea bicrenata bicrenata (Steeves) L X X X X L X L Decapoda Orconectes australis australis (Rhoades) ? L X Amphipoda Stygobromus sp. X L Turbellaria Tricladida Sphalloplana percoeca (Packard) X L Amphibia Caudata Gyrinophilus palleucus McCrady L X Total per cave: 4 12 11 11 7 6 8 10 11 17 258 Southeastern Naturalist Vol. 10, No. 2 caves, we calculated haplotype diversity (h), and nucleotide diversity (π) for every cave population for which we had two or more sequences, using DnaSP v.5 (Librado and Rozas 2009). Transition, tranversion, silent, and replacement nucleotide substitutions for each species were counted using MEGA v.4 (Tamura et al. 2007). To examine variability between cave populations, we calculated mean pairwise divergence (DXY) and population pair-wise FST using DnaSP. FST comparisons were only conducted between populations with three or more representative sequences. We also assessed variability and population structure for each species as a whole. For these assessments, we constructed phylogenetic networks using TCS 1.21 (Clement et al. 2000), and compared haplotype and nucleotide diversity calculated for the entirety of our samples for each species. Results Species diversity We observed 24 cave-obligate species within the campus of The University of the South. There were six arachnids, five diplopods, four insects, four non-insect hexapods, three malacostracans, one turbellarian, and an amphibian (Table 2). A total of 19 troglobites and 5 stygobites were observed. Only Ptomaphagus hatchi (Hatch’s Cave Beetle) was found in every cave sampled. Five species were found only in a single cave: Gyrinophilus palleucus (Tennessee Cave Salamander); a cavernicolous pselaphid beetle, Subterrochus steevesi; the troglobitic Anthrobia monmouthia (Mammoth Cave Spider); the cave flatworm, Sphalloplana percoeca; and an unidentified amphipod, Stygobromus sp. Previous surveys and taxonomic revisions identified 45 described caveobligate species (and several undescribed species) for the entirety of Franklin County, TN (Culver et al. 2000; Lewis 2005, 2009; Shear 2010). Our survey increased this total to 47 with the observation of two previously unrecorded species (Subterrochus steevesi and Anthrobia monmouthia). Subterrochus steevesi was previously known only from Marshall County, AL (Chandler 1997, Park 1960). Anthrobia monmouthia is known from scattered collections from Kentucky, Tennessee, Virginia, and Alabama (Miller 2005). An average of ten cave-obligate species were found in each cave on the campus (Table 2). Two species identifications from the Sewanee Blowhole are unconfirmed. We collected only immature Scoterpes individuals from this cave and so are unable to confirm the species. Other caves on the southern side of campus have S. stewartpecki, but it is also possible that these individuals represent the southernmost record of S. ventus (Table 2; Shear 2010). We also observed, but did not collect, a troglobitic crayfish in this cave. This specimen was probably Orconectes australis australis (Rhoades) (known from caves on the north side of campus), but as it was found on the southern side of campus, it may be Cambarus hamulatus (Cope) (known from caves in southern Franklin County) (Buhay and Crandall 2005, Buhay et al. 2007). We also collected an unidentified amphipod of the genus Stygobromus in Solomon’s Temple cave. 2011 G.B. Dixon and K.S. Zigler 259 Table 3. Focal taxa and their sequencing data. Number of individuals sequenced (n), the number of base pairs sequenced (bp), the number of conserved and variable sites, the number of transition (Transi.), transversion (Transv.), silent, and replacement (Repl.) mutations observed, and the number of stop codons present. There were two instances (marked with an asterisk) when a transition and a transversion were found at a single variable site. Conserved Variable Stop Species n bp sites sites Transi. Transv. Silent Repl. Codons Pseudotremia barri 9 626 620 6 4 2 6 0 0 Pseudotremia minos 10 626 620 6 5 1 6 0 0 Ptomaphagus hatchi 21 563 529 34 30* 5* 34 0 0 Spelobia tenebrarum 15 589 543 46 26* 21* 39 7 0 Caecidotea bicrenata 6 573 518 55 45 10 53 2 0 Nesticus barri 8 633 612 21 19 2 18 3 0 Total 69 3442 168 129 41 156 12 0 Five species of Stygobromus are reported from Franklin County, three of which are undescribed (Lewis 2005). Intraspecific genetic variation We acquired COI sequences from 69 individuals, and found 29 unique haplotypes among our six species of interest. Sequences ranged from 563 to 633 bp of the COI coding region. No stop codons or indels were observed (Table 3). Comparisons within species revealed a total of 168 variable sites: 156 were silent substitutions, and 12 were replacement substitutions. Across these sites we observed 129 transition mutations and 41 transversion mutations, with 2 variable sites showing both a transition and transversion (Table 3). Nucleotide diversity (π) of populations within caves across all species was generally low (mean πall populations = 0.25%) but varied across caves and taxa (Table 4). Seven populations were represented by a single haplotype, with nucleotide and haplotype diversities of zero. For populations represented by multiple halplotypes, haplotype diversity (h) ranged from 0.39 to 0.84 (Table 4). Haplotype diversities across all caves were high (mean h = 0.681), with five of our six focal species showing values >0.50. Population structure Divergence between populations averaged 3.7% and varied widely across species and the caves compared (Table 5). Notably, the sampled populations of C. bicrenata (from Walker Springs and Buckets of Blood), showed a divergence of 10.9%, whereas populations of other species, such as those of S. tenebrarum and P. minos, showed divergences of 0.3% and 0.4% between caves (Table 5). Interspecific divergence between P. minos and P. barri was 3.7%, which was similar to most intraspecific population comparisons. Of the 29 haplotypes observed across our six focal taxa, 25 were restricted to a single cave (Fig. 2). In three of the four cases in which haplotypes were shared between caves, the caves are within 3.5 km of one another (between TD 260 Southeastern Naturalist Vol. 10, No. 2 Table 4. Summary of genetic diversity data for focal taxa. For caves for which there were at least two representative sequences, the number of individuals sequenced (n), number of haplotypes per cave population, haplotype diversity and its standard deviation (h), and nucleotide diversity and its standard deviation (π) are reported. The “All” rows include all sequences for the species, including those from caves with fewer than two representative sequences not listed in the table. Species Cave n # haplotypes h π Pseudotremia barri Walker Springs 7 1 0 0 All 9 2 0.389 ± 0.164 0.0038 ± 0.0016 Pseudotremia minos Grapeville 3 1 0 0 Buckets of Blood 7 4 0.810 ± 0.130 0.0032 ± 0.0006 All 10 5 0.844 ± 0.080 0.0033 ± 0.0004 Ptomaphagus hatchi Grapeville 7 4 0.714 ± 0.181 0.0047 ± 0.0016 Buckets of Blood 4 3 0.833 ± 0.222 0.0021 ± 0.0007 Walker Springs 7 3 0.667 ± 0.160 0.0186 ± 0.0064 All 21 11 0.890 ± 0.049 0.0239 ± 0.0017 Spelobia tenebrarum Grapeville 2 1 0 0 Buckets of Blood 5 3 0.800 ± 0.164 0.0032 ± 0.0008 Walker Springs 6 2 0.533 ± 0.030 0.0019 ± 0.0006 Sewanee Blowhole 2 1 0 0 All 15 6 0.857 ± 0.003 0.0231 ± 0.0067 Caecidotea bicrenata Buckets of Blood 4 1 0 0 Walker Springs 2 1 0 0 All 6 2 0.533 ± 0.019 0.0583 ± 0.0188 Nesticus barri Grapeville 3 2 0.667 ± 0.314 0.0011 ± 0.0005 Sewanee Blowhole 4 1 0 0 All 8 3 0.571 ± 0.094 0.0182 ± 0.003 and Solomon’s Temple in Pseudotremia barri; between TD, Solomon’s Temple, and Walker Springs in Ptomaphagus hatchi; and between Grapeville and Buckets of Blood in Nesticus barri). In one case, the caves are on opposite sides of the campus (between Walker Springs and Buckets of Blood in Spelobia tenebrarum) (Fig. 2). FST values between caves were high (mean FST = 0.68), and only three of eight comparisons had FST less than 0.90 (Table 5). The lowest of these was between the Buckets of Blood and Walker Springs populations of S. tenebrarum (FST = 0.10; Table 5). Discussion Species diversity and distributions The campus of The University of the South supports an exceptionally rich collection of troglobitic and stygobitic species. As noted by Culver et al. (2000), the southern Cumberland Plateau is particularly rich in troglobitic 2011 G.B. Dixon and K.S. Zigler 261 species and is less rich in stygobitic species. This description correlates with our findings of 19 troglobitic species and 5 stygobitic species. Culver et al. (2000) noted that only 8 out of 3112 counties in the United States had more than 20 troglobitic species, yet here, on only 13,000 acres, we identified 19. The 24 total species found represent half of the 47 total cave-obligate species known from Franklin County. Also notable is the exceptionally species-rich Dry Cave, located just to the north of the campus. Seventeen cave-obligate species have been found in Dry Cave (Table 2), which is an impressive number given that only 36 caves worldwide are known to have 20 or more cave-obligate species (Culver and Pipan 2009). The Sewanee campus is transected by species range borders for eight of the 24 cave-obligate species we observed (Table 2). This large proportion reflects the small range sizes characteristic of troglobionts (Christman et al. 2005, Culver et al. 2000). The campus covers the southern range borders for Scoterpes ventus, Pseudotremia barri, Pseudanopthalmus humeralis, and Pseudanopthalmus intermedius, and the northern range borders for Scoterpes stewartpecki, Pseudotremia minos, Nesticus barri, and Tetracion jonesi (Table 2). The diplopod genera Scoterpes and Pseudotremia represent the clearest range borders, with one species from each genus found in caves on the northern side of the campus, and a second species from each genus found in caves on the southern side of the campus. For these genera, the distributions of the northern and southern species pairs on the campus are identical (Table 2). Whether this correlation is coincidental or the result of a shared evolutionary history remains to be determined. Table 5. Genetic divergence between populations for the focal taxa. Pair-wise distances (below diagonal) between cave populations represented by at least two sequences, and pair-wise FST (above diagonal, in bold) between caves represented by at least three sequences for Pseudotremia minos and P. barri (Ps), Ptomaphagus hatchi (Pt), Spelobia tenebrarum (Sp), Caecidotea bicrenata (Ca), and Nesticus barri (Nb). Cave Grapeville Buckets of Blood Sewanee Blowhole Walker Springs Grapeville 0.577 (Ps) 1.000 (Ne) 1.000 (Ps) 0.901 (Pt) 0.901 (Pt) Buckets of Blood 0.004 (Ps) 0.957 (Ps) 0.037 (Pt) 0.597 (Pt) 0.044 (Sp) 0.104 (Sp) Sewanee Blowhole 0.071 (Sp) 0.046 (Sp) 0.032 (Ne) Walker Springs 0.036 (Ps) 0.038 (Ps) 0.046 (Sp) 0.030 (Pt) 0.027 (Pt) 0.045 (Sp) 0.003 (Sp) 0.109 (Ca) 262 Southeastern Naturalist Vol. 10, No. 2 Gene flow between caves Troglobitic and stygobitic populations from a variety of taxa show high levels of genetic isolation between caves, even when they are found within a relatively small area (no caves sampled in this study are more than 7 km apart). This isolation is evident from the high proportion of haplotypes that are not shared between caves (25 of 29), the high pair-wise FST values (>0.55, with one exception) between caves (Table 5), and the restrictive clustering of haplotypes by cave (Table 4, Fig. 2). Even in cases where haplotypes are shared between caves, there is evidence that gene flow is limited by distance. In three of the four cases where haplotypes are shared between caves, the caves are adjacent to one another. The sole exception is in Spelobia tenebrarum, where a haplotype is shared between caves on opposite sides of the campus. As the only cave-obligate fly in North America, S. tenebrarum is clearly exceptional. It has a large range (Marshall and Peck 1985) and is one of only nine cave-obligate species recorded from more than 30 counties in the eastern United States (Christman Figure 2. Haplotype networks for six focal species: (A) Pseudotremia barri, (B) Pseudotremia minos, (C) Ptomaphagus hatchi, (D) Spelobia tenebrarum, (E) Caecidotea bicrenata, and (F) Nesticus barri. Circles represent haplotypes, and lines connecting them represent single nucleotide differences. Black bars and junctions between lines indicate unsampled and/or extinct haplotypes. Circle area is proportional to the number of sampled individuals with that haplotype (ranging from 1 to 7). Labels within circles indicate the cave (or caves) where that haplotype was found (abbreviated as in Table 1). Haplotypes found in more than one cave are indicated by the presence of more than one label within a circle. Unconnected networks or haplotypes differ by ten or more nucleotide differences from all others. 2011 G.B. Dixon and K.S. Zigler 263 and Culver 2001). Exactly how S. tenebrarum moves around is unclear, but it must have more ability to do so than most cave-obligate species. Despite this, our sample of S. tenebrarum shows evidence of isolation and divergence, as two of the four caves we sampled have unshared haplotypes that do not connect to the network that contains the haplotype shared between Buckets of Blood and Walker Springs (Fig. 2). Divergence between caves Despite low genetic diversity within caves (mean πall populations = 0.25%; Table 4) there are high levels of intraspecific divergence between caves (DXY ≥ 2.7% for 12 out of 14 comparisons, Table 5). Craft et al. (2010), using COI sequences to study lepidoptera from New Guinea, deemed intraspecific divergence greater than 2% as indicative of “cryptic lineage diversity”. According to this standard, 12 of our 14 intraspecific population comparisons identified cryptically distinct lineages. High levels of genetic diversity within species are also evident in the haplotype networks. For only two of the six focal taxa did all of the observed haplotypes connect into a single network (Fig. 2). It is generally assumed that aquatic subterranean habitats are better connected than terrestrial ones, facilitating greater dispersal ability for stygobites over troglobites (Culver et al. 2007, Lamoreaux 2004), and studies have shown that some stygobites are capable of wide-ranging dispersal (Buhay and Crandall 2005, Danielopol et al. 1994). Our limited molecular results for the aquatic isopod, C. bicrenata, were contrary to this pattern as they showed the greatest divergence of any in our sample group (DXY = 10.9%; Table 5). This divergence suggests that the two caves we sampled are not hydrologically connected, which may be the case as the stream from one of the caves drains to the south and the Tennessee River, whereas the other drains to the north and the Elk River. Further examination of this species would be interesting, particularly in light of its large range, which spans from northern Alabama through Tennessee, Kentucky, and into southern Illinois (Lewis 1982). Two other recent studies have examined COI diversity and divergence of cave animals on a small geographic scale. Carlini et al. (2009) studied the amphipod Gammarus minus Say, surveying COI diversity and divergence for five cave populations, including four that are no more than 20 km apart. They found low nucleotide diversity within populations (mean π = 0.70%), high divergence between cave populations (≈5% for most comparisons between caves), and isolation between caves. Snowman et al. (2010) studied Nesticus barri across its four-county range on the southern Cumberland Plateau, and found low COI diversity within caves, high divergence between caves (2–5%), and isolation between cave populations. Thus, the findings of these studies are consistent with our results, although we found these patterns on a smaller geographic scale, and for a variety of taxa. These results from small stygobitic and troglobitic invertebrates contrast with those from larger stygobites studied in this area (crayfish and cave salamanders), which found mitochondrial 264 Southeastern Naturalist Vol. 10, No. 2 haplotypes shared across counties (Buhay and Crandall 2005, Buhay et al. 2007, Niemiller et al. 2008). These large stygobites may be more mobile than troglobites or smaller stygobites, which may explain the patterns observed. Conclusion We identified 19 troglobites and 5 stygobites from the campus of The University of the South. For at least eight of these species, the campus is a northern or southern range boundary. Across six diverse taxa we found: (1) low genetic diversity within caves, (2) high genetic divergence between caves, and (3) little evidence for gene flow between caves. These patterns were generally consistent across the taxa studied, regardless of their mechanism of locomotion, ecological role, and habitat, suggesting these characteristics are common for cave-obligate species from this region. Particularly notable is that we observed these patterns within a small geographic area, where no surveyed cave was more than 7 km from another. Thus, the campus supports tremendous species diversity, and even more remarkable genetic diversity within those species on a very small scale. Our observation of high genetic divergence between caves, and little gene flow between them, highlights the importance of cave conservation on a regional scale, in order to conserve as much genetic diversity of cave-obligate species as possible. Acknowledgments We thank T. Barr, C. Carlton, L. Ferguson, J. Lewis, P. Paquin, and W. Shear for taxonomic assistance. We thank M. Williams, A. Sidik, G. Cooper, and J. Christopher for assistance in the field, and J. Benson and D. Durig for caving advice. We also thank N. Hollingshead, V. Moye, and A. Sidik for assistance with maps and figures. N. Berner, D. McGrath, A. Sidik, and A. Summers commented on the manuscript. This project was supported by Faculty Research Support funds from Sewanee: The University of the South and by a Yeatman Fellowship (to G.B. Dixon). Literature Cited Buhay, J.E., and K.A. Crandall. 2005. Subterranean phylogeography of freshwater crayfi shes shows extensive gene flow and surprisingly large population sizes. Molecular Ecology 14:4259–4273. Buhay, J.E., and K.A. Crandall. 2008. Taxonomic revision of cave crayfish in the genus Orconectes, subgenus Orconectes (Decapoda: Cambaridae) along the Cumberland Plateau, including a description of a new species—Orconectes barri. Journal of Crustacean Biology 28:57–67. Buhay, J.E., G. Moni, N. Mann, and K.A. Crandall. 2007. Molecular taxonomy in the dark: Evolutionary history, phylogeography, and diversity of cave crayfish in the subgenus Aviticambarus, genus Cambarus. Molecular Phylogenetics and Evolution 42:435–448. Carlini, D.B., J. Manning, P.G. Sullivan, and D.W. Fong. 2009. Molecular genetic variation and population structure in morphologically differentiated cave and surface populations of the freshwater amphipod Gammarus minus. Molecular Ecology 18:1932–1945. 2011 G.B. Dixon and K.S. Zigler 265 Chandler, D.S. 1997. A catalog of the coleoptera of America north of Mexico: Family: Pselaphidae. US Department of Agriculture, Washington, DC. 118 pp. Christman, M.C., and D.C. Culver. 2001. The relationship between cave biodiversity and available habitat. Journal of Biogeography 28:367–380. Christman, M.C., D.C. Culver, M.K. Madden, and D. White. 2005. Patterns of endemism of the eastern North American cave fauna. Journal of Biogeography 32:1441–1452. Clement, M., D. Posada, and K. Crandall. 2000. TCS: A computer program to estimate gene genealogies. Molecular Ecology 9:1657–1660. Craft, K.J., S.U. Pauls, K. Darrow, S.E. Miller, P.D.N. Herbert, L.E. Helgen, V. Novotny, and G.D. Weiblen. 2010. Population genetics of ecological communities with DNA barcodes: An example from New Guinea Lepidoptera. Proceedings of the National Academy of Science 107:5041–5046. Culver, D.C., and T. Pipan. 2009. The biology of caves and other subterranean habitats. Oxford University Press, New York, NY. 254 pp. Culver, D.C., L.L. Master, M.C. Christman, and H.H. Hobbs III. 2000. Obligate cave fauna of the 48 contiguous United States. Conservation Biology 14:386–401. Culver, D.C., M.C. Christman, W.R. Elliott, H.H. Hobbs III, and J.R. Reddell. 2003. The North American obligate cave fauna: Regional patterns. Biodiversity and Conservation 12:441–468. Culver, D.C., L. Deharveng, A. Bedos, J.J. Lewis, M. Madden, J.R. Reddell, B. Sket, P. Trontelj, and D. White. 2006. The mid-latitude biodiversity ridge in terrestrial cave fauna. Ecography 29:120–128. Culver, D.C., T. Pipan, and K. Schneider. 2007. Vicariance, dispersal, and scale in the aquatic subterranean fauna of karst regions. Freshwater Biology 54:918–929. Danielopol, D.L., P. Marmonier, A.J. Boulton, and G. Bonaduce. 1994. World subterranean ostracod biogeography: Dispersal or vicariance. Hydrobiologia 287:119–129. Folmer, O.M., M. Black, R. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology 3:294–299. Lamoreaux, J. 2004. Stygobites are more wide-ranging than troglobites. Journal of Cave and Karst Studies 66:18–19. Lewis, J.J. 1982. Systematics of the troglobitic Caecidotea (Crustacea: Isopoda: Asellidae) of the Southern Interior Low Plateaus. Brimleyana 8:65–74. Lewis, J.J. 2005. Southern Cumberland Plateau cave survey, Final Report. The Nature Conservancy. 155 pp. Lewis, J.J. 2009. Six new species of Pseudotremia from caves of the Tennessee Cumberland Plateau (Diplopoda: Chordeumatida: Cleidogonidae). Zootaxa 1080:17–31. Librado, P., and J. Rozas. 2009. DnaSP Version 5: A software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25:1451–1452. Marshall, S.A., and S.B. Peck. 1985. Distribution of cave-dwelling Sphaeroceridae (Diptera) of eastern North America. Proceedings of the Entomological Society of Ontario 115:37–41. Miller, J.A. 2005. Cave adaptation in the spider genus Anthrobia (Araneae, Linyphiidae, Erigoninae). Zoologica Scripta 34:565–592. Niemiller, M.L., B.M. Fitzpatrick, and B.T. Miller. 2008. Recent divergence with gene flow in Tennessee cave salamanders (Plethodontidae:Gyrinophilus) inferred from gene genealogies. Molecular Ecology 17:2258–2275. 266 Southeastern Naturalist Vol. 10, No. 2 Park, O. 1960. Cavernicolous pselaphid beetles of the United States. American Midland Naturalist 64:66–104 Peck, S.B. 1989. The cave fauna of Alabama: Part I. The terrestrial invertebrates (excluding insects). The NSS Bulletin 51:11–33. Peck, S.B. 1995. The cave fauna of Alabama. Part II: The insects. The NSS Bulletin 57:1–19. Porter, M.L. 2007. Subterranean biogeography: What have we learned from molecular techniques? Journal of Cave and Karst Studies 69:179–186. Shear, W.A. 2010. The millipede family Trichopetalidae, Part 2: The genera Trichopetalum, Zygonopus, and Scoterpes (Diplopoda: Chordeumatida, Cleidogonoiea). Zootaxa 2385:1–62. Snowman, C.V., K.S. Zigler, and M. Hedin. 2010. Caves as islands: Mitochondrial phylogeography of the cave-obligate spider species Nesticus barri (Araneae: Nesticidae). Journal of Arachnology 38:39–46. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24:1596–1599.