The study of spring- and subterranean-associated microsnail species in the Appalachian karst region has focused disproportionately on the northern Appalachian Valley and Ridge (AVR), leaving many areas in the southern Appalachians unexplored. Consequently, biological inventories of subterranean habitats have been initiated in the southern AVR, particularly in the state of Tennessee. In 2013 and 2018, several previously unknown populations of a microsnail species were discovered from caves in eastern Tennessee. Through both morphological and molecular analysis, we identified these populations as the Blue Ridge Springsnail, Fontigens orolibas. These newly discovered populations represent a significant range extension of F. orolibas. As such, we reassess the conservation status of F. orolibas under NatureServe criteria and emphasize the need for further sampling efforts in the southern AVR for microsnails.
INTRODUCTION
Currently, the freshwater snail genus Fontigens Pilsbry, 1933 is the only recognized North American group among the family Emmericiidae (subfamily Fontigentinae; Hershler et al. 1990; Wilke et al. 2013). Compared with other hydrobioid microsnails (i.e., freshwater snails ranging from 2 to 5 mm of the family Hydrobiidae sensu lato; Davis 1979), most Fontigens species exhibit broad geographic distributions but are known primarily from karst landscapes of the central and eastern United States (Hershler and Holsinger 1990; Hershler et al. 1990; Culver et al. 2003). At present, there are 10 Fontigens species that are fully or partially restricted to springs and subterranean habitats (Hershler et al. 1990; Liu et al. in press). Six of these species are from the Appalachian karst region within the Appalachian Valley and Ridge (AVR) physiographic province (Hershler et al. 1990). Five of the six are endemic to the northern Appalachian karst in the states of Maryland, Pennsylvania, Virginia, and West Virginia (Fig. 1), where the majority of research efforts on aquatic subterranean-associated animals (i.e., stygofauna) has occurred (Holsinger et al. 1976; Holsinger and Culver 1988; Fong et al. 2007; Fong and Culver 2018). These five species are as follows: Appalachian Springsnail Fontigens bottimeri; Virginia Springsnail Fontigens morrisoni; Blue Ridge Springsnail Fontigens orolibas; Organ Cavesnail Fontigens tartarea; and Greenbrier Cavesnail Fontigens turritella. The other species known from the Appalachians—the Watercress Snail Fontigens nickliniana Lea, 1838—is the most geographically widespread of all Fontigens, occurring throughout the AVR, including in Tennessee, Alabama, and Georgia (TAG) karst, and into the northeastern and central USA. Comparatively, underground habitats in the southern extent of the Appalachian karst (or TAG) have been studied very little (Niemiller and Zigler 2013; Niemiller et al. 2019). Only F. nickliniana has been described previously within TAG, from two spring sites.
Figure 1.
Geographic distributions of all Fontigens species occurring in the Appalachians. Newly discovered Fontigens orolibas populations are denoted with a star.
![img-z2-1_34.jpg](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z2-1_34.jpg)
Table 1.
Detailed site descriptions of Tennessee caves sampled in this study.
![img-z3-2_34.gif](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z3-2_34.gif)
Figure 2.
Fontigens orolibas specimens. Top: Cave specimens collected from Cruze Cave (TKN24), Knox Co., Tennessee (1, 2); Pedigo Cave (TKN103), Knox Co., Tennessee (3); and Eblen Cave (TRN6), Roane Co., Tennessee (4). Photo credit: N.S. Gladstone. Bottom: Live F. orobilas from TKN103. Photo credit: M.L. Niemiller.
![img-z4-1_34.jpg](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z4-1_34.jpg)
Since 2012, ongoing biological inventories of cave systems in the AVR of eastern Tennessee have uncovered several new populations of freshwater snails from cave streams. These include recently described stygobiotic species of the genus Antrorbis from two caves (Gladstone et al. 2019). The others were of a Fontigens-like snail, with two populations being found within the Melton Hill Lake watershed of the Clinch River, which flows into Watts Bar Lake and the Tennessee River, and a third within the Clinch River watershed of the Tennessee River. Using morphological and molecular data, we diagnose these three new Fontigens-like snail populations as the Blue Ridge Springsnail F. orolibas. In light of confirming these new populations as F. orolibas, we also reassess the conservation ranks of this species using NatureServe criteria (Master et al. 2009).
METHODS
Field Sampling
All biological surveys involved at least two, and as many as 12, researchers. Scientific research and collection in the caves were permitted, with renewals, by the Tennessee Department of Environment and Conservation (TDEC) and the Tennessee Wildlife Resource Agency (TWRA) (TDEC number 2013-026 and TWRA number 1605). Table 1 presents the timeline of population discovery and site summaries. A single population was first discovered in 2013 from a cave system in Knox Co., Tennessee (TKN24), and other populations were found in a different cave in Knox Co. (TKN103) and in a cave in Roane Co. (TRN6), Tennessee (Fig. 1). Monthly surveys were performed at the original locality (TKN24; physicochemical measurements, abundance data, and microhabitat descriptions can be found in Keenan et al. [2017]). We report only individual Tennessee Cave Survey (TCS) inventory numbers rather than exact geographic coordinates, but cave system descriptions and location data are maintained by the TCS ( http://www.subworks.com/tcs/) and are available from the authors upon request.
Table 2.
Mean shell measurements (mm) for specimens examined from each locality. Standard deviations in parentheses. * represents populations with measurements from Hershler et al. (1990).
![img-z5-2_34.gif](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z5-2_34.gif)
Morphological Analysis
Specimens collected from each site were preserved in 100% ethanol and transferred to the laboratory for morphological analysis and imaging. Upon completing the reevaluation of these materials, the specimens were deposited in the Auburn Museum of Natural History. We used a Jenoptik SUBRA full high-definition microscope camera to photograph and evaluate shells (Fig. 2). Standard shell measurements included in Hershler et al. (1990) were recorded for specimens, including number of whorls, shell height, shell width, whorl expansion rate, distance of generating curve, translation rate, and aperture shape (Table 2). We also compared the shell measurements of all individual Fontigens species included in Hershler et al. (1990) with three specimens from the TKN24 site via principal components analysis (PCA) with paleontological statistics software (Hammer et al. 2001). All seven shell measurements were included as components in the analysis and each individual specimen presented by Hershler et al. (1990) is included separately.
Table 3.
Sequence and locality information. All Fontigens sequences without a GenBank accession number were generated and provided by Liu et al. (in press). Numeric code next to locality information represents population identification shown in phylogeny.
![img-z5-17_34.gif](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z5-17_34.gif)
Figure 3.
Principal components analysis of Fontigens spp. morphological data from Hershler et al. (1990) and this study. The Tennessee F. orolibas specimens are shown with an open square. Each point represents an individual specimen. Component 1 accounts for 80.12% of variability and component 2 accounts for 15.3% of variability. Other abbreviations defined in the text.
![img-z6-1_34.jpg](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z6-1_34.jpg)
Molecular Methods and Phylogenetic Analysis
Genomic DNA was isolated from three specimens from the TKN24 population using the Qiagen DNeasy blood and tissue kit, following the manufacturer's protocol. We amplified a 638-base-pair fragment of mitochondrial cytochrome oxidase subunit 1 (CO1) locus using LC01490 and HC02198 primers (Folmer et al. 1994). PCR products were purified using ExoSAP-IT (Affymetrix, Santa Clara, CA, USA) and sequenced in both directions with BigDye chemistry at Eurofins MWG Operon (Louisville, KY, USA). Forward and reverse sequences were quality trimmed at the ends and assembled into contigs in DNA Baser v4.36 (Heracle BioSoft) and aligned using MAFFT (Katoh and Standley 2013). These sequences were then compared with CO1 sequences from several other Fontigens species generated by Liu et al. (in press) (Table 3).
The CO1 phylogeny was generated using a maximum-likelihood approach in IQ-TREE 2 (Minh et al. 2020) with the model-testing function to infer the best-fit substitution model for each codon partition under the corrected Akaike's information criterion. We implemented a general time-reversible model with corrections for a discrete gamma distribution (GTR + Γ) for the first and second codon positions, and the Hasegawa, Kishino, and Yano model with a discrete gamma distribution and a proportion of invariant sites (HKY + Γ + I) for the third codon position. Branch support was assessed with 10,000 ultrafast bootstrap replicates (Hoang et al. 2017).
Conservation Status Reassessment
On the basis of the taxonomic identity of these populations as F. orolibas, we reassessed the conservation statuses of the species using NatureServe criteria (Master et al. 2009). Conservation status, as designated by NatureServe, is calculated on the basis of several risk categories, including range/distribution, abundance/population condition, threat impacts, and population trends (for details on calculation procedure, see Faber-Langendoen et al. 2012). Each risk category as assessed through NatureServe was calculated with the NatureServe Rank Calculator v3.1932 (Faber-Langendoen et al. 2012). Abundance data for F. orolibas are virtually nonexistent, but the earlier efforts of Keenan et al. (2017) represent the first systematic survey of a single population to date. Consequently, population trends and viability information could not be completely assessed for the species. Geographic range size (as assessed by the geographic coordinates associated with each population) was calculated as extent of occurrence (EOO) and area of occupancy (AOO) using the web-based program GeoCAT (Bachman et al. 2011). To determine potential threats to F. orolibas throughout its range, we identified whether populations occurred on state or federally protected lands using the U.S. Geological Survey Protected Areas Database v1.3 (shapefile available at http://gapanalysis.usgs.gov/padus/). For each known population, we also examined history of disturbance of each site (if known), the adjacent human population density according to the 2018 U.S. Census from the U.S. Census Bureau (TIGER/Line®), and land cover associations according to the 2016 National Land Cover Database (Yang et al. 2018).
Figure 4.
Mitochondrial cytochrome oxidase subunit 1 (CO1) phylogeny of Fontigens species with sequence data generated from this study and by Liu et al. (in press). Internal values represent bootstrap support and the scale bar represents branch length in units of sequence divergence. The species Bythinella viridis was rooted as the outgroup. Location identification numbers are shown to the right of the species name and listed in Table 3.
![img-z7-1_34.jpg](ContentImages/Journals/fmbc/24/1/fmbc-d-20-00007/graphic/WebImages/img-z7-1_34.jpg)
RESULTS AND DISCUSSION
Shell measurements from all new populations in Tennessee showed considerable similarity with F. orolibas shells measured in Hershler et al. (1990). Results of the PCA also showed that our three Fontigens specimens from the TKN24 populations are within the range of morphological variability of F. orolibas. Principal component 1 accounts for 80.12% of variability and is highly influenced by translation rate, shell height, and number of whorls. Separation of taxa along principal component 2, which accounts for 15.3% of observed variability, is influenced by changes to shell height and shell width (Fig. 3). The resulting phylogeny shows 99% bootstrap support at the internal node for all F. orolibas individuals including the TKN24 population (Fig. 4), further supporting the species identification. These combined results confirm that the population at TKN24 is indeed F. orolibas, and as such we infer that the other two newly discovered populations are also F. orolibas on the basis of geographic proximity and near morphologic indistinguishability (Figs. 1, 2).
The population discoveries in eastern Tennessee extend the known geographic range of F. orolibas to 26 localities from 19 counties in four states and increase the EOO to 55,631 km2 and the AOO to 124 km2. The current NatureServe rank of F. orolibas is Vulnerable (G3), and our reassessment does not change this status. Despite a significant increase to its known range, there persists a near complete lack of information regarding threats, baseline monitoring of population trends, and population viability for this species. This information deficiency is likely due to the difficulty of surveying for and studying F. orolibas populations, making it problematic to identify both overarching and population-specific threats that could be mitigated by land conservation or management efforts. However, although no direct threat assessments have been reported for F. orolibas, freshwater snails restricted to such subterranean or spring habitats are broadly considered at elevated risk of local extirpation or extinction owing to habitat degradation such as groundwater pollution or extraction (Lysne et al. 2008; Hershler et al. 2014). Currently, eight of the 26 known localities of F. orolibas are within federally protected lands (i.e., Shenandoah National Park in Virginia), and these populations are most likely secure. Though almost all previous collection events of F. orolibas in its northern range are not accompanied by formal reports of abundance, the findings of Keenan et al. (2017) suggest that high population densities can occur within cave ecosystems, with the TKN24 site continuing to have hundreds of individuals throughout the cave stream since it was last visited in 2019 (N.S. Gladstone personal observation).
Some localities occur in areas of high human population density and increased urbanization, with ∼27% of F. orolibas populations occurring within 5 km2 of urban land cover, including two of the three new localities reported here. Both newly discovered cave populations in Knox Co., Tennessee are immediately adjacent to suburban neighborhoods, and both have increased levels of pollution and other forms of habitat degradation (Keenan et al. 2017). Given the lack of sampling within the AVR in eastern Tennessee, it is likely that other F. orolibas populations exist within the updated range extent and into the southern AVR. We hope the discovery of these previously unreported populations promotes further study of this species.
ACKNOWLEDGMENTS
We are grateful to all of the individuals who assisted in the Tennessee cave bioinventories. We also thank R. Hershler for his initial diagnosis of this snail as Fontigens and H. Liu for useful correspondence and providing supporting sequence data.