Desmognathus ocoee Nicholls, 1949

Holotype: USNM 128007 (field tag J.C.N. 1001; Fig. 1) by original designation; type locality “on the surface and in crevices of cliffs at Ship’s Prow Rock, in Ocoee Gorge, beside U.S. Highway 64, nine miles airline west of Ducktown, in Polk County, Tennessee,” collected 14 November 1948 by J.C. Nicholls, Jr.

Paratypes: A series of 28 specimens including the holotype was reported, so ostensibly 27, including untraceable specimens in the “ J. C. Nicholls, Jr. collection” and the “ S. C. Bishop collection, Rochester, New York,” AMNH A-54385–6, CM 29290–1, MCZ A-26589–90 & A-28307 (MCZ catalog notes received in “ Exch. Chicago N.H. Mus. ”—original number CNHM [now FMNH] 90035), FMNH (originally given as “Chicago Natural History Museum, Chicago, Ill.”— CNHM) 57313–4 & 90034–5 (FMNH 90035 now MCZ A-28307), and untraceable specimens at “Emory University, Emory University, Ga.”

Description: A small (~ 15–54mm SVL; n = 122) mountain dusky salamander exhibiting a wide variety of color patterns and ecomorphologies throughout its range. Diagnosable from Desmognathus abditus in the Cumberland Plateau by the presence of a modest keel on the distal portion of the tail and keratinized toe tips (Fig. 20; vs. absence of both, see Drukker et al. 2018). Usually diagnosable from D. adatsihi and D. balsameus by generally lacking a solid dorsal stripe (vs. often striped with straight, wavy, or undulating borders). Some individuals of the ocoee E lineage are diagnosable from all other species by the presence of red, yellow, or orange patches on the legs or very rarely both the legs and cheeks, and all species other than D. imitator and D. perlapsus by the presence of such patches on the cheeks. Neill (1950) suggested seven diagnostic characters differentiating D. ocoee from D. perlapsus, with D. ocoee being smaller, having a shorter and narrower head, longer limbs, no vomerine teeth in adult males, parasphenoid teeth in oblong patches, more-defined dorsal color pattern, and 5–6 pairs of spots (vs. 4) on the dorsum between the limb insertions. However, Valentine (1961) demonstrated exceptional variability in these traits in both species and questioned this diagnosis. Additional work is needed to diagnose this taxon morphometrically.

Range: Three disjunct population segments (Fig. 4); one restricted to a small portion of the Cumberland Plateau between Sewanee and Orme, Tennessee in Franklin and Marion Counties; a second occurring broadly on the Cumberland Escarpment (Walden Ridge and Sand Mountain) of south-central Tennessee, extreme northwestern Georgia, and northeastern Alabama (see Anderson & Tilley 2003); and a third in the Nantahala, Unicoi, and southern Blue Ridge Mountains of western North Carolina, southeastern Tennessee, and north-central Georgia. Additional populations probably referable to this species likely remain to be discovered and sampled. We examined a historical collection (MCZ A-143377–86) from Horse Cove, Rich Mountains (GA: Gilmer) which appears to be this species based on morphological gestalt and collected a single recent individual ourselves (Fig. 21). We have also received anecdotal reports from the Cohutta Mountains in Georgia (Fig. 4), due south of the type locality along the same general mountain range (S.P. Graham, pers. comm.), and their identity is of great interest.

Etymology: Named after the type locality in Ocoee Gorge. Various informal sources report “ocoee” as a Cherokee name for the Passionflower, Passiflora incarnata (e.g., the state symbol of Tennessee; https://www. tn.gov/about-tn/state-symbols.html). Name is a non-Latin noun used in apposition. The widely-used common name is “ Ocoee Salamander ” (Schmidt 1953).

Notes: Consists of the D. ocoee concept of Tilley & Mahoney (1996) and Anderson & Tilley (2003) in part, the ocoee E–H lineages of Kozak et al. (2005) and Beamer and Lamb (2020), and the apalachicolae A2 lineage of Beamer & Lamb (2008) as defined by Pyron et al. (2022c). Future work may reach more subtle and complex taxonomic conclusions regarding the specific identity of the ocoee E and apalachicolae A2 lineages. Extensive data on the biology of this species exist but will need to be carefully disambiguated from D. perlapsus by reference to geography; see bibliographies in Valentine (1964) and Camp & Tilley (2005).