Species Description: Nicholls, J. C., Jr. (1949) A new salamander of the genus Desmognathus from East Tennessee. Journal of the Tennessee Academy of Science 24: 127–129.
© 2011 Todd Pierson (1 of 85)
Desmognathus ocoee Nicholls, 1949
Carlos D. Camp1
1. Historical versus Current Distribution. Ocoee salamanders (Desmognathus ocoee) occur in two distinct regions. The larger region lies in the southwestern Blue Ridge and the adjacent Piedmont Physiographic Provinces. Blue Ridge populations include those in the Balsam, Blue Ridge, Cowee, Great Smoky, Nantahala, Snowbird, Tusquitee, and Unicoi mountains, as well as lower elevation populations in gorges of major rivers, including the Hiwassee, Ocoee, Tugaloo, and Tallulah (Valentine, 1961, 1964; Martof and Rose, 1963; Tilley and Mahoney, 1996). Populations also occur in adjacent areas of the Piedmont Province of Georgia (Martof and Rose, 1963) and South Carolina (Valentine, 1964). They occur as far south as the Gainesville Ridges (sensu Wharton, 1978) in the upper Piedmont of northeastern Georgia (Camp, 2000). A second cluster of populations occurs in the Appalachian Plateau of northeastern Alabama (Folkerts, 1968; Mount, 1975; Tilley and Mahoney, 1996). Ocoee salamanders occur over a greater elevational range (from low-lying gorges to mountain tops) than any other species of Desmognathus (Petranka, 1998). Populations of Ocoee salamanders in different mountain ranges have undergone considerable genetic differentiation and may ultimately be subdivided into several additional taxa (Tilley and Mahoney, 1996). Tilley (1997) speculated that interactions with seal salamanders may have contributed to the isolation and genetic differentiation of units of the D. ochrophaeus complex.
2. Historical versus Current Abundance. Ocoee salamanders are widespread and abundant in seepage areas, on wet rock faces, along banks of small streams, and in mesic forest floor habitats. Individuals often are found some distance from water in mesic forests at higher elevations. Low elevation populations are less terrestrial than those at high elevations; those on the Georgia Piedmont are restricted to areas of streambed except during wet weather (C.D.C., personal observations). Population densities on rock faces generally exceed those along streams. Rock face densities have been reported to be six adults/m2 and 18–25 total individuals/m2 (Huheey and Brandon, 1973; Tilley, 1980; Bernardo, 1994).
Petranka et al. (1993) demonstrated that clearcut timber harvesting negatively affects the number of total Desmognathus individuals, including members of the D. ochrophaeus complex. They estimated that, at the rate of clearcutting carried out during the 1980s and early 1990s, the Appalachian forests of North Carolina lost as many as 14 million salamanders of all species each year during that time. However, Ash and Bruce (1994) strongly disagreed with these estimates and did not consider clearcutting to have as strong an impact on native salamanders. Ocoee salamanders are significantly more abundant in cove hardwood stands > 85 yr of age than in younger stands (Ford et al., 2002). No significant change in abundance was noted in 20 yr of surveys of Ocoee salamanders carried out in undisturbed habitats in the Appalachian Mountains of southwestern North Carolina (Hairston and Wiley, 1993; Hairston, 1996).
3. Life History Features.
A. Breeding. Courtship occurs on land. Eggs are laid in or near flowing water.
i. Breeding migrations. Gravid females choose nesting sites as many as 2–3 wk before ovipositing (Forester, 1981).
ii. Breeding habitat. Courtship occurs on land from September–June (Martof and Rose, 1963; Huheey and Brandon, 1973; Forester, 1977; Houck et al., 1985).
i. Egg deposition sites. Females oviposit in cavities beneath rocks, in or under decaying logs, in leaf litter or under moss, under cover objects (e.g., rocks) near seepages, springs, and small streams, and behind vegetation or in crevices associated with wet cliffs (Pope, 1924; Martof and Rose, 1963; Folkerts, 1968; Forester, 1977; Bruce, 1990; Bernardo and Arnold, 1999; Camp, 2000). As is the case in other plethodontid species, females brood their eggs. Brooding appears to serve in defense against predators, remove dead eggs from the nest, reduce egg desiccation, reduce fungal attacks, and perhaps help the hatchlings escape (Tilley, 1972; Huheey and Brandon, 1973; Forester, 1978, 1979, 1984). Oviposition occurs from June–September (Martof and Rose, 1963; Folkerts, 1968), and some females may brood eggs during the winter (Tilley, 1977). Salamanders that nest near the headwaters of seepages risk the loss of reproductive investment due to failure to lay or the loss of eggs to desiccation during periods of drought (Camp, 2000).
ii. Clutch size. Females may congregate and oviposit their eggs in close proximity in areas of favorable habitat (Martof and Rose, 1963). Clutch sizes determined from ovarian follicles vary from 5–29 and are positively related to female body size in most populations (Martof and Rose, 1963). Clutch sizes determined from counting eggs laid in natural clutches averaged 12 (range = 5–18) in Alabama (Folkerts, 1968); 13 (range = 5–23) in North Carolina (Bruce, 1996); and 15 (range = 5–21) in Georgia (C.D.C., unpublished data). There is a strong, positive relationship between clutch size and the body size of the respective attending female in Georgia (C.D.C., unpublished data). Although clutch size and female body size are positively related, the same relationship may not be true between clutch size and female age (Tilley, 1980). Females are capable of oviposition in successive years (Forester, 1977; Tilley, 1977, 1980), but probably seldom reproduce annually throughout their lives (Tilley, 1977). Eggs measure about 2–3 mm in diameter (Martof and Rose, 1963; Folkerts, 1968; Huheey and Brandon, 1973; Bruce, 1990).
Clutches observed near Highlands, North Carolina, at 1,265 m hatched in 53–66 d (Tilley, 1972). Seven clutches observed at approximately 250 m on the Georgia Piedmont hatched in 32–39 days (C.D.C., unpublished data). Clutches near hatching have been found in August–September (Pope, 1924; Noble, 1927a; Martof and Rose, 1963; Tilley, 1972). Eggs on the Georgia Piedmont hatch from September–October (C.D.C., personal observations). Hatching occurs at approximately 11–12 mm SVL (Tilley, 1980; Bruce, 1989).
i. Length of larval stage. The larval period in montane populations lasts between 9–10 mo (Huheey and Brandon, 1973; Bruce, 1989). Larvae in a Georgia Piedmont population metamorphose in 7–8 mo (C.D.C., unpublished data). Beachy (1995) demonstrated that food and temperature can have environmental effects on the larval period under experimental conditions.
ii. Larval requirements.
a. Food. Larvae feed on small aquatic invertebrates. Individuals in captive environments readily fed on live Tubifex worms (Beachy, 1995).
b. Cover. Larvae are found in shallow water associated with seepages or in thin films of water flowing over rock faces (Huheey and Brandon, 1973).
iii. Larval polymorphisms. Ocoee salamanders are not known to exhibit distinct larval polymorphisms.
iv. Features of metamorphosis. Metamorphosis usually occurs in May and June, with animals averaging 13–14 mm SVL (Bruce, 1989; Bernardo, 1994). Newly metamorphosed individuals averaged 61–94 mg in weight under experimental conditions (Beachy, 1995).
v. Post-metamorphic migrations. Ocoee salamanders are not known to migrate.
vi. Neoteny. Neoteny is not known in this species.
D. Juvenile Habitat. Juveniles occur in seepages (Bernardo, 1994), on wet rock faces (Huheey and Brandon, 1973), and under cover along edges and in the streambeds of small streams (C.D.C., personal observations). Growth rates vary depending on elevation, microhabitat (rock faces versus forest floor), and size. Generally, growth rates are about 5–7 mm SVL/yr until animals reach sexual maturity, when growth slows (Tilley, 1977, 1980).
E. Adult Habitat. Ocoee salamanders have a strong affinity for the headwaters of first-order streams in montane regions (Tilley, 1997). Individuals are common in seepages, on wet rock faces, and in streambeds of larger streams (Petranka, 1998). Ocoee salamanders will move away from the streambeds under moist conditions, and in the mesic forests of higher elevations, individuals become components of the terrestrial salamander community (Hairston, 1987). Ocoee salamanders are substantially more abundant in cove hardwood forests of > 85 yr of age than in younger stands. They are also more abundant in cove forests with a significant amount of emergent rock. This possibly reflects the positive effect that shallow and emergent rock has on soil moisture retention as well as the abundance of refugia (Ford et al., 2002). Camp (2000) reported a seepage that dried due to drought, resulting in the failure to oviposit and the loss of clutches by female Ocoee salamanders that inhabited the seepage. Following an extended drought during which this same seepage dried every summer for 4 yr, the subpopulation collapsed, apparently due to desiccation-induced mortality (Camp, 2003b; C.D.C., unpublished data).
F. Home Range Size. Home ranges are small. In a study of salamanders inhabiting a wet rock face, Huheey and Brandon (1973) marked individuals that moved an average of 47 cm between observations. Movements centered around a small home range, to which displaced individuals successfully homed. Individual salamanders were randomly distributed on the rock face.
G. Territories. Adults are aggressive and defend space against conspecific intruders (Jaeger, 1988). Males are aggressive towards other males (Verrell and Donovan, 1991).
H. Aestivation/Avoiding Dessication. Unlikely. Individuals are active throughout the summer (Huheey and Brandon, 1973; Tilley, 1980).
I. Seasonal Migrations. Unknown. At lower elevations, individuals move into forest-floor habitats from streams during wet, mild weather (C.D.C., personal observations). It is possible that individuals in high elevation sites move between streams and surrounding forests in association with seasonal weather patterns.
J. Torpor (Hibernation). Adults and juveniles may aggregate in seepages or underground retreats during the winter (Shealy, 1975). Individuals become active in mild weather conditions throughout the winter; winter-collected specimens on a wet rock face will feed (Huheey and Brandon, 1973).
K. Interspecific Associations/Exclusions. Throughout their range, Ocoee salamanders are sympatric with several congeneric species. They often occur syntopically with black-bellied salamanders (D. quadramaculatus) and seal salamanders (D. monticola). They additionally may occur with seepage salamanders (D. aeneus). Except for the completely aquatic shovel-nosed salamander (D. marmoratus), southern Appalachian desmognathines assort by body size along the stream-forest interface with larger species being more aquatic and smaller ones occurring more terrestrially. This pattern is evident both along a horizontal gradient from stream–stream bank–forest and along a vertical gradient from stream–seepage–forest (Organ, 1961a). Ocoee salamanders are relatively small and occur more terrestrially than most of their sympatric congeners. The observed pattern of desmognathine assortment was explained initially as niche partitioning among competitors (Hairston, 1949; Organ, 1961a). Tilley (1968) and Hairston (1980c), however, suggested that interspecific predation was a more likely cause. A number of studies have attempted to determine which was the more probable factor (e.g., Kleeberger, 1984; Carr and Taylor, 1985; Hairston, 1986; Southerland, 1986a,b,d). They generally concluded that some combination of predation and aggressive interference was an important factor in interspecific desmognathine interactions. Hairston (1986) made the strongest case for predation, with competition being a secondary factor. His statistical methods have been criticized, however (Jaeger and Walls, 1989). Although large desmognathines readily eat small ones in artificial environments, dietary studies show that neither black-bellied salamanders nor seal salamanders are important predators of heterospecific congeners (Camp, 1997b). The lack of predation under natural conditions is probably a result of differential habitat selection and behavioral avoidance (perhaps involving chemical cues) of larger congeners by small individuals. Predation by large species may have been important historically in the organization of desmognathine communities. Alternative hypotheses based on abiotic factors, rather than biotic ones such as competition and predation, recently have been proposed to explain patterns of habitat preference among desmognathines (Bruce, 1996; Camp et al., 2000).
At elevations above 900 m in the Great Smoky Mountains, Ocoee salamanders are syntopic with imitator salamanders (D. imitator; Tilley, 1985), similar members of the D. ochrophaeus complex (Tilley et al., 1978). In these areas, Ocoee salamanders are more terrestrial than imitator salamanders (Tilley et al., 1978). Imitator salamanders may exclude Ocoee salamanders from lower elevation sites (Tilley et al., 1978; Bernardo, 2000). At Waterrock Knob in North Carolina, Ocoee salamanders may be largely excluded from both high and low elevation rock faces by Santeetlah dusky salamanders (D. santeetlah) and imitator salamanders, respectively (Tilley, 2000a). Ocoee salamanders and spotted dusky salamanders (D. conanti) are broadly sympatric but apparently largely exclude each other in areas of the Georgia Piedmont (C.D.C., personal observations) and the Appalachian Plateau region of Alabama (Folkerts, 1968; Mount, 1975).
L. Age/Size at Reproductive Maturity. Variable, and may increase with elevation (Tilley, 1977, 1980; Bernardo, 1994). In Bruce's (1990) study, both sexes apparently reached sexual maturity at 3 yr old; females first oviposit at 4 yr old. Females mature at about 29–30 mm SVL, males at or slightly smaller than 28 mm (Martof and Rose, 1963; Valentine, 1964; Huheey and Brandon, 1973). In skeletochronological studies of North Carolina populations, Castanet et al. (1996) and Bruce et al. (2002) found that males mature at 3–4 (generally 3) yr old, and females one year later (4–5 yr, but generally at 4 yr old).
M. Longevity. Using skeletochronological techniques, Castanet et al. (1996) and Bruce et al. (2002) estimated the ages of several males and females to be as long as 7 yr. They aged one male at 10 yr. Tilley (1977) recaptured several males 5 yr after their initial captures as large adults. These individuals must have been in at least their fifth adult year (8 yr of age) and were probably considerably older than that.
N. Feeding Behavior. As with most salamanders, Ocoee salamanders are generalist feeders, taking an array of small invertebrates, especially insects. Huheey and Brandon (1973) surveyed the stomachs of 54 individuals and reported flies (both larval and adult), ants, wasps, beetles, spiders, mites, an isopteran, a trichopteran, and a larval salamander (Desmognathus sp.). Folkerts (1968) additionally reported collembolans, lepidopterans, and a grasshopper.
Female Ocoee salamanders occasionally cannibalize their own eggs, the feeding response often being elicited by the presence of dead eggs (Tilley, 1972; Forester, 1979). They also occasionally eat conspecific hatchlings (Forester, 1981). A brooding female consumed its entire clutch of healthy eggs after she had been disturbed and moved to laboratory conditions (Bruce, 1990).
O. Predators. Woodland birds are probable predators (Petranka, 1998), as are various snakes, including ring-necked snakes (Diadophis punctatus), common garter snakes (Thamnophis sirtalis), and northern water snakes (Nerodia sipedon; Huheey and Brandon, 1973). On one occasion, a young northern water snake was found in a pile of loose shingles that had been dumped into a seepage on Rabun Bald Mountain in Georgia (C.D.C., personal observation). Ocoee salamanders were extremely abundant among the shingles, and the snake appeared to be gorged with salamanders. Although large desmognathine salamanders have been implicated as predators in experimental work (Hairston, 1986; Formanowicz and Brodie, 1993), dietary studies of seal salamanders and black-bellied salamanders indicate that neither species is a significant predator of Ocoee salamanders (Camp, 1997b), perhaps as a consequence of predator avoidance mechanisms. Spring salamanders (Gyrinophilus porphyriticus), on the other hand, are important predators of Ocoee salamanders (Bruce, 1979; Bernardo, 2002), explaining why Ocoee salamanders are more likely to flee from spring salamanders than from black-bellied salamanders (Hileman and Brodie, 1994).
P. Anti-Predator Mechanisms. Ocoee salamanders autotomized their tails when attacked by chickens in experimental trials (Labanick, 1984). These salamanders flee from predators such as spring salamanders (Hileman and Brodie, 1994). Ocoee salamanders may remain immobile to avoid detection (Dodd, 1990a). When attacked by snakes, Ocoee salamanders flip their bodies and bite, directing their bites to the face of the snake (Brodie et al., 1989).
Q. Diseases. Unknown.
R. Parasites. Goater et al. (1987) reported eight species of adult helminths, including nematodes (Capillaria inequalis, Thelandros magnavulvaris, Omeia papillocauda, Falcaustra plethodontis, Cosmocercoides dukae), flukes (Brachycoelium elongatum, Phyllodistomum solidum), and a tapeworm (Cylindrotaenia amaericana). They also reported two species of larval helminth, an ascaridoidid nematode and the plerocercoid of a proteocephalan tapeworm. Leeches (Oligobdella biannulata) are apparently rare parasites of Ocoee salamanders (Sawyer and Shelley, 1976; Goater, 2000).
4. Conservation. Ocoee salamanders are among the most common salamanders of the southern Appalachians (Petranka, 1998). In 20 yr of surveys of desmognathine salamanders carried out in undisturbed habitats in the Appalachian Mountains of North Carolina, Hairston (1996) found no significant change in the abundance of Ocoee salamanders. Petranka (1998) suggested that the isolated populations in northeastern Alabama may be more vulnerable to environmental degradation than other populations.
Because of the reliance of populations of this species on moist habitats, the greatest potential threat is probably the removal of the protective forest canopy through the harvesting of timber and the resulting desiccation of habitats. Petranka et al. (1993) estimated that the clearcut logging of Appalachian forests of North Carolina during the 1980s and early 1990s killed millions of salamanders, including members of the D. ochrophaeus complex, each year. Ash and Bruce (1993), however, disputed these estimates and considered them to be exaggerations of the actual number killed. Desmognathine salamanders currently are abundant in areas of the southern Appalachians (the Great Smoky Mountains) that have been logged extensively in the past (S.G.T., personal observations). Appalachian seepages occasionally dry up, negatively affecting both reproduction (Camp, 2000) and survival in Ocoee salamanders (C.D.C., unpublished data). It is not known how periodic drought may interact with timber harvesting to affect populations of Ocoee and other seepage-dwelling salamanders. Populations of Ocoee salamanders occupying high elevation sites in the southern Appalachians may also be vulnerable to the effects of acid precipitation.
Although Ocoee salamanders are abundant, their considerable genetic differentiation across mountain ranges indicates that relatively isolated genetic subunits may be vulnerable to declines and the potential loss of genetic diversity. Some of these subunits may ultimately be described as separate taxa (Tilley and Mahoney, 1996).
1Carlos D. Camp
2Stephen G. Tilley
Literature references for Amphibian Declines: The Conservation Status of United States Species, edited by Michael Lannoo, are here.
Feedback or comments about this page.
Citation: AmphibiaWeb. 2021. <https://amphibiaweb.org> University of California, Berkeley, CA, USA. Accessed 22 Oct 2021.
AmphibiaWeb's policy on data use.