Systematics: Originally named as Coluber erythrogaster by Johann Reinhold Forster in 1771, who based his identification on the snake illustrated in Plate 46 of Mark Catesby's (1731-1743) historical publication. Conant (1949) clarified the considerable confusion about the taxonomy of this species and restricted the type locality to Charleston County, South Carolina. This species had long been considered a subspecies of Natrix sipedon; it was first considered a full species within Natrix by Clark (1903). Rossman and Eberle (1977) determined that it was necessary to distinguish North American water snakes from Old World water snakes and restricted the genus Natrix to the latter group. The next available generic name for North American species was Nerodia, a name first used for these snakes by Baird and Girard (1853). Until this change, however, all authors in the Virginia literature used the combination Natrix erythrogaster for this species (e.g., Dunn, 1936; Carroll, 1950; Werler and McCallion, 1951; Conant, 1958, 1975; Delzell, 1979). Subsequent authors have used the current nomenclature (e.g., Mitchell, 1981a; Pague and Mitchell, 1991). Makowsky et al. (2010, Mol. Phylogenet. Evol.55: 985–995) demonstrated using mitochondrial data that this taxon represents a single widespread species with no concordance to any of the described subspecies. As such we do not recognize subspecies.

Description: A large, unpatterned water snake reaching a maximum total length of 1,575 mm (62.0 inches) (Conant and Collins, 1991). In Virginia, maximum known snout-vent length (SVL) is 1,189 mm (46.8 inches) and maximum total length is 1,474 mm (58.0 inches). In the present study, tail length/total length was 17.2-27.7% (ave. = 21.9 ± 2.5, n = 41).

Scutellation Ventrals: Ventrals 135-154 (ave. = 147.5 ± 3.0, n = 64); subcaudals 50-84 (ave. = 71.3 ± 7.8, n = 44); ventrals + subcaudals 200-232 (ave. = 218.7 ± 7.5, n = 44); dorsal scales keeled, scale rows usually 22-23 (80.4%, n = 51) at midbody, but may be 21, 24, or 25 (19.6%); anal plate divided; infralabials 10/10 (73.7%, n = 57) or other combinations of 9-12 (26.3%); supralabials 8/8 (97.0%, n = 57) or other combinations of 7-9 (3.0%); loreal scale present; preoculars 1/1; postoculars usually 3/3; temporal scales 1+3/1+3 (70.2%, n = 57) or other combinations of 1-3 (29.8%).

Coloration and Pattern: Dorsal pattern uniformly rich reddish brown; venter of body and tail uniformly orange to red; dorsal coloration encroaches onto margins of venter in some individuals; dorsum of head dark brown; rostrum and lateral portion of snout tan to reddish brown; chin and infralabials white to yellowish; supralabials reddish, tan, or white and may be outlined in dark brown; red color on venter fades to cream or light tan in preservative. These stout snakes are distinctly bicolored with a distinct head and large eyes. Eye diameter is equal to or greater than the distance between the eye and nostril (Conant, 1943a).

Sexual Dimorphism: Adult females averaged larger (738-1,189 mm, ave. = 958.0 ± 117.9, n = 29) than males (550-1,100 mm, ave. = 807.0 ± 141.9, n = 24) in body size, and reached a larger total length (to 1,474 mm) than males (to 1,235 mm). Sexual dimorphism index was 0.19. Females also reached a larger body mass (to 813 g) than males (to 556 g). Tail length/total length was similar between sexes (females 18.1-26.6%, ave. = 21.1 ± 2.4, n = 20; males 17.2-27.7%, ave. = 22.6 ± 2.4, n = 20), as was the number of ventral scales (females 145-154, ave. = 148.5 ± 2.1, n = 33; males 135-153, ave. = 146.3 ± 3.4, n = 30). The number of subcaudal scales was higher in males (50-84, ave. = 74.6 ± 8.5, n = 22) than in females (63-81, ave. = 67.4 ± 4.6, n = 21). Consequently, the number of ventrals + subcaudals was higher in males (200-232, ave. = 221.2 ± 8.7, n = 21) than females (208-230, ave. = 215.6 ± 4.6, n = 21). Males and females do not differ in color or pattern.

Juveniles: Juveniles are strongly patterned with a series of 4-6 dark crossbands anteriorly that change to middorsal blotches with alternating dark lateral blotches at the midbody region. Number of body crossbands and blotches averaged 33.4 ±2.7 (21-40, n = 52). The venter of the body and tail is uniformly cream to yellow. The chin and labials are usually cream in color, and the labial scales are outlined in dark brown. The rostrum and lateral portions of the snout are tan. The juvenile pattern becomes obscure with age. At birth, juveniles were 192-244 mm SVL (ave. = 218.5 ± 13.4, n = 74), 251-310 mm total length (ave. = 282.1 ± 17.1), and 5.7-9.8 g body mass (ave. = 8.4 ± 0.7, n = 44).

Confusing Species: Adults of this species may be confused with Pantherophis alleghaniensis and Coluber constrictor, which have weakly keeled or smooth scales, respectively, and have black and white or gray venters. Some darkly colored individuals of Nerodia sipedon in southeastern Virginia may be confused with N. erythrogaster, but they have a strongly patterned venter rather than a uniformly reddish venter. The differences in ventral pattern and coloration, as well as the larger number of anterior dorsal crossbands in N. sipedon, will distinguish juveniles of the two species of water snakes in southeastern Virginia.

Geographic Variation: No geographic variation in color, pattern, or scutellation occurs in Virginia.

Biology: Nerodia erythrogaster inhabits swamps; ditches; freshwater marshes; and low, wet areas in all vegetative habitats. It has been found in cypress swamps, fields traversed by drainage ditches, formerly forested areas with pools of standing water, and streams in pine forests. Plain-bellied Watersnakes are often seen lying on logs and vegetation in swampy habitats, but can also be found under surface objects, such as boards, litter, and logs. They are primarily diurnal, although they are known to move overland at night, especially during and after warm rains. Nerodia erythrogaster is active in Virginia 23 May through 18 September, according to museum records, but may be found during warm periods in other months. Body temperatures of two snakes found under boards were 27.1°C and 31.6°C. Body temperatures of two basking snakes were 22.0°C and 27.2°C; air temperatures were 25.8°C and 27.4°C respectively, indicating that they had only recently emerged from their hiding places.

Almost nothing is known of the prey of this snake in Virginia. One juvenile from the Dismal Swamp contained the tail of a salamander, probably the Red-backed Salamander (Plethodon cinereus). Wright and Wright (1957) listed fish, crayfish, frogs, and salamanders as prey. All prey are held by the long, recurved teeth and eaten alive. Predators include Snapping Turtles (Chelydra serpentina), hawks (Buteo spp.), and humans. Tupacz (1985) observed a red-shouldered hawk (Buteo lineatus) preying on an adult in the Dismal Swamp. Other predators and prey were summarized in Ernst and Barbour (1989b).

Plain-bellied Watersnakes are viviparous and bear 5-37 young alive in late summer (Fitch, 1970; Ernst and Barbour, 1989b). Linzey and Clifford (1981) stated that up to 50 young are born in a litter. Mating occurs April through June (Ernst and Barbour, 1989b); dates are unknown in Virginia.

Size at maturity for Virginia snakes is about 700 mm SVL for females and 500 mm SVL for males. Litter sizes for six females ranged from 11 to 25 (ave. = 20.8 ± 5.2). Birth dates were 6-21 September.

Nothing is known of the population ecology of red-bellied water snakes anywhere in their range. Werler and McCallion (1951) noted that they were fairly common in Seashore State Park, but there may have been a substantial decline in the population since the 1970s, possibly due to the low water conditions in the 1980s (C. A. Pague, pers. comm.).

Like all water snakes, this species will bite and discharge large quantities of musk from glands at the base of the tail when picked up.

Remarks: Other common names in Virginia are copper-belly watersnake (Dunn, 1936), copper-bellied moccasin (Carroll, 1950), copperbelly and moccasin (Linzey and Clifford, 1981), and red-bellied water moccasin (Brothers, 1992).

A large percentage of the Virginia sample (35.0% of 60) had partial tails. This stems from softening of the tissues in the tail during hibernation, apparently due to a disease causing splitting of the skin, and the habit of twisting around if a predator grabs them by the tail. Unlike in many lizards, however, a new tail will not grow back.

Linzey and Clifford (1981) noted that a specimen of this species was reported and verified from Tom Boo (1974) in Alexandria, Virginia. Since there have been no subsequent reports of this species in Northern Virginia this discovery is considered a liberated or escaped captive.

The western and northern margins of the range of this species in Virginia need delineation, and all aspects of its natural history should be reported.

U.S. Range