Systematics: Described originally as Coluber erytrogrammus by Ambroise Marie Frangois Joseph Palisot de Beauvois (in Sonnini and Latreille, 1801), based on a specimen from "North America." Harper (1940) restricted the type locality to near Charleston, South Carolina. Gray (1849) erected the genus Abastor for this species, and many authors in the Virginia literature used the long-accepted combination Abastor erythrogrammus (e.g., Dunn, 1918, 1936; Richmond, 1945a, 1954a; Conant, 1958; Burger, 1958). Neill (1964) pointed out that the correct scientific name was Farancia erytrogramma. Virginia authors have followed this nomenclature since then. Two subspecies are recognized: F. e. erytrogramma (Palisot de Beauvois) and F. e. seminola Neill. Mitchell (1982b) and Conant and Collins (1991) illustrated the distributions of these subspecies. Only the nominate subspecies occurs in Virginia.

Description: A large, robust snake reaching a maximum snout-vent length (SVL) of 1,518 mm (59.8 inches) and a maximum total length of 1,733 mm (68.2 inches). In this study, tail length/total length averaged 13.0 ± 2.0% (10.5-17.4, n = 18).

Scutellation: Ventrals 157-175 (ave. = 168.9 ± 4.9, n = 35); subcaudals 35-50 (ave. = 39.5 ± 3.8, n = 34); ventrals + subcaudals 203-214 (ave. = 208 ± 3.1, n = 34); dorsal scales smooth, except keeled above the anal region in some individuals, scale rows 19 (100%, n = 31) at midbody; anal plate divided (90.7%, n = 32) or entire (9.3%); infralabials 8/8 (73%, n = 30), 7/8 (13.3%), or 7/7 (13.3%); supralabials 7/7 (86.6%, n = 30) or 8/8 (13.4%); paired internasals; loreal present; no preoculars; postoculars 2/2; temporals usually 1+2/1+2 (73.3%, n = 15), l + l/l + l (13.3%), or 1 + 1/1+2 (13.3%).

Coloration and Pattern: Iridescent, with 3 red stripes on a black dorsum; middorsal stripe extends only to base of tail, whereas lateral stripes on scale row 6 (counting from ventrals on each side) extend onto dorsum of tail; coloration below 6th scale row usually black; venter yellow with 2 primary rows of black spots (there maybe a shorter, 3d midventral row in some specimens); dorsum of head black to slate blue with some scales edged in red, forming an artistic pattern; throat, chin, infralabials, and supralabials yellow; each supralabial and some infralabials and chin shields with distinct black spots; yellow and red coloration fades to white in preservative. These are stout, cylindrical snakes with no difference in the width of the head and neck. The tail has a sharp harmless spine.

Sexual Dimorphism: Sexual dimorphism is exhibited in size, pattern, and scutellation. Adult females reached longer SVLs (781-1,518 mm, ave. = 1,051.1 ± 220.9, n = 16) than males (722-730 mm, ave. = 726, n = 2). Sexual dimorphism index was 0.45. However, tail length/total length in males was higher (17.0-17.4%, ave. = 17.2, n = 2) than in females (10.5-15.8%, ave. = 12.3 ± 1.3, n = 16). Females had a higher number of ventrals (ave. = 170.8 ± 2.3, 167-175, n = 31) than males (159.2 ± 1.5,157-161, n = 6) but fewer subcaudals (females 36.9 ± 6.3, 35-41, n = 28; males 46.7 ± 2.2, 44-50, n = 6). Counts of ventrals + subcaudals for males (204-210, ave. = 206.3 ± 2.7, n = 6) averaged slightly less than counts for females (204-214, ave. = 208.8 ± 3.0, n = 28). The midventral spots were usually more distinct in males than in females. Richmond (1954a) reported that hatchlings from New Kent County were sexually dimorphic in number of ventrals (females 170-175, ave. = 172.0 ± 1.2, n = 39; males 155-162, ave. = 159.3 ± 1.5, n = 46), subcaudals (females 35-42, ave. = 37.6 ± 1.3; males 44-49, ave. = 46.5 ± 1.0), and ventrals + subcaudals (females 207-215, ave. = 209.7 ± 1.6; males 202-210, ave. = 205.8 ± 1.8). The midventral row of black spots is more pronounced in male hatchlings than females, and males have less distinct spotting on the subcaudals than females (Richmond, 1954a; Gibbons et al., 1977). However, these are statistical differences and cannot be used to assign gender to single individuals.

Juveniles: Juveniles are colored and patterned as adults. Hatchlings from Virginia averaged 196.3 ± 13.0 mm SVL (170-222, n = 81), 231.3 ± 13.6 mm total length (197-270, n = 81), and 5.9 ± 1.1 g body mass (3.6-8.7, n = 68).

Confusing Species: No other Virginia snake is as multicolored as F. erytrogramma. Eastern Mudsnakes (Farancia abacura) have stout bodies and are shiny, but are uniformly black dorsally and red ventrally with some overlap on the sides.

Geographic Variation: There is no geographic variation in pattern, color, or scutellation in Virginia. Neill (1964) noted that there is little variation in the ventral pattern from Virginia through the Florida panhandle. The number of ventrals (males 162-165, females 175-178) and subcaudals (males 48, females 39-41) in Alabama Common Rainbow Snakes were lower (Mount, 1975) than those in Virginia, but the sample sizes were small.

Biology: Common Rainbow Snakes are generally considered highly aquatic, but Richmond (1945a) found numerous adults and juveniles buried in sandy soil in flood-plain farm fields in New Kent County. They can also be found in freshwater creeks, rivers, ditches, and marshes and in brackish water marshes, rivers, and ditches. Adults and juveniles are terrestrial for unknown periods of time and have been found in pine woods, mixed hardwood and pine woods, and open fields. Juveniles have been found under all manner of surface objects. Rainbow snakes are active on the surface during rains. In Virginia, active specimens have been recorded in every month of the year (Richmond, 1945a). Museum records are about evenly spread over all months.

Common Rainbow Snakes eat freshwater eels (Anguilla rostrata) almost exclusively, although juveniles will eat tadpoles (Neill, 1964). Richmond (1945a) observed Common Rainbow Snakes eating eels in and out of water. Prey are eaten alive and usually swallowed headfirst. Enlarged posterior teeth occur on bones of the upper jaw in F. erytrogramma, presumably for holding onto slippery prey. Active foraging has been observed between 2100 and 2330 hours (Neill, 1964). Richmond (1945a) observed an unidentified hawk eating an adult on the ice in February, and noted that dead ones with heavy parasitic nematode loads were frequently found. De Rageot (1992) found a Common Rainbow Snake in Mackay Island National Wildlife Refuge that had been killed by an otter (Lutra canadensis). Skunks (Mephitis, Spilogale) and raccoons (Procyon lotor) eat eggs in nests, and skunks are known to kill and eat hatchlings (Ernst and Barbour, 1989b).

Little is known about reproduction in this species. The following is largely taken from N. D. Richmond's observations in New Kent County (Richmond, 1945a, and pers. comm.). Farancia erytrogramma is oviparous and lays 20-40 eggs (ave. = 29.4 ± 7.1, n = 9) in late June through July. One female was observed laying eggs in an open, dry, sandy field on 15 July. Eggs (37-40 x 22-29 mm; Ernst and Barbour, 1989b) are deposited in a cavity below the soil surface. One nest cavity measured was 10 cm deep, 20 cm long, and 15 cm wide. In South Carolina, females reached maturity at about 890 mm and males at about 680 mm total length (Gibbons et al., 1977). The smallest mature male I measured was 722 mm SVL (874 mm total length) and the smallest mature female was 781 mm SVL (886 mm total length). Females remain with their eggs in the nest, presumably to confer some protection against predation. Hatching apparently occurs in September, as nests with hatchlings were found on the 18th and 23d of that month. Upon hatching, juveniles burrow into the surrounding soil. Gibbons et al. (1977) found that juveniles entered the aquatic habitat in March and April in South Carolina. Thus, hatchlings may overwinter in the soil in or near the nest.

Nothing is known of the population ecology of this species anywhere in its range. Richmond (1945a) plowed up 20 rainbow snakes in a 10-acre field (4.1 hectares) on one day in New Kent County.

Common Rainbow Snakes will not bite, but will thrash about vigorously when picked up and attempt to scratch or poke the handler with the spine on the tail (it will not pierce the skin). These snakes appear to be oblivious to humans unless picked up. Richmond (1945a) described following an adult F. erytrogramma as it traveled up a bank and into a strip of woods, completely oblivious to his presence. He stated, "I crossed directly in front of it, but it merely paused long enough to let me pass, and then continued at its same steady rate, without even raising its head."

Remarks: Other common names in Virginia are striped horn snake (Dunn, 1936); sand snake and sand hog (Richmond and Goin, 1938); and mud snake, hoop snake, horn snake, and red swamp snake (Linzey and Clifford, 1981).

The horn snake myth, described for the mud snake, F. abacura, may apply to this species as well.

Conservation and Management: This species appears secure in Virginia despite the spotty occurrence of individual locality records. There is no information on loss of populations, but pollution of aquatic habitats is likely to have detrimental effects, especially if eels are affected. Protection of freshwater wetlands and large terrestrial buffers would increase the probability that this species will remain a part of our natural heritage for the long term.

U.S. Range