Crab-eating Snake, Fordonia leucobalia

1837a Homalopsis leucobalia Schlegel, Essai sur la Physionomie des Serpens. 2:345. Type locality: Timor. Holotype: RNHL 1161. Collector: unknown.
1842 Fordonia leucobalia – Gray, Zoology Miscellany, p. 67.
1849 Fordonia unicolor Gray, Catalogue of Specimens of Snakes in the Collection of the British Museum, p. 77. Type locality: Borneo. Syntypes: BMNH 111.22.2.a. Collector: Sir Belcher. BMNH 111.22.2.b Lowe’s collection.
1854 Hemiodontus leucobalia – Duméril, Bibron and Duméril, Érpétologie générale… reptiles, 7:884.
1863 Hemiodontus chalybaeus Jan, Elanco systematico degli ofidi, p. 79. Type locality: Singapore. Holotype: possibly the Milan Museum. Collector: unknown.
1868 Fordonia bicolor Theobald, Journal of the Linnaean Society of Zoology 10:56. Type locality: near Rangoon (= Yangon, southern Myanmar.). Holotype: apparently lost. Collector: unknown.
1877 Fordonia papuensis Macleay, Proceedings of the  Linnaean Society of New South Wales,2:35. Type locality: Kataw, New Guinea. Holotype: Australian Museum. Collector: unknown.
1878 Fordonia variabilis Macleay, Proceedings of the  Linnaean Society of New South Wales, 2:219. Type locality: Port Darwin, Australia. Syntypes: three specimens, Macleay Museum, Syndey? Collector: Mr. Spalding.
            The name leucobalia, is from the Greek leucomeaning white, and balios meaning spotted, dappled or piebald. This description fits several of the color morphs from northern Australia that may be black and white spotted, or a morph with red, black, and white spots; and it seems likely that the color morph is also present at the type locality, Timor, Indonesia.
            At the western edge of its range Fordoniais known from India’s Nicobar Islands, and on the coastal mainland from Bangladesh and Myanmar it ranges eastward along the coasts of islands and the mainland throughout Southeast Asia to the Philippines, and southward to New Guinea and Northern Australia. It should be emphasized that like other coastal marine homalopsids this snake probably does not occur inland or in open ocean environments accept as waifs.
In Southeast Asia Fordonia is almost always a uniform black with a white belly, however in Australia, New Guinea and possibly in extreme eastern Indonesia (Timor) it has multiple color morphs. Fordoniahas 23 – 27 scale rows at midbody but the most frequently encountered count is 25 rows (85%); it also has 5 – 7 (usually six) upper labials, a low count for a homalopsid; and it usually lacks a loreal scale.  Additionally, Fordonia has an internasal scale that completely separates the two nasal scales, a character state that will immediately separate if from all Enhydris. In its mangrove and mud flat habitat it is most likely to be confused with Gerarda prevostiana which also has a uniform black dorsum over most of its range. However, Gerarda has 17 scale rows at midbody. Fordonia has also been mistaken for Enhydris punctata that has a similar number of scale rows at midbody, but it possesses a loreal scale, has 12 or more upper labials, and nasal scales that contact each other.
            The largest male measured for this study had an 817 mm total length with a 103 mm tail. The largest female had a total length of 764 mm with an 89 mm tail. The smallest was an unsexed juvenile that had a total length of 192 mm with a 19 mm tail. Rooij (1917) mentioned a specimen that had a total length of 930 mm and a 110 mm tail. Smith (1943) reported a male that had a total length of 780 mm with a 100 mm tail and a female that was 1065 mm in total length with a 125 mm tail. Bergman (1960) suggested newborns are less than 180 mm SVL, and adults are sexually mature at 330 mm SVL. His largest male had a total length of 606 mm, and a 95 mm tail; his largest female had a total length of 637 mm with a 72 mm tail. Bergman (1960) noted males have longer tails than females, and reported that Kopstein found females to have 32 – 36 subcaudal scales; males have 39 – 45 subcaudal scales in the Java population. The tail/SVL ratio for the New Guinea population determined during this study was 14 – 18% in males and 12 – 15% in females (males n = 10, females n = 8).
            At the base of the tail the width is 75% of the height, based on the average of five specimens. Karns et al. (2002) studied this snake at Pasir Ris Park in Singapore and took measurements on 19 snakes; their data is summarized in Table 16.
External Morphology
The head is slightly distinct from the neck; it is short and very depressed. The eye diameter is slightly less that the eye-nostril distance. The lower jaw is countersunk. The tail is compressed.
On the head the rostral scale is rounded and barely visible from above. All of the head plates are imbricate. The nasals are semidivided (rarely entire), the nasal cleft may touch the prefrontal or the internasal, and in rare instances where a loreal is present the nasal cleft may touch that scale. The nasals are separated by the internasal. The internasal is single and slightly smaller in area than a single nasal. The prefrontals are about the same size as the nasal and they touch the loreal, the nasal, and the supraocular. The frontal is about 80% of the interorbital distance. The parietals are about 1.2 times the length of the frontal. The supraocular is elongated with rounded anterior and posterior margins, the preocular is single, there are two postoculars, and the bottom scale is larger than the upper scale. Loreal scales are usually absent. There are two primary temporal scales, the upper one is very large. Upper labials number five (rarely six), the largest is usually four or five, and the third (rarely the second) enters the orbit. On the chin the lower labials number 5 – 7 (usually six), and the first three contact the anterior chin shields. There are two pairs of chin shields (rarely one or three); the anterior pair is usually the largest; the second pair if present are in contact with each other. Gular scales number 3 – 6. On the body the dorsal scale rows number 25 – 27 on the neck; at mid-body dorsal scale rows number 23 – 27 (usually 25 – 85% of the time). Posterior dorsal scale rows are reduced to 21 – 23. The ranges of dorsal scales are somewhat misleading in that most specimens have 25 scale rows at mid-body, three (11.5%) specimens of 26 had 23 rows, and one (3.8%) specimen of 26 had 27 rows. The dorsal scales are imbricate, smooth, lanceolate (even the first row tends to be elongate), and scales have fine striations, otherwise they are smooth. Near the vent the scales become ovate and this shape continues onto the tail. The ventral scales number 140 – 150. They are wide, about three times the height of a nearby dorsal scale. No indication of sexual dimorphism or geographic variation in ventral scale counts was detected; males have 142 – 150 ventral scales, females have 140 – 150. The anal plate is divided; it is about twice the length of the preceding ventral scale, and the Singapore population has the ventral preceding the anal plate divided. Additionally, Karns et al (2002) reported sexual dimorphism in the anal plate. On the tail the subcaudal scales are usually divided, however some individuals have mixed subcaudal scales (some single and some divided scales), they number 28 – 41. In the Malaysian-Singapore population males have 29 – 34 subcaudal scales and females have 28 – 29. However, the New Guinea population has males with 35 – 41 subcaudal scales and females with 30 – 34. Gyi (1970) observed that, “Males have a slightly longer and more compressed tails than females. In males the tail is provided with a dorsal hump extending from the level of the fourth to about the 26th subcaudal. In females the tail gradually tapers to a point.”
            Color and Pattern Most of the specimens examined from Southeast Asia have a uniform black dorsum and a white belly. However, the Australia-New Guinea population shows dramatic color and pattern polymorphism. Gow (1989) reported six color morphs and O’Shea (1986) noted five different color morphs in the Western Province, Paupa New Guinea, and notes that no two specimens had the same color pattern. Color morphs I have seen include: (1) a uniform gray-black dorsum with the first two or three scale rows cream; (2) a black dorsum with white cross bars or spots on the vertebral line and lateral white spots or mottling; (3) a red dorsum with white crossbars each bordered in black; (4) a yellow or orange dorsum with black vertebral spots or crossbars.

This is a snake of the mangrove forest and associated mud flats. It may on occasion enter surrounding habits such as monsoon forest or open ocean (O’Shea, 1986; Campden-Main, 1970) but it is unlikely that these areas are supporting populations of this species.
            Fordoniauses the intertidal burrow system. Karns et al. (2002) monitored the movements of three male snakes over a period of five weeks using radiotelemetry in Singapore’s Pasir Ris Park, a mangrove forest. Snakes monitored for 7 – 10 days were relatively sedentary (46.6% of the days they were inactive) and when they did move it was only for short distances (1.8 – 14.0 m, <!–[if supportFields]> eq o(X,¯) <![endif]–><!–[if supportFields]><![endif]–>= 4.4 m). Two of the snakes were always located in mud lobster mounds (100% of the telemetric locations), but did not show a preference for lobster mounds of a particular size. A third snake used the mud-root tangle of the mangrove 59% of the time, and an area under a boardwalk the other 41% of the time. It was usually associated with two mud lobster mounds. While these individual snakes frequented the mud lobster mounds and the landward edge of the mangrove, they were observed foraging on tidal mud flats. Body temperatures (26.3 – 29.0°C, <!–[if supportFields]> eq o(X,¯) <![endif]–><!–[if supportFields]><![endif]–>= 28°C) for these three snakes were consistently above the temperature of the microhabitat they were using and significantly different. Of the three monitored snakes only one moved once during the day, and all other activity was nocturnal, and the snakes were active throughout the night.

Nt, Australia
            The literature supports the idea that this species uses mangrove forests, mud flats, and makes use of crustacean burrows across its distribution. Macnae (1968) described this species in the mangrove habitats of the Malayan Peninsula and Thailand. In Java, Kopstein (1931) stated that they live in crab burrows; and Hoesel (1959) described them hiding in holes in the mud. In Australia, Cogger (1981) reported mangrove habitat. And, Gow (1989) wrote, “A nocturnal aquatic species which inhabits mangrove flats along estuaries, creeks and rivers. At low tide it shelters among mangrove roots or down crab holes.” Worrell (1963) stated it occurs, “…in mangrove roots in colonies. I have seen them crawling along mudbanks at low tide…not often found in fresh water unless tidal.”  In New Guinea, Parker (1982) found them 250 m from the high tide mark; and O’Shea (1986) found them on a path, along a river; on mud flats; and in crab burrows, including, a crab burrow on a village street. Like other homalopsid species that use mudflats this species can use sidewinding locomotion (Cogger and Lindner, 1974).
Diet and Feeding Behavior
            Ever since Günther (1864) recognized that it feeds on crustaceans, most authors writing about this snake have discussed its unusual carcinophagus diet. And, several authors report their own observations. In Australia, Gow (1989) wrote, “It feeds upon small crustaceans and when hunting forages amongst dense mangrove roots, broken pools and channels. The author has recorded it feeding on fiddler crabs (genus Uca) and a mud lobster Thalassina anomala.” And, Shine (1991a) dissected 75 specimens and found 60 crabs, many were Uca, he also found one that contained a mud lobster, and two unidentified shrimp. In New Guinea, Parker (1982) reported small red and black mud crabs as food. McDowell (in Parker) reported nematodes in almost every stomach, and consistently found crabs, with the exception of one orthopteran insect in one specimen. Voris and Murphy (2002) reported the following crab taxa from Fordonia: Sarmiatium germaini, Macrophthalmus sp., and a sesarmine crab (Grapsidae) as well as Dotillopsis brevitarsis and Uca sp. (Octpodidae). Crab remains from five stomachs suggest these snakes use relatively small prey, 0.5 – 7.4% of the predator’s mass (Voris and Murphy, 2002). One specimen (MAGNT R.5270) a 44 cm SVL male was found carrying a 29.2 g mud lobster (Thalassina anamala) (carapace length 44.9 mm, total length 132 mm) in its mouth. Nobbs and Blamires (2004) describe diurnal feeding, and observed F. leucobalia coiling around a crab (Uca flammula), apparently in an attempt to secure it. They also observed two instances of F. leucobalia swallowing a mud lobster tail and separating the tail from the cephalothorax and report hearing audible crunching. In one of these instances one snake attempted to steal a mud lobster from another snake. Separation of mud lobster tails clearly does not occur all the time since this author has removed whole Thalassina from snake digestive systems.
            Glauert (1950) and Worrell (1963) mention frogs in its diet, this seems highly improbable, although in Southeast Asia the crab-eating frog Rana[=Fejervarya] cancrivora inhabits mangroves and uses mud lobster mounds, thus the opportunity is present, and it is not inconceivable that a frog eating a crab or a crab eating a frog could be encountered by a snake and ingested accidentally. Similarly, Hoesel (1959), Worrell (1963) and Campden-Main (1970) included fish in this snake’s diet, prey that seems highly improbable, but accidental ingestion is always a possibility.
            Several aspects of prey handling in Fordonia have gained attention because they seem to deviate from typical snake behavior. Hoesel (1959) wrote, “It is exciting to observe a Fordonia catching a crab. In a flash the crab is constricted and the snake waits in this position till the victim has quieted down by the influence of its poison.” Shine and Schwaner (1985) described Fordoniapressing crabs into the mud and holding crabs in a coil while removing the animal’s legs. And, O’Shea (1996) wrote, “Crabs are pinned in their burrows and ‘chewed’ until they are dismembered.” Shine (1985) and Voris and Murphy (2002) report them pinning crabs into the mud with their chin, and using the crab’s own defense behavior to immobilize the crustacean; swallowing small crabs with the strike, and Voris and Murphy (2002) suggested that the crabs may autonomize their own legs rather than having the snake “chew” them off.
            Savitzky (1983) proposed that Fordonia crushes its prey using “hypertrophied cranial kinesis.” Crushed crabs in snake stomachs, which would support this view, have not been found by me. However, the fangs of this species are particularly robust and the heavy musculature of the skull may be used to apply enough pressure to the crab’s exoskeleton via the fangs so that the fang can puncture the exoskeleton and deliver venom and/or digestive enzymes to the crab’s tissues. These robust fangs and associated musculature may also be used to remove the mud lobster tail from the rest of the body as reported by Nobbs and Blamires (2004).
Litter size. The range reported by Shine (1991a) of 2 – 17 (n = 15, <!–[if supportFields]> eq o(X,¯) <![endif]–><!–[if supportFields]><![endif]–>= 6.1) spans the ranges of all others reported in the other literature. Kopstein (1932; 1938) reported Indonesian females with 3 – 5 embryos (<!–[if supportFields]> eq o(X,¯) <![endif]–><!–[if supportFields]><![endif]–>= 4); Gyi reported a specimen with 13 eggs; Parker (1982) reported the range of 2 – 11. Gravid female SVL’s reported by Parker (1982) ranged from 521 – 720 mm.
Timing of reproduction was reported by McDowell (in Parker, 1982) based upon a series of snakes collected at Agats (Irian Jaya, Indonesia) in March, as well as specimens from the Western Province. Parker wrote,
He [McDowell] has reconstructed the reproductive calendar for the south coast as: eggs becoming mature June-July, when breeding takes place. Perhaps August to September through to December and January, embryonic development taking place. February and March fetuses identifiable and birth of young. April to June, eggs developing.
He found that a Queensland specimen agreed well with this timetable, but that Broome (Western Australia) specimens differed considerably, and that the Broome population had many morphological differences from the populations of southern New Guinea and Queensland.
One museum specimen examined (MAGNET R.6201), a female (425 mm SVL) gave birth to eight young on 24 November. This is the smallest gravid female reported to date.
Size of neonates was reported by Parker (1982) from specimens collected from January to March. They ranged in total length from 176 – 196mm (<!–[if supportFields]> eq o(X,¯) <![endif]–><!–[if supportFields]><![endif]–>= 188.5), and had tails that were 22 – 24mm (<!–[if supportFields]> eq o(X,¯) <![endif]–><!–[if supportFields]><![endif]–>= 23.5). Gow (1989) reported neonate size as 180 mm in length
Predators and Parasites
Lyle and Timms (1986) reported five snakes taken from the stomachs of four specimens of the nervous shark, Carcharhinus cautus, in Darwin Harbor, Northern Territory. An unnumbered MAGNT specimen of Varanus indicus, from Adelaide River Creek, NT contained a Fordonia leucobalia that was about 350 mm in total length. The lizard has an SVL length of about 40 cm. The lizard specimen is unregistered and on display at MAGNT. Guinea (in Greer, 1997) reported that the bird commonly called the jabiru (Xenorhynchus asiaticus) feeds on this snake. Loveridge (1948) reported the nematode Ortleppina longissima from the stomach of one specimen.
Populations, Abundance, and Activity
Parker (1982) wrote about the abundance of this species, “At times many hundreds of these snakes have been found in the mangroves of Daru and Bobo Islands [New Guinea]. At other times only odd individuals can be found. In some cases, these large numbers appeared to coincide with extra high tides. At those times, many of the snakes were gravid females with large eggs lacking embryonic development.” Karns et al. (2002) found no association between tides and the activity of this species at Pasar Ris, Singapore; but at this study site the tidal cycle had been modified by human alteration of the drainage system. However, they did get the impression that snake activity was higher on days with late afternoon or early evening showers.

%d bloggers like this: