Gerard’s Water Snake, Gerarda prevostiana

1837 Coluber (Homalopsis) prevostianusEydoux & Gervais, Magasin de Zoologie, 3:5, pl 15. Holotype: Location unknown. Type locality: “Manila.”  Collector: unknown.

1849 Gerarda bicolor Gray, Catalogue of Specimens of Snakes in the Collection of the British Museum, p. 77.  Type locality: “West Indies,” from Mr. Launa’s collection.  This type locality is in error. Holotype: BMNH 1946.1.2.33.

1854 Campylodon prevostianum – Duméril, Bibron and Duméril, Erpètologie Gènèrale…Reptiles, 7:946.

1884 Helipophis flavescens Müller, Verhandlungen der Naturforschenden Gesellschaft, 7:286. Type locality: “East Indies.”  Holotype: NMB 1771. Collectors: G. and F. Müller. 

1890 Gerardia prevostiana – Boulenger, Fauna of British India…Reptilia and Batrachia, p. 379.

1930 Gerarda prevostiana – Smith, Bulletin of the Raffles Museum, 3:62.
            Wall (1921) stated that this species was named in honor of Monsieur Prevost, Superintendent of the Paris Museum at the time specimen was named.
            Gerarda prevostiana is distributed in coastal areas from the vicinity of Bombay, India westward to the Philippines, this distribution may be discontinuous. While it seems probable that it is present in Indonesia I have seen no specimens or literature references to specific specimens that suggest its presence there. It also seems absent from tropical coastal China. Manthey and Grossmann (1997) have a map that suggests it is restricted to the west coast of Myanmar, Thailand, and the Malayan peninsula; however Taylor’s (1965) specimens (a series of 15) were collected in Chon Buri Province along the coast of the Gulf of Thailand.  Iskandar and Colijn (2001) consider it present in Indonesia but list no specific localities or specimens and they list the Philippine type locality as an error (probably following Gyi’s comments). However, Auth et al. (1990) verified that it is indeed present in the Philippines. The verification of its presence in the Philippines took 153 years (Murphy and Voris, 2002).  The distribution may be disjunct but this snake should be expected wherever mangrove forest is found. Due to its secretive habitats it has been considered rare by many authors, but because of its habits it cannot be easily deleted from a faunal list from any area with mangrove forest between Bombay and the Philippines. However, the apparent total lack of specimens from Indonesia is puzzling.  Its distribution overlaps with several other widespread coastal homalopsids, including its sister species Fordonia leucobalia and members of the genus Cerberus, except that Gerarda is absent from the Sahul Shelf (Australia and New Guinea).
            This is a small homalopsid with a dorsum that is uniform gray-brown or black. The nasal scales are separated by an internasal scale, and it has 17 rows of dorsal scales at midbody, the lowest count of any homalopsid. Its scales lack keels or striations, but may be pitted with irregular erosions of the scale surface. Its close relative, Fordonia leucobalia has 23 – 27 rows of dorsal scales at midbody, but it also has smooth scales and the nasal scales separated by the internasal. Cantoria violacea and Cantoria annulataalso have smooth dorsal scales (C. annulata females and juveniles have weakly keeled scales on rows 1 – 4 in the area of the vent) and the internasal separating the nasals. However these two snakes have 21 and 19 rows of scales respectively at midbody. Heurina has its nasal scales separated by the internasal and 27 scale rows at midbody; but it has a completely divided nasal scale. And, Bitia has its nasal scales separated by the internasal but it has 37 – 43 scale rows at midbody that are triangular in shape and have exposed skin between them.

            Boulenger (1896) gave its maximum length as 520 mm. Twelve specimens were measured for this study; the largest was a male with a total length of 514 mm, and a 65 mm tail. The largest female had a total length of 472 mm with a 57 mm tail. The smallest specimen examined had a total length of 230 mm with a 29 mm tail. An examination of 12 museum specimens from the eastern portion of the range (India to Thailand, five males and seven females) suggest that the tail/SVL ratio tends to be higher in males (14.4 – 17.6), than in females (13.7 – 15.8). This was also supported by data collected on the Singapore population by Karns et al. (2002).
External Morphology
            The head is slightly distinct from the neck and slightly depressed. The body is cylindrical to slightly depressed. The eyes are dorsolateral, but more lateral than dorsal.
            On the head the rostral is a constricted rectangle, about twice as broad as high, and is barely visible from above. The nasals are entire and separated completely by a large internasal that is equal or greater than the nasals. The internasal is not in contact with the loreal. The prefrontals are about equal to the nasals in size. The frontal is relatively short, its length being less than the interorbital distance, and the eye-nostril distance. The parietals are long, about 1.2 times the length of the frontal. The loreal is about as long as high, and in contact with the first three labials. The supraocular is close to being rectangular. The preocular is single and taller than long. There are two postocular scales (rarely three), the upper one is smaller, and the lower scale partially underlies the orbit. There are no subocular scales. The temporal formula can be 2 + 2 or 1 + 3. The upper labials number seven or eight, the fourth labial enters the orbit. The head shields on these snakes are imbricate and are frequently pitted or eroded.
            On the chin the lower labials number 7 – 9, the first four (rarely five) are in contact with the anterior pair of chin shields. Gulars number three to five, usually five.
            On the body the dorsal scale rows on the neck are in 17 – 19 rows; Thailand specimens have 17 – 19 rows on the neck, while those from Singapore and a single specimen from Sri Lanka had 17 rows on the neck.  The scales at mid-body are in 17 rows and the dorsal scales near the vent are in 13 – 15 rows.  Like the head scales the scales in the dorsal rows are pitted but there are no keels or striations.  The first row is ovate; they become more lanceolate toward the midline. The dorsal scales at posterior body are ovate and do not elongate toward the midline. The ventral scale count range was 141 – 157 and they are about three times wider than the height of a nearby dorsal scale. Males have 144 – 157 ventral scales, and females have 141 – 144. There is little or no indication of geographic variation in the ventral or subcaudal scales of this species from our sample which combines specimens from a variety of localities. The anal plate is divided, occasionally into three parts. It is about twice as long as the preceding ventral.
            On the tail the dorsal scales are similar to the ones above. The subcaudal scales are divided and number 30 – 35; males have 31 – 35; females have 30 – 32. At the base of the tail the width is 75% of the height based upon an average of five specimens.
            Color and Pattern. The crown of the head is uniform gray. Each of the upper labials is yellow or has a large yellow spot. The rostral is gray. The chin is mostly yellow with some scales having darker edges. The first three scale rows are mostly yellow, the rest are gray. Each ventral scale has diffuse chromatophores concentrated near the anterior edge of the scale. The ventral surface of the tail has pigmentation similar to the ventral scales.
            Karyotype.  Sing et al. (1970) report a 2n number of 36.
            This is a snake of the mangrove forest, although it may occasionally stray into adjacent environments. It is restricted to coastal areas and may be found in rivers that empty into the ocean, on mud flats, and along rocky shorelines, but it most likely reaches its greatest abundance in mangroves. Karns et al. (2002) found G.  prevostiana this species used a variety of microhabitats from the landward edge of the mangal to the mud pools in the lower tidal zone at Pasir Ris Park in Singapore. It has been excavated from mud lobster (Thalassinia anomala and T. gracilis) mounds (Karns et al., 2002; Voris and Murphy, 2002) which occur toward the landward edge of the mangal. However, Karns et al. (2002) found that it was not restricted to the area around mud lobster mounds. This species was found to be active throughout the night and more active at spring tides than at other times during the tidal cycle.
            The literature comments on this species all support its coastal distribution and its aquatic, intertidal habits. Gharpurey (1944) found it to be, “…entirely aquatic, haunting tidal rivers and estuaries, and often wanders out along the coasts.” Tweedie (1983) wrote, “It lives along sea coasts and in tidal rivers.” Wall (1905) stated, “It was lying sunning itself in a shallow pool of water, and made no movement when I walked over it snipe shooting.” He (Wall, 1921) later wrote, “It is entirely aquatic, haunting tidal rivers and estuaries, and straying out along the coasts. It is sometimes left on the shore by the receding tide, and I encountered one so stranded in Cannanore. It was lying in a shallow pool of water, partly submerged. Nicholson records one which had wandered into a dusty street in Rangoon.”  Mangrove habitats were also reported by Lim Boo Lait (1963) for Selangor, Malaysia; and by Taylor (1965) for the Thailand population.
Diet and Feeding Behavior
            Literature comments about the food of this species include Wall (1921) who wrote, “No observations have been made, but it is fairly certain to subsist upon fishes.” And, Cox (1991) stated, “Fish, shrimp, and other aquatic animals comprise the bulk of the diet.”
            Contrary to the literature none of the five museum specimens examined with stomach contents contained fish, they all contained crab remains (Voris and Murphy, 2002). Jayne et al. (2002) later found that it specializes in feeding on recently molted crabs, and that the snake will tear larger crabs apart in order to ingest them. This is the only species of snake known to tear its food into pieces that can be swallowed, all other snakes consume their prey whole (Fordoniais also a possible exception but in need of documentation). Tearing the prey into pieces is accomplished by a “loop and pull” behavior that involves the snake forming a loop of its body around the prey and pulling the prey with its mouth through the loop. The advantage of this is that it allows the snake to swallow a much larger prey, than it would otherwise be able to handle. G. prevostiana is known to feed on the crabs Dotillopsis sp. and Episesarma versicolor (Voris and Murphy, 2002; Jayne et al., 2002).
            Little is known about the reproduction of this snake, Gyi (1970) reported a specimen to contain five eggs.
Populations and Abundance
            Wall (1921) wrote, “Seemingly a very rare snake in Ceylon. It is also a very uncommon snake round the coasts of India, judging from the dearth of specimens in Museums…it is perhaps singular that it has not been recorded from the eastern coasts of India.” De Silva (1994) considered this snake extinct in Sri Lanka because according to him no specimens had been collected since 1888. However, Porej (2001) reported one from Muthurajawela Marsh (Sri Lanka). A clue to the habitat used by this snake was provided by Taylor (1965) when he reported finding them at night in a mangrove forest. This snake appears rare because it uses mud lobster mounds and other mangrove crustacean burrows as hiding places, a microhabitat that has been over-looked by herpetologists. Which is probably why Gharpurey (1944) wrote, “It is found all along the western coast of India, but not in large numbers.” Karns et al. (2002) found this species by nocturnal visits to mangrove forests. During two spring tide survey nights (full and new moon) they collected 16 snakes compared to eight neap tide (half moon) survey nights that produced only 12 snakes. This species made up about 15.9% of the snakes collected.