Chapter 11 - Life in the water

11. Life in the Water

…the gradations between the land and the fresh-water species, and between the latter and the salt-water snakes…are, like all other herpetological features, extremely close.
Catherine Hopley, 1882

                It was sunset over Thailand’s Andaman Sea coast, and we stood in water and mud over our knees searching the mangrove for homalopsid snakes. Four of us were spread out between mud lobster mounds waiting for dark. We had partially excavated several lobster mounds earlier in the day, uncovering a few lobster tunnels and a single, small lobster. Judging by the number of mounds and the presence of large trees, this habitat was relatively undisturbed. The mound I was standing next to was about five feet high with a tree growing out of its side. The entrance to the lobster’s burrow was offset from the top of the mound and crabs, mudskippers, and peanut worms were using the surface of the mound. Crab-eating frogs (Fejervarya cancrivora) were also here, using burrow openings to ambush passing crustaceans. It seemed likely that snakes were also using the lobster’s burrows for refuge and foraging. Graduate students at the National University of Singapore researching Mud Lobsters had reported finding snakes when they excavated the mounds. And, of course, somewhere within the pile of mud and muck were one or more Mud Lobsters (Thalassina anomala), the engineers that had constructed the mound. Mud Lobsters are the earthworms of the mangrove forest, filtering the soil for nutrients and piling up the undigested material to form the mounds. Most of the lobster mounds are located on the landward edge of the mangrove forest where some of the land is washed by the sea twice a day. Still, leaf litter had accumulated in areas the sea could not reach.
                Using flashlights we walked between and around the mounds, searching the shallow pools and the exposed mud for snakes. Shortly after our search began I heard leaves rustling to my right. Expecting to find a foraging snake, I was quite surprised to see an octopus moving overland through the leaf litter. After another half-hour of searching, one of my companions found an Asian Bockadam (Cerberus rynchops), one of the first species recognized to have grooved rear fangs and the species Hermann Schlegel refused to believe venomous (see Chapter 4). The snake moved from land to water, diving into the mud slurry and surfacing 10 meters away. It was the only snake we saw that night.
                Water changes the rules. Water supports an animal’s body weight and is slow to lose or gain heat.  Salt can render water undrinkable, but some coastal and marine snakes are believed to have a salt gland to remove excess salt. Salt also increases the density of water, making animals more buoyant.  Water provides other opportunities. Warm water reduces the need for basking. Water also lowers the risk of predation from terrestrial and aerial predators. Water provides abundant food resources in the form of invertebrates, fish, and amphibians. Water also presents a variety of new hiding places such as the intertidal burrow system. Of course aquatic environments also expose their inhabitants to a new set of predators and dangers from the physical environment.
                Virtually all snakes are capable of swimming, but some species are much better at it than others. Most all major lineages of snakes have some species adapted to aquatic habitats, and five lineages (file snakes, the homalopsids, elapids, dipsidids, and the natricids) contain many species highly adapted for life in water. Snake adaptations to water include modifications to virtually all organ systems. However, all adaptations to an aquatic existence do not occur in every aquatic snake, and while some species have extreme adaptations to life in water, others appear to have very few. In the introductory quote, Catherine Hopley recognized this and its significance. Dorsally positioned nostrils and eyes allow the snake to breathe and observe without exposing the head or body to air [Figure 11–1]. When diving, valves open and close the nostrils, and the trachea opens opposite the internal nares to exclude water from the respiratory tract. The body can be laterally compressed to increase the surface for swimming, the ribs may be less bowed out, and they may be elevated to exaggerate the surface on the sides of the snake. The tail may also be flattened; it may be compressed slightly at the base, or turned into a paddle with exaggerated fin-like flaps above and below the tail vertebrae. Ventral scales are broad in land dwelling snakes so the snakes can grip surfaces; in highly aquatic snakes, these are reduced so the belly scales are similar in size to the dorsal scales [Figure 11–2]. Undoubtedly, this reduction in ventral scales aids the snake in flattening its body for increased surface area when swimming. Additionally, the snakes most adapted for an aquatic existence swallow food while submerged and give birth in the water.

INSERT FIGURE 11-1.
Figure 11–1. A dorsal view of the head of Jagor’s Mud Snake, Enhydris jagorii. Note the eyes and nostrils are located on or near the top of the head and that the mottling makes this snake’s head very cryptic.

                Snakes using water to hunt are not always strong swimmers or even well adapted for swimming. Ambush predators may simply use the water for concealment and retain much of the morphology they used on land. Other snakes hunt shorelines and search small pools without entering the water, and a few snakes hunt from branches and strike prey from a perch over the water. The fact that arboreal snakes can use aquatic resources without morphological modifications to aquatic habitats is a great reminder of the impressive plasticity we see in life.

INSERT FIGURE 11-2.
Figure 11–2. A: Ventral scales of the semi-aquatic natricid, the Plain-bellied Water Snake (Nerodia erythrogaster) are very wide, rounded scales. B: Ventral scales of the highly aquatic homalopsid, the Tentacled Snake (Erpeton tentaculatus) are very narrow ventral scales that are hexagonal and keeled (arrow).

                Not surprisingly, the largest snakes often use water. The largest living boa, the Green Anaconda (Eunectes murinus), is highly aquatic and the largest pythons have strong affinities for water. Not only does water provide buoyancy for heavy bodies, but it allows the largest snakes to conceal themselves while hunting.

File Snakes
                Bizarre may be the best single word to describe file snakes of the family Acrochordidae. The advantage to their unusual anatomy was a puzzle for a very long time. Highly aquatic, file snakes are covered with loose skin and small, rough scales that have sensory organs. They have short, stubby, prehensile tails and the eyes and nostrils are located on top of the head. Their lack of muscle tone is spectacular, and can only be fully appreciated by an attempt to hold one.  The rough, file-like surface of the skin and its loose folds aid these snakes in holding slippery, mucus-covered fish while the snake maneuvers them into position for swallowing. Their maxillary teeth are similar in size; all tend to be long, and are fluted on the sides and posterior surfaces, a condition found in many fish-eating snakes. File snakes  usually have more than 100 teeth when they are counted on all the bones that carry them. Their thick bodies, slow movements, and complete reluctance to leave the water make file snakes unusual. The Little File Snake (Acrochordus granulatus) is the most marine of the three species and has been found at sea, 15–20 km offshore. It has the highest blood volume of any snake (13% of its body, most snakes are in the 5–6% range), and the largest proportion of red blood cell of any snake (50%, most snakes are in the 25–30% range). Not surprisingly, it has the largest oxygen-carrying capacity of any snake. Underwater the file snake’s heart beat is slow but, in anticipation of taking a breath as the snake surfaces, the heartbeat increases so that more oxygen can be taken up by the blood flowing through the lungs. File snakes and other aquatic snakes can also exchange gases through their skin, increasing their dive time. Freshwater file snakes do not have the unusually high blood volume or the oxygen-carrying capacity found in the marine species.
                Three file snake species range from coastal Mumbai, India eastward to Australia and the Solomon Islands and southward into the Indonesian Archipelago. But fossil file snakes dating to the the Miocene or early Pliocene, and found in Pakistan, India, and Nepal suggest the family had a wider distribution and a larger body size than they do today. Like all marine snakes, the file snakes probably evolved first in freshwater and transitioned to saltwater later in their evolution.
                The little file snake rarely reaches one meter in length; the mostly freshwater, Australasian, Arafura File Snake (A. arafurae) may slightly exceed 2 m; and the mostly freshwater [Figure 11–3] Southeast Asian Elephant Trunk Snake (A. javanicus) may reach 2.9 m, although this size record is old and may be an error, most large specimens reach about 2 m.

INSERT FIGURE 11-3.
Figure 11–3. The Arafura File Snake (Acrochordus arafurae) photographed near Darwin Australia.

                My experience with file snakes is limited. Despite months of field work on aquatic snakes in Thailand, I have only seen the elephant trunk snake in markets. However, Grant Husband at the Northern Territory Wildlife Park took me to a stream outside Darwin in Australia’s Northern Territory to demonstrate how to catch a file snake. The method consisted of walking the edge of a stream after dark with a flashlight and looking for snake snouts to locate the position of the snake. As the snake withdraws into the murky water, you plunge your hand and arm into the water and attempt to grab the snake. File snakes are relatively slow, and they depend on the dark water and holes in the stream bank to avoid capture.
                Ecological studies on the Arafura File Snake suggest that it may not have a specific home range; instead, they wander through the murky water. Some individual file snakes stay at one location for a day or more, while others may move almost a kilometer. The freshwater Arafura file snake lives in a seasonally wet climate, and their wetland habitat expands and contracts with the rains as they follow the water. During the dry season, populations may be highly concentrated in small areas, a situation that offers opportunity for mating. Multiple males compete for females in these situations and it seems likely that sperm competition is present in this species, as it is in most snakes. With the rains, the area covered by water expands and the snakes disperse to distant locations, lowering the density of the snakes. Like all animals file snakes are greatly impacted by climatic events, and Thomas Madsen and Richard Shine found high rainfall late in the wet season resulted in abundant fish populations and fat file snakes the following year.
                Like many aquatic snakes, file snakes laterally compress their body when swimming. This produces a keel on their belly that increases the body’s surface area and makes the snake a more efficient swimmer.
                Questions about file snake relationships were largely unresolved until DNA sequencing and genetic comparisons became possible. While some authors hypothesized file snakes were related to boas and pythons, others considered them colubrids, but DNA sequences suggest that file snakes are the sister species to all of the advanced snakes, the Caenophidia, and last shared an ancestor with them about 90.7 MYA. File snakes are not the only snake lineage adapted to a variety of water salinities, the homalopsids span the range of freshwater to full sea water.

Homalopsids
                Today Lake Songkhla is a large freshwater ecosystem in southern Thailand, but 150 years ago, however, it was a large bay separated from the Gulf of Thailand by barrier islands.  The openings between the barrier islands became blocked and, as freshwater begins to drain into the basin, the salt content of the water decreased, leaving a diverse fauna. Freshwater, brackish water, and salt water fishes co-exist in the lake. The aquatic snake fauna is also interesting: Little File Snakes, Pipe Snakes, Brook’s Sea Snake, Checkered Keelbacks, and five species of homalopsids feed and reproduce here. But one species numerically dominates the Songkhla system― the Rainbow Mud Snake (Enhydris enhydris)
                We used baited funnel traps to collect snakes and we would find the occasional Asian Pipe Snake (Cylindrophis ruffus), Puff -faced Water Snake (Homalopsis bucatta), Plumbeous Mud Snake (Enhydris plumbea), or Checkered Keelback (Xenochrophis flavipunctatus) in the traps. During the 1997 field season (13 days using 51 traps), we trapped 380 Rainbow Mud Snakes, and recaptured 144. Preliminary results suggested 406 to 567 Enhydris enhydris lived on the study site. Because this snake uses edge habitat, as opposed to open water, we calculated its density at one snake per 2 meters of shoreline. In spite of the fact that several of us worked the study site day and night over four field seasons, only once did we ever actually see a live snake that was not in a trap or in a ditch being drained by fishers. Despite its abundance, the Rainbow Mud Snake is difficult to observe in the field. Muddy water is the reason.
                Homalopsid snakes like shallow, muddy water. Several other aquatic ecosystems in Thailand were sampled and we found the same pattern: 3 to 5 species of homalopsid snakes were present, but one species always dominated the assemblage, the Rainbow Mud Snake. We know the E. enhydris occasionally leaves the water because of road kills. We also know it may colonize urbanized areas by using drainages ditches and drainage pipes when they are carrying water. Large shallow lakes, like Songkhla, however, produce dense populations of this snake. Enhydris enhydris dominates Lake Kabinburi in southeastern Thailand and the Great Lake of Cambodia, Tonlé Sap. All of these ecosystems have one commonality― they are all modified by human activity.
                The Rainbow Mud Snake has a small head and a large, bulky posterior body, and it is successful in part because it feeds on small fish. Humans have over-fished the large, shallow lakes of Southeast Asia, reducing the number of large predatory fish that would feed on snakes and smaller fish. With predators reduced and food supply increased, the Rainbow Mud Snake’s populations have exploded. This is most likely the result of the phenomenon biologists call mesopredator release. When large predators are removed from an ecosystem, the medium-sized and small-sized predators increase in abundance, hence meso (=middle-sized) predator release. Humans have also increased the Rainbow Mud Snake’s habitat with flooded paddy fields, shallow roadside ditches, lotus ponds, and other landscape water features that are so desirable in Southeast Asia.
                A small, black-to-olive colored homalopsid, the Plumbeous Mud Snake (Enhydris plumbea), is more terrestrial than the Rainbow Mud Snake. It lives in the mud-root tangle, and it was studied by Harold Voris and Daryl Karns in a wet pasture in Sabah on the island of Borneo. Snakes were collected by hand around the edges of buffalo wallows by pulling up grass mats at the edge of the wallow to expose the snakes. The Plumbeous Mud Snake hunts frogs, tadpoles, and fish in the security of the tangled grass roots, away from most predators.
Looking at museum specimens of homalopsids, it became clear that the snake labeled Enhydris jagorii in most collections were, in fact, composed of three different species. However, it was not until I examined the type specimen of Hypsirhina jagorii described by Wilhelm Peters in 1863 that I knew which species was the real Jagor’s Mud Snake. Peters had given the location of the type specimen as “Siam” and most of the specimens labeled jagorii were from Thailand (formerly Siam). All three species had 21 scale rows at mid body and a similar arrangement of scales on their head, but they had distinctly different color patterns, body proportions, and ventral scale counts. The most common of the three species labeled jagorii turned out to be widespread in the Mekong River drainage. It had been previously described as a subspecies of Rainbow Mud Snake, so it had a name― subtaeniata. One of the other species labeled as jagorii did not have a name, and all of the 12 specimens lacked locality data or said “Thailand” or “Bangkok.” Harold Voris and I named this snake after our friend and colleague Tanya Cha’nard of the Thailand Natural History Museum. Despite efforts to find an extant population of this snake, we have not. All existing specimens were collected in the first part of the 20th century. Is it possible that the snake that lived in the Bangkok area may now be extinct? Unlikely. A recent photograph published in a field guide labeled Enhydris jagorii was in fact Enhydris chanardi.
We were also interested in locating the real Jagor’s Mud Snake. Prior to our fieldwork, I had seen only nine museum specimens, all of them collected between 1860 and 1980, and, again, all of them with locality data suggested it lived in the Bangkok metropolitan area. Trapping in the canals, rice fields, and roadside ditches in and around Bangkok did not produce any specimens. Our efforts went un-rewarded until the summer of 2007 when Daryl Karns and I trapped snakes at several locations in Thailand’s Central Plain. During a wetland reconnaissance trip near the city of Uttaradit, we wandered into a small village, Bung Ka Lo. Some fishers lived along the road bordering the wetland, and we talked with them with the help of our driver and translator, Pon. It became obvious they were familiar with homalopsid snakes and were willing to collect them for us for a small sum per snake. The snakes frequently became entangled in their gill nets and the fishers considered them an annoyance. Daryl, Pon, and I had worked with many Thai fishers, but these people were exceedingly friendly, and some of them had no previous contact with westerners. By the time we left the first meeting, we had consumed the better part of a bottle of homemade whiskey and made some important field contacts. Since we were working on their time, it was going to be several days before we could get snakes; the villigers were celebrating a recent election and there was a weekend festival of food and kickboxing. When we arrived Monday morning they had several metal containers and bags full of snakes, including several Jagor’s Mud Snakes. In three days of working at Bung Ka Lo, we bought 262 snakes representing five species, including 26 E. jagorii.
Of the recognized homalopsid species, one is totally aquatic and reluctant to leave the water, the Tentacled Snake (Erpeton tentaculatus). Tentacled snakes are perhaps the most specialized aquatic snakes, possibly more specialized than the true sea snakes. They have two fleshy projections on their snout; very small ventral scales, strongly keeled dorsal scales, a long, prehensile tail, and unexpectedly large eyes for a homalopsid [Figure 11–4]. Picking one up will cause the snake to make its body exceptionally rigid and stick-like. It wraps its tail around the vegetation in turbid water with submergent plants and hangs in the water column. Typical hunting posture involves turning its head and neck back toward the body in a J-shape position. Rough, keeled scales provide a surface for the growth of algae and other protists, adding to the camouflage that makes this snake nearly impossible to see. In an aquarium stocked with floating hyacinths and small Fighting Fish (Betta splendens), the snake remained motionless while the fish graze the algal growth on its body. In 1999, Harold Voris and I obtained a freshly caught Erpeton from Thai fishermen and videotaped its hunting behavior. When reviewing the tape, we found the fish often disappeared with the strike. It was filmed at 30 frames per second and the fish frequently disappeared within one frame.  INSERT FIGURE 11-4.

Figure 11–4.   Views of the Tentacled Snake (Erpeton tentaculatus) using a scanning electron microscope. A: The head showing the eyes, nostrils, and tentacles. B: A single tentacle shows the forward facing scales. C: A single scale on the tentacle with striations but no pits, hair cells, ampullary organs or projections. Courtsey of Ken C. Catania.

Observers of the Tentacled Snake have disagreed on the function of the tentacles; some hypothesized they are sensory while others claim they are used for camouflage, breaking up the outline of the head. Close examination shows that the tentacles are covered with scales and that the anterior edge of the scale is free. The tentacles are also exceptionally flaccid and retractable.
Fish have a rapid-escape response called the C-start, a response that allows the fish to move left or right depending on the direction of the moving water stimulus. The water disturbance triggers the firing of the ipsilateral neuron, which, in turn, excites lateral motor neurons, causing the fish to turn away from a predator. Kenneth Catania discovered the Tentacled Snake exploit this escape reflex for capturing fish. About 1–3 milliseconds (ms) before initiating the strike, the snake creates a feint by muscle movements that start at the head and travel down the forebody in sequence; 5–12 ms later, the fish responds with a C-start. The snake’s subtle movement startles the fish, enabling the C-start response. More remarkably, the snake then predicts what the fish will do and directs the strike (before the fish responds with the C-start) not toward the head of the fish, but where the snake anticipates the fish to be when its open mouth arrives at the end of the strike. Erpeton is exploiting the fish’s escape response (the Mauthner-mediated escape response) to its own advantage and it raises the question: is the individual snake learning this behavior? [Figure 11–5].
After two centuries of speculation about how the tentacles function, Cantania and colleagues investigated further using staining techniques that allowed them to view the branches of the trigeminal nerve present in the tentacles. These nerves lie very near the tentacles’ surface. The researchers used von Frey hairs (small filaments of nylon of varying diameters used to test detection thresholds) on the tentacles and determined they were extremely sensitive, detecting forces as low as 0.008 grams. The labial scales (scales on the edge of the mouth) were much less sensitive. They also tested responses of the trigeminal ganglion and the sensitivity of the brain’s optic tectum. The results showed a close association between the tactile sense of the tentacles and the visual input from the eyes, suggesting a high level of integration between the two. When Cantania and colleagues placed fish outside of the aquarium where the snakes could not obtain any mechanical information, they oriented toward the prey; when shown a cartoon of a moving fish they struck at it, suggesting they can rely on visual information alone to hunt. This is not surprising. Most homalopsid snakes have very small, dorsally oriented eyes. The Tentacled snake is an exception; they have large eyes that are positioned laterally.
Cantania and co-authors compare the sensory systems of the Tentacled Snake to that of the Barn Owl (Tyto alba). Both have excellent vision, but when light is low or absent they can continue to hunt effectively. The Tentacled Snake uses vibrations in the water and the Barn Owl using sound cues. Erpeton can hunt day or night; it can hunt in clear water or in water heavily laden with sediment. Given that it lives in the drainage systems that carry water and sediments from the Himalayas, this snake is highly adapted for its environment.
Mike Alfaro and colleagues suggested that the Tentacled Snake and Bocourt's Mud Snake (Enhydris bocourti) shared an ancestor about 14.1 (9.3–18.8) MYA. Both snakes share similar lowland, shallow water habitats in the Indochinese peninsula, but have very different morphology and life styles.

INSERT FIGURE 11-5.
Figure 11–5. The Tentacled Snake’s strike and the fish’s response when the fish is parallel to the snake’s jaws. (A) Schematic of snakes position and the events during the strike. Numbers 1–3 show body movements prior to striking, arrows show the direction of the feint. (B) Frames from high speed video. (C)  Hydrophone recordings of strike showing pressure change (1–3) associated with feint, Y-axis units arbitrary, numbers show events illustrated in A. Note that latency to C-start is appropriate for body feint (7 milliseconds) but not for strike (4 milliseconds). (D) Percent turns toward head during the strike. (E) The long axis of the fish intersected line segment P. Courtesy of Kenneth Catania, Vanderbilt University.

                Many homalopsids feed exclusively on fish, a few feed on fish and frogs, and some are specialist for feeding on crustaceans; many are poorly studied, however, and their diets remain subjects for future investigations. What is known about homalopsid diets is that they take small prey, prey that is almost always less than 1% of their body weight, and they eat multiple prey in a single feeding session. Warm water allows them to do this. Small prey are more abundant than large prey and small prey has a high surface-to-volume ratio, making digested rapid. Therefore, these snakes can eat virtually around-the-clock. This is in stark contrast to boas, pythons, and vipers infrequently taking a single, large prey.      
                Swallowing whole crustaceans can be a challenge. Crustaceans have an exoskeleton of chitin, a complex carbohydrate that is impossible for most vertebrates to digest. To make things even more complicated, crustaceans also often have powerful pincers and live in submerged burrows. The Crab-eating Snake (Fordonia leucobalia) has long been known to have a diet composed exclusively of crustaceans. Fordonia takes crabs and Mud Lobsters and, while it often eats small prey that can be swallowed with the strike, it will eat larger crabs, dismembering them for easy swallowing. Crab-eaters have robust fangs for cracking through the exoskeleton and, while their venom has yet to be studied, it may contain the enzyme chitinase for the digestion of the chitin.  Crab-eaters are widespread in coastal habitats from Mumbai, India to Queensland Australia, a distribution almost identical to that of the mud lobster.
                Two other homalopsids were found to eat crustaceans during our study. Gerard’s Mud Snake (Gerarda prevostiana) is a down-sized version of the Crab-eating Snake and its sister species. Its distribution is not as well known as the Crab-eater, but it appears to occur in coastal habitats from Mumbai, India to the Philippines; and it, too, seems to be associated with the Mud Lobster. Edward Taylor collected a dozen or so specimens in a Thai mangrove forest and, when I examined them to see what they were eating, they contained only the remains of crabs. Bruce Jayne and colleagues studied this snake in Singapore and were interested in seeing how it handled the crabs. Captive snakes were offered a variety of species but showed no interesting in eating. Their appetite changed when they were given crabs that had just molted their exoskeleton. The snake seized the crab with its mouth, looped its body around the prey, and pulled the crab through the loop, tearing it into pieces. To date, this is the only species of snake known to tear its food into pieces. (As mentioned, its sister species, the crab-eating snake, does appear to chew the legs off larger crabs). The advantage to ripping crabs apart is the ability to swallow large prey that could otherwise not be consumed.
                Cantor’s Mud Snake (Cantoria violacea) also proved to be feeding on crustaceans, but we did not find crabs or mud lobsters in them. A mangrove and mudflat-dwelling species, Cantor’s Mud Snake is known from relatively few specimens. At first sight its rounded head, elongated body, and banded pattern suggested it may be a true sea snake in the genus Hydrophis. But it lacks a paddle tail and the front fangs found in elapids, and molecular studies place the species in the same clade with the other crustacean-eating homalopsids. Harold Voris and I have found only small snapping shrimp (Alpheus) in their digestive systems. Snapping shrimp inhabit the intertidal burrow system, and use their enlarged claw to produce a snapping sound by compressing a gas bubble that forms in the pincer. It seems possible that this snake may locate snapping shrimp by their odor, by the vibrations of the collapsing bubble, or the light generated by the pressurized gases in collapsing bubbles.
                Thirty-seven species of homalopsids inhabit the tropic and subtropical world from the Indus River Valley of Pakistan eastward to the Philippines and southward into India, Indochina, Indonesia, New Guinea, and northern Australia. Most species occur in the Indo-Chinese peninsula, Thailand, the Malaysian Peninsula, and the adjacent Sunda Shelf islands. They are absent from the fossil record, though various studies estimate the origin of the family between 65 and 22 MYA using the DNA clock.

Natricids
                The shallow, clear water at the edge of a flooded limestone quarry in the Chicago suburbs revealed a foraging Queen Snakes (Regina septemvittata). I observed the Northern Water Snake (Nerodia sipedon) catching fish across the quarry. Just a few meters from the shoreline, in wet grassy habitat shaded by trees, an Eastern Garter Snakes (Thamnophis sirtalis) was hunting for earthworms and frogs. The small, ground-dwelling Dekay’s Snake (Storeria dekayi) were abundant beneath limestone slabs, boards, and tar paper around the quarry where they feed on slugs and worms.
                All of these North American natricid species shared an ancestor that emigrated from Eurasia some time prior to the early Miocene (~19–18 MYA). We know this because the earliest known Western Hemisphere natricid fossils date to the early Miocene, and we know they shared an ancestor because of shared anatomy and DNA.
                Michael Alfaro and Steve Arnold compared three genes from many of the North American natricid snakes and found not only that they shared a distant ancestor, but that they formed three distinct clades: a water snake clade, a garter snake clade, and a novel semi-fossorial clade. Living around the suburban quarry were members of each of the three clades. Alfaro and Arnold’s work suggested that the most basal Western Hemisphere natricid is the small, fossorial Kirtland’s Snake (Clonophis kirtlandi) [Figure11–6a]. While Kirtland’s Snake is not known from the area around the quarry, there is a population within 10 miles. Both the Kirtland’s Snake [Figure 11–6b] and its sister, Dekay’s Snake, occur in wet prairie or grassland habitats. But Kirtland’s Snake is of particular interest because it is mostly fossorial, appears to use burrows of the chimney building crayfish as a hibernation site (suggesting it hibernates underwater or at least in the water), and its isolated populations are widely scattered over Illinois, Indiana, Ohio, and bordering states.
                North American natricids are perhaps the best known serpents. Garter snakes, water snakes, and brown snakes are often abundant in urban areas with remnant habitats. Natricids have adapted to aquatic habitats numerous times and then reverted to a life on land, only to return to the water. At least a few garter snakes are completely terrestrial, but many hunt along streams or in ponds, and others are almost totally aquatic.
                Jean-Claude Rage proposed that natricids and colubrids invaded North America from Europe via a land-bridge. Since the oldest natricid fossils are from the early Miocene, the migration and colonization of the land-bridge must have occurred prior to this date. An early Eocene land-bridge connecting Europe-Greenland-North America with a subtropical climate may explain the expansion of these snakes into the Western Hemisphere. But how probable is it that snakes used a high latitude land-bridge to disperse from Europe through Greenland and into continental North America in a mild or even cold climate?
                A snake den in an abandoned railroad yard in Portage County, Wisconsin provided a valuable clue. A friend had discovered the snake den near his summer house and took several of us for a tour. The circular cement foundation was mostly covered with rusted metal; it had been used as a giant turntable to switch railroad cars from one track to another, the cement cistern was about 2.4 m deep and partially filled with water. Shade from the metal covering had insulated the layer of ice and kept it from melting, but the ice cover was completely solid and transparent. Eastern Garter Snakes (Thamnophis sirtalis) [Figure 11–6d] and Western Fox Snakes (Pantherophis vulpinus) were submerged in water beneath the ice with their posterior bodies wrapped around debris. The snakes had been in the water since the previous fall; this was early May. With a solid ice cover in place since January or earlier, the snakes had not filled their lungs with air for at least five months.
                Experiments done by Jon Castanzo comparing submerged and non-submerged garter snakes from this Wisconsin den later found submerged snakes in total darkness at 5º C. These snakes had their oxygen consumption reduced 54% and heart rate reduced 77% over experimental snakes hibernating in air under the same conditions. While hibernating underwater, the snakes were living in acidic (5.5–6.5 pH), hypoxic (dissolved oxygen was 2.9 ppm) ground water. And yet, the snakes were carrying out aerobic respiration, using oxygen that diffused through their skin from the water for five months, a remarkable ability, but one that could have been anticipated. Charles Carpenter studied garter snakes in Michigan in the early 1950’s and observed five Eastern Garter Snakes hibernating in the burrow of a crayfish. He wrote,

…many snakes hibernate completely submerged in water in underground tunnels. It seems that this should be possible provided the water is very cold.

Costanzo’s experiments and the significance of snakes being able to spend months submerged in cold water have been long overlooked.
                The ability of natricids to survive submerged under ice for months at a time is not only valuable for surviving long Midwestern winters, but it could explain how snakes dispersed from Eurasia to North America by colonizing high latitude land bridges. Even if the land bridges had a seasonally cold climate, these snakes could have made the trip. Fox snakes were also under the ice and the literature reports Blue Racers and Black Ratsnakes can be at least partially submerged during hibernation. The Massasaugua is well known for using crayfish burrows as hibernacula, so it too can most likely used submerged hibernation. Vipers and colubrids, families composed primarily of terrestrial and arboreal species, also dispersed from Eurasia to North America prior to the early Miocene. Submerged hibernation may have also have been involved in their dispersal.
                More than 100 species of natricids use aquatic environments to varying degrees, and most feed on aquatic animals. Joke Bilcke and colleagues compared feeding behavior in natricids believed to be dietary generalists and specialists. The generalists were thought to forage for food by open-mouth searching, holding their mouth open and sweeping it through the water until they made contact with prey. These species tend to strike slowly to the side, perpendicular to the long axis of the body, and do not seem to depend upon vision. Specialists were thought to hunt from ambush, be visually alerted by the presence of prey, and strike forward, parallel to the long axis of their body. However, Bilcke and colleagues also looked for a correlation between prey density and strike behavior. Two European natricids’ (Natrix tesselata and Natrix maura) were compared to the North American Banded Water Snake (Nerodia fasciata), and they found prey-capture strategies in natricid snakes were not correlated with diet but with prey density. Snakes feeding on fish in a school or concentrated in shallow water use the open mouth technique, while snakes feeding on fish in low density strike from ambush.
                Toby Hibbits and Lee Fitzgerald noted that the narrow-headed garter snake (Thamnophis rufipunctatus) [Figure 11–6c] and Harter’s water snake (Nerodia harteri) have narrow heads and an elongated snout. Both snakes live in fast-moving water. The authors hypothesized that narrow-heads are more hydrodynamic when striking underwater in a swift current. They also posited a narrow snout increases binocular vision, allowing the snake to better judge the distance of the prey. Hibbits and Fitzgerald measured heads and calculated proportions for a variety of natricid species and found the two stream-dwellers had better binocular vision and experience less drag during a forward strike. Because these two snakes are in different clades, the similarity of head shape is considered the result of convergence.

INSERT FIGURE 11-6.
Figure 11–6. Representative North American natricid snakes. A. Kirtland’s Snake (Clonophis kirtlandi). B. Dekay’s Snake (Storeria dekayi). C. Narrow-headed Garter Snake (Thamnophis rufipunctatus). D. Eastern Garter Snake (Thamnophis sirtalis) feeding on a Leopard Frog (Lithobates pipiens). E. Northern Water Snake (Nerodia sipedon) feeding on a hybrid sunfish (Lepomis). F. Salt-marsh Snake (Nerodia compressicauda).

Dipsidids
                The South American snakes of the genus Helicops are ecologically and morphologically similar to the North American natricids but they are in a different lineage, the dipsidids. There are 15 species of Helicops with dorsal eyes, dorsal nostrils, and rear fangs. Compared to the North American watersnakes, the dipsidids are poorly studied but of interest because some species are live-bearing while others lay eggs, and one species (H. angulatus) appears to have populations that do both. Robson W. Àvila and colleagues examined more than 400 specimens of the spotted watersnake (Helicops leopardinus) from the Pantanal of central Brazil. The two most common prey items were knife fishes (Gymnotiformes) and treefrogs (Hylidae). This snake inhabits floating vegetation where they live in close contact with their prey and Àvila and colleagues suggest it actively forages for prey.
                Another poorly known aquatic group is the genus Hydrops; closely related to Helicops, Hydrops is widely distributed in the Amazon basin. They have small dorsal eyes and narrow ventral scales and are found in many aquatic habitats where they feed on fish including synbranchid eels. Perhaps the most interesting aquatic dipsidids, however, are the snakes in the genus Tretanorhinus, locally known as the Catívo in Cuba.  Four species are known from Ecuador, Central America, Cuba, and the Bahamas. The long head, square muzzle and long tail suggest this snake may be convergent with the Tentacled Snake (Erpeton tentaculatus) from Southeast Asia. Few herpetologists have examined these snakes since Wilfred Neil observed them in Cuba in 1949 and reported them hunting in shallow water and to be nocturnal and highly aquatic. They move very slowly and use freshwater and well as brackish, and possibly full sea water as well. Neill observed them snapping at small fish in shallow pools stating that they were quite successful in capturing their prey.

Aquatic Elapids
                American coral snakes are well known for their highly potent venom, contrasting patterns of red, black, and white or yellow rings. About 75 coral snake species are known for their leaf litter, crevice-dwelling, and burrowing habits. One species group, the triad coral snakes, has a pattern of rings arranged in groups of three, short parietal scales that are usually black, and very short tails. Triad corals are mostly Amazonian species and they are of particular interest because at least seven of the 20 species have habitat descriptions that include some mention of aquatic environments. One species, the Aquatic Coral Snake (Micrurus surinamensis), appears to spend much of its time in the water, but it has also been found above ground in vegetation. Its eyes and nostrils are located on top of the head, and its venom is particularly effective at killing knifefish. The Aquatic Coral Snake is the most basal species of the triad group, suggesting the ancestor of the triad coral snake group may, too, have been aquatic. The South American coral snakes (M. lemniscatus species complex) are also semi-aquatic or aquatic depending on the definition, and are known to feed on synbranchid eels and shallow water fish. When I collected this snake on Trinidad it was often near water, though I never had the opportunity to observe it in water.
                The kraits (genus Bungarus) of Southeast Asia also tend to be ground-dwelling, leaf litter snakes. They will hunt in streams and ponds as well as along the ocean’s shoreline, and they do not hesitate to enter the water. We have seen them in fishermen’s gill nets, and they will eat homalopsid snakes but their aquatic habitats are not well known. Shou-Hsian Mao found the Taiwan banned krait (Bungarus multicinctus) to feed on fish, and the semi-aquatic plumbeous mud snake (Enhydris plumbea). Kraits are in a different lineage of elapids than are the coral snakes and they also show aquatic tendencies.
                Much has been written about Lake Tanganyika, the African Great Lake. Perhaps it is most well known for the two British explorers, Richard Burton and John Speke, who discovered the Lake in 1858. Or perhaps its fame comes from its size, as the second deepest freshwater lake (~ 500 m), and the third largest in volume on the planet. A species swarm of more than 250 cichlid fish has made this lake famous among biologists and fish hobbyists, because the fish are a well-studied example of rapid evolution and popular in the pet trade. But there is a fish predator living here that is also of interest, Boulenger’s Water Cobra (Naja annulata).
                Arthur Loveridge, a Harvard University herpetologist, was one of the first to make this snake known. At Kasanga, rocky promontories extend into the lake, and one was built by the Germans for their Bismarckburg military base. These peninsulas are protected by a natural breakwater of jumbled rocks and it is here Loveridge observed the aquatic cobras. He wrote,

According to native reports, which my own experience confirmed in some points and contradicted in none, when the sun rises and strikes the rocks the cobras emerge from their retreats beneath them and bask for a short time on the tops of the rocks. Shortly afterward, ―and I found none on the rocks an hour and a half after sun up―they take to the water in search of fish. I was told on a calm day one might see as many as ten in the course of a morning’s fishing. We saw four in a little over three hours.

The snakes come out of the water in the evening and they are said to bask under the rocks. Loveridge suggested that this is probably correct because the evenings were cool and the snakes may absorb heat from the rocks and avoid the cool wind blowing in mid-May.
                Boulenger’s Water Cobra [Figure 11–7] is usually banded with strongly contrasting black-brown bands separated by yellow, but individuals vary and they may have a mostly grey-brown or yellow-brown dorsum.  They are quite large, reaching a size of 2.7 m, and the species is distributed across central Africa, from the Cameroons to Tanganyika, where it is often associated with flooded forest. It can dive to at least 8 m, stay submerged for 10 minutes or more, and this species may hunt in co-operative groups. When not looking for food, it uses rocks, bank holes, tree holes, and root clusters for refuge. Human encounters with water cobras are probably few, and the snake poses little danger to anyone except when it becomes entangled in fishermen’s nets or in the nets of herpetologists trying to trap it. Kate Jackson described a harrowing experience of trying to remove a large specimen from a gill net.

INSERT FIGURE 11-7.
Figure 11–7. The Water Cobra, Naja annulata from Brazzaville, in northern Congo. Photo by Kate Jackson.

                The water cobra is morphologically distinct from the common cobras and was placed in the genus Boulengerina, named after George Boulenger of the British Museum of Natural History. However, Wolfgang Wüster and colleagues’ molecular study found that the water cobra was, in fact, a part of the common cobra clade, and the researchers proposed that the aquatic cobra be placed in Naja. I follow that arrangement here. DNA suggests that the water cobra’s sister species is the Forest Cobra (Naja melanoleuca), a large species that shares a similar geographic distribution with Boulenger’s Water Cobra. The Forest Cobra shares aquatic foraging behavior as well as a diet that includes fish with the aquatic cobras.
                A second, poorly known species of water cobra inhabits the lower reaches of the Zaire River, Boulengerina christyi. A DNA sample of this snake was not included in the study done by Wüster and co-workers, and it is unclear where this species falls phylogenetically. Karl P. Schmidt erected a new genus, Limnonaja, for it in 1923 because he considered it morphologically distant from N. annulata.
                Sea kraits are widespread in the Indian and Pacific Oceans. The eight sea krait species represent an interesting evolutionary experiment in the terrestrial – marine transition. They may be viewed as a terrestrial snake’s incomplete attempt to invade the oceans, or they may be viewed as the fulfiment of a highly specialized niche, one that uses a unique microhabitat and lifestyle. Females tend to be larger than males and lay eggs in rock cavities, presumably above the high tide mark. The body size difference is also reflected in diet. Large females tend to feed in deeper water on members of the conger eel family (Congridae), while males hunt in shallow water and feed on moray eels (Family Muraenidae). Sexual dimorphism in sea krait feeding behavior may allow larger populations to exist than would otherwise be possible because the sexes are not competing for food.
                The freshwater Rennell’s Island Sea Krait (Laticauda crockeri) has already been mentioned in (Chapter 10). There is some evidence that of the other seven species some show habitat differences. Xavier Bonnet and colleagues examined microhabitats in New Caledonia and found St Girons’Sea Krait (L. saintgironsi) used a variety of terrestrial refuges including puffin burrows, tree root cavities, logs, and buildings. However, the Brown-lipped Sea Krait (Laticauda laticauda) was found only under beach rocks.
                Harvey Lillywhite and colleagues discovered sea kraits require freshwater. Prior to this study, sea snakes, including Laticauda, were thought to regulate their salt/water balance with salt glands located under the tongue. Standing on the deck of a wrecked ship off the coast of Papua, New Guinea in 1975, Lillywhite observed 100 or more emaciated-looking sea kraits stretched out on the ship’s deck. At the time, he thought it unusual that they could not find enough food. Now he has decided that hunger was not their problem. He wrote,

Although surrounded by the vast waters of the Pacific Ocean, they were most likely severely dehydrated. They might even have been early harbingers of climate change.

                Lillywhite and colleagues investigated the drinking habits and dehydration problems in three species of sea kraits. They found dehydrated snakes refused to drink sea water, but would drink freshwater or slightly brackish water. Furthermore, they found a correlation between the sea kraits’ distribution and locations where freshwater was available. A look at the distribution of all sea snakes revealed that there is greater species diversity in areas with high annual precipitation. Their need for freshwater may explain why the distribution of many sea snakes is so patchy.  Freshwater is available to sea snakes from two sources: streams and rivers draining into the ocean, and the lenses of low density freshwater that tend to stit on top of the more dense saltwater before they mix. The distribution of Laticauda coincides with low salinity surface water over most of its distribution. Sea kraits might be expected to rely on freshwater since they spend considerable time on land, but what about the true sea snakes, the hydrophiines? Lillywhite suspects they, too, may need freshwater. The most extreme sea snake is the Yellow-bellied Sea Snake (Pelamis platura) because it is pelagic. In the lab, it will drink freshwater and dehydrates rapidly when it is fasting in salt water and not obtaining metabolic water from its food. Living in the ocean is physiologically comparable to living in the desert, there is a shortage of freshwater in both habitats.
                The true sea snakes of the sub-family Hydrophiinae compose about 60 species of poorly studied marine snakes with paddle tails and front fangs. While more scientists have been examining specimens and diving with snakes to observe their behavior in the last few decades, there is still much to learn. These are the only living snakes to have successfully colonized the oceans, but, of these species, most are restricted to the waters of the continental shelf near sources of freshwater. All of the species studied to date are live-bearing and females produce relatively small litters, usually with less than 17 young and often as few as one to three young.
                Megan Kerford and colleagues studied the movements of the Bar-bellied Sea Snake (Hydrophis elegans) at Shark Bay in Western Australia. Bar-bellied Sea Snakes are specialist predators on snake eels (Family Ophichthidae) that live in burrows on open sand flats. During high tides, Tiger Sharks have access to most of the areas of Shark Bay, and Keford and co-workers discovered H. elegans move into adjacent sea grass beds at high tide. The dense grass cover provides few opportunities to forage on eels, but does provide cover for the snakes to avoid sharks.
                Photoreceptors are usually associated with the head of vertebrates, but Kenneth Zimmerman and Harold Heatwole found photoreceptors on the tail of the Olive Sea Snake (Aipysurus laevis). They observed the Olive Sea Snake’s tail was more often concealed during the day than at night, and that the tail will be pulled out of the light when it is exposed. By masking parts of the tail with tape so they could not be stimulated by light, they determined the photoreceptors were located mostly on the dorsal portion of the tail. The nature of the photoreceptors was not determined but the snake’s behavior revealed their presence.
                Only one sea snake has become truly pelagic, drifting with open ocean currents, the Yellow-bellied Sea Snake (Pelamis platura).  This is not to say that it does not occur over the continental shelf, because it occasionally gets washed up on beaches from Africa’s east coast to the coastlines of western North and Central America. The Yellow-bellied Sea Snake aggregates along slicks or drift lines. Floating debris accumulates in the slicks, and it may remain for days or weeks before a change in wind speed or current direction breaks them up. Aggregations of snakes in these drifts numbered from five to several thousands, and are composed of juvenile and adult snakes. Other animals inhabiting the drifts are jellyfish medusa, fish, porpoises, and sea turtles, with sea birds often following the lines of floating debris. Snakes aggregated here because the slicks are a useful place to locate food and mates. The degree to which these snakes have adapted to the marine environment is significant given they represent a recent evolutionary radiation.
                Kate Sanders and colleagues used mtDNA as well as nuclear genes to examine the relationships of the Australasian elapids and the sea snakes, and estimated the time sea snakes diverged from the other elapids. They found the sea kraits to be the sister to all other hydrophiines, and the Melanesian Small-eye Snake (Micropechis) the sister to the remaining species. They also recovered a clade containing the true sea snakes, the black swamp snakes (Hemiaspis), and the Tiger Snakes and their relatives (Notechis group). Which of these groups form the sister to the sea snakes was unclear. However, of interest is that tiger snakes are known to eat frogs and forage in water or along shorelines, and the black swamp snakes are semi-aquatic. Pre-adaptations to life in the water are wide spread in this clade. The time the ancestral sea snake diverged from the terrestrial or freshwater members is estimated at 6.2 MYA (7.9–4.7MYA) and the authors write that this is, “...an extremely brief interval to generate ~60 species of great ecological and morphological diversity.”

Ancient Aquatic Snakes of the Eocene
                The Eocene started 55.8 MYA and ended 33.9 MYA. During this time, the continents were drifting toward their present position. It is a period often described as a “greenhouse,” with global warming attributed to an increase in atmospheric carbon dioxide. The early Eocene is notable for species’ distributions now significantly different than those we see today. There were palm trees in Alaska and the northern Rockies, crocodiles on Ellesmer Island above the Arctic Circle, and forests covering much of Antarctica. Primates spread from Asia to Europe and through North America, and the rivers, estuaries, and oceans contained snakes, large aquatic snakes.
                Snakes in the family Palaeopheidae are known exclusively from their vertebrae and ribs, and they inhabited both hemispheres, from the Upper Cretaceous to the Eocene. Their fossils are always associated with rocks deposited in watery environments. The vertebrae tend to be tall and narrow, and the ribs are only slight curved, characteristics found in the most aquatic snakes living today, the sea snakes. Those who study Palaeopheidae fossils consider the group relatives of the boids, but others have suggested that they are close to the file snakes (the acrochordids). Size estimates for the paleophids range from 0.5 to 9 m or more and, while some lived in near-shore environments such as estuaries and mangroves, others were using open ocean habitats far from shore.
                Palaeophis colossaeus was described by Jean Claude Rage in 1983 based upon some 34 mm vertebrae collected in Mali. We don’t know how many vertebrae the snake had, but given that a typical boa or python has about 270, it is likely this snake could have been 9180 mm, or more than 30 feet. Another very large snake from this family is Pterosphenus schucherti, a species described from coastal North America from Texas to New Jersey. Remains from Florida indicate that during the late Eocene the snake died at least 300 km from the nearest mainland where it was buried with cartilaginous fish, bony fish, and an ancient whale.
                Why these ancient aquatic snakes disappeared near the end of the Eocene is uncertain, though their extinction is very likely linked to cimate change. Global cooling at the start of the Oligocene occurred as oceanic circulation was altered with Antarctica’s disconnection from Australia and South America and led to the formation of the South Pole’s continental ice sheet. The extinction of many warm water species undoubtedly followed.
                Aquatic snakes and semi-aquatic snakes occur in almost all major lineages of snakes, the exceptions appear to be the scolecophidians and the vipers. Scolecophidians seem to be specialized for burrowing and feeding on ants and termites, making aquatic adaptations less useful and unlikely to occur. Vipers have only one semi-aquatic species, the Cottonmouth. The poor representation of vipers in aquatic environments was hypothesized by Bruce Young to be due to the “…the poor hydrodynamic profile of the 'typical' viper head.” But, this would not preclude vipers from hunting from ambush in the water like an anaconda. Vipers tend to have a large bulky body, but so do anacondas and many homalopsids. As to why vipers have not invaded aquatic environments to the degree seen in other snake lineages remains largely unexplored.

Comments

Popular Posts