Tuesday, October 28, 2014

How to teach yourself about an obscure snake

This article will soon become available in Spanish

The world is full of obscure snakes. According to Darren Naish at Tetrapod Zoology, the more you know about them, the better a person you are. Writing this blog, and in my research, I am often confronted with the challenging task of finding out something - anything at all - about a species of snake that I've never heard of before. This post is a walk-through of the process that I usually use to track down even the most basic information about obscure snakes, although it could be used as an example of how to find trustworthy information about any species of plant or animal. I'll use as an example the species Liophidium mayottensis (Peters's Bright Snake) - a lamprophiid colubroid found on the island of Mayotte. If you're like me then you're filled with questions right away: Who was Peters? What is so bright about this snake? Where's Mayotte?

Wikipedia page for Liophidium mayottensis
as of October 2014
I needed to know about this snake as part of a project I'm doing where we compare endangered species of reptiles with those that aren't to try and figure out if there are traits or features that the endangered species have in common (and the same for invasive species and other special groups). This kind of thing has been done for birds and fish, but not really for reptiles. It's a much larger effort than just me, and my part in it is small, usually tracking down basic information about the reptiles so that we can build a database of reptile life history traits. I'm talking about things like size, sexual dimorphism, whether they lay eggs or give birth to live young, how many eggs or young they have at a time and how often, where and in what kind of habitats they live, what they eat, that kind of thing. Sounds simple, right? We'll just go to Wikipedia...well, as of 2014 that wasn't very helpful.

When faced with a species about which I know almost nothing - in this case a species I had never even heard of before - there are a couple of resources that I generally go to first in order to figure out how I should proceed. The first is always The Reptile Database. This wealth of information is curated by Peter Uetz, Jakob Hallermann, and Jiri Hosek, three individuals to whom the whole of the herpetological world is indebted. Using the advanced search feature, you can look up any species of living reptile using its common or scientific name, including by an old scientific name (a "synonym") that is no longer used. This is important because scientific names change all the time, and sometimes the same species has gone by 10 or 20 different names over the course of its taxonomic lifetime. It is particularly important to know about these names because the species may have gone by them for a long time in older literature, which is sometimes the most important literature there is.

Liophidium mayottensis
Before searching TRD, I sometimes try to use the scientific name itself to figure out a little bit about what I'm looking for. It helps to know some Latin and Greek, and a handy reference that I use a lot is Borror's Dictionary of Word Roots and Combining Forms. In this case, the genus name Liophidium told me that this was a snake with smooth scales (the Greek prefix lio- meaning smooth + the Greek root ophid meaning snake + the Latin suffix -ium normally used to form abstract nouns). The specific epithet mayottensis means "from Mayotte" (the -ensis suffix is a common way to form an adjective indicating spatial or geographic origin in Latin, similar to the English suffix -ese, as in Maltese, Chinese, or Portuguese). Although the Latin and Greek origins of the scientific name can be helpful, they can also be misleading (for example, the North American Racer is called Coluber constrictor even though it is not a constrictor) or unhelpful (another familiar North American snake, Storeria dekayi, is named after two 19th century herpetologists, David H. Storer and James E. DeKay), so don't rely too much on these.

The Madagascan Biogeographic Realm
Mayotte is the southeasternmost island in the
Comoros chain, although politically it is part of France.
Many interesting snakes inhabit this realm, including bolyeriids
When searching TRD, I always put the full binomial I'm looking for into the 'Synonym' field of TRD's Advanced Search, because the 'Genus' and 'Species epithet' fields only search the current names, and who knows what name it goes by now. Barring any misspellings, at least one record usually turns up, sometimes more if the name I've used has been split into multiple species. In this case, it's just one, and it matches the name I used. From this record, I can find out the currently accepted higher taxonomy of my species. In addition to being a snake (which I already knew), I can see that it's in the recently-erected family Lamprophiidae, a group of snakes found mostly in Africa. Furthermore, I can see that Liophidium mayottensis is in the subfamily Pseudoxyrhophiinae, a group of snakes found almost exclusively in Madagascar. Because Mayotte is an island in the Comoro Island chain, lying just northwest of Madagascar, this subfamilial designation makes sense - we think that lamprophiids colonized Madagascar, Socotra, and the Comoros from Africa about 30 million years ago, one of several radiations of snakes onto these islands. However, in this case knowing the subfamily doesn't help us much in our search for natural history information. Unlike certain instantly-recognizable groups of snakes such as pareatids or xenodermatids, pseudoxyrhophiines are diverse, including almost 90 species with a wide variety of lifestyles. I've written about the genus Langaha, which belongs to this group, before.

The BHL is also a great source of artwork
in the form of old plates, like this mudsnake
from Duméril's Erpétologie Générale,
which adorns the logo of this blog
In order to go further we need look at the rest of the TRD record. Since we're looking for a description of the species, one of the most helpful pieces of information is the location of the original description in the scientific literature. You'll find the name of the person who originally described the species and the year they did it in the TRD record, right next to the scientific name. This is called the authority, and it's presented in parentheses if the name that person used has subsequently been changed. For instance, 11 of the 139 reptile species described by Linnaeus, the father of modern taxonomy, still retain the original names he gave them. You can tell because these are the ones without parentheses. If you look to the bottom of the record, you'll find a citation for the book or article in which that first description resides, along with other literature pertinent to the species. This literature is usually focused on taxonomic changes, although sometimes more general ecology or natural history literature is included as well. Following up on this literature is easier in some cases than others. One thing TRD has done to make it simple is provide links to the full-text if it's available for free online somewhere. A lot of older literature is becoming available through the Biodiversity Heritage Library, a partnership of libraries that have digitized what they call the "legacy literature" of biodiversity.

Wilhlem C.H. Peters
Our species was originally described in 1874 by Wilhelm Carl Hartwig Peters, a German naturalist and explorer, which explains the first part of the common name Peters's Bright Snake (which was probably not applied until much later, since it's considered presumptuous to name a species for oneself). Peters called it Ablabes (Enicognathusrhodogaster var. mayottensis, a confusing mess if there ever was one. His description was published in the journal Monatsberichte der Königlichen Preussische Akademie des Wissenschaften zu Berlin (which is obfuscatingly abbreviated Monatsber. Königl. Akad. Wiss. Berlin.), which roughly translates to 'Monthly Reports of the Royal Prussian Academy of Sciences in Berlin'. Not exactly the most widely read journal, even if it has existed in one form or another since 1700 and is still around today. Anyway, it's in German, so it'll prove difficult for us to read Peters's original description even if we can find it (which thanks to the BHL, we can). There's also the little problem of whether the issue it's in was published in 1873 or 1874, because the citation in TRD lists both, but fortunately we can check both quickly since the page numbers are also given (it's '73). The article starts on this page and the description is on this one. These days descriptions of new species usually get their own stand-alone articles, but back then it was common practice to shoehorn them into checklists, expedition reports, and other types of articles. There's a description of a new chameleon in the same article. In a way, it's one explanation for the prolific output of Peters, who described 122 new genera and 649 new species of amphibians and reptiles in his lifetime, 281 of which are still recognized today (only four people, all his contemporaries, have described more). The high attrition is partly because many species were inadvertently described more than once. The guys at TRD have done a fabulous job keeping track of all this confusing literature, and I cannot commend them highly enough for their efforts.

Another difference between the 1800s and now is that species descriptions today are generally much more complete. You might be surprised to learn that the International Commission on Zoological Nomenclature, which advises, arbitrates, and recommends rules for the zoological community on describing new species of animals, stipulates only that in order for a species description to count as official, it must include at a bare minimum just "a description or definition that states in words characters that are purported to differentiate the taxon", and even this 'strict' definition applies only to names published after 1930. Peters's description of Liophidium mayottensis (translated) reads:
17. Ablabes (Enicognathus) rhodogaster Schlegel var. mayottensis: 
Two young specimens from Mayotte seem to me to belong to the above species, although they do not have red coloration on the belly. Frontal a little longer than high; 8 supralabials, of which the 4th and 5th touch the eye; temporals 1+2+2; infralabials 9, the first pair of which is in contact behind the tapered mental; two pairs of chin shields. Body scales smooth, without apical pits, in 19 longitudinal rows. Ventrals 190, divided anal, subcaudals 99 pairs. Above olive-brown, a little darker along the middle and fourth-to-last row of scales. From the snout through the eye and the frenal region there is a black napkin which is indistinct on the side of the neck and disappears in the penultimate row of scales on the side of the body. Under this there is a bright yellow band, which goes to the mouth. There are three black spots on the rostral and upper lip. The chin and infralabials are spotted or marbled with black and yellow. On the neck are fine yellowish transverse lines. Ventral scales with 4-6 black dots; posterior ventral scales and subcaudals yellowish-white.
Liophidium rhodogaster
Gold-collared Snake
So we've got counts of the scales and descriptions of their position relative to one another, which is considerably more than it took back in the 1870s to name a new species. No drawings, no information on size, habitat, reproduction, nothing. It's forgivable when you know that Peters, by then a museum curator, was merely reporting on a collection of amphibians and reptiles he had been sent from Madagascar and nearby islands by two guys named Pollen and van Dam. Peters thought the snakes they had collected on Mayotte were a variant of a species that had already been described, the Gold-collared Snake of Madagascar, known then as Enicognathus rhodogaster and today as Liophidium rhodogaster (rhodogaster meaning 'red belly'). We learn from TRD that twenty years later Belgian-British zoologist George Boulenger (writing, mercifully, in English) elevated it to its own species and changed the genus so that it was known as Polyodontophis mayottensis. Boulenger is even lighter on details than Peters, saying only that it is very similar to rhodogaster but differs in that it has one more pair of dorsal scale rows, about 11 more ventrals, and about 15 more subcaudals, and that its neck pattern includes the same yellow lines mentioned by Peters. Since he's not trying to describe a new species, it's OK, but it's frustrating since we're looking for more detail about the animal's ecology and natural history.

It's likely that neither Peters nor Boulenger ever saw Liophidium mayottensis, or many of the other species they described, alive, so we can forgive them for not mentioning its habitat or patterns of activity (although they could have at least measured the specimen). Sometimes museum specimens yield information about diet (via stomach contents) or reproduction (via eggs or embryos in utero), but this does not seem to be the case for Peters's Bright Snake. To learn about these things, we'll have to sleuth out some other papers. The other two listed at TRD don't look too promising - one is a biography of Peters that's only available in print, and the other focuses on a different genus, Sibynophis, that's superficially similar to Liophidium but distantly related.

We can do a little better by checking some other common sources of information on the web. We already know that Wikipedia's useless (although the links at the bottom of some pages can be quite useful), but a general Google search for the scientific name typically turns up links to the pages for a species on several authoritative sites that aggregate biodiversity information online. In no particular order, I often check the University of Michigan Museum of Zoology's Animal Diversity Web. This is a great student-authored resource but it's still incomplete, and it doesn't even have a page for our genus yet (but check out their detailed pages on all three Acrochordus species). Other similar sites include the Encyclopedia of Life (species page incomplete for L. mayottensis, but check out Laticauda colubrina for a fairly good page), DiscoverLife (which is mostly links with little original content, and is unhelpful for our species, although they host a cool ID guide for North American snakes), and Map of Life (which has lots of cool mapping capabilities but not for our species). Citizen science projects can be a rich source of information on distribution, but such projects are in their infancy for herps. Two of the best are iNaturalist and HerpMapper, neither of which has any data on our species. Remember that none of these sources are peer-reviewed, so they may propagate misinformation (although I have found this to be rare).

One of Pagale Bacha's Flickr photos of L. mayottensis
ARKive is a film and image archive that generally has pictures of rare species when most other websites fail, and that is the case here, but as of 2014 it contained no additional information (contrast with their excellent accounts for snakes like Natrix natrix and Macroprotodon cucullatus). Flickr can be a good source of images too, in this case providing us with four additional images, all taken by the same person of the same individual snake. I have noticed that a culture of accurate species identification exists on Flickr that isn't found elsewhere on the Internet. For instance, don't ever trust Google Images when searching for rare species - in this case, only one of the hundreds of images returned is actually of our snake. Earlier I mentioned the Biodiversity Heritage Library, one of the most consistently useful resources on the web, and their search feature leads us to one new resource: a mention in a paper by John Cadle from 1999, focusing on morphological taxonomy of  Malagasy snakes (which states that Liophidium are diurnal and led me to a paper describing the smooth, hinged, spatula-shaped teeth of Liophidium and other snakes, an adaptation for grasping and swallowing hard-bodied prey, such as skinks their teeth fold backwards when forces are applied to their leading surface, but lock into an erect position if forces come from behind).

Some L. mayottensis DNA. It looks just like the DNA
of any other species, although there's a lot it can tell us.
Two other online databases are more authoritative than those previously mentioned, in that they are reviewed by experts. One is GenBank, the NIH genetic sequence database. A GenBank search reveals that five genes have been sequenced from L. mayottensis, which is more than for most reptilesThese include four mitochondrial genes (ND4CO1, and cyt-b, which are essential to the electron transport chain of cellular respiration, and 16S, part of the protein synthesis machinery of ribosomes) and one nuclear gene (c-mos, which plays a role in mitosis). These genes were chosen for their conserved functions and relatively slow rates of evolution, which makes them useful for phylogenetic purposes (except for CO1, which evolves at just the right rate for DNA barcoding, a technique which is used, among other things, to monitor trade of reptiles without specialized expert knowledge). A phylogenetic analysis was done to determine the relationship of Liophidium pattoni, a new species discovered in Madagascar in 2009, to the other species in the genus. The results placed L. pattoni as sister to L. rhodogaster, and L. mayottensis as sister to two other Malagasy species, L. torquatum and L. chabaudi. This may seem like a dry, mundane detail, but it actually tells us something very interesting about our species: it probably colonized Mayotte from Madagascar after the ancestors of Liophidium had already radiated there. It also says that Peters, who thought that L. mayottensis was a subspecies of L. rhodogaster, was way off - it's actually more closely related to almost any other member of the genus (although to be fair to Peters, none of those other members had been described yet when he named L. mayottensis — and morphology might lead you to believe that L. mayottensis was the most basal member of the group, since it has 19 dorsal scale rows whereas every other species has 17).

Liophidium pattoni and its relationship to some of its closest relatives, including L. mayottensis
From Vieites et al. 2010
IUCN categories
The other more authoritative online database is the IUCN Red List. The Red List assesses the conservation status of species and often includes a distribution map (although not in this case), some ecological information, and a short bibliography focused on ecology and conservation rather than on taxonomy. The IUCN page contains several useful nuggets, most of which come to us by way of expert knowledge and may or may not be published elsewhere. For instance, we learn that our species is classified as Endangered under the IUCN categories, which are based on quite rigorous and quantitative criteria. Peters's Bright Snake qualifies as Endangered despite very limited data because all known records are from a forested area of about 65 km2 in the center of Mayotte, which is subject to a continuing decline in quality (criterion B2b(iii)) and within which the actual occurrence records of the snake suggest that its populations are severely fragmented (criterion B2a). Even if the area of occupancy is underestimated, the entire terrestrial area of Mayotte is only 365 km2, which is still less than the minimum of 500 km2 that a species must exceed unless both it and its habitat are known to be contiguous and stable.

Hinged teeth of Liophidium rhodogaster
From Savitzky 1981
The IUCN record also lists several other pieces of information. It tells that the known records are all between 144 and 653 meters above sea level. It states that "this snake is diurnal, ground-dwelling and very secretive", "observed in natural forests and plantations", and is egg-laying. This last tidbit is pretty helpful, and it's no surprise that we haven't encountered it before - it's from a field guide written in French by Danny Meirte, covering the terrestrial fauna of the Comoros, published in 1999 and updated in 2004. As for conservation, it says that our species is not used by humans for any known purpose, but that an introduced civet may be a threat. All native reptile species on Mayotte are protected by law, and several nature reserves may benefit L. mayottensis, but no data is available on the snake's occurrence at these sites.

Finally, the IUCN record notes that "the extreme scarcity of observations may be attributed to the cryptic habits of this snake, but also suggests that L. mayottensis is not common". No shock there. The short bibliography includes both the old and new editions of the field guide and a paper by Oliver Hawlitschek in the journal ZooKeys that used field surveys and remotely sensed data to assess the conservation status of Comoran reptiles, upon which most of the conservation assessment is based. The profile also cites another work in preparation by Hawlitschek, who was also an expert reviewer for the species and took the Arkive photograph. I visited his website and was able to learn that he is a German PhD student studying herp conservation & phylogeography in the Comoros.

Phylogenetic tree of Malagasy reptiles based on CO1 DNA barcodes
Liophidium is near the top right
From Nagy et al. 2012
Now we're getting somewhere, although we're still looking for body size and clutch size, two of the most basic species attributes. Usually, after checking all off the above sources, I repeat the whole process on Google Scholar and track down any promising articles. Often , I'll add a search term for the particular attribute I'm looking for (e.g., "clutch size", "svl") to see if that helps. In this case, even Google Scholar didn't turn up much specific to our species. I was about to give up when I decided to contact Oliver Hawlitschek. When I went to look up his email address, I noticed that he recently published a paper in the journal PLoS ONE, which of all places is known for its free and open accessibility to all. The paper, titled "Island Evolution and Systematic Revision of Comoran Snakes: Why and When Subspecies Still Make Sense", includes supplementary material that finally gives us the answer to our seemingly simple question of "how long is Liophidium mayottensis"? The average adult total length is about 80 cm for both sexes, maximum 1 meter  (3 feet), with the tail making up about 30% of the body. When I contacted Oliver he confirmed this, and he also told me that as far as he knew no information on clutch size was available (although he expected it would be small, like that of most other island snakes). From reading his paper, I also learned that this is by far the largest species of Liophidium (the next is L. therezieni at 72.6 cm) and the only one with 19 dorsal scale rows instead of 17. Oliver's paper suggested that Comoran Liophidium (and the snake Lycodryas and lizard Oplurus) are larger than their Malagasy congeners because they are released from competition with larger species that do not occur in the Comoros.

Liophidium mayottensis skull (with tooth closeup, inset)
Image by Cynthia Wang
Oliver also put me in touch with Cynthia Wang, another graduate student who is using high-resolution X-ray computed tomography to make 3-D scans of the skulls of snakes. Turns out she recently scanned a L. mayottensis skull. You can see the spatula-shaped, hinged teeth characteristic of the genus, although the connective tissue is missing. He also told me that he will be returning to the Comoros this November, and that L. mayottensis will be his #1 target while he's there. All in all, a pretty satisfying conclusion.

This was a long article; congratulations if you made it to the end! I justified the length partly in celebration of my birthday this month and partly in celebration of this blog reaching 250,000 views! I hesitated writing this article because I base a lot of my articles around obscure snakes and I was afraid that writing a how-to would amount to writing myself out of a lot of subject matter. On the other hand, I suppose I enjoy the chase, and I think this overly-long article's length goes to show just how much actually is out there, even for really obscure species, if you're willing to look (and there are certainly resources I've missed! Let me know about them in the comments.). I also think that this process is easily generalizable to non-reptiles - there are some great resources out there for amphibians, birds, algae, echinoderms, insects, and much else. Whatever you're interested in, happy researching!


Thanks to Oliver Hawlitschek, Cynthia WangHenry Cook, and Pagale Bacha for the use of their images.


Bauer, A. M., R. Günther, and M. Klipfel. 1995. The Herpetological Contributions of Wilhem CH Peters (1815-1883). SSAR Facsimile Reprints in Herpetology:114.

Boulenger, G. A. 1893. Catalogue of the snakes in the British Museum (Natural History). Trustees of the British Museum, London <link>

Cadle, J. E. 1999. The Dentition, Systematics and Phylogeny of Pseudoxyrhopus and Related Genera from Madagascar (Serpentes: Colubridae) with Descriptions of a New Species and a New Genus. Bulletin of the Museum of Comparative Zoology at Harvard College 155:381-443 <link>

Hawlitschek, O., B. Brückmann, J. Berger, K. Green, and F. Glaw. 2011. Integrating field surveys and remote sensing data to study distribution, habitat use and conservation status of the herpetofauna of the Comoro Islands. ZooKeys 144:21–78 <link>

Hawlitschek, O., Nagy, Z., & Glaw, F. 2012. Island evolution and systematic revision of Comoran snakes: why and when subspecies still make sense. PLoS ONE 7:e42970 <link>

Hedges, S. B. 2013. Revision shock in taxonomy. Zootaxa 3681:297-298 <link>

Meirte, D. 2004. Reptiles. Pages 201-224 in M. Louette, D. Meirte, and R. Jocqué, editors. La faune terrestre de l'archipel des Comores. MRAC, Tervuren.

Nagy, Z. T., U. Joger, M. Wink, F. Glaw, and M. Vences. 2003. Multiple colonization of Madagascar and Socotra by colubrid snakes: evidence from nuclear and mitochondrial gene phylogenies. Proceedings of the Royal Society of London. Series B: Biological Sciences 270:2613-2621 <link>

Peters, W. C. H. 1873. Über eine von Hrn. F. Pollen und van Dam auf Madagascar und anderen ostafrikanischen Inseln gemachte Sammlung von Amphibien. Monatsberichte der Königlichen Preussische Akademie des Wissenschaften zu Berlin 1873:792-795 <link>

Savitzky, A. H. 1981. Hinged teeth in snakes: an adaptation for swallowing hard-bodied prey. Science 212:346-349 <link>

Uetz, P. 2010. The original descriptions of reptiles. Zootaxa 2334:59-68 <link>

Vieites, D. R., F. M. Ratsoavina, R.-D. Randrianiaina, Z. T. Nagy, F. Glaw, and M. Vences. 2010. A rhapsody of colours from Madagascar: discovery of a remarkable new snake of the genus Liophidium and its phylogenetic relationships. Salamandra 46:1-10 <link>

Zaher, H., F. G. Grazziotin, R. Graboski, R. G. Fuentes, P. Sánchez-Martinez, G. G. Montingelli, Y. P. Zhang, and R. W. Murphy. 2012. Phylogenetic relationships of the genus Sibynophis (Serpentes: Colubroidea). Papeis Avulsos de Zoologia (Sao Paulo) 52:141-149 <link>

Creative Commons License

Life is Short, but Snakes are Long by Andrew M. Durso is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

Tuesday, September 23, 2014

Snake poop and the adaptive ballast hypothesis

Click here to read this post in Spanish
Haga clic aquí para leer este blog en español

Alternate title suggested by David SteenWhy snakes might benefit from holding it 

Most people probably spend as little time as possible thinking about poop, especially snake poop. Some animals produce enormous amounts of poop, like dairy cows. Others make lots of little poops - up to 50 a day in small birds.  In contrast, snakes don't poop much at all. In fact, because they eat so infrequently, snakes probably poop the least often of almost any animal. Anyone who has kept a snake as a pet can tell you that a few days after they're fed, most snakes tend to poop once (often in their water bowls, for some annoying reason), and they might poop again within a few more days. Like bird poop, snake poop is made up of two parts - the brown stuff (the fecal fragment, aka the actual poop) and the white stuff (the uric acid fragment, aka the pee, in a solid form). Also like birds, most reptiles use uric acid rather than urea to excrete their excess nitrogen, which helps them conserve water.

A young Racer (Coluber constrictor) that has eaten a
Ring-necked Snake (Diadophis punctatus) nearly 92% its length
You wouldn't think there would be much that's interesting about snake poop, but to a snake biologist everything about snakes is interesting. In 2002, Harvey Lillywhite, Pierre de Delva, and Brice Noonan published a chapter in the book Biology of the Vipers that detailed their studies on snake poop. Their most amazing finding was that some snakes can go for a really, really long time without pooping. As in, over a year. It's not because they're constipated though - these long fecal retention periods have actually evolved for a purpose in snakes. Here's what happens: most snakes eat very large meals, and they eat them all in one piece. That means that when a snake eats a meal, its body mass can more than double all at once, and it can only digest that meal from the outside in, because it hasn't chewed or cut it up into small pieces to increase its surface area. Even for the insane digestive tract of a snake, this is an incredible feat.

And the python's small heart grew two sizes that day
Figure from Riquelme et al. 2011
A well-publicized series of studies by Steve Secor and Jared Diamond on snake digestion is more than fascinating enough to warrant some digression. They revealed that some snakes actually let their digestive tracts atrophy between meals, and rebuild them (and many of their other organs, including their hearts, which double in size) each time they eat. If that sounds strange, remember that some snakes only eat a few times a year, unlike we mammals who must eat every day. In one paper on the subject, the authors used an analogy with driving a car in normal traffic vs. stopping at a railroad crossing. It's fine to keep the engine running during a brief stop, but turning the engine off saves fuel while waiting for a train to pass. By shrinking their organs, snakes are saving energy during the long fasts between meals. The flexibility of their body temperature and fundamental differences in their mitochondria are two of the ways in which snakes are able to endure these extreme fluctuations in their metabolic rate. As their gut size and metabolic rate change, so does their ability to uptake nutrients, which brings us back to the production of poop.

Uromacer oxyrhynchus just can't hold it's poop
Poop is what's left behind after your gut has extracted all the nutrients it can from a meal. The ability of a snake's gut to extract nutrients from its prey can change a lot as the gut itself is rebuilt following a meal. Specifically, it is highest following feeding and tapers off as physiology and morphology return to their pre-feeding states. Normally, once food has been reduced to poop, it doesn't hang around for long. This is true in mammals and birds and in some snakes, including ratsnakes, which normally take about two days between eating and pooping. Even that's relatively long compared with we humans, who are clinically constipated after three days. Other relatively slender or arboreal snakes such as bush and tree vipers (3-7 days) and tree pythons (~6 days) poop fairly regularly, and fecal retention time is at a bird-like minimum of 23 hours in the slender Hispaniolan Pointed-nosed Snake (Uromacer oxyrhynchus). But in other snakes, particularly in heavy-bodied species of henophidians and especially in terrestrial vipers, poop stays in the hindgut for months, even when they are fed often. The maximum values recorded by Lillywhite for boas and pythons fed mice are impressive: 76 days in an Emerald Tree Boa (Corallus caninus), 174 days in a Burmese Python (Python molurus), and 386 days in a Blood Python (P. curtus). For vipers, the figures are just as astounding: 116 days for a Puff Adder (Bitis arietans) and 286 for a Rhinoceros Viper (B. nasicornis) are among the longest, although nothing holds a candle to the heavyweight champion: one Gaboon Viper (B. gabonica) in Lillywhite's dataset that didn't poop for 420 days!

A Burmese Python intestine before (top), two days
after (middle), and 10 days after (bottom) eating.
From Secor 2008
The intestine of a snake can hold a lot of poop. Lillywhite & colleagues measured this by pumping (dead) snake intestines full of saline and found that an average viper hindgut can hold about twice as much total volume as that of a ratsnake. The cumulative mass of the poop stored by the vipers in their study totaled between 5 and 20% of the total body mass of the snakes. In humans, this kind of thing would cause an awful, awful death (some say that's what happened to Elvis). Why did these snakes do this? Lillywhite and colleagues put forth what they called the adaptive ballast hypothesis to explain their observations. When I first heard about the adaptive ballast hypothesis, I actually thought it would be that snakes held onto their poop so that they could use it defensively, in case they needed it to spray onto their would-be assailants during some future predation attempt or capture by a herpetologist. But in fact, it goes something like this:

Poop from this African Rock Python's last meal might help anchor it
as it laboriously swallows this wildebeest
Clearly, being heavy is not advantageous for arboreal snakes, so they poop on a regular basis shortly after eating. In terrestrial snakes, however, a little extra weight can give a snake a distinct advantage in capturing and handling large, potentially dangerous prey. Stored feces contribute an easily-altered component to the body's mass, an inert ballast that, unlike muscle, requires no energy to maintain (so long as the animal is sitting still and doesn't have to drag it around, a perfect fit for the sedentary lifestyle of pythons and vipers - no word yet on fecal retention in the sluggish elephant trunksnakes). This extra mass is concentrated in the posterior of the body, where it presumably increases both the inertia of that region and its friction with the ground. Essentially, the humongous mass of poop could anchor the back end of the snake during a strike or while constricting. Although no one has explicitly tested this idea, it's compelling, because the same evolutionary pressures that caused pythons and vipers to have heavy bodies in the first place could be selecting for these long retention times if they help the snakes more easily obtain food. What's more, the snakes could jettison their ballast quickly if it became a liability, such as following a new meal, before undertaking a long-distance movement, upon becoming pregnant, or prior to hibernation, thereby reducing their body mass by as much as 20% at one go.

In addition to providing ballast, the long time the fecal material spends inside the intestine could potentially increase the absorption of nutrients and water, although it probably doesn't take many months before the snake has got all it can out of its old meals. Uric acid and feces are normally mixed in snakes with short passage times, but in heavy-bodied viperids, boids, and pythons, feces are usually more compact and are more separate from the uric acid.

Few people have looked very deeply into these patterns of defecation (perhaps few would want to), so a lot of questions remain: does more frequent activity induce premature defecation? Do drinking or skin shedding influence defecation patterns? Do these patterns hold up in the field? What other functions might snake poop have? One study showed that captive snakes pooped more quickly after their cages were cleaned, whereas control animals whose cages were merely rearranged did not, which suggests that snakes might be using their feces for marking...something (we really don't know what since they aren't generally thought of as territorial, although they are a whole lot more social than most give them credit for). The mysteries are many.


Thanks to Pedro Rodriguez for allowing the use of his photograph.


Castoe, T. A., Z. J. Jiang, W. Gu, Z. O. Wang, and D. D. Pollock. 2008. Adaptive evolution and functional redesign of core metabolic proteins in snakes. PLoS ONE 3:e2201 <link>

Chiszar, D., S. Wellborn, M. A. Wand, K. M. Scudder, and H. M. Smith. 1980. Investigatory behavior in snakes, II: Cage cleaning and the induction of defecation in snakes. Animal Learning & Behavior 8:505-510 <link>

Cundall, D. 2002. Envenomation strategies, head form, and feeding ecology in vipers. Pages 149-162 in G. W. Schuett, M. Höggren, M. E. Douglas, and H. W. Greene, editors. Biology of the Vipers. Eagle Mountain Publishers, Eagle Mountain, UT <link>

Lillywhite, H. B., P. de Delva, and B. P. Noonan. 2002. Patterns of gut passage time and the chronic retention of fecal mass in viperid snakes. Pages 497-506 in G. W. Schuett, M. Höggren, M. E. Douglas, and H. W. Greene, editors. Biology of the Vipers. Eagle Mountain Publishers, Eagle Mountain, UT <link>

Riquelme, C. A., J. A. Magida, B. C. Harrison, C. E. Wall, T. G. Marr, S. M. Secor, and L. A. Leinwand. 2011. Fatty acids identified in the Burmese Python promote beneficial cardiac growth. Science 334:528-531 <link>

Secor, S. M. and J. Diamond. 1998. A vertebrate model of extreme physiological regulation. Nature 395:659-662 <link>

Secor, S. M. and J. M. Diamond. 2000. Evolution of regulatory responses to feeding in snakes. Physiological and Biochemical Zoology 73:123-141 <link>

Secor, S. M. 2008. Digestive physiology of the Burmese Python: broad regulation of integrated performance. Journal of Experimental Biology 211:3767-3774 <link>

Creative Commons License

Life is Short, but Snakes are Long by Andrew M. Durso is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

Saturday, August 30, 2014

Filesnakes, Wartsnakes, or Elephant Trunksnakes

Arafura Filesnake (Acrochordus arafurae)
In the swamps, marshes, streams, and estuaries of northern Australia and southeastern Asia live ancient snakes as thick as your arm, with tongues as thin as a thread, skin as rough as a file, and a disposition as gentle as a lamb. These snakes comprise the family Acrochordidae (from the Greek akrochordon, wart), and are known as filesnakes1, wartsnakes, or elephant trunksnakes. In Indonesian they are known as karung, which means 'sack'; in Thai, as ngū nguang-cĥāng, 'elephant-trunk snake'. There are three species, all in the genus Acrochordus: Javan (A. janavicus), Arafura (A. arafurae), and Little (A. granulatus). The largest, female Javans, grow up to 8 feet and over 20 pounds. Acrochordids are an old and highly distinct group of snakes, distantly related to colubroids, with which they share a common ancestor between 50 and 70, but possibly as long as 90 million years ago.

Close-up of A. arafurae scales
Filesnakes have strongly-keeled scales with the texture of sandpaper or a coarse file, after which they are named. They have very loose, baggy skin. I held one once and it felt like a human arm inside the sleeve of a very sturdy, very baggy rain poncho made out of chain mail. This loose skin is likely an adaptation that allows filesnakes to withstand the great force of the initial dash for freedom of their fish prey. Their sharp scales are used to help gain purchase on slimy fish skin during constriction. They have sensory, bristle-bearing tubercles on the skin between their scales, as well as sensory organs on the scales themselves, both of which presumably help them sense the underwater movements of nearby prey, analogous to the lateral line systems of fishes and some amphibians. Acrochordids sometimes ambush their prey, but more often they forage by searching slowly along the edges of mangroves, billabongs, and other water bodies at night, looking for sleeping fishes and crustaceans (although they don't tear them apart like some southeast Asian snakes). During the day, filesnakes hide in the shadows of overhanging trees, moving with them to remain concealed from predatory birds. They are nearly incapable of moving on land, and shed in the water using a knotting behavior similar to that of Pelamis platurus, the most completely aquatic sea snake.

Acrochordus arafurae regurgitates
an eel-tailed catfish (Tandanus tandanus)
whose spine has pierced its neck
Filesnakes occupy a unique phylogenetic position, not closely related to anything but somewhere in-between the colubroids ("advanced snakes") and the "henophidians" (boas, pythons, and other stem-group snakes). A few recent papers reanimate an old hypothesis that they might be closely related to dragonsnakes, but historically acrochordids have been considered the sister group to all colubroids, a group of >2,850 species (>80% of all living snakes) that includes dragonsnakes, asymmetrical slug-eaters, vipers, homalopsids, elapids, lamprophiids, and strict colubrids. Colubroidea and the three Acrochordus species together form the Caenophidia ("recent snakes"). Acrochordids share unspecialized head scales and undifferentiated ventral scales with boas and pythons, but they are united with colubroids in that they totally lack vestigial limbs and have spines on their hemipenes, a well-developed vomeronasal system, and several particular characteristics of skull morphology, including a coronoid bone. Other features of the skull and skeleton are unique to acrochordids, including the aforementioned skin sense organs, a passive joint between the frontal and parietal skull bones, the presence of certain holes in the vertebrae, the shape of the head of the ribs, and an ear region that most closely resembles the ears that other snakes have as embryos, but which forms in a different way. Acrochordids also have an unusual lung morphology, with a double row of holes leading from the trachea into individual small lunglets, and a more tangled intestinal tract than other snakes.

Acrochordus granulatus with algae growing on its back
Filesnakes have incredibly low metabolic rates, even for a snake, and cannot sustain rigorous physical activity for very long. In captivity, they "epitomize sluggishness in snakes", although radio telemetry has shown that in the wild they move around wetlands slowly but steadily, covering up to 450 feet per night. They can remain submerged for over an hour (record 2 h 20 min), and surface to take about 5 breaths, about one per minute. The first several of these breaths oxygenate the blood, and the last one fills the multi-chambered lung. In addition, Little Filesnakes have about twice as much blood as other snakes, and this voluminous blood is about twice as thick with red blood cells as even that of other diving snakes. Their hemoglobin has a very high affinity for oxygen, which results in their being able to store between three and fifteen times as much oxygen in their blood as a similarly-sized sea snake, and release it slowly over a long period of time. Many turtles also use this strategy. Also like turtles, filesnakes can both obtain oxygen from and release acidic carbon dioxide into the water through their skin, which helps prolong their dives.2 In fact, filesnakes are so well-adapted to sitting still that they are practically incapable of exercise, and get tired out quickly.

This slow theme carries over into filesnake life history. Male filesnakes mature around six years old, females around nine, and 8-10 years may elapse between consecutive births. Studies from northern Australia found that only a small proportion of females are reproductive in any given year, and that only the very largest females reproduce relatively frequently. Large Javan Filesnakes give birth to as many as 52 young at once, although the average is closer to 30. Arafura Filesnakes average about 16 (as few as 9 and as many as 25 have been reported), and Little Filesnakes about 6 (as few as 1 and as many as 12). Female filesnakes are courted by up to eight males at a time in shallow water. Their population dynamics are driven by rainfall in northern Australia. One captive filesnake gave birth to a single young after seven years of isolation, suggesting that filesnakes are either capable of parthenogenesis or of very prolonged sperm storage.

The Little Filesnake (Acrochordus granulatus) has
a banded pattern like a sea krait (Laticauda colubrina)
At first glance the three extant Acrochordus species seem quite similar, but in fact they exhibit striking differences in both anatomy and ecology. The Little Filesnake (Acrochordus granulatus) was described in 1799 and used to be classified in a separate, monotypic genus (Chersydrus). As its name suggests, it is the smallest acrochordid (~ 3 feet in total length) and the most widely distributed. It is found along the coast from northwestern India throughout southeast Asia and Indonesia, reaching east to the Solomon Islands. Its diverse habitats include freshwater lakes, rivers, mangroves, mudflats, reefs, and the open ocean, up to 6 miles offshore and over 60 feet deep. It is the most marine of the filesnakes, the most brightly patterned, and has a shorter, more laterally-compressed tail, more granular scales, more dorsally-oriented nostrils, and a salt excretion gland beneath its tongue.3

Acrochordus javanicus
The Javan Filesnake (Acrochordus javanicus) was the first to be described, in 1787, and is the largest and heaviest filesnake, sometimes reaching 8 feet and over 20 pounds. It is found in fresh and brackish water on the Malay Peninsula and on the islands of Sumatra, Java, and Borneo (and was introduced to Florida in the 1980s, although it does not appear to have established there). It is harvested for meat and for its skin, out of which is made fine leather; up to 2 million are exported from Indonesia annually. Unlike other filesnakes, the posterior-most teeth in its lower jaw have sharp edges. The Arafura Filesnake (Acrochordus arafurae) was thought to be the same species as the Javan until 1979. It grows as long but at the same size is only about half as heavy-bodied. It is found only in freshwater habitats in northern Australia and southern New Guinea. Surprisingly, A. arafurae is more closely related to A. granulatus than either is to A. javanicus, a relationship that is supported by genetics as well as morphology.

The long, thin tongue of Acrochordus javanicus
From Greene (1997)
Fossil Acrochordus have been found in Pakistan and Nepal, as well as within the extant range. These extinct filesnakes date from 5-20 million years ago during the Miocene, only a few million years after the Indian plate crashed into Asia to form the Himalayas. They grew larger than modern filesnakes, reaching at least 9 feet, and are the most well-represented snakes in the southern Asian fossil record, possibly because their habitat lends itself well to fossilization. The extinct species Acrochordus dehmi is represented by over 1000 fossils from over 100 different locations, and probably went extinct about 6 million years ago. Because it is so well-known, we can say with confidence that it is more closely related to A. javanicus than to the other two living species of AcrochordusMolecular clock methods suggest that the three modern species of Acrochordus and A. dehmi diverged from one another 16-20 million years ago, a timescale that usually justifies separation into family-level or higher categories. Despite their superficial similarities, the ecological and morphological differences among the three living Acrochordus species have been considered equivalent to differences among genera in other groups of snakes. Because no fossil acrochordids have been found in Australia, it is assumed that they evolved in Asia and spread to Australia in the last 5 million years. It is also likely that the ancestors of the Little Filesnake entered the ocean before sea snakes (~7 mya) and kraits (~13 mya) and just after marine homalopsids (~18 mya).

Acrochordus in contemporary aboriginal artwork by Chris Liddy (Moonggun),
showing the embryos inside the snake in the northern Australian style
One of the most interesting things about filesnakes is that Aboriginal Australians collect and eat them in some areas. Mostly this is done at the end of the Australian dry season, in November, when water levels are lowest and the snakes are easiest to find and capture. Although the snakes themselves don't generally put up much resistance, the old women who hunt them do so by wading into murky waters filled with crocodiles and feel under overhanging banks, weed beds, and logs, sometimes collecting over 30 snakes per person-hour. Often the snakes are killed immediately by biting their necks. The pregnant females are highly prized for their embryos, which are cooked on hot ashes, eaten like popcorn, and called 'cookies' by Aboriginal children.

1 Not to be confused with African Filesnakes (genus Mehelya), which are so-named not for their texture but for their cross-sectional shape, which resembles a triangular file.

2 Although the warm, shallow, slow-moving waters in which they live are fairly oxygen-poor and oxygen is difficult to extract out of salty water, so augmenting their ability to hold their breath using their massive blood oxygen reservoir is almost certainly of greater importance.

3 Little Filesnakes can excrete salt but gradually get dehydrated, so they must have a source of fresh water. They drink rain that falls on the ocean or migrate to areas where rivers flow into estuaries. This is because, like other marine reptiles, filesnakes "pee like a fish": they excrete nitrogen as ammonia, rather than as uric acid like other snakes or as urea like mammals. This is much more wasteful of water than the uric acid method, and it's not clear why they do this.


Thanks to Chris LiddyMatt Summerville, Darryl Houston, M. & P. Fogden, Jordan de Jong, Stephen Zozaya, Jason Isley, and Dick Bartlett for their photos, and to Rick Shine for information on tracking down Darryl Houston.


Boulenger, G. A. 1893. Catalogue of the snakes in the British Museum (Natural History). Trustees of the British Museum, London <link>

Feder, M. E. 1980. Blood oxygen stores in the file snake, Acrochordus granulatus, and in other marine snakes. Physiological Zoology 53:394-401 <link>

Heatwole, H. and R. Seymour. 1975. Pulmonary and cutaneous oxygen uptake in sea snakes and a file snake. Comparative Biochemistry and Physiology Part A: Physiology 51:399-405 <link>

Houston, D. and R. Shine. 1994. Movements and activity patterns of Arafura filesnakes (Serpentes: Acrochordidae) in tropical Australia. Herpetologica 50:349-357 <link>

Lillywhite, H. B. and T. M. Ellis. 1994. Ecophysiological aspects of the coastal-estuarine distribution of acrochordid snakes. Estuaries 17:53-61. <link>

Lillywhite, H. B., A. W. Smits, and M. E. Feder. 1988. Body fluid volumes in the aquatic snake, Acrochordus granulatus. Journal of Herpetology 22:434-438 <link>

Madsen, T. and R. Shine. 2000. Rain, fish and snakes: climatically driven population dynamics of Arafura filesnakes in tropical Australia. Oecologia 124:208-215 <link>

Magnusson, W. A. 1979. Production of an embryo by an Acrochordus javanicus isolated for seven years. Copeia 1979:744-745 <link>

McDowell, S. B. 1975. A catalogue of the snakes of New Guinea and the Solomons, with special reference to those in the Bernice P. Bishop Museum. Part II. Anilioidea and Pythoninae. Journal of Herpetology 9:1-79 <link>

McDowell, S. B. 1979. A catalogue of the snakes of New Guinea and the Solomons, with special reference to those in the Bernice P. Bishop Museum. Part III. Boinae and Acrochordoidea (Reptilia, Serpentes). Journal of Herpetology 13:1-92 <link>
Intertubercular papilla of Acrochordus granulatus
From Povel & Van Der Kooij 1996

Povel, D. and J. Van Der Kooij. 1996. Scale sensillae of the file snake (Serpentes: Acrochordidae) and some other aquatic and burrowing snakes. Netherlands Journal of Zoology 47:443-456 <link>

Pyron RA, Burbrink F, Wiens JJ, 2013. A phylogeny and revised classification of Squamata, including 4161 species of lizards and snakes. BMC Biology 13 <link>

Rasmussen, A. R., J. C. Murphy, M. Ompi, J. W. Gibbons, and P. Uetz. 2011. Marine Reptiles. PLoS ONE 6:e27373 <link>

Rieppel, O. and H. Zaher. 2001. The development of the skull in Acrochordus granulatus (Schneider)(Reptilia: Serpentes), with special consideration of the otico‐occipital complex. Journal of Morphology 249:252-266 <link>

Sanders KL, Mumpuni, Hamidy A, Jead J, Gower D, 2010. Phylogeny and divergence times of filesnakes (Acrochordus): inferences from morphology, fossils and three molecular loci. Molecular Phylogenetics and Evolution 56:857-867 <link>

Seymour, R., G. Dobson, and J. Baldwin. 1981. Respiratory and cardiovascular physiology of the aquatic snake, Acrochordus arafurae. Journal of Comparative Physiology 144:215-227 <link>

Shine R, 1995. Australian Snakes: A Natural History Ithaca, New York: Cornell University Press <link>

Shine, R. 1986. Sexual differences in morphology and niche utilization in an aquatic snake, Acrochordus arafurae. Oecologia 69:260-267 <link>

Shine, R. 1986. Ecology of a low-energy specialist: food habits and reproductive biology of the arafura filesnake (Acrochordidae). Copeia 10:424-437 <link>

Shine, R. 1986. Predation upon filesnakes (Acrochordus arafurae) by aboriginal hunters: selectivity with respect to body size, sex and reproductive condition. Copeia 10:238-239 <link>

Shine, R. and D. Houston. 1993. Acrochordidae. in C. Glasby, G. Ross, and P. Beesley, editors. Fauna of Australia. AGPS, Canberra <link>

Shine, R., P. Harlow, J. S. Keogh, and Boeadi. 1995. Biology and commercial utilization of acrochordid snakes, with special reference to karung (Acrochordus javanicus). Journal of Herpetology 29:352-360 <link>

Voris, H. K. and G. S. Glodek. 1980. Habitat, diet, and reproduction of the file snake, Acrochordus granulatus, in the straits of Malacca. Journal of Herpetology 14:108-111 <link>

Creative Commons License

Life is Short, but Snakes are Long by Andrew M. Durso is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.