Komodo Dragon: extending the saliva story

The previous post generated a number of emails on the use of biologically-active substances in saliva by the Komodo Dragon to help kill its prey, ranging from ‘unnecessary and highly speculative’ to ‘generally supportive but needs more work’. I was also told that the post seemed somewhat harsh. It was not meant that way but I think my style in writing it reflects the sort of debate that has arisen and the tone adopted by some authors and commentators. The comments/replies on a paper* are as robust as any I have seen in print for a long time.

Much of the controversy centres on use of the terms ‘venom’, ‘venomous’ and ‘venom glands’. I will not go into the arguments further here but there is a long-standing problem in naming structures or molecules for presumed or first-discovered function, potential function or partial function. The names of hormones and growth factors are well-known nightmares in this regard. There are actually two hypotheses, one an evolutionary question on the stage at which reptiles acquired the capacity to produce venom—the Toxicoferon Hypothesis—and the related but distinct question of how the Komodo Dragon (and possibly other extant and extinct monitor lizards) kills its prey. This latter question is susceptible to morphological, chemical, physiological and pharmacological observation and experiment as demonstrated by Bryan Fry and his international team in their original paper. I will restrict myself to expanding the dragon story.

Typical Komodo Dragon habitat on the island of Rinca, part of the Komodo
National Park, with gullies and open areas of grassland

My own view is that there is a chain of events by which biologically active substances in saliva could be used to decrease the time between initial attack and death of the large mammal prey. At present, though, I see a gap in knowledge between a scenario which describes what could happen and one that describes what does happen. I also see it as a gap that could be filled and which, if filled, would subdue further objections and a great deal of speculation on alternative roles for biologically active substances in the Dragon’s saliva.

There is strong evidence in favour of saliva doing something—and something deleterious to the life of the prey. The presence of a relatively large lumen in the mandibular gland where secretion can be stored is highly suggestive of the use of a ready store of saliva when a dragon attacks prey, and a store that can be delivered to the base of the teeth. The authors of the original papers also showed by direct biological assays that dragon saliva has components that act to cause vasodilatation and decrease blood pressure and to contain substances that are potent anticoagulants.

Therefore, a good case has been made for local effects of biologically-active substances on the wound inflicted by the dragon—parallel deep slashes inflicted by pulling after the initial bite. Diffusion into the tissues surrounding the wound would increase the rate of blood loss and keep the severed ends of large and small arteries open and the blood flowing from all severed vessels. That’s where the relatively low bite strength of the Dragon which was demonstrated by Fry and his collaborators would be of selective advantage. If the Dragon bites, pulls what have been described as its steak-knife teeth across blood vessels (‘grip-and-rip’), and holds that grip, the last thing it needs is a strong bite. Compression by a strong bite would compress the wound and slow the rate of blood loss. Moreover, it would decrease the rate of movement of substances from the saliva away from the wound into surrounding tissues and into the circulation through venous or lymphatic drainage.

So, as far as local actions of saliva are concerned, they must be present and I can think of experiments that would test how far diffusion would take them into surrounding tissues. 

Central effects, for example, in decreasing blood pressure, would be harder but not impossible to demonstrate, as would establishing the concentrations of substances from saliva in the peripheral circulation of the prey in order to determine whether they reach the levels required to exert the effect that can be demonstrated with particular concentrations in biological assay systems. The latter is not a trivial question with substances that would be appearing in the circulation relatively slowly from a wound and which may be relatively quickly broken down into inactive compounds.

The presence of several compounds known to be part of the arsenal of toxins deployed by classic venomous reptiles adds weight to the suggestion of some systemic role for Dragon saliva in the bitten animal. Others have suggested that may may fulfil other roles in the mouth or in digestion of the prey before ingestion and the action of stomach enzymes; I do not find the latter argument compelling although it could be tested experimentally.

So, while demonstration of central toxic events has not yet been attempted, there must be a local effect of saliva, at least, on the gripped-and-ripped prey. The question that remains on this local action is whether the effect is biologically significant in shortening the life or ability to resist of the animal that has been ambushed, or is so in a sufficiently large number of attacks as to confer a selective advantage.

Both local and systemic actions rely on getting saliva into the wound. Different durations of bites could have different effects. The mandibular gland implicated in storage seems to work in a similar way to that of back-fanged snakes in that the pressure exerted by chewing forces the secretion out. If that is so then a ‘grip-and-rip’ feeding attack could have an entirely different outcome from a quick offensive or defensive nip.

Some commentators have suggested testing the components of Dragon saliva on their natural prey. However, what is their natural prey? Water Buffalo were introduced relatively recently, Timor Deer in antiquity while domestic goats are, well, domestic. The Komodo Dragon and its extinct relatives appear to have been around the Indonesian islands, where they spread from Australia, even before the now extinct dwarf elephant appeared in their habitat so that fundamental mismatch between the bulk and apparent adaptations for killing large animals of Komodo Dragons and the size of potential prey species seems to be unexplained.

Other questions arise like are there differences between the sexes in the size of prey and feeding strategies, adult females being lighter than males? Do juvenile Dragons attack prey that is larger than that attempted by, say, Water Monitors (Varanus salvator) of similar size and, if so, would that make the use of saliva as a chemical weapon to help in subduing prey more likely? In other words, the advantage may come more to juveniles tackling large prey than to adults. Indeed, if other species such as the Lace Monitor or Goanna (Varanus varius) from Australia have a similar array of potential biological agents, which seems to be the case, and in quantitatively equivalent amounts, might the whole arrangement be an adaptation to extend in size the range of prey that can be killed and eaten by some or all monitor lizards?

So, after arguing with myself several times and learning lots about venomous and non-venomous reptiles, that’s may take on the Komodo Dragon’s feeding method in the light of present evidence. In essence, I argue that the available evidence points to a local chemical role for saliva in helping to subdue/kill large prey (while recognising that sheer physical force may often or sometimes be sufficient). As for toxins in saliva having a systemic effect on the prey, I argue that the gap between the prima facie potential effect and actual effect has not yet been filled by experimental data. And as to whether we should call Dragons ‘venomous’ or the gland the ‘venom’ gland…

...I will leave to others.

Whatever the final outcome on these fascinating and enigmatic lizards, the authors of the 2009 paper have done a signal service in examining how it and other extinct and extant monitor lizards kill their prey and have not shied from erecting hypotheses on possible mechanisms and their wider importance in the evolution of reptiles. 


*Fry BG, Casewell NR, Wüster W, Vidal N, Young B, Jackson TNW. 2012. The structural and functional diversification of the Toxicofera reptile venom system. Toxicon 60, 434-448. Weinstein SA, Keyler DE, White J. 2012. Replies to Fry et al. (Toxicon 2012, 60/4 434-448). Part A. Analyses of squamate reptile oral glands and their products: A call for caution in formal assignment of terminology designating biological function, Toxicon 60, 954-963. Kardong KV. 2013. Replies to Fry et al. (Toxicon 2012 60/4 434-448). Part B. Properties and biological roles of squamate oral products: The “venomous lifestyle” and preadaptation, Toxicon 63, 113-115. Jackson TNW, Casewell NR, Fry BG. 2013. Response to “Replies to Fry et al. (Toxicon 2012, 60/4, 434–448). Part A. Analyses of squamate reptile oral glands and their products: A call for caution in formal assignment of terminology designating biological function”, Toxicon 64, 106-112. Weinstein SA, White J, Keyler DE, Kardong KV. 2013. Response to Jackson et al. (2012), Toxicon 64, 116-127.

Komodo Dragons. A peaceful morning in the Komodo National Park and an acrimonious debate on reptilian venoms

Last September’s Expedition Cruise from Darwin through the Lesser Sunda islands of East Timor and Indonesia included stops at Rinca and Komodo. Ever since my fourth cousin once removed visited Komodo in 1956 I have wanted to see Komodo Dragons alive and in their natural habitat. They did not disappoint. We saw very large males (hanging out around the kitchen of the ranger station on Rinca hoping for a hand out), females and juveniles (but not the hatchlings which apparently take to the trees to avoid their predatory parents). One female was digging out the nest mound of a megapode, the Orange-footed Scrubfowl (Megapodius reinwardt) in which to lay her own eggs, September being the egg-laying time in their breeding cycle.

When we got back I had a chance to look at and think about the various hypotheses that have been advanced as to how Komodo Dragons—and, possibly, some other monitor lizards—kill their prey. The highly publicised but never properly tested proposal that Komodo Dragons are venomous in the sense that venomous snakes are venomous, i.e. toxins delivered by injection having a very or fairly rapid systemic effect on the prey, seems to be in the process of being discarded in the course of some pretty acrimonious arguments, along with the associated hypothesis of a single, early origin of venom in reptilian evolution—the Toxicofera hypothesis. The other idea, that pathogenic bacteria harboured in the mouth of dragons causes sepsis in prey animals that are bitten but escape the initial attack, doesn’t seem that convincing or special given the likelihood of infection from a bite from any animal, as any postman will testify, especially if that postman were to immerse his bitten ankle in fetid water, as non-native water buffalo do when bitten by a Dragon. If I were to bite a Timor Deer on the leg, there is every likelihood an infected wound that could impede mobility would ensue. 

Given the opposition to the Toxicofera hypothesis and the contentious nature of the evidence for Komodo Dragons being venomous, it seems a pity that the BBC repeated last November the screening of an episode from its 2011-12 season of the Natural World Series called Komodo—Secrets of the Dragon that was devoted virtually entirely to that proposition and its main protagonist. Had this been a programme about a topic other than science-led natural history the BBC would have been falling over itself to show someone with the opposite view, however perverse. As it is, the viewing public in Britain has been left with the impression that the venomous nature of the Dragon is accepted. At the very least the BBC should have been aware of what was going on (by reading Wikipedia for example) and before rescreening it should have added an annex to the original programme. That annex could have  explained the opposing views and evidence—some of which made news media reports*—that have been accumulating since 2009. Adam Hargreaves (Oxford), Abigail Tucker (King’s, London) and John Mulley (Bangor), who produced a devastating critique** of the toxicoferan hypothesis in 2015, should have been consulted and involved. But let’s leave the BBC and its usually excellent (but sometimes spectacularly poor) natural history programmes and return to the mouth of the dragon.

None of the criticisms of the venom hypothesis imply that the composition and quantity of saliva are not of selective advantage to the despatch of prey, subsequent swallowing and digestion or protection of the oral cavity against infection. Nor, indeed, do they refute the idea or evidence that saliva may have local beneficial—to the Dragon—effects on inflicted wounds, like the application of a secreted anticoagulant, for example.

The drooling mouths of some of the male Dragons we saw were impressive. The only comparable example I could think of was my late mother-in-law’s Boxer dogs given the slightest hint of finding something edible. With no obvious sign of food or feeding activity, is Dragon saliva being used for some other purpose like scent marking?

The accumulated data that I have found suggests to me that Komodo Dragons, as originally thought, kill their prey by sheer brute force from wounds inflicted by a very large mouth with very big teeth. In one study†, 17 attacks on large prey were observed, 12 were fatal. Of the 5 that escaped with injuries to their limbs and rump, 1 was quickly attacked and killed by a second Dragon, 2 died within hours, one fled being pursued by other Dragons and one limped away without being pursued. The eventual fate of the two that possibly survived is not known.

One may though ask how, if the Komodo Dragon originally preyed upon the now extinct dwarf elephant, Stegodon, of Flores, as suggested in 1987 by Jared Diamond, those beasts (smaller than a domestic water buffalo) were killed? Would biologically-active substances in saliva have been of selective advantage?

Nowhere have I found (but I am not have looked in the right place) any discussion of the selective benefits of a slow death of the prey to the predator. A snake with a fast-acting neurotoxin can quickly track and swallow its prey without that investment in metabolically expensive venom being lost to another snake or any old predator happening to find the corpse. But let’s say a Dragon does inject toxins with a small bite and the animal runs away to die. There is no guarantee that the investment in venom would pay off. With lots of other Dragons around, the prey could be lost entirely or the original killer would get only a small share in a communal but competitive feeding session. Surely, if monitor lizards are venomous at all then why have they not evolved a more effective venom to ensure a quick kill?

The venomous dragon and other monitor lizards hypothesis has been around now for ten years. I find it sad that so little has been done at the whole animal and tissue levels to test it. I found some of the original observations and experiments‡ unconvincing. But given the captive populations of Dragons and a local abundance on Komodo and Rinca, a rigorous examination of the whole question cannot be beyond the bounds of practical realisation or funding.

*e.g. Zimmer C. 2009. Chemicals in Dragon’s Glands Stir Venom Debate. New York Times, 18 May 2009. Yong E. 2015. A Venomous Fight Among Reptile Scientists. The Atlantic, 2 November 2015

*Hargreaves AD, Tucker, AS, Mulley JF. 2015. A critique of the toxicoferan hypothesis. In, Gopalakrishnakone P (ed), Malhotra M (ed). A critique of the toxicoferan hypothesis. In Evolution of Venomous Animals and Their Toxins: Toxicology. Springer Netherlands, p 1-15. DOI: 10.1007/978-94-007-6727-0_4-1

†Bull JJ, Jessop TS, Whiteley M. 2010. Deathly drool: evolutionary and ecological basis of septic bacteria in Komodo Dragon mouths. PLoS ONE 5(6): e11097. doi:10.1371/journal.pone.0011097

‡Fry BG Wroec S, Teeuwisse W, van Osch MJP, Moreno K, Ingle J, McHenry C, Ferrara T, Clausen P, Scheib H, Winter KL, Greisman L., Roelants K, van der Weer L, Clemente CJ, Giannakis E, Hodgson WC, Luz S, Martelli P, Krishnasamy K, Kochva E, Kwok H, Scanlon D, Karas J, Citron DM, Goldstein EJC, Mcnaughtan JE, Norman JA. 2009. A central role for venom in predation by Varanus komodoensis (Komodo Dragon) and the extinct giant Varanus (Megalania) priscus. Proceedings of the National Academy of Sciences of the USA 106, 8969-8974 doi 10.1073 pnas.0810883106