The tale of Toxicofera, part 6: the exaptation hypothesis

What is the conclusion of all this ruminating about the evolution of venom systems in reptiles? The conclusion is that there is no conclusion - the game of science is a game of discovery that never ends. Exciting, isn't it?

I started this series considering which animals are venomous by noting that academics love to argue and that this, rather than being a “bug”, is actually a feature of systems of knowledge acquisition. The only way we can gain a clearer understanding of which ideas are more (or less) valid is by putting the ideas out there and then engaging in collective criticism of them – by allowing them to do battle with one another in the “space of reasons”. We can see part of this concept embodied in a pithy quote from eighteenth century German philosopher, poet, and scientist Novalis:

“Hypotheses are nets: only he who casts will catch.”


Novalis: as pretty as his aphorisms were pithy. Engraving by Edouard Eichens.

But casting the hypothesis is just the first part. Twentieth century philosopher Karl Popper described the whole shebang as a never-ending series of “conjectures and refutations”. Conjectures are hypotheses, refutations are criticisms. Popper’s notion was that the “best” hypotheses are the ones that can withstand the most criticism. It was important for him that the criticism never end, however, because he did not believe in certainty or “ultimate truth”. The best we can do is to have a temporary “winner” – a current conception that is better supported by evidence and argument than any of its rivals, but which could nonetheless be refuted at any moment by a strong enough criticism (Popper called this “provisional acceptance”). Although Popper is mostly remembered as a philosopher of science, he understood that this process of conjectures and refutations extended to every kind of system that acquires knowledge about the world. For him, this included evolutionary systems of all kinds, including non-human organisms. Whilst it’s true that “epistemic” practices (those designed to acquire knowledge) are a particular and widespread form of evolutionary system, we’ll leave discussion of that for another time and place. For now, it’s only important to note that what distinguishes science from other epistemic practices that humans engage in is the use of controlled experiments as a special kind of criticism.

Popper unended quest

Karl Popper titled his autobiography "Unended Quest" to reflect his belief that the quest for knowledge was not about arriving at conclusions, but about continually refining our understanding in a neverending series of "conjectures and refutations".

The Toxicofera research programme

As you know if you’ve been following this blog series, “Toxicofera” has been a controversial hypothesis, that has been subjected to much criticism. Actually, there is no single “Toxicofera hypothesis”, but a series of interconnected hypotheses that have never been formally defined. Another twentieth century philosopher, Imre Lakatos, coined the term “research programme” to refer to such collections or sequences of hypotheses (or theories). It might be better, therefore, to think of there being a “Toxicofera research programme”.  Some of the hypotheses that are part of this research programme include:

1.  Toxicofera is a clade (see part one).

2.  Dental glands are an important synapomorphy (shared trait) of the Toxicofera (see part three).

3.  Monitor lizards are venomous (see part four).

4. The common ancestor of all toxicoferan reptiles was venomous.


It’s easy to see that there could be an almost limitless sequence of connected hypotheses within the research programme, and that they may exist at different levels in a hierarchy. For example, the hypothesis that Toxicofera is a clade is clearly much more general (hence closer to the foundation of the hierarchy) than the hypothesis that monitor lizards are venomous. As we’ve seen in the series so far, each of these hypotheses has been considered controversial and has been duly criticised (though rarely experimentally). Thus far, none of them has been “definitively refuted”, but that’s partly because some of them are not clearly defined and it’s hard to know what kinds of experiments or arguments could be used to “settle” the argument. Sometimes, as with both the question of whether Toxicofera is a clade and whether monitor lizards are venomous, the choice of evidence a scientist considers determinative is of paramount importance. As we’ve seen, scientists who favour phylogenies (hypotheses about relatedness) based on gene sequences have little doubt that Toxicofera is a clade, whereas those who feel morphology is a more important source of evidence remain unconvinced. Similarly, whether or not someone feels that monitor lizards should be labelled “venomous” usually depends on the sources of evidence they feel are most important. This is not uncommon in science, as a couple more quotes will suffice to demonstrate:

“All observations are theory-laden.” – Karl Popper

“The theory determines what we can observe.” – Albert Einstein

What both of these eminent thinkers are getting at is that our frame of reference, the “theory” that forms the lens through which we view the world (colloquially, our “worldview”) has a strong impact on what we see and how we interpret it. Sometimes, we like to imagine that a “scientific method” exists that enables us to negate the influence of the observer and thereby make “objective” (as opposed to “subjective”) claims about the nature of reality. Unfortunately, this is not entirely the case – science can’t give us pure objectivity, but it’s still the best system we have for getting us as close as possible, at least when it’s working the way it should.

The great Toxicofera debate 

Sheldonian Theatre

The Sheldonian Theatre at the University of Oxford: an auspicious venue for a scandalous (actually, not really) debate. Photo: Wikimedia Commons.  

In 2015, I was honoured to be invited to take part in a debate in the Sheldonian Theatre at the University of Oxford in the United Kingdom. The proposition to be debated was (something like) “venom evolved just once in the course of reptilian evolution” and I was to be on the affirmative team. Now, this is not a proposition I actually believe, but I agreed to take part because the purpose of a scientific debate is not to “win” or to “prove” one’s opponents wrong, but rather to shed further light on a contentious issue – “opponents” in a scientific debate are actually collaborators in the game of conjectures and refutations. At the event, I argued that since venom is a functional trait, differences in opinion concerning which organisms are venomous (and therefore how many times venom might have evolved) might be more to do with lack of functional data for many species and a lack of consistent application of the concept of function itself, than anything else. This may seem like prevarication or fence-sitting, but if you’ve been reading the blog you’ll know by now that I find fences quite comfortable to sit on…..or rather that I don’t think jumping to one side or the other is a productive move in and of itself. Regardless, the choice of proposition (and the fact that the affirmative side was outvoted) highlights the fact that hypothesis number 3 in my list above – “The common ancestor of all toxicoferan reptiles was venomous” – is the most controversial of all the hypotheses that are part of the Toxicofera research programme (although whether or not monitor lizards are venomous is up there too!).

Was the common ancestor of all toxicoferan reptiles venomous?

Boyds forest dragon

The spectacular Boyd's forest dragon of northern Queensland is non-venomous, but was it descended from a venomous ancestor? Probably not, but it's hard to be sure. Photo: Timothy Jackson.

This controversial hypothesis has never been formally proposed, but it has been implied with assertions that all of Toxicofera is “descended from a common venomous ancestor” and that there was a “single common origin of venom at the base of the clade”. Readers of this blog series should be able to see the problem with these statements immediately. After all, if we cannot even work out definitively if many extant species are “venomous” or not (monitor lizards, many non-front-fanged snakes), how can we possibly expect to make such a judgement about an extinct common ancestor?! Well, inferring ancestral states from extant diversity is common practice in evolutionary biology, but it’s absolutely true that when it comes to a slippery functional trait like venom, making conclusive judgements about extinct species is challenging and likely impossible (e.g. controversy about whether or not this extinct synapsid was venomous), unless venom systems recognisably similar to those of unequivocally venomous extant species are uncovered in the fossil record (as is the case for scorpions, centipedes etc). We have no such fossil evidence for the most recent common ancestor (MRCA) of the Toxicofera. Moreover, since a large number of extant toxicoferan reptiles (including one whole branch – Iguania) are almost certainly non-venomous, for the MRCA to have been venomous would require a large number of reversals. This doesn’t invalidate the inference, since we know there have indeed been many such reversals amongst snakes (some are reviewed here), but it certainly doesn’t lend it support.

The Toxicofera research programme has been highly productive

Radiated rat snake

A radiated rat snake (Coelognathus radiatus) making a threatening display. The discovery of a neurotoxin in the oral secretion of this species (secretion this very individual contributed to!) was one of the first "big discoveries" of the Toxicofera research programme. The species is a constrictor and not dangerous to humans, but it isn't shy about biting! Photo: Timothy Jackson

Even though criticism is a good thing if we want to learn, sometimes excessive criticism of trivial assertions (like the statements quoted above) can obscure the excellent work the statements are associated with. Think of this as similar to focussing all of your criticism on the marketing instead of the product – it may be necessary to criticise the marketing, but hyperbolic marketing doesn’t necessarily imply a poor product. All humans (even scientists!) are clumsy with language sometimes, and this clumsiness provides fodder for pedantic criticisms. Far be it from an arch-pedant like me to deny the value of pedantry, but sometimes focussing too much on the choice of words and ignoring what is being expressed leads to unproductive forms of criticism. Whether or not the MRCA of Toxicofera was “venomous” (there’s no reason to believe it was), there was something special about it that led to multiple distinct origins of specialised venom  systems within the clade. The work reconstructing the evolutionary history of venom systems within the Toxicofera since the early years of the twenty-first century has profoundly altered our understanding of the evolution of venom in reptiles and may have a substantial impact on our understanding of the evolutionary origins of functional traits in general. As I’ve stressed repeatedly in this series of articles, it is the detailed description of myriad transitional forms within the clade that are the truly significant data arising from this work, not the assertions that this or that species is (or is not) “venomous”. These transitional forms have been uncovered by an interdisciplinary research programme combining insights from molecular evolution, anatomical evolution, pharmacology/toxinology, and (a bit of) ecology. The product is astounding, never mind the marketing.

The exaptation hypothesis 

Velociraptor feathers

Feathers originally evolved for insulation - the dinosaur equivalent of the fur of mammals, as shown in this artist's impression of a Velociraptor. Feathers were later "exapted" for flight, by the ancestors of birds. Image: Wiki Commons.

So, what do I think is the best hypothesis to explain all the data we’ve reviewed in the course of this series? An “exaptation hypothesis”. We’ve discussed exaptation in previous posts and I’ve already highlighted its importance for understanding the evolution of venom and indeed traits of all kinds. The MRCA of Toxicofera was probably not venomous, but it did have attributes that “exapted it” for the subsequent evolution of venom, namely its dental glands. These dental glands likely expressed a wide range of genes encoding proteins that, with minimal evolutionary tinkering, could be weaponised as toxins in venom. Expression of such potentially toxic proteins in oral glands (and hence their presence in oral secretions) is not the sole preserve of toxicoferan reptiles, in fact it’s a widely exhibited trait (see mice link). In toxicoferan reptiles, however, the fact that these secretions are deposited at the base of the teeth, as well as the fact that reptiles in general tend to kill their prey with their mouths, increased the likelihood of the evolution of venom within this clade. This process reached fever pitch in the snakes, which, having lost their limbs, became increasingly reliant on oral means of prey subjugation (though it turns out that constriction is a pretty neat trick for a long tube of muscle, too).

So, the “exaptation hypothesis of venom evolution” – still not formally or “falsifiably” defined, but this is a blog post, after all, not a peer-reviewed publication! In any case, another of Popper’s evolutionary principles of epistemology (the “theory of knowledge” - basically how we come to know stuff) is that hypotheses are “ill-formed at birth”. A “theory” only takes shape through repeated rounds of criticism – it is refined or refuted by these criticisms in a similar way to an organismal lineage becoming more adapted or going extinct (phenotypic refutation!) as the result of selection pressures exerted by its environment (including other organisms). Nonetheless, it should be relatively easy to think of the kinds of experimental evidence that would refute the exaptation hypothesis, at least as it pertains to Toxicofera (it’s actually much more general, and we will be looking at the role of exaptation in the evolution of toxins in a future post). For example, if evidence came to light that Toxicofera was not a clade, or that there was nothing “special” about their dental glands or any other synapomorphy of the group that could be reasonably linked to the evolution of venom, then the exaptation hypothesis would be refuted. Personally, even though I think the exaptation hypothesis best explains the data we currently have, I’d be very happy to see it criticised and ultimately refuted, because that would mean we’d moved another rung up the ladder of knowledge. Even if we didn’t have a better hypothesis yet, we’d still know just a tiny bit more than we do now, because we’d know which explanations don’t work. Having said that, sometimes refutation of a new hypothesis leads to (provisional, as always) acceptance of an old hypothesis! Even then, we know just a little bit more than we did before we formulated and tested the new hypothesis. That’s the game of science!

More Information

Dr Timothy Jackson