Why is snake venom so toxic?

Puff-adders probably cause more human snake-bites than any other African
snake, but are rarely fatal. This is a juvenile, but don't think it's harmless.

After discovering all the amazing things about pedarin and the 'Nairobi Eye' last week, it set me thinking again about why so much wildlife is so incredibly toxic. Think about it - a little beetle small enough to crawl over you without you noticing at all, is more than toxic enough to kill a grown man - indeed, several. A snake like a black mamba can give a bite that's sufficient in toxicity and volume to kill an adult elephant. Many natural venoms aren't simply one chemical, but a mixture of nasty toxins with a whole range of activities - why go to the trouble of evolving a whole suite of nasty chemicals, when one is usually enough to kill most things? Why should it be so toxic? What's the purpose?


This is a topic that's puzzled lots of people for quite a long time, and a number of different answers have been proposed with no one clear winner. The two main threads of answers have been proposed: firstly, that there's no effective natural selection on snake venom, once evolved, venom toxicity just drifts along. Others suggest exactly the opposite: that it must be a trait that is subject to extremely high selection pressures to have evolved as it has. Now, the argument for a neutral drift stems mainly from the observation that non-venomous snakes exist and, indeed thrive: if one species manages without venom, but another has it and both do just fine, it would seem clear that evolution has nothing to do with it. On the surface it's a pretty convincing idea, but I wonder too if there's some slightly dubious thinking going on here. Remember the thorn story: many plants here have thorns, but the animals eat them anyway, so why should they have evolved? I think the same sort of idea is happening here: we're looking at a situation that may seem daft today, but wasn't before.

Female Boomslang - you'll probably die painfully
in 2-5 days if not treated. Happily, they're rather shy.

So let's look at the evidence suggesting an evolutionary benefit to the extreme toxicity of venom. Let's start by thinking why snakes have venom in the first place. They use it for two things (1) to kill their prey and (2) as a defence against threats. Certainly, for a defence mechanism alone, you don't need to kill - in fact, you might even be better off not doing so and just giving a warning that can be passed on to other potential predators. So let's concentrate on what seems the most likely initial reason to evolve venom in the first place: to kill and subdue prey. Clearly if you evolve venom for one reason (attack), it's perfectly reasonable to later use it for self defence, but it's also possible the attack reason is still driving the whole process. Now, a venomous snake finding something tasty to eat strikes quickly and usually lets go immediately, waiting for the prey to die at a safe distance. Not a bad idea if you want to take down something larger or stronger than yourself. So it's important to that snake that whatever it's bitten dies relatively quickly: too slow and the animal might first have had the chance to run away and the poor snake goes hungry. So although there might be enough venom in a mamba bite to kill an elephant, it takes hours and hours - not fast enough if the snake wanted to actually eat the elephant. So if it wanted to hunt a mouse it might only need a drop of around 0.185 mg (average of various estimates) to have a 50% chance of killing that mouse, it would probably have no chance whatsoever of actually eating the think unless a bite killed within seconds, which requires a much larger dose.

Boomslang showing rear fangs typical of colubrid snakes.
Don't try this at home folks...

Moreover, we know there's a large cost involved in producing venom - snakes that have  recently bitten and need to re-fill their venom glands have a metabolic rate 11% higher than snakes with no such need. This also explains why some snakes when acting in self-defence don't always inject venom - the bite itself is warning enough. And we also know that some snakes do even adjust the amount of venom they use according to the mass of the animal they're biting. So there's obviously a cost of generating venom that means it would be strongly selected against, if there was really no need for it, yet there seems remarkably little evidence that snakes have lost the ability to produce venom, whilst the ability appears to have evolved in at least three different groups separately. Now, although there's certainly a cost involved in producing venom, it's not clear to me that the costs involved in producing a rather less toxic venom would be significantly lower than those of producing a much more toxic venom - once you've paid the initial cost, producing different variations on a theme might not be additionally costly. So if it's relatively easy to vary the basic molecules involved in toxicity (and it seems it is), it would make perfect sense to evolve a whole suite of chemicals, each with slightly different action making the overall combination toxic in the extreme. What's more, this also fits with what we know about resistance.

Just as bacteria can evolve resistance to our antibiotics, so too can resistance evolve in prey to snake venom. Now, resistance is most likely to evolve in bacteria when the doses aren't kept high enough for long enough to kill everything - that's why it's crucial to complete a course of antibiotics (if you don't, there might be some bacteria that were nearly resistant still alive when you stop, and you've just selected very strongly for those strains that are close to evolving proper resistance). It's surely the same with snake venom: it's much harder to evolve resistance to a massive dose of a toxic nasty, than it is to a smaller dose that some individuals might just survive. We also use multiple antibiotics when we're trying to protect against bacteria developing resistance (it's much harder to simultaneously evolve resistance to two antibiotics than just one), which seems like a plausible reason for why venom is full of a whole range of toxins too. Despite this, resistance can still evolve in prey animals, so the doses need to keep getting larger and larger - just like thorns do - with the result that when used in self defence against an animals that has no evolved immunity because it's so rare to get bitten, the dose is incredibly toxic. All good reasons for why evolution might well lead to what looks like 'overkill' in snake venom.

Tiger snakes are only the tiniest bit venomous (no human risk), and use both
constriction and venom to subdue birds, bats, and other retiles.

It seems to me, therefore, that there's plenty of evidence that the massive toxicity of snakes (and some scorpions and those nasty cone shells too, of course) has an evolutionary benefit, and not much evidence that its a random drift. So why do I think it's only a minority of snakes that are venomous, if there's such an advantage to being venomous? Well, I'm guessing a bit, because I can't find an answer at the moment, but my feeling is simply that species specialise in different prey. Constrictors have gone down one avenue for killing large prey, venomous snakes another - both work well for dealing with big or fast moving or otherwise dangerous prey items. Non-venomous snakes (which are, remember, the majority of species) that aren't constrictors tend to go for smaller things that won't fight back - invertebrates, small reptiles and amphibians, etc. Things that once you've got your teeth into, are unlikely to be able to inflict any damage on you, or simply escape. It's venomous snakes that are able to hit birds, bats and other larger mammals where a quick bite will almost instantly immobilise and kill your chosen prey. And then, once you've evolved toxic venom there's plenty of good reasons why you should make that venom as toxic as possible - especially if it doesn't cost much more than the initial stage of having any venom.

Now, what we haven't talked about is what it actually feels like to be bitten by one of these, and it so happens that I know there's at least one person involved the blog who could give us a first hand account if he were feeling inspired in the comments? [He's in good company, by the way - estimates suggest about 43,000 venomous snake-bites each year in East Sub-saharan Africa.]

PS a word of advice if you do get bitten by a venomous snake - if you can identify it, great, if not please don't got thrashing about after it to kill it or get a better look. Most snake-bites can be identified sufficiently from their toxic effects and many anti-venin these days can target a group of snakes. If you (or your friends) go thrashing around after a venomous snake that's already scared the most likely outcome is someone else getting bitten too which doesn't help things at all. Do get yourself somewhere with anti-venin as fast as possible though...

Main references:
Mebs, D. (2001). Toxicity in animals. Trends in evolution? Toxicon, 39 (1), 87-96 DOI: 10.1016/S0041-0101(00)00155-0


Barlow, A., Pook, C., Harrison, R., & Wuster, W. (2009). Coevolution of diet and prey-specific venom activity supports the role of selection in snake venom evolution Proceedings of the Royal Society B: Biological Sciences, 276 (1666), 2443-2449 DOI: 10.1098/rspb.2009.0048