Occasional paper: The impossible predicament of the death newts

So a few days ago I posted about newts, and I mentioned that there was an American newt that was ridiculously toxic. But then (I said) there wasn’t space or time to go into why.  And of course I was immediately bombarded by many* comments and e-mails asking why. 

*three

Well, fine.  The world’s most toxic newt is Taricha granulosa, the Rough-Skinned Newt, a modest little amphibian native to the North American Pacific Northwest, west of the Cascades from around Santa Cruz, CA up to the Alaska Panhandle.

It’s so toxic that the poison from a single newt can easily kill several adult humans. You could literally die from licking this newt, just once.

(But note that the newt is toxic, not venomous. It doesn’t bite or sting. You could handle one safely, as long as you washed your hands thoroughly afterwards. Very, very thoroughly.)

Okay, but… why?  Lots of newts are mildly toxic.  Why is this particularly newt so extremely toxic?

Turns out this is a fairly deep rabbit hole!  I’ll try to teal deer it.

The simple version: the newt is in an arms race with the common garter snake, Thamnophis sirtalis.  The garter snake is a small-to-medium sized snake that is common all over North America.  It’s a slim, elegant little creature that is usually found in or near water.  And while it will eat pretty much anything it can catch, the garter snake particularly likes amphibians: small frogs, salamanders and newts.  And in the Pacific Northwest, it really likes snacking on the Rough-Skinned Newt.

So the garter snakes of the Pacific Northwest have been evolving resistance to tetrodotoxin.  (As you may recall, tetrodotoxin is the stuff that makes the blue-ringed octopus so deadly.  It’s produced by symbiotic bacteria that live in the newt’s body, mostly on its skin.)  And as the garter snakes evolve resistance, the newts have to evolve ever greater toxicity.  And as the newts get more toxic… right.  Feedback loop!  That’s the simple version.

Except this is biology, so of course it’s not that simple.

One thing to keep in mind is that nothing in nature is free. The newt’s toxicity comes with a cost: the metabolic load of supporting all those bacteria.  More toxicity means more bacteria means more load.  A very toxic newt has to consume more calories than its less-toxic cousin.

Meanwhile, evolving resistance also comes at a cost.  We don’t know that directly, but we can infer it pretty well.  If resistance to tetrodotoxin were cheap and easy, everything would evolve it. 

There have actually been attempts to measure the effect on the snakes!  They haven’t found them, but keep in mind that tetrodotoxin is a neurotoxin.  To resist it, you have to make changes to the biochemistry of your nervous system.  Even a small snake has a very very complex nervous system, where those changes might show up in ways that are hard to measure.  Like, if the resistant snakes were clumsier or had slower reflexes, sure, we could see that.  But maybe they’re suffering from much more subtle neurological effects, like being prone to insomnia or hallucinations or sexual dysfunction.  Or maybe they’re just a bit dim. 

We don’t know, but we’re pretty sure there must be something.   We know that garter snakes outside of the Pacific Northwest are much less resistant to tetrodotoxin.  They’ll drop dead from doses that their Oregon cousins simply ignore.  So evolving the resistance must have some cost or drawback.

(Note that this means the newt is double-whammied:  not only does it suffer the metabolic load of carrying around all those bacteria, but it also has to evolve resistance to tetrodotoxin, accepting whatever negative effects that brings.  And things are only going to get worse.)

Now: when the snakes eat Rough-Skinned Newts, they may sometimes show signs of discomfort. The snake may visibly gag.  It may writhe in obvious unease.  In some cases, it may go into respiratory distress.  Eating the newt looks pretty unpleasant.  Yet the snakes persist.

Okay then — if evolving toxin resistance carries a cost, and if even with resistance eating the newts is unpleasant, then why then do garter snakes insist on eating newts?  They cheerfully eat frogs, fish, and other non-toxic prey items.  Why don’t they just eat those, and leave the poor newts alone?

Turns out there is an answer:  the garter snakes sequester the tetrodotoxin, storing it in their livers.  This makes them toxic to their own predators.  (Of which there are plenty.  They’re not large snakes, so they’re hunted by everything from raccoons to ravens.)  But they don’t harbor the bacteria, so they don’t produce tetrodotoxin of their own. So eventually, the toxin that they’ve ingested breaks down.  And then they need to eat another newt to refresh their defense.

(If you are That Guy:  now you can name a snake that is poisonous without being venomous.)

So this explains why the snakes go after the newts particularly, preferring them to less toxic prey.  They want to eat toxic newts.  And it explains why the newts keep evolving to be more toxic: the snake may want to eat newts generally, but if an individual newt packs enough of a wallop, the snake may just retch it up and go after a different one.  Newts with weaker poison?  They get eaten.  Snakes with less resistance?  Have trouble finding newts they can choke down, and don’t get to steal their poison.  So the arms race continues.

This also explains why the newts, despite being very toxic to anything that’s not a garter snake, haven’t evolved aposematic coloring or signals.  “Aposematic” is just a fancy word for “warning”, meaning something — usually color — that tells the world Do Not Mess With Me:




[The Far Side by Gary Larsen, copyright 1984]

The Rough-Skinned Newt, like many newts, has a brightly colored underside that it can flash when threatened.  But the rest of it is a dull mottled grey, camouflage.  Presumably that’s because if the newt ever tried to evolve full-body colors like an Amazonian poison dart frog, it would just be a big “come eat me” sign to garter snakes.  Probably a certain number of newts get eaten by birds or fish or whatever anyway because they’re not aposematic enough.  The bird or fish may not survive the experience, but that’s not much comfort to the newt.

In sum: the unfortunate newt is not once, not twice, but three times screwed over here.  They have to be extra-toxic, carrying that metabolic load, just to maybe make the garter snakes think twice about eating them.  Then they have to evolve defenses against their own toxin.  But they can’t evolve aposematic coloring, because that’ll just lead to the snakes gobbling them all up.  And finally, they can’t go back to being not-very-toxic, because the snakes will just eat more of them to gain the same amount of tetrodotoxin.  They can’t win, they can’t break even, and they can’t leave the game.


[Copyright Gary Nafis, 2019]

There is even more going on here.  There are literally dozens of papers about this snake-newt interaction.  And we still haven’t reached the end of it!  Just a few examples:

— The newts that live in the northern end of their range, up in Alaska, don’t have garter snakes around.  As expected, they mostly aren’t very toxic.  But only mostly!  One paper found odd little pockets of unexpectedly toxic newts up there.  Why?  We don’t know.

— Meanwhile, the newts of Vancouver Island, offshore, also aren’t very toxic.  Because there are no snakes there?  Ha ha, no.  There are three different species of garter snake on Vancouver.  But  for some reason, the newts and snakes there seem to be living in harmony.  Well, relative harmony.  The snakes of Vancouver do eat newts.  But so far, the snakes and newts don’t seem to have started an arms race like they have on the mainland.  Why not?  We don’t know.

— A question that occurred, digging through these papers:  if the garter snakes are becoming toxic from eating newts, might the snake  begin to evolve aposematic coloring?  And sure enough, when I look at photos of Oregon garter snakes, a lot of them seem to have bright orange markings that look a bit aposematic:


[copyright Stephen Bol, 2023]

But on the other hand garter snake markings vary wildly.  Lots of garter snakes outside the Pacific Northwest have orange spots or stripes, and lots of snakes within the region don’t.  And as far as I can tell, nobody has even tried to research this yet.

— A thing to keep in mind: the Pacific Northwest, where all this is playing out, is a very young ecosystem in geological terms.  Just 20,000 years ago, during the last Ice Age, the Pacific coast of Washington and Oregon looked a lot like Greenland today: a thin coastal strip of cold tundra,  everything else covered by an ice cap.  The newts and the snakes are relatively recent colonists.  So we may be catching a snapshot of a relationship that is still evolving, and that may not be long-term stable.

— And finally, I have oversimplified this whole thing, because there are other species of newt in the genus Taricha that are pretty toxic — not as crazy deadly as their Rough-Skinned cousins, but still more toxic than normal newts — while the garter snake genus Thamnophis is just a taxonomic mess.  

In other words, we’ve learned a lot, but mysteries still abound. 

And that is probably enough about newts for now.