2004 | Lake Solano County Park, California

Ground squirrel (Spermophilus beecheyi) rattlesnake deterrence

In my last post, I made the (perhaps unreasonable) argument that the stomach stones of dinosaurs should be considered ancient tools. Not having learned from that, today I’m continuing to explore the borders of what might be considered tool-use, by taking a brief detour into the hinterland of chemical camouflage. Specifically, we’ll be looking at a clever squirrel that dresses itself up as a smelly snake.

Rattled

Our furry focus is on the Californian ground squirrel (Spermophilus beecheyi). They live in natural habitats that overlap with predatory rattlesnakes (Crotalus spp.), although they are far from passive prey. Over millions of years the squirrels have adapted—not only are they somewhat resistant to the venom of their neighbourhood rattlesnakes, but they will even actively and attack harass those snakes if needs be. California ground squirrels even add an infrared tail display to their harassment, which acts as a warning signal to the infrared-sensitive snakes.

We’re concentrating today on a 2008 US study by Barbara Clucas, Donald Owings and Matthew Rowe. Clucas and her team had noted reports of some ground squirrels adding yet another step to their set of snake responses. These squirrels were getting hold of shed rattlesnake skin, then methodically chewing on it and smearing the resultant mush—along with its rattlesnake scent—on their fur. Scent-application is unusual (although not unheard of) behaviour among rodents, but rarely has anyone systematically tested just what effect this olfactory armour actually has.

The scientists had several hypotheses going in. Perhaps the scent was a kind of camouflage, preventing the predator from detecting the prey, or maybe it stopped confused snakes from entering the squirrel’s burrow. Or the smell could be targeted not at the snakes but at other squirrels, alerting them to a snake hazard. It could even be a kind of insect repellent, mirroring the way that many birds rub crushed ants and other bugs on their feathers (although this ‘anting’ behaviour in birds also needs more study on its actual outcomes).

Here’s a two-minute video from PBS that shows the squirrel behaviour, with suitably dramatic shots of rattlesnakes lurking and licking and being all snake-y:

Clucas and her colleagues gave us the science behind that video. To narrow down their options, the researchers first had to test how much attention rattlesnakes even paid to squirrel scent. Without that evidence, all other smell-based arguments would have been missing the point. Wild-caught Pacific snakes (C. oreganus) were therefore exposed to one of three options: filter paper that had been rubbed on juvenile ground squirrels, filter paper that had been left with rattlesnake skin in a bag, and a mixed option with both squirrel and snake scents (replicating the squirrel-plus-snake scenario they were most interested in). The results were clear. The snakes spent more time and flicked their tongue more often near the squirrel-only scent than either of the other two.

This critical first step showed that the snakes definitely had an interest in smelly squirrels, and were less concerned with squirrel/snake hybrids. Next step was to test mammal vs reptile in a real world scenario. And so selected wild ground squirrels at Lake Solano County Park in California were given rattlesnake skins near their burrows, while the researchers sat back and watched. As a control, the same experiment was tried with rock squirrels (Spermophilus variegatus) in New Mexico.

The rock squirrels didn’t change their behaviour, whether or not they’d had access to snake skin. The same went for ground squirrels with no rattlesnake skin to chew on. But the skin-supplied ground squirrels were a different story. Those animals groomed themselves more and rested less after they’d applied the snake scent, which the researchers suggest could be the result of raising their own alertness level—if the scent of a predator follows you everywhere (even if you’re the one who put it there) then it heightens your natural awareness.

The smelly ground squirrels didn’t threaten snakes more or change how they acted around their furry family. It seems they weren’t getting dressed for battle, or putting on snake perfume as a way of alerting other squirrels to a possible threat. The most likely explanations remaining are that the snake-plus-squirrel scent is indeed a kind of camouflage, reducing snake interest, or that the critters are just keeping themselves alert.

But wait, wasn’t there also something about possible insect repellent? Don’t worry, the scientists didn't forget, although the squirrels might wish they had. The team collected 56 fleas from wild squirrels, again at Lake Solano County Park, then placed them in the centre of a tiny arena. On one side was a regular ground squirrel with added water, on the other a snake-scent-enhanced version. The flea got to choose which it preferred, and signal that preference by jumping the 15cm gap onto the hapless victim. The simple accompanying diagram sets up the tension nicely I think:

Still, after all that, it turns out the fleas don’t care. They went for the two squirrels equally, with no difference in how long it took them to make the choice either. If it’s a repellent, the snake scent isn’t for fleas. Clucas and her team didn’t report if they’ll be trying other creatures in the arena in future, although they do note that snake repellent/camouflage in theory might be targeted at other untested predators, like weasels or coyotes. They’ll need a bigger arena—a 15cm head start doesn’t seem enough if a coyote’s on your furry tail.

Nosy tools

Snakeskin anointing changes both snake and squirrel behaviour, so there’s no doubt that it has an effect. But is it tool-use? This is the bit where I get to make my latest tenuous argument. Although we might think of smells as invisible and therefore somehow non-physical, each one is very much made up of complex physical molecules. Those molecules travel via normal, non-magical means from whatever releases them to an animal’s scent-detection organ. In this way, I see them as not that different from a thrown spear, or even a rock dropped by a vulture onto an unsuspecting egg (or walrus). The fact that smells travel in unpredictable directions doesn’t change the fact that they are composed of effective, physical objects.

It’s debatable whether the squirrels know that the scent helps them avoid death by rattlesnake, or are aware that it juices their alertness like a strong espresso. In fact, there’s no evidence that they intend those consequences. And one of the problems with tool use definitions is that—because they started with humans and only then spread out into the wider animal kingdom—intention is given a lot of weight. If you don’t mean to use a tool, then are you doing so?

Personally, I’m not great at inferring intentions in other humans, let alone squirrels or crabs or elephants. My main research work on humans has involved people who died somewhere from decades to millennia ago, and from experience I know that inferring intent from archaeological remains can be just as difficult as trying to read the mind of a living, tool-using, wild crow. So my bias is to leave intent aside as much as possible, and instead focus on observable effects. Specifically, does the animal manipulate some moveable part of its environment to extend its natural abilities? In the case of ground squirrels and snake smells, I’m opting for yes.

Note that I’m not opening the door to a scent revolution. I’m not claiming that plants that produce chemicals to attract or repel insects and herbivores are tool users, because it’s part of their natural capacities. A rose planted in any suitable garden will smell rosy, and not all animal smells count either: skunks will do what skunks do with their own internally-made scents. However, a ground squirrel without a rattlesnake skin won’t produce that snaky smell, no matter how useful for survival it may be.

In the human world we can find parallel cases, such as car air-fresheners, the use of scented air in grocery stores to encourage buying, or the smell of baked goods in an open house to raise the attractiveness of a property. But that’s really another story (spoiler: I think those human-manipulated smells are tool use too). I expect there’ll be a post in the near future trying to flesh out this definition. For now, let’s just be glad that our fluffy squirrel friends have this extra trick against their slithering foes, whether they know it or not.

Sources: Clucas, B. et al. (2008) Donning your enemy’s cloak: ground squirrels exploit rattlesnake scent to reduce predation risk. Proceedings of the Royal Society B 275:847–852. || Bush, S. &, D. Clayton (2018) Anti-parasite behaviour of birds. Philosophical Transactions of the Royal Society B 373:20170196.

Main and second image credits: Cindy McIntyre, https://cindymcintyre.wordpress.com/2014/07/30/ground-squirrel-eats-snakeskin-true-story || Video credit: Nature on PBS; https://www.youtube.com/watch?v=U-tZiFZXI_0 || Third image credit: Clucas et al. (2008) Figure S1 || Fourth image credit: Connor Long, via Wikipedia; https://en.wikipedia.org/wiki/Crotalus_oreganus