Banded mongoose (Mungos mungo) millipede throwing

At what point, when you’re struggling to open a jar, do you give into frustration and just hurl it against the nearest hard surface? For the banded mongoose, it’s a fairly low threshold. Today, we’re taking a look at a behaviour that isn’t tool use under the strict definition, but it can certainly leave its mark on the landscape. And we’ll even see how those marks might help solve a 40 year old otter mystery in Zimbabwe.

A day at the zoo

We start with what I’m going to assume were a couple of bored New Yorkers, back in the mid 1960s. Thomas Eisner of Cornell University, and Joseph Davis of New York Zoo in the Bronx, had got hold of a small batch of giant African millipedes. These arthropods were from the genus Sphaerotherium, and there’s a clue in that name. In Latin it literally means ‘ball animal’, and it comes from the habit of these huge insects to curl up into a tight sphere when threatened.

Eisner and Davis knew that this tactic—shared with unrelated mammals such as armadillos and pangolins—was very likely defensive, but they decided that the theory needed testing. So they did what anyone with access to a zoo might do, and they decided to see which animals could break into a millipede ball. The needs of the millipedes were not, I’m sorry to say, part of their consideration.

Here’s a Sphaerotherium giganteum, one of the species that made the trip to the Bronx. Curled up, it’s a bit smaller than a ping pong ball. The second millepede species joining the experiment was its cousin, S. punctulatum. In all, nine millipedes ‘participated’ in this work.

The experiment started small. Eisner and Davis offered the tightly-wound millipedes to a colony of Florida harvester ants (Pogonomyrmex badius). The ants crawled all over them, attempting to bite or sting, but to no avail. The defences held.

So the zoologists upped the stakes. Next was a North American blue jay (Cyanocitta cristata). The bird pecked at the millepedes, but the only result was that its beak slipped off their shells, causing the arthropods to roll away.

Ok, let’s get mammalian, thought Eisner and Davis. What about a mouse? Another win for the millipedes: a southern US grasshopper mouse (Onychomys torridus) failed to make any impact. Even though grasshopper mice eat large insects such as cockroaches, their sharp teeth just couldn’t get purchase on the ball animals.

Which is when things took a turn for the worse for the nine Sphaerotherium. Because they met the banded mongoose.

This social species (Mungos mungo) is native to parts of central and southern Africa. Their common name comes from the distinctive stripes running across their backs, as you can see in the photo above. Their closest living relatives are the African meerkats, with whom they share an ability to stand up on their hind legs. Under natural conditions, the mongooses (yes that’s the plural) regularly pick up and throw down encased food, repeating this action over and over until they get what they want.

There are two common mongoose throwing actions. In the first they stand up and then, using their front paws, forcefully hurl the object onto the ground in front of them. In the second, however, they skilfully position their back against a rock or other hard surface, then throw the food backwards between their legs. That’s the image at the top of this post, a banded mongoose throwing a millipede backwards onto a rock at the New York Zoo. A second image from the paper shows the pre-throw stance:

Every one of the nine millipedes broke under pressure from the mongooses. As Eisner and Davis describe:

The predator responded instantly to the glomerid [millipede], sniffing it, and rolling it about with the paws. It seized it in the jaws, biting upon it with its sharp teeth, but the millipede was neither pierced nor crushed. Suddenly, the millipede was dropped from the jaws and grasped in the front paws.

The mongoose backed against a rocky ledge in the cage, assumed a partially erect stance, and—with a motion so quick as to be barely perceptible—hurled the millipede backward between its legs, smashing it against the rocks. Fatally injured, with its shell broken and its body torn apart, the millipede was promptly eaten.

The zoologists were satisfied. And out of millipedes. But just to be sure, they tried hazelnuts and various snails, and found that the mongooses were content to smash those too. You can also see a video at the end of this post of captive mongooses using the backwards throw technique on some colourful Easter eggs.

As Eisner and Davis reported, the natural African range of these millipedes overlaps with that of the banded mongoose, so it could be that they’ve had the time to evolve this specific attack pattern. But given that it’s not just the Sphaerotherium that triggers their throwing instinct, it would be premature to directly tie the two animals’ behaviour to each other.

What we do know is that banded mongoose feeding habits are learned early in life, in a manner resembling human culture. Young banded mongoose pups are reared by adults in their group that are not their parents (known as ‘escorts’), and what they learn from their escorts affects how they forage throughout their lifetime. It’s a case of nurture beating nature—or at least genetics. Whether those lessons include anti-millipede violence isn’t clear though.

An-otter matter

That’s it for the banded mongoose, but let’s finish up with the Zimbabwean mystery I mentioned back at the start. Less than a decade after Eisner and Davis told the world about their zoo trip, two other zoologists (R. Donnelly & J. Grobler) reported finding piles of broken freshwater mussel shells around rocks and logs in what was then Rhodesia. The rocks were exposed by low water levels, which in my mind recalls the tidal exposure of mussel-cracking rocks used by Californian sea otters that I’ve studied.

Donnelly and Grobler decided, based partly on tracks in the mud near these debris piles, that local Cape clawless otters (Aonyx capensis) were responsible for using the anvils. They didn’t actually see the animals though. And, while their conclusion has been cited as a clawless otter trait by other scientists in the forty years since, to this day no-one has seen the otters at work. Even during intensive field observations, the otters just aren’t using anvils to break mussels. Their teeth do the job just fine.

Finally, in 2018, otter specialist David Rowe-Rowe published a short note that made some sense of the situation: the otters weren’t the only locals that might want a mussel meal. The water mongoose (Atilax paludinosus) also lived in the area, and their tracks were even found at the same sites at which Donnelly and Grobler reported otter footprints.

Like it’s banded brethren, the water mongoose is happy to pick up and throw anything that gives it a bit of feeding trouble. Eggs, snails, mussels, you name it. Critically, this includes using anvils. Here’s one caught in the act, photographed by Rowe-Rowe:

The scientific record, including the animal behaviour record, often makes only slow progress. But in general it takes us towards a more accurate picture of the world we live in. Maybe it takes forty years to realise that it was destructive water mongooses, not clawless otters, that were littering the Zimbabwean landscape. And maybe nine millipedes get sacrificed in a mid-60s scientific game of ‘which animal would win’. Just as long as we keep recording what we see, and sharing it with the world, we’ll never run out of odd and wondrous things to learn.

Further viewing. Here’s the video I mentioned, of captive mongooses at Denver Zoo breaking Easter eggs. It’s from 2008, and if you need further proof of scientific progress, just compare the resolution of the camera from just over a decade ago with what we’re used to today…

Sources: Eisner, T. & J. Davis (1967) Mongoose Throwing and Smashing Millipedes. Science 155: 577-579. || Armstrong, A. & M. Hamer (2015) English names of the millipedes (Diplopoda) of KwaZulu-Natal. African Invertebrates 56: 147-159. || Sheppard, C. et al. (2018) Decoupling of Genetic and Cultural Inheritance in a Wild Mammal. Current Biology 28: 1846–1850. || Donnelly, R. & J. Grobler (1976). Notes on food and anvil using behaviour by the Cape clawless otter, Aonyx capensis, in the Rhodes Matopos National Park, Rhodesia. Arnoldia 7: 1-8. || Rowe-Rowe, D. (2018) Anvil-use by Aonyx capensis: a rebuttal. IUCN Otter Specialist Group Bulletin 35: 62-65.

Main image credit: Eisner & Davis (1967) || Millipede credits: Armstrong & Hamer (2015); http://portugal.inaturalist.org/taxa/426760-Sphaerotherium || Fourth image credit: Mike Unwin, https://travelafricamag.com/bands-on-the-run/ || Water mongoose credit: Rowe-Rowe (2018)