2007 | Shark Bay, Australia
Bottlenose dolphin (Tursiops sp.) shell carrying
On the last day of July, in 2007, Simon Allen and his colleagues watched from a boat as a female dolphin named CON poked her head out of the water. They were patrolling the western part of Shark Bay - the kind of place name that you’re probably not surprised to hear is in Western Australia.
Seeing dolphins was nothing new to the team. At that time, around 15,000 scientific dolphin sightings had been made in Shark Bay, one of the world’s leading locations for studying dolphin behaviour. What interested them was the large conch shell seemingly stuck on the end of the dolphin’s beak, or rostrum. That was unusual. As they photographed the scene, the dolphin shook its head up, down and sideways, while the shell remained fixed in place.
Was the mammal somehow caught in the shell? Was she signalling to other dolphins, or playing, or eating the large mollusc that builds and lives in such shells? No. She was using a tool to eat a fish.
A draining experience
When Allen and his team reported their discovery a few years later, in 2011, they’d gone back through their records. They found at least seven occasions when a dolphin had been seen waving around a big shell (including one in 2009, after their experience with CON). No one really knew whether those events were connected, or what has happening. But critically, those photographs from 2007 held the key. Here they are:
Did you spot the evidence? CON’s body is underwater, off to the left of the pictures, and the top of her head towards us. Her rostrum is well and truly stuck into the shell—of the genus Turbinella—as she agitates it. In the first image there’s something sticking diagonally down from where shell and dolphin meet. And in the second image it’s clearer: that’s a fish tail poking downwards, likely a member of the Lethrinus or emperor family.
The dolphin has brought the shell, with its fishy package included, up to the surface to drain the water out and get a tasty snack. It’s a clever use of the density difference between air and water, essentially using gravity to turn a fish’s watery hiding place into its own airy tomb.
Allen et al. named the behaviour they’d identified ‘conching’, in a parallel to the already known activity of ‘sponging’, in which a Shark Bay dolphin holds a marine sponge over its beak while foraging on the sandy seafloor. They didn’t explicitly discuss whether this was tool use, but they did note that it was a rare and innovative action. Whether it was repeatedly re-invented by the dolphins each time it had been seen, or was learned by calves from their mother (as sponging is), remained a mystery.
Mates rates
Fast forward a decade, and we have the answer. To find it, Allen and Michael Krützen (who was co-author on the earlier study) teamed up with Sonja Wild, who was researching dolphin innovation as part of her PhD at the University of Leeds, and William Hoppitt, and expert in animal networks.
In their report from 2020, Wild et al. had several more years of dolphin-watching data. They now knew of 42 conching events, performed by 19 different dolphins belonging to three different family groups. They also re-named conching to ‘shelling’, which allows for the possible use of large shells that aren’t specifically conches (baler shells, for example).
The aim of the new study was to work out exactly how the dolphins were learning this trick. Specifically, was it something that was prompted by the local environment (shallow seagrass beds, for instance), was it socially learned, or was it a series of independent inventions, also known as asocial learning? Wild and her colleagues calculated that they needed at least 11 sightings for a given dolphin to know which other dolphins they were hanging out with. That left them with a healthy sample of 310 individual animals, of which 15 were known shellers.
Network analysis compares each dolphin with a series of variables to determine which one(s) most likely contributed to the behaviour. The researchers constructed networks of which dolphins associated together—basically, their friend and extended family group—as well as which ones frequented specific parts of Shark Bay as a way of looking for environmental influence. Unusually for wild animal studies, the team also had a genetic library built up over several decades, allowing them to assign each dolphin to its specific lineage or haplotype.
By running each of these three networks through computer models, Wild found that only one scenario made sense: social transmission among the ‘friends and others’ group. The majority of the dolphins weren’t learning the activity from their mothers, and it wasn’t just the case that the shelling animals were hanging out in specific locales that happened to have the right conditions. In the words of the researchers:
social transmission includes transmission among associated individuals, which could be both horizontal and oblique but will be referred to as ‘‘non-vertical transmission’’ hereafter, as in, any social learning outside of the mother-calf bond.
This network-based diffusion analysis or NBDA allowed the team to go further, and estimate that around 57% of the shelling dolphins learned it from another animal, but not their mother. Their best model found that the other 43% likely innovated this activity themselves (asocial learning), although they couldn't rule out those dolphins having seen it done when the researchers weren’t floating nearby with their notebooks. They couldn’t identify exactly who started the craze, but one thing they were fairly sure of was that the western and eastern parts of Shark Bay had at least one independent shelling innovation each.
Nowhere to hide
Wild, Allen and their colleagues considered why the dolphins were finding fish inside giant shells in the first place. They decided the most plausible scenario is that the fish become trapped while sheltering in empty shells, either as general protection or (a slightly more sinister notion) because the dolphins were themselves chasing fish into those hiding places. They produced an animation outlining that idea, from which I’ve assembled this diagram:
The 2020 team was also more confident than in the 2011 study in referring to shellers as tool users. They found the activity to have all the necessary hallmarks, including manipulating an object to change the position of another organism:
Dolphins utilize the shell to first entrap the prey before exposing it to gravity by lifting the shell above the water surface and shaking it about in order to dislodge the prey, thereby fulfilling these conditions and qualifying shelling as tool use.
That makes shelling only the second form of wild dolphin tool use, alongside sponging. Even if the dolphins aren’t actively springing the shelly trap by herding the fish, the subsequent transport of the fish-in-a-shell might be similar to when we pick up a bag of potatoes at the supermarket. We didn’t put the potatoes in the bag, but we’re still using it as a tool to move them around.
A final thought from the Shark Bay dolpinologists covers when it might be more helpful to learn from your mother, and when learning from others is beneficial. Dolphin calves spend a lot of time with their mother, which gives them useful information about where and how to find food and behave in their mother’s home range. When the environment is stable, that information will be consistently valuable over time, and it’s worth copying. But in a period of instability, continuing to act on the previous generation’s information can become outdated, even counterproductive. In uncertain times, expanding your sources of knowledge and focusing on what works for others in the moment increases the chance that you’ll copy an appropriate strategy.
In 2011 a marine heatwave led a lot of the seagrass in Shark Bay to die back, possibly causing significantly more of the large gastropods that graze among that grass to die off and leave their empty shells behind. In turn, that could have changed the ways and places that local fish hide from marauding dolphins. The scientists did notice a steep increase in shelling in the two years after that heatwave, but they couldn’t say for certain that the environmental crisis directly prompted a shift to friend-copying behaviour. Unfortunately, though, it likely offers a preview of how these large-brained marine mammals will need to adjust to the continued ecological turmoil we’re inflicting on them and their habitats.
Further viewing. This summary of the Wild et al. study includes the shelling animation, as well as helpfully explaining network based diffusion analysis over a jaunty guitar track (3 minutes):
Sources: Allen, S. et al. (2011) Why do Indo-Pacific bottlenose dolphins (Tursiops sp.) carry conch shells (Turbinella sp.) in Shark Bay, Western Australia? Marine Mammal Science 27: 449–454. || Wild, S. et al. (2020) Integrating Genetic, Environmental, and Social Networks to Reveal Transmission Pathways of a Dolphin Foraging Innovation. Current Biology doi: 10.1016/j.cub.2020.05.069.
Main image credit: Wild et al. (2020) || Second image: Allen et al. (2011) || Third and fourth images: Wild et al. (2020) || Video credit: Cell Press, 2020.