This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
The largest animal behaviour conference in the UK (the Association for the Study of Animal Behaviour) was held last Thursday and Friday in my hometown of St Andrews. The theme of the conference was ‘Understanding Animal Intelligence’, encompassing animals from chimpanzees and humans, to bees and scrub jays. There was a tremendous amount of exciting new research being discussed, and I thought I’d share some snippets of it here.
Alex Thornton from Cambridge gave an entertaining talk on meerkat behaviour. Meerkats are unusual in that they go out of their way to teach each other things, something rarely seen in non-human animals. He argued that even though we often see social learning (learning from other individuals) as ‘intelligent’, it doesn’t necessarily have to be.
The most commonplace teaching in meerkats is from parents to young, where they often disable prey for young to learn how to kill themselves. As the young learn to kill the prey themselves, the parents will bring prey that are more able to fend for themselves, until the young are killing them entirely on their own.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Alex Thornton with the meerkats:
[youtube=https://www.youtube.com/watch?v=wMjalAElFXw]
But how do the parents know when the young are old enough to fend for themselves? An experiment showed that when the scientists played calls of older meerkats, the parents brought the pups prey that were less modified than when they heard the calls of younger pups.
Alex Thornton also discussed another experiment where meerkats were trained on two different ways to solve a task to get some food. Members of the meerkat group who watched a particular individual solving the task one way would then learn to solve it that way themselves. Thus it seems that these animals are able to learn from one another socially. However, it also seems that the meerkats do not necessarily ‘understand’ the tasks they are solving in the same way that a human might, nor do they use their knowledge (in the set-up tested) and apply it to a new situation.
Elisabetta Visalberghi gave a talk about a lifetime’s work on capuchin monkeys and their tool-using abilities. Capuchin monkeys use stones to break open nuts:
https://www.youtube.com/watch?v=wMjalAElFXw
A series of experiments in her lab showed that the capuchins were able to choose both appropriate tools, and appropriate places to crack the nuts on.
Elva Robinson gave a great talk on ants and how they make collective decisions about the best site to use. The experiments she discussed used a methodology which definitely left a lot of people talking about it afterwards: the ants were fitted with minute chips on their backs which could be scanned by a machine as they walked under it. This not only meant that individual ants could be identified by a simple scan, it also allowed for a particular experimental design which revealed how these creatures choose nest sites. By using these radio-frequency identification-tagged ants, the scientist could control which ants walked where. Some ants were allowed to check out a particular ‘good’ nest site, whereas other ants, which were not on the ‘guest list’ were scanned, and then a tiny door descended, denying them entry. The study found that a decision by the whole ant colony could emerge without individuals having to experience both the good and bad nest sites themselves: as long as some individuals experienced one of the nest sites, and other individuals the other, a group consensus could be made.
One of the definite highlights for me was a talk by Tom Seeley about swarm intelligence in bees. Swarm intelligence is the idea that multiple individuals can solve a task that one individual could not solve by itself. Perhaps the earliest example of swarm intelligence in humans was documented by Francis Galton in 1907, who wrote about a contest where 800 people had to guess the weight of an ox. The median of all their predictions was remarkably close to the actual weight. More modern examples would be Google and Wikipedia.
Seeley’s honeybees live in groups of around 10,000 worker bees, 300 to 500 of which are ‘scouts’. These scouts leave the group and go out searching for new places for the colony to live. The ‘dream home’ for them is a cavity high up in a living tree, large enough to hold the whole colony. When a scout bee discovers somewhere that seems appropriate, it comes back to the colony and does the ‘waggle dance‘ to demonstrate this. You may already know that this waggle dance of the scout bee tells the other bees in what direction and how far away the new potential home is. However, it also conveys to the other bees how good a location the scout bee thinks it is. This means that the scout bee will do a different dance if she thinks that the location is a top location to what she would do if it is mediocre.
Intelligent animals? Mark Biancaniello, who attracted 39.5 kilos (87lb) of bees onto his body for a bee-wearing contest.
However, there are hundreds of scouts, all exploring the world and returning to the colony with different pieces of information. This is where the swarm intelligence comes in. As multiple scouts return to the colony and share their information, there is a kind of ‘debate’ as the different ‘ideas’ are put on the table. It is not until a consensus is reached that the colony will move to the new spot. Through a series of experiments, Tom Seeley and his group have come a long way in understanding how such a consensus is reached. They set up a bee colony and placed various potential bee homes around (1 ‘dream home’ and 4 mediocre ones). In 9 out of 10 trials, the bees were able to choose the best one, and move the colony to it.
Through watching the colony, the scientists saw that when the scout bees sampled the box which was of higher quality, they did more circuits of their dance than when the potential new home was of lower quality. The other scouts payed attention to the dance, and checked out the box for themselves. Once a certain number of scouts ‘agree’ that it is a good place to live, and are all doing the dance, the colony will make a decision to move.
Now for the really cool bit: the scouts which are all rooting for a particular site, say the best box, will try and stop the scouts dancing for the other site. They interfere with their dance through a subtle signal: hitting the dancer. This inhibitory signal is an audible ‘beep’ as they hit into the dancing bee. It stops the dancer after somewhere between 5 and 25 ‘beeps’. In return, the bees which are rooting for that site will be trying to inhibit the bees dancing for the better site. Thus, through both excitatory (dancing) and inhibitory (‘beeping’) signals, a quorum develops, and the bees decide which location to move to.
Tom Seeley drew an interesting parallel at the end of the talk between this method of swarm intelligence and the brain. The bees, like individual neurons in a brain, combine together lots of small bits of information (excitatory and inhibitory) to come to a collective ‘decision’.
All in all, it was a stimulating and inspiring conference, and left me with a lot to think about.