Bees trained to score goals!

Bees have been trained to kick a ball around, displaying their surprising capability to learn.
28 February 2017

Interview with 

Clint Perry, Queen Mary, University of London


Bee dragging a ball


You may have heard of five a side, but what about hive-a-side? Using a sugar-water treat, scientists have discovered that they can train bees to kick a ball about! Ricky Nathvani spoke with Clint Perry to find out why…

Clint - One of the biggest findings is that bees were able to learn this very unnatural task. Normally bees, in nature, they move into flowers. They don’t really manipulate them to any complex degree, they just push forward into flowers to find the nectar or the pollen. Here we train them to actually roll a ball into a specific region to gain a reward - access to sugar water.

They were able to learn this socially. In the first experiment we used a plastic dummy bee to push the ball into the centre and have the bees watch and observe this demonstration. They were able to pick it up - all of the bees trained were able to pick this up. So we were then interested in what about this social learning was important for them?

So in a second experiment we trained them in a variety of manners. One they observed a live bee demonstrating how to roll the ball in the home to gain reward and in another situation we used a magnet underneath the platform to move the ball, so they didn’t see anyone or anything moving the ball, just the ball moving itself.

The bees who saw the live demonstration solved the task much more quickly than bees who saw just the magnet moved ball, but just the ball moving on it’s own was enough for the bees to learn.

Ricky - But actually, they pick up on that a lot faster if they see another bee doing that. Is that correct?

Clint - That’s right. And this was the purpose of the second experiment. Bees didn’t simply copy what they saw but they actually improved upon what they observed and the strategy that they saw.

So in experiment two, the live demonstrators always moved the furthest ball from the centre into the centre. They had three possibilities: one that was seven centimetres away, one that was five and one that was two, and the live demonstrators always moved the furthest ball. That’s what the observers saw each time during the training sessions, but during the tests, the observers moved the closest ball into the centre and, therefore, improved upon the strategy that they saw and didn’t simply copy what they observed.

Ricky - Wow! That’s really quite un-bee-leivable… sorry that was terrible. What I was going to say is it is quite impressive, all jokes aside, that you guys didn’t demonstrate this to the bees at all. They kind of naturally worked out that they could move one of the closer balls in and have a much more efficient solution to this problem.

Clint - Right. And an important point to that is that we ran controls to make sure they weren't using simpler mechanisms. One being just knowing that they weren’t paying attention to the position of the ball, they moved the closest ball into the centre. But also we changed the colour of the closest ball to black instead of yellow, which is what they saw during the training sessions and they still moved that black ball (the closest ball) in the vast majority of tests.

Ricky - And this is the first time that that kind of behaviour has been studied extensively?

Clint - Yes, as far as we know yes.

Ricky - So bees naturally, compared to loads of other animals that we’ve observed, have tiny brains. Would you necessarily expect bees to be capable of this kind of thing and what’s the significance of studying this kind of behaviour in bees, as opposed to larger animals?

Clint - Most of us, when we see bees or insects of any type, we look at those as genetically preprogrammed, unthinking machines and, often times as pests and what not, but there’s no real behaviour that’s been shown to require a large brain. Decades of research has shown that bees and other insects can learn, can solve complex tasks and can navigate complex environments.

I guess what’s important to note is why we’re studying this in insects? It’s the combination of bees having cognitive abilities as well as very small brains in order to access and to study. So we can record from and study individual neurons within the bee brain and, at the same time, looking at the entire brain. Whereas for larger animals, there are just so many neurons and so much stuff there that it’s hard to get at with the tools we have available today.


Add a comment