Scientists have been able to teach bees to do all kinds of things, like tug on a rope or count to four. But so far these skills have largely fallen into the realm of what scientists might expect is possible. So to really put bees to the test, and figure out exactly how much they can learn, researchers recently taught them a skill they would never need in their natural environment: playing 'soccer.'
"This was an attempt to kind of test the limits of bees' cognitive capacity," said Clint J. Perry, a cognitive neuroethologist (he studies animal brains) at Queen Mary University London, who worked on the experiment. "There are no flowers that exist that require a bee to bring a ball into it in order to get a reward."
The results were published in a study in Science Thursday. In the experiment, bumblebees learned to manipulate a little ball into a "goal," and when they did it, they'd get a reward: sugar water. Their behavior was compared to a control group of bees that did not get any training, and the difference was night and day, according to the study. When taught how to move the ball, and that it would trigger a reward, after just a few demonstrations the bees would start to solve the puzzle on their own. Control bees never quite figured it out:
Overall, the bees picked up the skill effortlessly, and even showed some ingenuity in how they solved the puzzle—something that's never been shown in insects before. It's a remarkable demonstration of how powerful a bee's sesame-seed-sized brain really is, and how incredible these important pollinators are.
Perry and his colleagues tested out a few different ways of teaching the bees. First, they used a really simple design (just one ball on a flat surface with a center "goal") and used a small puppet bee to demonstrate the desired behavior. After 10 to 20 demonstrations, all bees that watched this technique were able to perform the behavior to get the reward.
"But we wanted to know what elements of the demonstrations the bees were paying attention to," Perry told me. "Was it social? Was it the movement of the ball? Maybe they didn't care about the model bee. I mean it looked like a bee to us, but it wouldn't fool us, so was it fooling them?"
To test this, the researchers did two more versions of the experiment, this time using a more complicated stage with three balls. In one experiment, they had a fellow live bumblebee that had already been trained to solve the puzzle demonstrate for a naive bee. In another experiment, the researchers used a magnet to manipulate the ball from underneath the platform, kind of like a ghost bee was moving it, to try to teach the naive bees the trick.
The bees that watched a live demonstration had more successful trials and took less time to solve the puzzle than the ones with the "ghost" demonstrations, which shows that having a fellow bee teach them is a more effective way of learning. The researchers also tested if the bees were simply copying behavior or really putting two and two together. They had all the demonstrators (whether live bee or "ghost") always use the ball furthest away from the target when training, but noticed that the student bees would often go for the closest ball when it was their turn.
"If the bees were simply copying what they observed, they should have moved the furthest ball, but in almost every instance they moved the closest ball," Perry said. "This shows a behavioral flexibility that is unprecedented in insects."
Get six of our favorite Motherboard stories every day by signing up for our newsletter .