Bumblebees show advanced problem-solving skills in new experiment

A bumblebee alights on the bloom of a thistle in Berlin, Vt. on July 5, 2011. (AP / Toby Talbot)

German psychologist Wolfgang Köhler set up a famous experiment more than 100 years ago that changed how scientists understand animal intelligence and the power of insight — or spontaneous problem-solving.

Köhler made what he described as a playground for a group of chimpanzees with a banana hanging out of reach and various items — boxes, poles and sticks — lying around. The strewn objects offered opportunities for the animals to explore, and the food presented a challenge for them to unlock. After fruitlessly trying to snatch the banana, the chimps quickly started rearranging the items. The apes eventually stacked the boxes and easily grabbed the reward.

The experiment demonstrated that chimps were capable of insight. While most animals can do basic problem-solving, insight is a step up because it’s an understanding of cause and effect that does not rely on trial and error, copying others, or previous knowledge. Scientists have observed this cognitive ability in only a handful of species: great apes, elephants and some birds. There is an ongoing scientific debate over whether even more species — invertebrates such as octopuses and certain spiders — should also join the ranks of the spontaneous problem solvers.

Now, a study published Thursday in the journal Science suggests that bumblebees possess insight. In a lab experiment, the insects were able to roll a plastic foam ball underneath an artificial blue flower, climb over the ball and use it to reach the flower, obtaining a sugary reward. “We showed for the first time that bumblebees can solve a completely novel object-manipulation task, spontaneously and without being trained to do so, or without any trial and error,” said lead author Akshaye Bhambore, a doctoral researcher at the University of Oulu in Finland.

Bumblebees can use socially learned behaviours and logical reasoning to solve puzzles, previous studies have shown. In the new experiment, however, the researchers exposed the insects to the separate elements of the task but never trained them on the solution itself.

This result suggests that a tiny insect brain can support surprisingly flexible behaviour, according to James Nieh, a professor in the department of ecology, behaviour and evolution at the University of California San Diego, who was not involved with the study. “Bees do not normally move objects around to make platforms, so this is not a natural bumble bee behaviour,” he wrote in an email. “But the experiment shows that they can remember a hidden goal location and manipulate an object in relation to that goal.”

This exciting new study shows that insects can learn and change their behavior in ways scientists are only just starting to understand, Natalie Hempel de Ibarra, an associate professor of neuroethology at the University of Exeter in England, said in an email. Hempel de Ibarra was not part of the research. This flexibility could shape how bees and other pollinators interact with flowers, helping them cope with challenges as environments and landscapes change, she added.

Getting the ball rolling

The researchers built a circular arena about 10 centimetres (4 inches) in diametre and 3.2 centimetres (1.3 inches) tall, in which bumblebees could walk but not fly. In the centre, the team placed an artificial blue flower containing a sugary solution and let the bees explore it. Nearby, the scientists placed a small foam ball to familiarize the insects with the object and demonstrate that it wasn’t a threat.

A second scenario presented a different challenge: The ball now covered the blue flower, and the insects successfully pushed it away to access the reward. In a third and final scenario — the one designed to test insight — the team moved the flower from the floor to the ceiling, just above one of four pits shaped to accommodate the ball. A majority of the bees that had been exposed to the first two scenarios — 75 per cent of them — managed to roll the ball to the correct pit and climb on it to access the flower.

The researchers also presented the third scenario to two additional groups of bees: one that had only been exposed to the flower but not the ball, and another that was completely new to both the flower and the ball. Bees from these two additional groups were not able to solve the puzzle.

“We wanted to know how much previous information they needed in order to solve the task,” said study co-author Olli Loukola, a behavioural ecologist and senior research fellow at the University of Oulu. “We need to get rid of the neophobia, or fear of new objects, by giving them the ball and showing that it is a safe object. And they also need the motivation, or the association between a reward and the blue of the flower, because if they don’t have that, blue means nothing. But these two things together give them enough information to spontaneously solve the real problem, which is using the ball as a ladder to reach the blue flower.”

To rule out the possibility that the bees could be solving the problem by moving the ball randomly or simply reacting to the visual stimuli of the blue flower, the researchers repeated the experiment with more stringent conditions. The team created a scenario in which the flower was not visible from the starting position of the ball. The bees that were exposed to the first two experimental scenarios were still able to solve the problem and access the flower.

Loukola said that the bumblebees exhibited “true goal-directed behaviour” by using the ball as a ladder, unlike in the second scenario in which they needed to simply push the ball off the flower. Referring to that simpler task, he added, “The bees didn’t need to understand anything about that task and they could still learn to solve it.”

In the third scenario, however, understanding the objective was a requirement. “They knew that if they could not reach the flower on the ceiling, there was a ball they could move to make themselves bigger, so they needed to kind of understand the physics of the task, and they needed to have a goal in mind,” he explained.

However, he added, this does not mean that the bumblebees possess humanlike reasoning or consciousness, and the study stops short of calling the rolling of the ball “tool use,” a definition that is usually highly debated when it comes to animal behavior.

Nevertheless, Loukola said, the result is particularly significant because the bumblebees are “truly naive,” meaning nothing in their life experience could have prepared them to solve the problem they were presented. “We can be sure that none of the bumblebees have any earlier experience about these tasks, so we know that this is not innate behavior.”

Learn and change

The bees’ performance is even more impressive than that of Köhler’s chimps, since in some experiments they couldn’t see the target when they started moving the ball, according to Lars Chittka, a professor of sensory and behavioural ecology at Queen Mary University of London, who was not involved with the study.

“In a sense it’s like you and me entering a room, finding something on the ceiling that needs dealing with — perhaps changing the lightbulb of a lamp — seeing that we need a chair or ladder to get high enough, then going to a different room to fetch the chair or ladder, and coming back with the equipment to the correct destination,” he wrote in an email.

“All this really requires some understanding of the task at hand, keeping in mind where the target is, and taking appropriate action.”

He added that the results should prompt scientists to rethink how much intelligence can be squeezed into a small nervous system, and that humans as thinking beings are surrounded by all kinds of other thinking beings, “however radically different their modes of thinking might be.”