Bumblebees can be trained to solve problems in exchange for a food reward, according to a pair of new studies from the University of Guelph.

The first study, published in the journal Animal Cognition, tested bumblebees' problem solving abilities and the second, published in the journal Psyche , tested the bees' social learning abilities.

For the problem-solving test, Hamida Mirwan, a PhD student and Peter Kevan, a professor in Guelph's School of Environmental Sciences, presented the bees with a series of artificial flowers. Each of the flowers contained a sweet reward that could be obtained by figuring out how to access the treat. The difficulty of the puzzles progressed, requiring increasingly challenging strategies such as moving objects aside or upwards to gain access to the sugar syrup reward.

If the most complicated of the artificial flowers was presented to an inexperienced bee, the researchers found that the bee would be unable to solve the puzzle and get the reward. But if the bees were presented the artificial flowers in order of easiest to most difficult, they were able to solve the more challenging puzzles.

"Bees with experience are able to solve new problems that they encounter, while bees with no experience just give up," Mirwan said in a news release.

The researchers drew a parallel between the bees' learning ability with a concept in human behavior known as scaffold learning, which is usually so complex it is not applied to animals.

The second study tested bees' ability to learn by watching and communicating with other bees.

For this experiment, the researchers designed artificial flowers containing a sweet reward that could only be accessed by walking on the underside of the contraption.

Bees were allowed to feed from these trick flowers for several days, then the experiment chamber was partitioned with a mesh screen. A group of inexperienced bees were placed on one side of the partition and allowed to observe the experienced bees using the trick flowers. When the inexperienced bees were introduced to the trick flowers, it took just over one minute for them to get the reward. A control group of bees that had no prior introduction to the trick flowers could not access the food reward.

"Social learning in animals usually involves one individual observing and imitating another, although other kinds of communication can also be involved," said Mirwan. "They could try for up to 30 minutes, but most gave up before then."

The researchers say it's clear some form of social learning is taking place, but they are still uncertain of the exact communication method. It could be visual, through touch, vibration or smell, they said.

"We can't quite explain how bees that had never even seen an artificial flower were able to become adept so quickly at foraging on them, but clearly some in-hive communication took place," said professor Kevan. "It suggests that social learning in bumblebees is even more complex than we first expected."