Social BEEhaviour: how bumblebees overpower Pavlov
Humans are not the only animals to succumb to peer-pressure: even bumblebees show conformist behaviour, which may cause them to perform sub-optimally. What does this mean for the ongoing discussion on learning mechanisms?
Do you remember those times when you would forage from a certain flower simply because all your friends foraged from the same flower, and then your mom would say: “if all your friends jumped off a bridge, would you do it too”? Probably not, because your mom would never be so passive aggressive and, more importantly, you are not a bumblebee. However, if you had been one of these soft-haired pollinators, this scenario might have been more relatable. Because even though bumblebees are less likely to be persuaded into excess drinking or getting regrettable tattoos, they have been found to succumb to some form of peer-pressure.
Follow the BEEder
This has recently been shown by a research group that focussed on the negative effects of social learning in bumblebees. In their experiments they exposed individual bumblebees to different settings with artificial “flowers” (a romantic interpretation of coloured Perspex squares). The flowers could be divided into two groups: flowers with low and with high nutritional value. The former of these otherwise identical flowers were occupied by model bees. Ideally (for the bees at least, for the researchers it would have invalidated years of research) the bees would have learned to avoid the occupied flowers and go to the highly rewarding unoccupied flowers instead. Surprisingly, they continued to visit the occupied flowers as frequently as their more rewarding alternatives.
Now being the superior large-brained primates that we are we could stop here and be amused by the silly little bees and their inability to learn the simplest cues. This mindset of course will not get you published, and would be detrimental to science in general, and thus the researchers tested another condition. They used the same two groups of flowers, but now the poorly rewarding ones were cued by white blocks rather than model bees. And this is where things got interesting: after several rounds of exposure the bees ended up visiting the unoccupied flowers more than the others, which opposes the results from the model bee condition. With this it became clear that bumblebees are in fact capable of learning to associate cues with rewards, but this ability seems to be overpowered when social factors come into play.
The buzz on animal learning
So how are these results relevant to us? It is of course always nice to learn more about the most beloved of bee-species, but few people will lie awake at night wondering about the implications of the above research for the daily life of the bumblebee. But the implications for more intelligent animals, that is hot stuff, especially in a field as controversial as that of behavioural science. While it seems like a big leap to go from bumblebees to animals such as primates, it is exactly the distance between these two species that makes the found learning patterns relevant: they can be used in describing the most basic principles of learning that may be similar in an incredible variety of species.
To do this we are first going to travel back in time, to the early twentieth century. During this time the physiologist Pavlov found that animals can learn through conditioning, i.e., the association of neutral stimuli (e.g., bells) with potent ones (e.g., food). For many years people viewed Pavlov’s results as the absolute - and moreover exclusive – truth on learning principles in all non-human animals. This gave rise to the theory of behaviourism, which states that all behaviours of these animals are acquired through conditioning.
Although it seems increasingly inappropriate to apply the behaviourist belief to intelligent animals, it is understandable to use Pavlov’s theories to explain learning in animals with small brains (enter our friend the bumblebee). This is exactly how some years ago researcher Martin Giurfa made sense of findings of seemingly social learning in insects. These could be interpreted as indications for learning mechanisms that are more extensive than conditioning. Giurfa on the other hand, being sceptical of the cognitive abilities of insects, found another explanation. For one illustrative experiment in which bumblebees copied conspecific foraging behaviour, he stated that they still learned through conditioning: having learned to associate other bees with nectar during joint foraging, the bees will start to associate occupied targets with nectar. Hence the behaviourist theory stands strong…
… for a while. But now imagine if a bumblebee first learns to associate fellow bees with a poor reward (sound familiar?). This must mean that they will start to prefer unoccupied targets – at least according to the behaviourists. The attentive reader will remember that this is not the case: in this condition bumblebees have been shown to remain equally attracted to occupied flowers. This must mean that conspecifics are distinct from other stimuli in the process of associative learning. Thus there is probably a form of social learning innate in bumblebees, contrary to Giurfa’s statements.
From the bee to you and me
This gives reason to believe that learning in animals is more extensive than a black and white distinction with conditioning on one side and cognitive learning on the other; there are mechanisms in between that enable different forms of learning to take place.
The found results also imply that social behaviour is instinctive even in a species as relatively unintelligent as the bumblebee. As humans are also well known for showcasing this behaviour, it could be that the core features to this characteristic were already present in our common ancestors. It is, however, more likely that sociality is so important to survival that is has evolved independently in a wide range of species. Hence the next time someone accuses you of being a mindless follower, you can simply blame your actions on either your efficient evolution or (perhaps less convincingly) your common ancestry with the bumblebee.
Avargues-Weber, A., Lachlan, R., & Chittka, L. (2018). Bumblebee social learning can lead to suboptimal foraging choices. Animal Behaviour, 135, 209-214. doi:10.1016/j.anbehav.2017.11.022.
Giurfa, M. (2012). Social Learning in Insects: A Higher-Order Capacity? Frontiers in Behavioral Neuroscience, 6, 57. (2012). doi: 10.3389/fnbeh.2012.00057