Group size and cooperation

Do animals in larger groups always cooperate less?

The more people who observe an incident like a fire, the less likely it is that anyone will call for help. This is initially counter-intuitive: surely the more people see a problem, the more likely it is that someone will help? Except that sometimes, the more people that are present, the more everyone thinks “well someone else will do it, so I won’t” so each person becomes less likely to help. And this is something that we see in some scenarios in nature. When birds or antelope are in a larger group, they will spend less of their time looking for predators than they do when they are in smaller groups, even though if they look for predators, that is helpful to everyone (including the animal itself!). But when we investigated a similar phenomenon in a small tropical fish, we found something else.

In many small fish, their primary response to predators is to approach the predator. This seems a strange behaviour – why would individuals put themselves in danger by getting closer to a predator? Well, there is some evidence that when predators are approached in this way, they are less likely to attack. The small fish also gain information about the predator from this behaviour, which allows them to make better decisions about how to behave (should they flee? Or can they happily carry on foraging?). The interesting thing about this behaviour is that even if just one individual approaches the predator, everybody in the group benefits from the predator deterrence and the information gathered. This makes this predator inspection behaviour a ‘public good’.

So should we expect fish to behave in the same way as people observing a fire when they first see a predator? Does a larger group mean that the predator is less likely to be inspected by anyone? Does it mean that fish are more likely to wait for someone else to do it? According to our experiment, no. We found that larger groups actually had higher per capita rates of inspection. But why?

Well, we don’t really know why yet, but we can speculate. Inspecting a predator is much riskier to a fish than calling the fire brigade is to a person. But it also becomes much less risky when others get involved. When a big group of fish inspects a predator, this is much less risky for each fish in that group than if they inspect in just a small group. Additionally, a lone fish is at much greater risk anyway, even before we bring predators into the mix. This means that fish follow each other, they don’t particularly like to be alone. So in a big group, if one fish decides to inspect the predator, others are likely to follow and if most individuals are going in that direction, then it’s usually better to follow than strike out alone.

There are still lots of unanswered questions that our data weren’t precise enough to get at: are these inspections carried out equally by everyone, or do some individuals take on most of the labour? How are the costs and benefits shared and how does this change with group size? How does group size affect the information (the public good) that is produced during inspection?

There is still so much more to explore about predator inspection and public goods in general, and the nuanced differences between the behaviours that produce public goods and how they might cause the differences that we observe in nature. As it becomes possible to get very detailed data about animal behaviour, we will undoubtedly learn more about how and why animals behave in the ways that they do and how this matches (or doesn’t!) with our predictions.

Read the open access paper in Proceedings B.