Why study mechanisms of behaviour?

Behavioural ecology has long focused on “the evolutionary basis for animal behaviour due to ecological pressures”. With decades of work now showing that foraging, aggression, mating, and cooperation are elegantly adapted, why should we keep studying behaviour?

I think there are several reasons.

First, we often don’t know how behaviours work. How do seabirds circle the globe and return to the same island colony, year after year? How do fish make themselves invisible to predators in open water? How do dragonflies intercept flying prey, nine times out of ten? Many of these behaviours are abilities that we would like to be able to mimic technologically. We can learn from nature’s solutions to these problems that have been refined over millions of years of evolution. I think this will only become more important in the age of artificial intelligence, as we develop increasingly automated ways to solve sensory, motor, and decision-making problems – i.e., to create behaviours.

The applied side of behaviour needn’t be planned or fulfilled right away. For example, there has been a recent proliferation of optimization algorithms in engineering. My favourite? The “Bird Mating Optimizer”, based on theoretical monogamous, polygnous and promiscuous birds. It was inspired by classic work in behavioural ecology. Another example is reinforcement learning in machine learning, a type of artificial intelligence that can be traced back to Skinner’s behaviourist experiments using rewards to shape the actions of rats and pigeons.

A second reason to study behaviour is that it is fundamental to neuroscience. If we want to know how brains work, we have to start with robust, repeatable behaviours as the observable product of nervous systems. Consider the optomotor response, a reflex that allows organisms to visually stabilize locomotion. When flies are placed inside a rotating drum, they will turn in the same direction as the drum rotation. The discovery of this response began with behavioural studies of flies, mosquitoes and beetles in the early 20th century. This work led to the development of an elegant theory for a simplified circuit of cells that could compute motion cues from the retina, known as the elementary motion detector (EMD). Decades later, the theory has been verified and we know that EMDs really do exist in vivo.

Another example is bird song and neurogenesis. For over a century, it was assumed that new neurons are not generated in the adult brain. However, work on song birds in the mid- to late-20th century revealed that certain regions of the avian brain that are essential for song also grow (and later shrink) with each breeding season. This ultimately led to work that overturned the idea of static adult brains.

A third reason to study behaviour is that behavioural systems affect ecological and evolutionary dynamics. Animals that move bring a flow of nutrients, contaminants, diseases, and information. The movement of pollinators is crucial for ecology and agriculture. Behaviour also drives evolutionary change, since it acts as a selective force both within and among species. In other words, we need to understand not just how the environment shapes behaviour, but also how behaviour creates environments.

As a final point, I think there is tremendous value to being a student of behaviour because of its quantitative lessons. Most biological datasets are structured as a result of individuals and populations, time and space, and evolutionary relationships – and analysts have to take these relationships into account. But behavioural scientists also have to be mindful of additional sources of plasticity including motivation, response bias, learning, and habituation. Training in behaviour forces you to think explicitly about these sources of variation and thus provides a strong background in data science. “Data is the new oil”, and much of the current surge in data is focused on behaviour: e.g., of consumers, markets, and voters, to name a few. Behavioural science provides the foundation for understanding data in a rigorous way.