Context matters for bird personality

Date: 17 March 2017

Introduction: Lisa Arvidsson (UAntwerpen) shows traditional lab measurements of animal personality cannot be easily extrapolated to another environment.

More and more evidence point to the fact that animals’ personality influences whether and how far individuals disperse, how individuals cope with environmental change and how much offspring they produce (“fitness”). The most commonly used definition of animal personality refers to “consistent individual differences in behaviour across time and/or across contexts”. Personality traits are mostly measured in the laboratory.

However, the quantification of animal personality still battles with a number of questions. What is it we measure? And where should we measure it? A laboratory environment differs strongly from the wild, and extrapolating interpretations from the lab to the real world is difficult.

In her newest research study, Lisa Arvidsson (Global Change Ecology Centre, Research Group Evolutionary Ecology) and her colleagues studied one of the most commonly used personality traits in ecological and evolutionary studies: exploration behaviour.  Lisa: “We used great tits (Parus major) in our study. Exploration behaviour in great tits is typically measured in a small room with five artificial trees, and the number of hops and flights over two minutes are added up to a single exploration behaviour score. This measure is commonly interpreted to indicate the willingness to explore the environment. However, it could also be seen as just measuring the activity. The room is not complex and can be surveyed in its entireness in a single glance.”

Lisa and her colleagues therefore changed the traditional context.  They constructed a larger and more complex testing space, called ‘the arena’. The arena consists of 8 rooms connected by corridors. In the arena they measured both the total number of hops and flights similar to the traditional exploration behaviour score, and the number of rooms visited.

The traditional lab environment (top) with only one room and five trees, showed very different results than measurements in the new arena that was used by Lisa Arvidsson and colleagues (bottom).

Lisa: “We hoped to see whether an individuals’ exploration behaviour score in a traditional room was related to the exploration of the arena by visiting a different number of rooms, or just the activity in the rooms. We found that both measures in the arena were highly repeatable over time, telling us that we indeed measured relevant personality behaviour. However, to our surprise, no correlation was found between any behaviour in the arena, and behaviour in the traditional lab test. This shows that it is extremely difficult to extrapolate simple lab results to a complex environment."

Results were published in Animal Behaviour.

Movie of the experiments in the arena. Bird flies around minute 1:32.