Consistency of behavioral profiles in zebrafish: A machine learning approach to bold and shy individual differences

Abstract

Behavioral variation among individuals is ascribed to the species' biology and the life history of each one. Many study areas consider individual differences, acknowledging their significant impact on fitness. Although boldness remains the most extensively studied behavioral dimension of individual differences in animals, ongoing debates persist regarding the evaluation of behavioral consistency over time and between contexts, as well as the determination of which features are crucial for delineating profiles. In this study, we investigated which behavioral traits explain the profiles of shyness and boldness and assessed their temporal and contextual consistency. For this, we divided zebrafish into bold and shy profiles by applying an emergence test (black-to-white entrance) three consecutive times with the same population of fish. The two groups formed (bold and shy) went through five different behavioural tests: novel tank, open field, black and white preference, aggressiveness, and sociability, which were employed twice, with an interval of 30 days. Bold animals showed less anxiety-like behaviour and higher aggressiveness compared to shy animals, and this pattern remained consistent over time for the two contexts. This suite of related behavious were considered the main factors to classify zebrafish into bold and shy profiles. In addition, the consistency appeared to be context dependent. The differences noted in the behavioural profiles allowed us to understand aspects of behavioural syndromes and how individuals behave when facing environmental challenges in different situations.