Heritabilities and genetic correlations of Drosophila melanogaster locomotory behaviour traits: a high-throughput phenotyping approach

Szymon Marian Drobniak , Erian Macartney, Samantha Burke, Patrice Pottier, Catharina Vendl, Zine Hamoudi, Chloe Hart, Radiah Ahmed, Yong Qi Lin, Malgorzata Lagisz , Gregory Neely, Shinichi Nakagawa 

Genetic variance forms the basis for evolutionary inferences as it describes the evolutionary potential of traits. The major limitation of quantitative genetic studies is achieving sufficient power and sample sizes to estimate heritabilities with sufficient precision. This issue is especially important in the case of traits that are inherently susceptible to stochastic, nonbiological variation. Behavioural traits have long been associated with this group, often yielding quantitative estimates of genetic parameters that are subject to broad estimation errors, thereby hampering the discovery of genetic variation underlying such characters. Here, we used a well-established panel of inbred genetic lines of the fruit fly, Drosophila melanogaster, to estimate relevant genetic parameters in a range of behavioural traits associated with mobility and exploration. Using a high-throughput phenotyping approach and automated scoring of large numbers of individual animals, we provide precise estimates of the quantitative genetic background behind some basic characters associated with animal behaviour. Fruit flies turn out to harbour significant genetic variance in traits directly associated with mobility and substantially lower heritabilities of traits describing the temporal variability of Y-maze movements. Mobility traits also appeared to be only moderately genetically correlated, except for movement distance vs. variability traits, where we estimated strongly negative genetic correlations. In general, our results demonstrate the existence of evolutionary potential in behavioural trait proxies measured by high-throughput methods, additionally hinting at the potential for sex-specific effects. They also emphasise the growing importance of high-throughput phenotyping in modern behavioural biology and ecology.