Accelerometers as a tool to characterize reproductive behavior within social groups in long term experiments: the case of the Japanese Quail

BARBERIS L.; FLESIA, A. G.; SIMIAN C.; MARIN R. H.; KEMBRO, J. M.

Congreso; 2020 SMB Virtual Annual Meeting; 2020

Accelerometersare devices that convert movement into three signals belonging to eachcomponent of the acceleration vector at a high acquisition rate, up to 25 dataper second. When they are fixed to an animal, each action performed by theindividual leads to a particular shape in these signals that, when depicted ina computer, can be isolated and classified. Hence, accelerometer recordings canbe combined with machine learning techniques in order to automatically classifysignals into behavioral categories. This is particularly useful in the contextof long-term social behavior studies in large or natural environments wererecording from visual observation is difficult and time consuming. Herein, weplaced accelerometers on the back of adult male quails (Coturnix japonica) to register their activity when they arereleased into a home box containing two female quails during a 1-hour period.At the same time, the experiment was video-recorded to obtain a time series ofthe different behaviors performed by the male and their corresponding durationby direct inspection. The accelerometric signals and behavioral time seriesobtained were used to train a neuronal network. Our neuronal network was ableto classify reproductive behavior of males at high temporal resolution. Inparticular, we showed, first, that the duration of some reproductive events canbe much shorter than those reported previously and transitions betweendifferent behaviors are very fast (of the order of ~100ms). Second,reproductive behavior appears to begins earlier and finish later than it ispossible to observe visually using video recordings. Our results show thatcombining accelerometer recordings with neural network processing is a powerfulmethod to automatically register reproductive behaviors within social groupswith high accuracy. This is of particular importance given that it has the potentialto replace registering from visual observation of social behavior. Moreover,the long, high resolution reproductive time series obtained by this approachcan be useful for studding long-term reproductive behavioral rhythms inpoultry.