Long and winding road: Training progress and trainability variation across a psychoacoustic experiment in penguins

Abstract

Operant conditioning has been used in a variety of animal species both in captivity (laboratories, zoological institutions) and in the wild. Such training may enrich the environment of animals or facilitate their manipulation but it can also be used to answer scientific questions (so far primarily regarding sense modalities). However, the trainability of animals may vary because of cognitive differences between species or because of individual and external factors within species. The assumed difficulty of training seabirds presumably explains, why they have been mostly overlooked in studies based on operant conditioning, such as psycho-acoustic studies. Here, as part of a broad project examining the hearing capacity of penguins, we trained four captive Humboldt penguins (Spheniscus humboldti) for psychoacoustic trials, following well-established operant conditioning methods. In addition to describing the different steps of the training of penguins, we examined how the trainability of each individual (concentration, response to call and stationing quality) was affected by different life stages (moult, pairing interest) and environmental conditions (daylength, temperature, humidity, visitors). After 22 months and ca. 700 trials, all four penguins were ready for psychoacoustic measurements, which in turn extended over 22 additional months. However, the trainability of penguins was not stable over these 44 months, as their concentration decreased over time (presumably because of habituation) while the quality of their call response and stationing increased for most of them (presumably reflecting learning progress). Moreover, all trainability parameters were strongly reduced during moult while the effects of pairing interest and environmental conditions were more variable between individuals. Our study demonstrates that, despite the length and the instability of training as well as some interindividual differences in trainability, it is possible to train captive penguins for scientific purposes. As such, we hope that our study opens the path for further studies based on the training of animals, for which training is assumed to be challenging but for which scientific data are urgently needed because of their vulnerability in the face of environmental changes.