Résumé: Coping style is defined as a set of individual physiological and behavioural characteristics that are consistent across time and context. In the zebrafish Danio rerio, as well as in many other animals, several covariations have been established among behavioural, physiological and molecular responses. Nonetheless, not many studies have addressed the consistency in behavioural responses over time starting at the larval stage. Therefore, this study aimed to improve the understanding of behavioural consistency across contexts and over time in zebrafish from the larval to juvenile stages. Two distinct experiments were conducted: a larval stage experiment (from 8 to 21 days post fertilization, dpf) and a juvenile stage experiment (from 21 to 60 dpf). On one hand, the larval experiment allows to focus on the transition between 8 and 21 dpf, marked by significant morphological changes related to the end of larval stage and initiation of metamorphosis. On the other hand, the juvenile experiment allows to properly cover the period extending from the end of larval stage to the juvenile stage (60 dpf), including metamorphosis which is itself completed around 45 dpf. Within each experiment, boldness was determined using a group risk-taking test to identify bold and shy individuals. A novel environment test was then performed at the same age to evaluate consistency across contexts. Groups of fish (either bold or shy) were bathed in an alizarin red S solution for later identification of their initially determined coping style to evaluate behavioural consistency over time. Fish were then reared under common garden conditions and challenged again with the same behavioural tests at a later age (21 and 60 dpf in the larval and juvenile experiments, respectively). Behavioural consistency was observed across contexts, with bold fish being more active and expressing higher thigmotaxis regardless of age. There was, however, little behavioural consistency over age, suggesting behavioural plasticity during development. Moreover, the use of alizarin red S to conduct this experiment provides new perspectives for the further study of the longitudinal evolution of various traits, including behaviour, over life stages in fish.

Résumé: In various experiments under self-feeding conditions, sea bass groups could be divided into three categories regarding feeder actuation: high, low and zero-triggering fish. In all cases few high-triggering fish were responsible for a high percentage of the feed delivery. A question was raised about the role played by feeding motivation in such high-triggering status acquisition. It was approached by applying a 3-week fasting period in order to induce similar negative specific growth rate (SGR) in two groups of fish of similar mean weight but with either a low or a high coefficient of variation for weight (CVw)(T-low: CVw similar to 11%, 3 tanks of 60 fish each; T-high: CVw similar to 20%, 3 tanks of 60 fish each). These groups were created to test the consistency of behavioural responses in two different contexts (i.e. two population size-distributions). During the follow-up period of 40 days, the group level feed-demand behaviour was not strongly modified by the fasting period and there were no differences between T-low and T-high groups. Complete growth compensation was the same in all tanks as observed at the end of the experiment. At the individual level, high-triggering fish were exactly the same individuals before and after the fasting period. Up to four high-triggering fish could be observed according to the tank and when several fish were performing high-triggering activity, their rankings were sometimes reversed after the fasting period. High-triggering fish increased their activity levels after the fasting period showing behavioural plasticity. High-triggering status could neither be explained by an initial lower SGR nor a sex effect, nor by any of the measured physiological blood parameters. Thus, individual's triggering activity levels could be related to personality and/or metabolic traits but further research is required to confirm this assumption. (C) 2011 Elsevier B.V. All rights reserved.

Résumé: Fish learning and cognition are usually approached by testing single individuals in various devices such as mazes that have serious drawbacks, especially in gregarious species, including the stress induced by the test procedure. This might impair the results and lead to misinterpretation about the learning abilities of the targeted species. In order to provide an alternative to the individual-based tests, we investigated for the first time the operant conditioning of four similar groups (50 individuals per tank) of sea bass. We used two computerized self-feeder devices per tank, each coupled with individual electronic identification and that were alternately activated during varying positive appetitive reinforcement period of time (7 to 1 day). Learning abilities were examined at both group and individual levels. At the group level, the operant conditioning was demonstrated as the triggering activity significantly decreased when the device was turned off and increased when it was turned on, whatever the reinforcement period duration. The individual level analysis revealed a more complex situation with fish showing different learning performances that can be best explained through the producer-scrounger game theory.