Abstract
AbstractSize-selective harvesting evolutionary alters the life-history, behavioural and physiological traits in exploited fish populations. Changes in these traits may cause alteration in learning and decision-making abilities, either due to energetic trade-offs with brain investment that may vary across development or via correlations with boldness, sociability or other personality traits. Whether size-selective harvesting evolutionarily alters learning and decision-making abilities in fish remains unexplored, despite the global scale of fisheries. We tested the hypothesis that persistent removal of large-bodied individuals typical of many fisheries reduces learning ability in adults but not in juveniles, increases cognitive flexibility but reduces decision-making ability in adults. We examined associative learning through ontogeny, and reversal learning and collective decision-making in adults in three selection lines of zebrafish (Danio rerio) generated through positive, negative and random size-selective harvesting for five generations. Fish groups of each selection line were tested across ontogeny using a colour-discrimination paradigm with a food reward. The associative reversal task was conducted with a social reward, and the propensity to make group decisions was tested in an associative task. All selection lines showed significant learning ability and improved performance across ontogeny. Consistent with our hypothesis, the large-harvested line fish revealed a significantly reduced learning speed as subadults and adults, while the small-harvested line fish showed slower error rate compared to controls as 4-month old adults. We found no evidence of memory decay, and the selection lines did not vary in associative reversal ability. Decision-making speed did not vary across lines, but the large-harvested line made faster decisions during the probe trial. We conclude that size-selective harvesting typical of many fisheries evolutionarily alters learning and decision making. As this is likely to negatively and persistently affect resource acquisition and survival in exploited populations, we suggest that the cognition-related mechanism we identify may increase natural mortality.
Publisher
Cold Spring Harbor Laboratory