Chemically mediated learning in juvenile Atlantic salmon (Salmo salar) : testing the limits of acquired predator recognition under laboratory conditions and in the wild

• Main Author: Macnaughton, Camille J
• Format: Thesis
• Language: unknown
• Published: 2009
• Subjects: Atlantic salmon; Salmo salar
• Online Access: http://spectrum.library.concordia.ca/976206/1/MR63189.pdf

The assessment of predation risk is crucial to the survival of a prey individuals and the ability to gauge risk accurately will consequently be shaped by a suite of behavioural trade-offs. In salmonids, risk may be assessed through the detection of damage-released chemical cues. When these chemical cues are paired with a novel odour, covert antipredator responses are elicited upon subsequent exposure to the novel odour and learning occurs. My research focuses on the retention of newly acquired information (lemon odour), through sequential exposure to this same novel odour in both laboratory- reared and wild populations of juvenile Atlantic salmon ( Salmo salar ). Laboratory and field experiments consisted of a single conditioning day (AC + NO) followed by three recognition days (NO), in which antipredator responses were measured from the change in behaviour observed between the five minute pre-stimulus and post-stimulus observation periods. Significant short-term antipredator responses in the laboratory population were observed at the conditioning day, while they were absent at all subsequent recognition days. In particular, the foraging rate and the time spent moving decreased in response to the alarm cue treatment at the conditioning phase, but responses were not significantly different between treatments during any of the succeeding recognition phases. These results suggest that fish respond immediately and overtly to chemical cues, but may treat the information as irrelevant without subsequent exposure to the pairing of chemical cues with a novel odour. Conversely, my field experiment failed to confirm the laboratory results. Further work is required to elucidate any ecological processes that affect the learning mechanism in the current experiments.