Version 2 2024-06-04, 13:21Version 2 2024-06-04, 13:21
Version 1 2017-11-24, 18:36Version 1 2017-11-24, 18:36
journal contribution
posted on 2024-06-04, 13:21authored bySM Drenner, SG Hinch, EG Martins, D Robichaud, Timothy ClarkTimothy Clark, LA Thompson, DA Patterson, SJ Cooke, RE Thomson
Temperature is recognized as a key factor influencing physiology, behaviour and survival of anadromous salmonids, yet little is known about their thermal experience, nor factors affecting it, during marine homeward migrations. In 2006 and 2010, approximately 1000 Fraser River sockeye salmon Oncorhynchus nerka were captured and tagged in coastal marine waters, ∼215 km from the river mouth, during their spawning migration. Individual salmon were blood sampled, gastrically implanted with temperature loggers fixed to radio or acoustic tags, and released. We recovered 50 loggers from freshwater locales containing 14690 hourly temperature readings. Mixed-effects models were used to characterize marine thermal experience, and examine the association of thermal experience with initial physiological status as well as oceanographic and meteorological conditions. Sockeye salmon thermal experience was highly variable (8.4°C to 20.5°C), and we detected opposite diel patterns between study years that could be associated with moon phase, behavioural thermoregulation, olfactory/celestial navigation or predator avoidance. We were unable to find any relationships between thermal experience and environmental conditions or fish physiological state. Nonetheless, we found that the greatest variability in thermal experience was attributed to within-individual variation, suggesting that environmental and physiological variables need to be examined at different temporal and spatial scales, and/or additional environmental and physiological variables need to be assessed. Overall, the factors associated with the thermal experience of homing sockeye salmon in coastal marine environments are more complex than previously thought, and multiple year studies are needed before generalizing behavioural patterns observed from single year studies.