If and when: intrinsic differences and environmental stressors influence migration in brown trout (Salmo trutta).

Partial migration is a common phenomenon, yet the causes of individual differences in migratory propensity are not well understood. We examined factors that potentially influence timing of migration and migratory propensity in a wild population of juvenile brown trout (Salmo trutta) by combining experimental manipulations with passive integrated transponder telemetry. Individuals were subjected to one of six manipulations: three designed to mimic natural stressors (temperature increase, food deprivation, and chase by a simulated predator), an injection of exogenous cortisol designed to mimic an extreme physiological challenge, a sham injection, and a control group. By measuring length and mass of 923 individuals prior to manipulation and by monitoring tagged individuals as they left the stream months later, we assessed whether pre-existing differences influenced migratory tendency and timing of migration, and whether our manipulations affected growth, condition, and timing of migration. We found that pre-existing differences predicted migration, with smaller individuals and individuals in poor condition having a higher propensity to migrate. Exogenous cortisol manipulation had the largest negative effect on growth and condition, and resulted in an earlier migration date. Additionally, low-growth individuals within the temperature and food deprivation treatments migrated earlier. By demonstrating that both pre-existing differences in organism state and additional stressors can affect whether and when individuals migrate, we highlight the importance of understanding individual differences in partial migration. These effects may carry over to influence migration success and affect the evolutionary dynamics of sub-populations experiencing different levels of stress, which is particularly relevant in a changing world.