Temperature-dependent variation in alternative migratory tactics and its implications for fitness and population dynamics in a salmonid fish.

Abstract

Temperature-driven life-history modifications by adaptation occur in ectotherms, and therefore, life-history modifications by adaptation need to be taken into consideration when predicting population responses to the climate change. Partial migration is a common form of life-history diversity in which a population contains both migratory and resident behaviours. Salmonid fish exhibit a wide range of life-history diversity and, in particular, partial migration. We evaluated the effect of temperature-driven life-history modifications on population dynamics in partially migratory masu salmon (Oncorhynchus masou) by field observations and theoretical models. Field observations revealed that spatial patterns of alternative migratory tactics were associated with temperature gradients. The occurrence of resident males increased, whereas the proportion of migrant males and the proportion of delayed migrants including both sexes decreased with increasing temperature and, thereby, with improved early growth conditions. The expected fitness for each migratory tactic was computed in a life-history model with early growth conditions as a function. Individual fitness was maximized by adopting resident tactics under favourable early growth conditions, early migrant tactics under intermediate early growth conditions and delayed migrant tactics under unfavourable early growth conditions. The results suggest that individuals exhibited a status-dependent conditional strategy, that is, the adoption of alternative migratory tactics is influenced by the status of individuals to make the best of a situation. A simulation model suggests that increased residency by males to increased temperature leads to a substantial decrease in the number of migrants. Moreover, the decrease in the number of delayed (older) migrants with increasing temperature magnified fluctuations in abundance. Our findings indicate the importance of temperature-driven life-history modifications for predicting dynamics of natural populations under climate warming.