Every year, billions of animals across the globe embark on spectacular life-changing migrations following distinct seasonal patterns. Such migrations are expected to be in sync with environmental- and ontogenetic change of relative profitability of the alternative habitats and timing therefore differs between individuals both within- and between populations. But whereas theoretical predictions on such variation are fairly well developed, empirical studies on especially between-population differences are often limited by the lack of replication on population level. Furthermore, it is often not possible to determine whether it is the seasonal and ontogenetic profitability change in the natal- or in the migratory habitat that are driving the differences in migration patterns. Here, we are studying individual migration patterns of juvenile trout from 15 different populations migrating between natal streams and a shared migratory habitat, i.e. a large lake. We tagged several thousand juvenile trout with PIT-tags and recorded migrations from their natal streams to a shared migratory habitat, i.e. a large lake. We hypothesized that within-population differences are dependent on individual traits in concert with environmental conditions in both the natal and in the migratory habitat, and that between population variation are driven by differences between the natal habitats only.
Results/Conclusions
We find a significant ontogenetic influence on migratory timing, which is similar across populations, with large juveniles migrating earlier in the season smaller ones. This agrees well with predictions based on temporal development of size- and habitat specific predation risk-growth rate tradeoffs. However, differences between populations are mainly determined by the differences in environmental conditions between the natal streams. Specifically, trout from streams with warmer spring temperatures generally migrated earlier than trout from streams with colder spring temperatures. While this agrees with the concept of a species specific threshold temperature and a species specific number of degree days, it disagrees with the predictions of stream specific optimal time of migration calculated from the temporal development of habitat- and location specific predation risk-growth rate tradeoffs. This discrepancy is potentially due to stocking of trout of allopatric origin, i.e. not locally adapted, in the streams.