Ontogenetic shifts from social to experiential learning drive avian migration timing

Abstract

Migrating animals may benefit from social or experiential learning, yet whether and how these learning processes interact or change over time to produce observed migration patterns remains unexplored. Using 16 years of satellite-tracking data from 105 reintroduced whooping cranes, we reveal an interplay between social and experiential learning in migration timing. Both processes dramatically improved individuals' abilities to dynamically adjust their timing to track environmental conditions along the migration path. However, results revealed an ontogenetic shift in the dominant learning process, whereby subadult birds relied on social information, while mature birds primarily relied on experiential information. These results indicate that the adjustment of migration phenology in response to the environment is a learned skill that depends on both social context and individual age. Assessing how animals successfully learn to time migrations as environmental conditions change is critical for understanding intraspecific differences in migration patterns and for anticipating responses to global change.

Fig. 1. Whooping crane movement data and reintroduction program.

Daily positions of 105 satellite-tracked whooping cranes aged 1–6 during (a) autumn and (b) spring. Reintroduced juveniles (<age 1) were trained to perform their first autumn migration by (c) ultralight aircraft or (d) conspecifics; spring migrations and subsequent autumn migrations were performed in mixed-age groups of conspecifics (e).

Fig. 2. Environment, social learning, and experiential learning interact to affect latitudinal migration speed.

a Main effect coefficients and 95% confidence intervals of the effect of snow depth and NDVI on latitudinal speed during autumn and spring migrations. Both variables are standardized for comparability. b Response lines for interactions between age and snow depth (cm) during spring migration. c–e Response lines for interactions between age of the oldest bird(s) in the migratory group (i.e., group age; dashed lines), training method, and snow depth (cm) for subadults in their first independent autumn migration. Response lines for group age = 1, 5, 10, and 15 years are shown for demonstration purposes. f–h Response lines for interactions between individual age (solid lines), training method, and snow depth (cm) for all satellite-tracked birds ages 1–5 years during autumn migration. Lines for ages 4 and 5 in parent-reared birds are not shown due to lack of data coverage. Individual and group age were modeled as continuous variables in all analyses. Rugplots at the bottom of each panel show data distributions; shading represents 95% confidence intervals.

Fig. 3. Cumulative snow experienced during migration decreases with age.

a Total cumulative snow exposure experienced by conspecific-trained (hand- and parent-reared) birds during autumn and spring migrations by age (n = 242 migrations). Line and shading provide linear regression estimate and 95% confidence interval. b Cumulative snow experienced by three individuals ages 1, 2, and 3 years during spring migration in the year 2018. c Cumulative snow experienced by a single individual during spring migration as it aged from 1 to 5 years old.