Using an algorithmic model to reveal individually variable movement decisions in a wintering sea duck

Area of Science:

• Ornithology
• Animal Behavior
• Ecology

Background:

• Migratory birds are often assumed to be stationary in winter.
• Recent studies reveal mid-winter movements in various species, but the drivers are unclear.

Purpose of the Study:

• To investigate winter movements of king eiders (Somateria spectabilis) in the Bering Sea.
• To determine if environmental factors influence movements exceeding 50 km.

Main Methods:

• Satellite transmitters tracked 95 king eiders from 2002-2006.
• An algorithmic random forest model assessed predictive variables for site departure.

Main Results:

• High individual variability in winter movement decisions was observed.
• Individual bird identity was the most significant predictor, followed by location, date, and sea ice concentration.

Conclusions:

• Individual strategies interact with environmental conditions, creating diverse movement patterns.
• Winter movements may be exploratory, aiding in acquiring information on alternative sites for future survival.