Support vector machines for explaining physiological stress response in Wood mice (Apodemus sylvaticus)

Area of Science:

• Ecology
• Zoology
• Physiology

Background:

• The physiological stress response is a vital adaptive mechanism for prey species.
• Understanding factors influencing stress is crucial for wildlife conservation and ecological studies.
• Fecal glucocorticoid metabolites (FCMs) are reliable indicators of physiological stress in wild animals.

Purpose of the Study:

• To identify key environmental and individual factors explaining stress response in Wood mice (Apodemus sylvaticus).
• To analyze seasonal variations in stress levels over three years.
• To build predictive models for physiological stress using machine learning.

Main Methods:

• Seasonal collection of wild Wood mice fecal samples over three years.
• Analysis of fecal glucocorticoid metabolite (FCM) concentrations.
• Application of support vector machines to model FCM fluctuations based on environmental and individual variables.

Main Results:

• Support vector machines effectively modeled complex, context-dependent stressor interactions.
• Air humidity, temperature, and body weight explained 98% of FCM fluctuations.
• Environmental factors may relate to food availability and predation risk, while body weight may reflect reproductive or social stress.

Conclusions:

• Environmental conditions (humidity, temperature) and individual traits (body weight) are primary drivers of physiological stress in Wood mice.
• The stress response is a multifaceted process influenced by interacting factors.
• These findings enhance our understanding of wildlife adaptation to environmental changes and ecological pressures.