A framework for integrating inferred movement behavior into disease risk models.

Area of Science:

• Ecology and Disease Modeling
• Wildlife Movement Ecology
• Spatial Epidemiology

Background:

• Habitat selection by hosts influences pathogen exposure probability.
• Incorporating host movement behavior into disease risk models is often overlooked.
• Ecological Niche Models (ENMs) can be enhanced by behavioral data for disease risk assessment.

Purpose of the Study:

• To demonstrate how integrating inferred movement behavior impacts disease risk map construction.
• To map areas of likely Bacillus anthracis persistence using ecological and pathogen data.
• To analyze zebra movement behavior and its relation to anthrax exposure risk.

Main Methods:

• Utilized Maximum Entropy (MaxEnt) to model Bacillus anthracis persistence based on environmental variables.
• Employed Hidden Markov Models (HMM) to differentiate zebra foraging and non-foraging behavioral states.
• Developed Step-Selection Functions (SSFs) integrating behavioral states and MaxEnt predictions.

Main Results:

• Disease risk varied significantly between different zebra behavioral states (foraging vs. non-foraging).
• Analysis of full movement tracks obscured state-specific exposure risks.
• Behavioral state-specific analyses provided a more nuanced understanding of pathogen exposure.

Conclusions:

• Pathogen distribution models can be misleading without accounting for host behavioral states.
• Explicitly incorporating behavioral states into selection analyses improves exposure risk evaluation.
• Further research should build state-specific selection functions for accurate comparative risk assessment.