Deriving spatially explicit direct and indirect interaction networks from animal movement data.

Area of Science:

• Movement ecology
• Animal behavior
• Spatial analysis

Background:

• Understanding animal social structures requires quantifying spatiotemporal interactions.
• Global Positioning System (GPS) data offers insights but misses interactions between discrete locations.
• Existing methods underestimate interactions, especially with lower temporal resolution data.

Purpose of the Study:

• To develop a novel method for quantifying fine-scale spatiotemporal animal interactions using continuous-time movement models (CTMMs) and GPS data.
• To infer interactions occurring between observed GPS locations and account for indirect interactions.
• To assess the method's performance and applicability in disease ecology.

Main Methods:

• Applied CTMMs to infer high-resolution animal movement trajectories from GPS data.
• Developed a framework to estimate individual and spatial interaction patterns, including indirect interactions.
• Validated the method with simulations and applied it to wild pigs and mule deer for disease-relevant network inference.

Main Results:

• Simulations indicated significant underestimation of interactions with GPS data intervals exceeding 30 minutes.
• Empirical application revealed underestimation in both interaction rates and spatial distributions.
• The CTMM-Interaction method successfully recovered most true interactions, despite inherent uncertainties.

Conclusions:

• The developed CTMM-Interaction method accurately quantifies fine-scale spatiotemporal interactions from lower-resolution GPS data.
• This approach overcomes limitations of discrete GPS data for understanding social dynamics and disease transmission.
• The method provides a foundation for dynamic social network analysis and predictive ecological modeling.