Modeling the spread of invasive species using dynamic network models

Abstract

Spread dynamics of biological invasions are influenced by both the availability and spatial arrangement of suitable habitat. As such, invasive spread can be considered to occur across a network of nodes, representing patches of suitable habitat, with linkages representing the potential for movement between habitat patches. While static network models can provide valuable insight into the potential framework of nodes and linkages across which spread could occur, they offer little information on the actual spatiotemporal dynamics of range expansion processes. Here, we explore the development and application of dynamic network models (DNMs) to model the spread of invasive species. DNMs accommodate temporal dynamics in the utilization of nodes and the connections between them and can flexibly perform simulations at the spatial scales of observational data. As case studies, we develop DNMs to simulate the spread of a generalist forest pathogen and the hemlock woolly adelgid (Adelges tsugae Annand). We highlight the utility of DNMs for identifying habitat patches that contribute most to spread across the landscape and for visualizing emergent spread dynamics. While currently underutilized in ecology as compared to static network models, DNMs are potentially applicable to numerous research and management questions relevant to biological invasions and the more general phenomena of range expansion.