Epizootic Agent-Based Modeling: The Mobility Threshold for Disease Spread as a Critical Percolation Phenomenon

Abstract

Modeling of epidemic spread within human populations has been an active research area for a considerable time, yet relatively little work has been done in investigating the modeling dynamics of epizootic spread in livestock populations within intensive livestock production operations (ILPOs). Enclosed ILPOs are the standard environment for swine and poultry production in North America, and animal health is a key economic and food safety concern. This article investigates an agent-based model (ABM) overlayed on a modified dynamic percolation model to characterize epizootic spread within a poultry ILPO. The objective of the work was to develop a behavior-based disease modeling framework and, in doing so, to examine the relative impacts of infection longevity and agent mobility on epizootic spread and animal mortality in a poultry ILPO. Initial validation and systematic analysis of the model demonstrated that an ABM built on a regular percolation topology is a reasonable model to capture the macroscopic dynamics of disease transmission in a poultry ILPO. The results indicate that, at low population density, the infection duration has a greater impact on mortality than the impact of animal mobility on mortality. At higher population density, the impact of animal mobility rapidly dominates the impact of infection duration on mortality. The most significant insight arising from the study is that the percolation threshold, where a disease is able to spread throughout the entire ILPO environment, is greatly reduced when mobility is introduced. In other words, in addition to infection duration, animal mobility is a significant factor in accelerating disease spread. These insights and more detailed modeling can inform generalized livestock housing design considerations and management practices that may mitigate epizootic spread in an ILPO.