Skip to main content
ARS Home » Southeast Area » Fort Pierce, Florida » U.S. Horticultural Research Laboratory » Subtropical Plant Pathology Research » Research » Publications at this Location » Publication #335681
Title: Census-travel risk model to predict points of disease/pest introduction

Author
item Gottwald, Timothy
item LUO, WEIQI - North Carolina State University
item RILEY, TIMOTHY - Animal And Plant Health Inspection Service (APHIS)

Submitted to: Journal of Citrus Pathology
Publication Type: Abstract Only
Publication Acceptance Date: 12/20/2016
Publication Date: 5/18/2017
Citation: Gottwald, T.R., Luo, W., Riley, T. 2017. Census-travel risk model to predict points of disease/pest introduction. Journal of Citrus Pathology. 4(1):15/45.

Interpretive Summary: Imports and foreign travelers can unintentionally introduce exotic pathogens or pests, which domestic transportation can exacerbate further local spread. A model based on travel/trade data and the U.S. Census has been developed to estimate, rank and interactively map local areas for introduction of HLB, as well as other citrus diseases. The model has been integrated into other survey methodologies using local risk rankings for California, Texas and Arizona. A user-friendly front-end interface of the census travel model has been implemented.

Technical Abstract: There is increasing concern that invasive pests and diseases can be introduced into new areas due to international travel. For early detection of pathogens and their vectors, we have developed a predictive-system methodology that maximizes the prediction points of introduction and identifies high-risk areas for local transmission at an early epidemic phase. The model is versatile and independent of pathosystem, meaning it can be applied to various hosts, pathogens and vectors, and other insect pests. The model is constructed based on US Census and international travel data, partnered with knowledge of the epidemiological characteristics of the pathosystem. Combining existing foreign population habitat and international pathway data, the model generates a risk index map to identify locations (at various spatial scales) with ranked introduction potential for the disease. The foreign travel distribution is estimated for each of 176 individual source countries. This is then coupled with actual travel statistics from national travel and tourism offices to calculate the risk. The risk weighting for each individual source country is adjusted by the number of confirmed disease cases (e.g. log transformed) or the percentage of active area. The census travel model estimates the likely destination(s) of foreign travelers (e.g. connectivity to family, relatives & friends) in order to place a risk-bias for survey of interested areas. The census travel model is currently implemented in Shiny (a web application framework for R) to bring greater ease-of-use for regulatory agencies to detect high risk pathogens of agricultural significance at the early epidemic phase. The census travel model can parameterize and weigh risk contribution of international travelers in various categories (i.e. seasonal travel volume, visa class, age and gender) given there is a valid disease-associated concern. The census travel front-end allows users to run the model that best suits their needs without strong knowledge of background code and data. Risk maps have been generated for plant diseases (e.g. huanglongbing, citrus black spot, plum pox virus) and other human viruses (e.g. Ebola, malaria, dengue, Chagas and Zika). The model has been incorporated into risk-based Multi-Pest Survey (MPS) programs to provide an augmentative risk factor for continued risk introduction from human-mediated disease spread. In addition, the model can be linked with survey scenario estimators to balance efforts and costs within agency constraints and among emergency programs.