Download (PDF 13 KB): https://research.fs.usda.gov/download/treesearch/35075.pdf
Abstract
Landscape- to regional-scale models of plant epidemics are direly needed to predict largescale impacts of disease and assess practicable options for control. While landscape heterogeneity is recognized as a major driver of disease dynamics, epidemiological models are rarely applied to realistic landscape conditions due to computational and data limitations. Here we describe a stochastic susceptible-infectious epidemic model, applied to temporally and spatially heterogeneous landscape parameters, to predict the spread of the invasive forest pathogen Phytophthora ramorum, the cause of sudden oak death (SOD), in California (1990 to 2030). Three epidemiological processes (production of inoculum, dispersal, and infection) are modeled on a weekly time step across a 250 m by 250 m lattice composed of variable susceptible and infected host units. We describe how field, lab, and geospatial data were combined to parameterize and map the key system variables affecting transmission of
P. ramorum, including weather conditions, host infectiousness and availability, and a Markov Chain Monte Carlo estimated dispersal kernel. Replicated 1000 times to examine stochastic variability in epidemic outcomes, model predictions have a high degree of correspondence with 784 field plot observations that were collected across the pathogen’s potential geographic range to validate model performance. Results show that most disease spread occurs via local dispersal (
Parent Publication
Citation
Meentemeyer, Ross K.; Cunniffe, Nik; Cook, Alex; Rizzo, David M.; Gilligan, Chris A. 2010. Predicting the Spread of Sudden Oak Death in California (2010–2030): Epidemic Outcomes Under No Control. In: Frankel, Susan J.; Kliejunas, John T.; Palmieri, Katharine M. 2010. Proceedings of the Sudden Oak Death Fourth Science Symposium. Gen. Tech. Rep. PSW-GTR-229. Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station. pp. 121
