Temporal epidemiology of sudden oak death in Oregon

Abstract

An effort to eradicate Phytophthora ramorum, causal agent of sudden oak death, has been underway since its discovery in Oregon forests. Using an information-theoretical approach, we sought to model yearly variation in the size of newly infested areas and dispersal distance. Maximum dispersal distances were best modeled by spring and winter precipitation 2 years before detection, and infestation size the year prior. Infestation size was best modeled by infestation size and spring precipitation the year prior. In our interpretation, there is a 2-year delay between the introduction of inoculum and onset of mortality for a majority of sites. The year-long gap in between allows ample time for the production of inoculum contributing to the spread of P. ramorum. This is supported by epidemic development following changes in eradication protocols precipitated by an outbreak in 2011, attributable to a 2009 treatment delay and an uncharacteristically wet spring in 2010. Posteradication, we have observed an increase in the total area of new outbreaks and increased frequency in dispersal distances greater than 4 km. Although the eradication program has not eliminated P. ramorum from Oregon forests, it has likely moderated this epidemic, emphasizing the need for prompt treatment of future invasive forest pathogens.