I work with a multidisciplinary group studying Phytophthora ramorum, an invasive pathogen with over 137 host species throughout Europe and North America. This pathogen causes the disease sudden oak death, which resulted in mortality of millions of oak and tanoak trees in coastal forests of California and Oregon1. Impacts from this disease include significant alteration of resource availability, habitat modification, and the devaluation and loss of property, especially when interacting with other disturbances. My research is focused on the multiscale interactions between P. ramorum and its environment, particularly with other disturbances such as drought, wildfire, and land-use change. The substantial social, economic, and ecological impacts of this pathogen at local to global scales make detection and management major concerns for stakeholders from the individual landowners to international governments.
Visualizing disease intensity in space and time: Sudden oak death in Sonoma County, California
This video shows a spatiotemporal estimation of disease intensity from P. ramorum infection of California bay laurel leaves across a 275-km2 study area in Sonoma County, California. Dr. Ross Meentemeyer (NC State University) established 200 plots across this area from 2003 to 2004, and collected data on P. ramorum and SOD related symptoms each spring through 2012: symptomatic leaf count of bay laurel and canker infection on oak and tanoak. Using the symptomatic leaf count data, we rendered this map by clustering plots within 1500-m of each other in space and leaf counts within two years of each other in time. This method is called space-time kernel density estimation, and in this application it provides visualization of which regions of the study area experienced the highest disease intensity (“hot spots”) based on the data. The x-y axes depict space and the z-axis shows time starting with the data from 2004 at the base to the data from 2012 at the top. Further analysis is required to determine if the disease intensity in one plot is related to the disease intensity in nearby plots. Whalen Dillon (NC State University) and Eric Delmelle (UNC at Charlotte) developed this visualization using Voxler 3D Visualization Software (Golden Software, Golden, CO). Citeable on figshare.
Perspectives of spatial scale in a wildland forest epidemic (pdf)
In this research we used Phytophthora ramorum as a case study to examine host-pathogen-environment interactions across multiple spatial scales. Our ability to understand and manage wildland forest epidemics hinges on understanding these types of multiscale relationships. We established 20 multiscale field sites, using line transects and the point-quarter sampling method to establish five nested scales of observation.
Armed with disease and host data at each scale we conducted statistical analyses to model disease intensity from two perspectives: a focal view, where the disease intensity at the smallest scale was examined as a function of each of the broader scales, and an aggregate view, where disease intensity was modeled using the landscape conditions at the same spatial scale. We further compared models using direct field measurements of host density to models using remotely sensed estimates of host habitat as predictor variables. The models using direct measurements of host density performed better than those using remotely sensed variables across four spatial extents, but there was no difference in model performance at the individual level. Focal view models showed that host density declined in performance as the scale of measurement increased, whereas host habitat improved.
“These results illustrate how the scale of observation – both spatial extent and measurement detail – can influence conclusions drawn from epidemiological models of wildland pathosystems.” – Dillon et al. (2014) Perspectives of spatial scale in a wildland forest epidemic. Eur J Plant Pathol. 2014. doi:10.1007/s10658-013-0376-3.
Trail leading to study sites in Sonoma County (left), and P. ramorum symptoms on a bay laurel leaf (right).