The perils of naïve use of open-source data: A comment on “Spatiotemporal distribution of sudden oak death in the US and Europe”

Abstract

Kang et al. (2024) present a spatiotemporal analysis of Phytophthora ramorum outbreaks from 2005 to 2021 in the United States and Europe. However, the analysis and conclusions are flawed because of a lack of understanding of the pathosystems analyzed which led the authors to select improper methods for their analysis. The open-source data analyzed does not include sampling over all seasons of the year. Sampling is primarily conducted in the spring which makes the data unbalanced and inappropriate for examination of seasonality without transformation. Differences in characteristics, and significant driving factors (e.g., relative humidity) between the locations where infection clusters occur, irrigated nurseries with complex sources of inoculum and modified environments versus natural forests subject to only ambient environmental conditions, were not considered when analyzing relationships between moisture conditions and pathogen spread. Additional occurrence records exist for P. ramorum in the United States and the United Kingdom, but they were not included in the analysis. Clear descriptive language and proper study design are required to understand how environmental conditions influence P. ramorum establishment and spread so they can inform forest management and regulations to protect the resources at risk. An understanding of the temporal and spatial dynamics of Sudden Oak Death, Sudden Larch Death, Ramorum Blight and other diseases caused by P. ramorum is critical to serve as the basis for management strategies to limit losses and pathogen spread. The use of publicly available data presents specific challenges that need to be considered in spatiotemporal analyses to obtain meaningful results.