COS 302-3 - CANCELLED - Incorporating weather dynamics instead of climate suitability in forecasts of biological invasions improves model performance and realism.

Thursday, August 10, 2023

4:00 PM – 4:15 PM PDT

Location: D138

Presenting Author(s)

CJ

Chris M. Jones

North Carolina State University Center for Geospatial Analytics, North Carolina, United States

Coauthor(s)

BS

Ben Seliger

North Carolina State Center for Geospatial Analytics, North Carolina, United States
AP

Anna Petrasova

North Carolina State Center for Geospatial Analytics, North Carolina, United States
VP

Vaclav Petras

North Carolina State Center for Geospatial Analytics, North Carolina, United States
RM

Ross Meentemeyer

North Carolina State University, North Carolina, United States

Abstract: Emerging and invasive pathogens can cause loss of ecosystem services and diversity, threaten food security, and are difficult to manage. Given the threat these pathogens pose, models that predict pathogen suitability in areas outside the infected area are used to make policy decisions regarding surveying and treatment efforts. If these models are over- or under-predictive, it could result in expensive survey and management efforts, or stakeholders could be left unprepared. These models typically use species distribution models (e.g., MaxEnt) based on climate and host availability. Climate-suitability models frequently ignore the fact that weather events are the primary factor driving pathogen spread and establishment. For many pathogens, having precipitation during specific temperature ranges is the key to sporulation and, thus, spread and establishment. This study aims to demonstrate the importance of considering weather when modeling potentially suitable areas for pathogen establishment. We use the Pest or Pathogen Spread (PoPS) model to simulate the establishment potential of the EU1 lineage of Phytophthora ramorum in the conterminous US calibrated and validated from 2015 to 2021 in Oregon. PoPS simulates reproduction, dispersal, and establishment based on weather, host, and environmental conditions, allowing for predictions of when and where pathogens occur. We use nursery locations and assume all nurseries have infected hosts of P. ramorum that are allowed to sporulate for ten years whenever weather conditions are suitable. We use Daymet daily data to calculate weekly temperature and precipitation coefficients from 2011 to 2020. Given the uncertainty of hosts outside of California and Oregon to support P. ramorum sporulation, we use all hosts potentially capable of sporulation coupled with a range of host competency scores (ability to sporulate P. ramorum) for each species based on expert solicitation. We ran 10,000 simulations and calculated the number of simulations out of 10,000 that establishment occurs outside of the initial nursery infections to determine the potential risk of pathogen establishment in each cell. We compare our model to previously published models of P. ramorum suitability. We show that the total area suitable for P. ramorum establishment in the Eastern US is much lower than previously predicted and that the most at-risk locations are different than previously predicted. In conclusion, using weather drivers, instead of climate suitability, to determine the potential for forest pathogen establishment creates a more realistic understanding of the suitable area for emerging and invasive pathogens that can lead to better science-based decision-making.