Dataset and reproducible code for Firmly Rooted: Using host trees to predict new forestry pest invasions

Identifying which potentially invasive pests pose the greatest potential risk helps us prioritise prevention and early detection. However, there is often not enough location data describing the potential niche where these species occur, limiting our ability to quantify establishment risk. To address this, we developed and tested a novel quantitative approach using host tree location data in lieu of pest data to develop Species Distribution Models (SDMs) as a proxy for potential niche to predict the relative risk of establishment of potential pest species. This method was applied to pests of oak and pine host species on the island of Ireland and Sweden under current and future climate conditions, and validated using independent pest data. Our findings show that the host model methodology is effective at ranking the relative risk of pests, identifying opportunities for the method where locations have not yet been invaded and where pest data is incomplete. This dataset provides the data, code and figures for the manuscript titled: "Firmly Rooted: Using host trees to predict new forestry pest invasions." Data were made available to fully reproduce the results presented in the manuscript. Description of the data and file structure Species Distribution Models were developed for a set of oak and pine pest species using the occurrence data of their hosts as the training dataset and projected for the study areas of the island of Ireland and the country of Sweden. These models were then validated using independent pest data to assess the transferability of the host models in predicting the pest species. The data within contains the code to carry out this analysis, and produce the Figures & Tables for the associated manuscript "Firmly Rooted: Using host trees to predict new forestry pest invasions." Files and variables File: FirmlyRooted_Results_Code.R Description: Code to clean, process and present the output data from SDM_IRL.R and SDM_SWE.R as figures for the manuscript. File: SDM_IRL.R Description: Code to develop species distribution models and projections for the island of Ireland under current and future (2041-2060) climate conditions. File: SDM_SWE.R Description: Code to develop species distribution models and projections for the Sweden under current and future (2041-2060) climate conditions. File: FirmlyRooted_Figures.zip Description: Output figures from FirmlyRooted_Results_Code.R for the manuscript Firmly Rooted. File: Firmly_Rooted_Data_Output.zip Description: Outputs generated from SDM_IRL.R and SDM_SWE.R for use in FirmlyRooted_Results_Code.R to create the Figures & Tables for the manuscript Firmly Rooted as they are computationally intensive. Code/software R and R Studio was used to generate this data using R version 4.4.2 (2024-10-31) -- "Pile of Leaves" The workflow is: Input Data > SDM_IRL.R > SDM_SWE.R > FirmlyRooted_Results_Code.R > Manuscript Figures As SDM_IRL.R and SDM_SWE.R are computationally intensive, there is a break where Results_Code_Input_Data.RData can be run between steps 3 and 4 rather than waiting for extensive outputs from steps 2 and 3 to be generated. The following packages were used in the code, and should be installed prior to running. dplyr v1.1.4 readr v2.1.5 rgbif v3.8.1 data.table v1.17.0 dismo v1.3-16 tidyr v1.3.1 geodata v0.6-2 raster v3.6-31 mapview v2.11.2 tictoc v1.2.1sf v1.0-19 ggplot2 v3.5.1 Matrix v1.7-3 biomod2 v4.2-6-2 gridExtra v2.3 parallel v4.4.2 blockCV v3.1-5 tidyterra v0.7.0 tmap v4.0 terra v1.8-29 fuzzySim v4.29 ggridges v0.5.6 ggbump v0.1.0 gt v1.0.0 tidyverse v2.0.0 cowplot v1.1.3 viridis v0.6.5 MetBrewer v0.2.0 pROC v1.18.5 rnaturalearth v1.0.1 rnaturalearthdata v1.0.0 Access information Data was derived from the following sources: GBIF.org for occurrence location data worldclim.org for current and future bioclimatic variables soilgrids.org for pH and SOC data.