Climate matching models for Ceratapion basicorne (Coleoptera: Apionidae), a biocontrol agent of yellow starthistle

Access this dataset on Dryad: https://doi.org/10.5061/dryad.vmcvdnd2w

👥 Brittany Barker

Questions? 📧 [email protected]{.email} or [email protected]{.email}

Purpose

The code and files in this repository produce climate matching models for Ceratapion basicorne for the western U.S. (Barker et al. 2025).

Abstract

Ceratapion basicorne (Illiger) (Coleoptera: Apionidae), a weevil native to Europe and western Asia, shows promise for enhancing the control of yellow starthistle (Centaurea solstitialis L.), an invasive annual forb in the western U.S. However, a paucity of data on this biocontrol agent’s environmental constraints has made it difficult to assess the suitability of potential release locations. Climate matching models were developed for C. basicorne to help identify areas of the western U.S. with similar climates to the source area of breeding colonies being used for releases (home location). The models used climate variables derived from daily estimates of minimum temperature (Tmin), maximum temperature (Tmax), precipitation, and soil moisture for a 30-year period spanning 1991−2020 at 1-km^2^ resolution. Models indicated that the Central California Foothills, Eastern Cascades Foothills, Columbia Plateau, and mountainous parts of northcentral Utah had the most similar climates to the home location. Of these areas, the Eastern Cascades foothills in northeastern California and Wasatch Range in Utah occurred at a similar latitude as the home location, which may be important to consider if C. basicorne has photoperiodic diapause. The least similar climates occurred in very wet coastal areas, high-elevation (cold) mountains, and hot deserts. The development of process-based models for predicting the establishment of this agent will require a more detailed understanding of the agent’s requirements for development and survival.

Description of the data and file structure

Subfolders

The main folder (directory) contains the following subfolders:

• data: all data needed for modeling (see descriptions below).
• plots: image files (.png files) produced by models are saved here. Images produced by the CLIMEX-based model and Climatch models are saved to the CLIMEX_custom and Climatch folder, respectively.
• raster_outputs: rasters (.tif files) produced by models are saved here. Rasters produced by the CLIMEX-based model and Climatch models are saved to the CLIMEX_custom and Climatch subfolder, respectively.
• script: running the R script CEBA_climMatch.R re-produces all models, figures, and rasters presented in the manuscript (see below).

Data

Climate data for modeling: the CLIMEX-based model uses weekly climate data (/data/weekly/) whereas the Climatch model uses six bioclimatic variables (/data/bioclim/). For each model type, the data are split according to the 'home' location (Kilkis, Greece) and 'away' locations in the western U.S. Averages of climate for a 30-year period spanning 1991-2020 were derived from the Daymet and E-OBS datasets. In raster datasets, all locations in the western U.S. (extent: xmin = -125.0024, xmax = -101.9945,ymin = 31.18652, ymax = 49.40551) are at 1-km2 resolution and the coordinate reference system is WGS84 (EPSG:4326). Units are degrees Celsius for temperature (Tmin, Tmax), millimeters for precipitation, and m3 m-3 for soil moisture. Further details about the climatic data and variables used in models can be found in the publication for this study.

A summary of these datasets is below:

• /data/weekly/home/kilkis.csv: Data for the 'home' location used for the CLIMEX-based model. The columns correspond to averages of weekly climate for Tmin, Tmax, and soil moisture (sm) (52 weeks x 3 variables = 156 columns) (e.g., ppt01 = precipitation of the first week of the year), as well as average annual total precipitation (mm) (ppt_ann).
• /data/weekly/away/*tif: Data for the 'away' locations used for CLIMEX-based models. There are 156 rasters corresponding to weekly estimates of Tmin, Tmax, and soil moisture (sm) (52 weeks x 3 variables = 156 rasters), as well as an annual estimate of precipitation (ppt_ann).
• /data/bioclim/home/kilkis.csv: Data for the 'home' location used for the Climatch models. The columns correspond to bioclimatic variables derived from 30-year averages of climate. These included Tmax of the warmest month (bio5), Tmin of the coldest month (bio6), annual precipitation (bio12), highest monthly soil moisture (bio29), and lowest monthly soil moisture (bio30). The row corresponds to the source location for C. basicorne in Kilkis, Greece (latitude = 22.844, longitude = 40.994).
• /data/bioclim/away/*tif: Data for 'away' locations used for Climatch models (same bioclimatic variables as the home location).
• /data/YST_counties/YST_counties_9-19-24.shp: Geospatial vector data of counties in the western U.S. where yellow starthistle has been observed. Observations were obtained from the Early Detection and Distribution Mapping System (EDDMapS, 2024) and the Global Biodiversity Information Facility (GBIF.org, 2024). A single record was retained for each county, for a total of 182 records from 11 states. The geometry type is a multipolygon (extent: xmin = -125.0024, xmax = -101.9945,ymin = 31.18652, ymax = 49.40551) and the coordinate reference system is NAD83. The columns include: STATEFP (state FIPS code), COUNTYFP (U.S. county FIPS code), NAME (county name), geometry (longitude, latitude), n_number (number of records), and present [if starthisle has been detected (present) or not (absent)].

Sharing/Access information

Data were derived from the following sources:

• Daymet dataset for North America
• E-OBS dataset for Europe
• SiTHv2 global soil moisture dataset
• EDDMapS database
• GBIF | Global Biodiversity Information Facility

Citations for datasets can be found in the publication (Barker et al. 2025)

Code/Software

The R statistical software (version 4.3.2) was used to produce models.

Required R packages

The following packages must be installed: here, tidyverse, ggspatial, rnaturalearth, sf, terra, tidyterra, Euclimatch, GA, cowplot.

Instructions

(1) Clone the repository (or download the directory from Dryad). Don't move or delete any subfolders or datasets.
(2) Open the R project (CEBA_climMatch.Rproj) in RStudio/Posit.
(3) Open R script named CEBA_climMatch.R in the script subfolder.
(4) Install any necessary packages as listed above.
(5) Run the R script, which produces the climate matching models and associated outputs.

Acknowledgements

This work was funded by the U.S. Department of Defense Strategic Environmental Research and Development Program (U.S. Army Corps of Engineers, contract no. RC23-3611).

References

Barker, B. S. 2025. Climate matching models for Ceratapion basicorne (Coleoptera: Apionidae), a biocontrol agent of yellow starthistle. Journal of Economic Entomology. http:/doi.org/10.1093/jee/toae299.

EDDMapS. 2024. Early Detection & Distribution Mapping System. The University of Georgia - Center for Invasive Species and Ecosystem Health. Available online at: http://www.eddmaps.org (accessed 19 September 2024).

GBIF.org. 2024. (19 September 2024) GBIF Occurrence Download https://doi.org/10.15468/dl.sqmyns.