Prediction of the spatial distribution of vine weevil under climate change using multiple variable selection methods

Area of Science:

• * Ecology and Biogeography
• * Agricultural Entomology
• * Climate Change Impact Assessment

Background:

• * The vine weevil (Otiorhynchus sulcatus) is a significant polyphagous pest impacting numerous economically important crops globally.
• * Previous research has not comprehensively assessed the potential global distribution of this invasive pest.
• * Understanding pest distribution is crucial for developing effective agricultural management and biosecurity measures.

Purpose of the Study:

• * To predict the potential global distribution of the vine weevil under current and future climate change scenarios.
• * To evaluate the impact of different variable selection methods on species distribution model (SDM) accuracy.
• * To identify regions with the highest invasion risk for the vine weevil.

Main Methods:

• * Development of multiple species distribution models (SDMs) using the MaxEnt algorithm.
• * Application of diverse variable selection techniques: correlation, biological considerations, and principal component analysis.
• * Integration of individual SDMs into an ensemble model for robust prediction and ensemble modeling.

Main Results:

• * Ensemble SDMs using raw bioclimatic variables demonstrated robust performance (TSS 0.34–0.37, F1 score 0.60–0.67).
• * Reconstructed variables yielded lower model performance (TSS 0.13, F1 score 0.48) and divergent distribution patterns.
• * Predicted primary distribution in North America and Europe, with highest invasion risks identified in Far East Asia and northern India.

Conclusions:

• * The vine weevil's potential distribution is significantly influenced by the choice of predictor variables and modeling approaches.
• * Climate change is projected to facilitate northward habitat shifts for the vine weevil, particularly in existing ranges.
• * Human activities may contribute to the spread of the vine weevil into new geographical areas, necessitating enhanced monitoring and control efforts.