Many unreported crop pests and pathogens are probably already present

Area of Science:

• Agricultural Science
• Ecology
• Biosecurity

Background:

• Invasive species pose significant threats to global biodiversity, food security, and agricultural productivity.
• Effective pest risk analysis (PRA) is crucial for agricultural biosecurity but is hindered by incomplete data on pest and pathogen distributions.
• Existing global pest distribution databases often overlook valuable information from regional literature, such as the Chinese academic corpus.

Purpose of the Study:

• To develop and validate predictive models for current crop pest distributions at a subnational resolution.
• To assess the utility of regional literature, specifically the Chinese literature, for improving pest distribution data.
• To identify regions with a high probability of currently unobserved, but likely present, invasive pests.

Main Methods:

• Generalized linear models (GLMs) were employed to estimate pest presence probabilities for 1,739 pests using the CABI pest distribution database.
• Model predictions were validated against 100 newly observed pest occurrences in the People's Republic of China (PRC), sourced from Chinese literature.
• Predictive models were then applied globally to estimate occurrences of unobserved pests.

Main Results:

• Models demonstrated good predictive performance for pest presence in PRC provinces, with Area Under the ROC Curve (AUC) values of 0.75-0.76.
• Key predictors for pest presence include host availability, proximity to neighboring regions, per capita GDP, and global prevalence.
• A significant number of unreported pests (predicted presence probability >0.75) were identified in China, India, southern Brazil, and former USSR countries.
• The study successfully predicted pseudoabsences at a subnational level, highlighting the value of previously inaccessible data.

Conclusions:

• Generalized linear models are effective tools for predicting invasive pest presences and quantifying pseudoabsences, thereby enhancing pest risk analysis.
• Integrating regional literature, such as the Chinese academic corpus, significantly improves the accuracy and completeness of global pest distribution databases.
• The findings provide crucial insights for prioritizing biosecurity efforts and managing invasive species in agriculture worldwide.