Bayesian Networks to Compare Pest Control Interventions on Commodities Along Agricultural Production Chains.

Area of Science:

• Agricultural Science
• Phytosanitary Risk Assessment
• Computational Biology

Background:

• Agricultural commodity production involves multiple stages from planting to export.
• International trade relies on agreed-upon phytosanitary measures to prevent pest introduction.
• Quantifying pest infestation risk and control measure efficacy is crucial for trade facilitation.

Purpose of the Study:

• To develop a Bayesian network model for assessing potential pest infestation risk in agricultural commodities.
• To facilitate agreement between exporting and importing nations on effective phytosanitary measures.
• To provide a framework for comparing pest control strategies and communicating risk reduction.

Main Methods:

• Development of a Bayesian network representing pest infestation dynamics along the production chain.
• Incorporation of factors: pest challenge, commodity susceptibility, control measure impact, and efficacy uncertainty.
• Case study application using a fruit fly pest on dragon fruit in Vietnam to illustrate model functionality.

Main Results:

• The Bayesian network models the dependency of infestation levels on various production and control factors.
• The model allows for scenario analysis, comparing different pest control combinations and their performance.
• Illustrative results demonstrate potential infestation scenarios (most likely, alternative, worst-case) under specific measures.

Conclusions:

• The Bayesian network provides a quantitative framework for evaluating pest control measures in agricultural trade.
• It supports informed decision-making and communication between trading partners regarding phytosanitary risk.
• The model aids in achieving target levels of protection against regulated pests, facilitating international trade.