Predicting island biosecurity risk from introduced fauna using Bayesian Belief Networks

Area of Science:

• Ecology
• Conservation Biology
• Biosecurity

Background:

• Islands harbor unique endemic species but are vulnerable to invasive species due to human activities.
• Inefficient surveillance strategies hinder the detection and management of invasive species on islands.
• Prioritizing surveillance sites and identifying high-risk invasive species are critical biosecurity challenges.

Purpose of the Study:

• To develop and apply a risk assessment framework for prioritizing island surveillance against invasive species.
• To create automated, species- and site-specific biosecurity models for invasive species risk assessment.
• To identify key factors influencing the arrival and establishment of invasive fauna on islands.

Main Methods:

• Development of Bayesian Belief Networks (BBNs) using Java and GeNIe software.
• Integration of data on island attributes, visitor load, infrastructure, habitat, and dispersal mechanisms.
• Application of BBNs to assess the risk of 11 invasive faunal species establishing on 600 islands.

Main Results:

• Propagule pressure was the most influential factor for species arrival.
• Visitor numbers and animal swimming capabilities significantly impacted model outcomes.
• Models predicted an average of one invasive species arrival per 300 visitors across the studied species.

Conclusions:

• The developed biosecurity BBN is effective for identifying high-risk islands and invasive species within archipelagos.
• This approach enables efficient prioritization of surveillance activities in data-deficient regions.
• The BBN framework supports strategic biosecurity planning to protect island ecosystems.