Validation and uncertainty analysis of the match climates regional algorithm (CLIMEX) for Pest risk analysis

Abstract

Abstract Pest risk analysis (PRA) is conducted by plant health authorities to identify and implement phytosanitary measures to limit the accidental introduction and spread of species that are, or may become, pests. The ‘Match Climates Regional’ (MCR) algorithm is part of the CLIMEX species distribution modelling package and is a simple climate matching method that has often been applied in PRA to estimate the potential geographic distributions of weeds, insects and mites. However, there is a lack of studies that address its predictive validity and sensitivity to the input data. Here we investigated the validity of the MCR algorithm by comparing its predictions to empirical observations of the distributions of 30 species of biosecurity concern. We also evaluated MCR s sensitivity to its inputs, using data for a further 30 species, by creating numerous models for each species using different sized subsets of their distribution data, then comparing the results. We found that the algorithm showed overall high accuracy and its outputs were relatively insensitive to sample size. Thus, the results generally supported use of MCR for PRA.