Weed risk assessment validation: the Australian system and its potential use on introduced plant species in Chile

Abstract

Assessing the status of introduced species already established in the country is crucial to avoid new invasions. In this study, we validated a modified version of the Australian Weed Risk Assessment (AWRA) for application in Chile. We calculated invasion risk assessment scores for 369 species, adapting the AWRA protocol for Chile (WRA-Ch). We used expert ranking, and the Receiver Operating Characteristic (ROC) curve analysis method to validate. We calculated WRA-Ch scores for introduced species from the five categories proposed by Matthei (1995). Of the total species, 2 were accepted, 339 were rejected, and 28 resulted in second evaluation status. The expert categories differed significantly in the average WRA-Ch value. Major weeds and very serious weeds are the categories with the highest values. When evaluating the ability of the WRA-Ch to predict the invasive or non-invasive status of a species, the area under the ROC curve was 0.96, indicating an excellent ability of the WRA-Ch to differentiate between invasive and non-invasive plants. The validation of WRA-Ch for use in Chile represents an important step in preventing the introduction of new species. We suggest that WRA-Ch be used by government agencies in charge of biodiversity monitoring at the national level and border control. Validating WRA-Ch responds to the need to implement a national system of invasion risk analysis for species introduced into the country and those that are intentionally introduced or intercepted pre-border.