Summary: | 碩士 === 國立中興大學 === 農藝學系所 === 102 === Alien plant invasion threatens biological diversity and native environment, which impacts economic development in agriculture and forestry. To prevent introduction of alien plants into Taiwan and likely to become invasive species, establishment of the weed risk assessment (WRA) system predicting the invasiveness of alien species is very important. The main purpose of this study is to improve current WRA system in Taiwan. After comparing four WRA systems conducted by Australia, UK, USA and New Zealand, an appropriate candidate of WRA system for Taiwan has been proposed. By using 25 invasive plants and 11 non-invasive plants in Taiwan as database, the predictabilities of four WRA systems were compared based on the receiver operating characteristic (ROC) curve analysis model, and the Australian WRA (AWRA) system was selected. Subsequently, AWRA system was simplified by answering 10 questions only, which are critical and easy to answer. A modified AWRA system consisted of 10 basic questions, coupled with a score weighting process, was proposed, which has a high prediction ability to distinguish the invasive plants from non-invasive plants in Taiwan.
|