Predicting global ascidian invasions

Abstract

Aim: Many species of ascidians are invasive and can cause both ecological and economic losses. Here, we describe risk assessment for nineteen ascidian species and predict coastal regions that are more vulnerable to arrival and expansion. Location: Global. Methods: We used ensemble niche modelling with three algorithms (Random Forest, Support Vector Machine and MaxEnt) to predict ecologically suitable areas and evaluated our predictions using independent (area under the curve—AUC) and dependent thresholds (true skill statistics—TSS). Environmental variables were maximum and the range of sea surface temperature, mean salinity and maximum chlorophyll. We used our niche modelling results and a modified invasibility index to compare invasion risk among 15 coastal regions. Results: Currently, the most invaded regions are in temperate latitudes of the Northern Hemisphere and Temperate Australasia, which are regions most prone for new invasions. In the tropics, the West and Central Indo-Pacific are two regions of strong concern, the former with high risk of primary invasion by Botryllus schlosseri and Didemnum perlucidum. In the Southern Hemisphere, the Southwest and Southeast Atlantic are most at risk, both subject to invasion by Botrylloides violaceus, Didemnum vexillum, Molgula manhattensis and Styela clava among others. Regions most at risk of expansion of established invasive species are the Central Indo-Pacific, Northwest Pacific, Mediterranean and West Indo-Pacific. Main conclusions: All regions studied have areas that are suitable and connected to receive new ascidian introductions or that may permit the spread of already established species. Risk comparison of primary introductions and expansion of established introduced ascidians among regions will allow managers to prioritize species of concern for each region both for monitoring future introductions or to enforce control actions towards established species to decrease the risk of regional expansion.