Forecasting Invasions using Species Traits
Identifying potentially invasive species and preventing their introduction and establishment are of critical importance in invasion ecology and land management. Although an extensive body of research has been dedicated to identifying traits that confer invasiveness, our current knowledge is still often inconclusive due to limitations in geographic extent and/or scope of traits analyzed.
Using a comprehensive set of 45 traits, we performed a case study of invasive traits displayed by exotic woody plants in the United States (U.S.) by comparing 63 invasive and 794 non-invasive exotic woody plant species naturalized across the country.
Through a combination of multivariate and machine-learning methods, we were able to identify six key traits that can consistently predict invasiveness of exotic woody species across the United States. The ability to reproduce vegetatively in the wild and long-distance dispersal (via water, birds, mammals) were traits consistently associated with invasiveness in exotic woody species. Invasive and non-invasive woody species also differed in primary growth form, with invasive species displaying a higher proportion of lianas and a lower proportion of trees than non-invasive species. Boosted classification tree models based on these traits accurately predicted species invasiveness (86% accuracy on average).