Predicting Weed Establishment with Open Access Data

We have a new paper out in the Proceedings of the Royal Society B. Laure Gallien leads the project which uses open-access databases to develop indicators to predict the establishment success of alien species. In it we use our Acacia and Eucalypt phylogenies (available at Treebase and OpenTree) along with nearly 500,000 specimen occurrences records from the Australasian Virtual Herbarium (all available at GBIF). Laure's analysis indicates that with open-access databases we can successfully identify key features that predispose Australian plants to naturalize.

The four indicators that predict establishment are:

1. potential global range size

2. niche overlap with human-disturbed environments

3. number of palaeoinvasions

4. capacity for local dominance

From a phylogenetic view the number of paleoinvasions is intriguing. This means that lineages that have a history of dispersal across biogeographic boundaries are more likely to become established weeds. We provide an objective and cost-effective protocol for flagging high-risk introductions.