FLORAIDE, a new decision-making tool for weed species

A large number of plants have been introduced in North America, either accidentally or for food, medicinal or utilitarian purposes. However, most recent introductions were done to respond to an increasing demand for ornamental plants. Several of these plants could escape from gardens, naturalize, become invasive and eventually weeds. Gardeners should consider this possibility (and consequences) before introducing plant species. A new decision-making tool (FLORAIDE) can help to flag problematical plants before introduction. Using a different approach than other tools on invasive plants developed recently, FLORAIDE was developed from a database of 900 exotic naturalized plant species in Quebec (Canada). Biological and historical data were collected for each species. In addition, a committee of decision makers (professionals concerned by the problem of weeds) were asked to assign each plant to one of three weed categories: 1) presently a weed (for agriculture, horticulture, human health or environment), 2) potentially a weed (might become a weed, additional information required), or 3) not a weed (no evidence of problems associated). To establish a relation between plant characteristics and weed categories – an essential step toward a powerful predictive decision-making tool – a dominance-based rough set model and a logistic regression model were used. These models indicate that a limited number of biological (e.g., dispersal means) or historical (e.g., naturalization years) characteristics can be used to effectively predict which species might become (or not) weeds. Once available on a friendly-user website, FLORAIDE will represent a significant step toward the management of potential weeds before their introduction.