Project Number: 3625-21000-053-02
Project Type: Specific Cooperative Agreement

Start Date: Aug 23, 2007
End Date: Jul 31, 2010

Objective:
The objective of this cooperative research project is to produce valid risk-assessment models that can be applied to effective risk-management systems by all parties evaluating new woody-plant introductions. Three goals have been established to meet this overall objective: (1) to refine predictive models that integrate analyses of biological characteristics as well as native-range distributions to assess the relative risk that non-native, woody plants introduced for horticultural purposes will naturalize (survive and reproduce outside of cultivation) at a regional scale in the North Central US; (2) to expand upon our refined naturalization models and test the application of these same approaches for extracting biological characteristics and geographic data that are correlated with a non-native species' likelihood of becoming a invasive pest; and (3) to ensure that the resulting models are accurate and practical through a validation process that is both statistically sound and involves field personnel and pertinent decision-makers.

Approach:
We will conduct a three-year study through support of a post-graduate student at Iowa State University, which will allow us to evaluate the power and accuracy of three existing models to predict the likelihood that non-native woody plants can naturalize in Iowa. First we will test the existing models for Iowa species, and then expand the geographic analyses and population of test species to include two more states in the North Central region, Minnesota and Missouri. In addition, we will collect published data and the observations of land managers about negative environmental effects to support the coding of the non-native species to account for known invasiveness. We will use these additional data to develop a regional model to predict naturalization and to examine the factors that predict invasiveness. Important questions to be addressed include: Will the models developed in Iowa also serve to classify species accurately when using a broader regional data set? If so, which of the models will be most powerful and accurate? If not, can the models be easily modified to operate on a regional scale?