PARENT SESSION
Contributed Oral Session 158: Invasive Species: Management, Modeling, and Dynamics
Friday, August 12, 8:00 AM - 11:30 AM, Meeting Room 518 A, Level 5, Palais des congrès de Montréal

Plant species invasions: latitudinal gradients and filling in the gaps in the United States.

Crosier, Catherine^*,1, Stohlgren, Thomas¹, Barnett, David², Flather, Curtis³, Kartesz, John⁴, ¹ National Institute of Invasive Species Science, Fort Collins, CO² Natural Resource Ecology Laboratory, Fort Collins, CO³ Rocky Mountain Research Station, Fort Collins, CO⁴ Biota of North America Program, Chapel Hill, NC

ABSTRACT- Knowledge of patterns of native species richness, an important component of diversity, and non-native species invasions is lacking even though this knowledge is essential for conservation efforts. In addition, non-native species mitigation and control are very costly in the U.S. We cannot afford to wait for complete information on the distribution and abundance of harmful invasive species. We used Akaike^'s Information Criterion (AIC) to evaluate competing models predicting native and non-native plant species density (number of species/ km² in a county) from various combinations of biotic variables, environmental/topographic variables, and human variables. A previous study examined models using vascular plant data from 3000 counties to examine latitudinal gradients. There was little evidence of a latitudinal gradient. However, results may have been compromised by poor reporting in several counties. In this study, we used information from 192 counties in the coterminous U.S that were well-surveyed for vascular plant species to develop models to fill data gaps in poorly surveyed areas. The best model for native species density was the model including environmental/topographic and human variables was the best (AIC = -592; adj r² = 0.81), and for non-natives was a non-linear regression model of native plant density (AIC = -926.2; adj r² = 0.92). We showed that native species density is non-random and predictable, and that native species density was the best predictor of non-native species density. We found that the eastern U.S. agricultural and coastal areas are among the most invaded, and that the central U.S. appears to have the greatest proportion of non-native species relative to native species. These types of large scale models could also be applied to other taxa to set priorities for conservation and invasion mitigation, prevention, and control efforts across the U.S.

Key words: invasive species, modeling, biodiversity, species richness