The ability to predict the successful invasion of plant species into newly disturbed habitats has the potential to substantially increase the efficiency of early detection of nascent populations of key invaders. The integration of landscape ecology and predictive habitat modeling is a promising research area that may permit the prediction of large scale patterns of invasion. These techniques enable researchers to use landscape-scale predictor variables to model, statistically and spatially, habitat characteristics of invaded sites and then to estimate future sites of invasion by identifying suitable habitat across a landscape of interest. The present study sought to develop statistically based habitat models for anticipated invasion by Imperata cylindrica (cogongrass) following large-scale natural disasters.
Results/Conclusions
Data from field surveys for I. cylindrica were analyzed in the context of local- and landscape-scale data on environmental characteristics, including distance to roads, type of road, presence and type of local disturbance, and change in forest cover associated with Hurricane Katrina (August 2005). Analyses indicated that forest type and forest cover change between 2004 and 2006 were not statistically informative in predicting the presence of