PARENT SESSION
Poster Session # 16: Invasive Species.

Wednesday, August 6 Presentation from 5:00 PM to 6:30 PM. SITCC Exhibit Hall B.

Predicting early plant invasions in West Virginia public forests.

Huebner, Cynthia ^*,1, ¹ NE Research Station, USDA Forest Service, Morgantown, WV, USA

ABSTRACT- Western West Virginia public forests have experienced comparatively less invasion than forests in more urban regions and may serve as an optimal landscape for modeling early establishment (as opposed to spread) of exotic invasive plants. Documentation of the importance of disturbance and environmental variables as predictors of invasion in forests is lacking. I sampled the understory (herbs, shrubs, vines and tree seedlings) of 24 undisturbed (80 or more years old) and 24 disturbed (15-year old clear cuts) sites in the Monongahela National Forest Cheat Ranger District. Sites, which averaged 3 ha in size, were randomly selected such that three common ecological land types (sugar maple/basswood, sugar maple/red oak, and red oak dominated forests) were equally represented in both undisturbed and disturbed sites. Plant species were grouped into exclusive categories: exotic invasive, exotic non-invasive, native invasive, native non-invasive weed, and native non-invasive non-weed. Eleven environmental and 7 disturbance variables were measured. Data were analyzed using logistic regression analysis. In all sites, exotic invasive weeds had low ranking relative importance values, with Rosa multiflora and Berberis thunbergii being the most common among the 11 exotic invasives present. High herb/shrub/vine richness and diversity values and high levels of native non-invasive weeds and exotic non-invasive weeds best predicted invasion by exotic invasives in the undisturbed forests; only high herb/shrub/vine richness was a reliable predictor of invasion in disturbed sites. Undisturbed sites were more likely to be invaded the closer they were to a paved road. In the disturbed sites, northeast-facing slopes, lower moss/lichen cover, higher bare ground cover, shallower slopes, lower light levels, and shorter distances to gravel and paved roads best predicted invasion. A combined analysis showed that disturbed sites and sites with northeast facing slopes were most likely to be invaded. These results may be used with caution to predict invasion into other similar forests and indicate that, while disturbance is an important predictor, sites with certain environmental characteristics are more prone to invasion.

Key words: forest invasion, logistic regression, invasive exotic plants, predicting invasions