Populations of Microstegium vimineum (Japanese stiltgrass) located adjacent to closed canopy forests in West Virginia were monitored for three years (2005-2007) to estimate spread rates and population growth.  Microstegium vimineum is a shade tolerant, invasive, annual grass found throughout the mid-Atlantic area of the United States. Five populations were monitored in the Ridge and Valley Province (RV), five in the Allegheny Plateau Province in Tucker County (Tucker) and six in the Allegheny Plateau Province in Preston County (Preston).  Both Tucker and Preston are more mesic than RV, with Tucker having some of the most productive and diverse flora of the three areas.  Patches (1 m² or less in size) of M. vimineum were counted within a 30 m (parallel to the roadside population) X 50 m (perpendicular to the roadside) area or a total of 1500 potentially colonized patches. Stem counts per patch were also determined.  The invasive front (distance from the roadside into the forest interior) for each year was determined using two transects and measuring the distance at which M. vimineum patches were no longer contiguous along these transects.  Variables were compared using a mixed model analysis.

Results/Conclusions

Stem counts per patch were not significantly different among the populations.  Patch colonization and extinction rates were not constant over time, and extinction rates did not differ significantly among the three areas.  Tucker populations had significantly higher colonization rates than RV populations in 2006.  The front line was significantly further from the roadside in the RV populations than the Tucker populations in all three years and the Preston populations in 2007.  However, the RV populations had the lowest net reproductive rate (R₀), with some populations showing a negative rate.  Tucker populations had a significantly faster R₀ than both Preston and RV populations from 2005 to 2006.  Using  R_0(2005-2006) to estimate the intrinsic rate of growth  and assuming exponential growth, complete colonization of all 1500 patches would be expected in 12.0 and 11.1 years for the Preston and Tucker populations, respectively, and saturation of the patches would never occur for the RV populations. These results suggest that M. vimineum populations in drier regions may spread primarily along its front but grow at a slow rate with few disconnected patches.  In contrast, the populations in the mesic sites tend to grow at a more rapid rate via more abundant, disconnected patches.