Predicting change in invaded urban brownfield plant communities with remote sensing methods

Background/Question/Methods

We strive to develop and validate a non-destructive, combined remote sensing /transect sampling protocol that aims at the prediction of future vegetation changes in the face of invasion. Focusing on a unique field site, the naturally assembled urban “wild lands” on polluted brownfields of Liberty State Park (New Jersey), we tested whether measurements of primary productivity in border zones between interacting plant stands will allow predictions of future vegetations change.  In particular we tested whether remote sensing of border zones between stands native and non-native herbaceous plant showed a correlation between high productivity and future community change. Biophysical and reflectance measurements in transects running through transition zones between the non-native Artemisia vulgaris (mugwort) and the native Solidago canadensis (golden rod) were taken in two consecutive years (2006 and 2007) to study the effect of competition on plant primary production.  As biophysical characteristics, canopy cover, relative cover of both species, plant height and Leaf Area Index were measured.  Reflectance measurements were recorded using a handheld FieldSpec Pro Full Range spectrometer and vegetation indices (NDVI, SAVI) were calculated using the smoothed data set.
Results/Conclusions

Clear changes of productivity as measured by reflectance were recorded. Peaks of reflectance appeared to be correlated to community changes to be detected in the following year. Both NDVI and SAVI showed strong correlation with Artemisia and Solidago cover. Where Artemisia was dominant the productivity was relatively low and high where Solidago was dominant. In the following year Artemisia intruded into the formerly solid Solidago stands and both NDVI and SAVI showed significant decrease in productivity in the transition zone.  One possible interpretation is that as long as Solidago holds against Artemisia, its integrated clones allocate more energy into the border zone.  However, when one species out-competes the other, the latter might withdraw energy from the border zone, preserving more production capacity to be allocated in sections where the species is still dominant. These results and additional data from previous investigations suggest that (a) high productivity in border zones as measured by reflectance measurements correlates with community change and (b) border zones between native plant species tend to be less extensive and less dynamic then native – exotic plant pairs. The hypothesis follows that high productivity in border zones is a sign of intense competition that leads to community change.