Common buckthorn(Rhamnus cathartica) and glossy buckthorn(Frangula alnus) are invasive shrubs in North America that grow along forest edges and wetlands. Remote detection of this species is typically done using NIR aerial/satellite imagery. Spatial resolutions of 1-10m are high enough for detecting large patches of the shrub but too low to capture individual plants. Our study tested the utility of high resolution drone imagery to accurately map buckthorn distributions. Buckthorn’s unique boundary habitat, growing location, growing season, growth height, and color all assisted in mapping the species across Tobico Marsh in Saginaw Bay, MI. Drone imagery was collected in late October, a window in which deciduous trees in the area had begun to senescence while buckthorn stays a bright shade of green late into the season. The drone flight collected Red, Green, and Blue bands of imagery across the marsh and was processed to produce a DEM. Band indices were tested to determine which ones would most accurately detect buckthorn. Hue was selected to pull out the shade of green representing buckthorn in the imagery, ERGBVI Index separated green vegetation from senescence, and DEM was used to extract vegetation height representative of buckthorn(~3-7m). The resulting map was created using a thresholding technique across the selected bands and field reports indicate that the map detected buckthorn down to the individual plant, though misidentifies Honeysuckle and Tag Alder as buckthorn.
