Using photointerpretation and publicly available software to assess revegetation ground cover on a shaded hillside
This project examined the potential of three plant species, Heuchera micrantha, Polystichum acrostichoides and Pachysandra procumbens individually and in various combinations as potential ground cover for a heavily shaded north-facing slope on the Maymont Estate in Richmond, Virginia. Digital photographs were taken of the planting sites. Two software applications were used to make quantitative evaluations of ground cover for the test plantings: the USDA's point sampling application SamplePoint, and NIH's image analysis application ImageJ. GNU's Image Manipulation Program was used to edit the digital images as needed to prepare them for interpretation and OpenOffice.org was used to analyze data. Limiting factors for plant growth on these sites included topography and a dense tree canopy, which obstructed sunlight (May – September) and produced mats of leaf litter (October – March) which hindered quantification of plant growth. In order to compare the two image evaluation methods, individual plant species' ground cover averages recorded by ImageJ were summed to compute overall ground cover, and they were compared to the overall ground cover recorded in SamplePoint. The two ground cover evaluation methods gave similar results. The SamplePoint method can be applied by users with less technical training. However, due to consistent image quality and appropriate algorithms, ImageJ has great potential as a vegetation quantifier. SamplePoint results from August to May indicated H. micrantha initially covered an average of 6% in its quadrats in August 2009. May 2010 showed the average area had declined to 2%, while P. procumbens ground cover increased by from 3 to 4.5% and P. acrostichoides cover increased from 11 to 16.5%, making it the most effective species. Based on this study, P. acrostichoides will be used to revegetate the area, perhaps in combination with P. procumbens.