Assessing invasive plant infestation and disturbance gradients in a freshwater wetland using a GIScience approach

The assessment of aquatic ecosystems requires information on biological and disturbance gradients in order to evaluate quality. As a result decision makers need improved monitoring tools for characterizing relationships between invasive species infestation and disturbance to make informed choices regarding wetland condition and management plans. The overarching goal of this research was to assess invasive plant infestation and disturbance gradients using a GIScience approach. The study was conducted in a fresh-water, coastal wetland in the Muskegon River watershed, Michigan, USA. Airborne hyperspectral imagery (20 bands, 440-880 nm) was classified for Phragmites australis distribution using the Spectral Angle Mapper algorithm. Indicator semivariograms were utilized to define landscape structure and associated spatial scales, and assist in creating a transect scheme to generate landscape pattern metrics quantifying valued ecosystem attributes. Hydrological modifications, as measured by an area-weighted fractal dimension index, served as a proxy for human disturbance and was found to moderately influence Phragmites percent cover (R² = 0.4, n = 40), mean patch size (R² = 0.5), and patch shape (R² = 0.5). A general conclusion was that increased hydrological disturbances were correlated with increased infestation magnitude. The systematic approach executed in this study outlined how geospatial monitoring tools can be used as an assessment framework to provide more meaningful information that lends itself to comprehensive wetlands assessment.