USING REMOTE SENSING TO MONITOR ECOHYDROLOGICAL CHANGES IN GREAT LAKES INTERDUNAL WETLANDS/SLACKS

Secondary interdunal wetlands or slacks develop in the Great Lakes coastal dunes when the wind scours the sand to the water table within a dune. Slacks have a limited geographic distribution and are an imperiled ecosystem along the Great Lakes. Previously we used historic aerial photographs and satellite imagery of slacks at Saugatuck Harbor Natural Area on Lake Michigan’s eastern shore to examine the relationship between the Lake Michigan water level curve and the slack’s extent. Water levels and the nature of the slack’s vegetation were found to be tied to the water levels, larger pond areas in higher water level years and smaller ponds in lower water level years. We extended these studies along Lake Michigan’s coast to Warren Dunes and Ludington State Parks, Leelanau Peninsula, Crystal Lake and Point Betsie using historic images from 1952–2018. Using QGIS we drew polygons around three types of slack features, wet sand, standing water and slack vegetation on georeferenced photos, calculating perimeters and areas for each feature on each photo. Wetland perimeters and areas were summed annually for each feature. These values were then totaled on an annual basis (wetland extent) and compared to the Lake Michigan water level curve. Overall the wetland extent in Warren Dunes, Leelanau and Crystal Lake reflects the lake level curve, increasing as water levels rise and decreasing as they fall. Wetland extents at Point Betsie and Ludington follow the lake level curve with the exception of 1965 and 1993 at Point Betsie, and 1993–2012 at Ludington. Possible explanations being investigated for these variances include photo quality and/or a rain event(s). Overall this method provides a reliable way to evaluate historic ecohydrological changes in response to fluctuating lake levels. 3-band (red, green, near-infrared) drone imagery has also been used as lake levels rise (2017, 2018, and 2019) to examine the ecohydrological changes occurring in a large slack (>1 ha in size) at SHNA where ridges and pools create a diverse vegetation mosaic. Vegetation quadrat sampling (32 quadrats) was used to ground-truth the image interpretations. Vegetation indices using these data provide even greater detail on the ecological changes in the slack.