Modeling Geomorphic Effects on Eelgrass Before and After Restoration, Nisqually Delta, Washington

Abstract

Investigating native eelgrass (Zostera marina) distribution before and after a dike removal and estuary restoration in the Nisqually Delta, South Puget Sound, Washington, USA, through a comparison of mapping methods, provides a visible measure of geomorphic effects. Research aims were to use two mapping methods, image classification and digitization, to compare accuracy and efficiency at determining change in eelgrass distribution over time as a result of the 2009 Nisqually Delta dike removal and estuary restoration. Substrate analyses were used to show change before and after dike removal. Results found that supervised classifications of 2009, 2010, and 2011 false color near infrared aerial imagery of the tidal flats underestimated eelgrass presence while digitizations of dense eelgrass beds using a 30 meter minimum mapping unit overestimated eelgrass distribution. Central to the results of both mapping methods was that both the classification and the imagery were sensitive to eelgrass exposure at low tide. Because the IR band does not penetrate through water there was high absorption even at shallow depths. Varying amounts of eelgrass were detected by the imagery dependent on the tides when the imagery was taken, leading to inconsistency in determining eelgrass areal extent over time. Assessing changes in the distribution of eelgrass through these methods helps us to better understand how dike removal can affect important tidal wetland habitat, and findings may inform methods for eelgrass mapping and for monitoring eelgrass distribution over time.