Quantifying the Distribution and Shape of Mussel Mounds in Mesotidal Salt Marshes by Means of UAV LiDAR With Relevance to Salt Marsh Survival

Daniele Pinton1, Alberto Canestrelli2, Christine Angelini3, Benjamin Wilkinson4, Peter Ifju5, Collin Ortals2, Andrew Ortega6 and Sydney Williams7, (1)University of Florida, Civil and Coastal Engineering, Ft Walton Beach, FL, United States, (2)University of Florida, Civil and Coastal Engineering, Gainesville, FL, United States, (3)University of Florida, Engineering School of Sustainable Infrastructure & Environment (ESSIE), Gainesville, FL, United States, (4)University of Florida, Ft Walton Beach, FL, United States, (5)University of Florida, Gainesville, FL, United States, (6)University of Florida, Ft Walton Beach, United States, (7)University of Florida, Department of Environmental Engineering Sciences, Gainesville, FL, United States
Abstract:
One of the valuable functions performed by mussels is capturing organic and inorganic matter from the water column. Mussels excrete nutrients that are immediately available to plant life and deposit the remaining material in their proximity. This leads to the formation and growth of mounds. Recent studies show that mussel and marsh vegetation form an intimate interaction, known as mutualism, which benefits both species and may be critical to help these ecosystems bounce back from extreme climatic events, such as droughts. Due to the small spatial scale (~1 m) of the mounds and the presence of vegetation, neither standard satellite images nor photogrammetric techniques are able to define the 3D shape of these features. The present study aims at quantifying the distribution of mounds and the amount of material they store in meso-tidal salt marsh environments, by means of high resolution surveys. A point cloud of a vegetated tidal marsh at Sapelo Island, Georgia, USA was collected using a Velodyne 16 LiDAR sensor mounted on a DJI Matrice 600 UAV. An original object detection algorithm has been implemented which extracts all the mussel mounds from a LiDAR derived point cloud. The algorithm has been trained on 30 measured mussel mounds and validated on other 30 measurements. The algorithm provides the 3D shape of the mound, and the vegetation height and density on top of it. A comparison between 2016 satellite images and data from a 2019 LiDAR survey suggests a significant increase in mussel mound density in the study area. Our study shows that there is a significant increase in biomass on the mussel mounds. Moreover, material deposited by the mussels is a significant contribute to the marsh budget and could increase its resilience to sea level rise.