Quantification and spatial analysis of seagrass landscape structure through the application of aerial and acoustic remote sensing

Abstract

Biogenic components of the marine environment such as eelgrass (Zostera marina L.) provide numerous valuable ecosystem services and function as ecosystem engineers in the coastal environment. The spatial distribution and arrangement of eelgrass in the coastal landscape greatly influences ecological functions, necessitating mapping and monitoring of eelgrass habitat for effective management, though quantification of landscape structure has been hindered by challenges collecting and analyzing spatial data in the coastal zone. In this thesis, the spatial structure and distribution of eelgrass was studied through the application of acoustic and optical remote sensing and spatial analysis to quantify aspects of eelgrass landscape pattern. Single-beam acoustic data representing a seagrass landscape were collected and analyzed at multiple spatial scales through geostatistical methods and local spatial statistics (i.e., Getis-Ord Gi*), identifying areas of high and low cover in a spatially continuous seagrass bed. Acoustic data from the same site were compared to a satellite-derived dataset using remote sensing techniques for the evaluation of map accuracy. Performance of the satellite classification algorithm was found to vary depending on the spatial scale and degree of fragmentation in the landscape, highlighting the strengths and weaknesses of the method, and contrasting the landscape conceptualization of acoustic and aerial remote sensing. The spatial resolution of modern satellite data has greatly improved, though at pixel size of 2.4 m the ability to discern fine-scale patterns is limited. In contrast, very high-resolution aerial photography (pixel size ~3 cm) collected at a second site from a tethered helium-balloon platform was classified to depict a complex landscape mosaic comprised of eelgrass and blue mussels (Mytilus edulis L.). The application of landscape pattern metrics with high-resolution imagery additionally allowed tracking the temporal change in eelgrass patch metrics over a period of 26 months. The multidisciplinary approach of this thesis advance the application of spatial analysis in coastal ecosystems through the novel use of spatial statistics and remote sensing. Continued research and technological developments promise to improve management and provide new insights to the spatial dynamics and function of coastal landscapes.