Decoding intertidal oyster reef morphology: Insights from UAV photogrammetry and deep learning

Abstract

Global coverage of oyster reefs has substantially declined due to environmental changes driven by climate change and human activities, severely impairing their ecological functions. Quantitative information on oyster reef morphology is crucial for developing restoration strategies and understanding the ecomorphodynamics of these reefs. This study focused on the Liyashan oyster reef conservation area in the intertidal zone of Haimen, Jiangsu, China. We utilized Unmanned Aerial Vehicle (UAV) photogrammetry to obtain RGB orthoimages and Digital Elevation Models (DEMs) with centimeter-level resolution and accuracy. By integrating UAV photogrammetry and Deep Learning techniques, we efficiently and accurately identified reef footprints and created pixel-level reef height maps. Based on the reef height data, we introduced the Volume Balance Index (VBI) to assess reef fragmentation (degree of pitting), where a lower VBI value indicates higher relative fragmentation. Quantitative results at the reef block scale demonstrate a significant negative correlation between reef size (area and height) and overall fragmentation, with a strong logarithmic relationship between reef height and VBI. Generally, less degraded reefs are primarily distributed in the eastern and southern parts of the region, a pattern potentially related to local hydrodynamic conditions. This study provides a cost-effective and efficient solution for monitoring intertidal oyster reefs and lays the foundation for further research on their ecomorphodynamics.