Search

This study aims to quantify the dimensions of an oyster reef over two years via low-cost unoccupied aerial vehicle (UAV) monitoring and to examine the seasonal volumetric changes. No current study investigated via UAV monitoring the seasonal changes of the reef-building Pacific oyster (Magallana gigas) in the German Wadden Sea, considering the uncertainty of measurements and processing. Previous studies have concentrated on classifying and mapping smaller oyster reefs using terrestrial laser scanning (TLS) or hyperspectral remote sensing data recorded by UAVs or satellites. This study employed a consumer-grade UAV with a low spectral resolution to semi-annually record the reef dimensions for generating digital elevation models (DEM) and orthomosaics via structure from motion (SfM), enabling identifying oysters. The machine learning algorithm Random Forest (RF) proved to be an accurate classifier to identify oysters in low-spectral UAV data. Based on the classified data, the reef was spatially analysed, and digital elevation models of difference (DoDs) were used to estimate the volumetric changes. The introduction of propagation errors supported determining the uncertainty of the vertical and volumetric changes with a confidence level of 68% and 95%, highlighting the significant change detection. The results indicate a volume increase of 22 m³ and a loss of 2 m³ in the study period, considering a confidence level of 95%. In particular, the reef lost an area between September 2020 and March 2021, when the reef was exposed to air for more than ten hours. The reef top elevation increased from -15.5 ± 3.6 cm NHN in March 2020 to -14.8 ± 3.9 cm NHN in March 2022, but the study could not determine a consistent annual growth rate. As long as the environmental and hydrodynamic conditions are given, the reef is expected to continue growing on higher elevations of tidal flats, only limited by air exposure. The growth rates suggest a further reef expansion, resulting in an incr

The Pacific oyster (Magallana gigas) is an invasive species in the Wadden Sea transforming parts of it permanently. M. gigas, as an ecosystem engineer, builds reef structures that are characterized by highly complex and variable surfaces consisting of densely packed, sharp-edged individuals connected with cement-like bonds. To investigate the interactions between reef structure, shape and formation and wave as well as tidal currents, an understanding of the surface roughness is essential. This work reports on observations of oyster reefs for which seven new structural classes (Central Reef, Transitional Zone, Cluster I, Cluster II, Patch I, Patch II, and Garland) are proposed. For each class, high resolution Digital Elevation Models (DEMs) have been elaborated based on Structure-from-Motion (SfM) photogrammetry and analyzed using spatial statistics. By determining probability density functions (PDFs), vertical porosity distributions, abundances, orientations and second-order structure functions (SSFs), topographical parameters that influence the hydraulic bed roughness have been determined. The results suggest, that by applying the structural classification and their distinct topographical roughness parameters, the oyster reef surfaces can be described appropriately accounting for their complexity. The roughness accounts to a total roughness height kt = 103 ± 15 mm and root-mean-square roughness height krms = 23 ± 5 mm. These values were found similar across all structural classes, yet the shape of the PDFs reveal differences. With decreasing abundance, the distributions become more positively skewed and are characterized by more extreme outliers. This is reflected in the higher statistical moments, as the skewness ranges between Sk = 0.4–2.1 and the kurtosis between Ku = 2.2–11.5. The analysis of the orientations and the SSFs confirms anisotropic behavior across all structural classes. Further, the SSFs reveal the oyster shells as significant roughness elements wit