Numerical modeling of propagule dispersal for Sargassum bed restoration in Gamak Bay, Korea

Abstract

The Sargassum beds in Gamak Bay, Korea, have deteriorated over recent decades. We conducted numerical experiments to assess the applicability of artificial reefs and longlines for seaweed forest restoration in terms of propagule dispersal. We applied hydrodynamic (Environment Fluid Dynamics Code (EFDC)), wave (Simulating Waves Nearshore (SWAN)), and particle tracking models for tidal currents, wave-induced currents, and the dispersal of the propagules, respectively. The major factors for propagule dispersal such as forcing, settling velocity w s , and release height were included, and the results of the models were analyzed using dispersal kernel, analysis of variance (ANOVA) tests, and survival rate based on bed elevation. The dispersal distance kernel was fitted to Gaussian and polynomial functions and showed higher wave impacts, whereby the mean dispersal distance was simulated 6.3 m under pure tide and increased by 14 times under the combined tide and wave condition. Propagules were found to hit the bottom intensively during a low-to-flood phase, while wave forcing enhanced vertical dispersal and hindered the intensive sinking. A two-dimensional dispersal location kernel showed that dispersal distance increased with high release heights and its direction faced the shoreline due to shallow water depth and strong wave-induced currents. ANOVA tests regarding the mean distance indicated that w s , forcing, release heights, and the combined effect of forcing and w s contributed 49.5, 20.5, 5.4, and 22.9 %, respectively, yielding a total of 72 % from w s . Adjusting the release height to extend dispersal distances led to A, C, and B locational effects, in order from lowest to highest, while survival rates were 67, 70, and 83 % for A, B, and C, respectively. From the results, we concluded that C is the most suitable locational type for restoration.