Remote sensing monitoring of phytoplankton community structure variation in the typical North Yellow Sea marine ranching

Phytoplankton are the main contributors to primary productivity and bait supply for fish and shellfish in marine ranching. As filter-feeding shellfish characterized by selective feeding behavior, monitoring the temporal-spatial variation of phytoplankton community structure is of significant importance for understanding the levels and fluctuations of the natural food supply in marine ranching. Based on the filed measured algae density data collected from 2018 to 2023 in marine ranching and Sentinel-3 OLCI satellite images, we developed a Support Vector Machine (SVM) method to develop a remote sensing inversion algorithm to characterize the phytoplankton community in the marine ranching in North Yellow Sea. The results show that diatoms density and percentage content exhibited lower values in summer and autumn, and higher values in winter and spring. In contrast, the densities of dinoflagellates and chrysophytes followed the opposite trend. From 2018 to 2023, the densities of diatoms, dinoflagellates, and chrysophytes in the study area all showed increasing trend, however the percentage of diatom showed decreasing trend, which may lead to a decrease in the level of natural bait supply in marine ranching. The decrease in phosphate and mixed layer depth from winter to summer, along with the increase in sea surface temperature and photosynthetically active radiation, regulates the competition between dinoflagellates and diatoms. This results in a summer-high, winter-low pattern for the dinoflagellates-to-diatoms abundance ratio, with a peak value in summer six times higher than in winter. On a temporal scale, this leads to a significant increase in the dinoflagellates-to-diatoms abundance ratio at a rate of 7.94% per year.