A Phaeocystis bloom in the northern Beibu Gulf reflects nutrient-driven shifts in dominant phytoplankton taxa.

Large-scale Phaeocystis globosa blooms in the Beibu Gulf have adversely affected marine ecosystems and the functioning of a nuclear power plant. To understand the formation mechanisms for such large-scale blooms, we examined the P. globosa bloom dynamics in the northern Beibu Gulf from September 2015 to March 2016 in association with changes in phytoplankton community and environmental factors. Statistical analysis methods, such as change-point method, principal component analysis (PCA), redundancy analysis (RDA) and an artificial neural network (ANN) model, were utilized to identify environmental drivers on the dynamics of phytoplankton community during the P. globosa bloom. The bloom evolution can be categorized into three stages, delineated by the fluctuation in colony abundance of P. globosa. Stage I, immediately before the Pheaocystis bloom, was characterized by the dominance of cyanobacterium Synechococcus (Syn) and the presence of the pigment 19'-hexanoyloxyfucoxanthin (Hex-fuco) in phytoplankton. Stage II, the development stage of the Phaeocystis bloom, was characterized by the appearance of P. globosa colonies and the pigment 19'-butanoyloxyfucoxanthin (But-fuco). Stage III, the decline stage of the bloom, was characterized by the increase of diatoms and pico-sized eukaryotes. Dynamics of the P. globosa bloom reflected the shift of dominant phytoplankton taxa from diatoms to haptophytes during winter, closely related to the change of turbidity, silicate limitation and rebalancing of nitrogen and phosphate. This study is the first to examine the relationship of a Phaeocystis bloom with phytoplankton structure and environmental factors in the Beibu Gulf, providing critical insights for the bloom forecasting and mitigation.