Onset, intensification, and decline of phytoplankton blooms in the Southern Ocean

• Published in: ICES Journal of Marine Science
• Main Authors: Llort, Joan, Lévy, Marina, Sallée, Jean-Baptiste, Tagliabue, Alessandro
• Format: Article in Journal/Newspaper
• Language: English
• Published: Oxford University Press (OUP) 2015
• Subjects: Southern Ocean
• Online Access: http://dx.doi.org/10.1093/icesjms/fsv053; http://academic.oup.com/icesjms/article-pdf/72/6/1971/31225088/fsv053.pdf

Abstract The seasonal cycle of phytoplankton biomass in the Southern Ocean (SO) is characterized by a period of rapid accumulation, known as bloom, that is typical of high-latitude regions. Recent studies have illustrated how spatial and temporal dynamics of blooms in the SO are more complex than in other oceans. This complexity is likely related to differences in vertical mixing and the iron availability. In this work, we examine the sensitivity of bloom dynamics to changes in vertical mixing and iron availability using a biogeochemical model. Under idealized physical forcing, we produce seasonal cycles of phytoplankton for an ensemble of SO scenarios and we describe the bloom dynamics in terms of the net biomass accumulation rate. Based on this metric, we define three crucial bloom phases: the onset, the climax, and the apex. For the ensemble of modelled blooms, onsets always occur in winter and can be either bottom-up (increase in productivity) or top-down (decrease in grazing) controlled. Climaxes are mostly found in spring and their magnitudes are bottom-up controlled. Apexes are always found in late spring and strongly top-down controlled. Our results show that while a “strict” onset definition is consistent with a winter onset, the surface spring bloom is associated with the climax of the integrated bloom. Furthermore, we demonstrate that onset phase can be distinguished from climax phase using appropriate bloom detection methods based on surface satellite-based products. The ensemble of these results suggests that Sverdrup's blooming conditions are not indicative of the bloom onset but of the climax. We conclude that the recent bloom onset debate may partly be due to a confusion between what is defined here as the bloom onset and the climax, and that the SO observed complexity is due to the factors that control the climax.