Mismatch between coccolithophore-based estimates of particulate inorganic carbon (PIC) concentration and satellite-derived PIC concentration in the Pacific Southern Ocean

Coccolithophores are the main type of calcifying phytoplankton in the Southern Ocean (SO) and are key organisms in the production of particulate inorganic carbon (PIC). However, in situ studies of coccolithophores and in particular of their importance for the input of PIC in the SO are sparse in space and time due to the harsh weather conditions in the subantarctic realm. An alternative tool for monitoring PIC is the use of optical remote sensing, as coccolithophores account for most of the optical PIC backscattering in the sea. The aim of the present study is to provide coccolithophore-based estimates of PIC derived from Scanning Electron Microscope coccolith morphometric analyses and MODIS Aqua Level-2 and Level-3 PIC concentration values along two latitudinal transects from New Zealand to Antarctica and across the Drake Passage. In general, the coccolith-estimated PIC and satellite-derived PIC datasets show comparable trends in the Subantarctic and Polar Frontal zones of both transects, with coccolith-derived PIC values being generally lower than satellite PIC values. According to the coccolithophorid data, Emiliania huxleyi type A, over-calcified type A, and other taxa (e.g., Calcidiscus leptoporus) only contribute to coccolithophore PIC in the northernmost sampling locations, whereas E. huxleyi morphogroup B contributes substantially to the PIC content south of the Subantarctic Front in both transects. High satellite-derived PIC concentrations south of the Polar Front are not apparent in the coccolith-based PIC data. We suggest that the high reflectance signal in the Antarctic Zone may instead relate to the presence of small biogenic opal particles (e.g., diatoms, silicoflagellates, and/or small siliceous plankton) or other unknown highly reflective particles (such as Phaeocystis aggregations). Our results highlight the challenges presented by the lack of reliable satellite data in some parts of the SO as well as the importance of in situ measurements and methodological accuracy when estimating PIC values.