Abstract
Phytoplankton sinking is a major component of vertical ocean carbon and nutrient fluxes, and sinking is an integral component of phytoplankton biology and ecology. Much of our understanding of phytoplankton sinking derives from the settling column method (SETCOL) in which sinking speeds are calculated from the proportion of cells reaching the bottom of a water-filled column after a set time. Video-based methods are a recent alternative to SETCOL in which sinking speeds are measured by tracking the movement of individual cells in a salinity-stratified water column. In this study, we present the results of a meta-analysis showing that SETCOL produces significantly and consistently lower sinking speeds than the video method. Next, we perform a particle image velocimetry analysis, which shows that the observed discrepancy in sinking speeds between the two methods can probably be explained by weak convection currents in the SETCOLs. Finally, we discuss the implications of these results for the interpretation of past and future phytoplankton sinking speed measurements and models that rely on those measurements.