Sinking rate versus cell volume relationships illuminate sinking rate control mechanisms in marine diatoms

It has been shown that for dead marine diatom cells or diatom cells which are severely stressed metabolically, larger cells sink faster than small cells as dictated by Stokes' Law. In these cases, the slope of the sinking rate versus cell volume relationship within a culture reaches a maximum. Within cultures of rapidly dividing cells, larger cells sinking rate is reduced physiologically to that of smaller cells and the slope of this relationship approaches zero. In several marine diatom species between 5 and 100 μm in diameter, deviations from the maximum slope of the volume versus sinking rate relationship could be used to quantify the physiological reduction of sinking rates. This allowed us to differentiate 2 different components of sinking rate control, the ballasting component (driven by changes in cell composition and volume) which, when dominant, causes sinking rates to be proportional to cell volume and the energy-requiring, protoplast and vacuolar component which, when active, allows sinking rates to become independent of cell volume. Across the 9 species of diatoms examined, including the 3 single-celled species (Ditylum brightwellii, Thalassiosira pseudonana, and T. weissflogii), 4 chain-forming coastal bloom diatoms (T. aestivalis, Skeletonema costatum, Chaetoceros debilis and C. compressum) and 2 large floating open ocean species (Ethmodiscus sp. and entire Rhizosolenia spp. mats), there was a strong correlation between log cell volume and sinking rate only for cells that were metabolically inactivated either through extended dark treatment or through treatment with the respiratory inhibitor KCN. This was true both within and between cultures. However, no correlation between sinking rate and cell volume was found for rapidly growing cells maintained at saturating irradiances. This supports the notion that there is no obligate correlation between cell volume and sinking rate for metabolically active cells. This potential for cellular modification of the sinking rate versus volume relationship suggests that physiological state may be an important feature to include in models where carbon flux is predicted on the basis of particle size spectra. We suggest that the minimum cell volume necessary for active sinking rate control is ca 200 μm 3 , and that this represents a lower limit for Villareal's (1988; Deep Sea Res 35:1037-1045) theoretical minimum volume necessary for positive buoyancy.