Causes and biogeochemical implications of regional differences in silicification of marine diatoms

Diatoms facilitate the export of organic carbon and associated nutrient elements in the ocean because their dense opaline silica shells provide ballast to sinking particles. Marine ecosystem models generally assume that cellular silicification is either constant or varies solely due to physiological responses. Using a cell-specific technique, synchrotron-based X-ray fluorescence (SXRF) microscopy, we show that diatom cells in the cold, high-silicic-acid waters of the Antarctic Zone of the Southern Ocean (SOAZ) had 6 times more Si per volume than did those inhabiting the warm, low-silicic-acid waters of eastern equatorial Pacific (EEP). Ratios of Si: P and Si: S differed less than this because cellular P and S concentrations were higher in SOAZ cells. Resulting differences in excess density and frustule surface area-to-volume ratios should result in more efficient removal and slower dissolution of biogenic silica in the SOAZ compared to the EEP. Moreover, the difference between the excess densities of diatoms and nondiatoms was 15-fold greater in the SOAZ than in the EEP. Several possible causes of the regional differences in silicification are evaluated. Differences in cell volume between regions and additions of silicic acid and iron had minor effects on silicification. Instead, cellular silicification varied substantially among diatom morphological types within each region, suggesting that community composition largely determined the community silicification in these regions. We suggest that ecological processes may cause much larger systematic regional and temporal differences in cellular stoichiometry than is currently accommodated by ecosystem models.