Chesapeake Bay nutrient and plankton dynamics: III. The annual cycle of dissolved silicon

Abstract

Silicic acid (H 4SiO 4) flux from the sediment, H 4SiO 4 concentration and river flow were used to obtain an annual dissolved silicon budget for Chesapeake Bay. H 4SiO 4 concentrations vary seasonally in the estuary: for a 12-year period, mean H 4SiO 4 concentrations in the mesohaline region were high both in spring and in late summer to early fall, and were low in late spring—occasionally approaching levels potentially limiting to diatom growth. Most of the annual allochthonous H 4SiO 4 supply to the estuary derives from the three major rivers, but regenerative H 4SiO 4 flux from the sediment to the water column exceeds the total riverine input by a factor of at least five. Sediment H 4SiO 4 efflux exhibits seasonality and averages approximately 2–3 mol Si m −2 yr −1. The high rates of sediment dissolution and efflux appear to maintain high levels of H 4SiO 4 in the mesohaline region, and Si-limitation of diatom growth there seems unlikely. The relative rates of biogenic silica formation and dissolution do not vary synchronously: seasonal variations in diatom productivity, sedimentary release of H 4SiO 4 and river flow all contribute to the observed late winter and late summer seasonal maxima and late spring minimum in water column H 4SiO 4 concentrations. If the only source of Si to support sedimentary H 4SiO 4 efflux is biogenic particulate silica recently deposited from the water column and this silica in turn was produced by diatoms in a ratio of 8C:1 Si, the minimum annual primary production by diatoms is at least 260 g C m −2, approximately half of annual total plankton primary production. This estimate would be revised upwards according to the amount of particulate biogenic silica dissolving in the water column. Burial of biogenic silica amounts to from 2 to 84% of the sediment efflux of H 4SiO 4, depending on location in the bay. On an annual basis, burial represents from 60 to 100% of fluvial H 4SiO 4-Si inputs.