Underestimation of biogenic silicon flux due to dissolution in sediment trap samples

Samples obtained from sediment trap studies are exposed to a variety of biases. One is the solubilization of particles in the sampling jars. We examined the dissolved silicon concentrations in the supernatants of sediment trap samples compiled during long-term studies from the Northern North Atlantic and the Baltic Sea. Samples originated from shallow depths (130–500 m) to depths close to the seafloor (2000–3000 m). A considerable increase of the dissolved silicon concentration due to the solubilization of biogenic particulate silica (bPSi) at all depths was measured at time scales typical for annual mooring deployments and sample analysis. During periods of low bPSi flux more than 80% of the sedimented silicon can be converted to the dissolved phase. On an annual basis the dissolved fraction accounted for up to 60% of total sedimented bPSi during single years. Median values vary between 13% and 30% and considerable dissolution was encountered in the shallow as well as in the deep sediment traps. Based on our data we conclude that the flux of bPSi and quantities derived thereof are erroneous unless corrected for the dissolved fraction. We therefore strongly recommend to quantify the dissolved silicon in sediment trap samples and to use these values for the correction of the particulate silicon flux.