Daniel Conley

The biogeochemical cycling of silicon (Si) along the land-to-ocean continuum is studied by a variety of research fields and for a variety of scientific reasons. However, there is an increasing need to refine the methodology and the underlying assumptions used to determine biogenic silica (BSi) concentrations. Recent evidence suggests that contributions of nonbiogenic sources of Si dissolving during alkaline extractions, not corrected by standard silicate mineral dissolution correction protocols, can be substantial. The ratio between dissolved Si and aluminum (Al) monitored continuously during the alkaline extraction can be used to infer the origin of the Si fractions present. In this study, we applied both a continuous analysis method (0.5 M NaOH) and a traditional 0.1 M Na2CO3 extraction to a wide array of samples: (1) terrestrial vegetation, (2) soils from forest, cropland and pasture, (3) lake sediments, (4) suspended particulate matter and sediments from rivers, (5) sediments from estuaries and salt marshes and (6) ocean sediments. Our results indicate that the 0.1 M Na2CO3 extraction protocol can overestimate the BSi content, by simultaneously dissolving Si fractions of nonbiogenic origin that may represent up to 100% of the Si traditionally considered as biogenic, hampering interpretation especially in some deeper soil horizons, rivers and coastal oceanic sediments. Moreover, although the term amorphous Si was coined to reflect a growing awareness of nonbiogenic phases we show it is actually inappropriate in samples where silicate minerals may account for a large part of the extracted Si even after linear mineral correction.