An interlaboratory comparison for the measurement of biogenic silica in sediments

An interlaboratory comparison was made to assess the variability in the measurement of biogenic silica (BSI) in sediments among aquatic scientists. The purpose of this study was to (I) assess the present state of BSi analysis by aquatic scientists, (2) establish the range of variability in the measurement of BSi across the community of aquatic scientists (paleolimnologists, paleooceanographers, limnologists, estuarine scientists, and oceanographers), and (3) to determine if patterns in the measurement were related to specific methodologies used or to treatments. This study included 30 laboratories primarily using a variety of different wet chemical extraction techniques. The six samples used in the interlaboratory comparison were from modern freshwater and coastal marine depositional environments. The samples were chosen to cover a wide range of BSi concentrations, contain no sponge spicules or radiolarians, and were easy to digest relative to relict samples. Independently measured BSi concentrations of the same sediment ranged widely. BSi concentrations determined by X-ray diffraction were significantly higher than concentrations determined by wet chemical methods. The percent standard deviation of the mean in samples analyzed by wet chemical digestion techniques was highest in the samples with the lowest BSi concentration (Table 3) with 67.5% variation in sample 6 (1.31 +/- 0.88 wt.% as SiO2), and lowest at the highest BSI concentration with 21.2% variation in sample 2 (44.3 +/- 9.38 wt.% as SiO2,). An Analysis of Variance (ANOVA) was used to test for significant differences between treatments for the following variables: (1) pH of digestion solution, (2) digestion solution (Na,CO, or NaOH), and (3) whether a mineral correction was made. Although significant differences were found, inspection of the raw data suggests that several anomalous points may be responsible for the significant differences obtained. Therefore, the ANOVA was rerun following removal of 7 laboratories from the comparison. These laboratories were removed if they measured BSi concentrations more than Fl standard deviation from the mean in at least three of the six samples used in the intercomparison. For this analysis no significant differences were found between laboratories. Finally, no significant differences were found between the two most widely used techniques to measure BSi in sediments (DeMaster, D.J., 1979. The marine budgets of silica and Si-32. PhD Dissertation, Yale University, 308 pp.; Mortlock, R.A. and Froelich, P.N., 1989. A simple method for the rapid determination of biogenic opal in pelagic marine sediments. Deep-Sea Res., 36, 1415-1426.). The importance of optimizing the technique for the type of sample and of making a determination of slope for the mineral correction are explored. (C) 1998 Elsevier Science B.V. All rights reserved.