| Summary: | Detrital grain contamination in a diatom sample can considerably influence the delta(18)O(diatom) signal. In order to obtain a meaningful signal, pure samples must be used. This can be achieved via a series of cleaning stages including organic and carbonate material removal, sieving, differential settling and heavy liquid separation. The method described here works best for sediments with >20% diatom content. Based on testing various clean-up methods, we propose a sequence of four clean-up stages to produce pure diatom samples from a range of lake sediments types starting with a few grams of sediment. The diatom content and the oxygen isotope composition of the samples at each stage were measured in order to assess the effect of differential amounts of contamination. Results show that a four stage clean-up is necessary to produce clean diatom samples and that contamination by silt and clay causes lower delta(18)O(values).
|
|---|