An assessment of the sources of error in estimations of bulk sedimentary pigment concentrations and its implications for trophic status assessment

Abstract

Pigment concentrations in surface sediments in two lakes were analysed on ten occasions over a year. Statistical analysis showed that mean concentrations varied through the year and the relative contribution of several sources of variability (analytical, spatial, temporal) to the total was estimated. Using this information, the relative cost-effectiveness of several alternative sampling strategies was investigated. Measurement of sedimentary pigment concentrations on single cores on two or three different occasions would reduce the standard error of the mean by 30% and 42% respectively and is very cost effective. However replicate cores taken on a single occasion appeared to be a workable compromise.The data also allowed a comparison of the ability of the different pigment determinands to resolve two lakes of very different trophic status. Total carotenoid concentrations were found to be the most effective.