| Abstract: | This study investigates the potentially deleterious effect of prolonged sample storage on the reliability of calcareous nannofossil species abundance and community diversity measures and their subsequent interpretation as paleoceanographic proxies. Nannofossil assemblages from two freshly cored successions of hemipelagic clay were documented over a one-year period using a series of paired smear-slides, the first made from sediment samples within a day of coring and the second either 6 months or 1 year later. Diversity measured by Shannon's H is consistent between the two sets for the majority of samples, however significant changes, both positive and negative, were recorded in a number of cases and appear to be largely due to variations in smear-slide thickness and not to post-sampling dissolution. There is no significant change in nannofossil fragmentation or abundance over time, although the relative abundances of small, dissolution susceptible placoliths of Reticulofenestra minuta do appear to decrease with time in some samples. Counts of mineral grains show that there is a significant loss of fine, 1–2 μm, cubic pyrite within months of drilling, and this, combined with the decline in abundance of R. minuta, suggests that limited pyrite oxidation coupled to carbonate dissolution is occurring. Geochemical analyses confirm that these sediments contain significant concentrations of sulphate one year after sample recovery, which may be the result of gypsum formation associated with the oxidation of pyrite. Estimates of pyrite oxidation suggest that up to 3% of the original calcium carbonate has dissolved in the year after coring, which could account for the observed loss of small placoliths. Effects of this kind are almost certainly confined to hemipelagic sequences rich in organic matter and/or reduced iron but are frequently offset by the exceptional preservation of calcareous microfossils that can occur within such sediments. |
