Small-scale (<5.0 m) vertical heterogeneity in mudstones: implications for high-resolution stratigraphy in siliciclastic mudstone successions

48 mudstone samples, from the Peterborough Member, Oxford Clay Formation, have been analysed to determine the small-scale vertical and lateral lithofacies variability in this succession. The samples were collected every 0.15 m from Saxon and King's Dyke brickpits (Whittlesey, Peterborough, UK) over a short vertical interval (15 m). The sampled interval spans the obductum to the grossouvrei Subzones of the Callovian (Upper Jurassic). Lithofacies present were described using combined field logging, optical and backscattered electron optical techniques. The succession predominantly comprises silt-rich mudstones, with subordinate clay-rich mudstones, shell-beds and shell-pavements. The individual units are variously composed of detrital clay, quartz, pyrite, calcite with minor apatite, kaolinite and glauconite. In some units millimetre-scale upward-fining couplets are present; elsewhere, the original bedding structures have been destroyed by bioturbation. Comparison of successive thin sections reveals that upward-coarsening is present on a metre-scale and that shell-enriched units are located close to the tops of these upward-shoaling units. Comparison of samples collected laterally on a 0.01–1.0 m scale from the same level (3.0 m below the top of Bed 14) indicates that there is lateral lithofacies variability. The small-scale (0.001 m) couplets are interpreted to be beds deposited by waning flow currents. Significant breaks between these depositional events allowed a dysaerobic infauna to colonize the sediment. The metre-scale upward-coarsening intervals which are capped by shell-enriched units are interpreted to be parasequences. The observed small-scale lateral lithofacies variability is interpreted to be a product of small, but significant vertical miscorrelation errors rather that systematic lateral facies changes. These data suggest that producing reliable, bed-scale lateral correlations over distances of more than 1 km in this, and successions like it, will be very difficult.