METHODOLOGY

The six A. opercularis shells were obtained from five BGS cores taken in the Inner Silver and Sole Pits, two of a series of elongated troughs on the bed of the southern North Sea. Full locality details, specimen depths in the cores, and specimen identification numbers are given in Table 1; a location map and simplified core logs are provided in Figure 1 and Figure 2, respectively. All specimens used in this study are held at the BGS, Keyworth. Material from each shell was converted to graphite by Fe/Zn reduction at the NERC Radiocarbon Laboratory, East Kilbride, and ¹⁴C AMS-dated at the NSF Arizona Radiocarbon Laboratory. The uncorrected results are given in Table 2, together with values that have been corrected for the marine reservoir effect (Sutherland 1986). Two specimens from the same core (PB3 and PB4) yield absolute dates in accordance with their relative stratigraphic position. The remaining specimens, all from different cores, do not yield dates that relate closely to their respective depths below the sediment surface (for instance PB5, the youngest shell of all, is from the greatest depth). This is, however, far from surprising given the significant local variation in the overall rate of Holocene sedimentation in the southern North Sea (e.g., Caston 1979; figure 7.14). Indeed, the variable thickness and indeterminate age of the sediment fill within southern North Sea "pits" were the principal reasons for the coring and carbon-dating programme, in order accurately to determine depositional history (Land-Ocean Interaction Study 1994; p. 23).

All specimens exhibited excellent preservation of shell structure and showed little evidence of abrasion or long exposure on the sea floor (e.g., encrustation by epibionts on the inner surface), although the corer had damaged some shells. Two specimens (SP1 and PB5) had some pigmentation preserved (Figure 3). Five specimens were obtained as single valves only; the other (SP1 from the Inner Silver Pit) was found articulated and in life position (i.e., left valve uppermost), implying rapid post-mortem burial or death in conjunction with sediment deposition.

Specimens were scrubbed, then cleaned using an air-abrasive tool. The outer surface of cleaned specimens was then sequentially sampled along the dorso-ventral axis (average sample separation 1-1.5 mm) using a small drill, producing samples of approximately 0.5 mg weight. Isotopic analyses were performed at the NERC Isotope Geosciences Laboratory (NIGL), Keyworth, using a VG Isocarb + Optima system. Approximately 0.1 mg of sample was used in each analysis. Isotopic compositions were calculated by comparison with concurrently analyzed laboratory-standard carbonate, calibrated against the international standards NBS-19 and NBS-18. Analytical precision, expressed as 1 S.D. and based on laboratory standards and replicate sample analysis, was typically <0.12 for ¹⁸O and ¹³C; in the case of the former, this equates to an error of <0.5°C in temperature estimates.

Palaeotemperatures were calculated using the equation of O'Neil et al. (1969) for a calcite system:

T = 16.9 - 4.38 (¹⁸O_c - ¹⁸O_w) + 0.10 (¹⁸O_c - ¹⁸O_w)²

where T is water temperature (in °C), ¹⁸O_c is the ¹⁸O value of shell carbonate (vs. Vienna Peedee Belemnite [VPDB]) and ¹⁸O_w is the ¹⁸O value of ambient seawater (vs. Vienna Standard Mean Ocean Water [VSMOW]) (minus 0.26 to convert to VPDB; Coplen et al. 1983). The value of ¹⁸O_w used was 0.1, similar to modern North Sea values (cf. Hickson et al. 1999); use of this value is justified by the fact that fully marine conditions, like those of the present day, were established some 6,000 years BP, long prior to the date of the oldest shell investigated (Cameron et al. 1992). Given fully marine conditions, it is unlikely that ¹⁸O_w underwent anything more than the negligible seasonal variation observed at present.