We have developed a technique that combines a high temperature quartz furnace with inductively coupled plasma-mass spectrometry for the determination of Hg stored in the annual CaCO₃ bands found in coral skeletons. Substantial matrix effects, presumably due to the discontinuous introduction of CO₂ to the gas stream, were corrected for by simultaneously supplying a stream of argon containing highly enriched elemental ²⁰²Hg^o and observing peaks in the ²⁰⁰Hg/²⁰²Hg signal as the sample was decomposed. Primary signal calibration for Hg was achieved using gas injections from a saturated vapor standard. The absolute instrument detection limit was low (about 0.2 fmol), with a practical limit of detection (3σ of blanks) of 2 fmol. Reproducibility of samples was (RSD) 15–27%.

We applied this method to the determination of Hg concentrations in two colonies of Diploria labyrinthiformis collected from Castle Harbour, Bermuda, at a site about to be buried under the municipal waste landfill. The temporal reconstructions of Castle Harbour seawater Hg concentrations implied by the coral record show a decline throughout the period of record (1949–2008). The coral archived no apparent signal associated with waste disposal practices in the Harbour (bulk waste land-filling or, since 1994, disposal of waste incinerator ash), and mercury concentrations in the coral did not correlate to growth rate as assessed by linear extension. There was, however, a large and nearly exponential decrease in apparent Hg concentration in the Harbour which circumstantially implicates the dredging and/or landfilling operations associated with the construction of the airport on St. David’s Island.