The links between dissolved arsenic speciation, biological activity and the availabilities of the nitrogen and phosphorus plant nutrients have been investigated in a seasonal survey of Southampton Water (U.K.). Southampton Water (Hampshire, southern England) is an approximately 10 km long, and 2 km wide north-westerly extension of the Solent, receiving water hom the rivers Test and Itchen. It is a partially mixed estuary bordered by broad intertidal mudflats with shingle and sand on the eastern side, and a salt marsh to the west. Two sites were chosen: NW Netley Buoy is in a sheltered high-salinity estuarine environment whilst Calshot Buoy lies just outside Southampton Water and in a more exposed location of less-variable salinity. The first evidence of arsenic(III) production at both sites occurred in the second half of April, during the decay of a major Skeletonema costatum diatom bloom. Arsenic(III) levels rose as Skeletonema was replaced by a numerically smaller but more chlorophyll-rich bloom of another diatom, Rhizosolenia delicatula. Rhizosolenia is therefore implicated as a possible source of arsenic(III). Methylated arsenic was absent whilst the mater temperature was low and during the initial Skeletonema bloom, but a week later, during the growth phase of the succeeding bloom of the diatom R. delicatula, they became detectable. Methylated arsenic levels gradually increased through the spring to a broad maximum covering the mid-summer, when Mesodinium rubrum, Scrippsiella trochoidea and associated microflagellates also peaked. No subsequent single organism could be linked to the release of methylated arsenic into Southampton Water; organoarsenicals having been observed in the presence of flagellates, diatoms and ciliates. A large bacterial maximum was observed following blooms of S. trochoidea and M. rubrum but laboratory culture experiments of natural bacteria from Netley failed to produce significant changes in the concentration of any arsenic species. Phosphate depletion did not appear to be a prerequisite for arsenate assimilation. From the summer peak methylated arsenic levels then gradually diminished to undetectable levels in the winter months. Monomethylarsenic, present at concentrations approximately 50% those of dimethylarsenic, persisted longer through the summer. The arsenic species which can be measured using the hydride procedure may therefore represent intermediates in the decomposition of the bioarsenicals, such as arsenosugars, which are released, either actively as excretion/secretion products or passively as part of the decay process. Currently unidentified precursors of hydride-reducible arsenic species ('hidden' arsenic) may explain the poor link between planktonic activity and the levels of measurable arsenic species in the water column.
