DIURNAL CHANGES IN NEAR-SURFACE AMMONIUM CONCENTRATION - INTERPLAY BETWEEN ZOOPLANKTON AND PHYTOPLANKTON

Data from a cruise in 1996 in the Southern Ocean near the island of So uth Georgia indicate that ammonium concentrations in near-surface wate rs (top 30 m of the water column) varied diurnally by similar to 0.2-- 0.35 mmol m(-3)day(-1)(3 h mean values) in different regimes. Maximum values (up to 1.3 mmol m(-3)) occurred around local midnight and minim um values (down to 0.1 mmol m(-3)) close to noon. This cyclicity was n ot found in other nutrients analysed (silicate, nitrate, nitrite and p hos- phate). The potential for mixing ammonium-rich water from the pyc nocline into the surface mixed layer (SML) at night is examined. Budge t calculations for both ammonium and temperature suggest that complete mixing of the pycnocline water into the SML would increase concentrat ion to observed levels. However, it would also give rise to changes in pycnocline ammonium concentration and temperatures which do not agree with observations. Moreover, such a model will not account for daytim e drawdown, which is likely to be due to biological processes. A simpl e biological model combining night time excretion of ammonium by zoopl ankton and daytime uptake by phytoplankton simulates the observed chan ge, but only if (i) the daily-averaged phytoplankton growth rate was o f the order of 0.5-1 day(-1), (ii) this growth was strongly dependent on ammonium for its nitrogen nutrition, (iii) zooplankton grazing was sufficient to keep phytoplankton biomass at a constant level and (iv) much of the ingested nitrogen was excreted. The implications of these conditions are explored in the context of other observations during th e study, and on the basis of published data. It appears that the re su pply of inorganic nitrogen through zooplankton grazing may have been i mportant in sustaining primary production, indicating a very tight cou pling between grazers and their food supply. This conclusion is suppor ted by comparison with results from a similar cruise in 1994. Then the re was a reduction in the biomass of Antarctic krill, a major componen t of the zooplankton, to 14% of that observed in 1996. Diurnal change in ammonium concentration could not be detected and phytoplankton grow th rate had been slower (about 0.25 d(-1)), providing support for the suggestion that biological processes were more important than physical mixing in generating ammonium cyclicity. The potential consequences f or the biological cycling of carbon in high-and low-krill abundance ye ars are discussed.