The grazing pressure of micro-and nanozooplankton on phytoplankton was estimated in serial dilution experiments in the northwestern Arabian Sea and its adjacent areas (the Somali Current, the Somali Basin, the Gulf of Aden and the southern Red Sea) during the NE monsoon 1992-1993.Microzooplankton grazing rates @) on total phytoplankton (analyzed as chl a ) were generally exceeded by phytoplankton growth rates (g = 0.2 to 1.19 d-l, mean 0.48 d-'; p = 0.52 to 1.12 d-', mean 0.72 d-'1, resulting in an average daily consumption of 38% of the phytoplankton standing stock and 67% of the primary production. Microzooplankton grazing on 4 picophytoplankton groups (Prochlorococcus spp., Synechococcus spp., and 2 picoeukaryotes) analyzed by flow cytonletry showed growth (p = 0.27 to 0.92 d-l, mean 0.68 d") and grazing mortality rates @ = 0.26 to 0.73 d-l, mean 0.67 d-l) well in balance, with an average of 49% of the standing stock and 102% of the primary production grazed per day. Picophytoplankton growth and grazing mortality rates increased dramatically when grazers > l 0 pm were removed. These results suggest a control of the small grazers by larger ones (trophic cascade) and a close coupling between picoautotrophic prey and small grazers. The trophic cascade within the microbial food web of the nanoplankton encompasses 3 trophic levels: picoplankton -small HNF -larger flagellates and ciliates.

During a survey in the subtropical North Atlantic along 20" W longitude, up to 95 000 cells ml-' of an unknown picoplankter were observed. The species was identified as a prochlorophyte, sensu Chisholm et al. (1988), by shipboard flow-cytometry and HPLC-analysis of pigment composition. Enumeration of the picoplankton groups indicated that these prochlorophyte-like cells contributed most to the deep chlorophyll maximum (situated at 80 to 100 m). They were, however, also abundant in the upper water layers. The cellular content of their major pigments ('red shifted' or divinyl chlorophyll a, chlorophyll b and zeaxanthin) varied considerably over the water column and appeared to be dependent on the growth irradiance. The 'red-shifted' chlorophyll a concentration increased from 0.91 fg cell-' near the surface to 5.4 fg cell-' near the base of the euphotic zone (0.3 % I,,); chlorophyll b from 0.44 fg cell-' to 8.20 fg cell-'. The zeaxanthin concentration per cell remained almost constant with respect to depth (average 1.97 t 0.15 fg cell-'). A linear relationship was found between the sum of 'red-shifted' chlorophyll a and chlorophyll b and the flow-cytometrically obtained mean red fluorescence signal of the algal cell.

UVBR (ultraviolet-B radiation: 280 to 315 nm)-induced DNA damage, measured as cyclobutane pyrimidine dlmers (CPDs), was determined in size fractions of natural populatlons of bacterio-and phytoplankton collected in marine tropical waters. Mean biologically effective UVBR doses in the wind-mlxed layer were calculated from DNA dosimeter data. Phytoplankton species composition in these waters was monitored using flow cytometry and pigment analyses. In terms of (divinyl-)chlorophyll a concentrations, prochlorophytes and cyanobacteria comprised the largest fraction of the phytoplankton, except in a eutrophic bay at Curaqao, an island located in the southern Caribbean. In terms of cell numbers and amount of DNA, small prochlorophytes and marine bacteria dominated. Small but detectable levels of UVBR-induced DNA damage were found at all locations. In general, more DNA damage was found in the small size fraction (0.2 to 1 pm) than in the larger size fraction (1 to 10 pm). The greatest amount of damage was found in the small size fraction collected in the central Atlantic Ocean (20 C P~s / l 0~ nucleotides), despite the fact that UVBR doses were much higher at other locations. The calculated mean biologically effective UVBR doses In the wnd-mixed layer were 2 to l ? times lower as compared with incident UVBR doses. CPD levels determined in cultures of the cyanobacterium Synechococcus sp. subjected to UVBR doses similar to those in the wind-mixed layer corresponded with CPD levels measured in the 1 to 10 pm fractlon in the field. Our results indicate that UVBR vulnerability is size dependent. Furthermore, the low CPD levels observed in these field communities may be explained by the low mean biologically effective doses received by the cells as a result of wind-induced mixing.

In July 1993, an extensive study was made of a large bloom of the coccolithophorid Emiliania huxleyi in the North Sea halfway between the Shetland Islands and Norway. Here w e report on the hydrography, production and sedimentation of particulate organic carbon (POC) and calcite carbon (calcite-C) at 4 stations occupied for 24 h, 2 inside the bloom and 2 just outside. The coccolithophorid bloom was confined to North Sea waters, where a stable shallow mixed layer had been formed. Bloom development had entered the decaying phase, judged by the relatively low living cell number (maximally 1200 cells cm-3), the high number of loose coccoliths (up to 350000 coccoliths cm-3), and the fact that sedimentation of calcite-C exceeded production. In the top 15 m at the bloom stations, the mean daily production of coccoliths was 17 per cell. At the 2 stations outside the bloom, the dominant coccolithophore was a holococcolithophorid (up to 1400 cells cn1r3), wlth insignificant amounts of calcite produced per cell. At these stations, nutrients were present in non-lim~ting concentrations and production of POC was twice as high as at the bloom stations. In the bloom, mixed layer nitrate levels were below 0.2 PM. Faecal pellets collected in the sediment traps contained large numbers of coccoliths of E, huxleyi. Although the numbers of grazers at the 2 stations outside the bloom were not lower than those in the bloom, the volume of faecal matter sedimenting at 50 m was about 70 times lower. It is hypothesized that faecal pellets outside the bloom were so light in weight that they did not sink very far before degradation, whereas the pellets produced in the E. huxleyi bloom in general were exported rapidly due to their heavy load of calcite. This implies that recycling of materials in the mixed layer of this bloom is relatively low due to high downward flux rate. The ratio at which POC and calcite-C were sedimenting amounted to 1.3 on average for the 2 bloom stations at 50 m water depth.

During a survey of the end phase of an Emiliania huxleyi bloom in the northern part of the North Sea we measured total inorganic carbon (TIC) and the fugacity of CO2 (f CO2), as well as standing stocks of CaC03 and particulate organic carbon (POC). Production of CaC03 by E. huxleyi resulted in an immediate increase of f C 0 2 , but led to a long-term decrease in fCO, Observations during a surface survey and at 24 h stations showed a large increase of fCOz with the standing stock of CaCO,. The immediate increase of f C 0 2 is caused by a shift in the chemical equilibria In the inorganic carbon system when alkal~nity decreases relative to dissolved inorganic carbon (DIC). Average f C 0 2 in the high reflectance area (with high numbers of detached coccoliths) was lower than average fCOz In the reference areas, located outside the E. huxleyi bloom. The long-term decrease in fCO, is due to an enhdnced sedimentation of both organic and inorganic carbon in faecal pellets containing heavy calcite. This enhanced sedimentation is reflected in the vertical gradient of TIC between the surf~lce mixed layer and the aphotic zone, which increased from the POC-rich zone to the CaCO, maximum. The overall effect of production, air-sea exchange, mineralisation and sedimentation was a decreabr of f C 0 2 due to a net transport of carbon to below the pycnocline. We tentatively calculate an atmospheric carbon sink of 1.3 m01 m-2 for this bloom of E. huxleyi.