Prol-ocentrum minimum (formerly also known as P. mariae-lebouriae) is a common bloomforming, photosynthetic dinoflagellate in Chesapeake Bay, USA. It is also capable of ingesting other cells. In Chesapeake Bay, P. minimum usually CO-occurs with cryptophytes. Ingested cryptophyte material is observable in the dinoflagellate under an epifluorescent microscope as orange-fluorescent inclusions (OFI) During Aprll and May, the frequency of OF1 was 510% In both surface and pycnocline assemblages. In summer, up to 50'% of the P. m~njmum contained OFI. The frequency of OF1 was positively correlated with cryptophyte abundance, but OF1 were not frequent in all populations of P minimum when cryptophyte densities were high. On-deck experimental incubations were done to determine the conditions that influence feeding. Light level and inorganic nutrient availability over the previous 24 h affect feeding. Incidence of feeding is lower when populations are maintained In the dark for 24 h than on a natural 1ight:dark cycle. Addition of n~trate and phosphate together can inhibit feeding. Ingestion has a die1 pattern, with frequency of OF1 highest in the afternoon and evening and lowest in the morning. Feeding is influenced by a complex of factors, but the spatial-temporal pattern of ingestion and the experiments both suggest that feeding is primarily a mechanism for obtaining lim~hng inorganic nutrients rather than a mechanism for supplementing carbon nutrition d u r~n g light limitation. Ingestion of other protists, including competitors for light and nutrients, may be a n important strategy which allows bloom-forming dinoflagellates to dominate plankton assemblages for extended periods and dunng changing nutrient regimes.

In order to experimentally investigate feeding by mixotrophic dinoflagellates, we developed protocols for the use of live protistan prey as markers of ingestion. CMFDA (5-chloromethylfluorescein diacetate), a vital green fluorescent stain, was used to label cultures of photosynthetic nanoflagelIates, a diatom, and an oligotrichous ciliate. Cryptophytes were not readily stained with CMFDA, but phycoerythrin-containing members of this phylum havo a distinct yellow-orange fluorescence and thus can be used unstained to demonstrate ingestion. M'ith these complen~entary techniques, we qualitatively demonstrated feeding by the dinoflagellates Ceratium furca, G)/~nnodinium sanguineum, Gj,rodinium estuariale, Prorocentrum n~~n i m u m (= mariae-lebouriae) and Peridinium brevipesin natural dsscmblages from Chesapeake Bay, USA. LVe also used CMFDA-stained Isochrysis galbana (Prymncsiophyta) and unstained Cryptomonas sp (Cryptophyta) in laboratory and field studles, respectively, to examine prevalence of feeding by C estuariale as a function of prey dens~ty However, determination of in situ grazing rates for m~xotrophic dlnoflagellates proved difficult, as only a small percentage of cells contained labeled food vacuoles follolving short incubations (5 4 h) with stained prey added at tracer concentrations. The use of CMFDA-stained cells and phvcoerythrincontaining prey as markers of ingestion should also be applicable to species-specific feeding studies with other phagotrophic protists and micro-metazoa. The protocols prescmtcd here have advantaqas over the use of fluorescent microspheres or fluorescently labeled heat-killed algae (FLA) for investigating grazing or predation because many micrograzers do not readily ingest, or discriminate agalnst. inert particles.

Dunng the austral spring, a characteristic microbial community develops in the subsurface brine pockets and channels of the annual land-fast sea-lce in Mch4urdo Sound. This community IS distinct from the d~atom-dominated commun~ty that develops in the channels at the base of the scaice, at the seawater/icc interface, and In the platelet layer The photosynthetic biomass In the brlne pockets is domlnated by athecate dinoflagellates Chrysophyte statocysts (sometimes known as archaeomonads) and <5 l m photosynthetic flagellates are also charactel-~stlcally found in thls assemblage. In December, chlorophyll a content and blornass peak, and photosynthetic gymnodinloid dinoflagellates can reach densities of over 10' ml-' of brine The photosynthetic d~noflagellates form cysts (hypnozygotes) during late December and early January, and chrysophyte statocysts also become abundant DUI-lng austral summer, total autotroph~c biomass In the upper Ice brine decreases d u e to dilutlon by melt water, flushlng of brine Into the water column, and grazlng. By late summer, the annual sea-ice in McMurdo Sound has broken out. The yearly decay and retreat of sed-ice introduces a characteristic set of brlne protists and thelr cysts into Mcblurdo Sound.

ABSTRACT. The photosynthetic c h a t e Mesodinium rubrum is a common component of the plankton in estuarine, coastal and offshore areas. Unusually high photosynthetic rates [ 2 10 pghave been measured during vlsible blooms (red-waters) of this species, but llttle data were available on photosynthesis by Mesodmium during more routine conditions. We used single cell techniques to measure chlorophyll content and rates of photosynthesis in Mesodinium (16 to 18 X 21 to 22 pm in size) that were part of mixed-species phytoplankton assemblages in small estuaries and salt ponds. The carbon:chlorophyll a (wt:wt) ratio for Mesodinium ranged from 47 to 78. Light-saturated rates of photosynthesis ranged from 13 to 88 pg C cell-' h-' 11.8 to 8.6 pg C (pg chl a)-' h-']. These speciesspecific assimilation ratios were within the mid-range reported for community measurements made during Mesodinium red-waters and within the range reported for phytoplankton. Ik values for Mesodiniurn were 2-275 pE m-' S-' in all experiments. At saturating irradiance, carbon fixation ranged up to ca 14 % of body C h-' In our incubations, Mesodinium accounted for from < 1 to 2-70 % of the community primary production in surface water samples although at no time during our studies did it cause red-waters.

No abstract