Sinking rates of heterogeneous, temperate phytoplankton populations

Throughout the summer of 1978, the sinking rates of phytoplanktonwithin the Controlled Experimental Ecosystems (CEE's) were monitoredusing a technique based upon measurement of the transit timeof radioactively (¹⁴C) labeled cells. The experimental frameworkof FOODWEB 1 offered an unprecedented opportunity to documentthe sinking rates of heterogeneous phytoplankton of diversetaxonomic composition, growing under a variety of nutrient regimes;the absence of advective exchange in the CEE's provided knowledgeof the preconditioning history of the phytoplankton sampledat any given time. Sinking rates of whole phytoplankton assemblages (not size-fractioned)ranged from 0.32 – 1.69 m·day^–1; the averagerate (± s.d.) observed was 0.64 ± 0.31 m·day^–1.The most notable deviations from the mean value occurred whenthe population size distribution and taxonomic composition shifteddue to blooms. The relationship between phytoplankton sinkingand ambient nutrient levels was studied by following the ratesof a given size fraction (8–53 µm) for ten daysfollowing nutrient enrichment of a CEE. Over this time sinkingrates ranged from 1.08– 1.53 m·day^–1; decreasedrates occurred after nutrification, yet over the course of theentire trial sinking rates were not significantly (p >0.05)correlated to the ambient levels of any single nutrient species. The sinking rate implications of spore formation were also studied,and showed that sinking rates of Chaetoceros constrictus andC. socialis spores (2.75 ± 0.61 m·day^–1)were ca 5-fold greater than rates measured when the vegetativestages of these species dominated the population, reflectingthe influence of physiological mechanisms in controlling phytoplanktonbuoyancy. An example of the potential influence of colony formation uponbuoyancy was noted in observations of C. socialis which occasionallywas found to exist in large spherical configurations made ofcoiled cell chains and having low (0.40 m·day^–1)sinking rates. A hydrodynamic rationale is presented to showhow such a colony together with enveloped water may behave asa unit particle having lower excess density, and therefore lowobserved sinking rate, despite its conspicuously large size. Overall, sinking rates were not significantly correlated withany of the measured nutrient or photic parameters. There were,however, trials and comparisons which showed evidence for: (1)higher sinking rates in populations dominated by large cells,(2) decreased sinking rates after nutrient enrichment, and (3)buoyancy response to light levels. It is suggested that correlationof sinking rates with synoptic environmental measurements atany given time is not explicit because the associations mayinvoke lag times of physiological response. The interpretationmade from these findings is that the preconditioning historyof the population, rather than the prevailing conditions atthe time of a given measurement, is most closely associatedwith population buoyancy modifications.     

Grazing and colony size development in Phaeocystis globosa (Prymnesiophyceae): the role of a chemical signal

The bloom-forming prymnesiophyte Phaeocystis globosa forms hollow,spherical, mucilaginous colonies that vary from micrometresto millimetres in size. A recent paper gave the first empiricalevidence that colony size increase in P. globosa is a defensiveresponse against grazers, and knowing the signalling mechanism(s)behind this response will thus be a key to understanding thetrophodynamics in systems dominated by this species. I conductedexperiments with specially designed diffusion incubators, eachof which consists of a non-grazing chamber (with P. globosaonly) and a grazing chamber (grazers + phytoplankton) connectedby 2 µm polycarbonate membrane filters. The results showedthat physical contact with grazers was not required to initiatethe defensive response; instead, P. globosa colony size increasewas found to be stimulated by dissolved chemicals generatedby ambient grazing activities. This signal was non-species specific,such that various combinations of three species of grazers andfour species of phytoplankton in the grazing chambers all resultedin significant, but different extents of colony enlargementin P. globosa in the non-grazing chambers (30–300% largerthan the ‘grazer-free’ control). High concentrationsof ambient solitary P. globosa cells and other phytoplanktonseemed to suppress colony enlargement in P. globosa, and grazerswould help reduce this inhibition by removing the ambient solitaryP. globosa cells and other phytoplankton. These non-species-specificmechanisms would allow P. globosa to regulate colony size developmentand defend itself in diverse planktonic systems, which may helpto explain the global success of this species.     

A model of phytoplankton production in the lower River Rhine verified by observed changes in silicate concentration

The production of phytoplankton in the three main branches andsedimentation areas of the River Rhine in the Netherlands wasanalyzed using a simulation model describing the carbon andsilicate metabolism. This model is based on data derived froma sampling programme in which river water was followed duringdownstream transport. A ‘plug-flow model’ was developed,including sky irradiance and light attenuation in the water,and integrating photosynthetic rates determined in the laboratory.On the basis of the silicate content of diatom-dominated phytoplanktonand silicate regeneration in the river bottom, changes in silicateconcentrations were simulated and found to match observed changesin dissolved silicate. Low sìlicate concentrations wereshown to restrict the maximum population density of diatoms.Depth- and time-integrated rates of photosynthesis were shownto permit multiplication of the phytoplankton at a rate of upto one doubling day^–1 In the primary production periodApril-August 1988. values of 0.48–6.33 g C m^–2 day^–1,close to the few values reported for highly eutrophic riversand lakes, were observed. Model runs, including phytoplanktonproduction and losses, such as respiration, sedimentation andplanktonic grazing, were carried out to simulate the downstreamdevelopment of phytoplankton biomass. These simulations confirmthe view that a substantial part of the phytoplankton biomassand production is grazed or settles in the river delta despiteresidence times of only 52–97 h.     

Species-specific phytoplankton sedimentation in relation to primary production along an inshore--offshore gradient in the Baltic Sea

The temporal and spatial variability in the quality and quantityof settling phytoplankton material in relation to concurrentprimary production was studied using sediment traps at threecoastal stations from a semi-enclosed bay (Pojo Bay) throughthe outer archipelago to the open Gulf of Finland. The fluxof settling phytoplankton was high (9.3 g C m^–2period^–1)in Pojo Bay, especially in spring, and lower in the archipelago(8.1 g C m^–2 period^–1) and open-sea area (5.2 gC m"2 period"1), although the primary production followed theopposite pattern. A large influx of allochthonous material intoPojo Bay in spring brought allochthonous phytoplankton cellsinto the traps, but limited primary production. Diatoms werethe most abundant settled phytoplankton at all stations, butthe species composition varied between Pojo Bay (Aulacoseiraspp., Rhizosolenia minima) and the outer stations (Skeletonemacostatum, Chaetoceros spp.)At the outer stations, migratingdinoflagellates (Peridiniella catenate) comprised part of thesettling material in spring. The high settling flux of the cyanophyteAphanizomenon flos-aquae is discussed. The species compositionof the phytoplankton assemblage influenced the proportion ofthe total organic carbon sedimentation that consisted of phytoplanktoncarbon.     

Pigment-specific rates of phytoplankton growth and microzooplankton grazing in a subtropical lagoon

Phytoplankton growth and microzooplankton grazing rates wereevaluated in one station in Bahía Concepción,located in the middle region of the Gulf of California, México.We used high-performance liquid chromatography (HPLC) estimationsof phytoplankton pigment signatures to evaluate the annual variationof taxon-specific grazing and growth rates obtained with thedilution technique. Chlorophyll-a (Chl-a) concentrations variedwidely (0.34–3.32 µg L^–1) and showed two maxima,during late spring and autumn, associated with the transitionbetween mixed and stratified conditions. Phytoplankton growthrates varied seasonally with the lowest rates during summer(range: 0.01–2.55 day^–1 for Chl-a; 0.00–3.84day^–1 for Chl-b; 0.26–3.29 day^–1 for fucoxanthin;0.00–6.27 day^–1 for peridinin; 0.00–4.35 day^–1for zeaxanthin). Microzooplankton grazing was an important lossprocess (range: 0.0–1.89 day^–1 for Chl-a; 0.00–3.12day^–1 for Chl-b; 0.26–3.29 day^–1 for fucoxanthin;0.00–2.03 day^–1 for peridinin; 0.00–3.51 day^–1for zeaxanthin). Average grazing rates accounted 68–89%of estimated average phytoplankton pigment-specific growth rates.The analysis of pigment signatures indicates that diatoms anddinoflagellates were the dominant groups, and contrary to expectationfor typical subtropical lagoons, the specific growth rates inBahía Concepción showed a pronounced seasonalvariability, linked to transitional hydrographic conditions.Our results indicate a close coupling between the communitymicrozooplankton grazing and phytoplankton growth rates, withoutselective feeding behavior. These results suggest that microzooplanktonplay a critical role and may significantly modify the availabilityand efficiency of transfer of energy to higher trophic levels.     

Ecological investigations of blooms of colonial Phaeocystis pouchetti. II. The role of life-cycle phenomena in bloom termination

A bloom of the colonial stage of the prymnesiophyte Phaeocystispouchetii was studied for 2 months in a 1^3-m3 flow-through mesocosm.Phaeocystis increased in abundance for 6 weeks coincident withdeclining temperature and nutrient supply rates. Experimentssuggested that colony growth was primarily nitrogen-limitedduring this period. An extended period of subzero temperaturesand nutrient deprivation was associated with a mass exodus ofcells from the colonies. Previously non-motile cells developedflagella, became motile and emigrated out of the colonies, accompaniedby significant decreases in the chlorophyll a content and photosyntheticrates of the colonies. Concentrations of bacteria on the surfacesof such ‘ghost’ colonies were two orders of magnitudehigher than on ‘normal’ colonies. Growth rate studiesof field populations indicated that rapid declines in temperatureinduced development of motility and emigration from the colonies.Ancillary observations implied that chronic nutrient deprivationresulted in similar life-cycle events. Warming and nutrientaddition did not halt release of swarmers, suggesting that,once initiated, the process proceeds to completion. The combineddata indicate that blooms of colonial Phaeocystis, unlike manyother phytoplankton, are not necessarily terminated by grazingor sinking out of the euphotic zone. The physiological optionof motility and emigration provides Phaeocystis with an ecologicalalternative which has significant implications in interpretingthe structure and function of plankton communities.     

Sedimentation loss of phytoplankton cells from the mixed layer: effects of turbulence levels

In this paper, we describe a study of the role of turbulencein the loss by sedimentation of phytoplankton cells from themixed layer. The approach presented allows the quantificationof the sedimentation rate of phytoplankton in the whole rangeof turbulence levels of this layer. Two types of phytoplanktercan be distinguished according to the effect that turbulencecan exert on their sedimentation rate. The rate of those cellswhose settling velocity is lower than –1 m day^–1will not be modified by turbulence. The sedimentation rate ofcells with higher settling velocities can, however, be modifiedby the level of turbulence. A set of dimensionless numbers isgiven to delimit several processes that are important in thedynamics of phytoplankton sedimentation in a turbulent regime.The use of these dimensionless numbers suggests that an increasein the turbulence level in the mixed layer does not always implya decrease in the sedimentation rate of phytoplankton cells.     

Photosynthetic algal production, accumulation and release of phytoplankton storage carbohydrates and bacterial production in a gradient in daily nutrient supply

In a mesocosm experiment providing a gradient of semi-continuousaddition of mineral nutrient, production rates and mortalityof phytoplankton were estimated. Heterotrophic bacterial biomassand production rates and their responses to the mineral nutrientsadditions were also estimated. The purpose of the experimentwas to establish responses of the major biological factors asa function of nutrient amendments. Initial primary productionwas 0.47 µg C L^–1 day^–1. In the most fertilizedmesocosm, phytoplankton biomass increased at a specific rateof 0.4 day^–1 during the first week of the experiment,and on day 9 primary production reached a peak at 1027 µgC L^–1 day^–1. The responses in the other fertilizedmesocosms were intermediate, and in an unfertilized controlthe variables measured stayed almost constant throughout theexperiment. The termination of the blooms in the fertilizedmesocosms was a consequence of nitrogen limitation, and nitrogenlimitation subsequently induced storage of intracellular organicmaterial in the phytoplankton. In the mesocosm receiving thehighest daily dose of nutrients, strong post-bloom nutrientlimitation resulted in high phytoplankton mortality, and releaseof organic material from the algae supported the gradient’shighest heterotrophic bacterial production.     

Community structure, biomass and productivity of size-fractionated summer phytoplankton populations in lower Narragansett Bay, Rhode Island

Seventeen size-fractionation experiments were carried out duringthe summer of 1979 to compare biomass and productivity in the< 10, <8 and <5 µm size fractions with that ofthe total phytoplankton community in surface waters of NarragansettBay. Flagellates and non-motile ultra-plankton passing 8 µmpolycarbonate filters dominated early summer phytoplankton populations,while diatoms and dinoflagellates retained by 10 µm nylonnetting dominated during the late summer. A significant numberof small diatoms and dinoflagellates were found in the 10–8µm size fraction. The > 10 µm size fraction accountedfor 50% of the chlorophyll a standing crop and 38% of surfaceproduction. The <8 µm fraction accounted for 39 and18% of the surface biomass and production. Production by the< 8 µm fraction exceeded half of the total communityproduction only during a mid-summer bloom of microflagellates.Mean assimilation numbers and calculated carbon doubling ratesin the <8 µm (2.8 g C g Chl a^–1 h^–1; 0.9day^–1)and<5 µm(1.7 g C g Chl a^–1h^–1; 0.5day^–1)size fractions were consistently lower than those of the totalpopulation (4.8 g C g Chl a^–1 h^–1; 1.3 day^–1)and the <10 µm size fraction (5.8 g C g Chl a^–1h^–1; 1.4 day ^–1). The results indicate that smalldiatoms and dinoflagellates in fractionated phytoplankton populationscan influence productivity out of proportion to their numbersor biomass. ¹Present address: Australian Institute of Marine Science, P.M.B.No. 3, Townsville M.S.O., Qld. 4810, Australia.     

Phytoplankton community structure derived from HPLC analysis of pigments in the Faroe-Shetland Channel during summer 1999: the distribution of taxonomic groups in relation to physical/chemical conditions in the photic zone

A study of the phytoplankton community in the Faroe-ShetlandChannel was conducted in July 1999. Samples were collected atvarious depths in the photic zone along three transects (thenorthern entrance, the center and the southern entrance). Exceptfor a few easterly stations where nitrate and silicate werebelow 1 µM, all nutrients (phosphate, silicate, ammonium,nitrite and nitrate) were non-limiting for phytoplankton growth.HPLC pigment analysis revealed a pronounced (>50%) dominanceof Prymnesiophyceae at all stations. Their pigment ratio ofdiatoxanthin + diadinoxanthin/Chl a (DDX/Chl a) indicated thatthe phytoplankton community was controlled by light. Primaryproduction in the delayed spring bloom varied from 1.2 to 1.8g C m^–2 day^–1 along the northern transect. Alongthe other two transects, primary production ranged from 1.6to 3.8 g C m^–2 day^–1. Associated with the characteristicsindicating the establishment of a bloom, the relative contributionof diatoms and Prymnesiophyceae increased, whereas that of Prasinophyceae,Cryptophyceae, Chrysophyceae and Cyanobacteriaceae decreased.With respect to their vertical distribution, Cyanobacteriaceae,Chrysophyceae and Dinophyceae tended to have a higher abundance,relative to other taxonomic groups, in the surface layers. Therelative abundance of diatoms and Chlorophyceae increased withdepth. The DDX/Chl a ratio of the Prymnesiophyceae decreasedwith depth, indicating that vertical mixing in the upper 30m of the photic zone occurred less frequently than the timespan of physiological acclimation of cellular pigment composition.     

Productivity, growth rates and cell size distributions of phytoplankton in the SW Tasman Sea: implications for carbon metabolism in the photic zone

Data are presented on primary productivity, cell size distributionsand the standing stocks of living and detrital paniculate organiccarbon (POC) in the waters of the SW Tasman Sea. Primary productivitywas measured by both standard 4- and 12-h incubations as wellas time-series incubations. Data are presented for ¹⁴C uptakeand loss in 12L/12D methods. The importance of time zero anddark controls is demonstrated. The uptake of ¹⁴C in the lightwas linear and the loss of label in the following dark periodranged from zero to 39%. The loss of label in the dark was correlatedwith the particle size distribution, being greatest in oligotrophicwaters dominated by small cells (25–30%) and least inspring bloom areas (0–20%) dominated by large diatoms.Kinetic data were strongly supportive of a major grazing impactby microphagous organisms. The data were an experimental confirmationof recent theoretical models of ¹⁴C uptake and grazing. Sizedistributions of phytoplankton and detritus were measured byHIAC and by microscopic counting. The phytoplankton consistedof a ubiquitous group of picoplankton, and variable contributionsfrom small flagellates and diatoms. The distribution of totalcell volume was dominated by large cells in spring bloom areas.Chlorophyll concentrations were strongly correlated with themean cell size of the phytoplankton. Comparison of the resultsof ¹⁴C uptake experiments with the results of experiments todetermine changes in POC, by counting particles, gave good correlation.In detail, the comparison of the methods revealed systematicerrors with the greatest discrepancy between the methods atlow apparent growth rates. The detritus showed constant sizedistributions in surface waters. The mean size of detritus particlesreduced rapidly with depth and declined in a way suggestingbiological reprocessing and removal by grazing. These resultsare discussed in the context of the patterns of carbon metabolismin the photic zone, the role of living and detrital POC andthe balance of ‘new’ versus ‘regenerated’production in surface waters.     

Pyrosoma atlanticum (Tunicata, Thaliacea): grazing impact on phytoplankton standing stock and role in organic carbon flux

Pyrosomas are the large group of pelagic tunicates whose trophicrole in pelagic communities has not yet been sufficiently studied.We ran across a local area of high concentration of the mostwidespread and commonest species of pyrosomas, Pyrosoma atlanticum,450 miles off the Congo river mouth. The following was estimated:gut pigment content, defecation rate, organic carbon and pigmentcontent of fecal pellets, and sinking rate. Based on these dataand the measured number of pyrosomas colonies the grazing impacton phytoplankton and the fecal pellet flux were calculated.During the night swarms of 50–65 mm P.atlanticum removed53% of phytoplankton standing stock in the 0–10 m layer;sparsely distributed pyrosomas consumed only 4%. The grazingimpact in the 0–50 m layer was only 12.5 and <1% respectively.The fecal pellet flux resulting from nocturnal feeding of P.atlanticumwhile swarming made up 1.4–1.6 x 10⁶ pellets m^–210 h^–1 or 305–1035 mg C m^–2 10 h^–1 and1.4 x 10⁵ pellets m^–2 10 h^–1 or 87.4 mg C m^–210 h^–1 while non-swarming. Incubation experiments showedthe rapid degradation of fecal pellets at 23°C: the lossof pigment and carbon content was {small tilde}60–70%after 45 h. We believe that given the sinking rate of 70 m day^–1the main part of fecal material does not leave the upper watercolumn and is retained in the trophic web of the epipelagiclayer.     

Daily irradiance governs growth rate and colony formation of Phaeocystis (Prymnesiophyceae)

Phaeocystis was cultured at a range of ecologically significantdaily irradiances under nutrient-replete conditions. Below athreshold of 100 W h m^–1 day^–1, the cells were smalland flagellated, and remained solitary. Above this threshold,the cells were larger and able to form colonies. Growth rateand colony formation were maximum at sea surface irradiances(>700 W h m^–2 day^–2). Presumably, colonial growthis a strategy to maintain optimum growth rates in the watercolumn. Sinking, nutrient-stressed colonies reach low irradiancesand colonial cells can transform into small solitary flagellatedcells. These observations are important in understanding theecology and life cycle of Phaeocystis.     

Red tide assemblage formation in an estuarine upwelling ecosystem: Ria de Vigo

Red tides are conspicuous in the upwelling system of Galicia(NW Iberian Peninsula). At present, there are conflicting hypothesesabout the generation site of these phytoplankton assemblages.It is interesting to know whether the rias can be sites of redtide formation or if they act only as accumulation sites ofpopulations advected from shelf waters. A study in the Ra deVigo, carried out during late September 1990, showed the developmentof a red tide assemblage, composed of Alexandrium affinis, Ceraiiumfusus and Gymnodinium catenaium, during a 2 week upwelling-downwellingcycle. Growth occurred at the bottom of the thermocline-topof the nutricline. Above this assemblage, a diatom assemblage(large diatoms) was blooming. Prior to the formation of thered tide, a subsurface chlorophyll maximum made up of smalldiatoms (Nilzschia f. seriaia, Chaeloceros socialis), smallflagellates (<30 µm) and small gymnodinid forms (<30µm) was observed. In the nutrient-depleted upper layer,several autotrophic and large heterotrophic dinoflagellatesdominated. It is suggested that the ratio between the velocityof upward water movement and the depth of the stratified upperlayer (flushing rate, day^–1) is the critical parameterwhich triggers active phytoplankton growth. It can be concludedthat upward water velocities of {small tilde}2.5 m day^–1and a stratified upper layer of 10 m depth (flushing rate 0.25day^–1) are the main physical constraints for red tidedevelopment.     

Sinking losses of phytoplankton in closed limnetic systems

Specific algal recoveries from sediment traps of two differentdesigns and from mud surface deposits of large experimentalenclosures (Lund Tubes) were monitored during 1978 and are analyzedin relation to the vertical and temporal distribution of tendominant phytoplankton populations. Sedimentation accounts fordiffering proportions of the total loss of biomass for differentalgae: between 28 and 100% of diatoms; 15–95% of Eudorina;<4% of populations of small algae (spp. ofAnkyra, Chromulina,Cryptomonas). Rates of diatom loss are also derived from thecomparison of net rates of change (k_n) and the silica uptake-derivedgrowth rate (k¹); intrinsic sinking behaviour may be specificallyregulated in relation to growth conditions. Implications inthe calculation of sedimentary losses and their impact uponthe seasonal periodicity of phytoplankton are briefly discussed.     

Dominance of microflagellates over diatoms in the Antarctic areas of deep vertical mixing and krill concentrations

Phytoplankton data obtained during six summer Polish expeditionsto the Antarctic Peninsula area, are compared with concurrentlyrecorded data on water column stabilities and krill abundance.The results show that flagellates (1.5–20 µm) arenumerically dominant over diatoms in the areas of deep verticalmixing and/or extensive krill concentrations. Of 102 stationsdominated by flagellates, 85 (83.3%) are located in a well mixedwater column (>100 m) and correspond to a mean krill densityof 15–346 t Nm^–2. In the same areas, estimated flagellatecarbon biomass exceeds diatom carbon. On the other hand, ofthe 40 stations dominated by diatoms, 36 (90%) are located inareas of increased water column stability (upper mixed layerof 10–50 m) and correspond to a low mean krill biomassof 0.34–4.6 t Nm^–2. Positive correlations of flagellateto diatom (F:D) cell number ratios with the depth of the uppermixed layer suggest light limitation of diatom growth and anincreased sinking rate of diatoms relative to flagellates inthe areas of deep vertical mixing. The relationship of the F:Dratio with krill abundance suggests that krill prefer feedingon diatoms and are less efficient in grazing particles of thesize of microflagellates (<20 µm). Flagellates exceeddiatoms in an unstable water column when the phytoplankton populationsare low; both algal groups increase in numbers with growingstability. The results provide field evidence that deep verticalmixing and krill grazing create conditions for the dominanceof flagellates over diatoms. Both factors acting together arelikely to suppress diatom blooms in the Antarctic.     

Microzooplankton grazing in a coastal embayment off Concepcion, Chile, (~36{degrees} S) during non-upwelling conditions

The impact of grazing by natural assemblages of microzooplanktonwas estimated in an upwelling area (Concepción, Chile)during the non-upwelling season in 2003 and 2004. Seawater dilutionexperiments using chlorophyll a (Chl a) as a tracer were usedto estimate daily rates of phytoplankton growth and microzooplanktongrazing. Initial Chl a concentrations ranged from 0.4 to 1.4mg Chl a m^–3 and phytoplankton prey biomass and abundancewere numerically dominated by components <20 µm. Phytoplanktongrowth and microzooplankton grazing rates were 0.19–0.25day^–1 and 0.26–0.52 day ^–1, respectively.These results suggest that microzooplankton exert a significantremoval of primary production (>100%) during the non-upwellingperiod.     

Noctiluca scintillans MACARTNEY in the Northern Adriatic Sea: long-term dynamics, relationships with temperature and eutrophication, and role in the food web

The first ‘bloom’ of Noctiluca scintillans in theNorthern Adriatic Sea was recorded in 1977. The organism causedseveral red tides in the whole basin during the late 1970s,a period characterized by increasing nutrient loads. Duringthe 1980s and early 1990s, there was no ‘red tide’,but the species was an almost constant summer presence, associatedwith high temperatures. Noctiluca scintillans was almost completelyabsent from 1994 until May 1997, concurrent with a general planktondecrease. From summer 1997, N. scintillans was recorded againin the whole basin, although there was no other signal of increasingeutrophication. In contrast to all previous observations, duringwinter 2002–2003, N. scintillans was continuously sampledin the Gulf of Trieste. We estimated experimentally growth andgrazing rates of the dinoflagellate at 9–10°C in cultureand consuming the natural assemblage. Noctiluca scintillanswas able to reproduce actively at low temperatures, showingsimilar growth rates in both experiments (k = 0.2 day^–1).The values found were close to those reported in the literaturefor higher temperatures. The natural diet was mainly composedof phytoplankton (ingestion = 0.008 µg C Noctiluca ^–1day^–1), microzooplankton (ingestion = 0.008 µg CNoctiluca ^–1 day^–1) and bacteria (ingestion = 0.005µg C Noctiluca ^–1 day^–1) with an average carboncontent of 0.138 ± 0.020 µg C Noctiluca cell^–1.     

Short-term variability in the flux of rapidly sinking particles in the Antarctic marginal ice zone

A sediment trap deployment was made at a station (64°42'S, 139°59'E) at five depths (537, 796, 1,259, 1,722, 2,727 m) in the marginal ice zone (MIZ) of the Antarctic Ocean during a summer productive period from 26 December 1994 to 20 January 1995. This aim of the study was to reveal a possible occurrence of a sporadic bloom in surface layers and to evaluate the role of fast-sinking particles in transportation processes of bloom-derived material down to mesopelagic and bathypelagic layers. During the observation, a marked flux increase (70.5 mg C m^-2 day^-1, 7.7 mg N m^-2 day^-1) was observed at the depth of 537 m on 7-9 January. The increased flux at 537 m decreased with depth and time. The same mass of sinking particles forming the flux maximum at each depth sank down from the shallowest trap (537 m) to the deepest trap (2,727 m) within 4-11 days, indicating that these particles were transported downward to the bottom with the sinking rate of >200 m day^-1. Collected particles were composed of two major particle fractions; one was dominated by fecal pellets of macrozooplankton (mainly Euphausia superba) with relatively fast sinking rates (FSP; fast-sinking particles) and the other by minute diatoms of Fragilariopsis curta with slow sinking rates (SSP; slowly sinking particles). According to the comparison of time depth changes of these two fractions, the SSP had unexpectedly faster sinking rates comparable with the FSP during the periods of maximum fluxes, probably indicating the SSP were transformed from the FSP during sinking. The present result strongly suggests that a local bloom of F. curta and intensified zooplankton grazing activities occurred in surface layers in a few days in the MIZ, and then the egested fecal pellets were rapidly transported downward with fragmentation processes into small-sized minute particles in mesopelagic and bathypelagic layers.     

Buoyancy properties of the giant diatom Ethmodiscus

Net-collected Ethmodiscus spp. were examined and found to becapable of positive buoyancy with ascent rates averaging 1.4m h^–1 on individually measured cells, and 0.3 m h^–1using a homogeneous settling technique. These data, togetherwith published reports of positive buoyancy in Rhizosolenia,indicate that upward movement is a widespread phenomenon inlarge oceanic diatoms. These observations also suggest thatpublished size-rate relationships for marine diatoms requirerevision to include the effects of positive buoyancy.