Viral Lysis of Phaeocystis Pouchetii and Bacterial Secondary Production

AQUATIC MICROBIAL ECOLOGY Published October 15 Aquat Microb Ecol

Viral lysis of Phaeocystis pouchetii and bacterial secondary production

Gunnar ~ratbak'.', Anita Jacobsen2, Mikal Heldall

'~epartmentof Microbiology. University of Bergen, Jahnebakken 5, N-5020 Bergen. Norway 2~epartmentof Fisheries and Marine Biology, University of Bergen, Bergen High Technology Centre, N-5020 Bergen, Norway

ABSTRACT: In this experimental study we investigated the effect of viral infection on primary production and carbon flow in a phytoplankton-DOC-bacteria food chain during viral lysis of the phytoplankton population. The phytoplankter host-virus system used was Phaeocystis pouchetii (Pryrnnesio- phyceae) and the virus PpVO1. Viral infection allowed primary production in the cells to continue throughout most of the lytic cycle. In non-infected algal cultures, net production of DOC and bacterial biomass was low and at the end of the experiment the DOC concentration was 10 to 2096,and the bac- tenal biomass 0.5 to 4 % of the algal carbon biomass. The amount of DOC released during viral lysis of the algal cells implies that the enhre algal biomass was converted to DOC. Growth of bacteria succeeding cell lysis and release of DOC in virus infected cultures demonstrated that the net effect of the virus infection was an efficient conversion of algal biomass into bacterial biomass.

KEY WORDS: Phaeocystis pouchetii - Virus . DOC . Bacteria . Microbial food web

INTRODUCTION high and variable abundance of virus and virus-like particles in aquatic environments, and the occurrence The microbial loop, in which dissolved organic car- of such particles intracellular in a variable fraction of bon (DOC) lost from phytoplankton is assimilated by natural phytoplankton and bacterial communities. bacteria and transferred back to the grazing food chain The abundance of viruses and virus-like particles via bacterivorous microzooplankton, has been recog- (VLP)in aquatic marine ecosystems is typically in the nized as an important pathway in the microbial food order of 106 to 10' rnl-' (Maranger & Bird 1995). Most of web of aquatic ecosystems (Azam et al. 1983, 1993). these viruses are presumably bacteriophages (Bratbak Some 20 to 50% of the primary production is typically et al. 1990) but phycoviruses and viruses infecting found to pass through the microbial loop (Cole et al. other microorganisms may also be present in signifi- 1988). cant numbers (Proctor & Fuhrman 1990, Bratbak et al. Phytoplankton excretion and zooplankton grazing 1993, Suttle & Chan 1993, 1995, Waterbury & Valois and sloppy feeding have been considered as the quan- 1993). Electron microscopy studies of microorganisms titatively most important internal sources of DOC in in natural waters have revealed that 1 to 40% of the natural waters (Ducklow & Carlson 1992, Fuhrman bacterial community [Bratbak et al. 1994), and a vari- 1992). It has recently been suggested that viral infec- ety of eucaryotic microorganisms (van Etten et al. tion causing cell lysis may be of importance to carbon 1991, Reisser 1993, Zingone 1995) contain virus-like and nutrient flow in the microbial food web (Fuhrman particles. Total counts of viruses have been found to & Suttle 1993, Bratbak et al. 1994, Suttle 1994). The vary significantly in time and space, and have been observations which have led to this suggestion are the correlated to biological parameters such as chloro- phyll a concentration, bacterial abundance and bacterial activity (Bratbak et al. 1994, Maranger & Bird 'E-mail: [email protected] 1995). The methods used to assess the rate of viral lysis

O Inter-Research 1998 12 Aquat Microb Ecol 16: 11-16, 1998

of bacteria in natural waters are uncertain, but they MATERIALS AND METHODS indicate that a considerable fraction (Prymnesiophyceae) dal 1995, Hennes & Simon 1995, Mathias et al. 1995). strain AJO1, obtained from the culture collection at the There is also growing evidence that virus infection is a University of Bergen, Norway. This Phaeocystis strain quantitative significant cause of mortality in many does not form colonies under the growth conditions used phytoplankton populations (Proctor & Fuhrman 1990, in this study. The lytic virus PpVOl was isolated from Bratbak et al. 1993, 1995, Waterbury & Valois 1993, Raunefjorden, Western Norway (Jacobsen et al. 1996). Cottrell & Suttle 1995, Brussaard et al. 1996). Virus- The algal culture and virus lysate were not axenic. induced cell lysis may thus be a sizable source of DOC The cells were grown in a sterile filtered (0.2 pm) sea in natural environments. water medium prepared from aged sea water (33%0) A virus lytic to the marine haptophyte Phaeocystis with vitamin and trace element additions as for the f/2 pouchetii has recently been brought into culture medium (Guillard 1975). Nitrate and phosphate were (Jacobsen et al. 1996). Phaeocystis is an ecologically added from autoclaved stock solutions to give final important phytoplankton genus in many areas concentrations of 80 pM NaNO, and 5 pM KH2P0,. For (Lancelot et al. 1994) and may completely dominate all experiments we used 1 1 cultures growing in 2 1 spring blooms in regions as different in temperature, Erlenmeyer flasks. The incubation temperature was light conditions, nutrient status, and water column sta- 8°C and the illumination (continuous light) from white bility as e.g. the Barents Sea (Loeng et al. 1987) and the fluorescent tubes was 40 to 50 pm01 m-2 S-' measured North Sea coast of the European continent (Lancelot et at the surface of the cultures. The cultures were gently al. 1987). Phaeocystis from such blooms may sink swirled once a day to prevent sedimentation. (Wassmann et al. 1990, Wassmann 1994) or be con- Experimental design. For the carbon flow studies we sumed by predators (Admiraal & Venekamp 1986, used four 1 1 batch cultures of exponentially growing Huntley et al. 1987, Estep et al. 1990), but the collapse Phaeocystis pouchetii. Two of the cultures were inocu- of such blooms has also been found to leave high con- lated with a fresh virus lysate to give an initial virus-to- centrations of dissolved carbohydrates and organic host ratio of 0.1 to 0.2. The remaining 2 cultures served nitrogen compounds in the photic zone (Eberlein et al. as control and received lysate that had been 0.2 pm fil- 1985), as one would expect if blooms terminate due to tered to remove bacteria and heated to boiling in a massive cell lysis. Since the mechanism of bloom col- microwave oven to inactivate the viruses. lapse may be expected to strongly influence the post- To investigate the effect of virus infection on photo- bloom concentrations of organic and inorganic nutri- synthesis we measured the rate of 14C-CO2incorpora- ents in the photic zone, the whole post-bloom tion at several time-points before and after inoculating successional pattern of new phytoplankton species, an exponentially growing Phaeocystis pouchetii cul- heterotrophic bacteria, and protozoan predators would ture with a fresh virus lysate to give an initial virus-to- be expected to depend on the relative dominance of host ratio of about 10. A non-infected culture was used the different bloom-terminating mechanisms. as control. Samples for cell and virus counts, and for Virus infection and cell lysis of algae will inevitably analysis of inorganic and dissolved organic carbon, have a major impact on the cells' CO2 fixation and on were analyzed as described below. carbon flow in the foodweb. The purpose of this 14C incubations. CO2 incorporation was measured in experimental study was to investigate the carbon triplicate by adding 1 pCi '4C-C02 (New England flow and population dynamics in a phytoplankton- Nuclear) to 10 ml samples and incubating for 1 h under DOC-bacteria food chain dnring viral lysis of the the conditions described above. 5 m1 from each paral- phytoplankton population. The effect of viral infec- lel was transferred to scintillation vials, acidified with 2 tion on primary production was examined by measur- drops of 8 N HCI and left open for >24 h to remove ing 14C-CO2 fixation in infected and non-infected unincorporated CO,. Controls were acidified immedi- algal cultures. The specific questions we have ap- ately after isotope addition. Independent tests showed proached are related to timing of events and to mag- that no CO2 remained in the acidified samples after nitude and rate of carbon flow: At what stage in the 24 h. The samples were neutralized by adding 2 drops lytic cycle is the CO2 fixation shut down? How much of 8 N NaOH and the radioactivity measured in a of and how fast is the cellular biomass converted to Packard Tri-Carb scintillation counter after adding dissolved material during cell lysis? Is the organic 5 ml Ultima Gold XR (Packard) scintillation cocktail. material released during cell lysis readily available Enumeration and chemical analysis. Growth of the for bacterial secondary production or does it accumu- algae was monitored by live cell counts in a Fuchs late in the water? Rosenthal haematocytometer. Samples for total count- Bratbak et al.: Viral lysis of Phaeoc)/stispouchet~l 13

ing of bacteria and viruses were preserved with 2.5 % glutaraldehyde. Particles were harvested onto electron microscope grids (Ni, 400 mesh) by centrifugation and prepared for counting in the transmission electron microscope (TEM) basically as described earlier (Brat- bak & Heldal 1993). The centrifuge tubes we used were cut to accommodate ca 3 m1 of sample water and the samples were centrifuged for 30 min at 200 000 X g in a Beckrnan SW 41 swing-out rotor. The grids were positive stained with 2 % uranyl acetate and viewed in a Jeol 100CX or a Jeol 100s TEM at 20000x magnifi- cation. Carbon biomass of algae and bacteria was esti- 0 Phaeocysfis mated from volume measurements in phase contrast microscope (algae) and in TEM (bacteria). Bacteria DOC and inorganic carbon (IC) were measured in a Shimadzu TOC-5000 total organic carbon analyzer. Samples for DOC analysis (3 parallels of 5 ml) were filtered through 0.2 pm Supor 200 filters (Gelman Sci- ences), preserved with 50 p1 0.5 N HC1 and stored at 4°C until analysis. The initial DOC concentration in the seawater medium was 2.65 mgC 1-l. This amount was subtracted from all DOC values measured throughout the experiment to focus on net changes. Samples for IC T~rne,days analysis (3 parallels of 5 ml) were measured irnrnedi- Fig. 1.Carbon flow experiment. (a) Carbon flow and popula- ately after sampling. tion dynamics in 2 parallel cultures of Phaeocystis pouchetii infected with PpVOl virus at Day 8. (b) Carbon flow and pop- ulatlon dynamlcs In 2 parallel non-infected (control) cultures. A background DOC concentration in the medium of 2.65 mgC RESULTS AND DISCUSSION 1.' was subtracted from all DOC values to focus on net changes. Dotted vertical line marks the time of virus addition Biomass of algae and bacteria

Measured in the phase contrast microscope the Carbon flow mean size of living Phaeocystis pouchetii cells was about 5 pm in diameter (range 4 to 6 pm) from which Viral infection perturbed the exponential growth and we estimate a carbon content of 6.5 pgC cell-', assum- decimated the Phaeocystispopulation withn 3 d (Fig. la) ing a cell density of 1 and a 10 % carbon-to-wet weight while in the non-infected culture growth continued ratio. Measured in the TEM, the mean volume of glu- undisturbed (Fig. lb). Virus-like particles were never taraldehyde preserved bacteria was 0.3 + 0.03 pm3 observed in the non-infected cultures, suggesting that (mean SE, n = 19). Choosing a conservative volume there was no background of non-virus virus-like parti- to carbon conversion factor of 200 fgC pm-3 (cf. Duck- cles in the cultures and that the non-infected control low & Carlson 1992, Fuhrman 1992) we arrive at a cultures were indeed uninfected. The concentration of mean bacterial carbon content of 60 fgC cell-'. In pre- DOC and the abundance of bacteria showed no signifi- liminary experiments the carbon content in single cells cant changes in any of the cultures before the addition of P. pouchetii and of bacteria growing in the culture of viruses (Fig. la). The concentration of DOC in- was (Heldal et al. 1985, Norland et al. 1995) found to be creased when the algae lysed and this increase was fol- 5.6 + 0.8 pgC cell-' and 62 + 7 fgC cell-' (mean rt SE) lowed by an increase in bacterial abundance (Fig. la). respectively by quantitative X-ray microanalysis. The The same successional pattern between algae, DOC biomass estimates obtained by X-ray microanalysis and bacteria was observed during a lysis induced de- and by volume measurements were thus in agreement cline of a Phaeocystis spring bloom in the North Sea both for algae and for bacteria. The number of cells (van Boekel et al. 1992). At the end of the experiment sized was too low to reveal any significant changes in (Days 11to 13) the amount of DOC and of bacterial bio- cell volume throughout the experiment and we thus mass was l0 to 20 and 0.5 to 4 % respectively compared adopt 6 pgC algal cell-' and 60 fgC bacterial cell-' as to the algal carbon biomass. In the non-infected culture conversion factors to estimate the carbon content of the there was no substantial increase in DOC or bacterial respective populations from cell counts (Fig. 1). biomass compared to the algal biomass (Fig. lb). Aquat Microb Ecol 16: 11-16, 1998

The peak in algal biomass in the infected culture was 1.8 mgC 1-l, but 0.5 to 0.6 mgC 1-' was then already released as DOC due to viral lysis of the cells. The initial increase in viral abundance (Days 8.1 to 9.4) was 2.6 X 10' viruses ml-l which corresponds to a loss of ca 9 X 104 algal cells ml-' or 0.6 mgC 1-' assuming an average burst size of 270 viruses per cell. The latter fig- ure was estimated from the first net decrease in algal cells and the concurrent increase in viruses (Days 9.4 to 10.4). The estimated loss in algal biomass is in agreement with the observed net increase in DOC of 0.5 mgC 1-' between the addition of virus and the peak in algal biomass. The total algal biomass production may thus be estimated to have been about 2.4 mgC 1-' which compares well with the observed maximum DOC concentration of 2.45 mgC 1-l. The bacterial biomass was 1.8 mgC 1-' when sampling for this parameter was terminated (Day 12). The DOC concentration was at this time still above background concentration (i.e. above 2.65 mgC 1-') and fur- ther bacterial growth may thus have taken place. The ratio between DOC produced and algal biomass produced was close to 1, indicating that the entire algal biomass was converted to DOC upon cell lysis (Fig. 1). The ratio between bacterial biomass produced and DOC produced was as high as 0.7 although the bacterial biomass production may have been underestmated. In this carbon budget we have not accounted for extracellular organic polymers or slime which is not particulate in the sense of cellular nor dissolved in the sense that it passes the 0.2 pm filters used to separate particulate and dissolved material. Algal production and bacterial utiliza- tion of such material may explain the apparently high bacterial growth efficiency. The overall conclusion is nevertheless that algal carbon released upon viral lysis was efficiently converted to bacterial biomass. For viral infection of Aureococcus anophagefferens, Gobler et al. (1997) estimated that accumulated bacterial carbon secondary production was 22 % of the carbon in the cells lost during lysis. The amount of carbon released into the dissolved phase amounted to 15 % of the algal carbon. Compared to our results with Phaeo- Time (h) cystis pouchetti, this suggests that both the amount and bioavailability of organic material released from Fig. 2. Effect of virus infection on primary production. Dotted algae due to viral lysis, and the bacterial secondary vertical line marks the time of virus addition to the experi- production succeeding such lysis, may vary signifi- mental culture. (a) Abundance of Phaeocystis pouchetii in the cantly. infected (m) and the non-infected (control) (0) cultures. (b) Abundance of PpVOl virus (v)in the infected culture and of bacteria in infected (m)and the non-infected (control) (0) culture. Error bars are countlng error (l/dn, n = number of Effect of virus infection on CO, fixation viruses counted). (c) Primary production measured as I4C- COz fixation in the infected (m)and the non-infected (control) The changes in cell abundance in the infected and (0) culture. Error bars are standard error of mean of 3 parallels (errors less than the size of the symbols are not shown). non-infected cultures are shown in Fig. 2a. There was (d) Growth rate in the infected (a) and the non-infected (con- a lag in cell growth in the non-infected culture from trol) (0) culture estimated from primary production and a cel- 0 to 12 h which we interpret to have been caused by lular carbon content of 6 pgC cell-' Bratbak et al.: Viral lysis of Phaeoc)/stjs pouchetii

frequent sampling and disturbance of the culture. adapted to grow on organic material released from Fig. 2b shows that the abundance of free viruses these algae. Accordingly, the lag period before onset of increases rapidly between 12 and 24 h and reaches a bacterial growth after input of organic material from maximum between 24 and 30 h. With an initial virus- algal cell lysis may be expected to be short, and the to-host ratio of 10 and a lytic cycle of 12 to 18 h (Jacob- bacterial growth efficiency on this material may be sen et al. 1996) this indicates that most or all of the expected to be high. Natural bacterial communities algal cells were infected shortly after the addtion of that are not specifically selected and adapted for grow- the viruses. Virus-like particles were never observed ing on organic material released from lysed algal cells in the non-infected culture. may show a longer lag period before onset of growth The rate of CO2 fixation in both the virus infected and perhaps also a lower growth efficiency than and the non-infected culture (Fig. 2c) reflects the observed in these experiments. There is however no change in cell abundance in the 2 cultures. From reason to believe that the natural community will have increase in cell numbers we estimated the growth rate a qualitatively different response. (p, d-l) of uninfected algae (both cultures) to be 0.42 * Excretion and leaching from intact phytoplankton 0.02 d-' (mean i SE, n = l?). From the rate of CO2 fixa- cells and byproducts of zooplankton ingestion and tion (Fig. 2c) and the biomass, which was calculated digestion have been recognized as the main pathways from cell number (Fig. 2a) and a cell carbon content of of DOC from phytoplankton to bacteria (Jumars et al. 6 pgC cell-', we estimate a mean growth rate of 0.48 * 1989). Virus-induced cell lysis may nevertheless for 0.04 d-' (mean * SEM, n = 10) for the non-infected cul- short periods of time most certainly represent a much ture (Fig. 2d). The mean growth rate estimated for the larger input of DOC to the system. The difference infected culture was 0.48 * 0.07 d-' (mean * SEM, between these pathways, i.e. slow and steady release n = 7) until 12 h after infection, i.e. the same as in the of DOC versus massive cell lysis with sudden release of non-infected culture (Fig. 2d). The photosynthetic large amounts of DOC, may be of significant ecological apparatus of the infected cells does thus seem to be relevance. The former pathways may select for bacte- active during most of the lytic cycle which lasts for 12 ria growing steadily on a relatively low and constant to 18 h (Jacobsen et al. 1996). In this respect the Phaeo- input of carbon (K-strategy) while the latter may select cystis-PpV host-virus system resembles virus infection for those who have the ability to respond quickly and in Synechococcus where the photosynthetic rate in grow fast on relatively high concentrations of organic infected cultures was similar to that in non-infected material (r-strategy). Viral lysis of a blooming phyto- control cultures until near the onset of cell lysis (Suttle plankton con~munity may thus be hypothesized to & Chan 1993). This is however not a universal feature. result in a shift in the bacterial community composi- The rate of photosynthesis in Micromonas pusilla has tion. been found to be significantly reduced within 2 h after infection, but in part maintained until near the end of Acknowledgements. This work was supported by funding from The European Commission to the MAST-I11 Project the lytic cycle (Waters & Chan 1982). In ChlorelZa MEDEA, contract number MAS3-CT95-0016 and by funding infected with PBCV-1 virus, CO2 fixation is inhibited from The Research Council of Norway, project number almost immediately (van Etten et al. 1983). 11303?/120 and 1214254420. We thank Evy Foss Skjoldal, The difference between the growth rate estimated Jorunn Viken and Gunhild Bsdtker for their excellent techni- cal assistance. The EM work was done at the Laboratory for from increase in cell abundance and the growth rate Electron Microscopy, University of Bergen. estimated from CO2 fixation was insignificant. The former approach includes only production of new cells, LITERATURE CITED i.e. the particulate production, while the latter, which is based on total CO2 fixation, includes both particulate Admiraal W, Venekamp LAH (1986) Significance of tintinnid grazing during blooms of Phaeocystis pouchetli (Hapto- and dissolved production. We may thus interpret these phyceae) in Dutch coastal waters. Neth J Sea Res 20: results to indicate that the production of DOC, except 61-66 for that caused by viral lysis, was only a small fraction Azam F, Fenchel T. Field JG, Gray JS, Meyer-Reil LA, of the particulate production. Thingstad TF (1983) The ecological role of water-column microbes in the sea. Mar Ecol Prog Ser 10:257-263 Azam F, Smith DC, Steward GF, Hagstrom A (1993) Bacteria- organic matter coupling and its significance for oceanic Ecological considerations carbon cycling. Microb Ecol28:16?-179 Bratbak G, Egge JK. Heldal M (1993) Viral mortality of the The bacteria in these experiments are those which marine alga Emiliania huxleyi (Haptophyceae) and termi- have been carried over and CO-cultured with the nation of algal blooms. Mar Ecol Prog Ser 93:39-48 Bratbak G, Heldal M (1993) Total count of viruses in aquatic Phaeocystis since it was originally isolated. It may thus environments In: Kemp PF, Sherr BF, Sherr EB, Cole JJ be assumed that they are pre-selected and well (eds) Current methods in aquatic microbial ecology. Lewis 16 Aquat Microb Ecol 16: 11-16, 1998

Publishers, London, p 135-138 Lancelot C, Billen G, Sournla A, We~sseT, Colijn F, Veldhus Bratbak G, Heldal M (1995) V~ruses-the new players in the M, Davies A, Wassmann P (1987) Phaeocystis blooms and game, their ecolog~calrole and could they med~ategenetic nutrient enrichment In the continental coastal zones of the exchange by transduction? [n Joint I (ed) Molecular ecol- North Sea. Arnbio 16:38-46 ogy of aquahc microbes. NATO AS1 (Adv Sci Inst) Ser Ser Lancelot C, Wassmann P, Barth H (1994) Ecology of Phaeo- G Ecol Sci 38.249-264 cystis-dominated ecosystems J Mar Syst 5:l-4 Bratbak G, Heldal M, Norland S, Thingstad TF (1990) Viruses Loeng H, Rey F, Skloldal HR (1987) Physical and biological as partners In spring bloom microbial trophodynamics. oceanographic features across the polar front in the west- Appl Envlron Microbiol 56:1400-1405 ern Barents Sea. In: Ros J (ed) Persistent frontal systems. Bratbak G, Heldal M, Thingstad T (1994) Viruses and the 22nd Eur Mar Biol Symp 17-22 August 1987, Barcelona. microbial loop. Mlcrob Ecol 28:209-221 Sci Mar 53 Bratbak G, Levasseur M, Mlchaud S, Cantln G, Fernandez E, Maranger R, Bird DF (1995) Viral abundance in aquatlc sys- Heimdal BR, Heldal M (1995) Viral activity in relatlon to tems. a comparison between manne and fresh waters. Mar Emiliania huxleyl blooms. a possible mechanism of DMSP Ecol Prog Ser 121:217-226 release? Mar Ecol Prog Ser 128:133-142 Mathias CB, lrschner AKT, Velimrov B (1995) Seasonal Brussaard CPD, Kempers RS, Kop AJ, kegman R, Heldal M variation of vlrus abundance and vlral control of the bac- (1996) Virus like particles in a summer bloom of Emihania tenal production in a backwater system of the Danube huxleyi in the North Sea. Aquat Microb Ecol 10:105-113 River. Appl Environ Microblol 61:3734-3740 Cole JJ, Findlay F, Pace RL (1988) Bacterial production in Norland S, Fagerbakke KM, Heldal M (1995) Light element fresh and salt water ecosystems: a cross-system overview. analysis of Individual bacteria by X-ray mcroanalysis. Mar Ecol Prog Ser 43:l-10 Appl Environ Microbiol 61:1357-1362 Cottrell MT, Suttle CA (1995) Dynamics of a lytlc vlrus infect- Proctor LM, Fuhrman JA (1990) Viral mortality of manne bac- lng the photosynthetic marine picoflagellate &licromonas teria and cyanobacteria. Nature 343:60-62 pusilla. Limnol Oceanogr 40.730-739 Proctor LM, Okubo A, Fuhrman JA (1993) Calibrating estl- Ducklow HW, Carlson CA (1992) Oceanlc bacterial produc- mates of phage-induced mortality in manne bactena: tlon. Adv M~crobEcol 12:113-181 ultrastructural studles of marine bacteriophage develop- Eberlein K, Lea1 MT, Hammer KD, Hlckel W (1985) Dissolved ment from one-step growth experiments. Microb Ecol 25: organic substances during a Phaeocystis pouchetii bloom 161-182 in the German Bight (North Sea). Mar Biol 89.311-316 Reisser W (1993) Viruses and virus-like parhcles of freshwater Estep K, Nejstgaard JL, Skjoldal HR, Rey F (1990) Predation and manne eucaryotic algae-a review. Arch Protistenkd by copepods upon natural populations of Phaeocystis 143:257-265 pouch et^^ as a function of the physiological state of the Suttle CA (1994) The significance of viruses to mortality in prey. Mar Ecol Prog Ser 67:235-249 aquatic microbial commun~t~es.M~crob Ecol 28:237-243 Fuhrman J (1992) Bactenoplankton roles in cycllng of organic Suttle CA, Chan AM (1993) Marine cyanophages infecting matter: the microbial food web. In: Falkowslu PG, Wood- oceanic and coastal strains of Synechococcus: abundance, head AD (eds) Primary productivity and biogeochemical morphology, cross-infectivity and growth characteristics cycles in the sea. Plenum Press, New York, p 361-383 Mar Ecol Prog Ser 92:99-109 Fuhrman JA, Suttle CA (1993) Viruses in marine planktonic Suttle CA, Chan AM (1995) Vlruses ~nfectlng the marine systems. Oceanography 6(2).51-63 prymneslophyte Chrysochromul~na spp.: lsolatlon prelim- Gobler CJ, Hutchlns DA, Fisher NS, Cosper EM, Sanudo-Wil- inary charactenzation and natural abundance Mar Ecol helmy SA (1997) Release and bloavailability of C, N, P, Se Prog Ser 118:275-282 and Fe following viral lysis of a manne chrysophyte. Lim- van Boekel WHM, Hansen FC, R~egman R, Bak RPM (1992) no1 Oceanogr 42:1492-1504 Lysis-induced decline of a Phaeocystis spring bloom and Guillard RRL (1975) Culture of phytoplankton for feeding coupllng with the microbial foodweb Mar Ecol Prog Ser manne invertebrates. In: Smith WL, Chanley MH (eds) 81:269-276 Culture of marine invertebrate animals. Plenum Press, van Etten JL, Burbank DE, Xia Y, Meints RH (1983) Growth New York, p 29-60 cycle of a virus, PBCV-1, that infects Chlorella-like algae. Heldal M, Norland S, Tumyr 0 (1985) X-ray microanalyt~cal Virology 126:117-125 method for measurement of dry matter and elemental con- van Etten JL, Lane LC, Meints RH (1991) Viruses and vlrus- tent of individual bacteria. Appl Environ Mlcrobiol 50: llke particles of eucaryotic algae. Mlcrobiol Rev 55 1251-1257 586-620 Hennes KP, Simon M (1995) Signif~cance of bacteriophages Wassmann P (1994) Significance of sed~mentation for the ter- for controlling bactenoplankton growth in a mesotrophic mnatlon of Phaeocystis blooms J Mar Syst 5.81-100 lake. Appl Environ 1Microbiol61:333-340 Wassmann P, Vernet M, Mitchell BG, Rey F (1990) Mass sed- Huntley M, Tande K, Eilertsen HC (1987) On the trophic fate imentation of Phaeol-}.stls pouchetu in the Barents Sea of Phaeocystls pouchetii (Hariot). I1 Grazing rates of Mar Ecol Prog Ser bb:183-195 Calanus hyperboreus (Krsyer) on diatoms and different Waterbury JB, Valois FW (1993) Resistance to CO-occurnng size categones of Phaeocystis pouchetii. J Exp Mar B101 phages enables marine Synechococcus communities to Ecol 110:197-212 coexist wlth cyanophages abundant in seawater Appl Jacobsen A, Bratbak G, Heldal M (1996) Isolation and charac- Environ \llcrobiol 59.3393-3399 terisatlon of a virus infecting Phaeocystispouchel~ (Prym- Waters RE, Chan AT (1982) M~cromonas puslUa virus, the nesiophyceae). J Phycol32.923-927 virus growth cycle and associated physiolog~cal events Jumars PA, Penry DL, Baross JA, Perry MJ, Frost RW (1989) within the host cells, host range mutation. J Gen Virol 63: Closing the microbial loop: d~ssolved carbon pathwdy to 199-206 heterotrophic bactena from incomplete ingestion, diges- Zingone A (1995) The role of vlruses In the dynamics of phyto- t~onand adsorption in animals. Deep-Sea Res 36:483-495 plankton blooms G Bot Ital 129.415-423

Editorial responsibility: John Dolan, Submtted. March 31, 1998; Accepted: May 29, 1998 Villefranche-sur-Mer, Frdnce Proofs received from author[s]: September 29, 1998