University of Groningen
The role of viruses in marine phytoplankton mortality
Baudoux, Anne-Claire
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2007
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Baudoux, A-C. (2007). The role of viruses in marine phytoplankton mortality s.n.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 17-06-2017
The role of viruses in
marine phytoplankton mortality
The work presented in this thesis was carried out at the Department of
Biological Oceanography of the Royal Netherlands Institute for Sea
Research (NIOZ). Financial support was provided by the NWO-ALW and
the Treilles Foundation.
Cover design: Anne-Claire Baudoux
Printed by : Printon Trükikoda AS
RIJKSUNIVERSITEIT GRONINGEN
The role of viruses
in marine phytoplankton mortality
Proefschrift
ter verkrijging van het doctoraat in de
Wiskunde en Natuurwetenschappen
aan de Rijksuniversiteit Groningen
op gezag van de
Rector Magnificus, dr. F. Zwarts,
in het openbaar te verdedigen op
vrijdag 12 oktober 2007
om 14.45 uur
door
Anne-Claire Baudoux
geboren op 28 oktober 1977
te Amiens, Frankrijk
Promotor:
Prof. dr. G. J. Herndl
Copromotor:
Dr. C. P. D. Brussaard
Beoordelingscommissie:
Prof. dr. ir. H. J. W. de Baar
Prof. dr. G. Bratbak
Prof. dr. H. J. Laanbroek
ISBN: 978-90-367-3148-5
To Yann and to my parents, brother, and sisters.
Contents
Chapter 1
Introduction
Chapter 2
Virally induced mortality of Phaeocystis globosa during two
spring blooms in temperate coastal waters
Published in Aquatic Microbial Ecology (2006) 44: 207 – 217
31
Characterization of different viruses infecting the marine
harmful algal bloom species Phaeocystis globosa
Published in Virology (2005) 341: 80 – 90
51
Viruses as mortality agents of picophytoplankton in the deep
chlorophyll maximum layer during IRONAGES III
Accepted for publication in Limnology and Oceanography
71
Chapter 3
Chapter 4
Chapter 5
Phytoplankton losses in the North Sea during summer 2003
To be submitted to Aquatic Microbial Ecology
Chapter 6
Influence of irradiance on virus-algal host interactions
Submitted to Journal of Phycology
9
91
115
Synthesis
131
Summary / Samenvatting
137
Acknowledgments
147
Chapter 1
Introduction
Marine viruses, also referred to as virioplankton, are the most abundant and
diverse biological entities in the ocean. The ecological implication of marine viruses
goes beyond the mortality of their host. Viruses can substantially influence plankton
community structure. The cell lysis of the infected host may furthermore affect processes
of global significance, notably the cycling of nutrients. Viral pathogens are found to
infect all the major classes of phytoplankton. Algal viruses appear to be involved in the
disintegration of algal blooms, but only a limited number of studies has investigated
their role in population dynamics of non-bloom forming phytoplankton. Therefore, the
prevalence of virally mediated mortality in phytoplankton still remains elusive, and the
significance of viral lysis related to other phytoplankton losses is essentially unknown.
The purpose of this thesis is to address these issues in order to better understand the
ecological significance of algal viruses, specifically examining the impact of these
pathogens in marine environments with contrasting trophic status (eutrophic vs.
oligotrophic).
This introductory chapter covers different aspects of viral ecology, with special
emphasis on viruses infecting phytoplankton and the interaction with their host. The
chapters that follow describe the results from field and laboratory studies investigating
the ecological role of viral infection of phytoplankton as compared to other sources of
phytoplankton losses (e.g. microzooplankton grazing) in different marine systems.
1. Viruses in marine ecosystems
In essence viruses are simple. They are defined as small particles composed of
nucleic acids (either DNA or RNA) embedded in a protein shell (named caspid) that may
be surrounded by an envelope. Viruses do not respire, move, or grow. They do not have
inherent metabolism, therefore, they utilize the cellular machinery of a suitable host to
replicate. A given virus infects a restricted range of hosts. Most viruses described to date
are species-specific, i.e., they infect a single host species and sometimes even a single
9
Chapter 1
Introduction
strain within a species. As they do not move, viruses depend on passive diffusion to
contact a suitable host. Consequently the encounter rate between a virus and a host is
directly affected by their relative abundance.
1.1. Global abundance and diversity of marine viruses
In the years following the discovery of high virus abundance (typically 106 to
10 viruses mL-1), field studies have emphasized that marine viruses are also a dynamic
component of the planktonic community (Bergh et al. 1989). For example, virioplankton
abundance varies with depth (Hara et al. 1996, Culley & Welschmeyer 2002), along
trophic gradients (Weinbauer et al. 1993, Noble & Fuhrman 2000), and during the course
of phytoplankton blooming events (Bratbak et al. 1990, Castberg et al. 2001, Brussaard
et al. 2004b). It is now well accepted that viruses are present wherever life is found, and
current estimates of ≈ 1030 viruses in the world’s ocean indicate that they are the most
abundant marine biological entities (Suttle 2005).
The observation of natural virioplankton assemblages using transmission
electron microscopy (TEM) has revealed that, beyond their high abundance, viruses vary
in size and shape (Flint et al. 2000); most of them are polyhedral and range between 20
and 200 nm. The isolation and the characterization of viruses infecting specific microbial
host cultures, including prokaryotic and eukaryotic microbes, consistently showed that at
least one, and usually multiple, viruses can infect a single host species (see reviews by
Brussaard 2004, Weinbauer 2004). It is, therefore, sensible to assume that the richness of
viruses is at least as high as that of cellular life forms (Weinbauer 2004).
The development of culture-independent methodologies has considerably
advanced our understanding of global virus diversity. The use of pulsed field gel
electrophoresis (PGFE), that discriminates viruses based on their genome size, showed
that natural viral assemblage typically comprise 7 to 35 distinct viral genome size classes
ranging between 12 and 560 kilobases (Auguet et al. 2005 and reference therein). Using
this approach, pronounced variations in virus community structure were detected in
response to phytoplankton bloom formation (Castberg et al. 2001, Larsen et al. 2001),
water column stratification (Wommack et al. 1999), or salinity gradient (Sandaa et al.
2003). Overall, PFGE results confirm that viral populations are spatially and temporally
dynamic as previously predicted from changes in viral abundance.
The most striking evidence of high viral richness in the ocean was reported by
metagenomic analyses of uncultured marine viral communities (Breitbart et al. 2002,
Angly et al. 2006). The recent sequencing of shotgun libraries of 168 viral assemblages
collected from four major oceanic regions revealed that several hundred thousand
distinct viral species were dispersed in these waters (Angly et al. 2006). Most of these
viral genotypes were not similar to previously reported sequences, indicating that much
of the viral diversity is actually uncharacterized. An emerging view of viral diversity
contends that the vast majority of viral species is widespread and that the local
environmental conditions select for certain viral type through selective pressure.
8
10
Chapter 1
Introduction
1.2. Production vs. decay of marine viruses
In theory, all living marine organisms from ‘bacteria to whale’ are likely to be
infected by at least one virus. However, the host-virus encounter is an abundancedependent process; hence the majority of viruses probably infect the organism they most
frequently encounter, i.e., the bacteria and phytoplankton.
Viruses are produced by their host through four types of replication. The lytic
infection results in the release of virus progeny upon host lysis. The number of viruses
produced per infected cells is called the burst size. The magnitude of viral burst size has
important ecological implications as it directly influences the viral abundance, hence the
propagation of viral infection.
Viruses can also interact with their host through lysogeny (or latency) where the
viral genome is incorporated into the host cell genome (termed as prophage) and
propagates along with host replication until an inducing event triggers the lytic pathway.
The incidence of lytic or lysogenic replications may relate to ecological strategies.
Lysogenic replication may prevail over lytic infection when successful host-virus
encounter rate is too low to sustain lytic replication (Lenski 1988, Weinbauer 2004). The
importance and mechanisms underlying lytic vs. lysogenic replication are unclear and
therefore still need further investigation. So far lysogeny has only been reported for
prokaryotic microbes.
Although the lytic and lysogenic infections are the most investigated forms of
replication in marine environments, two other types of replication are described. These
include the chronic infection where viruses are released by budding or extrusion without
killing the host and the pseudolysogeny infection whereby the virus either enter a
dormant intracellular phase or proceed with lytic infection. This type of infection
resembles a true lysogenic infection except that the viral genome does not integrate into
host genome (as cited in Williamson et al. 2001).
The term ‘viral decay’ includes the reduction in viral abundance and infectivity.
Many different biotic and abiotic factors are involved in the loss of virus in the ocean.
Among these, solar radiations, particularly ultraviolet B radiations (UV-B), are
considered a major source of loss as they damage viral DNA (Suttle 2000, Wilhelm et al.
2002, Jacquet & Bratbak 2003). The effect of damaging sunlight can still be significant
at 10 m depth, as reported in the Gulf of Mexico offshore waters (Wilhelm et al. 2002).
Viruses may also be inactivated, at least temporarily, by adsorption to host cells, high
molecular weight dissolved organic matter, and transparent exopolymeric particles
(Noble & Fuhrman 1997, Brussaard et al. 2005b, Ruardij et al. 2005). Viral losses have
been observed due to grazing by nanoflagellates (Suttle & Chen 1992, Gonzalez & Suttle
1993) and adsorption of viruses to particles that sink out of the photic zone (Proctor &
Fuhrman 1991). The latter observation is consistent with the report of high virus
numbers in the sediments (Danovaro et al. 2001, Lawrence et al. 2001).
11
Chapter 1
Introduction
2. Ecological implications of marine viruses
2.1. The influence of viruses on the community composition
The majority of the marine viruses has a narrow host specificity, implying that
only a small subset of the community is infected by a given virus. The current consensus
is that virus will preferentially infect the most abundant host species (because of higher
encounter rate) and by “killing the winner” viruses maintain the coexistence of
competing species (Thingstad 2000). The incidence of changes in viral and microbial
community structure following lysis events strongly support the role of viruses as a
driving force for the interspecies competition and succession (Castberg et al. 2001,
Larsen et al. 2001, Brussaard et al. 2005b). Lately, several field studies demonstrated
that many marine viruses may even show intraspecies specificity, suggesting that viruses
may also influence the clonal composition of the host species (Tarutani et al. 2000,
Tomaru et al. 2004b, Mühling et al. 2005).
2.2. The effect of viruses on the biogeochemical cycles
Over the last two decades, it became evident that microbial processes largely
drive the cycle of matter and energy in the ocean. Whilst phytoplankton constitute the
base of the classical (grazing) food web, heterotrophic prokaryotes recycle dissolved
organic matter (DOM) to inorganic nutrients and bacterial biomass through the microbial
loop (Fig. 1). These microbes can be eaten by small predators, eventually leading back to
the classical (grazing) food web. In the original hypothesis of the ‘microbial loop’, the
primary source of DOM was assumed to derive from phytoplankton exudates and sloppy
feeding by zooplankton (Azam et al. 1983). Through the lysis of their host, viruses also
influence the cycling of DOM (Wilhelm & Suttle 1999). Prokaryotic and eukaryotic
viruses efficiently convert the particulate organic matter (i.e., living biomass) to DOM
that can be utilized by bacteria (Brussaard et al. 1996b, Middelboe et al. 2003). This
“viral shunt” results in diverting the flow of matter and nutrient away from the higher
trophic levels and, in turn, forces the food web towards a more regenerative system.
The incorporation of “viral module” in simple theoretical models systems
demonstrated that viral lysis enhanced bacterial respiration and production (Fuhrman
1999, Wilhelm & Suttle 1999) and reduced protist production (Fuhrman 1999). The first
mathematical ecosystem model based on empirical data confirmed that virally mediated
mortality of the bloom-forming algal species Phaeocystis globosa was an essential
regulating factor for the nutrient cycling (Ruardij et al. 2005). Experimental studies also
supported these models’ predictions. Viral mediated release of DOM can constitute a
significant supply of major nutrients (C, N, P) and trace nutrients (e.g. Fe) for other
12
Chapter 1
Introduction
photosynthetic and heterotrophic microorganisms (Middelboe et al. 1996, Göbler et al.
1997, Middelboe et al. 2003, Poorvin et al. 2004). Viral mediated nutrient cycling was
furthermore shown to influence bacteria and phytoplankton species composition and
succession (Göbler et al. 1997, Brussaard et al. 2005b).
Larger gazers
Bacteria
Debris,
exudates
Viruses
Viruses
Small grazers
(HNF, microzooplankton)
Inorganic
nutrients
DOM
Debris, exudates
Debris,
exudates,
sloppy
feeding
Viruses
Phytoplankton
Figure 1. Simplified diagram of the microbial loop. Through cell lysis of hosts, viruses divert
living biomass away from the higher trophic levels of the food web (microzooplankton,
heterotrophic nanoflagellates (HNF), larger grazers). Instead, living biomass is effectively
converted to dissolved organic matter (DOM), available for heterotrophic bacteria, hence forcing
the food web towards a more regenerative pathway. Black arrows indicate the flow of organic
matter and grey arrows represent the flow of inorganic nutrients.
Another possible effect of viruses on processes of global significance is to
accelerate the production of dimethylsulfide (DMS). The DMS is a biogas that
influences clouds formation, hence the global climate (Charlson et al. 1987). Many
phytoplankton species produce dimethylsulfoniopropionate (DMSP) that may be cleaved
into DMS and acrylic acid (AA) by the algal lyases and/or by the lyases of other
organisms. Laboratory studies have demonstrated that viral lysis of Micromonas pusilla,
Phaeocystis pouchetii, and Emiliania huxleyi was accompanied by a build up of DMSP
in the media (Hill et al. 1998, Malin et al. 1998, Wilson et al. 2002). Therefore, viral
lysis of phytoplankton may be an important source of DMSP in the environment.
13
Chapter 1
Introduction
3. Viruses and phytoplankton hosts
Soon after the discovery of high abundance of marine viruses, Suttle et al.
(1990, 1992) showed that adding virus concentrates to natural field samples could result
in decreased primary production. Complementing these studies, the observation of viral
infected algal cells suggested that viruses can account for significant phytoplankton
losses, particularly during blooming events (Bratbak et al. 1993, Nagasaki et al. 1994,
Brussaard et al. 1996b). With phytoplankton forming the base of the marine pelagic food
web and the awareness of the ecological and socio-economical consequences of algal
blooms (e.g., fisheries and tourisms activities), the ecology of algal viruses and their
contribution to phytoplankton mortality has gained considerable interest.
3.1 Taxonomy and phylogeny of algal viruses
Marine phytoplankton communities include a prokaryotic (cyanobacteria) and a
eukaryotic component. Currently, host specific viruses are reported for both groups of
phytoplankton. Viruses that infect cyanobacteria, referred to as cyanophages, were
reported in unicellular (Proctor & Fuhrman 1990) and colonial cyanobacteria (Ohki
1999, Hewson et al. 2004). The ecologically important marine Synechococcus sp. and
Prochlorococcus sp. are, by far, the most investigated cyanobacterial hosts (for reviews
see Suttle 2000, Mann 2003). All known cyanophages have tails, present double
stranded DNA, and belong to three families that also infect heterotrophic bacteria, the
Myoviridae, the Siphoviridae, and the Podoviridae. Besides morphological differences,
cyanophages belonging to these families also have variable “life styles”. For example,
the Myoviridae are typically lytic and have a broader host range than the other tailed
cyanophages. Conversely, the Podoviridae present the narrowest range of host. The
replication of Siphoviridae differs from other tailed phages as they can interact with their
host through lysogeny, and may thus propagate along with the host replication (see also
section 1.3).
Viruses have been isolated for the major existing classes of eukaryotic
phytoplankton. Unlike cyanophages, all known eukaryotic algal viruses propagate
through a lytic pathway. Until recently, most of these viruses were consistently assigned
to the family of large double stranded (ds)DNA viruses, the Phycodnaviridae (Wilson et
al. 2005b). Molecular based analysis using the highly conserved DNA polymerase (pol)
gene allowed distinguishing six genera within this family. With the increasing number of
viruses characterized, it became evident that viruses infecting phytoplankton are
extremely diverse with representatives in many more viral families than the
Phycodnaviridae. For example, picorna-like positive sense single stranded (ss)RNA
viruses were found to infect the diatom Rhizosolenia setigera (Nagasaki et al. 2004) and
the two toxic harmful algal bloom (HAB) species Heterocaspa circularisquama
14
Chapter 1
Introduction
(Tomaru et al. 2004a) and Heterosigma akashiwo (Tai et al. 2003). The recently
completed genomic sequence of the ssRNA H. akashiwo virus led to the creation of the
new, distinct viral family, the Marnaviridae (Lang et al. 2004). Previously unknown
nuclear inclusion viruses have also been reported to infect H. akashiwo (Lawrence et al.
2001) as well as the diatom Chaetoceros cf gracilis (Bettarel et al. 2005). Another
example of novel virus is the dsRNA virus that infects the cosmopolitan Micromonas
pusilla assigned to the Reoviridae family (Brussaard et al. 2004a, Attaoui et al. 2006).
These examples emphasize that many different viruses can infect the same algal species.
We are only starting to uncover the diversity of marine algal viruses. Many more viruses
need to be isolated and characterized in order to better evaluate algal virus richness.
3.2. Abundance and diversity of marine algal viruses
In the field, viral titer determination using the most probable number (MPN)
culture based method has been proven very useful in the study of algal virus ecology.
These studies indicated that infective algal viruses can be highly abundant (up to >105
mL-1), dynamic, and exhibit a high level of diversity (Waterbury & Valois 1993, Suttle
& Chan 1994, Cottrell & Suttle 1995a, Tomaru et al. 2004b).
The use of molecular tools allowed discriminating between different virus
isolates infecting the same species. For example, Cottrel and Suttle (1995b)
distinguished different Micromonas pusilla virus isolates using DNA hybridization. The
diversity of cyanophages (namely, myocyanophages) was estimated by the sequence
analysis of the gene encoding a structural protein g20 (Füller et al. 1998, Zhong et al.
2002, Mühling et al. 2005). Other genetic markers, such as the gene fragment of the
putative major caspid protein of Emiliania huxleyi viruses also revealed a high molecular
diversity among E. huxleyi viruses (Schroeder et al. 2002).
3.3 Algal viruses and phytoplankton mortality
The significance of algal viruses in terms of abundance, dynamics, and diversity
indicates a significant role of viral lysis in phytoplankton ecology. However, to fully
understand the role of viruses, it is essential to determine the extent of mortality they
impose on their host. The studies that have addressed this issue were mainly conducted
during conditions of high host cell abundance such as during phytoplankton blooms. In
theory, when the host cell abundance is high, the probability of collision between a host
and a virus increases, hence viruses may propagate rapidly through the host population.
This may result in bloom collapse if the viral lysis rates exceed the specific host growth
rates. This type of interactions is referred to as a control by “reduction” (Brussaard
2004). Several reports confirmed this theory and demonstrated that viruses are
profoundly involved in the disintegration of algal blooms. For example, high proportions
(10 - 50%) of algal cells were visibly (using TEM) infected at the end of a bloom of
15
Chapter 1
Introduction
Aureococcus anophagefferens (Sieburth et al. 1988, Gastrich et al. 2004), Heterosigma
akashiwo (Nagasaki et al. 1994), and Emiliania huxleyi (Bratbak et al. 1993, 1996,
Brussaard et al. 1996b). Other approaches determining cell lysis from the number of
putative algal viruses produced divided by an empirical viral burst size indicated that
viruses accounted for substantial mortality (7 - 100%) during the bloom of Phaeocystis
globosa (Brussaard et al. 2005a, Ruardij et al. 2005) and E. huxleyi (Jacquet et al. 2002).
While several studies have examined the role of viruses in controlling algal
bloom dynamics, fewer studies have investigated the potential role of viruses in
regulating non-blooming phytoplankton populations. Examinations of the picoeukaryote
Micromonas pusilla and its specific viruses indicated a turnover of host standing stock
between 2 and 25% d-1 in various marine systems (Cottrell & Suttle 1995a, Evans et al.
2003). Studies conducted on the cyanobacterium Synechococcus reported that virally
induced mortality daily removed <1 to 8% of the host population (Waterbury & Valois
1993, Suttle & Chan 1994, Garza & Suttle 1998). These results suggest a stable hostvirus coexistence, where the viruses seem to maintain host population size to nonblooming level rather than causing a rapid decline in host abundance. This type of
regulation is referred to as a “preventive” viral control (Brussaard 2004).
Overall, virally mediated mortality can occur at significant rates in
phytoplankton populations. However, our understanding of the global significance of
viral lysis is far from complete because non-blooming phytoplankton and more
generally, phytoplankton in oligotrophic environments have been inadequately sampled
as compared to bloom forming species, typically found in eutrophic (coastal) waters.
3.4. Potential factors regulating virally mediated mortality of
phytoplankton
The above referred field studies indicate that viral lysis can be responsible for
significant mortality in phytoplankton. Different factors can, however, regulate the
dynamics of virally mediated mortality. These regulatory parameters include
phytoplankton host abundance, morphology, physiology and their potential to develop
defense mechanisms.
As viral infection is a stochastic process, the rate of encounter depends on the
hosts and virus abundance and also on other morphological characteristics such as
particle size and motility (Murray & Jackson 1992). At a given virus concentration,
larger hosts will be more readily intercepted by a virus than the smaller counterparts.
Host cell motility enhances transport rates which, in turn, increase the probability of
collision with a given virus. Other host morphological characteristics can influence viral
infection rate. For example, field and laboratory evidence suggested that non-coccolithbearing Emiliania huxleyi are more readily infected than the lithed cells (Brussaard et al.
1996b, Jacquet et al. 2002). This is, however, not (yet) confirmed by controlled
experiments. Furthermore, a mesocosm study showed that Phaeocystis globosa cells
embedded in a colonial matrix tend to escape viral infection (Brussaard et al. 2005a,
16
Chapter 1
Introduction
Ruardij et al. 2005). Interestingly, this can be explained by the larger colonial size
(Ruardij et al. 2005).
The most efficient defense of phytoplankton against viral infection is to be
resistant. The incidence of resistant host strains has been reported for algal viruses in
culture (Thyrhaug et al. 2003) as well as in the field (Waterbury & Valois 1993). Theory
based on bacterial host-phages interactions suggests that resistance has a physiological
cost for the host cells, resistant cells may have a competitive disadvantage against
susceptible hosts (Levin et al. 1977). So far, the importance and the mechanisms
underlying resistance of phytoplankton against viruses remain largely unclear.
A potential phytoplankton chemical defense against viruses was recently
suggested by Evans et al. (2006). These authors related the negative effect of acrylic acid
(AA) and dimethylsulfide (DMS) on the titers of Emiliania huxleyi virus to the inability
to isolate viruses infecting E. huxleyi strains with high lyase activity (i.e., capable of
efficient conversion of dimethylsulfoniopropionate (DMSP) into the AA and DMS). It
was suggested that the cleavage of DMSP in DMS and AA during cell lysis of E. huxleyi
may reduce the titers of E. huxleyi viruses, and therefore decrease the probability of
infection of further cells. These observations led to argue that the DMSP system in
phytoplankton may operate as a chemical defense against viral infection. This study
supported the hypothesis that virucidal compounds can be produced alongside viruses
during viral infection, and, in turn, can reduce infection rates of other algal cells
(Thyrhaug et al. 2003). Another example of potential host defense strategy includes the
enhanced sinking rates of Heterosigma akashiwo cells when infected by a virus
(Lawrence & Suttle 2004). Viral infection may result in cells rapidly sink out of the
euphotic zone, which, in turn, may prevent viral infection of conspecifics.
As obligate parasites, viruses depend on the host cellular machinery to
propagate. Several studies have shown that the algal host growth stage may significantly
influence the lytic viral growth cycle. Reduction in viral burst size (Van Etten et al.
1991, Bratbak et al. 1998) and even prevention of viral infection were observed
(Nagasaki & Yamaguchi 1998) during the algal host stationary growth phase. The algal
host cell cycle stage may also influence the production of algal viruses (Thyrhaug et al.
2002). Viral infection of Pyramimonas orientalis at the onset of the light period led to a
higher viral production than when infected at the beginning of the dark period.
Conversely, Phaeocystis pouchetii infection was independent of the host cell cycle.
Different environmental variables known to influence phytoplankton growth
rates (i.e. light, nutrient, and temperature) may, furthermore, affect the viral growth
cycle. For instance, darkness could prevent viral infection of different prokaryotic and
eukaryotic algal hosts (MacKenzie & Haselkorn 1972, Allen & Hutchinson 1976,
Waters & Chan 1982). Temperature may alter the susceptibility of host species to virus
as shown for H. akashiwo (Nagasaki & Yamagushi 1998). Nutrient limitations were
found to have variable effects; phosphate depletion resulted in a reduction of the burst
size of P. pouchetii and Emiliania huxleyi viruses (Bratbak et al. 1993, 1998) or delayed
the cell lysis in Synechococcus (Wilson et al. 1996). Under nitrogen depletion, the
production of E. huxleyi viruses was delayed (Jacquet et al. 2002) and a reduction in the
17
Chapter 1
Introduction
viral burst size was observed for P. pouchetii (Bratbak et al. 1998).
In the ocean, phytoplankton cells experience strong fluctuations in natural
resources (e.g. light, nutrient, temperature). For instance, light and nutrient levels can
vary during phytoplankton bloom and across the water column of stratified systems (e.g.
open ocean). These variations may thus control the impact of viruses on the host
population. Furthermore, the contrasted nutrient conditions found in oligotrophic vs.
eutrophic environments may underlie differential virally mediated mortality of
phytoplankton in these respective environments.
4. Viral lysis and other sources of phytoplankton losses
4.1. Viral lysis and other sources of cell death by lysis
Algal cell lysis rates can be high and dynamic in marine environments
(Brussaard et al. 1995, 1996a, Agusti et al. 1998). Estimates up to 0.3 d-1 have been
reported not only during the termination of algal bloom (Brussaard et al. 1995, 1996a)
but also in oligotrophic marine environments (Agusti 1998). Algal cell lysis has
important implications on the trophic dynamics as the cell content released in
surrounding waters upon lysis provides DOM, potentially available for heterotrophic
bacteria. Several field studies indicated that DOM released upon algal cell lysis could be
sufficient to fulfill most of the bacterial carbon demand (Brussaard et al. 1996b, 2005b).
Viruses are considered important agents killing phytoplankton. Although
several studies have demonstrated that viruses can impose substantial mortality on their
host (see review Brussaard 2004 and section 3.3), the quantitative significance of viral
lysis in the ocean is not clear. One reason for this is that rates of virally mediated
mortality are assessed using different approaches; therefore results from these studies are
not necessarily comparable. More importantly, all known methodologies determining
virally mediated mortality rely on differing assumptions and conversion factors (Table
1). Very few studies have, thus far, determined the contribution of viral lysis to total
algal cell lysis. One recent mesocosm study has shown that viral lysis comprised 30 100% of the total lysis occurring during the bloom of Phaeocystis globosa (Brussaard et
al. 2005a).
In addition to viruses, several other mechanisms responsible for cell lysis are
currently described. For example, other pathogens (e.g., bacteria, fungi) are reported to
kill phytoplankton (Fukami et al. 1992, Mayali & Azam 2004). Another example
includes allelopathic interactions between phytoplankton species. In this type of
interaction, the production of a metabolite by a phytoplankton species has an inhibitory
effect on the growth or physiological function of another phytoplankton species that may
result in cell lysis (Vardi et al. 2002, Legrand et al. 2003).
18
19
- Rapid, inexpensive
- Direct observation of changes in
viral abundance/loss infectivity
- Rapid, inexpensive
- Direct changes in viral
abundance
- Rapid, inexpensive
- The only method excluding the
use of conversion factors
- Provides simultaneously viral
lysis and grazing rates
- Discriminates different algal
groups when combined with FCM
- All cells equally sensitive to viral infection
- All virus-host contact result in infection
- All cell lyse after viral infection
- Diffusion, adsorption rate, burst size, and
host cell size are constant
- Virus and host of interest can be
discriminated (Most Probable Number, MPN,
flow cytometry, FCM)
- Viral decay equals viral production
- Burst size is constant
- Virus of interest need be discriminated
(MPN, FCM)
- All virus produced from infected cells
- Virus of interest can be discriminated from
other virus (MPN, FCM)
- Burst size constant or within a stated range
- Algal growth rates unaffected by dilution
level and diluent
- Phytoplankton cell lysis starts after 12 h
- Losses are proportional to the dilution effect
of the loss agent (virus and grazers)
- No selective grazing on infected/
noninfected cells
Contact rates
Viral decay
Viral
production
Modified
dilution
method
19
Advantages
- No incubation required
Assumption
- All cells containing viruses will lyse
- The eclipse time (time from infection
to mature virus appearance) is constant
- Latent period equals host generation
- Host infection occurs continuously
Methodology
Frequency of
infected cells
using
transmission
electron
microscopy
(TEM)
- Only measures newly infected cells
- Initial phytoplankton concentration must be
high enough to allow up to 5-fold dilution
- A 24h incubation required
- Use of theoretical burst size
- MPN underestimates virus abundance
- FCM cluster may include other virus
- Viral loss not taken into account
- Use of theoretical burst size
- Relation between viral decay and
production is disputable
- Lytic and lysogenic virus not distinguished
- MPN underestimates virus abundance
- FCM cluster may include other virus
- Heavily dependent on theoretical
conversion factors (diffusion, adsorption
coefficient, cell size, burst size)
- MPN may underestimate virus abundance
- FCM cluster may include other viruses
- Cross infection is not taken into account
- Lytic and lysogenic virus not distinguished
Disadvantages
- Heavily dependent on theoretical
conversion factors (eclipse time, relationship
between latent period and host generation)
that are disputable
- Host of interest may be difficult to
discriminate in natural sample
- Selective losses of infected cells may occur
during sample preparation
- Time consuming
Evans et al. 2003
This thesis
Bratbak et al. 1993
Jacquet et al. 2002
Brussaard et al.
2005a
Suttle and Chan 1994
Cottrel and Suttle
1995a
Garza and Suttle
1998
Bongiorni et al. 2005
Suttle and Chan 1994
Garza and Suttle
1998
Source
Sieburth et al. 1988
Proctor et al. 1993
Nagasaki et al. 1994
Bratbak et al. 1993,
1996
Brussaard et al.
1996b
Binder 1999
Gastrich et al. 2004
Table 1. Summary of the assumptions, advantages, and disadvantages of the methods used to determine virally mediated mortality of
phytoplankton (adapted from Winget et al. 2005).
Chapter 1
Introduction
Non-pathogenic forms of algal cell lysis are also reported. For example, the
diatom Ditylum brightwellii was shown to experience a type of ‘intrinsic mortality’
under nitrogen controlled conditions using chemostat cultures (Brussaard et al. 1997).
Recently, another form of autocatalyzed cell death was shown to share similarities with
the programmed cell death (PCD) observed in higher plants and metazoans (Berges &
Falkowski 1998, Vardi et al. 1999, Berman-Frank et al. 2004). The PCD, unlike “natural
cell death” or “necrosis” refers to an active, genetically controlled degenerative process,
which involves series of apoptotic features such as morphological changes (e.g. cell
shrinkage, vacuolization) and complex biochemical events (e.g. activation of PCD
markers like caspases). The PCD ultimately leads to cell lysis. Laboratory studies
suggest that a wide range of phytoplankton is programmed to die in response to adverse
environmental conditions (see review by Franklin et al. 2006). The PCD pathway in
phytoplankton was found to operate under environmental stresses such as intense light
(Berman-Frank et al. 2004), darkness (Berges & Falkowski 1998), nutrient depletion
(Berman-Frank et al. 2004), CO2 limitation and oxidative stress (Vardi et al. 1999).
Some apoptotic features, possibly directing to PCD, have also been detected upon viral
infection (Berges and Brussaard unpubl. data, Lawrence et al. 2001). The complete
genome sequence of a virus infecting Emiliania huxleyi has recently revealed the
presence of genes encoding the biosynthesis of ceramide, which is known to suppress
cell growth and is an intracellular signal for apoptosis (Wilson et al. 2005a).
4.2. Viral lysis versus microzooplankton grazing
Prior to the recognition of algal cell lysis as an important loss factor,
phytoplankton cells were essentially treated as immortal unless they were preyed upon
by zooplankton or lost by sedimentation through the water column. Cell lysis,
sedimentation, and predation by zooplankton may thus, separately or in concert,
influence phytoplankton community structure.
Whether a phytoplankton cell sinks, is preyed upon, or undergoes lysis has
more implications than the structuring of phytoplankton community. As nicely
formulated by Kirchman (1999), “how phytoplankton die largely determines how other
marine organisms live”. The phytoplankton biomass that sinks is lost from the surface to
the benthic food web (Smetack 1985). In contrast, zooplankton grazing channels
phytoplankton biomass to the higher trophic levels whereas the release of cell
constituents mediated by lysis directly affects the standing stock of dissolved organic
carbon, forcing the food web towards a more regenerative pathway (Wilhelm & Suttle
1999, Brussaard et al. 2005b, Fig. 1). The quantification of cell lysis in relation to
sinking and grazing rates is, therefore, essential for an optimal understanding of the flow
of matter and energy in marine systems.
There are differential controls of phytoplankton in oligotrophic vs. eutrophic
environments. In oligotrophic waters (e.g. open ocean, surface coastal waters during
summer), the import rate of the controlling nutrient is low and the regeneration of this
limiting nutrient is critical to sustain high productivity. Small-sized picophytoplankton
20
Chapter 1
Introduction
dominate the autotrophic community due to their competitive growth characteristics.
Considering their micrometer size range, picophytoplankton cells are not prone to
sedimentation (Raven 1998). Instead, the rapidly growing small-sized predators
(heterotrophic nanoflagellates and microzooplankton) are thought to largely control this
phytoplankton biomass (Riegman et al. 1993, Kuipers & Witte 2000). As discussed
above (section 3.3), there is also evidence of significant viral lysis of smaller-sized
picophytoplankton. The relative importance of viral lysis as compared to grazing is,
however, largely unknown in these environments.
In eutrophic waters the import rate of the controlling nutrient is higher and
larger-sized phytoplankton can develop as they escape grazing by microzooplankton.
The larger-sized phytoplankton biomass may not be immediately controlled by larger
grazers as the development and the generation time of larger grazers is relatively long.
Size-selective escape of grazing or non-edible phytoplankton may form algal blooms.
Mass sedimentation can be involved in the disintegration of some of these blooms
(Smetack 1985, Brussaard et al. 1995). Cell lysis was also found to be responsible for
bloom termination (Brussaard et al. 1996a). As emphasized in this introductory chapter,
one possible agent causing the cell lysis is viral infection. Episodes of light and/or
nutrient limitations, typically occurring during algal blooms, may regulate the impact of
virus on host abundance. However, the extent to which viral lysis varies and the
underlying regulatory mechanisms remain poorly documented.
Summarizing the above, the understanding of the quantitative and qualitative
importance of phytoplankton viral lysis in the oligotrophic vs. eutrophic marine waters is
not clear. The present thesis aims to elucidate the role of algal viruses in these
contrasting environments. Therefore, virally induced mortality rates of different
phytoplankton groups were determined and related to microzooplankton grazing. In
order to compare results from different field studies, a single method assessing viral lysis
was used, namely an optimized version of the recently modified dilution method (Evans
et al. 2003). In its original form, the dilution method is routinely used to estimate grazing
on phytoplankton (Landry & Hassett 1982). The modified assay also includes virally
mediated losses of phytoplankton. Although this dilution method has some restrictions, it
has the benefits to exclude the use of conversion factors (i.e., it provides direct viral lysis
rates), to minimize the handling of sample, and it can be applied to the different
coexisting phytoplankton populations. In addition to these field studies, laboratory
experiment aimed to characterize specific virus-host model systems and to study the
virus-host interactions in relation to environmental relevant conditions.
21
Chapter 1
Introduction
5. This thesis
The aim of this thesis is to advance our knowledge on the ecological
significance of algal viruses for marine phytoplankton mortality. More specifically, this
research used a combination of field and laboratory approaches to explore three main
issues:
1.
2.
3.
The extent of virally induced lysis in phytoplankton mortality in marine
environments with contrasting trophic status (oligotrophic vs. eutrophic)
The comparison of viral lysis rates with other algal loss factors (mainly
microzooplankton grazing)
Possible factors regulating algal host-virus interactions.
Chapter 2 describes the significance of viruses during the course of an algal
bloom that developed in the eutrophic southern North Sea. The prymnesiophyte
Phaeocystis globosa is well known for its complex polymorphic life cycle (including
colonies and single cells) and its potential to develop dense blooms in temperate coastal
waters. The termination of P. globosa blooms is typically sudden. Earlier studies have
demonstrated that cell lysis can account for up to 75% of the bloom decline (Brussaard
et al. 1995). Recently, a mesocosm experiment showed that viruses could be a primary
cause of cell lysis for P. globosa (Brussaard et al. 2005a, Ruardij et al. 2005). However,
virally induced mortality of P. globosa has never been determined in the field. For
completeness and to allow the study of inter-annual variability, the significance of viral
lysis during 2 consecutive P. globosa blooms was investigated. To complement viral
lysis estimates of P. globosa, we monitored the total abundance of putative P. globosa
viruses (PgV) as well as the abundance of infective PgV.
Chapter 3 adds to the previous study by providing insight into the phenotypic
diversity of PgV. An earlier phylogenetic analysis of 24 PgV isolated from the Southern
North Sea revealed a close genetic relatedness among these isolates as they formed a
tight monophyletic group within the family of the Phycodnaviridae (Brussaard et al.
2004b). In order to address biodiversity issues, it was thus very challenging to explore
the phenotypic diversity among these isolates. Therefore, twelve of these isolates were
further characterized. This study includes a morphological (particle size and shape) and
molecular (genome size, major structural protein composition) characterization as well
as the investigation of ecologically relevant characteristics such as the length of the lytic
cycle, burst size, the range of host infected by these isolates, and their sensitivity to
temperature.
Chapters 4 and 5 investigate the role of viruses as mortality agents for
different picophytoplankton groups in oligotrophic waters. Chapter 4 describes a study
conducted in the northeastern subtropical Atlantic Ocean with a permanent oligotrophic
22
Chapter 1
Introduction
status whereas the study described in Chapter 5 was executed in the North Sea under
seasonal (summer) oligotrophic conditions. Sharp gradients of light, temperature, and
nutrient level are typically encountered across the water column in these environments.
At the bottom of the euphotic zone, an accumulation of phytoplankton, referred to as a
deep chlorophyll maximum (DCM), marks the transition between the nutrient-depleted
lit surface waters and the nutrient-enriched, light-depleted waters below the thermocline.
Different algal virus communities were observed in the surface and DCM waters (Zhong
et al. 2002). We investigated the role of algal viruses in the DCM (Chapter 4) and in the
surface waters (Chapter 5) and related rates of viral lysis to microzooplankton grazing
for 4 groups of picophytoplankton (including eukaryotes and prokaryotes).
Chapter 6 addresses the influence irradiance can have on virus-algal host
interactions. Given the changes in light conditions that a phytoplankton cell may
experience with depth or with time, investigating the effect of different irradiance on
host-virus interactions was timely. Chapter 6 describes a laboratory experiment testing
the effect of different light levels, including darkness, on two marine phytoplankton of
ecological relevance, namely the bloom former Phaeocystis globosa thriving in
eutrophic waters, and the picophytoplankter Micromonas pusilla ubiquitously distributed
including in oligotrophic environments.
Literature cited
Agusti S, Satta MP, Mura MP, Benavent E (1998) Dissolved esterase activity as a tracer
of phytoplankton lysis: Evidence of high phytoplankton lysis rates in the
northwestern Mediterranean. Limnol Oceanogr 43:1836-1849
Allen MM, Hutchinson F (1976) Effect of some environmental factors in cyanophage
AS-1 development in Anacystis nidulans. Arch Virol 110:55-60
Angly FE, Felts B, Breitbart M, Salamon P, Edwards RA, Carlson C, Chan AM, Haynes
M, Kelley S, Liu H, Mahaffy JM, Mueller JE, Nulton J, Olson R, Parsons R,
Rayhawk S, Suttle CA, Rohwer F (2006) The marine viromes of four oceanic
regions. Plos Biol 4: 2121-2131
Attaoui H, Jaafar FM, Belhouchet M, Micco P, de Lamballerie X, Brussaard CPD (2006)
Micromonas pusilla reovirus: a new member of the family Reoviridae assigned to
a novel proposed genus (Mimoreovirus). J Gen Virol 87:1375-1383
Auguet J-C, Montanié H, Lebaron P (2005) Structure of virioplankton in the Charente
Estuary (France): transmission electron microscopy versus pulsed field gel
electrophoresis. Microb Ecol 51:197-208
Azam F, Fenchel T, Field JG, Gray JS, Meyer-Reil LA, Thingstad F (1983) The
ecological role of water-column microbes in the sea. Mar Ecol Prog Ser 10:257–
263
23
Chapter 1
Introduction
Berges JA, Falkowski PG (1998) Physiological stress and cell death in marine
phytoplankton: induction of proteases in response to nitrogen or light limitation.
Limnol Oceanogr 43:129-135
Bergh Ø, Børsheim KY, Bratbak G, Heldal M (1989) High abundance of viruses found
in aquatic environments. Nature 340:467-468
Berman-Frank I, Bidle KD, Haramaty L, Falkowski PG (2004) The demise of the marine
cyanobacterium, Trichodesmium spp., via an autocatalyzed cell death pathway.
Limnol Oceanogr 49:997-1005
Bettarel Y, Kan J, Wang K, Williamson KE, Cooney S, Ribblett S, Chen F, Wommack
KE, Coats DW (2005) Isolation and preliminary characterization of a small
nuclear inclusion virus infecting the diatom Chaetoceros cf. gracilis. Aquat
Microb Ecol 40:103-114
Binder B (1999) Reconsidering the relationship between virally induced bacterial
mortality and frequency of infected cells. Aquat Microb Ecol 18:207–215
Bongiorni L, Magagnini M, Armeni M, Noble R, Danovaro R (2005) Viral production,
decay rates, and life strategies along a trophic gradient in the north Adriatic Sea.
Appl Environ Ecol 71:6644-6650
Bratbak G, Egge JK, Hedal M (1993) Viral mortality of the marine alga Emiliania
huxleyi (Haptophyceae) and termination of algal blooms. Mar Ecol Prog Ser
93:39-48
Bratbak G, Heldal M, Norland S, Thingstad TF (1990) Viruses as partners in spring
bloom microbial trophodynamics. Appl Environ Microbiol 56:1400-1405
Bratbak G, Heldal M, Thingstad TF, Tuomi P (1996) Dynamics of virus abundance in
coastal seawater. FEMS Microb Ecol 19:263-269
Bratbak G, Jacobsen A, Heldal M, Nagasaki K, Thingstad F (1998) Virus production in
Phaeocystis pouchetii and its relation to host cell growth and nutrition. Aquat
Microb Ecol16:1-9
Breitbart M, Salamon P, Andresen B, Mahaffy JM, Segall AM, Mead D, Azam F,
Rohwer F (2002) Genomic analysis of uncultured marine viral communities.
PNAS 99:14250-14255
Brussaard CPD (2004) Viral control of phytoplankton populations - a review. J Euk
Microbiol 51:125-138
Brussaard CPD, Gast GJ, Van Duyl FC, Riegman R (1996a) Impact of phytoplankton
bloom magnitude on a pelagic microbial food web. Mar Ecol Prog Ser 144:211221
Brussaard CPD, Kempers RS, Kop AJ, Riegman R, Heldal M (1996b) Virus-like
particles in a summer bloom of Emiliania huxleyi in the North Sea. Aquat Microb
Ecol 10:105-113
Brussaard CPD, Kuipers B, Veldhuis MJW (2005a) A mesocosm study of Phaeocystis
globosa population dynamics - I. Regulatory role of viruses in bloom. Harmful
Algae 4:859-874
Brussaard CPD, Mari X, Van Bleijswijk JDL, Veldhuis MJW (2005b) A mesocosm
study of Phaeocystis globosa (Prymnesiophyceae) population dynamics - II.
Significance for the microbial community. Harmful Algae 4:875-893
24
Chapter 1
Introduction
Brussaard CPD, Noordeloos AAM, Riegman R (1997) Autolysis kinetics of the marine
diatom Ditylum brightwellii (Bacillariophyceae) under nitrogen and phosphorus
limitation and starvation. J Phycol 33:980-987
Brussaard CPD, Noordeloos AAM, Sandaa RA, Heldal M, Bratbak G (2004a) Discovery
of a dsRNA virus infecting the marine photosynthetic protist Micromonas pusilla.
Virol 319:280-291
Brussaard CPD, Riegman R, Noordeloos AAM, Cadee GC, Witte H, Kop AJ,
Nieuwland G, Vanduyl FC, Bak RPM (1995) Effects of grazing, sedimentation
and phytoplankton cell lysis on the structure of a coastal pelagic food-web. Mar
Ecol Prog Ser 123:259-271
Brussaard CPD, Short SM, Frederickson CM, Suttle CA (2004b) Isolation and
phylogenetic analysis of novel viruses infecting the phytoplankton Phaeocystis
globosa (Prymnesiophyceae). Appl Environ Microbiol 70:3700-3705
Castberg T, Larsen A, Sandaa RA, Brussaard CPD, Egge JK, Heldal M, Thyrhaug R,
van Hannen EJ, Bratbak G (2001) Microbial population dynamics and diversity
during a bloom of the marine coccolithophorid Emiliania huxleyi (Haptophyta).
Mar Ecol Prog Ser 221:39-46
Charlson RJ, Lovelock JE, Andreae MO, Warren SG (1987) Oceanic phytoplankton,
atmospheric sulfur, cloud albedo and climate. Nature 326:655-661
Cottrell MT, Suttle CA (1995a) Dynamics of a lytic virus infecting the photosynthetic
marine picoflagellate Micromonas pusilla. Limnol Oceanogr 40:730-739
Cottrell MT, Suttle CA (1995b) Genetic diversity of algal viruses which lyse the
photosynthetic picoflagellate Micromonas pusilla (Prasinophyceae). Appl
Environ Microbiol 61:3088-3091
Culley AI, Welschmeyer NA (2002) The abundance, distribution, and correlation of
viruses, phytoplankton, and prokaryotes along a Pacific Ocean transect. Limnol
Oceanogr 47:1508-1513
Danovaro R, Dell'Anno A, Trucco A, Serresi M, Vanucci S (2001) Determination of
virus abundance in marine sediments. Appl Environ Microbiol 67:1384-1387
Evans C, Archer SD, Jacquet S, Wilson WH (2003) Direct estimates of the contribution
of viral lysis and microzooplankton grazing to the decline of a Micromonas spp.
population. Aquat Microb Ecol 30:207-219
Evans C, Malin G, Wilson W, Liss PS (2006) Infectious titers of Emiliania huxleyi virus
86 are reduced by exposure to millimolar dimethyl sulfide and acrylic acid.
Limnol Oceanogr 51:2468-2471
Flint SJ, Enquist LW, Krug RM, Racaniello VR, Skalka AM (2000) Principle of
virology - Molecular biology, pathogenesis, and control. ASM Press, Washington,
D.C.
Franklin DJ, Brussaard CPD, Berges JA (2006) What is the role and nature of
programmed cell death in phytoplankton ecology? Europ J Phycol 41:1-14
Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects.
Nature 399:541-548
Fukami K, Yuzawa A, Nishijima T, Hata Y (1992) Isolation and properties of a
bacterium inhibiting the growth of Gymnodinium nagasakiense. Nippon Suisan
25
Chapter 1
Introduction
Gakkaishi 58:1073-1077
Füller NJ, Wilson WH, Joint IR, Mann NH (1998) Occurrence of a sequence in marine
cyanophages similar to that of T4 g20 and its application to PCR-based detection
and quantification techniques. Appl Environ Microbiol 64:2051-2060
Garza DR, Suttle CA (1998) The effect of cyanophages on the mortality of
Synechococcus spp. and selection for UV resistant viral communities. Microb
Ecol 36:281-292
Gastrich MD, Leigh-Bell JA, Gobler CJ, Anderson OR, Wilhelm SW, Bryan M (2004)
Viruses as potential regulators of regional brown tide blooms caused by the alga,
Aureococcus anophagefferens. Estuaries 27:112-119
Göbler CJ, Hutchins DA, Fisher NS, Cosper EM, Sanudo-Wilhelmy SA (1997) Release
and bioavailability of C, N, P, Se, and Fe following viral lysis of a marine
chrysophyte. Limnol Oceanogr 42:1492-1504
Gonzalez JM, Suttle CA (1993) Grazing by marine nanoflagellates on viruses and virussized particles - ingestion and digestion. Mar Ecol Prog Ser 94:1-10
Hara S, Koike I, Terauchi K, Kamiya H, Tanoue E (1996) Abundance of viruses in
marine in deep oceanic waters. Mar Ecol Prog Ser 145:269-277
Hewson I, Govil SR, Capone DG, Carpenter EJ, Fuhrman J (2004) Evidence of
Trichodesmium viral lysis and potential significance for biogeochemical cycle in
the oligotrophic ocean. Aquat Microb Ecol 36:1-8
Hill RW, A. WB, Cottrell MT, Dacey JWH (1998) Virus-mediated total release of
dimethylsulfoniopropionate from marine phytoplankton: a potential climate
process. Aquat Microb Ecol 14:1-6
Jacquet S, Bratbak G (2003) Effects of ultraviolet radiation on marine virusphytoplankton interactions. FEMS Microbiol Ecol 44:279-289
Jacquet S, Heldal M, Iglesias-Rodriguez D, Larsen A, Wilson W, Bratbak G (2002)
Flow cytometric analysis of an Emiliania huxleyi bloom terminated by viral
infection. Aquat Microb Ecol 27:111-124
Kirchman DL (1999) Oceanography - Phytoplankton death in the sea. Nature 398:293294
Kuipers BR, Witte HJ (2000) Prochlorophytes as secondary prey for heterotrophic
nanoflagellates in the deep chlorophyll maximum layer of the (sub)tropical North
Atlantic. Mar Ecol Prog Ser 204:53-63
Landry MR, Hassett RP (1982) Estimating the grazing impact of marine
microzooplankton. Mar Biol 67:283-288
Lang AS, Culley AI, Suttle CA (2004) Genome sequence and characterization of a virus
(HaRNAV) related to picoma-like viruses that infects the marine toxic bloomforming alga Heterosigma akashiwo. Virology 320:206-217
Larsen A, Castberg T, Sandaa RA, Brussaard CPD, Egge J, Heldal M, Paulino A,
Thyrhaug R, van Hannen EJ, Bratbak G (2001) Population dynamics and
diversity of phytoplankton, bacteria and viruses in a seawater enclosure. Mar Ecol
Prog Ser 221:47-57
Lawrence JE, Chan AM, Suttle CA (2001) A novel virus (HaNIV) causes lysis of the
toxic bloom-forming alga Heterosigma akashiwo (Raphidophyceae). J Phycol
26
Chapter 1
Introduction
37:216-222
Lawrence JE, Suttle CA (2004) Effect of viral infection on sinking rates of Heterosigma
akashiwo and its implications for bloom termination. Aquat Microb Ecol 37:1-7
Legrand C, Rengefors K, Fistarol GO, Graneli E (2003) Allelopathy in phytoplankton:
biochemical, ecological and evolutionary aspects. Phycologia 42:406-419
Lenski RE (1988) Dynamics of interactions between bacteria and virulent bacteriophage.
Adv Microb Ecol 10:1 - 44
Levin BR, Steward FM, Chao L (1977) Resource-limited growth, competition, and
predation: a model and experimental studies with bacteria and bacteriophage. Am
Nat 111:3
MacKenzie JJ, Haselkorn R (1972) Photosynthesis and the development of a blue-green
algae virus SM-1. Virol 49:517-521
Malin G, Wilson WH, Bratbak G, Liss PS, Mann NH (1998) Elevated production of
dimethylsulfide resulting from viral infection of cultures of Phaeocystis pouchetii.
Limnol Oceanogr 43:1389-1393
Mann NH (2003) Phages of the marine cyanobacterial picophytoplankton. FEMS
Microbiol Rev 27:17-34
Mayali X, Azam F (2004) Algicidal bacteria in the sea and their impact on algal blooms.
J Euk Microbiol 51:139-144
Middelboe M, Jorgensen NOG, Kroer N (1996) Effects of viruses on nutrient turnover
and growth efficiency of noninfected marine bacterioplankton. Appl Environ
Microbiol 62:1991-1997
Middelboe M, Riemann L, Steward GF, Hansen V, Nybroe O (2003) Virus-induced
transfer of organic carbon between marine bacteria in a model community. Aquat
Microb Ecol 33:1-10
Mühling M, Fuller NJ, Millard A, Somerfield PJ, Marie D, Wilson WH, Scanlan DJ,
Post AF, Joint I, Mann NH (2005) Genetic diversity of marine Synechococcus and
co-occurring cyanophage communities: evidence for viral control of
phytoplankton. Environ Microbiol 7:499-508
Murray AG, Jackson GA (1992) Viral dynamics - A model of the effects of size, shape,
motion and abundance of single-celled planktonic organisms and other particles.
Mar Ecol Prog Ser 89:103-116
Nagasaki K, Ando M, Itakura S, Imai I, Ishida Y (1994) Viral mortality in the final stage
of Heterosigma akashiwo (raphidophyceae) red tide. J Plankton Res16:1595-1599
Nagasaki K, Tomaru Y, Katanozaka N, Shirai Y, Nishida K, Itakura S, Yamaguchi M
(2004) Isolation and characterization of a novel single-stranded RNA virus
infecting the bloom-forming diatom Rhizosolenia setigera. Appl Environ
Microbiol 70:704-711
Nagasaki K, Yamaguchi M (1998) Effect of temperature on the algicidal activity and the
stability of HaV (Heterosigma akashiwo virus). Aquat Microb Ecol 15:211-216
Noble RT, Fuhrman JA (1997) Virus decay and its causes in coastal waters. Appl
Environ Microbiol 63:77-83
Noble RT, Fuhrman JA (2000) Rapid virus production and removal as measured with
fluorescently labeled viruses as tracers. Appl Environ Microbiol 66:3790-3797
27
Chapter 1
Introduction
Ohki K (1999) A possible role of temperate phage in the regulation of Trichodesmium
biomass. Bull Inst Oceanogr Monaco 19:287-292
Proctor LM, Fuhrman JA (1990) Viral mortality of marine bacteria and cyanobacteria.
Nature 343:60-62
Proctor LM, Fuhrman JA (1991) Roles of viral-infection in organic particle-flux. Mar
Ecol Prog Ser 69:133-142
Proctor LM, Okubo A, Fuhrman JA (1993) Calibrating estimates of phage-induced
mortality in marine bacteria: ultrastructural studies of marine bacteriophage
development from one-step growth experiments. Microb Ecol 25:161–182
Raven JA (1998) The twelfth Tansley Lecture. Small is beautiful: the
picophytoplankton. Functional Ecology 12:503-513
Riegman R, Kuipers BR, Noordeloos AAM, Witte HJ (1993) Size-differential control of
phytoplankton and the structure of plankton communities. Neth J Sea Res 31:255265
Ruardij P, Veldhuis MJW, Brussaard CPD (2005) Modeling the bloom dynamics of the
polymorphic phytoplankter Phaeocystis globosa: impact of grazers and viruses.
Harmful Algae 4:941-963
Sandaa RA, Skjoldal EF, Bratbak G (2003) Virioplankton community structure along a
salinity gradient in a solar saltern. Extremophiles 7:347-351
Schroeder DC, Oke J, Malin G, Wilson WH (2002) Coccolithovirus (Phycodnaviridae):
characterization of a new large dsDNA algal virus that infects Emiliania huxleyi.
Arch Virol 147:1685-1698
Sieburth JMN, Johnson PW, Hargraves PE (1988) Ultrastructure and ecology of
Aureocococcus anophagefferens gen. et sp. nov. (Chrysophyceae): the dominant
picoplankter during a bloom in Narragansett Bay, Rhode Island. Summer 1985. J
Phycol 24:416-425
Smetack JS (1985) Role of sinking diatom life-history cycles: ecological, evolutionary
and geological significance. Mar Biol 84:239-251
Suttle CA (1992) Inhibition of photosynthesis in phytoplankton by the submicron size
fraction concentrated from seawater. Mar Ecol Prog Ser 87:105-112
Suttle CA (2000) Ecological, evolutionary and geochemical consequences of viral
infection of cyanobacteria and eukaryotic algae. In: Hurst CJ (ed) Viral Ecology.
Academic Press, p 247-296
Suttle CA (2005) Viruses in the sea. Nature 437:356-361
Suttle CA, Chan AM (1994) Dynamics and distribution of cyanophages and their effect
on marine Synechococcus spp. Appl Environ Microbiol 60:3167-3174
Suttle CA, Chan AM, Cottrell MT (1990) Infection of phytoplankton by viruses and
reduction of primary productivity. Nature 347:467-469
Suttle CA, Chen F (1992) Mechanisms and rates of decay of marine viruses in seawater.
Appl Environ Microbiol 58:3721-3729
Tai V, Lawrence JE, Lang AS, Chan AM, Culley AI, Suttle CA (2003) Characterization
of HaRNAV, a single-stranded RNA virus causing lysis of Heterosigma akashiwo
(Raphidophyceae). Journal of Phycology 39:343-352
Tarutani K, Nagasaki K, Yamaguchi M (2000) Viral impacts on total abundance and
28
Chapter 1
Introduction
clonal composition of the harmful bloom-forming phytoplankton Heterosigma
akashiwo. Appl Environ Microbiol 66:4916-4920
Thingstad TF (2000) Elements of a theory for the mechanisms controlling abundance,
diversity, and biogeochemical role of lytic bacterial viruses in aquatic systems.
Limnol Oceanogr 45:1320-1328
Thyrhaug R, Larsen A, Brussaard CPD, Bratbak G (2002) Cell cycle dependent virus
production in marine phytoplankton. J Phycol 38:338-343
Thyrhaug R, Larsen A, Thingstad TF, Bratbak G (2003) Stable coexistence in marine
algal host-virus systems. Mar Ecol Prog Ser 254:27-35
Tomaru Y, Katanozaka N, Nishida K, Shirai Y, Tarutani K, Yamaguchi M, Nagasaki K
(2004a) Isolation and characterization of two distinct types of HcRNAV, a singlestranded RNA virus infecting the bivalve-killing microalga Heterocapsa
circularisquama. Aquat Microb Ecol 34:207-218
Tomaru Y, Tarutani K, Yamaguchi M, Nagasaki K (2004b) Quantitative and qualitative
impacts of viral infection on a Heterosigma akashiwo (Raphidophyceae) bloom in
Hiroshima Bay, Japan. Aquat Microb Ecol 34:227-238
Van Etten JL, Lane LC, Meints RH (1991) Viruses and virus-like particles in eukaryotic
algae. Microbiol Rev 55:586-620
Vardi A, Berman-Frank I, Rozenberg T, Hadas O, Kaplan A, Levine A (1999)
Programmed cell death of the dinoflagellate Peridinium gatunense is mediated by
CO2 limitation and oxidative stress. Curr Biol 9:1061-1064
Vardi A, Schatz D, Beeri K, Motro U, Sukenik A, Levine A, Kaplan A (2002)
Dinoflagellate-cyanobacterium communication may determine the composition of
phytoplankton assemblage in a mesotrophic lake. Curr Biol 12:1767-1772
Waterbury JB, Valois FW (1993) Resistance to co-occurring phages enables marine
Synechococcus communities to coexist with cyanophages abundant in seawater.
Appl Environ Microbiol 59:3393-3399
Waters RE, Chan AT (1982) Micromonas pusilla virus: the virus growth cycle and
associated physiological events within the host cells; host range mutation. Virol
63:199–206
Weinbauer MG (2004) Ecology of prokaryotic viruses. FEMS Microbiol Rev 28:127181
Weinbauer MG, Fuks D, Peduzzi P (1993) Distribution of viruses and dissolved DNA
along a coastal trophic gradient in the northern Adriatic Sea. Appl Environ
Microbiol 59:4074-4082
Wilhelm SW, Jeffrey WH, Suttle CA, Mitchell DL (2002) Estimation of biologically
damaging UV levels in marine surface waters with DNA and viral dosimeters.
Photochem Photobiol 76:268-273
Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea. BioScience
49:781-788
Williamson SJ, McLaughlin MR, Paul JH (2001) Interaction of the ΦHSIC virus with its
host: lysogeny or pseudolysogeny? Appl Environ Microbiol 67:1682-1688
Wilson W, Tarran GA, Schroeder DC, Cox M, Oke J, Malin G (2002) Isolation of
viruses responsible for the demise of an Emiliania huxleyi bloom in the English
29
Chapter 1
Introduction
Channel. J Mar Biol Ass UK 82:369-377
Wilson WH, Carr NG, Mann NH (1996) The effect of phosphate status on the kinetics of
cyanophage infection in the oceanic cyanobacterium Synechococcus sp WH7803.
J Phycol 32:506-516
Wilson WH, Schroeder DC, Allen MJ, Holden MTG, Parkhill J, Barrell BG, Churcher
C, Hamlin N, Mungall K, Norbert H, Quail MA, Proce C, Rabbinowitsch E,
Walker D, Craignon M, Roy D, Ghazal P (2005a) Complete genome sequence
and lytic phase transcription profile of a Coccolithovirus. Science 309:1090-1092
Wilson WH, Van Etten JL, Schroeder DC, Nagasaki K, Brussaard CPD, Delaroque N,
Bratbak G, Suttle CA (2005b) Family: Phycodnaviridae. In: Fauquet CM, Mayo
MA, Maniloff J, Dusselberger U, Ball LA (eds) Virus Taxonomy, VIIIth ICTV
Report. Elsevier/Academic Press, London, p 163-175
Winget DM, Williamson KE, Helton RR, Wommack KE (2005) Tangential flow
diafiltration: an improved technique for estimation of virioplankton production.
Aquat Microb Ecol 41:221-232
Wommack KE, Ravel J, Hill RT, Chun JS, Colwell RR (1999) Population dynamics of
Chesapeake bay virioplankton: Total-community analysis by pulsed-field gel
electrophoresis. Appl Environ Microbiol 65:231-240
Zhong Y, Chen F, Wilhelm SW, Poorvin L, Hodson RE (2002) Phylogenetic diversity of
marine cyanophage isolates and natural virus communities as revealed by
sequences of viral capsid assembly protein gene g20. Appl Environ Microbiol
68:1576-158
30
Chapter 2
Virally induced mortality of
Phaeocystis globosa during two spring
blooms in temperate coastal waters 1
Anne-Claire Baudoux, Anna A. M. Noordeloos, Marcel J. W. Veldhuis
and Corina P. D. Brussaard
This study reports virally mediated mortality rates of Phaeocystis globosa
single cells in the southern North Sea during two consecutive spring blooms (2003 and
2004). An adapted dilution method was used to estimate simultaneously viral lysis and
microzooplankton grazing. Parallel dilution experiments were performed with 30 kDa
ultrafiltrate (virus and grazer-free diluent) and 0.2 µm filtered seawater (grazer-free,
but virus-containing diluent). Specific viral lysis rates were calculated from the
difference in P. globosa growth rates between the two dilutions series after 24 h
incubation under natural conditions. The validity of this method was tested using a
culture P. globosa infected with a known P. globosa virus (PgV). The field data show
that virally induced mortality can be a substantial loss factor for P. globosa single cells
(up to 0.35 d-1), comparable to microzooplankton grazing (up to 0.4 d-1). Viral lysis was
the major cause of total P. globosa cell lysis. Assuming no loss due to sinking, viral lysis
accounted for 5 to 66% of the total mortality of P. globosa single cells. Viral lysis and
total putative PgV abundance increased concomitantly with P. globosa single cell
abundance whilst the increase in infective PgV was delayed. This delay may be caused
by the formation of transparent exopolymeric particles that are generated when P.
globosa colonies disrupt and known to passively adsorb viruses. Viruses and
microzooplankton were shown to be major controlling agents for P. globosa single cells,
although their relative significance varied over the course of the bloom and between
years.
1
Published in Aquatic Microbial Ecology 44:207 – 217, 2006
31
Chapter 2
Virally induced mortality of Phaeocystis globosa
1. Introduction
With phytoplankton forming the basis of the pelagic marine food web, their
dynamics critically influence the functioning of marine ecosystems. Traditionally,
grazing and sinking are considered important source of phytoplankton mortality but over
the last the decade, cell lysis has also been recognized as a significant loss factor for
phytoplankton (Van Boekel et al. 1992, Brussaard et al. 1995, Agusti et al. 1998). Viral
infection is a major cause of phytoplankton cell lysis, which affects population dynamics
and diversity (for review see Brussaard 2004b). Successful infection depends on the
encounter between the virus and the host, which is directly affected by their abundance.
During algal bloom events, which are defined by high cell abundance, virally induced
mortality has indeed been reported as a substantial loss factor (Bratbak et al. 1993,
Brussaard et al. 1996b, Tomaru et al. 2004).
Phaeocystis globosa (Prymnesiophyte) is a world-widely distributed bloomforming phytoplankter. This marine microphytoplankter is well-known for its complex
polymorphic life cycle, including flagellated cells (5-7 µm in diameter) and colonies (up
to 1-2 centimeters), which consist of colonial cells embedded in a polysaccharide
(mucus) matrix. Typically in the southern North Sea, P. globosa develops high biomass
spring blooms, which contribute for the bulk of local primary production (Lancelot &
Billen 1984). These blooms also affect microbial food web dynamics and
biogeochemical processes (Stefels & Van Boekel 1993, Brussaard et al. 1995, Brussaard
et al. 1996a, Brussaard et al. 2005b).
Viruses infecting P. globosa (PgVs) have been isolated and characterized
(Brussaard et al. 2004, Baudoux & Brussaard 2005). A recent mesocosm experiment
demonstrated that virally mediated mortality of P. globosa accounted for 30-100% of the
total lysis of the P. globosa single cells. In contrast, cells embedded in a colonial matrix
tend to escape viral infection (Brussaard et al. 2005a, Ruardij et al. 2005). To our
knowledge, estimates of virally mediated mortality under natural P. globosa bloom
conditions do not exist. The dynamics in viral abundance, virus to P. globosa ratio and
total cell lysis rates of P. globosa during bloom events, however, suggest that viruses
play an important ecological role (Brussaard et al. 2004, Brussaard et al. 2005a).
Direct methods for estimating viral lysis of phytoplankton are scarce, and to
date, most of the viral lysis rates recorded in the literature rely on theoretical calculations
or are based on indirect measurements (for review see Brussaard 2004b). Recently, an
adaptation of the classical dilution approach (Landry & Hassett 1982) provided estimates
of the viral lysis of the picophytoplankter Micromonas pusilla in mesocosms (Evans et
al. 2003). The current study has applied this method to P. globosa under natural
conditions. Our work aims to elucidate the relative significance of viral lysis as
compared to microzooplankton grazing and total cell lysis of P. globosa single cells
during two consecutive spring blooms.
32
Chapter 2
Virally induced mortality of Phaeocystis globosa
2. Material and methods
Study site and sampling. Sampling of the coastal southern North Sea was
performed twice a week between March (Day 60) and June (Day 180) in 2003 and 2004
from the jetty of the Royal Netherlands Institute for Sea Research (NIOZ). Because the
jetty is located at the outer border of a major tidal inlet, samples were collected at
incoming high tide. Samples containing freshwater run-off (salinity < 27 ‰) were not
taken into account (1 out of 46 samples).
Chemical parameters. Nutrient samples (approx. 5 mL) were gently filtered
through 0.2 µm pore-size polysulfone filters (Acrodisc, Gelman Sciences) and stored at 50°C (or 4°C for the reactive silicate) until analysis. Analyses were performed using a
TrAAcs 800 autoanalyzer for dissolved orthophosphate (Murphy & Riley 1962),
nitrogen (nitrate, nitrite and ammonium; Grasshoff 1983, Helder & De Vries 1979), and
reactive silicate (Strickland & Parsons, 1968). The limit of detection was 0.03 µM for
phosphate, 0.1 µM for ammonium, 0.01 µM for nitrite, 0.15 µM for nitrate, and 0.05 µM
for silicate.
The concentration of transparent exopolymeric particles (TEP, in µg Equivalent
Gum Xanthan L-1) was measured according to Passow and Alldredge (1995). Replicate
samples (30-75 mL) were filtered through 0.4 µm pore-size polycarbonate filters
(Poretics). The particles retained on the filter were stained with 500 µL of a 0.02 %
solution of Alcian blue prepared in 0.06 % acetic acid (pH 2.5). After staining (< 2 s),
the filters were rinsed 3 times with MilliQ (Millipore) to remove excess dye. The filters
were immediately transferred into 20 mL glass tubes and soaked for 3 h in a solution of
80 % H2SO4 with gentle agitation every 30 min. The samples were analyzed
spectrophotometrically at 727 nm (U-3010 Hitachi).
Microbial abundances. Samples collected for phytoplankton pigments (150700 mL) were filtered onto GF/F glassfiber filters (Whatman) and stored at –50°C. The
extract from the filters was analyzed by high pressure liquid chromatography (HPLC)
after extraction in 4 mL of 100 % methanol buffered with 0.5 mol L-1 ammonium acetate
and homogenized for 15 s. The relative abundance of the taxonomic group
Prymnesiophyceae (specifically Phaeocystis globosa during our study) was determined
using CHEMTAX (Mackey et al. 1996, Riegman & Kraay 2001)
Phaeocystis globosa single cells were enumerated in 50 µm-sieved and unfixed
samples using a Beckman Coulter XL-MCL flow cytometer equipped with a 488 nm aircooled laser. Special care was taken to avoid rupture of P. globosa colonies during
sieving using a small volume of sample. Fixation of the sample resulted in the
disintegration of the colonial matrix, therefore the total abundance of P. globosa cells
(including both single and colonial cells) could be obtained from unfiltered samples that
33
Chapter 2
Virally induced mortality of Phaeocystis globosa
were fixed to a 1 % final concentration with formaldehyde:hexamine solution (18 %
v/v:10 % w/v). Fixation did not affect the P. globosa cell counts. These fixed samples
were frozen in liquid nitrogen and stored at -80ºC until flow cytometric analysis. P.
globosa cells were discriminated on the basis of their natural red chlorophyll
autofluorescence and forward scatter signal.
Green fluorescence (r.u.)
The abundance of virus-like-particles resembling P. globosa like viruses (PgV)
was determined on glutaraldehyde fixed samples (final concentration 0.5 %
glutaraldehyde, frozen in liquid nitrogen and stored at -80°C prior analysis) using a
Beckton-Dickinson FACSCalibur flow cytometer, with a 15 mW 488 nm air-cooled
argon-ion laser according to Brussaard (2004a). Thawed samples were diluted (dilution
factor >25) in 0.2 µm filtered sterile TE-buffer (pH 8) and stained with the nucleic acidspecific dye SYBR Green I at a final concentration of 0.5×10-4 of the commercial stock
(Molecular Probes, Eugene, OR). Putative PgV could be discriminated on the basis of
the green fluorescence and side scatter signature (Fig. 1), which was identical to that of
PgV isolates from the same geographical location and kept in culture at the Royal NIOZ
(Brussaard et al. 2004).
104
beads
bacteria
103
102
PgVs
101
other viruses
100
100
101
102
103
104
Side scatter (r.u.)
Figure 1. Flow cytometric signature of viruses infecting Phaeocystis globosa (PgV) from natural
seawater. PgVs were detected based on their green fluorescence and side scatter upon staining
with SYBR Green I.
The abundance of infectious PgV was estimated using the end-point dilution
approach (Most Probable Number, MPN, Suttle 1993). Natural seawater was filtered
through an 1 µm polycarbonate filter (Poretics) and serial diluted (8 titers, 5 replicates)
with exponentially growing P. globosa Pg-G (RUG culture collection, The Netherlands)
and Pg-01MD06 (NIOZ culture collection). To screen for rare PgV in 2004, an
additional natural sample was filtered through an 1 µm polycarbonate filter (Poretics),
34
Chapter 2
Virally induced mortality of Phaeocystis globosa
concentrated approx. 40 times using a VivaFlow 200 ultrafiltration system
(Vivascience), and added to a P. globosa host culture (20% v/v). Both P. globosa strains
originated from the North Sea and were chosen for their different sensitivity to PgV,
which was relatively broad for Pg-G and specific for Pg-01MD06 (Baudoux &
Brussaard 2005). The algae were grown in a 1:1 mixture of f/2 medium (Guillard 1975)
and enriched artificial seawater ESAW, (Harrison et al. 1980, Cottrell & Suttle 1991)
completed with Tris-HCl and Na2SeO3 (Cottrell & Suttle 1991). The dilution series were
incubated for 10 days at 15°C under a light:dark cycle of 16:8 h at 100 µmol photon m-2
s-1. Algal growth was monitored via in vivo chlorophyll fluorescence using a Turner
Designs fluorometer and compared to noninfected controls. Those dilutions that showed
signs of cell lysis were scored positive when PgV proliferation could be confirmed
(using flow cytometry as described above). The positive scores were converted to
abundance of infective PgV using a MPN assay computer program (Hurley & Roscoe
1983).
Loss parameters of P. globosa. Total cell lysis rates of P. globosa (d-1,
unspecific cause) were estimated using the dissolved esterase activity (DEA) assay
described in Brussaard et al. (1996a) and adapted by Riegman et al. (2002). Particulate
esterase activity was obtained by subtracting the dissolved esterase activity (0.2 µm
pore-size filtered) from the total esterase activity (unfiltered natural sample). The
dissolved esterase activity was corrected for non-enzymatic hydrolysis of the substrate,
as measured in natural sample filtered through 10 kDa (PES Vivaspin, Vivasciences),
and for a decay of esterase activity in seawater using a half-life time of 48 h (Riegman et
al. 2002). The P. globosa-specific particulate esterase activity was calculated by
multiplying the total particulate esterase activity by the contribution of P. globosa to
total chlorophyll based CHEMTAX pigment analysis described in the above section
(Brussaard et al. 2004). Data points were occasionally omitted when an unrealistically
high lysis rate was obtained from the ratio of low produced DEA to low P. globosa
specific PEA. This may occur at the onset of the bloom when P. globosa biomass is still
low.
Virally induced mortality of P. globosa single cells was estimated using the
viral lysis dilution assay according to Evans et al. (2003). Parallel dilution series of
natural seawater was performed with 0.2 µm filtered natural sample (Poretics, Millipore)
to obtain microzooplankton grazing rate (Landry & Hassett 1982), and with 30 kDa
filtered natural sample (polyether sulfone membrane, Pellicon filtration system,
Millipore) to obtain grazing and viral lysis rates. Viral lysis rates were determined from
the difference between the two dilutions series.
The viral lysis dilution assay has only been applied to Micromonas pusilla
(Evans et al. 2003) therefore we checked the validity of this method for P. globosa. The
test experiment was conducted using an exponentially growing P. globosa Pg-G (1×105
cells mL-1) in combination with the lytic virus PgV-07T (Brussaard et al. 2004) at a
multiplicity of infection (MOI) of 10, as determined by MPN assay (described above). In
order to simplify the interpretation of the test results no grazers were added, therefore
35
Chapter 2
Virally induced mortality of Phaeocystis globosa
identical net growth rates were recorded for all dilutions with the 0.2 µm pore-size
water, and thus the estimated grazing rate was not significantly different from zero (0.02
± 0.07 d-1, Fig. 2A). The dilution series with the <30 kDa diluent yielded a regression
slope of 1.3 ± 0.07 d-1, which corresponds to the viral lysis rate in this test since there
was no grazing. Knowing that P. globosa undergoes lysis 14-16 h after infection
(Baudoux & Brussaard 2005), we conclude that the lysis rate obtained during the 24 h
incubation originates from one lytic cycle. An independent one-step lytic growth cycle
experiment using the same strain of P. globosa and PgV validated the results of the
dilution assay. The viral lysis rate (1.4 d-1) calculated from this growth experiment were
comparable to those obtained with the laboratory viral lysis dilution assay (1.3 d-1).
These tests demonstrated the utility and validity of viral lysis dilution assay, allowing
this method to be applied in the field for P. globosa (Fig. 2B).
Apparent growth (d-1)
0.6
30 kDa
0.2 µm
0.3
0.0
-0.3
-0.6
-0.9
Apparent growth (d-1)
-1.2
0.4
30 kDa
0.2 µm
0.2
0.0
-0.2
-0.4
-0.6
0.0
0.2
0.4
0.6
0.8
1.0
Fraction of natural sample
Figure 2. Virally mediated mortality rates of P. globosa for (A) a test experiment using the virushost model system PgV-07T and Pg-G (no grazers were added) and (B) a typical field sample
(Day 132 in 2003). Parallel dilution experiments were performed in 0.2 µm filtered seawater
(grazer-free, but virus-containing diluent) and 30 kDa ultrafiltrate (virus and grazer-free diluent).
The regression coefficient of apparent growth rate vs. dilution factors resulting from the 0.2 µm
dilution series represents the microzooplankton grazing rate, and from the 30 kDa series
represents microzooplankton grazing as well as viral lysis. Viral lysis rates (d-1) were estimated
from the difference in regression coefficient of the two set of dilutions. For the readability of the
figures, we averaged the triplicate apparent growth for each dilution level. This operation does
not affect the estimated mortality rates.
36
Chapter 2
Virally induced mortality of Phaeocystis globosa
The experimental design of the dilution assay for field samples is described in
Fig. 3. All materials (carboys, tubing, bottles) used for this assay were cleaned for 24 h
with 0.1 N HCl, after which they were rinsed 3 times with MilliQ and once with the
sample. To prevent losses of virus, grazers or disruption of P. globosa colonies, sieving
and filtration were performed with special care by siphoning and avoiding air bubbling.
Polycarbonate Poretics filters (47 mm, Millipore) were exclusively used and replaced
frequently during the filtration to avoid loss of viral infectivity and abundance (Suttle et
al. 1991). Samples were processed under dimmed light (to prevent light stress) and in
situ temperature (4-18ºC). The seawater used for the dilution (10 L) was collected and
processed approx. 2 h before high tide in order to minimize the handling time of the
natural water to be diluted. The 10 L sample was pretreated by reverse sieving through
200 and 50 µm mesh (20 cm diameter) to remove larger grazers and P. globosa colonies.
Subsequently, the sample was filtered through 3 µm and 0.2 µm pore-size filters. A 5 L
aliquot of the 0.2 µm filtrate was used for generating the 0.2 µm dilution series. The
remaining 5 L were ultrafiltered through 30 kDa and used as diluent for the 30 kDa
dilution series.
At high tide, 20 L of natural seawater was de novo collected. The salinity of the
two batches of seawater used for the experiment was measured and found comparable in
all cases (difference <0.5 ‰). The sample was sieved through 200 µm mesh by reverse
sieving to remove mesozooplankton and immediately used to set up four levels of
dilution (20, 40, 70 and 100 % sample) in 2 L polycarbonate bottles that already
contained the 0.2 µm or the 30 kDa diluents. From the 2 L dilution bottle, 3 incubation
bottles (250 mL polycarbonate bottles) were carefully filled by siphoning and 5 mL
subsample were taken (T=0 h). These incubation bottles were refilled to the top with the
original dilution waters (remaining from the 2 L bottles) in order to avoid any air
bubbles being trapped inside upon closure. All bottles were incubated at near-surface
depth (approx. 1-2 m) under natural light and temperature conditions in a basket in the
NIOZ harbor (protected from wave-motion). Another sample of 5 mL was taken after a
24 h incubation period.
The set-up of the laboratory assay was similar to the field assay with the
exception that a P. globosa culture free of virus and grazers was used. Viruses infecting
P. globosa (0.2 µm filtered PgV-07T lysate, Poretics filters, Millipore) were added (MOI
= 10) to the diluent directly after the 0.2 µm filtration step and to the P. globosa culture
just prior to dilution. The P. globosa culture that had to be diluted was infected just
before setting up the dilutions in the 2 L bottles. All bottles were incubated at the host
culture’s growth conditions (15ºC under light:dark cycle of 16:8 h at 100 µmol photons
m-2 s-1).
For both field and laboratory assay, the single cells of P. globosa were
enumerated directly upon sampling after gentle sieving through 50 µm mesh-size using a
Beckman Coulter XL-MCL flow cytometer (three replicates of each sample). The
apparent growth rate (µ in d-1) of P. globosa single cells was calculated for each sample
from the changes in abundance during the incubation according to the equation
µ = ln Nt24 - ln Nt0,
37
Chapter 2
Virally induced mortality of Phaeocystis globosa
where Nt0 and Nt24 are the abundance of P. globosa single cells at T=0 and T=24
respectively. A typical field example is presented Fig. 2B. The regression coefficient of
apparent growth rate vs. dilution factors for the 0.2 µm dilution series represents the
microzooplankton grazing rate (Mg), whereas the regression coefficient resulting from
the 30 kDa series represents both microzooplankton grazing and viral lysis (M(g+v)).
Subsequently, mortality rate due to viral lysis (Mv) was calculated as Mv = M(g+v) – Mg.
Mg, M(g+v) and their respective standard errors (SEg and SE(g+v)) were calculated using
Sigma plot software. The standard error of Mv was calculated as the squared root of the
sum of squared SEg and SE(g+v).
1 PREPARATION DILUENTS
PREPARATION SAMPLES
2
Natural sample
Reverse sieving 200 µm
Reverse sieving 200 µm
Sieving 50 µm
Sample
< 200 µm
3 µm filtration
0.2 µm filtration
20%
< 0.2 µm
40%
70%
30 kDa ultrafiltration
100%
retentate
20%
filtrate
< 30 kDa
40%
70%
100%
Figure 3. Experimental design of the viral lysis dilution assay (field assay). (1) Diluents were
prepared approx. 2 h prior to (2) dilution with natural sample collected de novo at high tide.
Samples were processed under dimmed light and at in situ temperature. Sample transfers were
performed by siphoning or gentle pumping, avoiding damage to the organisms.
38
Chapter 2
Virally induced mortality of Phaeocystis globosa
3. Results
3.1. Chemical parameters
For the two consecutive years of study a comparable pattern in nutrient
dynamics was recorded from Day 60 to 180 (Fig. 4). Nitrate concentration declined
steadily from 60-80 µM at Day 60 to 0.3 µM at Day 107 in 2003 and 0.8 µM at Day 173
in 2004. For 2003, the nitrate concentration remained low until Day 119, after which it
increased slightly again. In contrast, growth-limiting nitrate concentrations were not
found during the sampling period in 2004. Inorganic phosphate concentrations declined
sharply for both years, from 0.8-1 µM at Day 60 to around 0.4 µM during Days 90-95.
The concentration of phosphate in 2003 did not increase until Day 112. In 2004 it
increased quickly again (Day 99), but a second decline was detected from Days 117 to
125 (< 0.1 µM). Both years, ammonium concentrations ranged between 0.5 and 5.8 µM
until Day 120, after which it increased substantially to maximum values of 9 and 16 µM
in 2003 and 2004 respectively. Silicate concentrations decreased steadily from the
beginning of the sampling period until Day 83 in 2003 and Day 92 in 2004, after which
the concentration stayed low (0.4 to 6 µM) for both years.
2003
A
100
2004
B
PO4
PO4
NO3
NO3
0.8
80
0.6
60
0.4
40
0.2
20
0.0
Nitrate (µM)
Inorganic phosphate (µM)
1.0
0
60
80
100
120
140
160 180
Day number
60
80
100
120
140
160 180
Day number
Figure 4. Concentrations of dissolved inorganic phosphate and nitrate during spring (A) 2003 and
(B) 2004. Grey bars under the x-axis indicate the duration of the P. globosa bloom.
39
Chapter 2
Virally induced mortality of Phaeocystis globosa
The TEP concentration in 2004 (Fig. 5) increased steadily during the sampling
period to a maximum of 1033 µg Equiv. GX L-1 on Day 128, after which it declined to a
level comparable to the start of the sampling period (164 µg Equiv. GX L-1 on Day 160).
TEP (µg Equiv. Gum Xanthan L-1)
1200
1000
800
600
400
200
0
60
80
100
120
140
160
180
Day number
Figure 5. Concentration of TEP in µg Equivalent Gum Xanthan (GX) L-1 during the P. globosa
spring bloom in 2004 as determined by the semi-quantitative method (Passow & Alldredge 1995).
Grey bars under the x-axis indicate the duration of the P. globosa bloom in 2004.
3.2. Phaeocystis globosa bloom dynamics
The bloom of prymnesiophytes represented up to 70 % and up to 40 % of the
total phytoplankton chlorophyll in 2003 and 2004 respectively (Figs. 6A and 6C). P.
globosa was likely the dominant prymnesiophyte during the experimental period. The
magnitude and the composition of the blooms differed between years. The bloom in
2003 occurred between Days 70 and 141, reached at the highest total cell abundance of
7.6×104 cells mL-1, and was generally dominated by the colonial cell morph (Fig. 6B).
Single cells dominated over cells embedded in a colonial matrix only during peak events
(e.g. Day 102 and 119). In 2004, P. globosa bloom occurred between Days 92 and 159,
reached only three fold lower cell abundance (2.1×104 cells mL-1) as compared to 2003,
and was dominated by the single cell morphotype (Fig. 6D).
40
Chapter 2
% Prymnesiophyceae
100
2003
A
2004
C
80
60
40
20
0
8
P. globosa (x104 mL-1)
Virally induced mortality of Phaeocystis globosa
B
D
Total
Single
6
4
2
0
60
80
100
120
140
160
180
Day number
60
80
100
120
140
160
180
Day number
Figure 6. Phaeocystis globosa bloom dynamics during spring 2003 and 2004. (A, C) Relative
contribution of Prymnesiophyceae to the total phytoplankton community based on their pigment ratio to
Chl a (Prymnesiophyceae are mostly, if not only, represented by P. globosa during to the sampling
periods. (B, D) Abundance of P. globosa single cells and total cells (single and colonial cells). Grey bars
under the x-axis indicate the duration of the P. globosa bloom.
41
Chapter 2
Virally induced mortality of Phaeocystis globosa
3.3. Phaeocystis globosa specific viruses
Despite the low abundance of P. globosa at the beginning of the sampling
period, there was a substantial build-up of standing stock of putative PgVs both in 2003
and 2004 (Fig. 7). Putative PgV abundance increased concomitantly with the
development of P. globosa biomass, largely corresponding to increased P. globosa
single cells and/or reduced abundance of colonial P. globosa cells (Figs. 6B and 7A).
PgV reached abundances > 4×105 mL-1 for both years.
The numerical increase of the infective PgVs (MPN method using P. globosa
strain Pg-G) was delayed compared to the total putative PgV abundance as detected by
flow cytometry (Fig. 7). Although not visible in Fig. 7, infective PgVs were recorded in
each sample tested using P. globosa strain Pg-G (< 1.8×103 mL-1 at Days 97, 105 and
113 in 2003; and < 25 mL-1 from Days 60-120 in 2004). Maximum number of infective
PgV was roughly comparable for 2003 and 2004 with 1.5-1.8×104 infective PgV mL-1 at
Days 133 and 145 respectively. For both years, the infective PgV accounted maximum
for 5 % of the total putative PgV population.
From a previous study, the P. globosa strain Pg-01MD06 was found to be
specifically infected by only certain PgV isolates in culture (PgV Group II, Baudoux &
Brussaard 2005). Therefore, this P. globosa strain was also assayed during 2004 to
determine whether different PgVs coexisted in the field. Cell lysis of the host due to viral
infection was recorded for Day 113 and 123 (Fig. 7B).
B
A
Putative PgV
Infective PgV for Pg-G
Infective PgV for Pg-01MD06
5
20
15
4
3
10
2
5
1
0
Infective PgV (x103 mL-1)
Putative PgV (x105 mL-1)
6
0
60
80
100
120
140
160
180
Day number
60
80
100
120
140
160
180
Day number
Figure 7. Abundance of putative viruses infecting P. globosa (PgV) and infective PgV in (A) 2003
and (B) 2004. Total PgV (mL-1) was obtained using flow cytometry, whereas the abundance of
infective PgV (mL-1) resulted from end-point dilution (MPN) using Phaeocystis globosa strain Pg-G
(triangles) with a broad sensitivity to PgV and strain Pg-01MD06 with a narrow sensitivity to PgV
(circles). Only positive scores of virally induced cell lysis of P. globosa are presented. Grey bars
under the x-axis indicate the duration of the P. globosa bloom.
42
Chapter 2
Virally induced mortality of Phaeocystis globosa
3.4. Viral lysis and grazing
The rates of viral lysis and grazing by microzooplankton of P. globosa single
cells were estimated over the course of the blooms (Fig. 8). Virally mediated mortality
rates in 2003 were low (0.01 to 0.03 d-1) until the collapse of P. globosa single cells,
when viral lysis was high (0.35 d-1 at Day 132, Fig. 8A). This enhanced viral lysis rate
concurred with the highest concentrations of total putative PgV as well as infective PgVs
(Figs. 7A and 8A). In contrast to viral lysis, microzooplankton grazing of P. globosa
single cells was shown to be an important source of mortality during the entire bloom of
P. globosa (Fig. 8A). During the course of the bloom, viral lysis accounted for 5 to 57 %
of the total mortality, assuming no losses of P. globosa single cells due to sinking.
Viral lysis rates were higher over the course of the bloom in 2004 (Fig. 8B, with
values of 0.29, 0.19 and 0.24 d-1 at Days 123, 127 and 135 respectively) as compared to
2003 when viral lysis only increased at the bloom termination (0.35 d-1 at Day 132).
These high P. globosa-specific viral lysis rates also coincided with increased abundances
of total putative and infective PgV. Microzooplankton grazing in 2004 was another
important loss factor for single cells of P. globosa, with rates ranging between 0.05 and
0.40 d-1.Viral lysis represented 44, 66 and 45 % of total losses at Days 123, 127 and 135,
respectively.
0.6
2003
A
Viral lysis
Grazing
Mortality rates (d-1)
0.5
B
2004
0.4
0.3
0.2
0.1
0.0
86
100
104
118
132
Day number
116
123
127
136
Day number
Figure 8. Viral lysis and microzooplankton grazing rates of Phaeocystis globosa single cells during
different stages of the P. globosa bloom in (A) 2003 and (B) 2004. Viral lysis (d-1) is represented by
the grey bars, and grazing (d-1) by the white bars. Error bars correspond to standard error.
43
Chapter 2
Virally induced mortality of Phaeocystis globosa
The total P. globosa specific cell lysis varied over the course of the bloom for
both years (Fig. 9). In 2003, total P. globosa cell lysis rates were < 0.1 d-1 until Day 121
and increased concomitantly with the collapse of the total P. globosa cell abundance (0.2
d-1 at Day 132). In 2004, total P. globosa cell lysis rates increased earlier (from Day
110). For both years total cell lysis rates reached maximum rates of about 0.2 d-1.
0.4
2003
A
0.3
Lysis rates (d-1)
B
2004
Total lysis
Viral lysis
0.2
0.1
0.0
60
80
100
120
140
160
180
Day number
60
80
100
120
140
160
180
Day number
Figure 9. Daily viral lysis and total cell lysis rates of P. globosa during the spring bloom in (A) 2003
and (B) 2004. Grey bars under the x-axis indicate the duration of the P. globosa bloom
4. Discussion
The present study revealed that virally mediated mortality, next to grazing, was
a major source of loss for P. globosa cells during two consecutive spring blooms (in
2003 and 2004). Incidentally, viral lysis rates recorded during the blooms were higher
than the microzooplankton grazing rates upon P. globosa cells. To our knowledge, this
study is the first direct assessment of viral lysis rates in P. globosa under natural
conditions. A recent study indicated that viruses can be significant mortality agents of P.
globosa cells, but the experiment was conducted under controlled conditions in
mesocosms and estimates were based on virus production and assumed burst size
(Brussaard et al. 2005a).
The Landry and Hassett dilution method (Landry & Hassett 1982) was
originally developed for the measurement of microzooplankton grazing and is now
routinely applied in a broad range of aquatic environments (Landry & Calbet 2004).
Recently, an extended version of the dilution method was successfully developed to
44
Chapter 2
Virally induced mortality of Phaeocystis globosa
specifically estimate viral lysis of the picophytoplankter Micromonas pusilla (Evans et
al. 2003). The suitability of the method was demonstrated for Phaeocystis using a
cultured P. globosa host and virus model system. Thereby, this test experiment validated
the two critical assumptions of the original dilution method; (1) that phytoplankton
growth rate is independent of the dilution factor, and (2) losses are proportional to the
dilution effect on the abundance of the predators (Landry & Hassett 1982, Landry et al.
1995). The obtained viral lysis rate (1.3 d-1) using this assay was comparable from that
of independent one-step lytic cycle experiment (1.4 d-1). In order to estimate the number
of P. globosa cells that underwent cell lysis during a 24 h period, the PgV produced in
the non-diluted samples during the 24 h incubation were divided by a theoretical burst
size ranging from 100-300 (Baudoux & Brussaard 2005). These values compared very
well with those obtained by multiplying the cell abundance at the start of the incubation
by the determined viral lysis rate, assuming no growth of the infected algal cells.
It is imperative to realize that this assay exclusively detects viral lysis of algal
hosts that are newly infected within the 24 hours of incubation. This incubation period is
essential to encompass the entire cellular diel cycle of phytoplankton that have
synchronized cell cycles. The lytic nature of all known PgVs, as well as the 14 to 16 h
required for P. globosa to undergo cell lysis (Baudoux & Brussaard, 2005) favor the
detection of P. globosa viral lysis using this dilution approach. However, the time
between successful viral infection of a P. globosa cell and its subsequent lysis is critical
since late infection during the incubation period (later than 10 h) will no longer result in
cell lysis within the duration of the incubation. The potential impact of viruses on the P.
globosa population is, therefore, likely to be underestimated. The lysis of P. globosa
cells happens 0-4 h after the first release of viral progeny of PgV (latent period of 10-16
h, Baudoux & Brussaard 2005). Thereby, a second round of infection by the newly
produced viruses during the incubation period should not affect the viral lysis rates since
the dilution method is governed by the enumeration of cells.
Viral lysis rates were compared with the P. globosa total cell lysis rates, which
also includes lysis due to causes other than viral infection (e.g. environmental stress).
This comparison reveals that virally induced mortality was the most important cause of
lysis and thereby supports the findings of P. globosa studies conducted in mesocosms
(Brussaard et al. 2005a). Differences between total cell lysis and viral lysis rates of P.
globosa likely represent other forms of cell lysis, e.g. automortality of the colonial cells
from nutrient depletion (Ruardij et al. 2005). Deviations due to the use of two different
parameters to estimate viral lysis and total cell lysis can, of course, not be excluded.
Viral lysis was determined from the difference in P. globosa cell abundance over 24
hours of incubation period using flow cytometry, whereas total cell lysis was estimated
from the ratio of dissolved to particulate esterase activity. It may also be possible that the
physiological status of cells influence the cellular esterase activity and thus the total lysis
rates. Reduced percentages of dying cells were recorded for virally infected P. pouchetii
cells in the early stationary growth phase upon staining with Calcein-AM, a fluorescent
dye revealing intracellular esterase activity (Brussaard et al. 2001). We speculate that the
lower total cell lysis rates than viral lysis rates at the end of the blooms might be
influenced by such methodological variations.
45
Chapter 2
Virally induced mortality of Phaeocystis globosa
The composition and magnitude of the P. globosa blooms differed for 2003 and
2004. The lower biomass of P. globosa colonies in 2003 and 2004 was likely due to the
lower standing stock of inorganic phosphate in 2004 which was half the concentration of
2003. One possible reason could be that silicate became growth-limiting later in 2004 as
compared to 2003 (Day 92 in 2004 as compared to Day 83 in 2003). Thus, dominance of
diatoms was prolonged and subsequently the concentration of phosphate was reduced
when P. globosa biomass finally developed (Jahnke 1989, Egge & Aksnes 1992).
Another reason might be that the reduced mean water irradiance during the bloom period
in 2004 (due to substantial cloud cover), limited colony formation (Peperzak 1993) and
thus lowered P. globosa biomass.
Virally induced mortality of P. globosa cells as well as PgV abundance
increased concomitantly with the development of the bloom, as can be expected since an
increasing abundance of host enhances the rate of successful viral infection. Despite the
higher P. globosa total cell abundance in 2003, the maximal abundance of PgV was
comparable in both years. It is suggested that P. globosa cells that are embedded inside a
colonial matrix are protected against viral infection (Brussaard et al. 2005a, Ruardij et al.
2005). Interestingly, our results show that the abundance of colonial cells was higher in
2003 than in 2004 but the abundance of single cells was comparable. This observation
suggests that PgV was mainly produced by P. globosa single cells and thus corroborates
the observations that colonies provide protection from viral infection. Hence, the
morphotype composition of a P. globosa bloom is an important factor underlying the
impact of viral infection for P. globosa.
When colonies disintegrate after experiencing nutrient depletion or light
deprivation (Veldhuis et al. 1986, Peperzak 1993), cells are released in the surrounding
waters and become readily infected (Brussaard et al. 2005a, Ruardij et al. 2005). In
2003, nitrate depletion (from Day 107 to 120) was most likely responsible for the major
collapse in the abundance of colonial cells after Day 115, thus increasing the impact of
viral lysis (0.35 d-1 at the end of the bloom). Although the cause of colonial
disintegration in 2004 is less obvious, the decline in colony abundance led to an increase
in single cell abundance and enhanced viral lysis (0.29 d-1 on Day 123).
The difference in colony abundance between the two years also likely
influenced the dynamics of transparent exopolymeric particles (TEP), which are
produced in high concentrations during colony disruption (Mari et al. 2005). TEP
formation has recently been acknowledged as a major inhibitor for viral infection
(Brussaard et al. 2005b, Ruardij et al. 2005). Viruses, like other microorganisms, tend to
passively adsorb to TEP and therefore are not available to infect algal cells. In 2004, the
period with the highest TEP concentration indeed coincided with a low abundance of
infective PgV and reduced viral lysis rates. Interestingly, the disintegration of TEP
concurred with an enhanced abundance of infective PgV. The concentrations of TEP
recorded for 2004 were lower than those reported during a P. globosa bloom with higher
colonial cell abundance (100 to 1000 µg equiv GX L-1 as compared to 100 to 2000 µg
GX L-1; Mari et al. 2005). Thus, the higher abundance of colonies in 2003 induced, in all
likelihood, a higher release of TEP and subsequently an enhanced impact of TEP on the
46
Chapter 2
Virally induced mortality of Phaeocystis globosa
fraction PgV still infective, which in turn resulted in reduced viral lysis rates. Therefore
it can be said that colonies play a controversial role acting as a potential viral lysis
inhibitor with TEP production but also as an enhancing agent of virally induced
mortality as they constitute a potential reservoir of single cells.
The present study, furthermore, indicates towards regulation of viral infection
on a finer scale, as the results from the end-point dilution assay (MPN) suggest the coexistence of different PgVs. The PgV population infecting Pg-01MD06 appeared and
disappeared within 2 weeks, implying that the PgV population in the field is diverse and
dynamic. This has also been proposed by other authors for different algal viruses
(Tarutani et al. 2000, Schroeder et al. 2003). The reason this PgV population is not
maintained in the water column is not clear. Possible reasons may be the removal of the
specific host from the water column or a loss of infectivity of this specific viral
population, but this needs further investigation.
In summary, this study shows that viral lysis and grazing by microzooplankton
are both major controlling agents for P. globosa single cells, although the relative
significance of each of these factors can vary during the course of the bloom and
between years. The application of the dilution method based on cell counts in
combination with total algal cell lysis rates and abundance of infective algal viruses
provides essential insight into the quantitative significance of viral lysis as compared to
other loss factors. The present study also gives insight into on the ecological role of viral
infection in relation to host population regulation, and some of the mechanisms
controlling successful infection.
Acknowledgments. We thank G. van Noort and C. Chenard for technical assistance, H.
Witte for statistical advice and T. Compton and C. Robertson for editing this manuscript.
Thanks to L. Peperzak for sharing data and to the anonymous reviewers for their
suggestions on the manuscript. Special thanks to B. Kuipers for general discussion and
to the editor for valuable comments on the manuscript. This work was supported by the
Research Council for Earth and Life Sciences (ALW) with financial aid from the
Netherlands Organization for Scientific Research (NWO).
Literature cited
Agusti S, Satta MP, Mura MP, Benavent E (1998) Dissolved esterase activity as a tracer
of phytoplankton lysis: Evidence of high phytoplankton lysis rates in the
northwestern Mediterranean. Limnol Oceanogr 43:1836-1849
Baudoux A-C, Brussaard CPD (2005) Characterization of different viruses infecting the
marine harmful algal bloom species Phaeocystis globosa. Virology 341:80-90
Bratbak G, Egge JK, Heldal M (1993) Viral mortality of the marine alga Emiliania
huxleyi (Haptophyceae) and termination of algal bloom. Mar Ecol Prog Ser 93:39-
47
Chapter 2
Virally induced mortality of Phaeocystis globosa
48
Brussaard CPD (2004a) Optimization of procedures for counting viruses by flow
cytometry. Appl Environ Microbiol 70:1506-1513
Brussaard CPD (2004b) Viral control of phytoplankton populations-a Review. J Euk
Microbiol 51:125-138
Brussaard CPD, Gast GJ, Van Duyl FC, Riegman R (1996a) Impact of phytoplankton
bloom magnitude on a pelagic microbial food web. Mar Ecol Prog Ser 144:211221
Brussaard CPD, Kempers RS, Kop AJ, Riegman R, Heldal M (1996b) Virus-like
particles in a summer bloom of Emiliania huxleyii in the North Sea. Aquat Microb
Ecol 10:105-113
Brussaard CPD, Kuipers B, Veldhuis MJW (2005a) A mesocosm study of Phaeocystis
globosa population dynamics I. Regulatory role of viruses in bloom control.
Harmful Algae 4:859-874
Brussaard CPD, Mari X, Van Bleijswijk JDL, Veldhuis MJW (2005b) A mesocosm
study of Phaeocystis globosa (Prymnesiophyceae) population dynamics - II.
Significance for the microbial community. Harmful Algae 4:875-893
Brussaard CPD, Marie D, Thyrhaug R, Bratbak G (2001) Flow cytometric analysis of
phytoplankton viability following viral infection. Aquat Microb Ecol 26:157-166
Brussaard CPD, Riegman R, Noordeloos AAM, Cadee GC, Witte H, Kop AJ,
Nieuwland G, Van Duyl FC, Bak RPM (1995) Effects of grazing, sedimentation
and phytoplankton cell lysis on the structure of a coastal pelagic food web. Mar
Ecol Prog Ser 123:259-271
Brussaard CPD, Short SM, Frederickson CM, Suttle CA (2004) Isolation and
phylogenetic analysis of novel viruses infecting the phytoplankton Phaeocystis
globosa (Prymnesiophyceae). Appl Environ Microbiol 70:3700-3705
Cottrell MT, Suttle CA (1991) Wide-spread occurrence and clonal variation in viruses
which cause lysis of a cosmopolitan eukaryotic marine phytoplankter,
Micromonas pusilla. Mar Ecol Prog Ser 78:1-9
Egge JK, Aksnes DL (1992) Silicate as regulating nutrient in phytoplankton competition.
Mar Ecol Prog Ser 83:281-289
Evans C, Archer SD, Jacquet S, Wilson WH (2003) Direct estimates of the contribution
of viral lysis and microzooplankton grazing to the decline of a Micromonas spp.
population. Aquat Microb Ecol 30:207-219
Grasshoff K (1983) Determination of nutrients. In: Grasshoff K, Ehrhardt M, Kremling
M (eds) Methods of seawateranalysis. Verlag Chemie, Weinheim, p 143–150
Guillard RRL (1975) Culture of phytoplankton for feeding marine invertebrates. In:
Chanley, MH (ed.) Culture of marine invertebrate animals. Plenum Press, New
York, p 29-60
Harrison PJ, Waters RE, Taylor FJR (1980) A broad spectrum artifical seawater medium
for coastal and open ocean phytoplankton. J Phycol 16:28-35
Helder W, De Vries R (1979) An automatic phenol-hypochlorite method for the
determination of ammonia in sea- and brackish waters. Neth J Sea Res 13:154160
48
Chapter 2
Virally induced mortality of Phaeocystis globosa
Hurley MA, Roscoe ME (1983) Automated statistical analysis of microbial enumeration
by dilution series. J Appl Bacteriol 55:159-164
Jahnke J (1989) The light and temperature-dependence of growth rate and elemental
composition of Phaeocystis globosa Scherffel and Phaeocystis pouchetii (Har)
Lagerh in Batch Cultures. Neth J Sea Res 23:15-21
Lancelot C, Billen G (1984) Activity of heterotrophic bacteria and its coupling to
primary production during the spring phytoplankton bloom in the southern bight
of the North Sea. Limnol Oceanogr 29:721-730
Landry MR, Calbet A (2004) Microzooplankton production in the oceans. ICES J Mar
Sci 61:501-507
Landry MR, Hassett RP (1982) Estimating the grazing impact of marine
microzooplankton. Mar Biol 67:283-288
Landry MR, Kirshtein J, Constantinou J (1995) A refined dilution technique for
measuring the community grazing impact of microzooplankton, with
experimental tests in the central equatorial Pacific. Mar Ecol Prog Ser 120:53-63
Mackey MD, Mackey DJ, Higgins HW, Wright SW (1996) CHEMTAX- a program for
estimating class abundances from chemical markers: application to HPLC
measurements of phytoplankton. Mar Ecol Prog Ser 144:265-283
Mari X, Rassoulzadegan F, Brussaard CPD, Wassmann P (2005) Dynamics of
transparent exopolymeric particles (TEP) production by Phaeocystis globosa
under N- or P-limitation: a controlling factor of the retention/export balance.
Harmful Algae 4:895-914
Murphy J, Riley JP (1962) A modified single solution method for the determination of
phosphate in natural waters. Anal Chim Acta 27:31-36
Passow U, Alldredge AL (1995) A dye-binding assay for the spectrophotometric
measurement of transparent exopolymer particles (TEP). Limnol Oceanogr
40:1326-1335
Peperzak L (1993) Daily irradiance governs growth-rate and colony formation of
Phaeocystis (Prymnesiophyceae). J Plankton Res 15:809-821
Riegman R, Kraay GW (2001) 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. J Plankton Res 23:191-205
Riegman R, van Bleijswijk JDL, Brussaard CPD (2002) The use of dissolved esterase
activity as a tracer of phytoplankton lysis - Comment. Limnol Oceanogr 47:916920
Ruardij P, Veldhuis MJW, Brussaard CPD (2005) Modeling the bloom dynamics of the
polymorphic phytoplankter Phaeocystis globosa: impact of grazers and viruses.
Harmful Algae 4:941-963
Schroeder DC, Oke J, Hall M, Malin G, Wilson WH (2003) Virus succession observed
during an Emiliania huxleyii bloom. Appl Env Microbiol 69:2484-2490
Stefels J, Van Boekel WHM (1993) Production of DMS from dissolved DMSP in axenic
cultures of the marine phytoplankton species Phaeocystis sp. Mar Ecol Prog Ser
97:11-18
49
Chapter 2
Virally induced mortality of Phaeocystis globosa
Strickland JDH, Parsons TR (1968) A practical handbook of seawater analysis. In:
Fisheries Research Board of Canada, Bull 167. Canadian Government Publishing
center, Ottawa, Canada, p 310
Suttle CA (1993) Enumeration and isolation of viruses. In: Current methods in aquatic
microbial ecology, P.F. Kemp, B.F. Sherr, E.F. Sherr and J.J Cole (eds), Lewis
Publ., Boca Raton, p. 121-134
Suttle CA, Chan AM, Cottrel MT (1991) Use of ultrafiltration to isolate viruses from
seawater which are pathogens of marine plankton. Appl Environ Microbiol
57:721-726
Tarutani K, Nagasaki K, Yamaguchi M (2000) Viral impacts on total abundance and
clonal composition of the Harmful bloom-forming phytoplankton Heterosigma
akashiwo. Appl Environ Microbiol 66:4916-4920
Tomaru Y, Tarutani K, Yamaguchi M, Nagasaki K (2004) Quantitative and qualitative
impacts of viral infection on a Heterosigma akashiwo (Raphidophyceae) bloom in
Hiroshima Bay, Japan. Aquat Microb Ecol 34:227-238
Van Boekel WHM, Hansen FC, Riegman R, Bak RPM (1992) Lysis induced decline of a
Phaeocystis spring bloom and coupling with the microbial food web. Mar Ecol
Prog Ser 81:269-276
Veldhuis MJW, Admiraal W, Colijn F (1986) Chemical and physiological-changes of
phytoplankton during the spring bloom, dominated by Phaeocystis pouchetii
(Haptophyceae) - Observations in dutch coastal waters of the North-Sea. Neth J
Sea Res 20:49-60
50
Chapter 3
Characterization of different viruses
infecting the marine harmful algal
bloom species Phaeocystis globosa 1
Anne-Claire Baudoux and Corina P. D. Brussaard
Twelve lytic viruses (PgV) infecting the marine unicellular eukaryotic harmful algal
bloom species Phaeocystis globosa were isolated from the southern North Sea in 20002001 and partially characterized. All PgV isolates shared common phenotypic features
with other algal viruses belonging to the family Phycodnaviridae, and could be
categorized in four different groups. Two main groups (PgV Group I and II) were
discriminated based on particle size (150 and 100 nm respectively), genome size (466
and 177 kb), and structural protein composition. The lytic cycle showed a latent period
of 10 h for PgV Group I, and latent periods of 12 h and 16 h for PgV Group IIA and IIB.
Host specificity and temperature sensitivity finally defined a fourth group (PgV Group
IIC). Our results imply that viral infection plays an important role not only in P. globosa
dynamics but also in the diversity of both host and virus community.
1. Introduction
Over nearly two decades, studies have underlined the numerical dominance,
ubiquity and worldwide distribution of viruses in marine pelagic environments
(Wommack & Colwell 2000). Algal viruses are diverse and dynamic within the
microbial community (Cottrell & Suttle 1991, Chen et al. 1996, 1995, Short & Suttle
2003, Schroeder et al. 2003, Brussaard et al. 2004b, Nagasaki et al. 2004, Tomaru et al.
2004). Most of the existing classes of photosynthetic eukaryotic unicellular algae
1
Published in Virology 341:80 – 90, 2005
51
Chapter 3
Viruses infecting Phaeocystis globosa
(phytoplankton) have been reported as hosts for virus-like particles (Van Etten et al.
1991, Reisser 1993, Brussaard 2004a). Previous studies indicate that algal viruses are
relevant mortality agents in marine environments, directly controlling the dynamics of
their host population (Bratbak et al. 1993, Brussaard et al. 1996b, Evans et al. 2003,
Tomaru et al. 2004, Brussaard et al. 2005a, Ruardij et al. 2005). Viral lysis of
phytoplankton indirectly affects the structure and functioning of the microbial foodwebs, especially when it involves bloom-forming algae (Fuhrman 1999, Wilhelm &
Suttle 1999, Castberg et al. 2001, Brussaard et al. 2005b).
An important bloom-forming phytoplankter is the globally distributed genus
Phaeocystis (Prymnesiophyceae). Phaeocystis has a polymorphic life cycle with
flagellated unicellular and non-motile cells that are embedded in colonies. Phaeocystis
blooms draw down atmospheric CO2 as well as produce dimethylsulphide (DMS), which
is involved in cloud formation (Stefels & Van Boekel 1993, Arrigo et al. 1999, DiTullio
et al. 2000). Thus, Phaeocystis is acknowledged as a microalga playing an important role
in global climate regulation, and is argued to be a key genus influencing the structure
and function of marine pelagic environments (Verity & Smetacek 1996).
Phaeocystis globosa has the potential to generate high biomass blooms in spring
occurring in the temperate waters of the southern North Sea (Lancelot et al. 1987, Cadée
& Hegeman 1991). Termination of these blooms can cause excessive production of
foam, which becomes a nuisance for socio-economical activities like tourism and
fisheries (Orton 1923, Pieters et al. 1980). Phaeocystis globosa is therefore considered a
harmful algal bloom species (HAB).
Cell lysis has been found to be an important loss factor for P. globosa cells and
can account for 75% of the decline of the bloom (Van Boekel et al. 1992, Brussaard et
al. 1995, 1996a). Brussaard and co-workers (2004b) observed that the decline of a
natural bloom was accompanied by a considerable increase of putative viruses infecting
P. globosa (PgVs), suggesting that viruses were a significant source of mortality for this
alga. Very recently, a mesocosm study showed that P. globosa population dynamics can
indeed be virally controlled (Brussaard et al. 2005a). However, to elucidate the
ecological implications of viral infection for P. globosa dynamics, more detailed
knowledge on the interactions between virus and algal host cell and the characteristics of
the viruses is needed.
In this study, twelve lytic viruses infecting P. globosa (PgVs) are partially
characterized. From these twelve PgVs, four groups were distinguished based on their
phenotype (morphology, genome size, structural proteins, latent period, host range and
temperature sensitivity).
52
Chapter 3
Viruses infecting Phaeocystis globosa
2. Material and methods
Algal cultures and growth conditions. The different unialgal P. globosa strains
and species representatives of different taxonomic groups (not axenic), used for virus
isolation and host specificity testing, are listed in Table 1. All algal species, except P.
pouchetii AJ01 and all Dinophyceae, were cultured in ESF2 medium, a 1:1 mixture of
f/2 medium (Guillard 1975) and enriched artificial seawater (ESAW, Harrison et al.
1980, Cottrell & Suttle 1991,) with the addition of Tris-HCl and Na2SeO3 (Cottrell &
Suttle 1991). The Bacillariophyceae were grown in ESF2 medium completed with
silicate (150 µM), P. pouchetii AJ01 was cultured in IMR1/2 medium (Eppley et al
1967), and the Dinophyceae were cultured in a medium specifically for dinoflagellates
(Hansen 1989). All cultures, except P. pouchetii AJ01, were incubated under a light:dark
cycle of 16:8 h at 100 µmol photons m-2s-1. P. pouchetii AJ01 was grown under a
light:dark cycle of 14:10 h at 40-50 µmol photons m-2 s-1. All cultures were grown at
15°C, except P. pouchetii AJ01, P. pouchetii Pp-13 and P. antarctica CCMP1871 which
were grown at 8°C and 2°C, respectively.
Virus isolation. Lytic viruses infecting P. globosa (PgV) were isolated from
natural seawater originating from the southern North Sea in 2000 (June-October) and
2001 (April) according to the procedure described in Brussaard et al. (2004b). Briefly,
filtered (Whatman GF/F) natural seawater was added to P. globosa cultures (10 – 20 %,
v/v) and incubated for 10 days at standard culture conditions of the host. Different host
strains of P. globosa were used for virus isolation in order to maximize the chance of
successful virus isolation. Algal growth was monitored via in vivo Chlorophyll
fluorescence (F0) using a Turner Designs fluorometer. The cultures that showed signs of
lysis as compared to noninfected controls were filtered through 0.2 µm pore-size
cellulose acetate filters (Schleicher and Schuell GmbH, Dassel, Germany), afterwhich
the lysate was used to reinfect an exponentially growing algal host culture. After
recurrent lysis and reinfection, viral isolates were made clonal by end-point dilution as
described by Brussaard and co-workers (2004b).
Transmission electron microscopy. The presence of virus-like particles was
confirmed for all virus isolates using transmission electron microscopy (TEM). Infected
algal cells (10 - 15 h after infection) were fixed with glutaraldehyde (0.1 % final
concentration, EM grade, Darmstadt, Germany) for 2 h on ice. Fixed cells were
harvested by low speed centrifugation (3,200 x g, 5 minutes, with a A-4-62 swing-out
rotor and using a 5810R centrifuge, Eppendorf), afterwhich pellets were resuspended in
6 % glutaraldehyde (final concentration) prepared in 0.1 M cacodylate buffer and
completed with 5 mM MgCl2 and 5 mM CaCl2 (pH 7.2, all products purchased at Sigma
Aldrich). Samples were kept on ice for 2 h and centrifuged as described above. Pelleted
cells were resuspended in 0.1 % glutaraldehyde (final concentration) and stored at 4°C
until postfixation. Prior to postfixation, cells were harvested, transferred to 1.5 ml
microtubes and washed twice in 0.1 M cacodylate buffer (pH 7.2) using a
53
Chapter 3
Viruses infecting Phaeocystis globosa
Table 1. List of phytoplankton species used to screen for virus-induced lysis by PgV strains in the
host range tests
Genus / Species
Strains
Prymnesiophyceae
Ph91mfa
Pg-G (A)b
Pg-Ib
Pg01MD-02c
Pg01MD-06c
SK 35d
Unknownd
Ph91hca
Pg-G (B)b
Ph Millera
Pg01MD-04c
Pg Kac 31e
Pp Kac 75e
AJ01f
Pp-13f
CCMP1871g
Phaeonap1h
B5h
Unknowni
CCMP 1323j
Unknownk
Unknownl
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis globosa
Phaeocystis pouchetii
Phaeocystis pouchetii
Phaeocystis pouchetii
Phaeocystis antarctica
Phaeocystis cordata
Phaeocystis jahnii
Emiliania huxleyi
Isochrysis galbana
Pavlova lutheri
Chrysochromulina polylepis
Bacillariophyceae
CCMP 469j
Unknownl
Cs-T01c
CCMP 1049j
CCMP 358j
Leptocylindrus danicus
Asterionellopsis glacialis
Chaetoceros socialis
Thalassiosira weissflogii
Ditylum brightwellii
Chlorophyceae
Unknownc
CCAP 251/2m
Dunaliella sp.
Nannochloris sp.
Prasinophyceae
Unknownc
CCMP 1192j
Tetraselmis sp.
Prasinococcus capsulatus
Cryptophyceae
CCMP 1319 j
Rhodomonas salina
Eustigmatophyceae
CCAP 849/4 m
Nannochloropsis salina
Dinophyceae
CCMP 1589j
Unknownc
Unknownn
Unknownc
Prorocentrum micans
Scrippsiella sp.
Amphidinium sp.
Gymnodinium simplex
Cyanophyceae
CCMP 839j
CCMP 1334j
Synechococcus sp.
Synechococcus sp.
OSD-RIKZ, The Netherlands, b Culture collection University of Groningen, The Netherlands, c Culture
collection of the Netherlands Institute for Sea Research, The Netherlands, d Alfred Wegener Institute,
Bremerhaven, Germany, e University of Kalmar, Sweden, f University of Bergen, Norway, g University
Libre of Bruxelles, Belgium, h Stazione Zoologica Anton Dohrn, Naples, Italy, i University of Leiden,
The Netherlands, j Provasoli-Guillard National Center for Culture of Marine Phytoplankton, Maine,
USA, k University Bergen, Norway, l University of Oldenburg, Germany, m Culture Collection of Algae
and Protozoa, Scotland, UK, n University of Copenhagen, Helsingør, Denmark.
a
54
Chapter 3
Viruses infecting Phaeocystis globosa
microcentrifuge (6,000 x g, 5 minutes, microcentrifuge model 5415C, Eppendorf).
Samples were carefully postfixed with 1% osmium tetroxide (Sigma Aldrich) prepared
in 0.1 M cacodylate buffer (1 - 2 h), afterwhich they were washed 3 times with 0.1 M
cacodylate buffer (pH 7.2). The postfixed cells were dehydrated in an ascending ethanol
series (from 70 to 100% absolute ethanol v/v, Fluka), and washed twice with propylene
oxide (Agar Scientific, Essex, UK). The supernatant was removed and the pellets were
infiltrated into a 1:1 mixture propylene oxide:agar resins (mixture of 12.6 g MNA, 12.6
g DDSA, 24.8 g agar resin and 0.5 g DMP, Agar Scientific, Essex, UK). The samples
were left with lid open overnight and placed at 60°C for 48 hours for polymerization of
the resin. Once the resins solidified, plugs were thin-sectioned using a Reichert
ultramicrotome. The thin sections were post-stained with 2% uranyl acetate and leadcitrate (Reynolds) before examination under a 100 CX transmission electron microscope
(final magnification from x 33500 to x 52000, JEOL, Tokyo). At least 10 viral-like
particles from each isolate were measured to estimate average particle diameter.
Virus growth cycle. An unialgal culture of P. globosa Pg-I was used to
determine the one-step virus growth cycle for each PgV isolate. This strain was chosen
because of its sensitivity to all the viral isolates studied. Exponentially growing P.
globosa cells (250 mL) were infected with a freshly produced PgV lysate at an initial
virus to host ratio of 20. In case of doubt (for example when finding deviating burst size)
the lytic growth experiment was repeated in order to confirm the results. Most probable
number examination of the viral lysates showed that the multiplicity of infection (MOI)
of the different viral isolates ranged between 13 and 20. Noninfected control cultures of
P. globosa received equal volume of medium. The samples were incubated at the host
culture standard conditions and sampled every 4 hours for a total period of 50 hours.
Algal and viral abundances were monitored by flow cytometry (FCM, Beckton
Dickinson FACScalibur equipped with a 15-mW, 488-nm air cooled argon-ion laser).
Algal samples were analyzed directly upon sampling whereas virus samples were fixed
with 25% glutaraldehyde (0.5% final concentration, EM grade, Merck) during 30
minutes at 4°C, followed by freezing in liquid nitrogen and storage at -80°C. Analysis of
the virus samples was performed using flow cytometry after dilution in TE and staining
with the nucleic acid-specific dye SYBR Green I (Molecular Probes, Eugene, OR)
according to Brussaard (2004b).
Host range. The host specificity of all PgV isolates was tested using a broad
range of phytoplankton species, including 12 different strains of P. globosa (see Table
1). Freshly produced PgV lysate was added to exponentially growing algal cultures (20
% v/v). The natural in vivo fluorescence of the cultures was monitored every 2 days for
10 days at standard culture conditions. Cultures that did not show signs of lysis as
compare to noninfected control cultures were considered resistant to the virus tested.
Cultures that underwent lysis were inspected for viral proliferation using flow cytometry.
Genome size and nature. For all virus isolates freshly produced viral lysate was
clarified of bacteria and cell debris by low speed centrifugation step (7,500 x g, 30
55
Chapter 3
Viruses infecting Phaeocystis globosa
minutes at 4°C with fixed angle rotor F-34-6-38, and using a 5810R centrifuge,
Eppendorf). Supernatant was concentrated by ultracentrifugation (141,000 x g, 2 h at
8°C, with a rotor TFF55.38 and using a Centrikon T-1080 ultracentrifuge, Kontron
Instruments). The viral pellets were resuspended in 150 µL of SM buffer (0.1 M NaCl, 8
mM MgSO4.7H2O, 50 mM Tris-HCl, 0.0005 % (w/v) glycerin, Wommack et al. 1999)
and stored at 4°C overnight. Equal volumes of virus concentrate and molten 1.5 % (w/v)
InCert agarose (Cambrex Bioscience, Rockland, ME USA) were dispensed into plug
moulds, and left to solidify for 3 minutes at -20°C. The plugs were then punched out of
the mould into microtubes containing 800 µL of lysis buffer (250 mM EDTA, 1 % SDS
(v/v), 1 mg mL-1 proteinase K, all products were purchased at Sigma-Aldrich) and
incubated overnight at 30°C. Next day, the digestion buffer was decanted and the plugs
were washed 4 times for 30 min each in TE 10:1 buffer (10 mM Tris-Base, 1 mM
EDTA, pH 8.0). Virus-agarose plugs were stored at 4°C in TE 20:50 (20 mM Tris, 50
mM EDTA, pH 8.0) until analysis.
To determine the nature of the viral isolates, virus agarose-plugs previously
prepared were treated with DNase RQ1 RNase-Free DNase (Promega) during 1 h at
37°C. Plugged samples and Lambda concatamers plugs (Bio-Rad, Richmond, CA) were
loaded onto a 1% SeaKem GTG agarose gel (Cambrex Bioscience, Rockland, ME USA)
prepared in 1× TBE gel buffer (90 mM Tris-Borate and 1 mM EDTA, pH 8.0). Wells of
the gel were overlaid with 1 % molten agarose and the gel was placed in the
electrophoretic cell containing 0.5× TBE tank buffer (45 mM Tris-Borate and 0.5 mM
EDTA, pH 8.0). Samples were electrophored using a Bio-Rad DR-II CHEF Cell unit
operating at 6 V cm-1 with pulse ramps of 20 to 45 s at 14°C during 22 h. After
electrophoresis, gels were stained for 1 h with SYBR Green I (1×10-4 of commercial
solution, Molecular Probes, Eugene, OR) and destained 10 minutes in MilliQ (Gradient
A10, Millipore) before a digital analysis for fluorescence using a FluorS imager (BioRad Instrument).
Protein characterization analysis. A 5 L freshly produced lysate was
concentrated using a 30 kDa MWCO ultrafiltration (Vivaflow 200, Vivascience). The
virus concentrate was clarified of bacteria and cell debris by low speed centrifugation
(7,500×g , 30 minutes at 4°C, with a fixed angle rotor type F-34-6-38, using a 5810R
centrifuge, Eppendorf) and further harvested by ultracentrifugation (141,000×g, 2 h at
8°C, with a TFT 55.38 rotor using a Centrikon T-1080 ultracentrifuge and, Kontron
Instruments). The pellets were resuspended in 150 µL SM buffer (0.1 M NaCl, 8 mM
MgSO4.7H2O, 50 mM Tris-HCl, 0.0005 % (w/v) glycerin, Wommack et al. 1999).
Viruses were purified on a 1.40 or 1.45 g mL-1 Cesium Chloride gradient (Molecular
Biology grade, Sigma-Aldrich). Samples were ultracentrifuged (111,000×g, 72 h at 8°C
with a SW41Ti swing out rotor, Beckman and using a Centrikon T-1080 ultracentrifuge,
Krontron Instrument). The visible viral bands were extracted, washed twice with PBS
(pH 8) using a 30kDa MWCO centrifugation filter Amicon Ultra (Millipore). The total
amount of protein in each Cesium Chloride bands was estimated using a BCA Protein
Assay Kit (Pierce, Rockford, USA) according to the manufacturer’s instructions. The
56
Chapter 3
Viruses infecting Phaeocystis globosa
purified viral particles were heated 4 minutes at 95°C in SDS sample buffer. A
subsample of 10 µL was loaded on a SDS-PAGE gel (Ready gel for polyacrylamide
electrophoresis, 10% TrisHCl, Bio-Rad, Hercules, CA, USA ) using a Mini Protean 3
Cell (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s protocol. Protein
molecular weight standards (Precision plus protein standard, Bio-Rad, Hercules, CA,
USA) were used for size calibration. The gel was stained for 30 min with a solution of
Sypro Orange (5×10-4 of the commercial stock, Molecular Probes, Eugene, OR) diluted
in 7.5% (final concentration) acetic acid. The gel was destained for 5 minutes in 7.5%
acetic acid prior to the analysis using an Imago imager (B&L Systems, Maarssen, The
Netherlands).
Stability against physiochemical treatment. To determine the viral stability at
low temperatures, duplicates of 0.5 mL viral lysate in 2 mL cryovials were placed at 196°C (liquid nitrogen), -80°C, -50°C and -20°C for 24 hours. Samples were thawed at
30°C, afterwhich they were quickly added to exponentially growing P. globosa Pg-I
cultures (10 % v/v). Heat stability was tested for temperatures ranging from 15 to 75°C
in steps of 5°C. A subsample of viral lysate of 1 mL was heated in a waterbath at the
specific temperature of interest for 10 minutes, afterwhich samples were cooled on ice
for 5 minutes. The subsamples were added in duplicate to exponentially growing algal
culture of host (10% v/v). All cultures were incubated for 10 days at the standard culture
condition of the host. The natural in vivo fluorescence of the algal cultures was
monitored during 10 days to detect algal cell lysis. An algal culture infected with a nontreated virus lysate was taken along as a positive control, and a noninfected culture of P.
globosa Pg-I served as a negative control.
3. Results
3.1. Viral morphology and flow cytometric signatures
For all virus isolates, virus-like particles were observed in the cytoplasm of the
host cell using TEM. Representatives of the two different virus morphologies are shown
in Fig. 1. Both types of viruses were tailless, non-enveloped and with a hexagonal
outline suggesting an icosahedral symmetry. The first morphological type (PgV Group
I), with a diameter of approximately 150 nm (mean value 153 ± 8 nm) and a thin outer
layer surrounding a layered inner core (Figs. 1B and C), was shared by 6 of the virus
isolates (PgV-06T, PgV-07T, PgV-09T, PgV-12T, PgV-13T and PgV-14T). The other
half of the virus isolates (PgV-01T, PgV-03T, PgV-04T, PgV-05T, PgV-10T and PgV11T) had the second morphological type (PgV Group II). These viral particles were
characterized by a diameter of 100 nm (mean value 106 ± 7 nm) and a thick outer layer
surrounding an electron-dense inner core (Figs. 1D and E).
A similar grouping of the virus isolates could be made on the basis of their flow
cytometric signature after staining with a green fluorescent nucleic acid-specific dye
57
Chapter 3
Viruses infecting Phaeocystis globosa
(representatives are shown in Fig. 2). The larger sized virus particles (PgV Group I, Fig.
2B) had a strongly enhanced green fluorescence compared to the relatively smaller sized
virus particles (PgV Group II, Fig. 2C).
c
n
c
Figure 1. Transmission electron micrographs of thin sections of infected and noninfected
Phaeocystis globosa Pg-I. For all virus isolates TEM micrographs were obtained, but only
representatives are shown here. P. globosa noninfected (A), infected with representative virus for
PgV Group I (B and C), and P. globosa infected with representative virus for PgV Group II (D
and E). Nucleus (n) and chloroplast (c) are indicated in the noninfected P. globosa.
58
Green fluorescence (r.u)
Chapter 3
Viruses infecting Phaeocystis globosa
beads
A
beads
B
bacteria
103
beads
C
bacteria
102
PgV Group I
101
PgV Group II
100
100
101
102
103
100
Side scatter (r.u)
101
102
103
Side scatter (r.u)
100
101
102
103
Side scatter (r.u)
Figure 2. Flow cytometric signatures of PgV after staining with the nucleic acid-specific dye
SYBR Green I of (A) noninfected algal control, (B) representative virus PgV-09T of PgV Group I),
and (C) representative virus PgV-11T of PgV Group II. Green fluorescence and side scatter are
expressed in relative units (r.u.).
3.2. Genome size and type
The isolates in PgV Group I harboured a large genome, on average 466 kb ± 4
kb (Fig. 3). Those in Group II harboured a genome of smaller size, on average 177 kb ±
3 kb (Fig. 3). All the viral genomes could be digested with DNase RQ1, indicating their
genetic nature to be DNA (data not shown). The large size of the viral genomes, the
DNA nature of the genomic material, and the staining with DAPI imply that the viral
genomes consisted of double stranded DNA.
kb
M
1
2
3
533.5
485.0
436.5
388.0
339.5
291.0
242.5
194.0
145.5
97.0
48.5
Figure 3. Genome sizes of all PgV isolates were determined by PFGE. Shown here are
representatives for PgV Group I and II. Lane M: Lambda concatamers ladder, Lane 1:
representative PgV-09T of PgV Group I, Lane 2: representative PgV-11T of PgV Group II, Lane
3: noninfected culture of P. globosa. The small-sized band (approximately 45 kb) as seen in lanes
1-3 correspond to bacteriophages since the algal cultures were not axenic.
59
Chapter 3
Viruses infecting Phaeocystis globosa
3.3. Lytic cycle
The isolates belonging to PgV Group I, with the large particle diameter and
genome size, had a latent period of around 10 h, according to their lytic cycle (Fig. 4A).
The decline in algal host abundance in the infected culture was slightly delayed
compared to the increase in extracellular free viruses (Fig. 4B). For the viruses of PgV
Group II, with the relatively small particle diameter and genome size, two different
latent periods were detected: 12 h for PgV-03T and PgV-05T (PgV Group IIA; Fig. 4C),
and 16 h for the other viruses (PgV Group IIB; Figs. 4E and G). The production of free
viral particles as well as algal lysis was, however, delayed for PgV-01 (Figs. 4G and H)
when compared to the other PgV Group IIB isolates (Figs. 4E and F).
From the maximum net decline in algal cell abundance and the concurrent
maximum increase in viral abundance, an average burst size for the PgV Group I of 248
viruses P. globosa cell-1 was estimated. There was, however, considerable variation in
burst sizes for the different isolates belonging to PgV Group I despite the fact that the
algal host cells were in exponential growth phase the moment of infection (127, 356, 77,
337, 252 and 337 viruses P. globosa cell-1 for PgV-06T, 07T, 09T, 12T, 13T and 14T).
The burst sizes of the virus isolates belonging to PgV Group II were less variable (274,
415, 378, 410, 376 and 360 viruses P. globosa cell-1 for PgV-03T, 05T, 01T, 04T, 10T
and 11T), with an average of 345 viruses P. globosa cell-1 for PgV Group IIA and 381
viruses P. globosa cell-1 for PgV Group IIB.
PgV
normalized to To
30
A
C
E
G
B
D
F
H
20
10
0
P. globosa
normalized to To
4
3
2
1
0
0
20
40
60 0
Time (hours)
20
40
60 0
Time (hours)
20
40
60 0
Time (hours)
20
40
60
Time (hours)
Figure 4. Abundance of free viral particles (A, C, E and G) and algal host P. globosa Pg-I (B, D,
F and H). Open diamonds represent PgV abundance, closed circles represent P. globosa
abundance in the control cultures, and closed triangles represent P. globosa abundance in the
infected cultures. Viral growth cycles were determined for all viral isolates. Presented here are
representative for PgV Group I (A and B, PgV-09T) and PgV Group II (C and D, PgV-03T; E and
F, PgV-11T; G and H, PgV-01T). The length of the latent period for PgV Group I was 10 h and for
PgV Group II 12 h (represented by PgV-03T) or 16 h (represented by PgV-11T and PgV-01T).
60
Chapter 3
Viruses infecting Phaeocystis globosa
3.4. Structural proteins
For the structural protein analysis, at least two representative clonal virus
isolates of each PgV group described above were selected. After isopycnic CsCl
centrifugation of PgV Group I, 3 bands with a buoyant density of 1.22, 1.23 and 1.275 g
mL-1 (respectively band 1, 2 and 3 in Fig. 5) were detected for each virus isolate tested.
All bands consisted of PgV Group I viruses with their typical high green fluorescence
signature after staining with a nucleic acid-specific dye in combination with flow
cytometry. The heaviest band 3 was relatively thicker than the other two and contained
70% of the total purified PgV Group I viruses. Besides, most of the protein bands
obtained after SDS-PAGE (Fig. 5) were common in all three CsCl fractions which
indicates that the bands represent the same virus strain. Either the virus particles are
unstable in CsCl and thus the different bands represent different forms of dissociated
virus, or some of the bands consist of immature virus particles. For all three bands the
regained viruses (after repeated wash steps with sterile seawater or PBS) had lost
infectivity.
The lightest density fraction (band 1) of the PgV Group I consisted of four
major polypeptides of approximately 257, 161, 111 and 52 kDa and five minor
polypeptides with molecular masses of 205, 94, 84, 42 and 41 kDa (Fig. 5). SDS-PAGE
did, however, reveal differences in the relative amount of the detected polypeptides for
each density fraction of PgV Group I. As compared to band 1, the intermediate (band 2)
Figure 5. SDS-PAGE of structural proteins from viral particles purified by isopycnic CsCl
centrifugation. Three distinct bands (bands 1-3, with increasing densities) of comparable protein
concentration (50-55 µg mL-1) were recorded for two representatives of PgV Group I (PgV-07T
and PgV-09T), whereas only one was recorded for representative viruses of PgV Group II (PgV01T, PgV-03T and PgV-11T). Lanes M: molecular weight marker, Lane 1: band 1 of
representative PgV-09T of PgV Group I, Lane 2: band 2 of representative PgV-09T of PgV Group
I, Lane 3: band 3 of representative PgV-09T of PgV Group I, Lane 4: representative PgV-11T of
PgV Group II .Not all bands may be visible on the gel shown.
61
Chapter 3
Viruses infecting Phaeocystis globosa
and the heaviest fractions (band 3) had a substantially higher amount of the 205 kDa
protein, in combination with a reduced relative amount of the 161, 111 and 52 kDa
proteins. Furthermore, the intermediate fraction showed an enhanced relative amount of
the 42 kDa polypeptide.
After CsCl equilibrium centrifugation of PgV Group II representatives, only one band
with a buoyant density of 1.37 g mL-1 was observed. Also here, the viruses had lost their
infectivity. SDS-PAGE of PgV Group II revealed 4 main polypeptides of 119, 99, 75
and 44 kDa, and 3 minor polypeptides of 60, 62 and 38 kDa.
3.5. Host range specificity
The virus isolates were specific for P. globosa as no other algal species tested,
including other Phaeocystis species were infected (Tables 1 and 2). The viruses of PgV
Group I had a slightly higher degree of strain specificity than the viruses of PgV Group
IIA and IIB (Table 2). PgV-01T was, however, an outlier as it was the only virus isolate
causing lysis of all P. globosa strains tested (including one from the west coast of the
USA, Table 2). This difference, in combination with the delayed algal host lysis and
production of viral particles, was striking enough to separate it into a new group (PgV
Group IIC).
Table 2. Phaeocystis globosa strains used to screen for virus-induced lysis by different PgV isolates.
Plus (+) indicates lysis and minus (-) indicates no lysis of the algal host culture upon infection with
PgV (20 % v/v)
PgV Groups
PgV-05T
PgV-04T
PgV-10T
PgV-11T
PgV-01T
Skagerrak
PgV-03T
CA, USA
IIC
PgV-14T
Unknown
IIB
PgV-13T
Ph91mf
Pg-G (A)
Pg-I
Pg01MD-02
Pg01MD-06
SK 35
Pg 1
Ph91hc
Pg-G (B)
Ph-Miller
Pg01MD-04
Pg Kac31
IIA
PgV-12T
Bight, North Sea
I
PgV-09T
not Southern
PgV-07T
strains
Strain origin if
PgV-06T
P. globosa
+
+
+
+
-
+
+
+
+
-
+
+
+
+
-
+
+
+
+
-
+
+
+
+
-
+
+
+
+
-
+
+
+
+
+
-
+
+
+
+
+
-
+
+
+
+
+
-
+
+
+
+
+
-
+
+
+
+
+
-
+
+
+
+
+
+
+
+
+
+
+
+
62
Chapter 3
Viruses infecting Phaeocystis globosa
3.6. Thermostability
The representatives of PgV Group I had different sensitivities to heat and
freezing treatments when compared to the PgV group IIA, IIB and IIC (Table 3). PgV
Group I became sensitive at temperature ≥ 35°C, with a complete inactivation of the
virus at 45°C. The viruses belonging to PgV Group IIA and Group IIB were negatively
affected by temperature ≥ 25°C, with a complete loss of infectivity at 35°C for PgV
Group IIA and IIB and 30°C for PgV Group IIC.
All virus isolates remained infective after storage for more than a year at 4°C in
the dark. The PgV Group I representative was stable when frozen for 24 h at all
temperatures tested (-20°C, -50°C, -80°C and -196°C). The viruses of Group IIA and IIB
were only stable when frozen at -80ºC and -196ºC. PgV Group IIC was the most
sensitive as it could not withstand freezing at any of the temperatures tested.
Table 3. Sensitivity of PgV isolates to temperature. Sensitivity was classified as not sensitive (-, no
loss of infectivity), sensitive but still lysis (+, delayed lysis of the host in comparison of non-treated
isolate), and very sensitive (++, complete loss of infectivity). The viral lysate was freshly prepared
and added to exponentially growing algal host. Control exposure temperature was set at 15°C,
and viral activity was assayed in duplicate. All PgV isolates stayed infective at 4°C for at least a
year. Treatments were performed on representatives of each PgV groups: PgV-09T for PgV Group
I, PgV-03T for PgV Group IIA, PgV-11T for PgV Group IIB and PgV-01T for PgV Group IIC.
PgV Groups
Temperature
I
IIA
IIB
IIC
20°C
25°C
30°C
35°C
40°C
45-75°C
+
+
++
+
+
++
++
++
+
+
++
++
++
+
++
++
++
++
20°C
-50°C
-80°C
-196°C
-
++
++
-
++
++
-
++
++
++
++
63
Chapter 3
Viruses infecting Phaeocystis globosa
4. Discussion
All twelve virus isolates infecting specifically P. globosa that were
characterized in the present study seem to belong to the virus family Phycodnaviridae:
they infect an eukaryotic algal species, are polyhedral in shape, do not have an envelope,
lack a tail, are large in diameter ( >100 nm) and contain large dsDNA genomes (>175
kb, http://www.ncbi.nlm.nih.gov/ICTVdB/51000000.htm; Van Etten & Meints 1999,
Brussaard 2004a). Moreover, our suggested classification of the PgVs into the
Phycodnaviridae is confirmed by a recent study examining the genetic relatedness
among seven PgV isolates of which six are described in the present study (Brussaard et
al. 2004b). Based on conservative DNA polymerase (pol) gene sequences, which have
been shown to be a good phylogenetic marker for inferring genetic relationships among
algal viruses, these authors showed that the PgVs formed a closely related monophyletic
group within the family Phycodnaviridae. Their results demonstrated that the DNA pol
fragments of the viruses examined were at least 96.9% identical to each other. Brussaard
and coworkers (2004b), however, did detect variation in the lysis patterns of P. globosa
based on the in vivo fluorescence algae infected by different PgVs. Our results confirm
that despite the similarity in inferred amino acid sequence phylogeny, PgV isolates differ
largely in their phenotypic characteristics. Thus, the present characterization provides
relevant additional information for a proper classification of these viruses.
We categorized two main groups of viruses infecting P. globosa (PgV Group I
and PgV Group II) which differed largely in genome size, particle diameter and protein
composition. The genome size of the PgV Group I viruses was more than 2.5 times
larger than the viruses belonging to PgV Group II (466 vs. 177 kb). Complementing this,
the particle size was 1.5 times larger (150 vs. 100 nm in diameter), and the maximum
size of the main structural proteins was about twice as large (257 vs. 119 kDa). These
results make it plausible that virus-host interactions, host range, and viral replication
might differ significantly. Host range specificity was, however, remarkably comparable
for PgV Group I and II, with many of P. globosa strains being infected by both groups of
viruses. In the case that no other characteristics affect successful infection, viruses with
the shortest latent period (PgV Group I) would have a competitive advantage. One of the
P. globosa strains (Pg01MD-06) was infected by the PgV Group II viruses, but not by
the PgV Group I viruses, which in turn might provide a niche for these viruses with a
longer latent period. We found relatively high algal host diversity in the field: indeed,
three P. globosa strains differing in their sensitivity to PgV infection have been isolated
in April 2001 (clonal Pg01MD-02, -04, and -06; Table 2). Pg01MD-02 was sensitive to
the infection by all PgV groups, in contrast to Pg01MD-06 and Pg01MD-04 which were
resistant to PgV Group I, and PgV Group I, IIA and IIB, respectively. Interestingly,
Pg01MD-04 had the tendency to flocculate (produce mucus) upon infection, as did all
other algal host strains that were not sensitive to infection by PgV Group I, IIA and IIB,
(with the exception of Pg01MD-06).
64
Chapter 3
Viruses infecting Phaeocystis globosa
Subgroup PgV Group IIC had a much broader host range than all the other PgV
isolates, being able to infect all P. globosa strains tested regardless of their geographical
origin, tendency to flocculate, and the presence of flagella (Pg 1 cells did not have
flagella for example). This indicates that this type of PgV would have had a higher
probability of encountering a suitable host in the field as compared to the other PgV
groups, potentially resulting in dominance despite its longer latent period. The PgV
group IIC was, nevertheless, isolated only once, whereas the other PgV groups could be
isolated more often and regardless of the absolute abundance of P. globosa algal host in
the waters (unpublished data, CB). Although we cannot rule out that some PgV groups
were more easily isolated, it could be that either production of (immature) viral particles
or loss of infectivity differed for the various PgV groups. Temperature has, for example,
been suggested as a relevant factor reducing the infectivity of phages (Weinbauer 2004),
but little is known about the temperature sensitivity of algal virus model systems.
Although scarcely studied, temperature sensitivity of algal virus isolates seems very
diverse (Van Etten et al. 1991, Cottrell & Suttle 1995, Nagasaki & Yamaguchi 1998,
Brussaard et al. 2004a). Of the PgV isolates characterized here, Group IIC was most
sensitive to a rise in temperature. Even after 10 min at temperatures above 20°C a loss of
infectivity was detected. Such temperature was observed in situ during the summer of
2000 (Van Aken 2001). As the incubation time in situ will be much longer than 10 min,
it can be speculated that the dynamics and potential dominance (due to the broad host
range) of PgV Group IIC viruses is strongly controlled by temperature.
Although we only investigated the ability of the PgV isolates to withstand
freezing for a short period of time (24 h), the possibility for cryopreservation of PgV
Group I, IIA and IIB without any additives is remarkable. We have indications that PgV
Group I can withstand cryopreservation for even longer periods (preliminary results),
which offers the opportunity to store the original virus, isolate the appropriate host, and
study the model system in detail with time. To our knowledge, there has only been one
other account of cryopreservation of an algal virus without the addition of
cryoprotectants (a dsRNA virus infecting Micromonas pusilla; Brussaard et al. 2004a).
Long-term cryopreservation has been reported for two algal viruses (infecting
Heterosigma akashiwo and Phaeocystis pouchetii) after addition of DMSO or sucrose,
but was found to be difficult and virus isolate-dependent (Nagasaki 2001). More detailed
research is needed to find out which viral characteristics accommodate successful
cryopreservation (for example of PgV Groups I, IIA and IIB in contrast to PgV Group
IIC).
The 4 distinct types of P. globosa viruses described during this study were
collected within a year from the same geographical location. We found different PgV
groups co-occurring in the same water sample, for example, during the decline of the
summer bloom in 2000 (PgV group I and group IIA). As a direct consequence of
coexisting viruses infecting the same host population, viral infection is argued to be one
of the most important factors regulating the abundance and clonal composition of
phytoplankton population occurring in the same water (Sahlsten 1998, Tarutani et al.
2000, Brussaard 2004a, Tomaru et al. 2004). Several P. globosa strains, differing in their
65
Chapter 3
Viruses infecting Phaeocystis globosa
sensitivity to PgV, were co-occurring with the characterized viruses, which confirms that
virus infection may regulate clonal diversity during an algal bloom.
However, our results also indicated that there can be a significant overlap in the
host range of the different PgVs. This suggests that distinct PgV groups co-occurring in
the same area do compete for the same specific host strain. Little is known about viral
competition for the same specific host strain to date, most likely because so far only a
few virus-systems infecting the same algal host strain have been brought into culture.
Thus, to what extent viral competition affects the diversity of the viral community, as
well as of the algal host populations in the field remains to be seen.
Acknowledgements. We are indebted to George McCartney (Queen’s University
Belfast) for the TEM images. We are grateful to Dr. A. Larsen (University of Bergen,
Norway), Dr. P. Salomon (Univeristy of Kalmar, Sweden), Dr. V. Schoemann
(University Libre de Bruxelles, Belgium) and Dr. A. Zingone (Stazione Zoologica “A.
Dohrn”, Naples, Italy) for providing Phaeocystis cultures. Special thanks to Dr. A.
Larsen for screening P. pouchetii AJ01 for the host range experiment. We, furthermore,
thank Anna Noordeloos, Marco Flohil and Govert Van Noort for their technical
assistance. This work was supported by the Netherlands Organization for Scientific
Research (NWO-ALW project 811.33.002).
Literature cited
Arrigo KR, Robinson, DH, Worthen DL, Dunbar RB, DiTullio GR, VanWoert M,
Lizotte MP (1999) Phytoplankton community structure and the drawdown of
nutrients and CO2 in the Southern Ocean. Science 283:365-367.
Bratbak G., Egge JK, Heldal M (1993) Viral mortality of the marine alga Emiliania
huxleyi (Haptophyceae) and termination of algal bloom. Mar Ecol Prog Ser 93:3948.
Brussaard CPD (2004a) Viral control of phytoplankton populations - a review. J Euk
Microbiol 51:125-138.
Brussaard CPD (2004b) Optimization of procedures for counting viruses by flow
cytometry. Appl Environ Microbiol 70:1506-1513.
Brussaard CPD, Gast GJ, Van Duyl FC, Riegman R (1996a) Impact of phytoplankton
bloom magnitude on a pelagic microbial food web. Mar Ecol Prog Ser 144:211221.
Brussaard CPD, Kempers RS, Kop AJ, Riegman R, Heldal M. (1996b) Virus-like
particles in a summer bloom of Emiliania huxleyi in the North Sea. Aquat Microb
Ecol 10:105-113.
Brussaard CPD, Kuipers B, Veldhuis MJW (2005a). A mesocosm study of Phaeocystis
globosa population dynamics I. Regulatory role of viruses in bloom control.
Harmful Algae 4:859-874
66
Chapter 3
Viruses infecting Phaeocystis globosa
Brussaard CPD, Mari X, Van Bleijswijk JDL, Veldhuis MJW (2005b). A mesocosm
study of Phaeocystis globosa (Prymnesiophyceae) population dynamics. II.
Significance for the microbial community. Harmful Algae 4:875-893
Brussaard CPD, Noordeloos AAM, Sandaa RA, Heldal M, Bratbak G (2004a).
Discovery of a dsRNA virus infecting the marine photosynthetic protist
Micromonas pusilla. Virology 319:280-291.
Brussaard CPD, Riegman R, Noordeloos AAM, Cadee GC, Witte H, Kop AJ,
Nieuwland G, Van Duyl FC, Bak RPM (1995) Effects of grazing, sedimentation
and phytoplankton cell lysis on the structure of a coastal pelagic food web. Mar
Ecol Prog Ser 123:259-271.
Brussaard CPD, Short SM, Frederickson CM, Suttle CA (2004b). Isolation and
phylogenetic analysis of novel viruses infecting the phytoplankton Phaeocystis
globosa (Prymnesiophyceae). Appl Environ Microbiol 70:3700-3705.
Cadée GC, Hegeman J (1991) Historical phytoplankton data of the Marsdiep. Hydrobiol
Bull 24:111-118.
Castberg T, Larsen A, Sandaa RA, Brussaard CPD, Egge J, Heldal M, Thyrhaug R, Van
Hannen EJ, Bratbak G (2001) Microbial population dynamics and diversity
during a bloom of the marine coccolithophorid Emiliania huxleyi (Haptophyta).
Mar Ecol Prog Ser 221:39-46.
Chen F, Suttle CA, Short SM (1996) Genetic diversity in marine algal virus communities
as revealed by sequence analysis of the DNA polymerase genes. Appl Environ
Microbiol 62:2869-2874.
Cottrell MT, Suttle CA (1991) Wide-spread occurrence and clonal variation in viruses
which cause lysis of a cosmopolitan, eukaryotic marine phytoplankter,
Micromonas pusilla. Mar Ecol Prog Ser 78:1-9.
Cottrell MT, Suttle CA (1995) Dynamics of a lytic virus infecting the photosynthetic
marine picoflagellate Micromonas pusilla. Limnol Oceanogr 40:730-793.
DiTullio GR, Grebmeier JM, Arrigo KR, Lizotte MP, Robinson DH, Leventer A, Barry
JP, VanWoert ML, Dunbar RB (2000). Rapid and early export of Phaeocystis
antarctica blooms in the Ross Sea, Antarctica. Nature 404:595-598.
Eppley RW, Holmes RW, Strickland JDH (1967) Sinking rates of the marine
phytoplankton measured with a fluorochrometer. J Exp Mar Biol Ecol 1:191-208.
Evans C, Archer SD, Jacquet S, Wilson WH (2003) Direct estimates of the contribution
of viral lysis and microzooplankton grazing to the decline of a Micromonas spp.
population. Aquat Microb Ecol 30:207-219.
Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects.
Nature 399:541-548.
Guillard RRL (1975) Culture of phytoplankton for feeding marine invertebrates. In
Chanley MH (Ed.), Culture of marine invertebrate animals. Plenum Press, New
York, pp 29-60.
Hansen PJ. (1989) The red tide dinoflagellate Alexandrium tamaranse: effects on
behaviour and growth of a tintinnid ciliate. Mar Ecol Prog Ser 53:105-116.
Harrison PJ, Waters RE, Taylor FJR (1980) A broad spectrum artifical seawater medium
for coastal and open ocean phytoplankton. J Phycol 16:28-35.
67
Chapter 3
Viruses infecting Phaeocystis globosa
Lancelot C, Billen G, Sournia A, Weisse T, Colijn F, Veldhuis MJW, Davies A,
Wassman P (1987) Phaeocystis blooms and nutrient enrichment in the continental
coastal zones of the North Sea. Ambio 16:38-46.
Nagasaki K (2001) Domestication of eukaryotic microalgal viruses from marine
environments. Microbes and Environments 16:3-8.
Nagasaki K, Yamaguchi M (1998) Effect of temperature on the algicidal activity and the
stability of HaV (Heterosigma akashiwo virus). Aquat Microb Ecol 15:211-216.
Nagasaki K, Tomaru Y, Nakanishi K, Hata N, Katanozaka N, Yamaguchi M (2004)
Dynamics of Heterocapsa circularisquama (Dinophyceae) and its viruses in Ago
Bay, Japan. Aquat Microb Ecol 34:219-226.
Orton JH (1923) The so-called “Baccy-juice” in the waters of the Thames oysters beds.
Nature 111, 773.
Pieters H, Kluytmans JH, Zandee DI (1980) Tissue composition and reproduction of
Mytilus edulis in relation to food availability. Neth J Sea Res14:349-361.
Reisser W (1993) Viruses and virus-like particles of freshwater and marine eukaryotic
algae review. Arch Protistenkunde 143:257-265.
Ruardij P, Veldhuis MJW, Brussaard CPD (2005) Modeling the bloom dynamics of the
polymorphic phytoplankter Phaeocystis globosa: impact of grazers and viruses.
Harmful Algae 4:941-963
Sahlsten E (1998) Seasonal abundance in Skagerrak-Kattegat coastal waters and host
specificity of viruses infecting the marine photosynthetic flagellate Micromonas
pusilla. Aquat Microb Ecol 16:103-108.
Schroeder DC, Oke J, Hall M, Malin G, Wilson W (2003) Virus succession observed
during an Emiliania huxleyi bloom. Appl Environ Microbiol 69:2484-2490.
Short SM, Suttle CA (2003) Temporal dynamics of natural communities of marine algal
viruses and eukaryotes. Aquat Microb Ecol 32:107-119.
Stefels J, Van Boekel WHM (1993) Production of DMS from dissolved DMSP in axenic
cultures of the marine phytoplankton species Phaeocystis sp. Mar Ecol Prog Ser
97:11-18.
Tarutani K, Nagasaki K, Yamaguchi M (2000) Viral impacts on total abundance and
clonal composition of the harmful bloom-forming phytoplankton Heterosigma
akashiwo. Appl Environ Microbiol 66 :4916-4920.
Tomaru Y, Tarutani K, Yamaguchi M, Nagasaki K (2004) Quantitative and qualitative
impacts of viral infection on a Heterosigma akashiwo (Raphidophyceae) bloom in
Hiroshima Bay, Japan. Aquat Microb Ecol 34:227-238.
Van Aken HM (2001) 140 years of daily observations in a tidal inlet (Marsdiep). ICES
Marine Science Symposia 219:359-631.
Van Boekel WHM, Hansen FC, Riegman R, Bak RPM (1992) Lysis-induced decline of
a Phaeocystis spring bloom and coupling with the microbial foodweb. Mar Ecol
Prog Ser 81:269-276.
Van Etten JL, Lane LC, Meints RH (1991) Viruses and virus-like particles of eukaryotic
algae. Microbiol. Rev. 55:586-620.
Van Etten JL, Meints RH (1999) Giant viruses infecting algae. Ann Rev Microbiol
53:447-494.
68
Chapter 3
Viruses infecting Phaeocystis globosa
Verity PG, Smetacek V (1996) Organism life cycles, predation, and the structure of
marine pelagic ecosystems. Mar Ecol Prog Ser 130:277-293.
Weinbauer MG. (2004) Ecology of prokaryotic viruses. FEMS Microbiology reviews
28:127-181.
Wilhelm WS, Suttle CA (1999). Viruses and nutrient cycles in the Sea. Biosciences 49,
781-788
Wommack KE, Colwell RR (2000) Virioplankton: viruses in aquatic ecosystems.
Microbiol Mol Biol Rev 64:69-114.
Wommack KE, Ravel J, Hill RT, Chun J, Colwell RR (1999) Population dynamics of
Chesapeake Bay virioplankton: total-community analysis by pulsed-field gel
electrophoresis. Appl Environ Microbiol 65:231-240.
69
Chapter 3
Viruses infecting Phaeocystis globosa
70
Chapter 4
Viruses as mortality agents of
picophytoplankton in the deep
chlorophyll maximum layer during
IRONAGES III 1
Anne-Claire Baudoux, Marcel J. W. Veldhuis, Harry J. Witte, and Corina P. D.
Brussaard
We report virally induced mortality rates of the picoeukaryotic (2 size classes)
and prokaryotic (cyanobacteria Prochlorococcus and Synechococcus) phytoplankton
during a cruise in the oligotrophic subtropical northeastern Atlantic (October 2002). A
dilution assay, simultaneously estimating viral lysis and microzooplankton grazing, was
applied around the deep chlorophyll maximum (DCM) at 6 stations. For the smallest
picoeukaryotes (group I) viral lysis was responsible for 50 to 100% of the total cell
losses, with rates ranging from 0.1 to 0.8 d-1. Viral lysis rates were positively linked to
the abundance and the contribution of large genome-sized (180 to 225 kb) putative algal
viruses. In contrast, the prokaryotic picophytoplankton did not seem to be controlled by
viral lysis. For Synechococcus, microzooplankton grazing dominated, with rates
between 0.1 and 0.25 d-1 (comparable to those for the eukaryotic algae). For
Prochlorococcus, both viral lysis and microzooplankton grazing rates were very low (0 0.1 d-1). Overall, the total carbon produced by the picophytoplankton community was
balanced by the combined losses due to viral lysis and microzooplankton grazing. Viral
lysis released 0.1 - 0.3 µg picophytoplanktonic C L-1 d-1, which comprised 21% of the
total carbon production by picophytoplankton
1
Accepted for publication in Limnology and Oceanography
71
Chapter 4
Viral lysis of picophytoplankton
1. Introduction
Phytoplankton communities in oligotrophic open oceans are usually dominated
by picophytoplanktonic cells (< 3 µm), including eukaryotes and prokaryotes (Partensky
et al. 1996). The prokaryotic component is represented by the cyanobacteria
Synechococcus and Prochlorococcus (Partensky et al. 1996). The picoeukaryotes are less
investigated than the prokaryotes but their contribution to the total carbon biomass and
the ecosystem productivity still is substantial (Li 1994; Partensky et al. 1996; Worden et
al. 2004). Recent publications describe dynamics in abundance (Worden et al. 2004) and
a high degree of species diversity within this group (Moon-van der Staay et al. 2001;
Veldhuis and Kraay 2004).
Picophytoplanktonic cells possess a high growth rate, despite the very the low
concentrations of major nutrients characterizing oligotrophic habitats (Partensky et al.
1996; Worden et al. 2004). Their ecological success in oligotrophic waters has mainly
been attributed to their small size, since the relatively high surface area to volume allows
maximal uptake to sustain the cell metabolism (Raven 1998). Among the loss factors
regulating picophytoplankton populations, grazing is considered important (Quevedo
and Anadòn 2001, Worden et al. 2004), whereas sedimentation is thought negligible
considering their micrometer size range (Raven 1998). Besides grazing, there are
indications that cell lysis may contribute to phytoplankton loss in oligotrophic systems
(Agusti et al. 1998). One of the factors causing cell lysis is viral infection and currently
host-specific viruses are reported for photosynthetic prokaryotes (for review see e.g.
Mann 2003) as well as picoeukaryotes (for review see e.g. Brussaard 2004a).
Phytoplankton losses caused by cells lysis and grazing influence the flow of nutrient and
energy in different ways. The release of cell constituents upon lysis directly affects the
standing stock of dissolved organic carbon and the recycling of nutrients, whereas
grazing channels phytoplankton biomass to the higher trophic levels (Wilhelm and Suttle
1999). Therefore, the relative effect of viruses and microzooplankton need to be assessed
for optimal understanding biogeochemical cycling.
Studies that encompass oligotrophic sites show high numerical abundance and a
diverse and dynamic virus community for both cyanophages and algal viruses (Short and
Suttle 2003; Mühling et al. 2005). This implies that viruses may be responsible for algal
mortality. Actual viral lysis rates of photosynthetic organisms in oligotrophic systems
are, however, poorly documented. The few existing studies suggest that between 0.6 and
8% d-1 of the standing stock of Synechococcus undergo viral lysis in oligotrophic waters
(Waterbury and Valois 1993, Garza and Suttle 1998). These values were however
obtained using theoretical conversion factors and calculations. A specific assay that
allows direct estimation of virally induced algal mortality has been developed recently
(Evans et al. 2003). This viral dilution assay has been applied successfully for the
picophytoplankter Micromonas pusilla (Evans et al. 2003) during a mesocosm study and
the nanophytoplankter Phaeocystis globosa during a field study in temperate eutrophic
coastal waters (Baudoux et al. 2006). The viral lysis dilution assay has, as yet, never
been applied in an oligotrophic ecosystem. During this study, we used the viral lysis
72
Chapter 4
Viral lysis of picophytoplankton
dilution assay to elucidate the relevance of virally induced mortality in
picophytoplankton of the oligotrophic subtropical northeastern Atlantic. Viral lysis rates
were determined across the different groups (prokaryotes as well as eukaryotes) forming
the picophytoplankton community and compared to microzooplankton grazing.
2. Material and methods
Study area and sampling. Six stations were investigated aboard the R.V.
Pelagia from 03 - 29 October 2002 during the IRONAGES III shipboard expedition (Fig.
1, Table 1). Samples were collected using 10 L NOEX bottles mounted on the Rosette
sampler equipped with a Seabird conductivity-temperature-depth (CTD) sensor and a
PAR detector. The sampling was performed around the deep chlorophyll maximum
(DCM) according to the highest in vivo fluorescence as detected by a fluorometer set up
on the CTD rosette (Table 1).
Discrete samples for nutrients (5 mL) were filtered through 0.2 µm pore-size
polysulfone filters (Acrodisc, Gelman Sciences). Analyses were performed onboard
using a TrAAcs 800 autoanalyzer for dissolved inorganic nutrients (N, P, and Si) as
described in Baudoux et al. (2006). Chlorophyll a (Chl a) samples of typically 1.5 L
were filtered onto a GF/F filter (Whatman) and stored at -80°C until analysis. Chl a was
extracted in 90% acetone and measured fluorimetrically.
-25°
-20°
-15°
35°
35
-10°
35°
N
Latitude (N)
8
30°
30°
27
40° N
33
15
25°
20
0
25°
20°
10°
0 1000 km
-30° -20°-10° 0° 10°
200 km
-25°
30°
-20°
-15°
Longitude (W)
-10°
Figure 1. Location of the sampling stations during the Ironages III cruise (Ponta Delgada,
Azores, Portugal – Valencia, Spain; 03 - 29 October 2002)
73
Chapter 4
Viral lysis of picophytoplankton
Table 1. Location and characteristics of the studied stations.
Station
8
15
20
27
33
35
Latitude (°N)
31.71
26.78
25.00
27.43
27.03
33.70
Longitude (°W)
Depth sample (m)
Temperature (ºC)
Salinity
NO3 (µmol L-1)
NO2 (µmol L-1)
NH4 (µmol L-1)
PO4 (µmol L-1)
Si (µmol L-1)
20.00
100
18.2
36.59
0.05*
0.02
0.10
0.02
0.61
24.00
100
20.1
36.86
0.06*
0.01
0.13
0.02
0.54
18.86
70
19.6
36.42
1.52
0.19
0.33
0.17
0.55
25.50
70
23.0
37.09
0.02*
0.01
0.02*
0.02
0.41
17.73
60
23.0
36.75
0.03*
0.03
0.14
0.02
0.42
13.22
80
17.7
36.39
0.01*
0.01
0.09
0.01
0.49
Chl a (µg L-1)
0.06
0.24
0.43
0.17
0.27
0.18
* Value below the detection limit; 0.008 µmol L-1 for PO4, 0.08 µmol L-1 for NO3, 0.008 µmol L-1 for NO2,
0.03 µmol L-1 for NH4, and 0.1 µmol L-1 for Si. Standard deviation between runs were 0.004 µmol L-1 for PO4,
0.05 µmol L-1 for NO3, 0.006 µmol L-1 for NO2, 0.04 µmol L-1 for NH4, and 0.07 µmol L-1 for Si.
Phytoplankton. Phytoplankton abundance from natural as well as experimental
samples was enumerated directly after sampling using a modified Beckman Coulter XLMCL flow cytometer. To increase the instrument sensitivity, the flow rate of the sheath
fluid was reduced and the band pass filter in the red detector was removed to increase the
spectral fluorescence band. The instrument was equipped with a laser with an excitation
wavelength of 488 nm (15mW) and emission bands for the chlorophyll autofluorescence
(> 630 nm) and phycoerythrin (PE 575 ± 20 nm). The discriminator for phytoplankton
was the red chlorophyll autofluorescence. Flow rate (135 ± 7 µL min-1) and machine
drift were checked every day using calibrated beads (Flow-Check Fluorospheres,
Beckman Coulter) as internal standard. A maximal volume of 1.5 mL per sample was
analyzed. Based on the pigment autofluorescence and forward scatter, we discriminated
the prokaryotes Prochlorococcus spp. and Synechococcus spp., as well as two
populations of picoeukaryotes (Fig. 2A). Synechococcus was discriminated from the
other phytoplankton based on the presence of their orange autofluorescence caused by
the accessory pigment phycoerythrin. The division of the picoeukaryotes was based on
their relative size using the approach of Veldhuis and Kraay (2004); the picoeukaryote
group I had a cell diameter ranging between 1.3 and 1.5 µm and group II between 1.5
and 2.5 µm.
74
Chapter 4
Viral lysis of picophytoplankton
Phytoplankton carbon biomass was derived from cellular carbon content of the
specific phytoplankton. We used an averaged cellular carbon content of 46 fg C cell-1 for
Prochlorococcus (Bertlisson et al. 2003). For Synechococcus and the picoeukaryotes, the
cellular carbon content was based on cell biovolume, which was estimated using the
calibration method of Veldhuis and Kraay (2004). Assuming phytoplankton cells to be
spherical, we used a biovolume to carbon conversion factor of 254 fg C µm-3 for
Synechococcus, derived from a carbon content of 213 fg C cell-1 (Bertlisson et al. 2003).
For the picoeukaryotes, we used a biovolume to carbon conversion factor of 239 fg C
µm-3, an average of the values obtained for Ostreococcus sp. CCE9901 (233 – 247 fg C
µm-3, Worden et al. 2004) and Micromonas pusilla CCMP 489 (238 fg C µm-3, DuRand
et al. 2002).
Figure 2. Flow cytometric dot plots of (A) the typical phytoplankton community and (B) the typical
viral community for the stations studied. The phytoplankton community is composed of the
cyanobacteria Prochlorococcus and Synechococcus as well as 2 picoeukaryotic populations with
the group I relatively smaller (1.3 - 1.5 µm) than the group II (1.5 - 2.5 µm). The viral community
is composed of three viral groups (V1, V2, and V3) discriminated based on the intensity of their
green fluorescence after staining with the nucleic acid-specific dye SYBR Green I. An internal
standard (yellow green fluorescent 1 µm beads) was added to the sample.
Viral lysis and microzooplankton grazing. The virally induced algal mortality
as well as the microzooplankton grazing on phytoplankton was assessed simultaneously
using an adapted version of the traditional dilution assay (Evans et al. 2003) after
modifications by Baudoux et al. (2006). A limitation of the dilution method is that initial
cell density should be sufficiently high to allow a 3 to 4 fold dilution with still sufficient
cells present for accurate counting. During the present cruise cell densities in surface
waters were extremely low and did not meet this criterion. In contrast, numbers at the
DCM were sufficiently high; therefore the present study was restricted to the deeper
waters.
75
Chapter 4
Viral lysis of picophytoplankton
The traditional dilution series of the natural seawater with 0.2 µm filtered
natural sample, which provided the microzooplankton grazing rate (Landry and Hassett
1982), was combined with a second dilution series with 30 kDa filtered natural sample,
which provided the loss rate due to both grazing and viral lysis. Viral lysis rates were
estimated from the difference between the two dilutions series. All material used for the
experiments was carefully cleaned with 1N HCl, rinsed with MilliQ and finally with the
same water from which the sample was collected. The experimental set up was
conducted in a controlled room adjusted at in situ temperature (17 - 23°C) and under
dimmed light. Around 08:00 h (local time), two 20 L-samples were carefully siphoned
from the NOEX bottles into two 20 L carboys which were darkened to prevent light
stress when the sample was brought on deck. The first 20 L sample was filtered through
a 0.2 µm pore-size polycarbonate filter (47 mm, Poretics, Millipore). A 5 L aliquot was
used as the 0.2 µm diluent, the remaining sample was ultrafiltered through a 30 kDa
polyether sulfone membrane filter (Pellicon filtration system, Millipore). The 30 kDa
filtrate was used for generating the 30 kDa dilution series. The second 20 L sample was
pre-sieved through 200 µm mesh to remove larger zooplankton and it was immediately
used to set up the dilution series (20, 40, 70, and 100% of natural water) with the 0.2 µm
and the 30 kDa diluent (3× 300 mL soft polycarbonate incubation bottles). Upon filling,
a 5 mL subsample was taken (T=0). The incubation bottles were squeezed and closed in
such a way that no air bubble was trapped in the bottle. All bottles were mounted on a
slowly rotating (0.5 rpm) plankton wheel and incubated at temperature (17 - 23°C) and
light intensity (10 - 37 µmol quanta m-2 s-1, light period of 12 h) adjusted to in situ
conditions given by the CTD and the PAR detectors. After 24 hours incubation, another
5 mL subsample was taken to monitor phytoplankton growth. The apparent growth rate
(µapp, in d-1) was calculated for each sample from the changes in abundance during the
incubation.
The regression coefficient of the apparent growth rate versus the dilution factor
for the 30 kDa series represents the phytoplankton losses due to microzooplankton
grazing and viral lysis (Mg+v, d-1), whereas the regression coefficient resulting from the
0.2 µm dilution series represents only the microzooplankton grazing rate (Mg, d-1).
Specific virally induced mortality rates (Mv, d-1) were thus obtained from the difference
between Mg+v and Mg. Specific growth rates (µ, d-1) were determined as the y-axis
intercept value of the regression line obtained with the 30 kDa series. The significance
(p) of the slope (Mg and Mg+v) and the intercept (µ) was determined performing a t-test
on the regression analysis. The significance (p) between the slopes of the regressions
lines (i.e., significance of Mv) was also estimated using a t-test.
A carbon budget was determined combining the cellular carbon content
estimates (above section) and data of the dilution experiments. For each specific
phytoplankton group, the carbon production (CP, in µg C L-1 d-1), losses due to grazing
(G, in µg C L-1 d-1) and, by adaptation, the losses due to viruses (V, in µg C L-1 d-1) were
calculated using the formulas of Landry et al. (2000); CP = µ × Pm ; G = Mg × Pm ; V =
Mv × Pm; and Pm = P0 × [e(µ - Mg+v) t - 1]/(µ - Mg+v)t, where µ (in d-1) is the dilution-based
specific growth rate (y intercept of the 30 kDa regression), Mg and Mv (in d-1) are the
76
Chapter 4
Viral lysis of picophytoplankton
dilution-based grazing and viral lysis rates, Po is the initial carbon biomass of
picophytoplankton, and Pm (in µg C L-1) is the geometric mean carbon biomass of
picophytoplankton during the incubation and t (in d) is the time of incubation.
Virus abundance. The abundance of viruses was determined on glutaraldehyde
fixed samples (final concentration 0.5% glutaraldehyde, frozen in liquid nitrogen and
stored at -80°C prior analysis) using a Beckton-Dickinson FACSCalibur flow cytometer,
with a 15 mW 488 nm air-cooled argon-ion laser according to Brussaard (2004b).
Thawed samples were diluted (dilution factor >50) in 0.2 µm filtered autoclaved TEbuffer (pH 8) and heated at 80°C for 10 min with the nucleic acid-specific dye SYBR
Green I at a final concentration of 5 × 10-5 of the commercial stock (Molecular Probes,
Invitrogen). Virus counts were corrected for the blank consisting of TE-buffer with
autoclaved 0.2 µm filtered seawater in the correct dilution. An internal standard (1 µm
yellow green fluorescent beads, Molecular Probes, Invitrogen) was added to the sample
prior to analysis. Different virus groups (V1, V2, and V3) could be clearly discriminated
on the basis of the green fluorescence and side scatter signature (Fig. 2B). Data were
analysed using the freeware CYTOWIN (http://sb-roscoff.fr/phyto/cyto.html). Due to
graphic software constraints, some viral particles may appear off-scale on the side scatter
signal (Fig. 2B). However, the discriminator was set on the green fluorescence signal,
thereby all particles that may appear off scale are actually computed in the total number
of viruses.
Virus diversity. Virus diversity was examined on a 5 L sample using pulsed
field gel electrophoresis (PFGE) as described by Larsen et al. (2001). Samples were
concentrated by 30 kDa MWCO ultrafiltration (Vivaflow 200, Vivascience), and
clarified of bacteria and cell debris by low speed centrifugation (10,000 × g, 30 min at
4°C, fixed angle rotor F-34-6-38, Eppendorf 5810R). Supernatant was harvested by
ultracentrifugation (141,000 × g, 2 h at 8°C, fixed angle rotor TFT 55.38 rotor,
Centrikon T-1080, Kontron Instruments) and pellets were resuspended in SM-buffer.
Three plugs of this concentrate were prepared in molten 1.5% (w/v) InCert agarose
(Cambrex Bioscience, Rockland, ME USA) and digested overnight at 30°C in a lysis
buffer (250 mmol L-1 EDTA, 1% SDS (v/v), 1 mg mL-1 proteinase K, Sigma-Aldrich).
Samples were loaded onto a 1% SeaKem GTG agarose gel (Cambrex Bioscience,
Rockland, ME USA) in 1× TBE buffer. The gel was run using a Bio-Rad DR-II CHEF
Cell unit operating at 6 V cm-1 at 14°C in 0.5× TBE tank buffer. Two different pulse
ramp settings were used for an optimal sizing of viral genomes. Besides a pulse ramp of
1-6 s for 20 h to examine the smaller virus genomes, a pulse ramp of 8-30 s for 20 h was
used to discriminate the larger virus genomes. After electrophoresis, gels were stained
for 1 h with SYBR Green I (1 × 10-4 of commercial solution, Molecular Probes,
Invitrogen) and destained 10 min in MilliQ (Gradient A10, Millipore) before a digital
analysis for fluorescence using a FluorS imager (Bio-Rad Instrument). Sizing of the viral
genomes was performed against a 5 kb lambda ladder or a lambda concatamers ladder
(both Bio Rad, Richmond, CA). The relative abundance of the different viral genome
77
Chapter 4
Viral lysis of picophytoplankton
sizes was estimated by normalizing the intensity of each detected band by the
determined genome size.
3. Results
3.1 Picophytoplankton
Total picophytoplankton abundance varied between 1.2 - 4.7 × 104 cells mL-1
(Fig. 3A). For all stations, the prokaryotic cyanobacteria (Prochlorococcus and
Synechococcus) numerically dominated the picophytoplankton community with an
abundance ranging from 1.1 to 3.8 × 104 cells mL-1 (76 to 95% of total numerical
abundance, Fig. 3A). Prochlorococcus was the main contributor, making up for 71 to
99% of the cyanobacterial abundance. The highest contribution of picoeukaryotes (18
and 24%, Fig. 3A) was found for the southeastern stations of the studied area (stations
20 and 33). The smaller sized picoeukaryote group I generally dominated over the group
II with abundance between 0.4 and 5.4 × 103 mL-1 (up to 3-fold that of the
picoeukaryotes group II).
Prochlorococcus
Synechococcus
picoeukaryotes I
picoeukaryotes II
Abundance (%)
100 A
80
60
40
20
Biomass (%)
0
100 B
80
60
40
20
0
8
15
20 27
Station
33
35
Figure 3. Contribution of the different picophytoplankton groups to (A) the total
picophytoplankton abundance and (B) the total carbon biomass of picophytoplankton. Numbers
above the bars indicate (A) the total picophytoplankton abundance (× 104 mL-1) and (B) the total
picophytoplankton carbon biomass (µg C L-1).
78
Chapter 4
Viral lysis of picophytoplankton
In terms of carbon biomass, the picophytoplankton ranged between 1.5 and 6.0
µg C L-1 with the highest values found at station 20 (Fig. 3B). The cyanobacteria
Synechococcus and Prochlorococcus accounted on average for 28 ± 14 % (range 3 44%) and 32 ± 17 % (range 20 – 65 %) of the total picophytoplankton carbon biomass
respectively (Fig. 3B). The picoeukaryotes accounted on average for 41 ± 9 % (range 30
– 53 %) of the total carbon biomass (Fig. 3B).
3.2. Dilution assay
Apparent growth (d-1)
Apparent growth (d-1)
The viral lysis dilution assay could be applied successfully to the oligotrophic
study site (Figs. 4 and 5, and Table 2). Figure 4 depicts the four representative
combinations of viral lysis and microzooplankton grazing obtained during the cruise:
substantial viral lysis rates with varying microzooplankton grazing (Fig. 4A and B) and
vice versa (Fig. 4C and D).
0.8
A
0.2 µm
30 kDa
B
0.4
0.0
-0.4
0.8
C
D
0.4
0.0
-0.4
0.0 0.2 0.4 0.6 0.8 1.0
0.0 0.2 0.4 0.6 0.8 1.0
Fraction of natural sample
Figure 4. Representative examples of the viral lysis dilution assay obtained during the cruise for
different grazing and viral lysis conditions. Each dilution was done in triplicate, and the linear
regressions were done through the independent 12 data points of 3 replicates of 4 dilutions each.
Parallel dilution experiments were performed in 30 kDa ultrafiltrate (no grazer, no virus) and 0.2
µm (no grazer) filtered seawater. Microzooplankton grazing rates correspond to the regression
slope obtained with the 0.2 µm dilution series. Viral lysis rates correspond to the difference of
regression coefficients of the 0.2 µm and 30 kDa series. (A) Viral lysis but no microzooplankton
grazing observed as for example recorded for the picoeukaryote group I at station 35; (B) Viral
lysis as well as microzooplankton grazing as recorded for the picoeukaryote group I at station 33;
(C) No viral lysis, no microzooplankton grazing observed as recorded for Prochlorococcus at
station 15; (D) Microzooplankton grazing but no viral lysis observed as recorded for
Synechococcus at station 33. N.B. For the readability of the graphs, we only show the averaged
value and their standard deviation at each dilution level; this does not affect the linear regressions
which were done using the independent values of triplicates at each of four dilutions.
79
Chapter 4
Viral lysis of picophytoplankton
As a result, virally induced mortality strongly varied depending on the
picophytoplankton group examined (Fig. 5, Table 2). The highest viral lysis rates were
observed for the picoeukaryote group I with rates ranging from 0.1 and 0.8 d-1. Viral
lysis rates of the other three groups (picoeukaryotes group II, Synechococcus and
Prochlorococcus) were smaller than 0.1 d-1. The microzooplankton grazing on the
picoeukaryote group I ranged between 0 and 0.2 d-1, and on the picoeukaryotes group II
between 0.1 and 0.4 d-1. Comparable grazing rates (0.1 - 0.3 d-1) were obtained for
Synechococcus, but grazing by microzooplankton on Prochlorococcus was generally
lower (max. 0.1 d-1).
Latitude (N)
35°
0.54
0.14
B) Picoeukaryotes II
0.81
0.00
-
-
30°
N
0.25
0.24
0.11 0.40
0.21
0.11 0.13
25°
0.15
0 200 km
35°
Latitude (N)
A) Picoeukaryotes I
C) Synechococcus
-
0.10
0.01
0.11 0.09
0.20
0.09 0.00
0.32
D) Prochlorococcus
-
0.12
0.25
0.06
0.01
30°
0.00
0.00
0.22 0.01
0.24
0.09 0.00
25°
0.07
-25°
-20°
-15°
0.02
0.00
0.08 0.00
0.05
0.00 0.00
0.06
-10°
Longitude (W)
-25°
-20°
-15°
-10°
Longitude (W)
Figure 5. Overview of viral lysis and microzooplankton grazing rates per station for (A) the
picoeukaryote group I, (B) the picoeukaryote group II, (C) Synechococcus, and (D)
Prochlorococcus. Viral lysis (d-1) is shown as the top number and grazing (d-1) is the bottom
number.
The growth rates for both picoeukaryote groups averaged 0.4 d-1 (Table 2), with
more variation for group I (0.2 - 0.9 d-1) than group II (0.3 - 0.6 d-1). The growth rate
averaged 0.2 d-1 for both Synechococcus (0.03 - 0.3 d-1) and Prochlorococcus (0.1 – 0.2
d-1), with the exception of Prochlorococcus at station 35 (1.3 d-1). Occasionally, the viral
dilution assay provided unsuccessful results for one or more of the specific algal groups
(station 8 and 35), which appears related to very low abundance and/or very low
apparent growth rates in the undiluted samples (the lowest encountered during this
study).
80
Chapter 4
Viral lysis of picophytoplankton
Table 2. Viral lysis, microzooplankton grazing, and growth rates calculated from the dilution
assay for the 4 phytoplankton groups. Coding listed in superscript correspond to the significance
(t-test) of the mortality and growth (- unsuccessful; a p <0.05; b p = 0.05-0.1; c p >0.1)
Station
8
15
20
27
33
35
Viral lysis rates (d )
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
0.54 a
-
0.40 a
0.09 c
0.01 c
0.00 c
0.13 c
0.00 c
0.00 c
0.00 c
0.25 b
0.10 c
0.00 c
0.02 c
0.24 a
0.01 c
0.00 c
0.00 c
0.81 a
0.12 a
0.06 c
Grazing rates (d-1)
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
0.14 b
-
0.11 a
0.09 c
0.09 c
0.00 c
0.15 b
0.36 a
0.07 b
0.06 c
0.11 c
0.11 c
0.22 a
0.08 c
0.21 a
0.20 a
0.24 a
0.05 a
0.00 c
0.25 a
0.01 c
Growth rates (d-1)
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
0.88 a
-
0.44 a
0.30 a
0.07 b
0.19 a
0.26 a
0.51 a
0.06 a
0.20 a
0.31 a
0.39 a
0.18 a
0.12 a
0.20 a
0.61 a
0.31 a
0.30 a
0.54 a
0.28 a
1.30 a
-1
3.3. Virus abundance and diversity
The total virus abundance was comparable for all stations (1.6 - 1.8 × 107 virus
mL ), except for station 35 that showed a higher abundance (2.5 × 107 mL-1). Within the
viral community, the group V1 (characterized by the lower nucleic acid green
fluorescence, Fig. 2B) dominated and comprised 64 - 73% of the total abundance. The
abundance of the group V3 (with the highest nucleic acid green fluorescence)
represented 5 - 11% of the total virus community. The highest abundance of this group
V3 was recorded at stations 8 and 35 (2 × 106 mL-1), whereas the lowest abundance was
found at station 20 (0.8 × 106 mL-1).
-1
81
Chapter 4
Viral lysis of picophytoplankton
The virus diversity analysis using PFGE displayed 4 to 7 genome sizes per
sample, ranging from 35 to 225 kilobases (kb, Fig. 6). The northern stations 8 and 35
showed intense bands for the larger virus genomes (185, 210, and 225 kb) representing
on average 10% of the total virus community. Stations 15, 20, 27, and 33 showed only a
single 185 kb band of moderate intensity contributing for 6% of the total virus
community. The 65 kb band was the thickest of the smaller viral genomes (35, 45, 65,
85, 90, and 100 kb), corresponding to an average 38% (28 - 62%) of the total virus
community.
Figure 6. Virus PFGE fingerprint for each station studied. (A) The upper electrophoretic profiles
represent large virus genomes (50 - 300 kb, including putative algal viruses). Their genome size
was determined using the phage lambda concatamers ladder (M, panel A). (B) The bottom profiles
show smaller virus genomes (10 - 100 kb), which were compared to a 5 kb lambda ladder (M,
panel B). Some of the bands may not be visible here, but could be seen on the original picture.
Each gel lane was aligned with the ladder and shown precisely as it fell in the gel it was actually
run in (and relative to the standard in that gel).
82
Chapter 4
Viral lysis of picophytoplankton
3.4. Daily carbon flow
The daily carbon production (CP) by the total picophytoplankton community
varied largely between stations (Table 3). Despite their low abundance, the
picoeukaryotes (group I and II) contributed for 61% of total carbon production (omitting
stations 8 and 35 because of incomplete data set). The total picophytoplanktonic carbon
losses balanced the total carbon production for the most part, comprising on average 74
% (± 11). The carbon losses due to virally induced mortality ranged between 0.1 and 0.3
µg C L-1 d-1, which almost exclusively originated from the picoeukaryotes (97 ± 4 %).
The total losses due to viral lysis comprised 21% of the total picophytoplankton carbon
biomass produced per day (again excluding stations 8 and 35). Microzooplankton
consumed 0.1 to 0.9 µg C L-1 d-1, which originated equally from the picocyanobacteria
and the picoeukaryotes. The highest grazing mediated carbon losses were observed for
stations 20 and 33 (0.9 and 0.7 µg C L-1 d-1 respectively).
Table 3. Daily carbon production and daily carbon losses mediated by viral lysis and
microzooplankton grazing for the respective picophytoplankton groups.
Station
C production (µg C L-1 d-1)
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
Total
Total C losses (µg C L-1 d-1)
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
Total
Viral lysis C losses (µg C L-1 d-1)
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
Total
Grazing C losses (µg C L-1 d-1)
Picoeukaryotes I
Picoeukaryotes II
Synechococcus
Prochlorococcus
Total
8
15
20
27
33
0.11
≥ 0.11
0.08
0.14
0.05
0.13
0.40
0.37
0.72
0.11
0.30
1.50
0.07
0.18
0.14
0.05
0.44
0.19
0.61
0.31
0.24
1.35
0.13
0.09
1.14
≥ 1.36
0.09
≥ 0.09
0.09
0.08
0.08
0.00
0.25
0.40
0.51
0.13
0.09
1.13
0.08
0.10
0.18
0.04
0.40
0.43
0.21
0.24
0.04
0.92
0.19
0.12
0.06
≥ 0.37
0.07
≥ 0.07
0.07
0.04
0.01
0.00
0.12
0.18
0.00
0.00
0.00
0.18
0.06
0.05
0.00
0.01
0.11
0.23
0.01
0.00
0.00
0.24
0.19
0.04
0.05
≥ 0.28
0.02
≥ 0.02
0.02
0.04
0.07
0.00
0.13
0.21
0.50
0.13
0.09
0.94
0.02
0.05
0.18
0.03
0.28
0.20
0.20
0.24
0.05
0.68
0.00
0.08
0.01
≥ 0.09
83
35
Chapter 4
Viral lysis of picophytoplankton
4. Discussion
The traditional dilution assay originally developed to estimate
microzooplankton grazing on phytoplankton (Landry and Hassett 1982) is routinely used
in a wide range of marine systems, including open ocean oligotrophic habitats (Calbet
and Landry 2004). This is, however, the first published report of a viral lysis dilution
assay applied to an oligotrophic environment. The present study shows that viral lysis
can be an important loss factor for the picophytoplankton (< 3 µm in diameter) in the
oligotrophic waters in the subtropical northeastern Atlantic Ocean. In particular, the
picoeukaryotes group I (< 1.5 µm in diameter) was prone to high viral mediated
mortality (with rates up to 0.8 d-1).
4.1. Methodological aspects
It is important to note that this method only detects lysis of algal hosts that are
newly infected during the incubation period (see Baudoux et al. 2006 for more detailed
discussion). An important assumption of the dilution assay is that phytoplankton losses
are proportional to the dilution effect on the abundance of the mortality agents. This
implies that a single round of infection should be detected and, thus, that the host’s cell
lysis must occur later than 12 h after infection but within the 24 h incubation. Most
studies of algal host-virus model systems (including prokaryotic algal hosts) show that
the time to cell lysis upon viral infection is indeed within 24 h (Mann 2003; Brussaard
2004a). It is, furthermore, assumed that there is no preferential grazing by
microzooplankton. In the case of substantial preferential grazing on infected cells, viral
lysis would be underestimated as suggested earlier (Ruardij et al. 2005). Another aspect
of concern is that there is no substantial loss of virus during the incubation. Grazing by
heterotrophic nanoflagellates (HNF) could conceivably cause viral loss, but reported
daily rates of virus removal by HNF are rather low (≤0.3% d-1, Gonzalez and Suttle
1993).
In spite of these considerations, this method has the benefit to exclude the use
of conversion factors and to minimize the handling of the sample. Its utilization has been
validated across different algal host taxa (Evans et al. 2003; Baudoux et al. 2006). The
consistency of the viral lysis rates obtained using this methodology with other means for
assessing cell lysis furthermore provides confidence on the suitability of this method to
infer the effect of virus on phytoplankton mortality (Baudoux et al. 2006). Interestingly,
we found a strong positive linear relationship (r = 0.82, n = 6, p = 0.01) between the viral
lysis rates of picoeukaryote group I and the abundance of virus presenting an enhanced
stained DNA fluorescence signal (group V3) that most likely comprises algal viruses
(Brussaard 2004b).
84
Chapter 4
Viral lysis of picophytoplankton
4.2. Picoeukaryotes
Viruses were substantial mortality agents for the picoeukaryotic community,
responsible on average for 71% and 26% of the mortality of the picoeukaryotes group I
and group II respectively. Acknowledging that our data set only includes 6 sampling
stations, we did record clear differences in virally induced mortality rates of the
picoeukaryotes and in the relative effect of viruses across the system. The highest viral
lysis rates were recorded for picoeukaryote group I at the northern stations 8 and 35 (0.5
and 0.8 d-1), and the southwestern station 15 (0.4 d-1). At these stations, viral lysis was
responsible for 80 - 100% of the total picoeukaryote cell losses. Stations 8, 15, and 35
also had the highest contribution of viruses with larger-sized genomes (185, 210, and
225 kb), a characteristic feature of the virus family Phycodnaviridae that infect
eukaryotic algae (ICTVdB - The Universal Virus Database, version 4.
http://www.ncbi.nlm.nih.gov/ICTVdb/ICTVdB/).
The growth rate of the picoeukaryote group I at these stations with increased
viral lysis rates was higher (0.5 - 0.9 d-1) as compared to the other stations. Earlier
studies conducted under controlled conditions showed that the algal host’s growth rate
can influence the interactions between virus and host, hence viral lysis rate (Bratbak et
al. 1998). In these studies optimal growth conditions of the algal host resulted in
enhanced virus production. Thereby, we speculate that the relatively high growth rates of
the picoeukaryotes I at stations 8, 15, and 35 rates may have enhanced the impact of
viruses as mortality agents (0.5 to 0.8 d-1).
The contribution of microzooplankton grazing as loss factor was highest in the
southeastern stations 20 and 33. At these stations, grazing comprised 50% of total
mortality for the picoeukaryote group I (0 - 30% at the other stations) and 95 - 100% for
the picoeukaryote group II (50% at the other stations). Although it is difficult to
conclude on the factors underlying the difference in mortality processes (viral lysis vs.
microzooplankton grazing), it is noteworthy that picoeukaryote abundance was 4 to 10
fold higher at these stations (20 and 33) than at other stations. These differences could be
due to different water types that may originate from the proximity of coastal areas and/or
the Canary Current that flows along the African coast and entrains upwelling waters. The
higher abundance in picoeukaryotes at these southeastern stations may exceed the
threshold level of prey for microzooplankton, resulting in enhanced grazing rates.
Selective grazing can also contribute to the different phytoplankton community.
4.3. Prokaryotes
Despite their numerical dominance, Synechococcus and Prochlorococcus
showed low virally induced mortality rates. Viral lysis rates averaged 0.02 ± 0.05 d-1 for
Synechococcus and 0.02 ± 0.03 d-1 for Prochlorococcus, which corresponds to a removal
of 1 and 3% of the standing stock per day. These estimates are comparable to previous
results indicating that c.a. 3% d-1 of Synechococcus population undergo viral lysis as
85
Chapter 4
Viral lysis of picophytoplankton
determined by indirect viral lysis assays (as cited in Garza and Suttle 1998). In offshore
oligotrophic waters, cyanobacteria are found in relatively high numbers but infectious
Synechococcus and Prochlorococcus phages occur in low abundance (Sullivan et al.
2003). Sullivan and collaborators (2003) even argued that these “low [cyanophages]
titers in areas of high host abundance seem to be a feature of the open ocean
ecosystems”. Such a situation will result in reduced contact rates between the potential
host and co-occurring virus and may have been a reason of the low reported viral lysis
rates for the prokaryotes. In agreement with the low viral lysis rates the percentage dead
cells, as determined using the SYTOX Green dye, was low (Veldhuis M. J. W. unpubl.).
Agusti (2004) also observed a good viability of the autotrophic prokaryotes in the same
geographical area.
The generally high genotypic diversity within the populations of Synechococcus
and Prochlorococcus (Scanlan and West 2002), as well as the ability of cyanobacteria to
acquire cyanophage resistance (Waterbury and Valois 1993) may have been other factors
reducing the effect of cyanophages as mortality agents. As hypothesized earlier for the
picoeukaryotes, low virally induced mortality may also be related to the moderately low
growth rates recorded for Synechococcus (0.2 ± 0.1 d-1) and Prochlorococcus (0.2 ± 0.05
d-1, with the exception of station 35 where growth reached 1.3 d-1) during our study.
Synechococcus typically distributes in the surface oceanic layers and lower growth rates
at Deep Chlorophyll Maximum depths are usually reported (Partensky et al. 1996, Liu et
al. 1998). Conversely, Prochlorococcus cells do extend into deeper waters where they
generally grow faster than Synechococcus (ca. 0.69 d-1, Partensky et al. 1999). Growth
rates as low as 0.1 d-1 have, however, been reported at similar depth in northern
subtropical Pacific Ocean (Liu et al. 1995), which substantiate the moderately low values
here reported.
Synechococcus was mainly controlled by microzooplankton grazing, and the
recorded rates (0.1 - 0.25 d-1) were consistent with those obtained in the same
geographical area at similar depth (on average 0.3 d-1, Quevedo and Anadòn 2001).
Grazing on Prochlorococcus was substantially lower (on average 0.04 ± 0.03 d-1) than
those observed for Synechococcus, indicating that both prokaryotes undergo different
loss mechanisms. The occurrence of differential grazing losses has already been reported
for Synechococcus and Prochlorococcus under controlled conditions. For example, some
ciliates showed a marked preference for Synechococcus against Prochlorococcus
(Christaki et al. 1999). Similar observations were reported for some heterotrophic
nanoflagellates (HNF), but preferential grazing against Prochlorococcus only occurred
when both preys coexisted at similar concentrations (Guillou et al. 2001). In the case of
HNF grazing on the autotrophic picocyanobacteria we cannot exclude that a trophic
cascade was induced that may have resulted in reduced grazing rates. The overall low
total loss rates in combination with the moderately low growth rates may partially
explain the numerical dominance of Prochlorococcus (60 - 94% of total abundance) at
the studied stations.
86
Chapter 4
Viral lysis of picophytoplankton
4.4. Implications for geochemical fluxes
The total carbon produced by the picophytoplankton at the DCM in the
northeastern subtropical Atlantic Ocean (October 2002) was largely balanced by the
losses due to grazing and viral lysis. This nicely illustrates a situation close to steadystate, as can be expected for oligotrophic ecosystems. The carbon losses mediated by
viral lysis accounted on average for 21% of the total CP by the picophytoplankton
community. This percentage is higher than the 2 - 10% loss assumed by Wilhelm and
Suttle (1999) applying the revised steady state model of Jumars et al. (1989). The latter
lower values may be caused by the limited number of studies that were incorporated into
the model, and/or by the restricted number of potential algal host taken into account. But
Wilhelm and Suttle (1999) also grouped all phytoplankton without making distinction
between eukaryotes and prokaryotes. Our results clearly show that the influence of viral
lysis on carbon cycling varied depending on the picophytoplankton group.
Picoeukaryotes were responsible for 97% of the cellular carbon released by viral lysis.
Hence, it appears essential to investigate viral lysis for all co-occurring phytoplankton
groups to obtain a better insight into the influence of algal viruses on the carbon cycling.
Next to the considerable release of carbon through picoeukaryotes viral lysis, a
parallel study conducted during IRONAGES III suggested that lysis of the
picoeukaryotes group I could be an important source of Fe-organic ligand at the Deep
Chlorophyll Maximum (Gerringa et al. 2006). Iron (Fe) availability is considered to be
an important co-limiting factor for the productivity of marine ecosystem. Most of the
total dissolved Fe in seawater is complexed by dissolved organic ligands, and
constitutes, as such, the largest potential pool of bioavailable Fe to marine plankton (Rue
and Bruland 1995). Because virally induced mortality was the main source of cell loss
for the picoeukaryotes group I (this study), viral lysis may well play a critical role in the
recycling of organically complexed Fe in the oligotrophic northeastern subtropical
Atlantic Ocean. Other authors have drawn similar conclusion for laboratory and
experimental studies (Poorvin et al. 2004).
In contrast to the picoeukaryotes, the prokaryotes did not substantially
contribute to the release of carbon through viral lysis. Instead, all of the prokaryotes
carbon losses are fully accounted for by microzooplankton grazing (97±5%). Overall,
microzooplankton consumed 52±14% of the total picophytoplanktonic carbon
production. This value is only slightly lower than the averaged estimate of 67% of
phytoplankton production consumed by microzooplankton which was based on a large
dataset (Calbet and Landry 2004). Our sampling at the Deep Chlorophyll Maximum
might be responsible for this difference (the sampling depth considered in Calbet and
Landry (2004) is not clear).
In summary, the adapted dilution assay that estimates both viral lysis and
microzooplankton grazing was successfully applied to the oligotrophic waters of the
northeastern subtropical Atlantic. The effect of viruses on picophytoplankton mortality
was algal group-specific and particularly high for the picoeukaryotes. Given that this
87
Chapter 4
Viral lysis of picophytoplankton
work represents the first quantification of viral mortality rates on picophytoplankton, the
prevalence of such rates for discrete populations should be explored in other marine sites
to verify their broader applicability.
Acknowledgments. We thank the captain and crew of the R.V. Pelagia, and especially
chief scientist Klaas Timmermans for the opportunity to join the IRONAGES III cruise.
We thank Marieke Bossink, Margriet Hiehle, Nelleke Krijgsman, Swier Oosterhuis and
the nutrient service lab for technical support. The reviewers as well as the associate
editor are acknowledged for their constructive comments on the manuscript. This work
was supported by the Research Council for Earth and Life Sciences (ALW) with
financial aid from the Netherlands Organization for Scientific Research (NWO).
Literature cited
Agusti S (2004) Viability and niche segregation of Prochlorococcus and Synechococcus
cells across the Central Atlantic Ocean. Aquat Microb Ecol 36:53-59.
Agusti S, Satta MP, Mura MP, Benavent E (1998) Dissolved esterase activity as a tracer
of phytoplankton lysis: Evidence of high phytoplankton lysis rates in the
northwestern Mediterranean. Limnol Oceanogr 43:1836-1849.
Baudoux A-C, Noordeloos AAM, Veldhuis MJW, Brussaard CPD (2006) Virally
induced mortality of Phaeocystis globosa during two spring blooms in temperate
coastal waters. Aquat Microb Ecol 44:207-217.
Bertlisson S, Berglund O, Karl DM, Chisholm SW (2003) Elemental composition of
marine Prochlorococcus and Synechococcus: Implications for the ecological
stoichiometry of the sea. Limnol Oceanogr 48:1721-1731.
Bratbak G, Jacobsen A, Heldal M, Nagasaki K, Thingstad F (1998) Virus production in
Phaeocystis pouchetii and its relation to host cell growth and nutrition. Aquat
Microb Ecol16:1-9.
Brussaard CPD (2004a) Viral control of phytoplankton populations - a review. J Euk
Microbiol 51:125-138.
Brussaard CPD (2004b) Optimization of procedures for counting viruses by flow
cytometry. Appl Env Microbiol 70:1506-1513.
Calbet A, Landry MR (2004) Phytoplankton growth, microzooplankton grazing, and
carbon cycling in marine systems. Limnol Oceanogr 49:51-57.
Christaki U, Jacquet S, Dolan JR, Vaulot D, Rassoulzadegan F (1999) Growth and
grazing on Prochlorococcus and Synechococcus by two marine ciliates. Limnol
Oceanogr 44:52-61.
DuRand MD, Green RE, Sosik HM, Olson RJ (2002) Diel variations in optical
properties of Micromonas pusilla (Prasinophyceae). J Phycol 38:1132-1142.
Evans C, Archer SD, Jacquet S, Wilson WH (2003) Direct estimates of the contribution
88
Chapter 4
Viral lysis of picophytoplankton
of viral lysis and microzooplankton grazing to the decline of a Micromonas spp.
population. Aquat Microb Ecol 30:207-219.
Garza DR, Suttle CA (1998) The effect of cyanophages on the mortality of
Synechococcus spp. and selection for UV resistant viral communities. Microb
Ecol 36:281-292.
Gerringa LJA, Veldhuis MJW, Timmermans KR, Sarthou G, de Baar HJW (2006) Covariance of dissolved Fe-binding ligands with phytoplankton characteristics in the
Canary Basin. Mar Chem 102:276-290.
Gonzalez JM, Suttle CA (1993) Grazing by marine nanoflagellate on viruses and virussized particles: ingestion and digestion. Mar Ecol Prog Ser 94:1-10.
Guillou L, Jacquet S, Chrétiennot-Dinet M-J, Vaulot D (2001) Grazing impact of two
small heterotrophic flagellates on Prochlorococcus and Synechococcus. Aquat
Microb Ecol 26: 201-207.
Jumars PA, Penry DL, Baross JA, Perry MJ, Frost BW (1989) Closing the microbial
loop - Dissolved carbon pathway to heterotrophic bacteria from incomplete
ingestion, digestion and absorption in animals. Deep-Sea Res I 36:483-495.
Landry MR, Hassett RP (1982) Estimating the grazing impact of marine microzooplankton. Mar Biol 67:283-288.
Landry MR, Constantinou J, Latasa M, Brown SL, Bidigare RR, Ondrusek ME (2000)
Biological response to iron fertilization in the eastern equatorial Pacific (IronEx
II). Dynamics of phytoplankton growth and microzooplankton grazing. Mar Ecol
Prog Ser 201:57-72.
Larsen A, Castberg T, Sandaa RA, Brussaard CPD, Egge J, Heldal M, Paulino A,
Thyrhaug R, van Hannen EJ, Bratbak G (2001) Population dynamics and
diversity of phytoplankton, bacteria and viruses in a seawater enclosure. Mar Ecol
Prog Ser 221:47-57.
Li WKW (1994) Primary production of prochlorophytes, cyanobacteria, and eukaryotic
ultraphytoplankton - Measurements from flow cytometric sorting. Limnol
Oceanogr 39:169-175.
Liu HB, Campbell L, Landry MR (1995) Growth and mortality rates of Prochlorococcus
and Synechococcus measured with a selective inhibitor technique. Mar Ecol Prog
Ser 116:277-287.
Liu HB, Campbell L, Landry MR, Nolla HA, Brown SL, Constantinou J (1998)
Prochlorococcus and Synechococcus growth rates and contributions to production
in the Arabian Sea during the 1995 Southwest and Northeast Monsoons. DeepSea Res II 45:2327-2352.
Mann NH (2003) Phages of the marine cyanobacterial picophytoplankton. FEMS
Microbiol Rev 27:17-34.
Moon-van der Staay SY, De Wachter R, Vaulot D (2001) Oceanic 18S rDNA sequences
from picoplankton reveal unsuspected eukaryotic diversity. Nature 409:607-610.
Mühling M, Fuller NJ, Millard A, Somerfield PJ, Marie D, Wilson WH, Scanlan DJ,
Post AF, Joint I, Mann NH (2005) Genetic diversity of marine Synechococcus and
co-occurring cyanophage communities: Evidence for viral control of
phytoplankton. Environ Microbiol 7:499-508.
89
Chapter 4
Viral lysis of picophytoplankton
Partensky F, Blanchot J, Lantoine F, Neveux J, Marie D (1996) Vertical structure of
picophytoplankton at different trophic sites of the tropical northeastern Atlantic
Ocean. Deep-Sea Res I 43:1191-1213.
Partensky F, Hess WR, Vaulot D (1999) Prochlorococcus, a marine photosynthetic
prokaryote of global significance. Microbiol Molec Biol Rev. 63:106-127.
Poorvin L, Rinta-Kanto JM, Hutchins DA, Wilhelm SW (2004) Viral release of iron and
its bioavailability to marine plankton. Limnol Oceanogr 49:1734-1741.
Quevedo M, Anadòn R (2001) Protist control of phytoplankton growth in the subtropical
northeast Altlantic. Mar Ecol Prog Ser 221:29-38.
Raven JA (1998) The twelfth Tansley Lecture. Small is beautiful: the
picophytoplankton. Funct Ecol 12:503-513.
Ruardij P, Veldhuis MJW, Brussaard CPD (2005) Modeling the bloom dynamics of the
polymorphic phytoplankter Phaeocystis globosa: impact of grazers and viruses.
Harmful Algae 4:941-963.
Rue EL, Bruland KW (1995) Complexation of iron (III) by natural organic-ligands in the
Central North Pacific as determined by a new competitive ligand equilibration
adsorptive cathodic stripping voltammetric method. Mar Chem 50:117-138.
Scanlan DJ, West NJ (2002) Molecular ecology of the marine cyanobacterial genera
Prochlorococcus and Synechococcus. FEMS Microb Ecol 40:1-12.
Short SM, Suttle CA (2003) Temporal dynamics of natural communities of marine algal
viruses and eukaryotes. Aquat Microb Ecol 32:107-119.
Sullivan MB, Waterbury JB, Chisholm SW (2003) Cyanophages infecting the oceanic
cyanobacterium Prochlorococcus. Nature 424:1047-1051.
Veldhuis MJW, Kraay GW (2004) Phytoplankton in the subtropical Atlantic Ocean:
towards a better assessment of biomass and composition. Deep-Sea Res. I 51:507530.
Waterbury JB, Valois FW (1993) Resistance to co-occurring phages enables marine
Synechococcus communities to coexist with cyanophages abundant in seawater.
Appl Environ Microb 59:3393-3399.
Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea. BioScience
49:781-788.
Worden AZ, Nolan JK, Palenik B (2004) Assessing the dynamics and ecology of marine
picophytoplankton: The importance of the eukaryotic component. Limnol
Oceanogr 49:168-179.
90
Chapter 5
Phytoplankton losses in the North Sea
during summer 2003 1
Anne-Claire Baudoux, Marcel J. W. Veldhuis, Anna A. M. Noordeloos, Govert van
Noort, and Corina P. D. Brussaard
This study investigated microzooplankton grazing and virally mediated
mortality of the eukaryotic (3 size classes) and prokaryotic (Synechococcus)
picophytoplankton across the North Sea during summer 2003. Our results showed that
the fate of picophytoplankton differed among algal groups and their area of occurrence.
Highest viral lysis rates (0.16 – 0.23 d-1) were recorded for specific picoeukaryotic
groups in the coastal region and the station investigated at the DCM. The virally
induced turnover rate of the host abundance was around 20% d-1, which corresponded
with the percentage dead cells observed for these respective groups (13 to 32%; using
the live/dead dye SYTOX Green). Viral lysis was insignificant at the other stations,
despite considerable percentage dead algal cells (up to 38%). Microzooplankton
actively grazed upon picoeukaryotes (average per group 0.2 - 0.4 d-1), independent of
the region studied. Grazing on Synechococcus was restricted to the coastal waters (0.7
d-1). Grazing by microzooplankton consumed on average 40 ± 27% of the carbon
produced by picophytoplankton, constituting in general the main loss factor. Viral lysis
released on average 8 ± 13% of the total picophytoplankton carbon production, and
locally up to 32%.
1
To be submitted to Aquatic Microbial Ecology
91
Chapter 5
Mortality of marine picophytoplankton
1. Introduction
Marine phytoplankton are responsible for up to half of the global primary
production on Earth, and hence play a critical role in global carbon cycling (Geider et al.
2001). Traditionally, phytoplankton cells were treated as immortal unless they were
preyed upon by zooplankton or lost by sedimentation through the water column. Over
the past decades, our view of phytoplankton mortality has been transformed with the
finding that algal cells can die by lysis. The reports of substantial algal cell lysis rates
(Brussaard et al. 1995, 1996a, Agusti et al. 1998) and high fraction of dead
phytoplankton cells (Veldhuis et al. 2001, Alonso-Laita & Agusti 2006) in marine
environments emphasized the significance of cell death for phytoplankton mortality.
Algal cell death by lysis can be caused by different mechanisms amongst which viral
infection (Suttle et al. 1990, Brussaard 2004a) and environmental stresses, such as
intense light (Berman-Frank et al. 2004), darkness (Berges & Falkowski 1998), or
nutrient depletion (Brussaard et al. 1997, Berman-Frank et al. 2004).
The partitioning of phytoplankton mortality into sinking, grazing, and cell lysis
is important because these loss factors affect the structure and the functioning of the
pelagic microbial food web differently. Phytoplankton sedimentation results in the
transfer particulate organic matter from the pelagic towards the benthic ecosystems
(Smetack 1985), whilst the grazing will mostly channel phytoplankton biomass to the
higher trophic levels in the pelagic food web. Through cell lysis, phytoplankton biomass
is converted to dissolved organic matter that becomes available for bacteria, hence
forcing the food web towards a more regenerative pathway (Wilhelm & Suttle 1999 and
reference therein). Knowledge on the relative contribution of these loss factors for
phytoplankton mortality is thus critical for an optimal understanding of the flow of
energy and nutrient in marine environments.
Extensive field studies conducted in the North Sea indicated a pronounced
temporal variation in the ecological relevance of these different loss factors (Riegman et
al. 1993, Brussaard et al. 1995, 1996a, Kuipers & Witte 1999). During the eutrophic
spring conditions in this temperate region, grazing by microzooplankton sets the limit of
the small-sized phytoplankton whereas larger counterparts can form blooms while they
escape grazing due to their size (Riegman et al. 1993, Brussaard et al. 1996a).
Sedimentation was reported to be responsible for the termination of the diatom bloom in
early spring, whereas algal cell lysis accounted for up to 75% of demise of the following
Phaeocystis globosa bloom (Brussaard et al. 1995). Recent investigations indicated that
viruses were the primary cause of cell lysis (Brussaard et al. 2005a, Baudoux et al.
2006).
The substantial nutrient consumption during these spring blooms forces the
system towards a more oligotrophic status by summer. In summer, smaller-sized
phytoplankton dominate the algal community (Kuipers & Witte 1999). These algal cells
are not likely to sink considering their micrometer range size (Raven 1998), and
microzooplankton is considered the major loss factor (Kuipers & Witte 1999). However,
virtually no data exist on the significance of virally mediated mortality for phytoplankton
92
Chapter 5
Mortality of marine picophytoplankton
in the North Sea during summer (Brussaard et al. 1996b).
Our study compared the relative contribution of both viral lysis and
microzooplankton grazing of the smaller-sized phytoplankton during oligotrophic
summer conditions and explored the variability of these loss factors across the North
Sea.
2. Material and methods
Study area and sampling. The data were obtained during the MOMAP-2
shipboard expedition in the North Sea from 8 to 20 July 2003, aboard the R.V. Pelagia
(Fig. 1). During this cruise, 11 pelagic stations were sampled for physical and chemical
parameters, microbial abundance, and phytoplankton viability (cell membrane integrity).
Among these 11 stations, 5 main stations were also investigated for microzooplankton
grazing and viral lysis. Samples were collected in 10 L NOEX bottles mounted on the
Rosette sampler equipped with Seabird conductivity-temperature-depth (CTD).
Typically, samples were collected at 6 different depths covering the upper 65 m of the
water column (except for station 2 and 12, as max. depth was 45 m). The euphotic zone
(1% light penetration) comprised the upper 23 to 45 m.
Latitude
-10°
60°
-5°
0°
5°
10°
6
60°
8
5
4
55°
9 10
12
13
3
14
2
0
50°
-10°
15°
-5°
0°
5°
55°
km
200 400
10°
50°
15°
Longitude
Figure 1. Location of the sampling stations during the MOMAP-2 expedition conducted in the
North Sea from 8 to 20 July 2003.
Physical and chemical parameters. The salinity, temperature and light
intensities were measured by the CTD Seabird mounted on the Rosette sampler equipped
with a PAR detector. A PAR- detector was also mounted on deck to determine incident
light. Nutrient samples (5 mL) were filtered through 0.2 µm pore-size polysulfone filters
(Acrodisc, Gelman Sciences). Analyses were performed on board using a TrAAcs 800
93
Chapter 5
Mortality of marine picophytoplankton
autoanalyzer for dissolved orthophosphate, nitrate, nitrite, ammonium, and silicate as
described in Baudoux et al. (2006). The limit of detection was 0.007 µM for phosphate,
0.15 µM for ammonium, 0.002 µM for nitrite, 0.03 µM for nitrate, and 0.02 µM for
silicate.
Phytoplankton community. Chlorophyll a (Chl a) samples (typically 1.5 l)
were filtered onto a GF/F filter (Whatman) and stored at -80°C until analysis. Chl a was
extracted in 90% acetone and measured fluorimetrically.
Phytoplankton abundance from natural as well as experimental samples was
enumerated directly after sampling using a Beckman Coulter XL-MCL flow cytometer.
The instrument was equipped with a laser with an excitation wavelength of 488 nm
(15mW) and emission bands for the chlorophyll autofluorescence (> 630 nm) and
phycoerythrin (PE 575 ± 20 nm). The discriminator for phytoplankton was the red
chlorophyll autofluorescence. Flow rate (135 ± 7 µL min-1) and machine drift were
checked every day using calibrated beads (Flow-Check Fluorospheres, Beckman
Coulter) as internal standard. Based on the pigment autofluorescence and forward
scatter, we discriminated the prokaryotes Synechococcus spp. as well as 5 populations of
eukaryotic cells. The division of the eukaryotes was based on their relative size as
determined by flow cytometry using a series of fractionated samples according to
Veldhuis & Kraay (2004). Using this procedure, the cyanobacterium Synechococcus had
an equivalent spherical diameter (esd) of 1.1 µm. Three picoeukaryotic groups were
discriminated with esd ranging between 1.1 and 1.5 µm (Pico I), 1.5 and 2.0 µm (Pico
II), and 2.0 and 3.0 µm (Pico III). Furthermore, we defined 2 nanoeukaryotic groups, in
lesser abundance, with esd ranging between 3.0 and 6.0 µm (Nano I), and 10 and 20 µm
(Nano II). The determination of the cellular carbon content of each phytoplankton group
was based on biovolume and assuming phytoplankton cells to be spherical. For
Synechococcus, we used an averaged biovolume to carbon conversion factor of 260 fg C
µm-3 derived Synechococcus WH8103 and WH8012 (average diameter 1.02 µm;
Bertlisson et al. 2003). For the picoeukaryotes, we used a biovolume to carbon
conversion factor of 239 fg µm-3, an average of the values obtained for Ostreococcus sp.
CCE9901 (233 – 247 fg C µm-3; Worden et al. 2004) and Micromonas pusilla CCMP
489 (238 fg C µm-3; DuRand et al. 2002). A biovolume to conversion factor of 160 fg
µm-3 was used for the nanoeukaryotes according to Verity et al. (1992). The resulting
averaged cellular carbon content were 0.18 ± 0.004 pg cell-1(mean ± SD) for
Synechococcus, 0.22 ± 0.02 pg cell-1 for Pico I, 0.52 ± 0.24 pg cell-1 for the Pico II, 2.3 ±
1.1 pg cell-1 for Pico III, and 8.0 ± 5.4 pg cell-1 for Nano I and 321 ± 707 pg cell-1 for
Nano II.
The abundance of dead algal cells was determined using the nucleic acidspecific stain SYTOX Green (Molecular Probes, Invitrogen) to test cell membrane
permeability (Brussaard et al. 2001). SYTOX Green can only penetrate and stain cells
with a compromised plasma membrane (i.e., by definition dying cells). Briefly, samples
were stained with SYTOX Green (final concentration 0.5 µM) for 15 min in the dark
prior to flow cytometric analysis using a Beckman Coulter XL-MCL flow cytometer (see
94
Chapter 5
Mortality of marine picophytoplankton
above section for instrument specification). Within each phytoplankton subgroup, cells
were separated in live and dead on the basis of their relative green fluorescence; dead
cells had >5 times the green autofluorescence fluorescence of live cells. Only the
numerically dominating picophytoplankton groups (Synechococcus, and Pico I, II, and
III) allowed proper analysis (typically down to a depth of 45 m). For practical reasons,
we restricted generally the analysis of live/dead algal cells to 1 station per region (region
1 was represented by station 10, region 2 by station 2, and region 3b by station 6). More
stations (3, 4, 5, and 8) were, however, analyzed for region 3a in order to test the results
consistency for different stations within one region.
DNA Green Fluorescence (r.u.)
Virus community. The abundance of putative algal viruses was determined on
glutaraldehyde fixed samples (final concentration 0.5% glutaraldehyde, frozen in liquid
nitrogen and stored at -80°C prior analysis) according to Brussaard (2004b), using a
Beckton-Dickinson FACSCalibur flow cytometer equipped with a 15 mW 488 nm aircooled argon-ion laser and a standard filter set up. The discriminator for virus
enumeration was the green fluorescence. Thawed samples were 100-fold diluted in 0.2
µm filtered sterile TE-buffer (pH 8) and stained with 5 × 10-5 of the commercial SYBR
Green for 10 min at 80°C in darkness. Counts were corrected for the blank consisting of
TE-buffer with sterile 0.2 µm filtered seawater in the correct dilution. Based on their
relative green fluorescence and side scatter signature, we distinguished 4 virus
subpopulations (V1, V2, V3, and V4, Fig. 2). Considering the viral FCM signatures of
algal viruses brought in culture (Brussaard 2004b), the viral group V3 included to some
extent putative algal viruses whereas group V4 consisted mainly of algal viruses.
104
beads
103
beads
bacteria
bacteria
102
V4
V4
V3
V2
V1
101
100
100
101
102
V3
V2
V1
103
104 100
Side scatter (r.u.)
101
102
103
104
Side scatter (r.u.)
Figure 2. Cytogram of natural viral community. A maximum of four viral groups (V1, V2, V3, and
V4) were discriminated based on the intensity of their green fluorescence after staining with the
nucleic acid-specific dye SYBR Green I and the side scatter. V4 was assigned to putative algal
viruses. V3 most likely also contains algal viruses. An internal standard (yellow green fluorescent
1 µm beads, Invitrogen – Molecular Probes) was added to the sample.
95
Chapter 5
Mortality of marine picophytoplankton
Viral diversity was examined in the surface (5 m) for all stations and at the
DCM of stations 2, 3, and 14 using pulsed field gel electrophoresis (PFGE) as described
by Larsen et al. (2001). Summarizing, a concentrate of 2 liters water samples was
plugged in molten InCert agarose (1.5 % (w/v), Cambrex Bioscience, Rockland, ME
USA) and digested overnight at 30°C in a lysis buffer. Samples were loaded onto a 1%
SeaKem GTG agarose gel (Cambrex Bioscience, Rockland, ME) and electrophored
using a Bio-Rad DR-II CHEF Cell unit operating at 6 V cm-1 at 14°C in 0.5× TBE tank
buffer. Two pulse ramp settings were used for an optimal sizing of viral genomes (1) 1-6
s for 20 h to examine the smaller virus genomes ranging between 10 – 100 kb and (2) 830 s for 20 h to discriminate the larger virus genomes comprised 50 – 400 kb. After
electrophoresis, gels were stained for 1 h with SYBR Green I (1 × 10-4 of commercial
solution) and destained 10 min in MilliQ before a digital analysis for fluorescence using
a FluorS imager (Bio-Rad Instruments). Viral genomes sizing was performed against a 5
kb lambda ladder or a lambda concatamers ladder (Bio-Rad, Richmond, CA). We
determined the relative abundance of the different viral genome sizes using the
Hyperladder VI (Gentaur) DNA standard with known size and amount of DNA. Large
genome-sized dsDNA viruses are most likely algal viruses (Mann 2003, Brussaard
2004a); therefore viruses with genome sizes >100 kb were identified as putative algal
viruses during this study. We realize that algal viruses may present genome size <100
kb, these smaller genome sized virus could not be included in the analysis because many
of the virus with genome <100 kb viruses are phages infecting heterotrophic
prokaryotes.
Phytoplankton mortality assays. Phytoplankton viral lysis and
microzooplankton grazing rates were, for practical reasons, determined in surface (5 m)
for one representative station per region only (region 1 was represented by station 10,
region 2 by station 2, and region 3b by station 6, Table 1), except for region 3a for which
2 stations were investigated (stations 3 and 5) to allow intraregion-specific variation.
Furthermore, only for station 3 an additional experiment could be performed at the deep
chlorophyll a maximum depth (DCM, 37 m). Sample collection was performed by
carefully siphoning from NOEX bottles into darkened carboys to prevent light stress
when the sample was brought on deck. All material used for these experiments were acid
cleaned (0.1N HCl), rinsed with MilliQ and finally with the same water as the sample.
The experimental set up was performed in a controlled room at in situ temperature (10 19 °C) and in dimmed light.
We estimated viral lysis and microzooplankton grazing rates using an adapted
dilution technique by Baudoux et al. (2006) which was based on the assay by Evans et
al. (2003). The data analysis was restricted to the numerically dominating
picophytoplankton groups (Synechococcus, Pico I, II, and III), as algal abundance needs
to be high enough to allow detection after a 5-fold dilution (standard dilutions 20, 40, 70,
and 100% of natural water). The natural seawater (<200 µm) was diluted with 0.2 µm
pore-size filtered diluent to obtain the microzooplankton grazing rate (Landry & Hassett
1982). An additional dilution series with 30 kDa filtered natural sample was performed
96
Chapter 5
Mortality of marine picophytoplankton
to provide the loss rate resulting from grazing and viral lysis. A detailed experimental
design can be found in Baudoux et al. (2006). A 5 mL subsample was taken (T=0) from
the soft polycarbonate incubation bottles (triplicate) upon filling, after which the bottles
were closed without trapping air bubbles inside. All bottles were mounted on a slowly
rotating (0.5 rpm) plankton wheel. Incubation temperature (10 – 19°C) and irradiance (3
– 100 µmol quanta m-2 s-1, light period of 14 h) were adjusted to the in situ conditions
given by the CTD and PAR detectors. After 24 h incubation, another 5 mL subsample
was taken to monitor phytoplankton growth using flow cytometry. The apparent growth
rate (µapp, d-1) was calculated for each sample from the changes in abundance during the
incubation. The regression coefficient of apparent growth rate vs. dilution factor for the
30 kDa series represents phytoplankton losses due to microzooplankton grazing and viral
lysis (Mg+v, d-1), whereas the regression coefficient resulting from the 0.2 µm dilution
series represents only the microzooplankton grazing rate (Mg, d-1). Specific virally
induced mortality rates (d-1) were thus obtained from the difference between Mg+v and
Mg. Specific growth rates (µ, d-1) were determined as the y-axis intercept value of the
regression line obtained with the 30 kDa series. The significance (p) of each slope (Mg
and Mg+v) and the intercept (µ) were determined performing a t-test on the regression
analysis. The significance (p) between the slopes of the regressions lines (i.e.,
significance of Mv) was also estimated using a t-test.
Table 1. Location and characteristics of the 5 main stations representative of the different water
masses. Region 1 is represented by station 10; region 2 by station 2; region 3a by stations 3 and 5,
and region 3b by station 6.
Station
2
3
3
5
6
10
Latitude (°N)
Longitude (°W)
Depth sample (m)
Temperature (ºC)
Salinity
NO3 (µM)
NO2 (µM)
NH4 (µM)
PO4 (µM)
Si (µM)
54.52
4.70
5
16.0
34.2
0.03
0.01
0.16
0.02
0.04
55.83
2.67
5
15.5
35.0
0.06
0.01
0.12
0.03
0.81
55.83
2.67
37
9.9
35.0
0.35
0.06
0.60
0.24
2.9
59.01
-0.02
5
14.4
35.2
0.24
0.01
0.27
0.03
0.21
60.48
-3.89
5
12.8
35.3
2.33
0.11
0.19
0.11
0.34
57.83
7.70
5
19.2
31.7
0.06
0.01
0.12
0.01
0.27
Chl a (µg L-1)
0.51
0.25
0.83
0.28
1.77
0.68
97
Chapter 5
Mortality of marine picophytoplankton
In parallel, the protistian grazing rates on cyanobacteria were also determined
using the fluorescently labeled prey approach. Labeling of Synechococcus sp. S1 (~ 1
µm diameter, NIOZ culture collection) was performed prior to the cruise with 4,6dichlorotriazin-2-yl aminofluorescein (DTAF, Sigma-Aldrich, St Louis, MO) according
to Sherr and Sherr (1993). The fluorescently labeled cyanobacteria (FLC) were
enumerated by epifluorescence microscopy (Zeiss, Axioplan 12500 X) and aliquots of
known concentration were stored at –20°C until use. On board, FLC were diluted to
<10% (v/v) of the natural cyanobacterial abundance by adding natural samples
(triplicate, 1 l polycarbonate bottles). A 20 mL subsample (T=0) was taken upon filling
and immediately fixed with 10% glutaraldehyde (1% final concentration). The fixed
sample was filtered onto a 0.2 µm pore-size black membrane polycarbonate filter
(Whatman) and stored at -20°C until analysis. Bottles were closed and incubated as
described for the modified dilution assay. After 24 h incubation, another 20 mL
subsample was taken and treated similarly for monitoring the abundance of FLC. The
estimation of the grazing rates (d-1) of Synechococcus was based on the decrease in
abundance of FLC.
A carbon budget was determined combining the cellular carbon content
estimates (see phytoplankton community section) and data of the dilution experiments.
For each specific phytoplankton group, the carbon production (CP, in µg C L-1 d-1),
losses due to grazing (G, µg C L-1 d-1) and, by adaptation, the losses due to viruses (V, in
µg C L-1 d-1) were calculated using the formulas of Landry et al. (2000); CP = µ × Pm ; G
= Mg × Pm ; V = Mv × Pm; and Pm = P0 × [e(µ - Mg+v) t - 1]/(µ - Mg+v)t, where µ (in d-1) is the
dilution-based specific growth (y intercept of the 30 kDa regression); Mg and Mv are the
dilution-based grazing and viral lysis rates (in d-1), P0 (in µg C l-1) is the initial carbon
biomass of picophytoplankton, Pm (in µg C l-1) is the geometric mean carbon biomass of
picophytoplankton during the incubation and t (in d) is the time of incubation.
3. Results
3.1. Physical and chemical characterization of the studied area
Based on the T-S diagrams, 4 hydrographic regions were distinguished in the
studied area (Fig. 3). Region 1 included the Norwegian coastal stations (9 and 10) and
was characterized by large gradients in salinity (29 to 34.5) and temperature (6 to
19.5°C). Region 2 corresponded to the southeastern North Sea (stations 2, 12, and 14),
characterized by intermediate salinity (34.2 to 34.5) and temperatures between 6 and
18°C. The central and northern North Sea waters (stations 3, 4, 5, 6, 8, and 13) defined a
third region, characterized by the most saline waters (34.7 to 35.5). Within this latter
region, a gradient of increasing salinity was observed in a northward direction. Based on
98
Chapter 5
Mortality of marine picophytoplankton
these criteria, we distinguished a region 3a including stations 3, 4, 5, 8, and 13 (salinity
range within 34.7 - 35.2 and temperature within 7.1 and 16.6°C) and a region 3b
represented by station 6 (salinity range within 35.0 - 35.4 and temperature within 9.4 and
12.7 °C). A thermal stratification was observed in regions 1, 2, and 3a but not in region
3b. Among these 3 stratified regions, only regions 2 and 3a presented a well defined
surface mixed layer extending to about 15 m.
Temperature (°C)
20
1
2
18
3a
16
14
12
3b
10
8
6
28
30
32
34
Salinity
station 2
station 3
station 4
station 5
station 6
station 8
station 9
station 10
station 12
station 13
station 14
36
Figure 3. Temperature-Salinity (T-S) diagrams of the different stations visited during the
MOMAP-2 expedition. Four regions (1, 2, 3a, and 3b) were determined based on the trend of
these data
The surface (down to 20 m) concentrations of inorganic nutrients were <0.04
µM for PO4, <0.35 µM for NO3, <1 µM for Si and <0.33 µM for NH4 in regions 1, 2,
and 3a. In region 3b, the concentration of PO4 and NO3 in the surface waters (down to 15
m) were somewhat higher (0.13 and 2.3 µM, respectively), while Si and NH4 levels were
comparable to that of the other stations. Below these depths, all nutrient levels increased
(max. 0.71 µM for PO4, 9.9 µM for NO3, and approx. 4.5 µM for Si and NH4). The
increase in PO4 and NO3 (0.20 and 0.49 µM, respectively) was, however, less
pronounced in region 2.
3.2. Phytoplankton distribution
All stations presented a deep chlorophyll a maximum (DCM) of 0.8 - 2.1 µg
Chl a L-1, located between 15 and 40 m (Fig. 4). The total phytoplankton cell abundance
exhibited a maximum typically located 5 to 10 m above the DCM (Fig. 4). The depth
integrated phytoplankton cell abundance over the upper 45 - 65 m ranged from 0.7 × 108
(region 2) to 3 × 108 cells cm-2 (regions 1 and 3a), of which 98 ± 2% were
picophytoplankton cells, independent of the region. The phytoplankton community in the
regions 1, 2, and 3a was numerically dominated by the cyanobacterium Synechococcus,
99
Chapter 5
Mortality of marine picophytoplankton
accounting for 82 ± 8% of the depth integrated phytoplankton abundance. The
concurring picoeukaryotic groups comprised 4 to 7% and the nanoeukaryotes
contributed for less than 2% of the depth integrated total abundance. The nutrient
enriched region 3b presented a different phytoplankton community structure with a
lower share of Synechococcus (23% of the depth integrated cell abundance). The Pico I,
II, and III accounted for 21, 35, and 12 % and the nanoeukaryotes comprised 10% of the
total community.
Chl a (µg L )
-1
0
1
4
-1
Total cells (x10 mL )
2
0
5
10
15
Region 1
Depth (m)
0
20
40
s9
s 10
60
Region 2
Depth (m)
0
20
40
s2
s 12
s 14
60
Region 3a
Depth (m)
0
20
s3
s4
s5
s8
s 13
40
60
Region 3b
Depth (m)
0
20
40
s6
60
Figure 4. Depth profiles of Chlorophyll a (Chl a, µg L-1) and total picophytoplankton abundance
(mL-1) for the different geographical regions distinguished.
100
Chapter 5
Mortality of marine picophytoplankton
The cell abundance of Synechococcus and Pico I distributed similarly for
regions 1, 2, and 3a (Fig. 5). Both groups developed a subsurface peak (15 – 35 m)
averaging 9 ± 3 × 104 cells mL-1 for Synechococcus and 10 ± 5 × 103 mL-1 for Pico I. By
comparison, Pico II and III abundance peaked deeper (35 - 40 m, 2 ± 1 × 103 mL-1 for
each group) in regions 2 and 3a. In region 1, the abundance of these latter groups was
high in the surface (10 and 2.5 × 103 cells mL-1 for Pico II and III, respectively) and
decreased with depth. In region 3b, all picophytoplankton groups showed similar
distributions with high abundance in the upper 15 m (~ 1 × 104 cells mL-1 for
Synechococcus, Pico I and II, and 3 × 103 cells mL-1for Pico III) and decreasing
gradually with depth.
The distribution of % dead algal cells differed among the picophytoplankton
groups and the area of occurrence (Fig 5). The % dead Synechococcus, Pico I, and Pico
II increased (up to 75%) below 15 m in the coastal region 1. The % dead Pico III
increased below 40 m to 60 to 80%, regardless of the region. Interestingly, we found a
negative and significant relation between the fraction of dead Synechococcus and their
cell abundance across the entire studied area (r = - 0.36, n = 40, p < 0.05, Spearman rank
order test).
In term of carbon biomass, the picophytoplankton groups comprised between
0.25 and 0.58 g C m-2, which corresponded to 51 ± 17 % of depth integrated
phytoplankton carbon biomass. Through the entire studied area, the carbon biomass of
the picoeukaryotes equaled (54 ± 16%) that of the cyanobacterium Synechococcus.
3.3. Virus community
The virus community was discriminated into 4 groups based on their DNA
green fluorescence and side scatter signature (Fig. 2). Total viral abundance was between
3 and 7 × 107 mL-1 in the top 15 - 30 m, below which it progressively declined (down to
1 – 3 × 107 mL-1). The abundance of the putative algal virus group (V4) in the surface
waters of region 3b was at least 4-fold higher (8 × 105 mL-1) than for the other regions
(Fig. 6). The viral group V3, including some algal viruses, showed increased surface
abundance at the northern stations of region 3a. Surface V3 abundance comprised up to
11% of the total virus abundance at station 5 while it was between 2 to 4% at the other
stations.
101
Chapter 5
Mortality of marine picophytoplankton
4
-1
3
Syn (x10 mL )
0
2
4
6
8 10
-1
3
Pico I (x10 mL )
0
5
10
15
-1
3
Pico II (x10 mL )
0
-1
Pico III (x10 mL )
5
10
15
0
25
50
75
0
1
2
3
4
5
Region 1
Depth (m)
0
20
40
-1
cell mL
% dead
60
Region 2
Depth (m)
0
20
40
60
Region 3a
Depth (m)
0
20
40
60
Region 3b
Depth (m)
0
20
40
60
0
25
50
75
% dead Syn
0
25
50
75
% dead Pico I
0
% dead Pico II
25
50
75
% dead Pico III
Figure 5. Depth profiles of the abundance of the 4 picophytoplankton groups (solid circle) and
their respective percentage of dead cells (%, open circle) for each regions of North Sea. Dead cells
were discriminated based on their membrane integrity using the nucleic acid stain SYTOX Green.
Region 1 is represented by station 10; region 2 by station 2; region 3a by stations 3, 4, 5 and 8
(averaged in figure); and region 3b by station 6. Errors bars correspond to SD.
102
Chapter 5
Mortality of marine picophytoplankton
6
V3 x 10 mL
0
0
1
2
-1
5
V4 x 10 mL
3
4 0
2
4
6
-1
8
10
Region 1
Depth (m)
10
20
30
40
50
s9
s 10
60
0
Region 2
Depth (m)
10
20
30
40
50
60
s2
s 12
s 14
0
Region 3a
Depth (m)
10
20
30
s3
s4
s5
s8
s 13
40
50
60
0
Region 3b
Depth (m)
10
20
30
40
50
s6
60
Figure 6. Depth profiles of viral group V4, assigned to putative algal viruses and group V3 most
likely containing putative algal viruses (virus mL-1) for the different geographical regions.
103
Chapter 5
Mortality of marine picophytoplankton
The virus diversity as determined by PFGE showed 5 to 20 distinct genome
sizes per sample ranging from 30 to 280 kb (Fig. 7). All samples displayed viral
genomes size ranging between 30 and 145 kb. Note that viral diversity in region 3b is
likely underestimated as only 20% of the plugged viruses migrated (reason unknown).
The viral community was dominated by small sized viral genomes (<100 kb) comprising
on average 98 ± 1 %. The largest viral genomes (105, 115, 145, 155, 165, and 180 kb)
accounted for 0.4 to 5.5% of the viral community, with the greatest variability in
richness and contribution found in region 3a. The viral diversity examined at the DCM
(stations 2, 3, and 14) showed an increased richness and abundance of viral genomes
ranging from 85 to 280 kb as compared to the surface sample.
< 1x107 mL
-1
1x108 - 4x108 mL
-1
Region 1
1x107 - 4x107 mL-1
5x107 - 9x107 mL-1
5x108 - 9x108 mL-1
>1x109 mL
3
2
-1
3b
Genome size (kb)
280
200
150
100
S 06 (5m)
S 13 (5m)
S 05 (5m)
S 03 (37m)
S 03 (5m)
S 14 (20m)
S 14 (5m)
S 12 (5m)
S 02 (20m)
S 02 (5m)
S 10 (5m)
S 09 (5m)
50
Figure 7. Schematic outline of the relative abundance (indicated by the area of the dot) of viral genome
sizes determined by PFGE fingerprinting. The PFGE was performed using two settings to optimally
determine the small sized genomes (30 – 100 kb; basically bacteriophages) and the large sized genomes
(50 - 400 kb, including the putative algal viruses).
104
Chapter 5
Mortality of marine picophytoplankton
3.4. Viral lysis and microzooplankton grazing
Viral lysis was substantial (0.16 - 0.23 d-1) and significant (p < 0.05) in the
surface the coastal Norwegian station 10 (region 1) and at the DCM of station 3, but only
for the picoeukaryotic groups Pico I and Pico III (Fig. 8 and Table 2). In the surface
waters of the offshore stations 2, 3, 5, and 6 (representatives of regions 2, 3a, and 3b),
viral lysis could be detected but estimates were not statistically significant (p > 0.05).
Microzooplankton grazing rates ranged widely (from insignificant to 0.72 d-1)
among the different picophytoplankton groups and geographical area (Fig 8, Table 2).
Synechococcus grazing was high at station 10 (0.72 d-1), whilst it was considerably lower
(0.05 ± 0.05 d-1) at the offshore stations 2, 3, 5, and 6. This considerable variation in
Synechococcus grazing rates was also reflected in the grazing rates obtained using the
independent FLC assay. The two grazing methodologies showed a strong correlation (r =
0.98, n = 5, p < 0.05, Spearman rank order test). Conversely to Synechococcus, the
picoeukaryotic groups were substantially grazed upon at all stations (average 0.13 ± 0.16
d-1 for Pico I, 0.18 ± 0.10 d-1 for Pico II, and 0.29± 0.16 d-1 for Pico III).
The algal growth rates in the surface waters varied considerably among the
picophytoplankton groups (from 0.1 – 2.1 d-1, Table 2) without a clear geographical
trend. The picocyanobacterium Synechococcus consistently presented the highest growth
rates (0.81 d-1 to 2.1 d-1) except for station 6 (region 3b, 0.17 d-1) and station 3 at the
DCM (-0.08 d-1). The growth rates of the picoeukaryotes varied on average between 0.25
and 0.41 d-1 (0.25 ± 0.15 d-1, 0.34 ± 0.16 d-1, and 0.41± 0.20 d-1 for Pico I, II, and III,
respectively).
3.5. Daily carbon production and losses
The total picophytoplankton carbon production (CP) varied greatly between
stations (0.5 to 15 µg C L-1 d-1, Table 2) with the highest value found at the coastal
station 10 (region 1, 15 µg C L-1 d-1). Overall, the picoeukaryotic groups substantially
contributed to the total picophytoplankton CP (on average 57 ± 35%).
The total picophytoplanktonic carbon losses due to viral lysis and
microzooplankton grazing (0.3 to 8.4 µg C L-1 d-1) balanced overall on average 49% of
the picophytoplanktonic CP with considerable variation among picophytoplankton
groups. Looking specifically at the picoeukaryotes, most of the CP was actually
counterbalanced by microzooplankton grazing and viral lysis (on average per group 60
to 120%). For Synechococcus only 10% of the CP was lost. Analysis per station shows
that only in the surface water of coastal station 10 (region 1) both viral lysis and
microzooplankton grazing were responsible for the total carbon loss (contributing for 29
and 71%, respectively). For the other stations microzooplankton alone accounted for the
total carbon loss in the surface waters (96 – 100%). At the DCM of station 3, however,
virally mediated carbon loss (0.2 µg C L-1 d-1) prevailed over grazing induced carbon
loss (0.1 µg C L-1 d-1), comprising 32% of the picophytoplankton CP (0.5 µg C L-1 d-1).
105
St 3 (37m)
St 5 (5m)
St 6 (5m)
µ apparent (d-1)
µ apparent (d-1)
µ apparent (d-1)
St 3 (5m)
µ apparent (d-1)
St 2 (5m)
µ apparent (d-1)
St 10 (5m)
Mortality of marine picophytoplankton
µ apparent (d-1)
Chapter 5
3.0
Synechococcus
picoeukaryote I picoeukaryotes II picoeukaryotes III
0.6
2.0
0.0
1.0
0.0
0.2 µm
30 kDa
-0.6
1.2
0.6
0.6
0.0
0.0
-0.6
1.2
0.6
0.6
0.0
0.0
-0.6
1.2
0.6
0.6
0.0
0.0
-0.6
1.2
0.6
0.6
0.0
0.0
-0.6
1.2
0.6
0.6
0.0
0.0
0.0
0.4
0.8
-0.6
0.0
0.4
0.8
0.0
0.4
0.8
0.0
0.4
0.8
Fraction of natural sample
Figure 8. Plots of apparent growth rate vs. fraction of natural water in the parallel dilution
experiment for the different picophytoplankton groups at the main stations. Region 1 is
represented by station 10; region 2 by station 2; region 3a by stations 3 and 5, and region 3b by
station 6. Parallel dilution experiments were performed in 30 kDa ultrafiltrate (no grazer, no
virus) and 0.2 µm (no grazer) filtered seawater. Microzooplankton grazing rates correspond to the
regression slope obtained with the 0.2 µm dilution series. Viral lysis rates correspond to the
difference of regression coefficients of the 0.2 µm and 30 kDa series. For the readability of the
figures, we averaged the triplicate apparent growth for each dilution level. This operation did not
affect the estimated mortality rates. The error bars reflect the SD between measurements.
106
107
1
Synechococcus
Picoeukaryote 1
Picoeukaryote 2
Picoeukaryote 3
Synechococcus
Picoeukaryote 1
Picoeukaryote 2
Picoeukaryote 3
Synechococcus
Picoeukaryote 1
Picoeukaryote 2
Picoeukaryote 3
Synechococcus
Picoeukaryote 1
Picoeukaryote 2
Picoeukaryote 3
Synechococcus
Picoeukaryote 1
Picoeukaryote 2
Picoeukaryote 3
Synechococcus
Picoeukaryote 1
Picoeukaryote 2
Picoeukaryote 3
To be submitted to Aquatic Microbial Ecology
St. 10 (5m)
St. 6 (5m)
St. 5 (5m)
St. 3 (37m)
St. 3 (5m)
St. 2 (5m)
Abundance
12
2.7
0.4
1.0
24
5.6
2.1
1.5
18
9.2
1.2
0.7
8.0
1.0
1.4
1.1
8.5
8.0
14
3.9
12
6.6
10
2.5
107
µ
1.20*
0.05
0.55*
0.71*
0.86*
0.12*
0.24*
0.06*
-0.08*
0.17*
0.18*
0.45*
0.81*
0.41*
0.51*
0.49*
0.17*
0.37*
0.23*
0.34*
2.10*
0.35*
0.31*
0.34*
G
0.12*
0.07
0.10
0.28*
0.01
0.15*
0.14*
0.43*
0.03
0.03
0.10
0.06
0.01
0.44*
0.31*
0.40*
0.07
0.11*
0.15*
0.41*
0.72*
0.01
0.29*
0.23*
V
0.00
-0.01
0.06
-0.08
0.01
0.03
-0.02
-0.05
0.01
0.05
0.04
0.20*
0.00
0.01
0.01
-0.04
-0.07
0.00
-0.04
0.00
-0.02
0.16*
0.02
0.13*
dead
23
6
27
31
24
9
10
15
38
18
25
20
25
3
21
5
38
9
15
33
20
13
23
32
CP
5.5
0.0
0.2
2.9
6.9
0.1
0.2
0.0
0.0
0.4
0.1
0.8
2.1
0.1
0.5
2.0
0.3
0.7
1.2
2.9
12.0
0.6
1.5
3.0
G : CP
0.10*
0.19
0.40*
0.01
1.25*
0.58*
2.87*
0.16
0.54
0.13
0.01
1.08*
0.63*
0.81*
0.34
0.28*
0.64*
1.20*
0.34*
0.00
0.93*
0.38*
V : CP
0.00
0.11
0.00
0.01
0.25
0.00
0.00
0.29
0.33
0.44*
0.00
0.02
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.43*
0.06
0.67*
Table 2. Dilution-based specific growth (µ, d-1), microzooplankton grazing (G, d-1), and viral lysis rates (V, d-1) for the picophytoplankton
groups and their abundance (× 103 mL-1) and fraction of dead cells (dead, %) at the onset of the experiment. Daily picophytoplanktonic
carbon production (CP, µg C l-1 d-1) and the fraction of the CP consumed by microzooplankton (G : CP) and viruses (V : CP) was calculated
for each experiment. Region 1 is represented by station 10; region 2 by station 2; region 3a by stations 3 and 5, and region 3b by station 6.
Asterix correspond to significant values (p ≤ 0.05).
Chapter 5
Mortality of marine picophytoplankton
4. Discussion
4.1. Grazing and viral lysis
The phytoplankton community in the North Sea during summer distributed into
6 size-classes (1 to 20 µm), amongst which the picophytoplankton component (<3µm)
dominated the community. The extent of the losses due to viral lysis and
microzooplankton varied widely among the picophytoplankton groups and their
geographical location. Microzooplankton were the main mortality agents (on average
0.19 ± 0.18 d-1), but their impact differed among the picophytoplankton groups.
Synechococcus was under the lowest grazing pressure (0.04 ± 0.05 d-1) except at station
10 (0.72 d-1). The strong correlation between the two independent grazing assays used
during this study (dilution technique and FLC method; r = 0.98, n = 5, p < 0.05) implies
that the low rates did not arise from experimental artefacts but were due to preferential
grazing. An earlier study executed in the same site indicated that 1-2 µm sized
phytoplankton were prone to lower grazing rates (0.07 d-1) than the 2-3 µm size class
(0.20 d-1, Kuipers & Witte 1999). Another study conducted in oligotrophic surface
waters has also consistently reported higher predation rates on picoeukaryotes (up to 3
fold) than Synechococcus (Worden et al. 2004). The differential grazing observed in
offshore vs. coastal station could result from distinct predator communities at these
stations that preferentially consumed certain picophytoplankton populations. For
instance, Synechococcus was found to be selectively egested or digested by some
heterotrophic nanoflagellates (Boenigk et al. 2001, Guillou et al. 2001) and ciliates
(Christaki et al. 1999).
Significant viral lysis rates (up to 0.23 d-1) were recorded along the Norwegian
coast (region 1) for picoeukaryotic groups Pico I and Pico III, and at the DCM of the
offshore station 3 for the picoeukaryote group Pico III (region 3). These results indicate
that about 20% of the abundance of each group underwent viral lysis on a daily basis.
Pigment analysis showed that prasinophytes distributed widely in the studied area,
comprising 20 and 30% of the total phytoplankton Chl a as compared to 0.1 - 5%
elsewhere (Brussaard C.P.D, unpubl. data). Interestingly, viruses infecting the
prasinophyte Micromonas pusilla were isolated from all regions, but the highest titer (1
× 103 mL-1) was found in region 1 and 3a (Brussaard C.P.D., unpubl. data). Flow
cytometry examination revealed a proliferation of viruses belonging to the group V3 in
all lysates. These observations suggest that viruses may be responsible for the mortality
in the M. pusilla population, which is consistent with an earlier mesocosm experiment
carried out in Norwegian coastal waters during late spring (M. pusilla turnover rates of 9
- 25% d-1; Evans et al. 2003). Also, viruses caused significant M. pusilla cell lysis along
the coast of the oligotrophic Gulf of Mexico (turnover rates 2 - 10% d-1; Cottrell & Suttle
1995).
108
Chapter 5
Mortality of marine picophytoplankton
The virally mediated turnover rates of about 20% of host abundance d-1 reported
for the specific picoeukaryotic groups at stations 10 and 3 matched nicely with the
fraction of dead algal cells in the respective groups (13% for the Pico I and 32% and
20% for Pico III at stations 10 and 3, respectively). Although this suggests that viral lysis
may be the cause of cell membrane permeabilization of picoeukaryotes, the finding of
relatively high % dead algal cells in combination with low viral lysis rates indicates that
there is no general relationship between viral lysis and % dead cells per se. A buildup in
standing stock of dead algal cells using the SYTOX Green live/dead assay cannot be
ruled out but other factors causing cell lysis may also be involved (Franklin et al., 2006).
Among these, our results indicated that water temperature might have altered
Synechococcus viability as reported by Alonso-Laita & Agusti (2006). Increasing
temperature significantly correlated to Synechococcus viability (r = 0.8, n = 6, p = 0.05,
Spearman rank order test) and growth rate (r = 1, n = 6, p < 0.01, Spearman rank order
test). The relatively low temperature at the DCM of station 3 (9.9°C) may have enhanced
the % of dead Synechococcus, however the negative effect of light limitation on
Synechococcus growth (Moore et al. 1995) and membrane integrity (Alonso-Laita &
Agusti 2006) cannot be excluded.
Studies on viral lysis conducted in oligotrophic environments are, thus far,
limited and mainly focused on Synechococcus. These studies consistently reported that
Synechococcus experience relatively low viral lysis (Waterbury & Valois 1993, Suttle &
Chan 1994, Garza & Suttle 1998). Our results support the relatively small impact of
viruses on Synechococcus mortality in surface oligotrophic waters. In contrast to an
earlier observation suggesting that dilution in virus-free water reduced the
Synechococcus growth rates (Suttle 1996), we recorded similar growth rates for
Synechococcus in both the 0.2 µm and the 30 kDa dilution series. These results,
combined with a previous study carried out in the oligotrophic northeastern subtropical
Atlantic Ocean (Baudoux et al. in press) show that that the modified dilution method can
be successfully used in oligotrophic environments.
The modified dilution method is, to date, the only method that provides viral
lysis rates of phytoplankton mortality directly. It is, however, unclear whether this
approach is sensitive enough to detect low viral lysis rates. In this study, viral lysis rates
down to 0.01 - 0.06 d-1 were recorded but they were not statistically significant. The
lowest significant (p < 0.05) virally induced mortality rate determined using this method
is 0.1 d-1 (Evans et al. 2003, Baudoux et al. submitted, this study). Acknowledging this
restriction, our results indicate differential viral control among picophytoplanktonic algal
groups and their area of occurrence.
Different reasons can account for the lack of viral lysis even in presence of the
relatively high standing stock of viral groups V4, assigned to putative algal viruses (8 ×
105 mL-1, station 6), and V3, likely containing algal virus (4 × 106 mL-1, station 5). One
reason for this observation could be that the viruses present in the sample were not
specific to the co-occurring Synechococcus or picoeukaryote populations. The generally
high genetic diversity of these hosts (Moon-van der Staay et al. 2001, Scanlan & West
2002), as well as the ability of Synechococcus to resist co-occurring viruses (Waterbury
109
Chapter 5
Mortality of marine picophytoplankton
& Valois 1993) could explain the absence of viral lysis. Another reason could be that the
algal virus community included viruses infecting nanoeukaryotes rather than
picoeukaryotes. Relatively high abundance of the prymnesiophytes Chrysochromulina
ericina and Emiliania huxleyi have been reported in waters adjacent to station 6
(Riegman & Kraay 2001 and references therein). Both of these nanoeukaryotes can be
subjected to viral lysis during summer blooms in the North Sea (Brussaard et al. 1996b).
Interestingly, pigment analysis revealed that prymnesiophytes comprised most of the Chl
a at station 6 (Brussaard C.P.D, unpubl. data). The FCM signature of the V4 viral group,
furthermore, resembled that of Chrysochromulina ericina virus (Brussaard 2004b) or
Emiliania huxleyi virus (Jacquet et al. 2002). Finally, the absence of viral lysis could be
caused by a reduced viral infectivity. In the surface layer of the ocean environmental
factors such as high solar radiation, and particularly UV radiation (UVR), can severely
alter the infectivity of algal viruses (Cottrell & Suttle 1995, Garza & Suttle 1998, Jacquet
& Bratbak 2003). It is noteworthy that viral lysis was exclusively detected in light
attenuated environments during the present study. The Norwegian coastal station had the
shallowest euphotic zone (23 m, irradiance approximately half of that at other stations),
whereas light intensity at the DCM at station 3 was very low (3 µmol quanta m-2 s-1).
Viral lysis rates of picoeukaryotes at the DCM were not only enhanced as compared to
the surface, but the richness and abundance of putative algal viruses (>100 kb) was also
higher. Such shift in viral community structure was confirmed for the other stations
where viral diversity was also studied at the DCM (stations 2 and 14). More detailed
research is required to test whether differences in algal virus diversity between surface
waters and the DCM are a more general feature. Also, it may be of ecological relevance
to investigate whether picophytoplankton viral lysis at the DCM is generally higher as
compared to the surface waters. Overall, our observations lead to speculate that the
ambient light level may underlie differential impact of viruses on picophytoplankton
mortality.
4.2. Implications for the carbon cycle
The distinction and quantification of the phytoplankton losses due to lysis and
microzooplankton is essential for an optimal understanding of the carbon pathway in
marine environments. Our results showed that viral lysis locally yielded substantial
carbon release, amounting to 0.2 and 2.4 µg C L-1 d-1 at the DCM of station 3 and in the
surface layer of station 10, respectively. The value obtained at the DCM compared well
with an earlier study executed in the DCM waters of the subtropical northeastern
Altantic Ocean (0.1 – 0.3 µg C L-1 d-1, Baudoux et al. in press). In contrast, the amount
of virally induced carbon release in the Norwegian coastal waters (2.4 µg C L-1 d-1)
largely exceeds those reported in the surface coastal waters of the Gulf of Mexico (0.12
– 0.35 µg C L-1 d-1, Wilhelm & Suttle 1999 and references therein). These lower values
may be caused by the limited number of potential host taken into account by these
authors (only M. pusilla and Synechococcus); therefore the impact of algal viruses on
carbon flow may be higher than previously assumed.
110
Chapter 5
Mortality of marine picophytoplankton
In addition to virally induced carbon release, the leakage of soluble compounds
from the cells with compromised membrane may constitute another relevant source of
dissolved organic carbon. However, the classification of cells as dead cannot be directly
converted to cell lysis rates, as the elapsed time between the physiological death and the
subsequent lysis is unknown. Therefore, the magnitude and quantification of this latter
process needs further investigation.
The carbon losses generated by microzooplankton grazing were generally
higher (range 0.1 – 6.0 µg C L-1 d-1) than those caused by viral lysis (range 0.2 - 2.4 µg
C L-1 d-1), except at the DCM of station 3. The picoeukaryotes constituted the primary
source of carbon for the microzooplankton, with on average 77% (range 55 - 140%) of
the picoeukaryotic carbon production lost by microzooplankton grazing (as compared to
10% for Synechococcus). Our results substantiate earlier studies suggesting the potential
of picoeukaryotes for the carbon transfer to the higher level of the pelagic marine food
web (Worden et al. 2004). Overall, microzooplankton consumed 40 ± 27 % d-1 of the
total picophytoplankton carbon production (CP) which is slightly lower than the general
estimate of CP consumption by microzooplankton of 59% d-1 as reported for the
temperate ecosystems (Calbet & Landry 2004). The analysis by Calbet & Landry (2004)
includes, however, all seasons and different geographical study sites.
In summary, the present study shows an important spatial variability and algal
group specificity of grazing and virally induced mortality rates. Such variability will
affect the structure of the plankton community and the carbon cycling differently and
should, therefore, be addressed in more detail in future analysis of phytoplankton
mortality and carbon cycling.
Acknowledgments. We thank the captain and crew of the R.V. Pelagia for excellent
shipboard support. We thank the nutrient service lab for technical support. Caroline
Chenard, Veronica Parada, and Dedmer Van de Waal are acknowledged for their
assistance during the cruise. This work was supported by the Research Council for Earth
and Life Sciences (ALW) with financial aid from the Netherlands Organization for
Scientific Research (NWO).
Literature cited
Agusti S, Satta MP, Mura MP, Benavent E (1998) Dissolved esterase activity as a tracer
of phytoplankton lysis: Evidence of high phytoplankton lysis rates in the
northwestern Mediterranean. Limnol Oceanogr 43:1836-1849
Alonso-Laita P, Agusti S (2006) Contrasting patterns of phytoplankton viabilily in the
subtropical NE Atlantic Ocean. Aquat Microb Ecol 43:67-78
Baudoux AC, Noordeloos AAM, Veldhuis MJW, Brussaard CPD (2006) Virally induced
111
Chapter 5
Mortality of marine picophytoplankton
mortality of Phaeocystis globosa during two spring blooms in temperate coastal
waters. Aquat Microb Ecol 44:207-217
Berges JA, Falkowski PG (1998) Physiological stress and cell death in marine
phytoplankton: induction of proteases in response to nitrogen or light limitation.
Limnol Oceanogr 43:129-135
Berman-Frank I, Bidle KD, Haramaty L, Falkowski PG (2004) The demise of the marine
cyanobacterium, Trichodesmium spp., via an autocatalyzed cell death pathway.
Limnol Oceanogr 49:997-1005
Bertlisson S, Berglund O, Karl DM, Chisholm SW (2003) Elemental composition of
marine Prochlorococcus and Synechococcus: Implications for the ecological
stoichiometry of the sea. Limnol Oceanogr 48:1721-1731
Boenigk J, Matz C, Jurgens K, Arndt H (2001) The influence of preculture conditions
and food quality on the ingestions and digestion process of three species of
heterotrophic nanoflagellates. Microb Ecol 42:168-176
Brussaard CPD (2004a) Viral control of phytoplankton populations - a review. J Euk
Microbiol 51:125-138
Brussaard CPD (2004b) Optimization of procedures for counting viruses by flow
cytometry. Appl Environ Microbiol 70:1506-1513
Brussaard CPD, Gast GJ, vanDuyl FC, Riegman R (1996a) Impact of phytoplankton
bloom magnitude on a pelagic microbial food web. Mar Ecol Prog Ser 144:211221
Brussaard CPD, Kempers RS, Kop AJ, Riegman R, Heldal M (1996b) Virus-like
particles in a summer bloom of Emiliania huxleyi in the North Sea. Aquat Microb
Ecol 10:105-113
Brussaard CPD, Kuipers B, Veldhuis MJW (2005a) A mesocosm study of Phaeocystis
globosa population dynamics - 1. Regulatory role of viruses in bloom. Harmful
Algae 4:859-874
Brussaard CPD, Marie D, Thyrhaug R, Bratbak G (2001) Flow cytometric analysis of
phytoplankton viability following viral infection. Aquat Microb Ecol 26:157-166
Brussaard CPD, Noordeloos AAM, Riegman R (1997) Autolysis kinetics of the marine
diatom Ditylum brightwellii (Bacillariophyceae) under nitrogen and phosphorus
limitation and starvation. J Phycol 33:980-987
Brussaard CPD, Riegman R, Noordeloos AAM, Cadée GC, Witte H, Kop AJ,
Nieuwland G, Vanduyl FC, Bak RPM (1995) Effects of grazing, sedimentation
and phytoplankton cell lysis on the structure of a coastal pelagic food web. Mar
Ecol Prog Ser 123:259-271
Christaki U, Jacquet S, Dolan JR, Vaulot D, Rassoulzadegan F (1999) Growth and
grazing on Prochlorococcus and Synechococcus by two marine ciliates. Limnol
Oceanogr 44:52-61
Calbet A, Landry MR (2004) Phytoplankton growth, microzooplankton grazing, and
carbon cycling in marine systems. Limnol Oceanogr 49:51-57.
Cottrell MT, Suttle CA (1995) Dynamics of a lytic virus infecting the photosynthetic
marine picoflagellate Micromonas pusilla. Limnol Oceanogr 40:730-739
DuRand MD, Green RE, Sosik HM, Olson RJ (2002) Diel variations in optical
112
Chapter 5
Mortality of marine picophytoplankton
properties of Micromonas pusilla (Prasinophyceae). J Phycol 38:1132-1142
Evans C, Archer SD, Jacquet S, Wilson WH (2003) Direct estimates of the contribution
of viral lysis and microzooplankton grazing to the decline of a Micromonas spp.
population. Aquat Microb Ecol 30:207-219
Franklin DJ, Brussaard CPD, Berges JA (2006) What is the role and nature of
programmed cell death in phytoplankton ecology? Europ J Phycol 41:1-14
Garza DR, Suttle CA (1998) The effect of cyanophages on the mortality of
Synechococcus spp. and selection for UV resistant viral communities. Microb
Ecol 36:281-292
Geider RJ, Delucia EH, Falkowski PG, Finzi AC, Grime JP, Grace J, Kana TM, La
Roche J, Long SP, Osborne BA, Platt T, Prentice IC, Raven JA, Schlesinger WH,
Smetacek V, Stuart V, Sathyendranath S, Thomas RB, Vogelmann TC, Williams
P, Woodward FI (2001) Primary productivity of planet Earth: biological
determinants and physical constraints in terrestrial and aquatic habitats. Global
Change Biology 7:849–882
Guillou L, Jacquet S, Chretiennotdinet MJ, Vaulot D (2001) Grazing impact of two
small heterotrophic flagellates on Prochlorococcus and Synechococcus. Aquat
Microb Ecol 26:201-207
Jacquet S, Bratbak G (2003) Effects of ultraviolet radiation on marine virusphytoplankton interactions. FEMS Microbiol Ecol 44:279-289
Jacquet S, Heldal M, Iglesias-Rodriguez D, Larsen A, Wilson W, Bratbak G (2002)
Flow cytometric analysis of an Emiliania huxleyi bloom terminated by viral
infection. Aquat Microb Ecol 27:111-124
Kuipers BR, Witte HJ (1999) Grazing impact of microzooplankton on different size
classes of algae in the North Sea in early spring and mid-summer. Mar Ecol Prog
Ser180:93-104
Landry MR, Hassett RP (1982) Estimating the grazing impact of marine
microzooplankton. Mar Biol 67:283-288
Landry MR, Constantinou J, Latasa M, Brown SL, Bidigare RR, Ondrusek ME (2000)
Biological response to iron fertilization in the eastern equatorial Pacific (IronEx
II). Dynamics of phytoplankton growth and microzooplankton grazing. Mar Ecol
Prog Ser 201:57-72.
Larsen A, Castberg T, Sandaa RA, Brussaard CPD, Egge J, Heldal M, Paulino A,
Thyrhaug R, van Hannen EJ, Bratbak G (2001) Population dynamics and
diversity of phytoplankton, bacteria and viruses in a seawater enclosure. Mar Ecol
Prog Ser 221:47-57
Mann NH (2003) Phages of the marine cyanobacterial picophytoplankton. FEMS
Microbiol Rev 27:17-34
Moon-van der Staay SY, De Wachter R, Vaulot D (2001) Oceanic 18S rDNA sequences
from picoplankton reveal unsuspected eukaryotic diversity. Nature 409:607-610
Moore LR, Goericke R, Chisholm SW (1995) Comparative physiology of
Synechococcus and Prochlorococcus: Influence of light and temperature on
growth, pigments, fluorescence and absorptive properties. Mar Ecol Prog Ser 116:
259-275
113
Chapter 5
Mortality of marine picophytoplankton
Raven
JA (1998) The twelfth Tansley Lecture. Small is beautiful: the
picophytoplankton. Funct Ecol 12:503-513
Riegman R, Kraay GW (2001) 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. J Plankton Res 23:191-205
Riegman R, Kuipers BR, Noordeloos AAM, Witte HJ (1993) Size-Differential Control
of Phytoplankton and the Structure of Plankton Communities. Neth J Sea Res
31:255-265
Scanlan DJ, West NJ (2002) Molecular ecology of the marine cyanobacterial genera
Prochlorococcus and Synechococcus. FEMS Microbiol Ecol 40:1-12
Sherr EB, Sherr BF (1993) Protistian grazing rates via uptake of fluorescently labeled
prey. In: Kemp PF, Sherr BF, Sherr EB, Cole JJ (eds) Microbial ecology. Lewis
Publishers, Boca Raton, p 695–701
Smetack JS (1985) Role of sinking diatom life-history cycles: ecological, evolutionary
and geological significance. Mar Biol 84:239-251
Suttle CA, Chan AM (1994) Dynamics and distribution of cyanophages and their effect
on marine Synechococcus spp. Appl Environ Microbiol 60:3167-3174
Suttle CA, Chan AM, Cottrell MT (1990) Infection of phytoplankton by viruses and
reduction of primary productivity. Nature 347:467-469
Suttle CA, Chan AM, Rodda KM, Short SM, Weinbauer MG, Garza DR, Wilhelm SW
(1996) The effect of cyanophages on Synechococcus spp. during a bloom in the
western Gulf of Mexico. EOS 76 (Suppl):OS207-OS208
Veldhuis MJW, Kraay GW (2004) Phytoplankton in the subtropical Atlantic Ocean:
towards a better assessment of biomass and composition. Deep-Sea Res I 51:507530
Veldhuis MJW, Kraay GW, Timmermans KR (2001) Cell death in phytoplankton:
correlation between changes in membrane permeability, photosynthetic activity,
pigmentation and growth. Europ J Phycol 36:167-177
Verity PG, Robertson CY, Tronzo CR, Andrews MG, Nelson JR, Sieracki ME (1992)
Relationships between cell volume and the carbon and nitrogen content of marine
photosynthetic nanoplankton. Limnol Oceanogr 37:1434-1446
Waterbury JB, Valois FW (1993) Resistance to co-occurring phages enables marine
Synechococcus communities to coexist with cyanophages abundant in seawater.
Appl Environ Microbiol 59:3393-3399
Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea. BioScience
49:781-788
Worden AZ, Nolan JK, Palenik B (2004) Assessing the dynamics and ecology of marine
picophytoplankton: The importance of the eukaryotic component. Limnol
Oceanogr 49:168-179
114
Chapter 6
Influence of irradiance on virus-algal
host interactions 1
Anne-Claire Baudoux and Corina P. D. Brussaard
The effect of different irradiance levels on the interactions between the algal
host and its virus were investigated for two marine phytoplankton, Phaeocystis globosa
Largerheim and Micromonas pusilla Manton and Parke. The algal cultures were
acclimated at 25, 100, and 250 µmol photons m-2 s-1 (LL, ML, and HL, respectively),
after which they were infected with a lytic virus (PgV and MpV) and monitored under
the appropriate irradiance and in darkness. The effect of irradiance levels on the hostvirus interactions was species-specific. For P. globosa, the LL-adapted cultures showed
a 4 h prolonged latent period (11-16 h), which may be related to the subsaturated
growth observed at this irradiance. The burst size was 50% reduced at LL and HL as
compared to ML (525 PgV cell-1). The fraction of infectious viruses was, however,
unchanged. Viral replication was prevented when the LL P. globosa cultures were kept
in darkness (up to 48h), but recovered when placed back into the light. PgV could still
replicate in the dark for the ML- and HL-adapted cultures, but viral yield was reduced
by 50 to 85%. For M. pusilla, the burst size (285 to 360 MpV cell-1), the infectivity, and
the latent period of MpV (7-11 h) were unaffected by the host’s photoacclimation.
Conversely, darkness not only inhibited MpV replication but also resulted in substantial
cell lysis of the noninfected cultures. Our study implies that irradiance level is an
important factor controlling algal host-virus interactions, hence the phytoplankton
population dynamics in the field.
1
Submitted to Journal of Phycology
115
Chapter 6
Light-affected viral lysis
1. Introduction
Marine algal viruses have been shown to influence phytoplankton dynamics, the
functioning of pelagic food webs and biogeochemical cycling (for review Wilhelm &
Suttle 1999, Brussaard 2004a). There is, however, still little documentation on
environmentally relevant factors that influence the interactions between the virus and its
host. Viral replication is governed by the host cellular machinery; therefore the factors
that affect the algal host’s physiology (e.g., nutrients, light) may influence the
interactions between algal hosts and viruses.
In marine environments phytoplankton are constantly subjected to fluctuations
in light intensities. These changes in irradiance can be generated, for instance, by cloud
coverage or water mixing on a short time scale, but they can also be associated with
stratification or seasonal variation in solar radiation on a longer time scale. In response
to these light fluctuations, phytoplankton have evolved a large variety of physiological
responses (photoacclimation) to optimize their growth (Falkowski & La Roche 1991).
To date, the few existing studies investigating the effect of light on virus-host
interactions were restricted to darkness and showed variable outcomes (Waters & Chan
1982, Van Etten et al. 1983, Bratbak et al. 1998, Suttle 2000). During the present study,
the effect of different irradiance levels, including darkness, on the interactions between
algae and their specific viruses were investigated for two ecologically important marine
phytoplankton, Phaeocystis globosa Largerheim and Micromonas pusilla Manton and
Parke.
Both P. globosa and M. pusilla are important for the structure of phytoplankton
community and the functioning of the system where they occur (Brussaard et al. 1995,
Not et al. 2004). These species can be subjected to virally induced mortality rates
comparable to grazing losses (Evans et al. 2003, Baudoux et al. 2006). Phaeocystis
globosa is known to generate dense and nearly monospecific spring blooms in temperate
coastal waters when sufficient light and nutrients are available (Cadée & Hegeman
2002). Phaeocystis globosa is characterized by a polymorphic life cycle composed of
unicellular flagellated cells (5 to 7 µm in diameter) and non-motile cells embedded in
colonies (up to 1 cm). In contrast, M. pusilla does not form high- biomass blooms and
only exists as small flagellated single cells (1 to 3 µm in diameter). Micromonas pusilla
has a worldwide distribution and is described as a major component of the
picophytoplankton community in many different coastal as well as oceanic waters
(Kuylenstierna & Karlson 1994, Throndsen & Zingone 1994, Not et al. 2004).
The different irradiance levels tested (0, 25, 100, and 250 µmol photons·m-2·s-1)
showed that light can strongly impact the interactions between the algal hosts and their
specific virus. The effects of light intensity on viral infection were, however, speciesspecific. The viral growth cycle of the virus infecting P. globosa (PgV) revealed that
both low and high irradiance reduced the burst size (number of produced viruses per host
cell). The latent period (time until virus progeny is released from the host cell) was,
furthermore, prolonged at low light. Darkness resulted in reduced burst sizes of PgV,
116
Chapter 6
Light-affected viral lysis
partially restored to the original level upon return to the light. The lytic growth cycle of
MpV was not affected by irradiance, but darkness stopped the viral production. These
results imply that light can affect the ecological role of (algal) viruses very differently.
2. Material and methods
Phytoplankton and virus cultures. Unialgal cultures of the prymnesiophyte
Phaeocystis globosa (Hariot) Largerheim strain G (Groningen University culture
collection, The Netherlands) and the prasinophyte Micromonas pusilla (Butscher)
Manton and Parke strain 1545 (Bigelow culture collection of marine phytoplankton,
USA) were used during this study. Both algal cultures were grown in a 1:1 mixture of f/2
medium (Guillard 1975) and ESAW (Cottrell & Suttle 1991). The algae were cultured at
15°C under a light:dark cycle of 16:8 h at light intensities of 25, 100, and, 250 µmol
photons m-2 s-1 (hereafter abbreviated LL, ML and HL for low, medium and high light
intensities respectively). The algal cultures were adapted at the appropriate irradiance by
repeated dilution (at least 5 volume changes) in fresh medium. Dilutions were performed
to keep cultures in early exponential phase at cell abundance between 1×104 and 1×106
cells mL-1. The cultures’ adaptation lasted until stabilization of their maximal growth
rate (µmax, Table 1).
The lytic viruses infecting P. globosa (PgV-07T, Baudoux & Brussaard 2005)
and M. pusilla (MpV-02T) were both isolated from surface water of the North Sea
according to Brussaard et al. (2004). In short, filtered (Whatman GF/F) natural seawater
was added to respective algal host culture (10 – 20 % v/v) and incubated for 10 days at
15°C under a light dark cycle 16:8 h receiving 100 µmol photons m-2 s-1 (standard
culture conditions of the hosts). Algal growth was monitored via in vivo chlorophyll
fluorescence (Fo) using a Turner Designs fluorometer (model 10-AU). The cultures that
showed signs of lysis as compared to noninfected controls were filtered through 0.2 µm
pore-size cellulose acetate filters (Schleicher and Schuell GmbH, Dassel, Germany),
after which the lysate was used to reinfect an exponentially growing algal host culture.
Cultures were checked for the presence of algal viruses using FCM and transmission
electron microscopy (TEM). After recurrent lysis and reinfection, viral isolates were
made clonal by repeated end-point dilution.
Experimental set up. The three different light-adapted P. globosa and M.
pusilla host cultures at a cell abundance of 1×105 cells mL-1 were split into four equal
subcultures of 220 mL. Two of these subcultures were inoculated with the corresponding
viral lysate (PgV-07T or MpV-02T) at an initial virus to host ratio of around 20. Most
probable number (MPN) examination (see below) of the viral lysate showed that the
multiplicity of infection (MOI) ranged between 10 and 20, which is sufficiently high to
allow a one step infection cycle. The other subculture was a noninfected control and
received an equal volume of medium. Each subculture was incubated at the respective
117
Chapter 6
Light-affected viral lysis
light regime (LL, ML, or HL) and sampled for algal and viral abundance, and
photochemical efficiency (Fv/Fm) every 4 h for a total period of 70 h.
A parallel experiment was simultaneously conducted in the dark using 50 mL
subcultures in triplicate. Earlier test showed that the use of a reduced volume (50 mL) as
compared to the experiment presented in the previous paragraph (220 mL) did not affect
the results. Darkness was achieved by wrapping the incubation flask completely into
three layers of aluminium foil. All samples were incubated at the standard culture
conditions. At T=0, 1, and 2 days, one of the dark triplicates was sampled for algal and
viral counts and photochemical efficiency (Fv/Fm). A 5 mL aliquot of this sample was
transferred into a borosilicate glass tube and exposed to the appropriate irradiance for 2
days (mixed by hand each day). After 2 days, subsamples of these tubes were taken for
viral abundance and the natural chlorophyll autofluorescence Fo, as a measure of algal
biomass.
Analyses. Algal abundance was monitored by flow cytometry using a Beckman
Coulter EPICS XL-MCL flow cytometer equipped with a laser with an excitation
wavelength of 488 nm (15mW) and emission bands for the chlorophyll autofluorescence
(> 630 nm), phycoerythrin fluorescence (575 ± 20 nm), and green fluorescence (515 ±
20 nm). For the algal abundance, fresh samples were diluted up to 10-fold in 0.2 µm
filtered sterile culture medium. Fluorescent microspheres of 0.95 µm (InvitrogenMolecular Probes, Eugene, OR, USA) were added as an internal standard. The trigger
was set on the red chlorophyll autofluorescence (RFL) and samples were analyzed for 1
min at a flow rate of 72 µL min-1. Scatter and fluorescent signals were normalized to the
signal of the internal standard beads.
For virus enumeration using flow cytometry (Brussaard 2004b), 1 mL samples
were fixed with 25 % glutaraldehyde (0.5 % final concentration, EM grade, SigmaAldrich, St Louis, MO, USA) during 30 minutes at 4°C, flash frozen in liquid nitrogen
and stored at -80°C until analysis. The thawed samples were diluted 100 to 1,000-fold in
autoclaved 0.2 µm filtered TE-buffer (pH 8.0) and stained with the nucleic acid-specific
dye SYBR Green I (Invitrogen-Molecular Probes, Eugene, OR, USA) for 10 min at
80°C. The trigger was set on the green fluorescence and the sample was delivered at a
rate of 20 µL min-1 and analyzed for 1 min. Virus counts were corrected for the blank
consisting of TE-buffer with autoclaved 0.2 µm filtered seawater in the correct dilution.
The abundance of infectious PgV and MpV was estimated using the end-point
dilution approach (Most Probable Number, MPN, Suttle 1993). Briefly, the freshly
produced lysate was 10-fold serial diluted (5 replicates, 12 dilution levels) with an
exponentially growing algal culture at the appropriate irradiance (LL, ML, or HL). The
dilution series were incubated for 10 days at 15°C under the respective culture regime.
Algal growth was monitored fluorometrically (Fo). Sample that underwent lysis were
scored positive and the resulting MPN of infectious viruses was calculated using a MPN
assay computer program (Hurley & Roscoe 1983).
The photochemical efficiency (Fv/Fm) of a 5 mL algal sample was measured
using a Turner Designs model 10-AU fluorometer. After dark-acclimation of the algal
cells for 5 min (Geider et al. 1993), the natural fluorescence (Fo) of the sample was
118
Chapter 6
Light-affected viral lysis
determined, afterwhich the maximal fluorescence (Fm) was induced by adding DCMU
(3’,4’-dichlorophenyl-1,1-dimethylurea, Sigma-Aldrich, St Louis, MO, USA) to a final
concentration of 10 µM. The variable fluorescence (Fv) was calculated as Fm minus Fo.
Maximal photochemical efficiency of the PSII reaction center of the algal host was
derived from the ratio of Fv over Fm. We occasionally (3 times) omitted unrealistically
high Fv/Fm values originating from the ratio of very low Fo values as found for the
cultures that underwent lysis at the end of the lytic cycle.
3. Results
3.1. Acclimation to different irradiance levels
The light acclimation of noninfected P. globosa resulted in a 2-fold reduced
growth rate at LL (0.6 d-1, Table 1) as compared to HL and ML (1.2 and 1.1 d-1,
respectively). The maximal growth rate of M. pusilla was also subsaturated at LL (0.5 d1
, Table 1) whereas no difference could be detected between the ML and HL cultures
(0.7 d-1).
Besides differential maximal growth rates, the light scattering properties of the
P. globosa cells increased in response to increasing irradiance (1.5 and 2-fold higher at
HL than at LL for the SSC and FSC signals, respectively). In contrast, the cellular RFL
and the culture’s photochemical efficiency (Fv/Fm) was comparable at all irradiance
levels (Table 1). For M. pusilla, the cellular intrinsic characteristics also differed
between the cultures, but in different way than for P. globosa. Increasing irradiance
resulted in enhanced FSC signals (1.5-fold higher at HL than at LL) whereas the RFL
decreased substantially (2.5-fold lower HL than LL). In contrast, the cellular SSC signals
and the culture’s Fv/Fm signals were unaffected by the different irradiance regimes
(Table 1).
Table 1. Maximal growth rates (µmax, d-1), photochemical efficiency (Fv/Fm) and flow cytometric
characteristics (FSC, SSC, RFL) for P. globosa and M. pusilla cultures incubated at HL, ML and
LL (being 250, 100, and 25 µmol photons m-2 s-1 respectively). All flow cytometric signals are
standardized to an internal standard (0.95 µm beads). Values are means of duplicate samples. The
coefficient of variation was at max. 6 % of the mean.
P. globosa
-1
µmax (d )
Fv/Fm
FSC
SSC
RFL
M. pusilla
HL
ML
LL
HL
ML
LL
1.17
0.6
8.1
3.6
19.2
1.12
0.6
6.3
2.5
23.0
0.63
0.6
5.0
1.7
19.6
0.69
0.5
2.0
0.5
1.2
0.73
0.5
1.8
0.5
2.0
0.54
0.5
1.5
0.5
3.1
119
Chapter 6
Light-affected viral lysis
3.2. Effect of irradiance levels on algal host - virus interactions
Viral infection of P. globosa resulted in a gradual loss in Fv/Fm signal and
subsequent cell lysis (Fig. 1). Both processes were, however, dependent on the irradiance
level the algal host was acclimated to. Cell lysis started 4 to 7 h post infection for the HL
cultures, 11 to 15 h for the ML and 15 to 20 h for the LL. Full lysis of P. globosa
cultures showed a similar pattern and was achieved after 30 h for the HL treatment, 45 h
for the ML, and >70 h for the LL culture.
The latent period of PgV was 7 to 11 h at HL and ML, but it was prolonged at
LL (11 to 16 h, Fig. 1E). The burst size of the HL and LL cultures, calculated as the ratio
of the maximum net virus produced over net maximum decline of algal host, was half
(265 and 260 PgV·cell-1) that of the ML-adapted culture (525 PgV cell-1). The fraction of
infectious PgV produced was, however, 100% regardless of the light treatment.
NONINFECTED
INFECTED
1.5
a
6
0.9
4
0.6
2
0.3
0
0.0
1.0
0.8
0.6
0.4
0.2
0.0
c
1.2
b
0
20 40 60
Time (h)
HL
ML
LL
80
PgV
Fv/Fm
P. globosa
8
1.0
0.8
0.6
0.4
0.2
0.0
d
40
e
30
20
10
0
0
20
40
60
80
Time (h)
Figure 1. Viral infection of Phaeocystis globosa grown at 250, 100, and 25 µmol photons m-2 s-1
(HL, ML, and LL respectively). Abundance and photochemical efficiency (Fv/Fm) of P. globosa in
the (a, b) noninfected and (c, d) infected cultures, and abundance of P. globosa viruses PgV in (e)
the infected cultures. Values are means of duplicate series, normalized to To.
120
Chapter 6
Light-affected viral lysis
As for P. globosa, the viral infection of the M. pusilla cultures induced a
progressive decline in Fv/Fm and host cell abundance as compared to the noninfected
algal hosts (Fig. 2). The time until full lysis was somewhat delayed for the LL treatment
(>60 h) as compared to the 45 h for the HL and ML regimes. In contrast to PgV, the
MpV growth cycle was not affected by the different irradiance levels; the latent period
was 7 to 11 h and the burst size was 285 to 360 MpV cell-1. Like the PgV, the infectivity
of the MpVs produced was maximal (100%) and unaffected by the different light
regimes.
NONINFECTED
INFECTED
1.5
a
8
1.2
6
0.9
4
0.6
2
0.3
0
0.0
b
1.0
0.8
0.6
0.4
0.2
0.0
0
20
40
Time (h)
HL
ML
LL
c
d
1.0
0.8
0.6
0.4
0.2
0.0
60
e
16
MpV
Fv/Fm
M. pusilla
10
12
8
4
0
0
20
40
Time (h)
60
Figure 2. Viral infection of Micromonas pusilla grown at 250, 100, and 25 µmol photons m-2 s-1
(HL, ML, and LL respectively). Abundance and photochemical efficiency (Fv/Fm) of M. pusilla in
the (a, b) noninfected and (c, d) the infected cultures, and abundance of M. pusilla viruses MpV in
(e) the infected cultures. Values are means of duplicate series, normalized to To.
121
Chapter 6
Light-affected viral lysis
3.3. Darkness-induced effects on algal host-virus interactions
The growth of the noninfected P. globosa cultures halted upon transfer in
complete darkness (Fig. 3A) but their Fv/Fm signal remained high (> 0.6, Fig. 3B). As
observed for the virally infected cultures incubated in the light, the infected cultures,
when placed in darkness, showed a faster decline in Fv/Fm signal and cell abundance in
the culture adapted at increasing irradiance (Fig. 3D). Cell lysis was, however, only
partial in all dark treatments. The PgV production in the ML and the LL was strongly
reduced as compared to the HL-adapted culture (Fig. 3E). The LL culture did not show
any PgV production when placed in the dark. The PgV yields in the HL and ML cultures
transferred in the dark were both lower than in the respective light-incubated cultures
(Fig. 1E). When the dark-incubated cultures were placed back to the appropriate
irradiance after 2 days of darkness, additional algal cell lysis and PgV production was
recorded in the ML and LL cultures. The final viral yield in the ML infected culture
remained, however, 6-fold lower than in the light-incubated culture. In contrast, the final
viral yield in the LL culture was comparable to that in the LL culture incubated in the
light.
INFECTED
1.2
c
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
b
1.0
0.8
0.6
0.4
0.2
0.0
0
d
1.0
0.8
0.6
0.4
0.2
0.0
24
Time (h)
HL
ML
LL
48
15 e
12
PgV
Fv/Fm
P. globosa
NONINFECTED
a
9
6
3
0
0
24
48
Time (h)
Figure 3. Effect of darkness on viral infection of Phaeocystis globosa acclimated at 250, 100, and
25 µmol photons m-2 s-1 (HL, ML, and LL respectively). Abundance and photochemical efficiency
(Fv/Fm) of P. globosa in the (a, b) noninfected and (c, d) infected cultures and abundance of P.
globosa viruses PgV in (e) the dark-infected cultures. Values are normalized to To.
122
Chapter 6
Light-affected viral lysis
NONINFECTED
1.2
a
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
INFECTED
1.2
c
1.0
0.8
0.6
0.4
0.2
0.0
b
0
24
Time (h)
HL
ML
LL
48
MpV
Fv/Fm
M. pusilla
The growth of the noninfected M. pusilla in the dark was not only halted, as
observed for P. globosa, but a large proportion (40 %) of the cells died within 24 h (Fig.
4A). An independent repetition of this experiment gave similar results (data not shown).
The remaining cells showed a drop of 20% in the Fv/Fm signal (Fig. 4B). Viral infection
in the dark resulted in a stronger decline in Fv/Fm and in cell abundance than in the
noninfected cultures (Figs. 4C, D). This additional decline, however, did not result in the
production of MpV (Fig. 4E). When the dark-incubated cultures were placed back to the
light, the Fo signal of noninfected M. pusilla increased again. In contrast, the Fo signal
of the infected algal culture did not increase and still no MpV production was observed
in any of the infected cultures.
d
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
e
0
24
Time (h)
48
Figure 4. Effect of darkness on viral infection of Micromonas pusilla acclimated at 250, 100, and
25 µmol photons m-2 s-1 (HL, ML, and LL respectively). Abundance and photochemical efficiency
(Fv/Fm) of M. pusilla in the (a, b) noninfected and (c, d) infected cultures; and abundance of M.
pusilla viruses (MpV) in (e) the dark-infected cultures. Values are means of duplicate series,
normalized to To.
123
Chapter 6
Light-affected viral lysis
4. DISCUSSION
This study constitutes the first detailed report on how different irradiance levels
influence the interactions between host and virus for two marine phytoplankton of
ecological relevance, P. globosa and M. pusilla. Change in irradiance influenced
species-specifically the studied algal hosts and convincingly influenced viral infection of
P. globosa. Indeed, irradiance strongly affected the viral growth, latent period and burst
size of PgV, whereas the viral infection of M. pusilla was most affected by darkness.
For P. globosa, the production of PgV was optimal at ML (100 µmol photons
m-2 s-1) with a 7-11 h latent period and a burst size of 525 viruses cell-1. This irradiance
level corresponds to the so called “host standard culture conditions” at which a previous
characterization of PgV-07T was executed (Baudoux & Brussaard 2005). The here
reported latent period matches nicely that of the PgV group I to which PgV-07T belongs.
The observed burst size was, however, higher than that reported earlier (approx. 300
PgVs cell-1), which may originate from the use of a different algal strain in the present
study. The acclimation of P. globosa at LL and HL (25 and 250 µmol photons m-2 s-1)
negatively affected the PgV burst size (half the burst size at ML). In the LL-adapted
cultures, the latent period was furthermore prolonged by 4 hours. These results suggest
that viral replication was photolimited and photoinhibited in the LL and HL cultures,
respectively.
The physiological adaptations of P. globosa host that influenced viral infection
are unclear. The reduced burst size was associated with a subsaturated P. globosa growth
rate in the LL-adapted culture whereas growth rate remained saturated in HL-adapted
culture. This indicates that growth rate per se was not responsible for the reduced viral
yield, contrarily as what has been suggested for Paramecium Chlorella host-virus
systems (Van Etten et al. 1991). Nonetheless, the reduced host growth rate could still
explain the prolonged latent period of PgV observed for the LL regime. Some of
phenotypic adjustments of P. globosa at HL and LL, such as the changes in cellular light
scattering signals FSC and SSC, may partially explain some of the observed difference
in viral infection. The light scattering signals provide information on the structure, the
internal granulometry and the biovolume of the cells. The considerable reduction of
these parameters in the LL-adapted culture may be related to a decrease in cell volume
as observed earlier (Buma et al. 1993, Moisan & Mitchell 1999). Such effect on cell
volume may, in certain case, reduce the viral yield per cell due to packaging constraint
(Brown et al. 2006). The elucidation of physiological reasons underlying such changes
would require further investigations. Nevertheless, the here reported results have
interesting ecological implications. Indeed, P. globosa found in shallow, turbulent and
turbid coastal marine environments may experience a large range of light intensities. The
lower PgV production obtained under high and low light intensity conditions can lead to
a reduced encounter rate between virus and host which, in turn, can give the opportunity
for P. globosa to thrive.
124
Chapter 6
Light-affected viral lysis
Even in complete darkness, we found that PgV could still be propagated when
the host was acclimated at ML and HL, although the viral yield was reduced by 50 to
85%. On a physiological aspect, these observations imply that PgV production occur
only partially in absence of photosynthesis as also reported for Chlorella sp. and
Phaeocystis pouchetii (Van Etten et al. 1983, Bratbak et al. 1998). P. globosa acclimated
at ML and HL should, thus, have sufficient energy stores or produce enough energy in
the dark to support viral replication. Studies conducted on the viral infection of the
raphidophyte Heterosigma akashiwo suggested that, in darkness, virus replication could
use the ATP reserves or production via respiration or the energy generated by the cyclic
photophosphorylation (Juneau et al. 2003, Lawrence & Suttle 2004). In contrast, the total
inhibition of PgV replication in the LL cultures when placed in the dark suggests that the
energy level of the LL-adapted host was too low for the viruses to replicate. The host’s
ability to propagate viruses was, however, fully restored upon return to the light while
photosynthesis proceeded de novo. Sedimentation and resuspension of P. globosa
colonies as observed in the field (Cadée 1996; L. Peperzak pers. com.) may affect the
viral infection processes. Under conditions of light depletion, settled P. globosa colonies
are likely to shed their single cells (Peperzak et al. 2000), which can, in turn, be infected
by viruses (Brussaard et al. 2005). Complete vial replication will, then again, be possible
upon resuspension of the single cells into shallower waters with enhanced light
intensities.
The light regimes here tested impacted viral infection of M. pusilla in a
different way than P. globosa. The interactions between M. pusilla and MpV were
unaffected by the different irradiance levels. The 7 to 11 h latent period observed during
this study was comparable to those reported earlier (Waters & Chan 1982, Brussaard et
al. 1999). The MpV burst size of 285 to 360 viruses cell-1 was, however, high as
compared to the literature (49 to 230 MpV cell-1; Waters & Chan 1982, Brussaard et al.
1999, CB unpubl. data). The observed differential effect of light intensity on viral
infection of the two host-virus model systems may be caused by the species-specific
photoacclimation of M. pusilla and P. globosa. We indeed observed differential
adjustment of their intrinsic cellular characteristics (FSC, SSC and RFL) and their
growth. It is noteworthy that the effect of light limitation on growth rate was less severe
for M. pusilla than P. globosa (only reduced by 25 % compared to 50% for P. globosa).
Based on this parameter, M. pusilla seems to better accommodate the low irradiance than
P. globosa, and interestingly, the characteristics of the MpV growth cycle are unaffected
by LL. These observations strongly suggest that M. pusilla can experience a constant
viral control over a large range of irradiance, which may prevent a sudden proliferation
of this phytoplankter.
In spite of this tolerance to low irradiance, M. pusilla could not support viral
replication in darkness. The photosynthetic processes did not take place in the dark;
thereby it would be tempting to suggest that MpV replication is a photosynthesisdependent process. An earlier study also suggested that viral infection of M. pusilla
requires light to proceed (Waters & Chan 1982, Suttle 2000). Our study showed that, in
125
Chapter 6
Light-affected viral lysis
contrast to P. globosa, darkness not only inhibited MpV replication, but resulted also in
a substantial impairment of the host’s physiology. Already after 1 day of darkness, the
noninfected culture showed significant cells lysis (40% loss). We have as yet no
explanation for this abrupt collapse of M. pusilla culture, although this raises the option
of apoptosis. Apoptotic features have been reported for other marine phytoplankton
species but only upon long-term light deprivation (6 days, Berges & Falkowski 1998).
The impairment of M. pusilla physiology makes it difficult to conclude whether MpV
requires host photosynthesis to replicate. Upon return into light, the growth of the
noninfected cells was restored but, in contrast to P. globosa, the capacity of M. pusilla to
propagate MpV production was not recovered. Darkness irreversibly prevented the
production of MpV. In nature, it is unlikely that M. pusilla cells sink out of the euphotic
zone considering their micrometer size range (Raven 1998). However, M. pusilla
distributes in polar and boreal latitudes (Throndsen 1970, Not et al. 2005) where cells
may experience prolonged darkness due to quite frequent deep-mixing, particularly at
spring and winter time. Based on our results we may expect a reduction or even a
prevention of viral infection of M. pusilla during prolonged dark conditions in natural
environments due to a poor physiological condition of the algal host.
In summary, this study revealed that both P. globosa and M. pusilla are
characterized by distinct strategies to accommodate different light intensities. The
growth rate of P. globosa was more affected by changing light intensities than that of M.
pusilla. Phaeocystis globosa, however, survived darkness (up to 2 days) whereas M.
pusilla was severely impaired by dark incubations. This species-specific
photoacclimation likely determined the differential effect of irradiance on viral infection.
The changes in light intensities that phytoplankton naturally experience can thus be a
significant process regulating viral lysis processes in the ocean. Further investigations on
different algal host-virus model systems would be required to know whether the
differential light effect can be related to the ecological strategies of the algal hosts.
Acknowledgments. We are indebted to Maaike Appeldorf for her technical assistance.
This work was supported by the Research Council for Earth and Life Sciences (ALW)
with financial aid from the Netherlands Organization for Scientific Research (NWO).
Literature cited
Baudoux A-C., Brussaard CPD (2005) Characterization of different viruses infecting the
marine harmful algal bloom species Phaeocystis globosa. Virology 341:80-90
Baudoux A-C, Noordeloos AAM, Veldhuis MJW, Brussaard CPD (2006) Virally
induced mortality of Phaeocystis globosa during two spring blooms in temperate
coastal waters. Aquat Microb Ecol 44:207-217
126
Chapter 6
Light-affected viral lysis
Berges JA, Falkowski PG (1998) Physiological stress and cell death in marine
phytoplankton: Induction of protease in response to nitrogen and light limitation. J
Phycol 43:129-135
Bratbak G, Jacobsen A, Heldal M, Nagasaki K, Thingstad F (1998) Virus production in
Phaeocystis pouchetii and its relation to host cell growth and nutrition. Aquat
Microb Ecol 16:1-9
Brown CM, Lawrence JE, Campbell DA (2006) Are phytoplankton population density
maxima predictable through analysis of host and viral genomic DNA content? J
Mar Biol Assoc UK 86:491-498
Brussaard CPD (2004a) Viral control of phytoplankton populations - a review. J Euk
Microbiol 51:125-138
Brussaard CPD (2004b) Optimization of procedures for counting viruses by flow
cytometry. Appl Environ Microb 70:1506-1513
Brussaard CPD, Kuipers B, Veldhuis MJW (2005). A mesocosm study of Phaeocystis
globosa population dynamics I. Regulatory role of viruses in bloom control.
Harmful Algae 4:859-874
Brussaard CPD, Riegman R, Noordeloos AAM, Cadee GC, Witte H, Kop A J,
Nieuwland G, van Duyl FC, Bak RPM (1995) Effects of grazing, sedimentation
and phytoplankton cell lysis on the structure of a coastal pelagic food-web. Mar
Ecol Prog Ser 123:259-271
Brussaard CPD, Short SM, Frederickson CM, Suttle CA (2004) Isolation and
phylogenetic analysis of novel viruses infecting the phytoplankton Phaeocystis
globosa (Prymnesiophyceae). Appl Environ Microb 70:3700-3705
Brussaard CPD, Thyrhaug R, Marie D, Bratbak G (1999) Flow cytometric analyses of
viral infection in two marine phytoplankton species, Micromonas pusilla
(Prasinophyceae) and Phaeocystis pouchetii (Prymnesiophyceae). J Phycol
35:941-948
Buma AGH, Noordeloos AAM, Larsen J (1993) Strategies and kinetics of
photoacclimation in three antarctic nanophytoflagellates. J Phycol 29:407-417
Cadée GC (1996) Accumulation and sedimentation of Phaeocystis globosa in the Dutch
Wadden Sea. J Sea Res 36:321-327
Cadée GC, Hegeman J (2002) Phytoplankton in the Marsdiep at the end of the 20th
century; 30 years monitoring biomass, primary production, and Phaeocystis
blooms. J. Sea Res. 48:97-110
Cottrell MT, Suttle CA (1991) Wide-spread occurrence and clonal variation in viruses
which cause lysis of a cosmopolitan, eukaryotic marine phytoplankter,
Micromonas pusilla. Mar Ecol Prog Ser 78:1-9
Evans C, Archer SD, Jacquet S, Wilson WH (2003) Direct estimates of the contribution
of viral lysis and microzooplankton grazing to the decline of a Micromonas spp.
population. Aquat Microb Ecol 30:207-219
Falkowski PG, La Roche J (1991) Acclimation to spectral irradiance in algae. J Phycol
27:8-14
Geider RJ, La Roche J, Greene RM, Olaizola M (1993) Response of the photosynthetic
apparatus of Phaeodactylum tricornutum (Bacillariophyceae) to nitrate,
127
Chapter 6
Light-affected viral lysis
phosphate, or iron starvation. J. Phycol. 29:755-766
Guillard RRL (1975) Culture of phytoplankton for feeding marine invertebrates. In
Chanley, MH (Ed.), Culture of marine invertebrate animals. Plenum Press, New
York, pp. 29-60
Hurley MA, Roscoe ME (1983) Automated statistical analysis of microbial enumeration
by dilution series. J Appl Bacteriol 55:159-164
Juneau P, Lawrence JE, Suttle CA, Harrison PJ (2003) Effects of viral infection on
photosynthetic processes in the bloom-forming alga Heterosigma akashiwo.
Aquat Microb Ecol 31:9-17
Kuylenstierna M, Karlson B (1994) Seasonality and composition of pico- and
nanoplanktonic cyanobacteria and protist in the Skagerrak. Bot Mar 37:17-33
Lawrence JE, Chan AM, Suttle CA (2002) Viruses causing lysis of the toxic bloomforming alga Heterosigma akashiwo (Raphidophiceae) are widespread in coastal
sediments of British Columbia, Canada. Limnol Oceanogr 47:545-550
Lawrence JE, Suttle CA (2004) Effect of viral infection on sinking rates of Heterosigma
akashiwo and its implications for bloom termination. Aquat Microb Ecol 37:1-7
Moisan TA, Mitchell BG (1999) Photophysiological acclimation of Phaeocystis
antarctica Karsten under light limitation. Limnol Oceanogr 44:247-258
Not F, Latasa M, Marie D, Cariou T, Vaulot D, Simon N (2004) A single species,
Micromonas pusilla (Prasinophyceae), dominates the eukaryotic picoplankton in
the western English Channel. Appl Environ Microbiol 70:4064-4072
Not F, Massana R, Latasa M, Marie D, Colson C, Eikrem W, Predró-Alió C, Vaulot D,
Simon N (2005) Late summer community composition and abundance of
photosynthetic piceukaryotes in Norwegian and Barents Seas. Limnol Oceanogr
50:1677-1686
Peperzak L, Colijn F, Vrieling EG, Gieskes WWC, Peeters JCH (2000) Observations of
flagellates in colonies of Phaeocystis globosa (Prymnesiophyceae); a hypothesis
for their position in the life cycle. J Plank Res 22:2181-2203
Raven JA (1998) The twelfth Tansley Lecture. Small is beautiful: the
picophytoplankton. Funct Ecol 12:503-513
Suttle CA (1993) Enumeration and isolation of viruses. In Kemp PF, Sherr BF, Sherr
EF, Cole JJ (Eds.) Current methods in aquatic microbial ecology. Lewis
Publishers, Boca Raton, pp. 121-134
Suttle CA (2000) The ecological, evolutionary and geochemical consequences of viral
infection of cyanobacteria and eukaryotic algae. In Hurst CJ (Ed.) Viral Ecology.
Academic Press, London, pp. 248-286.
Throndsen J (1970) Flagellates from Arctic waters. Nytt Magasin Bot. 17:49-57
Throndsen J, Zingone A (1994) Micromonads of the Mediterranean Sea. G Bot Ital
128:1031-1044
Van Etten JL, Burbank DE, Xia Y, Meints RH (1983) Growth cycle of a virus, PCBV-1,
that infects Chlorella-like algae. Virology 126:117-125
Van Etten JL, Lane LC, Meints RH (1991) Viruses and virus-like particles in eukaryotic
algae. Microbiol Rev 55:586-620
Waters RE, Chan AT (1982) Micromonas pusilla virus: The virus growth cycle and
128
Chapter 6
Light-affected viral lysis
associated physiological events within the host cells; host range mutation. J Gen
Virol 63:199-206
Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea. BioScience 49:
781-788
129
Chapter 6
Light-affected viral lysis
130
Synthesis
The synthesis of the field studies presented in this thesis focuses on the extent to
which viral mediated mortality of phytoplankton varies across marine systems with
contrasting trophic status and what these results may imply for the organic carbon
cycling.
1. Methodological insights
Prior to a more general analysis, some methodological issues should be
considered. As emphasized in the previous chapters, the lack of appropriate
methodology has long constrained advances in viral ecology. Several approaches have
been developed to determine virally mediated mortality and all of them suffer from
limitations (Brussaard 2004, Suttle 2005). Although not shown in this thesis, the utility
of virally mediated mortality methods other than the modified dilution assay were
examined.
Among these, we attempted to use transmission electron microscopy (TEM) to
determine the frequency of visibly infected cells (FVIC) that can be converted to
estimate of virally mediated mortality. However, several methodological caveats have
constrained the reliability of our analysis. Firstly, TEM sample preparation includes
multiple steps of fixation and centrifugation that lead to inevitable losses of material.
The extent of these losses was particularly problematic when processing samples with
either low algal abundance or small-sized phytoplankton. Secondly, the discrimination of
the algal host species of interest in the TEM thin-sections is essential but nearly
impossible for the smaller-sized algal species (e.g. flagellated picophytoplankton) in
natural samples. Thirdly, this methodology relies on the use of different conversion
factors to covert the FVIC to absolute estimate of virally mediated mortality. However,
the estimation of the value of the conversion factors is fraught with problems and
therefore may distort the clear picture (Proctor et al. 1993, Waterbury & Valois 1993,
Binder 1999). Finally, the sample processing on board of a research vessel, the access to
a TEM and laborious analysis are other practical constraints.
Different adaptations of the phage production approaches from Wilhelm et al.
(2002) and Winter et al. (2004) were, furthermore, tested and compared to the dilutionbased phytoplankton viral lysis rates. In brief, a natural phytoplankton concentrate was
131
Synthesis
resuspended in virus-free diluent and the algal virus abundance was monitored in parallel
light and dark incubations for 24 h. The reduction in viral abundance prior to incubation
lowered the incidence of new infection; hence any increase in viral abundance is
assumed to result from an earlier infection event. Using an empirical burst size, the algal
virus production can be converted to virally mediated mortality. Although routinely used
for the determination of viral impact in bacterial communities, the application of these
methodologies to phytoplankton was more difficult. Indeed, the extensive sample
handling could alter phytoplankton physiology, particularly for the large-sized
phytoplankton. Furthermore, the initial phytoplankton concentration step occasionally
led to concentrate viruses along, which violates the critical assumption of this method,
i.e., preventing the incidence of new infection during the incubation. Finally, the
generally low abundance of putative algal viruses discriminated from the rest of the
marine viruses (based on flow cytometry analysis) limited in most cases the accuracy of
the measurements. In contrast, we obtained statistically valid viral lysis rates using the
modified dilution technique. Keeping in mind the constraints of the modified dilution
technique (Chapters 2, 4, and 5) but realizing its higher resolution and the possibility to
accurately discriminate different phytoplankton groups, we considered this approach as
the most appropriate for determining viral lysis rates during our field studies.
Potential limitations of the dilution method have indeed been addressed in this
dissertation, but as stated earlier, all existing assays determining rates of viral lysis has
its restrictions. The adapted dilution technique is only applicable to the numerically
dominating phytoplankton, the use of flow cytometry restricts the analysis to the cells
smaller than 20-30 µm in diameter, and the detection limit of the dilution technique is
unclear (Chapter 5). Thus, the here reported results are likely conservative estimates of
virally induced mortality. Nonetheless, the consistency of the dilution-based estimates
with independent assessment of total cell lysis (Chapter 2), specific (or putative) algal
virus abundance (Chapters 2, 4, and 5), and specific algal virus production (Chapter 2)
strengthen our results.
2. Extent of viral lysis in eutrophic vs. oligotrophic
conditions
Our results showed that algal viruses imposed substantial mortality rates to
specific phytoplankton populations in eutrophic (range 0.01 – 0.35 d-1, Chapter 2) as
well as in oligotrophic environments (range 0 – 0.20 d-1 in surface and 0 – 0.80 d-1 in
DCM waters as shown in Chapters 4 and 5, respectively). In order to compare the extent
of viral lysis among the different areas, we computed the total virally mediated carbon
losses and related them to phytoplankton carbon production (CP) using the adapted
formulas by Landry et al. (2000; Chapters 4 and 5). Phytoplankton carbon production is
an important biological process in the ocean because it starts off the organic carbon
cycle. Thus, expressing viral and grazing mediated algal mortality as the proportion of
132
Synthesis
phytoplankton CP is a relevant manner to evaluate the impact of these loss factors on the
organic carbon flux. The absolute amount of carbon released by phytoplankton viral
lysis varied widely among the studied areas (Table 1).
Table 1. Summary of carbon losses due to viral lysis and microzooplankton grazing and
corresponding fraction of phytoplankton carbon production removed. Values between brackets
indicate the mean estimate per site and the number of studied stations (n).
Eutrophic a,b
Oligotrophic
surface c,e
DCM d,e
Carbon losses due to viral lysis
(µg C L-1 d-1)
0.8 – 43
(14, n = 9)
0.0 – 2
(0.5, n = 5)
0.1 – 0.5
(0.2, n = 6)
Carbon losses due to grazing
(µg C L-1 d-1)
3 – 79
(27, n = 9)
1–6
(3, n = 5)
0.1 – 1
(0.5, n = 6)
CP lost by viral lysis (% d-1)
2 – 280
(47, n = 9)
0 – 16
(4, n = 5)
11 – 32
(24, n = 6)
CP lost by grazing (% d-1)
9 - 217
(69, n = 9)
16 – 89
(44, n = 5)
17 – 65
(44, n = 6)
a
Southern North Sea, 2m, spring 2003 and 2004 (Chapter 2)
Calculation restricted to Phaeocystis globosa single cells that numerically dominated the community of
phytoplankton < 20 µm.
c
North Sea, 5m, July 2003 (Chapter 5)
d
Subtropical northeastern Atlantic, 60 – 100 m, October 2002 (Chapter 4) and North Sea, 37 m, July 2003
(Chapter 5)
e
Calculation including the four picophytoplankton groups numerically dominating the studied area. The
detailed calculation of this parameter is given in Chapters 4 and 5.
b
The highest value, on average 14 ± 16 µg C L-1 d-1 (about 47 ± 90 % of CP d-1),
were obtained in the eutrophic environment during the Phaeocystis globosa spring
bloom. Such a substantial impact of viruses on phytoplankton CP during the P. globosa
bloom was expected as the probability of successful infection increased during the
development of the bloom. By comparison, viral lysis led to a relatively lower carbon
release of 0.2 - 0.5 µg C L-1 d-1 (on average) in the oligotrophic environments. This
comprised, however, a considerable fraction of the phytoplankton CP, particularly in the
DCM water layer (on average 24 ± 12% d-1 as compared to 4 ± 6% d-1 of the CP in
surface waters).
The virally mediated mortality rates from the surface waters of the oligotrophic
waters (4 ± 6% d-1) matched closely the few estimates reported for specific
133
Synthesis
phytoplankton in the literature (Suttle & Chan 1994, Cottrell & Suttle 1995). This result
furthermore supports the value of 2 - 10% of the algal production lost by viral lysis in
steady-state pelagic systems as predicted in the revised model of Jumars et al. (1989)
described by Wilhelm & Suttle (1999). However, this predicted estimate is considerably
lower than the 24 ± 12% of the CP lost by viral lysis in DCM oligotrophic waters. These
observations suggest that the impact of virus on phytoplankton may be higher in DCM
than in surface waters. It is important to realize that the DCM and surface viral lysis
estimates were obtained at different oligotrophic sites. We should, therefore, be cautious
when extrapolating these results. The algal group specific effect of irradiance on virally
induced algal mortality (Chapter 6) confirms such need for caution. Nonetheless, the
prevalence of up to 24% of the CP removed daily by viruses suggests that the impact of
viruses on marine phytoplankton in steady-state pelagic environments might be higher
than previously assumed.
Interestingly, we noted that the picophytoplankton groups prone to the highest
rates of viral lysis in DCM waters of the oligotrophic subtropical northeastern Atlantic
Ocean were also suggested to be responsible for the release of Fe-organic ligands
(Gerringa et al. 2006). Recently, Fe released by virally mediated lysis was shown to be
highly bioavailable to marine plankton (Poorvin et al. 2004). Our observation supports
the idea that viruses may be involved in the Fe cycling and adds to it that specific
phytoplankton groups may play an essential role in this process. Given that large parts of
the world’s ocean are thought to be Fe-limited (Moore et al. 2002), this finding is of
prime importance and therefore requires further investigation.
In summary, the presented results indicate that, next to microzooplankton
grazing, viral lysis significantly influences the flow of organic carbon. The trophic status
of the ecosystem, and arguably relevant environmental variables, seems to affect the
extent of viral lysis. The relatively low rates in the open oligotrophic waters appeared to
be rather constant in contrast to the viral lysis during spring algal blooms in eutrophic
waters. This may result in a steady amount of photosynthetically fixed carbon shunted
towards the regenerative food web. Our results provide one of the first data sets on actual
viral lysis rates of natural phytoplankton in different ecosystems; data that are urgently
needed for a better understanding of global biogeochemical cycling. The high viral lysis
rates at the DCM of oligotrophic waters, and the algal group specificity of virally
mediated mortality rates are other interesting findings in this thesis that actually
strengthen the present call for more detailed studies on the role of viruses in the ocean.
Literature cited
Binder B (1999) Reconsidering the relationship between virally induced bacterial
mortality and frequency of infected cells. Aquat Microb Ecol 18:207-215
Brussaard CPD (2004) Viral control of phytoplankton populations - a review. J Euk
134
Synthesis
Microbiol 51:125-138
Cottrell MT, Suttle CA (1995) Dynamics of a lytic virus infecting the photosynthetic
marine picoflagellate Micromonas pusilla. Limnol Oceanogr 40:730-739
Gerringa LJA, Veldhuis MJW, Timmermans KR, Sarthou G, de Baar HJW (2006) Covariance of dissolved Fe-binding ligands with phytoplankton characteristics in the
Canary Basin. Mar Chem 102:276-290
Jumars PA, Penry DL, Baross JA, Perry MJ, Frost BW (1989) Closing the microbial
loop - Dissolved carbon pathway to heterotrophic bacteria from incomplete
ingestion, digestion and absorption in animals. Deep-Sea Res I 36:483-495
Landry MR, Constantinou J, Latasa M, Brown SL, Bidigare RR, Ondrusek ME (2000)
Biological response to iron fertilization in the eastern equatorial Pacific (IronEx
II). III. Dynamics of phytoplankton growth and microzooplankton grazing. Mar
Ecol Prog Ser 201:57-72
Moore JK, Doney SC, Glover DM, Fung IY (2002) Iron cycling and nutrient limitation
patterns in surface waters of the world ocean. Deep-Sea Res II 49:463-507
Poorvin L, Rinta-Kanto JM, Hutchins DA, Wilhelm SW (2004) Viral release of iron and
its bioavailability to marine plankton. Limnol Oceanogr 49:1734-1741
Proctor LM, Okubo A, Fuhrman JA (1993) Calibrating estimates of phage-induced
mortality in marine bacteria - Ultrastructural studies of marine bacteriophage
development from one-step growth experiments. Microb Ecol 25:161-182
Suttle CA (2005) Viruses in the sea. Nature 437:356-361
Suttle CA, Chan AM (1994) Dynamics and distribution of cyanophages and their effect
on marine Synechococcus spp. Appl Environ Microbiol 60:3167-3174
Waterbury JB, Valois FW (1993) Resistance to co-occurring phages enables marine
Synechococcus communities to coexist with cyanophages abundant in seawater.
Appl Environ Microbiol 59:3393-3399
Wilhelm SW, Brigden SM, Suttle CA (2002) A dilution technique for the direct
measurement of viral production: A comparison in stratified and tidally mixed
coastal waters. Microb Ecol 43:168-173
Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea. BioScience
49:781-788
Winter C, Herndl GJ, Weinbauer MG (2004) Diel cycles in viral infection of
bacterioplankton in the North Sea. Aquat Microb Ecol 35:207-216
135
136
Summary
Over the last two decades, evidence has accumulated that viruses may regulate
phytoplankton population dynamics. All the major classes of phytoplankton are infected
by viruses. The increasing number of algal viruses brought into culture indicates high
viral diversity. Observations of virally infected cells using transmission electron
microscopy furthermore suggest that viruses can account for significant phytoplankton
mortality. In spite of this awareness many aspect of virally mediated mortality are poorly
understood and the relative impact of viral lysis as compared to other phytoplankton loss
factors (e.g. grazing) is essential unknown. Furthermore, the extent of viral lysis in
environments with contrasting trophic status is far from complete. This lack of
knowledge constrains our understanding of the global significance of virally mediated
mortality.
The partitioning of phytoplankton mortality into cell lysis and grazing is
important because these loss factors influence the structure and functioning of the basis
of marine food webs in different ways. While grazing basically channels phytoplankton
biomass to the higher trophic levels, viral lysis shunts cellular material to the pool of
dissolved organic matter (DOM) which is subsequently regenerated by bacteria. In this
thesis, we have examined the role of algal viruses as compared to microzooplankton for
phytoplankton mortality in environments with contrasting trophic status (eutrophic vs.
oligotrophic). Details are given on the viruses infecting the bloom forming species
Phaeocystis globosa and on the role of irradiance in regulating virus-algal host
interactions.
In the eutrophic waters of the southern North Sea, Phaeocystis globosa typically
develops dense spring blooms including flagellated cells (5-7 µm) and colonies (up to 12 cm). The monitoring of two consecutive spring blooms (2003 and 2004) revealed that
viruses actively contributed to the demise of P. globosa single cells (Chapter 2).Viral
lysis was the major cause of total cell lysis with rates up to 0.35 d-1 and even prevailed
over microzooplankton grazing rates at the end of the blooms. The abundance of putative
P. globosa virus (PgV) increased during the development of the bloom, concomitantly
with the increase in virally induced mortality of P. globosa cells. Our results,
furthermore, showed that the increase in abundance of infective PgVs was delayed as
compared to total putative PgVs and viral lysis. Interestingly, this delay in infective
PgVs coincided with the presence of transparent exopolymeric particles (TEP), which
are generated when colonies disrupt. Because viruses can adsorb to TEP, the fraction of
infective PgVs available for successful infection may be strongly reduced. This first
137
Summary
simultaneous estimate of viral lysis and microzooplankton grazing in the field showed
that viruses are important loss agents for P. globosa single cells during natural blooms,
resulting in virally induced mortality rates of P. globosa comparable to
microzooplankton grazing.
Twelve viruses specifically infecting P. globosa (PgV) were isolated from the
southern North Sea and partially characterized (Chapter 3). All PgV isolates shared
common phenotypic features with other algal viruses belonging to the family
Phycodnaviridae. An earlier study investigating the sequence of the conservative DNA
polymerase (pol) gene also showed that the PgVs formed a closely related monophyletic
group within the family Phycodnaviridae. Despite this close genetic relatedness, we
could discriminate four groups of PgV based on phenotypic characteristics. Two main
groups were distinguished based on particle size (150 vs. 100 nm), genome size (466 vs.
177 kb) and structural protein composition. The lytic cycle showed a latent period of 10
h for one group, but for the other group, two different latent periods for PgV (12 h vs. 16
h) were recorded. Host specificity and temperature sensitivity finally defined a fourth
group. The 4 distinct types of P. globosa viruses described in this study were collected
within a year from the southern coastal North Sea. Interestingly, different PgV groups
were found to co-occur in the same water sample. The coexistence of viruses infecting
different strains within a host species might indicate that viral infection is not only
regulated by the overall P. globosa abundance but also by clonal composition of host
population. Our results also indicate that co-occurring PgV groups may be competing for
the same specific host strain. The outcome of these competitive interactions may
influence the diversity of natural PgV population. In summary, our results illustrate that
viral infection can play an important role not only in P. globosa population dynamics but
also in the diversity of both host and virus community.
Two other field studies executed in the subtropical northeastern (NE) Atlantic
Ocean during autumn (Chapter 4) and the North Sea during summer (Chapter 5)
revealed that viruses may not only substantially contribute to the mortality of
phytoplankton in eutrophic waters, but also in oligotrophic environments. The
magnitude of viral lysis varied widely among phytoplankton groups as well as location.
Our data suggest depth-variability of the extent of viral lysis of phytoplankton, but this
needs further testing.
In the subtropical NE Atlantic Ocean, viral lysis and microzooplankton grazing rates of
four specific picophytoplankton groups (2 size classes of picoeukaryotes, and the
cyanobacteria Synechococcus and Prochlorococcus) were determined in the deep
chlorophyll maximum (DCM). The DCM was chosen because algal abundance in the
overlying surface water was too low for proper analysis using this method. Viral lysis,
with rates up to 0.8 d-1, tightly controlled the smallest picoeukaryotic group.
Interestingly, these high viral lysis rates (0.5 – 0.8 d-1) positively related to putative algal
virus abundance and concurred with the highest host growth rates (0.4 – 0.9 d-1). In
comparison, the numerically dominating cyanobacteria Synechococcus and
Prochlorococcus experienced no relevant loss due to viral lysis (on average 0.02 d-1).
138
Summary
Microzooplankton grazing (0.1 - 0.2 d-1) appeared to be the main loss factor for
Synechococcus, but neither viruses nor microzooplankton (0 - 0.1 d-1) seem to control
Prochlorococcus. These relatively low loss rates in combination with moderate growth
rates (on average 0.4 d-1) may explain the numerical dominance of Prochlorococcus (60
- 94% of total abundance) in the studied area. Overall, phytoplankton viral lysis led to a
considerable carbon release (0.1 - 0.3 µg C L-1 d-1), which corresponded to an average of
21% of the total carbon biomass produced by picophytoplankton in the DCM. Because
the DCM represents a specific layer of the euphotic zone, these results cannot
extrapolated to the entire euphotic layer. Nevertheless, our data suggest that viruses can
impact a greater fraction of phytoplankton carbon production than previously assumed.
The significance of viruses as mortality agents for different picophytoplankton
groups (three size classes of picoeukaryotes and Synechococcus) was also investigated
across different regions of the seasonally (during summer) oligotrophic North Sea. Viral
lysis and microzooplankton grazing were assessed in the surface waters where sizeable
phytoplankton populations could be detected. For practical reasons, only one of the
stations was sampled at both the surface layer and the DCM. The extent of virally
induced mortality rates varied across the studied area. The highest rates of viral lysis
were found for specific picoeukaryotic groups in the DCM as well as in surface waters
of a Norwegian coastal station. At the three other stations (located in the offshore surface
waters of the North Sea) rates of viral lysis were insignificant. Conversely to viruses,
microzooplankton actively grazed upon picoeukaryotes (on average 0.3 d-1) across the
studied area, whereas grazing on Synechococcus appeared to be restricted to the
Norwegian coastal station. Microzooplankton constituted the main loss factor in the
North Sea during summer, consuming on average 40% of the carbon produced by
picophytoplankton. By comparison, viruses induced a daily release of picophytoplankton
carbon production ranging between 0 to 32% (average 8%). This study suggests that the
partitioning of algal mortality into viral lysis and grazing varied widely among different
regions of the North Sea during summer.
The observed differences in the magnitude of viral lysis rates in Chapters 4 and
5 led to hypothesize that the variable underwater light intensity may be involved in the
regulation of successful viral infection and, thus, the significance of virally mediated
algal mortality. We speculate that the high solar radiation in the surface waters might
have reduced (or even prevented) viral infection of picophytoplankton and/or affected
the kinetics of viral growth. Although not implicitly formulated in Chapter 2,
Phaeocystis globosa also experiences large variation in irradiance in the turbulent, turbid
coastal waters during spring. Information on how different light intensities affect virusalgal host interactions is largely lacking therefore, we investigated this issue in more
detail (Chapter 6). A laboratory experiment conducted with P. globosa, as representative
of phytoplankton thriving in eutrophic waters, and Micromonas pusilla, as representative
of picophytoplankton abundant under oligotrophic conditions, revealed that irradiance
level (0, 25, 100, and 250 µmol photons m-2 s-1) species-specifically affected viral lysis.
Both low and high irradiances (25 and 250 µmol photons m-2 s-1) prolonged the viral
latent period and/or reduced burst size of PgV. Hence, the occurrence of such light
139
Summary
intensities in nature may lead to a reduced encounter rate between virus and host
(because of reduced PgV abundance) which, in turn, can give the opportunity for P.
globosa to flourish. In contrast, the interactions between M. pusilla and MpV were
unaffected by the different irradiance levels, but darkness inhibited MpV replication.
Considering that the picoeukaryote M. pusilla and its viruses are abundant in
oligotrophic waters, these results support the finding of high rates of viral lysis in light
attenuated environments (as observed in Chapters 4 and 5). However, high irradiance
might not always lead to reduced viral infection in surface picophytoplankton as initially
hypothesized. Enhanced UV radiation might be another factor reducing the impact of
viruses in surface phytoplankton population. Overall, this study emphasized the potential
of solar radiation as a factor influencing virally mediated mortality.
The investigations presented in this thesis shed new light on the importance of
marine viruses as drivers of phytoplankton mortality. Our results clearly show that
viruses, next to microzooplankton, can be significance mortality agents for
phytoplankton across ecosystems with contrasting trophic status. Hence, viral lysis of
phytoplankton may substantially influence the nutrient cycling in the ocean. Last but not
least, solar radiation may be a relevant environmental factor not only affecting the
growth of phytoplankton, but also virally mediated mortality. Based on these studies, it
appears essential to perform more targeted studies of ecologically relevant
phytoplankton groups, of virus-host interactions among the different strains of viruses
and hoss, and of underlying factors influencing the extent of viral lysis. Only then will
we be able to obtain a better understanding of the role of viruses in marine
phytoplankton mortality and within biogeochemical processes.
140
Samenvatting
In de afgelopen twee decennia, is gebleken dat virussen de phytoplankton
populatie activiteit kunnen beïnvloeden. De voornaamste fytoplankton klasses worden
besmet door virussen. De toename in gekweekte algen virussen laat een hoge virus
diversiteit zien. Waarnemingen met behulp van transmissie elektronen microscopie van
virus geïnfecteerde cellen suggereert dat virussen zorgen voor significante fytoplankton
sterfte. Ondanks deze kennis, zijn vele aspecten van virus geïnduceerde algensterfte nog
niet bekend. Daarnaast is de relatieve invloed van virus gemedieerde sterfte ten opzichte
van andere oorzaken van fytoplankton sterfte (vb. begrazing) onderbelicht. Verder is de
omvang van de invloed van virus gemedieerde sterfte in ecosystemen met verschillende
trofische niveaus verre van bekend. Dit gebrek aan kennis beperkt ons inzicht ten
aanzien van de globale invloed van virus gecontroleerde sterfte.
Het is belangrijk om binnen de fytoplankton sterfte onderscheid te maken
tussen cel gemedieerde sterfte en begrazing, omdat deze verlies factoren op
verschillende manieren de structuur en functionaliteit van de basis van marine
voedselketens beïnvloeden. Begrazing kanaliseert voornamelijk fytoplankton biomassa
naar hogere trofische niveaus, terwijl virus sterfte cel materiaal afvoert naar de poel van
opgelost organisch materiaal (Dissolved Organic Matter), dat vervolgens door bacteriën
wordt geregenereerd. In dit proefschrift wordt de rol onderzocht van algen virussen in
vergelijking met microzoöplankton (begrazing) ten aanzien van fytoplankton sterfte in
ecosystemen met een contrasterend trofisch niveau (eutroof versus oligotroof). Details
worden gegeven van de virussen die de bloei van de soort Phaeocystis globosa
beïnvloeden en tevens de rol van belichting op het reguleren van virus-alg gastheer
interacties.
In de eutrofe wateren van het zuidelijk deel van de Noordzee ontwikkelt
Phaeocystis globosa zich middels een, snelle, typerende bloei fase in de lente bestaande
uit geflageleerde cellen (5-7 µm) en kolonies (tot 1-2 cm). Na observatie van twee
achtereenvolgende bloei perioden (2003-2004) bleek dat de activiteit van virussen deels
verantwoordelijk waren voor de achteruitgang van de eencellige P. globosa cellen
(hoofdstuk 2). Virale lysis was de voornamelijkste oorzaak van totale cel lysis, dit met
een snelheid van 0.35 d-1. Opgemerkt kan worden dat dit getal hoger is dan de afbraak
snelheid door microzooplankton aan het eind van de bloei periode. De hoeveelheid van
vemeend P. globosa virus (PgV) nam toe tijdens de ontwikkeling van de bloei,
gelijktijdig met de toename viral geinduceerde moraliteit van de P. globosa cellen. Onze
resultaten toonden verder aan dat de verhoging van de in overvloed aanwezige
141
Samenvatting
besmettelijke PgVs werd vertraagd in vergelijking tot totale vemeende PgVs en virale
lysis. Opgemerkt kon worden dat deze vertraging van besmettelijke PgVs samen viel
met de aanwezigheid van transparante exopolymeric deeltjes (TEP), die worden
geproduceerd wanneer er cellen van de kolonies losbreken. Omdat de virussen aan TEPs
kunnen adsorberen, kan de fractie van besmettelijke PgVs beschikbaar voor succesvolle
besmetting sterk worden verminderd. Deze eerste gelijktijdige raming van virale lysis en
afbraak door microzooplankton in het veld toonde aan dat de virussen belangrijke
verliesagenten zijn voor eencellige P. globosa tijdens natuurlijke bloei periodes, die door
virusveroorzaakte afsterving snelheid van P. globosa resulteren in vergelijkbare afbraak
snelheden door microzooplankton.
Twaalf virussen die specifiek P. globosa (PgV) besmetten zijn geïsoleerd vanuit
de zuidelijke Noordzee en vervolgens gedeeltelijk gekenmerkt (Hoofdstuk 3). Alle
onderzochte PgV deelden gemeenschappelijke phenotypic eigenschappen met andere
algen gerelateerde virussen die tot de familie Phycodnaviridae behoren. Een eerdere
studie, met als doel het onderzoeken van het conservatieve polymerase (pol.) gen toonde
ook aan dat PgVs een nauw verwante monophyletic groep binnen de familie
Phycodnaviridae vormde. Ondanks deze sterke genetische gerelateerdheid, konden wij
vier groepen PgV onderscheiden, gebaseerd op phenotypische kenmerken.. Twee
belangrijke groepen werden onderscheiden op basis van deeltjesgrootte (150 versus 100
nm), genoomgrootte (466 versus 177 kb) en structurele eiwitsamenstelling. De lytic
cyclus toonde een latente periode van 10 h voor één groep, maar voor de andere groep,
konden twee verschillende latente periodes voor PgV (12 h versus 16 h) worden
geregistreerd. Specificiteit van de gastheer en de temperatuurgevoeligheid bepaalden
uiteindelijk nog een vierde groep. De 4 verschillende soorten P. globosa virussen, die in
deze studie worden beschreven, werden verzameld binnen een jaar van de zuidelijke kust
van de Noordzee. Een opmerkelijk feit is dat er verschillende groepen PgV zijn
gevonden in hetzelfde water monster. De coëxistentie van virussen die verschillende
soorten besmetten binnen een gastheersoort kunnen erop wijzen dat de virale besmetting
niet alleen voornamelijk door de overvloed van aanwezige P. globosa wordt geregeld
maar ook door de samenstelling van klonen van de gastheer populatie. Onze resultaten
wijzen ook erop dat de mede-voorkomende groepen PgV voor dezelfde specifieke
gastheer soorten zouden kunnen concurreren. Het resultaat van deze concurrerende
interactie kan de diversiteit van natuurlijke populatie van PgV beïnvloeden. Samengevat
illustreren onze resultaten dat de virale besmetting een belangrijke rol speelt, niet alleen
in P. globosa populatie dynamica maar ook in de diversiteit van zowel gastheer als
virusgemeenschap
Twee studies, die werden uitgevoerd in het noordoosten van de Atlantische
Oceaan (herfst) en in de Noord Zee (zomer) (Hoofdtuk 5), hebben laten zien dat
fytoplankton sterfte door virussen niet alleen een belangrijke rol speelt in eutrofe
wateren maar ook in oligotrofe wateren. In de oligotrofe wateren, virus geïnduceerde
sterfte van fytoplankton verschilde per geografische locatie, maar belangrijker: het
varieerde voor de verschillende fytoplanktongroepen. Verder suggereert onze data een
142
Samenvatting
afhankelijkheid met de diepte in de water kolom, maar dat moet nog verder onderzocht
worden.
In het diepe chlorofyll maximum (DCM) van het subtropische noordoosten van
de Atlantische Oceaan hebben we de invloed van virussen en predatie door
microzooplankton op de sterfte van fytoplankton onderzocht in vier verschillende
fytoplankton groepen (twee verschillende groten picoplankton en de cyanobacterien
Synechococcus en Prochlorococcus). We hebben gekozen om dit onderzoek uit te
voeren in de DCM omdat de celdichtheid van het fytoplankton in het oppervlakte water
niet hoog genoeg was. Fytoplankton sterfte door virussen met snelheden van 0.8 d-1
controleren de populatie van de kleinste picoeukaryote groep. De hoge sterfte van
fytoplankton door virussen (0.5 – 0.8 d-1) is positief gerelateerd aan zowel de
aanwezigheid van mogelijke algen specifieke virussen als de specifieke groeisnelheid
van de gastheer (0.4 – 0.9 d-1). Ter vergelijking, de dominant aanwezige cyanobacterien
Synechococcus en Prochlorococcus vertoonden geen relevante sterfte in aanwezigheid
van virussen (gemiddeld 0.02 d-1). Sterfte onder Synechococcus werd voornamelijk
veroorzaakt door microzooplankton predatie (0.1 - 0.2 d-1). Echter, de Prochlorococcus
populatie lijkt noch door microzooplankton (0 - 0.1 d-1) noch door virussen
gecontroleerd te worden. De relatief lage sterfte in combinatie met matige
groeisnelheden (gemiddeld
0.4 d-1) zou de dominante aanwezigheid van
Prochlorococcus (60 - 94% van de totale fytoplankton aanwezigheid) in deze gebieden
kunnen verklaren. De sterfte van het fytoplankton door de virussen leidt tot een
significante hoeveelheid vrijgekomend koolstof (0.1 - 0.3 µg C L-1 d-1), overeenkomend
met een gemiddeld 21% van de totale koolstofproductie door picofytoplankton in de
DCM. Omdat de DCM een specifieke laag vormt in de eufotische zone kunnen de
resulaten van deze processen niet worden gebruikt om de hele eufotische zone te
bechrijven. Desalniettemin, onze resultaten laten zien dat virussen meer invloed
uitoefenen op de koolstof productie door fytoplankton dan was aangenomen.
Of virussen een belangrijke rol spelen in de sterfte van verschillende
picofytoplankton groepen (Synechococcus en drie verschillende groten in picoekaryoten)
was ook onderzocht in verchillende regionen van de oligotrofe Noordzee gedurende de
zomer. Op plaatsen met genoeg fytoplankton werd onderzocht of sterfte door virussen of
predatie door zooplankton een rol speelde. Om practische redenen werden deze procesen
slechts van één station in zowel het oppervlakte water als de DCM onderzocht. De
fytoplankton sterfte varieerde over het bestudeerde gebied. De fytoplankton sterfte door
virussen was het hoogst voor specifieke picoeukaryote groepen in zowel de DCM als in
de oppervlakte wateren van de Noordzee voor de kust van Noorwegen. Fytoplankton
sterfte door virussen was minimaal op de drie overige stations in de oppervlakte wateren
in het midden van de Noordzee. Naast sterfte door virussen was ook predatie van
picoeukaryoten door microzooplankton een belangrijke factor in het gehele bestudeerde
gebied. Predatie van Synechococcus door microzooplankton werd alleen waargenomen
langs de kust van Noorwegen. Microzooplankton vormde de belangrijkste factor in de
sterfte van fytoplankton gedurende de zomer in de Noordzee, namelijk voor gemiddeld
40% van het koolstof geproduceerd door het picoplankton. Ter vergelijking, virussen
zorgen voor het vrijkomen van 0 tot 32% (gemiddeld 8%) van de picoplankton koolstof
143
Samenvatting
productie. Deze studie suggereert dat er een grote variatie bestaat in de mate van sterfte
van fytoplankton door virussen of door predatie in de verschillende gedeelten van de
Noordzee gedurende de zomer.
De waargenomen verschillen in de snelheid van virus gemedieerde sterfte
genoemd in Hoofdstukken 4 en 5 leidden tot de hypothese dat met name de
lichtintensiteit onderwater van invloed is op de regulatie van een succesvolle virale
infectie. Hiermee wordt het belang duidelijk van de, door virus geïnduceerde,
algensterfte. Een hoge mate van zonnestraling op het oppervlaktewater kan enerzijds
zorgen voor een verminderde (of zelfs afwezige) virale infectie bij picophytoplankton
en/of anderzijds de kinetiek van de virale groei beïnvloeden. Alhoewel niet impliciet
beschreven in Hoofdstuk 2, wordt Phaeocystis globosa wel degelijk blootgesteld aan
grote verschillen in straling in de woelige, troebele kustwateren in de lente periode.
Omdat kennis over hoe verschil in lichtintensiteit van invloed is op de interactie tussen
virus en alg (de gastheer) grotendeels ontbreekt, gaan we in Hoofdstuk 6 dieper op deze
kwestie in. In een labexperiment hierover werd P. globosa geacht als representatief voor
picophytoplankton in eutroof water en Micromonas pusilla als representatief voor
picophytoplankton in oligotroof water. Het bleek dat lichtintensiteit (gemeten in 4
niveaus: 0, 25, 100 en 250 µmol photons m-2 s-1) virus gemedieerde sterfte soortspecifiek
beïnvloedde. In het geval van P. globosa werd de latente periode van virussen verlengd
door zowel extreem lage als hoge lichtintensiteit (respectievelijk 25 en 250 µmol
photons m-2 s-1). Tevens werd de grootte van de uitbraak van PgV verminderd. Het
voorkomen van zulke extreme lichtcondities in de natuur kunnen dus leiden tot een
verminderde kans op een ontmoeting van een virus met haar gastheer (door de lagere
hoeveelheid PgV). Hierdoor kan de gastheer, in dit geval P. globosa, floreren. In
tegenstelling tot P. globosa, was de lichtintensiteit niet aantoonbaar van invloed op de
interactie tussen M. Pusilla en MpV, alhoewel MpV zich niet meer kon repliceren in het
donker (0 µmol photons m-2 s-1). Doordat de picoeukaryoot M. pusilla in hoge
concentratie voorkomt in oligotroof water, ondersteunen deze resultaten de bevindingen
uit de Hoofdstukken 4 en 5, waarin wordt beschreven dat hoge mate van virus
gemedieerde sterfte met name voorkomt in een omgeving met een lage lichtintensiteit.
En hoge lichtintensiteit leidt echter niet altijd tot een verminderde virus infectie in
picophytoplankton levend in het oppervlaktewater, zoals in eerste instantie werd
aangenomen. Ook verhoogde UV straling kan van invloed zijn op de vermindering van
de virus infectie bij de phytoplankton populatie in het oppervlaktewater. Ter conclusie
legt deze studie de nadruk op de potentieel belangrijke invloed van zonnestraling op
virus gemedieerde sterfte.
Het onderzoek beschreven in dit proefschrift laat een nieuw licht schijnen over
de belangrijkheid van marine virussen als veroorzakers van fytoplankton sterfte. Onze
resultaten laten duidelijk zien, dat virussen, naast microzoöplankton, significante sterfte
middelen kunnen zijn in ecosystemen met verschillende trofische niveaus. Dus virus
gemedieerde sterfte van fytoplankton kan een substantiële invloed hebben op de nutrient
cyclus in de oceaan. Niet het minst belangrijk is de invloed van zonnestraling, dat een
144
Samenvatting
relevante omgevingsfactor is, niet alleen voor de groei van fytoplankton, maar ook voor
virus gemedieerde sterfte. Naar aanleiding van deze studie lijkt het essentieel om meer
gerichte onderzoeken uit te voeren; aan ecologische relevante fytoplankton groepen,
aan virus gastheer interacties tussen de verschillende spanningen tussen virussen en
gastheren en aan de onderliggende factoren, die de omvang van virus gemedieerde
sterfte beïnvloeden. Alleen dan kunnen we de rol van virussen in mariene fytoplankton
sterfte en biogeochemische processen beter begrijpen.
145
146
Acknowledgments
The past five years have been quite intensive but the achievement of this
dissertation would simply not have been possible without the help, the advice, and the
support of my colleagues and friends.
First of all, I like to express all my gratitude to my co-promotor, Corina.
Through the past five years, Corina has patiently guided me to unravel the mysteries of
algal virus ecology. Despite her overloaded schedule, Corina has always made time for
me to discuss scientific and more personal issues. The studies presented in this thesis are
not only the results of countless hours spent in the laboratory; each chapter has been
extensively discussed with Corina. Her rigorous comments and the knowledge she shared
with me has undoubtedly improved this thesis. I hope this long collaboration will last in
the future.
Gerhard, my promotor, has created a very stimulating working- and socializingatmosphere within the Bio department which he is heading. I have been fortunate to have
had the opportunity to work as graduate student in this group. I would like to thank
Gerhard for his support and advice, particularly at a final stage of my PhD.
Colleagues in the Bio department have always considered my work with much
interest and they always had helpful comments and advice. Although I cannot list them
all, I would like to particularly acknowledge Marcel, Anna, Harry, Bouwe, and Joaquin
for their invaluable help, discussions and support over the past few years.
The studies presented in this thesis would not have been possible without the
contribution of the many students who visited the lab. Many thanks go to Maaike,
Caroline, Ronnie, Gijs, and Klaske for their help and for bringing such a fresh and great
atmosphere in the lab.
I am furthermore very thankful to Wim and Hans from the computer department
and Nelleke from the repro who have (very patiently) helped me solving my (countless)
“computer” troubles over the past few years. Warm thanks also go to Evaline, Karel, and
Jan from the nutrient lab, Santiago Gonzalez, and Margriet, Ronald, and Taco from the
data management for their assistance and expertise.
Although most my time was spent onshore, parts of this thesis were conducted
onboard of the R. V. Pelagia where the crew has always been of excellent assistance.
Thanks for your help and fun onboard!
I am deeply embedded to Laëtitia and Vincent for their help in writing up a
proposal at a final stage of my PhD; your input and support has greatly contributed to the
achievement of this dissertation. Warm thanks go to the both of you.
Many friends have come and gone over the past five years. I would like to
infinitely thank my housemates Micha and Tanya, Ben and Joana, and Teresa and also
Yann, Phil, Furu, Isabel, Jéjé, Jasper, Nénette, Jokin, the Vézinas, Dephine, Pierrick,
Gaëlle, Denis, Khalid, Lorentz, Pedro, Marjolijn, Piet, Neven, Christian, Vero, Eva,
147
Thomas, Txetxu, Conny, Jorg, Wilhelm, Marta, and Craig for the great moments, laughs,
parties hosted by the 12 Balken (lots of thanks to Sander and Dirk!) and, most
importantly, for your help and your friendship. Very special thanks go to Tanya, Hélène,
Joaquin, Teresa, and Judith for the long counseling discussions, for your wise advice, for
keeping me running and smiling!
My stay on Texel would simply not have been the same without the company of
the three tenants of Haffelderweg 1, Phil, Furu, and Yann, and their regular visitors Isabel
and Jéjé. I spent unforgettable moments with Isabel; I just wish, a bit selfishly, that she
had stayed longer on “the island”. Special thanks to Jéjé for the wonderful Korean cuisine
accompanied with at least as wonderful Bordeaux wine that we enjoyed during cozy
evenings on the “continent”. Furu has been ever-present during my Texelse experience, I
am short in words to express how much her friendship means to me; thanks for all you
have done and shared with me. Phil is certainly one of the most important person I met on
Texel. Phil’s “company has been of a value that I couldn’t put into words here, nor need
to because hopefully he knows how important he is”; these words are his, but I wish I
wrote them first!! Thanks for everything, my Hosegood. Finally, Yann has probably
endured the worst aspects of my PhD: in spite of my doubt and apprehensions, he
provided me an unconditional support, he patiently encouraged me, he took care of me in
every possible ways and kept me going. I would certainly not have managed the past few
years without you besides me, nor would I in the future… I would like to deeply thank
you for your love and all you have done over these years. Some more French words are
for you at the end of these acknowledgments.
Durant ces cinq dernières années, mes visites à la maison ont été trop peu
nombreuses et surtout trop courtes. Toutefois, chacune de ces visites ont rendu les retours
en hollande un petit plus difficiles car j’y retrouvais des personnes exceptionnelles. Ont
fortement contribué à ces retours difficiles les retrouvailles avec mes potos Youri, Ludo,
Hélène, Thierry, Denis, la petite et la grande Sophie, Erwan et Sam que je tiens à
remercier pour leur fidèle amitié malgré mes longues absences.
Je souhaiterais très chaleureusement remercier Jacqueline et François Bozec qui
m’ont accueilli lors de mes passages en Bretagne. Du fond du cœur, merci pour votre
hospitalité, votre gentillesse, vos encouragements et tous les fabuleux petits plats !
La réalisation de cette thèse n’aurait certainement pas été possible sans le
soutien de mes parents, de mes trois sœurs et de mon frère. J’aimerais les remercier
d’avoir respecté mes décisions et surtout d’avoir été à mes côtés dans les moments de
doute, particulièrement durant la fin de cette thèse. Au travers ces quelques lignes, je
voudrais vous témoigner ma reconnaissance et mon amour ; ils sont infinis…
Je souhaiterais finalement remercier Yann qui a patiemment supporté, et bien
plus que quiconque, mes doutes et appréhensions au cours de ces dernières années. Grâce
à ta présence, tes encouragements, ton infinie patience (sûrement ton impatience aussi) et
ton amour, cette thèse est à présent terminée. Se faisant, nous tournons un chapitre
important de notre vie. J’aimerais refermer ce chapitre de vie / de thèse par ces quelques
mots : « … après ces longues années de thèse, ils continuèrent à vivre heureux, ensemble,
et pendant de nombreuses années… ». Merci pour tout, je t’aime…
148