MARINE BIOLOGY

Editor-in-Chief

John H. Steele

Marine Policy Center, Woods Hole Oceanographic Institution, Woods Hole,Massachusetts, USA

Editors

Steve A. Thorpe

National Oceanography Centre, University of Southampton,Southampton, UK

School of Ocean Sciences, Bangor University, Menai Bridge, Anglesey, UK

Karl K. TurekianYale University, Department of Geology and Geophysics, New Haven, Connecticut,

USA

Subject Area Volumes from the Second Edition

Climate & Oceans edited by Karl K. Turekian

Elements of Physical Oceanography edited by Steve A. Thorpe

Marine Biology edited by John H. Steele

Marine Chemistry & Geochemistry edited by Karl K. Turekian

Marine Ecological Processes edited by John H. Steele

Marine Geology & Geophysics edited by Karl K. Turekian

Marine Policy & Economics guest edited by Porter Hoagland, Marine Policy Center,

Woods Hole Oceanographic Institution, Woods Hole, Massachusetts

Measurement Techniques, Sensors & Platforms edited by Steve A. Thorpe

Ocean Currents edited by Steve A. Thorpe

ENCYLOPEDIAOF

OCEAN SCIENCES:MARINE BIOLOGY

Editor

JOHN H. STEELE

Boston • Heidelberg • London • New York • Oxford Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo

Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier32 Jamestown Road, London NW1 7BY, UK

30 Corporate Drive, Suite 400, Burlington, MA 01803, USA

525 B Street, Suite 1900, San Diego, CA 92101-4495, USA

Copyright ^ 2009 Elsevier Ltd. All rights reserved

Material in the work originally appeared in Encyclopedia of Ocean Sciences (Elsevier Ltd., 2001) and Encyclopedia of Ocean Sciences,2nd Edition (Elsevier Ltd., 2009), edited by John H. Steele, Steve A. Thorpe and Karl K. Turekian.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic,mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44)(0)

1865 843830; fax (+44)(0) 1865 853333; email: permissions@elsevier.com. Alternatively you can submit your request online by visiting

the Elsevier website at (http://elsevier.com/locate/permissions), and selecting Obtaining permissions to use Elsevier material

NoticeNo responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability,

negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein,

Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Control Number: 2009932544

ISBN: 978-0-08-096480-5

For information on all Elsevier publications

visit our website at www.elsevierdirect.com

PRINTED AND BOUND IN ITALY09 10 11 12 13 10 9 8 7 6 5 4 3 2 1

CONTENTS

Marine Biology: Introduction ix

PLANKTON & NEKTON

Plankton Overview M M Mullin 3

Marine Plankton Communities G-A Paffenhofer 5

Plankton Viruses J Fuhrman, I Hewson 13

Bacterioplankton H W Ducklow 21

Protozoa, Radiolarians O R Anderson 28

Protozoa, Planktonic Foraminifera R Schiebel, C Hemleben 33

Copepods R Harris 40

Gelatinous Zooplankton L P Madin, G R Harbison 51

Krill E J Murphy 62

Nekton W G Pearcy, R D Brodeur 71

Cephalopods P Boyle 78

Phytoplankton Size Structure E Maranon 85

Primary Production Methods J J Cullen 93

Primary Production Processes J A Raven 100

Primary Production Distribution S Sathyendranath, T Platt 105

Bioluminescence P J Herring, E A Widder 111

BENTHOS

Benthic Organisms Overview P F Kingston 123

Microphytobenthos G J C Underwood 132

Phytobenthos M Wilkinson 140

Benthic Foraminifera A J Gooday 147

Meiobenthos B C Coull, G T Chandler 159

Macrobenthos J D Gage 165

Deep-sea Fauna P V R Snelgrove, J F Grassle 176

Cold-Water Coral Reefs J M Roberts 188

Benthic Boundary Layer Effects D J Wildish 199

FISH BIOLOGY

Fish: General Review Q Bone 209

Antarctic Fishes I Everson 218

v

Coral Reef Fishes M A Hixon 222

Deep-sea Fishes J D M Gordon 227

Intertidal Fishes R N Gibson 233

Mesopelagic Fishes A G V Salvanes, J B Kristoffersen 239

Pelagic Fishes D H Cushing 246

Salmonids D Mills 252

Eels J D McCleave 262

Fish Ecophysiology J Davenport 272

Fish Feeding and Foraging P J B Hart 279

Fish Larvae E D Houde 286

Fish Locomotion J J Videler 297

Fish Migration, Horizontal G P Arnold 307

Fish Migration, Vertical J D Neilson, R I Perry 316

Fish Predation and Mortality K M Bailey, J T Duffy-Anderson 322

Fish Reproduction J H S Blaxter 330

Fish Schooling T J Pitcher 337

Fish Vision R H Douglas 350

Fish: Demersal Fish (Life Histories, Behavior, Adaptations) O A Bergstad 363

Fish: Hearing, Lateral Lines (Mechanisms, Role in Behavior, Adaptations to LifeUnderwater) A N Popper, D M Higgs 372

MARINE MAMMALS

Marine Mammal Overview P L Tyack 381

Baleen Whales J L Bannister 391

Sperm Whales and Beaked Whales S K Hooker 403

Dolphins and Porpoises R S Wells 411

Seals I L Boyd 424

Sirenians T J O’Shea, J A Powell 431

Sea Otters J L Bodkin 442

Sea Turtles F V Paladino, S J Morreale 450

Marine Mammal Diving Physiology G L Kooyman 458

Marine Mammal Evolution and Taxonomy J E Heyning 465

Marine Mammal Migrations and Movement Patterns P J Corkeron, S M Van Parijs 472

Marine Mammal Social Organization and Communication P L Tyack 481

Marine Mammal Trophic Levels and Interactions A W Trites 488

vi CONTENTS

BIRDS

Seabirds: An Overview G L Hunt, Jr. 497

Alcidae T Gaston 503

Laridae, Sternidae and Rynchopidae J Burger, M Gochfeld 510

Pelecaniformes D Siegel-Causey 522

Phalaropes M Rubega 531

Procellariiformes K C Hamer 539

Sphenisciformes L S Davis 546

Seabird Conservation J Burger 555

Seabird Foraging Ecology L T Balance, D G Ainley, G L Hunt Jr. 562

Seabird Migration L B Spear 571

Seabird Population Dynamics G L Hunt 582

APPENDIX

Appendix 9. Taxonomic Outline of Marine Organisms L P Madin 589

INDEX 601

CONTENTS vii

MARINE BIOLOGY: INTRODUCTION

This volume is a selection of articles from the second, electronic, edition of the Encyclopedia of OceanScience. It is one of nine volumes that focus on particular aspects of marine studies. Marine Biology not onlycovers a great variety of plant and animal species but refers to diverse aspects of their physical, chemical andhuman environment.

The volume is divided into the traditional sub-disciplines: Plankton, Benthos, Fish, Marine Mammals andSeabirds. Within each category, there are articles on the main taxonomic groups, but also articles dealingwith important processes such as primary production of phytoplankton, fish locomotion, feeding and for-aging, marine mammal diving physiology and seabird conservation.

Marine organisms have an intimate relation with their fluid environment. Ocean currents play a large rolein determining their migrations and vertical mixing and advection control the nutrient supply that regulatestheir food production. Longer term changes in these physical processes cause major stresses on populationsand communities. This close coupling of ocean physics and biology is a theme of many articles, especiallythose concerned with the impact of climatic changes on plankton, marine mammals and seabirds.

There are others stresses on marine communities. The general topic of marine pollution is dealt with in aseparate volume. The impact of fisheries not only on commercial stocks of fish but also on the remainder oftheir ecosystems, is considered at length in a companion volume to this one dealing broadly with ecologicalprocesses.

Each section of this volume opens with an ‘‘Overview’’ article written by the Section Editor responsible forthis theme within the Encyclopedia. These Section Editors were also involved in the selection of authors forthe individual topics. The Editors of the Encyclopedia are in their debt for their work in ensuring the qualityand coverage of these articles.

Given the breadth of topics under the rubric of Marine Biology and their inter-relation with other aspectsof ocean science, this one volume must be considered as a summary or introduction. For this reason eacharticle has, not only a further reading list, but also references to articles in the Encyclopedia or in othervolumes in this series.

The articles in this volume could not have been produced without the considerable help of the members ofthe Editorial Advisory Board of the Encyclopedia’s second edition, from which these articles were chosen.The board provided advice and suggestions about the content and authorship of particular subject areascovered in the Encyclopedia. In addition to thanking the authors of the articles in this volume, the Editorswish to thank the members of the Editorial Board for the time they gave to identify and encourage authors, toread and comment on (and sometimes to suggest improvements to) the written articles, and to make thisventure possible.

John H. SteeleEditor

ix

PLANKTON & NEKTON

PLANKTON OVERVIEW

M. M. Mullin, Scripps Institution of Oceanography, LaJolla, California, USA

Copyright & 2001 Elsevier Ltd.

The category of marine life known as plankton rep-resents the first step in the food web of the ocean(and of large bodies of fresh water), and componentsof the plankton are food for many of the fish har-vested by humans and for the baleen whales. Theplankton play a major role in cycling of chemicalelements in the ocean, and thereby also affect thechemical composition of sea water and air (throughexchange of gases between the sea and the overlyingatmosphere). In the parts of the ocean whereplanktonic life is abundant, the mineral remains ofmembers of the plankton are major contributors todeep-sea sediments, both affecting the chemistry ofthe sediments and providing a micropaleontologicalrecord of great value in reconstructing the earth’shistory.

‘Plankton’ refers to ‘drifting’, and describes or-ganisms living in the water column (rather than onthe bottom – the benthos) and too small and/or weakto move long distances independently of the ocean’scurrents. However, the distinction between planktonand nekton (powerfully swimming animals) can bedifficult to make, and is often based more on thetraditional method of sampling than on the organ-isms themselves.

Although horizontal movement of plankton atkilometer scales is passive, the metazoan zoo-plankton nearly all perform vertical migrations onscales of 10s to 100s of meters. This depth range cantake them from the near surface lighted waters wherethe phytoplankton grow, to deeper, darker andusually colder environments. These migrations aregenerally diurnal, going deeper during the day, orseasonal, moving to deeper waters during the wintermonths to return to the surface around the time thatphytoplankton production starts. The former patterncan serve various purposes: escaping visual predatorsand scanning the watercolumn for food. (It should benoted that predators such as pelagic fish also migratediurnally.) Seasonal descent to greater depths is acommon feature for several copepod species and mayconserve energy at a time when food is scarce in theupper layers. However, vertical migration has an-other role. Because of differences in current strength

and direction between surface and deeper layers inthe ocean, time spent in deeper water acts as atransport mechanism relative to the near surfacelayers. On a daily basis this process can take plank-ton into different food concentrations. Seasonally,this effective ‘migration’ can complete a spatial lifecycle.

The plankton can be subdivided along functionallines and in terms of size. The size category, pico-plankton (0.2–2.0 mm), is approximately equivalentto the functional category, bacterioplankton; mostphytoplankton (single-celled plants or colonies) andprotozooplankton (single-celled animals) are nano-or microplankton (2.0–20 mm and 20–200 mm, re-spectively). The metazoan zooplankton (animals, the‘insects of the sea’) includes large medusae and si-phonophores several meters in length. Size is moreimportant in oceanic than in terrestrial ecosystemsbecause most of the plants are small (the floatingseaweed, Sargassum, being the notable exception),predators generally ingest their prey whole (there isno hard surface on which to rest prey while dis-membering it), and the early life stages of many typesof zooplankton are approximately the same size asthe larger types of phytoplankton. Therefore, whilethe dependence on light for photosynthesis is char-acteristic of the phytoplankton, the concepts of‘herbivore’ and ‘carnivore’ can be ambiguous whenapplied to zooplankton, since potential plant andanimal prey overlap in size and can be equivalentsources of food. Though rabbits do not eat babyfoxes on land, analogous ontogenetic role-switchingis very common in the plankton.

Among the animals, holoplanktonic species arethose that spend their entire life in the plankton,whereas many benthic invertebrates have mer-oplanktonic larvae that are temporarily part of theplankton. Larval fish are also a temporary part of theplankton, becoming part of the nekton as they grow.There are also terms or prefixes indicating specialhabitats, such as ‘neuston’ to describe zooplanktonicspecies whose distribution is restricted to withina few centimeters of the sea’s surface, or ‘abysso-plankton’ to describe animals living only in thedeepest waters of the ocean. Groups of such speciesform communities (see below).

Since the phytoplankton depend on sunlight forphotosynthesis, this category of plankton occurs al-most entirely from the surface to 50–200 m of theocean – the euphotic depth (where light intensity is0.1–1% of full surface sunlight). Nutrients such as

3

nitrate and phosphate are incorporated into proto-plasm in company with photosynthesis, and returnedto dissolved form by excretion or remineralization ofdead organic matter (particulate detritus). Since muchof the latter process occurs after sinking of thedetritus, uptake of nutrients and their regenerationare partially separated vertically. Where and whenphotosynthesis is proceeding actively and verticalmixing is not excessive, a near-surface layer of lownutrient concentrations is separated from a layer ofabundant nutrients, some distance below the euphoticdepth, by a nutricline (a layer in which nutrient con-centrations increase rapidly with depth). Therefore,the spatial and temporal relations between theeuphotic depth (dependent on light intensity at thesurface and the turbidity of the water), the nutricline,and the pycnocline (a layer in which density increasesrapidly with depth) are important determinants of theabundance and productivity of phytoplankton.

Zooplankton is typically more concentratedwithin the euphotic zone than in deeper waters, butbecause of sinking of detritus and diel vertical mi-gration of some species into and out of the euphoticzone, organic matter is supplied and various types ofzooplankton (and bacterioplankton and nekton) canbe found at all depths in the ocean. An exception isanoxic zones such as the deep waters of the BlackSea, although certainly types of bacterioplanktonthat use molecules other than oxygen for their me-tabolism are in fact concentrated there.

Even though the distributions of planktonic spe-cies are dependent on currents, species are not uni-formly distributed throughout the ocean. Speciestend to be confined to particular large water masses,because of physiological constraints and inimicalinteractions with other species. Groups of species,from small invertebrates to active tuna, seem to‘recognize’ the same boundaries in the oceans, in thesense that their patterns of distribution are similar.Such groups are called ‘assemblages’ (when em-phasizing their statistical reality, occurring togethermore than expected by chance) or ‘communities’(when emphasizing the functional relations betweenthe members in food webs), though terms such as‘biocoenoses’ can be found in older literature. Thus,one can identify ‘central water mass,’ ‘subantarctic,’‘equatorial,’ and ‘boreal’ assemblages associatedwith water masses defined by temperature and

salinity; ‘neritic’ (i.e. nearshore) versus ‘oceanic’assemblages with respect to depth of water overwhich they occur, and ‘neustonic’ (i.e. air–sea inter-face), ‘epipelagic,’ ‘mesopelagic,’ ‘bathypelagic,’ and‘abyssopelagic’ for assemblages distinguished by thedepth at which they occur. Within many of thesethere may be seasonally distinguishable assemblagesof organisms, especially those with life spans of lessthan one year.

Regions which are boundaries between assem-blages are sometimes called ecotones or transitionzones; they generally contain a mixture of speciesfrom both sides, and (as in the transition zone be-tween subpolar and central water mass assemblages)may also have an assemblage of species that occuronly in the transition region.

Despite the statistical association between assem-blages and water masses or depth zones, it is far fromclear that the factor that actually limits distributionis the temperature/salinity or depth that physicallydefines the water mass or zone. It is likely that a fewimportant species have physiological limits confiningthem to a zone, and the other members of the as-semblage are somehow linked to those species func-tionally, rather than being themselves physiologicallyconstrained. Limits can be imposed on certain lifestage, such as the epipelagic larvae of meso- orbathypelagic species, creating patterns that reflect theenvironment of the sensitive life stage rather than theadult. Conversely, meroplanktonic larvae, such asthe phyllosome of spiny lobsters, can often be foundfar away from the shallow waters that are a suitablehabitat for the adults.

See also

Bacterioplankton. Gelatinous Zooplankton. Protozoa,Planktonic Foraminifera.

Further Reading

Cushing DH (1995) Population Production and Regulationin the Sea. Cambridge: Cambridge University Press.

Longhurst A (1998) Ecological Geography of the Sea. NewYork: Academic Press.

Mullin MM (1993) Webs and Scales. Seattle: University ofWashington Press.

4 PLANKTON OVERVIEW

MARINE PLANKTON COMMUNITIES

G.-A. Paffenhofer, Skidaway Institute ofOceanography, Savannah, GA, USA

& 2009 Elsevier Ltd. All rights reserved.

Introduction

By definition, a community is an interacting popu-lation of various kinds of individuals (species) in acommon location (Webster’s Collegiate Dictionary,1977).

The objective of this article is to provide generalinformation on the composition and functioning ofvarious marine plankton communities, which is ac-companied by some characteristic details on theirdynamicism.

General Features of a Plankton Community

The expression ‘plankton community’ implies thatsuch a community is located in a water column. Ithas a range of components (groups of organisms)that can be organized according to their size. Theyrange in size from tiny single-celled organisms suchas bacteria (0.4–1-mm diameter) to large predatorslike scyphomedusae of more than 1 m in diameter.A common method which has been in use fordecades is to group according to size, which here isattributed to the organism’s largest dimension; thusthe organisms range from picoplankton to macro-plankton (Figure 1). It is, however, the smallest di-mension of an organism which usually determineswhether it is retained by a mesh, since in a flow,elongated particles align themselves with the flow.

A plankton community is operating/functioningcontinuously, that is, physical, chemical, and bio-logical variables are always at work. Interactionsamong its components occur all the time. As onewell-known fluid dynamicist stated, ‘‘The surface ofthe ocean can be flat calm but below that surfacethere is always motion of the water at variousscales.’’ Many of the particles/organisms are movingor being moved most of the time: Those withoutflagella or appendages can do so due to processeswithin or due to external forcing, for example, fromwater motion due to internal waves; and those withflagella/cilia or appendages or muscles move or cre-ate motion of the water in order to exist. Orientedmotion is usually in the vertical which often resultsin distinct layers of certain organisms. However,physical variables also, such as light or density

differences of water masses, can result in layering ofplanktonic organisms. Such layers which are oftenhorizontally extended are usually referred to aspatches.

As stated in the definition, the components of aplankton community interact. It is usually the casethat a larger organism will ingest a smaller one or apart of it (Figure 1). However, there are exceptions.The driving force for a planktonic community origi-nates from sun energy, that is, primary productivity’s(1) direct and (2) indirect products: (1) autotrophs(phytoplankton cells) which can range from near 2 tomore than 300-mm width/diameter, or chemotrophs;and (2) dissolved organic matter, most of which isreleased by phytoplankton cells and protozoa asmetabolic end products, and being taken up by bac-teria and mixo- and heterotroph protozoa (Figure 1).These two components mainly set the microbial loop(ML; (see Bacterioplankton and Protozoa, PlanktonicForaminifera)) in motion; that is, unicellular organ-isms of different sizes and behaviors (auto-, mixo-, andheterotrophs) depend on each other – usually,but not always, the smaller being ingested by thelarger. Most of nutrients and energy are recirculatedwithin this subcommunity of unicellular organisms inall marine regions of our planet (see Bacterioplankton,Phytoplankton Size Structure, and Protozoa, Plank-tonic Foraminifera for more details, especially theML)These processes of the ML dominate the transferof energy in all plankton communities largely becausethe processes (rates of ingestion, growth, repro-duction) of unicellular heterotrophs almost alwaysoutpace those of phytoplankton, and also of meta-zooplankton taxa at most times.

The main question actually could be: ‘‘What is thecomposition of plankton communities, and how dothey function?’’ Figure 1 reveals sizes and relation-ships within a plankton community including theML. It shows the so-called ‘bottom-up’ and ‘top-down’ effects as well as indirect effects like theabove-mentioned labile dissolved organic matter(labile DOM), released by auto- and also by hetero-trophs, which not only drives bacterial growth butcan also be taken up or used by other protozoa.There can also be reversals, called two-way pro-cesses. At times a predator eating an adult metazoanwill be affected by the same metazoan which is ableto eat the predator’s early juveniles (e.g., well-grownctenophores capturing adult omnivorous copepodswhich have the ability to capture and ingest veryyoung ctenophores).

5

To comprehend the functioning of a planktoncommunity requires a quantitative assessment of theabundances and activities of its components. First,almost all of our knowledge to date stems from

in situ sampling, that is, making spot measurementsof the abundance and distribution of organisms inthe water column. The accurate determination ofabundance and distribution requires using meshes or

Dimensions(µm)

Macrozooplankton

2000

Mesoplankton

Microplankton

Nanoplankton

Picoplankton

200

20

2

0.2

Autotrophs Heterotrophs

Large copepods,salps, doliolids

ChaetognathsCtenophoresCoelenterates

}

Large diatomsDiatom chains}

}{

DiatomsDinoflagellates{

DiatomsDinoflagellates

Flagellates{

{

}

Dinoflagellates

Dinoflagellates

Ciliates

Ciliates

Nauplii

Flagellates

BacteriaProkaryotes

Labile DOM

{ }Copepodid stages andAdults of small copepodsNauplii of large copepods

Early zooids of salps and doliolids

Figure 1 Interactions within a plankton community separated into size classes of auto- and heterotrophs, including the microbial

loop; the arrows point to the respective grazer, or receiver of DOM; the figure is partly related to figure 9 from Landry MR and Kirchman

DL (2002) Microbial community structure and variability in the tropical Pacific. Deep-Sea Research II 49: 2669–2693.

6 MARINE PLANKTON COMMUNITIES

devices which quantitatively collect the respectiveorganisms. Because of methodological difficultiesand insufficient comprehension of organisms’ sizesand activities, quantitative sampling/quantificationof a community’s main components has been ofteninadequate. The following serves as an example ofthis. Despite our knowledge that copepods consistof 11 juvenile stages aside of adults, the majority ofstudies of marine zooplankton hardly considered thejuveniles’ significance and this manifested itself insampling with meshes which often collected merelythe adults quantitatively. Second, much knowledgeon rate processes comes from quantifying the re-spective organisms’ activities under controlled con-ditions in the laboratory. Some in situ measurements(e.g., of temperature, salinity, chlorophyll concen-trations, and acoustic recordings of zooplanktonsizes) have been achieved ‘continuously’ over time,resulting in time series of increases and decreases ofcertain major community components. To date thereare few, if any, direct in situ observations on theactivity scales of the respective organisms, frombacteria to proto- and to metazooplankton, mainlybecause of methodological difficulties. In essence,our present understanding of processes within plank-ton communities is incomplete.

Specific Plankton Communities

We will provide several examples of plankton com-munities of our oceans. They will include infor-mation about the main variables affecting them, theirmain components, partly their functioning over time,including particular specifics characterizing each ofthose communities.

In this section, plankton communities are pre-sented for three different types of marine environ-ments: estuaries/inshore, continental shelves, andopen ocean regions.

Estuaries

Estuaries and near-shore regions, being shallow, willrapidly take up and lose heat, that is, will be stronglyaffected by atmospheric changes in temperature,both short- and long-term, the latter showing in theseasonal extremes ranging from 2 to 32 1C in estu-aries of North Carolina. Runoff of fresh water,providing continuous nutrient input for primaryproduction, and tides contribute to rapid changes insalinity. This implies that resident planktonic taxaought to be eurytherm as well as – therm. Only veryfew metazooplanktonic species are able to exist insuch an environment (Table 1). In North Carolinian

estuaries, representative of other estuaries, they arethe copepod species Acartia tonsa, Oithona oculata,and Parvocalanus crassirostris. In estuaries of RhodeIsland, two species of the genus Acartia occur. Dur-ing colder temperatures Acartia hudsonica producesdormant eggs as temperatures increase and then isreplaced by A. tonsa, which produces dormant eggsonce temperatures again decrease later in the year.Such estuaries are known for high primary prod-uctivity, which is accompanied by high abundancesof heterotroph protozoa preying on phytoplankton.Such high abundances of unicellular organisms implythat food is hardly limiting the growth of the above-mentioned copepods which can graze on auto- aswell as heterotrophs. However, such estuaries areoften nursery grounds for juvenile fish like menhadenwhich prey heavily on late juveniles and adults ofsuch copepods, especially Acartia, which is not onlythe largest of those three dominant copepod speciesbut also moves the most, and thus can be seen mosteasily by those visual predators. This has resulted indiurnal migrations mostly of their adults, remainingat the seafloor during the day where they hardly eat,thus avoiding predation by such visual predators,and only entering the water column during darkhours. That then is their period of pronouncedfeeding. The other two species which are not heavilypreyed upon by juvenile fish, however, can be af-fected by the co-occurring Acartia, because fromearly copepodid stages on this genus can be stronglycarnivorous, readily preying on the nauplii of its ownand of those other species.

Nevertheless, the usually continuous abundanceof food organisms for all stages of the three co-pepod species results in high concentrations ofnauplii which in North Carolinian estuaries canreach 100 l�1, as can their combined copepodidstages. The former is an underestimate, becausesampling was done with a 75-mm mesh, which ispassed through by most of those nauplii. By com-parison, in an estuary on the west coast of Japan(Yellow Sea), dominated also by the genera Acartia,Oithona, and Paracalanus and sampling with 25-mm mesh, nauplius concentrations during summersurpassed 700 l�1, mostly from the genus Oithona.And copepodid stages plus adults repeatedly ex-ceeded 100 l�1. Here sampling with such narrowmesh ensured that even the smallest copepods werecollected quantitatively.

In essence, estuaries are known to attain among thehighest concentrations of proto- and metazoo-plankton. The known copepod species occur duringmost of the year, and are observed year after yearwhich implies persistence of those species beyonddecades.

MARINE PLANKTON COMMUNITIES 7

Table 1 Some characteristics of marine plankton communities

Estuaries Shelves Open ocean gyres

Subarctic Boreal Epipelagic subtropical

Pacific Atlantic

Atlantic/Pacific

Physical variables Wide range of

temperature and

salinity

Intermittent and seasonal

atmospheric forcing

Steady salinity, seasonal

temp. variability

Major seasonal variability

of temperature

Steady temperature and salinity,

continuous atmospheric forcing

Nutrient supply Continuous Episodic Seasonal Seasonal Occasional

Phytoplankton abundance High from spring to

autumn

Intermittently high Always low Major spring bloom Always low

Phytoplankton composition Flagellates, diatoms Flagellates, diatoms,

dinoflagellates

Nanoflagellates Spring: diatoms

Other: mostly

nanoplankton

Mostly prokaryotes, small nano- and

dinoflagellates

Primary Productivity High at most times Intermittently high Maximum in spring Max. in spring and

autumn

Always low

No. of metazoan species r5 B10–30 410 420 4100

Seasonal variability of

metazoan abundance

High spring and

summer, low winter

Highly variable High High Low

Copepod Ranges NaB10–500 l�1 o5–50 l�1 3–10 l�1

Abundance CopbB5–100 l�1 o3–30 l�1 Up to 1000 m�3 Up to 1000 m�3 300–1000 m�3

Neocalanus C. finmarchicus

Dominant metazoo-

plankton taxa

Acartia Oithona Neocalanus Calanus Oithona

Oithona Paracalanus Oithona Oithona Clausocalanus

Parvocalanus Temora Metridia Oncaea Oncaea

Doliolida

aNauplii.bCopepodids and adult copepods.

8M

AR

INE

PLA

NK

TO

NC

OM

MU

NIT

IES

Continental Shelves

By definition they extend to the 200-m isobath, andrange from narrow (few kilometers) to wide (morethan 100-km width). The latter are of interest be-cause the former are affected almost continuouslyand entirely by the nearby open ocean. Shelves areaffected by freshwater runoff and seasonally chan-ging physical variables. Water masses on continentalshelves are evaluated concerning their residence time,because atmospheric events sustained for more than1 week can replace most of the water residing on awide shelf with water offshore but less so from nearshore. This implies that plankton communities onwide continental shelves, which are often nearboundary currents, usually persist for limited periodsof time, from weeks to months (Table 1). They in-clude shelves like the Agulhas Bank, the CampecheBanks/Yucatan Shelf, the East China Sea Shelf, theEast Australian Shelf, and the US southeastern con-tinental shelf. There can be a continuous influx year-round of new water from adjacent boundary currentsas seen for the Yucatan Peninsula and Cape Canav-eral (Florida). The momentum of the boundary cur-rent (here the Yucatan Current and Florida Current)passing a protruding cape will partly displace wateralong downstream-positioned diverging isobathswhile the majority will follow the current’s generaldirection. This implies that upstream-producedplankton organisms can serve as seed populationstoward developing a plankton community on suchwide continental shelves.

Whereas estuarine plankton communities receivealmost continuously nutrients for primary pro-duction from runoff and pronounced benthic-pelagiccoupling, those on wide continental shelves in-frequently receive new nutrients. Thus they are atmost times a heterotroph community unless theyobtain nutrients from the benthos due to storms, orreceive episodically input of cool, nutrient-rich waterfrom greater depths of the nearby boundary currentas can be seen for the US SE shelf. Passing along theouter shelf at about weekly intervals are nutrient-richcold-core Gulf Stream eddies which contain plank-ton organisms from the highly productive Gulf ofMexico. Surface winds, displacing shelf surfacewater offshore, lead to an advance of the deep coolwater onto the shelf which can be flooded entirely byit. Pronounced irradiance and high-nutrient loads insuch upwellings result in phytoplankton bloomswhich then serve as a food source for protozoo- andmetazooplankton. Bacteria concentrations in suchcool water masses increase within several days by 1order of magnitude. Within 2–3 weeks most of thesmaller phytoplankton (c. o20-mm width) has been

greatly reduced, usually due to grazing by protozoaand relatively slow-growing assemblages of plank-tonic copepods of various genera such as Temora,Oithona, Paracalanus, Eucalanus, and Oncaea.However, quite frequently, the Florida Current whichbecomes the Gulf Stream carries small numbers ofThaliacea (Tunicata), which are known for inter-mittent and very fast asexual reproduction. Suchsalps and doliolids, due to their high reproductiveand growth rate, can colonize large water masses, thelatter increasing from B5 to 4500 zooids per cubicmeter within 2 weeks, and thus form huge patches,covering several thousands of square kilometers, asthe cool bottom water is displaced over much of theshelf. The increased abundance of salps (usually inthe warmer and particle-poor surface waters) anddoliolids (mainly in the deeper, cooler, particle-richwaters, also observed on the outer East China shelf)can control phytoplankton growth once they achievebloom concentrations. The development of suchlarge and dense patches is partly due to the lack ofpredators.

Although the mixing processes between the ini-tially quite cool intruding bottom (13–20 1C) and thewarm, upper mixed layer water (27–28 1C) are lim-ited, interactions across the thermocline occur, thuscreating a plankton community throughout thewater column of previously resident and newly ar-riving components. The warm upper mixed layeroften has an extraordinary abundance of earlycopepodid stages of the poecilostomatoid copepodOncaea, thanks to their ontogenetical migrationafter having been released by the adult femaleswhich occur exclusively in the cold intruding water.Also, early stages of the copepod Temora turbinataare abundant in the warm upper mixed layer; whileT. turbinata’s late juvenile stages prefer the cool layerbecause of the abundance of large, readily availablephytoplankton cells. As in estuaries, the copepodgenus Oithona flourishes on warm, temperate, andpolar continental shelves throughout most of theeuphotic zone.

Such wide subtropical shelves will usually be wellmixed during the cooler seasons, and then harbor,due to lower temperatures, fewer metazoplanktonspecies which are often those tolerant of wider orlower temperature ranges. Such wide shelves areusually found in subtropical regions, which explainsthe rapidity of the development of their planktoncommunities. They, however, are also found in coolerclimates, like the wide and productive Argentinian/Brazilian continental shelf about which our know-ledge is limited. Other large shelves, like the southernNorth Sea, have a limited exchange of water with theopen ocean but at the same time considerable influx

MARINE PLANKTON COMMUNITIES 9

of runoff, plus nutrient supply from the benthos dueto storm events, and thus can maintain identicalplankton communities over months and seasons.

In essence, continental shelf plankton communitiesare usually relatively short-lived, which is largely dueto their water’s limited residence time.

Open Ocean

The open ocean, even when not including oceanmargins (up to 1000-m water column), includes byfar the largest regions of the marine environment. Itsdeep-water columns range from the polar seas to theTropics. All these regions are under different at-mospheric and seasonal regimes, which affectplankton communities. Most of these communitiesare seasonally driven and have evolved along thephysical conditions characterizing each region. Thefocus here is on gyres as they represent specific oceancommunities whose physical environment can bereadily presented.

Gyres represent huge water masses extendinghorizontally over hundreds to even thousands ofkilometers in which the water moves cyclonically oranticyclonically. They are encountered in subpolar,temperate, and subtropical regions. The best-studiedones are:

• subpolar: Alaskan Gyre;

• temperate: Norwegian Sea Gyre, Labrador–Irminger Sea Gyre;

• subtropical: North Pacific Central Gyre (NPCG),North Atlantic Subtropical Gyre (NASG).

The Alaskan Gyre is part of the subarctic Pacific(Table 1) and is characterized physically by a shallowhalocline (B110-m depth) which prevents convectivemixing during storms. Biologically it is characterizedby a persistent low-standing stock of phytoplanktondespite high nutrient abundance, and several speciesof large copepods which have evolved to persist via alife cycle as shown for Neocalanus plumchrus. Bymidsummer, fifth copepodids (C5) in the upper 100 mwhich have accumulated large amounts of lipids beginto descend to greater depths of 250 m and beyondundergoing diapause, and eventually molt to femaleswhich soon begin to spawn. Spawning females arefound in abundance from August to January. Naupliiliving off their lipid reserves and moving upwardbegin to reach surface waters by mid- to late winter ascopepodid stage 1 (C1), and start feeding on theabundant small phytoplankton cells (probably pas-sively by using their second maxillae, but mostly byfeeding actively on heterotrophic protozoa which arethe main consumers of the tiny phytoplankton cells).The developing copepodid stages accumulate lipids

which in C5 can amount to as much as 50% of theirbody mass, which then serve as the energy source formetabolism of the females at depth, ovary develop-ment, and the nauplii’s metabolism plus growth.While the genus Neocalanus over much of the yearprovides the highest amount of zooplankton biomass,the cyclopoid Oithona is the most abundant meta-zooplankter; other abundant metazooplankton taxainclude Euphausia pacifica, and in the latter part ofthe year Metridia pacifica and Calanus pacificus.

In the temperate Atlantic (Table 1), the NorwegianSea Gyre maintains a planktonic community which ischaracterized, like much of the temperate oceanicNorth Atlantic, by the following physical features.Pronounced winds during winter mix the water col-umn to beyond 400-m depth, being followed by lesserwinds and surface warming resulting in stratificationand a spring bloom of mostly diatoms, and a weakautumn phytoplankton bloom. A major consumer ofthis phytoplankton bloom and characteristic of thisenvironment is the copepod Calanus finmarchicus,occurring all over the cool North Atlantic. This spe-cies takes advantage of the pronounced spring bloomafter emerging from diapause at 4400-m depth, bymoulting to adult, and grazing of females at highclearance rates on the diatoms, right away starting toreproduce and releasing up to more than 2000 fertil-ized eggs during their lifetime. Its nauplii start to feedas nauplius stage 3 (N3), being able to ingest diatomsof similar size as the adult females, and can reachcopepodid stage 5 (C5) within about 7 weeks in theNorwegian Sea, accumulating during that periodlarge amounts of lipids (wax ester) which serve as themain energy source for the overwintering diapauseperiod. Part of the success of C. finmarchicus is foundin its ability of being omnivorous. C5s either descendto greater depths and begin an extended diapauseperiod, or could moult to adult females, thus pro-ducing another generation which then initiates dia-pause at mostly C5. Its early to late copepodid stagesconstitute the main food for juvenile herring whichaccumulate the copepods’ lipids for subsequentoverwintering and reproduction. Of the other co-pepods, the genus Oithona together with the poeci-lostomatoid Oncaea and the calanoid Pseudocalanuswere the most abundant.

Subtropical and tropical parts of the oceans covermore than 50% of our oceans. Of these, the NPCG,positioned between c. 101 and 451N and movinganticyclonically, has been frequently studied. It in-cludes a southern and northern component, the latterbeing affected by the Kuroshio and westerly winds,the former by the North Equatorial Current and thetrade winds. Despite this, the NPCG has been con-sidered as an ecosystem as well as a huge plankton

10 MARINE PLANKTON COMMUNITIES

community. The NASG, found between c. 151 and401N and moving anticyclonically, is of similarhorizontal dimensions. There are close relations be-tween subtropical and tropical communities; for ex-ample, the Atlantic south of Bermuda is consideredclose to tropical conditions. Vertical mixing in bothgyres is limited. Here we focus on the epipelagiccommunity which ranges from the surface to about150-m depth, that is, the euphotic zone. The epipe-lagial is physically characterized by an upper mixedlayer of c. 15–40 m of higher temperature, belowwhich a thermocline with steadily decreasing tem-peratures extends to below 150-m depth. In thesetwo gyres, the concentrations of phytoplanktonhardly change throughout the year in the epipelagic(Table 1) and together with the heterotrophicprotozoa provide a low and quite steady food con-centration (Table 1) for higher trophic levels. Suchvery low particle abundances imply that almost allmetazooplankton taxa depending on them are livingon the edge, that is, are severely food-limited. Des-pite this fact, there are more than 100 copepodspecies registered in the epipelagial of each of thetwo gyres. How can that be? Almost all these co-pepod species are small and rather diverse in theirbehavior: the four most abundant genera have dif-ferent strategies to obtain food particles: the inter-mittently moving Oithona is found in the entireepipelagial and depends on moving food particles(hydrodynamic signals); Clausocalanus is mainlyfound in the upper 50 m of the epipelagial and al-ways moves at high speed, thus encountering nu-merous food particles, mainly via chemosensory;Oncaea copepodids and females occur in the lowerpart of the epipelagial and feed on aggregates; andthe feeding-current producing Calocalanus perceivesparticles via chemosensory. This implies that anycopepod species can persist in these gyres as long as itobtains sufficient food for growth and reproduction.This is possible because protozooplankton alwayscontrols the abundance of available food particles;thus, there is no competition for food among themetazooplankton. In addition, since total copepodabundance (quantitatively collected with a 63-mmmesh by three different teams) is steady and usu-ally o1000 m�3 including copepodid stages (pro-nounced patchiness of metazooplankton has not yetbeen observed in these oligotrophic waters), theprobability of encounter (only a minority of the zoo-plankton is carnivorous on metazooplankton) is verylow, and therefore the probability of predation lowwithin the metazooplankton. In summary, thesesteady conditions make it possible that in the epipe-lagial more than 100 copepod species can coexist, andare in a steady state throughout much of the year.

Conclusions

All epipelagic marine plankton communities are atmost times directly or indirectly controlled or affectedby the activity of the ML, that is, unicellular organ-isms. Most of the main metazooplankton species areadapted to the physical and biological conditions ofthe respective community, be it polar, subpolar, tem-perate, subtropical, or tropical. The only metazoo-plankton genus found in all communities mentionedabove, and also all other studied marine planktoncommunities, is the copepod genus Oithona. Thiscopepod has the ability to persist under adverse con-ditions, for example, as shown for the subarcticPacific. This genus can withstand the physical as wellas biological (predation) pressures of an estuary, thepersistent very low food levels in the warm openocean, and the varying conditions of the AntarcticOcean. Large copepods like the genus Neocalanus inthe subarctic Pacific, and C. finmarchicus in thetemperate to subarctic North Atlantic are adaptedwith respective distinct annual cycles in their re-spective communities. Among the abundant com-ponents of most marine plankton communities fromnear shore to the open ocean are appendicularia(Tunicata) and the predatory chaetognaths.

Our present knowledge of the composition andfunctioning of marine planktonic communities de-rives from (1) oceanographic sampling and timeseries, optimally accompanied by the quantificationof physical and chemical variables; and (2) labora-tory/onboard experimental observations, includingsome time series which provide results on small-scaleinteractions (microns to meters; milliseconds tohours) among components of the community. Opti-mally, direct in situ observations on small scales inconjunction with respective modeling would provideinsights in the true functioning of a plankton com-munity which operates continuously on scales ofmilliseconds and larger.

Our future efforts are aimed at developing in-strumentation to quantify in situ interactions of thevarious components of marine plankton com-munities. Together with traditional oceanographicmethods we would go ‘from small scales to the bigpicture’, implying the necessity of understanding thefunctioning on the individual scale for a com-prehensive understanding as to how communitiesoperate.

See also

Bacterioplankton. Copepods. GelatinousZooplankton. Phytoplankton Size Structure. PlanktonOverview. Protozoa, Planktonic Foraminifera.

MARINE PLANKTON COMMUNITIES 11

Further Reading

Atkinson LP, Lee TN, Blanton JO, and Paffenhofer G-A(1987) Summer upwelling on the southeasterncontinental shelf of the USA during 1981: Hydro-graphic observations. Progress in Oceanography 19:231--266.

Fulton RS, III (1984) Distribution and communitystructure of estuarine copepods. Estuaries 7: 38--50.

Hayward TL and McGowan JA (1979) Pattern andstructure in an oceanic zooplankton community.American Zoologist 19: 1045--1055.

Landry MR and Kirchman DL (2002) Microbialcommunity structure and variability in the tropicalPacific. Deep-Sea Research II 49: 2669--2693.

Longhurst AR (1998) Ecological Geography of the Sea,398pp. San Diego, CA: Academic Press.

Mackas DL and Tsuda A (1999) Mesozooplankton in theeastern and western subarctic Pacific: Communitystructure, seasonal life histories, and interannualvariability. Progress in Oceanography 43: 335--363.

Marine Zooplankton Colloquium 1 (1998) Future marinezooplankton research – a perspective. Marine EcologyProgress Series 55: 197--206.

Menzel DW (1993) Ocean Processes: US SoutheastContinental Shelf, 112pp. Washington, DC: US Depart-ment of Energy.

Miller CB (1993) Pelagic production processes in thesubarctic Pacific. Progress in Oceanography 32: 1--15.

Miller CB (2004) Biological Oceanography, 402pp.Boston: Blackwell.

Paffenhofer G-A and Mazzocchi MG (2003) Verticaldistribution of subtropical epiplanktonic copepods.Journal of Plankton Research 25: 1139--1156.

Paffenhofer G-A, Sherman BK, and Lee TN (1987)Summer upwelling on the southeastern continental shelfof the USA during 1981: Abundance, distribution andpatch formation of zooplankton. Progress in Ocean-ography 19: 403--436.

Paffenhofer G-A, Tzeng M, Hristov R, Smith CL, andMazzocchi MG (2003) Abundance and distribution ofnanoplankton in the epipelagic subtropical/tropicalopen Atlantic Ocean. Journal of Plankton Research 25:1535--1549.

Smetacek V, DeBaar HJW, Bathmann UV, Lochte K, andVan Der Loeff MMR (1997) Ecology and bio-geochemistry of the Antarctic Circumpolar Currentduring austral spring: A summary of Southern OceanJGOFS cruise ANT X/6 of RV Polarstern. Deep-SeaResearch II 44: 1--21 (and all articles in this issue).

Speirs DC, Gurney WSC, Heath MR, Horbelt W, WoodSN, and de Cuevas BA (2006) Ocean-scale modeling ofthe distribution, abundance, and seasonal dynamics ofthe copepod Calanus finmarchicus. Marine EcologyProgress Series 313: 173--192.

Tande KS and Miller CB (2000) Population dynamics ofCalanus in the North Atlantic: Results from the Trans-Atlantic Study of Calanus finmarchicus. ICES Journalof Marine Science 57: 1527 (entire issue).

Webber MK and Roff JC (1995) Annual structure of thecopepodcommunity and its associated pelagic environmentoff Discovery Bay, Jamaica. Marine Biology 123: 467--479.

12 MARINE PLANKTON COMMUNITIES

PLANKTON VIRUSES

J. Fuhrman, University of Southern California,Los Angeles, CA, USAI. Hewson, University of California Santa Cruz,Santa Cruz, CA, USA

& 2009 Elsevier Ltd. All rights reserved.

Introduction

Although they are the tiniest biological entities inthe sea, typically 20–200 nm in diameter, viruses areintegral components of marine planktonic systems.They are extremely abundant in the water column,typically 1010 per liter in the euphotic zone, and theyplay several roles in system function: (1) they areimportant agents in the mortality of prokaryotes andeukaryotes; (2) they act as catalysts of nutrient re-generation and recycling, through this mortality ofhost organisms; (3) because of their host specificityand density dependence, they tend to selectively at-tack the most abundant potential hosts, thus may‘kill the winner’ of competition and thereby fosterdiversity; and (4) they may also act as agents in theexchange of genetic material between organisms, acritical factor in evolution and also in relation to thespread of human-engineered genes. Although theseprocesses are only now becoming understood in anydetail, there is little doubt that viruses are significantplayers in aquatic and marine plankton.

History

It has only been in the past 25 years that micro-organisms like bacteria and small protists have beenconsidered ‘major players’ in planktonic food webs.The initial critical discovery, during the mid-1970s,was of high bacterial abundance as learned by epi-fluorescence microscopy of stained cells, with countstypically 109 l�1 in the plankton. These bacteria werethought to be heterotrophs (organisms that consumepreformed organic carbon), because they apparentlylacked photosynthetic pigments like chlorophyll(later it was learned that this was only partly right,as many in warm waters are in fact chlorophyll-containing prochlorophytes). With such high abun-dance, it became important to learn how fast theywere dividing, in order to quantify their function inthe food web. Growth rates were estimated primarilyby the development and application of methodsmeasuring bacterial DNA synthesis. The results of

these studies showed that bacterial doubling times intypical coastal waters are about 1 day. When thisdoubling time was applied to the high abundance, tocalculate how much carbon the bacteria are takingup each day, it became apparent that bacteria areconsuming a significant amount of dissolved organicmatter, typically at a carbon uptake rate equivalentto about half the total primary production. However,the bacterial abundance remains relatively constantover the long term, and they are too small to sinkout of the water column. Therefore, there must bemechanisms within the water to remove bacteria atrates similar to the bacterial production rate. In theinitial analysis, most scientists thought that grazingby protists was the only significant mechanismkeeping the bacterial abundance in check. This wasbecause heterotrophic protists that can eat bacteriaare extremely common, and laboratory experimentssuggested they are able to control bacteria at near-natural-abundance levels. However, some resultspointed to the possibility that protists are not theonly things controlling bacteria. In the late 1980s,careful review of multiple studies showed that graz-ing by protists was often not enough to balancebacterial production, and this pointed to the exist-ence of additional loss processes. About thatsame time, data began to accumulate that virusesmay also be important as a mechanism of removingbacteria. The evidence is now fairly clear that this isthe case, and it will be outlined below. This articlebriefly summarizes much of what is known abouthow viruses interact with marine microorganisms,including general properties, abundance, distribution,infection of bacteria, mortality rate comparisonswith protists, biogeochemical effects, effects on spe-cies compositions, and roles in genetic transfer andevolution.

General Properties

Viruses are small particles, usually about 20–200 nmlong, and consist of genetic material (DNA or RNA,single or double stranded) surrounded by a proteincoat (some have lipid as well). They have no me-tabolism of their own and function only via thecellular machinery of a host organism. As far as isknown, all cellular organisms appear to be suscep-tible to infection by some kind of virus. Culturestudies show that a given type of virus usually has arestricted host range, most often a single species orgenus, although some viruses infect only certain

13

subspecies and o0.5% may infect more than onegenus. Viruses have no motility of their own, andcontact the host cell by diffusion. They attach to thehost usually via some normally exposed cellularcomponent, such as a transport protein or flagellum.There are three basic kinds of virus reproduction(Figure 1). In lytic infection, the virus attaches to ahost cell and injects its nucleic acid. This nucleicacid (sometimes accompanied by proteins carried bythe virus) causes the host to produce numerousprogeny viruses, the cell then bursts, progeny arereleased, and the cycle begins again. In chronic in-fection, the progeny virus release is not lethal andthe host cell releases the viruses by extrusion orbudding over many generations. In lysogeny afterinjection, the viral genome becomes part of thegenome of the host cell and reproduces as geneticmaterial in the host cell line unless an ‘induction’event causes a switch to lytic infection. Inductionis typically caused by DNA damage, such as fromultraviolet (UV) light or chemical mutagens such asmitomycin C. Viruses may also be involved in killingcells by mechanisms that do not result in virusreproduction.

Observation of Marine Viruses

Viruses are so small that they are at or below theresolution limit of light microscopy (c. 0.1 mm).Therefore electron microscopy is the only way toobserve any detail of viruses. Sample preparationrequires concentrating the viruses from the wateronto an electron microscopy grid (coated with a thintransparent organic film). Because viruses are denserthan seawater, this can be done by ultracentrifuga-tion, typically at forces of at least 100 000� g for afew hours. It should be noted that under ordinarygravity, forces like drag and Brownian motionprevent viruses from sinking. To be observable theviruses must be made electron-dense, typically bystaining with uranium salts. The viruses are recog-nized by their size, shape, and staining properties(usually electron-dense hexagons or ovals, sometimeswith a tail), and counted. Typical counts are on theorder of 1010 viruses per liter in surface waters, withabundance patterns similar to those of heterotrophicbacteria (see below). Recently, it has been found thatviruses can also be stained with nucleic acid stainslike SYBR Green I, and observed and counted by

ChronicLytic

Lysogenic

Normal division

Induction

Normaldivision

Normal division continues unless induction occurs

Virus attachment to host

Nucleic acid (DNA or RNA) injected into hostViral nucleic acid injected

Host progeny continue to release viruses unless 'cured'Host lyses (bursts) to release progeny viruses

Host releases progeny viruses withoutlysing, by budding or extrusion

Host makes copies of viralnucleic acid and coat proteins

Host makes copies of viralnucleic acid and coat proteins

Viruses self-assemble inside host

Host with viral nucleic acid integrated intogenome or as extrachromosomal element

'Temperate' phage

Virus attachment to host

Figure 1 Virus life cycles. See text for explanation.

14 PLANKTON VIRUSES

epifluorescence microscopy. This is faster, easier, andless expensive than transmission electron microscopy(TEM). Epifluorescence viewing of viruses is shownin Figure 2, a micrograph of SYBR Green I-stainedbacteria and viruses, which dramatically illustratesthe high relative virus abundance. Epifluorescencemicroscopy of viruses is possible even though theviruses are below the resolution limit of light becausethe stained viruses are a source of light and appear asbright spots against a dark background (like starsvisible at night). Epifluorescence counts are similarto or even slightly higher than TEM counts fromseawater.

What Kinds of Viruses Occur inPlankton?

Microscopic observation shows the total, recogniz-able, virus community, but what kinds of virusesmake up this community, and what organisms arethey infecting? Most of the total virus community isthought to be made up of bacteriophages (virusesthat infect bacteria). This is because viruses lackmetabolism and have no means of actively movingfrom host to host (they depend on random diffusion),so the most common viruses would be expected toinfect the most common organism, and bacteria areby far the most abundant organisms in the plankton.Field studies show a strong correlation between viraland bacterial numbers, whereas the correlations be-tween viruses and chlorophyll are weaker. This

suggests that most viruses are bacteriophages ratherthan those infecting phytoplankton or other eu-karyotes. However, viruses infecting cyanobacteria(Synechococcus) are also quite common and some-times particularly abundant, exceeding 108 per liter insome cases. Even though most of the viruses probablyinfect prokaryotes, viruses for eukaryotic plankton arealso readily found. For example, those infecting thecommon eukaryotic picoplankter, Micromonaspusilla, are sometimes quite abundant, occasionallynear 108 per liter in coastal waters. Overall, the datasuggest that most viruses from seawater infect non-photosynthetic bacteria or archaea, but viruses in-fecting prokaryotic and eukaryotic phytoplanktonalso can make up a significant fraction of the total.

A great leap in our understanding of viral diversityhas occurred with the application of molecularbiological techniques. These involve the analysis ofvariability of DNA sequence of a single gene product(e.g., a capsid head protein, or an enzyme that makesDNA), or through a technique known as viriomics,whereby all genomes in a sample are cut up into smallpieces, then all the pieces sequenced and reassembled.These studies have revealed that: (1) there is a lot ofdiversity of closely related viruses infecting the sameor similar strains of microorganisms, a phenome-non known as ‘microdiversity’; (2) RNA-containingviruses comprise a small percentage of the viral mix inoceanic waters, and are similar to viruses that infectinsects and mollusks; and (3) that the diversity ofviruses in a liter of seawater is astonishingly high,estimated at 4104 different types for a sample fromcoastal California.

Virus Abundance

Total direct virus counts have been made in manyplanktonic environments – coastal, offshore, tempe-rate, polar, tropical, and deep sea. Typical virusabundance is 1–5� 1010 l�1 in rich near-shore sur-face waters, decreasing to about 0.1–1� 1010 l�1 inthe euphotic zone of offshore low-nutrient areas, andalso decreasing with depth, by about a factor of 10.A typical deep offshore profile is shown in Figure 3.Seasonal changes are also common, with virusesfollowing general changes in phytoplankton, bac-teria, etc.

Virus:prokaryote ratios also provide an interestingcomparison. In plankton, this ratio is typically 5–25,and commonly close to 10, even as abundance dropsto low levels in the deep sea. Why this ratio stays insuch a relatively narrow range is a mystery, but itdoes suggest a link and also tight regulatory mech-anisms between prokaryotes and viruses.

Figure 2 Epifluorescence micrograph of prokaryotes and

viruses from 16 km offshore of Los Angeles, stained with SYBR

Green I. The viruses are the very numerous tiny bright particles,

and the bacteria are the rarer larger particles. Bacterial size is

approximately 0.4–1mm in diameter.

PLANKTON VIRUSES 15

Viral Activities

Viruses have no physical activity of their own, so‘viral activity’ usually refers to lytic infection. How-ever, before discussing such infection, lysogeny andchronic infection are considered briefly. Lysogeny,where the viral genome resides in the host’s genome(Figure 1), is common. Lysogens (bacteria harboringintegrated viral genomes) can easily be found andisolated from seawater, and lysogeny, which is linkedto genetic transfer in a variety of bacteria, probablyimpacts microbial population dynamics and evo-lution. However, the induction rate appears to below and seasonally variable under ordinary naturalconditions, and lysogenic induction appears to be re-sponsible for only a tiny fraction of total virus pro-duction in marine systems for most of the year. On theother hand, at this time we simply do not know ifchronic infection is a significant process in naturalsystems. Release of filamentous (or other kinds ofbudding) viruses from native marine bacteria has notbeen noted in TEM studies of plankton, nor havesignificant numbers of free filamentous viruses (how-ever, filamentous viruses have been observed in mar-ine and freshwater sediments). But they could havebeen missed.

Regarding lytic infection, there are several studieswith a variety of approaches that all generally con-clude that viruses cause approximately 10–50% oftotal microbial mortality, depending on location,season, etc. These estimates are convincing, havingbeen determined in several independent ways.These include: (1) TEM observation of assembledviruses within host cells, representing the last stepbefore lysis; (2) measurement of viral decay rates;(3) measurement of viral DNA synthesis; (4) meas-urement of the disappearance rate of bacterial DNAin the absence of protists; (5) use of fluorescent virustracers to measure viral production and removalrates simultaneously; and (6) direct observation ofviral appearance in incubations where the viralabundance has been decreased several fold, yet hostabundances remain undiluted.

Comparison to Mortality from Protists

Because the earlier thinking was that protists are themain cause of bacterial mortality in marine plank-tonic systems, it is useful to ask how the contributionof viruses to bacterial mortality compares to that ofprotists. Multiple correlation analysis of abundances

Prokaryotes l−1D

epth

(m

)

108 109 10111010108107 109

Virus:prokaryote ratioViruses l−1

0

500

1000

1500

2000

3000

4000

2500

3500

0 5 10 15 20 25 30

Dep

th (

m)

0

500

1000

1500

2000

3000

4000

2500

3500

Dep

th (

m)

0

500

1000

1500

2000

3000

4000

2500

3500

Figure 3 Depth profile of total prokaryote (bacteria–archaea) counts, total viral counts, and virus:prokaryote ratios from the Coral

Sea (Apr. 1998), as determined by epifluorescence microscopy of SYBR Green-stained samples. Note the log scales of the counts.

16 PLANKTON VIRUSES

of bacteria, viruses, and flagellates showed virus-induced mortality of bacteria could occasionallyprevail over flagellate grazing, especially at highbacterial abundances. In more direct comparisons,measuring virus and protist rates by multiple in-dependent approaches, the total mortality typicallybalances production, and viruses are found to beresponsible for anything ranging from a negligibleproportion to the majority of total mortality.

To sum up these studies, the consensus is thatviruses are often responsible for a significant fractionof bacterial mortality in marine plankton, typicallyin the range 10–40%. Sometimes viruses may dom-inate bacterial mortality, and sometimes they mayhave little impact on it. It is unknown what controlsthis balance, but it probably includes variation inhost abundance, because when hosts are less com-mon, the viruses are more likely to be inactivatedbefore diffusing to a suitable host, as well as theexact types of bacteria that are present and theirpalatability. Application of new molecular techni-ques, based on ribosomal RNA sequences, andvariable regions within the host genomes have re-vealed that aquatic bacterial communities are typi-cally dominated by a handful of bacterial taxa, whilemost taxa make up a tiny proportion of cell num-bers. This would seem to support the notion thatmostly dominant taxa are targets of viral attack.

Roles in Food Web and GeochemicalCycles

The paradigm of marine food webs has been reviseda great deal in response to the initial discovery ofhigh bacterial abundance and productivity. It is nowwell established that a large fraction of the totalcarbon and nutrient flux in marine systems passesthrough the heterotrophic bacteria via the dissolvedorganic matter. How do viruses fit into this picture?Three features of viruses are particularly relevant:(1) small size; (2) composition; and (3) mode ofcausing cell death, which is to release cell contentsand progeny viruses to the surrounding seawater.

When a host cell lyses, the resultant viruses andcellular debris are made up of easily digested proteinand nucleic acid, plus all other cellular components, ina nonsinking form that is practically defined as dis-solved organic matter. This is composed of dissolvedmolecules (monomers, oligomers, and polymers),colloids, and cell fragments. This material is mostprobably utilized by bacteria as food. If it was abacterium that was lysed in the first place, then uptakeby other bacteria represents a partly closed loop,whereby bacterial biomass is consumed mostly by

other bacteria. Because of respiratory losses and in-organic nutrient regeneration connected with the useof dissolved organic substances, this loop has the neteffect of oxidizing organic matter and regeneratinginorganic nutrients (Figure 4). This bacterial–viralloop effectively ‘steals’ production from protists thatwould otherwise consume the bacteria, and segregatesthe biomass and activity into the dissolved and smal-lest particulate forms. The potentially large effect hasbeen modeled mathematically, and such models showthat significant mortality from viruses greatly increasesbacterial community growth and respiration rates.

Segregation of matter in viruses, bacteria, and dis-solved substances leads to better retention of nutrientsin the euphotic zone in virus-infected systems becausemore material remains in these small nonsinking forms.In contrast, reduced viral activity leads to more ma-terial in larger organisms that either sink themselves oras detritus, transporting carbon and inorganic nutrientsto depth. The impact can be particularly great forpotentially limiting nutrients like N, P, and Fe, whichare relatively concentrated in bacteria compared toeukaryotes. At least one study has demonstrated that incyanobacterial cultures that are starved for tracemetals, the Fe in viral lysate of another culture of equaldensity is taken up within an hour of supplement,which suggests a major role of viral lysis in the avail-ability of limiting nutrients. Therefore, the activity ofviruses has the possible effect of helping to supporthigher levels of biomass and productivity in theplanktonic system as a whole.

There are other potential geochemical effects ofviral infection and its resultant release of cell con-tents to the water, owing to the chemical and phys-ical nature of the released materials and the locationin the water column where the lysis occurs. For ex-ample, polymers released from lysed cells may fa-cilitate aggregation and sinking of material from theeuphotic zone. On the other hand, viral lysis ofmicroorganisms within sinking aggregates may leadto the breakup of the particles, converting somesinking particulate matter into nonsinking dissolvedmaterial and colloids at whatever depth the lysisoccurs. This contributes to the dissolution of sinkingorganic matter and its availability to free-livingbacteria in the ocean’s interior.

Viruses, particularly lysogens, have long beenknown to confer to hosts the ability to producetoxins – in fact cholera toxin is only produced by thecholera bacteria when lysogenized. Through study ofviriomics in marine plankton, it is now known thatviral genes may encode for several geochemicallyimportant enzymes including phosphorus uptake,and components of the photosynthetic apparatus.These appear to be used both under stable host–virus

PLANKTON VIRUSES 17

conditions, as well as during host replication, whenthe enhanced substrate utilization capabilities in-ferred to hosts is used to produce new virus particles.

Effects on Host Species Compositionsand Control of Blooms

Viruses mostly infect only one species or relatedspecies, and are also density dependent. Thus, themost common or dominant hosts in a mixed com-munity are believed to be most susceptible to in-fection, and rare ones least so. Lytic viruses canincrease only when the average time to diffuse fromhost to host is shorter than the average time that atleast one member from each burst remains in-fectious. Therefore, when a species or strain becomesmore abundant, it is more susceptible to infection.The end result is that viral infection works in op-position to competitive dominance. This may help tosolve Hutchinson’s ‘paradox of plankton’, whichasks us how so many different kinds of phyto-plankton coexist on only a few potentially limitingresources, when competition theory predicts one or afew competitive winners. Although there have beenseveral possible explanations for this paradox, viral

activity may also help solving it, because as statedabove, competitive dominants become particularlysusceptible to infection whereas rare species arerelatively protected. Extending this argument, onemight conclude that viruses have the potential tocontrol algal blooms, such as those consisting ofcoccolithophorids, and so-called ‘red tides’ of dino-flagellates. There is now evidence that at least undersome circumstances this may be true. Decliningblooms have been found to contain numerous in-fected cells.

Along similar lines, it is now commonly thoughtthat viral infection can influence the species com-position of diverse host communities even when theyare responsible for only a small portion of the hostmortality. This is again because of the near-speciesspecificity of viruses in contrast to the relativelyparticular tastes of protists or metazoa as grazers.This conclusion is supported by mathematical mod-els as well as limited experimental evidence.

Resistance

The development of host resistance to viral infectionis a common occurrence in laboratory and medical

Highertrophiclevels

Heterotrophicprokaryotes

'Futilecycle'

Viruses

Cellular debris

Grazers

InorganicnutrientsC, N, P, Fe, ...

Dissolved organicmatter

Phytoplankton

Figure 4 Prokaryote–viral loop within the microbial food web. Arrows represent transfer of matter.

18 PLANKTON VIRUSES

situations. Such resistance, where hosts mutate toresist the viral attack, is well known from nonmarineexperiments with highly simplified laboratory sys-tems. However, the existence of an apparently highinfection rate in plankton suggests that the rapiddevelopment resistance is not a dominant factor inthe plankton. How can the difference between la-boratory and field situations be explained?

Natural systems with many species and trophiclevels have far more interactions than simple la-boratory systems. One might expect that a specieswith a large fraction of mortality from one type ofvirus benefits from developing resistance. However,resistance is not always an overall advantage. It oftenleads to a competitive disadvantage from the loss ofsome important receptor, for example, involved insubstrate uptake. Even resistance to viral attach-ment, without any receptor loss, if that were pos-sible, would not necessarily be an advantage. For abacterium in a low-nutrient environment whosegrowth may be limited by N, P, or organic carbon,unsuccessful infection by a virus (e.g., stoppedintracellularly by a restriction enzyme, or with agenetic incompatibility) may be a useful nutritionalbenefit to the host organism, because the virus in-jection of DNA is a nutritious boost rich in C, N,and P. Even the viral protein coat, remaining out-side the host cell, is probably digestible by bacterialproteases. From this point of view, one might evenimagine bacteria using ‘decoy’ virus receptors to lureviral strains that cannot successfully infect them.With the proper virus and host distributions, theodds could be in favor of the bacteria, and if aninfectious virus (i.e., with a protected restriction site)occasionally gets through, the cell line as a wholemay still benefit from this strategy.

There are other reasons why resistance mightnot be an overall advantage. As described earlier,model results show that the heterotrophic bacteriaas a group benefit substantially from viral infection,raising their production by taking carbon and energyaway from larger organisms. Viruses also raise theoverall system biomass and production by helping tokeep nutrients in the lighted surface waters. How-ever, these arguments would require invoking somesort of group selection theory to explain how indi-viduals would benefit from not developing resistance(i.e., why not ‘cheat’ by developing resistance andletting all the other organisms give the group benefitsof infection?). In any case, evidence suggests thateven if resistance of native communities to viral in-fection may be common, it is not a dominant force,because there is continued ubiquitous existence ofviruses at roughly 10 times greater abundance thanbacteria and with turnover times on the order of a

day (as discussed above). Basic mass balance calcu-lations show that significant numbers of hosts mustbe infected and releasing viruses all the time. Forexample, with a typical lytic burst size of 50 andviral turnover time of 1 day, maintenance of a 10-fold excess of viruses over bacteria requires 20% ofthe bacteria to lyse daily. The lack of comprehensiveresistance might be due to frequent development ofnew virulent strains, rapid dynamics or patchiness inspecies compositions, or to a stable coexistence ofviruses and their hosts. All these are possible, andthey are not mutually exclusive.

Lysogeny, which is common in marine bacter-ioplankton, also conveys resistance to superinfection(i.e., infection by the same, or similar virus). Thismay be an advantage so long as the prophage re-mains uninduced, in that viral genomes often alsocontain useful genes involved in membrane trans-port, photosynthesis, etc., that benefit the host.However, because the prophage is carried around inthe host cell, this may also place an extra burden onthe host cell machinery (i.e., it must also replicate theviral genome in addition to the host genome).

Genetic Transfer

Viruses can also play central roles in genetic transferbetween microorganisms, through two processes. Inan indirect mechanism, viruses mediate genetictransfer by causing the release of DNA from lysedhost cells that may be taken up and used as geneticmaterial by another microorganism. This latter pro-cess is called transformation. A more direct process isknown as transduction, where viruses package someof the host’s own DNA into the phage head and theninject it into another potential host. Transduction inaquatic environments has been shown to occur ina few experiments. Although transduction usuallyoccurs within a restricted host range, recent dataindicate that some marine bacteria and phages arecapable of transfer across a wide host range. Al-though the extent of these mechanisms in naturalsystems is currently unknown, they could have im-portant roles in population genetics, by homogen-izing genes within a potential host population, andalso on evolution at relatively long timescales. Genetransfer across species lines is an integral componentof microbial evolution, as shown in the genomes ofmodern-day microbes that contain numerous genesthat have obviously been transferred from otherspecies. On shorter timescales, this process can beresponsible for the dissemination of genes thatmay code for novel properties, whether introducedto native communities naturally or via geneticengineering.

PLANKTON VIRUSES 19

Summary

It is now known that viruses can exert significantcontrol of marine microbial systems. A major effectis on mortality of bacteria and phytoplankton, whereviruses are thought to stimulate bacterial activity atthe expense of larger organisms. This also stimulatesthe entire system via improved retention of nutrientsin the euphotic zone. Other important roles includeinfluence on species compositions and possibly alsogenetic transfer.

See also

Bacterioplankton. Primary Production Distribution.Primary Production Methods. Primary ProductionProcesses.

Further Reading

Ackermann HW and Du Bow MS (1987) Viruses ofProkaryotes, Vol. I: General Properties of Bacterio-phages. Boca Raton, FL: CRC Press.

Azam F, Fenchel T, Gray JG, Meyer-Reil LA, andThingstad T (1983) The ecological role of water-columnmicrobes in the sea. Marine Ecology Progress Series 10:257--263.

Bratbak G, Thingstad F, and Heldal M (1994) Viruses andthe microbial loop. Microbial Ecology 28: 209--221.

Breitbart M, Salamon P, Andresen B, et al. (2002) Genomicanalysis of uncultured marine viral communities.Proceedings of the National Academy of Sciences of theUnited States of America 99: 14250--14255.

Culley A, Lang AS, and Suttle CA (2003) High diversity ofunknown picorna-like viruses in the sea. Nature 424:1054--1057.

Fuhrman JA (1992) Bacterioplankton roles in cycling oforganic matter: The microbial food web. In: Falkowski

PG and Woodhead AS (eds.) Primary Productivityand Biogeochemical Cycles in the Sea, pp. 361--383.New York: Plenum.

Fuhrman JA (1999) Marine viruses: Biogeochemical andecological effects. Nature 399: 541--548.

Fuhrman JA (2000) Impact of viruses on bacterialprocesses. In: Kirchman DL (ed.) Microbial Ecology ofthe Oceans, pp. 327--350. New York: Wiley-Liss.

Fuhrman JA and Suttle CA (1993) Viruses in marineplanktonic systems. Oceanography 6: 51--63.

Lindell D, Sullivan MB, Johnson ZI, Tolonen AC, RohwerF, and Chisholm SW (2004) Transfer of photosynthesisgenes to and from Prochlorococcus viruses. Proceedingsof the National Academy of Sciences of the UnitedStates of America 101: 11013--11018.

Maranger R and Bird DF (1995) Viral abundance in aquaticsystems – a comparison between marine and fresh waters.Marine Ecology Progress Series 121: 217--226.

Noble RT and Fuhrman JA (1998) Use of SYBR Green I forrapid epifluorescence counts of marine viruses andbacteria. Aquatic Microbial Ecology 14(2): 113--118.

Paul JH, Rose JB, and Jiang SC (1997) Coliphage andindigenous phage in Mamala Bay, Oahu, Hawaii. Appliedand Environmental Microbiology 63(1): 133--138.

Proctor LM (1997) Advances in the study of marine viruses.Microscopy Research and Technique 37(2): 136--161.

Suttle CA (1994) The significance of viruses to mortality inaquatic microbial communities. Microbial Ecology 28:237--243.

Thingstad TF, Heldal M, Bratbak G, and Dundas I (1993)Are viruses important partners in pelagic food webs?Trends in Ecology and Evolution 8(6): 209--213.

Wilhelm SW and Suttle CA (1999) Viruses and nutrientcycles in the sea – viruses play critical roles in thestructure and function of aquatic food webs. Bioscience49(10): 781--788.

Wommack KE and Colwell RR (2000) Virioplankton:Viruses in aquatic ecosystems. Microbiology andMolecular Biology Reviews 64(1): 69--114.

20 PLANKTON VIRUSES

BACTERIOPLANKTON

H. W. Ducklow, The College of Willian and Mary,Gloucester Point, VA, USA

Copyright & 2001 Elsevier Ltd.

Introduction

Marine bacteria, unicellular prokaryotic planktonusually less than 0.5–1 mm in their longest dimension,are the smallest autonomous organisms in the sea –or perhaps in the biosphere. The nature of their rolesin marine food webs and the difficulty of studyingthem both stem from their small size. A modernparadigm for bacterioplankton ecology was inte-grated into oceanography only following develop-ment of modern epifluorescence microscopy and theapplication of new radioisotopic tracer techniques inthe late 1970s. It was not until a decade later, withthe use of modern genomic techniques, that theiridentity and taxonomy began to be understood at all.Thus we are still in the process of constructing arealistic picture of marine bacterial ecology, consist-ent with knowledge of evolution, plankton dy-namics, food web theory, and biogeochemistry. Thelack of bacterioplankton compartments in most nu-merical models of plankton ecology testifies to outcurrent level of ignorance. Nevertheless, much isnow well known that was just beginning to beguessed in the 1980s.

Bacterioplankton are important in marine foodwebs and biogeochemical cycles because they are theprincipal agents of dissolved organic matter (DOM)utilization and oxidation in the sea. All organismsliberate DOM through a variety of physiologicalprocesses, and additional DOM is released whenzooplankton fecal pellets and other forms of organicdetritus dissolve and decay. By recovering the re-leased DOM, which would otherwise accumulate,bacterioplankton initiate the microbial loop, a com-plicated suite of organisms and processes based onthe flow of detrital-based energy through the foodweb. The flows of energy and materials through themicrobial loop can rival or surpass those flowspassing through traditional phytoplankton-grazer-based food chains. For further information on thetopics summarized here, the reader may consult theFurther Reading, especially the recent book edited byKirchman.

Identity and Taxonomy

Most bacterial species cannot be cultivated in thelaboratory and, until the development of culture-independent genomic methods, the identity of over90% of bacterial cells enumerated under the micro-scope was unknown. Only those few cells capable offorming colonies on solid media (agar plates) couldbe identified by classical bacteriological techniques.However, since the application of molecular genomicmethods to sea water samples in the mid-1980s, ourunderstanding of marine bacterial systematics andevolution has undergone a profound revolution. Inthis approach, plankton samples including bacter-ioplankton cells are collected and lysed to yield amixture of DNA strands reflecting the genetic com-position of the original assemblage. Then individualgenes on the DNA molecules can be cloned andamplified via the polymerase chain reaction (PCR)for further analysis. Theoretically, any gene complexcan be cloned, and several major groups of geneshave been studied to date – for example, genes con-trolling specific biogeochemical transformations likeammonium oxidation, nitrogen fixation, sulfate re-duction, and even oxidation of xenobiotic pollutantmolecules. The most useful and widely studied genesfor elucidating evolutionary relationships amongbacterioplankton have been the genes coding forsmall subunit ribosomal RNA (SSU rRNA), becausethey evolve relatively slowly and their charactershave been conserved across all life forms during thecourse of evolution. By sequencing the base pairsmaking up individual SSU rRNA molecules, thesimilarity of different genes can be established withgreat sensitivity. To date, nearly 1000 individualmicrobial SSU rRNA genes have been cloned andsequenced, yielding an entirely new picture of thecomposition of marine communities.

The most important aspect of our understanding isthat what we term ‘bacterioplankton’ really consistsof two of the fundamental domains of life: the Bac-teria and the Archaea (Figure 1). Domain Archaea isa group of microbial organisms with unique genetic,ultrastructural, and physiological characters that areabout as different, genetically, from the Bacteria aseither group is from higher life forms. Members ofthe Archaea may be typified by organisms from ex-treme habitats including anaerobic environments,hot springs, and salt lakes, but marine archaealgroups I and II are common in sea water. They makeup about 10% of the microbial plankton in the

21

surface waters of the oceans, and are relatively morenumerous at greater depths, where they approachabout half the total abundance. Since most of theseorganisms are known only from their RNA genes

and have never been cultured, their physiology androles in the plankton are almost entirely unknown.

Domain Bacteria contains all the familiar, cultur-able eubacterial groups and also a large number of

Cytophaga group

CyanobacteriaBacteria

0.10 substitutionsper nucleotide position

Pro

teo

bac

teri

a

Sulfolobus solfataricus

Desulfurococcus mobilis

Pyrodictium occultum

Thermoproteus tenax

Thermofilum

pendens

Archaea

Group I

Archaea

Met

hano

sarc

ina

bark

eri

Met

hano

spiri

llum

hun

gate

i

Hal

ofer

ax v

olca

nii

Hal

obac

teriu

m h

alob

ium

Arc

haea

Gro

up II

Ther

mop

lasm

a oc

idop

hilum

Thermococcus celer

Methanobacteriu

m form

icicum

Methan

obac

terium

ther

moauto

troph

icum

Methanococcus voltae

Methanococcus jannaschii

Aquifex pyrophilus

Thermotoga maritima

Geotoga subterranea

Herpetosiphon aurantiacus

Chloroflexus aurantiacus

Thermomicrobium roseum

SAR202

Nitrospira marinaRubrobacter xylanophilus

Marine ActinobacteriaCandidatus "

"

Microthrix parvicella

Atopobium minutumIsosphaera pallida

Pirellula staleyi

Planctomyces limnophilus

Verrucomicrobium spinosum

Chlamydophila

pneumoniae

Deinoc

occu

s rad

iodur

ans

Fibr

obac

ter s

uccin

ogen

es

Glo

eoba

cter

vio

lace

us

Syn

echo

cocc

usP

CC

6301

Mar

ine

Pic

op

hyto

pla

nkt

on

Bac

illus

sub

tilis

Lact

ococ

cus

lact

is

SAR

116

Azo

spiri

llum

bra

sile

nse

Mag

neto

spiri

llum

mag

neto

tact

icum

Erythr

obac

ter lo

ngus

Roseobacter/ SAR83

Rhodobacter capsulatusRhodovibrio salinarumSAR11

Alteromonas macleodiiOceanospirillum linum

SAR86Bathymodiolus thermophilus gill sym.

Rhodoferax fermentans

Methylophilus methylotrophus

Geobacter metallireducens

Bdellovibrio bacteriovorus

Marine G

roup B/SAR324

Bacteroides fragilis

Cellulophaga lyticaC

hlorobium vibrioform

eM

arine G

rou

p A

/SA

R406

Figure 1 Dendrogram showing relationships among the most widespread SSU rRNA gene clusters among the marine prokaryotes

(the ‘bacterioplankton’). (Modified after Giovannoni SJ, in Kirchman (2000).)

22 BACTERIOPLANKTON

unculturable, previously unknown groups. The mainculturable groups include members of the Proteo-bacteria, marine oxygenic, phototrophic Cyano-bacteria, and several other major groups includingmethylotrophs, planctomycetes, and the Cytophaga–Flavobacterium–Bacteroides group. But the mostabundant genes recovered so far are not similar tothose of the known culturable species. These includethe most ubiquitous of all groups yet recovered, theSAR-11 cluster of the alpha Proteobacteria, whichhave been recovered from every bacterial clone li-brary yet isolated. It appears to be the most widelydistributed and successful of the Bacteria. Thephotosynthetic Cyanobacteria, including Syneccho-coccus spp. and the unicellular prochlorophytes, arefunctionally phytoplankton and they dominate theprimary producer populations in the open sea, and attimes in coastal and even estuarine regimes. They aretreated elsewhere in this encyclopedia, so our dis-cussion here is limited to heterotrophic forms ofBacteria and to the planktonic Archaea, although wecannot specify what many (or most) of them do.Genomic techniques are now being used to investi-gate bacterial and archaeal species succession duringoceanographic events over various timescales, muchas phytoplankton and higher organism successionshave been observed for a century or more.

Nutrition and Physiology

Knowledge of the nutrition and physiology of nat-urally occurring bacterioplankton as a functionalgroup in the sea is based partly on laboratory studyof individual species in pure culture, but mostly onsea water culture experiments. Traditional labora-tory investigations show that bacteria can only util-ize small-molecular-weight compounds less thanB500 Daltons. Larger polymeric substances andparticles must first be hydrolyzed by extracellularenzymes. In the sea water culture approach, sampleswith natural bacterioplankton assemblages are in-cubated for suitable periods (usually hours to a fewdays) while bacterial abundance is monitored, theutilization of various compounds with 14C- or 3H-labelled radiotracers is estimated, and the net pro-duction or loss of metabolites like oxygen, CO2, andinorganic nutrients is measured. Such experiments,combined with size-fractionation using polycarbon-ate filters with precise and uniform pores of variousdiameter (0.2–10 mm), revealed that over 90% ofadded organic radiotracers are utilized by the smal-lest size fractions (o1 mm). Bacteria are over-whelmingly the sink for DOM in all habitats studiedto date. Nutrient limitation of bacterial growth can

be identified by adding various compounds (e.g.,ammonium, phosphate, or iron salts; monosacchar-ides and amino acids) singly or in combination toexperimental treatments and comparing growth re-sponses to controls. Using this approach, it has beenlearned that bacteria are effective competitors withphytoplankton for inorganic nutrients, includingiron, which bacteria can mobilize by producing iron-binding organic complexes called siderophores. Ingeneral, bacterial growth in the sea, from estuaries tothe central gyres, tends to be limited by organicmatter. Sea water cultures most often respond toadditions of sugars and amino acids, with the re-sponse sometimes enhanced if inorganic nutrients(including iron) are also added.

At larger scales, the ultimate dependence of bac-teria on organic matter supply is indicated by sig-nificant correlations between bacterial standingstocks or production (see below) and primary pro-duction (PP) across habitats (Figure 2). At within-habitat scales and shorter timescales, significant re-lationships are less common, indicating time lagsbetween organic matter production and its con-version by bacteria. Such uncoupling of organicmatter production and consumption is also shown bytransient accumulations of DOM in the upper ocean,where production processes tend to exceed utiliza-tion. It is not yet understood why DOM accumu-lates. Some fraction might be inherently refractory orrendered so by ultraviolet radiation or chemicalcondensation reactions in sea water. Deep ocean

100 1000 100001

10

100

1000

Primary production (mg C m d )_2 _1

Bac

teria

l pro

duct

ion

(mg

C m

d)

_2

_ 1

Figure 2 Bacterial production plotted against primary

production for the euphotic zone in several major ocean

regimes or provinces. The overall data set has a significant

regression, but the individual regions do not. J, Sargasso Sea;

K, Arabian Sea; B, equatorial Pacific; n, equatorial Pacific.

BACTERIOPLANKTON 23

DOM has a turnover time of centuries to millennia,and seems to become labile (vulnerable to bacterialattack) when the ocean thermohaline circulation re-turns it to the illuminated surface layer. Alternatively,bacterial utilization of marine DOM, which gener-ally has a high C:N ratio, might be limited byavailability of inorganic nutrients. The latter hy-pothesis is supported by observations that DOMaccumulation tends to be greater in the tropics andsubtropics, where nitrate and phosphate are depletedin surface waters.

The efficiency with which bacteria convert organicmatter (usually expressed in carbon units) into bio-mass can be estimated by comparing the apparentutilization of individual compounds or bulk DOMwith increases in biomass or with respiration. Res-piration is usually measured by oxygen utilizationbut precise new analytical techniques for measuringcarbon dioxide make CO2 production a preferableapproach. Bacterial respiration (BR) is difficult tomeasure because water samples must first be passedthrough filters to remove other, larger respiring or-ganisms, and because the resulting respiration ratesare low, near the limits of detection of oxygen andCO2 analyses. It is also not easy to estimate bacterialbiomass precisely (see below). The conversion effi-ciency or bacterial growth efficiency (BGE) is thequotient of net bacterial production (BP) and theDOM utilization:

BGE ¼ BP

DDOM¼ BP

BPþ BR½1�

Bacteria have rather uniform biomass C:N com-position ratios of 4–6. Intuitively, it seems reasonableto expect that they would utilize substrates with highC:N ratios at lower efficiency. Enrichment culturesinitiated from natural bacterial assemblages grow insea water culture in the laboratory on added sub-stances with efficiencies of 30–90%. The BGE is in-versely related to the C:N ratio of the organicsubstrate if just a single compound is being utilized,but when a mixture of compounds is present, as isprobably always the case in the environment, there isno discernible relationship between the chemicalcomposition of the materials being used and theBGE.

In the open ocean, BGE averages about 10–30%, arelatively low value that has important implicationsfor our understanding and modeling of organicmatter turnover and ocean metabolism. At largerscales, BGE appears to increase from B10% to 50%along an offshore-to-onshore gradient of increasingprimary productivity, probably reflecting greater or-ganic matter availability. This pattern has been used

to support an argument suggesting that in lake andoceanic systems with the lowest primary product-ivity, respiration exceeds production; that is, sucholigotrophic systems might be net heterotrophic.This possibility has also been supported by resultsfrom careful light–dark bottle studies in which oxy-gen consumption exceeds production. This finding,however, is inconsistent with a large amount ofgeochemical evidence, for instance showing netoxygen production at the basin and seasonal to an-nual scale. Resolution of this debate probably restson improved estimates of BGE.

Pure culture, sea water culture, and the latestgenomic studies indicate fundamental metabolic andgenetic differences among different bacterial popu-lations, which can generally be grouped into twobroad classes based on organic matter utilization.Native marine bacteria capable of utilizing DOM atconcentrations below 100 nmol l�1, termed oligo-trophs, cannot survive when DOM is greater thanabout 0.1–1 mmol l�1. Copiotrophic bacteria foundin some habitats with higher ambient DOM levelsthrive on concentrations far exceeding this threshold.Observations that copiotrophs shrink and have im-pressive survival capability under severe starvationconditions (thousands of days to, apparently, cen-turies) led some investigators to suggest that thedominant native marine bacteria are starving (non-growing) copiotrophs in a survival mode, awaitingepisodes of nutrient enrichment. A variable fractionof the total population usually does appear to bedormant, as indicated by autoradiography, vitalstaining, and RNA probes, but the timescales of thetransition from active growth to dormancy and backagain are not well defined. Maintenance of dormantcells in a population depends on strong predatorpreferences for actively growing cells and prey se-lection against the nongrowing cells. Most oligo-trophs so far isolated in the laboratory understringent low-DOM conditions appear to be un-related to known bacterial groups.

Bacterial Biomass, Growth, andProduction

The standing stock of bacteria is still most commonlyassessed by epifluorescence microscopy, followingstaining of the cells with a fluorochrome dye. Flowcytometric determination is gradually taking over,and has several key advantages over microscopy:faster sample processing, improved precision, anddiscrimination of heterotrophic and phototrophicbacteria. There is a gradient in bacterial abundanceproceeding from B1010 cells l�1 in estuaries to 109

24 BACTERIOPLANKTON

cells l�1 in productive ocean regimes and 108 cells l�1

in the oligotrophic gyres (Figure 3). These horizontalgradients parallel gradients in primary productionand organic matter fluxes, suggesting the overallimportance of bottom-up controls on bacterialabundance. Chlorophyll a concentrations, indicativeof phytoplankton biomass, vary somewhat morewidely than bacterial abundance over basin to globalscales, but within habitats, the variability of bacterialand phytoplankton biomass is about equal, reflectingthe generally close coupling between the two groupsand the similarity of removal processes (grazing,viral lysis, unspecified mortality) acting on them.

It is more difficult to estimate bacterial biomass,because we cannot measure the mass (e.g., as carbon)directly, and have to convert estimates of cell vol-umes to carbon instead. The best estimates now in-dicate 7–15� 10�15 g C cell�1 for oceanic cells and15–25� 10�15 g C cell�1 for the slightly larger cellsfound in coastal and estuarine habitats. Thus thebiomass gradient is steeper than the abundance gra-dient because the cells are larger inshore. Table 1shows data compiled from Chesapeake Bay and theSargasso Sea off Bermuda, two well-studied sites thatillustrate the contrasts in phytoplankton and bac-terioplankton from a nutrient-rich estuary to theoligotrophic ocean gyres. Bacterial and phyto-plankton biomass are much greater in the estuary, as

expected. Interestingly, assuming a mean euphoticzone depth of 1 m in the Bay and 140 m off Bermuda,we find that the standing stocks of bacteria in theseeuphotic zones are B10 and 50 mmol C m�2 in theestuaries and open sea, respectively. The oceaniceuphotic zone is somewhat more enriched in bacteriathan the more productive estuaries. Bacterial andphytoplankton stocks are nearly equal in the opensea, but phytoplankton exceeds bacterial biomassinshore. Carbon from primary producers appears tobe more efficiently stored in bacteria in oceanic sys-tems compared to estuarine ones.

Bacterial stocks in different environments can beassessed using the relationship

Bmax ¼ F=m ½2�

where Bmax is the carrying capacity in the absence ofremoval, F is the flux of utilizable organic matter tothe bacteria, and m is their maintenance efficiency(the specific rate of utilization when all of F is used tomeet cellular maintenance costs, with nothing left forgrowth). The problem is specifying values for F andm. The DOM flux can be evaluated by flow analysisand is about 20–50% of the net primary production(NPP) in most systems. Maintenance costs are poorlyconstrained and possibly very low if most cells arenear a starvation state, but 0.01 d�1 is a reasonablevalue for actively growing cells. Thus for the oligo-trophic gyres where the latest NPP estimates areabout 200–400 mg C m�2 d�1, we can calculate thatBmax should be about 4–8� 109 cells l�1, an order ofmagnitude greater than observed. Removal processesmust maintain bacterial stocks considerably belowtheir maximum carrying capacity.

Bacteria convert preformed organic matter intobiomass. This process is bacterial production, whichcan be expressed as the product of the biomass andthe specific growth rate (m)

BP ¼ dB=dt ¼ mB ½3�

Like biomass, BP cannot be measured directly inmass units. Instead, metabolic processes closelycoupled to growth are measured and BP is derivedusing conversion factors. The two most commonmethods follow DNA and protein synthesis using(3H)thymidine and (3H)leucine incorporation rates,respectively. The values for the conversion factors arepoorly constrained and hard to measure, leading touncertainty of at least a factor of two in the BP es-timates. Few measurements were performed in theopen sea before the 1990s. The Joint Global OceanFlux Study (JGOFS) time-series station at Bermuda isperhaps the best-studied site in the ocean (Table 1).

Hawaii

Sarga

sso

Ross S

ea

Gulf S

tream

Equat

orial

Pac

N. Atla

ntic

Chesa

peak

e Bay

66 94 136 30 37 27 16210

8

109

1010

Cel

ls l_ 1

Figure 3 Bacterial abundance in the euphotic zone of several

major ocean provinces. The box plots show the median, 10th,

25th, 75th and 90th centiles of the data. The number of samples

is listed for each region. There is no statistical difference among

the regions except for Chesapeake Bay.

BACTERIOPLANKTON 25

In the open sea, far removed from allochthonousinputs of organic matter, we can compare BP and PPdirectly, since all the organic matter ultimately de-rives from the PP. One difficulty is that BP itself is notconstrained by PP, since if the recycling efficiency ofDOM and the BGE are sufficiently high, BP canexceed PP. BP also commonly exceeds local PP inestuaries, where inputs of terrestrial organic matterare consumed by bacteria. Bacterial respiration,however, cannot exceed the organic matter supplyand serves as an absolute constraint on estimates ofBP. But as noted above, bacterial respiration is veryhard to measure and there are many fewer reliablemeasurements than for BP itself. BR is usually esti-mated from the BGE. Rearranging eqn [1],

BR ¼ ð1� BGEÞBP

BGE½4�

Most commonly, variations of eqn [1] have beenused to estimate the total bacterial carbon utilizationor demand (BCDBRþBP) from estimates or as-sumptions about BP and BGE. Earlier estimates andliterature surveys suggested that BP was as high as30% of PP. Combining this value with a BGE of 20%yields a BCD of 1.5 times the PP. This estimate initself is possibly acceptable, if recycling of DOM ishigh, but then eqn [4] yields a BR of 1.2 times thetotal PP – an impossibility. More recent estimates ofBP, typified by the Sargasso Sea data, suggest BP isabout 10% of PP in the open sea. Applying this valueand the mean BGE for the region (0.14), we find thatBR consumes about 55% of the primary productionin the Sargasso Sea, still a substantial figure. Similarcalculations for other well-studied ocean areas sug-gest that zooplankton (including protozoans andmicrozooplankton) and bacteria consume nearlyequal amounts of the total primary productivity.These estimates illustrate the biogeochemical im-portance of bacterioplankton in the ocean carboncycle: although their growth efficiency is low,

bacteria process large amounts of DOM. DOMproduced by a myriad of ecological and physio-logical processes must escape bacterial metabolismto enter long-term storage in the oceanic reservoir.

Role in Food Webs andBiogeochemical Cycles

The process of bacterivory (consumption of bacteriaby bacteriovores) completes the microbial loop.Bacterioplankton cells are ingested by a great diver-sity of predators, but, because of the small size of theprey, most bacteriovores are small protozoans, typi-cally o5 mm nanoflagellates and small ciliates. Bac-terial cells only occupy about 10�7 of the volume ofthe upper ocean, indicating the difficulty of en-countering these small prey. Larger flagellates, smallciliates, and some specialized larger predators canalso ingest bacterial prey. The most important of thelarger predators are gelatinous zooplankton likelarvaceans, which use mucus nets to capture bac-terial cells sieved from suspension. But most bacter-iovores are also very small. Nanoflagellates can clearup to 105 body volumes per hour, thus making aliving from harvesting small, rare bacterial prey,and generally dominating bacterivory in the sea.Protozoan bacterivory closely balances BP in less-productive oceanic regimes. Most crustacean zoo-plankton cannot efficiently harvest bacterioplanktonunless the latter are attached to particles, effectivelyincreasing their size. Bacterial prey enter marine foodwebs following ingestion by flagellates, and ingestionof the flagellates by other flagellates, ciliates, andcopepods. This means that bacteria usually enter thehigher trophic levels after several cycles of ingestionby consumers of increasing size, with attendantmetabolic losses at each stage. The microbial loopand its characteristic long, inefficient food chains canbe short-circuited by the gelatinous bacteriovores,which package bacterial cells into larger prey.

Table 1 The biomass (B) and production rates (P) of bacterioplankton and phytoplankton at estuarine and open ocean locationsa

Biomass (mmol m�3) Production rate (mmol m�3 d�1) P/B (d� 1)

Location Phytoplankton Bacteria Phytoplankton Bacteria Phytoplankton Bacteria

Chesapeake Bay 5–400 1–80 20–47 0.1–50 0.07–1.9 0.01–2

(56) (11) (33) (4) (0.34)

Sargasso Sea 0.3–3.2 0.2–0.6 0.06–0.9 0.002–0.07 0.1–1 0.01–0.16

(1.0) (0.4) (0.3) (0.02) (0.3) (0.06)

a The values are annual, euphotic zone averages derived from published reports. P/B is the specific turnover rate for the population.

The data are presented as ranges with the mean of various estimates in parentheses. Ranges encompass observations and as-

sumptions about conversion factors for deriving values from measurements (see text).

26 BACTERIOPLANKTON

Compared to phytoplankton and to bacteriovores,bacteria are enriched relative to body carbon in ni-trogen, phosphorus, protein, nucleic acids, and iron.Their excess nutritional content, coupled with themany trophic exchanges that bacterial biomass pas-ses through as it moves in food webs, means that themicrobial loop is primarily a vehicle for nutrient re-generation in the sea rather than an important sourceof nutrition for the upper trophic levels. The mainfunction of bacteria in the microbial loop is to re-cover ‘lost’ DOM, enrich it with macro- andmicronutrients, and make it available for regener-ation and resupply to primary producers. Lowerbacterial production estimates (see above) wouldalso tend to decrease the importance of bacteria as asubsidy for higher consumers.

In estuaries and other shallow near-shore habitats,BP is not as closely balanced by planktonic bacter-iovores as in ocean systems. In these productivehabitats, bacteria are larger and more often associ-ated with particles, so they are vulnerable to a widerrange of grazers. Bacteria can also be consumed bymussels, clams, and other benthic suspension feeders.External subsidies of organic matter mean that BP ishigher inshore, so bacteria are a more importantfood source in coastal and estuarine food webs thanin oceanic waters. In these productive systems wherebacterial abundance is greater, more of the bacterialstock is also attacked and lysed by viruses, resultingin release of DOM and nutrients instead of entry intofood webs. The relative importance of viruses andbacteriovores in removing bacteria is not yet wellknown, but has important implications for food webstructure.

Bacteria are the major engines of biogeochemicalcycling on the planet, and serve to catalyze majortransformations of nitrogen and sulfur as well as ofcarbon. They participate in the carbon cycle in sev-eral ways. Their principal role is to serve as a sink forDOM, and thus regulate the export of DOM fromthe productive layer. Bacteria also have intensivehydrolytic capability and participate in de-composition and mineralization of particles and ag-gregates. Bacteria rapidly colonize fresh particulatematter in the sea, and elaborate polymeric materialthat helps to cement particles together, so they bothreduce particle mass by enzymatic hydrolysis andpromote particle formation by fostering aggregation.The balance of bacterial activity for forming particlesand accelerating particle sedimentation or, in con-trast, decomposing particles and reducing it, is notclear. Larvaceans and other giant, specialized bac-teriovores, centimeters to meters in size, can

repackage tiny bacterial cells into large, rapidlysinking fecal aggregates, thus feeding the ocean’ssmallest organisms into the biological carbon pump.

Conclusions

Knowledge of the dynamics of bacterioplankton,their identity, roles in food webs and biogeochemicalcycles is now becoming better known and integratedin a general theory of plankton dynamics, but theseaspects are not yet common in plankton ecosystemmodels. The differential importance of bacteria inplankton food webs in coastal and oceanic systemsmight serve as a good test of our understanding inmodels. The dynamics of DOM are only crudelyparametrized in most models, and explicit formu-lation of bacterial DOM utilization may help inbetter characterizing DOM accumulation and ex-port. Other interesting problems such as the effectsof size-selective predation, bacterial communitystructure, and species succession are just beginning tobe explored. Exploration of marine bacterial com-munities together with molecular probes and nu-merical models should lead to a new revolution inplankton ecology.

See also

Copepods. Marine Mammal Trophic Levels andInteractions. Primary Production Methods. PrimaryProduction Processes. Protozoa, PlanktonicForaminifera. Protozoa, Radiolarians.

Further Reading

Azam F (1998) Microbial control of oceanic carbon flux:the plot thickens. Science 280: 694--696.

Carlson C, Ducklow HW, and Sleeter TD (1996) Stocksand dynamics of bacterioplankton in the northwesternSargasso Sea. Deep-Sea Research II 43: 491--516.

del Giorgio P and Cole JJ (1998) Bacterial growthefficiency in natural aquatic systems. Annual Review ofEcological Systems 29: 503--541.

Ducklow HW and Carlson CA (1992) Oceanic bacterialproductivity. Advances in Microbial Ecology 12:113--181.

Kemp PF, Sherr B, Sherr E, and Cole JJ (1993) Handbookof Methods in Aquatic Microbial Ecology. Boca Raton,FL: Lewis Publishers.

Kirchman DL (2000) Microbial Ecology of the Oceans.New York: Wiley.

Pomeroy LR (1974) The ocean’s food web, a changingparadigm. BioScience 24: 499--504.

BACTERIOPLANKTON 27

PROTOZOA, RADIOLARIANS

O. R. Anderson, Columbia University, Palisades,NY, USA

Copyright & 2001 Elsevier Ltd.

Introduction

Radiolarians are exclusively open ocean, silica-se-creting, zooplankton. They occur abundantly inmajor oceanic sites worldwide. However, some spe-cies are limited to certain regions and serve as indi-cators of water mass properties such as temperature,salinity, and total biological productivity. Abun-dances of total radiolarian species vary across geo-graphic regions. For example, maximum densitiesreach 10 000 per m3 in some regions such as thesubtropical Pacific. By contrast, densities range about3–5 per m3 in the Sargasso Sea. Radiolarians areclassified among the Protista, a large and eclecticgroup of eukaryotic microbiota including the algaeand protozoa. Algae are photosynthetic, single-celledprotists, while the protozoa obtain food by feedingon other organisms or absorbing dissolved organicmatter from their environment. Radiolarians aresingle-celled or colonial protozoa. The single-celled

species vary in size from o100 mm to very largespecies with diameters of 1–2 mm. The larger speciesare taxonomically less numerous and include mainlygelatinous species found commonly in surfacewaters. The smaller species typically secrete siliceousskeletons of remarkably complex design (Figure 1).The skeletal morphology is species-specific and usedin taxonomic identification. Larger, noncolonialspecies are either skeletonless, being enclosed only bya gelatinous coat, or produce scattered siliceousspicules within the peripheral cytoplasm and sur-rounding gelatinous layer. Colonial species containnumerous radiolarian cells interconnected by a net-work of cytoplasmic strands and enclosed within aclear, gelatinous envelope secreted by the radiolarian.The colonies vary in size from several centimeters tonearly a meter in length. The shape of the colonies ishighly variable among species. Some are spherical,others ellipsoidal, and some are elongate ribbon-shaped or cylindrical forms. These larger species ofradiolarians are arguably, the most diverse andlargest of all known protozoa. Many of the surface-dwelling species contain algal symbionts in the per-ipheral cytoplasm that surrounds the central cellbody. The algal symbionts provide some nutrition tothe radiolarian host by secretion of photosynthetic-ally produced organic products. The food resources

(A) (B)

Figure 1 Morphology of polycystine radiolaria. (A) Spumellarian with spherical central capsule and halo of radiating axopodia

emerging from the fusules in the capsular wall and surrounded by concentric, latticed, siliceous shells. (B) Nassellarian showing the

ovate central capsule with conical array of microtubules that extend into the basally located fusules and external axopodia protruding

from the opening of the helmet-shaped shell. Reproduced with permission from Grell K (1973) Protozoology. Berlin: Springer-Verlag.

28

are absorbed by the radiolarian and, combined withfood gathered from the environment, are used tosupport metabolism and growth. Radiolarians thatdwell at great depths in the water column where lightis limited or absent typically lack algal symbionts.The siliceous skeletons of radiolarians settle into theocean sediments where they form a stable and sub-stantial fossil record. These microfossils are an im-portant source of data in biostratigraphic andpaleoclimatic studies. Variations in the number andkind of radiolarian species (based on skeletal form)in relation to depth in the sediment provide infor-mation about climatic and environmental conditionsin the overlying water mass at the time the radio-larian skeletons were deposited at that geographiclocation. The radiolarians are second only to diatomsas a major source of biogenic opal (silicate) depositedin the ocean sediments.

Cellular Morphology

The radiolarian cell body contains a dense mass ofcentral cytoplasm known as the central capsule(Figure 2). Among the organelles included in thecentral capsule are the nucleus, or nuclei in specieswith more than one nucleus, most of the food re-serves, major respiratory organelles, i.e., mito-chondria, Golgi bodies for intracellular secretion,protein-synthesizing organelles, and vacuoles. Thecentral capsule is surrounded by a nonliving capsularwall secreted by the radiolarian cytoplasm. Thethickness of the capsular wall varies among species.It may be thin or in some species very reduced,consisting of only a sparse deposit of organic mattercontained within the surrounding cytoplasmic en-velope. In others, the wall is quite thick and opal-escent with a pearl-like appearance. The capsularwall contains numerous pores through which cyto-plasmic strands (fusules) connect to the extra-capsular cytoplasm. The extracapsular cytoplasmusually forms a network of cytoplasmic strands at-tached to stiffened strands of cytoplasm known asaxopodia that extend outward from the fusules in thecapsular wall. The central capsular wall and axo-podia are major defining taxonomic attributes ofradiolarians. A frothy or gelatinous coat typicallysurrounds the central capsule and supports theextracapsular cytoplasm. Algal symbionts, whenpresent, are enclosed within perialgal vacuoles pro-duced by the extracapsulum. In most species, thealgal symbionts are exclusively located in the extra-capsulum. Thus far, symbionts have been observedwithin the central capsular cytoplasm in only a fewspecies. Food particles, including small algae and

protozoa or larger invertebrates such as copepods,larvacea, and crustacean larvae, are captured by thesticky rhizopodia of the extracapsulum. The cyto-plasm moves by cytoplasmic streaming to coat andenclose the captured prey. Eventually, the prey isengulfed by the extracapsular cytoplasm and di-gested in digestive vacuoles (lysosomes). These typi-cally accumulate in the extracapsulum near thecapsular wall. Large prey such as copepods are in-vaded by flowing strands of cytoplasm and the morenutritious soft parts such as muscle and organ tissuesare broken apart, engulfed within the flowing cyto-plasm and carried back into the extracapsulumwhere digestion takes place. The siliceous skeleton,when present, is deposited within cytoplasmic spacesformed by extensions of the rhizopodia. This elab-orate framework of skeletal-depositing cytoplasm isknown as the cytokalymma. Thus, the form of theskeleton is dictated by the dynamic streaming andmolding action of the cytokalymma during the silicadeposition process. Consequently, the very elaborateand species-specific form of the skeleton is deter-mined by the dynamic activity of the radiolarian and

PV

N

SK

DV.

CW

Figure 2 Cytoplasmic organization of a spumellarian

radiolarian showing the central capsule with nucleus (N),

capsular wall (CW) and peripheral extracapsulum containing

digestive vacuoles (DV) and algal symbionts in perialgal vacuoles

(PV). The skeletal matter (SK) is enclosed within the

cytokalymma, an extension of the cytoplasm, that acts as a

living mold to dictate the shape of the siliceous skeleton

deposited within it. Reproduced with permission from Anderson

OR (1983) Radiolaria. New York: Springer-Verlag.

PROTOZOA, RADIOLARIANS 29

is not simply a consequence of passive physicalchemical processes taking place at interfaces amongthe frothy components of the cytoplasm as waspreviously proposed by some researchers.

Taxonomy

Radiolarians are included in some modern classifi-cation schemes in the kingdom Protista. However, thecategory of radiolaria as such is considered an arti-ficial grouping. Instead of the group ‘Radiolaria’, twomajor subgroups previously included in ‘Radiolaria’are placed in the kingdom Protista. These are thePolycystina and the Phaeodaria. Polycystina areradiolarians that contain a central capsule with poresthat are rather uniform in shape and either uniformlydistributed across the surface of capsular wall, orgrouped at one location. The Phaeodaria have cap-sular walls with two distinctive types of openings.One is much larger and is known as the astropyle withan elaborately organized mass of cytoplasm extendinginto the extracapsulum. The other type is composedof smaller pores known as parapylae with thin strandsof emergent cytoplasm. Some Phaeodaria also haveskeletons that are enriched in organic matter com-pared with the skeletons of the Polycystina. Amongthe Polycystina, there are two major taxonomicgroups, the Spumellaria and Nassellaria, assigned asorders in some taxonomic schemes. Spumellaria havecentral capsules that are usually spherical or nearly soat some stage of development and have pores dis-tributed uniformly over the entire surface of the cap-sular wall. All known colonial species are members ofthe Spumellaria. Although expert opinion varies, thereare two families and about 10 genera of colonialradiolarians. There are seven widely recognizedfamilies of solitary Spumellaria with scores of genera.Nassellaria have central capsules that are more ovateor elongated and the fusules are located only at onepole of the elongated capsular wall. This pore field iscalled a porochora and the fusules tend to be robustwith axopodia that emerge through outward-directedcollar-like thickenings surrounding the pore rim.Moreover, the skeleton of the Nassellaria, when pre-sent, tends to be elongated and forms a helmet-shapedstructure, often with an internal set of rods forming atripod to which the external skeleton is attached.Current systematics include seven major families withnumerous genera. Spumellarian skeletons are typicallymore spherical, or based on a form that is not derivedfrom a basic tripodal or helmet-like architectural plan.The shells of the Phaeodaria are varied in shape. Somespecies have ornately decorated open lattices resem-bling geodesic structures composed of interconnected,hollow tubes of silica. Other species have thickened

skeletons resembling small clam shells with closelyspaced pores on the surface. There are 17 majorfamilies with scores of genera. Since many species ofradiolarians were first identified from sediments basedsolely on their mineralized skeletons, much of the keytaxonomic characteristics include skeletal morph-ology. Increasingly, evidence of cytoplasmic finestructure obtained by electron microscopy and mo-lecular genetic analyses is being used to augmentskeletal morphology in making species discrimin-ations and constructing more natural evolutionaryrelationships. It is estimated that there are severalhundred valid living species of radiolarians.

Biomineralization

Biomineralization is a biological process of secretingmineral matter as a skeleton or other hardened prod-uct. The skeleton of radiolaria is composed of hydratedopal, an oxidized compound of silicon (nominallySiO2 � nH2O) highly polymerized to form a space-fill-ing, glassy mass incorporating a variable number (n) ofwater molecules within the molecular structure of thesolid. Electron microscopic evidence indicates thatsome organic matter is incorporated in the skeletonduring early stages of deposition, but on the whole, theskeleton is composed mostly of pure silica. Electronmicroscopic, X-ray dispersive analysis shows that asmall amount of divalent cations such as Ca2þ may beincorporated in the final veneer deposited on the sur-face to enhance the hardness of the skeleton. Duringdeposition of the skeleton, the cytoplasm forms theliving cytokalymma, i.e., the cytoplasmic silica-de-positing mold, by extension of the surface of the rhi-zopodia. The cytokalymma enlarges as silica isdeposited within it, gradually assuming a final formthat dictates the morphology of the internally secretedskeleton. Small vesicles are observed streaming out-ward from the cell body into the cytoplasm of thecytokalymma and these may bring silica to be de-posited within the skeletal spaces inside the cytoka-lymma. The cytoplasmic membrane surrounding thedeveloping skeleton appears to act as a silicalemma oractive membrane that deposits the molecular silica intothe skeletal space. During deposition, the dynamicmolding process is clearly evident as the living cyto-plasm continuously undergoes transformations inform, gradually approximating the ultimate geometryof the species-specific skeleton being deposited by theradiolarian cell. In general, species with multiple,concentric, lattice shells surrounding the central cap-sule appear to lay down the lattices successively, pro-gressing outward from the innermost shell.

The process of skeletal construction has beendocumented in fair detail for a few species, most

30 PROTOZOA, RADIOLARIANS

notably among the colonial radiolaria. Two forms ofgrowth have been identified. Bar growth is a processof depositing silica as rodlets within a thin tubularnetwork of cytoplasm formed by the cytokalymma.The rodlets become connected during silicogenesisand further augmented with silica to form a porouslattice with typically large polygonal pores. Thepores, once formed, may be further subdivided intosmaller pores by additional bar growth that spans theopening of the pore. Rim growth occurs by de-position of silica as curved plates that are differen-tially deposited at places to form rounded pores. Atmaturity, these are typically spherical skeletons withrather regular, rounded pores scattered across thesurface. For both types of skeletons, in some species,the ratio of the bar width between the pores to thepore diameter is a taxonomic diagnostic feature.

The rate of silica biomineralization in some specieshas been determined by daily observation of growthof individuals in laboratory culture using light mi-croscopy. The amount of silica in the skeleton of aliving radiolarian is mathematically related to thesize of the skeleton. For example, in the spumellarianspecies Spongaster tetras with a rectangular, spon-giose skeleton, the amount of silica (W) in micro-grams (mg) as related to the length of the majordiagonal axis of the quadrangular shell (L) in mi-crometers (mm) is approximated as follows:

W ¼ 3:338� 10�6� �

� L2:205 ½1�

The average daily growth in cultures of an S. tetras is3 mm with an average daily gain in weight of c. 8 ng.The total weight gain for one individual radiolarianduring maturation is about 0.1 mg. Silica depositionduring maturation appears to be sporadic and ir-regular, varying from one individual to another, withperiods of rapid deposition followed by plateaus ingrowth. The amount of skeletal opal produced by S.tetras alone in the Caribbean Sea, for example, is c.42 mg per m3 of sea water, with a range of 8–61 mgper m3. Peak production occurred in mid-summer(June to July). The rate of total radiolarian-producedbiogenic opal settling into the ocean sediments atvarying oceanic locations has been estimated in therange of 1–10 mg per m2 per day.

Reproduction

Protozoa reproduce by either asexual or sexual re-production. Asexual reproduction occurs by celldivision during mitosis to produce two or moregenetically identical offspring. Sexual reproductionoccurs by the release of haploid gametes (e.g., spermand egg cells) that fuse to produce a zygote with

genetic characteristics contributed by both of theparent organisms. Thus, sexual reproduction permitsnew combinations of genetic material and the off-spring are usually genetically different from theparents. There is evidence that some colonial radi-olaria have asexual reproduction. The central cap-sules within the colony have been observed to divideby fission. This increases the number of centralcapsules and allows the colony to grow in size. Thecolony may also break into parts, thus increasing thetotal numbers of colonies at a given location. In mostspecies of radiolaria, reproduction occurs by releaseof numerous flagellated swarmer cells that are be-lieved to be gametes. The nucleus of the parentradiolarian undergoes multiple division and the en-tire mass of the parent cell is converted into unin-ucleated flagellated swarmers. These are releasednearly simultaneously in a burst of activity, andpresumably after dispersal fuse to form a zygote. Thedetails of gamete fusion and the early ontogeneticdevelopment of radiolaria are poorly understood andrequire additional investigation. Ontogenetic devel-opment of individuals from very early stages to ma-turity has been documented in laboratory culturesand the stages of skeletal deposition are well under-stood for several species, as explained above in thesection on biomineralization.

Physiological Ecology andZoogeography

The physiological ecology of radiolaria has beenstudied by collecting samples of radiolaria and otherbiota at varying geographical locations in the worldoceans to determine what abiotic and biotic factorsare correlated with and predict their abundances, andby experimental studies of the physical and biologicalfactors that promote reproduction, growth, and sur-vival of different species under carefully controlledlaboratory conditions. Temperature appears to be amajor variable in determining abundances of somespecies of radiolaria. For example, high latitude spe-cies that occur abundantly at the North or South Polesare also found at increasing depths in the oceanstoward the equator. Since the water temperature ingeneral decreases with depth, these organisms popu-late broad depth regions within the water column thatmatch their physiological requirements. Species thatoccur in subtropical locations, where the water isintermediate in temperature based on a global range,are found at the equator at intermediate water depthsthat are cooler than the warm surface water. Somespecies are characteristically most abundant in onlywarm, highly productive water masses. For example,

PROTOZOA, RADIOLARIANS 31

some species of colonial radiolaria occur typically insurface water near the equator in the Atlantic Ocean,while others are most abundant at higher latitudes inthe Sargasso Sea where usually the water is also lessproductive. Upwelling regions where deep, nutrient-enriched sea water is brought to the surface are typi-cally highly productive regions for radiolaria, as oc-curs for example along the Arabian, Chilean, andCalifornia coast lines. Shallow-water dwelling specieshave been categorized into seven zoogeographic zonesbased on water mass properties: (1) SubArctic at highnorthern latitudes; (2) transition region as occurs inthe North Pacific drift waters; (3) north central region,typical of waters within the large anticyclonic circu-lation of the North Pacific; (4) equatorial region inlocations occupied by the North and South EquatorialCurrent systems; (5) south central water mass, as inthe South Pacific anticyclonic circulation pattern; (6)subAntarctic, a water regime bounded on the north bythe Subtropical Convergence and on the south by thePolar Convergence; and (7) Antarctic, bounded by thePolar Convergence on the north and the AntarcticContinent on the south.

The growth requirements of some species have beenstudied extensively in laboratory cultures. For example,the following three surface- to near-surface-dwellingspecies exhibit a range of optimal growth conditions.Didymocyrtis tetrathalamus, with a somewhat hour-glass-shaped skeleton (150mm), prefers cooler water(21–271C) and salinities in the range of 30–35 ppm.Dictyocoryne truncatum, a spongiose triangular-shapedspecies (300mm), is more intermediate in habitat re-quirements with optimal temperature of 281C and sal-inity of 35 ppm. Spongaster tetras, a quadrangular,spongiose species (300mm), prefers warmer, more salinewater (c. 281C and 35–40 ppm). The temperature tol-erance ranges (in 1C) for the three species also show asimilar pattern of increasing preference for warmerwater, i.e., 10–34, 15–28, and 21–31, respectively.

The prey consumed by radiolarians varies sub-stantially among species, but many of the polycystinespecies appear to be omnivorous, consuming bothphytoplankton and zooplankton prey. The smallerspecies consume microplankton and bacteria. Largerspecies are capable of capturing copepods and smallinvertebrates. Phaeodaria, especially those speciesdwelling at great depths in the water column, appearto consume detrital matter in addition to preying onplankton in the water column. The broad range ofprey accepted by many of the radiolarians studiedthus far suggests that they are opportunistic feedersand are capable of adapting to a broad range oftrophic conditions.

The role of algal symbionts, when present, hasbeen debated for some time – beginning with theirdiscovery in the mid-nineteenth century. At first, itwas supposed that the green symbionts may largelyprovide oxygen to the host. However, most radi-olaria dwell in fairly well-oxygenated habitats and itis unlikely that photosynthetically derived oxygen isnecessary. The other competing hypothesis was thatthe symbionts provide organic nourishment to thehost. Modern physiological studies have confirmedthat the algal symbionts provide photosyntheticallyproduced nutrition for the host. Biochemical ana-lyses combined with 14C isotopic tracer studies haveshown that stores of lipids (fats) and carbohydrates inthe host cytoplasm contain carbon derived from algalphotosynthetic activity. Well-illuminated, laboratorycultures of symbiont-bearing radiolaria survive forweeks without addition of prey organisms. Some ofthe algal symbionts are digested as food and can bereplaced by asexual reproduction of the algae, but itappears that much of the nutrition of the host comesfrom organic nutrients secreted into the host cyto-plasm by the algal symbionts. This readily available,‘internal’ supply of autotrophic nutrition makes sym-biont-bearing radiolaria much less dependent on ex-ternal food sources and may account in part for theirwidespread geographic distribution, including someoligotrophic water masses such as the Sargasso Sea.

Further Reading

Anderson OR (1983) Radiolaria. New York: Springer-Verlag.

Anderson OR (1983) The radiolarian symbiosis. In: Goff LJ(ed.) Algal Symbiosis: A Continuum of InteractionStrategies, pp. 69--89. Cambridge: Cambridge UniversityPress.

Anderson OR (1996) The physiological ecology of plank-tonic sarcodines with applications to paleoecology:Patterns in space and time. Journal of EukaryoticMicrobiology 43: 261--274.

Cachon J, Cachon M, and Estep KW (1990) PhylumAnctinopoda Classes Polycystina (¼Radiolaria) andPhaeodaria. In: Margulis L, Corliss JO, Melkonian M,and Chapman DJ (eds.) Handbook of Protoctista,pp. 334--379. Boston: Jones and Barlett.

Casey RE (1971) Radiolarians as indicators of past andpresent water masses. In: Funnell BM and Riedel WR(eds.) The Micropaleontology of Oceans, pp. 331--349.Cambridge: Cambridge University Press.

Steineck PL and Casey RE (1990) Ecology and paleobiologyof foraminifera and radiolaria. In: Capriulo GM (ed.)Ecology of Marine Protozoa, pp. 46--138. New York:Oxford University Press.

32 PROTOZOA, RADIOLARIANS

PROTOZOA, PLANKTONIC FORAMINIFERA

R. Schiebel and C. Hemleben, Tubingen University,Tubingen, Germany

Copyright & 2001 Elsevier Ltd.

Introduction

Planktonic foraminifers are single celled organisms(protozoans) sheltered by a test (shell) made of cal-cite, with an average test diameter of 0.25 mm. Theylive in surface waters of all modern open oceans anddeep marginal seas, e.g., Mediterranean, CaribbeanSea, Red Sea, and Japan Sea, and are almost absentfrom shelf areas including the North Sea and othershallow marginal seas. Planktonic foraminifers con-stitute a minor portion of the total zooplankton, butare the main producers of marine calcareous par-ticles deposited on the ocean floor and form the so-called ‘Globigerina ooze.’

Planktonic foraminifers (Greek: foramen¼ open-ing, ferre ¼ carry) first appeared in the middle Jur-assic, about 170 million years ago (Ma), and spreadsince the mid-Cretaceous over all world oceans.Times of main appearance of new species in theAptian (120 Ma), the Turonian (90 Ma), the Paleocene(55 Ma), and the Miocene (20 Ma), alternate withphases of main extinction in the Cenomanian (95 Ma),at the Cretaceous/Tertiary boundary (60 Ma), and inthe Upper Eocene (40 Ma). Modern planktonic fora-minifers have evolved since the early Tertiary, whenfirst spinose species occurred directly after the Cret-aceous/Tertiary boundary. Approximately 450 fossiland 50 Recent species are known, not includingspecies based on molecular biology investigations. Theappearance and radiation of new species seem tocorrelate with the development of new realms andniches, linked to plate tectonics and paleoceano-graphic changes. The geographical distribution andmain events in planktonic foraminiferal evolution areassociated in general with water mass properties, e.g.,availability of food or temperature. The reproductivestrategies depend highly on their life habitat in thephotic zone or slightly below. The life span of plank-tonic foraminifers varies between 14 days and a year,mostly linked to the lunar cycle. Most living speciesbear symbionts requiring a habitat in the upper tomiddle photic zone. Their feeding habit depends onthe spinosity (spinose versus nonspinose species) inrespect to the size and class of prey. Predators that

are specialized on planktonic foraminifers are notknown.

History

With the technological improvement of microscopesd’Orbigny in 1826 was able to describe the firstplanktonic foraminiferal species, Globigerina bul-loides, from beach sands, and classified it as acephalopod. In 1867 Owen described the planktoniclife habit of these organisms. Following the Chal-lenger Expedition (1872–1876) the surface-dwellinghabitat of planktonic foraminifers was recognized.Rhumbler first described the biology of foraminifersin 1911. In the first half of the twentieth century,foraminifers were widely used for stratigraphic pur-poses, and many descriptions were published, mainlyby Josef A. Cushman and co-workers. Studies on thegeographic distribution of individual foraminiferalspecies are based on samples from the living plank-ton since the work of Schott in 1935. Planktonicforaminifers have been used since the beginning ofthe twentieth century to date marine sediments dril-led by oil companies, and later on through the Deep-Sea Drilling and Ocean Drilling Programs. In add-ition, extensive studies on distribution, ecology oflive and fossil faunas were carried out to understandthe changing marine environment. The ecologicalsignificance has been applied in paleoecological andpaleoceanographic settings and yielded subtle infor-mation on ancient oceans and the Earth’s climate.Recent investigation still focuses on evolution andpopulation dynamics. Modern techniques, e.g.,polymerase chain reaction (PCR), are being usedto reveal the genetic code, and their relation tomorphological classification tests needs to bechecked.

Methods

Planktonic foraminifers are sampled from the watercolumn by plankton nets of various design, with amesh size of 0.063–0.2 mm, by employing planktonrecorders, water samplers, pumping systems, or col-lection by SCUBA divers. To study faunas fromsediment samples or consolidated rock, the sur-rounding sediment has to be disaggregated byhydrogen peroxide, tensids, acetic acid (pure), orphysical methods, and washed over a sieve (0.03–0.063 mm). Shells may be studied under a binocularmicroscope, or with a scanning electron microscope

33