General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

You may not further distribute the material or use it for any profit-making activity or commercial gain

You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Jun 03, 2020

Defense mechanisms in phytoplankton: traits and trade–offs

Mohr, Marina Pancic

Publication date:2019

Document VersionPublisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):Mohr, M. P. (2019). Defense mechanisms in phytoplankton: traits and trade–offs. Technical University ofDenmark.

Defense mechanisms in phytoplankton: traits and trade–offs

PhD Thesis

By Marina Pančić Mohr

DTU Aqua National Institute of Aquatic Resources

Defense mechanisms in phytoplankton:traits and trade–offs

PhD THESIS

Marina Panc̆ić Mohr

Panc̆ić, M.M. (2019) Defense mechanisms in phytoplankton: traits and trade–off

PhD thesis, Centre for Ocean Life, Technical University of Denmark, Denmark

The PhD project was part of the Centre for Ocean Life, a VKR centre of excel-lence that is founded by the Villum Foundation. Additional support was receivedby the Moore Foundation to Prof. Dr. Thomas Kiørboe, as well as Idella Foun-dation to Marina Panc̆ić Mohr.

Defense mechanisms in phytoplankton:traits and trade–offs

PhD THESIS

ACADEMIC THESIS

to obtain the Doctor of Philosophy Degreeat the Technical University of Denmark,

a public defense of the thesis will be given byMarina Panc̆ić Mohr

in building 303A, auditorium 45on Friday, March 8, 2019 at 13.00

in the presence of assessment committee

Principal supervisor:Prof. Dr. Thomas Kiørboe Centre for Ocean Life, DTU Aqua

Co–supervisor:Prof. Dr. Andre W. Visser Centre for Ocean Life, DTU Aqua

Assessment committee:Chairman:Dr. Sigrún H. Jónasdóttir DTU Aqua

Examiners:Prof. Dr. Marina Montresor Stazione Zoologica Anton DohrnAssist. Prof. Dr. Susanne Wilken University of Amsterdam

Summary

In aquatic ecosystems, phytoplankton form the base of the food webs. They gen-erate nearly half of the global primary productivity and consequently play a vitalrole in regulating Earth’s climate via carbon cycling. In addition, phytoplanktoncommunity composition impacts the biogeochemical cycles of inorganic elementssince different functional groups of phytoplankton have different requirementsand acquisition modes for nutrients. Therefore, since phytoplankton communitycomposition impacts the global climate and the functioning of aquatic ecosys-tems, it is important to understand what mechanisms in turn determine thecommunity composition.

Evidently, one such mechanism is predation and the subsequent employmentof defensive strategies, which promotes coexistence and species diversity. How-ever, the promotion of species diversity requires trade–offs. This implies that theadvantage of a defense mechanism must come at a cost, otherwise all specieswould evolve towards a state of equal defense and the community compositionwould no longer be promoted by predation. Based on the available evidence,it appears that the function of many proposed defensive traits in phytoplank-ton remains unclear and the trade–offs undocumented or unquantified. In manycases, experimental evidence even suggests that defenses are costless. Here, wepropose that some costs may materialize only under natural conditions, whileother may become evident under resource–deficient conditions when a rivalryfor limiting resources between growth and defense occurs. For that reason, amechanistic understanding of the hypothesized component processes is requiredfor evaluation of costs that are realized under natural conditions, while the mag-nitude of the costs must be assessed under conditions of resource limitation.

As stated above, the role of many proposed defenses remains elusive includ-ing the protective role of silicified cell walls in diatoms. While it seems intuitivethat the siliceous wall may have evolved as a mechanical protection against graz-ing, many other roles of a siliceous shell have been proposed and direct evidenceof defense is limited. In addition, the anticipated benefits must be traded offagainst the costs; since deposition of silica in diatoms is determined by theirgrowth rates, this dependency can be regarded as one of the costs of silicifica-tion. We experimentally demonstrate the protective role of silicified cell wallsagainst adult copepods and nauplii, with near inversely proportional relation-ship between predation risk and silica wall thickness. On the other hand, our

iv

empirical data reveal that the increased cell wall thickness is inconsequentialto protozoan grazers that engulf their prey. Additionally, we demonstrate thatthe deposition of silica in diatoms decreases with increasing growth rates, sug-gesting a possible cost to silicification. Overall, it appears that the silica wallis an efficient defense mechanism against copepods, implying that the plasticityof silicification in diatoms has likely evolved as a response to copepod grazingwhose specialized tools to break open silicified walls have co–evolved with theirprey.

Another common trait observed in phytoplankton is the production of chem-ical compounds, yet often the evolution and the function of such compoundsremain unclear. Often, a defensive role of toxin production is anticipated, andthis interpretation is supported by observations of increased toxin synthesis inthe presence of grazers, which in some cases leads to reduced predation mortal-ity. On the other hand, experiments have consistently failed to observe any costsof toxin production, but a mechanistic understanding of the underlying processesmay allow their quantification — i.e. that the production of nitrogen–rich com-pounds likely depends on ambient nitrogen levels. We demonstrate with a simplefitness optimization model that the cost of toxin production is indeed negligiblewhen nitrogen and light are plentiful. However, when nitrogen or light is limitingcell growth and when grazers are abundant, the model predicts substantial coststhat lead to a significant reduction in cell division rates. Our results indicatethat the investment in toxin production pays–off since defended cells experiencereduced grazing mortality, which consequently leads to the net growth rates thatare twice as high as of undefended cells.

In this PhD work, I provided an overview of defense mechanisms in phyto-plankton and associated trade–offs. In search for defense trade–offs, I identifiedwhich traits lack experimental evidence supporting their defensive function (e.g.silicified cell walls), and proposed ways to assess the costs in future experimentalor theoretical studies (e.g. toxin production).

Resumé

Fytoplankton udgør grundlaget for fødenettet i akvatiske økosystemer, gene-rer næsten halvdelen af den årlige globale netto primærproduktion og spilleren vigtig rolle i jordens klimaregulering via kulstofcykling. Sammensætningenaf fytoplanktonsamfund påvirker de bio–geokemiske processer idet funktionellefytoplankton–grupper har forskellige krav og optagelse af næringsstoffer. Pågrund af denne differentiering og den følgende påvirkning af det globale klima ogfunktioner i akvatiske økosystemer, er det vigtigt at forstå hvilke mekanismer,der påvirker samfundenes sammensætning.

Prædation og den derpå følgende udvikling af defensive strategier har vist sigat være afgørende for sammensætning af fytoplanktonsamfund. Et sådan “vå-benkapløb” er dog forbundet med “trade–offs”. Dette indebærer, at fordelene veden bestemt forsvarsmekanisme er forbundet med omkostninger. Hvis forsvaretvar gratis ville alle arter udvikle forsvar eller arter med forsvar ville udkonkur-rere arter uden forsvar, og prædations samfundsstrukturerende rolle ville ikkevære åbenlys. Mange forsvarsmekanismer og trade–offs i fytoplanktonsamfunder stadig ikke tilstrækkelig belyst eller kvantificeret, og mange eksperimentellestudier har vist, at forsvarsmekanismer tilsyneladende er udgiftsfri. Vi foreslår,at visse af disse udgifter kun materialiseres under naturlige forhold eller underforhold, hvor ressourcetilgængeligheden er lav og ressourcekonkurrence opstårmellem vækst og forsvarsmekanismer. En mekanistisk forståelse af disse fore-slåede processer er nødvendig for at evaluere og kvantificere udgifterne undernaturlige forhold og begrænsende ressourcetilgængelighed.

Som nævnt er mange af de foreslåede forsvarsmekanismer stadig ikke til-strækkelig belyst. En af disse, er de kiselholdige cellevægge i kiselalger. Til trodsfor at kiselholde cellevægge virker som en tilsyneladende intuitiv forsvarsmeka-nisme, har disse strukturer muligvis også andre roller og egentlige beviser forværdien som forsvarsmekanisme er utilstrækkelige. Ud over dette, må eventuellefordele ved kiselholdige vægge vejes op mod udgifter og trade–offs. Væksten hoskiselalger er også afhængig af kisel og der vil derfor være en afvejning mellemvækst og forsvar. Vi viser eksperimentelt, at kiselholdige vægge udgør en ef-fektiv forsvarsmekanisme mod vandlopper og vandloppe–nauplie. Vi fandt næromvendt proportionalitet mellem prædationsrisiko og godstykkelsen på celle-vægge hos kiselalger. Dog viste vores empiriske data også at tykkere cellevæggehos kiselalger ingen effekt har for prædationsraten fra encellede heterotrofe fla-

vi

gellater, hvor byttet sluges helt. Derudover viser vi at deponering af silikat ikiselalger aftager med stigende vækstrater, hvilket foreslår en muligvis trade–offmellem forsvarsmekanismen og væksten. Vi konkluderer, at kiselholdige celle-vægge udgør en effektiv forsvarsmekanisme mod vandlopper, og at plasticitetenaf silikatholdigheden i kiselalger er udviklet i et våbenkapløb med vandloppersevne til at knuse disse kiselholdige vægge.

Kemiske toxiner er en anden observeret forsvarsmekanisme i fytoplankton.Dog er det stadig uvist hvilke udgifter sådan en forsvarsmekanisme har. Enmekanistisk forståelse af de underlæggende processer kan muliggøre en kvanti-ficering af kosten af disse kvælstofholdige kemiske stoffer. Vi demonstrerer meden simpel fitness optimerings–model at produktionen af toxiner medfører ubety-delige omkostninger for cellen under forhold, hvor kvælstof og lys er tilgængeligti store mængder. Dog forudsiger modellen betydelige udgifter og lavere vækstra-ter under svage lysforhold og lave mængder kvælstof. Resultaterne indikerer, atproduktion af toxiner er en fordel, idet toxin–holdige celler oplever nedsat græs-ningstryk og derpå følgende højere nettovækstrater.

I denne PhD. afhandling giver jeg et overblik over foreslåede forsvarsme-kanismer i fytoplankton og de dertil hørende trade–offs. Jeg identificerer hvilkemekanismer, der ikke er tilstrækkeligt belyst (f.eks. kiselholdige vægge) og viser,at trade–offs kan kvantificeres gennem eksperimentelle eller teoretiske studier(f.eks. toxinproduktion).

Acknowledgements

“Life is like riding a bicycle. To keep your balance you must keep moving.”Albert Einstein

The three years as a PhD student at DTU Aqua have been a truly unforget-table experience, which beyond any doubt is also attributed to the stimulatingwork environment at the Centre for Ocean Life. Also, being surrounded with theexcellent mathematicians, physicists, chemists, theoretical ecologists, and biolo-gists readjusted my perception of science for which I am very grateful. I wouldespecially like to thank my supervisors, Thomas Kiørboe and Andy Visser fortheir support and guidance. Thomas, if my PhD was the bicycle then you wereone of the forces that helped moving it. I am beyond grateful for the opportunityto work with you — I enjoyed the feedback, discussions, and freedom you gaveme to pursue my ideas and learn from my own mistakes.

Furthermore, I am grateful to Rocio R. Torres for her utter effort, patienceand support in the lab, and to Colin Stedmon for his help with chemical analy-ses and resolving numerous issues encountered on the way. To Lasse T. Nielsen,I am grateful for his laboratory problem–solving skills, to Helge A. Thomsenfor his help with Adobe Photoshop and fluorescence microscopy, and to HansHansen for building new and repairing existing laboratory equipment wheneverneeded. I further thank Subhendu Chakraborty for coffee breaks during whichfruitful discussions took place and new ideas were born.

A big thank you also to the young researchers associated with the Centre forOcean Life for creating a fantastic social environment during the Friday beers,and the everyday 3–o’clock–cookie breaks. I would also like to thank the JournalClub for discussing science over a cup of coffee, and absolutely passionate con-versations with the Club organizer Sara Harðardóttir, which often led to manybrilliant ideas.

Very special thanks to my dear friends and colleagues Floor H. Soudijn andAurore Maureaud with whom I shared many great and sometimes less–greatmoments. Likewise I wish to thank Ana Ley for countless home–cooked dinners,support, and overall faith in me. My greatest thanks go to my family who hasprovided me with loving support, reliability, and help whenever something wasneeded. Last but not least, to my best friend and husband Sebastian Mohr, Iam thankful for his love, support, help, amazing travels, and all–in–all fantastictimes. It would not have been the same without you!

viii

List of publications

This PhD thesis is based on the following manuscripts:

I Panc̆ić, M. and Kiørboe, T. (2018). Phytoplankton defence mechanisms:traits and trade–offs. Biological Reviews, 93, 1269–1303

II Panc̆ić, M., Torres, R.R., Almeda, R., and Kiørboe, T. (2019). Silicifiedcell walls as a morphological defensive trait in diatoms. Proceedings of theRoyal Society B: Biological Sciences, 286

III Chakraborty, S., Panc̆ić, M., Andersen, K.H., and Kiørboe, T. (2018).The cost of toxin production in phytoplankton: the case of PST producingdinoflagellates. The ISME Journal, 1–12

Additionally, the following manuscripts represent minor contributions to thisthesis:

A Lundholm, N., Krock, B., John, U., Skov, J., Cheng, J., Panc̆ić, M.,Wohlrab, S., Rigby, K., Nielsen, T.G., Selander, E., and Harðardóttir, S.(2018). Induction of domoic acid production in diatoms — Types of grazersand diatoms are important. Harmful Algae, 1–10

B Visser, A.W., Brun, P., Jónasdóttir, S.H., van Denderen, P.D., Nielsen,L.T., Panc̆ić, M., Chakraborty, S., Thygesen, U.H., Heilmann, I., Kenitz,K., Dencker, T.S., Pécuchet, L., van Gemert, R., Schnedler–Meyer, N.A.,Holm, M.W., van Someren Gréve, H., Mariani, P., Payne, M.R., Kiørboe,T., Törnroos, A., and Andersen, K.H. Seasonality and annual routines inthe marine environment. (in preparation)

x

Contents

1 General introduction 11.1 Trait–based approach to ecology . . . . . . . . . . . . . . . . . . . 21.2 Inducible defenses . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.2.1 Terrestrial ecosystems . . . . . . . . . . . . . . . . . . . . . 61.2.2 Aquatic ecosystems . . . . . . . . . . . . . . . . . . . . . . 8

1.3 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2 Manuscript I: Phytoplankton defence mechanisms 17

3 Manuscript II: Silicified cell walls in diatoms 19

4 Manuscript III: Toxin production in dinoflagellates 21

5 Synopsis 235.1 Defensive traits and trade–offs . . . . . . . . . . . . . . . . . . . . 245.2 Defensive value and cost of silica walls . . . . . . . . . . . . . . . 255.3 Defensive value and cost of toxin production . . . . . . . . . . . . 275.4 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . 28

xii CONTENTS

Chapter 1

General introduction

2 1. General introduction

1.1 Trait–based approach to ecologyTrait–based ecology combines evolutionary ecology with classical population andcommunity ecology, and describes ecosystems where individual organisms ratherthan species or populations interact with each other (Kiørboe, Visser, and An-dersen, 2018). According to the trait–based approach, the individual organismsare described by various traits, and the characteristics that capture most of theirDarwinian fitness are the key traits. Hence, fundamental to the trait–based ap-proach is identification of the key traits and quantification of the benefits andcosts (i.e. trade–offs) associated with a particular trait.

The central measure of the individual’s fitness is the success of completingthree fundamental activities, namely, to acquire resources, survive, and repro-duce (Figure 1.1). However, the execution of any of these functions may interferewith one another as the energy obtained from the acquired resources may beallocated to different competing activities. Therefore, the optimal resource allo-cation strategy is the one that maximizes the fitness of the individual in a specificenvironment (Litchman, Ohman, and Kiørboe, 2013; Kiørboe et al., 2018).

Figure 1.1: Three main Darwinian activities in life: resource acquisition, survival,and reproduction. The execution of any of the three functions may conflict withthe others (i.e. trade–offs), as the energy obtained from acquired resources maybe allocated to different competing missions, which cannot be maximized at thesame time. Adapted from Litchman et al. (2013)

The fundamental life functions of acquiring resources and surviving to re-producing will of course be realized differently for different life forms and evenspecies. However, there seem to be a few traits that apply across all life forms,also termed the key traits. These are the body size, resource acquisition mode,and defense (Figure 1.2) (Kiørboe et al., 2018). Firstly, body size is often consid-ered a master trait, as it not only determines the order of magnitude of all vitalrates, such as growth and mortality, metabolism, feeding, and more, but it alsolargely governs the resource acquisition modes (Andersen et al., 2016). Consider

1.1 Trait–based approach to ecology 3

the size of unicellular phytoplankton as an example; the size of cells in itself of-fers partial protection against grazing, as large cells generally experience lowerpredation mortality than smaller ones (Smetacek, Assmy, and Henjes, 2004).The reasoning behind this is that small cells are grazed upon more readily bysmall herbivores, while large cells are consumed by larger herbivores, and the to-tal biomass of herbivores declines with increasing herbivore size (Kiørboe, 2011).Therefore, the predation mortality rate of phytoplankton declines with increas-ing cell size (Latasa et al., 1997). Concurrently, the diffusion–limited specificnutrient uptake (Kiørboe, 1993), the average chlorophyll–specific light absorp-tion coefficient (Raven and Kubler, 2002) as well as the growth rate (Marañón,2015) also decline with increasing cell size, generating the fundamental trade–offassociated with phytoplankton size.

The second key trait is the resource acquisition mode with few central types.They range from heterotrophic osmotrophy, as found in the smallest single–celled organisms that rely on diffusive uptake of dissolved organic matter, pho-totrophy in photosynthetic organisms that depend on light and diffusive uptakeof inorganic nutrients, to mixotrophy in organisms that can both photosynthe-size and phagocytosize, and finally phagotrophy in organisms that exclusivelyfeed on other organisms.

The final key trait is defense, which can be any morphological, physiologi-cal, or behavioral characteristic of organisms that will lead to reduced predation(or virus–mediated) mortality. A defense mechanism can be constitutive (al-ways present), or can be induced in response to the presence of pathogens orpredators. For instance, a diatom shell that may reduce predation mortality isalways present (Hamm et al., 2003), but the level of silicification increases inthe presence of mesozooplankton (Pondaven et al., 2007). Similarly, toxins syn-thesized by various species of dinoflagellates and diatoms are produced at alltimes, however, their production increases with exposure to mesozooplanktonicgrazers (Selander et al., 2006; Harðardóttir et al., 2015). Because the employ-ment of defenses requires resources, the induced defense mechanisms should befavored in the environments where predation pressure varies in time and space(Karban, 2011).

The abovementioned defenses are examples of so–called active defense mech-anisms, since organisms not only utilize part of the acquired resources to syn-thesize new structures or chemical compounds, but also change the structuresor concentrations of existing compounds as a response to specific biotic cues. Asopposed to the active defenses, different types of foraging behavior in mixotrophsand heterotrophs could be viewed as a passive defense mechanism. For example,ambush feeding zooplankton that wait for motile prey to pass within their sen-sory sphere will create low fluid disturbances, making them less susceptible torheotactic predators, which too perceive their prey from generated fluid distur-bances. On the other hand, active feeders actively encounter prey by swimmingthrough the water or by creating feeding currents, making their feeding more

4 1. General introduction

efficient but at the same time riskier due to larger fluid disturbances as well ashigher encounter rates with predators (Almeda, van Someren Gréve, and Kiør-boe, 2017, 2018; van Someren Gréve, Almeda, and Kiørboe, 2017). Therefore,the trait that optimizes the trade–off will be selected by the environment, forexample, ambush feeding will be favored in environments with high abundancesof both prey and predator, while active feeding in the environments where abun-dances of prey and predator are low (Kiørboe et al., 2018).

Figure 1.2: Diagram of a trait–based description of the plankton ecosystem withthe three generic key traits, namely body size, resource acquisition mode, and de-fense, all of which are interrelated through trade–offs. In addition, the life–formtransition from unicellularity to multicellularity is distinguished. Reproducedfrom Kiørboe et al. (2018)

Besides the aforementioned generic key traits, the transition of life form fromunicellularity to multicellularity also has important implications to the ecologyof organisms as well as dynamics of the system they inhabit. For example, theimportant difference between the two life forms is that the unicellular organ-isms mainly reproduce by cell division, and as a result, their size varies only bya small factor. On the other hand, the multicellular organisms produce offspringthat are smaller than the adults and may occupy different niches as well as feedon different prey. Therefore, life and developmental histories of organisms mayhave important implications to the co–existence and ecosystem dynamics. Boththe generic key traits and life form transition are argued to capture the mostimportant aspects of the organism’s ecology, and as such provide a sufficientbasis for a trait–based description of the ecosystem (Figure 1.2) (Kiørboe et al.,2018).

1.1 Trait–based approach to ecology 5

As these key traits may be realized differently for different life forms, themeans by which they materialize needs to be understood in order to allow quan-tification of the associated trade–offs. The trade–offs are described at the levelof individuals, and since the currencies of benefits and costs are reductions inmortality rates and growth rates, respectively, the consequent fitness of the in-dividuals will be the difference between the two (Pančić and Kiørboe, 2018).The evaluation of benefits and costs thus requires a mechanistic understandingof the underlying processes, and the assessment should be based on either ex-perimental observations that quantify the pros and cons of a specific trait, orthrough fitness–optimization models.

To exemplify the mechanistic method for evaluating benefits and costs fromexperiments, consider colony formation as a defensive trait in phytoplankton(Figure 1.3). Considerable changes in size can be achieved through formationof colonies, and this morphological change from unicells to large one–, two–, orthree–dimensional colonial structures can cause difficulties for herbivores with anarrow range of prey–size preferences. These responses seem to be adaptive sincemicrozooplankton associated infochemicals in many cases induce colony forma-tion in phytoplankton, leading to reduced clearance rates of the protistan grazers(Hessen and Van Donk, 1993; Jakobsen and Tang, 2002), while mesozooplank-ton associated infochemicals suppress coloniality as the grazing pressure of largerherbivores is lower on unicells than on colonies (Long et al., 2007; Bergkvist etal., 2012; Bjærke et al., 2015).

However, this impaired colony formation due to the mesozooplankton as-sociated infochemicals will on the other hand affect diel vertical migration ofmotile dinoflagellates, because swimming velocities of unicells and short chainsare lower than those of long chains. Long chains allow higher swimming veloc-ities, which enable the organisms to harvest the nutrients at depth during thenight, and light at the surface during the day. Therefore, the ability of unicellsand short chains to perform such migrations is reduced, which may consequentlylead to decreased growth rates (Selander et al., 2011). Another cost related tocoloniality is increased sinking of non–motile phytoplankton. In some organisms,the enhancement in sinking velocities is partly countered by lower cell densities ofcolonial cells relative to single cells, facilitated by a higher content of fatty acidsin the former (Lürling, De Lange, and Van Donk, 1997). While the energeticcost of the synthesis of lipids from carbohydrates together with the energeticvalue of the deposited lipids may seem trivial under light–sufficient conditions,the growth of colonies under light–limiting conditions would be considerablyaffected as a result of these costs (Pančić and Kiørboe, 2018). In addition, colo-niality can also reduce chlorophyll–specific light absorption coefficients (Lürling,1999) and nutrient uptake rates per cell (Ploug, Stolte, and Jørgensen, 1999),which may further decrease the growth if light or nutrients are limited.

This example illustrates that the trade–offs depend on the environment; thebenefits of coloniality in terms of reduced predation mortality depend on the

6 1. General introduction

type and abundance of grazers, whereas the costs of coloniality in terms of re-duced growth rates depend on the availability of resources.

Figure 1.3: Examples of colony formation in phytoplankton. (A) One–dimensional chain of the motile dinoflagellate Alexandrium catenella. (B) Chainof the diatom Thalassionema nitzschioides. (C) Spiny chain of the green algaScenedesmus quadricauda. (D) Chain of the diatom Thalassiosira rotula. Figurecredits: A, C, Gert Hansen, University of Copenhagen; B, D, Tara Ivanochko,University of British Columbia.

1.2 Inducible defenses

1.2.1 Terrestrial ecosystemsA number of organisms from across the plant and animal kingdom respond tospecific biotic cues from pathogens and predators by rapidly developing pheno-typic defenses (Harvell, 1990; Tollrian and Harvell, 1999). Induced defenses inorganisms grant resistance to pathogens and predators, and are in turn assumedto be accompanied by metabolic costs to the organisms (Tollrian and Harvell,1999). A trade–off between defense and other fitness components is directly as-sociated with the allocation of limited resources, meaning that a prey–organismwill allocate the limiting resources to the energy intensive biological processes

1.2 Inducible defenses 7

such as growth and reproduction when predators are absent, but will invest re-sources in defense mechanisms at the cost of reduced growth in the presence ofpredators (Mole, 1994).

While humans have been exploiting some of the natural products involved inplant defense, such as morphine, nicotine, cocaine, and more for centuries andeven millennia, the phenomenon of induced defenses in plants to herbivores orpathogens was only introduced in the early 20th century (Chester, 1933) (Figure1.4). Since then, the decades of work on induced changes in terrestrial plantresistance have provided new insights into the topic by investigating the un-derlying molecular mechanisms as well as developing theoretical models of theevolution of these responses (Cipollini and Heil, 2010 and references therein).

Figure 1.4: Examples of plants whose secondary metabolites are involved inplant defenses and also exploited by humans. Left: Papaver somniferum or theopium poppy plant (top) is the principal source of opium. Opium contains alka-loid morphine (bottom) that is not only used as a pain medication but also tomake other opioids such as heroin (images by Franz Eugen Köhler, distributedunder CC–PD–Mark license; and Neurotiker, distributed under PD–self license,respectively). Center: Nicotiana tabacum or the tobacco plant (top) is commer-cially processed into tobacco that contains nicotine (bottom) (images by Fran-cisco Manuel Blanco, distributed under CC–PD–Mark license; and Neurotiker,distributed under PD–self license, respectively). Right: Erythroxylum coca orthe coca plant (top) is used to extract alkaloid cocaine (bottom) (images byFranz Eugen Köhler, distributed under CC–PD–Mark license; and Neurotiker,distributed under PD–self license, respectively).

8 1. General introduction

Induced resistance has evolved as a result of adaptation to natural environ-mental selection pressures, and as opposed to constitutive resistance, it repre-sents a substantial feature of the phenotypic plasticity of plants. In other words,the changes in resistance phenotypes are related to the corresponding changes inlevels of employed defensive traits (Cipollini et al., 2004). Some studies have in-ferred the potential benefits of induced resistance to plant fitness by observationsof a reduction in succeeding pathogen or herbivore attacks or attacker perfor-mance that followed the induction (Zavala and Baldwin, 2004). Moreover, a fewstudies have quantified a direct increase in plant fitness (growth and reproduc-tion) after employing inducible defenses that was observed at a later time (Bald-win, 1998; Zavala and Baldwin, 2004; Traw, Kniskern, and Bergelson, 2007).Possessing induced defense mechanisms may also provide a long–term benefitto organisms by reshaping evolutionary responses of herbivores to plant defensetraits. For example, a modeling study based on empirical data showed that aplant population, which possesses inducible resistance, can significantly delaythe evolution of resistance to plant defenses in herbivore populations (Gardnerand Agrawal, 2002).

However, the employment of defense mechanisms is widely anticipated tocome with fitness costs, which are assumed to be the direct cost of the produc-tion of compounds and proteins associated with defense. Next to the direct cost,the additional cost that might contribute to the overall induction–expense isthe cost of an active down–regulation of photosynthetic apparatus, as limitingresources are allocated to defense production (Lou and Baldwin, 2004; Yan etal., 2007). Seemingly, the competition for limited resources as well as resourcespace is the major force determining the evolution of form and function in terres-trial plants (Smetacek, 2001), and the ability of organisms to adapt to varyingenvironments is crucial for their evolutionary success (Price, Qvarnström, andIrwin, 2003).

1.2.2 Aquatic ecosystemsIn the oceans, the major selective pressure shaping the composition of free–floating plankton of the pelagic zone is competition for limited resources andpredation (Quintana et al., 2015). Since numerous phytoplankton species coex-ist in the same space and at the same time on few resources, and are faced withstrong top–down control (Litchman and Klausmeier, 2008; Van Donk, Ianora,and Vos, 2011), a trade–off between the strategies maximizing growth and min-imizing losses is seemingly a fundamental property determining their ecologicalniches. In adapting to deterring predators, algal species have developed a setof rather sophisticated defensive phenotypes, viz. morphological, physiological,and behavioral defense mechanisms. Yet, unlike their terrestrial counterparts,an attack on a unicellular organism by the herbivore will not lead to a reductionin fitness due to wounding, but to its certain death (Van Donk, Lürling, and

1.2 Inducible defenses 9

Lampert, 1999). What is more, herbivorous zooplankton do not only directlyinfluence algal succession via grazing, but also indirectly via interaction withphytoplankton by making nutrients available to the “survivors”, and thus havean effect on the competition among algal species (Sterner, 1989). Consequently,this brings us to the implication that the combination of predation and defensemechanisms is a potential factor that allows for the coexistence of competingspecies (Darwin, 1859).

In the late 1960’s, ecologist Robert Paine observed that removal of the carniv-orous starfish Pisaster ochraceus from the rocky intertidal habitat reduced preyspecies diversity due to intense competition for space from the mussel Mytiluscalifornianus. In the study “Food web complexity and species diversity” (1966) hestated that the local species diversity is related to the abundance of predators orso–called “keystone species”, and their efficiency to prevent single species frommonopolizing limiting resources. In the absence of such predatory species, therewill be a “winner” in the competition for the limiting resource — in this casespace —, ultimately leading to reduced species diversity in the system (Paine,1966, 1969). Similar observations were later made in three–dimensional systems,where removal of herbivorous zooplankton led to increased level of competitionamong phytoplankton species, and as a result, reduced richness of primary pro-ducers (McCauley and Briand, 1979; Leibold et al., 2017). As such, predationevidently plays a major role in the maintenance of high algal diversity by en-abling the temporal and spatial coexistence of species that would otherwise likelybe eliminated by competitive exclusion.

Figure 1.5: Diagram of the general structure of the “killing the winner” prin-ciple, with a population of competition specialists and a population of defensespecialists competing for a shared limited resource. The limiting resource can beavailable directly to the two competing populations (free form), or is being seizedby the competitors and predators/parasites in the form of biomass. Reproducedfrom Winter et al. (2010)

10 1. General introduction

Similar to the concept of “keystone species” is the idea expressed in “killingthe winner” principle (Figure 1.5). Here, the assumption is that two prey pop-ulations, namely a competition specialist and a defense specialist, compete forshared limiting requisites. This concept explains that if the selective loss of thecompetition specialist due to predation prevents this population from seizing thelimiting resources, the resources will become available to the defense strategistas well (Thingstad et al., 2005; Winter et al., 2010). However, the promotion ofdiversity through predation and the subsequent employment of defense mecha-nisms requires trade–offs, meaning that the advantage of a defense mechanismmust come at a cost (Thingstad et al., 2014; Våge et al., 2014). In other words,if there was no cost to grazer resistance then all species would evolve towards astate of equal defense, and community diversity would no longer be promotedby predation.

1.3 ObjectivesGiven the importance of predation and defense mechanisms in relation to speciesdiversity, a deep understanding of defenses and associated trade–offs in promot-ing and sustaining phytoplankton diversity in aquatic environments is required.This thesis is based on three main manuscripts with the overarching aim ofidentifying defense mechanisms in phytoplankton and quantifying the associ-ated trade–offs. The following aspects are explored in particular (Figure 1.6):

1 The overview of suggested phytoplankton defensive traits and quantifica-tion of the trade–offs from available experimental evidence and theoreticalconsiderations (Chapter 2)

2 The protective role of silicified walls in diatoms and quantification of thebenefits and associated costs from experimental work (Chapter 3). Wehypothesize that grazing pressure exerted by adult copepods and naupliiwill decrease with increasing thickness of the walls, as they need to breakthe shells to access the contents, while the variability in silica content willbe inconsequential to dinoflagellates that engulf their prey. Moreover, wehypothesize that deposition of silica is not constant, and will decrease withincreasing growth of diatoms

3 The defensive role of toxin production in dinoflagellates and quantificationof the associated costs from theoretical work (Chapter 4). We hypothesizethat costs of toxin production will become significant in nitrogen–limitedenvironments and when grazing pressure is high, which will subsequentlylead to a reduction in cell division rates. However, we hypothesize that theinvestment in toxin production will pay–off because the predation mortal-ity on defended cells will decrease relative to undefended cells.

1.3 Objectives 11

In addition to the manuscripts, this thesis also includes a summary of the mainfindings as well as concluding remarks (Chapter 5).

Figure 1.6: Schematic representation of the main objectives of the PhD project,showing the relation between Project 1 with the aim to identify defensive traitsin phytoplankton and search for associated trade–offs, and Projects 2 and 3 withthe aim to quantify benefits and costs of selected defensive traits.

12 1. General introduction

References1. ALMEDA, R., VAN SOMEREN GRÉVE, H. and KIØRBOE, T. (2017) Be-

havior is a major determinant of predation risk in zooplankton. Ecosphere 8,e01668.

2. ALMEDA, R., VAN SOMEREN GRÉVE, H. and KIØRBOE, T. (2018) Preyperception mechanism determines maximum clearance rates of planktoniccopepods. Limnology and Oceanography 63, 2695–2707.

3. ANDERSEN, K.H., BERGE, T., GONÇALVES, R.J., HARTVIG, M., HEU-SCHELE, J., HYLANDER, S., JACOBSEN, N.S., LINDEMANN, C., MAR-TENS, E.A., NEUHEIMER, A.B., OLSSON, K., PALACZ, A., PROWE,A.E.F., SAINMONT, J., TRAVING, S.J., ET AL. (2016) Characteristic sizesof life in the oceans, from bacteria to whales. Annual Review of Marine Sci-ence 8, 217–241.

4. BALDWIN, I.T. (1998) Jasmonate-induced responses are costly but bene-fit plants under attack in native populations. Proceedings of the NationalAcademy of Sciences 95, 8113–8118.

5. BERGKVIST, J., THOR, P., JAKOBSEN, H.H., WANGBERG, S.A. andSELANDER, E. (2012) Grazer-induced chain length plasticity reduces graz-ing risk in a marine diatom. Limnology and Oceanography 57, 318–324.

6. BJÆRKE, O., JONSSON, P.R., ALAM, A. and SELANDER, E. (2015)Is chain length in phytoplankton regulated to evade predation? Journal ofPlankton Research 0, 1–10.

7. CHESTER, K.S. (1933) The problem of acquired physiological immunity inplants. The Quarterly Review of Biology 8, 129–154.

8. CIPOLLINI, D., ENRIGHT, S., TRAW, M.B. and BERGELSON, J. (2004)Salicylic acid inhibits jasmonic acid-induced resistance of Arabidopsis thaliana-text to Spodoptera exiguatext. Molecular Ecology 13, 1643–1653.

9. CIPOLLINI, D. and HEIL, M. (2010) Costs and benefits of induced resis-tance to herbivores and pathogens in plants. CAB Reviews: Perspectives inAgriculture, Veterinary Science, Nutrition and Natural Resources 5, 1–25.

10. DARWIN, C.R. (1859) On the origin of species by means of natural selection,or the preservation of favoured races in the struggle for life. In On the Originof Species, 5th edition. John Murray, Albemarle Street, London.

11. VAN DONK, E., IANORA, A. and VOS, M. (2011) Induced defences inmarine and freshwater phytoplankton: a review. Hydrobiologia 668, 3–19.

References 13

12. VAN DONK, E., LÜRLING, M. and LAMPERT, W. (1999) Consumer-induced changes in phytoplankton: inducibility, costs, benefits, and the im-pact on grazers. In The ecology and evolution of inducible defenses (eds R.TOLLRIAN and C.D. HARVELL), pp. 89–103. Princeton University Press.

13. GARDNER, S.N. and AGRAWAL, A.A. (2002) Induced plant defence andthe evolution of counter-defences in herbivores. Evolutionary Ecology Re-search 4, 1131–1151.

14. HAMM, C.E., MERKEL, R., SPRINGER, O., JURKOJC, P., MAIER, C.,PRECHTEL, K. and SMETACEK, V. (2003) Architecture and material prop-erties of diatom shells provide effective mechanical protection. Nature 421,841–843.

15. HARÐARDÓTTIR, S., PANČIĆ, M., TAMMILEHTO, A., KROCK, B.,MØLLER, E.F., NIELSEN, T.G. and LUNDHOLM, N. (2015) Dangerousrelations in the Arctic marine food web: interactions between toxin produc-ing Pseudo-nitzschia diatoms and Calanus copepodites. Marine drugs 13,3809–3835.

16. HARVELL, C.D. (1990) The ecology and evolution of inducible defenses. TheQuarterly Review of Biology 65, 323–340.

17. HESSEN, D.O. and VAN DONK, E. (1993) Morphological-changes in Scene-desmus induced by substances released from Daphnia. Archiv Fur Hydrobi-ologie 127, 129–140.

18. JAKOBSEN, H.H. and TANG, K.W. (2002) Effects of protozoan grazing oncolony formation in Phaeocystis globosa (Prymnesiophyceae) and the poten-tial costs and benefits. Aquatic Microbial Ecology 27, 261–273.

19. KARBAN, R. (2011) The ecology and evolution of induced resistance againstherbivores. Functional Ecology 25, 339–347.

20. KIØRBOE, T. (1993) Turbulence, phytoplankton cell size, and the structureof pelagic food webs. Advances in Marine Biology 29, 1–72.

21. KIØRBOE, T. (2011) How zooplankton feed: mechanisms, traits and trade-offs. Biological Reviews 86, 311–339.

22. KIØRBOE, T., VISSER, A. and ANDERSEN, K.H. (2018) A trait-basedapproach to ocean ecology. ICES Journal of Marine Science 75, 1849–1863.

23. LATASA, M., LANDRY, M.R., SCHLÜTER, L. and BIDIGARE, R.R. (1997)Pigment specific growth and grazing rates of phytoplankton in the centralequatorial Pacific. Limnology and Oceanography 42, 289–298.

14 1. General introduction

24. LEIBOLD, M.A., HALL, S.R., SMITH, V.H. and LYTLE, D.A. (2017) Her-bivory enhances the diversity of primary producers in pond ecosystems. Ecol-ogy 98, 48–56.

25. LITCHMAN, E. and KLAUSMEIER, C.A. (2008) Trait-based communityecology of phytoplankton. Annual Review of Ecology, Evolution, and Sys-tematics 39, 615–639.

26. LITCHMAN, E., OHMAN, M.D. and KIØRBOE, T. (2013) Trait-based ap-proaches to zooplankton communities. Journal of Plankton Research 35, 473–484.

27. LONG, J.D., SMALLEY, G.W., BARSBY, T., ANDERSON, J.T. and HAY,M.E. (2007) Chemical cues induce consumer-specific defenses in a bloom-forming marine phytoplankton. Proceedings of the National Academy of Sci-ences of the United States of America 104, 10512–10517.

28. LOU, Y. and BALDWIN, I.T. (2004) Nitrogen supply influences herbivore-induced direct and indirect defenses and transcriptional responses in Nico-tiana attenuata. Plant physiology 135, 496–506.

29. LÜRLING, M. (1999) Grazer-induced coenobial formation in clonal culturesof Scenendesmus obliquus (Chlorococcales, Chlorophyceae). Journal of Phy-cology 35, 19–23.

30. LÜRLING, M., DE LANGE, H.J. and VAN DONK, E. (1997) Changes infood quality of the green alga Scenedesmus induced by Daphnia infochemicals:biochemical composition and morphology. Freshwater Biology 38, 619–628.

31. MARAÑÓN, E. (2015) Cell size as a key determinant of phytoplanktonmetabolism and community structure. Annual Review of Marine Science 7,241–264.

32. MCCAULEY, E. and BRIAND, F. (1979) Zooplankton grazing and phyto-plankton species richness: field tests of the predation hypothesis. Limnologyand Oceanography 24, 243–252.

33. MOLE, S. (1994) Trade-offs and constraints in plant-herbivore defense theory:a life-history perspective. OIKOS 71, 3–12.

34. PAINE, R.T. (1966) Food web complexity and species diversity. The Ameri-can Naturalist 100, 65–75.

35. PAINE, R.T. (1969) A note on trophic complexity and community stability.The American Naturalist 103, 91–93.

36. PANČIĆ, M. and KIØRBOE, T. (2018) Phytoplankton defence mechanisms:traits and trade-offs. Biological Reviews 93, 1269–1303.

References 15

37. PLOUG, H., STOLTE, W. and JØRGENSEN, B.B. (1999) Diffusive bound-ary layers of the colony-forming plankton alga Phaeocystis sp.— implicationsfor nutrient uptake and cellular growth. Limnology and Oceanography 44,1959–1967.

38. PONDAVEN, P., GALLINARI, M., CHOLLET, S., BUCCIARELLI, E.,SARTHOU, G., SCHULTES, S. and JEAN, F. (2007) Grazing-induced changesin cell wall silicification in a marine diatom. Protist 158, 21–28.

39. PRICE, T.D., QVARNSTRÖM, A. and IRWIN, D.E. (2003) The role ofphenotypic plasticity in driving genetic evolution. Proceedings of the RoyalSociety of London, Series B 270, 1433–1440.

40. QUINTANA, X.D., ARIM, M., BADOSA, A., BLANCO, J.M., BOIX, D.,BRUCET, S., COMPTE, J., EGOZCUE, J.J., DE EYTO, E., GAEDKE, U.,GASCÓN, S., DE SOLÁ, L.G., IRVINE, K., JEPPESEN, E., LAURIDSEN,T.L., ET AL. (2015) Predation and competition effects on the size diversityof aquatic communities. Aquatic Sciences 77, 45–57.

41. RAVEN, J.A. and KUBLER, J.E. (2002) The scaling of metabolic rate withthe size of algae. Journal of Phycology 38, 11–16.

42. SELANDER, E., JAKOBSEN, H.H., LOMBARD, F. and KIØRBOE, T.(2011) Grazer cues induce stealth behavior in marine dinoflagellates. Pro-ceedings of the National Academy of Sciences of the United States of America108, 4030–4034.

43. SELANDER, E., THOR, P., TOTH, G. and PAVIA, H. (2006) Copepodsinduce paralytic shellfish toxin production in marine dinoflagellates. Proceed-ings of the Royal Society B: Biological Sciences 273, 1673–1680.

44. SMETACEK, V. (2001) A watery arms race. Nature 411, 745–745.

45. SMETACEK, V., ASSMY, P. and HENJES, J. (2004) The role of grazing instructuring Southern Ocean pelagic ecosystems and biogeochemical cycles.Antarctic Science 16, 541–558.

46. VAN SOMEREN GRÉVE, H., ALMEDA, R. and KIØRBOE, T. (2017)Motile behavior and predation risk in planktonic copepods. Limnology andOceanography 62, 1810–1824.

47. STERNER, R.W. (1989) The role of grazers in phytoplankton succession. InPlankton ecology (ed U. SOMMER), pp. 107–170. Springer, Berlin, Heidel-berg.

48. THINGSTAD, T.F., ØVREÅS, L., EGGE, J.K., LØVDAL, T. and HEL-DAL, M. (2005) Use of non-limiting substrates to increase size; a genericstrategy to simultaneously optimize uptake and minimize predation in pelagicosmotrophs? Ecology Letters 8, 675–682.

16 1. General introduction

49. THINGSTAD, T.F., VÅGE, S., STORESUND, J.E., SANDAA, R.-A. andGISKE, J. (2014) A theoretical analysis of how strain-specific viruses cancontrol microbial species diversity. Proceedings of the National Academy ofSciences 111, 7813–7818.

50. TOLLRIAN, R. and HARVELL, C.D. (1999) The ecology and evolution ofinducible defenses. Princeton University Press.

51. TRAW, M.B., KNISKERN, J.M. and BERGELSON, J. (2007) SAR increasesfitness of Arabidopsis thaliana in the presence of natural bacterial pathogens.Evolution 61, 2444–2449.

52. VÅGE, S., STORESUND, J.E., GISKE, J. and THINGSTAD, T.F. (2014)Optimal defense strategies in an idealized microbial food web under trade-offbetween competition and defense. PLoS ONE 9.

53. WINTER, C., BOUVIER, T., WEINBAUER, M.G. and THINGSTAD, T.F.(2010) Trade-offs between competition and defense specialists among uni-cellular planktonic organisms: the ‘Killing the Winner’ hypothesis revisited.Microbiology and Molecular Biology Reviews 74, 42–57.

54. YAN, Y., STOLZ, S., CHÉTELAT, A., REYMOND, P., PAGNI, M., DUBUGNON,L. and FARMER, E.E. (2007) A downstream mediator in the growth repres-sion limb of the jasmonate pathway. The Plant Cell 19, 2470–2483.

55. ZAVALA, J.A. and BALDWIN, I.T. (2004) Fitness benefits of trypsin pro-teinase inhibitor expression in Nicotiana attenuata are greater than their costswhen plants are attacked. BMC Ecology 4, 1–15.

Chapter 2

Phytoplankton defencemechanisms: traits andtrade–offs

Panc̆ić, M. and Kiørboe, T. (2018) Biological Reviews, 93, 1269–1303

Due to copyright, the original paper is not included in this version of the thesis.

18 2. Manuscript I: Phytoplankton defence mechanisms

Chapter 3

Silicified cell walls as a defensivetrait in diatoms

Panc̆ić, M., Torres, R.R., Almeda, R., and Kiørboe, T. (2019) Proceedings of theRoyal Society B: Biological Sciences, 286

Due to copyright, the original paper is not included in this version of the thesis.

20 3. Manuscript II: Silicified cell walls in diatoms

Chapter 4

The cost of toxin production inplankton: the case of PSTproducing dinoflagellates

Chakraborty, S., Panc̆ić, M., Andersen, K.H., and Kiørboe, T. (2018) ISMEJournal, 1–12

Due to copyright, the original paper is not included in this version of the thesis.

22 4. Manuscript III: Toxin production in dinoflagellates

Chapter 5

Synopsis

24 5. Synopsis

5.1 Identification of defensive traits and associ-ated trade–offs

Theoretical and experimental studies have demonstrated that, at steady state,the number of coexisting species in phytoplankton communities cannot exceedthe number of limiting resources unless additional mechanisms are involved. Onesuch mechanism is predation, which may increase species diversity by facilitatingthe evolution of defense mechanisms, provided that the employment of defenseshas an associated cost. However, given the importance of predation and defensemechanisms in relation to species diversity, it appears that the function of manyproposed defense mechanisms remains elusive and the costs and benefits are of-ten unquantified. Since many proposed defensive traits are inducible, meaningthat their intensity increases under the grazing pressure, this suggests that theproposed traits indeed have a defensive function. It also suggests the presenceof a cost — if there was no cost, why would phytoplankton only deploy defenseswhen needed?

In response to grazing pressure, phytoplankton have evolved various mor-phological defensive traits. While size in itself offers partial protection againstgrazing, alterations in size can be achieved through colony formation that canaffect grazing of herbivores with a narrow range of prey–size preferences. How-ever, such size variations are only beneficial against certain grazers, and canat the same time promote grazing by others. Associated costs may for exam-ple include increased sinking velocities, decreased nutrient uptake rates per cell,or reduced chlorophyll–specific light absorption coefficients due to competitionamong colonial cells, and these costs will depend strongly on the environmentalconditions. Next to coloniality, the ecological role of the calcite armor in coccol-ithophores and silica shells in diatoms has long been proposed to act as a defenseagainst predation. However, only recently has experimental evidence supportingthe defensive role of the calcite armor been provided, but the evidence support-ing mechanical protection against grazing by the shell remains limited (furtherexplored in Chapter 3). While both calcification and silicification seem to berather inexpensive processes in terms of energy expenditure, the dependency ofcoccolithophores and diatoms on the ambient concentrations of inorganic cal-cium and silicon represents a cost that can only be evaluated in the context ofprevailing environmental conditions.

In addition to the morphological defenses, several physiological traits suchas production of chemical substances, bioluminescence, and modified nutritionalvalue of phytoplankton, are suggested to represent physiological defense mech-anisms. While many of these defensive traits are constitutive, their increasedproduction can be activated by a range of predation cues, by grazing, or by mo-tion. A variety of chemical compounds with proposed defensive functions havevery variable effects on grazers: no effect, deterred grazing, and toxic effect. Thefirst type of response, where a grazer can feed on toxic algae without obvious

5.2 Defensive value and cost of silica walls 25

adverse effects, can be interpreted as a result of evolutionary history betweenthe predator and prey populations. However, it may also be due to the fact thatthe definition of ‘toxic algae’ rather refers to their toxic effects on humans thanon zooplankton. The second type of response, commonly referred to as deterredgrazing, enables the grazer to avoid potentially dangerous prey prior to ingest-ing harmful amounts. Here, the grazer has developed the ability to recognizeand avoid toxic prey either via sensing the prey or via capturing, tasting, andthen rejecting the prey. Either way, this will result in reduced grazing pressureon prey organisms that consequently gain a competitive advantage. Finally, thethird type of response, where the ingestion of toxic prey is required to affectthe grazer, cannot be considered a defensive trait since reduced grazing due totoxic effects on the grazer is equally beneficial to competitors, and thus doesnot provide the producer any competitive advantage. While induced productionof chemicals is in some cases advantageous to phytoplankton, the costs of syn-thesizing chemical compounds seem to be more difficult to assess. Based on theavailable evidence it appears that the production is costless, since induced anduninduced organisms, or toxic and non–toxic strains of the same species havesimilar growth rates, however, the production may depend on nutrient condi-tions. Since many of the substances produced by phytoplankton are alkaloidsrich in nitrogen, the costs of defense may be unmeasurably low when resourcesare plentiful but could become significant in nitrogen–limited environments (fur-ther explored in Chapter 4).

Finally, behavioral defenses such as resting stages and motility may functionas grazer–avoidance adaptations, and for these traits the assessment of bene-fits and costs is rather straightforward; delayed excystment due to unfavorablebiotic conditions will lead to lost opportunities for resource acquisition and pro-liferation, but will simultaneously reduce grazing mortality. Motile cells that arecaptured by grazers can break the contact with them by jumping or swimmingaway, thus reducing grazing mortality. However, fluid disturbances generated bythese swimming cells can be detected by rheotactic grazers, leading to increasedencounter rates and subsequently increased mortality.

5.2 Defensive value of silica walls in diatomsand the associated trade–offs

Diatoms are the major contributors to phytoplankton biomass because theirmortality rates are generally lower than those of other groups with similar growthrates. But what makes diatoms unique are their silicified cell walls, whose me-chanical strength relative to density is the highest among any known biologicalmaterial. As discussed in section 5.1, silicified cell walls may have evolved as amechanical protection against grazing, however, the direct evidence of defense islimited. While the silica shell may be able to withstand the forces exerted by the

26 5. Synopsis

crushing mechanisms of copepods and euphausiids, same crushing resistance isprobably of little consequence for protistan grazers whose ability to utilize preyis governed mainly by the size and shape of the prey. The anticipated benefits ofsilica walls must be traded off against the costs. Deposition of dissolved silica indiatoms depends on their growth, and since many diatoms do not store sufficientamounts of silicon for the formation of a new valve, they must harvest most ofthe required amounts immediately before the cell division. Thus, high growthrates under non–limiting ambient silicon concentrations will lead to a reducedperiod available for silicon uptake, resulting in lighter silicification of the walls.Therefore, if increased level of silicification in response to grazer cues requiresreduced growth, this may represent a direct cost of defense. Other costs appearto be of less importance, and include biochemical costs related to the uptakeand deposition of silicon, as well as the costs associated with the potential lossesdue to increased sinking velocities of heavy silicified cells.

In Chapter 3, we explored the relationship between the level of silicificationin diatoms and predation mortality exerted by micro– and mesozooplanktonicgrazers, and whether the survival of diatoms during the gut passage of copepodsis related to the silica content of the cells. Furthermore, we explored how themechanical properties of silica walls vary with cell size and degree of silicificationin the context of copepod predation. Finally, we explored the trade–off function,i.e. the relationship between the costs and benefits of the siliceous wall.

Since the level of silicification depends on the growth rates of diatoms, whichin turn are regulated by external factors, we used this property to manipulatesilica contents in cells by using different light intensities. As a result, we pro-duced diatoms with up to a 6–fold variation in silica contents. Our empiricaldata revealed that grazing of mesozooplankton (adult copepods and nauplii) isstrongly dependent on and near inversely proportional to the level of silicificationboth within and among diatom species (slope equals –0.79). This demonstratesthat the protective value of a shell increases with increasing silica content, whichis consistent with previous studies demonstrating that the mechanical strengthof shells also increases with increasing silica content. It is also consistent withthe scaling arguments explored in this study, on how the silica content mustincrease with size to be able to resist the same force before cracking or buckling.We showed that the fracture strength and the critical buckling load of a diatomshell increase approximately with the cell wall thickness raised to the power of alittle less than 3, which is identical to the observed size scaling of silica contentin diatoms. Furthermore, this scaling also revealed that the advantage of furtherthickening decreases as the silica content increases, because the unbreakabilityof cells rapidly increases with the silica content. Moreover, our data also showedthat the deposition rate of silica is not constant but decreases with increasinggrowth rates, meaning that at steady state, silica content should vary with thegrowth rate raised to a power of b that is smaller than –1 (slope equals –0.54 asestimated from the literature). This in turn implies that the grazing mortality

5.3 Defensive value and cost of toxin production 27

increases with the growth rate raised to the power of 0.43, suggesting that withdeclining growth, the reduction in predation mortality will accelerate. Whilethe silica wall reduces copepod grazing mortality, our empirical data furtherrevealed that the wall provides no protection against microzooplanktonic graz-ers that engulf their prey. Moreover, the data also showed that only a minutefraction of the ingested cells survive the gut passage of copepods, and that thesurvival is independent of the degree of silicification.

5.3 Defensive value of toxin production in di-noflagellates and the associated costs

Several phytoplankton species produce chemical compounds that are toxic tohuman consumers, therefore we consider these organisms ‘toxic algae’, but theevolution and the functional role of such chemicals remain unclear. Since sometoxic algae form blooms that are potentially promoted by toxicity and subse-quent deterrent effects on grazers, a defensive role of toxin production is oftenanticipated. This interpretation is supported by observations that the toxin pro-duction increases in the presence of grazers, which in some cases appears to beadvantageous to the algae as it leads to reduced predation mortality. Moreover,according to the optimal defense theory, which predicts that the inducibility ofdefenses should evolve only when the defense is costly and variable in time, ex-periments have consistently failed to demonstrate such costs (in terms of growthrates) in toxic phytoplankton. As discussed in section 5.1, many of these sub-stances are alkaloids rich in nitrogen, therefore the costs of toxin production maybe unmeasurably low when resources are abundant but could become significantin nitrogen–limited environments. Consequently, the cost may become manifesteither as a reduction in growth rates when nitrogen is limiting, since part of theavailable nitrogen is used for toxin production; or as a cessation of the toxinproduction in favor of higher growth rates and consequent loss of defense; or acombination of the two.

In Chapter 4, we explored — by means of a simple fitness optimizationmodel — the costs of chemical defense mechanisms in phytoplankton, using theparalytic shellfish toxin (PST) producing dinoflagellate Alexandrium spp. as anexample. We considered energy and material costs of PST production, their de-pendency on ambient concentrations of inorganic nutrients, and the reductionin predation mortality achieved via defense investments. Furthermore, we alsoexamined conditions under which the toxin production provides highest benefitto the producers.

In our resource allocation optimization model, the phytoplankton divisionrate depends on the uptake of inorganic nutrients as well as on the respiratoryexpenses, and subtracting the predation mortality from the division rate yieldsthe population growth rate. Moreover, the cell has the ability to optimize its

28 5. Synopsis

growth rate by regulating its resource allocation to the toxin production in away that maximizes its fitness, and the resulting population growth rate is ameasure of the phytoplankton fitness. The empirical data from the literaturewere used to calibrate the model parameters as well as to estimate the param-eter values related to the cell division rate and predation mortality, and thestoichiometry of PSTs and the biochemistry of PST synthesis were consideredto estimate the parameters related to the cost of toxin production. The modelsuccessfully reproduced all known features of PST production, such as that PSTproducing cells become less toxic in nitrogen–limited environments, that toxincontents are high in exponentially growing cells in the environments where ni-trogen and phosphorus are in balance, and that the cells become most toxic inphosphorus– or light–limited ambients. Finally, our model further conforms tothe experimental observations that the toxin production increases in the pres-ence of grazers, and that the cost of toxin production is negligible when nitrogenor light are plentiful. However, when nitrogen or light is limiting cell growth andwhen grazers are abundant, the model predicted that the costs associated withthe toxin production may be substantial and could lead to more than a 20%reduction in cell division rates. These costs are two–fold, and include metaboliccosts of the toxin biosynthesis as well as the costs related to the materials in-vested in the toxin. Nonetheless, the investment in the toxin production pays–offsince defended cells experience reduced grazing mortality, which subsequentlyleads to the net growth rates that are twice as high as of undefended cells.

5.4 Concluding remarksWhile many morphological, physiological, and behavioral defensive traits in phy-toplankton are rather well described, in most cases the associated costs remainunquantified because even the benefits of many deployed defenses are oftenundocumented, for example, the proposed protective role of siliceous wall in di-atoms. One reason for the lack of experimental evidence for the costs is that theyoften realize only under natural conditions, or when the deployment of defensemechanisms depends on the availability of specific resources. For that reason,a mechanistic understanding of the underlying processes is required, which to-gether with simple fitness optimization models may allow quantification of thecosts under specific environmental conditions. Another reason for the absenceof cost assessments is that some costs may become evident only when organismsare resource limited. However, ambient conditions to which experimental organ-isms are most often exposed to are resource–sufficient, and as such can supportresource demanding growth, reproduction, and defense, as there is no conflictin resource allocation between these functions. Therefore, in order to isolate rel-evant resources that should be limited in future experimental and theoreticalstudies, an understanding of the processes involved in the defense is required;for example, reduced growth of toxin producing organism may be found only

5.4 Concluding remarks 29

under nitrogen–limiting conditions.Although we now understand the defensive role of the siliceous wall, whose

silica contents are a ‘passive’ function of the diatom growth rates, it is still un-clear whether shell thickening can also be induced by biotic factors. Therefore, itremains to be demonstrated whether and to what extent the increased degree ofsilicification requires reduced growth rates, or whether the thickening is a resultof the increased rate of silica deposition in the wall. Hence, more elaborate ex-perimental studies are required to verify this and to (re–)quantify the trade–offfunction.

Similarly, the understanding of the processes involved in the PST productionin dinoflagellates led to the assessment of the costs associated with the toxin syn-thesis that are not only strongly dependent on the abundance of grazers but alsoon the resource availability. However, experimental studies on costs and bene-fits of toxin production are required to empirically test our model predictionsin which, for example, toxin production rates and growth rates are measuredunder steady state sufficient– and deficient–nitrogen concentrations.

Finally, the findings from this PhD research opened a number of new av-enues that could be explored in future studies. For example, while the benefitsof size alterations and the consequential physiological changes associated withcoloniality are rather well understood, it still remains to be demonstrated howthese changes affect growth rates in resource–limiting environments. Moreover,the costs related to the production of alkaloid substances, such as domoic acid indiatoms and microcystin in cyanobacteria, remain largely unquantified. Follow-ing a similar line of mechanistic reasoning as provided in this PhD work, thesecosts could eventually be measured in nitrogen–limited environments. Last butnot least, if silicified cell walls in diatoms provide protection against copepodgrazing, it still remains to be determined what defensive traits may reduce pre-dation from microzooplankton grazing. It is well understood that by formingchains diatoms escape grazing of microzooplankton that engulf their prey, how-ever, such increases in size are likely inconsequential to pallium feeding grazers.In this case, perhaps formation of spines or thorns that may pierce the palliumwould be a good mechanism, but that still remains to be determined.

Technical University of Denmark

DTU AquaKemitorvetDK-2800 Kgs. Lyngby

www.aqua.dtu.dk

In this project, we gained significant knowledge on the mechanism that promotes coexistence and

species diversity in phytoplankton – predation and the consequent evolution of defense mecha-

nisms. We provided an overview of defense mechanisms in phytoplankton and associated trade-

offs. In search for defense trade-offs, we identified which traits lack experimental evidence sup-

porting their defensive function, and proposed ways to assess the costs in future experimental or

theoretical studies. We experimentally demonstrated the defensive function of silicified cell walls,

whose effectiveness in reducing predation mortality may be one of the reasons for the success of

the single most important group of phytoplankton in the ocean – diatoms. Another defensive trait

observed in phytoplankton is production of toxins, for which – by means of a simple fitness opti-

mization model – we estimated the associated costs that appear to be strongly dependent on the

environmental conditions. In conclusion, quantification of the trade-offs is key to understanding

biological diversity, and may help us describe complex aquatic ecosystems as well as assess the

effects environmental changes may have on the structure of aquatic communities.