U N I V E R S I T Y O F C O P E N H A G E N F A C U L T Y O F S C I E N C E

Johannes Wilhelm Goessling

Marine Biological Section Department of Biology

Biophotonics of Linking frustule structure to photobiology

Biophotonics of diatoms Linking frustule structure to photobiology

Biophotonics of diatoms

Linking frustule structure to photobiology Johannes W. Goessling Ph.D. thesis

Advised by Michael Kühl

This PhD thesis has been submitted to the PhD school of Sciences, University of Copenhagen.

Herewith I declare that this thesis and the work presented in it are my own and have been generated by me as the result of my own original research.

Johannes W. Goessling Copenhagen, 2017

Author: Johannes Wilhelm Goessling, born in Hagen 1984, Germany

Title: Biophotonics of diatoms Subtitle: Linking frustule structure to photobiology

Academic advisor: Michael Kühl, PhD Full Professor Marine Biological Section University of Copenhagen, Denmark

Opponents: Bruno Jesus, PhD Associate Professor Remote Sensing and Benthic Ecology team - MMS University of Nantes

Mark Hildebrand, Dr. Full Professor Scripps Institution of Oceanography University of California, San Diego

Faculty opponent: Nina Lundholm, PhD Associate Professor, Curator Natural History Museum of Denmark University of Copenhagen

For my family

Front cover:

Silicate valves of different species generate structural colors in dark field microscopy. Valves are fixed on a permanent microscope slide in resin (“Diatom Cubed ®”; RI>1.70). The micrograph was recorded at an angle of 25° incident white light. Collection, decoration and species identification were conducted by Stefano Barone (www.diatomshop.com).

Species from top left to bottom right:

Surilella testudo ; Surirella utahensis ; Cymbella mexicana ; Diploneis smithii ; Pinnularia viridis ; Navicula oblonga ; Epithemia argus ; Anomoconeis polygramma ; Endyctia oceanica ; Arachnoidiscus sp. ; Triceratium castellatum ; Triceratium favus .

” I think it remarkable that these regular parallelograms are all of the same size, the longest side not exceeds 1/3 of an hairs breadth, and that the length is just double the breadth, the squares being visibly made up of two parallelograms joyn'd longwise. They seem very thin and the texture of every one is nearly the same... I took these branches at first for Salts, but finding them always of the same size, and that there was no sensible increase of their bulk while they continued in the Water, that after they had lain a day or two dry on a Glass Plate they alter'd not their Figure, and upon the addition of new Water (warm or cold) they had still the same appearance and cohesion, and that their adherence (tho touching only in the angular points) was so firm and rigid, that all mov'd together, and kept the same position in respect of one another, however agitated by the Water; these considerations, I say, persuade me, that they may be rather Plants than Salts, but they being so very minute that no judgment can be made of 'em but by the Eye, I shall not determine anything positively.”

The first certain record of a diatom communicated by an anonymous English country noblemen to the Philosophical Transactions of the Royal Society of London in 1703 1. He probably described the species Tabellaria floculosa 2 (micrograph on the right side).

The light microscopic image was downloaded from http://www.lenaturaliste.net/forum/viewtopic.php?t=9934 (10.05.2017).

Abstract

Diatoms are unicellular microalgae present in all aquatic environments on earth. Due to their high photosynthetic productivity and abundance, diatoms are main components of aquatic food webs and among the main contributors of global photosynthetic carbon fixation. A unique feature of diatoms is the encasement of the cell in a silicate frustule compounded of two valves and corresponding girdle bands. Photonic structures in the frustule, i.e. pores and chambers on the micro- to nanoscale, interact with electromagnetic radiation in the visible spectrum of light. It has therefore been proposed that the optical properties of frustules could mediate efficient diatom photosynthesis; however, due to lack of optical data of frustules in water and live cells, such links remained purely speculative. The current thesis investigates the potential implications of frustule biophotonics and photobiology also in living diatom cells.

We could show that the valve of the centric diatom species Coscinodiscus granii guides light in the horizontal plane, and redistributes photosynthetically productive radiation over the entire cell. Optical coupling of chloroplasts to the evanescent field of the valve induced photosynthesis in areas that were not directly illuminated with a laser beam focused onto the valve surface. The C. granii valve also strongly interacted with blue light which was scattered onto the chloroplasts inside the cell. However, this observation was restricted to an acute angle of light incidence in C. granii , while valves of some benthic diatom species showed similar phenomena at more obtuse-angled incidence. We hypothesized that the angle dependency of such phenomenon can be explained by differences in the light climate where these species live, i.e. sunlight might be incident at more acute angles in habitats close to the water surface, while the light climate is more diffusive inside sediments. Scattering of blue light by the frustules of pennate diatom species compensated for strong attenuation of such shorter wavelengths inside muddy sediment. Hence, variation of frustule photonic structures and optical properties between the species could have shaped niche differentiation of diatoms in habitats with different light climates. Data presented in this thesis also suggest that the different components of the frustule have various optical implications upon the living cell, i.e. we observed that photonic structures in girdle bands of the C. granii frustule have different optical properties than valves, such as iridescent coloration as a function of light incidence. We conclude that the different biophotonic properties of frustules contribute to the high photosynthetic flexibility of diatoms to various light conditions. We speculate that the photonic structures of frustules thereby enabled diatoms to inhabit environments with different light climates, and have hence influenced species diversification and evolution of diatoms.

Resumé

Kiselalger er unicellulære mikroalger der lever i alle vandmiljøer, er hovedkomponenter i vandmiljøer og nogle af de vigtigste bidragsydere til global fotosyntetisk kulstoffiksering. Photoniske krystalstrukturer kiselalgers i silikat frustule (vægstruktur) interagerer med elektromagnetisk stråling i den synlige den af lysets spektrum, men det er ikke klart hvornår og hvordan de optiske egenskaber hos frustulen påvirker den levende organisme. I denne afhandling er disse potentielle relationer undersøgt med bio-optiske og foto- fysiologiske metoder.

Det viste sig i vores undersøgelser at strukturer i frustulen fungerer som en fotonisk waveguide, som kan overføre lysenergi til kloroplaster ved hjælp af optisk kobling. Dette sås for eksempel i at lys, som var fokuseret på et lille område af frustulen med en laser, blev jævnt fordelt i hele cellen gennem effektiv waveguiding, således at fotosyntesen også blev induceret ide kloroplaster, som ikke var direkte oplyst af det indfaldende laserlys. Ved lysdiffraktion på de store porer på indersiden af frustule blev det fotosyntetisk mest aktive blå lys jævnt fordelt over kloroplasterne. I centriske, levende kiselalger blev denne virkning kun observeret når hvidt lys faldt ind i en spids vinkel. Dette kan være af stor betydning ved vandoverfladen, hvor sollyset ofte kommer ind vinkelret og fokuseret. I bentiske, sediment-levende kiselalger er den samme virkning kun synlig ved fladere indfaldsvinkler. Den således inducerede diffraktion af blåt lys modvirker den stærke dæmpning af korte bølgelængder i sedimentet. Mangfoldigheden af nanostrukturer i frustuler og deres underliggende optiske egenskaber kan dermed forklares ved hjælp af niche differentiering; et centralt element i klassisk økologisk teori. Yderligere tyder vores datapå at frustulen har forskellige optiske komponenter med delvist komplementære optiske egenskaber, i.e. frustulens bånd producerer iridescence. Med disse forskellige optiske egenskaber, forårsaget af fotoniske krystalstrukturer, har frustulen en central betydning for kiselalgers fotosyntetiske produktivitet, og bidrager dermed til kiselalgers mangfoldighed og evolutionære succes.

Acknowledgements

I am deeply grateful for the friendly environment at the Marine Biological Section. It would not have been possible to write this thesis without the open-minded spirit and the manifold scientific possibilities at KU’s Faculty of Science. A heartfelt thank-you to my dear colleagues at KU!

My sincere gratitude is owed to my academic advisor Michael Kühl, who has been an inspiration both scientifically and as a group leader. I am deeply obliged to his trust in me and my work, and to his scientific open-mindedness and high flexibility that allowed development of the thesis into a direction not initially planned. His input has significantly improved the current thesis.

My utmost respect and admiration is owed to Marianne Ellegaard who has been extremely supportive during my entire work in Copenhagen. The inspiring discussions and free exchange of information has significantly shaped this thesis in its current form.

I would like to express my sincere thanks to João Serôdio for his scientific advice and warm hospitality during my two-month stay at the University of Aveiro. His expertise and the support from his team have improved the current thesis.

Special thanks are owed to Marianne Ellegaard, Paulo Cartaxana, Klaus Benedikt Möllers and Ryan Pearl for revision and Danish translation of the abstract, and to the many coworkers and –authors involved in this project.

I would finally like to thank my dear friends and peers Klaus Benedikt Möllers, Ryan Pearl, Hari Charan Ayada, Patricia Freire, and my family Franziska, Ingeborg and Hubert Gößling for the endless trust, support and love.

Sincere thanks!

Johannes W. Goessling

Contents

Preface …………………………………………………………………………………….. 1 1. Introduction ……………………………...…………………………………………... 3 1.1. Peculiarities of photosynthesis in diatoms………………………….….…………. 5 1.2. Structural adaptation for photosynthesis…………………………..…….…...…… 9 1.3. Photonic structures in nature...……………….………………..……….…...... …... 11 1.4. Frustule morphogenesis and biomineralization………………………....…….….. 15 2. State of the art and objectives …….…………………………………………… 17 3. Main results and discussion …….……………………………...……………...... 19 3.1. Optical properties of C. granii frustules ………..………….……….………..…... 19 3.2. Stabilization of the cellular light field in C. granii …………………….…...….…. 21 3.3. Optical properties of benthic diatom valves ……………….……...…………..….. 25 4. Conclusion and outlook ………………….………………………………………. 29 5. References …………………...………...……………………………………………… 31 6. Included manuscripts ………………………………………………….…………. 41 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms……………………………………………………………………….…… 43 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules…………………………………………………………………………… 91 6.3. Photo-protection in the centric diatom Coscinodiscus granii is not promoted by chloroplast high-light avoidance movement……………………………….……... 115 6.4. Light and O 2 microenvironments in two contrasting diatom-dominated coastal sediments…………………………………………………………………….……. 127 6.5. Niche differentiation in the microphytobenthos is affected by the optical properties of raphid diatom valves ………………….…….……………………… 143

List of tables and figures

Figure 1: Evolution and diversity of diatoms 4 1. Figure 2: Ocean color 10 Introduction Figure 3: Photonic structures and structural coloration in nature 13 Figure 4: Frustule morphogenesis during asexual reproduction of the diatom cell 16 Figure 5: Structure of the C. granii frustule and valve 20 Figure. 6: Aquatic chlorophyll seasonal composite and Coscinodiscus granii 3. 23 recordings Conclusion Figure 7: Interference of diurnal solar irradiance and tidal cycles in estuarine intertidal 27 mudflats Figure 1. Spatial organization of the Coscinodiscus granii cell and frustule 69 Figure 2. Photonic waveguiding in the cleaned valve wall and optical coupling of 71 chloroplasts Figure 3. Stimulation of diatom photosynthesis by waveguiding in the valve and 73 optical coupling to chloroplasts Figure 4. Light-collecting isotropy of frustule valves as a function of angle of 75 incidence Figure 5. Influence of orientation and refractive index on the optical properties of 77 6.1. diatom valves Manuscript Figure 6. Effects of angle of incidence and refractive index on scattering properties of 79 1 cleaned valves and living cells Figure 7. Light attenuation, photopigments and PSII absorption in live diatoms 81 Suppl. Figure 1. Setup for laser induced single cell variable chlorophyll fluorimetry 83 Suppl. Figure 2. Spatial distribution of photosynthesis in single diatom as a function 85 of laser spot irradiance Suppl. Figure 3: Valve reflectance and in vivo light attenuation measurements on 87 diatoms Suppl. Figure 4. Light scattering of glass beads in water 89 Figure 1: Structure of the centric diatom frustule in C. granii 105 Figure 2: Optical effects of immersed valves, girdle bands and live diatom cells in the 107 centric diatom C. granii observed with light microscopy 6.2. Figure 3: Nanostructure of valve and girdle band in the centric diatom C. granii 109 Manuscript Figure 4: Light transmittance through valves and girdle bands of the centric diatom C. 2 111 granii Figure 5: Light reflectance on the surface of valves and girdle bands in the centric 113 species C. granii

Table 1: Photosynthetic characteristics of Coscinodiscus granii incubated under blue, 120 red, and white light Figure 1: Rapid chlorophyll fluorescence light-response curves of Coscinodiscus 121 granii under blue, red, and white light Figure 2: Chloroplast distribution in cells of Coscinodiscus granii 121 Figure 3: Proportion of Coscinodiscus granii cells with an altered chloroplast 122 distribution 6.3. Figure 4: Induction of altered chloroplast distribution in Coscinodiscus granii after 10, 122 Manuscript 30, and 60 min of exposure to increasing BL photon irradiance levels 3 Figure 5: Variable chlorophyll fluorescence in cells of Coscinodiscus granii showing an even chloroplast distribution or an altered chloroplast distribution after exposure to 123 high blue light and after recovery in the dark Figure 6: Maximal chlorophyll fluorescence yield, Fm ′ , and effective PSII quantum yield in Coscinodiscus granii cells during alteration of the chloroplast arrangement 123 under high BL photon irradiance Figure 7: Lethality test using Sytox® green staining on Coscinodiscus granii cells 124 with an even chloroplast distribution or an altered chloroplast distribution Table 1: Definition of abbreviation 131 Table 2. List of 28 diatom taxa found in the muddy coastal sediment 134 Table 3. List of 46 diatom taxa found in the sandy coastal sediment 135 Table 4. Depth-integrated photosynthesis, oxygen fluxes and respiration 137 6.4. Figure 1. Particle size fractions of the muddy and sandy coastal sediments 133 Manuscript Figure 2. Photon scalar irradiance depth microprofiles of photosynthetically available 136 4 light Figure 3. Depth profiles of spectral scalar irradiance 136 Figure 4. Steady-state O2 depth microprofiles 137 Figure 5. Volumetric gross photosynthesis depth microprofiles 137 Figure 6. Areal rates of net photosynthesis 138 Table 1: Diatom taxa collected from an estuarine intertidal mudflat 173 Figure 1: Organization of the frustule of motile raphid diatoms 175 Figure 2: Structural coloration of frustule valves in raphid diatoms 177 Figure 3: Optical properties of frustule valves and in vivo distribution of 6.5. 179 photosynthesis Manuscript Figure 4: Photosynthetic activity of raphid diatoms from an estuarine intertidal 5 181 mudflat Figure 5: Spectral attenuation of light in a diatom dominated intertidal estuarine 183 mudflat Figure 6: Light enhancement by nanostructured frustules of raphid diatoms 185

Preface

One of the first certain records of a diatom goes back to 1703, communicated in a paper to the Philosophical Transactions of the Royal Society of London. In his simple compound microscope, an unknown English country nobleman observed what he described as plants “made up of two parallelograms joyn’d longwise” 1,2 . His conclusion, that he was observing a phototrophic microorganism is remarkable, considering that all he had for comparison were macroscopic plants and algae 2. Perhaps even more surprising, he also precisely described the more or less bipolar structure inherent to virtually all diatoms, given by two symmetric shells (valves) made of almost pure glass. In concert with improvements in light microscopy in the following centuries, diatoms became a popular scientific fashion for educated European and American gentlemen 2, owed to the microscopic structural variety and the architectural beauty of the different diatom species valves. Diatom valves have ever since delighted hobby-microscopists, diatomists and scientists, and are still favorite objects used to calibrate compound microscopes today. A remarkable body of species descriptions has been acquired until the end of the 19 th century 2, before species identification significantly improved after the development of transmission electron microscopy in the 1960s 3. Better understanding of diatom physiology, biochemistry and its ecological importance expanded over the 20 th century, but only very recently – probably triggered by a paper published by Fuhrmann et al. in 2004 4 - research has been focused more upon the diatom valve optical properties and their structure-property relationships. Since then, it is debated enthusiastically whether or not the photonic properties of diatom valves may affect the living cell - foremost discussed in light of their potential implications upon photosynthesis 5,6 and literature therein . The two valves are of slightly different diameter and fit into one another; in principle like a Petri dish. One or more girdle bands are associated with each valve, encircling them at the overlapping region, thus keeping the two valves together. The entire construct named the frustule is made of almost pure glass (silicate), including the two valves and their corresponding girdle bands. The current thesis explores the potential implications of the diatom frustule optical properties upon photosynthesis at the interface of biophotonic research and diatom photobiology.

1

2

1. Introduction

Diatoms are single cell microalgae that are among the most abundant and diverse photosynthetic organisms on planet earth, present in the photic zone of all aquatic environments and many moist terrestrial habitats 7,8 . Diatoms are basal species of the marine food web and play a crucial role in the ocean’s carbon cycle 9,10 . Although it can only be estimated, diatoms may perform up to 20-25% of the world’s photosynthetic carbon fixation, while producing more oxygen on a global scale than tropical rainforests, to which it is often referred to as the green lungs of the world 11,12 . The term diatoms comprises a monophyletic group of organisms with extremely diverse appearance and life-forms 2,13 . The evolutionary origin of diatoms is still largely unknown, although genetic insights assured some more certainty in recent decades 14–16 (Fig. 1). Diatoms belong to the heterokonta (Bolidophycea), a superphylum comprising phototrophic divisions such as the Phaeophyceae (e.g. and other macroscopic ), Chrysophyceae (golden-brown algae) and diatoms 2, but also parasitic which have lost their ability to photosynthesize 17 . While were first located within the biological of plants, they were later moved to the kingdom of protists, a paraphyletic group of eukaryotic organisms that are not plants, fungi or animals 18 . Heterokonts are sometimes also specified as stramenopiles or chromista; terms that accentuate different characteristics inherent to organisms within the heterokonta superphylum. The term stramenopiles - from Latin stramen (straw) and pilus (hair) - refers to the two heterodynamic flagella, one of which is covered with tripartite hairs unique to some heterokonts 8,19 . However, these flagella have been lost in the diploid diatom cell, and is only preserved in the sperms of centric diatom species 20 . The term chromista comprises a paraphyletic group within the heterokonta 21 , referring to the unique pigmentation, i.e. chromista contain chlorophyll c instead of chlorophyll b, which is present in viridiplantae (algae with green chloroplasts and plants)22 .

It is assumed that a primary endosymbiosis event – the inclusion of a cyanobacterium into a heterotrophic prokaryotic host - happened only ones (maybe twice)21 around 1.5 billion years ago, then diverging into the three lineages of green algae, glaucophytes and red algae 23 . Heterokonts evolved thereafter by secondary endosymbiosis involving a red algae ancestor domesticated by a heterotrophic host 8. The red appearing pigment phycoerythrin, characteristic and eponymous for the red algae lineage, has been lost in diatoms 24 . Noticeable differences in the photosynthetic architecture are owed to the diverging evolutionary origin of viridiplantae and diatoms 25,26 , although “we do not know what the consequences of the secondary endocytobiosis on the metabolic regulations were, nor whether the physiological peculiarities of diatoms observed in the past reflect this different host-endosymbiont relationship” (in Wilhelm et al. , 2006) 27 . Many of these physiological peculiarities are associated with photosynthesis as the central metabolic pathway in the photoautotrophic diatom cell.

3

Figure 1: Evolution and diversity of diatoms.

More than 10.000 diatom species have been identified to date, all different in size, shape and life form 173 . Radial centric species evolved first followed by bipolar and multipolar centric species. Araphid and raphid species arose thereafter 13 . The here indicated gradual evolution from centric to pennate symmetry is still under debate 8.

Micrographs were downloaded (01.06.2017) from http://www.mikroskopie.de/mikroforum_2/index.php?topic=3711.0 (Hydrosera triquetra ); http://www.algaebase.org/search/species/detail/?species_id=b069a3580f92e3e88 (Rutilaria epsilon ); https://www.niwa.co.nz/media-gallery/detail/109727/29357 (Asterionellopsis glacialis ); http://tolweb.org/raphid_diatoms/ 125307 (Neidium sp. )

4 1. Introduction

1.1. Peculiarities of photosynthesis in diatoms

In principle, photosynthesis can be divided into the primary reaction – i.e. the conversion of light energy into chemical energy equivalents - and the secondary reaction – i.e. the consumption of these energy equivalents in order to fix atmospheric carbon in form of organic matter. The primary reaction is driven by two photosystems which harvest the energy provided by sunlight with accessory photopigments clustered together in light harvesting complexes located in the thylakoid membranes 28 . The main photosynthetic pigments in diatoms are chlorophyll a, fucoxanthin and chlorophyll c29,30 - in contrast to the pigmentation of viridiplantae, which contain large amounts of chlorophyll a and chlorophyll b31 . The diatom photopigment fucoxanthin broadens light harvesting to the green spectral range of light 32 . Light harvesting complexes collect the energy of striking light photons, which is then transmitted by Förster resonance transfer to the reaction centers of the corresponding photosystem 33 . The reaction centers are complex protein structures with a central pair of modified chlorophyll molecules that initiate electron transport along the thylakoid membrane via charge separation achieved by oxidation of primary electron acceptor molecules, i.e. pheophytin prior to plastoquinones 34 . The electron gap incurring at the reaction center of photosystem II is re-filled by oxidation of water at the water splitting side of photosystem II, furthermore resulting in formation of molecular oxygen. Reduced primary electron acceptors channel the excited electron to the cytochrome b6f-complex, which uses this energy to pump hydrogen ions from the chloroplast stroma into the luminal space 35 and literature therein .

Due to the impermeability of the thylakoid membrane, hydrogen ions cannot diffuse back resulting in the formation of an electrochemical gradient between stroma and lumen. This gradient is used to fuel a synthase protein located in the thylakoid membrane, which produces adenosine triphosphate (ATP) on the stroma side. A second photosystem is needed, because the energy of the electron funneled from photosystem II is consumed during protonation of the thylakoid lumen 28 . Therefore, an additional electron is released at the reaction center of photosystem I, following the principle of resonance energy transfer likewise to that occurring at photosystem II. This high-energy electron is passed to enzymes (Ferredoxin-NADP + reductase) producing nicotinamide adenine dinucleotide phosphate (NADPH) at the stroma side. The electron lost at photosystem I is re-recruited from photosystem II via the electron carrier protein plastocyanin. While plants and green algae adapt the ratio of antenna pigments associated with each photosystem in response to the availability of sunlight 36,37 , the ratio of photopigments associated with each photosystem is in diatoms not significantly altered under various light conditions 26 . However, it has been observed that the stoichiometry of photosystem II and photosystem I and associated factors can significantly vary in diatom species from pelagic and coastal regions, linked to the differences in light climates present in these habitats 38 . Short-term imbalances in available photosynthetic active radiation that can be caused by mixing events such as cloud shading or wave-focusing can be overcome by state transition, i.e. the reversible re-allocation of light

5 harvesting centers between photosystem II and photosystem I. While this mechanism is very effective in terrestrial plants, green algae and cyanobacteria 39,40 , it does not exist in diatoms 27 .

Reduction equivalents compounded during the primary reaction are consumed in the secondary reaction of 35 and literature therein photosynthesis . In this reductive pathway, atmospheric CO 2 is fixed in form of carbohydrates or other organic molecules. While starch is the major photo-assimilate in plants, diatoms store atmospheric carbon in form of lipids and the X-1,3 glucan chrysolaminarin, that in contrast to starch can affect the buoyancy of the cell determining the vertical position of the single cell along the water coumn 27 . Activity of the key enzyme in the secondary photosynthesis reaction named Ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) is controlled by a complex redox-regulatory feedback system. Rubisco is activated during the primary photosynthesis reaction, when ATP pools are filled and the stroma pH is low 41 . Sensitivity of Rubisco to its activation factors is mediated by the regulatory protein thioredoxin 42,43 . Thioredoxin shuttles electrons to photosystem I via ferredoxin and the ferredoxin-thioredoxin reductase, which in turn alters the catalytic activity of Rubisco. Thioredoxin thereby functions as an important redox- switch that down-regulates the Rubisco activity during energy-overloads emerging from high chlorophyll excitation states; however, diatoms seem to lack the thioredoxin redox-regulation pathway 44 . Low pH in the luminal space occurring during over-reduction of the electron transport chain may paralyze the ATPase- synthase activity and thereby depletes available ATP in the stroma, leading to inhibition of the Rubisco activation state 41 .

Over-reduction of the electron transport chain can be avoided by two essentially different strategies. Avoidance strategies comprise mechanisms to shield or dissipate excessive energy before it enters the electron transport chain, while removal strategies dissipate excessive energy from an over-reduced electron transport chain 45 . A central pathway of avoidance is provided by the xanthophyll-cycle, which is activated when excitation energy is pulled towards the reaction centers while oxidizing agents are missing and the luminal hydrogen ion concentration increases 46 . The xanthophyll-cycle is low-pH-driven, and initially activates an enzyme that catalyzes the reduction of the pigment violaxanthin to zeaxanthin. Zeaxanthin is a central component of the cellular photo-protective response, because it absorbs light in the actinic spectral range and reflects excessive radiation by thermal dissipation 45 . While only traces of the zeaxanthin cycle were found in diatoms, an equivalent pathway named the diadinoxanthin cycle is particularly effective in diatoms 47–50 . It was e.g. demonstrated that the capacity of diatoms to dissipate excessive light energy was three- to five-fold higher in the pennate species Phaeodactylum tricornutum compared to that of vascular plants under intermittent light regimes 51 . When excessive energy is not initially dissipated before it can enter the photosynthetic electron transport chain, photosystem II activity may be inhibited by phosphorylation when the plastoquinone pools remain in a reduced state 52 . When excessive excitation energy is then not dissipated by fluorescence or heat 53 , it can be transferred to alternative electron acceptors alongside the

6 1. Introduction

electron transport chain instead. Transfer of resonance energy from an excited photopigment to an oxygen molecule results in the formation of reactive oxygen species; during photosynthesis predominantly formed at the water splitting site of photosystem II, and by auto-oxidation of primary electron acceptors at the reducing site of photosystem I 54 . Although reactive oxygen species may be an inevitable by-product of the photosynthetic electron transport chain (with potentially important implications for the cell)55 , uncontrolled formation of reactive oxygen species can cause severe damages, ultimately leading to cell death 36,54 . To prevent oxidative stress, the cellular redox potential is controlled by antioxidants and the enzymes linking these metabolites 56 . Controlled formation of reactive oxygen species can for example be achieved with the Mehler reaction occurring at photosystem I, which transfers electrons to oxygen leading to the controlled formation of superoxide anions that can subsequently be converted into water 57 . This reaction may play a crucial role in the photo-protection of photosystem II under high photon flux densities 58 . While turnover of inactivated photosystem II occurs only during the light period in plants and cyanobacteria 59 , diatoms seem to continue the recycling of photosystem II also in the darkness, which could have evolved in consequence to rapidly changing light intenisites 60 .

7

8 1. Introduction

1.2. Structural adaptation for photosynthesis

Diatoms can be subjected to periodic and variable light environments caused by tidal rhythms, diurnal and seasonal variations in light intensity, water turbidity and currents, in addition to mixing events, cloud shading and wave-focusing of incident sunlight 61,62 . Changes in the visible spectral range of light occur as well, i.e. blue light can penetrate deep into oligotrophic waters, while shorter wavelengths attenuate more in turbid and nutrient-rich waters due to scattering on detritus and absorption by phytoplankton 63,64 (Fig. 2). The absorption of particular parts of the visible spectrum of light by the specific set of photo-pigments that evolved in phototrophs, facilitates niche differentiation along light as the resource spectrum 65 . Phototrophs can sense the extracellular light environment by photoreceptors 66 , that allow them to adapt and control photosynthesis under changing light conditions. The major photoreceptors of viridiplantae are also expressed in diatoms, i.e. the red light photoreceptors from the phytochrome family, and the blue light photoreceptors from the cryptochrome family 67 . Diatoms inherent an additional class of photoreceptors - the aureochromes - blue light photoreceptors unique to stramenopiles 68,69 . Aureochromes contain a light-oxygen-voltage (LOV) sensitive domain and a basic zinc-zipper motif, equivalent to the highly conserved domains in phototropin blue light photoreceptors present in higher plants and green algae 70 . Photoreceptors regulate kernels of development and photo-morphogenesis 71 and the chlorophyll biosynthesis 72 . They also mediate structural adaptation and light induced movement, like phototropism 73 , stomata opening 74 and chloroplast re-allocation movement 75 . The ability to re-arrange chloroplasts in the cell is controlled by the cytoskeleton 76 . Chloroplast re-arrangement movement improves photosynthetic efficiencies while photo-damages can be minimized when the photon flux density is too high 77–79 . Both, photoreceptors of the blue and the red light photoreceptor families are involved in the re-allocation movement of chloroplasts 79,80 , but foremost blue light and high solar irradiance induce high-light chloroplast avoidance movements (e.g. DeBlasio et al. 2003; Gabry ś 2004). Chloroplast re-allocation movement has been observed in different marine, benthic and freshwater diatom species 83–85 , and may hence be an important mechanism to regulate the light exposure of chloroplasts during photosynthesis in variable light environments.

Structural adaptation for photosynthesis is widespread among organisms benefitting from photosynthetic productivity, involving adjustments on different levels of organization 86 . Some halophytic terrestrial plant species expose epidermal bladder cells and salt glands on the leaf surface to back-reflect excessive sunlight 87,88 . Many terrestrial plants reduce illumination of the canopy by axial rotation of the leaves 89 . Famous examples of structural light adaptation on the community level are the spatial highly organized canopies of tropical rainforests, optimized for efficient light use throughout the photosynthetically active spectral range of light 90–92 . A unique form of structural adaption to light are photonic structures on the nano- to microscale, which are surprisingly widespread throughout the kingdoms of phototrophic organisms 93 . However, only little attention has been given to the potential implications of such structures upon physiology

9 Figure 2: Ocean color. A) Blue light penetrates deep in oligotrophic waters (Red Sea, Egypt). B and C) Blue light is attenuated in nutrient rich waters (B: Lysekil, Sweden; C: Little Belt, Denmark).

Photos courtesy of Susann Diercks and Felix Ivo Rossbach.

or photobiology. It has e.g. been demonstrated that symbiont bearing giant clams form crystal-like protein stacks that back-reflect harmful radiation while photosynthetic productive wavelengths are focused into deeper tissue where photosymbiotic microalgae are located 94 . Photonic structures have also been observed in shade-dwelling species of Begonia spp. (Begoniaceae) inhabiting in the twilight understory of tropical rain forests. Leaves of these species appear blue due to specialized chloroplasts named iridoplasts, located in the adaxial epidermis 95 . The iridescent brilliant-blue coloration of these eaves is caused by highly ordered internal layers of thylakoids with photonic crystal properties 96 . Photonic structures also evolved in the frustule surrounding diatoms, and it has been speculated that such structures could mediate efficient photosynthesis in diatoms 4–6,97–99 .

10 1. Introduction

1.3. Photonic structures in nature

In contrast to pigments, which appear colorful because they absorb parts of the visible spectrum of light, structural colors are caused by the physical interaction of incident light waves with structure, such as selective reflectance or scattering. The underlying materials are thereby often completely transparent100 . Famous, well-studied examples of natural structural coloration are the colorful chitin exoskeletons and wings of many insects, iridescent hair-like setae in polychaete worms, or the pied feathers of several bird species 101 . Structural colors are also widespread in aquatic ecosystems, and can be observed in the metallic appearing scales of fishes 102 , colorful shells of some mollusk species 103 and in the silicate frustule of diatoms 104 (Fig. 3). Singular optical phenomena have evolved in nature 105 ; one eye-grabbing example is iridescence, a phenomenon caused by multilayer reflectors or linear diffraction gratings 106 . Iridescence bases on interference of reflected light waves mirrored on at least two, but often more stacked layers of materials with alternating refractive indices. While a small part of incident light is reflected on the exterior surface, most of the light passes through the material, while another small part is reflected on the internal surface of each stacked layer. Constructive interference of coherent light waves – which have traveled different distances depending on the surface layer by which they were reflected, as well as a function of total layer thickness, number of layers, and layer refractive indices – then appear in different color at defined observation angles 107 . However, iridescence can be masked in the presence of pigments, thus becoming invisible to the naked eye 108,109 .

The perception of color is not only dependent on the objects physical properties, but on the visual system of the spectator; independently of whether the color is of structural nature or induced by pigments or bioluminescence 101 . This is owed to the different types of light perception organs that evolved in nature 110 . In consequence, organisms can perceive color differently in its basically three attributes hue, saturation and brightness 101 . The predator-prey relationship might have been an influential trigger for the evolution and diversity of natural structural coloration 93 . For example, iridescent coloration produced in the scales of hibiscus harlequin bugs appear repulsive to predating bird individuals, and is thereby an example of aposematism (warning coloration) caused by structural coloration 111 . Another trigger for evolution of natural structural coloration is apparently sexual attraction. In the latter case, a famous example is the male peacock’s colorful plumage that attracts females by its mesmerizing play of shiny and often iridescent colors 112 . The underlying mechanism bases on photonic bandgaps occurring in a 2-dimensional (2D) photonic crystal nanostructure inherent to the feather barbules. Variation of the 2D lattice constant changes the bandgap for particular wavelengths, thus causing reflection of different colors in the visible spectrum of light 113 . These bandgaps occur in consequence to differences in the refractive index between the periodic structures and the surrounding medium, which in nature is often air or water 114–116 .

11 Similar periodic structures evolved in the frustule of diatoms; an encasement of the cell made of almost pure 117 silicate (SiO 4) . The frustule may further contain traces of metals such as iron and aluminum, but amorphous silicate is the largest component with up to ~90-95% in oxidized frustules 117–119 (oxidized here implies chemically cleaned from organic matter). The frustule is shaped by its two almost identical valves with slightly different sizes, making them fit into one another in principle like a Petri dish 120 . The overlapping regions are connected with more or less circular silicate structures called girdle bands 121 . While girdle bands can be structured with pores, bridged with flat or undulated plates, or be poreless 120 and literature therein , the valve is always porous facilitating nutrient- and gas-exchanges between cell and environment 122 . Valve structures are usually arranged in square or hexagonal lattices, and comprise smaller and larger pores in addition to hexagonal chambers 4,120,123,124 . In many species, optical phenomena become visible to the human eye when uniform white light is shine on oxidized frustule valves; often restricted to the angle of incident light 125 . These phenomena can in some species be explained by wave diffraction on the semi- periodic porous structures or by Rayleigh scattering on small silicate elements inside the valve such as valve walls125 . Many studies have been focusing such diffraction phenomena on diatom frustules, i.e. diffraction of light waves on the grating structure of pores or slits present in most diatom frustule valves 5,125 and literature therein . Such interference could for example focus particular wavelength at defined distances behind the valve 126 , and may enhance photosynthetically productive wavelengths inside the frustule 127 .

12 1. Introduction

Figure 3: Photonic structures and structural coloration in nature. A) Blue structural coloration of the Morpho sp. butterfly (left hand) and its underlying photonic crystal structures (right hand) 115 . B) Male peacock ( Pavo cristatus ) presenting its plumage with mesmerizing structural colors (left hand). The peacock feather barbules have iridescent coloration (right hand) 113 . C) Fish scales in sardines (Sardina pilchardus ; left hand) appear metallic-blue due to photonic structures (magnified fish scales right hand) 101 . D) Begonia pavonina lives in the understory of tropical rainforests (left hand). The blue coloration is produced by specialized chloroplasts with stacked thylakoid membranes (right hand) 96 . E) Giant clams ( Tridacna spp. ) harbor symbiotic microalgae. Nano-scaled protein stacks back-reflect harmful blue radiation, while photosynthetically more productive radiation is transmitted (right hand) 94 . F) Structural coloration in oxidized valves of the centric diatom species Coscinodiscus granii observed in dark field microscopy in water (left hand), and its underlying nanostructures visible in a broken frustule valve (right hand). Scale bars: A: 1 µm; B: 50 µm; D: 1 µm; E: 2µm; F: 100 µm (left); 5 µm (right).

Images were provided by A, left) http://eol.org/pages/146992/details (10.05.2017); A, right) Kinoshita et al. (2002) 115 ; B, left) http://sobrefotos.com/2009/09/18/fotos-de-pavos-reales-esos-presumidos-animales/ (10.05.2017); B, right) https://www. wimp.com/photographer-uses-microscope-to-capture-the-incredible-beauty-of-peacock-feathers/ (10.05.2017); C, left) https:// www.stocksy.com/485036 (10.05.2017); C, right) https://www.pinterest.de /pin/382594930817285275/ (10.05.2017); D, left) https://www.pinterest.com/pin/78813062207827197/ (10.05.2017); D, right) Jacobs et al. (2016) 96 ; E, left) http:// otlibrary.com/giant-clam/ (10.05.2017); E, right) Holt et al. (2016) 94 .

13

14 1. Introduction

1.4. Frustule morphogenesis and biomineralization

There are two main morphotypes of diatom frustule shapes 2. Centric diatoms (or Centralis) are encased in circular or ellipsoid valves, bearing symmetrical pattern radiating from a central point 120 . Pennate diatoms (or Pennales) have usually elongated or round-shaped valves with bilateral symmetric ornamentation; however, the Pennales can be further divided into eight paraphyletic subdivisions by morphological characteristics 2. The frustule size can be extremely various, ranging from a few micrometers in small species to some millimeters in the largest yet discovered species named Ethmodiscus rex 128 . In contrast to most other algae, the cell wall formation is in diatoms semiconservative, which means that each asexually emerging cell comprises one parental and one newly synthesized daughter valve in addition to their respective girdle bands 120 (Fig. 4). The larger valve and its corresponding girdle is called epitheca, and the smaller valve hypotheca 120 . Because cytokinesis occurs inside the frustule, new valves are formed inside each parental valve. Hence, the cell size of one daughter cell decreases while its sibling, which inherits the larger parental epitheca, remains its cell size 2. In a natural diatom population, the cell size can thereby vary at a magnitude of two to five 129,130 . The maximum cell size is only restored after a minimum cell size is reached, which induces auxospore formation for sexual reproduction131 . One of these cell-size reduction-restitution cycles can go on over several tens to hundreds of generations 132 , leading to a typical saw tooth pattern of cell size in a diatom population over time 133 . It is unclear whether size reduction is essential, or an inevitable side effect of asexual reproduction in diatoms 131 , but cell size may have important implications upon sinking rates and predation escape 134,135 .

Frustule formation bases on the biomineralization of soluble silicic acid into amorphous, solid silicate. Although the whole silicification process in diatoms is unknown 136 , biomineralization in diatoms is initiated in specialized membrane-bound cell compartments called silica deposition vesicles 137 . It is furthermore known that material associated with the valve precursor, such as amino acids, peptides and carbohydrates, as well as the pH control inside the silica deposition vesicle 138 , influence silicification and patterning of frustule valve structures 139–141 . During cytokinesis, silica deposition vesicles emerge below the plasmalemma at the interior side of the parental valve 142 . After the inner valve is formed, silica deposition vesicles are excreted by exocytosis and the outer wall elements are completed 137 . Girdle bands can be formed together with the valve, but may be formed at later stages in some diatom species 120 . One intriguing process is the formation of periodic structures inside the silicate frustule valve. In centric diatoms, a silica deposition vesicle first emerges at a so called pattern center 136 , which is the nucleation point of the structural radial geometric symmetry. While the silica deposition vesicle expands, rib-like structures are formed along the y-z axis 137 . As these ribs become thicker due to silicate deposition over time, the in-between spaces are closed while only pores and chambers are left open. The frustule valve then grows along the z-axis, allowing formation of complex 3D structures 136 . The formation of pennate frustules follows similar basic principles, but is induced

15 along the apical axis before it expands perpendicularly 136. “Major questions still to be addressed concerning silicification in diatoms are, 1) how is higher order structure formed, 2) how is spatial and temporal control imparted over the process?” (in Hildebrand and Lerch, 2015) 136 . These questions may in future be addressed with exogenous manipulation of the silicification process and alteration of the frustule architecture in diatoms, such as demonstrated with light 143 , heavy metals 144,145 , different salt concentrations 146 and cytoskeleton inhibitors 147 , and complemented by genetic and transcriptomic analysis.

Figure 4: Frustule morphogenesis during asexual reproduction of the diatom cell. Center: Conception of the diatom frustule in an exploded scheme. Two valves and their corresponding girdle bands form a larger epitheca, and a slightly smaller hypotheca. 1) Cell before division. 2) After cytokinesis, newly formed biosilica (red) is deposited inside a silica deposition vesicle (yellow). 3) Expansion of the newly formed biosilica. 4) The new valves are fully developed. 5) The new valves are deposited by exocytosis. 6) Separation of the sibling cells. 7) Initiation of girdle band synthesis (red) mediated by newly formed silica deposition vesicles (yellow). 8) Progressing synthesis of further girdle bands. 9) Final stage. Two newly formed mature cells, now both of slightly different size. Figure and caption were adapted from Kröger and Poulsen (2008) 174 .

16

2. State of the art and objectives

Several hypotheses have been proposed to understand the evolutionary advantage of nanoporous structures in the diatom frustule. The network of pores facilitates the diffusive exchange of nutrients and metabolites between cell and environment 148 . The periodic architecture of chambers and pores generates an enormous mechanical stability, which could protect the cell from animal predation 149 . It was also proposed that the frustule could serve as a ballast, facilitating fast sinking of the cell when nutrients in the water column get depleted 135 . Based on numerous studies on the optical properties of diatom frustules it has also long been speculated that photonic structures in the frustule could affect diatom photosynthesis 5 and literature therein . It was for example observed that the frustule valve of the centric diatom species Coscinodiscus centralis enhanced light in the red spectral range, matching the red light absorption peak of chlorophyll a98. Valves of the freshwater diatom Melosira variance attenuated shorter wavelength in the blue spectral range of light, which led the authors to conclude that the valve could screen for more-energetic wavelengths in order to protect photosynthesis under high photon irradiance 99 . Diffraction gratings of large pores in valves of the centric species Coscinodiscus wailesii seem act as microlenses, focusing photosynthetically active radiation by superposition of diffracted waves at fixed distances inside the cell 127 . Theoretical assumptions predicted that the chambered frustule valve wall functions as a Fabry-Perot interferometer, i.e. diffraction of incident sunlight on the chamber walls could enhance photon irradiance by four orders of magnitude 150 . It was also predicted that incident light can be guided at the horizontal plane of the Coscinodiscus granii frustule, which could in live cells distribute actinic light homogenously over the diatom cell 4. Such waveguiding could be enabled by the low refractive index contrast of seawater (RI ~ 1.34) 151 inside the chambers of the valve and the silicate wall elements (RI ~ 1.43) 4. A non-propagating evanescent wave at the junction between valve and cytoplasm could then transmit light from the frustule to the chloroplasts via optical coupling 4.

While these hypotheses were consolidated on base of theoretical assumptions, or on the optical studies of oxidized valves in air, the current thesis explores frustule optics also in water and in live diatom cells. The potential implications of photonic structures in the frustule and the optical properties upon diatom photosynthesis were investigated with advanced microscopic techniques and microprobes. The earlier proposed waveguiding property of the C. granii frustule was tested experimentally with a newly developed microscopic setup allowing for simultaneous 1) stimulation of photosynthesis with a spectrally filtered, tunable and focusable laser light, while 2) measuring spatial and temporal changes of variable chlorophyll fluorescence of photosystem II with microscopy PAM chlorophyll fluometry 152 . Micro- and nanostructures of the frustule were determined with Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM). Optical properties were evaluated with fiber-optic microprobes (scalar irradiance 153 and field radiance probes 154 ) in addition to optical microscopy techniques, i.e. light microscopy and hyperspectral

17

imaging. Physiological measurements were performed on live cells with different methods of pulse amplitude modulated chlorophyll fluometry. We also investigated physiological structural mechanisms that can stabilize the cellular light field in cells of phototrophs, i.e. chloroplast re-allocation movement in response to light color and photon irradiance.

To better understand the ecological advantage of photonic structures of frustules upon diatom photosynthesis, we then analyzed the chemical and physical microenvironments of sandy and muddy sediments dominated by benthic diatom species prior to optical measurements on oxidized frustules. In particular, we studied oxygen evolution in response to light with fast responding Clarke-type microelectrodes, and determined spectral light attenuation in the sediments with fiber-optic scalar irradiance microprobes. Potential links of the optical properties of valves from motile pennate diatoms species and diatom photosynthesis were investigated as a function of light color and photon irradiance with pulse amplitude modulated chlorophyll fluometry.

18

3. Main results and discussion

3.1. Optical properties of C. granii frustules

The C. granii frustule is composed of two centric valves and their corresponding girdle bands. The valve exhibits three layers with photonic structures 4, i.e. small pores arranged in a hexagonal lattice on the exterior side, honey-comb chambers inside the valve wall and large foramina pores on the interior side of the valve (Fig. 5). We could confirm experimentally that the C. granii frustule behaves as a 2D slab photonic waveguide on the horizontal plane 4, guiding incident light waves that diffract on the chambered valve walls (Manuscript 1, Fig. 2). We speculate that propagating light waves are then transmitted to the chloroplasts by two possible mechanisms, i.e. 1) optical coupling to the evanescent field, facilitated by large field penetration depth due to low refractive index contrast between silicate (RI ~ 1.43) 4 and cytoplasm (RI ~ 1.37-1.38) 155 , and 2) by leakage of light through the large foramina pores located on the interior valve side (Manuscript 1, Fig. 3, 4). The asymmetric structure of the C. granii valve wall seems to be made for light guided in plane when light is incident on the exterior side. This is indicated by lower transmission through the exterior side, while higher light reflection on the interior valve side could keep and enhance light inside the frustule (Manuscript 1, Fig. 5). When white light was incident at more acute angles, forward scattering of blue light was visible in dark field microscopy. This phenomenon could be caused by diffraction of blue wavelength on the porous grating of foramina located on the interior valve side (Manuscript 1, Fig. 6). Blue light was absorbed more by the photopigments of C. granii , and transmitted more efficiently from antenna pigments to the reaction centers of photosystem II, compared to other colors in the photosynthetic action spectrum (Manuscript 1, Fig. 7).

An interesting observation was the variation of photonic structures of C. granii valves and girdle bands, i.e. the much smaller pores in the girdle band were arranged in square lattice pattern, while the C. granii valve had larger pores in hexagonal arrangement (Manuscript 2, Fig. 3). These different structures generated partially opposite optical effects such as stronger attenuation of red light in girdle bands, while valves attenuated blue light more (Manuscript 2, Fig. 1-4). Girdle bands also showed iridescent coloration in dependency of the angle at which white light was incident (Manuscript 2, Fig. 5).

19

Figure 5: Structure of the C. granii frustule and valve.

(A) The living diatom cell is encased in a silicate structure named frustule. The frustule consists of different structuring elements, i.e. two valves of slightly different size and corresponding girdle bands. The larger valve is named epivalve, and the smaller valve is named hypovalve. ( B) SEM micrograph of valve and girdle bands of different size. ( C) The C. granii valve wall bears three periodic elements each of which can be regarded to as a photonic crystal structure. The valve is on the exterior side perforated by small pores (here almost invisible due to their small size of ~10 nm). Honey-comb chambers form the center of the valve wall. Large pores named foramina are located on the interior side of the C. granii valve.

.

20 3. Main results and discussion

3.2. Stabilization of the cellular light field in C. granii

Diatoms inhabit various environments with light climates that can differ in intensity, spectral distribution and these factors in combination. Stabilization of the cellular light field, i.e. optimized light exposure for efficient photosynthesis, can be achieved on different levels of organization, ranging from biochemical feedback mechanisms to cellular structural adaptation for optimal light exposure. Diatom species living in environments with different light climates may in consequence use different mechanisms to stabilize the cellular light field, e.g. it was observed that the photosynthetic architectures of diatoms from coastal and oceanic environments are significantly different 38 . Compared to species inhabiting coastal regions, oceanic diatom species apparently reduced photosystem I and cytochrome b6f complexes, which contain considerable amounts of iron. This reduction was hence understood as a requirement to the lower iron availabilities in the 156 open ocean, which often limit phytoplankton growth . Cytochrome b6f controls the protonation of the lumen when NADP + is limited during photosynthesis, which is used as an agent to oxidize ferredoxin during 157,158 the light induced photosynthetic electron transport . In consequence, reduced b6f concentrations in oceanic species might lower the flexibility of photosynthesis to intermittent light regimes. It was therefore speculated that the reduction of cytochrome b6f in oceanic species was enabled by the more stable and diffusive light climate in oceanic habitats 38 .

The centric diatom species C. granii has been sampled in both oceanic and coastal waters all over the globe (Fig. 6 and literature therein). Light climates at these sampling sites might significantly differ in light intensity and stability of incident light, influenced by mixing events in the upper water layers. Such environments might also vary in the spectral composition of available sunlight affected by variation of phytoplankton concentrations which absorb parts of the sunlight spectrum for photosynthesis 159,160 . We observed that blue light induced higher electron transport rates in C. granii compared to that induced by comparable amounts of red photons at photosystem II (Manuscript 3, Fig. 1). Blue light also induced higher non-photochemical quenching as a function of photon irradiance, while red light induced only negligible non-photochemical quenching even at high photon flux densities at photosystem II. We assumed that such flexibility of photosynthesis in C. granii was supported by structural adaptation mechanisms, such as chloroplast high light avoidance movement, which has been observed in several other diatom species including centric species 85,161 . Although we observed an alteration of the chloroplast distribution as a function of light dose, cells with an altered chloroplast distribution showed apparent signs of death and did not re-arrange chloroplasts during recovery under low light or in the dark (Manuscript 3, Fig. 2, 3). Interestingly, such alteration of the cellular chloroplast constellation was foremost induced during illumination with blue light, while comparable photon flux densities at photosystem II of red light killed only few cells in the culture. However, even high doses of blue light induced such an altered chloroplast

21

distribution only in ~35% of cells, while most of the cells survived such high light treatment (Manuscript 3, Fig. 4).

In conclusion, C. granii apparently occurs in environments with different light climates, including variations in light intensity as well as in the spectral light availability. C. granii can cope with astonishing high doses of photon irradiance, while the response of photosynthesis differs as a function of the spectral composition of light. In addition to effective non-photochemical energy dissipation, such flexibility of photosynthesis could be further affected by structural adaption mechanisms, but apparently not by chloroplast high-light avoidance movement. For instance, the light collecting property of the frustule might serve to capture actinic light in diffusive light environments with limited availabilities of photosynthetically active radiation, while it could also re-distribute striking photons at the water surface, where sunlight is often collimated 61 . Perhaps interesting, photonic structures in the C. granii valve are affected by the spectral composition of light during the growth period, i.e. the foramen diameter decreases when cells are grown under monochromatic green light, and the number of foramina decreases under monochromatic blue or green light 162 . Such variation in the constant of photonic lattice structures in the frustule could affect the light-structure interaction and might facilitate efficient photosynthesis in environments with different light climates.

22 3. Main results and discussion

Figure 6: Aquatic chlorophyll seasonal composite and Coscinodiscus granii recordings.

Phytoplankton concentrations were estimated by remote sensing of chlorophyll fluorescence (NASA Earth Observation; data for April 2017). Information on C. granii recordings were received at http://www.algaebase.org (30.05.2017).

Locations : Africa: Europe: Egypt (Nassar and Khairy, 2014) 175 . Adriatic Sea (Vilicic et al., 2002) 176 Baltic Sea (Hällfors 2004) 177 Arctic Ocean: Black Sea (BSPC Editorial Board 2014) 178 Point Barrow, Alaska (Bursa, 1963) 179 Britain (Hartley et al. 1996) 180 France (Guilloux et al. 2013) 181 Antarctica: Helgoland (Hoppenrath 2004) 182 Scotia Sea (Ferrario et al. , 2008) 183 Oslofjord (Hasle and Lange)184 Romania (C ărăuş 2012) 185 Atlantic Islands: Turkey (Aysel 2005) 186 Canary Islands (Afonso-Carrillo, 2014) 187 . North America: Asia: Canada (Mather et al. 2010) 188 . Caspian Sea (Bagheri and Fallahi, 2014) 189 China (Liu, 2008) 190 South America: Taiwan (Shao, 2014) 191 . Argentina (Garibotti et al. 2011) 192 Brazil (Eskinazi-Leça et al. 2010) 193 Australia and New Zealand: Colombia (Lozano-Duque et al. 2011) 194 . Australia (McCarthy, 2013) 195 Mexico (Ferrario et al. 2008) 183 . New Zealand (Harper et al. , 2012) 196 . South-west Asia: Bangladesh (Ahmed et al. 2009) 197 Iraq (Maulood et al. 2013) 198 Lebanon (Lakkis 2013) 199 .

23

24 3. Main results and discussion

3.3. Optical properties of benthic diatom valves

While the light climate in the open ocean might be more homogenous, light environments in coastal habitats can locally change rapidly due to mixing events such as wave action and currents 63 . Intertidal habitats like estuarine mudflats are affected by tidal rhythms, that interfere with the diel light period (Fig. 7)163 . Benthic diatom species inhabiting such sediments can be roughly divided into two groups by life strategies, i.e. non- motile epipsammic species which are usually attached to sand grains, and motile epipelic species which can locate at different depth in the sediment to optimize light exposure 164,165 . Epipelic species bear a slit in the perapical plane of the frustule valve (named raphe), enabling movement inside the sediment by excretion of mucilage 166 .

Light attenuated differently in sandy and muddy sediments, i.e. scalar irradiance was significantly enhanced at sub-surface positions in the sandy sediment, while light attenuation coefficients were higher in the muddy sediment where light was attenuated exponentially by depth (Manuscript 4, Fig. 2). Although lower light acclimation of photosynthesis was observed in the muddy sediment which was dominated by epipelic species, net photosynthesis was significantly higher in muddy sediments compared to sandy sediments, which was assembled of same fractions of epipsammic and epipelic diatom species (Manuscript 4, Tab. 2, 3; Fig. 6). Such differences in the regulation of photosynthesis were explained by the contrasting life- strategies 167 , i.e. movement enabled epipelic species to position at optimal light conditions for efficient light harvesting and photoprotection 168 .

To evaluate whether photosynthesis was also affected by photonic structures in the frustule, we extracted epipelic diatoms from muddy sediment and examined the frustule optical properties of different species. We observed that the photonic structures and the optical properties of valves from different epipelic species varied, i.e. valves of the predominant species Gyrosigma fasciola appeared transparent when light was transmitted through the valve, while valves of Gyrosigma balticum and Pleurosigma angulatum interacted with shorter wavelengths in the blue spectral range (Manuscript 5, Fig. 2). We furthermore observed that all three tested species scattered blue light in dark field microscopy; similar to the observations made on valves of the centric species C. granii as described earlier. However, we observed that these phenomena were restricted to different angles of light incidence, i.e. valves of Gyrosigma balticum appeared blue at more acute angles of incidence, while valves of Gyrosigma fasciola and Pleurosigma angulatum appeared blue at more obtuse angles of incidence (Manuscript 5, Fig. 3). While conservation of such phenomena in all tested species could indicate an ecological advantage for diatoms, angle dependencies could point out that species may follow different strategies. For example, some species might position at diffusive light fields at subsurface locations, while other species may locate on the sediment surface where incident sunlight can be more collimated. A diatom suspension extracted from the same sediment confirmed that blue light induces

25

higher electron transport rates at photosystem II, while non-photochemical quenching was also higher at blue light compared to red light (Manuscript 5, Fig. 4). Interestingly, shorter wavelength in the visible spectrum of light attenuated more inside the sediment than longer wavelengths (Manuscript 5, Fig 5), giving rise to the assumption that the observed scattering properties of valves from epipelic species could play a role in enhancement of photosynthetically more productive blue light inside the sediment. We therefore measured scalar irradiance in a slurry mix of oxidized frustules and observed that light in the blue spectral range was indeed enhanced by up to 120% compared to downwelling irradiance (Manuscript 5, Fig. 6). We concluded that the photonic structures in the frustule of epipelic diatom species compensate for higher attenuation of blue light inside the sediment, facilitating efficient diatom photosynthesis in benthic habitats. We hypothesize that variation of photonic structures and the underlying optical properties of diatom frustules can be explained by niche differentiation in response to differences in the local light climate.

26 3. Main results and discussion

Figure 7: Interference of diurnal solar irradiance and tidal cycles in estuarine intertidal mudflats.

Estuarine intertidal mudflats are often dominated by motile (epipelic) diatom species. During high tide, epipelic species migrate into the sediment to avoid drift and predation. Upwards migration is controlled by combined low tide and light stimuli. Under these conditions photosynthesis might be more efficient compared to conditions at which diatoms locate at subsurface positions. Due to interference of diurnal and tidal cycles, the time span of efficient photosynthesis varies from day to day. Photosynthesis activity of microphytobenthos in intertidal mudflats is affected by the indicated parameters, i.e. amplitude of solar irradiance, tidal rhythm, and sediment altitude in meters above mean sea level (MAMSL).

Tidal data were derived from Tabuademares.com.

27

28 3. Main results and discussion

4. Conclusion and outlook

The current thesis presents first experimental evidence for cellular light propagation mechanisms to stimulate photosynthesis in live cells of the centric diatom species C. granii . Distribution of photosynthetically active radiation could be facilitated by waveguiding in plane of the valve, while propagating light could be coupled to chloroplasts located within the evanescent field. Forward scattering of blue light observed in dark field microscopy was also observed in valves of different pennate diatom species, suggesting an ecological advantage of such phenomena for diatoms, while angular variation between species could point out different life-strategies in response to the light climate, i.e. , diffusive or collimated light climates and light climates with different spectral characteristics. Frustules of benthic diatoms enhanced blue light against high attenuation of blue light inside the sediment. In all tested diatom species, blue light induced higher rates of electron transport at photosystem II, while it also induced more non-photochemical quenching. We suggest that the photonic structures in the frustule of diatoms inherit optical properties that modulate the light climate locally in favor for efficient diatom photosynthesis. Such properties include e.g. waveguiding and redistribution of photosynthetically productive radiation over the cell, as well as spectral enrichment of photosynthetically more productive blue light inside the cell. The frustule of diatoms is thereby an outstanding example of structural adaptation to light with potential implications upon diatom photosynthesis.

Although structure-based optical phenomena are surprisingly widespread in plants and other phototrophic organisms 93 , only few examples have been directly linked to the physiological performance, such as iridescent coloration of flowers attracting animal pollinators 169 . Little attention has been given to the potential implications of such phenomena upon photosynthesis, perhaps because they are often invisible to the naked eye or overcasted by the optical properties of pigments 101 . Photonic structures and photopigments are often located within the same, photosynthetically active, cellular components, such as in thylakoids which may form membrane stacks that can be regarded to as photonic multilayer structures 96 . Although it is not yet entirely clarified why such structures exist 170 , it has been observed that thylakoid stacking is induced under low light conditions in plants 171 . Interestingly, thylakoids are in diatoms always stacked in layers of three independently from the light intensity during growth 26 . In conclusion, more attention should be given to the photonic structures in plants and algae in respect to their potential implications upon photosynthesis and other physiological processes. The diatom frustule might be one example linking the biophotonic properties of nano-structure to photobiology, but similar links might exist in other biomineralizing microalgae, such as in calcifying coccolihophores. Exploration of biological photonic structures and their implications upon the physiology of organisms may also expand knowledge in the increasing field of bio-inspired optical materials 172 . For now, we conclude that the optical properties generated by photonic structures in the silicate frustule have the potential to improve photosynthesis of diatoms.

29

30

5. References

1. Anonymous. Two letters from a Gentleman in the Country, relating to Mr. Leeuwenhoek’s Letter in Transaction, no. 283. Philos. Trans. R. Soc. London B 23, 1494-1498-1501 (1702).

2. Round, F. E., Crawford, R. M. & Mann, D. G. The Diatoms: Biology and Morphology of the Genera. Cambridge Univ. Press. Cambridge, UK, 1990 (1990).

3. Werner, D. The biology of diatoms. University of California Press, Berkeley and New York (1977).

4. Fuhrmann, T., Landwehr, S., El Rharbi-Kucki, M. & Sumper, M. Diatoms as living photonic crystals. Appl. Phys. B Lasers Opt. 78, 257–260 (2004).

5. Tommasi, E. De. Light manipulation by single cells: the case of diatoms. J. Spectrosc. 2016, 1–13 (2016).

6. Ellegaard, M. et al. The fascinating diatom frustule – can it play a role for attenuation of UV radiation? J. Appl. Phycol. 28, 3295–3306 (2016).

7. Malviya, S. et al. Insights into global diatom distribution and diversity in the world’s ocean. Proc. Natl. Acad. Sci. 113, E1516–E1525 (2016).

8. Medlin, L. K. Evolution of the diatoms: major steps in their evolution and a review of the supporting molecular and morphological evidence. Phycologia 55, 79–103 (2016).

9. Ducklow, H. W., Steinberg, D. K. & Buesseler, K. O. Upper ocean carbon export and the biological pump. Oceanography 14, 50–58 (2001).

10. Field, C. B., Behrenfeld, M. J., Randerson, J. T. & Falkowski, P. Primary production of the biosphere: integrating terrestrial and oceanic components. Science (80-. ). 281, 237–240 (1998).

11. Geider, R. J. et al. Primary productivity of planet earth: biological determinants and physical constraints in terrestrial and aquatic habitats. Glob. Chang. Biol. 7, 849–882 (2001).

12. Falkowski, P. G. & Godfrey, L. V. Electrons, life and the evolution of Earth’s oxygen cycle. Philos Trans R Soc L. B Biol Sci 363, 2705–2716 (2008).

13. Armbrust, E. V. The life of diatoms in the world’s oceans. Nature 459, 185–192 (2009).

14. Guillou, L., Medlin, L. K., Oce, O. & Vaulot, D. Bolidomonas: A new genus with two species belonging to a new algal class, the (heterokonta). 381, 368–381 (1999).

15. Bowler, C. et al. The Phaeodactylum genome reveals the evolutionary history of diatom genomes. Nature 456, 239–244 (2008).

16. Sorhannus, U. A nuclear-encoded small-subunit ribosomal RNA timescale for diatom evolution. Mar. Micropaleontol. 65, 1–12 (2007).

17. Cavalier-Smith, T. Chloroplast evolution: Secondary symbiogenesis and multiple losses. Curr. Biol. 12, 62–64 (2002).

18. Pawlowski, J. et al. CBOL Protist working group: Barcoding eukaryotic richness beyond the animal, plant, and fungal kingdoms. PLoS Biol. 10, (2012).

19. Leipe, D. D. et al. The stramenopiles from a molecular prospective: 16S-like rRNA sequences from Labyrinthuloidesminuta and Cafeteria roenbergensis. Phycologia 33, 369–377 (1994).

20. von Stosch, H. A. Oogamy in a centric diatom. Nature 165, 531–532 (1950).

21. Bhattacharya, D. et al. Algae containing chlorophylls a + c are paraphyletic - molecular evolutionary analysis of

31

the chromophyta. Evolution (N. Y). 46, 1801–1817 (1992).

22. Maneveldt, G. W. & Keats, D. W. Chromista. In: Nature Encyclopedia of Life Sciences. Nature Publishing Group . 1-7 (2003)

23. Bhattacharya, D. & Medlin, L. K. The phylogeny of plastids: a review based on comparisons of small-subunit ribosomal RNA coding regions. J. Phycol. 31, 489–498 (1995).

24. Cavalier-Smith, T. Kingdoms Protozoa and Chromista and the eozoan root of the eukaryotic tree. Biol. Lett. 6, 342–345 (2010).

25. Anderson, J. M. & Aro, E. M. Grana stacking and protection of Photosystem II in thylakoid membranes of higher plant leaves under sustained high irradiance: An hypothesis. Photosynth. Res. 41, 315–326 (1994).

26. Lepetit, B., Goss, R., Jakob, T. & Wilhelm, C. Molecular dynamics of the diatom thylakoid membrane under different light conditions. Photosynth. Res. 111, 245–257 (2012).

27. Wilhelm, C. et al. The regulation of carbon and nutrient assimilation in diatoms is significantly different from green algae. Protist 157, 91–124 (2006).

28. Taiz, L. & Zeiger, E. Plant physiology and development. Sinauer Associates (2014).

29. Schellenberger Costa, B. et al. Blue light is essential for high light acclimation and photoprotection in the diatom Phaeodactylum tricornutum . J. Exp. Bot. 64, 483–493 (2013).

30. Kuczynska, P., Jemiola-Rzeminska, M. & Strzalka, K. Photosynthetic pigments in diatoms. Mar. Drugs 13, 5847–5881 (2015).

31. Lichtenthaler, H. K. [34] Chlorophylls and carotenoids: Pigments of photosynthetic biomembranes. Methods Enzymol. 148, 350–382 (1987).

32. Siefermann-Harms, D. The light-harvesting and protective functions of carotenoids in photosynthetic membranes. Physiol. Plant. 69, 561–568 (1987).

33. Förster, T. Energiewanderung und Fluoreszenz. Naturwissenschaften 33, 166–175 (1948).

34. Forbush, B. & Kok, B. Reaction between primary and secondary electron acceptors of photosystem II of photosynthesis. Biochim. Biophys. Acta 162, 243–53 (1968).

35. Bresinsky, A., Körner, C., Kadereit, J. W., Neuhaus, G. & Sonnewald, U. Strasburger’s Plant Sciences. Springer (2013).

36. Turan, S. Light Acclimation in Plants : Photoinhibition and Photoprotection. Adv. Biores. 3, 90–94 (2012).

37. Suorsa, M. et al. Light acclimation involves dynamic re-organization of the pigment-protein megacomplexes in non-appressed thylakoid domains. Plant J. 84, 360–373 (2015).

38. Strzepek, R. F. & Harrison, P. J. Photosynthetic architecture differs in coastal and oceanic diatoms. Nature 431, 689–692 (2004).

39. Williams, W. P. & Allen, J. F. State 1/State 2 changes in higher plants and algae. Photosynth. Res. 13, 19–45 (1987).

40. Mullineaux, C. W. & Allen, J. F. State 1-State 2 transitions in the cyanobacterium Synechococcus 6301 are controlled by the redox state of electron carriers between Photosystems I and II. Photosynth. Res. 23, 297–311 (1990).

41. Chen, J., Wang, P., Mi, H.-L., Chen, G.-Y. & Xu, D.-Q. Reversible association of ribulose-1, 5-bisphosphate carboxylase/oxygenase activase with the thylakoid membrane depends upon the ATP level and pH in rice without heat stress. J. Exp. Bot. 61, 2939–50 (2010).

32 5. References

42. Carmo-Silva, A. E. & Salvucci, M. E. The regulatory properties of rubisco activase differ among species and affect photosynthetic induction during light transitions. Plant Physiol. 161, 1645–1655 (2013).

43. Zhang, N. & Portis, a R. Mechanism of light regulation of Rubisco: a specific role for the larger Rubisco activase isoform involving reductive activation by thioredoxin-f. Proc. Natl. Acad. Sci. U. S. A. 96, 9438–43 (1999).

44. Michels, A. K., Wedel, N. & Kroth, P. G. Diatom plastids possess a phosphoribulokinase with an altered regulation and no oxidative pentose phosphate pathway. Plant Physiol. 137, 911–920 (2005).

45. Demming-Adams, B., Adams, W. W. & Mattoo, A. K. Photoprotection, photoinhibition, gene regulation and environment. Advances in Photosynthesis and Respiration (2006).

46. Bilger, W. & Björkman, O. Role of the xanthophyll cycle in photoprotection elucidated by measurements of light-induced absorbance changes, fluorescence and photosynthesis in leaves of Hedera canariensis. Photosynth. Res. 25, 173–185 (1990).

47. Lohr, M. & Wilhelm, C. Algae displaying the diadinoxanthin cycle also possess the violaxanthin cycle. Proc. Natl. Acad. Sci. U. S. A. 96, 8784–8789 (1999).

48. Schumann, A., Goss, R., Jakob, T. & Wilhelm, C. Investigation of the quenching efficiency of diatoxanthin in cells of Phaeodactylum tricornutum (Bacillariophyceae) with different pool sizes of xanthophyll cycle pigments. Phycologia 46, 113–117 (2007).

49. Stauber, J. L. & Jeffrey, S. W. Photosynthetic pigments in fifty-one species of marine diatoms. J. Phycol. 24, 158–172 (1988).

50. Goss, R. & Jakob, T. Regulation and function of xanthophyll cycle-dependent photoprotection in algae. Photosynth. Res. 106, 103–122 (2010).

51. Ruban, A. et al. The super-excess energy dissipation in diatom algae: Comparative analysis with higher plants. Photosynth. Res. 82, 165–175 (2004).

52. Dietzel, L., Bräutigam, K. & Pfannschmidt, T. Photosynthetic acclimation: State transitions and adjustment of photosystem stoichiometry - Functional relationships between short-term and long-term light quality acclimation in plants. FEBS J. 275, 1080–1088 (2008).

53. Maxwell, K. & Johnson, G. N. Chlorophyll fluorescence--a practical guide. J. Exp. Bot. 51, 659–68 (2000).

54. Apel, K. & Hirt, H. Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu. Rev. Plant Biol. 55, 373–399 (2004).

55. Torres, M. A., Jones, J. D. G. & Dangl, J. L. Reactive oxygen species signaling in response to pathogens. Plant Physiol. 141, 373–378 (2006).

56. Foyer, C. H., Lelandais, M. & Kunert, K. J. Photooxidative stress in plants. Physiol. Plant. 92, 696–717 (1994).

57. Alscher, R. G. Biosynthesis and antioxidant function of glutathione in plants. Physiol. Plant. 77, 457–464 (1989).

58. Waring, J., Klenell, M., Bechtold, U., Underwood, G. J. C. & Baker, N. R. Light-induced responses of oxygen photoreduction, reactive oxygen species production and scavenging in two diatom species. J. Phycol. 46, 1206– 1217 (2010).

59. Aro, E., Virgin, I. & Andersson, B. Photoinhibition of Photosystem II . Inactivation , protein damage and turnover. 1143, 113–134 (1993).

60. Li, G., Woroch, A. D., Donaher, N. A., Cockshutt, A. M. & Campbell, D. A. A Hard day’s night: Diatoms continue recycling photosystem II in the dark. Front. Mar. Sci. 3, 1–10 (2016).

33

61. Schenk, H. J. On the focusing of sunlight by ocean waves. J. Opt. Soc. Am. 47, 653–657 (1957).

62. Max, N. L. Light diffusion through clouds and haze. Comput. Vision, Graph. Image Process. 33, 280–292 (1986).

63. Kirk, J. T. O. Light and Photosynthesis in Aquatic Ecosystems. Cambridge University Press (2010).

64. Gilbert, M., Wilhelm, C. & Richter, M. Bio-optical modelling of oxygen evolution using in vivo fluorescence: Comparison of measured and calculated photosynthesis/irradiance (P-I) curves in four representative phytoplankton species. J. Plant Physiol. 157, 307–314 (2000).

65. Stomp, M., Huisman, J., Stal, L. J. & Matthijs, H. C. P. Colorful niches of phototrophic microorganisms shaped by vibrations of the water molecule. ISME J. 1, 271–282 (2007).

66. Wolken, J. J. Light detectors, photoreceptors, and imaging systems in nature. Oxford University Press (1995).

67. Depauw, F. A., Rogato, A., D’Alcalá, M. R. & Falciatore, A. Exploring the molecular basis of responses to light in marine diatoms. J. Exp. Bot. 63, 1575–1591 (2012).

68. Ishikawa, M. et al. Distribution and phylogeny of the blue light receptors aureochromes in . Planta 230, 543–552 (2009).

69. Takahashi, F. et al. AUREOCHROME, a photoreceptor required for photomorphogenesis in stramenopiles. Proc. Natl. Acad. Sci. U. S. A. 104, 19625–19630 (2007).

70. Briggs, W. R. et al. The phototropin family of photoreceptors. Plant Cell 13, 993–997 (2001).

71. Strasser, B., Sánchez-Lamas, M., Yanovsky, M. J., Casal, J. J. & Cerdán, P. D. Arabidopsis thaliana life without phytochromes. Proc. Natl. Acad. Sci. U. S. A. 107, 4776–4781 (2010).

72. Stephenson, P. G. & Terry, M. J. Light signalling pathways regulating the Mg-chelatase branchpoint of chlorophyll synthesis during de-etiolation in Arabidopsis thaliana . Photochem. Photobiol. Sci. 7, 1243–1252 (2008).

73. Galland, P. & Russo, V. E. Light and dark adaptation in Phycomyces phototropism. J. Gen. Physiol. 84, 101– 118 (1984).

74. Kinoshita, T., Doi, M. & Suetsugu, N. Regulation of stomatal opening. Nature 414, 0–4 (2001).

75. Kasahara, M., Kagawa, T., Sato, Y., Kiyosue, T. & Wada, M. Phototropins mediate blue and red light-induced chloroplast movements in Physcomitrella patens . Plant Physiol. 135, 1388–1397 (2004).

76. Takagi, S. Actin-based photo-orientation movement of chloroplasts in plant cells. J. Exp. Biol. 206, 1963–1969 (2003).

77. Park, Y. I., Chow, W. S. & Anderson, J. M. Chloroplast movement in the shade plant Tradescantia albiflora helps protect photosystem II against light stress. Plant Physiol. 111, 867–875 (1996).

78. Kasahara, M., Kagawa, T. & Oikawa, K. Chloroplast avoidance movement reduces photodamage in plants. Lett. to Nat. 420, 829–832 (2002).

79. Wada, M., Kagawa, T. & Sato, Y. Chloroplast movement. Annu. Rev. Plant Biol. 54, 455–68 (2003).

80. Luesse, D. R., DeBlasio, S. L. & Hangarter, R. P. Integration of Phot1, Phot2, and PhyB signalling in light- induced chloroplast movements. J. Exp. Bot. 61, 4387–97 (2010).

81. DeBlasio, S. L., Mullen, J. L., Luesse, D. R. & Hangarter, R. P. Phytochrome modulation of blue light-induced chloroplast movements in Arabidopsis. Plant Physiol. 133, 1471–1479 (2003).

82. Gabry ś, H. Blue light-in duced ori en ta tion move ments of chloroplasts in higher plants : Recent progress in the study of their mechanisms. 26, 473–478 (2004).

34 5. References

83. Kiefer, D. a. Chlorophyll a fluorescence in marine centric diatoms: Responses of chloroplasts to light and nutrient stress. Mar. Biol. 23, 39–46 (1973).

84. Chen, S.-T. & Li, C.-W. Relationships between the movements of chloroplasts and cytoskeletons in diatoms. Bot. Mar. 34, 505–511 (1991).

85. Furukawa, T., Watanabe, M. & Shihira-Ishikawa, I. Green- and blue-light-mediated chloroplast migration in the centric diatom Pleurosira laevis . Protoplasma 203, 214–220 (1998).

86. Ruban, A. V. Plants in light. Commun. Integr. Biol. 2, 50–55 (2009).

87. Flowers, T. J. The mechanism of salt tolerance in halphytes. Annu. Rev. Plant Physiol. 28, 89–121 (1977).

88. Shabala, S., Bose, J. & Hedrich, R. Salt bladders: Do they matter? Trends Plant Sci. 19, 687–691 (2014).

89. Berry, J. & Bjorkman, O. Photosynthetic response and adaptation to temperature in higher plants. Annu. Rev. Plant Physiol. 31, 491–543 (1980).

90. Endler, J. The color of light in forests and its implications. Ecol. Monogr. 63, 1–27 (1993).

91. Binkley, D., Campoe, O. C., Gspaltl, M. & Forrester, D. I. Light absorption and use efficiency in forests: Why patterns differ for trees and stands. For. Ecol. Manage. 288, 5–13 (2013).

92. Coopman, R. E. et al. Light energy partitioning in photosystems I and II during development of Nothofagus nitida growing under different light environments in the Chilean evergreen temperate rain forest. Trees 24, 247– 259 (2009).

93. Vukusic, P. & Sambles, J. R. Photonic structures in biology. Nature 424, 852–5 (2003).

94. Holt, A. L., Vahidinia, S., Gagnon, Y. L., Morse, D. E. & Sweeney, A. M. Photosymbiotic giant clams are transformers of solar flux. J. R. Soc. Interface 11, 20140678 (2014).

95. Sheue, C. R. et al. Bizonoplast, a unique chloroplast in the epidermal cells of microphylls in the shade plant Selaginella erythropus (Selaginellaceae). Am. J. Bot. 94, 1922–1929 (2007).

96. Jacobs, M., Lopez-Garcia, M., Phrathep, O.-P., Lawson, T., Oulton R. & Whitney H.M. Photonic crystal structure of Begonia chloroplasts enhances photosynthetic efficiency. Nat. Plants 2, 1–16 (2016). doi:10.1038/nplants.2016.162

97. De Tommasi, E. et al. Multi-wavelength study of light transmitted through a single marine centric diatom. Opt. Express 18, 12203–12212 (2010).

98. Romann, J. et al. Wavelength and orientation dependent capture of light by diatom frustule nanostructures. Sci. Rep. 5, 17403 (2015).

99. Yamanaka, S. et al. Optical properties of diatom silica frustule with special reference to blue light. J. Appl. Phys. 103, 1–6 (2008).

100. Parker, A. R. 515 Million years of structural colour. J. Opt. A Pure Appl. Opt. 2, R15–R28 (2000).

101. Sun, J., Bhushan, B. & Tong, J. Structural coloration in nature. RSC Adv. 3, 14862 (2013).

102. Land, M. F. The physics and biology of animal reflectors. Prog. Biophys. Mol. Biol. 24, 75–106 (1972).

103. Rama Swamy, S. X-ray analysis of the structure of iridescent shells. Proc. Indian Acad. Sci. - Sect. A 1, 19–30 (1935).

104. Parker, A. R. & Townley, H. E. Biomimetics of photonic nanostructures. Nat. Nanotechnol. 2, 347–353 (2007).

105. Swartzlander, G. A. Jr. & Gbur, G. Singular optical phenomena in nature. Proc. SPIE 7057, 705703-705703–9 (2008).

35

106. Parker, A. R. & Mcphedran, R. C. Aphrodite’s iridescence. 409, 36–37 (2001).

107. Seago, A. E., Brady, P., Vigneron, J.-P. & Schultz, T. D. Gold bugs and beyond: A review of iridescence and structural colour mechanisms in beetles (Coleoptera). J. R. Soc. Interface 2, 165–184 (2009).

108. van der Kooi, C. J. et al. Iridescent flowers? Contribution of surface structures to optical signaling. New Phytol. 203, 667–673 (2014).

109. van der Kooi, C. J., Dyer, A. G. & Stavenga, D. G. Is floral iridescence a biologically relevant cue in plant- pollinator signaling? New Phytol. 205, 18–20 (2015).

110. Gibson, J. J. The ecological approach to visual perception: classic edition. Boston, MA: Houghton-Mifflin (2014).

111. Fabricant, S. A., Exnerová, A., Ježová, D. & Štys, P. Scared by shiny? The value of iridescence in aposematic signalling of the hibiscus harlequin bug. Anim. Behav. 90, 315–325 (2014).

112. Dakin, R. & Montgomerie, R. Eye for an eyespot: How iridescent plumage ocelli influence peacock mating success. Behav. Ecol. 24, 1048–1057 (2013).

113. Zi, J. et al. Coloration strategies in peacock feathers. Proc. Natl. Acad. Sci. 100, 12576–12578 (2003).

114. Lienau, C., Noginov, M. a & Lon čar, M. Light–matter interactions at the nanoscale. J. Opt. 16, 110201 (2014).

115. Kinoshita, S., Yoshioka, S. & Kawagoe, K. Mechanisms of structural colour in the Morpho butterfly: cooperation of regularity and irregularity in an iridescent scale. Proc. Biol. Sci. 269, 1417–1421 (2002).

116. Kinoshita, S., Yoshioka, S. & Miyazaki, J. Physics of structural colors. Reports Prog. Phys. 71, 76401 (2008).

117. Desikachary, F. A. S. & Dweltz, N. E. The chemical composition of the diatom frustule. Encycl. Soil Sci. 53, 213–217 (1961).

118. Yang, W., Lopez, P. J. & Rosengarten, G. Diatoms: self assembled silica nanostructures, and templates for bio/chemical sensors and biomimetic membranes. Analyst 136, 42–53 (2011).

119. Aitken, Z. H., Luo, S., Reynolds, S. N., Thaulow, C. & Greer, J. R. Microstructure provides insights into evolutionary design and resilience of Coscinodiscus sp. frustule. Proc. Natl. Acad. Sci. U. S. A. 113, 2017–2022 (2016).

120. Cox, E. J. Morphology, cell wall, cytology, ultrastructure and morphogenetic studies: Overview and specific observations. The Diatom World 19 , 21–46 (2011).

121. Stosch, H. An amended terminology of the diatom girdle. Nov. Hedwigia, Beih 54–55, 113–166 (1975).

122. Losic, D., Rosengarten, G., Mitchell, J. G. & Voelcker, N. H. Pore architecture of diatom frustules: potential nanostructured membranes for molecular and particle separations. J. Nanosci. Nanotechnol. 6, 982–989 (2006).

123. Su, Y., Lundholm, N. & Ellegaard, M. Long-term cultivation of the diatom Coscinodiscus granii at different light spectra: effects on frustule morphology. J. Appl. Phycol. 3, 1–5 (2017)

124. Schmid, A.-M. M. & Volcani, B. E. Wall morphogenesis in Coscinodiscus Wailesii Gran and Angst. I. Valve Morphology and Development of Its Architecture. Journal of Phycology 19, 387–402 (1983).

125. Hoover, R. B. & Hoover, M. J. Optical phenomena in diatoms. Arkansas Acad. Sci. Proc. 1, 8–10 (1970).

126. Maibohm, C., Friis, S. M. M., Ellegaard, M. & Rottwitt, K. Interference patterns and extinction ratio of the diatom Coscinodiscus granii . Opt. Express 23, 9543 (2015).

127. De Stefano, L., Rea, I., Rendina, I., De Stefano, M. & Moretti, L. Lensless light focusing with the centric marine diatom Coscinodiscus walesii . Opt. Express 15, 18082–18088 (2007).

36 5. References

128. Gardner, J. V. & Burckle, L. H. Upper Pleistocene Ethmodiscus rex oozes from the eastern equatorial Atlantic. 21, 236–242 (2014).

129. Jewson, D. H. Size reduction, reproductive strategy and the life cycle of a centric diatom. Philos. Trans.: Biol. Sci. 336, 191–213 (1992).

130. Mann, D. G. Why didn´t Lund see sex in Asterionella ? A discussion of the diatom life cycle in nature. In: Algae and the aquatic environment. Biopress Bristol 83–412 (1988).

131. Chepurnov, V. A., Mann, D. G., Sabbe, K. & Vyverman, W. Experimental studies on sexual reproduction in diatoms. Int. Rev. Cytol. 237, 91–154 (2004).

132. Hense, I. & Beckmann, A. A theoretical investigation of the diatom cell size reduction-restitution cycle. Ecol. Modell. 317, 66–82 (2015).

133. Armbrust, E. V. & Chisholm, S. W. Patterns of cell size change in a marine centric diatom: Variability evolving from clonal isolates. Journal of Phycology 28, 146–156 (1992).

134. Edlund, M. B. & Stoermer, E. F. Ecological, evolutionary, and systematic significance of diatom life histories. J. Phycol. 33, 897–918 (1997).

135. Smetacek, V. S. Role of sinking in diatom life-hystory: ecological, evolutionary and geological significance. Mar. Biol. 84, 239–251 (1985).

136. Hildebrand, M. & Lerch, S. J. L. Diatom silica biomineralization: Parallel development of approaches and understanding. Semin. Cell Dev. Biol. 46, 27–35 (2015).

137. Schmid, A. M. M. & Schulz, D. Wall morphogenesis in diatoms: Deposition of silica by cytoplasmic vesicles. Protoplasma 100, 267–288 (1979).

138. Vrieling, E. G., Gieskes, W. W. C. & Beelen, T. P. M. Silicon deposition in diatoms: Control by the pH inside the silicon deposition vesicle. J. Phycol. 35, 548–559 (1999).

139. Kröger, N., Deutzmann, R. & Sumper, M. Polycationic peptides from diatom biosilica that direct silica nanosphere formation. Science (80-. ). 286, 1129–1132 (1999).

140. Brunner, E. et al. Chitin-based organic networks: An integral part of cell wall biosilica in the diatom Thalassiosira pseudonana . Angew. Chemie - Int. Ed. 48, 9724–9727 (2009).

141. Lechner, C. C. & Becker, C. F. W. Silaffins in silica biomineralization and biomimetic silica precipitation. Mar. Drugs 13, 5297–5333 (2015).

142. Dawson, P. A. Observations on the structure of some forms of Gomphonema parvulum Kütz. Br. Phycol. J. 7, 255–271 (1972).

143. Su, Y., Lundholm, N., Friis, S. M. M. & Ellegaard, M. Implications for photonic applications of diatom growth and frustule nanostructure changes in response to different light wavelengths. Nano Res. 8, 2363–2372 (2015).

144. Townley, H. E., Woon, K. L., Payne, F. P., White-Cooper, H. & Parker, A. R. Modification of the physical and optical properties of the frustule of the diatom Coscinodiscus wailesii by nickel sulfate. Nanotechnology 18, 295101 (2007).

145. Qin, T., Gutu, T., Jiao, J., Chang, C. & Rorrer, G. L. Biological Fabrication of Photoluminescent Nanocomb Structures by Metabolic Incorporation of Germanium into the Biosilica of the Diatom Nitzschia frustulum . 2, 1296-1304 (2008)

146. Vrieling, E. G. et al. Salinity-dependent diatom biosilicification implies an important role of external ionic strength. Proc. Natl. Acad. Sci. U. S. A. 104, 10441–10446 (2007).

147. Hildebrand, M. Prospects of manipulating diatom silica nanostructure. J. Nanosci. Nanotechnol. 5, 146–157

37

(2005).

148. Hale, M. S. & Mitchell, J. G. Functional morphology of diatom frustule microstructures: Hydrodynamic control of brownian particle diffusion and advection. Aquat. Microb. Ecol. 24, 287–295 (2001).

149. Hamm, C. E. et al. Architecture and material properties of diatom shells provide effective mechanical protection. Nature 421, 841–843 (2003).

150. Dossou, K. B. Large field enhancement obtained by combining Fabry-Perot resonance and Rayleigh anomaly in photonic crystal slabs. J. Opt. 19, 1–12 (2017).

151. Stanley, E. M. The refractive index of seawater as a function of temperature, pressure and two wavelengths. Deep. Res. Oceanogr. Abstr. 18, 833–840 (1971).

152. Trampe, E., Kolbowski, J., Schreiber, U. & Kühl, M. Rapid assessment of different oxygenic phototrophs and single-cell photosynthesis with multicolour variable chlorophyll fluorescence imaging. Mar. Biol. 158, 1667– 1675 (2011).

153. Rickelt, L. F., Lichtenberg, M., Trampe, E. C. L. & Kühl, M. Fiber-optic probes for small-scale measurements of scalar irradiance. Photochem. Photobiol. 92, 331–342 (2016).

154. Kühl, M. Optical microsensors for analysis of microbial communities. Methods Enzymol. 397, 166–199 (2005).

155. Ross, K. F. Measurement of the refractive index of cytoplasmic inclusions in living cells by the interference microscope. Nature 174, 836–837 (1954).

156. Geider, R. J. & La Roche, J. The role of iron in phytoplankton photosynthesis, and the potential for iron- limitation of primary productivity in the sea. Photosynth. Res. 39, 275–301 (1994).

157. Berg JM, Tymoczko JL, Stryer L, S. L. Biochemistry. Plant Systematics and Evolution (2007).

158. Laisk, A., Eichelmann, H., Oja, V. & Peterson, R. B. Control of cytochrome b6f at low and high light intensity and cyclic electron transport in leaves. Biochim. Biophys. Acta - Bioenerg. 1708, 79–90 (2005).

159. O’Reilly, J. E. et al. Ocean color chlorophyll algorithm for SeaWiFS. J. Geophys. Res. 103, 24937–24953 (1998).

160. Bricaud, A., Morel, A., Babin, M., Allali, K. & Claustre, H. Variations of light absorption by suspended particles with chlorophyll a concentrations for bio-optical models. J. Geophys. Res. 103, 31033–31044 (1998).

161. Goessling, J. W., Cartaxana, P. & Kühl, M. Photo-protection in the centric diatom Coscinodiscus granii is not controlled by chloroplast high-light avoidance movement. Front. Mar. Sci. 2, 1–10 (2016).

162. Su, Y. & Lundholm, N. Implications for photonic applications of diatom growth and frustule nanostructure changes in response to different light wavelengths Implications for photonic applications of diatom growth and frustule nanostructure changes in response to different light. Nano Res. 8, 2363-2372 (2015).

163. Palmer, J. D. Tidal rhythms: The clock control of the rhythmic physiology of marine organisms. Biol. Rev. 48, 377–418 (1973).

164. Round, F. E. the Epipsammon; a Relatively Unknown Freshwater Algal Association. Br. phycol. Bull 2, 456– 462 (1965).

165. Jenkerson, C. G. & Hickman, M. Interrelationships among the Epipelon, Epiphyton and Phytoplankton in a Eutrophic Lake. Int. Rev. der gesamten Hydrobiol. und Hydrogr. 71, 557–579 (1986).

166. Ruck, E. C. & Theriot, E. C. Origin and Evolution of the Canal Raphe System in Diatoms. Protist 162, 723–737 (2011).

167. Jesus, B. et al. Adaptations of microphytobenthos assemblages to sediment type and tidal position. Cont. Shelf Res. 29, 1624–1634 (2009).

38 5. References

168. Consalvey, M., Paterson, D. M. & Underwood, G. J. C. the ups and downs of life in a benthic biofilm: Migration of benthic diatoms. Diatom Res. 19, 181–202 (2004).

169. Glover, B. J. & Whitney, H. M. Structural colour and iridescence in plants : the poorly studied relations of pigment colour. Ann. Bot. 105, 505–511 (2010).

170. Trissl, H. W. & Wilhelm, C. Why do thylakoid membranes from higher plants form grana stacks? Trends Biochem. Sci. 18, 415–419 (1993).

171. Melis, A. Dynamics of photosynthetic membrane composition and function. BBA - Bioenerg. 1058, 87–106 (1991).

172. Yu, K., Fan, T., Lou, S. & Zhang, D. Biomimetic optical materials: Integration of nature’s design for manipulation of light. Prog. Mater. Sci. 58, 825–873 (2013).

173. Guiry, M. D. How many species of algae are there? J. Phycol. 1063, 1057–1063 (2012).

174. Kröger, N. & Poulsen, N. Diatoms-from cell wall biogenesis to nanotechnology. Annu. Rev. Genet. 42, 83–107 (2008).

175. Nassar, M.Z.A. & Khairy, H. M. Checklist of phytoplankton species in the Egyptian waters of the Red Sea and some surrounding habitats (1990-2010). Annu. Res. Rev. Biol. 4, 3566–3585 (2014).

176. Vilicic, D., Marasovic, I. & Miokovic, D. Checklist of phytoplankton in the eastern Adriatic Sea. Acta Bot. Croat. 61, 57–91 (2002).

177. Hällfors, G. Checklist of Baltic Sea phytoplankton species (including some heterotrophic protistan groups). Balt. Sea Environ. Proc. 95, 1–208 (204AD).

178. Board, B. E. Black Sea phytoplankton checklist. Available from http//phyto.bss.ibss.org.ua. pp. Accessed 19 April 2014 (2014).

179. Bursa, A. Phytoplankton in coastal waters of the arctic ocean at point Barrow, Alaska. Arctic 16, 238–262 (1963).

180. Hartley, B., Barber, H.G., Carter, J.R. & Sims, P. A. An atlas of British diatoms. Bristol Biopress Ltd. 1–106 (1996).

181. Guilloux, L. et al. An annotated checklist of Marine Phytoplankton taxa at the SOMLIT-Astan time series off Roscoff (Western English Channel, France): Data collected from 2000 to 2010. Cah. Biol. Mar. 54, 247–256 (2013).

182. Hoppenrath, M. A revised checklist of planktonic diatoms and dinoflagellates from Helgoland (North Sea, German Bight). Helgol. Mar. Res. 58, 243–251 (2004).

183. ME Ferrario, G Almandoz, S Licea, I. G. Species of Coscinodiscus (Bacillariophyta) from the Gulf of Mexico, Argentina and Antarctic waters: morphology. Nov. Hedwigia 133, 187–216 (2008).

184. Hasle, G. R. & Lange, C. B. Morphology and distribution of Coscinodiscus species from the Oslofjord, Norway, and the Skagerrak, North Atlantic. Diatom Res. 7, 37–68 (1992).

185. Cărăuş, I. & Ph, D. Algae of Romania: A distributional checklist of actual algae. Stud. şi Cercet. Univ. Bac ău (2012).

186. Aysel, V. Check-list of the freshwater algae of Turkey. J. Black Sea/Mediterranean Environ. 11, 1–124 (2005).

187. Afonso-Carrillo, J. Lista actualizada de las algas marinas de las islas Canarias, 2014. Las Palmas Elabor. para la Soc. Española Ficología (SEF). 1–64 (2014).

188. Mather, L., MacIntosh, K., Kaczmarska, I., Klein, G. & Martin, J. L. A checklist of diatom species reported (and presumed native) from Canadian coastal waters. Canadian Technical Report of Fisheries and Aquatic

39

Sciences (2010).

189. Bagheri, S. & Fallahi, M. Checklist of Phytoplankton Taxa in the Iranian Waters of the Caspian Sea. Casp. J. Environ. Sci. Casp. J. Env. Sci. Print. I.R. Iran [Research 12, 81–97 (2014).

190. Ruiyu, L. Checklist of biota of Chinese seas. Beijing Sci. Press. Acad. Sin. 1–1267 (2008).

191. Shao, K. T. Website. TaiBNET(Catalogue of Life in Taiwan) http://taibnet.sinica.edu.tw. (2014).

192. Garibotti, I. A., Ferrario, M. E., Almandoz, G. O. & Castaños, C. Seasonal diatom cycle in Anegada Bay, El Rincón estuarine system, Argentina. Diatom Res. 26, 227–241 (2011).

193. Eskinazi-Leça, E., Gonçalves da Silva Cunha, M. da G., Santiago, M.F., Palmeira Borges, G.C., Cabral de Lima, J.M., Da Silva, M.H., De Paula Lima, J. & Menezes, M. Bacillariophyceae. Catálogo de plantas e fungos do Brasil. Vol. 1. (2010).

194. Lozano-Duque, Y., Vidal, L.A. & Navas S, G. R. Listado de diatomeas (Bacillariophyta) registradas para el Mar Caribe colombiano. Bol. Investig. Mar. y Costeras 39, 83–116 (2011).

195. McCarthy, P. M. Census of Australian marine diatoms. Australian Biological Resources Study, Canberra. Canberra : Australian Biological Resources Study (2013). Available at: http://www.anbg.gov.au/abrs/Marine_Diatoms/index.html.

196. Harper, M.A., Cassie Cooper, V., Chang, F.H., Nelson, W.A. & Broady, P.A. Phylum : brown and golden-brown algae, diatoms, silicioflagellates, and kin. In: New Zealand inventory of biodiversity. Volume Three. Kingdoms Bacteria, Protozoa, Chromista, Plantae, Fungi. Canterbury University Press 114-163 (2012)

197. Ahmed, Z.U., Khondker, M., Begum, Z.N.T., Hassan, M.A., Kabir, S.M.H., Ahmad, M., Ahmed, A.T.A.. & Rhaman, A. K. A. Algae, Charophyta - Rhodophyta (Achnanthaceae - Vaucheriaceae). In: Encyclopedia of flora and fauna of Bangladesh. Dhaka: Asiatic Society of Bangladesh 9, (2009).

198. Maulood, B.K., Hassan, F.M., Al-Lami, A.A., Toma, J.J. & Ismail, A. M. Checklist of algal flora in Iraq. Minist. Environ. 94 (2013).

199. Lakkis, S. Flore et faune marines du Liban (Méditerranée orientale). Biologie, Biodiversité, Biogéographie. ARACNE editrice S.r.l. (2013).

40 6. Included manuscripts

41

42

6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms (in prep.)

Johannes W. Goessling 1,* , Yanyan Su 2, Paulo Cartaxana 3, Christian Maibohm 4, Lars Fledelius Rickelt 1,5 , Erik C. L. Trampe 1, Sandra Walby 2, Daniel Wangprasseurt 6, Xia Wu 7, Marianne Ellegaard 2, and Michael Kühl 1,8,*

1Marine Biology Section, Department of Biology, University of Copenhagen, Strandpromenaden 5, 3000 Helsingør, Denmark

2Section for Plant Glycobiology, Department for Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg, Denmark

3Departamento de Biologia & CESAM, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal

4International Iberian Nanotechnology Laboratory, 4715-330 Braga, Portugal

5Oxyguard International A/S, Farum Gydevej 64, DK-3520, Denmark / Zenzor, Krondrevet 31, DK-3140, Ålsgårde

6Department of Chemistry, University of Cambridge, Cambridge, UK

7Department of Chemistry, Paderborn University, Warburger Str. 100, 33098 Paderborn, Germany

8Climate Change Cluster, University of Technology Sydney, Australia

*Corresponding authors: Johannes W. Goessling and Michael Kühl

43

44 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Abstract

The optical properties of diatom silicate frustules have inspired photonics and nanotechnology research; however, whether light interaction with the nano-structuring of the frustule modulates photosynthesis of diatoms has remained unclear. Here we demonstrate that the optical properties of the frustule valves play a major role in light harvesting and photosynthesis in live cells of centric diatoms ( Coscinodiscus granii ). Microscale cellular mapping of photosynthesis around localized spot illumination demonstrated optical coupling of chloroplasts to the valve wall. Photonic structures of the three-layered valve wall facilitated light redistribution and efficient photosynthesis in cell regions distant from the directly illuminated area. The different porous structure of the two sides of the frustule valve also exhibited photon trapping and forward scattering of blue light, resulting in enhanced light absorption by diatom antennae pigments and a highly efficient photosystem II activity. Photonic structures in the diatom frustule may thereby have driven evolution and species diversification of diatoms.

45

46 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Introduction

Coloration generated by biological photonic structures often functions as sexual trigger, camouflage or warning signal 1–3. While most research have focused on how nano/micro-structures modulate optical properties of such biological materials 4, biological photonic structures can also affect physiological processes in plants and other aquatic phototrophic organisms 5–7. Recently, it was proposed that the brilliant blue iridescence caused by specialized chloroplasts in Begonia leaves can improve photosynthetic quantum efficiencies under low light conditions 8. On tropical coral reefs, outer tissue layers of giant clams harbor specialized iridescent protein stacks acting as Bragg reflectors focusing photosynthetically active radiation into deeper tissue regions harboring photosymbiotic microalgae, while screening out harmful radiation 9. Whether biological photonic structures also modulate light harvesting and photosynthetic efficiency in mineral depositing microalgae such as coccolihophores and diatoms remains unknown.

Diatoms are unicellular microalgae that are abundant in oceanic, benthic and freshwater environments, where they are important primary producers and main contributors to global photosynthetic carbon fixation 10,11 . Diatoms live in highly variable light environments spanning from seasonal variations to ms-long bursts of excess light via wave focusing 12,13 . Irradiance rapidly attenuates as a function of water depth due to scattering and absorption by water molecules, dissolved substances, biomass and abiotic particulate matter 14 . The ability of diatoms to adapt to a variable light environment is thought to be closely related to their high photosynthetic productivity and evolutionary success 15–17 . However, the photobiological mechanisms facilitating such photosynthetic flexibility and efficiency in diatoms remain largely unknown. Oceanic phototrophs often exhibit effective light harvesting of the blue-green part of the photosynthetically active radiation spectrum (PAR, wavelength λ = 400-700 nm), which penetrates deeper into oligotrophic waters 18 . While low light potentially limits photosynthesis, excess light can harm the organism due to impairment of photosynthesis and the formation of reactive oxygen species 19 . To prevent over-reduction of the photosynthetic electron transport chain, diatoms use effective energy dissipation mechanisms such as the xanthophyll cycle, which is controlled by blue light photoreceptors 20–22 . While blue and red light photoreceptor proteins are genuine to all photosynthetic organisms, diatoms employ an additional class of blue light photoreceptors called aureochromes 23–25 . These photoreceptors are essential in the perception of circadian and annual rythms and further control checkpoints in the diatom life cycle such as proliferation and reproduction 26 .

The >10,000 diatom species exhibit a large variety in size, shape and abundance 27 . Structural features of the frustule, i.e., the amorphous silicate cell walls surrounding the diatom cell, are used to classify diatoms. The frustule is composed of two valves that fit into each other like Petri dishes, each valve being connected with one or more girdle bands encircling the overlapping region (Fig. 1 a, b). Pennate diatom species are enclosed

47 in a linear or oval shaped frustule with a bilaterally symmetric, featherlike ornamental pattern, while frustules of centric diatom species exhibit radial symmetry with a line of pores radiating out from a central ring. The valve wall exhibits a complex three-dimensional asymmetric architecture with different structures in the micro- to nano-meter range (Fig. 1 c, d). The exterior and interior surfaces of the valve (valve orientations) have different ornamental design. In some centric diatoms the exterior valve surface (EVS) is perforated by small cribrum pores ( cribellum ) forming a sieve like structure. The cribellum is located over a hexagonal chamber in the central valve wall (CVW) which then opens into a single large areola on the interior valve surface (IVS), i.e. the foramen (pl. foramina )28 . The frustule provides mechanical stability and may protect the diatom against predation 29 , while the pores facilitate the diffusive exchange of gases, nutrients and other metabolites between the cell and the surrounding seawater 30 .

Photonic crystal structures in the diatom valve formed by the periodically arranged pore structures interact with light in the photosynthetically active wavelength range (Fig. 1) 28,31–33 . Based on optical measurements on isolated, cleaned valves in air, a range of studies have proposed links between the optical properties of frustules and the high photosynthetic productivity of diatoms 31,34–36 . Such hypotheses have hitherto remained speculative due to lack of experiments on live diatoms and frustules in water. We now show the differences of frustule optical properties in air and water, and we provide the first experimental evidence showing that the in vivo photosynthetic performance of intact living diatom cells is indeed shaped by the optical properties of their frustule valves.

48 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Results

Structure of the silicate frustule, and live diatom cell organization

Chloroplasts in live C. granii cells are evenly distributed in close proximity to the valve wall (Fig. 1a). The C. granii frustule comprises two slightly concave valves, each one joined by one or more girdle bands (Fig. 1b), where the exterior valve surface exhibits a sieve-like structure composed of six cribrum pores (9±2 nm across; mean ± SD; n=3) arranged in a hexagonal pattern around one central pore (Fig. 1c, d). The central hexagonal valve chamber exhibits diagonal diameters over three sides of 1277±69 nm, surrounded by 23±8 nm thick walls. The foramina on the interior valve surface have a diameter of 580±31 nm.

Waveguiding in cleaned valves and optical coupling to chloroplasts in live cells

To demonstrate light propagation inside the 2D lattice of the valve wall, a spectrally filtered and tunable laser beam (wavelengths λ = 500-700 nm) was focused locally onto a ~10 µm wide area of a cleaned valve, while light was measured with a scalar irradiance microprobe at a position on the opposite site of the illuminated spot on the diatom (Fig. 2a, b). Control measurements were done on a glass slide using a similar distance between spot and microprobe. We found higher light propagation inside valves than in a glass slide irrespective of valve orientation (Fig. 2c). In a live cell, chlorophyll fluorescence (maximum at λ = 685 nm) emitted from the chloroplasts of individual cells was visible in the valve wall, when the entire living cell was illuminated by weak blue light ( λ = ~440 nm; Fig. 2d). The coupling of light between chloroplasts, valve, and remote stimulation of chlorophyll fluorescence via the valve in live diatoms was also measured with a scalar irradiance microprobe detecting chlorophyll fluorescence around λ = 685 nm, when it was touching the diatom valve distant to the incident white ~10 µm laser beam spot ( λ = 450-650 nm) (Fig. 2e).

Stimulation of photosynthesis by localized laser illumination

By combining spectrally filtered laser spot illumination with variable chlorophyll fluorescence microscopy (Suppl. Fig. 1), we could demonstrate that light propagation in the valve affected quenching of chlorophyll fluorescence in the entire living cell (Fig. 3a). When the actinic (photosynthetically active) laser light ( λ = 650±10 nm) was focused as a ~10 µm wide spot on the diatom cell, a typical fluorescence induction behavior with an initial rise of yield followed by subsequent quenching of chlorophyll fluorescence was observed over the entire diatom cell. Chlorophyll fluorescence was quenched over the entire diatom cell during the illumination phase and started to recover, when the actinic laser light was switched off. Full recovery to

49 initial values was observed when the cell was kept in the dark for 15 min after the actinic laser was switched off (Fig. 3b). The effective photosystem II (PSII) quantum yields also decreased in regions of the diatom cell without direct laser spot illumination, and decreased further as a function of the photon irradiance red actinic light ( λ = 650±10 nm)(Fig. 3c; Suppl. Fig. 2).

Angular light collection properties of diatom frustule valves

To test light propagation in the diatom valves at different angles of incident light, a single cleaned frustule valve was fused to the tip of a tapered field radiance probe placed on a turntable (Fig. 4a, b). A full description of the construction of the turntable can be found in Rickelt et al. (2016) 37 . The presence of a frustule valve on the fiber tip reduced the collected amount of light but increased the angle of light collection (Fig. 4c, d). The valve in air showed a significant influence of the angular collection of light into the fiber, but such effects were minuscule when measured in water. The orientation of the fused valve towards the light source is indicated.

Influence of valve orientation and host refractive index on frustule optical properties

Transmittance: Cleaned diatom valves from C. granii exhibited tremendous differences in the intense and spectral light reflection and transmission properties in air as compared to water (Fig. 5; mean±SD; n=9; shown in % of incident PAR). In air, similar light transmission was observed when light was incident on the exterior and the interior valve surface (27±9% and 29±7%, respectively) (Fig. 5a). In water, transmittance was 78±3% with light incident on the interior valve surface (Fig. 5b). In this arrangement, minimum transmission was 61±6% at λ = 457 nm, when light was first transmitted through the exterior valve surface and 76±5% at λ = 452 nm when light was first transmitted through the interior valve surface. In summary, transmittance was higher in water when light was incident on the interior valve surface, but no spectral differences were determined by valve orientation.

Reflectance: In air, valve reflectance was highest at shorter wavelengths, while no distinct differences in reflectance between the interior and exterior valve’s surfaces were seen (Fig. 5c). The reflectance of PAR was similar for valves measured in air on the exterior valve surface (29±3%) and on the interior valve surface (28±3%). Reflectance in water was 2% on the exterior and 4±1% on the interior valve surface (Fig. 5d). Minimum reflectance was 2% at λ = 641 nm on the exterior and 3% at λ = 604 nm on the interior valve surface. Maximum reflectance was 3±1% at λ = 799 nm on the exterior valve surface and 6±1% at λ = 800 nm on the interior valve surface.

50 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

The reflection and transmission data mentioned above were all measured with a low numerical aperture (NA 0.13; 4x) objective. When a higher numerical aperture objective was used (NA 0.5; 20x), transmission in water was 92±2% for light incident on the exterior valve surface, and 97±4% for light incident on the interior valve surface (Fig. 5e). The numerical aperture alters the angle at which light can enter the objective, whereby wavelength-specific differences in light scattering or focusing on the frustule valve became evident (Fig. 5f).

Scattering properties of valves and living cells

The wavelength-dependent scattering properties of rinsed valves were different between measurements performed in air and water. Measurements were performed with dark-field microscopy using dark stops of different sizes to alter the angle of incident light in a controlled way (Fig. 6). In cleaned valves measured in water, a pronounced forward scattering of blue light was observed under light incident at an angle of 15°. Living cells showed forward scattering of green-yellow light independent from the angle of incidence, while pale blue color was only observed in large cells (~200 µm) under light incident at a 15° angle of incidence. Measurements performed on micro-glass beads (50-500 µm in diameter) showed that the wavelength dependency of light scattering was neither affected by the angle of incidence nor by the medium, but by the different shape and nanostructures in the frustule valves (Suppl. Fig. 4).

Light attenuation, photo-pigmentation and functional light absorption in living diatom cells

Measurements of light loss due to scattering and absorption in individual diatom cells were used as a proxy for light attenuation, as measured with a field radiance probe oriented at a 45° angle relative to a live diatom cell. Maximum light attenuation (28±7%) was measured at λ = 437±2 nm (mean±SD; n=9), while a second absorption maximum of 18±6% was observed at λ = 672±6 nm (mean±SD; n=9) (Fig. 7a; Suppl. Fig. 3). These attenuation maxima followed the diatom photo-pigmentation (Fig. 7b), where the most abundant photopigments in C. granii were Chlorophyll (Chl) a (195.7±1.2 pmol cell -1; mean±SD; n=3), fucoxanthin -1 -1 -1 (90.0±2.1 pmol cell ), Chl c1 (13.1±0.2 pmol cell ) and Chl c2 (8.6±0.1 pmol cell ; Fig. 7c). The PSII accessory pigment β-carotene was only present in low amounts (7.8±0.2 pmol cell -1). Photoprotective pigments of the diadinoxanthin-cycle, i.e. , diadinoxanthin (17.0±2.2 pmol cell -1) and diatoxanthin (1.9±1.0 pmol cell -1) and low amounts of violaxanthin (0.8±0.0 pmol cell -1) were also found as well as trace amounts of zeaxanthin (not quantified). The wavelength dependent functional light absorption cross section of photosystem II, i.e. , Sigma(II) λ, was highest in the blue region of visible light (Fig. 7d; n=5).

51

52 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Discussion

It has been proposed that the periodic lattice structure of the silicate valve theoretically behaves as a 2D photonic crystal, enabling slab waveguiding inside the valve wall on a horizontal plane 32 . Laser spot illumination of valves in water indeed led to enhanced light intensities distant from the illuminated area, and such light propagation was independent from the surface orientation of the valve (Fig. 2a-c). It was further suggested that the close proximity of chloroplasts to the valve wall could allow for coupling of the evanescent field of light to chloroplasts, facilitated by large field penetration depth due to the low refractive contrast between silicate and cytoplasm 32 . Experimental evidence for such mechanism was seen by transmission of Chl a emission from chloroplasts to the silicate valve wall, after stimulation of Chl a fluorescence in a living diatom with actinic blue light (Fig. 2d, e). Such optical coupling affected diatom photosynthesis as evident from a change in variable Chl fluorescence throughout the cell upon local illumination of the frustule with an actinic red ~10 µm wide laser spot (Fig. 3a). The Chl fluorescence kinetics measured in single chloroplasts distant to the area illuminated by the laser showed a typical photosynthesis induction behavior 38,41 , when the actinic laser light was switched on (Fig. 3b). The effective PSII quantum yield decreased throughout the cell as a function of increasing photon irradiance in the local laser illuminated spot (Fig. 3c).

Light coupling and guiding of propagation modes can occur when light enters the valves as a waveguide at an angle of incidence larger than the critical angle for total internal reflection. Albeit our measurements were performed at normal incidence of the localized laser beam illumination, some angles of light diffracted on the porous structures may fulfil the condition for guided modes in the valve wall. In this sense, the frustule valve can be regarded as a 2D subwavelength diffraction grating 39 . To further investigate the light-collecting properties of the valve, we fused a cleaned frustule valve onto the 10 µm tip of a tapered optical fiber (Fig. 4a-d). Compared to the bare fiber probe, the presence of a diatom valve on the fiber tip increased the acceptance angle over which light was collected. The extracellular matrix associated with the valve on the exterior side in live cells, as well as a layer of acidic polysaccharides (diatotepum) between valve and plasma membrane 40 may further affect waveguiding properties. Scattering and reflection of light by cellular constituents in the cytoplasm, e.g. lipids or proteins, may also affect light propagation in vivo 4; however, our data provide first experimental evidence that the frustule provides optical mechanisms to homogenize light exposure evenly over the frustule with potential implications in regards to photosynthesis in the living cell 32 .

The asymmetric and different design of the valve surface could lead to different reflection/transmission and different leakages of guided modes at the exterior and interior valve surfaces. We observed differences in both light transmittance and reflectance dependent on whether wet valves were facing their exterior or interior surface to the incident light, while such differences were absent in the measurements on dry valves

53 (Fig. 5 a-d). We speculate that the lower transmittance and reflectance of light incident from the exterior valve side than from the interior side indicates optimized light capture of diatoms. The before mentioned measurements were done with a low numerical aperture objective. When we measured the transmittance with an objective of higher numerical aperture (thus collecting a larger fraction of scattered light), valves in water appeared almost transparent without the relative lower transmittance in the blue region as measured with the objective of lower numerical aperture (Fig. 5 e, f). This result indicates that blue light apparently is scattered more by the valves.

Scattering on the gratings of foramina has also been observed in valves of C. wailesii and C. centralis , where light was focused by constructive interference in a distance of ~100 µm behind the valve 31,41 . Scattering by the valves can become visible under dark field illumination. In this case, direct illumination is blocked and the optical impact from the angle of incidence and host refractive index of the surrounding medium is more obvious. Cleaned valves showed different colors in air and water, and the angle of incidence showed strong effects on the diffraction properties of cleaned valves (Fig. 6 a, b). At acute angles of incidence (15°), scattering of blue light was visible from valves in water. This blue light could also be seen in large live cells when the chloroplast density was low, while light of the yellow-green spectral range was visible from more abundant chloroplasts (Fig. 6c). The light with these wavelengths are less absorbed by the typical array of photopigments in diatoms, and can hence be scattered by the chloroplasts 42–44 . We observed that living intact C. granii cells absorbed more light in the blue wavelength range as compared to red light (Fig. 7 a-c) 45;46 . Diatoms contain carotenoids that broaden light harvesting to the green spectral range of light 47, but also help dissipate excessive energy at high photon flux densities in form of heat 48 . Hence, although energy-richer blue light is usually absorbed more, photosynthetic rates can be lower at blue light compared to that at red light 49 . We confirmed earlier results suggesting that the functional light absorption at PSII in C. granii is higher at blue light in comparison to other colors (Fig. 7d), leading to higher photosynthetic rates under blue light as observed in many diatom species 50–52 .

We conclude that photonic crystal structures in the diatom frustule mediate efficient light harvesting and photosynthesis in aquatic environments. Our experimental results suggest that the C. granii valve has photonic properties that mediate efficient photosynthesis by redistribution of light via guided modes generated by diffraction at the pores and by enhanced scattering of more productive blue light. Light propagation by guided modes in the frustule valve distributes light to the chloroplasts by 1) optical coupling to the evanescent field, and 2) leakage through the large foramina pores. Light propagation has been demonstrated e.g. in glass sponges 53 and in the tissue of symbiont-bearing corals, leading to an efficient distribution of actinic radiation for photosymbiotic microalgae 54 . It has also been demonstrated that coupling to the evanescent field from an artificial waveguide can stimulate the productivity of photosynthetic biofilms 55 . Biological photonic structures can thus modulate biological functions, such as increasing the

54 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

quantum efficiency in phototrophic organisms 56,57 . The photonic properties of diatom valves reported here might inspire the design of more efficient solar panels or algal based biofuel reactors enabling focusing of higher energy photons in the blue wavelength range 58 . There are probably many more, yet unknown, examples that link optical properties of biological structures to efficient light-use in nature, with the potential to further inspire research in the growing field of biological photonics.

55

56 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Materials and methods

Diatom growth conditions and preparation of cleaned valves

The diatom species C. granii (strain K-1834) was provided by the Scandinavian Culture Collection for Algae and Protozoa (now a part of NIVA culture collection of algae, https://niva-cca.no). Samples for physiological measurements were grown at constant temperature (15˚C) in L1 medium with 250 µM Na 2SiO 3 and a salinity of 30‰. The cultures were grown under a photon irradiance (wavelengths λ = 400-700 nm) of 100 µmol photons m -2 s-1 provided by white light LEDs (Osram, Munich, Germany) in a 16 h light / 8 h dark cycle. Samples were investigated in the late exponential phase after six consecutive days of growth. Frustules were harvested from 10 mL of diatom culture grown at 300 µmol photons m -2 s-1 under white light LEDs. Thereafter, frustules were cleaned as follows: In a first step, 2 mL of sulfuric acid (30%) was added prior to addition of 10 mL saturated potassium permanganate. The solution was incubated for 24 h with gentle agitation after 6 and 20 h, respectively. Subsequently, ~10 mL of saturated oxalic acid solution was titrated until the solution became transparent. Samples were then centrifuged at 2,000 g for 20 min, and the supernatant was discarded. Chemical residuals were removed during two following rinsing steps before samples were kept in distilled water. The frustule cleaning procedure is described in more detail by Lundholm and Moestrup (2002) 59 .

Scanning electron microscopy and determination of valve dimensions

The structure of cleaned valves of C. granii was characterized with scanning electron microscopy. Valves were drop-casted onto a stub and air-dried. Stubs were coated with 15 nm platinum using a Polaron SC7640 sputter coater (Ernst Leitz GmbH, Germany) and examined in a scanning electron microscope (SEM FEI Quanta 200; FEI TM Corporate, USA). Pictures of 5 non-tilted valve samples were recorded and analyzed in the open source software Fiji ImageJ 60 .

Light microscopic determination of diatom valve orientation

The valve surface orientation could be determined by gently touching the valve on the sides with a dissecting needle positioned with a micromanipulator (MM33, Märzhäuser, Wetzlar, Germany), while observing the sample on a light microscope (AxioskopFS, Carl Zeiss, Germany). Valve orientation was defined in this way for >10 different specimens, which were subsequently imaged with a hyper-spectral camera system (VNIR- 100, Themis Vision Systems, St. Louis, USA) mounted on the microscope. The visible higher light

57 transmittance through valves presenting the exterior valve surface towards the microscope trinocular was then used to define the valve orientation in later subsequent experiments.

Waveguiding in cleaned valves in water, and optical coupling to chloroplasts in live cells

A white supercontinuum laser (SuperK COMPACT, NKT-Photonics, Denmark) was connected to a spectral filtration module (SuperK CONNECT, NKT Photonics, Denmark) that channeled filtered laser light via a single mode fiber to a special focusing unit (IS-OGP, Siskiyou Corporation, Grants Pass, USA). The latter unit was mounted between the filter cube slider and the eyepiece trinocular of an epi-fluorescence microscope (Axiostar Plus FL, Carl Zeiss, Germany)(Suppl. Fig. 1). The setup allowed focusing of the filtered laser beam through a 20x objective (NA 0.50) to a ~10 µm spot in the focal plane on live cells or cleaned valves, while imaging the distribution of scattered light or induced fluorescence over the field of view. The spectral filtering unit of the white laser allowed flexible adjustment of center wavelength and bandwidth of actinic light, and in this study we used photosynthetically active wavelengths in the range λ = 500-700 nm for measurements on cleaned valves and λ = 450-650 nm for measurements performed on live diatom cells. A fiber optic scalar irradiance microprobe (spherical tip diameter of 90 µm) 37 was mounted on a micromanipulator, which allowed positioning of the microprobe tip at an angle of 45° on the valve or diatom surface at defined distances from the incident laser spot. Local scalar irradiance at the position of the microprobe tip was recorded with a fiber-optic spectrometer (USB 2000+, Ocean Optics, Dunedin, USA). Waveguiding in cleaned valves and live cells was calculated in percent of incident light in the illuminated laser spot, measured with a microscope slide quantum sensor (Heinz Walz GmbH, Germany). A control experiment was performed where the sensor was placed touching the glass slide without valves or live cells being present, to exclude the effect of waveguiding in the glass slide itself.

Imaging of photosynthesis during localized laser illumination

The microscope setup described above was extended by mounting a variable chlorophyll fluorescence imaging system (RGB microscopy Imaging PAM, Heinz Walz GmbH, Germany) 61 onto the epifluorescence microscope (Suppl. Fig. 1), while using the filtered white laser coupled to the focusing unit as an actinic light source. The spectral filtering unit of the white laser allowed flexible adjustment of center wavelength and bandwidth of the actinic light. In this study we used photosynthetically active wavelengths of λ = 650±10 nm overlapping with the absorption spectrum of Chl a, and Chl c. Diatom samples were observed in well slides at 200x magnification. The actinic laser beam was focused on the edge of 100-200 µm wide living C. granii specimens, while the fluorescence yield was imaged across the field of view by application of weak non-

58 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

actinic pulses of blue measuring light ( λ = 450 nm) before and during a saturating blue light pulse to determine the distribution of variable chlorophyll fluorescence in single chloroplasts throughout the diatom cell. Effective quantum yields of photosystem II and non-photochemical quenching were calculated as described elsewhere 57,62 , and visualized in the open source software Fiji (ImageJ) 60 by using the Ratio Plus plug-in.

Angular effects of incident light on waveguiding in diatom frustule valves

An approximately 100 µm wide cleaned frustule valve was fused onto the 10 µm flat cut tip of a tapered fiber-optic field radiance microprobe (Zenzor, Denmark) 63 as follows. The fiber tip was first dipped into UV glue (Dymax, USA, RI 1.50) before a dried frustule valve was carefully picked under the microscope. The distal end of the fiber was connected to a UV light source (UV glue lamp, Dymax, Germany) and UV curing was applied for 30 min. The position of the valve on the fiber tip was recorded before fixing the microprobe on a custom-made turntable for measuring the angular light collection properties of the valve mounted on the fiber tip. A detailed description of the setup, usually used to determine the isotropic light collection properties of fiber-optic scalar irradiance probes, can be found in Rickelt et al. (2016). For measurements, the distal end of the fiber was connected to a spectrophotometer (USB 2000+, Ocean Optics, Dunedin, USA) controlled by the manufacturers spectral light acquisition software (Spectra Suite, Ocean Optics, Dunedin, USA). Control measurements were performed with the probe in the absence of a valve (Sensor only). To exclude effects of the optical cement, the fiber tip was also dipped in UV-glue in the control measurements. Data were calculated as the relative photon count normalized to the respective integration time.

Single cell in vivo light absorption

Light attenuation throughout a living C granii cell was estimated with a flat-cut field radiance microprobe (10 µm tip diameter; Zenzor, Denmark) placed at a 45° angle over a single diatom cell, while illumination was provided vertically from above (Suppl. Fig. 3). The sensor thereby recorded a proxy for the light loss due to scattering and absorption. The individual diatom cells were attached to a microscope glass slide coated with a black non-reflective tape. White light ( λ = 400-800 nm) was provided from a halogen lamp (KL 1500 HAL, Schott AG, Mainz, Germany) connected to a fiber-optic annular ring light system that enclosed the lens turret of a dissection stereo microscope (Stemi SV 6; Carl Zeiss, GER). Spectral reflectance was recorded with the microprobe connected to a fiber-optic spectrometer (USB 2000+, Ocean Optics, Dunedin, USA). Data were normalized to a 99% diffuse reflectance standard (Labsphere Inc., North Sutton, USA), measured at the same distance to the light source.

59 Orientation dependent light transmittance through the valve wall

Spectral transmission of light through dry and water covered C. granii valves was recorded with a hyper- spectral camera system (VNIR-100, Themis Vision Systems, St Louis, USA) mounted on a light microscope (AxioskopFS, Carl Zeiss, Germany). Samples were analyzed with objectives of different numerical apertures (UPlanFL N 4x/NA=0.13; UPlanFL N 20x/NA=0.50; Carl Zeiss, Germany). Hyperspectral images were recorded on cleaned valves placed on a microscope slide in a drop of water, or left to dry for at least two hours. Hyperspectral image stacks were calculated in % transmittance by normalizing the hyperspectral image stacks of transmitted light through the valves to image stacks of incident light; acquired in an image without valves. All data were corrected for dark noise prior to ratio calculations. Hyperspectral image acquisitions and analysis was done with the PhiLumina Hyperspectral Imaging System software (PhiLumina, University of Missisipi, USA).

Reflectance of diatom valves and in vivo light attenuation

Cleaned valve samples were placed in a drop of water, or left to dry for two hours on a glass slide coated with black tape to reduce background reflection. White light ( λ = 400-800 nm) was provided by a fiber-optic halogen lamp (KL 1500 HAL, Schott AG, Mainz, Germany) connected to a fiber-optic annular ring light system mounted around the objective of a stereo-microscope (Stemi SV 6; Carl Zeiss, Jena, Germany). A tapered flat-cut fiber optic field radiance probe with a tip diameter of 10 µm 63 was carefully positioned at an angle of 45° relative to the surface of the valve using a micromanipulator (MM33, Märtzhäuser, Wetzlar, Germany; Suppl. Fig. 3). Orientation of wet valves was determined by moving the valve with the sensor tip, while observing the curvature of the valve. Once the orientation was determined, valves were left to dry, and measurements were repeated in air. Spectral reflectance was recorded with the microprobe connected to a fiber-optic spectrometer (USB 2000+, Ocean Optics, Dunedin, USA). Measurements in water, and air were normalized to incident light by measuring reflected light from a 99% reflectance standard (Labsphere Inc., North Sutton, USA).

In a similar setup, spectral attenuation of light was measured on live C. granii cells placed on the 99% reflectance standard. Light was measured with the field radiance microprobe pointing on, but not touching the live cell. The probe was positioned at a 45° angle relative to the vertically incident light, a few µm over the diatom cell, and normalized to similar measurements of light reflected by the 99% reflectance standard in the absence the diatom cell.

60 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Dark field microscopy of diatom valves and living cells

Cleaned C. granii valves samples were observed in air, or distilled water, while live specimens where observed in L1 medium. Observations were performed on an optical compound microscope (BX41 Laboratory Microscope, Olympus, USA) under transmitted white light provided by the microscope lamp. Samples were observed at 40x magnification with an 4x objective (UPlanFL N; Carl Zeiss, Jena, Germany; NA = 0.13) using either bright field illumination, a dark field stop, or different dark field phase contrast filters using a phase contrast condenser turret (U-PCD2, Olympus, Tokyo, Japan). Images of the samples were recorded with a charge coupled device (CCD) camera (Color View Soft Imaging System, Olympus, Tokyo, Japan) mounted on the microscope.

Quantification of photopigments

Living diatoms were collected after six days of growth by filtering 20 mL of C. granii culture onto glass microfiber filters (GF/F, Whatman GE Healthcare, Buckinghamshire, UK) followed by immediate freezing in liquid N 2. Subsequent quantitative analysis of the C. granii pigmentation was performed in triplicates by certified High Performance Liquid Chromatography (HPLC) analysis (DHI Lab, Hørsholm, Denmark).

Wavelength dependent functional absorption of photosystem II

The wavelength-dependent functional absorption cross section of photosystem II in C. granii cells was measured with a multicolor pulse-amplitude modulated (PAM) variable chlorophyll fluorometer (MC-PAM, Heinz Walz GmbH, Effeltrich, Germany), using measuring light at wavelengths of λ = 440, 480, 540, 590 and 625 nm (all at non-actinic intensities of <0.5 µmol photons m -2 s-1). The wavelength dependent 64 functional absorption cross-section of PS(II), Sigma(II) λ, was measured with the script-file Sigma100.FTM using the system software PamWin3 (Heinz Walz GmbH, Germany). More details on this methodology can be found in Klughammer et al. (2011) 65 and Schreiber et al. (2012) 66 .

61

62 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Acknowledgements

We thank Gert Hansen for providing culture material from the Scandinavian Culture Collection (www.sccap.dk). We acknowledge the excellent technical assistance of Sofie Jakobsen. Michael Hansen (PLEN), Mads Lichtenberg and Klaus Koren (MARS) are thanked for fruitful discussions that significantly improved this work. This study was supported by a Sapere-Aude Advanced Grant from the Danish Council for Independent Research | Natural Sciences (MK), an instrument grant from the Carlsberg Foundation (MK), and a project grant from the Danish Research Council (project ALPHA 12-127569; grant to ME).

Author contribution

Experimental design (J.W.G., Y.S., L.R., P.C., M.E., M.K.). Culture growths and cleaning of valve material (Y.S., S.W., J.W.G.). Optical and photophysiological measurements (J.W.G.). Provided knowhow and experimental tools (C.M., D.W., E.C.L.T., L.R., M.K.). Analyzed data (J.W.G., X.W., C.M., M.E., M.K., Y.S.). Outlined manuscript (J.W.G., M.E., M.K.). Wrote the manuscript (J.W.G. with editorial assistance by all co-authors).

Competing financial interests

The authors declare no competing financial interests.

63

64 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Literature

1. Srinivasarao, M. Nano-Optics in the biological world: Beetles, butterflies, birds, and moths. Chem. Rev. 99, 1935–1962 (1999). 2. Parker, A. R. & Townley, H. E. Biomimetics of photonic nanostructures. Nat. Nanotechnol. 2, 347–353 (2007). 3. Stobbe, N. & Schaefer, H. M. Enhancement of chromatic contrast increases predation risk for striped butterflies. Proc. R. Soc. B Biol. Sci. 275, 1535–1541 (2008). 4. Vukusic, P. & Sambles, J. R. Photonic structures in biology. Nature 424, 852–5 (2003). 5. Sun, J., Bhushan, B. & Tong, J. Structural coloration in nature. RSC Adv. 3, 14862 (2013). 6. Gould, K. S. & Lee, D. W. Physical and ultrastructural basis of blue leaf iridescence in four Malaysian understory plants. Am. J. Bot. 83, 45–50 (1996). 7. Gkikas, D., Argiropoulos, A. & Rhizopoulou, S. Epidermal focusing of light and modelling of reflectance in floral-petals with conically shaped epidermal cells. Flora Morphol. Distrib. Funct. Ecol. Plants 212, 38–45 (2015). 8. Jacobs, M., Lopez-garcia, M., Phrathep, O.-P. & Lawson, T. Photonic crystal structure of Begonia chloroplasts enhances photosynthetic efficiency. Nat. Plants 1–16 (2016). 9. Holt, A. L., Vahidinia, S., Gagnon, Y. L., Morse, D. E. & Sweeney, A. M. Photosymbiotic giant clams are transformers of solar flux. J. R. Soc. Interface 11, 20140678 (2014). 10. Nelson, D. M. & Brzezinski, M. a. Diatom growth and productivity in an oligo-trophic midocean gyre: A 3-yr record from the Sargasso Sea near Bermuda. Limnol. Oceanogr. 42, 473–486 (1997). 11. Geider, R. J. et al. Primary productivity of planet earth: biological determinants and physical constraints in terrestrial and aquatic habitats. Glob. Chang. Biol. 7, 849–882 (2001). 12. Kirk, J. T. O. Light and Photosynthesis in Aquatic Ecosystems . Cambridge University Press (2010). 13. Schenk, H. J. On the Focusing of Sunlight by Ocean Waves. J. Opt. Soc. Am. 47, 653–657 (1957). 14. Krause-Jensen, D. & Sand-Jensen, K. Light attenuation and photosynthesis of aquatic plant communities. Limnol. Oceanogr. 43, 396–407 (1998). 15. Boyd, S. R. et al. Multiple growth events during diamond genesis: an integrated study of carbon and nitrogen isotopes and nitrogen aggregation state in coated stones. Earth Planet. Sci. Lett. 86, 341–353 (1987). 16. Thomas, D. N. & Dieckmann, G. S. Antarctic Sea ice--a habitat for extremophiles. Science 295, 641–644 (2002). 17. Mock, T. & Valentin, K. Photosynthesis and cold acclimation: Molecular evidence from a polar diatom. J. Phycol. 40, 732–741 (2004). 18. Gilbert, M., Wilhelm, C. & Richter, M. Bio-optical modelling of oxygen evolution using in vivo fluorescence: Comparison of measured and calculated photosynthesis/irradiance (P-I) curves in four representative phytoplankton species. J. Plant Physiol. 157, 307–314 (2000). 19. Apel, K. & Hirt, H. Reactive oxygen species: metabolism, oxidative stress, and signal

65 transduction. Annu. Rev. Plant Biol. 55, 373–399 (2004). 20. Schumann, A., Goss, R., Jakob, T. & Wilhelm, C. Investigation of the quenching efficiency of diatoxanthin in cells of Phaeodactylum tricornutum (Bacillariophyceae) with different pool sizes of xanthophyll cycle pigments. Phycologia 46, 113–117 (2007). 21. Goss, R. & Jakob, T. Regulation and function of xanthophyll cycle-dependent photoprotection in algae. Photosynth. Res. 106, 103–122 (2010). 22. Schellenberger Costa, B. et al. Blue light is essential for high light acclimation and photoprotection in the diatom Phaeodactylum tricornutum . J. Exp. Bot. 64, 483–493 (2013). 23. Furukawa, T., Watanabe, M. & Shihira-Ishikawa, I. Green- and blue-light-mediated chloroplast migration in the centric diatom Pleurosira laevis . Protoplasma 203, 214–220 (1998). 24. Montsant, A., Jabbari, K., Maheswari, U. & Bowler, C. Comparative genomics of the pennate diatom Phaeodactylum tricornutum . Plant Physiol. 137, 500–513 (2005). 25. Bowler, C. et al. The Phaeodactylum genome reveals the evolutionary history of diatom genomes. Nature 456, 239–244 (2008). 26. Lockhart, J. Blue light checkpoint: How blue light controls the onset of cell division in diatoms. Plant Cell 25, 1–1 (2013). 27. Guiry, M. D. How many species of algae are there? J. Phycol. 1063, 1057–1063 (2012). 28. Su, Y., Lundholm, N., Friis, S. M. M. & Ellegaard, M. Implications for photonic applications of diatom growth and frustule nanostructure changes in response to different light wavelengths. Nano Res. 8, 2363–2372 (2015). 29. Hamm, C. E. et al. Architecture and material properties of diatom shells provide effective mechanical protection. Nature 421, 841–843 (2003). 30. Hale, M. S. & Mitchell, J. G. Functional morphology of diatom frustule microstructures: Hydrodynamic control of brownian particle diffusion and advection. Aquat. Microb. Ecol. 24, 287–295 (2001). 31. Romann, J. et al. Wavelength and orientation dependent capture of light by diatom frustule nanostructures. Sci. Rep. 5, 17403 (2015). 32. Fuhrmann, T., Landwehr, S., El Rharbi-Kucki, M. & Sumper, M. Diatoms as living photonic crystals. Appl. Phys. B Lasers Opt. 78, 257–260 (2004). 33. Yamanaka, S. et al. Optical properties of diatom silica frustule with special reference to blue light. J. Appl. Phys. 103, 1–6 (2008). 34. Chen, X., Wang, C., Baker, E. & Sun, C. Numerical and experimental investigation of light trapping effect of nanostructured diatom frustules. Sci. Rep. 5, 11977 (2015). 35. Valmalette, J., Romann, J., Røyset, A. & Einarsrud, M. 3D-Hyperspectral mapping of light propagation through diatom frustule silica nanostructures. Opt. Lett. 40, 3–5 (2015). 36. Ellegaard, M. et al. The fascinating diatom frustule – can it play a role for attenuation of UV radiation? J. Appl. Phycol. 28, 3295–3306 (2016). 37. Rickelt, L. F., Lichtenberg, M., Trampe, E. C. L. & Kühl, M. Fiber-optic probes for small-scale measurements of scalar irradiance. Photochem. Photobiol. 92, 331–342 (2016). 38. Rohá ček, K. et al. Relaxation of the non-photochemical chlorophyll fluorescence quenching in diatoms: kinetics, components and mechanisms. Philos. Trans. R. Soc. Lond. B. Biol. Sci.

66 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

369, 20130241 (2014). 39. Popov, E., Nevière, M., Enoch, S. & Reinisch, R. Theory of light transmission through subwavelength periodic hole arrays. Phys. Rev. B 62, 16100–16108 (2000). 40. Yang, W., Lopez, P. J. & Rosengarten, G. Diatoms: self assembled silica nanostructures, and templates for bio/chemical sensors and biomimetic membranes. Analyst 136, 42–53 (2011). 41. De Stefano, L., Rea, I., Rendina, I., De Stefano, M. & Moretti, L. Lensless light focusing with the centric marine diatom Coscinodiscus walesii. Opt. Express 15, 18082–18088 (2007). 42. Roesler, C. S., Perry, M. J. & Carder, K. L. Modeling in situ phytoplankton absorption from total absorption spectra in productive inland marine waters. Limnol. Oceanogr. 34, 1510– 1523 (1989). 43. Astoreca, R. et al. Development and application of an algorithm for detecting Phaeocystis globosa blooms in the Case 2 Southern North Sea waters. J. Res. 31, 287–300 (2009). 44. Noernberg, M. A., Mizerkowski, B. D., Paloschi, N. G. & Bento, J. P. Hydrodynamics and bio-optical assessment of two pristine subtropical estuaries in southern brazil. Brazilian J. Oceanogr. 62, 265–278 (2014). 45. Lohr, M. & Wilhelm, C. Algae displaying the diadinoxanthin cycle also possess the violaxanthin cycle. Proc. Natl. Acad. Sci. U. S. A. 96, 8784–8789 (1999). 46. Millie, D. F., Schofield, O., Kirkpatrick, G. J., Johnsen, G. & Evens, T. J. Using absorbance and fluorescence spectra to discriminate microalgae. Eur. J. Phycol. 37, 313–322 (2002). 47. Kuczynska, P., Jemiola-Rzeminska, M. & Strzalka, K. Photosynthetic pigments in diatoms. Mar. Drugs 13, 5847–5881 (2015). 48. Ruban, A. et al. The super-excess energy dissipation in diatom algae: Comparative analysis with higher plants. Photosynth. Res. 82, 165–175 (2004). 49. Dougher, T. a & Bugbee, B. Differences in the response of wheat, soybean and lettuce to reduced blue radiation. Photochem. Photobiol. 73, 199–207 (2001). 50. Goessling, J. W., Cartaxana, P. & Kühl, M. Photo-protection in the centric diatom Coscinodiscus granii is not controlled by chloroplast high-light avoidance movement. Front. Mar. Sci. 2, 1–10 (2016). 51. Sanchez Saavedra, M. P. & Voltolina, D. The effect of different light quality on the food value of the diatom sp. for Artemia franciscana . Riv. Ital. Acquacolt. 30, 135– 138 (1995). 52. Mercado, J. M. et al. Blue light effect on growth, light absorption characteristics and photosynthesis of five benthic diatom strains. Aquat. Bot. 78, 265–277 (2004). 53. Sundar, V. C., Yablon, A. D., Grazul, J. L., Ilan, M. & Aizenberg, J. Fibre-optical features of a glass sponge. Nature 424, 899–900 (2003). 54. Wangpraseurt, D. et al. Lateral light transfer ensures efficient resource distribution in symbiont-bearing corals. J. Exp. Biol. 217, 489–498 (2014). 55. Ooms, M. D., Bajin, L. & Sinton, D. Culturing photosynthetic bacteria through surface plasmon resonance. Appl. Phys. Lett. 101, (2012). 56. Kitajima, M. & Butler, W. L. Quenching of chlorophyll fluorescence and primary photochemistry in chloroplasts by dibromothymoquinone. Biochim. Biophys. Acta -

67 Bioenerg. 376, 105–115 (1975). 57. Maxwell, K. & Johnson, G. N. Chlorophyll fluorescence--a practical guide. J. Exp. Bot. 51, 659–68 (2000). 58. Hill, S. P., Banerjee, T., Dilbeck, T. & Hanson, K. Photon Upconversion and Photocurrent Generation via Self-Assembly at Organic-Inorganic Interfaces. J. Phys. Chem. Lett. 6, 4510–4517 (2015). 59. Lundholm, N. & Moestrup, Ø. The marine diatom Pseudo-nitzschia galaxiae sp. nov. (Bacillariophyceae): morphology and phylogenetic relationships. Phycologia 41, 594–605 (2002). 60. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012). 61. Trampe, E., Kolbowski, J., Schreiber, U. & Kühl, M. Rapid assessment of different oxygenic phototrophs and single-cell photosynthesis with multicolour variable chlorophyll fluorescence imaging. Mar. Biol. 158, 1667–1675 (2011). 62. Genty, B., Briantais, J.-M. & Baker, N. R. The relationship between the quantum yield of photosynthetic electron transport and quenching of chlorophyll fluorescence. Biochim. Biophys. Acta - Gen. Subj. 990, 87–92 (1989). 63. Kühl, M. Optical microsensors for analysis of microbial communities. Methods Enzymol. 397, 166–199 (2005). 64. Schreiber, U., Klughammer, C. & Kolbowski, J. Assessment of wavelength-dependent parameters of photosynthetic electron transport with a new type of multi-color PAM chlorophyll fluorometer. Photosynth. Res. 113, 127–44 (2012). 65. Klughammer, C. & Schreiber, U. Saturation Pulse method for assessment of energy conversion in PS I. PAM Appl. Notes 11–14 (2008). 66. Schreiber, U., Klughammer, C. & Kolbowski, J. High-end chlorophyll fluorescence analysis with the MULTI-COLOR-PAM . I . Various light qualities and their applications . PAM Appl. Notes 1–21 (2011).

68 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Figures

Figure 1. Spatial organization of the Coscinodiscus granii cell and frustule.

(a) Color micrograph of a single C. granii cell in a slightly tilted view. The diatom chloroplasts show a brown photo-pigmentation, and the cell is encased in a silicate frustule. (b) Schematic illustration of the frustule with two valves and corresponding girdle bands. (c) SEM images and (d) schematic illustration of the C. granii valve with small cribrum pores on the exterior valve surface (EVS; green), honey comb-shaped structures forming the chambered valve wall (CVW; red), and large foramina on the interior valve surface (IVS; yellow). Scale bars: 100 µm (a); 1,000 µm (c); 500 nm (d).

69

70 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Figure 2. Photonic waveguiding in the cleaned valve wall and optical coupling of chloroplasts.

(a) Optical wide field image of a living diatom showing the setup in which photonic waveguiding was determined. A spectrally filtered laser beam was focused onto a ~10 µm wide spot on the periphery of the frustule, while measuring light on the opposite side of the frustule with a scalar irradiance microprobe (90 µm tip diameter). (b) Optical wide field image showing the transmission through different parts of the valve, while the valve orientation was deduced by their different dark-bright contrast in the light microscope. Arrows indicate the view on girdle bands (Gb), the interior valve surface (IVS) or the exterior valve surface (EVS), when light was first transmitted through the opposite valve side. (c) Waveguiding was observed independently from the valve orientation when a focused laser beam ( λ = 500-700 nm) illuminated the valve distant to the scalar irradiance microprobe. In a control experiment, the scalar irradiance probe touched a glass slide illuminated at a similar distance in the absence of valves. Data are normalized to incident light in the laser beam spot. (d) Fluorescence microscope image of chlorophyll fluorescence distribution induced by weak blue measuring light ( λ = 450 nm) upon local illumination of a ~10 µm spot of a live specimen of C. granii . Chloroplasts were located in close proximity to the valve (blue arrow), resulting in transmittance of chlorophyll fluorescence into the valve wall (white arrow). (e) Waveguiding in a living C. granii cell measured with a scalar irradiance microprobe distant to the local laser illumination (live cell). Incident light of the local laser illumination is indicated (localized laser). Inset: Chlorophyll fluorescence (maximum at a λ = 685 nm) in a live C. granii cell when stimulated by a focused laser beam (filtered at λ = 650 nm) and

71 transmitted into the valve wall as measured with a scalar irradiance microprobe. All measurements were done in water. Scale bars: 100 µm (a); 200 µm (b), 100 µm (c).

72 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Figure 3. Stimulation of diatom photosynthesis by waveguiding in the valve and optical coupling to chloroplasts.

(a) Living C. granii cell illuminated locally with a ~10 µm wide actinic laser spot ( λ = 650±10 nm; overlapping with Chl a and Chl c absorption), under simultaneous imaging of variable chlorophyll fluorescence over the whole diatom with weak non-actinic blue light pulses (8 Hz). When the laser was switched on, chlorophyll fluorescence was quenched in regions distant to the local illumination (position 1- 3), and chlorophyll fluorescence recovered to initial values when the actinic point illumination was switched off again. (b) Chlorophyll fluorescence traces and application of saturating light pulses (asterix) showed typical chlorophyll fluorescence induction kinetic when actinic point illumination was switched on. (c) Effective quantum yields of photosystem II decreased with increasing photon irradiance of the laser spot illumination. Error bars represent standard deviation of measurements on three individual C. granii cells. False color code scale bar in (a) represents 100 µm.

73

74 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Fig. 4. Light-collecting isotropy of frustule valves as a function of angle of incidence.

(a) A single cleaned frustule valve was glued with optical cement onto the 10 µm tip of a tapered field radiance probe. b) Light-collecting isotropy was tested on a turntable that allowed changing the angle of incident light from a collimated white light source relatively to the frustule valve glued onto the field radiance probe. Measurements of the light acceptance angle were performed on the same field radiance probe in the absence (Sensor only) or presence of a frustule valve (Valve on sensor). Measurements were conducted c) in air, and d) in water. Light intensity is shown in arbitrary units. Orientation of the frustule valve in relation to the angle of incident light is indicated in the respective white circle. Scale bar: 100 µm (a).

75

76 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Figure: 5. Influence of orientation and refractive index on the optical properties of diatom valves.

Hyperspectral imaging of (a; b) transmitted, and (c; d) reflected light incident on the exterior (EVS) or the interior (IVS) valve surface. Measurements were performed in (a; c) air and (b; d) water. Measurements were performed through an objective with (a; b) low numerical aperture (NA 0.13) and with (e) high numerical aperture (0.50). (f) Illustration of light collection with objectives of different numerical aperture. Constructive interference at the wavefront is dependent on the numerical aperture.

77

78 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Figure 6. Effects of angle of incidence and refractive index on scattering properties of cleaned valves and living cells.

Images of cleaned diatom valves obtained at 50x magnification with a low numerical aperture objective (NA 0.13). The angle of incident white light ( θ) was varied by inserting patch stops of different ring sizes between the microscope lamp and condenser, creating different angles of incident light. (a) Measurements were done on cleaned valves in air or water. (b) Illustration of the valve interaction with white light (yellow arrow) at an incidence angle of θ=15°, showing potential interaction and focusing of blue radiation. (c) Scattering properties of live cells measured in L 1 growth medium. Scale bars represent 200 µm.

79

80 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Figure 7. Light attenuation, photopigments and PSII absorption in live diatoms.

(a) Spectral light attenuation of single C. granii cells (average values for 5 independent specimens). (b) C. granii cells in apical and distal illustration. (c) Dominant photopigments in C. granii as quantified by HPLC analysis. (d) Wavelength dependent functional absorption cross-section of PSII, Sigma(II) λ, indicating preferences of the photosystem II antennae for shorter wavelengths (Average ± SD).

81

82 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Supplemental figures

Suppl. Figure 1. Setup for laser induced single cell variable chlorophyll fluorimetry.

An epi-fluorescence microscope (Axiostar Plus FL, Carl Zeiss, Jena, Germany) was combined with a chlorophyll fluorescence imaging system (RGB Microscopy Imaging PAM, Heinz Walz GmbH, Effeltrich, Germany) and an optical module ((IS-OGP, Siskiyou Corporation, Grants Pass, USA) enabling projection of a focused laser spot into the field of view. (1) A white super-continuum laser (SuperK COMPACT, NKT- Photonics, Denmark) was coupled to a spectral filtration module (SuperK CONNECT, NKT Photonics, Denmark) that channeled filtered laser light via a single mode fiber to optical module, which was mounted between the filter cube slider and the eyepiece trinocular of the microscope. The spectrally filtered laser light was projected into the light path of the microscope via a dichroic mirror (2) which allowed focusing of the laser beam through an 20x objective to a ~10 µm wide spot on the sample in the focal plane (3). Samples were observed at 200x magnification. The actinic laser light was focused on the edge of 100-200 µm wide living C. granii specimens, while weak non-actinic red measuring light (hatched blue line) and a saturating

83 red light pulses illuminated the sample via the microscopes epifluorescence port (4). This enabled variable chlorophyll fluorescence measurements (dark red line) and determination of effective PSII quantum yields 64 of photosystem II in single chloroplasts throughout the diatom cell during local illumination with the actinic red laser beam (red line). (5) Variable chlorophyll fluorescence and the position of the laser beam were recorded with the Imaging PAM CCD camera mounted on the microscope and processed in the software ImagingWin (Heinz Walz GmbH, Effeltrich, Germany).

84 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Suppl. Figure 2. Spatial distribution of photosynthesis in single diatom as a function of laser spot irradiance.

Chlorophyll fluorescence decreased over the entire diatom cell when the irradiance in the focused laser beam was increased (a). The effective PSII quantum yield, Y(II), decreased (b) and non-photochemical quenching, NPQ, increased (c) throughout the diatom cell with increasing irradiance in the laser beam spot. (c). Scale bar represents 200 µm.

85

86 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Suppl. Figure 3: Valve reflectance and in vivo light attenuation measurements on diatoms.

A tapered flat-cut fiber optic field radiance microprobe 65 with a tip diameter of 10 µm was attached on a micro-manipulator at a 45°C angle relative to the incident light. Illumination was provided vertically from above with a halogen lamp equipped with a fiber-optic ring light. The spectator controlled the position of the light sensor via microscopic observation. Light spectra were recorded with the field radiance microprobe tip positioned on the diatom/valve, while the distal end of the fiber probe was coupled to a fiber-optic spectrometer. Measurements on diatoms/valves were normalized to similar measurements on a white 99% reflectance standard.

87

88 6.1. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms

Suppl. Figure 4. Light scattering of glass beads in water.

Glass beads with a diameter of 50-200 µm were imaged with a low numerical aperture objective (NA 0.13; 5x). The angle of incident white light was varied with patch stops of different ring sizes applied between the microscope lamp and condenser. Measurements were performed in water. Scale bar indicates 500 µm.

89

90

6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules (in prep.)

Johannes W. Goessling 1* , Yanyan Su 2, Christian Maibohm 3, Marianne Ellegaard 2 and Michael Kühl 1,4

1Marine Biology Section, Department of Biology, University of Copenhagen, Strandpromenaden 5, 3000 Helsingør, Denmark

2Section for Plant Glycobiology, Department for Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg, Denmark

3International Iberian Nanotechnology Laboratory, Avenida Mestre José Veiga s/n, 4715-330 Braga, Portugal

4Climate Change Cluster, University of Technology Sydney, Australia

*Corresponding author: Johannes Goessling; [email protected]; Strandpromenaden 5, 3000 Helsingør, Denmark

91

92 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Abstract

Diatoms are phototrophic single celled microalgae encased in a frustule made of amorphous silicate. The frustule comprises two valves connected by a variable number of girdle bands, all exhibiting a periodic micro/nano-porous silicate structure. Using the large centric diatom Coscinodiscus granii in water, we studied the optical properties of girdle bands, a part of the frustule that has received less attention by the scientific community. The valves of this species have hexagonal lattice structures, while the pores in the girdle band are organized in a square lattice. We show that valves and girdle bands exhibit different, partially opposite, optical properties, as valves attenuate shorter wavelengths and girdle bands attenuate longer wavelengths of the visible light spectrum. Girdle bands also exhibited iridescence under certain incident light directions. While implications of the photonic properties upon photosynthesis have been proposed in earlier studies, the finding of differences in valve and girdle band optics indicates that the frustule of live diatoms can exhibit a variety of optical functions.

93

94 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Introduction

Diatoms are single-celled microalgae that perform approximately 20% of global carbon fixation via their photosynthesis, and they are responsible for up to 40% of the primary production in aquatic environments (Nelson & Brzezinski 1997, Geider et al. 2001). The diatom cell is encased in a nanoporous frustule made of almost pure, translucent silicate, and the shape and intricate ornamental patterns of the frustule are the main attributes used in the identification of diatoms. The term diatom ( διάτοµα , diátoma) is a blend of the Greek words for two ( δύο , dío) and indivisible pieces ( άτοµα , átoma), referring to its two more or less identical parts named the thecae. The term diatom may be misleading when taken literally, because the frustule comprises more than two indivisible parts, i.e., each theca is typically formed by one valve and one or more girdle bands (Cox 2011). The two valves are of slightly different diameter with the smaller hypovalve fitting into the larger epivalve like a Petri dish (Fig. 1). During cell division, two new valves develop inside the parental valves, while at least one girdle band (sg. copula; pl. copulae) is formed in concert with each daughter valve, and more girdle bands can be formed over time depending on the species. The sum of girdle bands associated with the valve is referred to as the girdle (sg. cingulum; pl. cingula); the epicingulum connected to the epivalve (together the epitheca) and the hypocingulum connected to the hypovalve (together the hypotheca). Taken together, the two girdles of the frustule can also be referred to as the girding (cinctura)(Round et al. 1990). The girding encircles the two valves at the overlapping region and keeps the frustule together. When the girding detaches, the frustule disassembles and the diatom dies and vice versa once the cell dies the girding detaches and the two valves drift apart.

While valves have been studied intensively in many diatom species, less attention has been given to the porous structures or to the structure-property relationship in the girdle (Cox 2011). In their monograph, Round et al. (1990) noted that illustrations of the diatom frustule often omit the girding despite its importance for diatom cell integrity. In some species, the girdle band structure varies from solid or chambered to porous, while most species have girdle bands with perforated strips of silica or channel-like structures (Cox 2011 and references therein). Pores in the valve wall are essential, allowing chemical communication and nutrient exchange between cell and environment (Hale & Mitchell 2001). It has also been suggested that the high mechanical strength of the valve protects the diatom cell from predation (Hamm et al. 2003, Aitken et al. 2016). Hamm et al. (2003) thus found that diatom valves resisted astonishing high levels of mechanical stress reaching 1-7 N mm -2 (equivalent to 100-700 tons m -2), while the high modulus of elasticity in the girdle bands caused resistance to the maximum mechanical stress that could be applied in their experimental setup, i.e., 560 N mm -2 (equivalent to a pressure 56,000 tons m -2). These exceptional mechanical properties were linked to the periodic nano- and micro-scale porous architecture of the valve and girdle.

95 The porous silicate matrix of the diatom frustule also exhibits photonic crystal-like properties that interact with light in the UV-VIS spectrum (Fuhrmann et al. 2004, Ellegaard et al. 2016) potentially modulating diatom photosynthesis via light focusing, waveguiding or by spectral filtration of photosynthetically productive radiation (Yamanaka et al. 2008, Toster et al. 2013, Romann et al. 2015, Tommasi 2016) (Goessling et al. 2017; in prep.). While the biophotonics of diatom valves have been studied intensively (e.g. Valmalette et al. ; De Stefano et al. 2007; De Tommasi et al. 2010; Maibohm et al. 2015a; Maibohm et al. 2015b; Romann et al. 2015), knowledge about the optical properties of the girdle band is very limited and mainly based on few anecdotal observations. Fuhrmann et al. (2004) observed that light emitted from the edges of the girdle band in the centric diatom Coscinodiscus granii appeared green when observed with a small numerical aperture (NA 0.40) objective, and blue-green when observed with a larger numerical aperture (NA 0.55) objective. In a review about the biomimetics of photonic nanostructures, Parker and Townley (2007) showed light microscopy pictures of apparently iridescent girdle bands in a centric diatom species, as an example of structural coloration in nature. The present study is the first detailed investigation of the girdle band nanostructure and its optical properties in the large (~50-200 µm) centric diatom species Coscinodiscus granii . We show that the girdle band exhibits optical properties that differ from those of the valves of C. granii , and we discuss implications of these optical properties for diatom photobiology.

96 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Material and Methods

Sample preparation

The centric diatom species C. granii (strain no. K-1843) was grown at 20°C in 50 mL culture flasks for 6 consecutive days under white light LEDs (opto semiconductors, Dragon1 PowerStar , OSRAM, UK) at a photon irradiance (400-700 nm) of 300 µmol photons m -2 s-1 in a 16/8 hours day/night cycle. After harvesting cells, the organic material was removed by oxidation following the procedure described in Lundholm and Moestrup (2002). Samples were stored in distilled water at 4°C prior to further analysis.

Structure analysis with TEM and SEM

Oxidized samples containing diatom valves and girdle bands were dropcasted on copper grids and examined with a transmission electron microscope (TEM; JEM-1010, Jeol, Japan). Other samples were mounted on a stub, before coating with 15 nm platinum using a sputter coater (Polaron SC7640; Ernst Leitz GmbH, Germany) and subsequent observation in a scanning electron microscope (SEM; FEI Quanta 200; FEITM Corporate, USA). Dimensions of the valve and girdle band nanostructure were determined on calibrated TEM/SEM pictures in the open source software Fiji (ImageJ 1.46r, Wayne Rasband, National Institutes of Health, USA)(Schindelin et al. 2012).

Analysis of light transmittance

Oxidized valves and girdle bands were placed on a microscope slide in a drop of distilled water and observed without a cover slip at 80x magnification (UPlanFL N 4x/NA=0.13; Carl Zeiss GmbH, Germany) under a light microscope (Axioskop FS, Carl Zeiss GmbH, Germany). Spectral image stacks of light transmitted through the samples were recorded with a hyperspectral camera system (VNIR-100, Themis Vision Systems, St Louis, USA) mounted on the C-mount of the microscope. Hyperspectral images were furthermore recorded with brightfield illumination through a drop of water without sample (100% light transmission), and in the dark (0% transmission). Hyperspectral image stacks of diatom frustules and girdle bands were corrected for the dark signal and normalized to the reference image stack to calculate the percent of light transmitted through the sample. Calibration and analysis of hyperspectral images was done with the manufacturers’ software (Hypervisual; PhiLumina, University of Missisipi, USA).

97 Analysis of light reflectance

Oxidized valves and girdle bands were placed in a black box (l=5 cm; h=2 cm; w=4 cm) submerged in distilled water. A fiber-optic halogen ring light (KL 1500 HAL, Schott AG, Mainz, Germany) was placed around the ocular of a dissection microscope (Stemi SV 6; Carl Zeiss GmbH, Jena, Germany), which focused white light (400-800 nm) onto the sample and allowed observation at 50x magnification. Reflectance was measured with a tapered, flat-cut fiber-optic field radiance microprobe (~10 µm tip diameter) as described in Kühl (2005). The probe was mounted in a micromanipulator (MM33, Märtzhäuser, Wetzlar, Germany) and was positioned towards the structure of interest at a 45° angle relative to the horizontal sample. Reflectance of valves was measured when the sensor tip was placed in the center of either the exterior or interior valve side. Reflectance of girdle bands was measured by touching the single girdle band and lifting it up. That was necessary to ensure that the sensor tip of ~10 µm in diameter was placed over the girdle band of approximately the same size. Reflected light spectra were recorded with the microprobe connected to a fiber- optic spectrometer (USB 2000+, Ocean Optics, Dunedin, USA), and data were normalized to measurements of reflected light from a spectrally neutral 99% white light reflectance standard (Labsphere Inc., North Sutton, USA) at the same relative distance to the light source as the frustule/girdle band samples.

Microscopic imaging

Oxidized valves and girdle bands were observed under an optical compound microscope (BX41 Laboratory Microscope, Olympus, USA) at 40x or 200x magnification (UPlanFL N 4x/NA=0.13; UPlanFL N 20x/NA=0.50; Carl Zeiss, Jena, GER). Samples were illuminated with white light provided by the microscope halogen lamp with the condenser in bright field mode or with a dark-field phase contrast filter (Ph2; 15° angle of incidence) using a phase contrast condenser turret (U-PCD2, Olympus, Tokio, Japan). Microscopic RGB images were recorded with a charge coupled device camera (Color View Soft Imaging System, Olympus, Tokyo, Japan) connected to a personal computer.

98 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Results

Optical effects of C. granii valves and girdle bands

In dark field microscopy, direct illumination is blocked and incident white light enters the specimen at an oblique angle of incidence. Thereby, light scattering structures that redirect light towards the observer become visible. Valves of C. granii scattered blue wavelengths when white light was applied at 15° angle of incidence (Fig. 2A). In girdle bands, light scattering of alternating blue and red radiation was observed under the same experimental conditions (Fig. 2B). In defocus, rainbow colors became visible over the girdle (Fig. 2C). In an intact live diatom cell, enhanced scattering of blue light was observed at the interior valve side (Fig. 2D). In a sample of cultured live diatoms, a specimen with aggregated chloroplasts showed scattering of blue light inside the frustule, when the girding was detached from the cell (Fig. 2E and F).

Structure and porosity of valves and girdle bands

The centric valves of C. granii are disc-shaped with pores and chambers arranged in a hexagonal pattern. The exterior surface is perforated with smaller pores (~10 nm), while a hexagonal chamber is located in the center that opens into a large pore on the interior side (~550 nm). More detailed information on the valve nanostructure in C. granii is presented elsewhere (Fuhrmann et al. 2004, Su et al. 2017, Goessling et al. 2017; in prep.).

Girdle bands of C. granii appeared as split rings, whose overlapping ends were slightly open when detached from the frustule (Fig. 3A). The overall diameter of the girdle band was dependent on the size of the corresponding valve, and varied from 50 to 200 µm. Large girdles reached a thickness of ~1-2 µm (Fig. 3B). The girdle band nanostructure was characterised by small pores with a diameter of >10 nm, arranged in a square lattice pattern (Fig. 3C, D). These pores may be the advalvar opening of channel-like structure perforating the girdle band.

Light transmittance through diatom valves and girdle bands

Transmittance spectra of valves (Fig. 4A) and a girdle band (Fig. 4B) showed different spectral characteristics, where girdle band transmittance decreased from 500-800 nm, while transmittance in this spectral region increased in the C. granii valves. However, the strength of this difference was dependent on the tested region of interest (ROI). We identified four different zones on the valve surface exhibiting different optical properties: i) a peripheral, ii) a medial, iii) a central core, and iv) a non-porous central region

99 (Fig. 4C). Transmittance through the valve was lowest in the medial core and in the non-porous center and highest in the peripheral core. We found a similar dependence on ROI position in the girdle band measurements (Fig. 4B, D). When measurements were performed on the side of a girdle band standing upright, the spectral modification was less pronounced (Fig. 4D).

Light reflectance of diatom valves and girdle bands

Approximately 2.5% of incident white light was reflected, when the exterior valve surface was oriented towards the incident light, while reflectance reached 4.5% with the interior valve surface oriented towards the incident light (Fig. 5A). The maximal reflectance of the girdle bands was higher, and their reflectance spectra exhibited a stronger variation as compared to measurements on valves. Girdle bands reflected between ~5-12% of the incident light with maxima at different wavelengths throughout the measured spectrum in the visible range of light (Fig. 5B). These differences could be explained by changes in the positioning of the girdle band (which was attached to the sensor tip) relative to the incident light source. The effect could also be recorded in light microscopy, where girdle bands showed different structural coloration depending on their relative positions towards the light source (Fig. 5C). These prism effects were even visible by the naked eye, when the spectator changed the position relatively to a sample of concentrated frustules placed in a drop of water (Fig. 5D).

100 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Discussion

Our data show that the optical properties of valves and girdle bands in the diatom C. granii are different and exhibit diverging and partially opposite optical effects in terms of spectral transmittance, reflectance and apparent iridescence (Fig. 2). This may be closely linked to the structure of these two different parts of the diatom frustule, where the pores in the C. granii valve wall are arranged in a hexagonal lattice pattern, while the girdle bands have a square lattice porous structure (Round et al. 1990; Fuhrmann et al. 2004; Fig. 3).

Both valves and girdle bands of C. granii can be regarded as photonic crystal-like structures acting as slab waveguides and exhibiting differences in their number of light propagation modes in the visible part of the spectrum (Fuhrmann et al. 2004). Hence, different light transmittance properties may be expected when incident white light interacts with the porous lattice of either valves or girdle bands. Transmittance spectra showed opposing trends, i.e., shorter wavelengths attenuated more when incident on the inside of the valve, while longer wavelengths attenuated more when incident on the inside of the girdle band (Fig. 4). These observations suggest that light of longer wavelengths was either reflected on the side of incident light, or guided more efficiently inside the girdle band, while shorter wavelengths were more likely linear transmitted through its thin network of pores. In contrast, Fuhrmann et al. (2004) argued that the much smaller pore network of the girdle band as compared to the frustule should prohibit coupling of red light into the waveguide of the girdle band. Indeed, when we measured transmittance at different regions of an upright standing girdle band, attenuation was more moderate when the girdle band’s surface was not aligned parallel to the objective. Similar effects were observed on the surface of a valve and were linked to the surface curvature, which might slightly change the angle of redirected light towards the objective. Changes in spectral light transmission were more distinct across the central parts of both the girdle band and the valve, where the surface was more or less aligned in parallel with the objective. In these regions, light may have been diffracted on nanoporous structures of the lattice in the valve and girdle band.

Iridescence is a widespread optical phenomenon in many different organisms. Classical examples are the shiny scales in many insects (Vukusic et al. 1999), hair-like structures found in some polychaete worms (McPhedran et al. 2001), or the colorful plumage of bird species such as the peacock (Dakin & Montgomerie 2013). Although iridescence is much more commonly encountered in terrestrial than in aquatic environments (Vukusic & Sambles 2003), it has also been observed e.g. in fish species (Gur et al. 2013) and in mollusks (Liu et al. 1999). In the mantle of giant clams, iridocytes act as Bragg reflectors that backscatter harmful radiation, while redistributing photosynthetically active radiation efficiently to deeper tissue layers containing symbiotic microalgae (Holt et al. 2014). Iridescence is induced when incident light is partially reflected by at least two stacked surfaces with refractive index contrast. In diatoms, the lower refractive index of water (relative to silicate) on the exterior could cause reflection on the advalvar side of the girdle

101 band, while light entering the girdle band could be reflected on the abvalvar internal surface. We could show that the girdle bands of C. granii produced iridescent colors dependent on their orientation towards the light source - in contrast to diatom valves, which reflected <5% throughout the spectrum of incident white light (Fig. 3).

The various optical effects observed in the valve and girdle band of C. granii may be linked to different biological challenges in the marine environment. For instance, reduced light reflection on the surface of diatoms valves, covering the by far larger surface area of the organism, would enhance light exposure, while the much smaller area covered by the girding might be of less importance for photosynthetic light harvesting. On the other hand, the partially opposing spectral effects of girdle bands as compared to the valve could also complement the valve´s optical properties to further affect light propagation and absorption in the living cell. Implications of the photonic properties of the frustule valve have been proposed earlier (Fuhrmann et al. 2004, Yamanaka et al. 2008, Romann et al. 2015, Ellegaard et al. 2016, Tommasi 2016) and could be linked to efficient light harvesting and photosynthesis by waveguiding in the frustule valve wall and spectral selectivity for more energetic blue light (Goessling et al. 2017; in prep). The optical properties from girdle bands have not been discussed in the light of a biological function to date. We speculate that the nanostructural differences of valves and girdle bands may have different optical functions in the live diatom cell. We conclude, that the diatom frustule is composed of different optical elements with complex biophotonic properties, the biological function of which still remains largely unknown.

102 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Literature

AITKEN Z. H., LUO S., REYNOLDS S. N., THAULOW C. & GREER J. R. 2016. Microstructure provides insights into evolutionary design and resilience of Coscinodiscus sp. frustule. Proceedings of the National Academy of Sciences of the United States of America 113: 2017–2022. COX E. J. 2011. Morphology, cell wall, cytology, ultrastructure and morphogenetic studies: Overview and specific observations. The Diatom World: 21–46. DAKIN R. & MONTGOMERIE R. 2013. Eye for an eyespot: How iridescent plumage ocelli influence peacock mating success. Behavioral Ecology 24: 1048–1057. ELLEGAARD M., LENAU T., LUNDHOLM N., MAIBOHM C., MICHAEL S., FRIIS M., ROTTWITT K. & SU Y. 2016. The fascinating diatom frustule – can it play a role for attenuation of UV radiation?. Journal of Applied Phycology 28: 3295–3306. Journal of Applied Phycology. FUHRMANN T., LANDWEHR S., EL RHARBI-KUCKI M. & SUMPER M. 2004. Diatoms as living photonic crystals. Applied Physics B: Lasers and Optics 78: 257–260. GEIDER R. J., DELUCIA E. H., FALKOWSKI P. G., FINZI A C., GRIME J. P., GRACE J., KANA T. M., LA ROCHE J., LONG S. P., OSBORNE B. A, PLATT T., PRENTICE I. C., RAVEN J. A, SCHLESINGER W. H., SMETACEK V., STUART V., SATHYENDRANATH S., THOMAS R. B., VOGELMANN T. C., WILLIAMS P. & WOODWARD F. I. 2001. Primary productivity of planet earth: biological determinants and physical constraints in terrestrial and aquatic habitats. Global Change Biology 7: 849–882. GOESSLING J.W., SU Y., CARTAXANA P., MAIBOHM C., RICKELT L.F. TRAMPE E.C.L.T, WALBY S., WANGPRASSEURT D., WU X., ELLEGAARD M. & KÜHL M. 2017. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms. In preparation. GUR D., POLITI Y., SIVAN B., FRATZL P., WEINER S. & ADDADI L. 2013. Guanine-based photonic crystals in fish scales form from an amorphous precursor. Angewandte Chemie - International Edition 52: 388–391. HALE M. S. & MITCHELL J. G. 2001. Functional morphology of diatom frustule microstructures: Hydrodynamic control of brownian particle diffusion and advection. Aquatic Microbial Ecology 24: 287–295. HAMM C. E., MERKEL R., SPRINGER O., JURKOJC P., MAIER C., PRECHTEL K. & SMETACEK V. 2003. Architecture and material properties of diatom shells provide effective mechanical protection.. Nature 421: 841–843. HOLT A. L., VAHIDINIA S., GAGNON Y. L., MORSE D. E. & SWEENEY A. M. 2014. Photosymbiotic giant clams are transformers of solar flux. Journal of The Royal Society Interface 11: 20140678. KÜHL M. 2005. Optical microsensors for analysis of microbial communities. Methods in Enzymology 397: 166–199. LIU Y., SHIGLEY J. & HURWIT K. 1999. Iridescent Color of a Shell of the Mollusk Pinctada Margaritifera Caused by Diffraction. Optics Express 4: 177. LUNDHOLM N. & MOESTRUP Ø. 2002. The marine diatom Pseudo-nitzschia galaxiae sp. nov. (Bacillariophyceae): morphology and phylogenetic relationships. Phycologia 41: 594–605. MAIBOHM C., FRIIS S. M. M., ELLEGAARD M. & ROTTWITT K. 2015a. Interference patterns and extinction ratio of the diatom Coscinodiscus granii. Optics Express 23: 9543. MAIBOHM C., FRIIS S. M. M., SU Y. & ROTTWITT K. 2015b. Comparing optical properties of different

103 species of diatoms 9360: 93600B. MCPHEDRAN R. C., NICOROVICI N. A., MCKENZIE D. R., BOTTEN L. C., PARKER A. R. & ROUSE G. W. 2001. The sea mouse and the photonic crystal. Australian Journal of Chemistry 54: 241–244. NELSON D. M. & BRZEZINSKI M. A. 1997. Diatom growth and productivity in an oligo-trophic midocean gyre: A 3-yr record from the Sargasso Sea near Bermuda. Limnology and Oceanography 42: 473–486. PARKER A. R. & TOWNLEY H. E. 2007. Biomimetics of photonic nanostructures.. Nature nanotechnology 2: 347–353. ROMANN J., VALMALETTE J.-C., CHAUTON M. S., TRANELL G., EINARSRUD M.-A. & VADSTEIN O. 2015. Wavelength and orientation dependent capture of light by diatom frustule nanostructures. Scientific Reports 5: 17403. Nature Publishing Group. ROUND F. E., CRAWFORD R. M. & MANN D. G. 1990. The Diatoms: Biology and Morphology of the Genera. Cambridge Univ. Press, Cambridge, UK, 1990. SCHINDELIN J., ARGANDA-CARRERAS I., FRISE E., KAYNIG V., LONGAIR M., PIETZSCH T., PREIBISCH S., RUEDEN C., SAALFELD S., SCHMID B., TINEVEZ J.-Y., WHITE D. J., HARTENSTEIN V., ELICEIRI K., TOMANCAK P. & CARDONA A. 2012. Fiji: an open-source platform for biological-image analysis. Nature Methods 9: 676–682. DE STEFANO L., REA I., RENDINA I., DE STEFANO M. & MORETTI L. 2007. Lensless light focusing with the centric marine diatom Coscinodiscus walesii.. Optics express 15: 18082–18088. TOMMASI E. DE 2016. Light manipulation by single cells: the case of diatoms. Journal of Spectroscopy 2016: 1–13. DE TOMMASI E., REA I., MOCELLA V., MORETTI L., DE STEFANO M., RENDINA I. & DE STEFANO L. 2010. Multi-wavelength study of light transmitted through a single marine centric diatom. Optics express 18: 12203–12212. SU Y., LUNDHOLM N. & ELLEGAARD M. 2017. Long-term cultivation of the diatom Coscinodiscus granii at different light spectra: effects on frustule morphology. Applied Phycology 1-5. TOSTER J., IYER K. S., XIANG W., ROSEI F., SPICCIA L. & RASTON C. L. 2013. Diatom frustules as light traps enhance DSSC efficiency. Nanoscale 5: 873–876. VALMALETTE J., ROMANN J., RØYSET A. & EINARSRUD M. 2015. 3D-Hyperspectral mapping of light propagation through diatom frustule silica nanostructures. Optics Letters 40: 3–5. VUKUSIC P. & SAMBLES J. R. 2003. Photonic structures in biology. Nature 424: 852–5. VUKUSIC P., SAMBLES J. R., LAWRENCE C. R. & WOOTTON R. J. 1999. Quantified interference and diffraction in single Morpho butterfly scales. Proceedings of the Royal Society of London. Series B: Biological Sciences 266: 1403–1411. YAMANAKA S., YANO R., USAMI H., HAYASHIDA N., OHGUCHI M., TAKEDA H. & YOSHINO K. 2008. Optical properties of diatom silica frustule with special reference to blue light. Journal of Applied Physics 103: 1–6.

104 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Figures

Figure 1: Structure of the centric diatom frustule in C. granii . Exploded scheme of the silicate frustule from the side showing its indivisible parts, i.e. two valves and two girdles ( cingula ) named epicingulum and hypocingulum . Each cingulum is associated with a larger valve (epivalve) and a smaller valve (hypovalve). Together, epivalve and epicingulum form the epitheca , while hypovalve and its corresponding hypocingulum form the hypotheca . The two cingula can also be described as the girding ( cinctura , after Round et. al. 1990). The shown example of a C. granii frustule comprises only one band per cingulum , while cingula can bear two to many bands in other species (Cox 2011). Chloroplasts and their location inside the living cell are indicated in brown color.

105

106 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Figure 2: Optical effects of immersed valves, girdle bands and live diatom cells in the centric diatom C. granii observed with light microscopy. A) Oxidized valves and girdle bands observed with dark field microscopy. B) Magnification of a girdle band with alternating blue and red coloration observed with dark field microscopy. C) Defocused oxidized intact theca observed with dark field microscopy, with visible blue coloration of the valve and iridescent colors produced by the girdle band. D) Live cell in lateral view showing scattering of blue light on the interior side of the valve (bottom left). The location of the girding is indicated by a white frame. E) Dying cell with detaching girdle band and aggregated chloroplasts observed with bright field microscopy. F) Cell from panel E observed with dark field microscopy. Images in panels A, B, C, and F were recorded with light at a 25° angle of incidence through an objective of small numerical aperture (N 4x/NA=0.13). the image in panel D was recorded with light at a 25° angle of incidence through an objective of large numerical aperture (N 20x/NA=0.5). The image in panel E was recorded with light at normal incidence through an objective of large numerical aperture (N 20x/NA=0.5).

107

108 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Figure 3: Nanostructure of valve and girdle band in the centric diatom C. granii . A) SEM image of highly concentrated valves and girdle bands. B) SEM image of a single valve and several girdle bands of different sizes. C) Magnification of a SEM image showing the interior valve side with large foramina pores (~550 nm), and the advalvar side of a girdle band with small pores (~10 nm). D) TEM image showing the nanoporous structures in a valve and a girdle band. E) Enlarged model of the hexagonal nanostructure in the C. granii valve, with small cribrum pores on the exterior side, a large chamber in the center, and a large foramen pore on the interior side. F) Enlarged model of the girdle band nanostructure with channel-like structures (diameter ~10 nm).

109

110 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Figure 4: Light transmittance through valves and girdle bands of the centric diatom C. granii . A) Transmittance spectra measured through the optically different areas on the valve surface. B) Transmittance spectra through a girdle band measured for different ROIs. C) Interior surface of an oxidized valve imaged with a hyperspectral camera in the light microscope, indicating optically different areas on the valve surface. D) Girdle band imaged with a hyperspectral camera in the light microscope. Optically different regions of interest (ROI) are indicated, while the focus plane is on the top of the standing girdle band.

111

112 6.2. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules

Figure 5: Light reflectance on the surface of valves and girdle bands in the centric species C. granii . A) Reflectance on the exterior and interior side of an oxidized valve. B) Reflectance on the girdle band at different angles of incidence. C) Photography of valves and girdle bands in a drop of water observed at 200x magnification in the light microscope, while incident white light was shine in a 30° angle. Valves appear translucent, but girdle bands reflected light in alternating colors as a function of angular location to the incident white light. D) Angular dependent transmittance of red, green or blue light on the surface of concentrated frustules in a drop of water was also visible with the naked eye.

113

114

1 6.3 Photo-protection in the centric diatom Coscinodiscus granii is not

2 promoted by chloroplast high-light avoidance movement

3 Johannes W. Goessling 1* , Paulo Cartaxana 1 and Michael Kühl 1,2

4

5 1Marine Biology Section, Department of Biology, University of Copenhagen, Strandpromenaden 5, 3000 Helsingør, Denmark

6 2 Climate Change Cluster, University of Technology Sydney, Australia

7

8 *Corresponding author: Johannes Goessling; [email protected]; Strandpromenaden 5, 3000 Helsingør, Denmark

115

116 ORIGINAL RESEARCH published: 08 January 2016 doi: 10.3389/fmars.2015.00115

Photo-Protection in the Centric Diatom Coscinodiscus granii is Not Controlled by Chloroplast High-Light Avoidance Movement

Johannes W. Goessling 1*, Paulo Cartaxana 1 and Michael Kühl 1, 2

1 Marine Biological Section, Department of Biology, University of Copenhagen, Helsingør, Denmark, 2 Plant Functional Biology and Climate Change Cluster, University of Technology Sydney, Sydney, Australia

Diatoms are important phototrophs in the worlds’ oceans contributing ∼40% of the global primary photosynthetic production. This is partially explained by their capacity to exploit environments with variable light conditions, but there is limited knowledge on how diatoms cope with changes in the spectral composition and intensity of light. In this study, the influence of light quality and high irradiance on photosynthesis in the centric diatom Coscinodiscus granii was investigated with microscopic imaging and variable chlorophyll fluorescence techniques. Determination of the wavelength-dependent functional absorption cross-section of photosystem (PS) II revealed that absorption of blue light (BL) and red light (RL) was 2.3- and 0.8-fold that of white light (WL), respectively. Hence, BL was more efficiently converted into Edited by: photo-chemical energy. Excessive energy from BL was dissipated via non-photochemical Karla B. Heidelberg, University of Southern California, USA quenching (NPQ) mechanisms, while RL apparently induced only negligible NPQ even Reviewed by: at high irradiance. A dose dependent increase of cells exhibiting an altered chloroplast Jean-David Rochaix, distribution was observed after exposure to high levels of BL and WL, but not RL. University of Geneva, Switzerland However, no effective quantum yield of PSII was measured in the majority of cells with Todd Kana, ® Horn Point Lab, USA an altered chloroplast distribution, and positive Sytox green death staining confirmed *Correspondence: that most of these cells were dead. We conclude that although C. granii can sustain Johannes W. Goessling high irradiance it does not perform chloroplast high-light avoidance movements for [email protected] photo-protection.

Specialty section: Keywords: PSII functional absorption cross-section, spectral quality, light stress, non-photochemical quenching, This article was submitted to photosynthesis, high-light avoidance movement, diatoms Aquatic Microbiology, a section of the journal Frontiers in Marine Science INTRODUCTION Received: 12 October 2015 Accepted: 10 December 2015 Diatoms are responsible for about 40% of the primary production in marine ecosystems and Published: 08 January 2016 account for up to 20% of global carbon fixation (Nelson and Brzezinski, 1997; Geider et al., Citation: 2001). The evolutionary success and high productivity of diatoms (Boyd et al., 2000; Thomas and Goessling JW, Cartaxana P and Dieckmann, 2002; Mock and Valentin, 2004) seem to be closely related to their ability to adapt Kühl M (2016) Photo-Protection in the to environmental fluctuations such as changes in irradiance and spectral composition of the light Centric Diatom Coscinodiscus granii is Not Controlled by Chloroplast field (Depauw et al., 2012). The in situ solar irradiance in aquatic systems can be affected on High-Light Avoidance Movement. different time scales by various factors such as vertical mixing, wave focusing, and varying cloud Front. Mar. Sci. 2:115. cover. Vertical shifts in the water column change the blue light (BL) to red light (RL) ratio, where doi: 10.3389/fmars.2015.00115 RL is more strongly absorbed by seawater, while BL penetrates deeper into the water column in

Frontiers in Marine Science | www.frontiersin.org 1 January 2016 | Volume 2 | Article 115

117 Goessling et al. Photo-Protection in Coscinodiscus granii the absence of large amounts of dissolved organic carbon and conspicuous ∼50–500 µm wide silica frustule surrounding the particles (Kirk, 1994). A unique feature of diatoms is the Coscinodiscus cell (Jeffryes et al., 2015). Hitherto, most studies encasement of the cell in a silicate frustule composed of two have focused on empty frustules and to our knowledge the overlapping sections known as thecae or valves (Schmid and in vivo light environment and how intact living Coscinodiscus Volcani, 1983). The frustule exhibits special optical properties cells absorb, propagate and dissipate light energy as a function and this has led to speculations that the frustule nanostructure of irradiance and spectral composition has not yet been could affect the photobiology of diatoms (Sumper and Brunner, analyzed. In this study, photosynthesis-light response curves, 2006; De Stefano et al., 2007; Parker and Townley, 2007; Su non-photochemical quenching (NPQ), and the wavelength- et al., 2015) and thus potentially modulate their photosynthetic dependent functional absorption cross-section of PS II efficiency. [Sigma(II)λ] were determined for blue, red and white actinic If the photosynthetic antennae absorb too much light, light in the centric diatom Coscinodiscus granii using variable the excessive energy load can cause severe damages to the chlorophyll fluorescence techniques. Mechanisms involved photosynthetic system, e.g., when electrons are transferred to in the stabilization of the cellular light environment and in other acceptors such as O2 causing formation of reactive oxygen photo-protection are discussed with emphasis on the observed species (ROS; Apel and Hirt, 2004). There are two main strategies alteration of the chloroplast distribution in response to high for phototrophic organisms to deal with excessive light energy levels of blue and white light. (Demming-Adams and Adams, 1992). The first strategy involves mechanisms to dissipate excessive energy by activation of MATERIALS AND METHODS alternative electron sinks such as photorespiration, and by photo- protective mechanisms such as the diadinoxanthin (Ddx) cycle Cultivation, Growth Conditions, and responsible for the remarkable potential of diatoms to release excessive energy as heat (Schumann et al., 2007; Goss and Jakob, Experimental Set-Up Semi-continuous cultures of C. granii strain K-1831 2010). The second strategy involves avoidance mechanisms in (Scandinavian Culture Collection, Copenhagen, Denmark) order to reduce the harvesting of incident light that cannot were grown at 15◦C in L1 medium with silica (250 µM Na SiO ) be used for photosynthesis. On the long term, such avoidance 2 3 in 35 permil sterile filtrated seawater (Guillard and Hargraves, can be achieved by changes in the pigmentation (e.g., Kiefer, 1993). White light fluorescence light tubes (18W/865, Osram, 1973), increased light reflection (e.g., Kasperbauer, 1987), or by Munich, Germany) provided illumination with an incident reduction of the light harvesting antenna complex (Demming- photon irradiance (PAR, 400–700nm) of 100 µmol photons m−2 Adams and Adams, 1992). On the short term, avoidance can s−1 from the bottom of a cultivation table in a 14/10 h day/night also be achieved by light-induced chloroplast re-allocation in the cycle. Microalgae were maintained in 50mL polyethylene flasks cell (Haupt, 1973; Kasahara et al., 2002). While photo-protection (Nunc cell culture treated EasYFlasks , Thermo Scientific, mechanisms are typically controlled by substrate and product ™ ™ Waltham, United States) without any forced aeration or rotation. feedbacks or redox switches (Schellenberger Costa et al., 2013), If not stated otherwise, experiments were performed on three light-induced movement is often triggered by photoreceptors biological replicates of C. granii in the exponential phase. (Strasser et al., 2010). High-light avoidance movement of chloroplasts is primarily PSII Absorption Cross Section and induced by BL and high irradiance, while back movement to a more evenly chloroplast distribution in the cell is mediated Chlorophyll Fluorescence Rapid Light by far RL (>710 nm) and under low irradiance (e.g., DeBlasio Curves et al., 2003; Gabry´s, 2004). Reduction of light exposure through To account for differences in the wavelength specific absorption chloroplast high-light avoidance movement can minimize photo- of light in subsequent experiments, the functional absorption damage and optimize photosynthetic performance (Park et al., cross section of PS(II), Sigma(II)λ, was determined. Sigma(II)λ 1996; Kasahara et al., 2002; Wada et al., 2003). Chloroplast was further used to calculate absolute electron transport rates high-light avoidance movement occurs in higher plants, mosses, as derived from chlorophyll fluorescence rapid light curves. ferns, green algae, and dinophytes. It has also been observed in For this, a 1 mL cell culture sample was transferred to a the pennate diatoms Biddulphia pellucida and Seminavis robusta quartz cuvette, which was then placed in the optical unit of (Gillard et al., 2008), as well as in centric diatom species such a multicolor pulse amplitude modulated (PAM) chlorophyll as Ditylum blightwellii, Pleurosira laevis, Odontella regia, and fluorescence photosynthesis analyzer (MC-PAM, Heinz Walz Lauderia borealis (Kiefer, 1973; Chen and Li, 1991; Furukawa GmbH, Effeltrich, Germany) with a long pass filter (RG et al., 1998). The ability of diatoms to re-arrange chloroplasts in 665, Heinz Walz GmbH, Effeltrich, Germany) attached to a fluctuating light environment might thus maximize their light the fluorescence detector. The photon irradiance levels of the use efficiency and contribute to the high productivity of diatoms LED array for specific program settings were calibrated with (Boyd et al., 2000; Thomas and Dieckmann, 2002; Mock and a spherical micro quantum sensor (US-SQS/WB, Heinz Walz Valentin, 2004). GmbH, Effeltrich, Germany). Values were corrected using a zero Members of the Coscinodiscaceae family recently observe offset measurement on L1 medium. A three-fold gain and a four- an increasing interest in the scientific community due to fold damping were set for all measurements. Non-actinic pulse- the optical properties and the industrial applicability of the amplitude modulated blue (BL), red (RL), or white measuring

Frontiers in Marine Science | www.frontiersin.org 2 January 2016 | Volume 2 | Article 115

118 Goessling et al. Photo-Protection in Coscinodiscus granii

light was used. During illumination with actinic light and during Using Sigma(II)λ values and PARmax data from ETR vs. application of a 0.6 s saturating light pulse, the measuring irradiance measurements described above, experimental photon light frequency was increased to 100 kHz. Blue measuring light irradiance levels were adjusted to 600 µmol photons m−2 and blue saturation pulses were applied during blue actinic s−1 BL, 1500 µmol photons m−2 s−1 RL, and 1250 µmol light measurements. Red measuring light and red saturating photons m−2 s−1 WL. In this way, an equal amount of pulses were used when red or white actinic light was used. photons entering PS(II) was ensured. Photosynthetic active The fluorescence yield (F) was recorded prior to a saturation radiation (PAR) was measured with a spherical irradiance pulse analysis yielding the maximum fluorescence yield in probe connected to a quantum irradiance meter (model ULM dark adapted cells (Fm) or the maximum fluorescence yield 500, Heinz Walz GmbH, Effeltrich, Germany). Except for a ′ under different levels of AL (Fm ). The effective photochemical control kept in the dark, actinic BL, RL, or WL was applied (Fm′−F) for 1 h. quantum yield of PS(II) was calculated as: Y (II) = Fm (Genty et al., 1989). Actin-mediated cytoskeleton movement was inhibited by Sigma(II)λ was derived in the fast acquisition mode of the adding 1 µL of 0.25 µM latrunculin A (LAT-A) solution per multicolor PAM, using the script-file Sigma100.FTM (Schreiber mL of culture (cat. no. L12370, Thermo Scientific, Waltham, et al., 2012) followed by a curve fit applied in the system software United States), while myosin-mediated cytoskeleton movement PamWin3 (Heinz Walz GmbH, Effeltrich, Germany). The change was inhibited with 100 µL of 50 mM 2,3-butanedione monoxime of Sigma(II)λ values for BL and RL in relation to WL were (BDM) solution (cat. no. B0753, Sigma-Aldrich, St. Louis, calculated [Sigma(II)λ BL or RL per Sigma(II)λ WL]. Sigma(II)λ Missouri, United States) per mL of cell culture (Cartaxana et al., values are presented as antenna per nm2 (Schreiber et al., 2012). 2008). LAT-A and BDM stock solutions were prepared in DMSO Parameters of the light dependent PSII electron transport were and culture medium, respectively. After the experiment, samples derived from chlorophyll fluorescence data during 10 s irradiance were kept in the dark for 20h to recover. steps, i.e., rapid light curves (Wight and Critchley, 1999). The A dose-dependent alteration of the chloroplast distribution magnetic stirrer in the quartz cuvette was switched off before was estimated by increasing BL from 0 to 1000 µmol photons − − the saturation pulse was applied. Absolute electron transport m 2 s 1. Samples were tested after 10, 30, or 60min of Y(II) illumination. Chloroplast distribution patterns were determined rates at PSII, ETR were calculated as ETR = PAR(II) Y(II)max . at 200x magnification using a light microscope (model Axiostar PAR(II) was calculated as PAR (II) = PAR0.6022 Sigma(II)λ, where 0.60221024 is Avogadro’s constant with the dimension of plus FL, Carl Zeiss, Jena, Germany). In each replicate, 50–100 mol−1 (Schreiber et al., 2012). cells were evaluated. Curve fits on ETR vs. irradiance curves were performed according to Eilers and Peeters (1988), using the Microsoft Excel Cell Imaging and Sytox® Green Lethality add-on Solver (Microsoft Corporation, Redmond, USA). The Stain initial slope, alpha, the irradiance level where PS(II) reaches light SytoxR green lethality stain was used to test the viability of cells saturation, Ek, the maximum electron transport rate of PSII, with an altered or an evenly distributed chloroplast pattern. ETRmax, and the irradiance at ETRmax, PARmax, were calculated Altered chloroplast distribution in C. granii kept in tissue from curve fits to the Sigma(II)λ corrected data according to culture plates (TC Plate 24 Well, StandardF, Sarstedt, Nümbrecht, Eilers and Peeters (1988). Germany) was induced by exposure to 600 µmol photons m−2 −1 R Light Stress Experiments s BL for 1h. Samples were then stained with 0.5 µM Sytox green (Invitrogen , Carlsbad, California) for 15min in the Light stress experiments were performed to estimate color- ™ dark according to Armbrecht et al. (2014), prior to counting and dose-dependent effects on an alteration of the chloroplast under epifluorescence light. SytoxR green stained cells showing distribution. Inhibitors were used to inhibit cytoskeleton an even or altered chloroplast distribution were determined mediated high-light avoidance movement and back-movement at 200x magnification using an Olympus BX50 epifluorescence toward an even chloroplast distribution (ECD). Light intensities microscope. were adjusted on the base of Sigma(II)λ values as described above. Two milliliter of cell culture (500–1000 cells mL−1) were transferred to each well of a tissue culture plate (TC Plate Single Cell Analysis of Variable Chlorophyll 24 Well, StandardF, Sarstedt, Nümbrecht, Germany). During Fluorescence the light stress experiments, the lid of the cell culture plate Variable chlorophyll fluorescence was perform in order to was removed. A monochromatic blue LED (473 ± 13 nm), estimate the photosynthetic capacity of cells with an altered a red LED (636 ± 11nm), and a white LED (452–647nm or with an even chloroplast distribution, respectively. Single- bandwidth) were used to subject the cells to high light stress cell variable chlorophyll fluorescence was investigated with and induce an altered chloroplast distribution, here defined a red-green-blue (RGB) pulse-amplitude-modulated (PAM) as when chloroplasts aggregated within the cortical cytoplasm. chlorophyll fluorescence imaging system (RGB Microscopy- Empty frustules and broken cells that released their chloroplasts IPAM, Heinz Walz GmbH, Germany) mounted on an epi- were not considered. The spectral range of the LED arrays was fluorescence microscope (model Axiostar plus FL, Carl Zeiss, determined with a calibrated spectroradiometer (SpectriLight, Jena, Germany). The setup is described in detail by Trampe International Light Technologies Inc., Peabody, USA). et al. (2011). Measurements were performed with a 20x

Frontiers in Marine Science | www.frontiersin.org 3 January 2016 | Volume 2 | Article 115

119 Goessling et al. Photo-Protection in Coscinodiscus granii

TABLE 1 | Photosynthetic characteristics of Coscinodiscus granii incubated under blue (BL), red (RL), and white (WL) light.

Description Parameter BL RL WL

Sigma determination Sigma 3.5 ± 0.4a 1.3 ± 0.2b 1.5 ± 0.10b ETR(II) curve fit parameter Alpha(ll) 0.44 ± 0.19a 0.08 ± 0.02b 0.06 ± 0.01b Pmax(ll) 41.4 ± 5.2a 35.1 ± 7.1a 32.3 ± 5.7a Ek(ll) 124 ± 42a 458 ± 21b 572 ± 70b PARmax(ll) 552 ± 40a 993 ± 71b 1145 ± 141b Experimental design Sigma fold WL 2.32 ± 0.41a 0.86 ± 0.17b 1.00 ± 0.00b LED spectral range [nm] 473 ± 13 636 ± 11 452–647 Intensities in Exp. [µmol m−2s−1 600 1500 1250

The wavelength dependent functional absorption cross-section of photosystem (PS) II [Sigma(II)λ ] was determined with a multicolor chlorophyll fluorescence analyzer. ETR(II) curve fit parameter were derived from ETR light curves using the equation of Eilers and Peeters (1988). Sigma(II)λ data were used to account for differences in the light absorbed by PS(II). Using the fold change to a WL source [Sigma(II)λ fold WL] similar light intensities were adjusted during the experiments. Statistically significant differences between the treatments are indicated as different small letters (p < 0.05).

− −2 −1 objective (Plan-Apochromate, Carl Zeiss GmbH, Germany). higher ETRmax under BL (41.4 ± 5.2 µmol e m s ) as Non-actinic blue measuring light was used and the effective compared to RL (35.1 ± 7.1 µmol e− m−2 s−1) and WL (32.3 ± PSII quantum yield (YII) was determined with a high intensity 5.7 µmol e− m−2 s−1) was not significant (p = 0.543). The saturating BL pulse. The maximum fluorescence yield (Fm) light saturation index Ik was significantly different between BL correction factor was set to 1.03, to account for signal (124.4 µmol photons m−2 s−1)andRL(458.2 µmol photons m−2 losses due to LED heating. The Y(II) values were calculated s−1; p < 0.001) and between BL and WL (572 ± 70 µmol using the system software (ImagingWin, Heinz Walz GmbH, photons m−2 s−1; p = 0.003), but not between RL and Germany). For each sample, quantitative analysis of variable WL (p = 0.149; Table 1). Significant differences in PARmax chlorophyll fluorescence was performed on 10 cells with evenly between BL (552 ± 40 µmol photons m−2 s−1; p = 0.003) distributed chloroplasts and 10 cells with an altered chloroplast and RL (993 ± 71 µmol photons m−2 s−1) and between BL distribution. Only up to 5 cells with an altered chloroplast and WL (1145 ± 141 µmol photons m−2 s−1; p = 0.010), distribution were analyzed in the control group because of lower but not between RL and WL (p = 0.250) were observed occurrence. (Table 1). Absolute ETR(II) was higher at BL than at RL and WL (Figure 1A). The dose-dependent increase of NPQ was higher Statistical Analysis with BL as compared to WL, while RL induced almost no NPQ Significant differences among treatment groups were determined (Figure 1B). with One- or Two-way analysis of variance (ANOVA) followed by Holm Sidak post-hoc tests. Data were log10-transformed to Light Stress Experiments normality if necessary. The ANOVA results are presented with An altered chloroplast distribution (Figures 2B,D) was observed F and p-values. Post-hoc results are indicated with p-values. in 32 ± 5% of cells after 1 h exposure to BL, which was Differences at the p < 0.05 level are reported as significant. significantly higher than in the control group (5 ± 2%; All statistical tests were carried out with SigmaPlot 11.0 (Systat p < 0.001) (Figure 3A). No significant differences were Software, Inc., Richmond, USA). found between RL-induced altered chloroplast distribution (8 ± 3%; p = 0.289) and the control group. An increase cells RESULTS with altered chloroplast distribution (to 19 ± 2%) after 1 h exposure to WL was significant (p < 0.001; Figure 3A). Functional PSII Absorption Cross Section No statistical significance was found when comparing the frequency of altered chloroplast distribution after 1 h of high and PSII Activity as a Function of light exposure and after 20h of recovery in the dark [F(1,23) = Irradiance. 0.819; p = 0.379; Figure 3A]. The frequency of cells with Significant differences in Sigma(II)λ were observed between BL altered chloroplast distribution in amounted to 22 ± 4% (3.5 ± 0.4 nm2) and both RL (1.3 ± 0.2 nm2; p < 0.001) and for BL, 9 ± 4% for RL, and 19 ± 15% for WL samples, WL (1.5 ± 0.1 nm2; p < 0.001), but not between RL and WL respectively. The presence of LAT-A did not significantly affect (p = 0.291; Table 1). Sigma(II)λ was 2.3 ± 0.4 fold higher in the altered chloroplast distribution as compared to untreated blue light than in white light, while Sigma(II)λ was 0.9 ± 0.2 samples [F(1,23) = 0.793; p = 0.386; Figure 3B]. After fold lower in red light than in white light (Table 1). Parameters of 20 h of recovery in the dark, altered chloroplast distribution rapid light curves derived from chlorophyll fluorescence analysis increased significantly in all samples treated with latrunculin revealed significantly higher alpha in BL (0.44 ± 0.19 µmol A (LAT-A) [F(1,23) = 31.825; p < 0.001]. Compared e− µmol−1 photons) compared to RL (0.08 ± 0.02; p = to untreated samples, presence of butanedione monoxime 0.006) and WL (0.06 ± 0.01; p = 0.03; Table 1). A trend of (BDM) significantly increased the altered chloroplast distribution

Frontiers in Marine Science | www.frontiersin.org 4 January 2016 | Volume 2 | Article 115

120 Goessling et al. Photo-Protection in Coscinodiscus granii

FIGURE 2 | Chloroplast distribution in cells of Coscinodiscus granii. (A,C) Even chloroplast distribution (ECD); (B,D) An altered chloroplast distribution (ACD) can be observed after harmful light doses in some of the cells. View from the top (A,B) and from the side (C,D).

Single Cell Variable Chlorophyll FIGURE 1 | Rapid chlorophyll fluorescence light-response curves of Coscinodiscus granii under blue, red, and white light. (A) Absolute PSII Fluorescence and Cell Mortality electron transport rates, ETR; (B) Non photochemical quenching, NPQ. Cells with altered chloroplast distribution exhibited a significant Average values and standard error (n = 3) are indicated. decrease in Y(II), independently from whether cells were treated with high levels of blue light (600 µmol m−2 s−1 BL) for 1 h, or if they were kept in the dark [F(1,19) = 727.167; p < 0.001; [F(1,23) = 31.778; p < 0.001; Figure 3C], and in the presence Figure 5A]. In untreated cells, Y(II) decreased from 0.44 ± 0.02 of BDM cells with an altered chloroplast distribution increased in cells with evenly distributed chloroplasts to 0.00 ± 0.01 in further after a 20h dark recovery period [F(1,23) = 29.153; p < cells with altered chloroplast distribution (p < 0.001; compare 0.001]. Figure 6). After BL exposure, Y(II) decreased from 0.35 ± 0.03 in cells with evenly distributed chloroplasts to 0.00 ± 0.08 in cells Time and Dose-Dependent Induction of with altered chloroplast distribution (p < 0.001). Recovery for Altered Chloroplast Distribution Patterns 20h in the dark did not significantly change the Y(II) of cells with altered chloroplast distribution [F = 0.199; p = 0.658; Both time and dose of BL-exposure did significantly increase (1,39) Figure 5B]. altered chloroplast distribution patterns [F(2,53) = 33.866; p < 0.001; Figure 4]. Compared to the control (dark), ® there was no significant increase in the altered chloroplast Sytox Green Lethality Staining R distribution pattern after 10 min of BL-exposure, even under Sytox green showed positive staining for only 7 ± 4% of a photon irradiance of 1000 µmol photons m−2 s−1 (p = cells with an even chloroplast distribution. The percentage cells 0.275). After 30 min of BL-exposure, there was a significantly showing an altered chloroplast distribution pattern that were R increased frequency of cells with altered chloroplast distribution stained with Sytox green was significantly higher [F(1,5) = [12 ± 5% at 700 µmol photons m−2 s−1 (p = 0.001), 27.505; p = 0.006] reaching 51 ± 9% (Figure 7). and of 18 ± 4% at 1000 µmol photons m−2 s−1 (p < 0.001)]. A significantly increased frequency of cells with altered DISCUSSION chloroplast distribution to 14 ± 3% after 60 min was found under lower photon irradiance of 300 µmol photons m−2 In the present study, effects of BL, RL, and WL on photosynthesis, s−1 of BL (p < 0.001). After 60 min of BL exposure, and light-induced alteration of the chloroplast distribution in the frequency of altered chloroplast distribution did further the centric diatom C. granii were investigated. Blue light was increase reaching maximum values of 37 ± 5% under a channeled much more efficiently to PSII than WL and RL based −2 −1 photon irradiance of 1000 µmol photons m s of BL (p < on Sigma(II)λ measurements (Table 1). This observation is in 0.001). accordance with studies on the pennate diatom Phaeodactylum

Frontiers in Marine Science | www.frontiersin.org 5 January 2016 | Volume 2 | Article 115

121 Goessling et al. Photo-Protection in Coscinodiscus granii

FIGURE 4 | Induction of altered chloroplast distribution (ACD) in Coscinodiscus granii after 10, 30, and 60 min of exposure to increasing BL photon irradiance levels. Average values and standard error (n = 3) are indicated.

efficient mechanisms to dissipate excessive irradiance and/or to prevent damages to the photosynthetic apparatus. BL, RL, and WL were used to induce an alteration of the chloroplast distribution in C. granii. Based on measurements of Sigma(II)λ we could adjust the actinic light levels under the different spectral regimes ensuring equal amounts of light absorbed by the PS(II) antennae, i.e., 600 µmol m−2 s−1 of BL, 1500 µmol m−2 s−1 of RL, and 1250 µmol m−2 s−1 of WL; all irradiance levels were well beyond the saturation point of photosynthesis in Coscinodiscus (Table 1). Under low light conditions, the chloroplasts were evenly distributed within the cortical stroma (Figures 2A,C). Only high doses of BL and WL promoted an alteration of the intracellular chloroplast arrangement in a significant number of cells by aggregation of chloroplasts in the cortical cytoplasm (Figures 2B,D). Interestingly, the number of cells showing an FIGURE 3 | Proportion of Coscinodiscus granii cells with an altered alteration of the chloroplast arrangement was not increased chloroplast distribution (ACD), after 1 h exposure to either 600 µmol when high RL was applied (Figure 3A), although incident RL −2 −1 −2 −1 −2 −1 m s BL, 1500 µmol m s RL, or 1250 µmol m s WL (after irradiances were much higher (Table 1). High-light avoidance exposure) and after 20 h of recovery in the dark (after 20 h recovery) (A) movement of chloroplasts and reversion toward an evenly in absence and presence of the cytoskeleton inhibitors LAT-A (B) and BDM (C). Average values and standard error (n = 3) are indicated. Statistically distribution under low light can reduce light-induced stress and significant differences between the treatments are indicated as different small optimize photosynthetic performance under fluctuating light letters (p < 0.05). regimes (Haupt, 1973). Chloroplast movement is mediated by photoreceptors that have been found in diatoms (Schnitzler Parker et al., 2004; Montsant et al., 2005; Bowler et al., 2008; tricornutum (Schellenberger Costa et al., 2013) and other Depauw et al., 2012). It was therefore expected that C. granii phytoplankton species (Gilbert et al., 2000). Photosynthetic performed chloroplast high-light avoidance movement under electron transport from PSII, ETR(II) saturated sooner under BL, high light stress. Yogamoorthi (2007) observed a dose-dependent and reached much higher levels than under RL or WL (Table 1; alteration of the chloroplast arrangement in Coscinodiscus gigas Figure 1A). We also observed a higher energy dissipation via toward a more aggregated state after treatment with UV- NPQ under BL (Figure 1B) while RL-induced NPQ was very B irradiation. In our study, an increase of cells exhibiting low and did not significantly increase with irradiance. Overall, an altered/aggregated chloroplast arrangement was observed high Ek and high ETRmax observed under all applied spectral depending on irradiance level, exposure time and light color. regimes indicate an outstanding photosynthetic capacity of However, we did not observe any reversion toward a more even C. granii (Table 1), indicating that this diatom can employ chloroplast distribution pattern even after a recovery of 20 h in

Frontiers in Marine Science | www.frontiersin.org 6 January 2016 | Volume 2 | Article 115

122 Goessling et al. Photo-Protection in Coscinodiscus granii

FIGURE 5 | Variable chlorophyll fluorescence in cells of Coscinodiscus granii showing an even chloroplast distribution (ECD) or an altered chloroplast distribution (ACD) after exposure to high blue light (A) and after recovery in the dark (B). Statistically significant differences between the treatments are indicated as different small letters (p < 0.05).

′ FIGURE 6 | Maximal chlorophyll fluorescence yield, Fm , and effective PSII quantum yield, Y(II), in Coscinodiscus granii cells during alteration of the chloroplast arrangement under high BL photon irradiance. In cells with an even chloroplast distribution (ECD) (A–D), Y(II) decreased upon illumination with BL for 30 min (B) and recovered after 30 min in the dark (C). After an alteration of the chloroplast distribution, Y(II) decreased after illumination with BL for 30 min (E) and did not recover after 30 min in the dark (F). the dark (Figure 3A). Such reversion could be expected under Krzeszowiec et al., 2007; Paves and Truve, 2007). In presence of low light conditions or upon far RL illumination (Kasahara inhibitors of both actin- and myosin-mediated movements, we et al., 2004) and in the diatom Pleurosira laevis, reversal from found an increasing number of C. granii cells with an altered an aggregated chloroplast distribution toward a more even chloroplast pattern (Figures 3B,C). These results are in contrast distribution has also been observed under green light (Shihira- to the observation of Shihira-Ishikawa et al. (2007), describing Ishikawa et al., 2007). inhibition of BL-induced high-light avoidance movement of Both, high-light avoidance movements and back movement chloroplasts with myosin-disrupting agents, and inhibition of are mediated through the cytoskeleton. While high-light green light (GL)-induced back movement using actin-inhibitors avoidance movement is mediated by myosin, back movement in the centric diatom Pleurosira laevis. After recovery in the is regulated by actin filaments (Sato et al., 2001; Takagi, 2003; dark, we observed that the amount of C. granii cells with

Frontiers in Marine Science | www.frontiersin.org 7 January 2016 | Volume 2 | Article 115

123 Goessling et al. Photo-Protection in Coscinodiscus granii

FIGURE 7 | Lethality test using Sytox® green staining on Coscinodiscus granii cells with an even chloroplast distribution (ECD) (A,C) or an altered chloroplast distribution (ACD) (B,D). (E) Proportion of cells that were positively tested on lethality. (C) Cells with an ECD always showed a pronounced chlorophyll auto-fluorescence. Average values and standard error (n = 3) are indicated. Statistically significant differences between the treatments are indicated as different small letters (p < 0.05). an altered chloroplast distribution even increased in presence period of >30min and then the alteration process happened of cytoskeleton inhibitors (Figures 3B,C), indicating potential within seconds, occasionally followed by a cracking of the frustule lethal effects of such long-term incubation. It was also surprising and the release of chloroplasts. This observation indicates a time- that the number of cells showing an alteration in chloroplast dependent accumulation of stress under high light exposure, arrangement did not increase within 10 min at high BL doses until a critical concentration threshold is reached triggering (Figure 4) and only after 30 min at >700 µmol photons m−2 rapid movement and aggregation of chloroplasts. We speculate s−1, did the number of cells with an altered chloroplast pattern that such stress could be related with the formation of reactive increase significantly. Earlier studies showed that high-light oxygen species, combined with a dramatic decrease of the luminal avoidance chloroplast movement is induced within a few minutes redox state to a level, which can no longer be buffered by in terrestrial phototrophs (Kasahara et al., 2002; Wada et al., protection mechanisms such as the Ddx cycle. We note that 2003) as well as in micro- and macro-algae (Haupt, 1973), the light doses needed to induce such an apparently irreversible including diatoms (e.g., Furukawa et al., 1998). Fast induction alteration of the chloroplast arrangement reflect irradiance of chloroplast high-light avoidance movement might prevent maxima occurring in the aquatic environment of Coscinodiscus. photo-induced damages of the photo apparatus on the short term Nevertheless, it is even more astonishing that a majority but if such movement would take too long, the system could of cells did neither show an alteration of their chloroplast already be damaged. arrangement (Figure 3A), nor a significant decrease in variable To estimate the photosynthetic state of cells with an altered chlorophyll fluorescence even after long term exposure of 1 h chloroplast distribution, measurements of variable chlorophyll (Figure 4). Such apparent high light tolerance of C. granii was fluorescence were performed on the cellular scale. Surprisingly, supported by the observation that cells with an evenly distributed cells with an altered chloroplast distribution showed a dramatic chloroplast pattern did not show a positive SytoxR green death loss PSII quantum yield, Y(II) (Figures 5A, 6). While a staining even after 1 h of exposure to high BL irradiance reduction could be interpreted as a physiological protection (Figure 7). to high light (Genty et al., 1989), total loss of Y(II) rather Our results suggest that C. granii can sustain high irradiance. indicates absence of photochemical quenching and an inactive However, this capacity is not promoted by chloroplast high- photosynthetic apparatus. This could be explained by severe light avoidance movements, and Coscinodiscus may thus use damage due to excessive light. Furthermore, Y(II) of cells other mechanisms to stabilize their cellular light environment. with an altered chloroplast distribution did not recover even The effective cross-sections for photochemistry, of large diatom after a dark period of 20h (Figure 5B) and application of the species such as Coscinodiscus is usually lower compared to SYTOXR green stain revealed that a high proportion of cells smaller cells, hence showing lower susceptibility to photo with an altered chloroplast arrangement were actually dead inactivation (Key et al., 2010). Furthermore, diatoms have been (Figure 7). shown to have exceptionally high capacity for rapid and large The cause for alterations in the chloroplast distribution in induction of the xanthophyll cycle under light stress leading to C. granii remains unclear. After exposure to high irradiance, the thermal dissipation of harmful excess energy (Ruban et al., alteration of the chloroplast arrangement occurred only after a 2004). The silica frustule stabilizes the diatom cell and protects

Frontiers in Marine Science | www.frontiersin.org 8 January 2016 | Volume 2 | Article 115

124 Goessling et al. Photo-Protection in Coscinodiscus granii against predation (Hamm et al., 2003), but the frustule of diatoms FUNDING also has unique optical properties (e.g., Fuhrmann et al., 2004; De Stefano et al., 2007; Lettieri et al., 2008; Chen et al., 2015), This study was supported by a Sapere-Aude Advanced grant from and recent work suggest that the silica frustule also plays a the Danish Council for Independent Research Natural Sciences role in adaptation to spectral light changes (Su et al., 2015). (MK) and by the Carlsberg Foundation (MK). High photosynthetic efficiencies in the centric diatom C. granii thus might be promoted by physiological mechanisms and by ACKNOWLEDGMENTS structural features on the microscale, but not by chloroplast high-light avoidance movement. We thank Gert Hansen for providing cultures from the Scandinavian Culture Collection and Erik Trampe for help and introduction to variable chlorophyll fluorescence microscopy. AUTHOR CONTRIBUTONS We further thank Sonia Cruz, Mads Lichtenberg, and Klaus Koren for valuable input during the project conception and Experimental design (JG, PC, MK), experimental work (JG, PC), Sofie L. Jakobsen for technical assistance. We appreciate data analysis (JG, PC, MK). JG wrote manuscript with input from the significant improvement of this manuscript by two PC and MK. reviewers.

REFERENCES Gabry´s, H. (2004). Blue light-in duced ori en ta tion move ments of chloroplasts in higher plants: re cent prog ress in the study of their mechanisms. Acta Physiol. Apel, K., and Hirt, H. (2004). Reactive oxygen species: metabolism, oxidative Plant. 26, 473–478. doi: 10.1007/s11738-004-0038-3 stress, and signal transduction. Annu. Rev. Plant Biol. 55, 373–399. doi: Geider, R. J., Delucia, E. H., Falkowski, P. G., Finzi, A. C., Grime, J. P., Grace, 10.1146/annurev.arplant.55.031903.141701 J., et al. (2001). Primary productivity of planet earth: biological determinants Armbrecht, L. H., Smetacek, V., Assmy, P., and Klaas, C. (2014). Cell death and physical constraints in terrestrial and aquatic habitats. Glob. Chang. Biol. 7, and aggregate formation in the giant diatom Coscinodiscus wailesii (Gran & 849–882. doi: 10.1046/j.1365-2486.2001.00448.x Angst, 1931). J. Exp. Mar. Biol. Ecol. 452, 31–39. doi: 10.1016/j.jembe.2013. Genty, B., Briantais, J.-M., and Baker, N. R. (1989). The relationship between 12.004 the quantum yield of photosynthetic electron transport and quenching Bowler, C., Allen, A. E., Badger, J. H., Grimwood, J., Jabbari, K., Kuo, A., et al. of chlorophyll fluorescence. Biochim. Biophys. Acta 990, 87–92. doi: (2008). The Phaeodactylum genome reveals the evolutionary history of diatom 10.1016/S0304-4165(89)80016-9 genomes. Nature 456, 239–244. doi: 10.1038/nature07410 Gilbert, M., Wilhelm, C., and Richter, M. (2000). Bio-optical modelling of oxygen Boyd, P. W., Watson, A. J., Law, C. S., Abraham, E. R., Trull, T., Murdoch, R., evolution using in vivo fluorescence: comparison of measured and calculated et al. (2000). A mesoscale phytoplankton bloom in the polar Southern Ocean photosynthesis/irradiance (P-I) curves in four representative phytoplankton stimulated by iron fertilization. Nature 407, 695–702. doi: 10.1038/35037500 species. J. Plant Physiol. 157, 307–314. doi: 10.1016/S0176-1617(00) Cartaxana, P., Brotas, V., and Serôdio, J. (2008). Effects of two motility 80052-8 inhibitors on the photosynthetic activity of the diatoms Cylindrotheca Gillard, J., Devos, V., Huysman, M. J. J., De Veylder, L., D’Hondt, S., Martens, C., closterium and Pleurosigma angulatum. Diatoms Res. 23, 65–74. doi: et al. (2008). Physiological and transcriptomic evidence for a close coupling 10.1080/0269249X.2008.9705737 between chloroplast ontogeny and cell cycle progression in the pennate diatom Chen, S. T., and Li, C. W. (1991). Relationships between the movements of Seminavis robusta. Plant Physiol. 148, 1394–1411. doi: 10.1104/pp.108.122176 chloroplasts and cytoskeletons in diatoms. Botanica Marina 34, 505–511. doi: Goss, R., and Jakob, T. (2010). Regulation and function of xanthophyll cycle- 10.1515/botm.1991.34.6.505 dependent photoprotection in algae. Photosyn. Res. 106, 103–122. doi: Chen, X., Wang, C., Baker, E., and Sun, C. (2015). Numerical and experimental 10.1007/s11120-010-9536-x investigation of light trapping effect of nanostructured diatom frustules. Sci. Guillard, R. R. L., and Hargraves, P. E. (1993). Stichocrysis immobilis is a diatom, Rep. 5, 11977. doi: 10.1038/srep11977 not a chrysophyte. Phycologica 32, 234–236. doi: 10.2216/i0031-8884-32-3- DeBlasio, S. L., Mullen, J. L., Luesse, D. R., and Hangarter, R. P. (2003). 234.1 Phytochrome modulation of blue light-induced chloroplast movements in Hamm, C. E., Merkel, R., Springer, O., Jurkojc, P., Maier, C., Prechtel, K., et al. Arabidopsis. Plant Physiol. 133, 1471–1479. doi: 10.1104/pp.103.029116 (2003). Architecture and material properties of diatom shells provide effective Demming-Adams, B., and Adams, W. W. (1992). Photoprotection and other mechanical protection. Nature 421, 841–843. doi: 10.1038/nature01416 responses of plants to high light stress. Annu. Rev. Plant Physiol. 43, 599–626. Haupt, W. (1973). Chloroplast movement. J. Bio Sci. 23, 289–296. doi: 10.1146/annurev.pp.43.060192.003123 Jeffryes, C., Agathos, S. N., and Rorrer, G. (2015). Biogenic nanomaterials Depauw, F. A., Rogato, A., D’Alcalá, M. R., and Falciatore, A. (2012). Exploring from photosynthetic microorganisms. Curr. Opin. Biotechnol. 33, 23–31. doi: the molecular basis of responses to light in marine diatoms. J. Exp. Bot. 63, 10.1016/j.copbio.2014.10.005 1575–1591. doi: 10.1093/jxb/ers005 Kasahara, M., Kagawa, T., and Oikawa, K. (2002). Chloroplast avoidance De Stefano, L., Rea, I., Rendina, I., De Stefano, M., and Moretti, L. (2007). Lensless movement reduces photodamage in plants. Lett. Nat. 420, 829–832. doi: light focusing with the centric marine diatom Coscinodiscus walesii. Opt. 10.1038/nature01213 Express 15, 18082–18088. doi: 10.1364/OE.15.018082 Kasahara, M., Kagawa, T., Sato, Y., Kiyosue, T., and Wada, M. (2004). Phototropins Eilers, P. H. C., and Peeters, J. C. H. (1988). A model for the relationship between mediate blue and red light-induced chloroplast movements in Physcomitrella light intensity and the rate of photosynthesis in phytoplankton. Ecol. Modell. patens. Plant Physiol. 135, 1388–1397. doi: 10.1104/pp.104.042705 42, 199–215. doi: 10.1016/0304-3800(88)90057-9 Kasperbauer, M. J. (1987). Far-red light reflection from green leaves and effects Fuhrmann, T., Landwehr, S., El Rharbi-Kucki, M., and Sumper, M. (2004). Diatoms on phytochrome-mediated assimilate partitioning under field conditions. Plant as living photonic crystals. Appl. Phys. B 78, 257–260. doi: 10.1007/s00340-004- Physiol. 85, 350–354. doi: 10.1104/pp.85.2.350 1419-4 Key, T., McCarthy, A., Campbell, D. A., Six, C., Roy, S., and Finkel, Furukawa, T., Watanabe, M., and Shihira-Ishikawa, I. (1998). Green- and blue- Z. V. (2010). Cell size trade-offs govern light exploitation strategies in light-mediated chloroplast migration in the centric diatomPleurosira laevis. marine phytoplankton. Environ. Microbiol. 12, 95–104. doi: 10.1111/j.1462- Protoplasma 203, 214–220. doi: 10.1007/BF01279479 2920.2009.02046.x

Frontiers in Marine Science | www.frontiersin.org 9 January 2016 | Volume 2 | Article 115

125 Goessling et al. Photo-Protection in Coscinodiscus granii

Kiefer, D. A. (1973). Chlorophyll a fluorescence in marine centric diatoms: Schreiber, U., Klughammer, C., and Kolbowski, J. (2012). Assessment of responses of chloroplasts to light and nutrient stress. Mar. Biol. 23, 39–46. doi: wavelength-dependent parameters of photosynthetic electron transport with a 10.1007/BF00394110 new type of multi-color PAM chlorophyll fluorometer. J. Photosyn. Res. 113, Kirk, J. T. O. (1994). Light and Photosynthesis in Aquatic Ecosystems. Cambrigde 127–144. doi: 10.1007/s11120-012-9758-1 University Press. Schumann, A., Goss, R., Jakob, T., and Wilhelm, C. (2007). Investigation of the Krzeszowiec, W., Rajwa, B., Dobrucki, J., and Gabry´s, H. (2007). Actin cytoskeleton quenching efficiency of diatoxanthin in cells of Phaeodactylum tricornutum in Arabidopsis thaliana under blue and red light. Biol. Cell 99, 251–260. doi: (Bacillariophyceae) with different pool sizes of xanthophyll cycle pigments. 10.1042/BC20060077 Phycologia 46, 113–117. doi: 10.2216/06-30.1 Lettieri, S., Setaro, A., De Stefano, L., De Stefano, M., and Maddalena, P. (2008). Shihira-Ishikawa, I., Nakamura, T., Higashi, S. I., and Watanabe, M. (2007). The gas-detection properties of light-emitting diatoms. Adv. Funct. Mater. 18, Distinct responses of chloroplasts to blue and green laser microbeam 1257–1264. doi: 10.1002/adfm.200701124 irradiations in the centric diatom Pleurosira laevis. Photochem. Photobiol. 83, Mock, T., and Valentin, K. (2004). Photosynthesis and cold acclimation: molecular 1101–1109. doi: 10.1111/j.1751-1097.2007.00167.x evidence from a polar diatom. J. Phycol. 40, 732–741. doi: 10.1111/j.1529- Strasser, B., Sánchez-Lamas, M., Yanovsky, M. J., Casal, J. J., and Cerdán, P. D. 8817.2004.03224.x (2010). Arabidopsis thaliana life without phytochromes. Proc. Natl. Acad. Sci. Montsant, A., Jabbari, K., Maheswari, U., and Bowler, C. (2005). Comparative U.S.A. 107, 4776–4781. doi: 10.1073/pnas.0910446107 genomics of the pennate diatom Phaeodactylum tricornutum. Plant Physiol. Su, Y., Friis, S. M. M., Lundholm, N., and Ellegaard, M. (2015). Implications for 137, 500–513. doi: 10.1104/pp.104.052829 photonic applications of diatom growth and frustule nanostructure changes in Nelson, D. M., and Brzezinski, M. A. (1997). Diatom growth and productivity response to different light wavelengths. Nano Res. 8, 2363–2372. doi: 10.1007/ in an oligo-trophic midocean gyre: a 3-yr record from the Sargasso Sea near s12274-015-0746-6 Bermuda. Limnol. Oceanogr. 42, 473–486. doi: 10.4319/lo.1997.42.3.0473 Sumper, M., and Brunner, E. (2006). Learning from diatoms: nature’s tools for Park, Y. I., Chow, W. S., and Anderson, J. M. (1996). Chloroplast movement in the production of nanostructured silica. Adv. Funct. Mater. 16, 17–26. doi: the shade plant tradescantia albiflora helps protect photosystem II against light 10.1002/adfm.200500616 stress. Plant Physiol. 111, 867–875. Takagi, S. (2003). Actin-based photo-orientation movement of chloroplasts in Parker, A. R., and Townley, H. E. (2007). Biomimetics of photonic nanostructures. plant cells. J. Exp. Biol. 206, 1963–1969. doi: 10.1242/jeb.00215 Nat. Nanotechnol. 2, 347–353. doi: 10.1038/nnano.2007.152 Thomas, D. N., and Dieckmann, G. S. (2002). Antarctic Sea ice–a habitat for Paves, H., and Truve, E. (2007). Myosin inhibitors block accumulation movement extremophiles. Science 295, 641–644. doi: 10.1126/science.1063391 of chloroplasts in Arabidopsis thaliana leaf cells. Protoplasma 230, 165–169. doi: Trampe, E., Kolbowski, J., Schreiber, U., and Kühl, M. (2011). Rapid assessment of 10.1007/s00709-006-0230-y different oxygenic phototrophs and single-cell photosynthesis with multicolour Ruban, A., Lavaud, J., Rousseau, B., Guglielmi, G., Horton, P., and Etienne, A. variable chlorophyll fluorescence imaging. Mar. Biol. 158, 1667–1675. doi: L. (2004). The super-excess energy dissipation in diatom algae: comparative 10.1007/s00227-011-1663-1 analysis with higher plants. Photosyn. Res. 82, 165–175. doi: 10.1007/s11120- Wada, M., Kagawa, T., and Sato, Y. (2003). Chloroplast movement. Annu. Rev. 004-1456-1 Plant Biol. 54, 455–468. doi: 10.1146/annurev.arplant.54.031902.135023 Sato, Y., Wada, M., and Kadota, A. (2001). Choice of tracks, microtubules and/or Wight, A. J., and Critchley, C. (1999). Rapid light curves: a new fluorescence actin filaments for chloroplast photo-movement is differentially controlled by method to assess the state of the photosynthetic apparatus. J. Photosyn. Res. phytochrome and a blue light receptor. J. Cell Sci. 114(Pt 2), 269–279. Retrieved 59, 63–72. doi: 10.1023/A:1006188004189 from http://www.ncbi.nlm.nih.gov/pubmed/11148129 Yogamoorthi, A. (2007). Artificial UV-B induced changes in pigmentation of Schellenberger Costa, B., Jungandreas, A., Jakob, T., Weisheit, W., Mittag, M., marine diatom Coscinodiscus gigas. J. Environ. Biol. 28, 327–30. Retrieved and Wilhelm, C. (2013). Blue light is essential for high light acclimation and from: http://www.ncbi.nlm.nih.gov/pubmed/17915774 photoprotection in the diatom Phaeodactylum tricornutum. J. Exp. Bot. 64, 483–493. doi: 10.1093/jxb/ers340 Conflict of Interest Statement: The authors declare that the research was Schmid, A.-M. M., and Volcani, B. E. (1983). Wall morphogenesis in conducted in the absence of any commercial or financial relationships that could coscinodiscus wailesii gran and angst. I. valve morphology and development be construed as a potential conflict of interest. of its architecture. J. Phycol. 19, 387–402. doi: 10.1111/j.0022-3646.1983. 00387.x Copyright © 2016 Goessling, Cartaxana and Kühl. This is an open-access article Schnitzler Parker, M., Armbrust, E. V., Piovia-Scott, J., and Keil, R. G. (2004). distributed under the terms of the Creative Commons Attribution License (CC BY). Induction of photorespiration by light in the centric diatom Thalassiosira The use, distribution or reproduction in other forums is permitted, provided the weissflogii (Bacillariophyceae): molecular characterization and physiological original author(s) or licensor are credited and that the original publication in this consequences. J. Phycol. 40, 557–567. doi: 10.1111/j.1529-8817.2004. journal is cited, in accordance with accepted academic practice. No use, distribution 03184.x or reproduction is permitted which does not comply with these terms.

Frontiers in Marine Science | www.frontiersin.org 10 January 2016 | Volume 2 | Article 115

126

6.4. Light and O 2 microenvironments in two contrasting diatom- dominated coastal sediments

Paulo Cartaxana 1,* , Lourenço Ribeiro 2,3 , Johannes W. Goessling 1, Sónia Cruz 1,4 and Michael Kühl 1,5

1Marine Biological Section, Department of Biology, University of Copenhagen, Strandpromenaden 5, 3000 Helsingør, Denmark

2Université de Nantes, Mer Molécules Santé EA2160, Faculté des Sciences et des Techniques, 44322 Nantes Cedex 3, France

3Centro de Ciências do Mar e Ambiente (MARE), Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal

4Departamento de Biologia & CESAM − Centro de Estudos do Ambiente e do Mar, Universidade de Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal

5Plant Functional Biology and Climate Change Cluster, University of Technology Sydney, Sydney, NSW 2007, Australia

*Corresponding author: Paulo Cartaxana; [email protected]; Campus de Santiago, 3810-193 Aveiro, Portugal

127

128 Vol. 545: 35–47, 2016 MARINE ECOLOGY PROGRESS SERIES Published March 8 doi: 10.3354/meps11630 Mar Ecol Prog Ser

Light and O2 microenvironments in two contrasting diatom-dominated coastal sediments

Paulo Cartaxana1,*, Lourenço Ribeiro2,3, Johannes W. Goessling1, Sónia Cruz1,4, Michael Kühl1,5

1Marine Biological Section, Department of Biology, University of Copenhagen, Strandpromenaden 5, 3000 Helsingør, Denmark 2Université de Nantes, Mer Molécules Santé EA2160, Faculté des Sciences et des Techniques, 44322 Nantes Cedex 3, France 3Centro de Ciências do Mar e Ambiente (MARE), Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal 4Departamento de Biologia & CESAM − Centro de Estudos do Ambiente e do Mar, Universidade de Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal 5Plant Functional Biology and Climate Change Cluster, University of Technology Sydney, Sydney, NSW 2007, Australia

ABSTRACT: The close coupling of photosynthesis and light was studied in 2 contrasting diatom- dominated coastal sediments (sand and mud flats) using O2 microelectrodes and fiber-optic micro- probes for scalar irradiance. The diatom community of the muddy sediment was composed almost exclusively of motile epipelic species, whereas in the sandy sediment similar contributions of epi - psammic and epipelic diatoms were observed. The attenuation coefficient of scalar irradiance (K0) was significantly higher in the mud, where light was attenuated exponentially with depth from the sediment surface. In the sand, scalar irradiance levels increased in the first 0.1–0.2 mm due to high scattering and low absorption. Attenuation of scalar irradiance was highest for wavelengths of absorption by major diatom photopigments (chlorophylls and carotenoids). Higher areal and volumetric rates of O2 respiration were found in illuminated sediments than those in the dark, resulting from an increase in both O2 concentration and penetration depth and a direct stimulation of heterotrophic processes. A lower light acclimation index (Ek) was observed for the muddy sediment community, indicating lower light acclimation compared to the sandy sediment commu- nity. Areal and volumetric rates of photosynthesis were ~3 times higher in the muddy sediment. We conclude that higher photosynthetic rates in the finer sediment were determined by (1) a thin- ner and more densely populated photic zone, where the contribution of active photopigments to total light absorption relative to that of photosynthetically inactive components was higher, and (2) differences in diatom species composition and dominant life-strategies, specifically the capacity of cells to actively search for optimal light microenvironments in the fine-grained sediment.

KEY WORDS: Microsensors · Diatoms · Migration · Photosynthesis · Respiration · Scalar irradiance

Resale or republication not permitted without written consent of the publisher

INTRODUCTION mary producers in coastal ecosystems (MacIntyre et al. 1996), playing a significant role as (1) facilitators Microphytobenthic communities inhabiting the of carbon transfer among trophic levels (Bellinger et intertidal and shallow subtidal mud and sand flats of al. 2009), (2) mediators of nutrient cycling and ex - estuaries and coastal zones are largely dominated by change across the sediment−water interface (Sund- diatoms and/or cyanobacteria. These organisms have bäck et al. 2000) and (3) efficient sediment stabilizers been identified as some of the most important pri- (Underwood & Paterson 2003).

*Corresponding author: [email protected] © Inter-Research 2016 · www.int-res.com

129 36 Mar Ecol Prog Ser 545: 35–47, 2016

The microenvironment of these densely populated are attached to sand grains, and therefore are found microphytobenthic communities is complex and more commonly in sandy substrata. As this nomen- characterized by steep physical and chemical gradi- clature represents an oversimplification, a more re - ents, including very strong light attenuation. Micro - fined classification has recently been proposed, un - environmental heterogeneity in benthic communities coupling motility or attachment from sediment type is more pronounced than in planktonic systems, at a (e.g. Ribeiro et al. 2013, Barnett et al. 2015). This in- spatial scale comparable to the size and distance cludes differentiation within the epipsammic group between the individual organisms (Underwood & between non-motile species, firmly attached (either Kromkamp 1999). Therefore, studies of microphyto- stalked or adnate) to sand particles, and motile forms benthic photosynthesis require specialized tools to that can move within the sphere of individual sand assess the distribution of light and photosynthetic grains. Motile epipelic diatoms may actively search activity at relevant spatial scales. The use of fiber- for optimal light microenvironments, exhibiting verti- optic scalar irradiance microprobes in combination cal migratory rhythms determined by diurnal and with O2 microsensors is an ideal way to resolve how tidal cycles (Round & Palmer 1966, Pinckney & Zing- the steep gradients of light intensity and spectral mark 1991, Underwood et al. 2005) and irradiance composition in sediments affect microbenthic photo- levels (Kromkamp et al. 1998, Perkins et al. 2010, synthesis and other aspects of the photobiology of Vieira et al. 2011). Epipsammon-dominated sand flat microphytobenthos in their natural environments communities do not show such migratory patterns (Kühl et al. 1996, Kühl 2005). and photoregulate exclusively through physiological Microphytobenthos colonizing sediments live in a mechanisms (Jesus et al. 2009, Cartaxana et al. 2011). diffuse light field with a strong component of scat- Although several studies have characterized light tered light that is harvested from all directions (Kühl and/or O2 microenvironments in benthic systems & Jørgensen 1992, 1994); thus, measurements of dominated by cyanobacteria and/or diatoms (e.g. downwelling irradiance can significantly underesti- Lassen et al. 1992, Ploug et al. 1993, Kühl et al. 1996, mate the light availability for photosynthesis in such Glud et al. 2002, Brotas et al. 2003, Hancke & Glud communities (Kühl et al. 1994). It is therefore of para- 2004, Al-Najjar et al. 2010, 2012), such measure- mount importance to relate photosynthesis at a given ments have never been done together in fine muddy point to scalar irradiance, i.e. the integral of radiance sediments with a typically migrating diatom commu- incident from all directions around a point in space nity in order to assess the close coupling of photo -

(Lassen et al. 1992, Kühl & Jørgensen 1994). Light in synthesis and light. In this study, O2 microelectrodes sedimentary environments is subject to intense ab - and fiber-optic microprobes for scalar irradiance sorption but also to a high degree of scattering due to were applied to resolve the vertical variability of O2, the high density of microalgae, detritus and sediment light and photosynthesis in 2 contrasting diatom- particles. This has several important consequences dominated coastal sediments: highly cohesive fine that may impact microphytobenthic photosynthesis, mud and coarse sandy sediment. We hypothesized namely (1) strong light attenuation resulting in an that differences in photosynthetic rates in diatom- extremely narrow photic zone, (2) local enhancement dominated sediments with distinct particle size com- of scalar irradiance at the sediment surface due to position would be determined by light absorption intense scattering, and (3) rapid change in light qual- patterns and dominant cell life-strategies. Oxygen ity with depth resulting from differential absorption: budgets and a detailed description of the diatom a decrease of the wavelengths coinciding with photo - taxo nomic composition of the 2 studied benthic com- pigment absorption maxima compared to other wave - munities are presented. lengths in the light spectrum (Lassen et al. 1992, Ploug et al. 1993, Kühl & Jørgensen 1994, Kühl et al. 1994, 1996). MATERIALS AND METHODS Benthic diatom communities have traditionally been divided into 2 main groups with regard to the Sediment sampling and experimental set-up life-strategies they exhibit: the epipelon and the epip- sammon (e.g. Round 1965, Admiraal 1984). The Sediment samples (2 cm depth) were collected epipelon consists of motile diatoms that can move from an estuarine mud flat in Lisbon, Portugal freely between sediment particles and usually domi- (38°47’46.7’’N, 09° 05’ 32.4’’W) and a sandy beach in nate in muddy sediments, whereas the epipsammon Helsingør, Denmark (56° 02’48.5’’ N, 12° 36’ 07.3’’ E) mostly comprises smaller, non-motile diatoms that using a rectangular acrylic corer (19.7 × 7.8 × 2 cm),

130 Cartaxana et al.: Light and O2 in diatom sediments 37

and were transferred with minimal disturbance to Table 1. Definitions of abbreviations custom-made acrylic flow-through chambers. In the laboratory, a stable laminar flow of aerated seawater Symbol Definition was maintained above the sediment surface (water α layer depth: 1.5 cm; temperature: 16°C; salinity: 30) Initial slope of the photosynthesis vs. irradiance curve using a submersible water pump (Rena Flow 400) Φ Sediment porosity connected to the flow chamber and immersed in a DBL Diffusive boundary layer 2 −1 temperature-controlled water reservoir. D0 Molecular diffusion coefficient of O2 (cm s ) 2 −1 Sediments were first exposed to an incident down- Ds Sediment diffusion coefficient of O2 (cm s ) −2 −1 E0 Photon scalar irradiance (µmol photons m s ) welling photon irradiance (Ed; Table 1 provides a list Ed Downwelling photon irradiance (µmol photons of abbreviations) of photosynthetically active radiation m−2 s−1) −2 −1 −2 −1 (PAR, 400–700 nm) of 75 µmol photons m s for ap- Ek Light acclimation index (µmol photons m s ) −2 −1 proximately 24 h. The sediments were illuminated Ec Compensation irradiance (µmol photons m s ) vertically from above with a fiber-optic tungsten halo- Jup Upward O2 flux through the DBL (= P , nmol O cm−2 s−1) gen lamp (KL-2500, Schott) equipped with a collimat- n 2 Jdown Downward O2 flux from the photic zone −2 −1 ing lens, and Ed was measured with a calibrated pho- (nmol O2 cm s ) ton irradiance meter (ULM-500, Walz) equipped with K0 Attenuation coefficient of scalar irradiance −1 a planar cosine collector (LI-190, LiCor). Sediments (mm ) PAR Photosynthetically active radiation were exposed at each experimental photon irradiance (400–700 nm, µmol photons m−2 s−1) −2 −1 (0, 250, 500 and 1000 µmol photons m s ) for at least Pz Volumetric gross photosynthesis at depth z −3 −1 45 min prior to O2 and scalar photon irradiance meas- (nmol O2 cm s ) −2 −1 urements. Changes in photon irradiance were Pg Gross areal photosynthesis (nmol O2 cm s ) Pn Net areal photosynthesis achieved without spectral distortion by adjusting the −2 −1 (= Jup, nmol O2 cm s ) aperture size on the fiber-optic halogen lamp. Pn,phot Net areal photosynthesis of the photic zone −2 −1 Sediment grain size composition was determined (= |Jup| + |Jdown|, nmol O2 cm s ) by sequential sediment sieving, evaluating the Pmax Maximal photosynthetic capacity −2 −1 relative contribution (% dry weight) of the size frac- (nmol O2 cm s ) Rlight Areal respiration in the light tions: >1000, 1000−500, 500−250, 250−125, 125−63 −2 −1 (= Pg − Pn, nmol O2 cm s ) −2 −1 and <63 µm. Rdark Areal respiration in the dark (nmol O2 cm s ) Rphot Areal respiration within the photic zone −2 −1 (= Pg − Pn,phot, nmol O2 cm s ) Microphytobenthos biomass and taxonomic composition euglenids and other microphytobenthic algal groups Sediment samples (approximately the upper 2 mm) (Ribeiro 2010). Total cell counts were made directly were collected for biomass estimation and taxonomic from the extract on an Olympus BX50 optical micro- composition analysis by scraping the surface with a scope, at 400× magnification. scalpel. Approximately 100 mg of sediment were ex- Diatom identification and cell counts were made tracted for 24 h with 3 ml of 90% acetone for biomass on permanent slides of cleaned diatom material, estimation. Chlorophyll a (chl a) concentrations were mounted in Naphrax™, on a Zeiss Axioskop 50 micro- determined on pigment extracts by spectrophoto - scope equipped with differential interference con- metry (UV-2101PC, Shimadzu) using the method of trast optical microscopy. The observation of the Jeffrey & Humphrey (1975). Approximately 3 ml of extracts showed that weakly-silicified diatoms, like sediment was collected for taxonomic composition Atheya and Cylindrotheca, were an important part of analysis and placed in 5 ml polypropylene tubes, to the assemblages. Therefore, to avoid the under - which 1 ml of a 2.5% glutaraldehyde solution was estimation of these fragile genera, 2 types of slides added, and stored at 4°C for later processing. Cells were made: one from diatom material incinerated in were extracted from the sediment following an iso - a muffle-furnace for 2 h at 450°C to preserve fragile pycnic separation technique using silica sol Ludox® specimens; and one from diatom material that was HS-40 (Sigma-Aldrich) that separates the organic oxidized by hydrogen peroxide (30%) at 90°C for 4 h, material from mineral particles, and is thus able to re - thus destroying the slightly silicified frustules, but move both migratory and non-migratory fractions of allowing better identification of very small taxa. In the diatom assemblages, as well as cyanobacteria, both cases, >400 frustules/valves were counted.

131 38 Mar Ecol Prog Ser 545: 35–47, 2016

After determining the relative percentage of the Volumetric gross photosynthesis rates (Pz, nmol O2 fragile diatoms (i.e. Cylindrotheca and Atheya spp.), cm−3 s−1) were measured at 0.1 mm depth intervals the relative abundance (percentage) of the other taxa at the higher experimental photon irradiance (400– was determined in the oxidized slide taking into con- 700 nm, 1000 µmol photons m−2 s−1) using the micro- sideration what was previously established for fragile electrode light−dark shift technique (Revsbech & Jør- specimens (Ribeiro 2010). Diatom identification gensen 1983). In this method, the gross rate of photo- mainly followed Ribeiro (2010) and Witkowski et al. synthesis is estimated as the initial O2 depletion rate (2000) and references therein. at a specific depth during the first few seconds after Biovolume assessment followed the estimates pre- light is briefly turned off (see also Glud et al. 1992). sented by Snoeijs et al. (2002) and Ribeiro et al. Depth integration over the sediment photic zone

(2013). In the cases of taxa that were not available in yielded areal gross photosynthesis rates (Pg, nmol O2 those works, biovolume calculations were made fol- cm−2 s−1). lowing Hillebrand et al. (1999). Size-classes and life- strategy information were primarily taken from Ri - beiro (2010) and references herein, but followed the Flux calculations of net photosynthesis growth-form groups proposed by Barnett et al. and respiration (2015): epipelon (EPL), motile epipsammon (EPM-M) and non-motile epipsammon (EPM-NM) (no tycho- Sediment net areal photosynthesis (Pn) was calcu- planktonic species — those that can live both in the lated from steady-state O2 profiles in the light using sediment and in the water column — were observed). the flux of O2 (Jup) into the overlaying water through Diversity of the 2 studied microphytobenthic commu- the diffuse boundary layer (DBL) as calculated by nities was calculated using the Shannon index: Fick’s first law of diffusion (Kühl et al. 1996): S Jup = Pn = −D0 dC/dz (2) Hpp'–= ∑ ii loge (1) i=1 where D0 is the molecular O2 diffusion coefficient at experimental salinity and temperature (from Table

Microprofiling of O2 concentration and gross for seawater and gases, Unisense A/S) and dC/dz is photosynthesis the linear concentration gradient in the DBL. Net photosynthetic rates as a function of the incident

Depth profiles of dissolved O2 concentrations were photon irradiance, Pn(E), were fitted with an expo- measured with fast responding (t90 < 0.5 s) Clark-type nential model (Webb et al. 1974), with an extra term, microelectrodes (tip diameter ~25 µm, OX-25, Uni - R, to account for O2 consumption (Spilling et al. sense A/S) at vertical steps of 0.1 mm. The O2 micro- 2010): electrodes were connected to a pA meter (Unisense −αE/Pmax Pn(E) = Pmax (1 − exp ) + R (3) A/S) and signals were recorded via a USB-interfaced

A/D-converter (DCR16, Pyro Science) through de - where Pmax represents the maximum photosynthetic dicated PC-controlled data acquisition software capacity and α is the initial slope of the photo-

(Profix, Pyro Science). The O2 microsensors were lin- synthesis vs. irradiance curve. The compensation early calibrated at experimental temperature and irradiance (Ec), i.e. the irradiance at which total O2 salinity from measurements in aerated seawater as production and consumption in the sediment bal- well as in seawater made anoxic by the addition of anced each other, was determined when Pn(E) = 0. sodium sulphite. The O2 microsensors were mounted The light ac climation index, Ek, i.e. the irradiance at on a motor-driven micromanipulator (MU1, Pyro Sci- onset of photosynthesis saturation was calculated ence), which was interfaced to a desktop computer from Ek = Pmax/α. and controlled with the Profix software. Surface posi- Sediment areal respiration in the dark (Rdark) was tioning of the microsensors was done while observ- calculated by Eq. (2) as the linear flux of O2 into the ing the sediment with a PC-interfaced USB digital sediment through the DBL. The areal respiration of microscope (AD7013MZT Dino-Lite, AnMo Elec- the sediment in the light (Rlight) was calculated as the tronics). The microsensors were inserted into the difference between gross and net photosynthesis sediment at an angle of 45° relative to the vertical (Pg − Pn). Net photosynthesis and respiration were incident light beam to avoid self-shading. The soft- also calculated for the photic zone alone (Kühl et al. ware automatically accounted for the sensor inser- 1996). The areal net photosynthesis of the photic tion angle; all depths are given in vertical distances. zone (Pn,phot) was calculated as the absolute sum of

132 Cartaxana et al.: Light and O2 in diatom sediments 39

the flux of O2 into the overlaying water across the DBL (Jup) and the flux into the biofilm below the photic zone (Jdown). Downward O2 flux at the lower boundary of the photic zone was calculated as:

Jdown = −Φ Ds dC/dz (4) where Φ is the sediment porosity, and Ds is the sedi- ment diffusion coefficient, which was calculated from the porosity and D0 according to Iversen & Jørgensen (1993). Sediment porosity was calculated from core slicing and subsequent drying of known volumes of sediment. The areal respiration of the photic zone

(Rphot) was then calculated as Pg − Pn,phot. Average volumetric rates of O2 respiration of distinct zones of Fig. 1. Particle size fractions (%) of the muddy (mud) and the sediment were calculated by dividing areal respi- sandy (sand) coastal sediments ration rates determined from the flux calculations with the depth of the photic zone or the O2 penetra- tion depth (Kühl et al. 1996). RESULTS

Sediment particle size and porosity Light measurements The 2 studied coastal sediments differed signifi-

Spectral measurements of scalar irradiance (E0) cantly regarding particle size composition. More were done at 0.1 mm vertical depth intervals using a than 99% of the muddy sediment was composed of scalar irradiance microprobe connected to a fiber- particles <63 µm (Fig. 1), with a porosity of 0.9. In the optic spectrometer (USB2000+, Ocean Optics) inter- sandy sediment, the particle size distribution was faced to a PC running spectral acquisition software dominated by medium sand of 250–500 µm with 56% (Spectra Suite, Ocean Optics). The scalar irradiance dry weight, followed by coarse and fine sand (500− microprobe consisted of a small diffusing sphere 1000 and 125−250 µm size-classes) with 18 and 16% (80 µm diameter) cast on the coated tip of a tapered dry weight, respectively. Small stones and broken optical fiber (Rickelt et al. 2016). The position and in- shells (>1000 µm) were also a significant part (9%) of sertion angle of the scalar irradiance probe was con- the sediment. Particles <125 µm represented less trolled as described above for the O2 microelectrodes. than 1% of the sediment (Fig. 1), with a porosity of Photon scalar irradiance of PAR was obtained by 0.6. integrating scalar irradiance spectra from 400 to 700 nm. The downwelling spectral scalar irradiance

(i.e. Ed) at the sediment surface was measured by Microphytobenthos biomass and community positioning the scalar irradiance microprobe over a composition black non-reflective light well at the same position relative to the vertically incident, collimated light Concentrations of chl a were higher in the muddy field as the sediment surface. Scalar irradiance spec- sediment, but only when expressed on a per dry tra at various depths in the sediment were normal- weight basis: 229 ± 12 and 94 ± 23 µg chl a g−1 for the ized to the known incident downwelling photon mud and sandy sediment, respectively. On a volu- irradiance at the sediment surface. Attenuation co - metric basis, similar chl a concentrations were ob - efficients of photon scalar irradiance were calculated served: 108 ± 6 and 138 ± 34 µg chl a cm−3, for the from the depth profiles of spectral scalar irradiance mud and sandy sediment, respectively. Diatoms dom- as (Kühl 2005): inated the microphytobenthic communities of both studied coastal sediments, with a relative abundance K0 = ln(E1/E2)/(z2 − z1) (5) of 95.6 and 90% of total cells in the muddy and sandy where K0 is the vertical spectral attenuation coeffi- sediment, respectively. In the muddy sediment, cyano - cient of scalar irradiance, and E1 and E2 are the spec- bacteria of the genera Oscillatoria and Merismopedia tral scalar irradiance measured at depths z1 and z2 in and euglenophytes represented only 1.5 and 2.9% of the sediment (where z2 > z1). the total cell number. In the sandy sediment, the

133 40 Mar Ecol Prog Ser 545: 35–47, 2016

Table 2. Diatoms (28 taxa) collected from the muddy coastal sediment site (Lisbon, Portugal), including details of their relative abundance (%), biovolume (µm3) and life-form (EPL: epipelon; EPM-M: motile epipsammon; EPM-NM: non-motile epipsam- mon). Relative abundances were also allocated to 4 size-classes which comprised the average cell biovolumes of <100, 100−250, 250−1000 and >1000 µm3

Taxon Abundance Biovolume Size-class Life-form

Biremis lucens (Hustedt) Sabbe, Witkowski & Vyverman 0.1 439 250−1000 EPM-NM Cylindrotheca closterium (Ehrenberg) Reimann & Lewin 15.7 247 100−250 EPL Cylindrotheca cf. gracilis (Brébisson in Kützing) Grunow 5.5 639 250−1000 EPL Entomoneis paludosa var. paludosa (W. Smith) Reimer in Patrick & Reimer 6.3 2869 >1000 EPL Fallacia subforcipata (Hustedt) Mann 0.1 204 100−250 EPM-M Gyrosigma acuminatum (Kützing) Rabenhorst 2.0 20 790 >1000 EPL Gyrosigma fasciola (Ehrenberg) Griffith & Henfrey 9.4 6614 >1000 EPL Gyrosigma cf. limosum Sterrenburg & Underwood 0.3 7436 >1000 EPL Gyrosigma wansbeckii (Donkin) Cleve 0.2 27104 >1000 EPL Luticola mutica (Kützing) Mann 0.2 425 250−1000 EPL Navicula arenaria Hustedt 0.2 3290 >1000 EPL Navicula cf. biskanterae Hustedt 0.1 507 250−1000 EPL Navicula flagellifera Hustedt 0.3 757 250−1000 EPL Navicula gregaria Donkin 3.2 271 250−1000 EPL Navicula pargemina Underwood & Yallop 0.1 74 <100 EPL Navicula perminuta Grunow in van Heurck 0.3 126 100−250 EPL Navicula phyllepta Kützing 0.3 228 100−250 EPL Navicula cf. phyllepta Kützing 18.6 137 100−250 EPL Navicula spartinetensis Sullivan & Reimer 34.5 299 250−1000 EPL Nitzschia cf. dubia W. Smith 0.2 5529 >1000 EPL Nitzschia cf. distans Gregory 0.2 1343 >1000 EPL Nitzschia sp.1 0.2 82 <100 EPL Petrodictyon gemma (Ehrenberg) Mann in Round, Crawford & Mann 0.5 22764 >1000 EPL Plagiotropis vitrea (W. Smith) Grunow 0.3 6676 >1000 EPL Pleurosigma angulatum sensu W. Smith emend. Sterrenburg 0.3 38 153 >1000 EPL Staurophora salina (W. Smith) Mereschkowsky 0.5 2278 >1000 EPL Surirella sp.1 sensu Ribeiro (2010) 0.3 3529 >1000 EPL Tryblionella gracilis W. Smith 0.1 2880 >1000 EPL same 2 genera of cyanobacteria were present at a rel- volume <250 µm3. Nevertheless, medium sized dia - ative abundance of 8.3%, while green microalgae toms (250 to 1000 µm3) dominated both muddy and and euglenophytes only made minor contributions sandy communities in terms of relative abundance (1.6 and 0.1%, respectively). (in both cases, ca. 45%). Diatom species richness (46) and diversity (H’ = The microphytobenthic communities of the 2 stud- 2.87) were higher in the sand than in the muddy sed- ied sediments were also quite distinct in terms of iment (28 and H’ = 1.99, respectively) (Tables 2 & 3). diatom composition, sharing only 2 species (Cylindro - The sandy sediment community was also more di - theca closterium and Navicula gregaria). The muddy verse in terms of life-strategies, with motile (18 spe- sediment was dominated by species of the genera cies) and non-motile epipsammon (11) coexisting Navicula (57.6% relative abundance), particularly N. with epipelic (17) diatoms (Table 3). In the mud, spartinetensis (34.5%) and N. cf. phyllepta (18.6%) epipelic species were mainly observed (26 out of a (Table 2). In the sand, the most representative spe- total of 28 species) (Table 2). In terms of relative cies were the epipelic Navicula sp.1 (15.6%) and the abundance, the muddy sediment was almost exclu- non-motile epipsammic Anorthoneis vortex (12.4%) sively inhabited by epipelic diatoms (99.8%), while and Attheya decora (11.2%) (Table 3). in the sand, epipelic and epipsammic diatoms exhib- ited a similar contribution (~50%) to the microphyto- benthic community (Tables 2 & 3). In the muddy Scalar irradiance sediment, most of the identified species were large diatoms, with 14 species having a biovolume of Depth profiles of photon scalar irradiance were >1000 µm3. In contrast, most of the identified diatoms rather different between the 2 studied coastal sedi- in the sand were small, with 33 species having a bio- ments (Fig. 2): while scalar irradiance levels at the

134 Cartaxana et al.: Light and O2 in diatom sediments 41

Table 3. Diatoms (46 taxa) collected from the sandy coastal sediment site (Helsingør, Denmark), including details of their relative abundance (%), biovolume (µm3) and life-form (EPL: epipelon; EPM-M: motile epipsammon; EPM-NM: non-motile epipsammon). Relative abundances were also allocated to 4 size-classes which comprised the average cell biovolumes of <100, 100−250, 250−1000 and >1000 µm3

Taxon Abundance Biovolume Size-class Life-form

Biremis lucens (Hustedt) Sabbe, Witkowski & Vyverman 0.1 439 250−1000 EPM-NM Achnanthes cf. amoena Hustedt 0.2 85 <100 EPM-M Amphora subacutiuscula Schoeman 3.9 1194 >1000 EPM-M Amphora cf. tenerrima Aleem & Hustedt 0.2 120 100−250 EPL Amphora wisei (Salah) Simonsen 0.2 110 100−250 EPM-M Anaulus balticus Simonsen 0.4 52 <100 EPM-NM Anorthoneis vortex Sterrenburg 12.4 790 250−1000 EPM-NM Astartiella punctifera (Hustedt) Witkowski & Lange-Bertalot 2.4 468 250−1000 EPM-M Attheya decora T. West 11.2 2363 >1000 EPM-NM Biremis ambigua (Cleve) Mann 0.2 23488 >1000 EPM-NM Cavinula sp. 0.2 104 100−250 EPM-M Cocconeis hauniensis Witkowski 0.7 54 <100 EPM-NM Cocconeis pelta A. Schmidt 0.2 221 100−250 EPM-NM Cocconeis peltoides Hustedt 0.4 175 100−250 EPM-NM Cylindrotheca closterium (Ehrenberg) Reimann & Lewin 0.9 247 100−250 EPL Fallacia cryptolyra (Brockmann) Stickle & Mann 0.2 162 100−250 EPM-M Fragilaria cf. cassubica Witkowski & Lange-Bertalot 0.2 81 <100 EPM-NM Grammatophora oceanica Ehrenberg 0.2 4126 >1000 EPM-NM Halamphora cf. abuensis (Foged) Levkov 0.4 110 100−250 EPL Navicula aleksandrae Lange-Bertalot, Bogaczewicz-Adamczak & Witkowski 1.1 76 <100 EPM-M Navicula biskanterae Hustedt 0.2 98 <100 EPM-M Navicula cf. celinei Witkowski, Metzeltin & Lange-Bertalot 0.2 122 100−250 EPM-M Navicula cf. perminuta Grunow in van Heurck 2.0 68 <100 EPL Navicula diserta Hustedt 5.9 54 <100 EPM-M Navicula germanopolonica Witkowski & Lange-Bertalot 2.6 132 100−250 EPM-M Navicula gregaria Donkin 8.4 831 250−1000 EPL Navicula sp.1 15.6 381 250−1000 EPL Navicula sp.2 0.4 158 100−250 EPL Navicula sp.3 0.2 64 <100 EPL Navicula sp.4 0.4 205 100−250 EPL Navicula sp.5 0.2 338 250−1000 EPL Navicula sp.6 0.9 58 <100 EPL Navicula viminoides Giffen 0.7 96 <100 EPM-M Nitzschia aurariae Cholnoky 0.7 64 <100 EPL Nitzschia cf. distans Gregory 2.8 1343 >1000 EPL Nitzschia dubiiformis Hustedt 0.2 1189 >1000 EPL Nitzschia sp.2 0.7 354 250−1000 EPL Opephora guenter-grassii (Witkowski & Lange-Bertalot) Sabbe & Vyverman 0.2 62 <100 EPM-NM Opephora mutabilis (Grunow) Sabbe & Vyverman 0.2 167 100−250 EPM-NM Parlibellus cf. calvus Witkowski, Metzeltin & Lange-Bertalot 5.5 252 250−1000 EPL Planothidium delicatulum (Kützing) Round & Bukhtiyarova morphotype 1 3.1 140 <100 EPM-M Planothidium delicatulum (Kützing) Round & Bukhtiyarova morphotype 3 0.2 91 100−250 EPM-M Planothidium deperditum (Giffen) Witkowski 0.4 54 <100 EPM-M Planothidium cf. lemmermannii (Hustedt) Morales 0.9 75 <100 EPM-M Planothidium sp.1 0.7 100 100−250 EPM-M Planothidium sp.2 0.2 108 100−250 EPM-M Seminavis cf. strigosa (Hustedt) Danielidis & Economou-Amilli 10.9 171 100−250 EPL

sediment surface were similar in both, reaching 0.1–0.2 mm to 130−150% of incident downwelling ~120% of incident downwelling photon irradiance, irradiance, followed by an exponential light attenua- light was attenuated differently. In the muddy sedi- tion starting from ~0.2 to 0.4 mm depth, depending ment, light was attenuated exponentially with depth on the incident downwelling irradiance (Fig. 2). Light starting from the surface, while photon scalar irradi- attenuation coefficients of PAR were distinctively dif- ance levels in the sand increased over the first ferent between the 2 coastal sediments, with K0 val-

135 42 Mar Ecol Prog Ser 545: 35–47, 2016

with a thicker DBL (0.5–0.6 mm) and

with peak O2 concentrations at 0.5– 1.0 mm below the sediment surface, depending on the incident down- welling photon irradiance. At higher photon irradiance, the upper part of

the sandy sediment reached O2 con- centrations of ~3 times air saturation (Fig. 4).

Pz showed the highest rates of O2 production in the surface of the muddy sediment (Fig. 5), reaching −3 −1 30.3 ± 6.3 nmol O2 cm s under an incident downwelling photon irradi- ance of 1000 µmol photons m−2 s−1. Photosynthesis rates decreased rap- idly with depth, and photosynthetic

O2 production was no longer de- tected below 0.6 mm. In the sandy Fig. 2. Photon scalar irradiance depth microprofiles of photosynthetically sediment, a subsurface maximum of −2 −1 available light (E0, PAR; µmol photons m s ) under different incident gross photosynthesis rate of 8.5 ± downwelling photon irradiances (250, 500 and 1000 µmol photons m−2 s−1) of −3 −1 3.8 nmol O2 cm s was observed the muddy (mud) and sandy (sand) coastal sediments at 0.1 mm under an incident down- ues of 8.65 ± 0.40 and 1.52 ± 0.06 mm−1 for the muddy and sandy sediment, respectively. Spectral attenuation of scalar irradiance increased from the infrared towards the blue part of the spec- trum in both sediments (Fig. 3). Light attenuation for shorter wavelengths (400–50 nm) and around 675 nm (the absorption maxima of carotenoids and chlorophylls) was more pronounced in the muddy sediment. Spectral wavelengths outside the absorp- tion bands of major diatom photopigments, espe- cially in the near infrared region, penetrated deeper into the sediments (Fig. 3).

O2, photosynthesis and respiration

The O2 concentration profiles also differed be - tween the 2 sediment types (Fig. 4). In the muddy sediment, O2 penetration increased with increasing irradiance from about 1.5 mm in the dark to 2.4 and −2 −1 2.7 mm at 250 and 1000 µmol photons m s . The O2 concentration increased towards the muddy sedi- ment surface, defining an approximately 0.3–0.4 mm thick DBL, and O2 concentrations reached a maxi- mum immediately below the sediment surface (0.1– 0.2 mm depth). At higher irradiance, the upper part Fig. 3. Depth profiles of spectral scalar irradiance (E0) nor- malized to the downwelling irradiance at the sediment of the muddy sediment reached O2 concentrations of surface (Ed) of the muddy (mud) and sandy (sand) coastal ~6 times air saturation (Fig. 4). The O2 penetration in sediments. Data are presented on a logarithmic scale and the sandy sediment was deeper than 4 mm (Fig. 4), numbers indicate depth in mm

136 Cartaxana et al.: Light and O2 in diatom sediments 43

for the muddy and sandy sediments,

respectively. Pg rates were higher for the muddy sediment (0.698 nmol −2 −1 O2 cm s ) compared to the sand −2 −1 (0.242 nmol O2 cm s ). Jup and Jdown were higher in the muddy sed- iment, representing a total O2 export −2 −1 of 0.659 nmol O2 cm s , compared −2 −1 to only 0.109 nmol O2 cm s in the sandy sediment (Table 4).

Pn increased with incident photon irradiance for both studied sediments (Fig. 6). For the muddy sediment, net photosynthesis rates were much higher, reaching values of 0.56 nmol −2 −1 O2 cm s , and saturating at higher photon irradiance than in the sandy sediment, where the highest net pho- tosynthesis rates achieved were −1 −2 −1 Fig. 4. Steady-state O2 depth microprofiles (µmol l ; mean ± SD, n = 3) in the 0.10 nmol O2 cm s (Fig. 6). From dark, and under different incident downwelling photon irradiances (250, 500 curve-fitting of Pn vs. photon irradi- and 1000 µmol photons m−2 s−1) of the muddy (mud) and sandy (sand) coastal ance data, Pmax was estimated as 0.63 sediments −2 −1 and 0.17 nmol O2 cm s for the muddy and sandy sediments, respec- welling photon irradiance of 1000 µmol photons tively. The corresponding Ek indexes, estimated from −2 −1 m s , followed by a decreasing photosynthetic O2 the interception of Pmax and α, were 314 and production with depth until 1.3 mm below the sand 966 µmol photons m−2 s−1. The compensation irradi- sediment surface. Under an incident downwelling ances (i.e. Ec; the light at which the sediments −2 −1 photon irradiance of 1000 µmol photons m s , became net O2-producing), were 20 and 60 µmol average volumetric gross photosynthesis rates in photons m−2 s−1 for the muddy and sandy sediments, −3 −1 the photic zone were 10.5 and 3.0 nmol O2 cm s respectively. Rdark was highest for the muddy sediment, whereas Rphot rates were highest in the sandy sediment (Table 4). Comparable areal respiratory rates under illuminated conditions were found in the 2 types of sediment (Table 4). Areal respiration rates measured in the light were considerably higher than in the dark. At a photon irradiance of 1000 µmol photons

Table 4. Depth-integrated photosynthesis, oxygen fluxes −2 −1 and respiration (nmol O2 cm s ) estimated for 1000 µmol photons m−2 s−1 of incident downwelling irradiance in the 2 studied diatom-dominated coastal sediment sites

−2 −1 Areal rates (nmol O2 cm s ) Mud Sand

Photosynthesis (Pg) 0.698 0.242 Upward O2 flux (Jup) 0.562 0.097 Downward O2 flux (Jdown) 0.097 0.012 Total O2 export (Pn,phot) 0.659 0.109 Fig. 5. Volumetric gross photosynthesis depth microprofiles Respiration in the dark (Rdark) 0.038 0.010 −3 −1 (Pz; nmol O2 cm s , mean ± SD, n = 3) measured under an Respiration in the light (Rlight) 0.136 0.145 incident downwelling photon irradiance of 1000 µmol Photic zone respiration (Rphot) 0.039 0.133 −2 −1 photons m s for the muddy (mud) and sandy (sand) Aphotic zone respiration (Raphot) 0.097 0.012 coastal sediments

137 44 Mar Ecol Prog Ser 545: 35–47, 2016

intensity (i.e. the scalar irradiance) available for microbenthic phototrophs. Kühl et al. (1994) showed that measurements of downwelling irradiance can underestimate light availability by >100%. Although this represents an extreme value measured in highly scattering sand with little absorption, scalar irradi- ance in the visible part of the spectrum typically reaches about 110 to 160% of the incident irradiance at the surface of most photosynthetic biofilms and sediments (Kühl et al. 1997). Scalar irradiance was attenuated exponentially with depth starting from the surface in the fine muddy sediment. The scalar irradiance attenuation coefficient reported here for the muddy sediment −2 −1 Fig. 6. Areal rates of net photosynthesis (Pn; nmol O2 cm (8.65 ± 0.40 mm ) is among the highest described for s−1, mean ± SD, n = 3) in the dark, and under different benthic systems, comparable only to other highly incident downwelling photon irradiances (250, 500 and cohesive, fine intertidal sediments (Cartaxana et al. 1000 µmol photons m−2 s−1) for the muddy (mud) and sandy (sand) coastal sediments. Solid lines represent curve fits to 2011) and microbial mats (Kühl et al. 1997). In both the data: R2 (mud) = 0.998; R2 (sand) = 0.999 sediments, strongest light attenuation was observed for wavelengths around 675 nm and in the 400–550 nm range, corresponding to absorption maxima of diatom m−2 s−1, areal respiration rates were 3.6 and 14.5 photopigments (chlorophylls and carotenoids), espe- times higher than the areal dark respiration in mud - cially in the muddy sediment. Such differential light dy and sandy sediments, respectively. In the mud, absorption led to a decrease with depth of the wave- −2 areal respiration in the dark was 0.038 nmol O2 cm lengths more relevant for photosynthesis, thus limit- −1 s , whereas O2 penetrated down to 1.7 mm. There- ing photosynthetic activity in deeper sediment lay- fore, the average volumetric O2 respiration in the ers. Ploug et al. (1993) related these changes in light −3 −1 dark was 0.038/0.17 = 0.22 nmol O2 cm s . Ac - quality to the vertical zonation of a population of pen- cordingly, considering an O2 penetration of 4 mm, nate diatoms over a dense underlying population of average volumetric O2 respiration in the dark was filamentous cyanobacteria that largely sustained −3 −1 0.03 nmol O2 cm s in the sandy sediment. Areal O2 their oxygenic photosynthesis via phycobiliproteins respiration rates within the photic zone were 0.039 with absorption characteristics complementary to −2 −1 and 0.133 nmol O2 cm s for mud and sand, respec- chlorophylls. At an incident downwelling photon tively (Table 4). Considering the thicknesses of the irradiance of 1000 µmol photons m−2 s−1, photon photic zones (0.6 and 1.3 mm for mud and sand, res- scalar irradiance values of about 6 and 200 µmol pho- −2 −1 pectively), average volumetric O2 respiration within tons m s were measured at the lower boundary of the corresponding productive zones were 0.65 and the photic zones in the muddy and sandy sediments, −3 −1 1.02 nmol O2 cm s . respectively. The relatively high light availability at the lower boundary of the sandy sediment photic zone may indicate that parameters other than light DISCUSSION intensity may have limited photo synthesis. Accord- ing to Kühl et al. (1996), the lower boundary of the Photon scalar irradiance at the surface of both photic zone in cyanobacterial mats could result from coastal sediment types was higher (about 120%) than reduced light quality, nutrient limitation or the pres- the incident downwelling irradiance due to intense ence of toxic compounds such as sulfide. light scattering in the uppermost sediment layers. In Significantly higher areal rates of O2 consumption the sandy sediment, due to low light absorption and were found in illuminated sediments than in the high scattering at the uppermost layers, a subsurface dark. This could be the result of the observed gradual scalar irradiance maximum was observed reaching increase of both O2 concentration and penetration 150% of incident downwelling irradiance. This sug- depth with irradiance as well as enhanced volumetric gests that measurements of downwelling irradiance O2 respiration in and below the photic zone, resulting (the most commonly measured light parameter in from a direct stimulation of heterotrophic processes benthic studies) can underestimate the total light (Kühl et al. 1996). A similar pattern of light-enhanced

138 Cartaxana et al.: Light and O2 in diatom sediments 45

respiration has been observed in cyanobacterial mats tion compared to epipsammic-dominated sand flat (Kühl et al. 1996, Epping & Kühl 2000), although the communities (Jesus et al. 2009, Pniewski et al. 2015). volumetric rates of respiration reported were 3 to 4 Both studied coastal sediments were dominated by times higher than in both sediments studied here, diatoms exhibiting relative abundances >90%. How- indicating a lower heterotrophic activity in diatom- ever, the species composition was very distinct in that dominated communities. the 2 different microphytobenthic communities only Areal and volumetric rates of photosynthesis were shared 2 common diatom species. Sediment particle ~3 times higher in the muddy sediment, although the composition, defined largely by hydrodynamic forces microphytobenthos biomass (chl a per volume on that control local sediment deposition and erosion, 2 mm depth samples) was similar in mud and sand. has been shown to be the most important parameter Hence, differences in the distribution of the biomass determining diatom species composition in coastal along the sediment profile and a thinner euphotic sediments (Sabbe & Vyverman 1991, Hamels et al. zone may explain the higher productivity of the 1998, Paterson & Hagerthey 2001, Ribeiro et al. muddy sediment. Microphytobenthos in muddy sedi- 2013). Our results confirm previous reported patterns ments has been shown to be highly stratified in the indicating that larger epipelic diatom and smaller mud, with most of the chl a occurring in the top epipsammic diatom species are more abundant in 0.5 mm, while in sandy sediments relatively constant muddy and sandy sediments, respectively. However, concentrations were found throughout the sediment half of the sand-dwelling diatoms were part of the profile down to a depth of several mm (Cartaxana et epipelon, and medium size diatoms (250 to 1000 µm3) al. 2006). In relatively flat benthic systems, photo - dominated both muddy and sandy communities in synthetic efficiency has been shown to be higher in terms of relative abundance. Diatom species richness sediments and biofilms with a thinner and more and diversity were higher in the sandy sediment with densely populated photic zone (Al-Najjar et al. 2012), the co-occurrence of motile epipelic and various as the one found in the muddy sediment we studied. forms of non-motile (i.e. adnate, stalked) and motile The rationale is that the contribution of active pho- epipsammon. This could be related to the more topigments to total light absorption relative to that of diverse particle size composition of the sandy sedi- photosynthetic inactive components increases with ment and the sand-grain microtopography itself, augmented biomass and a thinner photic zone. On allowing higher niche differentiation compared to the other hand, light attenuation by sand grains and the more homo geneous muddy sediment (Ribeiro et mineral precipitates was probably highly relevant in al. 2013). the sandy sediment, where the microphytobenthic Although both diatom communities were domi- cells were more dispersed and the photic zone was nated by motile species in terms of relative abun- wider. Hence, maximum volumetric photosynthesis dance, the importance of motility for the photo - has been shown to correlate positively with light at - physio logy of the muddy sediment community was tenuation in sediments/biofilms and in other aquatic much more prevalent. The light gradient in the mud plant communities (Al-Najjar et al. 2012, Krause- was much steeper, and typical free-moving epipelic Jensen & Sand-Jensen 1998). Our results contrast diatoms would experience significant changes in with those reported by Billerbeck et al. (2007), who photon irradiance simply by covering distances com- showed higher photosynthetic rates in coarse and parable to their cell size. On the other hand, the light fine sands than in a muddy sediment. environment in the sandy sediment was much more Volumetric and areal rates of photosynthesis in the homogeneous, requiring cells to travel much longer diatom-dominated muddy sediment were similar to distances to actively search for a different light maximum in situ rates measured in another mud flat micro environment. Furthermore, movement in mo - of the Tagus estuary (Brotas et al. 2003). Maximum tile epipsammic species is highly limited, i.e. re - photosynthetic rates at saturating photon irradiances stricted to the sphere of individual sand grains or were higher in the muddy sediment than in cyano- very slow movement from sand grain to sand grain bacterial mats or cyanobacteria−diatom mixed bio- (Round 1979). Finally, the higher hydrodynamic films (Kühl et al. 1996, Al-Najjar et al. 2010, 2012), forcing in the sandy sediment causes a continuous but within the range compiled by Krause-Jensen & re arrangement and sorting of the sand grains that Sand-Jensen (1998) for microalgal mats. A lower distributes the epipsammon more evenly, whilst level of Ek was observed for the muddy sediment, destroying or disrupting large accumulations of epi- corroborating the results of previous studies that mud pelic diatoms at certain depths (Delgado et al. 1991, flat diatom communities exhibit lower light acclima- Paterson & Hagerthey 2001). Hence, the higher areal

139 46 Mar Ecol Prog Ser 545: 35–47, 2016

and volumetric rates of photosynthesis and higher Billerbeck M, Røy H, Bosselmann K, Huettel M (2007) Ben- maximum photosynthetic capacity observed in the thic photosynthesis in submerged Wadden Sea intertidal flats. Estuar Coast Shelf Sci 71: 704−716 muddy sediment could also be related to the capacity Brotas V, Risgaard-Petersen N, Serôdio J, Ottosen L, Dals- of epipelic diatoms in this sediment to actively search gaard T, Ribeiro L (2003) In situ measurements of photo- for optimal light microenvironments. Kromkamp et synthetic activity and respiration of intertidal benthic al. (1998) proposed a ‘micromigration’ mechanism, in microalgal communities undergoing vertical migration. Ophelia 57:13−26 which cells would be continuously replaced by others Cartaxana P, Mendes CR, van Leeuwe MA, Brotas V (2006) at the sediment surface, increasing light-use effi- Comparative study on microphytobenthic pigments of ciency and preventing photoinhibition and CO2 limi- muddy and sandy intertidal sediments of the Tagus estu- tation. Micromigration could thus represent a rapid, ary. Estuar Coast Shelf Sci 66: 225−230 flexible and energetically cheap way for benthic Cartaxana P, Ruivo M, Hubas C, Davidson I, Serôdio J, Jesus B (2011) Physiological versus behavioral photoprotection diatoms to optimize productivity in fine grained sedi- in intertidal epipelic and epipsammic benthic diatom ments (Serôdio et al. 2001). communities. J Exp Mar Biol Ecol 405: 120−127 This work describes the major differences in light Delgado M, de Jonge VN, Peletier H (1991) Effect of sand movement on the growth of benthic diatoms. J Exp Mar and O2 microenvironments of 2 diatom-dominated Biol Ecol 145:221−231 microphytobenthic communities inhabiting distinct Epping E, Kühl M (2000) The responses of photosynthesis substrata. We conclude that the higher photosyn- and oxygen consumption to short-term changes in tem- thetic rates observed in the fine muddy sediment perature and irradiance in a cyanobacterial mat (Ebro were related to (1) the compaction of the productive Delta, Spain). Environ Microbiol 2:465−474 biomass into a thin photic zone which reduces the Glud RN, Ramsing NB, Revsbech NP (1992) Photosynthesis and photosynthesis-coupled respiration in natural bio- absorption of light by inactive sediment components, films quantified with oxygen microsensors. J Phycol 28: and (2) differences in diatom species composition 51−60 and dominant life-strategies, specifically the capacity Glud RN, Kühl M, Wenzhöfer F, Rysgaard S (2002) Benthic of diatom cells to change their light microenviron- diatoms of a high Artic fjord (Young Sound, NE Green- land): importance for ecosystem primary production. Mar ment by moving within a steep irradiance gradient. Ecol Prog Ser 238: 15−29 Hamels I, Sabbe K, Muylaert K, Barranguet C, Lucas C, Acknowledgements. This study was supported by a grant Herman P, Vyverman W (1998) Organisation of micro- from the Danish Council for Independent Research | Natural benthic communities in intertidal estuarine flats, a case Sciences (M.K.). S.C. was supported by Fundação para a study from the Molenplaat (Westerschelde estuary, The Ciência e a Tecnologia (IF/00899/2014). We thank Sofie Netherlands). Eur J Protistol 34: 308−320 Jakobsen for technical assistance, Lars Rickelt for manufac- Hancke K, Glud RN (2004) Temperature effects on respira- turing scalar irradiance microsensors, Mads Lichtenberg for tion and photosynthesis in three diatom-dominated ben- help with the microprofiling set-up, and Carla Gameiro with thic communities. Aquat Microb Ecol 37: 265−281 sediment sampling in the Tagus estuary. We thank 2 anony- Hillebrand C, Dürselen CD, Kirschtel D, Pollingher U, Zo ha - mous reviewers for their critical comments on earlier drafts ry T (1999) Biovolume calculation for pelagic and benthic of the manuscript. microalgae. J Phycol 35: 403−424 Iversen N, Jørgensen BB (1993) Diffusion coefficients of sul- fate and methane in marine sediments: influence of LITERATURE CITED porosity. Geochim Cosmochim Acta 57:571−578 Jeffrey SW, Humphrey GF (1975) New spectrophotometric Admiraal W (1984) The ecology of estuarine sediment- equations for determining chlorophylls a, b, c1 and c2 in inhabiting diatoms. In: Chapman DJ, Round FE (eds) higher plants, algae and natural phytoplankton. Biochem Progress in phycological research. Biopress, Bristol, Physiol Pflanz 167:191−194 p 269−322 Jesus B, Brotas V, Ribeiro L, Mendes CR, Cartaxana P, Al-Najjar MA, de Beer D, Jørgensen BB, Kühl M, Polerecky Paterson DM (2009) Adaptations of microphytobenthos L (2010) Conversion and conservation of light energy as semblages to sediment type and tidal position. Cont in a photosynthetic microbial mat ecosystem. ISME J 4: Shelf Res 29:1624−1634 440−449 Krause-Jensen D, Sand-Jensen K (1998) Light attenuation Al-Najjar MA, de Beer D, Kühl M, Polerecky L (2012) Light and photosynthesis of aquatic plant communities. Limnol utilization efficiency in photosynthetic microbial mats. Oceanogr 43:396−407 Environ Microbiol 14:982−992 Kromkamp J, Barranguet C, Peene J (1998) Determination Barnett A, Méléder M, Blommaert L, Lepetit B and others of microphytobenthos PSII quantum efficiency and (2015) Growth form defines physiological photoprotec- photo synthetic activity by means of variable chlorophyll tive capacity in intertidal benthic diatoms. ISME J 9: fluorescence. Mar Ecol Prog Ser 162: 45−55 32−45 Kühl M (2005) Optical microsensors for analysis of microbial Bellinger BJ, Underwood GJC, Ziegler SE, Gretz MR (2009) communities. Methods Enzymol 397: 166−199 Significance of diatom derived polymers in carbon flow Kühl M, Jørgensen BB (1992) Spectral light measurements dynamics within estuarine biofilms determined through in microbenthic phototrophic communities with a fiber- isotopic enrichment. Aquat Microb Ecol 55: 169−187 optic microprobe coupled to a sensitive diode array

140 Cartaxana et al.: Light and O2 in diatom sediments 47

detector. Limnol Oceanogr 37: 1813−1823 trasting shores: a case study from the Tagus estuary. ➤ Kühl M, Jørgensen BB (1994) The light-field of microbenthic J Phycol 49: 258−270 communities: radiance distribution and microscale op - ➤ Rickelt LF, Lichtenberg M, Trampe E, Kühl M (2016) Fiber- tics of sandy coastal sediments. Limnol Oceanogr 39: optic probes for small scale measurements of scalar irra- 1368−1398 diance. Photochem Photobiol, doi: 10.1111/php.12560 ➤ Kühl M, Lassen C, Jørgensen BB (1994) Light penetration ➤ Round FE (1965) The epipsammon; a relatively unknown and light intensity in sandy marine sediments measured freshwater algal association. Brit Phycol Bull 2:456−462 with irradiance and scalar irradiance fiber-optic micro- ➤ Round FE (1979) A diatom assemblage living below the sur- probes. Mar Ecol Prog Ser 105: 139−148 face of intertidal sand flats. Mar Biol 54: 219−223 ➤ Kühl M, Glud RN, Ploug H, Ramsing NB (1996) Micro - ➤ Round FE, Palmer JD (1966) Persistent, vertical-migration environmental control of photosynthesis and photo - rhythms in benthic microflora. II. Field and laboratory synthesis-coupled respiration in an epilithic cyano - studies on diatoms from the banks of the River Avon. bacterial biofilm. J Phycol 32: 799−812 J Mar Biol Assoc UK 46: 191−214 ➤ Kühl M, Lassen C, Revsbech NP (1997) Simple light meter Sabbe K, Vyverman W (1991) Distribution of benthic diatom for measurements of PAR (400 to 700 nm) with fiber-optic assemblages in the Westerschelde (Zeeland, The Nether- microprobes: application for P vs E0(PAR) measurements lands). Belg J Bot 124: 91−101 in a microbial mat. Aquat Microb Ecol 13: 197−207 ➤ Serôdio J, Marques da Silva J, Catarino F (2001) Use of in ➤ Lassen C, Ploug H, Jørgensen BB (1992) Microalgal photo- vivo chlorophyll a fluorescence to quantify short-term synthesis and spectral scalar irradiance in coastal marine variations in the productive biomass of intertidal micro- sediments of Limfjorden, Denmark. Limnol Oceanogr 37: phytobenthos. Mar Ecol Prog Ser 218:45−61 760−772 ➤ Snoeijs P, Busse S, Potapova M (2002) The importance of ➤ MacIntyre HL, Geider RJ, Miller DC (1996) Microphyto - diatom cell size in community analysis. J Phycol 38: benthos: the ecological role of the ‘secret garden’ of un - 265−272 vegetated, shallow-water marine habitats. 1. Distribu- ➤ Spilling K, Titelman J, Greve TM, Kühl M (2010) Micro - tion, abundance and primary production. Estuaries 19: sensor measurements of the external and internal micro- 186−201 environments of Fucus vesiculosus (Phaeophyceae). Paterson DM, Hagerthey SE (2001) Microphytobenthos in J Phycol 46: 1350−1355 contrasting coastal ecosystems: biology and dynamics. ➤ Sundbäck K, Miles A, Goransson E (2000) Nitrogen fluxes, In: Reise K (ed) Ecological comparisons of sedimentary denitrification and the role of microphytobenthos in shores. Springer-Verlag, Berlin, p 105−25 microtidal shallow-water sediments: an annual study. ➤ Perkins RG, Lavaud J, Serôdio J, Mouget JL and others Mar Ecol Prog Ser 200: 59−76 (2010) Vertical cell movement is a primary response of ➤ Underwood GJC, Kromkamp J (1999) Primary production intertidal benthic biofilms to increasing light dose. Mar by phytoplankton and microphytobenthos in estuaries. Ecol Prog Ser 416: 93−103 Adv Ecol Res 29: 93−153 ➤ Pinckney J, Zingmark RG (1991) Effects of tidal stage and ➤ Underwood GJC, Paterson DM (2003) The importance of sun angles on intertidal benthic microalgal productivity. extracellular carbohydrate production by marine epi - Mar Ecol Prog Ser 76: 81−89 pelic diatoms. Adv Bot Res 40:183−240 ➤ Ploug H, Lassen C, Jørgensen BB (1993) Action spectra of ➤ Underwood GJC, Perkins RG, Consalvey MC, Halon ARM, microalgal photosynthesis and depth distribution of Oxborough K, Baker NR, Paterson DM (2005) Patterns in spectral scalar irradiance in a coastal marine sediment of microphytobenthic primary productivity: species-spe- Limfjorden, Denmark. FEMS Microbiol Ecol 12: 69−78 cific variation in migratory rhythms and photosynthetic ➤ Pniewski FF, Biskup P, Bubak I, Richard P, Latala A, Blan- efficiency in mixed-species biofilms. Limnol Oceanogr chard G (2015) Photoregulation in microphytobenthos 50: 755−767 from intertidal mudflats and non-tidal coastal shallows. ➤ Vieira S, Utkin A, Lavrov A, Santos NM, Vilar R, Marques da Estuar Coast Shelf Sci 152: 153−161 Silva J, Cartaxana P (2011) Effects of intertidal micro- ➤ Revsbech NP, Jørgensen BB (1983) Photosynthesis of ben- phytobenthos migration on biomass determination via thic microflora measured with high spatial resolution by laser-induced fluorescence. Mar Ecol Prog Ser 432: 45−52 the oxygen microprofile method: capabilities and limita- ➤ Webb WL, Newton M, Starr D (1974) Carbon-dioxide tions of the method. Limnol Oceanogr 28: 749−756 exchange of Alnus rubra: a mathematical model. Oeco- Ribeiro L (2010) Intertidal benthic diatoms of the Tagus estu- logia 17: 281−291 ary: taxonomic composition and spatial-temporal varia- Witkowski A, Lange-Bertalot H, Metzeltin D (2000) Diatom tion. PhD dissertation, University of Lisbon flora of marine coasts, Vol 1. In: Lange-Bertalot H (ed) ➤ Ribeiro L, Brotas V, Rincé Y, Jesus BM (2013) Structure and Iconographia Diatomologica: annotated diatom micro- diversity of intertidal benthic diatom assemblages in con- graphs. Koeltz Scientific Books, Königstein

Editorial responsibility: Ronald Kiene, Submitted: November 4, 2015; Accepted: January 22, 2016 Mobile, Alabama, USA Proofs received from author(s): February 22, 2016

141

142

6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms (in prep.)

Johannes W. Goessling 1*, Silja Frankenbach 2, Lourenço Ribeiro 3;4 , João Serôdio 2 and Michael Kühl 1;5

1Marine Biological Section, Department of Biology, University of Copenhagen, Strandpromenaden 5, 3000 Helsingør, Denmark

2Centro de Estudos do Ambiente e do Mar (CESAM), Departamento de Biologia, Universidade de Aveiro, 3810-193 Aveiro,

Portugal

3Centro de Ciências do Mar e Ambiente (MARE), Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749-016

Lisboa, Portugal

4Université de Nantes, Mer-Molécules-Santé EA 2160, Faculté des Sciences et des Techniques, 2, Rue de la Houssinière, BP 92 208,

44322 Nantes, France

5Plant Functional Biology and Climate Change Cluster, University of Technology Sydney, Australia

*Corresponding author: Johannes W. Goessling; [email protected]; Strandpromenaden 5, 3000 Helsingør, Denmark

143

144 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Abstract

Differences in the spectral light absorption of photopigments allow for niche differentiation and coexistence of cyanobacteria and diatoms in light exposed habitats. Evolution in the microphytobenthos could have further been shaped by the photonic crystal-like structures in the frustule of diatoms, interacting with electromagnetic radiation in the visible spectrum of sunlight. We extracted motile (raphid) diatoms from an intertidal estuarine mudflat and studied their frustule valve optical properties in water. We observed forward scattering of blue light in the frustule valve, while the angle of incident white light at which this phenomenon was visible differed between the three tested raphid species. A slurry of oxidized diatom frustules from the sediment enhanced scalar irradiance in the blue spectral range by up to 120% compared to downwelling irradiance. Blue light induced higher relative electron transport rates and non-photochemical quenching at photosystem II as compared to red actinic light. Based on our experimental data, we hypothesize that structures in the pennate diatom frustule modulate microalgal light absorption for efficient photosynthesis and thereby contribute to niche differentiation of diatoms in the microphytobenthos

145

146 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Introduction

Diatoms are microscopic photosynthetic protists, present in the photic zone of oceans, in the benthos, intertidal habitats and many moist terrestrial ecosystems (Vanormelingen et al. 2007). Due to their high abundance and diversity, diatoms are regarded as an evolutionary extremely successful group (Boyd et al. 2000, Thomas & Dieckmann 2002, Mock & Valentin 2004). This can partially be explained by efficient light use for photosynthesis (Tanada 1951) in combination with effective photo-protection mechanisms (Ruban et al. 2004, Zhu & Green 2010), facilitating high diatom productivity in dynamic light environments (Mortain- Bertrand 1989, Lavaud 2007). Wave-focusing and shifting cloud cover can change solar irradiance locally within milliseconds (Schenk 1957, Max 1986). Spectral differences occur as a function of depth in the water column as influenced by suspended matter and water trophic states (Kirk 2010). Blue light may penetrate deep into oligotrophic waters, while longer wavelengths of the visible light spectrum are attenuated e.g. by the inherent water absorption of red light and near infrared radiation. Blue light attenuates more in turbid or nutrient-rich waters due to scattering and absorption of particulate and dissolved organic matter and absorption by phytoplankton (Atlas & Bannister 1980), utilizing foremost blue and red wavelengths of the visible light spectrum (Gilbert et al. 2000). Diatoms inhabiting the microphytobenthos are exposed to a diffused light field caused by scattering on detritus, inorganic particles and organic matter (Baker & Lavelle 1984). Scattering of light affects both the intensity and quality of available photosynthetic active radiation (PAR) inside the sediment, while it can even enhance PAR locally (Kühl & Jorgensen 1994). In fine sediments, light of shorter wavelengths usually attenuates more due to Rayleigh scattering on densely packed small particles and light absorption by the microphytobenthos, but exhibits higher subsurface scalar irradiance maxima (Kühl et al. 1994, Cartaxana et al. 2016).

Diatoms contain large amounts of Chlorophyll (Chl)-a and Chl-c for efficient light harvesting in the blue and red spectral range of light (Mann & Myers 1968). Carotenoids such as xanthophyll pigments expand the light absorption spectrum of diatoms towards the green wavelength range (Kuczynska et al. 2015). Carotenoids also shield against excessive radiation by heat dissipation under high irradiance; a mechanism that in diatoms is particularly effective under energy-rich blue light (Siefermann-Harms 1987). Phototrophic organisms evolved different mechanisms to cope with high solar irradiance, which can be roughly classified as removal and prevention strategies (Demmig-Adams & Adams 1992). Removal strategies include surveillance and scavenging of reactive oxygen species (Apel & Hirt 2004), photo electron circulation by sinks other than photosynthesis such as photorespiration (Wu et al. 1991), and dissipation of energy in form of heat (Bilger & Björkman 1990). The xanthophyll cycle in diatoms is an effective mechanism to remove excessive energy (Schumann et al. 2007), enabling efficient regulation of photosynthesis under intermittent light regimes (Ruban et al. 2004). Prevention strategies involve structural adaptation enabling a reduction of light exposure

147

via reflection or attenuation of harmful radiation (Haupt 1973, Demmig-Adams & Adams 1992). Prevention strategies may also be mediated by structural features such as biological photonic crystal structures, e.g. it was demonstrated that specialized iridescent protein grana in the mantle of giant clams act as Bragg- reflectors back-scattering harmful radiation and reshaping the propagation of beneficial wavelengths for better exposure of their photosymbiotic microalgae situated in deeper tissue layers (Holt et al. 2014). While coloration of photopigments is based on selective absorption, structural coloration via iridescence can be due to a variety of optical effects such as diffraction, selective mirror-effects, photonic crystals or surface gratings (Kinoshita et al. 2008, Sun et al. 2013).

A unique feature of diatoms is their encasement in a nanostructured silicate frustule, which has been shown to have photonic crystal-like properties (Fuhrmann et al. 2004) interacting with electromagnetic radiation in the ultra-violet (UV) to visible (VIS) spectrum of light (Ellegaard et al. 2016). The term frustule comprises two valves and their corresponding girdle bands (Round et al. 1990), while the term frustule valve refers to the single part of the frustule. The term frustule valve wall refers to the complex asymmetric 3D structures in the valve including chambers and pores of different size in the micro- to nanometer range (De Stefano et al. 2007, Chen et al. 2015, Romann et al. 2015, Valmalette et al. 2015). Diatoms can be classified by the shape and ornamentation of their frustule valve, where centric frustule valves bear a radial symmetry and pennate frustule valves are usually elongated with bilateral symmetric ornamentations (Round et al. 1990). Different functions have been proposed with regards to the frustule material properties and its micro- and/or nano- structure: 1) The complex architecture of the frustule valve provides an enormous physical strength and could serve as predation defense for the living diatom cell (Hamm et al. 2003, Aitken et al. 2016); 2) The frustule might affect sinking rates in diatoms under nutrient depletion (Smetacek 1985); 3) The pH buffering property of biosilica enables the enzymatic conversion of bicarbonate to CO2 inside the cell, thus enabling photosynthetic carbon assimilation (Milligan 2002); 4) The photonic crystal-like structures in the frustule valve could modulate PAR for efficient photosynthesis (Tommasi 2016), i.e. by focusing PAR onto the chloroplasts (De Stefano et al. 2007) or by screening incident light for harmful radiation and excess light energy (Yamanaka et al. 2008). In some diatom species, the biosilica of the frustule itself absorbs UV light that can inhibit photosynthesis and may damage the living cell (Ellegaard et al. 2016). Mycosporine-like amino acids associated with the frustule valve provide an additional UV-protection on the exterior side of the frustule (Ingalls et al. 2010). The photonic properties of frustules may be generated by different optical phenomena such as photonic waveguiding, Rayleigh scattering, diffraction and Fabry-Pérot interference (Hoover & Hoover 1970, Fuhrmann et al. 2004, Dossou 2016). However, the photonic crystal-like structures and frustule optics may significantly vary between the >10000 described diatom species (Guiry 2012) and it is presently complicated to link frustule optical properties to the different life forms of diatoms, i.e., planktonic forms in surface waters, or non-motile and motile life forms inside the microphytobenthos. Only

148 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

raphid diatoms are motile, due to a slit in their frustule valve (the raphe), allowing for locomotion upon excretion of adhesive strands of mucilage (Edgar & Pickett-Heaps 1983).

In this study, we report on the frustule optical properties of different raphid diatom species from an estuarine intertidal mudflat. We describe the photosynthetic activity of this community as a function of light color and photon irradiance, and provide detailed information about the light attenuation characteristics in this particular environment. Motile raphid diatoms were extracted with the lens-tissue technique after induction of upwards migration with light and low-tide stimuli (Consalvey et al. 2004). Frustule optics were determined in different motile diatom species to evaluate whether differences in their frustule optics could affect niche differentiation inside the microphytobenthos.

149

150 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Material and methods

Sampling site

A meso-tidal mudflat was sampled at low tide during August and September 2016 at the Portuguese Atlantic Coast in the estuary system “Ria de Aveiro” (PT, 40°35 ′ N, 8°41 ′ W)(Serodio et al. 2008). Sediment samples were transferred to polyethylene trays and kept at photon irradiances of 100 µmol photons m-2 s-1 provided as white light from a halogen lamp in 10:14 hours day-night cycles. Sea water from the sampling site was added by the end of the daytime period, and drained off again before illumination was started the next day.

Harvesting of motile diatom species, preparation of rinsed frustules and taxonomic identification

Sediment from the intertidal mudflat was brought to the laboratory, where motile raphid diatom species were sampled for the experiments within 3 consecutive days. Immediately after the onset of illumination in the morning, a lens cleaning tissue (WhatmanTM, GE Healthcare, UK) was placed on the sediment surface to collect migrating microphytobenthos. Lens cleaning tissue was removed after 2-4 hours, and the accumulated microphytobenthos was washed and stored in autoclaved sea water for subsequent physiological characterization. Diatom suspensions were further collected for extraction of oxidized frustules. For this, the diatom suspension was centrifuged at 5,000 G prior to removal of organic matter by nitric acid oxidation, which was monitored by the fading color of bromophenol blue added to the suspension. Cleaned diatom frustules were washed in distilled water before a drop of the solution was placed on a coverslip and mounted in NaphraxTM (Northern Biological Supplies, UK). Taxonomic identification and cell counts were done at 100x using an oil immersion objective on a Leica DM2500 LED microscope equipped with differential interference contrast (DIC). A total of 40 random ocular fields were screened and 668 valves were counted. Species were identified according to Witkowski et al. (2000) and references therein.

Scanning electron microscopy

Rinsed frustule valves from the diatom suspension were air dried on metal stubs covered with a thin pellicle of graphite, coated with gold-palladium and observed with a scanning electron microscope (JEOL-JSM 5400, Jeol Ltd., Tokyo, Japan), operated at 10-20 kV.

151

Hyperspectral imaging and image analysis

A water drop containing cleaned diatom frustules was transferred to a microscope glass slide and covered with a glass cover slip. Transmittance was measured with an objective with 10x magnification (UPlanFL N 10x/0.3; Carl Zeiss, Germany) on a light microscope (Axiostar Plus, Carl Zeiss, Germany) connected to a hyper-spectral camera system (VNIR-100, Thermis Vision Systems, St Louis, USA). Hyperspectral images were calculated in percent transmission, after correction for dark noise and normalization to full light transmittance images acquired in the absence of samples. Data acquisition from hyperspectral image stacks was performed with the PhiLumina Hyperspectral Imaging System software (PhiLumina, University of Mississippi, USA).

Dark field microscopy

Cleaned frustule valves from the suspension were observed in a drop of water with an optical compound microscope (BX41 Laboratory Microscope, Olympus, USA) at 40x or 200x magnification, respectively (UPlanFL N 4x/NA=0.13; Carl Zeiss, Jena, Germany; UPlanFL N 20x/NA=0.5; Carl Zeiss, Jena, Germany). Dark stops of different size were applied between light source and condenser (U-PCD2, Olympus, Tokyo, Japan) to generate different angles of incident white light, i.e., θ = 15°, θ = 25°, and θ = 50° . Images were recorded with a charge coupled device camera using the manufacturers’ software (Color View Soft Imaging System, Olympus, Tokyo, Japan).

Variable chlorophyll fluorescence imaging

One mL of diatom suspension was incubated with 1 µL of 0.25 µM latrunculin-A (LAT-A) solution (Thermo Scientific, Waltham, United States). It was earlier demonstrated that LAT-A inhibits motility of raphid diatoms, without affecting photosynthesis (Cartaxana & Serôdio 2008). A drop of inhibited diatom suspension was placed on a microscope glass slide and covered with a glass coverslip. The coverslip was mounted on an epifluorescence microscope (Axiostar Plus, Carl Zeiss, Germany) equipped with a variable chlorophyll fluorescence imaging system (RGB Microscopy I-PAM, Heinz Walz GmbH, Germany; Trampe et al. 2011). Keeping the sample in the dark, blue non-actinic measuring light ( λ = 450 nm) was used to determine minimum fluorescence yields (Fo), while a strong saturating blue light pulse was used to the determine the maximum fluorescence yield in the dark (Fm). Data were recorded using the I-PAM system software (ImagingWin; Heinz Walz GmbH, Germany), while subsequent calculations were done on exported image files of Fo and Fm (in TIF format). Variable fluorescence (Fv) was calculated as Fv=Fm-Fo (Maxwell & Johnson 2000), prior to calculation of the maximum quantum yield of PSII ( ϕPSII)(Genty et al. 1989) as

152 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

ϕPSII =Fv/Fm, using the Ratio Plus plug in and image calculator functions of the freeware Fiji (ImageJ)(Schindelin et al. 2012).

PSII absorption cross section and steady state light curves of PSII electron transport vs. irradiance

The functional wavelength-dependent light absorption cross section of PSII (Sigma λII) and the effective PSII quantum yield was quantified with a multicolor variable chlorophyll fluorescence analyzer (Multi-Color PAM; Heinz Walz GmbH, Germany). Sigma λII was measured at monochromatic light as provided by LEDs at wavelengths of λ = 440, 480, 540, 590, and 625 nm, respectively. Measurements were performed with the script file Sigma1000.FTM executed in the firmware PamWin_3.Ink as described in Schreiber et al. (2012)(Schreiber et al. 2012). Thereafter, samples were dark acclimated for 30 min before steady state light curves were derived from measurements of the effective PSII quantum yield over a set of increasing actinic irradiance levels. Photon irradiance levels of 15, 45, 100, 200, 250, 350, 530, 770, 1050 and 1800 µmol photons m-2 s-1 of actinic blue light ( λ = 440 nm), or 10, 40, 100, 200, 360, 550, 800, 1100, and 1800 µmol photon m-2 s-1 of actinic red light ( λ = 625 nm) were applied for each 12 min prior to saturation pulse measuring the effective PSII quantum yield. Photosynthetic electron transport rates at PSII (ETRII) were calculated under consideration of Sigma λII for the respective colors (Schreiber et al. 2012) as described elsewhere (Genty et al. 1989, Maxwell & Johnson 2000). Non-photochemical energy quenching (NPQ) at PSII was calculated as described by Bilger and Björkman (1990).

Modification of the multicolor PAM measurement head

The measuring head unit of a multicolor PAM (MCP-BK Optics ,Heinz Walz GmbH, Germany) was modified to allow for simultaneous illumination of sediments with colored light, while measuring scalar irradiance depth profiles with a fiber-optic microprobe. A 3D model of the MCP-BK Optics unit was generated in the open source software Autodesk 123D design (Autodesk Inc., Mill Valley, USA) using the dimensions provided in the multicolor PAM manual. In the modified version, an open space was left out close to the measuring window. Consequently, a fiber optic scalar irradiance microprobe could be inserted into the sediment, while the same spot could be illuminated with the PAM emitter rod of the unit (Suppl. Fig. 1). The 3D printed model was made out of PLA nylon filament (Nylon Black, BEEverycreative, Aveiro, Portugal) in a full metal dual extruder (model helloBEEprusa, BEEverycreative, Aveiro, Portugal). More detailed information on how to use this 3D printer is published elsewhere (Serôdio et al. 2017). The modified

153

3D model produced in the software Autodesk 123D design is freely available from the authors (see supplementary information).

Scalar irradiance in sediments with or without diatoms, and in suspensions of oxidized diatom frustules

Autoclaved sediment was baked in an incubator at 70°C for 3 days. Thereafter, sediment was filled into 50 mL tubes and left over night until saturated with 1) autoclaved seawater, or 2) diatom suspension extracted from the intertidal sediment. Attenuation of scalar irradiance was measured with a fiber optic scalar irradiance microprobe(Rickelt et al. 2016) inserted into the sediment surface at an angle of 45° relative to the vertically incident light. Surface positioning of the sensor was controlled by observation through a digital 400x microscope (Dino-Lite AM4013MZT4, Electronics Corporation, USA). The sensor was inserted into the sediment in vertical increments of 0.1 mm using a micromanipulator (MM33, Märtzhäuser, Germany) controlled by an electrical motor (Model 18011 Controller, Oriel Encoder Mike and Oriel, Oriel New Port Cooperation, USA). Downwelling spectral irradiance (Ed) ( λ = 400-800 nm) from a halogen lamp (KL 2500 LCD, Schott AG, Jena, Germany), or from a modified multicolor PAM LED panel (see description above; determined at wavelength λ = 440 and 625 nm, respectively) was recorded with the scalar irradiance microprobe connected to a PC-controlled spectrophotometer (USB 2000+, Ocean Optics, Dunedin, USA) running the manufacturers spectral acquisition software (Spectra Suite, Ocean Optics, Dunedin, USA). Spectral scalar irradiance spectra measured in different depth below the sediment surface were normalized to the incident downwelling irradiance spectra measured over a black light well, i.e. a black taped box. The same setup was used to measure scalar irradiance spectra in a mix of densely packed oxidized frustules in a droplet of water, where the measurements were subsequently normalized to scalar irradiance spectra measured in a droplet of water in the absence of frustules (Fig. 6B).

Statistical analysis

Significance was determined at the P < 0.05 level using one way analysis of variance (ANOVA) followed by the Holm Sidak post-hoc test. Significant differences of Sigma λII were tested on the effects of different wavelengths, i.e. λ = 440, 480, 540, 590 and 625 nm. Significant differences of light attenuation in sediments were tested on one level, i.e. wavelength of λ = 440 nm and λ = 625 nm at different depths, respectively. Significant differences of scalar irradiance in mixed oxidized frustules were tested at wavelengths λ = 440 nm and λ = 625 nm on one level, i.e. scalar irradiance in percent of incident irradiance. All data were normally distributed and passed the Equal Variance test. Degrees of freedom are indicated as subscript letters

154 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

behind F values, while results from the post-hoc test are indicated as statistical power (p)-values. All statistics were performed with SigmaPlot 11.0 (Systat Software Inc., USA).

155

156 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Results

Composition of raphid diatoms in the microphytobenthos

The silicate frustule valves of raphid diatom species bear a raphe slit facilitating motility by excretion of mucilage (Fig. 1). The raphid fraction of diatoms extracted from an estuarine intertidal mudflat was dominated by the species Gyrosigma fasciola (32%), Staurophora sp. 1 (25%) and Navicula sp. 2 (17%). In total, 21 species with an abundance of >0.15% were identified. The large diatom species Gyrosigma balticum and Pleurosigma angulatum were very rare (<0.1%), but an examination of a permanent sample slide at lower magnification (200x) confirmed that they were an integral part of the diatom assemblage (Table 1).

Spectral light transmittance through single frustule valves of raphid diatoms in water

Transmittance spectra through frustule valves of the species Gyrosigma balticum and Pleurosigma angulatum in water showed stronger attenuation of shorter wavelengths, when an objective with lower numerical aperture (NA=0.14) was used. This was not the case in the most abundant species Gyrosigma fasciola , which appeared almost transparent in hyperspectral imaging in the same optical setup (exemplary spectra; Fig. 2B). All three tested species showed a square lattice pattern of large areolae on the exterior side of the frustule valve (Fig. 2C).

Light scattering properties of single frustule valves of raphid diatoms in water

In dark field microscopy, forward scattering of blue light was observed on frustule valves of Gyrosigma fasciola , Gyrosigma balticum and Pleurosigma angulatum . However, the angle of incident white light ( θ) at which this phenomenon was visible differed between the species. Forward scattering of blue light was visible at incident light angles of θ=50° on frustule valves of Gyrosigma fasciola and Pleurosigma angulatum , and at incident light angles of θ=15° and θ=25° on frustule valves of Gyrosigma balticum (Fig. 3A). Although other coloration was possible at different angles, e.g. greenish coloration on the frustule valves of Gyrosigma balticum at a θ=50° angle of incidence, forward scattering of blue light was the most abundant observation in all tested species (Fig. 3A).

157

Spatial distribution of photosynthesis in diatom cells

Differences between the three diatom species were also observed in the distribution of chloroplasts at the single cell level, as determined with a microscope imaging variable chlorophyll fluorescence. We observed differences in the cellular spatial distribution of maximum quantum yields of PSII ( ΦPSII). In cells of Gyrosigma fasciola , ΦPSII was distributed in bilateral symmetry about the apical plane. Cells of Gyrosigma balticum hosted a single large chloroplast filling almost the entire cell in close proximity to the frustule. Cells of Pleurosigma angulatum hosted two clearly defined chloroplasts (Fig. 3B).

Photosynthesis activity of raphid diatoms in suspension

Light color significantly affected the functional wavelength dependent absorption cross-section of PSII (Sigma λII; F4,14 = 33; p<0.001). Sigma λII was maximal at λ = 440 nm with 6.6±0.6 nm2, followed by λ = 480 nm with 4.8±0.7 nm2 and 3.4±0.5 nm2 at λ = 590 nm (n=5; Mean ± SD). Sigma λII was lowest at λ = 540 nm with 2.2±0.3 nm2 and λ = 625 nm with 2.7±0.6 nm2.

Electron transport rates of photosystem II (ETRII) were determined under consideration of Sigma λII as described in Schreiber et al. (2012). Blue light ( λ = 440 nm) induced highest ETRII of 320±23 µmol m-2 s-1 at an irradiance of 750 µmol m-2 s-1 photons, while red light ( λ = 625 nm) induced ETRII of 213±13 µmol m-2 s-1 at an irradiance of 1600 µmol m-2 s-1 photons (n=3; Mean±SD; Fig. 4B).

Non-photochemical energy quenching (NPQ) at photosystem II was higher under blue light as compared to NPQ determined at comparable photon irradiances of red light (n=3; Mean±SD; Fig. 4C).

Scalar irradiance in sediments in the absence and presence of live diatoms

Sediments from an intertidal estuarine mudflat were first autoclaved to determine scalar irradiance (E0) and light attenuation in the absence of live diatoms and cyanobacteria. Depth attenuation of blue light ( λ = 440 nm) was significantly higher compared to that at red light ( λ = 625; n=3; Mean ± SD; F1,53 = 37; p<0.001; Fig. 5A). When diatoms were spiked into the autoclaved sediment, attenuation of shorter wavelength in the tested range λ = 450-800 nm increased. In addition, attenuation in the red spectral range maximum at λ = 664 nm was observed (n=3; Mean ± SD; Fig. 5B). When normalizing scalar irradiance spectra (E0) measured at 0.1 mm depth in sediments spiked with diatoms to similar measurements in autoclaved sediment, a typical diatom light absorption spectrum was calculated. The same light absorption profile was determined in the diatom suspension (n=3; Mean ± SD; Fig. 5C).

158 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Scalar irradiance in a water drop containing mixed frustules from raphid diatoms

The presence of frustules from raphid diatoms significantly enhanced scalar irradiance (in % of incident irradiance) at shorter wavelengths (~1.3 fold) as compared to that of longer wavelength light, i.e. E0 at λ = 440 was 117±8% while E0 was 91±5% at λ = 625 (n=9; Average ± SD; F1,19 = 77; p<0.001; Fig. 6).

159

160 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Discussion

While diatom frustules usually appear transparent in light microscopy when an objective with larger numerical aperture is used (Hoover & Hoover 1970)(Fig. 2A), we observed attenuation of light in the blue spectral range in frustule valves of Gyrosigma balticum and Pleurosigma angulatum when using lower numerical aperture objectives. This differed in frustule valves of the main abundant species Gyrosigma fasciola , which appeared transparent throughout the visible spectrum of light (Fig. 2B). Higher attenuation of blue light by diatom frustule valves has been shown before, e.g. in the freshwater diatom Melosira varians , where it was speculated that more energetic blue light could be absorbed by the biosilica of the frustule valve in order to protect photosynthesis at high irradiance (Yamanaka et al. 2008). Frustule valves of some diatoms can absorb wavelengths in the UV range (Ellegaard et al. 2016), while the amorphous silicate in the diatom frustule (Sumper 2002) is nearly transparent throughout the visible spectrum of light (Fanderlick 1983). Hence, attenuation of blue wavelengths is apparently caused by diffractive interactions of light with the photonic crystal-like structures of the frustule valve wall (Hoover & Hoover 1970). Light can for example diffract on the network of pores on the frustule valve surface (De Stefano et al. 2007)(Fig. 2C), or scatter on the complex silicate structures inside the frustule valve wall (Hoover & Hoover 1970). Such scattering structures can be seen in dark field microscopy where direct illumination is blocked. We observed forward scattering of blue light on the frustule valves in all tested species, however, the angle of incident white light at which this phenomenon was visible varied between the species (Fig. 3A). We speculate that the apparent conservation of blue light scattering observed on the frustule valves of different species could indicate that this phenomenon has an ecological advantage for the diatom cell. This was further promoted by the observed variation of the effect in dependency of the angle of light incidence, indicative for different light micro- environments inhabited by these species inside the microphytobenthos, where some species might position at subsurface positions at diffused light fields (Baker & Lavelle 1984, Kühl & Jorgensen 1994), while other species might experience more collimated sunlight at the sediment surface (Kirk 2010). Different life- strategies might also be indicated by variation of the spatial distribution of photosynthetic compartments in the cell, i.e. number and arrangement of chloroplasts (Fig. 3B). In contrast, pelagic centric diatom species have multiple chloroplasts that are usually evenly distributed over the cell cortical periphery (Yogamoorthi 2007, Armbrecht et al. 2014, Goessling et al. 2016). The structural arrangement of chloroplasts might hence indicate different life-strategies in consequence to the respective light micro-environment. Links between light climate, life-strategy and the biochemical photosynthetic architecture of diatoms were proposed earlier by Strzepek and Harrison (2004), who observed that pelagic species have widely forfeit their photosynthetic flexibility to intermittent light regimes in order to reduce iron requirements (Strzepek & Harrison 2004). We speculate that the variation of frustule valve optical properties and its underlying photonic crystal-like structures are indicative for different life-strategy of diatom species, and furthermore that such structures

161

promote photosynthesis at varying light environments. However, more detailed bio-optical studies at the single cell level are needed to unravel the exact optical mechanisms and their effect on diatom photosynthesis.

Absorption of light in diatoms is dependent on their photopigments, which typically absorb more light in the blue spectral range (Kuczynska et al. 2015). On the other hand, whether or not absorbed light is available for photosynthesis is not only dependent on the type and assembly of pigments at the photosystems, but also on the antenna pigment-connectivity, which can be determined as the wavelength dependent functional light absorption cross-section (Sigma λII)(Schreiber et al. 2012). We confirmed earlier results observed in the centric diatom species Coscinodiscus granii , showing that Sigma λII was significantly higher at blue light compared to that at red light (Goessling et al. 2016)(Fig. 4A). Higher PSII electron transport rates (ETRII) at λ = 450 nm as compared with rates induced by red light (Fig. 4B) may explain the higher diatom productivity when grown under blue light treatments, as observed in phytoplanktonic (Sanchez Saavedra & Voltolina 1995, Goessling et al. 2016) and benthic diatom species (Mercado et al. 2004). However, another study found that diatom growth rates in all eight tested benthic species did not increase under blue light when compared to similar irradiances at other colors (Correa-Reyes et al. 2001). Mercado et al. (2004) therefore concluded that the effect of blue light on the growth of benthic diatoms could be species dependent (Mercado et al. 2004). Energy dissipation by carotenoids can also reduce photosynthetic rates under blue light, which is energy-richer and usually more absorbed (Dougher & Bugbee 2001). We could confirm that blue light induced higher non-photochemical energy quenching (NPQ) compared to that induced by red light (of similar photon irradiance) in a suspension of benthic diatoms (Fig. 4C). Brunet et al. (2014) showed that the xanthophyll pool size and the NPQ amplitude increased when the pennate diatom species Pseudonitzschia multistriata was grown under blue light for several days (Brunet et al. 2014). Hence, the high response of NPQ under blue light observed in our samples may indicate that diatoms from the suspension were adapted to considerable amounts of blue light inside the sediment. Interestingly, we observed that although blue light attenuated more rapidly with depth (Fig. 5A), diatoms were able to harvest blue light more efficiently than longer wavelengths. Thus, a light absorption spectrum typical for diatoms was recorded in vitro, displaying the effective use of blue light by diatoms located at different depths (Fig. 5B, C). To investigate the potential implication of the optical properties of nanostructured diatom frustules upon the light field inside the sediment, we also measured attenuation of incident light in a drop of water containing slurry of oxidized frustules. Here, blue light was significantly enhanced, which might indicate that the frustule optical properties could also modulate PAR on the community level inside the microphytobenthos (Fig. 6).

Stomp et al. (2000) point out that the specific sets of pigments that evolved on Earth can be regarded as instrumental for niche differentiation among phototrophs in relation to the solar spectrum as an energy resource, and such differentiation can lead to reduced competition and coexistence of species (Stomp et al.

162 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

2007). Selective filtration of sunlight by photo-pigments leaves out parts of the spectrum available for other phototrophic organisms. The canopies of rainforests (Endler 1993, Binkley et al. 2013) and stratified microbial mats (e.g. Kühl and Fenchel 2000 ) are well-studied examples of such efficient niche differentiation in regards to absorption and light use throughout the spectrum of sunlight. Biological photonic crystal structures, which are surprisingly widespread among photosynthetic organisms (Glover & Whitney 2010, Vignolini et al. 2013), could further broaden the spectral use of sunlight by local enhancement and redistribution of particular wavelength within the PAR range. Such optical modulation has e.g. been demonstrated in shade dwelling Begonia spp. living on the floor of tropical forests, enhancing the quantum yield of PSII by increased light capture to the green spectral range of light (Jacobs et al. 2016). Photonic crystal structures might thus constitute an advantage in strongly light-limited environments, such as in diatom-dominated microphytobenthos (Watermann et al. 1999). Cyanobacteria, which often occur in combination with diatoms in intertidal mudflats (Underwood & Smith 1998, Stal 2010) may leave colorful niches by lower absorption in the blue-cyan spectral range of light (Nicklisch 1998), where rapid attenuation of these wavelengths inside the sediment might be compensated by photonic crystal structures in the frustule of raphid diatom species.

In conclusion, this first biophotonic study of raphid diatom frustules showed apparently species-specific variation in the optical properties of photonic crystal-like structures in the frustule valve. We speculate that such variation can facilitate niche differentiation inside the microphytobenthos, and we propose that similar relationships might exist in other life-forms of diatoms. Although the frustule may be one outstanding example of structural light manipulation in nature, we propose that photonic structures can also be linked to niche differentiation and efficient photosynthesis in other phototrophic organisms. For now we conclude that the optical properties of the frustule of diatoms modulate visible light in the PAR range inside the cell, and inside the sediment of an estuarine intertidal mudflat.

163

164 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Acknowledgement

Thus study was funded by a Sapere-Aude Advanced grant from the Danish Council for Independent

Research ǀ Natural Sciences (MK) and an instrument grant from the Carlsberg Foundation (MK). This work was also supported by the Fundação para a Ciência e a Tecnologia (FCT) [doctoral fellowship

SFRH/BD/86788/2012 to SF]. We thank Lars Rickelt (www.zenzor.eu) for providing custom-made fiber optic scalar irradiance microprobes. We acknowledge the assistance of William Schmidt during sediment sampling. We are grateful for fruitful discussions with Jörg Frommlet, Matthew Nitschke, João Simões,

Gregor Christa and William Schmidt, and for the excellent technical support during the experimental work at the CESAM lab.

Author contributions

Experimental design (JWG, MK and JS). Diatom sampling (JWG and SF). Optical and physiological experiments (JWG). Identification of diatom taxa (LR). Scanning electron microscopy (JS). Data analysis

(JWG). Wrote the manuscript (JG with editorial assistance of all co-authors).

Competing financial interests

The authors declare no competing financial interest.

165

166 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Literature

Aitken ZH, Luo S, Reynolds SN, Thaulow C, Greer JR (2016) Microstructure provides insights into evolutionary design and resilience of Coscinodiscus sp. frustule. Proc Natl Acad Sci U S A 113:2017– 2022

Apel K, Hirt H (2004) Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu Rev Plant Biol 55:373–399

Armbrecht LH, Smetacek V, Assmy P, Klaas C (2014) Cell death and aggregate formation in the giant diatom Coscinodiscus wailesii (Gran & Angst, 1931). J Exp Mar Bio Ecol 452:31–39

Atlas D, Bannister TT (1980) Dependence phytoplankton of mean spectral extinction coefficient on depth , water color , and species of. Limnol Oceanogr 25

Baker ET, Lavelle JW (1984) The Effect of Particle Size on the Light Attenuation Coefficient of Natural Suspensions. J Geophys Res 89:8197–8203

Bilger W, Björkman O (1990) Role of the xanthophyll cycle in photoprotection elucidated by measurements of light-induced absorbance changes, fluorescence and photosynthesis in leaves of Hedera canariensis. Photosynth Res 25:173–185

Binkley D, Campoe OC, Gspaltl M, Forrester DI (2013) Light absorption and use efficiency in forests: Why patterns differ for trees and stands. For Ecol Manage 288:5–13

Boyd PW, Watson a J, Law CS, Abraham ER, Trull T, Murdoch R, Bakker DC, Bowie a R, Buesseler KO, Chang H, Charette M, Croot P, Downing K, Frew R, Gall M, Hadfield M, Hall J, Harvey M, Jameson G, LaRoche J, Liddicoat M, Ling R, Maldonado MT, McKay RM, Nodder S, Pickmere S, Pridmore R, Rintoul S, Safi K, Sutton P, Strzepek R, Tanneberger K, Turner S, Waite a, Zeldis J (2000) A mesoscale phytoplankton bloom in the polar Southern Ocean stimulated by iron fertilization. Nature 407:695–702

Brunet C, Chandrasekaran R, Barra L, Giovagnetti V, Corato F, Ruban A V. (2014) Spectral radiation dependent photoprotective mechanism in the diatom Pseudo-nitzschia multistriata. PLoS One 9

Cartaxana P, Ribeiro L, Goessling J, Cruz S, Kühl M (2016) Light and O2 microenvironments in two contrasting diatom-dominated coastal sediments. Mar Ecol Prog Ser 545:35–47

Cartaxana P, Serôdio J (2008) Inhibiting diatom motility: a new tool for the study of the photophysiology of intertidal microphytobenthic biofilms. Limnol Oceanogr Methods 6:466–476

Chen X, Wang C, Baker E, Sun C (2015) Numerical and experimental investigation of light trapping effect of nanostructured diatom frustules. Sci Rep 5:11977

Consalvey M, Paterson DM, Underwood GJC (2004) the Ups and Downs of Life in a Benthic Biofilm: Migration of Benthic Diatoms. Diatom Res 19:181–202

Correa-Reyes JG, Sanchez-saavedra MDELP, Siqueiros-beltrones DA, Flores-acevedo N (2001) Growth of Eight Strains of Benthic Under Two Light Conditions. 20:603–610

Demmig-Adams B, Adams WW (1992) Responses of plants to high light stress. Annu Rev Plant Physiol Plant Mol Biol 43:599–626

Dossou KB (2016) Large field enhancement obtained by combining Fabry-Perot resonance and Rayleigh

167

anomaly in photonic crystal slabs. 45101

Dougher T a, Bugbee B (2001) Differences in the response of wheat, soybean and lettuce to reduced blue radiation. Photochem Photobiol 73:199–207

Edgar L a., Pickett-Heaps JD (1983) The Mechanism of Diatom Locomotion. I. An Ultrastructural Study of the Motility Apparatus. Proc R Soc B Biol Sci 218:331–343

Ellegaard M, Lenau T, Lundholm N, Maibohm C, Michael S, Friis M, Rottwitt K, Su Y (2016) The fascinating diatom frustule – can it play a role for attenuation of UV radiation? J Appl Phycol 28:3295– 3306

Endler J (1993) The color of light in forests and its implications. Ecol Monogr 63:1–27

Fanderlick I (1983) Optical properties of glass. Glas Sci Technol:5–8

Fuhrmann T, Landwehr S, Rharbi-Kucki M El, Sumper M (2004) Diatoms as living photonic crystals. Appl Phys B Lasers Opt 78:257–260

Genty B, Briantais J-M, Baker NR (1989) The relationship between the quantum yield of photosynthetic electron transport and quenching of chlorophyll fluorescence. Biochim Biophys Acta - Gen Subj 990:87–92

Gilbert M, Wilhelm C, Richter M (2000) Bio-optical modelling of oxygen evolution using in vivo fluorescence: Comparison of measured and calculated photosynthesis/irradiance (P-I) curves in four representative phytoplankton species. J Plant Physiol 157:307–314

Glover BJ, Whitney HM (2010) Structural colour and iridescence in plants : the poorly studied relations of pigment colour. Ann Bot 105:505–511

Goessling JW, Cartaxana P, Kühl M (2016) Photo-protection in the centric diatom Coscinodiscus granii is not controlled by chloroplast high-light avoidance movement. Front Mar Sci 2:1–10

Guiry MD (2012) How many species of algae are there? 1063:1057–1063

Hamm CE, Merkel R, Springer O, Jurkojc P, Maier C, Prechtel K, Smetacek V (2003) Architecture and material properties of diatom shells provide effective mechanical protection. Nature 421:841–843

Haupt W (1973) Chloroplast Movement. 23:289–296

Holt AL, Vahidinia S, Gagnon YL, Morse DE, Sweeney AM (2014) Photosymbiotic giant clams are transformers of solar flux. J R Soc Interface 11:20140678

Hoover RB, Hoover MJ (1970) Optical Phenomena in Diatoms. Arkansas Acad Sci Proc 1:8–10

Ingalls AE, Whitehead K, Bridoux MC (2010) Tinted windows: The presence of the UV absorbing compounds called mycosporine-like amino acids embedded in the frustules of marine diatoms. Geochim Cosmochim Acta 74:104–115

Jacobs M, Lopez-garcia M, Phrathep O-P, Lawson T (2016) Photonic crystal structure of Begonia chloroplasts enhances photosynthetic efficiency. Nat Plants:1–16

Kinoshita S, Yoshioka S, Miyazaki J (2008) Physics of structural colors. Reports Prog Phys 71:76401

Kirk JTO (2010) Light and Photosynthesis in Aquatic Ecosystems.

Kuczynska P, Jemiola-Rzeminska M, Strzalka K (2015) Photosynthetic pigments in diatoms. Mar Drugs

168 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

13:5847–5881

Kühl M, Jorgensen BB (1994) The light-field of microbenthic communities- radiance distribution and microscale optics of sandy coastal sediments. Limnol Oceanogr 39:1368–1398

Kühl M, Lassen C, Jorgensen BB (1994) Light penetration and light intensity in sandy marine sediments measured with irradiance and scalar irradiance fiber-optic microprobes. Mar Ecol Prog Ser 105:139– 148

Kühl M, Fenchel T (2000) Bio-optical characteristics and the vertical distribution of photosynthetic pigments and photosynthesis in an artificial cyanobacterial mat. Microbial Ecology 40: 94-103.

Lavaud J (2007) Fast Regulation of Photosynthesis in Diatoms: Mechanisms, Evolution and Ecophysiology. Functional Plant Science and Biotechonology 1: 267-287

Mann J., Myers J (1968) On pigments, growth, and photosynthesis of Phaedactylum tricornutum. JPhycology 4:349–355

Max NL (1986) Light diffusion through clouds and haze. Comput Vision, Graph Image Process 33:280–292

Maxwell K, Johnson GN (2000) Chlorophyll fluorescence--a practical guide. J Exp Bot 51:659–68

Mercado JM, Sánchez-Saavedra MDP, Correa-Reyes G, Lubián L, Montero O, Figueroa FL (2004) Blue light effect on growth, light absorption characteristics and photosynthesis of five benthic diatom strains. Aquat Bot 78:265–277

Milligan a. J (2002) A Proton Buffering Role for Silica in Diatoms. Science (80- ) 297:1848–1850

Mock T, Valentin K (2004) Photosynthesis and cold acclimation: Molecular evidence from a polar diatom. J Phycol 40:732–741

Mortain-Bertrand A (1989) Effects of light fluctuations on the growth and productivity of Antartic diatoms in culture. Polar Biol 9:245–252

Nicklisch A (1998) Growth and light absorption of some planktonic cyanobacteria, diatoms and Chlorophyceae under simulated natural light fluctuations. J Plankton Res 20:105–119

Rickelt LF, Lichtenberg M, Trampe ECL, Kühl M (2016) Fiber-Optic Probes for Small-Scale Measurements of Scalar Irradiance. Photochem Photobiol 92:331–342

Romann J, Valmalette J-C, Chauton MS, Tranell G, Einarsrud M-A, Vadstein O (2015) Wavelength and orientation dependent capture of light by diatom frustule nanostructures. Sci Rep 5:17403

Round FE, Crawford RM, Mann DG (1990) The Diatoms: Biology and Morphology of the Genera. Cambridge Univ Press Cambridge, UK, 1990

Ruban A, Lavaud J, Rousseau B, Guglielmi G, Horton P, Etienne AL (2004) The super-excess energy dissipation in diatom algae: Comparative analysis with higher plants. Photosynth Res 82:165–175

Sanchez Saavedra MP, Voltolina D (1995) The effect of different light quality on the food value of the diatom Chaetoceros sp. for Artemia franciscana. Riv Ital Acquacolt 30:135–138

Schenk HJ (1957) On the Focusing of Sunlight by Ocean Waves. J Opt Soc Am 47:653–657

Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, Tinevez J-Y, White DJ, Hartenstein V, Eliceiri K, Tomancak P, Cardona A (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682

169

Schreiber U, Klughammer C, Kolbowski J (2012) Assessment of wavelength-dependent parameters of photosynthetic electron transport with a new type of multi-color PAM chlorophyll fluorometer. Photosynth Res 113:127–44

Schumann A, Goss R, Jakob T, Wilhelm C (2007) Investigation of the quenching efficiency of diatoxanthin in cells of Phaeodactylum tricornutum (Bacillariophyceae) with different pool sizes of xanthophyll cycle pigments. Phycologia 46:113–117

Serôdio J, Schmidt W, Frankenbach S (2017) A chlorophyll fluorescence-based method for the integrated characterization of the photophysiological response to light stress. J Exp Bot:erw492

Serodio J, Vieira S, Cruz S (2008) Photosynthetic activity, photoprotection and photoinhibition in intertidal microphytobenthos as studied in situ using variable chlorophyll fluorescence. Cont Shelf Res 28:1363– 1375

Siefermann-Harms D (1987) The light-harvesting and protective functions of carotenoids in photosynthetic membranes. Physiol Plant 69:561–568

Smetacek VS (1985) Role of sinking in diatom life-hystory: ecological, evolutionary and geological significance. Mar Biol 84:239–251

Stal LJ (2010) Microphytobenthos as a biogeomorphological force in intertidal sediment stabilization. Ecol Eng 36:236–245

Stefano L De, Rea I, Rendina I, Stefano M De, Moretti L (2007) Lensless light focusing with the centric marine diatom Coscinodiscus walesii. Opt Express 15:18082–18088

Stomp M, Huisman J, Stal LJ, Matthijs HCP (2007) Colorful niches of phototrophic microorganisms shaped by vibrations of the water molecule. ISME J 1:271–282

Strzepek RF, Harrison PJ (2004) Photosynthetic architecture differs in coastal and oceanic diatoms. Nature 431:689–692

Sumper M (2002) A phase separation model for the nanopatterning of diatom biosilica. Science 295:2430– 2433

Sun J, Bhushan B, Tong J (2013) Structural coloration in nature. RSC Adv 3:14862

Tanada T (1951) The Photosynthetic Efficiency of Carotenoid Pigments in Navicula minima. Am J Bot 38:276

Thomas DN, Dieckmann GS (2002) Antarctic Sea ice--a habitat for extremophiles. Science 295:641–644

Tommasi E De (2016) Light manipulation by single cells: the case of diatoms. J Spectrosc 2016:1–13

Trampe E, Kolbowski J, Schreiber U, Kühl M (2011) Rapid assessment of different oxygenic phototrophs and single-cell photosynthesis with multicolour variable chlorophyll fluorescence imaging. Mar Biol 158:1667–1675

Underwood GJC, Smith DJ (1998) Predicting epipelic diatom exopolymer concentrations in intertidal sediments from sediment chlorophyll a. Microb Ecol 35:116–125

Valmalette J, Romann J, Røyset A, Einarsrud M (2015) 3D-Hyperspectral mapping of light propagation through diatom frustule silica nanostructures. Opt Lett 40:3–5

Vanormelingen P, Verleyen E, Vyverman W (2007) The diversity and distribution of diatoms: From

170 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

cosmopolitanism to narrow endemism. Biodivers Conserv 17:393–405

Vignolini S, Moyroud E, Glover BJ, Steiner U, Glover BJ (2013) Analysing photonic structures in plants.

Watermann F, Hillebrand H, Gerdes G, Krumbein WE, Sommer U (1999) Competition between benthic cyanobacteria and diatoms as influenced by different grain sizes and temperatures. Mar Ecol Prog Ser 187:77–87

Witkowski A, Lange-Bertalot H MD (2000) Diatom flora of marine coasts. In: H L-B (ed) Iconographia Diatomologica: annotated diatom micrographs. Koeltz Scientific Books

Wu J, Neimanis S, Heber U (1991) Photorespiration is More Effective than the Mehler Reaction in Protecting the Photosynthetic Apparatus against Photoinhibition. Bot Acta 104:283–291

Yamanaka S, Yano R, Usami H, Hayashida N, Ohguchi M, Takeda H, Yoshino K (2008) Optical properties of diatom silica frustule with special reference to blue light. J Appl Phys 103:1–6

Yogamoorthi a (2007) Artificial UV-B induced changes in pigmentation of marine diatom Coscinodiscus gigas. J Environ Biol 28:327–30

Zhu SH, Green BR (2010) Photoprotection in the diatom Thalassiosira pseudonana: Role of LI818-like proteins in response to high light stress. Biochim Biophys Acta - Bioenerg 1797:1449–1457

171

172 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Tables

Table 1: Diatom taxa collected from an estuarine intertidal mudflat (Aveiro, Portugal).

The taxa are listed in alphabetical order with the 3 most abundant species marked by grey shading.

Abundance Taxon Rank Size-class (length in µm) (%)

Amphora sp. 1 8 12 <150 Amphora sp. 2 0.15 12 <50 Amphora sp. 3 0.3 11 <100 Entomoneis sp. 1 0.15 12 <100 Fallacia sp. 1 0.15 12 <100 Gyrosigma fasciola (Ehrenberg) J. W. Griffith & Henfrey 31.89 1 <50 Navicula cf. phyllepta Kützing 3.89 7 <50 Navicula gregaria Donkin m.2 4.79 6 <50 Navicula phyllepta Kützing 0.45 10 <50 Navicula salinarum Grunow cf 0.15 12 <50 Navicula sp. 1 5.84 4 <50 Navicula sp. 2 17.37 3 <20 Navicula sp. 3 3.29 8 <20 Navicula sp. 4 0.45 9 <50 Navicula sp. 5 0.15 12 <50 Nitzschia frustulum (Kützing) Grunow 0.3 11 <100 Nitzschia pararostrata (Lange-Bertalot) Lange-Bertalot 0.15 12 <20 Nitzschia sp. 1 0.3 11 <50 Staurophora salina (W. Smith) Mereschkowsky 4.49 5 <50 Staurophora sp. 1 25.45 2 <20 Tryblionella apiculata Gregory 0.15 12 <50

173

174 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Figures

Figure 1: Organization of the frustule of motile raphid diatoms.

Raphid frustule scheme in (A) lateral view and (B) distal view. (C) SEM image of the raphid pennate diatom species Gyrosigma fasciola . (D) Close-up on the exterior surface of the Gyrosigma fasciola frustule valve observed in SEM. (E) Illustration of the frustule valve wall nanostructure in cross-section. Abbreviations:

Central area (ca); sternum (s); raphe (r). Scale bars: 5 µm (C, D); ~500 nm (E)

175

176 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Figure 2: Structural coloration of frustule valves in raphid diatoms.

(A) Micrograph of Gyrosigma balticum in differential interference contrast microscopy. (B) Transmission spectra of single frustule valves of Gyrosigma fasciola (green), Gyrosigma balticum (grey) and Pleurosigma angulatum (black) observed with hyperspectral imaging. (C) SEM images of the corresponding species showing the arrangement of large areolae on the exterior frustule valve side. Square lattice patterns are indicated in red color. Scale bars: 20 µm (A); 1 µm (C).

177

178 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Figure 3: Optical properties of frustule valves and in vivo distribution of photosynthesis.

(A) Light scattering properties of frustule valves in water observed with objectives of different numerical aperture (NA) in dark field microscopy. The angle of incidence ( θ) was varied with dark stops of different size. (B) Distribution of the maximum quantum yield of PSII (see color bar for scale) in Gyrosigma fasciola ,

Gyrosigma balticum and Pleurosigma angulatum , as observed by variable chlorophyll fluorescence microscopy (Trampe et al. 2011). The spatial arrangement of chloroplasts (chlpst) and the frustule outlines are indicated. Scale bars: 100 µm (A, B).

179

180 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Figure 4: Photosynthetic activity of raphid diatoms from an estuarine intertidal mudflat.

(A ) Functional wavelength-dependent absorption cross section of PSII (Sigma λII) determined at different colors in the visible spectrum of light. Significant differences (p=0.05) are indicated by use of small letters.

(B) Steady-state electron transport rates at PSII (ETR II ) measured as a function of photon irradiance and light color. (C) Non-photochemical energy quenching at PSII (NPQ) as a function of photon irradiance and light color.

181

182 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Figure 5: Spectral attenuation of light in a diatom dominated intertidal estuarine mudflat.

(A) Attenuation of blue and red scalar irradiance in autoclaved sediment. Asterix indicate significance at the p≤0.05 (*), p ≤0.01 (**), and p ≤0.001 level (***). (B) Spectral scalar irradiance ( E0) in different depths normalized to the downwelling irradiance at the sediment surface ( Ed), as measured in autoclaved sediment in the absence (Autoclaved sediment) and presence of diatoms (Spiked diatoms). (C) Light absorption spectrum of diatoms calculated as the ratio of scalar irradiance in autoclaved sediment with and without diatoms at 0.1 mm depth, respectively. The Optical density (OD Lambda ) of a diatom suspension at different

183

wavelengths is shown as bars.

184 6.5. Niche differentiation in the microphytobenthos shaped by the optical properties of frustule valves of raphid diatoms

Figure 6: Light enhancement by nanostructured frustules of raphid diatoms.

(A) Spectral scalar irradiance measured in a water drop containing clean oxidized frustules of raphid diatoms collected from an intertidal estuarine mudflat. The inset shows the highly concentrated frustule slurry. (B)

Schematic illustration of the setup: The light collecting sphere (dark grey circle) of a scalar irradiance microprobe collected diffused light (small yellow arrows) scattered on the frustules from different raphid diatom species (light grey) placed in a drop of water (blue). The microprobe was inserted into the droplet at an angle of 45° relative to the vertically incident collimated white light (large yellow arrow).

185

186

List of included manuscripts:

1. Goessling, J.W., Su, Y., Cartaxana, P., Maibohm, C., Rickelt, L.F., Trampe, E.C., Walby, S., Wangprasseurt, D., Wu, X., Ellegaard, M. & Kühl, M. Optical properties of the silicate frustule mediate efficient photosynthesis in diatoms. In prep.

2. Goessling, J.W., Su, Y., Maibohm, C., Ellegaard, M. & Kühl M. Optical properties of the nanoporous girdle band of Coscinodiscus granii frustules. In prep.

3. Goessling, J.W., Cartaxana, P., & Kühl, M. Photo-protection in the centric diatom Coscinodiscus granii is not promoted by chloroplast high-light avoidance movement. Front. Mar. Sci. 2, 1-10 (2016)

4. Cartaxana, P., Ribeiro, L., Goessling, J.W., Cruz, S. & Kühl M. Light and O2 micro- environments in two contrasting diatom-dominated coastal sediments. Mar. Ecol. Progr. Ser. 545, 35-47 (2016)

5. Goessling, J.W., Frankenbach, S., Ribeiro, L., Serodio, J. & Kühl M. Niche differentiation in the microphytobenthos is affected by the optical properties of raphid diatom frustules. In prep.

João Serôdio