Biogeography, Species Diversity and Stress Tolerance of Aquatic and Terrestrial Diatoms

Ghent University Faculty of Sciences, Department of Biology Research group of Protistology and Aquatic Ecology

Biogeography, species diversity and stress tolerance of aquatic and terrestrial diatoms

ȱě

ȱĴȱȱȱęȱȱ Promotor: Prof. Dr. Wim Vyverman the requirements for the degree of Doctor in Co-promotor: Prof. Dr. Koen Sabbe Sciences - Biology Exam Committee

ȱȱȱȱĴ Prof. Jens Boenigk (University of Duisburg-Essen, Germany) Dr. Eileen J. Cox (Natural History Museum, London, UK) Dr. Frederik Leliaert (Ghent University)

ȱȱȱ¡ȱĴ Prof. Dr. Luc Lens (Chairman, Ghent University) Prof. Dr. Wim Vyverman (Promotor, Ghent University) Prof. Dr. Koen Sabbe (Co-promotor, Ghent University) Prof. Dr. Jens Boenigk (University of Duisburg-Essen, Germany) Dr. Eileen J. Cox (Natural History Museum, London, UK) Dr. Frederik Leliaert (Ghent University) Dr. Bart Van de Vijver (National Botanic Garden of Belgium) Dr. Elie Verleyen (Ghent University)

2011 Universiteit Gent In science we resemble children collecting a few pebbles at the beach of knowledge, while the wide ocean of the unknown unfolds itself in front of us. Isaac Newton

Contents

Chapter 1 General introduction 1

PART 1 PATTERNS IN DIATOM DISTRIBUTION

Poles apart: Interhemispheric contrasts in diatom diversity driven Chapter 2 by climate and tectonics 29

Distance decay and species turnover in terrestrial and aquatic Chapter 3 diatom communities across multiple spatial scales in three sub- 45 Antarctic islands

PART 2 SPECIATION IN TIME AND SPACE

Time-calibrated multi-gene phylogeny of the diatom genus Chapter 4 Pinnularia 73

Phylogenetic signals in the valve and plastid morphology of the Chapter 5 diatom genus Pinnularia 103

Molecular evidence for the existence of distinct Antarctic lineages Chapter 6 in the cosmopolitan terrestrial diatoms Pinnularia borealis and 149 Hantzschia amphioxys

PART 3 TOLERANCE LEVELS FOR DISPERSAL-RELATED STRESSES

Tolerance of benthic diatoms from temperate aquatic and Chapter 7 terrestrial habitats to experimental desiccation and temperature 177 stress

Resistance of freshwater benthic diatom resting stages Chapter 8 to experimental desiccation and freezing depends on the 205 temporality of their habitat

Chapter 9 General discussion 227

References 241

Summary 271

Samenvatting 276

Acknowledgments 283

1. General introduction

1.1 Biogeography 1.1.1 General biogeographical processes Biogeography studies the geographical distribution of species over space and time. While before the 20th century biogeography was part ȱȱ¢ȱȱ¢ȱȱȱȱęȱȱȱ “where” and “when” organisms live and had lived (e.g. Linnaeus, 1781; de Candolle, 1820; Wallace, 1876), it changed throughout the 20th century into a more explanatory research area, trying to answer the questions “why” a particular organism lives or had lived at a particular place, and “why not”. This change was stimulated by ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ e.g. area, immigration and extinction rate (Arrhenius, 1921; MacArthur & Wilson, 1963) and the discovery of plate tectonics (Wegener, 1924), supplying explanatory mechanisms for the observed historical and ȱȱĴȱȱȱǯȱȱǰȱȱȱȱ done to further understand the processes underlying biogeographical Ĵȱǯȱ¢ȱȱ ȱǰȱȱȱȱ of a species starts with its formation in a suitable habitat in a certain region, dispersal into and eventually colonization of new regions resulting in an expansion of its biogeographical range, further ȱ ȱ ȱ ¡ȱ ȱ Ěȱȱ ȱ Ȭȱ environmental changes and biotic interactions, and ultimately ends with the extinction of the species or further evolution and ǯȱ ȱȱ¢ȱȱȱ ȱȱȱȱěȱȱ the distribution of an organism, namely historical factors and local environmental factors (Lomolino, 2010).

Historical factors ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱȱȱȱȱȱȱǻĚȱȱȱȱ among these patches), the geographical position and isolation of regions and larger geographical areas, and historical changes in

General introduction 1 ȱęȱȱȱȱȱȱǻĚȱȱȱ and sea level changes) and plate tectonics. On the other hand, historical factors also include features of the species itself such as its area of origin, its evolutionary background and the resulting phylogenetic constraints on the characteristics of the species (stress tolerance levels, adaptive and competitive potential, environmental niche, etc.Ǽȱ ȱ ȱěȱȱȱȱȱ£ȱǯȱ The dispersal ability of a species depends on a whole range of factors, including its morphological and physiological characteristics (such as active locomotion or features enhancing passive motility, dormancy and tolerance levels for adverse conditions during Ǽǰȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ medium or habitat they have to bridge during dispersal (physical barriers, heterogeneity, environmental factors, biotic interactions). The colonization ability of a species depends on its competitive strength, niche breadth and adaptive abilities. The combined action ȱȱȱȱ¡ȱȱȱ ȱ¢ȱěȱȱ the dispersal capacities and environmental niche of the species and thus its realized biogeographical range, but can also result in subsequent speciation or in extinction (Coyne & Orr, 2004).

Local environmental factors Local environmental factors comprise the present biotic and abiotic conditions in a habitat patch, including positive and negative interactions with other organisms and the chemical characteristics ȱȱǯȱȱȱ¢ȱȱȱȱȱ factors in a local patch is called “species sorting” (Leibold et al., 2004) and can result in the local extinction or unsuccessful colonization of species for which the habitat patch is not suitable. The interplay between local environmental factors (species ǼȱȱȱȱęȱȱȱȱȱȱȂȱ occurrences and community structure is explicitly incorporated in the metacommunity concept (Leibold et al., 2004). A metacommunity ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ dispersal of multiple potentially interacting species, which are available in the regional species pool. Using this concept, several studies have recently tried to estimate the relative contribution of ȱȱȱȱȱȱęȱȱȱȱ patches in determining contemporary local community composition, over a range of spatial scales and organism groups (e.g. Potapova &

2 General introduction ǰȱŘŖŖŘǲȱĴȱǰȱŘŖŖśǲȱ£ȱet al.ǰȱŘŖŖśǲȱȱȱ ȱet al., 2007).

Speciation and extinction ȱȱ¡ȱȱ ȱȱȱȱ¢ǯȱ ȱȱȱȱȱȱȱȱȱǻȱ£ǰȱ ŘŖŖŝǼǰȱȱȱ ȱȱĚȱ ȱ ȱȱȱȱȱ species is reduced or prevented by a geographical, reproductive and/or ecological barrier (Cook, 1906, 1908; Coyne & Orr, 2004). The geographical distance between populations or the occurrence of a physical barrier (any unsuitable habitat type such as oceans, continents, rivers or mountains) can reduce dispersal and thus gene Ěȱ ȱ ȱǯȱȱȱĚȱ ȱȱĜȱȱ¢ȱǰȱ random genetic drift and/or ecological adaptation will occur and eventually result in the formation of genetically separated lineages, which will become reproductive isolated over time (allopatric and ȱǰȱ¢ȱǭȱǰȱŘŖŖŚǼǯȱȱȱȱĚȱ ȱ can also occur inside a single area when populations get ecologically ȱ¢ȱǰȱȱȱȱȱȱȱĚȱ ȱ ȱȱ¢ǰȱ ȱȱȱěȱȱ (sympatric speciation, Cook, 1908; Coyne & Orr, 2004). Because of the crucial position of dispersal in allopatric speciation, which is supposed to be the main mode of speciation (Coyne and Orr, 2004), ȱȱȱȱȱȱěȱȱȱȱ¢ǯȱȱ ǰȱȱȱȱȱĚȱȱȱȱȱȱȱ processes including biogeography (Macarthur & Wilson, 1963), population dynamics (Johst & Brandl, 1997) and genetics (Hanski, 1999), metacommunity dynamics (Leibold et al., 2004), and speciation (Coyne & Orr, 2004).

1.1.2 Microbial biogeography Knowledge of the general processes determining the biogeographical range of species mainly comes from studies on macroorganisms (see for example the classic studies compiled in Lomolino et al., 2004). For microorganisms, much less is known on their biogeographies, mainly because their small sizes require the use of microscopes, culturing or molecular techniques to detect them and because it was ȱ ȱ ȱ Ȃȱ ȱ ȱ ȱ structure anyway (Martiny et al., 2006). The combination of high local population abundances (and thus a higher chance to get dispersed) and small sizes (and thus low weight, increasing the

General introduction 3 chance and distance of passive transport) was believed to increase the dispersal probabilities of microorganisms to such an extent that no geographical barriers would exist and only species sorting would ěȱȱȱȱǻǰȱŗşŗřǲȱȬǰȱŗşřŚǼǯ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ Ĵȱȱ ȱ Ĵȱȱ understanding of the ecology and evolution of microorganisms. ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ Ĵȱȱ in several eukaryotic (e.g. Fawley et al.ǰȱ ŘŖŖŚǲȱ ȱ et al., 2004; Pringle et al.ǰȱ ŘŖŖśǲȱ ȱ et al., 2006; Taylor et al., 2006; Evans et al., 2009; Boo et al., 2010; Casteleyn et al., 2010) and prokaryotic microorganisms (e.g.ȱȱǭȱǰȱŘŖŖŖǲȱȱet al., 2003; Whitaker et al., 2003), while other species turn out to have cosmopolitan distributions (e.g.ȱ ȱ et al., 2000; Brandao et al., 2002; Ward ǭȱȂǰȱŘŖŖŘǲȱ et al.ǰȱŘŖŖśǲȱȱet al., 2010). This indicates that microbial biogeographies are very likely shaped by the same processes as assessed for macroorganisms (Martiny et al., 2006), being history and dispersal limitation, next to species sorting.

Problems related to studying microbial biogeography Besides the geographical component, these recent studies also ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ¢ȱ ȱȱĴȱȱ¢ȱȱ¢ȱ species sorting. First and most importantly, the taxonomic resolution used in microbial studies is usually much lower compared to the studies on macroorganisms (Martiny et al., 2006). Indeed, leading ȱȱȱȱȱȱȱȱȱǻe.g. Fawley et al., 2004; Boenigk et al.ǰȱ ŘŖŖŜǼȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ǻe.g.ȱ ȱ et al., 2004), ȱȱȱȱȱȱȱ¡ȱȱȱȱȱȱ too conservative to distinguish closely related species (Behnke et al., ŘŖŖŚǼǯȱǰȱęȱȱȱȱȱȱȱ characteristics is commonly used in eukaryotic microorganisms. However, many studies have reported within single morphospecies the presence of distinct genetic lineages or species without or with ¢ȱ¢ȱȱȱěȱȱ ȱȱǻ¢ȱ and pseudocryptic species, respectively; e.g.ȱ ȱ et al., 2006; Beszteri et al.ǰȱ ŘŖŖŝǲȱ ȱ et al., 2008). The oldest lineages of the ĚȱȱȱȱȱMicromonas pusilla (Butcher) ȱǭȱȱȱȱȱȱȱȱȱȱȱŜśȱȱ

4 General introduction ¢ȱǻȱet al., 2006). This further biases our understanding of microbial biogeography. ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ of microorganisms (Martiny et al., 2006). The dispersal rates and distances of an organism depend on one hand on intrinsic factors such as the number of cells dispersing (and thus population sizes), their morphology (weight and special structures promoting transport) and their physiological tolerance to the adverse conditions encountered during transport and colonization; and ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ picked up by a dispersal vector and the chance to land in a suitable habitat patch. Apart from that, successful dispersal also depends on organism characteristics allowing it to compete with the local communities and populations already established in a patch (such as competitive ability, dormancy, adaptive potential). Of all these ǰȱ ǰȱ Ĵȱȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱ microorganisms certainly can have extremely large population sizes, this is most probably not the case for all microbial species (see ȱ¡ȱ ȱǭȱǰȱŘŖŗŖǼǯȱ¢ǰȱȱȱȱȱȱ the dispersal probabilities of all microorganisms are high enough to overcome genetic divergence between populations (Martiny et al., 2006). Many bacteria and eukaryotic microorganisms do form dormant stages (Hutchinson, 1967; Hargraves & French, ŗşŞřǲȱȱet al., 2000; Anderson, 2010), but this is for several taxa still unknown and there are not always data available on the ȱȱȱȱȱȬȱǯȱ ȱǰȱ variation in niche breadth, competitiveness and adaptive capacities ȱȱȱȱȱȱǯȱǰȱȱȱȱ ȱęȱȱȱȱ¡ȱȱȬȱȱǻe.g. ǰȱŗşśŞǲȱȱet al., 2008; Müller et al., 2010), suggesting that taxa will also vary in their dispersal probabilities, dispersal rates and realized biogeographies (Martiny et al.ǰȱŘŖŖŜǼǯȱȱ ȱęȱȱȱ ȱ ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ range of scales on which geographical structure has been detected, and would further corroborate the current opinion that microbial biogeography is dictated by the same processes as macroorganisms ǻ ȱǭȱǰȱŘŖŖŜǲȱ¢ et al., 2006).

General introduction 5 Dispersal mechanisms and dispersal barriers of microorganisms ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ location to another, away from its parental population. An individual of a species can disperse from its parental population into another population, or disperse from its parental population towards an empty patch, colonize this patch and start a new population (if propagating asexually) or await a second individual of opposite mating type and start a new population (for sexual organisms). For ¢ǰȱȱȱȱȱĚȱȱ ȱȱȱ ȱ ȱ ȱ ȱ ȱ ǻȱ ȱ ȱ Ěȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ Ȭȱ system) and when the movement is towards a new patch (thus expanding a biogeographical range or enhancing the number of ȱǼǯȱ ¢ǰȱȱěȱȱȱȱȱ subsequent reproduction and the onset of a population. Two main types of dispersal mechanisms are generally recognized, being active and passive dispersal. Active dispersal will depend on local factors (local population size, resource limitation, habitat size, etc.) forcing organisms to actively move out of a ęȱȱȱǻ ȱǭȱǰȱŘŖŖśǼǯȱȱ¢ȱǰȱ however, organisms depend on the kinetic energy occurring in the environment (animal and plant vectors, wind, water currents, gravity) for their movement (Maguire, 1963). For microorganisms, the active movement of a single individual will be spatially restricted and unlikely to bridge the gap between populations. Therefore, passive dispersal is the main dispersal type of microorganisms. Microorganisms are thus dispersed through animal and plant vectors, wind and water currents. ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ organism and include geographical barriers and physiological ǯȱȱȱĜȱȱȱȱǰȱȱȱȱ (rivers, oceans, land barriers, mountains, etc.) only preclude the dispersal of an individual because of physiological restrictions. ǰȱȱȱȱȬęȱȱǻȱ ȱȱ¢ȱ ȱǼǯȱȱǰȱȱȱȱȱȱ conditions will be encountered, ranging from among others an ȱȱȱǻȂȱǰȱȱȱ ȱȱ¡Ǽǰȱ adverse environmental conditions in the crossed habitat patches, to ȱǯȱȱȱȱȬȱȱ ȱ ȱĚȱȱȱȱȱȱȱȱȱ¢ȱ of surviving a dispersal event. Besides this, the availability

6 General introduction of individuals (and thus population size) and morphological ȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ dispersed passively, while the availability of suitable habitat patches ȱ£ȱȱ ȱěȱȱȱ¢ȱȱȱȱȱ ȱȱȱȱ£ȱǯȱ ǰȱĴȱȱȱ ȱȱ the extent in which dispersal limitation occurs in microorganisms ȱȱȱȱȱȱȬȱǯ

1.2 Organisms under study: benthic freshwater and terrestrial diatoms 1.2.1 Diatoms: general introduction ȱ ǻ¢ǰȱ ¢Ǽȱ ȱ ȱ ȱ ȱ unicellular microalgae characterized by a siliceous cell wall (the frustule) consisting of two “valves” connected by girdle elements (Fig. 1.1), and a diplontic life cycle involving gradual cell size reduction by vegetative divisions followed by cell size restitution through sexual reproduction (Round et al., 1990; Chepurnov et al., ŘŖŖŚǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ siliceous cell walls, which have traditionally been used to delineate species (Round et al., 1990). Moreover, as this silica wall is relatively resistant to corrosion, diatoms have an extensive fossil record, going ȱ ȱ ȱ ǰȱ ŗşŖȱ ȱ ¢ȱ ȱ ǻĵȱǰȱ ŗŞşŜǰȱ ŗşŖŖȱ ȱ ȱ et al., 2006). Within the more than 200,000 extant species ȱȱ¡ȱǻȱǭȱǰȱŗşşŜǼǰȱ ȱȱȱ groups are traditionally delineated (Fig. 1.1), being the “centric” diatoms with radially organized valves and the “pennate” diatoms with bilaterally organized valves. A subgroup of pennate diatoms has evolved a longitudinal slit along the apical axis of the valve, called the raphe (Fig. 1.1), through which polysaccharide strands can be extruded onto the substrate, allowing the movement over ȱ ȱ ǻȱ ǭȱ ǰȱ ŗşŜŜǼǯȱ ȱ ȱ ȱ ȱ pennate diatoms (having no raphe) are immotile and live suspended ȱ ȱ ȱ Ĵȱȱ ȱ ǰȱ ȱ ȱ ȱ ȱ are motile and live mostly benthic in sediments and on substrates. ȱȱȱȱȱŝŖȱȱȱǻĴȱǰȱŗŞŞŜǲȱǰȱŗŞŞşǲȱ Ȭ et al.ǰȱ ŘŖŖŘǲȱ ȱ et al., 2006), raphid diatoms have ęȱȱ¢ȱȱȱȱȱ ȱȱ ȱȱȱ¢ȱȱȱ¡ȱȱǰȱȱȱȱ

General introduction 7 Fig. 1.1 Living diatom cells as visualized by light microscopy (A-B) and the cleaned siliceous frustule visualized by scanning electron microscopy (C-D). Diatoms can be elongated and bilateral symmetric (A-C) and termed pennate diatoms, or radially symmetric (B-D) and termed centric diatoms. The “raphe” is the longitudinal slit running over the apical axis of the valve of raphid pennate diatoms, such as in C. A) ¢ȱǯ, B) Cyclotella sp., C) Pinnularia borealis, D) a multipolar centric. ȱȱěȱȱ¢ȱȱȱȱȱȱǻȱet al., 2006). Being photoautotrophic, diatoms thrive worldwide in the photic zone of the marine and freshwater plankton and benthos (Round et al., 1990). They are ecologically very important as primary ǰȱȱȱȱȱȱȱŘŖȬŘśȱƖȱȱȱȱ primary production in oceans (Werner, 1977). To a lesser extent, diatoms also occur in terrestrial habitats and live in and on soils and on wet rocks, mosses, tree barks and other humid substrates (Round et al., 1990). However, not much is known about the ecology of terrestrial diatoms, as most ecological, physiological, biochemical and molecular studies have focused on marine or freshwater diatoms. Concerning biogeography and dispersal, several studies have been performed and some general trends and work hypotheses can be ȱ ȱ ȱ ǻȱ ȱ ŗǯŘǯŘǼǯȱ ǰȱ ȱ ęȱȱ mechanisms controlling diatom distributions are still unknown, and many questions remain. Moreover, similar to other microbial taxa,

8 General introduction several species complexes containing (pseudo)cryptic species have been reported (e.g.ȱȱet al.ǰȱŘŖŖśǲȱ£ȱet al., 2007; Mann & ǰȱŘŖŖŝǲȱȱet al.ǰȱŘŖŖŞǲȱȱet al., 2009). Because cryptic diversity hampers the interpretation of geographical ranges, this issue has to be taken into account when reading previous ¢Ȭȱȱȱȱ ȱȱ ȱ ones.

1.2.2 Diatom biogeography: an interplay of species sorting and dispersal limitation? ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ geographical distribution of diatom species is only determined by species sorting, or by an interaction between local ecological ȱ ȱ ȱ ȱ ǻȱ ǭȱ ǰȱ ŘŖŖŘǲȱ ȱ et al., 2004; Telford et al.ǰȱŘŖŖŜǲȱ¢ȱet al.ǰȱŘŖŖŝǲȱ¢ȱet al., ŘŖŖşǼǯȱȱȱȱȱȱȱȱȱ reveal that their geographical distributions are explained by local environmental variables and spatial variables, indicating that both species sorting and dispersal limitation account for the observed geographical structure. This result is further supported by reports of endemic diatom species and genera [e.g. species of Muelleria ǻǼȱ ȱ ǻȱ ȱ ȱ et al., 2010), and the genus Eunophoraȱ¢ǰȱȱǭȱȱǻ¢ȱet al., 1998)] and ¢ȱ ȱ ȱ ȱ ȱ ȱ Ěȱ ȱ ȱ populations of the marine planktonic species Ȭĵȱȱ pungensȱǻ ȱ¡ȱǼȱ ȱǻ¢ et al., 2010), Skeletonema marinoiȱȱǭȱȱǻ ȱǭȱ ãǰȱŘŖŗŖǼȱȱDitylum brightwelliiȱ ǻǼȱ ȱ ǻȱ ȱ ȱ ȱ ȱ ȱ however unclear) (Rynearson et al., 2009), and of the freshwater species Sellaphora capitataȱȱǭȱȱǻȱet al., 2009)]. There are no reliable data available on the geographical distribution of terrestrial diatom species, but it is hypothesized that the biogeographical ranges of terrestrial diatoms would be larger ȱȱȱȱȱǻȱet al., 2010). Compared to lakes which are islands of suitable habitat within a hostile matrix, terrestrial habitats are presumably less fragmented on a regional scale, increasing the dispersal probabilities and dispersal rates of terrestrial species. Moreover, because terrestrial habitats ȱ £ȱ ¢ȱ ȱ ¢ȱ Ěȱȱ ȱ ȱ ȱ ȱȱǻȱet al.ǰȱŗşŞŗǲȱ ȱet al., 2008) compared to

General introduction 9 permanent lakes, it is assumed that terrestrial species are adapted to overcome these more hostile conditions [analogous to green ȱǻȱet al., 2008)], which would also increase their tolerance to Ȭȱȱȱȱǯȱ¢ǰȱȱ¢ȱ sediment is more susceptible to be picked up by air currents (Chepil, ŗşśŜǼǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ wind compared to lacustrine diatoms. The combination of the less fragmented terrestrial habitat, higher susceptibility for wind dispersal and higher tolerance for desiccation would enhance the dispersal probabilities of terrestrial diatoms compared to aquatic ȱǻǰȱŗşŜşǲȱȱǭȱ ǰȱŗşŝśǼǯȱȱ ȱȱȱǰȱȱȱȱĚȱ ȱȱȱȱȱ also result in lower speciation rates, and it is thus expected that terrestrial diatoms are less diverse than aquatic diatoms. As said, no ȱȱ¢ȱȱȱęȱȱȱȱȱ¢ǯ While it has been indicated that dispersal limitation does play a role in at least lacustrine and marine diatom distributions, it remains unknown on which scale and at what rate dispersal and ȱĚȱ ȱǰȱȱ ȱȱȱȱǽȱȱȱȱ morphological evolution of Stephanodiscus niagarae Ehrenberg to S. yellowstonensis ȱǭȱȱȱca. 4,000 years (Theriot et al.ǰȱ ŘŖŖŜǼǾǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱȱȱǰȱȱȱěȱȱȱȱȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ǯȱ Furthermore, while there are some clues on the general tolerances of diatoms during transport by air currents and animals (see section ŗǯŚǼǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ěȱȱȱȱǯȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ěȱ ȱ ȱ populations of a single species, two methods are currently available based: phylogeographical studies and microsatellite analysis. These are based on the genetic information stored in individual genomes, and on the fact that, independently of environmental selection, neutral mutations will accumulate with time in the genomes (called “genetic ȄǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ genetic drift through the exchange of genetic material. By studying ȱ¢ȱȱȱǻěȱȱȱȱȱȱǼȱ ȱ and between populations one will gather an approximation of the amount of dispersal between these populations. When analyzing this in an explicit geographical framework as done in phylogeographical studies, one will gather information on both the amount and

10 General introduction ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ǯȱ ȱ ȱ ȱȱȱȱȱȱȱĚȱ ȱ ȱȱ ȱȱȱȱǰȱȱȱȱǻȬ coding sequences of 1 to 6 base pairs repeated 10 to 100 times in Ǽȱȱȱǻȱȱ¢ǰȱȱȱ ȱȱȱȱȱȱȱȱȱǼȱ have to be analyzed.

Phylogeographies ȱ ȱ Ȭǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ǰȱ phylogeographical studies will evaluate more recent divergences between and within species in an explicit geographical framework ȱȱȱěȱȱȱȱǻǰȱŘŖŖŖǼǯȱ¢ȱ ȱǰȱȱȱȱęȱȱȱȱȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ǯȱȱ¢ȱȱȱ¡ȱȱȱȱȱ the colonization routes after the last glacial maximum 10,000 years ago from the glacial refugia to the present locations (e.g. King & ǰȱ ŗşşŞǲȱ ȱ ǭȱ ǰȱ ŘŖŖŘǲȱ ȱ et al., 2006). For diatoms, phylogeographical studies are scarce and limited to the marine planktonic diatoms Skeletonemaȱ ȱǻ ȱet al.ǰȱŘŖŖśǼȱȱ ȱ Ěȱȱǻ Ǽȱ¢¡ȱǭȱ ȱǻȱet al., 2010).

Microsatellite studies ȱȱȱȱȱĚȱ ȱ ȱ¢ȱȱ populations, microsatellite frequencies can be analyzed. For ǰȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ǽPseudo- ĵȱȱ (Casteleyn et al., 2010), Skeletonema marinoiȱǻ ȱ ǭȱ ãǰȱ ŘŖŗŖǼǰȱ ȱ ȱ Ditylum brightwellii (Rynearson et al., 2009) in which the species boundaries were however unclear] and in a single freshwater species [(Sellaphora capitata (Evans et al., 2009)] reveal that gene exchange is restricted in more or lesser extent between geographically separated populations on even small geographical scales, while a study on P. pungens on smaller scale ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ Ěȱ ȱ ȱ populations (Casteleyn et al., 2009).

General introduction 11 1.3 Species diversity and speciation 1.3.1 Species as fundamental units of biology ȱ ȱ ȱ ȱ ęȱȱ ȱ ¢ǰȱ ȱ ¢ǰȱ species are used as basic biological study unit. A species (or lineage) is a group of individuals structured as a metapopulation that ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǻȱ £ǰȱ ŘŖŖŝǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻȱ Ǽȱ ȱ ȱ ǻǯȱ ŗǯŘǼǯȱ ȱ ȱ ǰȱ ěȱȱ ȱ ȱ ȱ acquired by these evolving lineages: the lineages will become ¢ȱǰȱ¢ȱěȱǰȱ¢ȱ incompatible, and genetically reciprocally monophyletic, amongst ǯȱȱȱȱȱȱěȱȱȱȱȱǻȱ operational criteria) for species delimitation, and were traditionally ȱ ȱ ěȱȱ ȱ ȱ ǰȱ ȱ the morphological, ecological, biological and phylogenetic species concepts. It is important to clearly distinguish between the theoretical concept of species (i.e. separately evolving metapopulation lineages) and the operational criteria (morphology, ecology, genetics) used ȱ ¢ȱ ȱ ȱ ǯȱ ȱ ȱ ěȱȱ ȱȱȱȱěȱȱȱȱȱǰȱȱ is a grey zone (Fig. 1.2) between the initial lineage separation and the fully completed lineage separation during which alternative ȱȱȱȱĚȱȱȱȱ ȱ ȱȱ a single species are present. On each outer side of this grey zone, there will be unanimous agreement on the number of species. For ȱ ǰȱ ȱ ěȱȱ ȱ ȱ ǰȱ ȱ ¡ȱ morphological, genetic and physiological data, will enhance the probability to delineate the species correctly. In diatom taxonomy, valve morphology is traditionally the main basis on which species are delineated, to a lesser extent supplemented with cytological data (Round et al., 1990). Consistently ¢ȱěȱȱȱȱȱȱȱ into “morphospecies” as an approximation of the real species (as ȱ¢ȱǯȱȱȱȱ¢ȱǼǯȱ ȱȱǰȱ this morphological or typological concept is predominantly used in several study areas, including ecology and biogeography. Recently, ȱ¢ȱȱȬȱȱȱȱȱȱ of nucleotide information as additional line of evidence for species delimitation. Phylogenetic analyses of this genetic information will group strains into lineages depending on the nucleotide composition

12 General introduction Fig. 1.2 Diagram visualizing lineage separation and divergence (speciation) and species concepts (after de Queiroz, 2007). A single lineage (species) splits to form two lineages (species). The horizontal lines labeled SC (species criterion) 1 to 9 represent the times at ȱȱȱȱěȱȱȱǻ¢ȱ¢ǰȱ¢ȱ incompatible, ecologically distinct, etc.). In the grey zone, alternative species concepts come ȱĚȱǯȱȱȱȱȱȱ¢ȱ£ǰȱȱ ȱȱȱȱȱȱ number of species.

General introduction 13 and visualize this on a phylogenetic tree in which branch lengths are proportional to the number of nucleotide substitutions (Felsenstein, 2004). Individuals with minor variation in nucleotide composition will cluster together, while individuals which vary in nucleotide composition will be placed on separated branches, the branch lengths ȱȱȱȱȱǯȱȱȱĴȱȱ occur using independent sources of genetic data (for example chloroplast genes vsǯȱ ȱ ȱ ȱ ǰȱ ȱ ěȱȱ genes from a single genome source), the probability to correctly delineate species increases, and the same holds for morphological data.

1.3.2 Diatom diversity and cryptic species ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ǰȱ ȱȱȱŘŖŖǰŖŖŖȱ¡ȱȱǻȱǭȱǰȱŗşşŜǼǯȱ They thrive in a wide range of habitats and ecological conditions, which all contain a whole range of species resulting in a high overall species diversity. The unique characteristics of the diatom genome, combining genes present in bacteria, higher plants and higher animals next to their own set of genes, is hypothesized to result in a high adaptive potential (Bowler et al., 2008), which might facilitate ecological adaptation and subsequent speciation. While the causal relationship between ecological change and speciation is still uncertain, it has been shown that closely related (pseudo) cryptic diatom species can be ecologically diverged, as seen for ¡ȱȱěȱȱȱ¢ȱȱȱȱNavicula phyllepta ûĵȱ ȱȱ¡ȱǻȱet al.ǰȱŘŖŖşǼȱȱěȱȱ in temperature optima in the Cylindrotheca closterium (Ehrenberg) ȱ ǭȱ ȱ ¡ȱ ǻȱ et al., unpubl.). The morphological evolution from Stephanodiscus niagarae Ehrenberg to the endemic S. yellowstonensisȱȱǭȱȱ ȱȱ with progressive warming (Theriot et al., 2006). Ecological niche ěȱȱȱȱȱȱȱȱȱȱȱ and further account for the high diversity. The high number of estimated diatom species is partly a ȱȱȱȱȱ¢ȱȱǻȱǭȱǰȱŗşşŜǼǯȱȱ ȱǰȱȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱ ȱȱęȱȱȱȱȱěȱȱ ȱǻǼȱȱȱ ȱ¢ȱěȱȱǽe.g. Skeletonema costatumȱ ǻ Ǽȱ ȱ ǻȱ et al.ǰȱ ŘŖŖśǼǰȱ Cyclotella meneghinianaȱ ûĵȱ ȱ ǻ£ȱ et al., 2007), Sellaphora pupula

14 General introduction ǻ ûĵȱǼȱ¢ȱǻȱǭȱǰȱŘŖŖŝǼǰȱNavicula phyllepta ûĵȱ ȱ ǻȱ et al., 2009), Eunotia bilunaris (Ehrenberg) ȱǻȱet al., 2008), ĵȱȱȱǻ ûĵȱ Ǽȱ ȱǻȱet al., 2009)]. While until now cryptic species have only been reported for a limited number of diatom morphospecies, it is expected that this is a widespread phenomenon in diatoms (Mann, 1999).

1.3.3 Speciation over time and space As discussed for macrobiota (see section 1.2.1), speciation and ¡ȱ ȱ ȱ ȱ ȱ ȱ ¢ǯȱ ȱ rates, periods and locations will therefore give further insight into the ȱȱȱ ǯȱ ȱȱȱȱ ȱȱȱ¢ǰȱȱȱȱȱȱĚȱȱ are often very well dated and can therefore be linked to dated ȱǯȱȱǰȱȱȱȱ Ȭȱ fossil sequence showing the abrupt morphological evolution from Stephanodiscus niagarae Ehrenberg to the endemic S. yellowstonensis ȱǭȱȱȱȱȱȱȱŚǰŖŖŖȱ¢ȱȱȱ with progressive warming following the retreat of the glaciers 10,000 years ago (Theriot et al., 2006). Because the discovery of such Ȭȱȱȱ¡¢ȱȱȱȱȱ¢ȱ applicable, other methods have been devised to estimate the time ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ often coding for very conservative proteins or having no function in protein coding, which are assumed not to be under ecological ȱȱȱ¢ȱ¢ȱȱȱȱǯȱȱȱ will evolve over time and will further diverge between species due to genetic drift, resulting in less variation between recently evolved species and more variation between phylogenetically distant ǰȱ ȱȱȱȱĚȱ ȱ ȱȱ ȱȱǻȱ drastically reduce) this divergence. ȱȱȱȱȱȱȱȱȱǰȱ these can evolve slowly, thus giving more information on the deeper and older divergences (e.g.ȱŗŞȱȱȱȱǼȱ for example between larger groups of species or genera; or evolve faster and being of no use for deeper divergences but only for more recent speciation events (e.g.ȱ ŘŞȱ ǰȱ rbcL, coxŗǰȱ Ǽǯȱ ȱ genetic markers evolve even faster and point mutations occur also within species (e.g. ȱ ȱ rbcL), these can be used to

General introduction 15 trace variations between geographically separated populations of a ȱǯȱȱǰȱȱȱȱȬȱȱ ȱȱ ǰȱȱěȱȱȱȱȱȱȱǯȱȱ¢ȱ ȱȱȱȱȱȱȱĚȱ ȱ ȱȱ ȱȱȱȱǰȱȱȱȱǻȬ coding sequences of 1 to 6 base pairs repeated 10 to 100 times in Ǽȱȱȱǻȱȱ¢ǰȱȱȱ ȱȱȱȱȱȱȱȱȱǼȱ have to be analyzed (see section 1.2.2).

Time-calibrated molecular phylogenies ȱȱȬǰȱȱȱȱȱ ȱȱȱ between larger groups of species can be investigated by combining ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ¢ǯȱ ȱ Ȭȱ ¢ȱ ȱ as a starting point to interpret species divergences in a geological ȱ ǻȱ ǭȱ ǰȱ ŘŖŖŝǼǯȱ ȱ ¡ǰȱ ȱ ȱ marine planktonic foraminifer Neogloboquadrina Bandy, Frerichs & ǰȱȱȬȱ¢¢ȱȱȱȱȱȱ ěȱȱȱ ȱȱȱȱ ȱ¢ȱ species in the Arctic and Antarctic were correlated with changes in ice masses, but that exchange between the two poles still occurred ȱȱȱŘŖŖǰŖŖŖȱ¢ȱǻȱǭȱǰȱŘŖŖŞǼǯȱȱǰȱ such explicit assessments of geological timeframes have only been done for the larger taxonomic groups of diatoms (Kooistra & Medlin, 1996; Medlin et al.ǰȱ ŗşşŝǲȱ ǰȱ ŘŖŖŝǲȱ ȱ ǭȱ ǰȱ 2010) trying to estimate the geological processes triggering the ęȱȱȱȱȱȱȱȱȱ ǰȱȱȱȱȱ marine planktonic species [Ȭĵȱȱ (Casteleyn et al., 2010)] estimating the divergence within the P. pungens species complex during the Pleistocene glacial cycles.

1.4 Stress tolerance and dispersal limitation As summarized in section 1.1, dispersal is a key process controlling biogeographical ranges and knowledge on dispersal probabilities of organisms is thus indispensable to correctly interpret observed ȱĴȱǯȱȱȱ¢ȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ the medium or habitat they have to bridge (physical barriers,

16 General introduction heterogeneity, environmental factors, biotic interactions) but also on its morphological and physiological characteristics such as active locomotion or features enhancing passive motility, dormancy and tolerance levels for adverse conditions during dispersal. The stress tolerance of an organism is therefore an important factor determining its dispersal potential and subsequently its potential geographical range. However, as reviewed by Martiny et al. (2006), knowledge ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ęȱȱȱȱȱȱȱȱȱ¢ȱ lacking for microorganisms, and the same holds for diatoms. ȱȱȱȱ¡ȱȬȱȱǰȱȱȱ important to have a good knowledge on the dispersal mechanisms diatoms can use and subsequently deduce the problems diatoms ȱȱǯȱȱȱȱȱǰȱȱęȱȱ hypotheses can be put forward on dispersal tolerances of freshwater and terrestrial diatoms.

1.4.1 Freshwater diatom dispersal ȱ ȱ ȱ ȱ £ǰȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ macroorganisms, and this has to be taken into account. Most diatoms do not move actively. However, the raphid diatoms do move on smaller scales, but this movement is aimed to leave less suitable patches and search randomly for a more suitable patch with ȱ ȱ ȱ ǻǯȱ ǯȱ ǯȱ Ǽǰȱ ȱ ȱ higher light conditions (Admiraal et al., 1984; Perkins et al., 2010), or ȱȱȱǻǯȱǯȱ ǯȱ Ǽǰȱȱǯȱ¢ȱ ȱǰȱ¢ȱȱȱȱȱȱŗȬŘśȱΐȱȱȱǻǰȱ ŗşŝŗǲȱȱǭȱĴȱȬ ǰȱŗşŞŚǼǯȱ ǰȱȱǰȱȬȱ ȱȱȱ ȱȱěȱȱȱȱǯȱȱǰȱ over multiple generations a diatom species could eventually extend its range over larger scales by active movement, but by that time it is likely that genetic divergence from the source population has occurred, except if a second, faster dispersal mechanism has allowed ĜȱȱȱȱĚȱ ȱ ȱȱǯȱȱȱ is thus not likely to be the key dispersal mechanism of diatoms to extent their geographical range and this will mainly be achieved by passive dispersal. ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ¢ȱ a supportive medium (air currents, water currents or another organism) without consuming itself additional energy for movement. However, the dispersed organism is entirely dependent

General introduction 17 on the transport medium for the rate, direction, distance and point ȱȱȱȱǻ ȱǭȱ ǰȱŘŖŖŗǼǯȱȱȱȱȱ ȱ of the passively dispersed organism the transport is random, but in some cases the movement of the medium is less random than ¡ǰȱȱȱȱȱȱǰȱȱ ȱĚȱ ȱ or migration routes of animals, and will largely determine the main dispersal routes of the passively dispersed organisms. As such, the distribution of some passively dispersed zooplankton ȱȱȱȱȱȱȱȱȱ ȱ ȱǻ ȱǭȱǰȱŘŖŖśǼǯȱ ǰȱȱȱȱ scales, passive dispersal will be random. Traditionally, short and long distance dispersal has been distinguished, using a more or less arbitrary distance as boundary, for example 10 km as boundary ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱ ŘŖŖśǼǰȱ ȱ ęȱȱ ȱ as respectively dispersal inside a habitat type and dispersal over ěȱȱȱǻ ȱǭȱ ǰȱŘŖŖŗǼǯȱ ǰȱȱȱ ȱ on the distances over which diatoms are transported generally, nor on the heterogeneity encountered by a diatom cell inside a habitat ¢ǰȱȱȱ¢ȱȱęȱȱȱȱȱȱȱ diatom dispersal.

1.4.2 Dispersal vectors and dispersal-related stress factors Three types of transport vectors account for the passive dispersal of diatoms, being water currents, air currents and living organisms (Kristiansen, 1996).

Dispersal by water currents Freshwater currents in rivers or as ground water, and oceanic currents, the thermohaline circulation and other water currents all transport algae passively, including diatoms (Finlay & Esteban, ŘŖŖŚǼǯȱ ȱ £ȱ ¡ȱ ȱ ȱ ęȱȱ ȱ ȱ connected to an existing pond showed that over short time periods the same species occured as in the source pond, while new species, transported over land, only occurred after several years (Atkinson, 1988). Water currents are the fastest way of dispersal for algae, but are constrained by the position and availability of sources and destinations (Kristiansen, 1996). Of course – and this applies to all ȱȱȮȱěȱȱȱȱ¢ȱȱ ȱȱ

18 General introduction propagule arrives alive in a suitable patch and is able to divide and start a new population.

Dispersal by organisms ȱ ¢ȱ ȱ ȱ ȱ ¢ȱ ȱ ǯȱ Macroalgae and macrophytes will transport the cells during their passive dispersal through water currents and animals. Actively ȱ ȱ ȱ ęȱǰȱ ǰȱ ȱ ȱ ȱ can transport the cells internally (endozoochory) or externally (epizoochory). For freshwater habitats several organisms have been shown to contain living diatoms in or on their bodies. Fish guts ȱ ȱ ȱ ȱ ǻǰȱ ŗşŚŖǼǰȱ ȱ ȱ ȱ is restricted to the watercourse. Less restricted but still over short distances is external transport on water beetles (Migula, 1888), while ȱȱȱȱǻȱȱŞśŖȱǼȱȱȱ¢ȱ ȱ ¡¢ȱ ȱ ȱ Ěȱȱ ǻȱ et al., 1966) and aquatic ȱ ǻ ··Ȭǰȱ ŗşřŞǼǯȱ ǰȱ ȱ ȱ ȱ dispersers of freshwater algae are waterfowl (Kristiansen, 1996; ȱǭȱǰȱŘŖŖśǼȱ ȱȱ¢ȱȱȱ¢ȱȱ ŘŖȬřŖȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱŘŖŖśǼǯȱ¡¢ǰȱȱȱȱȱȱȱȱ ǰȱȱȱȱȱȱǻǰȱŗşŜŖǲȱȱ ȱǻŗŞŞŞǼȱ in Kristiansen, 1996), but algal cells having slime sheets or forming ȱȱ¢ȱȱ¢ȱȱȱȱǻǰȱŗşŜŖǼȱ because desiccation during external transport limits the survival of ȱȱǻȱǭȱ ǰȱŘŖŖŘǼǯȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ǻǰȱ ŗşśşǼǰȱ ȱ ȱ diatoms already die when the moisture content of the surrounding ȱ ȱ ȱ śŖƖȱ ǻǰȱ ŗşśşǼǰȱ ȱ ȱ ȱ ȱ rarely longer than 30 minutes on the feet of ducks, decreasing the ¢ȱȱȱȱȱ¢ȱȱ ¢ȱǻǰȱŗşŜŖǼǯ ȱȱěȱȱ ȱ ȱ ¢ȱ ǰȱ ȱ ȱ through the guts should be survived, which depends on the physical, chemical and bacterial components of the birds gut, but also on the thickness and composition of the cell wall of the dispersed cell ǻ ǰȱ ŗşşŜǲȱ ȱ ǭȱ ǰȱ ŘŖŖŘǲȱ ȱ ǭȱ ǰȱŘŖŖśǼǯȱȱȱȱȱȱȱȱȱȱ guts lies however much lower than the amount of algae carried ¡¢ȱǻǰȱŗşŜŖǼȱȱȱȱęȱȱȱȱȱ only a single survived digestion experiments (Atkinson, 1971). However, internal transport could occur over longer distances than

General introduction 19 external transport (Proctor, 1966), and living Asterionella Hassall has been found in faeces two hours after feeding, corresponding to a ĚȱȱȱŘŘŖȱȱǻǰȱŗşŞŖǼǯȱ Also humans disperse microalgae and diatoms between ȱȱȱȱȱ¡ȱęȱȱȱȱǻ¢ȱ ǭȱ ǰȱ ŗşŞŞǼȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ £ȱȱǻǰȱŗşśŗǼǯȱ¢ǰȱȱȱȱ indications that diatoms are internally and externally transported by various organisms, but knowledge on the frequency and distance of transport (Figuerola et al.ǰȱŘŖŖśǼǰȱȱȱȱȱȱȱ ȱȱȱěȱȱȱǻȱǭȱǰȱ 2002) is extremely scarce.

Dispersal by air currents In general it is assumed that wind transport plays an important role for the dispersal of microalgae (Round, 1981; Chalmers et al.ǰȱ ŗşşŜǲȱ ǰȱ ŗşşŜǼǯȱ ǰȱ ȱ ȱ Ĵȱȱ ȱ ȱ predominantly terrestrial algae, including cyanobacteria, green algae and to lesser extent terrestrial diatoms, are transported through the air, while there is uncertainty on the frequency, distance and extent of aerial transport. Concerning diatoms, it is already longer known ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ ǰȱ ŗŞŚśǲȱȱǭȱǰȱŗşřśǲȱ¢ǰȱŗşŜŞǲȱ et al., 1996), ȱȱȱȱȱȱȱȱȱǯȱȱ ȱǻŗşřŝǼǰȱȱǻŗşŜŗǰȱŗşŜŚǼǰȱ ȱet al.ȱǻŗşŜŚǼȱȱ¢Ȭ Ocotla & Carrera (1993) assessed the presence of living microalgae and found predominantly cyanobacteria and coccoid green algae, besides some sporadic diatom cells from the genera Navicula Bory ȱȬǰȱ ĵȱȱ ǰȱĵȱ Hassall and Melosira Agardh. Their conclusions are that principally green algae such as Chlorococcum Meneghini, Chlorellaȱ ȱ ȱ Klebsormidium ǰȱĴȱ¡ȱǭȱ ȱȱȱ¢ȱ ȱ ȱȱȱ ȱȱȱǻȱǰȱŗşřŝǲȱǰȱŗşŜŗǲȱ ȱ et al.ǰȱŗşŜŚǲȱǰȱŗşŜŚǲȱ¢ȬȱǭȱǰȱŗşşřǼǰȱȱȱ mainly terrestrial taxa are recovered (Brown et al.ǰȱŗşŜŚǲȱ¢Ȭȱ ǭȱǰȱŗşşřǼǯȱ ǰȱ ȱǭȱ ěȱȱǻŗşŜŜǼȱǰȱȱ the typical green algae, a large amount of living diatom species in their air traps, qualitatively and quantitatively corresponding to the nearby situated river Havel. These data contradict the previous opinion that diatoms are not common in air currents, and that only a limited amount of taxa are found, but no data are given on the

20 General introduction Ȭǰȱȱ¡ȱȱȱȱȱȱȱ ȱǯȱ ȱȱȱ¢ǰȱȱȱȱŗȱȱŘȱȱȱȱ m³ are found, which is not much compared to the on average 142,3 pollen per m³ (Brown et al., 1964; Tormo et al., 2001). ȱ ȱȱ¢ȱȱȱȱȱȱȱ ȱȱ ǻ ȱ ǭȱ ǰȱ ŘŖŖŗǼǯȱ ȱ ȱ ȱ ȱ ȱ humidity, wind velocity and wind shearing, air temperature, atmospheric pressure and the occurrence of turbulences or eddies ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ǰȱ ȱ period and distance it will remain in the air and move, and the moment of landing into a new patch. As dry soils are more wind ȱȱ ĴȱȱȱǻǰȱŗşśŜǼǰȱȱȱȱȱ predominantly terrestrial algae are found in the air, and that the highest abundances of dispersed algae are found in dryer periods ǻȱ ǰȱ ŗşřŝǲȱ et al., 2001). Besides meteorological conditions, the intrinsic features of the transported organism will determine the ease to be picked up and transported by air currents (such as size, morphological characteristics, local population abundances and habitat) and the capacity to survive the stresses ¡ȱȱȱȱȱȱȬǰȱȱ ȱ¡ȱȱǻǰȱŗşŜŗǲȱ¢ǰȱŗşŞŜǲȱȱ ǭȱǰȱŗşŞŝǲȱ¢ȬǰȱŗşşŗǼǯȱ¢ǰȱȱ and cytological characteristics will enhance survival probabilities, ȱȱȱȱȱȱȱǻǰȱŗşŞŝǼǰȱȱȱ in concentrations of pigments (Imshenetsky & Kondrateva, 1987) or compatible compounds (Welsh, 2000), or the formation of resting ȱǻ ȱǭȱǰȱŗşŝśǲȱ ěȱǰȱŗşşŜǼǯ

1.4.3 Dispersal rates and probabilities ȱĚȱȱȱȱȱȱȱȱȱ only on its prevalence, but also on its rate. To be able to overcome ȱȱȱȱȱǰȱȱĚȱ ȱ ȱȱ must be high enough to overcome genetic drift and selection. It is thus possible that very low levels of dispersal are not able to ȱ ȱ ěȱǯȱ ȱ ȱ ȱ ȱ ȱ ȱ local population abundances in the colonized patch, but also on the colonization and competition abilities of the colonizing cells. ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ dispersal. However, based on the literature we can estimate their ȱ ȱ ǰȱ ȱ ȱ ęȱǰȱ ȱ ȱ ȱ ȱ ěȱȱȱǯȱ

General introduction 21 First, terrestrial habitats are assumed to be less fragmented ȱȱȱǰȱȱȱ¢ȱȱěȱȱ ȱȱȱǯȱǰȱ¢ȱȱȱ are transported by air currents, whereas lacustrine diatoms will be transported by waterfowl, but the presence of living diatoms is for both vectors quite low. Based on the knowledge that diatoms are not very commonly transported by waterfowl and air currents (see above), we can assume that their dispersal rates will in general not be extremely high. Third, dispersal probabilities decrease with ǰȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ǯȱ ȱ ȱ assumptions, we can hypothesize that for both lacustrine and ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ Ĝȱȱ¢ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ and prevent the formation of geographical structure, while dispersal over short distances does very likely occur. For terrestrial diatoms ȱ ȱ ȱ ¢ȱ Ȭǰȱ ȱ ¢£ȱ ȱ ¢ȱ will disperse farther away than lacustrine diatoms due to the less fragmented terrestrial habitat and the occurrence of aerial dispersal, albeit not on a world wide scale, creating thus geographical structure on a somewhat broader scale.

1.5 Aims and thesis outline The overall aim of this study is to improve our understanding of ȱȱȱȱȱȱȱȱĴȱȱ in lacustrine and terrestrial diatoms. This is done by using multiple lines of evidence, including the analysis of large and regional ȱ ¢ȱ Ĵȱǰȱ Ȭȱ ¢ȱ ȱ ȱ evidence, and ecophysiological studies. Our primary study region ȱ ȱ ȱ ȱ Ȭȱ ǯȱ ȱ ȱ ȱ ¢ȱ Ȭȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ geographical isolation of many habitats, this region is amenable to ȱǰȱȱȱȱȱȱǰȱȬ ȱȱȱ¢ȱĴȱȱȱ¢ȱ ȱȱȱ ȱ ȱ ¢ȱ Ȭȱ ȱ ȱ ȱ ¢ȱȱȱȱȱ£ȱȱǽŜśǯśȱȱ¢ȱǻǼȱȱ till present] during which the pennate diatoms radiated.

22 General introduction The thesis is organized in an introduction, three parts with our main research results, and a general discussion:

INTRODUCTION ǐȱChapter 1 gives a general introduction on biogeography and microbial biogeography, and focuses further on diatom biogeography, species diversity and the interaction of dispersal probabilities and stress tolerance.

PART 1: PATTERNS IN DIATOM DISTRIBUTION ǐȱIn Chapter 2 we combine analyses of biogeographical Ĵȱȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ Ȭ calibrated molecular phylogeny to provide evidence that adaptive radiations through allopatric speciation, survival in glacial refugia, and regional extinction operate in microorganisms at similar spatial and temporal scales as observed in macroscopic organisms. ȱęȱȱ¢ȱȱȱȱȱȱȱ biogeography that diatoms are not limited in their dispersal and consequently do not have biogeographies. ǐȱIn Chapter 3 we examine if terrestrial and lacustrine diatoms ȱ ěȱȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ȱ scale, using distance decay analysis of community similarities and ȱȱȱȱ ȱȱȬȱȱ £ȱȱ ȱȱȱȱȱ ȱǰȱȱȱ ȱȱȱȱȱȱǯȱ¢ȱȱ decrease over space for both terrestrial and aquatic habitats, but no ęȱȱ ěȱȱ ȱ ȱ ȱ ǯȱ ¢ȱ ȱ ȱ ȱǰȱ ȱǰȱȱȱȱȱęȱ¢ȱ lower turnover between sites, suggesting higher dispersal rates and/ or broader niches for terrestrial diatoms.

PART 2: SPECIATION IN TIME AND SPACE ǐȱTo further explore the importance of time in speciation events, ȱ ȱ ȱ ȱ Ĵȱǰȱ ȱ ȱ ȱ Chapter 4ȱȱȬȱ¢¢ȱȱȱȱȱPinnularia Ehrenberg, and calibrate it in time using fossil material. Our analysis ȱȱęȱȱȱȱȱȱca. 60 Ma ago, 10 to řŖȱȱȱȱ¢ȱǯȱȱȱęȱȱȱřȱ ȱȱȱȱȬǰȱȱȱȱȱ ȱ ȱȱȱ¢ȱǰȱȱȱȱ clades have dispersed successfully worldwide.

General introduction 23 ǐȱȱ Ȭȱ ¢¢ȱ ȱ ȱ ȱ Pinnularia from chapter 4 allows us to investigate the evolution of morphological characters and to compare the phylogenetic signal present in genetic and morphological data in Chapter 5. This is done by using ancestral state reconstruction of the morphological data and by performing cladistic analyses on the morphological data. While some ȱȱȱȱȱȱȱěȱȱ clades, other character states were restricted to a certain clade. The ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ relationships, but only the genetic data are able to fully resolve these relationships. However, similar clades are recovered, indicating that a similar phylogenetic signal is present in both datasets. ǐȱIn Chapter 6 we select the cosmopolitan terrestrial morphospecies Pinnularia borealis Ehrenberg and ĵȱȱ amphioxysȱǻǼȱ ȱȱȱȱȱȱȱ ěȱȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ the plastid marker rbcȱȱȱȱȱŘŞȱǯȱȱȱ the divergences of P. borealisȱȱȱȱȬȱȱȱ ȬȱȱȱȱPinnulariaǯȱȱȱȱěȱȱ ȱ ȱ ȱ ȱ ¢ȱ ěȱȱ ȱ for both taxa. The results indicate that in both taxa the Antarctic ȱ ȱ¢ȱȱ¢ȱěȱǰȱȱ the occurrence of allopatric speciation on the Antarctic.

PART 3: TOLERANCE LEVELS FOR DISPERSAL-RELATED STRESSES ǐȱIn Chapter 7 we investigate for a selection of benthic freshwater and terrestrial diatom strains the tolerance levels of ȱ ȱ ȱ ȱ Ȭȱ ȱ ǰȱ ¢ȱ desiccation and extreme temperatures. We observe a very high sensitivity of both aquatic and terrestrial diatoms for desiccation, while terrestrial diatoms are more tolerant for extreme temperatures. ǐȱBecause resting stages are in many organisms the dominant dispersing stages, we analyze in Chapter 8 the tolerance levels of diatom resting cells for desiccation and freezing, and this for a selection of taxa occurring over a gradient from permanent aquatic habitats to terrestrial soils. While all taxa formed resting stages, only those from species that occurred mainly or exclusively outside water bodies were tolerant for desiccation and freezing, showing that terrestrial diatoms are adapted to their environment, and increasing their dispersal probabilities compared to aquatic diatoms.

24 General introduction GENERAL DISCUSSION ǐȱIn Chapter 9 the results of the previous chapters are discussed in the general framework of biogeography and dispersal probabilities.

General introduction 25

Part 1. Patterns in diatom distribution

Part 1: Patterns in diatom distribution 27

2. Poles apart: Interhemispheric contrasts in diatom diversity driven by climate and tectonics

Elie Verleyen1, Bart Van de Vijver2ǰȱȱěȱ1, Koen Sabbe1, Dominic A. Hodgson3, Linda Nedbalová4, Pieter Vanormelingen1, Dermot Antoniades5 and Wim Vyverman1

1 ǐȱ Ghent University, Protistology and Aquatic Ecology, Krijgslaan 281 S8, 9000 Gent, Belgium 2 ǐȱ National Botanic Garden of Belgium, Domein van Bouchout, 1860 Meise, Belgium 3 ǐȱ British Antarctic Survey, Natural Environment Research Council, High Cross Madingley Road, CB3 0ET, Cambridge, UK 4 ǐȱ Charles University in Prague, Department of Ecology, ²¤ȱŝǰȱŗŘŞȱŚŚȱȱŘǰȱ£ȱ 5 ǐȱ ȱȱÇȱ¢ȱęȱÇȱȱęȱǰȱàȱ Limnología, Universidad de la República, Iguá 4225, CP11400, Montevideo, Uruguay

Ĵȱȱ Contribution of CS: part of the statistical analysis and molecular phylogeny

Interhemispheric contrasts in diatom diversity 29 Abstract ȱȱȱĚȱȱȱȱěȱȱ¢ȱȱȱȱ ȱȱ¢ȱǻĴȱȱǭȱǰȱŘŖŖřǲȱ¢ȱǭȱ ǰȱ ŘŖŖŝǼǰȱ Ěȱȱ ȱ ěȱȱ ȱ ȱ history and landmass connectivity (Barnes et al.ǰȱŘŖŖŜǼǯȱ ȱǰȱ ȱ ȱ ȱ ¢ȱ ǻ¢ǰȱ ŘŖŖŘǼȱ ȱ that Arctic and Antarctic ecosystems should harbour the same microbial communities, with allopatric speciation being rare as a result of the ubiquitous dispersal of microbes. Here we show that, ¢ȱȱȱǰȱĴȱȱȱ¢ȱȱȱȱ ȱȱĚȱȱȱȱȱȱȱȱ¢ȱǯȱ Antarctic communities are impoverished and imbalanced relative ȱȱȱǰȱȱȱ£ȱ¢ȱȱȱȱ ǰȱȱȱȱ¢ȱȱȱȱȱȱ taxa, an overrepresentation of terrestrial lineages, and a general paucity of globally successful genera. Comparison of contemporary ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ¢ȱȱęȱȱĴȱȱȱȱȱȱȱȱ extinction during Neogene and Quaternary glacial maxima, in combination with radiations through allopatric speciation in glacial refugia. We propose that processes controlling the distribution and ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ and temporal scales as those for macroscopic organisms, leading to ¢ȱȱȱĴȱǯ

30 Interhemispheric contrasts in diatom diversity 2.1 Introduction ȱĴȱȱȱȱ¢ȱȱȱ¢ȱ¢ȱ ęȱȱȱ¡ȱȱȱȱȱȱȱȱ ȱȱȱȱ¢ȱǻǰȱŘŖŗŖǼǯȱȱȱ particularly evident in polar terrestrial biota, where interhemispheric ěȱȱȱȱ¢ȱǻȱet al.ǰȱŘŖŖŗǼȱȱȱȱȱ ęȱȱȱȱȱĚȱȱȱȱĴȱȱȱ contemporary plant and animal biodiversity (Barnes et al.ǰȱ ŘŖŖŜǼǯȱ Most Arctic biota are thought to have evolved in the high mountain ȱȱȱȱȱ£ȱȱȱȱȱȱ ȱ¡ȱȱȱǻĴȱȱet al.ǰȱŘŖŖŖǼȱ¢ȱȱȱȱ ȱȱȱȱȱȱřȱȱȱǻȱet al.ǰȱŘŖŖŗǼǯȱ Other species originated during the Holocene, or, alternatively, are descendents from the forest biota that occupied the Arctic during ȱȱ¢ȱǻĴȱȱǭȱǰȱŘŖŖřǼǯȱȱȱȱ ice ages, many of these cold-adapted species probably proliferated in vast tundra populations further south or in high latitude glacial ȱǻ ·ǰȱŗşřŝǼǰȱ ȱ¢ȱȱȱȱȬ ȱȱȱǰȱ ȱȱȱęȱȱȱ ȱȱ¢ȱǻȱǭȱ¢ǰȱŘŖŖŞǼǯȱȱȱȱȱȱ high connectivity between landmasses of the Northern Hemisphere ȱȱȱȱȱȱǰȱȱ¡ȱȱȱȬ ȱȱǻĴȱȱǭȱǰȱŘŖŖřǼǯȱ In the Southern Hemisphere, by contrast, the isolated and fragmented nature of landmasses prevented such large-scale ȱȱȱȬȱǯȱȱȱ¢ȱ have revealed that taxa started to diverge into endemic lineages through allopatric speciation shortly after the isolation of Antarctica from the other continents of the southern hemisphere (Convey ȱ ǯǰȱ ŘŖŖŞǼǯȱ ¢ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ Ȭ Miocene cooling event (ca.ȱŗŚȱǼǰȱ ȱȱȱȱ ȱȱ advanced and full polar climate conditions were established (Lewis ȱǯǰȱŘŖŖŞǼǯȱȱȱȱȱ¢ȱȱȱȱ Ěȱȱȱȱȱȱȱȱǻ¢ȱǭȱǰȱŘŖŖŝǼȱ as restrictions placed on dispersal by continental isolation prevented £ȱȱȱȱȱȬȱȱȱ interglacial periods. ȱȱȱȱȱĚȱȱȱǰȱȱȱ of dispersal limitation, allopatric speciation, and long-term tectonic and climate dynamics have been considered to be of minor relevance in shaping contemporary polar microbial communities.

Interhemispheric contrasts in diatom diversity 31 Indeed, prevailing theory predicts that similar habitats at both poles should share the same microbial communities as a result of their ȱ ȱ ȱ ǻ¢ǰȱ ŘŖŖŘǼǯȱ ȱ ȱ factors are therefore believed to act as the major process structuring their metacommunities through lineage sorting (Van der Gucht et al.ǰȱ ŘŖŖŝǼȱ ȱ ȱ Ĵȱȱ Ěȱȱ ȱ ȱ changes and tectonic processes are predicted to be absent (Martiny et al.ǰȱŘŖŖŜǼǯȱȱȱȱȱȱȱȱ idea and revealed that dispersal limitation may lead to allopatric divergence of populations, which may ultimately result in the formation of new microbial species (Evans et al.ǰȱ ŘŖŖşǲȱ ¢ȱ et al.ǰȱŘŖŗŖǼǯȱȱȱȱȱȱȱȱ ¡ȱȱȱȱȱȱȱȱĴȱȱȱ diversity (Telford et al.ǰȱŘŖŖŜǲȱ¢ȱet al.ǰȱŘŖŖŝǼǰȱ¢ȱ structure (Verleyen et al.ǰȱ ŘŖŖşǼȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻĵȱȱet al.ǰȱ ŘŖŖŝǲȱ ȱ et al.ǰȱ ŘŖŖŝǼǯȱ Here we test if speciation promoted by geographical isolation is ȱȱȱȱȱȱęȱȱȱȱȱȱȱ ȱ ǯȱ ęȱ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ predominant allopatric speciation, the greater degree of geographic ȱȱȱȱ ȱ ȱȱĚȱȱȱǻǼȱȱ ȱȱȱĚȱȱȱ ȱȱȱ ȱȱȱǰȱȱǻǼȱȱȱȱȱȱȱȱ isolated areas.

2.2 Results and discussion ȱȱȱȱȱȱȱĚȱȱ ȱ¢ȱȱ for possible variation in diatom assemblages due to limnological ěȱȱ ȱ ȱ ȱ ȱ ǻǯȱ ŘǯŗǼǯȱ ȱ ȱ ȱ ȱ ęȱȱ ǰȱ ȱ ȱ ȱ Ěȱȱ ȱ ¢ȱ ȱ ȱ ȱ ǻŗŘřȱ ȱ ȱ Şřȱ Ǽȱ ȱ ȱ ȱȱ ȱ ǻŘŘřȱ ȱ ȱ śŜȱ ǰȱ ǯȱ ŘǯŗǼǯȱ ȱ ȱĚȱȱȱȱȱȱ ȱȱȱȱ ȱ and abundantly present in the majority of the freshwater systems ǯȱȱȱĚȱȱȱȱȱȱȱȱ to typically terrestrial genera and globally successful genera are ǯȱ ȱ Ȭ¢ȱ ȱ ȱ Ěȱȱ ȱ ȱ ¢ȱ ěȱȱ ȱ ȱ ca. 14ȱ ȱ ȱ £ȱ ȱ preserved in McMurdo Dry Valley deposits (Lewis et al.ǰȱ ŘŖŖŞǼǰȱ

32 Interhemispheric contrasts in diatom diversity Fig.2.1 Number of species per diatom genus in Arctic and Antarctic lakes sharing the same limnological properties, showing the highly impoverished and disharmonic nature of the ȱȱĚȱǰȱ ȱȱ¢ȱȱȱȱȱȱȱȱ ǯȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ ȱ successful genera.

Interhemispheric contrasts in diatom diversity 33 ȱ ȱ£ȱ¢ȱȱ¡ȱ¢ȱȱȱ ¢ȱȱȱ¢ȱȱĚȱǯȱ ȱȱȱȱȱȱȱȱĚȱȱ ȱ assessed using a dataset of 392 Antarctic and sub-Antarctic freshwater ȱȱȱȱȱȱȱȱȱ ȱ ȱ ǻǯȱ ŘǯŘǼȱ £ȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ¢ǰȱ ŘŖŖŝǼǯȱ ȱ ¢ȱ ȱ ȱ ȱ geographical structuring of diatom biodiversity is similar to that of ȱȱǻǯȱŘǯřǼǯȱǰȱȱȱ ȱȱ ¢ȱǻǯȱŘǯŘȱȱǯȱŘǯřǼȱȱȱȱȱȱ provinces showed a high degree of endemism (i.e. 43%, 35% and 51%, respectively, for Sub-Antarctica, maritime Antarctica and ȱǼǯȱ ȱ ȱ ¢ȱ ȱ ȱ Ĵȱȱ ȱȱȱȱǻ ȱǭȱ¢ǰȱŘŖŖŝǼǰȱ ȱ approximately 50% of the lichen, tardigrade and dipteran species are endemic to Antarctica, as are the majority of mites and springtails, and possibly all nematodes (Convey et al.ǰȱŘŖŖŞǼǯȱȱȱȱȱ ȱȱęȱȱȱȱȱȱĚȱȱȱȱȱ ¢ȱȱęȱȱȱ ȱȱȱȱȱ ȱȱ¡ȱǻŶƽŖǯśşşŞǲȱpƽŖǯŖŖŘǰȱƽŗřǲȱǯȱŘǯŚǼǰȱȱȱ with our second prediction, and by a time-calibrated phylogeny of the globally distributed morphospecies Pinnularia borealis. The molecular phylogeny demonstrates that continental Antarctic ȱȱȱȱȱȱȱȱȱŝǯŜŝȱȱȱ ȱȱȱȱǻǯȱŘǯŚǼǯȱȱȱȱȱ ȱȱ consequences. First, our morphology-based estimates of endemism ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ Ȭ phylogenetic analyses are applied to other presumably ubiquitous ¡ȱǻ£ȱet al.ǰȱŘŖŖŝǲȱȱet al.ǰȱŘŖŖşǼǯȱ¢ǰȱȱȱȱ ęȱȱȱȱȱ¡ȱȱȱȱȱȱȱ Antarctic springtails (Stevens et al.ǰȱŘŖŖŜǼȱȱȱǻȱet al., ŘŖŗŗǼǰȱȱȱȱȱȱȱȱǻȱet al., ŘŖŖŜǼǯȱ ȱ ęȱȱ ȱ ȱ ¢ȱ ȱ ¢ȱ have occurred simultaneously in non-related groups of Antarctic organisms, in response to tectonic and climate change and increasing habitat fragmentation resulting from the expanding ice sheets. ȱȱȱȱȱȱȱĚȱǰȱȱȱ levels of endemism and the timing of speciation all point to an ȱ¢ȱ¢ȱȱȱǰȱ£ȱ ¢ȱ ȱ ¡ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ genera, and selective survival of taxa in glacial refugia since the

34 Interhemispheric contrasts in diatom diversity Fig.2.2ȱȱȱȱȱ ȱ ȱȱěȱȱȱȱ ȱ ȱ ȱ ȱ ǻĴȱȱ Ǽȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ versus endemic taxa. Maritime Antarctic species occur on the western side of the Antarctic Peninsula down to ca. 72°S and on the eastern side north of 65° S, as well as in the South Shetland and South Orkney islands. Continental Antarctic endemics are restricted to all other areas and islands of Antarctica south of 65°S. Sub-Antarctic endemics occur on the islands or island groups between 46° and 55° S. Antarctic endemics s.l. are restricted to the Southern Ocean region but occur in at least two provinces.

Fig.2.3 Biplot of a principle component analysis of Hellinger transformed species data ȱ ȱ ěȱȱ ȱ the diatom communities of 392 lakes and the congruence with the biogeographical zonation observed in multicellular taxa. The ȱǻǼȱȱȱ¡ȱȱ the axes.

Interhemispheric contrasts in diatom diversity 35 Fig.2.4 Time-calibrated molecular phylogeny of the diatom morphospecies Pinnularia borealis showing a divergence of the continental Antarctic lineage (in boldǼȱȱȱ ȱ European lineage around 7.67 Ma [1.95-14.77 Ma, 95% Highest Posterior Probability ȱǻ ǼǾǰȱȱȱȱrbcȱȱȱȱǻŗȬŘȱǼȱȱ ǯȱ ȱ ȱ ȱ ěȱȱ ¢ȱ ǯȱ ȱ ȱ ȱ ȱ clones per lineage is given between brackets. Bars represent the 95% HPD intervals. The grey encircled nodes were time-constrained during the analysis.

36 Interhemispheric contrasts in diatom diversity mid-Miocene cooling event, followed by widespread allopatric speciation. Moreover, we have shown that these processes operate at similar spatial and temporal scales as in macroscopic organisms. ȱ ȱ ȱ ȱ ¡ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱȱ ȱȱȱȱ¡ǰȱ ȱȱȱȱȱ escaped glacial overriding were probably permanently covered with ȱȱǻ et al.ǰȱŘŖŖśǼǰȱȱȱȱȱȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱ¢ȱǻ¢ǰȱŘŖŖŘǼȱȱ have important implications for the response of polar microbial communities in the face of the accelerated rates of warming, which have already impacted on the primary productivity and functioning of lacustrine ecosystems (Quayle et al.ǰȱ ŘŖŖŘǼǯȱȱ ȱ ȱ are particularly prone to extinction because of their restricted distribution in relatively small geographic regions, stringent ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ ȱ ȱ ȱ to protect them against the introduction of alien species is highly ęȱǯȱ

2.3 Materials and methods Sample collection and diatom analysis ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ deepest part using a gravity corer. Diatom samples were collected ȱ ȱ ȱ śȱ ǯȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ¢ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱȱ ȱŖǯśȱȬȱŗȱȱȱȱĴȱȱ£ǯȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ŘŘȱ ǻřŖȱ ƖǼȱ ȱ ř ȱ ǻşśƖǼǯȱ ȱ ȱ ¢ǰȱ ȱ ȱ ȱ dried onto a glass coverslip, mounted in Naphrax® and studied ȱȱȱȱ ȱěȱȱȱǯȱ ȱ ȱ ȱ ¢ǰȱ ¡£ȱ ȱ ȱ ȱ ¢ȱ Ȭȱ ȱ ȱ ȱ ȱ ĴȱȬǰȱ ȱ ¡ȱ ȱȱ ȱ ȬŞŚŖȱȱȱŗśȱǯȱȱȱ ȱ¢ǰȱŗŖȱΐȱȱȱ¡£ȱȱ ȱȱ on formvar-coated copper slot grids. Grids were examined with a ȱ ȱŗŖŗŖȱȱȱŜŖȱǯȱ The dataset was developed by combining newly obtained ȱ¢ȱ ȱ¡ȱȱǻȱȱŘǯŗǼǯȱȱ¢ȱ diatom data consisted of enumerations of species relative abundances in surface sediment samples, each of which typically represents a

Interhemispheric contrasts in diatom diversity 37 ȱȱȱ ȱȱȱȱȱȱȱȱȱ habitats spanning several years. Between 400 and 600 specimens were counted per sample. Taxonomic consistency between datasets ȱȱȱȱȱȱ¡ȱ ȱȱěȱȱ analysts. When the identity of a species could not be determined with certainty it was considered to be a non-endemic. Hence the ȱȱȱȱȱȱȱȱ¢ȱȱȱ an underestimation. The geographic distribution of the diatom species was based on literature data provided with unambiguous illustrations and/or descriptions and our own observations. The endemic species were subdivided according to the biogeographical province in which they occur. For the bipolar comparison Eunotia and Gomphonema were excluded from the analysis because these genera are in need of taxonomic revision.

Time-calibrated molecular phylogeny To develop a time-calibrated molecular phylogeny for Pinnularia borealisǰȱ śŗȱ ȱ ȱ ŝȱ ȱ ȱ ǰȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ŗȬŘȱ ȱ ȱ ȱ ŘŞȱ ȱ ǻǼȱ and plastid gene rbcȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ et al. ǻĴȱǼǯȱRbcL was partitioned into codon position (1st +2nd Ǽǲȱȱ ȱȱȱȱȱǯȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ şśƖȱ ȱ ȱ ¢ȱ ǻ ǼȱȱȱȱȱȬȱȱȱPinnularia ¢¢ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻěȱȱ et al.ǰȱ ĴȱǼǯȱ ȱ ȱ ȱ ȱ ǻ¢ȱ ȱ ȱ ǯȱŘǯŚǼȱȱȱȱȱȱȱȱȱşśƖȱ HPD. The outgroup consists of representatives of the most closely related Pinnulariaȱȱȱȱ¢ȱȱęȱȬȱȱ phylogeny for the genus Pinnulariaȱǻěȱȱet al.ǰȱĴȱǼǯ Ȭȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱȱŗǯśǯŚȱǻȱǭȱǰȱŘŖŖŝǼȱȱȱ ȱƸȱ̆ȱ ǻŚǼȱȱȱȱȱȱȱ¡ǰȱȱȱ ȱȱȱȱȱǯȱȱȱ¢ǰȱȱ started by a UPGMA tree, were run for 100 million generations and sampled every 1,000th generation. Burn-in values of 30 to 40% ȱȱȱȱǯŗǯśȱǻȱǭȱǰȱŘŖŖŝǼǯȱȱ post-burn in trees were combined after which the maximum clade credibility chronogram with mean node heights was calculated using TreeAnnotator v.1.5.4.

38 Interhemispheric contrasts in diatom diversity Statistical analysis ȱ ȱ ȱ ¢ȱ ȱ £ȱ ȱ ȱ ȱ ȱ ǻ ǰȱ ¢ȱ ȱ ȱ Ǽȱ ȱ ȱȱȱȱȱȱȱ ȱȱȱ ȱǻǯȱŘǯŗǼǯȱȱȱȱȱ ȱȱȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ analysis which resulted in the selection of 83 Antarctic and 56 Arctic ȱ ǯȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ transformed relative abundance data was used to assess the ȱ£ȱȱȱȱĚȱȱȱȱȱǯȱ All ordinations were performed in Canoco 4.5 for windows (Ter ǰȱŘŖŖŘǼǯ

Acknowledgements The Australian Antarctic Division, the British Antarctic Survey, the French Institute for Polar Research, the Japanese Institute for Polar Research, the Polar Continental Shelf Program and the South African National Antarctic Program provided logistic support. Louis Beyens, ȱ ǰȱ ȱǯȱ ǰȱ ȱǰȱȱ ǯȱǰȱ ȱǯȱǰȱȱ ǰȱȱȱ ȱȱȱ for providing samples and/or help during sampling campaigns. ȱ ǯȱ ǰȱȱǰȱȱȱȱ ȱȱ ȱ ȱ ǯȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǯȱǰȱǰȱȱȱȱȱ¢ȱȱȱȱęȱȱȱ ȱ ǻǼǰȱ ǯȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱȱ ȱȱȱȱȱ ǯŖśřřǯŖŝǯȱ

Interhemispheric contrasts in diatom diversity 39 6XSSOHPHQWDU\¿JXUHV

Fig. S2.1 Principal component analysis of standardized and centred environmental ȱǻ¢ǰȱ ǰȱƸǰȱ ƸǰȱŘƸǰȱȬǰȱǼȱȱȱȱǻǼȱȱȱ ǻǼȱȱ ȱǯȱȱȱȱȱȱȱȱĚȱȱěȱȱ ȱȱȱȱǻǯȱŘǯŗǼȱȱȱ¢ȱȱȱȱ£ȱ¢ȱȱ ȱȱȱęȱȱ¡ȱ ȱȬŗȱȱŖǯŞǰȱȱȱȱȱ¡ȱ ȱŖǯŝȱȱȬŖǯśǯ

Fig. S2.2 Species accumulation curves showing the total number of ȱȱǻǼȱȱǰȱ ȱ ǻǼȱ ȱ ȱ lakes.

40 Interhemispheric contrasts in diatom diversity Fig. S2.3 Species accumulation curves showing the total number of ȱ ȱ ǻǼȱ Ȭǰȱ ȱ ǻǼȱȱǯ

Fig. S2.4ȱĴȱȱȱ ȱȱȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱǻȱ¡Ǽǯ

Interhemispheric contrasts in diatom diversity 41 Supplementary table Table S1 References to the diatom datasets used. * Indicate datasets which were newly developed for the present study † Indicate datasets for which the taxonomy was revised

Island/oasis Island group/region Province Reference to original dataset Livingston Island† South Shetland Islands Maritime Antarctica Jones et al. 1993 Signy Island† South Orkney Islands Maritime Antarctica Jones et al. 1993 Beak Island* Prince Gustav Channel Maritime Antarctica Sterken et al. in prep. James Ross Island* Prince Gustav Channel Maritime Antarctica Kopalová et al. in prep Horseshoe Island* Marguerite Bay Maritime Antarctica Hodgson et al. in prep. Pourquoi Pas Island* Marguerite Bay Maritime Antarctica Hodgson et al. in prep. West Ongul Island* Lützow Holm Bay Continental Antarctica Tavernier et al. in prep. Skarvsnes* Lützow Holm Bay Continental Antarctica Tavernier et al. in prep. Schirmacher Oasis* Schirmacher Oasis Continental Antarctica Verleyen et al. in prep. Vestfold Hills† Prydz Bay Continental Antarctica Roberts & McMinn 1996 Bølingen Islands† Prydz Bay Continental Antarctica Sabbe et al. 2004 Larsemann Hills† Prydz Bay Continental Antarctica Verleyen et al. 2003 South Georgia South Georgia Sub-Antarctica – South Van de Vijver & Beyens Atlantic Ocean 1996 Kerguelen Kerguelen Sub-Antarctica – South Van de Vijver et al. Indian Ocean 2001 Crozet Crozet Sub-Antarctica – South Van de Vijver & Beyens Indian Ocean 1999 Prince Edward Island Prince Edward Island Sub-Antarctica – South Van de Vijver et al. Indian Ocean 2008 Marion Island* Marion Island Sub-Antarctica – South Van de Vijver et al. Indian Ocean 2008 Heard Island Heard Island Sub-Antarctica – South Van de Vijver et al. Indian Ocean 2004 Ellef Ringnes Canadian Arctic Canadian Arctic Antoniades et al. 2008 Northern Ellesmere Canadian Arctic Canadian Arctic Antoniades et al. 2008 Prince Patrick Island Canadian Arctic Canadian Arctic Antoniades et al. 2008

42 Interhemispheric contrasts in diatom diversity Supplementary references of Table S1 Antoniades, D., Hamilton, P., Douglas, M.S.V. & Smol, J.P. (2008) Freshwater diatoms of the Canadian High Arctic Islands: Ellef Ringnes, northern Ellesmere and Prince Patrick islands. Iconographia Diatomologica 17. A.R.G. Gantner Verlag, Ruggell, 706 pp. Jones, V. J., Juggins, S. & Ellisevans, J. C. (1993) The relationship between water chemistry and surface sediment diatom assemblages in Maritime Antarctic lakes. Antarctic Science, 5, 339-348 Roberts, D. & McMinn, A. (1996) Relationships between surface sediment diatom assemblages and water chemistry gradients in saline lakes of the Vestfold Hills, Antarctica. Antarctic Science, 8, 331-341 Sabbe, K., Hodgson, D. A., Verleyen, E., Taton, A., Wilmotte, A., Vanhoutte, K. & Vyverman, W. (2004) Salinity, depth and the structure and composition of microbial mats in continental Antarctic lakes. Freshwater Biology 49:296-319 Sabbe, K., Verleyen, E., Hodgson, D.A., Vanhoutte, K. & Vyverman, W. %HQWKLFGLDWRPÀRUDRIIUHVKZDWHUDQGVDOLQHODNHVLQWKH/DUVHPDQQ+LOOV and Rauer Islands, East Antarctica. Antarctic Science, 15, 227-248 Van de Vijver B. & Beyens L.(1996) Freshwater diatom communities of the Strømness Bay area, South Georgia. Antarctic Science, 8, 359-368 Van de Vijver B. & Beyens L.(1999) Freshwater diatoms from Ile de la Possession (Crozet Archipelago, Subantarctica): an ecological assessment. Polar Biology, 22, 178-188 Van de Vijver, B., Ledeganck, P. & Beyens, L. (2001) Habitat preferences in freshwater diatom communities from Kerguelen (TAAF, Subantarctica). Antarctic Science, 13, 28-36 Van de Vijver, B., Frenot, Y., Beyens, L. (2002) Freshwater Diatoms from Ile de la Possession (Crozet Archipelago, Subantarctica). Bibliotheca Diatomologica 46. J. Cramer, Stutgart, 412 pp. Van de Vijver, B., Beyens, L., Vincke, S. & Gremmen, N. (2004). Moss- inhabiting diatom communities from Heard Island, sub-Antarctic. Polar Biology, 27, 532-543 Van de Vijver, B., Gremmen, N. & Smith, V. (2008). Diatom communities from the sub-Antarctic Prince Edward Islands: diversity and distribution patterns. Polar Biology, 31, 795-808 Verleyen, E., Hodgson, D. A., Vyverman, W., Roberts, D., McMinn, A., Vanhoutte, K. & Sabbe, K. (2003) Modelling diatom responses to climate induced ÀXFWXDWLRQVLQWKHPRLVWXUHEDODQFHLQFRQWLQHQWDO$QWDUFWLFODNHVJournal of Paleolimnology, 30, 195-215

Interhemispheric contrasts in diatom diversity 43

3. Distance decay and species turnover in terrestrial and aquatic diatom communities across multiple spatial scales using three sub- Antarctic islands: an exploratory study

ȱěȱ1, Bart Van de Vijver2, Niek J.M. Gremmen3, Pieter Vanormelingen1, Elie Verleyen1, Koen Sabbe1 and Wim Vyverman1

ǐȱ 1 Laboratory of Protistology and Aquatic Ecology, Ghent University, Krijgslaan 281 - S8, 9000 Gent, Belgium ǐȱ 2 Department of Cryptogamy, National Botanic Garden of Belgium, Domein van Bouchout, 1860 Meise, Belgium ǐȱ 3 Department of Spatial Ecology, Centre for Estuarine and Marine Ecology, P.O. Box 140, 4400 AC Yerseke, The Netherlands

Unpublished manuscript

Species turnover of diatom communities 45 Abstract The decrease in community similarity with geographical distance or “distance decay” is the result of various processes, including environmental species sorting and dispersal limitation. The relative ȱȱȱȱ ȱěȱȱȱȱȱȱǯȱ For example, organisms with higher dispersal capacities will display shallower slopes. Terrestrial diatoms are assumed to have a higher dispersal potential than lacustrine diatoms due to the higher susceptibility to be picked up by air currents, the higher connectivity in their habitat, and/or their higher tolerances for desiccation. We tested this hypothesis for both habitats using a taxonomically intercalibrated dataset covering two South Indian Ocean archipelagos (Kerguelen and Crozet) and a South Atlantic Ocean island (South Georgia). Both terrestrial and aquatic diatom community similarities decrease with distance over intermediate (ca. 1,500 km) and large (ca.ȱŜǰŖŖŖȱǼȱȱǰȱȱȱ¢ȱ ȱ¢ȱęȱȱȱ the largest scale for the terrestrial communities. The rate of distance ¢ȱ ȱ ȱ ȱ ȱ ȱ ęȱ¢ȱ ěȱȱ from that in lacustrine communities at none of the spatial scales, ȱȱȱȱȱȱȱ ȱĚȱĴȱȱȱȱ higher community similarities in terrestrial communities and the ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ and/or broader niches for terrestrial diatoms. Within islands, species ȱǻΆȬ¢Ǽȱ ȱ¢ȱȱȱȱ ȱȱ ȱȱȱ Ȭȱǻ΅Ǽȱ¢ǰȱ¢ȱĚȱȱȬ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ¢ȱěȱǯȱ

46 Species turnover of diatom communities 3.1 Introduction Community similarity decreases with geographic distance. This ȱ Ĵȱȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻe.g. Nekola & White, 1999; Condit et al., 2002; Tuomisto et al., 2003; Wiersma & Urban, 2005), eukaryote microorganisms (e.g. Hillebrand et al., 2001; Green et al., 2004; Shurin et al., 2009) and even bacteria (e.g. Cho & Tiedje, 2000; Whitaker et al., 2003; Horner-Devine et al., 2004). The decrease in similarity of communities with increasing ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱǰȱȱȬ¢ȱǻĴȱǰȱŗşŜŖǰȱŗşŝŘǲȱȱ et al., 2005), and is accounted for by at least three mechanisms. First, environmental characteristics change over distance causing a decrease in biotic similarity due to species sorting (Nekola & White, 1999; Leibold et al., 2004). Second, limited dispersal capacity will result in a decrease in community similarity with distance (Hubbell, ŘŖŖŗǼǯȱǰȱȱȱȱěȱȱ¢ȱȱȱęȱȱ of the landscape between patches, and the presence of dispersal barriers will result in a decrease in community similarity (Garcillan & Ezcurra, 2003). These three mechanisms are not mutually exclusive, and various data indicate that in most communities they jointly control distance decay (Tuomisto et al.ǰȱŘŖŖřǲȱĴȱǰȱŘŖŖśǼǯ The rate (or slope) of spatial decrease in community similarity depends on several factors, including organismal features such as type of dispersal (passive vs. active) (Soininen et al.ǰȱŘŖŖŝǼǰȱ£ȱ (Hillebrand et al., 2001; Soininen et al.ǰȱ ŘŖŖŝǼǰȱ ȱ ȱ and formation of resting stages (Martiny et al., 2006) and dispersal ability (Ozinga et al., 2005), but also includes the spatial scale under study (Nekola & White, 1999; Soininen et al.ǰȱŘŖŖŝǼǰȱȱěȱȱ and taxonomic resolution (Green & Bohannan, 2006; Soininen et al.ǰȱ ŘŖŖŝǼǰȱ ȱ ȱ ¢ȱ ǻǰȱ ȱ ȱ Ǽȱ ȱ ȱ ěȱȱ ȱ ȱ ęȱȱ ǻȱ et al.ǰȱ ŘŖŖŝǼǯȱ Microorganisms are shown to have lower rates of distant decay than macroorganisms (Hillebrand et al., 2001; Green et al., 2004), but the lower taxonomic resolution used for microbial studies such as ȱȱȱȱȱȱȱěȱȱȱȱȱĴȱȱ ȱĚȱĴȱȱȱǻ ȱǭȱǰȱŘŖŖŜǲȱ¢ȱet al., 2006). Similarly, terrestrial organisms have lower rates of distance decay than freshwater organisms, plausibly due to the more fragmented ęȱȱȱǰȱȱȱȱȱĚȱĴȱȱȱȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ (Soininen et al.ǰȱŘŖŖŝǼǯȱȱȱȱȱȱȬȱ

Species turnover of diatom communities 47 ǰȱęȱȱȬȱěȱȱȱ¡ǲȱ however, knowledge on dispersal ability and niche is, in particular for microorganisms, mostly lacking (Martiny et al., 2006). Because distance decay unites several ecological phenomena, it is valuable for visualizing spatial turnover across sites and a starting point to ȱȱȱȱȱĴȱȱǻȱet al.ǰȱŘŖŖŝǼǯ Lacustrine diatom community similarities are known to show distance decay (Hillebrand et al.ǰȱŘŖŖŗǲȱ£¢ǰȱŘŖŖŘǼǰȱȱȱęȱȱ with results from multivariate studies showing that both spatial and local environmental factors shape these communities (Telford et al., 2006; Vyverman et al.ǰȱŘŖŖŝǲȱ¢ȱet al., 2009). Furthermore, recent studies on diatom biogeography revealed a high degree of endemism (Verleyen et al.ǰȱĴȱǼȱȱȱȱĚȱ ȱǻȱ et al., 2009), indicating that dispersal limitation does play a role in shaping the distribution of lacustrine diatoms. This idea is further strengthened by reports that lacustrine diatoms are very sensitive ȱ ȱ ȱ ¡ȱ ȱ ǻěȱȱ et al., 2010; ěȱȱ et al., unpubl.), very likely decreasing their dispersal potential. All these data indicate that the biogeographies of lacustrine diatoms are shaped by the same factors as higher organisms, namely a combination of historical factors (climatic and environmental history, geological processes, evolutionary background, isolation and dispersal limitation) and local ecological factors (Lomolino, 2010). For terrestrial diatoms, however, much less is known. Terrestrial diatoms thrive in soils and on mosses, rocks, tree barks and other substrates (Round et al., 1990) and compared with lacustrine habitats such as ponds and lakes, which form islands in a matrix of unsuitable habitats, the availability of terrestrial habitats seems much higher and less fragmented (however, this could not apply to very ęȱȱȱȱȱȱǼȱ ȱȱȱȱȱȱ by wind currents is more likely (Chepil, 1956). Furthermore, soil- inhabiting diatoms are shown to be more tolerant for temperature ¡ȱȱȱȱǻěȱȱet al., 2010) while their ȱȱȱȱȱȱȱǻěȱȱet al., unpubl. b), both potentially enhancing their dispersal capacity. Indeed, the small amount of viable diatom species reported from air traps are ¢ȱȱ¡ȱǻȱǰȱŗşřŝǲȱǰȱŗşŜŗǲȱ Brown et al., 1964; Schlichting, 1964; Roy-Ocotla & Carrera, 1993) while aquatic diatoms are only rarely encountered (but see Geissler ǭȱ ěȱǰȱ ŗşŜŜǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ

48 Species turnover of diatom communities higher susceptibility to aerial dispersal, and a higher tolerance to desiccation raises the probability that terrestrial diatoms have higher ȱĚȱ ǰȱȱȱȱȱȱȱȱ than lacustrine diatoms (Spaulding et al., 2010). Despites these indications, no data are available on the dispersal and biogeography of terrestrial diatoms and no studies have yet explicitly compared ěȱȱ ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ diatoms. The only dataset available covering both lacustrine and terrestrial diatom communities is a survey from B. Van de Vijver & N.J.M. Gremmen on multiple sub-Antarctic islands. The sub-Antarctic region has had a minor human historical impact compared to other geographically less remote regions (Frenot et al., 2001), being the rationale behind studying this area. Using oceanic islands for our study has the drawback that both the terrestrial and lacustrine habitats are equally fragmented on intermediate spatial scale. However, within each island the fragmentation level ȱ ȱ ȱ ȱ ȱ ȱ ěȱǰȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ the framework of environmental variation or dispersal limitation. From this large dataset, we selected the data of the best sampled islands, being the archipelagos of Kerguelen, Crozet and South Georgia, to analyze the distance decay of terrestrial and aquatic diatom communities (Fig. 3.1). While the three archipelagos all lay between 55° and 46° S (South Georgia: 54°S; Îles Crozet: 46°S; Îles Kerguelen: 49°S; spanning a latitude comparable from mid-France ȱȱǼǰȱ¢ȱȱěȱȱȱȱǯȱȱ Georgia, lying in the South Atlantic Ocean, is of continental origin and belongs to the Scotia island arch (related to the subduction of the South American Plate to its east), is geologically characterized by a gneiss and schist bedrock, and is colder than the two other islands (average annual temperature 1.85°C at Grytviken). The archipelagos of Crozet and Kerguelen are situated in the South Indian Ocean, are both of volcanic (hotspot) origin resulting in a basalt bedrock, ȱȱȱȱȱȱȱȱśǯśŜǚȱȱŚǯŝşǚȱ (at Alfred Faure and Port-aux-Prince), respectively. South Georgia is the oldest island (min. 120 Ma), while Kerguelen (min. 40 Ma) and £ȱǻȱŞǯŝȱǼȱȱȱ¢ǯȱ The South Atlantic Ocean and South Indian Ocean form two ȱ ȱ ǰȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱǻĴȱǰȱŗşŜŖǲȱ ȱǰȱŗşŞŚǲȱȱȱĴȱȱ

Species turnover of diatom communities 49 et al., 2010). Situated 1,300 km (South Georgia), 2,500 km (Îles Crozet) and 3,900 km (Îles Kerguelen) from the nearest continent, these three islands are very isolated. The archipelagos of Kerguelen and Crozet, separated by 1,420 km ocean, are more similar in geological, ȱȱȱǰȱȱ¢ȱȱěȱȱ in rainfall levels (Crozet: 2,400 mm/year; Kerguelen: >3,200 mm/ year on the west side, <800 mm/year on the east side) (Frenot et al., ŗşşŞǼǰȱȱȱǻ£ǱȱȱȱŗŚŝƸŗřŖȱŶǲȱ ǱȱŝǰŘŖŖȱ ŶǼǰȱ ȱ ¡ȱ ȱ ǻ£Ǳȱ ŗǰŖśŖȱ ǲȱ Ǳȱ ŗǰŞśŖȱ m) and therefore also in present ice cover (Crozet: 0%; Kerguelen: ŗŘƖǼȱ ǻ£ȱ ȱ ȱȱ Ĵȱȱ et al., 2010). However, they both have a volcanic history with several outbreaks, and during the Last Glacial Maxima the lower parts of both archipelagos remained Ȭȱǻ ȱȱȱȱĴȱȱet al., 2010). A thorough study comparing the environmental variables on both islands is thus necessary before we can use these two archipelagos as a case- ¢ȱ ȱ ¡ȱ ěȱȱ ȱ ȱ ¢ȱ ȱ Ȭȱ and lacustrine diatom communities at intermediate scale while assuming the availability of similar habitat types. Meanwhile, we can assume that the environmental conditions of Îles Crozet and Îles Kerguelen are more alike compared to South Georgia, which would result in more comparable diatom communities. The distance

Fig. 3.1 Map of the South Polar Region including the Antarctic and sub-Antarctic. The island groups of South Georgia, Crozet and Kerguelen are indicated by grey boxes.

50 Species turnover of diatom communities between these South Indian Ocean Islands and South Georgia (5,900 and 6,600 km from Crozet and Kerguelen, respectively) forms an ȱĴȱȱȱȱȱ¢ȱȱȱȱ on even larger scales. Following the assumption that soil-inhabiting diatoms (termed “terrestrial” below) have larger dispersal abilities than lacustrine diatoms (termed “aquatic” below) we hypothesize that the terrestrial communities will be characterized by lower species turnover and higher community similarities than the aquatic communities 1) within islands; and 2) between Crozet and Kerguelen. However, due to the larger environmental variations and geographical distance between the South Indian and the South Atlantic Ocean, we predict that 3) both terrestrial and aquatic communities will display pronounced distance decays at this larger scale.

3.2 Materials & Methods Sampling procedure Samples were taken in such a way as to assure that all habitat types were included, across a wide range of altitudes, moisture levels and intensities of sea spray and manuring and animal trampling. In waterbodies such as lakes, pools and rivers, very wet mires ȱ ȱ ǰȱ ȱ Ĵȱȱ ȱ Ȧȱ ȱ £ȱ ȱ ¢¢ȱ ȱ ȱ ȱ ȱ Ĵȱȱ ȱ ęȱ¡ȱ ȱ řƖȱ formaldehyde. A few moss plants were included in the sample. In drier habitats, bryophytes (or where these were absent, the top soil layer) were collected and air-dried in paper bags. In the case of seepage areas on rocks the algal mat was collected and air-dried. More information regarding the sampling can be found in Van de ȱǭȱ¢ȱǻŗşşŝǼǰȱȱȱȱet al. (2002; 2004b) and Van de Vijver & Mataloni (2008).

Slide preparation and counting Diatom samples for observation by light microscopy (LM) were ȱ ȱȱȱȱȱȱȱěȱȱǻŗşśśǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ řŝƖȱ ŘŘȱ and heating to 80° C for about 1 h. The reaction was completed by addition of KMnO4. Following digestion and centrifugation (three ȱȱŗŖȱȱȱřǰŝŖŖȱ¡ȱǼǰȱȱȱ ȱȱ ȱ

Species turnover of diatom communities 51 distilled water to avoid excessive concentrations of diatom valves on the slides. Cleaned diatom material was mounted in Naphrax®. The samples and slides are stored at the Department of Bryophytes and Thallophytes, National Botanic Garden of Belgium. In each sample, śŖŖȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ŗǰŖŖŖȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ¢ȱśŗȱǰȱȱ ȱěȱȱ ȱ Contrast (Nomarski) and Colorview I Soft Imaging System. In some ǰȱȱȱĜȱȱȱȱȱȱǯȱȱȱȱ several entire slides at x1,000, some samples yielded less than 250 valves. Considering the extreme environments involved, these ȱȱȱǻȱȱȱǭȱ¢ǰȱŗşşŝǼǯȱȱ ęȱȱ ȱȱȱȱ¢ǰȱȱȱȱ ȱȱ¢ȱȱȱȱȱȱęȱȱ ȱ ȱȱȱǯȱ ęȱȱȱȱȱȱ based on descriptions by Bourrelly & Manguin (1954), Le Cohu & Maillard (1983, 1986), Schmidt et al. (1990), Oppenheim (1994), Van de Vijver et al. (2002; 2004a) and Le Cohu (2005). Nomenclature follows Bukhtiyarova & Round (1996), Round & Bukhtiyarova (1996), Lange-Bertalot (2001) and Krammer (2000, 2002).

Data analysis Based on the samples of Crozet, Kerguelen and South Georgia, a species list was compiled using presence-absence and relative abundance data. Sites were subdivided into “aquatic”, “terrestrial” and “semi-terrestrial” habitat classes. The aquatic class contained all permanent lakes. The terrestrial class contained all dry soil samples and thus the true soil-inhabiting diatom communities. All other ȱ ȱęȱȱȱȬǰȱȱ¢ȱȱȱ types including peats and bogs, seepages and mosses. While aquatic sites are the least susceptible to dry out, semi-terrestrial habitats are intermediately susceptible to dry out, and terrestrial habitats are the most likely to undergo desiccation. Datasets of aquatic, terrestrial and semi-terrestrial sites were compiled for each island and analyzed separately.

Community similarity and distance decay For each habitat class, pairwise community similarities of sites were calculated between sites 1) within islands, and 2) among the three ěȱȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ

52 Species turnover of diatom communities were based on the Jaccard Index, using presence-absence data as we are mainly interested in dispersal limitation and therefore in the absence or presence of species. Similarities were calculated in R using the vegan package (Oksanen et al., 2011) after which the ǰȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱȱȱ¡ȱǻȱĜȱȱǼǯȱěȱȱȱ similarity between terrestrial and aquatic datasets were statistically ȱ ȱ Ȭȱ ǻȱ ŝǰȱ Ǽǯȱ ȱ ȱ Ȭȱ comparisons, the median similarity values of the three islands were taken as replicates. For the among-island comparisons, we only compared between oceans (i.e. largest spatial scale); the median similarity values of the pairwise comparisons of Crozet vs. South Georgia and Kerguelen vs. South Georgia were taken as replicates. To analyze the distance decay per habitat class, similarities ȱĴȱȱȱȱȱ ȱȱȱǯȱȱ ȱ Ȭȱȱ ȱȱĴȱȱȱȱȱȱ of 1 km because we had no data yet on the geographical coordinates of the individual sites. The median between-island similarities were Ĵȱȱȱȱȱȱ ȱȱȱǯȱȱȱ of the straight lines connecting the median similarity values (per habitat class) for all between-island pairs were calculated using ȱȱȱ¡ȱǻȱĜȱȱȱŘŖŗŖǼȱȱ¢ȱ transformed similarity and distance data. For the statistical analyses, the distance decay slopes for the data points of Kerguelen and Crozet were taken as replicates per distance comparison (intermediate and ȱ Ǽǯȱ ęȱȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ the aquatic and terrestrial datasets was analyzed using t-tests in ȱŝȱǻǼǯ

Diversity partitioning Diversity partitioning is a tool to partition the total diversity of a dataset or region (i.e.ȱ ·Ȭ¢Ǽȱ ȱ ȱ ȱ ǰȱ ȱ the average local (sample/site) diversity (i.e.ȱ ΅Ȭ¢Ǽȱ ȱ ȱ average turnover in diversity between localities (samples/sites) (i.e. ΆȬ¢Ǽǯȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱȱȱȱ¢ȱ¡ȱȱǯȱȱȱ ȱěȱȱ diversity indices: species richness (Q=0) and Shannon diversity (Q=1). Based on the gamma and alpha-diversities, beta-diversity is ȱȱΆƽ·Ȭ΅ȱǻȱȱȱǼȱȱȱΆƽ·Ȧ΅ȱǻȱ multiplicative species richness and for Shannon diversity) (Jost, ŘŖŖŝǼǯȱȱ¢ȱȱ ȱȱȱȱȱ

Species turnover of diatom communities 53 ȱȱȱȱȱěȱȱȱǱȱŗǼȱ ȱ islands, 2) among islands in each pairwise comparison, 3) among the South Indian Ocean (i.e. the combination of datasets from Crozet and Kerguelen) and the South Atlantic Ocean (i.e. South Georgia). Analyses were performed for each habitat class separately based on relative abundances using the software PARTITION v3 (Veech & Christ, 2009). Samples were not weighted, and randomization was individual-based using 1,000 randomizations. The “among islands” and “among oceans” analyses were hierarchically analyzed, such that the total beta-diversity was decomposed into the beta-diversity ȱȱ ȱȱȱȱ ȱȱȱǻΆŗǼǰȱȱȱ Ȭ¢ȱȱȱȱȱȱȱȱǻΆŘǼǯȱȱ ȱ ȱȱȱȱȱ ¢ǰȱ ȱ·ƽ΅ƸΆŗƸΆŘȱǻǰȱ 1996). Additive partitioning has the advantage over multiplicate ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ be compared with each other and with the components of other communities (Lande, 1996; Veech et al., 2002). Shannon diversity can only be partitioned multiplicatively, and to make the alpha, beta and gamma components relate additively, we transformed these components into Shannon entropies using the natural logarithm ǻ ǰȱŘŖŖŝǼǯ

3.3 Results Dataset properties In total 556 samples were analyzed, containing a total of 339 species. The number of samples and number of species per island is given in Table 3.1. South Georgia was less intensively sampled, and we therefore predominantly focus on the island groups of Kerguelen ȱ£ǯȱȱȱȱ¢ȱȱȱȱěȱǰȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ǻȱ et al., ŘŖŖŝǼǯȱ

Similarities and distance decay of similarity Median similarities were higher within islands than between oceans ǻǯȱřǯŘǲȱǯȱřǯřǼȱ ȱȱȱȱ¢ǰȱȱȱěȱȱ ȱ¢ȱ¢ȱęȱȱȱȱȱȱǻǯȱřǯŘȱ B, pƽŖǯŖŗŘŝǼǯȱȱȱȱȱǰȱ ȱǰȱȱȱ sites showed a lower median similarity compared with the terrestrial and semi-terrestrial sites when measured per island (Fig. 3.2 A;

54 Species turnover of diatom communities 96 (30) action of 1) -0.174 -0.0074 -0.0502 -0.1439 number of species semi-terrestrial 32 -0.1505 -0.0089 -0.0536 -0.1275 terrestrial 54 (8) Slopes of distance decay 4 -0.115 -0.093 -0.0321 -0.0913 aquatic 115 (59) Aquatic Terrestrial Semi-terrestrial 6,600 km 1,420 km 1,420 km 5,900 km Distance 35 # sites species # # sites species # # sites species # 166 Total # species Crozet Kerguelen South Georgia South Georgia 71 Crozet Crozet Kerguelen Kerguelen Crozet 304 246 97 174 (20) 91 199 (32) 116 188 (14) Basis island (A) island (B) Compared Kerguelen 181 220 42 128 (14) 13 124 (4) 127 204 (49) South Georgia Island group Total # sites occurring only in that particular habitat is given between brackets. between occurring only in that particular habitat is given Table 3.1 Dataset characteristics indicating the number of sites and species per island group habitat on each group. The Table distance (km) and similarities (as fr calculated between 3.2 Summary of the slopes distance decays. Linear regressions were Table s were both logarithmically and the compared island (B). Distance similaritie s were the basis island (A) and similarities between within the basis island (A) transformed.

Species turnover of diatom communities 55 Fig. 3.2 Percentages of community similarity (Jaccard Index) per habitat type summarized ǼȱȱȱȱȱǰȱȱǼȱȱ¡ȱȱȱȱȱěȱȱȱ as replicates (within island similarities) or the pairwise similarities between South Georgia and Kerguelen respectively Crozet as replicates (between oceans similarities). * indicates ȱěȱȱȱ¢ȱ¢ȱȱǻp<0.05).

56 Species turnover of diatom communities ǯȱřǯřǼǯȱ ǰȱȱěȱȱ ȱȱȱȱ ¢ȱ ȱ ȱ ȱ ¢ȱ ęȱȱ ǻp>0.05) (Fig. 3.2 B). At intermediate scale (i.e. when comparing Crozet and Kerguelen; ca. 1,500 km), the aquatic communities had again a lower similarity compared with the terrestrial and semi-terrestrial communities (Fig. 3.2 A; Fig. 3.3) but because no replicate island ȱ ȱȱȱȱȱęȱȱȱȱȱ ǯȱ ȱ Ĵȱȱ ȱ ȱ ǻǯȱ řǯřǼǰȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ĚȱĴȱȱ ¢ȱ ȱ ¢ǰȱ while the aquatic communities decreased stronger in similarity, but the slopes of the linear regressions between similarity and distance ǻȱřǯŘǼȱȱȱěȱȱęȱ¢ȱ ȱȱ¢ȱǻp>0.05). At the largest spatial scale when comparing the South Indian Ocean islands (Crozet and Kerguelen) with the South Atlantic Ocean island South Georgia (ca. 6,000 km), similarities were very low

Fig. 3.3 Graph showing median similarity (Jaccard Index) of the within-island and between- ȱ¢ȱĴȱȱȱǯ

Species turnover of diatom communities 57 for all habitat types (Fig. 3.2; Fig. 3.3). Among oceans, similarities decreased drastically for all habitat types, but more slowly for the aquatic communities (Table 3.2; Fig. 3.3), as their initial within-island ȱ ȱ ǯȱǰȱ ȱ ęȱȱ ěȱȱ ȱ slopes were found (p>0.05).

Beta-diversity within islands ȱ ȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ǻ΅Ȭ¢Ǽȱ ȱ ęȱ¢ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ǻΆȬ¢ǰȱǼȱ ȱȱȱȱȱǻȱ datasets: p=0.00001; terrestrial datasets: p=0.0168) (Fig. 3.4 A-B, Fig. 3.5 A, C), indicating that, compared to the total diversity ǻ·Ȭ¢Ǽȱǰȱȱȱȱȱ ȱȱȱ ǰȱȱȱěȱȱ¢ȱȱȱȱǯȱȱĴȱȱ was opposite when calculated as Shannon diversity (Fig. 3.4 C, ǯȱřǯśȱǰȱǼǯȱȱ΅Ȭ¢ȱȱȱȱ ȱȱ ęȱ¢ȱȱȱȱΆȬ¢ȱǻpƽŖǯŖŘśŜǼǰȱȱȱęȱȱ ěȱȱ ȱȱȱȱȱǯȱȱȱ ¡ȱȱ accounts for abundances and evenness in the communities and can be interpreted as the chance that the next species is identical to the previous one. The opposite results based on species richness and the Shannon Index suggests that the large turnover in species based on species richness is mainly due to variations in absence or presence of ȱȱ ȱȱěȱȱǯȱȱȱȱ ȱ ȱ ΆȬ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻǯȱřǯŚȱȬǲȱǯȱřǯśȱȬǼǰȱȱȱ ȱȱęȱȱǻǯȱřǯśȱǰȱǲȱ p>0.05).

Beta-diversity among islands on intermediate scale (ca. 1,500 km) Focusing on Crozet and Kerguelen, a larger turnover in species is ȱ ȱȱȱȱǻΆŗǼȱȱ ȱȱ ȱȱ ȱ ǻΆŘǼȱ ǻǯȱ řǯŜȱ Ȭǰȱ Ǽȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ǯȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻ·Ȭ¢Ǽȱ was thus more determined by the single islands, meaning that a large part of the diversity was already present on both islands. The terrestrial community had a lower turnover within islands ǻΆŗǼȱȱȱǰȱȱ¢ȱǯȱ ǰȱ it had larger among-island turnover than the aquatic community when using species richness as diversity index (Fig. 3.6 B, left), and

58 Species turnover of diatom communities lower among-island turnover when using Shannon diversity (Fig. 3.6 C, left), which indicates that species turnover in the terrestrial communities was mainly due to changes in the rare taxa. However, because no replicate islands are available on this intermediate scale ȱȱęȱȱȱȱǯ

Beta-diversity among islands on larger scale (ca. 6,000 km) When looking individually at the turnover between South Georgia and Crozet respectively Kerguelen (Fig. 3.6 A-C, middle and right Ǽǰȱȱ ȱȱȱȱǻΆŘǼȱȱȱȱȱȱ ȱ ȱ ȱ £ȱ ǻęȱȱ ȱ ȱ Ǽȱ ȱ ȱ ȱ¢ǯȱ ȱȱȱ ȱȱȱǻΆŗǼǰȱȱ communities had a lower species turnover between sites than aquatic ȱ ǻǯȱ řǯŜȱ ȬǼǰȱ ȱ ȱ ęȱȱ ȱ ȱ species richness (species richness p=0.0222, Shannon Index p=0.0545). However, contrary to our expectations, the species turnover between £ȱȱ ȱȱȱ ȱǻΆŘǼȱ ȱȱȱȱ ȱȱȱȱȱǻǯȱřǯŝȱȬǼȱȱ ęȱȱ ȱȱȱȱǻpƽŖǯŖŗŞŘǼȱȱȱęȱȱ ȱȱȱȱ ¡ǯȱȱȱȱȱĴȱȱ ȱ¢ȱ ęȱȱȱȱȱȱȱȱȱȱȱ observed for terrestrial communities was mainly due to variations in rare species. When giving more weight to abundances and evenness by using the Shannon Index, aquatic and terrestrial communities ȱ ȱ ȱ ȱ ǻǯȱ řǯŝȱ Ǽǯȱ ǰȱ ȱ ȱ ȱ¢ǰȱΆŘȬ¢ȱȱȱȱȱǻǯȱřǯŜȱǲȱǯȱ řǯŝȱǼȱ ȱȱȱ ȱȱȱȱ ¡ȱǻǯȱřǯŜȱǲȱǯȱ řǯŝȱǼǰȱȱȱȱȱȱȱȱȱȱ ȱ¢ȱȱȱěȱȱȱȱǯ

Beta-diversity among the South Atlantic and South Indian Ocean When performing a hierarchical partitioning over the two oceans ǻǯȱ řǯŞȱȬǼǰȱ ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ islands, but no replicate ocean combinations were available and ȱ ȱ ȱ ¢ȱ ȱ ȱ ǯȱ ȱ ΆŘȬ¢ǰȱ the turnover in diversity between the oceans, tended to be larger for terrestrial communities than for aquatic communities and again

Species turnover of diatom communities 59 Fig. 3.4 Diversity partitioning within islands for each habitat type shown per island as absolute values of additive species richness (A) and percentages ȱ ȱ ·Ȭ¢ȱ based on additive species richness (B) and Shannon Index (C). Total diversity consists ȱ ȱ ΅Ȭ¢ȱ or average diversity within sites, and ΆȬ¢ȱ ȱ average diversity between sites within the island. Aq = aquatic communities, Ter = terrestrial communities, Semi = semi-terrestrial communities.

60 Species turnover of diatom communities Fig. 3.5ȱ¡ȱȱȱ΅Ȭ¢ȱȱȱ Ȭȱ¢ȱǻǰȱǼȱȱΆȬ¢ȱȱ average between-sites diversity (C, D) for the aquatic and terrestrial communities using the three islands Crozet, Kerguelen and South Georgia as replicates. Diversity was partitioned using species richness (A, C) and using Shannon Index (B, D).

Species turnover of diatom communities 61 Fig. 3.6 Diversity partitioning among islands in the three pairwise combinations, shown per combination as absolute values of additive species richness (A) and percentages of the ·Ȭ¢ȱ ȱ on additive species richness (B) and Shannon Index (C) for the three habitat types. Total diversity ȱ ȱ ȱ ΅ŗȬ diversity or average within-site diversity ȱ ǰȱ ΆŗȬ diversity or average among-site diversity within each island, ȱ ΆŘȬ¢ȱ ȱ average between- island diversity. Aq = aquatic communities, Ter = terrestrial communities, Semi = semi-terrestrial communities.

62 Species turnover of diatom communities Fig. 3.7ȱ¡ȱȱȱΆŗȬ¢ȱȱȱ Ȭȱ¢ȱǻǰȱǼȱȱΆŘȬ¢ȱ or average between-island diversity (C, D) for the aquatic and terrestrial communities using Kerguelen/Crozet versus South Georgia as replicates. Diversity was partitioned using ȱȱǻǰȱǼȱȱȱȱ ¡ȱǻǰȱǼǯȱȘȱȱ¢ȱěȱȱ datasets (p<0.05).

Species turnover of diatom communities 63 Fig. 3.8 Diversity partitioning among the South Atlantic and South Indian oceans shown for the three habitat types as absolute values of additive species richness (A) and percentages ȱ ȱ ·Ȭ¢ȱ based on additive species richness (B) and Shannon Index (C). Total diversity ȱ ȱ ȱ ΅ŗȬ diversity or average within-site diversity ȱ ǰȱ ΆŗȬ diversity or average among-site diversity within each ocean, ȱ ΆŘȬ¢ȱ ȱ average between- ocean diversity.

64 Species turnover of diatom communities larger when using species richness compared to the Shannon Index. ȱΆŗȬ¢ǰȱȱȱ ȱȱ ȱȱǰȱ was lower for the terrestrial habitat than for the aquatic habitat, and again lower when using Shannon Index.

3.4 Discussion ȱȱ¢ȱ¢ȱęȱȱȱȱ¢ȱ ȱȱȱ to test, and due to the lack of replicate data we were not able to statistically evaluate some observed trends. First, within the studied islands, terrestrial diatom communities had higher similarities and lower turnover between sites than aquatic communities. While this ȱȱęȱȱ ȱ¢£ȱȱȱ¢ȱǻǯȱřǯśȱ ȬǼǰȱȱ Ȭȱ¢ȱǻΆŗǼȱȱȱȱ ȱ ęȱ¢ȱȱ ȱ¢£ȱ£Ȧ ȱversus South ȱǻǯȱřǯŝȱȬǼǯȱȱȱȱȱ ȱȱ turnover of terrestrial communities within islands could be a result of the higher connectivity between the sites, lower environmental variation between the sites, broader niches of terrestrial diatoms, a higher dispersal ability of terrestrial diatoms over short distances, or a combination of these. While we were not able to disentangle these ȱȱȱǰȱȱĴȱȱȱȱȱȱ ȱ results of a meta-analysis by Soininen et al.ȱǻŘŖŖŝǼȱȱȱȱ smallest scale similarities were highest in the terrestrial realm and lowest for freshwater communities. This higher similarity between terrestrial communities within islands is most likely related to the less fragmented terrestrial habitat matrix (Soininen et al.ǰȱŘŖŖŝǼǰȱȱȱȱȱȱ and colonization probabilities and higher prevalence of terrestrial species. We can assume that this is also the case for diatoms because there are no a priori reasons to expect that terrestrial diatoms would have broader niches, or soils would be less variable with respect to the abiotic environment than lake sediments. The extent of variation in environmental conditions of dry soils and lakes could be assessed, ǰȱ ȱ ȱ ǯȱ ǰȱ ȱ ęȱ¢ȱ ȱȱǻΆȬ¢Ǽȱ ȱȱȱȱȱ ȱȱȱ¢ȱ ȱȱǻ΅Ȭ¢Ǽȱȱȱȱ of species sorting and environmental small-scale heterogeneity, ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ěȱȱ ǻi.e.ȱęȱȱ ȱ have an advantage and dominate the community) (Drake, 1991). ǰȱȱ¢ȱěȱȱ ȱȱȱȱȱ¢ȱ

Species turnover of diatom communities 65 between sites, except if a small number of species are dominant, more abundant and would therefore have higher dispersal rates ȱ ȱ ȱ ȱ ȱ ęȱȱ £ȱ ȱ ǯȱ ȱ ȱ ȱ in terrestrial communities could thus indicate higher dispersal rates Ȧȱȱ¢ȱěȱȱȱȱȱǯȱ For all habitats, the turnover was higher when based on species richness than when based on the Shannon Index. Because species richness uses presence-absence data, it gives a disproportionally higher weight to rare species, while Shannon diversity also accounts for relative abundances and evenness of communities. The higher turnover based on species richness indicates that the rare species contribute most to the variations between sites. In a second hypothesis we predicted that terrestrial communities would have lower turnover and higher similarities at intermediate spatial scales (ca. 1,500 km) between the island groups Kerguelen and Crozet when compared to the aquatic communities. Indeed, lower similarities were observed for the aquatic communities, but species ȱǻΆŘȬ¢Ǽȱ ȱȱȱ ȱ¢ȱ ȱȱȱ terrestrial communities when using Shannon diversity and actually higher when using species richness (Fig. 3.6). Because no replicate island combinations were available at this intermediate scale the ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱȱĴȱȱȱ ȱȱǰȱȱ ȱ¢ȱ between the aquatic sites could be the result of environmental ǰȱȱȱȱ¢ȱěȱǰȱ ȱȱȱ similarities of terrestrial communities could be due to larger ȱǰȱȱȱěȱȱ ȱȱȱ the two islands, or a more generalist life-style of terrestrial diatoms. ȱȱȱǰȱĚȱȱȱȱ¢ȱ ȱȱ archipelagos of Kerguelen and Crozet, it is possible that these island groups share similar terrestrial and aquatic habitat types (but as noted in the introduction, this should be further analyzed). Assumed ȱȱȱȱȱ ȱȱěȱȱȱ aquatic and terrestrial sediments in a similar extent (as assumed in Hillebrand et al.ǰȱ ŘŖŖŗǼǰȱ ȱ ȱ ¢£ȱ ȱ ȱ ěȱȱ in similarity level between the aquatic and terrestrial communities could be a result of higher dispersal capacities of terrestrial diatoms due to higher habitat availability (or higher chance to arrive in a suitable patch), higher susceptibility to dispersal (Schlichting, 1961; Brown et al., 1964; Schlichting, 1964; Roy-Ocotla & Carrera, 1993) ȱ ȱ ȱ ȱ ȱ ǻěȱȱ et al., unpubl. b).

66 Species turnover of diatom communities However, this should be further analyzed using environmental and community composition data. ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ȱ richness and Shannon diversity among the terrestrial and aquatic communities of Crozet and Kerguelen needs to be interpreted in the light of what both indices account for. As stated above, species richness gives a disproportionally higher weight to rare species. The lower turnover in terrestrial communities based on the Shannon diversity indicates that the dominant, abundant species in terrestrial ȱȱȱȱȱěȱȱȱȱȱǰȱ ȱ it are mainly the rare species which account for the higher turnover between the islands detected using species richness. For aquatic communities, the opposite is true, and variation in community composition is thus mainly determined by the more abundant species. This could indicate that the distribution of aquatic species ȱȱĚȱȱ¢ȱȱǰȱȱȱ ȱ¢ȱȱ ȱȱȱȱ ȱ¢ȱęȱȱȱȱȱȱ at other sites. Alternatively, this could indicate that aquatic diatoms are more dispersal limited, and species that arrived by single, unlikely dispersal events will thrive at that location, but will not be able to be easily dispersed to a next, suitable lake, or that priority ěȱȱȱȱȱȱȱǯȱ For the terrestrial communities, abundant species could be the ȱȱǰȱȱȱěȱȱȱȱ ȱȱȱȱȱȱȱȱ¢ȱęȱȱǯȱ Alternatively, the abundant terrestrial species, by having larger population densities, are more likely to successfully disperse over the two islands, creating a low turnover in these dominant species, while the rare, less abundant species are less likely to successfully disperse. Alternatively, it is possible that the rare terrestrial species are species that do not have high tolerances to dispersal, while the ȱ ȱ ǯȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ range of possible mechanisms, it is not easy to interpret these and further study is needed to reveal the processes at work. However, ȱ Ĵȱȱ ȱ ęȱȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ additional isolated islands. In agreement with our last hypothesis, we observed a low community similarity between the South Indian Ocean and South Atlantic Ocean islands (ca. 6,000 km) and high distance decay for both terrestrial and lacustrine communities, which was only ęȱȱ ȱ ȱ ȱ ȱ ǻǯȱ řǯŘȱ Ǽǯȱ ȱ ȱ ȱ

Species turnover of diatom communities 67 agreement with similar discontinuities over these oceans reported ȱȱȱǻȱȱĴȱȱet al., 2010), Coleoptera (Morrone, 1998) and insects and birds (Greve et al., 2005), and the distinction of ȱȱȱȱ ȱȱȱǻĴȱǰȱ ŗşŜŖǲȱ ȱǰȱŗşŞŚǲȱȱȱĴȱȱet al., 2010). The distribution of macroorganisms on the sub-Antarctic islands is determined by environmental variables such as temperature and geological features including island age and area, but also distance to the nearest continent and thus dispersal limitation (Chown et al., 1998). While ȱȱȱȱȱȱĚȱȱȱȱȱȱ variation on the distribution of terrestrial and aquatic diatoms yet (Borcard et al., 1992; Borcard & Legendre, 1994), recent studies on lacustrine diatom biogeography underscore the importance of ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ Ěȱȱ of dispersal limitation on diatom distribution (Telford et al., 2006; Vyverman et al.ǰȱŘŖŖŝǲȱ¢ȱet al., 2009), and our data suggest that this also applies for the sub-Antarctic diatom communities. It is clear that many questions still remain and need further study using additional replicate locations to enable statistical testing ȱȱȱĴȱǰȱȱȱȱȱȱ data. While we focused on very isolated islands, it would be ȱȱȱȱĴȱȱ ȱȱȱȱ communities in a more continuous landscape. Furthermore, the Ěȱȱȱȱȱȱȱȱȱȱ communities should be assessed to elucidate to what degree the observed distance-decay relationship is related to environmental ȱȱȱȱȱȱěȱȱȱȱǲȱ and many aspects of the ecology of soil-inhabiting diatoms are still ǯȱȱȱęȱȱǰȱȱȱ ȱȱȱȱȱ and lacustrine diatoms, distance does play a role in their distribution on isolated islands, and that terrestrial and aquatic communities ȱěȱȱȱȱǯ

Acknowledgements ȱȱ ȱęȱ¢ȱȱ¢ȱȱȱȱęȱȱ Research – Flanders (FWO-Flanders, Belgium, funding of CS, PV ȱ Ǽǰȱ ȱ Ȭȱ ř ȬŖśřřȬŖŝȱ ȱ ¡ȱ ȱ ȱ terrestrial and aquatic diatom communities, and the FWO-project 3G-0419-08 on thermal adaptation.

68 Species turnover of diatom communities Part 2. Speciation in time and space

Part 2: Speciation in time and space 71

4. Time-calibrated multi-gene phylogeny of the diatom genus Pinnularia

ȱ ěȱ1, Heroen Verbruggen2, Alexander P. Wolfe3, Pieter Vanormelingen1, Peter A. Siver4, Eileen J. Cox5, David G. Mann6, Bart Van de Vijver7, Koen Sabbe1 and Wim Vyverman1

ǐȱ 1 Laboratory of Protistology and Aquatic Ecology, Ghent University, Krijgslaan 281 - S8, 9000 Gent, Belgium ǐȱ 2 Phycology Research Group, Ghent University, Krijgslaan 281 – S8, 9000 Gent, Belgium ǐȱ 3 Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB T6G 2E3, Canada ǐȱ 4 Department of Botany, Connecticut College, New London, CT 06320, USA ǐȱ 5 Department of Botany, The Natural History Museum, Cromwell Road, London SW7 5BD, United Kingdom ǐȱ 6 Royal Botanic Garden Edinburgh, Edinburgh EH3 5LR, Scotland, United Kingdom ǐȱ 7 Department of Cryptogamy, National Botanic Garden of Belgium, Domein van Bouchout, 1860 Meise, Belgium

ȱĴȱȱȱMolecular Phylogenetics and Evolution

Time-calibrated phylogeny of Pinnularia 73 Abstract Pinnularia is an ecologically important and species-rich genus of freshwater diatoms (Bacillariophyceae) having considerable ȱ ȱ ȱ ¢ǯȱ ęȱȱ ¢ȱ relationships were inferred for 36 Pinnulariaȱ¡ȱȱȱęȱȬȱ dataset. A range of fossil taxa, including newly discovered Middle Eocene forms of Pinnularia, was used to calibrate a relaxed molecular clock analysis and investigate the temporal aspects of the genus’ ęȱǯȱȱȬȱȱȱȱȱ Ȭȱ phylogeny of three major clades and several subclades that were frequently, but not universally, delimited by valve morphology. The genus Caloneisȱ ȱȱȱȱ¢ǰȱęȱȱǰȱȱ currently delimited, this genus is not evolutionarily meaningful and should be merged with Pinnularia. The Pinnularia-Caloneis complex is estimated to have diverged between the Upper Cretaceous and the early Eocene, implying a ghost range of at least 10 million year (Ma) in the fossil record. This result predicts an earlier origin for the raphid diatoms, 10-35 Ma before that inferred by other published molecular clock studies.

Key words: Molecular phylogenetics, relaxed molecular clock, fossil record, raphid diatoms, Pinnularia, Bacillariophyceae, Eocene

74 Time-calibrated phylogeny of Pinnularia 4.1 Introduction Diatoms are an extremely diverse group of unicellular algae that are uniquely characterized by a siliceous cell wall (the frustule) consisting of two valves (Round et al., 1990) and a diplontic life cycle involving gradual size reduction during vegetative divisions and rapid size restitution usually through sexual reproduction (Chepurnov et al., 2004). In the so-called pennate diatoms the valve Ĵȱȱȱ£ȱ¢ȱȱȱȱ¡ǰȱȱȱ most cases the valve is elongate. Raphid pennate diatoms possess a longitudinal slit along the apical axis (the raphe) (Fig. 4.1), from which extracellular polymeric substances are exuded and used in locomotion and adhesion to the substratum (Round et al., 1990). The raphe is considered a derived character state that distinguishes the raphid diatoms from the more ancestral araphid pennate forms that lack this structure and the oldest known forms, the radially organized “centric” taxa (Sims et al., 2006). Based on fossil remains, ȱȱȱęȱȱȱȱȱȱȱǻca. 75 Ma; Chambers, 1966; Hajós & Stradner, 1975), and the raphe-bearing ȱȱǰȱȱŝŖǯŜȮśśǯŞȱȱǻĴȱǰȱŗŞŞŜǲȱǰȱ 1889; Chacon-Baca et al., 2002; Singh et al., 2006). Monophyly of pennate diatoms as a whole, as well as the raphid pennates, has been documented using SSU rDNA and rbcL sequences (e.g. Kooistra et al., 2003; Sorhannus, 2004, 2007). Since their origin, raphid pennate ȱȱęȱȱ¢ȱȱȱȱȱ¢ȱ of the over 200,000 extant species estimated to exist (Mann & Droop, 1996), indicating the evolutionary advantages conferred by the raphe (Sims et al., 2006). Despite the diversity and ecological success of raphid pennate diatoms, relatively few detailed molecular phylogenetic reconstructions exist. Phylogenies applied at genus to ordinal levels have yielded partly unsupported results (e.g. Bruder & Medlin, 2008; Trobajo et al., 2009), in part due to very limited taxon sampling and the use of a limited number of genetic markers (Mann & Evans, 2007). In addition, molecular phylogenies of individual genera have ȱ¢ȱȱȱęȱȱȱ¢ȱ¢ȱȱȱ the elucidation of evolutionary relationships between lineages (e.g. Lundholm et al., 2006; Beszteri et al., 2007; Evans et al., 2008). Except for a few studies of diatoms as a whole, or the wider heterokont group (Kooistra & Medlin, 1996; Medlin et al., 1997; Sorhannus, 2007; Brown & Sorhannus, 2010), or even eukaryotes (Berney & Pawlowski, 2006), there are no explicit time-calibrated phylogenies and few

Time-calibrated phylogeny of Pinnularia 75 76 Time-calibrated phylogeny of Pinnularia analyses formalize the evolutionary associations between the timing ȱȱĴȱȱȱǰȱǰȱ¢ȱȦ or reproductive strategies, life cycles and geographical distributions (but see Casteleyn et al., 2010). Furthermore, despite important ȱȱǰȱȱȱ ȱęȱȱȱ raphid diatoms may be older than previously suspected (Siver & Wolfe, 2007; Siver et al., 2010), numerous gaps remain in the fossil record. As a result, the overall course of evolution in raphid pennate diatoms is not known and the relationships between many groups remain uncertain (Mann & Evans, 2007). Pinnularia Ehrenberg (1843) is one of the most species-rich genera of raphid pennate diatoms, with 2462 taxon names present in Algaebase of which 412 are currently accepted (Guiry & Guiry, 2011). It occurs globally in freshwater habitats of varying pH and trophic status, and to a lesser extent in moist soils, peatlands, spring seeps and marine coastal environments (Round et al., 1990; Krammer, 2000). Members of Pinnularia and the closely related genus Caloneis Cleve (1894) have linear-lanceolate, occasionally capitate valves with a central raphe system (Fig. 4.1) that terminates internally

Fig. 4.1 Morphological variation in Pinnularia and representative illustrations of strains included in the multi-gene phylogeny. The valve construction of typical frustules belonging to the P. divergens group [strain (Tor7)c] is shown by scanning electron micrographs [outside (a), inside (b) and side (= girdle) view (c)]. Cultured and sequenced strains are illustrated by a series of light micrographs (d-x), divided, where appropriate, into subclades recovered in the phylogeny with vertical dashed lines. Clade A (d-f) includes Caloneis lauta (d) as well as P. divergens grade representatives (Tor7)c in (e) and (Tor1)b in (f). Clade B (g-p) includes the “grunowii” subclade (g-i) with Pinnularia sp. (Tor4)i in (g), P. subanglica Pin650 in (h), and P. cf. marchica (Ecrins4) a in (i); the “nodosa” subclade (j-k) with P. acrosphaeria (Val1)b in (j) and P. nodosa Pin885 in (k); and the “subgibba” subclade (l-p) represented by P. parvulissima Pin887 in (l), Pinnularia sp. “gibba-group” (Tor7)f in (m), P. subcapitata var. elongata (Wie)c in (n), P. sp. (Tor4)r in (o), and Pinnularia sp. “gibba-group” (Tor8)b in (p). Clade C (q-x) comprises the “viridis” subclade (q-r) represented by P. neglectiformis Pin706 in (q) and P. viridiformis (Enc2)a in (r); with its sister P. acuminata Pin876 in (s); the “subcommutata” subclade (t) represented by P. subcommutata var. nonfasciata ŗŖȱȱǻǼǲȱȱȱȱȱ¢ȱęȱȱȱ Ȭęȱȱȱȱ¢ȱP. sp (Wie)a in (u); the “borealis-microstauron” subclade (v-w) including P. cf. microstauron (B2)c in (v) and P. borealis Alka1 in (w); and subclade C1 (x) represented by P. cf. altiplanensis (Tor11)b in (x). Live cells of Pinnularia sensu lato (i.e. including Caloneis) also show two distinct plastid arrangements, which are illustrated by light (y, a’, and b’) and laser-scanning confocal ¢ȱȱȱĚȱȱȱǻ£ǰȱȂǼǯȱȱ¡ǰȱȱȱȱP. subcommutata and gibba taxa have parallel plastids on either side of the apical axis (y, £Ǽǰȱ ȱCaloneis silicula (a’) and P. grunowii (b’, c’) have plastids that are joined by a central bridge. See text for details. All scale bars are 10 μm, and images (d-x) are reproduced ȱȱȱęȱȱȱȱ£ȱȱ ȱ¡ǯ

Time-calibrated phylogeny of Pinnularia 77 78 Time-calibrated phylogeny of Pinnularia in helictoglossae at the poles (Fig. 4.1, b; Fig. 4.2, b, d, f, h). They possess two plate-like plastids (Fig. 4.1, y-z) or a single H-shaped plastid (Fig. 4.1, a’-c’). Pinnularia is characterized by the presence of a chambered, double-walled valve structure in which the outer surface is ornamented by multiple rows of small pores forming a multiseriate stria (detail in Fig. 4.2, c), while the inner wall of each chamber or alveolus is perforated by a large transapically elongate aperture (Fig. 4.1, b; Fig. 4.2, b,d,f,h) (Round et al., 1990). Caloneis is similar, except that the inner aperture is smaller and often circular, and sometimes there are two apertures per stria. Despite their species ǰȱ ȱ ¢ȱ ȱ ȱ ęȱȱ ȱ freshwater and terrestrial ecosystems, the evolutionary relationships among Pinnularia and Caloneis species are poorly known. Bruder et al. (2008) constructed a molecular phylogeny of Pinnularia and Caloneis using 18S, 28S and rbcL genes for 15 species. However, this resulted in an overall low support for the clades, suggesting that a more exhaustive sampling with respect to both taxa and genetic markers is needed to produce a well-resolved phylogeny. ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ Pinnularia are also poorly documented, despite the fact that numerous fossils have been reported for the genus. To date, the oldest known diatoms reliably assigned to Pinnularia originate from the Wagon Gap Formation of Wyoming, U.S.A. (Lohman & Andrews, 1968). The age of these sediments is not known precisely because diatom- containing sediment clasts have been redeposited in a carbonate conglomerate, but their age is situated between the Late Eocene and Early Oligocene (35-32 Ma). Dating from the same period, the Oamaru diatomite deposits also contain two species of Pinnularia (Desikachary & Sreelatha, 1989). From the Early Miocene on, diverse morphological types of Pinnularia are reported from freshwater

Fig. 4.2 Supporting SEM micrographs showing the shape of the central raphe endings (left column: a, c, e, g) and the extension of the alveolar opening (right column: b, d, f, h). Central raphe endings can be linear (a, e), drop-like (c) or round (g). Linear raphe endings occur in Caloneis silicula (a), the subcommutata clade (e), P. acuminata, (Wie)a and P. cf. altiplanensis. Drop-like endings are present in clade B (c), the borealis-microstauron clade and the divergens group; while round endings occur in the viridiformis clade (g). The alveolar openings can be small (b), large (d, h) or intermediate (f). Small openings are typical for Caloneis silicula (b) but also occur in P. acrosphaeria; intermediate openings (f) occur in the subcommutata and viridiformis clades and P. acuminata; while large openings (d, h) are present in all other species, including the divergens group, clade B (without P. acrosphaeria), the borealis-microstauron clade with P. borealis (h) and the isolates (Wie)a and P. cf. altiplanensis. Scale bars represent 2 μm (a, b, c, e, g) and 5 μm (d, f, h).

Time-calibrated phylogeny of Pinnularia 79 deposits or as freshwater inwash in marine deposits (for Miocene deposits including Pinnularia see for e.g. Pantocsek, 1886; Héribaud, 1902; Hajós, 1986; Servant-Vildary et al., 1988; Vanlandingham, 1991; Ognjanova-Rumenova & Vass, 1998; Saint Martin & Saint Martin, 2005; Yang et al., 2007; Lewis et al., 2008; Li et al., 2010). The origin of Pinnularia thus predates the Late Eocene, and the occurrence of several taxa within the Wagon Gap material suggests an even earlier origin for the genus. As such, despite the well-established ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ Pinnularia ȱěȱȱȱ¢ǰȱȱȱȱȱȱ¢ȱ scarce and too fragmentary to provide detailed information about ȱ ȱ ȱ ¢ȱ ęȱȱ ǯȱ ȱ ȱ ȱ ǰȱ ȱ given that ghost lineages are anticipated to be common in algae in general and diatoms in particular (Brown & Sorhannus, 2010), it is particularly relevant to produce a well-resolved, time-calibrated phylogeny for this genus. The goal of the present study is to reconstruct a molecular phylogeny for a representative selection of Pinnularia taxa spanning the morphological variability of the genus, and to infer a time-calibrated phylogeny constrained by accurately dated fossil representatives. To achieve this, we sequenced two nuclear markers (18S rDNA, 28S rDNA), two plastid markers (rbcL, psbA) and a mitochondrial marker (cox1) from 36 species of the genus and inferred phylogenies using partitioned models in a likelihood framework. We present new Pinnularia fossils from the Middle Eocene of Canada that are included as constraints in the relaxed molecular clock analyses.

4.2 Material and Methods Taxon sampling We selected strains belonging to 36 Pinnularia taxa (Tables 4.1-2), covering the range of morphological variation within the genus (Fig. 4.1). Because the relationship of Pinnularia and Caloneis remains ambiguous (Cox, 1988; Round et al., 1990; Mann, 2001; Bruder & Medlin, 2008; Bruder et al., 2008), we added three sequences of Caloneis. Five taxa from the genera Sellaphora, Eolimna and Mayamaea were selected as outgroups based on their apparent phylogenetic relatedness (Kooistra et al., 2003; Bruder & Medlin, 2008; Bruder et al., 2008; Evans et al., 2008). A list of the cultured and sequenced taxa is provided in Table 4.2, together with their geographical origin

80 Time-calibrated phylogeny of Pinnularia and morphometric data. Summary photomicrographs of the strains considered are shown in Fig. 4.1, and more detailed morphological descriptions of living and oxidized material are presented elsewhere ǻěȱȱet al.ǰȱǯȱǼǯȱ ęȱȱȱȱȱ ȱ (2000) for Pinnularia and Krammer and Lange-Bertalot (1986) for Caloneis. Voucher slides of oxidized material of all natural samples ȱȱȱȱȱ¡ȱȱęȱȱȱȱȱȱ the Laboratory of Protistology & Aquatic Ecology (Ghent University) and are available upon request. '1$H[WUDFWLRQDPSOL¿FDWLRQDQGVHTXHQFLQJ DNA was extracted from centrifuged diatom cultures following Zwart et al. (1998) using a bead-beating method with phenol extraction and ethanol precipitation. After extraction, DNA was ęȱȱ ȱ ȱ £țȱ ȱ Ȭȱ ¢ȱ ǻǼǯȱ Sequences of the nuclear 18S and the D1-D2 region of 28S, the two plastid genes rbcL and psbA, and the mitochondrial gene cox1 were ęȱȱȱȱȱȱȱȱǻ et al., 1994; Daugbjerg & Andersen, 1997; Guillou et al., 1999; Van Hannen et al., 1999; Yoon et al., 2002; Evans et al.ǰȱŘŖŖŝǲȱĴȱ et al., 2008). PCR ȱ ȱȱȱ ȱȱęȱȱ ȱǻǰȱ Hilden, Germany) following the manufacturer’s instructions. The sequencing reaction was performed by cycle sequencing (initial step of 1 min at 96°C, 30 cycles of 10 s at 96°C, 10 s at 50°C and 1 min 15 s at 60°C) using the ABI Prism BigDye V 3.1 Terminator Cycle Sequencing kit (Applied Biosystems). The resulting sequencing reaction products were analysed on a Perkin-Elmer ABI Prism 3100 automated DNA sequencer (Applied Biosystems). Primer sequences of both PCR and sequencing reactions and PCR temperatures are listed in Appendix A.4. All newly generated sequences have been deposited in GenBank (Table 4.1).

6HTXHQFHDOLJQPHQW The sequences of 18S, 28S, rbcL, psbA and cox1 were edited separately and automatically aligned using ClustalW (Thompson et al., 1994), as implemented in BioEdit 7.0.3 (Hall, 1999). Plastid and mitochondrial markers aligned unambiguously without any gaps. The 18S and 28S alignments were corrected manually using the secondary structure of Toxarium undulatum (Alverson et al., 2006) and Apedinella radians (Ben Ali et al., 2001), respectively, after which ambiguously aligned regions were removed.

Time-calibrated phylogeny of Pinnularia 81 1 ------cox - L AM710506 AM710461 AM710427 AM710435 rbc AM710470 A - - - - psb - 28S AM502039 AM710595 AM501994 AM710550 AM501962 AM710516 AM502003 AM710559 18S - ”)

microstauron Krammer Krammer (Hustedt) Lange-Bertalot AM501969 AM710524 Krammer (Hustedt) Lange-Bertalot

subislandica

permitis borealis subislandica

permitis var. (Grunow) Krammer var. var. (Ehrenberg) Cleve (“southern Cleve (Ehrenberg) (Grunow) Lange-Bertalot Lange-Bertalot Krammer Carter & Bailey-Watts (Ehrenberg) Cleve (Ehrenberg) (Ehrenberg) W. Smith Krammer Ilka Schönfelder Krammer W. Smith Krammer Krammer Krammer Ehrenberg var. Ehrenberg cf. var. Ehrenberg var. Ehrenberg (Ehrenberg) W. Smith W. (Ehrenberg)

altiplanensis isselana marchica microstauron

. .

cf cf grunowii

Caloneis budensis Caloneis lauta Caloneis silicula Eolimna minima atomus Mayamaea Mayamaea atomus acuminata P. borealis P. borealis P. borealis P. borealis P. P. c f. P. c f. P. P. P. P. mesolepta neglectiformis P. neomajor P. neomajor P. P. n o d o s a parvulissima P. Cal 890 TM Pin 706 F (Wes2)f Pin 876 TM Alka 1 (Ecrins4)a (B2)c (Tor11)b Strain Taxon Corsea 2 Pin 885 TM (Tor12)d Pin 889 MG (Ecrins7)a (Tor3)a (Tor1)a Pin 877 TM Cal 878 TM : Taxon list with Genbank accession numbers. Missing sequences are indicated a dash. Data taken from GenBank in bold . 4.1 : Taxon Table

82 Time-calibrated phylogeny of Pinnularia 1 - - - - - cox - L AM710490 rbc AM710503 A - - psb - 28S - AM502023 AM710579 - AM502036 AM710592 18S Krammer Van de Vijver, Cahttová & Metzeltin Cahttová de Vijver, Van - Krammer D.G. Mann & S. Droop Mann & S. D.G. D.G. Mann & S. Droop Mann & S. D.G. W. Smith nonfasciata Krammer subcapitata

elongata var. var. var. var. var. var. Krammer Krammer Krammer (Nitzsch) Ehrenberg (divergens-group) (divergens-group) (gibba-group) (gibba-group) (subcommutata-group) (subcommutata-group)

P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . australogibba P. P. s p . P. s p . subanglica P. subcapitata P. subcommutata P. P. substreptoraphe viridiformis P. P. v i r i d i f o r m i s P. viridis Pinnularia acrosphaeria blackfordensis Sellaphora blackfordensis Sellaphora (Enc2)b (Enc2)a (Wie)c Pin 870 MG Strain(Tor4)r Taxon (Wie)a PinnC7 (Tor1)b (Tor7)c (Tor7)f (Bfp04)02 (Tor8)b (W045)b Pin 883 Pin 650 K Pin 649 K Corsea 10 (Tor4)i Pin 873 TM (Bfp5x8)F1-3

Time-calibrated phylogeny of Pinnularia 83 μ m 10 μ m) # stria per and morphometric C. Souffreau 151.4 ± 3.6 ± 0.4 24.9 8.2 ± 0.4 C. Souffreau ± 0.5 17.9 ± 0.3 4.5 20.0 ± 0.0 D.G. MannD.G. MannD.G. 41.9 ± 0.8 105.3 ± 3.9 ± 0.2 7.9 18.5 ± 0.5 13.0 ± 0.8 8.1 ± 0.2 C. Souffreau 22.8 ± 1.3 10.6 ± 0.4 12.1 ± 0.3 C. Souffreau ± 0.4 27.3 5.4 ± 0.4 15.6 ± 0.7 Collector Length ( μ m) Width ( Podkova (Czech Republic) – 19.IV.2006 Republic) (Czech Podkova A. Poulícková ± 1.3 37.4 8.2 ± 0.6 ± 0.3 4.6 Threipmuir Resr. (Scotland) – 02.III.2008 Resr. Threipmuir Mann D.G. 33.4 ± 1.1 ± 0.3 7.4 12.1 ± 0.3 Threipmuir Resr. (Scotland) – 02.III.2008 Resr. Threipmuir Mann D.G. 38.9 ± 2.6 11.5± 0.8 10.0 ± 0.0 Glen Corse Resr. (Scotland) – 06.IV.2008 Corse Resr. Glen Vanormelingen P. 169.2 ± 1.9 28.4 ± 0.5 ± 0.3 7.8 Balerno (Scotland) – 02.III.2008 Loch (Scotland) – 23.I.2008 Figgate (Scotland) – 02.III.2008 Resr. Threipmuir Mann D.G. 58.3 ± 1.5 9.8 ± 0.3 10.0 ± 0.0 De Panne (Belgium) – 03.III.2008De Panne C. Souffreau ± 0.1 8.7 3.9 ± 0.2 22.4 ± 1.9 Torres del Paine (Chile) – 30.I.2007 del Paine Torres C. Souffreau ± 0.6 27.3 9.6 ± 0.3 6.0 ± 0.0 Seno Otway (Chile) – 02.II.2007 Seno Otway C. Souffreau 36.8 ± 1.6 9.5 ± 0.5 5.0 ± 0.0 Torres del Paine (Chile) – 30.I.2007 del Paine Torres Ailfroide (France) – 23.IX.2006 (France) Ailfroide C. Souffreau 33.0 ± 1.7 8.1 ± 0.3 5.3 ± 0.4 Locality and date of collection – I.2006 Peninsula) (Antarctic Beak Island E. Verleyen 38.9 ± 1.4 9.9 ± 0.3 ± 0.4 14.0

borealis subislandica

permitis W. Smith W. (Chile) – 22.I.2007 Ovalle var. var. (Ehrenberg) Cleve Cleve (Ehrenberg) Lange-Bertalot (Chile) – 01.II.2007 del Paine Torres Krammer (Ehrenberg) Cleve (Ehrenberg) (Scotland) – 02.III.2008 Resr. Threipmuir Mann D.G. 48.8 ± 3.7 13.0 ± 0.9 ± 0.5 17.4 Krammer Ilka Schönfelder – 22.IX.2006 (France) Ailfroide K. Krammer W. Smith W. Krammer Krammer Krammer Krammer Ehrenberg var. Ehrenberg cf. var. Ehrenberg var. Ehrenberg (Ehrenberg) W. Smith W. (Ehrenberg) (Scotland) – 02.III.2008 Resr. Threipmuir Mann D.G. 42.1 ± 1.2 ± 0.2 6.7 10.0 ± 0.0 (Hustedt) Lange-Bertalot Krammer Krammer Caloneis silicula atomus Mayamaea Pinnularia acrosphaeria acuminata P. borealis P. borealis P. subislandica borealis P. borealis P. altiplanensis cf. P. isselana cf. P. marchica cf. P. microstauron cf. P. (“southern microstauron”) grunowii P. neglectiformis P. neomajor P. neomajor P. P. n o d o s a parvulissima P. Pin 706 F Cal 890 TM (Enc2)b (Tor11)b (Ecrins7)a (Tor12)d Pin 885 TM (Tor3)a (Tor1)a Pin 877 TM Pin 876 TM Strain Taxon Pin 889 MG (Ecrins4)a (Wes2)f Alka 1 Cal 878 TM (B2)c Corsea 2 measurements (length, width and stria density from 10 valves per taxon, ± standard deviation, n.m. = not measured). measurements (length, width and stria density from 10 valves Table 4.2 : List of the 38 taxa cultured and sequenced for this study, with source locality date collection, identity collector Table

84 Time-calibrated phylogeny of Pinnularia μ m 10 μ m) # stria per C. SouffreauC. Souffreau 61.2 ± 1.6 40.0 ± 0.5 8.9 ± 0.2 5.9 ± 0.3 10.1 ± 0.5 11.8 ± 0.4 C. SouffreauC. Souffreau 51.5 ± 2.1 88.4 ± 1.7 10.4 ± 0.3 12.5 ± 0.4 10.8 ± 0.8 10.4 ± 0.5 C. SouffreauC. Souffreau 33.3 ± 1.3 42.6 ± 0.4 9.4 ± 0.5 5.8 ± 0.3 ± 0.0 14.0 12.4 ± 0.5 D.G. MannD.G. 80.5 ± 2.3 ± 0.3 16.7 9.2 ± 0.3 C. Souffreau ± 1.8 74.9 ± 0.5 17.7 8.8 ± 0.4 Collector Length ( μ m) Width ( Torres del Paine (Chile) – 30.I.2007 del Paine Torres (Chile) – 31.I.2007 del Paine Torres – 02.XII.2007 Billabong (Australia) Kew (Scotland) – 02.III.2008 Resr. Threipmuir K.M. Evans Mann D.G. 53.0 ± 3.2 63.1 ± 2.8 ± 0.3 11.7 ± 0.1 14.1 10.1 ± 0.2 10.3 ± 0.5 Ovalle (Chile) – 22.I.2007 Ovalle Balerno (Scotland) – 02.III.2008 Kew Billabong (Australia) – 02.XII.2007 Billabong (Australia) Kew K.M. Evans 46.2 ± 0.3 8.2 ± 0.2 11.1 ± 0.7 Glen Corse Resr. (Scotland) – 06.IV.2008Glen Corse Resr. Vanormelingen P. 50.1 ± 1.8 ± 0.1 10.7 11.8 ± 0.4 Torres del Paine (Chile) – 31.I.2007 del Paine Torres (Chile) – 31.I.2007 del Paine Torres Amsterdam Island – 06.XII.2007Amsterdam Island de Vijver Van B. ± 0.8 37.1 ± 0.2 8.7 11.2 ± 0.6 Locality and date of collection collection BCCM/DCGCulture collection BCCM/DCGCulture - - 38.1 ± 13.6 9.0 ± 1.4 n.m. 20.2 ± 1.1 n.m. n.m. Torres del Paine (Chile) – 31.I.2007 del Paine Torres (Chile) – 31.I.2007 del Paine Torres De Wieden (The Netherlands) – 26.II.2007 (Scotland) – 02.III.2008 Resr. Threipmuir Vanormelingen P. – XII.2004 (Sub-Antarctica) Iles Crozet ± 0.9 46.7 Mann D.G. ± 0.2 9.7 de Vijver Van B. 28.3 ± 2.8 11.8 ± 0.8 12.1 ± 0.2 10.2 ± 0.4 6.3 ± 0.3 12.3 ± 0.5 13.8 ± 0.6

Van Krammer De Wieden (The Netherlands) – 26.II.2007 Vanormelingen P. 53.3 ± 0.6 6.3 ± 0.3 11.9 ± 0.3 D.G. Mann D.G. Mann D.G.

nonfasciata subcapitata

elongata var. var. var. var. var. var. Krammer Krammer Krammer (divergens-group) (divergens-group) (gibba-group) (gibba-group) (subcommutata-group) (subcommutata-group) Krammer de Vijver, Cahttová & Metzeltin Cahttová de Vijver, P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . P. s p . australogibba P. P. s p . P. s p . subanglica P. subcapitata P. subcommutata P. P. v i r i d i f o r m i s P. v i r i d i f o r m i s blackfordensis Sellaphora Droop & S. blackfordensis Sellaphora Droop & S. (Tor4)i (Bfp04)02 (Enc2)a Pin 649 K Corsea 10 Strain Taxon (Bfp5x8) F1-3 Pin 873 TM PinnC7 (Tor1)b (Tor4)r (Wie)a Pin 870 MG (Wie)c Pin 883 TM Pin 650 K (Tor8)b (W045)b (Tor7)c (Tor7)f

Time-calibrated phylogeny of Pinnularia 85 Model testing and phylogenetic analyses Nine alternative partitioning strategies were tested: partitioning into genes, codon positions, stem and loop regions of rDNA regions, functionality and some combinations of these. Selection of a suitable ȱ ¢ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ based on the Bayesian Information Criterion (BIC, Schwarz, 1978). For each partitioning strategy six substitution models were optimized (JC, F81, K80, HKY, SYM, GTR) with or without a proportion of ȱ ȱ Ȧȱ ȱ ȱ ȱ ȱ ȱ rate variation across sites. All parameters were unlinked between partitions. The preferred model and partitioning strategy was a GTR Ƹȱ̆Śȱȱ ȱŗŞȱȱŘŞȱȱȱȱȱȱǰȱ while plastid and mitochondrial genes were separated by genome and were both partitioned into three codon positions. Single genes and the complete concatenated dataset were analysed by maximum likelihood phylogenetic inference using RAxML 7.2.6 (Stamatakis, 2006) under the preferred model and partition strategy with 10,000 independent tree searches from randomized MP starting trees. Maximum likelihood bootstrap analyses (Felsenstein, 1985) consisted of 1,000 replicates. Bayesian phylogenetic inference was carried out with MrBayes v.3.1.2 (Huelsenbeck & Ronquist, 2001; Ronquist & Huelsenbeck, 2003) under the same model and partitioning scheme. Using default Ĵȱǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȬȱȬȱȱȱ ȱȱȱęȱȱ million generations for individual genes and 70 million generations for the concatenated dataset. Runs were sampled every 1,000th generation, and convergence and stationarity of the log-likelihood and parameter values was assessed using Tracer v.1.5 (Rambaut ǭȱǰȱŘŖŖŝǼǯȱȱęȱȱȱȱȱŗŖȱȱ generations were discarded as burn-in for the single-gene and multi-gene analyses, respectively. Runs for psbA converged only when using a heating factor of 0.1. We therefore also analysed the concatenated dataset with a heating factor of both 0.2 and 0.1 but ȱ ěȱȱ ȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱǯȱȱȬȬȱȱȱȱěȱȱȱ ȱ summarized and posterior probabilities (PPs) were calculated in MrBayes using the sumt command. All analyses were performed using the computer cluster Bioportal (Kumar et al., 2009).

86 Time-calibrated phylogeny of Pinnularia Fossil Pinnularia In order to assist with the temporal calibration of the multi-gene phylogeny, we document a new fossil occurrence of the genus Pinnularia from Middle Eocene lacustrine facies in northwestern ǯȱ ȱ ěȱȱ ȱ ȱ ¢ȱ ǻŜŚɃŚŞȂǰȱ ŗŗŖɃŖŚȂǼȱ contains unpermineralized diatom-rich organic sediments, which ȱȱȬȱęȱȱȱȱȂȱȱǯȱȱ sediments are dated between 40 and 48 Ma (Lutetian Stage), and ȱ¢ȱȱęȱȱȱȱȱȱȱ ȱ diatom lineages (Siver & Wolfe, 2007; Wolfe & Siver, 2009; Siver et al.ǰȱŘŖŗŖǼǯȱȱ¡ȱȱȱ ěȱȱPinnularia specimens is illustrated by a range of light and scanning electron micrographs (Fig. 4.3).

Relaxed molecular clock analysis A time-calibrated phylogeny was inferred using a relaxed molecular clock method as implemented in BEAST v.1.5.4 (Drummond & Rambaut, 2007). An uncorrelated lognormal clock model and Yule ȱȱ ȱęȱȱȱ ȱȱȱȱȱ and models of sequence evolution used for the phylogenetic ǯȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ record were carried out to assess congruence between calibration points. The root node calibration prior was varied (uniform, gamma and truncated normal probability distribution), and all internal calibration points had a uniform prior (Ho & Phillips, 2009). All calibration points and strategies are summarized in Table 4.3 and fossil constraints are explained below. For each calibration strategy, three independent runs were carried out, starting from a user- ęȱǰȱ£ȱȬȱȱȱȱȱŞȱ (Sanderson, 2003) based on the ML tree of the phylogenetic analysis. Markov chains were run for 50 million generations and sampled every 1,000 generations. Convergence and stationarity of log-likelihood and parameter values were assessed using Tracer v.1.5 (Rambaut & Drummond, 2007) and 10% of the generations were discarded as burn-in. The post-burn-in trees of three independent runs were combined, after which a maximum clade credibility chronogram with mean node heights was calculated with TreeAnnotator v.1.5.4. Fossil diatoms were used to calibrate the tree in geological time using the following constraints. The root node of Pinnularia was given a minimum age of 40 Ma following the recovery of Pinnularia

Time-calibrated phylogeny of Pinnularia 87 Table 4.3: List of calibration points and alternative calibration schemes used to date the phylogenetic tree, with indication of the minimum (min) and maximum (max) age ǯȱȱȱȱȱȱȱȱȱ¢ȱȱǻǯȱŚǯŚǼǯ

Calibration scheme with age node Constraints (Ma) Bauplan Reference root A B C E min 40 40 40 40 40 1 root node Pinnularia spp. this study max 75 75 75 75 100

neomajor- min - 11.7 11.7 11.7 11.7 Saint Martin & Saint 2 neglectiformis P. viridis Martin (2005) clades max - 40 40 40 40

nodosa min - 14.5 14.5 14.5 14.5 pers. comm. A. 3 P. n o d o s a acrospaeria max - 40 40 40 40 Menicuci

grunowii min - - 11.7 11.7 - Saint Martin & Saint 4 P. mesolepta mesolepta max - - 40 40 - Martin (2005) min - - - 13 - borealis Servant-Vidary et 5 P. borealis microstauron max - - - 70 - al. (1988)

ȱ ȱ ȱ ȱ ěȱȱ ¢ȱ ǻǯȱ ŚǯřǼǯȱ ȱ ěȱȱ ¡ȱȱȱ ȱǯȱȱęȱȱ¡ȱȱȱ the root node was set at 100 Ma (Albian, Lower Cretaceous), based ȱȱȱȱȱȱĚȱȱȱȱȱȱ sediments containing a range of centric morphologies but neither raphid nor araphid pennate taxa (Gersonde & Harwood, 1990; Harwood & Gersonde, 1990). This date can be assumed to be highly conservative, and therefore we also considered an alternate upper boundary for Pinnularia at 75 Ma (Campanian, Upper Cretaceous), the period for which the earliest araphid pennate diatoms have ȱęȱȱǻǰȱŗşŜŜǲȱ àȱǭȱǰȱŗşŝśǼǯȱȱ these fossils are abundant but not diverse, it is believed that araphid pennate forms had evolved only recently by this time (Sims et al., 2006). Following this reasoning, we assume that raphid pennates had not yet evolved. Four internal calibration points were also used (Fig. 4.4), ȱ ȱ ęȱȬȱ ȱ ȱ ȱ ȱ types or Baupläne recovered from late-Oligocene to mid-Miocene freshwater deposits. The following morphological types were used: diatoms morphologically similar to extant P. viridis (11.7 Ma; Saint Martin & Saint Martin, 2005), P. mesolepta (11.7 Ma; Saint Martin & Saint Martin, 2005), P. nodosa (14.5 Ma; pers. comm. A. Menicucci) and P. borealis (13.0 Ma; Servant-Vildary et al., 1988). These fossil occurrences were used as minimum estimates for the

88 Time-calibrated phylogeny of Pinnularia Fig. 4.3 Middle Eocene Pinnulariaȱȱȱȱ ěȱȱȱȱȱ¢ȱ with a minimum age of 40 Ma. In light microscopy, valve (a-c) and girdle (d) views illustrate morphologies consistent with the genus as currently circumscribed. Under scanning electron microscopy, the valve overview (e), close-ups of proximal (f) and distal external ȱȱǻǼǰȱȱ ȱȱȱȱȱȱǻȬǼȱȱȱǻǼȱȱȱȱęȱȱ ȱĜȱȱ¢ȱ ȱ¡ȱǯȱȱȱȱȱȱȱȱǻǼȱȱ directly comparable with numerous modern congeners, here shown is a representatives of the P. subgibba subclade [(Tor4)r] in (l). Scale bars are 10 μm (a-d), 20 μm (e), 5 μm (f,i,j), and 2 μm (g,h,k,l).

Time-calibrated phylogeny of Pinnularia 89 Fig. 4.4 Phylogenetic relationships within the genus PinnulariaȱȱȱȱęȱȬȱ DNA alignment using maximum likelihood under a partitioned model. Numbers at nodes indicate statistical support, ML bootstrap proportions - BI posterior probabilities (both given as percentages). Encircled numbers represent nodes constrained in the relaxed molecular clock analysis (see Table 4.3).

90 Time-calibrated phylogeny of Pinnularia corresponding Bauplan, thus constraining the MRCA (most recent common ancestor) of clades containing species within each of these ǻǯȱŚǯŚǼǯȱȱ¡ȱȱȱȱěȱȱ§ȱ ȱȱȱ 75 Ma when using the conservative calibration scheme, and 40 Ma for the less conservative calibration strategy. Although specimens originating from the Wagon Gap Formation clearly belong to Pinnularia (Lohman & Andrews, 1968), in the absence of electron ȱȱ¢ȱȱȱȱ ȱęȱȱ to any of the modern morphological types used in our phylogeny. Linked to the uncertain dating of this locality, Wagon Gap Pinnularia were not used as an explicit constraint for our phylogeny.

4.3 Results Dataset properties ȱŚŚȬ¡ǰȱęȱȬȱȱȱŞŝƖȱęȱȱǻŗşŘȱǲȱȱ 4.1) and includes 4852 sites of which 1012 (21%) are parsimony- informative. The most complete markers are 28S rDNA and rbcL (0 of 44 missing), followed by 18S rDNA (4 sequences or 9% missing), psbA (8 or 18% missing) and cox1 (16 or 36% missing). The 18S rDNA sequences provided most characters (1706) followed by rbcL (1386), psbA (762), cox1 (615) and 28S rDNA (383). The percentage of parsimony-informative characters varied between genes, with the highest percentages for 28S rDNA (42%) and cox1 (37%), the lowest for psbA (10%), and 18S rDNA (18%) and rbcL (17%) being intermediate.

Phylogenetic relationships ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ěȱǲȱ therefore we show only the results of the concatenated analysis. Similarly, ML and BI analyses of the concatenated dataset produced identical topologies so we show only the ML phylogeny with both ML bootstrap support (BS) and BI posterior probabilities (PP) (Fig. 4.4). Based on our concatenated dataset, taxa representing Pinnularia and Caloneis form a monophyletic group comprising three robustly supported clades that each contain several well-supported subclades. Clade A comprises Caloneis silicula, C. lauta, and two species of Pinnularia cf. divergens. No apparent morphological synapomorphies unite these taxa, Caloneis silicula being characterized by linear external central raphe endings (Fig. 4.2, a), small alveolar apertures

Time-calibrated phylogeny of Pinnularia 91 (Fig. 4.2, b) and parallel striae (Fig. 4.1, d), while the divergens-group has drop-like external central raphe endings (Fig. 4.2, c), large alveolar apertures (Fig. 4.2, d) and oblique striae (Fig. 4.1, e, f). Clade B includes small, linear Pinnularia species (Fig. 4.1, g–p) that have drop-like external central raphe endings (Fig. 4.2, c). Clade B is subdivided into three well-supported subclades: the grunowii, nodosa and subgibba subclades. The grunowii subclade contains P. grunowii, P. subanglica and P. marchica. Species in this group contain a single H-shaped plastid with two pyrenoids (Fig. 4.1, b’-c’), ǰȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ǻȱ ěȱȱ et al., unpubl. a, for detailed information) all other clades of the tree are characterized by two linear, strip-like plastids (Fig. 4.1, y–z). The nodosa subclade includes P. nodosa and P. acrosphaeria, which are both characterized by irregular wart-like structures on the external valve face (Fig. 4.1, j, k; detail in Fig. 2, c). The subgibba subclade ȱěȱȱȱȱP. subgibba, P. parvulissima, and P. subcapitata, all characterized by a combination of an elongated ȱȱ ȱȱȱǻȱȱȱ¢ȱęȱȱȱ ȱ silica) and a broad, non-porous, central area (fascia) (Fig. 4.1, l–p). Clade C is composed of two subclades. Subclade C1 contains Caloneis budensis and Pinnularia cf. altiplanensis (Fig. 4.1, x), but is poorly supported (BS=56; PP=87). Subclade C2 is further subdivided into moderately to highly supported groups, all of which have characteristic morphologies. The borealis-microstauron group includes the P. borealis species complex, which is easily recognizable by its relatively broad cells and coarse striae (Fig. 4.1, w), and P. microstauron with its relatively robust shape and wedge-shaped ends (Fig. 4.1, v). The remaining groups (the subcommutata and viridiformis clades) contain large, elongated-elliptical species, with almost parallel striae and small central areas (Fig. 4.1, q-t), undulate ¡ȱ ȱ ęȱǰȱ Ȭ£ȱ ȱ ȱ ǻǯȱ 4.2, f), and plastids without pyrenoids. The subcommutata group contains Pinnularia species with linear central raphe endings (Fig. 4.2, e) while species of the viridiformis group have round central raphe endings (Fig. 4.2, g) and a complex raphe system, visible as three parallel (Fig. 4.1, q) or twisting lines (Fig. 4.1, r) with light microcopy (Round et al., 1990; Krammer, 1992, 2000). Sister to the viridiformis and subcommutata groups is the isolate (Wie)a (Fig. 4.1, u), an undescribed Pinnularia species morphologically similar to P. microstauron. In the single-gene trees, (Wie)a is always retrieved in clade C, but with a closer (unsupported) relationship to the viridiformis-group based on

92 Time-calibrated phylogeny of Pinnularia 18S and 28S rDNA, or with a sister relationship to P. cf. divergens and P. borealis based on rbcL and psbA.

Newly-observed fossil Pinnularia specimens Pinnulariaȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ¢ȱ ȱȱĜȱȱ¢ȱȱȱȱȱȱ ǯȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ Ĝȱȱȱ numbers have been observed to document their morphology in light and scanning electron microscopy (Fig. 4.3). The specimens are characterized by an alveolate valve structure (Fig. 4.3, k), placing ȱ ęȱ¢ȱ ȱ ȱ ȱ Pinnulariaǯȱ ǰȱ ȱ ěȱȱ Pinnularia forms have large alveolar openings (Fig. 3, k), identical to the Pinnularia species from clades A, B, C1 and the microstauron- borealis group of clade C (e.g. Fig. 4.3, l; Fig. 4.2, d,h). However, the particular combination of morphological characters of the fossil specimens, including wart-like structures on the valves (similar to P. acrosphaeria), the very wide axial area (similar to P. acuminata) and the distinctly lanceolate or “naviculoid” shape (not encountered in our species) meant that we were not able to place these specimens within a known Pinnularia Bauplan in the present phylogeny nor, to our knowledge, with any currently described species. Their distribution in samples spanning tens of meters of core, typically associated with chrysophyte and euglyphid siliceous microfossils and benthic eunotioid diatoms, and their occurrence within the mudstone matrix discounts the possibility that they are contaminants. ȱȱȱ¢ȱȱȱ¢ȱȱ ěȱȱȱ ȱ ȱ ȱ ǻȱ ǭȱ ǰȱ ŘŖŖşǼǯȱ ȱ ǃŚŖȱ ȱ ȱ ǰȱ these Pinnularia specimens represent the earliest known forms of Pinnularia, providing us with a fossil calibration point to constrain the basal node of our phylogeny temporally.

Time-calibrated phylogeny ȱ ěȱȱ ȱ ȱ ¢£ȱ ȱ ȱ ȱ ȱ identical relationships to the RAxML and MrBayes analyses with comparable posterior probabilities. The average node ages and şśƖȱ ȱǻ ȱȱ¢Ǽȱȱȱȱěȱȱ calibration schemes are given in Fig. 4.5. Estimates using the internal calibration points based on the frustule morphologies of viridis, nodosa and borealis were consistent with estimates using only a root calibration. However, adding the internal calibration point based on

Time-calibrated phylogeny of Pinnularia 93 Fig. 4.5 Resulting mean node ages (indicated as “x”; in Ma) and 95% HPD (error bars; in Ma) for nine of the alternative calibration ȱ ǻȱ ȱ ŚǯřǼǰȱ £ȱ ȱ ȱ ěȱȱȱȱǻȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ŚǯŚǼǯȱ The preferred calibration scheme used for ǯȱŚǯŜȱȱȱȱ¢ǯȱȃȄȱȱ were only constrained on the root. Scheme C used a less conservative calibration, with a maximum of 75 Ma and all internal calibration points. Scheme E used a very conservative calibration with a maximum of 100 Ma and all internal calibration ǯȱȱĚȱȱȱȱȱȱ on the grunowii-mesolepta Bauplan is ¢ȱȱȱȱěȱȱ ȱȱ estimations based on “root” and “scheme C” or “scheme E”.

94 Time-calibrated phylogeny of Pinnularia grunowii-mesolepta resulted in larger 95% HPD intervals and higher average node ages (plus 5–12 Ma) at all nodes. Using only the root constraint, the grunowii-mesolepta node is estimated as being 3.5– 12.6 Ma. Constraining this node to have a minimum age of 11.7 Ma therefore pushes its own time-estimate, and that of all other nodes, back in time (Fig. 4.5). Using the conservative maximum root age of 100 Ma resulted in a 7–11 Ma higher average root node age and a 12–20 Ma higher 95% HPD maximum root age compared with the alternative 75 Ma maximum age boundary, but for the internal nodes, changes in mean age and 95% HPD were only 1–6 Ma. Furthermore, the use of ěȱȱ¢ȱȱȱȱȱȱǻǰȱ ȱȱȱǼȱȱȱȱěȱȱȱ average node ages and 95% HPD up to 6 Ma for the less-conservative 75 Ma schemes and up to 10 Ma for the conservative 100 Ma scheme. Because the Eocene fossil record is so fragmentary, we chose the conservative uniform prior probability distribution as the preferred calibration strategy. Fig. 4.6 presents our preferred relaxed molecular clock as applied to the Pinnularia phylogeny, constrained by all four internal calibration points (scheme C = viridis, borealis, nodosa and mesolepta), with the root age constrained between 40 and 75 Ma using a uniform distribution. This estimate places the origin of Pinnularia around 64 Ma, between the Campanian and Early Eocene [75–50]. The Pinnularia complex began to diversify between the Maastrichtian and Middle Eocene 60 Ma [73–45] and continued through to the Pliocene, with no ȱȱȱȱȱȱęȱȱȱ¢ȱȱ ǯȱ ȱȱ ȱ ȱ ęȱȱ ¢ȱ ȱ ȱ Ȯ Eocene (respectively, 49 Ma [65–33] and 52 Ma [64–38]), while clade B started to diverge between the Eocene and Oligocene (39 Ma [50– 28]).

4.4 Discussion Relationships within Pinnularia The current study presents a multi-gene phylogeny of the genus Pinnularia and extends the sampling of Bruder et al. (2008) three- ǯȱȱęȱȬȱ¢ȱ¢ȱȱ Ȭȱ¢ȱ tree, with Pinnularia and Caloneis resolved in a single monophyletic clade, in accordance with the results of Bruder and Medlin (2008) and Bruder et al. (2008). Our phylogeny shows that Caloneis (as

Time-calibrated phylogeny of Pinnularia 95 Fig. 4.6 Chronogram of the genus Pinnularia from a Bayesian relaxed molecular clock analysis performed with BEAST (Drummond and Rambaut, 2007), time-constrained by four internal fossil calibration points (viridis, borealis, nodosa and mesolepta) and a 40 Ma minimum and 75 Ma maximum root age with uniform probability distribution (scheme Ǽȱȱȱȱȱȱȱȱ ěȱȱȱǯȱȱȱȱȱȱȱ ages and grey bars represent 95% HPD (highest probability density) intervals.

96 Time-calibrated phylogeny of Pinnularia ¢ȱ ęȱǼȱ ȱ ȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ were recovered in various places among the Pinnularia species. ȱ ęȱȱ ȱ ȱ ȱ ȱ ¡ȱ ǻŗşŞŞǼǰȱ ȱ et al. (1990) and Mann (2001) who concluded that the current separation of Caloneis from Pinnularia was not supported by plastid and frustule structure. Cleve (1894) delimited Caloneis from Pinnularia using ȱ¢ǰȱȱȱĴȱȱȱȱǰȱȱ ȱȱǰȱȱȱȱȱȱȱęȱǰȱȱ and the presence of “longitudinal lines” (the margins of the alveolar apertures on the inner side of the valve) was not consistent. While Krammer & Lange-Bertalot (Krammer & Lange-Bertalot, 1986; Krammer, 1992) argued that Caloneis can indeed be distinguished from Pinnulariaȱȱȱȱęȱȱȱȱǰȱȱ et al. (1990) suggested that, although Pinnularia–Caloneis is a natural group characterized by the double-walled, chambered (alveolated) valve structure, any split of the Pinnularia–Caloneis complex would not follow the traditional boundary between the two genera. Despite the fact that our molecular phylogeny contained only three Caloneis species, there is no apparent genetic boundary between Pinnularia and Caloneis. Nevertheless, ultrastructural characters, e.g. ȱ£ȱȱȱǰȱȦȱȱȱȱ¢ȱȱ used, alongside molecular synapomorphies, to delimit groups or genera within this complex. However, much further genetic and morphological investigation of the Pinnularia–Caloneis complex ȱ ȱ ȱ ¢ȱ Ĵȱȱ ȱ ȱ ȱ ¢ȱ ȱ ǯȱ Alternatively, subdivision could be considered unnecessary, given the highly characteristic general morphology of the complex. Pinnularia is a morphologically diverse genus and the ęȱȱȱȱȱȱęȱȬȱȱ¢ȱ ȱ ȱ Ĵȱǯȱ ȱ ȱ ȱ ȱ ǰȱ well-supported clades that were also recovered with lower or no support in the concatenated (18S, 28S, rbcL) dataset of Bruder et al. (2008). With the exception of clade A, which contains two Caloneis species and the Pinnularia divergens group, the remaining clades are ¢ȱ ȬęȱǯȱȱȱȱȱPinnularia taxa of intermediate size that have a fascia, and is subdivided into three ¢ȱȱȱ¢ȱ Ȭęȱȱǯȱȱ structural similarities between P. acrosphaeria and P. nodosa have been described by Cleve (1895) and more recently by Krammer & Lange-Bertalot (1986). The grunowii and subgibba subclades are also Ȭęȱȱ¢ȱ Ȭȱȱȱȱ ȱǰȱ¢ǯȱ

Time-calibrated phylogeny of Pinnularia 97 Clade C contains the larger and more robust forms of Pinnularia, and is subdivided into well-supported subclades, each with a distinct morphology. The combined viridiformis and subcommutata group, characterized by intermediate alveolar openings (Fig. 4.2, f), was previously described by Cleve (1895) as the Maiores and Complexae, and more recently as a single group by Krammer & Lange-Bertalot (1986). ȱ ȱ Ȭęȱȱ ȱ Ȭȱ ǰȱ ȱ clusters are less easily interpreted. Some enigmatic relationships within clade A can only be further resolved and understood through additional taxon sampling. This is also the case for the positions of the undescribed species (Wie)a and P. cf. altiplanensis, and the relationship between P. microstauron and P. borealis. Because only a small subset of the described Pinnularia species were included in this analysis, it is plausible that they represent more elaborate clusters. Adding taxa in those parts of the tree could therefore improve our understanding of the evolution of morphological features.

Evolution of the genus Pinnularia Based on our preferred time-calibration, Pinnularia originated between the Campanian (Late Cretaceous, ca. 75 Ma) and Early Eocene (ca. 50 Ma). This estimate is 10 to 35 Ma older than the ca. 40 Ma origin deduced from the chronogram of Sorhannus (2007). ǰȱ ȱ ȱ ȱ ȱ Ĵȱȱ ¢ȱ ȱ ȱ ȱ ȱ divergences within major diatom lineages, and furthermore only two Pinnularia taxa were included and analyzed with a single genetic marker (nuclear-encoded SSU rDNA). More comprehensive taxon sampling (Sanderson, 1990), the use of multiple genes (Rodríguez-Trelles et al., 2003), and full integration of the fossil record (Donoghue & Benton, 2007) have all been shown to improve the accuracy of molecular dating. In the case of PinnulariaǰȱȱǃŚŖȱ ȱ ěȱȱȱ¢ȱȱȱȱȱȱȱ estimated ages for the origination of the genus (i.e. Late Cretaceous). ȱ ȱ ȱ ¢ȱ ȱ ěȱȱPinnularia forms to ¡ȱ¡ǰȱ ȱȱęȱȱȱȱȱȱ¡ǯȱȱȱ ěȱȱȱǰȱȱǰȱȱȱȱȱ Ȭ dated sediments, they do not inform the search for morphologically primitive Pinnularia, or its immediate ancestor. The new estimate for the origin of Pinnularia implies a considerable gap in the fossil record of freshwater diatoms. The magnitude of this ghost lineage (ca. 10-35 Ma) should not be

98 Time-calibrated phylogeny of Pinnularia surprising given that freshwater deposits are underrepresented and understudied in the palaeontological record. Similar gaps apply to nearly all extant raphid diatom lineages, and the scarcity of fossil raphid diatoms of Late Cretaceous to Paleocene age has resulted in a lack of consensus on plausible minimum ages for most genera (but see Singh et al.ǰȱŘŖŖŜǼǯȱ ȱȱȱĜȱȱȱȱȱȱȱ age for Pinnularia in a broader context, although generally our observations are consistent with the preferred model of Berney & Pawlowski (2006), in which pennate diatoms arose some 98 (110– 77) Ma ago, leaving the entire Late Cretaceous for innovations such as the raphe. Apart from Pinnularia, no raphid diatom genera have yet been dated by phylogenetic methods, but the time-calibrated phylogeny of Sorhannus (2007) suggests that several raphid genera are as old as, or even older than, our estimate for Pinnularia. Given the available fossil and phylogenetic records, our estimated timing of 60 Ma for the origin of Pinnularia seems plausible. Since its origin, Pinnulariaȱȱęȱȱ ¢ȱȱȱ large number of species and a range of morphologies. The evolution of particular morphological innovations appears to have sparked the appearance of new lineages, such as the evolution of the “complex” raphe, which may have allowed the development of larger cells, as seen in the viridiformis group, or the evolution of H-shaped plastids in the grunowii group. Further research will be needed to integrate the ecological, morphological, genetic and evolutionary processes ¢ȱȱęȱȱȱȱȱȱȱPinnularia. Our time-calibrated multi-gene phylogeny of Pinnulariaȱȱȱęȱȱ important step towards this goal by providing a temporal context in ȱȱȱȱęȱȱȱȱǰȱȱȱ¢ȱȱ possible by integrating molecular and palaeontological information.

Acknowledgments ȱ ȱ ȱ Ç²¤ȱ ȱ ȱ ȱ ȱ Pinnularia borealis and Elie Verleyen for providing live environmental samples. Alex Ball enthusiastically assisted with confocal microscopy. Sequencing was done by Andy Vierstrate. Funding was provided ¢ȱȱȱȱęȱȱȬȱǻȬǲȱȱ fellowship to CS, post-doctoral fellowships to HV and PV, and ȱȱř ȦŖśřřȦŖŝǼǯȱȱȱȱȱȱ¢ȱ supported by the U.S. National Science Foundation (DEB-0716606 to PAS) and the Natural Sciences and Engineering Council of Canada

Time-calibrated phylogeny of Pinnularia 99 (APW). Morphological analyses at the Natural History Museum, London, were supported by an EU framework 7 SYNTHESYS grant (GB-TAF-77) to CS.

100 Time-calibrated phylogeny of Pinnularia Appendix 4.A DNA regions, primer sequences and PCR conditions. F= forward, R= reverse.

DNA region Primer sequence PCR protocol (°C, min:sec) 18S rDNA P2F: CTG GTT GAT TCT GCC AGT 94°C, 3:00 + 40 x (94°C, 1:00 – P4R : TGA TCC TTC YGC AGG TTC 55°C, 1:00 – 72°C, 1:50) + 72°C, AC 10:00 P12: CGG CCA TGCACC ACC P14: CGG TAA TTC CAG CTC C P15 13: TTT GAC TCA ACA CGG G P16: GWA TTA CCG CGG CKG CTG 28S rDNA DIR-F: ACC CGC TGA ATT TAA GCA 95°C, 5:00 + 35 x (94°C, 1:00 – TA 45°C, 1:00 – 74°C, 1:00) + 72°C, D2C-R: CCT TGG TCC GTG TTT 10:00 CAA GA rbcL DPrbcL1F: AAG GAG AAA THA ATG 94°C, 3:00 + 40 x (94°C, 1:00 – TCT 55°C, 1:00 – 72°C, 1:50) + 72°C, DPrbcL7R: AAC AAC CTT GTG TAA 5:00 GTC TC rbcL13F: CGT TTA GAA GAT ATG CGT ATT C NDrbcL6F: GTA AAT GGA TGC GTA NDrbcL12R: GCA CCT AAT AAT GG 15R: ACA CCA GAC ATA CGC ATC CA 17R: TGA CCA ATT GTA CCA CC psbA psbAF: ATG ACT GCT ACT TTA GAA 94°C, 3:00 + 35 x (94°C, 1:00 – AGA CG 46°C, 1:00 – 72°C, 1:00) + 72°C, psbAR: GCT AAA TCT ARW GGG 10:00 AAG TTG TG cox1 GazF2: CAA CCA YAA AGA TAT WGG 95°C, 3:00 + 35 x (95°C, 0:30 – TAC 50°C, 1:00 – 72°C, 1:50) + 72°C, KedtmR: AAA CTT CWG GRT GAC 5:00 CAA AAA

Time-calibrated phylogeny of Pinnularia 101

5. Phylogenetic signals in the valve and plastid morphology of the diatom genus Pinnularia

ȱěȱ1, Eileen J. Cox2, Koen Sabbe1 and Wim Vyverman1

ǐȱ 1 Laboratory of Protistology and Aquatic Ecology, Ghent University, Krijgslaan 281 - S8, B-9000 Gent, Belgium ǐȱ 2 Department of Botany, The Natural History Museum, Cromwell Road, London SW7 5BD, United Kingdom

Unpublished manuscript

Morphology of Pinnularia 103 Abstract The diatom genus Pinnularia is characterized by broad variation in frustule and plastid morphology, making it an interesting genus in which to analyze the phylogenetic signal of these characters ȱ ȱ ȱ ęȱȱ ǯȱ ȱ ȱ ȱ cladistic analysis on 48 strains using 17 frustule and 3 plastid characters and made a comparison with a multi-gene phylogeny of the same taxa. Independent phylogenetic analyses of molecular and morphological data generated similar clades, and Mantel ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱ ¢ȱ ȱ ȱ ęȱȱ molecular markers containing in total 1012 parsimony-informative sites were fully resolved, the lower number of phylogenetically informative morphological characters inevitably resulted in fewer clades being recovered, and we were never able to fully resolve the ęȱȱȱȱPinnularia based on morphology. The ȱ ȱ ȱ ¢Ȭȱ ȱ ěȱȱ ȱ acquired by analyzing all morphological characters simultaneously. Mainly “non-structural” frustule characters were found to be homoplasious while “structural” characters (such as central raphe endings, raphe complexity and alveolus openings) and one plastid character (number of pyrenoids) were informative for reconstructing the phylogenetic relationships within the genus. Some problems of homology were encountered and are discussed.

Key words: ancestral state reconstruction, Caloneis, cladistics, nuclear genes, Pinnularia, phylogeny, plastid genes, plastid morphology, valve morphology

104 Morphology of Pinnularia 5.1 Introduction ȱ ¡¢ȱ ȱ ęȱȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ Ȭęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ revealed by light and scanning electron microscopy (Round et al., 1990). To a lesser extent, cytological information, such as chloroplast morphology and organelle location, has been incorporated into diatom systematics (Mereschkowsky, 1902; Cox, 1981; Round et al., 1990; Cox & Williams, 2006). Genetic information in the form of nucleotide sequences from (parts of) faster and slower evolving genes are increasingly being used to infer relationships in this microalgal group (reviewed in Alverson, 2008), but as with morphology and cytology, there is the risk that some level of parallel evolution and convergence in base composition obscures the real relationships. For DNA nucleotides, the knowledge on their biological function (i.c. the coding of amino acids used to construct proteins) and chemical properties (e.g. the stronger hydrogen bound between guanine and cytosine than between adenine and thymine) allowed to model their evolution using mathematical language. By applying such models on a larger number of nucleotides originating from ěȱȱ ¡ǰȱ ȱ ¢ȱ ȱ ȱ ȱ ¡ȱ (including probability levels) can be estimated (Felsenstein, 2004). In contrast, morphological and cytological features are expected to be generated by the coordinated action of multiple genes (by analogy with higher plants, e.g. Mathur, 2006; Qian et al., 2009) making the selection of the used characters and character states, ȱ ȱȱěȱȱȱȱǰȱȱȱȱȱ gains and losses of character states more ambiguous, and thus more prone to interpretational (e.g.ȱȱȱěȱȱȱȱȱ ȱȱǰȱȱěȱȱǼȱȱ¢ȱǯȱ Moreover, the low number of morphological characters commonly used to infer relationships (see for example Edgar & Theriot, 2004; Bleidorn et al., 2009; Kooistra et al., 2010 and compare the number of morphological characters with the number of nucleotides) enhances the impact of potentially misleading characters on the outcome, making the reconstruction of phylogenetic relationships based on morphological datasets very challenging. Meanwhile, cladistic analyses have provided useful results at higher (intergeneric) (e.g. Kociolek & Stoermer, 1986; Cox & Williams, 2000; Marivaux et al., 2004; Cox & Williams, 2006; Ohl & Spahn, 2010) and lower (intrageneric) taxonomic levels (e.g. Julius & Tanimura, 2001; Uriz & Carballo, 2001; Edgar & Theriot, 2004; Kooistra et al.,

Morphology of Pinnularia 105 2010). In addition, phylogenetic analysis of morphological datasets can provide insights into the phylogenetic signal and resolution of ěȱȱȱ¢ȱȱȱȱǻ¡ȱǭȱǰȱ 2000; Maximino, 2008), reveal the evolution of morphological stages (Maximino, 2008) and the selective pressure behind morphological evolution (Stark & O’grady, 2010). Morphology is also the link that allows the integration of fossil taxa and large molecular datasets (Edgar & Theriot, 2004; Smith & Turner, 2005; Giribet, 2010), and thus ȱ ȱ ȱ ¢ȱ Ĵȱȱ ȱ ȱ ǻ¡ȱ et al.ǰȱ ŘŖŖŚǼǯȱ ¢ȱ ¢ȱ ȱ ȱ Ĵȱȱ ȱ from molecular data (e.g. Jacobs et al., 2008; Mizuno, 2008; Ivanovic et al., 2009) and, interestingly, molecular-phylogenetic analyses of diatom genera often support previous morphological observations (Lundholm et al., 2002; 2003; Edgar & Theriot, 2004; Lundholm et al., 2006; Chen et al., 2007; Sarno et al., 2007). There is also evidence ȱěȱȱȱȱȱȱ¢ȱȱȱ and morphological datasets can enhance the resolution (Wortley & Scotland, 2006) and overall probability of the phylogenetic signal (Lee & Camens, 2009). For all these reasons morphological, cytological, biochemical, ecological and developmental data remain important components for the inference of evolutionary relationships and understanding of evolutionary processes. Despite the recognized importance of using only phylogenetically informative data when reconstructing relationships, there is a lack of knowledge on the phylogenetic value of morphological characters in diatoms. Morphological data are mostly analyzed using cladistics, a technique that describes evolutionary history based on shared (homologous), derived characters (Hennig, 1965, 1966), but identifying homologous characters and ęȱȱ ȱ ȱ ȱ ȱ ¢ȱ ǯȱ ȱ ȱ ȱ to use positional, structural and developmental information to interpret potential phylogenetically informative characters (Cox, ŘŖŗŖǼȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ěȱȱ ¢ȱ (Cox, 2001) while contrasting features may share developmental pathways (Cox, 2002). Even if the systematics and cladistic analysis of diatoms have been dominated by frustule characters, cytological features such as shape and number of chloroplasts have also been used to infer relationships (Mereschkowsky, 1902; Kociolek & Stoermer, 1986; Round et al., 1990; Cox & Williams, 2000; Cox & Reid, 2004; Cox & Williams, 2006). Partitioned analyses of protoplast and frustule data have shown that both datasets are informative at

106 Morphology of Pinnularia ěȱȱȱǻ¡ȱǭȱǰȱŘŖŖŖǼǯȱȱȱȱPinnularia exhibits a broad variety of plastid morphology potentially useful for subgeneric groupings (Cox, 1988), as well as large variation in frustule characters (Round et al., 1990; Krammer, 2000), it is an ideal genus with which to analyze the phylogenetic signal of both valve and plastid characters at the infrageneric level using a cladistic approach. Moreover, the availability of a molecular-phylogenetic ȱȱȱȱȱȱęȱȱȱȱǻěȱȱet al., ĴȱǼȱ ȱȱȱȱȱȱȱ ȱȱ molecular one. The genus Pinnularia Ehrenberg contains biraphid, benthic diatoms and occurs world-wide, mainly in freshwater habitats. It was described by Ehrenberg (1843) based on Pinnularia viridis ǻĵȱǼȱȱȱȱȱȱȱȱȱȱǯȱ At the time of writing, Algaebase (Guiry & Guiry, 2011) includes 2462 species names and varieties of which 413 are currently accepted taxonomically, but in the light of recent discoveries of pseudocryptic species (Sarno et al., 2005; Mann & Evans, 2007; Vanormelingen et al., 2008) it is believed that the worldwide species diversity of Pinnularia could exceed this last number by almost ten-fold (pers. comm. Bart Van de Vijver). The fossil record of the genus dates back to the Middle-Eocene (ca.ȱŚŖȱȱǼȱǻȱǭȱ ǰȱŗşŜŞǲȱěȱȱ et al.ǰȱĴȱǼȱȱȱȱȱȱ ȱ diatoms yet recovered, and a fossil-calibrated molecular clock ȱȱȱȱȱŜŖȱȱǻěȱȱet al.ǰȱĴȱǼǯ Pinnularia is morphologically characterized by the possession of alveolate or “chambered” valves in which the striae (as seen using light microscopy) correspond to separate alveoli (chambers) (Round et al., 1990; Krammer, 2000). The outer wall of each chamber is a porous plate with many rows of small round poroids (occluded by hymenes), while the inner wall consists of a plain plate perforated by one (large or small) aperture. The raphe system is central and the ȱęȱȱȱȱǯȱPinnularia is readily distinguished from other diatom genera by these features, including the most closely related genera Sellaphora and Eolimna (Round et al., 1990; Bruder & Medlin, 2008; Evans et al., 2008), but not from Caloneis Cleve, which also has an alveolated valve (Round et al., 1990; ǰȱŗşşŘǰȱŘŖŖŖǼǯȱȱęȱȱȱCaloneis, Cleve (1894) noted the similarities with Pinnularia. The controversy still exists, and while Cox (1988), Round et al. (1990) and Mann (2001) consider it no longer possible to make a clear distinction along the traditional

Morphology of Pinnularia 107 boundary between these two genera, Krammer & Lange-Bertalot (1985, 1986) and Krammer (1992, 2000) consider that the combination of a lanceolate valve outline, parallel or radial orientation of the striae at the apices and marginal longitudinal lines separate Caloneis from Pinnularia. In molecular phylogenies (Bruder & Medlin, 2008; Bruder et al., 2008) Pinnularia and Caloneis were recovered as a single clade, but with low bootstrap support. The present study does not aim to resolve this issue, but we have added some strains of Caloneis to complete the sampling. Given its distinguishing morphological characteristics, Pinnularia exhibits considerable morphological diversity, including very small to very large forms (Round et al., 1990; Krammer, 2000). Valve shape varies from linear to undulate, apices can be rounded to capitate, and the raphe system can be simple and linear or very complex with a tongue and groove system visible as multiple parallel or twisting lines in LM (Round et al., 1990; Krammer, 1992, 2000). Plastid morphology varies from two girdle-appressed, linear chloroplasts per cell to a single H-shaped plastid with varying numbers of pyrenoids (Cox, 1988). Relationships within Pinnularia have never been resolved based on morphology. A number of ȱȱ£ȱěȱȱȱȱ ȱȱ genus based on valve size and outline and stria orientation (Cleve, 1895; Krammer & Lange-Bertalot, 1985, 1986), but these have not ȱȱȱȱęȱȱǽ¡ȱȱȱǭȱȱǻŗşŜŜǼǾȱȱ were only intended to group the broad variety of morphologies for ęȱȱ ǯȱ ȱ Ȭ¢ȱ ¢ȱ ȱ ȱŚŚȱ¡ȱǻěȱȱet al.ǰȱĴȱǼȱȱȱȱǰȱ ȱȱȱȱěȱǰȱ¢ȱȱǰȱ ȱ ȱ¢ȱȱȱ ¢ȱ¢ȱęȱǯ ȱ ȱȱȱȱ¢ȱ ȱȱ¢£ȱȱ¢ȱȱěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ relationships in the genus Pinnularia and to compare this with the ȱȱȱȱęȱȱȱȱȱěȱ et al. ǻĴȱǼǯȱǰȱ ȱȱȱȱ¢ȱȱȱȱȱ and plastid morphology to reconstruct evolutionary relationships in the genus Pinnularia and to identify which characters contribute most to the reconstruction of these relationships. Secondly, the congruence between the morphological and multi-gene phylogeny ǻěȱȱet al.ǰȱĴȱǼȱ ȱȱ¢ȱȱȱǻǰȱ 1967; Mantel & Valand, 1970). Finally, we performed ancestral state reconstructions of the morphological characters based on the multi-

108 Morphology of Pinnularia ȱ¢¢ȱȱěȱ et al.ȱǻĴȱǼȱȱȱȱ results with the available fossil record.

5.2 Materials & Methods 6WUDLQLVRODWLRQDQGLGHQWL¿FDWLRQ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ¢ȱ Ĵȱȱ ȱ ¢ȱ the lens-tissue technique from environmental samples of aquatic ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ between 2006 and 2008 (Table 5.1). All cultures were kept in liquid Ȭȱǽǻ ȱǭȱ£ǰȱŗşŝŘǼȱȱ ȱȱȱ ǾȱȱȱȱȱŗŞǚȱƹȱŘǚǰȱŘśȬřŖȱΐȱȱ m-² s-1 light intensity and a 12:12 hours light:dark period. Cultures were re-inoculated in fresh medium when reaching late exponential phase. For morphological and genetic characterization, cells were harvested in exponential phase and concentrated by centrifugation. ȱ ȱ ȱ £ȱ ȱ ȱ ȱ ȬŘŖǚȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ et al.ȱǻĴȱǼǲȱ ȱ ȱ morphological study by light microscopy (LM) were instantly oxidized in hydrogen peroxide and, after several washing steps, embedded in Naphrax®. Pictures of the cleaned valves were taken using a Zeiss Axioplan 2 microscope equipped with an AxioCam Mrm camera. Valve length, width and stria density of 10 valves per ȱ ȱ ȱ ȱ ȱ ŗǯřŝȱ ǯȱ ęȱȱ of the cleaned valves was based on Krammer (2000) for Pinnularia, and Krammer & Lange-Bertalot (1986) for Caloneis. Voucher slides of all natural samples and cultures are kept in the Laboratory of Protistology & Aquatic Ecology and are available upon request. To cover the wide range of morphological variation of the genus Pinnularia, a total of 48 morphologically dissimilar strains were selected for further analysis.

0RUSKRORJLFDOFKDUDFWHUL]DWLRQ The morphology of cleaned frustules was examined and photographed using both LM and scanning electron microscopy (SEM). For SEM, the oxidized material was coated with gold for 60 sec with a JEOL JFC-1200 Fine Coater and examined using a JEOL JSM-5600LV SEM. Cytological observations of plastids and pyrenoids were made on living cells of non-synchronized, exponentially growing strains using a Zeiss Axioplan 2 microscope equipped with

Morphology of Pinnularia 109 an AxioCam Mrm camera. To improve the visibility of the plastid shape and pyrenoids, Leica Confocal Laser Scanning Microscopy ǻǼȱ ȱȱȱ£ȱȱĚȱȱȱȱȱ ȱ ȱ ȱ ȱ ŘƖȱ ¢Ȭęȱ¡ȱ ǯȱ ¢ȱ ȱ ȱ ȱ ¡ȱ ȱ Ĵȱȱ ǻŗşŝŜǼȱ ȱ ęȱȱ £Ȭ ȱ solution (Sigma, Germany) prepared following Gerlach (1977).

&KDUDFWHUFRGLQJ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ criteria: characters must (i) show minimal variation within a taxon, (ii) be variable between taxa, (iii) be independent, (iv) be discrete, (v) be heritable (Pimentel & Riggins, 1987; Farris, 1990). Table 5.2 gives an overview of the 17 frustule and 3 plastid characters and their ȱȱǯȱȱȱȱěȱȱȱ are given with comments and micrographs in Appendix 5.1-3 and terminology is based on Ross et al. (1979), Cox & Ross (1981) and Round et al. (1990). Character states were coded from simple to complex. To assign the primitive states we did not use the outgroup method (Kociolek, 1986) because the intergeneric relationships of Pinnularia and Caloneis are not clear and character states of the non- alveolate genera Sellaphora, Eolimna or Mayamaea (see molecular ȱ ȱ ěȱȱ et al.ǰȱ ĴȱǼȱ ȱ ȱ ȱ ȱ those of Pinnularia to be workable. The ancestral states were thus based on the earliest known fossil material of Pinnularia described by Lohman & Andews (1968), and the same character states apply for the Middle Eocene Pinnulariaȱȱȱȱǻěȱȱet al.ǰȱĴȱǼȱȱȱȱȱȱȱȱȱ the cuneate apices. The taxon/character matrix is given in Table 5.3.

&ODGLVWLFDQDO\VLV Frustule characters included both “structural” features, such as type of raphe slit, extent and position of alveolus opening and girdle band areolae, which are expected to have a genetic background and thus phylogenetic value, and “non-structural” characters, such as valve outlines, shapes of hyaline areas and stria orientation, which ȱȱȱȱȱȱǻęȱȱȱȱ¢ȱ constrained) valve components but are independent of the type of these valve components (e.g. stria orientation is independent of the type of stria and pore construction, but will be constrained by the ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ Ǽȱ ǻ¡ǰȱ

110 Morphology of Pinnularia Table 5.1 Overview of the strains used in the morphological and molecular phylogenetic ¢ȱ ȱȱęȱǰȱȱȱȱǯ 1 Geographic location in case the cultures were isolated by the authors from environmental ǲȱȱȱȱǰȱȱȱ ȱȱȱȱȱǯ 2ȱȱȱ ȱȱȱŗŖȱ¡£ȱȱȱǰȱȱ are given as average ± the standard deviation. Strain Taxon Origin1 Length Width # striae (μm)2 (μm)b per 10 μmb % F 3FIPLFURVWDXURQ $QWDUFWLF3HQLQVXOD 38.9 ± 1.4 9.9 ± 0.3 14.0 ± 0.4 (“southern – Beak Island PLFURVWDXURQ”) (B2)g10 3FIPLFURVWDXURQ $QWDUFWLF3HQLQVXOD 40.8 ± 1.2 10.7 ± 0.3 14.1 ± 0.5 (“southern – Beak Island PLFURVWDXURQ”) (FULQV D 3FIPDUFKLFD Ilka )UDQFH±/HV(FULQV 27.3 ± 0.4 5.4 ± 0.4 15.6 ± 0.7 6FK|QIHOGHU (FULQV G &DORQHLVVLOLFXOD )UDQFH±/HV(FULQV 24.2 ± 2.1 8.3 ± 0.7 20.7 ± 1.0 (Ehrenberg) Cleve (FULQV D P. borealis Ehrenberg )UDQFH±/HV(FULQV 33.0 ± 1.7 8.1 ± 0.3 5.3 ± 0.4 (QF D P. v i r i d i f o r m i s Krammer Chile – Valle de 74.9 ± 1.8 17.7 ± 0.5 8.8 ± 0.4 (QFDQWR (QF E 3DFURVSKDHULD W. Chile – Valle de 22.8 ± 1.3 10.6 ± 0.4 12.1 ± 0.3 Smith (QFDQWR (No5)j P. isselana Krammer Norway – Bergen 31.2 ± 8.4 8.3 ± 0.4 12.0 ± 0.0 (Pinn P. s p . 6XE$QWDUFWLF,OHV 11.8 ± 0.8 6.3 ± 0.3 13.8 ± 0.6 C-Noel)7 &UR]HW (Tor1)a P. neomajor Krammer Chile – Torres del 151.4 ± 3.6 24.9 ± 0.4 8.2 ± 0.4 Paine (Tor1)b P. s p . (divergens- Chile – Torres del 51.5 ± 2.1 10.4 ± 0.3 10.8 ± 0.8 group) Paine (Tor1)g P. neomajor Krammer Chile – Torres del 102.8 ± 2.6 24.7± 1.3 8.3 ± 0.5 Paine (Tor11)b 3FIDOWLSODQHQVLV Chile – Torres del 17.9 ± 0.5 4.5 ± 0.3 20.0 ± 0.0 Lange-Bertalot Paine (Tor12)d P. borealis Ehrenberg Chile – Torres del 36.8 ± 1.6 9.5 ± 0.5 5.0 ± 0.0 Paine (Tor3)a P. borealis Ehrenberg Chile – Torres del 27.3 ± 0.6 9.6 ± 0.3 6.0 ± 0.0 Paine (Tor4)i P. s p . Chile – Torres del 33.3 ± 1.3 9.4 ± 0.5 14.0 ± 0.0 Paine (Tor4)r P. s p . Chile – Torres del 42.6 ± 0.4 5.8 ± 0.3 12.4 ± 0.5 Paine 7RU F P. s p . (divergens- Chile – Torres del 88.4 ± 1.7 12.5 ± 0.4 10.4 ± 0.5 group) Paine (Tor7)f P. s p . (gibba-group) Chile – Torres del 61.2 ± 1.6 8.9 ± 0.2 10.1 ± 0.5 Paine (Tor8)b P. s p . (gibba-group) Chile – Torres del 40.0 ± 0.5 5.9 ± 0.3 11.8 ± 0.4 Paine (Tor8)d P. s p . (gibba-group) Chile – Torres del 28.2 ± 0.5 5.6 ± 0.2 12.2 ± 0.6 Paine

Morphology of Pinnularia 111 Strain Taxon Origin1 Length Width # striae (μm)2 (μm)b per 10 μmb (Val1)b 3DFURVSKDHULD W. Chile – Valdivia 45.7 ± 1.3 10.4 ± 0.3 11.3 ± 0.5 Smith (W045)b P. s p . (gibba-group) 6XE$QWDUFWLFD 37.1 ± 0.8 8.7 ± 0.2 11.2 ± 0.6 Amsterdam Island (W123)a P. s p . 6XE$QWDUFWLFD 23.5 ± 1.2 6.8 ± 0.1 13.7 ± 0.3 Amsterdam Island (Wie)a P. s p . The Netherlands – 46.7 ± 0.9 9.7 ± 0.2 12.1 ± 0.2 De Wieden :LH F 3VXEFDSLWDWD var. The Netherlands – 53.3 ± 0.6 6.3 ± 0.3 11.9 ± 0.3 elongata Krammer De Wieden (Yde1)a 3VXEFRPPXWDWD Belgium – Ter Yde 62.3 ± 2.0 11.4 ± 0.4 10.1 ± 0.2 Krammer Alka 1 P. borealis Ehrenberg Prof. A. Pouli?ková 37.4 ± 1.3 8.2 ± 0.6 4.6 ± 0.3 Cal 769 B &DORQHLVVLOLFXOD Prof. D.G. Mann 27.5 ± 2.9 10.5 ± 0.5 18.6 ± 0.5 (Ehrenberg) Cleve Cal 856 L &DORQHLVVLOLFXOD Prof. D.G. Mann 27.4 ± 3.2 8.7 ± 0.3 19.4 ± 0.5 (Ehrenberg) Cleve Cal 878 TM P. FI. isselana Krammer Prof. D.G. Mann 33.4 ± 1.1 7.4 ± 0.3 12.1 ± 0.3 Cal 890 TM &DORQHLVVLOLFXOD Prof. D.G. Mann 48.8 ± 3.7 13.0 ± 0.9 17.4 ± 0.5 (Ehrenberg) Cleve Corsea 10 3VXEFRPPXWDWD var. 6FRWODQG 50.1 ± 1.8 10.7 ± 0.1 11.8 ± 0.4 QRQIDVFLDWD Krammer Corsea 11 3VXEFRPPXWDWD var. 6FRWODQG 51.0 ± 1.8 10.7 ± 0.1 11.8 ± 0.4 QRQIDVFLDWD Krammer Corsea 2 P. neomajor Krammer 6FRWODQG 169.2 ± 1.9 28.4 ± 0.5 7.8 ± 0.3 Corsea 6 3VXEFRPPXWDWD var. 6FRWODQG 46.5 ± 2.0 11.3 ± 0.3 11.5 ± 0.5 QRQIDVFLDWD Krammer Pin 12 TM P. parvulissima Prof. D.G. Mann 33.2 ± 2.4 8.9 ± 0.2 11.4 ± 0.5 Krammer Pin 649 K P. s p . VXEFRPPXWDWD Prof. D.G. Mann 53.0 ± 3.2 11.7 ± 0.3 10.1 ± 0.2 group) Pin 650 K 3VXEDQJOLFDKrammer Prof. D.G. Mann 46.2 ± 0.3 8.2 ± 0.2 11.1 ± 0.7 Pin 706 F 3QHJOHFWLIRUPLV Prof. D.G. Mann 105.3 ± 3.9 18.5 ± 0.5 8.1 ± 0.2 Krammer Pin 867 inv P. grunowii Krammer Prof. D.G. Mann 42.2 ± 0.3 8.2 ± 0.2 11.1 ± 0.7 Pin 870 MG P. v i r i d i f o r m i s Krammer Prof. D.G. Mann 80.5 ± 2.3 16.7 ± 0.3 9.2 ± 0.3 Pin 873 TM P. s p . Prof. D.G. Mann 28.3 ± 2.8 10.2 ± 0.4 12.3 ± 0.5 Pin 876 TM 3DFXPLQDWD W. Smith Prof. D.G. Mann 38.9 ± 2.6 11.5± 0.8 10.0 ± 0.0 Pin 877 TM P. parvulissima Prof. D.G. Mann 58.3 ± 1.5 9.8 ± 0.3 10.0 ± 0.0 Krammer Pin 883 TM P. s p . VXEFRPPXWDWD Prof. D.G. Mann 63.1 ± 2.8 14.1 ± 0.1 10.3 ± 0.5 group) Pin 885 TM P. n o d o s a (Ehrenberg) Prof. D.G. Mann 42.1 ± 1.2 6.7 ± 0.2 10.0 ± 0.0 W. Smith Pin 888 MG P. n o d o s a (Ehrenberg) Prof. D.G. Mann 37.7 ± 1.5 7.3 ± 0.4 10.6 ± 0.7 W. Smith Pin 889 MG P. grunowii Krammer Prof. D.G. Mann 41.9 ± 0.8 7.9 ± 0.2 13.0 ± 0.8

112 Morphology of Pinnularia Table 5.2 Summary of the 17 frustule and 3 plastid characters and their respective character states and codes used for the cladistic analyses. cȱǰȱȬȱȱȱȱ ȱǯȱȱȱȱȱ ȱȱȱȱȬǯȱȱȱ¡ȱȱȱǯ d Descriptions of the characters and character states can be found in Appendix 5.1. e Coding refers to the code of the character states used in the cladistic analyses and mentioned in Table 5.3.

Character typec Characterd Character statesd Codinge 1 1RQVWUXFWXUDO Margins Convex 0 Parallel 1 Undulate 2 &RQVWULFWHGLQWKHPLGGOH 3 ,QÀDWHGLQWKHPLGGOH 4 2 1RQVWUXFWXUDO $SLFHV Cuneate 0 Rounded 1 Rostrate 2 Capitate 3 3 6WUXFWXUDO Path of external raphe slit Curved 0 Undulate 1 4 6WUXFWXUDO Central raphe endings Elongated drop-like 0 Round 1 Linear 2 5 6WUXFWXUDO +HOLFWRJORVVD Tongue-like 0 Hooked 1 6 6WUXFWXUDO 5DSKHFRPSOH[LW\ Simple 0 Three lines visible 1 Complex 2 7 1RQVWUXFWXUDO Axial area: relative width at Intermediate: 1/4 – 1/6 0 PLGEUDQFK Narrow: < 1/6 1 Broad: > 1/4 2 8 1RQVWUXFWXUDO Axial area: relative width at Narrow: < 1/2 0 FHQWUDOUDSKHHQGLQJV Broad: 1/2 – 1/1 1 Valve width: 1/1 2 9 1RQVWUXFWXUDO )DVFLD Absent 0 Unilateral 1 Bilateral 2 10 1RQVWUXFWXUDO Central area: shape (OOLSWLFDO 0 Diamond 1 5HFWDQJXODU 2 11 6WUXFWXUDO Ghost striae Absent 0 Present 1 12 6WUXFWXUDO 6XUIDFHLUUHJXODULWLHV Absent 0 Present 1 13 1RQVWUXFWXUDO 6WULDRULHQWDWLRQFHQWUH Radiate 0 Transverse 1 14 1RQVWUXFWXUDO Stria orientation: apex Convergent 0 Radiate 1 Transverse 2 15 6WUXFWXUDO Alveolar opening length: > 2/3 0 proportion of half valve width 2/3 – 1/4 1 < 1/4 2 16 6WUXFWXUDO $OYHRODURSHQLQJORFDWLRQ Central 0 Distal 1

Morphology of Pinnularia 113 Character typec Characterd Character statesd Codinge 17 6WUXFWXUDO Girdle bands: shape of areolae Elongate 0 Round 1 18 &\WRORJLFDO Plastid: shape Linear 0 H-shaped 1 19 &\WRORJLFDO Plastid: extension under valve Narrow 0 VXUIDFH Broad 1 20 &\WRORJLFDO Pyrenoids: number per plastid One 0 Two 1 Absent 2

Table 5.3 Taxon/character matrix of the 48 strains. The a priori determined ancestral states are given in the last line (“ancestral”).

Character Strain 1234567891011121314151617181920 % F 02000000010000000111 (B2)g10 02000000010000000111 (FULQV D 02000001210000000111 (FULQV G 01021010100011211110 (FULQV D 01000000000000000010 (QF D 11110210000000100002 (QF E 01000021000112210000 (No5)j 01120000000000100002 (Pinn CNoel)7 01000012210000000010 (Tor1)a 40110100000000100002 (Tor1)b 01000000210000000010 (Tor1)g 00110100000000100002 (Tor11)b 0102000222000000011? (Tor12)d 01000000000001000010 (Tor3)a 01000000000012000010 (Tor4)i 01000010200000000111 (Tor4)r 32000002210000000010 7RU F 41000002210000000010 (Tor7)f 01100022201000000010 (Tor8)b 33000022211000000010 (Tor8)d 01000002211000000010 (Val1)b 01000021000112210000 (Wie)a 01020000000000000111 :LH F 33100002210000000010 (Yde1)a 01120000000000100002 Cal 769 01021010000011211110 Cal 856 01021010000011211110 Cal 878 0012000000000010011?

114 Morphology of Pinnularia Character Strain 1234567891011121314151617181920 Cal 890 40021000000011211110 Corsea 10 01120000000000100002 Corsea 11 01120000000000100002 Corsea 2 41110100000000100??? Corsea 6 01120000000000100002 Pin 12 01000002211000000010 Pin 649 00110100000000100012 Pin 650 13000002211000000111 Pin 706 01110200000000100012 Pin 867 23000011211000000111 Pin 870 11110200000000100??? Pin 873 01020010000010100012 Pin 876 100200210000021000?? Pin 877 01000022201000000010 Pin 883 00110100000000100012 Pin 885 22000001201000000010 Pin 888 32000001201000000010 Pin 889 23000001210000000111 W045b 01000002211000000010 W123a 01000002210000000010 $QFHVWUDO 00000000000000000???

2010). Based on this, the morphological dataset was subdivided into three subsets: all frustule characters (characters 1-17 in Table 5.2), only structural frustule characters (3-6, 11-12, 15-17), and plastid characters (19-20), which were analysed separately and in combination. Cladistic analyses were performed in PAUP* 4.0b10 ǻ ěȱǰȱ ŘŖŖŘǼȱ ȱ ȱ ¡ȱ ¢ȱ ǻǼȱ ȱ search with branch swapping and tree bisection-reconnection (TBR) rearrangement followed by bootstrap analysis with 1,000 replicates. All characters were unordered and unweighted. Because a large number of most parsimonious trees was recovered in all analyses, strict consensus trees were computed, together with the consistency index (CI) and retention index (RI) of these trees and the single characters. The CI measures the relative amount of homoplasy in a character or cladogram (with CI=1 for non-homoplasious characters/ cladograms), while the RI measures the amount of synapomorphy expected from a data set which is retained as synapomorphy on a cladogram (with RI=1 for strongly informative characters/ cladograms).

Morphology of Pinnularia 115 ¢ǰȱ ȱ ȱ ȱ ěȱȱ ȱ ¡ȱ ȱ Ȭ informative characters, we selected for every analysis the characters ȱȱ ȱǃȱśȱȱȱȱ ȱȱȱȱ ȱȱȱśǰȱȱ ȱ ǃȱ śȱ ȱ ȃȄȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ cladistic analysis. Afterwards, tree length, CI and RI were compared for both analyses (all characters/only informative characters) of ȱ ěȱȱ ȱ ǯȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ morphological characters were also computed based on a molecular ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ěȱȱet al.ȱǻĴȱǼǯ

Mantel tests To evaluate whether morphological and molecular distances ȱ ȱ ȱ ęȱ¢ȱ ȱ ȱ ȱ ȱ ȱ (Mantel, 1967; Mantel & Valand, 1970). Mantel tests were computed using the ZT software tool (Bonnet & Van De Peer, 2002) with 100,000 permutations. The genetic distances of the concatenated molecular data set were derived from the most likely molecular ȱ¢ȱęȱȱǻ ȱet al., 2004). The morphological distances between the strains were calculated as total distance by PAUP* ŚǯŖŗŖȱǻ ěȱǰȱŘŖŖŘǼȱȱȱȱ¡Ȧȱ¡ȱǻȱ 5.3) for all subsets (all characters, and frustule, structural frustule and plastid characters separately and in pairwise combinations). ȱęȱȱȱ ȱp=0.05.

$QFHVWUDOVWDWHUHFRQVWUXFWLRQ Ancestral state estimations were performed by Mesquite 2.73 (Maddison & Maddison, 2010). As a backbone we used the most likely molecular tree recovered by the concatenated DNA ¢ȱ ȱ ȱ ěȱȱ et al.ȱǻĴȱǼȱ ȱ all morphologically dissimilar strains per species (see Table 5.1). ȱȱ ȱęȱȱȱȱȱȱȱȱȱ GenBank. In a phylogenetic tree, branch lengths are proportional to the amount of molecular evolution in the genes used to build the ǯȱȱȱěȱȱǰȱȱȱȱȱȱ ȱ occur. Because the genes we used in our study are not assumed to be linked to the morphological characters we investigate, this rate variation was smoothed out by constructing an ultrametric tree in which branch lengths are roughly proportional to evolutionary time instead of to amounts of molecular evolution. The original

116 Morphology of Pinnularia phylogenetic tree was therefore smoothed into an ultrametric tree using penalized likelihood (Sanderson, 2002, 2003). The same character coding was used as for the cladistic analyses, and ancestral state reconstructions were using both maximum parsimony (MP) ȱ¡ȱȱǻǼȱȱ ȱȱĴȱǯ

5.3 Results 6WUDLQVHOHFWLRQ ¢ȱ ȱȱȱŚŞȱ¢ȱěȱȱǰȱŚŚȱ ȱ ȱ ȱęȱȱȱPinnularia and 4 as Caloneis silicula (Table 5.1, Figs 5.1-50). A total of 31 morphological species (not including varieties) were recognized, of which only 15 could be assigned to a described species. Of the remaining 16 taxa, 11 could not even be assigned to a morphologically closely resembling taxon and can be considered as undescribed species. Table 5.1 gives an overview of the ȱ ȱ ȱ ȱ ęȱǰȱ ȱ ȱ ȱ characteristics. Oxidized material of all strains is shown in Figs. 5.1- 50.

0RUSKRORJLFDOFKDUDFWHUV Table 5.2 gives an overview of the 20 selected characters and their character states. In total 17 frustule characters (Table 5.2; 1-17) and 3 cytological characters (Table 5.2; 18-20) were used and most characters had more than two character states (Table 5.2). For frustule characters, we discriminated between structural and non-structural wall characters as described above (see Materials and methods).

&ODGLVWLFDQDO\VHV All cladistic analyses resulted in a large number of most parsimonious trees, and therefore strict consensus trees (SCTs) were constructed based on these most parsimonious trees to examine the resulting Ĵȱȱȱǯȱȱȱȱȱȱȱȱ clades but are in general very low. An overview of the recovered ȱȱȱěȱȱ¢ȱȱȱȱȱśǯŚǯȱȱȱȱ trees, CI and RI of SCTs and characters are summarized in Table 5.5. Cladistic analysis of all characters revealed four clades in addition to a cluster containing the two stains of P. cf. microstauron (Fig. 5.51; Table 5.4). Three of the four clades corresponded closely ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢¢ȱ ǻěȱȱ et

Morphology of Pinnularia 117 118 Morphology of Pinnularia Figs 5.1-34 Oxidized material of the strains used for the cladistic analyses and reconstruction ȱȱȬ¢ȱǯȱȱŗȬŜȱ ȱȱȱȱǯȱȱŝȬŘśȱ ȱȱ ȱȱǰȱ ȱȱȱȱȱȃgrunowiiȄȬȱǻȱŝȬŗŗǼǰȱȱȃnodosaȄȬȱ ǻȱŗŘȬŗśǼȱȱȱȃsubgibbaȄȬȱǻȱŗŝȬŘśǼǯȱȱŘŞȬřŚȱ ȱȱȱȱȱȱ ȱȱǰȱȱȱȱȃborealis-microstauronȄȬȱǻȱŘşȬřřǼȱȱ ȱȱ ȱǻȱŘŞǰȱřŚǼǯȱȱŘŜȬŘŝȱȱ ȱȱȱȱȱȱȱȱȱ ancestral Pinnularia resulting from ancestral state reconstruction (ASR) (see text). ǯȱ ŗǯȱ Pinnularia sp. (divergensȬǼȱ ǻŝǼǯȱ ǯȱ Řǯȱ Pinnularia sp. (divergensȬ ǼȱǻŗǼǯȱǯȱřǯȱCaloneis siliculaȱǻȱŞşŖǼǯȱǯȱŚǯȱCaloneis siliculaȱǻȱŝŜşǼǯȱ ǯȱśǯȱCaloneis siliculaȱǻȱŞśŜǼǯȱǯȱŜǯȱCaloneis siliculaȱǻŚǼǯȱǯȱŝǯȱPinnularia subanglicaȱǻȱŜśŖǼǯȱǯȱŞǯȱPinnularia sp.ȱǻŚǼǯȱǯȱşǯȱPinnularia cf. marchica ǻŚǼǯȱǯȱŗŖǯȱPinnularia grunowiiȱǻȱŞŜŝǼǯȱǯȱŗŗǯȱPinnularia grunowii (Pin ŞŞşǼǯȱǯȱŗŘǯȱPinnularia acrosphaeiraȱǻŘǼǯȱǯȱŗřǯȱPinnularia acrosphaeria (Val1) ǯȱǯȱŗŚǯȱPinnularia nodosaȱǻȱŞŞśǼǯȱǯȱŗśǯȱPinnularia nodosaȱǻȱŞŞŞǼǯȱǯȱŗŜǯȱ Pinnularia subcapitata var. elongateȱǻǼǯȱǯȱŗŝǯȱPinnularia sp.ȱǻŚǼǯȱǯȱŗŞǯȱ Pinnularia sp. ǻȱ ȬǼŝǯȱ ǯȱ ŗşǯȱ Pinnularia sp. ǻŗŘřǼǯȱ ǯȱ ŘŖǯȱ Pinnularia sp. (subgibbaȬǼȱǻŞǼǯȱǯȱŘŗǯȱPinnularia sp. (subgibbaȬǼȱǻŞǼǯȱǯȱ 22. Pinnularia sp. (subgibbaȬǼȱǻŝǼǯȱǯȱŘřǯȱPinnularia sp. (subgibbaȬǼȱ ǻŖŚśǼǯȱ ǯȱ ŘŚǯȱ Pinnularia sp. (subgibbaȬǼȱ ǻȱ ŗŘǼǯȱ ǯȱ Řśǯȱ Pinnularia sp. (subgibbaȬǼȱ ǻȱ ŞŝŝǼǯȱ ǯȱ ŘŜǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ancestral Pinnulariaǰȱ ȱ ȱ ¢ȱǯȱ ǯȱ Řŝǯȱ ȱ ȱ ȱ ȱ ȱ of a hypothetical ancestral Pinnulariaǰȱ ȱ ȱ ¢ȱ ǯȱ ǯȱ ŘŞǯȱ Pinnularia cf. altiplanensisȱǻŗŗǼǯȱǯȱŘşǯȱPinnularia cf. microstauronȱǻŘǼǯȱǯȱřŖǯȱPinnularia cf. microstauronȱ ǻŘǼŗŖǯȱ ǯȱ řŗǯȱ Pinnularia borealisȱ ǻřǼǯȱ ǯȱ řŘǯȱ Pinnularia borealisȱǻŝǼǯȱǯȱřřǯȱPinnularia borealisȱǻŗŘǼǯȱǯȱřŚǯȱPinnularia sp. (Wie) a. Scale bar represents 10 μm.

Morphology of Pinnularia 119 120 Morphology of Pinnularia Figs 5.35-50 Oxidized material of the strains used for the cladistic analyses and ȱȱȱȬ¢ȱǰȱ ȱȱȱȱȱȱǯȱ ȱřśȬŚŗȱ ȱȱȱȱȱȃsubcommutataȄǰȱǯȱŚŘȱ ȱP. acuminataǰȱ ȱȱŚřȬśŖȱ ȱȱȱȱȱȃviridiformisȄǯȱǯȱřśǯȱPinnularia sp. (Pin ŞŝřǼǯȱǯȱřŜǯȱPinnularia cf. isselanaȱǻȱŞŝŞǼǯȱǯȱřŝǯȱPinnularia subcommutata var. nonfasciataȱǻȱŗŖǼǯȱǯȱřŞǯȱPinnularia subcommutata var. nonfasciata (Corsea ŗŗǼǯȱǯȱřşǯȱPinnularia subcommutata var. nonfasciataȱǻȱŜǼǯȱǯȱŚŖǯȱPinnularia isselanaȱ ǻśǼǯȱ ǯȱ Śŗǯȱ Pinnularia subcommutataȱ ǻŗǼǯȱ ǯȱ ŚŘǯȱ Pinnularia acuminataȱǻȱŞŝŜǼǯȱǯȱŚřǯȱPinnularia sp. (subcommutataȬǼȱǻȱŞŞřǼǯȱǯȱŚŚǯȱ Pinnularia sp. (subcommutataȬǼȱǻȱŜŚşǼǯȱǯȱŚśǯȱPinnularia neglectiformis ǻȱŝŖŜǼǯȱǯȱŚŜǯȱPinnularia viridiformisȱǻȱŞŝŖǼǯȱǯȱŚŝǯȱPinnularia viridiformis ǻŘǼǯȱǯȱŚŞǯȱPinnularia neomajorȱǻŗǼǯȱǯȱŚşǯȱPinnularia neomajor (Tor1)a. ǯȱśŖǯȱPinnularia neomajor (Corsea 2). Scale bars represent 10 μm.

Morphology of Pinnularia 121 ȱ characters ȱȱȱ Structural + plastid taxa only 2 -Pin876 -Pin873, x -Pin870, characters xxxx xxxx -Pin876 Structural frustule -Pin873, x -Pin870, Plastid characters x -Cal878 x x taxa only 2 ȱȱȱȱȱȱ¢ȱȱȱȱě x x x taxa only 2 xxxx xxxx xx (+Wiea)x xx xxx All characters Frustule characters x -Tor4i [7RUF All char. Inform. All char. Inform. All char. All char. Inform. All char. Inform. ´FODGH ´FODGH ´VXEFODGH ´FODGH ´FODGH ´FODGH ´VXEFODGH ȱǼǯȱȱȃ¡Ȅȱȱȱȱȱȱ¡ȱǻȱȱȱ¢Ǽǯȱȱȱȱȱǻȃ¡ȱƸȄǼȱȱȱȱ

&VLOLFXOD P. n o d o s a borealis P.

grunowii divergens 3DFURVSKDHULD subgibba 3PLFURVWDXURQ “ “ “ Molecular clades Character subset viridiformis “ VXEFRPPXWDWD

DFURVSKDHULDQRGRVD “ “ et al. ȱǻĴ

VXEFRPPXWDWDDFXPLQDWDYLULGLIRUPLV ȱȱ “ Table 5.4 ȱ ȱȱȱȱȱȱȱȱȱTable ȱǻȱǰȱȱǰȱȱǰȱȱȱǰȱȱȱƸȱȱǼȱ ȱ¢£ȱ characters (Inform.). The morphological clades are compared with the molecular of all selected characters (All char.) or only the “informative” ě (“x –“) compared to the molecular clades is indicated by light grey shading.

122 Morphology of Pinnularia al.ǰȱ ĴȱǼȱ ȱ ȱ ȱ ȱ ȃgrunowii”, “Ȭ Ȭ” and “subgibba” clade, respectively. The “grunowii” clade contained smaller strains with H-shaped plastids while the “ȬȬ” clade contained the more robust Pinnularia’s. The “subgibba” clade contained the “subgibba”-like Pinnularia’s, smaller Pinnularia’s characterized by ghost striae, together with (Tor7)c and P. nodosa which were placed separately in the molecular phylogeny. The “Ȭ” clade contained Caloneis silicula and P. acrosphaeria and was never recovered as a single clade in the molecular phylogeny. When plastid characters were left out of the analysis (Fig. 5.52; Table 5.4), the “grunowii” and “subgibba” clades were not recovered, and (Wie)a was included in the “ȬȬ” clade. Analysis of structural characters only (Fig. 5.53; Table 5.4) resulted in the “Ȭ” and “ȬȬ viridiformis” clade, but without Pin870 and Pin873. Interestingly, a “viridiformis”-subclade (identical to the molecular “viridiformis” subclade) was found inside this clade, separating the “viridiformis” strains from the “subcommutata” strains. Analysis of the structural and plastid characters only revealed the “Ȭ” clade, and no clades were formed when using the three plastid characters only (Figs not shown).

&KDUDFWHUVVXSSRUWLQJWKHPRUSKRORJLFDOFODGHV The “grunowiiȄȱ ȱ ǽ ȱ ǻŚǼȱ ȱ ȱ ȱ ȱ Ǿȱ ȱȱ¢ȱȱȱǰȱȱȬȱȱ area and H-shaped plastids with two pyrenoids, none of which were synapomorphies. This clade disappears when plastid characters are discarded, indicating the importance of these characters to recover the relationship. The same applies to the “subgibba” clade (+ (Tor7) c and P. nodosa compared to the molecular clade) which was partly supported by a bilateral fascia, but also by linear plastids having each a single pyrenoid. The two other clades were supported by structural characters and were therefore also recovered when plastid or non-structural ȱ ȱ Ĵȱǯȱ ȱ ȃȬ” clade was characterized by a transverse orientation of the apical striae and the synapomorphic small, distal opening of the alveoli. The characters supporting the “ȬȬ” clade were ȱȱ¡ȱȱęȱǰȱȱȬȬȱȱȱ endings, the intermediate length of the alveolar opening, and the

Morphology of Pinnularia 123 Fig. 5.51 Strict consensus tree (SCT) from the most parsimonious trees of the cladistic ¢ȱȱȱȱŘŖȱȱǰȱ ȱȱȱȱȱǯȱ ȱȱȱȱǰȱȱȱȱǰȱȱȱȱȱȱȱ noted on the right of the cladogram and indicated by a star (*) on the tree.

124 Morphology of Pinnularia Fig. 5.52 Strict consensus tree (SCT) from the most parsimonious trees of the cladistic ¢ȱ ȱ ȱ ȱ ŗŝȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱȱȱǰȱȱȱȱǰȱȱȱȱȱȱȱ noted on the right of the cladogram and indicated by a star (*) on the tree.

Morphology of Pinnularia 125 Fig. 5.53 Strict consensus tree (SCT) from the most parsimonious trees of the cladistic ¢ȱȱȱȱşȱȱȱǰȱ ȱȱȱȱȱǯȱ ȱȱȱȱǰȱȱȱȱǰȱȱȱȱȱȱȱ noted on the right of the cladogram and indicated by a star (*) on the tree.

126 Morphology of Pinnularia absence of pyrenoids. Because all these characters, except for pyrenoids, were structural characters, this clade is also supported in the cladistic analysis of those characters. Support for the “viridiformis” subclade in the structural character analysis resulted from the synapomorphic round central raphe endings, undulate ¡ȱȱęȱȱȱ¡ȱȱȱȱ ȱřȱȱ lines.

(IIHFWRIGHOHWLQJQRQLQIRUPDWLYHFKDUDFWHUV ȱśǯśȱ£ȱȱ ȱȱ ȱȱȱěȱȱȱȱ all analyses. Because none of the plastid characters was recovered as ȃȄȱȱȱęȱȱ¢ȱȱȱȱ ȱȱ ǰȱȱȱ analysis including only “informative” characters was impossible. In all analyses the most informative characters (CI=1, RI=1) were the helictoglossa, surface irregularities, alveolus opening and location, and girdle band areolae, all structural frustule characters. Stria orientation and the number of pyrenoids had a CI of 0.5. The CI of all other characters was lower than 0.5. However, selecting only the more “informative” characters based on the CI and RI had a negative Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻȱ śǯŚǼǰȱ ¢ȱ partly due to the even lower number of characters included in the analysis. This shows that characters with low CI and RI can still make a valuable contribution to the recovery of relationships. When analyzing the morphological character matrix using the ¢ȱȱȱȱǰȱȱ ȱȱ ȱǃȱŖǯśȱ ȱȱȱ helictoglossa, surface irregularities, alveolus opening and location, girdle band areolae and number of pyrenoids, but additionally also central raphe endings, raphe complexity and fasciae. Again, most of these features are structural characters. Note that stria orientation had a low CI when based on the molecular tree topology, and that the molecular topology was much less parsimonious (see the higher tree length in Table 5.5) and thus contained many more changes in character states of the morphological characters over the tree, compared to the MPTs resulting from the maximally parsimonious cladistic analyses (Table 5.5).

$QFHVWUDOVWDWHUHFRQVWUXFWLRQ Fig. 5.54 shows ultrametric molecular phylogenies on which the ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ Ĵȱǰȱ ȱ ȱ ȱ ȱ ȱ ǻǼȱ ȱ ȱ ȱ

Morphology of Pinnularia 127 $OOFKDUDFWHUV )UXVWXOHFKDUDFWHUV &KDUDFWHU1/tree All7 Inf.8 Mol.9 Allg Inf.h Mol.i CI RI CI RI CI RI CI RI CI RI CI RI 1 margins 0.3 0.1 0.3 0.1 0.3 0.0 0.3 0.1 DSLFHV 0.2 0.3 0.2 0.3 0.2 0.1 0.2 0.3 3 ext. raphe path 0.2 0.7 0.2 0.7 0.3 0.8 0.2 0.7 0.3 0.9 0.3 0.8 FHQWUDOUDSKHHQGLQJV 0.2 0.5 0.4 0.9 0.5 0.9 0.2 0.6 0.4 0.9 0.5 0.9 KHOLFWRJORVVDH 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 UDSKHFRPSOH[LW\ 0.3 0.0 0.5 0.7 0.3 0.0 0.5 0.7 D[LDODUHDPLGEUDQFK 0.2 0.3 0.2 0.1 0.2 0.3 0.2 0.1 D[LDODUHDFHQWUDO 0.3 0.7 0.1 0.3 0.3 0.7 0.2 0.4 0.3 0.7 IDVFLD 0.3 0.8 0.2 0.7 0.5 0.9 0.3 0.8 0.2 0.7 0.5 0.9 FHQWUDODUHD 0.3 0.7 0.2 0.4 0.2 0.6 0.2 0.6 0.2 0.4 0.2 0.6 11 ghost striae 0.1 0.3 0.2 0.6 0.1 0.2 0.2 0.9 VXUIDFHLUUHJXODULWLHV 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 VWULDHFHQWUDO 0.5 0.9 0.5 0.9 0.3 0.6 0.5 0.9 0.5 0.9 0.3 0.6 14 striae apex 0.5 0.7 0.5 0.7 0.4 0.6 0.5 0.7 0.5 0.7 0.4 0.6 15 alveolar opening 1.0 1.0 1.0 1.0 0.7 1.0 1.0 1.0 1.0 1.0 0.7 1.0 DOYHRODUORFDWLRQ 1.0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 1.0 0.5 0.8 17 girdle areolae 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 18 plastid shape 0.2 0.6 0.1 0.2 0.2 0.7 19 plastid extension 0.1 0.3 0.3 0.8 20 pyrenoid number 0.5 0.9 0.2 0.6 0.7 0.9 SCT10 0.3 0.6 0.3 0.7 --0.3 0.6 0.3 0.7 -- MPT11’s 0.4 0.8 0.5 0.9 0.3 0.7 0.5 0.8 0.6 0.9 0.3 0.7 SCTi: length 119 79 - 114 79 - MPTk’s: length 80 41 105 66 26 94

Table 5.5 Summary of the CI and RI values of the morphological characters based on the ȱȱǰȱȱ ǰȱ ȱȱȱȱȱȱȱȱǻǼȱȱȱ ȱȱǻȂǼȱȱȱȱ¢ǰȱȱȱȱȱȱȱ ȱǻȱǯǼȱȱ¢ȱȱȱȱǻȱǯǼǰȱȱȱȱ ȱ¢ȱȱȱȱ¢¢ȱǻȱǼǯȱȱǃȱŖǯśȱȱȱȱȱȱ boldǰȱȱǀȱŖǯśȱȱȱȱ¢ǯȱȱȱȱȱȱ ȱȱȱ ȱȱȱǯȱȬ¢ȱȱȱȱȱ ȱȱȱ and were excluded during subsequent analyses using informative characters only. Ŝ See Table 5.2 and Appendix 5.1 for more information on the characters.

128 Morphology of Pinnularia 6WUXFWXUDOIUXVWXOHFKDUDFWHUV 3ODVWLGFKDUDFWHUV 6WUXFWXUDOSODVWLGFKDUDFWHUV Allg Inf.h Mol.i Allg Mol.i Allg Inf.h Mol.i CI RI CI RI CI RI CI RI CI RI CI RI CI RI CI RI

0.3 0.9 0.5 0.9 0.3 0.8 0.1 0.0 0.3 0.8 0.3 0.8 0.3 0.8 0.5 0.9 0.1 0.2 0.5 0.9 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.5 0.7 0.5 0.7 0.5 0.7 0.3 0.0 0.5 0.3

0.1 0.0 0.2 0.6 0.1 0.0 0.2 0.1 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

0.5 0.9 0.5 0.9 0.7 1.0 0.1 0.2 0.7 0.6 1.0 1.0 1.0 1.0 0.5 0.8 0.5 0.8 1.0 1.0 0.5 0.4 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.1 0.0 0.2 0.7 0.1 0.2 0.2 0.1 0.1 0.0 0.3 0.8 0.1 0.1 0.3 0.3 0.1 0.0 0.7 0.9 0.1 0.0 0.7 0.6 0.4 0.8 0.5 0.9 --0.1 0.0 --0.1 0.2 1.0 1.0 -- 0.8 1.0 0.8 1.0 0.5 0.8 0.8 1.0 0.4 0.8 0.6 0.9 1.0 1.0 0.4 0.8 32 21 - 44 - 116 4 - 15132551126436

7ȱƽȱȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ during the cladistic analysis 8 Inf.= Informative characters; only the informative characters (based on the CI and RI of the ¢ȱ ȱȱȱȱȱęȱȱǼȱ ȱȱȱȱȱ¢ şȱǯƽȱȱǲȱȱ¢ȱȱȱȱȱȱěȱȱet al.ȱǻĴȱǼȱ ȱ ȱȱȱȱ ȱȱ ȱȱȱȱěȱȱȱ 10ȱƽȱȱȱǰȱ11 MPT= Most parsimonious tree

Morphology of Pinnularia 129 nodes. Based on the ancestral state reconstruction (ASR) using both MP and LM, ancestral Pinnularia (see Figs 5.26-27) had valves with convex margins and rounded apices, a simple raphe with curved external path with elongated central raphe endings, an intermediate width axial area mid-branch, but narrow axial area by the central raphe endings. The central area was elliptical and a bilateral fascia was present, without ghost striae or surface irregularities. Striae were radiate at the centre, convergent near the apices, with large internal alveolar openings. The helictoglossa was simple, tongue- like and the girdle bands had elongated areolae. Plastids would have been linear with broad extensions and one pyrenoid per plastid (Fig. 5.27). These recovered ancestral states were also estimated a priori by us as being the primitive frustule character states (see Table 5.2 and 5.3) based on the morphology of the fossils reported by Lohman & Andrews (1968), except for the rounded apices and the presence of a fascia. Most of the fossil species of Lohman & Andrews (1968) had cuneate apices rather than rounded ones and no fascia, however, both alternative character states were also present in a single species. The fossil Pinnularia reported as being the earliest representative ȱȱȱ¢ȱȱǻěȱȱet al.ǰȱĴȱǼȱȱ lacked a fascia and even had rostrate rather than rounded apices, and surface irregularities. Based on this ASR, most of the character states evolved multiple times during the evolution of Pinnularia, while other character states evolved only once, thereby often characterizing a single molecular clade. Of the 20 characters studied, only 8 contained character states which evolved once in the present phylogeny (central raphe endings, helictoglossae, girdle band areolae, surface irregularities, raphe complexity, length and location of alveolus openings, and pyrenoids). Round central raphe endings were synapomorphic for the “viridiformis” clade (Fig. 5.54, character 4), while linear central raphe endings evolved at the base of the cluster containing the “ȬȬ” clade, (Wie)a and (Tor11)b. Caloneis silicula also has linear central raphe endings, but note that ȱ¢ȱȱ ȱȱęȱȱȱȱȱȱ (SEM photographs: Appendix 2, n-n´). Hooked helictoglossae and round girdle band areolae were synapomorphic for Caloneis silicula, while the presence of surface irregularities only evolved in P. acrosphaeria (Figs not shown). Raphes with three parallel lines visible in LM (Appendix 5.2, r) or three twisting lines (complex raphe, Appendix 5.2, s) evolved

130 Morphology of Pinnularia once at the base of the “viridiformis” clade (Fig. 5.54, character 6). However, the complex, twisting raphe structure evolved multiple times within this clade. Because we also observed raphes with a complex, twisting structure and with three parallel lines within monoclonal strains, those two characters states are probably part of a continuum and should be regarded as a single character. Intermediate alveolus openings were synapomorphic for the “ȬȬ” clade (Fig. 5.54, character 15), while small and distantly located alveolar openings occurred in both P. acrosphaeria and C. silicula. Also the absence of pyrenoids evolved at the base of the “ȬȬ” clade (Fig. 5.54, characters 20).

Mantel tests Mantel tests between the morphological pairwise distance matrices ȱȱȱȱȱȱęȱȱȱ between the datasets (p=0.00001) for all morphological subsets.

5.4 Discussion Independent phylogenetic analyses of molecular and morphological ȱ ȱ ŚŞȱ ¢ȱ ěȱȱ ȱ ȱ ȱ ȱ Pinnularia and Caloneis recovered similar clades, and morphological ȱ ȱ ȱ ȱ ęȱ¢ȱ ǯȱ ǰȱ ȱ ¢ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ǻěȱȱ et al.ǰȱ ĴȱǼǰȱ ȱ ȱ generated fewer clades, due to the lower number of phylogenetically informative characters compared to the molecular data, and was ȱȱȱ¢ȱȱȱęȱȱȱȱPinnularia. ȱ¢ȱȱȱȱȱ¢ȱĴȱȱȱȱ ěȱȱȱȱǰȱ ȱ¢ȱȬȱ frustule characters were found to be homoplasious, several ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ relationships. However, excluding the characters with low CI and RI in a subsequent analysis did not improve the resolution of the phylogeny. Inferences of the most likely ancestral states of characters ȱ ȱ ěȱȱ ȱ ȱ Pinnularia are fairly congruent with the currently known sequence of appearances in the fossil record. Our molecular and morphological datasets supported similar clades. Most of the clades retrieved by the morphological data were

Morphology of Pinnularia 131 Fig. 5.54ȱȱȱȱȱȬ¢ȱȱȱȱȱȱ ěȱȱet al.ȱǻĴȱǼȱȱŗŞǰȱŘŞǰȱrbcǰȱpsbA and coxŗǰȱ ȱȱȱȱ ancestral state reconstruction of 18 of the 20 morphological characters using maximum ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ shaded circle. Shading legend is given for each tree separately. Bootstrap support – Posterior ¢ȱȱȱȱȱęȱȱȱǻȱŗǼǯ

132 Morphology of Pinnularia Morphology of Pinnularia 133 134 Morphology of Pinnularia Morphology of Pinnularia 135 136 Morphology of Pinnularia also recovered by molecular-genetic analyses: the “grunowii” and “subgibba” clades, and the “viridiformis” and “subcommutata” clades; ȱ ȱ ȱ £ȱ ¢ȱ ȱ ęȱȱ ȱ ȱ genetic and morphological distances using Mantel tests. Congruence between morphological and molecular data has been encountered in a wide variety of organisms (e.g. Marvaldi et al., 2002; Scotland et al., 2003; Edgar & Theriot, 2004; Lundholm et al., 2006; Jacobs et al., 2008) and retrieving independent support for clades strengthens ȱ ¢ȱ ȱ ȱ ¢ȱ ¢ȱ ǻĴȱǰȱ ŗşŞŘǼǯȱ ǰȱ ȱ ȱ ȱ ¢ǰȱ ȱ ȱ ȱ ęȱȱ ȱ resolution in morphological data compared to molecular sequences (Wortley & Scotland, 2006). Morphological analyses could not assign ȱ¡ȱȱȱǰȱȱěȱȱȱȱȱȱȱ analysis could not be recovered with the morphological dataset, e.g. the kinship of the P. borealisȱȱȱȱĜȱȱ¢ȱ ȱP. nodosa and P. acrosphaeria, nor were the deeper relations between the clades resolved. Moreover, bootstrap support was never high. The low resolution and support of morphological phylogenies is likely caused by the low number of morphological characters (Scotland et al., 2003; Giribet, 2010) and, because of the low character/taxon ratio, bootstrap percentages of morphological studies are inherently low (Bremer et al.ǰȱ ŗşşşǼǯȱ ǰȱ Ĝȱȱȱ ȱ ȱ ȱ with molecular data when seeking to recover relationships with a small number of nucleotides or non-informative markers (Rokas et al., 2003). For the current molecular phylogeny of Pinnularia ęȱȱ ěȱȱ ȱ ǻȱ ŚŞśŘȱ ǰȱ ȱ ŗŖŗŘȱ parsimony-informative sites) were needed to obtain support at the ȱȱǻěȱȱet al.ǰȱĴȱǼǯȱ In this study we used 20 morphological characters, mostly with two or three character states, which is low compared to the 48 taxa studied. Only discrete variables were used, but incorporating continuous data for example for size measurements and shape could increase the number of characters and enhance the phylogenetic resolution. Morphometric data are known to improve phylogenetic estimates (Edgar & Theriot, 2004) and new algorithms have been ȱ ȱ ¢£ȱ ȱ ȱ ǻǰȱ ŘŖŖŗǲȱ ěȱȱet al., 2006; Gonzalez-Jose et al., 2008). Nevertheless, at the intrageneric level relatively low variation in morphology is to be expected since diatom genera are delimited based on morphological similarity and homogeneity (Kociolek, 1997). Cladistic analyses are mostly successful at intergeneric levels (Kociolek & Stoermer, 1986; Cox

Morphology of Pinnularia 137 & Williams, 2000; Marivaux et al., 2004; Cox & Williams, 2006; Ohl & Spahn, 2010), while using only morphological data to estimate ęȱȱ ȱ ȱ ȱ ȱ ȱ ǻȱ et al., 2009; Denk & Grimm, 2009; Ohl & Spahn, 2010) as is further demonstrated here. It has been argued that incorporating as much ȱ ȱ ȱ ȱ ěȱȱ ǰȱ ȱ ǰȱ biochemical, ecological and molecular ones, best approximates ȱ ȱ ¢¢ȱ ǻǰȱ ŗşşŞǼǯȱ ȱ ěȱȱ ȱ Ȭȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ data - improves the resolution and support of phylogenies due ȱ ęȱȱ ȱ ȱ ȱ ȱ ǻȱ et al., 2002; Edgar & Theriot, 2004; Wortley & Scotland, 2006; Lee & Camens, 2009), while examining data partitions independently gives insight ȱȱěȱȱȱȱ¢ȱȱǻǰȱŗşşŞǼȱȱ ȱ¢ȱȱȱěȱȱȱȱȱ (Cox & Williams, 2006). ȱ ¢£ȱ ȱ ěȱȱ ȱ ȱ ȱ characters: plastid, frustule appearance (LM) and frustule structure ǻǼǯȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ number of resolved clades. Frustule structures contained most of the phylogenetically informative characters (external raphe path, central raphe endings, raphe structure, alveolus opening). However, excluding plastid characters prevented the resolution of the “subgibba” and “grunowii” clades, supporting the phylogenetic value of plastid characters emphasized by Cox (1988) and Cox & Williams (2000). The highest number of resolved clades was obtained by ¢£ȱ ȱ ȱ ¢ǰȱ ęȱȱ ȱ ȱ ěȱȱȱȱǯ Based on the cladistic analyses, the most informative characters (CI=1, RI=1) for Pinnularia were the helictoglossa, girdle band areolae, surface irregularities, and alveolus opening ȱ ǯȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ Caloneis silicula and P. acrosphaeria and were not informative for resolving ȱęȱȱǯȱȱ ȱ ȱȱȱ ¢ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ȃȬ Ȭ” clade is characterized by a synapomorphic intermediate alveolus opening length. Krammer & Lange-Bertalot (1985) used the alveolus openings to distinguish three groups within the Pinnularia/Caloneis complex: open alveoli (in our terminology having a length of > 1/3 valve width), half-closed alveoli (1/8 < x < 1/3; intermediate length), and almost completely closed alveoli (< 1/8

138 Morphology of Pinnularia valve width), seen in LM as more central or marginal longitudinal bands (Krammer, 2000). The clade characterized by intermediate alveolus opening length in our phylogenies corresponded to their group of “ȱǭȱ¡, P. gibba, C. amphisbaena, C. latiuscula”. The two other groups were not recovered. Of course, we did not sample all Pinnularia and Caloneis species and further taxon sampling is needed to support this preliminary result. Based on both the morphological and molecular analyses, homoplasious characters include valve margins, apical shape, stria orientation (both central and apical), all non-structural characters, and plastid extension. All other characters showed ȱ ¢ȱ Ĵȱǯȱ ȱ Ȭȱ ȱ ȱ important for the delineation of phylogenetic groups, such as the axial area and presence of fascia for the “subgibba” clade, but on the whole structural characters were more informative. The number of ¢ȱȱ ȱȱȱĴȱǯȱȱȱ¢ȱ in the P. viridis group had already been noted by Schmid (2001) and we retrieved this in our “ȬȬ” clade. ¡ȱ ǻŗşŞŞǼȱ ȱ ȱ ȱ ȱ ¢ȱ ęȱȱ of pyrenoids in this genus, and also remarked the great variety in plastid shape. Similarly, we found taxa with linear and H-shaped plastids. However, it was not always clear whether H-shaped plastids were truly single chloroplasts, or the result of integration of the outer plastid membranes (pers. comm. David Mann). This can only be resolved by thorough examination of the complete cell- cycle and/or TEM observations. Therefore, it is not clear whether the H-shaped plastids characteristic of the “grunowii” clade are homologous to the H-shaped plastid of C. silicula or P. microstauron (see Figs in Appendix 2, q-q’’). It should be noted that the plastids of the “grunowii” clade and P. cf. microstauron have two pyrenoids, compared to one pyrenoid in C. silicula, which would indeed suggest a two-plastid origin for the former. Problems in establishing homology were also encountered with respect to the elongated central raphe endings of the “viridiformis” clade and C. silicula (compare Appendix 5.2, n and n’), and in the short alveolar openings of P. acrosphaeria and C. silicula. Determination of homology of morphological characters is one of the most problematic issues in cladistics (Pimentel & Riggins, 1987; Scotland & Pennington, 2000; Wagner, 2001; Rieppel & Kearney, ŘŖŖŘǼȱȱȱȱȱ¡ȱȱȱĴȱȱ (Maximino, 2008; Cox, 2010) and an understanding of the processes

Morphology of Pinnularia 139 underlying morphological evolution (Scotland et al., 2003). However, in diatoms developmental pathways have not been investigated for most frustule and plastid characters (but see Cox, 1999). ȱĴȱȱȱȱ¢ȱ¢ȱȱȱȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ¢ȱ Ĵȱȱ (Pagel, 1997). Ancestral state reconstruction (ASR) is a statistical method using MP or ML to estimate the character states of ancestral nodes based on the present distributions (Cunningham et al., 1998). In our analysis, the most parsimonious and most likely ancestral states for Pinnularia were mostly congruent with the morphology ȱȱȱ ȱȱǻȱǭȱ ǰȱŗşŜŞǲȱěȱȱ et al.ǰȱĴȱǼǰȱ¡ȱȱȱȱȱȱȱǰȱ although a single fossil species (P. spatula Lohman & Andrews) did have a fascia. Complex raphe systems and intermediate alveolar openings, which are characteristic for the “ȬȬ viridiformis” clade, are all derived character states for the genus, and this is in agreement with the presently known fossil sequence. Based on the fossil record, the larger viridis-like taxa are suggested to have evolved between the late Eocene and middle Miocene (Lohman & Andrews, 1968; Ognjanova-Rumenova & Vass, 1998). It is important to remember that phylogenies are only hypotheses of genealogies, ȱ ¢ȱ ȱ ȱ ¢ȱ ȱ Ěȱȱ ȱ reliability of the ASR (Pagel, 1997). The most problematic drawback of the present phylogeny of Pinnularia is the low taxon sampling. ǰȱȱȱȱęȱȱȱȱȱęȱȱȱ and broader clades within this large genus, and further taxon sampling of Pinnularia and Caloneis will undoubtedly improve these preliminary phylogenetic estimates.

Acknowledgements ȱ Ç²¤ǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ provided diatom cultures. Bart Van de Vijver, Elie Verleyen, Joris ěȱȱ ȱ ȱ ȱȱ ȱ ȱ samples. Alex Ball assisted with confocal laser scanning microscopy ȱ ȱ ȱ ¢ǯȱ ȱ ȱ ȱ ęȱ¢ȱ ȱ ¢ȱ Ȭȱ ǻȱ ȱ ęȱȱ ȱ Ȯȱ Flanders) through funding of CS, by the FWO-project 3G/0533/07, and by a European Union SYNTHESYS grant (GB-TAF-77) to CS.

140 Morphology of Pinnularia Appendix 5.1 Description of and comments on the selected frustule and plastid characters

Frustule outline 1. Margins: convex (0); parallel (1), undulate (2); constricted in the ȱǻřǼǲȱĚȱȱȱȱȱǻŚǼǯȱ The margins of the frustules in valve view were delimited from halfway along one raphe branch, beside the central area, to halfway along the second raphe branch. The character, margin, therefore describes the central portion of the valve. The relationship between the opposite margins can be convex (App. 2, a), parallel (App. 2, b), ȱǻǯȱŘǰȱǼǰȱȱȱȱȱǻǯȱŘǰȱǼȱȱĚȱȱ in the middle (App. 2, e). 2. Apices: cuneate (0); rounded (1); rostrate (2); capitate (3). ȱ¡ȱȱęȱȱȱȱȱȱȱȱǰȱȱ ¢ȱ along each raphe branch. In valve view, the outline of the apices can be abruptly narrowing toward the tip, like a wedge (cuneate; App. 2, f) or smoothly rounded (rounded; App. 2, g). Valve apices that have undulating margins towards the apex can be divided into two groups: the valve may narrow, but never become narrower than the apex itself (rostrate; App. 2, h), or it may narrow and then widen ȱ ȱȱ¡ȱǰȱȱȱȱěȱȬȱȱȱȱ “head” (capitates; App. 2, i).

5DSKHFKDUDFWHUV řǯȱȱȱȱ¡ȱȱęȱǱ curved (0); undulate (1). The line of the raphe slit which is visible on the outer side of ȱȱȱȱȱȱȱ¡ȱȱęȱǯȱȱęȱȱ can be continuously curved to one side of the valve only (curved; App. 2, j), or can switch its course between the two sides of the valve, and undulate over the valve (undulate; App. 2, k). 4. Central raphe endings: elongated drop-like (0); round (1), linear (2). The central raphe endings were observed using SEM and can be elongated and drop-like (App. 2, l), round (App. 2, m) or linear (App. 2, n-n’). 5. Helictoglossa: tongue-like (0); hooked (1). The helictoglossa (internal, apical raphe ending) was observed using SEM and can have an elongated, tongue-like form (tongue-

Morphology of Pinnularia 141 like; App. 2, o) or be hooked (App. 2, p). 6. Raphe complexity: simple (0); three lines visible (1); complex (2). In cross section, the raphe slit has a parallel “groove” and a ȃȄȱǻȱ ȱŗşşŘȱȱŘȱęȱȱŗȬŚǼǰȱ ȱęȱȱȱȱȱ other (App. 2, t). The edge of the tongue and the two edges (external and internal) of the groove are visible in LM as thin, longitudinal lines. The number of visible lines depends on the relative positions of these margins. Usually, the outer and inner margins of the groove are superimposed and visible as a single line. In this case, two thin lines are visible: the groove margins and the tongue edge. Using LM, the course of the groove margin can be compared with the straight course of the tongue margin. If the groove margin follows the course of the tongue margin without crossing it, the raphe is termed “simple” and two thin, parallel lines are visible (App. 2, q). If the ȱȱȱȱȱȱȱȱěȱȱȱ ȱȱȱȱȱǰȱȱȱȱęȱȱȱȃ¡Ȅȱ (App. 2, s). A special case of the simple raphe occurs when the two margins of the groove are not superimposed but are visible as separate lines. In LM, three non-crossing, thin lines are then visible (two lines of the groove and one line of the tongue). Because of the ¢ȱȱěȱǰȱ ȱęȱȱȱȱȱȱ character state (three thin lines; App. 2, r).

Axial area 7. Axial area: ȱ ȱȱȬǱȱǱȱŗȦŚȱȮȱŗȦŜȱ (0); narrow: < 1/6 (1); broad: > 1/4 (2). The relative width of the axial area was determined halfway along the raphe branches in LM and compared with the total width of the valve at the same position. 8. Axial area: relative width at central raphe endings: narrow: < 1/2 (0); broad: between 1/2 and 1/1 (1); valve width: 1/1 (2). The relative width of the axial area was determined at the junction between the central raphe ending and the raphe branch in LM, and compared with the total width of the valve at the same position.

Central area 9. Fascia: absent (0); unilateral (1); bilateral (2). ȱȱȱȱȱęȱȱǻǼǰȱȱȱȱǯȱ A fascia can be absent (App. 3, a), only present on one side of a valve

142 Morphology of Pinnularia (unilateral; App. 3, b), or present at both sides of a valve (bilateral; App. 3, c). If both unilateral and bilateral fasciae were present on ěȱȱȱȱȱȱǰȱ ȱęȱȱȱȱȱȱ a “bilateral” fascia. 10. Central area: shape: elliptical (0); diamond (1); rectangular (2). We delimited the central area by the junctions between the central pores and the raphe branches, thus including the central pores. The shape of the central area is determined by the arrangement of the adjacent striae. The striae can lengthen very gradually towards the apices, forming a convex edge and creating an elliptical central area (App. 3, d), or can lengthen more abruptly, forming a concave rim and creating a diamond-shaped central area (App. 3, e). If striae are absent or no variation in stria length is observed, the central area is rectangular (App. 3, f).

6XUIDFH 11. Ghost striae: absent (0); present (1). ȱ ȱ ȱ ¢ȱ ęȱȱ ȱ ȱ ȱ ȱ ¢ȱ ęȱȬȱ ȱ ȱ ǯȱ ȱ ȱ ȱ present, but the outlines of the striae are more or less visible and create parallel darker shadows. Ghost striae are for example typical for the gibba-group of Pinnularia and can be seen as four spots in the central area (indicated by arrows in App. 3, g). ŗŘǯȱȱǱ absent (0); present (1). On the outside of the valve, the central and axial areas can be ȱ ȱȱęȱȱ ȱǻǯȱřǰȱǼǯ

Striae 13. Stria orientation: centre: radiate (0); transverse (1). The striae at the centre of the valve can be oriented parallel to the transapical axis of the valve (transverse; App. 3, j) or oblique to the transapical axis of the valve, the proximal (to the raphe) ends of the striae being closer to the valve centre than the distal ends (radiate; App. 3, i). 14. Stria orientation: apex: convergent (0); radiate (1); transverse (2). The striae near the valve apices can be oriented parallel to the transapical axis of the valve (transverse; App. 3, k). They can also be oblique to the transapical axis of the valve, distal ends of the striae being closer to the valve centre than the proximal ends (convergent; App. 3, i), or the proximal (to the raphe) ends of the striae being

Morphology of Pinnularia 143 closer to the valve centre than the distal ends (radiate; App. 3, l).

Alveoli ŗśǯȱȱȱǱȱproportion of half valve width: > 2/3 ǻŖǼǲȱŘȦřȱȮȱŗȦŚȱǻŗǼǲȱǀȱŗȦŚȱǻŘǼǯ The alveoli open on the inner side of the valves and are visible using SEM as openings in the chamber roofs above the striae. The length of the alveolus opening was determined halfway along the raphe branches and was calculated relative to the distance between the raphe slit and the valve margin at the same location. These ȱȱȱǁȱŘȦřȱǻǯȱřǰȱǼǲȱŘȦřȱȮȱŗȦŚȱǻǯȱřǰȱǼǲȱǀȱŗȦŚȱ (App. 3, o) ŗŜǯȱȱǱ location: central (0); distal (1). Alveolus openings located halfway between the margin and the raphe are termed “central” (App. 3, m-n), while alveolus openings located adjacent to the valve margins are termed “distal” (App. 3, o).

Girdle bands 17. Girdle bands: shape of areolae: elongated (0); round (1). The girdle bands were visualized using SEM and bear a single row of areolae. Areolae can be elongated or roundish in shape.

Plastids 18. Plastid: shape: linear (0); H-shaped (1). Plastids were observed in non-synchronized, exponentially growing cultures using LM and confocal laser scanning microscopy (CLSM). Cells could contain two large, linear plastids lying along each side of the girdle and extending under the valves (linear; App. 3, p-p’’), or a single plastid, comprising two linear plates lying along the girdle and extending under the valves, connected by a narrow central bridge against one valve face, above the nucleus ǻ ȬǲȱǯȱřǰȱȬȂȂǼǯȱȱȱȱȱȱĜȱȱȱȱȱȱ ȱǰȱȱ ȱȱȱęȱȱȱǯȱȱ were categorized as having an H-shaped plastid if a bridge was observed in all examined cells. 19. Plastid: extension under valve face: narrow (0); broad (1). Both linear and H-shaped plastids have margins that extend under the valve face. The degree to which the margins extended was described as “narrow” if it was less than or equal to 1/4 valve width,

144 Morphology of Pinnularia and described as “broad” if it occupied more than 1/4 of the valve width. 20. Pyrenoids: number per plastid: one (0); two (1); absent (2). Pyrenoids are proteinaceous bodies located in the plastid, which contain high levels of the enzyme ribulose-1,5-biphosphate ¡¢Ȧ¡¢ǰȱ ȱ ȱ ŘȬęȱ¡ǯȱ ȱ ȱ material is present around the pyrenoids in diatoms. Using regular ȱȱǰȱ¢ȱ ȱȱȱȱǻǼȱȱȬĚȱȱ (CLSM) regions in the plastid (indicated by arrows in App. 3, p-q’). We used additional staining with Azocarmine G to improve visualization of the pyrenoids in LM. If no staining was observed, ȱȱ ȱęȱȱȱȱ¢ȱǻǼǯ

Morphology of Pinnularia 145 Appendix 5.2 Appendix 5.2ȱȱęȱȱȱȱȱȱȱȱǯ

146 Morphology of Pinnularia Appendix 5.3 Appendix 5.3ȱȱęȱȱȱȱȱȱȱȱǯȱ

Morphology of Pinnularia 147

6. Molecular evidence for the existence of distinct Antarctic lineages in the cosmopolitan terrestrial diatoms Pinnularia borealis and Hantzschia amphioxys

ȱ ěȱ1, Pieter Vanormelingen1, Bart Van de Vijver2, Tsvetelina Isheva1, Elie Verleyen1, Koen Sabbe1 and Wim Vyverman1

Unpublished manuscript

Genetic divergence in P. borealis and H. amphioxys 149 Abstract Recent studies based on morphology revealed that the Antarctic continent has a high percentage of endemic freshwater diatom species, besides a considerable percentage of presumed cosmopolitan species. Given the widespread (pseudo)cryptic species diversity in diatoms, we used two of these cosmopolitan taxa, Pinnularia borealis and ĵȱȱ¡¢, to assess the molecular divergence between Antarctic strains and strains from other locations. Molecular phylogenies based on the plastid gene L and the nuclear 28S rDNA (D1-D2 region) revealed that P. borealis and ǯȱ¡¢ are two species complexes consisting of multiple genetically distinct lineages, each including a distinct Antarctic lineage. A molecular time-calibration estimates the origin of the species complex P. borealis řśǯŜȱȱ¢ȱǻǼȱǰȱȱȱęȱȱ¡ȱȱĴȱȱȱ 22.0 Ma, making this the oldest known diatom species complex. The Antarctic P. borealis lineage is estimated to have diverged 7.7 (15-2) Ma ago from a European sister lineage, largely after the ęȱȱȱȱȱȱȱ ȱȱȱȱ isolation of the continent, and after the start of the Mid-Miocene cooling event which resulted in the formation of a permanent ice sheet and full polar conditions. In addition, the Antarctic lineages of both P. borealis andȱ ǯȱ¡¢ diverged physiologically from most lineages from more temperate regions and tend to have a higher relative growth rate at low temperature, a lower optimal temperature and lower lethal upper temperature, indicating niche ěȱǯȱȱǰȱȱȱȱȱ¢ȱȱȱ presumed cosmopolitan Antarctic diatom species are in fact species complexes, possibly containing Antarctic endemic species.

Key words: thermal adaptation, L, LSU rDNA, cryptic diversity, diatoms, biogeography, Antarctica, dispersal limitation, allopatric speciation

150 Genetic divergence in P. borealis and H. amphioxys 6.1 Introduction The Antarctic continent is a geographically and thermally isolated area. Since the breakup of Gondwana 180 Ma ago it gradually drifted away from other land masses, resulting in a strong geographical isolation. During the Eocene (50-32 Ma), the gradual opening of the ocean between Antarctica and Australia respectively South America triggered the establishment of circumpolar winds and ocean currents which blocked heat transfer from the tropics. This resulted in the thermal isolation of the Antarctic continent and the formation of a continental ice sheet (Lawver & Gahagan, 2003; Barker & Thomas, 2004; Brown et al., 2006). Since at least 14 Ma, the Antarctic continent is therefore a cold and windy desert, characterized by extreme climatic conditions constituting an ecological challenge for any resident or colonizing organism (Block et al., 2009). The remaining ice-free regions ȱȱȱ¢ȱĴȱǰȱ¢ȱȱȱǰȱ ȱ¢ȱĚȱȱ¢ȱȱ¢ǰȱȱȱȱȱ extinction risks and low colonization rates (Pugh & Convey, 2008). As a result, the Antarctic region, and especially continental Antarctica, ȱ£ȱ¢ȱȱ¢ȱȱĚȱȱȱȱǻ ȱ Smith, 1984) and a high incidence of endemism in both freshwater and terrestrial habitats for macroscopic organisms such as dipterans (Currie & Adler, 2008), pycnogonids (Munilla & Membrives, 2009) and Collembola and Oribatida (Greenslade, 1995; Marshall & Pugh, 1996). Based on discontinuities in species distributions for several of the above taxonomic groups, the continent is subdivided into two biogeographic regions, namely the continental and maritime ȱǰȱ ȱȱȱ¢ȱȱȬȱ Ĵȱȱȱ situated in the South of the Antarctic Peninsula (Chown & Convey, 2007). For microalgae, inferring species’ geographic distributions was until recently hampered by undersampling and taxonomical problems due to the limited number of available morphological characters and the uncertain interpretation of small morphological ěȱǯȱ ȱ ȱ ǰȱ ¢ȱ ǻŗşşŜǼȱ ȱ ȱ the Antarctic diversity is low, and lower in continental than in maritime Antarctica, with only a small element of endemic species. This conclusion is now contradicted by molecular phylogenies with separate Antarctic lineages of cyanobacteria and green algae, suggesting a long history of isolation on the continent (Taton et al., 2003; Taton et al., 2006; De Wever et al., 2009). For lacustrine diatoms, a substantial body of work in the (sub-)Antarctic based on careful

Genetic divergence in P. borealis and H. amphioxys 151 morphological analysis showed (1) a highly impoverished diatom Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ¢ȱ et al., 2007), (2) an overrepresentation of typical terrestrial genera and poverty of globally successful genera and planktonic taxa (Verleyen et al.ǰȱ ĴȱǼǰȱ ǻřǼȱ ȱ ȱ ȱ ȱ ȱ ȱ provinces, being the sub-, maritime, and continental Antarctic (Verleyen et al.ǰȱ ĴȱǼǰȱ ȱ ǻŚǼȱ ȱ ȱ ȱ ȱ ǻȬǼ Antarctic endemic species (27-63%, Verleyen et al.ǰȱ ĴȱǼȱ ȱĴȱȱȱ ȱȱȱĚȱȱǻȱȱȱet al., 2005). Together, this suggests that colonization is limited on both a regional and continental scale, and that allopatric speciation, or ȱȱȱĚȱȱȱȱȱȦȱȱ selection during long-term geographic isolation (after rare long-term ȱȱȦȱȱȱȱȱȱȱ or continental drift) (Mayr, 1963; Coyne, 1992; Rice & Hostert, 1993), ȱȱȱȱęȱȱȱȱȱȬȱȱ microorganisms, contrary to the hypothesis that microorganisms have unlimited dispersal probabilities (Beijerinck, 1913; Finlay, 2002; Finlay et al., 2002). However, besides this high number of endemic morphospecies, there also appears to be a substantial fraction of cosmopolitan diatom morphospecies on the Antarctic continent (Verleyen et al.ǰȱĴȱǼǯȱ How this should be interpreted is unclear, given the prevalence of (pseudo)cryptic species diversity in diatoms (e.g. Sarno et al., 2005; Evans et al., 2008; Trobajo et al., 2009). They might be truly generalistic, cosmopolitan species which have (recently) colonized the Antarctic, or it might concern cryptic species complexes containing genetic lineages with potentially restricted geographic distributions (Boo et al., 2010) and ecological niches (Vanelslander et al., 2009; Vanelslander et al., unpubl.). In that respect, it is well-known that many Antarctic algae have evolved special ecophysiological properties in response to the extreme climatological conditions (Kirst & Wiencke, 1995). For instance, several marine diatoms are obligately psychrophilic: optimum temperatures are at 3°-5° C and temperatures above 6°- 8° C are lethal (Fiala & Oriol, 1990). Cryptic lineages (or species, ȱ£ǰȱŘŖŖŝǼȱȱȱęȱȱȱȱ¢ȱ based on species-level molecular markers (e.g. De Vargas et al., 1999; Beszteri et al., 2007; Boo et al.ǰȱ ŘŖŗŖǼǯȱ ȱ Ĝȱȱ¢ȱ ǰȱ such phylogenies can be time-calibrated using the fossil record, feasible for diatoms given their resistant siliceous cell walls (Sims et al., 2006), and combined with geographical distributions and niche

152 Genetic divergence in P. borealis and H. amphioxys £ȱȱȱȱĴȱȱȱȱȱ¢ȱ history of these (cryptic) lineages (see e.g. Darling et al., 2007; Verbruggen et al., 2009). The diatom morphospecies Pinnularia borealis Ehrenberg and ĵȱȱ¡¢ (Ehrenberg) Grunow are two common taxa in mainly terrestrial habitats on all continents, including Antarctica (Krammer & Lange-Bertalot, 1986, 1988). For both taxa, some morphological variants have been described as varieties, but there is morphological overlap and as a result their taxonomic status is uncertain (Cleve-Euler, 1952; Krammer & Lange-Bertalot, 1986; Krammer, 2000). In this study, we (1) test whether both taxa are species complexes with a distinct Antarctic lineage or not, (2) estimate divergence times for any such lineage(s) in P. borealis based on a time-calibrated Pinnulariaȱ ¢¢ȱ ǻěȱȱ et al.ǰȱ ĴȱǼǰȱ and (3) determine the temperature preferences of Antarctic strains in comparison to their more temperate counterparts. To this end, we isolated monoclonal strains of P. borealis and ǯȱ¡¢ from continental Antarctica and (cold) temperate regions in Europe, the Americas and Asia. These were used to construct molecular phylogenies based on L and the D1-D2 region of 28S rDNA, calculate lineage divergence times using molecular-clock methods for P. borealis, and assess strain growth rates over a temperature gradient from 3°-32°C.

6.2 Materials & Methods Taxon sampling We used 52 monoclonal strains of Pinnularia borealis (Table 6.1) and 31 of ĵȱ, including ǯȱ¡¢, H. abundans and ĵȱȱ ǯȱ ǻȱ ŜǯŘǼȱ ȱ ȱ ěȱȱ ǰȱ ȱ ȱ Schirmacher Oasis on continental Antarctica. Geographic origin of the strains is listed in Tables 6.1 and 6.2. In the future, more detailed information about sampling dates and localities will be available in the barcode database (www.boldsystems.org) in which all used strains will be included. Per location, multiple environmental samples were used. Environmental samples were taken by scraping the upper half cm of bare soil using a sterile falcon tube. For the non-Antarctic samples, a small amount of soil was incubated upon arrival in the lab in liquid WC medium (Guillard & Lorenzen, 1972, but without pH adjustment and vitamin addition) at standard culture conditions of 18°C, 25-30 μmol ph m-² s-1 light

Genetic divergence in P. borealis and H. amphioxys 153 exp. Temp. L LSU

rbc Striae density (#/10 μ m) μ m) Width ( Morphometric data (± stdv.) ( μ m) Length 33.0±1.733.0±1.7 8.1±0.3 7.9±0.4 5.3±0.4 5.2±0.3 x x x x x x 40.9±0.7 8.1±0.333.7±0.8 6.0±0.2 7.6±0.3 x 6.0±0.0 - x x x 34.9±0.8 7.4±0.4 6.6±0.5 x - 41.9±1.1 8.5±0.5 5.4±0.4 x x 31.6±0.831.1±1.0 7.1±0.332.6±1.1 8.2±0.335.4±1.4 6.5±0.7 7.5±0.332.0±1.2 5.7±0.5 8.0±0.328.3±1.2 x 6.0±0.0 7.9±0.3 x 5.5±0.5 8.4±0.3 x x 6.0±0.0 x x 5.6±0.5 x - x x x - - - x x 39.6±1.2 7.6±0.335.9±0.6 6.2±0.4 8.2±0.3 x 5.8±0.4 - x x - Origin Belgium - Gent Belgium - Gent Belgium - Gent Belgium - Gent Belgium - Gent Canada – Banff Canada – Banff Canada – Banff Canada – Banff Canada – Banff Canada – Banff France – Ailfroide France France – Ailfroide France – Ailfroide France rbc ȱȱǯȱȱȱȱȱȱ¢ȱȱǯȱ ȱȱȱȱ Krammer Krammer Krammer Mongolia – Khangai-Nuruu Krammer Mongolia – Khangai-Nuruu 55.2±0.9 Mongolia – Khangai-Nuruu 56.8±1.0 Mongolia – Khangai-Nuruu 10.5±0.3 55.2±0.7 10.9±0.7 55.0±0.6 5.1±0.2 10.9±0.4 5.0±0.0 10.9±0.3 x 5.1±0.2 x 5.0±0.0 x x x x x x x x Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) FDWLRQ

borealis borealis borealis borealis borealis borealis borealis borealis borealis islandica islandica islandica islandica

scalaris scalaris scalaris scalaris scalaris var. var. var. var. var. var. var. var. var. var. var. var. var. var. var. ,GHQWL¿ var. var. var. var. var. var.

P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P.

P. borealis P. borealis P. borealis P. borealis P. borealis P. (E7)c (E7)a (E7)b (St1)i (St1)k (St1)d (St1)b (St1)g (Mo1)j (Mo1)f Strain (Mo1)c (Mo1)k (Ban2)c (Ban1)a (Ban2)a (Ban2)e (Ban1)g (Ban2)d Table 6.1 List of strains Pinnularia borealis ȱȱȱȱ¢ǰȱȱ ȱȱęȱǰȱȱǰȱȱȱǻƹȱȱ Table Ǽǰȱȱ¢ȱǻ¡Ǽȱȱȱȱȱ ȱ¡ȱȱȱȱȱȱǯ

154 Genetic divergence in P. borealis and H. amphioxys exp. Temp. L LSU xx x x xx xx x x x x x x

rbc Striae density (#/10 μ m) μ m) Width ( Morphometric data (± stdv.) ( μ m) Length 37.7±0.9 8.8±0.3 5.6±0.5 x x 37.1±0.8 8.7±0.5 6.3±0.4 x x 34.2±1.7 9.1±0.4 5.6±0.6 x - 34.6±0.7 7.2±0.4 6.7±0.5 x x 39.7±1.028.7±1.0 9.9±0.4 8.4±0.3 5.6±0.5 6.5±0.4 x x x - 34.4±2.4 9.0±0.6 6.3±0.5 x - 36.8±1.6 9.5±0.5 5.0±0.0 x x 38.6±1.142.3±0.6 7.4±0.3 8.9±0.4 6.2±0.4 5.0±0.0 x x x x x x Origin Belgium - Gent Belgium - Gent Belgium - Gent Belgium - Gent Chile – Seno Otway Chile – Seno Otway Chile – Seno Otway Chile – Seno Otway Chile – Seno Otway Chile – Seno Otway Chile – Torres del PaineChile – Torres 27.3±0.6 9.6±0.3 6.0±0.0 x x x Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Antarctica – Schirmacher Oasis Antarctica – Schirmacher Oasis (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) Rabenhorst (Ehr.) – Podkova republic Czech Rabenhorst (Ehr.) – Podkova republic Czech Rabenhorst (Ehr.) 37.4±1.3 – Podkova republic Czech 38.6±0.8 – Podkova republic Czech 8.2±0.6 37.4±1.0 8.2±0.2 38.1±1.3 5.7±0.4 8.1±0.3 5.4±0.5 8.2±0.4 x 5.8±0.3 x 5.2±0.4 x x x x x x x x FDWLRQ

borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis var. unknown var.

scalaris scalaris scalaris scalaris scalaris scalaris scalaris var. var. var. var. var. var. var. var. var. var. var. var. var. var. ,GHQWL¿ var. var. var. var. var. var. var. var. P. borealis

P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P.

P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. E1 D7 D6 D8 D14 D10 Alka2 Alka3 Alka4 Alka1 (St6)c (St6)a (St8)a (St8)d Strain (Tor3)a (Tor12)i (Tor12)f (Tor12)e (Tor12)g (Tor12)d (Tor12)h

Genetic divergence in P. borealis and H. amphioxys 155 exp. Temp. L LSU xx x x xx - x x x x xx xx x x x x x x x x xx xx x x x x

rbc Striae density (#/10 μ m) μ m) Width ( Morphometric data (± stdv.) ( μ m) Length Origin Ehrenberg Ehrenberg Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Ehrenberg Antarctica – Schirmacher Oasis Antarctica – Schirmacher Oasis Antarctica – Schirmacher Oasis FDWLRQ

borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis borealis var. var. var. var. var. var. var. var. var. var. var. var. var. var. ,GHQWL¿

P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. borealis P. P8 V2 V4 P10 P12 P14 P17 S18 E10 E13 V15 V18 V20 Strain

156 Genetic divergence in P. borealis and H. amphioxys exp. Temp. L LSU x -

rbc μ m) (#/10 Fibulae density μ m) Striae (#/10 density ( μ m) Width Morphometric data (± stdv) ( μ m) Length 37.7±0.7 6.3±1.0 21.9±0.6 7.9±0.7 x x 42.9±0.4 6.2±0.1 25.2±0.9 1.1±0.7 x42.8±0.6 x 6.4±0.236.0±0.6 24.7±1.1 6.0±0.438.1±0.8 x 7.7±0.7 25.8±0.4 6.0±0.3 8.0±0.9 24.1±0.9 x 9.3±0.7 x x x x x x x 48.3±0.9 9.0±0.436.8±1.8 21.6±0.7 8.7±0.3 6.9±0.9 21.2±0.936.8±0.5 6.9±0.7 6.6±0.5 x 23.7±1.0 x 9.2±0.7 x x x x x 39.2±1.1 7.5±0.4 21.7±0.738.1±1.1 8.3±0.9 7.9±0.643.6±0.8 21.4±0.7 7.7±0.440.4±0.5 x 6.8±1.2 21.1±0.9 7.8±0.3 7.9±0.4 x 21.0±1.1 x 7.2±0.9 x x x x x Origin Canada – Banff Canada – Banff Canada – Banff Canada – Banff Canada – Banff Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent rbc ȱȱǯȱȱȱȱȱȱ¢ȱȱǯȱ ȱȱȱȱ Mongolia – Kangai-NuruuMongolia – Kangai-NuruuMongolia – Kangai-Nuruu 85.2±2.6Mongolia – Kangai-Nuruu 8.8±0.3 86.4±3.1Mongolia – Kangai-Nuruu 20.0±0.8 8.3±0.5 88.4±0.5 6.9±1.0 20.2±1.2 8.5±0.5 87.6±1.3 7.2±0.8 19.4±0.7 8.7±0.6 87.9±4.5 x 7.1±0.3 19.9±0.5 8.9±0.5 x 7.2±0.5 x 19.8±0.8 x 7.8±0.6 x x x x x x ǯȱȱȱȱ¢ǰȱȱ ȱȱęȱǰȱȱǰȱȱȱǻƹȱȱ ȱ ȱ (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) FDWLRQ (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.)

H. sp. 1 H. sp. 1 H. sp. 1 H. sp. 1 H. sp. 1 H. sp. 2 H. sp. 2 H. sp. 2 H. sp. 2 H. sp. 2 H. sp. ,GHQWL¿ mphioxys

amphioxys cf. cf. a cf. H. amphioxys H. amphioxys H. amphioxys H. amphioxys H. amphioxys H. cf. amphioxys H. cf. H. H. (St1)f (St3)c (St3)a (St1)e (St4)a (St4)b (St2)d (St1)h (Mo1)l Strain (Mo1)e (Mo1)a (Mo1)h (Ban1)f (Mo1)m (Ban1)e (Ban1)d (Ban1)b (Ban1)h Table 6.2 List of strains ĵ Table Ǽǰȱȱ¢ȱǻ¡Ǽȱȱȱȱȱ ȱ¡ȱȱȱȱȱȱǯ

Genetic divergence in P. borealis and H. amphioxys 157 exp. Temp. L LSU - - x x x x xx - xx x x x x x x x

rbc μ m) (#/10 Fibulae density μ m) Striae (#/10 density ( μ m) Width Morphometric data (± stdv) ( μ m) Length 46.8±0.7 8.2±0.3 20.3±1.3 7.6±0.7 x x 68.1±2.8 9.7±0.4 17.1±1.0 6.6±0.5 x x x 56.1±0.2 7.1±0.651.8±0.2 20.4±0.5 7.3±0.3 6.8±0.4 19.6±0.9 7.6±0.9 x x x x 52.3±0.5 7.2±0.452.3±0.4 19.7±0.7 7.3±0.4 6.8±1.0 18.8±1.0 7.9±1.0 x x x x Origin Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent Belgium – Gent Belgium – De Panne Belgium – De Panne Chile – Torres del PaineChile – Torres 68.1±0.9 8.1±0.5 18.8±1.4 7.2±1.2 x x (Ehr.) Grunow (Ehr.) FDWLRQ (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) – Schirmacher Oasis Antarctica Grunow (Ehr.) – Schirmacher Oasis Antarctica Grunow (Ehr.) – Schirmacher Oasis Antarctica – Schirmacher Oasis Antarctica – Schirmacher Oasis Antarctica Lange-Bertalot Lange-Bertalot Lange-Bertalot

H. sp. 3 H. sp. 3 H. sp. ,GHQWL¿

amphioxys cf. cf. H. amphioxys H. amphioxys H. abundans H. abundans H. abundans H. amphioxys H. amphioxys H. amphioxys H. amphioxys H. amphioxys H. S11 S13 S16 S15 S10 (St6)f (St8)c (St5)c (St6)e (St6)b Strain (Tor3)c (Wes3)b (Wes1)b

158 Genetic divergence in P. borealis and H. amphioxys intensity and a 12:12 hours light:dark period. Within the next week, cells resembling P. borealis and ǯȱ¡¢ were isolated under a ȱȱȱȱȱĴȱǯȱȱȱ ȱ maintained under standard culture conditions and re-inoculated when reaching late exponential phase. Antarctic samples were kept cold during transport and were incubated in WC medium at 5°C and low light intensities (ǯ 5 μmol ph m-² s-1) to prevent the death of possibly psychrophilic diatoms. Isolations of cells using needle and Ĵȱȱ ȱȱ ȱȱ¡ȱŗŖȱ¢ǯȱȱ ȱ ȱęȱȱȱȱŞǚȱȱȱȱŗśǚȱȱȱȱ demonstrated that the Antarctic cultures tolerated this temperature.

Morphological observations Cultures were harvested for morphological observations in late exponential phase, oxidized using hydrogen peroxide and embedded in Naphrax®. Pictures were taken using a Zeiss Axioplan 2 microscope equipped with an AxioCam Mrm camera. Valve length, width and stria density of 10 valves per strain were measured using ȱ ŗǯřŝȱ ǯȱ ęȱȱ ȱ ȱ ȱ ȱ (2000) for Pinnularia borealis and Lange-Bertalot (1993) for ĵȱ. ęȱȱȱȱȱȱȱȱȱŜǯŗȱȱ 6.2, respectively. Voucher material is kept in the Laboratory of Protistology & Aquatic Ecology (Gent University, Belgium) and is available upon request.

Molecular-genetic analyses Cultures were harvested for genetic analyses in exponential phase, and DNA was extracted from centrifuged diatom cultures following Zwart et al. (1998) using a bead-beating method with phenol extraction ȱȱǯȱȱ¡ǰȱȱ ȱęȱȱ ȱȱ Wizard® DNA Clean-up system (Promega). Sequences of the D1-D2 region of the nuclear 28S rDNA and of the plastid gene L were ęȱȱ ȱȱȱȱȱȱǻȱ et al., 1994; Daugbjerg & Andersen, 1997; Jones et al., 2005). Primer sequences, PCR conditions and sequencing protocol are described in ȱȱěȱȱet al. ǻĴȱǼǯ Outgroups of P. borealis were based on a genus phylogeny of Pinnulariaȱǻěȱȱet al.ǰȱĴȱǼȱȱȱǯȱǯȱ (B2)c, ǯȱǯ (Wie)a, ǯȱǯȱPin873, ǯȱ var. elongata Krammer Corsea10, ǯȱ W. Smith Pin876, ǯȱ Krammer

Genetic divergence in P. borealis and H. amphioxys 159 Corsea 2, ǯȱ Krammer Pin706. For ǯȱ ¡¢, all available strains of ĵȱȱęȱȱȱȱȱ ǯȱ¡¢ were used as outgroup (Table 6.2). Sequences were separately edited and automatically aligned using ClustalW (Thompson et al., 1994) implemented in BioEdit 7.0.3 (Hall, 1999). Plastid sequences aligned unambiguously without any gaps. The alignment of 28S was corrected manually using the secondary structure of ȱ radians (Ben Ali et al., 2001) after which ambiguously aligned regions were removed. For P. borealis, sequence lengths of 1,428 and 511 sites were used for L and 28S rDNA, respectively. For ĵȱ this was 1,457 and 525 sites, respectively. The L alignement was ȱȱȱęȱȱ ǰȱ ȱȱȱȱȱ a separate third codon position. The 28S was analyzed as a single partition. Datasets of Pinnularia borealis and ĵȱ ¡¢ were analyzed separately. Model selection was performed using TreeFinder (Jobb et al., 2004) on the concatenated dataset using a three-partition strategy (28S, and two codon positions of L) and on the single genes, and based on the Bayesian Information Criterion (BIC, Schwarz, 1978) as a selection criterion. Individual genes and concatenated datasets were analysed by maximum likelihood phylogenetic inference using RAxML 7.2.6 ǻǰȱŘŖŖŜǼȱȱȱȱ ƸɫŚȱȱ ȱŗŖǰŖŖŖȱ independent tree searches from randomized MP starting trees. Maximum likelihood bootstrap analyses (Felsenstein, 1985) consisted of 1,000 replicates. Bayesian inference was applied using MrBayes v.3.1.2 (Huelsenbeck & Ronquist, 2001; Ronquist & Huelsenbeck, 2003) with the same models and partition schemes. Using default Ĵȱǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ Metropolis-coupled Monte-Carlo Markov Chains were run for ten million generations for the individual genes and 30 million generations for the concatenated datasets. Runs were sampled every 1,000th generation and convergence and stationarity of the log- likelihood and parameter values was assessed using Tracer v.1.5 ǻȱ ǭȱ ǰȱ ŘŖŖŝǼǯȱ ȱ ęȱȱ ŗŖƖȱ ȱ ȱ discarded as burn-in for the individual gene and multi-gene analyses. ȱ ȬȬȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ £ȱ and posterior probabilities (PPs) were calculated in MrBayes using the sumt command. All analyses were performed using the publicly available computer resource Bioportal (Kumar et al., 2009).

160 Genetic divergence in P. borealis and H. amphioxys Molecular time-calibration Based on the molecular dataset, the phylogenetic relationships of P. borealis were placed in a time frame by constructing a time-calibrated phylogeny. Calibration of the molecular clock was based on the 95% highest posterior desities (HPD) resulting from the preferred time- calibration of the Pinnulariaȱ ȱ ¢¢ȱ ȱ ěȱȱ et al. ǻĴȱǼȱ ȱ ȱȱ¢ȱȱȱǯȱȱȱ of the phylogeny of P. borealis were constrained (gray circles in Fig. 6.2) using each a uniform distribution of their recovered 95% HPD. L was partitioned into codon position (1st +2nd combined), LSU was used as a single partition. Metropolis-coupled Monte-Carlo Markov Chains were run by BEAST v1.5.4 (Drummond & Rambaut, ŘŖŖŝǼȱ ȱ ȱ ȱ Ƹȱ ̆Śȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ǰȱ uncorrelated lognormal clock model and Yule tree prior. Three independent analyses, each started by a UPGMA tree, were run for 100 million generations and sampled every 1,000th generation. Burn- in values of 30 till 40% were assessed using Tracer v.1.5 (Rambaut & Drummond, 2007). All post-burn-in trees were combined after which the maximum clade credibility chronogram with mean node heights was calculated using TreeAnnotator v.1.5.4.

Temperature preference experiments For P. borealis, we selected per environmental sample two genetically characterized strains for the experiments, giving in total 22 strains (Table 6.1). The cells of the cultures from the Chilean samples Tor12 and Tor3 were too small to be used. For ĵȱ, we selected ¢ȱ ȱ ȱ ęȱȱ ȱ ǯȱ ¡¢ and their genetically nearest sister taxon (St3)a. All Antarctic strains and 1 or 2 strains per lineage were analyzed (depending on the availability), giving in total 11 strains (Table 6.2). Before experimental treatment, all strains were grown in standard culture conditions at 15°C during at least 2 months. The growth rate of the strains was assessed over a temperature range from +2.9°C to +32.4°C using a temperature ȱȱǻǰȱǯǼȱȱĚȱȱȱȱȬ adapted cells (F0) as a proxy for density (Consalvey et al., 2005). ȱĚȱȱǰȱȱ ȱȬȱȱ 15 minutes at 18°C, and F0 was measured daily during 7 days by ȱ ȱ ȱ ǻǼȱ Ěȱȱ ȱ ȱ ĵȱȱ ȱ Ȭȱ ǻȬǼȱ ȱ ȱ Ĵȱǰȱ ¢ȱ ȱ 12, gain of 8 and damping of 1. The experiments were carried out in

Genetic divergence in P. borealis and H. amphioxys 161 ȱȱ ȱȱŘŚȬ ȱȱȱȱȱ ȱęȱȱ ȱşȱ series of 4 plates. Each of the four rows of wells in a series of plates ȱȱȱęȱȱǰȱ ȱȱȱȱŘŚȱ ȱǻŚȱ plates in a series, 6 wells per plate) were available per temperature in ȱȱ¡ǯȱ ȱȱŗŘȱěȱȱ¡ȱȱ were selected, spread over the 9 series of plates. Per temperature, strains were randomly inoculated in the wells. Replicates were ȱȱěȱȱ¡ȱȱȱ¢ȱȱěȱȱȱ experiment. ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ¢ȱ condition, all strains were prior to the experiment inoculated at ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ Ŗȱ ȱ Ěȱȱ ȱ ȱ ŖǯŖśŖȱ ȱ ȱ Ȭȱ Ĵȱȱ ȱ ȱ during 5 days in standard culture conditions in 6-well plates. Next, all strains were again re-inoculated at an F0-value of 0.050 and grown for 5 days during which the culture medium was refreshed daily to keep them in exponential phase. Together, this treatment ensures growth in early to mid-exponential phase for all strains for in total approximately 5 generations. The strains were harvested by Ĵȱȱȱȱȱȱ ȱȱȱȱȱ¢ȱȱ 0.025 (F0-value), after which they were placed on the gradient table. ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ Ěȱȱ ȱ ȱǯȱȱǰȱȱěȱȱ ȱȱȱ each temperature series were placed back in random order at their appropriate temperature on the table. The temperature of each well was measured using a calibrated precision thermometer (Ebro) at the end of each experiment, and average temperature was calculated ȱȱȱȱȱęȱ¢ȱȱȱȱȱ experiments per assessed temperature. To calculate the growth rate for each temperature per strain, F0-values were log 2 transformed, ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ĵȱȱ ȱ ȱ ǻȱ hour), and the slope of the exponential growth curve was estimated ȱȱȱȱ¡ȱŘŖŖŝȱǻȱĜȱȱǼǯ

6.3 Results Taxon sampling In total 52 strains of P. borealis were studied, including the morphological varieties borealis, and (Table 6.1). For ĵȱ, from the 31 strains only 11 strains could be

162 Genetic divergence in P. borealis and H. amphioxys morphologically assigned to ĵȱȱ¡¢ (stria density of ŘŖȬŘşȦŗŖȱΐǼǰȱ ȱȱȱȱȱȱ ȱęȱȱȱH. abundansȱǻȱ¢ȱȱŗśȬŘŖȦŗŖȱΐǼȱǻȱŜǯŘǼǯȱȱȱ ȱȱȱęȱȱȱȱ ȱȱǰȱȱ ȱ reported as ĵȱȱǯŗ, ǯȱŘ and ǯȱř or ǯȱǯȱ¡¢.

Molecular-phylogenetic analyses of P. borealis Analyses using ML and BI gave identical results based on the individual genes and concatenated dataset. Therefore only the ML phylogeny of the concatenated analysis is shown in Fig. 6.1. ȱ ȱ ¢ȱ ǻěȱȱ et al.ǰȱ ĴȱǼǰȱ P. borealis was recovered as monophyletic. Within the P. borealis lineage, eight highly supported lineages can be clearly delineated (indicated by grey boxes in Fig. 6.1). These same lineages were all resolved with high support in the individual gene phylogenies, enhancing the ¢ȱȱȱȱȱȱěȱȱǯȱȱ ȱȱȱȱȱȱěȱȱȱǽȱ + French Alps; Czech Republic + Canada (1); Belgium + Canada (2)]. There appears to be some (phylo)geographic signal within these ȱȱȱȱěȱȱȱǻȱvs. Europe, and Belgium vs.ȱȱǼȱȱěȱȱǯȱȱȱȱ ȱ ȱ ȱ ǻȱ ƽȱ ǰȱ Ǽȱ ȱ ěȱȱ ȱ ȱ detected and a single sediment sample (St6) contained two lineages, showing that at least some of the lineages occur sympatrically. Only a single morphological variety, var. , is restricted to a single ȱǻǯȱŜǯŗǼǰȱȱ ȱȱȱȱĴȱȱȱ ȱ ¢¢ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ęȱȱ have no evolutionary meaning. Based on our taxon sampling, the P. borealisȱȱ¡ȱȱȱ¢ȱŘŘȱȱȱǽŗśǯśȬŘŝǯŘȱ ȱşśƖȱȱȱȱǻ ǼǾȱǻǯȱŜǯŘǼǯȱȱěȱȱ lineages diverged on average between 11.9 and 5.0 Ma ago, with a minimum and maximum 95% HPD of 0.45 and 17.91 Ma over all lineages.

Molecular-phylogenetic analyses of H. amphioxys Again, we only show the ML phylogeny based on the concatenated dataset since both ML and BI and the two individual genes gave

Genetic divergence in P. borealis and H. amphioxys 163 Fig. 6.1ȱ¡ȱȱ¢¢ȱȱȱrbcȱȱȱŗȬŘȱȱȱŘŞȱȱ for P. borealisȱ ȱȱȱȱȱȬȱȱǯȱȱȱ ȱ¢ȱ¢ȱ¡ǰȱ ȱȱ¢ȱ¡ȱȱȱȱȱ¢ȱȱ ȱǯ

164 Genetic divergence in P. borealis and H. amphioxys identical results (Fig. 6.3). All ĵȱȱ ¡¢ strains were recovered as a monophyletic clade. In total 5 ǯȱ¡¢ lineages can be delineated (indicated by grey boxes in Fig. 6.3), of which 4 were isolated from a single site in Belgium (St = Gent, Sterre). Compared to the P. borealis phylogeny, the branch lengths between these lineages are much shorter. Outside the ǯȱ ¡¢ clade, ȱ ȱ ȱ ȱ ȱ ęȱ¢ȱ ȱ H. ¡¢ (H. ǯȱ ¡¢), two of which were placed in a large clade containing most other ĵȱ strains, another one as sister to the ǯȱ¡¢ clade (Fig. 6.3). Also for other strains, e.g. those ęȱȱ ȱ H. abundans or ǯȱ ǯȱ ŗǰȱ ȱ ęȱȱ ȱ ȱ morphology in LM often didn’t match the separation in lineages in the phylogeny.

Thermal adaptation in P. borealis Average growth rate per lineage versus temperature is shown in Fig. 6.4 The optimal growth temperature (i.e. the temperature at which the highest growth rate is observed) of the Antarctic lineage lays around 15°C, while for most other lineages this is situated around 20-24°C. The Antarctic lineage is thus clearly adapted to colder temperatures. Note however that the lineage containing the Mongolian strains ȱȱȱ ȱȱęȱȱȱȱȱǯȱȱȱ upper lethal temperature is situated around 24°C for the Antarctic and Mongolian strains, while this lies between 25 and 29°C for the other lineages, indicating that the Antarctic strains have lost the capacity to survive at higher temperatures. The lineage Canada(1)- Czech Republic and the lineage from the French Alps are intermediate concerning their upper lethal temperature. Meanwhile, because the strains of the French Alps underwent cell size enlargement in culture (the mechanism was not observed), these results should be interpreted with caution. Despite this, the two colder-adapted ȱǻȱȱǼȱȱȱȱęȱȱȱȱ of P. borealis (Fig. 6.1, upper clade), while the higher temperature tolerant lineages all form a second subclade (Fig. 6.1, lower clade).

Thermal adaptation in H. amphioxys Similar to the P. borealis species complex, the optimal and upper lethal temperatures of the Antarctic lineage are lower compared to the other lineages (Fig. 6.5), but to a lesser extent than in the P. borealis species complex (Fig. 6.4). The optimal temperature lays

Genetic divergence in P. borealis and H. amphioxys 165 Fig. 6.2ȱȱȱȱ¢¢ȱȱP. borealis based on rbcȱȱȱŗȬŘȱ ȱȱŘŞȱǰȱ ȱȱȱȱȱȱȱȱȱȱȱ ¢ǯȱȱȱ¢ȱȱȱȱȱȱȱȱşśƖȱ ȱȱȱ ȱȬȱ¢¢ȱȱȱȱPinnulariaȱȱěȱȱet al. ǻĴȱǼȱȱ ȱȱǯȱ ¢ȱȱȱȱşśƖȱ ȱǯȱȱȱȱȱ ȱȱȱȱȱ ȱǯ

166 Genetic divergence in P. borealis and H. amphioxys around 20°C for the Antarctic lineage, and around 23°C for the other lineages. The temperature preference of the single ǯȱǯȱ¡¢ strain (St3)a sister to the ǯȱ¡¢ clade was similar to that of the temperate ǯȱ¡¢ strains.

6.4 Discussion Based on the molecular phylogenies, P. borealis and ǯȱ ¡¢ consist of multiple genetically diverged lineages, including a distinct Antarctic lineage. Based on a molecular clock estimate, the species complex P. borealis originated between 30 and 47 Ma ago, while the Antarctic lineage of P. borealis is estimated to have diverged around 7.7 Ma ago from its sister lineage. In addition, the Antarctic lineages of both P. borealis and ǯȱ¡¢ diverged in optimal temperature and in upper lethal temperature from the lineages from more temperate regions. Taken together, despite our doubtlessly non- exhaustive taxon sampling, this indicates that the high percentage of endemics found on the Antarctic continent as reported by Verleyen et al.ȱǻĴȱǼȱȱȱȱȱȱȱȱȱ reality of the other presumed cosmopolitan species in the Antarctic. (Pseudo)cryptic speciation has been reported in a variety of diatom genera (Sarno et al., 2005; Mann & Evans, 2007; Vanormelingen et al., 2008; Trobajo et al., 2009; Vanelslander et al., 2009) and we gathered genetic indications that Pinnularia borealis and ĵȱȱ ¡¢ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ Ĵȱȱ ȱ ȱ molecular phylogenies is composed of groups of strains with few nucleotide substitutions (i.e. lineages) compared to the large number of substitutions separating them, which is typical for the species level (e.g. Jargeat et al., 2010; Tronholm et al., 2010). The fact that identical ȱ ȱȱ¢ȱ ȱěȱǰȱȱȱǻin L and 28S rDNA), even when it concerned sympatric strains, ȱȱǯȱ ȱȱȱ¡ǰȱȱěȱȱ lineages have no clearly distinct morphologies, and only correspond ¢ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ¢ǯȱ ȱ doubt whether further morphological examination could reveal ǻȱ ȱ Ǽȱ ȱ ȱ ȱ ęȱ¢ȱ ¢ȱ ȱ ěȱȱ ǯȱ ȱ ǻ ¢ȱ ǭȱ ¢ǰȱ ŘŖŗŗǼȱ ȱ Śśȱ varieties of P. borealis, while the 7 varieties mentioned in Krammer (2000) already overlap in described morphological characteristics. Concerning ǯȱ¡¢, 25 varieties are mentioned in AlgaeBase (Guiry & Guiry, 2011), and Cleve-Euler (1952) characterized ǯ 30

Genetic divergence in P. borealis and H. amphioxys 167 Fig. 6.3ȱ¡ȱȱ¢¢ȱȱȱrbcȱȱȱŗȬŘȱȱȱŘŞȱȱ for H. amphioxysȱ ȱȱȱȱȱȬȱȱǯȱȱȱ ȱ¢ȱ¢ȱ¡ǯ

168 Genetic divergence in P. borealis and H. amphioxys ęȱȱ¡ȱ ȱȱ ǯȱ¡¢, while Krammer & Lange-Bertalot (1988) recognized the extreme morphological diversity in ǯȱ¡¢ but estimated the disentanglement of this complex beyond reach. On the other hand, advanced morphometrics has been useful for separating pseudocryptic species in a few of the more diverse diatom species complexes (Droop et al., 2000; Mann et al.ǰȱ ŘŖŖŚǼǯȱ ȱ ǰȱ ȱ ęȱȱ ȱ ȱ ȱ ¢ȱȱ¢ȱ ȱ ¢ȱȱĜȱȱȱȱǯȱ DNA barcoding, a diagnostic technique in which short characteristic ȱǻǼȱȱȱȱȱęȱȱǻ ȱet al., 2003; Savolainen et al.ǰȱ ŘŖŖśǼǰȱ ȱ ěȱȱ ȱ ȱ ǻȱ et al., 2010). The genetic divergence between the Antarctic and other P. borealis and ǯȱ ¡¢ lineages indicates that dispersal is limited enough to allow lineage divergence or speciation on the Antarctic continent, even for such abundant terrestrial diatoms. While the sampling of strains from the non-Antarctic locations ȱ ȱ ¡ǰȱ ȱ ¢ȱ ȱ ¢ȱ ěȱȱ lineages in P. borealisȱȱȱěȱȱȱȱȱǰȱ ȱ ȱ ǯȱ ¢ȱ ȱ ȱ Ěȱ ȱ ȱ ȱ suggested by the observation that strains within the same lineage ȱȱěȱȱȱȱȱȱȱȱ mutations, indicating that there is phylogeographic structure within lineages between and perhaps even within continents. However, this should be investigated in more detail by increasing the number of sampled locations and by increasing the number of individual strains sampled per population. Furthermore, a high number of sequenced individuals per meta-population (i.e. per species) is ¢ȱȱȱȱȱ¢ȱ ȱȱęȱȱǯȱ That way, algorithmic sequence-based species delimitation which determines the boundary between species and populations can be performed (Pons et al., 2006; Monaghan et al., 2009). On a chronogram, the transition from the species-level processes (speciation and extinction) to population-level processes (coalescence) is association with a sudden increase in branching rate (i.e. a sudden increase in ȱȱěȱȱ¢Ǽǯ Compared with freshwater habitats, soils are characterized by a higher spatial availability and continuity, while the dryer sediment is more susceptible to be picked up by air currents (Chepil, 1956). Combined with the relatively high tolerance of soil-inhabiting diatoms, including P. borealis and ǯȱ¡¢, compared to their

Genetic divergence in P. borealis and H. amphioxys 169 Fig. 6.4ȱ Ȭȱ ȱ ȱ ȱ P. borealisǯȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱȱȱȱŘŘȱP. borealisȱǰȱȱȱ ȱ ȱȱǯ

Fig. 6.5ȱ Ȭȱ ȱ ȱ ȱ H. amphioxysǯȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱȱȱȱŗŗȱH. amphioxys ǰȱ ȱȱȱȱ ȱȱǯ

170 Genetic divergence in P. borealis and H. amphioxys ȱȱȱȱ¡ȱǻěȱȱet al., 2010) ȱȱȱǻěȱȱet al., unpubl. b), this will enhance the dispersal potential and thus colonization chances of these diatoms. Living cells of ĵȱȱ¡¢ have been encountered ȱȱȱȱ ȱȱǻ ȱǭȱ ěȱǰȱŗşŜŜǼȱȱȱ a colonization experiment on the roof of a university building (C. ěȱǰȱǯȱǼǯȱĴȱȱȱ ȱȱǻ Ǽȱȱȱ ȱ ěȱȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ encountered at low heights, close to the soil (Van Overeem, 1937; Schlichting, 1961; Brown et al., 1964), suggesting that short distance ȱȱȱ ȱȱȱęȱ¢ȱȱ ȱȬȱ dispersal might be associated with much higher mortalities. In any case, if any diatom species had a high chance of being cosmopolitan, it would have been P. borealis and ǯȱ¡¢, making it likely that many of the other so-called cosmopolitan species on (sub-)Antarctica are also species complexes with a distinct (sub-)Antarctic lineage. The Antarctic continent is highly isolated, and we can realistically assume that the surrounding oceans are a relatively strong dispersal barrier for diatom cells. The age of the species complex P. borealis, being between 30 and 47 Ma, has allowed it to colonize all continents, including the Antarctic. Based on our taxon sampling the current Antarctic lineage of P. borealis present in the Schirmacher Oasis is estimated to have diverged between 15 and 2 Ma ago from its sister lineage, which is fairly recent compared to the ȱȱȱȱ¡ǰȱȱȱ¢ȱȱȱȱęȱȱ opening of the Drake Passage between the Antarctic Peninsula and South America around 14 Ma (Lagabrielle et al., 2009) and after the start of the Mid-Miocene cooling event of the Antarctic continent, which resulted in the increase of the East Antarctic ice sheet and ȱȱȱȱȱȱȱǻ ȱǭȱ Ĵȱǰȱ 1994). This corresponds to divergent dates of Antarctic mites and springtails from the sub-Antarctic lineages (Stevens et al., 2006; Mortimer et al., 2011). These time estimates indicate the establishment of P. borealis on the geographically isolated continent through long- distance dispersal, and subsequent lineage divergence. It is yet unclear whether any P. borealis inhabiting the continent before the Mid-Miocene cooling went extinct, or if there still is a second more ancient P. borealis lineage on the continent, which started diverging after opening of the Drake passage but is currently regionally absent from the Schirmacher Oasis. Alternatively, for P. borealis, the upper limit of divergence for the Antarctic lineage (15 Ma) is correct and

Genetic divergence in P. borealis and H. amphioxys 171 ȱȱȱȱȱȱȱȱȱ ȱȱęȱȱ opening of the Drake passage or the Mid Miocene cooling event triggering allopatric speciation. The growth rate experiments show that the Antarctic lineages of P. borealis and ǯȱ¡¢ have a relatively low optimal growth temperature and low upper temperature limit compared to lineages from more temperate areas. However, they are not psychrophilic, in contrast to a range of marine planktonic diatoms which die at temperatures no higher than 12-15°C (Bunt et al., 1966; Van Baalen & Odonnell, 1983; Fiala & Oriol, 1990; Thomas et al., 1992; Smith et al., 1994). This might be related to the fact that their habitat is less ěȱȱ ȱ ȱ ǯȱ ȱ ȱ Ȭ¡ȱ shallow water and wet soils or seepages, temperatures can be as high as 15°C even on continental Antarctica. In the Antarctic H. ¡¢, there is a modest shift in temperature preference with the most closely related temperate lineages, implying the loss of the ability to grow well at temperatures well above 20°C. For Antarctic P. borealis, the most closely related lineages from the French Alps and the Mongolian Khangai mountain range have quite similar ȱęȱǯȱȱȱ ȱ ȱȱȱ ȱ above 20°C group into another lineage suggesting that there might be some phylogenetic signal in temperature preference in P. borealis. Species from one clade might be adapted to warmer climates, while species from the other clade, including the Antarctic lineage, might inhabit colder climates. Thermal adaptation related to climatic zones ȱ ȱ ȱ ȱ ȱ ȱ Ěȱȱ (Boenigk et al., 2007) and the marine diatom ¢ (Vanelslander et al., unpubl.), and our results contribute to the idea that protist species ȱȱȱȱȱȱȱęȱȱȱ zones. Here, we have shown for two cosmopolitan diatom ȱȱ¢ȱȱȱěȱȱȱ ȱȱȱ continental Antarctic lineage which has diverged in its temperature preference from more temperate counterparts. However, many ȱ ȱ ȱ ȱ ęȱȱ ȱ ¢ǯȱ Further research should therefore focus on assessing the geographic distribution of P. borealis and ǯȱ ¡¢ lineages, especially in the (sub-)Antarctic and Arctic, to further unravel the evolutionary history of (polar) diatom lineages. More complete molecular phylogenies through more extensive taxon sampling, in combination with niche modelling, should also allow determining phylogenetic

172 Genetic divergence in P. borealis and H. amphioxys signals in temperature preference. Shorter-term regional extinction- colonization dynamics due to climate change in the Antarctic should be investigated using phylogeographic markers such as SNPs, or with ǯȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ĴȱȱȱȱĚȱ ȱȱȱȱȱȱȱȱ rates and distances. However, the correct interpretation of such studies requires a decent knowledge on species boundaries, and by doing so, this study lays a basis for future molecular research on the biogeography of polar diatoms.

Acknowledgments ȱ ȱ ȱ Ç²¤ȱ ȱ ȱ ȱ ȱ Pinnularia borealis from Czech Republic. We are grateful to Tine Verstraete for performing the DNA extractions and PCR reactions of the Antarctic strains, and Andy Vierstraete for the sequencing. This research was ęȱ¢ȱȱ¢ȱȱȱȱęȱȱȱȮȱȱ (FWO-Flanders, Belgium, funding of CS, PV and EV), the FWO- project 3G-0533-07 on taxonomic turnover in terrestrial and aquatic diatom communities, and the FWO-project 3G-0419-08 on thermal ǯȱȱȱȱȱ ȱęȱ¢ȱȱ¢ȱ the InBev-Baillet-Latour Fund project DELAQUA and the BelSPO Project Beldiva.

Genetic divergence in P. borealis and H. amphioxys 173

Part 3. Tolerance levels for dispersal- related stresses

Part 3: Tolerance levels for dispersal-related stresses 175

7. Tolerance of benthic diatoms from temperate aquatic and terrestrial habitats to experimental desiccation and temperature stress

ȱ ěȱǰȱ ȱ ǰȱ ȱ ¢ǰȱ ȱ ȱȱȱ¢

ǐȱȱ¢ȱȱ¢ȱȱȱ¢ǰȱ ȱ ¢ǰȱ ȱŘŞŗȱȬȱŞǰȱȬşŖŖŖȱ ǰȱ

ęȱȱȱȱȱȱPhycologiaȱǻŘŖŗŖǼȱŚşǱȱřŖşȬřŘŚ

Stress tolerance of vegetative diatom cells 177 Abstract ȱěȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ Ěȱǰȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱǯȱȱȱȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ řŚȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ǯȱ ¡ȱěȱȱȱȱ ȱǱȱȱȱȱȱ ƸřŖǚȱ ȱ ƸŚŖǚǰȱ ȱ ȱ ȱ ƸŚŖǚǰȱ £ȱ ȱ ȬŘŖǚǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ƸřŖǚǯȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ęȱ¢ȱ ȱ ȱ ȱ ȱ ȱȱȱǯȱȱȱȱȱǰȱ £ȱ ȱȱ¢ȱ¢ȱȱǰȱȱȱȱ ȱ ęȱ¢ȱȱȱȱȱǰȱȱȱ¢ȱ ȱ¢ȱȱȱȱȱȱȱȱȱǯȱ ȱȱȱȱ¡ȱǻƸŚŖǚȱȱȬŘŖǚǼȱ ȱȱȱ ȱ¡ȱȬęȱǰȱȬęȱȱěȱȱȱȱ ȱ ȱ ǯȱ ¢ȱ ȱ ȱ ȱ £ȱ ȱ ȱȱ ȱȱȱȱȱȱȱƸŚŖǚǰȱȱ ȱȱȱȱȱȱȱȱȱȱ ¡ǯȱǰȱȱ ȱȱ ȱȱȱȱȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ƸŚŖǚǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻŗǼȱ ȱ ¢ȱ ȱȱ ǰȱ £ȱ ȱ ȱ ǰȱ ȱ ǻŘǼȱ ȱ ȱ Ȭ ȱȱȱȱ¡ǯȱȱȱȱ ȱȱȱȱȱȱȱȱȱ ȱĴȱȱȱȱȱǯ

Key wordsǱȱǰȱȱǰȱǰȱȱ ¢ǰȱȱǰȱȱ¡ǰȱ

178 Stress tolerance of vegetative diatom cells 7.1 Introduction ȱ¢ȱȱȱȱȱęȱȱȱȱȱ ǻǰȱ ŗşŝŝǼǯȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱȱȱȱȱȱ ȱǻǰȱŗşŝŝǼǰȱ ȱ¢ȱ ȱȱ ȱ¢ȱ ȱ ȱ ȱ ǻĴȱȱ ǭȱ §ǰȱ ŗşşśǼǯȱ ȱ ȱ ȱȱ£ȱ¢ȱ£ȱȱǻĴȱȱǭȱ §ǰȱ ŗşşśǲȱ ǰȱ ŗşşşǲȱ ȱ et al.ǰȱ ŘŖŖŖǲȱ ȱȱ ȱ et al.ǰȱ ŘŖŖŚǼǯȱ ȱ ȱ ȱ ǰȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ ȱǭȱǰȱŘŖŖŚǲȱȱet alǯǰȱŘŖŖŞǼǯȱȱȱȱȱȱ ȱȱȱ¢ȱȱęȱȱȱȱȱ ȱȱǻȱet al.ǰȱŘŖŖŞǼǯȱȱȱǰȱ ȱǰȱȱ¢¢ȱ ȱȱȱĚȱȱȱ ǰȱ ȱȱȱ¢ȱȱȱȬȱ¢ȱ ǻȱet al.ǰȱŗşŞŗǲȱ ȱet al.ǰȱŘŖŖŞǼǯ ȱȱȱȱ¢ǰȱȱȱȱȱȱȱ ȱ ȱ ¡ȱ Ěȱȱ ȱ ȱ ǯȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȬȱȱȱȱ ȱ ȱ ǻȱ et al.ǰȱ ŗşşśǲȱ Ĵȱǰȱ ŗşşşǲȱ ȱ ǭȱ ǰȱ ŘŖŖŜǲȱ ¢ȱet al.ǰȱŘŖŖŝǼǰȱȱȱǻȱet al.ǰȱŘŖŖśǼǰȱ£ȱ ǻȱǭȱǰȱŘŖŖŜǼȱȱȱǻȱet al.ǰȱŘŖŖŘǼǯȱȱȱ ȱ¢ȱȱȱȱȱ¡ȱ ¢ȱ ¢ȱ ǻĴȱǰȱ ŗşşşǲȱ ȱ et al.ǰȱ ŘŖŖśǼǰȱ ȱ ȱ ȱǻǰȱŗşśŞǰȱŗşśşǼȱȱ¢ȱȱȱȱ ǻǰȱ ŗşśŞǰȱ ŗşśşǲȱ ǰȱ ŗşŞŞǲȱ ěȱǰȱ ŗşşŜǼǯȱ ȱ ȱ ȱ ǰȱȱȱȱȱȱȱ¡ȱ ȱ ȱ ȱ ǯȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ Ĵȱȱ ǭȱ ǰȱ ŗşŝŖǼǰȱ ȱ ȱ ȱ¡ȱȱȱȱȱ¢ȱȱȱ ȱȬęȱȱȱȱȱȱȱȱȱȱ ȱȱȱ ȱȱȱȱǻǰȱ ŗşśŞǰȱ ŗşśşǲȱ Ĵȱȱ ǭȱ ǰȱ ŗşŝŖǼǯȱ ȱ ȱ ȱ ȱȱȱǰȱȱȱȱȱ ȱƸřŖǚȱȱƸŚŖǚȱǻ£ȱǭȱǰȱŗşşśǲȱĴȱ ȱet al.ǰȱ ŘŖŖśǼǯȱ ȱȱȱȱȱȱȱ¢ȱ ǯȱȱȱȱȱȱȱȱ¢ȱȱȱ ȱȱȱȱ¢ȱ ȱȱȱȱȱ ȱ¢ȱȱǻǰȱŗşśŞǰȱŗşśşǼǯȱ¢ȱȱ

Stress tolerance of vegetative diatom cells 179 ȱȱȱ¢ȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱǽ ȱȱ ȱǭȱȱǻŗşŞřǼȱȱȱ ǭȱ ȱǻŗşşŜǼǾǰȱȱ¢ȱȱȱ¡ȱȱȱȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱ ŗşŝśǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱ¢ȱȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱȱǻ ǰȱŗşśřǰȱŗşŝŗǲȱǰȱ ŗşŝşǲȱȱȱǭȱǰȱŗşŝşǼǯȱ ¢ȱȱȱȱǰȱ Craticula cuspidataȱǻ ûĵȱ Ǽȱ ǯ ǯȱ ǰȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱ¢ȱǻǰȱŘŖŖşǼǯȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱȱȱȱȱȱȱȱ¢ȱȱȱ ȱȱȱȱȱȱȱȱ ¢ǯȱȱ ¢ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱǻȱǭȱ ǰȱŗşşŜǼǰȱȱȱȱȱȱȱ ȱȱǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱǰȱȬȱȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱ ŘŖŖŗǲȱ ȱ ǭȱ ǰȱŘŖŖŘǲȱȱet al.ǰȱŘŖŖŝǼǯȱȱȱȱȱ ǻ ȱ¢ȱ ǰȱŗşşŜǼǰȱ ȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻǰȱ ŗşŜŖǲȱ ǰȱŗşşŜǲȱȱǭȱ ǰȱŘŖŖŘǼǯȱȱȱȱĴȱȱ ȱȱ¢ȱȱ¢ȱȱǻȱǭȱ ǰȱŘŖŖŘǼȱȱ ȱȱȱȱȱȱ¡¢ȱȱȱȱȱ ȱȱȱȱȱǰȱȱȱȱ¢ǯȱ ȱȱȱȱ¢ȱȱȱȱȱȱ ȱȱǻȱǰȱŗşřŝǲȱǰȱŗşŜŗǲȱ ȱet al.ǰȱŗşŜŚǲȱ ǰȱŗşŜŚǲȱ¢Ȭȱǭȱǰȱŗşşřǲȱȱǭȱǰȱ ŗşşŝǼȱȱȱȱȱȱȱ¢ȱȱ ¢ȱ ȱ ȱ ǻ ȱ et al.ǰȱ ŗşŜŚǲȱ ǰȱ ŗşŜŚǲȱ ȱ ǭȱ ǰȱŗşşŝǼǯȱ ȱȱ Ȭȱȱ Ȭȱȱ ¢ǰȱ ȱ ȱ ȱ ¢ȱ ǯȱ ȱ ȱ ¢ȱȱ¡ȱ¢ȱȱ ȱȱȱȱȱ ȱȱ¢ȱȱȱȱǻ¢ǰȱŗşşŗǲȱȱet al.ǰȱŘŖŖŜǼǰȱȱȱȱȱȱȱȱ ȱȱȱȱȱ ȱȱȱȱȱȱǯȱ ǰȱěȱȱȱȱȱ ȱȱȱȱ ȱȱȱěȱȱȱǽe.g.ȱȱȱȱ

180 Stress tolerance of vegetative diatom cells ǰȱ ȱ ȱ ȱ ȱ ȱ ǻ ¢ȱ et al.ǰȱ ŘŖŖŝǼǾȱ ¢ȱ ȱ ȱ ěȱȱȱǰȱȱȱȱěȱȱȱ ȱȱȱȱȱȱȱǯ ȱĴȱȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ǯȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ¡ȱȱȱȱǰȱȱȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ǻȱ et al.ǰȱ ŘŖŖŞǼǯȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱȱȱȱȱȱȱȱǻ¢ȱet al.ǰȱŘŖŖŘǼǰȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱǻȱȱȱ Ǽȱȱ¢ȱĴȱȱǻȱet al.ǰȱŘŖŖŜǲȱ¢ et al.ǰȱ ŘŖŖŝǼǯȱȱȱȱȱ¢ȱȱȱȱȱȱȱ ǻȱet al.ǰȱŘŖŖŞǼǯȱȱȱȱȱȱȱ ȱȱȱĴȱȱȱȱȱ¢ȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱěȱȱȱȱ ȱȱȱȱȱ ȱȱȱȱȱȱĴȱǯ ȱȱ¡ȱ¢ǰȱ ȱȱȱ ȱȱȱ ȱȱȱȱ¢ȱȱȱȱ¡ȱ ǻŗǼȱȱ ȱ¡ȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ǰȱ ȱ ǻŘǼȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ĵȱȱ ȱȱ¡ǯ

7.2 Materials & Methods Sampling ȱ ȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ěȱȱȱȱǱȱȃȱȄȱǻǼȱȱ ǰȱȱ ȱȃȱȄȱǻǼȱȱȱǰȱȱȱȃȱȄȱ ǻǼȱ ȱ ǰȱ ȱ ȱ ȃȱ Ȅȱ ǻǼȱ ȱ ǰȱȱȱȃ Ȅȱǻ Ǽȱȱǯȱȱȱȱ ȱȱȱȱȱ ȱȱȱȱȱ¢ȱǲȱǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ǲȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ £ȱ ¢ȱ ȱ Ȭȱ ¢ȱ ȱ ǻ Ȭǰȱ ŘŖŖŗǼǯȱ ȱ ŝǯŗȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱ¢ȱ ǰȱȱǰȱȱȱ

Stress tolerance of vegetative diatom cells 181 ȱǯȱȱȬȱ ȱȱȱȱȱ ŘŗȱȱŘŖŖŝǲȱȱȱȱȱřȱȱŘŖŖŞǯȱ Cultures ȱȱȱŘŖȱȱȱȱǻŗśǼȱȱȱǻśǼȱ ȱȱȱ ȱ¢ȱȱȱȱȱȬȱǻ ȱǭȱ£ǰȱ ŗşŝŘǼȱȱȱȱȱŗŞǚȱƹȱŘǚǰȱŘśȬřŖȱΐȱȱȬŶȱȬŗȱ ȱ¢ȱȱȱŗŘǱŗŘȱȱǱȱǯȱȱȱ ȱ ȱǰȱȱȱȱ ȱȱ¢ȱĴȱȱȱ ȱȱȱȱǻǼȱȱȱȱǯȱȱ ȱȱȱȱȱěȱȱ£ȱȱȱȱȱȱ ǰȱȱȱĴȱȱȱȱȱȱǰȱ ȱa priori ȱǯȱȱǰȱȱȱ ȱȱ ȱȱ ȱǯȱȱ ȱȬȱȱȱȱ ȱ¢ȱ ȱȱ¡ȱǯȱȱȱ¢ǰȱȱ ȱ¡£ȱȱ¢ȱ¡ȱȱȱȱ¡țǯȱ ȱ ȱȱȱȱȱ¡ȱŘȱȱȱ ȱ ȱ ¡ȱ ȱ ǯȱ ȱ ǰȱ ȱ ȱ ȱ ¢ȱȱŗŖȱȱȱȱ ȱȱȱ ȱŗǯřŝȱ ǯȱ ęȱȱ ȱȱȱ ȱǭȱȬȱ ǻŗşŞŜǰȱŗşŞŞǰȱŗşşŗǰȱǼȱȱ ȱǻŘŖŖŖǼȱȱȱȱPinnularia ǯȱȱ ĵȱȱ¡¢ȱǻǼȱ ǰȱRhopalodia gibbaȱǻǼȱǯȱûȱȱ¢ȱȱ ǰȱȱ ȱȱȱ¢ȱǻ ǰȱȱ¢ǰȱȱ Ǽȱ ȱȱ ȱȱǯȱȱȱȱȱȱ ¢ȱ ȱ ǰȱ ȱ ȱ ȱ ¢ȱ ȱ ȱȱȱ¢ȱȱ ȱěȱȱȱ ǰȱǰȱ¢ȱȱȱȱǻe.g. ǰȱŗşşşǲȱȱet al.ǰȱŘŖŖŚǲȱȱet al.ǰȱŘŖŖŜǲȱȱ et al.ǰȱŘŖŖŞǼǰȱȱȱ ȱȱȱěȱȱ¢ȱ ǻȱŝǯŘǼȱȱȱȱȱǯȱȱ¡¢ȱȱ ȱ¢ǰȱ ǯȱȱȱȱȱȱȱȱ ȱȱȱȱȱ¢ȱȱ¢ȱǭȱȱ¢ȱ ȱȱȱȱǯȱ¡¢Ȭȱȱȱȱȱ ȱȱ ȱȱȱȱȱȱ¡ǯȱ ȱȱ¡£ȱȱȱȱěȱȱȱȱȱ ȱ¢ȱȱȱȱȱŝǯŗȬřśǯ

182 Stress tolerance of vegetative diatom cells Carex Carex 10% 75% OP 10% 5% OPPRVV none moss 5% moss 5% moss 5% &KDUD

arenaria PLFURELDOELR¿

Claytonia perfoliata Claytonia 3KUDJPLWHVDXVWUDOLV PLFURELDOELR¿ Vegetation type & cover PLFURELDOELR¿OPPRVV OP QHRUJDQLF silt sand sandy material sandy-silt moss 5% VDQGELR¿ Sediment type KXPXVULFKVDQG VDQG¿ VDQG\VLOWZLWKKXPXV PRLVW VDQG\VLOWZLWKKXPXV ) OPPRLVW VDQG\VLOWZLWKKXPXV PRLVW LQPHDGRZPRLVW VDQG\VLOWZLWKKXPXV Taxus moist Betulus &KDPDHF\SDUHV

Quercus aquatic 200 m² VODFNDTXDWLFPð Habitat description +LSSRSKDHDUKDPQRLGHV VVXUHEHWZHHQWZRFRQFUHWHVODEV ( EDUHVRLOXQGHU ZHWVDQGZLWKELR¿OPLQGXQHVODFN KXPXVULFKPRLVWVRLOLQGXQHVKUXE OOHG¿ EDUHVRLOXQGHU FRQFUHWHSDYHPHQWZLWKELR¿ JURRYHVEHWZHHQSDYHPHQWWLOHVPRLVW ZHWFDUWUDFNLQPHDGRZZLWKJUDVVHVDQG EDUHVRLOXQGHU sandy sediment of temporary puddle in dune sandy sediment of temporary VRLO¿ VDQG\VHGLPHQWRIVKDOORZODNHLQGXQHVODFN 1 T TT dry sand and moss on dune sand moss 100% T T T T T T dry sand and moss on dune sand moss 30% T T A A Type 02°34’57.3”E 02°34’37.5”E 02°34’43.4”E 03°42’45.2”E 02°34’26.7”E 03°43’04.6”E 02°34’65.3”E 03°43’06.2”E 03°42’50.2”E 03°42’52.8”E 03°42’55.2”E 03°42’58.4”E 02°34’59.2”E 51°05’27.2”N 51°05’27.2”N 51°01’44.5”N 51°01’44.5”N 51°05’49.9”N 51°05’49.9”N 51°01’40.3”N 51°01’40.3”N 51°05’43.9”N 51°05’43.9”N 51°05’33.7”N 51°05’33.7”N 51°01’26.9”N 51°04’96.8”N 51°01’33.3”N 51°01’30.8”N 51°01’28.0”N 51°01’38.6”N 51°05’28.6”N coordinates Geographical Wes1 Wes2 Wes3 Wes4 Wes5 Wes6 Sterre St8 Sterre St2 Sterre St4 Sterre St6 Sterre St1 Sterre St3 Sterre St5 Location Sample :HVWKRHN :HVWKRHN :HVWKRHN :HVWKRHN :HVWKRHN :HVWKRHN ȱȱȱȱȱǰȱȱȱȱǯ Table 7.1 ȱȱȱȱȱȱȱ ȱȱ ȱȱȱ ȱǯ Table 1

Stress tolerance of vegetative diatom cells 183 ; 10% ; total 75% VXEPHUVHG total 20% moss 100% PDFURSK\WHV

&DUH[/X]XOD Carex arenaria Carex Vegetation type & cover

PRVV 3KUDJPLWHV/X]XOD

&KDUD3KUDJPLWHV&HUDWRSK\OOXP sand sand sand moss 50% sand moss 30% (none in depression) sand) Sediment type RUJDQLFPDWHULDOOLWWOH none (small amount of sand + organic material sand + organic sand + organic material sand + organic OP VKSRQGDTXDWLFPð 5m² LQGXQHVODFN SODLQLQGXQHVODFN Habitat description FRYHUHGSODLQLQGXQHVODFN VHGLPHQWVDPSOHRIODNHDTXDWLF ZHWVDQGLQGXQHVODFNZLWKELR¿ ZHWVDPSOHRIZHWVDQGIURPPRVVFRYHUHG VHGLPHQWVDPSOHRI¿ moist sample of moss from moss-covered plain moss-covered moist sample of moss from VHGLPHQWVDPSOHRISXGGOHLQPHDGRZDTXDWLF PRLVWWLOOZHWVDPSOHRIKXPLGVDQGIURPPRVV 1 T T T T A A A Type 03°09’38.7”E 03°09’93.4”E 03°28’93.9”E 02°34’38.3”E 02°41’86.1”E 02°41’86.1”E 02°41’86.1”E 51°05’44.4”N 51°05’44.4”N 51°19’31.7”N 51°19’31.7”N 51°08’18.4”N 51°19’50.1”N 51°04’53.8”N 51°08’18.4”N 51°08’18.4”N coordinates Geographical Wes7 Ter YdeTer Yde1 Ter YdeTer Yde2 Ter YdeTer Yde3 Location Sample :HVWKRHN Fonteintjes Fon3 Fonteintjes Fon1 Kraenepoel Kra1

184 Stress tolerance of vegetative diatom cells 4 + 30°C + desic. desic. 10 min. 00 0 00 0 00 0 00 0 - 20°C 100 0 abrupt 79 – 86 72 – 87 0 0 0 + 40°C 2 – 7 0 0 0 0 65 – 70 60 – 74 0 0 0 39 – 48 0 0 0 0 83 – 91 65 – 69 64 – 72 0 0 0 + 40°C gradual 65 – 79 65 – 79 54 – 71 72 – 87 0 0 0 Minimal – maximal % of viable cells after treatment 98 – 99 94 – 98 89 – 93 80 – 90 91 – 95 94 – 98 98 – 99 9987 – 94 86 – 91 85 – 88 89 – 9398 – 99 91 – 97 85 – 90 98 98 – 9993 – 96 91 – 95 82 – 92 94 – 95 84 – 93 88 – 94 97 – 99 96 – 99 Control 30°C + / 3 μ m 10 # striae

3 ( μ m) Width

3 ( μ m) Length 35.5 ± 0.7 8.2 ± 0.4 18.0 ± 0.0 38.1 ± 0.8 6.0 ± 0.3 ± 0.9 24.1 35.0 ± 0.8 ± 0.2 7.5 21.2 ± 0.4 (St3)c 42.8 ± 0.6 6.4 ± 0.2 ± 1.1 24.7 (St1)e 42.9 ± 0.4 6.2 ± 0.1(St3)a 25.2 ± 0.9 36.8 ± 0.5 6.6 ± 0.5 ± 1.0 23.7 (St4)a 36.0 ± 0.6 6.0 ± 0.4 25.8 ± 0.4 71 – 76 74 – 82 73 – 77 6W E Origin (Yde2)a 36.0 ± 0.6 ± 0.2 7.3 20.4 ± 0.5 (Yde2)e ± 1.2 24.9 ± 0.3 7.8 13.4 ± 0.9 (Yde2)d 10.4 ± 0.3 3.0 ± 0.0 28.0 ± 1.2 (Wes3)f ± 1.2 28.7 ± 0.5 7.0 16.4 ± 1.1 :HV E .W]LQJ &]DUQHFNL 2 *UXQRZPRUSK %UpELVVRQ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ *UXQRZLQ9DQ+HXUFN *UXQRZ .UDPPHU *UXQRZ .UDPPHU

Caloneis molaris Caloneis molaris &\PEHOODVXEDHTXDOLV $FKQDQWKHVFRDUFWDWD +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V

&UDWLFXODFIKDORSKLOD $FKQDQWKLGLXPPLQXWLVVLPXP 1 T T T T T T T T T T T ȱǯȱȱ¢ǯ ȱȱȱȱȱȱƹȱȱǰȱƽŗŖǯ ȱȱȱȱƖȱȱȱȱȱȱǯȱȱȱȱȱȱȱȱȱƖǱȱȱŖƖȱǻȱǼǰȱŗȬŘŚƖȱǻŗŖƖȱ Invisible in LM. ȱ ȱȱȱȱȱǯ ȱȱȱęȱǯ ȱȱȱȱȱȱȱȱǯ Habitat ǼǰȱŘśȬŚşƖȱǻŘśƖȱǼǰȱśŖȬŝŚƖȱǻśŖƖȱǼȱȱŝśȬŗŖŖ ƖȱǻŝśƖȱǼǯȱǯȱȱǯȱ Table 7.2 ȱȱ ȱȱȱ¡ȱȱȱ ȱȱȱȱȱȱȱȱȱǯ Table ȱȱȱǰȱȱȱȱȱȱǯ 2 3 4 ś 6 ŝ 8

Stress tolerance of vegetative diatom cells 185 4 0 + 30°C + desic. 6 00 00 desic. 10 min. 6 6 00 0 00 0 00 0 0 – 1 - 20°C 11 – 64 0 – 5 6 – 25 0 0 0 abrupt 74 – 80 0 0 0 30 – 49 0 – 3 0 0 85 – 90 16 – 40 0 0 0 20 – 46 0 – 2 + 40°C 00000 00000 5 – 44 1 – 3 0 0 0 85 – 88 93 – 96 47 – 64 41 – 45 0 0 0 49 – 85 51 – 76 2 – 3931 – 43 19 – 54 0 10 – 18 0 0 0 + 40°C gradual Minimal – maximal % of viable cells after treatment 100100 100 100 100 99 – 100 98 – 99 96 – 97 82 – 87 77 – 81 75 – 80 97 – 98 92 – 93 89 – 91 88 – 92 98 – 99 98 – 99 9996 – 98 97 – 98 96 – 97 96 – 98 91 – 95 85 – 89 81 – 88 80 – 88 89 – 92 81 – 83 97 – 98 91 – 96 98 – 99 100 98 – 99 76 – 78 97 – 98 97 – 98 Control 30°C + 99 – 100 93 – 98 97 – 99 99 – 100 97 – 99 88 – 98 / 3 μ m 10 # striae

3 ( μ m) Width

3 ( μ m) Length 40.4 ± 0.9 6.2 ± 0.2 ± 0.9 24.6 56.1 ± 0.2 ± 0.6 7.1 20.4 ± 0.5 51.8 ± 0.2 ± 0.3 7.3 19.6 ± 0.9 (St1)f 48.3 ± 0.9 9.0 ± 0.4 21.6 ± 0.7 (St6)f 52.3 ± 0.4 ± 0.4 7.3 18.8 ± 1.0 (St8)c 46.8 ± 0.7 8.2 ± 0.3 20.3 ± 1.3 62 – 80 55 – 62 (St5)a 40.5 ± 0.5 5.6 ± 0.2 ± 0.0 24.0 (St6)e 52.3 ± 0.5 ± 0.4 7.2 ± 0.7 19.7 Origin (Wes2)f ± 0.1 8.7 3.9 ± 0.2 22.4 ± 1.9 (Wes5)c 22.6 ± 0.4 5.4 ± 0.2 15.0 ± 0.7 (Wes5)a 41.5 ± 1.5 ± 0.2 9.7 12.0 ± 0.0 (Wes3)e 32.3 ± 0.4 6.6 ± 0.2 13.0 ± 0.0 (Wes2)a 39.5 ± 0.7 6.0 ± 0.3 ± 0.7 24.0 73 – 79 66 – 72 43 – 72 66 – 68 0 0(Wes1)a 0 56.6 ± 0.8 ± 0.4 7.2 20.0 ± 0.0 (Wes3)a 52.6 ± 0.7 ± 0.4 7.3 20.0 ± 0.0 (Wes2)e 8.6 ± 0.3 3.8 ± 0.2 20.4 ± 0.9 :HV E :HV E :HV E (Wes3)d 45.0 ± 1.1 9.6 ± 0.2 11.3 ± 0.4 6FKRHPDQ /DQJH%HUWDORW .XW]LQJ .W]LQJ 2 (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ (KUHQEHUJ *UXQRZ Taxon PRUSK PRUSK PRUSK PRUSK PRUSK PRUSK PRUSK PRUSK PRUSK PRUSK PRUSK .W]LQJ /DQJH%HUWDORWYDU .W]LQJ /DQJH%HUWDORWYDU +XVWHGW /DQJH%HUWDORW +XVWHGW /DQJH%HUWDORW

Navicula radiosa Navicula radiosa Navicula

1DYLFXODOLERQHQVLV

permitis permitis

Navicula cryptotenella Navicula

+DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V +DQW]VFKLDDPSKLR[\V

0D\DPDHDDWRPXV 0D\DPDHDDWRPXV 1 T T T T T T T T T T T T T T T T T Habitat

186 Stress tolerance of vegetative diatom cells 6 6 4 0 + 30°C + desic. 6 desic. 10 min. 00 0 00 0 00 0 00 0 00 0 00 0 3 – 12 0 0 - 20°C abrupt 79 – 85 68 – 82 15 – 2167 – 71 22 – 47 0 0 0 – 10 0 62 – 70 0 0 0 78 – 83 86 – 88 76 – 80 44 – 60 0 0 0 + 40°C 00000 0 0 0 0 0 – 84 53 – 66 5 – 32 0 0 0 63 – 94 31 – 59 6 – 24 0 0 0 62 – 70 62 – 70 2 – 5 0 0 71 – 82 15 – 38 3 – 16 0 0 0 21 – 33 34 – 61 0 0 0 + 40°C gradual 100 80 – 86 89 – 91 62 – 81 1 – 3 0 – 1 6 0 0 0 77 – 82 71 – 85 69 – 88 Minimal – maximal % of viable cells after treatment 94 – 98 91 – 98 93 – 97 96 – 81 – 85 85 – 8880 – 85 91 – 93 85 – 92 86 – 90 81 – 87 87 – 90 84 – 88 81 – 82 97 – 98 98 – 99 89 – 93 90 – 92 81 – 96 95 – 98 96 – 99 91 – 95 92 – 94 79 – 80 90 – 92 88 – 89 85 – 94 89 – 91 97 – 98 98 – 99 94 – 98 98 – 99 96 – 98 96 – 97 95 – 97 87 – 92 90 – 92 Control 30°C + / 3 μ m 10 LQYLVLEOH # striae

3 ( μ m) Width

3 ( μ m) Length 23.0 ± 0.7 3.6 ± 0.2 ± 0.7 14.1 33.7 ± 0.833.7 ± 0.3 7.6 6.0 ± 0.0 46 – 55 26 – 28 1 – 20 2 – 8 0 0 – 10 39.6 ± 1.2 ± 0.3 7.6 6.2 ± 0.4 65 – 82 74 – 77 48 – 63 63 – 73 14 – 23 0 0 16.2 ± 0.4 ± 0.1 4.9 13.0 ± 0.7 43.0 ± 1.6 9.8 ± 0.2 ± 0.4 11.7 (St1)i 35.9 ± 0.6 8.2 ± 0.3 5.8 ± 0.4 45 – 52 20 – 37 1 – 8 1 – 12 0 0 0 (St6)c ± 0.8 37.1 ± 0.5 8.7 6.3 ± 0.4 67 – 75 (St5)c ± 0.6 40.7 ± 0.5 7.9 6.0 ± 0.0 65 – 84 56 – 68 64 – 74 47 – 65 0 0 0 6W N (St2)a 29.0 ± 0.6 5.4 ± 0.3 ± 0.5 14.5 (St8)a 38.6 ± 1.1 ± 0.3 7.4 6.2 ± 0.4 (St6)a ± 0.7 34.6 ± 0.4 7.2 ± 0.5 6.7 (St8)e ± 0.7 32.7 2.9 ± 0.2 ± 0.5 29.7 74 – 85 64 – 74 62 – 86 66 – 80 0 0 0 (St2)d ± 1.7 27.1 ± 0.4 7.1 12.8 ± 0.4 6W E (St6)g 28.5 ± 1.4 6.9 ± 0.3 6.6 ± 0.6 (St8)d 42.4 ± 0.6(St1)d 8.9 ± 0.4 40.9 ± 0.6 5.0 ± 0.0 8.1 ± 0.3 6.0 ± 0.2 76 – 86 56 – 75 45 – 48 22 – 31 33 – 47 0 0 0 Origin (Yde3)f 22.2 ± 0.2 22.2 ± 0.2 15.6 ± 0.5 (Yde1)f 65.8 ± 0.7 10.9 ± 0.3 10.3 ± 0.3 (Yde1)c 18.0 ± 0.5 2.9 ± 0.1 28.0 ± 0.7 64 – 73 73 – 78 72 – 78 (Yde1)a 62.3 ± 2.0 11.4 ± 0.4 10.1 ± 0.2 (Yde1)e 42.2 ± 0.8 8.4 ± 0.5 ± 0.5 7 7.6 (Fon1)d 16.4 ± 1.7 3.3 ± 0.2 20.2 ± 0.4 (Yde1)d 41.4 ± 0.7 8.1 ± 0.6 7 ± 0.7 7.5 (Wes2)c 18.0 ± 0.4 ± 0.3 4.5 13.2 ± 0.8 46 – 63 (Wes7)e ± 0.4 44.0 ± 0.5 4.3 :HV E Krammer var. var. Krammer var. Krammer .UDVVNH *UXQRZ &OHYH 6W E Hustedt Krammer (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ (KUHQEHUJ .W]LQJ .W]LQJ 2 /DQJH%HUWDORW

nonfasciata nonfasciata

Navicula veneta Navicula Navicula radiosa Navicula 3LQQXODULDREVFXUD 1LW]VFKLDDXVWULDFD Pinnularia isselana 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV 3LQQXODULDERUHDOLV

1LW]VFKLDSDOHD Navicula 3LQQXODULDPDUFKLFD $PSKRUDSHGLFXOXV 3LQQXODULDVXEFRPPXWDWD 3LQQXODULDVXEFRPPXWDWD Pinnularia divergentissima 1LW]VFKLDSHUPLQXWXP 5KRSDORGLDJLEED 5KRSDORGLDJLEED 1 T T T T T T T T T T T T T T T T T T T T T T T T A Habitat

Stress tolerance of vegetative diatom cells 187 4 0 + 30°C + desic. 6 desic. 10 min. 00 0 00 0 - 20°C abrupt 46 – 67 0 0 0 58 – 65 0 0 0 59 – 80 0 0 0 + 40°C 00000 00000 00000 000000 0 0 0 – 74 00000 00000 81 – 90 + 40°C gradual 65 – 74 45 – 66 40 – 61 30 – 45 0 0 0 Minimal – maximal % of viable cells after treatment 99 99 98 97 – 85 – 91 86 – 91 88 – 91 80 – 90 96 – 97 89 – 97 87 – 90 96 – 98 90 – 95 86 – 95 88 – 91 84 – 91 84 – 90 86 – 92 92 – 96 85 – 91 91 – 93 85 – 91 77 – 84 79 – 82 90 – 92 85 – 93 94 – 95 92 – 98 Control 30°C + 99 – 100 98 / 3 μ m 10 # striae

3 ( μ m) Width

3 ( μ m) Length 26.7 ± 0.626.7 ± 0.2 4.4 ± 0.9 14.4 52.0 ± 1.6 9.1 ± 0.4 8.6 ± 0.5 7 Origin (Fon3)i 56.3 ± 1.4 ± 0.3 4.5 16.0 ± 1.6 (Kra1)c 25.5 ± 0.9 ± 0.3 4.2 15.0 ± 1.0 (Fon3)f 66.1 ± 1.0 10.3 ± 0.3 10.9 ± 0.7 (Kra1)a 50.3 ± 0.9 10.3 ± 0.1 11.0 ± 0.0 .UD E (Fon1)a 53.9 ± 0.9 9.1 ± 0.3 8.8 ± 0.8 7 71 – 77 69 – 80 0 0 0 0 0 (Fon3)e ± 1.4 64.3 10.3 ± 0.4 11.0 ± 0.6 (Fon3)g 56.3 ± 1.6 ± 0.5 4.3 15.4 ± 0.9 70 – 75 28 – 42 49 – 63 15 – 40 0 0 0 (Wes6)c ± 0.9 24.3 6.4 ± 0.2 ± 0.4 13.7 (Wes4)a 68.9 ± 0.5 10.9 ± 0.2 11.1 ± 0.2 (Wes4)e 32.1 ± 0.1 6.6 ± 0.1 13.0 ± 0.0 (Wes6)d 21.5 ± 0.4 ± 0.1 5.7 16.0 ± 0.0 (Wes4)d ± 0.7 44.4 9.5 ± 0.1 11.8 ± 0.4 Krammer (Fon3)a 86.4 ± 2.1 18.6 ± 0.7 8.9 ± 0.2 .W]LQJ .W]LQJ .W]LQJ .W]LQJ .W]LQJ 6FKRHPDQ .W]LQJ 2 1|USHOHWDO 1|USHOHWDO *UXQRZPRUSK

radiosa radiosa

Taxon (KUHQEHUJ 6FKDDUVFKPLGW (KUHQEHUJ 6FKDDUVFKPLGW (KUHQEHUJ 20OOHUPRUSK (KUHQEHUJ 20OOHUPRUSK )RQ E

Navicula radiosa Navicula radiosa Navicula Navicula Navicula radiosa Navicula veneta Navicula

Eunotia implicata Eunotia implicata 1DYLFXODOLERQHQVLV viridiformis Pinnulara

&\PEHOODVXEDHTXDOLV

(XQRWLDELOXQDULV (XQRWLDELOXQDULV

5KRSDORGLDJLEED 5KRSDORGLDJLEED 1 A A A A A A A A A A A A A A A Habitat

188 Stress tolerance of vegetative diatom cells Experimental set-up ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǱȱǻŗǼȱȱǻȱ ȱǼǲȱǻŘǼȱȱȱ ȱƸřŖǚǲȱǻřǼȱȱȱȱƸŚŖǚǲȱǻŚǼȱȱȱȱƸŚŖǚǲȱ ǻśǼȱ £ȱ ȱ ȬŘŖǚǲȱ ǻŜǼȱ ȱ ȱ ŗŖȱ ǲȱ ȱ ǻŝǼȱ ȱ¢ȱȱȱȱƸřŖǚȱ ȱ¢ȱȱ ȱŗŖȱȱǻȱŝǯřǼǯȱȱȱřŖǚȱȱŚŖǚȱ ȱ ȱȱȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ǻ£ȱ ǭȱ ǰȱ ŗşşśǲȱ Ĵȱ ȱ et al.ǰȱ ŘŖŖśǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱǻe.g.ȱȱet al.ǰȱŗşşŞǲȱ¢¢ȱet al.ǰȱ ŘŖŖŗǲȱ ȱ et al.ǰȱ ŘŖŖřǲȱ ȱ et al.ǰȱ ŘŖŖŝǼǰȱ ȱ ȱ ǻȱŞǰȱȱet al.ǰȱŘŖŖŚǼǯȱȱȱȱ¢ȱȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ¢ȱȱǯ ȱ¡ȱ ȱȱȱŘŚȬ ȱǯȱȱȱ ȱȱ ȱȱȱȱȱȱȱǰȱȱȱ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱȱȱȱȱȱȱǯȱ¢ȱ¢ȱ¡ȱȱȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱęȱȱ ȱŗǯśȱȱȱȱȱȱȱȱȱ ¡¢ȱ ȱȱ ȱȱȱȱ¡ȱ ȱ ȱ ¢ȱ ȱ ȱ ǯȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ǻȱ ȱ ȱ ŗŘŖȱ ȱ ŗŞśǰŖŖŖȱȱȱ Ǽȱȱȱȱȱ ¢ȱ ȱ ȱȱȱěȱȱȱ ȱǯ ȱ ŝǯřȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ¡ȱ ȱ ȱęȱȱ ȱǯ Temperature treatment Treatment Duration of dark condition (duration) control 18°C KK & K & KPLQ +30°C gradual KKPLQK & K & K & KPLQ +40°C gradual KKPLQKKPLQ & K & K &DEUXSW & KPLQ K & KPLQ & IUHH]LQJ& & KPLQ K & KPLQ & K & K & desiccation 10 minutes air-dry desiccation (10min) 10min & K & KPLQ +30°C + desiccation 10 minutes & K GHVLFFDWLRQ KKPLQKPLQ KPLQ

Stress tolerance of vegetative diatom cells 189 ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ŗŞǚȱ ȱ ŗśȱȱȱȱȱřȱȱ¡ȱȱȱȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ǰȱ ȱ ȱ ȱ ȱ ęȱȱ ȱȱ¢ȱǯȱȱȱ¡ȱȱȱ ȱěȱȱȱȱȱȱȱŝǯřǯȱȱȱ£ȱ ǰȱȱȱ ȱȱȱȱ£ȱȱȬŘŖǚȱȱŚŚśǯȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱĴȱǰȱȱ ȱȱ ȱ ȱ¢ȱȱ ȱȱȱȱȱ ȱęȱȱȱ ȱȱȱȱ ȱȱȱȱȱȱ ȱȱǯȱȱȱȱŗŖȱ ǰȱŗǯśȱȱȱȱȬȱ ȱǯȱȱ ȱȱ ȱȱ¢ȱȱȱȱȱȱȱ ƸřŖǚȱ ȱ ŗŞřŖȱ ǻȱŝǯřǼǯȱȱ ȱ ȱ ȱȱ ȱȱȱȱȬ¡ȱǯȱȱǰȱȱȱ ȱȱȱȱǯ ¢ȱ ȱ ȱ ǻŖǼȱ ȱ ȱ ȱ ŘŖŖȱ ȱ ȱ ȱ ȱȱȱȱȱ¡ȱȱȱȱ ȱǻȱȱȱȱŶǼȱ ȱǯȱȱŖȱȱȱ ȱ ȱǯȱȱŗŚȱ¢ȱǻŗŚǼȱȱȱȱ ȱ ȱ ȱ Ǽȱ ȱ ȱ ǻȱ ȱ ȱ ȱ ȱȱ¢ȱǼȱȱǼȱȱȱǻȱȱȱǼǯȱ ǰȱ ȱȱȱȱȱŘŖŖȱȱȱ ȱȱȱŗŚȱȱ ǯȱ ǰȱȱȱȱȱȱȱ ȱȱ ȱȱǰȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱȱȱȱȱȱȱ¢ȱȱȱ ȱ ȱŖȱȱ ŗŚȱȱȃ Ȅȱȱȱȱȱȱǯȱȱȱ ȱȱȃȄȱȱȱȱȱȱȱ ȱȱȱȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȃȄȱȱȱȱȱȱȱȱȱȱȱ ȱȱ ǯȱȱȱȱȱȱǻȱƖǼȱȱ ŗŖŖȱ¡ȱǽŗȱȬȱǻȱȱȱȱȱŶȱȱŗŚȱȦȱȱȱȱ ȱȱŶȱȱŖǼǾǯȱ ȱȱĴȱȱȱȱȱȱěȱȱȱ ěȱȱ ȱȱȱȃȱ¢Ȅȱǻi.e.ȱƖȱȱȱ ȱȱȱǼȱȱȱȱȱ¢ȱǻ ȱ ȱȱȱȱȱ¢ȱȱȱȱ¢Ǽǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȃȱ ƖȄȱ ¢ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ Ɩȱ ȱ ȱǯȱȱȱȱȱƖȱ ȱȱ ¢£ǯ

190 Stress tolerance of vegetative diatom cells Statistical analysis ȱ ȱ ǰȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱǱȱǻǼȱȱȱȱȱȱǻȦȱȱ ǼǰȱȱǻǼȱȱȱǰȱȱȱȱȱȱ ȱȱȱȱȱȱǻƖǼǯȱȱȱȱ¢ȱȱȱ ȱȱȱȱƖǯȱȱȱȱ ȱȱ¢ȱ ǰȱ ȱȱȬȱȱǯȱ ȱ¢ȱ ȱȱȱ ȱȱǻȱȱǼǯȱȱ ȱȱȱȱȱȱěȱȱȱ¢ȱ ȱȱ ȱěȱȱ ȱȱ ȱȱ ȱȱ ǯȱȱ ȱ¢ȱ ȱȱ ȱȱŝǯŖȱǻǰȱ ǯǰȱ ǰȱǰȱǼǰȱȱȱęȱȱȱ ȱȱȱpȱƽȱŖǯŖśǯ ǰȱ ȱȱȱȱȱȱ¡ȱȱȱ ȱȱ ȱȱȱȱ ȱȱƖȱȱȱěȱȱȱ ȱȱȱȱȱȱȬȱ¡ȱ ȱ ȱ ȱ ǻȱ ǭȱ ǰȱ ŘŖŖŖǼǯȱ ǰȱ ¡ȱ ȱȱȱ ȱȱȱȱȱěȱȱȱƖȱ ȱ ȱ ěȱȱ ȱ ǯȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Pinnularia borealisȱ ȱ ȱ ĵȱȱ ¡¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱǻȱȱȱǭȱ¢ǰȱŗşşŝǲȱȱ ȱet al.ǰȱŘŖŖŖǼǯȱ ǰȱ ȱ¢ȱȱȱȱȬȱ ȱ ȱȱȱƖȱȱȱǯȱȱ ȱ ȱȱȱȱȱȱȱȱȱȱPinnularia borealis and ĵȱȱ¡¢ȱ¢ǯ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻȦȱ ȱ Ǽȱ ȱ ȱ ěȱȱ ęȱ¢ȱ ȱȱ¢ǯȱȱȱǰȱȱ¢ȱ ȱ ȱȱȱȱȱȱȱȱȱȱ ¢ȱ ȱȱȱ¢£ȱ ȱȱȬȱȂȱȱ ȱȱǻȱǭȱǰȱŘŖŖŖǼȱȱȱȱěȱȱȱȱ¢ǯȱ ȱȱ ȱ ȱȱȱȱȱȱȱ ȱ¢ȱȬǯ ¢ǰȱȱȱȱ ȱȱ ȱȱȱƖȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ Ȭȱ Ȭ¢ȱȱȱǻȱǭȱǰȱŘŖŖŖǼǯ

Stress tolerance of vegetative diatom cells 191 7.3 Results Diatom strains ȱȱŜşȱȱȱřŚȱȱǻŘśȱȱȱşȱǼȱ ȱǰȱ ȱ ȱ ŗřȱ ȱ ǻȱ ŝǯŘȱ ȱ ȱ ŝǯŗȬřśǼǯȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱȱȱPinnularia borealisǰȱ ĵȱȱ¡¢ǰȱĵȱȱ austriacaȱ ǰȱ ǯȱ ȱ ǻ Ǽȱ ǯȱ ǰȱ ǯȱ ǻ ûĵȱǼȱǯȱǰȱȱ¢ȱȱǻ ûĵȱ Ǽȱ ȱǯȱ ȱ ǻ Ǽȱ Ȭȱ ǻ¢ȱ ȱ ȱ ȱ Ǽȱ ǻĴȱȱ ǭȱ §ǰȱ ŗşşśǼǯȱ ¢ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻǰȱ ŗşŝŝǲȱ ȱ et alǯǰȱ ŗşşŞǲȱ ȱ ȱ et al.ǰȱ ŘŖŖŖǲȱ ȱ ȱ ȱet al.ǰȱŘŖŖŚǼǯȱȱȱȱȱȱȱȱ ȱȱȱȱǻ ûĵȱ Ǽȱ ȱȱȱ ãȱet al.ȱȱȱęȱȱȱ ȱȱȱȱǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ûĵȱ ȱ ǻȱ ŝǯŗśȬŗŝǼǰȱ ǯȱ ȱ ȱ ǻȱ ŝǯŗŘȬŗřǼȱ ȱ ǯȱ ûĵȱ ȱǻȱŝǯşǰȱŝǯŗŗǼȱ ȱ¢ȱǰȱ ȱȱ ȱ ¢ȱ ȱ ǻȱ ŝǯřŗȬřŘǼȱ ȱ Rhopalodia gibbaȱǻȱ ŝǯŘŞȬŘşǼȱěȱȱȱȱ ǰȱȱȱȱȱęȱȱ¢ǰȱ ¢ȱǻȱŝǯŘǼǯ

Survival in control conditions ȱȱȱȱȱȱȱƖȱȱȱȱşŖƖȱȱȱȱ ǯȱ ǰȱ ȱ ȱ Ȯȱ ȱ ¡ȱ ȱ ȱ ȱ ĵȱȱ¡¢ǰȱ¡ȱȱȱP. borealisǰȱP. obscura and ĵȱȱ austriacaȱ ȱ Ɩȱ ȱ ȱ ȱ ŝŖƖǯȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ P. obscuraȱ ǰȱ ȱ ȱ P. borealisǰȱȱȱȱȱ ǯȱ¡¢ȱ¢ȱŗȱ ȱŘǰȱȱȱȱȱȱȱȱȱ ȱǯȱȱȱȱȱƖȱȱȱȱ ȱ ȱ ȱ ęȱ¢ȱ ěȱȱ ȱ ȱ ¢ȱ (pƽŖǯŝŞȱȱȱȱȱȱȱǼǯȱ

7ROHUDQFHWRKHDWLQJ ȱ ȱ ȱ ¢ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻȱ ŝǯŘǼǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱ¢ȱ ȱ Ɩȱ ȱ ȱ ȱ ȱ ȱ ǻpȱ ǀȱ ŖǯŖŖŗǼȱ ǻǯȱ ŝǯřŜǼǯȱ ȱ ȱȱěȱȱȱƖȱ ȱȱěȱȱȱ

192 Stress tolerance of vegetative diatom cells Figs 7.1-17ȱȱȱȱ¡£ȱȱȱȱȱȱȱȱȱȱȱ ȱ¡ȱ ȱȱȱȱȱęȱȱȱȱ¢ǯȱȱ ȱƽȱŗŖȱΐǯȱȱȱȱȱ¢ȱȱǰȱȱȱȱ¢ǯ ǯȱŝǯŗȱȱȱǻřǼǲȱǯȱǯȱŝǯŘȱȱ var. ȱǻŗǼǲȱǯȱǯȱŝǯřȱȱȱǻŗǼǲȱǯȱǯȱŝǯŚȱȱ ǻŘǼǲȱǯȱǯȱŝǯśȱȱȱǻŘǼǲȱǯȱǯȱŝǯŜȱȱȱǻŘǼǲȱ ǯȱǯȱŝǯŝȱȱȱǻśǼǲȱǯȱǯȱŝǯŞȱȱȱǻŘǼǲȱ ǯȱǯȱŝǯşȱȱȱǻŜǼǲȱǯȱǯȱŝǯŗŖȱȱȱ¢ȱǻśǼǲȱǯȱ ǯȱŝǯŗŗȱȱȱǻřǼǲȱǯȱǯȱŝǯŗŘȱȱȱǻřǼǲȱǯȱǯȱŝǯŗřȱ ȱȱǻŚǼǲȱǯȱǯȱŝǯŗŚȱȱǯȱȱǻŝǼǲȱǯȱǯȱŝǯŗśȱ ȱȱǻśǼǲȱǯȱǯȱŝǯŗŜȱȱȱǻřǼǲȱǯȱǯȱŝǯŗŝȱȱ ȱǻŚǼǲȱǯ

Stress tolerance of vegetative diatom cells 193 Figs 7.18-35ȱȱȱȱ¡£ȱȱȱȱȱȱȱȱȱȱȱ ȱ¡ȱ ȱȱȱȱȱęȱȱȱȱ¢ǯȱȱ ȱƽȱŗŖȱΐǯȱȱȱȱȱ¢ȱȱǰȱȱȱȱ¢ǯ ǯȱ ŝǯŗŞȱ ĵȱȱ ¡¢ȱ ¢ȱ Śȱ ǻŗǼǲȱ ǯȱ ǯȱ ŝǯŗşȱ ĵȱȱ ¡¢ȱ ¢ȱ řȱ ǻŜǼǲȱ ǯȱ ǯȱ ŝǯŘŖȱ ĵȱȱ ¡¢ȱ ¢ȱ Řȱ ǻŞǼǲȱǯȱǯȱŝǯŘŗȱ ĵȱȱ¡¢ȱ¢ȱŗȱǻŗǼǲȱǯȱǯȱŝǯŘŘȱĵȱ ȱ ȱ ǻŝǼǲȱ ǯȱ ǯȱ ŝǯŘřȱ ĵȱ ȱ ȱ ǻŞǼǲȱ ǯȱ ǯȱ ŝǯŘŚȱ ĵȱ ȱ ȱǻŗǼǲȱǯȱǯȱŝǯŘśȱȱȱǻŗǼǲȱǯȱǯȱŝǯŘŜȱȱǯȱ ȱǻŗǼǲȱǯȱǯȱŝǯŘŝȱȱȱǻřǼǲȱǯȱǯȱŝǯŘŞȱȱ ¢ȱ Řȱ ǻŗǼǲȱ ǯȱ ǯȱ ŝǯŘşȱ ȱ ȱ ¢ȱ ŗȱ ǻŗǼǲȱ ǯȱ ǯȱ ŝǯřŖȱȱȱǻřǼǲȱǯȱǯȱŝǯřŗȱ¢ȱȱ¢ȱŗȱ ǻŘǼȱǲȱǯȱǯȱŝǯřŘȱ¢ȱȱ¢ȱŘȱǻŜǼǲȱǯȱǯȱŝǯřřȱȱ ȱǻ ŗǼǲȱǯȱǯȱŝǯřŚȱȱȱǻŘǼǲȱǯȱǯȱŝǯřśȱ ¢ȱ var. ȱǻŘǼǲȱǯ

194 Stress tolerance of vegetative diatom cells Fig. 7.36ȱ ¡ȱ ȱ ȱ ȱ percentages of viable cells ǻƖǼȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ǯȱ ƽřŚȱ ȱ ȱ ǯ

Fig. 7.37ȱȱȱȱȱǻȱ ȱȱȱȱ ȱȱȱȱ ȱȱȱȱȱǼȱȱȱ¢ȱȱȱǯȱȱȱȱ ȱ¡ǰȱȱȱȱȱ¡ǯȱȱȱ¡ȱƽŘśǰȱȱȱ¡ȱ ƽşǯ

Stress tolerance of vegetative diatom cells 195 ȱ ǻpǀŖǯŖśȱ ȱ ȱ ȱ ȱ ȱ ƖǼȱ ǻǯȱ ŝǯřŜǼǯȱȱȱȱȱȱȱƸřŖǚȱǻǯȱŝǯřŝǼȱ ȱ ȱęȱȱěȱȱȱƖȱ ȱȱ¢ȱǻǯȱŝǯřŞǼǯ ȱȱȱƸŚŖǚȱ ȱȱ¢ȱȱęȱ¢ȱ ȱ ȱȱȱȱȱȱȱȱǻpƽŖǯŖŗřǼȱ ǻǯȱŝǯřŝǼǯȱȱȱȱǰȱȱȱȱ ȱȱȱȱȱȱȱȱ¢ȱ and Rhopalodia gibbaȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻȱŝǯŘǼǯȱ ȱ ǰȱȱȱȱȱ¢ȱ and Rhopalodia gibbaȱ ȱ ȱ ȱ ȱ ŚŖǚǯȱ ȱ Ɩȱ ȱ ¢ȱ ȱȱȱȱȱȱȱ ȱ¡ȱ ȱ ŖȱȱşŞƖȱǻȱŝǯŘȱȱǯȱŝǯřŞǼǰȱ ȱȱęȱȱěȱȱ ȱȱ ȱǯ ȱ ȱ ȱ ƸŚŖǚȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ǳȱ ȱ ȱǻ ûĵȱ Ǽȱ £ǰȱ ȱȱǻ Ǽȱȱȱ¢ȱ ǻȱŝǯŘǼǯȱȱȱ ȱȱȱȱȱ¢ȱ (pƽŖǯŗŚŘǼȱ ǻǯȱ ŝǯřŝǼȱ ǰȱ ȱ ȱ ȱ ȱ ǰȱ ȱ Ɩȱ ȱ ¢ȱ ȱ ȱ ȱ ęȱȱ ěȱȱ ȱ ȱ ȱȱ¢ȱǻǯȱŝǯřŞǼǯȱȱȱȱ¢ȱǰȱȱ ęȱ¢ȱ ȱƖȱ ȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ƸŚŖǚȱ ǻpƽŖǯŖŗŝȱ ȱ ȱ ȱ ȱ ȱƖǼȱǻǯȱŝǯřŜǼǯ

7ROHUDQFHWRIUHH]LQJ £ȱȱȬŘŖǚȱ ȱȱȱȱȱȱȱǻȱŝǯŘǼǯȱȱ ȱȱȱȱǯȱ¢ȱȱȱȱȱȱ ȱ ȱ ȱ ǰȱ ȱ Pinnularia borealisȱ ǻśȱ ȱ ȱ ŗŖȱǼǰȱ¢ȱȱǯȱȱǻȱȱǼȱȱ ĵȱȱ¡¢ȱ¢ȱŘȱǻŗȱȱȱŚǼȱȱ¢ȱŚȱǻŗȱ ȱȱŘǼȱǻȱŝǯŘǼǯȱȱƖȱĚȱȱ¢ȱ ȱǰȱ ȱŘȱȱŜŚƖȱǻȱŝǯŘǼǯȱȱęȱȱěȱȱ ȱȱ ȱȱ¢ȱȱȱȱȱǻpƽŖǯŘŗŝǼȱǻǯȱ ŝǯřŝǼȱȱȱƖȱǻpǁŖǯŖśȱȱȱȱȱȱƖǼȱ ǻǯȱŝǯřŞǼǯ

Tolerance to desiccation ȱȱȱȱ¢ȱȱǰȱ ȱȱ ȱ ȱǻȱŝǯŘȱȱǯȱŝǯřŝǼǯȱ ȱ ȱȱȱȱ

196 Stress tolerance of vegetative diatom cells Fig. 7.38ȱ¡ȱȱȱȱȱȱȱǻǰȱ¡Ǳȱęȱȱȱȱǰȱ Ǳȱȱȱ¡Ǽȱȱȱěȱȱȱȱȱȱ ȱǻȱŗŞǚǰȱƸřŖǚǰȱƸŚŖǚȱǰȱƸŚŖǚȱǰȱȬŘŖǚǰȱǰȱƸřŖǚȱ ǭȱ Ǽȱ ȱ ȱ ȱ ǻ ȱ Ǽȱ ȱ ȱ ȱ ǻȱ Ǽǯȱ ȱ ȱ¡ȱƽŘśǰȱȱȱ¡ȱƽşǯ ȱȱȱęȱȱȱȱȱěȱȱǰȱPinnularia borealisȱǻŘȱǼǰȱ ĵȱȱ¡¢ȱ¢ȱŚȱǻŗȱǼȱȱ ȱȱǻŘȱǼǯȱȱęȱȱ ȱȱ¢ȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱǻȱŝǯŘǼǯȱƖȱȱȱȱ ȱ ȱśȱȱŞŚƖȱǻȱŝǯŘǼǯ

,QÀXHQFHRIYDOYHOHQJWK ¢ǰȱȱȱȱȱȱ¡ȱƖȱȱ ȱȱ¢£ȱȱǻŶȱȱȱȱŖǯŖŞǰȱpǁŖǯŖśǼǰȱȱ ȱƖȱȱȱȱȱP. borealis and ǯȱ¡¢ȱ¢ǯ

Stress tolerance of vegetative diatom cells 197 7.4 Discussion ȱȱȱȱȱȱȱȱ¢ȱ ¢ȱȱȱȱȱȱȱǰȱ ȱȱȱȱȱǻƖǼȱęȱ¢ȱȱȱȱ ¡ȱȱȱȱȱȱǯȱǰȱȱ ȱȱȱȱ¢ȱȱ ȱȱȱȱ ǰȱ£ȱȱǯȱ¢ȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ƸŚŖǚǰȱ ¢ȱ ȱ ȱ ȱ £ǰȱȱȱȱȱȱ¢ȱȱǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ řŖȬŚŖǚȱǻ£ȱǭȱǰȱŗşşśǲȱĴȱ ȱet al.ǰȱŘŖŖśǼǰȱȱȱ ȱ ȱ¡ȱȱȱǻ ȱǭȱǰȱŗşŝŖǼǯȱ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ǰȱ ȱȱȱȱȱȱȱȱ ȱȱȱȱ¢ȱȱȱ ȱǻĴȱǰȱŗşşşǲȱȱ ǭȱǰȱŘŖŖŜǲȱ ¢ȱet al.ǰȱŘŖŖŝǼǰȱȱ£ȱȱȱ ȱȬŚŖǚȱȱȱȱǻ et al.ǰȱŘŖŖśǲȱȱǭȱǰȱ ŘŖŖŜǼǯ ȱȱȱ£ȱȱȱȱȱ¢ȱ ȱȱǻǰȱŘŖŖŖǼǰȱȱȱȱȱȱ ȱȱ ȱȱȱȱ£ǯȱȱȱ ȱȱ ȱȱȱȱȱȱȱȱ£ȱǻȱet al.ǰȱŘŖŖŜǲȱȱǭȱǰȱŘŖŖŜǼǯȱ ¢ǰȱȱȱȱȱ ȱȱ ǯȱ¡¢ and P. borealisȱȱȱ£ȱ ȱȱȱȱȱǯȱȱȱȱȱȱ ȱ ȱ¢ȱ ǰȱȱ¢ȱȱȱȱȱȱȱȱ ȱ ȱȱ¢ȱȱȱ ȱȱȱȱȱȱ ȱ ǻ ȱřŖŖȱȱŗŖŖŖǼǯȱȱȱȱȱȱȱȱȱ ȱȱȱȱ£ȱȱǰȱ¢ȱ¢ȱȱ ȱěȱȱ¡ȱȱȱȱȱȱȱ ¢ȱǻǰȱŘŖŖŖǼǰȱȱȱǻȱet al.ǰȱŗşşŞǲȱȱ et al.ǰȱ ŘŖŖŚǼǰȱ ¡ȱ ȱ Ĵȱ¢ȱ ȱ ȱ ǻ ȱ et al.ǰȱŘŖŖŘǲȱȱet al.ǰȱŘŖŖřǲȱȱet al.ǰȱŘŖŖŚǲȱȱet al.ǰȱŘŖŖŝǼǰȱ ȱ¢ȱȱȱ¢ȱȱǻ ȱet al.ǰȱŗşşŘǼǯȱ ȱǰȱ ȱ ȱȱǯȱȱ ȱȱȱȱȱȱ ŝŚȱȱŞŚƖȱȱȱȱǯȱ¢ǰȱȱȱ ȱȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ǰȱ ȱȱȱȱ ȱĴȱȱȱȱȱȱȱ¡ȱ ǯȱ ǰȱ ȱ ȱ ȱ ȱ ȱ

198 Stress tolerance of vegetative diatom cells ȱȱȱȱȱȱȱȱȱȱ ȱȱȱ¢£ǯȱȱȱȱȱȱ ȱȱȱȱȱǻȱȱȬęȱǼȱȱȱ ȱȱȱȱ¢ȱȱȱ ȱ¢ȱȱȱȱ ȱ£ǰȱȱȱȱȱȱȱȱȱȱȱȱ ȱȱǯ ȱȱȱȱȱȱǰȱȱ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ śŖƖȱ ȱ ȱ ǻǰȱ ŗşśşǼǯȱ ȱ ȱ ȱ ȱ ȱȱȱ ȱȱȱȱ ȱȱ¢ȱǻ ȱ et al.ǰȱŘŖŖŞǼȱȱȱȱȱȱȱȬȱ ȱ ȱ ¢ȱ ȱ ¢ȱ ȱ ǻǰȱ ŗşśşǼȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱ ȱǰȱȱ ȱȱȱȱȱ ȱ ȱȱȱȱǻǰȱŗşśşǲȱ Ĵȱȱǭȱ ǰȱŗşŝŖǼǰȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻȱ et al.ǰȱ ŘŖŖŜǼȱ ȱ ȱ ¢ȱ ȱ ȱ ǻȱ ǭȱ ǰȱ ŗşşŜǼǯȱ ȱ ȱ ¢ǰȱȱ¢ȱȱȱȱȱƸřŖǚȱȱȱȱ ȱȱȱǰȱȱȱ ȱȱĚȱȱ ȱȬȱȱȱȱ¢ȱȱȱȱȱ ȱ ȱ ȱ ȱ ǻȱ et al.ǰȱ ŗşşŞǲȱ ¢¢ȱ et al.ǰȱŘŖŖŗǲȱȱet al.ǰȱŘŖŖřǲȱȱet alǯǰȱŘŖŖŝǼǯȱ ǰȱȱȱ ȱȱȱ ȱȱěȱȱȱǰȱȱȱ ȱȱȱȱȱȱȱȱŚŖǚǯȱǰȱ ȱȱȱȱȱȱȱȱĚȱȱȱƖȱ ȱȱǯȱȱȱȱǰȱȱȱȱȱȱ Rhopalodia gibbaǰȱ ĵȱ and ǯȱ ¡¢ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱȱȱǰȱ ȱȱȱȱȱȱȱȱȱȱȱȱȱȱ ȱǯȱ ȱ¢ȱȱ ȱȱȱ£ǰȱȱ ȱ ȱ ȱ ¢ȱ Ěȱȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱ ŘŖŖŖǲȱ ȱ ǭȱ ǰȱ ŘŖŖŖǼǰȱ ȱȱȱȱȱȱȱȱȱ¢ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱǻȱet al.ǰȱŘŖŗŖǼǯȱǰȱȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ¢ȱ ȱ ȱ ȱ Ȯȱ

Stress tolerance of vegetative diatom cells 199 ȱȱȱȮȱȱ¢ȱȱȱȱȱȱȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻȱ ǰȱ ŗşřŝǲȱ ǰȱ ŗşŜŗǲȱ ȱet al.ǰȱŗşŜŚǲȱǰȱŗşŜŚǲȱ¢ȬȱǭȱǰȱŗşşřǼȱȱ ȱ ȱǻǰȱŗşŜŗǼǯȱȱȱȱȱ¢¢ȱ ȱȱȱȱ ȱȱȱȬȱ ȱ¢ȱ ȱȱǯȱȱ ȱ¢ȱȱ¢ȱȱȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱ ȱȱȱǰȱȱ ȱȱ ¢ȱȱŗŖŖȱȱȱȱȱȱǻ et al.ǰȱŘŖŖşǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ǰȱ ȱȱȱȱ¢ȱȱȱȱȱȱ¢ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ûĵȱ ȱ ǻȱȱ ȱ et al.ǰȱ ŘŖŖŘǼǰȱ Luticolaȱ ǯ ǯȱ ȱ ǻȱȱȱǭȱǰȱŘŖŖŞǼǰȱMuelleriaȱǻǼȱȱ ǻȱet al.ǰȱŗşşşǼǰȱȱȱȱǻȱȱȱet al.ǰȱŘŖŖśǼǯȱ ǰȱȱȱȱȱȱȱȱȱ ȱ¡ȱȱȱȱȱȱǰȱȱȱȱ ȱ ȱȱȱȱȱȱȱȱȱȱȱ ȱ¢ȱǯȱȱǰȱ ȱ¢ȱȱȱȱ ¢ȱȱȱȱ¢ȱ ǰȱřřȱȱȱȱȱȱ ǰȱ Ŝşȱěȱȱȱȱ ȱǰȱ¢ȱȱȱ ȱ ȱ ȱ ȱ ǻȱ ǭȱ ǰȱ ŗşŝŖȱ ȱ ǰȱ ŗşşŜǼǯȱǰȱ ȱȱǰȱȱȱȱ ȱȱ¢ȱȱǰȱȱȱȱ ȱȱȱ ¢ȱȱ¢ȱ ȱȱ ǯ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱǰȱȱȱȱȱȱȱȱȱ ¢ȱȱȱȱȱȱȱȱ ȱ ȱ ǯȱ ȱ ǰȱ ȱ ȱ ǰȱ ȱ ¢ȱ ȱȱǰȱǰȱȱȱȱǰȱȱ£ǯȱ ȱ ¢ȱ ȱ ȱ ȱ Ɩȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ Ŗȱ ȱ ŗŖŖƖȱȬȱ ȱ ¢ȱ Ȭęȱȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱǯȱȱȬęȱȱȱȱ ȱ¢ȱȱȱȱȱǻĴȱ et al.ǰȱŘŖŖśǲȱ ¢ȱet al.ǰȱŘŖŖŝǼǯȱ ȱ ȱ ȱ¢ȱȱȱȱȱȱȱ ȱȱȱȱȱȬȱȱěȱȱ ȱǰȱ ȱȱȱ¡¢ȱȱȱęȱȱȱ ȱ Ɩǯȱȱ ȱ ¡ȱ ȱ ȱ ȱ £ȱ

200 Stress tolerance of vegetative diatom cells ȱȱ ȱęȱȱȱȱȱĴȱǰȱȱȱȱ ȱ ¢ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ ȱ ǰȱ ȱȱȱȱȱěȱȱȱȱȱ ȱȱ ȱȱȱǯȱȱǰȱȱȱȱ ǯȱ¡¢ȱǯȱ Śȱ ȱ £ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱȱȱP. borealisǰȱȱ ȱȱȱȱȱȱȱȱ £ǯȱ ȱ ȱ ȱ ȱ ȱ ƸŚŖǚȱ ǰȱ ěȱȱ ȱ Ɩȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ¡ȱȱȱȱȱȱ¡ȱȱęȱȱ ȱȱȱȱǯ ǰȱ ȱ ȱ ȱ Ȭȱ ȱ ęȱȱ ěȱǰȱȱȬȱȱȱȱȱȱ ƸŚŖǚȱȱ£ȱȱȱȱȱĴȱȱȱȱȱȱ ȱȱȱȱȱȱȱ¡ȱǯȱǰȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ƸŚŖǚȱ ȱȱȱǰȱ ȱȱ¢ȱȱȱȱȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ǻ ȱet al.ǰȱŘŖŖŞǼǯȱȱȱȱȱȱȱȱȱ Rhopalodia gibba and ¢ȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭęȱȱ ěȱǯȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ¢ǯȱ ȱ ěȱȱ ȱ ȱ ȱȱěȱȱǰȱȱȱȱ¢¢ȱ ȱ¢ȱȱȱȱȱȱȱȱ ȱ ǻ ȱ ǭȱ ǰȱ ŘŖŖśǲȱ ¢ȱ et al.ǰȱ ŘŖŖŝǲȱ ȱ et al.ǰȱ ŘŖŖŞǼǯȱ ȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǻǼ¢ȱ ȱ ¢ȱ ¡ȱ ǻȱ et al.ǰȱ ŘŖŖśǲȱ ȱ ǭȱ ǰȱ ŘŖŖŝǲȱ et al.ǰȱŘŖŖŞǼǰȱȱ¢ȱȱȱȱ¢ȱ ěȱȱ ȱ ǻȱ et al.ǰȱ ŘŖŖşǼǯȱ ¢ǰȱ ȱ ¢ȱ ȱěȱȱ¢ȱȱǰȱȱȱȱȱ ȱȱȱ ¢ȱȱȱȱȱ ȱȱ¢ǯȱ ȱ ǰȱȱȱȱȱȱȱ¢ȱȱ ȱ¡¢ȱȱȱ ¢ȱȱȱȱ¡ȱȱ ȱ ǻȱȱ¢Ǽȱȱȱěȱǯȱȱǰȱǯȱȱ¢ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱȱȱȱȱȱȱȱȱȱȱȱ radiosaǯȱȱȱȱ¢ǰȱ ǯ ȱ ȱ ȱ ȱ ȱ Ȭęȱȱ ȱ ȱ ȱ Ȯȱ ȱȱ¢ȱȱȮȱȱȱȱ£ȱ ¡ǯȱȱȱȱȱȱȱȱȱ¢ȱ ȱ¡ȱ ĵȱȱ¡¢ǰȱPinnularia borealis and ¢ȱ

Stress tolerance of vegetative diatom cells 201 ȱǯȱȱǻĴȱȱǭȱ §ǰȱŗşşśǼǯȱ¢ǰȱȱ¢ȱȱ ȱȱȱȱȱ£ȱȱȱȱȱ ancepsȱȱȱȱȱȱȱǻ Ĵȱȱ ǭȱ ǰȱŗşŝŖǼǯȱȱȱȱȱ£ȱ¢ȱȱȱ ȱȱȱȱȱȱ¡ȱȱȱȱȱ ȱȱǻ ȱet al.ǰȱŘŖŖŞǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱȱȱǯȱȱȱȱ ȱȱȱ ȱ ǰȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǻ Ĵȱȱ et al.ǰȱ ŘŖŖŜǼǯȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱǻe.g.ȱȱet al.ǰȱŘŖŖřǲȱ ǰȱŘŖŖŞǲȱȱet al.ǰȱŘŖŖŞǲȱȱǭȱǰȱŘŖŖşǼǯȱȱǰȱ ȱ¢ȱȱěȱȱȱȱȱ ȱȱ ȱȱȱȱȱ¢ȱȱȱȱȱȱ ȱ¢ȱǰȱȱ¢ȱȱȱȱȱ¡ȱǯ ȱ ǰȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ £ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ¢ǰȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱ¡ȱȱ ȱȱǰȱ¢ȱȱȱȱȱȱȱ ¡ȱȱǯȱȱȱȱȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ¢ȱ ȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ¢ǰȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ěȱȱ ȱ ȱ £ȱȱȱȱȱ¢ȱȱǯ

Acknowledgements ȱȱȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱǯȱȱȱ ȱ ęȱ¢ȱȱ¢ȱȱȱȱęȱȱȱȮȱȱ ǻȬǰȱǰȱȱȱȱȱǼǰȱȱȬȱ ř ȦŖśřřȦŖŝȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ

202 Stress tolerance of vegetative diatom cells ȱǰȱȱȱȬȱř ȦŖŚŗşȦŖŞȱȱȱ ǯ

Stress tolerance of vegetative diatom cells 203

8. Resistance of freshwater benthic diatom resting stages to experimental desiccation and freezing depends on the temporality of their habitat

ȱěȱǰȱȱȱȱȱ¢

ǐȱȱ¢ȱȱ¢ȱȱȱ¢ǰȱ ȱ ¢ǰȱ ȱŘŞŗȱȬȱŞǰȱşŖŖŖȱ ǰȱ

Stress tolerance of diatom resting cells 205 Abstract ǰȱ ¢ȱ ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ ȱ ¡ȱ ȱ ¢ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ǯȱȱȱȱ¢ȱȱ£ȱ¢ȱěȱȱ ȱǰȱȱȱȱ ȱȱȱȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ǰȱȱ ȱȱȱȱȱȬǯȱ ȱ ¡¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ȱȱȱȱǰȱȱȱȱȱǰȱ ȱŗŝȱȱ ȱȱȱȱȱȱ ȱȱȱȱȱȱǰȱȱȱęȱȱ ȱǯȱȱȱȱ ȱȱ¢ȱȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ £ǯȱ ¢ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ £ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱȱȱśȱȱȱǯȱ ȱǰȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ¢ȱ ȱ ȱ¢ȱȱȱǰȱȱȱȱȱ¡ȱȱ £ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱǯȱ¢ǰȱ¢ȱȱȱȱȱȱ ¡ǰȱ i.e.ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ¢ȱȱǻȱ¢ȱŚǼȱȱ¢ȱ¡¢ȱȱ ȱ ȱǻ¢ȱśǼȱȱȱȱ£ǰȱȱȱ ȱȱȬęȱȱȱȱȱȱǯȱ ȱ ȱ ȱ ȱ ȱ ǻȱ Ǽȱ ȱ ȱ ȱ ǰȱ ȱ ǰȱ ȱ ȱ ȱ ȱȱȱȱȱǯ

Key words:ȱ ȱ ǰȱ ǰȱ ǰȱ ȱ ǰȱǰȱ£ǰȱȱǰȱ

206 Stress tolerance of diatom resting cells 8.1 Introduction ȱȱȱȱȱȱȱǰȱȱȱ ȱǰȱȱȱ¢ȱȱȱȱȬȱ ȱȱȱȱȱȱȱ¢ȱȱ ȱȱȱȱ ǻȱ et al.ǰȱ ŗşşŖǼǯȱ ȱ ¢ȱ ȱ ¢ȱ ¢ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ǻȬǼȱ ȱ ȱ ěȱȱ ȱ ȱǻȱet al.ǰȱŗşŞŗǲȱ ȱet al.ǰȱŘŖŖŞǼǯȱȱȱ¡ȱȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱȱǻǰȱŗşşŚǼǯȱȱȱǰȱȱȱȱ ȱȱȱȱȱ ȱȱȱȱȱ ȱ¢ȱęȱȱȱȱȱ ȱǰȱęȱȱȱ ȃȄȱȱǻǰȱŗşřśǲȱĴȱȱǭȱ §ǰȱŗşşśǼǰȱȱ ȱ ȱ Achnanthes coarctataȱ ǻ·Ǽȱ ǰȱ Luticola spp. ǻǼǰȱ ĵȱȱ ¡¢ȱ ǻǼȱ ǰȱ Pinnularia borealisȱ ȱ ȱ ¢ȱ ȱǻ ûĵȱ Ǽȱ Ȭȱ ǻĴȱȱǭȱ §ǰȱŗşşśǲȱ ǰȱŗşşşǲȱȱ ȱet al.ǰȱŘŖŖŖǲȱ ȱȱȱet al.ǰȱŘŖŖŚǲȱǰȱŘŖŖŜǼǯȱȱȱet al.ȱǻŗşşŚǼȱ ȱęȱȱȱȱȱȱȱȱ ȱ ȱȱȱȱȱȱ¢ȱȱȱ ȱȱ¢ȱ ȱȱȱȱ ȱǯ ȱȱȱ¢ȱȱȱȱȱ ȱȱȱ ȱȱȱĚȱǰȱȱȱȱ ȱȱȱ ȱȱȱȱȱȱȱ ǯȱ ȱȱȱ¢ȱȱ ȱȱȱȱ ȱȱȱ ȱ¡ǯȱěȱȱ¢ȱȬ¡ȱȱ ȱǯȱǰȱȱȱȱȱȱȱ ȱȱȱȱȱȱȱǰȱȱȱȱ Ȭ¢ȱǻĴȱǰȱŗşşşǲȱȱet al.ǰȱŘŖŖśǼȱȱȱȱȱ ȱȱǻǰȱŘŖŖŖǲȱȱet al.ǰȱŘŖŖŚǲȱ ¢ ȱet al.ǰȱ ŘŖŖŝǲȱȱet al.ǰȱŘŖŖŝǼǯȱ ǰȱȱȱȱȱ ȱ ȱ ¢ȱ ¢ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻěȱȱet al.ǰȱŘŖŗŖǼǰȱȱȱȱȱȱ¢ȱ ȱȱȱȱȱȱǻǰȱŗşśŞǰȱŗşśşǲȱ Ĵȱȱ ǭȱ ǰȱŗşŝŖǲȱ£ȱǭȱǰȱŗşşśǲȱĴȱ et al.ǰȱŘŖŖśǼǯȱ ȱȱȱȱȱȱ¡ȱȱȱ¢ȱǰȱ ȱȱȱȱȱśŖƖȱȱȱȱȱȱ ȱ ǻǰȱ ŗşśŞǰȱ ŗşśşǼǯȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ

Stress tolerance of diatom resting cells 207 ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭ ȱ ȱ ȱȱȱ¢ȱȱȱǻȱet al.ǰȱŘŖŗŖǼǯȱ¢ǰȱ ȱȱȱȱȱȱȱ¢ȱȱ ȱ¡ȱȱȱȱȱǻǼ£ȱȱ ¢ȱȱ ȱȱȱȱȱǯȱ ǰȱ ȱ¡Ȭ£ȱ¢ȱȱ¢ȱȱȱȱ ȱȱȱȱȱ ȱȱȬȱ ȱǻ ǰȱŗşşŜǲȱ ȱǭȱ ǰȱŘŖŖŗǼǯȱȱęȱȱ¢ȱȱ ȱȱȱȱȱǰȱ ȱ ȱȱ¢ȱ ȱ ȱȱȱȱ¢ȱȱȱȱ£ȱȱ ȱ ȱǯ ȱȱȱ¢ȱęȱȱȱȱȱȱ¢ȱȱȱ ȱȱ ȱ ȱȱȱȱȱȱ ȱ Ȃȱ ǯȱ ¢ȱ ȱ £ȱ ¢ȱ ȱ ȱ ȱ ¢ǰȱȱȱ¢¢ȱȱȱȱȱȱ ǯȱȱȱȱ ȱȱȱǰȱ ȱȱȱǻe.g.ȱȱet al.ǰȱŘŖŖŖǼǰȱȱ ȱ¢ȱǻe.g.ȱȱet al.ǰȱŗşŝşǼǰȱȱ¢ȱȱȱ ȱ ȱ ȱ ǻe.g.ȱ ǰȱ ŗşŝśǲȱ ȱ ǭȱ ǰȱ ŘŖŗŖǲȱ ǰȱŘŖŗŖǲȱûȱet al.ǰȱŘŖŗŖǼȱȱȱ£ȱȱȱ ǻe.g.ȱǰȱŗşśŚǲȱ ǰȱŗşŜŝǲȱ ȱǭȱǰȱŗşşśǼǰȱ ȱȱȱȱǻe.g.ȱǰȱŗşŜşǲȱ¢ǰȱŗşŞŖǲȱȱǭȱ ěȱǰȱŗşşřǼȱȱȱȱȱȱǻe.g.ȱ ȱet al.ǰȱŗşşśǼǯȱ ȱȱȱȱǯȱ ȱ¢ȱȱȱȱȱ ǲȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ǻ ȱ ȱ ȱǭȱ ǰȱŗşşŜǼǰȱ ȱȱȱȱ¢ȱ ȱȱȱȱȱǻǰȱŗşśřǲȱǰȱŗşŝśǲȱȬ ȱet al.ǰȱŗşŞŜǼǯȱȱȱȱȱĚȱ¡ȱȱ¢ȱȱ ȱȱȱȱȂȱȱȱȱȱǰȱ ȱ ȱȱȱȱȱȱȱȱ¢ȱǻ ȱ ǭȱǰȱŗşşŖǲȱ ȱet al.ǰȱŗşşřǼǯȱȱ¢ȱȱȱȱ ȱ¢ȱȱȱȱȱȱȱȱǰȱȱ ȱȱȱȱȱȱȱȱ¢ǰȱȱȱ ȱȱȱěȱȱ¡ȱȱǻǰȱŗşŝśǲȱ ȱ ǭȱǰȱŗşŝśǲȱ ȱet al.ǰȱŗşŞŖǲȱȬ ȱet al.ǰȱŗşŞŜǼǯȱȱ ȱȱ¢ȱȱȱȱǰȱ ȱȱȱ ¡ȱ ǻ ȱ ȱ ȱ ǭȱ ǰȱ ŗşŝśǲȱ ȱ ǭȱ ǰȱ ŗşşŜǲȱ ǰȱ ŘŖŖşǼǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱǻȬ ȱet al.ǰȱŗşŞŜǲȱȱǭȱ ǰȱŗşşŜǼȱ ȱȱȱȱȱȱȱǯ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ

208 Stress tolerance of diatom resting cells ȱȱȱȱȱǻǰȱŗşśřǲȱȱǭȱ ǰȱŗşşśǲȱ ȱet al.ǰȱŗşşśǲȱǰȱŘŖŖŘǲȱ ȱet al.ǰȱŘŖŖŜǲȱÇ¤ȱ et al.ǰȱŘŖŖŞǼǰȱ ȱ¢ȱ¡ȱȱȱȱȱȱȱ ȱ ȱ ȱ ǻ ǰȱ ŗşŗŘǲȱ ǰȱ ŗşśřǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ǰȱ ȱ ȱ ¡¢ȱ ȱ ȱ ȱ ȱ ȱ ǻe.g. ȱ ǭȱ ǰȱ ŗşŝśǲȱ ȱ et al.ǰȱ ŗşŞŗǲȱ ȱ et al.ǰȱ ŗşşřǲȱ ȱ ǭȱ ǰȱ ŗşşśǲȱ ȱet al.ǰȱŘŖŖŜǼǯȱ ȱ¢ȱȱǰȱ ǰȱȱ ȱȱȱȱěȱǰȱȱȱȱȱ ¡ǯȱȱȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱȱȱǻǰȱŗşśŞǰȱ ŗşśşǼǰȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ǻǼȱ ȱ ȱ ȱȱȱȱȱȱǯȱ ȱȱ¢ȱ ȱȱ ¡¢ȱ ȱ ȱ ȱ ¢ǲȱ ǻŗǼȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ǰȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ǻŘǼȱ ¢ȱ ȱ ȱ ȱȱ ȱȱ£ȱȱȱȱȱȱ ȱȱ ǰȱȱȱěȱȱȱȱȱ¢ȱȱ ȱ ȱ ȱ ¢ǯȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱȱȱȱǻȱǭȱ ǰȱŗşşśǼȱȱśŖȱ ȱȱŗŝȱȱȱȱȱǰȱȱ ȱȱęȱȱȱȱȱȱȱet al. ǻŗşşŚǼǯȱ¡ǰȱȱ ȱ ȱ £ȱ ȱ ȱ ǻ ȱ ȱ ȱ ȱ ȱǼȱ ȱȱȱȱȱȱȱȱ ȱȱȱȱȱǯ

8.2 Materials & Methods 7D[RQVDPSOLQJDQGLGHQWL¿FDWLRQ ȱ ȱ śŖȱ ȱ ȱ ȱ ŗŝȱ ȱ ȱ ȱ ȱ ȱ ǻȱ ŞǯŗǼǰȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ¢ȱȱȱet al.ȱǻŗşşŚǼȱǻȱŞǯŘǼǯȱ¢ȱȱ ȱ ȱȱȱ ȱȱ ȱȱ ȱǰȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ǯȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ǰȱ ȱ £ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱěȱȱet al.ȱǻŘŖŗŖǲȱĴȱǼǰȱ et al. ǻŘŖŖŜǼǰȱȱet al. ǻŘŖŖşǼǰȱȱet al.ȱǻŘŖŖŞǼȱȱȱet al.ȱǻŘŖŖşǼȱǻȱ ȱ ȱ ȱ ŞǯŗǼǯȱ ȱ ȱ ǻŖŝŖǼȱ ȱ ǻŖŚřǼȱ ȱ ȱȱȱȱěȱȱet al.ȱǻŘŖŗŖǼȱȱȱȱȱ

Stress tolerance of diatom resting cells 209 Ǽȱ RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 -20°C -20°C ȱŗǯŖȱȱŗŖǯŖƖǲȱȱ RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 +30°C +desiccation

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 +30°C +desiccation RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 ȱǰȱȱȱȱȱȱȱǻȬ¡

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 RC Control Desiccation Desiccation 98.7-100 95.0-95.6 Origin BCCM/DCG BCCM/DCG Dr. D.G. Mann D.G. Dr. Mann D.G. Dr. 86.4-94.5 96.3-99.9 Dr. D.G. Mann D.G. Dr. Mann D.G. Dr. 99.9-99.9 99.9-100 Culture collection Culture collection Belgium – De Panne 94.1-95.3 Belgium – De Panne 99.9-100 Belgium – De Panne 99.7-99.7

(Ehr.) (Ehr.) (Ehr.) Cleve Cleve Taxon Grunow Grunow Krammer Krammer . ȱǼȱ¢ȱȱȱȱȱȱȱ¢ȱȱȱȱȱ Lange-Bertalot Lange-Bertalot Lange-Bertalot

Nitzschia fonticola Nitzschia fonticola Pinnularia grunowii Pinnularia grunowii Caloneis silicula Caloneis silicula cryptotenella Navicula cryptotenella Navicula cryptotenella Navicula c c c b b a a a ȱǯ ȱǻĴ ȱǯȱ ǻŘŖŗŖǼȱ ¢ 49 ǱȱřŖşȬřŘŚǯ ȱȱǯ ȱǻŘŖŖŜǼȱ ȱȱ¢¢ 42 ǱȱŗřśřȬŗřŝŘǯ ȱȱǯ ȱǻŘŖŖşǼȱ ¢ 48 ǱȱŚŚřȬŚśşǯȱ ȱ ȱ ȱ ȱȱ ȱȱ 1 Cal 856 1 Clone A 1 Clone C 1 867 Pin 1 889a Pin 2(Wes5)c 2(Pan2)c 2(Pan2)d 1 Cal 769 M.C. Strain ȱȱǯ ȱǻŘŖŖşǼȱ 160 ǱȱřŞŜȬřşŜǯ ȱ ȱ ȱě ȱě ȱǯ ȱǻŘŖŖŞǼȱ 159 ǱȱŝřȬşŖǯ ȱȱ ȱ ȱȱŗŖǯŖƖǯȱȱȱȱȱ ȱȱȱȱ¡£ȱȱȱ ȱȱȱȱǯ Table 8.1 ȱȱ ȱȱȱȱ¢ȱǻǯǯǰȱȱȱŞǯŘǼǰȱę Table a b c d e f ȱȱěȱȱȱȱȱȱȱǻǼȱȱȱȱǻǼǯȱȱȱȘȱȱȱȱ ȱǰ ȱ ȱȘȘȱȱ ȱȱ ȱǯȱȱȱȱȱ¡ȱȱȱ ȱŗǯŖƖǲȱȱȱ

210 Stress tolerance of diatom resting cells RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 -20°C -20°C RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 +30°C +desiccation

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 +30°C +desiccation RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 RC Control Desiccation Desiccation 99.8-100 99.9-100 98.6-99.7 99.8-99.9 99.9-99.9 99.8-99.9 99.3-99.5 99.4-99.6 99.4-99.5 99.9-99.9 99.0-99.1 Origin Belgium – Belgium – Belgium – Belgium – Belgium – Belgium – BCCM/DCG BCCM/DCG BCCM/DCG BCCM/DCG BCCM/DCG Blankenberge Blankenberge Blankenberge Blankenberge Blankenberge Oostduinkerke Oostduinkerke Oostduinkerke Oostduinkerke Oostduinkerke Culture collection Culture collection Culture collection Culture collection Culture collection Belgium – De Panne 99.7-99.7 ” ” ” ”

Mann Mann

(Ehr.) (Ehr.) labile (Ehr.) labile (Ehr.) (Ehr.) robust robust Kützing Kützing Belgium – De Panne Kützing Belgium – De Panne 90.7-94.4 Kützing Belgium – De Panne 99.8-100 Kützing 99.7-99.7 Belgium – De Panne 99.6-99.6 Kützing Belgium – De Panne Kützing 99.4-99.9 (Kützing) (Kützing) Taxon W. Smith W. Smith & McDonald & McDonald (Kützing) Grunow (Kützing) Grunow Nitzschia frustulum Nitzschia frustulum

Eunotia bilunaris Schaarschmidt “ Eunotia bilunaris Schaarschmidt “ Eunotia bilunaris Eunotia bilunaris Schaarschmidt “ Schaarschmidt “ radiosa Navicula radiosa Navicula radiosa Navicula radiosa Navicula radiosa Navicula veneta Navicula veneta Navicula Sellaphora capitata Sellaphora capitata Sellaphora Nitzschia palea Nitzschia palea e c c c c c c c c e c c c f c d d 2Mf38 3(Fon3)g 3(Fon3)i 3 DM22-17 3 DM22-4 3(Fon3)f 3(Wes4)d 3(Wes5)a 3(Wes5)b 3(Pan2)b 3(Wes6)d 3(Yde3)f 3(Yde1)c 3(Yde1)o 3 (02)7D 3(Wes7)e 2Mf18 M.C. Strain

Stress tolerance of diatom resting cells 211 RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0.1* 0.2-0.7 0.5-1.0 0.1-0.5 0-0.08* 7.2-22.6 0.07-0.5 0.02-0.1 0-0.02** 0.006-8.2

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 -20°C -20°C 2.1-39.3 10.3-18.3 RC 0-0 0-0 0-0 0-0 0-0.7* 1.1-0.7 0.7-1.8 1.1-2.2 +30°C 1.4-15.8 0.08-0.6 2.2-19.6 0-0.05** 27.4-35.3 36.8-69.0 0.04-0.06 0.002-0.004 +desiccation

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 +30°C +desiccation RC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0.3-2.2 5.6-13.8 28.9-54.8 13.4-31.0 28.9-43.5 33.6-36.5 0.001-0.002 0.001-0.002

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 0-0 RC Control Desiccation Desiccation Origin Belgium – Gent 99.7-99.7 Belgium – Gent 99.7-99.8 Belgium – Gent 99.7-99.8 Belgium – Gent 99.4-99.6 Belgium – Gent 99.8-99.9 Belgium – Gent 99.6-99.6 Belgium – Gent 99.8-99.8 Belgium – Gent Belgium – Gent 100-100 99.9-99.9 Belgium – Gent 98.2-99.2 Belgium – Gent 99.7-99.8 Belgium – Gent 99.8-99.8 Amsterdam Island 99.8-99.9 Amsterdam Island 99.9-99.9 Belgium – De Panne 98.4-98.7 Belgium – De Panne 99.9-99.9

sp sp. Grunow Grunow Taxon Grunow Grunow Ehrenberg (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.) Grunow (Ehr.)

Mayamaea atomus Mayamaea atomus Mayamaea Pinnularia borealis Nitzschia perminuta Nitzschia perminuta Nitzschia communis Nitzschia communis Hantzschia amphioxys Hantzschia amphioxys Hantzschia amphioxys Hantzschia amphioxys Hantzschia amphioxys Hantzschia amphioxys Hantzschia amphioxys (Grunow) M. Peragallo (Grunow) M. Peragallo Hantzschia Hantzschia (Kützing) Lange-Bertalot (Kützing) Lange-Bertalot c c b b c c c c c c c c c c c c 4(St5)b 3(St8)i 4(St1)e 4(St2)d 4(Wes2)e 4 (W043) 4(St2)d 4(St4)a 4(St5)c 4(St1)h 4 (M070) 4(St5)a 4(Wes2)f 4(St1)f 4(St3)c 3(St8)e M.C. Strain

212 Stress tolerance of diatom resting cells RC 0-0.8* 0.2-1.6 0-0.06* 0.08-0.8 0.07-0.2 27.0-30.5 88.5-90.0 95.2-96.9

VC 0-0 0-0 0-0

-20°C -20°C 0.3-1.6 2.1-3.0 3.1-12.4 4.6-16.2 13.9-23.2 RC 0-0 0-0 4.4-6.1 2.4-4.8 0.2-0.5 0.1-0.4 1.6-1.8 +30°C 0.06-0.4 +desiccation

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0

+30°C 2.0-3.9 +desiccation RC 0-0 0-0 0-0 0-0 0-0.8* 3.9-5.7 1.2-2.0 0-0.08**

VC 0-0 0-0 0-0 0-0 0-0 0-0 0-0

1.0-2.7 RC Control Desiccation Desiccation Origin Canada – Banff Canada – Banff 99.1-99.3 Belgium – Gent Belgium – Gent 99.6-100 99.7-99.8 Belgium – Gent 97.7-98.7 Belgium – Gent Belgium – Gent 100-100 99.7-99.8 Belgium – Gent 99.7-99.8 Belgium – De Panne 97.1-97.2

Taxon Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg Ehrenberg

Pinnularia borealis Pinnularia borealis Pinnularia borealis Pinnularia borealis Pinnularia borealis Pinnularia borealis (Brébisson) Grunow (Brébisson) Grunow (Brébisson) Achnanthes coarctata Achnanthes coarctata c c c c c c c 4(St6)a 4(St8)a 4(St8)d 5(Ban2)b 4(St1)b 5(Wes3)f 4(St1)i 4(St6)g M.C. Strain

Stress tolerance of diatom resting cells 213 Table 8.2ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǻŗşşŚǼȱ ȱ ȱ ȱ ȱ ȱ ǰȱȱȱȱȱȱȱȱȱ¢ȱȱȱȱȱ ¡ȱȱȱ¢ȱȱȱȱ¢ǯȱȱȱ¡ȱȱȱȱȱ¢ȱȱȱ ȱǯȱǻŗşşŚǼȱ ȱŗǰřŖşǰȱȱ ȱŝŘşȱ ȱȱȱȱȱ¢ǯ

Moisture # of taxa # of taxa in Description category in the list this study 1 never, or only very rarely, occurring outside water bodies 167 3 mainly occurring in water bodies, sometimes on wet 2 148 2 places mainly occurring in water bodies, also rather regularly on 3 282 6 wet and moist places mainly occurring on wet and moist or temporarily dry 4 116 5 places 5 nearly exclusively occurring outside water bodies 16 1

ȱ ǰȱȱȱȱŘŖŖŝǰȱȱ ȱęȱȱ ȱ ȱǭȱȬȱǻŗşŞŞǼǯȱȱǻŘǼȱ ȱȱ ȱȱȱȱȱȱȱȱȱěȱȱǻǼȱ ȱ ȱ ¢ȱ ŘŖŖŞǰȱ ȱ ęȱȱ ȱ ȱ ǭȱ Ȭ ȱǻŗşşŗǼǯȱȱ ȱȱȱȱȱǻ ȱǭȱ £ǰȱŗşŝŘǲȱȱ ȱ ȱȱȱȱǼȱȱ ȱȱȱȱŗŞǚȱƹȱŘǚǰȱŘśȬřŖȱΐȱȱȬŶȱȬŗȱȱ ȱŗŘǱŗŘȱȱǱȱǰȱȱȱ¢ȱ ȱ ȱ ȱȱȱ¡ȱǯ

Experimental set-up ȱ¡ȱȱȱȱěȱȱȱȱ ȱȱ ȱȱȱȱȱȱȱȱ ȱ¡ǱȱǻŗǼȱ ȱǻȱ ȱǼǲȱǻŘǼȱ£ȱȱȬŘŖǚȱȱ ŚŚśǲȱǻřǼȱȱȱśȱǲȱȱǻŚǼȱȱ¢ȱ ȱȱȱƸřŖǚȱ ȱ¢ȱȱȱśȱǯȱȱ ȱȱȱȱȱȱȱȱȱěȱȱet al. ǻŘŖŗŖǼǰȱ¡ȱȱȱ ȱȱȱśȱȱȱ ȱŗŖǯȱȱȱȱȱȱȱȱȱěȱȱ et al.ȱǻŘŖŗŖǼǯȱȱȱȱȱ ȱȱȱȱȱ ¡ȱȱȱȱȱȱŗȱȱȱŜȱ ȱǻȱǼǯȱȱ¡ȱȱȱȱȱȱ ȱȱ ȱȱȱȱȱŘŚȬ ȱȱȱȱ ȱȱȱȱȬ¡ȱǯȱ ȱ ȱȱȱȱ ȱȱȱȱǭȱ ȱ ǻŗşşśǼǰȱ ȱȱȱȱȱȱȱȱȱȱ ȱȱŗŖȱ¢ǰȱȱ ȱȱȱȱȱȱŗŚȱ¢ǯȱȱ

214 Stress tolerance of diatom resting cells ȱȱȱȱ ȱȱ¢ȱȱȱȱȬ ȱȱȱȱȱȱȱȱǻśǚǼȱǯȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱȱ¢ȱȱȱ¡ȱ ǯȱȱȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻǼȱ Ěȱȱ ȱȱĵȱȱ ȱ ȬȱǻȬǼȱ ȱȱĴȱǰȱ ¢ȱ Ŝǰȱ ȱ Śȱ ȱ ȱ ŗǯȱ ȱ ȱ Ěȱȱ ȱ Ŗȱ ȱȱŗśȱȱȬȱ ȱȱȱ¡¢ȱȱȱ ¢ȱǻ¢ȱet al.ǰȱŘŖŖśǼǯȱȱȱȱȱȱ ȱ¢ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ¡ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ŖƽŖǯŖřŖȱ ȱȱ ȱ¢ȱ¡ȱ ȱȱŚȱ¢ǯȱ¡¢ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ Ȭ ȱǰȱȱȱȱȱ¡ȱ ȱȱȱȱ ¢ȱȱŖƽŖǯŖřŖȱȱ ȱȱŘȱȱȱȬȱǯȱȱ ȱȱ ȱȱȱȱȱȱȱȱ ¢ȱȱ ȱȱȱȱ¡ȱȱȱȱȱȱ ¢ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ŘŖŖȱ ȱęȱȱȱȱȱȱȱȱȱ¢ǯȱȱ ¢ȱȱȱ ȱȱȱȱȱȱśǚȱȱ ȱ ȱ ȱȱȱȱȱȱȱǻȱǭȱ ǰȱŗşşśǼǯ ȱ ȱ ȱ ǰȱȱ¡ȱȱȱȱȱ ȱ ¢ȱ ȱ ȱ ȱ ¢ȱ ęȱȱ ȱ ȱ ěȱȱet al.ȱǻŘŖŗŖǼǯȱȱȱěȱȱ¡ȱȱ ȱȱȱȱ¢ȱȱȱȱ¡ȱ ȱǰȱȱ ȱȱ ȱȱȱȱȱȱśǚȱȱ ȱ ȱ ȱ ęȱ£ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ǯȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ǰȱȱȬȱȱ ȱȱȱȱȱ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱŗŞǚǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ¢ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱȱȱǯȱ¢ǰȱȱȱȱ ȱ ȱ ŘŚȬ ȱ ȱ ȱ ęȱȱ ȱ Řȱ ȱ ȱ ȱ ȱ ȱ ȱ¢ȱȱŖƽŖǯŖřŖǯȱȱ ȱȱȱȱȱ ȱȱ ȱ¢ȱȱ ȱȱȱȱȱȱȱ ȱȱȬǰȱȱ ȱȱȱȱ ȱ ȱȱ ȱȱȱǯȱȱ ȱȱȱȱȱ ȱŗŞǚȱȱȱȱȱȱěȱȱȱ ȱȱ ǰȱȱ ȱ¢ȱȱȱȱȱȱȱȱȱ

Stress tolerance of diatom resting cells 215 ȱȱȱȱ ȱȱȱȱȱȱȱ ȱ ȱȱ¢ȱȱȱǰȱȱȱ ȱȱȱǯ ȱ ȱȱȱȱȱȱȱ ȱ ȱȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ¡ȱ ǻŖǼȱ ȱ ȱ ŝȱ ¢ȱ ȱ ȱǻŝǼǯȱȱȱȱȱ ȱŘǰřŖŖȱȱŗřŜǰŖŖŖȱ ȱ ȱ ǰȱ ȱ ȱ ȱ řřǰŖŖŖȱ ȱ ȱ ȱ ȱ ȱ Ŗǯȱ ȱ¢ȱȱǰȱȱ ȱȱȱȱȱȱǰȱ ȱȱȱȱŘŖŖȱȱȱ ǯȱ ȱȱȱȱȱ¢ȱ ȱȱȱȱȱȱ¢ȱȱȱȱȱ ȱȱȱ ȱȱȱȱȱ ǯȱȱȱȱ ȱ ȱ ȱ¢ȱȱȱȱ¡ȱȱȱ ȱ ǯȱȱȱȱȱȱȱŖȱ ȱȱȱŗŖŖȱ¡ȱǽŗȱ ȬȱǻȱȱȱȱȱŶȱȱŝȱȦȱȱȱȱȱȱ ŶȱȱŖǼǾǯȱ

8.3 Results Strain selection ȱȱ ȱ¢ȱȱȱ¢ȱŗȱǻǰȱ ȱ¢ȱ¢ȱ¢ǰȱȱȱ ȱǼǰȱ ȱȱ¢ȱ Řȱǻ¢ȱȱȱ ȱǰȱȱȱ ȱǼǰȱ¡ȱȱ ¢ȱřȱǻ¢ȱȱȱ ȱǰȱȱȱ¢ȱ ȱ ȱȱȱǼǰȱęȱȱȱ¢ȱŚȱǻ¢ȱȱȱ ȱȱȱȱ¢ȱ¢ȱǼȱȱȱȱȱȱ ¢ȱśȱǻ¢ȱ¡¢ȱȱȱ ȱǼǯȱȱ ȱȱ ȱ¢ȱ ȱȱȱȱȱȱȱȱ¢ȱśȱ ȱȱȱȱ ȱȱȱȱȱȱ¢ȱǻŗŜǲȱȱŞǯŗǼǯȱ ǰȱȱȱȱȱȱȱǰȱȱ ȱ ĵȱȱ¡¢ or Pinnularia borealisǰȱ ȱȱȱȱ ¢ǯ

Cytological changes in resting cells and survival of resting cell induction ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱȱȱȱ ȱ ȱȱȱȱȱśǚǰȱȱ ěȱȱ¢¢ȱǻǯȱŞǯŗȱȬǼǯȱȱ¢ȱ ȱȱȱ ȱ ȱ ǰȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱȱ ȱȱȱȱȱȱ ȱȱǻǯȱŞǯŗǼǯȱ

216 Stress tolerance of diatom resting cells Fig. 8.1 ȱ¢ȱȱȱȱȱȱȱŖǯȱǼȱȱ ǻȱŞŞşǼǰȱǯǯƽŗǲȱǼȱȱ¢ȱǻŘǼǰȱǯǯƽŘǲȱǼȱȱ ǻřŞǼǰȱǯǯƽŘǲȱǼȱȱȱǻŘŘȬŗŝǼǰȱǯǯƽřǲȱǼȱĵȱ ȱȱǻŖŘǼŝǰȱ ǯǯƽřǲȱǼȱĵȱ ȱȱǻŗǼǰȱǯǯƽřǲȱǼȱȱȱǻŞǼǰȱǯǯƽŚǲȱ ǼȱȱȱǻŘǼǰȱǯǯƽśǯ

Stress tolerance of diatom resting cells 217 ȱ ȱ ǰȱ ȱ ¢ȱ ǰȱ ȱ ȱ ȱȱǲȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǰȱȱ ȱȱȱ£ȱȱȱȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱ ȱȱȱŗŖŖȱȱȱǰȱȱȱȱ ȱȱȱȱȱȱȱȱȱ ȱȱȱǻȱŞǯŗǼǯ

Tolerance to desiccation ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ Achnanthes coarctataȱǻȱ¢ȱśǼǰȱȱȱȱȱ ȱśȱǰȱ ȱȱ¢ȱȱȱȱȱȱǻǯȱŞǯŘȱ ȬǼǯȱȱȱ ȱȱȱȱȱȱȱȱ ȱȱȱŗȬřǲȱȱȱȱǻȱ¢ȱȱ ȱǼȱǻǯȱŞǯŘȱȬǼǯȱ ȱȱǰȱȱȱȱȱȱ ȱȱȱȱȱȱȱȱŚȱȱ śǰȱ¡ȱȱȱ ȱĵȱȱȱ ȱǰȱȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǻǯȱ ŞǯŘȱ Ǽǯȱ ȱ ȱǰȱ ȱȱȱǰȱȱȱ ĵȱȱ ȱ sp.ȱ ǯȱ ȱ ȱ ȱ ȱ Ȭȱ ȱȱȱȱȱȱȱȱȱ ǯȱ¡¢ ȱP. borealisȱǻȱŞǯŗǼǯȱ ȱǰȱ ǰȱȱȱȱ ȱ ȱ ȱ ǻȱ ¢ȱ ȱ ȱ Ǽȱ ȱ ǰȱȱ¢ȱȱȱȱǻǁŘśƖǼȱȱȱȱ ĵȱȱ ¡¢ȱǻȱŞǯŗǼǯȱȱȱ ȱ¢ȱȱȱ ȱ ȱȱȱǰȱ¡ȱȱ ȱ ǯȱ¡¢ ȱ ȱ ȱȱȱĴȱȱǻȱŞǯŗǼǯ

Tolerance to freezing ȱȱȱȱȱȱȱȱȱ ȱȱȱŗȬřȱȱȱȱ£ȱǻǯȱŞǯŘȱǼǯȱ ȱ ȱǰȱ£ȱ ȱȱ¢ȱȱȱȱȱȱ ȱȱȱȱȱȱȱȱȱȱ ȱŚȱȱśǯȱȱȱǰȱȱȱȱȱȱ ȱȱ£ǰȱȱȱ¡ȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ £ǰȱ ȱ ¡ȱ ȱ ȱ ȱ ĵȱȱ ȱ ǯȱ ȱ ȱ ȱ ¢ȱ ǰȱ ȱȱȱȱ¡ȱA. coarctata ȱ ȱȱȱ

218 Stress tolerance of diatom resting cells ¢ȱȱȱȱȱȱȱ£ȱǻȱȱşřƖǼǯȱ ȱ ȱȱȱ Ȭȱȱȱ£ȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ¡¢ǰȱ P. borealisǰȱ ȱ ȱȱȱA. coarctataȱǻȱŞǯŗǼǯȱȱȱ ȱ ȱ¢ȱȱȱȱȱȱȱȱȱǻȱ ŞǯŗǼǰȱȱȱP. borealisȱȱ¢ȱȱȱȱȱ £ȱȱ ȱȱȱȱȱ¢ȱ ȱȱȱ ȱȱǯȱ

8.4 Discussion ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ŗŝȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ǰȱ ȱ ǰȱ ǯȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱ¢ȱȱȱ ȱȱȱǰȱȱ¢ȱȱȱ ¢ȱ ȱǯȱ¢ǰȱȱȱ¢ȱ ȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ǰȱ ȱ ȱ ȱ¢ȱ ȱȱǯȱǰȱȱȱ ȱ £ȱ Ĵȱȱ ȱ ȱ ǯȱ ¢ǰȱ ¢ȱ ȱ ȱȱȱȱȱȱŚȱǻ¢ȱȱ ȱȱȱȱ ¢ȱ¢ȱǼȱȱśȱǻȱ¢ȱ¡¢ȱȱ ȱ Ǽȱ ȱ ȱ ȱ £ǰȱ ȱ ȱ ȱȱȬęȱȱȱȱȱȱǯ ȱ ȱȱ¢ȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ǰȱ ȱ ¢ȱ ¢ȱ ȱ ǭȱ ȱǻŗşşśǼǰȱȱȱȱȱȱ ȱǯȱ ǰȱȱȱ¢ȱȱȱȱŗŚȱ¢ȱȱȱ ȱȱȱǰȱȱȱȱȱ ȱȱ ȱȱȱ¢ȱȱȱ¢ȱ¢ǯȱ ȱ¢ȱȱȱȱ¢ȱȱȱŗȱȱŘȱ¢ȱȱ ȱȱȱǻȱet al.ǰȱŗşŞŜǲȱ££ȱet al.ǰȱŗşşŖǲȱ ȱ et al.ǰȱ ŘŖŖŞǼǰȱ ¢ȱ Ȭȱ ǻȱ et al.ǰȱ ŗşŞŜǼȱ ȱ ¢ȱȱȱȱȱǻ ¢ȱet al.ǰȱŘŖŗŖǼǯȱ ȱȱȱȱ ȱȱ ȱǻǼȱȱȱȱ¢ȱȱ ȱȱȱȱȱȱȱǯȱȱȱȱ ȱȱ¢ȱ ȱ¢¢ȱ ¢ȱ¢ȱěȱȱȱ ȱȱǰȱȱěȱȱ ȱ¡Ȭȱ

Stress tolerance of diatom resting cells 219 Fig. 8.2ȱ ȱ ȱȱȱƖȱȱȱȱȱȱȱȱ ǻǼȱȱȱȱǻ¢Ǽǰȱȱȱȱęȱȱȱȱȱȱȱ ȱǯȱǻŗşşŚǼǯȱǼȱǲȱǼȱȱ ȱȱȱǲȱǼȱ£ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ɩȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ǯȱȱȱlogarithmic scaleȱȱȱȱƖȱȱȱȱǯ

220 Stress tolerance of diatom resting cells ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ǰȱ ¢ȱ ¢ǰȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ěȱȱ ¡ȱ ȱ ȱ ěȱȱ ¢ȱ ȱ ȱȱȱǰȱ ȱȱȱȱȱȱ ¢ȱȱȱȱ¢ȱȱȱȱȱĴȱ¢ȱȱ ȱȱȱȱȱȱȱȱȱȱǯȱ ǰȱȱȱěȱȱȱ¢ȱȱ ȱ ȱȱȱǰȱȱ ȱȱȱ£ȱ ȱȱȱȱȱǰȱ ȱǯ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ£ǯȱ¢ȱȱȱȱȱAchnanthes coarctataȱȱ ȱǻ ȱȱȱǼȱȱȱǰȱȱ ȱȱȱ¡ȱȱȱȱŚȱȱśȱȱȱ £ȱȱȱȱ ȱ ȱȱȱȱǯȱ £ȱȱȱȱȱȱȱȱȱȱȱȱ ȱȱȱ¢ȱěȱȱet al.ȱǻŘŖŗŖǼǯȱȱ ȱȱ ȱȱȱȱȱǻȱǼȱȱŚȱȱȱŜȱȱ ȱȱȱ¢ȱȱȱǯȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻP. borealisȱ ǻŞǼȱ ȱ ǻŞǼǼȱ ȱ ȱ ȱ ȱ ȱ ¢ǰȱ ǯȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱȱ£ȱ

Stress tolerance of diatom resting cells 221 ȱȱȱȱȱǯ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱǻȱ ¢ȱȱȱǼȱȱȱȱȱ¡ȱȱȱ ȱȱǰȱŚȱǻśǼȱ¡ȱȱȱŜȱȱȱȱȱȱ ǻȱ¡Ǽǯȱȱ ȱȱȱȱ¡ȱȱȱȱ£ǰȱȱ ȱȱȱęȱȱȱ¡ȱ ȱȱȱȱȱǯȱ ǰȱ ȱ¢ȱȱȱȱȱȱȱ£ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱȱȱȱȱȱȱȱȱ¢ȱśǰȱA. coarctataȱȱ ȱȱȱȱ£ȱȱȱǰȱ ȱ ȱ ¢ȱ ȱ £ȱ ȱ ǰȱ ȱ ȱ şŜƖǯȱ ȱȱȱǰȱȱP. borealisǰȱȱǯȱȱȱȱ £ȱȱȱȱȱȱǻǼȱ ȱȱȱǰȱ ȱȱ¢ȱȱȱȱěȱȱ¢ȱȱȱ £ǯȱȱȱȱȱȱȱȱȱ Ȧȱ £ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱ ŗşŝśǲȱ ǰȱ ŘŖŖşǼȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ǻ et al.ǰȱŗşŝşǲȱȱǭȱǰȱŘŖŖŖǲȱ ȱet al.ǰȱŘŖŖřǼǰȱ ¢£ȱȱǻȱǭȱǰȱŗşŝŚǼǰȱ£ȱȱ ȱǻ ȱǭȱǰȱŗşşśǲȱ¢ǰȱŗşşŜǼȱȱȱ¢ȱǻûȱ et al.ǰȱŘŖŗŖǼǯȱȱ ȱȱȱȱȱȱȱ¢¢ȱ ȱ¢ȱȱȱ ȱȱȱ¢ǰȱ ȱȱȱȱ ȱȱȱȱǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱȱ¢ȱ ǰȱȱȱ ȱȱȱȱȱ ȱ ȱȱȱȱ ȱȱȱ ȱȱǯȱ ȱȱǰȱȱȱȱȱȱ ȱȱȱȱśƖȱȱȱȱȱ ȱȱŖǯŖŖŗƖǯȱȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱǯȱȱ ȱȱȱȱ ȱȱȱȱȱȱȱȱȱȱȱ¡ȱ ȱ¢ȱǻ ȱǭȱ ǰȱŘŖŖŜǲȱȱǭȱǰȱ ŘŖŖŜǼȱȱȱȱȱǻȱet al.ǰȱŘŖŖŞǼǯȱȱ£ȱȱ ȱȱȱ¢ȱȱȱȱȱȱȱȱȱ ȱȱȱȱǰȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ęȱȱ ěȱȱ ȱ ȱȱȱȱȱ ȱȱȱȱȱ

222 Stress tolerance of diatom resting cells ȱǻĴȱ ȱet al.ǰȱŘŖŖśǲȱȱǭȱǰȱŘŖŖŜǲȱ ¢ȱet al.ǰȱ ŘŖŖŝǼǯȱ ȱȱȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ Ȭ ǯȱ ¢ȱ ȱ ȱȱȱȱȱ¡ȱǻȱet al.ǰȱŘŖŖśǲȱȱ ǭȱǰȱŘŖŖŝǲȱȱet al.ǰȱŘŖŖŞǼǰȱȱP. borealisȱȱ ǯȱ¡¢ȱǻěȱȱet al.ǰȱǯȱǼǯȱȱȱȱȱȱȱ¢ȱ ȱȱ ȱ¡ȱȱȱęȱȱȱȱęȱȱ ǯȱ ȱȱęȱȱǰȱȱȱȱȱȱȱȱ ȱȱȱ ȱȱȱǰȱȱȱȱȱȱ ȱȱȱȱȱȱȱȱȱ ȱǯȱ ȱȱȱǰȱȱȱȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ęȱȱ ȱ ȱęȱȱȱȱȱǯȱȱȱȱȱȱ ȱȱȱȱ¢ȱȱȱęȱǰȱȱ ȱȱȱȱȱȱȱȱȱȱ ȱ ȱȱȱȱ ĵȱȱǯȱȱ ǯȱ¡¢ȱǽǻŗǼȱ ȱǻřǼǾȱȱȱȱȱȱǻěȱȱet al.ǰȱǯȱǼȱ ȱȱȱ£ȱȱȱȱȱȱP. borealis ȱȱȱȱȱǽǻŗǼǰȱǻŗǼȱȱǻŜǼǾǯȱȱȱ ȱȱǯȱ¢ǰȱȱ ȱȱȬěȱȱȱȱȱ ȱȱȱ£ȱǰȱ¢ȱȱȱ ȱȱȱȱȱȱȱȱǻȱȱǼȱ ȱȱǻǰȱŘŖŖŖǲȱ et al.ǰȱŘŖŖŜǼǯ ȱ ȱȱȱȱȱȱȱȱȱȱȱȱ ȱǰȱ¢ȱȱȱȱȱȱ¡ȱȱȱ ȱŚȱȱśȱǻ ȱȱȱet al.ǰȱŗşşŚǼǰȱȱ¢ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ¢ȱ ȱ ȱ ¢ȱ ¡¢ȱ ȱ ȱ ǯȱ ȱ ȱ ¡ȱ ȱ ¢ȱ Śǰȱ ĵȱȱ ǰȱ ȱ ȱ ȱ ¢ȱ ǰȱ ȱ ȱ ȱ ȱ ȱȱȱȱ¡ȱǯȱ ȱȱȱȱǻǼ ¢ȱ ȱ ȱ ȱ ǻȱ et al.ǰȱ ŘŖŖśǲȱ ȱ ǭȱ ǰȱ ŘŖŖŝǲȱ ȱ et al.ǰȱ ŘŖŖŞǼȱ ȱ ȱ ¢ȱ ȱ ȱ ěȱȱ ȱ¢ȱȱȱǻȱet al.ǰȱ ŘŖŖşǼǰȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ et al.ȱ ǻŗşşŚǼȱ ȱ ȱ ȱ¢ǯȱȱȱȱ¢ȱȱȱȱȱȱN. ȱȱ ȱȱȱȱȱǯȱǰȱ ȱȱȱȱȱȱȱȱ£ȱȱ¢ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱęȱȱȱȱ¢ȱ

Stress tolerance of diatom resting cells 223 ȱǯȱȱȱȱ ȱȱȂȱ ȱȱȱ¢ȱȱȱȱȱȱǰȱ ȱ ȱȱȱ ȱ£ȱȱȱǯ ȱ ȱ ȱȱȱȱȱȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱȱȱȱĴȱȱȱǰȱȱ¢ȱ ȱ ȱ ȱ ǻȱ et al.ǰȱ ŘŖŖśǼȱ ȱ ȱ ȱ ǻȱ ǭȱ ǰȱ ŘŖŗŖǼǰȱ ȱ ȱ ȱ ¢ȱȱȱȱȱȱ ȱȱȱȱ¢ȱȱȱ ȱǻ ǰȱŗşşŜǼǯȱȱȱǻǼȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ǻȱ Ǽȱ ȱ ȱ ǻ ȱ et al.ǰȱ ŗşŜŚǼȱȱȱȱȱȱȱȱȱǻǰȱŗşśŞǰȱ ŗşśşǼȱȱȱǰȱ ȱȱȱ¢ȱȱȱȱ ȱȱȱȱȱ¡ȱȱȱȱȱǻȱǰȱ ŗşřŝǲȱ ǰȱ ŗşŜŗǲȱ ȱ et al.ǰȱ ŗşŜŚǲȱ ǰȱ ŗşŜŚǲȱ ¢Ȭ ȱ ǭȱ ǰȱ ŗşşřǼǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ¢ȱ ȱ ȱ¢ȱȱǯȱȱȱȱȱȱ ȱȱȱȱȱȱ ȱěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱet al. ǻŘŖŗŖǼȱ¢£ȱȱȱ ȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱ ȱ ǯȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱȱǽȱȱȱ ȱȱȱȱȱȱȱȱȱǻǰȱ ŘŖŖŞǼǾȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ¢ȱȱȱȱȱĴȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱȱ ȱȱȱȱȱǯȱ

224 Stress tolerance of diatom resting cells Acknowledgments ȱȱ ȱęȱ¢ȱȱ¢ȱȱȱȱęȱȱ ȱ Ȯȱ ȱ ǻȬǰȱ ǰȱ ȱ ȱ ȱ ȱ Ǽǰȱ ȱ Ȭȱ ř ȬŖśřřȬŖŝȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ Ȭȱ ř ȬŖŚŗşȬŖŞȱȱȱǯ

Stress tolerance of diatom resting cells 225

9. General discussion

9.1 Diatom biogeography 9.1.1 Geographical distribution patterns Given the controversial importance of historical factors for microbial distributions (reviewed in Martiny et al., 2006), several recent studies ȱ ȱ ȱ ȱ ȱ ȱ Ĵȱȱ ȱ disentangle the relative importance of historical factors and local factors on freshwater benthic diatom community structure and diversity (reviewed in Vanormelingen et al., 2008b). These showed that, on large spatial scales (hundreds to thousands of km), distance explains a substantial fraction of the variation in local community structure independent of environmental factors (Potapova & Charles, 2002; Soininen et al., 2004; Verleyen et al., 2009). This is mainly due to species with restricted distributions which are not readily explainable by environmental limitations. Moreover, local diversity is limited by habitat availability (Telford et al., 2006), and even regional genus diversity can be strongly limited in isolated regions, such as those found at high latitudes of the Southern hemisphere (Vyverman et al., 2007). This indicates that dispersal limitation is an important ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ǰȱ from incomplete local recruitment from the regional species pool to restrictions on geographical species ranges, and shows that reports ȱȱȱȱ¢ȱȱĴȱȱȱȱ¢ȱ restricted occurrence of suitable habitat. In this thesis we gathered further evidence for the importance of historical factors for diatom distributions and the occurrence of endemic species in the (sub-)Antarctic based on morphological surveys. 1. First, we show that there is a high level of endemicity in lacustrine diatoms (27-63%, chapter 2), similar to the levels found for macroorganisms in the (sub-)Antarctic (Allegrucci et al., 2006; Stevens et al., 2006; Convey et al., 2008). Moreover, ȱ ęȱȱ ȱ ¢ȱ ȱ ȱ ȱ

General discussion 227 using molecular phylogenies in two of the cosmopolitan morphospecies (chapter 6) suggests that the percentage of (sub-)Antarctic endemic species is in fact (much) higher than ȱ Ȭȱ ęȱǯȱ ȱ ȱ ȱ ȱ regions (the sub-, maritime and continental Antarctic, Chown & Convey, 2007) can be distinguished for lacustrine diatoms (chapter 2). Together, these strikingly congruent ȱĴȱȱ ȱȱȱȱ ȱȱȱȱęȱȱȱȱȱ the (sub-) Antarctic operate at similar spatial and temporal scales as in macroscopic organisms. 2. Secondly, regional lacustrine diatom diversity in the Southern hemisphere is not only reduced (Vyverman et al., 2007) but is also imbalanced with an overrepresentation of predominantly terrestrial lineages and a general poverty of globally successful genera and planktonic lineages (chapter 2). This can only be explained by the presence of historical limitations in diatom lineage dispersal into the (sub-)Antarctic. 3. Finally, the low community similarity for both lacustrine and terrestrial diatoms among the sub-Antarctic island groups Crozet and Kerguelen as compared to the within-island similarities (chapter 3) indicate the presence of a geographic signal in diatom distributions, even within the sub-Antarctic. This is congruent with previous reports on species endemic to single islands or groups of islands in the sub-Antarctic (Van de Vijver et al., 2002; Van de Vijver et al., 2004; Van de Vijver & Mataloni, 2008; Van de Vijver et al., 2010), but it is still unclear ȱȱ¢ȱěȱȱȱȱȱȱ ěȱȱ ȱȱ ȱȱȱȱǰȱ or to other factors including dispersal limitation and priority ěȱǯ

9.1.2 Terrestrial vs. lacustrine diatoms As mentioned in the introduction, we hypothesized that terrestrial diatoms have wider geographical ranges compared to their aquatic counterparts resulting from higher dispersal probabilities due to the less fragmented terrestrial habitat (Spaulding et al., 2010), a higher probability of being picked up by wind, and their assumed higher tolerance for desiccation and temperature extremes than aquatic taxa (Schlichting, 1969; Ehresmann & Hatch, 1975), which was ęȱȱ¢ȱȱ¡ȱǻchapters 7 and 8). In line with our

228 General discussion expectations, community similarity analyses on a dataset covering the sub-Antarctic island groups Crozet and Kerguelen indicated that terrestrial diatom communities were more similar within and between islands than the lacustrine communities on an intermediate scale (ca. 1,500 km) (chapter 3), indicating higher dispersal rates within and between islands for terrestrial taxa. However, while these two islands are similar in terms of climate and geology, we ȱ ȱ ¡¢ȱ ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱȱȱĴȱǰȱȱȱȱȱȱȱȱ higher community similarities found for terrestrial communities were a result of higher dispersal, wider realized niches for terrestrial species or less environmental variability in the terrestrial habitat. ȱȱȱȱĴȱȱȱ¢ȱȱȱȱ organisms show less distance decay than those of aquatic organisms, ȱ ȱ ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ȱ ęȱȱ of terrestrial habitats (Soininen et al., 2007). However, also higher dispersal capacities and wider realized niches could contribute to ȱ Ĵȱǰȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ Ěȱȱ of environmental and spatial variables are needed to fully interpret the observed decreases in community similarity with distance (Legendre et al., 2005; Soininen et al., 2007). At the same time, even ȱȱȱ ȱȱȱĴȱȱȱȱ spatial scales. First, terrestrial community similarities decreased ęȱ¢ȱȱȱȱ ȱȱȱȱȱȱ (chapter 3). Secondly, the Antarctic strains of Pinnularia borealis and ĵȱȱ¡¢ȱ ȱ¢ȱȱ¢ȱěȱȱ from the non-Antarctic strains (chapter 6). While lacustrine and ȱȱȱȱȱȱȱěȱȱȱ ĴȱȱȱȱȱȱǻȱȱŗǰśŖŖȱǼǰȱ¢ȱȱȱ ȱȱǯȱȱĴȱȱȱȱȱȱȱȱȱ probabilities and stress tolerance further below in section 9.3.

9.2 Species diversity and speciation in the (sub-)Antarctic (Pseudo)cryptic species diversity is widespread in diatoms, with sometimes tens of species discovered in a single species complex (Behnke et al., 2004; Kooistra et al., 2005; Sarno et al., 2005; Beszteri et al., 2007; Mann & Evans, 2007; Vanormelingen et al., 2008a; Evans et al., 2009; Trobajo et al., 2009; Vanelslander et al.ǰȱŘŖŖşǼǰȱȱęȱȱȱ predictions of Mann & Droop (1996). Almost at the same time as these

General discussion 229 ęȱǰȱȱȱȱĴȱȱȱȱȱĴȱȱ to phenotypic plasticity has reversed and thousands of new species have been described including lots of potential endemics, many of which would previously be regarded to be part of a morphological continuum within species (reviewed in Vanormelingen et al., 2008b). The current study on P. borealis and ǯȱ¡¢ (chapter 6) shows that species diversity is still underestimated even when applying ȱ ęȱȬȱ ȱ ȱ ȱ ǻcf. chapter 2). The P. borealis lineage has an estimated age of no less than 42 Ma, and ȱ ȱ ȱ ȱ Ĵȱȱ ȱ ȱ ȱ ŘŘȱ ȱ ǰȱ making it the oldest known cryptic diatom species complex known to date. Estimated lineage divergence times for the marine diatom Ȭĵȱȱ were 0.5-0.9 Ma (Casteleyn et al., 2010) and the age of the ȱȬ complex is at least 12 Ma (Evans et al., 2008). The high species diversity encountered calls for studies into diatom speciation. It is currently not well known what the contribution of various historical and ecological processes is for ȱ ęȱǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ǯȱ ȱȱȱǰȱȱȱȱȱȱȱęȱ¢ȱ higher synonymous mutation rates compared to the less diverse families based on phylogenetic analyses, indicating that there could ȱȱȱȱ ȱȱȱȱȱęȱȱǻi.e. speciation minus extinction) (Lanfear et al.ǰȱŘŖŗŖǼǯȱȱȱ¢ȱ on polyploidy speciation in a marine planktonic diatom highlighted the potential importance of polyploidization (Koester et al., 2010), known to be involved in 15% of plant speciation and 31% of fern speciation events (Wood et al., 2009). The high percentage of diatom species endemic to (parts of) the (sub-)Antarctic suggests that there is ample opportunity for allopatric speciation in the region (chapters 2 and 6). Molecular phylogenies are not only useful for (cryptic) species delimitation but ǰȱ ȱĜȱȱ¢ȱȱȱȬȱȱȱǰȱ ȱȱȱȱȱ¢ǰȱȱȱȱȱȱĴȱǯȱ When combined with data on ecological niches and geographical distributions, they allow to reconstruct the evolutionary history of taxa and thus get an idea of the driving (historical and ecological) forces behind lineage splits (see ǯǯ Darling et al., 2007; Verbruggen et al., 2009; Casteleyn et al., 2010). Darling et al. (2007), for instance, reconstructed the evolutionary history of the polar planktonic foraminifera species complex ȱ¢ sinistral using time-calibrated molecular phylogenies and geographic

230 General discussion ȱ ȱ ȱ ěȱȱ ǯȱ ȱ ȱ ȱ ȱ ȱȱęȱȱȱ ȱȱȱȱȱ Atlantic Arctic and Antarctic populations 1.5 - 1.8 Ma ago. Further ęȱȱȱȱȱ ǰȱȱȱȱ of an extreme polar lineage capable of living in the sea ice and a currently isolated lineage in the Benguela upwelling system, could be linked to glacial-interglacial climate dynamics. Together with other such studies, a rather complete picture is emerging on the biogeography of planktonic foraminifers (reviewed in Darling & Wade, 2008). While it is clear that we are still far from achieving such a goal for freshwater or terrestrial diatoms in general and more ęȱ¢ȱ ȱ ȱ ǻȬǼǰȱ ȱ ȱ ȱ ȱ been made towards achieving this goal with this PhD-thesis. These are (1) the reconstruction of a time-calibrated molecular phylogeny of the globally important diatom genus Pinnularia, allowing to date lineage splits for any sampled taxon in this genus (chapter 4 and chapter 5), (2) the construction of a molecular phylogeny for the globally distributed diatom Pinnularia borealis (chapter 6) showing the presence of a number of distinct lineages including a continental Antarctic lineage with an estimated split from a northern temperate lineage 7.7 Ma ago, remarkably similar to divergence times of Antarctic lineages of some other organism groups (Stevens et al., 2006; Mortimer et al., 2011), and (3) an assessment of temperature preferences of these lineages showing that Antarctic P. borealis has a preference for relatively low temperatures but not to an extreme extent, similar to the Antarctic lineage of ǯȱ ¡¢ (chapter 6Ǽǯȱȱěȱȱȱȱȱȱȱ¢ȱȱȱ geographic areas (especially in the (sub-)Antarctic), at the same time assessing the geographic distribution of these lineages, and further niche characterization. Together, this will allow a reconstruction of the evolutionary history of P. borealis and potentially other Pinnularia lineages in the Southern Hemisphere, including a view ȱȱȱȱȱ ȱȱęȱǯ

9.3 Stress tolerance and consequences for dispersal and biogeography ȱěȱ¢ȱȱȱ£ȱȱ ȱǰȱȱȱ to survive the transport. The survival probability of dispersal depends on the tolerance levels of the species and individual cells

General discussion 231 for the adverse conditions encountered during transport. This critical point in dispersal is not easy to quantify. First, the conditions ȱȱĚȱȱȱȱȱȱȱȱȱ¢ȱ several factors (Isard & Gage, 2001). For example, desiccation rates during external transport by birds or by wind will depend on the relative humidity and temperature of the air, the wind rate and solar incidence, next to the presence of mud around the cells and the presence of protecting grooves in the feet epidermis of ducks. It is therefore not easy to set up experimental conditions that cover the range of possible conditions, while at the same time covering a whole range of taxa. Second, to quantify propagule viability in natural transport conditions, knowledge on starting densities are ȱ ǰȱ ȱ ȱ ǰȱ ¢ȱ ȱ ¢ȱ ȱ ěȱȱ ȱ ǯȱ ǰȱ ȱ ȱ ęȱȱ ȱ ȱ viabilities are equally important as the viability level itself, it is important to assess the tolerance of as many species as possible. In our experiments (chapters 7 and 8), we decided to put more weight on the number of taxa instead of the number of conditions, mainly ȱ ȱ ȱ ¢ȱ Ĝȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱęȱȱȱęȱȱȱ ȱȱȱ detected using a low number of stress conditions, and because our ȱȱ ȱȱȱěȱȱȱȱȱ ȱ terrestrial and aquatic species. Two main results came out of our stress tolerance experiments on lacustrine and terrestrial diatoms. First, vegetative cells are very sensitive for desiccation and extreme temperatures (chapter 7). Second, also resting stages are very sensitive to these stresses, except the resting stages of typical terrestrial diatoms which can survive short-term (5 minutes) desiccation, albeit generally with low survival percentages (chapter 8). This has important consequences for diatom dispersal rates, and consequently also for population structures, speciation and geographical distributions. First, freshwater diatoms inhabit aquatic “islands” in an “ocean of inhospitable land”, and dispersal among such “islands” might not be very common given the general low desiccation tolerance of both vegetative and resting diatom cells and the resulting extremely low probabilities of transport of viable cells by air. Even if the conditions encountered ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ¡ȱ conditions we used (chapters 7 and 8), desiccation is known to be a key factor during both wind and bird transport (Schlichting, 1960; Ehresmann & Hatch, 1975; Kristiansen, 1996; Figuerola & Green,

232 General discussion 2002). The absence of resistant dispersal stages in aquatic diatoms is in agreement with the single microsatellite study on freshwater diatoms (and even microalgae), showing that populations separated ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ¢ȱ ȱ microsatellite allele frequencies, indicative of a very low connectivity between populations (Evans et al., 2009). Current microsatellite studies on freshwater diatoms are directed at pinpointing at exactly how small the spatial scale is at which connectivity breaks down (pers. comm. P. Vanormelingen). Interestingly, marine diatom populations ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ (Casteleyn et al., 2009; Casteleyn et al., 2010; Godhe & Harnström, 2010), but additional studies are necessary to assess the generality of ȱĴȱǯȱǰȱȱȱěȱȱȱȱȱ between aquatic and terrestrial diatoms it can be hypothesized that ȱȱȱ ȱȱȱěȱȱ than aquatic diatoms at the same spatial scale. Secondly, dessication and freezing tolerant resting stages were only present in terrestrial and not in aquatic taxa (chapter 8) ȱȱȱȱęȱ¢ȱȱȱȱǯȱȱ Ĵȱȱȱȱȱȱȱ¡ȱ ȱȱȱȱȱ representatives, such as in green algae (Zoe et al., 2008). Furthermore, the tolerance of these terrestrial resting stages to desiccation will enhance their survival probabilities of dispersal by animals and ȱȱǰȱȱȱȱ¡ȱȱĴȱȱȱȱ dispersal by wind than aquatic taxa. It should be noted however that it is not clear how desiccation survival depends on the desiccation period, since cultures were only desiccated for 5 minutes, which is of course critical for determining the distances over which dispersal can take place. In any case, the tolerance of only terrestrial diatoms ȱȱȱ ȱȱęȱȱȱ¢ȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻȱ ǰȱ ŗşřŝǲȱ Schlichting, 1961; Brown et al.ǰȱŗşŜŚǲȱ¢ȬȱǭȱǰȱŗşşřǼǰȱ albeit this is most likely a combination of the higher susceptibility of terrestrial material to get suspended in the air, and the desiccation survival of terrestrial diatom resting stages. The observed higher presence of terrestrial diatoms in the air also corresponds to the observation that terrestrial diatom community similarities might decrease less over intermediate distance (within islands and up to 1,500 km) compared to the aquatic communities (chapter 3, Soininen et al.ǰȱŘŖŖŝǼǯȱ ǰȱȱȱĴȱȱȱȱȱȱ of the combination of three mutually non-exclusive factors: the less

General discussion 233 fragmented terrestrial habitat, the higher susceptibility of terrestrial organisms to get picked up by air currents or to disperse over land, and the adaptation of terrestrial organisms to a terrestrial life-style with its requirements of tolerance to at least desiccation. Thirdly, the desiccation tolerance of terrestrial diatom resting cells (chapter 8) emphasizes the importance of resting stages for dispersal. Resting stages are widely known to be the dominant dispersal stages in several organism groups (Hutchinson, 1967; Levins, 1969; Sutherland et al., 1979; Vleeshouwers et al., 1995; Nicholson et al.ǰȱ ŘŖŖŖǲȱ ǰȱ ŘŖŗŖǼȱ ȱ ȱ ȱ ęȱȱ adaptations enhancing dispersal, such as wings or sticky hairs on the seeds of higher plants, a protective outer ectocarp, or the thick-walled protective ephippium enclosing the resting eggs of zooplankton. Besides for dispersal through space, resting stages are also very important as “seed bank” to overcome adverse conditions in time (Vleeshouwers et al., 1995; Figuerola et al., 2003; Poulícková et al.ǰȱŘŖŖŞǲȱ ȱǭȱǰȱŘŖŗŖǼǰȱȱ¢ȱȱęȱȱȱ¢ȱ role in the long-term persistence of populations and species (Jones & Lennon, 2010). For marine and freshwater planktonic diatom populations the importance of resting cells and spores to overcome seasonally unfavorable light, nutrient and temperature conditions is well-established (Gran, 1912; Hargraves & French, 1975, 1983; McQuoid & Hobson, 1995, 1996). Terrestrial habitats are subjected to ȱ¢ȱȱȱĚȱȱȱȱȱ¢ȱ (Starks et al., 1981; Gao et al., 2008), and the overall higher stress tolerance of terrestrial diatom resting cells compared to vegetative cells (chapter 8) indicates that they are important for population persistence under unfavorable conditions. However, there are no data on the population dynamics of terrestrial diatoms, and it is therefore not known whether terrestrial diatom populations show a metapopulation structure with frequent extinctions and recolonizations from other populations in the metapopulation once the habitat patch becomes suitable again, or whether a local seed bank guarantees long-term population survival. The presence of resistant resting stages, however, does indicate that both a metapopulation structure with regular exchange between patches and the presence of a local seed bank are possible.

234 General discussion 9.4 Biogeographical patterns of terrestrial and aquatic diatoms In line with their higher tolerance to desiccation compared to aquatic taxa (chapter 8Ǽǰȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ of diatom communities in aquatic and terrestrial habitats indicate that terrestrial diatom communities tend to have a lower decrease in community similarity over distances up to 1,500 km (chapter 3), ȱ ȱ ęȱ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ than aquatic communities (chapter 3). At larger spatial scales this community similarity decreases strongly in a way similar to the aquatic communities (chapter 3). We therefore hypothesize that ȱȱȱĴȱȱȱȱȱȱȱ shorter distances due to a certain degree of desiccation tolerance (chapter 8), a higher susceptibility for wind dispersal and a less fragmented terrestrial habitat on a within-island scale, but that this ěȱȱ ȱȱȱȱȱȱȱȱ distances. This might be due to the fact that such long-distance dispersal is extremely rare also for terrestrial diatoms, resulting in very large community dissimilarities. However, further research ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱǯȱȱěȱȱȱȱȱ seem similar for both types of habitats within and between the sub- Antarctic islands studied, it can’t be excluded that terrestrial habitats ȱȱȱěȱȱȱȱȱȱȱ lakes, or that terrestrial diatoms have wider niches.

9.5 studying dispersal During this thesis, we gathered data indicating that diatom distributions are similar to those recorded for macroorganisms (chapter 2), that presumed cosmopolitan diatom species may not be cosmopolitan after all (chapter 6), and that both vegetative and resting cells of diatoms are very sensitive to desiccation (chapters 7 and 8). However, while the initial subject of this thesis was “dispersal” of diatoms, we did not estimate dispersal rates between natural populations. We therefore present three more or less feasible approaches to further study natural diatom dispersal, each with its strengths and weaknesses. Ideally, they are combined to provide a complete view on dispersal.

General discussion 235 1. Finding dispersing cells. By sampling and quantifying the ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ as animals and air currents, we can further improve our knowledge on dispersed taxa and abundances. Previous knowedge has been summarized in section 1.4.2, but we still lack quantitative information on species-level. It should be stressed that directly oxidizing the transported diatom material renders it useless since it doesn’t allow to distinguish living cells from dead frustules, which are well-known to be easily taken up by air and water currents and deposited away from their origin. Therefore, culture-based methods or decent ęȱ¡ȱȱȱǯȱȱȱȱȱȱȱ ȱȱȱȱȱěȱȱȱ ȱȱ ȱǰȱȱȱĜȱȱȱȱȱȱȱȱȱȬ distance dispersal events which might be responsible for the colonization of new habitats. 2. Colonization and transplant experiments. The only way to prove (but not disprove) dispersal limitation in diatoms is a “transplant” experiment. This involves placing diatoms from one area to a geographically separated, ecologically similar, area and assess if the transplanted diatoms can establish populations in that new area, which would imply that the only factor previously preventing those diatoms from occurring in the other area was dispersal limitation. While this is ethically ȱ ¢ȱ ęȱǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱȱȱ ȱȱȱȱĴȱȱǰȱȱ ȱȱĴȱȱȱȱȱȱǽȱǯǯ the transfer of organisms with ballast water of oceanic ships, or the transfer ȱ ȱ ¢ȱ ȱ ęȱȱ ǻ ěȱȱ ǭȱ ǰȱ 1992; Klein et al. 2010)]. In a colonization experiment, a new habitat is created and the build-up of diversity is followed. This way, it can be determined how long it takes before new communities are saturated with species, and thus how long dispersal limitation plays a role in these communities. It should be noted however that decent control communities are necessary in which dispersal limitation is prevented by ęȱȱǯ 3. ȱ ȱ ȱ ěȱǯȱ ěȱȱ ȱ populations in allele frequencies for neutral molecular markers (ǯǯ microsatellites) can be used to infer the current level of ȱĚȱ ȱ ȱȱǯȱȱȱǰȱ ȱȱ

236 General discussion ȱ ȱȱȱȱĚȱ ȱȱǰȱȱȱȱ ěȱȱ ȱȱȱ¢ȱȱȱ ȱȱȱěȱȱȱȱ ȱȱ ȱǻ Ĝȱȱȱet al., 1996). For such interpretations to be made, it is absolutely necessary that neutral markers are used and that these are not linked to genes under selection (which would be the case when more than one, cryptic, species is included in the study). In such a case, natural selection (species sorting) will contribute to the observed “population” ěȱȱȱȱȱȱȱȱȱȱ ǯȱȱȱȱ¡ȱȱȱȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱǻȱȱ ȱȱȱȱĚȱ ȱȱȱ¢ȱ reached), which may be quite common in various freshwater organisms (see De Meester et al., 2002). However, when one ȱ Ĝȱȱ¢ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ highly valuable method to get an idea of the actual dispersal rates and dispersal scales of diatoms.

9.6 Final thoughts ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ biogeographies are shaped by both species sorting and historical factors, many questions remain. First of all, as stated by Green & Bohannan (2006) we can shift from the question “is microbial ¢ȱ¢ȱěȱ¢ȱȱȱȱȱ macroorganisms”, towards the question “on which spatial scale and using which taxonomical resolution microbial biogeography ȱ ȱ ȱ Ȅǯȱ ęȱȱ ȱ ȱ turnover will thus be important in the future, and this also applies for diatom research. Local abundances of (living) terrestrial diatoms and seasonal dynamics of population density are unknown. Dispersal rates and scales, and speciation rates and mechanisms are all needed to infer the threshold levels at which speciation will occur between populations. Microsatellite studies and the frequencies ȱ ȱ ȱ ȱ ȱ ěȱȱ ęȱȱ ȱ ȱ ěȱȱȱȱ ȱȱȱȱȱȱ ȱ dispersal occurs, provided that a correct single genetic lineage is used (i.e. a single species) and that distinctions are made within species ¡ȱ ǽȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ¢ȱ (Rynearson et al., 2009)]. While only a single microsatellite

General discussion 237 ¢ȱȱȱȱȱ ȱȱǽȱ (Evans et al., 2009)] and a second and third are under way (pers. comm. P. Vanormelingen), terrestrial diatoms remain unexplored. Compared to aquatic diatoms, much less is known about the ecology of terrestrial diatoms, but their life style and population dynamics are as much intriguing. Several questions concerning terrestrial diatoms have been put forward in the previous sections. ȬȱȱȱȱȱĴȱȱȱȱȱ of terrestrial diatoms to overcome adverse conditions and the mechanisms they use, besides the importance of resting stages. Despite the fact that terrestrial diatoms have been thriving under our feet since we evolved, they are in many perspectives still a closed book.

238 General discussion References

Abbott, R.J., Smith, L.C., Milne, R.I., Crawford, R.M.M., Wolff, K. & Balfour, J. (2000) MAbbott, R.J., Smith, L.C., Milne, R.I., Crawford, R.M.M., Wolff, K. & Balfour, J. (2000) Molecular analysis of plant migration and refugia in the Arctic. Science, 289, 1343-1346 Abbott, R.J. & Brochmann, C. (2003) History and evolution of the Arctic ÀRUDLQWKHIRRWVWHSVRI(ULF+XOWHQ0ROHFXODU(FRORJ\, 12, 299-313 Admiraal, W., Peletier, H. & Brouwer, T. (1984) The seasonal succession SDWWHUQVRIGLDWRPVSHFLHVRQDQLQWHUWLGDOPXGÀDWDQH[SHULPHQWDODQDO\VLV Oikos, 42, 30-40 Allegrucci, G., Carchini, G., Todisco, V., Convey, P. & Sbordoni, V. (2006) A molecular phylogeny of Antarctic Chironomidae and its implications for biogeographical history. Polar Biology, 29, 320-326 $OYH( *ROGVWHLQ67 'LVSHUVDOVXUYLYDODQGGHOD\HGJURZWKRI benthic foraminiferal propagules. Journal of Sea Research, 63, 36-51 $OYHUVRQ$-&DQQRQH--*XWHOO55 7KHULRW(& 7KH evolution of elongate shape in diatoms. Journal of Phycology, 42, 655-668 Alverson, A.J. (2008) Molecular systematics and the diatom species. Protist, 159, 339-353 Anderson, O.R. (1975) Ultrastructure and cytochemistry of resting cell formation in Amphora coffaeformis (Bacillariophyceae). Journal of Phycology, 11, 272-281 Anderson, O.R. (2010) Field and laboratory studies of encysted and trophic stages of naked amoebae, including a perspective on population life cycle dynamics. Acta Protozoologica, 49, 1-8 Arrhenius, O. (1921) Species and area. -RXUQDORI(FRORJ\, 9, 95-99 $WNLQVRQ.0 )XUWKHUH[SHULPHQWVLQGLVSHUVDORISK\WRSODQNWRQE\ birds. Wildfowl, 22, 98-99 $WNLQVRQ.0 ([SHULPHQWVLQGLVSHUVDORISK\WRSODQNWRQE\GXFNV British Phycological Journal, 15, 49-58 Atkinson, K.M. (1988) The initial development of net phytoplankton in Con *UHHQ5HVHUYRLU 8SSHU7HHVGDOH DQHZLPSRXPHQWLQ1RWKHUQ(QJODQGAlgae and the aquatic environment HGE\)(5RXQG SS%LRSUHVV%ULVWRO Avise, J.C. (2000) 3K\ORJHRJUDSK\WKHKLVWRU\DQGIRUPDWLRQRIVSHFLHV. Harvard University Press, Cambridge. $\ORU'( $IUDPHZRUNIRUH[DPLQLQJLQWHUUHJLRQDODHULDOWUDQVSRUW of fungal spores. Agricultural and Forest Meteorology, 38, 263-288 $]RYVN\$, 6L]HGHSHQGHQWVSHFLHVDUHDUHODWLRQVKLSVLQEHQWKRV is the world more diverse for microbes? (FRJUDSK\, 25, 273-282 Baas-Becking, L.G.M. (1934) Geobiologie of inleiding tot de miliekunde. 9DQ6WRFNNXP =RRQ'HQ+DDJ

References 241 %DNHU3' %HOOLIHPLQH' (QYLURQPHQWDOLQÀXHQFHVRQ akinete germination of Anabaena circinalis and implications for management of cyanobacterial blooms. Hydrobiologia, 427, 65-73 %DUNHU3) 7KRPDV( 2ULJLQVLJQDWXUHDQGSDODHRFOLPDWLF LQÀXHQFHRIWKH$QWDUFWLF&LUFXPSRODU&XUUHQW(DUWK6FLHQFH5HYLHZV, 66, 143- 162 %DUQHV'.$+RGJVRQ'$&RQYH\3$OOHQ&6 &ODUNH$ ,QFXUVLRQDQGH[FXUVLRQRI$QWDUFWLFELRWDSDVWSUHVHQWDQGIXWXUHGlobal (FRORJ\DQG%LRJHRJUDSK\, 15, 121-142 Bayley, M., Petersen, S.O., Knigge, T., Kohler, H.R. & Holmstrup, M. (2001) 'URXJKWDFFOLPDWLRQFRQIHUVFROGWROHUDQFHLQWKHVRLOFROOHPERODQFolsomia candida. Journal of Insect Physiology, 47, 1197-1204 %HKQNH$)ULHGO7&KHSXUQRY9$ 0DQQ'* 5HSURGXFWLYH FRPSDWLELOLW\DQGU'1$VHTXHQFHDQDO\VHVLQWKHSellaphora pupula species FRPSOH[ %DFLOODULRSK\WD Journal of Phycology, 40, 193-208 %HLMHULQFN0: 'HLQIXVLHVHQGHRQWGHNNLQJGHUEDNWHULsQ Jaarboek van de Koninklijke Akademie van Wetenschappen. Müller, Amsterdam. %HQ$OL$'H%DHUH59DQ'HU$XZHUD*'H:DFKWHU5 9DQ'H Peer, Y. (2001) Phylogenetic relationships among algae based on complete large- subunit rRNA sequences. ,QWHUQDWLRQDO-RXUQDORI6\VWHPDWLFDQG(YROXWLRQDU\ Microbiology, 51, 737-749 Berney, C. & Pawlowski, J. (2006) A molecular time-scale for eukaryote evolution recalibrated with the continuous microfossil record. Proceedings of the Royal Society B-Biological Sciences, 273, 1867-1872 Beszteri, B., John, U. & Medlin, L.K. (2007) An assessment of cryptic genetic diversity within the Cyclotella meneghinianaVSHFLHVFRPSOH[ %DFLOODULRSK\WD EDVHGRQQXFOHDUDQGSODVWLGJHQHVDQGDPSOL¿HGIUDJPHQW length polymorphisms. (XURSHDQ-RXUQDORI3K\FRORJ\, 42, 47-60 %LHUNHQV-9DQ'H3HUUH: 0DHV- (IIHFWRIGLIIHUHQW environmental variables on the synthesis of Hsp70 in Raphidocelis subcapitata. Comparative Biochemistry and Physiology a-Molecular and Integrative Physiology, 120, 29-34 %LWWQHU/3D\UL&(&RXORX[$&UXDXG&'H5HYLHUV% 5RXVVHDX ) 0ROHFXODUSK\ORJHQ\RIWKH'LFW\RWDOHVDQGWKHLUSRVLWLRQZLWKLQWKH 3KDHRSK\FHDHEDVHGRQQXFOHDUSODVWLGDQGPLWRFKRQGULDO'1$VHTXHQFHGDWD 0ROHFXODU3K\ORJHQHWLFVDQG(YROXWLRQ, 49, 211-226 %OHLGRUQ&+LOO1(UVHXV& 7LHGHPDQQ5 2QWKHUROH RIFKDUDFWHUORVVLQ2UELQLLGSK\ORJHQ\ $QQHOLGD 0ROHFXOHVvs. morphology. 0ROHFXODU3K\ORJHQHWLFVDQG(YROXWLRQ, 52, 57-69 %ORFN:6PLWK5,/ .HQQHG\$' 6WUDWHJLHVRIVXUYLYDODQG UHVRXUFHH[SORLWDWLRQLQWKH$QWDUFWLFIHOO¿HOGHFRV\VWHPBiological Reviews, 84, 449-484 Boenigk, J., Pfandl, K., Garstecki, T., Harms, H., Novarino, G. & Chatzinotas, $ (YLGHQFHIRUJHRJUDSKLFLVRODWLRQDQGVLJQVRIHQGHPLVPZLWKLQD protistan morphospecies. $SSOLHGDQG(QYLURQPHQWDO0LFURELRORJ\, 72, 5159-5164 %RHQLJN--RVW66WRHFN7 *DUVWHFNL7 'LIIHUHQWLDOWKHUPDO adaptation of clonal strains of a protist morphospecies originating from different climatic zones. (QYLURQPHQWDO0LFURELRORJ\, 9, 593-602 %RQQHW( 9DQ'H3HHU< ]WDVRIWZDUHWRROIRUVLPSOHDQG partial Mantel tests. Journal of Statistical Software, 7, 1-12

242 References Boo, S.M., Kim, H.S., Shin, W., Boo, G.H., Cho, S.M., Jo, B.Y., Kim, J.H.,

References 243 Bruder, K. & Medlin, L.K. (2008) Morphological and molecular investigations of naviculoid diatoms. II. Selected genera and families. 'LDWRP5HVHDUFK, 23, 283- 329 Bruder, K., Sato, S. & Medlin, L.K. (2008) Morphological and molecular investigations of naviculoid diatoms IV. Pinnularia vs. Caloneis. 'LDWRP, 24, 8-24 Brzezinski, M.A., Olson, R.J. & Chisholm, S.W. (1990) Silicon availablility and cell-cycle progression in marine diatoms. 0DULQH(FRORJ\3URJUHVV6HULHV, 67, 83-96 %XNKWL\DURYD/ 5RXQG)( 5HYLVLRQRIWKHJHQXVAchnanthes s.l.DQHZJHQXVEDVHGRQA. marginulatum. 'LDWRP5HVHDUFK, 11, 1-30 %XQW-62ZHQV29+ +RFK* ([SORUDWRU\VWXGLHVRQ physiology and ecology of a psychrophilic marine diatom. Journal of Phycology, 2, 96-& %XVKQHOO-+ 5DR.6 'RUPDQWRUTXLHVFHQWVWDJHVDQG structures among (FWRSURFWD. Physical and chemical factors affecting viability and germination of statoblasts. Transactions of the American Microscopical Society, 93, 524-543 %XWWHUZLFN&+HDQH\6, 7DOOLQJ-) 'LYHUVLW\LQWKHLQÀXHQFH of temperature on the growth rates of freshwater algae, and its ecological relevance. Freshwater Biology, 50, 291-300 &DVWHOH\Q*(YDQV.0%DFNHOMDX7'¶+RQGW6&KHSXUQRY9 Sabbe, K. & Vyverman, W. (2009) Lack of population genetic structuring in the marine planktonic diatom Pseudo-nitzschia pungens (Bacillariophyceae) in a heterogeneous area in the Southern Bight of the North Sea. Marine Biology, 156, 1149-1158 &DVWHOH\Q*/HOLDHUW)%DFNHOMDX7'HEHHU$(.RWDNL<5KRGHV //XQGKROP16DEEH. 9\YHUPDQ: /LPLWVWRJHQHÀRZLQD cosmopolitan marine planktonic diatom. Proceedings of the National Academy of Sciences of the United States of America, 107, 12952-12957 &KDFRQ%DFD(%HUDOGL&DPSHVL+&HYDOORV)HUUL]656.QROO$+ *ROXELF6 0DQRQPDULQHGLDWRPVIURPQRUWKHUQ0H[LFRGeology, 30, 279-281 Chalmers, M.O., Harper, M.A. & Marshall, W.A. (1996) An illustrated catalogue of airborne microbiota from the Maritime Antarctic. British Antarctic Survey, Cambridge. Chambers, P.M. (1966) Late Cretaceous and Palaeocene marine diatom ÀRUDV,QS8QLYHUVLW\&ROOHJH/RQGRQ/RQGRQ Charalambidou, I. & Santamaria, L. (2002) Waterbirds as endozoochorous GLVSHUVHUVRIDTXDWLFRUJDQLVPVDUHYLHZRIH[SHULPHQWDOHYLGHQFHActa 2HFRORJLFD,QWHUQDWLRQDO-RXUQDORI(FRORJ\, 23, 165-176 Chen, G.F., Wang, G.C., Zhang, B.Y. & Fan, X.L. (2007) Morphological and phylogenetic analysis of Skeletonema costatum-like diatoms (Bacillariophyta) from the China Sea. (XURSHDQ-RXUQDORI3K\FRORJ\, 42, 163-175 &KHSLO:6 ,QÀXHQFHVRIPRLVWXUHRQHURGLELOLW\RIVRLOE\ZLQG Soil Science Society of America Proceedings, 20, 288-292 &KHSXUQRY9$0DQQ'*6DEEH. 9\YHUPDQ: ([SHULPHQWDOVWXGLHVDQVH[XDOUHSURGXFWLRQLQGLDWRPVInternational Review of Cytology - a Survey of Cell Biology, Vol. 237SS(OVHYLHU$FDGHPLF3UHVV ,QF6DQ'LHJR

244 References Cho, J.C. & Tiedje, J.M. (2000) Biogeography and degree of endemicity of ÀXRUHVFHQWPseudomonas strains in soil. $SSOLHGDQG(QYLURQPHQWDO0LFURELRORJ\, 66, 5448-5456 &KRZQ6/*UHPPHQ1-0 *DVWRQ.- (FRORJLFDO ELRJHRJUDSK\RIVRXWKHUQRFHDQLVODQGV6SHFLHVDUHDUHODWLRQVKLSVKXPDQ impacts, and conservation. American Naturalist, 152, 562-575 Chown, S.L. & Convey, P. (2007) Spatial and temporal variability across OLIH¶VKLHUDUFKLHVLQWKHWHUUHVWULDO$QWDUFWLFPhilosophical Transactions of the Royal Society B-Biological Sciences, 362, 2307-2331 Cleve, P.T. (1894) Synopsis of the naviculoid diatoms. Part I. Kongliga Svenska Vetenskapsakademiens Handlingar, 26, 1-194 Cleve, P.T. (1895) Synopsis of the naviculoid diatoms. Part 2. Kongliga Svenska Vetenskapsakademiens Handlingar, 27, 1-219 &OHYH(XOHU$ 'LH'LDWRPHHQYRQ6FKZHGHQXQG)LQQODQG Kungl. Svenska Vetenskaps Akad. Handl., 1-153 Cofey, B.T. & Miller, S.T. (1988) Hydrodictyon reticulatum (L.) Lagerheim &KORURSK\WD DQHZJHQXVUHFRUGHGIURP1HZ=HDODQGNew Zealand Journal of Botany, 26 &RQGLW53LWPDQ1/HLJK(*&KDYH-7HUERUJK-)RVWHU5% 1XQH]3$JXLODU69DOHQFLD59LOOD*0XOOHU/DQGDX+&/RVRV( Hubbell, S.P. (2002) Beta-diversity in tropical forest trees. Science, 295, 666-669 &RQVDOYH\03HUNLQV5*3DWHUVRQ'0 8QGHUZRRG*-& 3DPÀXRUHVFHQFH$EHJLQQHUVJXLGHIRUEHQWKLFGLDWRPLVWV'LDWRP5HVHDUFK, 20, 1-22 Convey, P. & Stevens, M.I. (2007) Antarctic biodiversity. Science, 317, 1877-1878 &RQYH\3*LEVRQ-$(+LOOHQEUDQG&'+RGJVRQ'$3XJK3-$ Smellie, J.L. & Stevens, M.I. (2008) Antarctic terrestrial life - challenging the history of the frozen continent? Biological Reviews, 83, 103-117 Cook, O.F. (1906) Factors of species-formation. Science, 23, 506-507 &RRN2) (YROXWLRQZLWKRXWLVRODWLRQAmerican Naturalist, 42, 727-731 Cottenie, K. (2005) Integrating environmental and spatial processes in ecological community dynamics. (FRORJ\/HWWHUV, 8, 1175-1182 &R[(- 7KHXVHRIFKORURSODVWVDQGRWKHUIHDWXUHVRIWKHOLYLQJ FHOOLQWKHWD[RQRP\RIQDYLFXORLGGLDWRPVProceedings of the 6th International Symposium on recent and fossil diatoms, Budapest, 1980, pp. 115-133. &R[(- 5RVV5 7KHVWULDHRISHQQDWHGLDWRPVProceedings of the 6th International Symposium on recent and fossil diatoms, Budapest, 1980 pp. 267-278. &R[(- 9DULDWLRQZLWKLQWKHJHQXVPinnularia(KUHQEIXUWKHU evidence for the use of live material in diatom systematics? Proceedings of the 9th ,QWHUQDWLRQDO'LDWRP6\PSRVLXP%ULVWRO, pp. 437-447. &R[(- 9DULDWLRQLQSDWWHUQVRIYDOYHPRUSKRJHQHVLVEHWZHHQ UHSUHVHQWDWLYHVRIVL[ELUDSKLGGLDWRPJHQHUD %DFLOODULRSK\FHDH Journal of Phycology, 35, 1297-1312 &R[(- :LOOLDPV'0 6\VWHPDWLFVRIQDYLFXORLGGLDWRPVWKH LQWHUUHODWLRQVKLSVRIVRPHWD[DZLWKDVWDXURV(XURSHDQ-RXUQDORI3K\FRORJ\, 35, 273-282

References 245 &R[(- :KDWFRQVWLWXWHVDVWDXURV"$PRUSKRJHQHWLFSHUVSHFWLYH Festschrift für H. Lange-Bertalot. (ed. by R. Jahn, A. Witkowski and P. Compère), pp. 303-316. &R[(- 'LDWRPV±WKHHYROXWLRQRIPRUSKRJHQHWLFFRPSOH[LW\LQ single-celled plants. 'HYHORSPHQWDOJHQHWLFVDQGSODQWHYROXWLRQ (ed. by Q.C.B. Cronk, R.M. Bateman and J.A. Hawkins), pp. 459-492. Taylor & Francis, London. &R[(- 5HLG* *HQHULFUHODWLRQVKLSVZLWKLQWKH1DYLFXOLQHDH ±DSUHOLPLQDU\FODGLVWLFDQDO\VLV3URFHHGLQJVRIWKHWK,QWHUQDWLRQDO'LDWRP Symposium., pp. 49-62. Biopress Ltd, Bristol. &R[(- :LOOLDPV'0 6\VWHPDWLFVRIQDYLFXORLGGLDWRPV %DFLOODULRSK\WD DSUHOLPLQDU\DQDO\VLVRISURWRSODVWDQGIUXVWXOHFKDUDFWHUVIRU IDPLO\DQGRUGHUOHYHOFODVVL¿FDWLRQSystematics and Biodiversity, 4, 385-399 &R[(- 0RUSKRJHQHWLFLQIRUPDWLRQDQGWKHVHOHFWLRQRIWD[RQRPLF characters for raphid diatom systematic. 3ODQW(FRORJ\DQG(YROXWLRQ, 143, 271- 277 Coyne, J.A. (1992) Genetics and speciation. Nature, 355, 511-515 Coyne, J.A. & Orr, R.J.S. (2004) Speciation. Sinauer Associates, Sunderlands. &XQQLQJKDP&:2PODQG.( 2DNOH\7+ 5HFRQVWUXFWLQJ DQFHVWUDOFKDUDFWHUVWDWHVDFULWLFDOUHDSSUDLVDO7UHQGVLQ(FRORJ\ (YROXWLRQ, 13, 361-366 &XUULH'& $GOHU3+ *OREDOGLYHUVLW\RIEODFNÀLHV 'LSWHUD Simuliidae) in freshwater. Hydrobiologia, 595, 469-475 'DUOLQJ.).XFHUD0 :DGH&0 *OREDOPROHFXODU phylogeography reveals persistent Arctic circumpolar isolation in a marine planktonic protist. Proceedings of the National Academy of Sciences of the United States of America, 104, 5002-5007 'DUOLQJ.) :DGH&$ 7KHJHQHWLFGLYHUVLW\RISODQNWLF foraminifera and the global distribution of ribosomal RNA genotypes. Marine Micropaleontology, 67, 216-238 'DUZLQ& The voyage of the Beagle. Chpater I. St. Jago - Cape de Verde Islands. John Murray, London. 'DXJEMHUJ1 $QGHUVHQ5$ $PROHFXODUSK\ORJHQ\RIWKH heterokont algae based on analyses of chloroplast-encoded rbcL sequence data. Journal of Phycology, 33, 1031-1041 'DYH\0& 7KHVHDVRQDOSHULRGLFLW\RIDOJDHRQ$QWDUFWLFIHOO¿HOG soils. +RODUFWLF(FRORJ\, 14, 112-120 GH&DQGROOH$3 (VVDLpOpPHQWDLUHGHJpRJUDSKLHERWDQLTXH 'LFWLRQQDLUHGHVVFLHQFHVQDWXUHOOHV. Flevrault, Strasbourg. de Queiroz, K. (2007) Species concepts and species delimitation. Systematic Biology, 56, 879-886 de Vargas, C., Norris, R., Zaninetti, L., Gibb, S.W. & Pawlowski, J. (1999) Molecular evidence of cryptic speciation in planktonic foraminifers and their relation to oceanic provinces. Proceedings of the National Academy of Sciences of the United States of America, 96, 2864-2868 'H:HYHU$/HOLDHUW)9HUOH\HQ(9DQRUPHOLQJHQ39DQGHU *XFKW.+RGJVRQ'$6DEEH. 9\YHUPDQ: +LGGHQOHYHOVRI SK\ORGLYHUVLW\LQ$QWDUFWLFJUHHQDOJDHIXUWKHUHYLGHQFHIRUWKHH[LVWHQFHRIJODFLDO refugia. Proceedings of the Royal Society B-Biological Sciences, 276, 3591-3599

246 References 'HQN7 *ULPP*: 7KHELRJHRJUDSKLFKLVWRU\RIEHHFKWUHHV Review of Palaeobotany and Palynology, 158, 83-100 'HVLNDFKDU\79 6UHHODWKD30 Oamaru diatoms. J. Cramer, Stuttgart. 'RGGV:.*XGGHU'$ 0ROOHQKDXHU' 7KHHFRORJ\RINostoc. Journal of Phycology, 31, 2-18 'RQJ<: 6RPHUR*1 7HPSHUDWXUHDGDSWDWLRQRIF\WRVROLF malate dehydrogenases of limpets (genus Lottia GLIIHUHQFHVLQVWDELOLW\DQG function due to minor changes in sequence correlate with biogeographic and vertical distributions. -RXUQDORI([SHULPHQWDO%LRORJ\, 212, 169-177 'RQRJKXH3&- %HQWRQ0- 5RFNVDQGFORFNVFDOLEUDWLQJWKH Tree of Life using fossils and molecules. 7UHQGVLQ(FRORJ\ (YROXWLRQ, 22, 424- 431 'UDNH-$ &RPPXQLW\DVVHPEO\PHFKDQLFVDQGWKHVWUXFWXUHRIDQ H[SHULPHQWDOVSHFLHVHQVHPEOHAmerican Naturalist, 137, 1-26 'URRS6-00DQQ'* /RNKRUVW*0 6SDWLDODQGWHPSRUDO stability of demes in 'LSORQHLVVPLWKLL'IXVFD (Bacillariophyta) supports a narrow species concept. Phycologia, 39, 527-546 'UXP5 +RSNLQV- 'LDWRPORFRPRWLRQ$QH[SODQDWLRQ Protoplasma, 62, 1-33 'UXPPRQG$- 5DPEDXW$ %($67%D\HVLDQHYROXWLRQDU\ analysis by sampling trees. %PF(YROXWLRQDU\%LRORJ\, 7, 8 'XQODS&$(YDQV.27KHHOHQ%%RHNKRXW7 6FKLVOHU'$ 2VPRWLFVKRFNWROHUDQFHDQGPHPEUDQHÀXLGLW\RIFROGDGDSWHG Cryptococcus ÀDYHVFHQV OH 182.9, previously reported as C. nodaensis, a biocontrol agent of Fusarium head blight. Fems Yeast Research, 7, 449-458 (GJDU/$ 3LFNHWW+HDSV-' 'LDWRPORFRPRWLRQProgress in Phycology Research, 3, 47-88 (GJDU60 7KHULRW(& 3K\ORJHQ\RIAulacoseira (Bacillariophyta) based on molecules and morphology. Journal of Phycology, 40, 772-788 (KUHQEHUJ&* 0LWWKHLOXQJHQEUHQHXHDVLVWLVFKH/DJHUIRVVLOHU ,QIXVRULH(UGHQDXVGHPUXVVLVFKHQ7UDQV.DXNDVLHQ *UXVLHQ XQG6LEHULHQBe. Bekanntm. Verh. Königl. Preuss. Akad. Wiss. Berlin, 1834, 43-49 (KUHVPDQQ': +DWFK07 (IIHFWRIUHODWLYHKXPLGLW\RQ survival of airborn unicellular algae. Applied Microbiology, 29, 352-357 (OVWHU-'HJPD3.RYDFLN/9DOHQWRYD/6UDPNRYD. 3HUHLUD$% )UHH]LQJDQGGHVLFFDWLRQLQMXU\UHVLVWDQFHLQWKH¿ODPHQWRXVJUHHQDOJD Klebsormidium from the Antarctic, Arctic and Slovakia. Biologia, 63, 843-851 (WWO+ 'LH*DWWXQJChlamydomonas(KUHQEHUJNova Hedwigia Beihefte, 49, 1-1122 (WWO+ *lUWQHU* Syllabus der Boden-, Luft- und Flechtenalgen. Gustav Fisher Verlag, Stuttgart. (YDQV-+ 7KHVXUYLYDORIIUHVKZDWHUDOJDHGXULQJGU\SHULRGV An investigation of the algae of 5 small ponds. -RXUQDORI(FRORJ\, 46, 149-167 (YDQV-+ 7KHVXUYLYDORIIUHVKZDWHUDOJDHGXULQJGU\SHULRGV 'U\LQJH[SHULPHQWV6WUDWL¿FDWLRQRIDOJDHLQSRQGPDUJLQOLWWHUDQGPXG -RXUQDORI(FRORJ\, 47, 55-81

References 247 (YDQV.0:RUWOH\$+ 0DQQ'* $QDVVHVVPHQWRISRWHQWLDO diatom “barcode” genes (FR[1, rbc/6DQG,76U'1$ DQGWKHLUHIIHFWLYHQHVVLQ determining relationships in Sellaphora (Bacillariophyta). Protist, 158, 349-364 (YDQV.0:RUWOH\$+6LPSVRQ*(&KHSXUQRY9$ 0DQQ'* $PROHFXODUV\VWHPDWLFDSSURDFKWRH[SORUHGLYHUVLW\ZLWKLQWKHSellaphora pupula VSHFLHVFRPSOH[ EDFLOODULRSK\WD Journal of Phycology, 44, 215-231 (YDQV.0&KHSXUQRY9$6OXLPDQ+-7KRPDV6-6SHDUV%0 0DQQ'* +LJKO\GLIIHUHQWLDWHGSRSXODWLRQVRIWKHIUHVKZDWHUGLDWRP Sellaphora capitata suggest limited dispersal and opportunities for allopatric speciation. Protist, 160, 386-396 )DUULV-6 3KHQHWLFVLQFDPRXÀDJHCladistics-the International Journal of the Willi Hennig Society, 6, 91-100 Fawley, M.W., Fawley, K.P. & Buchheim, M.A. (2004) Molecular diversity among communities of freshwater microchlorophytes. 0LFURELDO(FRORJ\, 48, 489- 499 )HOVHQVWHLQ- &RQ¿GHQFHOLPLWVRQSK\ORJHQLHVDQDSSURDFKXVLQJ the bootstrap. (YROXWLRQ, 39, 783-791 Felsenstein, J. (2004) Inferring phylogenies. Sinauer Associates, Sunderland. Fiala, M. & Oriol, L. (1990) Light-temperature interactions on the growth of Antarctic diatoms. Polar Biology, 10, 629-636 )LJXHUROD- *UHHQ$- 'LVSHUVDORIDTXDWLFRUJDQLVPVE\ ZDWHUELUGVDUHYLHZRISDVWUHVHDUFKDQGSULRULWLHVIRUIXWXUHVWXGLHVFreshwater Biology, 47, 483-494 Figuerola, J., Green, A.J. & Santamaria, L. (2003) Passive internal transport RIDTXDWLFRUJDQLVPVE\ZDWHUIRZOLQ'RQDQDVRXWKZHVW6SDLQ*OREDO(FRORJ\ and Biogeography, 12, 427-436 )LJXHUROD-*UHHQ$- 0LFKRW7& ,QYHUWHEUDWHHJJVFDQÀ\ (YLGHQFHRIZDWHUIRZOPHGLDWHGJHQHÀRZLQDTXDWLFLQYHUWHEUDWHVAmerican Naturalist, 165, 274-280 Finlay, B.J. (2002) Global dispersal of free-living microbial eukaryote species. Science, 296, 1061-1063 )LQOD\%-0RQDJKDQ(% 0DEHUO\6& +\SRWKHVLV7KHUDWH and scale of dispersal of freshwater diatom species is a function of their global abundance. Protist, 153, 261-273 )LQOD\%- (VWHEDQ*) 8ELTXLWRXVGLVSHUVDORIIUHHOLYLQJ microorganisms. Microbial diversity and bioprospecting (ed. by A.T. Bull), pp. 216- $603UHVV:DVKLQJWRQ'& Flechtner, V.R., Johansen, J.R. & Clark, W.H. (1998) Algal composition of PLFURELRWLFFUXVWVIURPWKH&HQWUDO'HVHUWRI%DMD&DOLIRUQLD0H[LFRGreat Basin Naturalist, 58, 295-311 Flower, B.P. & Kennett, J.P. (1994) The middle Miocene climatic transition (DVW$QWDUFWLFLFHVKHHWGHYHORSPHQWGHHSRFHDQFLUFXODWLRQDQGJOREDOFDUERQ cycling. Palaeogeography Palaeoclimatology Palaeoecology, 108, 537-555 Frenot, Y., Gloaguen, J.C., Cannavacciuolo, M. & Bellido, A. (1998) Primary succession on glacier forelands in the subantarctic Kerguelen Islands. Journal of Vegetation Science, 9, 75-84 Frenot, Y., Gloaguen, J.C., Masse, L. & Lebouvier, M. (2001) Human activities, ecosystem disturbance and plant invasions in subantarctic Crozet, Kerguelen and Amsterdam Islands. Biological Conservation, 101, 33-50

248 References )U\HU* 'LDSDXVHDSRWHQWIRUFHLQWKHHYROXWLRQRIIUHVKZDWHU crustaceans. Hydrobiologia, 320, 1-14 Gao, Z., Horton, R., Wang, L., Liu, H. & Wen, J. (2008) An improved force- restore method for soil temperature prediction. (XURSHDQ-RXUQDORI6RLO6FLHQFH, 59, 972-981 *DUFLOODQ33 (]FXUUD( %LRJHRJUDSKLFUHJLRQVDQGEHWDGLYHUVLW\ of woody dryland legumes in the Baja California peninsula. Journal of Vegetation Science, 14, 859-868 *HLVVOHU8 *HUORII- 'DVYRUNRPPHQYRQ'LDWRPHHQLQ menschlichen Organen und in der Luft. Nova Hedwigia, 10, 565-577 *HLWOHU/ $EKlQJLJNHLWGHU0HPEUDQELOGXQJYRQGHU=HOOWHLOXQJEHL 'LDWRPHHQXQGGLIIHUQWLHOOH7HLOXQJHQLP]XVDPPHQKDQJPLWGHU,QQHQVFKDOHQ Planta, 43, 75-82 *HLWOHU/ 'LHLQlTXDOH7HLOXQJEHLGHU%LOGXQJGHU,QQHQVFKDOHQ von Meridion circulare. Osterreichesche Botanische Zeitschrift, 119, 442-446 *HR9ODDQGHUHQ 'LJLWDOVRLOPDSRI)ODQGHUV85/DYDLODEOHDW KWWSZZZDJLYEHJLV *HUODFK' %RWDQLVFKH0LNURWHFKQLN(LQH(LQIKUXQJ. Georg Thieme Verlag, Stuttgart. *HUVRQGH5 +DUZRRG'0 /RZHU&UHWDFHRXVGLDWRPVIURP 2'3/HJVLWH :HGGHOO6HD 3DUWYHJHWDWLYHFHOOVProceedings of the 2FHDQ'ULOOLQJ3URJUDP6FLHQWL¿F5HVXOWV, pp. 365-402. *LOEHUW-- 6FKUHLEHU'. ,QGXFWLRQRIGLDSDXVLQJDPLFWLFHJJV in Synchaeta pectinata. Hydrobiologia, 313, 345-350 *LOODUG-'HYRV9+X\VPDQ0--'H9H\OGHU/'¶+RQGW60DUWHQV &9DQRUPHOLQJHQ39DQQHUXP.6DEEH.&KHSXUQRY9$,Q]H' Vuylsteke, M. & Vyverman, W. (2008) Physiological and transcriptomic evidence for a close coupling between chloroplast ontogeny and cell cycle progression in the pennate diatom Seminavis robusta. Plant Physiology, 148, 1394-1411 Giribet, G. (2010) A new dimension in combining data? The use of morphology and phylogenomic data in metazoan systematics. Acta Zoologica, 91, 11-19 *ORFNQHU)2=DLFKLNRY(%HONRYD1'HQLVVRYD/3HUQWKDOHU- Pernthaler, A. & Amann, R. (2000) Comparative 16S rRNA analysis of lake bacterioplankton reveals globally distributed phylogenetic clusters including an abundant group of actinobacteria. $SSOLHGDQG(QYLURQPHQWDO0LFURELRORJ\, 66, 5053-+ Godhe, A. & Harnström, K. (2010) Linking the planktonic and benthic KDELWDWJHQHWLFVWUXFWXUHRIWKHPDULQHGLDWRPSkeletonema marinoi. Molecular (FRORJ\, 19, 4478-4490 Goloboff, P.A., Mattoni, C.I. & Quinteros, A.S. (2006) Continuous characters analyzed as such. Cladistics, 22, 589-601 *RQ]DOH]-RVH5(VFDSD,1HYHV:$&XQHR5 3XFFLDUHOOL+0 (2008) Cladistic analysis of continuous modularized traits provides phylogenetic signals in Homo evolution. Nature, 453, 775-U4 Gran, H.H. (1912) Pelagic plant life. The depths of the ocean (ed. by J. Murray and J. Hjort). Macmillan, London. *UD\':/HZLV/$ &DUGRQ=* 3KRWRV\QWKHWLFUHFRYHU\ following desiccation of desert green algae (Chlorophyta) and their aquatic relatives. 3ODQW&HOODQG(QYLURQPHQW, 30, 1240-1255

References 249 *UHHQ-/+ROPHV$-:HVWRE\02OLYHU,%ULVFRH''DQJHU¿HOG M., Gillings, M. & Beattie, A.J. (2004) Spatial scaling of microbial eukaryote diversity. Nature, 432, 747-750 Green, A.J. & Figuerola, J. (2005) Recent advances in the study of long- distance dispersal of aquatic invertebrates via birds. 'LYHUVLW\DQG'LVWULEXWLRQV, 11, 149-156 Green, J. & Bohannan, B.J.M. (2006) Spatial scaling of microbial biodiversity. 7UHQGVLQ(FRORJ\ (YROXWLRQ, 21, 501-507 Greenslade, P. (1995) Collembola from the Scotia Arc and Antarctic Peninsula including descriptions of two new species and notes on biogeography. 3ROVNLH3LVPR(QWRPRORJLF]QH, 64, 305-319 Greve, M., Gremmen, N.J.M., Gaston, K.J. & Chown, S.L. (2005) 1HVWHGQHVVRI6RXWKHUQ2FHDQLVODQGELRWDVHFRORJLFDOSHUVSHFWLYHVRQD biogeographical conundrum. Journal of Biogeography, 32, 155-168 *ULI¿WKV$-)0LOOHU-)6X]XNL'7/HZRQWLQ5& *HOEDUW:0 (1996). Introduction to Genetic AnalysisWK(GLWLRQ:+)UHHPDQ1HZBolidophyceae (Heterokonta). Journal of Phycology, 35, 368-381 *XLU\0' *XLU\*0 $OJDH%DVH:RUOGZLGHHOHFWURQLF SXEOLFDWLRQ1DWLRQDO8QLYHUVLW\RI,UHODQG*DOZD\KWWSZZZDOJDHEDVHRUJ searched on 19 February 2011. Gupta, S. & Agrawal, S.C. (2006) Survival of blue-green and green algae under stress conditions. Folia Microbiologica, 51, 121-128 Hajós, M. & Stradner, H. (1975) Late Cretaceous Archaeomonadaceae, 'LDWRPDFHDHDQG6LOLFRÀDJHOODWDHIURPWKH6RXWK3DFL¿F2FHDQ'HHS6HD'ULOOLQJ Project Leg 29, site 275. ,QLWLDO5HSRUWVRIWKH'HHS6HD'ULOOLQJ3URMHFW, 29, 913- 1109 +DMyV0 6WUDWLJUDSK\RI+XQJDU\¶V0LRFHQHGLDWRPDFHRXVHDUWK deposits. Geologica Hungarica, Series. Palaeontologica, 49, 1-339 +DOO7$ %LR(GLWDXVHUIULHQGO\ELRORJLFDOVHTXHQFHDOLJQPHQW HGLWRUDQGDQDO\VLVSURJUDPIRU:LQGRZV17Nucl. Acids. Symp. Ser., 41, 95-98 Hallegraeff, G.M. & Bolch, C.J. (1992) Transport of diatom and GLQRÀDJHOODWHUHVWLQJVSRUHVLQVKLSV¶EDOODVWZDWHULPSOLFDWLRQVIRUSODQNWRQ biogeography and aquaculture. Journal of Plabkton Research, 14, 1067-1084 Hanski, I. (1999) Metapopulation ecology2[IRUG8QLYHUVLW\3UHVV2[IRUG +DUJUDYHV3( )UHQFK): 2EVHUYDWLRQVRQWKHVXUYLYDORI diatoms resting spores. Nova Hedwigia Beihefte, 53, 229-238 +DUJUDYHV3( )UHQFK): 'LDWRPUHVWLQJVSRUHVVLJQL¿FDQFH and strategies. Survival strategies of the algae HGE\*$)U\[HOO SS Cambridge University Press, New York.

250 References +DUZRRG'0 *HUVRQGH5 /RZHU&UHWDFHRXVGLDWRPVIURP 2'3/HJ6LWH :HGGHOO6HD 3DUWUHVWLQJVSRUHVFKU\VRSK\FHDQF\VWV DQHQGRVNHOHWDOGLQRÀDJHOODWHDQGQRWHVRQWKHRULJLQRIGLDWRPVProceedings of WKH2FHDQ'ULOOLQJ3URJUDP6FLHQWL¿F5HVXOWV, 113, 403-425 +D\ZDUG6$/5LQHKDUW-36DQGUR/+/HH5( 'HQOLQJHU '/ 6ORZGHK\GUDWLRQSURPRWHVGHVLFFDWLRQDQGIUHH]HWROHUDQFHLQWKH Antarctic midge Belgica antarctica. -RXUQDORI([SHULPHQWDO%LRORJ\, 210, 836-844 +HEHUW3'15DWQDVLQJKDP6 GH:DDUG-5 %DUFRGLQJDQLPDO OLIHF\WRFKURPHFR[LGDVHVXEXQLWGLYHUJHQFHVDPRQJFORVHO\UHODWHGVSHFLHV Proceedings of the Royal Society of London Series B-Biological Sciences, 270, S96-S99 Hennig, W. (1965) Phylogenetic systematics. $QQ5HY(QWRPRO, 10, 97- 116 Hennig, W. (1966) Phylogenetic Systematics. University of Illinois Press, Urbana. +pULEDXG- /HVGLDWRPpHVIRVVLOHVG¶$XYHUJQH, Clermont-Ferrand. Hillebrand, H., Watermann, F., Karez, R. & Berninger, U.G. (2001) 'LIIHUHQFHVLQVSHFLHVULFKQHVVSDWWHUQVEHWZHHQXQLFHOOXODUDQGPXOWLFHOOXODU organisms. Oecologia, 126, 114-124 Ho, S.Y.W. & Phillips, M.J. (2009) Accounting for calibration uncertainty in phylogenetic estimation of evolutionary divergence times. Systematic Biology, 58, 367-380 +REDQ0$)U\[HOO*$ %XFN.5 %LGGXOSKLRLGGLDWRPV resting spores in Antarctic (XFDPSLD and Odontella. Journal of Phycology, 16, 591-602 +RGJVRQ$(YDQV'-%URZQ05: *LOEHUW3 ,QYROYHPHQW of cellular division cycle in the susceptibility of (VFKHULFKLDFROL to cold-shock and osmotic shock. Microbios, 72, 175-182 +RGJVRQ'$9HUOH\HQ(6DEEH.6TXLHU$+.HHO\%-/HQJ M.J., Saunders, K.M. & Vyverman, W. (2005) Late Quaternary climate-driven HQYLURQPHQWDOFKDQJHLQWKH/DUVHPDQQ+LOOV(DVW$QWDUFWLFDPXOWLSUR[\ evidence from a lake sediment core. Quaternary Research, 64, 83-99 Hoffmann, L. (1996) Geographic distribution of freshwater blue-green algae. Hydrobiologia, 336, 33-40 +ROOLEDXJK-76HLEHUW'/5 7KRPDV:+ 2EVHUYDWLRQV on the survival and germination of resting spores of 3 Chaetoceros (Bacillariophyceae) species. Journal of Phycology, 17, 1-9 +ROPVWUXS0+HGOXQG. %RULVV+ 'URXJKWDFFOLPDWLRQDQG lipid composition in Folsomia candidaLPSOLFDWLRQVIRUFROGVKRFNKHDWVKRFNDQG acute desiccation stress. Journal of Insect Physiology, 48, 961-970 Hori, K., Okamoto, J., Tanji, Y. & Unno, H. (2003) Formation, sedimentation and germination properties of Anabaena akinetes. %LRFKHPLFDO(QJLQHHULQJ Journal, 14, 67-73 +RUQHU'HYLQH0&/DJH0+XJKHV-% %RKDQQDQ%-0 $ WD[DDUHDUHODWLRQVKLSIRUEDFWHULDNature, 432, 750-753 +RVWHWWHU+3 +RVKDZ5: (QYLURQPHQWDOIDFWRUVDIIHFWLQJ resistance to desiccation in the diatom Stauroneis anceps. American Journal of Botany, 57, 512-518 Hubbell, S.P. (2001) 7KHXQL¿HGWKHRU\RI%LRGLYHUVLW\DQG%LRJHRJUDSK\. Princeton University Press, New Yersey.

References 251 +XHOVHQEHFN-3 5RQTXLVW) 0U%D\HV%D\HVLDQLQIHUHQFHRI phylogenetic trees. Bioinformatics, 17, 754-755 +XOWpQ( Outline of the history of Arctic and Boreal biota during the Quarternary period. Lehre J Cramer, New York. +XWFKLQVRQ*( A treatise on limnology. Wiley, New York. +X\VPDQ0--0DUWHQV&9DQGHSRHOH.*LOODUG-5D\NR(+HLMGH 0%RZOHU&,Q]H'9DQGH3HHU<'H9H\OGHU/ 9\YHUPDQ: Genome-wide analysis of the diatom cell cycle unveils a novel type of cyclins involved in environmental signaling. Genome Biology, 11, 19 Imshenetsky, A.A. & Kondrateva, T.F. (1987) Resistance of polyploid cultures of Candida utilis to the action of ultraviolette rays. Microbiology, 56, 698- 700 ,UpQpH0DULH) )ORUHGHVPLGLDOHGHODUpJLRQGH0RQWUpDO. Laprairie, Montreal. Isard, S.A. & Gage, S.H. (2001) )ORZRIOLIHLQWKHDWPRVSKHUHDQDLUVFDSH appraoch to understanding invasive organisms. Michigan State University Press, (DVW/DQVLQJ ,YDQRYLF$6RWLURSRXORV.']XNLF* .DOH]LF0/ 6NXOOVL]H DQGVKDSHYDULDWLRQYHUVXVPROHFXODUSK\ORJHQ\DFDVHVWXG\RIDOSLQHQHZWV (Mesotriton alpestris, Salamandridae) from the Balkan Peninsula. Zoomorphology, 128, 157-167 -DFREV%'RQRJKXH0-%RXPDQ)+X\VPDQV6 6PHWV( (YROXWLRQDQGSK\ORJHQHWLFLPSRUWDQFHRIHQGRFDUSDQGVHHGFKDUDFWHUVLQ Viburnum $GR[DFHDH International Journal of Plant Sciences, 169, 409-431 Jargeat, P., Martos, F., Carriconde, F., Gryta, H., Moreau, P.A. & Gardes, M. (2010) Phylogenetic species delimitation in ectomycorrhizal fungi and implications IRUEDUFRGLQJWKHFDVHRIWKHTricholoma scalpturatumFRPSOH[ %DVLGLRP\FRWD 0ROHFXODU(FRORJ\, 19, 5216-5230 -DZRUVNL*- /XQG-:* 'URXJKWUHVLVWDQFHDQGGLVSHUVDORI Asterionella formosa Hass. Nova Hedwigia Beihefte, 31, 37-48 -HZVRQ'+/RZU\6) %RZHQ5 &RH[LVWHQFHDQGVXUYLYDORI diatoms on sand grains. (XURSHDQ-RXUQDORI3K\FRORJ\, 41, 131-146 -REE*9RQ+DHVHOHU$ 6WULPPHU. 75((),1'(5DSRZHUIXO graphical analysis environment for molecular phylogenetics. %PF(YROXWLRQDU\ Biology, 4, 9 -RKDQVHQ-5 &KDSWHU'LDWRPVRIDHULDOKDELWDWV7KHGLDWRPV applications for the environmental and earth sciences HGE\()6WRHUPHUDQG J.P. Smol), pp. 264-273. Cambridge University Press, New York. Johst, K. & Brandl, R. (1997) The effect of dispersal on local population dynamics. (FRORJLFDO0RGHOOLQJ, 1, 87-101 -RQHV+06LPSVRQ*(6WLFNOH$- 0DQQ'* /LIHKLVWRU\ and systematics of Petroneis (Bacillariophyta), with special reference to British waters. (XURSHDQ-RXUQDORI3K\FRORJ\, 40, 61-87 -RQHV6( /HQQRQ-7 'RUPDQF\FRQWULEXWHVWRWKH maintenance of microbial diversity. Proceedings of the National Academy of Sciences of the United States of America, 107, 5881-5886 Jost, L. (2007) Partitioning diversity into independent alpha and beta components. (FRORJ\, 88, 2427-2439

252 References Julius, M.L. & Tanimura, Y. (2001) Cladistic analysis of plicated Thalassiosira (Bacillariophyceae). Phycologia, 40, 111-122 .LQJ5$ )HUULV& &KORURSODVW'1$SK\ORJHRJUDSK\RIAlnus glutinosa (L.) Gaertn. 0ROHFXODU(FRORJ\, 7, 1151-1161 .LUVW*2 :LHQFNH& (FRSK\VLRORJ\RISRODUDOJDHJournal of Phycology, 31, 181-199 .OHLQ*0DF,QWRVK..DF]PDUVND, (KUPDQ-0 'LDWRP VXUYLYRUVKLSLQEDOODVWZDWHUGXULQJWUDQV3DFL¿FFURVVLQJVBiological Invasions, 12, 1031-1044 .RFLROHN-3 'HWHUPLQLQJFKDUDFWHUSRODULW\LQGLDWRPVDQ alternative appraoch. 'LDWRP5HVHDUFK, 1, 189-192 .RFLROHN-3 6WRHUPHU() 3K\ORJHQHWLFUHODWLRQVKLSV DQGFODVVL¿FDWLRQRIPRQRUDSKLGGLDWRPVEDVHGRQSKHQHWLFDQGFODGLVWLF methodologies. Phycologia, 25, 297-303 Kociolek, J.P. (1997) Historical constraints, species concepts and the search IRUDQDWXUDOFODVVL¿FDWLRQRIGLDWRPV'LDWRP, 13, 3-8 Kociolek, J.P. & Spaulding, S.A. (2000) Freshwater diatom biogeography. Nova Hedwigia, 71, 223-241 .RHVWHU-$6ZDOZHOO-(YRQ'DVVRZ3 $UPEUXVW(9 Genome size differentiates co-occurring populations of the planktonic diatom 'LW\OXPEULJKWZHOOLL (Bacillariophyta). %PF(YROXWLRQDU\%LRORJ\, 10 .RRLVWUD: 0HGOLQ/. (YROXWLRQRIWKHGLDWRPV (bacillariophyta). 4. Reconstruction of their age from small subunit rRNA coding regions and the fossil record. 0ROHFXODU3K\ORJHQHWLFVDQG(YROXWLRQ, 6, 391-407 .RRLVWUD:'H6WHIDQR00DQQ'*6DOPD1 0HGOLQ/. Phylogenetic position of 7R[DULXP, a pennate-like lineage within centric diatoms (Bacillariophyceae). Journal of Phycology, 39, 185-197 Kooistra, W., Hargraves, P., Andersen, R.A., Balzano, S., Zingone, A. & 6DUQR' 3VHXGRFU\SWLFGLYHUVLW\DQGSK\ORJHRJUDSK\LQWKHFHQWULF diatom Skeletonema. Phycologia, 44, 134 .RRLVWUD:6DUQR'+HUQDQGH]%HFHUULO'8$VVP\3'L3ULVFR& & Montresor, M. (2010) Comparative molecular and morphological phylogenetic DQDO\VHVRIWD[DLQWKH&KDHWRFHURWDFHDH %DFLOODULRSK\WD Phycologia, 49, 471- 500 Kottek, M., Grieser, J., Beck, C., Rudolf, B. & Rubel, F. (2006) World map of WKH.RSSHQ*HLJHUFOLPDWHFODVVL¿FDWLRQXSGDWHGMeteorologische Zeitschrift, 15, 259-263 Krammer, K. & Lange-Bertalot, H. (1985) Naviculaceae, neue und wenig EHNDQQWH7D[DQHXH.RPELQDWLRQHQXQG6\QRQ\PHVRZLH%HPHUNXQJHQ]X einigen Gattungen, Braunschweig. Krammer, K. & Lange-Bertalot, H. (1986) %DFLOODULRSK\FHDH7HLO Naviculaceae. Gustav Fisher Verlag, Stuttgart. Krammer, K. & Lange-Bertalot, H. (1988) %DFLOODULRSK\FHDH7HLO (SLWKHPLDFHDH%DFLOODULDFHDH6XULOOHOODFHDH. Gustav Fisher Verlag, Stuttgart. Krammer, K. & Lange-Bertalot, H. (1991a) %DFLOODULRSK\FHDH7HLO &HQWUDOHV)UDJLODULDFHDH(XQRWLDFHDH. Gustav Fischer Verlag, Stuttgart. Krammer, K. & Lange-Bertalot, H. (1991b) %DFLOODULRSK\FHDH7HLO $FKQDQWKDFHDH.ULWLVFKH(UJlQ]XQJHQ]X1DYLFXOD /LQHRODWD XQG*RPSKRQHPD. Gustav Fisher Verlag, Stuttgart.

References 253 Krammer, K. (1992) 3LQQXODULDHLQH0RQRJUDSKLHGHUHXURSlLVFKHQ7D[D. J. Cramer, Berlin. Krammer, K. (2000) The genus Pinnularia. A.R.G. Gantner Verlag, Ruggell. Krammer, K. (2002) The genus Cymbella. A.R.G. Gantner Verlag, Ruggell. .ULVWLDQVHQ- 'LVSHUVDORIIUHVKZDWHUDOJDH$UHYLHZ Hydrobiologia, 336, 151-157 .XPDU66NMDHYHODQG$2UU5-6(QJHU35XGHQ70HYLN%+ %XUNL)%RWQHQ$ 6KDOFKLDQ7DEUL]L. $,5$EDWFKRULHQWHGZHE program package for construction of supermatrices ready for phylogenomic analyses. BMC Bioinformatics, 10, 7 Kuwata, A. & Takahashi, M. (1990) Life-form population responses of a marine planktonic diatom, Chaetoceros pseudocurvisetus, to oligothophication in regionally upwelled water. Marine Biology, 107, 503-512 .XZDWD$+DPD7 7DNDKDVKL0 (FRSK\VLRORJLFDO characterization of 2 life forms, resting spores and resting cells, of a marine phytoplanktonic diatom, Chaetoceros pseudocurvisetus, formed under nitrogen depletion. 0DULQH(FRORJ\3URJUHVV6HULHV, 102, 245-255 /DJDEULHOOH<*RGGHULV<'RQQDGLHX<0DODYLHLOOH- 6XDUH]0 7KHWHFWRQLFKLVWRU\RI'UDNH3DVVDJHDQGLWVSRVVLEOHLPSDFWVRQJOREDO climate. (DUWKDQG3ODQHWDU\6FLHQFH/HWWHUV, 279, 197-211 Lande, R. (1996) Statistics and partitioning of species diversity, and similarity among multiple communities. Oikos, 76, 5-13 /DQIHDU5+R6<:/RYH' %URPKDP/ 0XWDWLRQUDWHLV OLQNHGWRGLYHUVL¿FDWLRQLQELUGVProceedings of the National Academy of Sciences of the United States of America, 107, 20423-20428 Lange-Bertalot, H. (1993) QHXHWD[DXQGEHUZHLWHUHQHX GH¿QLHUWHWD[DHUJlQ]HQG]XU6VVZDVVHUÀRUDYRQ0LWWHOHXURSD9RO. J. Cramer, Berlin. Lange-Bertalot, H. (2001) Navicula sensu stricto, 10 new genera separated from Navicula sensu lato, Frustulia. A.R.G. Gantner Verlag, Ruggell. /DZYHU/$ *DKDJDQ/0 (YROXWLRQRI&HQR]RLFVHDZD\VLQWKH circum-Antarctic region. Palaeogeography Palaeoclimatology Palaeoecology, 198, 11-37 /H&RKX5 0DLOODUG5 /HVGLDWRPpHVPRQRUDSKLGpHVGHV,OHV Kerguelen. Ann. Limnol., 19, 143-167 /H&RKX5 0DLOODUG5 'LDWRPpHVG¶HDXGRXFHGHV,OHV .HUJXHOHQ jO¶H[FOXVLRQGHV0RQRUDSKLGpHV Ann. Limnol., 22, 99-118 /H&RKX5 5pYLVLRQGHVSULQFLSDOHVHVSqFHVGXOoDTXLFROHV G¶$FKQDQWKDOHV %DFLOODULRSK\WD GHVvOHVVXEDQWDUFWLTXHVGH.HUJXHOHQAlgol. Stud., 116, 79-114 Lee, M.S.Y. & Camens, A.B. (2009) Strong morphological support for the molecular evolutionary tree of placental mammals. -RXUQDORI(YROXWLRQDU\%LRORJ\, 22, 2243-2257 /HJHQGUH3%RUFDUG' 3HUHV1HWR35 $QDO\]LQJEHWD GLYHUVLW\3DUWLWLRQLQJWKHVSDWLDOYDULDWLRQRIFRPPXQLW\FRPSRVLWLRQGDWD (FRORJLFDO0RQRJUDSKV, 75, 435-450

254 References Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J.M., +RRSHV0)+ROW5'6KXULQ-%/DZ57LOPDQ'/RUHDX0 *RQ]DOH] $ 7KHPHWDFRPPXQLW\FRQFHSWDIUDPHZRUNIRUPXOWLVFDOHFRPPXQLW\ ecology. (FRORJ\/HWWHUV, 7, 601-613 /HYLQV5 'LDSDXVHDVDQDGDSWLYHVWUDWHJ\Symposia of the 6RFLHW\RI([SHULPHQWDO%LRORJ\, 23, 1-10 Lewis Smith, R.I. (1984) Terrestrial plant biology of the sub-Antarctic and Antarctic. $QWDUFWLF(FRORJ\ (ed. by R.M. Laws), pp. 61-162. Academic Press, London. Lewis, L.A. & Flechtner, V.R. (2004) Cryptic species of Scenedesmus (Chlorophyta) from desert soil communities of Western North America. Journal of Phycology, 40, 1127-1137 Lewis, L.A. & Lewis, P.O. (2005) Unearthing the molecular phylodiversity of desert soil green algae (Chlorophyta). Systematic Biology, 54, 936-947 /HZLV$50DUFKDQW'5$VKZRUWK$&+HGHQDV/+HPPLQJ65 -RKQVRQ-9/HQJ0-0DFKOXV0/1HZWRQ$(5DLQH-,:LOOHQEULQJ -.:LOOLDPV0 :ROIH$3 0LG0LRFHQHFRROLQJDQGWKHH[WLQFWLRQRI tundra in continental Antarctica. Proceedings of the National Academy of Sciences of the United States of America, 105, 10676-10680 /L<0)HUJXVRQ'.:DQJ<) /L&6 3DOHRHQYLURQPHQWDO inferences from diatom assemblages of the middle Miocene Shanwang Formation, Shandong, China. Journal of Paleolimnology, 43, 799-814 /LJKWKDUW% 0RKU$- (VWLPDWLQJGRZQZLQGFRQFHQWUDWLRQVRI viable airborne microorganisms in dynamic atmospheric conditions. Applied and (QYLURQPHQWDO0LFURELRORJ\, 53, 1580-1583 Linnaeus, C. (1781) On the increase of the habitable earth. Translation by F.J. Brandt. Amoenitates Academicae, 2, 17-27 /RKPDQ.( $QGUHZV*: /DWH(RFHQHQRQPDULQHGLDWRPV IURPWKH%HDYHU'LYLGHDUHD)UHPRQW&RXQW\:\RPLQJUnited States Geological Survey Professional Paper, 593-E /RPROLQR096D[') %URZQ-+ Foundations of biogeography. Classic papers with commentaries. The university of Chicago Press, London. Lomolino, M.V., Riddle, B.R., Whittaker, R.J., Brown, J.H. (2010) Biogeography. Sinauer Associates, Inc. Lund, J.W.G. (1953) The seasonal cycle of the plankton diatom Melosira italica (KU .XW]subsp. subarctica O. Mull. -RXUQDORI(FRORJ\, 42, 151-179 /XQGKROP1'DXJEMHUJ1 0RHVWUXS2 3K\ORJHQ\RIWKH Bacillariaceae with emphasis on the genus Pseudo-nitzschia (Bacillariophyceae) EDVHGRQSDUWLDO/68U'1$(XURSHDQ-RXUQDORI3K\FRORJ\, 37, 115-134 /XQGKROP10RHVWUXS2+DVOH*5 +RHI(PGHQ. A study of the 3VHXGRQLW]VFKLDSVHXGRGHOLFDWLVVLPDFXVSLGDWDFRPSOH[ %DFLOODULRSK\FHDH :KDWLVP. pseudodelicatissima? Journal of Phycology, 39, 797-813 0DF$UWKXU5+ :LOVRQ(2 $QHTXLOLEULXPWKHRU\IRULQVXODU zoogeography. (YROXWLRQ, 17, 373-387 0DGGLVRQ:3 0DGGLVRQ'5 0HVTXLWHDPRGXODUV\VWHPIRU HYROXWLRQDU\DQDO\VLV9HUVLRQKWWSPHVTXLWHSURMHFWRUJ

References 255 0DJUL'9HQGUDPLQ**&RPSV%'XSDQORXS,*HEXUHN7*RPRU\ '/DWDORZD0/LWW73DXOH/5RXUH-07DQWDX,YDQGHU.QDDS:2 Petit, R.J. & de Beaulieu, J.L. (2006) A new scenario for the Quaternary history of (XURSHDQEHHFKSRSXODWLRQVSDODHRERWDQLFDOHYLGHQFHDQGJHQHWLFFRQVHTXHQFHV New Phytologist, 171, 199-221 Maguire, B. (1963) Passive dispersal of small aquatic organisms and their colonization of isolated bodies of water. (FRORJLFDO0RQRJUDSKV, 33, 161-185 0DQQ'* 'URRS6-0 %LRGLYHUVLW\ELRJHRJUDSK\DQG conservation of diatoms. Hydrobiologia, 336, 19-32 0DQQ'* 7KHVSHFLHVFRQFHSWLQGLDWRPVPhycologia, 38, 437- 495 0DQQ'* $GLVFXVVLRQRICaloneis and related genera. 'LDWRP, 17, 29-36 0DQQ'*0F'RQDOG60%D\HU00'URRS6-0&KHSXUQRY9$ /RNH5(&LREDQX$ 'X%XI-0+ 7KHSellaphora pupula species FRPSOH[ %DFLOODULRSK\FHDH PRUSKRPHWULFDQDO\VLVXOWUDVWUXFWXUHDQGPDWLQJ GDWDSURYLGHHYLGHQFHIRU¿YHQHZVSHFLHVPhycologia, 43, 459-482 0DQQ'* (YDQV.0 0ROHFXODUJHQHWLFVDQGWKHQHJOHFWHGDUW of diatomics. 8QUDYHOLQJWKHDOJDH±WKHSDVWSUHVHQWDQGIXWXUHRIDOJDPROHFXODU systematics (ed. by J. Brodie and J.M. Lewis), pp. 321-265. CRC Press, London. 0DQQ'*6DWR67UREDMR59DQRUPHOLQJHQ3 6RXIIUHDX& '1$EDUFRGLQJIRUVSHFLHVLGHQWL¿FDWLRQDQGGLVFRYHU\LQGLDWRPVCryptogamie Algologie, 31, 557-577 0DQWHO1 'HWHFWLRQRIGLVHDVHFOXVWHULQJDQGDJHUQHUDOL]HG regression approach. Cancer Research, 27, 209-& Mantel, N. & Valand, R.S. (1970) A technique of non-parametric multivariate analysis. Biometrics, 26, 547 0DULYDX[/9LDQH\/LDXG0 -DHJHU-- +LJKOHYHOSK\ORJHQ\RI HDUO\7HUWLDU\URGHQWVGHQWDOHYLGHQFHZoological Journal of the Linnean Society, 142, 105-134 0DUVKDOO'- 3XJK3-$ 2ULJLQRIWKHLQODQG$FDULRI&RQWLQHQWDO $QWDUFWLFDZLWKSDUWLFXODUUHIHUHQFHWR'URQQLQJ0DXG/DQGZoological Journal of the Linnean Society, 118, 101-118 Marshall, W.A. & Chalmers, M.O. (1997) Airborne dispersal of antarctic terrestrial algae and cyanobacteria. (FRJUDSK\, 20, 585-594 Martiny, J.B.H., Bohannan, B.J.M., Brown, J.H., Colwell, R.K., Fuhrman, J.A., *UHHQ-/+RUQHU'HYLQH0&.DQH0.UXPLQV-$.XVNH&50RULQ3- Naeem, S., Ovreas, L., Reysenbach, A.L., Smith, V.H. & Staley, J.T. (2006) Microbial ELRJHRJUDSK\SXWWLQJPLFURRUJDQLVPVRQWKHPDSNature Reviews Microbiology, 4, 102-112 0DUYDOGL$(6HTXHLUD$62¶EULHQ&: )DUUHOO%' 0ROHFXODU DQGPRUSKRORJLFDOSK\ORJHQHWLFVRIZHHYLOV &ROHRSWHUD&XUFXOLRQRLGHD 'RQLFKH VKLIWVDFFRPSDQ\GLYHUVL¿FDWLRQ"Systematic Biology, 51, 761-785 Mathur, J. (2006) Trichome cell morphogenesis in ArabidopsisDFRQWLQXXP of cellular decisions. &DQDGLDQ-RXUQDORI%RWDQ\5HYXH&DQDGLHQQH'H%RWDQLTXH, 84, 604-612 0D[LPLQR& (YROXWLRQDU\FKDQJHVLQWKHFRPSOH[LW\RIWKHWHFWXP RIQRQWHWUDSRGV$FODGLVWLFDSSURDFKPlos One, 3, 7 0D\QDUG1* 6LJQL¿FDQFHRIDLUERUQHDOJDHZeitschrift Fur Allgemeine Mikrobiologie, 8, 225-&

256 References 0D\U( Animal speciation and their evolution. Harvard University Press, Cambridge. McQuoid, M.R. & Hobson, L.A. (1995) Importance of resting stages in diatom seasonal succession. Journal of Phycology, 31, 44-50 0F4XRLG05 +REVRQ/$ 'LDWRPUHVWLQJVWDJHVJournal of Phycology, 32, 889-902 Medlin, L.K., Kooistra, W., Gersonde, R., Sims, P.A. & Wellbrock, U. (1997a) ,VWKHRULJLQRIWKHGLDWRPVUHODWHGWRWKHHQG3HUPLDQPDVVH[WLQFWLRQ"Nova Hedwigia, 65, 1-11 0HGOLQ/..RRLVWUD:3RWWHU'6DXQGHUV*: $QGHUVHQ5$ E 3K\ORJHQHWLFUHODWLRQVKLSVRIWKHµJROGHQDOJDH¶ KDSWRSK\WHVKHWHURNRQW chromophytes) and their plastids. 3ODQW6\VWHPDWLFVDQG(YROXWLRQ, 187-219 Meier, F.C. & Lindbergh, C.A. (1935) Collecting microorganisms from the Arctic atmosphere. 7KH6FLHQWL¿F0RQWKO\, 40, 5-20 0HUHVFKNRZVN\& /HVW\SHVGHO¶HQGRFKURPHFKH]OHV'LDWRPpHV Scripta Botanica, Horti Universitalis Imperialis Petropolitanae, 21, 107-193 0LJXOD: 'LH9HUEUHLWXQJVZHLVHGHU$OJHQBiol. Zentr., 7, 514-517 0L]XQR0 (YROXWLRQRIFHQWULFGLDWRPVLQIHUUHGIURPSDWWHUQVRI oogenesis and spermatogenesis. Phycological Research, 56, 156-165 0RQDJKDQ07:LOG5(OOLRW0)XMLVDZD7%DONH0,QZDUG'- /HHV'&5DQDLYRVROR5(JJOHWRQ3%DUUDFORXJK7* 9RJOHU$3 Accelerated species inventory on Madagascar using coalescent-based models of species delineation. Systematic Biology, 58, 298-311 0RUURQH-- 2Q8GYDUG\¶V,QVXODQWDUFWLFDSURYLQFHDWHVWIURPWKH ZHHYLOV &ROHRSWHUD&XUFXOLRQRLGHD Journal of Biogeography, 25, 947-955 0RUWLPHU(9DQ9XXUHQ%-/HH-(0DUVKDOO'-&RQYH\3 Chown, S.L. (2011) Mite dispersal among the Southern Ocean Islands and $QWDUFWLFDEHIRUHWKHODVWJODFLDOPD[LPXPProceedings of the Royal Society B-Biological Sciences, 278, 1247-1255 0OOHU+$FKLOOHV'D\8(0 'D\-* 7ROHUDQFHRIWKHUHVWLQJ cysts of &ROSRGDLQÀDWD (Ciliophora, Colpodea) and Meseres corlissi (Ciliophora, Spirotrichea) to desiccation and freezing. (XURSHDQ-RXUQDORI3URWLVWRORJ\, 46, 133-142 Munilla, T. & Membrives, A.S. (2009) Check-list of the pycnogonids from $QWDUFWLFDQGVXE$QWDUFWLFZDWHUV]RRJHRJUDSKLFLPSOLFDWLRQVAntarctic Science, 21, 99-111 Nekola, J.C. & White, P.S. (1999) The distance decay of similarity in biogeography and ecology. Journal of Biogeography, 26, 867-878 Nicholson, W.L., Munakata, N., Horneck, G., Melosh, H.J. & Setlow, P. (2000) Resistance of BacillusHQGRVSRUHVWRH[WUHPHWHUUHVWULDODQGH[WUDWHUUHVWULDO environments. Microbiology and Molecular Biology Reviews, 64, 548-+ 2JQMDQRYD5XPHQRYD1 9DVV' 3DOHRHFRORJ\RIWKH/DWH Miocene maar lakes, Podrecany Basalt Formation, Southern Slovakia, on the basis of siliceous microfossils. Geologica Carpathica, 49, 351-368 Ohl, M. & Spahn, P. (2010) A cladistic analysis of the cockroach wasps EDVHGRQPRUSKRORJLFDOGDWD +\PHQRSWHUD$PSXOLFLGDH Cladistics, 26, 49-61

References 257 2NVDQHQ-%ODQFKHW%.LQGW5/HJHQGUH32¶+DUD5%6LPSVRQ */6RO\PRV36WHYHQV0++ :DJQHU+ 9HJDQ&RPPXQLW\ (FRORJ\3DFNDJHKWWS&5$15SURMHFWRUJSDFNDJH YHJDQ, R package version 1.17-6 Oldenhof, H., Wolkers, W.F., Bowman, J.L., Tablin, F. & Crowe, J.H. (2006) Freezing and desiccation tolerance in the moss Physcomitrella patens$Qin situ Fourier transform infrared spectroscopic study. %LRFKLPLFD(W%LRSK\VLFD$FWD General Subjects, 1760, 1226-1234 2OVRQ5-9DXORW' &KLVKROP6: (IIHFWVRIHQYLURQPHQWDO stresses on the cell-cylce of 2 marine phytoplankton species. Plant Physiology, 80, 918-925 2SSHQKHLP'5 7D[RQRPLFVWXGLHVRIAchnanthes (Bacillariophyta) in freshwater maritime Antarctic lakes. Canadian Journal of Botany-Revue &DQDGLHQQH'H%RWDQLTXH, 72, 1735-1748 Ozinga, W.A., Schaminee, J.H.J., Bekker, R.M., Bonn, S., Poschlod, P., Tackenberg, O., Bakker, J. & van Groenendael, J.M. (2005) Predictability of plant species composition from environmental conditions is constrained by dispersal limitation. Oikos, 108, 555-561 Paerl, H.W. (1988) Growth and reproductive strategies of freshwater blue- green algae (Cyanobacteria). Growth and reproductive strategies of freshwater phytoplankton HGE\&'6DQGJUHQ SS&DPEULGJH8QLYHUVLW\3UHVV Cambridge. Pagel, M. (1997) Inferring evolutionary processes from phylogenies. Zoologica Scripta, 26, 331-348 Pantocsek, J. (1886) %HLWUlJH]XU.HQQWQLVVGHUIRVVLOHQ%DFLOODULHQ 8QJDUQV,7KHLO0DULQH%DFLOODULHQ. J. Platzko, Nagy-Tapolscány, Hungary. Pantocsek, J. (1889) %HLWUlJH]XU.HQQWQLVVGHUIRVVLOHQ%DFLOODULHQ Ungarns. Teil II Brackwasser Bacillarien. Julíus Platzko, Nagy-Tapolcsány, Hungary. 3DSNH575DPVLQJ1%%DWHVRQ00 :DUG'0 Geographical isolation in hot spring cyanobacteria. (QYLURQPHQWDO0LFURELRORJ\, 5, 650-659 3DUVRQV:06FKOLFKWLQJ+( 6WHZDUW.: ,QÀLJKWWUDQVSRUWRI algae and protozoa by selected Odonata. Transaction of the American Microscopic Society, 85, 520-527 Patrick, R. & Reimer, C.W. (1966) The diatoms of the United States, H[FOXVLYHRI$ODVNDDQG+DZDLL9ROXPH)UDJLODULDFHDH(XQRWLDFHDH Achnanthaceae, Naviculaceae. 3DWULFN5 (FRORJ\RIIUHVKZDWHUGLDWRPVDQGGLDWRPFRPPXQLWLHV The biology of diatoms HGE\':HUQHU SS%ODFNZHOO6FLHQWL¿F 3XEOLFDWLRQV2[IRUG Patterson, C. (1982) Morphological characters and homology. Problems of Phylogenetic Reconstruction (ed. by G.P. Wagner), pp. 21-74. Academic Press, London. 3HUNLQV5*/DYDXG-6HURGLR-0RXJHW-/&DUWD[DQD35RVD P., Barille, L., Brotas, V. & Jesus, B.M. (2010) Vertical cell movement is a primary UHVSRQVHRILQWHUWLGDOEHQWKLFELR¿OPVWRLQFUHDVLQJOLJKWGRVH0DULQH(FRORJ\ Progress Series, 416, 93-103 Persson, A. (2002) Proliferation of cryptic protists and germination of UHVWLQJVWDJHVIURPXQWUHDWHGVHGLPHQWVDPSOHVZLWKHPSKDVLVRQGLQRÀDJHOODWHV Ophelia, 55, 151-166

258 References 3HWHUVHQ-% 6WXGLHVRQWKHELRORJ\DQGWD[RQRP\RIVRLODOJDH 'DQVN%RW$UNLY, 8, 1-180 3HW]:9DOERQHVL$6FKLIWQHU84XHVDGD$ (OOLV(YDQV-& &LOLDWHELRJHRJUDSK\LQ$QWDUFWLFDQG$UFWLFIUHVKZDWHUHFRV\VWHPVHQGHPLVPRU global distribution of species? )HPV0LFURELRORJ\(FRORJ\, 59, 396-408 Pimentel, R.A. & Riggins, R. (1987) The nature of cladistic data. Cladistics, 3, 201-209 Pommier, T., Canback, B., Riemann, L., Bostrom, K.H., Simu, K., Lundberg, P., Tunlid, A. & Hagstrom, A. (2007) Global patterns of diversity and community structure in marine bacterioplankton. 0ROHFXODU(FRORJ\, 16, 867-880 3RQV-%DUUDFORXJK7**RPH]=XULWD-&DUGRVR$'XUDQ'3 +D]HOO6.DPRXQ66XPOLQ:' 9RJOHU$3 6HTXHQFHEDVHGVSHFLHV GHOLPLWDWLRQIRUWKH'1$WD[RQRP\RIXQGHVFULEHGLQVHFWVSystematic Biology, 55, 595-609 3RWDSRYD0* &KDUOHV') %HQWKLFGLDWRPVLQ86$ULYHUV distributions along spatial and environmental gradients. Journal of Biogeography, 29, 167-187 Potts, M. (1999) Mechanisms of desiccation tolerance in cyanobacteria. (XURSHDQ-RXUQDORI3K\FRORJ\, 34, 319-328 3RXOLþNRYi$/\VDNRYD0+DVOHU3 /HONRYD( )LVKSRQG sediments - the source of palaeoecological information and algal “seed banks”. Nova Hedwigia, 86, 141-153 3ULQJOH$%DNHU'03ODWW-/:DUHV-3/DWJH-3 7D\ORU-: (2005) Cryptic speciation in the cosmopolitan and clonal human pathogenic fungus Aspergillus fumigatus. (YROXWLRQ, 59, 1886-1899 3URFWRU9: 'LVSHUVDORIGHVPLGVE\ZDWHUELUGVPhycologia, 5, 227-232 3XJK3-$ &RQYH\3 6XUYLYLQJRXWLQWKHFROG$QWDUFWLF endemic invertebrates and their refugia. Journal of Biogeography, 35, 2176-2186 Qian, P.P., Hou, S.W. & Guo, G.Q. (2009) Molecular mechanisms controlling pavement cell shape in Arabidopsis leaves. Plant Cell Reports, 28, 1147-1157 4XD\OH:&3HFN/63HDW+(OOLV(YDQV-& +DUULJDQ35 ([WUHPHUHVSRQVHVWRFOLPDWHFKDQJHLQ$QWDUFWLFODNHVScience, 295, 645-645 5DMHQGUDQ80.DWKLUYHO( $QDQG1 'HVLFFDWLRQ LQGXFHGFKDQJHVLQDQWLR[LGDQWHQ]\PHVIDWW\DFLGVDQGDPLQRDFLGVLQWKH cyanobacterium 7RO\SRWKUL[VF\WRQHPRLGHV. World Journal of Microbiology & Biotechnology, 23, 251-257 5DPEDXW$ 'UXPPRQG$- 7UDFHU$YDLODEOHIURP KWWS EHDVWELRHGDFXNWUDFHU! 5HQDXG607KLQK/9/DPEULQLGLV* 3DUU\'/ (IIHFWRI temperature on growth, chemical composition and fatty acid composition of tropical Australian microalgae grown in batch cultures. Aquaculture, 211, 195-214 5LFH:5 +RVWHUW(( /DERUDWRU\H[SHULPHQWVRQVSHFLDWLRQ What have we learned in 40 years. (YROXWLRQ, 47, 1637-1653 Rieppel, O. & Kearney, M. (2002) Similarity. Biological Journal of the Linnean Society, 75, 59-82 5RGUtJXH]7UHOOHV)7DUUtR5 $\DOD)- 0ROHFXODUFORFNV whence and whither? 7HOOLQJWKHHYROXWLRQDU\WLPHPROHFXODUFORFNVDQGWKHIRVVLO record HGE\0-'RQRJKXHDQG036PLWK SS7D\ORU )UDQFLV/RQGRQ

References 259 Rokas, A., Williams, B.L., King, N. & Carroll, S.B. (2003) Genome-scale approaches to resolving incongruence in molecular phylogenies. Nature, 425, 798-804 5RQTXLVW) +XHOVHQEHFN-3 0U%D\HV%D\HVLDQSK\ORJHQHWLF LQIHUHQFHXQGHUPL[HGPRGHOVBioinformatics, 19, 1572-1574 5RVV5&R[(-.DUD\HYD1,0DQQ'*3DGGRFN7%%6LPRQVHQ R. & Sims, P.A. (1979) An amended terminology for the siliceous components of the diatom cell. Nova Hedwigia Beihefte, 64, 513-533 5RVVLWHU1 .DZDQDEH+ $QFLHQWODNHV%LRGLYHUVLW\HFRORJ\ and evolution - Advances in ecological research - Preface. $GYDQFHVLQ(FRORJLFDO Research, Vol 31, pp. XII-XVI. Academic Press Ltd, London. 5RXQG)( %XNKWL\DURYD/ )RXUQHZJHQHUDEDVHGRQ Achnanthes (Achnanthidium WRJHWKHUZLWKDUHGL¿QLWLRQRIAchnanthidium. 'LDWRP5HVHDUFK, 11, 245-261 5RXQG)( %HQWKLFPDULQHGLDWRPVOceanography and Marine %LRORJ\DQ$QQXDO5HYLHZ, 9, 83-139 5RXQG)( The ecology of algae. Cambrige University Press, Cambridge. 5RXQG)(&UDZIRUG50 0DQQ'* 7KHGLDWRPVELRORJ\DQG morphology of the genera. Cambridge University Press. 5RXVFK-0%LQJKDP6( 6RPPHUIHOG05 &KDQJHVLQIDWW\ DFLGSUR¿OHVRIWKHUPRLQWROHUDQWDQGWKHUPRWROHUDQWPDULQHGLDWRPVGXULQJ temperature stress. -RXUQDORI([SHULPHQWDO0DULQH%LRORJ\DQG(FRORJ\, 295, 145-156 5RXVFK-0%LQJKDP6( 6RPPHUIHOG05 3URWHLQH[SUHVVLRQ during heat stress in thermo-intolerant and thermo-tolerant diatoms. Journal of ([SHULPHQWDO0DULQH%LRORJ\DQG(FRORJ\, 306, 231-243 Roy-Ocotla, G. & Carrera, J. (1993) Aeroalgae - responses to some aerobiological questions. Grana, 32, 48-56 5\QHDUVRQ7$/LQ(2 $UPEUXVW(9 0HWDSRSXODWLRQVWUXFWXUH in the planktonic diatom 'LW\OXPEULJKWZHOOLL (Bacillariophyceae). Protist, 160, 111-121 6DEDFND0 (OVWHU- 5HVSRQVHRIF\DQREDFWHULDDQGDOJDHIURP Antarctic wetland habitats to freezing and desiccation stress. Polar Biology, 30, 31-37 Saint Martin, S. & Saint Martin, J.P. (2005) The diatom assemblages as environmental evolution recording of the Paratethys area during Sarmatian times. Comptes Rendus Palevol, 4, 191-201 6DQGHUVRQ0- (VWLPDWLQJUDWHVRIVSHFLDWLRQDQGHYROXWLRQD bias due to homoplasy. Cladistics-the International Journal of the Willi Hennig Society, 6, 387-391 6DQGHUVRQ0- (VWLPDWLQJDEVROXWHUDWHVRIPROHFXODUHYROXWLRQ DQGGLYHUJHQFHWLPHV$SHQDOL]HGOLNHOLKRRGDSSURDFKMolecular Biology and (YROXWLRQ, 19, 101-109 6DQGHUVRQ0- UVLQIHUULQJDEVROXWHUDWHVRIPROHFXODUHYROXWLRQ and divergence times in the absence of a molecular clock. Bioinformatics, 19, 301-302

260 References 6DUQR'.RRLVWUD:0HGOLQ/.3HUFRSR, =LQJRQH$ 'LYHUVLW\LQWKHJHQXVSkeletonema (Bacillariophyceae). II. An assessment of the WD[RQRP\RIS. costatum-like species with the description of four new species. Journal of Phycology, 41, 151-176 6DUQR'.RRLVWUD:%DO]DQR6+DUJUDYHV3( =LQJRQH$ 'LYHUVLW\LQWKHJHQXV Skeletonema %DFLOODULRSK\FHDH ,,,3K\ORJHQHWLFSRVLWLRQ and morphological variability of Skeletonema costatum and Skeletonema grevillei, with the description of Skeletonema ardens sp. nov. Journal of Phycology, 43, 156-170 Savolainen, V., Cowan, R.S., Vogler, A.P., Roderick, G.K. & Lane, R. (2005) 7RZDUGVZULWLQJWKHHQF\FORSDHGLDRIOLIHDQLQWURGXFWLRQWR'1$EDUFRGLQJ Philosophical Transactions of the Royal Society B-Biological Sciences, 360, 1805- 1811 Scheckenbach, F., Wylezich, C., Weitere, M., Hausmann, K. & Arndt, H. 0ROHFXODULGHQWLW\RIVWUDLQVRIKHWHURWURSKLFÀDJHOODWHVLVRODWHGIURP VXUIDFHZDWHUVDQGGHHSVHDVHGLPHQWVRIWKH6RXWK$WODQWLFEDVHGRQ668U'1$ $TXDWLF0LFURELDO(FRORJ\, 38, 239-247 Schelske, C.L., Carrick, H.J. & Aldridge, F.J. (1995) Can wind-induced resuspension of meroplankton affect phytoplankton dynamics? Journal of the North American Benthological Society, 14, 616-630 6FKOLFKWLQJ+(- 7KHUROHRIZDWHUIRZOLQWKHGLVSHUVDORIDOJDH Transaction of the American Microscopic Society, 79, 160-166 6FKOLFKWLQJ+(- 9LDEOHVSHFLHVRIDOJDHDQGSURWR]RDLQWKH atmosphere. Lloydia, 24, 81-& 6FKOLFKWLQJ+(- 0HWHRURORJLFDOFRQGLWLRQVDIIHFWLQJGLSVHUVDORI airborne algae + protozoa. Lloydia, 27, 64 6FKOLFKWLQJ+(- 7KHLPSRUWDQFHRIDLUERUQHDOJDHDQGSURWR]RD Air Pollution Control Association, 19, 946-951 6FKPLG$00 ,QÀXHQFHRIHQYLURQPHQWDOIDFWRUVRQWKH development of the valve in diatoms. Protoplasma, 99, 99-115 Schmid, A.M.M. (2001) Value of pyrenoids in the systematics of WKHGLDWRPVWKHLUPRUSKRORJ\DQGXOWUDVWUXFWXUHProceedings of the 16th international diatom symposium, pp. 1-32. Schmid, A.M.M. (2009) Induction of resting-spores in the pennate diatom Navicula (Craticula) cuspidata by uncoupling of the cell and plastid cycles. Nova Hedwigia, 85-101 6FKPLGW50lXVEDFKHU5 0OOHU- +RORFHQHGLDWRPÀRUDDQG stratigraphy from sediment cores of two Antarctic lakes (King George Island). Journal of Paleolimnology, 3, 55-74 6FKROLQ&$+HU]RJ06RJLQ0 $QGHUVRQ'0 ,GHQWL¿FDWLRQ RIJURXSVSHFL¿FDQGVWUDLQVSHFL¿FJHQHWLFPDUNHUVIRUJOREDOO\GLVWULEXWHG $OH[DQGULXP 'LQRSK\FHDH 6HTXHQFHDQDO\VLVRIDIUDJPHQWRIWKH/68 ribosomal RNA gene. Journal of Phycology, 30, 999-1011 6FKROQLN'$ 6HDVRQDOYDULDWLRQDQGGLXUQDOÀXFWXDWLRQLQ ephemeral desert pools. Hydrobiologia, 294, 111-116 6FKZDU]* (VWLPDWLQJGLPHQVLRQRIDPRGHOAnnals of Statistics, 6, 461-464 Scotland, R.W. & Pennington, R.T. (2000) +RPRORJ\DQGV\VWHPDWLFV coding characters for phylogenetic analysis. Taylor and Francis, London.

References 261 Scotland, R.W., Olmstead, R.G. & Bennett, J.R. (2003) Phylogeny UHFRQVWUXFWLRQ7KHUROHRIPRUSKRORJ\Systematic Biology, 52, 539-548 Servant-Vildary, S., Paicheler, J.C. & Semelin, B. (1988) Miocene lacustrine diatoms from Turkey. 3URFHHGLQJVRIWKHWK,QWHUQDWLRQDO'LDWRP6\PSRVLXP Bristol, 1986, pp. 165-180. Biopress Ltd., Koenigstein. 6KDUPD1.5DL$.6LQJK6 %URZQ50 $LUERUQHDOJDH Their present status and relevance. Journal of Phycology, 43, 615-627 6KHSKDUG./ (YDSRUDWLRQRIZDWHUIURPWKHPXFLODJHRID gelatinous algal community. British Phycological Journal, 22, 181-185 Shurin, J.B., Cottenie, K. & Hillebrand, H. (2009) Spatial autocorrelation and dispersal limitation in freshwater organisms. Oecologia, 159, 151-159 6LFNR*RDG/6WRHUPHU() )DKQHQVWLHO* 5HMXYHQDWLRQRI Melosira granulata %DFLOODULRSK\FHDH UHVWLQJFHOOVIURPWKHDQR[LFVHGLPHQWV RI'RXJODVVODNH0LFKLJDQ/LJKWPLFURVFRS\DQG&XSWDNHJournal of Phycology, 22, 22-28 6LPRQ(01DQQH\'/ 'RHUGHU)3 7KH³Tetrahymena pyriformis´FRPSOH[RIFU\SWLFVSHFLHVBiodiversity and Conservation, 17, 365-380 6LPV3$0DQQ'* 0HGOLQ/. (YROXWLRQRIWKHGLDWRPV insights from fossil, biological and molecular data. Phycologia, 45, 361-402 Sinclair, B.J., Addo-Bediako, A. & Chown, S.L. (2003) Climatic variability and the evolution of insect freeze tolerance. Biological Reviews, 78, 181-195 6LQJK566WRHUPHU() .DU5 (DUOLHVWIUHVKZDWHU diatoms from the deccan intertrappean (Maastrichtian) sediments of India. Micropaleontology, 52, 545-551 Siver, P.A. & Wolfe, A.P. (2007) (XQRWLDVSS(Bacillariophyceae) from 0LGGOH(RFHQHODNHVHGLPHQWVDQGFRPPHQWVRQWKHRULJLQRIWKHGLDWRPUDSKH &DQDGLDQ-RXUQDORI%RWDQ\5HYXH&DQDGLHQQH'H%RWDQLTXH, 85, 83-90 6LYHU3$:ROIH$3 (GOXQG0% 7D[RQRPLFGHVFULSWLRQVDQG HYROXWLRQDU\LPSOLFDWLRQVRI0LGGOH(RFHQHSHQQDWHGLDWRPVUHSUHVHQWLQJWKH H[WDQWJHQHUD2[\QHLV, Actinella and Nupela (Bacillariophyceae) 3ODQW(FRORJ\DQG (YROXWLRQ, 143, 340-351 Skottsberg, C. (1960) Remarks on the plant geography of the southern cold temperate zone. Proceedings of the Royal Society of London Series B-Biological Sciences, 152, 447-457 âODSHWD-/RSH]*DUFLD3 0RUHLUD' *OREDOGLVSHUVDODQG ancient cryptic species in the smallest marine eukaryotes. Molecular Biology and (YROXWLRQ, 23, 23-29 Slobodkin, L.B. (1954) Population dynamics in 'DSKQLDREWXVD Kurz. (FRORJLFDO0RQRJUDSKV, 24, 69-88 6PLWK5(+6WDSOHIRUG/& 5LGLQJV56 7KHDFFOLPDWHG response of growth, photosynthesis, composition, and carbon balance to temperature in the psychrophilic ice diatom Nitzschia seriata. Journal of Phycology, 30, 8-16 6PLWK1' 7XUQHU$+ 0RUSKRORJ\¶VUROHLQSK\ORJHQ\ UHFRQVWUXFWLRQ3HUVSHFWLYHVIURPSDOHRQWRORJ\Systematic Biology, 54, 166-173 6RLQLQHQ-0F'RQDOG5 +LOOHEUDQG+ 7KHGLVWDQFHGHFD\RI similarity in ecological communities. (FRJUDSK\, 30, 3-12

262 References Soininen, J., Paavola, R. & Muotka, T. (2004) Benthic diatom communities LQERUHDOVWUHDPVFRPPXQLW\VWUXFWXUHLQUHODWLRQWRHQYLURQPHQWDODQGVSDWLDO gradients. (FRJUDSK\, 27, 330-342 Sokal, R.R. & Rohlf, F.J. (2000) %LRPHWU\WKHSULQFLSOHVDQGSUDFWLFHRI statistics in biological research, 3 edn. W.H. Freeman, New York. 6RUKDQQXV8 'LDWRPSK\ORJHQHWLFVLQIHUUHGEDVHGRQGLUHFW optimization of nuclear-encoded SSU rRNA sequences. Cladistics-the International Journal of the Willi Hennig Society, 20, 487-497 Sorhannus, U. (2007) A nuclear-encoded small-subunit ribosomal RNA timescale for diatom evolution. Marine Micropaleontology, 65, 1-12 6RUKDQQXV82UWL]-':ROI0 )R[0* 0LFURHYROXWLRQDQG speciation in 7KDODVVLRVLUDZHLVVÀRJLL (Bacillariophyta). Protist, 161, 237-249 6RXIIUHDX&9DQRUPHOLQJHQ39HUOH\HQ(6DEEH. 9\YHUPDQ W. (2010) Tolerance of benthic diatoms from temperate aquatic and terrestrial KDELWDWVWRH[SHULPHQWDOGHVLFFDWLRQDQGWHPSHUDWXUHVWUHVVPhycologia, 49, 309- 324 6RXIIUHDX&&R[(-6DEEH. 9\YHUPDQ: XQSXEOD 3K\ORJHQHWLF signals in the valve and plastid morphology of the diatom genus Pinnularia. Souffreau, C., Vanormelingen, P. & Vyverman, W. (unpubl. b) Resistance of IUHVKZDWHUEHQWKLFGLDWRPUHVWLQJVWDJHVWRH[SHULPHQWDOGHVLFFDWLRQDQGIUHH]LQJ depends on the temporality of their habitat. Souffreau, C., Vanormelingen, P., Van de Vijver, B., Isheva, T., Verleyen, (6DEEH. 9\YHUPDQ: XQSXEOF 0ROHFXODUHYLGHQFHIRUWKHH[LVWHQFH of distinct Antarctic lineages in the cosmopolitan terrestrial diatoms Pinnularia borealis and +DQW]VFKLDDPSKLR[\V. Souffreau, C., Verbruggen, H., Wolfe, A.P., Vanormelingen, P., Siver, P.A., &R[(-0DQQ'*9DQ'H9LMYHU%6DEEH. 9\YHUPDQ: VXEPLWWHG $ time-calibrated multi-gene phylogeny of the diatom genus Pinnularia. Submitted to 0ROHFXODU3K\ORJHQHWLFVDQG(YROXWLRQ 6SDXOGLQJ6$.RFLROHN-3 :RQJ' $WD[RQRPLFDQG systematic revision of the genus Muelleria (Bacillariophyta). Phycologia, 38, 314- 341 6SDXOGLQJ6$9DQGH9LMYHU%+RGJVRQ$0F.QLJKW'09HUOH\HQ ( 6WDQLVK/ 'LDWRPVDVLQGLFDWRUVRIHQYLURQPHQWDOFKDQJHLQ$QWDUFWLF and subantarctic fresh waters. 7KHGLDWRPVDSSOLFDWLRQVIRUWKHHQYLURQPHQWDO and earth sciences HGE\-36PRODQG()6WRHUPHU SS 6WDPDWDNLV$ 5$[0/9,+3&0D[LPXPOLNHOLKRRGEDVHG SK\ORJHQHWLFDQDO\VHVZLWKWKRXVDQGVRIWD[DDQGPL[HGPRGHOVBioinformatics, 22, 2688-2690 6WDUN-% 2¶JUDG\30 0RUSKRORJLFDOYDULDWLRQLQWKHIRUHOHJV RIWKH+DZDLLDQ'URVRSKLOLGDH,7KH$0&&ODGHJournal of Morphology, 271, 86-103 6WDUNV7/6KXEHUW/( 7UDLQRU)5 (FRORJ\RIVRLODOJDHD review. Phycologia, 20, 65-80 6WHYHQV0,*UHHQVODGH3+RJJ,' 6XQQXFNV3 6RXWKHUQ KHPLVSKHUHVSULQJWDLOV&RXOGDQ\KDYHVXUYLYHGJODFLDWLRQRI$QWDUFWLFD" 0ROHFXODU%LRORJ\DQG(YROXWLRQ, 23, 874-882 6XQJ'<.DSODQ)/HH.- *X\&/ $FTXLUHGWROHUDQFHWR WHPSHUDWXUHH[WUHPHVTrends in Plant Science, 8, 179-187

References 263 6XWKHUODQG-0+HUGPDQ0 6WHZDUW:'3 $NLQHWHVRIWKH cyanobacterium Nostoc-Pcc-7524. Macromolecular composition, structure and control of differentiation. Journal of General Microbiology, 115, 273-287 Suzuki, Y. & Takahashi, M. (1995) Growth responses of several diatom species isolated from various environments to temperature. Journal of Phycology, 31, 880-888 6ZRIIRUG'/ 3$83 3K\ORJHQHWLF$QDO\VLV8VLQJ3DUVLPRQ\ DQG other methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Talling, J.F. (1951) The element of chance in pond populations. The Naturalist, 1951, 157-170 Tamaru, Y., Takani, Y., Yoshida, T. & Sakamoto, T. (2005) Crucial role of H[WUDFHOOXODUSRO\VDFFKDULGHVLQGHVLFFDWLRQDQGIUHH]LQJWROHUDQFHLQWKHWHUUHVWULDO cyanobacterium Nostoc commune. $SSOLHGDQG(QYLURQPHQWDO0LFURELRORJ\, 71, 7327-7333 7DWRQ$*UXELVLF6%UDPELOOD('H:LW5 :LOPRWWH$ &\DQREDFWHULDOGLYHUVLW\LQQDWXUDODQGDUWL¿FLDOPLFURELDOPDWVRI/DNH)U\[HOO 0F0XUGRGU\YDOOH\V$QWDUFWLFD $PRUSKRORJLFDODQGPROHFXODUDSSURDFK $SSOLHGDQG(QYLURQPHQWDO0LFURELRORJ\, 69, 5157-5169 7DWRQ$*UXELVLF6%DOWKDVDUW3+RGJVRQ'$/D\ERXUQ3DUU\- Wilmotte, A. (2006) Biogeographical distribution and ecological ranges of benthic F\DQREDFWHULDLQ(DVW$QWDUFWLFODNHV)HPV0LFURELRORJ\(FRORJ\, 57, 272-289 Taylor, F. (1980) Optimal switching to diapause in relation to the onset of winter. Theoretical Population Biology, 18, 125-133 7D\ORU-:7XUQHU(7RZQVHQG-3'HWWPDQ-5 -DFREVRQ' (XNDU\RWLFPLFUREHVVSHFLHVUHFRJQLWLRQDQGWKHJHRJUDSKLFOLPLWVRI VSHFLHVH[DPSOHVIURPWKHNLQJGRP)XQJLPhilosophical Transactions of the Royal Society B-Biological Sciences, 361, 1947-1963 7HOIRUG5-9DQGYLN9 %LUNV+-% 'LVSHUVDOOLPLWDWLRQVPDWWHU for microbial morphospecies. Science, 312, 1015-1015 Telford, R.J., Vandvik, V. & Birks, H.J.B. (2006b) How many freshwater diatoms are pH specialists? A response to Pither & Aarssen (2005). (FRORJ\ Letters, 9(( Templeton, A.R. (2008) The reality and importance of founder speciation in evolution. Bioessays, 30, 470-479 Ter Braak, C.J.F., Smilauer, P. (2002) CANOCO reference manual and &DQR'UDZIRU:LQGRZVXVHU¶VJXLGHVRIWZDUHIRUFDQRQLFDOFRPPXQLW\RUGLQDWLRQ (version 4.5). . Microcomputer Power, Ithaca, New York, USA. 7KHULRW(&)ULW]6&:KLWORFN& &RQOH\'- /DWH4XDWHUQDU\ rapid morphological evolution of an endemic diatom in Yellowstone Lake, Wyoming. Paleobiology, 32, 38-54 7KRPDV'1%DXPDQQ0(0 *OHLW]0 (I¿FLHQF\RIFDUERQ assimilation and photoacclimation in a small unicellular Chaetoceros species from WKH:HGGHOO6HD $QWDUFWLFD LQÀXHQFHRIWHPSHUDWXUHDQGLUUDGLDQFHJournal of ([SHULPHQWDO0DULQH%LRORJ\DQG(FRORJ\, 157, 195-209 7KRPSVRQ-'+LJJLQV'* *LEVRQ7- &OXVWDO:,PSURYLQJ the sensitivity of progressive multiple sequence alignment through sequence ZHLJKWLQJSRVLWLRQVSHFL¿FJDSSHQDOWLHVDQGZHLJKWPDWUL[FKRLFHNucleic Acids Research, 22, 4673-4680

264 References Tomanek, L. (2008) The importance of physiological limits in determining ELRJHRJUDSKLFDOUDQJHVKLIWVGXHWRJOREDOFOLPDWHFKDQJHWKHKHDWVKRFN response. Physiological and Biochemical Zoology, 81, 709-717 7RUPR55HFLR'6LOYD, 0XQR]$) $TXDQWLWDWLYH investigation of airborne algae and lichen soredia obtained from pollen traps in south-west Spain. (XURSHDQ-RXUQDORI3K\FRORJ\, 36, 385-390 7UREDMR50DQQ'*&KHSXUQRY9$&ODYHUR( &R[(- 7D[RQRP\OLIHF\FOHDQGDX[RVSRUXODWLRQRINitzschia fonticola (Bacillariophyta). Journal of Phycology, 42, 1353-1372 7UREDMR5&ODYHUR(&KHSXUQRY9$6DEEH.0DQQ'*,VKLKDUD 6 &R[(- 0RUSKRORJLFDOJHQHWLFDQGPDWLQJGLYHUVLW\ZLWKLQWKH widespread bioindicator Nitzschia palea (Bacillariophyceae). Phycologia, 48, 443- 459 Tronholm, A., Steen, F., Tyberghein, L., Leliaert, F., Verbruggen, H., Siguan, 0$5 GH&OHUFN2 6SHFLHVGHOLPLWDWLRQWD[RQRP\DQGELRJHRJUDSK\RI 'LFW\RWDLQ(XURSH 'LFW\RWDOHV3KDHRSK\FHDH Journal of Phycology, 46, 1301- 1321 7XRPLVWR+5XRNRODLQHQ.

References 265 Van der Gucht, K., Cottenie, K., Muylaert, K., Vloemans, N., Cousin, S., 'HFOHUFN6-HSSHVHQ(&RQGH3RUFXQD-06FKZHQN.=ZDUW*'HJDQV +9\YHUPDQ: 'H0HHVWHU/ 7KHSRZHURIVSHFLHVVRUWLQJ/RFDO factors drive bacterial community composition over a wide range of spatial scales. Proceedings of the National Academy of Sciences of the United States of America, 104, 20404-20409 9DQGHU3XWWHQ19HUEUXJJHQ&2FK\UD59HUOH\HQ( )UHQRW < 6XEDQWDUFWLFÀRZHULQJSODQWVSUHJODFLDOVXUYLYRUVRUSRVWJODFLDO immigrants? Journal of Biogeography, 37, 582-592 9DQ+DQQHQ(-0RRLM:09DQ$JWHUYHOG03*RQV+- /DDQEURHN+- 'HWULWXVGHSHQGHQWGHYHORSPHQWRIWKHPLFURELDO FRPPXQLW\LQDQH[SHULPHQWDOV\VWHP4XDOLWDWLYHDQDO\VLVE\GHQDWXULQJJUDGLHQW gel electrophoresis. $SSOLHGDQG(QYLURQPHQWDO0LFURELRORJ\, 65, 2478-2484 Van Kerckvoorde, A., Trappeniers, K., Nijs, I. & Beyens, L. (2000) Terrestrial soil diatom assemblages from different vegetation types in Zackenberg (Northeast Greenland). Polar Biology, 23, 392-400 Van Overeem, M.A. (1937) On green organisms occuring in the lower troposhere. 5HFXHLOGHV7UDYDX[%RWDQLTXH1HHUODQGDLV, 34, 388-442 9DQHOVODQGHU%&UHDFK99DQRUPHOLQJHQ3(UQVW$&KHSXUQRY9$ 6DKDQ(0X\]HU*6WDO/-9\YHUPDQ: 6DEEH. (FRORJLFDO differentiation between sympatric pseudocryptic species in the estuarine benthic diatom Navicula phyllepta (Bacillariophyceae). Journal of Phycology, 45, 1278- 1289 9DQHOVODQGHU%5XsGLJHU).DUVWHQ86DEEH. 9\YHUPDQ: (unpubl.) Thermal niche evolution in the marine diatom genus Cylindrotheca. Vanlandingham, S.L. (1991) Precision dating by means of traditional biostratigraphic methods for the middle Miocene diatomaceous interbeds within the middle Yakima (Wanapum) Basalt of south-central Washington (U.S.A.). Nova Hedwigia, 53, 349-368 9DQRUPHOLQJHQ3&KHSXUQRY9$0DQQ'*6DEEH. 9\YHUPDQ W. (2008a) Genetic divergence and reproductive barriers among morphologically heterogeneous sympatric clones of (XQRWLDELOXQDULVVHQVXODWR (Bacillariophyta). Protist, 159, 73-90 9DQRUPHOLQJHQ39HUOH\HQ( 9\YHUPDQ: E 7KHGLYHUVLW\DQG GLVWULEXWLRQRIGLDWRPVIURPFRVPRSROLWDQLVPWRQDUURZHQGHPLVPBiodiversity and Conservation, 17, 393-405 9DXORW'2OVRQ5- &KLVKROP6: /LJKWDQGGDUNFRQWURORI the cell-cylce in 2 marine phytoplankton species. ([SHULPHQWDO&HOO5HVHDUFK, 167, 38-52 Veech, J.A., Summerville, K.S., Crist, T.O. & Gering, J.C. (2002) The DGGLWLYHSDUWLWLRQLQJRIVSHFLHVGLYHUVLW\UHFHQWUHYLYDORIDQROGLGHDOikos, 99, 3-9 9HHFK-$ &KULVW72 3$57,7,21VRIWZDUHIRUKLHUDUFKLFDO partitioning of species diversity, version 3.0. ZZZXVHUVPXRKLRHGXFULVWWR partition.htm Velasques, G.T. (1940) On the viability of algae obtained from he digestive tract of the Gizzad Shad, 'RURVRPDFHSHGLDQXP. American Midland Naturalist, 22, 376-412

266 References Verbruggen, H., Tyberghein, L., Pauly, K., Vlaeminck, C., Van 1LHXZHQKX\]H..RRLVWUD:/HOLDHUW) 'H&OHUFN2 0DFURHFRORJ\ PHHWVPDFURHYROXWLRQHYROXWLRQDU\QLFKHG\QDPLFVLQWKHVHDZHHGHalimeda. *OREDO(FRORJ\DQG%LRJHRJUDSK\, 18, 393-405 9HUOH\HQ(9\YHUPDQ:6WHUNHQ0+RGJVRQ'$'H:HYHU $ -XJJLQV69DQGH9LMYHU%-RQHV9-9DQRUPHOLQJHQ35REHUWV')ORZHU R., Kilroy, C., Souffreau, C. & Sabbe, K. (2009) The importance of dispersal related DQGORFDOIDFWRUVLQVKDSLQJWKHWD[RQRPLFVWUXFWXUHRIGLDWRPPHWDFRPPXQLWLHV Oikos, 118, 1239-1249 9HUOH\HQ(9DQGH9LMYHU%6RXIIUHDX&6DEEH.+RGJVRQ'$ 1HGEDORYD/9DQRUPHOLQJHQ3$QWRQLDGHV' 9\YHUPDQ: XQSXEO 3ROHV DSDUWLQWHUKHPLVSKHULFFRQWUDVWVLQ$UFWLFDQG$QWDUFWLFGLDWRPGLYHUVLW\ 9LFUH0)DUUDQW-0 'ULRXLFK$ ,QVLJKWVLQWRWKHFHOOXODU mechanisms of desiccation tolerance among angiosperm resurrection plant species. 3ODQW&HOODQG(QYLURQPHQW, 27, 1329-1340 9LOODQXHYD9' 'LDWRPGLVWULEXWLRQLQDWHPSRUDU\SRQGIURPWKH Patagonian Andes (Fantasma Pond) during drought. Nova Hedwigia, 83, 459-472 9OHHVKRXZHUV/0%RXZPHHVWHU+- .DUVVHQ&0 5HGH¿QLQJ VHHGGRUPDQF\$QDWWHPSWWRLQWHJUDWHSK\VLRORJ\DQGHFRORJ\Journal of (FRORJ\, 83, 1031-1037 9RQ6WRVFK+$ )HFKHU. µ,QWHUQDOWKHFDH¶RI(XQRWLDVROHUROLL %DFLOODULRSK\FHDH GHYHORSPHQWVWUXFWXUHDQGIXQFWLRQDVUHVWLQJVSRUHV Journal of Phycology, 15, 223-243 9\YHUPDQ:6DEEH.0DQQ'*.LOUR\&9\YHUPDQ59DQKRXWWH . +RGJVRQ' (XQRSKRUDJHQQRY(Bacillariophyta) from Tasmania and 1HZ=HDODQGGHVFULSWLRQDQGFRPSDULVRQZLWK(XQRWLDand amphoroid diatoms. (XURSHDQ-RXUQDORI3K\FRORJ\, 33, 95-111 9\YHUPDQ:9HUOH\HQ(6DEEH.9DQKRXWWH.6WHUNHQ0 +RGJVRQ'$0DQQ'*-XJJLQV6'H9LMYHU%9-RQHV9)ORZHU5 5REHUWV'&KHSXUQRY9$.LOUR\&9DQRUPHOLQJHQ3 'H:HYHU$ Historical processes constrain patterns in global diatom diversity. (FRORJ\, 88, 1924-1931 :DJQHU*3 &KDUDFWHUVXQLWVDQGQDWXUDONLQGVDQLQWURGXFWLRQ The character concept in evolutionary biology (ed. by G.P. Wagner), pp. 1-10. $FDGHPLF3UHVV6DQ'LHJR Wallace, A.R. (1876) The geographical distribution of animals. McMillan, London. :DUG%% 2¶0XOODQ*' :RUOGZLGHGLVWULEXWLRQRINitrosococcus oceaniDPDULQHDPPRQLDR[LGL]LQJJDPPDSURWHREDFWHULXPGHWHFWHGE\ PCR and sequencing of 16S rRNA and amoA genes. $SSOLHGDQG(QYLURQPHQWDO Microbiology, 68, 4153-4157 Wegener, A.L. (1924) The origin of continents and oceans. Translated by J.G.A. Skerl. Methuen, London. :HOVK'7 (FRORJLFDOVLJQL¿FDQFHRIFRPSDWLEOHVROXWH DFFXPXODWLRQE\PLFURRUJDQLVPVIURPVLQJOHFHOOVWRJOREDOFOLPDWHFems Microbiology Reviews, 24, 263-290 :HUQHU' The biology of diatoms. University of California Press. :KLWDNHU5-*URJDQ': 7D\ORU-: *HRJUDSKLFEDUULHUV isolate endemic populations of hyperthermophilic archaea. Science, 301, 976-978

References 267 Whittaker, R.H. (1960) Vegetation of the Siskiyou mountains, Oregon and California. (FRORJLFDO0RQRJUDSKV, 30, 279-338 :KLWWDNHU5+ (YROXWLRQDQGPHDVXUHPHQWRIVSHFLHVGLYHUVLW\ 7D[RQ, 21, 213-251 Wiens, J.J. (1998) Combining data sets with different phylogenetic histories. Systematic Biology, 47, 568-581 :LHQV-- &KDUDFWHUDQDO\VLVLQPRUSKRORJLFDOSK\ORJHQHWLFV Problems and solutions. Systematic Biology, 50, 689-699 :LHUVPD<) 8UEDQ'/ %HWDGLYHUVLW\DQGQDWXUHUHVHUYH system design in the Yukon, Canada. Conservation Biology, 19, 1262-1272 Witt, O.N. (1886) Ueber den Polierschiefer von Archangelsk-Kurojedowo im Gouv. Simbirsk. Verhandlungen der Russisch-Kaiserlichen Mineralogischen Gesellschaft zu St Petersburg, series 2, pp. 137-177. :ROIH$3 6LYHU3$ 7KUHHH[WDQWJHQHUDRIIUHVKZDWHU 7KDODVVLRVLURLGGLDWRPVIURPPLGGOH(RFHQHVHGLPHQWVLQQRUWKHUQ&DQDGD American Journal of Botany, 96, 487-497 :RRG7(7DNHED\DVKL1%DUNHU060D\URVH,*UHHQVSRRQ3% Rieseberg, L.H. (2009) The frequency of polyploid speciation in vascular plants. Proceedings of the National Academy of Sciences of the United States of America, 106, 13875-13879 Wortley, A.H. & Scotland, R.W. (2006) The effect of combining molecular and morphological data in published phylogenetic analyses. Systematic Biology, 55, 677-685 :\QQZLOOLDPV'' $HURELRORJ\DQGFRORQL]DWLRQLQ$QWDUFWLFD The Biotas program. Grana, 30, 380-393 Xiao, X.W., Xu, X. & Yang, F. (2008) Adaptive responses to progressive drought stress in two Populus cathayana populations. Silva Fennica, 42, 705-719 Yang, J., Wang, Y.F., Spicer, R.A., Mosbrugger, V., Li, C.S. & Sun, Q.G. (2007) Climatic reconstruction at the miocene shanwang basin, China, using OHDIPDUJLQDQDO\VLVFODPSFRH[LVWHQFHDSSURDFKDQGRYHUODSSLQJGLVWULEXWLRQ analysis. American Journal of Botany, 94, 599-608

268 References Summary / Samenvatting

Summary Until recently, the geographical distribution of microorganisms was assumed to be only controlled by local ecological factors (species sorting); their high local abundances and small size would increase ȱȱȱĜȱȱ¢ȱȱȱȱȱ barriers (Finlay et al.ǰȱŘŖŖŘǼǯȱȱǰȱȱȱȱ¢ȱ suggested and shown that when using a higher taxonomical ǰȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ěȱȱ ¢ȱ ȱ ȱ ǻ ȱ ȱ ȱ ǭȱ Bohannan, 2006; Martiny et al. 2006). Also for diatoms, recent studies ȱěȱȱȱȱ ȱȱȱȱȱ ȱȱȱȱ¢ȱȱȱȱȱ ȱȱǻȱǭȱǰȱŘŖŖŘǲȱȱet al., 2004; Telford et alǯǰȱŘŖŖŜǲȱ¢ȱet al., 2007; Verleyen et al.ǰȱŘŖŖşǲȱȱ et al.ǰȱŘŖŖşǲȱ¢ et al., 2010). ȱȱǰȱ ȱȱȱȱȱȱȱ ȱȱȱȱȱȱȱȱĴȱȱ in lacustrine and terrestrial diatoms by analyzing global and regional ȱ ¢ȱ Ĵȱǰȱ Ȭȱ ¢ȱ ȱ ȱ ǰȱ ȱ ¢ȱ ¢ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ Ȭȱ ǯȱ ȱ ¢ȱ ȱ ȱ ȱ ǻȬǼȱȱȱȱȱ¢ȱȬȱȱ Ěȱȱȱȱȱȱȱȱȱěȱȱȱ scales. In chapter 2ǰȱ ȱ ¢£ȱ ȱ ȱ Ĵȱȱ ȱ lacustrine diatoms in the Antarctic and Arctic and combined these ȱ ȱȱȱȱȱȬȱȱ¢¢ǯȱ ȱȱȱȱǰȱȱȱȱ ȱ ȱ ȱ ǰȱ ȱ ȱ £ȱ ¢ȱ ȱ ȱȱǰȱȱȱȱ¢ȱȱȱȱȱ ȱ ¡ǰȱ ȱ ȱ ȱ ȱ ǰȱ ȱ a general paucity of globally successful genera. A comparison of ¢ȱ ȱ Ěȱȱ ȱ ȱ ȱ ȱ

Summary / Samenvatting 271 ȱ ȱ ȱ ȱ ¢ȱ ȱ ęȱȱ Ĵȱȱ ȱ to high rates of local extinction during Neogene and Quaternary glacial maxima, in combination with radiations through allopatric speciation in glacial refugia. This indicates that processes controlling ȱȱȱęȱȱȱȱȱȱȱȱ spatial and temporal scales as those for macroscopic organisms, ȱȱȱȱĴȱǯȱ In chapter 3, we examined if terrestrial and lacustrine diatoms ȱěȱȱĴȱȱȱȱȱȱěȱȱȱ scales, using distance decay analysis of community similarities ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ £ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ǰȱȱȱȱȱȱǯȱȱȱȱ ȱ ȱ ȱ ȱ ǻΆȬ¢Ǽȱ ȱ ¢ȱ ȱ ȱȱ ȱȱȱȱȱ Ȭȱǻ΅Ǽȱ¢ǰȱ ¢ȱ Ěȱȱ Ȭȱ ȱ ěȱȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ěȱǯȱ ǰȱ ȱ ȱȱȱęȱ¢ȱ ȱȱ ȱȱȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ Ȧȱ ȱȱȱȱȱȱ ȱȱǯȱȱ ȱȱǰȱ ȱȱȱ ȱȱȱǻca. 1,500 Ǽǰȱȱ ȱȱȬęȱȱȱ ȱȱ¢ȱ ȱȱ ȱȱȱȱǰȱ ȱȱ the largest spatial, among the two oceanic basins (ca.ȱŜǰŖŖŖȱǼǰȱȱ communities showed a strong decrease in community similarity, ȱ¢ȱęȱȱȱȱȱǯȱȱęȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ěȱȱ ȱ ȱ dispersal probabilities at small to intermediate scales, but further ȱ ȱ ȱ ȱ ȱ ȱ Ěȱȱ ȱ ȱ ǯȱ In chapter 4ȱ ȱ ȱ ȱ Ȭȱ Ȭȱ phylogeny of the diatom genus Pinnularia to explore the importance ȱȱȱȱǯȱȱȱęȱȬȱȱ ȱȱ ȱ ęȱȱ ¢ȱ ȱ ȱ řŜȱ Pinnularia taxa. ȱȱȱȱ¡ǰȱȱ ¢Ȭȱȱȱ forms of Pinnularia, were used to calibrate a relaxed molecular ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȃȱ ęȱǯȱȱȬȱȱȱȱȱ Ȭȱ ¢¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ǰȱ ȱ ȱ ¢ǰȱ ȱ ¢ȱ ȱ ¢ǯȱęȱȱȱȱȱȱȱȱȱȱ

272 Summary / Samenvatting ca.ȱŜŖȱȱǰȱŗŖȱȱřŖȱȱȱȱȱȱȱǰȱ ȱȱȱȱȱȱȱǰȱ ȱȱȱ successfully disperse worldwide. In chapter 5 we used the molecular phylogeny created in chapter 4 to compare the phylogenetic signal of molecular and ȱǯȱȱȱȱȱ¢ȱȱŚŞȱȱ ȱŗŝȱȱȱřȱȱǯȱ ȱ¢ȱ analyses of molecular and morphological data generated similar ǰȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱ ¢ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ŗǰŖŗŘȱ ¢Ȭȱ ȱ ȱ ¢ȱ ǰȱ ȱ ȱ ȱ ȱ ¢¢ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱȱȱ¢ȱȱȱęȱȱȱȱPinnularia ȱ ȱ ¢ǯȱ ȱ ȱ ȱ ȱ ¢Ȭȱ ȱěȱȱ ȱȱ¢ȱ¢£ȱȱȱ characters simultaneously. Mainly “structural” characters (such as ȱ ȱ ǰȱ ȱ ¡¢ȱ ȱ ȱ Ǽȱ ȱȱȱȱǻȱȱ¢Ǽȱ ȱȱ for reconstructing the phylogenetic relationships within the genus. In chapter 6, we used two presumed cosmopolitan taxa, Pinnularia borealis and ĵȱȱ¡¢, to assess the molecular ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ locations. Molecular phylogenies based on the plastid gene rbcL ȱ ȱ ŘŞȱ ȱ ŗȬŘȱ ȱ ȱ ȱ P. borealis and ǯȱ ¡¢ are two species complexes consisting of multiple ¢ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ species complex P. borealis ȱřśǯŜȱȱ¢ȱǻǼȱǰȱȱȱ ęȱȱȱĴȱȱȱȱ¡ȱ¡ȱȱŘŘǯŖȱǰȱȱȱ ȱȱ ȱȱȱ¡ǯȱȱȱP. borealis ȱȱȱȱȱȱŝǯŝȱǻŗśȬŘǼȱȱȱȱȱȱ ǰȱȱȱęȱȱȱȱȱȱǰȱȱȱ ȱȱȱȬȱȱȱ ȱ¢ȱȱ speciation. In addition, the Antarctic lineages of both P. borealis and ǯȱ¡¢ȱȱȱȱȱ ȱȱȱ ȱǰȱ a lower optimal temperature and lethal upper temperature than most lineages from more temperate regions, indicating thermal ȱěȱǯȱȱȱȱȱ¢ȱȱȱȱ cosmopolitan Antarctic diatom species are in fact species complexes,

Summary / Samenvatting 273 possibly containing Antarctic endemic lineages. In chapter 7ǰȱ ȱȱȱ¢ȱ¡ȱȱřŚȱ ȱ ȱȱȱȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ǰȱ¢ȱȱȱ¡ȱǯȱ¡ȱěȱȱ ȱ ȱ ȱ Ǳȱ ȱ ȱ ȱ ȱ ƸřŖǚȱ ȱ ƸŚŖǚǰȱȱȱȱƸŚŖǚǰȱ£ȱȱȬŘŖǚǰȱȱȱ ȱȱ ȱȱȱƸřŖǚǯȱȱȱ ȱȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ¡ȱ ǻƸŚŖǚȱ ȱ ȬŘŖǚǼȱ ȱ ȱ ȱ ȱ ¡ȱ Ȭęȱǯȱ ǰȱ ¢ȱ ȱ ȱ ȱ £ȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱƸŚŖǚǰȱȱȱ at a higher tolerance of terrestrial diatoms to temperature extremes, and indicating that terrestrial diatoms are adapted to their habitat. In chapter 8, we experimentally assessed the tolerance for desiccation and freezing of diatom resting cells, in comparison to ȱǰȱȱŗŝȱȱ ȱȱȱ ȱȱȱȱȱȱȱȱȱǰȱ ȱ ȱ ęȱȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱȱ£ǯȱ¢ȱȱȱȱȱ¡ȱȱ £ȱȱȱȱ ȱȱȱȱȱśȱȱ of desiccation. In contrast, resting cells showed a higher tolerance for desiccation, especially when preceded by a heat treatment, ȱȱȱȱ¡ȱȱ£ǰȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ¢ǰȱ only the resting cells of terrestrial taxa, i.e. those occurring mainly in wet and moist or temporarily dry places (moisture category 4) ȱ ¢ȱ ¡¢ȱ ȱ ȱ ȱ ǻ¢ȱ śǼȱ ȱ ȱȱ£ǰȱȱȱȱȱȬęȱȱ ȱȱȱȱǯ From the results gathered in this thesis, we can conclude that diatom biogeographies are shaped by the same processes ȱȱǯȱȱĴȱȱȱȱȱ scales indicate that historical factors and dispersal limitation are at ǰȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ¢ȱȱȱȱȱȱȱȱĴȱǯȱȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ¢ȱ ȱ

274 Summary / Samenvatting high dispersal probabilities of diatoms, despite their high local ǯȱȱȱȱȱȱǰȱ ȱȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ compared to lacustrine diatom communities, which suggests higher ȱȱȱȱȱȱęȱȱȱȱȱȱ their resting cells.

Summary / Samenvatting 275 Samenvatting ȱȱȱ ȱȱȱȱęȱȱȱ ȱȬȱȱ ȱȱȱȱȱ ȱǻȱǼǲȱȱȱȱȱȱȱ ȱ £ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ¸ȱȱȱȱ£ȱ£ȱǻ¢ et al., 2002). ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ǰȱ Ȭȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻ ȱ ǭȱ ǰȱ ŘŖŖŜǲȱ ¢ȱ et al. 2006). Met ȱ ȱ ºȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ºȱ ȱ ȱ ȱ £ ȱ ȱ ȱ ȱȱǻȱǭȱǰȱŘŖŖŘǲȱȱet al., 2004;. Telford et al.ǰȱ ŘŖŖŜǲǯȱ ¢ȱ et al., 2007.; Verleyen et al., 2009;. ȱet al.ǰȱŘŖŖşǲǯȱ¢ȱet al., 2010). ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ºȱ ȱ ȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ¢ºȱ ȱ ȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ dispersiegerelateerde factoren. Er werd een primaire focus op ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ ȱ ºĚȱȱ ȱ ȱ ȱ ęȱȱ ȱȱȱȱȱȱȱǯ In hoofdstuk 2ȱ ¢ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ºȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ£ȱȱȱȱȱȱȱȬ ȱȱ¢ǯȱȱ£ȱȱȱȱ ȱ£ȱȱȱºȱ ȱȱ ǰȱȱ ȱ£ȱȱȱȱȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ £ȱ ȱ ¡ǰȱ ȱ ȱ ȱȱǰȱȱȱȱȱȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱĚȱȱȱȱȱȱȱȱȱȱ ȱ Ȭ¢ȱ ȱ ęȱȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ ȱ ȱ ȱ ȱ

276 Summary / Samenvatting glaciale maxima, in combinatie met radiaties door allopatrische ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱȱęȱȱȱºȱȱ ȱȱȱȱȱȱ ȱȱ£ȱ ȱȱǰȱȱȱȱȱ ęȱȱǯ In hoofdstuk 3 onderzochten we of terrestrische en lacustriene ºȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȃȱ ¢Ȅȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȃȄȱ ȱ ºȱ ȱ ȱ Ȭ ȱ ȱ £ȱ ȱ ȱ ȱ ȱ ȱ Ȭ ȱǰȱȱȱ ȱȱȱȱȬȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǻΆȬǼȱ ȱ ȱ ȱ ȱ ȱ ȱ Ȭȱ ǻ΅Ǽȱ ǰȱ ȱ ȱ £ȱ ȱ ěȱȱ ȱ ȱ ȱȱȱǰȱȱȱȱěȱǯȱȱ ºȱȱȱęȱȱȱȱȱȱ ȱ£ȱȱȱǰȱ ȱ ȱȱȱ ȱ Ȧȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ǯȱȱȱǰȱȱȱȬ ȱȱ (ca. ŗśŖŖȱǼǰȱ ȱȱȱȱȬęȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱȱȱȱȱǰȱȱ ȱ ȱȱǻca.ȱŜŖŖŖȱǼǰȱȱȱĴȱ¢ȱȱ ȱȱȱȱ ȱ ȱȱ ȱȱęȱȱ ȱȱȱȱǯȱ£ȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ºȱ ȱ£ȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ £ȱȱȱȱȱȱȱȱȱ controleren. In hoofdstuk 4ȱ ȱ ȱ ȱ Ȭǰȱ ȱ¢ȱȱȱȱPinnularia om het belang ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ Ȭ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱȱřŜȱPinnulariaȱ¡ǯȱȱȱȱȱ¡ǰȱȱȱ ȱȱȱȱȱȬǰȱ ȱȱȱȱ ȱȱ¢ȱȱȱȱȱȱȱȱ ȱęȱȱȱȱȱȱ£ǯȱȱȬȱȱ resulteerde in een goed opgeloste fylogenie, bestaande uit drie grote

Summary / Samenvatting 277 ȱȱȱȱȱǰȱȱȱǰȱȱ ȱȱȱǯȱęȱȱȱȱȱȱȱ Ĵȱ ca.ȱŜŖȱȱȱǰȱŗŖȱȱřŖȱȱȱȱ ȱ ȱȱȱ ǯȱȱȱȱȱȱ ȱȱǰȱ ȱ£ȱ ȱȱȱ ǯ In hoofdstuk 5ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ Śȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ŚŞȱ ȱ ȱ ȱ¢ȱȱȱŗŝȱȬȱȱřȱǯȱ Ģȱȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ śȱ ȱ ȱ ȱ ȱ ȱ ȱ ŗŖŗŘȱ Ȭȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ¢ȱ ȱȱȱȱȱȱȱ ȱȱǯȱȱȱȱȱȱȱȱ ȱ ęȱȱ ȱ ȱ Pinnulariaȱ ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ȱ ǯȱ ȱ ȃȄȱ ȱ ȱ ǻ£ȱ ȱȱǰȱȱ¡ȱȱȱǼȱ ȱ··ȱȱǻȱ¢Ǽȱ ȱȱ ȱȱȱȱȱ¢ȱȱȱȱǯ In hoofdstuk 6ȱȱ ȱ ȱȱȱ taxa, Pinnularia borealis en ĵȱȱ¡¢, om de moleculaire ȱȱȱȱȱȱȱȱ Ĥȱȱȱȱǯȱȱȱ¢ºȱȱ ȱȱȱ rbcȱȱȱȱŘŞȱȱŗȬŘȱȱ ȱ ȱ ȱ P. borealis en ǯȱ ¡¢ twee soortscomplexen £ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱȱȱȱǯȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ¡ȱ P. borealisȱ ȱ řśǰŜȱ ȱȱǻǼǰȱȱȱȱȱȱŘŘǰŖȱȱȱȱȱ ȱȱº¡ȱǯȱȱȱP. borealisȱȱȱȱĴȱȱŝǰŝȱǻŗśȬŘǼȱȱȱȱ ȱ ȱ £ǰȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ Ȭȱ ȱ ȱ ȱ ȱ ȱ ǯȱ ȱ £ȱ ȱ ȱȱȱ£ ȱP. borealis als ǯȱ¡¢ genetisch

278 Summary / Samenvatting ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ£ȱȱȱȱȱȱȱȱȱȱȱǰȱ een lagere optimale temperatuur en een lagere lethale maximum ǰȱ ȱ ȱȱȬěȱǯȱ£ȱȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ºȱ ¡ȱ ȱ £ǰȱ ȱ ȱ endemische Antarctische lineages. In hoofdstuk 7 ȱ ȱȱ¡ȱȱ ȱ řŚȱ ȱ £ ȱ ȱ ȱ ºȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ £ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ¡ȱ ǯȱ ȱ ȱ ȱ ȱ Ǳȱ ȱ ȱ ȱ ƸřŖǚȱ ȱ ƸŚŖǚǰȱ ȱ ȱ ȱ ƸŚŖǚǰȱ £ȱ ȱ ȬŘŖǚǰȱ ȱ ȱ ȱ ȱ £ȱ ȱ ȱ ƸřŖǚǯȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ºȱ £ȱ ȱ £ȱ ȱ ǰȱ £ȱ ȱ ȱ ȱ ȱ £ȱ ȱ ǯȱȱȱ¡ȱȱǻƸŚŖǚȱȱȬŘŖǚǼȱ ȱ ȱ ȱ ȱ ęȱǯȱ ǰȱ ȱ ȱ ȱ ȱ £ȱ ȱ ȱ ºȱ ȱ ȱȱȱȱĴȱȱȱƸŚŖǚǰȱȱ ȱ ȱȱȱȱȱȱºȱȱ¡ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ºȱ aangepast zijn aan hun habitat. In hoofdstuk 8 onderzochten we experimenteel de tolerantie ȱ ȱ ȱ ºȱ ȱ ȱ ȱ £ȱ ȱ ȱȱȱȱǰȱȱȱȱŗŝȱȱ ȱȱȱºȱȱȱȱȱȱ ȱǰȱȱȱȱºǯȱ ȱ ȱȱȱºȱȱȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱȱȱȱ£ȱȱȱȱȱ£ǯȱ ȱȱȱȱȱ¡ȱȱ£ȱ ȱȱȱǰȱ ȱȱ··ȱȱśȱȱȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǲȱ ȱ ȱ ȱ ȱ ¡ȱ ȱ £ǰȱ£ȱȱȱȱȱȱȱȱȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ¡ǰȱǯ ǯ£ȱ¡ȱȱ£ȱȱĴȱȱȱȱ ȱȱȱȱǻȱŚǼȱȱȱȱ

Summary / Samenvatting 279 ȱ ȂȱǻȱśǼȱǰȱȱ ȱ £ȱ ǰȱ ȱ ȱ ȱ ęȱȱ aanpassingen aan deze ongunstige omstandigheden. ȱ ȱ ȱ £ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ęȱȱ ȱ ºȱ ȱ ȱȱ£ȱȱȱȱǯȱȱęȱȱ patronen op wereldschaal tonen aan dat historische factoren en ȱȱȱȱ£ǰȱ ȱ¢ȱ soorten en een te lage taxonomische resolutie nog steeds een aantal ȱȱęȱȱȱǯȱȱȱȱȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ºȱ ȱ ǰȱ ȱ ȱȱȱǯȱȱȱȱȱǰȱ ȱ ȱȱȱȱȱȱȱȱ ȱ ȱ £ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ £ȱ ȱ £ȱ ȱ ȱ ȱ ȱȱȱȱȱȱǯ

280 Summary / Samenvatting Acknowledgments

There’s one important thing I’ve learned during these four years of ęȱȱ¢ǰȱǰȱȱȱǱȱnever make a PhD on your ownǯȱ ȱȱȱȱȱȱęȱȱ ȱ¢ǰȱȱ¢ȱǯȱ ǰȱȱȱȱȱ¢ȱȱȱȱ¢ȱ path.

ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱ¢ȱȱȱ ȱ ȱ¢ȱȱȱȱȱ ȱǯȱȱȱȱǰȱȱȱ¢ȱȱ¢ǰȱȱ ȱȱȱ ǰȱȱǰȱȱȱȱǰȱ Ȃȱ very grateful.

ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ¢ȱȱȱȱ ȱȱȱȱȱȱȱ ǯȱ ȱ ȱ ęȱȱ ȱ ȱ ȱ Ƿȱ ȱ ¢ȱ ȱ ¢ȱ positivism.

ȱȱȱȱȱȱȱȱȱȱȱȱ ȱȱȱȱȱȱȱǯȱȱǰȱ ¢ȱ ȱȱȱ ȱȱȱȱȱȱ ȱ ȱ ȱĴȱȱȱȱ ȱ¢ȱȱȱǰȱȱȱ ȱ¢ȱ ȱȱȱȱȱ¢ȱȱ ȱȱ ȱ ȱ ȱ ȱȱȱȱȱȱȱǯȱ ȱȱȱȱȱǰȱ ȱ ȱ¢ȱȱȱȱǯ

ȱ¢ǰȱȱ¢ȱȱȱȱȱȱȱ ȱȱ ȱȱȱȱȱȱ¢ȱȱȱȱ ȱȱȱǯ

ȱȱȱǰȱȱ¢ȱȱ¢ȱȱǰȱ ¢ȱ¢ȱȱ ȱȱȱǰȱȱ¢ȱ¢ȱȱȱ ȱ¡ȱȱȬȱǱȱȱ¢ȱ¢ȱǯȱ ȱ¢ǰȱ ȱ ȱ ȱȱȱ¡¢ǯ

ȱ ¡ǰȱ ȱ ¢ȱ ȱ ¢ȱ ȱ ȱ ȱ ȱ ȱ

Acknowledgments 283 ¢ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱ ǯȱ ȱ ȱ ȱ Ĵȱȱ ¡ȱ ȱ ¢ȱ ěȱȱ Ǳȱȱȱȱǰȱȱ¢Ȭȱ·ȱ ȱȱǰȱȱȱȱȱȱȱ ȱ ȱȱȱ ǰȱȱȱȱȱȱ¢ȱȱȱȱ ¢Ƿ

ȱ ǰȱ ¢ȱ ¢ȱ ȱ ȱ ǯȱ Ȃȱ¢ȱȱȱ¢ȱȱ¢ȱ¢¢ȱ¢ȱȱ ¢ȱ ȱǯȱ ¡ȱǰȱ¢ȱ¢ȱ¢ȱȱȱȱȱȱȱȱ ǰȱȱȱȱȱȱȱǯȱ

ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ¢ȱ ȱ ȱ ȱ ǰȱ ȱ ȱ ȱȱęȱȱȂǰȱȬȱǰȱȱȱȱ ȱ ǯȱ ȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ȱȱȱȱ ǯ

¢ȱ ȱ Ǳȱ ȱ ǰȱ ȱ ȱ ǰȱȱ ȱ ȱ ȱ ¢ǯȱ ȱȱ ȱ ȱȱȱȱȱȱęȱȱȱǰȱȱ¢ȱȱȱȱȱȱȱ £ǰȱ ǰȱ £ǰȱ ǰȱ ǰȱ ȱ ȱȱȱ¢ȱ ǯȱ ȱ ȱȱ¢ȱĴȱȱȱȱȱǳ

ȱȱȱȱȱȱȱȱȱȱ¢ȱ¢ǰȱǰȱ ¢ǰȱ¢ǰȱ¢ȱȱȱǯȱȱȱ¢ȱ¢ȱ ǱȱȱȱȱȱȱȱǷ

ȱȱȱȱȱȱȱȱȱȱȱȱ¢ȱ ȱǯȱȱȱȱȱ ȱȱ¢ȱȱȱ¢ȱ ȱȱȱȱȱȱȱǰȱȱ¢ȱ ȱ¢ȱǯȱȱȱȱ ȱ ȱȱȱȱȱ¢Ǳȱ ȱȱ ȱ ȱȱȱȱȱǯȱ

ȱ ȱ ȱ ȱ ¢ȱ ȱǰȱ ȱ ȱ ¢ȱ ȱ ȱ ȱȱȱǯȱȱȱǰȱȱȱ ȱȱȱȱȱǻȱǼȱȱęȱȱȱ ǰȱȱȱ ¢ȱȱȱȱȱȱ¢ȱȱǰȱȱ ¢ȱ¢ȱǷȱǰȱȂȱȱȱȱȱȱȱ¢ȱǷȱ

284 Acknowledgments ¢ȱȃ¢ȬȬ ȄȱȱǻȱĜȱȱǼȱȬȬ ȱ[£Ǳȱ ȱ¢ȱȱȱ¢ȱ ȱ ȱ¢ȱȱǯȱ

ǰȱȱȱ ȱȱȱȱ ȱ¢ȱȱ ȱęȱȱǯȱȱȱȱȱȱǰȱȱȱȱ ȱȱǰȱȱȱȱȱȱ¢ȱǻȱ Ǽǰȱ ȱȱ¢ȱǯȱȱ ȱȱȱȱȱȱ Ěȱ ǰȱȱȱȱȱ££¢ȱȱǰȱ¢ȱȱ¢ȱ ȱȱȱȱȱȱǰȱ ȱȱȱǯȱȱȱ ȱȱȱ ȱȱȱȱȱȱ¢ȱȱǷ

¢ȱ ȱ ȱ ǭȱ Ǳȱ ȱ ȱ ȱ ȱ ȱ ȱ ȱȱǰȱȱȱ Ĵȱȱ¢ȱȱ ȱȱȱ ȱ¢ǰȱȱ¢ȱȱ Ƿ ȱ ¢ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ¢ȱ ȱ ǯȱ ȱ ȱ ȱ ȱ £ȱ ȱ ǰȱ ȱ ȱ ȱ ȱ ȱ ǰȱ ǰȱ ¢ȱ ȱ ȱ ǰȱȱǰȱȬȁȬȱȱȱȂǯȱȱ¢ȱ ȱȱȱ¢ȱ£¢ȱȱȱǻȱǰȱȱ ȂȱǷǼȱ ȱ Ȭȱ ȱ ȱ ǯȱ ȱ ¢ǰȱ ȱ ȱ ȱ ȱ ȱȱǰȱ ȱȱȱ ǯ

ǰȱǰȱ Ȃȱ¢ȱȱȱ¢ȱǯȱȱȱ ȱȱȱȱȱȱȱȱ ȱȱȱȱȱ ȱ ǯǯǯ

ȱěȱ Gent, 15 May 2011

ȱȱ ȱęȱ¢ȱȱ¢ȱȱȱȱęȱȱȱ ȮȱȱǻȬǼǯ

Acknowledgments 285