Biology, Ecology and Invasion Characteristics Of

Charles University in Prague, Faculty of Science Department of Botany

Biology, ecology and invasion characteristics of Campylopus introﬂexus in the Czech Republic Biologie, ekologie a charakteristiky invaze druhu Campylopus introﬂexus v Ceskéˇ Republice

Mgr. Eva Mikulášková

Study programme: Biology; Botany Ph.D. thesis Brno 2012

Supervisor: RNDr. ZdenˇekSoldán, CSc.

DECLARATION

I hereby declare, that I made this thesis independently, using the listed references, or in the cooperation with other authors of the papers (for my contribution to particular papers see chapter 6 Author´s contribution). I did submit neither the thesis nor its any part to acquire any other academic title.

Brno, 2012

Mgr. Eva Mikulášková

ABSTRACT

Ecological and economic impact of invasive plants to natural ecosystems is the subject of many studies; however, invasive bryophytes have been studied only marginally. Campylopus introflexus (Hedw.) Brid. is one of the most strongly invasive bryophyte species in Europe. The species appears to be native in the Southern Hemisphere. In Europe, it was collected for the first time in the British Isles in 1941. The moss has expanded eastward and the first collection in the Czech Republic is dated to 1988. This thesis found that more than 70 localities were known known in the Czech Republic in 2006, and more than 100 localities became known by 2011. It has been further demonstrated that the Czech Republic was colonized repeatedly by generative spores and all populations have a unique genetic composition. Genetic variation of the populations is low, the genetic diversity of populations within the Czech Republic is not correlated with their geographic position or with any of the monitored environmental variables. At a fine scale within particular localities, the species disperses by vegetative diaspores, while it uses generative spores for spreading over longer distances. In Central Europe, C. introflexus prefers open coniferous forests, especially plantations of either spruce or pine. It colonizes clearings, pathsides, forests edges and disturbed peatbogs. It may form large and compact stands at these habitats. It is able to colonize microhabitats with very heterogeneous ecological conditions. It inhabits acidic, nutrient-poor, bare soil without strong competition with vascular plants, lichens and other bryophytes. Cultivation experiments as well as vegetation-data analyses showed that the species thrives best on organic and sandy soils and avoids limestone and strongly waterlogged soils. C. introflexus represents no significant risk to natural plant communities in Central Europe presently, because it colonizes mainly human-disturbed habitats. It can form stable mixed stands with the domestic pioneer species. It occupies free patches more rapidly than some native species, but it not able to outcompete them directly. The ratio of coverage of C. introflexus and native pioneer species is partially dependent on the amount of both generative and vegetative diaspores during initial colonization. In some cases C. introflexus can block the natural succession.

v ABSTRAKT(INCZECH)

Ekologický a ekonomický dopad invazních rostlin na pˇrirozené ekosystémy je pˇredmˇetemmnoha studií, avšak invazní mechorosty jsou studovány pouze okrajovˇe. Campylopus introflexus (Hedw.) Brid. je nejvýznamnˇejšíinvazní druh mechorostu v Evropˇe.Pochází z jižní polokoule a v Evropˇebyl poprvé zaz- namenán v roce 1941 na Britských ostrovech, odkud se postupnˇešíˇrísmˇerem na východ. První údaj z Ceskéˇ republiky pochází z roku 1988. V rámci této práce bylo zjištˇeno,že v roce 2006 bylo v Ceskéˇ republice známo pˇres 70 lokalit, v roce 2011 už bylo známo více než 100 lokalit. Dále je ukazováno, že Ceskáˇ republika byla kolonizována opakovanˇepomocí gener- ativních spor, všechny populace mají unikátní genotyp. Populace mají malou genetickou variabilitu, genetická diverzita v rámci Ceskéˇ republiky není ko- relována s geografickou pozicí ani s žádnou ze sledovaných promˇenných prostˇredí. V rámci jemného mˇeˇrítkajedné dílˇcílokality se rozšiˇrujepomocí vegetativních diaspor, zatímco pro šíˇrení na vˇetšívzdálenosti využívá gener- ativní spory. Ve stˇrední Evropˇe C. introflexus preferuje jehliˇcnatéprosvˇetlenélesy, zvláštˇe monokultury smrku ˇciborovice. Druh kolonizuje paseky, bˇrehy cest, okraje porost ˚ua narušená rašeliništˇe.Na tˇechtomístech m ˚užetvoˇrit rozsáhlé a kompaktní porosty. Je schopný osídlit mikrostanovištˇes velice r ˚uznorodými ekologickými podmínkami. Osidluje holou, kyselou, na živiny chudou p ˚udu bez veliké kompetice cévnatých rostlin, lišejník ˚ua ostatních mechorost ˚u. Kultivaˇcníexperimenty stejnˇejako analýzy vegetaˇcníchdat ukázaly, že druh nejlépe prosperuje na organických a písˇcitých p ˚udácha vyhýbá se p ˚udám vápenatým a silnˇepodmáˇceným. C. introflexus v souˇcasnédobˇenepˇredstavuje významné riziko pro pˇrirozená spoleˇcenstva stˇrední Evropy, protože kolonizuje pˇrevážnˇeantropogennˇenarušená stanovištˇe.S domácími pionýrskými druhy m ˚užetvoˇritstabilní smˇesnéporosty. Volný neosídlený prostor osidluje mnohem rychleji než ostatní domácí druhy, nicménˇeje není schopen pˇrímovytlaˇcovat. Pomˇerpokryvnosti druhu C. introflexus a domácích pionýrských druh ˚uje ˇcásteˇcnˇe závislý na množství pohlavních i nepohlavních diaspor pˇripoˇcáteˇcníkolonizaci. V nˇekterých pˇrí- padech m ˚uže C. introflexus blokovat pˇrirozenou sukcesi na stanovištích.

vi Nature preserves the species, and cares but very little for individuals. — Voltaire

ACKNOWLEDGMENTS

First of all, I would like to thank Michal Hájek, Jon and Blanka Shaw for motivation, support, and many indispensable advices. I thank to my Thesis adviser, ZdenˇekSoldán for help starting my work and for help in the ﬁeld. I am grateful to Michal Hájek, Zuzana Fajmonová and Tomáš Fér for their comments on the manuscript of this thesis. My thanks are due to Matthew Johnson and Jon Shaw for the revision of English style. For some consultations and valuable information for on work in the DNA lab, I thank to Veronika Kuˇcabová, and I also thank to JiˇríVáˇnafor consultations about bryology. I thank Ondˇrej Hájek for the creation of maps. For the precise assistance and kind atmosphere, I thank the staff of cryp- togamic section and the staff of the DNA lab of Department of Botany, Fac- ulty of Science, Charles University and staff of the Mire ecology group of Department of Botany and Zoology, Faculty of Science, Masaryk University. I would like to thank my colleagues and friends from the Czech and Slo- vak lichenological community for nice experiences and kind atmosphere throughout the years. I deeply thank JiˇríMikulášek, Jindˇriškaand my family for their patience and all support. I thank to all babysitters who took care about Jindˇriškaduring writing this thesis. My work was supported by Grant Agency of Charles University (project no. 258/2004/B-BIO/PrF) and institutional support of Masaryk University.

vii

CONTENTS i introduction1 1 general introduction3 1.1 Bryophytes ...... 5 1.2 Heath star moss Campylopus introflexus (Hedw.) Brid...... 6 1.2.1 Taxonomy ...... 6 1.2.2 Anatomy, morphology and reproduction ...... 7 1.2.3 Ecology ...... 9 1.2.4 Phytosociology ...... 11 1.2.5 Biogeography – native and invasive distribution . . . . . 11 ii main thesis 15 2 the object and the aims of the study 17 3 outline of the thesis 19 3.1 The main ideas and results of papers ...... 19 3.1.1 Habitats colonized by C. introflexus in the Central Eu- rope (Paper 1)...... 19 3.1.2 C. introflexus dispersal and genetic diversity (Paper 2). 20 3.1.3 Molecular methods optimalization (Paper 3)...... 21 3.1.4 The role of interspecific competition in C. introflexus spreading (Paper 4)...... 21 3.1.5 Distribution of C. introflexus in the Czech Republic (Pa- per 5)...... 22 3.2 The results of papers in the broader context of species biology . 22 3.2.1 Dispersal mode ...... 22 3.2.2 Invaded habitats ...... 23 3.2.3 Distribution ...... 25 4 conclusion 27 iii papers 37 5 list of papers 39 6 author´s contribution 41 7 habitats colonized by c. introflexus in the central europe 43 7.1 Introduction ...... 44 7.2 Methods ...... 46 7.2.1 Data sampling, vegetation data ...... 46 7.2.2 Data analyses ...... 48 7.3 Results ...... 50 7.3.1 Distribution of Campylopus introflexus in the Czech Re- public ...... 50 7.3.2 Vegetation composition of C. introflexus habitats . . . . 51 7.3.3 Ecological characteristic of the habitats – field data . . . 53 7.3.4 Ecological characteristic of the habitats – cultivation experiment ...... 57 7.3.5 Invasibility of vegetation by C. introflexus ...... 58

ix x contents

7.4 Discussion ...... 59 7.4.1 Distribution pattern in the Czech Republic ...... 59 7.4.2 Recently invaded habitats and their ecological characteristics ...... 61 7.4.3 Differences in habitat affiliation between the Czech Re- public and Western Europe ...... 61 7.4.4 Future invasion potential ...... 63 7.4.5 Conclusions ...... 63 8 c. introflexus dispersal and genetic diversity (paper 2) 71 8.1 Introduction ...... 72 8.2 Methods ...... 74 8.2.1 Samples ...... 74 8.2.2 DNA extraction and AFLP fingerprinting ...... 77 8.2.3 Isozyme analysis ...... 77 8.2.4 Data analysis ...... 79 8.3 Results ...... 80 8.3.1 Clonal structure ...... 82 8.3.2 Genetic diversity ...... 83 8.3.3 Genetic structure ...... 83 8.3.4 Genetic relationships at the population scale ...... 85 8.4 Discussion ...... 85 8.4.1 Distribution mechanisms of C. introflexus ...... 89 8.4.2 Genetic variation among and within populations . . . . 90 8.4.3 Spatial genetic structure in the Czech Republic . . . . . 91 8.4.4 Conclusions ...... 92 9 molecular methods optimalization (paper 3) 101 9.1 Introduction ...... 101 9.2 Methods ...... 103 9.2.1 Sample collection ...... 103 9.2.2 DNA extraction ...... 103 9.2.3 AFLP fingerprinting ...... 104 9.2.4 Assessment of AFLP error rate ...... 106 9.3 Results ...... 106 9.3.1 DNA extraction ...... 106 9.3.2 AFLP fingerprinting ...... 107 9.4 Discussion ...... 108 9.4.1 Optimization of DNA extraction ...... 108 9.4.2 AFLP fingerprinting ...... 113 9.4.3 Reproducibility tests ...... 114 10 the role of interspecific competition in c. introflexus spreading 119 10.1 Introduction ...... 120 10.2 Methods ...... 122 10.2.1 Field design ...... 122 10.2.2 Data analysis ...... 123 10.3 Results ...... 123 10.3.1 Colonization of unvegetated patches ...... 123 10.3.2 Colonization rate ...... 125 contents xi

10.3.3 Outcompetition of native species ...... 127 10.4 Discussion ...... 128 10.4.1 Dynamics of establishment of new populations . . . . . 129 10.4.2 Difference between C. introflexus and native species in increase of cover ...... 129 10.4.3 Interspecific interactions with native species ...... 130 11 distribution of campylopus introflexus in the czech republic (paper 5) 137 11.1 Introduction ...... 137 11.2 Distribution of C. introflexus in the Czech Republic ...... 138 11.3 Discussion and conclusion ...... 146 a optical disc 151 LISTOFFIGURES

Figure 1.1 Genus Campylopus – tree ...... 8 Figure 1.2 C. introflexus – photo ...... 9 Figure 1.3 World distribution ...... 12 Figure 1.4 European distribution ...... 13 Figure 3.1 Distribution in Czech Republic ...... 26 Figure 7.1 Distribution of potential habitats...... 52 Figure 7.2 DCA ordination diagram...... 56 Figure 7.3 Ellenberg indicator values for pine forests...... 57 Figure 7.4 Cultivation experiment...... 59 Figure 8.1 Map of investigated populations...... 74 Figure 8.2 Neighbor-joining tree – dataset A ...... 78 Figure 8.3 STRUCTURE analysis...... 86 Figure 8.4 Principal coordinate analysis...... 87 Figure 8.5 Spatial autocorrelation analysis...... 88 Figure 9.1 Duplicated samples – banding pattern...... 109 Figure 9.2 Four storaged DNA types – banding pattern...... 110 Figure 10.1 Area covered by C. introflexus in different habitats . . 125 Figure 10.2 Area covered by C. introflexus and native species . . . 126 Figure 11.1 Distribution in the Czech Republic...... 138 Figure 11.2 Altitudinal distribution...... 147 Figure 11.3 Increase of records over the years...... 148

LISTOFTABLES

Table 1.1 Alien bryophytes in Europe ...... 7 Table 7.1 Linear mixed effects models – AIC ...... 51 Table 7.2 Classiﬁcation of dataset A...... 54 Table 7.3 Characteristics of pine forests – dataset B...... 55 Table 7.4 Linear mixed effects models – results...... 58 Table 7.5 Vegetation types with similar species composition. . . . 60 Table 8.1 List of populations...... 75 Table 8.2 Genetic diversity pattern – dataset C...... 81 Table 8.3 Genetic diversity pattern – datasets A, B...... 82 Table 8.4 Analysis of molecular variance – datasets A, B..... 84 Table 8.5 Spatial autocorrelation analysis...... 88 Table 9.1 List of collectionsites...... 104 Table 9.2 Comparison among extraction approaches...... 107 Table 9.3 Modiﬁcations to the AFLP protocol...... 111 Table 9.4 Summary of count of error rate...... 111 Table 9.5 Storaged DNA – percentage of polymorphic markers. . 112 Table 10.1 Studied populations ...... 124

xii Part I

INTRODUCTION

GENERALINTRODUCTION 1

Invasive species and their environmental affects have been studied by biol- ogist around the world from the second half of 20th century (Elton, 1958). Plant invasions are considered to be one of the major threats for ecosystem diversity (Simberloff and Rejmánek, 2011). Numerous studies summarized the impact of invasive species on native species and community structure and have identified a set of invasion-threatened habitats (e. g. Williamson, 1996; Chytrý et al., 2009a; Hejda et al., 2009). They point to a fact, that invasive species can dramatically change current species composition or influence natural succession of plant communities. In Europe, there are more than five thousand alien species including many invasive species with huge ecological and economical impacts (Daisie, 2009). A plant species may be invasive either because it shares habitats with resident native species and outcompetes them or because it possesses habitats different from those of native species and thus occupies free niches (Sakai et al., 2001). Invasive species spreading, after the first establishment, depends on their ability to colonize new habitats and/or their competition with native species. These factors are therefore important in evaluating the actual threat for native flora. The broad-scale analyses of functional plant traits, habitat characteristics and vegetation composition can be used as a good predictor for future spreading of an invader (Pyšek and Richardson, 2007; Chytrý et al., 2008; Van Kleunen et al., 2010). Habitat filters play an important role in plant invasions (Pyšek et al., 2005; Carranza et al., 2011; Pinke et al., 2011). There are several factors making some habitats prone to an invasion of a non-native species. First, disturbance was often linked with invasiveness of communities (Burke and Grime, 1996; Elton, 1958; Williamson, 1996). In addition, habitats subjected to strong human influence; e. g. riverine corridors and lowland sandy areas, are often invaded. Habitats in low altitudes are more invaded than in mountains. Coastal and riverine habitats are frequently invaded as well because of the absence of dispersal barriers. This pattern is very similar in all European regions in spite of different invasive plants species present (Pyšek and Prach, 1993; Pyšek et al., 1998; Chytrý et al., 2005, 2009b). Conditions under which a species invades may differ from those under which it usually occurs in its native area. Successful invaders occupy a wider range of habitats in the invaded than in native range, what is probably related to changes in biological features of invasive species (Hejda et al., 2009). In an invaded area, the range of occupied habitats increases with time after initial colonization, depending on invader´s ability to establish in new communities (Sakai et al., 2001). A lag time is often observed between the first colonization of habitat and rapid spreading (Pyšek and Prach, 1993). Due to the lag time, a species can be present in its new area for many years before an invasive behavior is detected (Crooks and Soulé, 1999; Lockwood et al., 2007).

3 4 generalintroduction

Species-rich communities has been believed to be more resistant to invasions, because most or all of the available resources are utilized. Levine and D’Antonio (1999) found that this prediction is often supported by field and experimental data, but other studies report the opposite, probably due to ecological factors associated with high diversity (Levine and D’Antonio, 1999; Levine, 2000; Lonsdale, 1999). Research on invasive plants has been focused mainly on their actual or predicted distributions (e. g. Chytrý et al., 2009a) and their ecology with impact to native communities (e. g. Starfinger et al., 1998; Daisie, 2009). Less work has been reported on mating systems (e. g. Herben, 1994), genetic structure (e. g. Wang et al., 2008; Dlugosch and Parker, 2008), or physiology (e. g. Spar- rius and Kooijman, 2011). Both genetic structure and dispersal mechanisms could reflect environmental conditions and therefore may provide predictions about future directions and speed of expansion (Gunnarsson et al., 2007). The detection of specific genotypes within invasive species should help for management decisions (Hufbauer, 2004; Howard et al., 2008). How- ever, connection between molecular loci and phenotypic traits is not direct for most organisms; an additional research is necessary. Key use of molecular data for invasive species is identifying species/genotypes, determining their origins, and understanding mechanisms and rates of spread (Hufbauer, 2004). The success of invasive species can depend on their ability to adapt to local environmental conditions (Sakai et al., 2001; Prentis et al., 2008). Evolu- tion of these adaptations will be promoted by standing genetic variation for ecologically relevant traits (Colautti et al., 2010), as well as new mutations. Genetic structure in invasive populations may be weak compared to their native ranges because founder effects lead to decreased genetic diversity (Ward, 2006). On the other hand, many authors show high levels of genetic variation in introduced populations as a consequence of multiple introductions (e. g. Sakai et al., 2001; Pairon et al., 2010; Dlugosch and Parker, 2008). However, genetically distinct and spatially separated populations at the regional scale may be the product of different founding events, or they may result from one initial founding event and the subsequent evolution of populations or subpopulations in partial isolation (Torimaru et al., 2003). High percentages of unique genetic patterns have been recorded for several invasive species (e. g. Walker et al., 2003; Chapman et al., 2004; Kolbe et al., 2004; Durka et al., 2005; Wang et al., 2008). The investigation of genetic diversity of invasive species within small areas could represent a random sample of genetic variation in the invaded area overall, derived by independent colonization from disparate source populations (Ramaiya et al., 2010; Sundberg, 2005). Invasive plants are typically dispersed by wind and have high propagule production (Moravcová et al., 2010). Also human-dispersed species have high potential to become invasive. In species with vegetative as well as sexual reproduction, fine scale genetic structuring provides key information about patterns of colonization and spread, such as the relative importance of clonal growth versus sexual reproduction for population expansion (Has- sel and Söderström, 2005; Ward, 2006). Most invasive plants are primarily selfing or asexual (e. g. Husband and Barrett, 1991; Amsellem et al., 2001; 1.1 bryophytes 5

Wang et al., 2008; Zhang et al., 2010). Some plants even switch to higher levels of inbreeding or vegetative reproduction during their invasions (Pellegrin and Hauber, 1999; Amsellem et al., 2001).

1.1 bryophytes

Mosses (Bryophyta) are non-vascular plants and compose the second most diverse phylum of terrestrial plants with approximately 13 000 species (Goffinet et al., 2008). Despite their small size and inconspicuous nature, bryophytes are very successful and important parts of many ecosystems. As descendants of the early land plants, bryophytes have retained the ability to adapt to a variety of habitats and environments. There are various factors which determine species ecological behaviour; e. g. climate; substrate chemistry, physical factors such as surface texture, shading, the coverage of the surrounding vascular plants, pollution level. Bryophytes show a wide range of interactions with other organisms (Glime, 2012). Part of their success is due to the fact that they can avoid direct competition with higher plants, either because they live in areas where higher plants cannot grow at all, or because they are small enough to inhabit microhabitats that are not suitable for other plants. Interspecific competition being a factor shaping species co-occurrence patterns has been documented within bryophytes as well (e. g. Rydin, 1997; Kotowski et al., 2006; Löbel et al., 2006; Mälson and Rydin, 2009). Bryophytes show a wide variety of distribution patterns. Wide areas of distribution are consequences of effective and successful dispersal (Frahm, 2012). A high degree of diversity in vegetative diaspores of bryophytes is unparalleled among land plants (Duckett and Ligrone, 1992; Frahm, 2009). Abundance of small spores that survive long distance dispersal (e. g. Van Zanten and Pócs, 1981; Muñoz et al., 2004) can lead to elevated dispersal and establishment rates (Herben et al., 1991; Herben, 1994). Species with small spores may be better colonizers with higher probability of becoming invasive, if established outside its native range (Söderström, 1992). These life- history characteristics may contribute to high invasive potential for some bryophytes. Although a large amount of studies have been conducted to explain and predict plant invasions, the majority of them are focused on vascular plants. Studies on invasions of non-flowering plants, such as bryophytes, are rather rare (but see e. g. Hedenäs et al., 1989a,b; Hasse, 2007; Klinck, 2008). Bryophytes have rarely been intentionally introduced to areas outside their natural range when compared to vascular plants (Pyšek et al., 2011) because of their limited uses for humans. Many introduced bryophytes require protected conditions in greenhouses or indoor aquaria (Glime, 2007). The first neophyte bryophytes were recognized in the 19th century (Frahm, 1974). The number of recognized alien bryophytes exponentially increased in the 20th century due to a greater attention to these species by bryologists in this period. As compared to vascular plant alien species (ca 45 % of European flora is classified as alien either in Europe or at least in some European countries), 6 generalintroduction

alien species form a minor part (ca 2 %) of European bryoflora (Daisie, 2009). This surprisingly low number of neophyte species could be in fact much higher due to difficulties in determination if a species is native, immigrated or introduced in given area. There are 45 bryophyte species that are considered to be alien at least in some parts of Europe, including 13 cryptogenic species with unclear histories of European occurrence (Essl and Lambdon, 2009; Essl et al., 2011). Only five alien and three cryptogenic species have invaded more than five countries (Table 1.1). Eight species are native in some part of Europe (e. g. Riccia rhenana Lorb. and Lunularia cruciata (L.) Lindb.) and they are spreading to other country/countries as alien species. Other species originated outside of Europe; the native range of these species frequently lies on southern hemisphere. Only two neophytes, Campylopus introflexus (Hedw.) Brid. and Orthodontium lineare Schwägr., are widespread over a large part of Europe. The majority of the alien bryophytes, similar to vascular plants (Chytrý et al., 2009a), prefer human-disturbed habitats (Essl and Lambdon, 2009) and they have colonized regions with oceanic climates. For alien bryophytes, much less is known about ecological and economical impacts of invasions than for vascular plants. Competition with native species was observed in some vegetation communities (e. g. Hedenäs et al., 1989b; Biermann and Daniëls, 1997; Hasse and Daniëls, 2006; Klinck, 2008). C. introflexus forms dense compact cushions which negatively affect the germination of spores and seeds of other species in drained bogs (Equihua and Usher, 1993). The changes of vegetation composition by alien species negatively affect other groups of organisms; e. g. carabid beatles and spiders (Schirmel et al., 2011).

1.2 heath star moss campylopus introflexus (hedw.) brid.

1.2.1 Taxonomy

The genus Campylopus belongs to the family Dicranaceae. It was first described by J. Hedwig in “Species Muscorum Frondosorum”, as Dicranum introflexus (Richards, 1963) in 1801. In 1819, the species was recombined by S. E. Bridel-Brideri to Campylopus introflexus. The European Campylopus pilifer Brid. (= C. polytrichoides De Not.) was for a long time considered as a synonym of Campylopus introflexus. These two species were considered conspecific until the study of Giacomini (1955), who verified that a number of morphological characters and distribution distin- guish them (e. g. Frahm, 1974; Richards, 1963; Gradstein and Sipman, 1978). The geographic distributions of C. introflexus and C. pilifer differ, but overlap in southern Africa, western USA, on the British Isles and western south- western part of Europe. Differentiation between the two species was confirmed by Stech and Dohrmann (2004) and Stech and Wagner (2005) based on their analysis of rDNA ITS1, ITS2, and the atpB-rbcL spacer genetic markers (Figure 1.1). Authors described C. introflexus as monophyletic, little diversified lineage. Intermediate populations with morphological characteristics of C. introflexus (characterized by low lamellae) and C. pilifer (with straight hair point) 1.2 heath star moss campylopus introflexus (hedw.) brid. 7

Table 1.1: Alien bryophytes in Europe, ranked by decreasing number of invaded countries / regions. Only species invading > 3 countries / regions are shown. (Essl et al., 2011, p. 31)

were found, and Frahm and Stech (2006) classified such populations as Campylopus pilifer var. brevirameus (Dix.) J.-P. Frahm & Stech. This taxon was found on several places in western part of Europe, South Africa, Seychelles, Réunion, and Argentina. In some localities, it grows together with C. introflexus. It cannot be confirmed nor excluded that Campylopus pilifer var. brevirameus is of hybridogenous origin (Frahm and Stech, 2006).

1.2.2 Anatomy, morphology and reproduction

Campylopus introflexus grows in mixed stands with other bryophytes or in monodominant stands, that can cover up to 100 % of microhabitats. It forms dense, yellowish to dark olive green or almost black cushions that look hoary when dry. The cushions show a layering, with fresh green parts on top, and brown decaying material underneath. Plants are 0.5–5 cm high, tomentum is present or almost absent. Stems form thickened nodes on stem with slender sections among them. Leaves are 3–6 mm long, lanceolate, erect and straight when moist, and more appressed with the hair point often reflexed to 90 ° when dry (Figure 1.2). They are relatively wide at the base or just above, with entire margins. The nerve width varies from 30 to 70 % of the leaf width; transverse section of the nerve in the central part of the leaf shows adaxial hyalocysts and abaxial stereids; the nerve is shortly lamellose at back with ribs 1–2 cells high (Frahm, 2007). The laminal cells are thin-walled, incras- sate, shortly rectangular to oblique, the alar cells are lacking or formed by thin-walled, hyaline to reddish, inflated cells. Asexual reproduction occurs 8 generalintroduction

Figure 1.1: Tree obtained by a maximum likelihood analysis of combined ITS1, ITS2, and atpB-rbcL spacer sequences of genus Campylopus, with Pilopogon as outgroup representative. Bootstrap values > 50 % are depicted above the branches. (Stech and Dohrmann, 2004, pp. 428) 1.2 heath star moss campylopus introflexus (hedw.) brid. 9

Figure 1.2: Cushions of Campylopus introﬂexus with reﬂexed hairpoints and with (a) sporophytes or (b) vegetative diaspores. Photo E. Mikulášková.

(a) (b) frequently by the fragile leafy stem tips which easily break off (Van der Meulen et al., 1987). Fertile plants are frequent. The seta is 7–12 mm long, yellowish brown to brownish if old, curved or sinuose; one to several sporophytes often grow from the stem. It is a dioicous species; capsules are brown, 1.5 mm long, slightly asymmetric and curved when empty. The calyptra is ciliate at base. Spores are small, 12–14 µm (Frahm, 2007). In Europe, C. introflexus is an easily recognizable species. It could be con- fused with the closely allied Campylopus pilifer Brid., but both species can be recognised by character of stems. C. pilifer has less frequently reflexed hair- point to a right angle and has lamellae on abaxial side of costa of 3–4 cells high. The problems with determination may rarely arise with plants from shaded habitats, in which the hairpoints are lacking or so short that they are not reflexed (Frahm, 2007). Campylopus introflexus reproduces easily, generatively from small spores and vegetatively from fragile stem tips. Both spores and stem tips are blown away by the wind and establish if the habitat is suitable (Van der Meulen et al., 1987; Söderström, 1992). The dry moss carpet is often seen to fragment and break loose from the ground and could serve as diaspores (Equihua and Usher, 1993). This could be due to mechanical tension within the moss as it grows longer (Van der Meulen et al., 1987), or due to animals looking for food in the underlying humus layer (Ketner-Oostra and Sýkora, 2008). Stem tips or carpet fragments are large and therefore contribute mainly to short distance spreading, but they could be transported by animals, or due to human activities, e. g. on vehicles, over larger distances.

1.2.3 Ecology

Although C. introﬂexus is among the 100 most invasive species in Europe (including all groups of organisms; Daisie, 2009), knowledge of its ecology is sketchy. All research has been focused into those parts of Europe with oceanic climates (e. g. Hasse and Daniëls, 2006; Sparrius and Kooijman, 2011). Ecological aspects of the Campylopus introﬂexus invasion into diverse 10 generalintroduction

habitats and its competitive potential have been investigated from the 1980s (e. g. Berg, 1985). In contrast to most alien bryophytes, C. introflexus has naturalized in near- natural vegetation such as coastal dunes, grey dunes, bogs, moist or humus- rich heaths, acidic grasslands (Klinck, 2008; Essl and Lambdon, 2009). In the colder climate of Iceland, the South Sandwich Islands and Antarctica Campylopus introflexus can be found on geothermal ground with tempera- tures at the surface of the moss cover above 40 °C (Kennedy, 1996; Convey and Lewis Smith, 2006). In a geothermal area in Italy, its maximum abundance was found on soils of temperature (5 cm below surface) at about 45 °C (Chiarucci et al., 2008). Campylopus introflexus was also found growing on sites where temperature 2.5 cm into the vegetation reached 50–75 °C, but at these hot sites only the top 1–2 cm of the moss vegetation was alive (Convey and Lewis Smith, 2006). Although it colonizes a wide spectrum of habitats in Europe, the microhabitat requirements are more uniform. The species can be found in a wide range of altitudes, but it prefers habitats in low elevations (Richards, 1963). Generally, C. introflexus prefers dry or moderately wet, slightly humose, nutrient-poor and acidic substrates as peat or sandy soils (Zerbe and Wirth, 2006), but it has also been found on the primarily calcareous dunes (Van der Meulen et al., 1987). The species occurs rarely on calcium-rich bedrock, especially where a peat horizon develops (Schlüsslmayr, 2005; Hasse and Daniëls, 2006). Experimental data from Western Europe show better survival under enhanced ammonium concentrations (Sparrius and Kooijman, 2011). The species frequently occurs in pioneer phases of vegetation (e. g. Hasse and Daniëls, 2006), Biermann and Daniëls (1997) documented a rapid increase of C. introflexus cover after disturbances in the vascular-plant or lichen canopy. Disturbance thus seems to be an important factor promoting C. introflexus invasion. C. introflexus characteristically forms extremely dense turfs extending over hundreds of square meters in sandy areas of the Netherlands, in which almost no other bryophyte species can survive. Frequent observations showed inhibition effects of C. introflexus on the colonization of the native species (Van der Meulen et al., 1987; Biermann and Daniëls, 1995, 1997; Ketner- Oostra and Sýkora, 2000, 2004; Sparrius and Kooijman, 2011). However, there is no evidence that a moss carpet of C. introflexus causes permanent inhibition of the long-term development of sandy vegetation (Hasse, 2007). It is because the lichens are probably strong competitors, forcing C. introflexus into disturbed open patches (Biermann and Daniëls, 1997). They are even able to re-colonize the dead as well as living moss carpets developed at disturbed patches (Hasse, 2007; Daniëls et al., 2008; but see Zerbe and Wirth, 2006), or can establish themselves between the mosses in a mixed carpet of C. introflexus and Polytrichum piliferum (Ketner-Oostra and Sýkora, 2008). This is supported by (Minarski and Daniëls, 1996) who observed that after approximately 10 years of Campylopus-dominance in Corynephorus canescens grassland, the lichen vegetation recovered during progressive succession. C. introflexus can be a potential major threat to the native vegetation if it covers a majority of the area and the amount of lichen diaspores in surrounding ar- 1.2 heath star moss campylopus introflexus (hedw.) brid. 11 eas is low (Hasse, 2007). It is estimated that succession to native vegetation to undisturbed localities may take 15–20 years (Daniëls et al., 2008; Biermann and Daniëls, 2001). However, increasing nitrogen deposition may alter the competitive hierarchy (Sparrius and Kooijman, 2011).

1.2.4 Phytosociology

Phytosociology of bryophytes is not frequently studied, but there are several comprehensive works (e. g. Marstaller, 1993; Schlüsslmayr, 2005). Com- munities with C. introflexus are included in the phytosociological order Poly- trichetalia piliferi v. Hübschm. 1975, in the alliance Ceratodonto-Polytrichion piliferi (Waldh. 1947) v. Hübschm 1967. Vegetation from this alliance occurs on acidic soils on silicate bedrock mainly in the subalpine and alpine altitudinal belt. It forms dense carpets with large cover on dry, sunny and sandy sites. The association Campylopodetum polytrichoidis Giac. 1951 was noted for the first time by Marstaller (1993) with the uncertain occurrence of C. introflexus. In his other studies (e. g. Marstaller, 2003, 2004) he distinguished the association Cladonia gracilis-Campylopodetum introflexi Marst. 2001, which was found in slate dump and was characterized by a rich layer of lichen, especially of genus Cladonia, and the dominance of C. introflexus in a moss layer. C. introflexus is frequently found in the another association of the order, Racomitrio-Polytrichetum piliferi v. Hübschm. 1967 (subassociation campylo- podetosum introflexi). This association occurs on bare humus and is charac- terised by the dominance of Polytrichum piliferum. Other frequently occurring bryophyte species are Cephaloziella divaricata, Pohlia nutans and several Cladonia species (Marstaller, 1989). The presence of C. introflexus is occasionally noted in the samples from other associations of the order Ceratodonto-Polytrichion piliferi (e. g. Campy- lopus flexuosus-community, Polytrichetum pallidiseti Marst. 2002, Polytrichetum juniperini v. Krus. 1945 dicranetosum scoparii; Marstaller, 2003, 2004). All this communities are usually found on schist in Central Europe.

1.2.5 Biogeography – native and invasive distribution

Campylopus introﬂexus almost (!) certainly originated from the temperate southern hemisphere — from South America (Chile, Argentina, SE Brazil), southern part of Africa, parts of Australia and several south Paciﬁc, At- lantic and Indian ocean islands (New Caledonia, Subantarctic Islands, New Zealand; Figure 1.3). In the native area it inhabits savannahs (Daisie, 2009), fynbos (Mikulášková, pers. obs.), Eucalyptus or Melaleuca forests, shrublands, grasslands, dry and wet heathlands (Atlas of Living Australia). It prefers temporarily dry to humid, non-calcareous, humus-rich or mineral soil in open habitats, at low altitudes, but it can colonize bases of trees or rotten wood (Gradstein and Sipman, 1978). It often colonizes human-disturbed habitats (e. g. burnt areas, road surroundings, forest clearings and edges, excavated heatlands and peatbogs; Streimann, 2002; Clarkson et al., 2011). 12 generalintroduction

Figure 1.3: The present distribution of Campylopus introﬂexus. Black colour – native distribution, red colour – invasive distribution. (Klinck, 2008, p. 21)

Campylopus introflexus was introduced into Europe probably through ships sailing from the Southern Hemisphere, either on shoes of travellers or in boxes for articles covered with the moss. It was reported for the first time, from Brittany in France, based on a collection made in 1954 (Størmer, 1958). However, an earlier specimen was discovered in the British Isles, collected in 1941 (Richards, 1963). Other earlier specimens were incorrectly determined (Richards, 1963). It was found throughout the U.K. and Ireland by the 1960s and since the beginning of the 1970s it has been rapidly spreading throughout Europe. The recent distribution of the species in Europe ranges from Iceland in the north to Italy in the south, and from Ireland and Portugal in the west to Estonia, Poland and Hungary in the east (Figure 1.4; Hassel and Söderström, 2005; Essl and Lambdon, 2009; Klinck, 2008, 2010). Three localities were known in the Czech Republic in 1996 (Soldán, 1996). It is possible that the moss may have invaded Europe from several points independently (Richards, 1963). The first record in North America dates from 1975 (Frahm, 1980) and was made on a gravel roof of a building in Arcata, California. Previously determined specimens belong to C. pilifer or other species of the genus Campylopus (Frahm, 2007). In North America the species presently spreads in along western coast; it is known form Oregon, USA (Christy et al., 1982) and British Columbia in Canada (Taylor, 1997). 1.2 heath star moss campylopus introflexus (hedw.) brid. 13

Figure 1.4: Neophytic distribution in Europe (red colour). (Klinck, 2008, p. 22; Klinck, 2010, p. 4)

Part II

MAINTHESIS

THEOBJECTANDTHEAIMSOFTHESTUDY 2

This study is focused on moss Campylopus introﬂexus (Hedw.) Brid. in the Czech Republic (Central Europe) — its habitat demands, biological features and genetic background associated with its competitive potential. To a lesser extent, I also studied the distribution of the species in the Czech Republic, and its soil requirements. The synthesis of all obtained information is used for possible prediction of future spread. The overall aim of the thesis was to explore behaviour of Campylopus in- troﬂexus in the Czech Republic and characterize its colonization rate and dispersal ability. The principal objectives can be characterized as follows:

1. To describe habitats colonized by C. introﬂexus in term of vegetation composition and conservation value and to characterize its ecological demands (Paper 1).

2. To characterize mating system, dispersal mode and genetic structure of the species in the Czech Republic (Paper 2, Paper 3).

3. To evaluate the threat for native species and communities (Paper 1, Paper 4).

4. To update the actual distribution pattern of C. introﬂexus in the Czech Republic (Paper 1, Paper 5).

OUTLINEOFTHETHESIS 3

All objectives and planned outputs of the thesis were focused on Campylopus introflexus and its invasion in the Czech Republic. During the first period of the study, the distribution of populations in the Czech Republic and Slovakia was recorded (Paper 5). Vegetation composition of habitats invaded by C. introflexus in the Czech Republic was recorded and respective soil samples were collected. Permanent plots and reciprocal transplantation experiments were established in selected localities in order to compare dispersal and competition ability of the species with native species (Paper 4). Gametophores of C. introflexus were collected from the field and cultivated ex situ and in vitro. Herbarium specimens of most of the populations were gathered. The vegetation plots were classified and the communities typically inhab- ited by C. introflexus were described. Comparison of the vegetation plots with C. introflexus and the plots stored in the Czech National Phytosocio- logical database allowed to predict a future distribution of C. introflexus in the Czech Republic in term of habitat vegetation composition (Paper 1). Cul- tivated plants were used for investigation of soil and water requirements (Paper 1). The plants as well as herbarium specimens were used for genetic analysis (Paper 2, 3). Amplified Fragment Length Polymorphisms (AFLP) and analysis of isozymes appeared to be an easily applicable and accurate method for research of genetic structure and diversity (Paper 3). We showed that protocols for DNA extraction and AFLP needed to be improved, with a particular focus on to calculating a genotyping error rate (Paper 3) for valid evaluation of AFLP fingerprints. The genetic structure of C. introflexus in the Czech Republic was investigated (Paper 2). The thesis is composed of five papers (see list of papers). In this outline, I try

1. to brieﬂy present the main ideas and results of the papers (objectives),

2. to discuss the results in the broader context of Campylopus introﬂexus research (as outlined in Introduction).

3.1 the main ideas and results of papers

3.1.1 Habitats colonized by C. introﬂexus in the Central Europe (Paper 1)

The ﬁrst paper is focused on habitat ecology of C. introﬂexus. In Central Europe, the species colonizes different habitats than in Western and North- ern Europe, where its invasion causes ecological problems in native vegetation. Describing suitable habitats for colonization could allow prediction of its future spread to the east and evaluation of ecological impact of its

19 20 outline of the thesis

invasion in Central Europe. We investigated microhabitat conditions (vegetation composition and ecological characteristics) in Central European habitats by a set of vegetation plots with C. introflexus to document the diversity of colonized habitats. Further, we calculated similarity in species composition between the vegetation-plot records with C. introflexus and a dataset of 26 998 vegetation-plot records without C. introflexus from the Czech National Phytosociological Database (Chytrý and Rafajová, 2003) in order to predict potential habitats to be colonized. C. introflexus frequently colonized certain types of pine forests in Central Europe. We also compared ecological characteristics based on species composition of microhabitats with and without C. introflexus in pine forests, a habitat which is frequently colonized by the species. Ability of C. introflexus to grow in different soil and moisture conditions was tested by cultivation experiments. We found that C. introflexus invades dry, nutrient-poor, acidic soils in a range of vegetation types, most commonly in coniferous forest plantations (with low cover of oaks) and in drained bogs, where it colonizes open patches with low competition pressure, created by anthropogenic disturbances. Data from a Twinspan analysis shows that the Cladonio gracilis-Campylopodetum introflexi could be only part of other, widespread communities in which this neophytic species become dominant. The habitats that appeared to be similar in species composition to that with C. introflexus and which are therefore suitable for colonization are distributed in the major part of the Czech Republic. Differences between the colonized habitats in Western (Northern) Europe and Central Europe are caused by the absence of similar habitats; however, C. introflexus displays an affinity to similar microhabitats. We conclude that although the species will become common in Central Europe because of wide range of potential habitats with favourable vegetation composition and ecological parameters, it presently represents no risk for endangered plant species and communities.

3.1.2 C. introﬂexus dispersal and genetic diversity (Paper 2)

The ability for effective dispersal is a basic prerequisite for plant species to become invasive. C. introflexus profits from both sexual and asexual dispersion. The Czech Republic is on the edge of the invasive area with many distinct populations, so it is a suitable area to study dispersal mechanisms. The genetic structure and diversity of C. introflexus reflect its invasive history in the territory. We can detect multiple colonization events and founder effects. We used isozymes and amplified fragment length polymorphism (AFLP) for studying the genetic structure of populations in the Czech Republic. Both of these markers are targeted to study the entire genome, so they could capture the variability in a this species which is generally low in genetic variability. The individual populations within the Czech Republic had unique genetic patterns, and genetic variation was nearly equally divided among and within populations. Most of the genetic variance occurred among populations within broader regions whereas differentiation among the regions was slight. Relatively low genetic variability within populations of the invaded areas reflected generally low variability of species and/or founder effects in invasive populations. Individual populations in the Czech Republic are 3.1 the main ideas and results of papers 21 mainly multiclonal and appear to originate from multiple colonizing spores. Short-distance spreading by vegetative diaspores occurs mainly within individual cushions. Genetic structure of C. introflexus was not correlated with geography, altitude, pH, age or habitats of populations. We concluded that populations in Czech Republic had multiple origins from the rest of Europe, and the unique genetic patterns of individual populations is a result of prevailing sexual long-distance dispersal.

3.1.3 Molecular methods optimalization (Paper 3)

The genetic structure of C. introflexus was investigated using nucleotide sequences from several segments of DNA. Because the total variability of the species proved to be very small, it could be usefull to use another method which targets the whole genome and includes highly polymorphic markers. Amplified fragment length polymorphism (AFLP) become a standard method for investigating genetic variation in plants, but only in few bryophyta. Published protocols for bryophyta did not give reproducible results for C. introflexus. We first tested a small samples of DNA extraction protocols, because AFLP analysis is sensitive to the quality of template DNA. We compared the yield and purity for DNA stored in different ways (new cultivated plants, fresh plants from field, dried speciments). Then we optimized the protocol for AFLP itself. We tested the reproducibility of AFLP fingerprints produced by the final optimized protocol. The Invisorb Plant Mini Kit produced the highest DNA amount and purity. Newly cultivated stems gave the purest DNA; however, dry herbarium specimens (up to two years old) were suitable too. The replicability of AFLP profiles was not dependent on the way plants were stored. An increased amount of restriction enzymes and prolonged restriction and ligation of up to 10 h were the most important modifications for improving results. The modified protocol was applied to 30 samples and four selective primer combinations, and gave an average genotyping error rate of 0.0451, which could be used for evaluating data for population study in Paper 2.

3.1.4 The role of interspeciﬁc competition in C. introﬂexus spreading (Paper 4)

One of the most important concerns associated with a new invasive species is whether it represents a threat for native species. Presence of invasive species can have an important impact to vegetation composition of native communities. The observation of interactions among invader and native species provide key information for predicting future spread and threats to native communities. Plant species may be invasive either because they share habitats with resident native species and outcompete them, or because they grow in habitats different from those of native species and thus occupies vacant niches. First, we monitored the population dynamics in new patches and estimated the colonization rate of C. introﬂexus in experimental plots on bare soil without the presence of potential competitors. Then we tested whether the colonization rate differs between C. introﬂexus and native species and if 22 outline of the thesis

C. introflexus is able to outcompete native species. For these goals, reciprocal transplantations were established where the inner blocks of cushions were changed between C. introflexus and a native moss species. The behaviour of C. introflexus is dependent of habitat. After initial colonisation, a short lag phase occurred and from the second year C. introflexus started to spread by frequently formed vegetative diaspores. Ceratodon purpureus and Pohlia nutans had similar colonisation rates in unvegetated patches. Hypnum cupressiforme and Polytrichastrum formosum colonised new patches more slowly than C. introflexus, but they were not outcompeted by C. introflexus from mixed patches during the five-year experiment. We observed three patterns in the interaction between C. introflexus and the native species. If development of C. introflexus is not affected by competition with native pioneer species or by disturbance, C. introflexus stands become dense and compact within a few years. We concluded that C. introflexus is not able to outcompete native species, but it could influence community structure by extraordinarily rapid occupation of disturbed unvegetated patches

3.1.5 Distribution of C. introﬂexus in the Czech Republic (Paper 5)

The first report of C. introflexus in the Czech Republic was published in 1990. Then the number of known localities slowly increased, but many specimens were stored only in private herbaria. The distribution of C. introflexus in the Czech Republic was insufficiently known; therefore, update was necessary. The number of localities rapidly increased after 1998, because of both faster spreading rates (more diaspores in surroundings) and a greater attention to this species by bryologists. In 2006, 71 localities of C. introflexus were recorded in the Czech Republic. The species occurs more frequently in the western and southern regions of the Czech Republic, while only scattered occurrences have been recorded from the eastern part of the country. Analysis of distribution in the Czech Republic showed that most populations are distributed in moderately warm regions (Mesophyticum) in the altitudes between 230 and 800 m a. s. l.

3.2 the results of papers in the broader context of species biology

3.2.1 Dispersal mode

Dynamics of colonization of new microhabitats correspond strongly to the classic colonist life strategy. Both frequent vegetative and generative reproduction are important prerequisites for invasive behaviour of bryophyte species. Forming high numbers of vegetative diaspores is probably induced by ecological factors taking place in field. We cultivated in vitro or ex situ plants on a suitable substrate in standard laboratory conditions (see Paper 1) for three years, and used them as a source material for genetic analysis (Pa- per 2, 3). These plants did not form any vegetative diaspores. In the field, vegetative diaspores started to form during the second year at the latest. The differences between field- and experimentally-grown plants could re- 3.2 the results of papers in the broader context of species biology 23

flect a trade-off between forming secondary protonema and forming vegetative diaspores. Cultivated plants frequently formed secondary protonema from stems and leaves, and use it for spreading into cultivation containers or Petri dishes. New gametophores were frequently formed from buds on protonema. Both latter ways of spreading are clonal, but energetic cost for forming protonema or diaspores as well as their resistance are different. It is obvious that the species forms big vegetative diaspores (whose formation consumes more energy) as a main dispersal mode for short distances in natural conditions with fluctuating moisture, while when water is well available it forms mainly secondary protonema, which forming require less energy but on the other hand is more sensitive to lack of water. If the development of C. introflexus is not hampered by competition with native species or by disturbance, C. introflexus stands become dense and compact within a few years. Theoretically predicted dispersal of vegetative diaspores by animals or humans has not been confirmed. If the species profits from zoochory or antropochory, we could expect a large number of populations close to each other in a small area (e. g. edge of animal forest paths or along touristic paths and trails). This distribution pattern was not confirmed; populations in the Czech Republic are mostly discrete and of small size. Additional evidence for low levels of zoochory or antropochory is the lack of correlation between genetic and geographical distance. The species appears to be dispersed among suitable sites by spores, as indicated by because the fact that each population has a unique genotypic pattern. The great distances among populations resulted in lower gene flow among them and increasing their genetic isolation and limited possibilities of sporophyte formation; the species is dioicous and spermatozoids move only to very short distances. However, genetic isolation and a lack of sporophytes were not observed. Almost a third of population in the Czech Republic are fertile and populations are genetically very close. This pattern can be explained by the simultaneous spread of a group of coherent spores, which enable the early formation of sporophytes in populations.

3.2.2 Invaded habitats

The invasion into new regions may be accompanied by changes in the genome, which enable invasive species to adapt to new conditions. Then individual genotypes can be linked to a specific habitat type and therefore we can observe tight coincidence between colonised habitat and genetic structure. We can speculate whether invasion history of C. introflexus in Europe is sufficiently long to be reflected in genome structure. The phase of genetic adaptation is often accompanied by the lag stage when the species does not spread rapidly. C. introflexus is successfully spreading eastward within decades and is able to colonize a wide range of habitats. Despite that, it has low genetic variability and it is not possible to trace the correlation between commonly colonized habitats (Paper 1) and genetic structure (Paper 2). This result could indicate that C. introflexus is able to tolerate very diverse en- 24 outline of the thesis

vironments in its native area, and that specific genetic adaptations to new habitats was not necessary. Despite the seeming heterogeneity of invaded habitats, one common feature can be found – in all cases the species grew on bare soil with low competitive pressure from vascular plants, as in its source area (Southern Hemi- sphere). The species is able to survive in habitats with prolonged frosts (Gi- ant Mountains, edges of high raised peatbogs), but there forms only small populations. The low frost tolerance could be a reason why the expansion seems to weaken eastwards. The number of newly colonised localities has increased more slowly in the Czech Republic and Slovakia in recent years than in previous decades (Paper 1, 5). This could be explained by the intense interest of bryologists after discovery of the first finding in 1988 in the Czech Re- public. In spite of this, it seems that the species spreads eastwards with less intensity than in previous decades even if there is huge amount of suitable habitats in Central Europe (Paper 1) and huge numbers of spores produced in Western Europe. One possible explanation could be frequent occurrence of black frost in more continental areas, which may limit C. introflexus spread, but direct evidence is lacking. Distribution of C. introflexus is concentrated in the hilly areas of temperate deciduous forests and only few populations were found in mountains. Our research confirmed (Paper 1, 4) that the species prefers low competition from vascular plants. In many Central-European mountains there are ecologically suitable microhabitats, including low-competition patches. Despite this, C. introflexus has not colonized these areas recently and a lack of frost tolerance is the most plausible explanation. The same explanation may be applied on the species absent in the sub-continental Pannonian region (south-eastern Czech Republic), together with high amount of lime and salts in soils, and extreme drought in summer. The strongest plant and animal invasive species in Europe successfully outcompete native species from their natural environment, what has consequences for ecosystem functioning and may have economic impacts. C. introflexus is the example that even the most invasive bryophyte in Europe has a much lower impact on native plants and animals as compared to vascular plants. Its impact on natural communities of Central Europe is low, mainly because C. introflexus prospers in human influenced habitats with low conservation value (Paper 1) and in Central European habitats it is not able to overgrow and displace native species (3). The biggest danger lies in the rapidly forming dense cushions that may block natural succession. This risk usually acts only at very fine scale of few square meters, so it does not endanger the occurrence of natural early-succession communities. The only habitat where C. introflexus is a serious problem in Central Europe is in degraded peatbogs, where the species blocks succession during restoration, even at large scales. Meadows and pristine mires are not impacted as they have a dense canopy of vascular plants and bryophytes, or a high groundwater level. Generally, invasive species predominantly colonize edges of roads and rivers. C. introflexus inhabits less frequently invaded habitats such as coniferous forests. Due to the fact that C. introflexus is not a strong competitor, it 3.2 the results of papers in the broader context of species biology 25 prefers habitats where it meets only few native species. It is the reason why it prefers human-disturbed habitats where autogenic succession is blocked. Most of the Czech Republic is intensively managed by man, with prevailing monodominant plantations. Therefore, theoretically C. introflexus can grow there in the thousands of microhabitats. Because suitable habitats are frequent in the Czech Republic, the invasion intensity of C. introflexus is dependent mainly on the amount of both generative and vegetative diaspores in diaspore bank. In the past, Czech Republic was repeatedly colonised by C. introflexus from source populations lying probably in Western Europe and we can expect a similar flow of spores in the future. The amount of source spores will grow and increase the probability that populations grow, and its distribution in Czech Republic will be large. Effective dispersal of male and female spores will also promote the formation of sporophytes. This would allow more rapid colonization of the eastern parts of the country and accel- erate the invasion towards the east.

3.2.3 Distribution

C. introflexus has been recorded in the entire area of the Czech Republic, which is situated on the eastern edge of the invaded range. The oldest localities are randomly scattered across the entire area of the Czech Republic. Increasing numbers of recorded localities with C. introflexus indicate that spread of species is currently unfinished. The number of localities increased since three published records in 1996 (Soldán, 1996), up to almost 80 known records in 2006. More than one hundred herbarium specimens and published as well as unpublished records have been collected to date (Figure 3.1). The increase of the number of localities occurs mainly in the western part of the country, while in the eastern part the records are still rare. One of the possible reasons for this distribution pattern is (besides lower amount of suitable habitats in the southeast) the earlier colonization of the western part of the Czech Republic. Due to the fact that the Czech Republic was colonized in the past repeatedly, we can assume that in Western Europe there is a stable source of spores from which C. introflexus can continue to spread eastward. It can be assumed that a large number of undetected populations still exist in our area, because Campylopus introflexus grows in habitats that are neglected by bryologists. The potentially invaded plant communities are distributed across most of the Czech Republic, therefore we can expect that the number of known sites will increase, especially in the western part of the country. 26 outline of the thesis

Figure 3.1: Campylopus introﬂexus present distribution in Czech Republic. CONCLUSION 4

The main results of the thesis can be outlined as follows: • Most of the Czech Republic has the potential to be colonized by C. introflexus. • Both native and non-native pine forests include many microhabitats but only some of them are suitable for colonization by C. introflexus. Its occurrence is most likely in dry acidic pine forests both with and without developed lichen layer. It prefers stands with little representation of oak in the tree layer. • C. introflexus is able to survive on a wide range of soils, but thrives best on the acid organic to sandy soils with low nutrient content. It grows on habitats with low to medium moisture. Limestone soils and waterlogged soils have a harmful effect on his gametophores, but on waterlogged soil is able to survive on a elevated bare soil on disturbed places. • Suitable microhabitat conditions are essential for the occurrence of C. introflexus. In Central Europe, it is pioneer species that colonizes patches with bare soil without high competitive pressure from vascular plants, lichens and bryophytes. • It seems that C. introflexus is not a threat to native species. It colonizes habitats where common bryophyte species widespread in Central Eu- rope occurs. C. introflexus does not outcompete native species from their natural habitats. It influences the composition of communities, because it occupies a substrates for native pioneer species and blocks subsequent succession. • Native species were not able to displace C. introflexus from mixed patches within five years and vice versa. • It has been verified that AFLPs is suitable molecular technique for the study of genetic variability within less variable species such as C. introflexus. • It was confirmed, that not-too-old herbarium bryophyte speciments can be successfully used for DNA extractions. • The Czech Republic was colonized by spores from minimally two genetically different sources. There is no correlation between genetic structure within country and geographic position, altitude and measured ecological factors. • Lateral spreading is ensured by both sexual spores and vegetative diaspores. Dispersion by vegetative diaspored is confirmed only among cushions within a single locality.

27 28 conclusion

• C. introﬂexus is in the Czech Republic low in genetic variability, what could reﬂect a low general variability within the species, and/or founder effect within European populations. conclusion 29 references

Amsellem, L., J.L. Noyer and M. Hossaert-McKey (2001). “Evidence for a switch in the reproductive biology of Rubus alceifolius (Rosaceae) towards apomixis, between its native range and its area of introduction.” American Journal of Botany 88, pp. 2243–2251. Berg, C. (1985). “Zur Ökologie der neophytischen Laubmoosart Campylopus introflexus (Hedw.) Brid. in Mecklenburg.” Archiv der Freunde der Naturge- schichte 25, pp. 117–126. Biermann, R. and F.J.A. Daniëls (1995). “Campylopus introflexus (Dicranaceae, Bryopsida) in Flechtenreichen Silber-Grasfluren Mitteleuropas.” In: Flechten Follmann. Contributions to lichenology in Honour of Gerhard Follmann. Ed. by F.J.A. Daniëls, M. Schulz and J. Peine. Cologne: Botanical Institute, Univer- sity of Cologne, pp. 493–500. –(1997). “Changes in a lichen-rich dry sand grassland vegetation with special reference to lichen synusiae and Campylopus introflexus.” Phytocoenolo- gia 27, pp. 257–273. –(2001). “Vegetationsdynamik im Spergulo-Corynephoretum unter besonderer Berücksichtigung des neophytischen Laubmooses Campylopus introflexus.” Braunschweiger Geobotanische Arbeiten 8, pp. 27–37. Burke, M.J.W. and J.P. Grime (1996). “An experimental study of plant community invasibility.” Ecology 77.3, pp. 776–790. Atlas of Living Australia. Campylopus introflexus (Hedw.) Brid. – Atlas of Living Australia. url: http://www.ala.org.au (visited on 05/06/2012). Carranza, M.L., C. Ricotta, M. Carboni and A.T.R. Acosta (2011). “Habitat selection by invasive alien plants. A bootstrap approach.” Preslia 83, pp. 529– 536. Chapman, H., B. Robson and M.L. Pearson (2004). “Population genetic structure of a colonising, triploid weed, Hieracium lepidulum.” Heredity 92, 182– 188. Chiarucci, A., M. Calderisi, F. Casini and I. Bonini (2008). “Vegetation at the limits for vegetation: Vascular plants, bryophytes and lichens in a geothermal field.” Folia Geobotanica 43, pp. 19–33. Christy, J.A., J.H. Lyford and D.H. Wagner (1982). “Checklist of Oregon Mosses Checklist of Oregon Mosses.” Bryologist 85, pp. 22–36. Chytrý, M. and M. Rafajová (2003). “Czech National Phytosociological Data- base: basic statistics of the available vegetation-plot data.” Preslia 75, pp. 1– 15. Chytrý, M., P. Pyšek, L. Tichý, I. Knollová and J. Danihelka (2005). “Invasions by alien plants in the Czech Republic: a quantitative assessment across habitats.” Preslia 77, pp. 339–354. Chytrý, M., L.C. Maskell, J. Pino, P. Pyšek, M. Vilà, X. Font and S.M. Smart (2008). “Habitat invasions by alien plants: a quantitative comparison among Mediterranean, subcontinental and oceanic regions of Europe.” Journal of Applied Ecology 45, pp. 448–458. Chytrý, M., P. Pyšek, J. Wild, J. Pino, L.C. Maskell and M. Vilà (2009a). “Eu- ropean map of alien plant invasions based on the quantitative assessment across habitats.” Diversity and Distributions 15, pp. 98–107. 30 conclusion

Chytrý, M., J. Wild, P. Pyšek, L. Tichý, J. Danihelka and I. Knollová (2009b). “Maps of the level of invasion of the Czech Republic by alien plants.” Preslia 81.3, pp. 187–207. Clarkson, B.R., M.C. Smale, P.A. Williams, S.K. Wiser and R.P. Buxton (2011). “Drainage, soil fertility and fire frequency determine composition and structure of gumland heaths in northern New Zealand.” New Zealand Jour- nal of Ecology 35.1, p. 96. Colautti, R.I., C.G. Eckert and S.C.H. Barrett (2010). “Evolutionary constraints on adaptive evolution during range expansion in an invasive plant.” Pro- ceedings of the Royal Society B: Biological Sciences 277, pp. 1799–1806. Convey, P. and R.I. Lewis Smith (2006). “Geothermal bryophyte habitats in the South Sandwich Islands, maritime Antarctic.” Journal of Vegetation Sci- ence 17.4, pp. 529–538. Crooks, J.A. and M.E. Soulé (1999). “Lag times in population explosions of invasive species: Causes and implications.” In: Invasive Species and Biodiver- sity Management. Ed. by O.T. Sandlund, P.J. Schei and A. Viken. Trondheim: Kluwer Academic Publishers, pp. 103–125. Daisie (2009). Handbook of alien species in Europe. Vol. 3. Invading nature. Dor- drecht: Springer, p. 399. Daniëls, F.J.A., A. Minarski and O. Lepping (2008). “Dominance pattern changes of lichen-rich Corynephorus grassland in the inland of the Nether- lands.” Annali di Botanica 8, pp. 9–19. Dlugosch, K.M. and I.M. Parker (2008). “Founding events in species invasions: genetic variation, adaptive evolution, and the role of multiple introductions.” Molecular Ecology 17, pp. 431–449. Duckett, J. G and R. Ligrone (1992). “A survey of diaspore liberation mechanisms and germination patterns in mosses.” Journal of bryology 17, pp. 335– 354. Durka, W., O. Bossdorf, D. Prati and H. Auge (2005). “Molecular evidence for multiple introductions of garlic mustard (Alliaria petiolata, Brassicaceae) to North America.” Molecular Ecology 14, pp. 1697–1706. Elton, C.S. (1958). The ecology of invasions by animals and plants. London: Me- thuen. Equihua, M. and M.B. Usher (1993). “Impact of carpets of the invasive moss Campylopus introflexus on Calluna vulgaris regeneration.” Journal of Ecology 81, pp. 359–365. Essl, F. and P.W. Lambdon (2009). “Alien bryophytes and lichens of Europe.” In: Handbook of Alien Species in Europe. Ed. by Daisie. Dordrecht: Springer, pp. 29–41. Essl, F., W. Rabitsch and P. Lambdon (2011). “Bryophytes and lichens.” In: Encyclopedia of biological invasions. Ed. by D. Simberloff and M. Rejmánek. Berkeley: University of California Press, pp. 81–85. Frahm, J.P. (1974). “Zur Unterscheidung und Verbreitung von Campylopus introflexus (Hedw.) Brid. und C. polytrichoides De Not.” Revue bryologique et lichénologique 40, pp. 33–44. –(1980). “Synopsis of the genus Campylopus in North America north of Mex- ico.” Bryologist 83, pp. 570–588. conclusion 31

–(2007). “Campylopus introflexus.” In: Flora of North America North of Mex- ico. Ed. by Flora of North America Editorial Committee. Vol. 27. New York and Oxford, p. 371. url: http://www.efloras.org/florataxon.aspx? flora_id=1&taxon_id=105474 (visited on 09/20/2011). –(2009). “Diversity, dispersal and biogeography of bryophytes (mosses).” In: Protist Diversity and Geographical Distribution. Ed. by W. Foissner and D.L. Hawksworth. Netherlands: Springer, pp. 43–50. –(2012). “The phytogeography of European bryophytes.” Botanica Serbica 36, pp. 23–36. Frahm, J.P. and M. Stech (2006). “The taxonomic status of intermediate forms of Campylopus introflexus (Hedw.) Brid. and C. pilifer Brid.(Dicranaceae, Bry- opsida) newly discovered in Europe.” Cryptogamie. Bryologie 27, pp. 213– 223. Giacomini, V. (1955). “Sull’autonomia specifica e sul ciclo di forme de Campy- lopus polytrichoides De Not.” Atti Istituto di Botanica, Università Pavia, Labo- ratorio Di Crittogama 13, pp. 45–83. Glime, J.M. (2007). “Volume 5. Uses.” In: Bryophyte ecology. Ebook sponsored by Michigan Technological University and the International Association of Bryologists. url: http://www.bryoecol.mtu.edu (visited on 05/12/2012). –(2012). “Volume 2. Bryological interactions.” In: Bryophyte ecology. Ebook sponsored by Michigan Technological University and the International As- sociation of Bryologists. url: http://www.bryoecol.mtu.edu (visited on 05/12/2012). Goffinet, B., W.R. Buck and A.J. Shaw (2008). “Morphology, anatomy and classification of the Bryophyta.” In: Bryophyte Biology. Ed. by A.J. Shaw and B. Goffinet. 2nd ed. Cambridge: Cambridge University Press, pp. 55– 138. Gradstein, S.R. and H.J.M. Sipman (1978). “Taxonomy and world distribution of Campylopus introflexus and C. pilifer (= C. polytrichoides): a new synthesis.” Bryologist 81, pp. 114–121. Gunnarsson, U., A.J. Shaw and M. Lönn (2007). “Local-scale genetic structure in the peatmoss Sphagnum fuscum.” Molecular Ecology 16, pp. 305–312. Hasse, T. (2007). “Campylopus introflexus invasion in a dune grassland: succession, disturbance and relevance of existing plant invader concepts.” Herzogia 20, pp. 305–315. Hasse, T. and F.J.A. Daniëls (2006). “Species responses to experimentally induced habitat changes in a Corynephorus grassland.” Journal of Vegetation Science 17, pp. 135–146. Hassel, K. and L. Söderström (2005). “The expansion of the alien mosses Orthodontium lineare and Campylopus introflexus in Britain and continental Europe.” Journal of the Hattori Botanical Laboratory 97, pp. 183–193. Hedenäs, L., T. Herben, H. Rydin and L. Söderström (1989a). “Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and population structure.” Holoartic Ecology 12, pp. 163–172. –(1989b). “Ecology of the invading moss species Orthodontium lineare in Sweden: substrate preference and interactions with other species.” Journal of Bryology 15, pp. 565–581. 32 conclusion

Hejda, M., P. Pyšek, J. Pergl, J. Sádlo, M. Chytrý and V. Jarošík (2009). “Inva- sion success of alien plants: do habitat affinities in the native distribution range matter?”: Global Ecology and Biogeography 18, pp. 372–382. Herben, T. (1994). “Local rate of spreading and patch dynamics of an invasive moss species, Orthodontium lineare.” Journal of Bryology 18, pp. 115– 125. Herben, T., H. Rydin and L. Söderström (1991). “Spore establishment probability and the persistence of the fugitive invading moss, Orthodontium lineare - a spatial simulation-model.” Oikos 60, 215–221. Howard, R.J., S.E. Travis and B.A. Sikes (2008). “Rapid growth of a Eurasian haplotype of Phragmites australis in a restored brackish marsh in Louisiana, USA.” Biological Invasions 10.3, pp. 369–379. Hufbauer, R.A. (2004). “Population Genetics of Invasions: Can We Link Neu- tral Markers to Management?”: Weed technology 18, pp. 1522–1527. Husband, B.C. and S.C.H. Barrett (1991). “Colonization history and population genetic structure of Eichhornia paniculata in Jamaica.” Heredity 66, pp. 287–296. Kennedy, A.D. (1996). “Antarctic fellfield response to climate change: a tri- partite synthesis of experimental data.” Oecologia 107.2, pp. 141–150. Ketner-Oostra, R. and K.V. Sýkora (2000). “Vegetation succession and lichen diversity on dry coastal calciumpoor dunes and the impact of management experiments.” Journal of Coastal Conservation 6, pp. 191–206. –(2004). “Decline of lichen-diversity in calcium-poor coastal dune vegetation since the 1970s, related to grass and moss encroachment.” Phyto- coenologia 34, pp. 521–549. –(2008). “Vegetation change in a lichen-rich inland drift sand area in the Netherlands.” Phytocoenologia 38.4, pp. 267–286. Klinck, J. (2008). “The alien invasive moss Campylopus introflexus in the Dan- ish coastal dune system.” MA thesis. Depon. in: Copenhagen University, Copenhagen, p. 106. –(2010). NOBANIS – Invasive Alien Species Fact Sheet – Campylopus introflexus. url: http://www.nobanis.org (visited on 05/11/2012). Kolbe, J.J., R.E. Glor, L.R. Schettino, A.C. Lara, A. Larson and J.B. Losos (2004). “Genetic variation increases during biological invasion by a Cuban lizard.” Nature 431, pp. 177–181. Kotowski, W., W. Thörig, R. Diggelen and M.J. Wassen (2006). “Competition as a factor structuring species zonation in riparian fens-a transplantation experiment.” Applied Vegetation Science 9, pp. 231–240. Levine, J.M. (2000). “Species diversity and biological invasions: relating local process to community pattern.” Science 288.5467, pp. 852–854. Levine, J.M. and C.M. D’Antonio (1999). “Elton revisited: a review of evidence linking diversity and invasibility.” Oikos 87, pp. 15–26. Lockwood, J.L., M.F. Hoopes and M.P. Marchetti (2007). Invasion ecology. Wiley-Blackwell publishing. Lonsdale, W.M. (1999). “Global patterns of plant invasions and the concept of invasibility.” Ecology 80.5, pp. 1522–1536. Löbel, S., J. Dengler and C. Hobohm (2006). “Species richness of vascular plants, bryophytes and lichens in dry grasslands: the effects of en- conclusion 33

vironment, landscape structure and competition.” Folia Geobotanica 41.4, pp. 377–393. Marstaller, R. (1989). “Die Moosgesellschaften des Verbandes Ceratodonto- Polytrichion piliferi (Waldheim 1947) v. Hübschmann 1967. 38. Beitrag zur Moosvegetation Thüringens.” Gleditschia 17, pp. 107–120. –(1993). “Synsystematische Übersicht über die Moosgesellschaften Zentraleu- ropas.” Herzogia 9.3-4, pp. 513–541. –(2003). “Die Moosgesellschaften des Schieferbruches Kirchberger Gluck bei Reichenbach (Kreis Saalfeld-Rudolstadt). 92. Beitrag zur Moosvegeta- tion Thüringens.” Herzogia 16, pp. 221–238. –(2004). “Bryosoziologische Studien auf der Rehbachhalde bei Schmiede- bach (Landkreis Saalfeld-Rudolstadt). 98. Beitrag zur Moosvegetation Thü- ringens.” Herzogia 17, pp. 245–267. Minarski, A. and F.J.A. Daniëls (1996). “Veränderungen om Dominanzmuster von Kryptogamen-Synusien und Gräsern in einem Sandtrockenrasen-Bestand in den Niederlanden om Zeitraum von 1981 bis 2004.” Arbeiten aus dem In- stitut für Landschaftsökologie, Münster 15, pp. 39–41. Moravcová, L., P. Pyšek, V. Jarošík, V. Havlícková and P. Zákravský (2010). “Reproductive characteristics of neophytes in the Czech Republic: traits of invasive and non-invasive species.” Preslia 82, pp. 365–390. Muñoz, J., Á.M. Felicísimo, F. Cabezas, A.R. Burgaz and I. Martínez (2004). “Wind as a long-distance dispersal vehicle in the Southern Hemisphere.” Science 304, pp. 1144–1147. Mälson, K. and H. Rydin (2009). “Competitive hierarchy, but no competitive exclusions in experiments with rich fen bryophytes.” Journal of Bryology 31, pp. 41–45. Pairon, M., B. Petitpierre, M. Campbell, A. Guisan, O. Broennimann, P.V. Baret, A.L. Jacquemart and G. Besnard (2010). “Multiple introductions boosted genetic diversity in the invasive range of black cherry (Prunus serotina; Rosaceae).” Annals of botany 105, pp. 881–890. Pellegrin, D. and D.P. Hauber (1999). “Isozyme variation among populations of the clonal species, Phragmites australis (Cav.) Trin. ex Steudel.” Aquatic Botany 63.3-4, pp. 241–259. Pinke, G., P. Karácsony, B. Czúcz and Z. Botta-Dukát (2011). “Environmental and land-use variables determining the abundance of Artemisia ambrosiifo- lia in arable ﬁelds in Hungary.” Preslia 83, pp. 219–236. Prentis, P.J., J.R.U. Wilson, E.E. Dormontt, D.M. Richardson and A.J. Lowe (2008). “Adaptive evolution in invasive species.” Trends in Plant Science 13, pp. 288–294. Pyšek, P. and K. Prach (1993). “Plant Invasions and the Role of Riparian Habitats: A Comparison of Four Species Alien to Central Europe.” Journal of Biogeography 20.4, pp. 413–420. Pyšek, P. and D.M. Richardson (2007). “Traits associated with invasiveness in alien plants: where do we stand?”: In: Biological Invasions. Ed. by W. Nentwig. Berlin: Springer, pp. 97–125. Pyšek, P., K. Prach and B. Mandák (1998). “Invasions of alien plants into habitats of Central European landscape: an historical pattern.” In: Plant invasions ecological mechanisms and human responses. Ed. by U. Starﬁnger, 34 conclusion

K. Edwards, I. Kowarik and M. Williamson. Leiden: Backhuys Publishing, pp. 23–32. Pyšek, P., V. Jarošík, M. Chytrý, Z. Kropáˇc,L. Tichý and J. Wild (2005). “Alien plants in temperate weed communities: Prehistoric and recent invaders occupy different habitats.” Ecology 86.3, pp. 772–785. Pyšek, P., V. Jarošík and J. Pergl (2011). “Alien plants introduced by different pathways differ in invasion success: unintentional introductions as a threat to natural areas.” PLoS One 6.9, e24890. Ramaiya, M., M.G. Johnson, B. Shaw, J. Heinrichs, J. Hentschel, M. Von Kon- rat, P.G. Davison and A.J. Shaw (2010). “Morphologically cryptic biological species within the liverwort Frullania asagrayana.” American Journal of Botany 97, pp. 1707–1718. Richards, P.W. (1963). “Campylopus introflexus (Hedw.) Brid. and C. polytrichoides De Not. in the British Isles; a preliminary account.” Transactions of the British Bryological Society 4, pp. 404–417. Rydin, H. (1997). “Competition among bryophytes.” Advances in Bryology 6, pp. 135–168. Sakai, A.K., F.W. Allendorf, J.S. Holt, D.M. Lodge, J. Molofsky, K.A. With, S. Baughman, R.J. Cabin, J.E. Cohen, N.C. Ellstrand et al. (2001). “The Popu- lation Biology of Invasive Specie.” Annual Review of Ecology and Systematics 32, pp. 305–332. Schirmel, J., L. Timler and S. Buchholz (2011). “Impact of the invasive moss Campylopus introflexus on carabid beetles (Coleoptera: Carabidae) and spiders (Araneae) in acidic coastal dunes at the southern Baltic Sea.” Biological Invasions 13, pp. 605–620. Schlüsslmayr, G. (2005). “Soziologische Moosflora des südöstlichen Oberöster- reich.” Stapfia 84, pp. 1–695. Simberloff, D. and M. Rejmánek, eds. (2011). Encyclopedia of biological invasions. London: Univiversity of California Press. Soldán, Z. (1996). “Rozšíˇrení neofytických mech ˚u Campylopus introflexus a Or- thodontium lineare v Ceskéˇ republice [Distribution of Campylopus introflexus a Orthodontium lineare in Czech Republic].” Brynora 18, pp. 10–19. Sparrius, L.B. and A.M. Kooijman (2011). “Invasiveness of Campylopus introflexus in drift sands depends on nitrogen deposition and soil organic matter.” Applied Vegetation Science 14, pp. 221–229. Starfinger, U., K. Edwards, I. Kowarik and M. Williamson, eds. (1998). Plant invasions: ecological mechanisms and human responses. Leiden: Backhuys Pub- lishing. Stech, M. and J. Dohrmann (2004). “Molecular relationships and biogeography of two Gondwanan Campylopus species, C. pilifer and C. introflexus (Dicranaceae).” Monographs in Systematic Botany from the Missouri Botanical Garden 98, pp. 415–431. Stech, M. and D. Wagner (2005). “Molecular relationships, biogeography, and evolution of Gondwanan Campylopus species (Dicranaceae, Bryopsida).” Taxon 54, pp. 377–382. Streimann, H. (2002). The Mosses of Norfolk Island. Canberra: ABRS. conclusion 35

Størmer, P. (1958). “Some mosses from the phytogeographical excursion 1-9 through the Armorican massive in 1954.” Revue bryologique et lichénologique 27, pp. 13–16. Sundberg, S. (2005). “Larger capsules enhance short-range spore dispersal in Sphagnum, but what happens further away?”: Oikos 108, pp. 115–124. Söderström, L. (1992). “Invasions and range expansions and contractions of bryophytes.” In: Bryophytes and lichens in a changing environment. Ed. by J.W. Bates and A.M. Farmer. New York: Clarendon Press, Oxford University Press, pp. 131–158. Taylor, T. (1997). “Campylopus introﬂexus — Moss introduced in British Col- umbia.” Botanical Electronic News 162. Torimaru, T., N. Tomaru, N. Nishimura and S. Yamamoto (2003). “Clonal diversity and genetic differentiation in Ilex leucoclada M. patches in an old- growth beech forest.” Molecular ecology 12, pp. 809–818. Van der Meulen, F., H. Van der Hagen and B. Kruijsen (1987). “Campylopus introﬂexus. Invasion of a moss in Dutch coastal dunes.” Proceedings of the Koninklijke. Nederlandse Akademie van Wetenschappen. 90, pp. 73–80. Van Kleunen, M., E. Weber and M. Fischer (2010). “A meta-analysis of trait differences between invasive and non-invasive plant species.” Ecology Let- ters 13.2, pp. 235–245. Van Zanten, B.O. and T. Pócs (1981). “Distribution and dispersal of bryophytes.” Advances in bryology 1, pp. 479–562. Walker, N.F., P.E. Hulme and A.R. Hoelzel (2003). “Population genetics of an invasive species, Heracleum mantegazzianum: implications for the role of life history, demographics and independent introductions.” Molecular Ecology 12, pp. 1747–1756. Wang, T., Y. Su and G. Chen (2008). “Population genetic variation and structure of the invasive weed Mikania micrantha in Southern China: consequences of rapid range expansion.” Journal of Heredity 99, pp. 22–33. Ward, S. (2006). “Genetic analysis of invasive plant populations at different spatial scales.” Biological Invasions 8, pp. 541–552. Williamson, M.H. (1996). Biological invasions. New York: Chapman and Hall. Zerbe, S. and P. Wirth (2006). “Non-indigenous plant species and their ecological range in Central European pine (Pinus sylvestris L.) forests.” Annals of forest science 63, pp. 189–203. Zhang, Y.Y.E., D.A.Y. Zhang and S.C.H. Barrett (2010). “Genetic uniformity characterizes the invasive spread of water hyacinth (Eichhornia crassipes), a clonal aquatic plant.” Molecular Ecology 19, pp. 1774–1786.

Part III

PAPERS

LISTOFPAPERS 5

My thesis is based on the following ﬁve papers, mentioned in the previous text as Papers 1–5.

Paper 1 Mikulášková E., Fajmonová Z. & Hájek M. 2012. Invasion of the moss Campy- lopus introﬂexus into central European habitats. Preslia [after major revisions]; see chapter 7

Paper 2 Mikulášková E. & Fér T.: Patterns of genetic diversity and dispersal of Campy- lopus introﬂexus (Bryophyta) in the edge of invasive area. Ms. [presently under review in Journal of Bryology]; see chapter 8

Paper 3 Mikulášková E., Fér T. & Kuˇcabová V. 2012: The effect of different DNA isolation protocols and AFLP ﬁngerprinting optimizations on error rate estimates in the bryophyte Campylopus introﬂexus. Lindbergia, 35:7-17; see chapter 9

Paper 4 Mikulášková E., Hájek M. & Kintrová K: The role of interspeciﬁc competition in spreading of an invasive moss species: transplantation experiments. Ms.; see chapter 10

Paper 5 Mikulášková E. 2006. Vývoj rozšíˇrení neofytického mechu Campylopus intro- ﬂexus v Ceskéˇ republice. Bryonora, 38:1-10. [Development in distribution of the neophytic moss Campylopus introﬂexus in the Czech Republic — in Czech, translated to english]; see chapter 11

AUTHOR´S CONTRIBUTION 6

Paper 1 E. Mikulášková was responsible for the idea and planning the study. M. Há- jek and Z. Fajmonová helped with planning data analysis. E. Mikulášková sampled vegetation plots with C. introflexus, collected soil samples and cultivated C. introflexus. Z. Fajmonová was responsible for classification and large-scale analysis of vegetation plots. E. Mikulášková interpreted results and led the writing.

Paper 2 E. Mikulášková was responsible for the idea and planning the study. She collected and determinated the moss samples, cultivated gametophores for isozyme analysis, performed DNA isolation and following AFLP and isozymes analysis, she wrote the paper with contribution of T. Fér.

Paper 3 The study was planned jointly. E. Mikulášková collected and determinated the moss samples, cultivated gametophores, performed with help of V. Kuˇcabová DNA isolation and following AFLP analysis. She wrote the paper, T. Fér helped to improve the text of the manuscript.

Paper 4 E. Mikulášková was responsible for the idea and planning the study. She determinated the moss samples, established permanent plots and monitored them within ﬁve years. K. Kintrová was responsible for statistical analysis. E. Mikulášková led the writing, both co-authors contributed to writing of the manuscript.

INVASIONOFTHEMOSS CAMPYLOPUSINTROFLEXUS 7 INTOCENTRALEUROPEANHABITATS

Eva Mikulášková1,2, Zuzana Fajmonová1,3& Michal Hájek1,3 1Dept. of Botany and Zoology, Faculty of Science, Masaryk University, CZ-61137 Brno, Czech Republic; [email protected]. 2Dept. of Botany, Faculty of Science, Charles University in Prague, CZ-128 01 Praha 2, Czech Republic. 3Department of Vegetation Ecology, Institute of Botany, Academy of Sciences of the Czech Republic, Lidická 25/27, CZ-657 20 Brno, Czech Republic. abstract

Although vascular plant invasions have been frequently studied, little is known about invasive bryophytes. Campylopus introflexus is an invasive moss endangering natural vegetation in Western Europe and extending its secondary area eastwards. Therefore, we studied its ecology in the Czech Re- public (Central Europe). We updated its distribution, described colonised habitats in term of vegetation composition, investigated substrate and water demands, and investigated which habitats in the Czech Republic are potentially endangered by C. introflexus invasion. The first dataset contained 78 vegetation plots with C. introflexus across the whole investigated area covers the diversity of colonised habitats. The second dataset contained published vegetation plots of pine forests both with and without C. introflexus for fine scale distinguishing of habitat preferences within this habitat. Vegetation-plot records from both datasets were numerically classified. We further calculated similarity in species composition between vegetation-plot records with C. introflexus and dataset of 26 998 vegetation-plot records without C. introflexus stored in a large database in order to predict potential habitats to be colonised. Ecological demands were characterized by in situ research (soil samples from 52 recorded vegetation plots) and ecological interpretation of the pine forest dataset. Further, cultivation experiment was established with populations from 20 study sites in order to test the ability of C. introflexus to grow in different soil and moisture conditions; data were evaluated by linear mixed effect models. We found that C. introflexus does invade dry, nutrient poor acidic soils in a range of vegetation types, being most common in coniferous forest plantations and in drained bogs, where it colonises open patches, with low competition pressure, created by anthropogenic disturbances. The databased vegetation plots that appeared to be similar in species composition to that with C. introflexus represent mainly forest habitats broadly distributed in the Czech Republic. Cultivation experiments showed that the species has low fitness when growing on lime-enriched or waterlogged soils.

43 44 habitatscolonizedby c. introflexus in the central europe

The species invades different habitats in Western and Central Europe, but it displays an afﬁnity to similar microhabitats. The difference therefore seems to be caused largely by the absence of sandy habitats with oceanic climate in central Europe. We conclude that although the species will became common in Central Europe because of wide range of potential habitats with favourable vegetation composition and ecological parameters, it presently represents no risk for endangered plant species and communities.

keywords

Campylopus introﬂexus, invasive species, plant communities, cultivations, threat, Central Europe.

7.1 introduction

Plant invasions are considered to be one of the major threats for ecosystem diversity (Simberloff and Rejmánek, 2011). Numerous studies have summarized the impact of invasive species on native species and community structure and have identified a set of invasion-threatened habitats (e. g. Will- iamson, 1996; Chytrý et al., 2009a; Hejda et al., 2009). They point to a fact that invasive species can dramatically change current species composition or affect natural succession of plant communities. A plant species may be invasive either because it shares habitats with resident native species and outcompetes them or because it possesses habitats different from those of native species and thus occupies vacant niches (Sakai et al., 2001). Spread of invasive species after the first establishment is affected by their ability to colonise new habitats and/or their competition with native species. These factors are thus important in evaluation of actual threat for native flora. Habitat filters play an important role in plant invasions (Carranza et al., 2011; Pinke et al., 2011). Conditions under which a species invades may differ from those under which it usually occurs in its native area. In an invaded area, the range of occupied habitats increases with time after initial colonisation, depending on invader´s ability to establish in new communities (Sakai et al., 2001). This may lead to higher threat of native habitats by an invader and contribute to its successful naturalisation. The broad-scale analyses of functional plant traits, habitat characteristics and vegetation composition can be used as a good predictor for future spreading of an invader (Pyšek and Richardson, 2007; Chytrý et al., 2008). Although a large amount of studies have been conducted on explaining and predicting plant invasions, a huge majority of them are focused on vascular plants. Ecological studies on invasions of non-flowering plants, such as bryophytes, are rather rare. As compared to vascular plant alien species (ca 45 % of European flora is classified as alien either in Europe or at least in some European country), bryophyte alien species form a minor part (ca 2 %) of European bryoflora (Daisie, 2009). Bryophytes have small spores, which allow long distance dispersal across continents. Consequently, direct classification as “alien species” is sometimes problematic when based only on a disjunct occurrence. Therefore, the classification is often based on indi- 7.1 introduction 45 rect evidence as e. g. biogeography, genetics and history of records. There are 45 bryophyte species that are considered to be alien at least in some parts of Europe, including 13 cryptogenic species with unclear history of European occurrence (Söderström, 1992; Essl and Lambdon, 2009; Essl et al., 2011). Only five alien and three cryptogenic species have invaded more than five countries. The native range of these species frequently lies on southern hemisphere, but few species (e. g. Lunularia cruciata (L.) Lindb) are native in some European region and invade other European countries. The majority of the alien bryophytes have colonised regions with oceanic climates. Only two neophytes, Campylopus introflexus (Hedw.) Brid. and Orthodontium lineare Schwägr., are widespread over a large part of Europe. Campylopus introflexus has a native range in the Southern Hemisphere (South America, southern part of Africa and Australia, New Zealand; Frahm, 1984). The first finding in Europe was recorded in 1941 in Great Britain (Richards, 1963) and later in 1954 in French Bretagne (Størmer, 1958). It is also possible that the moss may have invaded Europe from several points independently (Richards, 1963). It has been introduced probably through ships sailing from the Southern Hemisphere, either on shoes of travellers or in boxes for articles covered with the moss. Then it started to spread eastward and recently it occurs in 21 countries (Hassel and Söderström, 2005; Essl and Lambdon, 2009). Ecological aspects of the Campylopus introflexus invasion to diverse habitats and its competitive potential have been investigated from the 1980s (e. g. Berg, 1985). In contrast to most alien bryophytes colonising predominantly human-disturbed habitats (gardens, roadsides, walls), C. introflexus has naturalised in near natural vegetation such as coastal dunes, bogs, and pine forests (Essl and Lambdon, 2009). C. introflexus forms dense compact cushions which could negatively affect the germination of spores and seeds of other species (Equihua and Usher, 1993). Invaded native communities could change into communities with only few mosses or lichens, with strong dominance of C. introflexus. C. introflexus characteristically forms extremely dense turfs extending over hundreds of square meters in sandy areas of the Netherlands, in which almost no other bryophyte or lichen species can survive (Van der Meulen et al., 1987; Biermann and Daniëls, 1997; Ketner-Oostra and Sýkora, 2000, 2004; Sparrius and Kooijman, 2011). The changes of vegetation composition negatively affect other groups of organisms, e. g. carabid beatles and spiders (Schirmel et al., 2011). Disturbance seems to be an important factor promoting C. introflexus invasion. However, present knowledge about the species colonisation dynamics is concentrated to West European dry sand grassland vegetation. Biermann and Daniëls (1997) recorded rapid increase of C. introflexus cover after formation of gaps in vascular-plant or lichen canopy. C. introflexus appears in initial successional phases (e. g. Hasse and Daniëls, 2006). Frequent observations showed inhibition effect of C. introflexus on the colonisation of the native species (Biermann and Daniëls, 1995, 1997). C. introflexus was reported from the Czech Republic first in 1988 from southern Bohemia (Novotný, 1990a). In the following years, the number of new localities has exponentially increased; more than 80 localities are known 46 habitatscolonizedby c. introflexus in the central europe

recently. The rapid increase of known localities after 1998 is attributed to both a faster spreading rate after the moss population became fertile and to a greater attention to this species by bryologists (Soldán, 1996, 1997; Miku- lášková, 2006). Although C. introflexus belongs to the 100 of the most invasive species in Europe (including all groups of organisms; Daisie, 2009) the knowledge of its ecology is sketchy. All research has been focused into those parts of Europe with oceanic climates (e. g. Hasse and Daniëls, 2006; Sparrius and Kooijman, 2011). Although it is known that C. introflexus colonises wide spectrum of habitats, all studies have been conducted in sandy dunes and grasslands dominated by Cladonia lichens and Corynephorus canescens. There is no information about species´ habitat and vegetation preferences. Moreover, Hasse (2007) found out that general invader hypotheses for vascular plants are only partially applicable to C. introflexus. In Central Europe, which has been colonised rather recently, the only information concerns its geographic distribution. Nevertheless, the mechanisms that influence its spatial pattern could reflect regionally specific environmental conditions, which are different in oceanic and continental climatic parts of Europe. It is therefore worth investigating possible influence and threat to native communities in inland Europe. The Czech Republic is a good example of Central European region recently colonised by C. introflexus, with a sufficient number of populations. Aims of the study are (1) to update the actual distribution pattern of C. introflexus in the Czech Republic; (2) to describe habitats colonised by C. introflexus in term of vegetation composition and conservation value (3) to characterize ecological demands of C. introflexus;(4) for the prediction of future spread, to investigate which habitats in the Czech Republic are potentially prone to C. introflexus invasion.

7.2 methods

7.2.1 Data sampling, vegetation data

Distribution of Campylopus introflexus in the Czech Republic was updated on the basis of own records, review of the literature, revisions of herbarium specimens, and verbal information of other collectors. Each locality was geo-referenced. Size (total area covered by all moss cushions in sampled vegetation type) and fertility (fertile / sterile, i. e. occurrence of sporophytes) of C. introflexus populations as well as type of habitat were recorded. In order to describe vegetation and environmental conditions of habitats colonised by C. introflexus, two datasets were used. Dataset A (AppendixA) consisted of 78 vegetation plots of 1 m2 area sampled in all different vegetation types with the occurrence of C. introflexus across the entire area of the Czech Republic. The plots were placed preferentially according to major concentration of cushions of C. introflexus in each locality. Selection of localities within the whole study area reflected the actual general knowledge about the species distribution. The aim of the sampling was to cover the complete diversity of habitats colonised by C. introflexus in the country. In each plot, all species of vascular plants, bryophytes and lichens were recorded and their 7.2 methods 47 percentage cover was estimated. Population size of C. introflexus in sampled vegetation type and its fertility status were recorded. The following environmental variables were identified in each plot in situ: total percentage cover within vegetation layers, slope inclination, aspect, and needle to broadleaf litter ratio. In 52 plots, soil samples were taken under moss tufts to the depth of 3 cm. The following parameters were determined in the samples: a ration of organic and mineral components by loss-of-ignition treatment, pH, and concentrations of NH4, NO3 and PO4 ions. The soil samples were mixed and air-dried. For determination of ratio between organic and mineral components, the samples were annealed in the cups at 300 °C for 2 hours, 400 °C for 1 h, 500 °C for 3 h, 800 °C for 1 h to constant weight. Relation between organic and mineral components was calculated from difference of soil weight before and after annealing. Measurement of pH(H2O) was performed in a soil suspension in the volume ratio soil:demineralized water 1 : 5. With the help of a mechanical shaker soil suspension was intensively stirred 5 minutes. After standing for 4 hours measurement was carried out by pH glass electrode. The ions were determined spectrophotometrically using the flow analyzer FIA-Star (Tecator, Sweden). Further, ability of soil to keep water was researched with help of metal rollers with exact volume of 100 cm3. The rollers were filled by soil directly under cushion of C. introflexus, then they were let to saturate by distilled water during 5–6 days, dried at 120 °C for 6 days and finally at 90 °C for 1 day. The rollers were weighed after satu- rating by water (W) and after drying (D). The ability of soil to keep water (Vmax) was calculated as a percentage by weight Vmax = 100 ∗ (W − D)/D. Five types of soil were distinguished on the basis of this calculation: < 5 % very weakly water retaining soil, 5–10 % weakly water retaining soil, 10–30 % medium water retaining soil, 30–50 % strongly water retaining soil, > 50 % very strongly water retaining soil. Dataset B contained 68 vegetation plots of pine forests (alliance Dicrano- Pinion), sampled in the Plzeˇnregion (West Bohemia). The sampled forests were situated around the city of Plzeˇnat a distance of 3.5–9 km (Pecháˇcková and Peksa, 2010). The plots were sampled preferentially, in order to cover entire diversity of pine forests in the area. From the total number of plots, 13 contained C. introflexus. The plot size was 15 × 15 m. The cover of all species of vascular plants, bryophytes and lichens was recorded on the nine-grade ordinal Braun-Blanquet scale (Westhoff and Van der Maarel, 1978). Potential of habitats in the Czech Republic to be invaded by C. introflexus was investigated by using dataset C. The dataset consisted of vegetation plots collected from the Czech National Phytosociological Database (Chytrý and Rafajová, 2003). All plots assigned to syntaxa at least at the level of the class and with the ground layer recorded were selected. The geographical distribution of the plots was uneven because some regions as well as some habitats were investigated more intensively than others (Chytrý and Rafajová, 2003), which may lead to overestimation of geographical over ecological variation. Geographical stratification of the dataset was therefore performed to obtain geographically balanced distribution of habitats. Only one plot from each vegetation unit was selected from a grid cell of 1.25’ longitude and 0.75’ latitude (approximately 1.5 × 1.4 km). Recently recorded plots 48 habitatscolonizedby c. introflexus in the central europe

were preferred in the selection. The stratified dataset C contained 26 998 vegetation-plot records. Besides field studies, a cultivation experiment was established during summer 2005 in order to test the ability of C. introflexus to grow in different soil and moisture conditions. A total number of 20 populations sampled in different localities were used for the cultivation. Small moss cushions (10 gametophores) from each population were cultivated in standard laboratory conditions (12h light–dark cycle; constant room temperature 18 °C, 60 % moisture) during one year in rounded ceramic pots (9 cm in diameter). The number of living gametophores was counted at the end of experiment. Different soil conditions were simulated by seven types of substrate: sand (organic components max 1 %), commercial substrate (Peaty-soil substrate for plants, seedlings and others; Jiránek company, Beˇcicen. Jizerou, Reg. nr. 0427/2000, organic components min 25 %, components above 20 mm up to 5 %, moisture up to 65 %), sandy soil (mixture of sand and commercial substrate 1 : 1, organic components min 10 %), peaty soil (mixture of peat and commercial substrate 1 : 1, organic components min 50 %), peat (Garden peat without additives; Jiránek company, Beˇcicen. Jizerou, Reg. nr. 0738/2000, organic components min 70 %, components above 10 mm up to 5 %, moisture 45–65 %), spruce litter and commercial substrate enriched with crushed limestone. The effect of different soil moisture on growth of C. introflexus was tested independently of the experiment with different soil substrates. Plants were cultivated on the commercial substrate and three levels of watering were used: permanently wet (watering twice a week), medium-wet (watering once a week) and dry (watering once in three weeks).

7.2.2 Data analyses

All vegetation datasets were imported into the software JUICE (Tichý, 2002). Different species abundance scales were transformed into percentages. Rec- ords of plants not determined at the species level were deleted. Seedlings of trees and high shrubs in herb layer were also deleted because they had not been recorded in all plots. Multiple records of species in different layers were combined so that all species appeared in respective dataset only once. Records of critical taxa in sense of determination were combined. This was the case for Betula spec. div., Empetrum nigrum s. lat. (E. nigrum, E. hermaphrodi- tum), Molinia caerulea s. lat. (M. caerulea, M. arundinacea), Oxycoccus palustris s. lat. (O. palustris, O. microcarpus), Sphagnum recurvum s. lat. (S. angustifolium, S. brevifolium, S. fallax, S. flexuosum), and Hypnum cupressiforme s. lat. (H. cupressiforme, H. jutlandicum). Nomenclature of taxa follows Kubát et al. (2002) for vascular plants, Kuˇceraand Váˇna(2005) for bryophytes and Vˇezdaand Liška (1999) for lichens. Vegetation plots of the datasets A and B were classified separately to plant communities on the basis of species composition. In both cases, the TWINSPAN algorithm (Hill, 1979) with modified stopping rules according to cluster heterogeneity (Roleˇceket al., 2009) was used, with pseudospecies cut levels set at 0, 5, and 25 % of species cover and Whittaker’s beta diversity as a dissimilarity measure for assessing cluster heterogeneity. C. introflexus 7.2 methods 49 was deleted from the datasets prior to the analyses. The plant communities were described with the help of diagnostic, constant and dominant species. Diagnostic species were determined as the species with a high fidelity to a given community with the phi coefficient as the fidelity measure (Chytrý et al., 2002). For the calculation of the phi coefficient, numbers of plots in the communities were standardized to an equal size (Tichý and Chytrý, 2006). The species with phi > 0.2 were considered diagnostic. Constant species were defined as the species with frequency of occurrence in a given community > 50 % and dominant species as the species occurring in at least 10 % of plots of a given community with a cover > 25 %. Variability of vegetation with C. introflexus and its relationships to environmental variables was expressed on 52 samples from the dataset A with all soil parameters determined by detrended correspondence analysis (DCA), using CANOCO 4.5 package (Ter Braak and Šmilauer, 2002). The percentage frequency of species was log-transformed and rare species were down- weighted. Pine forest plant communities (dataset B) were ecologically interpreted by means of Ellenberg indicator values (EIV) (Ellenberg et al., 1992). Ellenberg indicator values were developed for Central European vascular plant species. Based on field observations, they reflect ecological behaviour of particular species with respect to light, moisture, nutrients, soil reaction, temperature and continentality. Unweighted averages of EIV were calculated for each vegetation plot in the program JUICE. As different growth abilities and responses to substrate and/or watering level might have been expected for individual moss populations, the cultivation experiment was analysed by means of linear mixed effects models, following the protocol of Zuur et al. (2009) page 90–92, with population included as a random factor to account for various population means, slopes and/or variances. First, several models with the most complex fixed effect part (experiment treatment factor) and different random effects were built: (0) linear model without random effects, (1) model with random intercept (allowing for different means of gametophores per population – localities), (2) model with random intercept and random slope (allowing for different means and response rates per population), (3) model with allowance for different variances within watering / soil groups, (4) model with random intercept and allowance for different variances within watering / soil groups, (5) model with random intercept and random slope and allowance for different variances within watering / soil groups. The parameters of random effects were estimated by REML (Restricted maximum likelihood approach) and the combination of random effects resulting in the lowest model AIC (Akaike’s information criterion) was selected (Table 7.1). Afterwards, the fixed part of the models was simplified following data exploration: several new models were built with the treatment levels with the most similar growth abilities grouped together, while the random effects were kept the same (those selected in the previous step). The models were fit with ML (Maximum Likeli- hood) during the fixed part simplification, and model AICs were calculated and compared. The most parsimonious model was refit with REML and considered the final. The R program, version 2.12.0 (R Development Core Team) 50 habitatscolonizedby c. introflexus in the central europe

and the ’nlme’ library (Pinheiro et al., 2011) were used for the mixed effects models construction and testing. Invasibility of habitats in the Czech Republic by C. introflexus was assessed by calculation of similarity in species composition between vegetation plots of all habitats in the Czech Republic (dataset C) and six groups of plots with recorded occurrence of C. introflexus representing all invaded habitats in the area. From them, four groups originated by the classification of plots of the dataset A, one group was composed of 13 plots of pine forests from the dataset B, and one group of 16 plots of drained peatbog vegetation with the occurrence of C. introflexus (Konvalinková, 2010). All records of lichens were deleted for this analysis because they were recorded only in some of the plots from the dataset C. The similarity was calculated by Frequency In- dex within function Matching in the program JUICE. The Frequency Index (FQI) is defined as a sum of frequencies of species i (FQi) from assessed vegetation plot in a given vegetation unit (group of plots) divided by sum of frequencies over all species in the vegetation unit (Formulae 7.1). Species present in the assessed plot are indicated as i–R and species present in the vegetation unit as i–C (Tichý, 2005). Vegetation plots with similarity > 40 % to any of the groups were considered as similar in their species composition to the vegetation where C. introflexus does occur, so that they represent vegetation potentially endangered by its invasion. The similar plots were grouped to the phytosociological classes according to their original assignment and the groups were characterized by species with frequency of occurrence > 50 %. Finally, we created a map of potentially suitable habitats in which we mapped both the locations of the vegetation-plot records that appeared to be similar to already invaded vegetation and the distribution of habitat that corresponds to corresponding phytosociological units. The distribution of particular habitats has been taken manually from the Habi- tat Catalogue of the Czech Republic (Chytrý et al., 2010). Nomenclature of sytaxonomical units follows Moravec (1995, 1998, 2000, 2002).

 FQi  i∈R FQI = 100 ×   (7.1) PFQi i∈C P

7.3 results

7.3.1 Distribution of Campylopus introﬂexus in the Czech Republic

C. introﬂexus has been recorded in the entire area of the Czech Republic; more than one hundred herbarium specimens and published as well as unpublished records have been collected to date (Fig. 7.1). Its distribution cen- ter lies in the West and South Bohemia (western part of the Czech Republic), only 12 localities have been found in Moravia (eastern part of the Czech Republic). Its distribution is concentrated in area of temperate deciduous forests, ranging from hilly country to montane belt (280–1140 m a. s. l.). Most populations in the Czech Republic are sterile, but the reports about occurrence of sporophytes are recently increasing (about 30 fertile populations 7.3 results 51

Table 7.1: Akaike’s information criterion (AIC) values of the linear mixed effects models for the cultivation experiment of C. introﬂexus in different soil and moisture conditions. The AIC values were used to select the most parsimonious random components (in bold), while the model ﬁxed part was kept constant. The degrees of freedom – df used for parameter estimation. For random components’ types (0)–(5), see Methods 7.2.

Random component type Water supply Soil type df AIC df AIC (0) no random effect 4 348 8 753 (1) random intercept 5 337 9 743 (2) random intercept and slope 10 323 36 749 (3) different variances 6 337 14 757 (4) r. intercept + diff. var. 7 333 15 738 (5) r. intercept and slope + diff. var. 12 327 42 765

in 2011). Vegetative propagation by deciduous stem apices is common. Size of populations varies from few plants to cushions of more than 100 m2, but smaller discontinuous tufts up to few square metres are more frequent. The most frequent habitats of C. introflexus in the Czech Republic are edges of spruce plantations and pine forests with the young outplanting, forest clearings and bare parts of damaged peat bogs. All these habitats are influenced by human disturbance. C. introflexus grows on bare soil or in gaps created by animals or by man. Field observations show low resistance of the moss to permanently waterlogged substrate; the species avoids permanently wet bogs or damaged bogs with restored water regime.

7.3.2 Vegetation composition of C. introﬂexus habitats

On the basis of the Twinspan classification, four plots from the dataset A were excluded from analyses because of outlying species composition. These plots originated from natural mixed beech forest (one plot), waterlogged forest clearing (two plots) and brook bank (one plot). The classification of the other 74 plots of the dataset A distinguished four plant communities with the occurrence of C. introflexus (Table 7.2). Group 1 associates microhabitats in pine forests and peat bogs, which are characterized by pine in tree layer, Vaccinium myrtillus in herb layer and several species of the genus Cladonia in ground layer. This group also includes two plots from oak forest and one population from spoil tip. Group 2 is composed of plots from very different habitats. They include edges of spruce forests, peat bogs with birch, drained peat bogs, fishpond banks, spruce- beech and birch-pine forests. Plots of the group 2 usually contain birch in tree layer, C. introflexus and Pohlia nutans in ground layer and higher percentage of broadleaf litter. Group 3 represents spruce forests, mainly forestry plantations. As compared to other groups, this habitat is the most strongly 52 habitatscolonizedby c. introflexus in the central europe

Figure 7.1: Distribution of habitats which are suitable for the Campylopus introflexus and species present distribution in Czech Republic in grid cell of 3’ longitude and 7’ latitude. Hatched squares show the distribution of vegetation plots from the Czech National Phytosociological Database, which appeared to be similar with the plots already containing C. introflexus. Grey squares show the complete distribution of the most often invaded habitats and habitats with similar species composition (see Results 7.3 for the details). Black dot – cell with vegetation plot sampled (Dataset A), empty dot – cell without vegetation plot sampled, but with recorded occurrence of C. introflexus. 7.3 results 53 shaded by tree layer (Picea abies), and the herb layer is not well developed. Only a few bryophytes and lichens with low cover occur, and the major part of land is bare with needle litter. Group 4 associates vegetation developed under very open tree canopy composed of pine, spruce and/or birch and higher competition pressure of other bryophytes and lichens. This vegetation was usually found in rather dry spruce-forest edges and trails, but occurred also along forest trails, in forest rides and clearings and at bog edges. Species associated with C. introflexus in the vegetation plots of the dataset A (Table 7.2) belong to common elements of European flora with continental climate. The species are widespread in temperate-boreal forests, with wide realised ecological niche. C. introflexus was rarely found in other habitats, as indicated by only four outliers. They were sampled in (i) beech forest, where the species grew close to the edge of forest clearing, (ii) moist clearings in a spruce forest and (iii) brook bank, where the species was associated with hygrophilous bryophyte species. In addition (occurrence without sampled plot), the species was once recorded in boulder scree with relict flora and once on fallen trunk in ad- vanced stage of decay. Despite seeming heterogeneity of all these habitats, one common feature can be found — in all cases the species grew on bare soil with low competitive pressure of vascular plants. The vegetation samples from pine forests (dataset B) were classified into five well-distinguished plant communities (Table 7.3). They are differentiated by the composition of tree layer and by different nutrient availability and soil moisture: (1) nutrient-rich forests with oak, without Vaccinium myrtillus,(2) humid forests with oak and Vaccinium myrtillus,(3) extremely- species-poor dry lichen forests (4) dry lichen forests, (5) pine forests with spruce. The first two types of pine forests do not contain plots with C. introflexus, while this species is present in some plots of the other three communities.

7.3.3 Ecological characteristic of the habitats – ﬁeld data

The habitats where C. introflexus was found (dataset A) have usually low cover of the tree layer. Cover of other layers is similar across all plant communities. C. introflexus grows generally on acid substrate, or on alkaline substrate if at least the small amount of spruce or pine needle litter is present (Table 7.2). There is no difference in pH range among plant communities. Needle litter prevails in 75 % of the localities, the ratio of needle:broadleaf litter is more equal in mixed forests with birch. Mean values of PO4 and NH4 ion contents are only slightly different among vegetation types derived from the dataset A (Table 7.2), but the range of values is high. Amount of NH4 ions in soil under cushions varies from 20.5 to 248.02 mg/kg, amount of PO4 ions varies from 4.4 to 12.7 mg/kg, amount of NO3 ions (data not shown) varies from 0.56–3.87 mg/kg. Proportion of soil organic and mineral components varies in all communities, and loss-of-ignition ranges between 2 % (pine forests) and 98 % (bog communities). C. introflexus grows on medium up to strongly water-retaining soils, but it is able to grow on very slightly 54 habitatscolonizedby c. introflexus in the central europe

Table 7.2: Characteristics of habitats with C. introflexus (dataset A) classified by modified Twinspan analysis. Threshold value of phi coefficient for diagnostic species was 0.2 (0.5), minimum frequency of occurrence for constant species was 50 %(80 %), for dominant species 10 % with cover at least 25 %. E3 tree layer, E1 herb layer, E0 ground layer; all values are averages for group with (±) standard deviation. 1 1 − − 2 . 0 mg kg ± mg kg 7 % % . 7 . % 26 5 % . 3 58 6 20 26 2 245 ± 30 O ± ± ± ± ± 2 ± 20 1 2 ± 3 . : . 35 80 3 8 81 52 1 0 E Pohlia nutans E E Avenella flexuosa Avenella flexuosa Cladonia fimbriata Cladonia fimbriata ions ions 4 Dicranella heteromalla soil pH-H 4 Polytrichastrum formosum Polytrichastrum formosum PO NH Cladonia fimbriata, Avenella flexuosa, 1 1 − − 2 . s. lat. 0 mg kg ± mg kg 9 % % % . 7 . 8 23 80 . % 3 46 19 24 23 1 227 670 ± 26 O ± ± ± ± ± 2 ± 15 7 ± 4 . . : 30 22 74 7 80 53 3 1 0 Picea abies Picea abies Picea abies E E E Avenella flexuosa Avenella flexuosa Cladonia fimbriata ions ions Cladonia fimbriata 4 soil pH-H 4 Hypnum cupressiforme PO NH Group , 1 1 − − 4 . 0 ± mg kg mg kg % % % 7 7 . . 7 40 85 . % 3 21 30 29 1 15 238 608 ± 30 O ± ± ± ± ± 2 spec. div. ± spec. div. 40 ± 4 4 . : . 12 35 60 7 45 3 1 0 56 E E E Pohlia nutans Pohlia nutans Pohlia nutans Avenella flexuosa Betula Betula ions ions 4 soil pH-H 4 Polytrichastrum formosum Polytrichastrum formosum PO NH 1 1 − − 2 . 0 mg kg ± mg kg 3 % % . 7 % . 1 36 60 . % 3 21 22 22 2 232 670 12 ± 35 O ± ± ± ± 2 ± ± 25 16 12 21 20 1 ± 7 . 8 . : 30 70 6 3 51 480 55 1 0 80 E E E Calluna vulgaris Luzula luzuloides Cladonia pyxidata Pinus sylvestris ions ions Calamagrostis villosa 4 soil pH-H 4 pine forests and peatbogs mixed forests with birch spruce forests forest edges and paths PO NH Soil layers in soil species species species Needle : Altitude Cover of burnable Constant leaf litter [ m a. s. l.] Dominant vegetation Diagnostic Number of Contents of populations components characteristic 7.3 results 55

Table 7.3: Characteristics of pine forests (dataset B) by diagnostic, constant and dominant species. Threshold value of phi coefﬁcient for diagnostic species was 0.2 (0.5), minimum frequency of occurrence for constant species was 50 %(80 %), for dominant species 10 % with cover at least 25 %.

Diagnostic species Constant species Dominant species Poa nemoralis Pinus sylvestris Pinus sylvestris Brachythecium rutabulum Avenella flexuosa Avenella flexuosa Plagiomnium affine Plagiomnium affine Quercus petraea agg. Quercus petraea agg. Brachythecium rutabulum Pleurozium schreberi Impatiens parviflora Quercus Calamagrostis epigejos Arrhenaterum elatius petraea agg. Brachythecium rutabulum Galeopsis speciosa Hypnum cupressiforme s. lat. Sorbus aucuparia Dicranum scoparium Hieracium laevigatum Poa nemoralis nutrient-rich pineforest Mycelis muralis Sorbus aucuparia Moehringia trinervia Quercus robur Quercus rubra Pleurozium schreberi Festuca rubra Calamagrostis epigejos Calamagrostis epigejos Betula spec. div. Quercus robur Festuca ovina

Rubus idaeus Vaccinium myrtillus Pinus sylvestris Luzula pilosa Pinus sylvestris Avenella flexuosa Frangula alnus Avenella flexuosa Vaccinium myrtillus Plagiothecium suculentum Pleurozium schreberi Pleurozium schreberi Molinia caerulea s. lat. Hypnum cupressiforme s. lat. Hypnum cupressiforme s. lat. Dryopteris carthusiana Vaccinium vitis-idaea Plagiothecium suculentum Lysimachia vulgaris Rubus idaeus Molinia caerulea s. lat. Eurhynchium angustirete Plagiomnium affine Betula spec. div. humid pineforest Calliergon cordifolium Dicranum scoparium Melampyrum pratense Betula spec. div. Vaccinium myrtillus Polytrichastrum formosum Carex pilulifera Pohlia nutans Plagiomnium affine Melampyrum pratense Hieracium lachenalii Frangula alnus Dryopteris carthusiana Aulacomnium androgynum Pinus sylvestris Pinus sylvestris Cladonia furcata Hypnum cupressiforme s. lat. Dicranum scoparium Ptilidium ciliare Dicranum scoparium Pleurozium schreberi Hieracium pilosella Avenella flexuosa Festuca ovina Pleurozium schreberi Rumex acetosella Pohlia nutans Betula spec. div. Ptilidium ciliare Cladonia furcata lichen pineforest, species poor

Cladonia arbuscula Pinus sylvestris Pinus sylvestris Cladonia gracilis Dicranum scoparium Pleurozium schreberi Cetraria islandica Pleurozium schreberi Cladonia macilenta Vaccinium vitis-idaea Leucobryum glaucum Vaccinium myrtillus Dicranum spurium Hypnum cupressiforme s. lat. Cladonia uncialis Avenella ﬂexuosa Cladonia squamosa Pohlia nutans lichen pineforest Vaccinium vitis-idaea Calluna vulgaris Cladonia rangiferina Leucobryum glaucum Calluna vulgaris Dicranum polysetum Cladonia furcata

Picea abies Pinus sylvestris Pinus sylvestris Hylocomium splendens Avenella flexuosa Pleurozium schreberi Cladonia ochrochlora Pohlia nutans Vaccinium myrtillus Pohlia nutans Pleurozium schreberi Avenella flexuosa Melampyrum pratense Hypnum cupressiforme s. lat. Campylopus introflexus Dicranum scoparium Polytrichastrum formosum Picea abies

spruce pineforest Vaccinium myrtillus Betula spec. div. Vaccinium vitis-idaea Dicranum polysetum Calluna vulgaris Polytrichastrum formosum 56 habitatscolonizedby c. introflexus in the central europe

Figure 7.2: DCA ordination diagram of 52 vegetation plots of dataset A and environmental variables passively projected onto the ordination. The first and second ordination axes are displayed. The groups of plots were classified by modified Twinspan analysis. Identification of groups: (1) pine forests and bog edges, (2) mixed forests with birch, (3) spruce forests, (4) spruce forest edges and trails.

water-retaining soil with maximum water content 10 %. Slope inclination is low in all localities, with one exception of boulder scree in the Krkonoše Mts, with no preferences to slope orientation (data not shown). Mean altitude for particular vegetation types derived from dataset A reflects the distribution of respective plant communities in Czech Republic; pine forests and drained peatbogs prevail in lower altitudes (Table 7.2). In the ordination space, the vegetation groups are well distinguished along the first two ordination axes (Fig. 7.2). Distribution of the plots along the first ordination axis correlates with altitude, while distribution along the second ordination axis correlates with concentration of PO4 and NH4 ions, pH and character of litter (needles versus leaves). Cover, population size and fertility of Campylopus introflexus decreases along the first axis, from lowland pine forests and bog edges to higher-altitude spruce forests. Ellenberg indicator values indicate ecological differences among particular types of pine forests (dataset B). The first two groups differ from the others by higher both soil reaction (Fig. 7.3a) and nutrients (Fig. 7.3b). No apparent differences were found for the other EIV parameters (data not shown). C. introflexus was not recorded in any plot of the first two groups, therefore it seems that soil reaction and amount of nutrients in soil can be limiting environmental factors for C. introflexus in pine forests. 7.3 results 57

Figure 7.3: Average Ellenberg indicator values for nutrients and soil reaction for pine forest plant communities (dataset B). The higher EIV, the higher soil reaction (from acidic to basic) and nutrients demands of plant species in a community. For both parameters, the values range from one to nine. Identiﬁcation of groups: (1) nutrient-rich forests with oak, without Vac- cinium myrtillus,(2) wet forests with oak and Vaccinium myrtillus,(3) dry lichen forests, species-poor (4) dry lichen forests, (5) pine forests with spruce.

7.3.4 Ecological characteristic of the habitats – cultivation experiment

The results of the cultivation experiment confirmed the results of the field studies. Protonema started to grow after two weeks from establishment of experiment on major part of soil types and levels of watering tested. Protonema covered the whole surface of cultivation substrates during the first month. Gametophores started to grow during the first and second month, and covered the whole surface of pots after a few months. On sand, sandy soil and dry treatment protonema started to grow during the second month, with new gametophores growing adjacently to initial cushion. On the lime- enriched substrate the protonema growth was inhibited and protonema was formed mainly on initial plants. Most gametophores died on lime-enriched soil within six months; when gametophores grew, they were high, weak, and had a few small leaves. Under permanently wet treatment, the protonema started to grow during the first month but the young gametophores died soon. Statistically significant effects of soil type and moisture regime on the viability and fitness of C. introflexus were observed. For testing the effect of soil type, the random component (4) of the linear mixed effects model was the most parsimonious (Table 7.1). Four soil types with similar growth rates (commercial substrate for plants, peaty soil, peat, spruce litter) could be grouped together into an “organic soil group” without any loss of information, as judged by AIC. However, any further groupings led to a considerably higher model AIC, suggesting different growth abilities of the moss in the remaining treatment groups (the final model F = 100.54, p < 0.0001, numerator DF = 3, and denominator DF = 99). The mean number of developed gametophores decreased from organic soils, through sandy soil and sand to the lime-enriched soil (Table 7.4, Fig. 7.4a). Gametophore production was 58 habitatscolonizedby c. introflexus in the central europe

Table 7.4: Coefﬁcients of the linear mixed effects models showing the estimated mean, its standard error (SE) and relative variance of the number of gametophores per treatment groups. Different variances per treatment groups were allowed in the soil experiment only. Organic soil group represents the four soil treatments grouped together during the analysis: commercial substrate for plants, peaty soil, peat, and spruce litter.

Estimated mean number of Experiment Treatment gametophore (SE) Variance estimates organic soil group 26.279412 (0.9395467) 0.5378913 sandy soil 21.647059 (1.2224823) 0.4898299 soil type sand 11.529412 (2.0660512) 1.0000000 lime-enriched soil 4.470588 (1.4617967) 0.6449329 permanently wet 1.80 (0.489898) - water supply medium-wet 25.45 (1.170863) - dry soil 19.95 (1.134332) -

5 times lower on the soil mixed with limestone compared to that on organic soils, evidencing that lime enrichment considerably harmed C. introflexus. For testing the effect of watering, the random component (2) was chosen (Table 7.1). Any treatment groupings caused a considerable increase of AIC, suggesting significant differences in the moss growth abilities among the treatment groups (the final model F = 220.87, p < 0.0001, numerator DF = 2, and denominator DF = 38). Fewer gametophores developed on permanently wet soil (~10 and 13 times, respectively) than developed on medium-wet and dry soils (Table 7.4, Fig. 7.4b), evidencing that very wet conditions considerably harmed C. introflexus. On the other hand, the number of gametophores on the medium-wet soil was slightly higher compared to the dry soil, imply- ing that C. introflexus benefited from moderate moisture.

7.3.5 Invasibility of vegetation by C. introﬂexus

From the total number of 26 998 vegetation plots in the dataset C, 938 were chosen as ﬂoristically similar based on Frequency Index to the vegetation invaded by C. introﬂexus (Fig. 7.1). They were originally assigned to nine vegetation classes (Table 7.5). Major part of the plots belongs to Querco- Fagetea, Vaccinio-Piceetea and Quercetea robori-petrae. These classes also have the highest proportion of chosen plots to all plots from the respective class in the dataset C. The selected plots were assigned mostly to alliances Dicrano- Pinion, Piceion excelsae, Genisto germanicae-Quercion, Luzulo-Fagion and Fagion. Except for forest clearings of Epilobietea angustifolii, non-forest vegetation is represented rather marginally. Distribution of selected plots as well as complete distribution of corresponding habitats is shown in Fig. 7.1. The following habitats were mapped based on data in the Habitat Catalogue (Chytrý et al., 2010): L5.4. Acidophilous beech forests; 7.4 discussion 59

Figure 7.4: The effect of soil type and watering on the number of gametophores. Level of water supply – dry, medium-wet, permanently wet. Types of substrate - organic soil group = commercial substrate for plants + peaty soil + peat + spruce litter, sandy soil, sand, lime-enriched soil.

L7 Acidophilous oak-forests; L8 Dry pine forests; L9.1. Montane Cala- magrostis spruce forests, T5 Sand and shallow soil grasslands (except for the unit T5.5 Acidophilous grasslands on shallow soils, which represents widespread, but small-scale open roadside and forest-margin siliceous grasslands on ranker soils, but where Campylopus introflexus was never observed) and R3.4. Degraded raised bogs. Species with the highest frequency of occurrence in chosen vegetation plots of respective classes (Table 7.5) characterize vegetation similar to the vegetation invaded by C. introflexus. All of these species are widespread and common in central Europe. Picea abies, Avenella flexuosa, Vaccinium myrtillus, Polytrichastrum formosum, and Dicranum scoparium were recorded in more than 75 % of all selected plots (Table 7.5).

7.4 discussion

7.4.1 Distribution pattern in the Czech Republic

Colonisation rate of an invasive species is regulated by production, transport and establishment of diaspores. C. introﬂexus reproduces both sexually and asexually in the Czech Republic and in the western part, the concentration of populations is higher. The colonisation of suitable habitats in the western part is higher than in the eastern part because of higher propagule pressure. The uncolonised area and distances between populations increase eastward even though there are still a high number of potentially suitable habitats in the east (Fig. 7.1). Taking into account that the species is spreading from Western Europe eastwards, the pattern suggests that the long distance spreading by spores is more important for initial colonisation. For subsequent spreading, importance of asexual reproduction increases. This colonisation pattern characterises initial phase of the invasion when colonisation of new areas is mainly dispersal-limited, in contrast to already widely dis- 60 habitatscolonizedby c. introflexus in the central europe

Table 7.5: Vegetation types that have similar species composition as vegetation already invaded by C. introflexus. The similarity was calculated by Fre- quency Index within function Matching in the program JUICE, using stratified Czech National Phytosociological Database. Minimum frequency of occurrence for constant species was 50 %(80 %). ; %, %, 47 29 , Vaccinium , Festuca ovina, Vaccinium , Vaccinium Polytrichastrum Polytrichastrum Calamagrostis , Plagiothecium laetum, . s. lat. %, , Ptilidium pulcherrimum Vaccinium myrtillus 35 Oxalis acetosella Fagus sylvatica s. lat , Dryopteris dilatata, Hieracium %, %, 76 30 Vaccinium myrtillus , Vaccinium vitis-idaea, Dicranum , Dicranum scoparium, Pleurozium , Avenella flexuosa ., Pleurozium schreberi, Pohlia nutans, agg. Pinus sylvestris s. lat , Oxycoccus palustris %, 36 Luzula luzuloides Polytrichum formosum Dryopteris dilatata, Dryopteris filix-mas, Geranium s. lat. , Eriophorum angustifolium, Eriophorum vaginatum, Dicranum scoparium %, 30 %, Calamagrostis villosa, Dryopteris dilatata, Epilobium Quercus petraea , Calamagrostis villosa, Dryopteris dilatata, Epilobium , Eriophorum vaginatum, s. lat. 80 Picea abies, Avenella flexuosa Pohlia nutans % %, , Pohlia nutans, , Pohlia nutans 27 40 Avenella flexuosa, agg. Sorbus aucuparia s. lat. Dicranum scoparium, Hypnum cupressiforme Vaccinium vitis-idaea, Calluna vulgaris Avenella flexuosa Avenella flexuosa, Avenella flexuosa %, Vaccinium vitis-idaea, Calluna vulgaris, Dicranum polysetum, Dicranum scoparium, s. lat. 33 Polytrichastrum formosum Polytrichum formosum , %, myrtillus, Pinus mugo, Picea abies, Avenellarostrata, flexuosa, Empetrum Calamagrostis nigrum villosa, Carex nigra,Juncus Carex filiformis, Molinia caerulea Picea abies; Avenella flexuosa, Vacciniumscoparium, myrtillus Polytrichastrum formosum Betula species, Pinus sylvestris, Avenella flexuosa, Melampyrum pratense, Nardus stricta, Potentilla erecta, myrtillus, Vaccinium vitis-idaea, Drepanocladus fluitans,nutans, Pleurozium Polytrichum schreberi, commune, Pohlia Polytrichum strictum,recurvum Sphagnum majus, Sphagnum Abies alba, Fagus sylvatica, Picea abies, angustifolium, Rubus idaeus, Vaccinium myrtillus,formosum Dicranum scoparium, Picea abies, angustifolium, Rubus idaeus, Vaccinium myrtillus,formosum Dicranum scoparium, Picea abies, uliginosum, robertianum, murorum, Oxalis acetosella, Vaccinium myrtillus, Dicranums. scoparium, lat. Hypnum cupressiforme Sorbus aucuparia, schreberi, Pohlia nutans, Polytrichum commune,s. Sphagnum lat. magellanicum, Sphagnum recurvum 82 Polytrichastrum formosum Pleurozium schreberi Pleurozium schreberi Hieracium murorum, Luzula luzuloides, MelampyrumDicranum pratense, scoparium, Hypnum cupressiforme Senecio nemorensis Rubus idaeus %, %, Hypnum cupressiforme 27 33 Picea abies %, %, 42 85 Habitat Constant species in similar samples serpentines forests vegetation peatbogs Vaccinium vitis-idaea Hieracium murorum Dryopteris dilatata %, %, %, 27 Vaccinium myrtillus 34 % other peatbogs %% pine and spruce forests acid oak forests % beech forests and oak % forest clearings 45 % heaths % pine% forests on high/low raised % mountain highherbal %, 37 20 30 46 . . . 12 067 41 95 15 90 . . . . ,. . Percentage of similar samples Abies alba %, 34 Avenella flexuosa All samples in stratified database villosa Maianthemum bifolium Pleurozium schreberi Similar samples 34893 984 355 35 26 91 814 1448 1 0 428 41572 10 14 27 354 7 3 35 469 7 8 254 3 Vegetation class Vaccinio- Piceetea Quercetea robori- Nardo- Callunetea Scheuchzerio fuscae Species with the highest selected samples from the database: Querco- Fagetea Erico- Pinetea Oxycocco- Sphagnetea Epilobietea angustifolii Mulgedio- petrae percentage frequency of Aconitetea -Caricetea 7.4 discussion 61 tributed species with habitat limitation (Mulligan and Gignac, 2001). In contrast, Orthodontium lineare, the second strong invasive moss species in Cen- tral Europe, seems to be habitat-limited (Herben, 1994) rather than dispersal- limited (Hedenäs et al., 1989a,b). Taking into account the difference in dispersal ability among different taxonomic groups of organisms, with bryophytes being better dispersers than vascular plants or some animals (Hájek et al., 2011), we can expect that the first phase of the C. introflexus invasion will be rather short, and that the distribution of the species will become habitat- limited.

7.4.2 Recently invaded habitats and their ecological characteristics

Using both the large-scale field study from the Czech Republic and the cultivation experiment we demonstrated the great affinity of C. introflexus to moderately wet acidic soils, with a high proportion of organic matter (peat, forest humus). When the species occurs at calcium-rich bedrock, an at least shallow acidic litter or peat horizon must be developed (Schlüsslmayr, 2005; Hasse and Daniëls, 2006). At the habitat level, we found that C. introflexus occurs frequently in coniferous forests without a shrub layer, but with abundant dwarf shrubs, with low cover of herb and ground layer. Such coniferous forests are represented predominantly by spruce and pine plantations at low altitudes, but natural pine forests are invaded as well. Fre- quent co-occurrence of C. introflexus and pine is possible due to more open tree layers in pine forests, compared to forests dominated by other conifers or broadleaved trees. Another reason could be frequent occurrence of pine in initial stages of succession, which are preferred by C. introflexus as well (Klinck, 2008). However, we found that some pine forests are not invaded, and they either have higher soil reaction and nutrient availability, or they are waterlogged (e. g. bog woodlands with undisturbed water regime). These observations fit the results of our cultivation experiment evidencing that either the presence of lime or strong waterlogging limits the C. introflexus fitness, as well as the results of Zerbe and Wirth (2006) who found that C. introflexus invaded predominantly dry, nutrient poor and strongly acid lichen pine forests. Besides natural pine forests and forestry plantations C. introflexus colonised floristically similar drier edges of peat bogs and abandoned drained peat bogs. Equihua and Usher (1993) found that in drained bogs, C. introflexus can inhibit the natural succession of native bryophytes and vascular plants. Nevertheless, this risk is relevant only in drained bogs or in drier bog edges at the transition into the spruce forests or Nardus grasslands, because in well-hydrated parts with restored water regime, C. introflexus is limited by waterlogging.

7.4.3 Differences in habitat afﬁliation between the Czech Republic and Western Europe

The habitats which are invaded by Campylopus introﬂexus in the Czech Re- public correspond neither to habitats which are usually invaded by other 62 habitatscolonizedby c. introflexus in the central europe

species in Central Europe (Chytrý et al., 2005, 2008; Hejda et al., 2009) nor to native habitats of the invader (Frahm, 2007). They also differ from the habitats which are frequently invaded by Campylopus introflexus in the northern and western Europe, such as open sand dunes and sandy grasslands (Biermann and Daniëls, 1997, 2001; Ketner-Oostra and Sýkora, 2000, 2004; Hasse, 2007), where Campylopus introflexus frequently forms large cushions and represents a risk for native flora (Equihua and Usher, 1993; Klinck, 2008; Sparrius and Kooijman, 2011). In contrast, we demonstrated that in Central Europe the species prefers rather nutrient-poor wooded habitats, or drained bogs, and when cultivated it shows the worst fitness on sand and sandy soils. However, this seemingly different ecological behaviour is probably not caused by changes in genome that would allow the extension of the spectrum of invaded habitats (Bossdorf et al., 2005). Well-dispersing bryophytes often have uniform genetic structure (e. g. Karlin et al., 2011), as was preliminary confirmed also for C. introflexus by Stech and Dohrmann (2004), Stech and Wagner (2005) and Mikulášková et al. (unpublished data). In addition, differences between Central and Western Europe are smaller at the microhabitat scale. The species generally prefers bare soil or the dieback of grass turfs, heaths or moss carpets (see also Biermann and Daniëls, 2001; Piessens et al., 2008), and in West-European open sands the species prefers patches with higher amounts of organic carbon (Sparrius and Kooijman, 2011), con- forming to the macroscale ecological behaviour in Central Europe. One possible explanation of the observed difference between Western and Central Europe is simply the absence of coastal sand dunes in inland Europe. Al- though sandy habitats occur in the Czech Republic as well, they are ecologically different because of climatic reasons. Being located more inland, they experience more frost in winter and less moisture in summer. As most of the native range of the species is frost-free, C. introflexus is believed to be rather frost-intolerant, but experimental evidence is missing. In an observation study of Chiarucci et al. (2008), the species is reported from acidic geothermal fields. Concerning moisture, our cultivation experiments suggest that the species can survive long-term desiccation, but on extremely dry sands it grows worse than on organic soils. The data from Western Eu- rope (Sparrius and Kooijman, 2011) further suggest better survival under enhanced ammonium concentrations. Hence, in Central Europe the species finds suitable habitat conditions in habitats that are non-calcareous, acidic, moderately moist, rather ammonium-rich and not experiencing black frosts. Because such habitats are usually overgrown by dense growths of vascular plants or competitively strong bryophytes (such as Sphagnum in nitrogen- polluted bogs, Hájková et al., 2011), C. introflexus occupies disturbed patches, which are common in drained bogs or coniferous-tree plantations. The success of C. introflexus in Central Europe thus cannot be explained by strong competitive ability as in the case of some vascular plant invaders. Instead, the disturbance-mediated competition (Schooler et al., 2010) or, purely an ability to colonise quickly bare acidic soil after anthropogenic disturbances, explain the recent spread of the species in Central Europe. 7.4 discussion 63

7.4.4 Future invasion potential

Although the pattern of C. introflexus invasion rather suggests no threat for native flora, in contrast to Western Europe (Biermann and Daniëls, 1997; Hasse and Daniëls, 2006; Klinck, 2008; Sparrius and Kooijman, 2011), this may be only a matter of initial phase of the invasion. Some other habitats whose ecological conditions allow the occurrence of C. introflexus may be invaded later. Therefore we analysed the similarity between species composition of already invaded habitats and all other Czech habitats, using vegetation samples stored in the Czech National Phytosociological Database. Vegetation similar to that already colonised by C. introflexus was assigned into nine vegetation classes. Though this represents almost a quarter of all vegetation classes distinguished in the Czech Republic (Moravec, 1995), most samples belong only to two of them. Although C. introflexus very frequently appears on disturbed peat bogs, no vegetation of drained peatbogs was selected for several reasons — samples from this habitat are not well represented in the Czech National Phytosociological Database or these samples were not included in the dataset C because they did not meet requirements for selection. Some samples of later successional stages of drained peatbogs may also have been assigned to Dicrano-Pinion. In contrast, many vegetation plots assigned to the Fagion alliance were selected. Rather then typical beech forests, the plots represent vegetation of spruce forestry plantations or mixed forests classified probably on the basis of potential vegetation of the stand to the Fagion alliance. The potentially invaded plant communities are distributed across most of the Czech Republic (Fig. 7.1; Chytrý et al., 2010). They represent widespread communities, composed of quite common species, rather than rare communities with unique species composition where C. introflexus invasion might cause a conservation problem Our results further suggest that even pine plantations, the most frequently invaded habitat, will not be invaded equally. Wet pine forests, calcium-rich pine forests and oak-pine forests will probably not be invaded, except for small dry open patches. Invasion of lichen-rich pine forests will depend on the intensity of competition between C. introflexus and lichens. Generally, lichens are stronger competitors forcing C. introflexus into disturbed open patches (Biermann and Daniëls, 1997), and they are even able to re-colonise the moss carpets developed at disturbed patches (Hasse, 2007; but see Zerbe and Wirth, 2006). However, increasing nitrogen deposition may alter this competitive hierarchy (Sparrius and Kooijman, 2011).

7.4.5 Conclusions

Campylopus introﬂexus has already colonised the whole area of the Czech Re- public, but it is more frequent in the western part (Fig. 7.1). This is due to both the earlier colonisation of the western part of the Czech Republic and somewhat lower connectivity of suitable habitats in the eastern part with missing suitable habitats in a signiﬁcant part of the southeast. The species invades predominantly coniferous forests and plantations and disturbed 64 habitatscolonizedby c. introflexus in the central europe

(drained) peatbogs, where it occupies patches with competition pressure, which are most often created by anthropogenic disturbances. Cultivation experiments confirmed that the species grows best at organic, moderately wet soils, and has a low fitness on lime-enriched and waterlogged soils. Analysis of species composition of already invaded communities, and its comparison with a large phytosociological database, suggests that Central Europe has a large amount of suitable habitat for the C. introflexus invasion and that although we can expect geographically wide distribution of the species in the future, there is only small threat for the native flora presently.

souhrn

I když jsou rostlinné invaze ˇcastostudovány, vˇetšinaprací je zamˇeˇrena na cévnaté rostliny. Tato práce je zamˇeˇrena na studium ekologických poža- davk ˚uinvazního mechu Campylopus introflexus v Ceskéˇ republice. Studovali jsme ekologické a vegetaˇcnípreference druhu a aktualizovali data o jeho souˇcasnémrozšíˇrení. Ekologické nároky druhu jsme zkoumali jak in situ, tak i pomocí in vitro experiment ˚u.Pro testování schopnosti druhu C. introflexus r ˚ustza r ˚uzných p ˚ud- ních a vlhkostních podmínek jsme založili kultivaˇcníexperimenty, které jsme vyhodnotili pomocí smíšených lineárních model ˚u.Pomocí klasifikace a ordi- nace fytocenologických snímk ˚us výskytem druhu jsme popsali diverzitu vegetace, do které druh u nás vstupuje. Na základˇefloristické podobnosti existujících snímk ˚us výskytem druhu a všech snímk ˚uuložených v Ceskéˇ národní fytocenologické databázi jsme obecnˇecharakterizovali již invadovaná stanovištˇea predikovali, jaké biotopy mohou být v budoucnu v Ceskéˇ republice invadovány. Z našich výsledk ˚uvyplývá, že C. introflexus není extrémnˇespecializován a vyskytuje se v r ˚uzných rostlinných spoleˇcenstvech. Kultivaˇcníexperimenty ukázaly, že špatnˇeprosperuje na vápníkem obohacených a zamokˇrených p ˚udách.Na rozdíl od západní Evropy, kde C. introflexus kolonizuje písˇcité p ˚udya vytlaˇcuje ohrožené p ˚uvodní druhy, v Ceskéˇ republice expanduje pˇrevážnˇev jehliˇcnatých monokulturách a na narušených rašeliništích. V tˇech- to biotopech osídluje otevˇrená místa s nízkou konkurencí cévnatých rostlin, která jsou ˇcastovytvoˇrená lidskými zásahy. Invadované biotopy ve stˇrední Evropˇese liší od tˇechv západní Evropˇe,což je zp ˚usobenopˇredevším absencí podobných biotop ˚uzp ˚usobenourozdílným klimatem. Došli jsme k závˇeru, že aˇckolise C. introflexus pravdˇepodobnˇestane bˇežnýmdruhem bryoflóry stˇrední Evropy, v souˇcasné dobˇe nepˇredstavuje vážné riziko pro domácí druhy.

acknowledgments

We are grateful to Vít Syrovátka for help with analysis of data from cultivation experiment, Milan Chytrý and David Zelený for valuable comments, Ondˇrej Hájek for creation of maps and ZdenˇekSoldán for help with ﬁeld assistance. Our special thanks go to all colleagues who provided their unpublished reports on local species occurrence. This research was supported 7.4 discussion 65 by Grant Agency of Charles University (project no. 43203154), institutional support of Masaryk University and long-term development project of the Czech Academy of Sciences no. RVO 67985939. references

Berg, C. (1985). “Zur Ökologie der neophytischen Laubmoosart Campylopus introflexus (Hedw.) Brid. in Mecklenburg.” Archiv der Freunde der Naturge- schichte 25, pp. 117–126. Biermann, R. and F.J.A. Daniëls (1995). “Campylopus introflexus (Dicranaceae, Bryopsida) in Flechtenreichen Silber-Grasfluren Mitteleuropas.” In: Flechten Follmann. Contributions to lichenology in Honour of Gerhard Follmann. Ed. by F.J.A. Daniëls, M. Schulz and J. Peine. Cologne: Botanical Institute, Univer- sity of Cologne, pp. 493–500. –(1997). “Changes in a lichen-rich dry sand grassland vegetation with special reference to lichen synusiae and Campylopus introflexus.” Phytocoenolo- gia 27, pp. 257–273. –(2001). “Vegetationsdynamik im Spergulo-Corynephoretum unter besonderer Berücksichtigung des neophytischen Laubmooses Campylopus introflexus.” Braunschweiger Geobotanische Arbeiten 8, pp. 27–37. Bossdorf, O., H. Auge, L. Lafuma, W.E. Rogers, E. Siemann and D. Prati (2005). “Phenotypic and genetic differentiation between native and introduced plant populations.” Oecologia 144, pp. 1–11. Carranza, M.L., C. Ricotta, M. Carboni and A.T.R. Acosta (2011). “Habitat selection by invasive alien plants. A bootstrap approach.” Preslia 83, pp. 529– 536. Chiarucci, A., M. Calderisi, F. Casini and I. Bonini (2008). “Vegetation at the limits for vegetation: Vascular plants, bryophytes and lichens in a geothermal field.” Folia Geobotanica 43, pp. 19–33. Chytrý, M. and M. Rafajová (2003). “Czech National Phytosociological Data- base: basic statistics of the available vegetation-plot data.” Preslia 75, pp. 1– 15. Chytrý, M., L. Tichý, J. Holt and Z. Botta-Dukát (2002). “Determination of diagnostic species with statistical fidelity measures.” Journal of Vegetation Science 13, pp. 79–90. Chytrý, M., P. Pyšek, L. Tichý, I. Knollová and J. Danihelka (2005). “Invasions by alien plants in the Czech Republic: a quantitative assessment across habitats.” Preslia 77, pp. 339–354. Chytrý, M., L.C. Maskell, J. Pino, P. Pyšek, M. Vilà, X. Font and S.M. Smart (2008). “Habitat invasions by alien plants: a quantitative comparison among Mediterranean, subcontinental and oceanic regions of Europe.” Journal of Applied Ecology 45, pp. 448–458. Chytrý, M., P. Pyšek, J. Wild, J. Pino, L.C. Maskell and M. Vilà (2009a). “Eu- ropean map of alien plant invasions based on the quantitative assessment across habitats.” Diversity and Distributions 15, pp. 98–107. Chytrý, M., T. Kuˇcera,M. Koˇcí,V. Grulich and P. Lustyk, eds. (2010). Kata- log biotop˚u Ceskéˇ republiky [Habitat Cataloque of the Czech Republic]. 2nd ed. Prague: Agentura ochrany pˇrírody a krajiny CR.ˇ 66 habitatscolonizedby c. introflexus in the central europe

Daisie (2009). Handbook of alien species in Europe. Vol. 3. Invading nature. Dor- drecht: Springer, p. 399. Ellenberg, H., H.E. Weber, R. Düll, V. Wirth, W. Werner and D. Paulißen (1992). “Zeigerwerte von pflanzen in Mitteleuropa.” Scripta Geobotanica 18, pp. 1–258. Equihua, M. and M.B. Usher (1993). “Impact of carpets of the invasive moss Campylopus introflexus on Calluna vulgaris regeneration.” Journal of Ecology 81, pp. 359–365. Essl, F. and P.W. Lambdon (2009). “Alien bryophytes and lichens of Europe.” In: Handbook of Alien Species in Europe. Ed. by Daisie. Dordrecht: Springer, pp. 29–41. Essl, F., W. Rabitsch and P. Lambdon (2011). “Bryophytes and lichens.” In: Encyclopedia of biological invasions. Ed. by D. Simberloff and M. Rejmánek. Berkeley: University of California Press, pp. 81–85. Frahm, J.P. (1984). “Phytogeography of European Campylopus species.” Pro- ceedings of the Third Meeting of Bryologists from Central and East Europe. Ed. by J. Váˇna.Praha: Univerzita Karlova, pp. 191–212. –(2007). “Campylopus introflexus.” In: Flora of North America North of Mex- ico. Ed. by Flora of North America Editorial Committee. Vol. 27. New York and Oxford, p. 371. url: http://www.efloras.org/florataxon.aspx? flora_id=1&taxon_id=105474 (visited on 09/20/2011). Hasse, T. (2007). “Campylopus introflexus invasion in a dune grassland: succession, disturbance and relevance of existing plant invader concepts.” Herzogia 20, pp. 305–315. Hasse, T. and F.J.A. Daniëls (2006). “Species responses to experimentally induced habitat changes in a Corynephorus grassland.” Journal of Vegetation Science 17, pp. 135–146. Hassel, K. and L. Söderström (2005). “The expansion of the alien mosses Orthodontium lineare and Campylopus introflexus in Britain and continental Europe.” Journal of the Hattori Botanical Laboratory 97, pp. 183–193. Hedenäs, L., T. Herben, H. Rydin and L. Söderström (1989a). “Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and population structure.” Holoartic Ecology 12, pp. 163–172. –(1989b). “Ecology of the invading moss species Orthodontium lineare in Sweden: substrate preference and interactions with other species.” Journal of Bryology 15, pp. 565–581. Hejda, M., P. Pyšek, J. Pergl, J. Sádlo, M. Chytrý and V. Jarošík (2009). “Inva- sion success of alien plants: do habitat affinities in the native distribution range matter?”: Global Ecology and Biogeography 18, pp. 372–382. Herben, T. (1994). “Local rate of spreading and patch dynamics of an invasive moss species, Orthodontium lineare.” Journal of Bryology 18, pp. 115– 125. Hill, M.O. (1979). TWINSPAN: A FORTRAN Program for Arranging Multivari- ate Data in an Ordered Two-way Table by Classification of Individual and At- tributes. New York: Cornell University Press Ithaca. Hájek, M., J. Roleˇcek,K. Cottenie, K. Kintrová, M. Horsák, A. Poulíˇcková, P. Hájková, M. Fránková and D. Dítˇe(2011). “Environmental and spatial controls of biotic assemblages in a discrete semi-terrestrial habitat: com- 7.4 discussion 67

parison of organisms with different dispersal abilities sampled in the same plots.” Journal of Biogeography 38, pp. 1683–1693. Hájková, P., M. Hájek, K. Rybníˇcek,M. Jiroušek, L. Tichý, Š. Králová and E. Mikulášková (2011). “Long-term vegetation changes in bogs exposed to high atmospheric deposition, aerial liming and climate fluctuation.” Journal of Vegetation Science 22, pp. 891–904. Karlin, E.F., R.E. Andrus, S.B. Boles and A.J. Shaw (2011). “One haploid par- ent contributes 100% of the gene pool for a widespread species in north- west North America.” Molecular ecology 20, pp. 753–767. Ketner-Oostra, R. and K.V. Sýkora (2000). “Vegetation succession and lichen diversity on dry coastal calciumpoor dunes and the impact of management experiments.” Journal of Coastal Conservation 6, pp. 191–206. –(2004). “Decline of lichen-diversity in calcium-poor coastal dune vegetation since the 1970s, related to grass and moss encroachment.” Phyto- coenologia 34, pp. 521–549. Klinck, J. (2008). “The alien invasive moss Campylopus introflexus in the Dan- ish coastal dune system.” MA thesis. Depon. in: Copenhagen University, Copenhagen, p. 106. Konvalinková, P. (2010). “Spontaneous vegetation succession in mined peatlands.” PhD thesis. Depon. in: University of South Bohemia, Ceskéˇ Budˇe- jovice. Kubát, K., L. Hrouda, J. Chrtek Jr, Z. Kaplan, J. Kirschner and J. Štˇepánek, eds. (2002). Key to the flora of the Czech Republic. Prague: Academia. Kuˇcera,J. and J. Váˇna(2005). “Check and Red List of bryophytes of the Czech Republic.” Pˇríroda 23, pp. 1–104. Mikulášková, E. (2006). “Vývoj rozšíˇrení neofytického mechu Campylopus introflexus v Ceskéˇ republice [Development in distribution of the neophytic moss Campylopus introflexus in the Czech Republic].” Bryonora 38, pp. 1–10. Moravec, J., ed. (1995). Rostlinná spoleˇcenstva Ceskéˇ Republiky a jejich ohrožení [Red list of plant communities of the Czech Republic and their endangerment]. 2nd ed. Litomˇeˇrice:Severoˇceskoupˇrírodou. – ed. (1998). Pˇrehled vegetace Ceskéˇ republiky, svazek 1, Acidofilní doubravy [Veg- etation Survey of the Czech Republic – Acidophilous Oak Forests, vol. 1]. Prague: Academia. – ed. (2000). Pˇrehled vegetace Ceskéˇ republiky, svazek 2, Hygrofilní, mezofilní a xe- rofilní opadavé lesy [Vegetation Survey of the Czech Republic – Hygrophilous, mesophilous and xerophilous deciduous forests, vol. 2]. Prague: Academia. – ed. (2002). Pˇrehled vegetace Ceskéˇ republiky, svazek 3, Jehliˇcnatélesy [Vegetation Survey of the Czech Republic – Coniferous forests, vol. 3]. Prague: Academia. Mulligan, R.C. and L.D. Gignac (2001). “Bryophyte community structure in a boreal poor fen: reciprocal transplants.” Canadian Journal of Botany 79, pp. 404–411. Novotný, I. (1990a). “Campylopus introflexus (Hedw.) Brid. – nový druh mechu v Ceskoslovensku.”ˇ 2. Bryologicko–lichenologické dny v Opavˇe(19.–21.9.1989) – souhrny referát˚u. Ed. by Anonymus. Vol. 4, pp. 4–5. Pecháˇcková, S. and O. Peksa (2010). “Fytocenologická studie acidofilních bor ˚uPlzeˇnska[Phytosociological study of acidophilic pine forests in Pilsen 68 habitatscolonizedby c. introflexus in the central europe

region (Czech Republic).].” Final project report. Depon in: West-Bohemian Museum, Pilzen. Piessens, K., H. Stieperaere, O. Honnay and M. Hermy (2008). “Effects of management and adjacent forest on the heathland bryophyte layer.” Basic and Applied Ecology 9, pp. 253–262. Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar and R Development Core Team (2011). “nlme: Linear and Nonlinear Mixed Effects Models.” R package version 3.1-101. Pinke, G., P. Karácsony, B. Czúcz and Z. Botta-Dukát (2011). “Environmental and land-use variables determining the abundance of Artemisia ambrosiifo- lia in arable fields in Hungary.” Preslia 83, pp. 219–236. Pyšek, P. and D.M. Richardson (2007). “Traits associated with invasiveness in alien plants: where do we stand?”: In: Biological Invasions. Ed. by W. Nentwig. Berlin: Springer, pp. 97–125. R Development Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, url: http://www.r- project.org (visited on 09/24/2011). Richards, P.W. (1963). “Campylopus introflexus (Hedw.) Brid. and C. polytrichoides De Not. in the British Isles; a preliminary account.” Transactions of the British Bryological Society 4, pp. 404–417. Roleˇcek,J., L. Tichý, D. Zelený and M. Chytrý (2009). “Modified TWINSPAN classification in which the hierarchy respects cluster heterogeneity.” Jour- nal of Vegetation Science 20, pp. 596–602. Sakai, A.K., F.W. Allendorf, J.S. Holt, D.M. Lodge, J. Molofsky, K.A. With, S. Baughman, R.J. Cabin, J.E. Cohen, N.C. Ellstrand et al. (2001). “The Popu- lation Biology of Invasive Specie.” Annual Review of Ecology and Systematics 32, pp. 305–332. Schirmel, J., L. Timler and S. Buchholz (2011). “Impact of the invasive moss Campylopus introflexus on carabid beetles (Coleoptera: Carabidae) and spiders (Araneae) in acidic coastal dunes at the southern Baltic Sea.” Biological Invasions 13, pp. 605–620. Schlüsslmayr, G. (2005). “Soziologische Moosflora des südöstlichen Oberöster- reich.” Stapfia 84, pp. 1–695. Schooler, S.S., T. Cook, G. Prichard and A.G. Yeates (2010). “Disturbance- mediated competition: the interacting roles of inundation regime and mechanical and herbicidal control in determining native and invasive plant abundance.” Biological Invasions 12, pp. 3289–3298. Simberloff, D. and M. Rejmánek, eds. (2011). Encyclopedia of biological invasions. London: Univiversity of California Press. Soldán, Z. (1996). “Rozšíˇrení neofytických mech ˚u Campylopus introflexus a Or- thodontium lineare v Ceskéˇ republice [Distribution of Campylopus introflexus a Orthodontium lineare in Czech Republic].” Brynora 18, pp. 10–19. –(1997). “Invazní mechorosty. [Invasive bryophytes].” Zprávy Ceskéˇ botanické spoleˇcnosti 14, pp. 33–39. Sparrius, L.B. and A.M. Kooijman (2011). “Invasiveness of Campylopus introflexus in drift sands depends on nitrogen deposition and soil organic matter.” Applied Vegetation Science 14, pp. 221–229. 7.4 discussion 69

Stech, M. and J. Dohrmann (2004). “Molecular relationships and biogeography of two Gondwanan Campylopus species, C. pilifer and C. introflexus (Dicranaceae).” Monographs in Systematic Botany from the Missouri Botanical Garden 98, pp. 415–431. Stech, M. and D. Wagner (2005). “Molecular relationships, biogeography, and evolution of Gondwanan Campylopus species (Dicranaceae, Bryopsida).” Taxon 54, pp. 377–382. Størmer, P. (1958). “Some mosses from the phytogeographical excursion 1-9 through the Armorican massive in 1954.” Revue bryologique et lichénologique 27, pp. 13–16. Söderström, L. (1992). “Invasions and range expansions and contractions of bryophytes.” In: Bryophytes and lichens in a changing environment. Ed. by J.W. Bates and A.M. Farmer. New York: Clarendon Press, Oxford University Press, pp. 131–158. Ter Braak, C.J.F. and P. Šmilauer (2002). Canoco 4.5: Reference Manual and Canodraw for Windows. User’s Guide: Software Form Canonical Community Ordination (version 4.5). Ithaca: Microcomputer Power. Tichý, L. (2002). “JUICE, software for vegetation classification.” Journal of Vegetation Science 13, pp. 451–453. –(2005). “New similarity indices for the assignment of relevés to the vegetation units of an existing phytosociological classification.” Plant Ecology 179, pp. 67–72. Tichý, L. and M. Chytrý (2006). “Statistical determination of diagnostic species for site groups of unequal size.” Journal of Vegetation Science 17, pp. 809– 818. Van der Meulen, F., H. Van der Hagen and B. Kruijsen (1987). “Campylopus introflexus. Invasion of a moss in Dutch coastal dunes.” Proceedings of the Koninklijke. Nederlandse Akademie van Wetenschappen. 90, pp. 73–80. Vˇezda,A. and J. Liška (1999). Katalog lišejník˚u Ceskéˇ republiky [Catalogue of lichens of the Czech Republic]. Pr ˚uhonice:Institute of Botany, Academy of Sciences of the Czech Republic, Westhoff, V. and E. Van der Maarel (1978). “The Braun-Blanquet approach.” In: Classification of plant communities. Ed. by R.H. Whittaker. The Hague: W. Junk, pp. 289–399. Williamson, M.H. (1996). Biological invasions. New York: Chapman and Hall. Zerbe, S. and P. Wirth (2006). “Non-indigenous plant species and their ecological range in Central European pine (Pinus sylvestris L.) forests.” Annals of forest science 63, pp. 189–203. Zuur, A.F., E.N. Ieno, N.J. Walker, A.A. Saveliev and G.M. Smith (2009). Mixed effects models and extensions in ecology with R. New York: Springer.

PATTERNSOFGENETICDIVERSITYANDDISPERSALOF 8 CAMPYLOPUSINTROFLEXUS (BRYOPHYTA)INTHE EDGEOFINVASIVEAREA

Eva Mikulášková1,2 and Tomáš Fér2 1Dept. of Botany and Zoology, Faculty of Science, Masaryk University, CZ-61137 Brno, Czech Republic; [email protected]. 2Dept. of Botany, Faculty of Science, Charles University in Prague, CZ-128 01 Praha 2, Czech Republic. abstract

Although vascular plant invasions have been frequently studied, little is known about invasive bryophytes. This research is focused on population genetic structure, dispersal and genetic diversity in the invasive moss Campy- lopus introflexus (Hedw.) Brid. in the Czech Republic, which is on the eastern edge of its invasive area. Genetic patterns in the Czech populations were studied using isozymes and amplified fragment length polymorphism (AFLP). Three AFLP primer combinations provided 253 scorable loci from which 199 (78.66 %) were polymorphic. Seven enzymatic systems provided 21 alleles across eight scorable loci. The numbers of clones (genotypes) within populations depend on (1) whether the population was fertile or sterile and (2) distances among cushions. Populations are mainly multiclonal and appear to originate from multiple colonizing spores. Short-distance spreading by vegetative propagules occurs mainly within individual cushions. Genetic diversity between populations is high and does not decrease with increasing longitude. There are no differences in gene diversity among regions in the Czech Republic except that it was higher Western Bohemia region. Analyses using the software STRUCTURE, indicate that two groups of populations best fit the data. AMOVA analysis revealed that genetic variation is nearly equally divided among (54 %) and within (46 %) populations. Most of the genetic variance occurs among populations within broader regions whereas differentiation among the regions was slight. Genetic structure is not correlated with geography, altitude, pH, age or habitats of populations. We conclude that populations in the Czech Republic have multiple origins. keywords

AFLP, bryophytes, Campylopus introﬂexus, invasive species, isozymes, genetic structure, population biology.

71 72 c. introflexus dispersal and genetic diversity (paper 2)

8.1 introduction

Invasions by alien plants are considered one of the major threats to the diversity of ecosystems (Simberloff and Rejmánek, 2011). Rapid temporal changes in species diversity are features of many plant communities including bryophytes (Hedenäs et al., 1989b); however, alien species could dramatically influence these changes (Drake et al., 1989). Therefore the prediction of future expansion is a powerful instrument for management and conservation of native communities based on the precautionary principle (Ruiz and Carl- ton, 2003; Pyšek and Richardson, 2010). It is necessary to know the biology and ecology of invasive species in detail for modelling the conditions of their future spread. Although these characteristics have been repeatedly studied for vascular plants (e. g. Amsellem et al., 2001; Pellegrin and Hauber, 1999; Sakai et al., 2001; Ward, 2006) little is known about alien bryophytes. Inva- siveness of alien bryophytes is rare in comparison to vascular plants (Essl and Lambdon, 2009; Essl et al., 2011). In Europe, moss invasions are most frequent in oceanic regions and only two alien species, Campylopus introflexus (Hedw.) Brid. and Orthodontium lineare Schwägr., are widespread over large parts of Europe. Invasive plants are typically dispersed by wind and have high propagule production (Moravcová et al., 2010). These characteristics are typical features of bryophytes. The diversity of asexual propagules in bryophytes is unparalleled among land plants (Duckett and Ligrone, 1992). Abundance of small spores that survive long distance dispersal (e. g. Van Zanten and Pócs, 1981; Muñoz et al., 2004) can lead to elevated dispersal and establishment rates (Herben et al., 1991; Herben, 1994). These life history characteristics may contribute to high invasive potential for bryophytes. Characterizing genetic structure in an invasive moss species can provide information about reproductive and colonizing behaviour of the invader. Research on invasive bryophytes has been focused mainly on their ecology (e. g. Biermann and Daniëls, 2001; Hedenäs et al., 1989b; Hasse, 2007), but little work has been reported on mating systems, genetic structure or physiology (e. g. Sparrius and Kooijman, 2011; Herben, 1994). Both genetic structure and dispersal mechanisms could reflect environmental conditions and therefore may provide predictions about future directions and speed of expansion (Gunnarsson et al., 2007). For colonization and maintenance of patches by bryophytes, clonal growth or vegetative reproduction may hence be of fundamental relevance and strongly affect the genetic diversity and constitution of populations (Hassel et al., 2005). High dispersal potential may be a strong homogenizing force that prevents genetic divergence in bryophytes; however, differentiation of cryptic species are also frequent (Shaw, 2001). Genetic structure in invasive populations may be weak compared to their native ranges because founder effects lead to decreased genetic diversity (Thingsgaard, 2001; Stenøien and Såstad, 1999). On the other hand, multiple introductions could lead to increased genetic diversity in invasive species (e. g. Sakai et al., 2001; Pairon et al., 2010; Dlugosch and Parker, 2008). The invasion of Campylopus introflexus (Hedw.) Brid. in Europe has re- ceived much attention and its expansion is so well documented that it may 8.1 introduction 73 serve as a model for the bryophyte invasion process. It is a neophytic species whose natural range is in the southern hemisphere (South America, southern part of Africa and Australia, New Zealand; Frahm, 1984). The habitats of this species in its native range are coastal sand dunes along the coast, open scrubs, fynbos and bare soil along trails Frahm, 2007. It has been uninten- tionally introduced, probably with commodities to France and Great Britain (Richards, 1963; Størmer, 1958). It started to spread quickly eastwards and is now known to occur in 21 countries (Hassel and Söderström, 2005; Essl and Lambdon, 2009). C. introflexus was reported for the first time from the Czech Republic (Central Europe) in 1988, from South Bohemia (Novotný, 1990a). The numbers of new localities exponentially increased in the following years and today it is known from more than 80 localities (Soldán, 1996, 1997; Mikulášková, 2006). It invades acidic, dry habitats with little competition from vascular plants and mosses (Mikulášková et al., 2012a). C. introflexus is dioicous and both sexual and asexual dispersal is thought to be common (Hassel and Söderström, 2005). Inbreeding can occur in dioicous bryophytes. Different haploid gametophores may have originated from same capsule so mating between such plants correspond to “self-fertilization” as it is commonly labelled in seed plants. Therefore we could expect losses of heterozygosity when mating occurs among related gametophytes, even in dioicous species with unisexual gametophytes. The taxonomic relationships of C. introflexus have been intensively studied using molecular methods (e. g. Stech and Dohrmann, 2004; Stech and Wagner, 2005). The molecular data resolve C. introflexus as a monophyletic taxon with low infraspecific sequence variation. However, sequencing specific genes may not be sufficient for detecting genetic variability on fine scales. More sensitive genetic markers could elucidate the population structure of this moss and could provide new information about the process of its European invasion. Molecular studies using isozymes and Amplified Fragment Length Polymorphisms (AFLPs; Vos et al., 1995) have detected genetic differentiation in bryophytes even on small spatial scales (compare, e. g. Cronberg, 2002; Selkirk et al., 1997; Vanderpoorten and Tignon, 2000; Pfeiffer et al., 2006). AFLP analyses have facilitated distinguishing vegetative and sexual origins of individually populations, helped identify clones and clarified the structure of cushions (Pfeiffer et al., 2006). AFLPs seem to produce very polymorphic markers, useful for investigations of population genetic structure in bryophytes (e. g. Vanderpoorten and Tignon, 2000; Fer- nandez et al., 2006; Pfeiffer et al., 2006; McDaniel, Shaw et al., 2005; Gunnars- son et al., 2005; Zartman et al., 2006). Isozymes have been used extensively for genetic analyses of bryophyte populations (e. g. Cummins and Wyatt, 1981; Yamazaki, 1981; Wyatt et al., 2005) Applying two types of molecular markers (AFLP fingerprinting, isozymes) we investigate genetic variation in Campylopus introflexus in the Czech Re- public. In particular we aim to answer the following questions. (1) Are cushions and populations multiclonal or uniclonal? (2) Is dispersal among populations predominantly by spores, or by vegetative propagules? Are clones shared among populations? (3) Are there differences in genetic variability among regions within the Czech Republic? (4) Is there spatial genetic struc- 74 c. introflexus dispersal and genetic diversity (paper 2)

Figure 8.1: Map of investigated populations of Campylopus introﬂexus in Czech Re- public showing geographical “Regions” and “Zones”.

ture within the Czech Republic? Are geographically close populations genetically more similar than more widely separated populations?

8.2 methods

8.2.1 Samples

Material was sampled from 57 populations (51 from the Czech Republic, two from Poland, one from Slovakia, and three from South Africa; Table 8.1, Fig. 8.1). Sampling covered all known habitats of C. introﬂexus (peatbogs, spruce and pine forests, sand dunes) and the full range of known altitudes (from 200 to 1150 m a. s. l.). Populations were placed into groups according to their geographical origin at two hierarchical levels. First, nine groups were created (“Regions”; Africa, Poland, Moravia, Czech–Moravian Highlands, South Bo- hemia, Central Bohemia, North Bohemia, Bohemian Forest, and West Bo- hemia). Subsequently, these regions were pooled into three larger groups — “Zones” (Czech Republic East, Czech Republic West, South Africa + Poland; see Table 8.1 for details). 8.2 methods 75

Table 8.1: Localities of studied populations of Campylopus introﬂexus. Numbers in 1.– 3. column — number of sampled gametophytes, CZ – Czech Republic, PL – Poland, SK – Slovakia, ZA – Republic of South Africa, PLA – Protected Land Area, NP – National Park, NR – Nature reserve, [. . . ] – code in Fig. 8.1.

Dataset Dataset B Dataset C Fertile / Zone / Geographical region Locality WGS 84 – WGS 84 – Altitude A (AFLP) (AFLP) (iso- sterile N [°] E [°] [m a. s. l.] zymes) 1 F Africa / – ZA; NP Table Mountain, Cape Town. −33.95964 18.40372 1000 1 S Africa / – ZA; NP Tsitsikama, Storm river. −34.02263 23.89436 300 1 1 S CZ East / C–M Highlands CZ; Ceskáˇ Kanada Nature Park, Albeˇr. 49.02867 15.15497 548 1 S CZ East / C–M Highlands CZ; Ceskáˇ Kanada Nature Park, H ˚urky. 49.04654 15.15868 654 1 1 S CZ East / C–M Highlands CZ; Ceskomoravskáˇ vrchovina Mts., Popice. 49.34095 15.538 619 1 S CZ East / C–M Highlands CZ; Toulovcovy Maštale NR, Bor u Skutˇce. 49.81686 16.14192 522 3 1 F CZ East / C–M Highlands CZ; Toulovcovy Maštale NR, Bor u Skutˇce. 49.81686 16.14192 505 [Osli] 3 S CZ East / C–M Highlands CZ; Toulovcovy Maštale NR, Budislav. 49.80661 16.15167 522 S CZ East / C–M Highlands CZ; Toulovcovy Maštale NR, Budislav. 49.80517 16.15333 522 1 S CZ East / C–M Highlands CZ; Toulovcovy Maštale NR, Zderaz. 49.83019 16.11375 425 1 3 S CZ East / Moravia CZ; Benešov. 49.49239 16.76822 662 1 S CZ East / Moravia CZ; Hˇrebeˇc. 49.72772 16.57578 616 1 S CZ East / Moravia CZ; PLA Beskydy, Hostašovice. 49.50627 17.98976 400 5 3 1 S CZ East / Moravia CZ; PLA Beskydy, Hostašovice. [Valmez_I] 49.50533 17.99122 400 1 S CZ East / Moravia CZ; PLA Moravský Kras, Vilémovice. 49.37972 16.72861 430 1 1 S CZ West / Bohemian Forest CZ; Kralovické louky NR, Nebahovy. 49.00494 14.08364 630 1 S CZ West / Bohemian Forest CZ; Šumava Mts., Gerl ˚uvpotok I. zone of NP. 49.16717 13.29913 970 1 3 S CZ West / Bohemian Forest CZ; Šumava Mts., Kašperské Hory. 49.15708 13.556 667 1 S CZ West / Bohemian Forest CZ; Šumava Mts., Nový Brunst I. zone of NP. 49.17880 13.2712 972 1 1 S CZ West / Bohemian Forest CZ; Šumava Mts., Nový Brunst I. zone of NP. 49.17884 13.27127 972 5 S CZ West / Bohemian Forest CZ; Šumava Mts., Popelná. 49.09574 13.58812 942 1 1 S CZ West / Bohemian Forest CZ; Šumava Mts., Raˇcín. 48.70867 14.04464 732 1 1 F CZ West / Bohemian Forest CZ; Šumava Mts., Soumarský most. 48.90787 13.83383 750 1 F CZ West / Bohemian Forest CZ; Šumava Mts., V MokˇrináchI. zone of NP. 49.12822 13.39072 850 1 S CZ West / Bohemian Forest CZ; Šumava Mts., Zh ˚uˇrskélouky I. zone of NP. 49.18697 13.31267 931 1 F CZ West / Central Bohemia CZ; Džbán Nature Park, Bílichov. 50.25411 13.90811 352 1 S CZ West / Central Bohemia CZ; Džbán Nature Park, Holedec. 50.26506 13.58033 291 1 S CZ West / Central Bohemia CZ; Džbán Nature Park, Žichovec. 50.27528 13.92364 379 1 S CZ West / Central Bohemia CZ; Petrov. 49.89304 14.4603 400 1 F CZ West / Central Bohemia CZ; Psáry. 49.92131 14.50839 409 1 1 F CZ West / Central Bohemia CZ; Radlík. 49.92131 14.50839 442 1 F CZ West / Central Bohemia CZ; Radlík. 49.91526 14.5115 442 1 S CZ West / North Bohemia CZ; Dolní Rasnice.ˇ 50.96231 15.15478 400 1 S CZ West / North Bohemia CZ; Krkonoše Mts., Pec pod Snˇežkou. 50.72689 15.72383 1033 1 1 F CZ West / North Bohemia CZ; NP Ceskéˇ Švýcarsko, Kyjov. 50.91911 14.44447 431 2 F CZ West / North Bohemia CZ; NP Ceskéˇ Švýcarsko, Mezná. 50.87331 14.29647 269 1 1 S CZ West / North Bohemia CZ; PLA Lužické hory, Jedlová. 50.84975 14.57194 568 1 S CZ West / South Bohemia CZ; PLA Tˇreboˇnsko,Hluboká u Borovan. 48.89242 14.69145 475 1 2 F CZ West / South Bohemia CZ; PLA Tˇreboˇnsko,NR Borkovická blata, 49.23847 14.62239 420 Mažice. 1 2 F CZ West / South Bohemia CZ; PLA Tˇreboˇnsko,NR Borkovická blata, 49.23739 14.62597 428 Mažice. 3 1 2 F CZ West / South Bohemia CZ; PLA Tˇreboˇnsko,NR Borkovická blata, 49.23334 14.61666 420 Mažice. [Bork_III] 3 1 1 S CZ West / South Bohemia CZ; PLA Tˇreboˇnsko,NR V rájích, Spolí. [Raje] 48.98583 14.70894 445 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Brtná. 50.00686 12.50961 650 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Brtná. 50.00603 12.51319 650 1 F CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Brtná. 50.00483 12.51686 630 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Karlova Hut’. 49.55967 12.60647 691 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Mariánské Láznˇe. 49.92872 12.50769 756 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Rybník. 49.50178 12.68553 593 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Tachovská Hut’. 49.93195 12.55037 724 4 3 F CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Tachovská Hut’. [Tisina_I] 49.93183 12.55058 724 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Vysoká. 49.96211 12.49056 810 1 S CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Vysoká. 49.95928 12.49608 850 1 F CZ West / West Bohemia CZ; PLA Ceskýˇ Les, Železná. 49.60394 12.56614 517 5 3 F Poland / – PL; NP Słowi´nskiPark Narodowy, Izica. 54.697 17.50682 15 [PL_II] 2 1 1 F Poland / – PL; NP Słowi´nskiPark Narodowy, Łeba. [PL_I] 54.74192 17.37655 2 1 S Slovakia / – SK; Prievaly. 48.57765 17.3339 208

8.2 methods 77

One to ﬁve gametophytes were collected in each population (see below). Samples were divided into three datasets for analyses (Table 8.1). Dataset A (AFLP) contained 25 gametophytes from seven populations. From three populations gametophytes were collected from more than one cushion (coded as A, B, C in Fig. 8.2). Other populations are represented by a single cushion each. From all cushions at least two gametophytes were collected to detect possible genetic variability within cushions (coded as 1, 2, 3 in Fig. 8.2). Pop- ulations included in dataset A are a subset from dataset B and included populations with at least two gametophytes from cushions. This dataset was used to elucidate genetic and clonal diversity within population and cushions. Dataset B (AFLP) contained 39 gametophytes from 33 populations. From three populations more than one cushion was collected (see dataset A). This dataset was analysed in order to get an overview of genetic variation in the studied area. Dataset C (isozymes) contained 51 gametophytes from 39 populations. From eight populations more than one gametophyte was collected.

8.2.2 DNA extraction and AFLP ﬁngerprinting

Dry stems (gametophyte tissue) from herbarium specimens (max. 2 years old) or fresh stems brought directly from the field were used for DNA extractions using Invisorb Spin Plant Mini Kit (Invitek). Brown tissue and inorganic matter were manually removed from all samples. Gametophytes were rehydrated and washed several times in distilled water. About 100 to 400 mg of clean wet tissue from one gametophyte was used for each extraction. The AFLP method (Vos et al., 1995) with several modifications (Mikuláš- ková et al., 2012b) was applied using AFLP Core Plant Reagent Kit I and AFLP Pre-Amp Primer Mix I (both Invitrogen). Thirty-six different combinations of selective primers were tested. Three primer combinations (EcoRI- ACA / MseI-CAT, EcoRI-AAC / MseI-CAA and EcoRI-AGG / MseI-CAA) amplified evenly distributed and easily scorable DNA fragments over the range of 50–500 bp, and were used for analysis of all samples.

8.2.3 Isozyme analysis

Fresh plants from ﬁeld, cultivated under standard conditions on sterile soil were used for isozyme extractions. Brown tissue and inorganic matter were manually removed from all samples. About 30–50 mg of clean tissue was homogenized in extraction buffer (Akiyama, 1994) using a grinding mortar with pestle that were continuously cooled with ice. 100 µl extraction buffer was used per 10 mg of moss tissue. About 25 to 40 µl of centrifuged supernatant was used for electrophoresis. Isozymes were vertically separated on polyacrylamid gels using HOEFER SE 600 apparatus (1 hour 180 V, than 350 V to full separation; Kirschner et al., 1994; Boisselier-Dubayle and Bis- chler, 1994). The whole system was continuously cooled with ice to 4 °C. Vi- sualization of enzymatic system followed Soltis and Soltis (1989); Kirschner et al. (1994). 78 c. introflexus dispersal and genetic diversity (paper 2)

Figure 8.2: Neighbor-joining tree based on AFLP pattern of 25 samples of Campy- lopus introﬂexus from 7 populations (dataset A). Bootstrap values higher than 50 % are indicated (based on 3000 permutations). Samples from the Czech Republic East in italics; Czech Republic West in bold; A–C label of cushion; 1–3 label of individual plant. 8.2 methods 79

8.2.4 Data analysis

AFLP electropherograms were analysed and manually scored using Gene- Marker v1.8 (SoftGenetics, LLC., State College, PA, U.S.A.). Distinct polymorphic peaks ranging from 50 to 500 bp were scored as present or absent. Peaks with unambiguous determinations for all electropherograms were included in the analyses. All isozyme gels were digitally photographed and manually scored. Banding patterns were interpreted in terms of discrete alleles of separate putative loci. Loci with unambiguous determinations for all electropherograms were included in the analysis. Since haploid gametophytes were analysed one allele per locus and sample was expected. Samples with two alleles in one locus were omitted due to possible mixture of more samples. For datasets A and B the numbers of genotypes, proportions of variable markers and average gene diversity (with 95% confidence intervals based on 1000 of bootstrap permutations; Nei, 1973) within populations, regions and zones were computed using AFLPdat (Ehrich, 2006). A pairwise population matrix of Nei’s genetic distances was computed in GENALEX, ver. 6.41 (Peakall and Smouse, 2006). For all calculations individuals with up to five different bands were treated as identical (based on an average genotyping error rate of 0.0451; see Mikulášková et al., 2012b). Populations were randomly selected so total gene diversity should not be biased by the different numbers of populations representing regions. For dataset C number of alleles (Na), estimates of effective number of alleles (Ne), Shannon’s diversity index (I), haploid genetic diversity (h), unbiased haploid diversity (uh), and pairwise population matrix of Nei’s genetic identities and Nei’s genetic distances were accomplished in GENALEX, ver. 6.41 (Peakall and Smouse, 2006). For datasets A and B, matrices of inter-individual distances (= 1−Jaccard’s similarity) were calculated and neighbor-joining (NJ; Saitou and Nei, 1987) trees were generated using FAMD ver. 1.2β Schlüter and Harris, 2006. Branch support was evaluated using 3000 bootstrap permutations and bootstrap values > 50 % are indicated in the NJ tree. Resulting trees were visualized using FigTree (Ver. 1.3.1, Andrew Rambaut, Institute of Evolutionary Biology, Uni- versity of Edinburgh, http://tree.bio.ed.ac.uk). Relationships among individuals from dataset B were visualized using principal coordinates analysis (PCoA; Gower, 1966) based upon Jaccard’s similarities (computed in FAMD ver. 1.2β; (Schlüter and Harris, 2006)). The graph was plotted in OpenOffice.org Calc (ver. 3.2.1, Oracle). For datasets B and C, Nei’s genetic distance (Nei, 1973) matrices were generated for AFLP and isozyme data, respectively, and the correlation between these matrices was calculated by Mantel tests (Mantel, 1967) using GENALEX. Significance of correlations was tested using of 999 permutations. Genetic distance matrices were calculated with the “interpolate missing data” feature off. Matrices were generated for 17 samples with data for both AFLP and isozymes. Analyses of molecular variance (AMOVA) were computed in order to understand the partitioning of genetic variation within and between pop- 80 c. introflexus dispersal and genetic diversity (paper 2)

ulations (dataset A) and across geographical regions and zones (dataset B). Four AMOVAs were computed: (1) among and within populations of dataset A,(2) among and within three zones (dataset B), (3) among and within eight regions (dataset B), and (4) and among-within zones and regions simultaneously (dataset B). Percentages of total variance at each level and PhiPT, an analogue of the fixation index (FST), were computed using Arlequin ver. 3.5.1.2 (Excoffier and Lischer, 2010). Significance of all values were tested using 1000 permutations. To test correlations between genetic (Jaccard’s) and geographic distances among individuals from datasets B and C, Mantel tests (Mantel, 1967) based on Pearson’s correlations (Mantel, 1967) were calculated using zt software (Bonnet and Van de Peer, 2002). Significance of correlations was tested using 1000 permutations. Moreover, spatial autocorrelation of kinship coefficients (Hardy et al., 2003) was calculated for dataset B using SPAGeDi ver. 1.3 (Hardy and Vekemans, 2002). Inter-individual distances were automatically divided into five equifrequent distance classes, with an additional class representing comparisons within populations. Average values for each distance class were computed and standard errors for the multilocus estimates of the kinship coefficients per distance class were estimated using a jackknife procedure over the loci. Significance of all coefficients were tested using 1000 permutations within distance classes. For dataset B, we tested for genetic admixture among samples and correspondence of genetic structure with defined regions using STRUCTURE ver. 2.3.2 (Pritchard et al., 2000). This program applies a Bayesian model- based clustering method, which uses a Markov chain Monte Carlo (MCMC) algorithm to organize genetically similar individuals into clusters using multilocus genotype data. As AFLP are dominant markers, a recessive allele model was used. Admixture analyses were run with K = 2 through K = 7, with 10 replicate runs of 1 million generations (following a burn-in of 250000 generations) at each value of K with allele frequencies uncorrelated. The optimum number of clusters (K) was assessed using the ∆-K method (Evanno et al., 2005). Replicate runs within in each value of K were summarized using CLUMPP ver. 1.1.2 (Jakobsson and Rosenberg, 2007) and visualized using a custom R script (Matthew G. Johnson, DUKE University).

8.3 results

Three AFLP primer combinations provided 253 scorable loci from which 199 (78.66 %) were polymorphic. Seven from the ten tested enzymatic systems were polymorphic and two were monomorphic (AAT, SOD). In total, 21 alleles across eight scorable loci were detected (SHDH, ADH, 6-PGDH, EST, PGM-1, PGM-2, PGI, LAP). Other loci were omitted because of too much missing data or ambiguous bands. The average number of alleles per locus is 1.625 (±0.075 SE). The effective number of alleles (the number of equally frequent alleles in an ideal population with homozygosity equivalent to the actual population) was 1.326 (±0.045 SE) (Table 8.2). The percentage of polymorphic loci was 58.93 % (±6.52 %. 8.3 results 81

Table 8.2: Regional variability within Campylopus introﬂexus populations at izozymes loci. N - number of samples, Na - number of alleles, Ne - effective number of alleles (estimate of the number of equally frequent alleles if the population were ideal), I – Shannon’s information index, h – haploid genetic diversity (the probability that two haploid individuals are different), uh – unbiased haploid diversity, and %P – percentage of polymorphic loci.

Locus Region Total %P SHDH ADH 6-PGDH EST PGM-1 PGM-2 PGI LAP N 13 13 13 13 13 13 13 12 103 75.00 % Na 1 1 2 2 3 2 2 2 15 Ne 1.000 1.000 1.166 1.166 1.374 1.988 1.166 1.600 10.459 I 0.000 0.000 0.271 0.271 0.536 0.690 0.271 0.562 h 0.000 0.000 0.142 0.142 0.272 0.497 0.142 0.375 Bohemian Forest uh 0.000 0.000 0.154 0.154 0.295 0.538 0.154 0.409 N 88 88 8 8846062.50 % Na 2 1 2 1 2 2 1 2 13 Ne 1.600 1.000 1.600 1.000 1.600 2.000 1.000 1.600 11.400 I 0.562 0.000 0.562 0.000 0.562 0.693 0.000 0.562 h 0.375 0.000 0.375 0.000 0.375 0.500 0.000 0.375 C–M Highlands uh 0.429 0.000 0.429 0.000 0.429 0.571 0.000 0.500 N 66 66 6 6664862.50 % Na 2 1 2 2 2 2 1 1 13 Ne 1.385 1.000 1.385 1.385 1.385 1.800 1.000 1.000 10.338 I 0.451 0.000 0.451 0.451 0.451 0.637 0.000 0.000 h 0.278 0.000 0.278 0.278 0.278 0.444 0.000 0.000 Central Bohemia uh 0.333 0.000 0.333 0.333 0.333 0.533 0.000 0.000 N 66 66 6 6664850.00 % Na 1 1 2 1 1 2 2 2 12 Ne 1.000 1.000 1.800 1.000 1.000 1.385 1.800 1.800 10.785 I 0.000 0.000 0.637 0.000 0.000 0.451 0.637 0.637 Moravia h 0.000 0.000 0.444 0.000 0.000 0.278 0.444 0.444 uh 0.000 0.000 0.533 0.000 0.000 0.333 0.533 0.533 N 66 66 6 6664875.00 % Na 2 2 2 1 2 2 1 3 15 Ne 1.385 1.385 1.385 1.000 1.800 1.800 1.000 2.000 11.754 I 0.451 0.451 0.451 0.000 0.637 0.637 0.000 0.868 h 0.278 0.278 0.278 0.000 0.444 0.444 0.000 0.500 North Bohemia uh 0.333 0.333 0.333 0.000 0.533 0.533 0.000 0.600 N 88 88 8 8886462.50 % Na 2 1 2 1 2 2 1 2 13 Ne 1.882 1.000 1.280 1.000 1.280 1.600 1.000 1.280 10.322 I 0.662 0.000 0.377 0.000 0.377 0.562 0.000 0.377 h 0.469 0.000 0.219 0.000 0.219 0.375 0.000 0.219 South Bohemia uh 0.536 0.000 0.250 0.000 0.250 0.429 0.000 0.250 N 44 44 4 4443225.00 % Na 2 1 1 1 1 1 2 1 10 Ne 1.600 1.000 1.000 1.000 1.000 1.000 1.600 1.000 9.200 I 0.562 0.000 0.000 0.000 0.000 0.000 0.562 0.000 h 0 375 0 000 0 000 0 000 0 000 0 000 0 375 0 000 West Bohemia ...... uh 0.500 0.000 0.000 0.000 0.000 0.000 0.500 0.000 N 51 51 51 51 51 51 51 46 403 58.93 % Na 3 2 3 3 3 2 2 3 21 Ne 1.487 1.040 1.381 1.082 1.377 1.800 1.169 1.501 10.837

Total I 0.593 0.097 0.538 0.193 0.523 0.637 0.275 0.629 h 0.328 0.038 0.276 0.076 0.274 0.444 0.145 0.334 uh 0.334 0.039 0.282 0.078 0.279 0.453 0.147 0.341 82 c. introflexus dispersal and genetic diversity (paper 2)

Table 8.3: Genetic diversity pattern (number of genotypes, proportion of variable markers and average gene diversity with 95% conﬁdence intervals) of Campylopus introﬂexus populations from dataset A and B. N – number of sampled plants / cushions, CZ – Czech Republic, F/S – fertile or sterile population.

Zone Population / region N Nb of Mean of Proportion Average Nei’s gene genotypes longitude of variable diversity (95 % conﬁdence (WGS-84) markers interval) (1) Dataset A – within populations Poland 7 6 0.537 0.250 (0.199 − 0.298) PL (F) 7/4 6 17.474 ° 0.537 0.250 (0.204 − 0.295) CZ – east 8 7 0.758 0.358 (0.314 − 0.401) Osli (F) 3/1 3 16.142 ° 0.168 0.112 (0.063 − 0.161) Valmez_I (S) 5/3 3 17.991 ° 0.442 0.234 (0.181 − 0.291) CZ – west 10 7 0.695 0.316 (0.270 − 0.363) Bork_III (F) 3/1 2 14.622 ° 0.337 0.225 (0.161 − 0.288) Raje (S) 3/1 1 14.709 ° 0.074 0.049 (0.014 − 0.091) Tisina_I (F) 3/1 4 12.551 ° 0.316 0.184 (0.130 − 0.244) (2) Dataset B – among populations Africa+Poland 6 6 0.518 0.237 (0.201 − 0.270) Africa 2 2 21.152 ° 0.106 0.106 (0.065 − 0.151) Poland 4 4 17.507 ° 0.337 0.182 (0.145 − 0.220) CZ – east 10 10 0.759 0.307 (0.278 − 0.336) Moravia 5 5 17.99 ° 0.513 0.250 (0.214 − 0.284) C-M Highlands 5 5 15.497 ° 0.568 0.285 (0.251 − 0.322) CZ – west 23 23 0.955 0.330 (0.309 − 0.350) South Bohemia 4 4 14.643 ° 0.503 0.281 (0.243 − 0.322) North Bohemia 2 2 14.508 ° 0.291 0.291 (0.231 − 0.357) Bohemian Forest 8 8 13.594 ° 0.688 0.267 (0.236 − 0.295) West Bohemia 9 9 12.541 ° 0.839 0.362 (0.332 − 0.388)

8.3.1 Clonal structure

Based on the AFLP data (after corrections for genotyping error; see Meth- ods ), four cushions from which more than one sample was analysed were composed of one clone (Table 8.3, Fig. 8.2). In four populations cushions comprising more genotypes were detected. No clones were shared between two or more cushions within populations, in three populations each cushion possessed its own genotype. The numbers of clones (genotypes) within populations was similar with whether the population was fertile or sterile. No clones were shared between two or more populations, each population possessed unique genotypes. Based on the isozymes data, 23 haplotypes were shared among populations. Identical multilocus genotypes occurred in both sterile and fertile populations. In three populations cushions comprising multiple genotypes were detected. Five cushions from which more than one samples was analysed were composed of one multilocus isozyme genotype. The numbers of clones (genotypes) detected within a population depended on distances among cushions. Cushions up to 20 meters from each other shared clones frequently. With increasing distance among sampled cushions there was an increased probability that cushions had different genotypes. 8.3 results 83

8.3.2 Genetic diversity

All populations of C. introflexus in the Czech Republic have unique AFLP banding patterns (i. e. there is no sharing of haplotypes among populations). The percentage of polymorphic loci (%) ranged from 7 to 53 % within populations (dataset A) and from 10 to 83 % within regions (dataset B). The most polymorphic zone is the western part of the Czech Republic (zone CZ west, 95 %) with nine samples; the lowest polymorphism occurred in the eastern part of the country (zone CZ east), with only 76 % (Table 8.3). Overall genetic diversity (h) measured from the AFLP’s data is 0.161 (±0.005 SE). Average Nei’s gene diversity was 0.237, 0.307 and 0.330 within Africa+Poland, CZ east and CZ west zones, respectively (Table 8.3). The Africa+Poland samples have significantly lower genetic diversity than both Czech zones (95% confidence intervals were not overlapping). Gene diversity within regions in the Czech Republic is very similar (Table 8.3), with exception of West Bohemian region. This region has significantly higher gene diversity (0.362) than any of the other sampled areas. Gene diversity within populations ranged between 0.049 to 0.250; in geographical regions between 0.106 to 0.362; in zones between 0.518 to 0.995. There is no clear pattern of differences in genetic diversity among populations (confidence intervals of average gene diversity overlap). Genetic diversity does not decrease (in either direction) with longitude (see Table 8.3). Pairwise population comparison of Nei’s genetic distance matrices generated from the AFLP data shows significant differentiation both Bohemian Forest and Moravia regions (PhiPT = 0.103 − 0.383; P < 0.05). There is low variability in isozymes among populations. Diversity statistics are provided separately for eight geographical regions (Table 8.2). The percentage of polymorphic loci (%P) ranged from 25 % (West Bohemia region) to 75 % (North Bohemia, Bohemian Forest regions; dataset C; Table 8.2). Pairwise population comparisons of Nei’s genetic identity shows a high degree of genetic similarity between regions (Nei’s identity values 0.869 − 0.997) with low overall variation. The West Bohemia region differs significantly from most other regions in genetic diversity (P < 0.05), however the difference is not significant in comparison with Moravia and South Bohemia. Overall diversity (h) measured from isozyme data is 0.309 (±0.037 SE). Shannon’s diversity index among the eight regions ranged between I = 0.141 (±0.092 SE; West Bohemia) to I = 0.437 (±0.108 SE; North Bohemia), averaged across loci. We detected one private allele each for North Bohemia (ADH-b), Central Bohemia (EST-a) and Bohemian Forest (EST-c) regions. A total of 21 alleles were detected in the zone CZ east, compared with 15 in the CZ west. No allele was private for the CZ east zone. Six alleles were private for the CZ west zone (SDH-c, ADH-b, EST-a,c, PGM-1-a, LAP-c).

8.3.3 Genetic structure

The AMOVA for dataset A (Table 8.4) revealed that genetic variation is nearly equally divided among (54 %) and within (46 %) populations. There is a high level of genetic differentiation among the seven populations (PhiPT = 84 c. introflexus dispersal and genetic diversity (paper 2)

Table 8.4: Results of the hierarchical analysis of molecular variance (AMOVA) of populations of Campylopus introﬂexus – dataset A and B. Levels of signiﬁ- cance tests are based on 1023 permutations.

Source of variation d.f. Sum of Variance Percentage Fixation Index P value squares components of variation (1) among and within populations of dataset A Among populations 5 270.260 10.98905 Va 53.92 FST = 0.53916 > 0.00000 Within populations 19 178.460 9.39261 Vb 46.08 Total 24 448.720 20.38165 (2) among and within 3 zones: Czech Republic West, Czech Republic East, Africa+Poland (dataset B) Among zones 2 104.065 1.91913 5.83 FST = 0.05835 < 0.006 Within zones 36 1114.961 30.97114 94.17 Total 38 1219.026 (3) among and within 8 regions: Africa, Poland, Moravia, C-M Highlands, Bohemian Forest, South, North, West Bohemia (dataset B) Amongregions 7 354.548 4.83234 14.77 FST = 0.14769 < 0.00000 Withinregions 31 864.478 27.88638 85.23 Total 38 1219.026 32.71872 (4) three-level AMOVA for zones and regions (dataset B) Among zones 2 104.065 0.15819 0.48 > 0.2 Among regions within zones 5 250.483 4.72704 14.42 FST = 0.14907 < 0.00000 Within regions 31 864.478 27.88638 85.09 < 0.00000 Total 38 1219.026 32.77161

0.54). There are high levels of variation within the three geographic zones represented in dataset B (Czech Republic West, Czech Republic East, Africa + Poland; 94 %), but plants in the three zones are only slightly differentiated (i. e. only 6 % of the total variation is partitioned among zones). Similarly, only about 15 % of the variance is attributable to differentiation among regions, with the majority of variance (85 %) within regions. (Table 8.4). The three-level AMOVA (regions nested within zones) corroborated this pattern; most variation within regions and little among zones. Nevertheless, all components of variation are significantly greater than zero when the samples were partitioned among the three levels of variation, with the exception of differences among zones (P = 0.283). The Bayesian analysis of the complete dataset using STRUCTURE produced consistent results for only those runs in which K = 2 (similarity coefficient = 0.999, among 10 repeats; Fig. 8.3). Thus the analysis identified two genetic clusters. Twenty four out of the 39 gametophores show less than 10 % genetic admixture between these two clusters. Only four gametophores from one region seems to be genetically pure for one cluster whereas ten individuals from four regions are pure for the second cluster. Genetic structure (clusters) does not clearly correspond to geographic regions or zones. First cluster corresponds to the individuals from North Bohemia, Africa, major part of Bohemian Forest, and several individuals from West Bohemia regions. Second cluster comprises some (slightly admixed) individuals from Moravia, Poland, Bohemian Forest, C–M Highlands and West Bohemia). Two regions (Bohemian Forest and West Bohemia) were clearly colonized by multiple genetic clusters with low levels of subsequent gene flow among them. Other regions show genetic admixture in samples (South Bohemia, Moravia, 8.4 discussion 85

Poland, C–M Highlands) which suggests recombination among the major clusters within region or origin from already recombined populations. Pop- ulations in the Czech Republic originated from at least two genetic sources. African populations are not separated from populations in the Czech Repub- lic as a third cluster.

8.3.4 Genetic relationships at the population scale

Genetic relationships among populations in dataset A are illustrated in the NJ tree (Fig. 8.2). All Polish populations are grouped in a separate cluster. Populations from the western and eastern parts of the Czech Republic do not cluster separately. Samples from individual populations are clones or genetically close and group together (except the oldest population in the Czech Republic — Bork_III; Novotný, 1990b). Major clusters are supported with bootstrap value higher than 50 %. Clustering of dataset B (tree not shown) based on AFLP markers reveals no clear pattern in relatedness among regions; bootstrap support was absent for most of the clades (< 50 %), and the Mantel test revealed no significant correlation between geographic and genetic distance (r = −0.081777, p = 0.173483). The PCoA of this wide dataset suggested only weak structure among samples from different populations (Fig. 8.4). Neither populations nor geographic areas form separate groups. The principal coordinate axes are not correlated with any of the clearly distinguishable environmental gradients (geographical spread from west to east, altitude, pH, habitat; size or fertility of population; age). Only populations from Africa (part of the native range of C. introflexus) form a well separated group in relation to the third axis. The first axis explains 19.77 % of the total variance, the second and third explaining only 9.76 % and 9.50 %. The NJ tree based on isozyme markers (dataset C, data not shown) corresponds well with results obtained from dataset B (AFLPs), but there is low support for branches. Genetic similarities among populations do not correspond with geographic position, as with the AFLP data. The correlation between matrices of Nei’s genetic distances based on isozymes versus AFLP’s markers was not significant (r = 0.014, p = 0.459). There was a low but significant positive spatial autocorrelation at local scales within populations (Table 8.5, Fig. 8.5) generally indicating the occurrence of similar or identical genotypes within populations. Values in other distance classes were negative and not significant. A significant negative correlation over the longest distances implies a slight isolation-by-distance pattern. However, absolute values of the kinship coefficient are very low.

8.4 discussion

The success of invasive species can depend on their ability to adapt to local environmental conditions (Sakai et al., 2001; Prentis et al., 2008). Evo- lution of these adaptations will be promoted by standing genetic variation for ecologically relevant traits (Colautti et al., 2010), as well as new mutations. Previous sequencing of several genes shows that the C. introﬂexus is 86 c. introflexus dispersal and genetic diversity (paper 2)

Figure 8.3: Plot of ∆-K in relation to number of clusters resolved in STRUCTURE analysis. Population structure of gametophytes of Campylopus introﬂexus (39 individuals from dataset B) using the program STRUCTURE. The graph shows the estimated membership for each individual in each cluster for K = 2 (the best K according to ∆-K). The individuals are grouped according to geographical “Regions”. 8.4 discussion 87

Figure 8.4: Principal coordinate analysis (PCoA) of Campylopus introﬂexus — dataset B based on AFLP markers. Geographical "Regions" are labeled by different symbols, "Zones": white – Czech Republic West, grey – Czech Repub- lic East, black – Africa+Poland. 88 c. introflexus dispersal and genetic diversity (paper 2)

Table 8.5: Summary of spatial autocorrelation analysis of 37 individuals of Campy- lopus introflexus based on AFLP data (dataset B without african samples). All pair-wise distances were divided into 6 equifrequent distance classes, and the kinship coefficient for dominant markers was calculated for each distance class. The first distance class refers to the correlation within populations. Maximum distance (km), mean distance (km), estimator of the coefficient after the jackknifing procedure over loci, standard error (SE) and significance of the coefficient estimator after 1000 permutations are provided for each distance class.

Distance class 1 2 3 4 5 6 Max distance [km] within sites 90 158 225 400 710 Mean distance [km] 0 39 123 191 312 625 Kinship coefﬁcient (jacknifed) 0.0474 −0.0129 −0.011 −0.0227 −0.0185 −0.0036 SE 0.0057 0.005 0.0044 0.0058 0.0031 0.0002 P-value < 0.001 0.2927 0.3826 0.031 0.023 < 0.001

Figure 8.5: Results of spatial autocorrelation analysis of 37 individuals of Campy- lopus introflexus (dataset B) based on AFLP data. All pair-wise distances were divided into 6 equifrequent distance classes (Table 8.5). The first distance class refers to the correlation within populations. 8.4 discussion 89 very uniform species without detected variability in both native and invaded areas (Stech and Dohrmann, 2004; Stech and Wagner, 2005). This genetic uniformity can be explained either by high gene flow between populations in native and invaded range or general lack of genetic variation. The question of how variable C. introflexus is and how standing variation is partitioned within and among populations can be better addressed using more highly variable molecular markers such as AFLP’s We detected substantial variability within Europe, which confirms the suitability of AFLP fingerprints for C. introflexus. There is a general assumption that the spreading of C. introflexus is has been from west to east within Europe. Spores of C. introflexus are dispersed by wind, so the major direction of gene flow will be correlated with south-western prevailing winds in Europe (Ferrel, 1856; Muñoz et al., 2004). However this pattern is not detectable through the Czech Republic; the northern and southern populations are grouped together. Polish populations form a genetically separate group relative to Czech populations (Fig. 8.2). This fact could be a consequence of lower gene flow among more distant than nearby regions, especially between south and north regions. We interpret the differentiation between Czech and Polish populations to suggest that the majority of populations in Central Europe is colonized from southern and central Germany, while the northern Polish populations originated from another area. However these hypothesized source populations must be genetically similar, because neither the PCoA nor the STRUCTURE analyses detected an additional genetic cluster.

8.4.1 Distribution mechanisms of C. introﬂexus

Many bryophyte species often spread both vegetatively and sexually (i. e. spores). In species with vegetative as well as sexual reproduction, fine scale genetic structuring provides key information about patterns of colonization and spread, such as the relative importance of clonal growth versus sexual reproduction for population expansion (Ward, 2006). Bryophytes are usually considered to be good dispersers, as suggested by the low levels of geographical differentiation revealed by many phylogeographical studies (e. g. Van der Velde and Bijlsma, 2003) and indirect measurements of intercontinental gene flow (McDaniel and Shaw, 2003; Szövényi et al., 2008; Vanderpoorten et al., 2008). Low levels of endemism and cosmopolitan distributions are likely consequences of long-distance dispersal and establishment (Van Zanten and Pócs, 1981). However, some moss are characterized by restricted dispersal ranges and dispersal limitation at the landscape scale (Snäll et al., 2004; Hutsemekers et al., 2010). High genetic diversity among populations without sharing of clones in the Czech Republic suggests that here is a regional spore rain that forms populations, and sexual reproduction within local populations may not be so important for colonisation of new sites (see a similar pattern in Ramaiya et al., 2010). Nevertheless, genetic variation within populations of C. introflexus was low in comparision to some predominantly clonal species (Pohjamo et al., 2008). More or less equal partitioning of variation between and within populations 90 c. introflexus dispersal and genetic diversity (paper 2)

suggests that sexual reproduction contributes to short- as well as long- dis- tance colonization. This is contrary to the finding that most invasive plants are primarily selfing or asexual (e. g. Husband and Barrett, 1991; Amsellem et al., 2001; Wang et al., 2008). Some plants even switch to higher levels of inbreeding or vegetative reproduction during their invasions (Pellegrin and Hauber, 1999; Amsellem et al., 2001). Our results are in accordance with general expectation that sexual reproduction tends to increase population genetic variation (Hamrick and Godt, 1996). At the regional and zone scales, no geographical signature is revealed in the pattern of AFLP variation among introduced populations of C. introflexus. This result may be mainly attributed to dispersal of diaspores and high gene flow within central Europe. This indicates that sexual reproduction is responsible for spreading populations and it is the dominant reproductive system. Asexual reproduction in combination with random genetic drift reduces levels of genetic diversity and may increase between-population genetic differentiation. The lower level of differentiation found within sterile populations in the Czech Republic may be a consequence of clonality and small population sizes combined with genetic drift. These reproductive features combined with genetic bottlenecks have the potential to severely limit genetic diversity in invasive populations (Zhang et al., 2010).

8.4.2 Genetic variation among and within populations

The estimates of genetic differentiation among populations reflect, among other factors, the amount of gene flow between populations. All forms of data analyses indicated that C. introflexus has preserved a substantial proportion of its genetic variability among populations (FST = 0.54). This differentiation among populations can be attributed to four main factors including mating system, dispersal ability, invasion history and diversifying selection. Limited genetic differentiation among regions and zones indicates significant gene flow in the area, as in another expanding moss species Pogonatum dentatum (Hassel et al., 2005). Almost half of the genetic variability within C. introflexus occurs within populations. This within population variation reflects differences between cushions, since most individual cushions appear to be clonal. The observed global FST value is, in fact, comparable with that found in autogamous angiosperms (Hamrick, Godt et al., 1990) and other bryophytes (e. g. Thingsgaard, 2001; McDaniel and Shaw, 2003; Szövényi et al., 2008; Laenen et al., 2011; Hutsemekers et al., 2010). It is comparable also with other invasive plant species (Prinz et al., 2009). However, not all bryophytes have most of their genetic variability partitioned among populations (compare with Spagnuolo et al., 2007; Pohjamo et al., 2008). In C. introflexus, genetic diversity within populations is similar to that observed in other bryophytes (Cronberg, 2002; Thingsgaard, 2001). The amount of genetic diversity was higher among the western regions as compared to the rest of the Czech Republic. 8.4 discussion 91

8.4.3 Spatial genetic structure in the Czech Republic

AFLP markers displayed a low level of genetic polymorphism among the investigated populations of C. introflexus (the lowest in Poland/Africa, higher in the Czech Republic), but all populations in the Czech Republic have a unique genetic signature. Similarly, high percentage of unique genetic patterns have also been recorded in the populations of other invasive plants (e. g. Walker et al., 2003; Chapman et al., 2004; Durka et al., 2005; Wang et al., 2008). Kolbe et al. (2004) suggested that such high levels of genetic variation in introduced populations can result from multiple introductions. However, genetically distinct and spatially separated populations at the regional scale may be the product of different founding events, or they may result from one initial founding event and the subsequent evolution of populations or subpopulations in partial isolation (Torimaru et al., 2003). There probably has not been enough time to develop new haplotypes in C. introflexus by mutation since start of colonisation the Czech Republic, so the pattern we observed more likely reflects multiple colonizations. Our results show no significant correlation between genetic and geographical distances in majority of distance classes. This supports the hypothesis that genetically differentiated populations are likely the products of multiple, genotypically distinct founder events. Populations in the Czech Republic include a broad sample of genetic types found within Europe; genetic diversity in the Czech Republic appears to represent a random sample of genetic variation in the invaded area overall, derived by independent colonization from disparate source populations. This phenomenon has been observed in the hepatic Frullania asagrayana (Ramaiya et al., 2010) and in Sphagnum species (Sundberg, 2005). Additional indicators of multiple origins of populations in the Czech Republic include the observation that genetic diversity does not increase from west to east, as would be expected if migration through the country were by stepwise migration from a single source. If populations from different geographical locations cluster together in terms of genetic similarities, we might propose multiple origins for each population. On the contrary, when different plants from the same population cluster together, this likely reflects clonal spreading of C. introflexus. The PCoA did not show any significant differentiation of geographic groups in Czech populations (Fig. 8.4). Similarity among populations from eastern and western regions of the Czech Republic in the NJ trees could reflect that some eastern populations may have originated from fertile western populations, or from the same source population. Another explanation is that populations on the eastern border of the invasion area more probably originated from more distant regions of Europe. Sundberg (2005) showed that with increasing isolation of populations, a higher proportion of wind-blown spores would originate from sources farther away than the nearest sources. This could explain the high genetic diversity and lack of genetic-geographic correlation among populations in the Czech Republic. NJ trees constructed from all datasets show low support for individual branches. This probably reflects homoplasy and/or recombination. This may 92 c. introflexus dispersal and genetic diversity (paper 2)

also explain why the relationships among plants constructed from AFLP and isozymes markers are not correlated. Another explanation for this could be the contrasting levels of variability in AFLP marker as opposed to isozymes markers. Most of the genetic diversity among Czech plants of C. introflexus is partitioned among populations, which is consistent with separate origins for many populations. Populations in one region were probably formed multiple times from different sources in western Europe; results from both the NJ and STRUCTURE analyses suggest little recombination among genetic types within individual populations despite movement between sites. The most genetically diversity within C. introflexus is in the western part of the Czech Republic, which borders closely with likely source areas. This may reflect frequent colonization of new localities by spores. There are no significant differences in FST values among the first recorded sites where C. introflexus is known to have occurred in the Czech Republic (Novotný, 1990b), and more recent populations. We found no correlation between genetic diversity and population size, which would imply genetic drift. This is contrary to observations on the invasive plant Suaeda maritima (Prinz et al., 2009).

8.4.4 Conclusions

The invasion process in C. introflexus appears to include the common founding of populations by spores followed by local spreading by vegetative propagules. Relatively low genetic variability within populations of the invaded areas reflects generally low variability of species and/or founder effects in invasive populations. The unique genetic patterns of all populations in the Czech Republic and high level of among and within population variability indicates multiple colonisations events. Our results have to be interpreted with care because of the low number of individuals sampled per population, and the low overall level of overall variability. But the observed patterns suggest that the investigated populations and regions are connected by considerable gene flow, and through long-distance dispersal.

acknowledgments

This project was supported by the Grant Agency of Charles University (project no. 258/2004/B-BIO/PrF), the Ministry of Education, Youth and Sports of the Czech Republic (grant no. MSM0021620828) and institutional support of Masaryk University. We are grateful to members of Shaw lab for valuable comments, in addition to Jonathan A. Shaw for language corrections and Matthew G. Johnson for help with data analysis.

references

Akiyama, H. (1994). “Allozyme variability within and among populations of the epiphytic moss Leucodon (Leucodontaceae: Musci).” American Journal of Botany 81, pp. 1280–1287. 8.4 discussion 93

Amsellem, L., J.L. Noyer and M. Hossaert-McKey (2001). “Evidence for a switch in the reproductive biology of Rubus alceifolius (Rosaceae) towards apomixis, between its native range and its area of introduction.” American Journal of Botany 88, pp. 2243–2251. Biermann, R. and F.J.A. Daniëls (2001). “Vegetationsdynamik im Spergulo- Corynephoretum unter besonderer Berücksichtigung des neophytischen Laub- mooses Campylopus introﬂexus.” Braunschweiger Geobotanische Arbeiten 8, pp. 27–37. Boisselier-Dubayle, M.C. and H. Bischler (1994). “A combination of molecular and morphological characters for delimitation of taxa in European Porella.” Journal of Bryology 18, p. 11. Bonnet, E. and Y. Van de Peer (2002). “zt: a software tool for simple and partial Mantel tests.” Journal of Statistical software 7, pp. 1–12. Chapman, H., B. Robson and M.L. Pearson (2004). “Population genetic structure of a colonising, triploid weed, Hieracium lepidulum.” Heredity 92, 182– 188. Colautti, R.I., C.G. Eckert and S.C.H. Barrett (2010). “Evolutionary constraints on adaptive evolution during range expansion in an invasive plant.” Pro- ceedings of the Royal Society B: Biological Sciences 277, pp. 1799–1806. Cronberg, N. (2002). “Colonization dynamics of the clonal moss Hylocomium splendens on islands in a Baltic land uplift area: reproduction, genet distribution and genetic variation.” Journal of Ecology 90, pp. 925–935. Cummins, H. and R. Wyatt (1981). “Genetic variability in natural populations of the moss Atrichum angustatum.” Bryologist 84, pp. 30–38. Dlugosch, K.M. and I.M. Parker (2008). “Founding events in species invasions: genetic variation, adaptive evolution, and the role of multiple introductions.” Molecular Ecology 17, pp. 431–449. Drake, J.A., H.A. Mooney, F. Di Castri, R.H. Groves, F.J. Kruger, M. Re- jmánek and M. Williamson (1989). Biological invasions: a global perspective. New York, USA: John Wiley and Sons. Duckett, J. G and R. Ligrone (1992). “A survey of diaspore liberation mechanisms and germination patterns in mosses.” Journal of bryology 17, pp. 335– 354. Durka, W., O. Bossdorf, D. Prati and H. Auge (2005). “Molecular evidence for multiple introductions of garlic mustard (Alliaria petiolata, Brassicaceae) to North America.” Molecular Ecology 14, pp. 1697–1706. Ehrich, D. (2006). “AFLPdat: a collection of R functions for convenient han- dling of AFLP data.” Molecular Ecology Notes 6, pp. 603–604. Essl, F. and P.W. Lambdon (2009). “Alien bryophytes and lichens of Europe.” In: Handbook of Alien Species in Europe. Ed. by Daisie. Dordrecht: Springer, pp. 29–41. Essl, F., W. Rabitsch and P. Lambdon (2011). “Bryophytes and lichens.” In: Encyclopedia of biological invasions. Ed. by D. Simberloff and M. Rejmánek. Berkeley: University of California Press, pp. 81–85. Evanno, G., S. Regnaut and J. Goudet (2005). “Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study.” Molecular ecology 14, pp. 2611–2620. 94 c. introflexus dispersal and genetic diversity (paper 2)

Excoffier, L. and H.E.L. Lischer (2010). “Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows.” Molecular Ecology Resources 10, pp. 564–567. Fernandez, C.C., J.R. Shevock, A.N. Glazer and J.N. Thompson (2006). “Cryp- tic species within the cosmopolitan desiccation-tolerant moss Grimmia laevigata.” Proceedings of the National Academy of Sciences of the United States of America 103, pp. 637–642. Ferrel, W. (1856). “An Essay on the Winds and the Currents of the Ocean.” Nashville Journal of Medicine and Surgery 11, pp. 287–301. Frahm, J.P. (1984). “Phytogeography of European Campylopus species.” Pro- ceedings of the Third Meeting of Bryologists from Central and East Europe. Ed. by J. Váˇna.Praha: Univerzita Karlova, pp. 191–212. –(2007). “Campylopus introflexus.” In: Flora of North America North of Mex- ico. Ed. by Flora of North America Editorial Committee. Vol. 27. New York and Oxford, p. 371. url: http://www.efloras.org/florataxon.aspx? flora_id=1&taxon_id=105474 (visited on 09/20/2011). Gower, J.C. (1966). “Some distance properties of latent root and vector methods used in multivariate analysis.” Biometrika 53.3-4, pp. 325–338. Gunnarsson, U., K. Hassel and L. Söderström (2005). “Genetic structure of the endangered peat moss Sphagnum angermanicum in Sweden: A result of historic or contemporary processes?”: The Bryologist 108, pp. 194–203. Gunnarsson, U., A.J. Shaw and M. Lönn (2007). “Local-scale genetic structure in the peatmoss Sphagnum fuscum.” Molecular Ecology 16, pp. 305–312. Hamrick, J.L. and M.J.W. Godt (1996). “Effects of life history traits on genetic diversity in plant species.” Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 351, pp. 1291–1298. Hamrick, J.L., M.J.W. Godt et al. (1990). “Allozyme diversity in plant species.” In: Plant population genetics, breeding, and genetic resources. Ed. by A.H.D. Brown, M.T. Clegg, A.L. Kahler and B.S. Weir. Sunderland: Sinauer Asso- ciates Inc., pp. 43–63. Hardy, O.J. and X. Vekemans (2002). “SPAGeDi: a versatile computer program to analyse spatial genetic structure at the individual or population levels.” Molecular Ecology Notes 2, pp. 618–620. Hardy, O.J., N. Charbonnel, H. Fréville and M. Heuertz (2003). “Microsatel- lite allele sizes: a simple test to assess their significance on genetic differentiation.” Genetics 163, pp. 1467–1482. Hasse, T. (2007). “Campylopus introflexus invasion in a dune grassland: succession, disturbance and relevance of existing plant invader concepts.” Herzogia 20, pp. 305–315. Hassel, K. and L. Söderström (2005). “The expansion of the alien mosses Orthodontium lineare and Campylopus introflexus in Britain and continental Europe.” Journal of the Hattori Botanical Laboratory 97, pp. 183–193. Hassel, K., S.M. Såstad, U. Gunnarsson and L. Söderström (2005). “Genetic variation and structure in the expanding moss Pogonatum dentatum (Poly- trichaceae) in its area of origin and in a recently colonized area.” American journal of botany 92, pp. 1684–1690. Hedenäs, L., T. Herben, H. Rydin and L. Söderström (1989b). “Ecology of the invading moss species Orthodontium lineare in Sweden: substrate pref- 8.4 discussion 95

erence and interactions with other species.” Journal of Bryology 15, pp. 565– 581. Herben, T. (1994). “Local rate of spreading and patch dynamics of an invasive moss species, Orthodontium lineare.” Journal of Bryology 18, pp. 115– 125. Herben, T., H. Rydin and L. Söderström (1991). “Spore establishment probability and the persistence of the fugitive invading moss, Orthodontium lineare - a spatial simulation-model.” Oikos 60, 215–221. Husband, B.C. and S.C.H. Barrett (1991). “Colonization history and population genetic structure of Eichhornia paniculata in Jamaica.” Heredity 66, pp. 287–296. Hutsemekers, V., O.J. Hardy, P. Mardulyn, A.J. Shaw and A. Vanderpoorten (2010). “Macroecological patterns of genetic structure and diversity in the aquatic moss Platyhypnidium riparioides.” New Phytologist 185, pp. 852–864. Jakobsson, M. and N.A. Rosenberg (2007). “CLUMPP: a cluster matching and permutation program for dealing with label switching and multi- modality in analysis of population structure.” Bioinformatics 23, pp. 1801– 1806. Kirschner, J., L. Kirschnerová, J. Štˇepánekand M. Tichý (1994). “Analýza isoenzym ˚uv populaˇcníbiologii rostlin [Analysis of isozymes in plant population biology].” Depon in: BÚ AVCR,ˇ Pr ˚uhonice,Prague. Kolbe, J.J., R.E. Glor, L.R. Schettino, A.C. Lara, A. Larson and J.B. Losos (2004). “Genetic variation increases during biological invasion by a Cuban lizard.” Nature 431, pp. 177–181. Laenen, B., A. Désamoré, N. Devos, A.J. Shaw, J.M. González-Mancebo, M.A. Carine and A. Vanderpoorten (2011). “Macaronesia: a source of hidden genetic diversity for post-glacial recolonization of western Europe in the leafy liverwort Radula lindenbergiana.” Journal of Biogeography 38, pp. 631– 639. Mantel, N. (1967). “The detection of disease clustering and a generalized regression approach.” Cancer research 27.2 Part 1, p. 209. McDaniel, S.F. and A.J. Shaw (2003). “Phylogeographic structure and cryptic speciation in the trans-antarctic moss Pyrrhobryum mnioides.” Evolution 57, pp. 205–215. McDaniel, S.F., A.J. Shaw et al. (2005). “Selective sweeps and intercontinental migration in the cosmopolitan moss Ceratodon purpureus (Hedw.) Brid.” Molecular ecology 14, p. 1121. Mikulášková, E. (2006). “Vývoj rozšíˇrení neofytického mechu Campylopus introflexus v Ceskéˇ republice [Development in distribution of the neophytic moss Campylopus introflexus in the Czech Republic].” Bryonora 38, pp. 1–10. Mikulášková, E., Z. Fajmonová and M. Hájek (2012a). “Invasion of the moss Campylopus introflexus into central European habitats.” Preslia. [under review]. Mikulášková, E., T. Fér and V. Kuˇcabová (2012b). “The effect of different DNA isolation protocols and AFLP fingerprinting optimizations on error rate estimates in the bryophyte Campylopus introflexus.” Lindbergia 351, pp. 7–17. 96 c. introflexus dispersal and genetic diversity (paper 2)

Moravcová, L., P. Pyšek, V. Jarošík, V. Havlícková and P. Zákravský (2010). “Reproductive characteristics of neophytes in the Czech Republic: traits of invasive and non-invasive species.” Preslia 82, pp. 365–390. Muñoz, J., Á.M. Felicísimo, F. Cabezas, A.R. Burgaz and I. Martínez (2004). “Wind as a long-distance dispersal vehicle in the Southern Hemisphere.” Science 304, pp. 1144–1147. Nei, M. (1973). “Analysis of gene diversity in subdivided populations.” Pro- ceedings of the National Academy of Sciences 70, p. 3321. Novotný, I. (1990a). “Campylopus introflexus (Hedw.) Brid. – nový druh mechu v Ceskoslovensku.”ˇ 2. Bryologicko–lichenologické dny v Opavˇe(19.–21.9.1989) – souhrny referát˚u. Ed. by Anonymus. Vol. 4, pp. 4–5. –(1990b). “The moss Campylopus introflexus (Hedw.) Brid. new to Czechoslo- vakia.” Acta Musei Moraviae, Scientiae Naturales 75, pp. 237–238. Pairon, M., B. Petitpierre, M. Campbell, A. Guisan, O. Broennimann, P.V. Baret, A.L. Jacquemart and G. Besnard (2010). “Multiple introductions boosted genetic diversity in the invasive range of black cherry (Prunus serotina; Rosaceae).” Annals of botany 105, pp. 881–890. Peakall, R.O.D. and P.E. Smouse (2006). “GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research.” Molecular Ecology Notes 6, pp. 288–295. Pellegrin, D. and D.P. Hauber (1999). “Isozyme variation among populations of the clonal species, Phragmites australis (Cav.) Trin. ex Steudel.” Aquatic Botany 63.3-4, pp. 241–259. Pfeiffer, T., S. Fritz, M. Stech and W. Frey (2006). “Vegetative reproduction and clonal diversity in Rhytidium rugosum (Rhytidiaceae, Bryopsida) inferred by morpho-anatomical and molecular analyses.” Journal of plant research 119, pp. 125–135. Pohjamo, M., H. Korpelainen and N. Kalinauskait (2008). “Restricted gene flow in the clonal hepatic< i> Trichocolea tomentella in fragmented land- scapes.” Biological Conservation 141, pp. 1204–1217. Prentis, P.J., J.R.U. Wilson, E.E. Dormontt, D.M. Richardson and A.J. Lowe (2008). “Adaptive evolution in invasive species.” Trends in Plant Science 13, pp. 288–294. Prinz, K., K. Weising and I. Hensen (2009). “Genetic structure of coastal and inland populations of the annual halophyte Suaeda maritima (L.) dumort. in Central Europe, inferred from amplified fragment length polymorphism markers.” Plant Biology 11, pp. 812–820. Pritchard, J.K., M. Stephens and P. Donnelly (2000). “Inference of population structure using multilocus genotype data.” Genetics 155, pp. 945–959. Pyšek, P. and D.M. Richardson (2010). “Invasive species, environmental change and management, and health.” Annual Review of Environment and Resources 35, pp. 25–55. Ramaiya, M., M.G. Johnson, B. Shaw, J. Heinrichs, J. Hentschel, M. Von Kon- rat, P.G. Davison and A.J. Shaw (2010). “Morphologically cryptic biological species within the liverwort Frullania asagrayana.” American Journal of Botany 97, pp. 1707–1718. 8.4 discussion 97

Richards, P.W. (1963). “Campylopus introflexus (Hedw.) Brid. and C. polytrichoides De Not. in the British Isles; a preliminary account.” Transactions of the British Bryological Society 4, pp. 404–417. Ruiz, G.M. and J.T. Carlton, eds. (2003). Invasive species: Vectors and management strategies. Washington, D.C.: Island Press. Saitou, N. and M. Nei (1987). “The neighbor-joining method: a new method for reconstructing phylogenetic trees.” Molecular biology and evolution 4, pp. 406–425. Sakai, A.K., F.W. Allendorf, J.S. Holt, D.M. Lodge, J. Molofsky, K.A. With, S. Baughman, R.J. Cabin, J.E. Cohen, N.C. Ellstrand et al. (2001). “The Popu- lation Biology of Invasive Specie.” Annual Review of Ecology and Systematics 32, pp. 305–332. Schlüter, P.M. and S.A. Harris (2006). “Analysis of multilocus fingerprinting data sets containing missing data.” Molecular Ecology Notes 6, pp. 569–572. Selkirk, P.M., M. Skotnicki, K.D. Adam, M.B. Connett, T. Dale, T.W. Joe and J. Armstrong (1997). “Genetic variation in Antarctic populations of the moss Sarconeurum glaciale.” Polar Biology 18, pp. 344–350. Shaw, A.J. (2001). “Biogeographic patterns and cryptic speciation in bryophytes.” Journal of Biogeography 28, pp. 253–261. Simberloff, D. and M. Rejmánek, eds. (2011). Encyclopedia of biological invasions. London: Univiversity of California Press. Snäll, T., J. Fogelqvist, P.J. Ribeiro Jr and M. Lascoux (2004). “Spatial genetic structure in two congeneric epiphytes with different dispersal strategies analysed by three different methods.” Molecular Ecology 13, pp. 2109–2119. Soldán, Z. (1996). “Rozšíˇrení neofytických mech ˚u Campylopus introflexus a Or- thodontium lineare v Ceskéˇ republice [Distribution of Campylopus introflexus a Orthodontium lineare in Czech Republic].” Brynora 18, pp. 10–19. –(1997). “Invazní mechorosty. [Invasive bryophytes].” Zprávy Ceskéˇ botanické spoleˇcnosti 14, pp. 33–39. Soltis, D.E. and P.S. Soltis (1989). Isozymes in plant biology. Ed. by D.E. Soltis and P.S. Soltis. London: Chapman and Hall. Spagnuolo, V., L. Muscariello, S. Cozzolino, R. Castaldo Cobianchi and S. Giordano (2007). “Ubiquitous genetic diversity in ISSR markers between and within populations of the asexually producing moss Pleurochaete squar- rosa.” Plant Ecology 188, pp. 91–101. Sparrius, L.B. and A.M. Kooijman (2011). “Invasiveness of Campylopus introflexus in drift sands depends on nitrogen deposition and soil organic matter.” Applied Vegetation Science 14, pp. 221–229. Stech, M. and J. Dohrmann (2004). “Molecular relationships and biogeography of two Gondwanan Campylopus species, C. pilifer and C. introflexus (Dicranaceae).” Monographs in Systematic Botany from the Missouri Botanical Garden 98, pp. 415–431. Stech, M. and D. Wagner (2005). “Molecular relationships, biogeography, and evolution of Gondwanan Campylopus species (Dicranaceae, Bryopsida).” Taxon 54, pp. 377–382. Stenøien, H.K. and S.M. Såstad (1999). “Genetic structure in three haploid peat mosses (Sphagnum).” Heredity 82, pp. 391–400. 98 c. introflexus dispersal and genetic diversity (paper 2)

Størmer, P. (1958). “Some mosses from the phytogeographical excursion 1-9 through the Armorican massive in 1954.” Revue bryologique et lichénologique 27, pp. 13–16. Sundberg, S. (2005). “Larger capsules enhance short-range spore dispersal in Sphagnum, but what happens further away?”: Oikos 108, pp. 115–124. Szövényi, P., S. Terracciano, M. Ricca, S. Giordano and A.J. Shaw (2008). “Re- cent divergence, intercontinental dispersal and shared polymorphism are shaping the genetic structure of amphi-Atlantic peatmoss populations.” Molecular Ecology 17, pp. 5364–5377. Thingsgaard, K. (2001). “Population structure and genetic diversity of the amphiatlantic haploid peatmoss Sphagnum affine (Sphagnopsida).” Hered- ity 87, pp. 485–496. Torimaru, T., N. Tomaru, N. Nishimura and S. Yamamoto (2003). “Clonal diversity and genetic differentiation in Ilex leucoclada M. patches in an old- growth beech forest.” Molecular ecology 12, pp. 809–818. Van der Velde, M. and R. Bijlsma (2003). “Phylogeography of five Polytrichum species within Europe.” Biological Journal of the Linnean Society 78, pp. 203– 213. Van Zanten, B.O. and T. Pócs (1981). “Distribution and dispersal of bryophytes.” Advances in bryology 1, pp. 479–562. Vanderpoorten, A. and M. Tignon (2000). “Amplified fragments length polymorphism between populations of Amblystegium tenax exposed to contrasting water chemistries.” Journal of Bryology 22, pp. 252–262. Vanderpoorten, A., N. Devos, B. Goffinet, O.J. Hardy and A.J. Shaw (2008). “The barriers to oceanic island radiation in bryophytes: insights from the phylogeography of the moss Grimmia montana.” Journal of biogeography 35, pp. 654–663. Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. Van De Lee, M. Hornes, A. Friters, J. Pot, J. Paleman and M. Kuiper (1995). “AFLP: a new technique for DNA fingerprinting.” Nucleic acids research 23, pp. 4407–4414. Walker, N.F., P.E. Hulme and A.R. Hoelzel (2003). “Population genetics of an invasive species, Heracleum mantegazzianum: implications for the role of life history, demographics and independent introductions.” Molecular Ecology 12, pp. 1747–1756. Wang, T., Y. Su and G. Chen (2008). “Population genetic variation and structure of the invasive weed Mikania micrantha in Southern China: consequences of rapid range expansion.” Journal of Heredity 99, pp. 22–33. Ward, S. (2006). “Genetic analysis of invasive plant populations at different spatial scales.” Biological Invasions 8, pp. 541–552. Wyatt, R., I.J. Odrzykoski and N. Cronberg (2005). “High levels of genetic variation in the haploid leafy liverwort Porella platyphylla from the southeastern United States.” Journal of Bryology 27, 247–252. Yamazaki, T. (1981). “Genetic variabilities in natural population of haploid plant, Conocephalum conicum. I. The amount of heterozygosity.” The Japanese journal of genetics 56, pp. 373–383. Zartman, C.E., S.F. McDaniel and A.J. Shaw (2006). “Experimental habitat fragmentation increases linkage disequilibrium but does not affect genetic 8.4 discussion 99

diversity or population structure in the Amazonian liverwort Radula ﬂac- cida.” Molecular ecology 15, pp. 2305–2315. Zhang, Y.Y.E., D.A.Y. Zhang and S.C.H. Barrett (2010). “Genetic uniformity characterizes the invasive spread of water hyacinth (Eichhornia crassipes), a clonal aquatic plant.” Molecular Ecology 19, pp. 1774–1786.

THEEFFECTOFDIFFERENTDNAISOLATION 9 PROTOCOLSANDAFLPFINGERPRINTING OPTIMIZATIONSONERRORRATEESTIMATESINTHE BRYOPHYTE CAMPYLOPUSINTROFLEXUS

Eva Mikulášková1,2, Tomáš Fér2 and Veronika Kuˇcabová2 1Dept. of Botany and Zoology, Faculty of Science, Masaryk University, CZ-61137 Brno, Czech Republic; [email protected]. 2Dept. of Botany, Faculty of Science, Charles University in Prague, CZ-128 01 Praha 2, Czech Republic. abstract

Amplified fragment length polymorphism (AFLP) has become a standard method for investigating genetic variation in plants. Nevertheless, only a few applications in bryophytes have been published and there is still a need to optimize the method. We optimized DNA isolation and AFLP protocols for Campylopus introflexus (Hedw.) Brid. As DNA quality is crucial for successful AFLP analysis, three different DNA extraction protocols were compared and the Invisorb Plant Mini Kit produced the highest DNA amount (15, 38 and 47 ng µl−1 for dry, fresh and new cultivated tissues, respectively) and purity. Newly grown stems gave the purest DNA (absorbance at the wavelengths of 260 nm and 280 nm was 1.86). However, banding patterns obtained from dry herbarium specimens (up to two years old) corresponded with those from fresh material at a similarity level of 94.71 %. The replicability of AFLP profiles was not dependent on the way plants were stored. We compared commercial kits and the influence of modifications of the protocols for obtaining reliable results. We tested the reproducibility of AFLP fingerprints produced by the final optimized protocol. An increased amount of restriction enzymes and prolonged restriction and ligation of up to 10 h were the most important modifications for improving results. The modified protocol was applied to 30 samples and four selective primer combinations, and gave an average genotyping error rate of 0.0451. keywords

AFLP optimization, bryophytes, Campylopus introﬂexus, DNA isolation, error rate assessment, PCR

9.1 introduction

Ampliﬁed fragment length polymorphisms (AFLPs) (Vos et al., 1995), which are polymerase chain reaction (PCR) based markers for the rapid screening of genetic diversity, have become a standard method for population genetics

101 102 molecular methods optimalization (paper 3)

in plants. AFLPs allow rapid generation of hundreds of highly reproducible DNA markers in almost any target organism. Endonuclease digestion of total genomic DNA followed by selective PCR amplifications and electrophoresis of a subset of the fragments results in a unique fingerprint for each genetic individual (Meudt and Clarke, 2007; Mueller and Wolfenbarger, 1999). AFLP consists of restriction, ligation and two rounds of PCR. First, DNA is cut with two restriction enzymes, a rare cutter (e. g. EcoRI) and a frequent cutter (e. g. MseI). Second, adaptors with known sequences are ligated to the ends of DNA fragments to generate a DNA template for PCR amplification. The number of fragments is reduced by using adaptor-specific primers that amplify only fragments with both EcoRI and MseI ends. Adding one to three selective nucleotide(s) to these primers is an accurate and efficient way of selecting a specific set of restriction fragments for amplification. Specific selection is mostly achieved by two consecutive PCRs. In the first step (pream- plification) the restricted/ligated DNA is amplified with primers, both of which have a single selective nucleotide. The second PCR (selective amplification) uses both primers with two or three selective nucleotides (Karp et al., 1996; Blears et al., 1998; Weising et al., 2005; Meudt and Clarke, 2007). AFLP is especially suitable when studying intraspecific polymorphisms or relationships among closely related organisms (Karp et al., 1996; Arens et al., 1998). It requires no prior molecular knowledge of the target organism (Meudt and Clarke, 2007). AFLP is mostly insensitive to the starting DNA concentration (Vos et al., 1995; Rosendahl and Taylor, 1997) but requires relatively pure DNA without contamination from proteins and secondary compounds (Jones et al., 1997; Meudt and Clarke, 2007). When optimized it is generally highly replicable with a genotyping error rate of less than 2 %(Vos et al., 1995; Arens et al., 1998; Mueller and Wolfenbarger, 1999; Bonin et al., 2004). However, AFLP is a dominant marker that does not allow homo- and heterozygotes to be distinguished. Fingerprints are scored as a binary matrix (presence/absence of bands). Problems associated with possibly subjective data scoring are largely overcome with a large number of loci (Bonin et al., 2004). Error rates were estimated in few of the studies using AFLP. But, for the numbers to be useful, it is necessary to have reference ranges for error rates, especially when publishing surprising or controversial results (Bonin et al., 2004). Publishing of error rates in bryophyte studies has been very rare (Fernandez et al., 2006); therefore estimates of other reference ranges are useful. Whereas many studies dealing with genetic variation in higher plants have been published (Chauhan et al., 2004; Mahmud et al., 2007; Arens et al., 1998; Pelser et al., 2003; Schönswetter et al., 2004), the use of AFLP in bryophytes has been limited (Vanderpoorten and Tignon, 2000; Pfeiffer et al., 2006). However, AFLPs have been proved sufficiently polymorphic at the intra-specific level in bryophytes for studies of population genetic processes (Snäll et al., 2004; Zartman et al., 2006), cryptic speciation (Fernandez et al., 2006), clone identification (Pfeiffer et al., 2006), species differentiation (Rown- tree et al., 2010) or linkage mapping (McDaniel et al., 2007). Several studies have also reported that bryophytes require special modification of DNA isolation protocols (Schlink and Reski, 2002; Fernandez et al., 2006) because the 9.2 methods 103 cells of mosses contain a lot of secondary metabolites (Mittmann et al., 2007; Xie and Lou, 2009). Although the extraction process influences the quality of isolated DNA (which is one of the most important factors necessary for successful and reliable AFLP fingerprinting), no study on bryophytes has dealt with this aspect. AFLP is still not a standard method in bryophytes and no essential modifications of the standard protocol or suggestions for special treatment have been published to date. Therefore it is still necessary to optimize the original protocol (Vos et al., 1995) for each newly investigated species. In this study we optimized the DNA isolation and AFLP protocols for Campylopus introflexus (Hedw.) Brid.In particular, we aimed to (a) compare different DNA extraction protocols with respect to obtaining a sufficient quantity and quality of DNA for AFLP, and (b) test the reproducibility of AFLP fingerprints after different modifications to the protocol.

9.2 methods

9.2.1 Sample collection

Campylopus introflexus is one of the most aggressively invasive mosses in Eu- rope, originating from the Southern Hemisphere (Frahm, 1984). During the last few decades it spread from western Europe to the north and east, and in some habitats it has became a dominant (Hassel and Söderström, 2005; Daisie, 2009). For the optimization work within this study we used samples from the Czech Republic (16 populations), Poland (1 population) and South Africa (1 population) (Table 9.1). From five populations we sampled more than one cushion. Samples were obtained from all the usual habitats of C. introflexus (peatbogs, spruce and pine forests, sand dunes). We confirmed that in four populations individual cushions (to 3 × 3 cm) contain only one AFLP genotype. We analysed the upper parts (1–3 cm) of 3–10 leafy stems per cushion from a total of 30 C. introflexus cushions. Preliminary analyses showed that all stems from one cushion are genetically identical. As the starting material for DNA extraction, gametophyte tissue in three different conditions was used: (a) dry stems from herbarium specimens, max. 2 years old, (b) fresh stems collected from the field, (c) new stems from ex situ cultivation on sterile soil (1–3 months at 22 °C).

9.2.2 DNA extraction

Brown tissue and inorganic matter were manually removed from all samples. Samples were washed several times in distilled water. The moss tissue (gametophytes) was used dried, wet or rehydrated in distilled water. About 100–800 mg of clean tissue was homogenized 1) in a mixer mill MM-200 (Retsch) using microcentrifuge tubes with 3mm glass beads, then ground to a ﬁne powder (in the case of dry gametophytes), or 2) using a grinding mortar with a pestle that was placed in liquid nitrogen and then ground 104 molecular methods optimalization (paper 3)

Table 9.1: Collectionsites for experimental material of C. introﬂexus. Code Locality GPS coordinates [WGS-84] Afrika South Africa, Cape Town: NP Table Mountain 16°65’51.2”S, 14°30’26”E Bork Czech Republic, Sobˇeslav: Nature reserve Borkovická blata 49°23’72.2”N, 14°62’58.3”E Bor Czech Republic, Skuteˇc:Nature reserve Maštale 49°81’68.6”N, 16°14’19.2”E Gerl Czech Republic, PancíˇrMt.: Gerl ˚uvpotok I. zone of NP Šumava 49°16’72”N, 13°29’91”E Hrebec Czech Republic, Svitavy: Hˇrebeˇcvillage 49°72’77.2”N, 16°57’57.8”E Jilov Czech Republic, Jílové u Prahy: Petrov village 49°89’30.4”N, 14°46’3”E Kasper Czech Republic, Kašperské hory 49°15’70.8”N, 13°55’60”E Kralov Czech Republic, Prachatice: Nature reserve Kralovické louky 49°00’49.4”N, 14°08’36.4”E Kyjov Czech Republic, Kyjov 50°91’91.1”N, 14°44’44.7”E NBrunst Czech Republic, PancíˇrMt.: Nature reserve Nový Brunst 49°17’8”N, 13°27’13”E Osika Czech Republic, Albeˇr 49°02’86.7”N, 15°15’49.7”E Osli Czech Republic, Proseˇc:Nature reserve Maštale 49°81’68.6”N, 16°14’19.2”E PL dun Poland, Smołdzino: Słowi´nskiPark Narodowy 54°74’81”N, 17°40’34”E Popice Czech Republic, Jihlava: Popice village 49°34’9.5”N, 15°53’80”E Raje Czech Republic, Tˇreboˇn:Nature reserve V rájích 48°98’60.8”N, 14°70’91.8”E Tisina Czech Republic, Tachovská Hut’: Tišina Mt. 49°93’18.3”N, 12°55’5.8”E Valmez Czech Republic, Hostašovice 49°50’53.3”N, 17°99’12.2”E Zvonk Czech Republic, Pˇrední Zvonková: Racín village 48°70’86.7”N, 14°04’46.4”E

to a fine powder manually (in the case of wet or rehydrated gametophytes). Three extraction protocols were tested: a) CTAB extraction protocol (Doyle, 1990), b) DNeasy Plant Mini Kit (Qiagen), and c) Invisorb Spin Plant Mini Kit (Invitek) with modifications described below. In the CTAB extraction protocol (Doyle, 1990), samples were incubated in 700 µl of extraction buffer with 0.2% beta-mercaptoethanol for 30 min at 60 °C with constant shaking, with the addition of 0.02 g PVP (MW 40000) in the first few minutes of shaking. After that 500 µl of chloroform:isoamylalcohol (24:1) was added to the liquid phase of each sample than several times in- verted and incubated for 5 min at room temperature. This step was repeated twice. Samples were centrifuged for 6 min at 13500 rpm and the supernatant was precipitated by 500 µl isopropanol for 30 min at −20 °C. Pellets were washed in 96% and then 70% ethanol, dried down, and resuspended in 50 µl 1 × TE buffer. For extraction using the Invisorb Kit (Invitek) the manufacturer’s protocol was followed. Ten µl proteinase K (10 mg ml−1) was added to the mixture in the lysis step. The final DNA product was eluted in 50 µl pre-heated Elution Buffer D and incubated for 45 min. Extraction using the DNeasy Plant Mini Kit (Qiagen) followed the manufacturer’s protocol. Extracted DNA was visualized on a 0.8% TAE agarose gel and the DNA concentration was measured photometrically using a BioPhotometer (Eppen- dorf) (as absorbance of UV light of wavelength of 260 and 280 nm [= A260, A280]). DNA concentration was adjusted to 20 ng µl−1. Purity was expressed as the ratio of A260/A280. Clean DNA should have a ratio between 1.8 and 2.

9.2.3 AFLP ﬁngerprinting

The AFLP methods largely followed the procedures described by Vos et al. (1995). Commercial AFLP kits (Invitrogen) were used for the restriction, lig- 9.2 methods 105 ation and pre-amplification steps in order to ensure standard concentrations and conditions across samples. These kits have successfully been used repeatedly to generate AFLP fingerprints in vascular plants and fungi (Bless et al., 2006; Prohens et al., 2006; Mahmud et al., 2007; Bory et al., 2008). The AFLP Core Plant Reagent Kit I (Invitrogen) was used for restriction and ligation. The original protocol (Invitrogen (AFLP® Core Reagent Kit )) was followed with a few modifications which are routinely used in our laboratory: incubation of the restriction mixture for 15 min at 70 °C after restriction was eliminated, ligation was incubated at 37 °C, and the ligation mixture was not diluted with TE buffer. As the plant samples were very small all steps were scaled down to a final volume of restriction mix before incubation of 10 µl. Complete restriction of DNA was tested on 1.8% TBE agarose gels. Pre-amplification was carried out using the AFLP PreAmp Primer Mix I (Invitrogen) following the manufacturer’s protocol (Invitrogen (AFLP® In- struction Manual )). Whole reaction was made in ten times reduced volume (final volume of pre-amplification mix before PCR was 5.1 µl), in order to save material. Pre-selective PCR started with an initial step of 72 °C for 2 min, followed by 20 cycles of 10 s at 94 °C, 30 s at 56 °C and 120 s at 72 °C. Final elongation was carried out at 60 °C for 30 min. The product was visualized on 1% TBE agarose gel after being diluted ten times with ddH2O. A total of 2.3 µl of pre-amplified DNA was added to the selective amplification premix (5.1 µl ddH2O, 1 µl 10× polymerase buffer, 0.2 mM dNTP, 0.5 pmol EcoRI primer, 0.5 pmol MseI primer, 0.2 U DNA polymerase). Selec- tive PCR amplification started at 92 °C for 2 min, 65 °C for 30 s and 72 °C for 2 min. A touchdown protocol was applied in the following eight cycles of 1 s at 94 °C, 30 s at 64 °C (1 °C decrease each cycle), and 60 s at 72 °C. This was followed by 23 cycles of 1 s at 94 °C, 30 s at 56 °C and 120 s at 72 °C. Final elongation took place at 60 °C for 30 min. For all PCR amplifications a Touchgene gradient (Techne) thermocycler was used and the ramping time was lowered to 0.9 °C s−1. Two precipitations produced the final purification. First, PCR products with 1 µl of sodium acetate and 25 µl of 96% ethanol were chilled for 20 min at −20 °C. Precipitated products were spun at 4 °C for 30 min at 12500 rpm and the supernatant was discarded. Second, 100 µl of 70% ethanol was added and samples were spun at 4 °C for 5 min at 12500 rpm. Purified products were desiccated at 65 °C for 10 min. Just before the products were run on the sequencer, 10 µl of the mixture HiDi formamide: GeneScan-500 Rox (20:1, Applied Biosystems) was added to each sample. Fragment analysis was performed on an ABI 3100 Avant automated sequencer (Applied Biosystems). Raw data were analysed and scored using GeneMarker ver. 1.8 (SoftGenetics). The protocol described above gave very inconsistent results that were not readily scoreable. In the next step we therefore tested a few modifications of the standard protocol to improve resolution of banding patterns. Some modifications were succesfully used in previous AFLP studies (Snäll et al., 2004; Pfeiffer et al., 2006; Rowntree et al., 2010), but explicit tests of how they influenced the results were not described. The following modifications to the AFLP protocol were subsequently used during optimization and only the best results (assessed by PCR success and after that by the most distinct 106 molecular methods optimalization (paper 3)

AFLP banding patterns) from a particular modiﬁcation step were included in the subsequent testing:

1. Denaturation time increased to 10 seconds in all preampliﬁcation and selective ampliﬁcation steps.

2. Different DNA polymerases (REDTaq DNA Polymerase (Sigma), Jump- Start REDTaq DNA Polymerase (Sigma), Immolase DNA Polymerase (Bioline).

3. Amount of template DNA for restriction (50, 100, 200, 600 and 1000 ng)

4. Amount of mixture (AFLP Core Plant Reagent Kit I) of restriction enzymes EcoRI/MseI(1.0 U of each – amount in original protocol, 1.25 U of each – increased amount, both in a 10 µl restriction mixture).

5. Duration of restriction and ligation (4, 6 and 10 h)

6. Dilution of the pre-ampliﬁcation product (1 : 9.7, 1 : 6.8, 1 : 4)

After optimization, 36 different primer combinations were investigated using the optimized protocol. Finally, four primer combinations that produced polymorphic and evenly distributed fragments in the range of 50–500 bp were selected for the ﬁnal analysis of 30 samples and for error rate assessment.

9.2.4 Assessment of AFLP error rate

Distinct polymorphic peaks in the 50- to 500-bp range were scored as present or absent, and the only characters included in the analysis were those for which unambiguous determinations could be made in all electropherograms. We followed all the suggestions made by Bonin et al. (2004) to minimize error rate. We assessed the reproducibility of the AFLP reactions by repeating the entire AFLP procedure on 30 individuals following the optimized protocol. We estimated the genotyping error rate (Bonin et al., 2004) after scoring by comparing the 1/0 matrices obtained for the duplicated samples. Differences detected at this point could be due either to the technical work, and/or to the subjectivity introduced during the scoring process. Samples with large intensity differences for some peaks are often erroneously scored when using automated peak detection. In weak samples it is necessary to correct automatic detection manually, otherwise the genotyping error rate increases considerably. The error rate was estimated by dividing the number of differences by the total number of comparisons among all pairs. This was done separately for each primer combination.

9.3 results

9.3.1 DNA extraction

DNA extraction protocols gave different results for the amount of DNA obtained and its purity (Table 9.2). The cheapest CTAB method yielded, on 9.3 results 107

Table 9.2: Comparison among extraction approaches in terms of the average concentration of DNA obtained, (±SE [standard error of the mean]), minimum and maximum value (ng µl−1 ) and average purity measured as the ratio of absorbance (A260/A280). Extraction protocol Source of DNA Dry stems Fresh stems New stems average (±SE) min/max purity average min/max purity average min/max purity CTAB 74.04 ± 50.86 7/180 1.58 ± 0.4 31 ± 36.77 5/57 1.65 ± 0.45 40.29 ± 34.72 10/80 1.85 ± 0.39 DNeasy Kit 3.67 ± 1.53 2/5 1.3 ± 0.38 0.33 ± 1.53 0/2 –––– Invisorb Kit 15.3 ± 8.96 3/70 1.79 ± 0.25 38.76 ± 46.73 5/224 1.77 ± 0.18 47.64 ± 51.51 7/280 1.86 ± 0.04 average, 74, 31 and 40 ng µl−1 for dry, fresh and new stems, respectively. However, the amount of DNA obtained varied widely (from 7 to 180 ng µl−1) and the purity was inappropriate for subsequent AFLP analysis in most of the samples. The DNeasy Kit gave very low DNA concentrations (about 3.6 ng µl−1 on average), with variable purity. Such low amounts of DNA were insufﬁcient for AFLP. The Invisorb Kit gave most stable results of all the protocols tested and yielded, on average, 15, 38 and 47 ng µl−1 for dry, fresh and new stems, respectively. The amount of obtained DNA varied among samples but was sufﬁcient for subsequent AFLP analysis in most cases. The highest DNA purity (A260/A280 ratio 1.86) was achieved when new stems were used as the starting material for extraction. Dry herbarium specimens were also suitable for use (after soaking in distilled water for 8 hours) but the amount of DNA was considerably lower (15 ng µl−1 on average).

9.3.2 AFLP ﬁngerprinting

Good quality starting DNA is one of the most important prerequisites for successful AFLP analysis. For that reason only DNA extracted using the standard CTAB protocol and Invisorb Kit was used. For the testing alterna- tive protocols the source DNA was extracted from dry stems (in the case of CTAB extraction) and from fresh and new stems (Invisorb Kit). There were no differences between fresh and new stems from Invisorb Kit extractions during monitoring the impact of modifications in AFLP protocol. During successive optimization steps, CTAB-extracted DNA always gave worse (less clear) banding patterns (or non-amplification during PCR) than Invisorb Kit-extracted DNA. Thus for the final optimized AFLP protocol we used the Invisorb Kit-extracted DNA. The standard (unoptimized) protocol (Methods) produced unpredictable results for about 50 % of the samples with an error rate of about 0.5. Consistent results (clear scorable banding patterns) were only achieved by incorporation of all six additional optimization steps (Table 9.3). First, extension of the denaturation step (94 °C) in the pre-amplification PCR to 10 s had no effect on successful PCR amplification. Second, the use of REDTaq and JumpStart Taq DNA polymerases improved the analysis; RED- Taq gave slightly sharper bands for CTAB-extracted samples and JumpStart Taq for Invisorb Kit-extracted samples. Third, using template DNA between 100 and 200 ng positively influenced the results. Fourth, increased amounts of restriction enzymes during the restriction step (EcoRI and MseI both at 1.25 U in a 10 µl restriction mixture) appeared to be crucial and markedly 108 molecular methods optimalization (paper 3)

improved the results. Fifth, prolonging the restriction and ligation to 6 or 10 h (for each step) (the latter being better) again improved the results. Sixth, the use of less diluted pre-amplification products (1 : 4) gave better results. The optimized protocol is therefore as follows (Table 9.3): Invisorb Kit- extracted DNA, extension of denaturation time in pre-amplification PCR to 10 s, use of JumpStart Taq polymerase (0.1 µl) and 100 ng of template DNA, increased amount of restriction enzymes EcoRI(0.5 µl) and MseI(0.5 µl), prolonged restriction and ligation steps to 10 h each, and decreased dilution of pre-amplification products to 1:4. This optimized protocol was used to test the utility of 36 primer combinations and for reproducibility tests with error rate assessment. Nine primer combinations (EcoRI-ACA / MseI-CAT, EcoRI-ACA / MseI- CTC, EcoRI-AAG / MseI-CAA, EcoRI-AGG / MseI-CAA, EcoRI-AAG / MseI- CTA, EcoRI-AAC / MseI-CAA, EcoRI-AGC / MseI-CAC, EcoRI-AAC / MseI- CTG, EcoRI-AAC / MseI-CTT) appeared to yield polymorphic bands in our dataset (Table 9.1). Only EcoRI-ACA / MseI-CAT, EcoRI-AAG / MseI-CAA, EcoRI-AAC / MseI-CAA and EcoRI-AGG / MseI-CAA amplified evenly distributed and easily scorable DNA fragments over the range of 50–500 bp. These four primer combinations provided 253 scorable fragments including 189 polymorphic markers, and were thus chosen for analysis of the whole data set. The error rate assessment showed high reproducibility of AFLP profiles (Fig. 9.1). A total of 2372 comparisons were made and 107 differences in band patterns were found (Table 9.4). The genotyping error rate for individual primer combinations varied from 0.038 (EcoRI-ACA/MseI-CAT) to 0.054 (EcoRI-AGG/MseI-CAA), and the average for all samples was 0.0451. All comparisons were made on AFLP profiles from plants with same storage history and using the same type of DNA isolation. We standardized storage history in these analyses because repeatability of AFLP profiles is affected by whether DNA was extracted from fresh, rehydrated, or dry plants. This was tested on DNA isolated from same cushion stored as herbarium specimens, as field-grown, and as cultivated tissues (Fig. 9.2, Table 9.5). The genotyping error rates when comparing AFLP profiles in these cases varied from 0.033 to 0.097. There seemed to be some consistent differences in banding patterns between plants with different storage histories but sample sizes were insuf- ficient to determine if these were significant. The genotyping error rate for herbarium specimens was 0.059, for fresh field-collected plants 0.058, and for cultivated tissue 0.043. These values do not differ from estimates of genotyping error rates for the whole set of populations.

9.4 discussion

9.4.1 Optimization of DNA extraction

Our results demonstrate the importance of selecting the appropriate DNA extraction protocol for bryophyte species. Both the quantity and purity of the DNA differed markedly between the extraction protocols tested in this study. The best scorable AFLP proﬁles were obtained using the Invisorb Plant Mini 9.4 discussion 109

Figure 9.1: Polymorphic fragments over 70–370-bp range for six different pairs of duplicated samples (Kasper, Valmez_B, Valmez_C, Tisina, PL_dun, Osika), primer EcoRI-ACA/ MseI-CAT (Genographer). Major polymorphic fragments are marked. 110 molecular methods optimalization (paper 3)

Figure 9.2: Banding pattern over 50–500-bp range of primer combination EcoRI- AAC/ MseI-CAA for one cushion (population code Bork) – DNA isolation by 1) Invisorb from plants from cultivation, 2) Invisorb from fresh plant from ﬁeld, 3) Invisorb from herbarium speciments, 4) CTAB from herbarium speciments. 9.4 discussion 111

Table 9.3: Modifications to the AFLP protocol (tested in the same order as in the table, with successful steps added to subsequent analyses; order of tested modifications correspond with descriptions in Methods). Effects were assessed by PCR success and after that, by distinctness of the AFLP banding pattern: no effect – the results were not different from unmodified protocol, + modification produced more distinct patterns than unmodified protocol, ++ the patterns after modification were the most distinct, – the results were ambiguous.

Extraction protocol + included modifications + → →testing modified step CTAB Invisorb (dry (fresh stems) stems) Source protocol from Methods (SP) – – 1. SP increasing time of all denaturation no effect no effect steps during PCR to 10 s 2. SP + denaturation 10 s polymerases JumpStart 0.1 µl + ++ JumpStart 0.2 µl + + REDTaq 0.1 µl ++ + Immolase 0.1 µl not tested no effect 3. SP + denaturation 10 s + JumpStart 0.1 µl more template DNA 50 ng + + for restriction 100 ng + ++ 200 ng ++ + 600 ng − + 1000 ng − − 4. SP + denaturation 10 s + JumpStart 0.1 µl increased amount of restriction − ++ + 100 ng template DNA enzymes EcoRI(1.25 U) and MseI (1.25 U) 5. SP + denaturation 10 s + JumpStart 0.1 µl changes in duration 4 hours − − + 100 ng template DNA + EcoRI(0.1 µl ) of restriction and 6 hours − + 0 5 and MseI( . µl ) ligation 10 hours − ++ 6. SP + denaturation 10 s + JumpStart 0.1 µl decreasing dilution 1 : 9.7 no effect no effect + 100 ng template DNA + EcoRI(0.5 µl ) of product after 1 : 6.8 no effect no effect 0 5 10 and MseI( . µl ) + hod to r/l pre-amplification 1 : 4 no effect + SP + denaturation 10 s + JumpStart0.1 µl+ 100 ng template DNA + EcoRI (0.5 µl) and MseI(0.5 µl) + 10 hod to r/l + dilution 1 : 4 = final protocol

Table 9.4: Summary of count of error rate. Primer combination No. of tested No. of assessed No. of Error rate pairs of bins non-replicated replicated bins samples EcoRI-ACA/MseI-CAT 19 719 28 0.0389 EcoRI-AGG/ MseI-CAA 10 442 24 0.0543 EcoRI-AGG/ MseI-CAA 4 63 3 0.0476 EcoRI-AAC/ MseI-CAA 25 1148 52 0.0453 Total 58 2372 107 0.0451 112 molecular methods optimalization (paper 3)

Table 9.5: Percentage of polymorphic markers among plants from the same cushion depending on the way the specimen was stored (H – herbarium speci- ment, C – cultivated plants, F – fresh plants from ﬁeld; primer combination EcoRI/MseI). ACA/CAT AGG/CAA AAC/CAA AAG/CAA scorable fragments 74 41 79 59 H–F 4.05 % 4.88 % 3.80 % 5.08 % H–C 5.41 % 7.32 % 5.06 % 3.39 % F–C 4.05 % 7.32 % 3.80 % 5.08 % F–C–H 6.76 % 9.76 % 6.33 % 6.78 % replicated F 4.05 % 7.32 % 5.06 % 6.78 % replicated H 5.41 % 4.88 % 6.33 % 6.78 % replicated C 4.05 % 4.88 % 5.06 % 3.39 %

Kit. Isolation of bryophyte DNA using different kits is common (Snäll et al., 2004; Pfeiffer et al., 2006), but extraction yields are rarely reported. Several previous studies have proposed modifications to the standard extraction protocols (Schlink and Reski, 2002; Fernandez et al., 2006). Different methods yielded 20–30 ng per sample (Physcomitriella patens; Schlink and Reski 2002; Mittmann et al. 2007), 50 ng (Ceratodon purpureus, Mittmann et al. 2007), or up to 400 ng (Grimmia laegivata; Fernandez et al. 2006) using the CTAB protocol, with modifications that greatly increased the success of the PCR. In our work we obtained more than 1000 ng of DNA per sample using the CTAB protocol but the purity was lower than for kit-extracted samples. Quality of DNA from CTAB isolation could be increased by purification (Rowntree et al., 2010). As demonstrated here, purity is one of the most important factors influencing the success and reliability of AFLPs. Unfortunately, purity of DNA samples used for other studies has rarely been described so it is dif- ficult to compare our results. Interestingly, only Invisorb-extracted samples were usable in our study, whereas the DNeasy Plant Mini Kit produced very low amounts of DNA in all samples. Thus, although this extraction kit is frequently and successfully used for many vascular plants species (Weising et al., 2005) it might not be appropriate for some bryophyte species. On the other hand, Werner et al. (2002) successfully used this kit for 17 bryophyta species. Other extraction methods have been used for mosses very rarely (NaOH extraction or direct amplification – Werner et al. 2002; Xin et al. 2003; Mittmann et al. 2007). We also tested the ability to extract DNA from dry herbarium specimens and the quality of the DNA obtained for AFLP fingerprinting. Lower yields might be caused by differences in the amounts of secondary compounds, intraand extracellular water content, and/or DNA degradation. However, DNA from dry and rehydrated stems is suitable for AFLP analysis. This is consistent with the results of Fernandez et al. (2006) who found that even old herbarium specimens of Grimmia laevigata were suitable for high quality DNA extraction. Werner et al. (2002) and Pedersen et al. (2006) successfully used herbarium specimens of several bryophyte species up to even 20 years 9.4 discussion 113 old to produce DNA sufficient for PCR, but they used another technique to extract DNA (CTAB protocol, NAOH extraction, another commercial kits). This provides an advantage for bryophytes over vascular plants, for which the use of fresh or silica gel-dried leaves is recommended in order to obtain DNA with sufficient quality for AFLP (Weising et al., 2005) and for which herbarium specimens are often unsuitable because of DNA degradation. It is appropriate to use DNA from specimens stored prior to DNA extraction in the same way, because of increasing genotyping error rates from samples stored in different manners. The differences could reflect the occurrence of faint bands, especially from dried material, resulting from partial DNA degradation. However, the differences (genotyping error rate) in AFLP profiles for plants stored in different ways were similar to differences from replicated samples stored in the same way. No manner of storage is much better or worse than another in terms of repeatability. Another source of artifactual variation in bryophytes could be from surface contamination of fresh (not in vitro cultivated) gametophytes by fungi or other organisms. Although all stems were thoroughly washed with distilled water prior to DNA extraction some remnant contamination could theoretically influence the AFLP results. That this level of contamination is not a real problem was demonstrated by the high correspondence between AFLP profiles from washed and unwashed stems (about 95 %; Fernandez et al. 2006). Potential endophytic and epiphytic or other surface contaminants (fungi, algae) have 100× smaller genomes than bryophyta and therefore their preence s in samples is unlikely to induce significant problems. Several studies reported the need for in vitro stem cultivation for obtaining DNA suitable for PCR (Amblystegium tenax, Vanderpoorten and Tignon 2000; Physcomitrella patens, Schlink and Reski 2002). In our study higher yields were obtained from cultivated plants than from fresh field-collected plants. Moreover, cultivated plants had a number of advantages: clean new parts without old dead leaves, neither soil nor other surface contamination on leaves, easier homogenization of young growing parts (upper 1–2 cm).

9.4.2 AFLP ﬁngerprinting

AFLP have not been used commonly for systematic or population studies in bryophytes. All the studies published to date followed the original AFLP protocol (Vos et al., 1995) with slight modifications, or used commercially available kits. Fernandez et al. (2006) used the Applied Biosystems protocol (AFLP Plant Mapping Kit) without fundamental modifications. Vander- poorten and Tignon (2000) used the AFLP Analysis system I (Life Technolo- gies). We used the AFLP Core Plant Reagent Kit and AFLP Pre-Amp Primer Mix I (both Invitrogen) in our study and tested the impact of substantial modifications. There were no differences in the AFLP analyses using Invisorb-extracted DNA from rehydrated versus new (in vitro cultivated) stems. The first modification (increased denaturation time in pre-amplification and selective amplification PCR cycles) helped to stabilize subsequent steps. Similar protocols were used in several previous studies (Pfeiffer et al., 2006; Rowntree et al., 114 molecular methods optimalization (paper 3)

2010), but without comparisons across methodologies. The selection of Taq polymerase also seemed to be an important factor in our study; the best results were achieved with hot start polymerase (JumpStart, Sigma). This is not surprising since hot start polymerases generally perform better. The amount of template DNA for endonuclease restriction was optimized to 100 ng. This is a low level relative to kitmanufacturer recommendation. DNA quantities in previous moss studies are highly variable, from 10 ng (Snäll et al., 2004) to 500 ng (Pfeiffer et al., 2006). Lower quantities gave inconsistent results, whereas a higher amount of DNA probably led to imperfect restriction, producing results with a background smear. The next improvement was an increase in the amount of restriction endonucleases. Bless et al. (2006) used the same AFLP kit as in our study (Invitrogen) and also had to increase enzyme (EcoRI/MseI) amounts to produce good results. Similarly, Pfeiffer et al. (2006) had to increase enzyme amounts to improve results in their study of the bryophytes, Hypopterygium tamarisci and Rhytidium rugosum. It has repeatedly been reported that lower amounts of restriction enzymes are sufficient when the template DNA is clean, however (Weising et al., 2005). Complete but specific restriction and ligation are essential for the whole AFLP analysis. Therefore we tested the effect of a prolonged duration of both restriction and ligation (up to 10 hours each) to ensure their completeness. This is a frequently used modification of the general AFLP protocol (Snäll et al., 2004; Pfeiffer et al., 2006; Zartman et al., 2006) that contributes to the stability of reaction. The substantial improvement that we observed might be due to the relative sensitivity of restriction enzymes to impurities (secondary metabolites, etc.) present in the DNA solution (Weising et al., 2005), and the increased amounts of enzymes helped to stabilize the restriction step without the occurrence of non-specific fragments. Although the duration of both the restriction and ligation steps was markedly prolonged in our experiments we found no non-specific restriction products but rather very good reproducibility of the results. Zartman et al. (2006) also reported that prolonging the duration of the restriction step was necessary for reliable results. The results obtained from the optimized protocol could be used for population studies in a broad range of mosses and liverworts. Both UPGMA (unweighted pair group method with arithmetic mean) and NJ (neighbour- joining) cluster trees were constructed from similarity matrices by using Jac- card’s similarity coefficient in FAMD software (Schlüter and Harris, 2006). However, clustering of samples from different primer combinations of the scored AFLP fragments was not congruent. This may be because genetic variation of C. introflexus was limited and so various primer combinations produced seemingly different banding patterns but the differences were likely not significant.

9.4.3 Reproducibility tests

Every genetic data set includes some erroneous genotypes that can cause problems in the interpretations and conclusions. One of the classical genotyping errors is caused by sensitivity to sample contamination by fungal 9.4 discussion 115

DNA (Dyer and Leonard, 2000). In four of 93 samples we found an increased amount of non-replicable bands (the genotyping error rate for these samples was 16–28 %), presumably due to fungal contamination. These samples were easy to recognize and were excluded from further scoring. Although AFLP is more reproducible than e. g. RAPDs, Bonin et al. (2004) pointed out the need to state the error rate in molecular studies. They distinguished between the technical difference rate (due only to laboratory procedures), genotyping error rate (due to laboratory work and differences in scoring between two replicated samples), and scoring differences between two different people evaluating the data. The first two error sources should not exceed 5 %. All our data were scored by one person with no attempt to quantify the latter error source. We calculated the genotyping error rate for our modified protocol and obtained results comparable with error rates in studies devoted to the reproducibility of AFLP profiles (Jones et al., 1997; Bonin et al., 2004) and with those obtained for mosses using the standard AFLP protocol (Fernandez et al., 2006). We repeated the entire AFLP procedure for 30 individuals from 18 populations and achieved a genotyping error rate of 4.51 % (for comparing genotyping error rate mentioned for Betula nana in Bonin et al. 2004 was 2.6 %, for Grimmia laevigata in Fernandez et al. 2006 was 5.9 %). Overall, AFLP can be sucessfully and effectively used as molecular markers for genetic studies of bryophytes. Our optimized protocol is expected to work well for species closely related to C. introflexus. We did not, however, test how well our optimized protocol applies to species increasingly distant phylogenetically from our study organism. For some species, modification of the protocol might improve results. At the start of the optimization process it is sufficient to try various DNA isolation approaches – CTAB, diverse commercial kits, purification – because secondary metabolites could inhibit subsequent PCR. Optimum amounts of DNA for restriction should be identified because too low or too high concentrations may cause problems. Low concentrations of restriction enzymes and/or short time of restriction/ligation can also be a source of variation. The restriction and ligation steps are probably the most crucial in the whole AFLP process. Particular care should be given to different peak intensities and difference in the way investiga- tors score presence / absence of bands. The first problem could be solved by manual scoring and the juxtaposition of samples with various intensity of profiles. Ideally, a given study should be scored by a single individual. Repeatability for bryophytes appears to be similar to that found for other organisms (up to to 5 % error). acknowledgements

This project was supported by the Grant Agency of Charles University (project no. 258/2004/B-BIO/PrF/) and the Ministry of Education, Youth and Sports of the Czech Republic (grant no. MSM 0021620828 and MSM0021622416). We are grateful to Jonathan Shaw for valuable comments. 116 molecular methods optimalization (paper 3)

references

Arens, P., H. Coops, J. Jansen and B. Vosman (1998). “Molecular genetic analysis of black poplar (Populus nigra L.) along Dutch rivers.” Molecular Ecology 7, 11–18. Blears, M.J., S.A. De Grandis, H. Lee and J.T. Trevors (1998). “Amplified fragment length polymorphism (AFLP): a review of the procedure and its applications.” Journal of industrial microbiology & biotechnology 21, pp. 99– 114. Bless, C., H. Palmeter and M.M. Wallace (2006). “Identification of Acer rubrum using Amplified Fragment Length Polymorphism.” Journal of forensic sciences 51, pp. 31–38. Bonin, A., E. Bellemain, P. Bronken Eidesen, F. Pompanon, C. Brochmann and P. Taberlet (2004). “How to track and assess genotyping errors in population genetics studies.” Molecular Ecology 13, pp. 3261–3273. Bory, S., P. Lubinsky, A.M. Risterucci, J.L. Noyer, M. Grisoni, M.F. Duval and P. Besse (2008). “Patterns of introduction and diversification of Vanilla planifolia (Orchidaceae) in Reunion Island (Indian Ocean).” American Journal of Botany 95, pp. 805–815. Chauhan, N., M.S. Negi, V. Sabharwal, D.K. Khurana and M. Lakshmiku- maran (2004). “Screening interspecific hybrids of Populus (P. ciliata × maxi- mowiczii) using AFLP markers.” Theoretical and Applied Genetics 108, pp. 951– 957. Daisie (2009). Handbook of alien species in Europe. Vol. 3. Invading nature. Dor- drecht: Springer, p. 399. Doyle, J.J. (1990). “Isolation of plant DNA from fresh tissue.” Focus 12, pp. 13– 15. Dyer, A.T. and K.J. Leonard (2000). “Contamination, error, and nonspecific molecular tools.” Phytopathology 90, pp. 565–567. Fernandez, C.C., J.R. Shevock, A.N. Glazer and J.N. Thompson (2006). “Cryp- tic species within the cosmopolitan desiccation-tolerant moss Grimmia laevigata.” Proceedings of the National Academy of Sciences of the United States of America 103, pp. 637–642. Frahm, J.P. (1984). “Phytogeography of European Campylopus species.” Pro- ceedings of the Third Meeting of Bryologists from Central and East Europe. Ed. by J. Váˇna.Praha: Univerzita Karlova, pp. 191–212. Hassel, K. and L. Söderström (2005). “The expansion of the alien mosses Orthodontium lineare and Campylopus introflexus in Britain and continental Europe.” Journal of the Hattori Botanical Laboratory 97, pp. 183–193. Invitrogen. AFLP® Core Reagent Kit. url: http://tools.invitrogen.com/ content/sfs/manuals/10482016.pdf (visited on 05/12/2012). – AFLP® Instruction Manual. url: http://tools.invitrogen.com/content/ sfs/manuals/aflpi_man.pdf (visited on 05/11/2012). Jones, C.J., K.J. Edwards, S. Castaglione, M.O. Winfield, F. Sala, C. Van de Wiel, G. Bredemeijer, B. Vosman, M. Matthes, A. Daly et al. (1997). “Repro- ducibility testing of RAPD, AFLP and SSR markers in plants by a network of European laboratories.” Molecular breeding 3, pp. 381–390. 9.4 discussion 117

Karp, A., O. Seberg and M. Buiatti (1996). “Molecular techniques in the assessment of botanical diversity.” Annals of Botany 78, pp. 143–149. Mahmud, M.A., H. Kitaura, M. Fukuda and A. Yamada (2007). “AFLP analysis for examining genetic differences in cultivated strains and their single- spore isolates and for confirming successful crosses in Agaricus blazei.” Mycoscience 48, pp. 297–304. McDaniel, S.F., J.H. Willis and A.J. Shaw (2007). “A linkage map reveals a complex basis for segregation distortion in an interpopulation cross in the moss Ceratodon purpureus.” Genetics 176, pp. 2489–2500. Meudt, H.M. and A.C. Clarke (2007). “Almost forgotten or latest practice? AFLP applications, analyses and advances.” Trends in plant science 12, pp. 106– 117. Mittmann, F., S. Dienstbach and G. Wagner (2007). “Large scale extraction of high quality moss DNA.” Russian Journal of Plant Physiology 54, pp. 564– 568. Mueller, U.G. and L.L.R. Wolfenbarger (1999). “AFLP genotyping and fingerprinting.” Trends in Ecology & Evolution 14, pp. 389–394. Pedersen, N., S.J. Russell, A.E. Newton and S.W. Ansell (2006). “A novel molecular protocol for the rapid extraction of DNA from bryophytes and the utility of direct amplification of DNA from a single dwarf male.” The Bryologist 109, pp. 257–264. Pelser, P.B., B. Gravendeel and R. Van der Meijden (2003). “Phylogeny recon- struction in the gap between too little and too much divergence: the closest relatives of Senecio jacobaea (Asteraceae) according to DNA sequences and AFLPs.” Molecular Phylogenetics and Evolution 29, pp. 613–628. Pfeiffer, T., S. Fritz, M. Stech and W. Frey (2006). “Vegetative reproduction and clonal diversity in Rhytidium rugosum (Rhytidiaceae, Bryopsida) inferred by morpho-anatomical and molecular analyses.” Journal of plant research 119, pp. 125–135. Prohens, J., G.J. Anderson, J.M. Blanca, J. Cañizares, E. Zuriaga and F. Nuez (2006). “The implications of AFLP data for the systematics of the wild species of Solanum section Basarthrum.” Systematic botany 31, pp. 208–216. Rosendahl, S. and J.W. Taylor (1997). “Development of multiple genetic markers for studies of genetic variation in arbuscular mycorrhizal fungi using AFLP.” Molecular Ecology 6, pp. 821–829. Rowntree, J.K., R.S. Cowan, M. Leggett, M.M. Ramsay and M.F. Fay (2010). “Which moss is which? Identification of the threatened moss Orthodon- tium gracile using molecular and morphological techniques.” Conservation Genetics 11, pp. 1033–1042. Schlink, K. and R. Reski (2002). “Preparing high-quality DNA from moss (Physcomitrella patens).” Plant Molecular Biology Reporter 20, p. 423. Schlüter, P.M. and S.A. Harris (2006). “Analysis of multilocus fingerprinting data sets containing missing data.” Molecular Ecology Notes 6, pp. 569–572. Schönswetter, P., A. Tribsch, I. Stehlík and H. Niklfeld (2004). “Glacial history of high alpine Ranunculus glacialis (Ranunculaceae) in the European Alps in a comparative phylogeographical context.” Biological Journal of the Linnean Society 81, pp. 183–195. 118 molecular methods optimalization (paper 3)

Snäll, T., J. Fogelqvist, P.J. Ribeiro Jr and M. Lascoux (2004). “Spatial genetic structure in two congeneric epiphytes with different dispersal strategies analysed by three different methods.” Molecular Ecology 13, pp. 2109–2119. Vanderpoorten, A. and M. Tignon (2000). “Amplified fragments length polymorphism between populations of Amblystegium tenax exposed to contrasting water chemistries.” Journal of Bryology 22, pp. 252–262. Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. Van De Lee, M. Hornes, A. Friters, J. Pot, J. Paleman and M. Kuiper (1995). “AFLP: a new technique for DNA fingerprinting.” Nucleic acids research 23, pp. 4407–4414. Weising, K., H. Nybom, K. Wolff and G. Kahl (2005). DNA fingerprinting in plants: Principles, methods, and applications. 2nd ed. Boca Raton: CRC press, p. 472. Werner, O., R.M. Ros and J. Guerra (2002). “Direct amplification and NaOH extraction: two rapid and simple methods for preparing bryophyte DNA for polymerase chain reaction (PCR).” Journal of Bryology 24, pp. 127–131. Xie, C.F. and H.X. Lou (2009). “Secondary metabolites in bryophytes: an ecological aspect.” Chemistry & biodiversity 6, pp. 303–312. Xin, Z.G., J.P. Velten, M.J. Oliver and J.J. Burke (2003). “High-throughput DNA extraction method suitable for PCR.” Biotechniques 34, pp. 820–826. Zartman, C.E., S.F. McDaniel and A.J. Shaw (2006). “Experimental habitat fragmentation increases linkage disequilibrium but does not affect genetic diversity or population structure in the Amazonian liverwort Radula flac- cida.” Molecular ecology 15, pp. 2305–2315. THEROLEOFINTERSPECIFICCOMPETITIONIN 10 SPREADINGOFANINVASIVEMOSSSPECIES: TRANSPLANTATIONEXPERIMENTS.

Eva Mikulášková1,2, Michal Hájek1 & Kateˇrina Kintrová1 1Dept. of Botany and Zoology, Faculty of Science, Masaryk University, CZ-61137 Brno, Czech Republic; [email protected]. 2Dept. of Botany, Faculty of Science, Charles University in Prague, CZ-128 01 Praha 2, Czech Republic. abstract

One of the most important concerns associated with a new invasive species is whether it represents a threat for native species. In contrast to vascular plants, there are few invasive bryophyte species in Europe, with Campy- lopus introflexus being considered as the most invasive one. The ecological and economical impact of C. introflexus invasion has been investigated only in northern and western Europe, while information about its behaviour in the rest of Europe is missing. Here we investigated competitive potential of C. introflexus in Central–European habitats. We investigated the dynamics of colonisation of new patches, differences in colonisation rate between C. introflexus and native species and their competitive interactions in mixed patches. Three experimental designs were applied. First, permanent plots were established on bare soil bordered with C. introflexus cushion to investigate how rapidly and in which manner C. introflexus forms a compact cushion in unvegetated patches. Second, a set of permanent plots was established on bare soil between a C. introflexus cushion and a cushion of a native moss species common at a particular locality to test whether the colonisation rate differs between C. introflexus and native species. Third, a reciprocal transplantation experiment was established, where inner blocks of cushions were transplanted between C. introflexus and a native moss species in order to investigate whether C. introflexus is able to outcompete native species. Increase of cover of C. introflexus in newly colonised unvegetated patches was dependent on amounts of diaspores in the surroundings and on the presence of its competitive equivalents (r-strategy species Ceratodon purpureus or Pohlia nutans), but not on habitat. After initial colonisation, a short lag phase occurred and from the second year C. introflexus started to spread by frequently formed vegetative diaspores. Four out of six native species colonised new patches more slowly than C. introflexus, but they were not outcompeted by C. introflexus from mixed patches during the five-year experiment . We conclude that rapid invasion of C. introflexus in Central Europe is determined by its ability to quickly colonise disturbed patches rather than by competitive exclusion of native species.

119 120 the role of interspecific competition in c. introflexus spreading

keywords

bryophytes, Campylopus introﬂexus, competition, invasive species, transplantation experiment

10.1 introduction

Plant invasions are considered as one of the major threats for ecosystem diversity (Simberloff and Rejmánek, 2011). Invasive species generally have high propagule production and they are rapidly and effectively dispersed, typically by wind (Moravcová et al., 2010). In addition, they are often strong competitors and hence they can endanger persistence of native species in a community. Spreading of invasive species is influenced by their ability to colonise new habitats as well as their competition with native species. Within plants, bryophytes are the organisms for which both mechanisms, the effective dispersal and the competitive hierarchy are characteristic life history traits. The abundance of small spores that survive long distance dispersal (e. g. Van Zanten and Pócs, 1981; Muñoz et al., 2004) causes elevated dispersal and establishment rates (Herben et al., 1991; Herben, 1994). In addition, the diversity of asexual propagules in bryophytes is unparalleled among land plants (Duckett and Ligrone, 1992; Frahm, 2009). Inter- specific competition as a factor shaping species co-occurrence patterns has been documented within bryophytes as well (Rydin, 1997; Mälson and Ry- din, 2009). These life history characteristics may contribute to a high invasive potential of bryophytes. On the other hand, bryophytes generally have broader natural distributional ranges than vascular plants, often span- ning more continents (Frahm, 2009) and there is therefore lower probability of recent human-mediated species invasions into other continents. In addition, most alien bryophytes colonise predominantly urban habitats (gardens, roadsides, walls) and therefore they are not in focus of studies that assess risks of invasions to biodiversity of natural habitats. One of the few excep- tions is Campylopus introflexus (Hedw.) Brid., the invasive moss belonging to the 100 of the most invasive species in Europe (including all groups of organisms; Daisie, 2009), which has naturalized in near-natural vegetation such as coastal dunes, peatbogs and pine forests (Essl and Lambdon, 2009; Mikulášková et al., 2012a). Ecological aspects of the C. introflexus invasion and the species´ competitive potential have been investigated from the 1980s (e. g. Berg, 1985), but knowledge of its ecology is still sketchy. C. introflexus characteristically forms extremely dense turfs extending over hundreds of square meters in sand- dune areas in the Netherlands, in which almost no other bryophyte or lichen species can survive (Van der Meulen et al., 1987; Biermann and Daniëls, 1997; Ketner-Oostra and Sýkora, 2000, 2004; Sparrius and Kooijman, 2011). How- ever, there is no evidence for an ability of C. introflexus to outcompete native species in inland Europe. Because of increasing occurrence of C. introflexus in Central–European habitats, the research of competitive interactions with 10.1 introduction 121 local native bryophyte species can provide basis for a prediction of future behaviour of the invader including interactions with native flora. C. introflexus prosper on habitats with low competition of vascular plants and bryophytes (Mikulášková et al., 2012a). It benefits from frequent sexual as well as asexual spreading. It frequently appears in disturbed patches and initial successional phases (e. g. Hasse and Daniëls, 2006). Frequent observations showed an inhibition effect of C. introflexus on the colonisation of disturbed patches by native species (Biermann and Daniëls, 1995, 1997). This implies a certain level of competition with native bryophyte colonists. Because the habitats invaded by C. introflexus are rather species-poor (Miku- lášková et al., 2012a), initial density of diaspores will not play as big of a role compared to species-rich communities (compare Zamfir and Goldberg, 2000). The competition with native species may have two levels, either the direct competitive exclusion of a native species from a community or the faster colonisation of unvegetated patches by an invasive species. In order to understand the invasive behavior of C. introflexus, and bryophyte distributions generally, it is worth investigating not only the direct competitive interactions between the co-existing species, but also the interspecific differences in colonisation rate. Field transplantation experiments represent a useful tool for experimental testing of these effects. Glasshouse competitive experiments are frequently used for bryophytes as well (McAlister, 1995; Zamfir and Goldberg, 2000; Mälson and Rydin, 2009), but the advantage of transplantation experiments over species cultivation in controlled environment is that it ensures a com- plexity of environmental conditions that occur in nature. In laboratory, overall soil and water chemistry of a specific site can be sometimes maintainable only with difficulty. Moreover, experiments studying the role of competition are better designed in the field. The mid-term (3–5 years) field transplantation experiments were found to be sufficient to monitor changes in growth rate and competition (Granath et al., 2009) and the results may serve as a prediction of long-term changes (cf.Rydin, 1993a). The research of this phenomenon is focused mainly to fen bryophytes nowadays (e. g. Kooijman and Kanne, 1993; Kooijman and Bakker, 1995; Såstad et al., 1999; Mulligan and Gignac, 2002; Kotowski et al., 2006; Mälson and Rydin, 2009), but it could be successfully used in other habitats as well (Scandrett and Gimingham, 1989; Marino, 1991; Van der Hoeven and During, 1997). The aim of our study is to use the field transplantation experiments to address the following questions:

1. How rapidly and in which manner does C. introﬂexus form a compact dense cushion?

2. Do the colonisation rates differ between C. introﬂexus and native species?

3. Is C. introﬂexus able to outcompete native species? 122 the role of interspecific competition in c. introflexus spreading

10.2 methods

10.2.1 Field design

C. introflexus was reported from the Czech Republic for the first time in 1988 from South Bohemia (Novotný, 1990a). During the following years, the number of new localities has increased exponentially resulting in more than 80 recently known localities (Soldán, 1996, 1997; Mikulášková, 2006). The Czech Republic is a good example of a Central European region recently colonised by C. introflexus with a sufficient number of populations. The species invades predominantly coniferous forests and plantations and disturbed (drained) peatbogs. Accompanying species are those that are widespread and abundant in Central Europe (Mikulášková et al., 2012a). There are no threatened bryophyte species frequently occurring in known localities of C. introflexus in the Czech Republic, therefore we chose a few common native species for this study. Twenty-three localities of C. introflexus differing in habitat (pine forest, spruce forest, drained peatbog) were selected in the Czech Republic (Central Europe; Table 10.1). Each locality was geo-referenced. Occurrence of sporophytes, vegetative diaspores and vegetation composition were recorded. We established a series of field transplantation experiments in summer 2004 and 2005. Total cover of the species in a plot (cm2; A, B design; see lower) or differences in size of cushions (cm2; C design), and fitness of plants were monitored after first, second, third and fifth years (up to 2009 and 2010). Three types of experimental design were applied:

(A) Plots of 10 × 10 cm of bare soil just outside C. introﬂexus cushion were ﬁxed at nine localities and spreading of the target species was recorded. Altogether 9 plots were established for the experiment A.

(B) A experimental plot of 10 × 10 cm on bare soil was established directly between C. introﬂexus cushion and a cushion of its potential competitor — one native moss species common at a particular locality (Cer- atodon purpureus, Pohlia nutans, Polytrichastrum formosum, Hypnum cupressiforme). Spreading of both moss species into the plot was recorded; 2 replicates were established for each native species in two different habitats (drained peatbog, pine forest or spruce forest). Altogether 16 plots were established for the experiment B.

(C) A reciprocal transplantation experiment was established, in which inner blocks of cushions (10 × 10 cm) were transplanted between C. introflexus and a native moss species. Pairs of plots (C. introflexus inside a cushion of one of native species and the native species inside C. introflexus cushion) were monitored. Growing pattern of both moss species in a mixed patch was recorded. Nine native species were included. One pair of plots was established in pine forest for Campylopus flexuosus, Pleurozium schreberi, Dicranum polysetum; one pair in spruce forest for Aulacomnium palustre and two replicates of plot pairs were established for Ceratodon purpureus, Dicranum scoparium, Hypnum cupressiforme and Pohlia nutans in two habitats (pine forest, spruce forest; together 16 pairs of plots). 10.3 results 123

Three replicates of plot pair were established for Polytrichastrum formosum in three habitats (pine forest, spruce forest, drained peatbog; together 9 plots pairs). Ten plots were established as control plots, in which C. introﬂexus and all nine native species were transplanted into their own growth to distance 1 m. In total, 39 plots pairs were established for the experiment C.

10.2.2 Data analysis

Data were analysed by repeated measures ANOVA with repetition in time (experiment A) and species (i.e. C. introﬂexus vs. native species; experiment B, C). The post-hoc multiple comparisons were tested by Tukey’s honestly signiﬁcant difference test with equal or unequal sample sizes. In some cases, there was a negligible variability in the data and small number of replica- tions (particularly species in experiment C except for Polytrichastrum formosum). In these cases, only growing patterns and changes in cover were described, without statistical testing. Statistical analyses were performed using the Statistica for Windows (Statsoft Inc., 2011, Version 10) and R software (R Development Core Team).

10.3 results

10.3.1 How rapidly and in which manner does Campylopus introﬂexus form a compact cushion in unvegetated patches?

In the experiment A, C. introflexus spread laterally from an original cushion to bare soil after the first year. However, the difference in increase of cover among nine experimental plots was great. The cover of C. introflexus in experimenal plots had ranged from 5 to 50 cm2 after the first year and from 10 to 100 cm2 after the second year (Fig. 10.1). During the second and third year increases in cover were the highest. On average, the covered area enlarged by 16.5 cm2 per year and the rate of colonisation decreased over time. After five years, experimenal plots were overgrown from 40 to 100 cm2. No native moss species colonised experimenal plots during that time. The effect of habitat (spruce forest, pine forest, drained peatbog) on the species´ colonisation rate was not significant neither as a single factor (F2,18 = 0.3, p = 0.72) nor in interaction with particular years (F6,18 = 0.9, p = 0.52; Fig. 10.1). Focused on reproductive organs, young plants of C. introflexus were usually either without or with only few vegetative diaspores during the first year and all of them lacked sporophytes. During the second year the new plants formed vegetative diaspores with increasing frequency and during the third year almost all new plants formed vegetative diaspores. No population of C. introflexus formed sporophytes in experimental plots within five years. Both sterile and fertile populations dispersed predominantly by vegetative diaspores, but we cannot exclude that some new plants in experimental plots germinated from spores coming from mature plants growing outside the experimental plots. However, the rate of colonisation of experimental plots 124 the role of interspecific competition in c. introflexus spreading

Table 10.1: Localities of studied populations of Campylopus introﬂexus where experiments were established. A, B, C — experiment type see Methods, CZ – Czech Republic, PLA – Protected Land Area, NP – National Park, NR – Nature reserve [m a. s. l.] – E [°] Altitude 4603 400 5115 442 92364 379 56614 517 99122 400 49056 810 83383 750 72861 430 0446450839 732 409 85117 510 15333 522 2964758812 269 942 . 44447 431 50839 442 908116914557194 352 . 475 568 51686 630 62239 420 14192 522 76822 662 84 ...... – N [°] WGS 27528 13 60394 12 50533 17 96211 12 90787 13 37972 16 7086792131 14 14 7783689304 12 80517 14 16 8733109574 14 13 91911 14 9152692131 14 14 254118924284975 13 14 14 00483 12 23847 14 81686 16 49239 16 84 ...... 50 49 49 49 48 49 48 49 49 49 49 50 49 50 49 49 50 48 50 50 49 49 49 CF CS CS CS CS CS CF CF CF CF CS B, C F B, C S B, C S B, C S B, C S A, C F A, B, C S A, B, C F A, B, C F A, B, C S Ceratodon purpureus Polytrichastrum formosum Ceratodon purpureus Dicranum scoparium Pohlia nutans Polytrichastrum formosum Hypnum cupressiforme Ceratodon purpureus Pohlia nutans Ceratodon purpureus Hypnum cupressiforme Pleurozium schreberi Campylopus ﬂexulosus Dicranum scoparium Aulacomnium palustre Ceratodon purpureus Hypnum cupressiforme Polytrichastrum formosum Hypnum cupressiforme Dicranum polysetum Pohlia nutans Polytrichastrum formosum Polytrichastrum formosum Polytrichastrum formosum Polytrichastrum formosum Polytrichastrum formosum Pohlia nutans Dicranum scoparium Hypnum cupressiforme Hypnum cupressiforme Dicranum scoparium ˇ ˇ ˇ Ceský Les,Železná village. spruce forest Ceský Les, Vysoká village. spruce forest Ceský Les, Brtná village. spruce forest ˇ ˇ Ceské Švýcarsko, Mezná village. pine forest Ceské Švýcarsko, Kyjov village. pine forest CZ; Žichovec village. pine forest CZ; PLA CZ; PLA Beskydy, Hostašovice village. spruce forest CZ; PLA CZ; NP Šumava Mts., Soumarský most village. drained peatbog CZ; Radlík village.CZ; PLA Moravský Kras, Vilémovice village. spruce forest spruce forest CZ; Petrov village.CZ; PLA Šumava Mts., Raˇcínvillage. pine forest drained peatbog CZ; NR Toulovcovy Maštale, Budislav village. spruce forest CZ; NP Šumava Mts., Popelná village.CZ: Mže Stˇríbro, river, Ošelín village spruce forest pine forest CZ; NP CZ; NP CZ; Radlík village spruce forest CZ; PLA u Tˇreboˇnsko,Hluboká Borovan village.CZ; PLA pine Lužické forest hory, Jedlová village.CZ; Radlík village. spruce forest spruce forest A S CZ; Džbán Nature Park, Bílichov village. pine forest CZ; PLA CZ; PLA Borkovická Tˇreboˇnsko,NR blata. drained peatbog A F CZ; NR Toulovcovy Maštale. pine forest LocalityCZ; Benešov village. spruce forest Habitat Native species Experiment Fertile/sterile WGS 10.3 results 125

Figure 10.1: Medians of area covered by Campylopus introflexus in different habitats. Dark: pine forests, light: drained peatbogs, grey: spruce forests. Boxes define first and third quartiles, whiskers represent extremes. Only covers within the experimental plots (100 cm2) were measured.

was not signiﬁcantly different between fertile and sterile populations of the species (post-hoc multiple comparisons, p > 0.79).

10.3.2 Do the colonisation rate differ between C. introﬂexus and native species?

Increase in cover of C. introflexus and one native moss species common at a locality (Ceratodon purpureus, Pohlia nutans, Polytrichastrum formosum, Hypnum cupressiforme) was studied using the experiment B. Three plots (H. cupressiforme – pine forest, P. nutans – drained peatbog, P. formosum – spruce forest) were destroyed by animals during the course of the experiment. The cover of C. introflexus in experimental plots significantly increased over time (F3,27 = 57.1, p < 0.001). Generally, C. introflexus spread faster than native species (F3,27 = 14.7, p < 0.001, without differentiation among native species). Neither habitat nor identity of native moss species affected significantly the increase of cover of C. introflexus. There were marked differences in forming vegetative diaspores among experimental plots of C. introflexus during the first two years. Some populations spread only negligibly, especially those growing in dry pine forests where the species started to spread not until the third year when first vegetative diaspores were formed. In contrast, some populations (especially those in spruce forests and drained peatbogs) formed vegetative diaspores in high amount already during the first year and they were quickly spread to the experimental plots where C. introflexus became dominant already during the second and third year. Looking at the increase in cover of the native moss species, there was not a significant difference among species neither in an average cover (F3,27 = 2.5, 126 the role of interspecific competition in c. introflexus spreading

Figure 10.2: Medians of area covered by Campylopus introflexus and particular native bryophyte species. Dark: C. introflexus, light: native species specified in the title. Boxes define first and third quartiles, whiskers represent extremes. Small letters in third and fourth panels describe significant differencies among means of covers for particular species and years.

p = 0.13) nor with respect to particular years (F9,27 = 0.6, p = 0.81). The type of habitat did not significantly affect the propagation of the native moss species (single factor: F2,30 = 1.1, p = 0.36; interaction with years: F6,30 = 0.3, p = 0.94). Although the differences in cover increase of either C. introflexus or native moss species among plots were not significant, the comparison of the increase in cover between C. introflexus and the native species at the same plot detected different patterns for particular native moss species (Fig. 10.2). The best competitor with C. introflexus was Ceratodon purpureus whose increase of cover was as fast as the invasive C. introflexus (F3,9 = 0.1, p = 0.96; Fig. 10.2). Three of four plots were almost overgrown by both competitors after five years of the experiment. C. introflexus reached the least mean cover (37.5 cm2) in plots with C. purpureus in the fifth year. A second statistically good competitor with the invasive C. introflexus was Pohlia nutans. Its mean cover after the first year was slightly higher than C. introflexus (33.3 cm2). During the following years P. nutans lost its priority but the difference in cover increase between P. nutans and C. introflexus was not significant (F3,6 = 1.3, p = 0.37, Fig. 10.2). After five years, all the plots were completely covered by admixture of both species. The final mean cover of C. introflexus was 60 cm2. 10.3 results 127

C. introflexus became dominant in all plots bordered by Hypnum cupressiforme cushions. H. cupressiforme formed long stems which colonised the unvegetated plot or overgrew C. introflexus plants during the first year, however its increase in cover was lower than that of C. introflexus (F3,6 = 4.8, p = 0.05, Fig. 10.2). After five years, the plot area was dominantly covered by C. introflexus (53.3 cm2) with an admixture of H. cupressiforme (23.3 cm2). Polytrichastrum formosum was the weakest competitor with C. introflexus. It did not spread markedly for two years. During the next three years its cover increased a bit as a result of several new gametophytes, but the increase in cover of C. introflexus was much higher (F3,6 = 13.1, p = 0.005, Fig. 10.2). A majority of the area (60–90 cm2) of plots bordered by P. formosum were covered by C. introflexus after five years (76.7 cm2 in average). Two experimental plots with P. formosum were colonised by P. nutans, which formed a mixed stand with C. introflexus, but covered only small area (10–20 cm2) of the plots.

10.3.3 Is C. introﬂexus able to outcompete native species?

The reciprocal transplantation experiment (design C) explored the ability of established plants of C. introflexus to displace a native moss species. Six pairs of plots (with Dicranum scoparium, Pohlia nutans and Polytrichastrum formosum) were destroyed by animals during the course of the experiment. Control plots showed no detectable changes in fitness or cover for all included species within five years. We observed three types of patterns in the interaction between C. introflexus and native species. These patterns were same across all habitats:

1. Competitive equivalents — the same cover of both species. All transplanted cushions of Ceratodon purpureus, Dicranum scoparium and Pohlia nutans into C. introflexus cushions as well as all transplanted cushions of C. introflexus into cushions of these native species did not change cover during the first year. During the second year, a 1–3 cm wide mixed zone had established, but since the third year there were no detectable changes in covers of all species. Measured area of all species stayed approximately same as at the start of the experiment (ca 100 cm2). Similar interspecific interaction was observed in plots with Campylopus flexuosus, but mixed zone was formed already during the first year.

2. C. introflexus dominates. Transplanted cushions of Aulacomnium palustre, Dicranum polysetum and Polytrichastrum formosum into C. introflexus had dried during the first year in the most plots. These cushions died during the second year and C. introflexus expanded to the unvegetated patches. Transplanted cushions of C. introflexus into the cushions of native species showed no visible changes after the first year; later, particular native species started to differ in their interaction with C. introflexus. P. formosum did not change its cover during second year. A three cen- timeters wide mixed zone was established during the third year. The gametophytes of A. palustre and D. polysetum dried on borders of transplanted cushions of C. introflexus, and therefore C. introflexus started to 128 the role of interspecific competition in c. introflexus spreading

expand. Its area increased of 30 cm2 at the expense of A. palustre and of 10 cm2 at the expense of D. polysetum. Covers of all above mentioned species in all plots became stable within third or four year and we did not detect any change in the ﬁfth year.

3. Native species dominates. Transplanted cushions of Hypnum cupressiforme into C. introflexus spread by new stems of maximum 5 cm length and overgrew surrounding cover of C. introflexus. However, within the next four years, H. cupressiforme increased its cover only slightly and it was not able to form a more dense carpet that could displace C. introflexus. Similar patterns were observed in cushions of C. introflexus transplanted into H. cupressiforme. C. introflexus was a little subsided during the first year because few stems of H. cupressiforme overgrew its transplanted cushions, but H. cupressiforme was not able displace the dense cushion of C. introflexus. The area of transplanted C. introflexus decreased to 75–90 cm2. The cushion of Pleurozium schreberi transplanted into C. introflexus started to dry within two years and C. introflexus expanded inwards in these plots. The area of transplanted P. schreberi decreased to 60 cm2 during five years. The transplanted cushion of C. introflexus into P. schreberi showed no visible changes after the first year, but its borders started to be shaded by high gametophytes of P. schreberi. C. introflexus was overgrown by them until second and third year. The area of transplanted C. introflexus decreased to 70 cm2.

10.4 discussion

The magnitude of competitive exclusion among species with high niche overlap has been questioned by plant ecologists over the last decades. Invasive species often share a niche with some native species. Our results suggest that exclusion due to interspecific competition between native and invasive bryophytes is either a slow process or may not occur at all. For fen bryophytes, Mälson and Rydin (2009) also found a competitive hierarchy, but no competitive exclusion during the duration of a glasshouse experiment. We monitored the transplantation experiment for five years. Changes in interactions between the species were observed within the first two or three years. After that time the equilibrium has been established. Risk of transplantation experiments is in their interpretation. Bryophytes are sensitive to changes of microhabitats conditions. The low survival of transplanted species might be actually caused by presence of a superior competitor, but might be also caused simply by transport to non-ideal microhabitats (Kooijman and Kanne, 1993; Van der Hoeven, 1999; Rydin, 1993a,b). This could explain Aulacomnium palustre, Dicranum polysetum, Pleurozium schreberii and Polytrichastrum formosum quickly dried after moving into C. introflexus, which prefer drier habitats (Mikulášková et al., 2012a) and bare patches inside their cushions accelerated drying. Transplanted C. introflexus colonised these new unvegetated patches. Again, it illustrates for this species that an ability to colonise unvegetated patches rapidly is more important than competition ability. 10.4 discussion 129

10.4.1 Dynamics of establishment of new populations

Colonisation of new habitats requires successful reproductive strategy. Spec- ies with multiple reproductive strategies are in general good colonists (Huen- neke and Vitousek, 1990). The number of diaspores and dispersal mode are critical factors regulating the spread of invasive species. C. introflexus benefits from both vegetative and generative reproduction. Generative reproduction starts after five years from initial colonisation at earliest, because creation of sporophytes on new plants in new patches was not observed in any monitored population (in both fertile and sterile populations). However, we detected that C. introflexus frequently formed vegetative diaspores (fragile upper part of leafy stems) within two years after the occurrence of new patches. Then, the new plants produced vegetative diaspores immediately and stands become dense and compact within next three years. It seems that the occurrence of free patches close to established cushions causes increasing production of vegetative diaspores on old gametophytes. After initial successful colonisation of a microhabitat, the next stage of invasion starts and is characterized by establishment of a viable, self-sustaining population. This period represents a short lag phase, a common feature of an invasion; the onset of rapid population growth and range expansion (Mack, 1985). Lag phase is expected to occur because of adaptations to a new habitat. We detect approximately one year long lag phase in C. introflexus cushions. Once initial colonisation and establishment have occurred, C. introflexus was able continue to long-distance dispersal by spores (in originally fertile populations), and to start short-distance dispersal by vegetative diaspores with lateral expansion of the established population (Davis and Thompson, 2000). Continued spread of the established population often occurs because of ex- cellent adaptation of the species to disperse (Sakai et al., 2001). We observed that during the period about two years after initial colonisation C. introflexus spreads the most quickly into new microhabitats. It is a phase when it could represent a threat for native communities, because it could quickly occupy free space and block further succession. Dynamics of colonisation of new microhabitats corresponds mainly with colonist life strategy (During, 1979) with high reproductive effort, late sporophytes, first asexual reproduction after few month, short turf growth form and small spores. Establishment in a natural community may require different traits than those required upon entering a human disturbed habitat (Horvitz et al., 1998). Human disturbance of natural communities may have broadened the range of characteristics leading to successful colonisation and thus increased the frequency of invasion into existing communities (Vitoušek et al., 1996). Observed dynamics of colonisation of unvegetated patches showed that C. introflexus could prosper mainly in human influenced habitats, where initial cover of vascular plants and other bryophytes is reduced.

10.4.2 Difference between C. introﬂexus and native species in increase of cover

Competition avoidance is a fundamental strategy of many moss species (Slack, 1990). It could be achieved through efﬁcient dispersal, high spore out- 130 the role of interspecific competition in c. introflexus spreading

put and rapid development (the ruderal strategy, During, 1990). In contrast, stress-tolerant species occupy habitats that are unsuitable for stronger competitors. Thomas et al. (1994) shows that two species (Ceratodon purpureus, Pohlia nutans) which were the most able to compete with C. introflexus in speed of colonisation and establishment, combine ruderal and stress-tolerant life strategy and are able to behave as pioneer species also under toxic conditions. Many studies have documented that invaders show a superior ability to exploit local resources as compared to native natives (Melgoza et al., 1990; Petren and Case, 1996) or as compared to non-invading introduced species (Thebaud et al., 1996). The competitive interactions in the juvenile stage might be more important than those of the established plants (Wat- son, 1980; Jonsson and Esseen, 1990). In accordance with ruderal and stress- tolerant life strategy, C. introflexus is more frequent on bare soils without strong competition of vascular plants and other bryophytes (Mikulášková et al., 2012a). Invasive plants are typically dispersed by wind and have high diaspore production (Moravcová et al., 2010). The speed of colonisation of new patches is therefore dependent on amount of diaspores in regional diaspore bank (During, 2001). The experiment investigating colonisation speed of C. introflexus as compared with native species showed that the speed of colonisation is similar for C. introflexus, Pohlia nutans and Ceratodon purpureus. The amount of diaspores in regional diaspore bank for both these species is higher than for C. introflexus, because they are more common species. There- fore we can conclude that successful colonisation of new patches by C. introflexus is a result of coincidence between availability of bare soil with suitable ecological conditions (compare Mikulášková et al., 2012a), presence of diaspores of C. introflexus and low amount of diaspores of Pohlia nutans or Ceratodon purpureus. Although a high dispersal ability of bryophytes is generally assumed, several studies (Söderström and Jonsson, 1989; Kimmerer, 1991; Sundberg, 2005) show that a large proportion of bryophyte spores are dispersed only within a few meters from the source. Therefore, once a locality is colonised by C. introflexus, the species spreads into the surrounding area. However, further spread of C. introflexus is also dependent on bryophyte species composition and vascular plant canopy openness. Two native species (Hypnum cupressiforme and Polytrichastrum formosum) were not able to compete with C. introflexus in the increase of cover, but C. introflexus were not able to displace them from mixed patches.

10.4.3 Interspeciﬁc interactions with native species

Interspecific interactions and competitive ability are other traits that increase a success of an invasive species. The great number of taxonomically allied species in dense bryophyte communities supports the observation that morphologically similar coexisting species are more likely to be competitive equivalents (e. g. Slack, 1984; Marino, 1991; Marino, 1997). C. introflexus forms mixed stands with Pohlia nutans, Ceratodon purpureus, Dicranum scoparium and Ceratodon flexuosum. All species are acrocarpous and form short or 10.4 discussion 131 tall turfs. They are morphologically similar and belong to the same order Di- cranales, except for P. nutans (Bryales). Their competitive ability is supposed to be similar and they colonise similar habitats. The cause of their coexistence in mixed patches might be explained by colonisation and recolonisa- tion events in unvegetated patches in suitable habitats (preemptive competition). This effect of priority in colonisation was previously observed in peatbog communities (Mälson and Rydin, 2009), where an established species prevents newcomers from germinating or greatly decreases their chance on survival (Rydin, 1997). Connell and Slatyer (1977) formulated three models of succession, where interspecific competition played an important role. Many studies of colonisation ability of bryophytes show that bryophyte communities could be characterized by the Tolerance model of succession, in which later colonists (climax species) could grow already on new patches, but their growth rate is lower than that of earlier colonists (see review in Rydin, 1997). On the other hand, observations of competition among bryophytes e. g. (Breeuwer et al., 2008; Granath et al., 2010) shows that competitive ability may differ only little among species, and priority of colonisation is therefore the crucial force in forming species composition (Rydin, 1993a,b). Dependence of the species composition on primary colonisation is a typical feature of the Inhibition (invasive) model of succession. Our results show that on suitable habitats C. introflexus is one of the most successful colonists and also long- term dominant species inhibiting succession. Most native species were not able to outcompete it, therefore the presence of C. introflexus in our bryoflora influences species composition of native early succession stages and conse- quent communities. Once it formed a pure dense carpet it is very resistant to the succession of other bryophytes or vascular plants (Equihua and Usher, 1993). Negative effects on the establishment of vascular plant seedlings by dense bryophyte layer was recorded also in non-invading species (Otsus and Zobel, 2002), but positive effects have been documented as well (Van Tooren, 1988; Ryser, 1993). Dense carpets of C. introflexus which inhibit succession are known from coastal sand dunes (Klinck, 2010), drained peatbogs (novotny1990Campylopus) or dry pine forests (Mikulášková et al., 2012a). This behavior is similar to that of Racomitrium lanuginosum on lava fields in Iceland (Bjarnason, 1991). Both species represent the Invasive Competitive Model (Connell and Slatyer, 1977). The support of the Invasive Competition Model as relevant for C. introflexus could be seen in an ability to colonise and spread in suboptimal ecological conditions such as wet soils. Although the species avoids high water levels (Mikulášková et al., 2012a), its patches transplanted into wet cushions of Aulacomnium palustre and Dicranum polysetum survived and spread there. The reciprocal transplants show that the ability of C. introflexus to outcompete native species directly is low. The temporarily decreasing number gametophores of P. formosum at the beginning of the experiment was caused rather by disturbance (compare Hasse, 2007). Similar development, with tem- porary expansion of a species and renewed equilibrium after few years, was observed in the case of Hypnum cupressiforme transplanted into C. introflexus. The only species which was able to overgrow the cushions of C. introflexus 132 the role of interspecific competition in c. introflexus spreading

was Pleurozium schreberi because of its high gametophytes and thus a higher ability to compete for light. However, this species was not able to displace an established dense carpet of C. introflexus during the experiment. Despite these our results, we still cannot completely exclude that competitive exclusion between C. introflexus and native species in a mixed patch might occur in a long-term perspective. Hasse (2007) showed that presence of C. introflexus could cause a decrease of Polytrichum piliferum cover within five years. However, our results for Pleurozium schreberi rather suggest that in inland Europe C. introflexus is a superior competitor over high-gametophyte species even in the long-term perspective. Based on the result of all three experiments we can therefore conclude that the behavior of C. introflexus corresponds with the Invasive competition model, but its rapid invasion is determined more by an ability to quickly colonise disturbed patches than by competitive superiority.

references

During, H.J. (1990). “Clonal growth patterns among bryophytes.” In: Clonal growth in plants: regulation and function. Ed. by J. Van Groenendael and H. De Kroon. The Hague: SPB Academic Publishing, pp. 153–176. –(2001). “New frontiers in bryology and lichenology - Diaspore banks.” Bryologis 104, pp. 92–97. Equihua, M. and M.B. Usher (1993). “Impact of carpets of the invasive moss Campylopus introﬂexus on Calluna vulgaris regeneration.” Journal of Ecology 81, pp. 359–365. Essl, F. and P.W. Lambdon (2009). “Alien bryophytes and lichens of Europe.” In: Handbook of Alien Species in Europe. Ed. by Daisie. Dordrecht: Springer, pp. 29–41. Frahm, J.P. (2009). “Diversity, dispersal and biogeography of bryophytes (mosses).” In: Protist Diversity and Geographical Distribution. Ed. by W. Foiss- ner and D.L. Hawksworth. Netherlands: Springer, pp. 43–50. Granath, G., J. Strengbom, A. Breeuwer, M.M.P.D. Heijmans, F. Berendse and H. Rydin (2009). “Photosynthetic performance in Sphagnum transplanted along a latitudinal nitrogen deposition gradient.” Oecologia 159, pp. 705– 715. Granath, G., J. Strengbom and H. Rydin (2010). “Rapid ecosystem shifts in peatlands: linking plant physiology and succession.” Ecology 91, pp. 3047– 3056. Hasse, T. (2007). “Campylopus introﬂexus invasion in a dune grassland: succession, disturbance and relevance of existing plant invader concepts.” Herzogia 20, pp. 305–315. Hasse, T. and F.J.A. Daniëls (2006). “Species responses to experimentally induced habitat changes in a Corynephorus grassland.” Journal of Vegetation Science 17, pp. 135–146. Herben, T. (1994). “Local rate of spreading and patch dynamics of an invasive moss species, Orthodontium lineare.” Journal of Bryology 18, pp. 115– 125. Herben, T., H. Rydin and L. Söderström (1991). “Spore establishment probability and the persistence of the fugitive invading moss, Orthodontium lineare - a spatial simulation-model.” Oikos 60, 215–221. Horvitz, C.C., J.B. Pascarella, S. McMann, A. Freedman and R.H. Hofstetter (1998). “Functional roles of invasive non-indigenous plants in hurricane- affected subtropical hardwood.” Ecological Applications 8, pp. 947–974. Huenneke, L.F. and P.M. Vitousek (1990). “Seedling and clonal recruitment of the invasive tree Psidium cattleianum: implications for management of native Hawaiian forests.” Biological Conservation 53, pp. 199–211. Jonsson, B.G. and P.A. Esseen (1990). “Treefall Disturbance Maintains High Bryophyte Diversity in a Boreal Spruce Forest.” Journal of Ecology 78, pp. 924– 936. Ketner-Oostra, R. and K.V. Sýkora (2000). “Vegetation succession and lichen diversity on dry coastal calciumpoor dunes and the impact of management experiments.” Journal of Coastal Conservation 6, pp. 191–206. –(2004). “Decline of lichen-diversity in calcium-poor coastal dune vegetation since the 1970s, related to grass and moss encroachment.” Phyto- coenologia 34, pp. 521–549. 134 the role of interspecific competition in c. introflexus spreading

Kimmerer, R.W. (1991). “Reproductive ecology of Tetraphis pellucida. II. Dif- ferential success of sexual and asexual propagules.” Bryologist, pp. 284– 288. Klinck, J. (2010). NOBANIS – Invasive Alien Species Fact Sheet – Campylopus introflexus. url: http://www.nobanis.org (visited on 05/11/2012). Kooijman, A.M. and C. Bakker (1995). “Species replacement in the bryophyte layer in mires: the role of water type, nutrient supply and interspecific interactions.” Journal of Ecology 83, pp. 1–8. Kooijman, A.M. and D.M. Kanne (1993). “Effects of water chemistry, nutrient supply and interspecific interactions on the replacement of Sphagnum subnitens by S. fallax in fens.” Journal of bryology 17, pp. 431–438. Kotowski, W., W. Thörig, R. Diggelen and M.J. Wassen (2006). “Competition as a factor structuring species zonation in riparian fens-a transplantation experiment.” Applied Vegetation Science 9, pp. 231–240. Mack, R.N. (1985). “Invading plants: their potential contribution to population biology.” In: Studies on plant demography: John L. Harper festschrift. Ed. by J. White. London: Academic Press, pp. 127–142. Marino, P. (1997). “Competition, dispersal and coexistence of Splachnaceae in patchy habitats.” Advances in Bryology 6, p. 63. Marino, P.C. (1991). “Competition between mosses (Splachnaceae) in patchy habitats.” The Journal of Ecology, pp. 1031–1046. McAlister, S. (1995). “Species interactions and substrate specificity among log-inhabiting bryophyte species.” Ecology 76, pp. 2184–2195. Melgoza, G., R.S. Nowak and R.J. Tausch (1990). “Soil water exploitation after fire: competition between Bromus tectorum (cheatgrass) and two native species.” Oecologia 83, pp. 7–13. Mikulášková, E. (2006). “Vývoj rozšíˇrení neofytického mechu Campylopus introflexus v Ceskéˇ republice [Development in distribution of the neophytic moss Campylopus introflexus in the Czech Republic].” Bryonora 38, pp. 1–10. Mikulášková, E., Z. Fajmonová and M. Hájek (2012a). “Invasion of the moss Campylopus introflexus into central European habitats.” Preslia. [under review]. Moravcová, L., P. Pyšek, V. Jarošík, V. Havlícková and P. Zákravský (2010). “Reproductive characteristics of neophytes in the Czech Republic: traits of invasive and non-invasive species.” Preslia 82, pp. 365–390. Mulligan, R.C. and L.D. Gignac (2002). “Bryophyte community structure in a boreal poor fen II: interspecific competition among five mosses.” Canadian journal of botany 80, pp. 330–339. Muñoz, J., Á.M. Felicísimo, F. Cabezas, A.R. Burgaz and I. Martínez (2004). “Wind as a long-distance dispersal vehicle in the Southern Hemisphere.” Science 304, pp. 1144–1147. Mälson, K. and H. Rydin (2009). “Competitive hierarchy, but no competitive exclusions in experiments with rich fen bryophytes.” Journal of Bryology 31, pp. 41–45. Novotný, I. (1990a). “Campylopus introflexus (Hedw.) Brid. – nový druh mechu v Ceskoslovensku.”ˇ 2. Bryologicko–lichenologické dny v Opavˇe(19.–21.9.1989) – souhrny referát˚u. Ed. by Anonymus. Vol. 4, pp. 4–5. 10.4 discussion 135

Otsus, M. and M. Zobel (2002). “Small-scale turnover in a calcareous grassland, its pattern and components.” Journal of Vegetation Science 13, pp. 199– 206. Petren, K. and T.J. Case (1996). “An experimental demonstration of exploitation competition in an ongoing invasion.” Ecology 77, pp. 118–132. R Development Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, url: http://www.r- project.org (visited on 09/24/2011). Rydin, H. (1993a). “Interspecific competition between Sphagnum mosses on a raised bog.” Oikos, pp. 413–423. –(1993b). “Mechanisms of interactions among Sphagnum species along water- level gradients.” Advances in Bryology 5, pp. 153–185. –(1997). “Competition among bryophytes.” Advances in Bryology 6, pp. 135– 168. Ryser, P. (1993). “Influences of neighbouring plants on seedling establishment in limestone grassland.” Journal of Vegetation Science 4, pp. 195–202. Sakai, A.K., F.W. Allendorf, J.S. Holt, D.M. Lodge, J. Molofsky, K.A. With, S. Baughman, R.J. Cabin, J.E. Cohen, N.C. Ellstrand et al. (2001). “The Popu- lation Biology of Invasive Specie.” Annual Review of Ecology and Systematics 32, pp. 305–332. Scandrett, E. and C.H. Gimingham (1989). “Experimental investigation of bryophyte interactions on a dry heathland.” The Journal of Ecology 77, pp. 838– 852. Simberloff, D. and M. Rejmánek, eds. (2011). Encyclopedia of biological invasions. London: Univiversity of California Press. Slack, N.G. (1984). “A new look at bryophyte community analysis: field and statistical methods.” Journal of the Hattori Botanical Laboratory 55, pp. 113– 32. –(1990). “Bryophytes and ecological niche theory.” Botanical Journal of the Linnean Society 104.1-3, pp. 187–213. Soldán, Z. (1996). “Rozšíˇrení neofytických mech ˚u Campylopus introflexus a Or- thodontium lineare v Ceskéˇ republice [Distribution of Campylopus introflexus a Orthodontium lineare in Czech Republic].” Brynora 18, pp. 10–19. –(1997). “Invazní mechorosty. [Invasive bryophytes].” Zprávy Ceskéˇ botanické spoleˇcnosti 14, pp. 33–39. Sparrius, L.B. and A.M. Kooijman (2011). “Invasiveness of Campylopus introflexus in drift sands depends on nitrogen deposition and soil organic matter.” Applied Vegetation Science 14, pp. 221–229. Sundberg, S. (2005). “Larger capsules enhance short-range spore dispersal in Sphagnum, but what happens further away?”: Oikos 108, pp. 115–124. Såstad, S.M., B. Pedersen and K. Digre (1999). “Habitat-specific genetic effects on growth rate and morphology across pH and water-level gradients within a population of the moss Sphagnum angustifolium (Sphagnaceae).” American journal of botany 86, pp. 1687–1698. Söderström, L. and B.G. Jonsson (1989). “Spatial pattern and dispersal in the leafy hepatic Ptilidium pulcherrimum.” Journal of Bryology 15, pp. 793–802. 136 the role of interspecific competition in c. introflexus spreading

Thebaud, C., A.C. Finzi, L. Affre, M. Debussche and J. Escarre (1996). “As- sessing why two introduced Conyza differ in their ability to invade Mediter- ranean old fields.” Ecology 77, pp. 791–804. Thomas, P.A., M.C.F. Proctor and E. Maltby (1994). “The ecology of severe moorland fire on the North York moors: chemical and physical constraints on moss establishment from spores.” Journal of Ecology 82, pp. 457–474. Van der Hoeven, E.C. (1999). “Do Chalk Grassland Bryophytes Compete?” PhD thesis. Depon in: Utrecht University, Utrecht. Van der Hoeven, E.C. and H.J. During (1997). “The effect of density on size frequency distributions in chalk grassland bryophyte populations.” Oikos 80, pp. 533–539. Van der Meulen, F., H. Van der Hagen and B. Kruijsen (1987). “Campylopus introflexus. Invasion of a moss in Dutch coastal dunes.” Proceedings of the Koninklijke. Nederlandse Akademie van Wetenschappen. 90, pp. 73–80. Van Tooren, B.F. (1988). “The fate of seeds after dispersal in chalk grassland: the role of the bryophyte layer.” Oikos 53, pp. 41–48. Van Zanten, B.O. and T. Pócs (1981). “Distribution and dispersal of bryophytes.” Advances in bryology 1, pp. 479–562. Vitoušek, P.M., C.M. D’Antonio, L.L. Loope and R. Westbrooks (1996). “Bi- ological invasions as global environmental change.” American Scientist 84, pp. 468–478. Watson, M.A. (1980). “Patterns of habitat occupation in mosses-relevance to considerations of the niche.” Bulletin Of The Torrey Botanical Club 107, pp. 346–372. Zamfir, M. and D.E. Goldberg (2000). “The effect of initial density on interactions between bryophytes at individual and community levels.” Journal of Ecology 88, pp. 243–255. CHANGESINDISTRIBUTIONOFTHEMOSS 11 CAMPYLOPUSINTROFLEXUS INTHECZECHREPUBLIC

Eva Mikulášková Dept. of Botany, Faculty of Science, Charles University in Prague, CZ-128 01 Praha 2, Czech Republic; [email protected]. abstract

The neophytic moss Campylopus introﬂexus is known from 71 localities in the Czech Republic. Geographical and altitude of the localities in the country are reviewed and the development in the number of collections over the time is reported as well. keywords

Invasive species, Campylopus introﬂexus, Czech Republic, distribution map.

11.1 introduction

Campylopus introflexus (Hedw.) Brid. is one of the most prominent invasive bryophytes in Europe. It is a neophytic species with an original distribution in the southern hemisphere (South America, southern parts of Africa and Australia, New Zealand; Frahm 1984). It was first reported from Brittany in France from a collection made in 1954 (Størmer, 1958). However, an earlier specimen from the British Isles was discovered, collected in 1941 (Richards, 1963). It was likely introduced from ships transporting goods from South America in crates padded by mosses. It also could have been introduced via spores attached to shoes of travellers. C. introflexus quickly started spreading eastward, and today it is a dominant species in some biotopes of Western Europe. The eastern edge of its distribution in Europe is currently from Kaliningrad area of western Russia, through southeastern Poland to Slovakia and Austria (Hassel and Söderström, 2005). Campylopus introflexus is closely related and very similar to C. pilifer Brid. These two species were considered conspecific until the study of Giacomini (1955). Other authors have compared the two species and published world- wide distributions (i. e., Frahm 1974; Gradstein and Sipman 1978). Records of C. introflexus discoveries were published from the individual European countries; published are notes about the first records, but about not next spreading in country. A comprehensive work describing expansion of the distribution was published only recently (Hassel and Söderström, 2005). Ecological aspects of the invasion into diverse biotopes and the potential of C. introflexus to compete with native species was studied by various

137 138 distribution of campylopus introflexus in the czech republic (paper 5)

Figure 11.1: Distribution map of Campylopus introﬂexus in the Czech Republic (white circles: localities know up to 2001, white squares: localities recorded in 2001–2003, black squares: localities recorded during 2004–2006).

authors since the second half of 1980’s (e. g., Berg 1985; Equihua and Usher 1993; Biermann and Daniëls 1997). Campylopus introflexus is currently under intensively study using methods of molecular biology and taxonomy. Stech and Dohrmann (2004) described it as monophyletic, little diversified, lineage, based on analysis of rDNA ITS1, ITS2, and the atpB-rbcL spacer. Subsequent work by Stech and Wagner (2005) confirmed this result. Unlike C. introflexus, the closely related and similar species, C. pilifer, is much more genetically variable, and populations form paraphyletic or even polyphyletic taxon. These results can be interpreted such that C. introflexus evolved from one of the C. pilifer lineages (populations from Arabia, Macaronesia, Europe and Africa). Intermediate populations with morphological characteristics of C. introflexus (characterized by low lamellae) and C. pilifer (with straight hair point) were found, and Frahm and Stech (2006) classified such populations as Campylopus pilifer var. brevirameus (Dix.) J.-P. Frahm & Stech.

11.2 distribution of c. introflexus in the czech republic

Campylopus introﬂexus was found in the Czech Republic for the ﬁrst time in 1988 at the South Bohemian site “Borkovicka Blata” (Novotný, 1990a,b). The number of localities grew in subsequent years, and today it is known from 71 sites (Fig. 11.1). Reports were published mostly in the bulletin Bryonora, with summaries published by Soldán (1996, 1997). This contribution updates the current distribution in the Czech Republic. It is likely that the number of sites will increase in future. The list of localities is based on published records, available herbarium specimens including private collections1, and personal communication with

1 herb. E. Loskotová = herb. E. Mikulášková 11.2 distribution of c. introflexus in the czech republic 139 collectors. Locality description is translated to Czech and occasionally made more accurate after visits by the author. Localities are ordered according to phytogeographical divisions of the Czech Republic (Skalický, 1988). Some localities occur at borders between phytogeographical regions and their precise placements can be disputed. References of previously published sites are abbreviated.

Thermophyticum

2. Stˇrední Poohˇrí— a) Žatecké Poohˇrí • Liboˇcany(Nˇemcová, 2000) [not found in 2005]

6. Džbán • údolí Bílichovického potoka (Jupa and Soldán, 1993) [veriﬁed 2. 5. 2005, WGS-84:N50.25411° E13.90811°, 352 m, cover 12 m2, with abundant sporophytes] • Žichovec: poblíž silnice vedoucí z vesnice do Panenského Týnce, cca 600 m SZS od Žichovce na kraji, WGS-84:N50.27528° E13.92364°, 379 m, 2. 5. 2005 leg. E. Loskotová & Z. Soldán, herb. E. Loskotová. [scattered, cover ca. 20 m2, sterile] • Holedec: v borovém hájku u lesního parkovištˇeu silnice 227 ze Žatce do Rakovníka, 2 km JV od vesnice Holedec, WGS-84: N50.26506° E13.58033°, 291 m, 2. 5. 2005 leg. E. Loskotová & Z. Soldán, herb. E. Loskotová. [cover 20 cm2, sterile]

15. Východní Polabí — c) Pardubické Polabí • NPR Bohdaneˇcskýrybník a rybník Matka (Marková, 2002) [not found in 2004] • Chvaletice: okraj výsypek u rudného odkalištˇe,u cesty, WGS-84: N50.03636° E15.44481°, ca 250 m, 17. 11. 2005 leg. O. Peksa & Z. Soldán, herb. E. Loskotová. [cover 2 cushions 25 + 25 cm2, sterile]

Mesophyticum

28. Tepelské vrchy — c) Mnichovské hadce • Prameny: NPR Pluh ˚uvbor, zahlinˇenýkámen v kulturní smrˇcinˇe na severozápadním okraji MZCHÚ; WGS-84:N50° 03.309’E12° 46.817’, 31. 7. 2005 leg. R. Mudrová, herb. Muzeum Ceskéhoˇ lesa v Tachovˇe.

28. Tepelské vrchy — f) Svojšínská pahorkatina • Ošelín: smrkem osázená paseka s brusnicovými spoleˇcenstvyca 0.7 km SSZ od obce, kv. 6243a, 510 m, 25. 8. 2004 leg. R. Mudrová, herb. R. Mudrová, herb. Muzea Ceskéhoˇ lesa v Tachovˇe.[veriﬁed 16. 8. 2005, WGS-84:N49.77836° E12.85117°, 510 m, 9 cushions 4– 100 cm2, sterile] 140 distribution of campylopus introflexus in the czech republic (paper 5)

31. a) Plzeˇnskápahorkatina vlastní • Plzeˇn:u obce Valcha v borovém lese za vˇeznicíBory, 14. 4. 2006 leg. J. Váˇna,herb. PRC. [population ca. 100 cm2, sterile]

36. Ceskýˇ les • Dolní Žandov: po levé stranˇelesní silniˇckyvedoucí do obce Paliˇc, ca. 800 m JZ od Brtné. WGS-84:N50.00483° E12.51686°, 630 m, 16. 8. 2005 leg. E. Loskotová & J. Mikulášek, herb. E. Loskotová. [cover 5 m2, one quarter frequently fruiting, along road to Paliˇc are two additional population up to 10 cm2, sterile, N50.00603° E12.51319°, N50.00686° E12.50961°] • Paliˇc:samoty Nový Svˇet,ve smrˇcinˇe 1.1 km SV od samot, kv. 5941c, 650 m, 20. 7. 2001 leg. R. Mudrová, herb. Muzeum Ceskéhoˇ lesa v Tachovˇe.[not found in 2005] • Vysoká, vrch Dyleˇn(k. 940.3): brusnicová spoleˇcenstva na býv. „drátech“ ca. 1.2 km JZ od vrcholu, 6040b, ca. 850 m, 10. 8. 2002 leg. R. Mudrová, herb. R. Mudrová. [verified 17. 8. 2005, WGS- 84:N49.95928° E12.49608°, 850 m, scattered cushions up to 10 cm2, sterile] • Vysoká: pˇešinapo bývalých drátech S Kosího potoka, 1 km ZJZ od hájovny Háj, lesní mýtina. WGS-84:N49.96089° E12.52650°, 715 m, 17. 8. 2005 leg. E. Loskotová & J. Mikulášek, herb. E. Loskotová. [cover 3 cm2, sterile] • Vysoká, vrch Dyleˇn(k. 940.3): lokalita „Stˇred Evropy“na státní hranici JZ od Dylenˇe,na zahlinˇenémbalvanu, kv. 6040, 830 m, 14. 7. 1999 leg. R. Mudrová, det. M. Vondráˇcek,herb. Muzea Ces-ˇ kého lesa v Tachovˇe.[verified 17. 8. 2005, WGS-84:N49.96211° E12.49056°, 810 m, cover 5 m2, sterile] • Býv. osada Slatina (Ceskýˇ Les): pˇešinapo státní hranici u hraniˇcní- ho kamene 15/10 (na smrkové opadance pˇriokraji pˇešiny), 6041c, 700 m, 13. 7. 2002 leg. R. Mudrová, herb. R. Mudrová. [verified 17. 8. 2005, WGS-84:N49.92872° E12.50769°, 756 m, cover 0.1 m2 plus several minute cushions, sterile] • Tachovská Hut’: okraj rašeliníkové smrˇcinyna pramenech beze- jmenného levostranného pˇrítokuLesního potoka pod vrchem Ti- šina, 6041c, 720 m, 9. 8. 2001 leg. R. Mudrová, herb. R. Mudrová. [verified 17. 8. 2005, WGS-84:N49.93183° E12.55058°, 724 m, 3 mic- rolocalities of 400 cm2, sterile] • Branka: Hvozdný rybník na Lískovém potoce, zrašelinˇelámísta na východním bˇrehu, kv. 6141c, 620 m, 9. 7. 1998 leg. R. Mudrová, det. M. Vondráˇcek,herb. R. Mudrová. [not found in 2005] • Železná: okraj lesní silniˇcky(u kóty 504.6) 2.4 km SZ od Železné, 6341c, 500 m, 25. 6. 2002 leg. R. Mudrová, herb. R. Mudrová. [verified 18. 8. 2005, WGS-84:N49.60394° E12.56614°, 517 m, in vicinity of the crossroad scattered ca. 10 m2, with abundant sporoplytes] 11.2 distribution of c. introflexus in the czech republic 141

• Karlova Hut’: U svícnové jedle (k. 730.2), okraj staré buˇcinyna Z úboˇcívrchu, 6441b, 685 m, 28. 7. 2002 leg. R. Mudrová, herb. R. Mudrová. [veriﬁed 18. 8. 2005, WGS-84:N49.55967° E12.60647°, 691 m, 2 mikrolokalities of 3 m2, sterile] • Rybník: okraj vzrostlé smrˇciny 0.4 km JV od kˇrižovatky silnic Ryb- ník–Závist–Šidlákov, kv. 6442c, 610 m, 15. 8. 2004 leg. R. Mudrová, herb. R. Mudrová. [veriﬁed 18. 8. 2005, WGS-84:N49.50178° E12.68553°, 593 m, several mikrolokalities in 100 m line, populations always less than 2 m2, sterile]

37. Šumavsko-novohradské podh ˚uˇrí— h) Prachatické Pˇredšumaví • Prachatice: PR Kralovické louky, u paty bˇrízyna J okraji menší ˇcástirezervace. Vlhˇcímýtina s coverem bˇrízya smrku 500 m JZ, cca 620 m, 7. 6. 2003 leg. J. Košnar, herb. J. Košnar. [veriﬁed 27. 4. 2005, WGS-84:N49.00494° E14.08364°, 630 m, 3 cushions 60 + 4 + 1 cm2, sterile]

37. Šumavsko-novohradské podh ˚uˇrí— l) Ceskokrumlovskéˇ Pˇredšumaví • Hoˇricena Šumavˇe: Certˇ ˚uvMlýn (3 km SV od Hoˇric),borový les 200–250 m SV od vesnice, na okraji lesa S-42:N3442.08 E5406.15, 660 m, 4. 8. 2000 leg. J. Kuˇcera,herb. J. Kuˇcera.

38. Budˇejovická pánev • Kˇrída:mladší smrk.- bor. les u silnice k osadˇePapírna, asi 1.2 km JV od Kˇrídy, ca. 150 m nad rozcestím s polní cestou k Dobronic ˚um. [S-42:E3463.80,N5469.01], 437 m, na kyselé lesní zemi, polostín, 1. 7. 2006 leg. J. Kuˇcera,herb. J. Kuˇcera.[cca 1 m2, sterile] • Bechynˇe: 2.5 km JZ obce Nuzice, les Výˇrice,pˇrechod mezi stˇrednˇe vzrostlou smrˇcinou a pasekou staršího data založení. WGS-84: N49.26631° E14.44228°, 405 m, 23. 7. 2006 leg. Z. Soldán, herb. E. Loskotová. [cover 3–4 m2, with abundant sporophytes]

39. Tˇreboˇnskápánev • PR Borkovická blata (Novotný, 1990a,b; Hillermannová, 1994) [verified 9. 11. 2004, WGS-84:N49.23739° E14.62597°, 428 m, in the area of the bog tens to hundreds of m2, with abundant sporophytes] • Borovany: 1 km JV obce Hluboká u Borovan, PR Žemliˇcka.Lesní mýtina. S-42:N3477.5 E5417.85, 475 m, 16. 10. 1999, 16. 11. 2001 leg. J. Kuˇcera,herb J. Kuˇcera.[verified 11. 10. 2004, cover 1 m2, sterile] • Tˇreboˇn: 1.2 km S obce Spolí; 0.5 km JZ Spolského mlýna v PR V rájích na kraji malého palouku s rašeliništˇem.S-42:N3478.84 E5428.26, 440 m, 31. 3. 2001, 23. 10. 2001 leg. J. Kuˇcera,herb. J. Kuˇcera.[verified 13. 10. 2004, cover 1 m2, sterile] • Dvory nad Lužnicí: NPR Žofinka (Kuˇcerová et al., 2000) [up to 1 % of 400 m2] 142 distribution of campylopus introflexus in the czech republic (paper 5)

41. Stˇrední Povltaví • Dˇedovice: vytˇežené rašeliništˇe "Jezero", ca. 1.5 km SSZ obce, Z ˇcást(530 m SZ od hájenky ‘Doupata’), S-42:N3438.762 E5475.211, 406 m, 11. 10. 2003 leg. J. Kuˇcera,herb. J. Kuˇcera.

46. Labské pískovce — d) Jetˇrichovické skalní mˇesto • Mezná (Nˇemcová, 2000) [veriﬁed 16. 5. 2005, WGS-84:N50.87331° E14.29647°, 269 m, by the road 2 cushions 1 + 0.5 m2, with abundant sporophytes, by clearcut 3 m2, with abundant sporophytes] • Kyjov: na Z svahu rokle na pravém bˇrehu ˇríˇckyKˇrenice 1.5 km SZ obce Kyjov (cesta odboˇcujícíu ˇreky vlevo od cesty k Jeskyni víl, WGS-84:N50.91911° E14.44447°, 431 m, 16. 5. 2005 leg. E. Losko- tová & I. Marková, herb. E. Loskotová. [scattered in area of 10 m2, 1 cushion with sporophytes] • 500 m SZ Kyjova, na okraji porostu u Vˇresové cesty, WGS-84: 50°54’59’’N; 014°27’18’’E, 430 m, 4. 10. 2005, leg. I. Marková, herb. I. Marková. • 1.3 km SV Turistického mostu, na okraji smrkového porostu u lesní cesty 20 m nad od odboˇckoudo Rákosového dolu (smˇerVlˇcí hrádek), WGS-84: 50°55’’38’’N; 014°26’12’’E, 350 m. 1. 9. 2005, leg. I. Marková, herb. I. Marková. • Kyjovské údolí, boˇcní strž na levé stranˇepotoka. Lokalita je v horní ploché ˇcástinad strží na kraji vzrostlé smrˇcinypoblíž lesní cesty, WGS-84:N50°55’07.3’’, E14°26’50.0’’, 970 m, 16. 4. 2005 leg. Z. Soldán, herb. E. Loskotová. [cover 4–5 cm2, sterile, not found in 2006] •NP CŠ:ˇ ob. Janov, ve smrˇcinˇe 1.5 km SV obce Janov, na lesní cestˇe v mladé smrˇcinˇe, 245 m, 10. 11. 2005, leg. I. Marková, herb. I. Marková. •NP CŠ:ˇ ob. Mezní Louka, lesní porosty 900 m S obce Mezní Louka, na lesní cestˇeu smrkového mlází, WGS-84: 50°52’50’’N; 014°19’13’’E; 50°52’39’’N, 014°18’52’’E; 50°52’45’’N, 014°18’52’’E; 270–360 m, 13. 10. 2005, leg. I. Marková, herb. I. Marková.

49. Frýdlantská pahorkatina • Dolní Rasnice:ˇ lesní komplex mezi Dolní Rasnicíˇ a Bulovkou, paseka nad potokem, asi 30 m od lesní cesty. WGS-84:N50.96231° E15.15478°, 400 m, 3. 5. 2005 leg. O. Peksa, herb. E. Loskotová. [cover 200 cm2, sterile]

50. Lužické hory • Jedlová (Gerišová, 1999) [veriﬁed 16. 5. 2005, WGS-84:N50.84975° E14.57194°, 568 m, 3 cushions of 0.5 m2, ca. 30 small cushions up to 5 cm2, sterile, additional populations ca. 200 m N50.84947° E14.57250° — 2 cushions of 1.5 m2, and 8 cushions 2–10 cm2, sterile] 11.2 distribution of c. introflexus in the czech republic 143

51. Polomené hory • Kokoˇrínsko: PR StránˇeTruskavenského dolu, Truskavenský d ˚ul 2.5 km Z Kokoˇrína, exponovaná hlava pískovcových skal, 20 m nad dnem rokle, S-42,M33:N5589.565°, E3466.964°, 274 m, 22– 23. 9. 2006 leg. J. Kuˇcera& Z. Soldán, herb. J. Kuˇcera(no. 12639). [min. 2 mikrolokalities 0.5 m2, sterile]

52. Ralsko-bezdˇezskátabule • Ceskolipsko,ˇ Veselí (Zmrhalová and Nˇemcová, 1994)

62. Litomyšlská pánev • PR Maštale, skalní útvar Kazatelna (Košnar, 2004) [verified 29. 10. 2004, WGS-84:N49.83019° E16.11375°, 425 m, cover 25 cm2, sterile; r. 2005 cover 10 cm2] • PR Maštale, údolí V Kvíˇcalnici(Košnar, 2004) [not found in 2004] • údolí Proseˇcského potoka (Košnar, 2004) [verified 29. 10. 2004, WGS-84:N49.81769° E16.12678°, 492 m, cover 0.5 m2, sterile; verified 2005 cover 1 + 0.1 + 0.1 m2, with sporophytes] • Proseˇc: PR Maštale — 1.5 km Z obce Budislav — 1 km SV od hájovny Posekanec, 200 m od Nového rybníka, WGS-84:N49.80517° E16.15333°, 522 m, 29. 10. 2004 leg. E. Loskotová, herb. E. Losko- tová. [16 cushions of max. 9 cm2, sterile; in 2005 10 new cushions up to 5 cm2 found scattered in the area, 200 m further by the road, N49.80661° E16.15167° new population of 0.5 m2, sterile] • Proseˇc:PR Maštale — 1.5 km JV obce Bor u Skutˇce— 20 m od ústí Oslí chodby, WGS-84:N49.81686° E16.14192°, 522 m, 29. 10. 2004 leg. E. Loskotová, herb. E. Loskotová. [cover 1 m2, sterile; in 2005 carpet disturbed, 10 cm2 survived]

63. Ceskomoravskéˇ mezihoˇrí— i) Hˇrebeˇcovská vrchovina • Svitavy: poblíž obce Hˇrebeˇc(2.1 km J obce) na okraji mladé výsad- by smrku a modˇrínuuprostˇred vzrostlého smrkového lesa u lesní cesty, leg. S. Kubešová, herb. BRNM. [veriﬁed 25. 7. 2005, WGS-84: N49.72772° E16.57578°, 616 m, 0.2 m2, sterile]

64. Ríˇcanskáplošinaˇ — b) Jevanská plošina • Radlík (Franklová, 1998) [veriﬁed 25. 10. 2004, WGS-84:N49.91526° E14.51150°, 442 m, scattered on 2 m2, with abundant sporophytes, on road 100 m apart N49.91515° E14.51272° additional 0.4 m2, capsules found rarely; in 2005 cover on clearcut 4 m2, along road 0.5 m2] • Jílové u Prahy: 2.1 km VSV obce Petrov — štˇerkové parkovištˇe v lese u silnice mezi Jílovým a Petrovem. Pˇrechod vzrostlé smr- ˇcinydo mladé borovicové výsadby cca 200 m JV od parkovištˇe, 2003 leg. H. Franklová, herb. H. Franklová. [veriﬁed 25. 10. 2004, WGS-84:N49.89304° E14.46030°, 400 m, cover 2 mikrolokalities 10 + 1 m2, sterile; in 2005 cover 17 + 2 m2] 144 distribution of campylopus introflexus in the czech republic (paper 5)

• Psáry: poblíž silnice z Psár do Radlíka na lesní mýtinˇe, 2005 leg. Z. Soldán, herb. E. Loskotová. [veriﬁed 5. 5. 2005, WGS-84:N49.92131° E14.50839°, 409 m, cover 1 m2, rarely with sporophytes]

67. Ceskomoravskáˇ vrchovina • Popice u Jihlavy (Berka, 2006) [veriﬁed 12. 11. 2004, WGS-84: N49.34095° E15.53800°, 619 m, cover 1 m2, sterile] • Blato: na S bˇrehu H ˚ureckého rybníka na obnažené rašelinˇe,WGS- 84:N49.04654°, E15.15868°, 650 m, 8. 6. 2006 leg. E. Loskotová, herb. E. Loskotová. [1 cushion 50 cm2 + 6 small cushions up to 20 cm2, sterile] • Albeˇr: na JV bˇrehu rybníka Osika mezi smrkovým zmlazením, WGS-84:N49.02867°, E15.15497°, 548 m, 9. 6. 2006 det. E. Losko- tová, herb. E. Loskotová. [2 cushions 15 + 45 cm2, sterile]

70. Moravský kras • Vilémovice: pˇrihranˇePustého žlebu, N49°22’47’’, E16°43’43’’, tráv- ník s keˇrina okraji lesa, 6. 8. 2005, leg. S. Kubešová, herb. BRNM. [veriﬁed 22. 10. 2005, ca. 430 m, cover 0.3 m2, sterile]

71. Drahanská vrchovina — b) Drahanská plošina • Boskovice: Horní rybník 2.5 km J obce Benešov, 300 m od levé strany silnice do Suchého. Na SV bˇrehu rybníka na okraji vzrostlé- ho smrkového lesa, 1998 leg. S. Kubešová, herb. BRNM. [veriﬁed 1. 11. 2004, WGS-84:N49.49239° E16.76822°, 662 m, 14 cushions up to 10 cm2, sterile; in 2005 at least 16 new cushions found, up to 4 cm2; in 2006 ca. 35 cushions present]

80. Stˇrední Pobeˇcví— b) Veˇrovické vrchy • Hostašovice (Duda, 2002) [veriﬁed 22. 10. 2005, WGS-84: N49.50533° E17.99122°, 400 m, 10 cushions 4–100 cm2, sterile]

83. Ostravská pánev • Trnávka (Duda and Duda, 2006)

Oreophyticum

85. Krušné hory • Hora Sv. Šebestiána: Novodomské rašeliništˇe,ˇcástJezerní rašelini- štˇe,okraj lesní cesty, 2003 leg. L. Nˇemcová, herb. L. Nˇemcová. • Pˇrebuz: rybník Licha, menší porosty na ulehlé rašelinˇev prostoru SV bˇrehu i na samotné hrázi. WGS-84:N50.39117° E12.62938°; 10. 9. 2005 leg. R. Mudrová, herb. Muzeum Ceskéhoˇ lesa v Tachovˇe.

88. Šumava — a) Královský Hvozd 11.2 distribution of c. introflexus in the czech republic 145

• Keply: 0.5 km od silnice ˇc. 27 z Klatov do Železné Rudy, v I. zónˇe Zh ˚uˇrské louky na pravém bˇrehu ˇreky Kˇremelné, na kraji lesa. WGS-84:N49.18697° E13.31267°, 931 m, 29. 4. 2005 leg. E. Losko- tová, herb. E. Loskotová. [cushion 9 cm2, sterile; in 2006 two cushions of 30 cm2, 100 m apart additional 0.1 m2] • I. zóna Nový Brunst (Holá, 2006) [veriﬁed 15. 4. 2005, scattered cover 10 m2, sterile; in 2006 cover 12 m2, additional 2 population in the neighborhood of 6 + 1.5 + 1 m2, and 400 + 25 cm2] • nad I. zónou Gerl ˚uvpotok(Holá, 2006) [veriﬁed 28. 9. 2005, 2 cushions 15 + 800 cm2, sterile] • I. zóna V mokˇrinách(Holá, 2006) [in 2006 scattered in area of 2 m2, with sporophytes, 20 m appart additional one cushion 12 cm2, sterile]

88. Šumava — b) Šumavské Plánˇe • Popelná: ve svahu 300 m JZ pod vrcholem Valy na pˇrechodu sta- rého smíšeného porostu do mladého smrkového lesa na malé pasece. WGS-84:N49.09574° E13.58812°, 942 m, 12. 4. 2005 leg. E. Loskotová, herb. E. Loskotová. [2 mikrolokalities 4 + 1 m2, sterile, in 2006 cover 8 + 2 m2] • Kvilda: Na J hranici I. zóny Tetˇrevská slat’, pˇrechod staré mýtiny ve vzrostlý remízek u slati, obnažený kus rašeliny, N49.02395° E13.54857°, 1140 m, 30. 10. 2006 leg. E. Mikulášková, herb. E. Mi- kulášková. [2 cushions 25 + 42 cm2, sterile] • Strážný: I. zóna Castá,ˇ 30 m J od „Jelení stezky“ na levém bˇrehu Castéˇ na pˇrechodu smrkového remízku a louky, WGS-84 cca N48.91801° E13.67263°, 905 m, 1. 11. 2006 leg. E. Mikulášková, herb. E. Mikulášková. [3 cushions 50 + 25 + 15 cm2, sterile]

88. Šumava — c) Javorník • Kašperské Hory: v údolí potoka mezi KH a hradem Kašperkem (pravý pˇrítok Otavy u Radešova). Na strhnutém pravém bˇrehu potoka asi 150 m po proudu od cesty na hrad. WGS-84:N49.15708° E13.55600°, 667 m, 13. 4. 2005 leg. E. Loskotová, herb. E. Loskotová. [scattered cover 20 cm2, sterile]

88. Šumava — g) Hornovltavská kotlina • Pˇrední Zvonková: Vytˇeženézarostlé rašeliništˇe 1.6 km V obce Racín na pravém bˇrehu vodní nádrže Lipno. 730 m, 29. 3. 1999 leg. J. Kuˇcera,herb. J. Kuˇcera.[veriﬁed 13. 10. 2004, WGS-84:N48.70867° E14.04464°, 732 m, cover 0.25 m2, sterile] • Lenora: rozlehlé vytˇeženéa v souˇcasnédobˇerevitalizované rašeli- ništˇeu železniˇcnízastávky Soumarský most. WGS-84:N48.90787° E13.83383°, ca. 750 m, 10. 11. 2005 leg. E. Loskotová, herb. E. Losko- tová. [tens to hundreds of m2, with abundant sporophytes]

91. Žd’árské vrchy 146 distribution of campylopus introflexus in the czech republic (paper 5)

• Žd’árské vrchy: NPR Radostínské rašeliništˇe, 1.5 km SVS stˇredu obce Radostín, na okraji borovicového lesíka, na obnažené rašelin- né p ˚udˇe[WGS-84 49°39’43 − 55’’N, 15°53’01 − 20’’E, kv. 636lad], ca. 615 m, 18. 10. 2002 leg. S. Kubešová, herb. BRNM. [not found in 2004]

93. Krkonoše • Pec pod Snˇežkou: Certovaˇ zahrádka — V svah Studniˇcníhory, 3 km S od Pece pod Snˇežkou,sut’ové pole 3 m pod rulovým ma- sívem. S-42:N3551.244 E5622.096, 1030 m, 24. 9. 2002 leg. J. Kuˇcera & B. Buryová, herb. J. Kuˇcera.[veriﬁed 2. 11. 2004, 6 cushions up to 10 cm2, sterile; in 2005 8 cushions up to 10 cm2 plus groups if individual stems on soil in ﬁssures in area 1 m2]

94. Teplicko–adršpašské skály • Vlˇcírokle (Nˇemcová, 1999) [not found in 2006]

11.3 discussion and conclusion

Campylopus introflexus is currently known from 71 sites in the Czech Re- public. The majority of the localities are part of Czech–Moravian Meso- phyticum, about 9 % occur in Czech Thermophyticum, 20 % occur in Czech Oreophyticum. Only two sites are known from Carpathian Mesophyticum. Campylopus introflexus grows mostly in hilly to submontane areas (77 % of collections), less often in lowland areas (for example, Chvaletice, údolí Bíli- chovického potoka, Bohdaneˇcskýrybník) or in the mountains (for example Studniˇcníhora, Adršpaško–teplické skály, Kvilda). The altitudinal distribution of sites is summarized in Fig. 11.2. It is currently assumed that the species spread to the Czech Republic from Germany. Fig. 11.1 clearly shows that majority of the sites occurs in the south–western and central part of Bo- hemia. On the contrary, only isolated occurrences are known from Moravia (for example, in CHKO Beskydy or in CHKO Moravský Kras). Most of the records are in areas of the south Bohemian basin, south Bohemian mountains, north Bohemian sandstones, east Bohemian chalk basin, and in so called Sázavsko–chrudimský okruh (Skalický, 1988). There are three localities currently known from Slovakia, near the border with Czech Repub- lic (Holotová and Šoltés, 1997; Janovicová, 1998; Janovicová and Kubinská, 2002). Only one of the three sites was confirmed/relocated during field investigation of the author in 2004. Trends in the continuous colonization of the Czech Republic are apparent in the distribution map. The oldest localities are randomly scattered in the whole area of the Czech Republic. This is consistent with the assumption of predominant spreading by light-weight spores by wind. Given the random distribution of just a few initial populations, it can be deduced that the concentration of spores in the atmosphere probably wasn’t very high; source populations in close proximity to the Czech Republic were not as often bearing sporophytes as they are today. The subsequent phase shows an increase in site numbers, especially in the western part of the country, and again ran- 11.3 discussion and conclusion 147

Figure 11.2: Altitudinal distribution of Campylopus introﬂexus localities in the Czech Republic.

dom expansion into central and eastern parts. This trend is most apparent in recent years. An increase in frequency of discovery of new locations in areas with sporophyte-bearing populations is also shown. It can be assumed that a large number of undetected populations exists in our area, because Campylopus introﬂexus grows in undercollected habitats not often visited by bryologists. It is likely that we will meet this moss more and more often at margins of spruce and pine forests (especially pine forests on sandy soils), and in disturbed peatlands. The presence of multiple populations producing sporophytes in our country is one of the main reasons of spreading in recent time (Fig. 11.3). Short distance colonization is achieved mainly by vegetative reproduction. Vegeta- tive diaspores are formed very soon after establishment at a new site; they enable more effective colonization of all suitable habitats in the immediate vicinity. This is apparent for example at sites in Ceskýˇ Les, or in disturbed peatbogs (“Borkovická blata”, “Soumarský most”), where Campylopus intro- ﬂexus forms large compact carpets and covers large areas. Increases in the number of new sites in last ten years is likely also caused by increased attention to this species. acknowledments

I would like to thank all the collectors for making their specimens and information about the ﬁndings available; especially to ZdenˇekSoldán for col- laboration on this project. Thanks to Jan Kuˇcerafor valuable comments on the text. This work was ﬁnantially supported by Grant Agency of Charles University, project no. 258/2004, and partially by Grant Agency of the Czech Republic, project no. 206/03/H 137. 148 distribution of campylopus introflexus in the czech republic (paper 5)

Figure 11.3: Cumulative graph showing increase of Campylopus introﬂexus records over the years.

summary

The neophytic moss Campylopus introﬂexus was for the ﬁrst time reported from the Czech Republic in 1988 (Novotný, 1990a), from a locality in South Bohemia. Nowadays it is known from 71 localities in the Czech Republic — 71 % of them lying in the moderately warm regions (Mesophyticum), 20 % in the regions of montane vegetation (Oreophyticum), and 9 % in the regions of thermophilous vegetation (Termophyticum). It is more frequent in the west and south parts of the Czech Republic, only scattered occurrences have been recorded from the eastern part of the country. Most localities lie in the altitudes between 230 and 800 m a.s.l. The rapid increase of localities after 1998 is attributed to both faster spreading rate after the population became fertile and a greater attention to this species by bryologists in the last years.

references

Berg, C. (1985). “Zur Ökologie der neophytischen Laubmoosart Campylopus introflexus (Hedw.) Brid. in Mecklenburg.” Archiv der Freunde der Naturge- schichte 25, pp. 117–126. Berka, T. (2006). “Mechorosty Brtnické vrchoviny [Bryophytes of Brtnická vrchovina highlands].” MA thesis. Depon in: Palackého University, Olo- mouc, p. 113. Biermann, R. and F.J.A. Daniëls (1997). “Changes in a lichen-rich dry sand grassland vegetation with special reference to lichen synusiae and Campy- lopus introflexus.” Phytocoenologia 27, pp. 257–273. Duda, J. (2002). “Campylopus introflexus (Hedw.) Brid. na východní Moravˇe [Campylopus introflexus (Hedw.) Brid. in Eastern Moravia].” Casopisˇ Slezského Muzea, Ser. A 51, p. 185. 11.3 discussion and conclusion 149

Duda, J. and J. Duda (2006). “Mechorosty území mezi Pˇríborem a Brušperkem (severní Morava) [Bryophytes between Pˇríborand Brušperk villages (northern Moravia)].” Casopisˇ Slezského Zemského Muzea, Ser. A 55, pp. 83–86. Equihua, M. and M.B. Usher (1993). “Impact of carpets of the invasive moss Campylopus introflexus on Calluna vulgaris regeneration.” Journal of Ecology 81, pp. 359–365. Frahm, J.P. (1974). “Zur Unterscheidung und Verbreitung von Campylopus introflexus (Hedw.) Brid. und C. polytrichoides De Not.” Revue bryologique et lichénologique 40, pp. 33–44. –(1984). “Phytogeography of European Campylopus species.” Proceedings of the Third Meeting of Bryologists from Central and East Europe. Ed. by J. Váˇna. Praha: Univerzita Karlova, pp. 191–212. Frahm, J.P. and M. Stech (2006). “The taxonomic status of intermediate forms of Campylopus introflexus (Hedw.) Brid. and C. pilifer Brid.(Dicranaceae, Bry- opsida) newly discovered in Europe.” Cryptogamie. Bryologie 27, pp. 213– 223. Franklová, H. (1998). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 22. Ed. by Z. Soldán, pp. 16–17. Gerišová, Z. (1999). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 23. Ed. by Z. Soldán, pp. 9–10. Giacomini, V. (1955). “Sull’autonomia specifica e sul ciclo di forme de Campy- lopus polytrichoides De Not.” Atti Istituto di Botanica, Università Pavia, Labo- ratorio Di Crittogama 13, pp. 45–83. Gradstein, S.R. and H.J.M. Sipman (1978). “Taxonomy and world distribution of Campylopus introflexus and C. pilifer (= C. polytrichoides): a new synthesis.” Bryologist 81, pp. 114–121. Hassel, K. and L. Söderström (2005). “The expansion of the alien mosses Orthodontium lineare and Campylopus introflexus in Britain and continental Europe.” Journal of the Hattori Botanical Laboratory 97, pp. 183–193. Hillermannová, K. (1994). “Mechorosty Borkovických blat [Bryophytes of Borkovická blata peatbog].” MA thesis. Depon in: Charles University, Prague, p. 53. Holotová, E. and R. Šoltés (1997). “Campylopus introflexus new moss species to the Slovakian moss flora.” Biologia 52, pp. 494–494. Holá, E. (2006). “Bryoflóra horního toku Kˇremelné na Šumavˇe[Bryophytes of upper part of Kˇremelná river in Bohemian Forest].” Silva Gabreta 12, pp. 109–131. Janovicová, K. (1998). “Zaujímavé bryofloristické nálezy zo Slovenska [Inter- esting bryophytes records from Slovakia].” Bryonora 21, pp. 2–3. Janovicová, K. and A. Kubinská (2002). “Súˇcasnýstav poznania bryoflóry Záhorskej nížiny (JZ Slovensko) [Bryophytes of Záhorská nížina lowland].” Bulletin Slovenskej Botanickej Spoloˇcnosti 24, pp. 55–59. Jupa, I. and Z. Soldán (1993). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 11. Ed. by Z. Soldán, p. 13. Košnar, J. (2004). “Bryofloristicko-ekologická studie pˇrírodní rezervace Maš- tale [Bryofloristic and ecological study of Maštale Nature reserve].” MA thesis. Depon in: University of South Bohemia, Ceskéˇ Budˇejovice, p. 67. 150 distribution of campylopus introflexus in the czech republic (paper 5)

Kuˇcerová, A., L. Rektoris and K. Pˇribáˇn(2000). “Vegetation changes of Pinus rotundata bog forest in the Žofinka” Nature Reserve, TˇreboˇnBiosphere Reserve.” Pˇríroda 17, pp. 119–138. Marková, I. (2002). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 29. Ed. by B. Buryová, pp. 21–22. Novotný, I. (1990a). “Campylopus introflexus (Hedw.) Brid. – nový druh mechu v Ceskoslovensku.”ˇ 2. Bryologicko–lichenologické dny v Opavˇe(19.–21.9.1989) – souhrny referát˚u. Ed. by Anonymus. Vol. 4, pp. 4–5. –(1990b). “The moss Campylopus introflexus (Hedw.) Brid. new to Czechoslo- vakia.” Acta Musei Moraviae, Scientiae Naturales 75, pp. 237–238. Nˇemcová, L. (1999). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 24. Ed. by Z. Soldán, p. 29. –(2000). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 26. Ed. by Z. Soldán and B. Buryová, p. 13. Richards, P.W. (1963). “Campylopus introflexus (Hedw.) Brid. and C. polytrichoides De Not. in the British Isles; a preliminary account.” Transactions of the British Bryological Society 4, pp. 404–417. Skalický, V. (1988). “Regionálnˇe fytogeografické ˇclenˇení. [Regional phytogeographical division].” In: Kvˇetena Ceskéˇ socialistické republiky 1. Ed. by S. Hejný and B. Slavík. Prague: Academia, pp. 103–121. Soldán, Z. (1996). “Rozšíˇrení neofytických mech ˚u Campylopus introflexus a Or- thodontium lineare v Ceskéˇ republice [Distribution of Campylopus introflexus a Orthodontium lineare in Czech Republic].” Brynora 18, pp. 10–19. –(1997). “Invazní mechorosty. [Invasive bryophytes].” Zprávy Ceskéˇ botanické spoleˇcnosti 14, pp. 33–39. Stech, M. and J. Dohrmann (2004). “Molecular relationships and biogeography of two Gondwanan Campylopus species, C. pilifer and C. introflexus (Dicranaceae).” Monographs in Systematic Botany from the Missouri Botanical Garden 98, pp. 415–431. Stech, M. and D. Wagner (2005). “Molecular relationships, biogeography, and evolution of Gondwanan Campylopus species (Dicranaceae, Bryopsida).” Taxon 54, pp. 377–382. Størmer, P. (1958). “Some mosses from the phytogeographical excursion 1-9 through the Armorican massive in 1954.” Revue bryologique et lichénologique 27, pp. 13–16. Zmrhalová, M. and L. Nˇemcová (1994). “Zajímavé nálezy. Campylopus introflexus (Hedw.) Brid. [Interesting records].” Bryonora 14. Ed. by Z. Soldán, pp. 20–21. OPTICALDISC A

• PhD. Thesis (PDF)

• Abstract in English (PDF)

• Abstract in Czech (PDF)

• Dataset A for Paper 1 (XLS)

151