Biodiversity of Klebsormidium (Streptophyta) from alpine biological soil crusts (Alps, Tyrol, Austria, and Italy)

Forty Klebsormidium strains isolated from soil crusts of mountain regions (Alps, 600–3,000 m elevation) were analyzed. The molecular phylogeny (internal transcribed spacer rDNA sequences) showed that these strains belong to clades B/C, D, E, and F. Seven main (K. flaccidum, K. elegans, K. crenulatum, K. dissectum, K. nitens, K. subtile, and K. fluitans) and four transitional morphotypes (K. cf. flaccidum, K. cf. nitens, K. cf. subtile, and K. cf. fluitans) were identified. Most strains belong to clade E, which includes isolates that prefer humid conditions. One representative of the xerophytic lineage (clade F) as well as few isolates characteristic of temperate conditions (clades B/C, D) were found. Most strains of clade E were isolated from low/middle elevations (<1,800 m above sea level; a.s.l.) in the pine‐forest zone. Strains of clades B/C, D, and F occurred sporadically at higher elevations (1,548–2,843 m a.s.l.), mostly under xerophytic conditions of alpine meadows. Comparison of the alpine Klebsormidium assemblage with data from other biogeographic regions indicated similarity with soil crusts/biofilms from terrestrial habitats in mixed forest in Western Europe, North America, and Asia, as well as walls of buildings in Western European cities. The alpine assemblage differed substantially from crusts from granite outcrops and sand dunes in Eastern Europe (Ukraine), and fundamentally from soil crusts in South African drylands. Epitypification of the known species K. flaccidum, K. crenulatum, K. subtile, K. nitens, K. dissectum, K. fluitans, K. mucosum, and K. elegans is proposed to establish taxonomic names and type material as an aid for practical studies on these algae, as well as for unambiguous identification of alpine strains. New combination Klebsormidium subtile (Kützing) Mikhailyuk, Glaser, Holzinger et Karsten comb. nov. is made.

vascular plants is limited by low water availability and extremes of temperature (Belnap and Lange 2001).
The algal species composition of cryptogamic crusts in different regions is astonishingly similar and typically includes taxa that are adapted to extreme environments, mostly cyanobacteria (West 1990, Sant'Anna and Azevedo 1991, Rosentreter and Belnap 2001. However, some studies have indicated that the climatic conditions of some regions influence the composition of crusts. Biological soil crusts from deserts of North America (temperate zone) include many green algae (Johansen et al. 1982), whereas significant parts of cryptogamic crusts from Antarctica are composed of prasiolacean green algae (Green and Broady 2001). Filamentous streptophycean green algae (Zygogonium K€ utzing and Klebsormidium P.C. Silva, Mattox & Blackwell) dominate the soil crusts of temperate Europe (Luke sov a and Kom arek 1987, Hoppert et al. 2004), while diatoms and conjugating green algae are the main components of soil crusts in the tundra zone (Skuja 1964).
The Alps constitute an extreme habitat for photosynthetic organisms, including terrestrial algae, due to their harsh climatic and environmental conditions: wide seasonal and diurnal temperature fluctuations, occasional frost in summer, strong impact of wind causing drought and abrasion, and rarefied atmosphere, as well as high levels of insolation including intense ultraviolet radiation that increases with elevation (L€ utz and Engel 2007). Investigation of algae present in alpine soil crusts is important because of their possible adaptation to withstand the extreme environmental conditions that are typical for mountains, and because of their essential role in alpine terrestrial ecosystems. Nevertheless, data on the composition and ecology of these algal assemblages are sparse in comparison with the knowledge of soil crusts from arid and polar regions (T€ urk andG€ artner 2001, Karsten and. Investigations of the species composition, distribution, and ecology of algae from alpine soils began in the 1960s (Pitschmann 1963, Reisigl 1964, 1969, Trenkwalder 1975, Vinatzer 1975. The information was completed and summarized by Reisigl (1964) and later by T€ urk and G€ artner (2001) and G€ artner (1995, 2014). Some data on the species composition and characteristics of alpine soil algae are included in the contributions of G€ artner (2004), Tschaikner et al. (2007Tschaikner et al. ( , 2008 and Tschaikner (2008). As a result, new species of terrestrial algae from different genera (especially Heterococcus Chodat, Myrmecia Printz, Leptosira Borzi, Botrydiopsis Borzi, Trochisciopsis Vinatzer, Coelastrella Chodat, and others) were described, and some data on the ecology and distribution of known taxa were provided (Ettl and G€ artner 2014). T€ urk and G€ artner (2001) provided information about the species composition of algae of biological soil crusts in the Alps, which contain abundant filamentous streptophycean and xanthophycean algae, along with cyanobacteria.
Members of the filamentous green alga genus Klebsormidium (Klebsormidiophyceae, Streptophyta) are one of the essential components of soil crusts. These algae are widely distributed in terrestrial habitats worldwide (Hoffmann 1989, Lokhorst 1996, Rindi et al. 2008. The reasons for the ability of Klebsormidium to survive and develop high biomass under extremely dry, insolated, hot, or cold terrestrial conditions are not completely understood, but more recent publications indicate a high potential for acclimation to fluctuations in water availability, temperature, and solar radiation (Holzinger and , Karsten and Holzinger 2014, Kitzing et al. 2014. A recent transcriptomic approach revealed that all prerequisites for living in a terrestrial habitat (e.g., ROS protection mechanisms, and up-regulation of enzymes involved in the biosynthesis of the raffinose family of oligosaccharides for osmotic protection) are present in Klebsormidium crenulatum (K€ utzing) Lokhorst . The presence of these prerequisites was further supported by a genome-sequencing study of Klebsormidium flaccidum (K€ utzing) P.C. Silva, Mattox & W.H. Blackwell (Hori et al. 2014). However, the taxonomy of Klebsormidium is problematic because of high morphological uniformity and plasticity as well as, probably, a high degree of hidden cryptic diversity (Rindi et al. 2008, Skaloud and Rindi 2013. Despite the many investigations on morphology, ontogeny, ultrastructure, and phylogeny of Klebsormidium (Stewart and Mattox 1975, Lokhorst and Star 1985, Lokhorst 1996, Skaloud 2006, Rindi et al. 2008, Sluiman et al. 2008, Skaloud and Rindi 2013, Skaloud et al. 2014, Ry s anek et al. 2015 and references therein), unambiguous identification of species in the genus remains difficult. Species delimitation within Klebsormidium and even the phylogenetic position of the type species, K. flaccidum, are still under debate (Rindi et al. 2011, Skaloud and Rindi 2013, Skaloud et al. 2014. The type material of most of the known Klebsormidium species is represented by herbarium sheets (Lokhorst 1996, Rindi et al. 2011, and it is urgently necessary to provide epitypification and designation of the various taxa based on algal strains that are deposited and accessible in culture collections. Our investigation is part of a broader study on the ecology and ecophysiological performance of Klebsormidium as a component of alpine biological soil crusts of the Tyrolean Alps , Holzinger et al. 2011, Kaplan et al. 2012, Karsten and Holzinger 2012, Kitzing et al. 2014), highlighting its biodiversity using an integrative approach. We isolated strains of Klebsormidium from biological soil crusts collected in mountain regions at different elevations (Tyrolean Alps, Austria and Italy, between 600 and 3,000 m a.s.l.), to undertake morphological identifications, to evaluate their genetic diversity, and finally to correlate the biodiversity with the elevational and ecological/biogeo-BIODIVERSITY OF KLEBSORMIDIUM FROM ALPINE SOIL CRUSTS graphic distributions. Another goal was the epitypification of some well-described Klebsormidium species, to link names to type material, as an improved baseline for future taxonomic studies on original and newly isolated material.

MATERIAL AND METHODS
Collection sites, strain isolation and culture conditions. Most of the 40 strains of Klebsormidium were isolated from samples of the top 5 mm of alpine biological soil crust collected at different locations in Tyrol (Austria, Italy) during spring 2009; a few samples were collected as biofilms covering rock surfaces or artificial stone substrates. Two strains were isolated in 2007, and five isolates were provided by Prof. Georg G€ artner, University of Innsbruck, Austria. The strain number, origin, and habitat of all the Klebsormidium isolates are provided in Table S1 in the Supporting Information.
Klebsormidium from the field samples were purified and established as unialgal cultures by the procedure of Tschaikner (2008). All Klebsormidium cultures were cultured on solid (1.5% agar) and liquid modified Bold's Basal Medium (Starr and Zeikus 1993)  Other Klebsormidium strains involved in the investigation. Epitypification of known Klebsormidium species was undertaken to link names to their respective type specimens, thus allowing consistent species identification of strains isolated from alpine soil crusts. Eight strains from the Sammlung von Algenkulturen, University of G€ ottingen, Germany (SAG: Friedl and Lorenz 2012, www.epsag.uni-goettingen.de) were used for comparison with the alpine isolates. Comprehensive information on these strains was previously presented by Rindi et al. (2011).
Light microscopy and morphological characterization. Young (2to 3-week old) and old (2-to 3-month old) cultures of all Klebsormidium strains were characterized morphologically using an Olympus IX70 light microscope (Olympus Europa Holding, Hamburg, Germany) with Nomarski differential interference optics. Filament morphology was documented with a ColorView II camera (Soft Imaging System GmbH, M€ unster, Germany) using the imaging software analySIS (Soft Imaging System GmbH). The identification keys of Starmach (1972), Moshkova (1979), G€ artner (1995), Hind ak (1996) and Lokhorst (1996) were used to identify the taxa prior to the morphological studies. Filament and cell shape and size, morphology of chloroplasts and pyrenoids, presence of H-like fragments of cell wall and mucilage, growth habit on solid and in liquid medium, as well as modifications of all these characters during the life cycle were documented.
DNA isolation, PCR, sequencing, and phylogenetic analyses of the Klebsormidium strains. Genomic DNA was extracted using the DNeasy Plant Mini Kit (Qiagen GmbH, Hilden, Germany). Internal transcribed spacer (ITS) rDNA was amplified in a thermocycler (T gradient Thermoblock, Biometra, Germany) according to Luo et al. (2006) using the Taq PCR Mastermix Kit (Qiagen GmbH) with the primers (EAF3 and ITS055R) published by Marin et al. (2003); PCR products were purified using the Qiagen PCR purification kit (Qiagen GmbH), following the instructions provided by the manufacturer; purified PCR products were sequenced with an ABI 3730 sequencer using the primers 1400F, ITS2F, GF, and GR (Marin et al. 2003, Pr€ oschold et al. 2005. Nucleotide sequences were deposited in GenBank under the accession numbers given in Figure 1 and Table S1. Sequences of strains marked with an exclamation point in Figure 1 were previously published by Rindi et al. (2011), but without 5.8S rDNA. These sequences were completed or corrected by one of us (TM) and were resubmitted to GenBank. The new accession numbers of these sequences are given in Figure 1.
Forty-nine sequences of Klebsormidium and Interfilum Chodat strains were used for comparison with strains from alpine soil crusts. These sequences were published by Sluiman et al. (2008), Rindi et al. (2008), and Skaloud and Rindi (2013). Multiple alignments of the newly determined ITS1 and ITS2 rDNA sequences and other sequences selected from the Gen-Bank databases were made using ClustalW and then corrected manually using Bioedit software (Hall 1999). The resulting alignments of the 89 Klebsormidium and Interfilum strains were a concatenated data set (611 bp) of ITS-1 (365 bp) and ITS-2 (246 bp) rDNA sequences according to other researchers working on these genera (Rindi et al. 2008, Skaloud and Rindi 2013. To determine the evolutionary model that best fit the data set, the program MEGA version 6 (Tamura et al. 2013) was used. For maximum likelihood (ML), the GTR model with the proportion of invariable sites (I), and the gamma shape parameter (G) resulted in the lowest Akaike Information Criterion (Akaike 1974). For Bayesian analyses, GTR+G had the lowest Bayesian Information Criterion.
The unrooted phylogenetic tree was constructed in MrBayes 3.2.2 Ronquist 2001, Ronquist andHuelsenbeck 2003) using the GTR+G model with 5,000,000 generations. Two runs of four Monte Carlo Markov Chains were calculated simultaneously, with trees sampled every 500 generations. Split frequency between the runs was below 0.01 at the end of the calculation. The trees sampled before the likelihood scores reached saturation were discarded afterward. The robustness of the tree topology was confirmed by ML (GTR+I+G) performed in GARLI 2.0 (March 2011), and bootstrap support was calculated with 1,000 replicates.
Statistical analyses. Statistical analyses were done in R software (Version 3.1, R Development Core Team 2009). To visualize the dissimilarities in the composition of Klebsormidium clades between different habitats, the non-metric multidimensional scaling (nMDS) plot was calculated based on the Bray-Curtis dissimilarity index (Bray and Curtis 1957). Goodness of fit was estimated based on the threshold recommended by Clarke and Ainsworth (1993). The distribution patterns of Klebsormidium clades along the elevation gradient were visualized by boxplots, also calculated with R software.

RESULTS
Eleven distinct morphotypes were identified among the alpine Klebsormidium strains studied. Seven morphotypes represented known Klebsormidium species, according to their morphology: K. flaccidum K. elegans Lokhorst, K. crenulatum, K. dissectum (F. Gay) H. Ettl & G. G€ artner, K. nitens (K€ utzing) Lokhorst, K. subtile (K€ utzing) Tracanna ex Tell and K. fluitans (F. Gay) Lokhorst. Four morphotypes were impossible to identify unambiguously with reference to known species, because they were transitional morphological forms between described species. These isolates were identified as follows: K. cf. flaccidum (transitional morphotype between K. flaccidum and K. dissectum), K. cf. nitens (K. nitens and Klebsormidium (K.) based on ITS-1 and ITS-2 rDNA sequence comparisons. Phylogenetic tree was inferred by Bayesian method with Bayesian Posterior Probabilities (PP) and maximum likelihood (ML) bootstrap support (BP) indicated at nodes. From left to right, support values correspond to Bayesian PP and ML BP; BP values lower than 50% and PP lower than 0.8 not shown. Strains marked in bold are sequences of Klebsormidium strains from alpine soil crusts. Strains marked with asterisk (*) are proposed as epitypes. Strains marked with exclamation mark (!) are resubmissions of corrected or completed sequences previously published by Rindi et al. (2011). Clade designations follow Rindi et al. (2011). BIODIVERSITY OF KLEBSORMIDIUM FROM ALPINE SOIL CRUSTS and K. dissectum), K. cf. subtile (K. nitens and K. subtile), and K. cf. fluitans (K. subtile and K. fluitans).
Klebsormidium is characterized by a high level of morphological plasticity, and hence we often found different morphotypes over the course of repeated observations on the same strain, reflecting different culture ages and developmental stages. The representatives with thin or medium-sized filaments had in general a high level of variability. This variation in what are considered informative taxonomic characters, such as cell length and filament width and the degree of its disintegration, influenced the general appearance of Klebsormidium filaments. Descriptions and images of morphotypes of alpine Klebsormidium strains are presented in Table 1 and Figures 2-4.
The Bayesian phylogenetic tree of ITS rDNA sequences is presented in Figure 1. Six previously described clades are shown on the tree: A, B/C, D, E, F, and G. The ITS phylogeny did not clearly differentiate between clades B and C. Clade E, which included the majority of strains, had weak statistical support and limited resolution of some subclades.
Forty of the strains from alpine soil crusts were distributed among the main phylogenetic lineages of Klebsormidium: clades B/C, D, E, and F. An exception was clade G, which is composed mostly of strains isolated from arid regions. The majority of alpine strains (31, 77.5%) were included in clade E, which contains the largest number of taxa ( Fig. 5): K. dissectum, K. nitens, K. cf. nitens, K. subtile, K. cf. subtile, K. fluitans, and K. cf. fluitans. Many fewer strains were distributed among clades B/C (6 strains, 15.0%), D (2 strains, 5.0%) and F (1 strain, 2.5%). Clade B/C united two morphotypes: K. flaccidum and K. cf. flaccidum. Clades D and F included one morphotype each: K. elegans and K. crenulatum, respectively.
The distribution of Klebsormidium morphotypes and lineages along the elevation gradient showed some clustering (Fig. 6). Strains from the largest clade E were distributed evenly over the elevations sampled. However, a closer look revealed that strains of the K. dissectum and K. cf. fluitans morphotypes were collected only at lower elevations, whereas K. nitens occurred at middle elevations, and morphotype K. cf. nitens at high elevations. Four of six members of clade B/C, along with all strains of clades D and F, were collected at high elevations (Fig. 6).

DISCUSSION
Distribution of alpine Klebsormidium strains among different phylogenetic clades and along elevation gradients. The phylogeny presented in Figure 1 corresponds well with the ITS phylogeny of Klebsormidium and Interfilum published by Rindi et al. (2011) and with the ITS-rbcL phylogeny from more recent publications ( Skaloud et al. 2014, Ry s anek et al. 2015. However, clade A (Interfilum), which usually appears as a sister group to clade B/C, is distant from other Klebsormidium clades (B/C, D, E, F, and G). As the largest number of strains isolated from the alpine habitats were representatives of clade E it formed the base of biodiversity for the alpine Klebsormidium strains (77.5%). In addition, this group also united the largest number of Klebsormidium morphotypes (Fig. 5). Clade E is the most common group worldwide, and is typical for terrestrial habitats of Europe, North America, and Asia (Ry s anek et al. 2015). Alpine strains were found in all known Klebsormidium lineages, with the exception of clade G. This clade consists mostly of strains isolated from arid regions such as in Africa (Rindi et al. 2011), and hence did not contribute to the Klebsormidium biodiversity of the Alps. More recently, however, several strains of Klebsormidium from acidic soils in Europe were added to clade G ( Skaloud et al. 2014). Although the number of samples is too small for a proper correlation analysis, at least some conspicuous trends among the Klebsormidium morphotypes/clades and their elevational distributions were apparent (Figs. 6 and 7). Representatives of clades B/C, D, and F were collected mainly from soil crusts at high elevations near the pine-forest line and above on alpine meadows and in the nival belt. In contrast, strains from clade E were mostly found at low and middle elevations in the forest zone. In total, 20 Klebsormidium strains from clade E (64.5%) were isolated <1,800 m a.s.l. within the zone of pine forests, and 11 strains (35.5%) were found above this level. Therefore, despite their morphological and ecological plasticity, strains of clade E in general may be more typical of humid and shaded habitats globally (e.g., Skaloud and Rindi 2013), such as pine forest. Strains of clades B/C, D, and F appear to be more adapted to xerophytic habitats characteristic of higher elevations exposed to greater amounts of solar radiation.
Comparing distribution patterns of Klebsormidium phylogenetic clades from alpine soil crusts to terrestrial habitats from other regions. We compared our biodiversity data on the alpine Klebsormidium isolates with those from other regions to verify our proposed ecological preferences of Klebsormidium strains from different phylogenetic lineages, as well as to reveal species composition patterns in terrestrial habitats of the Alps. We selected studies that reflected the most comprehensive Klebsormidium biodiversity of some terrestrial habitats, using an integrative approach. Although the monograph of Kostikov et al. (2001) contains only morphological data, it was very helpful because of the detailed descriptions of easily identifiable species (K. crenulatum / K. mucosum (J.B. Peteresen) Lokhorst complex).
The results of our analysis are presented in Figure 8. Differences in species composition of Klebsormidium in alpine soil crusts appear related more to different habitats (forest or meadows) than to elevations (see Fig. 7). Therefore the assemblage of Klebsormidium from alpine soil crust was divided into two groups, according to their site of collection in either forest or meadows. The species composition 754  Rindi et al. 2011, or clades 5, 6, 8, 9, 13, and 14 according to Skaloud and Rindi 2013). Some ecophysiological data on E clade strains indicate a high sensitivity to desiccation Rindi 2010, Karsten and. Therefore, it is reasonable to assume that this group of Klebsormidium species is, in general, adapted to more hydrophilic habitats, because it is widely distributed in humid and shaded habitats of Western Europe, North America, and Asia, as well as at middle elevations in the Alps, especially in forest belts where adequate humidity is always available.
Biofilms from granite outcrops of steppe slopes and from sand dunes of the Ukraine showed a different composition of Klebsormidium species: a predominance of strains from clade B/C, with a strong contribution of clade F in the assemblage ( Fig. 8; Kostikov et al. 2001. Members of clade F undoubtedly represent a xerophytic adapted lineage of Klebsormidium, according to their characteristic morphological and ultrastructural features (thick cell walls, narrow cells, and a tendency of filaments to fold in braids, which aids in self-protection by preventing excessive water evaporation) along with a high desiccation tolerance , Holzinger et al. 2011, Kaplan et al. 2012. It is possible that representatives of clade B/ C are more xerophytic than strains of clade E, regardless of the morphological similarity of this genetically separated lineage. Particularly the more xerophytic clades of Klebsormidium (B/C, F) have a wider distribution in terrestrial habitats of open landscapes of the Ukraine, due to the more continental (drier) climate and more insolated conditions on steppe slopes and sand dunes. Recently, it was reported that clade B/C does not include acidophilic strains, which were found in all other lineages of Klebsormidium ( Skaloud et al. 2014). The soil crusts from savannas and deserts of South Africa have a unique assemblage of Klebsormidium species that may reflect the specific climatic conditions of these habitats (B€ udel et al. 2009(B€ udel et al. , Rindi et al. 2011). Most of these strains of Klebsormidium belong to the unique and recently discovered phylogenetic lineage clade G. This lineage is xerophytic as well, as its representatives have morphological similarities to species of clade F (thick cell walls and narrow cells, and strongly curved filaments arranged in ball-like aggregations and cluster-like colonies). Strains of clade E, typical of rather humid conditions, are found in the savannas and deserts of South Africa only sporadically. Strain VOE5 was previously identified as a transitional morphotype between K. subtile and K. subtilissimum, based on morphological characters ). g Strain SIE2 was previously identified as a transitional morphotype between K. nitens and K. dissectum, based on morphological characters .
h Strain ASIB V103 was previously identified as K. fluitans, based on morphological characters ) and ITS phylogeny (Kitzing et al. 2014).
The nMDS plot (Fig. 9) clearly confirms the expected similarity of the Klebsormidium species composition among alpine soil crusts from forest habitats of Western Europe, North America, and Asia (stress = 0.06). The Klebsormidium assemblages of alpine soil crusts from meadows and from forests in the U.S. occupy a transitional position between those of alpine soil crusts in Western Europe, North America, and Asia, and those of open landscapes in Eastern Europe. The species composition of African Klebsormidium points to a very high dissimilarity compared with all other habitats, due to the presence of members of clade G (Fig. 9).
Klebsormidium is a cosmopolitan genus, and its members seem to be easily dispersed via air transport (Hoffmann 1989, Ettl and G€ artner 1995, Lokhorst 1996, Rindi et al. 2011. Ecological differentiation has been described for some genetic lineages and clades (Rindi et al. 2011, Skaloud andRindi 2013), showing that not geographic, but rather ecological factors determine the distribution of members of the genus, because no endemic lin-eages could be detected (Ry s anek et al. 2015). However, it appears that ecological and geographic factors are interdependent, as climatic peculiarities determine the characteristics of a habitat. Although geographic barriers do not affect the distribution of microscopic organisms (Finlay et al. 1996, Finlay 2002, the geographic position determines environmental conditions. Our analysis indicates the influence of both parameters (geographic and ecological) on the distribution of Klebsormidium species in soil crusts. The composition of Klebsormidium assemblages from shaded or open habitats (e.g., forests or steppes, savannas) in different geographic regions exhibits similarities due to the influence of comparable microclimatic conditions. Despite statements that the species composition of organisms forming soil crusts (lichens, mosses, and algae) in different regions is similar because of adaptation to extreme environments (West 1990, Sant'Anna and Azevedo 1991, Rosentreter and Belnap 2001, T€ urk and G€ artner 2001, our data showed that a key genus of soil crusts, Klebsormidium, contains different Identification of alpine strains, problems of species delimitation, morphological plasticity. The phylogenetic analysis of Klebsormidium alpine isolates (Fig. 1) showed that only a small number of the described morphotypes clearly corresponded to known taxa. Alpine strains assigned to the K. flaccidum and K. dissectum morphotypes, along with one strain of the K. subtile morphotype showed high similarity to the respective epitypes (see below). Two strains of the morphotypes K. elegans and K. crenulatum were included in clades D and F, respectively, but each formed a separate subclade. Strains belonging to the morphotypes K. fluitans, K. nitens, and K. subtile (in part) appeared in several subclades of E, far distant phylogenetically from the corresponding epitype strains. Data concerning the polyphyletic position of these morphotypes were partially reported in previous papers (Rindi et al. 2011, Skaloud and Rindi 2013, Skaloud et al. 2014. Some of our alpine strains represented transitional morphotypes (K. cf. flaccidum, K. cf. nitens, K. cf. subtile, and K. cf. fluitans) that do not clearly correspond to existing morphological species. The strains of the K. cf. flaccidum morphotype belonged to clade B/C, but formed separate subclades (Fig. 1). Morphotypes of K. cf. nitens, K. cf. subtile, and K. cf. fluitans were all distributed within clade E, in no clear order, but with a gradual transition from one morphotype to another (Fig. 1). BIODIVERSITY OF KLEBSORMIDIUM FROM ALPINE SOIL CRUSTS Some alpine strains of Klebsormidium were mentioned in other publications, where they were identified based on morphology or ITS/rbcL phylogeny (see Table 1). While most identifications could be confirmed, two strains BOT2 (SAG 2417) and OBE1 were different: K. nitens as described by Kaplan et al. (2012) represents K. dissectum, and K. fluitans as described by Kitzing et al. (2014) is a member of K. nitens. This misidentification is related to the absence of reference strains for most species of Klebsormidium. Identification based purely on molecular data, as in the above papers, without comparison to morphology is problematic. Therefore, a comprehensive taxonomic revision of Klebsormidium, combining molecular, morphological, and ecological data is needed.
Recent data of other authors showed the existence of a high number of cryptic Klebsormidium species within clade E ( Skaloud and Rindi 2013), as determined on the basis of ITS-rcbL phylogeny and some ecological preferences. Sequences of several strains published by these authors are included in our phylogenetic tree. Alpine strains formed separate subclades and corresponded to some cryptic species according to Skaloud and Rindi (2013), but were distributed among other, previously known strains in no clear order (Fig. 1). Therefore, it is FIG. 7. Boxplots showing median, 25%-75% percentiles and range of distribution of Klebsormidium phylogenetic clades in alpine soil crusts along elevation gradient.
FIG. 8. Comparison of distribution pattern of phylogenetic clades within Klebsormidium collected from alpine soil crusts ("Alps, forest" and "Alps, meadows") and other sites. "Europ. cities, walls" refers to algal crusts from building walls of Western European cities (Rindi et al. 2008); "Washington, forest," "Ohio, forest," "Connecticut, forest," "Czech Rep., forest," "Wales, forest" and "Japan, forest" are terrestrial habitats under mixed forest of northern temperate zones of United StatesA, Western Europe and Japan (Ry s anek et al. 2015); "Ukraine, granite" stands for algal crusts from granite outcrops in Ukraine ; "Ukraine, dunes" refers to soil crusts from dunes along Dnieper River Dnipro sand dunes in Ukraine (Kostikov et al. 2001); "Africa, drylands" to soil crusts from South African drylands (B€ udel et al. 2009(B€ udel et al. , Rindi et al. 2011 760 difficult to reach a definite conclusion on the taxonomic position of some alpine isolates, or on their ecological preferences, because they are distributed among taxa identified by other authors as cryptic species characteristic of artificial subaerial substrates or freshwater habitats ( Skaloud and Rindi 2013). We agree with the statement of these authors that the morphological, genetic, and ecological variability in clade E members is difficult to address because of high plasticity with respect to these parameters.
We compared our data with those of an early publication (Reisigl 1964) on the diversity of Klebsormidium from alpine soils, which was later cited by T€ urk and G€ artner (2001) and G€ artner (1995, 2014). Four morphotypes of Klebsormidium (Types 1, 2, 3, and 4) referred to Hormidium flaccidum (K€ utzing) A. Brown were found in alpine soils at elevations between 3,457 and 3,739 m a.s.l. (Reisigl 1964). Reisigl used a wide species concept for H. flaccidum and mentioned a high morphological plasticity in culture as well as taxonomic problems with the genus. Although a precise evaluation of Reisigl's data is difficult because of the brief descriptions and the absence of illustrations of the four morphotypes, most probably all of these morphotypes belonged to clade E, as they consisted of thin filaments (~5.5 lm diameter). Another species, K. montanum (Hansgirg) S.Watanabe, found in soil in South Tyrol, Italy G€ artner 1995, 2014) represents a morphotype very similar to the K. crenulatum/K. mucosum complex.
For taxa assigned to Klebsormidium, we propose using species' protologues as well as Lokhorst (1996) to designate epitypes. Lokhorst (1996) provided the most comprehensive morphological treatment of Klebsormidium species in Western Europe, based primarily on cultured material. His work is valuable because Lokhorst (1996) designated holotypes, lectotypes, or neotypes of all Klebsormidium species that he investigated. Unfortunately, Lokhorst's monograph does not refer to any strain numbers, and type material cannot be used to study morphological plasticity and molecular phylogeny (Rindi et al. 2011, Skaloud et al. 2014 or even to identify species Leliaert 2007, Friedl andRybalka 2012). According to Article 9.8 of the ICN (McNeill et al. 2012), "An epitype is a specimen or illustration selected to serve as an interpretative type when the holotype, lectotype, or previously designated neotype, or all original material associated with a validly published name, is demonstrably ambiguous and cannot be critically identified for purposes of the precise application of the name to a taxon." For Klebsormidium species, as for many other species of microalgae (e.g., Darienko et al. 2010, Bock et al. 2011, Demchenko et al. 2012, Rybalka et al. 2013, epitypification is necessary to unambiguously link names to sequenced specimens. In most cases, the strains isolated and investigated by Lokhorst (1996) are proposed as epitypes, except for K. subtile that he did not treat. For each recognized species, we usually accept Lokhorst's (1996) heterotypic synonyms.
Comments: Klebsormidium flaccidum has a simple morphological appearance and therefore could be identified as strains of several genetic lineages of Klebsormidium (clades B, C, or E; Rindi et al. 2011, Skaloud et al. 2014. The herbarium type material and the incomplete original diagnosis cannot be used to clarify the situation with this species. Therefore, the designation of an epitype specimen based on a subjective choice is the only feasible solution for a reassessment for this species (Rindi et al. 2011, Skaloud et al. 2014. We propose a strain of K. flaccidum isolated by Lokhorst and now preserved in the SAG collection (SAG 2307) as the epitype. SAG 2307 was isolated from clayey soil in a field of beets near Niederkruechten (Germany), original number KL 1. It does not conflict with the original description, is~400 km from the type locality (K€ utzing 1849), and generally corresponds to the emended description. Minor differences include the slightly wider filaments, 7.8(8.8) lm (in the emended description-(5.6)6.5-7.4 lm). The epitype strain rarely has H-like fragments of cell walls in agar culture and has a prominent starch envelope composed of several layers of small starch grains surrounding the pyrenoid (the starch envelope of the pyrenoid is visible in Lokhorst's micrographs (Lokhorst 1996, figs. 44, 45), although the description mentions "without distinct starch envelope").
Type locality: On wet and warm wall of a bath house, Padua (Italy).
Epitype: Strain SAG 37.86 designated here to support the lectotype specified above and the authentic strain of K. crenulatum that is permanently preserved in a metabolically inactive state (cryopreserved in liquid nitrogen) in the SAG.
Comments: Strain SAG 37.86 was isolated by H. Trenkwalder (original number T 93) from a soil from Brixen, South Tyrol (Italy) and was previously identified as Ulothrix tenuissima K€ utzing. Strain SAG 6.96, isolated by Lokhorst (original number KL 64) is no longer available. The phylogenetic analysis provided by Rindi et al. (2011) showed a close relationship between the two isolates. Strain SAG 37.86 does not conflict with the original description and type locality (K€ utzing 1845) and generally corresponds well with the emended description, but differs in having slightly thicker filaments, reaching 16.6(17.8) lm (the original description gives the maximum width as 14.0(15.8) lm). The taxonomic combination Klebsormidium crenulatum (K€ utzing) H. Ettl & G. G€ artner (Ettl and G€ artner 1995) is invalid, because of priority rules. Type locality: In a mill-course. Emended description: Filaments long, with some tendency to fragmentation, sometimes curved, thin or medium in width, in young culture cells are long and cylindrical, in mature and old cultures sometimes isodiametric, (5.1)6.0-6.6(7.0) 9 (4.7)5.8-10.3(11.5) lm (length/width-1.0-1.8), filaments are slightly bead-like, constricted near cross walls, Hlike pieces of cell walls sometimes present; chloroplast covering 2/3 of the cell inner surface, with smooth or undulating margins; the pyrenoid is small, round, compact, surrounded by a layer of starch grains. It forms submerged tufts and a superficial layer in liquid culture, and smooth or slightly undulating colonies on an agar plate.
Epitype (designated here): Strain SAG 384-1 (proposed here as the authentic strain of K. subtile) is permanently preserved in a metabolically inactive state (cryopreserved in liquid nitrogen) in the SAG.
Comments: K. subtile is morphologically and ecologically close to K. subtilissimum (Rabenhorst) P.C. Silva, Mattox et Blackwell. Both these taxa may represent a single species inhabiting water bodies or humid terrestrial habitats. They were initially described as representatives of Ulothrix K€ utzing (K€ utzing 1845, Rabenhorst 1857). As K. subtile was described earlier than K. subtilissimum (as Ulothrix subtilis K€ utzing 1845), the former name has priority. Strain SAG 384-1, initially identified as K. subtilissimum, is proposed as the epitype for K. subtile. Strain SAG 384-1 was isolated by R.A. Lewin from snow (United States). It does not conflict with the original description (K€ utzing 1845), the description in Tell (1976) and thus generally corresponds to the emended description. The combination Klebsormidium subtile (K€ utzing) Tracanna ex Tell (Tell 1976, p. 535) is invalid because although the basionym is listed in this study, a complete and direct reference to its author and place of valid publication including page reference and date is missing (see Article 41.5 of the ICN (McNeill et al. 2012)).
Epitype: Strain SAG 13.91, designated here to support the lectotype specified above. This authentic strain of K. nitens is permanently preserved in a metabolically inactive state (cryopreserved in liquid nitrogen) in the SAG.
Comments: The K. nitens morphotype is present in several lineages of clade E (Fig. 1), but most of the investigated strains with this morphotype can be grouped in E2 (Rindi et al. 2011, Skaloud andRindi 2013). Strains of clade E2 were proposed as good candidates for designation as K. nitens ( Skaloud et al. 2014). We propose as the epitype strain SAG 13.91 that belongs to E2 (Rindi et al. 2011), and its morphological and ecological features correspond with the diagnosis of K. nitens as emended by Lokhorst (1996). SAG 13.91 was isolated by E.A. Flint from Tekoa soil (New Zealand), original number No 60/74. It does not conflict with the original description (K€ utzing 1849) and generally corresponds to the emended description. SAG 13.91 is also genetically close to Lokhorst's strain of K. nitens (KL 37), which was lost (Rindi et al. 2011).
Epitype: Strain SAG 2417 is designated here to support the neotype specified above and proposed as the authentic strain of K. dissectum. It is permanently preserved in a metabolically inactive state (cryopreserved in liquid nitrogen) in the SAG.
Comments: Strain SAG 2155 isolated by Lokhorst from the neotype locality (original number KL 2) cannot be used as the epitype, because it does not correspond to emended description of K. dissectum. SAG 2155 morphologically and genetically corresponds to K. nitens (Rindi et al. 2011). Perhaps, this strain was mislabeled. We propose a strain SAG 2417 that was isolated by U. Karsten (originally labeled BOT2) from the concrete basement of a BIODIVERSITY OF KLEBSORMIDIUM FROM ALPINE SOIL CRUSTS water level of the river Dommel near Valkenswaard (the Netherlands). It does not conflict with the original description (Petersen 1915) and generally corresponds to the emended description. The only difference is the slightly thinner filaments, not exceeding 20.0 lm (in original description-up to 23.3 lm).
Epitype: Strain SAG 7.96 designated here to support the type specified above and proposed as the authentic strain of K. elegans is permanently preserved in a metabolically inactive state (cryopreserved in liquid nitrogen) in the SAG.
Comments: Strain SAG 7.96 was isolated by Lokhorst from the type locality (original number KL 24) and generally corresponds to the original description. The only difference consists of the slightly thinner filaments-(7.8)8.1-9.3(10.0) lm (in original description-(8.4)9.3-10.2(13.0) lm). Klebsormidium bilatum is proposed as a synonym of K. elegans because the ITS and rbcL sequences as well as the morphological characters of the strain isolated by Lokhorst (SAG 5.96) are close to the proposed epitype strain (see Rindi et al. 2011 and present paper). The rbcL sequences of both strains (SAG 7.96 and SAG 5.96) are identical, but ITS 2 sequences have differences in one nucleotide.
Our sincere thanks to Dr. Thomas Pr€ oschold for providing sequences of the alpine Klebsormidium strains and fruitful discussion on taxonomic aspects, to Prof. Thomas Friedl for making it possible to compare our material with strains in the SAG collection, and to Dr. Maike Lorenz and the SAG staff for their help with cryopreservation of epitype strains. T.M. thanks the DAAD for a short-term research fellowship and the Alexander von Humboldt Foundation for a Georg-Forster research fellowship at the University of Rostock. The support of this study by Austrian Science Fund (FWF) grant P 24242-B16 is acknowledged by A.H. Finally, U.K. thanks the DFG for financial support (KA899/16-1/4) and the University of Innsbruck as host during his sabbatical.