Introduction
During summer, when persistent snowfields become waterlogged, they represent basic ecosystems that often include protozoa and rotifers, but are mainly populated by psychrophilic bacteria, fungi and eukaryotic photoautotrophs. These organisms are a well-known phenomenon occurring not only in polar, but also in high alpine regions (Kol Reference Kol1968, Hoham & Duval Reference Hoham and Duval2001, Komárek & Nedbalová Reference Komárek and Nedbalová2007). In Antarctica, Ling (Reference Ling1996) identified a total of 24 different snow algal species in the Windmill Islands region. There, they can cause green, orange or red snow, depending on the dominant pigments. They have to be adapted to different harsh environmental conditions like permanently low temperatures, high irradiation (ultraviolet and visible) at the snow surface, a short growth period and limited water availability due to frequent freeze-thaw cycles in the habitat. Additionally, low pH and desiccation after snowmelt play a role. As a consequence, many of these organisms have developed special cytological strategies such as the formation of robust secondary cell walls for mechanical protection (Remias et al. Reference Remias, Karsten, Lütz and Leya2010a), or the accumulation of secondary metabolites such as carotenoids (Leya et al. Reference Leya, Rahn, Lütz and Remias2009) or phenolics (Remias et al. Reference Remias, Schwaiger, Aigner, Leya, Stuppner and Lütz2012). Both are able to shield the chloroplast and the nucleus against excessive irradiation, but their metabolism has to be adapted to cold conditions, which has been demonstrated previously with photosynthesis and respiration measurements for algae causing red snow (e.g. Bidigare et al. Reference Bidigare, Ondrusek, Kennicutt, Iturriaga, Harvey, Hoham and Macko1993).
In 1912, Fritsch (Reference Fritsch1912) found fusiform cells with flanged cell walls causing yellow snow on the South Orkney Islands, Maritime Antarctica, and described them as Scotiella polyptera Fritsch. His assumption included a constant morphology, thus daughter cells should have the same outer shape. Decades later, careful observations of the life cycle by light microscopy (LM) revealed that the ornamented cells in fact are immotile cyst stages of a Chlamydomonadacean flagellate, and the daughter cells in fact have smooth cell walls. Consequently, the species was renamed Chloromonas polyptera (Hoham et al. Reference Hoham, Mullet and Roemer1983). The situation was similar for several further “Scotiella-like” snow algae such as Chloromonas rosae var. psychrophila Hoham, Bonome, Martin & Leebens-Mack (former Scotiella cryophila Chodat; Hoham et al. Reference Hoham, Bonome, Martin and Leebens-Mack2002), Chloromonas pichinchae (Lagerheim) Wille (former Scotiella tatrae Kol; Hoham Reference Hoham1975) or Chloromonas nivalis (former Scotiella nivalis (Chodat) Fritsch; Hoham & Mullet Reference Hoham and Mullet1978). The latter species was used by Remias et al. (Reference Remias, Karsten, Lütz and Leya2010a) to verify the identity of flanged cysts and smooth flagellates with molecular methods. Chloromonas polyptera has been reported so far from Antarctic locations on the Windmills Islands (Ling & Seppelt Reference Ling and Seppelt1998), Rumpa Island (Akiyama Reference Akiyama1979) and Petermann Island (Kol Reference Kol1968). All these sites were coastal and in the vicinity of penguin rookeries. Putative discoveries in North America (Garric Reference Garric1965, Hoham et al. Reference Hoham, Mullet and Roemer1983) later turned out to be populations of a morphologically similar but smaller psychrophilic species, Chloromonas hohamii Ling & Seppelt (Ling & Seppelt Reference Ling and Seppelt1998). The same reference gives the most detailed report about C. polyptera so far, including information about morphology and life cycle. However, physiological measurements of photosynthesis or respiration, as well as an analysis of pigments or soluble carbohydrates have not been performed before. The aim of this work was to elucidate the kind of secondary pigmentation which protects this snow alga against high irradiation at the snow surface, discover whether soluble carbohydrates can play a role against intracellular freezing, and to compare these results with the alpine relative Chloromonas nivalis (Remias et al. Reference Remias, Karsten, Lütz and Leya2010a). It was expected that photosynthesis assays would show the metabolic activity at low temperatures typical for snow. Finally, molecular sequencing should provide further insights in the still intricate phylogeny of Chlamydomonadacean algae living in snow.
Material and methods
Sampling and sample preparation
Superficial orange snow caused by C. polyptera (also compare Ling & Seppelt (Reference Ling and Seppelt1998) to its closely related species C. hohamii) was collected at two sites along the Antarctic Peninsula: first, on Goudier Island (situated in Port Lockroy next to Wiencke Island) on 31 December 2008, near the British “Base A” station (64°49.51′S, 63°29.69′W, 13 m a.s.l.), and second, in the vicinity of the Chilean Presidente Gabriel Gonzáles Videla Station (at Paradise Harbour) on 03 January 2009, (64°49.45′S, 62°51.50′W, 9 m a.s.l., sample DRAnt023). The electrical conductivity of the meltwater was measured with a Cond 340i, and the pH with an Inolab (WTW instruments, Germany). The samples were harvested with 500 ml sterile plastic bags and cooled at 4°C for transportation to Jubany Station (Dallmann Laboratory) at King George Island (South Shetland Islands). At the station samples were freeze-dried for 48 hours on 47 mm Whatman GF/C glass fibre filters for later pigment and carbohydrate analysis. For an Antarctic vs alpine comparison, field samples of Chloromonas nivalis causing orange snow in Kühtai, Tyrol, Austrian Alps, 2300 m a.s.l. (compare Remias et al. Reference Remias, Karsten, Lütz and Leya2010a), were harvested and analysed in a similar manner.
Photosynthesis measurements
Photosynthesis and respiration of samples from Paradise Harbour were measured with a Fibox 3 optode (PreSens, Germany) in a transparent 3 ml acryl-chamber with integrated magnetic stirrer. Two ml of algal suspension were mixed with 1 ml of a 0.1 M HCO3- source before measurement and allowed to equilibrate for 15 min in darkness. Different light levels were established with Hansatech A5 neutral filters and calibrated with a Hansatech QRT1 photosynthetically active radiation (PAR) sensor (compare Remias et al. Reference Remias, Albert and Lütz2010b). The oxygen production/consumption per time of three replicates was normalized to the chlorophyll content of the samples, which was determined spectrophotometrically (Perkin Elmer Lambda 20) according to Porra et al. (Reference Porra, Thompson and Kriedemann1989) after grinding and extracting the cells with 1 ml N-N-dimethylformamide (DMF, Sigma-Aldrich).
Light- and electron microscopy
In Antarctica, field samples were observed with a Zeiss Axiolab light microscope and a Nikon Coolpix 8400 camera. Samples which were transported to Austria for laboratory long-term observation (kept at 4°C and 50 μmol PAR m-2 s-1) were monitored with an Zeiss Axiovert 200 M, equipped with a Zeiss MRc 5 camera. Fluorescence images were taken with a Zeiss 09 filter (excitation 450–490 nm, emission 515 nm long pass). Cell sizes were digitally measured with Zeiss Axiovision software. Samples for scanning electron microscopy (SEM) were dried at 50°C for 12 h and subsequently contrasted without forgoing dehydration and critical-point drying for direct analysis in a Phillips SEM XL20.
Pigment and carbohydrate analysis
Filters with freeze-dried algae were ground with a grinding mill (Dismembrator S, Sartorius, Germany) with quartz balls and Teflon jars pre-cooled in liquid nitrogen. The sample for pigment analysis was extracted in 3 ml DMF, those for soluble carbohydrates in 80% ethanol. Analysis was performed with an Agilent ChemStation 1100 HPLC with diode array detector for pigments and refractive index detector for carbohydrates. The pigments were separated with a LiChrospher column (250 x 4.6 mm) thermostated to 30°C using a binary solvent gradient with acetonitrile, methanol, tris-buffer and hexane (Remias & Lütz Reference Remias and Lütz2007). Identification was carried out by compound retention time and spectral absorption of the peak in comparison with available carotenoid and chlorophyll standards (Sigma-Aldrich; CaroteNature). Minor, unknown carotenoids were roughly quantified with the calibration formula of beta-carotene. Carotenoids were quantified at 440 nm, chlorophyll b at 648 and chlorophyll a at 662 nm. Soluble carbohydrates were separated with a Phenomenex Rezex RCM - monosaccharide Ca2+ column (300 x 8 mm) at 80°C and isocratic aqueous conditions at 0.6 ml min-1. The sample was dissolved in pure water, centrifuged and filtered prior to injection. The compounds were identified and quantified considering the following calibration standards (Sigma-Aldrich). Aldoses and ketoses: sucrose, galactose, glucose, mannose, trehalose, maltose, cellubiose, raffinose, xylose, rhamnose, fucose, fructose, verbascose, stachyose, 1-kestose and ribose. Sugaralcohols: galactinol, meso-erythritol, mannitol, arabitol, dulcitol, xylitol, sorbitol, adonitol and glycerol.
Phylogenetic analyses
Sequences of the 18S rRNA gene used in this phylogenetic study were obtained either from field samples collected during expeditions by two of the authors: Daniel Remias (prefix “sample DR”) or Thomas Leya (prefix “sample CCCryo”), or from existing entries in the National Center for Biotechnology Information (NCBI) database. Genomic DNA was isolated from field samples using the DNeasy Plant Kit (Qiagen, Hilden, Germany). Sequences from field samples were directly derived from PCR products using the same primer sets for PCR and sequencing reactions as stated in Remias et al. (Reference Remias, Karsten, Lütz and Leya2010a) without further cloning. New sequences for the resting stages used in the phylogeny of this paper are available on the internet (e.g. http://www.ncbi.nlm.nih.gov/nuccore) under accession numbers JQ790556–JQ790560. All gene sequences used were assembled and aligned using the software CLC Combined Workbench software (version 6.6.1, CLC bio, Aarhus, Denmark). The alignment was imported into the phylogenetic software PAUP* 4.0 (beta version 10, Sinauer Associates Inc, Sunderland, MA, USA). To choose the best-fit evolutional model for nucleotide substitution the data were analysed using version 3.7 of the software Modeltest (Posada & Crandall Reference Posada and Crandall1998, Posada & Buckley Reference Posada and Buckley2004). Final data analysis, analyses of bootstrap values, and phylogenetic tree building were performed with the PAUP software.
Results
At both sites, patches of orange snow occurred around coastal gentoo penguin rookeries not far away from the sea (Fig. 1a–d). The locations were evidently influenced by guano, showing pH values of 7.4 (Goudier Island) and 7.5 (Presidente Gabriel Gonzáles Videla Station). The electrical conductivities of the melted snow were 94.8 μS cm-1 and 42.0 μS cm-1, respectively. The orange colour reached about 5–10 cm deep, sometimes replaced by green snow deeper in the snowpack (Fig. 1c). The green snow consisted of several unidentified algae, including green flagellates and Chlorella-like species (data not shown). The orange snow harvested from the surface contained almost exclusively immotile cyst stages of Chloromonas polyptera (Fig. 2). In the light microscope, cytosolic parts were of orange colour, interspersed by several green spots caused by the chloroplast (Fig. 2a). This also could be visualized when switching from bright field to fluorescence mode, with the spherical to ellipsoidal disc-like chloroplast sections fluorescing in red colour (Fig. 2b & c). Some cells were kept cool and illuminated for several weeks to study any changes. After approximately three months, the protoplast contracted (Fig. 2d), resulting in a subsequent cleavage into four smooth-walled daughter cells (Fig. 2e). The complex, flanges-like structures of the cyst cell wall are depicted by means of SEM in Fig. 2f & g. However, sometimes the ornamentation could also be observed by LM. The cysts measured 25.7 ± 2.5 μm long and 19.5 ± 2.3 μm wide.
Fig. 1 a. View of the first sampling site at Goudier Island with the British “Base A” in the foreground and a glacier of Wiencke Island in the background. Note the penguins at snow-free places. b. View of the second sampling site next to the Chilean Presidente Gabriel Gonzáles Videla Station, Paradise Harbour. c. Close-up of orange snow caused by Chloromonas polyptera at Goudier Island. Note the dug hole, revealing a green colouration caused by algae deeper in the snow. d. Coastal orange snow caused by C. polyptera in the vicinity of the Presidente Gabriel Gonzáles Videla Station.
Fig. 2 a. Light microscopy of field samples of Chloromonas polyptera from Goudier Island, containing immotile cysts with ornamented cell walls. Two cells appear roundish due to an upright position, where the cell wall flanges can be seen (arrows). Orange cytoplasmic regions containing astaxanthin alternate with greenish spots of the chloroplast. b. Bright field, and c. fluorescence image of the same two cells, the latter revealing disk-like structures of chloroplast portions. d. Contraction of the protoplast of an older cyst in advance of the formation of four elongate daughter cells (three of them visible) with e. smooth cell walls. f. Longitudinal, and g. angular view of the cyst cell wall by means of SEM. Many flanges reach from pole to pole, others merge or divide in between.
The relative amount of plastidal pigments in the sample from Presidente Gabriel Gonzáles Videla Station can be seen in Fig. 3. The keto-carotenoid astaxanthin was the only pigment detected in the class of secondary carotenoids (c. 51%). About 90% occurred as esters, thus resulting in additional, more lipophilic peaks with later retention times but the same absorption spectrum (λmax = 478 nm). Due to the spectral absorption, the astaxanthin was estimated to be in the all-trans configuration, only trace amounts of non-esterified cis-isomers were detected, which have an additional absorption shoulder in the UVA region and a maximum shifted to shorter wavelengths (λmax = 466 nm). Chlorophylls a and b were the second largest group of pigments (c. 41%), followed by primary carotenoids like lutein, beta-carotene and neoxanthin. In addition, the three xanthophyll cycle pigments violaxanthin (V), antheraxanthin (A) and zeaxanthin (Z) were measured, also belonging to the group of primary pigments (primary carotenoids in total, 5%). The de-epoxidation state of the xanthophyll cycle, defined as (A + Z)/(V + A + Z), was 0.871. Furthermore, about 3% of the pigments were minor unknown carotenoids. Additionally, Fig. 3 depicts the relative quantities of mature cysts of the alpine Chloromonas nivalis (from Remias et al. Reference Remias, Karsten, Lütz and Leya2010a), which contained less secondary carotenoids (21%), but more chlorophylls (60%) and primary carotenoids (19%). The main soluble carbohydrates of C. polyptera were saccharose (1.4 mg g DW-1) and glycerol (0.9 mg g DW-1). The total amount of quantifiable carbohydrates, including further sugars like glucose and sugar alcohols, accounted for 4.4 mg g DW-1. In contrast, the most abundant sugar alcohol of Chloromonas nivalis was glycerol (0.7 mg g DW-1), and saccharose was the second most frequent compound (0.2 mg g DW-1).
Fig. 3 Proportions of pigment classes (chlorophylls, primary and secondary carotenoids) between Antarctic Chloromonas polyptera (left) and European Alpine Chloromonas nivalis (right). The values of the latter were taken from Remias et al. (Reference Remias, Karsten, Lütz and Leya2010a).
The oxygen consumption and production of C. polyptera in darkness, and at different light levels was measured at conditions close to ambient (1°C) (Fig. 4). Generally, the alga was performing physiologically well at a temperature typical for snow. The mean oxygen consumption during dark respiration was 28.01 μmol O2mgchl-1h-1. At low light (48 μmol PAR m-2 s-1), net photosynthesis was already positive, and a physiological light compensation point of c. 30.2 μmol PAR m-2 s-1 was calculated. Even at higher irradiation levels (746 and 1362 μmol PAR m-2 s-1), no photoinhibition due to potential high light stress took place, in fact oxygen production was still rising.
Fig. 4 Dark respiration and light-dependent photosynthesis of Chloromonas polyptera measured at 1°C, expressed as oxygen turnover per time and amount of chlorophyll.
Phylogeny
Figure 5 depicts a phylogenetic tree based on a maximum likelihood analysis of the 18S rRNA gene sequences of different Chlamydomonadaceae involved in green, orange or red snow and closely related taxa. The tree resulted from an analysis under the Akaike Information Criterion using the transitional model TIM+I+G with proportions of invariable sites I = 0.6059 and an estimated gamma shape G = 0.8133 (values corresponding to the general time reversible model GTR+G+I). The Antarctic C. polyptera (sample DRAnt023) groups close together with the alpine Chloromonas nivalis (sample DRAlp024) in the Chloromonas-snow clade, which contains ten snow algae from different regions. This specific subclade is part of the A2 clade described by Hoham et al. (Reference Hoham, Bonome, Martin and Leebens-Mack2002).
Fig. 5 Phylogenetic analysis of Arctic, Antarctic and European snow algae within the Chlamydomonadaceae based on 1523 nucleotides of the nuclear-encoded 18S rRNA gene. Two Trebouxiophyceae were used as an outgroup. Bootstrap values at the branches were determined for distance (top left), maximum parsimony (top right) and maximum likelihood (bottom). Cr. = Chloromonas, Cd. = Chlamydomonas. Chloromonas polyptera is marked in bold. The tree is explained in detail in the discussion section. Clade assignments (A1, A2) according to Hoham et al. (Reference Hoham, Bonome, Martin and Leebens-Mack2002).
Discussion
Chloromonas polyptera seems to be typical for coastal Antarctica at locations where seasonal snowfields are influenced by eutrophication due to animal colonies. This fact can be seen in the slightly basic pH values and elevated meltwater conductivities, most probably caused on the one hand by guano input and by sea spray on the other hand. In contrast, snowfields in the European Alps, where the morphological similar species Chloromonas nivalis lives, have a tentatively acidic pH and much lower conductivities (Remias et al. Reference Remias, Karsten, Lütz and Leya2010a), the latter most probably reflecting a lower nutrient content. Generally, mountainous snowfields populated with algae have acidic pH values and show lower electrical conductivities than coastal populations (Hoham & Duval Reference Hoham and Duval2001). Chloromonas polyptera was morphologically identical or at least very similar to the earlier description of Ling & Seppelt (Reference Ling and Seppelt1998). Chloromonas hohamii morphologically is also very similar to C. polyptera, but is regarded as a separate species based on its sole reports from North America (Ling & Seppelt Reference Ling and Seppelt1998). The former has a sexual reproductive cycle, lives in snow with acidic pH and low nutrients, and is found only in North America, whereas the latter is asexual, found in alkaline snow with high nutrients, and is from Antarctica. This is the first report measuring the pigment composition of C. polyptera. The metabolite causing the orange colour of this species, the carotenoid astaxanthin, has been reported from many other snow algae as well, e.g. Chlamydomonas cf. nivalis (F.A. Bauer) Wille (Bidigare et al. Reference Bidigare, Ondrusek, Kennicutt, Iturriaga, Harvey, Hoham and Macko1993, Remias & Lütz Reference Remias and Lütz2007, Fujii et al. Reference Fujii, Takano, Kojima, Hoshino, Tanaka and Fukui2010). The macroscopic intensity of the orange or red colour observed in the field obviously depends on the relative concentration of astaxanthin compared to the chlorophylls in the single cells. For example, Chloromonas nivalis contains less astaxanthin than Chloromonas polyptera and thus causes a different (less intense) shade of orange instead of the clear orange snow observed in the latter (Remias et al. Reference Remias, Karsten, Lütz and Leya2010a). Accordingly, Chlamydomonas nivalis with relatively high astaxanthin content (82% of all pigments; Remias et al. Reference Remias, Lütz-Meindl and Lütz2005) causes crimson red snow. In general, the accumulation of secondary carotenoids is believed to protect algae against unfavourable environmental conditions (Bidigare et al. Reference Bidigare, Ondrusek, Kennicutt, Iturriaga, Harvey, Hoham and Macko1993, Lemoine & Schoeffs Reference Lemoine and Schoefs2010).
Interestingly, the chloroplasts of C. polyptera contained relatively high amounts of A and Z causing a de-epoxidation-state (=(A+Z)/(V+A+Z)) of 0.871. Most probably, the xanthophyll cycle of the chloroplast was induced by high light conditions, because the astaxanthin content of this species could be too low for shading the thylakoids effectively against excessive visible irradiation. This would go along with the report that the high astaxanthin content of Chlamydomonas nivalis causes much lower de-epoxidation ratios of 0.27 or lower (Remias et al. Reference Remias, Lütz-Meindl and Lütz2005), measured in samples taken from highly irradiated alpine snow surfaces.
Generally, soluble carbohydrates of snow algae have been analysed less frequently than pigments. Roser et al. (Reference Roser, Melick, Ling and Seppelt1992) and Chapman et al. (Reference Chapman, Roser and Seppelt1994) found glucose, saccharose, trehalose and the sorbitol in a Chloromonas species causing red snow in Antarctica. However, they did not detect glycerol as we did in this study. The latter supports the integrity of biomembranes during drought, thus delimiting cytological damages caused by water stress, which is the case whenever the snow habitat becomes frozen (Thomas et al. Reference Thomas, Fogg, Convey, Fritsen, Gili, Gradinger, Laybourn-Parry, Reid and Walton2008), and trehalose has the same functions (Elbein et al. Reference Elbein, Pan, Pastuszak and Carroll2003).
As shown in this study by photosynthesis experiments the cysts are not physiological “resting stages” and their active carbon fixation at 1°C, even at high irradiances, can be regarded as an adaptation to the harsh habitat conditions. Since such cysts do not proliferate during the rest of the season, they can invest energy in the accumulation of reserve metabolites to oppose upcoming severe conditions. This is necessary because cells passively accumulate on the snow surface due to ongoing snowmelt in summer, where they are subject to high irradiation and to freeze-thaw cycles, whilst later at the Antarctic sites where the populations were observed, the snow pack may completely melt away. Consequently, in late summer the cells remain on bare rock surfaces and may be subject to drought and relatively high temperatures when they are exposed to light.
Cells deposited in dry surface dust can play a crucial role in long distance dispersal of snow algal cysts by wind, as shown by Marshall & Chalmers (Reference Marshall and Chalmers1997) at Signy Island. After a period of dormancy, such cysts may germinate again the next year when they are transported to a site with suitable conditions.
The phylogenetic tree (Fig. 5) depicts the taxonomic relationship of different Chlamydomonadaceae involved in coloured snow as well as closely related taxa. The main genera responsible are Chloromonas (Cr.) on one side (Chloromonas-snow clade) and cf. Chlamydomonas (Cd.) on the other side (‘Chlamydomonas’-snow clade). For the former group the genus Chloromonas has long been identified, whilst the correct genus affiliation responsible for red snow from the latter clade is still under debate. Sequence data for this clade are very sparse and live cultures only exist of closely related strains (CCCryo 086-99 and CCCryo 101-99). The ‘Chlamydomonas’-snow clade comprises taxa from Spitsbergen (CCCryo samples) and the European Alps (DRAlp027), which produce either spherical cysts with smooth (‘little orange’) or slightly ornamented (‘warty red’) wall surfaces. Especially the warty red cysts are typical for red snow in polar regions and the European Alps, and these are also generally richer in the red pigment astaxanthin than representatives of the Chloromonas-snow clade as already discussed above. The identification of the different taxa involved is still under investigation, although usually such spherical red cysts are assigned to the collective name of Chlamydomonas nivalis. However, as Pröschold et al. (Reference Pröschold, Marin, Schlösser and Melkonian2001) showed, the clade of Chlamydomonas reinhardtii P.A. Dangeard as the type species is pointing towards the true genus Chlamydomonas, leaving the neighbouring snow clade isolated. Strains CCCryo 086-99 and CCCryo 101-99 provide the only related algae, but even their identification to date remains unclear. In addition, these strains appear to be robustly separated from the ‘Chlamydomonas’-snow clade, leaving the latter behind without a genus designation so far.
The three Chloromonas clades in this phylogeny cover the mesophilic Chloromonas augustae (Skuja) Pröschold, Marin, Schlösser & Melkonian, the Chloromonas reticulata (I.N. Goroschankin) Wille clade (both marked as A1 clade in Hoham et al. Reference Hoham, Bonome, Martin and Leebens-Mack2002) and the Chloromonas-snow clade. The C. reticulata clade contains strains isolated from snow habitats, although only one appears as being truly psychrophilic (Chloromonas rosae var. psychrophila strain UTEX B SNO38). Others strains (UTEX LB 1969, SAG 29.83) formerly had been misidentified either as the snow algal taxa Chlamydomonas nivalis or Chlamydomonas yellowstonensis Kol, or proved not be psychrophilic (T. Pröschold, personal communication 2012).
The Chloromonas-snow clade comprises psychrophilic Chloromonas species found solely on snow habitats. All species of this clade produce spindle shape or ellipsoidal cysts (or zygospores) with different forms of ornamentations/flanges on the outer cell wall. The only exceptions seem to be Chloromonas tughillensis Hoham, Bergman, Rogers, Felio, Ryba & Miller, of which only smooth spherical zygotes are described (Hoham et al. Reference Hoham, Berman, Rogers, Felio, Ryba and Miller2006), and Chloromonas platystigma Korshikov in Pascher, with polyhedral, yellowish zygotes with corrugations (see p. 477f. in Ettl Reference Ettl1983). Regarding snow colours, the species of this Chloromonas-snow clade cause rather different shades, ranging from green and yellow to orange. The tentative subclades, especially the A2 subclade, partly show poor resolution and cannot always be separated robustly against each other, disregarding whether the 18S or the rbcL gene was used for the phylogenetic analysis (compare also Hoham et al. Reference Hoham, Bonome, Martin and Leebens-Mack2002, Muramoto et al. Reference Muramoto, Kato, Shitara, Hara and Nozaki2008). Well supported in our tree is the subclade comprising C. platystigma and Chloromonas alpina Wille of which only gloeocysts and aplanospores of unknown colour are known (p. 504 in Ettl Reference Ettl1983). Wille (Reference Wille1903) depicted ellipsoidal aplanospores (?) with a spiked membrane for C. alpina. Chloromonas rostafinskii (Starmach & Kawecka) forming zygotes resulting in yellow-green snow (p. 520 in Ettl Reference Ettl1983) stands separately. A second subclade comprises strain CCCryo 005-99 of Chloromonas nivalis with orange zygotes and cysts forming orange and green snow (Remias et al. Reference Remias, Karsten, Lütz and Leya2010a) together with another Chloromonas strain (CCCryo 047-99) slightly distinct in the morphology of the sporangia and amount of zoospores formed. Resting stages of the latter strain have not been observed yet. This small clade groups together with a field sample of Chloromonas nivalis (DRAlp024) of which sequences were obtained from orange coloured cysts and with the sample of C. polyptera (DRAnt023) forming orange snow. The close relation of C. polyptera to Chloromonas nivalis is reflected by the low bootstrap value at the basis of these clades (only 57 for the distance model). Whether sample “DRAlp024” actually was correctly identified as Chloromonas nivalis (compare Remias et al. Reference Remias, Karsten, Lütz and Leya2010a) or whether it might actually represent C. polyptera remains unclear until a more detailed and more comprehensive analysis is performed.
The third subclade covers different species with resting stages of different colours. Chloromonas brevispina (Fritsch) Hoham, Roemer & Mullet forms green snow from vegetative cells and immature zygotes as well as snow of different shades (green to red) from blunt-spiked zygotes which change their colour from yellow to orange or pink depending on the zygotes’ exposure to direct sunlight (Hoham et al. Reference Hoham, Roemer and Mullet1979). Chloromonas tughillensis forms zygospores green to orange in colour (Hoham et al. Reference Hoham, Berman, Rogers, Felio, Ryba and Miller2006), and those of Chloromonas pichinchae result in yellow-green snow (Hoham Reference Hoham1975). According to Hoham et al. (Reference Hoham, Berman, Rogers, Felio, Ryba and Miller2006) this subclade also comprises Chloromonas chenangoensis Hoham, Berman, H.S. Rogers, Felio, Ryba & P.R. Miller forming green snow, however, their maximum parsimony phylogramme based on rbcL sequences already suggests a separate clade (bootstrap value of 88 in the A2 clade). This is confirmed by Muramoto et al. (Reference Muramoto, Kato, Shitara, Hara and Nozaki2008) who also used rbcL gene sequences. In their phylogeny C. chenangoensis also stands separated from the former clade of C. brevispina, C. tughillensis and C. pichinchae. Interestingly, our Antarctic sample DRAnt023 of C. polyptera directly neighbours the clade containing C. hohamii (compare Fig. 5 in Muramoto et al. Reference Muramoto, Kato, Shitara, Hara and Nozaki2008) and thus confirms the observation by Ling & Seppelt (Reference Ling and Seppelt1998) that the American C. hohamii and the Antarctic C. polyptera are genetically different species and reproductively isolated from one another even though morphologically similar. Unfortunately no sequences of the Antarctic C. polyptera have been analysed so far elsewhere, as a live culture of this species is not available. This problem, that snow algal samples or isolates were first assigned to one single species according to their phenotype, but later had to be split into two separate species due their genotype, which is also justified by the geographic separation (and differences in their life cycle), might also apply to Chloromonas nivalis (syn. C. cryophila) (Hoham & Mullet Reference Hoham and Mullet1977, Reference Hoham and Mullet1978). Undoubtedly Chloromonas nivalis is closely related to C. polyptera (and most probably also to C. hohamii). Apart from our sequences of Chloromonas nivalis (strain CCCryo 005-99 and possibly also CCCryo 047-99, as well as the field sample DRAlp024) various other sequences assigned to this species have been analysed so far. Hoham et al. (Reference Hoham, Bonome, Martin and Leebens-Mack2002) have assigned their strain CU563D (=UTEX B SNO66) to this species and in their phylogenetic analyses this strain neighbours to C. brevispina and C. pichinchae in their A2 clade. In an analysis of Novis et al. (Reference Novis, Hoham, Beer and Dawson2008) using rbcL data, the same strain also occurs between C. tughillensis and C. chenangoensis. Muramoto et al. (Reference Muramoto, Kato, Shitara, Hara and Nozaki2008) found the same strain UTEX SNO66 at a similar position grouping together with their sequence derived from a field sample of red zygotes from Japan (Gassan-NIV1) which they identified as Chloromonas nivalis as well. However, difficulties with the identification of this species based on zygotes from field material become clear when looking at the position of the other sequence (Gassan-NIV2) in the same publication. In the same publication they assigned this sample to Chloromonas nivalis, and it occurs somewhat distant from the former sequence. Unfortunately none of our strains from Spitsbergen or the European Alps were used in the analyses of the above authors. Regarding the different positions of the various Chloromonas nivalis isolates, obviously all resembling each other morphologically, one has to question if there is a similar effect for this species as for C. polyptera and C. hohamii, namely distinct genotypes from different geographic locations. A more detailed and combined analysis of the strains involved is necessary to resolve this and the assignment to different species might be a consequence.
Muramoto et al. (Reference Muramoto, Nakada, Shitara, Hara and Nozaki2010) included another species found on snow from Mount Gassan in Japan, Chloromonas miwae (Fukushima) Muramoto et al., in their snow algal phylogeny. Their sequence is also found in the A2 clade mentioned above. In their rbcL phylogeny the Chloromonas nivalis strain UTEX B SNO66 has a similar, somewhat separated position within the A2 clade. Chloromonas hohamii groups together with C. brevispina and C. tughillensis, but again unfortunately, as no sequence of C. polyptera could be incorporated at that time into their phylogeny, our finding of the close relationship of C. polyptera with Chloromonas nivalis, respectively its cyst samples, cannot be resolved any further. To complete the phylogeny of species in this Chloromonas-snow clade, two species from the genus Chlainomonas also deserve mention: Chlainomonas rubra (Stein & Brooke) and Chlainomonas kolii (Hardy & Curl) are the only species in this group producing truly red coloured cells forming blooms of pink-red snow (Novis et al. Reference Novis, Hoham, Beer and Dawson2008). The rbcL phylogenies of Novis et al. (Reference Novis, Hoham, Beer and Dawson2008) as well as of Muramoto et al. (Reference Muramoto, Nakada, Shitara, Hara and Nozaki2010) place these unusual quadriflagellate snow algae in a well supported separate clade neighbouring C. chenangoensis within the A2 clade depicted by Hoham et al. (Reference Hoham, Bonome, Martin and Leebens-Mack2002).
In summary, C. polyptera can be regarded as a microalga typical for Antarctic coastal snowfields. Despite the vast geographical distance, this species is closely related to other snow algae from the northern hemisphere. Physiological and cytological characteristics are similar to other species (C. hohamii and Chloromonas nivalis), such as the disk-like chloroplast sections, the formation of cysts with ornamented cell walls, and the accumulation of the carotenoid astaxanthin causing the orange colour. Phylogenetically C. polyptera is separately located in the A2 clade within the Chloromonas-snow clade and is closely related to Chloromonas nivalis from Spitsbergen (strain CCCryo 005-99) and from the European Alps (sample DRAlp024). Its close relationship to C. hohamii from North America on the other side becomes mainly visible in the very similar cyst shape. The exact position of this species, but also a more precise phylogeny of all the different species covered by the Chloromonas-snow clade unfortunately still remains unclear, as until now, no all-encompassing phylogenetic analyses have been performed and for many strains only either rbcL or 18S rRNA gene sequences are available from the public databases. Thus, a more detailed comparison of our findings based on 18S sequences with those based on the rbcL gene (Hoham et al. Reference Hoham, Bonome, Martin and Leebens-Mack2002, Reference Hoham, Berman, Rogers, Felio, Ryba and Miller2006, Novis et al. Reference Novis, Hoham, Beer and Dawson2008, Muramoto et al. Reference Muramoto, Nakada, Shitara, Hara and Nozaki2010) was not possible in this study. For future analyses of this very complex algal group, the use of ITS gene sequences in addition to rbcL and 18S sequence information might help to differentiate between geographic populations and to prove species identity.
Acknowledgements
We thank the German AWI (Bremerhaven) for hosting at Dallmann Laboratory (Jubany Station), Austrian Science Fund (FWF) P20081 to CL, and Werner Kofler (Institute of Botany, University of Innsbruck) for taking the SEM images. We gratefully acknowledge the constructive comments of the reviewers.
