Four Achnanthidium species (Bacillariophyta) formerly identified as Achnanthidium minutissimum from the Antarctic Region

. Four taxa belonging to the complex of species around Achnanthidium minutissimum were found during the ongoing taxonomic revision of the Antarctic freshwater and limno-terrestrial diatom flora. Two taxa were previously described as Achnanthidium lailae and A. sieminskae . Two others were formerly identified as A. minutissimum but detailed light and scanning electron microscopical observations revealed sufficient morphological differences compared to the type of A. minutissimum , to justify their separation and description as new taxa: Achnanthidium indistinctum and A. maritimo-antarcticum . The morphology and ecology of all four taxa are discussed comparing the species with morphologically similar taxa. The biogeographical consequences of the splitting of the former A. minutissimum complex in the Antarctic Region are discussed.


Introduction
Based on the list of all published diatom records (40 entries) from the Antarctic Region in Kellogg & Kellogg (2002), Achnanthes minutissima Kütz.(Kützing 1833) is one of the most commonly observed diatom taxa in the (sub-)Antarctic Region.Apart from the Antarctic Continent, the species has been reported from all investigated localities on the sub-Antarctic Islands in the southern Indian Ocean (Bourrelly & Manguin 1954;Le Cohu 2005;Van de Vijver et al. 2002) and the Maritime Antarctic Region (e.g., South Shetland Islands, South Orkney Islands) (Jones et al. 1993;Oppenheim 1994).The species often forms large populations on the sub-Antarctic islands, dominating the aquatic diatom communities in lakes and pools (Van de Vijver et al. 2001, 2002, 2004, 2008).All findings contributed to the generally accepted idea that A. minutissima was a typical cosmopolitan species showing a worldwide distribution on all continents (Ector 2011).
The correct identity and biogeographical distribution of this taxon has for a long time been obscured by force-fitting and taxonomic drift.As a correct interpretation of the type of A. minutissima was virtually unknown, all identifications were based on the short description in Kützing (1833).In 1994, Czarnecki transferred the species to the genus Achnanthidium as Achnanthidium minutissimum (Kütz.)Czarnecki (Czarnecki 1994).The type material of Achnanthes minutissima remained unstudied until Potapova & Hamilton (2007) published for the first time several SEM images from the type material.Following a taxonomic workshop in Luxemburg in 2009 (Ector 2011), it was clear that the old concept of A. minutissimum was no longer acceptable and that a revision of all populations formerly identified as A. minutissimum was necessary.A special volume of Algological Studies (published in 2011) describing the workshop results, was entirely dedicated to the study of the Achnanthidium minutissimum-complex.A large number of new taxa were described, split off from the former catch-all taxon A. minutissimum and some of its varieties.Since 2009, more than 15 new Achnanthidium species have been described such as A. caravelense Novais & Ector (Novais et al. 2011) (Wojtal et al. 2011).Simultaneously to the revision of the European and North American Achnanthidium minutissimum complex, several Antarctic populations have been re-investigated which resulted in the separation of two new species with a purely Antarctic distribution.In 2009, Zidarova et al. described Achnanthidium lailae Van de Vijver (Zidarova et al. 2009) from James Ross Island located in the northern Weddell Sea, close to the Antarctic Peninsula.Three years later, Achnanthidium sieminskae Witkowski, Kulikovskiy & Riaux-Gobin (Witkowski et al. 2012) was described from the sub-Antarctic Iles Kerguelen in the southern Indian Ocean.
During a survey of the freshwater diatom flora from the Maritime Antarctic Region, two other taxa have been found that were previously identified as A. minutissimum.Analysis of their ultrastructure using scanning electron microscopy revealed significant differences with the type of Achnanthidium minutissimum and with all other recently split-off taxa justifying their description as new taxa.The present paper gives a formal description for both new taxa, Achnanthidium maritimo-antarcticum sp.nov.and A. indistinctum sp.nov., and discusses the morphology, ecology and distribution of the two other Antarctic taxa, A. lailae and A. sieminskae.

Material and methods
Samples used in this study were collected during several field trips from different aquatic habitats on the South Shetland Islands (Livingston Island), James Ross Island and the sub-Antarctic Islands in the southern Indian (Iles Crozet, Iles Kerguelen, Prince Edward Islands) and Atlantic Ocean (South Georgia) (Fig. 1).Details on sampling methods and physico-chemical analyses can be found in Van de Vijver & Beyens (1996), Van de Vijver et al. (2001, 2002, 2004, 2008), Kopalová et al. (2013) and Kopalová & Van de Vijver (2013).
Diatom samples for LM observation were prepared following the method described in Van der Werff (1955).In total, ten sampling sites from various Antarctic localities were examined in this study, with all samples listed in Table 1.Subsamples of the original material were oxidized using 37% H 2 O 2 and heating to 80°C for approximately 1h.The reaction was further completed by the addition of KMnO4.Following digestion and centrifugation (three times 10 minutes at 3700x g), the material free of organic matter was diluted with distilled water for sample mounting to avoid excessive concentrations of diatom valves and frustules on the slides.A subsample from the organic-free material was mounted in Naphrax® for diatom community studies.The slides were analysed using an Olympus BX53 microscope, equipped with Differential Interference Contrast (Nomarski) and an Olympus UC30 digital camera.Samples and slides are stored at the BR-collection, property of the Belgian federal government and given in permanent loan to the Botanic Garden Meise (Belgium).For scanning electron microscopy (SEM), parts of the oxidized suspensions were filtered through a 1-µm Isopore™ polycarbonate membrane filter (Merck Millipore).The stubs were sputter-coated with a Gold-Palladium layer of 10 nm and studied in a ZEISS ULTRA SEM microscope at 3 kV (Natural History Museum London, UK).Diatom terminology follows Ross et al. (1979), Round et al. (1990) and Potapova & Hamilton (2007).For comparison, the following publications were consulted: Krammer & Lange-Bertalot (1991), Kobayashi (1997), Potapova & Ponader (2004), Ivanov & Ector (2006), Potapova (2006), Monnier et al. (2007), Ponader & Potapova (2007), Potapova & Hamilton (2007), Zidarova et al. (2009), Hlúbiková et al. (2011), Morales et al. (2011), Van de Vijver et al. (2011a, 2011b) and Witkowski et al. (2012).

Etymology
The specific epithet was given because the species is rather small and can be easily overlooked.

Ecology, distribution and associated diatom flora
So far, A. indistinctum has only been found on Livingston Island (South Shetland Islands).The type population was observed living on submerged mosses collected in a large lake on the central plateau, located between Midge Lake and Limnopolar lake.The lake had an alkaline pH (7.35) with a low specific conductance value (97 µS/cm) and a high phosphate level (5.2 mg/l).The flora in the sample was  (Bukhtiyarova & Round 1996)], Nitzschia cf.perminuta (Grunow) M.Peragallo (Peragallo 1903), Brachysira minor (Krasske) Lange-Bert. in Lange-Bertalot & Moser (Lange-Bertalot & Moser 1994), Sellaphora seminulum (Grunow) D.G.Mann (Mann 1989) and Amphora sp.It is possible that the species is also present on Signy Island (South Orkney Islands) based on the illustrations in Oppenheim (1994).The distribution in the Antarctic region and other geographical areas is unclear due to the insufficient discrimination between species in the Achnanthidium minutissimum-complex.All published records should be verified in order to know the exact biogeography of this and all other taxa.49) slightly convex with very narrow, linear axial area, widening to form a weakly elliptically, never transapically elongated central area, never forming a fascia or subfascia.Central striae slightly more distantly spaced.Striae slightly radiate throughout, only slightly more denser near the apices, 30-32 in 10 µm.

Etymology
The specific epithet refers to the geographical area where the new species was found: Maritime Antarctic Region.The hyphen can and should be maintained in the species name based on art.60.9 of the International Code of Nomenclature for algae, fungi and plants (McNeill et al. 2012).

Ecology, distribution and associated diatom flora
The type population of A. maritimo-antarcticum was found in a large lake located on the central plateau of Byers Peninsula.The lake is characterized by a weakly alkaline pH (7.5) and a low conductivity (130 µS/cm).The ion composition in the lake is mainly dominated by Na + (20.0 mg/l) and Ca 2+ (12.1mg/l).Both nutrients (N-NO 3 +N-NO 2 ) and phosphate were rather low (25 µg/l and 10 µg/l resp.).The shoreline of the lake was completely covered with partly submerged mosses.Dominant taxa in the sample included Fragilaria capucina s.

Ecology, distribution and associated diatom flora
The type population was found in a large circumneutral lake (pH = 7.1) on Ulu Peninsula on James Ross Island (Zidarova et al. 2009).Since then, several other large populations were discovered in lakes on Clearwater Mesa, a volcanic tableland on James Ross Island next to Ulu Peninsula.All populations were observed in alkaline lakes (pH 8.1-8.7) with relatively high conductivity (1000-2000 µS/cm).The samples were dominated by Halamphora sp., Pinnularia australomicrostauron Zidarova et al. (Zidarova et al. 2012)

Ecology, distribution and associated diatom flora
Achnanthidium sieminskae was originally described as a brackish-littoral taxon from a tidal flat near the Kerguelen Islands (Witkowski et al. 2012) but the authors already expressed some doubts about this observation.Most likely, their population was the result of a riverine deposition in the marine Bossière fjord on Kerguelen.As they stated themselves, the only way to find out would be the analysis of the living material which was unfortunately no longer available at the time of their publication.We reported this taxon (as A. minutissimum) from a large number of purely freshwater habitats ranging from small pools to larger lakes and even streams from all sub-Antarctic islands in both the Indian and Atlantic Ocean (Van de Vijver & Beyens 1996; Van de Vijver et al. 2001, 2002, 2004, 2008).In many cases, very large populations could be observed, occasionally up to 60-80% of all counted valves.These data indicate that this is typical limno-terrestrial taxon preferring wet moss vegetations in pools, bogponds and lakes with a pH between 5.2 and 7.4 and a low conductivity (< 160 µS/cm).These data confirm the final conclusion in Witkowski et al. (2012) that the observed valves in their study are the result of the redeposition of a freshwater species in a marine environment.Given the large amount of available data and material, it is a pity the original authors did not verify their rather unusual observation more carefully.

Discussion
All taxa discussed here clearly belong to the complex of species around Achnanthidium minutissimum based on the simple, straight distal raphe endings contrary to the species in the A. pyrenaicum group that have clearly deflected to even hooked distal raphe fissures (Kobayashi 1997;Potapova & Hamilton 2007).In the (even recent) past, all taxa discussed were identified as Achnanthidium minutissimum (e.g., Oppenheim 1994;Van de Vijver et al. 2002;Le Cohu 2005;Kopalová & Van de Vijver 2013).Table 2 summarizes all important morphological features of all four taxa formerly identified as A. minutissimum.The application of a more fine-grained taxonomy based on a narrow species concept (Mann 1999), the detailed analysis of type material of so-called cosmopolitan taxa and as a consequence no longer force-fitting (Tyler 1996)  A similar conclusion can now be made for the genus Achnanthidium confirming once more the unique nature of the Antarctic diatom flora and contradicting the Ubiquity hypothesis (Finlay & Clarke 1999) stating the cosmopolitan nature of all micro-organisms.As a consequence, the genus Achnanthidium shows some interesting biogeographical aspects.At present, only six Achnanthidium taxa are known from the Antarctic Region.Apart from the four taxa discussed in this paper, two other Achnanthidium taxa, Achnanthidium modestiforme (Lange-  (Taylor et al. 2014), a taxon recently described in the complex of species around A. exiguum (Grunow) Czarnecki (Czarnecki 1994), were confirmed from the Region.All six taxa are so far exclusively found in the Antarctic Region despite records of A. modestiforme from Argentina, Bolivia, Brazil and Chile (Maidana 1996;Díaz Villanueva & Maidana 1999;Rumrich et al. 2000;Tremarin et al. 2009;Morales et al. 2011) . australexiguum), all of which so far never observed outside this Region.It is however unclear whether this biogeography represents the complete distribution of these taxa or is just the consequence of undersampling and underreporting.Kellogg & Kellogg (2002) list almost 40 records of Achnanthidium minutissimum (usually as Achnanthes minutissima) from mostly the Maritime Antarctic Region.All records should be verified to complete our biogeographical knowledge of the species discussed in this paper.Most literature records lack however (good) illustrations and even the retrieval of the original material is not always very straightforward making new sampling in these areas a more obvious choice to clarify the biogeography of these taxa.
Although molecular data would be interesting to confirm the separation of the taxa of the A. minutissimum complex, it is clear that all taxa discussed here are morphologically distinct from each other andare not conspecific with the former catch-all taxon A. minutissimum.Potapova & Hamilton (2007) already presented some SEM images of the type material of A. minutissimum s.s.Based on these published images and additional (but so far unpublished) observations on the type material of A. minutissimum (Novais et al. submitted), it is clear that none of the observed populations in the Antarctic Region represent A. minutissimum s.s.
Achnanthidium indistinctum can be separated from A. minutissimum based on several features (Potapova & Hamilton 2007;Hlúbiková et al. 2011) The second new taxon, A. maritimo-antarcticum can be separated from all other Achnanthidium taxa by its more linear valve outline in combination with the protracted, broadly rounded apices.The striae are solely composed of rounded to transapically elongated areolae, never presenting slit-like areolae as often found in several Achnanthidium taxa such as A. hoffmannii, A. minutissimum and A. lailae.Zidarova et al. (2009) already discussed the differences between A. lailae, A. minutissimum and several other narrow, linear Achnanthidium taxa such as A. strictum Reichardt (Reichardt 2004) and A. lineare W.Sm. (Smith 1855).The analysis of additional populations of A. lailae on Clearwater Mesa, James Ross Island, confirm entirely the original separation of this taxon from the above mentioned taxa A. minutissimum, A. strictum and A. lineare, based on the weakly deflected distal raphe fissures and the higher number of areolae per stria in A. lailae.The only feature that need to be adjusted compared to the original description, is the presence of slit-like areolae near the valve margins.In the original description, the absence of these slit-like areolae was mentioned (Zidarova et al. 2009).
Our observations on the often very large freshwater populations of Achnanthidium sieminskae from Iles Kerguelen refine our knowledge on the morphology of this species.According to the original description (Witkowski et al. 2012), the number of areolae per stria on the rapheless valve was set to 3, sometimes 2 areolae although in the accompanying table, 3-5 areolae are mentioned.Based on our results, and a more careful reanalysis of the illustrations in Witkowski et al. (2012), it is clear that the striae are composed of 2-5 rounded, sometimes slit-like areolae.Moreover, the populations observed on Iles Kerguelen and the other sub-Antarctic islands in the southern Indian Ocean, fine-tuned the morphometric data for this species.The observed length and width range in the freshwater and limno-terrestrial populations is larger than in the original description.Some raphe valves seem to lack one or two striae in the central area.The typical presence of 1-2 striae in the central area of A. sieminskae was used by Witkowski et al. (2012) as a discriminating feature to separate the species from A. minutissimum.It is clear that this feature should no longer be used.The separation of both taxa can also no longer be based on the denser striae near the apices as analysis of the Kützing type material of A. minutissimum (Novais et al. submitted) showed the presence of dense striae near the apices in the latter.Conspecificity of both taxa is however to be excluded but the separation of both taxa is therefore only possible based on the valve outline with more linear to linear-lanceolate valves (contrary to the typical lanceolate valves in A. minutissimum), the presence of clearly rostrate to capitate apices and differences in the number of areolae on the rapheless valve.In their analysis of the type of A. minutissimum, Novais et al. (submitted) show a rapheless valve having at least 5-6 rounded to weakly elongated areolae per stria whereas in the A. sieminskae valves shown in this paper, only 2-3(-4) rounded and one slit-like areola are shown.It is also clear that the raphe in A. sieminskae is usually more distinctly visible in LM than in A. minutissimum.Achnanthidium sieminskae shows some similarity to a few other (usually recently described) Achnanthidium taxa.

Conclusions
The present paper highlights the need for thorough morphological revisions of the entire Antarctic freshwater and limno-terrestrial diatom flora applying a narrower species concept.The description of two new Achnanthidium taxa split off from Achnanthidium minutissimum s.l.improves our knowledge about the biogeography of species in this complex and allows their use in more precise ecological and biogeographical studies.
in the framework of the terrestrial program 136 (Marc Lebouvier & Yves Frenot).Samples on Byers Peninsula were taken in the framework of the IPY-Limnopolar Project POL2006-06635 (Ministerio de Ciencia y Tecnología, Spain).The authors would also like to thank the members of expeditions to the Czech J.G. Mendel Antarctic Station for field support and assistance.K. Kopalová benefited from the GA UK grant nr.394211.Part of the research was funded within the Belspo CCAMBIO project and an EU Synthesys grant to BVDV to visit the National History Museum in London, UK.Alex Ball, the staff of the EMMA laboratory and Eileen J. Cox at the Natural History Museum are thanked for their help with the scanning electron microscopy.

Fig. 1 .
Fig. 1.The Antarctic region with the location of the sampled islands and archipelagos.

Figs 2 -
Figs 2-28.Achnanthidium indistinctum Van de Vijver & Kopalová sp.nov.Light and scanning electron micrographs of the type population on Byers Peninsula (Livingston Island).2-4.LM views of some frustules in girdle view.5-15.LM views of raphe valves.16-24.LM views of rapheless valves.25.SEM external view of an entire rapheless valve.26.SEM external view of an entire raphe valve.27.SEM internal view of an entire rapheless valve.28.SEM internal view of an entire raphe valve.Scale bars represent 10 µm.
l. Desm.(Desmazières 1825), Staurosirella sp., Navicula cremeri Van de Vijver & Zidarova in Van de Vijver et al. (Van de Vijver et al. 2011d) and Nitzschia cf.perminuta.Smaller populations were found in other lakes such as Limnopolar Lake but these populations never exceed 1% of the total diatom composition.Due to force-fitting and taxonomic drift, the distribution of the new taxon in the Antarctic Region is unclear.Achnanthidium lailae Van de Vijver in Zidarova et al. (Zidarova et al. 2009) Figs 54-77 Morphological observations Light microscopy (Figs 54-74) Frustules in girdle view narrow, rectangular, bent around the transapical axis, apices weakly recurved (Figs 54-56).Valves linear to very slightly linear-lanceolate with almost parallel margins and nonprotracted, broadly rounded, never rostrate or capitate apices (Figs 57-74).Valve dimensions (n=30): length 10-14 µm, valve width 1.8-2.5 µm.Raphe valve (Figs 57-64) concave with a rather narrow, linear to linear-lanceolate axial area, widening towards the central area.Central area forming a typical rectangular fascia.Shortened marginal striae occasionally present in the central area.Raphe straight to weakly undulating with inconspicuous straight proximal raphe endings.Distal raphe fissures not discernible in LM.Striae weakly but still distinctly radiate near the valve center, becoming more radiate near the apices, 30-33 in 10 µm.Rapheless valve (Figs 65-74) slightly convex with moderately broad, clearly lanceolate axial area, widening near the valve centre.Central area elongated, rhombic lanceolate, never expanding into a fascia due to several longer marginal striae in the central area.Striae parallel to weakly radiate near the valve centre, more radiate near the apices, 28-30 in 10 µm.

,
Figs 54-77.Achnanthidium lailae Van de Vijver.Light and scanning electron micrographs of a population on Clearwater Mesa (James Ross Island).54-56.LM views of some frustules in girdle view.57-64.LM views of raphe valves.65-74.LM views of rapheless valves.75.SEM external view of an entire rapheless valve.76.SEM internal view of an entire rapheless valve.77.SEM external (right) and internal (left) view of an entire raphe valve.Scale bars represent 10 µm.
margins and clearly protracted, rostrate to even capitate apices (Figs 81-99).Valve dimensions (n=25): length 9.5-18.0µm, width 1.9-3.1 µm.Raphe valve (Figs 81-91) concave with a very narrow linear axial area, almost not discernible near the valve apices.Central area elliptical to rounded, very small, almost indistinct, formed by 2-3 more widely spaced striae.Central striae often lacking extending the central area up to the valve margin.Raphe straight, filiform with simple, straight proximal raphe endings.Distal raphe hardly discernible in LM.Striae weakly radiate near the valve center, becoming parallel and even convergent near the apices, well visible in LM, ca.32 in 10 µm, more densely spaced near the valve ends.Rapheless valve (Figs 92-99) slightly convex with narrow, linear axial area, widening to form a weakly elliptically, never transapically elongated central area.Fascia never present.Central 2-3 striae more distantly spaced.Striae weakly radiate throughout, becoming more parallel to even convergent near the apices, ca.35 in 10 µm.
and taxonomic drift, clearly influenced our understanding of the biogeography of Antarctic diatoms as shown in the taxonomic revision of several genera such as Muelleria (Van de Vijver et al. 2010), Luticola (Van de Vijver & Mataloni 2008, Van de Vijver et al. 2011c), Hantzschia (Zidarova et al. 2010) and Navicula (Van de Vijver et al. 2011d).
, A. acerosum Van de Vijver, Lange-Bert.& Jarlman (Van de Vijver et al. 2011a) and A. acsiae Wojtal, E.Morales, Van de Vijver & Ector Raphe valve (Figs 5-15) concave with a very narrow linear axial area, almost invisible towards the valve apices, almost not widening towards the central part of the valve.Central area very small, indistinct, bordered by several (usually two) more widely spaced striae.Raphe straight, filiform, almost invisible, with simple proximal raphe endings.Distal raphe fissures not discernible in LM.Striae moderately radiate throughout the entire valve, becoming indiscernible (due to higher density) near the apices.Rapheless valve (Figs 16-24) slightly convex with narrow, linear axial area, not or very slightly widening towards the valve centre.Central area almost non-existing, never forming a fascia or subfascia.Central striae very weakly more distantly spaced.Striae weakly radiate throughout, only very slightly more densely spaced near the apices, 30-35 in 10 µm.

Table 2 .
Overview of the main morphological features of all four taxa formerly identified as A. minutissimum s.l. reported from the Antarctic Region.Bert.)Van de Vijver in Van de Vijver et al. (Van de Vijver et al. 2002) and A. australexiguum Van de Vijver in Taylor et al.
that are clear examples of force-fitting and most likely represent other, yet undescribed taxa.Two of these six taxa have a distribution that is at present only restricted to the sub-Antarctic Region (including all islands and archipelagos in the southern Indian Ocean and South Georgia in the southern Atlantic Ocean): A. sieminskae and A. modestiforme, lacking at present any confirmed records from other parts of the Antarctic Region.On the other hand, the Maritime Antarctic Region counts four Achnanthidium taxa (A.lailae, A. maritimo-antarcticum sp.nov., A. indistinctum sp.nov.and A Van de Vijver et al. (Van de Vijver et al. 2011a)arrower valves(1.8-2.2 µm vs. 2.5-3.1 µm in A. minutissimum).Additionally, A. minutissimum has a higher number of areolae per stria, usually 4-5(Novais et al. submitted), a high number never observed in A. indistinctum where 2-3(-4) is the rule.The mantle areolae in A. indistinctum are often enlarged making them easily visible in LM on frustules in girdle view, similarly to for instance A. hoffmanniiVan de Vijver et al. (Van de Vijver et al. 2011a).Finally, the valve outline of A. minutissimum shows a combination of more convex valve margins together with clearly rostrate to subcapitate apices, compared to the weakly compared convex margins and weakly rostrate margins in A. indistinctum.
(Slate & Stevenson 2007)n de Vijver et al. 2011a) with A. jackii Rabenh.(Rabenhorst1861)and A. lailae, although any confusion with the latter is entirely to be excluded based on valve outline, dimensions and stria structure.Achnanthidium ertzii Van de Vijver & Lange-Bert.inVan de Vijver et al. (Van de Vijver et al. 2011a), described from Madeira, shows a similar elongated valve outline but has narrower valves with a lower number of areolae per stria (2-3 vs. 3-5 in A. sieminskae) making conspecificity highly unlikely.Achnanthidium maritimo-antarcticum, described in this paper, has comparable valve dimensions but a more linear-lanceolate valve outline (compared to the typical lanceolate outline in A. sieminskae), a lower number of areolae per stria (max.3) and less protracted, rostrate to subcapitate apices.The correct taxonomic and morphological identity of other Achnanthidium taxa such as A. caledonicum (Lange-Bert.)J.E.Slate & R.J.Stevenson(Slate & Stevenson 2007)is at present unclear making a comparison between A. sieminskae and these taxa rather difficult.Achnanthidium caledonicum as it is currently known, is much longer and has clearly capitate apices.