An illustrated and annotated checklist of freshwater diatoms (Bacillariophyta) from Livingston, Signy and Beak Island (Maritime Antarctic Region)

1Ghent University, Protistology and Aquatic Ecology, Krijgslaan 281 S8, BE-9000 Gent, Belgium 2University College London, Department of Geography, Pearson Building, Gower Street, London, WC1E 6BT, United Kingdom 3British Antarctic Survey, Natural Environment Research Council, High Cross, Madingley Road, Cambridge, CB3 0ET, United Kingdom 4Botanic Garden Meise, Department of Bryophyta, Thallophyta, Nieuwelaan 38, BE-1860 Meise, Belgium 5University of Antwerp, Department of Biology, ECOBE, Universiteitsplein 1, BE-2610 Wilrijk, Belgium *Author for correspondence: bart.vandevijver@botanicgardenmeise.be


INTRODUCTION
Diatoms are one of the most species-rich and widespread algal groups in sub-Antarctic and Antarctic non-marine environments (Jones 1996, Van de Vijver & Beyens 1999).They are commonly used as indicators for monitoring the ecological status of contemporary environments and reconstructing past environmental and climatic conditions, since many of their species have distinct ecological optima and narrow tolerances, and because their silica cell walls are often well-preserved in fossil sediments.In the ice-free regions of Antarctica, where phanerogamic vegetation is often absent, diatoms are therefore of paramount importance for paleo-ecological Pl.Ecol. Evol. 148 (3), 2015 research (e.g.Jones et al. 2000, Sterken et al. 2012, Verleyen et al. 2003).
Recent revisions of diatom systematics and biodiversity in the Antarctic (e.g.Van de Vijver & Mataloni 2008, Van de Vijver et al. 2010a, 2010b, 2010c, 2011a, 2011b, Kopalová et al. 2011, Zidarova et al. 2012) have shown that in the past many species have been wrongly identified, usually by forcefitting to known European or North-American species (Tyler 1996, see Sabbe et al. 2003 for a detailed discussion of this issue).This practice has had serious repercussions on our understanding of diatom biodiversity and biogeography in the Antarctic, revealing a higher species richness and level of endemism than previously accepted.
This paper presents the results of a detailed taxonomic intercalibration of non-marine diatom materials from three islands in the vicinity of the Antarctic Peninsula, viz.Signy Island (South Orkneys), Livingston Island (South Shetlands) and Beak Island (James Ross Island group).The checklist is based on a re-analysis of samples from Livingston, Signy Island, which have previously been used for the construction of diatom transfer functions (cf.Jones et al. 1993, Jones & Juggins 1995), supplemented with newly collected recent and fossil material from lakes and ponds on Beak Island (NE-Antarctic Peninsula, Sterken et al. 2012).Recently, numerous new studies on the taxonomy of (sub-)Antarctic diatoms have been published, and therefore a re-analysis and intercalibration of these materials was timely.The results of this intercalibration form the basis for several transfer functions (e.g. of nutrients, Sterken 2009) that may be applied in a wider geographical area, especially after intercalibration with new, larger datasets, which will be published elsewhere.All entries are annotated, illustrated and provided with original morphometric data.

Study area
Livingston Island (62°40′S 61°00′W) is the second-largest island of the South Shetland Islands (fig.1).The island has a maritime Antarctic climate strongly influenced by westerly winds, resulting in higher precipitation rates and milder temperatures than in continental Antarctica (Kopalová & Van de Vijver 2013 and references therein).Livingston's geology consists of Jurassic-Cretaceous shales, sandstones and volcanic rocks (Hobbs 1968).The snow-free ground is usually barren and rocky with sparse vegetation, mainly consisting of lichens and mosses in inland sites, and a more extensive cover of lichens, mosses and two flowering plants in the coastal areas (Jones et al. 1993, Kopalová & Van de Vijver 2013).Samples were obtained from lakes in Byers Peninsula, the largest ice-free area consisting of an inland, upland plateau (between 20 and 193 m a.s.l.) and a low-lying coastal area (Jones et al. 1993).In summer, most of the lakes on Byers Peninsula, situated on the barren central plateau, are ice-free and well-mixed by wind.They are generally nutrient poor, except for the brackish coastal lakes, which are heavily influenced by sea spray and in some cases by large animal populations (Jones et al. 1993, Kopalová & Van de Vijver 2013).
Signy Island (60°43′S 45°38′W) is a low-lying (max.altitude 279 m) island belonging to the South Orkney Islands (fig.1).Approximately 32% of the island is covered with a thin, climate-sensitive ice-cap (Hodgson & Convey 2005).Similarly to Livingston Island, Signy has a maritime Antarctic climate.The island mainly consists of metamorphic rocks (schists), and the soils are generally more acid than those on Livingston Island (Jones et al. 1993).The ice-free areas of Signy Island are well vegetated with patches of mosses, lichens, and some flowering plants (Smith 1972), and the catchments of the lakes are generally more densely vegetated than those on Livingston Island (Jones et al. 1993).Signy Island lakes are ice-free and well-mixed in summer, but experience a (several weeks) shorter period of ice cover than those on Livingston Island, due to the lower snow accumulation, which causes a lower albedo at the lake surfaces (Jones et al. 1993).Lakes on Signy Island are generally more nutrientrich than those on Livingston Island, which is mainly due to the influence of animals (birds, fur seals) in the Signy Island lake catchments (Jones et al. 1993).A detailed description of the geology and vegetation of both Signy and Livingston islands, and the physical, chemical and biological characteristics of the study lakes are given in Jones et al. (1993), Jones & Juggins (1995), Heywood et al. (1979Heywood et al. ( , 1980) ) and references in these papers.Except for Oppenheim (1994), who described and illustrated several Achnanthes (sensu lato, s.l.) species, none of the other limnological and/or diatom studies in these areas have provided diatom illustrations.
Beak Island (63°36′S 57°20′S) is a small, ice-free island located in the northern Prince Gustav Channel, between Eagle Island and the Tabarin Peninsula (fig.1).The island is assumed to either have a similar continental climate regime as James Ross Island (50 km SW of Beak, with mean monthly temperatures below 0°C (Björck et al. 1996) or a climate in between maritime and continental (Hawes & Brazier 1991), as winter temperatures only occasionally fall below -20°C (data for nearby Seymour Island, Jones & Limbert 1989).James Ross Island is influenced by the rain shadow effect of the mountains of the Antarctic Peninsula (AP), and is probably characterized by yearly maximum precipitation values of 100-200 mm yr -1 (Sugden 1982).Beak Island is composed of Miocenic volcanic rocks (James Ross Volcanic Island group; Bibby 1966).The island contains a few lakes and shallow ponds, with small, barren catchments, partly vegetated by moss banks (see Sterken et al. 2012).Details about the lakes and their limnology can be found in Sterken et al. (2012).

Sampling and microscopic analyses
For details of sampling and sample preparation of the Livingston and Signy Island materials from lacustrine (recent and (sub)fossil) surface sediments, see Jones & Juggins (1995) and Jones et al. (1993).Detailed information on the Beak Island material (recent and (sub)fossil samples) can be found in Sterken et al. (2012) and Sterken (2009).Samples from Beak Island were oxidized using hydrogen peroxide (Renberg 1990); oxidized materials were mounted in Naphrax.Original light microscope (LM) pictures of all materials were taken with an Olympus DP50 camera on an Olympus CX 41 microscope, and an Axiocam on a Zeiss Axioplan II microscope, at 10x100x magnification.Samples and slides are stored at the Protistology and Aquatic Ecology Lab, Ghent University, Belgium and at UCL, London, UK.
Morphometric data (L= length, W = width, D = diameter, S = stria density, F = fibula density) are given as ranges and number of specimens measured (n).Stria densities (S) were measured alongside the raphe between the central area and the apex.

DISCUSSION
A total of 102 diatom taxa (including varieties and undescribed forms) belonging to at least 34 different genera has been observed in recent and (sub)fossil sediments from 66 lakes in these islands.As a result of recent major revisions of the diatom flora of the Maritime Antarctic Region, almost all diatoms found in this study could be identified to species level.Significantly, 41 of these have been described very recently (since 2000) from the Antarctic region.A large number of these taxa were previously force-fitted into European or North-American taxa.Despite this renewed research effort on Antarctic diatom taxonomy there are still many unknown taxa remaining and species complexes that are in urgent need of a revision.This is especially the case for the genera Gom phonema and all Nitzschia species, at present mostly tentatively identified (hence 'cf.').Several species could as yet not be identified and are being described as new species (e.g.Navicula sp. 1, B. Van de Vijver, unpubl.res.) whereas of the remainder, several closely resemble (hence 'cf.') known species or varieties but display slight but possibly significant differences with the type and/or other populations.Other taxa (e.g.Pinnularia borealis s.l., Fragiliaria capucina s.l., Nitzschia paleacea, Encyonema minutum) appear to display a considerable degree of intraspecific variation (noted in other studies), which requires further study.Given the fact that recent diatom studies have revealed extensive semicryptic diversity in many established diatom taxa (e.g.Behnke et al. 2004, Beszteri et al. 2007;see Mann (1999) for a review), each case needs to be carefully studied (at least including morphometry of different populations and SEM analyses, and if at all possible, molecular analyses) to assess its true identity.It is highly likely that in several cases, this will lead to the splitting of the species in several independent (newly to describe) taxa.
The most species-rich genera in our study were Pinnula ria (twelve taxa) Chamaepinnularia, Luticola, Planothidium, Psammothidium and Stauroneis (seven taxa each), Nitzschia (six taxa), Humidophila and Navicula (five taxa each).Most taxa were found on Livingston Island (86 taxa) whereas on the more southerly located Beak Island, only 54 taxa were observed.It is generally established that in the Antarctic diatom diversity decreases when moving southwards (Jones 1996).Our results seem to confirm this trend, although it cannot be ruled out that the lower number of taxa on Beak Island is due to the smaller number of lakes on the island.Thirty-four taxa were shared by all three islands while 38 taxa were only found on one island.Of these, Livingston had the largest number of unique taxa (24) and Signy Island the lowest.However, on Signy Island, all unique taxa are shared with the sub-Antarctic Region such as Staurosirella circula and Adlafia bryophila.These taxa have never been found in previous diatom surveys in the Maritime Antarctic Region (see for instance Kopalová & Van de Vijver 2013, Kopalová et al. 2012, 2013, 2014).On Beak Island, several unique taxa such as Chamaepinnularia cymatopleura, Craticula antarc tica or Pinnularia splendida are shared with neighboring James Ross Island or the Antarctic Continent but have never been found in the northern part of the Maritime Antarctic Region.These results suggest that bioregionalism may occur within the Antarctic diatom flora.

Fig. 3AO-AP & AV-AW
Dimensions -L: 8.0-14.0µm; W: 3.0-4.0µm; S: 24-32 in 10 µm (n = 20).Remarks -Our specimens resemble Navicula (?) sp. 2 in Sabbe et al. (2003) but the specimens illustrated there are narrower (2.6-3.0 µm).They are also similar to Chamae pinnularia evanida (Hust.)Lange-Bert.(cf.Krammer & Lange-Bertalot 1986) but are wider, have a slightly larger central area and on average a higher stria density (N.evanida: L: 8-10 µm; W: 3 µm; S: 24 in 10 µm, in Hustedt 1942; L: 6-10 µm; W: 2.5-3 µm; S: 24-28 in 10 µm, in Krammer & Lange-Bertalot 1986; L: 7-10 µm; W: 2-3 µm; S: 23-25 per 10 µm, in Van de Vijver et al. 2002).Chamaepinnularia sp. 1 also resembles Chamaepinnularia spec. in Van de Vijver et al. (2002: pl. 86 , Figs 15-22, W: 2-3 µm; S: 23-26 in 10 µm) but these are narrower and generally less finely striated.Wetzel et al. (2013) revised several small Chamaepinnularia taxa but none of them entirely match the Antarctic populations.Most likely these Antarctic populations represent more than one new species but detailed SEM investigations will be necessary to clarify their identity.Distribution -Livingston Island, Signy Island.Dimensions -L: 23.5-29.5 µm; W: 6.0-7.5 µm; S: 17-20 in 10 µm (n = 9).Remarks -Craticula antarctica has been recently described and discussed in Van de Vijver et al. (2010c).The species shows a rather broad geographic distribution on the Antarctic Continent and on some southerly located islands in the Maritime Antarctic Region such as James Ross Island (Ko palová et al. 2013).However, due to confusion with so-called cosmopolitan taxa such as Craticula molestiformis (Hust.)Lange-Bert., its precise distribution is currently uncertain.Distribution -Beak Island.Rabenhorst (1862Rabenhorst ( : no. 1261).Dimensions -L: 10.0-23.0µm; W: 4.5-6.0µm; S: 14-20 in 10 µm (n = 16).Remarks -Our specimens fit Encyonema minutum as described in Krammer (1997) and Kelly et al. (2005), both with respect to valve characteristics and dimensions.This is the first confirmed record in the entire Antarctic Region.A recent taxonomic revision of the cymbelloid diatoms of the Antarctic did not reveal any populations of this taxon so far (B.Van de Vijver, Botanic Garden Meise, Belgium, pers.obs.).Kellogg & Kellogg (2002) list a large number of records for the sub-Antarctic and Maritime Antarctic Region but so far none of them could be confirmed.It is possible that the observed specimens may represent more than one species but since only a few populations were observed, this is too early to say without detailed SEM analysis.Distribution -Livingston Island, Signy Island.Remarks -Our specimens closely resemble Eolimna (Na viculadicta) elorantana as illustrated in Van de Vijver et al. (2002).It is not certain whether some smaller specimens also belong to this taxon as the central area is much smaller.Our longest specimens are longer than those in Van de Vijver et al. (2002).Eolimna elorantana also resembles Sellaphora (Naviculadicta) seminulum (Grunow) D.G.Mann (Van de Vijver et al. 2002) but this taxon typically has elliptical valves with non-protracted apices.The true taxonomic identity of the Antarctic populations is as yet not clear and detailed SEM research will be necessary.It is highly likely that they represent a new species.Distribution -Signy Island.

Morphotype B Fig. 2F-H
Dimensions -L: 13.5-26.0µm; W: 3.0-4.5 µm; L/W ratio: 3.8-7.1;S: 17-19 in 10 µm (n = 12).Remarks -Fragilaria capucina s.l.clearly represents a complex of taxa that are all very similar to each other and at present it is not possible to identify individual taxa at the species level.Morphotype A corresponds well with the description of Fragilaria capucina morphotype 2 in Van de Vijver et al. (2002: 44, pl. 7, Figs 19-31) although the longest specimens might even belong to their morphotype 3. Morphotype B on the other hand shows similar characteristics as F. capucina morphotype 1 described in Van de Vijver et al. (2002: 41, pl. 7, Figs 1-13).As in Van de Vijver et al. (2002), no difference in stria density was found between the different F. capucina morphotypes.It is highly likely that these morphotypes represent new species but due to a lot of confusion regarding the different types within the F. capucina s.l.complex, the exact taxonomic identity remains uncertain.Distribution -Beak Island, Livingston Island, Signy Island.Gomphonema spp.Fig. 4A-H Dimensions -L: 13.5-40.0µm, W: 4.0-8.0µm; S: 10-26 in 10 µm (n = 124).Remarks -It is currently not possible to identify the Maritime Antarctic Gomphonema population at the species level.It is highly likely that several separate species are present but detailed morphometric and SEM analysis will be necessary to elucidate variation patterns in this complex (J.P. Kociolek, University of Boulder, USA, pers.comm.).Our specimens belong to the variable species complex around Gomphonema angustatum (Kütz.)Rabenh.and varieties (e.g.G. angusta tum var.productum) and similar species such as G. micropus Kütz.(Reichardt 1999), G. gracile Ehrenb.and G. parvulum Kütz.They are particularly common in the Antarctic (e.g.Wasell & Håkansson 1992, Sabbe et al. 2003).Some valves are very similar to Gomphonema signyensis described by Kociolek & Jones (1995) but the variability of this species is currently unsufficiently known to make clear taxonomic decisions.Distribution -Beak Island, Livingston Island, Signy Island.Van 2014b) recently revised all Halamphora taxa present in the Antarctic Region.Two closely related taxa were observed in the Maritime Antarctic Region, separated from each other by valve width and differences in stria structure.Halamphora ausloosiana typically shows biseriate striae whereas H. oligotraphenta has uniseriate striae.This feature is however only discernible in SEM.Based on valve width, most observed valves belong to H. ausloosiana as all observed H. oligotraphenta populations in the Maritime Antarctic Region have a maximum valve width of only 4.5 µm.However, detailed SEM anal-ysis of all populations will be necessary to unambiguously distinguish between both taxa.The heavily silicified valves (fig.4N) often found in the populations, have so far not been observed in H. ausloosiana (Van de Vijver et al. 2014b).Distribution -Beak Island, Livingston Island, Signy Island.

Halamphora ausloosiana
Hantzschia amphioxys (Ehrenb.)Grunow f. muelleri Ko-bayashi Fig. 5M Original description - Ko-Bayashi (1963: 62).Dimensions -L: 28.5-53.0µm; W: 4.5-6.0µm; S: 22-24 in 10 µm; F: 6-9 in 10 µm (n = 3).Remarks -The Hantzschia amphioxys species complex in the Antarctic Region is characterized by a high degree of morphological variability (see for instance Ko-bayashi 1965) and is in need of revision (Sabbe et al. 2003).Recently, Zidarova et al. (2010) described five new Hantzschia species and one new variety from Livingston Island.Detailed SEM analysis is necessary to enable a clear separation between the different taxa but due to the extreme paucity of Hantzschia valves in our study, it is impossible to perform a good SEM analysis.The specimens observed in our study most likely represent Hantzschia amphioxys f. muelleri.Zidarova et al. (2010) discuss the taxonomic history of this taxon and its distribution in the Antarctic Region.Distribution -Livingston Island.
Humidophila nienta (Carter) R.Lowe, Kociolek, J.R.Johans., Van de Vijver, Lange-Bert.& Kopalová Fig. 2BL-BN Original description -Navicula nienta Carter (1966: 464).Dimensions -L: 8.5-15.0µm; W: 2.0-2.5 µm; S: not discernible in LM (n = 10).Remarks -There has been confusion in the past regarding this taxon as it was described in 1966 by Carter Krasske (1939: 384).Dimensions -L: 10.5-13.5 µm; W: 4.5-5.0µm; S: 24-28 in 10 µm (n = 7).Remarks -Our specimens agree well with the description of the type of this species in Lange-Bertalot et al. (1996), although they are slightly smaller and narrower (L: 14 -17 µm and W: 5.5-6 µm in Lange-Bertalot et al. 1996).Distribution -Livingston Island.Remarks -Our specimens closely match the description of the type of L. cohnii given in Van de Vijver et al. (2011a) and Levkov et al. (2013).Detailed studies of larger populations will be necessary to assess the morphological variability of the Antarctic specimens.Distribution -Livingston Island, Signy Island.Van  Remarks -The species was recently described from Livingston Island, separating the typical capitate specimens from those of the more broadly rounded L. muticopsis.One of the distinct features of L. katkae is the presence of a clear, irregularly shaped raised axial sternum, visible in LM as an irregular thickening in the axial area.Distribution -Livingston Island.

Luticola vermeulenii Van de Vijver
Microcostatus naumannii (Hust.)Lange-Bert.& Genkal Fig. 2BU-BW Original description -Navicula naumanii Hust.(Hustedt 1942: 115).Dimensions -L: 9.5-14.0µm; W: 4.0-4.5 µm; S: 22-26 in 10 µm (n = 8).Remarks -Recently all Antarctic M. naumannii populations, including one from Livingston Island, have been screened and compared with the type material (Van de Vijver et al. 2010a).The results did not show any significant differences, although the Antarctic populations seemed to show a larger morphometric range, as is also shown by the specimens observed in this study.Our specimens differ slightly from the holotype of M. naumannii illustrated in Simonsen (1987) and Van de Vijver et al. (2010a) in the shape of the axial and central areas, which are respectively less elongate and wider than the holotype, but the differences are so minor that they do not justify the separation of the Antarctic populations as a new species.Distribution -Beak Island, Livingston Island, Signy Island.Remarks -This species was recently described based on material collected from nearby King George Island (Van de Vijver et al. 2011b).The species was frequently reported under different names (e.g.Navicula sp. 1 by Jones et al. 1993) and proved to be quite widespread in the Maritime Antarctic Region (Kopalová et al. 2012, 2013, Kopalová & Van de Vijver 2013).Distribution -Livingston Island.

Nitzschia cf. gracilis Hantzsch Fig. 5AD-AF
Original description -Hantzsch (1860: 40).Dimensions -L: 26.0-43.0µm; W: 2.5-3.5 µm; S: indistinct in LM; F: 14-19 in 10 µm (n = 15).Remarks -Our specimens resemble Nitzschia gracilis Hantzsch (cf. Krammer & Lange-Bertalot 1988, Van de Vijver et al. 2002).The valves in our study are slightly smaller and have narrower and longer apices than those illustrated in Van de Vijver et al. (2002).It is however unclear whether all populations in the Maritime Antarctic Region belong to N. gracilis or in fact represent one or more separate species.A revision on all Nitzschia species in the South Shetland Islands and James Ross Island is ongoing (K.Kopalová, Charles University in Prague, Czech Republic, unpubl. res.).Distribution -Beak Island, Livingston Island, Signy Island.
Planothidium quadripunctatum (Oppenheim) Sabbe Fig. 2AB-AC Original description -Achnanthes quadripunctata Oppenheim (1994Oppenheim ( : 1742)).Dimensions -L: 7.5-9.0µm; W: 2.5-4.5 µm; S: 15-21 in 10 µm (n = 15).Remarks -Our specimens correspond well to those described in Oppenheim (1994), Van de Vijver et al. (2002) and Sabbe et al. (2003).It is often difficult to distinguish this species from Planothidium renei (see below), which has a higher stria density (> 20 in 10 µm) and a different ultrastructure (terminal raphe endings deflected in the same direction in P. quadripunctatum and in opposite directions in P. renei, and the striae are composed of 3-4 instead of 2 transapical rows of areolae in the latter).Preliminary SEM investigations have shown that some populations which were assigned to P. quadripunctatum on the basis of valve dimen-sions displayed ultrastructural features typical of P. renei.There also appears to be an overlap in stria densities in some populations.Further SEM and morphometric analyses are needed to resolve the relationship between these two species.Distribution -Beak Island, Signy Island.

Figure 1 -
Figure 1 -Map of the study area, showing Livingston, Signy and Beak Island in the northern tip of the Antarctic Peninsula.