﻿Two new Halamphora (Bacillariophyta) species from the marine coasts off Livingston Island, Antarctica

﻿Abstract During a survey of the marine benthic diatom flora on the coasts off Livingston Island (South Shetland Islands, Maritime Antarctic Region), two Halamphora species that could not be identified based on the currently available literature, were observed. Detailed light and scanning electron microscopy observations and thorough comparison with similar taxa in the literature revealed that both taxa should be described as new species. The first taxon, Halamphorakenderovianasp. nov., was most likely misidentified in past Antarctic studies, and included within the range of another taxon, Halamphoracoffeaeformis. Analysis of literature data showed that the second new taxon, Halamphoramonchevianasp. nov., has been previously reported from the Antarctic Continent (but as an unidentified species). The new taxa are compared with similar Halamphora taxa worldwide. Data on their ecology and distribution are also provided.


Introduction
In the past two decades considerable effort has been undertaken to improve our understanding of the diversity, species identities and distribution of the terrestrial and freshwater diatoms (Bacillariophyta) in the Antarctic realm (Zidarova et al. 2016 and references therein). In contrast, the marine benthic diatom flora of the region remained far less studied. Although in the past 30 years three new marine benthic diatom genera have been described from islands in the Southern Ocean [Tabulariopsis D.M.Williams (Williams 1988), Brandinia L.F. Fernandes (in Fernandes et al. 2007), and Australoneis J.M. Guerrero & Riaux-Gob. (in Guerrero et al. 2021)], and several new species have been recognized and described within existing genera [including Berkeleya Grev. (Medlin 1990), Cocconeis Ehrenb. (Al-Handal et al. 2008, 2010, Gomphonemopsis Medlin (Al-Handal et al. 2018), Nitzschia Hassall (Al-Handal et al. 2019), Melosira C.Agardh (Fernandes and de Souza-Mosimann 2001), Pteroncola R.W.Holmes & Croll (Almandoz et al. 2014) and Rhoicosphenia Grunow (Ligowski et al. 2014)], a recent study on the marine benthic diatom flora from South Bay (Livingston Island, South Shetland Islands) showed that a large number of the recorded taxa could not be identified with certainty up to species level, with some of these taxa most likely being new to science (Zidarova et al. 2022).
The present paper describes two new marine species in the genus Halamphora (Cleve) Levkov, observed in several recently collected samples from the coasts off Livingston Island, part of the archipelago of the South Shetland Islands (Maritime Antarctic Region).
The genus Halamphora (Cleve) Levkov, originally described in 1895 by Cleve as a subgenus of Amphora Ehrenb. ex Kütz., was raised to genus level in 2009 (Levkov 2009). The characteristic features of the genus include a moderately to strongly dorsiventral valve outline, an eccentric raphe system, uni-to biseriate striae composed of round to elliptical and even transapically elongated areolae, internally occluded by hymenes, and a girdle composed of numerous open copulae. Halamphora species can be found both in freshwater and marine ecosystems (Levkov 2009). In the Antarctic Region, Van de Vijver et al. (2014) revised the freshwater Halamphora species, describing several new taxa, but in the marine realm data on Halamphora (or former Amphora) species are scarce. One of the observed species in the present study has most likely been misidentified and reported in the earlier Antarctic literature under the name of another, presumably widespread marine Halamphora species, H. coffeaeformis (C.Agardh) Kütz., whereas the second taxon is unknown. Following extensive light and scanning electron microscopy observations and comparisons of their morphology with similar taxa from all over the world and from the Antarctic Region, both taxa are described as new species: Halamphora kenderoviana sp. nov. and Halamphora moncheviana sp. nov. A survey of the Antarctic literature with iconographic material provided additional information on their basic ecology and Antarctic distribution.

Materials and methods
Livingston Island is the second largest of the South Shetland Islands, located ca. 130 km north of the Antarctic Peninsula (Fig. 1). Samples were collected in December 2018 and February 2020 from the epilithon of small pools on coastal rocks at the southern coasts of the island, including Hannah Point area and the eastern shores of South Bay (Fig. 1). These pools are small water basins, formed during low tide on or between larger rocks on the coasts, and having variable water temperature and salinity levels (Zidarova et al. 2022 and references therein). The biofilm covering the bottom or sides of the pools was collected using a toothbrush and preserved with 3% formaldehyde in situ. Samples, with their environmental parameters measured with a handheld multi-parameter meter WTW3410 during sample collection, are listed in Table 1. For light microscopy (LM), diatom samples were prepared following the method of Hasle and Fryxell (1970). Cleaned material was mounted in Naphrax. LM observations were conducted using an Olympus BX51 light microscope at 1000× magnification (N.A. 1.30), equipped with Differential Interference Contrast (DIC) optics and Olympus digital imaging system. For scanning electron microscopy (SEM), part of the suspension was filtered through 5 μm Isopore polycarbonate membrane filters (Merck Millipore), after air-drying pieces of which were affixed with double sided carbon stickers (Agar Scientific Carbon Tabs) on 12.7 mm Ø aluminium stubs (Agar Scientific Ltd), coated with a platinum layer of 20 nm and studied using a JEOL-JSM-7100F field emission scanning electron microscope at 2 kV. Slides and stubs are stored at the BR-collection (Meise Botanic Garden, Belgium). Plates with the microphotographs of the species were prepared using Adobe Photoshop. For stria number, measurements were done starting from the valve middle. Terminology for taxa descriptions follows Round et al. (1990), Levkov (2009), and Stepanek and Kociolek (2018). In a search for more ecological and distributional data for the taxa we describe, a survey of the earlier Antarctic literature was conducted, including the major older works (e.g. Van Heurck 1909;Peragallo 1921;Frenguelli and Orlando 1958;Simonsen 1992), as well as more recent reports with iconographic material, such as Roberts and McMinn (1999), Cremer et al. (2003), and others.
Etymology. The new species is named after our colleague Dr Lyubomir Kenderov, hydrobiologist at the Faculty of Biology, University of Sofia, with whom RZ shared two Antarctic seasons, and who was often a helping hand during field work in Antarctica.
Ecology, Antarctic distribution and associated diatom flora. Halamphora kenderoviana was typically observed in tidal pools (Zidarova et al. 2022, as Amphora sp.5), but only found in abundance (17.5% of the counted valves) in the type locality, a tidal pool with alkaline water and a salinity level of 33.1 PSU (Table 1). Other common taxa in the sample are Parlibellus rhombicus W.Greg., Tabulariopsis australis (Perag.) D.M.Williams, and several Navicula species, including N. aff. perminuta Grunow and N. glaciei Van Heurck. So far, H. kenderoviana is known to be present with certainty on the marine coasts of the South Shetland Islands (Livingston Island). Earlier, Cremer et al. (2003) reported a very similar taxon as Amphora coffeaeformis (Cremer et al. 2003, fig. 13) from sediment cores in Windmill Island, East Antarctica. SEM observations will be needed to confirm that it is conspecific with H. kenderoviana, but it seems likely that at least some of the records of A. coffeaeformis, transferred to Halamphora by Levkov (2009)    Description. LM description (Fig. 3A-H). Valves weakly silicified, broadly semielliptic, with a more or less straight ventral and distinctly convex dorsal margin. Apices protracted, subcapitate in larger valves (Fig. 3A), becoming only weakly protracted, rostrate in smaller valves ( Fig. 3F-G). Valve dimensions (n = 23): length 16.0-27.5 μm, width 5.0-7.0 μm. Raphe straight. Central raphe endings straight, enlarged (Fig. 3A, F-H). Terminal raphe fissures not discernible in LM. Axial area narrow, central area absent. Dorsal striae parallel to weakly radiate in the middle, becoming more radiate towards the apices, 24-27 in 10 μm, crossed by several undulating longitudinal lines (Fig. 3A-H).
Etymology. The new species is named after Prof Dr Snejana Moncheva, phycologist and former Director of the Institute of Oceanology at the Bulgarian Academy of Sciences, to thank her for considering our (RZ, NDzh) employment and career possibilities at the Institute.
Ecology, Antarctic distribution and associated diatom flora. Halamphora moncheviana was most abundant in the epilithon of a small coastal pool, having a relatively low salinity (6.5 PSU, sample LT10, Table 1), where it was found together with Craspedostauros laevissimus (W.West & G.S.West) Sabbe and several Nitzschia, Melosira and Navicula species. Roberts and McMinn (1999, Pl. 1, figs 10-11) recorded the same taxon as Amphora sp. d from the Vestfold Hills on the Antarctic Continent, although their reported valves were slightly larger (length 30-35 μm, width 5-8 μm) and with a slightly coarser striation of "approximately" 22 striae in 10 μm. Nevertheless, the SEM photo of the species, identified as Amphora sp. d in Roberts and McMinn (1999, plate 1, fig. 11), presenting a valve externally with striae, composed of a few transapically elongated areolae on the dorsal side and forming irregular longitudinal lines on the valve face, confirms the conspecificity between the species observed on the Antarctic Continent, and H. moncheviana. Roberts and McMinn (1999) reported the species from hypersaline lakes. Based on their and our findings, H. mocheviana is apparently a very tolerant species to changes in salinity. Likely the same taxon was also depicted by Priddle and Belcher (1981, fig. 3l, as Amphora sp.), which they observed in the epilithon of a large, shallow pool (Pool 7) situated near the sea, together with several species of marine origin, including Craspedostauros laevissimus (reported as Tropidoneis laevissima W.West & G.S.West).

Discussion
Based on the observed set of morphological features of H. kenderoviana and H. moncheviana, both these species from the coasts of Livingston Island clearly belong to the genus Halamphora, as defined in Levkov (2009). The morphological analysis also showed that the combination of features in both taxa is sufficiently unique to justify their description as new species.
Halamphora kenderoviana is one of the many Halamphora species, having a valve outline with protracted apices and biseriate striae, similarly to Halamphora coffeaeformis and its related taxa (Stepanek and Kociolek 2018). The morphology of H. coffeaeformis, an often-misinterpreted species, was studied in detail by Archibald and Schoeman (1984), and later discussed in Levkov (2009) and in Stepanek and Kociolek (2018). In contrast to H. coffeaeformis, where the biseriate dorsal striae continuing onto the mantle are interrupted by more or less developed dorsal ridge, the mantle striae in H. kenderoviana are often reduced to only a single enlarged areola after the narrow dorsal ridge. Moreover, the biseriate striae in H. coffeaeformis are composed of very small (fine) and closely positioned areolae (see for instance figs 102, 105 and 108 in Archibald andSchoeman 1984, andPl. 46 in Stepanek andKociolek 2018), compared to the relatively large and distantly spaced areolae in the dorsal striae of H. kenderoviana (Fig. 2J). In H. coffeaeformis the two rows of areolae in each stria are also positioned very close together (e.g. figs 108, 116, 121, 142, 155, etc., in Archibald andSchoeman 1984, andPl. 46, figs 6, 7 in Stepanek andKociolek 2018), leaving wide virgae between the striae, whereas the two rows of areolae in the biseriate striae in H. kenderoviana are clearly widely spaced, leaving narrower virgae between the striae (e.g. Fig. 2J, M). Halamphora coffeaeformis is also a larger taxon, with a width of usually above 5 μm, and all populations, considered to be identical with the type, show a more or less arched raphe (e.g. Pl. 91, figs 1-14 in Levkov 2009 andPl. 45, figs 1-8 in Stepanek andKociolek 2018). Two other Halamphora species with biseriate striae, H. aponina (Kütz.) Levkov and H. isumiensis Stepanek et al. possess fine, rounded areolae, with biseriate striae continuing onto the mantle (Levkov 2009, Pl. 233, figs 2, 7 andKociolek 2018, Pl. 53, fig. 1, respectively). Halamphora aponina has slightly longer valves (> 23 μm) with a slightly finer striation of 20-22 striae in 10 μm, and an almost linear raphe ledge, only expanded near the apices (Levkov 2009), but not in the valve middle, whereas H. isumiensis has a much finer striation on the ventral side (ca. 36 striae in 10 μm vs. 27-28 in H. kenderoviana), and a very distinct dorsal marginal ridge (Stepanek and Kociolek 2018), not observed in H. kenderoviana. Halamphora kenderoviana also differs from all the above-mentioned species with the presence of prominently raised, though narrow internal costae. Kellogg and Kellogg (2002), who compiled a list of diatom taxa, recorded in the Antarctic by the year 2000, listed Halamphora coffeaeformis and some (earlier considered) infraspecific taxa of the latter, showing a similar valve outline and comparable valve dimensions to H. kenderoviana, such as Amphora coffeaeformis var. borealis (Kütz.) Cleve. Amphora coffeaeformis var. borealis is treated as a synonym of A. borealis Kütz. (see Kellogg and Kellogg 2002, p. 71 and references therein), a species now transferred to Halamphora as H. borealis (Kütz.) Levkov (Levkov 2009). However, H. borealis has uniseriate dorsal striae (Levkov 2009), contrary to the biseriate striae in H. kenderoviana. Other, and more recently described taxa, similar in valve outline and possessing biseriate dorsal striae (as both H. coffeaeformis and H. kenderoviana), include H. bistriata Stepanek & Kociolek (Stepanek and Kociolek 2018), H. tumida (Hustedt) Levkov (Sar et al. 2004;Levkov 2009) and H. americana Kociolek (Kociolek et al. 2014). When compared to H. kenderoviana, they all generally have a larger valve width of above 4 μm and biseriate striae in only part of the valve face dorsally.
The most similar taxa in LM in terms of valve outline and striation pattern include Halamphora nagumoi Stepanek et al. and H. banzuensis Stepanek et al. However, under SEM, H. nagumoi, described from the Pacific coasts, presents very closely positioned central raphe endings and a prominent dorsal ridge (Stepanek and Kociolek 2018, Pl. 55, fig. 1), contrary to H. kenderoviana, lacking these features. Halamphora banzuensis from the Banzu flat near Tokyo (Japan) has a much finer striation on the ventral side (40-43 vs. 27-28 striae in 10 μm in H. kenderoviana), a well-developed dorsal ridge, and striae dorsally are separated by raised virgae on the valve exterior (Stepanek and Kociolek 2018, Pl. 70, figs 3, 4), features not present in H. kenderoviana. Finally, based on valve outline and striation, Amphora cognata Cholnoky is also morphologically similar, but the latter has a slightly finer striation on the dorsal side (22-24 striae vs. 18-20 striae in 10 μm in H. kenderoviana), and a slightly larger width (5-6 μm vs. up to 4.5 μm in H. kenderoviana) (Cholnoky 1966). As SEM observations are lacking for A. cognata, it is not possible to compare the ultrastructure of both species at present. They, however, differ in ecology. Amphora cognata was described from warm springs in South Africa (Cholnoky 1966), a very different habitat, compared to the cold-water Antarctic marine coasts where H. kenderoviana was discovered, and it is rather unlikely that the two taxa are conspecific. The Argentinean Amphora (Halamphora) capitellata Freng., whose original description and drawing are provided by Sar et al. (2009), is larger (valve width 5-6 μm), and based on the drawing in Sar et al. (2009, p. 49), possesses a less convex dorsal side and a small central area on the dorsal side, contrary to H. kenderoviana.
Halamphora moncheviana, showing internally rectangular areolar openings arranged in regular rows, and externally irregular longitudinal lines on the dorsal side in LM, can hardly be confused with any other Amphora or Halamphora taxa. In valve outline, with its shortly protracted subcapitate apices, it bears only a slight resemblance to the South American brackish species Halamphora mira (Krasske) Levkov. The latter is a much larger taxon, with a length exceeding 35 μm, a width above 9 μm, and with a strongly arched raphe (Lange-Bertalot et al. 1996, as Amphora mira Krasske; Levkov 2009), contrary to H. moncheviana. The smaller H. miroides Levkov also presents a strongly arched raphe, and finely punctate striae. Moreover, it is a typically freshwater (and not marine) species, known mostly from Africa (Levkov 2009 fig. 9K and fig.11E). Amphora (Halamphora) eunotia Cleve var. striolata Freng. from Argentina has larger valves with a much coarser striation of 14-18 striae in 10 μm (Sar et al. 2009  Stepanek & Kociolek is a smaller taxon, with a width of only 2.5-4.5 μm (vs. > 5 μm in H. moncheviana), with a very dense striation near the apices (> 29 striae in 10 μm), and lacks the clearly rectangular internal areolar openings (Stepanek and Kociolek 2013, fig. 85), present in H. moncheviana. Finally, the valves shown on the original drawing of Amphora kuehniae Schoeman in the diatom collection files of the Academy of Natural Sciences, Philadelphia , show some similarity to H. moncheviana in their striation pattern; however, the valves from Lesotho, which were examined and depicted under LM by Levkov (2009), who also transferred the species to the genus Halamphora (as H. kuehniae (Schoeman) Levkov), differ from H. moncheviana with their strongly dorsiventral valves with coarse, but more rounded and not arranged in longitudinal rows areolae. Halamphora kuehniae is also a much larger taxon with an arched (and not straight) raphe (Levkov 2009).
Two other marine taxa, Amphora antarctica Hust. (Hustedt 1958) and A. barrei Manguin (Manguin 1960), were described from Antarctica. They share a similar valve outline with strongly protracted apices, overlapping dimensions and a very fine striation pattern, suggesting they might be conspecific, as noted earlier by Hargraves (1968). Both, however, clearly differ from H. moncheviana based on their dense and almost indiscernible striation pattern in LM and the presence of much more protracted apices (see also figs 5, 6 in Cremer et al. 2003 for A. antarctica), excluding all conspecificity.