Marine macroalgae of the Balleny Islands and Ross Sea

Abstract The macroalgae of the Balleny Islands (66°15′S–67°35′S and 162°30′E–165°00′E) have been infrequently collected and the flora remains poorly known. This chain of islands is located on the edge of the Antarctic Circle in the northern Ross Sea, ~250 km north of the coast of northern Victoria Land, and it represents the most northerly land in the Ross Sea region. As well as being very remote, access to these islands is difficult given the highly variable prevailing ice conditions. We summarize the macroalgal floras of the Balleny Islands and the Ross Sea, including reporting new records, extending the known distribution of other taxa and highlighting the need for further taxonomic research on some of the most common and widespread species. Many of the taxa reported have been collected on few occasions and, as a consequence, there is insufficient material available, including reproductively mature samples, for some species to be fully documented. While these collections are providing intriguing insights into the relationships between the macroalgae found around the Antarctic continent, the full biodiversity of the Balleny Islands remains to be investigated, and further collections are required to enable detailed comparisons with other parts of the Antarctic region.


Introduction
The Balleny Islands sit on the edge of the Antarctic Circle in the northern Ross Sea, ∼250 km north of the coast of northern Victoria Land (Fig. 1). Along with Scott Island, the Balleny Islands are the only oceanic islands and the most northerly land in the Ross Sea region. There are three main islands in the Balleny group (Sturge, Buckle and Young) and three smaller islands (Row, Borradaile and Sabrina), as well as numerous offshore rock stacks. These volcanic islands are orientated in a north-west to south-east direction and span over a degree of latitude (66°15'S-67°35'S), several degrees of longitude (162°30'E-165°00'E) and ∼190 km (Bradford-Grieve & Fenwick 2002). The Balleny Islands have never been inhabited and are ice bound throughout the winter. The sea between the Balleny Islands and mainland Antarctica is frequently impenetrable due to heavy ice, and, although the area generally becomes ice free from early to mid-January until April, pack ice may survive through a significant part of the summer (Bradford-Grieve & Fenwick 2002).
Early accounts documenting macroalgal collections from the Ross Sea and East Antarctica include the work of Barton (1902), Foslie (1905Foslie ( , 1907, Gepp & Gepp (1905a, 1905b, 1907, 1917, Lucas (1919), Levring (1945) and Skottsberg (1953). Macroalgal collections from the western Ross Sea were made by J.S. Zaneveld in 1963Zaneveld in -1964Zaneveld in , 1964Zaneveld in -1965Zaneveld in and 1967, including sampling undertaken on the combined New Zealand-United States Expedition to the Ross Sea, Balleny Islands and Macquarie Ridge from the United States Coast Guard icebreaker Glacier (January-March 1965). Data from these collections were subsequently published in a series of reports (Neushul 1965, Zaneveld 1966a, 1966c, 1968, Zaneveld & Sanford 1980, Wagner & Zaneveld 1988. The track of the Glacier voyage was provided in Zaneveld & Sanford (1980) and the specimens are housed in the Rijksherbarium, Leiden, The Netherlands.
Over the past ∼30 years, a range of ecological studies on macroalgae have been undertaken in the Ross Sea (e.g. Miller & Pearse 1991, Schwarz et al. 2003, Norkko et al. 2004, 2007, as well as research on the benthic algal flora in Terra Nova Bay (Cormaci et al. 1992(Cormaci et al. , 1998(Cormaci et al. , 2000. Wiencke & Clayton (2002) summarized knowledge of Antarctic seaweeds reporting 116 species from the continent, with only 30 of these recorded from the Ross Sea, and specific reference to only 2 species (Chaetomorpha mawsonii, Georgiella confluens) from the Balleny Islands. Cormaci et al. (2000) collated records of macroalgae from the Ross Sea, reporting 37 taxa, many of which have limited

299
distributions or are rarely found. Cormaci et al. (2000) observed 'a much poorer flora (only 5-10%) is found in latitudes which are higher than 70°S (such as in Terra Nova Bay, Ross Sea) to the minimum of seven taxa recorded by Miller & Pearse (1991) from McMurdo Sound'. With 17 taxa recorded from Terra Nova Bay, Cormaci et al. (2000) concluded that this 'can be considered the richest area in species both in this area and in the stations located to the North'. More recently, Oliveira et al. (2020) compiled records of macroalgae from Antarctica, South Shetland Islands, King George Island and Adelaide Island with 151 species in total (85 Rhodophyta, 34 Ochrophyta and 32 Chlorophyta), incorporating the results of a number of other reports (including Mystikou et al. 2014, Wiencke et al. 2014, Pellizzari et al. 2017. Additional new records of green and red algae have been reported by Dubrasquet et al. (2018Dubrasquet et al. ( , 2021 from the South Shetland Islands and the Antarctic Peninsula. Thus, only ∼25-30% of the total Antarctic flora has been reported from East Antarctica.
There are many unresolved taxonomic questions relating to the species recorded from the Antarctic region. Many records are based on drift or dredged specimens, a number of species are known from few collections and names of morphologically similar taxa from outside the region have been applied to Antarctic specimens. Caution therefore is required when comparing lists of taxa recorded from the region. In this account, we describe the macroalgal diversity from the Balleny Islands, Scott Island and the Ross Sea, updating records reported in Nelson et al. (2010Nelson et al. ( , 2017, and compare the diversity of macroalgal and benthic communities and habitats with reports from other parts of the region.

Materials and methods
This study is based primarily on collections lodged in the Herbarium of the Museum of New Zealand Te Papa Tongarewa (WELT; Thiers 2021). Figure 1 shows the localities that specimens were collected from and Table S1 provides full details of collection sites, as well as collection methods and depths, along with reference numbers and GenBank numbers. Samples were collected both opportunistically and as part of targeted research at the Balleny Islands and in the north-western Ross Sea between 2001 and 2006 during expeditions by RV Tangaroa, RV Italica and yacht Tiama. Additional opportunistic collections have been made available to us from the Ross Sea, including material from Franklin Island (76°04'48.0"S, 168°19'12.0"E), Cape Evans (77°38'05.7"S, 166°24'50.6"E) and the McMurdo jetty (77°50'51.5"S, 166°39'27.7"E). A number of the samples examined in this study came from depths well below the euphotic zone and using equipment that was targeting the collection of other organisms.
Macroalgal samples were either pressed in the field or preserved in either 5-10% formalin/seawater or 90% ethanol or were frozen. Some specimens were subsampled for subsequent molecular sequencing, with small portions (∼1 × 2 cm) of tissue removed and placed in silica gel. Liquid-preserved specimens were rinsed, examined for epiphytes and then pressed on herbarium paper. Specimens that had been frozen were highly problematic, and on thawing many of the samples disintegrated. Fragmentation was lessened by immersing samples in trays of 70% alcohol as they thawed, followed by rapid preparation of herbarium specimens.
Specimens were examined microscopically by either hand sectioning with a razor blade or by preparing whole mounts. Some slides were stained for several minutes in an acidified aniline blue solution (one part 1% aniline blue solution: nine parts 7% acetic acid solution) and permanently mounted in a 50% aqueous Karo mixture (with phenol crystals added to prevent microbial growth). Permanent slides were examined by light microscopy using a Zeiss AxioCam HRc camera with accompanying AxioVision software mounted on a Zeiss Axiovert microscope.
In addition to the field collections, we examined underwater images taken in 2001 and 2006 from Tangaroa and images taken in 2006 from Tiama to compare with dredge and scuba records.

Molecular sequencing data
Samples of algae were sequenced during this study, including the red algae initially identified as belonging to the genera Ballia, Gainia, Iridaea, Palmaria, Phyllophora and Phycodrys (reported in Lin & Nelson 2010), four samples of Hapalidiaceae and a sample of the green algal genus Prasiola (reported in Heesch et al. 2012). DNA was extracted following a modified cetyl trimethylammonium bromide (CTAB) protocol (Zuccarello & Lokhorst 2005) or, for coralline algae, using the Qiagen DNeasy Blood and Tissue DNA Extraction Kit (Qiagen GmbH, Hilden, Germany), using a modified version of the protocol described by Broom et al. (2008). Amplifications were performed in 25 μl polymerase chain reactions (PCRs) containing 1 μl of 1:50 diluted DNA extract, 5 μl of 5 × reaction buffer, 1.5 mmol MgCl 2 , 10 nmol of each deoxynucleotide triphosphate, 25 pmol of each primer and 0.1 U Kapa 2 G Robust DNA polymerase (Custom Science, Auckland, New Zealand). The plastid-encoded large subunit of the ribulose-1,5-bisphosphate carboxylase/ oxygenase (rbcL) gene was amplified from the six red 300 algal genera named above in two parts, using a variety of primer combinations: F145-R898 and F762-R1442 (Kim et al. 2010) or F57, F492, F753 (Freshwater & Rueness 1994) or F7 (Gavio & Fredericq 2002) combined with the reverse primers R753 or RrbcS (Freshwater & Rueness 1994), annealing at 45°C. The photosystem II thylakoid membrane protein D1 ( psbA) gene was amplified from the Hapalidiaceae specimens and from one specimen initially identified as a Phyllophoraceae sp. specimen using primers psbAF1 paired with psbAR1 (Yoon et al. 2002). Amplifications were performed in MJ Research PTC-200 (GMI, Ramsey, MN, USA), using an initial denaturation step (95°C, 3 min), followed by a 35 cycles of denaturation (95°C, 15 s), annealing (44-50°C, 20 s) and extension (72°C, 1 min), with a final extension of 2 min at 72°C. The PCR products were visualized on 1% agarose gels, purified by exonuclease I/alkaline phosphatase digestion using standard methods and commercially sequenced in both directions (Macrogen, Inc., Seoul, Korea). Sequences were assembled and aligned using Geneious Prime 2019.0.4 (https://www.geneious.com). Sequences obtained were compared with accessions in GenBank of the National Center for Biotechnology Information (NCBI) using BLAST searches (https://blast.ncbi.nlm. nih.gov/Blast.cgi; Altschul et al. 1990), and they also were compared with unpublished sequences from current research projects.
Alignments were produced using newly generated data, previous data ) and available sequences from GenBank. The rbcL and psbA datasets were aligned using MAFFT in Geneious. Maximum-likelihood analyses were implemented using IQ-TREE (Trifinopoulos et al. 2016). IQ-TREE was used to select the molecular evolution models (ModelFinder; Kalyaanamoorthy et al. 2017) under the Bayesian information criterion. Genes were partitioned by codon as appropriate. The datasets were subjected to nonparametric bootstrap analysis (500 replicates ;Felsenstein 1985) in IQ-TREE. Trees were annotated in FigTree v1.4.4 (http://tree.bio.ed.ac.uk/ software/figtree/) and Inkscape 1.1 (http://inkscape.org).

Field observations
Underwater photography at the Balleny Islands revealed macroalgae occupying a range of habitats, with dense stands of Himantothallus grandifolius and Desmarestia menziesii forming forests in the nearshore. Macroalgae were not restricted to hard substrate and shallow water (<30 m). There were also macroalgae growing on cobbles and on small patches of harder substrate amongst soft sediment environments at depths of up to ∼100+ metres.
The red algae Palmaria decipiens and Ballia sp. were the most commonly seen species on underwater video from sites at Borradaile Island, while the green algae Monostroma hariotii and Lambia antarctica were also identified along video transects. Encrusting non-geniculate coralline algae were noted on what appeared to be stable cobble beds. The large brown algae H. grandifolius and D. menziesii, the red alga Georgiella confluens and bladed red algae (potentially Notophycus fimbriatus, Halymeniaceae and Iridaea sp.) were also noted.

Species list
We report 30 taxa, 27 recorded from the Balleny Islands and 18 taxa recorded from the western Ross Sea. Fifteen of the western Ross Sea taxa are also recorded from the Balleny Islands, with three taxa recorded solely from the Ross Sea.   recognized. Ricker (1987) discussed differences between sub-Antarctic and Antarctic collections.
Type locality: Melchior Island, Antarctic Peninsula, on Leptosarca simplex, coll. I.M. Lamb, 3.ii.1945 (Ricker 1987 General comments: We consider that the genus Palmaria in the Antarctic region requires taxonomic investigation, as the sequences from the Balleny Islands and East and West Antarctica form a clade that is distinct from that of Palmaria palmata, the type of the genus (Fig. S3). Guillemin et al. (2018) found very low genetic diversity in samples from the Antarctic Peninsula, reflecting the impacts of historical perturbations on macroalgae in the region. General comments: The material from the Ross Sea and Balleny Islands is tentatively placed in this species pending the availability of further material and more detailed study. Although Gepp & Gepp (1907) described Spongoclonium orthocladum from Cape Adare, this species has not been subsequently collected and remains poorly known. Moe & Silva (1979) examined the type specimen of S. orthocladum and concluded that it is not a species of Spongoclonium but rather 'seems closely related to Antarcticothamnion'. As also noted by the Gepps, the heavy settlement of diatoms obscured the branching pattern.

306
were found to differ by > 3% rbcL base pair distance (Hommersand & Fredericq 2003), suggesting that they belonged in separate species. Recent research has shown that I. cordata from southern South America differs from the species present on the Antarctic Peninsula, which is a single genetic entity present in sites ranging 700 km across the Antarctic Peninsula and South Shetland Islands , Ocaranza-Barrera et al. 2019. Specimens collected from the Balleny Islands and Ross Sea are part of this Antarctic Iridaea species (Fig. S5) (Schwarz et al. 2003, Norkko et al. 2004, Thrush et al. 2006. Samples examined came predominantly from scuba collections from 15 to 20 m depth, and most material was fragmentary. General comments: As previously observed by Fredericq & Ramirez (1996), phylogenetic analysis of Antarctic samples of Phyllophora antarctica (now including material from the Ross Sea; Fig. S6) indicates that members of this taxon form a clade distinct from that of Phyllophora crispa (Huds.) P.S.Dixon, and further work is required to determine the correct generic placement for this taxon.
General comments: A sequence from WELT A023659 (ON211594) forms a clade with Phyllophorella. This substantially extends the known distribution of this genus: the three species of Phyllophorella described to date are all known solely from Peru (Calderon & Boo 2016) (Fig. S6).
Habitat: Subtidal, two scuba collections from 15 to 16 m depth, dredge samples from 63 to 1570 m depth.

Discussion
In this account, we report a total of 30 species, 27 from the Balleny Islands and 18 from the Ross Sea. Earlier treatments of the flora reported a total of 11 species from the Balleny Islands (Zaneveld 1968, Zaneveld & Sanford 1980, Wagner & Zaneveld 1988, and 37 species have been reported previously from the Ross Sea (Cormaci et al. 2000). The samples from the Ross Sea that are included in this report were gathered largely opportunistically yet include five new records for the region: Antarcticothamnion polysporum, Callithamnion sp., Elachista antarctica, Notophycus fimbriatus and 'Pugetia'. In order to interpret these findings and compare them with previously published accounts, it is important to consider the reliability of the taxonomy and systematic understanding of the flora. Three of the records reported here are placed provisionally within families of the red algae based on the material available (Delesseriaceae, Dumontiaceae, Halymeniaceae), and we have also recorded as yet undetermined non-geniculate coralline algae. A further 10 taxa have been provisionally placed in genera, requiring further work before their placements are confirmed and/or they can be identified to the species level. In some cases, the record in this report is based on a very small number of samples that may be incomplete or sterile, precluding more thorough documentation until further collections are available (e.g. Hymenocladiopsis sp., 'Pugetia' sp.). For the taxa reported here, five of the records are based on a single collection, and a further five records are based on two to three collections, clearly indicating that collections have not yet reached saturation in terms of flora coverage.
For some of the taxa presented here, recent taxonomic and phylogenetic investigations have been conducted, and we consider the identifications are soundly based (e.g. Phycodrys antarctica, Prasiola crispa, Tethysphytum antarcticum; Lin & Nelson 2010, Heesch et al. 2012, Sciuto et al. 2021. For other groups of taxa, the published literature is not reliable, with significant underpinning nomenclatural and systematic issues that are far from resolved (e.g. non-geniculate coralline algae). Sequence data have confirmed the need for further work on the genera Ballia and Iridaea and on the correct generic placement of species currently recorded as Phyllophora antarctica, Phyllophorella sp. and Palmaria decipiens in Antarctica.
Algal diversity has been observed previously to decrease by midway along the western Ross Sea coast in the Terra Nova Bay region (Cormaci et al. 1992), and in the southern Ross Sea the species diversity is low, with three red macroalgae dominating habitats (Iridaea sp., P. antarctica and non-geniculate corallines; Miller & Pearse 1991). The observations and collections presented here support these earlier accounts, with the five new records for the Ross Sea sampled from the Outer Ross Sea region (71-72°S), indicating that further biodiversity surveys in the wider Ross Sea are warranted. The flora of West Antarctica is much better known than that of East Antarctica.
A relatively high proportion of the macroalgal flora recorded from the Antarctic region is considered to be endemic to Antarctica: ∼35% of the total flora. This includes 44% of the Ochrophyta and 36% of the Rhodophyta (Wiencke et al. 2014). This high degree of endemism is attributed to the isolation of the continent and the 'significant tectonic, oceanographic and climatic changes' that have affected the region over the past 50 million years (discussed by Guillemin et al. 2020). Molecular sequence data and analyses are providing insights into the relationships of the floras of Antarctica, southern South America and some of the peri-Antarctic islands, including enabling the recognition of cryptic species (e.g. Hommersand et al. 2009, Billard et al. 2015, Pellizzari et al. 2017, Ocaranza-Barrera et al. 2019. The impact of glaciation has been investigated for selected macroalgae on the Antarctic Peninsula, revealing evidence of genetic bottlenecks associated with the Last Glacial Maximum ∼18,000 years ago, with low genetic diversity and evidence of recent population expansion (Guillemin et al. , 2020. Guillemin et al. (2020) speculated on the role of glacial refugia in the peri-Antarctic islands south of the Antarctic Polar Front (APF), including referencing the Balleny Islands as an area that warrants further sampling.
Macroalgae form significant beds along the coastal regions of Antarctica to depths of 100 m, contributing as both primary producers and as ecosystem engineers. The flux of materials from these extensive beds provides an important source of energy supporting a significant proportion of secondary production (Marina et al. 2018, Quartino et al. 2020. Momo et al. (2020) summarized the critical role played by macroalgae in Antarctic coastal food webs. Drift or stranded macroalgae play a key part in these food chains. To date, 39 species of macroalgae have been recorded drifting, stranding or floating in Antarctica, including some crossing the APF (Macaya et al. 2020). The ecological isolation of Antarctica is predicted to be strongly influenced by changes in the frequency and intensity of storms, with an increasing number of incursions crossing the APF predicted (Fraser et al. 2018), Fraser et al. (2020 evaluated the biogeographical processes influencing the floras of the Antarctic and sub-Antarctic, focusing on the role of buoyant macroalgal rafts and their contributions to species dispersal. There is increasing focus being given to the changes to the biota in the Antarctic region, either through the introduction of species directly by human-mediated vectors (vessel traffic, ballast water) or through the effects of changing climate on algal seasonality, depth distributions and habitat changes and potentially influencing competitive relationships between resident species, as well as range expansions and the survival of sub-Antarctic or cold temperate species, with consequences for the native Antarctic biota (e.g. Mystikou et al. 2014, Chown et al. 2015, Schoenrock et al. 2016, Valdivia 2020. Improved baseline information about the macroalgal flora is needed to fully understand changes over time. As noted by Mystikou et al. (2014), 'species numbers from limited collections alone cannot be considered as a reliable proxy to estimate changes in algal communities impacted by climate change over a time span of several decades'.
The full biodiversity of the Balleny Islands remains poorly known, and detailed comparisons with the floras of East and West Antarctica are not possible with the material currently available. Based on the number of taxa known from single or few collections, we suggest that more diversity will be discovered in this region when thorough and well-targeted collections are made. Based on collections currently available, many of the taxa cannot be fully documented as there is insufficient mature material available. In addition, the taxonomic resolution for the Antarctic flora as a whole requires further research. The continued use of names that are known not to apply to Antarctic species confuses the literature and inhibits analysis of relationships.
Despite the inadequacies in the data available for the Balleny Islands, it is clear that the macroalgal biodiversity of the Balleny Islands and the outer Ross Sea has been underestimated. The Balleny Islands occupy a unique position in relation to the Antarctic continent and there are some intriguing patterns of species distribution that warrant further investigation. Further targeted collections in this region would enable better definition of the biota and community structure and enable investigation of relationships and connections within the wider Antarctic region. survey of the western Ross Sea and Balleny Amanda Kelley (University of California) and Ross Sea samples provided by Vonda Cummings (NIWA). We thank Jenn Dalen and Antony Kusabs (Herbarium, Museum of New Zealand Te Papa Tongarewa) for their assistance with the registration of algal specimens. We thank Quantarctica and the Norwegian Polar Institute for provision of their simple base map. We also thank two anonymous reviewers for their comments.

Financial support
This work has been supported by funding from NIWA SSIF to the Coasts and Oceans National Centre, Marine Biodiversity Programme.

Author contributions
WAN processed and identified specimens, analysed data and led the writing of the manuscript. KFN processed specimens and data, contributed to the approach of the article, prepared maps and figures and prepared and edited the manuscript prior to submission. RD contributed to taxonomic identifications, performed DNA extraction, amplification, molecular analyses and data interpretation and contributed to writing the manuscript. JES contributed to sequencing and initial analyses and commented on the text.

Supplemental material
Seven supplemental figures and one supplemental table will be found at https://doi.org/10.1017/S0954102022000220.