Hotspots in the cold: Continent-wide meta-analysis of Antarctic genetic data reveals diverse evolutionary histories and highlights critical knowledge gaps

This is the latest contribution in what has been a rapidly moving field in the last 1-2 decades. It is the first such cross-taxon synthesis specifically of molecular data that I am aware of, although the underlying concepts the paper recognises (long-term isolation, limited connectivity) have received considerable attention and recognition in both taxon-specific research and more general syntheses involving multiple researchers in this period (including and not limited to E. Biersma, A. Carapelli, G. Collins, P. Convey, C. Fraser, P. Greenslade, I. Hogg, N. Iakovenko, M. Stevens, E. Verleyen, W. Vyverman), so are not of themselves ‘new’ conclusions. Nevertheless this is an extremely interesting subject and a timely synthesis.

The abstract gives a good overview of the paper, though brings a couple of questions to mind. First, ‘environmental drivers’ are mentioned at l11, and conclusions drawn about them at l21, but nothing is said explicitly about where this information on drivers was obtained from – at l10 the paper ‘presents a synthesis of published genetic studies..’ and in my experience very few of these report on detailed environmental conditions? Second, at l16 the bias towards flowering plants is highlighted (52% of floral datasets), immediately followed by the geographic bias to the Antarctic Peninsula and Victoria Land. The latter most likely reflects the spatial distribution of stations, researchers and research that has been done. The former seems more surprising given that only two flowering plants are present on the continent (Antarctic Peninsula), and might imply a bias in genetic research to the sub-Antarctic islands, where there are more flowering plants. It might be useful to define the ‘area’ of study early in both the Abstract and Introduction. Can this be made clearer (accepting that the Abstract is word limited)? For instance, while the Abstract does mention sub-Antarctic connectivity with lower latitudes, all of the opening section of the Introduction refers explicitly only to the Antarctic continent and Peninsula (the latter, with the Scotia Arc archipelagos, making up the maritime Antarctic region); it is only at the very end of the Introduction that the sub-Antarctic gets its first mention. At the outset, my impression of the existing literature is that relatively few molecular phylogeographic studies have focused on or included sub-Antarctic material.

L41 – an additional Convey et al. citation that brings this statement more up to date, and covers research in wider range of taxonomic groups that has become available in that time, is Convey, P., Biersma, E.M., Casanova-Katny, A. & Maturana, C.S. 2020. Refuges of Antarctic Diversity. Chapter 10 in: Past Antarctica (eds. Oliva, M. & Ruiz-Fernández, J.), pp. 181-200. Academic Press, Burlington, doi: 10.1016/B978-0-12-817925-3.00010-0.

L54-59 – another line of evidence supporting this lies in the high numbers of ‘bipolar’ species in certain groups, particularly mosses (e.g. see Biersma et al. 2017 Roy Soc Open Sci) and various microbial groups (papers by D. Pearce, J. Kleinteich).

L74-76 – this sentence doesn’t capture the true complexity of this change in view. I am also not sure that it is strictly correct with reference to C. antarcticus – I don’t believe this species has been regarded as ‘pan-Antarctic’, and rather the name has been used to refer to material from the maritime and sub-Antarctic, not the continental Antarctic (and it is very likely that each ‘species’ found on different sub-Antarctic islands will soon be elevated to true species level, but I don’t believe that has happened yet). So, I think it is incorrect to say that C. antarcticus has been considered to be present on both sides of the Gressitt Line? The picture with F. grisea is more complex, as explained in the Carapelli et al paper referred to (and there is a more recent Stevens et al paper that should be cited too, I think in ZooKeys) – this species name has been used to refer to specimens from the sub-Antarctic (South Georgia, type locality), maritime Antarctic and continental Antarctic. The most recent molecular work now assigns all maritime Antarctic material previously known as F. grisea to F. antarctica (with F. grisea itself now limited to South Georgia alone). However, all continental Antarctic material (i.e the other side of the Gressitt Line) is referred to several new species, none of which occur in the same region within continental Antarctica – in essence they are all endemic to different (parts of) different ACBRs, and are deeply divergent from each other and F. antarctica, rather than simply being on the other side of the Gressitt Line.

L90 – this general point was also made by Chown & Convey 2007, as well as in Chown, S.L. & Convey, P. 2012. Spatial and temporal variability in terrestrial Antarctic biodiversity. In: Antarctica: An Extreme Environment in a Changing World, Chapter 1, eds. Rogers, A.D., Johnston, N.M., Murphy, E. & Clarke, A., Blackwell, Oxford, pp. 13-43.

L97-105 – there is a need to posit hypotheses, but the way these are stated here they almost read as too obvious or simplistic. It is generally well recognised in Antarctic biology that there is a close link between research station location and biological sampling, both historically and today. Likewise, classical biodiversity research already underpins the conclusion that maritime Antarctic (Ant Pen and Scotia Arc) diversity is greater in many groups and most locations than in the continental Antarctic, and that sub-Antarctic diversity is a step up again.

L149-150 – this statement seems odd – Terauds et al. 2012 provided the original definition of the ACBRs, which by definition lie within the Antarctic Treaty area south of 60 degrees latitude. They do not include the sub-Antarctic region, which lies at lower latitude outside the Antarctic Treaty area. If an erroneous definition and understanding of the sub-Antarctic is being used here, this will need carefully correcting and updating throughout the paper – I have a concern that the Antarctic Peninsula and/or Scotia Arc archipelagos (other than South Georgia) are being incorrectly defined as sub-Antarctic here, as this sentence implies. The problem of regional definitions is not helped by Figure 1, which does include markers for several correct sub-Antarctic islands, but then also lower latitude groups such as the Falklands, southern South America and New Zealand. There needs to be table or other form of clear summary that makes area definitions clear from the outset. Further, not all the symbols used in Figure 1 are explained on the figure or the caption. I note that in Table 1 the areas are defined more correctly, though would recommend inserting a line or a further subheading after Prince Edward Island, as the remaining locations (Australia etc) are not sub-Antarctic, and also moving the Falkland Islands into this latter section. With reference to the French Indian Ocean Islands, if this group includes Amsterdam Island then that is also not sub-Antarctic. For the widely accepted definitions of the continental, maritime and sub-Antarctic regions see chapter by RIL Smith in Laws (1984) Antarctic Ecology book, or the chapter by Convey in editions of the Encyclopedia of Biodiversity (most recent version 2017).

L171 – in reality this is quite a small proportion of overall diversity in Antarctica – it might give further strength to the study’s conclusion that much more sampling is required to give here some figure for the overall recognised diversity of the selected groups in Antarctica.

L220 – as noted in Abstract, where were these environmental data obtained from? There is a risk that using values of, e.g., general means of standard meteorological datasets provides such a coarse environmental description as to be of little use in linking with biodiversity data.

L260 – depending what location is referred to in Ellsworth Land, note that BAS operate Sky Blu (not a research station but a logistic hub) at 75 deg S, and the ALE Union Glacier camp is at about 79 deg S in the Ellsworth Mountains, again providing a hub of human activity, and supporting both tourism and national operator activities (Brazilian and Chilean summer station facilities are now installed close by).

L305-308 – there is further confusion in definition of areas here, with South America referred to as ‘sub-Antarctic’, as well as the Scotia Arc archipelagos (this time with the exception of the South Shetlands) also referred to as ‘sub-Antarctic’. With reference to South America, while the very southern part is justifiably referred to by some authors as the Magellanic Sub-Antarctic (basically the Cape Horn archipelago and islands south of the Beagle Channel, and Diego Ramirez), the range of locations marked on Figure 1, which I assume represent sampling locations/data sources extends far north of the Magellanic sub-Antarctic.

Para at l309 – I find the use of ‘population’ a little confusing in this para.

Figure 3 – standard journal instructions are that figures and their captions need to standalone, so the various abbreviations on this figure need explaining in the caption or in a key. It is a little unfortunate that molecular taxonomic and phylogeographic studies of both Friesea species and some Antarctic tardigrades have become available since the time cutoff used in this study. See Carapelli et al. 2020 Diversity, Stevens et al. ZooKeys, Short et al. 2022 Molecular Phylogenetics and Evolution. This also applies to the genus Sterotydeus in Victoria Land – see Brunetti et al. 2021 articles in Diversity and Taxonomy. There are also more recent phylogeographic studies of Colobanthus, see Biersma et al. 2020 J Biogeog and Alban et al. 2022 Plant Systematics and Evolution.

L360 – intuitively a positive relationship between diversity and elevation seems surprising, given most authors recognised that greatest diversity is generally found in Antarctica in areas close to the coast, and declines rapidly with altitude. I wonder if this finding could be artefactual? – there are very few locations or studies that have looked at, or even had access to, diversity across a wide altitudinal range in a given location or even region. I note the comment that interpretations in this section need to be taken cautiously. However, again, I wonder given the very small proportion of overall diversity considered in the paper, plus the inevitably coarse nature of the environmental variables used, whether trying to include these analyses at all is really a step too far.

L385-390 – I can see what the authors are saying in this paragraph about different foci of research, but these sentences do appear to summarise two very large fields of research in a very simple way and with reference to only two citations.

Para at l413 – while I agree with the thrust of this para (again with caveat about correct use of ‘sub-Antarctic’), in a way it seems to be giving a circular argument…in reality sufficient data were available for the study from the three regions within Antarctica mentioned in the para, the genetic diversity data from these regions were distinct for some groups, and hence it is argued they have long term evolutionary isolation and refugial persistence. In essence this is the same conclusion, but based on a much more limited dataset and areal coverage, as has been drawn by the multiple studies (and others) mentioned early in the Introduction. Similarly, the conclusion of plants (in particular mosses) having more northern distributions in Antarctica is also widely recognised (e.g. see Peat et al. 2007 J Biogeog), and would appear to be more directly underlain by physiological limitations of the group as a whole than an indicator of a different evolutionary history.

An earlier salient citation for Victoria Land diversity would be Adams et al. 2006 Soil Biol Biochem. Also, for the sentence a l434, I Hogg has recently published a study giving genetic support for such dispersal from refugia. However, I think the final sentence of this para is another step too far as, but its own admission, the current study simply does not have enough data and sampling locations available to support this conclusion. There is much work that remains to be done in this field, but the deep divergences reported in multiple molecular studies of various groups and from various regions increasingly support very long term persistence in most regions examined as the ‘norm’ – see various of the studies and syntheses noted above in this review.

L447 – survive what for long periods?

Section at l448 – this section, throughout, again has problems caused by incorrect use of ‘sub-Antarctic’, in particular incorrectly including the maritime Antarctic South Sandwich and South Orkney Islands.

L521-522 – this is particularly the case for lichens – the lichen flora of the South Sandwich Islands (max age 3 million years) includes multiple species that are currently regarded as being endemic to Antarctica, ie only otherwise occur at more southern latitudes. Given the age of the South Sandwich Is, it seems most plausible that these lichens will have colonised them from the south (discussed by Convey et al. 2000 J Biogeog).

Para at l529 – counter to the argument proposed in this paragraph, recent studies by G Collins, I Hogg, A Carapelli and their various colleagues in Victoria Land have provided genetic evidence for strong dispersal limitation even at very small spatial scales (along with suggestions already provided, see Collins et al. 2019 Frontiers in Ecology and Evolution, 2020 PNAS). As already noted, I am not at all convinced that this attempt at linking the genetic diversity data with these environmental variables/features is really justified. This is almost recognised by the authors at l550-555.

A section of the Supplementary Material (p179 and following) appears to be inappropriately formatted.

I appreciate the detailed and considered responses given to the original reviews, and am satisfied with those to my own comments. The revisions made to the ms I think improve its balance and flow.
One very minor comment on Fig 1 (bottom left panel) is that there appears to be inconsistency in the use of 'Is' or 'I' as abbreviation for island(s).
And one on Figure 2 caption - algae are plants as well, so is it appropriate to include them with lichens/fungi rather than plants?