Survival strategies of Antarctic vegetation during extensive glacial expansion across the Oligocene/Miocene Transition

Duncan, Bella J.; McKay, Robert; Levy, Richard; Prebble, Joseph G.; Naish, Timothy; Seki, Osamu; Kraus, Christoph; Moossen, Heiko; Ventura, G. Todd; Kulhanek, Denise K.; Bendle, James

doi:10.5194/egusphere-2024-4021

Preprints

https://doi.org/10.5194/egusphere-2024-4021

Preprints

09 Jan 2025

| 09 Jan 2025

Survival strategies of Antarctic vegetation during extensive glacial expansion across the Oligocene/Miocene Transition

Bella J. Duncan, Robert McKay, Richard Levy, Joseph G. Prebble, Timothy Naish, Osamu Seki, Christoph Kraus, Heiko Moossen, G. Todd Ventura, Denise K. Kulhanek, and James Bendle

Abstract. Antarctica’s terrestrial ecosystems are at risk from a rapidly changing climate. Investigating how Antarctica’s vascular plants responded to major climatic variations in the geological past, especially under atmospheric CO₂ values similar to modern and future projections, may provide insight into how organisms could migrate across the continent as conditions change. Here, we investigate vegetation trends across the Oligocene/Miocene Transition (OMT, ~23 Myr), one of the largest transient glaciations of the Cenozoic. Despite extensive ice sheet expansion, Antarctic vegetation survived throughout this glacial episode. We use compound specific isotope trends (δ¹³C and δ²H) of plant waxes in an Antarctic proximal sediment core from the Ross Sea (Deep Sea Drilling Project site 270) to investigate the response and survival mechanisms of Antarctic vegetation during this event. We detect the first observation of a marked negative n-alkane δ¹³C excursion over the OMT, coupled with a shift to more positive n-alkane δ²H. We interpret this as plants sacrificing water use efficiency to maintain photosynthesis and carbon uptake during increasing glacial conditions, as atmospheric CO₂ decreased and orbital configurations favoured shorter, colder growing seasons with lower light intensity. We consider further drivers of these isotopic trends to be enhanced aridity, and a shift to a stunted, low elevation vegetation. These findings establish the adaptability of ancient Antarctic vegetation under atmospheric CO₂ conditions comparable to modern, and mechanisms that allowed vegetation to keep a foothold on the continent despite prolonged hostile conditions.

Received: 18 Dec 2024 – Discussion started: 09 Jan 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 1313 KB)

Supplement (59 KB)

Download & links

Bella J. Duncan, Robert McKay, Richard Levy, Joseph G. Prebble, Timothy Naish, Osamu Seki, Christoph Kraus, Heiko Moossen, G. Todd Ventura, Denise K. Kulhanek, and James Bendle

Status: closed

RC1:
'Comment on egusphere-2024-4021', Anonymous Referee #1, 22 May 2025
In this paper, the authors evaluate changes in Antarctic vegetation during the latest Oligocene to early Miocene (~25 -20 Ma). They evaluate this by measuring the distribution and compound specific stable carbon and hydrogen isotopic composition of long-chain n-alkanes in marine sediments from DSDP Site 270. The authors document a gradual long-term negative excursion in n-alkane d13C values and positive shift in d2H values during the late Oligocene. The authors interpret this to represent: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms.
Whilst the biomarker data appears robust, the interpretation and structure of this manuscript requires improvement. My main concerns are listed below:
The discussion focuses heavily on using leaf wax carbon isotope values to reconstruct changes in carbon isotope discrimination (Δ) across the O/M boundary – but this was not introduced prior to the discussion and I was left wondering why this was important? Was there a specific hypothesis that you wanted to test? To resolve this, I suggest some restructuring is required (e.g., talk about Δ during the introduction so that the reader knows the importance…)

The manuscript explores the many potential controls on carbon isotope discrimination values and states that “…low seasonality and decreased light intensity during the OMT would contribute to a negative δ13C excursion and associated higher Δ” – however, it was unclear what evidence you had to support either low seasonality and/or decreased light intensity during the OMT. Thus, this section currently feels speculative – this is a big part of your manuscript, thus key to resolve.

The authors suggest a “marked negative carbon isotope excursion … for all chain lengths across the OMT” but the data in Figure 3 and 4 does not appear to show a “marked” decrease at the OMT (i.e. 23 Ma). There is clearly a long-term shift towards more 13C-depleted values from ~25-23 Ma but the OMT data looks pretty similar to values at 24 Ma and 20.5 Ma.

There was no discussion on how the age model was constructed and the methods would benefit from discussion this further.

Other comments:
The Methods (L56) is referred to as section 4 (and lipid biomarkers as 4.1) – but assume this meant to be section 2?
L62: after 3:1 add (v/v)
L63: should be total lipid extract, not total solvent extract
L69: no need to capitalise Mass
After line 86, equation for stable isotopes states dD not d2H
L113: I would argue that a CPI >1-2 is not a reasonable threshold for modern plant material – most modern plants typically have CPI values between 3 to 30 – anything below 2 is likely affected by diagenesis. I suggest looking at the supplements in Bush and McInernery and re-assessing this in modern plants (e.g, calculate the CPI for 90% of modern plant data) …
L114: very rare to get CPI < 1 except in hypersaline environments
L137: not sure you can report to 2 SFs (i.e. -38.13)
L137: when you say “most values”, define how many
L137: “One sample each for n-C25, n-C27 and n-C29, and three samples for n-C31 were excluded as outliers for those specific chain lengths, as they were different from adjacent samples by more than 2 ‰” – I don’t understand the rationale for excluding this data. Can you explain why you excluded this?
Figure 3: suggest showing individual data points on each black line as this helps show data resolution
Discussion 3.1 – I wasn’t totally sure what the aim of this section was – perhaps better to start by discussing you ACL data and THEN comparing this to pollen data. Or might be better to fold in the palynology/vegetation discussion into d2H interpretation
L169: How does a CPI value of 2.4 argue against reworking?
L298: “plant wax δ2H values trend from more negative in the late Oligocene to more positive over the OMT” – please state the values (i.e. from -xxx to -xxx per mil)
L334-338: “We attribute n-alkane δ13C and δ2H trends to several environmental changes that affected Antarctic vegetation during the OMT (Fig. 5.). These are: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms” - what is evidence for shorter colder, more arid growing seasons with less light, a decrease in catchment area, and more shrub like growth forms?
Citation: https://doi.org/10.5194/egusphere-2024-4021-RC1
- AC1:
  'Reply on RC1', Bella Duncan, 30 Jul 2026
  In this paper, the authors evaluate changes in Antarctic vegetation during the latest Oligocene to early Miocene (~25 -20 Ma). They evaluate this by measuring the distribution and compound specific stable carbon and hydrogen isotopic composition of long-chain n-alkanes in marine sediments from DSDP Site 270. The authors document a gradual long-term negative excursion in n-alkane d13C values and positive shift in d2H values during the late Oligocene. The authors interpret this to represent: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms.
  We thank the reviewer for their thoughtful review.
  Whilst the biomarker data appears robust, the interpretation and structure of this manuscript requires improvement. My main concerns are listed below:
  The discussion focuses heavily on using leaf wax carbon isotope values to reconstruct changes in carbon isotope discrimination (Δ) across the O/M boundary – but this was not introduced prior to the discussion and I was left wondering why this was important? Was there a specific hypothesis that you wanted to test? To resolve this, I suggest some restructuring is required (e.g., talk about Δ during the introduction so that the reader knows the importance…)
  
  Thank you for this comment. We agree, and it was a point also noted by Reviewer 2. We have undertaken substantial restructuring of the manuscript to address this concern. Most notably, we have shifted text describing the background context of carbon and hydrogen isotopes from the discussion to the introduction, and have also introduced the concept of carbon isotope discrimination here too (see revised lines 60-82):
  The manuscript explores the many potential controls on carbon isotope discrimination values and states that “…low seasonality and decreased light intensity during the OMT would contribute to a negative δ13C excursion and associated higher Δ” – however, it was unclear what evidence you had to support either low seasonality and/or decreased light intensity during the OMT. Thus, this section currently feels speculative – this is a big part of your manuscript, thus key to resolve.
  
  Thank you for your suggestion that we need to clarify this. The evidence for this comes from intervals of the late Oligocene and OMT where orbital parameters of low eccentricity and low obliquity occur concurrently. Low seasonality driven by this orbital configuration has long been implicated as a primary cause of the OMT ice sheet advance (Zachos et al., 2001). Low obliquity and eccentricity are associated with reduced seasonality, where the contrast between the seasons becomes less pronounced due to the Earth’s axial tilt being smaller and the Earths orbit being more circular. This manifests as more mild winters but also cooler summers, as well are reducing the intensity of light that reaches the high latitudes throughout the year. To highlight we are working off a long-established framework and orbital theory for identifying low seasonality at this time we have adjusted text in the discussion to better reflect this:
  ‘The length of the summer growing window for ancient Antarctic plants would have been dependent on regional temperature. The late Oligocene saw regional cooling in Antarctica, starting from ~24.5 Ma, and inferred to be driven by decreasing atmospheric pCO₂ concentrations and reduced radiative forcing due to a combination of orbital parameters (low eccentricity and obliquity) that favour low seasonality (Fig. 4; Zachos et al., 2001; Pälike et al., 2006; Duncan et al., 2022). Low obliquity and eccentricity are associated with reduced seasonality due to a smaller axial tilt and more circular orbit under these conditions. This manifests as more mild winters but also cooler summers, as well as reducing the intensity of light that reaches the high latitudes throughout the year. This timing coincides with a shift to more positive carbon isotope discrimination in DSDP 270. The Mi-1 glaciation at 23.01 Ma is associated with a further decline in atmospheric pCO₂ concentrations to ~265 ppm (Greenop et al., 2019, CenCO2PIP, 2023), and another period of extended cool summers due to low obliquity and eccentricity (Fig. 4; Pälike et al., 2006). We therefore suggest that the trend to more elevated carbon isotope discrimination initiates at 24.4 Ma due to increasingly constrained summer growing seasons and continues against the backdrop of a cooling regional climate through the late Oligocene before culminating with the extensive Mi-1 glaciation (Fig. 5).’
  The authors suggest a “marked negative carbon isotope excursion … for all chain lengths across the OMT” but the data in Figure 3 and 4 does not appear to show a “marked” decrease at the OMT (i.e. 23 Ma). There is clearly a long-term shift towards more 13C-depleted values from ~25-23 Ma but the OMT data looks pretty similar to values at 24 Ma and 20.5 Ma.
  
  We agree that some of the terminology we have used to describe the timing of change throughout the manuscript is sometimes confusing. Similar to the comment above, we have made sure to adjust our descriptions to clarify that the compound-specific isotope trends initiate at ~24.5 Ma. For example, we have changed the text you describe above to:
  ‘A marked negative carbon isotope excursion is apparent for all chain lengths across the late Oligocene and OMT, initiating at ~24.5 Ma and culminating during the Mi-1 glaciation at 23 Ma, with the magnitude of the excursion varying between -2.6 ‰ for n-C25, -1.5 ‰ for n-C27, -3.5 ‰ for n-C29 and -5.0 ‰ for n-C31 (Fig. 3).’
  We also more clearly acknowledge that the early Miocene data does not return to values comparable to before 24.5 Ma, as follows:
  ‘In early Miocene sediments (~20.5-20 Ma), carbon isotope discrimination remains elevated compared to values from before 24.5 Ma, suggesting that the OMT potentially represents a step-change in Antarctic vegetation as plants adjust to more enduring cooler conditions under lower atmospheric CO₂ concentrations in the early Miocene (CenCO2PIP, 2023).’
  There was no discussion on how the age model was constructed and the methods would benefit from discussion this further.
  
  The age model for DSDP 270 was presented and discussed in detail in a previous publication (Kulhanek et al., 2019). To make this clearer in the text we have added the following section to the start of the methods:
  ‘2.1. Samples and age model
  Samples from DSDP 270 were obtained from the IODP Gulf Coast Repository at Texas A&M University and from archive samples held at GNS Science, New Zealand. Samples (n=30) were collected every 5-25 m down the core, depending on availability, sample quality and core recovery. We use the preferred age model of Kulhanek et al. (2019), who revised the chronostratigraphy of DSDP 270 using biostratigraphy, magnetostratigraphy, Sr-isotope stratigraphy, and K-Ar dating of glauconite. The late Oligocene and OMT (25.4-22.9 Ma) is represented by an extended package of sediments between 352 and 121-112 mbsf, with short-lived unconformities at 246, 198 and 149-146 mbsf. A longer unconformity occurs at 121-112 mbsf with sediments overlying this to 20 mbsf representing the early Miocene (20.6-19.7 Ma).’
  Other comments:
  The Methods (L56) is referred to as section 4 (and lipid biomarkers as 4.1) – but assume this meant to be section 2?
  Section numbering has been corrected.
  L62: after 3:1 add (v/v)
  Corrected
  L63: should be total lipid extract, not total solvent extract
  Corrected
  L69: no need to capitalise Mass
  Corrected
  After line 86, equation for stable isotopes states dD not d2H
  Corrected
  L113: I would argue that a CPI >1-2 is not a reasonable threshold for modern plant material – most modern plants typically have CPI values between 3 to 30 – anything below 2 is likely affected by diagenesis. I suggest looking at the supplements in Bush and McInernery and re-assessing this in modern plants (e.g, calculate the CPI for 90% of modern plant data) …
  Thank you for this comment. We have reassessed the data and interpretations presented in Bush and McInerney (2013), and note that they find 81.2% of the 1722 species they survey have a CPI of greater than 2. While they comment that a CPI of >1 might be a reasonable threshold to determine a relatively unmodified terrestrial plant source for sedimentary n-alkanes, they suggest a more rigorous threshold would be >2. As such we have adjusted our text to reflect this and shifted some text previously in the discussion, which describes a lack of reworked pollen taxa in the core, to here:
  ‘A survey of modern leaf wax material demonstrates that a CPI of >2 is a reasonable threshold value indicative of relatively unmodified terrestrial plant material (Bush and McInerney, 2013), while sediments containing CPI values of <1 usually indicate that the sediment has been exposed to elevated burial temperatures which have modified the n-alkane distribution, or there has been an input of organic matter that has been altered by diagenetic or catagenetic processes (Bray and Evans, 1961). CPI values for n-C₂₅ to n-C₃₁ range from 2.4 to 4.4 (avg. = 3.3), indicating they are predominantly sourced from relatively unmodified terrestrial plants (Fig. 2). This is supported by previously published palynological assemblages from DSDP 270 which contain minimal reworked taxa (Kulhanek et al., 2019).’
  L114: very rare to get CPI < 1 except in hypersaline environments
  Yes we agree that CPI <1 is rarely associated with unmodified plant material, please see our previous response.
  L137: not sure you can report to 2 SFs (i.e. -38.13)
  Thank for noticing this, we have adjusted it to 1 SF inline with the rest of the discussed δ¹³C values.
  L137: when you say “most values”, define how many
  This has been adjusted to define this as “all but 4 values”.
  L137: “One sample each for n-C25, n-C27 and n-C29, and three samples for n-C31 were excluded as outliers for those specific chain lengths, as they were different from adjacent samples by more than 2 ‰” – I don’t understand the rationale for excluding this data. Can you explain why you excluded this?
  Thank you for highlighting that we have not sufficiently explained our reasoning here. The three n-C₃₁ data points are all significantly more negative than those in adjacent samples, by 5, 7 and 8‰. After reassessing these samples we note that the n-C₃₁ alkane has partial coelution with the C₂₉ αβ hopane, and in these three cases we believe that the measured value may instead be reflecting the C₂₉ αβ hopane, as these can display more depleted δ¹³C values than n-alkanes (Inglis et al., 2019). The three values we removed for n-C₂₅, n-C₂₇ and n-C₂₉ all come from the same sample, which had a low recovery of n-alkanes compared to other samples, and was the only sample analysed for compound-specific isotopes in a separate batch. After reassessing this, we have decided that the δ¹³C values in this sample are likely less reliable and have removed the n-C₃₁ value from our data set too. We have added some text to explain our rationale here:
  ‘A single sample (at 194.8 mbsf) was removed from the data set as it was considered less reliable due to differences of more than 2 ‰ compared to adjacent samples (more negative for n-C₂₅ and more positive for n-C₂₇ and n-C₂₉). This sample was associated with low extract recovery and was also the only sample analysed for δ¹³C in a later batch than the rest of the samples. The n-C₃₁ values of three samples were 5, 7 and 8 ‰ more negative than adjacent samples. We noted this compound partially coeluted with C₂₉ αβ hopane, which can display more negative δ¹³C values than n-alkanes when sourced from methanotrophic bacteria (Inglis et al., 2019). We therefore suspect that in these cases partial hopane coelution biased the n-C₃₁ values and we have removed them from the data set.’
  Figure 3: suggest showing individual data points on each black line as this helps show data resolution
  This has been changed to add data points to the black lines for ACL and CPI.
  Discussion 3.1 – I wasn’t totally sure what the aim of this section was – perhaps better to start by discussing you ACL data and THEN comparing this to pollen data. Or might be better to fold in the palynology/vegetation discussion into d2H interpretation
  Thank you for your suggestion, we’ve made some alterations to this paragraph including adding an introductory sentence to describe why we consider this data before discussing the plant wax isotope trends.
  ‘To set our plant wax-based compound-specific isotope trends in context, it is first important to consider evidence for changes in the presence and diversity of vegetation over the OMT in DSDP 270. Previously published palynological analysis on DSDP 270 indicates Antarctic vegetation in the sediment core catchment was represented by a low diversity, shrubby tundra dominated by Nothofagidites, podocarps and bryophytes including mosses (Kulhanek et al., 2019). Pollen assemblages are sparse with low pollen counts, but available data indicate only minor changes in vegetation diversity throughout the record, with a relative increase in spores sourced from mosses and other bryophytes during the Mi-1 event compared to the upper Oligocene (Fig. 4; Kulhanek et al., 2019). Average n-alkane chain length values (Eq. 2) show limited variation (between 27.6 and 28), with lowest values occurring in Mi-1 and the early Miocene, likely driven by a combination of climate (i.e. cooler temperatures) and minor vegetation changes that pollen assemblages suggest occurred during this interval. In summary, palynological evidence and n-alkane average chain length data suggests limited changes in vegetation diversity throughout the record.’
  L169: How does a CPI value of 2.4 argue against reworking?
  Please see earlier comment.
  L298: “plant wax δ2H values trend from more negative in the late Oligocene to more positive over the OMT” – please state the values (i.e. from -xxx to -xxx per mil)
  This sentence has been changed to ‘In DSDP 270 δ²H values trend from more negative in the late Oligocene to more positive in the lead up to and over the OMT (δ²H values for n-C₂₅ to n-C₃₁ of ~-188 ‰ at 25.2 Ma moving to -178 ‰ at 24.4 Ma and -183 ‰ at 23.0 Ma)’.
  L334-338: “We attribute n-alkane δ13C and δ2H trends to several environmental changes that affected Antarctic vegetation during the OMT (Fig. 5.). These are: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms” - what is evidence for shorter colder, more arid growing seasons with less light, a decrease in catchment area, and more shrub like growth forms?
  We have adjusted the wording in the conclusion paragraph to briefly summarise the evidence we use to infer these interpretations, with the details discussed in greater depth in the discussion:
  ‘We attribute the trend to increased carbon isotope discrimination and more positive plant wax δ²H to several environmental changes during the OMT which impacted mass dependent isotope abundances during plant growth (Fig. 5.). These are: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, as evidenced by cooling regional temperatures, declining atmospheric CO₂ and periods of orbital configurations which favour low seasonality with cooler summers (Pälike et al., 2006, Duncan et al., 2022, CenCO2PIP, 2023); 2) we infer a cooling climate and ice sheet growth also likely resulted in a decrease in catchment area and reduction in growing region (both space and elevation), and 3) a shift to more shrub-like growth forms, which we infer based on fossilised vegetation from younger Miocene and Pliocene Antarctic outcrops (Francis and Hill, 1996; Ashworth and Cantrill, 2004). We suggest the impact of growing season length and conditions on the isotopic trends demonstrated during the late Oligocene and OMT likely indicates Antarctic plants prioritised continued photosynthesis and uptake of carbon over water use efficiency during short growing seasons.’
  
  References
  Inglis, G. N., Naafs, B. D. A., Zheng, Y., Schellekens, J., & Pancost, R. D. (2019). δ¹³C values of bacterial hopanoids and leaf waxes as tracers for methanotrophy in peatlands. Geochimica et Cosmochimica Acta, 260, 244-256.
  Zachos, J. C., Shackleton, N. J., Revenaugh, J. S., Pälike, H., & Flower, B. P. (2001). Climate response to orbital forcing across the Oligocene-Miocene boundary. Science, 292(5515), 274-278.
  
  Citation: https://doi.org/10.5194/egusphere-2024-4021-AC1
EC1: 'Comment on egusphere-2024-4021', Yves Godderis, 24 Jun 2025

Given that I haven't been able to find a second reviewer, and I've let time slip by, I've personally revised this submission, despite not being a specialist in these matters.
The article is very technical. I don't have any specific comments. It describes a significant dataset.
I think it's missing a section containing a real confrontation with climate change, even if speculative. But that's just a suggestion.
I recommend the authors respond to all of reviewer 1's questions.

Citation: https://doi.org/10.5194/egusphere-2024-4021-EC1
RC2:
'Comment on egusphere-2024-4021', Irina Rogozhina, 18 Jun 2026

Survival strategies of Antarctic vegetation during extensive glacial expansion across the Oligocene/Miocene Transition by Duncan et al., in review by CP
Review by Irina Rogozhina
First of all (and on behalf of co-editors in chief of Climate of the Past, CP), I would like to apologize for the long duration of the review and open discussion stages. The handling editor could not find a second reviewer, even though it was a pre-condition for publishing with EGU journals. Although the research methods used in this study are well-established, they don´t seem to attract enthusiastic reviewers. The editor contacted 14 potential referees, and only one of them, in addition to agreeing to reviewing the manuscript, had actually submitted their review. Then I (as a co-editor in chief) agreed to provide a second review, even though it did require quite some homework from my side, since I am on the numerical modeling side of things. However, in the meantime I got sick and ended up in hospital. With this first paragraph I conclude my apology and come on to the review itself.
In general, I have a mixed impression from reading and thinking about this manuscript – I believe that both the writing style and structure of chapters/analyses should improve and even more so the interpretation of the results in terms of scientific findings. Messages should be bold and clear, with the appropriate discussion of the uncertainties.
Writing:
The manuscript is written in such a technical way that anyone who is not directly involved in the analyses in question would grind their teeth trying to extract the most important scientific findings buried under all the equations, formulas and symbols. There are many numbered equations, but these numbers are never mentioned in the text. Also, the manuscript introduces an overwhelming number of abbreviations and notations that may be familiar to experts but for a reader with a more general background, such as mine, it is difficult to follow through more technical sections such as methods and especially results, but also parts of the discussion. While it is rather common for the methods section, it is much less so for the other two above-mentioned sections. Look at for example, section 3.1 - in one paragraph there are 3 acronyms without reminding what they mean. In the texts that are so abundant in special symbols and equations I would avoid introducing even more acronyms.
From my point of view, when submitting a biology-heavy manuscript to a journal like CP, one needs to make sure that it is written in a relatable and easy-to-grasp manner so that climate modelers, paleoglaciologists, geomorphologists, and other interested parties can understand and use these results. The subject and the results of this manuscript are clearly within the scope of CP, but its presentation style is not so much. I encourage the authors to polish their text and also visual materials.
Then I was surprised to find quite a lot of grammatical errors and typos, since the author team includes a lot of native speakers – I even consulted with a native speaker from my team. The manuscript was probably written in a hurry, which unfortunately did not result in fast publication. Also, there are some assumptions that are made in the text about the prior knowledge of literature. For example, in lines 201 – 204 there is a hidden link to some earlier literature discussing an internal parameter from an equation published elsewhere. At least I have not found where it was shown in this manuscript.
Structure:
The structure of the discussion section is quite confusing. It feels like the authors are pulling rabbits from a magician´s hat, and each rabbit is from a different continent. There is no cohesion in the narrative of this section, while it is clearly much needed, especially in the discussion. It would be better to avoid mixing actual scientific findings with new equations, introduction of new symbols and concepts, and in some places long discussions of contradictions in earlier studies and current assumptions. Some of these belong to the methods and results (with references to the equation numbers where these symbols are explained). Other issues would be better addressed in the introduction and then referred to in the discussion section. Finally, you should design a separate sub-section in the discussion that describes and analyzes uncertainties in your findings.
In conclusion: Emphasize your scientific findings and innovations. Currently it looks like a methodological publication, scientific results are poorly represented. Explain how your data synthesis was happening.
Interpretation:
I do understand that going into the deep past is not an easy task and many analogues for the past plant types that we can use today are not ideal matches for such interpretations. However, I also spotted some confusing discussions that are going back and forth on whether a plant-state indicator should increase or decrease in response to wetter/drier or colder/warmer climate conditions and also depending on the latitudes and sun light regimes. The discussion has to become more structured and relatable for a reader with a more general background to be useful for the future scientific progress.
Could sea level drop also influence plant migration to lower elevations?
You are presenting a description of many measured factors but with a very limited interpretation of their climate significance. This might be confusing for the reader.
Specific comments:
Lines 137-138: Is there any idea about the origin of this deviation?
Lines 203-204: Has this been demonstrated or only assumed?
Line 206: Why temperate?
Lines 207-2010 and similar: Why not place these numbers in tables?
Line 250: Permafrost could have also developed during Mi-1.
Paragraph of line 246+: There are significant differences between plant types in the Northern and Southern Hemispheres even now. Doesn´t it have to be discussed?
Lines 263-271 (and similar): There are so many different factors that affect delta (and other indicators). How reliable is the presented interpretation?

Citation: https://doi.org/10.5194/egusphere-2024-4021-RC2
- AC2:
  'Reply on RC2', Bella Duncan, 30 Jul 2026
  First of all (and on behalf of co-editors in chief of Climate of the Past, CP), I would like to apologize for the long duration of the review and open discussion stages. The handling editor could not find a second reviewer, even though it was a pre-condition for publishing with EGU journals. Although the research methods used in this study are well-established, they don´t seem to attract enthusiastic reviewers. The editor contacted 14 potential referees, and only one of them, in addition to agreeing to reviewing the manuscript, had actually submitted their review. Then I (as a co-editor in chief) agreed to provide a second review, even though it did require quite some homework from my side, since I am on the numerical modeling side of things. However, in the meantime I got sick and ended up in hospital. With this first paragraph I conclude my apology and come on to the review itself.
  We are grateful that the reviewer has been able to provide a review and are sorry to hear that they have been unwell. Thank you for your review.
  In general, I have a mixed impression from reading and thinking about this manuscript – I believe that both the writing style and structure of chapters/analyses should improve and even more so the interpretation of the results in terms of scientific findings. Messages should be bold and clear, with the appropriate discussion of the uncertainties.
  Writing:
  The manuscript is written in such a technical way that anyone who is not directly involved in the analyses in question would grind their teeth trying to extract the most important scientific findings buried under all the equations, formulas and symbols. There are many numbered equations, but these numbers are never mentioned in the text. Also, the manuscript introduces an overwhelming number of abbreviations and notations that may be familiar to experts but for a reader with a more general background, such as mine, it is difficult to follow through more technical sections such as methods and especially results, but also parts of the discussion. While it is rather common for the methods section, it is much less so for the other two above-mentioned sections. Look at for example, section 3.1 - in one paragraph there are 3 acronyms without reminding what they mean. In the texts that are so abundant in special symbols and equations I would avoid introducing even more acronyms.
  Thank you for this feedback. We have moved the equations and associated descriptions for carbon preference index and average chain length into the Methods. We agree that this helps to streamline the Results and Discussion sections. We have also moved the text describing the calculation and trends in carbon isotope discrimination from the Discussion to the Results section, as we felt that this was best placed directly after describing the carbon isotope results rather than placing it in the Methods. We also removed one equation from the text to simplify how carbon isotope discrimination is described. This leaves 3 equations in the Methods and one in the Results, which is not excessive relative to comparable studies we cite, as well as other plant wax compound-specific isotope-based papers published in Climate of the Past. Each equation is discussed in the surrounding text, but we have adjusted some of this text to remove potential ambiguity, as well as referencing the equation numbers later in the manuscript.
  We appreciate that plant wax-based isotope proxies can be complex and are associated with a standard but large range of abbreviations and notations, and we have attempted to minimise these as much as possible. As described in later responses, we have tried to clarify the reasoning for the steps that we have taken to make the manuscript more accessible to non-experts but feel it is also important not to remove details that an expert would expect to see when reading a manuscript describing plant-wax isotope data.
  Where previously defined acronyms are used again for the first time in the Results and Discussion sections we have spelled them out in full to help remind the reader what they stand for, with the exception of the drill core name (DSDP 270), as this is referred to extensively through the text and is standard terminology for this well-known drilling program. Regarding the paragraph referred to by the reviewer, we have adjusted the text to no longer refer to CPI and HMW, and agree it is helpful to spell out ACL in full as “average chain length”.
  From my point of view, when submitting a biology-heavy manuscript to a journal like CP, one needs to make sure that it is written in a relatable and easy-to-grasp manner so that climate modelers, paleoglaciologists, geomorphologists, and other interested parties can understand and use these results. The subject and the results of this manuscript are clearly within the scope of CP, but its presentation style is not so much. I encourage the authors to polish their text and also visual materials.
  We are sorry that the reviewer has found the presentation of the manuscript difficult to relate to but agree that it is well within the scope of CP, which has an extensive record of publishing similar papers based on plant-wax isotopes. As discussed in other replies, we have endeavoured to improve the readability of the manuscript, while ensuring that the details that an expert would expect to find in a manuscript that described plant wax isotope data remain included.
  Then I was surprised to find quite a lot of grammatical errors and typos, since the author team includes a lot of native speakers – I even consulted with a native speaker from my team. The manuscript was probably written in a hurry, which unfortunately did not result in fast publication.
  We can assure the reviewer that their assumption that the manuscript was written in a hurry is incorrect and was instead the culmination of several years of work, which we concede can sometimes make it hard to ‘see the wood for the trees’ in the final submission phase. No examples of errors or typos have been provided, and so we have thoroughly assessed the manuscript using both manual proofreading as well as double-checking using an AI (Claude) prompt to “identify and list grammatical and typo errors”. This process identified five typos as listed below. We have endeavoured to re-read the manuscript carefully to ensure the language is as clear as possible.
  Line 66: ‘an flame ionisation detector’ changed to ‘a flame ionisation detector.
  Line 79: ‘analysed with the temperature was held’ corrected to ‘analysed with the temperature held’.
  Line 186: ‘overwhelming comprise C₃ vegetation’ corrected to ‘overwhelmingly comprise C₃ vegetation’.
  Line 203: ‘derived from for benthic foraminifera’ corrected to ‘based on benthic foraminifera’.
  Line 254: ‘the Antarctica vegetation’ corrected to ‘the Antarctic vegetation’.
  Several instances where the hyphen was missing from ‘compound-specific isotopes’ have also been corrected. We have also made adjustments throughout the text to improve general readability, as well as the more detailed changes described in other comments.
  Also, there are some assumptions that are made in the text about the prior knowledge of literature. For example, in lines 201 – 204 there is a hidden link to some earlier literature discussing an internal parameter from an equation published elsewhere. At least I have not found where it was shown in this manuscript.
  We assume that the internal parameter is referring to the use of the term ‘δ¹⁸O_sw’, and we apologise that this was not clear. We have adjusted these lines to now read as follows:
  ‘To calculate Equation 4, we derive δ¹³C_atm using values published in Tipple et al. (2010) based on benthic foraminifera. We use δ¹³C_atm calculated from benthic foraminifera as they are not affected by differences in production depth, the impact of photosymbionts or seasonal variability (Tipple et al., 2010). A source of uncertainty in using these values derives from the method by which they were calculated in Tipple et al. (2010), which requires knowledge of the oxygen isotopic value of seawater (δ¹⁸O_sw) in the past, a parameter closely linked to global ice volume (Chappell and Shackleton, 1986). Tipple et al. (2012) apply the same δ¹⁸O_sw value (-0.5 ‰) over the late Oligocene and early Miocene during the calculation of δ¹³C_atm, despite this value varying through glacial and interglacial cycles in response to ice volume changes (Chappell and Shackleton, 1986). While this could affect the reliability of calculated δ¹³C_atm, Tipple et al. (2010) investigate the influence of changing this parameter on calculated δ¹³C_atm and find the impact to be only minor’
  Structure:
  The structure of the discussion section is quite confusing. It feels like the authors are pulling rabbits from a magician´s hat, and each rabbit is from a different continent. There is no cohesion in the narrative of this section, while it is clearly much needed, especially in the discussion. It would be better to avoid mixing actual scientific findings with new equations, introduction of new symbols and concepts, and in some places long discussions of contradictions in earlier studies and current assumptions. Some of these belong to the methods and results (with references to the equation numbers where these symbols are explained). Other issues would be better addressed in the introduction and then referred to in the discussion section. Finally, you should design a separate sub-section in the discussion that describes and analyzes uncertainties in your findings.
  In conclusion: Emphasize your scientific findings and innovations. Currently it looks like a methodological publication, scientific results are poorly represented. Explain how your data synthesis was happening.
  We thank the reviewer for their feedback. This goes back to our comment earlier about sometimes struggling to “see the wood for the trees” after a significant period of time working on a manuscript. We agree with the reviewer’s feedback and have made a range of adjustments to reflect their suggestions, and believe it has much improved the flow of the paper:
  We have shifted and streamlined text describing the background context of carbon and hydrogen isotopes from the Discussion to the Introduction and have also introduced the concept of carbon isotope discrimination here too.
  
  We have moved the text and equations describing the calculation of carbon preference index and average chain length from the Results to the Methods.
  
  We have shifted the text describing the calculation of carbon isotope discrimination from the Discussion section to the Results.
  
  We have made a range of adjustments through the Discussion to simplify and focus the text on our key interpretations and findings.
  
  We investigated creating a separate subsection describing uncertainties but found that to provide sufficient context for each statement resulted in too much repetition of content described elsewhere. Instead, we have more clearly acknowledged uncertainties throughout the Results and Discussion, i.e.:
  Line 232: ‘A source of uncertainty in using these values derives from the method by which they were calculated in Tipple et al. (2010), which requires knowledge of the oxygen isotopic value of seawater (δ¹⁸O_sw) in the past, a parameter closely linked to global ice volume (Chappell and Shackleton, 1986). Tipple et al. (2012) apply the same δ¹⁸O_sw value (-0.5 ‰) over the late Oligocene and early Miocene during the calculation of δ¹³C_atm, despite this value varying through glacial and interglacial cycles in response to ice volume changes (Chappell and Shackleton, 1986). While this could affect the reliability of calculated δ¹³C_atm, Tipple et al. (2010) investigate the influence of changing this parameter on calculated δ¹³C_atm and find the impact to be only minor’
  
  Line 372: ‘We acknowledge that a key uncertainty with this approach is that Arctic vegetation consists of Northern Hemisphere species that were not present in ancient Antarctica, where vegetation was dominated by species similar to those found in modern South America, New Zealand and Tasmania (Kulhanek et al., 2019). However, the climatic conditions experienced by Arctic vegetation compare well to ancient Antarctica, and it is an approach used previously when considering growing conditions for Antarctic plant fossils (Francis and Hill, 1996).’
  
  Line 430: ‘In summary, changes to carbon isotope discrimination in both modern and ancient plants can be driven by a range of environmental variables, making it challenging to disentangle the key drivers, and resulting in a degree of inherent uncertainty in our interpretations.’
  
  Line 487: ‘As with carbon isotope discrimination, the range of environmental variables that can influence plant wax δ²H introduces some uncertainty regarding the interpretation of the key drivers. However, we observe that the same environmental conditions described here, namely colder, shorter, and drier growing seasons, combined with a shift toward lower-altitude vegetation and more stunted growth forms, can account for both the increasing carbon isotope discrimination and more positive δ²H values we see in DSDP 270, which lends greater confidence in our interpretations.’
  
  Interpretation:
  I do understand that going into the deep past is not an easy task and many analogues for the past plant types that we can use today are not ideal matches for such interpretations. However, I also spotted some confusing discussions that are going back and forth on whether a plant-state indicator should increase or decrease in response to wetter/drier or colder/warmer climate conditions and also depending on the latitudes and sun light regimes. The discussion has to become more structured and relatable for a reader with a more general background to be useful for the future scientific progress.
  We have adjusted text throughout the Discussion to summarise some of the relevant research more succinctly and avoid further confusion. For example, at lines 357 to 360, instead of going back and forth between studies that show varying responses of modern plants to changes in precipitation, we have summarised this as ‘Modern equivalents of the Nothofagidites species and Podocarpidites found in DSDP 270 also demonstrate significant variability in the response of Δ to changes in precipitation, with a range of studies finding different sites or species can show either higher or lower Δ as precipitation increases (i.e. Read and Farquhar, 1991; Read et al., 2010; Peri et al., 2012; Griener et al., 2013; Brett et al., 2014; Londoño et al., 2024)’. As discussed in other responses, we have also moved some content from the Discussion into the Introduction, Methods and Results, and streamlined and focused the remaining Discussion.
  Could sea level drop also influence plant migration to lower elevations?
  While growth of a large ice sheet usually results in a sea level fall at distal sites, localities close to the ice sheet can see an opposite signal whereby glacial isostatic adjustment can result in land subsidence proximal to a thickening and advancing ice sheet, and a relative sea level increase due to gravitational effects (Whitehouse et al., 2019). Galeotti et al. (2016) investigated this concept in a western Ross Sea drill core over major glaciation at the Eocene/Oligocene boundary and found that the sea level signal can be complex depending on the location of the drill core, and the location of thickening ice. As such, we consider that while the impact of glacial isostatic adjustment in the Ross Sea during ice advance at Mi-1 is difficult to precisely constrain, glacial isostatic adjustment theory indicates that land around the coast of Antarctica may have seen a sea level increase rather than a fall. Additionally, seismic reflection data in the Ross Sea indicates that the late Oligocene and early Miocene was also marked by an extended period of regional subsidence (De Santis et al., 1995, McKay et al., 2025), which is supported by paleodepth estimates from foraminifera in DSDP 270 which indicate gradually increasing water depth at the core site (Leckie and Webb, 1983; Kulhanek et al., 2019).
  You are presenting a description of many measured factors but with a very limited interpretation of their climate significance. This might be confusing for the reader.
  As described in previous responses, we have made an array of changes to streamline the text and focus on our findings and their significance.
  Specific comments:
  Lines 137-138: Is there any idea about the origin of this deviation?
  Thank you for highlighting that we have not sufficiently explained our reasoning here. The three n-C₃₁ data points are all significantly more negative than those in adjacent samples, by 5, 7 and 8‰. After reassessing these samples we note that the n-C₃₁ alkane has partial coelution with the C₂₉ αβ hopane, and in these three cases we believe that the measured value may instead be reflecting the C₂₉ αβ hopane, as these can display more depleted δ¹³C values than n-alkanes (Inglis et al., 2019). The three values we removed for n-C₂₅, n-C₂₇ and n-C₂₉ all come from the same sample, which had a low recovery of n-alkanes compared to other samples, and was the only sample analysed for compound-specific isotopes in a separate batch. After reassessing this, we have decided that the δ¹³C values in this sample are likely less reliable and have removed the n-C₃₁ value from our data set too. We have added some text to explain our rationale here:
  ‘A single sample (at 194.8 mbsf) was removed from the data set as it was considered less reliable due to differences of more than 2 ‰ compared to adjacent samples (more negative for n-C₂₅ and more positive for n-C₂₇ and n-C₂₉). This sample was associated with low extract recovery and was also the only sample analysed for δ¹³C in a later batch than the rest of the samples. The n-C₃₁ values of three samples were 5, 7 and 8 ‰ more negative than adjacent samples. We noted this compound partially coeluted with C₂₉ αβ hopane, which can display more negative δ¹³C values than n-alkanes when sourced from methanotrophic bacteria (Inglis et al., 2019). We therefore suspect that in these cases partial hopane coelution biased the n-C₃₁ values and we have removed them from the data set.’
  Lines 203-204: Has this been demonstrated or only assumed?
  We appreciate that our wording here was not clear. This is a well-defined relationship and commonly used proxy, and we have now adjusted the text to reference the relationship between ice volume and δ¹⁸O_sw back to an early paper which describes it:
  ‘A source of uncertainty in using these values derives from the method by which they were calculated in Tipple et al. (2010), which requires knowledge of the oxygen isotopic value of seawater (δ¹⁸O_sw) in the past, a parameter closely linked to global ice volume (Chappell and Shackleton, 1986). Tipple et al. (2012) apply the same δ¹⁸O_sw value (-0.5 ‰) over the late Oligocene and early Miocene during the calculation of δ¹³C_atm, despite this value varying through glacial and interglacial cycles in response to ice volume changes (Chappell and Shackleton, 1986).’
  Line 206: Why temperate?
  Thank you for this comment, we understand that this may have been confusing. Available data for the isotopic fractionation between plant tissue and individual n-alkanes comes from temperate climate zones instead of high latitudes. We have reworded this sentence to reflect this more clearly:
  ‘To calculate this, we apply n-alkane specific averages from published data for a wide range of C₃ plants from temperate climate zones, as ε_lipid data in high latitude plants is limited in the literature (Collister et al., 1994; Chikaraishi and Naraoka, 2003; Diefendorf et al., 2011).’
  Lines 207-2010 and similar: Why not place these numbers in tables?
  We thank the reviewer for this suggestion but as there are not many numbers discussed (i.e. in the lines suggested there are two parameters with five values each), we prefer the current format. While this does create a list of numbers, we do not think it is excessive in this section and is a more efficient use of space. However, if the editor decides a table is better, then we are also comfortable with that.
  Line 250: Permafrost could have also developed during Mi-1.
  Yes, we agree that Antarctica likely had permafrost present during the late Oligocene and early Miocene, which leads us to consider the Arctic as an analogue for ancient Antarctica i.e. in text:
  ‘Like the modern Arctic, at latitudes south of the Antarctic Circle, vegetation growth would have been restricted to a short period during summer when there was sunlight, a thawed layer of soil and warmer temperatures for photosynthesis.’
  Paragraph of line 246+: There are significant differences between plant types in the Northern and Southern Hemispheres even now. Doesn´t it have to be discussed?
  This is a good point and we now more carefully acknowledge this uncertainty:
  ‘We acknowledge that a key uncertainty with this approach is that Arctic vegetation consists of Northern Hemisphere species that were not present in ancient Antarctica, where vegetation was dominated by species similar to those found in modern South America, New Zealand and Tasmania (Kulhanek et al., 2019). However, the climatic conditions experienced by Arctic vegetation compare well to ancient Antarctica, and it is an approach used previously when considering growing conditions for Antarctic plant fossils (Francis and Hill, 1996).’
  Lines 263-271 (and similar): There are so many different factors that affect delta (and other indicators). How reliable is the presented interpretation?
  We understand the concern here, but note that almost all paleo-proxies have multiple solutions. We believe it is best practice to explain a range of possible interpretations and then narrow them down through comparison with other proxies and logical arguments. If non-unique solutions still remain, it is valid to present these, as it helps to inform possible feedbacks and Earth system responses that lead to changes in carbon and hydrogen isotopes that could be tested by models or future data-based studies. We now more clearly acknowledge the uncertainties in plant wax δ²H interpretations; however, the methods we use and interpretations throughout the text are based on a framework that is well established in the field: ‘As with carbon isotope discrimination, the range of environmental variables that can influence plant wax δ²H introduces some uncertainty regarding the interpretation of the key drivers. However, we observe that the same environmental conditions described here, namely colder, shorter, and drier growing seasons, combined with a shift toward lower-altitude vegetation and more stunted growth forms, can account for both the increasing carbon isotope discrimination and more positive δ²H values we see in DSDP 270, which lends greater confidence in our interpretations.’
  
  References
  De Santis, L., Anderson, J. B., Brancolini, G., & Zayatz, I. (1995). Seismic record of late Oligocene through Miocene glaciation on the central and eastern continental shelf of the Ross Sea. Geology and Seismic Stratigraphy of the Antarctic Margin, 68, 235-260.
  Leckie, R. M., & Webb, P. N. (1983). Late Oligocene–early Miocene glacial record of the Ross Sea, Antarctica: Evidence from DSDP site 270. Geology, 11(10), 578-582.
  Kulhanek, D. K., Levy, R. H., Clowes, C. D., Prebble, J. G., Rodelli, D., Jovane, L., Morgans, H. E. G., Kraus, C., Zwingmann, H., Griffith, E. M., Scher, H. D., McKay, R. M. & Naish, T. R. (2019). Revised chronostratigraphy of DSDP Site 270 and late Oligocene to early Miocene paleoecology of the Ross Sea sector of Antarctica. Global and Planetary Change, 178, 46-64.
  McKay, R., Cockrell, J., Shevenell, A. E., et al. (2025). Miocene ice sheet dynamics and sediment deposition in the central Ross Sea, Antarctica. Geological Society of America Bulletin, 137(3-4), 1267-1291.
  Whitehouse, P. L., Gomez, N., King, M. A., & Wiens, D. A. (2019). Solid Earth change and the evolution of the Antarctic Ice Sheet. Nature communications, 10(1), 503.
  
  Citation: https://doi.org/10.5194/egusphere-2024-4021-AC2

Status: closed

RC1:
'Comment on egusphere-2024-4021', Anonymous Referee #1, 22 May 2025
In this paper, the authors evaluate changes in Antarctic vegetation during the latest Oligocene to early Miocene (~25 -20 Ma). They evaluate this by measuring the distribution and compound specific stable carbon and hydrogen isotopic composition of long-chain n-alkanes in marine sediments from DSDP Site 270. The authors document a gradual long-term negative excursion in n-alkane d13C values and positive shift in d2H values during the late Oligocene. The authors interpret this to represent: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms.
Whilst the biomarker data appears robust, the interpretation and structure of this manuscript requires improvement. My main concerns are listed below:
The discussion focuses heavily on using leaf wax carbon isotope values to reconstruct changes in carbon isotope discrimination (Δ) across the O/M boundary – but this was not introduced prior to the discussion and I was left wondering why this was important? Was there a specific hypothesis that you wanted to test? To resolve this, I suggest some restructuring is required (e.g., talk about Δ during the introduction so that the reader knows the importance…)

The manuscript explores the many potential controls on carbon isotope discrimination values and states that “…low seasonality and decreased light intensity during the OMT would contribute to a negative δ13C excursion and associated higher Δ” – however, it was unclear what evidence you had to support either low seasonality and/or decreased light intensity during the OMT. Thus, this section currently feels speculative – this is a big part of your manuscript, thus key to resolve.

The authors suggest a “marked negative carbon isotope excursion … for all chain lengths across the OMT” but the data in Figure 3 and 4 does not appear to show a “marked” decrease at the OMT (i.e. 23 Ma). There is clearly a long-term shift towards more 13C-depleted values from ~25-23 Ma but the OMT data looks pretty similar to values at 24 Ma and 20.5 Ma.

There was no discussion on how the age model was constructed and the methods would benefit from discussion this further.

Other comments:
The Methods (L56) is referred to as section 4 (and lipid biomarkers as 4.1) – but assume this meant to be section 2?
L62: after 3:1 add (v/v)
L63: should be total lipid extract, not total solvent extract
L69: no need to capitalise Mass
After line 86, equation for stable isotopes states dD not d2H
L113: I would argue that a CPI >1-2 is not a reasonable threshold for modern plant material – most modern plants typically have CPI values between 3 to 30 – anything below 2 is likely affected by diagenesis. I suggest looking at the supplements in Bush and McInernery and re-assessing this in modern plants (e.g, calculate the CPI for 90% of modern plant data) …
L114: very rare to get CPI < 1 except in hypersaline environments
L137: not sure you can report to 2 SFs (i.e. -38.13)
L137: when you say “most values”, define how many
L137: “One sample each for n-C25, n-C27 and n-C29, and three samples for n-C31 were excluded as outliers for those specific chain lengths, as they were different from adjacent samples by more than 2 ‰” – I don’t understand the rationale for excluding this data. Can you explain why you excluded this?
Figure 3: suggest showing individual data points on each black line as this helps show data resolution
Discussion 3.1 – I wasn’t totally sure what the aim of this section was – perhaps better to start by discussing you ACL data and THEN comparing this to pollen data. Or might be better to fold in the palynology/vegetation discussion into d2H interpretation
L169: How does a CPI value of 2.4 argue against reworking?
L298: “plant wax δ2H values trend from more negative in the late Oligocene to more positive over the OMT” – please state the values (i.e. from -xxx to -xxx per mil)
L334-338: “We attribute n-alkane δ13C and δ2H trends to several environmental changes that affected Antarctic vegetation during the OMT (Fig. 5.). These are: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms” - what is evidence for shorter colder, more arid growing seasons with less light, a decrease in catchment area, and more shrub like growth forms?
Citation: https://doi.org/10.5194/egusphere-2024-4021-RC1
- AC1:
  'Reply on RC1', Bella Duncan, 30 Jul 2026
  In this paper, the authors evaluate changes in Antarctic vegetation during the latest Oligocene to early Miocene (~25 -20 Ma). They evaluate this by measuring the distribution and compound specific stable carbon and hydrogen isotopic composition of long-chain n-alkanes in marine sediments from DSDP Site 270. The authors document a gradual long-term negative excursion in n-alkane d13C values and positive shift in d2H values during the late Oligocene. The authors interpret this to represent: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms.
  We thank the reviewer for their thoughtful review.
  Whilst the biomarker data appears robust, the interpretation and structure of this manuscript requires improvement. My main concerns are listed below:
  The discussion focuses heavily on using leaf wax carbon isotope values to reconstruct changes in carbon isotope discrimination (Δ) across the O/M boundary – but this was not introduced prior to the discussion and I was left wondering why this was important? Was there a specific hypothesis that you wanted to test? To resolve this, I suggest some restructuring is required (e.g., talk about Δ during the introduction so that the reader knows the importance…)
  
  Thank you for this comment. We agree, and it was a point also noted by Reviewer 2. We have undertaken substantial restructuring of the manuscript to address this concern. Most notably, we have shifted text describing the background context of carbon and hydrogen isotopes from the discussion to the introduction, and have also introduced the concept of carbon isotope discrimination here too (see revised lines 60-82):
  The manuscript explores the many potential controls on carbon isotope discrimination values and states that “…low seasonality and decreased light intensity during the OMT would contribute to a negative δ13C excursion and associated higher Δ” – however, it was unclear what evidence you had to support either low seasonality and/or decreased light intensity during the OMT. Thus, this section currently feels speculative – this is a big part of your manuscript, thus key to resolve.
  
  Thank you for your suggestion that we need to clarify this. The evidence for this comes from intervals of the late Oligocene and OMT where orbital parameters of low eccentricity and low obliquity occur concurrently. Low seasonality driven by this orbital configuration has long been implicated as a primary cause of the OMT ice sheet advance (Zachos et al., 2001). Low obliquity and eccentricity are associated with reduced seasonality, where the contrast between the seasons becomes less pronounced due to the Earth’s axial tilt being smaller and the Earths orbit being more circular. This manifests as more mild winters but also cooler summers, as well are reducing the intensity of light that reaches the high latitudes throughout the year. To highlight we are working off a long-established framework and orbital theory for identifying low seasonality at this time we have adjusted text in the discussion to better reflect this:
  ‘The length of the summer growing window for ancient Antarctic plants would have been dependent on regional temperature. The late Oligocene saw regional cooling in Antarctica, starting from ~24.5 Ma, and inferred to be driven by decreasing atmospheric pCO₂ concentrations and reduced radiative forcing due to a combination of orbital parameters (low eccentricity and obliquity) that favour low seasonality (Fig. 4; Zachos et al., 2001; Pälike et al., 2006; Duncan et al., 2022). Low obliquity and eccentricity are associated with reduced seasonality due to a smaller axial tilt and more circular orbit under these conditions. This manifests as more mild winters but also cooler summers, as well as reducing the intensity of light that reaches the high latitudes throughout the year. This timing coincides with a shift to more positive carbon isotope discrimination in DSDP 270. The Mi-1 glaciation at 23.01 Ma is associated with a further decline in atmospheric pCO₂ concentrations to ~265 ppm (Greenop et al., 2019, CenCO2PIP, 2023), and another period of extended cool summers due to low obliquity and eccentricity (Fig. 4; Pälike et al., 2006). We therefore suggest that the trend to more elevated carbon isotope discrimination initiates at 24.4 Ma due to increasingly constrained summer growing seasons and continues against the backdrop of a cooling regional climate through the late Oligocene before culminating with the extensive Mi-1 glaciation (Fig. 5).’
  The authors suggest a “marked negative carbon isotope excursion … for all chain lengths across the OMT” but the data in Figure 3 and 4 does not appear to show a “marked” decrease at the OMT (i.e. 23 Ma). There is clearly a long-term shift towards more 13C-depleted values from ~25-23 Ma but the OMT data looks pretty similar to values at 24 Ma and 20.5 Ma.
  
  We agree that some of the terminology we have used to describe the timing of change throughout the manuscript is sometimes confusing. Similar to the comment above, we have made sure to adjust our descriptions to clarify that the compound-specific isotope trends initiate at ~24.5 Ma. For example, we have changed the text you describe above to:
  ‘A marked negative carbon isotope excursion is apparent for all chain lengths across the late Oligocene and OMT, initiating at ~24.5 Ma and culminating during the Mi-1 glaciation at 23 Ma, with the magnitude of the excursion varying between -2.6 ‰ for n-C25, -1.5 ‰ for n-C27, -3.5 ‰ for n-C29 and -5.0 ‰ for n-C31 (Fig. 3).’
  We also more clearly acknowledge that the early Miocene data does not return to values comparable to before 24.5 Ma, as follows:
  ‘In early Miocene sediments (~20.5-20 Ma), carbon isotope discrimination remains elevated compared to values from before 24.5 Ma, suggesting that the OMT potentially represents a step-change in Antarctic vegetation as plants adjust to more enduring cooler conditions under lower atmospheric CO₂ concentrations in the early Miocene (CenCO2PIP, 2023).’
  There was no discussion on how the age model was constructed and the methods would benefit from discussion this further.
  
  The age model for DSDP 270 was presented and discussed in detail in a previous publication (Kulhanek et al., 2019). To make this clearer in the text we have added the following section to the start of the methods:
  ‘2.1. Samples and age model
  Samples from DSDP 270 were obtained from the IODP Gulf Coast Repository at Texas A&M University and from archive samples held at GNS Science, New Zealand. Samples (n=30) were collected every 5-25 m down the core, depending on availability, sample quality and core recovery. We use the preferred age model of Kulhanek et al. (2019), who revised the chronostratigraphy of DSDP 270 using biostratigraphy, magnetostratigraphy, Sr-isotope stratigraphy, and K-Ar dating of glauconite. The late Oligocene and OMT (25.4-22.9 Ma) is represented by an extended package of sediments between 352 and 121-112 mbsf, with short-lived unconformities at 246, 198 and 149-146 mbsf. A longer unconformity occurs at 121-112 mbsf with sediments overlying this to 20 mbsf representing the early Miocene (20.6-19.7 Ma).’
  Other comments:
  The Methods (L56) is referred to as section 4 (and lipid biomarkers as 4.1) – but assume this meant to be section 2?
  Section numbering has been corrected.
  L62: after 3:1 add (v/v)
  Corrected
  L63: should be total lipid extract, not total solvent extract
  Corrected
  L69: no need to capitalise Mass
  Corrected
  After line 86, equation for stable isotopes states dD not d2H
  Corrected
  L113: I would argue that a CPI >1-2 is not a reasonable threshold for modern plant material – most modern plants typically have CPI values between 3 to 30 – anything below 2 is likely affected by diagenesis. I suggest looking at the supplements in Bush and McInernery and re-assessing this in modern plants (e.g, calculate the CPI for 90% of modern plant data) …
  Thank you for this comment. We have reassessed the data and interpretations presented in Bush and McInerney (2013), and note that they find 81.2% of the 1722 species they survey have a CPI of greater than 2. While they comment that a CPI of >1 might be a reasonable threshold to determine a relatively unmodified terrestrial plant source for sedimentary n-alkanes, they suggest a more rigorous threshold would be >2. As such we have adjusted our text to reflect this and shifted some text previously in the discussion, which describes a lack of reworked pollen taxa in the core, to here:
  ‘A survey of modern leaf wax material demonstrates that a CPI of >2 is a reasonable threshold value indicative of relatively unmodified terrestrial plant material (Bush and McInerney, 2013), while sediments containing CPI values of <1 usually indicate that the sediment has been exposed to elevated burial temperatures which have modified the n-alkane distribution, or there has been an input of organic matter that has been altered by diagenetic or catagenetic processes (Bray and Evans, 1961). CPI values for n-C₂₅ to n-C₃₁ range from 2.4 to 4.4 (avg. = 3.3), indicating they are predominantly sourced from relatively unmodified terrestrial plants (Fig. 2). This is supported by previously published palynological assemblages from DSDP 270 which contain minimal reworked taxa (Kulhanek et al., 2019).’
  L114: very rare to get CPI < 1 except in hypersaline environments
  Yes we agree that CPI <1 is rarely associated with unmodified plant material, please see our previous response.
  L137: not sure you can report to 2 SFs (i.e. -38.13)
  Thank for noticing this, we have adjusted it to 1 SF inline with the rest of the discussed δ¹³C values.
  L137: when you say “most values”, define how many
  This has been adjusted to define this as “all but 4 values”.
  L137: “One sample each for n-C25, n-C27 and n-C29, and three samples for n-C31 were excluded as outliers for those specific chain lengths, as they were different from adjacent samples by more than 2 ‰” – I don’t understand the rationale for excluding this data. Can you explain why you excluded this?
  Thank you for highlighting that we have not sufficiently explained our reasoning here. The three n-C₃₁ data points are all significantly more negative than those in adjacent samples, by 5, 7 and 8‰. After reassessing these samples we note that the n-C₃₁ alkane has partial coelution with the C₂₉ αβ hopane, and in these three cases we believe that the measured value may instead be reflecting the C₂₉ αβ hopane, as these can display more depleted δ¹³C values than n-alkanes (Inglis et al., 2019). The three values we removed for n-C₂₅, n-C₂₇ and n-C₂₉ all come from the same sample, which had a low recovery of n-alkanes compared to other samples, and was the only sample analysed for compound-specific isotopes in a separate batch. After reassessing this, we have decided that the δ¹³C values in this sample are likely less reliable and have removed the n-C₃₁ value from our data set too. We have added some text to explain our rationale here:
  ‘A single sample (at 194.8 mbsf) was removed from the data set as it was considered less reliable due to differences of more than 2 ‰ compared to adjacent samples (more negative for n-C₂₅ and more positive for n-C₂₇ and n-C₂₉). This sample was associated with low extract recovery and was also the only sample analysed for δ¹³C in a later batch than the rest of the samples. The n-C₃₁ values of three samples were 5, 7 and 8 ‰ more negative than adjacent samples. We noted this compound partially coeluted with C₂₉ αβ hopane, which can display more negative δ¹³C values than n-alkanes when sourced from methanotrophic bacteria (Inglis et al., 2019). We therefore suspect that in these cases partial hopane coelution biased the n-C₃₁ values and we have removed them from the data set.’
  Figure 3: suggest showing individual data points on each black line as this helps show data resolution
  This has been changed to add data points to the black lines for ACL and CPI.
  Discussion 3.1 – I wasn’t totally sure what the aim of this section was – perhaps better to start by discussing you ACL data and THEN comparing this to pollen data. Or might be better to fold in the palynology/vegetation discussion into d2H interpretation
  Thank you for your suggestion, we’ve made some alterations to this paragraph including adding an introductory sentence to describe why we consider this data before discussing the plant wax isotope trends.
  ‘To set our plant wax-based compound-specific isotope trends in context, it is first important to consider evidence for changes in the presence and diversity of vegetation over the OMT in DSDP 270. Previously published palynological analysis on DSDP 270 indicates Antarctic vegetation in the sediment core catchment was represented by a low diversity, shrubby tundra dominated by Nothofagidites, podocarps and bryophytes including mosses (Kulhanek et al., 2019). Pollen assemblages are sparse with low pollen counts, but available data indicate only minor changes in vegetation diversity throughout the record, with a relative increase in spores sourced from mosses and other bryophytes during the Mi-1 event compared to the upper Oligocene (Fig. 4; Kulhanek et al., 2019). Average n-alkane chain length values (Eq. 2) show limited variation (between 27.6 and 28), with lowest values occurring in Mi-1 and the early Miocene, likely driven by a combination of climate (i.e. cooler temperatures) and minor vegetation changes that pollen assemblages suggest occurred during this interval. In summary, palynological evidence and n-alkane average chain length data suggests limited changes in vegetation diversity throughout the record.’
  L169: How does a CPI value of 2.4 argue against reworking?
  Please see earlier comment.
  L298: “plant wax δ2H values trend from more negative in the late Oligocene to more positive over the OMT” – please state the values (i.e. from -xxx to -xxx per mil)
  This sentence has been changed to ‘In DSDP 270 δ²H values trend from more negative in the late Oligocene to more positive in the lead up to and over the OMT (δ²H values for n-C₂₅ to n-C₃₁ of ~-188 ‰ at 25.2 Ma moving to -178 ‰ at 24.4 Ma and -183 ‰ at 23.0 Ma)’.
  L334-338: “We attribute n-alkane δ13C and δ2H trends to several environmental changes that affected Antarctic vegetation during the OMT (Fig. 5.). These are: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, 2) a decrease in catchment area and reduction in growing region (both space and elevation), 3) a shift to more shrub-like growth forms” - what is evidence for shorter colder, more arid growing seasons with less light, a decrease in catchment area, and more shrub like growth forms?
  We have adjusted the wording in the conclusion paragraph to briefly summarise the evidence we use to infer these interpretations, with the details discussed in greater depth in the discussion:
  ‘We attribute the trend to increased carbon isotope discrimination and more positive plant wax δ²H to several environmental changes during the OMT which impacted mass dependent isotope abundances during plant growth (Fig. 5.). These are: 1) a shift in climate to colder, shorter and more arid growing seasons with lower light intensity, as evidenced by cooling regional temperatures, declining atmospheric CO₂ and periods of orbital configurations which favour low seasonality with cooler summers (Pälike et al., 2006, Duncan et al., 2022, CenCO2PIP, 2023); 2) we infer a cooling climate and ice sheet growth also likely resulted in a decrease in catchment area and reduction in growing region (both space and elevation), and 3) a shift to more shrub-like growth forms, which we infer based on fossilised vegetation from younger Miocene and Pliocene Antarctic outcrops (Francis and Hill, 1996; Ashworth and Cantrill, 2004). We suggest the impact of growing season length and conditions on the isotopic trends demonstrated during the late Oligocene and OMT likely indicates Antarctic plants prioritised continued photosynthesis and uptake of carbon over water use efficiency during short growing seasons.’
  
  References
  Inglis, G. N., Naafs, B. D. A., Zheng, Y., Schellekens, J., & Pancost, R. D. (2019). δ¹³C values of bacterial hopanoids and leaf waxes as tracers for methanotrophy in peatlands. Geochimica et Cosmochimica Acta, 260, 244-256.
  Zachos, J. C., Shackleton, N. J., Revenaugh, J. S., Pälike, H., & Flower, B. P. (2001). Climate response to orbital forcing across the Oligocene-Miocene boundary. Science, 292(5515), 274-278.
  
  Citation: https://doi.org/10.5194/egusphere-2024-4021-AC1
EC1: 'Comment on egusphere-2024-4021', Yves Godderis, 24 Jun 2025

Given that I haven't been able to find a second reviewer, and I've let time slip by, I've personally revised this submission, despite not being a specialist in these matters.
The article is very technical. I don't have any specific comments. It describes a significant dataset.
I think it's missing a section containing a real confrontation with climate change, even if speculative. But that's just a suggestion.
I recommend the authors respond to all of reviewer 1's questions.

Citation: https://doi.org/10.5194/egusphere-2024-4021-EC1
RC2:
'Comment on egusphere-2024-4021', Irina Rogozhina, 18 Jun 2026

Survival strategies of Antarctic vegetation during extensive glacial expansion across the Oligocene/Miocene Transition by Duncan et al., in review by CP
Review by Irina Rogozhina
First of all (and on behalf of co-editors in chief of Climate of the Past, CP), I would like to apologize for the long duration of the review and open discussion stages. The handling editor could not find a second reviewer, even though it was a pre-condition for publishing with EGU journals. Although the research methods used in this study are well-established, they don´t seem to attract enthusiastic reviewers. The editor contacted 14 potential referees, and only one of them, in addition to agreeing to reviewing the manuscript, had actually submitted their review. Then I (as a co-editor in chief) agreed to provide a second review, even though it did require quite some homework from my side, since I am on the numerical modeling side of things. However, in the meantime I got sick and ended up in hospital. With this first paragraph I conclude my apology and come on to the review itself.
In general, I have a mixed impression from reading and thinking about this manuscript – I believe that both the writing style and structure of chapters/analyses should improve and even more so the interpretation of the results in terms of scientific findings. Messages should be bold and clear, with the appropriate discussion of the uncertainties.
Writing:
The manuscript is written in such a technical way that anyone who is not directly involved in the analyses in question would grind their teeth trying to extract the most important scientific findings buried under all the equations, formulas and symbols. There are many numbered equations, but these numbers are never mentioned in the text. Also, the manuscript introduces an overwhelming number of abbreviations and notations that may be familiar to experts but for a reader with a more general background, such as mine, it is difficult to follow through more technical sections such as methods and especially results, but also parts of the discussion. While it is rather common for the methods section, it is much less so for the other two above-mentioned sections. Look at for example, section 3.1 - in one paragraph there are 3 acronyms without reminding what they mean. In the texts that are so abundant in special symbols and equations I would avoid introducing even more acronyms.
From my point of view, when submitting a biology-heavy manuscript to a journal like CP, one needs to make sure that it is written in a relatable and easy-to-grasp manner so that climate modelers, paleoglaciologists, geomorphologists, and other interested parties can understand and use these results. The subject and the results of this manuscript are clearly within the scope of CP, but its presentation style is not so much. I encourage the authors to polish their text and also visual materials.
Then I was surprised to find quite a lot of grammatical errors and typos, since the author team includes a lot of native speakers – I even consulted with a native speaker from my team. The manuscript was probably written in a hurry, which unfortunately did not result in fast publication. Also, there are some assumptions that are made in the text about the prior knowledge of literature. For example, in lines 201 – 204 there is a hidden link to some earlier literature discussing an internal parameter from an equation published elsewhere. At least I have not found where it was shown in this manuscript.
Structure:
The structure of the discussion section is quite confusing. It feels like the authors are pulling rabbits from a magician´s hat, and each rabbit is from a different continent. There is no cohesion in the narrative of this section, while it is clearly much needed, especially in the discussion. It would be better to avoid mixing actual scientific findings with new equations, introduction of new symbols and concepts, and in some places long discussions of contradictions in earlier studies and current assumptions. Some of these belong to the methods and results (with references to the equation numbers where these symbols are explained). Other issues would be better addressed in the introduction and then referred to in the discussion section. Finally, you should design a separate sub-section in the discussion that describes and analyzes uncertainties in your findings.
In conclusion: Emphasize your scientific findings and innovations. Currently it looks like a methodological publication, scientific results are poorly represented. Explain how your data synthesis was happening.
Interpretation:
I do understand that going into the deep past is not an easy task and many analogues for the past plant types that we can use today are not ideal matches for such interpretations. However, I also spotted some confusing discussions that are going back and forth on whether a plant-state indicator should increase or decrease in response to wetter/drier or colder/warmer climate conditions and also depending on the latitudes and sun light regimes. The discussion has to become more structured and relatable for a reader with a more general background to be useful for the future scientific progress.
Could sea level drop also influence plant migration to lower elevations?
You are presenting a description of many measured factors but with a very limited interpretation of their climate significance. This might be confusing for the reader.
Specific comments:
Lines 137-138: Is there any idea about the origin of this deviation?
Lines 203-204: Has this been demonstrated or only assumed?
Line 206: Why temperate?
Lines 207-2010 and similar: Why not place these numbers in tables?
Line 250: Permafrost could have also developed during Mi-1.
Paragraph of line 246+: There are significant differences between plant types in the Northern and Southern Hemispheres even now. Doesn´t it have to be discussed?
Lines 263-271 (and similar): There are so many different factors that affect delta (and other indicators). How reliable is the presented interpretation?

Citation: https://doi.org/10.5194/egusphere-2024-4021-RC2
- AC2:
  'Reply on RC2', Bella Duncan, 30 Jul 2026
  First of all (and on behalf of co-editors in chief of Climate of the Past, CP), I would like to apologize for the long duration of the review and open discussion stages. The handling editor could not find a second reviewer, even though it was a pre-condition for publishing with EGU journals. Although the research methods used in this study are well-established, they don´t seem to attract enthusiastic reviewers. The editor contacted 14 potential referees, and only one of them, in addition to agreeing to reviewing the manuscript, had actually submitted their review. Then I (as a co-editor in chief) agreed to provide a second review, even though it did require quite some homework from my side, since I am on the numerical modeling side of things. However, in the meantime I got sick and ended up in hospital. With this first paragraph I conclude my apology and come on to the review itself.
  We are grateful that the reviewer has been able to provide a review and are sorry to hear that they have been unwell. Thank you for your review.
  In general, I have a mixed impression from reading and thinking about this manuscript – I believe that both the writing style and structure of chapters/analyses should improve and even more so the interpretation of the results in terms of scientific findings. Messages should be bold and clear, with the appropriate discussion of the uncertainties.
  Writing:
  The manuscript is written in such a technical way that anyone who is not directly involved in the analyses in question would grind their teeth trying to extract the most important scientific findings buried under all the equations, formulas and symbols. There are many numbered equations, but these numbers are never mentioned in the text. Also, the manuscript introduces an overwhelming number of abbreviations and notations that may be familiar to experts but for a reader with a more general background, such as mine, it is difficult to follow through more technical sections such as methods and especially results, but also parts of the discussion. While it is rather common for the methods section, it is much less so for the other two above-mentioned sections. Look at for example, section 3.1 - in one paragraph there are 3 acronyms without reminding what they mean. In the texts that are so abundant in special symbols and equations I would avoid introducing even more acronyms.
  Thank you for this feedback. We have moved the equations and associated descriptions for carbon preference index and average chain length into the Methods. We agree that this helps to streamline the Results and Discussion sections. We have also moved the text describing the calculation and trends in carbon isotope discrimination from the Discussion to the Results section, as we felt that this was best placed directly after describing the carbon isotope results rather than placing it in the Methods. We also removed one equation from the text to simplify how carbon isotope discrimination is described. This leaves 3 equations in the Methods and one in the Results, which is not excessive relative to comparable studies we cite, as well as other plant wax compound-specific isotope-based papers published in Climate of the Past. Each equation is discussed in the surrounding text, but we have adjusted some of this text to remove potential ambiguity, as well as referencing the equation numbers later in the manuscript.
  We appreciate that plant wax-based isotope proxies can be complex and are associated with a standard but large range of abbreviations and notations, and we have attempted to minimise these as much as possible. As described in later responses, we have tried to clarify the reasoning for the steps that we have taken to make the manuscript more accessible to non-experts but feel it is also important not to remove details that an expert would expect to see when reading a manuscript describing plant-wax isotope data.
  Where previously defined acronyms are used again for the first time in the Results and Discussion sections we have spelled them out in full to help remind the reader what they stand for, with the exception of the drill core name (DSDP 270), as this is referred to extensively through the text and is standard terminology for this well-known drilling program. Regarding the paragraph referred to by the reviewer, we have adjusted the text to no longer refer to CPI and HMW, and agree it is helpful to spell out ACL in full as “average chain length”.
  From my point of view, when submitting a biology-heavy manuscript to a journal like CP, one needs to make sure that it is written in a relatable and easy-to-grasp manner so that climate modelers, paleoglaciologists, geomorphologists, and other interested parties can understand and use these results. The subject and the results of this manuscript are clearly within the scope of CP, but its presentation style is not so much. I encourage the authors to polish their text and also visual materials.
  We are sorry that the reviewer has found the presentation of the manuscript difficult to relate to but agree that it is well within the scope of CP, which has an extensive record of publishing similar papers based on plant-wax isotopes. As discussed in other replies, we have endeavoured to improve the readability of the manuscript, while ensuring that the details that an expert would expect to find in a manuscript that described plant wax isotope data remain included.
  Then I was surprised to find quite a lot of grammatical errors and typos, since the author team includes a lot of native speakers – I even consulted with a native speaker from my team. The manuscript was probably written in a hurry, which unfortunately did not result in fast publication.
  We can assure the reviewer that their assumption that the manuscript was written in a hurry is incorrect and was instead the culmination of several years of work, which we concede can sometimes make it hard to ‘see the wood for the trees’ in the final submission phase. No examples of errors or typos have been provided, and so we have thoroughly assessed the manuscript using both manual proofreading as well as double-checking using an AI (Claude) prompt to “identify and list grammatical and typo errors”. This process identified five typos as listed below. We have endeavoured to re-read the manuscript carefully to ensure the language is as clear as possible.
  Line 66: ‘an flame ionisation detector’ changed to ‘a flame ionisation detector.
  Line 79: ‘analysed with the temperature was held’ corrected to ‘analysed with the temperature held’.
  Line 186: ‘overwhelming comprise C₃ vegetation’ corrected to ‘overwhelmingly comprise C₃ vegetation’.
  Line 203: ‘derived from for benthic foraminifera’ corrected to ‘based on benthic foraminifera’.
  Line 254: ‘the Antarctica vegetation’ corrected to ‘the Antarctic vegetation’.
  Several instances where the hyphen was missing from ‘compound-specific isotopes’ have also been corrected. We have also made adjustments throughout the text to improve general readability, as well as the more detailed changes described in other comments.
  Also, there are some assumptions that are made in the text about the prior knowledge of literature. For example, in lines 201 – 204 there is a hidden link to some earlier literature discussing an internal parameter from an equation published elsewhere. At least I have not found where it was shown in this manuscript.
  We assume that the internal parameter is referring to the use of the term ‘δ¹⁸O_sw’, and we apologise that this was not clear. We have adjusted these lines to now read as follows:
  ‘To calculate Equation 4, we derive δ¹³C_atm using values published in Tipple et al. (2010) based on benthic foraminifera. We use δ¹³C_atm calculated from benthic foraminifera as they are not affected by differences in production depth, the impact of photosymbionts or seasonal variability (Tipple et al., 2010). A source of uncertainty in using these values derives from the method by which they were calculated in Tipple et al. (2010), which requires knowledge of the oxygen isotopic value of seawater (δ¹⁸O_sw) in the past, a parameter closely linked to global ice volume (Chappell and Shackleton, 1986). Tipple et al. (2012) apply the same δ¹⁸O_sw value (-0.5 ‰) over the late Oligocene and early Miocene during the calculation of δ¹³C_atm, despite this value varying through glacial and interglacial cycles in response to ice volume changes (Chappell and Shackleton, 1986). While this could affect the reliability of calculated δ¹³C_atm, Tipple et al. (2010) investigate the influence of changing this parameter on calculated δ¹³C_atm and find the impact to be only minor’
  Structure:
  The structure of the discussion section is quite confusing. It feels like the authors are pulling rabbits from a magician´s hat, and each rabbit is from a different continent. There is no cohesion in the narrative of this section, while it is clearly much needed, especially in the discussion. It would be better to avoid mixing actual scientific findings with new equations, introduction of new symbols and concepts, and in some places long discussions of contradictions in earlier studies and current assumptions. Some of these belong to the methods and results (with references to the equation numbers where these symbols are explained). Other issues would be better addressed in the introduction and then referred to in the discussion section. Finally, you should design a separate sub-section in the discussion that describes and analyzes uncertainties in your findings.
  In conclusion: Emphasize your scientific findings and innovations. Currently it looks like a methodological publication, scientific results are poorly represented. Explain how your data synthesis was happening.
  We thank the reviewer for their feedback. This goes back to our comment earlier about sometimes struggling to “see the wood for the trees” after a significant period of time working on a manuscript. We agree with the reviewer’s feedback and have made a range of adjustments to reflect their suggestions, and believe it has much improved the flow of the paper:
  We have shifted and streamlined text describing the background context of carbon and hydrogen isotopes from the Discussion to the Introduction and have also introduced the concept of carbon isotope discrimination here too.
  
  We have moved the text and equations describing the calculation of carbon preference index and average chain length from the Results to the Methods.
  
  We have shifted the text describing the calculation of carbon isotope discrimination from the Discussion section to the Results.
  
  We have made a range of adjustments through the Discussion to simplify and focus the text on our key interpretations and findings.
  
  We investigated creating a separate subsection describing uncertainties but found that to provide sufficient context for each statement resulted in too much repetition of content described elsewhere. Instead, we have more clearly acknowledged uncertainties throughout the Results and Discussion, i.e.:
  Line 232: ‘A source of uncertainty in using these values derives from the method by which they were calculated in Tipple et al. (2010), which requires knowledge of the oxygen isotopic value of seawater (δ¹⁸O_sw) in the past, a parameter closely linked to global ice volume (Chappell and Shackleton, 1986). Tipple et al. (2012) apply the same δ¹⁸O_sw value (-0.5 ‰) over the late Oligocene and early Miocene during the calculation of δ¹³C_atm, despite this value varying through glacial and interglacial cycles in response to ice volume changes (Chappell and Shackleton, 1986). While this could affect the reliability of calculated δ¹³C_atm, Tipple et al. (2010) investigate the influence of changing this parameter on calculated δ¹³C_atm and find the impact to be only minor’
  
  Line 372: ‘We acknowledge that a key uncertainty with this approach is that Arctic vegetation consists of Northern Hemisphere species that were not present in ancient Antarctica, where vegetation was dominated by species similar to those found in modern South America, New Zealand and Tasmania (Kulhanek et al., 2019). However, the climatic conditions experienced by Arctic vegetation compare well to ancient Antarctica, and it is an approach used previously when considering growing conditions for Antarctic plant fossils (Francis and Hill, 1996).’
  
  Line 430: ‘In summary, changes to carbon isotope discrimination in both modern and ancient plants can be driven by a range of environmental variables, making it challenging to disentangle the key drivers, and resulting in a degree of inherent uncertainty in our interpretations.’
  
  Line 487: ‘As with carbon isotope discrimination, the range of environmental variables that can influence plant wax δ²H introduces some uncertainty regarding the interpretation of the key drivers. However, we observe that the same environmental conditions described here, namely colder, shorter, and drier growing seasons, combined with a shift toward lower-altitude vegetation and more stunted growth forms, can account for both the increasing carbon isotope discrimination and more positive δ²H values we see in DSDP 270, which lends greater confidence in our interpretations.’
  
  Interpretation:
  I do understand that going into the deep past is not an easy task and many analogues for the past plant types that we can use today are not ideal matches for such interpretations. However, I also spotted some confusing discussions that are going back and forth on whether a plant-state indicator should increase or decrease in response to wetter/drier or colder/warmer climate conditions and also depending on the latitudes and sun light regimes. The discussion has to become more structured and relatable for a reader with a more general background to be useful for the future scientific progress.
  We have adjusted text throughout the Discussion to summarise some of the relevant research more succinctly and avoid further confusion. For example, at lines 357 to 360, instead of going back and forth between studies that show varying responses of modern plants to changes in precipitation, we have summarised this as ‘Modern equivalents of the Nothofagidites species and Podocarpidites found in DSDP 270 also demonstrate significant variability in the response of Δ to changes in precipitation, with a range of studies finding different sites or species can show either higher or lower Δ as precipitation increases (i.e. Read and Farquhar, 1991; Read et al., 2010; Peri et al., 2012; Griener et al., 2013; Brett et al., 2014; Londoño et al., 2024)’. As discussed in other responses, we have also moved some content from the Discussion into the Introduction, Methods and Results, and streamlined and focused the remaining Discussion.
  Could sea level drop also influence plant migration to lower elevations?
  While growth of a large ice sheet usually results in a sea level fall at distal sites, localities close to the ice sheet can see an opposite signal whereby glacial isostatic adjustment can result in land subsidence proximal to a thickening and advancing ice sheet, and a relative sea level increase due to gravitational effects (Whitehouse et al., 2019). Galeotti et al. (2016) investigated this concept in a western Ross Sea drill core over major glaciation at the Eocene/Oligocene boundary and found that the sea level signal can be complex depending on the location of the drill core, and the location of thickening ice. As such, we consider that while the impact of glacial isostatic adjustment in the Ross Sea during ice advance at Mi-1 is difficult to precisely constrain, glacial isostatic adjustment theory indicates that land around the coast of Antarctica may have seen a sea level increase rather than a fall. Additionally, seismic reflection data in the Ross Sea indicates that the late Oligocene and early Miocene was also marked by an extended period of regional subsidence (De Santis et al., 1995, McKay et al., 2025), which is supported by paleodepth estimates from foraminifera in DSDP 270 which indicate gradually increasing water depth at the core site (Leckie and Webb, 1983; Kulhanek et al., 2019).
  You are presenting a description of many measured factors but with a very limited interpretation of their climate significance. This might be confusing for the reader.
  As described in previous responses, we have made an array of changes to streamline the text and focus on our findings and their significance.
  Specific comments:
  Lines 137-138: Is there any idea about the origin of this deviation?
  Thank you for highlighting that we have not sufficiently explained our reasoning here. The three n-C₃₁ data points are all significantly more negative than those in adjacent samples, by 5, 7 and 8‰. After reassessing these samples we note that the n-C₃₁ alkane has partial coelution with the C₂₉ αβ hopane, and in these three cases we believe that the measured value may instead be reflecting the C₂₉ αβ hopane, as these can display more depleted δ¹³C values than n-alkanes (Inglis et al., 2019). The three values we removed for n-C₂₅, n-C₂₇ and n-C₂₉ all come from the same sample, which had a low recovery of n-alkanes compared to other samples, and was the only sample analysed for compound-specific isotopes in a separate batch. After reassessing this, we have decided that the δ¹³C values in this sample are likely less reliable and have removed the n-C₃₁ value from our data set too. We have added some text to explain our rationale here:
  ‘A single sample (at 194.8 mbsf) was removed from the data set as it was considered less reliable due to differences of more than 2 ‰ compared to adjacent samples (more negative for n-C₂₅ and more positive for n-C₂₇ and n-C₂₉). This sample was associated with low extract recovery and was also the only sample analysed for δ¹³C in a later batch than the rest of the samples. The n-C₃₁ values of three samples were 5, 7 and 8 ‰ more negative than adjacent samples. We noted this compound partially coeluted with C₂₉ αβ hopane, which can display more negative δ¹³C values than n-alkanes when sourced from methanotrophic bacteria (Inglis et al., 2019). We therefore suspect that in these cases partial hopane coelution biased the n-C₃₁ values and we have removed them from the data set.’
  Lines 203-204: Has this been demonstrated or only assumed?
  We appreciate that our wording here was not clear. This is a well-defined relationship and commonly used proxy, and we have now adjusted the text to reference the relationship between ice volume and δ¹⁸O_sw back to an early paper which describes it:
  ‘A source of uncertainty in using these values derives from the method by which they were calculated in Tipple et al. (2010), which requires knowledge of the oxygen isotopic value of seawater (δ¹⁸O_sw) in the past, a parameter closely linked to global ice volume (Chappell and Shackleton, 1986). Tipple et al. (2012) apply the same δ¹⁸O_sw value (-0.5 ‰) over the late Oligocene and early Miocene during the calculation of δ¹³C_atm, despite this value varying through glacial and interglacial cycles in response to ice volume changes (Chappell and Shackleton, 1986).’
  Line 206: Why temperate?
  Thank you for this comment, we understand that this may have been confusing. Available data for the isotopic fractionation between plant tissue and individual n-alkanes comes from temperate climate zones instead of high latitudes. We have reworded this sentence to reflect this more clearly:
  ‘To calculate this, we apply n-alkane specific averages from published data for a wide range of C₃ plants from temperate climate zones, as ε_lipid data in high latitude plants is limited in the literature (Collister et al., 1994; Chikaraishi and Naraoka, 2003; Diefendorf et al., 2011).’
  Lines 207-2010 and similar: Why not place these numbers in tables?
  We thank the reviewer for this suggestion but as there are not many numbers discussed (i.e. in the lines suggested there are two parameters with five values each), we prefer the current format. While this does create a list of numbers, we do not think it is excessive in this section and is a more efficient use of space. However, if the editor decides a table is better, then we are also comfortable with that.
  Line 250: Permafrost could have also developed during Mi-1.
  Yes, we agree that Antarctica likely had permafrost present during the late Oligocene and early Miocene, which leads us to consider the Arctic as an analogue for ancient Antarctica i.e. in text:
  ‘Like the modern Arctic, at latitudes south of the Antarctic Circle, vegetation growth would have been restricted to a short period during summer when there was sunlight, a thawed layer of soil and warmer temperatures for photosynthesis.’
  Paragraph of line 246+: There are significant differences between plant types in the Northern and Southern Hemispheres even now. Doesn´t it have to be discussed?
  This is a good point and we now more carefully acknowledge this uncertainty:
  ‘We acknowledge that a key uncertainty with this approach is that Arctic vegetation consists of Northern Hemisphere species that were not present in ancient Antarctica, where vegetation was dominated by species similar to those found in modern South America, New Zealand and Tasmania (Kulhanek et al., 2019). However, the climatic conditions experienced by Arctic vegetation compare well to ancient Antarctica, and it is an approach used previously when considering growing conditions for Antarctic plant fossils (Francis and Hill, 1996).’
  Lines 263-271 (and similar): There are so many different factors that affect delta (and other indicators). How reliable is the presented interpretation?
  We understand the concern here, but note that almost all paleo-proxies have multiple solutions. We believe it is best practice to explain a range of possible interpretations and then narrow them down through comparison with other proxies and logical arguments. If non-unique solutions still remain, it is valid to present these, as it helps to inform possible feedbacks and Earth system responses that lead to changes in carbon and hydrogen isotopes that could be tested by models or future data-based studies. We now more clearly acknowledge the uncertainties in plant wax δ²H interpretations; however, the methods we use and interpretations throughout the text are based on a framework that is well established in the field: ‘As with carbon isotope discrimination, the range of environmental variables that can influence plant wax δ²H introduces some uncertainty regarding the interpretation of the key drivers. However, we observe that the same environmental conditions described here, namely colder, shorter, and drier growing seasons, combined with a shift toward lower-altitude vegetation and more stunted growth forms, can account for both the increasing carbon isotope discrimination and more positive δ²H values we see in DSDP 270, which lends greater confidence in our interpretations.’
  
  References
  De Santis, L., Anderson, J. B., Brancolini, G., & Zayatz, I. (1995). Seismic record of late Oligocene through Miocene glaciation on the central and eastern continental shelf of the Ross Sea. Geology and Seismic Stratigraphy of the Antarctic Margin, 68, 235-260.
  Leckie, R. M., & Webb, P. N. (1983). Late Oligocene–early Miocene glacial record of the Ross Sea, Antarctica: Evidence from DSDP site 270. Geology, 11(10), 578-582.
  Kulhanek, D. K., Levy, R. H., Clowes, C. D., Prebble, J. G., Rodelli, D., Jovane, L., Morgans, H. E. G., Kraus, C., Zwingmann, H., Griffith, E. M., Scher, H. D., McKay, R. M. & Naish, T. R. (2019). Revised chronostratigraphy of DSDP Site 270 and late Oligocene to early Miocene paleoecology of the Ross Sea sector of Antarctica. Global and Planetary Change, 178, 46-64.
  McKay, R., Cockrell, J., Shevenell, A. E., et al. (2025). Miocene ice sheet dynamics and sediment deposition in the central Ross Sea, Antarctica. Geological Society of America Bulletin, 137(3-4), 1267-1291.
  Whitehouse, P. L., Gomez, N., King, M. A., & Wiens, D. A. (2019). Solid Earth change and the evolution of the Antarctic Ice Sheet. Nature communications, 10(1), 503.
  
  Citation: https://doi.org/10.5194/egusphere-2024-4021-AC2

Bella J. Duncan, Robert McKay, Richard Levy, Joseph G. Prebble, Timothy Naish, Osamu Seki, Christoph Kraus, Heiko Moossen, G. Todd Ventura, Denise K. Kulhanek, and James Bendle

Supplement

https://doi.org/10.5194/egusphere-2024-4021-supplement

Bella J. Duncan, Robert McKay, Richard Levy, Joseph G. Prebble, Timothy Naish, Osamu Seki, Christoph Kraus, Heiko Moossen, G. Todd Ventura, Denise K. Kulhanek, and James Bendle

Viewed

Total article views: 7,237 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
6,192	850	195	7,237	342	311	429

HTML: 6,192
PDF: 850
XML: 195
Total: 7,237
Supplement: 342
BibTeX: 311
EndNote: 429

Views and downloads (calculated since 09 Jan 2025)

Month	HTML	PDF	XML	Total
Jan 2025	436	165	10	611
Feb 2025	215	20	5	240
Mar 2025	158	10	0	168
Apr 2025	120	50	10	180
May 2025	158	181	20	359
Jun 2025	320	20	20	360
Jul 2025	161	55	5	221
Aug 2025	665	5	0	670
Sep 2025	2,606	32	8	2,646
Oct 2025	205	30	10	245
Nov 2025	210	99	30	339
Dec 2025	210	35	15	260
Jan 2026	216	55	20	291
Feb 2026	104	23	21	148
Mar 2026	253	36	10	299
Apr 2026	64	13	2	79
May 2026	33	8	2	43
Jun 2026	19	1	0	20
Jul 2026	23	10	3	36
Aug 2026	16	2	4	22

Cumulative views and downloads (calculated since 09 Jan 2025)

Month	HTML	PDF	XML	Total
Jan 2025	436	165	10	611
Feb 2025	215	20	5	240
Mar 2025	158	10	0	168
Apr 2025	120	50	10	180
May 2025	158	181	20	359
Jun 2025	320	20	20	360
Jul 2025	161	55	5	221
Aug 2025	665	5	0	670
Sep 2025	2,606	32	8	2,646
Oct 2025	205	30	10	245
Nov 2025	210	99	30	339
Dec 2025	210	35	15	260
Jan 2026	216	55	20	291
Feb 2026	104	23	21	148
Mar 2026	253	36	10	299
Apr 2026	64	13	2	79
May 2026	33	8	2	43
Jun 2026	19	1	0	20
Jul 2026	23	10	3	36
Aug 2026	16	2	4	22

Viewed (geographical distribution)

Total article views: 7,176 (including HTML, PDF, and XML) Thereof 7,176 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 13 Aug 2026

Short summary

We use plant wax compound specific stable isotopes to investigate how ancient Antarctic vegetation adapted to glacial conditions 23 million years ago. We find plants became less water efficient to prioritise photosynthesis during short, harsh growing seasons. Ecosystem changes also included enhanced aridity, and a shift to a stunted, low elevation vegetation. This shows the adaptability of ancient Antarctic vegetation under atmospheric CO₂ conditions comparable to modern.


Total:	0
HTML:	0
PDF:	0
XML:	0