Oliveira and co-authors present a new pollen record from the Bay of Bengal, documenting Indian vegetation and monsoon variability across MIS 11, complemented by climate model simulations aimed at improving our understanding of orbital- to millennial-scale variability. As a new referee for this manuscript, I acknowledge the considerable effort the authors have made in addressing the comments raised by the previous two reviewers, and the clarity of both the text and figures has improved substantially. However, the manuscript currently lacks a coherent chronological framework and a clear separation of orbital and millennial-scale variability, which in turn undermines both the interpretation of vegetation changes and the justification for the included modelling results. Alongside numerous technical issues detailed below, I have three major concerns that need to be addressed before the manuscript can be further considered for publication in Climate of the Past.
1. whether a consistent and synchronised age model has been established to support multi-proxy comparisons;
2. the lack of a clear and robust demonstration of orbital-scale signals in the presented records; and
3. the rationale, necessity, and added value of the modelling results within the overall framework of the study.
I therefore recommend a major revision.
L107-110: The core of this study lies in the new pollen record combined with the improved age model. I recommend providing a dedicated figure illustrating the age model, together with its propagated uncertainties. This would constitute a key foundation for comparing the pollen record with other proxy records. In addition, please specify the age models used for the comparison records, for example, whether they are based on independent chronologies or synchronised to LR04. Without such a foundation, statements—particularly those concerning millennial-scale variability—should be treated with caution.
L192-194: Do the model simulations cover the full study interval, i.e. MIS 11? If so, please present the full simulated time span (370–428 ka) in Fig. 4. Notably, I am surprised that no millennial-scale variability or AMOC reduction is evident in the current modelling results. As shown in Fig. 1 of Yin et al. (2021), pronounced abrupt AMOC reductions have been documented. Therefore, if the present manuscript indeed uses the same model outputs as Yin et al. (2021), one would expect to see an associated abrupt precipitation change around ~398 ka. Please clarify this apparent inconsistency.
L201-205: The interpretation in Clement et al. (2024), that shifts in major ecological groups primarily reflect changes in ISO rainfall, is based on two warm interglacial periods characterised by relatively stable CO₂ concentrations and thus limited changes in the background temperature state. This context should be carefully considered when discussing orbital-scale vegetation variability in the present study, which also spans glacial inception periods. For example, in lines 206–209, the dominance of semi-arid steppe vegetation may also reflect colder climatic conditions. I do not suggest that the current interpretation is necessarily incorrect, but I encourage the authors to provide a more explicit and unequivocal statement regarding what the records specifically indicate.
L216: I also recommend specifying the exact figure panels being referred to. For example, “Fig. 2” should be revised to “Fig. 2e, d” here. There are many similar instances throughout the manuscript that should be corrected.
L219: Based on Fig. 2e and d, the maximum abundance of tropical forest and wet evergreen forest taxa appears to peak at ~410 ka. Could the authors clarify why a broader interval of “419–408 ka” is given here? Is there a formal definition of maximum/minimum intensity that has not been described in the text? If not, I recommend specifying ~410 ka rather than using such a wide time window.
L223: The term “climatic thresholds” is introduced without prior definition. Please clarify what is meant here.
L233-236: How are the expansion periods defined here? Based on visual inspection, I do not find that the two proposed expansion phases coincide with decreases in δD. Please clarify.
L241: “Forest event” should be capitalised as “Forest Event”.
L295-297: it would be helpful to explicitly spell out the proxy records referred to here (e.g. δ¹⁸O, δD, or Antarctic ice-core records). Ideally, a figure comparing these multiple records should be provided, which would likely serve as a key reference figure for understanding the broader climate evolution during MIS 11c.
L309: What is meant by “two comparable peaks” here? Is this referring to Fig. 2h?
L316-317: What are the “several distinct patterns” referred to here?
L317-321: On orbital timescales, the two speleothem δ¹⁸Oₛₚ records (Sanbao and Yongxing) exhibit similar patterns. Moreover, vegetation variability on orbital timescales is influenced not only by precipitation but also by temperature and other processes. Therefore, it is challenging to conclude that pollen records necessarily provide a better proxy for local hydrological changes than speleothem δ¹⁸O records, although I agree that the new pollen record offers evidence supporting Tzedakis’ hypothesis. If this interpretation is not intended, please clarify. In addition, the Jingfo record should be included in Fig. 2.
L322-338: I find this section difficult to follow because millennial- and orbital-scale variability appear to be mixed in the presented records (Fig. 2). I strongly recommend extracting the orbital-scale signal prior to this discussion, for example by applying a 5-kyr running mean to represent orbital-scale variability.
L334-336: Are humidity/aridity here assumed to be equivalent to strong/weak monsoon rainfall or circulation? In addition, the authors highlight an important feature of Asian cave records: higher summer insolation does not necessarily lead to more depleted isotopic values (e.g. MIS 5e vs. 5c and MIS 11c vs. 11a). Could the pollen record provide further insight into this issue?
L355-357: What mechanism could account for the prolonged or more frequent El Niño conditions around ~408 ka? Several studies have documented AMOC disruption events during this interval (e.g. Galaasen et al., 2020, *Science*, 367, 1485–1489). Could this provide an alternative explanation?
L406-408: Were the sharp rises in CO₂ driven by abrupt AMOC intensification?
L435-436: I do not observe a stable sea level during the early to mid MIS 11c, but rather a pronounced rise toward the highest sea-level stand during mid MIS 11c. Please clarify and revise the related statements accordingly.
L438: Is the age model synchronised with other proxy records? If not, this statement appears too strong and should be toned down. In addition, based on visual inspection of Fig. 4, I do not identify a clear lead of CO₂ maxima relative to ice-volume minima. Please clarify.
L434-456: Nearly all statements in this section are inferences without direct support from the modelling results, which makes the purpose of including the modelling component unclear. In particular, LOVECLIM includes relevant experiments that could help disentangle the respective roles of CO₂, ice sheets, and insolation in orbital-scale ISM variability. As noted earlier, the full MIS 11 simulation output should be shown in Fig. 4. Furthermore, if CO₂ and ice sheets are indeed the primary drivers of orbital-scale variability in the Indian CMZ, one would expect a relatively weak precessional signal in the record. Please explicitly separate the orbital-scale signal and confirm this. Finally, in Fig. 4f, why is the simulated PI level much weaker than during MIS 11, especially given that CO₂ levels are lower than PI values?
L458: As noted earlier, is orbital-scale ISM variability unequivocally represented by pollen-based vegetation changes? What is the role of temperature in driving vegetation variability, and does ISM variability itself also involve temperature changes? Please clarify.
L464-478: The rationale for the modelling experiments remains unclear and somewhat confusing. The modelling results do not appear to support the pollen-based interpretation that emphasises the roles of CO₂ and ice-sheet dynamics in driving orbital-scale vegetation changes in the Indian CMZ. Under these circumstances, the inclusion of the modelling component may not strengthen the manuscript’s conclusions, but instead risks confusing readers about the main findings. Please clarify the purpose and added value of the modelling experiments. |
This article is a high-quality piece of work with some important new data spanning an interesting climatic transition with material being delivered from central India and providing critical new constraints on the development of environment during interglacial periods within the core monsoon area. I believe that it is worthy of publication but it will need some rewriting before it could be accepted. I actually found the paper really difficult to follow and understand. The writing is very dense and there is so much use of abbreviations that it makes it very difficult to follow. Not everyone is going to be as familiar with all these abbreviations as the authors and I found their use obstructive to my understanding of the manuscript. I’m sure the data is a good quality and if I understood this correctly it suggests that during warm periods there was a greater prevalence of tropical forest in central India compared to other times but that the intensity of the Indian monsoon and the vegetation is regulated by orbital processes but also the extent of ice sheets in high latitudes. One of the other things that stopped me understanding the manuscript more fully was the use of the MIS stages rather than telling me whether it’s warmer or a colder period. Again, the number of the stages and sub stages are probably very familiar to the author but not to all the readers and I think it might be helpful if they were to try and better describe what each provide more information about what each stage number means. The authors assume too much prior knowledge. I was a little confused about the comparison of the climate model and the data itself. The conclusion appeared to be that the model wasn’t very good unless you accounted also for glacial variations which at least in my reading suggests at the model isn’t that good and so I wonder why so much time was spent discussing the results. I provide here below a number of smaller editorial comments and questions to help improve the understanding of the manuscript.
Line 18 - Marine Isotope Stage (MIS) 11 - Should say how old this is in ka
Line 20 - core monsoon zone (CMZ) -What is that?
Line 28 - MIS 11c - Again, we need the ages of these substages
Line 36 - ISM weakening could also occur under similarly warm future conditions – Triggered by what? Be specific
Line 43 - Core Monsoon Zone (CMZ) – Label this on a map
Line 73 – IODP – Define abbreviation. And SST
Line 79 - during MIS 5e - provide age in ka
Line 101 - Integrated Ocean Drilling Program (IODP)- Say this on first use
Line 106 - LOVECLIM Earth System Model – Need a reference here
Line 111 - δ18Ob – Superscript “18”. Is “b” right?
Line 140 - rapid delivery of continental material – Depends what you mean by “rapid”. Could be a whole 100 ky glacial cycle for coarse grained sediment. There is likely buffering in the flood plain for fine material including the pollen.
Figure 1 - Text in inset North Atlantic map needs to have the same font as the rest of the figure. Text needs to be larger.
Line 154 - A total of 62 levels were sampled for pollen analysis, - A total of 62 samples were taken for pollen analysis
Line 163 - main sum - I’m not quite sure what this means
Line 163 - excluding Himalayan taxa – Why are there Himalayan taxa in the sample. Is the sediment really from the Mahanadi or perhaps the Ganges?
Line 171 - tropical forest (TF) - Not sure this abbreviation is necessary
Line 174 - tropical forest (TF) – If you are going to abbreviate you only need to define this once.
Line 185 - primary indicator of ISM rainfall variability - But the monsoon is seasonal. Couldn’t this be found in a tropical, non-seasonal environment?
Line 189 - Fig. 4 - Call out figures in numerical order
Line 196 - over the India's CMZ (17°-27.5° N, 67.5°-90° E) - over the Indian CMZ. Also mark this on the map figure.
Figure 2 – This chart is very hard to read.
Line 231 - abundance tropical deciduous forest – abundance of tropical deciduous forest
Line 238 - increasing dryness. – and seasonality?
Line 241 - Site U1446 shows a gradual TF decline- pollen from Site U1446 shows a gradual TF decline
Line 246 - PZ MIS 11a-1, - You use too many abbreviations. I already forgot what PZ means
Line 250 - substantial less humid condition - substantially less humid conditions
Line 250 - The two forest expansion - What does that mean?
Line 258 - decrease of 22.3% - decrease to 22.3%?
Line 296 - long records – long duration records
Line 300 - unusually high global terrestrial biosphere – Not sure what this means
Line 305 - contribution of tropical forests to the global terrestrial biosphere - you did not estimate volumes of forest or the volumes of carbon that would be captured by the forest so I don’t think you really understand what proportion of the global budget this represents.
Line 314 - asymmetric “M-shaped” structure – I’m unclear about what you mean. Are you talking about the shape of proxy record? Which one in particular?
Line 325 - EDC ice core – what does EDC mean? And where was this core taken?
Line 327 - imprint on the CH4 record - what record are you talking about? What is a proxy for methane?
Line 339 - tropical wetland extent – how can a speleothem record tell you about the extent of wetlands?
Line 343 – speleothem data alone cannot be used as a definitive proxy - or that these data indicate complicated hydrological system
Figure 5 - I suggest you use different colours when you’re plotting two sets of data on the same plot. Using two shades of blue is a little confusing. Why don’t you just use something more distinctive?
Line 381 - The asymmetric M-shaped pattern – I don’t know what you mean. You’re going to need to label that quite clearly on one of your figures.
Line 386 – NHSI - This is a confusing abbreviation. If you told us what it was earlier in the manuscript, then I’ve already forgotten. I suggest you don’t use this.
Line 395 - gradually declines - gradually declined
Line 426 - the absence of ice-sheet forcing and associated freshwater flux in the simulations - this seems like a major omission to me. Is it necessary to ignore the ice sheets in the model? Seems like a bad model.
Line 444 - Yongxing Cave -but that is east Asian monsoon and not entirely clear that it’s appropriate.
Line 448 - MIS 11c optimum in southern Europe – I don’t understand. Site U1446 and the Yongxing Cave are both not in Europe
Line 472 - involves distinct in-cave processes – also, they may be issues related to regional rainfall patterns
Line 483 – soutprashern - Spelling
Line 492 - monsoon Hadley circulation patterns – I don’t understand. Monsoon circulation is exactly the opposite of Hadley circulation at least in South Asia.
Line 498 - still highly discussed - strongly debated? Controversial?
Line 505 - carbon dioxide jumps (CDJ) - this is entirely inappropriate. You have far too many abbreviations. The manuscript is extremely hard to read anyway without you adding this. Nobody understands these abbreviations when there are so many..
Line 503 - DO-like - More confusion
Line 526 - Ice-Rafted Detritus (IRD) - Please stop
Line 541 - the asymmetric M-shaped pattern in our tropical forest record -I never really understood what you meant by this.
Line 548 - IN the CH₄ record - IN?
Line 555 - CH₄ overshoots – I don’t understand what that means