This paper is one in a series by the authors which look at the use of the bivalve Arthritica helmsi as an indicator of past hydrologic and climate change. It builds upon previous papers which look at modern and museum specimens of this species and its suitability for isotopic studies. This study is located in an area of significant interest in Australia, with regards water management, cultural connections and understanding past hydrologic conditions. The study is well-constructed, and the findings clearly presented. The establishment of the local reservoir effect for radiocarbon dating is particularly to be commended. The results are of significance not only in addressing a long-running debate in the palaeohydrology of the Coorong Lagoon, but also in adding another high-resolution climate record for southeast Australia over the past ~2000 years, allowing more robust comparisons regionally and globally.

Specific comments

Line 12: advise giving some context to the conservation and restoration efforts and how palaeoclimatic records can be useful in addressing these.

R: We agree that more context would clarify how studies such as this one can benefit conservation. We will add context to the Introduction section of this manuscript to keep the Abstract succinct.

Line 30: the reference to arid and semi-arid regions here seems a little out of place. Perhaps set the geographic context of the study first.

R:  We will delete the reference to separate regions to simplify the text.

Line 39-40: where are these located in relation to the study site – geographically and climatically? Would you expect these to be congruent?

R: We will add a map showing the locations of the study sites from these studies in relation to the site in the current study. The sites are mostly from southeast Australia so the authors would expect that the locations are influenced by the similar climate drivers as the site in the present study. A map illustrating the proximity of records will be a good addition to the manuscript – thank you for the suggestion.

Line 130-1 and Figure 3: You are to be commended for do analysis on individual valves as well as the bulk samples. However, the individual valves show a very broad distribution in oxygen isotope values, at any given depth. What is the justification for 5 valves per sample being representative?

R: Thank you for this comment, we agree this information is important and should be added to the text. Five valves was selected as it meant there would be consistency across the study as the number of valves in each sediment layer was variable. It was also found in the lead authors Honours Thesis (Chamberlayne, 2015) that five valves was representative of the trace element value in a larger sample size of 16 valves. The authors will add this information to the text of the present study.

Line 210: relatively dry in the context of the record or in comparison with other areas?

R: The authors are stating that it is relatively dry in the context of the record. We will edit the text to make this clear.

Line 239: Perhaps show on a map where the records in Dixon et al, 2019 are in relation to the current record. Climatically, would you expect the same response or not?

R: We will add a map showing the locations of the study sites from these studies in relation to the site in the current study. The sites are mostly from southeast Australia so the authors would expect that the locations are influenced by the similar climate drivers as the site in the present study. A map illustrating the proximity of records will be a good addition to the manuscript – thank you for the suggestion.

Line 335-337: I would recommend expanding on this correlation a little. How are you defining the region here and how do each of the drivers you mention here relate to wetter or drier conditions? Can you unpick the influences of each of these with respect to the variability in your record – and the regional context? How does it help build the story?

R: The authors will add text to the manuscript stating the influence of the mentioned drivers on the climate of the region. We will clarify the region and how similarities in the record to others in the region show a common influence. The next comment suggests splitting the conclusion section into two paragraphs. The authors will implement this suggestion which will provide more space for the expansion on conclusion statements as suggested in this comment.

Conclusions: Suggest splitting into two sections – Firstly the palaeoclimate, and then the Coorong and management implications as separate paragraphs.

R: The authors will split the conclusion into two sections for clarity of ideas.

Figure 1: I would suggest including a map showing major climatic zones or influences of major climatic drivers. May also be worth including a map showing the locality of this site in relation to others mentioned in the text – both in SE Australia and globally.

R: The authors will edit Figure 1 to contain both the major climate drivers in the region and the locations of the study sites from other studies mentioned in the text.

Figure 4: Consider annotating which of these records were utilised in the Dixon et al., 2019 compilation

R: The authors will add this information to the figure caption

The authors wish to thank the reviewer for their positive and constructive comments which have improved the clarity of this manuscript. We will add context to the sections suggested by the reviewer in addition to editing Figure 1 to include the locations for the studies included in the Dixon et al. 2019 paper.

References:

Chamberlayne, B., 2015. Late Holocene seasonal and multicentennial hydroclimate variability in the Coorong Lagoon, South Australia: evidence from stable isotopes and trace element profiles of bivalve molluscs, honours thesis, University of Adelaide, Adelaide.

The two primary concerns I have are 1) if this is a temperature or hydroclimate record and 2) in relation to the age model.

Firstly – the authors provide fairly weak evidence that despite establishing a d18O and temperature relationship for this site, they believe their d180 record better represents rainfall/flow. The authors need to provide further evidence/discussion of this through analysis between contemporary flow/rainfall and d180 in the top layers. This is crucial as an alternative interpretation of the results (based on the stated temperature relationships) is that that 500-1100 was cool rather than wet and 1100-1750 was warm rather than dry.

R: The authors thank the reviewer for this valid comment, but we do not believe it would be possible to add to this study due to the lack of samples in the top centimetres of the core from the South Lagoon. Furthermore, while temperature may contribute to changes in d18O, the change required to result in the range of d18O values measured in A. helmsi valves is unrealistic for the region (~10 °C). This is discussed in section 4.1 of the manuscript. The authors can edit this section to make a stronger argument.

Secondly – The top 500 years of the age model doesn’t appear to be well constrained and the authors exclude some dates based on being outliers. There is a sedimentary horizon at 40cm with a date below of 1444 and a date above (20cm) of 1783. This is a big gap. Could there be a hiatus in between these dates or a change in deposition rate? Another question is if the shells were in dead or live position as I am aware they can burrow into sediments and therefore may ‘move into older layers’. Also the marine reservoir effect may vary over time. For example, a bushfire could result in a high influx of young carbon, so subtracting 800 years in this instance would be erroneous. If the lake has a lot of vegetation and organic matter surrounding it, this may act to offset the carbon - groundwater age. While this may not be able to be completely addressed, the authors should discuss this in their discussion and conclusions as potential sources of uncertainty.

R: The points made by the reviewer in this comment are valid and we agree that a more detailed discussion of these potential uncertainties should in included in the text. Resolving these uncertainties is outside the scope of this study, but the authors will highlight the potential for further research to better constrain the timing of changes in the d18O record. We will add detail to the discussion and conclusions of the manuscript highlighting this source of uncertainty.

Other comments:

In the first line of the abstract, the authors mention the resilience of aquatic ecosystems, however, palaeoclimate data is relevant to all ecosystems including terrestrial.

R: We agree and will remove the word aquatic

Line 35 – highlights the lack of decadal scale records. However this statement is true for both high and low frequency reconstructions Reference to Dixon et al 2007 as ‘recent’ is probably not quite right. There have been advancements in the last 5 years. Particularly for Tasmania for example but also WA.

R: Dixon et al was published in 2017. We will add a sentence highlighting some recent papers from TAS and WA.

Of the 9 records in Dixon, how many are on the mainland in SE Australia – it is worth noting for context for this study.

R: We will add a map to Figure 1 showing the locations of the study sites from these studies in relation to the site in the current study.

Line 40 – I don't think it is appropriate to lump the ANZDA in with reconstructions based on a single remote proxy. The ANZDA is based 176 tree-ring chronologies and one coral series from both Australia and wider Pacific.

R: The authors will highlight the number of studies included in this reconstruction in the text and clarify any comparisons with single proxy data studies.

Line 47 – A flood is not a decadal phenomenon. Floods tend to build rapidly, peak and subside within a week or so. Droughts can last seasons to years. The term pluvial or flood dominated epoch would be more appropriate.

R: Agree, we will instead refer to drought or flood dominated epochs.

It would be good if the authors could provide a stronger case as to why low frequency reconstructions are useful. In the first line of the introduction the authors state that “Multi-decadal to centennial records of past hydroclimate variability are crucial for understanding long term climate drivers, for calibrating and validating climate models, for assessing hydroclimate sensitivity to external drivers and for estimating the probability of multi-decadal climate extremes”, however much of this requires annual or sub annual data. I am not suggesting low frequency reconstructions that cover longer periods are not useful but please spell out why they are an important piece of the puzzle in the introduction.

R: The authors will add context to this statement to build a case as to why both low and high frequency records are valuable.

Line 70 – ‘Moreover’ should be ‘however’

R: Amended

Line 91 – Remove “formally speaking”

R: Amended

The authors mention the use of Pinus Pollen in identifying when modern section of the record. How is the pollen actually identified? Please include details

R: The authors will clarify in the text that the pollen collection and identification method is outlined in another paper (Krull et al. 2009)

Line 135 – missing “the”

R: Added

The authors mention some samples with erroneous dates, where they were identified as older than the sequence in which they were in. is this evidence of the aged carbon offset not being constant? If not, why might then be in error?

R: The erroneous dates may have been a consequence of mixing in the top sediments during storm activity or similar. Another possibility is that old shells were washed into the lagoon. Along the banks of the lagoons are past shorelines which contain abundant amounts of A. helmsi shells. It is possible that these were transported by wind, water or wildlife. We will add this information to the text.

Section 4.1 – Perhaps the authors could compare to Tasmanian temperature reconstructions by Allen et al? Or the Lake Tay reconstruction by Cullen and Grierson 2009

R: The authors will include references to these papers in section 4.1

Figure 1 – Add a box around Australia as it is a different scale to the rest of the map

R: The authors will add a box around Australia in Figure 1.