Dear Editor,

The manuscript titled "Coupled Otolith and Foraminifera Oxygen and Carbon Stable Isotopes Evidence Paleoceanographic Changes and Fish Metabolic Responses" by Agiadi et al. presents a novel and significant contribution towards understanding the impact of extreme events, particularly the preconditioning stages of the Mediterranean Salinity Crisis, on marine fish life. This study stands out for its incorporation of higher organisms, adding substantial value to the research. As the authors state, the Mediterranean is indeed a great example of how different groups of organisms cope with changes in oxygen levels, nutrients and salinity and therefore I believe that this work is of extreme interest for many scientists dealing with the evolution of the Mediterranean, both from a biological and geological perspective. The manuscript is clear, well written and the obtained data are coherently interpreted.

Apart from some minor changes to be applied to the text some clarifications are needed:

• Where does the age model come from? Has it been already published in previous works?
• Elaborate or make more explicit why the 13C trend in otoliths does not reflect the Mediterranean or Global trend of decrease. Is it because the strong influence that nutrients have?, should then the trend be opposite in this specific case? Does it mean that they reflect more local changes and cannot be used for global changes in climate? Do they highlight punctual more abrupt changes?

Line 113: parenthesis missing for reference (Whitley, 1941)

Line 115: has been reported

Line 119: Today, it has a preferred temperature….

Line 244: Strange sentence, not clear, n that it requires substitute

This work is novel as it attempts to explore physiological responses of higher trophic level organisms (fishes) to large scale climatic fluctuations. Generally in deep time studies, physiological responses are inferred (e.g. as a response to temperature or via changes in body size or species compositions). Direct reconstruction of organism physiology across time series of environmental change is an exciting development.

I would comment in general that co-incidence of physiological change with temperature / salinity change does demonstrate causation and likely cannot be used to unequivocally identify drivers of local extirpation– but it is certainly interesting and informative to explore physiological responses.

Another important angle to the work which is not developed in the introduction is the perspective that deep time studies can bring to modern ecological-evolutionary studies – for instance evolutionary adaptation within populations is expected to reduce long term physiological effects of drivers (e.g. temperature) at least within the organisms’ thermal window. Deep time studies provide ability to test the capacity for species and communities to resist physiological effects of directed environmental change – this could be brought out a bit more in the intro / conclusion.

I do have some requests for clarification around the data analyses.  

I would find it helpful in the data analysis section to define the purpose / hypothesis behind each analysis. I understand the test for size-dependency on otolith d¹³C values, but the reasoning underpinning the correlation tests (lines 141-145) were less clear to me. There is temporal auto-correlation expected – presumably the analyses are essentially looking for the effect of temporal autocorrelation rather than an effect of (say d¹⁸O foram on d¹⁸O fish…).  Perhaps because I didn’t fully understand the rationale for these correlations, I found the description of their relative strength– which are not visualised anywhere- difficult to follow, or to grasp the significance or otherwise of different correlations.  Could the whole analysis be analysed in a single model with lithology as a fixed effect  - and with some consideration of temporal structure?  In this way, fixed effects of taxon (and interactions) could be used to explore similarities or differences in their responses to either salinity or temperature.

It is difficult to determine if there are sufficient samples to validate the correlations in each parameterization.

I appreciate the novelty of attempting to recover metabolic responses to environmental drivers that forms the central argument to the study. The metabolism story of progressively increasing separation between d13Cforam and d¹³C otolith esp in the upper part of the sequence is really interesting, as that does imply that whatever causes the surface water DIC shift (argued for reducing surface productivity)is coincident with decreasing metabolic activity in fish –  this is potentially a really nice observation.

I am struggling a little to be certain that the shorter-term excursions, sometimes indicated by single data points can be reliably interpreted. The decoupling between pelagic and benthic fish time series is promising. Temporal co-incidence between d13Cshifts and other proxies is interesting, but I would at least want to see replication of the effect and to have things like diagenetic effects ruled out. 

more techical comments:

Line 60 – final sentence here doesn’t follow – respiratory (metabolic) C has lower d13C value than DIC because of preferential incorporation of 12C during photosynthetic fixation of C at the base of the food chain.

Materials and methods

A little more detail of exactly which horizons were sampled is needed. What mass / volume of sediment was sampled? Are these from the marl or sapropel (or both) horizons? Was any chemical used to break down sediment for wet sieving?

L80 – and elsewhere  - numbers of replicates is potentially an issue here – why only one otolith per horizon?

L83 – how was diagenetic condition assessed? Are the otoliths still purely aragonitic? Is there any secondary calcite? Line 93 clarifies that Smicroscopywas used, ok, but what do you expect for a well vs poorly preserved otolith? I guess this is much more poorly understood than for forams..

L85 – fish taxa selected due to their‘well established ecology’ – please summarise what is known… OK provided in line 119 (maybe move up as this is prior knowledge rather than newly inferred so doesn’t need to be among the methods)

L88-89 – I’m not clear what is meant by species specific differences in vital effects, or why selecting adult samples would mitigate against this. Indeed the premise of using d13C as a metabolic proxy is surely to quantify species specific differences in ‘vital effects’ (vital effects being a rather undefined geochemist term usually reflecting processes associated with metabolic rate…)

Line 95-100 Clarify - were whole otoliths crushed and homogenized, or was the outer surface analysed? This is critical to establish. Assuming whole otoliths were crushed and sub-sampled, d13C and d18O values reflect whole life conditions, likely averaged towards early life stages where otolith growth is fastest.

What mass of sample was analysed?  Presumably gas was evolved with phosphoric acid (temperature and reaction time?)

Line 100 (per mille not per mil)

Throughout, please use d1O ‘values’ and d13C ‘values’..

Line 200 – careful with language here – text implies a ‘response of zooplankton and fishes’ to salinity or temperature – but actually you show a response of biomineral d¹⁸O values – which implies that the animals are not responding to temperature /salinity (i.e. they remain in place and their minerals record the fluctuations in the water chemistry).

Line 205-215. I am a little uncomfortable with interpretations based on very small numbers of fish otolith data. For instance the shift in d18Ob.albini around 6.814 (if I understand the sampling correctly) is based on a single point / analysis?

Line 230 – I am happy with the idea of using d¹³Cotolith as a proxy for metabolic responses if the d¹³C of seawater is controlled. Why not report the difference between d13Cotolith and d13Cforam for each sample?