Mid-Cretaceous marine Os isotope evidence for heterogeneous cause of oceanic anoxic events

During the mid-Cretaceous, the Earth experienced several environmental perturbations, including an extremely warm climate and Oceanic Anoxic Events (OAEs). Submarine volcanic episodes associated with formation of large igneous provinces (LIPs) may have triggered these perturbations. The osmium isotopic ratio (187Os/188Os) is a suitable proxy for tracing hydrothermal activity associated with the LIPs formation, but 187Os/188Os data from the mid-Cretaceous are limited to short time intervals. Here we provide a continuous high-resolution marine 187Os/188Os record covering all mid-Cretaceous OAEs. Several OAEs (OAE1a, Wezel and Fallot events, and OAE2) correspond to unradiogenic 187Os/188Os shifts, suggesting that they were triggered by massive submarine volcanic episodes. However, minor OAEs (OAE1c and OAE1d), which do not show pronounced unradiogenic 187Os/188Os shifts, were likely caused by enhanced monsoonal activity. Because the subaerial LIPs volcanic episodes and Circum-Pacific volcanism correspond to the highest temperature and pCO2 during the mid-Cretaceous, they may have caused the hot mid-Cretaceous climate.

The authors have provided an improved version of their work which takes into account the suggestions of the reviewers. I have especially found improvements in the discussion which are now more complete. I do recommend the publication.
In the current version I spotted a few minor things reported as follows: lines 165 and 167 "represent" is repeated twice.
line 240 "is the more" I do not think this is correct in english, please check line 270 "Previous studies have revealed THAT the onsets of the major Cretaceous OAEs (OAE1a, Wezel, Fallot, and OAE2) in the Tethyan region correspond to…." The authors have addressed the most significant points raised in the reviews. I note some minor points in the text that need attention and some more substantive issues. Page 21: in their discussion of Cenomanian-Turonian temperatures the authors state that no longterm evidence for hydrothermal activity is recorded and they reference The manuscript is well written. The results are well summarized. The paper is well structured. Nice story! There are only a few points that I do not find so successful. The interpretations for the different intervals are not always consistent. For example, I would like you to look at the OAE 1b in a more differentiated way. I think there is more music in your data than you express. The statement that OAE 1b is to be regarded like OAE 2 or the Aptian OAEs is too simple. Jacob seems to be below the negative excursion. Kilian is characterized by a clear volcanogenic signal. For Leenhardt and Co the data is weak. I propose to discuss this time interval in more detail. Reply Thank you for many valuable comments. As you pointed out, OAE1b is composed of many organic-rich layers having different geochemical features. Thus, we have dealt with OAE1b separately in the different paragraphs and added a more detailed explanation as following "Among the mid-Cretaceous OAEs, OAE1b is a problematic example. In the Umbria-Marche Basin OAE1b is composed of several major organic-rich horizons (Jacob, Kilian, Urbino, and Leenhardt Levels)

1-2. Comment
Your statement that submarine volcanism does not contribute to the CO2 rise in the atmosphere is brave! Is there any evidence for this -perhaps references from recent submarine volcanism or modeling data? One should not let this statement stand alone and

undiscussed. Reply
Thank you for the comments. The volcanic events under the deep-sea condition were totally different from the subaerial eruption and the outgassing of volatile is suppressed by the hydrostatic pressures. We added the recent articles explaining the difference between the subaerial and submarine volcanism and added some explanations on this point as "When a basaltic plateau was emplaced under submarine conditions, outgassing from submarine volcanism and the expansion of the volatile to shallower waters could have been suppressed by high hydrostatic pressure59, and, thus, they may not have contributed to the long-term increase in the pCO2." (L. 363-366).

1-3. Comment
L.1: "This title is misleading a bit. Why not mentioning the first long term Os-Isotope record"

Reply
Thank you for the suggestion. We modified the title of this manuscript as "Mid-Cretaceous marine Os isotope stratigraphy: evolutional history of hydrothermal activity" (L. 1)

1-4. Comment
L. 140-142: "I am not sure this makes a lot of sense. Of course you may have a high percentage of CaCO3 based soleley on the presence of authigenic carbonates. Better delete as the sentence doesnt contribute to your story anyway" Reply L.146. Thank you for the suggestion. We removed this part.

1-5. Comment
L. 248: "Looks as if the Os-isotopes decrease happens after Niveau Jacob, right? I've never been a big fan of grouping the different horizons together one OAE 1b (and than differentiating between OAE 1a and Livello Wezel). Mayb you should try to tell a more differentiated story fpr the diffenet black shale levels based on your data -now.

Reply
As we explained above, we have discussed OAE1b in the different sections (L. 342-350).

1-6 Comment
L. 307-311: "I am not an expert of submarine volcanism and its influence on pCO2, but here I would like to see some quotes that confirm this theory!" Reply L. 317. Thank you for the comment. We added a detailed explanation and a reference of a recent article explaining the difference between the submarine and subaerial eruptions supporting our hypothesis as "When a basaltic plateau was emplaced under submarine conditions, outgassing from submarine volcanism and the expansion of the volatile to shallower waters could have been suppressed by high hydrostatic pressure 59 , and, thus, they may not have contributed to the long-term increase in the pCO2." (L. 363-366).

Reviewer #2
The manuscript by Matzumoto and co-authors provides the first compilation of Os187/Os188(i) record across the Late Barremian-Late Cenomanian time interval by combining many published data and new data collected for the late Albian-early Cenomanian.
The dataset is very good and represents an important reference for this time interval. The work has potential for publication. Please find below my general comments and some minor comments/suggestions in the pdf attached.

2-1. Comment
The main issue I see is that the authors can improve the discussion with respect to what we already know from other works about the paleoceanographic conditions occurred during this time interval. The discussion chapter lacks a deeper discussion with respect to literature especially regarding the ocean-atmosphere system evolution across the mid-Cretaceous and the studied OAEs. Reply Thank you for the comment. We added discussions on the detailed atmosphere and oceanographic conditions during OAEs and mid-Cretaceous and added some references as: (L. 270-292) "Previous studies have revealed the onsets of the major Cretaceous OAEs (OAE1a,Wezel,Fallot,and OAE2) in the Tethyan region correspond to unradiogenic Os isotopic shifts 13,16,19-21 , which  Therefore, the eruption style (e.g., submarine or subaerial) and its duration could have potentially influenced not only the temperature variations but also the diversity of calcareous planktons during the mid-Cretaceous.".

2-2. Comment
lines 248-257: this is not something new. Moreover, literature data do not support a "productivity" model to explain all the OAEs mentioned by the authors. I suggest to revise this part taking into account what is known from literature and, if possible, also trying to go a bit further with the interpretation since now the authors have a nice long-term curve of the Os ratio variations. Reply As you have indicated, the references do not fully support that these OAEs are productivity OAEs. Thus, we modified the meaning of this sentence to avoid misleading interpretation as following (L. 270-273) "Previous studies have revealed the onsets of the major Cretaceous OAEs (OAE1a,Wezel,Fallot,and OAE2) in the Tethyan region correspond to unradiogenic Os isotopic shifts 13,16,[19][20][21] , which is consistent with synchronicity between massive submarine volcanism and OAE.".
In addition, we added the more precise discussion on the Os behavior and its implication of the emission of the volatile elements and its relation to onset of OAEs as "During these OAEs, unradiogenic Os shifts are often accompanied by the negative carbon isotopic excursions [19][20][21]29 , implying the volcanic events supply mantle-derived CO2 with negative carbon isotopic values. Besides, 2 to 16 times increase in the input of mantle-derived Os are required to explain these unradiogenic Os isotopic shifts. Considering that Os could have been supplied in highly volatile oxidized form (OsO4), enormous amounts of other volatile trace metal elements could have been also injected into the ocean-atmosphere system during the most prominent unradiogenic Os isotopic shifts in these OAEs (OAE1a,Wezel and Fallot events,and OAE2). This possibility supports the linkage between biolimiting trace mental input and the high productivity 52 ." (L. 273-282).

2-3. Comment
lines 264-274: The new datasets confirms what has already been said in some published works about "OAE 1 c" and OAE 1d which were not induced by volcanic activity but rather by other imposed paleoclimatic and paleoecological conditions. I suggest to revise this part underlying this aspect and citing the papers who already said that on the basis of other proxies. Please, be also carful that, following some works, OAE 1d black shales were not marked by high productivity (see my comments in the text) Reply Thank you for the comment. We added the explanation on the previous studies and mentioned the productivity during OAE1d as following (L. 295-321) "Therefore, we consider that the onsets of OAE1c and OAE1d were unrelated to massive submarine volcanism, unlike other major mid-Cretaceous OAEs. Mercury anomaly has been reported just below the OAE1d horizon at the Youxia section, the eastern Tethys, which has been interpreted as the submarine volcanic eruption at the Kerguelen Plateau 53 .  (Fig. 3) 8,22 . During Quaternary, astronomically modulated monsoonal activity cyclically enhanced the hydrology of the Mediterranean Sea at low latitude, which supplied freshwater and nutrients to the peri-continental ocean. The resulting input of terrigenous organic matter, stratification, and slightly enhanced productivity led to oxygen-depleted bottom-water conditions and the deposition of organic-rich sediments dominated by the terrigenous origin [54][55] . Thus, the lack of the unradiogenic Os isotopic shift and the cyclic deposition of thin black shale layers during OAE1c and OAE1d may suggest a regional-scale weak marine anoxia caused by monsoonal activity modulated by astronomical cycles as proposed by previous studies 55 rather than an episodic large volcanic event 53 . The increase in the primary productivity was not significant at the Tethyan region during OAE1d 55 . However, a small positive carbon isotopic excursion during OAE1d suggests a slight increase in the primary production (Fig. 2). In addition, organic-rich sediments are reported from the Calera Limestone in California, which was deposited in the Pacific Ocean, and thus the oxygen-depleted condition could have prevailed in the East Pacific as well 14 . Thus, the latter process can also cause a supraregional increase in productivity to some extent.". Which time scale did you use? This is critical with respect to the position of the LIP ages.

2-4. Comment
Please specify and eventually discuss that in the text. Reply

2-13. Comment
L. 272-273: "be careful as Bornemann et al 2005 and Gambacorta et al 2020, for example, do not find high productivity during OAE 1d." Reply Thank you for the comments. As you have pointed out, there are few data suggesting the high productivity during OAE1d and1c. However, in case of OAE1d, the possibility of slightly enhanced productivity cannot be ruled out because it is accompanied by the slight positive carbon isotopic excursion and organic-rich sediments have been discovered in the Pacific region. Therefore, we modified "high productivity" to "slightly enhanced productivity" (L. 313) and added more detailed explanation on this point as (L. 298-324) "Therefore, we consider that the onsets of OAE1c and OAE1d were unrelated to massive submarine volcanism, unlike other major mid-Cretaceous OAEs. Mercury anomaly has been reported just below the OAE1d horizon at the Youxia section, the eastern Tethys, which has been interpreted as the submarine volcanic eruption at the Kerguelen Plateau 53 .

However, considering the lack of Os isotopic perturbations around OAE1d, this mercury enrichment is probably more related to local perturbation with limited influence on global climate. Major Cretaceous OAEs (OAE1a, Wezel, Fallot, and OAE2) are represented by thick organic-rich intervals, whereas the sedimentary expression of OAE1c and OAE1d in the Umbria-Marche Basin consist of cyclic alternations of thin black shales 8 . Similar cyclic intercalations of thin black shale layers in a carbonate sequence have been observed in the Valanginian-Barremian, Albian, and upper Cenomanian in the Umbria-
Marche Basin (Fig. 3) 8,22 . During Quaternary, astronomically modulated monsoonal activity cyclically enhanced the hydrology of the Mediterranean Sea at low latitude, which supplied freshwater and nutrients to the peri-continental ocean. The resulting input of terrigenous organic matter, stratification, and slightly enhanced productivity led to oxygen-depleted bottom-water conditions and the deposition of organic-rich sediments dominated by the terrigenous origin [54][55] . Thus, the lack of the unradiogenic Os isotopic shift and the cyclic deposition of thin black shale layers during OAE1c and OAE1d may suggest a regional-scale weak marine anoxia caused by monsoonal activity modulated by astronomical cycles as proposed by previous studies 55 rather than an episodic large volcanic event 53 . The increase in the primary productivity was not significant at the Tethyan region during OAE1d 55 . However, a small positive carbon isotopic excursion during OAE1d suggests a slight increase in the primary production (Fig. 2). In addition, organic-rich sediments are reported from the Calera Limestone in California, which was deposited in the Pacific Ocean, and thus the oxygen-depleted condition could have prevailed in the East Pacific as well 14 . Thus, the latter process can also cause a supraregional increase in productivity to some extent.". Besides, we added important references which seem to contradict our arguments and explanations on the points. For example, OAE1d was widespread and organic-rich sediments have been reported in the Pacific region. Thus, we added explanations on this point as (L. 318-324) "The increase in the primary productivity was not significant at the Tethyan region during OAE1d 55 . However, a small positive carbon isotopic excursion during OAE1d suggests a slight increase in the primary production (Fig. 2). In addition, organic-rich sediments are reported from the Calera Limestone in California, which was deposited in the Pacific Ocean, and thus the oxygen-depleted condition could have prevailed in the East Pacific as well 14 . Thus, the latter process can also cause a supraregional increase in productivity to some extent.".

2-14. Comment
We have also added some explanations on the sulfur isotopic fluctuations during OAE2 because it seems concordant to our hypothesis (L. 262-268): "δ 34 Sbarite data around OAE2 are scarce, but δ 34 S of pyrite and carbonate-associated sulfates (CAS) around OAE2 51 has been intensively investigated instead. δ 34 SCAS and δ 34 Spyrite showed a positive excursion (2-4‰) across the OAE2, suggesting an enhanced sulfate reduction 51 . Considering the global oceanic anoxia and short duration of unradiogenic Os isotopic shift during OAE2 (~600 kyr), the effect of the sulfate reduction could have overwhelmed the effect of volcanic sulfur input.".
Finally, we mentioned the paper discussing the volcanic events during OAE1d based on mercury as following (L. 298-304) "Therefore, we consider that the onsets of OAE1c and OAE1d were unrelated to massive submarine volcanism, unlike other major mid-Cretaceous OAEs. Mercury anomaly has been reported just below the OAE1d horizon at the Youxia section, the eastern Tethys, which has been interpreted as the submarine volcanic eruption at the Kerguelen Plateau 53 . However, considering the lack of Os isotopic perturbations around OAE1d, this mercury enrichment is probably more related to local perturbation with limited influence on global climate.".

3-4. Comment
Line 57: Robinson et al. recognize an OAE1d black shale in the Calera Limestone of California, which derives from the paleo-Pacific (Geological Society of America Bulletin, 120, pp.1416-1426). Hence, the reach of this event may be greater than just Tethys/Atlantic. The Pacific also has a record of OAE 1a and OAE 2 black shales that reflect major events. How does this Calera Limestone occurrence affect the interpretation of OAE 1d as monsoon-related? Note also the new mercury evidence bearing on OAE 1d: Yao, H., Chen, X., Yin, R., Grasby, S.E., Weissert, H., Gu, X. and Wang, C., 2021.
Mercury evidence of intense volcanism preceded oceanic anoxic event 1d. Geophysical Research Letters, 48, p.e2020GL091508. Some reconsideration is warranted here, even if the absence of an Os-isotope signal may be telling us something. Reply Thank you for the constructive comments. We have added an explanation of the Pacific record of OAE1d in the Introduction as (L. 57 to 59) "Additionally, other minor OAEs (e.g.,OAE1b,OAE1c,and OAE1d), which have been reported mainly from the Tethys and Atlantic Oceans 9,11-13 and a part of the Pacific region 14 , are regarded as regional to supraregional marine anoxic events." and discussion section as (L.318-324) "The increase in the primary productivity was not significant at the Tethyan region during OAE1d 55 .
However, a small positive carbon isotopic excursion during OAE1d suggests a slight increase in the primary production (Fig. 2). In addition, organic-rich sediments are reported from the Calera Limestone in California, which was deposited in the Pacific Ocean, and thus the oxygen-depleted condition could have prevailed in the East Pacific as well 14 . Thus, the latter process can also cause a supra-regional increase in productivity to some extent.".
Besides, we added the discussion of mercury anomaly in the main text. Considering the studied site was located near the Kerguelen Plateau, this mercury anomaly could represent the local small volcanic events at the Kerguelen Plateau. We added the explanation as following (L. 299-304) "Mercury anomaly has been reported just below the OAE1d horizon at the Youxia section, the eastern Tethys, which has been interpreted as the submarine volcanic eruption at the Kerguelen Plateau 53 . However, considering the lack of Os isotopic perturbations around OAE1d, this mercury enrichment is probably more related to local perturbation with limited influence on global climate.".

3-5. Comment
Line 104: I agree that considering OAE 1b as a cluster of events is the most useful approach. The Kilian is now taken to mark the Aptian-Albian boundary (GSSP: see Kennedy et al.,Episodes,vol. 40).

Reply
Thank you for the comments. We tried to add the explanation and citation on the Aptian-Albian boundary at this point. However, the description of the just AAB seems not important in our paper. Besides, the number of references is limited, we gave up citing the reference here.

3-6. Comment
Line 107: it is not clear to me what 'partially records OAE 1c' means. How is (or was) OAE 1c defined? Maybe go back to Arthur, M.A., et al., 1990. Stratigraphy, geochemistry, and paleoceanography of organic carbon-rich Cretaceous sequences. In Cretaceous resources, events and rhythms (pp. 75-119). -who suggested the OAE nomenclature.

Reply
We modified the explanation and reference of Amadeus segment and OAE1c in the main text as (L.111-114) "A peculiar ~2-m-thick interval in the upper Albian, called the Amadeus segment 24 , is located in the middle part of OAE1c that spans almost the entire Biticinella breggiensis planktonic foraminiferal Zone 8,25 .".

3-7. Comment
Line 132: Reference 9 does not discuss carbon-isotope data. Refer to the paper of Gambacorta et al. (Newsletters on Stratigraphy,48, for carbon-isotope patterns in the Upper Cretaceous of Marche-Umbria.
Reply L.138. We added the reference here.

3-8. Comment
Line 184: I could not find a reference to modelling of weathering of Ontong Java in Tejada et al., although they suggest various types of basalt-seawater interaction to produce the Os-isotope excursion . Presumably warm-to low-temperature submarine weathering of basalt could be part of this. Reply Thank you for the comment. As you have indicated, the hydrothermal activity includes the weathering under the submarine condition though the estimation of its effect is unclear at present. We modified the discussion as (L. 174-179) "these unradiogenic Os isotopic shifts correspond to the radiometric ages of the Ontong Java, Manihiki, and Hikurangi Plateaus, which once formed a single large oceanic plateau called "Ontong Java Nui" (OJN) (Fig. 3a, e), these unradiogenic Os isotopic shifts were likely triggered by a massive input of mantle-derived unradiogenic Os through hydrothermal activity and warm-and low-temperature submarine weathering at OJN 13,20,21,34 . " .

Reply
Line 194: We modified here to "during the early Albian"

3-12. Comment
Line 217: in what way is the geochemical behaviour of Os different from that of Sr in this context?

Reply
When Os was oxidized, it is volatile and easy to move. This effect during the LIPs eruption is a very critical factor to determine the Os behavior during the eruption. We added the explanation in the main text as (L. 222) "and the volatile feature of highly oxidized form of OsO4.".

3-13. Comment
Line 228: presumably you exclude the possibility that the positive sulphur-isotope excursion was due to increased pyrite burial in black shales?? As illustrated in Fig. 3c, most of the increase in S-isotope values took place in the Cenomanian, post-dating OAE 1d; in so doing, S isotopes rise in parallel with established carbon-isotope curves (e,g. Jarvis et al. for the English Chalk) likely signifying gradually increasing global marine carbon burial in the run-up to OAE 2. The alternative to the idea of a decrease in hydrothermal input of sulfur, namely that the data reflect synchronous increased burial of isotopically light carbon and sulphur, needs to be discussed.

Reply
Thank you for the comment. We checked the carbon isotopic excursions during Cenomanian-Turonian at the English Chalk (Jarvis et al., 2006 Geological Magazine).
The onset of the positive carbon isotopic excursion occurs after the positive sulfur isotopic excursion at the English Chalk. Since the residence time of sulfur is much longer than carbon, the fact that carbon isotopic excursion postdates the sulfur isotopic ratio is hard to explain. Besides, in the Umbria-Marche Basin, abundant organic-rich sediments are observed during the Albian and the latest Cenomanian after the mid-Cenomanian event, which does not match the timing of sulfur isotopic excursion. Therefore, we consider that the cease in the volcanic sulfur is the most likely candidate for triggering the positive sulfur isotopic excursion as suggested in Laakso et al. (2020). The time lag between Os and sulfur isotopic ratio is derived from the large differences in the residence time.
Because of the extremely long residence time of sulfur compared to Os, the sulfur isotopic ratio may have postdated the Os isotopic curve. Also, the sulfur isotopic data during Albian is scarce, and more detailed data will be required for further discussions on this point. We have explained and discussed in the main text as following (L.235-244) "The positive excursion of δ 34 Sbarite can be also explained by an increase in sulfur reduction during the early Cenomanian. However, considering the organic-rich sediments are more pronounced during Albian than Cenomanian at the Umbria-Marche Basin, sulfate reduction should also have been more significant during Albian than Cenomanian.
Therefore, we consider that the decrease in the volcanic sulfur is the more important factor for the positive δ 34 S excursion rather than the sulfate reduction. The positive excursion of δ 34 Sbarite during the Cenomanian postdates the cease of the Os isotopic fluctuations (Fig. 3a,c). Since the residence time of sulfur in the ocean is longer than Os, the onset of the changes of δ 34 Sbarite could have been more gradual and possibly postdated the radiogenic Os isotopic shift.".

3-14. Comment
Line 244: Sulfur-isotope evolution around OAE 2 (data from the English Chalk and the