Revision of manuscript “ Anoxia and salinity changes : a new Permian catastrophe record ”

This paper by Cassel et al. focuses on the early Permian Irati Formation in Brazil. Authors try to determine and constrain potential anoxic episodes in several cores from this basin, and to link these events to other major known bio or geological Permian events known elsewhere. At a first view, the used dataset and analyses appear numerous and some interpretations valid. Some obtained information are maybe also important although local. Overall, I was highly disconcerted by the form of the text and the discussion/ conclusion included in this work. In my opinion, the paper requires numerous significant modifications and additions. To be brief:

view, the used dataset and analyses appear numerous and some interpretations valid. Some obtained information are maybe also important although local. Overall, I was highly disconcerted by the form of the text and the discussion/ conclusion included in this work. In my opinion, the paper requires numerous significant modifications and additions. To be brief: 1-I fully understand that it is very difficult for some authors to write in English, but this text is not at all up to the level of a standard publication. There are too many mistakes and typos pervading the ms (thus I did not list them hereafter). It took me a very long time to read it (several times) and (try to) to understand ideas contained in many paragraphs and sentences. It is necessary that authors have their text corrected by a native-English speaker because, as it stands, very few passages are simply understandable by the reader.
AC: The manuscript is currently with a native English proofreader.
Regarding the distribution of data and whether they are regionally representative, we have made an additional figure that demonstrates the regional extent of the described facies (see Figure 4 at the end of the document).
2-Title, discussion and conclusion focus on major Permian extinction events, such as during the latest Permian, and link them to major geological events, such as the Siberian traps. However, the Irati Formation is not coeval to any of these events and thus, this link made by the authors is not understandable. None of the data shown in this work can be linked to any mass extinction. This is interesting to document that the Irati Formation potentially records anoxic conditions but they are local or regional and should be discussed with coeval events worldwide (if identified). Authors should also provide more biostratigraphical data allowing to constrain the age of this formation.
Overall, the entire structure of the paper has to be changed.
We do not link to siberian traps. There are other trigger possibilities for of the Permian extinction beyond the Siberian traps. There pulses of environmental disasters causing stress to the previous P-T boundary. The data brought here may instead be linked to these catastrophic Permian episodes, which in turn may have caused the deaths. The data shown in this study, as answered in the previous item, are of regional magnitude. Additional dating data were cited in all figures that mention any stratigraphic units.
The mass extinction that occurred at the end of the Permian is the most severe biotic crisis in Earth's history. Numerous papers report that this event was responsible for decimating more than 90% of marine species and about 70% of continental vertebrate families (Erwin, 1994;Retallack, 1995;Knoll et al., 1996;Knol et al., 2007) . The pattern of disappearance of these species is quite complex, some species disappeared before and others after the Permo-Triassic (P-T) limit.
Permian cause-of-death models demonstrate that the extinction of individuals occurred due to hypercapnia, hypoxia and oceanic acidification due to generalized anoxic and euxinic conditions, Some authors point out that anoxia occurred in much shallower waters than previously recognized and show that sections of high latitude boreal oceans were also affected by the anoxic event.
Some models indicate that the Permian extinction, as well as the Neoproterozoic, occurred through the repeated occurrence of anomalous conditions. Thus, such a model predicts that high-resolution stratigraphy within basins will show rapid pulsed extinctions, with the latest clustered appearances associated with C isotopic excursions and other evidence for anoxic deep ocean turnover (Knoll et al, 1996;Knoll et al , 2007). This model is in line with Wignall and Twitchett (1996) observation that Permian extinction can be a considerably more complex event, involving an initial extinction separated from the most recent mass slaughter of the late Permian. Therefore, it was not a single extinction event that occurred at ~ 252.3 My, but rather extinction pulses (Song et al., 2013). In addition, environmental (and catastrophic) stress events would have preceded the major extinction ( Knoll et al., 2007;Ward et al., 2005). An example is the extinction that occurred about 10 Myr before the great extinction, preserved in the geological record between the Guadalupian and the Lopingian (Stanley and Yang, 1994). Already in Kungurian there are radical episodes of anoxia. These are explained by climate trends between the tropics and Gondwanna, related to the Late Paleozoic Icehouse-greenhouse transition. This phenomenon caused oceanic stagnation and consequent anoxia (Liu et al 2017).
Thus, the Irati Formation presents these anomalous characteristics typical of the Permian catastrophic environmental crises and is characterized as another record of this singular period in the Earth history. Such a relation had not yet been reported in this way, especially once in Gondwana.
3-Authors claimed that 23 cores were studied but data from only two sites are presented. Authors should provide data from these other sites with their corresponding environments as the reader can follow interpreted environmental fluctuations in space and time.
AC: To answer this request, additional figures were made to clarify the correlation of the described facies in all cores (see Figure 4 at  Overall, mainly based on the poor quality of the English and of the false links made between observed local/regional environmental changes and Permian extinctions, requiring a complete rewriting of the paper, I recommend to reject the paper. AC: We greatly appreciate the review and suggestions we accepted. As previously written, the text is already is currently with a native English proofreader, and we hope to have explained more clearly that this is not a false linkage between the Irati Formation and the Permian catastrophes, which are triggers for the extinction.
Below are the figures with the modifications made and new figures added as requested by the referee.