In this manuscript, Karatsolis and Henderiks generated two calcareous nannofossil records from International Ocean Discovery Program (IODP) Sites U1463 and U1464, located in the NW Australian shelf, in order to reconstruct long-term changes in ocean circulation, seasonality and nutrient availability from ~ 6-3.5 million years ago (Ma). The authors characterised different periods of change in stratification and nutrient availability in the study area by analysing the shifts in the calcareous nannofossil dominant taxa and comparing them with palaeotemperature gradients between the NW Australian shelf and the eastern Indian Ocean.

Karatsolis and Henderiks found a marked regional change in the oceanographic conditions that affected the ecology of calcareous nannofossils across the Miocene to Pliocene boundary (5.4–5.2 Ma), which they attributed to an increase in seasonality and general intensification of the upper water column mixing. The authors also put the observed local variations in a more global context, considering events, such as the extinction of Sphenolithus spp. (~3.54 Ma) and the termination of the late Miocene to early Pliocene biogenic bloom in the eastern Indian Ocean (4.6-4.4 Ma).

General comments

This manuscript represents a substantial contribution to scientific progress within the scope of Climate of the Past and it is of interest for the coccolithophore, calcareous nannofossil, palaeoceanographic and micropalaeontological communities.

It is well written, logically structured, and presents a new calcareous nannofossil dataset.

The title reflects the contents of the manuscript and Karatsolis & Henderiks present an adequate summary of their work in the abstract. The state of the art and the main aims of this work are properly introduced in the first section.

The methods used in this piece of research seem adequate and are described section 2 of the manuscript. In my opinion, mathematical formulae, symbols, abbreviations and units are correctly defined and used through the text.

The interpretation and conclusions have been logically derived from their findings, and supported by the original data shown in section 3 (Results).

My main concern is that the authors should highlight more the variability between proxy types. They use different proxies, such as GDGT-based TEX86 temperature and alkenone-based U37k´ SST or Mg/Ca derived SSTs from Trilobatus sacculifer. I like that they use gradients, but in the manuscript, the uncertainty of working with different paleotemperature indicators need to be more clearly addressed. Perhaps adding some reference(s).

Figures and tables are in general clear (a very good example of this is Figure 7). I just have minor suggestions regarding the figures (see specific comments/technical corrections). I would recommend merging some of them (4 and 6).

I find that the references cited in the manuscript are adequate. I just found some typos.  

The supplementary material is also adequate, but it could be improved (in some of the plots there are just too many wiggles). I would probably move some of the supplementary figures/plates to the main manuscript (see specific comments).

The Karastolis and Hendericks study is an interesting contribution in a region and period that need to improve knowledge, particularly.

The use of CN offers a unique opportunity to characterize surface water masses and to monitor their evolution during the Mio-Pliocene transition interval.

The state of the art is well stablished and objectives of interest.

The chosen technique is appropriate, based on previous initiatives. The use of sphenoliths as main taxonomic group linked to stratification is fine, although should be more correct to include too discoasters (although its proportion is low). Others species clold be also considered in the interpretation. In fact there are some that are present in the plots but not mentioned/considered. If they are not taken into account, there is no point in including them.

One aspect repeatedly used is seasonality. In the text it is discussed that with this resolution a direct correspondence with present day conditions it is not feasible.

In this sense, should be more correct talking about persistent conditions (e.g. stratification), rather than enhance seasonality.

In fact, the signal that is manifested is the dominant one in a sufficiently broad period to refer its dynamics to seasonality (although it is obvious that these processes may be the triggers), but there are not enough arguments to focus these changes on seasonal variation only.

I suggest a modification in this sense, avoiding such a direct reference to seasonality.

Other aspect that must be considered in Discussion and Conclusions, is provide more clear information about the ITF and LC, marked as objective, trying to ling the signal observed with characteristics of LC (for example). In fact this is considered, but plots should be explained in this sense, in order to follow better the arguments. Discussion refers the features in a general way without focusing on the evolution of those water masses. In this sense, I cannot see clearly the utility of the H Index (not linked to these features) and the CA (also superfluous. With the available data (T) and stratification index, should be possible to identify/define the involved water-masses, and consequently its evolution along the considered period.

Concerning the potential mechanisms related to paleoenvironmental aspects (3 options), the authors should consider the most reasonable possibility, taking into account the available data.

The comments referred in line 395 is too speculative: should provide extra information to propose this mechanism, or afford the explanation in a more general way (mixing!).

Section 4.3 refer an interesting global feature. Here is considered lightly, being necessary a better explanation of the processes and records. The link with the rest of the text is not clear, need better justification.