Interactive comment on “ Role of the North Atlantic circulation in the mid-Pleistocene transition ”

General Comments Martin-Garcia use foraminiferal assemblages from a drill core in the eastern subtropical Atlantic (U1385) to reconstruct glacial/interglacial variations in surface hydrography during the latter stages of the mid-Pleistocene transition (MPT) (Marine Isotope Stages 20-14, ∼800-530Ka). Their overall conclusion, as stated in the abstract, is that beginning with MIS 16, there is an increase in poleward warm surface water transport in the North Atlantic Current, which would act as a positive feedback prolonging the more extreme, 100 kyr-paced glacial maxima.

In the context of Figure 3, why not show the N. pachy counts from Site 607? Interpretations regarding heat transport are based on spatial thermal gradients, yet none of the ï ň Ągures show such gradients.The reader is asked to ï ň Ągure this out from the SST records shown in Figure 4.It is also really difï ň Ącult to follow the argumentation in the discussion because statements are not followed-up with appropriate call-outs to ï ň Ągures.The new Figure 3 includes the N. pachyderma sin record from Site 607 (see Fig. 3c,yellow graphic).In this way, comparison with sites 980 and U1385 is clearer.Both latitudinal and longitudinal thermal gradients have been calculated for the North Atlantic, using data from the studied sites.The estimation of such gradients is described in the Methods section, and the gradients themselves are included in Figure 3g.To better highlight the thermal variation along the time series, the statistical mean has been calculated for each MIS, in both latitudinal and longitudinal gradients, and represented in the same figure.Call-outs to figures have been corrected in the text.
There are a few statements in the text that seem to go against what is generally known about deep water circulation on glacial/interglacial time scales.For example, in the abstract the authors imply that NADW is strong during MIS 16 (lines 20-27)?To my knowledge, and shown in Figure 3b, the relative ï ň Ćux of NADW increased during the deglaciation.So perhaps this is just a matter of carefully rewording the pertaining sentences.There are numerous other instances in the text where the wording of the sentences does not clearly communicate the message (see details below).Following suggestion, the text has been changed as follows: ". ..and the increase in the North Atlantic Deep Water (NADW) formation respect to previous glacials" Interpreting changes in percentages is complicated by the fact that an increase in one species results in an apparent decrease in another, when, in fact, there may not be a change at all in the accumulation of the latter species.The authors should address this so-called 'closed sum' problem.It is clear that the closed sum effect exists, but there is no better way to show the results about the planktonic foraminifer assemblages.Sev-C2 eral authors (e.g., Bé, 1977;Ottens, 1991;) have studied present-day North Atlantic water masses and identified the dominant planktonic foraminiferal species (in percentages) for each of them.In the same way, fossil assemblages have been associated to specific water masses (e.g., Cayre et al., 1999;Vautravers et al., 2004;Salgueiro et al., 2008) SST reconstructions are also based on assemblage's composition (measured in percentages) Regarding the description of deep-water mass changes, I suggest rewording the sentences to make it clear that it is the relative ï ň Ćuxes of NADW and AABW that are changing.The text has been changed as follows: ". ..mid-latitudeNorth Atlantic sites registered a relative decrease of the AABW during glacials, and subtropical sites recorded the presence of NADW at depths previously occupied by the AABW" How do these results compare with Alonso-Garcia et al. ( 2011) speciï ň Ącally?The time intervals of study overlap, so there is potential to make more of this comparison.Or, are the interpretations of the shifting fronts based on their ï ň Ąndings?In this case the study should be cited in the discussion.
Both Alonso-Garcia et al ( 2011), and Hernandez-Almeida et al ( 2013) studied site U1314, situated too north-westward for being useful in the study of variations of the NAC through glacials.This site, as well as others located northward 980 -like 984, studied by Wright and Flower (2002) together with the 980, register advances of the AF very early in glacials, both before and after the MPT.Particularly, site U1314 was compared to U1385 in Martin-Garcia et al. (2015), and SST differences between both sites, studied for the interval 780-490 ka.This study demonstrated that the NAC did not reach site U1314 since glacial inceptions, both before and after MIS16.Site 980, on the other hand, lies in the path of the NAC and thus, at a key location to register both advances of the AF and presence of the NAC during glacials, as can be observed in Fig. 3.

C3 Speciï ň Ąc Comments
Lines 59-62: include Alonso-Garcia et al 2011 in the list of citations?These lines refer to the mid-latitude NAtlantic, not to the subpolar one, which is why this citation has not been included Line 64: Alonso-Garcia's record begins with MIS 19.Therefore, it is no entirely appropriate to cite their study in the context of something that "began" during MIS 21?This citation has been removed Line 66: Why abbreviate the reference to Wright and Flower (2002) withW&F02?None of the other citations are abbreviated.The text has been changed as suggested Line 92: "to obtain an conclusion" seems awkward.Perhaps replace with: to reach basin-wide conclusions?Or to obtain a basin-wide picture/view/reconstruction?The text has been changed as suggested: "reach basin-wide conclusions" Line 96-97: Awkward sentence.Do you mean that the records extend far back into the past, or that they have been studied for a long time?The text has been changed: "for paleoclimate and oceanographic research on the Quaternary" Line 134: "generally present" is vague.Figure 2b shows that N. pachy are present throughout the entire study interval, but their relative abundance increases during these glacial intervals.I suggest specifying what percentages are considered signiï ň Ącant and why.For example, there is also a peak during MIS 15.The new Results section explains the variation of this species through the time series, comparing its relative abundances during glacials/interglacials, and also the occurrence of peak percentages Line 149-151: This reads as if you are implying that MIS 20 is an interglacial interval.The text has been changed: ". ..even more abundant than during interglacials, like in MIS20, when it reaches the highest percentages of the whole study interval".Line 156: I would suggest changing the section heading to specify that the focus is on MIS 20 and MIS 18 The new heading is: "5.1 North Atlantic circulation during glacials C4 MIS20 and MIS18" Line 169-170: The sentence needs a speciï ň Ąc ï ň Ągure call-out.I found the info in Figure 4c and d.The figure call-out has been added.The information is now in the new Fig.3f Line 181: Vague: What is the difference between very low and relatively low?And, it is confusing to read about low ice volume in the context of glacial intervals (MIS 20 and 18).This sentence has been removed.
Line 192: Deï ň Ąne what the thermal gradient is.What does it mean when it is negative in terms of the temperature difference between the sites?Once this is established, it is easier to follow the interpretation with respect to heat transport.The method to calculate the thermal gradient is now fully explained in the "Materials and Methods" section ("2.3.Estimation of thermal gradients").This section also explains the meaning of a positive and a negative gradient between sites.
Line 214: I am not sure that I see that the thermal gradient was signiï ň Ącantly different during MIS 18 from MIS 16.This is true only for some intervals of time, but not consistently.For example, the same SSTs are recorded by the sites during MIS 16 at ∼640-650 Ka.In any case, signiï ň Ącance, which is a statistical term, is not demonstrated in this data set.The ambiguous term has been changed.The new Fig.3g, includes thermal gradients.As the average value has been calculated separately for each stage, it is easier to see that the latitudinal thermal gradient in the NAtlantic was higher during MIS16, and MIS14, than during the whole interval MIS20-MIS18.
Line 220: It is really difï ň Ącult to follow how these records show a negative thermal gradient.Would it be possible to just calculate the SST difference between the records to support this point?Thermal gradients have been calculated between the records, and represented in figure 3g Line 226: I am not sure I detect a repeating pattern in the data set.MIS 14 has quite C5 a bit of variability, as you point out, so which pattern are you referring to?The text has been modified: "While in the older glacials SST decreased towards glacial maxima, this trend is not observed during MIS16 and MIS14, and warm SST was recorded also during glacial maxima".
Line 251: Is this correct?Do these studies really show that AABW is reduced during glacial intervals?There is a study by Lang et al., 2016 (Nature Geosciences) that shows % NADW for the past 3 million years.They show that NADW goes to zero, or almost zero during glacial intervals beginning around 0.9 Ma.We are comparing conditions during glacials.It is proved that there is and increasing trend in the NADW formation rate since MIS22, but it is during glacials that, the difference in the AMOC rate influences the mass of water present in the deep mid-latitude North Atlantic.The text has been changed to explain this better: ". ..data from the sub-polar North Atlantic (Wright and Flower, 2002;Hodell et al., 2008) document a long-term increase in the NADW formation rate, that initiated in MIS22 and culminated in MIS14.This enhanced the southward flux of the NADW and, since MIS17, mid-latitude North Atlantic sites registered a relative decrease of the AABW during glacials, and subtropical sites recorded the presence of NADW at depths previously occupied by the AABW (e.g., Poirier and Billups, 2014;Hodell et al., 2015)".
Technical Comments The following is an incomplete list of editorial-type ï ň Ąxes.Line 23: "At" the surface Line 30: Blocking Line 68: during interglacial periods Line 69: related "to" Line 86: "...which makes it an ideal location: : :." Line 99 meters: : :.At the surface: : :; at depth: : :.Line 122: on average Line 123: commas before and after 1 cm thick?Line 131: associated with Line 141: replace 'since then' with 'after' Line 215: higher All the type fixes indicated by reviewer 1 have been taken into account.