The authors present a data set of sedimentary alkenones, retrieved from continental margin and drift sites that is of interest because it contains combined proxy measurements for sea surface temperatures (SST) and productivity with size-fraction analysis and 14C dating. It is argued that the alkenone signals from the bulk sediment are altered depending on the size-fraction distribution of the sediment. Each grain size is affected by different hydrodynamic processes and exposure to oxic degradation: alkenones in clay and fine silt are affected by lateral transport, coarse sand alkenones are more prone to oxic degradation. This finding is important because it has implications for the use of alkenones as a paleotemperature and paleoproductivity proxy.

The manuscript is well written, the references are up-to-date and the data support the conclusions. I think the paper should be published in Biogeosciences after revisions. Although my suggested changes are minor, they are important, especially regarding the figures. The reader would profit from a more consistent naming of the axes and more descriptive figure captions.

 Specific comments:

Line 116: Table 1. the cores are in very different water depths, can you acknowledge that somewhere in the discussion, especially knowing that many deep-time paleotemperature and -productivity records come from open ocean deep-sea sediments.

Line 127: for the quantification, what standard and at what concentration was used? Did you run replicates?

Line 129: Mention the exact variables that were used for the Bayspline code (e.g. prior standard deviation). Also state measurement/calibration error. Is it possible to show this error in figure 5?

Line 152: Even though this study seems to have better recoveries than other previous studies using wet-sieves (85%, Magill et al., 2018, Tesi et al., 2016), how can you be sure that there wont be any preferential loss of C37:3 or C37:2, which could ultimately affect your SST results? Could you comment on that?

Line 153: Can you explain the significant loss of alkenones during manual columns? Instead of comparing the fraction-weighted values to the bulk values did you also use an internal standard to account for the loss/recovery of alkenones?

Line 238: Can you state if you recorded C37:4 in BER, which also reflects advection of high-latitude alkenones (Rosell-Melé et al., 1998; Bendle and Rosell-Melé, 2004; Bendle et al., 2005)?

Minor comments/ technical corrections:

Line 16: spell out OM and SST as first mention in manuscript

Line 28:(C37:2 + C37:2) should be (C37:2 + C37:3)

Line 33: also add latitude-dependent (Lutz et al., 2007)

Line 42: define 37:2, 37:3, e.g. UK37′=[C37:2]/([C37:2]+[C37:3]), where [C37:2] and [C37:3] are the concentrations of di- and triunsaturated C37 alkenones, respectively.

Line 44: reported precision in Prahl and Wakeham, 1987 is ±0.5°C

Line 55: wrong format for reference

Line 65: which hydrodynamic processes are we talking about? Maybe give an example in brackets?

Line 68: define OC

Line 95-103: for consistency, add oxic/suboxic conditions to NAT, NAM and BER in text.

Line 109: Figure 1: Out of curiosity: why was SHAK06 not described in this study? It was analyzed in Ausin et al., 2021 for OC in bulk sediment on continental margin sediments

Line 114: add abbreviation to clay, similar to fine silt and coarse silt, which is later used in figures and tables.

Line 115: The author refers in Ausín et al., 2021 to Magill et al., 2018 for the method. Might be worth mentioning here as well.

Line 168: Figure caption 2: Figure 2. C) should be C37:3 and D) should be C37:2

Define abbreviations for CS, FS, C

Line 171: briefly mention, why alkenone 14C ages in SBB, NAF, BER were not measured

Line 185: Figure 3 would be easier to read if you would change TOC-foraminifera to another symbol. It currently might be confusing because you use open symbols in panel A for bomb 14C.

Line 190: Figure 4: Define abbreviations for CS, FS, C

Line 192: Fig.5. to be consistent with the naming in figure 6C, rename y-axis in panel B “ΔSST_{bulk-annual mean}”

Line 275: Figure 7: In panel A, x-axis label is missing. To be consistent with the nomenclature of your manuscript rename label for y-axis to OC_%. Also, insert is hard to read, maybe its enough to state the R² from Ausín et al., 2021 in the text, instead of having the insert. In panel B, add C37 to y-axis to be consistent with naming (in comparison to Fig. 2A)

Review of ‚The influence of lateral transport on sedimentary paleoproxy signals’ by Ausín et al.

Ausín et al. report OC and alkenone contents as well as corresponding 14C ages in grain-size fractions from several surface sediment samples from continental margin sites and compare these to ages of planktonic foraminifera and alkenone-based SST estimates. The SST estimates from the grain-size fractions in association with their 14C ages are used to assess the influence of lateral transport on alkenone SST estimates in bulk samples. This new dataset has value as it presents SST estimates and 14C ages of grainsize fractions to shed more light into the processes of grain-size specific lateral advection effects on proxy signals in sediment archives. I recommend publication after major revision.

Major comments:

Concept vs new findings: The concept of the influence of lateral transport on the age and SST estimates of alkenones is more than 20 years old (see Benthien & Müller, 2000 and diverse follow-up papers). Now this manuscript presents grain-size specific SST and 14C data which support this concept and point to the fine silt fraction as being mainly responsible for the observed effects. My main concern with the Ausín et al. paper is the way this is presented. The history of this concept is described in the introduction but it nevertheless reads as if the whole concept is new and Ausín et al. are indeed the first to present the concept. I am referring to sentences like: “Our results demonstrate that selective association of alkenones with mineral surfaces and associated hydrodynamic mineral sorting processes can alter alkenone signals encoded in marine sediments (14C age, content, and distribution) and confound corresponding proxy records (productivity and SST) in the spatial and temporal domain.”. In this last sentence of the abstract, the first half of the sentence reads as if this was not known before. In contrast, this has been clear before and is just supported by the new data. The second half of the sentence, in contrast, is just an inference so far. It is logically to expect these effects but it is not actually demonstrated in the manuscript. Another example is the last sentence of the conclusions: “Our results highlight the importance of considering the influence of hydrodynamic processes (e.g., lateral transport) on sedimentary alkenone signatures (amount, age, and temperature) and their relationship to surface waters overlying the depositional location.”. Since the works from Mollenhauer, Ohkouchi and others from decades ago it is known that the influence of hydrodynamic sorting on organic paleo-signals in sediments has to be considered. Such sentences thus read like ‘constructing a strawman’ to oversell own findings. That said, I think the new grain-size specific data in itself carry enough value to be reported and the manuscript does not benefit from overselling and should be adjusted to be toned down.

Biological oceanography: Another point which seems to be wrong is the interpretation of alkenones being indicative for highest productivity in the Peruvian upwelling system used for explaining a warm bias (line 323 to line 325). This cannot be true as upwelling activity is driven by trade wind strength which is highest in austral winter leading to deep Ekman pumping which brings dissolved Si into the surface waters causing outcompeting of haptophytes by diatoms. The warm bias thus likely arises from the fact that alkenones are actually not produced during strongest upwelling, the latter associated with lowest SST.

Errors and precision: Please state what the analytical and propagated errors are of compound quantifications and SST estimates. For instance, in table 2 alkenone concentrations (C37:2 and C37:3 and combined) are reported to the last digit. Is this reasonable with a usual GC-FID error of at least 10% for long-chain alkenone quantification? How does this error propagate to the UK37’-SST estimates? I guess that with a good error handling many of the reported ‘biases’ will actually be within error and only a few significant offsets will remain. Also, error bars should be added to all plots showing SST estimates and, preferentially, also instrumental SST data.

Minor comments:

Generally, please replace ‘warmer bias’ by ‘warm bias’

Line 3: Please check affiliations. Bruni and Eglinton are not in Salamanca.

Line 9: “…gaps remain on…” - consider wording

Line 29: Are alkenones really a large component of total OC of Emiliania huxleyi? Please check, I doubt this statement.

Line 106: “…in contrast with…” – consider wording ‘contrast to’

Line 113: how was the grain-size fractionation done? Wet or dry sieving?

Line 201: I would presume that the statement that only SST estimates from SBB and SMB are comparable to atlas data is not true when considering errors. Avoid arguing with ‘comparability’ and argue with errors instead.

Line 204: I doubt that the statement of a general warm bias is actually true when considering uncertainties. Looks like only true for 2 out of 7 samples.

Line 208: The statement that FS overall shows the smallest temperature offsets with bulk is not true. Please look at the data from NAT. I do not understand the significance of the following statement on larger/smaller offsets. Consider removing.

Line 318-321: In upwelling areas coccolithophores are outcompeted by diatoms during strong upwelling. See major comment above.

Line 380: should read SMB

Tables:

Table 1: Namibian core – MC or BC?

Table 2: see comment on precision of data given.

Figures:

Figure 2: Please add comment in caption about sand fraction in BER

Figure 4: I do not see the significance of panel B, just another representation of the same data (also no reference to B in caption).

Figure 5: panel B: see comment on errors. I suspect that data from SBB, SMB, NAM and NAF are actually within error with instrumental SST. Preferentially add error bars to SST estimates and instrumental SST data.

Figure 6: captions for panel B and C are mixed up.

Figure 8: What is the significance of the solid regression line considering all data. Consider removing.