the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Aug 2026
Why does the Northwest Pacific Ocean (Kamchatka Margin) have a different carbonate time series than the rest of the Pacific? Regional carbonate production is greater than deep Pacific dissolution at the start of interglacials
Abstract. Pleistocene records of sedimentary carbonate from most of the Pacific Ocean have a strong 100-ky cyclicity, attributed for the most part to glacial-interglacial variation in carbonate dissolution. Pleistocene glacial intervals throughout the Pacific Ocean have high CaCO3 burial and preservation and accelerated dissolution in interglacial intervals. Records from the equatorial Pacific, Ontong Java Plateau, Shatsky Rise, and along the California continental margin all have this pattern, attributed to changes in corrosiveness of deep waters in the Pleistocene. Surprisingly, CaCO3 records from the far NW Pacific near Kamchatka (ODP Site 882 and cores nearby) have high CaCO3 during early interglacial intervals and no CaCO3 in the glacials, opposite to the rest of the Pacific. We synthesise evidence to suggest that sea ice and/or low salinity surface waters over Site 882 impairs glacial carbonate production sufficiently so that no CaCO3 is buried during glacial intervals. CaCO3 is found in sediments at the beginnings of interglacial intervals, likely because of changes in surface CaCO3 production as oceans reorganize and the Bering Strait opens to shunt sea ice and low salinity water north into the Arctic. The deposition of CaCO3, consistently associated with interglacial intervals at Site 882, implies that glacial sea ice and the associated low salinity surface layer was a factor affecting NW Pacific carbonate deposition during glacial-interglacial transitions throughout the Pleistocene.
- Preprint
(2740 KB) - Metadata XML
-
Supplement
(2516 KB) - BibTeX
- EndNote
Status: open (until 23 Oct 2026)
- RC1: 'Comment on egusphere-2026-3515', R. Wilkens, 14 Sep 2026 reply
-
RC2: 'Comment on egusphere-2026-3515', Anonymous Referee #2, 01 Oct 2026
reply
This manuscript presents high-resolution CaCO3 records from ODP Site 882 and uses them to explore why calcium carbonate accumulation in the northwestern Pacific margin differs from the broader Pacific pattern. The authors propose that cold, relatively fresh surface waters and associated sea ice, potentially linked to Bering Strait restriction during glacial intervals, suppressed carbonate production near the Kamchatka margin and that carbonate accumulation increased when a North Pacific-Arctic connection developed during deglaciation. The high-resolution carbonate record is potentially valuable, and the effort to locate and reintegrate a missing section of the Site 882 splice is a useful contribution. The proposed connection among sea level, Bering Strait exchange, surface-water properties, and carbonate production is also intriguing. However, I have several major concerns, as enumerated below, that have led to my recommendation of major revisions.
Major comments
1. The central framing of the carbonate record is misleading
The manuscript repeatedly presents the far northwestern Pacific as having a carbonate record that is “opposite” to the rest of the Pacific: high CaCO3 in early interglacial intervals and little or no CaCO3 during glacials. This is not an accurate description. The Site 882 record appears to be characterized by generally low CaCO3 values, with pronounced peaks during deglaciations. Those peaks may occur near the beginning of interglacials, but their occurrence during the transition out of glacial conditions is the important observation. This is not the inverse of the typical Pacific pattern, which generally shows cyclic glacial-interglacial variability with higher carbonate percentages during glacials and lower values during interglacials. It is instead a record with strong deglacial excursions superimposed on a low background. This distinction is not semantic: the manuscript’s overarching motivation, hypothesis, and interpretation are built around a mischaracterization of the record. The authors need to correct this premise throughout the abstract, introduction, results, discussion, and title before the proposed mechanism can even be evaluated.
The geographic scope is also overstated. The manuscript describes the far northwestern Pacific or the Kamchatka Margin “as epitomized by Site 882”, but it does not provide sufficient evidence that this single site is regionally representative. In practice, this is a site-specific paper about why the carbonate record at ODP Site 882 has unusual deglacial peaks. If the authors intend to make a regional argument, they need to present and evaluate additional records from the Kamchatka Margin and explain why Site 882 can serve as a regional reference point. Otherwise, the title and repeated regional claims should be narrowed. The title is also unnecessarily long and should be revised for clarity and precision.
2. The GRAPE record is used three times in ways that are not independent
The most serious methodological concern is the extent to which the manuscript reuses the GRAPE dataset for three different parameters. First, the GRAPE record is used to construct the age model (done in 1995 by Tiedemann). The same GRAPE measurements are then calibrated (in this manuscript) once to biogenic silica and separately to bulk density. Finally, the calibrated biogenic silica concentration, bulk density, and sed rate (based on the age model) are multiplied together to calculate biogenic silica mass accumulation rates – which in essence can be reduced to GRAPE multiplied by itself three times. Thus, one measurement is being used to establish the chronology, to generate all three terms entering the mass accumulation rate calculation, and to provide the basis for interpreting the resulting records. This is far beyond simply using one dataset for several descriptive purposes, and the manuscript does not address the resulting statistical dependence.
The problem is that the effective degrees of freedom are unclear and likely much smaller than the analysis implies. All downstream quantities share the measurement error and any depth-dependent biases in the GRAPE signal. Calibration uncertainty in the biogenic silica and density relationships is then compounded when the two calibrated products are multiplied, and the resulting mass accumulation rates are not independent of the GRAPE-based age model. Shared errors can produce apparent structure or covariance among the derived records and can make the final proxy appear more quantitatively constrained than it is. The authors need to account for this dependency explicitly and demonstrate that the main conclusions are robust to it.
At minimum, the authors should show the actual calibration scatter plots, regression equations, R2 values, RMSE, sample sizes, and fully propagated uncertainties, especially for biogenic silica. I went into the supplementary data to look for myself, and the relationship between biogenic silica and GRAPE derived wet bulk density was not very strong (R2 = 0.41). Using the smoothed, decompacted version of wet bulk density improved the relationship slightly (R2 = 0.52), but that still leaves a whole lot of variance unexplained. Given the scatter, the reconstructed biogenic SiO2 values may carry uncertainties on the order of several weight percent once regression error and RMSE are included. The authors need to propagate these uncertainties through the density and sedimentation rate calculations, and they should test the conclusions using independently measured density and biogenic silica data wherever possible. Without such an analysis, the GRAPE-derived mass accumulation rates should not be presented as precise quantitative records.
3. A robust age model is necessary to evaluate the proposed mechanism
Given the importance of this chronology to the manuscript’s central argument, the age model requires much stronger justification. At present, it appears to be based primarily on tuning the GRAPE record to precession, as was done over 30 years ago shortly after the core was recovered. How confident should we be in an age model constructed in this way, particularly when GRAPE is also used to reconstruct biogenic silica and bulk density? The authors should consider developing an independent age model based on the available benthic δ¹⁸O record, ideally using an objective correlation approach such as BIGMACS. At minimum, they should demonstrate that the proposed chronology is robust to reasonable alternative tuning choices and provide realistic age uncertainties.
Establishing the timing of the carbonate peaks is essential for evaluating whether the North Pacific record is genuinely anomalous. If the peaks occur during deglaciations, then the record is not simply the inverse of the typical Pacific pattern. It may instead represent a largely low-carbonate record punctuated by deglacial maxima, which are expected due to CO2 outgassing. By contrast, peaks that occur clearly within the early interglacial would have different implications for the proposed mechanism.
The treatment of Bering Strait flooding is also too simplistic. The timing of flooding cannot be inferred solely from the point at which a global sea-level curve exceeds the sill depth. Jakobsson et al. (2017) emphasize that the shallowest barriers are not necessarily located at the modern strait itself, and they note that isostatic changes, tectonic movements, and sediment deposition or erosion introduce uncertainty into the relevant threshold. Their new sediment-core evidence indicates substantial Pacific-Arctic exchange at approximately 11 cal ka BP, associated with the late deglacial flooding event, but it does not justify treating the global sea-level curve as a direct proxy for local flooding at Site 882 (as in Figure 3).
Pico et al. (2020) make the problem even more explicit. Their compilation distinguishes between evidence for an initial shallow connection at approximately 13.3 ka and evidence for substantial throughflow at approximately 11.5-11 ka. Their GIA modeling allows an initial breach near 13.3 ka followed by more complete inundation and substantial northward flow near 11.5 ka, with local relative sea level varying substantially across the Bering Shelf. This two-stage interpretation is directly relevant to the manuscript, because a carbonate response to throughflow, freshening, or nutrient transport would not necessarily coincide with the first moment at which a shallow sill was breached. The authors should address with this distinction, explain whether their proposed carbonate peaks correspond to initial inundation or establishment of substantial flow, and avoid assigning a causal connection between Bering Strait opening and carbonate accumulation that their current chronology cannot support.
4. The carbonate and biogenic-silica records require more careful treatment
The interpretation of organic carbon and biogenic silica also needs revision. The sediment contains very large amounts of opal in some intervals, in places approaching 50 wt%, but relatively little preserved organic carbon. This could indicate high productivity coupled with substantial remineralization or respiration within the sediment, processes that could also affect carbonate preservation (e.g., metabolic dissolution, Emerson & Bender, 1980). The authors should consider whether the relationship between CaCO3, biogenic silica, and organic carbon is consistent with changes in productivity alone, or whether post-depositional respiration and dissolution may be important. At present, the manuscript treats the proxy relationships too simplistically, especially given the uncertainty in the reconstructed biogenic silica record.
5. The literature review and regional comparisons are not sufficiently current or relevant
The manuscript relies heavily on a relatively small number of older references and frequently cites the authors’ own work when it may not be the most appropriate or relevant. For example, observations from the California margin (usually made by the authors in previous publications) are transferred to the northwestern Pacific despite the fact that the regions are hydrographically quite different. Nutrient transport and upwelling along the California margin are governed by an eastern boundary current system and cannot simply be used as a direct analogue for the hydrographic conditions near Kamchatka. The manuscript should cite observations and modeling studies that are actually relevant to the northwestern Pacific and Bering Sea. aSimilarly, the discussion of C37:4 alkenones would be more to focus on the Bering Sea relationship rather than relying primarily on results from the Norwegian Sea. Harada et al. (2003) report a different relationship in the Bering Sea, yet the manuscript does not explain its implications. Moreover, most of the Site 882 values appear to fall outside the salinity and C37:4 range over which the cited relationship is demonstrated. The authors should determine whether the biomarker signal reflects cold temperature, low salinity, or both, and should evaluate whether newer studies (since 2003) have refined that relationship. The issue is not that older studies are inappropriate, but that the manuscript does not demonstrate sufficient engagement with more recent work or with the broader literature.
More generally, the manuscript often advances qualitative associations without providing direct evidence. For example, the apparent coincidence of moderate winter SST and relatively warm summer or fall SST with higher CaCO3 production does not establish a causal relationship. The authors need to distinguish observations from hypotheses and support proposed mechanisms with regional data, modern analogues, or model results.
6. Figures
The figures require substantial redesign for readability and scientific clarity. Across the manuscript, dense cross-hatching, gray bands, overlying text, and low-contrast colors make the data difficult to read. The authors should use simpler graphical treatments and color palettes that are accessible to readers with color-vision deficiencies. Captions should identify every subplot, core, plotted field, and symbol without requiring the reader to infer what is being shown. The maps should devote more space to the actual study region and make the core locations legible (and colorblind safe). The inset map in Figure 2 is distracting and it is unclear what it is adding to the figure. Figure 3 is not really a “hypothesis” figure; it is a data plot. If the authors want to present a conceptual hypothesis, they should add a schematic showing the proposed links among sea level, Bering Strait exchange, surface-water properties, sea ice, productivity, and carbonate preservation. That would actually be a helpful addition.
Citation: https://doi.org/10.5194/egusphere-2026-3515-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 160 | 64 | 46 | 270 | 43 | 50 | 47 |
- HTML: 160
- PDF: 64
- XML: 46
- Total: 270
- Supplement: 43
- BibTeX: 50
- EndNote: 47
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study provides a solution to a puzzle: why does the Northwest Pacific (Kamchatka Margin) exhibit a carbonate burial signal opposed to the rest of the Pacific Ocean. While standard Pleistocene Pacific records feature high carbonate abundance during glacial intervals due to reduced abyssal dissolution, ODP Site 882 displays pronounced carbonate spikes exclusively during early interglacials and near or total absence during glacials.
The authors present a meticulous re-evaluation of the late Pleistocene record at Site 882. By incorporating a corrected splice, generating high-resolution particulate organic carbon and carbonate profiles and calibrating wet bulk density to biogenic silica, they construct a robust record of mass accumulation rates spanning 650 kyr.
The primary strength of the manuscript lies in its rigorous multi-proxy approach to test deep-water dissolution models. By integrating alkenone-derived sea surface temperatures and C37 alkenone percentages alongside biogenic fluxes, the authors persuasively demonstrate that surface productivity controls—rather than corrosive abyssal waters—drive the regional carbonate anomaly.
Overall, this manuscript is an exemplary piece of paleoceanographic synthesis that resolves an anomalous regional record of carbonate accumulation. I don't have any substantive technical or scientific questions. I do have one stylistic suggestion. In the first sentence of the manuscript the authors define carbonate with “CaCO3” and then proceed in almost the entire rest of the manuscript to use the chemical formula instead of the word “carbonate”. For whatever reason, while I am reading, I say to myself the entire chemical formula, which I find somewhat disruptive. I would much prefer they use the word carbonate rather than the formula for it. It will read more smoothly.