the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 30 Jun 2026
Refined CT-Based Proxy for Foraminiferal Diagenesis Shows Evidence for Shallow Dissolution in North Atlantic Sediment Cores
Abstract. Geochemical reconstructions (e.g., δ18O, Mg/Ca) derived from the shells of the polar and subpolar planktonic foraminifera Neogloboquadrina pachyderma and Neogloboquadrina incompta are foundational to our understanding of high-latitude climate history. However, these species are highly vulnerable to diagenetic alteration—specifically partial dissolution and secondary calcification—which can significantly bias paleoceanographic signals. We introduce nCTDX%, a refined X-ray micro-computed tomography (μCT) proxy designed to quantify foraminiferal diagenesis with increased objectivity and time efficiency. By implementing mathematical thresholding and normalization to a density standard, nCTDX% allows for the direct comparison of shell density across different scans and study sites. Our results from North Atlantic, Nordic Sea, and North Pacific sediment cores confirm that both species dissolve from the inside out, preferentially removing internal structures previously shown to be rich in Mg and leaving behind resistant, low-Mg gametogenic crusts. While core-top nCTDX% correlates with bottom-water carbonate saturation (Δ[CO32-]), we demonstrate that undersaturated pore waters can initiate significant dissolution within the top 2 cm of the sediment column, even at locations where overlying bottom waters are supersaturated. Furthermore, we show that nCTDX% is a sensitive indicator of secondary inorganic overgrowth, which appears in SEM images as low-density rhombohedral crystals on shell exteriors. The potential for pore water dissolution and inorganic calcification to overprint the bottom water Δ[CO32-] signal contained in nCTDX% and other μCT and dissolution-based proxies suggests extreme caution must be taken when they are applied as proxies for bottom water carbonate ion saturation. Instead, given that even partial dissolution of internal shell walls can bias reconstructed temperatures by 1–2 °C, our findings suggest that μCT and dissolution-based proxies are better utilized as indicators of in situ diagenetic state. We therefore advocate for the routine application of nCTDX% to evaluate the preservation state of downcore samples and ensure the fidelity of geochemical paleoclimate records.
- Preprint
(1546 KB) - Metadata XML
-
Supplement
(2379 KB) - BibTeX
- EndNote
Status: open (until 25 Aug 2026)
- RC1: 'Comment on egusphere-2026-3333', Anonymous Referee #1, 06 Aug 2026 reply
-
RC2: 'Comment on egusphere-2026-3333', Anonymous Referee #2, 06 Aug 2026
reply
Review “Refined CT-Based Proxy for Foraminiferal Diagenesis Shows Evidence for Shallow Dissolution in North Atlantic Sediment Cores.”
This is an interesting article using X-ray micro-computed tomography to look at diagenesis (dissolution and inorganic secondary calcification).
The paper shows interesting results, confirms (albeit with low sample numbers) the potential effect of undersaturated seawater on core top samples, and that of porewaters on down-core sediment samples.
General comment. nCTDX% forms the main data for this manuscript. While it is interesting to see the summary figures (e.g. figure 5 and 7), it would be helpful to see some of your main inferences evidenced by figures in the main manuscript. For example, in the abstract you write: “Furthermore, we show that nCTDX% is a sensitive indicator of secondary inorganic overgrowth, which appears in SEM images as low-density rhombohedral crystals on shell exteriors.” If this is one of your important conclusions, you need to provide a figure with nCTDX% cross sections showing where this is the case (and where not!) in the profile to the main manuscript. It would also be helpful if you indicated in your supplementary figures which figures show dissolution, which secondary growth and indicate where. It would furthermore be insightful if these figures could then somehow be related to environmental conditions!
Other comments
Line 73 “In addition to partial dissolution, water supersaturated in [CO32-] can also result in diagenesis that influences shell geochemistry.” Replace result in with cause?
Line 120 what does remineralization refer to (do you mean dissolution resulting from low carbonate ion due to remineralisation of organic material?)
Figure 1. Not sure what white symbols refers to (there are purple symbols).
Line 195 Samples size:
It is confusing how the decision regarding minimum sample size was reached by Iwasaki et al. (2022). In their supplementary information they say that test samples were obtained from multiple cores which does not make sense; perhaps multiple samples down-core…
If you look at their a (PS97/114) sample there still is quite a bit variability in st.dev of %low CT volume at a sample number of 15. Iwasaki et al. looked at Globigerina bulloides, whereas here you are looking at other species. To ensure similar conditions apply, it would have been good if you had carried out a similar check.
Section 3.2
Lines 290-292: generalised statement. Is it not possible that the different species have different calcite matrix make-up and thus response differently to in situ ddCO32-? For N.pachyderma it looks like some downcore samples have higher nCTDx%, but not all (different station 3 and 2, where 2 has slightly higher ddCO32-). The baselines look different for the two species, so that for N. incompta one could argue that only station (downcore sample) 19 (with uncertain ddCO32-) is affected by a higher nCTDX%.
Section 3.2
Lines 369-371: these are only a few datapoints; with only one (53MC) showing this (Station 19 has a very wide uncertainty for dd13CO32-). Why do you not comment on N. incompta’s insensitivity to lower carbonate ion? Does this mean it resistant to dissolution?
Please add a figure that shows some clear examples in the article.
Line 428-435: when considering down core samples, you need to consider bioturbation effects, effects of burrows etc as well; some of your specimens could be 200 years, the other thousands of years. And bottom water and pore water carbonate ion concentrations will vary seasonally, in relation to organic carbon flux to the seafloor.
Lines 487-489: please rephrase this. You seem to have different threshold values to the two species, so it will likely be different for other species as well. Plus you cannot be sure that the relationship will be the same in different environments (for example methane seeps).
Lines 576-578: some words missing.
Line 588 basedproxies.
Line 594-594: why do rapid sedimentation rates lead to higher remineralization rates of organic matter? Does it not depend on what the sediment is made out of?
Citation: https://doi.org/10.5194/egusphere-2026-3333-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 51 | 26 | 15 | 92 | 14 | 18 | 9 |
- HTML: 51
- PDF: 26
- XML: 15
- Total: 92
- Supplement: 14
- BibTeX: 18
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Summary of study
This is a generally well written study presenting microCT scans of subpolar /polar planktic foraminifera species N pachyderma (left and right coiling) (13 samples) and N. incompta (7 samples) from core-tops and sediment cores from the Nordic Seas, subpolar North Atlantic and the North Pacific.
The study builds on previous work which has used the technique of uCT on samples from sites with different Δ[CO32-] values in order to investigate post depositional changes and partial dissolution of the tests of planktic foraminifera. This study is small, but would add new, previously not published species to the knowledge base.
Samples come from a mixture of Atlantic and Pacific core tops (9 samples), where bottom water Δ[CO32-] was between ~ +45 to -13 umol/kg, and down core (to 2 cm) (5) samples. One novel feature of the study is they calculate pore water Δ[CO32-] for the downcore samples. This was always considerably (up to ~30 umol/kg) less than the bottom water, with values of ~+15 to -11 umol/kg.
CT data is presented in the form of a dissolution index, nCTDX%, based on the attenuation values of the scans. The method is based on that of Iwasaki et al. (2022, CTDX%), but modified by using fragments of a standard material (NBS19) to try to standardize where the thresholds of the scan output should be selected. There is an intrinsic problem involved in trying to assess CT scans of foraminifera in that, test wall (particularly if it is partially dissolved), sediment trapped in the test and material reprecipitated inside the test are not completely distinguishable from each other. The method of thresholding used in this index, where there is a cut off for low density material, but the remaining selected pixels are dilated slightly seems a reasonable approach
Conclusions fit generally with previous findings, that inner material of planktic foram tests is the most susceptible to dissolution; that partially dissolved / partially reprecipitated material on the inside of partially dissolved tests shows as low-density material in CT scans; and that environments of Δ[CO32-] below some threshold can be identified with this technique
The methodology seems fine and study aims seem worthwhile. There are some areas which need clarification. The ms is also a bit long for the amount of data presented, it could be made more concise. I put my comments / queries / suggestions to improve the ms below.
Core tops and down core samples
I would like more information about these cores and sites. There are very large differences between calculated Δ[CO32-] of bottom water and pore water, even from sediment shallow in the core (0.5 to 1 cm). Why is this? Is this typical for the average ocean or is it particular to these sites? Are the sites particularly high in organic matter? Do they contain abundant calcite? Are the porewaters anoxic? Is the pH of the porewaters still controlled by carbonate system chemistry, or could it be influenced by acids produced by respiration of organic matter?
What were the default values used for total phosphate and total silica? Zero? Is this likely to bias results?
Actual porewater values came from depth x and then the values were interpolated between that pore water value and bottom water for each sample depth. All the data used to calculate Δ[CO32-] – alkalinity, actual depth of pore-water data, pH– should all be in the Sup Mat.
It is stated (line 70) that usually in these kinds of studies it is the Δ[CO32-] of bottom water that is used, not pore-water and this might introduce inaccuracy. In the open ocean, this might be completely fine. It may not be a good assumption for your sites. Say why. This topic needs more discussion.
Is it concluded that Δ[CO32-] from porewaters fits better with the dissolution index than using bw values? It is never clearly stated
NBS19 standard
Say more about this standard. What is it made of? Where did it come from? Why are the particle sizes so different?
CT is of course a semi-quantitative technique, attenuation intensity of a material depends not only on the material of the sample, but its form (porosity, shape). Then there is some variation in the scanner from day to day, and different scanners no doubt give different values.
I was surprised to see that the values of repeated scans of std A shown in Fig S2 are quite so variable. But I also wonder if this is actually a meaningful number. The intensity of the background should be added here. Does it vary in the same way? I would expect that the difference between the background and the standard would be more consistent, and more useful.
The implication is ( line 111 ) that CT gives extremely variable results between scans and that previous studies, which did not use NBS19 standard, used thresholds “visually chosen” for each sample and no attempt was made to provide consistency between samples. This is not the case.
Line by line comments
Abstract
Add that dissolution index nCTDX% presented here is based on CTDX% (Iwasaki)
Add information about number of samples
Line 21 – how does the refined method add to time efficiency, where is this presented?
Line 26 – n..% correlates with Δ[CO32-], what is the correlation. Add to discussion and to Conclusion
Introduction
Add information about the sites. Why were these sites chosen for investigation, what is special about them
Line 43 – state that foram tests start to dissolve in waters undersaturated wrt calcite, this is not state of the knowledge. They start to dissolve in supersaturated waters, many references for this, some even cited later
Line 91 – state that (n)CTDX% is a dissolution index and that high values represent less dense partially dissolved material. (Usually for CT the convention is that high values mean more dense material (higher greyscale, lighter colours), so it would be good to clarify for people who might be familiar with CT but not dissolution indices.) Same goes for the figures. Add an arrow to nCTDX% axes to show which direction means better preservation
Line 105 - the problems of distinguishing between partially dissolved test, trapped sediment and newly crystallising calcite is not only described by Iwasaki et al., (2022), it is mentioned in all studies using CT of forams
Methods
Describe orientation of tests in the scanner. It appears random with some on their sides, some aperture up.
All 8 tests from one sample are scanned together?
Section 2.3 image processing (line188)
There are some unexplained initials in this section eg TMX, expand
Line 209 – all slices are used to define the background? But all slices will contain std and sample, not just background. Unclear
Line 259 - Authors state they do not want to get rid of pixels lower intensity than the air peak because it comes from beam hardening (softening). What is the advantage of including scan artefacts
Line 282 –more details are needed in Sup Mat of how you calculated Δ[CO32-], but I suppose it is meant with respect to calcite (not eg aragonite)
Results
Does the offset pore and bottom water seem reasonable? Can it be explained
Line 312 - All the sample names and results should be in a results table, not just in the sup figs
Line 394 – several samples are mentioned by sample name, do they share similar sites / features?
Discussion
Line 440 - authors mention that N. incompta may be more dissolution susceptible than N. pachyderma because it has higher Mg/Ca. There are several sets of CT scans published for various species. Some from Iwasaki are plotted in Fig 7. Does the observation hold true, does N pachyderma appear have lower dissolution index than tropical species for similar Δ[CO32-] values ?
Line 451 – the reader is directed towards a large figure, need more direction as what is supposed to be seen here
Line 475 – similar to above, make a fig showing what is meant
Line 485 – why should time be a factor, and how was it concluded it is not
Line 492 – high values of dissolution index n..% does not identify between dissolution and diagenetic overgrowth. Preservation should be “evaluate preservation state of the internal walls”. How is this then to be evaluated?
Section 4.6 (line 521)
This part reads like a literature review. I don’t think it adds to the ms. No geochemical data was measured on these samples so no need to speculate so much.
Conclusions
Line 586 – how reliably can you identify tests from sites where Δ[CO32-] < 0 from the index. Looking at fig 7, it is not an exact method.
Was there a better match between the dissolution index and Δ[CO32-] for the core top or the down core data?
Line 595 – states that n..% identifies post deposition precipitation on exterior of test as shown in the sem images (Fig). I do not see convincing evidence of this in the CT scans. If it is there, direct the reader to it with a figure. If by diagenetic overgrowth, it is meant the low density material inside the test which has been described and discussed in previous studies as a partial dissolution feature, then this is different to what Groeneveld 2008 shows, which is precipitation onto well preserved tests.
Table 1
Is there a mistake, CE23012?
Figures
Fig 4 could go to sup mat
Fig 5 - Adding error bars to each symbol in the legend detracts from clarity
It would be more visually pleasing to place Δ[CO32-] on the y axis and CT on the x axis
Add information to CTDX% scale on y-axis to show in which direction preservation is better or worse
Fig 6 – add arrows to guide the reader to the post mortem calcite
Could you show CT scans of the same features
Figure 7 – it is stated this data matches well with Iwasaki data. Fig 7 shows data on two different scales. How were the y axes matched up? By assuming similar Δ[CO32-] gave similar index values???
Supplement
Figure S4, S5 add information to scale bar
0 100
less dense – more dense
What is the white diamond on the scale bar?
“Average station ..DX% increases … right”
does this have some significance? Is there some difference in the stations to the right
All the results presented in S6 etc should be in a table