the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Dec 2024
Sea ice and mixed layer depth influence on nitrate depletion and associated isotopic effects in the Drake Passage – Weddell Sea region, Southern Ocean
Abstract. The regions near the Antarctic Peninsula in the Southern Ocean are highly productive, with notable phytoplanktonic blooms in the ice-free season. The primary productivity is sustained by the supply of nutrients from convective mixing with nitrate-rich subsurface waters, which promotes rapid phytoplankton growth as the sea ice melts in spring and summer. Surface waters are marked by the contrast between the warmer Drake Passage and the colder Weddell Sea, and seasonal duration of sea ice cover varies accordingly. Sea ice exerts multiple controls over primary production, by shading the light entering the ocean and stratifying the upper ocean with freshening by ice melt. However, the interaction between sea ice and productivity remains poorly characterized because satellites are unable to quantify biomass in partially ice-covered ocean, and direct measurements are too scarce to characterize the seasonally varying productivity. Here we evaluate productivity by assessing removal of nitrate from surface waters by biological nutrient utilization, and study the associated change in δ15N of nitrate. We use a combination of bottle samples and in situ nitrate measurements from published databases, completed by two transects with isotopic measurements. The timing of sea ice melt date conditions the initiation of nitrate drawdown, but the annual minimum of nitrate only weakly correlates with sea ice concentration. As previously reported, we observe that δ15N of nitrate increases with nitrate depletion. Interestingly, the lowest nitrate depletion and δ15N values are found in the central region of N-S transects, where intermediate temperature and sea ice conditions prevail. Deeper mixing in waters that passed through the northern Bransfield Strait may explain higher nitrate concentration due to both a greater nitrate resupply and reduced productivity under light limitation in deeply mixed waters, confirmed by isotopic fractionation effects during nitrate uptake and nitrogen isotope modelling. This highlights the importance of oceanographic controls on productivity patterns in sea-iced regions in the Southern Ocean.
- Preprint
(3758 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 23 Jan 2025)
-
RC1: 'Comment on egusphere-2024-3687', Anonymous Referee #1, 03 Jan 2025
reply
General comments
Motivation/coherence
While I’m not an expert on isotope chemistry, I don’t see any obvious flaw in the methodologies of the paper. It is very descriptive, but that is okay as long as the overall motivation is clear. However, the motivation isn’t always clear, and it becomes somewhat less clear as more parts of the analyses are being discussed. My concern is that the individual pieces of the analysis aren’t well connected. Or rather: that it’s not made very clear how they are connected, and why they’re needed to answer the main questions. Is the isotope data mainly confirming that the approach used for the nitrogen deficit was valid (L436-438)? Or is it mainly helping to understand what’s happening in the deep mixed layers in the Leg B transect, hinting at light limitation? Or both? I’m also not sure I understand what the model adds to the overall interpretation of the data. So I recommend making the connections and motivations clearer, leading the reader through the story that the data and analyses tell.
Additionally, if some of the analyses are merely supporting main conclusions coming from one aspect of the work, the sections pertaining to the supporting analyses could maybe be shortened/more focussed.
Specific comments
Why this region?
The general question of what influence sea ice exerts over nitrate uptake is a worthwhile one to try and answer, but if one was interested in a universal answer, the analysis should encompass the whole sea ice zone in the Southern Ocean, not just this area around the Antarctic Peninsula. So why the focus on this area, what is special about it? If anything, it seems like an oceanographically complex region, so whatever the answer regarding the influence of sea ice that is found here, it’s likely not going to be universally applicable. That’s fine, but it would be good to motivate more clearly the interest in this question in this region.
What about iron limitation?
I agree that nitrogen data can be used in iron-limited regions to learn about productivity, there’s no problem there. However, the first sentence of the conclusions states that the goal of this study was to better understand environmental controls on primary productivity - but iron limitation was not investigated at all. In the introduction the authors state that the Bransfield Strait isn’t iron-limited, but the area investigated is much larger than the Bransfield Strait, and iron limitation is very likely to play a role in this larger area. So any investigation of the “environmental drivers” that doesn’t consider iron limitation is missing a big potential piece of the puzzle. Indeed, the fact that the sea ice seasonal range does have a measurable influence on nitrate drawdown but the MLD is not the driver (Figure 5) could be interpreted as an indication that another aspect of sea ice melt, for example release of dissolved iron, may be responsible for the observed relationship. So I recommend that the authors reconsider, and discuss in more depth, the potential influence of iron limitation on productivity in the region.
Some comments in order of appearance:
L 19: Since nitrate is not limiting in the region, this should probably refer to the re-supply of iron? See also comment about iron from sea ice melt above.
L 52: Argo data are now allowing us to look under the sea ice, so satellites are not the only means to remotely measure the ocean under ice.
L 186: Refer to figure 3 here? This would help the reader understand that these analyses are used for the maps in this figure
L 186-194: Consider citing some more recent studies that have used this approach with the WW (e.g. Moreau et al.)
L 242-245: Aren’t the two sentences basically saying the same thing? Maybe this could be streamlined.
L 294-295: I can’t see how this summary is derived from the Figure.
Figure 4b: How is ice melt defined here? When sea ice concentration is 15% or less? Please specify in the Figure caption.
L 316-317: Is it worth also testing other metrics for seasonal changes in sea ice cover (other than the standard deviation)? For example, the amplitude (between max and min sea ice cover) could be a metric to try.
L 327: Please add “in this region” to the end of the last sentence.
L 328, 331: When saying SIC here, do you mean variability (std) of SIC? That’s what the figure shows that is being referred to.
L 354 (Eq 1): What is taken as the initial here?
L 364-366: Can these two goals really be mixed? If there are deviations from the expectation, how can you tell whether that has to do with environmental forcings versus a violation of the relationship? I have trouble imagining what it would mean/look like if “the relationship between nitrate concentration and isotopes” were not respected. So maybe it’s my ignorance but if that’s the case, then a few explanatory words might help.
L 406-407: For someone new to this type of analysis, a bit more detail about how to think of this analysis/interpret it would be useful here. Why is this analysis (Figure A2) done station by station, what’s the reasoning for that? If there’s a disconnect assumed between MLD and the deeper waters, it’s not obvious to me that the deep data at a given station should be used? Does the noise around the slope hold any information that is worth interpreting, or is it just the slope (e.g. LB07, which is also one of the 2 stations with the high epsilon)?
L 412-426: Could this paragraph be tightened to make the main finding (light limitation in the profiles with deep MLD) clearer and relate it back to the motivation stated on lines 394-395? Does this result then exclude the other explanation for the high nitrate concentration mentioned on lines 392-394? And could there be any other explanations for the high observed epsilon (other than light limitation)?
L 440-446: Can the locations that were modelled please be shown in Fig 1 or Fig 2? Also, how/why were these specific locations chosen, what was the motivation?
L 503-504: What the model results show for the biological pump will depend highly on the parameterization. So I suggest to either cut that bit or lay out more explicitly what the parameterization for the biological pump was, and how well it’s expected to work in this region.
L 517-518: Is this one of the main results of this study? If so, can this be explored a bit more? How does this compare to what we already knew about the region?
L 529-535: The nitrate deficit approach is hardly qualitative, it’s quantitative. This whole paragraph could use more focus/tightening.
Technical corrections
L 203-204: Consider rewording the sentence starting with “Therefore”
L 216: Replace “international” with “interannual”?
L 263: Should this refer to Figure 3 instead of Figure 2?
L 285: Consider replacing the word “non-segregated”
Fig 6: Consider switching the colour bar for the top panels so it matches the bottom? Darker colours are also usually associated with higher concentrations, not lower.
L 491-492: Not sure what is meant by “divergence of nitrate concentration” and also what the word “differs” refers to
L 498: Phytoplankton can also grow at depth, and “advected” is probably a more appropriate term here than “dragged”
Citation: https://doi.org/10.5194/egusphere-2024-3687-RC1 -
RC2: 'Comment on egusphere-2024-3687', Anonymous Referee #2, 09 Jan 2025
reply
This study examines the influence of sea ice and mixed layer depth on productivity in the Drake Passage and Weddell Sea regions—a significant and ongoing area of debate. While the manuscript is well-written and most analyses are sound, I find that the connections and added value of each section are sometimes unclear and redundant, which could be improved. Additionally, a thorough comparison with other sectors and studies is lacking. Specifically, are these findings representative of the seasonal ice zone as a whole? Addressing these aspects would enhance the manuscript's clarity and contextual relevance.
Minor comments:
Line 58: Such an approach has been described well before Moreau et al. (2020), e.g., Pondaven et al. (2000; doi:10.1038/35012046) or Nelson et al. (2002, doi:10.1016/S0967-0645(02)00005-X).
Lines 138-144: You can briefly mention (1-3 sentences) that you measured nitrate+nitrite δ15N instead of nitrate-only δ15N. This approach is entirely appropriate for evaluating nitrate assimilation in the Antarctic Zone, given the observed differences between nitrate+nitrite δ15N and nitrate-only δ15N, which are attributed to the interconversion between nitrate and nitrite. (e.g., Kemeny et al., 2016, https://doi.org/10.1002/2015GB005350; Fripiat et al., 2019, https://doi.org/10.1016/j.gca.2018.12.003).
Line 188: A reference such as Spira et al. (2024, https://doi.org/10.1029/2024JC021017) may be more appropriate to describe the winter water.
Line 234: Are you confident that a four-year spin-up period is sufficient? Have you ensured that a steady state has been reached? Kemeny et al. (2016, https://doi.org/10.1029/2018PA003478) employed a spin-up phase lasting over a hundred years with a similar model.
Line 279: Here is an example where a more comprehensive analysis could have been conducted, such as a circumpolar-scale comparison. For instance, by comparing satellite estimates of primary production with the timing of sea-ice retreat.
Line 290: See also the study of Savoye et al. (2004, https://doi.org/10.1029/2003GL018946) or Sambrotto and Mace (2000, https://doi.org/10.1016/S0967-0645(00)00071-0).
Lines 312-314: Did you check the effect between sea ice (area covered or std sea ice) on MLD? This could be worthed to mention.
Lines 316-319: While I found the metrics “std sea ice” convincing, I would also test other metrics. Have you tried other metrics (e.g., difference in sea-ice concentration, time difference between maximum and minimum sea-ice extent …)?
Section 4.2: The authors should consider that some of the studies referenced used nitrate-only δ15N to estimate the isotope effect, while others used nitrate+nitrite δ15N (see Fripiat et al., https://doi.org/10.1016/j.gca.2018.12.003). Given the evidence for nitrate-nitrite interconversion in the Antarctic Zone, where this process occurs in a closed system, nitrate+nitrite δ15N values are more representative of assimilated δ15N and should be used to estimate the isotope effect. Based on this approach, no significant difference in isotope effect estimates was found in the Antarctic Zone, consistent with Fig. 8 (i.e., the two data points at 60 and 80 MLD are not significantly different from the one at below 50 MLD). This discussion should be somewhere mentioned in the text.
Section 4.2.2: The added-value and motivation for the model are unclear.
Citation: https://doi.org/10.5194/egusphere-2024-3687-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 149 | 67 | 6 | 222 | 0 | 1 |
- HTML: 149
- PDF: 67
- XML: 6
- Total: 222
- BibTeX: 0
- EndNote: 1
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1