the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Oct 2024
Controls on brGDGT distributions in the suspended particulate matter of the seasonally anoxic water column of Rotsee
Abstract. Developing reliable methods for quantifying past temperature changes is essential for understanding Earth's climate evolution and predicting future climatic shifts. The degree of methylation of branched tetraethers (MBT’5ME), of branched glycerol dialkyl glycerol tetraethers (brGDGTs), a group of bacterial membrane lipids, has become a widely accepted tool for lacustrine paleothermometry. To allow this, an empirical calibration was developed, based on MBT’5ME values of surface sediments across large spatial scales. As these sediments integrate variability across several years to decades, the sensitivity of MBT’5ME to seasonal and short-term environmental changes in the water column remains underexplored. Here, we present a record of brGDGTs in suspended particulate matter (SPM) from a monomictic, eutrophic temperate lake (Rotsee, Switzerland) over a 10-month period, examining both core lipids and intact polar lipids. Rotsee offers an ideal setting for this study due to its strong seasonal variations in temperature, conductivity, and dissolved oxygen caused by summer warming and associated stratification. In the oxic epilimnion, a minor increase in MBT'5ME during stratified summer months was caused by a rise in brGDGT Ia concentration. A similar increase in concentration of 6-methyl brGDGTs indicates a sensitivity to water temperature. In the seasonally anoxic hypolimnion, MBT'5ME correlated with water pH rather than temperature, suggesting that water chemistry influences this ratio, complicating its use as a temperature proxy. The production of intact polar lipid (IPL) tetraethers was observed exclusively in the anoxic hypolimnion during stratification, confirming anoxia as a key trigger for IPL tetraether production. Surface sediment samples along a depth gradient have a distinct depth-dependent distribution. Sediments below the oxic water column showed lower MBT'5ME values, likely due to the sedimentary production of brGDGTs IIa and IIIa. Sediments from seasonally anoxic areas reflected average epilimnion SPM values, suggesting the deposition of epilimnion brGDGTs into the sediments. This study of brGDGTs in Rotsee SPM and sediments thus indicates that temperature, pH and oxygen concentration impact GDGT distribution, with significant implications for using MBT'5ME as a temperature proxy in sediments from stratified lakes.
- Preprint
(1803 KB) - Metadata XML
-
Supplement
(647 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-3052', Anonymous Referee #1, 06 Nov 2024
The study of Ajallooeian et al. reports branched GDGT data and genetic data from SPM obtained from a large part of a single season in lake Rotsee, as well as comparisons with surface sediments and soils. The discussions follow the often-used strategy for interpretation of branched GDGT data, i.e. one correlates concentrations and indices with all environmental or genetic quantitative parameters which happened to be measured and discuss all significant, and some non-significant, correlations. As often happens, some of the correlations agree with previous studies and some do no leading to the often-made conclusion that multiple environmental parameters by influence brGDGT distributions.
The data are useful and add to a growing collection of such data which are not easy to obtain. The unfortunate thing is that I do not see (yet) how this data improved our understanding of lacustrine branched GDGT except ‘it is complicated’. There have been more then plenty of those kind of studies on lake branched GDGTs (the far majority perhaps) so this is not a new insight. This is not the authors fault but one would like to see studies where more clarity is obtained and solid conclusions can be drawn. Can we still apply branched GDGTs as a lake temperature proxy? Should we just stop altogether in applying it? Is this single seasonal study of SPM sufficient to say that a brGDGT record from Rotsee is not useful for paleoreconstructions?
My main criticism of the study (details below) is that it fully relies on correlations. For this to work, the quantitation, number of data points and representativeness of the samples must be sufficient. I am worried about these aspects. For example, as far as I can see there has not been a ‘true’ replicate analysis done (i.e. independent work-up of 2 parts of the same filter with addition of IS to the raw extract) so there is no way of knowing the real error in the quantitation. Some of the changes in concentration or indices are fairly small and we have no way of knowing if they are real changes or not. The number of data points is reasonable (12, but seems less for e.g. nutrients, i.e. 7 if I counted correctly) but I am unclear how representative is this season for Rotsee? Would another season have shown the same? Several studies have shown that biomarker lipid patterns can vary from season to season. I realize this is additional effort but it is a concern which needs to be addressed.
Detailed comments:
Line 40-41. ….has become a widely accepted tool for lacustrine paleothermometry…. Unsure about that wide acceptance for lakes but at least nothing is mentioned in the introduction on ‘success stories’ of this proxy in lake temperature reconstructions. Perhaps a few examples of apparently nice temperature records, as well as clearly wrong/biased records, would be useful in the introduction.
Lin 112. I would have expected some lake DNA (metagenome) studies as a reference.
Line 124. Are these the only seasonal studies of lake brGDGTs? I thought there were more. If so, did they reach similar conclusions?
Line 206-209. This error is not the error in quantification as the ionization differences between IS and GDGTs were not corrected (I think? If so, the ng amounts calculated are just a complete guess) and the IS was added at the really final stage of fraction preparation (errors due to hydrolysis+workup are not included). At best this is a repeatability error of the instrument. I cannot therefore agree with the statement at these lines because this error really represents the (unrealistic) best case scenario and ignores the complete workup and sample inhomogeneity errors. I would recommend splitting a sediment sample, or even better a filter cut in two pieces, and work this up completely separately to obtain a better constraint of the overall quantification error.
Line 375-380 and line 483. IPLs are nearly always a small fraction of the total lipids. Since direct measurements of IPLs were not done, this could even be an overestimate, i.e. if some of the GDGTs released by acid hydrolysis were not derived from IPLs but from matrix-bound material. Why are these IPL abundances so low, if the assumption is that all this material is derived from living biomass? Do branched GDGTs mostly occur as CL in the cell? Is cell lysis so quick or happening during filtration?
Line 385 and further. Coming back to the 15% error in quantification. It is not indicated here what the errors in brGDGT indices are. Are the changes observed larger than the assumed errors? At least replicate analysis could be done but preferably replicate work up.
Fig. 2. There is a clear mismatch in the no. of data points with only 7 data points for the cat- and anion concentrations. Can we really make solid conclusions on so few datapoints?
Citation: https://doi.org/10.5194/egusphere-2024-3052-RC1 -
RC2: 'Comment on egusphere-2024-3052', Anonymous Referee #2, 11 Nov 2024
Ajallooeian et al. presented a record of brGDGTs in suspended particulate matter (SPM) from a monomictic, eutrophic temperate lake (Rotsee, Switzerland) over a 10-month period, examining both core lipids and intact polar lipids, in addition to surface sediments and soils. The authors aimed to elucidate which environmental variables, such as water temperature, water chemistry (e.g., dissolved oxygen, pH, alkalinity, and conductivity), or bacterial community composition, best explain seasonal variations in brGDGT distributions, and thus examine the sensitivity of MBT’5ME to seasonal and short-term environmental changes in the water column. Overall, the study underscores the influences of temperature, pH, and oxygen on brGDGT distribution, raising important considerations for using MBT’5ME in temperature reconstructions from stratified lake sediments. The manuscript is well-organized, but several issues need to be addressed before acceptance.
As environmental factors influencing brGDGT production and distribution, temperature, conductivity, dissolved oxygen, pH, and alkalinity were considered in this study. Among these, the authors suggested that in the oxic epilimnion, MBT'5ME was associated with temperature, while in the seasonally anoxic hypolimnion, MBT'5ME correlated with water pH. However, considering Tables S2 and S5, it appears that in the epilimnion, MBT'5ME is related not only to temperature but also to conductivity. In the hypolimnion, MBT'5ME is influenced not only by pH but also by temperature. Does this imply that temperature, pH, and oxygen are not only complicating factors for the use of MBT'5ME, but also conductivity?
The authors proposed that the IR represents a stronger dependency on temperature in the epilimnion, highlighting the potential of using this proxy to identify brGDGT distributions dominantly sourced from the epilimnion within the water column by comparing MBT'5ME and IR in parallel. According to Table S2, IR also correlates with conductivity, similar to MBT'5ME. Would this suggest that although brGDGTs sourced from the epilimnion can be identified by comparing MBT'5ME and IR, MBT'5ME still does not fully reflect the influence of temperature alone?
Sediments from seasonally anoxic areas reflected average epilimnion SPM values, suggesting the deposition of epilimnion brGDGTs into the sediments. Does this suggest that the seasonal contribution from the anoxic hypolimnion plays a minor role in the application of MBT'5ME in such a lake? What could be the reason that the brGDGTs produced in the hypolimnion were not deposited or well-preserved in the sediments?
Other comments:
51-53: “The production of intact polar lipid (IPL) tetraethers was observed exclusively in the anoxic hypolimnion during stratification, confirming anoxia as a key trigger for IPL tetraether production.” – If so, how should we interpret the IPL brGDGTs data shown in Figure 3D for the epilimnion?
494-495: “The production of IPL brGDGTs in the hypolimnion is limited to anoxic conditions. This finding unequivocally highlights the role of anoxia as a key trigger for in-situ IPL brGDGT production.” – However, in Figure 1, bottom water anoxic conditions also occurred in June and November, yet IPL brGDGTs were not produced. What is the reason for this?
499: “… brGDGT IIIa’’, a compound which was not observed in Rotsee)” – The compound brGDGT IIIa’’ is mentioned here but is not shown in Figure S1, nor is there any explanation for it. Is it necessary to mention it here at all?
531: “During the epilimnion mixing season, a decrease in brGDGT Ia and an increase in
brGDGT IIIa are observed, reflecting the GDGT distribution found in the hypolimnion (Fig. 3C)”. – When is the mixing season in Figure 3? It would be helpful to add this information to Figure 3, as mentioned in Figure 1: (i) isothermal mixing (December and February), (ii) stratification onset (June), (iii) stratified water column (August and September), and (iv) post-stratification conditions (November).
537-539” “However, although the increase in concentration of Ia is observed in warm stratified months in the epilimnion, the absence of a correlation between Ia and temperature during colder months, contributes to the non-significant dependency between MBT’5ME and temperature (r= 0.59, p= 0.10).” – I guess in this text, Ia is referring to Ia-CL in Figure 3A. This should be clarified first. Nonetheless, in my view, it is not clear in Figure 3A how the concentration of Ia increases during the warm, stratified months in the epilimnion.
542-544: “Furthermore, the negative loadings of brGDGT IIIa’ on epilimnion PCA axis 1 (Ia: -0.24, IIIa': -0.28) align with the loading of the temperature vector (Fig. 7B).” – In my view, this statement does not fit with the figure shown.
573-574: “however, dissolved oxygen content (and conductivity and alkalinity to a lesser extent) seems to drive increases in cyclopentane-containing and 6-methyl brGDGTs (Supp. Table S2).” – However, there was no relationship between IR and conductivity in Table S2.
577: “… as well as the IR, CBT’ and DC’ (r= -0.65 to -0.78, p< 0.05).” – delete IR in this sentence.
577-588: “This correlation is also observed in the IR (r=-0.64, r= 0.66, p< 0.05, for dissolved oxygen and alkalinity respectively).” – r=-0.64 should be r=-0.64 in Table S2.
637-638: “While CL brGDGTs are produced throughout the water column, the production of IPL brGDGTs seems confined to the anoxic hypolimnion.” Does this mean that IPL brGDGTs are produced specifically in the anoxic hypolimnion, implying that they are not produced in the epilimnion or other oxygenated parts of the water column? If so, how should we interpret the IPL brGDGTs data shown in Figure 3D for the epilimnion? This point should be clarified to avoid any confusion.
In many places in the text, "GDGT" was used instead of "brGDGTs." It would be better to use "brGDGTs" consistently.
For "Rotsee" or "Lake Rot," it would be best to use one term consistently throughout the text and figure captions.
Fig. 7: In the figures, the notation for A), B), and C) is missing, and the colors of the circles are difficult to distinguish.
Fig. 8. The layer legend (i.e., 1-Surface, 2-Bottom) can be replaced with 1-Epilimnion and 2-Hypolimnion to ensure consistency.
Fig. 9: For figure, using different symbols for soil and surface sediment would be helpful.
Citation: https://doi.org/10.5194/egusphere-2024-3052-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 179 | 58 | 29 | 266 | 38 | 3 | 4 |
- HTML: 179
- PDF: 58
- XML: 29
- Total: 266
- Supplement: 38
- BibTeX: 3
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1