the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Endogenic methylmercury in a eutrophic lake during the formation and decay of seston
Laura Balzer
Carluvy Baptista-Salazar
Sofi Jonsson
Harald Biester
Abstract. Anoxic microniches in sinking particles in lakes have been identified as important water phase production zones of monomethylmercury (MeHg) (endogenic MeHg). However, the production and decay of MeHg during organic matter (OM) decomposition in the water column and its relation to the total Hg concentration in seston are poorly understood. We investigated Hg speciation and chemical changes in sinking seston in a small and shallow (12-m-deep) eutrophic lake during phytoplankton blooms from April to November 2019. The results show that MeHg proportions are high in seston at the water surface (up to 22 %) and at the oxic-suboxic redox boundary (up to 26 %). During suboxic OM decomposition, and with decreasing redox-potential, the concentration and proportion of MeHg in seston strongly decrease (< 0.5 %) as the water depth increases. Under these conditions, total Hg concentrations show a 3.8 to 26-fold increase. In the hypolimnion environment, changes in MeHg proportions were minimal in sinking seston, and samples collected by sediment traps had MeHg values similar to those measured at the sediment-water interface, though higher MeHg concentrations were found deeper in the sediment. Our results indicate that cycling of MeHg and total Hg (THg) in seston within small productive lakes is largely controlled by the decomposition processes of settling seston and that the endogenic MeHg pool appears to be largely disconnected from the sedimentary MeHg pool.
- Preprint
(1311 KB) -
Supplement
(1754 KB) - BibTeX
- EndNote
Laura Balzer et al.
Status: closed
-
RC1: 'Comment on bg-2022-170', Anonymous Referee #1, 29 Sep 2022
General comments on the manuscript
The paper aims to assess the role of seston in the production of methylmercury in a eutrophic lake. The paper is based on water, seston, trap and sediment sampling at seven dates between April and November 2019.
The paper is well written and structured, methods thoroughly described, and results generally well presented. In general, I like this paper, but I find that there are weaknesses in the design of the research and then over-interpretation of some results.
- MeHg has not been measured in the dissolved phase, then there is no clear picture of the overall situation. Based on partition coefficients, and on the concentration of seston in water column, it appears that usually most the MeHg in the raw water column is in the dissolved (or colloidal phase) (see Gallorini and Loizeau 2022, Chemosphere).
- The authors rule out the possibility of diffusion of MeHg from the sediment, without clear evidence, as there is no measurement of seston and MeHg in the water column below RTZ. They invoke “that mixing is minor during times of high productivity (line 265)”, however diffusion seems to occur and shown by the Mn profiles.
- The “pronounced” maximum concentration at the RTZ that is at the base of the all discussion and interpretation is not so evident. On line 250 it reads “During periods in which the RTZ was clearly defined, MeHg concentrations in seston showed a pronounced maximum at the RTZ that did not occur in April, when no RTZ was observed (Fig. 3).” The pronounced maximum is clear only in Aug. 19.
- Some discussed variations are very small and probably within uncertainties (e.g. C/N ratio). But the authors do not give uncertaintiy of the measurements, so it is impossible to evaluate the relevance of the variations.
Detailed remarks
L25 “The methylation of inorganic divalent forms of Hg(Hg(II)) to toxic MeHg is carried out…”. The sentence implies that Hg(II) is not toxic, which is not the case.
L 29 “ being influenced by temporal and spatial variabilities”. It is not clear to which processes these variabilities refer.
L 32 all these references on marine environment (19) are too much. Better to select the most relevant for your purpose.
L50 About MeHg formation in lake snow, see Gallorini and Loizeau 2022, Chemosphere.
L84 As a pump and tubing have been used to sample water and seston, how potential contaminations (mainly for disssolved THg in water) have been evaluated?
L89 text reads “PE Falcon tubes for.” The end of the sentence is missing.
L94 text reads “but in most cases covered the upper 4 m”. However most figures indicate that the lower sample is at 5 m depth. Moreover, it is not clear why samples below 5 m were not collected.
L97 Is electrical conductivity corrected for temperature? Explain how it and other parameters were measured ? From CTD or on the boat?
L99 The exposure time of the sediment trap (141 days) is very long, and then the material experience early diagenesis if no preservative was added. Then it is not clear why this sample was collected.
L162 change “the” to “then”
L182 and following. It should be better stressed how the author link parameters to productivity. For instance, L188 text reads “Chlorophyll a concentrations were 2.5 to 2.8 8 μg, indicative of low phytoplankton productivity.” Chl a is not a measurement of the productivity, as other factors may change the Chl a concentrations (for instance grazing). Chl a may be a direct proxy of algal biomass, not productivity.
L202 the profile of Fe in May is strange, as Fe(III) is essentially insoluble. So what is the "dissolved" species found in the upper layer in May? Then what happened in August 19, Fe dropped from > 500 to 100 ug/L and increase again > 500 in September.
L203 text reads “After mixing in November, the Mn and Fe concentrations were uniformly low”. From Fig 1, Mn isn’t low in November, with values much higher than in surface waters measured the other months.
L207 C/N ratio compare organic carbon to organic nitrogen in samples. Is all C in the sample from organic matter? For instance, the sediment trap results indicate C concentration of ~9%, that is 18 to max 30% of the sediment is organic matter. What is the composition of the remaining 70% of the sediment? Does it contain C as carbonates? This point should be clarified.
L213 The decrease of C/N ratio explained by mineralization is not obvious. A reference is needed here, as usually it is the reverse that is observed as mentioned the given reference Meyers and Lallier Vergès 1999. Moreover, is the decomposition the only processes, what about selected grazing or change in composition of the seston (phyto vs. zooplankton) to explain the C/N variation?
L260 text reads “This would explain the continuous increase in MeHg concentrations with depth…” What is the explanation? The absence of microniches does not explain the formation of MeHg at depth, where O2 saturation is still > 20%. Diffusion from sediments?
L294 Mass loss is the only explanation of the THg increase with depth. However, C concentration decrease by max a factor 3.9, whereas THg increase is a factor 26. Then the mass loss cannot account totally for the increase in THg concentration.
L314 text reads “The sulphide produced may form insoluble complexes with Hg (Shanks and
Reeder, 1993; Bianchi et al., 2018), such as Hg sulphides (HgS), meaning that Hg becomes less available for methylation” It is not so clear that the presence of S decreases the bioavailability of Hg. Barrouilhet et al 2022 ESPR show that methylation potential increases with S concentration before to decrease at high S concentration.
L318 text reads “Thus, THg and MeHg fluxes to the sediment are largely determined by changes in OM composition and mass loss during decomposition.” While these processes may change the MeHg fluxes, it is not clear why these processes change the flux of THg: i) if the authors are correct, the increase in THg concentration is due to mass loss in OM, then the quantity of THg remain the same, so the flux, and transformation of THg to MeHg will not change the flux of THg as MeHg is included in THg.
L341 MeHg concentrations in the sediments are not sufficient to assess fluxes from sediments to interstitial water to overlying water.
L354 “Water column MeHg formation and degradation in eutrophic lakes appears to be intense and occurs rapidly and at rates similar to what we observed within the bottom sediments” This statement is not supported by the data/discussion. No rate has been determined neither in the seston nor in sediments.
Fig 1. The scales do not cover the entire range of the results.
Citation: https://doi.org/10.5194/bg-2022-170-RC1 -
AC1: 'Reply on RC1', Laura Balzer, 18 Nov 2022
We would like to thank reviewer 1 for the detailed comments and suggestions to improve the quality of our work.
The paper aims to assess the role of seston in the production of methylmercury in a eutrophic lake. The paper is based on water, seston, trap and sediment sampling at seven dates between April and November 2019.
The paper is well written and structured, methods thoroughly described, and results generally well presented. In general, I like this paper, but I find that there are weaknesses in the design of the research and then over-interpretation of some results. MeHg has not been measured in the dissolved phase, then there is no clear picture of the overall situation. Based on partition coefficients, and on the concentration of seston in water column, it appears that usually most the MeHg in the raw water column is in the dissolved (or colloidal phase) (see Gallorini and Loizeau 2022, Chemosphere).
We thank the reviewer for this positive assessment.
We are aware that analyses of MeHg in the dissolved phase would have been useful. However, our paper is focused on seston and specifically the fate of MeHg during decay of algae derived organic matter. Our intend was not to resolve the entire biogeochemical Hg/MeHg cycle in this lake. We further believe, that the situation is different in eutrophic lakes compared to oligotrophic lakes (such as lake Geneva in Gallorini and Loizeau 2022) regarding the partition of MeHg between the dissolved and the solid phase as there is so much more organic matter present during algae blooms that the dissolved phase is of minor importance here. Previous studies have shown, that dissolved Hg is depleted after algae blooms due to water phase Hg scavenging by sinking seston (Schütze et al, 2021).
The authors rule out the possibility of diffusion of MeHg from the sediment, without clear evidence, as there is no measurement of seston and MeHg in the water column below RTZ. They invoke “that mixing is minor during times of high productivity (line 265)”, however diffusion seems to occur and shown by the Mn profiles. The “pronounced” maximum concentration at the RTZ that is at the base of the all discussion and interpretation is not so evident. In line 250 it reads “During periods in which the RTZ was clearly defined, MeHg concentrations in seston showed a pronounced maximum at the RTZ that did not occur in April, when no RTZ was observed (Fig. 3).” The pronounced maximum is clear only in Aug. 19.
The reason why there is no data from below the RTZ in some of the profiles is that there was not enough suspended matter below the RTZ which could be sampled with our method (25 µm net several 2 hours pumping) (see L254). From our data in the solid phase, we assumed that MeHg diffussion from the sediment is unlikely, but we agree with the reviewer that we cannot rule this out. We have now added the depth profiles of DOC from the different sampling days. MeHg released from bottom sediments is most likely bound to DOM as chloride concentrations in lakes are too low to be competetive. The DOC profiles clearly indicate that DOC release from the sediment occurs as indicated by the highest DOC concentrations found in the deepest water samples and it is likey that MeHg released from decaying organic matter in the uppermost sediment layers is bound to DOM and distributed in the water column during lake mixing. However, DOC profiles do not show diffussion gradients during the summer months when the algae blooms occur and concentrations were even higher in the upper water layers indicating DOC release from decomposing algae organic matter which suggests rather MeHg formation in the water phase (labile algae derived DOM supports microbial MeHg formation in the water phase (Bravo et al....)) than uptake of MeHg released from the sediment although both is possible. We will revise the manuscript accordingly.
Some discussed variations are very small and probably within uncertainties (e.g. C/N ratio). But the authors do not give uncertainty of the measurements, so it is impossible to evaluate the relevance of the variations.
We will add uncertainities of the measurements, which, however, cannot explain the observed variations and trends in the data. We will also tone the interpretation of the relatively small variations of C/N ratios in the upper water layers down and will focus on the changes in the RTZ.
Detailed remarks
L25 “The methylation of inorganic divalent forms of Hg(Hg(II)) to toxic MeHg is carried out…”. The sentence implies that Hg(II) is not toxic, which is not the case.
We agree. Will be changed accordingly.
L 29 “ being influenced by temporal and spatial variabilities”. It is not clear to which processes these variabilities refer.
We refer to the variability in redox-conditions. Will be revised
L 32 all these references on marine environment (19) are too much. Better to select the most relevant for your purpose.
Will be changed accordingly.
L50 About MeHg formation in lake snow, see Gallorini and Loizeau 2022, Chemosphere.
Reference will be added
L84 As a pump and tubing have been used to sample water and seston, how potential contaminations (mainly for disssolved THg in water) have been evaluated?
Pump and tubing has been cleaned (acid washed) thoroughly. Blanks will be added.
L89 text reads “PE Falcon tubes for.” The end of the sentence is missing.
Will be corrected
L94 text reads “but in most cases covered the upper 4 m”. However most figures indicate that the lower sample is at 5 m depth. Moreover, it is not clear why samples below 5 m were not collected.
Will be changed to 5 m. We were not able to gain sufficient material for solid phase analyses during 2 h sampling/ pumping as the amount of suspended matter below 5 m was, in most cases, very low. Longer pumping was not possible due to overheating of the pumps etc.
L97 Is electrical conductivity corrected for temperature? Explain how it and other parameters were measured? From CTD or on the boat?
Water parameters have been measured on the boat and are corrected for temperature.
L99 The exposure time of the sediment trap (141 days) is very long, and then the material experience early diagenesis if no preservative was added. Then it is not clear why this sample was collected.
The idea of the sediment-trap approach was to get an idea what the integrated material looks like after a period of some months of decomposition and if/how it differs from bottom sediments.
L162 change “the” to “then”
Will be changed
L182 and following. It should be better stressed how the author link parameters to productivity. For instance, L188 text reads “Chlorophyll a concentrations were 2.5 to 2.8 8 μg, indicative of low phytoplankton productivity.” Chl a is not a measurement of the productivity, as other factors may change the Chl a concentrations (for instance grazing). Chl a may be a direct proxy of algal biomass, not productivity.
We agree, will be changed
L202 the profile of Fe in May is strange, as Fe(III) is essentially insoluble. So what is the "dissolved" species found in the upper layer in May? Then what happened in August 19, Fe dropped from > 500 to 100 ug/L and increase again > 500 in September.
We did not analyse dissolved Fe-species, but we assume that the small amount of dissolved Fe found in the upper water layers is organically bound Fe, probably release during algae matter decay or from zooplankton. DOM-Fe is soluble under oxic conditions. The appearance of dissolved (reduced) Fe changes in the deep water layers between August 12, 19 and Sep.02 are most likely due to a change in redox zonation caused by more or less amount of suspended organic matter and differences in productivity/amount of algae biomass produced. Note, that pH is higher on Aug.12 = higher productivity compared to August 19. Similar, pH and Chl a at Sept 02 is higher (higher productivity) than on Aug. 19.
L203 text reads “After mixing in November, the Mn and Fe concentrations were uniformly low”. From Fig 1, Mn isn’t low in November, with values much higher than in surface waters measured the other months.
We agree, although the message is clear that there is no more redox zonation. Text will be changed.
L207 C/N ratio compare organic carbon to organic nitrogen in samples. Is all C in the sample from organic matter? For instance, the sediment trap results indicate C concentration of ~9%, that is 18 to max 30% of the sediment is organic matter. What is the composition of the remaining 70% of the sediment? Does it contain C as carbonates? This point should be clarified.
Will be clarified. The remaining material in the traps is mineral matter. There is no carbonate formation in the lake. In addition, samples have been decarbonated prior to carbon analysis. In addition, in the deep layers where the sediment trap was installed the lower pH will cause dissolution of calcite.
L213 The decrease of C/N ratio explained by mineralization is not obvious. A reference is needed here, as usually it is the reverse that is observed as mentioned the given reference Meyers and Lallier Vergès 1999. Moreover, is the decomposition the only processes, what about selected grazing or change in composition of the seston (phyto vs. zooplankton) to explain the C/N variation?
We agree with the reviewer that changes in C/N ratio above the RTZ are probably too small to undoubtedly indicate organic matter decomposition. We also agree, that some of the small changes seen here could have been caused by the occurrence of zooplankton. We will therefore restrict the interpretation of C/N ratios as a measure for organic matter decomposition to the values within or below the RTZ.
L260 text reads “This would explain the continuous increase in MeHg concentrations with depth…” What is the explanation? The absence of microniches does not explain the formation of MeHg at depth, where O2 saturation is still > 20%. Diffusion from sediments?
This sentence is indeed not clear and will be clarified. We believe that the increase in MeHg concentrations with depth in April is mainly caused by mass loss due to progressive organic matter decomposition (comparable to what has been described by Gallorini et al, 2022) although we cannot exclude MeHg formation by Mn reducing bacteria.
L294 Mass loss is the only explanation of the THg increase with depth. However, C concentration decrease by max a factor 3.9, whereas THg increase is a factor 26. Then the mass loss cannot account totally for the increase in THg concentration.
We are not sure if we understand this comment correctly. THg and C do not necessarily have to increase by the same exent because C (and other elements) is lost during mineralisation, but Hg is not. An additional explanation might be that some Hg released to the water phase during organic matter decomposition is scavenged by sinking seston as it has been observed in marine studies.
L314 text reads “The sulphide produced may form insoluble complexes with Hg (Shanks and Reeder, 1993; Bianchi et al., 2018), such as Hg sulphides (HgS), meaning that Hg becomes less available for methylation” It is not so clear that the presence of S decreases the bioavailability of Hg. Barrouilhet et al 2022 ESPR show that methylation potential increases with S concentration before to decrease at high S concentration.
Our data indicates that there is no sulfate reduction and thus formation of sulphide. The increase in S concentration in sesteon is thus rather due to mass loss during organic matter decomposition. This was an assumption which we could not proof in the frame of this study (only based on the increase of S concentration). The amount of material gained was too small to do Hg-thermo-desorption analyses or similar. The study of Barrouilhet et al 2022 is quite different from what we did and we could hardly say if their findings do apply here.
L318 text reads “Thus, THg and MeHg fluxes to the sediment are largely determined by changes in OM composition and mass loss during decomposition.” While these processes may change the MeHg fluxes, it is not clear why these processes change the flux of THg: i) if the authors are correct, the increase in THg concentration is due to mass loss in OM, then the quantity of THg remain the same, so the flux, and transformation of THg to MeHg will not change the flux of THg as MeHg is included in THg.
We agree, will be revised
L341 MeHg concentrations in the sediments are not sufficient to assess fluxes from sediments to interstitial water to overlying water.
We agree, statement will be toned down
L354 “Water column MeHg formation and degradation in eutrophic lakes appears to be intense and occurs rapidly and at rates similar to what we observed within the bottom sediments” This statement is not supported by the data/discussion. No rate has been determined neither in the seston nor in sediments.
We agree will be changed or removed
Fig 1. The scales do not cover the entire range of the results.
Will be adapted
Citation: https://doi.org/10.5194/bg-2022-170-AC1
-
RC2: 'Comment on bg-2022-170', Anonymous Referee #2, 03 Nov 2022
General comments.
This paper builds on several prior studies that show that the water column of lakes and oceans can be an important site for MeHg formation. It differs from most water column studies by focusing on a eutrophic urban lake and by specifically targeting MeHg abundance in bulk seston at different depths and dates for clues about formation and decay mechanisms. Unfortunately, the sampling technique lumped zooplankton in with seston, potentially introducing bias due to biomagnification. And the sampling scheme was spatially inconsistent, which makes the comparison of depth profiles on different dates difficult. The reason that the entire water column was sampled on one date and only the upper water column on most other dates is unexplained, and it compromises the authors’ conclusions about what’s going on as particles sink (especially in the hypolimnion since it was rarely sampled). Among other things (below), the authors need to justify their sampling methods and revisit the interpretation of changes in Hg speciation across depth and time. They also need to reconsider conclusions about links between climate change, productivity and bioaccumulation. This will require major revision.
Specific comments.
- The term “endogenic” should be reconsidered. It means “within the system”, which for lakes technically includes sediments. “Water column” would be better, unless they mean “within the seston” – in which case the title and text need to be re-worded
- Line 89 is an incomplete sentence
- Line 90: why a 25um net? It would allow many cyanophytes and chlorophytes to pass through, and bias collection toward zooplankton (which are not “seston”). Why not a clean pump-and-sieve/filter system instead?
- L220-225. The seston samples collected on those dates are not really much closer to the sediment surface. There’s just one hypo sample and it’s directly beneath the RTZ. You’d need to sample more depths to justify. Revise.
- L235. But peak concentrations of MeHg in seston occur in the suboxic RTZ on 4 of the 5 dates when the lake was strongly stratified. On the remaining date, seston MeHg concentrations are highest in the upper hypolimnion. During stratification, MeHg is never highest in the oxic epilimnion. If anything, these finding suggest that MeHg production is associated with microbial respiratory pathways that are less energy efficient than O2 reduction (e.g. sulfate reduction, Fe reduction). Revise.
- L240-245. Alternatively, low MeHg during high productivity may reflect biodilution in the larger phytoplankton biomass (i.e. parental seston). Lacking sound data, one can’t distinguish zooplankton bias from biodilution in microplankton, and neither necessarily point to sestonic microniches. Revise
- L255-263. They could also be explained by the presence of free-water microbes that possess the methylation gene pair hgcAB and occupy the O/A boundary. DOM rather than POM could be their carbon source. Revise.
- L275-284. Sestonic MeHg in the 20% range is not atypical for unpolluted temperate lakes. What’s unusual is the very low %MeHg in April
- L346. Actually, this was first shown in Little Rock Lake, which is only 10m deep (but the eutrophic part may be right).
- L346-end. Note that the range of Hg and MeHg in the seston of this eutrophic lake is on the low end of seston data reported for mesotrophic to oligotrophic North American lakes, both for MeHg concentration and %MeHg. High productivity is not necessarily conducive to abnormally high rates of MeHg accumulation in bioseston. In fact, most data suggest the opposite due to biodilution. It may be true that higher amounts of OM decomposition in eutrophic lakes does indeed exacerbate O2 depletion and enhance methylation in suboxic water, but that was not measured here. It seems that the most you can say with the data presented here is that the opposing forces of high biodilution and high decomposition need to be reconciled before addressing the impact of climate change. Revise
Citation: https://doi.org/10.5194/bg-2022-170-RC2 -
AC2: 'Reply on RC2', Laura Balzer, 18 Nov 2022
We would like to thank reviewer 2 for the constructive and helpful comments.
General comments.
This paper builds on several prior studies that show that the water column of lakes and oceans can be an important site for MeHg formation. It differs from most water column studies by focusing on a eutrophic urban lake and by specifically targeting MeHg abundance in bulk seston at different depths and dates for clues about formation and decay mechanisms. Unfortunately, the sampling technique lumped zooplankton in with seston, potentially introducing bias due to biomagnification. And the sampling scheme was spatially inconsistent, which makes the comparison of depth profiles on different dates difficult. The reason that the entire water column was sampled on one date and only the upper water column on most other dates is unexplained, and it compromises the authors’ conclusions about what’s going on as particles sink (especially in the hypolimnion since it was rarely sampled). Among other things (below), the authors need to justify their sampling methods and revisit the interpretation of changes in Hg speciation across depth and time. They also need to reconsider conclusions about links between climate change, productivity and bioaccumulation. This will require major revision.
We agree with reviewer that there might be potential bias in the interpretation of our MeHg depth profile. Our focus was on bulk seston and to our knowledge zooplankton is part of seston. We also believe, that the distribution of zooplankton alone cannot explain the THg and MeHg depth profiles in our lake. The separation of phyto- and zooplankton is useful in studies on biomagnification, which was not our topic. In this case, a qualitative separation of both fraction in small amounts is sufficient. In case of bulk seston sampling, eg. by means of a pump-and-sieve/filter system (0.45 µm?) during algae blooms as suggested by the reviewer such qualitative separation is nearly impossible (agglomeration) if larger sample volume is needed. Moreover, we believe that our interpretation of the distribution of MeHg and THg in the water phase is supported by our data on algae biomass, (Chl 1, pH) is sound. We regret that our sampling was imperfect, we made a lot of effort to gain in all cases sufficient material, however, this is a natural system with sometimes unpredictable changes of conditions. The reason why sampling is inconsistent through time and space is that we could not get sufficient material from the hypolimnion within the possible pumping time (~ 2 h per layer). We already mentioned this in the text but will explain this in more detail (see L254).
We will tone down on our conclusion regarding the link between climate change, productivity and bioaccumulation. We assume that the reviewer based this comments on his assumption that we mainly see biodilution. We have commented on this below.
Specific comments.
The term “endogenic” should be reconsidered. It means “within the system”, which for lakes technically includes sediments. “Water column” would be better, unless they mean “within the seston” – in which case the title and text need to be re-worded
This term has been introduced in other studies. For example, in Gallorini et al. 2021: We will define it within the manuscript as “production within the water column”.
Line 89 is an incomplete sentence
Will be changed
Line 90: why a 25µm net? It would allow many cyanophytes and chlorophytes to pass through, and bias collection toward zooplankton (which are not “seston”). Why not a clean pump-and-sieve/filter system instead?
We agree with the reviewer that some of the small fraction of phytoplankton might have got lost during water pumping through a 25 µm net. We tried a pump-and –sieve filter system before. However, this took to long to gain sufficient material from each water layer to do the solid phase analyses needed here at a resolution of 1 m within a single day (filter clogging etc.). Because of this, we decided to pump the water through a 25 µm net. Although it would have been the best option to sample all phytoplankton fraction, we believe that the lack of the fraction < 25 µm has no significant influence of the overall results and conclusions of this study (it just means more phytoplankton). To our knowledge zooplankton is part of the seston, too.
L220-225. The seston samples collected on those dates are not really much closer to the sediment surface. There’s just one hypo sample and it’s directly beneath the RTZ. You’d need to sample more depths to justify. Revise.
The reviewer is right, the deepest sample was taken just below the RTZ in those months. We will revise this statement. As explained above. We were not able to gain sufficient material from deeper layers in those month within the possible pumping time (2 h)
L235. But peak concentrations of MeHg in seston occur in the suboxic RTZ on 4 of the 5 dates when the lake was strongly stratified. On the remaining date, seston MeHg concentrations are highest in the upper hypolimnion. During stratification, MeHg is never highest in the oxic epilimnion.
We agree that this sentence is misleading and will clarify this section.
If anything, these finding suggest that MeHg production is associated with microbial respiratory pathways that are less energy efficient than O2 reduction (e.g. sulfate reduction, Fe reduction). Revise.
Not clear what the reviewer means here and what should be revised. We discussed in the ms that MeHg formation is releated to redox conditions in the water column specifically to Mn reduction (similar to what has been shown by Petersen et. al 2020(in a lake) and by Kohler et la., 2022 (in the arctic ocean) Fe reduction is of minor importance in this lake and sulphate reduction does not take place in the water column (compare Fig. S3). Data on eutrophic lakes are rare and to our knowledge not available yet at similar high resolution. A major aim of this study is to show changes in MeHg in seston at this comparatively high temporal and spatial (depth) resolution to understand the evolution of MeHg and THg concentrations and proportion during sinking through the water column.
L240-245. Alternatively, low MeHg during high productivity may reflect biodilution in the larger phytoplankton biomass (i.e. parental seston). Lacking sound data, one can’t distinguish zooplankton bias from biodilution in microplankton, and neither necessarily point to sestonic microniches. Revise
We agree with the reviewer that the variation in MeHg concentration might be a result of biodilution, zooplankton influence and/ or the formation of sestonic microniches. Based on ph, O2 and chlorophyll a concentation, we have classified the sampling days of May and November as periods of lower productivity and the sampling days in April and between June to September as periods of higher productivity (L. 181-190). Our data shows that the time of lower productivity is likley accompanied by higher amounts of zooplankton (higher N concentrations and lower C/N ratios in the upper two meters) and the time of higher productivity is likley accompanied by lower amounts of zooplankton (lower N concentraions and higher C/N ratios in the upper two meters) (L 238-246; Fig. 4). But we agree with the reviewer that we cannot fully rule out that the lower MeHg concentrations in times of higher productivity are caused to a certain extent by biodilution and vise versa the higher MeHg concentrations in times of lower productivity by bioconcentration. But we believe that biodilution/bioconcentration is not the major process in our lake as our data point to zooplankton influence between the sampling days. If biodilution/bioconcentration would be the the major process determining the MeHg concentration in our seston samples, we would also see a dilution or an increase in seston THg concentration, which we do not. The THg concentration do not change in the same way as the MeHg concentration. So there has to be another process than biodilution/bioconcentration controlling the MeHg concentration. However, as disscused in the ms the distribution of zooplankton -alone cannot explain the THg and MeHg depth profiles in our lake (L267-275). Obviously, we cannot completely exclude that zooplankton contributes to the maximum MeHg concentration within the RTZ. Our idea is that more organic matter rich seston (as a result of algae blooms) favours MeHg formation in sestonic unoxic microniches as oxic microbial pathways of MeHg formation are not yet known and MeHg get enriched during decay of the phytoplankton-dominated organic material. We will revise this section to make our point clearer.
L255-263. They could also be explained by the presence of free-water microbes that possess the methylation gene pair hgcAB and occupy the O/A boundary. DOM rather than POM could be their carbon source. Revise.
We discussed this point in L 266. It is likely that free-water microbial Hg methylation occurs, specifically because there is predominantly easy accessible DOM in the water column. But to our knowledge, oxic microbial pathways of MeHg formation are not yet known. Many papers point to methylation within anoxic microniches. However, our data suggest that free-water microbial Hg methylation is rather not the dominant process here as high MeHg concentration only occur during times of a pronounced RTZ (compare April when production is already high but MeHg is low because redox-zonation is not yet established) We will try to make this point clearer in the ms.
L275-284. Sestonic MeHg in the 20% range is not atypical for unpolluted temperate lakes. What’s unusual is the very low %MeHg in April
We can tone this statement down. However, we think that 20 % is a lot regarding the high biomass and that there is no influence from soil derived DOC/MeHg-rich inflow, which is typical for many oligotrophic lakes. We believe that the low MeHg proportions and concentrations in April (compared to the summer months) supports our conclusion that the MeHg is predominantly formed in the water phase along redox-gradients/micro-niches and the role of the RTZ. In April, algae biomass and productivity as indicated by high Chl a and high pH is already high, but the redox gradients are only weakly developed (only weak Mn-reduction) (Fig. S1), most likely because organic matter decomposition in the water column is still low. If most MeHg is originated from the sediment we should see this MeHg in the seston also in April.
We will try to make this point clearer in the ms.
L346. Actually, this was first shown in Little Rock Lake, which is only 10m deep (but the eutrophic part may be right).
We will refer to this point in the ms. However, we actually pointed out that our study is focused on eutrophic lakes, where data is rare.
L346-end. Note that the range of Hg and MeHg in the seston of this eutrophic lake is on the low end of seston data reported for mesotrophic to oligotrophic North American lakes, both for MeHg concentration and %MeHg. High productivity is not necessarily conducive to abnormally high rates of MeHg accumulation in bioseston. In fact, most data suggest the opposite due to biodilution. It may be true that higher amounts of OM decomposition in eutrophic lakes does indeed exacerbate O2 depletion and enhance methylation in suboxic water, but that was not measured here. It seems that the most you can say with the data presented here is that the opposing forces of high biodilution and high decomposition need to be reconciled before addressing the impact of climate change. Revise
We will revise this section, although it is not clear to us what exactly the reviewer will tell us here. We speculate that there would be much lower MeHg in this lake without high productivity regarding the total amount of MeHg formed in this lake. See answer above regarding biodilution. We want to show the distribution and changes in THg and MeHg concentration and proportion in seston with depth and how they are controlled by redox-conditions and decomposition under eutrophic conditions. How many studies on North American lakes show the distribution of THg and MeHg in seston at high temporal and spatial resolution including redox conditions, Chl a data etc.?.
Citation: https://doi.org/10.5194/bg-2022-170-AC2
Status: closed
-
RC1: 'Comment on bg-2022-170', Anonymous Referee #1, 29 Sep 2022
General comments on the manuscript
The paper aims to assess the role of seston in the production of methylmercury in a eutrophic lake. The paper is based on water, seston, trap and sediment sampling at seven dates between April and November 2019.
The paper is well written and structured, methods thoroughly described, and results generally well presented. In general, I like this paper, but I find that there are weaknesses in the design of the research and then over-interpretation of some results.
- MeHg has not been measured in the dissolved phase, then there is no clear picture of the overall situation. Based on partition coefficients, and on the concentration of seston in water column, it appears that usually most the MeHg in the raw water column is in the dissolved (or colloidal phase) (see Gallorini and Loizeau 2022, Chemosphere).
- The authors rule out the possibility of diffusion of MeHg from the sediment, without clear evidence, as there is no measurement of seston and MeHg in the water column below RTZ. They invoke “that mixing is minor during times of high productivity (line 265)”, however diffusion seems to occur and shown by the Mn profiles.
- The “pronounced” maximum concentration at the RTZ that is at the base of the all discussion and interpretation is not so evident. On line 250 it reads “During periods in which the RTZ was clearly defined, MeHg concentrations in seston showed a pronounced maximum at the RTZ that did not occur in April, when no RTZ was observed (Fig. 3).” The pronounced maximum is clear only in Aug. 19.
- Some discussed variations are very small and probably within uncertainties (e.g. C/N ratio). But the authors do not give uncertaintiy of the measurements, so it is impossible to evaluate the relevance of the variations.
Detailed remarks
L25 “The methylation of inorganic divalent forms of Hg(Hg(II)) to toxic MeHg is carried out…”. The sentence implies that Hg(II) is not toxic, which is not the case.
L 29 “ being influenced by temporal and spatial variabilities”. It is not clear to which processes these variabilities refer.
L 32 all these references on marine environment (19) are too much. Better to select the most relevant for your purpose.
L50 About MeHg formation in lake snow, see Gallorini and Loizeau 2022, Chemosphere.
L84 As a pump and tubing have been used to sample water and seston, how potential contaminations (mainly for disssolved THg in water) have been evaluated?
L89 text reads “PE Falcon tubes for.” The end of the sentence is missing.
L94 text reads “but in most cases covered the upper 4 m”. However most figures indicate that the lower sample is at 5 m depth. Moreover, it is not clear why samples below 5 m were not collected.
L97 Is electrical conductivity corrected for temperature? Explain how it and other parameters were measured ? From CTD or on the boat?
L99 The exposure time of the sediment trap (141 days) is very long, and then the material experience early diagenesis if no preservative was added. Then it is not clear why this sample was collected.
L162 change “the” to “then”
L182 and following. It should be better stressed how the author link parameters to productivity. For instance, L188 text reads “Chlorophyll a concentrations were 2.5 to 2.8 8 μg, indicative of low phytoplankton productivity.” Chl a is not a measurement of the productivity, as other factors may change the Chl a concentrations (for instance grazing). Chl a may be a direct proxy of algal biomass, not productivity.
L202 the profile of Fe in May is strange, as Fe(III) is essentially insoluble. So what is the "dissolved" species found in the upper layer in May? Then what happened in August 19, Fe dropped from > 500 to 100 ug/L and increase again > 500 in September.
L203 text reads “After mixing in November, the Mn and Fe concentrations were uniformly low”. From Fig 1, Mn isn’t low in November, with values much higher than in surface waters measured the other months.
L207 C/N ratio compare organic carbon to organic nitrogen in samples. Is all C in the sample from organic matter? For instance, the sediment trap results indicate C concentration of ~9%, that is 18 to max 30% of the sediment is organic matter. What is the composition of the remaining 70% of the sediment? Does it contain C as carbonates? This point should be clarified.
L213 The decrease of C/N ratio explained by mineralization is not obvious. A reference is needed here, as usually it is the reverse that is observed as mentioned the given reference Meyers and Lallier Vergès 1999. Moreover, is the decomposition the only processes, what about selected grazing or change in composition of the seston (phyto vs. zooplankton) to explain the C/N variation?
L260 text reads “This would explain the continuous increase in MeHg concentrations with depth…” What is the explanation? The absence of microniches does not explain the formation of MeHg at depth, where O2 saturation is still > 20%. Diffusion from sediments?
L294 Mass loss is the only explanation of the THg increase with depth. However, C concentration decrease by max a factor 3.9, whereas THg increase is a factor 26. Then the mass loss cannot account totally for the increase in THg concentration.
L314 text reads “The sulphide produced may form insoluble complexes with Hg (Shanks and
Reeder, 1993; Bianchi et al., 2018), such as Hg sulphides (HgS), meaning that Hg becomes less available for methylation” It is not so clear that the presence of S decreases the bioavailability of Hg. Barrouilhet et al 2022 ESPR show that methylation potential increases with S concentration before to decrease at high S concentration.
L318 text reads “Thus, THg and MeHg fluxes to the sediment are largely determined by changes in OM composition and mass loss during decomposition.” While these processes may change the MeHg fluxes, it is not clear why these processes change the flux of THg: i) if the authors are correct, the increase in THg concentration is due to mass loss in OM, then the quantity of THg remain the same, so the flux, and transformation of THg to MeHg will not change the flux of THg as MeHg is included in THg.
L341 MeHg concentrations in the sediments are not sufficient to assess fluxes from sediments to interstitial water to overlying water.
L354 “Water column MeHg formation and degradation in eutrophic lakes appears to be intense and occurs rapidly and at rates similar to what we observed within the bottom sediments” This statement is not supported by the data/discussion. No rate has been determined neither in the seston nor in sediments.
Fig 1. The scales do not cover the entire range of the results.
Citation: https://doi.org/10.5194/bg-2022-170-RC1 -
AC1: 'Reply on RC1', Laura Balzer, 18 Nov 2022
We would like to thank reviewer 1 for the detailed comments and suggestions to improve the quality of our work.
The paper aims to assess the role of seston in the production of methylmercury in a eutrophic lake. The paper is based on water, seston, trap and sediment sampling at seven dates between April and November 2019.
The paper is well written and structured, methods thoroughly described, and results generally well presented. In general, I like this paper, but I find that there are weaknesses in the design of the research and then over-interpretation of some results. MeHg has not been measured in the dissolved phase, then there is no clear picture of the overall situation. Based on partition coefficients, and on the concentration of seston in water column, it appears that usually most the MeHg in the raw water column is in the dissolved (or colloidal phase) (see Gallorini and Loizeau 2022, Chemosphere).
We thank the reviewer for this positive assessment.
We are aware that analyses of MeHg in the dissolved phase would have been useful. However, our paper is focused on seston and specifically the fate of MeHg during decay of algae derived organic matter. Our intend was not to resolve the entire biogeochemical Hg/MeHg cycle in this lake. We further believe, that the situation is different in eutrophic lakes compared to oligotrophic lakes (such as lake Geneva in Gallorini and Loizeau 2022) regarding the partition of MeHg between the dissolved and the solid phase as there is so much more organic matter present during algae blooms that the dissolved phase is of minor importance here. Previous studies have shown, that dissolved Hg is depleted after algae blooms due to water phase Hg scavenging by sinking seston (Schütze et al, 2021).
The authors rule out the possibility of diffusion of MeHg from the sediment, without clear evidence, as there is no measurement of seston and MeHg in the water column below RTZ. They invoke “that mixing is minor during times of high productivity (line 265)”, however diffusion seems to occur and shown by the Mn profiles. The “pronounced” maximum concentration at the RTZ that is at the base of the all discussion and interpretation is not so evident. In line 250 it reads “During periods in which the RTZ was clearly defined, MeHg concentrations in seston showed a pronounced maximum at the RTZ that did not occur in April, when no RTZ was observed (Fig. 3).” The pronounced maximum is clear only in Aug. 19.
The reason why there is no data from below the RTZ in some of the profiles is that there was not enough suspended matter below the RTZ which could be sampled with our method (25 µm net several 2 hours pumping) (see L254). From our data in the solid phase, we assumed that MeHg diffussion from the sediment is unlikely, but we agree with the reviewer that we cannot rule this out. We have now added the depth profiles of DOC from the different sampling days. MeHg released from bottom sediments is most likely bound to DOM as chloride concentrations in lakes are too low to be competetive. The DOC profiles clearly indicate that DOC release from the sediment occurs as indicated by the highest DOC concentrations found in the deepest water samples and it is likey that MeHg released from decaying organic matter in the uppermost sediment layers is bound to DOM and distributed in the water column during lake mixing. However, DOC profiles do not show diffussion gradients during the summer months when the algae blooms occur and concentrations were even higher in the upper water layers indicating DOC release from decomposing algae organic matter which suggests rather MeHg formation in the water phase (labile algae derived DOM supports microbial MeHg formation in the water phase (Bravo et al....)) than uptake of MeHg released from the sediment although both is possible. We will revise the manuscript accordingly.
Some discussed variations are very small and probably within uncertainties (e.g. C/N ratio). But the authors do not give uncertainty of the measurements, so it is impossible to evaluate the relevance of the variations.
We will add uncertainities of the measurements, which, however, cannot explain the observed variations and trends in the data. We will also tone the interpretation of the relatively small variations of C/N ratios in the upper water layers down and will focus on the changes in the RTZ.
Detailed remarks
L25 “The methylation of inorganic divalent forms of Hg(Hg(II)) to toxic MeHg is carried out…”. The sentence implies that Hg(II) is not toxic, which is not the case.
We agree. Will be changed accordingly.
L 29 “ being influenced by temporal and spatial variabilities”. It is not clear to which processes these variabilities refer.
We refer to the variability in redox-conditions. Will be revised
L 32 all these references on marine environment (19) are too much. Better to select the most relevant for your purpose.
Will be changed accordingly.
L50 About MeHg formation in lake snow, see Gallorini and Loizeau 2022, Chemosphere.
Reference will be added
L84 As a pump and tubing have been used to sample water and seston, how potential contaminations (mainly for disssolved THg in water) have been evaluated?
Pump and tubing has been cleaned (acid washed) thoroughly. Blanks will be added.
L89 text reads “PE Falcon tubes for.” The end of the sentence is missing.
Will be corrected
L94 text reads “but in most cases covered the upper 4 m”. However most figures indicate that the lower sample is at 5 m depth. Moreover, it is not clear why samples below 5 m were not collected.
Will be changed to 5 m. We were not able to gain sufficient material for solid phase analyses during 2 h sampling/ pumping as the amount of suspended matter below 5 m was, in most cases, very low. Longer pumping was not possible due to overheating of the pumps etc.
L97 Is electrical conductivity corrected for temperature? Explain how it and other parameters were measured? From CTD or on the boat?
Water parameters have been measured on the boat and are corrected for temperature.
L99 The exposure time of the sediment trap (141 days) is very long, and then the material experience early diagenesis if no preservative was added. Then it is not clear why this sample was collected.
The idea of the sediment-trap approach was to get an idea what the integrated material looks like after a period of some months of decomposition and if/how it differs from bottom sediments.
L162 change “the” to “then”
Will be changed
L182 and following. It should be better stressed how the author link parameters to productivity. For instance, L188 text reads “Chlorophyll a concentrations were 2.5 to 2.8 8 μg, indicative of low phytoplankton productivity.” Chl a is not a measurement of the productivity, as other factors may change the Chl a concentrations (for instance grazing). Chl a may be a direct proxy of algal biomass, not productivity.
We agree, will be changed
L202 the profile of Fe in May is strange, as Fe(III) is essentially insoluble. So what is the "dissolved" species found in the upper layer in May? Then what happened in August 19, Fe dropped from > 500 to 100 ug/L and increase again > 500 in September.
We did not analyse dissolved Fe-species, but we assume that the small amount of dissolved Fe found in the upper water layers is organically bound Fe, probably release during algae matter decay or from zooplankton. DOM-Fe is soluble under oxic conditions. The appearance of dissolved (reduced) Fe changes in the deep water layers between August 12, 19 and Sep.02 are most likely due to a change in redox zonation caused by more or less amount of suspended organic matter and differences in productivity/amount of algae biomass produced. Note, that pH is higher on Aug.12 = higher productivity compared to August 19. Similar, pH and Chl a at Sept 02 is higher (higher productivity) than on Aug. 19.
L203 text reads “After mixing in November, the Mn and Fe concentrations were uniformly low”. From Fig 1, Mn isn’t low in November, with values much higher than in surface waters measured the other months.
We agree, although the message is clear that there is no more redox zonation. Text will be changed.
L207 C/N ratio compare organic carbon to organic nitrogen in samples. Is all C in the sample from organic matter? For instance, the sediment trap results indicate C concentration of ~9%, that is 18 to max 30% of the sediment is organic matter. What is the composition of the remaining 70% of the sediment? Does it contain C as carbonates? This point should be clarified.
Will be clarified. The remaining material in the traps is mineral matter. There is no carbonate formation in the lake. In addition, samples have been decarbonated prior to carbon analysis. In addition, in the deep layers where the sediment trap was installed the lower pH will cause dissolution of calcite.
L213 The decrease of C/N ratio explained by mineralization is not obvious. A reference is needed here, as usually it is the reverse that is observed as mentioned the given reference Meyers and Lallier Vergès 1999. Moreover, is the decomposition the only processes, what about selected grazing or change in composition of the seston (phyto vs. zooplankton) to explain the C/N variation?
We agree with the reviewer that changes in C/N ratio above the RTZ are probably too small to undoubtedly indicate organic matter decomposition. We also agree, that some of the small changes seen here could have been caused by the occurrence of zooplankton. We will therefore restrict the interpretation of C/N ratios as a measure for organic matter decomposition to the values within or below the RTZ.
L260 text reads “This would explain the continuous increase in MeHg concentrations with depth…” What is the explanation? The absence of microniches does not explain the formation of MeHg at depth, where O2 saturation is still > 20%. Diffusion from sediments?
This sentence is indeed not clear and will be clarified. We believe that the increase in MeHg concentrations with depth in April is mainly caused by mass loss due to progressive organic matter decomposition (comparable to what has been described by Gallorini et al, 2022) although we cannot exclude MeHg formation by Mn reducing bacteria.
L294 Mass loss is the only explanation of the THg increase with depth. However, C concentration decrease by max a factor 3.9, whereas THg increase is a factor 26. Then the mass loss cannot account totally for the increase in THg concentration.
We are not sure if we understand this comment correctly. THg and C do not necessarily have to increase by the same exent because C (and other elements) is lost during mineralisation, but Hg is not. An additional explanation might be that some Hg released to the water phase during organic matter decomposition is scavenged by sinking seston as it has been observed in marine studies.
L314 text reads “The sulphide produced may form insoluble complexes with Hg (Shanks and Reeder, 1993; Bianchi et al., 2018), such as Hg sulphides (HgS), meaning that Hg becomes less available for methylation” It is not so clear that the presence of S decreases the bioavailability of Hg. Barrouilhet et al 2022 ESPR show that methylation potential increases with S concentration before to decrease at high S concentration.
Our data indicates that there is no sulfate reduction and thus formation of sulphide. The increase in S concentration in sesteon is thus rather due to mass loss during organic matter decomposition. This was an assumption which we could not proof in the frame of this study (only based on the increase of S concentration). The amount of material gained was too small to do Hg-thermo-desorption analyses or similar. The study of Barrouilhet et al 2022 is quite different from what we did and we could hardly say if their findings do apply here.
L318 text reads “Thus, THg and MeHg fluxes to the sediment are largely determined by changes in OM composition and mass loss during decomposition.” While these processes may change the MeHg fluxes, it is not clear why these processes change the flux of THg: i) if the authors are correct, the increase in THg concentration is due to mass loss in OM, then the quantity of THg remain the same, so the flux, and transformation of THg to MeHg will not change the flux of THg as MeHg is included in THg.
We agree, will be revised
L341 MeHg concentrations in the sediments are not sufficient to assess fluxes from sediments to interstitial water to overlying water.
We agree, statement will be toned down
L354 “Water column MeHg formation and degradation in eutrophic lakes appears to be intense and occurs rapidly and at rates similar to what we observed within the bottom sediments” This statement is not supported by the data/discussion. No rate has been determined neither in the seston nor in sediments.
We agree will be changed or removed
Fig 1. The scales do not cover the entire range of the results.
Will be adapted
Citation: https://doi.org/10.5194/bg-2022-170-AC1
-
RC2: 'Comment on bg-2022-170', Anonymous Referee #2, 03 Nov 2022
General comments.
This paper builds on several prior studies that show that the water column of lakes and oceans can be an important site for MeHg formation. It differs from most water column studies by focusing on a eutrophic urban lake and by specifically targeting MeHg abundance in bulk seston at different depths and dates for clues about formation and decay mechanisms. Unfortunately, the sampling technique lumped zooplankton in with seston, potentially introducing bias due to biomagnification. And the sampling scheme was spatially inconsistent, which makes the comparison of depth profiles on different dates difficult. The reason that the entire water column was sampled on one date and only the upper water column on most other dates is unexplained, and it compromises the authors’ conclusions about what’s going on as particles sink (especially in the hypolimnion since it was rarely sampled). Among other things (below), the authors need to justify their sampling methods and revisit the interpretation of changes in Hg speciation across depth and time. They also need to reconsider conclusions about links between climate change, productivity and bioaccumulation. This will require major revision.
Specific comments.
- The term “endogenic” should be reconsidered. It means “within the system”, which for lakes technically includes sediments. “Water column” would be better, unless they mean “within the seston” – in which case the title and text need to be re-worded
- Line 89 is an incomplete sentence
- Line 90: why a 25um net? It would allow many cyanophytes and chlorophytes to pass through, and bias collection toward zooplankton (which are not “seston”). Why not a clean pump-and-sieve/filter system instead?
- L220-225. The seston samples collected on those dates are not really much closer to the sediment surface. There’s just one hypo sample and it’s directly beneath the RTZ. You’d need to sample more depths to justify. Revise.
- L235. But peak concentrations of MeHg in seston occur in the suboxic RTZ on 4 of the 5 dates when the lake was strongly stratified. On the remaining date, seston MeHg concentrations are highest in the upper hypolimnion. During stratification, MeHg is never highest in the oxic epilimnion. If anything, these finding suggest that MeHg production is associated with microbial respiratory pathways that are less energy efficient than O2 reduction (e.g. sulfate reduction, Fe reduction). Revise.
- L240-245. Alternatively, low MeHg during high productivity may reflect biodilution in the larger phytoplankton biomass (i.e. parental seston). Lacking sound data, one can’t distinguish zooplankton bias from biodilution in microplankton, and neither necessarily point to sestonic microniches. Revise
- L255-263. They could also be explained by the presence of free-water microbes that possess the methylation gene pair hgcAB and occupy the O/A boundary. DOM rather than POM could be their carbon source. Revise.
- L275-284. Sestonic MeHg in the 20% range is not atypical for unpolluted temperate lakes. What’s unusual is the very low %MeHg in April
- L346. Actually, this was first shown in Little Rock Lake, which is only 10m deep (but the eutrophic part may be right).
- L346-end. Note that the range of Hg and MeHg in the seston of this eutrophic lake is on the low end of seston data reported for mesotrophic to oligotrophic North American lakes, both for MeHg concentration and %MeHg. High productivity is not necessarily conducive to abnormally high rates of MeHg accumulation in bioseston. In fact, most data suggest the opposite due to biodilution. It may be true that higher amounts of OM decomposition in eutrophic lakes does indeed exacerbate O2 depletion and enhance methylation in suboxic water, but that was not measured here. It seems that the most you can say with the data presented here is that the opposing forces of high biodilution and high decomposition need to be reconciled before addressing the impact of climate change. Revise
Citation: https://doi.org/10.5194/bg-2022-170-RC2 -
AC2: 'Reply on RC2', Laura Balzer, 18 Nov 2022
We would like to thank reviewer 2 for the constructive and helpful comments.
General comments.
This paper builds on several prior studies that show that the water column of lakes and oceans can be an important site for MeHg formation. It differs from most water column studies by focusing on a eutrophic urban lake and by specifically targeting MeHg abundance in bulk seston at different depths and dates for clues about formation and decay mechanisms. Unfortunately, the sampling technique lumped zooplankton in with seston, potentially introducing bias due to biomagnification. And the sampling scheme was spatially inconsistent, which makes the comparison of depth profiles on different dates difficult. The reason that the entire water column was sampled on one date and only the upper water column on most other dates is unexplained, and it compromises the authors’ conclusions about what’s going on as particles sink (especially in the hypolimnion since it was rarely sampled). Among other things (below), the authors need to justify their sampling methods and revisit the interpretation of changes in Hg speciation across depth and time. They also need to reconsider conclusions about links between climate change, productivity and bioaccumulation. This will require major revision.
We agree with reviewer that there might be potential bias in the interpretation of our MeHg depth profile. Our focus was on bulk seston and to our knowledge zooplankton is part of seston. We also believe, that the distribution of zooplankton alone cannot explain the THg and MeHg depth profiles in our lake. The separation of phyto- and zooplankton is useful in studies on biomagnification, which was not our topic. In this case, a qualitative separation of both fraction in small amounts is sufficient. In case of bulk seston sampling, eg. by means of a pump-and-sieve/filter system (0.45 µm?) during algae blooms as suggested by the reviewer such qualitative separation is nearly impossible (agglomeration) if larger sample volume is needed. Moreover, we believe that our interpretation of the distribution of MeHg and THg in the water phase is supported by our data on algae biomass, (Chl 1, pH) is sound. We regret that our sampling was imperfect, we made a lot of effort to gain in all cases sufficient material, however, this is a natural system with sometimes unpredictable changes of conditions. The reason why sampling is inconsistent through time and space is that we could not get sufficient material from the hypolimnion within the possible pumping time (~ 2 h per layer). We already mentioned this in the text but will explain this in more detail (see L254).
We will tone down on our conclusion regarding the link between climate change, productivity and bioaccumulation. We assume that the reviewer based this comments on his assumption that we mainly see biodilution. We have commented on this below.
Specific comments.
The term “endogenic” should be reconsidered. It means “within the system”, which for lakes technically includes sediments. “Water column” would be better, unless they mean “within the seston” – in which case the title and text need to be re-worded
This term has been introduced in other studies. For example, in Gallorini et al. 2021: We will define it within the manuscript as “production within the water column”.
Line 89 is an incomplete sentence
Will be changed
Line 90: why a 25µm net? It would allow many cyanophytes and chlorophytes to pass through, and bias collection toward zooplankton (which are not “seston”). Why not a clean pump-and-sieve/filter system instead?
We agree with the reviewer that some of the small fraction of phytoplankton might have got lost during water pumping through a 25 µm net. We tried a pump-and –sieve filter system before. However, this took to long to gain sufficient material from each water layer to do the solid phase analyses needed here at a resolution of 1 m within a single day (filter clogging etc.). Because of this, we decided to pump the water through a 25 µm net. Although it would have been the best option to sample all phytoplankton fraction, we believe that the lack of the fraction < 25 µm has no significant influence of the overall results and conclusions of this study (it just means more phytoplankton). To our knowledge zooplankton is part of the seston, too.
L220-225. The seston samples collected on those dates are not really much closer to the sediment surface. There’s just one hypo sample and it’s directly beneath the RTZ. You’d need to sample more depths to justify. Revise.
The reviewer is right, the deepest sample was taken just below the RTZ in those months. We will revise this statement. As explained above. We were not able to gain sufficient material from deeper layers in those month within the possible pumping time (2 h)
L235. But peak concentrations of MeHg in seston occur in the suboxic RTZ on 4 of the 5 dates when the lake was strongly stratified. On the remaining date, seston MeHg concentrations are highest in the upper hypolimnion. During stratification, MeHg is never highest in the oxic epilimnion.
We agree that this sentence is misleading and will clarify this section.
If anything, these finding suggest that MeHg production is associated with microbial respiratory pathways that are less energy efficient than O2 reduction (e.g. sulfate reduction, Fe reduction). Revise.
Not clear what the reviewer means here and what should be revised. We discussed in the ms that MeHg formation is releated to redox conditions in the water column specifically to Mn reduction (similar to what has been shown by Petersen et. al 2020(in a lake) and by Kohler et la., 2022 (in the arctic ocean) Fe reduction is of minor importance in this lake and sulphate reduction does not take place in the water column (compare Fig. S3). Data on eutrophic lakes are rare and to our knowledge not available yet at similar high resolution. A major aim of this study is to show changes in MeHg in seston at this comparatively high temporal and spatial (depth) resolution to understand the evolution of MeHg and THg concentrations and proportion during sinking through the water column.
L240-245. Alternatively, low MeHg during high productivity may reflect biodilution in the larger phytoplankton biomass (i.e. parental seston). Lacking sound data, one can’t distinguish zooplankton bias from biodilution in microplankton, and neither necessarily point to sestonic microniches. Revise
We agree with the reviewer that the variation in MeHg concentration might be a result of biodilution, zooplankton influence and/ or the formation of sestonic microniches. Based on ph, O2 and chlorophyll a concentation, we have classified the sampling days of May and November as periods of lower productivity and the sampling days in April and between June to September as periods of higher productivity (L. 181-190). Our data shows that the time of lower productivity is likley accompanied by higher amounts of zooplankton (higher N concentrations and lower C/N ratios in the upper two meters) and the time of higher productivity is likley accompanied by lower amounts of zooplankton (lower N concentraions and higher C/N ratios in the upper two meters) (L 238-246; Fig. 4). But we agree with the reviewer that we cannot fully rule out that the lower MeHg concentrations in times of higher productivity are caused to a certain extent by biodilution and vise versa the higher MeHg concentrations in times of lower productivity by bioconcentration. But we believe that biodilution/bioconcentration is not the major process in our lake as our data point to zooplankton influence between the sampling days. If biodilution/bioconcentration would be the the major process determining the MeHg concentration in our seston samples, we would also see a dilution or an increase in seston THg concentration, which we do not. The THg concentration do not change in the same way as the MeHg concentration. So there has to be another process than biodilution/bioconcentration controlling the MeHg concentration. However, as disscused in the ms the distribution of zooplankton -alone cannot explain the THg and MeHg depth profiles in our lake (L267-275). Obviously, we cannot completely exclude that zooplankton contributes to the maximum MeHg concentration within the RTZ. Our idea is that more organic matter rich seston (as a result of algae blooms) favours MeHg formation in sestonic unoxic microniches as oxic microbial pathways of MeHg formation are not yet known and MeHg get enriched during decay of the phytoplankton-dominated organic material. We will revise this section to make our point clearer.
L255-263. They could also be explained by the presence of free-water microbes that possess the methylation gene pair hgcAB and occupy the O/A boundary. DOM rather than POM could be their carbon source. Revise.
We discussed this point in L 266. It is likely that free-water microbial Hg methylation occurs, specifically because there is predominantly easy accessible DOM in the water column. But to our knowledge, oxic microbial pathways of MeHg formation are not yet known. Many papers point to methylation within anoxic microniches. However, our data suggest that free-water microbial Hg methylation is rather not the dominant process here as high MeHg concentration only occur during times of a pronounced RTZ (compare April when production is already high but MeHg is low because redox-zonation is not yet established) We will try to make this point clearer in the ms.
L275-284. Sestonic MeHg in the 20% range is not atypical for unpolluted temperate lakes. What’s unusual is the very low %MeHg in April
We can tone this statement down. However, we think that 20 % is a lot regarding the high biomass and that there is no influence from soil derived DOC/MeHg-rich inflow, which is typical for many oligotrophic lakes. We believe that the low MeHg proportions and concentrations in April (compared to the summer months) supports our conclusion that the MeHg is predominantly formed in the water phase along redox-gradients/micro-niches and the role of the RTZ. In April, algae biomass and productivity as indicated by high Chl a and high pH is already high, but the redox gradients are only weakly developed (only weak Mn-reduction) (Fig. S1), most likely because organic matter decomposition in the water column is still low. If most MeHg is originated from the sediment we should see this MeHg in the seston also in April.
We will try to make this point clearer in the ms.
L346. Actually, this was first shown in Little Rock Lake, which is only 10m deep (but the eutrophic part may be right).
We will refer to this point in the ms. However, we actually pointed out that our study is focused on eutrophic lakes, where data is rare.
L346-end. Note that the range of Hg and MeHg in the seston of this eutrophic lake is on the low end of seston data reported for mesotrophic to oligotrophic North American lakes, both for MeHg concentration and %MeHg. High productivity is not necessarily conducive to abnormally high rates of MeHg accumulation in bioseston. In fact, most data suggest the opposite due to biodilution. It may be true that higher amounts of OM decomposition in eutrophic lakes does indeed exacerbate O2 depletion and enhance methylation in suboxic water, but that was not measured here. It seems that the most you can say with the data presented here is that the opposing forces of high biodilution and high decomposition need to be reconciled before addressing the impact of climate change. Revise
We will revise this section, although it is not clear to us what exactly the reviewer will tell us here. We speculate that there would be much lower MeHg in this lake without high productivity regarding the total amount of MeHg formed in this lake. See answer above regarding biodilution. We want to show the distribution and changes in THg and MeHg concentration and proportion in seston with depth and how they are controlled by redox-conditions and decomposition under eutrophic conditions. How many studies on North American lakes show the distribution of THg and MeHg in seston at high temporal and spatial resolution including redox conditions, Chl a data etc.?.
Citation: https://doi.org/10.5194/bg-2022-170-AC2
Laura Balzer et al.
Laura Balzer et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 296 | 89 | 14 | 399 | 34 | 2 | 3 |
- HTML: 296
- PDF: 89
- XML: 14
- Total: 399
- Supplement: 34
- BibTeX: 2
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1