the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Jun 2022
Contrasts in dissolved, particulate and sedimentary organic carbon from the Kolyma River to the East Siberian Shelf
Lisa Bröder
Tommaso Tesi
Kirsi H. Keskitalo
Nikita Zimov
Anna Davydova
Philip Pika
Negar Haghipour
Timothy I. Eglinton
Jorien E. Vonk
Abstract. Arctic rivers will be increasingly affected by the hydrological and biogeochemical consequences of thawing permafrost. During transport, permafrost-derived organic carbon (OC) can either accumulate in floodplain and shelf sediments or be degraded into greenhouse gases prior to final burial. Thus, the net impact of permafrost OC on climate will ultimately depend on the interplay of complex processes that occur along the source-to-sink system. Here, we focused on the Kolyma River, the largest watershed completely underlain by continuous permafrost, and marine sediments of the East Siberian Sea as a transect to investigate the fate of permafrost OC along the land-ocean continuum. Three pools of riverine OC were investigated for the Kolyma main stem and five of its tributaries: dissolved OC (DOC), suspended particulate OC (POC), and riverbed sediment OC (SOC) and compared to earlier findings in marine sediments. Carbon isotopes (δ13C, Δ14C), lignin phenol, and lipid biomarkers show a contrasting composition and degradation state of these different carbon pools. Dual isotope source apportionment calculations imply that old permafrost-OC is mostly associated with sediments (SOC; contribution of 68 ± 10 %), and less dominant in POC (38 ± 8 %), while autochthonous primary production contributes around 44 ± 10 % to POC in the main stem and up to 79 ± 11 % in tributaries. Biomarker degradation indices suggest that Kolyma DOC is relatively degraded, regardless of its generally young age shown by previous studies. In contrast, SOC shows the lowest Δ14C signal (oldest OC), yet relatively fresh compositional signatures. Furthermore, decreasing mineral surface area-normalised OC- and biomarker loadings suggest that SOC is reactive along the land-ocean continuum supporting the idea that floodplain and shelf sediments are efficient reactors. A better understanding of DOC and POC dynamics in Arctic rivers is still necessary, however, this study highlights that sedimentary dynamics play a crucial role when targeting permafrost-derived OC in aquatic systems. Chemical and physical processes (e.g. degradation, sorption) along fluvial-marine transects will determine to what degree thawed permafrost OC may be destined for long-term burial, therewith attenuating further global warming.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2343 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2343 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Dirk J. Jong et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-516', Anonymous Referee #1, 22 Jul 2022
Jong et al. offer a look at the organic carbon (OC) pools across the rapidly changing land-ocean interface associated with the Kolyma River. They use stable and radio isotopes of OC as well as lignin, phenol and lipid biomarkers to demonstrate how the composition of the dissolved (DOC), suspended particulate (POC) and sedimentary (SOC) OC pools varied from Kolyma River tributaries, along the river itself and out to the East Siberian Sea. They find considerable variability in the age and sources of OC between these pools and across their transect.
Overall, I enjoyed reading the manuscript. It is clearly written and importantly, integrates terrestrial (freshwater) and marine measurements along a coastal margin experiencing rapid environmental change, something that is still not often done. I do believe that the manuscript will be of interest to the readers of Biogeosciences, particularly those interested in carbon dynamics at the coastal margin and/or in Arctic regions. The most substantive changes that I have suggested relate to the inclusion of more information in the Methods section to enable replicability of the methods, particularly as it relates to sample processing and analysis. Note also one editorial suggestion related to the description of isotope ratios that will involve a detailed look throughout the manuscript.
General considerations for the authors:
- Until the reader consults Vonk et al. (2012), Wild et al. (2019) and Bröder et al. (2020), how this study differs from those published previously by some of the author team is not entirely clear. It would be helpful for the reader if this was made more explicit in the manuscript itself. There is obviously great value in using previously published data to answer new questions, but what distinguishes this study could be more clear.
- There is a temporal offset between the collection of the riverine (2018) and marine (2008 & 2014) data. Do the authors think that this temporal offset could be important? Did anything important happen within the watershed during that time that might be reflected in the organic carbon pool? In 2021, for example (evidently outside of the sampling time period but likely not an isolated incident), widespread wildfires occurred within the Kolyma River watershed. Wildfires are just one example of events that are known to impact both permafrost thaw dynamics but also organic carbon pools.
Additional details in the method:
- Section 2.1: Could more information on the receiving ocean environment be provided? Is the Kolyma River at station K6, for example, tidally influenced? How do waters circulate within the East Siberian Sea? How was from the edge of the continental shelf from the sampling site furthest from land.
- Were replicates collected/run for any of the analyses?
- L178-179: How many subsamples were collected for each sample?
- L189: How were the filters subsampled for the radiocarbon analyses?
- L211: How was it determined to select one, two or three GF/F filters for the analysis?
- I’d be curious to know why two different acidification techniques (direct acidification and fumigation) were used to remove inorganic carbon from the SOC/POC samples for the stable isotope and radioisotope analyses, respectively? Do the authors have confidence that the two were equally effective in removing inorganic carbon?
- Were the samples for stable isotope analysis rinsed or neutralized following HCl addition?
- L238-240 and L247-251: How many samples contributed to the mean +/- SD used for the permafrost OC and primary production end-member values?
- L245-247: Why did the authors choose to include vegetation and soil OC as one end-member instead of two as in the source publication?
- Please specify whether means and standard deviations are indicated throughout the manuscript or if not, what metrics of average and variance are used.
- It is not clear until Table 2 that SOC was not collected at all sites.
- L260 – 263: How was convergence of the Bayesian model assessed?
Figures and figure captions:
- Figure 1 caption: Include the names of the tributaries and their abbreviations as a key in the caption.
- Figure 1c: It may be helpful for the reader to change the colours of marine sampling stations to distinguish between the two sampling campaigns. If possible, it would be great to see a bathymetry layer added to the figure, which would help in describing the receiving marine environment (as above).
- Figure 2 caption: Indicate whether the “average” refers to a mean or median.
- Figure 4 caption: What values are presented in the figure? Are these Bayesian median credible intervals? Means?
- Figure 5 caption: It does not appear as though any POC samples have been included in the figure, though POC (triangles) is included in the caption.
- Possible additional supplementary figure: It might be helpful to include a hydrograph of the Kolyma River as a supplementary figure and outline the time period over which the presented samples were collected. This would help to give hydrological context to the samples presented.
Results and Discussion:
- L277: Indicate the range of DOC concentrations observed in ESS surface waters from the literature for the reader to be able to make the comparison.
Editorial changes:
- L134: Change “or” to “of”
- L182: Remove “in” between “acidified” and “as described”.
- L246: Change “weighed” to “weighted”.
- Throughout the manuscript (example usages on L324, 334, 335, 336, 342, etc.): This is a matter of semantics, but δ13C and Δ14C are ratios and the ratio itself cannot inherently be more enriched or depleted. For δ13C, for example, the sample is more enriched or depleted in the heavier isotope 13C or in the lighter isotope 12 Alternatively, the ratios can be described as higher or lower, but not enriched or depleted without specifying to which of the two isotopes these modifiers refers. See the guide to Common Mistakes in Stable Isotope Terminology and Phraseology: http://dx.doi.org/10.6084/m9.figshare.1150337
Citation: https://doi.org/10.5194/egusphere-2022-516-RC1 -
AC1: 'Reply on RC1', Dirk Jong, 15 Sep 2022
We want to thank the reviewers for their thorough evaluation of the manuescript, and their helpful comments, questions and suggestions for improvement. Attached you can find the original 'referee comment' file, with author responses added in blue italics.
-
RC2: 'Comment on egusphere-2022-516', Anonymous Referee #2, 15 Aug 2022
The Jong et al. manuscript contained an enriched dataset of organic matter in different forms, including dissolved, suspended and sedimentary, from samples collected along the Kolyma River to the East Siberian Shelf. A comprehensive list of parameters was measured on these samples, including carbon stable and radio-isotopes, lignin phenols, lipid biomarkers, mineral specific surface area etc. They also used a mixing model to quantify the contribution of organic matter from three endmembers to these samples. The main conclusion was that DOC, POC and SOC along the transect have distinct compositional and degradation patterns, with significant contributions from permafrost-derived OC, particularly for SOC and DOC. It was also concluded that degradation occurred along the river to ocean transit based on biomarkers and OC loadings on minerals, among other minor conclusions.
Clearly this data is much more comprehensive than what has been published about the Kolyma River, or other Arctic rivers in general, as they included all three phases of organic carbon, and bulk and specific parameters. These data will be of value to the community, thus need to be published. The conclusions are solid, although I have to say that they are kind of expected and it is hard to find anything particular novel from what we already know.
It is great that DOC, POC and SOC were all measured in a same study, but the authors need to acknowledge the fact that SOC may be in totally different time scales in terms of mobilization and transport than DOC and POC. DOC and POC are co-transported with water flow, but SOC is likely not unless in a storm fasion. In other words, their resience times are way different. It is also not clear the depth of riverbed sediment was collected. This is important to know, as one could imagine surface 1cm could be very different from 10cm, in terms of not only the transport but also the level of dissolved oxygen which would affect degradation. The authors need to factor this in to the text.
Despite the comprehensives of this dataset, I still feel that there are a couple of key parameters missing, which would strengthen their arguments. For example, production was attributed to be the major contributor to the POC, but why not directly quantify the Chla concentration? This would direct address riverine production. 14C-DOC was not measured, either. They offered a couple of references, but I think this is a key parameter to have, particularly because its changes along the transect would offer further insights into the OC dynamics. The situation may not be as simple as cited, “earlier studies show that Kolyma River and tributary DOC is relatively young…”. Similarly, I am not sure why lignin phenols were not measured on POC?? This would directly address the contribution of terrestrial plants…
One of the motivations for conducting this work was the elusive nature of cycling and degradation of POD during the lateral transport through the whole watershed, as set up in the Introduction by the authors. However, when all the data are integrated, say from Figures 3-7, the degradation signals were most pronounced from the river mouth to East Siberian Sea, regardless of the end member contribution (Fig. 4), normalized biomarker centration (Fig. 6), or biomarker degradation (Fig. 7). In a sense, I think that these data collectively mean that the estuary section is more important than the river stream itself in terms of organic matter processing. Yet, this was not discussed but should be (even though you may not agree with me).
Line 60: delete the “.” before “degradation”
Line 68: should be “Hilton et al. (21015)”
Lines 121-130: it is a bit awkward to have a table and figure in the introduction. I would suggest that this be moved to the next section.
Line 153: how deep did the sampler penetrate? This may be important information (see my comment above).
Line 174: change to “according to Deirmendjian et al. (2020).”
Line 252: it’s not clear what you meant by “…our own algal sample”. How do you know it was algal bloom? And there would be other types of organic matter in a riverine sample!
Line 499: it could be simply due to the conversion of aldehyde to acid during oxidation, not necessarily selective degradation.
Citation: https://doi.org/10.5194/egusphere-2022-516-RC2 -
AC2: 'Reply on RC2', Dirk Jong, 15 Sep 2022
We want to thank the reviewers for their thorough evaluation of the manuscript, and their helpful comments, questions and suggestions for improvement. Attached you can find the original 'referee comment' file, with author responses added in blue italics.
-
AC2: 'Reply on RC2', Dirk Jong, 15 Sep 2022
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-516', Anonymous Referee #1, 22 Jul 2022
Jong et al. offer a look at the organic carbon (OC) pools across the rapidly changing land-ocean interface associated with the Kolyma River. They use stable and radio isotopes of OC as well as lignin, phenol and lipid biomarkers to demonstrate how the composition of the dissolved (DOC), suspended particulate (POC) and sedimentary (SOC) OC pools varied from Kolyma River tributaries, along the river itself and out to the East Siberian Sea. They find considerable variability in the age and sources of OC between these pools and across their transect.
Overall, I enjoyed reading the manuscript. It is clearly written and importantly, integrates terrestrial (freshwater) and marine measurements along a coastal margin experiencing rapid environmental change, something that is still not often done. I do believe that the manuscript will be of interest to the readers of Biogeosciences, particularly those interested in carbon dynamics at the coastal margin and/or in Arctic regions. The most substantive changes that I have suggested relate to the inclusion of more information in the Methods section to enable replicability of the methods, particularly as it relates to sample processing and analysis. Note also one editorial suggestion related to the description of isotope ratios that will involve a detailed look throughout the manuscript.
General considerations for the authors:
- Until the reader consults Vonk et al. (2012), Wild et al. (2019) and Bröder et al. (2020), how this study differs from those published previously by some of the author team is not entirely clear. It would be helpful for the reader if this was made more explicit in the manuscript itself. There is obviously great value in using previously published data to answer new questions, but what distinguishes this study could be more clear.
- There is a temporal offset between the collection of the riverine (2018) and marine (2008 & 2014) data. Do the authors think that this temporal offset could be important? Did anything important happen within the watershed during that time that might be reflected in the organic carbon pool? In 2021, for example (evidently outside of the sampling time period but likely not an isolated incident), widespread wildfires occurred within the Kolyma River watershed. Wildfires are just one example of events that are known to impact both permafrost thaw dynamics but also organic carbon pools.
Additional details in the method:
- Section 2.1: Could more information on the receiving ocean environment be provided? Is the Kolyma River at station K6, for example, tidally influenced? How do waters circulate within the East Siberian Sea? How was from the edge of the continental shelf from the sampling site furthest from land.
- Were replicates collected/run for any of the analyses?
- L178-179: How many subsamples were collected for each sample?
- L189: How were the filters subsampled for the radiocarbon analyses?
- L211: How was it determined to select one, two or three GF/F filters for the analysis?
- I’d be curious to know why two different acidification techniques (direct acidification and fumigation) were used to remove inorganic carbon from the SOC/POC samples for the stable isotope and radioisotope analyses, respectively? Do the authors have confidence that the two were equally effective in removing inorganic carbon?
- Were the samples for stable isotope analysis rinsed or neutralized following HCl addition?
- L238-240 and L247-251: How many samples contributed to the mean +/- SD used for the permafrost OC and primary production end-member values?
- L245-247: Why did the authors choose to include vegetation and soil OC as one end-member instead of two as in the source publication?
- Please specify whether means and standard deviations are indicated throughout the manuscript or if not, what metrics of average and variance are used.
- It is not clear until Table 2 that SOC was not collected at all sites.
- L260 – 263: How was convergence of the Bayesian model assessed?
Figures and figure captions:
- Figure 1 caption: Include the names of the tributaries and their abbreviations as a key in the caption.
- Figure 1c: It may be helpful for the reader to change the colours of marine sampling stations to distinguish between the two sampling campaigns. If possible, it would be great to see a bathymetry layer added to the figure, which would help in describing the receiving marine environment (as above).
- Figure 2 caption: Indicate whether the “average” refers to a mean or median.
- Figure 4 caption: What values are presented in the figure? Are these Bayesian median credible intervals? Means?
- Figure 5 caption: It does not appear as though any POC samples have been included in the figure, though POC (triangles) is included in the caption.
- Possible additional supplementary figure: It might be helpful to include a hydrograph of the Kolyma River as a supplementary figure and outline the time period over which the presented samples were collected. This would help to give hydrological context to the samples presented.
Results and Discussion:
- L277: Indicate the range of DOC concentrations observed in ESS surface waters from the literature for the reader to be able to make the comparison.
Editorial changes:
- L134: Change “or” to “of”
- L182: Remove “in” between “acidified” and “as described”.
- L246: Change “weighed” to “weighted”.
- Throughout the manuscript (example usages on L324, 334, 335, 336, 342, etc.): This is a matter of semantics, but δ13C and Δ14C are ratios and the ratio itself cannot inherently be more enriched or depleted. For δ13C, for example, the sample is more enriched or depleted in the heavier isotope 13C or in the lighter isotope 12 Alternatively, the ratios can be described as higher or lower, but not enriched or depleted without specifying to which of the two isotopes these modifiers refers. See the guide to Common Mistakes in Stable Isotope Terminology and Phraseology: http://dx.doi.org/10.6084/m9.figshare.1150337
Citation: https://doi.org/10.5194/egusphere-2022-516-RC1 -
AC1: 'Reply on RC1', Dirk Jong, 15 Sep 2022
We want to thank the reviewers for their thorough evaluation of the manuescript, and their helpful comments, questions and suggestions for improvement. Attached you can find the original 'referee comment' file, with author responses added in blue italics.
-
RC2: 'Comment on egusphere-2022-516', Anonymous Referee #2, 15 Aug 2022
The Jong et al. manuscript contained an enriched dataset of organic matter in different forms, including dissolved, suspended and sedimentary, from samples collected along the Kolyma River to the East Siberian Shelf. A comprehensive list of parameters was measured on these samples, including carbon stable and radio-isotopes, lignin phenols, lipid biomarkers, mineral specific surface area etc. They also used a mixing model to quantify the contribution of organic matter from three endmembers to these samples. The main conclusion was that DOC, POC and SOC along the transect have distinct compositional and degradation patterns, with significant contributions from permafrost-derived OC, particularly for SOC and DOC. It was also concluded that degradation occurred along the river to ocean transit based on biomarkers and OC loadings on minerals, among other minor conclusions.
Clearly this data is much more comprehensive than what has been published about the Kolyma River, or other Arctic rivers in general, as they included all three phases of organic carbon, and bulk and specific parameters. These data will be of value to the community, thus need to be published. The conclusions are solid, although I have to say that they are kind of expected and it is hard to find anything particular novel from what we already know.
It is great that DOC, POC and SOC were all measured in a same study, but the authors need to acknowledge the fact that SOC may be in totally different time scales in terms of mobilization and transport than DOC and POC. DOC and POC are co-transported with water flow, but SOC is likely not unless in a storm fasion. In other words, their resience times are way different. It is also not clear the depth of riverbed sediment was collected. This is important to know, as one could imagine surface 1cm could be very different from 10cm, in terms of not only the transport but also the level of dissolved oxygen which would affect degradation. The authors need to factor this in to the text.
Despite the comprehensives of this dataset, I still feel that there are a couple of key parameters missing, which would strengthen their arguments. For example, production was attributed to be the major contributor to the POC, but why not directly quantify the Chla concentration? This would direct address riverine production. 14C-DOC was not measured, either. They offered a couple of references, but I think this is a key parameter to have, particularly because its changes along the transect would offer further insights into the OC dynamics. The situation may not be as simple as cited, “earlier studies show that Kolyma River and tributary DOC is relatively young…”. Similarly, I am not sure why lignin phenols were not measured on POC?? This would directly address the contribution of terrestrial plants…
One of the motivations for conducting this work was the elusive nature of cycling and degradation of POD during the lateral transport through the whole watershed, as set up in the Introduction by the authors. However, when all the data are integrated, say from Figures 3-7, the degradation signals were most pronounced from the river mouth to East Siberian Sea, regardless of the end member contribution (Fig. 4), normalized biomarker centration (Fig. 6), or biomarker degradation (Fig. 7). In a sense, I think that these data collectively mean that the estuary section is more important than the river stream itself in terms of organic matter processing. Yet, this was not discussed but should be (even though you may not agree with me).
Line 60: delete the “.” before “degradation”
Line 68: should be “Hilton et al. (21015)”
Lines 121-130: it is a bit awkward to have a table and figure in the introduction. I would suggest that this be moved to the next section.
Line 153: how deep did the sampler penetrate? This may be important information (see my comment above).
Line 174: change to “according to Deirmendjian et al. (2020).”
Line 252: it’s not clear what you meant by “…our own algal sample”. How do you know it was algal bloom? And there would be other types of organic matter in a riverine sample!
Line 499: it could be simply due to the conversion of aldehyde to acid during oxidation, not necessarily selective degradation.
Citation: https://doi.org/10.5194/egusphere-2022-516-RC2 -
AC2: 'Reply on RC2', Dirk Jong, 15 Sep 2022
We want to thank the reviewers for their thorough evaluation of the manuscript, and their helpful comments, questions and suggestions for improvement. Attached you can find the original 'referee comment' file, with author responses added in blue italics.
-
AC2: 'Reply on RC2', Dirk Jong, 15 Sep 2022
Peer review completion
Journal article(s) based on this preprint
Dirk J. Jong et al.
Dirk J. Jong et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 323 | 145 | 15 | 483 | 5 | 6 |
- HTML: 323
- PDF: 145
- XML: 15
- Total: 483
- BibTeX: 5
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2343 KB) - Metadata XML