The marine reservoir age of Greenland coastal waters

Pearce, Christof; Özdemir, Karen Søby; Forchhammer, Ronja Cedergreen; Detlef, Henrieka; Olsen, Jesper

doi:https://doi.org/10.5194/gchron-2023-7

Preprints

https://doi.org/10.5194/gchron-2023-7

Preprints

17 Apr 2023

| 17 Apr 2023

Status: a revised version of this preprint was accepted for the journal GChron and is expected to appear here in due course.

The marine reservoir age of Greenland coastal waters

Christof Pearce, Karen Søby Özdemir, Ronja Cedergreen Forchhammer, Henrieka Detlef, and Jesper Olsen

Abstract. Knowledge of the marine reservoir age is fundamental for creating reliable chronologies of marine sediment archives based on radiocarbon dating. This age difference between the ¹⁴C age of a marine sample and that of its contemporaneous atmosphere is dependent on several factors, among others ocean circulation, water mass distribution, terrestrial runoff, upwelling, sea-ice cover and is therefore spatially heterogenous. Anthropogenic influence on the global isotopic carbon system, mostly through atmospheric nuclear tests, has complicated the determination of the regional reservoir age correction ΔR, which therefore can only be measured on historic samples of known age. In this study we expand on the few existing measurements of ΔR for the coastal waters around Greenland, by adding 92 new radiocarbon dates on mollusks from museum collections. All studied mollusk samples were collected during historic expeditions of the late 19^th and early 20^th centuries and besides coastal sites around Greenland, the dataset also includes localities from the western Labrador Sea, Baffin Bay, and the Iceland Sea. Together with existing measurements, the new results are used to calculate average ΔR values for different regions around Greenland, all in relation to Marine20, the most recent radiocarbon calibration curve. To support further discussions and comparison with previous datasets, we introduce the term ΔR₁₃ where the suffix 13 refers to the previous calibration curve Marine13. Our study explores the links between the marine reservoir age and oceanography, sea ice cover, water depth, mollusk feeding habits, and the presence of carbonate bedrock. Although we provide regional averages, we encourage people to consult the full catalogue of measurements and determine a suitable ΔR for each case individually, based on the exact location including water depth. Despite this significant expansion of the regional reservoir age database around Greenland, data from the northern coast, directly bordering the Arctic Ocean remains missing.

Received: 04 Apr 2023 – Discussion started: 17 Apr 2023

Download & links

Preprint (PDF, 2148 KB)

Supplement (6224 KB)

Download & links

Christof Pearce et al.

Status: closed

RC1:
'Comment on gchron-2023-7', Paula Reimer, 13 May 2023

Marine radiocarbon calibration requires an estimate of the reservoir offset from the marine calibration curve (ΔR). These estimates can be based on ¹⁴C measurements of pre-nuclear weapons testing, known age shells, independently dated coral, or contemporaneous marine and terrestrial samples. Until now the ΔR values for coastal Greenland have been sparse. The authors have significantly enlarged the dataset of known age shell measurements from coastal Greenland and neighboring regions of the Arctic. They have carefully selected samples from museum specimens to ensure the mollusks were collected alive. The effects of sea ice cover, water depth and mollusk feeding habits were investigated and discussed. Regional averages were calculated for zones based on “prevailing currents and water masses” although most of the zones have overlapping values. The authors also compared ΔR values for a limited number of samples stored in ethanol to dry samples.
Specific comments/questions:
Wet vs dry sample comparison: This comparison is based on only 6 dry samples and 4 wet samples from one region (Suppl. Fig 2). This is a rather small dataset to reach the conclusion that dry samples are not reliably collected alive. It is difficult to tell which dry samples were used in the comparison but, of the 5 dry samples from Kaiser Frans Joseph Fjord, 4 were species with unknown feeding habits or deposit feeders. It is well known, and also shown in this manuscript, that deposit feeders may incorporate older carbon from their environment. This comparison apparently forms the basis for one of the stated criteria for sample selection (line 412-413): ‘Museum sample storage: As the exact age of samples from “dry” collections is possibly unknown, only samples with soft tissue present, stored in “wet” collections, should be used for ΔR evaluation’. Samples stored in ethanol may be ideal to ensure live collection but this criterion would exclude many of the existing values in the literature. In some cases, the museum documentation is unambiguous about live collection but there are also other indications of whether “dry” bivalves in collections were most likely collected live or shortly after death. These include fragile mollusks that would have been abraded if transported to a beach as well as those with residual ligament, muscle or periostracum (O’Connor et al. 2010). In addition, some species have colours that are light sensitive so would be bleached if not collected alive and stored in the dark (Angulo et al. 2007).
The study also makes use samples from water depths that would not be considered surface ocean in general. The low ΔR values for these samples provide a very interesting and useful observation for these locations which are ‘characterized by convection and formation of North Atlantic Deep Water and Labrador Sea Water’. The authors advise that: ‘When calibrating benthic dates from deeper sites one could therefore consider excluding extreme values obtained from surface ocean samples when making the choice of which reservoir correction to apply’. This seems like valid advice for these regions however it should be noted that surface ocean ΔR values are not generally applicable for benthic dates in other regions where deep water can be very depleted in ¹⁴C. Ideally one would have ΔR values from deep water samples to use for radiocarbon calibration of benthic samples but these are scarce in the literature. Also is there an explanation for the low ΔR values for relatively deep samples in NW Greenland zone 5? Is the West Greenland Current fed by Labrador Sea water?
Technical comments:
Line 18: ” Marine20, the most recent radiocarbon calibration curve” Insert “marine” ahead of radiocarbon.
Line 19 and 74: ‘we introduce the term ΔR₁₃”. This term has been previously introduced in Heaton et al. 2023. I would suggest replacing ‘introduce’ with ‘use’
Line 51: ‘to a lesser extent, injection of 14C-depleted CO2 from the burning of fossil fuels’ Although this is a common perception and definitely true for reservoir ages relative to the atmosphere, for ΔR this is insignificant. ΔR is the difference between the marine radiocarbon age and the marine calibration curve which is modelled with input from the atmosphere so includes the Suess effect.
Line 54: ‘tephrochronology (Pearce et al., 2017; Austin et al., 1995; Olsen et al., 2014), or paired marine/terrestrial dating’ ΔR values may also be determined by U-Th dated coral (e.g. Hua et al. 2015).
Line 57: ‘Several hundred different studies were made to study the local reservoir age’. Replace ‘were’ with ‘have been’.
Line 127: ‘the most commonly used value for the reservoir age correction (prior to publication of Marine20), ΔR = 0 14C years’ Since ΔR without a subscribe is defined earlier as relative to Marine20 ,it would be better if this written here as 'Rxx = 0 14C were xx =04, 09 or 13.
Line 184: ‘Wet samples were placed in a drying oven at 40 °C for several days’ It would be worth stating that this is to remove any ethanol from the shell since contamination from the ethanol might be a concern.
Line 189: ‘milliQ water’ Trademark symbol needed
Line 239: ‘where ΔRi and σi are the mean value and uncertainty of calculated local reservoir age offset’. Add ‘of sample i’ to clarify.
Line 243: ‘Where the subscript w indicates that the uncertainty is calculate using the error each ΔRi’ Change ‘error’ to ‘uncertainty’ and ‘is calculate’ to ‘is calculated’
Line 368: ‘no ΔR values higher than 50 years are found, and where ΔR values exceed 160 years,’ ΔR values should be given as ‘14C yrs’ rather than ‘years’
Line 370: ‘there is also a positive correlation between sea-ice cover and reservoir age’. Are the correlations significant?
Line 429: ‘these values remain only valid for the modern situation’ Insert ‘pre-bomb’ before modern because the values would not be valid for post-bomb samples.
Fig. 1 caption: Need to define WGC, NFL, EGC.
Also. ‘Areas of deep convection in the Labrador Sea and north of Iceland are colored yellow’. These look light green on top of the blue background - perhaps 'shaded light green' would be better
Fig 2. Given the results, is there justification for separate zones for the Greenland coastal waters since ΔR values overlap?
Suppl. Fig 2. Sample numbers on Suppl. Fig 2 would be helpful for comparison of species and feeding habits
References:
Angulo, R. J., Reimer, P. J., De Souza, M. C., Scheel-Ybert, R., Tenório, M. C., Disaró, S. T. & Gaspar, M. D. 2007. A tentative determination of upwelling influence on the paleo-surficial marine water reservoir effect in southeastern Brazil. Radiocarbon, 49, 1-5.
Heaton, T. J., Bard, E., Bronk Ramsey, C., Butzin, M., Hatté, C., Hughen, K. A., Köhler, P. & Reimer, P. J. 2023. A response to community questions on the MARINE20 radiocarbon age calibration curve: marine reservoir ages and the calibration of 14c samples from the oceans. Radiocarbon, 65, 247-273.
Hua, Q., Webb, G. E., Zhao, J.-X., Nothdurft, L. D., Lybolt, M., Price, G. J. & Opdyke, B. N. 2015. Large variations in the Holocene marine radiocarbon reservoir effect reflect ocean circulation and climatic changes. Earth and Planetary Science Letters, 422, 33-44.
O'Connor, S., Ulm, S., Fallon, S. J., Barham, A. & Loch, I. 2010. Pre-bomb marine reservoir variability in the Kimberley region, Western Australia. Radiocarbon, 52, 1158-1165.

Citation: https://doi.org/10.5194/gchron-2023-7-RC1
- AC1: 'Reply on RC1', Christof Pearce, 22 Aug 2023
  
  We thank Paula Reimer for her valuable comments and suggestions. Please, find our detailed response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC1
RC2:
'Comment on gchron-2023-7', Anonymous Referee #2, 14 Jun 2023

Pearce et al. present about 100 new marine radiocarbon (¹⁴C) reservoir ages (MRA) of coastal and shelf waters around Greenland, Baffin Island, Newfoundland, and Iceland. The data result from ¹⁴C measurements on pre-bomb molluscs retrieved from museums. The MRA results are binned to seven regions and discussed with respect to the global Marine20 ¹⁴C calibration curve in terms of the regional MRA correction, ∆R₂₀. The authors also discuss their ∆R₂₀ results in the light of specific factors such as sample depth, sea ice cover and feeding habits.
The manuscript is well written, the presentation is clear, and the dataset is an important contribution to the MRA / ∆R data base. However, there are a few minor issues that should be addressed before publication in GChron (L = line):
L 30: The half-life of ¹⁴C has been slightly revised to 5700 years (e.g., Audi et al., 2003; Bé and Chechev, 2012; Kutschera, 2013)
L 124 "marine mammals": "marine" should be removed
Figure 1:

(i) Add a depth scale (such as in Fig. 1 by Pieńkowski et al. 2022)

(ii) "NFL", "WGC", and "EGC" should be also explained in the caption.
L 214-216 (and Figure 2): Is there a hard objective criterion to separate the three southernmost data points in East Greenland from region 7?
Figure 2: Explain "CS"
L 353: Explain "mwd"
Figure 3: Would it make sense to indicate the positions of the outliers in the inserted map?
L 376: Explain "mwd"
Figure 4:

(i) As ∆R depends on the sea ice concentration, the coordinate axes should be swapped. The situation is different from Figure 3 where the independent variable (usually plotted along the horizontal axis) is depth (typically plotted in vertical direction).

(ii) Can you quantify the trends, and are they significant? I wonder if the trends are still visible once the coordinate axes have been swapped.

(iii) Would it make sense to indicate the position of the outlier in the inserted map?

References:
Audi, G., Bersillon, O., Blachot, J., and Wapstra, A. H.: The Nubase evaluation of nuclear and decay properties, Nuclear Physics A, 729, 3–128, https://doi.org/10.1016/j.nuclphysa.2003.11.001, 2003.
Bé, M.-M. and Chechev, V. P.: 14C - Comments on evaluation of decay data, Laboratoire National Henri Becquerel, Gif-sur-Yvette, http://www.lnhb.fr/nuclides/C-14_com.pdf, 2012.
Kutschera, W.: Applications of accelerator mass spectrometry, International Journal of Mass Spectrometry, 349–350, 203–218, https://doi.org/10.1016/j.ijms.2013.05.023, 2013.

Citation: https://doi.org/10.5194/gchron-2023-7-RC2
- AC2: 'Reply on RC2', Christof Pearce, 22 Aug 2023
  
  We thank anonymous reviewer R2 for their valuable comments and suggestions. Please, find our detailed response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC2
CC1:
'Comment on gchron-2023-7', Elisabeth Michel, 23 Jun 2023

The authors present new ¹⁴C reservoir ages for surface and deep waters of the North Atlantic and Nordic seas : Labrador sea, Baffin Bay and Iceland Sea, from shell museum collections. The shells have been collected from 1865 to 1931. They present a nice review of existing reservoir ages.
First, they compare the results from shells that were preserved in ethanol in museum collections and those who were dry samples. They found that the mean dry samples ¹⁴C reservoir age is much higher than the mean of ethanol preserved samples and argue that the dry samples might be dead since a long time when they were collected.
The authors propose regional ¹⁴C reservoir ages within 7 different geographic zones, considering both their new results and 14C reservoir ages from the Marine Reservoir Age Database (Reimer and Reimer 2001) considering only samples preserved in ethanol.
For the relevance of the results, the authors also consider the results of deposit feeders compared to suspension feeder.
For the interpretation of the regional ¹⁴C reservoir age they consider the depth of collection of the different samples and shortly discuss the impact of ocean circulation and sea ice.

This paper is mainly a data paper, the discussion of the result is rather short and do not discuss in depth the different factors that could impact their regional ¹⁴C reservoir age.
Following are some detailed comments and also some ideas for a more complete discussion concerning the regional results.

Considering dry samples, I wonder if there is any evidence on the shell, muscle marks or the like, to tell whether the specimen was collected alive or could have been dead for a long time.
For the deposit and suspension feeders, the authors should compare the results zone by zone as they indicated that the ∆R was very different from one zone to another. They could also check the dispersion for species for which the feeding habit is unknown. It would be better to discuss first the aspect linked to the mollusk : dry and ethanol preserved samples, feeding habitat and after all the physical parameters: sea ice, depth and circulation.
One question that is not addressed, do the author have an idea of the mean lifetime of the different mollusk?
It seems that the authors choose to include only ¹⁴C ∆R measured on molluks. I wonder why they do not compare their results with 14C measurements made directly on DIC of sea water in the early fifties like for example Fonselius and Östlund, 1959 Tellus.

What is the most impressive is the dispersion of the ¹⁴∆R data within some of the geographic zones. The authors discuss the impact of sea-ice checking if a relationship exist between the annual average sea ice concentration of a sample location and its ¹⁴C reservoir age (fig. 4). The regressions and their statistics for the different geographic zones are necessary if the authors want to demonstrate that the regional relationships are significant. Furthermore during formation of sea ice the carbon sink in the ocean might be effective thus the impact of non-perennial sea-ice is not obvious.
The authors argue that Heaton et al., 2020, explain that the Marine20 does not apply to the polar regions because of sea ice. Heaton et al, 2020 is as much about ocean circulation as it is about sea ice.
The role of Ocean Circulation could be considered considering fluxes along the different straits. Furthermore the influence of Atlantic and Artic water masses might changes with time, for example linked to North Atlantic Oscillation. Thus a time evolution of ¹⁴C ∆R within the geographical zones could be also discussed and might explain partly the large dispersion of the results?
∆R could be also influence by continental waters with old ¹⁴C DIC coming from under the ice like in the Ross Sea (Mikucki et al., 2009). This point is not discussed.

Figures: even if the projections does not make it easy and they will not be regularly spaced, it would be nice to have some latitudinal and longitudinal tics on the borders of figures 1, 2 and suppl. Fig.1.

Citation: https://doi.org/10.5194/gchron-2023-7-CC1
- AC3: 'Reply on CC1', Christof Pearce, 22 Aug 2023
  
  We appreciate the valuable comments and suggestions of Elisabeth Michel to our manuscript. Please, find our detailed response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC3
RC3:
'Comment on gchron-2023-7', Matt O'Regan, 29 Jun 2023

This is a very nicely written paper presenting 92 new radiocarbon dates on pre-bomb mollusks collected from around Greenland (with the exception of its northern Arctic Ocean margin). In addition to the utility of these new dates for constraining regional reservoir corrections, I think the manuscript is timely in presenting a nice practical discussion (and examples) on the need to update reservoir corrections when using the new Marine20 calibration curve.
The comparisons of dR with water depth and sea ice coverage are interesting in highlighting patterns, although somewhat inconclusive in identifying a cause/explanation for the variability. I do not think this limits the scientific contribution made by the paper, and certainly sets the stage for future work needed to understand this variability. I believe this would require a considerable amount of work, and could potentially start with moving away from water depth and looking at the variability in Temperature-Salinity space to see if ages cluster in specific water masses. However, I do not think this is a necessary addition to this work, which very well suited for publication in Geochronology in its current form.
I do feel one aspect the paper is missing is a discussion on the limited, but rather informative data on Holocene dR values from the central Arctic Ocean. Specifically the inferred differences between the age of Pacific and Atlantic waters that are found in the interior Arctic, and should be impacting the age of surface waters(?) in northern Baffin Bay. For example, West et al (2022), Geochemistry, Geophysics, Geosystems (doi: 10.1029/2021GC010187) used tephra from the Aniakchak eruption circa 3.6 ka in two cores from the Chukchi Sea - one at 50 m depth (Pacific water) and one at 120 m depth (likely Atlantic water) - to show that the dR (using Marine 20) for benthic foraminifera and mollusks at these sites was about 330 years for Pacific waters and 205 years for Atlantic waters. These seem to be somewhat consistent with the larger dR values in sections 3 and 4 from Northern Baffin Bay. It would be nice to see some discussion about the influence of Arctic outflow and the water masses involved on the dR values in Northern Baffin Bay. Currently these are described simply as 'outflow' from the Arctic, which could easily be expanded to detail the role and age of Pacific and Atlantic waters in this outflow.
Overall, I feel this is a great contribution that will provide significant support to future paleoceanographic work around Greenland.

Citation: https://doi.org/10.5194/gchron-2023-7-RC3
- AC4: 'Reply on RC3', Christof Pearce, 22 Aug 2023
  
  We thank Matt O'Regan for his encouraging comments and valuable suggestions to improve our manuscript. Please, find our short response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC4

Status: closed

RC1:
'Comment on gchron-2023-7', Paula Reimer, 13 May 2023

Marine radiocarbon calibration requires an estimate of the reservoir offset from the marine calibration curve (ΔR). These estimates can be based on ¹⁴C measurements of pre-nuclear weapons testing, known age shells, independently dated coral, or contemporaneous marine and terrestrial samples. Until now the ΔR values for coastal Greenland have been sparse. The authors have significantly enlarged the dataset of known age shell measurements from coastal Greenland and neighboring regions of the Arctic. They have carefully selected samples from museum specimens to ensure the mollusks were collected alive. The effects of sea ice cover, water depth and mollusk feeding habits were investigated and discussed. Regional averages were calculated for zones based on “prevailing currents and water masses” although most of the zones have overlapping values. The authors also compared ΔR values for a limited number of samples stored in ethanol to dry samples.
Specific comments/questions:
Wet vs dry sample comparison: This comparison is based on only 6 dry samples and 4 wet samples from one region (Suppl. Fig 2). This is a rather small dataset to reach the conclusion that dry samples are not reliably collected alive. It is difficult to tell which dry samples were used in the comparison but, of the 5 dry samples from Kaiser Frans Joseph Fjord, 4 were species with unknown feeding habits or deposit feeders. It is well known, and also shown in this manuscript, that deposit feeders may incorporate older carbon from their environment. This comparison apparently forms the basis for one of the stated criteria for sample selection (line 412-413): ‘Museum sample storage: As the exact age of samples from “dry” collections is possibly unknown, only samples with soft tissue present, stored in “wet” collections, should be used for ΔR evaluation’. Samples stored in ethanol may be ideal to ensure live collection but this criterion would exclude many of the existing values in the literature. In some cases, the museum documentation is unambiguous about live collection but there are also other indications of whether “dry” bivalves in collections were most likely collected live or shortly after death. These include fragile mollusks that would have been abraded if transported to a beach as well as those with residual ligament, muscle or periostracum (O’Connor et al. 2010). In addition, some species have colours that are light sensitive so would be bleached if not collected alive and stored in the dark (Angulo et al. 2007).
The study also makes use samples from water depths that would not be considered surface ocean in general. The low ΔR values for these samples provide a very interesting and useful observation for these locations which are ‘characterized by convection and formation of North Atlantic Deep Water and Labrador Sea Water’. The authors advise that: ‘When calibrating benthic dates from deeper sites one could therefore consider excluding extreme values obtained from surface ocean samples when making the choice of which reservoir correction to apply’. This seems like valid advice for these regions however it should be noted that surface ocean ΔR values are not generally applicable for benthic dates in other regions where deep water can be very depleted in ¹⁴C. Ideally one would have ΔR values from deep water samples to use for radiocarbon calibration of benthic samples but these are scarce in the literature. Also is there an explanation for the low ΔR values for relatively deep samples in NW Greenland zone 5? Is the West Greenland Current fed by Labrador Sea water?
Technical comments:
Line 18: ” Marine20, the most recent radiocarbon calibration curve” Insert “marine” ahead of radiocarbon.
Line 19 and 74: ‘we introduce the term ΔR₁₃”. This term has been previously introduced in Heaton et al. 2023. I would suggest replacing ‘introduce’ with ‘use’
Line 51: ‘to a lesser extent, injection of 14C-depleted CO2 from the burning of fossil fuels’ Although this is a common perception and definitely true for reservoir ages relative to the atmosphere, for ΔR this is insignificant. ΔR is the difference between the marine radiocarbon age and the marine calibration curve which is modelled with input from the atmosphere so includes the Suess effect.
Line 54: ‘tephrochronology (Pearce et al., 2017; Austin et al., 1995; Olsen et al., 2014), or paired marine/terrestrial dating’ ΔR values may also be determined by U-Th dated coral (e.g. Hua et al. 2015).
Line 57: ‘Several hundred different studies were made to study the local reservoir age’. Replace ‘were’ with ‘have been’.
Line 127: ‘the most commonly used value for the reservoir age correction (prior to publication of Marine20), ΔR = 0 14C years’ Since ΔR without a subscribe is defined earlier as relative to Marine20 ,it would be better if this written here as 'Rxx = 0 14C were xx =04, 09 or 13.
Line 184: ‘Wet samples were placed in a drying oven at 40 °C for several days’ It would be worth stating that this is to remove any ethanol from the shell since contamination from the ethanol might be a concern.
Line 189: ‘milliQ water’ Trademark symbol needed
Line 239: ‘where ΔRi and σi are the mean value and uncertainty of calculated local reservoir age offset’. Add ‘of sample i’ to clarify.
Line 243: ‘Where the subscript w indicates that the uncertainty is calculate using the error each ΔRi’ Change ‘error’ to ‘uncertainty’ and ‘is calculate’ to ‘is calculated’
Line 368: ‘no ΔR values higher than 50 years are found, and where ΔR values exceed 160 years,’ ΔR values should be given as ‘14C yrs’ rather than ‘years’
Line 370: ‘there is also a positive correlation between sea-ice cover and reservoir age’. Are the correlations significant?
Line 429: ‘these values remain only valid for the modern situation’ Insert ‘pre-bomb’ before modern because the values would not be valid for post-bomb samples.
Fig. 1 caption: Need to define WGC, NFL, EGC.
Also. ‘Areas of deep convection in the Labrador Sea and north of Iceland are colored yellow’. These look light green on top of the blue background - perhaps 'shaded light green' would be better
Fig 2. Given the results, is there justification for separate zones for the Greenland coastal waters since ΔR values overlap?
Suppl. Fig 2. Sample numbers on Suppl. Fig 2 would be helpful for comparison of species and feeding habits
References:
Angulo, R. J., Reimer, P. J., De Souza, M. C., Scheel-Ybert, R., Tenório, M. C., Disaró, S. T. & Gaspar, M. D. 2007. A tentative determination of upwelling influence on the paleo-surficial marine water reservoir effect in southeastern Brazil. Radiocarbon, 49, 1-5.
Heaton, T. J., Bard, E., Bronk Ramsey, C., Butzin, M., Hatté, C., Hughen, K. A., Köhler, P. & Reimer, P. J. 2023. A response to community questions on the MARINE20 radiocarbon age calibration curve: marine reservoir ages and the calibration of 14c samples from the oceans. Radiocarbon, 65, 247-273.
Hua, Q., Webb, G. E., Zhao, J.-X., Nothdurft, L. D., Lybolt, M., Price, G. J. & Opdyke, B. N. 2015. Large variations in the Holocene marine radiocarbon reservoir effect reflect ocean circulation and climatic changes. Earth and Planetary Science Letters, 422, 33-44.
O'Connor, S., Ulm, S., Fallon, S. J., Barham, A. & Loch, I. 2010. Pre-bomb marine reservoir variability in the Kimberley region, Western Australia. Radiocarbon, 52, 1158-1165.

Citation: https://doi.org/10.5194/gchron-2023-7-RC1
- AC1: 'Reply on RC1', Christof Pearce, 22 Aug 2023
  
  We thank Paula Reimer for her valuable comments and suggestions. Please, find our detailed response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC1
RC2:
'Comment on gchron-2023-7', Anonymous Referee #2, 14 Jun 2023

Pearce et al. present about 100 new marine radiocarbon (¹⁴C) reservoir ages (MRA) of coastal and shelf waters around Greenland, Baffin Island, Newfoundland, and Iceland. The data result from ¹⁴C measurements on pre-bomb molluscs retrieved from museums. The MRA results are binned to seven regions and discussed with respect to the global Marine20 ¹⁴C calibration curve in terms of the regional MRA correction, ∆R₂₀. The authors also discuss their ∆R₂₀ results in the light of specific factors such as sample depth, sea ice cover and feeding habits.
The manuscript is well written, the presentation is clear, and the dataset is an important contribution to the MRA / ∆R data base. However, there are a few minor issues that should be addressed before publication in GChron (L = line):
L 30: The half-life of ¹⁴C has been slightly revised to 5700 years (e.g., Audi et al., 2003; Bé and Chechev, 2012; Kutschera, 2013)
L 124 "marine mammals": "marine" should be removed
Figure 1:

(i) Add a depth scale (such as in Fig. 1 by Pieńkowski et al. 2022)

(ii) "NFL", "WGC", and "EGC" should be also explained in the caption.
L 214-216 (and Figure 2): Is there a hard objective criterion to separate the three southernmost data points in East Greenland from region 7?
Figure 2: Explain "CS"
L 353: Explain "mwd"
Figure 3: Would it make sense to indicate the positions of the outliers in the inserted map?
L 376: Explain "mwd"
Figure 4:

(i) As ∆R depends on the sea ice concentration, the coordinate axes should be swapped. The situation is different from Figure 3 where the independent variable (usually plotted along the horizontal axis) is depth (typically plotted in vertical direction).

(ii) Can you quantify the trends, and are they significant? I wonder if the trends are still visible once the coordinate axes have been swapped.

(iii) Would it make sense to indicate the position of the outlier in the inserted map?

References:
Audi, G., Bersillon, O., Blachot, J., and Wapstra, A. H.: The Nubase evaluation of nuclear and decay properties, Nuclear Physics A, 729, 3–128, https://doi.org/10.1016/j.nuclphysa.2003.11.001, 2003.
Bé, M.-M. and Chechev, V. P.: 14C - Comments on evaluation of decay data, Laboratoire National Henri Becquerel, Gif-sur-Yvette, http://www.lnhb.fr/nuclides/C-14_com.pdf, 2012.
Kutschera, W.: Applications of accelerator mass spectrometry, International Journal of Mass Spectrometry, 349–350, 203–218, https://doi.org/10.1016/j.ijms.2013.05.023, 2013.

Citation: https://doi.org/10.5194/gchron-2023-7-RC2
- AC2: 'Reply on RC2', Christof Pearce, 22 Aug 2023
  
  We thank anonymous reviewer R2 for their valuable comments and suggestions. Please, find our detailed response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC2
CC1:
'Comment on gchron-2023-7', Elisabeth Michel, 23 Jun 2023

The authors present new ¹⁴C reservoir ages for surface and deep waters of the North Atlantic and Nordic seas : Labrador sea, Baffin Bay and Iceland Sea, from shell museum collections. The shells have been collected from 1865 to 1931. They present a nice review of existing reservoir ages.
First, they compare the results from shells that were preserved in ethanol in museum collections and those who were dry samples. They found that the mean dry samples ¹⁴C reservoir age is much higher than the mean of ethanol preserved samples and argue that the dry samples might be dead since a long time when they were collected.
The authors propose regional ¹⁴C reservoir ages within 7 different geographic zones, considering both their new results and 14C reservoir ages from the Marine Reservoir Age Database (Reimer and Reimer 2001) considering only samples preserved in ethanol.
For the relevance of the results, the authors also consider the results of deposit feeders compared to suspension feeder.
For the interpretation of the regional ¹⁴C reservoir age they consider the depth of collection of the different samples and shortly discuss the impact of ocean circulation and sea ice.

This paper is mainly a data paper, the discussion of the result is rather short and do not discuss in depth the different factors that could impact their regional ¹⁴C reservoir age.
Following are some detailed comments and also some ideas for a more complete discussion concerning the regional results.

Considering dry samples, I wonder if there is any evidence on the shell, muscle marks or the like, to tell whether the specimen was collected alive or could have been dead for a long time.
For the deposit and suspension feeders, the authors should compare the results zone by zone as they indicated that the ∆R was very different from one zone to another. They could also check the dispersion for species for which the feeding habit is unknown. It would be better to discuss first the aspect linked to the mollusk : dry and ethanol preserved samples, feeding habitat and after all the physical parameters: sea ice, depth and circulation.
One question that is not addressed, do the author have an idea of the mean lifetime of the different mollusk?
It seems that the authors choose to include only ¹⁴C ∆R measured on molluks. I wonder why they do not compare their results with 14C measurements made directly on DIC of sea water in the early fifties like for example Fonselius and Östlund, 1959 Tellus.

What is the most impressive is the dispersion of the ¹⁴∆R data within some of the geographic zones. The authors discuss the impact of sea-ice checking if a relationship exist between the annual average sea ice concentration of a sample location and its ¹⁴C reservoir age (fig. 4). The regressions and their statistics for the different geographic zones are necessary if the authors want to demonstrate that the regional relationships are significant. Furthermore during formation of sea ice the carbon sink in the ocean might be effective thus the impact of non-perennial sea-ice is not obvious.
The authors argue that Heaton et al., 2020, explain that the Marine20 does not apply to the polar regions because of sea ice. Heaton et al, 2020 is as much about ocean circulation as it is about sea ice.
The role of Ocean Circulation could be considered considering fluxes along the different straits. Furthermore the influence of Atlantic and Artic water masses might changes with time, for example linked to North Atlantic Oscillation. Thus a time evolution of ¹⁴C ∆R within the geographical zones could be also discussed and might explain partly the large dispersion of the results?
∆R could be also influence by continental waters with old ¹⁴C DIC coming from under the ice like in the Ross Sea (Mikucki et al., 2009). This point is not discussed.

Figures: even if the projections does not make it easy and they will not be regularly spaced, it would be nice to have some latitudinal and longitudinal tics on the borders of figures 1, 2 and suppl. Fig.1.

Citation: https://doi.org/10.5194/gchron-2023-7-CC1
- AC3: 'Reply on CC1', Christof Pearce, 22 Aug 2023
  
  We appreciate the valuable comments and suggestions of Elisabeth Michel to our manuscript. Please, find our detailed response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC3
RC3:
'Comment on gchron-2023-7', Matt O'Regan, 29 Jun 2023

This is a very nicely written paper presenting 92 new radiocarbon dates on pre-bomb mollusks collected from around Greenland (with the exception of its northern Arctic Ocean margin). In addition to the utility of these new dates for constraining regional reservoir corrections, I think the manuscript is timely in presenting a nice practical discussion (and examples) on the need to update reservoir corrections when using the new Marine20 calibration curve.
The comparisons of dR with water depth and sea ice coverage are interesting in highlighting patterns, although somewhat inconclusive in identifying a cause/explanation for the variability. I do not think this limits the scientific contribution made by the paper, and certainly sets the stage for future work needed to understand this variability. I believe this would require a considerable amount of work, and could potentially start with moving away from water depth and looking at the variability in Temperature-Salinity space to see if ages cluster in specific water masses. However, I do not think this is a necessary addition to this work, which very well suited for publication in Geochronology in its current form.
I do feel one aspect the paper is missing is a discussion on the limited, but rather informative data on Holocene dR values from the central Arctic Ocean. Specifically the inferred differences between the age of Pacific and Atlantic waters that are found in the interior Arctic, and should be impacting the age of surface waters(?) in northern Baffin Bay. For example, West et al (2022), Geochemistry, Geophysics, Geosystems (doi: 10.1029/2021GC010187) used tephra from the Aniakchak eruption circa 3.6 ka in two cores from the Chukchi Sea - one at 50 m depth (Pacific water) and one at 120 m depth (likely Atlantic water) - to show that the dR (using Marine 20) for benthic foraminifera and mollusks at these sites was about 330 years for Pacific waters and 205 years for Atlantic waters. These seem to be somewhat consistent with the larger dR values in sections 3 and 4 from Northern Baffin Bay. It would be nice to see some discussion about the influence of Arctic outflow and the water masses involved on the dR values in Northern Baffin Bay. Currently these are described simply as 'outflow' from the Arctic, which could easily be expanded to detail the role and age of Pacific and Atlantic waters in this outflow.
Overall, I feel this is a great contribution that will provide significant support to future paleoceanographic work around Greenland.

Citation: https://doi.org/10.5194/gchron-2023-7-RC3
- AC4: 'Reply on RC3', Christof Pearce, 22 Aug 2023
  
  We thank Matt O'Regan for his encouraging comments and valuable suggestions to improve our manuscript. Please, find our short response in the attached file.
  
  Citation: https://doi.org/10.5194/gchron-2023-7-AC4

Christof Pearce et al.

Supplement

https://doi.org/10.5194/gchron-2023-7-supplement

Christof Pearce et al.

Viewed

Total article views: 593 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
377	186	30	593	37	6	4

HTML: 377
PDF: 186
XML: 30
Total: 593
Supplement: 37
BibTeX: 6
EndNote: 4

Views and downloads (calculated since 17 Apr 2023)

Month	HTML	PDF	XML	Total
Apr 2023	85	64	4	153
May 2023	61	26	4	91
Jun 2023	87	25	10	122
Jul 2023	41	16	0	57
Aug 2023	56	20	9	85
Sep 2023	25	14	2	41
Oct 2023	15	15	0	30
Nov 2023	7	6	1	14

Cumulative views and downloads (calculated since 17 Apr 2023)

Month	HTML	PDF	XML	Total
Apr 2023	85	64	4	153
May 2023	61	26	4	91
Jun 2023	87	25	10	122
Jul 2023	41	16	0	57
Aug 2023	56	20	9	85
Sep 2023	25	14	2	41
Oct 2023	15	15	0	30
Nov 2023	7	6	1	14

Viewed (geographical distribution)

Total article views: 582 (including HTML, PDF, and XML) Thereof 582 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 14 Nov 2023

Short summary

Reliable chronologies lie at the base of paleoclimatological reconstructions. When working with marine sediment core, the most common dating tool for recent sediments is radiocarbon, but this requires calibration to convert to calendar ages. This calibration requires knowledge of the marine radiocarbon reservoir age and this is known to vary in space and time. In this study we provide 92 new radiocarbon measurements to improve our knowledge of the reservoir age around Greenland.


Total:	0
HTML:	0
PDF:	0
XML:	0