<< Previous article

The heterogeneous nature of Fe delivery from melting icebergs

M.J. Hopwood1,

1GEOMAR, Helmholtz Centre for Ocean Research, 24148 Kiel, Germany

C. Cantoni2,

2CNR-ISMAR, Marine Science Institute, 34149 Trieste, Italy

J.S. Clarke1,

1GEOMAR, Helmholtz Centre for Ocean Research, 24148 Kiel, Germany

S. Cozzi2,

2CNR-ISMAR, Marine Science Institute, 34149 Trieste, Italy

E.P. Achterberg1

1GEOMAR, Helmholtz Centre for Ocean Research, 24148 Kiel, Germany

Affiliations  |  Corresponding Author  |  Cite as  |  Funding information

Hopwood, M.J., Cantoni, C., Clarke, J.S., Cozzi, S., Achterberg, E.P. (2017) The heterogeneous nature of Fe delivery from melting icebergs. Geochem. Persp. Let. 3, 200–209.

European Commission (OCEAN-CERTAIN, FP7- ENV-2013-6.1-1; no: 603773)

Geochemical Perspectives Letters v3, n2  |  doi: 10.7185/geochemlet.1723
Received 27 September 2016  |  Accepted 25 April 2017  |  Published 7 June 2017

Keywords: icebergs, iron, Fe, Svalbard, Arctic



top

Abstract


The micronutrient iron (Fe) can be transported from marine terminating glaciers to the ocean by icebergs. There are however few observations of iceberg Fe content, and the flux of Fe from icebergs to the offshore surface ocean is poorly constrained. Here we report the dissolved Fe (DFe), total dissolvable Fe (TdFe) and ascorbic acid extractable Fe (FeAsc) sediment content of icebergs from Kongsfjorden, Svalbard. The concentrations of DFe (range 0.63 nM – 536 nM, mean 37 nM, median 6.5 nM) and TdFe (range 46 nM – 57 µM, mean 3.6 µM, median 144 nM) both demonstrated highly heterogeneous distributions and there was no significant correlation between these two fractions. FeAsc (range 0.0042 to 0.12 wt. %) was low compared to both previous measurements in Kongsfjorden and to current estimates of the global mean. FeAsc content per volume ice did however, as expected, show a significant relationship with sediment loading (which ranged from < 0.1 – 234 g L-1 of meltwater). In the Arctic, icebergs lose their sediment load faster than ice volume due to the rapid loss of basal ice after calving. We therefore suggest that the loss of basal ice is a potent mechanism for the reduction of mean TdFe and FeAsc per volume of iceberg. Delivery of TdFe and FeAsc to the ocean is thereby biased towards coastal waters where, in Kongsfjorden, DFe (18 ± 17 nM) and TdFe (mean 8.1 µM, median 3.7 µM) concentrations were already elevated.

Figures and Tables

Figure 1 Surface fjord sample locations in Kongsfjorden.

Figure 2 Median FeAsc (wt. %) with 25/75th (boxes) and 10/90th (whiskers) percentiles (outliers also shown) for iceberg embedded (n = 34) and iceberg surface (n = 20) sediment.

Figure 3 DFe and TdFe (both nM, plotted as log10, TdFe shown minus DFe) for 28 discrete iceberg samples showed no clear relationship.

Table 1 Comparing data for Kongsfjorden from this and prior work suggests a critical difference in both FeAsc (wt. %) and in the scaling of FeAsc to iceberg sediment load (g L-1 of ice melt). *The suggested 0.5 g L-1 sediment loading is used for data from Raiswell et al. (2016). **For our study, measured sediment loadings were used for each sample. As sediment-rich ice was specifically targeted, the calculated mean/median should be over-estimates.

Figure 4 FeAsc (µmol L-1 melted ice) increased with sediment load (g L-1 melted ice), but it is unclear if the relationship remains linear at high (>50 g L-1) loadings.

Figure 1 Figure 2 Figure 3 Table 1 Figure 4

View all figures and tables  


Supplementary Figures and Tables

Table S-1 Analysis of Certified Reference Materials for Fe concentration (± standard deviation of at least 6 measurements).

Table S-2 FeAsc concentration reported for various sediment samples collected around Kongsfjorden. For iceberg embedded sediment with sediment loadings >0.1 g L-1, FeAsc is reported both as wt. % and per volume of meltwater.

Table S-3 Dissolved (<0.2 µm) Fe and total dissolvable Fe for 15 surface water samples collected in Kongsfjorden (July 2015).

Table S-4 Dissolved (<0.2 µm) Fe and total dissolvable Fe for 28 randomly collected Kongsfjorden iceberg samples (July 2015).

Table S-1 Table S-2 Table S-3 Table S-4

View all supplementary figures and tables  


top

Introduction


Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990

Martin, J.H., Gordon, R.M., Fitzwater, S.E. (1990) Iron in Antarctic waters. Nature 345, 156–158.

; De Baar et al., 1995

De Baar, H.J.W., De Jong, J.T.M., Bakker, D.C.E., Loscher, B.M., Veth, C., Bathmann, U., Smetacek, V. (1995) Importance of iron for plankton blooms and carbon dioxide drawdown in the Southern Ocean. Nature 373, 412–415.

; Loscher et al., 1997

Loscher, B.M., De Baar, H.J.W., De Jong, J.T.M., Veth, C., Dehairs, F. (1997) The distribution of Fe in the Antarctic Circumpolar Current. Deep Sea Research Part II: Topical Studies in Oceanography 44, 143–187.

) and particulate (>0.2 µm) phases (Hart, 1934

Hart, T.J. (1934) Discovery Reports. Discovery Reports 8, 1–268.

; Lin et al., 2011

Lin, H., Rauschenberg, S., Hexel, C.R., Shaw, T.J., Twining, B.S. (2011) Free-drifting icebergs as sources of iron to the Weddell Sea. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1392–1406.

; Shaw et al., 2011

Shaw, T.J., Raiswell, R., Hexel, C.R., Vu, H.P., Moore, W.S., Dudgeon, R., Smith Jr., K.L. (2011) Input, composition, and potential impact of terrigenous material from free-drifting icebergs in the Weddell Sea. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1376–1383.

). Icebergs should thus constitute a source of the micronutrient Fe to offshore polar waters (Raiswell et al., 2008

Raiswell, R., Benning, L.G., Tranter, M., Tulaczyk, S. (2008) Bioavailable iron in the Southern Ocean: the significance of the iceberg conveyor belt. Geochemical Transactions 9, doi:10.1186/1467-4866-9-7.

). As the rate of iceberg calving in polar seas oscillates on glacial to inter-glacial timescales (Bond et al., 1992

Bond, G., Heinrich, H., Broecker, W.S., Labeyrie, L., McManus, J.F., Andrews, J.T., Huon, S., Jantschik, R., Clasen, S., Simet, C., Tedesco, K., Klas, M., Bonani, G., Ivy, S. (1992) Evidence for massive discharges of icebergs into the North Atlantic ocean during the last glacial period, Nature 360, 245–249.

), and recent climate change has increased the discharge of calved ice volume from both the Antarctic (Paolo et al., 2015

Paolo, F.S., Fricker, H.A., Padman, L. (2015) Volume loss from Antarctic ice shelves is accelerating. Science 348, 327–331.

) and Greenlandic (Bamber et al., 2012

Bamber, J., van den Broeke, M., Ettema, J., Lenaerts, J., Rignot, E. (2012) Recent large increases in freshwater fluxes from Greenland into the North Atlantic. Geophysical Research Letters 39, L19501, doi:10.1029/2012gl052552.

) ice sheets, Fe delivery from icebergs may also change. Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990

Martin, J.H., Gordon, R.M., Fitzwater, S.E. (1990) Iron in Antarctic waters. Nature 345, 156–158.

, 1991

Martin, J.H., Gordon, R.M., Fitzwater, S.E. (1991)The case for iron. Limnology and Oceanography 36, 1793–1802.

; Moore et al., 2013

Moore, C.M., Mills, M.M., Arrigo, K.R., Berman-Frank, I., Bopp, L., Boyd, P.W., Galbraith, E.D., Geider, R.J., Guieu, C., Jaccard, S.L., Jickells, T.D., La Roche, J., Lenton, T.M., Mahowald, N.M., Maranon, E., Marinov, I., Moore, J.K., Nakatsuka, T., Oschlies, A., Saito, M.A., Thingstad, T.F., Tsuda, A., Ulloa, O. (2013) Processes and patterns of oceanic nutrient limitation. Nature Geoscience 6, 701–710.

), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007

Smith Jr., K.L., Robison, B.H., Helly, J.J., Kaufmann, R.S., Ruhl, H.A., Shaw, T.J., Twining, B.S., Vernet, M. (2007) Free-drifting icebergs: Hot spots of chemical and biological enrichment in the Weddell Sea. Science 317, 478–482.

; Schwarz and Schodlok, 2009

Schwarz, J.N., Schodlok, M.P. (2009) Impact of drifting icebergs on surface phytoplankton biomass in the Southern Ocean: Ocean colour remote sensing and in situ iceberg tracking. Deep Sea Research Part I: Oceanographic Research Papers 56, 1727–1741.

; Smith et al., 2011

Smith, K.L., Sherman, A.D., Shaw, T.J., Murray, A.E., Vernet, M., Cefarelli, A.O. (2011) Carbon export associated with free-drifting icebergs in the Southern Ocean. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1485–1496.

), a change in iceberg Fe supply could significantly affect marine primary productivity. Yet there remain large uncertainties concerning the magnitude of iceberg Fe supply and its effect(s) on marine ecosystems. For example, calculated phytoplankton Fe utilisation is considerably less than present estimates of iceberg Fe supply to the Weddell Sea (Boyd et al., 2012

Boyd, P.W., Arrigo, K.R., Strzepek, R., Van Dijken, G.L. (2012) Mapping phytoplankton iron utilization: Insights into Southern Ocean supply mechanisms. Journal Geophysical Research: Oceans 117, doi:10.1029/2011JC007726.

). The reason for this is unclear, yet it demonstrates the difficulty in isolating the contribution of icebergs to the marine Fe cycle.

Observations of DFe concentrations in iceberg meltwater are sparse, but the available data does suggest a heterogeneous distribution, with DFe ranging 4–600 nM in Antarctic (Lin et al., 2011

Lin, H., Rauschenberg, S., Hexel, C.R., Shaw, T.J., Twining, B.S. (2011) Free-drifting icebergs as sources of iron to the Weddell Sea. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1392–1406.

) and 3–300 nM in Greenlandic (Hopwood et al., 2016

Hopwood, M.J., Connelly, D.P., Arendt, K.E., Juul-Pedersen, T., Stinchcombe, M., Meire, L., Esposito, M., Krishna, R. (2016) Seasonal changes in Fe along a glaciated Greenlandic fjord. Frontiers in Earth Sciences 4, doi:10.3389/feart.2016.00015.

) iceberg melt. The distribution of particulate Fe (which includes FeAsc) is also expected to be heterogeneous due to the presence of embedded sediment-rich layers that account for only a small fraction of total iceberg volume (Lin et al., 2011

Lin, H., Rauschenberg, S., Hexel, C.R., Shaw, T.J., Twining, B.S. (2011) Free-drifting icebergs as sources of iron to the Weddell Sea. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1392–1406.

; Raiswell, 2011

Raiswell, R. (2011) Iceberg-hosted nanoparticulate Fe in the Southern Ocean: Mineralogy, origin, dissolution kinetics and source of bioavailable Fe. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1364–1375.

; Raiswell et al., 2016

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

). Whilst TdFe data for icebergs is sparse, iceberg FeAsc content has previously been estimated in multiple catchments worldwide (Raiswell et al., 2016

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

) producing a mean global content of 2.7–17 µM. However, FeAsc content and offshore iceberg FeAsc fluxes are normally calculated using a mean sediment loading (0.5 g L-1 is widely used as outlined in Raiswell et al. (2016)

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

) with considerable uncertainty generally acknowledged in this value. Here we combine the analysis of DFe, TdFe, FeAsc and iceberg sediment load in order to provide a well constrained assessment of iceberg-Fe content within a single catchment.

top

Methods


A FeAsc dataset was compiled for icebergs in Kongsfjorden with visible embedded or surface sediment sampled from small boats in July 2015 and August 2016. Sediment from pro-glacial streambeds in the catchment, embedded sediment from Kongsvegen glacier surface, and embedded sediment ~100 m inside an ice crevasse (on Midtre Lovénbreen glacier) was also collected (Fig. 1) for comparative purposes. FeAsc leaches were conducted on wet sediment as per Raiswell et al. (2010)

Raiswell, R., Vu, H.P., Brinza, L., Benning, L.G. (2010) The determination of labile Fe in ferrihydrite by ascorbic acid extraction: Methodology, dissolution kinetics and loss of solubility with age and de-watering. Chemical Geology 278, 70–79.

, with leached Fe determined by measuring absorbance (λ = 562 nm) before and after the addition of ferrozine (as detailed in Supplementary Information Methods).

Separately, ice samples (1–2 kg) were randomly collected from small boats (July 2015). The meltwater was acidified to pH < 2. After storage for 12 months, DFe and TdFe were measured by inductively coupled plasma mass spectroscopy (further details in Supplementary Information Methods).


Figure 1 Surface fjord sample locations in Kongsfjorden.
Full size image | Download in Powerpoint

top

Results


The FeAsc concentration is reported for 116 different sediment samples (Table S-2) including 58 iceberg samples collected from ice with visible embedded sediment. FeAsc ranged from 0.0042 to 0.12 wt. % in iceberg embedded sediment (Fig. 2). Ice sediment content ranged from <0.1 to 234 g L-1 of meltwater, close to the 0.2–200 g L-1 range previously reported in Svalbard by Dowdeswell and Dowdeswell (1989)

Dowdeswell, J.A., Dowdeswell, E.K. (1989) Debris in Icebergs and Rates of Glaci-Marine Sedimentation: Observations from Spitsbergen and a Simple Model. The Journal of Geology 97, 221–231.

. Combining FeAsc (wt. %) and sediment load (g L-1 of melted ice) produced a median FeAsc ice content of 2.5 µM. Given that our sampling strategy was to target sediment-rich ice, this should thereby be an over-estimate of median iceberg FeAsc content (L-1 of melted ice) in Kongsfjorden.


Figure 2 Median FeAsc (wt. %) with 25/75th (boxes) and 10/90th (whiskers) percentiles (outliers also shown) for iceberg embedded (n = 34) and iceberg surface (n = 20) sediment.
Full size image | Download in Powerpoint

The DFe and TdFe concentrations are reported in parallel for 28 randomly collected iceberg samples (Table S-4). TdFe ranged from 46 nM to 57 µM (mean 3.6 µM, median 144 nM) and the range of 0.63 nM – 536 nM for DFe was similarly broad (mean 37 nM, median 6.5 nM). There was no significant correlation between particulate Fe (TdFe minus DFe) and DFe in these samples (Fig. 3), suggesting that DFe was not specifically associated with sediment laden ice. For comparison, DFe in surface fjord waters averaged 18 ± 17 nM (15 surface stations, Fig. 1) and TdFe ranged widely from 1.1 to 52 µM (mean 8.1 µM, median 3.7 µM) (Table S-3).

top

Discussion


As has been demonstrated in this study and elsewhere (e.g., Markussen et al., 2016

Markussen, T.N., Elberling, B., Winter, C., Andersen, T.J. (2016) Flocculated meltwater particles control Arctic land-sea fluxes of labile iron. Scientific Reports 6, 24033.

), surface waters in stratified, glaciated fjords can exhibit extremely high TdFe concentrations due to the presence of glacially derived particle plumes. TdFe concentrations in surface waters of Kongsfjorden (mean 8.1 µM, median 3.7 µM) exceeded those in icebergs (3.6 µM and 144 nM, respectively). In the Arctic, a large fraction of iceberg melt occurs in these inshore, high TdFe waters before icebergs are able to deliver Fe to the offshore environment. Accounting for this near-shore loss in flux calculations for iceberg derived Fe supply to the open ocean is difficult. In two Greenlandic catchments, Ilulissat Fjord and Sermilik Fjord, overall in-fjord iceberg volume losses were estimated to be >50 % (Enderlin et al., 2016

Enderlin, E.M., Hamilton, G.S., Straneo, F., Sutherland, D.A. (2016) Iceberg meltwater fluxes dominate the freshwater budget in Greenland’s iceberg-congested glacial fjords. Geophysical Research Letters 43, 11287–11294.

), tentatively supporting the 50 % inshore iceberg volume loss used to estimate offshore FeAsc fluxes by Raiswell et al. (2016)

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

. However this assumes that changes in total iceberg Fe content are directly proportional to changes in total ice volume.

All measured Fe phases (DFe, TdFe and FeAsc) in Kongsfjorden were very heterogeneously distributed within the ice. For TdFe and FeAsc (but not DFe, Fig. 3), this can specifically be attributed to the heterogeneous distribution of ice embedded sediment. In the Arctic, iceberg-borne sediment is known to be lost from icebergs faster than ice volume (Mugford and Dowdeswell, 2010

Mugford, R.I., Dowdeswell, J.A. (2010) Modeling iceberg-rafted sedimentation in high-latitude fjord environments. Journal of Geophysical Research: Earth Surface 115, doi:10.1029/2009JF001564.

) due to its association with basal ice. Thus we expect that the mean TdFe content per volume of an iceberg should decline with time after calving. A model for Kangerdlugssuaq Fjord (Greenland) shows that whilst icebergs lose 20–30 % ice volume within this fjord, the corresponding in-fjord sediment loss is 70–85 % (Mugford and Dowdeswell, 2010

Mugford, R.I., Dowdeswell, J.A. (2010) Modeling iceberg-rafted sedimentation in high-latitude fjord environments. Journal of Geophysical Research: Earth Surface 115, doi:10.1029/2009JF001564.

). Only a relatively small iceberg volume loss (<20 %) is thereby likely required for the majority of TdFe content to be lost from icebergs. In Kongsfjorden, where summer melting of calved ice is quite rapid due to relatively warm surface seawater (4–5 °C throughout July-August 2016), the post-calving age of an iceberg is therefore likely a critical factor in determining its TdFe content. Sediment loss should also affect mean FeAsc content in the same way, however FeAsc losses may be offset from TdFe losses if significant processing of surface sediment occurs on the timescale of iceberg Fe delivery (Raiswell et al., 2016

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

).


Figure 3 DFe and TdFe (both nM, plotted as log10, TdFe shown minus DFe) for 28 discrete iceberg samples showed no clear relationship.
Full size image | Download in Powerpoint

In Kongsfjorden, Raiswell et al. (2016)

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

reported a FeAsc range of 0.016–0.37 wt. % (n = 14), with a mean of 0.14 wt. % and median of 0.092 wt. %; equivalent to 1.4–33 µM, 12 µM and 8.2 µM, respectively when using the suggested mean sediment loading of 0.5 g L-1. Comparing our data both as wt. % and as a µM concentration (calculated using measured sediment loading for each sample, range 0.1–234 g L-1, Table S-2), our FeAsc (wt. %) is consistently lower. Yet our mean FeAsc per volume is much higher (51 µM), because our measured sediment loadings were often greater than the assumed mean of 0.5 g L-1. These differences generally highlight the very high spatial variability in iceberg sediment load and thus TdFe and FeAsc content even within a single fjord.

Table 1 Comparing data for Kongsfjorden from this and prior work suggests a critical difference in both FeAsc (wt. %) and in the scaling of FeAsc to iceberg sediment load (g L-1 of ice melt). *The suggested 0.5 g L-1 sediment loading is used for data from Raiswell et al. (2016)

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

. **For our study, measured sediment loadings were used for each sample. As sediment-rich ice was specifically targeted, the calculated mean/median should be over-estimates.

a This studyb Raiswell et al. (2016)

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

a/b %
FeAsc / wt. %Mean0,0210,1416

Median0,0150,09217
FeAsc / µM (per litre of ice melt)*Mean<59 **12480

Median<2.5 **8,231
Download in Excel

Some methodological differences between this study and previous work could be important for the difference in FeAsc (wt. %) (Table 1). In our study, the sediment was not sieved to remove anomalous large particles. Yet a relatively large sub-sample mass was used with good reproducibility demonstrated. For glacial flour particles of <1 mm, it has previously been demonstrated that the change in FeAsc (wt. %) with particle size is not pronounced (Hopwood et al., 2014

Hopwood, M.J., Statham, P.J., Tranter, M., Wadham, J.L. (2014) Glacial flours as a potential source of Fe(II) and Fe(III) to polar waters. Biogeochemistry 118, 443–452.

; Raiswell et al., 2016

Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.

), but this may not be the case for larger particles. Moreover, in this study sediment was processed in Svalbard with no prolonged storage between collection and analysis. Whilst dried sediment exhibits a rapid decline in FeAsc wt. % (Raiswell et al., 2010

Raiswell, R., Vu, H.P., Brinza, L., Benning, L.G. (2010) The determination of labile Fe in ferrihydrite by ascorbic acid extraction: Methodology, dissolution kinetics and loss of solubility with age and de-watering. Chemical Geology 278, 70–79.

), it is not clear how storage of ice or wet sediments affects FeAsc.

Furthermore, there are the critical issues of heterogeneity and of the non-linearity between iceberg sediment and iceberg volume losses. In-fjord iceberg volume loss should correspond to a disproportionately high loss of iceberg embedded sediment (Mugford and Dowdeswell, 2010

Mugford, R.I., Dowdeswell, J.A. (2010) Modeling iceberg-rafted sedimentation in high-latitude fjord environments. Journal of Geophysical Research: Earth Surface 115, doi:10.1029/2009JF001564.

), and thereby also FeAsc and TdFe. There are no quantitative measures of iceberg age or volume loss for our dataset and the residence time of ice in Kongsfjorden is strongly affected by meteorological conditions and thus subject to high short-term variability. Nonetheless, a difference in the post-calving age of ice sampled between different datasets could correspond to large shifts in iceberg FeAsc and TdFe content. Increased iceberg age would be expected to correspond to lower mean sediment load, and thereby lower TdFe and FeAsc per volume. A reduction in basal sediment load could also explain a difference in FeAsc (wt. %) content if FeAsc (wt. %) is enriched in basal ice compared to non-basal ice. FeAsc (µmol L-1) is correlated with sediment load (Fig. 4), but assessing whether changes in sediment load affect FeAsc (wt. %) is complicated by the lack of any parameter to account for the post-calving age of ice and by the highly variable bedrock composition across Kongsfjorden (see for example Hjelle, 1993

Hjelle, A. (1993) The geology of Svalbard: Oslo. Polarhåndbok nr. 6. Norsk Polarinstitutt, Oslo.

).


Figure 4 FeAsc (µmol L-1 melted ice) increased with sediment load (g L-1 melted ice), but it is unclear if the relationship remains linear at high (>50 g L-1) loadings.
Full size image | Download in Powerpoint

top

Conclusions


Whilst median DFe (6.5 nM) and TdFe (144 nM) concentrations in Kongsfjorden were within the range of concentrations reported elsewhere globally, the median FeAsc concentration (2.5 µM) measured was considerably lower than that reported previously in Kongsfjorden, and compared to present estimates of the global mean, despite the very high sediment loadings observed (<0.1 – 234 g L-1). Generally in the Arctic, a sharp decline in the mean FeAsc and TdFe per volume of meltwater from icebergs with time after calving would be expected due to the preferential loss of iceberg basal ice, as modelled by Mugford and Dowdeswell (2010)

Mugford, R.I., Dowdeswell, J.A. (2010) Modeling iceberg-rafted sedimentation in high-latitude fjord environments. Journal of Geophysical Research: Earth Surface 115, doi:10.1029/2009JF001564.

. Iceberg derived fluxes of TdFe and FeAsc are thereby biased towards delivery in near-shore waters and offshore fluxes are likely much less than if TdFe and FeAsc were homogeneously distributed throughout icebergs.

top

Glossary


‘Fe’ refers to all iron phases.
‘DFe’, dissolved Fe, refers to all Fe phases <0.2 µm.
‘FeAsc’ is the ferrihydrite content of sediment, defined by Raiswell et al. (2010)

Raiswell, R., Vu, H.P., Brinza, L., Benning, L.G. (2010) The determination of labile Fe in ferrihydrite by ascorbic acid extraction: Methodology, dissolution kinetics and loss of solubility with age and de-watering. Chemical Geology 278, 70–79.

.
‘TdFe’ is all Fe soluble at pH < 2, inclusive of DFe and should also include any FeAsc present in unfiltered meltwater.

top

Acknowledgements


Financial aid from the European Commission (OCEAN-CERTAIN, FP7- ENV-2013-6.1-1; no: 603773) is gratefully acknowledged. 2016 fieldwork was conducted during the CNR Dirigibilite Italia hosted project ‘pH in Svalbard’.

Editor: Liane G. Benning

top

References



Bamber, J., van den Broeke, M., Ettema, J., Lenaerts, J., Rignot, E. (2012) Recent large increases in freshwater fluxes from Greenland into the North Atlantic. Geophysical Research Letters 39, L19501, doi:10.1029/2012gl052552.
Show in context

As the rate of iceberg calving in polar seas oscillates on glacial to inter-glacial timescales (Bond et al., 1992), and recent climate change has increased the discharge of calved ice volume from both the Antarctic (Paolo et al., 2015) and Greenlandic (Bamber et al., 2012) ice sheets, Fe delivery from icebergs may also change.
View in article


Bond, G., Heinrich, H., Broecker, W.S., Labeyrie, L., McManus, J.F., Andrews, J.T., Huon, S., Jantschik, R., Clasen, S., Simet, C., Tedesco, K., Klas, M., Bonani, G., Ivy, S. (1992) Evidence for massive discharges of icebergs into the North Atlantic ocean during the last glacial period, Nature 360, 245–249.
Show in context

As the rate of iceberg calving in polar seas oscillates on glacial to inter-glacial timescales (Bond et al., 1992), and recent climate change has increased the discharge of calved ice volume from both the Antarctic (Paolo et al., 2015) and Greenlandic (Bamber et al., 2012) ice sheets, Fe delivery from icebergs may also change.
View in article


Boyd, P.W., Arrigo, K.R., Strzepek, R., Van Dijken, G.L. (2012) Mapping phytoplankton iron utilization: Insights into Southern Ocean supply mechanisms. Journal Geophysical Research: Oceans 117, doi:10.1029/2011JC007726.
Show in context

For example, calculated phytoplankton Fe utilisation is considerably less than present estimates of iceberg Fe supply to the Weddell Sea (Boyd et al., 2012).
View in article


De Baar, H.J.W., De Jong, J.T.M., Bakker, D.C.E., Loscher, B.M., Veth, C., Bathmann, U., Smetacek, V. (1995) Importance of iron for plankton blooms and carbon dioxide drawdown in the Southern Ocean. Nature 373, 412–415.
Show in context

Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990; De Baar et al., 1995; Loscher et al., 1997) and particulate (>0.2 µm) phases (Hart, 1934; Lin et al., 2011; Shaw et al., 2011).
View in article


Dowdeswell, J.A., Dowdeswell, E.K. (1989) Debris in Icebergs and Rates of Glaci-Marine Sedimentation: Observations from Spitsbergen and a Simple Model. The Journal of Geology 97, 221–231.
Show in context

Ice sediment content ranged from <0.1 to 234 g L-1 of meltwater, close to the 0.2–200 g L-1 range previously reported in Svalbard by Dowdeswell and Dowdeswell (1989).
View in article


Enderlin, E.M., Hamilton, G.S., Straneo, F., Sutherland, D.A. (2016) Iceberg meltwater fluxes dominate the freshwater budget in Greenland’s iceberg-congested glacial fjords. Geophysical Research Letters 43, 11287–11294.
Show in context

In two Greenlandic catchments, Ilulissat Fjord and Sermilik Fjord, overall in-fjord iceberg volume losses were estimated to be >50 % (Enderlin et al., 2016), tentatively supporting the 50 % inshore iceberg volume loss used to estimate offshore FeAsc fluxes by Raiswell et al. (2016).
View in article


Hart, T.J. (1934) Discovery Reports. Discovery Reports 8, 1–268.
Show in context

Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990; De Baar et al., 1995; Loscher et al., 1997) and particulate (>0.2 µm) phases (Hart, 1934; Lin et al., 2011; Shaw et al., 2011).
View in article


Hjelle, A. (1993) The geology of Svalbard: Oslo. Polarhåndbok nr. 6. Norsk Polarinstitutt, Oslo.
Show in context

FeAsc (µmol L-1) is correlated with sediment load (Fig. 4), but assessing whether changes in sediment load affect FeAsc (wt. %) is complicated by the lack of any parameter to account for the post-calving age of ice and by the highly variable bedrock composition across Kongsfjorden (see for example Hjelle, 1993).
View in article


Hopwood, M.J., Connelly, D.P., Arendt, K.E., Juul-Pedersen, T., Stinchcombe, M., Meire, L., Esposito, M., Krishna, R. (2016) Seasonal changes in Fe along a glaciated Greenlandic fjord. Frontiers in Earth Sciences 4, doi:10.3389/feart.2016.00015.
Show in context

Observations of DFe concentrations in iceberg meltwater are sparse, but the available data does suggest a heterogeneous distribution, with DFe ranging 4–600 nM in Antarctic (Lin et al., 2011) and 3–300 nM in Greenlandic (Hopwood et al., 2016) iceberg melt.
View in article


Hopwood, M.J., Statham, P.J., Tranter, M., Wadham, J.L. (2014) Glacial flours as a potential source of Fe(II) and Fe(III) to polar waters. Biogeochemistry 118, 443–452.
Show in context

For glacial flour particles of <1 mm, it has previously been demonstrated that the change in FeAsc (wt. %) with particle size is not pronounced (Hopwood et al., 2014; Raiswell et al., 2016), but this may not be the case for larger particles.
View in article


Lin, H., Rauschenberg, S., Hexel, C.R., Shaw, T.J., Twining, B.S. (2011) Free-drifting icebergs as sources of iron to the Weddell Sea. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1392–1406.
Show in context

Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990; De Baar et al., 1995; Loscher et al., 1997) and particulate (>0.2 µm) phases (Hart, 1934; Lin et al., 2011; Shaw et al., 2011).
View in article
Observations of DFe concentrations in iceberg meltwater are sparse, but the available data does suggest a heterogeneous distribution, with DFe ranging 4–600 nM in Antarctic (Lin et al., 2011) and 3–300 nM in Greenlandic (Hopwood et al., 2016) iceberg melt.
View in article
The distribution of particulate Fe (which includes FeAsc) is also expected to be heterogeneous due to the presence of embedded sediment-rich layers that account for only a small fraction of total iceberg volume (Lin et al., 2011; Raiswell, 2011; Raiswell et al., 2016).
View in article


Loscher, B.M., De Baar, H.J.W., De Jong, J.T.M., Veth, C., Dehairs, F. (1997) The distribution of Fe in the Antarctic Circumpolar Current. Deep Sea Research Part II: Topical Studies in Oceanography 44, 143–187.
Show in context

Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990; De Baar et al., 1995; Loscher et al., 1997) and particulate (>0.2 µm) phases (Hart, 1934; Lin et al., 2011; Shaw et al., 2011).
View in article


Markussen, T.N., Elberling, B., Winter, C., Andersen, T.J. (2016) Flocculated meltwater particles control Arctic land-sea fluxes of labile iron. Scientific Reports 6, 24033.
Show in context

As has been demonstrated in this study and elsewhere (e.g., Markussen et al., 2016), surface waters in stratified, glaciated fjords can exhibit extremely high TdFe concentrations due to the presence of glacially derived particle plumes.
View in article


Martin, J.H., Gordon, R.M., Fitzwater, S.E. (1990) Iron in Antarctic waters. Nature 345, 156–158.
Show in context

Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990; De Baar et al., 1995; Loscher et al., 1997) and particulate (>0.2 µm) phases (Hart, 1934; Lin et al., 2011; Shaw et al., 2011).
View in article
Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990, 1991; Moore et al., 2013), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007; Schwarz and Schodlok, 2009; Smith et al., 2011), a change in iceberg Fe supply could significantly affect marine primary productivity.
View in article


Martin, J.H., Gordon, R.M., Fitzwater, S.E. (1991)The case for iron. Limnology and Oceanography 36, 1793–1802.
Show in context

Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990, 1991; Moore et al., 2013), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007; Schwarz and Schodlok, 2009; Smith et al., 2011), a change in iceberg Fe supply could significantly affect marine primary productivity.
View in article


Moore, C.M., Mills, M.M., Arrigo, K.R., Berman-Frank, I., Bopp, L., Boyd, P.W., Galbraith, E.D., Geider, R.J., Guieu, C., Jaccard, S.L., Jickells, T.D., La Roche, J., Lenton, T.M., Mahowald, N.M., Maranon, E., Marinov, I., Moore, J.K., Nakatsuka, T., Oschlies, A., Saito, M.A., Thingstad, T.F., Tsuda, A., Ulloa, O. (2013) Processes and patterns of oceanic nutrient limitation. Nature Geoscience 6, 701–710.
Show in context

Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990, 1991; Moore et al., 2013), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007; Schwarz and Schodlok, 2009; Smith et al., 2011), a change in iceberg Fe supply could significantly affect marine primary productivity.
View in article


Mugford, R.I., Dowdeswell, J.A. (2010) Modeling iceberg-rafted sedimentation in high-latitude fjord environments. Journal of Geophysical Research: Earth Surface 115, doi:10.1029/2009JF001564.
Show in context

In the Arctic, iceberg-borne sediment is known to be lost from icebergs faster than ice volume (Mugford and Dowdeswell, 2010) due to its association with basal ice.
View in article
A model for Kangerdlugssuaq Fjord (Greenland) shows that whilst icebergs lose 20–30 % ice volume within this fjord, the corresponding in-fjord sediment loss is 70–85 % (Mugford and Dowdeswell, 2010).
View in article
In-fjord iceberg volume loss should correspond to a disproportionately high loss of iceberg embedded sediment (Mugford and Dowdeswell, 2010), and thereby also FeAsc and TdFe.
View in article
Generally in the Arctic, a sharp decline in the mean FeAsc and TdFe per volume of meltwater from icebergs with time after calving would be expected due to the preferential loss of iceberg basal ice, as modelled by Mugford and Dowdeswell (2010).
View in article


Paolo, F.S., Fricker, H.A., Padman, L. (2015) Volume loss from Antarctic ice shelves is accelerating. Science 348, 327–331.
Show in context

As the rate of iceberg calving in polar seas oscillates on glacial to inter-glacial timescales (Bond et al., 1992), and recent climate change has increased the discharge of calved ice volume from both the Antarctic (Paolo et al., 2015) and Greenlandic (Bamber et al., 2012) ice sheets, Fe delivery from icebergs may also change.
View in article


Raiswell, R. (2011) Iceberg-hosted nanoparticulate Fe in the Southern Ocean: Mineralogy, origin, dissolution kinetics and source of bioavailable Fe. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1364–1375.
Show in context

The distribution of particulate Fe (which includes FeAsc) is also expected to be heterogeneous due to the presence of embedded sediment-rich layers that account for only a small fraction of total iceberg volume (Lin et al., 2011; Raiswell, 2011; Raiswell et al., 2016).
View in article


Raiswell, R., Benning, L.G., Tranter, M., Tulaczyk, S. (2008) Bioavailable iron in the Southern Ocean: the significance of the iceberg conveyor belt. Geochemical Transactions 9, doi:10.1186/1467-4866-9-7.
Show in context

Icebergs should thus constitute a source of the micronutrient Fe to offshore polar waters (Raiswell et al., 2008).
View in article


Raiswell, R., Vu, H.P., Brinza, L., Benning, L.G. (2010) The determination of labile Fe in ferrihydrite by ascorbic acid extraction: Methodology, dissolution kinetics and loss of solubility with age and de-watering. Chemical Geology 278, 70–79.
Show in context

FeAsc leaches were conducted on wet sediment as per Raiswell et al. (2010), with leached Fe determined by measuring absorbance (λ = 562 nm) before and after the addition of ferrozine (as detailed in Supplementary Information Methods).
View in article
Whilst dried sediment exhibits a rapid decline in FeAsc wt. % (Raiswell et al., 2010), it is not clear how storage of ice or wet sediments affects FeAsc.
View in article
‘FeAsc’ is the ferrihydrite content of sediment, defined by Raiswell et al. (2010).
View in article


Raiswell, R., Hawkings, J.R., Benning, L.G., Baker, A.R., Death, R., Albani, S., Mahowald, N., Krom, M.D., Poulton, S.W., Wadham, J., Tranter, M. (2016) Potentially bioavailable iron delivery by iceberg-hosted sediments and atmospheric dust to the polar oceans. Biogeosciences 13, 3887–3900.
Show in context

The distribution of particulate Fe (which includes FeAsc) is also expected to be heterogeneous due to the presence of embedded sediment-rich layers that account for only a small fraction of total iceberg volume (Lin et al., 2011; Raiswell, 2011; Raiswell et al., 2016).
View in article
Whilst TdFe data for icebergs is sparse, iceberg FeAsc content has previously been estimated in multiple catchments worldwide (Raiswell et al., 2016) producing a mean global content of 2.7–17 µM.
View in article
However, FeAsc content and offshore iceberg FeAsc fluxes are normally calculated using a mean sediment loading (0.5 g L-1 is widely used as outlined in Raiswell et al. (2016)) with considerable uncertainty generally acknowledged in this value.
View in article
In two Greenlandic catchments, Ilulissat Fjord and Sermilik Fjord, overall in-fjord iceberg volume losses were estimated to be >50 % (Enderlin et al., 2016), tentatively supporting the 50 % inshore iceberg volume loss used to estimate offshore FeAsc fluxes by Raiswell et al. (2016).
View in article
Sediment loss should also affect mean FeAsc content in the same way, however FeAsc losses may be offset from TdFe losses if significant processing of surface sediment occurs on the timescale of iceberg Fe delivery (Raiswell et al., 2016).
View in article
In Kongsfjorden, Raiswell et al. (2016) reported a FeAsc range of 0.016–0.37 wt. % (n = 14), with a mean of 0.14 wt. % and median of 0.092 wt. %; equivalent to 1.4–33 µM, 12 µM and 8.2 µM, respectively when using the suggested mean sediment loading of 0.5 g L-1.
View in article
Table 1 [...] *The suggested 0.5 g L-1 sediment loading is used for data from Raiswell et al. (2016).
View in article
For glacial flour particles of <1 mm, it has previously been demonstrated that the change in FeAsc (wt. %) with particle size is not pronounced (Hopwood et al., 2014; Raiswell et al., 2016), but this may not be the case for larger particles.
View in article


Schwarz, J.N., Schodlok, M.P. (2009) Impact of drifting icebergs on surface phytoplankton biomass in the Southern Ocean: Ocean colour remote sensing and in situ iceberg tracking. Deep Sea Research Part I: Oceanographic Research Papers 56, 1727–1741.
Show in context

Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990, 1991; Moore et al., 2013), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007; Schwarz and Schodlok, 2009; Smith et al., 2011), a change in iceberg Fe supply could significantly affect marine primary productivity.
View in article


Shaw, T.J., Raiswell, R., Hexel, C.R., Vu, H.P., Moore, W.S., Dudgeon, R., Smith Jr., K.L. (2011) Input, composition, and potential impact of terrigenous material from free-drifting icebergs in the Weddell Sea. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1376–1383.
Show in context

Icebergs contain higher Fe concentrations than seawater, both in the dissolved (<0.2 µm) (Martin et al., 1990; De Baar et al., 1995; Loscher et al., 1997) and particulate (>0.2 µm) phases (Hart, 1934; Lin et al., 2011; Shaw et al., 2011).
View in article


Smith, K.L., Sherman, A.D., Shaw, T.J., Murray, A.E., Vernet, M., Cefarelli, A.O. (2011) Carbon export associated with free-drifting icebergs in the Southern Ocean. Deep Sea Research Part II: Topical Studies in Oceanography 58, 1485–1496.
Show in context

Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990, 1991; Moore et al., 2013), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007; Schwarz and Schodlok, 2009; Smith et al., 2011), a change in iceberg Fe supply could significantly affect marine primary productivity.
View in article


Smith Jr., K.L., Robison, B.H., Helly, J.J., Kaufmann, R.S., Ruhl, H.A., Shaw, T.J., Twining, B.S., Vernet, M. (2007) Free-drifting icebergs: Hot spots of chemical and biological enrichment in the Weddell Sea. Science 317, 478–482.
Show in context

Particularly in the Southern Ocean, where DFe deficiency extensively limits primary production (Martin et al., 1990, 1991; Moore et al., 2013), and icebergs cause chemical and biological enrichment of surrounding waters (Smith Jr. et al., 2007; Schwarz and Schodlok, 2009; Smith et al., 2011), a change in iceberg Fe supply could significantly affect marine primary productivity.
View in article



top

Supplementary Information


Supplementary Methods


Ice sample collection
All ice samples were collected from calved ice masses within Kongsfjorden east of 11.89° E. Sampled icebergs had a maximum length (visible above the waterline) of between 0.4 and 20 m, and a maximum height above the waterline of between 10 cm and 4 m. For FeAsc, icebergs with visible sediment layers (darkened layers approximately 0.5 to 5 cm thick) were targeted whereas sample collection for DFe/TdFe was random.

FeAsc
To measure FeAsc, 1–2 kg ice pieces were returned to the laboratory in insulated plastic boxes. As a precaution against contamination, ice was rinsed with de-ionised water (Milli-Q, Millipore, conductivity 18.2 MΩ cm-1). Ice was then melted in low density polyethylene (LDPE) bags with the first (1–2 hr later) meltwater discarded. After melting ice overnight, the total volume was recorded and then excess water was removed slowly (<5 mL min-1) by filtration (0.45 µm, polyvinylidene fluoride, Millipore) to concentrate the sediment. All plastic vials and filtration equipment for sediment collection/handling was pre-cleaned (1 M HCl, 3 de-ionised water rinses) prior to use. FeAsc was determined as per Raiswell et al. (2010) but using larger (~100 mg) sediment samples without sieving. Leached Fe was determined by measuring absorbance (λ = 562 nm) before, and after, the addition of ferrozine (Stookey, 1970) using a 1 cm cell with a USB4000 Fiber-optic Spectrometer and a LS-1 tungsten halogen light source (Ocean Optics). Fe standards were made by spiking DFe from an acidified (pH 2) 1 mM Fe stock solution into aliquots of ascorbic leaching solution producing a linear response (R2 > 0.99, derived molar adsorption coefficient 24,500 M-1 cm-1) over the absorbance range 0–0.6. Sample absorbance was then corrected for background absorbance and dilution by reagents.

A sub-sample of sediment was retained (except for sediment loadings <0.1 g L-1 where this was not possible) and air dried to constant mass in order to calculate FeAsc content g-1 (of sediment) and L-1 (of meltwater). FeAsc measurements were duplicated for 11 randomly selected samples producing a relative standard deviation of 4.5 %. The propagated standard deviation for FeAsc L-1 of meltwater (obtained from combining sediment load with FeAsc wt. %) was approximately 15 %.

DFe and TdFe
125 mL LDPE bottles (Nalgene) were pre-cleaned in a 3 stage process with 3 de-ionised water rinses after each stage (detergent, 1.2 M HCl, 1.2 M HNO3). After melting ice in LDPE bags, as above, 125 mL meltwater was retained without filtration for TdFe, and 125 mL was syringe filtered for DFe analysis (0.20 µm, polyvinylidene fluoride, Millipore). TdFe and DFe samples were acidified (with HCl, UPA, ROMIL) to pH < 2.0 and subsequently stored for 12 months to ensure complete recovery of soluble Fe (Edwards and Sedwick, 2001). DFe and TdFe were analysed by ICP-MS (ELEMENT XR, ThermoFisherScientific) after dilution with 1 M HNO3 (distilled in house using a DST-1000, Savillex, from SPA grade HNO3, ROMIL) and calibrated by standard addition with a linear peak response from 0–1000 nM Fe (R2 > 0.99). The analytical blank (DFe and TdFe) was always <0.6 nM Fe. To verify that the dilution technique yielded reproducible and accurate DFe results, NASS-7 and CASS-6 Certified Reference Materials (CRMs, National Research Council Canada) were analysed for Fe. Both CRMs yielded reproducible Fe concentrations within the certified ranges (Table S-1).

Table S-1 Analysis of Certified Reference Materials for Fe concentration (± standard deviation of at least 6 measurements).
Certified Reference MaterialFe (± SD) / nMCertified Fe concentration / nM
NASS-76.21 (± 0.77)6.29 (± 0.47)
CASS-626.6 (± 0.71)27.9 (± 2.1)
Download in Excel

A linear plot of all paired DFe and TdFe minus DFe (n = 28) yielded R2 = 0.78. However, R2 declined to 0.01 when the 3 highest TdFe data were removed. A Spearman Rank Order Correlation suggested there was no significant relationship between DFe and TdFe minus DFe (P value 0.21). A linear plot of all FeAsc (mol L-1) data and corresponding ice sediment load (g L-1) yielded R2 = 0.88. A Spearman Rank Order Correlation Coefficient of 0.95 (P value 2 × 10-7, n = 58) demonstrates that the relationship is significant and remains significant if the 3 highest sediment loadings are removed (Coefficient 0.94, P value 2 × 10-7, n = 55) - regardless of whether the sediment loads reported as <0.1 g L-1 are treated as 0.1 g L-1, 0.01 g L-1, or excluded. Statistics were performed in SigmaPlot 13.

Water column
In July 2015, 15 samples of fjord surface water (depth <0.2 m) were collected by hand upstream of a small boat (locations shown Fig. 1). As per ice samples for DFe and TdFe, 125 mL was retained without filtration and 125 mL was syringe filtered for DFe analysis (0.20 µm, polyvinylidene fluoride, Millipore). Analysis was conducted via ICP-MS after storage for 12 months at pH < 2, and dilution with 1 M HNO3 (as per melted ice, above). The temperature and salinity of surface fjord water were recorded using a LF 325 conductivity meter (WTW) which was calibrated before use with a KCl solution.

In July–August 2016 conductivity, temperature, depth profiles (CTDs) were acquired at 20 stations in the area of the fjord where ice was observed and collected (east of 11.89° E and proceeding to within approximately 400 m of each marine terminating glacier in Kongsfjorden, Fig. 1). Seawater temperature (recorded at 1 m depth) was consistently 4.0–5.0 °C. A broader range of surface temperatures were observed in July 2015 (2.6–8.8 °C at <0.2 m), but are not directly comparable because of the depth difference.

Supplementary Material


Table S-2 FeAsc concentration reported for various sediment samples collected around Kongsfjorden. For iceberg embedded sediment with sediment loadings >0.1 g L-1, FeAsc is reported both as wt. % and per volume of meltwater.
Sample numberOriginFeAsc wt. %*FeAsc µmol / L*Sediment loading g / L
1Embedded sediment , glacier crevasse (Midtre Lovénbreen)0.014

2Embedded sediment , glacier crevasse (Midtre Lovénbreen)0.0057

3Embedded sediment , glacier crevasse (Midtre Lovénbreen)0.17

4Embedded sediment , glacier crevasse (Midtre Lovénbreen)0.023

5Embedded sediment , glacier crevasse (Midtre Lovénbreen)0.0038

6Embedded sediment , glacier crevasse (Midtre Lovénbreen)0.053

7Glacier surface sediment (Kongsvegen)0.016

8Glacier surface sediment (Kongsvegen)0.016

9Glacier surface sediment (Kongsvegen)0.012

10Glacier surface sediment (Kongsvegen)0.010

11Glacier surface sediment (Kongsvegen)0.016

12Glacier surface sediment (Kongsvegen)0.012

13Glacier surface sediment (Kongsvegen)0.014

14Iceberg embedded sediment0.00911069
15Iceberg embedded sediment0.00933420
16Iceberg embedded sediment0.00769066
17Iceberg embedded sediment0.00745239
18Iceberg embedded sediment0.00851711
19Iceberg embedded sediment0.0162.50.91
20Iceberg embedded sediment0.00891.20.76
21Iceberg embedded sediment0.0853.90.25
22Iceberg embedded sediment0.0149.63.9
23Iceberg embedded sediment0.020820230
24Iceberg embedded sediment0.0343.90.65
25Iceberg embedded sediment0.0112613
26Iceberg embedded sediment0.0261000230
27Iceberg embedded sediment0.0239623
28Iceberg embedded sediment0.00423850
29Iceberg embedded sediment0.02216041
30Iceberg embedded sediment0.01611038
31Iceberg embedded sediment0.00823121
32Iceberg embedded sediment0.0245212
33Iceberg embedded sediment0.009511065
34Iceberg embedded sediment0.00739170
35Iceberg embedded sediment0.0155.01.8
36Iceberg embedded sediment0.013177.5
37Iceberg embedded sediment0.011340170
38Iceberg embedded sediment0.0277.81.6
39Iceberg embedded sediment0.0261.20.25
40Iceberg embedded sediment0.025112.4
41Iceberg embedded sediment0.123.00.14
42Iceberg embedded sediment0.0282.50.51
43Iceberg embedded sediment0.013198.1
44Iceberg embedded sediment0.0118039
45Iceberg embedded sediment0.0203.60.99
46Iceberg embedded sediment0.0261.20.26
47Iceberg embedded sediment0.0122.11.0
48Iceberg embedded sedimentn/d0.26< 0.1
49Iceberg embedded sedimentn/d0.26< 0.1
50Iceberg embedded sedimentn/d0.87< 0.1
51Iceberg embedded sedimentn/d0.65< 0.1
52Iceberg embedded sedimentn/d0.27< 0.1
53Iceberg embedded sedimentn/d0.50< 0.1
54Iceberg embedded sedimentn/d0.28< 0.1
55Iceberg embedded sedimentn/d0.63< 0.1
56Iceberg embedded sedimentn/d0.40< 0.1
57Iceberg embedded sedimentn/d0.69< 0.1
58Iceberg embedded sedimentn/d0.37< 0.1
59Iceberg embedded sedimentn/d0.43< 0.1
60Iceberg embedded sedimentn/d0.68< 0.1
61Iceberg embedded sedimentn/d0.87< 0.1
62Iceberg embedded sedimentn/d0.27< 0.1
63Iceberg embedded sedimentn/d0.39< 0.1
64Iceberg embedded sedimentn/d1.8< 0.1
65Iceberg embedded sedimentn/d0.40< 0.1
66Iceberg embedded sedimentn/d0.77< 0.1
67Iceberg embedded sedimentn/d0.35< 0.1
68Iceberg embedded sedimentn/d0.29< 0.1
69Iceberg embedded sedimentn/d0.32< 0.1
70Iceberg embedded sedimentn/d0.20< 0.1
71Iceberg embedded sedimentn/d0.15< 0.1
72Iceberg surface sediment0.0065

73Iceberg surface sediment0.0068

74Iceberg surface sediment0.0091

75Iceberg surface sediment0.0072

76Iceberg surface sediment0.024

77Iceberg surface sediment0.0071

78Iceberg surface sediment0.032

79Iceberg surface sediment0.010

80Iceberg surface sediment0.012

81Iceberg surface sediment0.021

82Iceberg surface sediment0.0093

83Iceberg surface sediment0.0077

84Iceberg surface sediment0.059

85Iceberg surface sediment0.0066

86Iceberg surface sediment0.0080

87Iceberg surface sediment0.0080

88Iceberg surface sediment0.029

89Iceberg surface sediment0.032

90Iceberg surface sediment0.0087

91Iceberg surface sediment0.0098

92Pro-glacial stream, glacial flour0.0090

93Pro-glacial stream, glacial flour0.0083

94Pro-glacial stream, glacial flour0.0063

95Pro-glacial stream, glacial flour0.0079

96Pro-glacial stream, glacial flour0.0072

97Pro-glacial stream, glacial flour0.0070

98Pro-glacial stream, glacial flour0.0073

99Pro-glacial stream, glacial flour0.019

100Pro-glacial stream, glacial flour0.031

101Pro-glacial stream, glacial flour0.010

102Pro-glacial stream, glacial flour0.0078

103Pro-glacial stream, glacial flour0.010

104Pro-glacial stream, glacial flour0.0066

105Pro-glacial stream, glacial flour0.0089

106Pro-glacial stream, glacial flour0.0096

107Pro-glacial stream, glacial flour0.0041

108Pro-glacial stream, glacial flour0.0032

109Pro-glacial stream, glacial flour0.0057

110Pro-glacial stream, glacial flour0.0079

111Surface glacier (Kongsvegen) embedded sediment0.044

112Surface glacier (Kongsvegen) embedded sediment0.071

113Surface glacier (Kongsvegen) embedded sediment0.037

114Surface glacier (Kongsvegen) embedded sediment0.011

115Surface glacier (Kongsvegen) embedded sediment0.061

116Surface glacier (Kongsvegen) embedded sediment0.11

* Iceberg embedded sediment only
n/d not determined as dry mass insufficient (<0.1 g)

Download in Excel

Table S-3 Dissolved (<0.2 µm) Fe and total dissolvable Fe for 15 surface water samples collected in Kongsfjorden (July 2015).
NameDFe / nMTdFe / nMTemperature / °CLatitude °NLongitude °ESalinity
F128.572268.478.93811.99220.0
F29.515487.078.93712.00332.1
F319.618066.778.94211.99031.7
F46.6124448.278.93712.01814.2
F519.711757.578.93111.95731.7
F647.154638.878.92811.92622.6
F715.5524952.678.88012.52826.9
F810.344923.778.89312.51828.1
F94.634254.178.90312.49328.6
F104.551264.978.91112.51823.1
F1160.1155584.878.91712.57023.1
F126.834084.878.91712.47626.7
F1327.836915.078.91912.41927.6
F146.926004.678.92412.33029.8
F155.616055.478.92912.22329.7
Download in Excel

Table S-4 Dissolved (<0.2 µm) Fe and total dissolvable Fe for 28 randomly collected Kongsfjorden iceberg samples (July 2015).
Sample labelOriginCatchmentDFe / nMTdFe / nMPFe (TdFe-DFe) / nM
Ice 1Ny Alesund, Svalbard, Summer 2015Kongsfjorden19.2261242
Ice 2Ny Alesund, Svalbard, Summer 2015Kongsfjorden9.59787
Ice 3Ny Alesund, Svalbard, Summer 2015Kongsfjorden1.3369367
Ice 4Ny Alesund, Svalbard, Summer 2015Kongsfjorden7.910131005
Ice 5Ny Alesund, Svalbard, Summer 2015Kongsfjorden4.9331326
Ice 6Ny Alesund, Svalbard, Summer 2015Kongsfjorden0.6133133
Ice 7Ny Alesund, Svalbard, Summer 2015Kongsfjorden3.2377374
Ice 8Ny Alesund, Svalbard, Summer 2015Kongsfjorden2.0109107
Ice 9Ny Alesund, Svalbard, Summer 2015Kongsfjorden2.15452
Ice 10Ny Alesund, Svalbard, Summer 2015Kongsfjorden6.4897890
Ice 11Ny Alesund, Svalbard, Summer 2015Kongsfjorden16.522742257
Ice 12Ny Alesund, Svalbard, Summer 2015Kongsfjorden6.6470464
Ice 13Ny Alesund, Svalbard, Summer 2015Kongsfjorden9.72325923250
Ice 14Ny Alesund, Svalbard, Summer 2015Kongsfjorden8.0221213
Ice 15Ny Alesund, Svalbard, Summer 2015Kongsfjorden3.74945
Ice 16Ny Alesund, Svalbard, Summer 2015Kongsfjorden3.95955
Ice 17Ny Alesund, Svalbard, Summer 2015Kongsfjorden11.16251
Ice 18Ny Alesund, Svalbard, Summer 2015Kongsfjorden9.54637
Ice 19Ny Alesund, Svalbard, Summer 2015Kongsfjorden12.46553
Ice 20Ny Alesund, Svalbard, Summer 2015Kongsfjorden5.47973
Ice 21Ny Alesund, Svalbard, Summer 2015Kongsfjorden536.35740156864
Ice 22Ny Alesund, Svalbard, Summer 2015Kongsfjorden54.77722
Ice 23Ny Alesund, Svalbard, Summer 2015Kongsfjorden4.1156151
Ice 24Ny Alesund, Svalbard, Summer 2015Kongsfjorden1.49997
Ice 25Ny Alesund, Svalbard, Summer 2015Kongsfjorden1.4107106
Ice 26Ny Alesund, Svalbard, Summer 2015Kongsfjorden3.48884
Ice 27Ny Alesund, Svalbard, Summer 2015Kongsfjorden16.5830814
Ice 28Ny Alesund, Svalbard, Summer 2015Kongsfjorden266.11087610610
Download in Excel

Supplementary Information References


Edwards, R., Sedwick, P. (2001) Iron in East Antarctic snow: Implications for atmospheric iron deposition and algal production in Antarctic waters. Geophysical Research Letters 28, 3907–3910.


Raiswell, R., Vu, H.P., Brinza, L., Benning, L.G. (2010) The determination of labile Fe in ferrihydrite by ascorbic acid extraction: Methodology, dissolution kinetics and loss of solubility with age and de-watering. Chemical Geology 278, 70–79.


Stookey, L.L. (1970) Ferrozine- a new spectrophotometric reagent for iron. Analytical Chemistry 42, 779–781.
top