Review article: How does glacier discharge affect marine biogeochemistry … · 2020-06-22 · Fjords act as highly stratiﬁed estuaries and provide a path-way for the exchange of

The Cryosphere 14 1347ndash1383 2020httpsdoiorg105194tc-14-1347-2020copy Author(s) 2020 This work is distributed underthe Creative Commons Attribution 40 License

Review article How does glacier discharge affect marinebiogeochemistry and primary production in the ArcticMark J Hopwood1 Dustin Carroll2 Thorben Dunse34 Andy Hodson35 Johnna M Holding6 Joseacute L Iriarte7Sofia Ribeiro8 Eric P Achterberg1 Carolina Cantoni9 Daniel F Carlson14 Melissa Chierici510 Jennifer S Clarke1Stefano Cozzi9 Agneta Fransson11 Thomas Juul-Pedersen12 Mie H S Winding12 and Lorenz Meire1213

1GEOMAR Helmholtz Centre for Ocean Research Kiel Kiel Germany2Moss Landing Marine Laboratories San Joseacute State University Moss Landing CA3Western Norway University of Applied Sciences Sogndal Norway4The University of Oslo Oslo Norway5The University Centre in Svalbard Longyearbyen Svalbard6Department of Bioscience Aarhus University Silkeborg Denmark7Instituto de Acuicultura and Centro Dinaacutemica de Ecosistemas Marinos de Altas Latitudes ndash IDEALUniversidad Austral de Chile Puerto Montt Chile8Geological Survey of Denmark and Greenland Copenhagen Denmark9CNR-ISMAR Istituto di Scienze Marine Trieste Italy10Institute of Marine Research Fram Centre Tromsoslash Norway11Norwegian Polar Institute Fram Centre Tromsoslash Norway12Greenland Climate Research Centre Greenland Institute of Natural Resources Nuuk Greenland13Royal Netherlands Institute for Sea Research and Utrecht University Yerseke the Netherlands14Institute of Coastal Research Helmholtz-Zentrum Geesthacht Centre for Materials and Coastal ResearchGeesthacht Germany

Correspondence Mark J Hopwood (mhopwoodgeomarde)

Received 3 June 2019 ndash Discussion started 18 June 2019Revised 27 January 2020 ndash Accepted 27 February 2020 ndash Published 24 April 2020

Abstract Freshwater discharge from glaciers is increas-ing across the Arctic in response to anthropogenic climatechange which raises questions about the potential down-stream effects in the marine environment Whilst a combi-nation of long-term monitoring programmes and intensiveArctic field campaigns have improved our knowledge ofglacierndashocean interactions in recent years especially with re-spect to fjordocean circulation there are extensive knowl-edge gaps concerning how glaciers affect marine biogeo-chemistry and productivity Following two cross-cutting dis-ciplinary International Arctic Science Committee (IASC)workshops addressing the importance of glaciers for the ma-rine ecosystem here we review the state of the art con-cerning how freshwater discharge affects the marine en-vironment with a specific focus on marine biogeochem-istry and biological productivity Using a series of Arc-

tic case studies (Nuup KangerluaGodtharingbsfjord Kongsfjor-den Kangerluarsuup SermiaBowdoin Fjord Young Soundand Sermilik Fjord) the interconnected effects of freshwa-ter discharge on fjordndashshelf exchange nutrient availabilitythe carbonate system the carbon cycle and the microbialfood web are investigated Key findings are that whetherthe effect of glacier discharge on marine primary produc-tion is positive or negative is highly dependent on a com-bination of factors These include glacier type (marine-or land-terminating) fjordndashglacier geometry and the lim-iting resource(s) for phytoplankton growth in a specificspatio-temporal region (light macronutrients or micronu-trients) Arctic glacier fjords therefore often exhibit dis-tinct dischargendashproductivity relationships and multiple case-studies must be considered in order to understand the net ef-fects of glacier discharge on Arctic marine ecosystems

Published by Copernicus Publications on behalf of the European Geosciences Union

1348 M J Hopwood et al Effects of glaciers in the Arctic

1 Introduction

Annual freshwater discharge volume from glaciers has in-creased globally in recent decades (Rignot et al 2013 Bam-ber et al 2018 Mouginot et al 2019) and will continue todo so across most Arctic regions until at least the middleof this century under a Representative Concentration Path-way (RCP) 45 climate scenario (Bliss et al 2014 Huss andHock 2018) This increase in discharge (surface runoff andsubsurface discharge into the ocean) raises questions aboutthe downstream effects in marine ecosystems particularlywith respect to ecosystem services such as carbon seques-tration and fisheries (Meire et al 2015 2017 Milner et al2017) In order to understand the effect of glaciers on thepresent-day marine environment and under future climatescenarios knowledge of the physical and chemical perturba-tions occurring in the water column as a result of glacier dis-charge and the structure function and resilience of ecosys-tems within these regions must be synthesized

Quantifying the magnitude of environmental perturbationsfrom glacial discharge is complicated by the multiple concur-rent and occasionally counteracting effects that glacial dis-charge has in the marine environment For example ice-rockabrasion means that glacially fed rivers can carry higher sed-iment loads than temperate rivers (Chu et al 2009 Overeemet al 2017) Extensive sediment plumes where glacier dis-charge first enters the ocean limit light penetration into thewater column (Murray et al 2015 Halbach et al 2019)and ingestion of glacial flour particles can be hazardous oreven fatal to zooplankton krill and benthic fauna (Whiteand Dagg 1989 Włodarska-Kowalczuk and Pearson 2004Arendt et al 2011 Fuentes et al 2016) However theseplumes also provide elevated concentrations of inorganiccomponents such as calcium carbonate which affects sea-water alkalinity (Yde et al 2014 Fransson et al 2015)and dissolved silicic acid (hereafter Si) (Brown et al 2010Meire et al 2016a) and iron (Fe) (Statham et al 2008 Lip-piatt et al 2010) which can potentially increase marine pri-mary production (Gerringa et al 2012 Meire et al 2016a)

The impacts of glacier discharge can also depend upon thespatial and temporal scales investigated (van de Poll et al2018) In semi-enclosed Arctic coastal regions and fjord sys-tems summertime discharge typically produces strong near-surface stratification This results in a shallow nutrient-poorlayer which reduces primary production and drives phyto-plankton biomass deeper in the water column (Rysgaard etal 1999 Juul-Pedersen et al 2015 Meire et al 2017)On broader scales across continental shelves freshening cansimilarly reduce vertical nutrient supply throughout summer(Coupel et al 2015) but may also impede the breakdown ofstratification in autumn thereby extending the phytoplank-ton growing season (Oliver et al 2018) Key research ques-tions are how and on what spatial and temporal timescalesthese different effects interact to enhance or reduce marineprimary production Using a synthesis of field studies from

glacier catchments with different characteristics (Fig 1) weprovide answers to three questions arising from two interdis-ciplinary workshops on the importance of Arctic glaciers forthe marine ecosystem under the umbrella of the InternationalArctic Science Committee (IASC)

1 Where and when does glacial freshwater discharge pro-mote or reduce marine primary production

2 How does spatio-temporal variability in glacial dis-charge affect marine primary production

3 How far-reaching are the effects of glacial discharge onmarine biogeochemistry

2 Fjords as critical zones for glacierndashocean interactions

In the Arctic and sub-Antarctic most glacial discharge entersthe ocean through fjord systems (Iriarte et al 2014 Straneoand Cenedese 2015) The strong lateral gradients and sea-sonal changes in environmental conditions associated withglacial discharge in these coastal environments differentiatethese ecosystems from offshore systems (Arendt et al 2013Lydersen et al 2014 Krawczyk et al 2018) Fjords can beefficient sinks for organic carbon (Smith et al 2015) andCO2 (Rysgaard et al 2012 Fransson et al 2015) sustainlocally important fisheries (Meire et al 2017) and are criticalzones for deep mixing which dictate how glacially modifiedwaters are exchanged with the coastal ocean (Mortensen etal 2014 Straneo and Cenedese 2015 Beaird et al 2018)Fjord-scale processes therefore comprise an integral part ofall questions concerning how glacial discharge affects Arcticcoastal primary production (Arimitsu et al 2012 Renner etal 2012 Meire et al 2017)

Fjords act as highly stratified estuaries and provide a path-way for the exchange of heat salt and nutrients betweennear-glacier waters and adjacent coastal regions (Mortensenet al 2014 2018 Straneo and Cenedese 2015) In deepfjords such as those around much of the periphery of Green-land warm saline water is typically found at depth (gt200 m) overlaid by cold fresher water and during summera thin layer (sim 50 m or less) of relatively warm near-surfacewater (Straneo et al 2012) The injection of freshwater intofjords from subglacial discharge (Xu et al 2012 Carroll etal 2015) and terminus (Slater et al 2018) and iceberg melt(Moon et al 2018) can drive substantial buoyancy-drivenflows in the fjord (Carroll et al 2015 2017 Jackson et al2017) which amplify exchange with the shelf system as wellas submarine melting and the calving rates of glacier terminiTo date such modifications to circulation and exchange be-tween glacier fjords and shelf waters have primarily beenstudied in terms of their effects on ocean physics and melt-ing at glacier termini yet they also have profound impacts onmarine productivity (Meire et al 2016a Kanna et al 2018Torsvik et al 2019)

The Cryosphere 14 1347ndash1383 2020 wwwthe-cryospherenet1413472020

M J Hopwood et al Effects of glaciers in the Arctic 1349

Figure 1 Locations of five key Arctic field sites where extensive work bridging the glacier and marine domains has been conducteddiscussed herein in order to advance understanding of glacierndashocean interactions 1 Kongsfjorden (Svalbard) 2 Young Sound (E Greenland)3 Sermilik (SE Greenland) 4 Nuup KangerluaGodtharingbsfjord (SW Greenland) 5 Bowdoin FjordKangerluarsuup Sermia (NW Greenland)

While renewal of fjord waters from buoyancy-drivenprocesses is mainly thought to occur over seasonal tosub-annual timescales (Gladish et al 2014 Mortensenet al 2014 Carroll et al 2017) energetic shelf forcing(ie from coastalkatabatic winds and coastally trappedwaves) can result in rapid exchange over synoptic timescales(Straneo et al 2010 Jackson et al 2014 Moffat 2014)and similarly also affect marine productivity (Meire et al2016b) Katabatic winds are common features of glaciatedfjords Down-fjord wind events facilitate the removal oflow-salinity surface waters and ice from glacier fjords aswell as the inflow of warmer saline waters at depth (Johnsonet al 2011) The frequency direction and intensity of windevents throughout the year thus adds further complexity tothe effect that fjord geometry has on fjordndashshelf exchangeprocesses (Cushman-Roisin et al 1994 Spall et al 2017)Topographic features such as sills and lateral constrictionscan exert a strong control on fjordndashshelf exchange (Gladishet al 2014 Carroll et al 2017 2018) Ultimately circu-lation can thereby vary considerably depending on fjordgeometry and the relative contributions from buoyancy windand shelf forcing (Straneo and Cenedese 2015 Jacksonet al 2018) Some variability in the spatial patterns ofprimary production is therefore expected between Arcticglacier fjord systems as differences in geometry and forcingaffect exchange with the shelf and water column structure

These changes affect the availability of the resources whichconstrain local primary production (Meire et al 2016bArimitsu et al 2016 Calleja et al 2017)

Fjordndashshelf processes also contribute to the exchange ofactive cells and microbial speciesrsquo resting stages thus pre-conditioning primary production prior to the onset of thegrowth season (Krawczyk et al 2015 2018) Protists (uni-cellular eukaryotes) are the main marine primary produc-ers in the Arctic This highly specialized and diverse groupincludes species that are ice-associated (sympagic) andorpelagic Many protists in fjords and coastal areas of the Arc-tic maintain diverse seed banks of resting stages which pro-motes the resilience and adaptability of species on timescalesfrom seasons to decades (Ellegaard and Ribeiro 2018) Yetseawater inflow into fjords can still change the dominantspecies within a single season In Nuup Kangerlua (Godtharingb-

wwwthe-cryospherenet1413472020 The Cryosphere 14 1347ndash1383 2020

Figure 2 Primary production for Arctic glacier fjord systems in-cluding Disko Bay (Andersen 1977 Nielsen and Hansen 1995Jensen et al 1999 Nielsen 1999 Levinsen and Nielsen 2002)Godtharingbsfjord (Juul-Pedersen et al 2015 Meire et al 2017)Kangerlussuaq (Lund-Hansen et al 2018) Kongsfjorden (Hop etal 2002 Iversen and Seuthe 2011 Hodal et al 2012 van de Pollet al 2018) NordvestfjordScoresby Sund (Seifert et al 2019)Hornsund (Smoła et al 2017) Young Sound (Rysgaard et al1999 Meire et al 2017 Holding et al 2019) the Canadian Arc-tic Archipelago (Harrison et al 1982) and Glacier Bay (Reis-dorph and Mathis 2015) Circles represent glacier fjords trian-gles are sites beyond glacier fjords and bold markers are lt 80 kmfrom a marine-terminating glacier Error bars are standard devia-tions for stations where multiple measurements were made at thesame station Dashed line is the pan-Arctic mean primary produc-tion (MarchndashSeptember) Shaded area is the pan-Arctic shelf rangeof primary production for MayndashAugust (Pabi et al 2008)

sfjord) the spring phytoplankton bloom is typically domi-nated by Fragilariopsis spp diatoms and Phaeocystis spphaptophytes Unusually prolonged coastal seawater inflowin spring 2009 led to the mass occurrence of chain-formingThalassiosira spp diatoms and the complete absence of thenormally abundant Phaeocystis spp (Krawczyk et al 2015)ndash a pattern which has been found elsewhere in the Arcticincluding Kongsfjorden (Hegseth and Tverberg 2013)

3 Pelagic primary production in Arctic glacier fjords

Key factors controlling rates of primary production acrossArctic marine environments are light availability nutrientavailability and grazing (Nielsen 1999 Taylor et al 2013Arrigo and van Dijken 2015 Tremblay et al 2015) Sea-sonal changes in the availability of bioessential resources thestructure of the water column and the feeding patterns of zoo-plankton thereby interact to produce distinct bloom periodsof high primary production shouldered by periods of low pri-mary production In glacier fjords strong lateral and verticalgradients in some or all of these factors create a far more dy-namic situation for primary producers than in the open ocean(Etherington and Hooge 2007 Arendt et al 2010 Murrayet al 2015)

Large inter- and intra-fjord differences in primary produc-tion are demonstrated by field observations around the Arcticwhich show that glacier fjords range considerably in produc-tivity from very low (lt 40 mg C mminus2 dminus1) to moderately pro-ductive systems (gt 500 mg C mminus2 dminus1) during the meltwaterseason (eg Jensen et al 1999 Rysgaard et al 1999 Hop etal 2002 Meire et al 2017) For comparison the pan-Arcticbasin exhibits a mean production of 420plusmn 26 mg C mminus2 dminus1

(mean MarchndashSeptember 1998ndash2006) (Pabi et al 2008)which has increased across most regions in recent decadesdue to reduced summertime sea-ice coverage (Arrigo and vanDijken 2015) and summertime (MayndashAugust) Arctic shelfenvironments exhibit a range of 360ndash1500 mg C mminus2 dminus1

(Pabi et al 2008) So is it possible to generalize how pro-ductive Arctic glacier fjords are

Extensive measurements of primary production through-out the growth season in glacier fjords are only availablefor Godtharingbsfjord (Juul-Pedersen et al 2015 Meire et al2017) Young Sound (Rysgaard et al 1999 Meire et al2017 Holding et al 2019) Glacier Bay (Alaska Reisdorphand Mathis 2015) Hornsund (Svalbard Smoła et al 2017)and Kongsfjorden (Iversen and Seuthe 2011 van de Poll etal 2018) Observations elsewhere are sparse and typicallylimited to summertime-only data Generalizing across mul-tiple Arctic glacier fjord systems therefore becomes chal-lenging due to the paucity of data and the different ge-ographic and seasonal context of individual primary pro-duction data points (Fig 2) Furthermore there are poten-tially some methodological implications when comparingdirect measurements of primary production using 14C up-take (eg Holding et al 2019) with estimates derived fromchanges in water column macronutrient (eg Seifert et al2019) or dissolved inorganic carbon (eg Reisdorph andMathis 2015) inventories

Nevertheless some quantitative comparison can be madeif we confine discussion to months where a meltwater signalmay be evident in most glaciated regions (JulyndashSeptember)All available data for Arctic glaciated regions can then bepooled according to whether it refers to primary productionwithin a glacier fjord and whether or not it could plausiblybe influenced by the presence of a marine-terminating glacier(see Sect 5) For the purposes of defining the spatial extentof individual glacier fjords we consider broad bay areas suchas the lower and central parts of Glacier Bay (Etheringtonand Hooge 2007 Reisdorph and Mathis 2015) ScoresbySund (Scoresby Sound in English Seifert et al 2019) andDisko Bay (Jensen et al 1999 Nielsen 1999) to be be-yond the scale of the associated glacier fjords on the basis ofthe oceanographic interpretation presented in the respectivestudies Defining the potential spatial influence of marine-terminating glaciers is more challenging Using observationsfrom Godtharingbsfjord where primary production is found to beaffected on a scale of 30ndash80 km down-fjord from the marine-terminating glaciers therein (Meire et al 2017) we define

a region lt 80 km downstream of calving fronts as being po-tentially influenced by marine-terminating glaciers

Four exclusive categories of primary production data re-sult (Table 1) Primary production for group I is significantlyhigher than any other group and group II is also signifi-cantly higher than group IV (p lt 0025) Primary productionis higher in regions designated as having a potential marine-terminating glacier influence On the contrary other near-glacier regions (ie with land-terminating glaciers) seem tohave low summertime primary productivity irrespective ofhow mean Arctic primary production is defined (Table 1)What processes could lead to such differences In the nextsections of this review we discuss the biogeochemical fea-tures of glacier-affected marine regions that could potentiallyexplain such trends if they do not simply reflect data defi-ciency

4 Effects of glacial discharge on marine resourceavailability

One of the most direct mechanisms via which glacial dis-charge affects downstream marine primary production is byaltering the availability of light macronutrients (such as ni-trate NO3 phosphate PO4 and silicic acid Si) andor mi-cronutrients (such as iron and manganese) in the ocean Thechemical composition of glacial discharge is now relativelywell constrained especially around Greenland (Yde et al2014 Meire et al 2016a Stevenson et al 2017) Alaska(Hood and Berner 2009 Schroth et al 2011) and Svalbard(Hodson et al 2004 2016) Whilst high particle loads (Chuet al 2012 Overeem et al 2017) and Si are often associ-ated with glacially modified waters (Fig 3a) around the Arc-tic (Brown et al 2010 Meire et al 2016a) the concentra-tions of all macronutrients in glacial discharge (Meire et al2016a) are relatively low and similar to those of coastal sea-water (Fig 3a b and c)

Macronutrient concentrations in Arctic rivers can behigher than in glacier discharge (Holmes et al 2011)(Fig 3d e and f) Nevertheless river and glacier meltwateralike do not significantly increase the concentration of PO4in Arctic coastal waters (Fig 3c and f) River water isrelatively a much more important source of NO3 (Cauwetand Sidorov 1996 Emmerton et al 2008 Hessen et al2010) and in river estuaries this nutrient can show a sharpdecline with increasing salinity due to both mixing andbiological uptake (Fig 3e) Patterns in Si are more variable(Cauwet and Sidorov 1996 Emmerton et al 2008 Hessenet al 2010) Dissolved Si concentration at low salinity ishigher in rivers than in glacier discharge (Fig 3a and d)yet a variety of estuarine behaviours are observed acrossthe Arctic Peak dissolved Si occurs at a varying salinitydue to the opposing effects of Si release from particles anddissolved Si uptake by diatoms (Fig 3d)

A notable feature of glacial freshwater outflows into theocean is the high turbidity that occurs in most Arctic glacierfjords High turbidity in surface waters within glacier fjordsarises from the high sediment transport in these drainage sys-tems (Chu et al 2012) from iceberg melting and also fromthe resuspension of fine sediments (Azetsu-Scott and Syvit-ski 1999 Zajaczkowski and Włodarska-Kowalczuk 2007Stevens et al 2016) The generally high sediment load ofglacially derived freshwater is evident around Greenlandwhich is the origin of sim 1 of annual freshwater dischargeinto the ocean yet 7 ndash9 of the annual fluvial sedimentload (Overeem et al 2017) Sediment load is however spa-tially and temporally variable leading to pronounced inter-and intra-catchment differences (Murray et al 2015) Forexample satellite-derived estimates of sediment load for 160Greenlandic glacier outflows suggest a median sediment loadof 992 mg Lminus1 but some catchments exhibit gt 3000 mg Lminus1

(Overeem et al 2017) Furthermore it is suggested that gt25 of the total annual sediment load is released in a singleoutflow (from the Sermeq glacier) (Overeem et al 2017)

The extent to which high turbidity in glacier outflows lim-its light availability in downstream marine environments istherefore highly variable between catchments and with dis-tance from glacier outflows (Murray et al 2015 Mascaren-has and Zielinski 2019) The occurrence and effects of sub-surface turbidity peaks close to glaciers is less well studiedSubsurface turbidity features may be even more spatially andtemporally variable than their surface counterparts (Stevenset al 2016 Kanna et al 2018 Moskalik et al 2018) Ingeneral a spatial expansion of near-surface turbid plumesis expected with increasing glacier discharge but this trendis not always evident at the catchment scale (Chu et al2009 2012 Hudson et al 2014) Furthermore with long-term glacier retreat the sediment load in discharge at thecoastline is generally expected to decline as proglacial lakesare efficient sediment traps (Bullard 2013 Normandeau etal 2019)

In addition to high turbidity the low concentration ofmacronutrients in glacier discharge relative to saline watersis evidenced by the estuarine mixing diagram in Kongsfjor-den (Fig 3) and confirmed by extensive measurements offreshwater nutrient concentrations (eg Hodson et al 20042005) For PO4 (Fig 3c) there is a slight increase in concen-tration with salinity (ie discharge dilutes the nutrient con-centration in the fjord) For NO3 discharge slightly increases

Table 1 JulyndashSeptember marine primary production (PP) data from studies conducted in glaciated Arctic regions PP data points are cate-gorised into four groups according to whether or not they are within 80 km of a marine-terminating glacier and whether or not they are withina glacier fjord Data sources as per Fig 2 n is the number of data points where studies report primary production measurements at the samestation for the same month at multiple time points (eg Juul-Pedersen et al 2015) a single mean is used in the data compilation (ie n= 1irrespective of the historical extent of the time series)

Mean PP(plusmn standard deviation)

Category mg C mminus2 dminus1 n Data from

(I) Marine-terminating glacierinfluence non-fjord

847plusmn 852 11 Disko Bay Scoresby Sund GlacierBay North Greenland Canadian ArcticArchipelago

(II) Marine-terminating glacierinfluence glacier fjord

480plusmn 403 33 Godtharingbsfjord Kongsfjorden ScoresbySund Glacier Bay Hornsund

(III) No marine-terminating glacierinfluence non-fjord

304plusmn 261 42 Godtharingbsfjord Young Sound ScoresbySund Disko Bay Canadian ArcticArchipelago

(IV) No marine-terminating glacierinfluence glacier fjord

125plusmn 102 35 Godtharingbsfjord Young Sound Kangerlus-suaq Disko Bay

Figure 3 (a) Si (b) NO3 and (c) PO4 distributions across the measured salinity gradient in Kongsfjorden in summer 2013 (Fransson et al2016) 2014 (Fransson et al 2016) 2015 (van de Poll et al 2018) and 2016 (Cantoni et al 2019) Full depth data are shown with a linearregression (black line) for glacially modified waters (S lt 342) during summer 2016 The position of stations varies between the datasetswith the 2016 data providing the broadest coverage of the inner fjord Linear regression details are shown in Table S1 in the Supplement(d) Si (e) NO3 and (f) PO4 distributions in surface waters of three major Arctic river estuaries the Lena Mackenzie and Yenisey (Cauwetand Sidorov 1996 Emmerton et al 2008 Hessen et al 2010) Note the different y- and x-axis scales

the concentration in the upper-mixed layer (Fig 3b) For Si asteady decline in Si with increasing salinity (Fig 3a) is con-sistent with a discharge-associated Si supply (Brown et al2010 Arimitsu et al 2016 Meire et al 2016a) The spa-tial distribution of data for summer 2013ndash2016 is similar andrepresentative of summertime conditions in the fjord (Hop etal 2002)

Whilst dissolved macronutrient concentrations in glacialdischarge are relatively low a characteristic of glaciatedcatchments is extremely high particulate Fe concentrationsHigh Fe concentrations arise both directly from glacier dis-charge (Bhatia et al 2013a Hawkings et al 2014) and alsofrom resuspension of glacially derived sediments throughoutthe year (Markussen et al 2016 Crusius et al 2017) Total

dissolvable Fe (TdFe) concentrations within Godtharingbsfjordare high in all available datasets (May 2014 August 2014 andJuly 2015) and strongly correlated with turbidity (linear re-gression R2

= 088 R2= 056 and R2

= 088 respectivelyHopwood et al 2016 2018) A critical question in oceanog-raphy in both the Arctic and Antarctic is to what extent thislarge pool of particulate Fe is transferred into open-ocean en-vironments and thus potentially able to affect marine primaryproduction in Fe-limited offshore regions (Gerringa et al2012 Arrigo et al 2017 Schlosser et al 2018) The mech-anisms that promote transfer of particulate Fe into bioavail-able dissolved phases such as ligand-mediated dissolution(Thuroczy et al 2012) and biological activity (Schmidt et al2011) and the scavenging processes that return dissolved Feto the particulate phase are both poorly characterized (Tagli-abue et al 2016)

Fe profiles around the Arctic show strong spatial vari-ability in TdFe concentrations ranging from unusually highconcentrations of up to 20 microM found intermittently close toturbid glacial outflows (Zhang et al 2015 Markussen etal 2016 Hopwood et al 2018) to generally low nanomo-lar concentrations at the interface between shelf and fjordwaters (Zhang et al 2015 Crusius et al 2017 Cape etal 2019) An interesting feature of some of these profilesaround Greenland is the presence of peak Fe at sim 50 mdepth perhaps suggesting that much of the Fe transportaway from glaciers may occur in subsurface turbid glaciallymodified waters (Hopwood et al 2018 Cape et al 2019)The spatial extent of Fe enrichment downstream of glaciersaround the Arctic is still uncertain but there is evidence ofglobal variability downstream of glaciers on the scale of 10ndash100 km (Gerringa et al 2012 Annett et al 2017 Crusius etal 2017)

41 Non-conservative mixing processes for Fe and Si

A key reason for uncertainty in the fate of glacially derivedFe is the non-conservative behaviour of dissolved Fe in salinewaters In the absence of biological processes (ie nutrientassimilation and remineralization) NO3 is expected to ex-hibit conservative behaviour across estuarine salinity gradi-ents (ie the concentration at any salinity is a linear functionof mixing between fresh and saline waters) For Fe how-ever a classic non-conservative estuarine behaviour occursdue to the removal of dissolved Fe (DFe1) as it flocculatesand is absorbed onto particle surfaces more readily at highersalinity and pH (Boyle et al 1977) Dissolved Fe concen-trations almost invariably exhibit strong (typically sim 90 )non-conservative removal across estuarine salinity gradients(Boyle et al 1977 Sholkovitz et al 1978) and glaciatedcatchments appear to be no exception to this rule (Lippiattet al 2010) Dissolved Fe in Godtharingbsfjord exhibits a re-

1For consistency dissolved Fe is defined throughout opera-tionally as lt 02 micro m and is therefore inclusive of ionic complexedand colloidal species

moval of gt 80 DFe between salinities of 0ndash30 (Hopwoodet al 2016) and similar losses of approximately 98 forKongsfjorden and 85 for the Copper riverestuary (Gulfof Alaska) system have been reported (Schroth et al 2014Zhang et al 2015)

Conversely Si can be released from particulate phases dur-ing estuarine mixing resulting in non-conservative additionto dissolved Si concentrations (Windom et al 1991) al-though salinityndashSi relationships vary between different estu-aries due to different extents of Si release from labile particu-lates and Si uptake by diatoms (eg Fig 3d) Where evidentthis release of dissolved Si typically occurs at low salinities(Cauwet and Sidorov 1996 Emmerton et al 2008 Hessenet al 2010) with the behaviour of Si being more conser-vative at higher salinities and in estuaries where pronounceddrawdown by diatoms is not evident (eg Brown et al 2010)Estimating release of particulate Si from Kongsfjorden data(Fig 3c) as the additional dissolved Si present above theconservative mixing line for runoff mixing with unmodifiedsaline water that is entering the fjord (via linear regression)suggests a Si enrichment of 13plusmn 2 (Fig 3a) This isbroadly consistent with the 6 ndash53 range reported for es-tuarine gradients evident in some temperate estuaries (Win-dom et al 1991) Conversely Hawkings et al (2017) sug-gest a far greater dissolution downstream of Leverett Glacierequivalent to a 70 ndash800 Si enrichment and thus proposethat the role of glaciers in the marine Si cycle has been under-estimated Given that such dissolution is substantially abovethe range observed in any other Arctic estuary the apparentcause is worth further consideration

The general distribution of Si in surface waters for Kongs-fjorden (Fransson et al 2016) Godtharingbsfjord (Meire et al2016a) Bowdoin Fjord (Kanna et al 2018) Sermilik (Capeet al 2019) and along the Gulf of Alaska (Brown et al2010) is similar Si shows pseudo-conservative behaviour de-clining with increasing salinity in surface waters The limitedreported number of zero-salinity or very low salinity end-members for Godtharingbsfjord and Bowdoin are significantlybelow the linear regression derived from surface nutrient andsalinity data (Fig 4) In addition to some dissolution of par-ticulate Si another likely reason for this is the limitation ofindividual zero-salinity measurements in dynamic fjord sys-tems where different discharge outflows have different nu-trient concentrations (Kanna et al 2018) especially giventhat subglacial discharge is not directly characterized in ei-ther location (Meire et al 2016a Kanna et al 2018) As

Figure 4 Dissolved Si distribution vs salinity for glaciated Arc-tic catchments Data are from Bowdoin Fjord (Kanna et al 2018)Kongsfjorden (Fransson et al 2016 van de Poll et al 2018) Ser-milik Fjord (Cape et al 2019) Kangerlussuaq (Hawkings et al2017 Lund-Hansen et al 2018) Godtharingbsfjord (Hopwood et al2016 Meire et al 2016b) and the Gulf of Alaska (Brown et al2010) Linear regressions are shown for large surface datasets onlyLinear regression details are shown in Table S1 Closed markers in-dicate surface data (lt 20 m depth) and open markers indicate sub-surface data

demonstrated by the two different zero-salinity Si endmem-bers in Kongsfjorden (iceberg melt ofsim 003 microM and surfacerunoff of sim 59 microM) pronounced deviations in nutrient con-tent arise from mixing between various freshwater endmem-bers (surface runoff ice melt and subglacial discharge) Forexample total freshwater input into Godtharingbsfjord is 70 ndash80 liquid with this component consisting of 64 ice sheetrunoff 31 land runoff and 5 net precipitation (Langenet al 2015) and being subject to additional inputs from ice-berg melt along the fjord (sim 70 of calved ice also meltswithin the inner fjord Bendtsen et al 2015)

In a marine context at broad scales a single freshwa-ter endmember that integrates the net contribution of allfreshwater sources can be defined This endmember includesiceberg melt groundwater discharge surface and subsur-face glacier discharge and (depending on location) sea-icemelt which are challenging to distinguish in coastal waters(Benetti et al 2019) Close to glaciers it may be possibleto observe distinct freshwater signatures in different watercolumn layers and distinguish chemical signatures in wa-ter masses containing subglacial discharge from those con-taining primarily surface runoff and iceberg melt (eg inGodtharingbsfjord Meire et al 2016a and Sermilik Beaird et

al 2018) but this is often challenging due to mixing andoverlap between different sources Back-calculating the inte-grated freshwater endmember (eg from regression Fig 4)can potentially resolve the difficulty in accounting for data-deficient freshwater components and poorly characterized es-tuarine processes As often noted in field studies there is ageneral bias towards sampling of supraglacial meltwater andrunoff in proglacial environments and a complete absence ofchemical data for subglacial discharge emerging from largemarine-terminating glaciers (eg Kanna et al 2018)

Macronutrient distributions in Bowdoin Godtharingbsfjordand Sermilik unambiguously show that the primarymacronutrient supply to surface waters associated withglacier discharge originates from mixing rather than fromfreshwater addition (Meire et al 2016a Kanna et al 2018Cape et al 2019) which emphasizes the need to considerfjord inflowoutflow dynamics in order to interpret nutrientdistributions The apparently anomalous extent of Si dissolu-tion downstream of Leverett Glacier (Hawkings et al 2017)may therefore largely reflect underestimation of both thesaline (assumed to be negligible) and freshwater endmem-bers rather than unusually prolific particulate Si dissolutionIn any case measured Si concentrations in the Kangerlus-suaq region are within the range of other Arctic glacier estu-aries (Fig 4) making it challenging to support the hypothesisthat glacial contributions to the Si cycle have been underesti-mated elsewhere (see also Tables 2 and 3)

42 Deriving glacierndashocean fluxes

In the discussion of macronutrients herein we have focusedon the availability of the bioavailable species (eg PO4 NO3and silicic acid) that control seasonal trends in inter-annualmarine primary production (Juul-Pedersen et al 2015 vande Poll et al 2018 Holding et al 2019) It should be notedthat the total elemental fluxes (ie nitrogen phosphorus andsilicon) associated with lithogenic particles are invariablyhigher than the associated macronutrients (Wadham et al2019) particularly for phosphorus (Hawkings et al 2016)and silicon (Hawkings et al 2017) Lithogenic particles arehowever not bioavailable although they may to some extentbe bioaccessible depending on the temporal and spatial scaleinvolved This is especially the case for the poorly quantifiedfraction of lithogenic particles that escapes sedimentation ininner-fjord environments either directly or via resuspensionof shallow sediments (Markussen et al 2016 Hendry et al2019) It is hypothesized that lithogenic particle inputs fromglaciers therefore have a positive influence on Arctic marineprimary production (Wadham et al 2019) yet field data tosupport this hypothesis are lacking A pan-Arctic synthesisof all available primary production data for glaciated regions(Fig 2 and Table 1) spatial patterns in productivity alongthe west Greenland coastline (Meire et al 2017) popula-tion responses in glacier fjords across multiple taxonomicgroups (Cauvy-Fraunieacute and Dangles 2019) and sedimentary

records from Kongsfjorden (Kumar et al 2018) consistentlysuggest that glaciers or specifically increasing volumes ofglacier discharge have a net negative or negligible effecton marine primary producers ndash except in the specific case ofsome marine-terminating glaciers where a different mecha-nism seems to operate (see Sect 5)

Two linked hypotheses can be proposed to explain theseapparently contradictory arguments One is that whilstlithogenic particles are potentially a bioaccessible source ofFe P and Si they are deficient in bioaccessible N As NO3availability is expected to limit primary production acrossmuch of the Arctic (Tremblay et al 2015) this creates aspatial mismatch between nutrient supply and the nutrientdemand required to increase Arctic primary production A re-lated alternative hypothesis is that the negative effects of dis-charge on marine primary production (eg via stratificationand light limitation from high turbidity) more than offset anypositive effect that lithogenic particles have via increasingnutrient availability on regional scales prior to extensive sed-imentation occurring A similar conclusion has been reachedfrom analysis of primary production in proglacial streams(Uehlinger et al 2010) To some extent this reconciliationis also supported by considering the relative magnitudes ofdifferent physical and chemical processes acting on differentspatial scales with respect to global marine primary produc-tion (see Sect 10)

The generally low concentrations of macronutrients anddissolved organic matter (DOM) in glacier discharge relativeto coastal seawater (Table 2) have an important methodolog-ical implication because what constitutes a positive NO3PO4 or DOM flux into the Arctic Ocean in a glaciologicalcontext can actually reduce short-term nutrient availabilityin the marine environment It is therefore necessary to con-sider both the glacier discharge and saline endmembers thatmix in fjords alongside fjord-scale circulation patterns inorder to constrain the change in nutrient availability to ma-rine biota (Meire et al 2016a Hopwood et al 2018 Kannaet al 2018)

Despite the relatively well constrained nutrient signatureof glacial discharge around the Arctic estimated fluxes ofsome nutrients from glaciers to the ocean appear to be sub-ject to greater variability especially for nutrients subject tonon-conservative mixing (Table 3) Estimates of the Fe fluxfrom the Greenland Ice Sheet for example have an 11-folddifference between the lowest (gt 26 Mmol yrminus1) and highest(290 Mmol yrminus1) values (Hawkings et al 2014 Stevenson etal 2017) However it is debatable if these differences in Feflux are significant because they largely arise in differencesbetween definitions of the flux gate window and especiallyhow estuarine Fe removal is accounted for Given that thedifference between an estimated removal factor of 90 and99 is a factor of 10 difference in the calculated DFe fluxthere is overlap in all of the calculated fluxes for GreenlandIce Sheet discharge into the ocean (Table 3) (Statham et al2008 Bhatia et al 2013a Hawkings et al 2014 Stevenson

et al 2017) Conversely estimates of DOM export (quanti-fied as DOC) are confined to a slightly narrower range of 7ndash40 Gmol yrminus1 with differences arising from changes in mea-sured DOM concentrations (Bhatia et al 2013b Lawson etal 2014b Hood et al 2015) The characterization of glacialDOM with respect to its lability C N ratio and implicationsfor bacterial productivity in the marine environment (Hood etal 2015 Paulsen et al 2017) is however not readily appar-ent from a simple flux calculation

A scaled-up calculation using freshwater concentrations(C) and discharge volumes (Q) is the simplest way ofdetermining the flux from a glaciated catchment to theocean However discharge nutrient concentrations varyseasonally (Hawkings et al 2016 Wadham et al 2016)often resulting in variable CndashQ relationships due to changesin mixing ratios between different discharge flow pathspost-mixing reactions and seasonal changes in microbialbehaviour in the snowpack on glacier surfaces and inproglacial forefields (Brown et al 1994 Hodson et al2005) Therefore full seasonal datasets from a range ofrepresentative glaciers are required to accurately describeCndashQ relationships Furthermore as the indirect effectsof discharge on nutrient availability to phytoplankton viaestuarine circulation and stratification are expected to be agreater influence than the direct nutrient outflow associatedwith discharge (Rysgaard et al 2003 Juul-Pedersen etal 2015 Meire et al 2016a) freshwater data must becoupled to physical and chemical time series in the coastalenvironment if the net effect of discharge on nutrientavailability in the marine environment is to be understoodIndeed the recently emphasized hypothesis that nutrientfluxes from glaciers into the ocean have been significantlyunderestimated (Hawkings et al 2016 2017 Wadham et al2016) is difficult to reconcile with a synthesis and analysis ofavailable marine nutrient distributions (Sect 4) in glaciatedArctic catchments especially for Si (Fig 4)

A particularly interesting case study concerning thelink between marine primary production circulation anddischarge-derived nutrient fluxes is Young Sound It was ini-tially stipulated that increasing discharge into the fjord in re-sponse to climate change would increase estuarine circula-tion and therefore macronutrient supply Combined with alonger sea-ice-free growing season as Arctic temperatures

Table 2 Measuredcomputed discharge and saline endmembers for well-studied Arctic fjords (ND not determinednot reported BD belowdetection)

Fjord Dataset Salinity NO3 (microM) PO4 (microM) Si (microM) TdFe (microM)

Kongsfjorden Summer 2016 00 (ice melt) 087plusmn 10 002plusmn 003 003plusmn 003 338plusmn 100(Svalbard) (Cantoni et al 2019) 00 (surface discharge) 094plusmn 10 0057plusmn 031 591plusmn 41 74plusmn 76

3450plusmn 017 125plusmn 049 020plusmn 006 100plusmn 033 ND

Nuup Kangerlua Summer 2014 00 (ice melt) 196plusmn 168 004plusmn 004 13plusmn 15 031plusmn 049Godtharingbsfjord (Hopwood et al 2016 00 (surface discharge) 160plusmn 044 002plusmn 001 122plusmn 163 138(Greenland) Meire et al 2016a) 3357plusmn 005 115plusmn 15 079plusmn 004 80plusmn 10 ND

Sermilik Summer 2015 00 (subglacial discharge) 18plusmn 05 ND 10plusmn 8 ND(Greenland) (Cape et al 2019) 00 (ice melt) 097plusmn 15 ND 4plusmn 4 ND

349plusmn 01 128plusmn 1 ND 615plusmn 1 ND

Bowdoin Summer 2016 00 (surface discharge) 022plusmn 015 030plusmn 020 BD ND(Greenland) (Kanna et al 2018) 343plusmn 01 147plusmn 09 11plusmn 01 195plusmn 15 ND

Young Sound Summer 2014 (Runoff JulyndashAugust) 12plusmn 074 029plusmn 02 952plusmn 38 ND(Greenland) (Paulsen et al 2017) (Runoff SeptemberndashOctober) 10plusmn 07 035plusmn 02 2957plusmn 109 ND

336plusmn 01 (JulyndashAugust) 64plusmn 11 118plusmn 05 666plusmn 04 ND335plusmn 004 (SeptemberndashOctober) 56plusmn 02 062plusmn 02 65plusmn 01 ND

increase this would be expected to increase primary pro-duction within the fjord (Rysgaard et al 1999 Rysgaardand Glud 2007) Yet freshwater input also stratifies the fjordthroughout summer and ensures low macronutrient availabil-ity in surface waters (Bendtsen et al 2014 Meire et al2016a) which results in low summertime productivity in theinner and central fjord (lt 40 mg C mminus2 dminus1) (Rysgaard etal 1999 2003 Rysgaard and Glud 2007) Whilst annualdischarge volumes into the fjord have increased over the pasttwo decades resulting in a mean annual 012plusmn005 (practicalsalinity units) freshening of fjord waters (Sejr et al 2017)shelf waters have also freshened This has potentially im-peded the dense inflow of saline waters into the fjord (Booneet al 2018) and therefore counteracted the expected increasein productivity

43 How do variations in the behaviour and location ofhigher-trophic-level organisms affect nutrientavailability to marine microorganisms

With the exception of some zooplankton and fish speciesthat struggle to adapt to the strong salinity gradients andorsuspended particle loads in inner-fjord environments (Wccedils-lawski and Legezytnska 1998 Lydersen et al 2014)higher-trophic-level organisms (including mammals andbirds) are not directly affected by the physicalchemicalgradients caused by glacier discharge However their foodsources such as zooplankton and some fish species aredirectly affected and therefore there are many examplesof higher-level organisms adapting their feeding strategieswithin glacier fjord environments (Arimitsu et al 2012Renner et al 2012 Laidre et al 2016) Strong gradientsin physicalchemical gradients downstream of glaciers par-ticularly turbidity can therefore create localized hotspots of

secondary productivity in areas where primary production islow (Lydersen et al 2014)

It is debatable to what extent shifts in these feeding pat-terns could have broadscale biogeochemical effects Whilstsome species are widely described as ecosystem engineerssuch as Alle alle (the little auk) in the Greenland North Wa-ter Polynya (Gonzaacutelez-Bergonzoni et al 2017) for changesin higher-trophic-level organismsrsquo feeding habits to have sig-nificant direct chemical effects on the scale of a glacier fjordsystem would require relatively large concentrations of suchanimals Nevertheless in some specific hotspot regions thiseffect is significant enough to be measurable There is am-ple evidence that birds intentionally target upwelling plumesin front of glaciers as feeding grounds possibly due to thestunning effect that turbid upwelling plumes have upon preysuch as zooplankton (Hop et al 2002 Lydersen et al 2014)This feeding activity therefore concentrates the effect ofavian nutrient recycling within a smaller area than wouldotherwise be the case potentially leading to modest nutri-ent enrichment of these proglacial environments Yet withthe exception of large concentrated bird colonies the effectsof such activity are likely modest In Kongsfjorden bird pop-ulations are well studied and several species are associatedwith feeding in proglacial plumes yet still collectively con-sume only between 01 and 53 of the carbon producedby phytoplankton in the fjord (Hop et al 2002) The esti-mated corresponding nutrient flux into the fjord from birds is2 mmol mminus2 yrminus1 nitrogen and 03 mmol mminus2 yrminus1 phospho-rous

Table 3 Flux calculations for dissolved nutrients (Fe DOC DON NO3 PO4 and Si) from Greenland Ice Sheet discharge Where a fluxwas not calculated in the original work an assumed discharge volume of 1000 km3 yrminus1 is used to derive a flux for comparative purposes(ASi amorphous silica LPP labile particulate phosphorous) For DOM PO4 and NO3 non-conservative estuarine behaviour is expectedto be minor or negligible Note that whilst we have defined ldquodissolvedrdquo herein as lt 02 microm the sampling and filtration techniques usedparticularly in freshwater studies are not well standardized and thus some differences may arise between studies accordingly Clogging offilters in turbid waters reduces the effective filter pore size DOP DON NH4 and PO4 concentrations often approach analytical detectionlimits which alongside fieldanalytical blanks are treated differently low concentrations of NO3 DON DOP DOC NH4 and DFe are easilyinadvertently introduced to samples by contamination and measured Si concentrations can be significantly lower when samples have beenfrozen

Freshwaterendmember

concentrationNutrient (microM) Flux Estuarine modification Data

Fe 013 gt 26 Mmol yrminus1 Inclusive gt 80 loss Hopwood et al (2016)164 39 Mmol yrminus1 Assumed 90 loss Stevenson et al (2017)

0053 53 Mmol yrminus1 Discussed not applied Statham et al (2008)370 180 Mmol yrminus1 Assumed 90 loss Bhatia et al (2013a)071 290 Mmol yrminus1 Discussed not applied Hawkings et al (2014)

DOC 16ndash100 67 Gmol yrminus1 Not discussed Bhatia et al (2010 2013b)12ndash41 11ndash14 Gmol yrminus1 Not discussed Lawson et al (2014b)

15ndash100 18 Gmol yrminus1 Not discussed Hood et al (2015)2ndash290 24ndash38 Gmol yrminus1 Not discussed Csank et al (2019)27ndash47 40 Gmol yrminus1 Not discussed Paulsen et al (2017)

DON 47ndash54 5 Gmol yrminus1 Not discussed Paulsen et al (2017)17 07ndash11 Gmol yrminus1 Not discussed Wadham et al (2016)

Si 13ndash28 22 Gmol yrminus1 Inclusive Meire et al (2016a)96 4 Gmol yrminus1 Discussed Hawkings et al (2017)

(+190 Gmol yrminus1 ASi)

PO4 023 010 Gmol yrminus1 Discussed Hawkings et al (2016)(+023 Gmol yrminus1 LPP)

026 026 Gmol yrminus1 Not discussed Meire et al (2016a)

NO3 14ndash15 042 Gmol yrminus1 Not discussed Wadham et al (2016)05ndash17 05ndash17 Gmol yrminus1 Not discussed Paulsen et al (2017)

5 Critical differences between surface and subsurfacedischarge release

Critical differences arise between land-terminating andmarine-terminating glaciers with respect to their effects onwater column structure and associated patterns in primary

production (Table 1) Multiple glacier fjord surveys haveshown that fjords with large marine-terminating glaciersaround the Arctic are normally more productive than theirland-terminating glacier fjord counterparts (Meire et al2017 Kanna et al 2018) and despite large inter-fjord vari-ability (Fig 2) this observation appears to be significantacross all available primary production data for Arctic glacierfjords (Table 1) A particularly critical insight is that fjord-scale summertime productivity along the west Greenlandcoastline scales approximately with discharge downstream ofmarine-terminating glaciers but not land-terminating glaciers(Meire et al 2017) The primary explanation for this phe-nomenon is the vertical nutrient flux associated with mixingdriven by subglacial discharge plumes which has been quan-tified in field studies at Bowdoin glacier (Kanna et al 2018)

Sermilik Fjord (Cape et al 2019) Kongsfjorden (Halbach etal 2019) and in Godtharingbsfjord (Meire et al 2016a)

As discharge is released at the glacial grounding linedepth its buoyancy and momentum result in an upwellingplume that entrains and mixes with ambient seawater (Car-roll et al 2015 2016 Cowton et al 2015) In Bowdoin Ser-milik and Godtharingbsfjord this nutrient pump provides 99 97 and 87 respectively of the NO3 associated withglacier inputs to each fjord system (Meire et al 2016aKanna et al 2018 Cape et al 2019) Whilst the pan-Arctic magnitude of this nutrient pump is challenging toquantify because of the uniqueness of glacier fjord systemsin terms of their geometry circulation residence time andglacier grounding line depths (Straneo and Cenedese 2015Morlighem et al 2017) it can be approximated in genericterms because plume theory (Morton et al 1956) has beenused extensively to describe subglacial discharge plumes inthe marine environment (Jenkins 2011 Hewitt 2020) Com-puted estimates of subglacial discharge for the 12 Greenlandglacier fjord systems where sufficient data are available tosimulate plume entrainment (Carroll et al 2016) suggestthat the entrainment effect is at least 2 orders of magni-tude more important for macronutrient availability than di-rect freshwater runoff (Hopwood et al 2018) This is con-sistent with limited available field observations (Meire et al2016a Kanna et al 2018 Cape et al 2019) As macronu-trient fluxes have been estimated independently using differ-ent datasets and plume entrainment models in two of theseglacier fjord systems (Sermilik and Illulissat) an assessmentof the robustness of these fluxes can also be made (Table 4)(Hopwood et al 2018 Cape et al 2019) Exactly how theseplumes and any associated fluxes will change with the com-bined effects of glacier retreat and increasing glacier dis-charge remains unclear (De Andreacutes et al 2020) but maylead to large changes in fjord biogeochemistry (Torsvik et al2019) Despite different definitions of the macronutrient flux(Table 4 ldquoArdquo refers to the out-of-fjord transport at a definedfjord cross-section window whereas ldquoBrdquo refers to the ver-tical transport within the immediate vicinity of the glacier)the fluxes are reasonably comparable and in both cases un-ambiguously dominate macronutrient glacier-associated in-put into these fjord systems (Hopwood et al 2018 Cape etal 2019)

Whilst large compared to changes in macronutrient avail-ability from discharge without entrainment (Table 3) itshould be noted that these nutrient fluxes (Table 4) are stillonly intermediate contributions to fjord-scale macronutrientsupply compared to total annual consumption in these en-vironments For example in Godtharingbsfjord mean annual pri-mary production is 1037 g C mminus2 yrminus1 equivalent to biolog-ical consumption of 11 mol N mminus2 yrminus1 Entrainment fromthe three marine-terminating glaciers within the fjord is con-servatively estimated to supply 001ndash012 mol N mminus2 yrminus1

(Meire et al 2017) ie 1 ndash11 of the total N supply re-quired for primary production if production were supported

exclusively by new NO3 (rather than recycling) and equallydistributed across the entire fjord surface Whilst this is con-sistent with observations suggesting relative stability in meanannual primary production in Godtharingbsfjord from 2005 to2012 (1037plusmn178 g C mminus2 yrminus1 Juul-Pedersen et al 2015)despite pronounced increases in total discharge into the fjordthis does not preclude a much stronger influence of entrain-ment on primary production in the inner-fjord environmentThe time series is constructed at the fjord mouth over 120 kmfrom the nearest glacier and the estimates of subglacial dis-charge and entrainment used by Meire et al (2017) are bothunrealistically low If the same conservative estimate of en-trainment is assumed to only affect productivity in the mainfjord branch (where the three marine-terminating glaciers arelocated) for example the lower bound for the contribution ofentrainment becomes 3 ndash33 of total N supply Similarlyin Kongsfjorden ndash the surface area of which is considerablysmaller compared to Godtharingbsfjord (sim 230 km2 compared to650 km2) ndash even the relatively weak entrainment from shal-low marine-terminating glaciers (Fig 5) accounts for approx-imately 19 ndash32 of N supply An additional mechanismof N supply evident there which partially offsets the inef-ficiency of macronutrient entrainment at shallow groundingline depths is the entrainment of ammonium from shallowbenthic sources (Halbach et al 2019) which leads to unusu-ally high NH4 concentrations in surface waters Changes insubglacial discharge or in the entrainment factor (eg froma shift in glacier grounding line depth Carroll et al 2016)can therefore potentially change fjord-scale productivity

A specific deficiency in the literature to date is the ab-sence of measured subglacial discharge rates from marine-terminating glaciers Variability in such rates on diurnal andseasonal timescales is expected (Schild et al 2016 Fried etal 2018) and intermittent periods of extremely high dis-charge are known to occur for example from ice-dammedlake drainage in Godtharingbsfjord (Kjeldsen et al 2014) Yetdetermining the extent to which these events affect fjord-scale mixing and biogeochemistry as well as how these rateschange in response to climate forcing will require furtherfield observations Paradoxically one of the major knowl-edge gaps concerning low-frequency high-discharge eventsis their biological effects yet these events first became char-acterized in Godtharingbsfjord after observations by a fishermanof a sudden Sebastes marinus (Redfish) mortality event inthe vicinity of a marine-terminating glacier terminus Theseunfortunate fish were propelled rapidly to the surface by as-cending freshwater during a high-discharge event (Kjeldsenet al 2014)

A further deficiency yet to be specifically addressed inbiogeochemical studies is the decoupling of different mixingprocesses in glacier fjords In this section we have primar-ily considered the effect of subglacial discharge plumes onNO3 supply to near-surface waters downstream of marine-terminating glaciers (Fig 5) Yet a similar effect can arisefrom down-fjord katabatic winds which facilitate the out-

Table 4 A comparison of upwelled NO3 fluxes calculated from fjord-specific observed nutrient distributions (A) (Cape et al 2019) andusing regional nutrient profiles with idealized plume theory (B) (Hopwood et al 2018) ldquoArdquo refers to the out-of-fjord transport of nutrientswhereas ldquoBrdquo refers to the vertical transport close to the glacier terminus

Field (A) Calculated (B) Idealizedcampaign(s) out-of-fjord NO3 NO3 upwelling

Location for A export Gmol yrminus1 Gmol yrminus1

Ilulissat Icefjord 2000ndash2016 29plusmn 09 42(Jakobshavn Isbraelig)Sermilik (Helheim Glacier) 2015 088 20Sermilik (Helheim Glacier) 2000ndash2016 12plusmn 03

Figure 5 The plume dilution (entrainment) factor relationship withglacier grounding line depth as modelled by Carroll et al (2016)for subglacial freshwater discharge rates of 250ndash500 m3 sminus1 andgrounding lines of gt 100 m (shaded area) Also shown are theentrainment factors determined from field observations for Kro-nebreen (Kongsfjorden Kr Halbach et al 2019) Bowdoin (BnKanna et al 2018) Saqqarliup Sermia (SS Mankoff et al 2016)Narsap Sermia (Ns Meire et al 2016a) Kangerlussuup Sermia(KS Jackson et al 2017) Kangiata Nunaata Sermia (KNS Bendt-sen et al 2015) Sermilik (Sk Beaird et al 2018) and Nioghalvf-jerdsfjorden Glacier (the ldquo79 N Glacierrdquo 79N Schaffer et al2020) Note that the 79 N Glacier is unusual compared to the otherArctic systems displayed as subglacial discharge there enters a largecavity beneath a floating ice tongue and accounts for only 11 ofmeltwater entering this cavity with the rest derived from basal icemelt (Schaffer et al 2020)

of-fjord transport of low-salinity surface waters and the in-flow of generally macronutrient-rich saline waters at depth(Svendsen et al 2002 Johnson et al 2011 Spall et al2017) Both subglacial discharge and down-fjord windstherefore contribute to physical changes affecting macronu-trient availability on a similar spatial scale and both pro-cesses are expected to be subject to substantial short-term(hours-days) seasonal and inter-fjord variability which ispresently poorly constrained (Spall et al 2017 Sundfjordet al 2017)

51 Is benthicndashpelagic coupling enhanced by subglacialdischarge

The attribution of unusually high near-surface NH4 concen-trations in surface waters of Kongsfjorden to benthic releasein this relatively shallow fjord followed by upwelling closeto the Kronebreen calving front (Halbach et al 2019) raisesquestions about where else this phenomenon could be im-portant and which other biogeochemical compounds couldbe made available to pelagic organisms by such enhancedbenthicndashpelagic coupling The summertime discharge-drivenupwelling flux within a glacier fjord of any chemical which isreleased into bottom water from sediments for example FeMn (Wehrmann et al 2014) dissolved organic phosphorous(DOP) dissolved organic nitrogen (DON) (Koziorowska etal 2018) or Si (Hendry et al 2019) could potentially beincreased to varying degrees depending on sediment com-position (Wehrmann et al 2014 Glud et al 2000) and theinterrelated nature of fjord circulation topography and thedepth range over which entrainment occurs

Where such benthicndashupwelling coupling does occur closeto glacier termini it may be challenging to quantify from wa-ter column observations due to the overlap with other pro-cesses causing nutrient enrichment For example the mod-erately high dissolved Fe concentrations observed close toAntarctic ice shelves were classically attributed mainly to di-rect freshwater inputs but it is now thought that the directfreshwater input and the Fe entering surface waters from en-trainment of Fe-enriched near-bottom waters could be com-parable in magnitude (St-Laurent et al 2017) although withlarge uncertainty This adds further complexity to the roleof coastal fjord and glacier geometry in controlling nutri-ent bioaccessibility and determining the significance of suchcoupling is a priority for hybrid modelndashfield studies

52 From pelagic primary production to the carbonsink

Whilst primary production is a major driver of CO2 draw-down from the atmosphere to the surface ocean much of thisC is subject to remineralization and following bacterial orphotochemical degradation of organic carbon re-enters the

atmosphere as CO2 on short timescales The biological Cpump refers to the small fraction of sinking C which is se-questered in the deep ocean or in sediments There is nosimple relationship between primary production and C ex-port into the deep ocean as a range of primary-productionndashC-export relationships have been derived globally with theunderlying cause subject to ongoing discussion (Le Moigneet al 2016 Henson et al 2019)

Irrespective of global patterns glacier fjords are notablefor their extremely high rates of sedimentation due to highlithogenic particle inputs (Howe et al 2010) In additionto terrestrially derived material providing additional organiccarbon for burial in fjords (Table 3) ballasting of sinkingPOC (particulate organic carbon) by lithogenic material gen-erally increases the efficiency of the biological C pump byfacilitating more rapid transfer of C to depth (Iversen andRobert 2015 Pabortsava et al 2017) With high sedimentloads and steep topography fjords are therefore expected tobe efficient POC sinks especially when normalized with re-spect to their surface area (Smith et al 2015) Organic car-bon accumulation rates in Arctic glacier fjords are far lowerthan temperate fjord systems likely due to a combination ofgenerally lower terrestrially derived carbon inputs and some-times lower marine primary production but Arctic fjordswith glaciers still exhibit higher C accumulation than Arcticfjords without glaciers (Włodarska-Kowalczuk et al 2019)

The limited available POC fluxes for Arctic glacier fjordssupport the hypothesis that they are efficient regions of POCexport (Wiedmann et al 2016 Seifert et al 2019) POCequivalent to 28 ndash82 of primary production was foundto be transferred to gt 100 m depth in Nordvestfjord (westGreenland) (Seifert et al 2019) This represents medium-to-high export efficiency compared to other marine environ-ments on a global scale (Henson et al 2019) High lithogenicparticle inputs into Arctic glacier fjords could therefore beconsidered to maintain a low-primary-productionndashhigh-C-export-efficiency regime On the one hand they limit lightavailability and thus contribute to relatively low levels of pri-mary production (Table 1) but concurrently they ensure thata relatively high fraction of C fixed by primary producers istransferred to depth (Seifert et al 2019)

Beyond the potent impact of high sedimentation on ben-thic ecosystems (Włodarska-Kowalczuk et al 2001 2005)which is beyond the scope of this review and the ballast-ing effect which is sparsely studied in this environment todate (Seifert et al 2019) relatively little is known about theinteractive effects of concurrent biogeochemical processeson glacier-derived particle surfaces occurring during theirsuspension (or resuspension) in near-shore waters Chem-ical processes occurring at turbid freshwaterndashsaline inter-faces such as dissolved Fe and DOM scavenging onto par-ticle surfaces and phosphate or DOM co-precipitation withFe oxyhydroxides (eg Sholkovitz et al 1978 Charette andSholkovitz 2002 Hyacinthe and Van Cappellen 2004) haveyet to be extensively studied in Arctic glacier estuaries where

they may exert some influence on nutrient availability andC cycling

6 Contrasting Fe- and NO3-limited regions of the ocean

Whether or not nutrients transported to the ocean surfacehave an immediate positive effect on marine primary pro-duction depends on the identity of the resource(s) that limitsmarine primary production Light attenuation is the ultimatelimiting control on marine primary production and is exacer-bated close to turbid glacial outflows (Hop et al 2002 Arim-itsu et al 2012 Murray et al 2015) However the spatialextent of sediment plumes andor ice meacutelange which limitlight penetration into the water column is typically restrictedto within kilometres of the glacier terminus (Arimitsu et al2012 Hudson et al 2014 Lydersen et al 2014) Beyond theturbid light-limited vicinity of glacial outflows the proximallimiting resource for summertime marine primary productionwill likely be a nutrient the identity of which varies with lo-cation globally (Moore et al 2013) Increasing the supplyof the proximal limiting nutrient would be expected to havea positive influence on marine primary production whereasincreasing the supply of other nutrients alone would not ndash apremise of ldquothe law of the minimumrdquo (Debaar 1994) Al-though proximal limiting nutrient availability controls totalprimary production organic carbon and nutrient stoichiome-try nevertheless has specific effects on the predominance ofdifferent phytoplankton and bacterial groups (Egge and Ak-snes 1992 Egge and Heimdal 1994 Thingstad et al 2008)

The continental shelf is a major source of Fe into the ocean(Lam and Bishop 2008 Charette et al 2016) and this re-sults in clear differences in proximal limiting nutrients be-tween Arctic and Antarctic marine environments The iso-lated Southern Ocean is the worldrsquos largest high-nitrate low-chlorophyll (HNLC) zone where Fe extensively limits pri-mary production even in coastal polynyas (Sedwick et al2011) and macronutrients are generally present at high con-centrations in surface waters (Martin et al 1990a b) Con-versely the Arctic Ocean is exposed to extensive broad shelfareas with associated Fe input from rivers and shelf sed-iments and thus generally has a greater availability of Ferelative to macronutrient supply (Klunder et al 2012) Fe-limited summertime conditions have been reported in partsof the Arctic and sub-Arctic (Nielsdottir et al 2009 Ryan-Keogh et al 2013 Rijkenberg et al 2018) but are spatiallyand temporally limited compared to the geographically ex-tensive HNLC conditions in the Southern Ocean

However few experimental studies have directly assessedthe nutrient limitation status of regions within the vicin-ity of glaciated Arctic catchments With extremely high Feinput into these catchments NO3 limitation might be ex-pected year-round However PO4 limitation is also plausibleclose to glaciers in strongly stratified fjords (Prado-Fiedler2009) due to the low availability of PO4 in freshwater rel-

ative to NO3 (Ren et al 2019) Conversely in the South-ern Ocean it is possible that Fe-limited conditions occurextremely close to glaciers and ice shelves (Fig 6) High-NO3 low-Fe water can be found in the immediate vicin-ity of Antarcticarsquos coastline (Gerringa et al 2012 Marsayet al 2017) and even in inshore bays (Annett et al 2015Houmlfer et al 2019) Macronutrient data from Maxwell Bay(King George Island South Shetland Islands) for examplesuggest that Fe from local glaciers mixes with high-NO3high-Si ocean waters providing ideal conditions for phyto-plankton blooms in terms of nutrient availability The lowestsurface macronutrient concentrations measured in MaxwellBay in a summer campaign were 17 microM NO3 14 microM PO4and 47 microM Si (Houmlfer et al 2019) Similarly in Ryder Bay(Antarctic Peninsula) the lowest measured annual macronu-trient concentrations ndash occurring after strong drawdown dur-ing a pronounced phytoplankton bloom (22 mg mminus3 chloro-phyll a) ndash were 25 microM NO3 and 04 microM PO4 (Annett et al2015) This contrasts starkly with the summertime surfacemacronutrient distribution in glaciated fjords in the Arcticincluding Kongsfjorden (Fig 3) where surface macronutri-ent concentrations are typically depleted throughout summerThese differences may explain why some Antarctic glacierfjords have significantly higher chlorophyll and biomass thanany of the Arctic glacier fjord systems considered herein(Mascioni et al 2019) However we note a general lack ofseasonal and interannual data for Antarctic glacier fjord sys-tems precludes a comprehensive inter-comparison of thesedifferent systems

For a hypothetical nutrient flux from a glacier the sameflux could be envisaged in two endmember scenarios oneseveral kilometres inside an Arctic fjord (eg Godtharingbsfjordor Kongsfjorden) and one at the coastline of an isolatedSouthern Ocean island such as the Kerguelen (Bucciarelliet al 2001 Bowie et al 2015) Heard (van der Merwe etal 2019) or South Shetland Islands (Houmlfer et al 2019) Inthe Arctic fjord a pronounced Fe flux from summertime dis-charge would likely have no immediate positive effect uponfjord-scale marine primary production because Fe may al-ready be replete (Hopwood et al 2016 Crusius et al 2017)This is consistent with the observation that Fe-rich dischargefrom land-terminating glaciers around west Greenland doesnot have a positive fjord-scale fertilization effect (Meire etal 2017) and may possibly be associated with a negativeeffect (Table 1) Conversely the same Fe input into coastalwaters around the Kerguelen Islands would be expected tohave a pronounced positive effect upon marine primary pro-duction because the islands occur within the worldrsquos largestHNLC zone Where Fe is advected offshore in the wake ofthe islands a general positive effect on primary productionis expected (Blain et al 2001 Bucciarelli et al 2001) eventhough there are marked changes in the phytoplankton com-munity composition between the Fe-enriched bloom region(dominated by microphytoplankton) and the offshore HNLCarea (dominated by small diatoms and nanoflagellates) (Uitz

Figure 6 Contrasting nutrient properties of water on the (a) south-east Greenland shelf (data from Achterberg et al 2018) with (b) theRoss Sea shelf (data from Marsay et al 2017) Note the differentscales used on the x axes

et al 2009) However even in these HNLC waters there arealso other concurrent factors that locally mitigate the effect ofglacially derived Fe in nearshore waters because light limita-tion from near-surface particle plumes may locally offset anypositive effect of Fe fertilization (Wojtasiewicz et al 2019)

61 The subglacial discharge pump frommacronutrients to iron

The effect of the subglacial discharge nutrient pump maysimilarly vary with location Contrasting the NO3 and DFeconcentrations of marine environments observed adjacent todifferent glacier systems suggests substantial variations inthe proximal limiting nutrient of these waters on a globalscale (Fig 7) In Antarctic shelf regions such as the westernAntarctic Peninsula a high log-transformed ratio of summer-time NO3 DFe (median value 2) is indicative of Fe limita-tion Across the Arctic there is a broader range of ratios (me-dian values minus12 to 13) indicating spatial variability in thebalance between Fe and NO3 limitation (Fig 7) Variation isevident even within specific regions The range of NO3 DFeratios for both the Gulf of Alaska (log10minus25 to 17) and thesouth Greenland shelf (log10minus15 to 18) includes values thatare indicative of the full spectrum of responses from NO3limitation to FeNO3 co-limitation to Fe limitation (Brown-ing et al 2017) This suggests a relatively rapid spatial tran-sition from excess to deficient DFe conditions

How would the marine-terminating glacier upwelling ef-fect operate in an Fe-limited system The physical mecha-nism of a nutrient pump would be identical for glaciers withthe same discharge and grounding line one in a high-Fe low-NO3 Arctic system and one in a low-Fe high-NO3 Antarc-tic system However the biogeochemical consequences withrespect to marine primary production would be different (Ta-ble 5) In the case of subglacial discharge for simplicity weconsider a mid-depth glacier (grounding line of 100ndash250 m

Figure 7 Variations in the ratio of dissolved NO3 and Fe in surfacewaters (lt 20 m) adjacent to glaciated regions whiskers show the10th and 90th percentiles bars shows the median 25th percentileand 75th percentile and dots show all outliers Data from the west-ern Antarctic Peninsula (WAP Annett et al 2017 Ducklow et al2017) the south Greenland shelf (Achterberg et al 2018 Tonnardet al 2020) Godtharingbsfjord (Hopwood et al 2016) Kongsfjorden(Hopwood et al 2017) the Gulf of Alaska (Lippiatt et al 2010)and the NE Greenland shelf (Hopwood et al 2018) For Kongs-fjorden NO3 and Fe data were interpolated using the NO3ndashsalinityrelationship

below sea level) with a constant discharge rate of 250 m3 sminus1An entrainment factor of 6ndash10 would then be predicted byplume theory (Fig 5) (Carroll et al 2016) In a Greenlandfjord with no sill to constrain circulation and a residence timeshort enough that inflowing nutrient concentrations were notchanged significantly prior to entrainment an average NO3concentration of 5ndash12 microM is predicted in the entrained wa-ter compared to sim 2 microM in glacier discharge (Hopwood etal 2018) Over a 2-month discharge period this would pro-duce a NO3 flux of 40ndash160 Mmol NO3 with 2 ndash6 of theNO3 flux arising from meltwater discharge and 94 ndash98 from plume entrainment Complete utilization of this NO3 byphytoplankton according to the Redfield ratio (106 C 16 N)(Redfield 1934) would correspond to a biological sink of027ndash10 Gmol C

In an analogous HNLC environment surface NO3 require-ments would already vastly exceed phytoplankton require-ments (Fig 7) due to extensive Fe limitation of primary pro-duction Thus whilst the upwelled NO3 flux would be largerin an Fe-limited system due to higher concentrations of NO3in the water column (see Fig 6) the short-term biological ef-fect of upwelling NO3 alone would be negligible More im-portant would be the upwelling of the proximal limiting nu-trient Fe If we assume that dissolved Fe in the marine watercolumn is in a stable bioavailable form and that additionaldissolved Fe from freshwater is delivered to the marine en-vironment with a 90 ndash99 loss during estuarine mixing(Table 3) the upwelled Fe flux can be estimated Upwelledunmodified water from a depth of 100ndash250 m would be ex-

pected to contain 006ndash012 nM Fe (Marsay et al 2017)The freshwater endmember in the context of an Antarcticcalving ice front would largely consist of ice melt (ratherthan subglacial discharge Hewitt 2020) so we use an in-termediate freshwater Fe endmember of 33ndash680 nM in icemelt (Annett et al 2017 Hodson et al 2017) Upwellingvia the same hypothetical 250 m3 sminus1 discharge as per theArctic scenario would generate a combined upwelled anddischarge flux (after estuarine removal processes) of 089ndash89 kmol Fe with 2 ndash52 of the Fe arising from upwellingand 48 ndash98 from freshwater Using an intermediate Fe C value of 5 mmol Fe molminus1C which is broadly applicableto the coastal environment (Twining and Baines 2013) thiswould correspond to a biological pool of 0019ndash19 Gmol CIt should be noted that the uncertainty on this calculation isparticularly large because unlike NO3 upwelling there is alack of in situ data to constrain the simultaneous mixing andnon-conservative behaviour of Fe

For a surface discharge of 250 m3 sminus1 nutrient entrain-ment is assumed to be negligible In the case of Fe outflowinto a low-Fe high-NO3 system we assume that the glacieroutflow is the dominant local Fe source over the fertilizedarea during the discharge period (ie changes to other sourcesof Fe such as the diffusive flux from shelf sediments are neg-ligible) For the case of surface discharge into a low-NO3high-Fe system this is not likely to be the case for NO3Stratification induced by discharge decreases the vertical fluxof NO3 from below thus negatively affecting NO3 supply al-though there are to our knowledge no studies quantifying thischange in glacially modified waters

It is clear from these simplified discharge scenarios (Ta-ble 5) that both the depth at which glacier discharge is re-leased into the water column and the relative availabilities ofNO3 and Fe in downstream waters could be critical for de-termining the response of primary producers The responseof primary producers in low-Fe regimes is notably subject tomuch larger uncertainty mainly because of uncertainty in theextent of Fe removal during estuarine mixing (Schroth et al2014 Zhang et al 2015) Whilst the effects of the marine-terminating glacier nutrient pump on macronutrient fluxeshave been defined in numerous systems its effect on Feavailability is poorly constrained (Gerringa et al 2012 St-Laurent et al 2017 2019) Furthermore Fe bioavailability isconceptually more complicated than discussed herein as ma-rine organisms at multiple trophic levels affect the speciationbioaccessibility and bioavailability of Fe as well as the trans-fer between less-labile and more-labile Fe pools in the ma-rine environment (Poorvin et al 2004 Vraspir and Butler2009 Gledhill and Buck 2012) Many microbial species re-lease organic ligands into solution which stabilize dissolvedFe as organic complexes These feedbacks are challenging tomodel (Strzepek et al 2005) but may exert a cap on the lat-eral transfer of Fe away from glacier inputs (Lippiatt et al2010 Thuroczy et al 2012) To date Fe fluxes from glaciersinto the ocean have primarily been constructed from an inor-

Table 5 Suppositional effect of different discharge scenarios calculated from the Redfield ratio 106 C 16 N 1 P 0005 Fe (Redfield 1934Twining and Baines 2013) A steady freshwater discharge of 250 m3 sminus1 is either released from a land-terminating glacier or from a marine-terminating glacier at 100ndash250 m depth in both cases for two months into Fe-replete NO3-deficient or Fe-deficient NO3-replete marineenvironments Freshwater endmembers are defined as 2 microM NO3 and 33ndash675 nM dissolved Fe (Annett et al 2017 Hodson et al 2017Hopwood et al 2018) Ambient water column conditions are defined as Greenland (Achterberg et al 2018) (ie high-Fe low-NO3) andRoss Sea (Marsay et al 2017) (ie low-Fe high-NO3) shelf profiles

Surface discharge Subglacial discharge

High-Fe low-NO3environment(predominant Arctic condition)

eg Young Soundlt 0ndash0017 Gmol C

eg Bowdoin FjordSermilik027ndash10 Gmol C

Low-Fe high-NO3 environment(predominant Antarcticcondition)

eg Antarctic Peninsula0009ndash19 Gmol C

ganic freshwater perspective (Raiswell et al 2006 Raiswelland Canfield 2012 Hawkings et al 2014) Yet to understandthe net change in Fe availability to marine biota a greater un-derstanding of how ligands and estuarine mixing processesmoderate the glacier-to-ocean Fe transfer will evidently berequired (Lippiatt et al 2010 Schroth et al 2014 Zhang etal 2015)

7 Effects on the carbonate system

Beyond its impact on inorganic nutrient dynamics glacialdischarge also affects the inorganic carbon system com-monly referred to as the carbonate system in seawater Thecarbonate system describes the seawater buffer system andconsists of dissolved CO2 and carbonic acid bicarbonateions and carbonate ions These components buffer pH andare the main reason for the oceanrsquos capacity to absorb at-mospheric CO2 The interaction between these chemicalspecies which varies with physical conditions including tem-perature and salinity (Dickson and Millero 1987) dictatesthe pH of seawater and the saturation state of biologicallyimportant carbonate minerals such as aragonite and calcite(Ar and Ca respectively) Discharge generally reducesthe total alkalinity (TA buffering capacity) of glacially mod-ified waters mainly through dilution (Fig 8) which resultsin a decreased carbonate ion concentration Since carbonateions are the main control on the solubility of CaCO3 de-creasing carbonate ion availability due to meltwater dilutionnegatively impacts the aragonite and calcite saturation state(Doney et al 2009 Fransson et al 2015) Glacier dischargecan also moderate the carbonate system indirectly as higherprimary production leads to increased biological dissolvedinorganic carbon (DIC) uptake lower pCO2 and thus higherpH in seawater Therefore increasing or decreasing primaryproduction also moderates pH and the aragonite and calcitesaturation state of marine surface waters

Total alkalinity measurements of glacial discharge acrossthe Arctic reveal a range from 20 to 550 micromol kgminus1 (Yde et

al 2005 Sejr et al 2011 Rysgaard et al 2012 Evans etal 2014 Fransson et al 2015 2016 Meire et al 2015Turk et al 2016) Similar to Si concentrations the broadrange is likely explained by different degrees of interactionbetween meltwater and bedrock with higher alkalinity corre-sponding to greater dischargendashbedrock interaction (Wadhamet al 2010 Ryu and Jacobson 2012) and also reflects localchanges in bedrock geology (Yde et al 2005 Fransson etal 2015) However in absolute terms even the upper end ofthe alkalinity range reported in glacial discharge is very lowcompared to the volume-weighted average of Arctic rivers1048 micromol kgminus1 (Cooper et al 2008) In an Arctic contextmeltwater is therefore relatively corrosive In addition to lowtotal alkalinity glacier estuaries can exhibit undersaturationof pCO2 due to the non-linear effect of salinity on pCO2(Rysgaard et al 2012 Meire et al 2015) This undersatura-tion arises even when the freshwater endmember is in equi-librium with atmospheric pCO2 and thus part of the CO2drawdown observed in Arctic glacier estuaries is inorganicand not associated with primary production In Godtharingbs-fjord this effect is estimated to account for 28 of total CO2uptake within the fjord (Meire et al 2015)

By decreasing the TA of glacially modified waters (Fig 8)glacier discharge reduces the aragonite and calcite saturationstates thereby amplifying the effect of ocean acidification(Fransson et al 2015 2016 Ericson et al 2019) High pri-mary production can mitigate this impact as photosyntheticCO2 uptake reduces DIC and pCO2 (eg Fig 9) in surfacewaters and increases the calcium carbonate saturation state(Chierici and Fransson 2009 Rysgaard et al 2012 Meire etal 2015) In relatively productive fjords the negative effectof TA dilution may therefore be counter balanced Howeverin systems where discharge-driven stratification is responsi-ble for low productivity increased discharge may create apositive feedback on ocean acidification state in the coastalzone resulting in a lower saturation state of calcium carbon-ate (Chierici and Fransson 2009 Ericson et al 2019)

Figure 8 Total alkalinity in Kongsfjorden during the meltwater sea-son (data from Fransson and Chierici 2019) A decline in alkalin-ity is evident with increasing freshwater fraction in response to thelow alkalinity concentrations in glacier discharge Freshwater frac-tion was calculated using an average marine salinity endmemberof 3496 hence some slightly negative values are calculated in theouter fjord associated with the higher salinity of unmodified At-lantic water Linear regression details are shown in Table S1

Low-calcium carbonate saturation states (lt 1 ie cor-rosive conditions) have been observed in the inner part ofGlacier Bay (Alaska) demonstrating that glaciers can am-plify seasonal differences in the carbonate system and nega-tively affect the viability of shell-forming marine organisms(Evans et al 2014) Low Ar has also been observed inthe inner parts of Kongsfjorden coinciding with high glacialdischarge (Fransson et al 2016) Such critically low Ar(lt 14) conditions have negative effects on aragonite-shell-forming calcifiers such as the pteropod Limacina helicina(Comeau et al 2009 2010 Lischka et al 2011 Lischka andRiebesell 2012 Bednaršek et al 2014) Under future cli-mate scenarios in addition to the effect of increased glacierdrainage in glacier fjords synergistic effects with a combina-tion of increased ocean CO2 uptake and warming will furtheramplify changes to the ocean acidification state (Fransson etal 2016 Ericson et al 2019) resulting in increasingly pro-nounced negative effects on calcium carbonate shell forma-tion (Lischka and Riebesell 2012)

8 Organic matter in glacial discharge

In addition to inorganic ions glacial discharge also con-tains many organic compounds derived from biological ac-tivity on glacier surfaces and overridden sediments (Barkeret al 2006 Lawson et al 2014b) Organic carbon stimu-lates bacterial activity and remineralization of organic mat-ter is a pathway to resupply labile nitrogen and phosphorousto microbial communities Similar to macronutrient concen-trations DOM concentrations in glacial discharge are gen-erally low (Table 2) compared to runoff from large Arcticrivers which have DOM concentrations 1ndash2 orders of mag-nitude higher (Dittmar and Kattner 2003 Le Fouest et al

2013) This is evidenced in Young Sound where dissolvedorganic carbon (DOC) concentrations increase with salinityin surface waters demonstrating that glaciers are a relativelyminor source of DOM to the fjord (Paulsen et al 2017)

While DOM concentrations are low in glacial dischargethe bioavailability of this DOM is much higher than its ma-rine counterpart (Hood et al 2009 Lawson et al 2014bPaulsen et al 2017) This is likely due to the low C N ratioof glacial DOM as N-rich DOM of microbial origin is gen-erally highly labile (Lawson et al 2014a) It has been sug-gested that as glaciers retreat and the surrounding catchmentsbecome more vegetated DOC concentrations in these catch-ments will increase (Hood and Berner 2009 Csank et al2019) However DOM from non-glacial terrestrial sourceshas a higher composition of aromatic compounds and thus isless labile (Hood and Berner 2009 Csank et al 2019) Fur-thermore glacier coverage in watersheds is negatively corre-lated with DOC DON ratios so a reduction in the lability ofDOM with less glacial coverage is also expected (Hood andScott 2008 Hood and Berner 2009 Ren et al 2019)

While DOC is sufficient to drive bacterial metabolismbacteria also depend on nitrogen and phosphorus for growthIn this respect bacteria are in direct competition with phy-toplankton for macronutrients and increasing additions oflabile DOM downstream of glaciers could give bacteria acompetitive edge This would have important ecological con-sequences for the function of the microbial food web andthe biological carbon sink (Larsen et al 2015) Experimentswith Arctic fjord communities including Kongsfjorden haveshown that when bacteria are supplied with additional sub-sidies of labile carbon under nitrate limitation they outcom-pete phytoplankton for nitrate (Thingstad et al 2008 Larsenet al 2015) This is even the case when there is an additionof excess Si which might be hypothesized to give diatoms acompetitive advantage The implications of such competitionfor the carbon cycle are however complicated by mixotro-phy (Ward and Follows 2016 Stoecker et al 2017) An in-creasing number of primary producers have been shown tobe able to simultaneously exploit inorganic resources and liv-ing prey combining autotrophy and phagotrophy in a singlecell Mixotrophy allows protists to sustain photosynthesis inwaters that are severely nutrient limited and provides an ad-ditional source of carbon as a supplement to photosynthesisThis double benefit decreases the dependence of primary pro-ducers on short-term inorganic nutrient availability More-over mixotrophy promotes a shortened and potentially moreefficient chain from nutrient regeneration to primary pro-duction (Mitra et al 2014) Whilst mixotrophy is sparselystudied in Arctic glacier fjords both increasing temperaturesand stratification are expected to favour mixotrophic species(Stoecker and Lavrentyev 2018) and thus an understandingof microbial food web dynamics is vital to predict the impli-cations of increasing discharge on the carbon cycle in glacierfjord systems

Regardless of the high bioavailability of DOM fromglacial discharge once glacial DOM enters a fjord and isdiluted by ocean waters evidence of its uptake forming asignificant component of the microbial food web in the Arc-tic has yet to be observed Work from several outlet glacierfjords around Svalbard shows that the stable isotopic C ra-tio of bacteria does not match that of DOC originating fromlocal glaciers suggesting that glacially supplied DOC is aminor component of bacterial consumption compared to au-tochthonous carbon sources (Holding et al 2017 Paulsenet al 2018) Curiously a data synthesis of taxonomic pop-ulations for glaciated catchments globally suggests a signif-icant positive effect of glaciers on bacterial populations inglacier fjords but a negative effect in freshwaters and glacierforefields (Cauvy-Fraunieacute and Dangles 2019) This suggeststhat multiple ecological and physicalndashchemical processes areat play such that a simplistic argument that increasing glacialsupply of DOC favours bacterial activity is moderated byother ecological factors This is perhaps not surprising as dif-ferent taxonomic groups may respond differently to perturba-tions from glacier discharge leading to changes in food webdynamics For example highly turbid glacial waters haveparticularly strong negative effects on filter-feeding (Arendtet al 2011 Fuentes et al 2016) and phagotrophic organisms(Sommaruga 2015) and may also lead to reduced viral loadsin the water column due to adsorption onto particle surfaces(Maat et al 2019)

Whilst concentrations of DOM are low in glacier dis-charge DOM-sourced nitrogen and phosphorous could stillbe relatively important in stratified outlet glacier fjords sim-ply because inorganic nutrient concentrations are also low(eg Fig 3) Refractory DON in rivers that is not directly de-graded by bacteria can be subsequently broken down by pho-toammonification processes releasing ammonium (Xie et al2012) In large Arctic rivers this nitrogen supply is greaterthan that supplied from inorganic sources (Le Fouest et al2013) For glacier discharge processing of refractory DOMcould potentially produce a comparable nitrogen flux to in-organic sources (Table 2 Wadham et al 2016) Similarly inenvironments where inorganic PO4 concentrations are lowDOP may be a relatively more important source of phospho-rous for both bacteria and phytoplankton Many freshwaterand marine phytoplankton species are able to synthesize theenzyme alkaline phosphatase in order to efficiently utilizeDOP (Hoppe 2003 Štrojsovaacute et al 2005) In the context ofstratified low-salinity inner-fjord environments where inor-ganic PO4 concentrations are potentially low enough to limitprimary production (Prado-Fiedler 2009) this process maybe particularly important ndash yet DOP dynamics are understud-ied in glaciated catchments with limited data available (Stibalet al 2009 Hawkings et al 2016)

Finally whilst DOC concentrations in glacier dischargeare low POC concentrations which may also impact micro-bial productivity in the marine environment and contribute tothe C sink within fjords are less well characterized Down-

stream of Leverett Glacier mean runoff POC concentrationsare reported to be 43ndash346 microM ndash 5 times higher than DOC(Lawson et al 2014b) However the opposite is reported forYoung Sound where DOC concentrations in three glacier-fed streams were found to be 7ndash13 times higher than POCconcentrations (Paulsen et al 2017) Similarly low POCconcentrations of only 5 microM were found in supraglacial dis-charge at Bowdoin glacier (Kanna et al 2018) In summaryrelatively little is presently known about the distribution fateand bioavailability of POC in glaciated catchments

9 Insights into the long-term effects of glacier retreat

Much of the present interest in Arctic icendashocean interactionsarises because of the accelerating increase in discharge fromthe Greenland Ice Sheet captured by multi-annual to multi-decadal time series (Bamber et al 2018) This trend is at-tributed to atmospheric and oceanic warming due to anthro-pogenic forcing at times enhanced by persistent shifts in at-mospheric circulation (Box 2002 Ahlstroumlm et al 2017)From existing observations it is clear that strong climatevariability patterns are at play such as the North Atlantic Os-cillationArctic Oscillation and that in order to place recentchange in context time series exceeding the satellite era arerequired Insight can be potentially gained from research intopast sedimentary records of productivity from high-latitudemarine and fjord environments Records of productivity andthe dominance of different taxa as inferred by microfos-sils biogeochemical proxies and genetic records from thosespecies that preserve well in sediment cores can help es-tablish long-term spatial and temporal patterns around thepresent-day ice sheet periphery (Ribeiro et al 2012) AroundGreenland and Svalbard sediment cores largely corroboraterecent fjord-scale surveys suggesting that inner-fjord watercolumn environments are generally low-productivity systems(Kumar et al 2018) with protist taxonomic diversity andoverall productivity normally higher in shelf waters than ininner-fjord environments (Ribeiro et al 2017)

Several paleoclimate archives and numerical simulationssuggest that the Arctic was warmer than today during theearly to mid-Holocene thermal maximum (sim 8000 yearsago) which was registered by sim 1 km thinning of the Green-land Ice Sheet (Lecavalier et al 2017) Multiproxy analy-ses performed on high-resolution and well-dated Holocenemarine sediment records from contrasting fjord systems aretherefore one approach to understand the nature of suchpast events as these sediments simultaneously record climateand some long-term biotic changes representing a uniquewindow into the past However while glacialndashinterglacialchanges can provide insights into large-scale icendashocean inter-actions and the long-term impact of glaciers on primary pro-duction these timescales are of limited use to understandingmore recent variability at the icendashocean interface of fjord sys-tems such as those mentioned in this review The five well-

characterized Arctic fjords used as case studies here (Fig 1Bowdoin Kongsfjorden Sermilik Godtharingbsfjord and YoungSound) for example did not exist during the Last GlacialMaximum sim 19 000 years ago (Knutz et al 2011)

On long timescales glacierndashocean interactions are sub-ject to marked temporal changes associated with glacialndashinterglacial cycles In the short term the position of glaciertermini shifts inland during ice sheet retreat or outwards dur-ing ice sheet expansion and in the long-term proglacial re-gions respond to isostatic uplift and delta progradation Theuplift of fine-grained glaciomarine and deltaic sediments is anotable feature of landscape development in fjord environ-ments following the retreat of continental-scale ice sheets(Cable et al 2018 Gilbert et al 2018) This results in thegradual exposure and subsequent erosion of these sedimentinfills and their upstream floodplains releasing labile organicmatter to coastal ecosystems Whilst the direct biogeochem-ical significance of such chemical fluxes may be limited inthe marine environment on interannual timescales (Table 2)potentially more important is the Fe fertilization followingwind erosion and dust emittance from glacial floodplains

Ice core records from Greenland and Antarctica span-ning several climatic cycles suggest that aeolian depositionrates at high latitudes were as much as 20 times greater dur-ing glacial than interglacial periods (Kohfeld and Harrison2001) Elevated input of terrigenous Fe during windy glacialepisodes and associated continental drying has thereforebeen hypothesized to stimulate oceanic productivity throughtime and thus modify the oceanic and atmospheric CO2 bal-ance (Martin 1990) While there seems to be a pervasivedustndashclimate feedback on a glacialndashinterglacial planetaryscale (Shaffer and Lambert 2018) glacier retreat also ex-poses new areas of unconsolidated glacial sediments leadingto an increase in both dust storm events and sediment yieldsfrom glacial basins locally The spatial scale over which thisglacially derived dust can be transported (100ndash500 km) farexceeds that of discharge-carried nutrients (Crusius et al2011 Prospero et al 2012 Bullard 2013)

10 A need for new approaches

The pronounced temporal and spatial variations evident inthe properties of glacially modified waters emphasize theneed for high-resolution data on both short (hourly to daily)and long (seasonal to interannual) timescales in order tounderstand glacial processes and their downstream effectsIn Godtharingbsfjord Juul-Pedersen et al (2015) provide a de-tailed study of seasonal primary production dynamics Thismonthly monitoring programme captures seasonal annualand interannual trends in the magnitude of primary produc-tion Whilst such a time series clearly highlights a stronginterannual stability in both seasonal and annual primaryproduction (1037plusmn 178 g C mminus2 yrminus1 Juul-Pedersen et al2015) it is unable to fully characterize shorter (ie days to

weeks) timescale events such as the spring bloom period Yethigher data resolution cannot feasibly be sustained by ship-board campaigns

Low-frequency high-discharge events are known to occurin Godtharingbsfjord and other glacier fjords (Kjeldsen et al2014) but are challenging to observe from monthly reso-lution data and thus there is sparse data available to quan-tify their occurrence and effects or to quantify the short-term variation in discharge rates at large dynamic marine-terminating glaciers Consequently modelled subglacial dis-charge rates and glacier discharge derived from regionalmodels (eg RACMO Noeumll et al 2015) which underpinour best-available estimates of the subglacial nutrient pump(eg Carroll et al 2016) do not yet consider such variabil-ity Time lapse imagery shows that the lifetimes and spatialextents of subglacial discharge plumes can vary considerably(Schild et al 2016 Fried et al 2018) While buoyant plumetheory has offered important insights into the role of sub-glacial plumes in the nutrient pump buoyant plume theorydoes not characterize the lateral expansion of plume watersFurthermore determining the influence of discharge beyondthe immediate vicinity of glacial outflows is a Lagrangianexercise yet the majority of existing observational and mod-elling studies have been conducted primarily in the Eule-rian reference frame (eg ship-based profiles and mooredobservations that describe the water column at a fixed loca-tion) Moving towards an observational Lagrangian frame-work will require the deployment of new technology such asthe recent development of low-cost GPS trackers which es-pecially when combined with in situ sensors may improveour understanding of the transport and mixing of heat fresh-water sediment and nutrients downstream of glaciers (Carl-son et al 2017 Carlson and Rysgaard 2018) For exam-ple GPS trackers deployed on ldquobergy bitsrdquo have revealedevidence of small-scale retentive eddies in Godtharingbsfjord(Carlson et al 2017) and characterized the surface flow vari-ability in Sermilik Fjord (Sutherland et al 2014)

Unmanned aerial vehicles and autonomous sur-faceunderwater vehicles can also be used to observethe spatio-temporal variability of subglacial plumes athigh resolution (Mankoff et al 2016 Jouvet et al 2018)Complementing these approaches are developments in therapidly maturing field of miniaturized chemical sensorssuitable for use in cryosphere environments (Beaton et al2012) Such technology will ultimately reduce much ofthe uncertainty associated with glacierndashocean interactionsby facilitating more comprehensive more sustainable fieldcampaigns (Straneo et al 2019) with reduced costs andenvironmental footprints (Nightingale et al 2015 Grand etal 2017 2019) This is evidenced by a successful prolongedmooring deployment in the Santa Ineacutes Glacier fjord system(Fig 9)

The Santa Ineacutes Glacier fjord sits adjacent to the open waterof the Straits of Magellan in southwest Patagonia Mooredhigh-resolution measurements are now collected in situ us-

Figure 9 Winterndashspring dynamics of salinity pH and pCO2 atthe Santa Ineacutes Glacier fjord Ballena (Patagonia) High-resolutionpCO2 and pH measurements (every three hours) were taken in situusing autonomous SAMI-CO2 and SAMI-pH sensors (as per Ver-gara-Jara et al 2019) (Sunburst Sensors LLC) starting in the aus-tral autumn (March 2018) All sensors were moored at 10 m depth

ing sensor technology and a mooring within the fjord Mea-surements include the carbonate system parameters pCO2and pH The 2018 winter to spring time series (Fig 9)demonstrates a sharp decline in pCO2 and corresponding in-crease in pH associated with the onset of the spring bloomin early October Such a pronounced event occurring oversim 2 weeks would be impossible to characterize fully withmonthly sampling of the fjord Over winter pH and pCO2were more stable but sensor salinity data still reveal short-term dynamics within the fjordsrsquo surface waters (Fig 9) Ageneral decline in salinity is evident moving from winterinto spring Short-term changes on diurnal timescales ndash pre-sumably linked to tidal forcing ndash and also on dailyndashweeklytimescales ndash possibly linked to weather patterns ndash are alsoevident (Fig 9) Much work remains to be done to deducethe role of these short-term drivers on primary production

Finally we note that the different scales over which theprocesses discussed herein operate raises the critical ques-tion of how importantly the different effects of glacial dis-charge on the marine environment are perceived in differ-ent research fields Herein we have largely focused on local-to regional-scale processes operating on seasonal to inter-annual timescales in the marine environment at individualfield sites (Fig 1) A very different emphasis may have beenplaced on the relative importance of different processes if adifferent spatialtemporal perspective had been adopted for

Figure 10 A scale comparison of the significance of differentchemicalphysical processes driven by glacial discharge in termsof the resulting effects on annual marine primary production (PP)or CO2 drawdown (units Tg C yrminus1) Bold lines indicate mean esti-mates based on multiple independent studies dashed lines are basedon only one Greenndashblue colours are positive grey colours are neg-ative Calculated changes (largestndashsmallest) are determined fromglacial discharge superimposed on a modelled global RCP85 sce-nario (Kwiatkowski et al 2019) pCO2 uptake due to meltwater-induced undersaturation scaled to the Greenland Ice Sheet (Meire etal 2015) computed upwelled NO3 fluxes (assuming 100 utiliza-tion at Redfield ratio Hopwood et al 2018) mean freshwater NO3(Greenland) inventory (Table 3) NO3 anomaly due to upwelling inSermilik Fjord (Cape et al 2019) and contrasting the mean PP forgroups II and IV (Table 1) for a fjord the size of Young Sound

example considering the decadalndashcentennial effects of in-creasing meltwater addition to the Atlantic Ocean or con-versely the seasonal effect of meltwater solely within terres-trial systems One conceptual way of comparing some of thedifferent processes and effects occurring as a result of glacialdischarge is to consider a single biogeochemical cycle on aglobal scale for example the carbon drawdown associatedwith marine primary production (Fig 10)

A net decrease in primary production is predicted overthe 21st century at the Atlantic scale on the order of gt60 Tg C yrminus1 mmminus1 of annual sea-level rise from Green-land due solely to the physical effects of freshwater ad-dition (Kwiatkowski et al 2019) An example of a po-tential negative effect on primary production operating ona much smaller scale would be the retreat of marine-terminating glaciers and the associated loss of NO3 up-welling (Torsvik et al 2019) The effect of switching a mod-est glacier fjord the size of Young Sound from being a higher-productivity marine-terminating glacier fjord environmentto a low-productivity glacier fjord environment receivingrunoff only from land-terminating glaciers (using mean pri-mary production values from Table 1) would be a changeof sim 001 Tg C yrminus1 Conversely potential positive effects ofglacier discharge on primary production can be estimated us-

ing the Redfield ratio (Redfield 1934) to approximate howmuch primary production could be supported by NO3 sup-plied to near-surface waters from meltwater-associated pro-cesses Adding all the NO3 in freshwater around Greenland(Table 3) into the ocean in the absence of any confound-ing physical effects from stratification would be equivalentto primary production of sim 009 Tg C yrminus1 Using the samearbitrary conversion to scale other fluxes the primary pro-duction potentially supported by upwelling of NO3 at Ser-milik (Cape et al 2019) is approximately 013 Tg C yrminus1

and that supported by upwelling of NO3 at 12 large Green-landic marine-terminating systems (Hopwood et al 2018)is approximately 13 Tg C yrminus1 Finally the inorganic CO2drawdown due to pCO2 undersaturation in glacier estuariesaround Greenland is approximately 18 Tg C yrminus1 (Meire etal 2015)

These values provide a rough conceptual framework forevaluating the relative importance of different processes op-erating in parallel but on different spatial scales (Fig 10)Whilst a discussion of glacial weathering processes is be-yond the scope of this review we note that these estimatesof annual C fluxes (Fig 10) are comparable to or largerthan upper estimates of the CO2 drawdownrelease associ-ated with weathering of carbonate silicate and sulfide miner-als in glaciated catchments globally (Jones et al 2002 Tran-ter et al 2002 Torres et al 2017) The implication of thisis that shifts in glacierndashocean inter-connectivity could be im-portant compared to changes in weathering rates in glaciatedcatchments in terms of feedbacks in the C cycle on inter-annual timescales

101 A link between retreating glaciers and harmfulalgal blooms

Shifts between different microbial groups in the ocean canhave profound implications for ecosystem services For ex-ample addition of DOM can induce shifts in the microbialloop to favour bacteria in their competition with phytoplank-ton for macronutrient resources which directly affects themagnitude of CO2 uptake by primary producers (Thingstadet al 2008 Larsen et al 2015) Similarly changing theavailability of Si relative to other macronutrients affects theviability of diatom growth and thus due to the efficiency withwhich diatom frustules sink potentially the efficiency of thebiological carbon pump (Honjo and Manganini 1993 Dug-dale et al 1995)

A particularly concerning hypothesis recently proposedfrom work across Patagonian fjord systems and the first eval-uations of harmful algal bloom (HAB)-associated speciesaround Greenland is that changes in glacier discharge andassociated shifts in stratification and temperature could af-fect HAB occurrence (Richlen et al 2016 Leoacuten-Muntildeoz etal 2018 Joli et al 2018) In the Arctic very little workhas been done to specifically investigate HAB occurrenceand drivers in glacier-discharge-affected regions Yet HAB-

associated species are known to be present in Arctic wa-ters (Lefebvre et al 2016 Richlen et al 2016) includingAlexandrium tamarense which has been implicated as thecause of toxin levels exceeding regulatory limits in scallopsfrom west Greenland (Baggesen et al 2012) and Alexan-drium fundyense cysts of which have been found at lowconcentrations in Disko Bay (Richlen et al 2016) AroundGreenland low temperatures are presently thought to be amajor constraint on HAB development (Richlen et al 2016)Yet increasing meltwater discharge into coastal regionsdrives enhanced stratification and thus directly facilitatesthe development of warm surface waters through summerThis meltwater-driven stratification has been linked to theoccurrence of HAB species including the diatoms Pseudo-nitzschia spp (Joli et al 2018) Thus increasing freshwaterdischarge from Greenland could increase HAB viability indownstream stratified marine environments (Richlen et al2016 Joli et al 2018 Vandersea et al 2018) potentiallywith negative impacts on inshore fisheries

Given the ongoing intensification of climate change andthe interacting effects of different environmental drivers ofprimary production in glacier fjord systems (eg surfacewarming carbonate chemistry light availability stratifica-tion nutrient availability and zooplankton distribution) itis however very challenging to predict future changes onHAB event frequency and intensity Furthermore differ-ent HAB-associated groups (eg toxin-producing diatomand flagellate species) may show opposite responses to thesame environmental perturbation (Wells et al 2015) More-over many known toxin-producing species in the Arcticare mixotrophic further complicating their interactions withother microbial groups (Stoecker and Lavrentyev 2018)Fundamental knowledge gaps clearly remain concerning themechanisms of HAB development and there are practi-cally no time series or studies to date investigating changesspecifically in glaciated Arctic catchments Given the socio-economic importance of glacier-fjord-scale subsistence fish-eries especially around Greenland one priority for futureresearch in the Arctic is to establish to what extent HAB-associated species are likely to benefit from future climatescenarios in regions where freshwater runoff is likely to besubject to pronounced ongoing changes (Baggesen et al2012 Richlen et al 2016 Joli et al 2018)

11 Understanding the role of glaciers alongside othermanifestations of climate change

In order to comprehensively address the questions posed inthis review it is evident that a broader perspective than anarrow focus on freshwater discharge alone and its regionalbiogeochemical effects is required (Fig 10) Freshwater dis-charge is not the sole biogeochemical connection betweenthe glaciers and the ocean (Fig 11) Dust plumes fromproglacial terrain supply glacial flour to the ocean on scales

Figure 11 The approximate spatial scale over which glaciers di-rectly affect different drivers of marine primary production (PP)compared to the likely limiting resources constraining primary pro-duction

of gt 100 km and thus act as an important source of Fe to theocean at high latitudes where other atmospheric dust sourcesare scarce (Prospero et al 2012 Bullard 2013) Similarlyicebergs have long been speculated to act as an importantsource of Fe to the offshore ocean (Hart 1934 Raiswell etal 2008 Lin et al 2011) and induce mixing of the sur-face ocean (Helly et al 2011 Carlson et al 2017) Whilstfreshwater discharge is a driver of biogeochemical changesin nearshore and fjord environments downstream of glaciers(Arimitsu et al 2016) the distant (gt 100 km scale) biogeo-chemical effects of glaciers on the marine environment arelikely dominated by these alternative mechanisms (Fig 11)Furthermore the distal physical effects of adding increas-ingly large volumes of glacier discharge into the Atlanticmay have biogeochemical feedbacks which whilst poorlystudied are potentially far larger than individual regional-scale processes discussed herein (Fig 10) (Kwiatkowski etal 2019)

Discharge-derived effects must also be interpreted in thecontext of other controls on primary production in the high-latitude marine environment Sea-ice properties and particu-larly the timing of its breakup and the duration of the ice-freeseason are a key constraint on the seasonal trend in primaryproduction in the Arctic (Rysgaard et al 1999 Rysgaardand Glud 2007) Similarly whilst discharge affects multi-ple aspects of the three-dimensional water column includ-ing fjord-scale circulation and mixing (Kjeldsen et al 2014Carroll et al 2017) stratification (Meire et al 2016b Oliveret al 2018) and boundary current properties (Sutherland et

al 2009) other changes in the Earth system including windpatterns (Spall et al 2017 Sundfjord et al 2017 Le Bras etal 2018) sea-ice dynamics regional temperature increases(Cook et al 2016) and other freshwater sources (Benetti etal 2019) are driving changes in these parameters on simi-lar spatial and temporal scales (Stocker et al 2013 Hop etal 2019)

Several key uncertainties remain in constraining the roleof glaciers in the marine biogeochemical system Outletglacier fjords are challenging environments in which togather data and there is a persistent deficiency of bothphysical and biogeochemical data within kilometres of largemarine-terminating glacier systems where glacier dischargefirst mixes with ocean properties Subglacial discharge plumemodelling and available data from further downstream canto some extent evade this deficiency for conservative phys-ical (eg salinity and temperature) and chemical (eg noblegases NO3 and PO4) parameters in order to understand mix-ing processes (Mortensen et al 2014 Carroll et al 2017Beaird et al 2018) However the mixing behaviour of non-conservative chemical parameters (eg pH Si and Fe) ismore challenging to deduce from idealized models Further-more the biogeochemical effects of low-frequency high-discharge events and small-scale mixing such as that in-duced around icebergs remain largely unknown There is acritical need to address this deficiency by the deployment ofnew technology to study marine-terminating glacier mixingzones and downstream environments

The uniqueness of individual glacier fjord systems dueto highly variable fjord circulation and geometry is it-self a formidable challenge in scaling up results from Arc-tic field studies to produce a process-based understandingof glacierndashocean interactions A proposed solution whichworks equally well for physical chemical and biologicalperspectives is to focus intensively on a select number ofkey field sites at the landndashocean interface rather than mainlyon large numbers of broadscale summertime-only surveys(Straneo et al 2019) In addition to facilitating long-termtime series focusing in detail on fewer systems facilitatesgreater seasonal coverage to understand the changes in cir-culation and productivity that occur before during and af-ter the melt season However the driving rationale for theselection of key glacier field sites to date was in manycases their contribution to sea-level rise Thus well-studiedsites account for a large fraction of total Arctic glacier dis-charge into the ocean but only represent a small fractionof the glaciated coastline For example around the Green-land coastline the properties of over 200 marine-terminatingglaciers are characterized (Morlighem et al 2017) Yet just5 glaciers (including Helheim in Sermilik Fjord) accountfor 30 of annual combined meltwater and ice dischargefrom Greenland and 15 account for gt 50 (year 2000 dataEnderlin et al 2014) The relative importance of individ-ual glaciers changes when considering longer time periods(eg 1972ndash2018 Mouginot et al 2019) yet irrespective

of the timescale considered a limited number of glaciersaccount for a large fraction of annual discharge Jakob-shavn Isbraelig and Kangerlussuaq for example are among thelargest four contributors to ice discharge around Greenlandover both historical (1972ndash2018) and recent (2000ndash2012)time periods (Enderlin et al 2014 Mouginot et al 2019)Whilst small glaciated catchments such as Kongsfjordenand Young Sound are far less important for sea-level risesimilar ldquosmallrdquo glaciers occupy a far larger fraction of thehigh-latitude coastline and are thus more representative ofglaciated coastline habitat

12 Conclusions

121 Where and when does glacial freshwaterdischarge promote or reduce marine primaryproduction

In the Arctic marine-terminating glaciers are associatedwith the enhanced vertical fluxes of macronutrients whichcan drive summertime phytoplankton blooms throughout themeltwater season

In the Arctic land-terminating glaciers are generally asso-ciated with the local suppression of primary production dueto light limitation and stratification impeding vertical nutri-ent supply from mixing Primary production in Arctic glacierfjords without marine-terminating glaciers is generally lowcompared to other coastal environments

In contrast to the Arctic input of Fe from glaciers aroundthe Southern Ocean is anticipated to have a positive effect onmarine primary production due to the extensive limitation ofprimary production by Fe

In some brackish inshore waters DOM from glaciatedcatchments could enhance bacterial activity at the expenseof phytoplankton but a widespread effect is unlikely due tothe low DOM concentration in freshwater

Glacier discharge reduces the buffering capacity ofglacially modified waters and amplifies the negative effectsof ocean acidification especially in low-productivity sys-tems which negatively affects calcifying organisms

122 How does spatio-temporal variability in glacialdischarge affect marine primary production

Glacier retreat associated with a transition from marine- toland-terminating systems is expected to negatively affectdownstream productivity in the Arctic with long-term in-land retreat also changing the biogeochemical compositionof freshwater

Low-frequency high-discharge events are speculated to beimportant drivers of physical and biogeochemical processesin the marine environment but their occurrence and effectsare poorly constrained

HAB viability may increase in future Arctic glacier fjordsin response to increasing discharge driving enhanced strati-

fication but there are very limited data available to test thishypothesis

A time series in Godtharingbsfjord suggests that on inter-annual timescales fjord-scale primary production is rela-tively stable despite sustained increases in glacier discharge

123 How far-reaching are the effects of glacialdischarge on marine biogeochemistry

Local effects of glaciers (within a few kilometres of the ter-minus or within glacier fjords) include light suppression im-pediment of filter-feeding organisms and influencing the for-aging habits of higher organisms

Mesoscale effects of glaciers (extending tens to hundredsof kilometres from the terminus) include nutrient upwellingFe enrichment of seawater modification of the carbonate sys-tem (both by physical and biological drivers) and enhancedstratification

Remote effects are less certain Beyond the 10ndash100 kmscale over which discharge plumes can be evident othermechanisms of material transfer between glaciers and theocean such as atmospheric deposition of glacial flour andicebergs are likely more important than meltwater (Fig 11)Fully coupled biogeochemical and physical global modelswill be required to fully assess the impacts of increasing dis-charge into the ocean on a pan-Atlantic scale (Fig 10)

Data availability Data sources are cited within the text Forprimary production data see Andersen (1977) Nielsen andHansen (1995) Jensen et al (1999) Nielsen (1999) Levin-sen and Nielsen (2002) Juul-Pedersen et al (2015) Meire etal (2017) Lund-Hansen et al (2018) Hop et al (2002) Iversenand Seuthe (2011) Hodal et al (2012) van de Poll et al (2018)Seifert et al (2019) Smoła et al (2017) Rysgaard et al (1999)Holding et al (2019) Harrison et al (1982) and Reisdorph andMathis (2015) For chemical data and associated fluxes see Frans-son et al (2016) van de Poll et al (2018) Cantoni et al (2019)Cauwet and Sidorov (1996) Emmerton et al (2008) Hessen etal (2010) Hopwood et al (2016 2017 2018) Kanna et al (2018)Cape et al (2019) Hawkings et al (2014 2017) Lund-Hansen etal (2018) Meire et al (2015 2016a) Brown et al (2010) Paulsenet al (2017) Stevenson et al (2017) Statham et al (2008) Bha-tia et al (2010 2013a 2013b) Lawson et al (2014b) Hood etal (2015) Csank et al (2019) Wadham et al (2016) Achterberget al (2018) Marsay et al (2017) Annett et al (2017) Ducklowet al (2017) Tonnard et al (2020) Lippiatt et al (2010) Franssonand Chierici (2019) Vergara-Jara et al (2019) and Kwiatkowski etal (2019) For discharge plume properties see Carroll et al (2016)Halbach et al (2019) Kanna et al (2018) Mankoff et al (2016)Meire et al (2016b) Jackson et al (2017) Bendtsen et al (2015)Beaird et al (2018) and Schaffer et al (2020)

Supplement The supplement related to this article is available on-line at httpsdoiorg105194tc-14-1347-2020-supplement

Author contributions TD coordinated workshop activities and de-signed questions to structure the review paper MJH coordinatedmanuscript writing All authors contributed to writing at least onesection of the review and assisted with the revision of other sectionsDC edited all figures

Competing interests The authors declare that they have no conflictof interest

Acknowledgements The authors thank all conveners and partici-pants of the IASC cross-cutting activity ldquoThe importance of Arcticglaciers for the Arctic marine ecosystemrdquo hosted by the CryosphereWorking GroupNetwork on Arctic Glaciology and the MarineWorking Group IASC funding to support early career scientist at-tendance is gratefully acknowledged Figure 7 and all linear regres-sions were produced in SigmaPlot

Financial support Mark Hopwood was financed by the DFG(award number HO 63211-1) Andy Hodson was supportedby Joint Programming Initiative (JPI-Climate Topic 2 RussianArctic and Boreal Systems) award 71126 and Research Council ofNorway grant 294764 Johnna Holding was supported by MarieCurie grant GrIS-Melt (752325) Lorenz Meire was supportedby the VENI program from the Dutch Research Council (NWOgrant 016Veni192150) Joseacute L Iriarte received support from theFONDECYT 1170174 project Sofia Ribeiro received support fromGeocenter Denmark (project GreenShift) Thorben Dunse wassupported by the Nordforsk-funded project (GreenMAR)

The article processing charges for this open-accesspublication were covered by a ResearchCentre of the Helmholtz Association

Review statement This paper was edited by Evgeny A Podolskiyand reviewed by Jon Hawkings and Kiefer Forsch

References

Achterberg E P Steigenberger S Marsay C M LemoigneF A C Painter S C Baker A R Connelly D P MooreC M Tagliabue A and Tanhua T Iron Biogeochemistry inthe High Latitude North Atlantic Ocean Sci Rep 8 1283httpsdoiorg101038s41598-018-19472-1 2018

Ahlstroslashm A P Petersen D Langen P L Citterio M andBox J E Abrupt shift in the observed runoff from thesouthwestern Greenland ice sheet Sci Adv 3 e1701169httpsdoiorg101126sciadv1701169 2017

Andersen O G N Primary production illumination and hydrog-raphy in Joslashrgen Broslashnlund Fjord North Greenland in Med-delelser om Groslashnland Nyt Nordisk Forlag Koslashbenhavn 1977

Annett A L Skiba M Henley S F Venables H J Mered-ith M P Statham P J and Ganeshram R S Compara-tive roles of upwelling and glacial iron sources in Ryder Bay

coastal western Antarctic Peninsula Mar Chem 176 21ndash33httpsdoiorg101016jmarchem201506017 2015

Annett A L Fitzsimmons J N Seacuteguret M J M LagerstroumlmM Meredith M P Schofield O and Sherrell R M Controlson dissolved and particulate iron distributions in surface watersof the Western Antarctic Peninsula shelf Mar Chem 196 81ndash97 httpsdoiorg101016jmarchem201706004 2017

Arendt K E Nielsen T G Rysgaard S and Tonnesson K Dif-ferences in plankton community structure along the Godthab-sfjord from the Greenland Ice Sheet to offshore waters MarEcol Prog Ser 401 49ndash62 httpsdoiorg103354meps083682010

Arendt K E Dutz J Jonasdottir S H Jung-Madsen SMortensen J Moller E F and Nielsen T G Effects ofsuspended sediments on copepods feeding in a glacial in-fluenced sub-Arctic fjord J Plankton Res 33 1526ndash1537httpsdoiorg101093planktfbr054 2011

Arendt K E Juul-Pedersen T Mortensen J Blicher ME and Rysgaard S A 5-year study of seasonal pat-terns in mesozooplankton community structure in a sub-Arctic fjord reveals dominance of Microsetella norvegica(Crustacea Copepoda) J Plankton Res 35 105ndash120httpsdoiorg101093planktfbs087 2013

Arimitsu M L Piatt J F Madison E N Conaway J S andHillgruber N Oceanographic gradients and seabird prey com-munity dynamics in glacial fjords Fish Oceanogr 21 148ndash169httpsdoiorg101111j1365-2419201200616x 2012

Arimitsu M L Piatt J F and Mueter F Influence of glacierrunoff on ecosystem structure in Gulf of Alaska fjords MarEcol Prog Ser 560 19ndash40 httpsdoiorg103354meps118882016

Arrigo K R and van Dijken G L Continued increases in Arc-tic Ocean primary production Prog Oceanogr 136 60ndash70httpsdoiorg101016jpocean201505002 2015

Arrigo K R van Dijken G L Castelao R M Luo H Renner-malm Aring K Tedesco M Mote T L Oliver H and YagerP L Melting glaciers stimulate large summer phytoplanktonblooms in southwest Greenland waters Geophys Res Lett 446278ndash6285 httpsdoiorg1010022017GL073583 2017

Azetsu-Scott K and Syvitski J P M Influence of melting ice-bergs on distribution characteristics and transport of marine par-ticles in an East Greenland fjord J Geophys Res 104 5321httpsdoiorg1010291998JC900083 1999

Baggesen C Moestrup Oslash and Daugbjer N Molecular phylogenyand toxin profiles of Alexandrium tamarense (Lebour) Balech(Dinophyceae) from the west coast of Greenland HarmfulAlgae 19 108ndash116 httpsdoiorg101016jhal2012060052012

Bamber J L Tedstone A J King M D Howat I M En-derlin E M van den Broeke M R and Noel B Land IceFreshwater Budget of the Arctic and North Atlantic Oceans1 Data Methods and Results J Geophys Res-Ocean 1231827ndash1837 httpsdoiorg1010022017JC013605 2018

Barker J D Sharp M J Fitzsimons S J andTurner R J Abundance and dynamics of dissolvedorganic carbon in glacier systems Arct AntarctAlp Res 38 163ndash172 httpsdoiorg1016571523-0430(2006)38[163aadodo]20co2 2006

Beaird N L Straneo F and Jenkins W Export ofstrongly diluted Greenland meltwater from a ma-jor glacial fjord Geophys Res Lett 43 4163ndash4170httpsdoiorg1010292018GL077000 2018

Beaton A D Cardwell C L Thomas R S Sieben V J LegiretF E Waugh E M Statham P J Mowlem M C and MorganH Lab-on-Chip Measurement of Nitrate and Nitrite for In SituAnalysis of Natural Waters Environ Sci Technol 46 9548ndash9556 httpsdoiorg101021es300419u 2012

Bednaršek N Tarling G A Bakker D C E FieldingS and Feely R A Dissolution Dominating Calcifi-cation Process in Polar Pteropods Close to the Pointof Aragonite Undersaturation PLoS One 9 e109183httpsdoiorg101371journalpone0109183 2014

Bendtsen J Mortensen J and Rysgaard S Seasonal surfacelayer dynamics and sensitivity to runoff in a high Arctic fjord(Young SoundTyrolerfjord 74 N) J Geophys Res-Ocean119 6461ndash6478 httpsdoiorg1010022014JC010077 2014

Bendtsen J Mortensen J Lennert K and Rysgaard SHeat sources for glacial ice melt in a west Greenlandtidewater outlet glacier fjord The role of subglacial fresh-water discharge Geophys Res Lett 42 4089ndash4095httpsdoiorg1010022015GL063846 2015

Benetti M Reverdin G Clarke J S Tynan E Holli-day N P Torres-Valdes S Lherminier P and YashayaevI Sources and distribution of fresh water around CapeFarewell in 2014 J Geophys Res-Ocean 124 9404ndash9416httpsdoiorg1010292019JC015080 2019

Bhatia M P Kujawinski E B Das S B Breier C F Hender-son P B and Charette M A Greenland meltwater as a signifi-cant and potentially bioavailable source of iron to the ocean NatGeosci 6 274ndash278 httpsdoiorg101038ngeo1746 2013a

Bhatia M P Das S B Xu L Charette M A Wadham JL and Kujawinski E B Organic carbon export from theGreenland ice sheet Geochim Cosmochim Acta 109 329ndash344httpsdoiorg101016jgca201302006 2013b

Bhatia M P Das S B Longnecker K Charette MA and Kujawinski E B Molecular characterization ofdissolved organic matter associated with the Greenlandice sheet Geochim Cosmochim Acta 74 3768ndash3784httpsdoiorg101016jgca201003035 2010

Blain S Treguer P Belviso S Bucciarelli E Denis MDesabre S Fiala M Jezequel V M Le Fevre J Mayzaud PMarty J C and Razouls S A biogeochemical study of the is-land mass effect in the context of the iron hypothesis KerguelenIslands Southern Ocean Deep Res Part I 48 163ndash187 2001

Bliss A Hock R and Radic V Global response of glacier runoffto twenty-first century climate change J Geophys Res-EarthSurf 119 717ndash730 2014

Boone W Rysgaard S Carlson D F Meire L Kir-illov S Mortensen J Dmitrenko I Vergeynst L andSejr M K Coastal Freshening Prevents Fjord Bottom Wa-ter Renewal in Northeast Greenland A Mooring StudyFrom 2003 to 2015 Geophys Res Lett 45 2726ndash2733httpsdoiorg1010022017GL076591 2018

Bowie A R van der Merwe P Queacuteroueacute F Trull T FourquezM Planchon F Sarthou G Chever F Townsend A T Ober-nosterer I Salleacutee J-B and Blain S Iron budgets for threedistinct biogeochemical sites around the Kerguelen Archipelago

(Southern Ocean) during the natural fertilisation study KEOPS-2 Biogeosciences 12 4421ndash4445 httpsdoiorg105194bg-12-4421-2015 2015

Box J E Survey of Greenland instrumental temperaturerecords 1873ndash2001 Int J Climatol 22 1829ndash1847httpsdoiorg101002joc852 2002

Boyle E A Edmond J M and Sholkovitz E R Mecha-nism of iron removal in estuaries Geochim Cosmochim Acta41 1313ndash1324 httpsdoiorg1010160016-7037(77)90075-81977

Brown G H Sharp M J Tranter M Gurnell A M andNienow P W Impact of post-mixing chemical reactions onthe major ion chemistry of bulk meltwaters draining the hautglacier drsquoarolla valais Switzerland Hydrol Process 8 465ndash480 httpsdoiorg101002hyp3360080509 1994

Brown M T Lippiatt S M and Bruland K W Dis-solved aluminum particulate aluminum and silicic acid innorthern Gulf of Alaska coastal waters Glacialriverine in-puts and extreme reactivity Mar Chem 122 160ndash175httpsdoiorg101016jmarchem201004002 2010

Browning T J Achterberg E P Rapp I Engel A BertrandE M Tagliabue A and Moore C M Nutrient co-limitationat the boundary of an oceanic gyre Nature 551 242ndash246httpsdoiorg101038nature24063 2017

Bucciarelli E Blain S and Treguer P Iron and manganese in thewake of the Kerguelen Islands (Southern Ocean) Mar Chem73 21ndash36 2001

Bullard J E Contemporary glacigenic inputs to thedust cycle Earth Surf Process Landf 38 71ndash89httpsdoiorg101002esp3315 2013

Cable S Christiansen H H Westergaard-Nielsen A KroonA and Elberling B Geomorphological and cryostratigraphicalanalyses of the Zackenberg Valley NE Greenland and signifi-cance of Holocene alluvial fans Geomorphology 303 504ndash523httpsdoiorg101016jgeomorph201711003 2018

Calleja M L Kerherveacute P Bourgeois S Kedra M LeynaertA Devred E Babin M and Morata N Effects of increaseglacier discharge on phytoplankton bloom dynamics and pelagicgeochemistry in a high Arctic fjord Prog Oceanogr 159 195ndash210 httpsdoiorg101016jpocean201707005 2017

Cantoni C Hopwood M Clarke J Chiggiato J AchterbergE P and Cozzi S Hydrological biogeochemical and carbon-ate system data in coastal waters and in glacier drainage systemsin Kongsfjorden (Svalbard) during JulyndashAugust 2016 Data setPANGAEA httpsdoiorg101594PANGAEA904171 2019

Cape M R Straneo F Beaird N Bundy R M and CharetteM A Nutrient release to oceans from buoyancy-driven up-welling at Greenland tidewater glaciers Nat Geosci 12 34ndash39httpsdoiorg101038s41561-018-0268-4 2019

Carlson D F and Rysgaard S Adapting open-source drone au-topilots for real-time iceberg observations MethodsX 5 1059ndash1072 httpsdoiorg101016jmex201809003 2018

Carlson D F Boone W Meire L Abermann J and RysgaardS Bergy Bit and Melt Water Trajectories in Godtharingbsfjord (SWGreenland) Observed by the Expendable Ice Tracker Front MarSci 4 276 httpsdoiorg103389fmars201700276 2017

Carroll D Sutherland D A Shroyer E L Nash J DCatania G A and Stearns L A Modeling TurbulentSubglacial Meltwater Plumes Implications for Fjord-Scale

Buoyancy-Driven Circulation J Phys Oceanogr 45 2169ndash2185 httpsdoiorg101175JPO-D-15-00331 2015

Carroll D Sutherland D A Hudson B Moon T Catania GA Shroyer E L Nash J D Bartholomaus T C FeliksonD Stearns L A Noeumll B P Y and van den Broeke M RThe impact of glacier geometry on meltwater plume structureand submarine melt in Greenland fjords Geophys Res Lett 439739ndash9748 httpsdoiorg1010022016GL070170 2016

Carroll D Sutherland D A Shroyer E L Nash J D CataniaG A and Stearns L A Subglacial discharge-driven renewalof tidewater glacier fjords J Geophys Res-Ocean 122 6611ndash6629 httpsdoiorg1010022017JC012962 2017

Carroll D Sutherland D A Curry B Nash J D ShroyerE L Catania G A Stearns L A Grist J P LeeC M and de Steur L Subannual and Seasonal Vari-ability of Atlantic-Origin Waters in Two Adjacent WestGreenland Fjords J Geophys Res-Ocean 123 6670ndash6687httpsdoiorg1010292018JC014278 2018

Cauvy-Fraunieacute S and Dangles O A global synthesis of biodiver-sity responses to glacier retreat Nat Ecol Evol 3 1675ndash1685httpsdoiorg101038s41559-019-1042-8 2019

Cauwet G and Sidorov I The biogeochemistry of Lena Riverorganic carbon and nutrients distribution Mar Chem 53 211ndash227 httpsdoiorg1010160304-4203(95)00090-9 1996

Charette M A and Sholkovitz E R Oxidative precipita-tion of groundwater-derived ferrous iron in the subterraneanestuary of a coastal bay Geophys Res Lett 29 85httpsdoiorg1010292001GL014512 2002

Charette M A Lam P J Lohan M C Kwon E Y HatjeV Jeandel C Shiller A M Cutter G A Thomas ABoyd P W Homoky W B Milne A Thomas H Ander-sson P S Porcelli D Tanaka T Geibert W Dehairs Fand Garcia-Orellana J Coastal ocean and shelf-sea biogeo-chemical cycling of trace elements and isotopes lessons learnedfrom GEOTRACES Philos Trans R Soc A 374 20160076httpsdoiorg101098rsta20160076 2016

Chierici M and Fransson A Calcium carbonate saturation inthe surface water of the Arctic Ocean undersaturation infreshwater influenced shelves Biogeosciences 6 2421ndash2431httpsdoiorg105194bg-6-2421-2009 2009

Chu V W Smith L C Rennermalm A K ForsterR R Box J E and Reeh N Sediment plume re-sponse to surface melting and supraglacial lake drainageson the Greenland ice sheet J Glaciol 55 1072ndash1082httpsdoiorg103189002214309790794904 2009

Chu V W Smith L C Rennermalm A K Forster R R andBox J E Hydrologic controls on coastal suspended sedimentplumes around the Greenland Ice Sheet The Cryosphere 6 1ndash19 httpsdoiorg105194tc-6-1-2012 2012

Comeau S Gorsky G Jeffree R Teyssieacute J-L and GattusoJ-P Impact of ocean acidification on a key Arctic pelagicmollusc (Limacina helicina) Biogeosciences 6 1877ndash1882httpsdoiorg105194bg-6-1877-2009 2009

Comeau S Jeffree R Teyssieacute J-L and Gattuso J-P Re-sponse of the Arctic Pteropod Limacina helicina to Pro-jected Future Environmental Conditions PLoS One 5 e11362httpsdoiorg101371journalpone0011362 2010

Cook J Oreskes N Doran P T Anderegg W R L Ver-heggen B Maibach E W Carlton J S Lewandowsky S

Skuce A G and Green S A Consensus on consensus a syn-thesis of consensus estimates on human-caused global warm-ing Environ Res Lett 11 48002 httpsdoiorg1010881748-9326114048002 2016

Cooper L W McClelland J W Holmes R M RaymondP A Gibson J J Guay C K and Peterson B J Flow-weighted values of runoff tracers (δ18O DOC Ba alkalin-ity) from the six largest Arctic rivers Geophys Res Lett 35L18606 httpsdoiorg1010292008GL035007 2008

Coupel P Ruiz-Pino D Sicre M A Chen J F LeeS H Schiffrine N Li H L and Gascard J CThe impact of freshening on phytoplankton production inthe Pacific Arctic Ocean Prog Oceanogr 131 113ndash125httpsdoiorg101016jpocean201412003 2015

Cowton T Slater D Sole A Goldberg D and Nienow PModeling the impact of glacial runoff on fjord circulation andsubmarine melt rate using a new subgrid-scale parameteriza-tion for glacial plumes J Geophys Res-Ocean 120 796ndash812httpsdoiorg1010022014JC010324 2015

Crusius J Schroth A W Gasso S Moy C M Levy R Cand Gatica M Glacial flour dust storms in the Gulf of AlaskaHydrologic and meteorological controls and their importance asa source of bioavailable iron Geophys Res Lett 38 06602httpsdoiorg1010292010gl046573 2011

Crusius J Schroth A W Resing J A Cullen J andCampbell R W Seasonal and spatial variabilities in north-ern Gulf of Alaska surface water iron concentrations drivenby shelf sediment resuspension glacial meltwater a Yaku-tat eddy and dust Global Biogeochem Cy 31 942ndash960httpsdoiorg1010022016GB005493 2017

Csank A Z Czimczik C I Xu X and Welker J MSeasonal patterns of riverine carbon sources and export inNW Greenland J Geophys Res-Biogeosci 124 840ndash856httpsdoiorg1010292018JG004895 2019

Cushman-Roisin B Asplin L and Svendsen H Up-welling in broad fjords Cont Shelf Res 14 1701ndash1721httpsdoiorg1010160278-4343(94)90044-2 1994

De Andreacutes E Slater D A Straneo F Otero J Das Sand Navarro F Surface emergence of glacial plumes de-termined by fjord stratification The Cryosphere Discusshttpsdoiorg105194tc-2019-264 in review 2020

Debaar H J W VonLiebig Law of the minimum and plank-ton ecology (1899ndash1991) Prog Oceanogr 33 347ndash386httpsdoiorg1010160079-6611(94)90022-1 1994

Dickson A G and Millero F J A comparison of the equi-librium constants for the dissociation of carbonic acid inseawater media Deep Sea Res Part A 34 1733ndash1743httpsdoiorg1010160198-0149(87)90021-5 1987

Dittmar T and Kattner G The biogeochemistry of the river andshelf ecosystem of the Arctic Ocean a review Mar Chem83 103ndash120 httpsdoiorg101016S0304-4203(03)00105-12003

Doney S C Fabry V J Feely R A and Kley-pas J A Ocean Acidification The Other CO2Problem Ann Rev Mar Sci 1 169ndash192httpsdoiorg101146annurevmarine010908163834 2009

Ducklow H W Vernet M and Prezelin B Dissolved inorganicnutrients including 5 macro nutrients silicate phosphatenitrate nitrite and ammonium from water column bottle sam-

ples collected during annual cruise along western AntarcticPeninsula 1991ndash2019 ver 8 Environmental Data Initiativehttpsdoiorg106073pasta9611089f65d3cbd0801204d174eee9472017

Dugdale R C Wilkerson F P and Minas H J The role of asilicate pump in driving new production Deep Res I 42 697ndash719 1995

Egge J K and Aksnes D L Silicate as regulating nutrient inphytoplankton competition Mar Ecol Prog Ser 83 281ndash2891992

Egge J K and Heimdal B R Blooms of phytoplanktonincluding Emiliania huxleyi (Haptophyta) Effects of nutri-ent supply in different N P ratios Sarsia 79 333ndash348httpsdoiorg10108000364827199410413565 1994

Ellegaard M and Ribeiro S The long-term persistence of phyto-plankton resting stages in aquatic lsquoseed banksrsquo Biol Rev 93166ndash183 httpsdoiorg101111brv12338 2018

Emmerton C A Lesack L F W and Vincent W F Nutrient andorganic matter patterns across the Mackenzie River estuary andshelf during the seasonal recession of sea-ice J Mar Syst 74741ndash755 httpsdoiorg101016jjmarsys200710001 2008

Enderlin E M Howat I M Jeong S Noh M-J van AngelenJ H and van den Broeke M R An improved mass budgetfor the Greenland ice sheet Geophys Res Lett 41 866ndash872httpsdoiorg1010022013GL059010 2014

Enderlin E M Carrigan C J Kochtitzky W H Cuadros AMoon T and Hamilton G S Greenland iceberg melt variabil-ity from high-resolution satellite observations The Cryosphere12 565ndash575 httpsdoiorg105194tc-12-565-2018 2018

Ericson Y Falck E Chierici M Fransson A and KristiansenS Marine CO2 system variability in a high arctic tidewater-glacier fjord system Tempelfjorden Svalbard Cont Shelf Res181 1ndash13 httpsdoiorg101016jcsr201904013 2019

Etherington L L and Hooge P N Oceanography of Glacier BayAlaska Implications for biological patterns in a glacial fjord es-tuary Estuar Coast 30 927ndash944 2007

Evans W Mathis J T and Cross J N Calcium carbonate cor-rosivity in an Alaskan inland sea Biogeosciences 11 365ndash379httpsdoiorg105194bg-11-365-2014 2014

Fransson A and Chierici M Marine CO2 system data forthe Svalbard fjord Kongsfjorden and the West-Spitsbergenshelf in July 2012ndash2014 Data set] Norwegian Polar Institutehttpsdoiorg1021334npolar2019e53eae53 2019

Fransson A Chierici M Nomura D Granskog M A Kris-tiansen S Martma T and Nehrke G Effect of glacialdrainage water on the CO2 system and ocean acidificationstate in an Arctic tidewater-glacier fjord during two con-trasting years J Geophys Res-Ocean 120 2413ndash2429httpsdoiorg1010022014JC010320 2015

Fransson A Chierici M Hop H Findlay H S Kris-tiansen S and Wold A Late winter-to-summer changein ocean acidification state in Kongsfjorden with implica-tions for calcifying organisms Polar Biol 39 1841ndash1857httpsdoiorg101007s00300-016-1955-5 2016

Fried M J Catania G A Stearns L A Sutherland D ABartholomaus T C Shroyer E and Nash J Reconcilingdrivers of seasonal terminus advance and retreat at 13 centralwest Greenland tidewater glaciers J Geophys Res-Earth 1231590ndash1607 2018

Fuentes V Alurralde G Meyer B Aguirre G E Canepa AWoumllfl A-C Hass C H Williams G N and Schloss I RGlacial melting an overlooked threat to Antarctic krill Sci Rep6 27234 httpsdoiorg101038srep27234 2016

Gerringa L J A Alderkamp A-C Laan P Thuroczy C-E De Baar H J W Mills M M van Dijken G Lvan Haren H and Arrigo K R Iron from melting glaciersfuels the phytoplankton blooms in Amundsen Sea (SouthernOcean) Iron biogeochemistry Deep Res Part Ii 71ndash76 16ndash31httpsdoiorg101016jdsr2201203007 2012

Gilbert G L OrsquoNeill H B Nemec W Thiel C Chris-tiansen H H and Buylaert J-P Late Quaternary sedi-mentation and permafrost development in a Svalbard fjord-valley Norwegian high Arctic Sedimentology 65 2531ndash2558httpsdoiorg101111sed12476 2018

Gladish C V Holland D M Rosing-Asvid A Behrens JW and Boje J Oceanic Boundary Conditions for Jakob-shavn Glacier Part I Variability and Renewal of Ilulis-sat Icefjord Waters 2001ndash14 J Phys Oceanogr 45 3ndash32httpsdoiorg101175JPO-D-14-00441 2014

Gledhill M and Buck K N The organic complexation of ironin the marine environment a review Front Microbiol 3 69httpsdoiorg103389fmicb201200069 2012

Glud R N Risgaard-Petersen M Thamdrup B Fossing Hand Rysgaard S Benthic carbon mineralization in a high-Arcticsound (Young Sound NE Greenland) Mar Ecol Prog Ser 20659ndash71 httpsdoiorg103354meps206059 2000

Gonzaacutelez-Bergonzoni I L J K Anders M Frank L ErikJ and A D T Small birds big effects the little auk (Allealle) transforms high Arctic ecosystems P Roy Soc B 28420162572 httpsdoiorg101098rspb20162572 2017

Grand M M Clinton-Bailey G S Beaton A D SchaapA M Johengen T H Tamburri M N Connelly D PMowlem M C and Achterberg E P A Lab-On-Chip Phos-phate Analyzer for Long-term In Situ Monitoring at Fixed Ob-servatories Optimization and Performance Evaluation in Estu-arine and Oligotrophic Coastal Waters Front Mar Sci 4 255httpsdoiorg103389fmars201700255 2017

Grand M M Laes-Huon A Fietz S Resing J A Obata HLuther G W Tagliabue A Achterberg E P Middag RTovar-Saacutenchez A and Bowie A R Developing AutonomousObserving Systems for Micronutrient Trace Metals Front MarSci 6 35 httpsdoiorg103389fmars201900035 2019

Halbach L Vihtakari M Duarte P Everett A Granskog MA Hop H Kauko H M Kristiansen S Myhre P I PavlovA K Pramanik A Tatarek A Torsvik T Wiktor J MWold A Wulff A Steen H and Assmy P Tidewater Glaciersand Bedrock Characteristics Control the Phytoplankton GrowthEnvironment in a Fjord in the Arctic Front Mar Sci 6 254httpsdoiorg103389fmars201900254 2019

Harrison W G Platt T and Irwin B Primary Production andNutrient Assimilation by Natural Phytoplankton Populations ofthe Eastern Canadian Arctic Can J Fish Aquat Sci 39 335ndash345 httpsdoiorg101139f82-046 1982

Hart T J Discovery Reports Discov Reports VIII 1ndash268 1934Hawkings J Wadham J Tranter M Telling J Bagshaw E

Beaton A Simmons S-L Chandler D Tedstone A andNienow P The Greenland Ice Sheet as a hot spot of phosphorus

weathering and export in the Arctic Global Biogeochem Cy30 191ndash210 httpsdoiorg1010022015GB005237 2016

Hawkings J R Wadham J L Tranter M Raiswell R Ben-ning L G Statham P J Tedstone A Nienow P Lee Kand Telling J Ice sheets as a significant source of highly reac-tive nanoparticulate iron to the oceans Nat Commun 5 3929httpsdoiorg101038ncomms4929 2014

Hawkings J R Wadham J L Benning L G Hendry K RTranter M Tedstone A Nienow P and Raiswell R Icesheets as a missing source of silica to the polar oceans Nat Com-mun 8 14198 httpsdoiorg101038ncomms14198 2017

Hegseth E N and Tverberg V Effect of Atlantic water inflowon timing of the phytoplankton spring bloom in a high Arcticfjord (Kongsfjorden Svalbard) J Mar Syst 113ndash114 94ndash105httpsdoiorg101016jjmarsys201301003 2013

Helly J J Kaufmann R S Stephenson Jr G R and VernetM Cooling dilution and mixing of ocean water by free-driftingicebergs in the Weddell Sea Deep Res Part I 58 1346ndash1363httpsdoiorg101016jdsr2201011010 2011

Hendry K R Huvenne V A I Robinson L F Annett ABadger M Jacobel A W Ng H C Opher J Picker-ing R A Taylor M L Bates S L Cooper A Cush-man G G Goodwin C Hoy S Rowland G SamperizA Williams J A Achterberg E P Arrowsmith C Brear-ley J A Henley S F Krause J W Leng M J Li TMcManus J F Meredith M P Perkins R and Wood-ward E M S The biogeochemical impact of glacial meltwa-ter from Southwest Greenland Prog Oceanogr 176 102126httpsdoiorg101016jpocean2019102126 2019

Henson S Le Moigne F and Giering S Drivers of Carbon Ex-port Efficiency in the Global Ocean Global Biogeochem Cy33 891ndash903 httpsdoiorg1010292018GB006158 2019

Hessen D O Carroll J Kjeldstad B Korosov A A PetterssonL H Pozdnyakov D and Soslashrensen K Input of organic carbonas determinant of nutrient fluxes light climate and productivityin the Ob and Yenisey estuaries Estuar Coast Shelf Sci 88 53-62 httpsdoiorg101016jecss201003006 2010

Hewitt I J Subglacial Plumes Annu Rev Fluid Mech 52145ndash169 httpsdoiorg101146annurev-fluid-010719-0602522020

Hodal H Falk-Petersen S Hop H Kristiansen S and ReigstadM Spring bloom dynamics in Kongsfjorden Svalbard nutri-ents phytoplankton protozoans and primary production PolarBiol 35 191ndash203 httpsdoiorg101007s00300-011-1053-72012

Hodson A Mumford P and Lister D Suspended sediment andphosphorus in proglacial rivers bioavailability and potential im-pacts upon the P status of ice-marginal receiving waters Hy-drol Process 18 2409ndash2422 httpsdoiorg101002hyp14712004

Hodson A Nowak A and Christiansen H Glacial andperiglacial floodplain sediments regulate hydrologic transfer ofreactive iron to a high arctic fjord Hydrol Process 30 1219ndash1229 httpsdoiorg101002hyp10701 2016

Hodson A Nowak A Sabacka M Jungblut A NavarroF Pearce D Aacutevila-Jimeacutenez M L Convey P and VieiraG Climatically sensitive transfer of iron to maritime Antarc-tic ecosystems by surface runoff Nat Commun 8 14499httpsdoiorg101038ncomms14499 2017

Hodson A J Mumford P N Kohler J and Wynn PM The High Arctic glacial ecosystem New insightsfrom nutrient budgets Biogeochemistry 72 233ndash256httpsdoiorg101007s10533-004-0362-0 2005

Houmlfer J Giesecke R Hopwood M J Carrera V Alarcoacuten Eand Gonzaacutelez H E The role of water column stability andwind mixing in the productionexport dynamics of two bays inthe Western Antarctic Peninsula Prog Oceanogr 174 105ndash116httpsdoiorg101016jpocean201901005 2019

Holding J M Duarte C M Delgado-Huertas A Soetaert KVonk J E Agustiacute S Wassmann P and Middelburg J J Au-tochthonous and allochthonous contributions of organic carbonto microbial food webs in Svalbard fjords Limnol Oceanogr62 1307ndash1323 httpsdoiorg101002lno10526 2017

Holding J M Markager S Juul-Pedersen T Paulsen M LMoslashller E F Meire L and Sejr M K Seasonal and spatialpatterns of primary production in a high-latitude fjord affectedby Greenland Ice Sheet run-off Biogeosciences 16 3777ndash3792httpsdoiorg105194bg-16-3777-2019 2019

Holmes R M McClelland J W Peterson B J Tank S EBulygina E Eglinton T I Gordeev V V Gurtovaya T YRaymond P A Repeta D J Staples R Striegl R G Zhuli-dov A V and Zimov S A Seasonal and Annual Fluxes ofNutrients and Organic Matter from Large Rivers to the Arc-tic Ocean and Surrounding Seas Estuar Coast 35 369ndash382httpsdoiorg101007s12237-011-9386-6 2011

Honjo S and Manganini S J Annual biogenic particle fluxesto the interior of the North Atlantic Ocean studied at 34 N21W and 48 N 21W Deep Sea Res Part II 40 587ndash607httpsdoiorg1010160967-0645(93)90034-K 1993

Hood E and Berner L Effects of changing glacial cover-age on the physical and biogeochemical properties of coastalstreams in southeastern Alaska J Geophys Res 114 G03001httpsdoiorg1010292009jg000971 2009

Hood E and Scott D Riverine organic matter and nutrients insoutheast Alaska affected by glacial coverage Nat Geosci 1583ndash587 httpsdoiorg101038ngeo280 2008

Hood E Fellman J Spencer R G M Hernes P J EdwardsR DrsquoAmore D and Scott D Glaciers as a source of ancientand labile organic matter to the marine environment Nature 4621044ndash1047 httpsdoiorg101038nature08580 2009

Hood E Battin T J Fellman J Orsquoneel S and Spencer R GM Storage and release of organic carbon from glaciers and icesheets Nat Geosci 8 91ndash96 httpsdoiorg101038ngeo23312015

Hop H Pearson T Hegseth E N Kovacs K M WienckeC Kwasniewski S Eiane K Mehlum F Gulliksen BWlodarska-Kowalczuk M Lydersen C Weslawski J MCochrane S Gabrielsen G W Leakey R J G Loslashnne OJ Zajaczkowski M Falk-Petersen S Kendall M WaumlngbergS-Aring Bischof K Voronkov A Y Kovaltchouk N A Wik-tor J Poltermann M Prisco G Papucci C and Gerland SThe marine ecosystem of Kongsfjorden Svalbard Polar Res 21167ndash208 2002

Hop H Assmy P Wold A Sundfjord A Daase M DuarteP Kwasniewski S Gluchowska M Wiktor J M Tatarek AWiktor J Kristiansen S Fransson A Chierici M and Vih-takari M Pelagic Ecosystem Characteristics Across the AtlanticWater Boundary Current From Rijpfjorden Svalbard to the Arc-

tic Ocean During Summer (2010ndash2014) Front Mar Sci 6 181httpsdoiorg103389fmars201900181 2019

Hoppe H-G Phosphatase activity in the sea Hydrobiologia 493187ndash200 httpsdoiorg101023A1025453918247 2003

Hopwood M J Connelly D P Arendt K E Juul-Pedersen T Stinchcombe M C Meire L Espos-ito M and Krishna R Seasonal changes in Fe alonga glaciated Greenlandic fjord Front Earth Sci 4 15httpsdoiorg103389feart201600015 2016

Hopwood M J Cantoni C Clarke J S Cozzi S andAchterberg E P The heterogeneous nature of Fe deliveryfrom melting icebergs Geochem Perspect Lett 3 200ndash209httpsdoiorg107185geochemlet1723 2017

Hopwood M J Carroll D Browning T J Meire LMortensen J Krisch S and Achterberg E P Non-linear re-sponse of summertime marine productivity to increased melt-water discharge around Greenland Nat Commun 9 3256httpsdoiorg101038s41467-018-05488-8 2018

Howe J A Austin W E N Forwick M Paetzel MHarland R and Cage A G Fjord systems and archivesa review Geol Soc London Spec Publ 344 5ndash15httpsdoiorg101144SP3442 2010

Hudson B Overeem I McGrath D Syvitski J P MMikkelsen A and Hasholt B MODIS observed increase in du-ration and spatial extent of sediment plumes in Greenland fjordsThe Cryosphere 8 1161ndash1176 httpsdoiorg105194tc-8-1161-2014 2014

Huss M and Hock R Global-scale hydrological response tofuture glacier mass loss Nat Clim Change 8 135ndash140httpsdoiorg101038s41558-017-0049-x 2018

Hyacinthe C and Van Cappellen P An authigenic iron phosphatephase in estuarine sediments composition formation and chem-ical reactivity Mar Chem 91 227ndash251 2004

Iriarte J L Pantoja S and Daneri G OceanographicProcesses in Chilean Fjords of Patagonia From smallto large-scale studies Prog Oceanogr 129 1ndash7httpsdoiorg101016jpocean201410004 2014

Iversen K R and Seuthe L Seasonal microbial processes ina high-latitude fjord (Kongsfjorden Svalbard) I Heterotrophicbacteria picoplankton and nanoflagellates Polar Biol 34 731ndash749 httpsdoiorg101007s00300-010-0929-2 2011

Iversen M H and Robert M L Ballasting effects ofsmectite on aggregate formation and export from a nat-ural plankton community Mar Chem 175 18ndash27httpsdoiorg101016jmarchem201504009 2015

Jackson R H Straneo F and Sutherland D A Exter-nally forced fluctuations in ocean temperature at Greenlandglaciers in non-summer months Nat Geosci 7 503ndash508httpsdoiorg101038ngeo2186 2014

Jackson R H Shroyer E L Nash J D Sutherland D A Car-roll D Fried M J Catania G A Bartholomaus T C andStearns L A Near-glacier surveying of a subglacial dischargeplume Implications for plume parameterizations Geophys ResLett 44 6886ndash6894 httpsdoiorg1010022017GL0736022017

Jackson R H Lentz S J and Straneo F The Dynamics of ShelfForcing in Greenlandic Fjords J Phys Oceanogr 48 2799ndash2827 httpsdoiorg101175JPO-D-18-00571 2018

Jenkins A Convection-Driven Melting near the Grounding Linesof Ice Shelves and Tidewater Glaciers J Phys Oceanogr 412279ndash2294 httpsdoiorg101175JPO-D-11-031 2011

Jensen H M Pedersen L Burmeister A and WindingHansen B Pelagic primary production during summer along65 to 72 N off West Greenland Polar Biol 21 269ndash278httpsdoiorg101007s003000050362 1999

Johnson H L Muumlnchow A Falkner K K and MellingH Ocean circulation and properties in Petermann FjordGreenland J Geophys Res-Ocean 116 C01003httpsdoiorg1010292010JC006519 2011

Joli N Gosselin M Ardyna M Babin M Onda D FTremblay J-Eacute and Lovejoy C Need for focus on mi-crobial species following ice melt and changing freshwa-ter regimes in a Janus Arctic Gateway Sci Rep 8 9405httpsdoiorg101038s41598-018-27705-6 2018

Jones I W Munhoven G Tranter M Huybrechts P and SharpM J Modelled glacial and non-glacial HCOminus3 Si and Ge fluxessince the LGM little potential for impact on atmospheric CO2concentrations and a potential proxy of continental chemical ero-sion the marine GeSi ratio Global Planet Chang 33 139ndash153 httpsdoiorg101016S0921-8181(02)00067-X 2002

Jouvet G Weidmann Y Kneib M Detert M SeguinotJ Sakakibara D and Sugiyama S Short-lived icespeed-up and plume water flow captured by a VTOLUAV give insights into subglacial hydrological system ofBowdoin Glacier Remote Sens Environ 217 389ndash399httpsdoiorg101016jrse201808027 2018

Juul-Pedersen T Arendt K E Mortensen J Blicher M ESoslashgaard D and Rysgaard S Seasonal and interannual phy-toplankton production in a sub-Arctic tidewater outlet glacierfjord SW Greenland Mar Ecol Prog Ser 524 27ndash38httpsdoiorg103354meps11174 2015

Kanna N Sugiyama S Ohashi Y Sakakibara D Fuka-machi Y and Nomura D Upwelling of macronutrientsand dissolved inorganic carbon by a subglacial fresh-water driven plume in Bowdoin Fjord northwesternGreenland J Geophys Res-Biogeosci 123 1666ndash1682httpsdoiorg1010292017JG004248 2018

Kjeldsen K K Mortensen J Bendtsen J Petersen D LennertK and Rysgaard S Ice-dammed lake drainage cools andraises surface salinities in a tidewater outlet glacier fjordwest Greenland J Geophys Res-Surf 119 1310ndash1321httpsdoiorg1010022013JF003034 2014

Klunder M B Bauch D Laan P de Baar H J W van HeuvenS and Ober S Dissolved iron in the Arctic shelf seas andsurface waters of the central Arctic Ocean Impact of Arc-tic river water and ice-melt J Geophys Res 117 C01027httpsdoiorg1010292011jc007133 2012

Knutz P C Sicre M-A Ebbesen H Christiansen Sand Kuijpers A Multiple-stage deglacial retreat of thesouthern Greenland Ice Sheet linked with Irminger Cur-rent warm water transport Paleoceanography 26 PA3204httpsdoiorg1010292010PA002053 2011

Kohfeld K E and Harrison S P DIRTMAP the ge-ological record of dust Earth-Science Rev 54 81ndash114httpsdoiorg101016S0012-8252(01)00042-3 2001

Koziorowska K Kulinski K and Pempkowiak J Depositionreturn flux and burial rates of nitrogen and phosphorus in the

sediments of two high-Arctic fjords Oceanologia 60 431ndash445httpsdoiorg101016joceano201805001 2018

Krawczyk D W Witkowski A Juul-Pedersen T Arendt K EMortensen J and Rysgaard S Microplankton succession in aSW Greenland tidewater glacial fjord influenced by coastal in-flows and run-off from the Greenland Ice Sheet Polar Biol 381515ndash1533 httpsdoiorg101007s00300-015-1715-y 2015

Krawczyk D W Meire L Lopes C Juul-Pedersen TMortensen J Li C L and Krogh T Seasonal succession dis-tribution and diversity of planktonic protists in relation to hy-drography of the Godtharingbsfjord system (SW Greenland) PolarBiol 41 2033ndash2052 httpsdoiorg101007s00300-018-2343-0 2018

Kumar V Tiwari M and Rengarajan R Warming in the Arc-tic Captured by productivity variability at an Arctic Fjordover the past two centuries PLoS One 13 e0201456httpsdoiorg101371journalpone0201456 2018

Kwiatkowski L Naar J Bopp L Aumont O Defrance D andCouespel D Decline in Atlantic primary production acceleratedby Greenland ice sheet melt Geophys Res Lett 46 11347ndash11357 httpsdoiorg1010292019GL085267 2019

Laidre K L Twila M Hauser D D W McGovern R Heide-Joslashrgensen M P Rune D and Hudson B Use of glacial frontsby narwhals (Monodon monoceros) in West Greenland BiolLett 12 20160457 httpsdoiorg101098rsbl201604572016

Lam P J and Bishop J K B The continental margin is a keysource of iron to the HNLC North Pacific Ocean Geophys ResLett 35 L07608 httpsdoiorg1010292008gl033294 2008

Langen P L Mottram R H Christensen J H Boberg FRodehacke C B Stendel M van As D Ahlstroslashm AP Mortensen J Rysgaard S Petersen D Svendsen KH Aethalgeirsdoacutettir G and Cappelen J Quantifying energyand mass fluxes controlling Godtharingbsfjord freshwater input ina 5-km simulation (1991ndash2012) J Climate 28 3694ndash3713httpsdoiorg101175JCLI-D-14-002711 2015

Larsen A Egge J K Nejstgaard J C Di Capua I ThyrhaugR Bratbak G and Thingstad T F Contrasting response tonutrient manipulation in Arctic mesocosms are reproduced bya minimum microbial food web model Limnol Oceanogr 60360ndash374 httpsdoiorg101002lno10025 2015

Lawson E C Bhatia M P Wadham J L and Kujawinski EB Continuous Summer Export of Nitrogen-Rich Organic Matterfrom the Greenland Ice Sheet Inferred by Ultrahigh ResolutionMass Spectrometry Environ Sci Technol 48 14248ndash14257httpsdoiorg101021es501732h 2014a

Lawson E C Wadham J L Tranter M Stibal M Lis G PButler C E H Laybourn-Parry J Nienow P Chandler Dand Dewsbury P Greenland Ice Sheet exports labile organiccarbon to the Arctic oceans Biogeosciences 11 4015ndash4028httpsdoiorg105194bg-11-4015-2014 2014b

Le Bras I A-A Straneo F Holte J and Holliday N P Sea-sonality of Freshwater in the East Greenland Current SystemFrom 2014 to 2016 J Geophys Res-Ocean 123 8828ndash8848httpsdoiorg1010292018JC014511 2018

Lecavalier B S Fisher D A Milne G A Vinther B MTarasov L Huybrechts P Lacelle D Main B Zheng JBourgeois J and Dyke A S High Arctic Holocene tem-perature record from the Agassiz ice cap and Greenland ice

sheet evolution P Natl Acad Sci USA 114 5952ndash5957httpsdoiorg101073pnas1616287114 2017

Lefebvre K A Quakenbush L Frame E Huntington KB Sheffield G Stimmelmayr R Bryan A Kendrick PZiel H Goldstein T Snyder J A Gelatt T GullandF Dickerson B and Gill V Prevalence of algal toxinsin Alaskan marine mammals foraging in a changing arc-tic and subarctic environment Harmful Algae 55 13ndash24httpsdoiorg101016jhal201601007 2016

Le Fouest V Babin M and Tremblay J-Eacute The fate of river-ine nutrients on Arctic shelves Biogeosciences 10 3661ndash3677httpsdoiorg105194bg-10-3661-2013 2013

Le Moigne F A C Henson S A Cavan E GeorgesC Pabortsava K Achterberg E P Ceballos-Romero EZubkov M and Sanders R J What causes the inverse re-lationship between primary production and export efficiencyin the Southern Ocean Geophys Res Lett 43 4457ndash4466httpsdoiorg1010022016GL068480 2016

Leoacuten-Muntildeoz J Urbina M A Garreaud R and Iriarte J LHydroclimatic conditions trigger record harmful algal bloomin western Patagonia (summer 2016) Sci Rep 8 1330httpsdoiorg101038s41598-018-19461-4 2018

Levinsen H and Nielsen T G The trophic role of ma-rine pelagic ciliates and heterotrophic dinoflagellatesin arctic and temperate coastal ecosystems A cross-latitude comparison Limnol Oceanogr 47 427ndash439httpsdoiorg104319lo20024720427 2002

Liestoslashl O The glaciers in the Kongsfjorden area Spitsber-gen Nor Geogr Tidsskr ndash Nor J Geogr 42 231ndash238httpsdoiorg10108000291958808552205 1988

Lin H Rauschenberg S Hexel C R Shaw T J and TwiningB S Free-drifting icebergs as sources of iron to the WeddellSea Deep Res Part Ii-Topical Stud Oceanogr 58 1392ndash1406httpsdoiorg101016jdsr2201011020 2011

Lippiatt S M Lohan M C and Bruland K WThe distribution of reactive iron in northern Gulf ofAlaska coastal waters Mar Chem 121 187ndash199httpsdoiorg101016jmarchem201004007 2010

Lischka S and Riebesell U Synergistic effects of oceanacidification and warming on overwintering pteropodsin the Arctic Global Chang Biol 18 3517ndash3528httpsdoiorg101111gcb12020 2012

Lischka S Buumldenbender J Boxhammer T and Riebesell UImpact of ocean acidification and elevated temperatures on earlyjuveniles of the polar shelled pteropod Limacina helicina mor-tality shell degradation and shell growth Biogeosciences 8919ndash932 httpsdoiorg105194bg-8-919-2011 2011

Lund-Hansen L C Hawes I Holtegaard Nielsen M DahlloumlfI and Sorrell B K Summer meltwater and spring sea ice pri-mary production light climate and nutrients in an Arctic estu-ary Kangerlussuaq west Greenland Arctic Antarct Alp Res50 S100025 httpsdoiorg10108015230430201714144682018

Lydersen C Assmy P Falk-Petersen S Kohler J Kovacs KM Reigstad M Steen H Stroslashm H Sundfjord A VarpeOslash Walczowski W Weslawski J M and Zajaczkowski MThe importance of tidewater glaciers for marine mammals andseabirds in Svalbard Norway J Mar Syst 129 452ndash471httpsdoiorg101016jjmarsys201309006 2014

Maat D S Prins M A and Brussaard C P D Sedi-ments from Arctic Tide-Water Glaciers Remove Coastal Ma-rine Viruses and Delay Host Infection Viruses 11 123httpsdoiorg103390v11020123 2019

Mankoff K D Straneo F Cenedese C Das S B Richards CG and Singh H Structure and dynamics of a subglacial dis-charge plume in a Greenlandic Fjord J Geophys Res-Ocean121 8670ndash8688 httpsdoiorg1010022016JC011764 2016

Markussen T N Elberling B Winter C and Ander-sen T J Flocculated meltwater particles control Arc-tic land-sea fluxes of labile iron Sci Rep 6 24033httpsdoiorg101038srep24033 2016

Marsay C M Barrett P M McGillicuddy D J and Sed-wick P N Distributions sources and transformations of dis-solved and particulate iron on the Ross Sea continental shelfduring summer J Geophys Res-Ocean 122 6371ndash6393httpsdoiorg1010022017JC013068 2017

Martin J H Glacial-interglacial CO2 change The iron hypothe-sis Paleoceanography 5 1ndash13 1990

Martin J H Fitzwater S E and Gordon R M Iron deficiencylimits phytoplankton growth in Antarctic waters Global Bio-geochem Cy 4 5ndash12 1990a

Martin J H Gordon R M and Fitzwater S E Iron in Antarcticwaters Nature 345 156ndash158 httpsdoiorg101038345156a01990b

Mascarenhas V J and Zielinski O Hydrography-Driven Opti-cal Domains in the Vaigat-Disko Bay and Godthabsfjord Ef-fects of Glacial Meltwater Discharge Front Mar Sci 6 335httpsdoiorg103389fmars201900335 2019

Mascioni M Almandoz G O Cefarelli A O Cusick AFerrario M E and Vernet M Phytoplankton composi-tion and bloom formation in unexplored nearshore waters ofthe western Antarctic Peninsula Polar Biol 42 1859ndash1872httpsdoiorg101007s00300-019-02564-7 2019

Meire L Soslashgaard D H Mortensen J Meysman F J RSoetaert K Arendt K E Juul-Pedersen T Blicher M Eand Rysgaard S Glacial meltwater and primary production aredrivers of strong CO2 uptake in fjord and coastal waters adja-cent to the Greenland Ice Sheet Biogeosciences 12 2347ndash2363httpsdoiorg105194bg-12-2347-2015 2015

Meire L Meire P Struyf E Krawczyk D W Arendt K EYde J C Juul Pedersen T Hopwood M J Rysgaard Sand Meysman F J R High export of dissolved silica fromthe Greenland Ice Sheet Geophys Res Lett 43 9173ndash9182httpsdoiorg1010022016GL070191 2016a

Meire L Mortensen J Rysgaard S Bendtsen J Boone WMeire P and Meysman F J R Spring bloom dynamics in asubarctic fjord influenced by tidewater outlet glaciers (Godtharingb-sfjord SW Greenland) J Geophys Res-Biogeosci 121 1581ndash1592 httpsdoiorg1010022015JG003240 2016b

Meire L Mortensen J Meire P Juul-Pedersen T Sejr MK Rysgaard S Nygaard R Huybrechts P and MeysmanF J R Marine-terminating glaciers sustain high productiv-ity in Greenland fjords Glob Chang Biol 23 5344ndash5357httpsdoiorg101111gcb13801 2017

Milner A M Khamis K Battin T J Brittain J E BarrandN E Fuumlreder L Cauvy-Fraunieacute S Giacuteslason G M Jacob-sen D Hannah D M Hodson A J Hood E LencioniV Oacutelafsson J S Robinson C T Tranter M and Brown

L E Glacier shrinkage driving global changes in down-stream systems P Natl Acad Sci USA 114 9770ndash9778httpsdoiorg101073pnas1619807114 2017

Mitra A Flynn K J Burkholder J M Berge T CalbetA Raven J A Graneacuteli E Glibert P M Hansen P JStoecker D K Thingstad F Tillmann U Varingge S WilkenS and Zubkov M V The role of mixotrophic protists inthe biological carbon pump Biogeosciences 11 995ndash1005httpsdoiorg105194bg-11-995-2014 2014

Moffat C Wind-driven modulation of warm water supply to aproglacial fjord Jorge Montt Glacier Patagonia Geophys ResLett 41 3943ndash3950 httpsdoiorg1010022014GL0600712014

Moon T Sutherland D A Carroll D Felikson D KehrlL and Straneo F Subsurface iceberg melt key to Green-land fjord freshwater budget Nat Geosci 11 49ndash54httpsdoiorg101038s41561-017-0018-z 2018

Moore C M Mills M M Arrigo K R Berman-Frank I BoppL Boyd P W Galbraith E D Geider R J Guieu C Jac-card S L Jickells T D La Roche J Lenton T M Ma-howald N M Maranon E Marinov I Moore J K Nakat-suka T Oschlies A Saito M A Thingstad T F Tsuda Aand Ulloa O Processes and patterns of oceanic nutrient limita-tion Nat Geosci 6 701ndash710 httpsdoiorg101038ngeo17652013

Morlighem M Williams C N Rignot E An L Arndt J EBamber J L Catania G Chaucheacute N Dowdeswell J ADorschel B Fenty I Hogan K Howat I Hubbard A Jakob-sson M Jordan T M Kjeldsen K K Millan R Mayer LMouginot J Noeumll B P Y OrsquoCofaigh C Palmer S Rys-gaard S Seroussi H Siegert M J Slabon P Straneo F vanden Broeke M R Weinrebe W Wood M and ZinglersenK B BedMachine v3 Complete Bed Topography and OceanBathymetry Mapping of Greenland From Multibeam EchoSounding Combined With Mass Conservation Geophys ResLett 44 11051ndash11061 httpsdoiorg1010022017GL0749542017

Mortensen J Lennert K Bendtsen J and Rysgaard S Heatsources for glacial melt in a sub-Arctic fjord (Godthabsfjord)in contact with the Greenland Ice Sheet J Geophys Res 116C01013 httpsdoiorg1010292010jc006528 2011

Mortensen J Bendtsen J Lennert K and Rysgaard SSeasonal variability of the circulation system in a westGreenland tidewater outlet glacier fjord Godtharingbsfjord(64 N) J Geophys Res-Earth Surf 119 2591ndash2603httpsdoiorg1010022014JF003267 2014

Mortensen J Rysgaard S Arendt K E Juul-Pedersen T Soslash-gaard D H Bendtsen J and Meire L Local Coastal Wa-ter Masses Control Heat Levels in a West Greenland TidewaterOutlet Glacier Fjord J Geophys Res-Ocean 123 8068ndash8083httpsdoiorg1010292018JC014549 2018

Morton B R Taylor G and Turner J S Turbulent gravitationalconvection from maintained and instantaneous sources ProcR Soc A 234 1ndash23 httpsdoiorg101098rspa195600111956

Mouginot J Rignot E Bjoslashrk A A van den Broeke M Mil-lan R Morlighem M Noeumll B Scheuchl B and WoodM Forty-six years of Greenland Ice Sheet mass balance from

1972 to 2018 P Natl Acad Sci USA 116 9239ndash9244httpsdoiorg101073pnas1904242116 2019

Moskalik M Cwiakała J Szczucinski W DominiczakA Głowacki O Wojtysiak K and Zagoacuterski P Spa-tiotemporal changes in the concentration and compositionof suspended particulate matter in front of Hansbreen atidewater glacier in Svalbard Oceanologia 60 446ndash463httpsdoiorg101016joceano201803001 2018

Murray C Markager S Stedmon C A Juul-PedersenT Sejr M K and Bruhn A The influence of glacialmelt water on bio-optical properties in two contrastingGreenlandic fjords Estuar Coast Shelf Sci 163 72ndash83httpsdoiorg101016jecss201505041 2015

Nielsdottir M C Moore C M Sanders R Hinz D J andAchterberg E P Iron limitation of the postbloom phytoplank-ton communities in the Iceland Basin Global Biogeochem Cy23 GB3001 httpsdoiorg1010292008gb003410 2009

Nielsen T G Plankton community structure and carbon cyclingon the western coast of Greenland during the stratified summersituation I Hydrography phytoplankton and bacterioplanktonAquat Microb Ecol 16 205ndash216 1999

Nielsen T G and Hansen B Plankton community structure andcarbon cycling on the western coast of Greenland during and af-ter the sedimentation of a diatom bloom Mar Ecol Prog Ser125 239ndash257 1995

Nightingale A M Beaton A D and Mowlem M C Trendsin microfluidic systems for in situ chemical analysis of nat-ural waters Sensors Actuators B Chem 221 1398ndash1405httpsdoiorg101016jsnb201507091 2015

Noeumll B van de Berg W J van Meijgaard E Kuipers MunnekeP van de Wal R S W and van den Broeke M R Evalua-tion of the updated regional climate model RACMO23 summersnowfall impact on the Greenland Ice Sheet The Cryosphere 91831ndash1844 httpsdoiorg105194tc-9-1831-2015 2015

Normandeau A Dietrich P Hughes Clarke J Van WychenW Lajeunesse P Burgess D and Ghienne J-F Re-treat Pattern of Glaciers Controls the Occurrence of Tur-bidity Currents on High-Latitude Fjord Deltas (Eastern Baf-fin Island) J Geophys Res-Earth Surf 124 1559ndash1571httpsdoiorg1010292018JF004970 2019

Oliver H Luo H Castelao R M van Dijken G L Mat-tingly K Rosen J J Mote T L Arrigo K R Renner-malm Aring K Tedesco M and Yager P L Exploring the Po-tential Impact of Greenland Meltwater on Stratification Pho-tosynthetically Active Radiation and Primary Production inthe Labrador Sea J Geophys Res-Ocean 123 2570ndash2591httpsdoiorg1010022018JC013802 2018

Overeem I Hudson B D Syvitski J P M Mikkelsen AB Hasholt B Van Den Broeke M R Noel B P Y andMorlighem M Substantial export of suspended sediment to theglobal oceans from glacial erosion in Greenland Nat Geosci10 859ndash863 httpsdoiorg101038NGEO3046 2017

Pabi S van Dijken G L and Arrigo K R Primary productionin the Arctic Ocean 1998ndash2006 J Geophys Res-Ocean 113C08005 httpsdoiorg1010292007JC004578 2008

Pabortsava K Lampitt R S Benson J Crowe C McLachlanR Le Moigne F A C Mark Moore C Pebody C ProvostP Rees A P Tilstone G H and Woodward E M S Carbon

sequestration in the deep Atlantic enhanced by Saharan dust NatGeosci 10 189ndash194 httpsdoiorg101038ngeo2899 2017

Paulsen M L Nielsen S E B Muumlller O Moslashller E FStedmon C A Juul-Pedersen T Markager S Sejr MK Delgado Huertas A Larsen A and Middelboe MCarbon Bioavailability in a High Arctic Fjord Influenced byGlacial Meltwater NE Greenland Front Mar Sci 4 176httpsdoiorg103389fmars201700176 2017

Paulsen M L Muumlller O Larsen A Moslashller E F Middelboe MSejr M K and Stedmon C Biological transformation of Arc-tic dissolved organic matter in a NE Greenland fjord LimnolOceanogr 64 1014ndash1033 httpsdoiorg101002lno110912018

Poorvin L Rinta-Kanto J M Hutchins D A and Wilhelm SW Viral release of iron and its bioavailability to marine plank-ton Limnol Oceanogr 49 1734ndash1741 2004

Prado-Fiedler R Winter and summer distribution of dissolved oxy-gen pH and nutrients at the heads of fjords in Chilean Patagoniawith possible phosphorus limitation Rev Biol Mar Oceanogr44 783ndash789 2009

Prospero J M Bullard J E and Hodgkins R High-Latitude Dust Over the North Atlantic Inputs from Ice-landic Proglacial Dust Storms Science 80 1078ndash1082httpsdoiorg101126science1217447 2012

Raiswell R and Canfield D E The Iron biogeochemi-cal Cycle Past and Present Geochem Perspect 1 1ndash220httpsdoiorg107185geochempersp11 2012

Raiswell R Tranter M Benning L G Siegert M Dersquoath RHuybrechts P and Payne T Contributions from glacially de-rived sediment to the global iron (oxyhydr)oxide cycle Impli-cations for iron delivery to the oceans Geochim CosmochimActa 70 2765ndash2780 httpsdoiorg101016jgca2005120272006

Raiswell R Benning L G Tranter M and TulaczykS Bioavailable iron in the Southern Ocean the signifi-cance of the iceberg conveyor belt Geochem Trans 9 7httpsdoiorg1011861467-4866-9-7 2008

Redfield A C On the proportions of organic derivations in seawater and their relation to the composition of plankton in JamesJohnstone Memorial Volume edited by R J Daniel 177ndash192University Press of Liverpool Liverpool 1934

Reisdorph S C and Mathis J T Assessing net community pro-duction in a glaciated Alaskan fjord Biogeosciences 12 5185ndash5198 httpsdoiorg105194bg-12-5185-2015 2015

Ren Z Martyniuk N Oleksy I A Swain A and Hotaling SEcological Stoichiometry of the Mountain Cryosphere FrontEcol Evol 7 360 httpsdoiorg103389fevo2019003602019

Renner M Arimitsu M L Piatt J F and Rochet M-J Struc-ture of marine predator and prey communities along environmen-tal gradients in a glaciated fjord Can J Fish Aquat Sci 692029ndash2045 httpsdoiorg101139f2012-117 2012

Ribeiro S Moros M Ellegaard M and Kuijpers A Climatevariability in West Greenland during the past 1500 years evi-dence from a high-resolution marine palynological record fromDisko Bay Boreas 41 68ndash83 httpsdoiorg101111j1502-3885201100216x 2012

Ribeiro S Sejr M K Limoges A Heikkilauml M Andersen TJ Tallberg P Weckstroumlm K Husum K Forwick M Dals-

gaard T Masseacute G Seidenkrantz M-S and Rysgaard S Seaice and primary production proxies in surface sediments froma High Arctic Greenland fjord Spatial distribution and impli-cations for palaeoenvironmental studies Ambio 46 106ndash118httpsdoiorg101007s13280-016-0894-2 2017

Richlen M L Zielinski O Holinde L Tillmann U Cem-bella A Lyu Y and Anderson D M Distribution of Alexan-drium fundyense (Dinophyceae) cysts in Greenland and Icelandwith an emphasis on viability and growth in the Arctic MarEcol Prog Ser 547 33ndash46 httpsdoiorg103354meps116602016

Rignot E Jacobs S Mouginot J and Scheuchl B Ice-Shelf Melting Around Antarctica Science 80 266ndash270httpsdoiorg101126science1235798 2013

Rijkenberg M J A Slagter H A Rutgers van der Lo-eff M van Ooijen J and Gerringa L J A DissolvedFe in the Deep and Upper Arctic Ocean With a Focus onFe Limitation in the Nansen Basin Front Mar Sci 5 88httpsdoiorg103389fmars201800088 2018

Ryan-Keogh T J Macey A I Nielsdottir M C Lucas MI Steigenberger S S Stinchcombe M C Achterberg E PBibby T S and Moore C M Spatial and temporal develop-ment of phytoplankton iron stress in relation to bloom dynamicsin the high-latitude North Atlantic Ocean Limnol Oceanogr58 533ndash545 httpsdoiorg104319lo20135820533 2013

Rysgaard S and Glud R N Carbon cycling and climate changePredictions for a high-Arctic marine ecosystem (Young SoundNE Greenland) Meddelelser om Groenland Bioscience 58206ndash213 2007

Rysgaard S Nielsen T and Hansen B Seasonal varia-tion in nutrients pelagic primary production and grazingin a high-Arctic coastal marine ecosystem Young SoundNortheast Greenland Mar Ecol Prog Ser 179 13ndash25httpsdoiorg103354meps179013 1999

Rysgaard S Vang T Stjernholm M Rasmussen B WindelinA and Kiilsholm S Physical conditions carbon transport andclimate change impacts in a northeast Greenland fjord ArctAntarct Alp Res 35 301ndash312 httpsdoiorg1016571523-0430(2003)035[0301pcctac]20co2 2003

Rysgaard S Mortensen J Juul-Pedersen T Soslashrensen L LLennert K Soslashgaard D H Arendt K E Blicher M ESejr M K and Bendtsen J High airndashsea CO2 uptake ratesin nearshore and shelf areas of Southern Greenland Tem-poral and spatial variability Mar Chem 128ndash129 26ndash33httpsdoiorg101016jmarchem201111002 2012

Ryu J-S and Jacobson A D CO2 evasion from the Greenland IceSheet A new carbon-climate feedback Chem Geol 320ndash32180ndash95 httpsdoiorg101016jchemgeo201205024 2012

Schaffer J Kanzow T von Appen W von Albedyll L Arndt JE and Roberts D H Bathymetry constrains ocean heat supplyto Greenlandrsquos largest glacier tongue Nat Geosci 13 227ndash231httpsdoiorg101038s41561-019-0529-x 2020

Schild K M Hawley R L and Morriss B F Sub-glacial hydrology at Rink Isbraelig West Greenland inferredfrom sediment plume appearance Ann Glaciol 57 118ndash127httpsdoiorg101017aog20161 2016

Schlosser C Schmidt K Aquilina A Homoky W B Cas-trillejo M Mills R A Patey M D Fielding S AtkinsonA and Achterberg E P Mechanisms of dissolved and labile

particulate iron supply to shelf waters and phytoplankton bloomsoff South Georgia Southern Ocean Biogeosciences 15 4973ndash4993 httpsdoiorg105194bg-15-4973-2018 2018

Schmidt K Atkinson A Steigenberger S Fielding S LindsayM C M Pond D W Tarling G A Klevjer T A Allen CS Nicol S and Achterberg E P Seabed foraging by Antarctickrill Implications for stock assessment bentho-pelagic couplingand the vertical transfer of iron Limnol Oceanogr 56 1411ndash1428 httpsdoiorg104319lo20115641411 2011

Schroth A W Crusius J Chever F Bostick B C and RouxelO J Glacial influence on the geochemistry of riverine ironfluxes to the Gulf of Alaska and effects of deglaciation GeophysRes Lett 38 L16605 httpsdoiorg1010292011gl0483672011

Schroth A W Crusius J Campbell R W and Hoyer IEstuarine removal of glacial iron and implications for ironfluxes to the ocean Geophys Res Lett 41 3951ndash3958httpsdoiorg1010022014GL060199 2014

Sedwick P N Marsay C M Sohst B M Aguilar-Islas A MLohan M C Long M C Arrigo K R Dunbar R B SaitoM A Smith W O and DiTullio G R Early season depletionof dissolved iron in the Ross Sea polynya Implications for irondynamics on the Antarctic continental shelf J Geophys Res116 C12019 httpsdoiorg1010292010JC006553 2011

Seifert M Hoppema M Burau C Elmer C Friedrichs AGeuer J K John U Kanzow T Koch B P Konrad Cvan der Jagt H Zielinski O and Iversen M H Influ-ence of Glacial Meltwater on Summer Biogeochemical Cyclesin Scoresby Sund East Greenland Front Mar Sci 6 412httpsdoiorg103389fmars201900412 2019

Sejr M K Krause-Jensen D Rysgaard S Soslashrensen L LChristensen P B and Glud R N Airndashsea flux of CO2in arctic coastal waters influenced by glacial melt water andsea ice Tellus B 63 815ndash822 httpsdoiorg101111j1600-0889201100540x 2011

Sejr M K Stedmon C A Bendtsen J Abermann J Juul-Pedersen T Mortensen J and Rysgaard S Evidence of localand regional freshening of Northeast Greenland coastal watersSci Rep 7 13183 httpsdoiorg101038s41598-017-10610-9 2017

Shaffer G and Lambert F In and out of glacial extremes by wayof dustminusclimate feedbacks P Natl Acad Sci USA 115 2026ndash2031 httpsdoiorg101073pnas1708174115 2018

Sholkovitz E R Boyle E A and Price N B The removal ofdissolved humic acids and iron during estuarine mixing EarthPlanet Sci Lett 40 130ndash136 httpsdoiorg1010160012-821X(78)90082-1 1978

Slater D A Straneo F Das S B Richards C GWagner T J W and Nienow P W Localized PlumesDrive Front-Wide Ocean Melting of A Greenlandic Tide-water Glacier Geophys Res Lett 45 12312ndash350358httpsdoiorg1010292018GL080763 2018

Smith R W Bianchi T S Allison M Savage C and GalyV High rates of organic carbon burial in fjord sediments glob-ally Nat Geosci 8 450ndash453 httpsdoiorg101038ngeo24212015

Smoła Z T Tatarek A Wiktor J M Wiktor J M W KubiszynA and Wesławski J M Primary producers and production inHornsund and Kongsfjorden ndash comparison of two fjord systems

Polish Polar Res 38 351ndash373 httpsdoiorg101515popore-2017-0013 2017

Sommaruga R When glaciers and ice sheets melt conse-quences for planktonic organisms J Plankton Res 37 509ndash518httpsdoiorg101093planktfbv027 2015

Spall M A Jackson R H and Straneo F Katabatic Wind-Driven Exchange in Fjords J Geophys Res-Ocean 122 8246ndash8262 httpsdoiorg1010022017JC013026 2017

Statham P J Skidmore M and Tranter M Inputs of glaciallyderived dissolved and colloidal iron to the coastal ocean and im-plications for primary productivity Global Biogeochem Cy 22Gb3013 httpsdoiorg1010292007gb003106 2008

Stevens L A Straneo F Das S B Plueddemann A J KukulyaA L and Morlighem M Linking glacially modified watersto catchment-scale subglacial discharge using autonomous un-derwater vehicle observations The Cryosphere 10 417ndash432httpsdoiorg105194tc-10-417-2016 2016

Stevenson E I Fantle M S Das S B Williams H M andAciego S M The iron isotopic composition of subglacialstreams draining the Greenland ice sheet Geochim CosmochimActa 213 237ndash254 httpsdoiorg101016jgca2017060022017

Stibal M Anesio A M Blues C J D and TranterM Phosphatase activity and organic phosphorus turnoveron a high Arctic glacier Biogeosciences 6 913ndash922httpsdoiorg105194bg-6-913-2009 2009

St-Laurent P Yager P L Sherrell R M Stammerjohn S Eand Dinniman M S Pathways and supply of dissolved iron inthe Amundsen Sea (Antarctica) J Geophys Res-Ocean 1227135ndash7162 httpsdoiorg1010022017JC013162 2017

St-Laurent P Yager P L Sherrell R M Oliver H Din-niman M S and Stammerjohn S E Modeling the Sea-sonal Cycle of Iron and Carbon Fluxes in the Amundsen SeaPolynya Antarctica J Geophys Res-Ocean 124 1544ndash1565httpsdoiorg1010292018JC014773 2019

Stocker T F Qin D Plattner G-K Tignor M Allen S KBoschung J Nauels A Xia Y Bex V and Midgley P MClimate change 2013 The physical science basis Contributionof working group I to the fifth assessment report of the intergov-ernmental panel on climate change 1535 2013

Stoecker D K and Lavrentyev P J Mixotrophic Plankton inthe Polar Seas A Pan-Arctic Review Front Mar Sci 5 292httpsdoiorg103389fmars201800292 2018

Stoecker D K Hansen P J Caron D A and Mitra A Mixotro-phy in the Marine Plankton Annu Rev Mar Sci 9 311ndash335httpsdoiorg101146annurev-marine-010816-060617 2017

Straneo F and Cenedese C The Dynamics of GreenlandrsquosGlacial Fjords and Their Role in Climate Annu Rev MarSci 7 89ndash112 httpsdoiorg101146annurev-marine-010213-135133 2015

Straneo F Hamilton G S Sutherland D A Stearns L ADavidson F Hammill M O Stenson G B and Rosing-Asvid A Rapid circulation of warm subtropical waters in amajor glacial fjord in East Greenland Nat Geosci 3 182ndash186httpsdoiorg101038ngeo764 2010

Straneo F Curry R G Sutherland D A Hamilton G SCenedese C Varingge K and Stearns L A Impact of fjorddynamics and glacial runoff on the circulation near HelheimGlacier Nat Geosci 4 322ndash327 2011

Straneo F Sutherland D A Holland D Gladish CHamilton G S Johnson H L Rignot E Xu Yand Koppes M Characteristics of ocean waters reach-ing Greenlandrsquos glaciers Ann Glaciol 53 202ndash210httpsdoiorg1031892012AoG60A059 2012

Straneo F Sutherland D A Stearns L Catania G Heim-bach P Moon T Cape M R Laidre K L Barber DRysgaard S Mottram R Olsen S Hopwood M J andMeire L The Case for a Sustained Greenland Ice Sheet-Ocean Observing System (GrIOOS) Front Mar Sci 6 138httpsdoiorg103389fmars201900138 2019

Štrojsovaacute A Vrba J Nedoma J and Šimek K Extracellularphosphatase activity of freshwater phytoplankton exposed to dif-ferent in situ phosphorus concentrations Mar Freshw Res 56417ndash424 httpsdoiorg101071MF04283 2005

Strzepek R F Maldonado M T Higgins J L Hall J Safi KWilhelm S W and Boyd P W Spinning the ldquoFerrous WheelrdquoThe importance of the microbial community in an iron budgetduring the FeCycle experiment Global Biogeochem Cy 19GB4S26 httpsdoiorg1010292005GB002490 2005

Sundfjord A Albretsen J Kasajima Y Skogseth R Kohler JNuth C Skarethhamar J Cottier F Nilsen F Asplin L Ger-land S and Torsvik T Effects of glacier runoff and wind onsurface layer dynamics and Atlantic Water exchange in Kongs-fjorden Svalbard a model study Estuar Coast Shelf Sci 187260ndash272 httpsdoiorg101016jecss201701015 2017

Sutherland D A Pickart R S Peter Jones E Azetsu-ScottK Jane Eert A and Oacutelafsson J Freshwater composi-tion of the waters off southeast Greenland and their linkto the Arctic Ocean J Geophys Res-Ocean 114 C05020httpsdoiorg1010292008JC004808 2009

Sutherland D A Roth G E Hamilton G S Mernild S HStearns L A and Straneo F Quantifying flow regimes in aGreenland glacial fjord using iceberg drifters Geophys ResLett 41 8411ndash8420 httpsdoiorg1010022014GL0622562014

Svendsen H Beszczynska-Moslashller A Hagen J O LefauconnierB Tverberg V Gerland S Oslashrboslashk J B Bischof K PapucciC Zajaczkowski M Azzolini R Bruland O Wiencke CWinther J-G and Dallmann W The physical environmentof KongsfjordenndashKrossfjorden an Arctic fjord system in Sval-bard Polar Res 21 133ndash166 httpsdoiorg101111j1751-83692002tb00072x 2002

Tagliabue A Aumont O DeAth R Dunne J P DutkiewiczS Galbraith E Misumi K Moore J K Ridgwell A Sher-man E Stock C Vichi M Voumllker C and Yool A Howwell do global ocean biogeochemistry models simulate dis-solved iron distributions Global Biogeochem Cy 30 149ndash174 httpsdoiorg1010022015GB005289 2016

Taylor R L Semeniuk D M Payne C D Zhou JTremblay J-Eacute Cullen J T and Maldonado M T Col-imitation by light nitrate and iron in the Beaufort Seain late summer J Geophys Res-Ocean 118 3260ndash3277httpsdoiorg101002jgrc20244 2013

Thingstad T F Bellerby R G J Bratbak G Boslashrsheim K YEgge J K Heldal M Larsen A Neill C Nejstgaard J Nor-land S Sandaa R-A Skjoldal E F Tanaka T ThyrhaugR and Toumlpper B Counterintuitive carbon-to-nutrient cou-

pling in an Arctic pelagic ecosystem Nature 455 387ndash390httpsdoiorg101038nature07235 2008

Thuroczy C-E Alderkamp A-C Laan P Gerringa L J AMills M M Van Dijken G L De Baar H J W and Ar-rigo K R Key role of organic complexation of iron in sus-taining phytoplankton blooms in the Pine Island and AmundsenPolynyas (Southern Ocean) Deep Res Part Ii 71ndash76 49ndash60httpsdoiorg101016jdsr2201203009 2012

Tonnard M Planquette H Bowie A R van der Merwe P Gal-linari M Desprez de Geacutesincourt F Germain Y Gourain ABenetti M Reverdin G Treacuteguer P Boutorh J Cheize MLacan F Menzel Barraqueta J-L Pereira-Contreira L Shel-ley R Lherminier P and Sarthou G Dissolved iron in theNorth Atlantic Ocean and Labrador Sea along the GEOVIDEsection (GEOTRACES section GA01) Biogeosciences 17 917ndash943 httpsdoiorg105194bg-17-917-2020 2020

Torres M A Moosdorf N Hartmann J Adkins J Fand West A J Glaciers sulfide oxidation and the car-bon cycle P Natl Acad Sci USA 114 8716ndash8721httpsdoiorg101073pnas1702953114 2017

Torsvik T Albretsen J Sundfjord A Kohler J Sandvik A DSkarethhamar J Lindbaumlck K and Everett A Impact of tide-water glacier retreat on the fjord system Modeling present andfuture circulation in Kongsfjorden Svalbard Estuar Coast ShelfSci 220 152ndash165 httpsdoiorg101016jecss2019020052019

Tranter M Huybrechts P Munhoven G Sharp M J Brown GH Jones IW Hodson A J Hodgkins R and Wadham J LDirect effect of ice sheets on terrestrial bicarbonate sulphate andbase cation fluxes during the last glacial cycle minimal impacton atmospheric CO2 concentrations Chem Geol 190 33ndash44httpsdoiorg101016S0009-2541(02)00109-2 2002

Tremblay J-Eacute Anderson L G Matrai P Coupel P BeacutelangerS Michel C and Reigstad M Global and regional driversof nutrient supply primary production and CO2 drawdown inthe changing Arctic Ocean Prog Oceanogr 193 171ndash196httpsdoiorg101016jpocean201508009 2015

Turk D Bedard J M Burt W J Vagle S Thomas H Azetsu-Scott K McGillis W R Iverson S J and Wallace D WR Inorganic carbon in a high latitude estuary-fjord system inCanadarsquos eastern Arctic Estuar Coast Shelf Sci 178 137ndash147httpsdoiorg101016jecss201606006 2016

Twining B S and Baines S B The Trace Metal Compositionof Marine Phytoplankton Ann Rev Mar Sci 5 191ndash215httpsdoiorg101146annurev-marine-121211-172322 2013

Uehlinger U Robinson C Hieber M and Zah R The physico-chemical habitat template for periphyton in alpine glacialstreams under a changing climate Hydrobiologia 657 107ndash121101007s10750-009-9963-x 2010

Uitz J Claustre H Griffiths F B Ras J Garcia Nand Sandroni V A phytoplankton class-specific primaryproduction model applied to the Kerguelen Islands region(Southern Ocean) Deep Sea Res Part I 56 541ndash560httpsdoiorg101016jdsr200811006 2009

van de Poll W H Kulk G Rozema P D Brussaard C P DVisser R J W and Buma A G J Contrasting glacial melt-water effects on post-bloom phytoplankton on temporal and spa-tial scales in Kongsfjorden Spitsbergen Elem Sci Anth 6 50httpsdoiorg101525elementa307 2018

van der Merwe P C Wuttig K Holmes T Trull TChase Z Townsend A Goemann K and Bowie AR High lability Fe particles sourced from glacial ero-sion can meet previously unaccounted biological demandHeard Island Southern Ocean Front Mar Sci 6 332httpsdoiorg103389fmars201900332 2019

Vandersea M W Kibler S R Tester P A Holderied K Hon-dolero D E Powell K Baird S Doroff A Dugan Dand Litaker R W Environmental factors influencing the dis-tribution and abundance of Alexandrium catenella in Kachemakbay and lower cook inlet Alaska Harmful Algae 77 81ndash92httpsdoiorg101016jhal201806008 2018

Vergara-Jara M J DeGrandpre M D Torres R Beatty C MCuevas L A Alarcoacuten E and Iriarte J L Seasonal changesin carbonate saturation state and air-sea CO2 fluxes during anannual cycle in a stratified-temperate fjord (Reloncaviacute FjordChilean Patagonia) J Geophys Res-Biogeosci 124 2851ndash2865 httpsdoiorg1010292019JG005028 2019

Vraspir J M and Butler A Chemistry of Marine Lig-ands and Siderophores Annu Rev Mar Sci 1 43ndash63httpsdoiorg101146annurevmarine010908163712 2009

Wadham J L Tranter M Skidmore M Hodson AJ Priscu J Lyons W B Sharp M Wynn P andJackson M Biogeochemical weathering under iceSize matters Global Biogeochem Cy 24 GB3025httpsdoiorg1010292009GB003688 2010

Wadham J L Hawkings J Telling J Chandler D AlcockJ OrsquoDonnell E Kaur P Bagshaw E Tranter M TedstoneA and Nienow P Sources cycling and export of nitrogenon the Greenland Ice Sheet Biogeosciences 13 6339ndash6352httpsdoiorg105194bg-13-6339-2016 2016

Wadham J L Hawkings J R Tarasov L Gregoire L JSpencer R G M Gutjahr M Ridgwell A and Kohfeld KE Ice sheets matter for the global carbon cycle Nat Commun10 3567 httpsdoiorg101038s41467-019-11394-4 2019

Ward B A and Follows M J Marine mixotrophy increasestrophic transfer efficiency mean organism size and verti-cal carbon flux P Natl Acad Sci USA 113 2958ndash2963httpsdoiorg101073pnas1517118113 2016

Wccedilslawski W J M and Legezytnska J Glaciers causedzooplankton mortality J Plankton Res 20 1233ndash1240httpsdoiorg101093plankt2071233 1998

Wehrmann L M Formolo M J Owens J D RaiswellR Ferdelman T G Riedinger N and Lyons T WIron and manganese speciation and cycling in glaciallyinfluenced high-latitude fjord sediments (West Spitsber-gen Svalbard) Evidence for a benthic recycling-transportmechanism Geochim Cosmochim Acta 141 628ndash655httpsdoiorg101016jgca201406007 2014

Wells M L Trainer V L Smayda T J Karlson B SO Trick C G Kudela R M Ishikawa A Bernard SWulff A Anderson D M and Cochlan W P Harmful al-gal blooms and climate change Learning from the past andpresent to forecast the future Harmful Algae 49 68ndash93httpsdoiorg101016jhal201507009 2015

White J R and Dagg M J Effects of suspended sediments onegg production of the calanoid copepod Acartia tonsa Mar Biol102 315ndash319 httpsdoiorg101007BF00428483 1989

Wiedmann I Reigstad M Marquardt M Vader A andGabrielsen T M Seasonality of vertical flux and sink-ing particle characteristics in an ice-free high arcticfjordmdashDifferent from subarctic fjords J Mar Syst 154192ndash205 httpsdoiorg101016jjmarsys201510003 2016

Windom H Byrd J Smith R Hungspreugs M Dhar-mvanij S Thumtrakul W and Yeats P Trace metal-nutrient relationships in estuaries Mar Chem 32 177ndash194httpsdoiorg1010160304-4203(91)90037-W 1991

Włodarska-Kowalczuk M and Pearson T H Soft-bottom mac-robenthic faunal associations and factors affecting species dis-tributions in an Arctic glacial fjord (Kongsfjord Spitsbergen)Polar Biol 27 155ndash167 httpsdoiorg101007s00300-003-0568-y 2004

Włodarska-Kowalczuk M and Weslawski J M Impact of climatewarming on Arctic benthic biodiversity a case study of two Arc-tic glacial bays Clim Res 18 127ndash132 2001

Włodarska-Kowalczuk M Pearson T H and Kendall M ABenthic response to chronic natural physical disturbance byglacial sedimentation in an Arctic fjord Mar Ecol Progr Ser303 31ndash41 httpsdoiorg103354meps303031 2005

Włodarska-Kowalczuk M Mazurkiewicz M Goacuterska BMichel L N Jankowska E and Zaborska A Organiccarbon origin benthic faunal consumption and burial insediments of northern Atlantic and Arctic fjords (60ndash81 N) J Geophys Res-Biogeosci 124 3737ndash3751httpsdoiorg1010292019JG005140 2019

Wojtasiewicz B Trull T W Clementson L Davies D MPatten N L Schallenberg C and Hardman-Mountford NJ Factors Controlling the Lack of Phytoplankton Biomassin Naturally Iron Fertilized Waters Near Heard and McDon-ald Islands in the Southern Ocean Front Mar Sci 6 531httpsdoiorg103389fmars201900531 2019

Xie H Beacutelanger S Song G Benner R Taalba A Blais MTremblay J-Eacute and Babin M Photoproduction of ammoniumin the southeastern Beaufort Sea and its biogeochemical implica-tions Biogeosciences 9 3047ndash3061 httpsdoiorg105194bg-9-3047-2012 2012

Xu Y Rignot E Menemenlis D and Koppes M Nu-merical experiments on subaqueous melting of greenlandtidewater glaciers in response to ocean warming and en-hanced subglacial discharge Ann Glaciol 53 229ndash234httpsdoiorg1031892012AoG60A139 2012

Yde J C Knudsen N T and Nielsen O B Glacier hy-drochemistry solute provenance and chemical denuda-tion at a surge-type glacier in Kuannersuit KuussuatDisko Island West Greenland J Hydrol 300 172ndash187httpsdoiorg101016jjhydrol200406008 2005

Yde J C Knudsen N T Hasholt B and Mikkelsen A BMeltwater chemistry and solute export from a Greenland IceSheet catchment Watson River West Greenland J Hydrol 5192165ndash2179 httpsdoiorg101016jjhydrol201410018 2014

Zajaczkowski M and Włodarska-Kowalczuk M Dynamic sed-imentary environments of an Arctic glacier-fed river estu-ary (Adventfjorden Svalbard) I Flux deposition and sed-iment dynamics Estuar Coast Shelf Sci 74 285ndash296httpsdoiorg101016jecss200704015 2007

Zhang R John S G Zhang J Ren J Wu Y Zhu ZLiu S Zhu X Marsay C M and Wenger F Trans-port and reaction of iron and iron stable isotopes in glacialmeltwaters on Svalbard near Kongsfjorden From rivers toestuary to ocean Earth Planet Sci Lett 424 201ndash211httpsdoiorg101016jepsl201505031 2015

Abstract

Introduction

Fjords as critical zones for glacierndashocean interactions

Pelagic primary production in Arctic glacier fjords

Effects of glacial discharge on marine resource availability

Non-conservative mixing processes for Fe and Si

Deriving glacierndashocean fluxes

How do variations in the behaviour and location of higher-trophic-level organisms affect nutrient availability to marine microorganisms

Critical differences between surface and subsurface discharge release

Is benthicndashpelagic coupling enhanced by subglacial discharge

From pelagic primary production to the carbon sink

Contrasting Fe- and NO3-limited regions of the ocean

The subglacial discharge pump from macronutrients to iron

Effects on the carbonate system

Organic matter in glacial discharge

Insights into the long-term effects of glacier retreat

A need for new approaches

A link between retreating glaciers and harmful algal blooms

Understanding the role of glaciers alongside other manifestations of climate change

Conclusions

Where and when does glacial freshwater discharge promote or reduce marine primary production

How does spatio-temporal variability in glacial discharge affect marine primary production

How far-reaching are the effects of glacial discharge on marine biogeochemistry

Data availability

Supplement

Author contributions

Competing interests

Acknowledgements

Financial support

Review statement

References