Volume 46, Issue 11 p. 5959-5967
Research Letter
Free Access

Algal Export in the Arctic Ocean in Times of Global Warming

Catherine Lalande

Corresponding Author

Catherine Lalande

Département de biologie, Université Laval, Québec City, Canada

Correspondence to: C. Lalande,

[email protected]

Search for more papers by this author
Eva-Maria Nöthig

Eva-Maria Nöthig

Polar Biological Oceanography, Alfred-Wegener Institute Helmholtz Centre for Polar and Marine Research, Bremerhaven, Germany

Search for more papers by this author
Louis Fortier

Louis Fortier

Département de biologie, Université Laval, Québec City, Canada

Search for more papers by this author
First published: 20 May 2019
Citations: 45


Satellite-derived data suggest an increase in annual primary production following the loss of summer sea ice in the Arctic Ocean. The scarcity of field data to corroborate this enhanced algal production incited a collaborative project combining six annual cycles of sequential sediment trap measurements obtained over a 17-year period in the Eurasian Arctic Ocean. Here we present microalgal fluxes measured at ~200 m to reflect the bulk of algal carbon production. Ice algae contributed to a large proportion of the microalgal carbon export before complete ice melt and possible detection of their production by satellites. In the northern Laptev Sea, annual microalgal carbon fluxes were lower during the 2007 minimum ice extent than in 2006. In 2012, early snowmelt led to early microalgal carbon flux in the Nansen Basin. Hence, a change in the timing of snowmelt and ice algae release may affect productivity and export over the Arctic basins.

Key Points

  • The majority of algal cells were exported before complete ice melt
  • Early snowmelt led to the early release of ice algae prior to the record minimum sea ice extent of 2012
  • Phytoplankton carbon contributed up to ~40% of the particulate organic carbon fluxes during short intense export events

Plain Language Summary

An increase in algal production has been reported following the recent decline in sea ice cover in the Arctic Ocean. However, very few field measurements have been conducted to corroborate these remote sensing-derived observations. Algal cell fluxes collected with sediment traps at ~200 m during six annual cycles between 1995 and 2012 were combined to determine changes in the timing, magnitude, and composition of the algal export in the Eurasian Arctic Ocean. Algal fluxes, mostly composed of ice algae, reflected a large fraction of the surface phytoplankton carbon production. Ice algae were nearly all collected during snowmelt and before complete ice melt, preventing the detection of their production by satellites. While relatively high algal export was observed under thick ice cover in 1996, early algal export in 2012 suggested an early release of ice algae due to early snowmelt. Overall, snow cover played an important role in regulating the timing and magnitude of ice algae release in the Eurasian Arctic Ocean.

1 Introduction

Aside from the occasional ice camps or overwintering expeditions, biological measurements in the Arctic Ocean usually take place several weeks after the onset of the productive season during the sea ice minimum period. This limitation results in a lack of data on the magnitude, composition, and duration of the algal production, especially in the remote Central Arctic Ocean. A strategy to fill these gaps is to use moored sequential sediment traps deployed over annual cycles to obtain continuous biological measurements indirectly reflecting pelagic processes not measurable by ship nor satellite. Such moorings with sequential sediment traps have been deployed in the Eurasian Arctic Ocean on six occasions over a 17-year period spanning from 1995 to 2012. Moored sediment traps were deployed by the Alfred-Wegener Institute Helmholtz Centre for Polar and Marine Research (AWI) of Germany on the Lomonosov Ridge (LR: 1995-1996; Fahl & Nöthig, 2007; Fahl & Stein, 2012; Zernova et al., 2000), in the Amundsen Basin (AB: 2011-2012), and in the Nansen Basin (NB: 2011-2012), and by the Network of Centres of Excellence of Canada ArcticNet in the northern Laptev Sea (LS: 2005-2006 and 2006-2007; Lalande, Bélanger, et al., 2009; Lalande, Forest, et al., 2009) and in the East Siberian Sea (2007-2008; Figure 1 and Table 1). Here we present seasonal and annual variations in the magnitude, timing, and composition of algal fluxes collected with these sediment traps in an area that experienced a large decline in seasonal ice cover over the past decade.

Details are in the caption following the image
Mooring positions in the Eurasian Arctic Ocean Lomonosov Ridge (LR; 1995-1996), Amundsen Basin (AB; 2011-2012), Nansen Basin (NB; 2011-2012), Northern Laptev Sea (LS; 2005-2006 and 2006-2007), and East Siberian Sea (ESS; 2007-2008).
Table 1. Location, Sampling Period, and Deployment Information of the Sediment Traps Moored in the Eurasian Arctic Ocean
Mooring Latitude (°N) Longitude (°E) Deployment period Trap model Trap depth (m) Water depth (m)
Lomonosov Ridge 81° 4.083′ 138° 54′ 15 September 1995 to 16 August 1996 Kiel 150 1,710
Northern Laptev Sea 79° 55.717′ 142° 21.812′ 19 September 2005 to 28 August 2006 Technicap PPS3/3 175 1,350
Northern Laptev Sea 79° 55.717′ 142° 21.812′ 16 September 2006 to 31 August 2007 Technicap PPS3/3 180 1,355
East Siberian Sea 79° 45.407′ 159°20.006′ 25 September 2007 to 31 August 2008 Technicap PPS3/3 260 1,460
Nansen Basin 82° 38.735′ 109° 1.228′ 15 September 2011 to 15 August 2012 Kiel 285 3,600
Amundsen Basin 83° 16.230′ 124° 46.050′ 15 September 2011 to 15 August 2012 Kiel 240 4,240

2 Material and Methods

2.1 Remote Sensing

Daily averaged sea ice concentrations were retrieved at a 12.5 km resolution from the CERSAT service of the French Research Institute for Exploitation of the Sea. Snow depth on top of sea ice was retrieved at a 25 km resolution derived from the Scanning Multichannel Microwave Radiometer and the Special Sensor Microwave/Imager of the National Aeronautics Space Agency. Apparent outliers and flagged data were removed. Daily sea ice concentration and snow depth were averaged for a delimited region (1° latitude × 1° longitude) above each mooring.

2.2 Sediment Traps

Modified automatic Kiel sediment traps (0.5 m2 aperture) were deployed and recovered on board RV Polarstern in 1995, 1996, 2011, and 2012. Sample cups rotated at intervals ranging from 10 to 31 days and were filled with filtered seawater adjusted to a salinity of 40 with NaCl and poisoned with HgCl2 (0.14% final solution) to preserve samples during deployment and after recovery. Technicap PPS3/3 sediment traps (0.125 m2 aperture) were deployed and recovered during the Nansen and Amundsen Basins Observational System (NABOS) expeditions on board Kapitan Dranitsyn in 2005, 2006, and 2008 and on board Viktor Buynitskiy in 2007. Sample cups rotated at intervals ranging from 7 to 31 days and were filled with filtered seawater poisoned with formalin (5% v/v) buffered with sodium borate and adjusted to a salinity of 35 with NaCl. Because some of the moorings were recovered before the completion of the last rotation, measurements from the open collection cups were discarded from the study.

2.3 Algal Identification and Phytoplankton Carbon Estimations

Subsamples were used for the enumeration of algal cells according to the Utermöhl method (Utermöhl, 1958). A minimum of 300 algal cells were counted and identified using phase contrast microscopy at different magnifications depending on cell size. Diatoms were grouped into intact cells (containing chloroplasts), empty cells (valves only), and resting spores. Cell sizes of 20 representative cells of the different groups of phytoplankton were measured using an object micrometer to then calculate the biovolume and carbon value (pg C/cell) for all size classes according to Edler (1979). The phytoplankton carbon (PPC) content was calculated by multiplying the cell counts of individual cells with chloroplasts and resting spores by the carbon values associated to each group.

2.4 Particulate Organic Carbon Measurements

Swimmers (zooplankton actively entering the cups) were removed from samples with forceps and rinsed to retrieve all particles using a dissecting microscope. Subsamples for particulate organic carbon (POC) measurements were filtered onto GF/F filters (nominal pore size: 0.7 μm) precombusted at 500°C for 4 h. Filters were dried at 60°C, exposed to concentrated HCl fumes for 12 h or soaked in 0.1 N HCl for removal of inorganic carbon, and dried again at 60°C. POC measurements were conducted on a Perkin Elmer CHNS 2400 Series II analyzer or a Carlo Erba CHN elemental analyzer and should be considered as minimum values not corrected for dissolution of organic material. The different sediment trap models, deployment depths, and CHN analyzers potentially introduced bias when comparing the magnitude of the fluxes.

3 Results and Discussion

While small algal cells dominate the low phytoplankton biomass in the upper water column during late summer in the Central Arctic Ocean (Booth & Horner, 1997; Gosselin et al., 1997), small-sized algal groups such as flagellates, dinoflagellates, silicoflagellates, and coccolithophores were discernible only during periods of very low algal fluxes (data not shown). Instead, diatoms clearly dominated algal export at all sites sampled in the Eurasian Arctic Ocean (Figure 2). The algal flux composition at ~200 m suggests that distinct physical processes are required for the export of small cells in the stratified waters of the Central Arctic Ocean (Assmy et al., 2017; Lalande et al., 2011; Wollenburg et al., 2018). Among the diatoms observed in the sinking material, some thrive in ice and water while a few are exclusively ice-associated or exclusively pelagic (Booth & Horner, 1997; Gran, 1904; Poulin et al., 2011). In this study, the exclusively ice-associated diatoms Nitzschia frigida and Melosira arctica and the exclusively pelagic diatoms Chaetoceros spp. and Thalassiosira spp. were selected to monitor seasonal and interannual variations in algal export and to evaluate the relative contribution of ice-associated and pelagic algae to PPC and POC export in the Eurasian Arctic Ocean.

Details are in the caption following the image
Diatom and particulate organic carbon fluxes during the six mooring deployments: black indicates diatom cells with chloroplasts, grey indicates empty diatom cells, white indicates resting spores, and dashed red line indicates POC. The dashed grey areas represent the periods not sampled, the blue lines represent sea ice concentration, the grey lines represent snow depth, and the blue areas represent the melt period. Note the different diatom and POC flux scales at the Laptev Sea site.

3.1 Export of Ice-Associated Diatoms

Ice algal growth during spring depends on irradiance through snow and ice cover. Despite the attenuation properties of sea ice cover, the overlying snow layer primarily controls the magnitude of sunlight reaching the bottom ice due to its high albedo and greater capacity to scatter light (Campbell et al., 2015). Snow also insulates the bottom ice from the warming atmosphere during spring, therefore regulating bottom ice temperature. Whereas the exact mechanism leading to ice algal release following snow removal is unconfirmed, it is likely through a combination of bottom ice warming and ablation, exposure to increasing levels of irradiance, and biological ice melt, which is the absorption and conversion of radiation to heat by algal cells (Campbell et al., 2015; Fortier et al., 2002; Juhl & Krembs, 2010).

Nitzschia frigida, a chain-forming pennate diatom widely distributed in the Arctic Ocean (Poulin et al., 2011; Szymanski & Gradinger, 2016), displays a net-like structure well-suited for aggregation and rapid sedimentation (Michel et al., 1993). Nitzschia frigida was consistently among the first diatom species exported at the onset of snowmelt in the Eurasian Arctic Ocean (Figure 3a), similar to observations made in the Beaufort Sea (Dezutter et al., 2019). The collection of the rapidly-sinking N. frigida at ~200 m several weeks before ice breakup suggested that snowmelt triggered its sloughing and export (Cota et al., 1991; Fortier et al., 2002; Lalande et al., 2007; Legendre et al., 1981; Michel et al., 1993; Welch & Bergmann, 1989). While peaks in N. frigida fluxes were observed at the end of June in 1996, 2006, and 2007, N. frigida fluxes peaked in early June in the NB in 2012, suggesting that early snowmelt prior to the 2012 record ice minimum extent resulted in an early release of ice algae. While N. frigida fluxes were very low in the AB in 2012, their first collection at the end of May also suggest an earlier release of ice algae (Figure 3a). Observations made in 1957 at the US-IGY drifting station Alpha, at similar latitudes than measurements made in 2012, indicated an increase in under-ice chlorophyll a concentrations once snow cover vanished in early July (English, 1961), further highlighting the early occurrence of snowmelt in 2012.

Details are in the caption following the image
Fluxes of (a) Nitzschia frigida, (b) Melosira arctica, (c) Chaetoceros spp., and (d) Thalassiosira spp. collected during the six mooring deployments; black indicates cells with chloroplasts, grey indicates empty cells, and white indicates resting spores. The dashed grey areas represent the periods not sampled, the blue lines represent sea ice concentration, the grey lines represent snow depth, and the blue areas represent the melt period. Note the different scales. Empty panels indicate the absence of the species.

Melosira arctica, a centric diatom forming long brownish-green filaments, has commonly been observed attached to the underside of the ice or as free-floating filaments in the meltwater layer of the Arctic Ocean (Booth & Horner, 1997; Dickie, 1852; Gran, 1904; Melnikov & Bondarchuk, 1987; Poulin et al., 2014; Szymanski & Gradinger, 2016). Fluxes of M. arctica consistently peaked a few weeks after N. frigida, often in July (Figure 3b). This delayed export likely reflected the retention of M. arctica within the ice, as M. arctica cells produce large amounts of extracellular polymeric substances that increase their capacity to anchor to the underside of the ice (Krembs et al., 2011). Alternatively, the delay may be due to the formation and trapping of gas bubbles produced through photosynthesis within the mucous matrix, providing a buoyancy mechanism that maintains M. arctica afloat for a longer period (Fernández-Méndez et al., 2014).

A widespread deposition of M. arctica strands was observed on the seafloor of the AB and NB in August and September 2012 (Boetius et al., 2013), a few weeks following the collection of M. arctica in the sediment traps in July and August 2012. While M. arctica fluxes reached 46 mg C m-2 yr-1 in the NB in 2012 (Table 2), ice algal biomass deposition on the seafloor of the Arctic basins was estimated to range from 1 to 156 g Cm-2 (Boetius et al., 2013), reflecting a large spatial variability in the export of M. arctica strands that were not collected intact in the sediment traps. Elevated summertime fluxes of M. arctica at the LR in 1996 suggest that episodes of large M. arctica export occurred over the central basins before significant sea ice retreat, thinning, and increase in meltpond coverage in the climate record (Stroeve & Notz, 2018) and were likely a recurrent feature in the Eurasian Arctic Ocean, contrary to previous conclusions (Boetius et al., 2013).

Table 2. Annual Particulate Organic Carbon (POC) Fluxes, Phytoplankton Carbon (PPC) Contribution to the POC Fluxes, and Phytoplankton Carbon (PPC) Contribution of the Selected Exclusively Sympagic and Exclusively Pelagic Algae to the PPC Fluxes in the Eurasian Arctic Ocean
POC PPC N. frigida M. arctica Chaetoceros spp. Thalassiosira spp.
Mooring mg Cm-2 yr-1 mg C m-2yr-1 (% of POC flux) mg C m-2 yr-1 (% of PPC flux) mg Cm-2yr-1 (% of PPC flux) mg C m-2 yr-1 (% of PPC flux) mg C m-2 yr-1 (% of PPC flux)
Lomonosov Ridge 954 68(7.1) 6.6 (9.8) 51.3 (75.9) 0.6 (0.9) not present
Northern Laptev Sea (2005-2006) 3,858 85 (2.2) 27 (31.3) 0.1 (0.1) 0.2 (0.2) 0.9 (1.1)
Northern Laptev Sea (2006-2007) 8,661 60 (0.7) 22(36.7) 0.6 (0.9) 0.1 (0.2) 0.3 (0.5)
East Siberian Sea 395 11 (2.9) not present 6.5 (57.5) 0.02 (0.2) not present
Nansen Basin 587 85 (14.5) 17(20.0) 46 (54.1) 0.3 (0.4) 3.1 (3.6)
Amundsen Basin 527 19 (3.6) 0.3 (1.6) 2.3 (12.2) 6 (31.7) 0.7 (3.7)

3.2 Export of Pelagic Diatoms

Most of the exclusively pelagic centric diatoms Chaetoceros spp. and Thalassiosira spp. were exported as cells without chloroplasts or resting spores (Figures 3c and 3d). These resting spores or empty cells may have been incorporated into sea ice during ice formation and released during melt several months later or may have been grazed upon and exported within fecal pellets. Low fluxes of Chaetoceros spp. and Thalassiosira spp. cells with chloroplasts were also observed in June and/or July at nearly all sites. As Chaetoceros spp. and Thalassiosira spp. usually dominate spring blooms in the northeast Atlantic and on Arctic shelves (Degerlund & Eilertsen, 2010), fluxes of intact Chaetoceros spp. cells at the end of June in 2006 and 2007 in the northern LS may have reflected advection from the nearby shallow continental shelf. However, the export of intact cells in July over the NB and AB suggested moderate local under-ice pelagic production away from the continental shelf (Figures 3c and 3d). Early explorers reported that Chaetoceros spp. and Thalassiosira spp. were abundant in upper waters north of the New Siberian Islands in October 1893 (Gran, 1904) and dominated the North Pole drift station in August and early September 1939 (Usachev, 1961). Therefore, their presence at ~200 m reflected under-ice pelagic production over the basins, in agreement with historic observations.

3.3 PPC Versus POC Fluxes

The exclusively ice-associated algae N. frigida and M. arctica contributed to a large proportion of the annual PPC fluxes in the Eurasian Arctic Ocean (Figure 4 and Table 2), supporting previous observations that ice algae dominate algal export over the central basins (Bauerfeind et al., 1997; Boetius et al., 2013; Fahl & Nöthig, 2007; Gosselin et al., 1997; Gradinger, 2009; Poulin et al., 2014; Zernova et al., 2000). In contrast, exclusively ice-associated algae rarely contribute significantly to PPC fluxes on Arctic shelves and slopes (Juul-Pedersen et al., 2010). The generally low contribution of the exclusively pelagic algae Chaetoceros spp. and Thalassiosira spp. to the annual PPC fluxes may also reflect higher grazing pressure later in the productive season (Kosobokova & Hirche, 2009).

Details are in the caption following the image
Relative phytoplankton carbon (PPC) content of the main algal species and groups including the visible (mostly shell bearing cells) and intact cells (containing chloroplasts) to the particulate organic carbon (POC) fluxes collected at ~200 m in the Eurasian Arctic Ocean

Although only six years of observations were obtained from five different sites in the Eurasian Arctic Ocean, similar algal communities at these sites suggest low spatial variability over the deep basins. Relatively high annual PPC fluxes at the LR, in the northern LS, and in the NB likely reflected higher nutrient availability in these regions than in the AB and ESS (Table 2; Lalande et al., 2014). In contrast, POC fluxes were much higher near the LS shelf (Figure 2; Table 2), in agreement with previous POC flux measurements in the Central Arctic Ocean (Cai et al., 2010; Lalande et al., 2014; Roca-Martí et al., 2016). Annual PPC fluxes ranged from 0.7 to 14.5 % of annual POC fluxes in the Eurasian Arctic Ocean, clearly indicating that algal cells contribute to a small portion of the POC fluxes and that POC export must be considered with caution and even discarded as an indicator of primary production (Figure 4 and Table 2). Contrary to a previous report (Lalande, Bélanger, et al., 2009), the very high POC fluxes recorded in the northern LS in July 2007 were not the result of an increase in marine primary production associated with the sea ice minimum extent of 2007 (Figure 2). As the annual PPC flux at that site was lower in 2007 than in 2006 (Table 2) and algal fluxes were nearly absent following ice retreat (Figure 2), elevated POC fluxes in July 2007 may instead have reflected a large release of ice-rafted particulate matter during ice melt due to a complete melt in July 2007 in contrast to a partial melt in 2006 (Lalande et al., 2014; Wegner et al., 2005). Whereas POC measurements include several components of the flux ranging from terrestrial matter to fecal pellets, PPC fluxes specifically reflect a large fraction of the primary production through the rapid export of diatoms. Therefore, although PPC fluxes do not reflect small-size phytoplankton groups that tend not to sink nor the impact of zooplankton grazing, they are a better indicator of marine primary production than POC fluxes.

3.4 A Rapidly Changing Arctic Ocean

Rapid sea ice retreat, sea ice thinning, and increase in incident light (Comiso, 2012; Hill et al., 2018; Nicolaus et al., 2012; Serreze et al., 2007) have the potential to modify the magnitude, timing, and duration of the algal growth phase, and these modifications are likely already ongoing in the Eurasian Arctic Ocean where sea ice retreat has recently been considerable (Arrigo et al., 2008). Our results indicated that nearly all algal export occurred after snowmelt and before ice melt, further suggesting that snow cover regulates light transmission, algal growth, and ice algae release. Significant export of ice algae following snowmelt in 1996, before the recent reduction in ice cover extent and thickness, further support the role of snow cover. As model projections suggest either a rapid decline of Arctic snow cover accumulation and duration (Hezel et al., 2012) or an increase in Arctic snow precipitation (Bintanja & Selten, 2014), it remains unclear how changes in snow cover will affect the magnitude and timing of algal production and export over the deep basins.

Satellite-derived trends toward earlier phytoplankton blooms and increased annual primary production have been reported for the Arctic Ocean, mostly in areas with reduced ice concentrations such as the Siberian Arctic Ocean (Arrigo et al., 2008; Kahru et al., 2011). However, satellite-derived estimates exclude under-ice and subsurface production that contribute significantly to annual primary production in the Arctic Ocean (Ardyna et al., 2013; Arrigo et al., 2012; Fortier et al., 2002; Lalande et al., 2007; Mundy et al., 2009). Algal fluxes obtained in the Eurasian Arctic Ocean further showed that a large proportion of algal export occurred before complete ice melt and possible detection by satellites. Despite the mooring sites being located at higher latitudes in 2012, an early release of ice algae was observed over the NB and AB. Hence, sediment trap-derived fluxes hint at a trend toward earlier ice algal export taking place several weeks prior to satellite-derived observations. As benthic megafauna and sediment bacteria of the deep Arctic basins benefit from ice algae deposition (Boetius et al., 2013), early ice algal export may affect benthic biomass and remineralization rates. These results show the importance of maintaining long-term observatories to improve our understanding of biogeochemical fluxes in the Arctic Ocean.


We thank the crew of RV Polarstern (expeditions ARKXI-1, ARKXII, ARKXXVI-3, and ARKXXVII-3), Kapitan Dranitsyn, and Viktor Buynitskiy for their support at sea. Logistical and financial support was provided by AWI and ArcticNet. We thank the mooring teams and technicians at AWI and ArcticNet for fieldwork and laboratory support. This is a joint contribution to AWI, the Long-Term Oceanic Observatories project, and the Nansen and Amundsen Basins Observational System (NABOS) project. Data are available at https://doi.pangaea.de/10.1594/PANGAEA.901800.