Export of Algal Biomass from the Melting Arctic Sea Ice

See allHide authors and affiliations

Science  22 Mar 2013:
Vol. 339, Issue 6126, pp. 1430-1432
DOI: 10.1126/science.1231346

Diatom Fall

2012 saw the greatest Arctic ice minimum ever recorded. This allowed unprecedented access for research vessels deep into the Arctic Ocean to make high-latitude observations of ice melt and associated phenomena. From the RV Polarstern between 84° to 89° North, Boetius et al. (p. 1430, published online 14 February; see the cover) observed large-scale algal aggregates of the diatom Melosira arctica hanging beneath multiyear and seasonal ice across a wide range of latitudes. The strands of algae were readily dislodged and formed aggregates on the seabed up to 4400 meters below, where the algae are consumed by large mobile invertebrates, such as sea cucumbers and brittle stars. Although Nansen observed sub-ice algae in the Arctic 100 years ago, the extent of this bloom phenomenon was unknown. The dynamics of such blooms must impinge on global carbon budgets, but how the dynamics will change as ice melt becomes more extensive remains unclear.


In the Arctic, under-ice primary production is limited to summer months and is restricted not only by ice thickness and snow cover but also by the stratification of the water column, which constrains nutrient supply for algal growth. Research Vessel Polarstern visited the ice-covered eastern-central basins between 82° to 89°N and 30° to 130°E in summer 2012, when Arctic sea ice declined to a record minimum. During this cruise, we observed a widespread deposition of ice algal biomass of on average 9 grams of carbon per square meter to the deep-sea floor of the central Arctic basins. Data from this cruise will contribute to assessing the effect of current climate change on Arctic productivity, biodiversity, and ecological function.

Primary productivity in the central Arctic is limited by light and nutrients. Photosynthetically active radiation (PAR) for under-ice primary production is only available from May to August but is locally restricted by ice thickness and snow cover (14). Owing to stratification (5, 6), the mixed layer depth is limited to 10 to 30 m in summer (Table 1), which constrains the nutrient supply for algal growth (7). Hence, average estimates for primary production (PP) in the ice-covered central Arctic are low, on the order of 1 to 25 g C m−2 year−1 (8, 9). The contribution of ice algae is not well constrained, ranging from 0 to 80% (1013). However, as a consequence of Arctic warming, primary productivity in and under the ice may be boosted by higher light transmission through thinning sea ice (3, 14, 15) and the increase in melt-pond coverage during summer (4, 16).

Table 1

Distribution of algal aggregates and characteristics of sea-ice stations investigated. Methods are provided in the supplementary materials. Where available, averages and standard deviations are given. FYI, first-year ice; MYI, multiyear ice; n.d., not determined.

View this table:

Assessing the consequences of current climate change in the central Arctic regions remains difficult because reliable baselines for Arctic productivity, biodiversity, and ecological function are lacking [reviewed in (17)]. During the 2012 sea-ice minimum, research vessel (RV) Polarstern visited the ice-covered eastern-central basins between 82° to 89°N and 30° to 130°E (Fig. 1). In this area, thick multiyear sea ice has been largely lost as a result of melt by atmospheric heat (18). Our airborne electromagnetic measurements confirmed that first-year ice dominated (>95%), with an average modal thickness of less than a meter and a melt-pond cover of 30 to 40%.

Fig. 1

Ice conditions during RV Polarstern Expedition IceArc (ARK27-3, 2 August to 8 October 2012).

(A) Ice cover in July 2012 in percentages. Ice stations with fresh and degraded algal deposits are marked by green and brown circles, respectively. White indicates no deposits. (B) Aerial image of station 3 in mid-August. (C) Aerial image of station 6 in mid-September.

Previous investigations of the underside of Arctic sea ice found that the diatom Melosira arctica grows meter-long filaments, anchoring in troughs and depressions under ice floes and covering up to 40 to 80% of the underside of undisturbed ice floes (12, 1924) (Fig. 2). Warming and melting leads to their rapid sedimentation (2023). Deposition of Melosira strands had been observed on the sea floor of Arctic shelves (12, 21), but their contribution to carbon export in the ice-covered basins remains unknown (25, 26). Particulate organic carbon flux to the deep sea, measured by sea-floor carbon demand (25) and by sediment traps moored in the Amundsen Basin (27), was around 1 g C m−2 year−1 (>1500 m) in the 1990s, with a peak contribution of sub-ice algae of up to 28% in August (27). Repeated measurements during the first Arctic-wide sea-ice minimum in 2005–2007 showed an increased carbon flux of 6.5 g C m−2 year−1 (850 m), peaking in July (28).

Fig. 2

M. arctica aggregations. Strands (~20 cm) of Melosira (A) under ice (station 7), (B) recovered from the sea floor (station 7), and (C) photographed in situ with K. hyalina grazing on deposits (station 3). (D to F) Microscopic images of Melosira cells from (A), (B), and (C) (extract of Kolga gut), respectively.

During the expedition IceArc in summer 2012, we observed in seven out of eight regions sea-floor deposits of fresh M. arctica strands and other sub-ice algae at 3500- to 4400-m water depth (Fig. 1, fig. S1, and movies). Patches of algae of 1 to 50 cm in diameter covered up to 10% of the sea floor. This attracted opportunistic megafauna—such as the deep-sea holothurians Kolga hyalina (29) and Elpidia heckeri and the ophiurid Ophiostriatus striatus—which were observed to feed on the Melosira strands. Based on their color, chlorophyll a content, and chloroplast morphology, the freshest algal deposits were observed at the northernmost stations, 7 and 8 (>87°N). Stations 4 to 6 (82° to 85°N), north of the Laptev Sea margin, showed degraded algal deposits. In this area, megafauna biomass was substantially elevated, as was the pigment concentration of holothurian gut content (Table 1). The larger body sizes (>6 cm) and apparent fecundity of the Kolga population (based on gonad sizes) in this area suggested that sources of food had been available for at least 2 months and that the main algal flux had occurred before June. This matches observations of rapid melt and export of ice from the Laptev Sea as early as May 2012. By July, large open water areas had appeared within the ice zone up to 85°N (Fig. 1), causing a rapid decline of the sea-ice cover, reflected in 1 to 2 m of melt-water content above the winter thermocline (Table 1).

Our surveys showed shreds of M. arctica (Table 1), indicating their melt-out earlier in the season (23). At 3500- to 4400-m depth, deposits of coiled Melosira strands (diameters of 5 to 12 cm) covered 0.1 to 10% of the sea floor. The carbon deposition by sub-ice algae was estimated to be equivalent to 1 to 156 g C m−2 (median 9 g C m−2) (Table 1). For comparison, the 2012 pelagic new production in the same regions was estimated to be 7 to 16 g C m−2 (median 11 g C m−2) (Table 1), with a contribution by diatoms of 36% based on silicate inventories (Table 1). Melosira strands are not used as food in the pelagial and sink rapidly to the sea floor (23). This results in a contribution of at least 45% of total primary production and >85% of carbon export in 2012.

The algal deposits at the sea floor and extracts of Kolga gut at stations 3, 4, 7, and 8 contained living Melosira cells with green chloroplasts and lipid vesicles (Fig. 2). The algal deposits had variable high concentrations of chloroplast pigment equivalents (CPE) (27 ± 21 μg cm−3; n = 18 aggregate samples) and a high chlorophyll a to total pigment ratio (51 ± 18%). In comparison, pigment contents of bare sediments next to the patches were low at 0.8 ± 0.3 μg cm−3, matching concentrations found in the 1990s (25). The gut contents of Kolga specimens showed even higher pigment concentrations of, on average, 51 ± 47 μg cm−3 (Chla/CPE ratio of 41 ± 14%; n = 15 gut samples), and algae recovered from guts were still photosynthesizing when exposed to light (30).

Previous investigations focusing on oligotrophic deep-sea sediments have found a direct relationship between carbon flux, benthic biomass, and remineralization rates (3135). However, despite the widespread deposition of algae observed in the eastern-central basins, apparently only sediment bacteria (as estimated from respiration rates) (fig. S2) and large mobile megafauna had profited from the ice-algae deposition. Infauna burrows and tubes were rare, indicating an absence of the sediment-dwelling macrofauna characteristic of other deep-sea basins with seasonally sedimenting phytoplankton blooms [reviewed in (36)]. Furthermore, the bare sediments next to the algal deposits maintained oxygen fluxes of only 0.3 to 0.4 mmol O2 m−2 day–1, equivalent to a carbon demand of 1 to 2 g C m−2 year−1. Such low rates are typical for oligotrophic deep-sea sediments (37, 38) and match carbon export fluxes measured in the 1990s in this area (25, 27). In contrast, in situ and ex situ microprofiling of diffusive oxygen fluxes into sediments covered by algal aggregates showed elevated rates of 5 to 6 mmol O2 m−2 day−1, equivalent to carbon fluxes of 25 g C m−2 year−1 (stations 7 and 8) (fig. S2). This suggests considerable microbial respiration (13 to 60%) of the algal carbon input. Accordingly, in cores covered by Melosira strands, oxygen penetration in the sediment was reduced to a few millimeters compared with the surrounding sediment, where oxygen penetrated >50 cm (fig. S1). Hence, if high exports of sea-ice algae had occurred regularly before 2012, oxygen penetration depth would have been less than observed, independent of the fresh Melosira deposits (30). Hence, we conclude that massive algal falls were rare.

Arctic climate models predict a further decline in the sea-ice cover, toward a largely ice-free summer in the Arctic in coming decades (39). Our observations support the hypothesis (14) that the current sea-ice thinning and increasing melt-pond cover may be enhancing under-ice productivity and ice-algae export, with ecological consequences from the surface ocean to the deep sea.

Affiliations and contributions of the RV Polarstern ARK27-3-Shipboard Science PartyWriting team: Antje Boetius1,2,3 with all coauthors. Ice physics and remotely operated vehicle (ROV) surveys: Stefan Hendricks,1 Christian Katlein,1 Thomas Krumpen,1 Marcel Nicolaus1; Sea-ice biology: Mar Fernández-Méndez,1,2 Ilka Peeken1,3; Oceanography and nutrients: Karel Bakker,5 Catherine Lalande,1 Benjamin Rabe,1 Raquel Somavilla1; Deep-sea surveys, sampling and measurements: Sebastian Albrecht,4 Christina Bienhold,1 Antje Boetius,1 Janine Felden,1 Antonina Rogacheva,6 Elena Rybakova,6 Frank Wenzhöfer1

Fieldwork and scientific discussions: Shipboard Scientific Party, including other contributors listed in the supplementary materials.1Alfred Wegener Institute, Helmholtz Center for Polar and Marine Research, 27515 Bremerhaven, Germany.2Max Planck Institute for Marine Microbiology, 28359 Bremen, Germany.3MARUM, Center for Marine Environmental Sciences, University Bremen, 28334 Bremen, Germany.4FIELAX Gesellschaft für wissenschaftliche Datenverarbeitung mbH, 27568 Bremerhaven, Germany.5NIOZ Royal Netherlands Institute for Sea Research, 1790 AB Den Burg, Netherlands.6P. P. Shirshov Institute of Oceanology, Russian Academy of Sciences, 117997 Moscow, Russia.

Supplementary Materials

Materials and Methods

Figs. S1 and S2

Movies S1 and S2

References (4050)

References and Notes

  1. Materials and methods are available as supplementary materials on Science Online.
  2. Acknowledgments: We thank the captain and crew of RV Polarstern expedition IceArc (ARK27-3) as well as our helicopter and meteorology teams for their excellent support with work at sea. This study was funded by the PACES (Polar Regions and Coasts in a Changing Earth System) program of the Helmholtz Association. Additional funds were made available to A.B. by the European Research Council Advanced Investigator grant 294757 and the Leibniz program of the Deutsche Forschungsgemeinschaft, and to B.R. for the Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie project 03F0605E. Supplementary data are available at
View Abstract

Stay Connected to Science

Navigate This Article