During a year-round occupation of Amundsen Gulf in the Canadian Arctic Archipelago dissolved inorganic and organic carbon (DIC, DOC), total alkalinity (TA), partial pressure of CO 2 (pCO 2 ) and related parameters were measured over a full annual cycle. A two-box model was used to identify and assess physical, biological, and chemical processes responsible for the seasonal variability of DIC, DOC, TA, and pCO 2 . Surface waters were undersaturated with respect to atmospheric CO 2 throughout the year and constituted a net sink of 1.2 mol C m 22 yr 21 , with ice coverage and ice formation limiting the CO 2 uptake during winter. CO 2 uptake was largely driven by under ice and open-water biological activity, with high subsequent export of organic matter to the deeper water column. Annual net community production (NCP) was 2.1 mol C m 22 yr 21 . Approximately one-half of the overall NCP during the productive season (4.1 mol C m 22 from Apr through Aug) was generated by under-ice algae and amounted to 1.9 mol C m 22 over this period. The surface layer was autotrophic, while the overall heterotrophy of the system was fueled by either sedimentary or lateral inputs of organic matter.

[1] From sea-ice formation in November 2007 to onset of ice melt in May 2008, we studied the carbonate system in first-year Arctic sea ice, focusing on the impact of calcium-carbonate (CaCO 3 ) saturation states of aragonite (XAr) and calcite (XCa) at the ice-water interface (UIW). Based on total inorganic carbon (C T ) and total alkalinity (A T ), and derived pH, CO 2 , carbonate ion ( [CO 3 22 ]) concentrations and X, we investigated the major drivers such as brine rejection, CaCO 3 precipitation, bacterial respiration, primary production and CO 2 -gas flux in sea ice, brine, frost flowers and UIW. We estimated large variability in sea-ice C T at the top, mid, and bottom ice. Changes due to CaCO 3 and CO 2 -gas flux had large impact on C T in the whole ice core from March to May, bacterial respiration was important at the bottom ice during all months, and primary production in May. It was evident that the sea-ice processes had large impact on UIW, resulting in a five times larger seasonal amplitude of the carbonate system, relative to the upper 20 m. During ice formation, [CO 2 ] increased by 30 mmol kg 21 , [CO 3 22 ] decreased by 50 mmol kg 21, and the XAr decreased by 0.8 in the UIW due to CO 2 -enriched brine from solid CaCO 3 . Conversely, during ice melt, [CO 3 22 ] increased by 90 mmol kg 21 in the UIW, and X increased by 1.4 between March and May, likely due to CaCO 3 dissolution and primary production. We estimated that increased ice melt would lead to enhanced oceanic uptake of inorganic carbon to the surface layer.

The Arctic sea-ice-scape is rapidly transforming. Increasing light penetration will initiate earlier seasonal primary production. This earlier growing season may be accompanied by an increase in ice algae and phytoplankton biomass, augmenting the emission of dimethylsulfide and capture of carbon dioxide. Secondary production may also increase on the shelves, although the loss of sea ice exacerbates the demise of sea-ice fauna, endemic fish and megafauna. Sea-ice loss may also deliver more methane to the atmosphere, but warmer ice may release fewer halogens, resulting in fewer ozone depletion events. The net changes in carbon drawdown are still highly uncertain. Despite large uncertainties in these assessments, we expect disruptive changes that warrant intensified long-term observations and modelling efforts.

Antarctic pack ice is inhabited by a diverse and active microbial community reliant on nutrients for growth. Seeking patterns and overlooked processes, we performed a large-scale compilation of macro-nutrient data (hereafter termed nutrients) in Antarctic pack ice (306 ice-cores collected from 19 research cruises). Dissolved inorganic nitrogen and silicic acid concentrations change with time, as expected from a seasonally productive ecosystem. In winter, salinity-normalized nitrate and silicic acid concentrations (C*) in sea ice are close to seawater concentrations (C w ), indicating little or no biological activity. In spring, nitrate and silicic acid concentrations become partially depleted with respect to seawater (C* < C w ), commensurate with the seasonal build-up of ice microalgae promoted by increased insolation. Stronger and earlier nitrate than silicic acid consumption suggests that a significant fraction of the primary productivity in sea ice is sustained by flagellates. By both consuming and producing ammonium and nitrite, the microbial community maintains these nutrients at relatively low concentrations in spring. With the decrease in insolation beginning in late summer, dissolved inorganic nitrogen and silicic acid concentrations increase, indicating imbalance between their production (increasing or unchanged) and consumption (decreasing) in sea ice. Unlike the depleted concentrations of both nitrate and silicic acid from spring to summer, phosphate accumulates in sea ice (C* > C w ). The phosphate excess could be explained by a greater allocation to phosphorus-rich biomolecules during ice algal blooms coupled with convective loss of excess dissolved nitrogen, preferential remineralization of phosphorus, and/or phosphate adsorption onto metal-organic complexes. Ammonium also appears to be efficiently adsorbed onto organic matter, with likely consequences to nitrogen mobility and availability. This dataset supports the view that the sea ice microbial community is highly efficient at processing nutrients but with a dynamic quite different from that in oceanic surface waters calling for focused future investigations.