Estimates of nutrient demand by dense mats of ice algae in the high Arctic indicate that substantial nutrient fluxes are necessary to satisfy the observed growth over the 2-month bloom. In our study area, Barrow Strait, the quantity of nutrients in the surface-mixed layer is about 3-10 times greater than estimates of total demand during the bloom, and nutrient fluxes in the water column are estimated to be of the same order of magnitude as algal demand. The fluxes in the water column are predicted to vary by more than an order of magnitude over the fortnightly tidal cycle, assuming that fluxes depend upon the strength of tidal currents and the vertical nutrient gradients. In the latter half of the bloom, when biomass levels are high, it appears that established populations of ice algae may experience cyclic conditions of nutrient limitation during neap tides when nutrient fluxes are minimal. Contributions from regeneration and brine exclusion from the ice sheet appear to satisfy only a portion of the bloom's total requirement for nutrients. INTRODUCTIONDense growths of ice algae are a ubiquitous feature beneath annual sea ice in polar regions during the springtime. In the high Arctic these microalgae are largely concentrated at the ice-water interface on or within the skeletal layer of the congelation ice. MacroScopically, they resemble a mottled goldenbrown carpet about 1-2 cm thick. We have found that these algae grow for at least 2 months (April-May) in the central portion of the Northwest Passage. Moreover, the ice alga. e often achieve high levels of biomass (80-100 mg chlorophyll m -e) that are similar to water column values integrated over the depth of the euphotic zone in many planktonic systems.

Abundance and photosynthetic activity of ice algae in Resolute Passage in the Canadian high Arctic were measured in relation to in situ irradiance throughout the main growth season in 1985 and 1986. A simple model was used to calculate in situ production rates and the theoretical maximum (light limited) size of crops and production rates. Both the observed and the maximum possible crop sizes and production rates varied directly with irradiance over the natural range of snow cover, and crops attained the theoretical maximum imposed by self-shading (77 to 225 mg m-' chlorophyll a under thin snow cover) in both years. Calculated in situ production of ice algae under thin snow cover (S to 23 gC m-2 yr-l) could equal or exceed typical values for Arctic plankton. Comparison against observed biomass accumulation in the ice indicated that as much as 65 % of the production could be exported from the ice during the growth season. Where light was artificially increased by maintaining snow-free areas, observed crops were much less than the theoretical maximum despite the abscence of photosynthetic photoinhibition. Crops reported from some other arctic sites were also much less than their corresponding theoretical maxima. Low irradiance often limits ice algal growth, but our results suggest that losses associated with excessive irradiance and with grazing by amphipods at near-shore sites are additional factors determining algal abundance and production.

Except in "bottom ice" (lowest few centimetres) and surface waters impacted by autotrophs, the major inorganic nutrients behave conservatively in seawater and sea ice. From mid- to late spring, steep and persistent nutrient gradients were observed in the "well-mixed surface layer" with minima near the ice–water interface. Nitrate, ammonium, and phosphate are highly concentrated in heavily colonized bottom ice relative to seawater and the remainder of the ice sheet; concentrations in darkened, weakly colonized bottom ice are similar to the ice sheet. These nutrients also display strong vertical stratification over millimetre scales. Nitrate and phosphate in the bottom ice layer display strong positive relationships with chlorophyll. The accumulation of these nutrients in bottom ice must be biologically mediated and constitutes a significant sink. In contrast, silicic acid concentrations in bottom ice are close to those expected for sea ice formed from the source seawater, are only weakly related to algal biomass, and vary much less seasonally. Ice algae are apparently shocked osmotically and release their intracellular pools of dissolved nutrients. Intracellular pools of nitrate averaged 1.4–9.5% of total particulate nitrogen. Nitrient stresses, during periods of high biomass and sluggish supply, may be alleviated by pooling.