Future arctic Ocean Seasonal ice Zones and implications for pelagic-Benthic Coupling

USAGE Permission is granted to copy this article for use in teaching and research. Republication, systematic reproduction, or collective redistribution of any portion of this article by photocopy machine, reposting, or other means is permitted only with the approval of The Oceanography Society. Send all correspondence to: info@tos.org or The Oceanography Society, PO Box 1931, Rockville, MD 20849-1931, USA. Oceanography THE OFFICIAl MAGAzINE OF THE OCEANOGRAPHY SOCIETY

The display of color in the southern Barents Sea north of Norway reveals a bloom of tiny marine plants called phytoplankton.The colors can be produced by a variety of pigments, including chlorophyll, that the plants use to harness sunlight for photosynthesis.This moderate resolution imaging Spectroradiometer (mOdiS) image was captured by the aqua satellite on July 19, 2003.From http://visibleearth.nasa.gov/view_rec.php?id=16521 iNTrOduCTiON In the Arctic Ocean, productivity, biogeochemical cycling, and pelagic-benthic coupling (i.e., the link between pelagic and benthic systems) are primarily determined by the distribution, thickness, and melt dynamics of sea ice.Global warming is reflected in decreased extent and thickness of sea ice in the Arctic Ocean (Comiso, 2003;Kwok and Rothrock, 2009).Ice cover extent has shown an overall negative trend for 1979-2006 (Stroeve et al., 2007), decreasing at an average rate of 10% per decade (Comiso et al., 2008;Polyakov et al., 2010).
Melting accelerated in 2007, but a slower and still negative trend was re-established in 2008-2010. Nonetheless, the Arctic Ocean may be largely ice-free in late summer in two to three decades, with a winter cover consisting mainly of firstyear ice.In addition, the average ice thickness has decreased steadily, and the Arctic Ocean may have lost over 50% of its sea ice volume (Kwok and Rothrock, 2009).Simultaneously, freshwater inputs have increased (McPhee et al., 2009;Yamamoto-Kawai et al., 2009), and ice transport toward Fram Strait has grown in both volume and velocity (von Eye et al., 2009).Along with the reduction of sea ice cover, there has been an increase in the area of low surface salinity and stratification in the pan-Arctic meltwater band (i.e., the seasonal ice zone).These changes have been accompanied by surface-layer warming.The amount of photosynthetic and UV radiation that reaches the water column has increased in most of the seasonal ice zone.
Conversely, total incident radiation may have decreased in shelves that receive river discharges and diffuse runoff from land because of increased particle content owing to permafrost melting on land.
Thus, the entire physical forcing that determines productivity, biogeochemical cycling, and the relationship between algae and grazers has already changed and continues to do so steadily (see Box 1).Clearly, these environmental alterations have also had an impact on pelagic-benthic coupling (e.g., export of fresh algae, zooplankton fecal pellets, and other detritus from the water column to the bottom).The largest share of pelagic-benthic coupling is composed of vertical export or gravitational flux.
Vertical flux depends, first of all, on primary production, but also upon the feeding intensity of zooplankton in the upper layers that simultaneously impoverishes their feeding grounds (Wassmann, 1998).Vertical flux of organic matter also influences atmospheric carbon dioxide drawdown, which is of global significance in the Arctic Ocean (Bellerby et al., 2005).
When the Arctic Ocean is subjected to new physical regimes and rapid changes, ecological responses and adjustments of both pelagic and bottom-dwelling organisms as well as changes in biogeochemical cycling are likely to ensue.aBSTr aCT.Despite concerns about rapid changes in Arctic Ocean physical forcing and ecosystem function, quantitative knowledge and time series are scarce.
The number of reliable physical-biological coupled models and models based on remote sensing is small.To improve our comprehension of carbon flux in the most prominent Arctic Ocean feature, the seasonal ice zone, a possible first step is to evaluate how biogeochemical cycling might develop in the future by examining conceptual models that address climate warming and seasonality in ecosystem development.Here we present three conceptual models of biogeochemical cycling and climate warming in the seasonal ice zone of the Arctic Ocean.They are designed to enhance, in a conceptual and semiquantitative manner, understanding of the possible temporal sequence of future primary production development, its spatial variation, and food availability in the most productive part of the future Arctic Ocean, including pelagic-benthic coupling.We speculate that the largest changes will take place in (a) the northern portions of today's seasonal ice zone, which will expand to cover the entire Arctic Ocean, and (b) the southern portions, which will be exposed to more thermal stratification.The former change increases and the latter change decreases productivity and supply to the bottom.Lack of nutrient availability means that new production in the central Arctic Ocean will remain low.Blooms of ice and plankton algae may start earlier, depending on snow cover, providing more continuity in food supply for grazers in the upper water column.Weakening of today's highly episodic primary production in the seasonal ice zone will result in lower average food concentrations for pelagic heterotrophs.We suggest that more of the available energy will be recycled in the pelagic zone, and that vertical export of biogenic matter will be less variable and of reduced quality.
Paul Wassmann (paul.wassmann@uit.no)(Wassmann et al., 2011).The future ecology of the seasonal ice zone and pelagic-benthic coupling are thus difficult to depict, let alone predict.In addition, the Arctic Ocean entire region is to apply mathematical models developed from and validated by existing measurements of the physical, chemical, and biological oceanography from areas that have been investigated.
The number of well-documented changes in planktonic and benthic systems in the Arctic Ocean is surprisingly low (Wassmann et al., 2011).Likewise, the number of available physical-biological coupled models for the region is low, but this field is rapidly developing (e.g., Popova et al., 2010;Zhang et al., 2010;Slagstad et al., 2011).Despite the alarming nature of warming and its potentially strong effects on the Arctic Ocean, little research is being done to evaluate the impacts of climate change in a balanced manner over the entire region.

gOal aNd iNTeNTiONS
To determine how productivity and pelagic-benthic coupling in the Arctic Ocean will evolve in the near future, and to develop more realistic mathematical models, we apply future ecosystem scenarios as the basis for designing suitable dedicated investigations.Conceptual models for the Arctic Ocean have been presented previously (e.g., Hunt and Stabeno, 2002;Carmack and Wassmann, 2006;Leu et al., 2011;Wassmann, 2011).Some of them are re-edited and compiled here, with the goal of shedding light on the fate of carbon in the Arctic Ocean by focusing upon rapid transitions and future ecosystem states, in particular.To this end, we display semiquantitative scenarios that focus on the physical forcing of primary production and pelagic-benthic coupling.
For a "bird's-eye" perspective of future primary production in the Arctic Ocean, see Box 3. not included in the scenarios presented here.An all-encompassing ecosystem development theory for the Arctic Ocean has still to be developed.
how Will global Warming Change the Timing of primary production in the ice-Covered arctic Ocean?
The growth of ice algae depends primarily on light availability, as determined by solar angle, ice thickness, and snow cover.Ice algae production is patchy and highly variable, averaging 5-10 g C m -2 yr -1 ; the concomitant production of Arctic phytoplankton is higher, averaging 12-50 g C m -2 yr -1 , depending on latitude and the duration of ice-free periods (Legendre et al., 1992;Gosselin et al., 1997).In areas with more extensive ice cover, ice algae are of comparatively greater importance.In the central Arctic Ocean's multiyear ice pack, for instance, ice algae contribute, on average, 57% of the total primary production (Gosselin et al., 1997) almost all life on earth is directly or indirectly reliant on primary production.
The organisms responsible for primary production are known as primary producers or autotrophs, and they form the base of the food chain.in aquatic ecoregions, algae are primarily responsible for primary production.We distinguish net and gross primary production.Net primary production is the dynamic balance between gross primary production and cell respiration.
Net primary production creates the base of new production that is determined by the availability of the limiting nutrient (e.g., nitrate; Eppley and Peterson, 1979).Organisms ultimately metabolize nitrogenous organic molecules, which are returned to the water column as ammonium, in a process known as regeneration.Total primary production is thus comprised of new production (nitrate) and regenerated production (ammonium).
New production can also be designated harvestable production (i.e., the maximum biomass that can be extracted from the system without destroying its carrying capacity).The balance between gross and net primary production has direct implications for ecosystems, biogeochemical cycling, pelagicbenthic coupling, and fisheries.The maximum marine harvest and the annual vertical carbon export from the upper layers (also termed export production) are limited upward by net or new production.
meter-long threads that hang from multiyear ice.Their existence in an environment where heterotrophs' need for food far exceeds the amount of food available (Olli et al., 2007) implies that these algae must be difficult for planktivores to graze or digest.However, for organisms at the seafloor, these ice algae are reported to constitute a food source (e.g., Carroll and Carroll, 2003).Further information on annual and seasonal phytoplankton production in several Arctic Ocean ecosystems can be found in, for example, Sakshaug (2004), Tremblay et al. (2006), andAppolonio andMatrai (2010).Two distinct pulses of biogenic matter panel F projects future primary production at 70°N after global warming leads to increasing thermal stratification and decreased primary production.Modified from Figure 1 in Leu et al. (2011) dominated by autotrophic cells (green) are followed by less-distinct pulses of degrading matter (reddish).
Climate warming extends the icefree period, with melt starting earlier in spring and freeze up occurring later in fall (Figure 2B).Consequently, both the Arctic ice and plankton algae blooms start earlier (Kahru et al., 2011;Perrette et al., 2011) rich benthos (e.g., Piepenburg et al., 2011) on the shelves.Consequently, the heavily phased nature of today's seasonal ice zone will be replaced by more evenly distributed (nutrient-limited) primary production and food availability.
During winter, by far the longest period of the year, when the ice is thick and snow-covered, and the sun is below the horizon, productivity is negligible, suspended biomass accumulation is minimal, and vertical carbon export from the upper layers is usually very low (Olli et al., 2002;Forest et al., 2008).
The fate of the ice algae blooms in early spring is still not well understood, but they are clearly an important food source for zooplankton (Leu et al., 2011) and benthos (Carroll and Carroll, 2003).It is generally assumed that a considerable amount of the biomass sinks ungrazed as high-quality input to the benthos (Figure 2A).As the season proceeds, with more light, ice melting, and the development of a pelagic bloom in the marginal ice zone (Reigstad et al., 2002;Søreide et al., 2010;Leu et al., 2011), grazing intensity will determine this bloom's fate.High grazing impact can reduce vertical export in terms of carbon and the quality of the settling organic matter, but usually a peak in vertical carbon export is observed following the pelagic bloom and its large component of phytoplankton cells (Figure 2A; Olli et al., 2002;Reigstad et al., 2008).The post-bloom and summer periods are characterized by regenerated production, and they are controlled by stratification and recycling of nutrients in the euphotic zone.This ecosystem structure facilitates recycling, allowing less material to be exported below the euphotic zone, where nutrients will then be lost to further incorporation (Wassmann, 1998).Investigations in the Barents Sea seasonal ice zone revealed that the vertically exported material is more degraded during this period (Figure 2A; Reigstad et al., 2008).Zonation, functional ocean types (alpha/beta), and ice cover are indicated at the top.primary production, the depth of the euphotic zone, and the mixed layer are also shown, as are phytoplankton concentrations and ice algae.Below each scenario, the principal profiles of vertical export of particulate organic carbon for winter (red), spring bloom/ episodic mixing (green), and summer (blue) are illustrated in a semiquantitative manner.The stippled lines indicate variability in phytoplankton bloom strength (ii) and the result of episodic mixing (iii).(B) The generic scheme after shrinkage and disappearance of the multiyear ice zone (i) in a future arctic Ocean.The entire arctic Ocean turns into a seasonal ice zone (ii).erosion of stratification in the outer section of the seasonal ice zone results in expansion of the vertically mixed alpha ocean northwards (iii), but there is a simultaneous increase of thermal stratification in the south, which is a new and presently non-existent scenario (iV).Below each scenario the principal profiles of vertical export of particulate organic carbon for winter (red), spring bloom/episodic mixing (green), and summer (blue) are illustrated in a semi-quantitative manner.Partly redrawn and modified from Carmack and Wassmann (2006) and Wassmann (2011) most productive region of the southern Barents Sea toward the end of the century (Slagstad et al., 2011).
Vertical carbon export and retention of sinking material in the upper layers create the characteristic vertical flux curve, which is described as an attenuation curve, similar to light attenuation (Wassmann et al., 2003;Buesseler et al., 2007).Grazing by heterotrophic organ- • Increased radiation will increase new (or harvestable) production only to a limited degree in the arctic Ocean (e.g., Slagstad et al., 2011) • More consumption than production in parts of the Arctic Ocean may be caused by: 1. increased import of mesozooplankton from warming subarctic regions (e.g., Olli et al., 2007) 2. increased respiration of heterotrophs (in particular, microbes;Vaquer-Sunyer et al., 2010) 3. relatively small increases in primary production (low supply of nutrient from nutrient-rich waters) • With increasing temperature, respiration increases much faster than primary production, creating a scenario where the arctic Ocean turns from a sink into a producer of atmospheric carbon dioxide (e.g., Vaquer-Sunyer et al., 2010;Kritzberg et al., 2010) • With increasing temperature and decreasing salinity, the cell size of autotrophs decreases, providing more strength to the microbial loop subregions (e.g., Tremblay and Gagnon, 2009).The processes that make nutrients available to the euphotic zone-mixing, diffusion, tidal movements, and wind stress-also need considerable attention (e.g., Sundfjord et al., 2007).

Remotely sensed information on phytoplankton distribution in the Arctic
Ocean (e.g., Pabi et al., 2008) needs quality control and validation, and the algorithms that convert chlorophyll into primary production need to be carefully considered for the specific Arctic Ocean conditions.The few existing physicalbiological coupled models that cover the Arctic Ocean (e.g., Popova et al., 2010;Zhang et al., 2010;Slagstad et al., 2011) must be improved with regard to physical forcing (e.g., how available light for photosynthesis is affected by low sun angle, an atmosphere with high probability for fog and haze, the thickness of ice, and variable snow cover).There is also a need for more information on temperature-dependent respiration and metabolism at low, but increasing, temperatures (e.g., Kritzberg et al., 2010;Vaquer-Sunyer et al., 2010), and on how key zooplankton conquer new or lose former habitats (e.g., Kosobokova et al., 2011).The processes depicted in

Future
arctic Ocean Seasonal ice Zones and implications for pelagic-Benthic Coupling B Y pa u l Wa S S m a N N a N d m a r i T r e i g S Ta d BOx 1 | phYSiCal FOrCiNg iNFlueNCiNg CarBON Flux iN The arCTiC OCeaN: aN OVerVieW annual primary production by ice algae and phytoplankton in the seasonal ice zone is determined by nutrient availability (generally low winter accumulated concentrations, except in regions of advection or shelf breaks), light (determined primarily by ice, snow cover, and atmospheric conditions), upper-layer stratification (depending mainly on ice melt, but in some regions also on river discharge), and type of algae present (ice or phytoplankton algae).Based upon predicted changes in climate, we can identify factors that would be expected to increase primary production in the future arctic Ocean: • Episodic nutrient availability (upwelling at shelf breaks and low-pressure passages) • Increased light availability due to ice melt and reduced snow cover (due to rain and warm-weather spells) • Increased nutrient discharge from rivers By the same token, we can identify factors that would decrease primary production: • Increased stratification (ice melt and river discharge) • Increased denitrification on the shallow shelves in the Pacific sector • Decrease in incident light (more cloudy weather in the low-pressure belt) • Increased turbidity in river discharge regions (permafrost melt, beach erosion, river discharge, wind-driven resuspension) is complex, composed of five ecosystems that contribute to productivity and biogeochemical cycling.There are three different shelf ecosystems types (Carmack and Wassmann, 2006): inflow (Barents and Bering Seas), outflow (Fram Strait and Canadian Archipelago), and interior (Siberian and Beaufort shelves).And, there are two deep basins (the Nansen/Amundsen and Canadian Basins), separated by the Lomonosov Ridge, that function differently.In a region as remote, vast, and inaccessible as the Arctic Ocean, the only practical method for addressing climate change and primary production over the How can pan-Arctic changes in primary production (see Box 2 for terminology details) and ecosystem function be described and understood?This timely question is presently difficult, if not impossible, to answer adequately.Historical impediments-practical and political-have prevented intensive research in the Arctic so that ongoing and future change in the Arctic must be measured against comparatively weak baseline knowledge . The relative contribution of ice algae to primary production and vertical export in the various sectors of the seasonal ice zone is uncertain, as it has not been quantified.For an overview of ice algae and phytoplankton in the Arctic, see Poulin et al. (2011).Some ice algae diatoms form We focus mostly upon the European Arctic Corridor (Fram Strait to Kara Sea) and adjacent basins, the climate "motor" of the Arctic Ocean.More than 80% of the total water exchange between the Arctic Ocean and the adjacent Atlantic and Pacific Oceans takes place within this corridor.The Barents Sea alone, comprising about 30% of the total shelf area in the Arctic Ocean, provides over half of the ocean's total primary production (Sakshaug, 2004; Wassmann et al., 2010).Thus, we focus largely on the Arctic Ocean's most important carbon cycling region, and less on other regions such as the Bering Strait/ Chukchi Sea and the Siberian shelves.Also, little attention is paid here to the interior shelves whose significant terrigenous supply of biogeochemical matter, turbidity, and shallowness has particular primary production and pelagic-benthic coupling constraints.Alternative scenarios of climate forcing and ecosystem function for today's and the future Arctic Ocean exist.One of them, the Oscillating Control Hypothesis (e.g., Hunt and Stabeno, 2002), characterizes the shallow parts of the Bering Sea shelf (50-100 m depth), but may have application for adjacent regions.The region is characterized by low Arctic latitude, ~ 55° N, an unusual range of Arctic temperatures (-1.8 to 14°C), strong wind forcing (when ice-free), and zooplankton species that can cope with extensive environmental variability.Thus, the ice-covered Bering Strait is, by definition, a part of the Arctic Ocean, but has noticeable subarctic and boreal features.These conditions deviate strongly from the core of the Arctic Ocean and the Oscillating Control Hypothesis and are, therefore, BOx 2 | NOT SO eaSY: The VariOuS TermS aNd aSpeCTS OF primarY prOduC TiON primary production is the production of organic compounds from atmospheric or aquatic carbon dioxide, principally through the process of photosynthesis.

Figure 1
Figure 1 depicts the present-day temporal development of ice and plankton algae along a transect through

Figure 1 .
Figure 1.Timing of ice algae and phytoplankton bloom development along a latitudinal axis of the open water-seasonal ice zone region (ranging from 75-85°N) with long to short productive periods in open water (70-75°N) and heavy ice-covered regions (> 73-75°N) in the european arcticCorridor, respectively.present-day scenario (left) and predicted future scenario with a warmer climate (right) along the same latitudes.Notice how today's bloom development in scenario a disappears while scenario F enters the latitudinal gradient in the future.panels e and F exemplify the course of primary production in the scenario of continuously open water in the Barents Sea, characterized by no major freshwater source and weak and slow development of surface water stratification.The variable production in June (panel e) arises through variations in nutrient supply caused by vertical mixing events triggered by low-pressure passage after the end of the spring bloom.panel F projects future primary production at 70°N after global warming leads to increasing thermal stratification and decreased primary production.Modified from Figure1inLeu et al. (2011)

A
Figure3.(a) generic scheme illustrating the basic features and function of the multiyear ice zone (i), seasonal ice zone(ii), and permanently open water outside the seasonal ice zone continuum (iii), with the marginal ice zone toward open water being the most conspicuous feature of zone ii.Zonation, functional ocean types (alpha/beta), and ice cover are indicated at the top.primary production, the depth of the euphotic zone, and the mixed layer are also shown, as are phytoplankton concentrations and ice algae.Below each scenario, the principal profiles of vertical export of particulate organic carbon for winter (red), spring bloom/ episodic mixing (green), and summer (blue) are illustrated in a semiquantitative manner.The stippled lines indicate variability in phytoplankton bloom strength (ii) and the result of episodic mixing (iii).(B) The generic scheme after shrinkage and disappearance of the multiyear ice zone (i) in a future arctic Ocean.The entire arctic Ocean turns into a seasonal ice zone (ii).erosion of stratification in the outer section of the seasonal ice zone results in expansion of the vertically mixed alpha ocean northwards (iii), but there is a simultaneous increase of thermal stratification in the south, which is a new and presently non-existent scenario (iV).Below each scenario the principal profiles of vertical export of particulate organic carbon for winter (red), spring bloom/episodic mixing (green), and summer (blue) are illustrated in a semi-quantitative manner.Partly redrawn and modified fromCarmack and Wassmann (2006) andWassmann (2011) isms, biological degradation by bacteria, or physical breakup of sinking particles determine vertical flux attenuation.The attenuation curvature therefore depends on (a) the vertical distribution of biomass that can potentially settle through the upper layers, (b) the distribution of grazers of sinking material, and (c) the upper-layer mixing regime(Carmack and Wassmann, 2006).As biomass accumulation is restricted to the euphotic zone, and grazers tend to accumulate at layers with high food concentrations, the physical characteristics of the upper water column are important.Ice, wind exposure, heat loss, and stratification strength determine nutrient supply in the euphotic zone(Reigstad et al., 2008).A subsurface chlorophyll a maximum often develops at the nutricline, determined by the light, and the highest export rates are often observed at this depth.Physical environments from multiyear ice, to the seasonal ice zone, to open water form a gradient that is also reflected in the vertical flux attenuation curve and, consequently, affects pelagicbenthic coupling.In different phases of the productive season the vertical flux curve differs along the north-south gradient (lower panel of Figure 3A).During winter, vertical export is low throughout the water column, reflecting low biomass and low export at all depths.This is similar along the entire gradient from open water to multiyear ice, but with lowest vertical export in the multiyear ice zone.The pre-and post-bloom scenarios are often reflected in slightly higher export rates in surface waters (close to the nutricline and the bottom of the euphotic zone) and a noticeable but relatively small reduction with depth (Figure 3A).Vertical export is higher than during winter, but still moderate.The bloom scenarios with high biomass accumulation result in different vertical attenuation patterns determined by the depth and intensity of upper-watercolumn stratification.In the multiyear ice zone (I in Figure 3A), attenuation takes place at shallower depths compared to the marginal ice zone (II), where the euphotic zone is deeper.Export from the upper layers in the multiyear ice may vary considerably depending on nutrient availability, biomass accumulation, and BOx 3 | primarY prOduCTiON CarBON Flux iN The upper COlumN OF The arCTiC OCeaN OF The FuTure: a Bird'S-eYe perSpeCTiVe OF BaSiC FaCTS• As the upper layers of the Arctic Ocean receive more radiation, heat, and freshwater, vertical mixing processes no longer provide enough nutrients for additional phytoplankton growth (e.g.,Tremblay and gagnon, 2009) schemes presented here primarily depict ecosystem function and development within the European sector, they may guide research in the entire central Arctic Ocean.There is a particularly urgent need for information regarding the horizontal and vertical distributions of nutrients, which are currently not well known, except in a few Arctic Ocean shelf Figures 1-3 cannot be studied using permanent moorings or remote sensing.Norklima and IPY programmes (e.g., iAOOS-Norway; http://www.iaoos.no), the Arctic Tipping Points project (http://www.eu-atp.org)funded by FP7 of the European Union (contract #226248), and the project Fate of organic material in the ocean: Controlling mecha-