2.1. The North Water Polynya ecosystem
The North Water Polynya is in northern Baffin Bay, between Ellesmere Island (Canada) and northwest Greenland, connected to Lancaster Sound and Jones Sound in the west, to Kane Basin and the Lincoln Sea in the north, and to central Baffin Bay in the south (
Fig. 1). The average depth of the North Water Polynya is 300 m and ranges from 170 to 600 m (
Bâcle 2000;
Tremblay and Smith 2007). The polynya ecosystem is strongly influenced by season, availability of light, ice breakup, period of open water, and timing of the spring phytoplankton bloom (
Hornby et al. 2021). A detailed description of the current understanding of the North Water Polynya ecosystem is given in
Hornby et al. (2021) and in the International North Water Polynya Study (e.g.,
Klein et al. 2002;
Ringuette et al. 2002;
Tremblay et al. 2002).
The North Water Polynya is one of the most productive Arctic ecosystems (
Klein et al. 2002), partly due to the early onset of open water conditions that extends the light exposure of primary producers, leading to an early spring bloom (
Lewis et al. 1996;
Tremblay et al. 2006). The energy provided by the spring bloom largely accumulates in surface waters, where it is intensely grazed upon by herbivorous zooplankton (e.g.,
Calanus hyperboreus). It is estimated that only 27% of the particulate primary production during a spring bloom leaves the upper 50 m, of which 1%–7% reaches the benthos, depending on the water depth (
Tremblay et al. 2006). Lower trophic level (TL) prey, such as Arctic cod and other meso-zooplankton, play a key role in transferring energy to higher TLs such as seabirds and marine mammals (
Hobson et al. 2002).
Defining the exact extent of the North Water Polynya can be challenging due to its dynamic boundaries. Sea-ice melting in the North Water Polynya starts with the return of daylight in the spring in the southeast—due to the influence of the warmer West Greenland Current—and progresses northwest. By May, extensive open water occurs throughout the region (
Barber et al. 2001;
Barber and Massom 2007), reaching its maximum extent by late June or early July (
Dunbar 1969;
Barber et al. 2001;
Pikialasorsuaq Commission 2017). The North Water Polynya continues to expand until it eventually merges with open water northward from Davis Strait, dissolving the North Water Polynya, and creating a largely open ocean by August (
Barber et al. 2001;
Preußer et al. 2015). To capture the period of polynya formation, full open water extent, and dissolvement of the polynya, the model represents the time frame from April to October. The study area boundary ranges from 76°–78.5°N to 80.5°–65°W, covering the area of the North Water Polynya, with a size of ∼85 000 km
2, representing the peak observed extension of the polynya in 2000 (
Moshøj 2015).
2.4. Functional groups
We used 20 functional groups to characterize the food web, which were parameterized using qualitative and quantitative studies from the North Water Polynya or other Arctic regions if local data were not available (
Tables 1 and
A3). Selection of functional groups included species of interest in terms of role in Inuit subsistence and the polynya ecosystem. Where available, biomass (
B) was estimated from survey reports and peer-reviewed research for the North Water Polynya. Production to biomass ratios (
P/
B) values were either calculated using total morality rates (natural mortality rate (
M) + fisheries mortality rate (
F)) or derived via the life-history tool in FishBase (
www.fishbase.org), which combines mortality, parameters of the Bertalanffy growth function, and mean temperature to calculate
M (
Palomares and Pauly 1998).
M estimates were taken from published literature for the North Water Polynya or other Arctic ecosystems.
F, if not explicitly available in the literature, was calculated as catch over biomass, based on reported subsistence catches from Canada and Greenland. Consumption to biomass ratios (
Q/
B) were largely obtained for North Water Polynya taxa when available. In the absence of information from the North Water Polynya, we used estimates of values from the literature or other Ecopath models that were most appropriate to be used for North Water Polynya ecosystem or from other Arctic regions. A diet matrix (
Table 2) was constructed based on published diet studies for North Water Polynya taxa, when available. In the absence of diet studies from the North Water Polynya, we used estimates of values from the literature that were the most appropriate to be used for North Water Polynya or similar Arctic species. Details on data sources for each parameter and functional group are in
Table A3.
Marine mammals—marine mammals were represented in five functional groups: polar bear, beluga, narwhal, walrus, and ringed seal. Data for the relevant populations or subpopulations were obtained from empirical studies and other Arctic Ecopath models (
Tables 1 and
A4). Based on the assumption that all marine mammals inhabited the North Water Polynya throughout the model period. To calculate biomass (
B), the number of individuals was multiplied by the average weight per individual and divided by the total model area (tonnes/km
2). For polar bears,
B estimates were based on an average population size of 90 bears (2005–2007) from the Kane Basin subpopulation, which was similar to the estimate of 60 bears for the North Water Polynya based on aerial spring surveys in 2009 and 2010 (
Heide-Jørgensen et al. 2013).
P/
B was calculated using
M estimates for the Baffin Bay subpopulation and
F for the Kane Basin subpopulation, derived from aerial or capture–recapture data provided by Canadian and Greenland government agencies (
York et al. 2016).
Q/
B estimates for polar bears were derived from an Ecopath model for the Western Baffin Bay (
Pedro et al. 2023).
B estimates for beluga were based on aerial observations in the spring of 2008–2010, as direct observations for the North Water Polynya were available only for that timeframe. Similarly,
B estimates for narwhal relied on aerial observations in the spring of 2009, 2010, and 2014.
P/
B for belugas considered
M from the Beaufort Sea Shelf Ecopath model (
Hoover et al. 2021) and
F based on North Atlantic Marine Mammal Commission (NAMMCO) catch statistics (
NAMMCO 2016). Due to a lack of data for narwhal in the North Water Polynya, we let Ecopath estimate
P/
B, based on our calculated
B, and
Q/
B and
EE estimates from the Western Baffin Bay Ecopath model (
Pedro et al. 2023).
F for narwhal was based on the annual reported catches for the municipality of Qaanaaq referenced in
Heide-Jørgensen et al. (2013).
B estimates for walrus were based on aerial survey counts integrated in a Bayesian assessment model (
Witting and Born 2005).
P/
B was calculated accounting for
M and
F.
M of 0.02 year
−1 was taken from
Witting and Born (2014), who used Bayesian statistical models to calculate
M for western Greenland walrus populations, including Baffin Bay and the North Water region. Overall, natural mortality in walruses is considered low, due to low productivity and high longevity (∼40 years) (
NAMMCO 2016).
F was based on NAMMCO catch statistics, reflecting regional catches by Inughuit walrus hunting for food and walrus ivory (
Born 2017;
NAMMCO 2018).
Q/
B estimates for walrus were estimated by an Ecopath model for the Western Baffin Bay (
Pedro et al. 2023).
B estimates for ringed seal were based on a population model for Baffin Bay and associated waters (Kane Basin, Jones Sound, eastern Lancaster sound, Ungava Bay, and eastern Hudson Strait). Because no direct estimates for the North Water Polynya exist, the population size for the fraction of the study area was calculated, assuming an even distribution in the region.
P/
B was calculated accounting for
M and
F.
M estimates were derived from an Ecopath model for the Beaufort Sea Shelf (
Hoover et al. 2021).
F was based on NAMMCO catch statistics (
NAMMCO 2016).
Q/
B estimates for ringed seal was derived from an Ecopath model for the Beaufort Sea Shelf (
Hoover et al. 2021).
Seabirds—little auk (
Alle alle) represents seabirds in the model. Little auk is a key species in the North Water Polynya ecosystem, exhibiting the largest colony on Earth on the eastern side of the North Water Polynya (
Karnovsky and Hunt 2002;
Wojczulanis-Jakubas et al. 2022). In future model iterations, additional bird species could be added, such as Ivory Gull (
Pagophila eburnea), Thick-billed Murre (
Uria lomvia), and Black-legged Kittiwake (
Rissa tridactyla).
B for little auk was calculated based on population estimates of 66 million birds from aerial video recordings and aerial photographs in the Thule area (eastern North Water Polynya.
P/
B and
Q/
B were based on the Ecopath model for the Newfoundland and Labrador Shelf ecosystem (
Tam and Bundy 2019). Annual mean harvest between 1998 and 2013 ranged between 7772 and 75 712 birds (
Mosbech et al. 2018). To calculate
F, we calculated annual mean harvest of 33 971 birds.
Fish—there is limited data on relative abundance of fish, spatial distributions, of most marine life histories, spatial connectivity, and seasonal migrations in the North Water Polynya (
Hornby et al. 2021). Fish were divided into two functional groups—Age 1 + Arctic cod (Arctic cod inhabiting waters <100 m representing their benthic–pelagic life stage;
Geoffroy et al. 2016) and other fish. We assumed that the other fish group contained fish that are present within the North Water Polynya with different feeding ecology than that of Arctic cod and may represent a food source to higher TLs in the North Water Polynya, including Greenland halibut (
Reinhardtius hippoglossoides), Arctic eelpout (
Lycodes reticulatus), snailfish (
Liparis spp.), and Thorny skate (
Amblyraja radiata) (
Hornby et al. 2021).
B for Arctic cod (Age 1+) was based on biomass estimates from hydroacoustic surveys in the North Water Polynya for 2005–2007 (
Herbig et al. 2023).
P/
B was calculated accounting form
M and
F.
M was derived from the FishBase life-history tool.
F was set to 0.0001 year
−1 as directed, commercial, and subsistence fisheries for this species are negligible (
Steiner et al. 2019;
Geoffroy et al. 2023).
Q/
B was derived from the FishBase life-history tool. For the other fish group, we let Ecopath estimate
B by setting
EE to 0.85. Because the biomass for this group was unknown during model creation, an assumed
EE value of 0.85 indicated that 85% of fish in this group is consumed within the system (
Christensen et al. 2005). This assumption is reasonable for a group set to include pelagic, demersal, and benthic species with various predation vulnerabilities (
Table A7).
P/
B and
Q/
B was derived from published literature values, that make ecological sense for assumed species in this group (
Table A3).
Benthic invertebrates—for the benthic invertebrate community, species were aggregated into four functional groups: arthropods, bivalves, echinoderms, and worms. These represent species that are abundant in the North Water Polynya, such as clams, scallops, crabs, polychaetas, and sea stars (
Roy et al. 2015a,
2015b;
Mäkelä et al. 2017).
B estimates for each invertebrate group were provided by the Archambault lab (Pers. comm.) and literature based on trawl surveys in the North Water Polynya in 2013–2015. Therefore,
B for all benthic invertebrate groups assumes that the respective biomass does not vary substantially over time.
P/
B and
Q/
B for each group were based on other Arctic marine ecosystem models (
Table A3), as life-history parameters for the benthic invertebrate groups within the study area were not available.
Zooplankton and ichthyoplankton—copepods are the most important group of zooplankton by biomass in the North Water Polynya (
Ringuette et al. 2002). Herbivorous, omnivorous, and carnivorous copepods are all present in the North Water (
Hornby et al. 2021). Gelatinous zooplankton are also present in the North Water Polynya and have higher grazing rates than copepods later in summer but were not included due to data limitations (
Deibel et al. 2017). Hence, our model represents one hypothesis of how the ecosystem may be structured and provides a basis for future hypothesis testing, including those around gelatinous zooplankton. Zooplankton and ichthyoplankton were represented by four functional groups: large copepods (
C. hyperboreus, Calanus glacialis, and
Metridia longa), medium copepods (
Pseudocalanus spp. and
Calanus finmarchicus), other meso-zooplankton, and Arctic cod (Age 0; (Arctic cod inhabiting waters >100 m;
Geoffroy et al. 2016), as a key representative for ichthyoplankton in the region.
B for large and medium copepods, and other meso-zooplankton were provided by Gérald Darnis (
Darnis et al. 2022; Pers. comm.).
B for Arctic cod (Age 0) was based on biomass estimates from hydroacoustic and trawl surveys in the North Water Polynya for 2005–2007 (
Herbig et al. 2023).
P/
B ratios for large and medium copepods were taken from the Canadian Beaufort Sea Shelf Ecopath model (
Hoover et al. 2021).
P/
B Arctic cod (Age 0) was calculated based on daily
M estimates from the Chukchi Sea.
Q/
B for large and medium copepods was initially set to 45 year
−1, to allow for a
P/
Q value of 0.4;
EE was set to 0.95 (adapted from
Hoover et al. 2021).
Q/
B for Arctic cod (Age 0) was set to be 105 year
−1, based on values for zooplankton <2 mm (
Tam and Bundy 2019), assuming a similar diet and a body size of ≤1.6 mm of the sampled Arctic cod (Age 0). We assumed no cannibalism of Arctic cod (Age 0) by Arctic cod (Age 1+) occurred, since current literature from other Arctic regions suggest it to be limited (
Walkusz et al. 2013;
Majewski et al. 2016). Biomass for the other meso-zooplankton group was unknown during model creation; hence, an assumed
EE value of 0.85 was set to indicate that nearly all the species included are consumed within the system. The
P/
B and
Q/
B were set to 15 and 60 year
−1, higher than the copepod group.
Primary Producers—primary producers are represented by two functional groups, large pelagic producers (>=5 µm) and small pelagic producers (0.7–5 µm). Dry weight biomass values for each size class were converted to wet weight using a conversion factor of 1 gC = 9 g wet weight after
Pauly and Christensen (1995). The modelled period (April—October) represents the growth period of phytoplankton in the North Water Polynya (
Klein et al. 2002). We did not include ice algae as a separate primary production functional group because grazing impact of ice algae is negligible in early spring and the low standing stock of sea ice meiofauna feeding on ice algae is not considered an important food source for higher TLs in the North Water Polynya (
Nozais et al. 2001).
B and
P/
B for both groups were based on biomass and productivity estimates from water samples in the North Water Polynya in Summer–Fall (2005–2007).
Detritus—detritus is represented by two functional groups, pelagic and benthic detritus. The pelagic detritus group represents detritus that is retained within the water column, primarily derived from the spring bloom (large and small pelagic producers) and is a food source for zooplankton (
Hoover et al. 2021). The benthic detritus group represents detritus that sinks quickly from the water column, originating primarily from plankton and feeding by benthic invertebrates (
Hoover et al. 2021).
B for the two detritus groups were based on an Ecopath model for the Beaufort Sea (
Hoover et al. 2021), as no estimates exist for the North Water Polynya ecosystem.