Climate-driven changes in the Subarctic will directly impact capelin populations and the ecosystem they inhabit, including their predators, prey, and physical habitats. Consequently, incorporating ecosystem considerations in future capelin fisheries management is crucial. In this study, a multidisciplinary group of experts critically evaluated whether the current capelin stock assessment and management frameworks for the four main capelin stocks in the Barents Sea (BS), Iceland-East Greenland-Jan Mayen (IEGJM), Newfoundland and Labrador shelf (NL) and Alaska (AK) align with the principles of an Ecosystem Approach to Fisheries Management (EAFM). An evidence-based ranking of our knowledge on current capelin dynamics across ecological, economic, and social dimensions was conducted, using expert knowledge supported by literature. This exercise also identified data currently used for assessment and management, which highlighted that the existing capelin assessment frameworks include varying degrees of EAFM elements across stocks, such as considerations of trophic interactions, bottom-up processes, accounting for ecosystem uncertainty, and stakeholder engagement in the advisory process. Nonetheless, there is room for improvement where data and knowledge are lacking. We provide some key tactical (short-term) and strategic (long-term) recommendations from our perspective on what is required to ensure the sustainable management of capelin in the circumpolar region over the coming decades.

There have been several large capelin stock declines (collapses) in the Barents Sea since the 1980s. A recent study by Olsen et al., using a state-space model and parameters from a previously published paper in 2018 covering the years 1980–2015, concluded that the authors can confidently say that they did not find support for the hypothesis that high levels of herring predation are a necessary condition for the capelin collapses. However, in the 2018 article, where some of the same authors from the 2024 publication contributed, they found a negative association of juvenile herring with capelin age-0 abundance. Previous studies from the Barents Sea show that all the known capelin collapses are linked to high abundance of juvenile herring, contradicting the statements by Olsen et al. The capelin collapses are all associated with high abundance of juvenile herring in the Barents Sea. However, the converse is not true; high herring biomass does not always lead to the collapse of the capelin stock, probably due to interannual variation in spatial overlap between the stocks. Extended data series of capelin and herring up to 2023 and before 1980 from the Barents Sea further provide evidence supporting previous findings. We contend that the capelin collapses are most likely caused by herring predation and that the study by Olsen et al. does not shed new light on causes for capelin stock collapses.

Statistical distributions of spatial abundance data of animal populations can be informative about underlying processes and mechanisms that govern the spatial distributions. Here, we examine a large data set from annual 0-group fish surveys in the Barents Sea (1980-2017) collected with standardised trawl sampling from a regular grid with 30-35 nautical miles between stations. The primary data are expressed as swept-area density of 0-group individuals for six fish species. The annual data series are highly right-skewed with a tail of high values on a linear scale but are close to lognormal with log-transformed data. The 0-group density spans typically 4-5 orders of magnitude for the annual series. When stations are ranked in order from highest to lowest density, they show a common pattern for the six species across the annual series. With log-transformed data, this pattern is a near log-linear relationship but with an upswing in the high-density end revealing higher variability for the top stations. The cumulative version with linear-scale data is approximated by a simple asymptotic model. Between 14 and 42% of the total number of sampled individuals of a year-class were on average contained in the highest ranked station for the six species, while 50% of the sampled individuals were found in the first 2-7 highest ranked stations. The common and consistent pattern of the ranked distributions is interpreted to reflect the opposing forces of physical dispersion, on the one hand and aggregation by swimming and schooling, on the other.

Acoustic surveys of mixed fish aggregations often rely on manual allocation of acoustic energy to target species, with auxiliary information such as trawl catches and target strength distributions aiding interpretation. However, different teams of experts may use the same auxiliary data differently. In this experiment, seven teams of experts interpreted the same acoustic data, and we compared their classification to target categories to quantify uncertainty in the manual classification process. The teams largely agreed on the total acoustic energy attributable to organisms, but there was significant variation in how the teams split this energy on different target categories. This was caused by differences in applied thresholds for separating plankton and fish, as well as disagreements in species classification. For all target categories, the variation due to teams was lower than the overall variability across acoustic segments, but when scaled up to the number of segments in a typical survey, the team effect either dominated or was of similar magnitude as the segment variability. These results imply a need for further standardisation and uncertainty estimation of expert evaluations in acoustic surveys involving manual interpretation of echo sounder data.

We present acoustic observations obtained by bottom mounted echosounders at two locations in the northern Barents Sea and a third on the northern continental slope of the Barents Sea. Data collected over a period of approximately two years reveal significant variability in the density and depth distribution of pelagic macrozooplankton and fish. Variability over diurnal to seasonal timescales is related to light conditions, prevailing currents, water column characteristics (temperature, stability, and pycnocline depth) and to sea ice conditions. Of particular importance is the relative volume of Polar Water and Atlantic influenced water present at the mooring sites. On the northern continental shelf, above average concentrations of macrozooplankton and fish were observed during late summer and early autumn following ice melt, and during winter. Minimum densities of macrozooplankton and fish were observed during spring and early summer. These data demonstrate the persistent presence of a macrozooplankton community throughout the winter months in the northern Barents Sea. On the Great Bank full depth diel vertical migrations were observed throughout the Polar Night and under sea ice. At the two more northerly stations the density of the sound scattering layers performing diel vertical migrations and the vertical range of the migrations was greatest during the transition period between the polar night and mid-night sun periods. Superimposed on the seasonal variability in density distribution was significant shorter-term variability driven by hydrographic processes occurring over synoptic time scales. Significant mixing events, or changes in the prevailing current direction that lead to a change in water mass characteristics, forced changes in the concentration and depth distribution of macrozooplankton over periods of hours that were on occasion of similar magnitude to the seasonal variability. On the northern continental slope, a mesopelagic sound scattering layer more than 300 m thick at its seasonal maximum, was located for most of the year within the warm core of the Atlantic Water Boundary Current, ascending into the epipelagic zone for 5-6 weeks during summer. The density and thickness of this sound scattering layer were greatest during the late summer/early autumn and lowest during early spring. Fish-like scatterers were present, concentrated within the deeper regions of the sound scattering layer, during the winter months and during July.

The rapid ongoing changes in the Central Arctic Ocean call for baseline information on the pelagic fauna. However, sampling for motile organisms which easily escape vertically towed nets is challenging. Here, we report the species composition and catch weight of pelagic fishes and larger zooplankton from 12 trawl hauls conducted in ice covered waters in the Central Arctic Ocean beyond the continental slopes in late summer. Combined trawl catches with acoustics data revealed low amounts of fish and zooplankton from the advective influenced slope region in the Nansen Basin in the south to the ice-covered deep Amundsen Basin in the north. Both arctic and subarctic-boreal species, including the ones considered as Atlantic expatriate species were found all the way to 87.5o N. We found three fish species (Boreogadus saida, Benthosema glaciale and Reinhardtius hippoglossoides), but the catch was limited to only seven individuals. Euphausiids, amphipods and gelatinous zooplankton dominated the catch weight in the Nansen Basin in the mesopelagic communities. Euphausiids were almost absent in the Amundsen Basin with copepods, amphipods, chaetognaths and gelatinous zooplankton dominating. We postulate asymmetric conditions in the pelagic ecosystems of the western and eastern Eurasian Basin caused by ice and ocean circulation regimes.

During several surveys covering the north-western and northern Svalbard waters, and the deeper Yermak Plateau north of Svalbard during the period 2012-2020, 291 standardized hauls with a demersal trawl were made. All fishes in the catches were identified to the lowest possible taxonomic level, mostly to species. In addition to the data generated from the trawl catches, bottom temperature, salinity, and depth were recorded at each trawl station. The eelpouts were the most species rich family, with 15 species, followed by codfishes and sculpins with six species each. The other 13 families were represented with one to four species each. Atlantic cod (Gadus morhua) dominated by weight in the catches, while polar cod (Boreogadus saida) dominated by numbers. In the deeper areas including the Yermak Plateau, Greenland halibut (Reinhardtius hippoglossoides) dominated by weight. Zoogeographically, 23 species were Arctic, 5 were Mainly Arctic, 4 Arctoboreal, 25 Mainly Boreal, 9 Boreal, and 2 Widely Distributed. The Arctic species dominated in the deeper areas (Yermak Plateau and slope >500 m) and on the shallow eastern shelf, whereas the Mainly Boreal species dominated along the slope and on the western and northern shelves < 500 m. A hierarchical cluster analysis revealed three prominent station clusters consistent with the zoogeographical classifications which reflected the oceanographic conditions, water masses and sea ice. The hierarchical cluster analysis additionally separated the cold-water species into two groups consistent with the bathymetry of the region. Diet studies of Atlantic cod revealed that it mainly fed on hyperiids, and that its preferred food further south, capelin, was not prominent in the diet. Regional index terms: Northeast Atlantic, Fram Strait, Svalbard, Yermak Plateau

Arctic cod (Boreogadus saida) is the most abundant forage fish in the Arctic Ocean. Here we review Arctic cod habitats, distribution, ecology, and physiology to assess how climate change and other anthropogenic stressors are affecting this key species. This review identifies vulnerabilities for different life stages across the entire distribution range of Arctic cod. We explore the impact of environmental (abiotic and biotic) and anthropogenic stressors on Arctic cod with a regional perspective in a scenario up to the year 2050 and identify knowledge gaps constraining predictions. Epipelagic eggs and larvae are more vulnerable to climate change and stressors than adults. Increased water temperatures, sea-ice decline, altered freshwater input, acidification, changing prey field, increased interspecific competition, new predators, and pollution are the principal stressors that will affect Arctic cod populations. Detrimental effects are likely to be greater in regions characterized by the advection of warmer Atlantic and Pacific waters. In contrast, Arctic cod may benefit from ocean warming in colder areas of the High Arctic. The risk from fisheries is moderate and primarily limited to bycatch. Overall, a decrease in suitable habitat and an associated decline in total Arctic cod biomass are predicted. In most Arctic seas, the relative abundance of Arctic cod within the fish community will likely fluctuate in accordance with cold and warm periods. A reduced abundance of Arctic cod will negatively affect the abundance, distribution, and physiological condition of certain predators, whereas some predators will successfully adapt to a more boreal diet. Regional management measures that recognize the critical role of Arctic cod are required to ensure that increased anthropogenic activities do not exacerbate the impacts of climate change on Arctic marine ecosystems. Ultimately, the mitigation of habitat loss for Arctic cod will only be achieved through a global reduction in carbon emissions.

Young-of-the-year (0-group) fish in the Barents Sea have been investigated in an annual joint Norwegian-Russian pelagic trawl survey in autumn, using a standardized procedure since 1980. We use a conceptual framework of `upstream' spawning areas and 'downstream' nursery areas, recorded as 0-group distribution in the Barents Sea, to address spatial (geographical) and temporal (1980-2017) variation in 0-group length. Four boreal species (cod Gadus morhua, haddock Melanogrammus aeglefinus, herring Clupea harengus, and deepwater redfish Sebastes mentella) tended to have smaller 0-group individuals in the northern and eastern parts of the Barents Sea, with the largest individuals found in the central part where they were also most abundant. We interpret this to reflect slower growth as the lore-runners' of the seasonal cohort of juveniles are transported into colder waters (through lateral mixing). The Arctic species (capelin Mallotus villosus and polar cod Boreogadus saida) showed a different pattern with increasing 0-group length with increasing distance away from the spawning areas, seen most clearly for capelin. The longer juveniles in northern areas are probably older and stemming from early spawning. There was temporal covariation in 0-group length between the six species over the 38-year time series, with highest correlation between cod and haddock. The covariation reflected similar fluctuations in four decadal 'waves', with maxima in 0-group length in the early/mid 1980s, 1990s, 2000s, and 2010s. There was a high degree of spatial consistency in the temporal patterns of variation in 0-group length, with synchronous variations in different geographical areas. There were also increasing linear trends over the time series for cod, haddock, and polar cod, which represented increase of about 20%, 40%, and 15% of the initial length for the three species, respectively. The fluctuations and trends in 0-group length were positively correlated with seawater temperature, which suggests a strong effect of climate variability and warming (by 1.5-2.0 degrees C since 1980) on 0-group length. The clear differences among the species, and the limited fraction of variance explained by temperature, suggest that other factors such as food play additional roles. Zooplankton biomass integrated over the water column had low explanatory power, but this may reflect intrinsic limitations in the data (e.g., depth-integrated, end of season) in providing an adequate representation of feeding conditions, rather than suggesting that food was not important.

Zooplankton biomass has been monitored on joint Norwegian-Russian surveys in late summer and autumn since the 1980s. We report here on zooplankton biomass in three size fractions (<1, 1-2, and > 2 mm in screen mesh opening) obtained with WP-2 plankton net (180 mu m mesh size) hauled vertically over the water column from near bottom to the surface for the period 1989-2020. The number of samples (stations) collected each year has been about 100-200, with a total number of 4543 stations for the whole data set. The size composition of zooplankton reflected by the three fractions has shown remarkable stability, with about 50% of biomass con-tained in the medium fraction (made up largely of Calanus species), about 1/3 in the small fraction (36%), and 16% in the large fraction. The depth integrated biomass was generally larger in basins compared to shallower bank areas. The temporal (interannual) pattern of change was characterized by a marked peak in biomass in 1994 and 1995 with values up to >20 g dry weight (dw) m- 2, driven to large extent by the small size fraction. Subsequently the biomass decreased to lower values but with a divergence of relatively high values (10-15 g dw m-2) in the inflow area of Atlantic water in southwest, and low values (2-6 g dw m-2) in the central area. The difference is interpreted to reflect an increase in a second summer generation of Calanus finmarchicus in the Atlantic water and a decrease of C. glacialis in the central area. The zooplankton biomass fluctuated inversely with the biomass of the Barents Sea capelin (Mallotus villosus) stock, reflecting a top-down predation effect. However, biomass was also negatively correlated with temperature of the Atlantic water, suggesting an addi-tional and confounding effect of climate variability and change. The decrease in biomass of the central area used as a forage area by capelin, was associated with a shift to dominance by the small size fraction. This is likely an effect of predation and could be associated with a lower trophic conversion efficiency from phytoplankton to planktivorous fish and higher trophic levels by smaller zooplankton (smaller copepods such as Pseudocalanus and others) compared to the larger Calanus species.

In September 2016, a marine ecosystem survey covered all trophic levels from phytoplankton to seals in the Arctic Ocean to the west and north of Svalbard. At the ice edge, 26 harp seals were sampled to assess whether recent environmental changes had affected their diets and body condition by comparing our current results with previous investigations conducted 2-3 decades ago in the northern Barents Sea, when the ice edge was located much further south. Our results suggest that the body condition was slightly but significantly lower for one year and older seals in 2016 compared with seals sampled in the early 1990s. Furthermore, we confirmed previous findings that polar cod (Boreogadus saida) and the pelagic hyperiid amphipod Themisto libellula still dominate the seal diet. One consequence of current ice edge localisation north of Svalbard is that the water depth underneath is now 500 m and deeper, which probably explains the absence of bottom associated species, and the presence of species such as Atlantic cod (Gadus morhua) and blue whiting (Micromesistius poutassou) as alternative species in addition to polar cod and T. libellula in the seal diets. Stable isotope data also suggest possible long-term importance in the seal diet of T. libellula and of low trophic level benthopelagic prey such as the squid Gonatus fabricii over mid-trophic level pelagic fishes, but with a strong component of small, benthopelagic fish such as blue whiting. The long-term importance of pelagic crustaceans was also suggested from the fatty acid analyses. Assessment of the abundance of prey showed that T. libellula was by far the most abundant prey species in the upper water layers, followed by krill (mainly Thysanoessa inermis), Atlantic cod and polar cod. Prey-preference analyses indicated that polar cod was the most preferred prey species for the seals.

Sustainable human exploitation of living marine resources stems from a delicate balance between yield stability and population persistence to achieve socioeconomic and conservation goals. But our imperfect knowledge of how oceanic oscillations regulate temporal variation in an exploited species can obscure the risk of missing management targets. We illustrate how applying a management policy to suppress fluctuations in fishery yield in variable environments (prey density and regional climate) can present unintended outcomes in harvested predators and the sustainability of harvesting. Using Atlantic cod (Gadus morhua, an apex predatory fish) in the Barents Sea as a case study we simulate age-structured population and harvest dynamics through time-varying, density-dependent and density-independent processes with a stochastic, process-based model informed by 27-year monitoring data. In this model, capelin (Mallotus villosus, a pelagic forage fish), a primary prey of cod, fluctuations modulate the strength of density-dependent regulation primarily through cannibalistic pressure on juvenile cod survival; sea temperature fluctuations modulate thermal regulation of cod feeding, growth, maturation, and reproduction. We first explore how capelin and temperature fluctuations filtered through cod intrinsic dynamics modify catch stability and then evaluate how management to suppress short-term variability in catch targets alters overharvest risk. Analyses revealed that suppressing year-to-year catch variability impedes management responses to adjust fishing pressure, which becomes progressively out of sync with variations in cod abundance. This asynchrony becomes amplified in fluctuating environments, magnifying the amplitudes of both fishing pressure and cod abundance and then intensifying the density-dependent regulation of juvenile survival through cannibalism. Although these transient dynamics theoretically give higher average catches, emergent, quasicyclic behaviors of the population would increase long-term yield variability and elevate overharvest risk. Management strategies that overlook the interplay of extrinsic (fishing and environment) and intrinsic (life history and demography) fluctuations thus can inadvertently destabilize fish stocks, thereby jeopardizing the sustainability of harvesting. These policy implications underscore the value of ecosystem approaches to designing management measures to sustainably harvest ecologically connected resources while achieving socioeconomic security.

The 3.3 million km(2) marine ecosystem around the North Pole, defined as the Central Arctic Ocean (CAO), is a blind spot on the map of the world's fish stocks. The CAO essentially comprises the permanently ice-covered deep basins and ridges outside the continental shelves, and is only accessible by ice-breakers. Traditional trawling for assessing fish stocks is impossible under the thick pack ice, and coherent hydroacoustic surveys are unachievable due to ice-breaking noise. Consequently, nothing is known about the existence of any pelagic fish stocks in the CAO, although juveniles of Boreogadus saida richly occur at the surface associated with the sea ice and iceassociated Arctogadus glacialis has been reported as well. We here present a first indication of a possible mesopelagic fish stock in the CAO. We had the opportunity to analyse a geophysical hydroacoustic data set with 13 time windows of usable acoustic data over a transect from 84.4 degrees N in the Nansen Basin, across the North Pole (90.0 degrees N), to 82.4 degrees N in the Canada Basin. We discovered a deep scattering layer (DSL), suggesting the presence of zooplankton and fish, at 300-600 m of depth in the Atlantic water layer of the CAO. Maximum possible fish abundance and biomass was very low; values of ca. 2,000 individuals km(-2) and ca. 50 kg km(-2) were calculated for the DSL in the North-Pole area according to a model assuming that all acoustic backscatter represents 15-cm long B. saida and/or A. glacialis. The true abundance and biomass of fish is even lower than this, but cannot be quantified from this dataset due to possible backscatter originating from pneumatophores of physonect siphonophores that are known to occur in the area. Further studies on the DSL of the CAO should include sampling and identification of the backscattering organisms. From our study we can conclude that if the central Arctic DSL contains fish, their biomass is currently too low for any sustainable fishery.

Identifying spawning sites of fish often involves extensive egg and larval sampling surveys over potential spawning sites, or by backward-tracking advected larvae to their source. Due to the vastness of the Barents Sea capelin spawning areas, back-tracking methods have limited application. Egg and larval surveys that provide information about spawning sites have also been discontinued in recent years. This paper aims at using alternative data sources to egg and larval distribution information, to infer potential spawning regions of the Barents Sea capelin during the period 1994–2020. We use the K-Means clustering technique to cluster historical spawning sites into spawning regions, and the Self Organizing Map (SOM) algorithm to define observed data clusters, which we assign to specific regions. The observation data consists of survey data sets from capelin pre-spawning and post-spawning periods during winter and spring respectively, as well as data from the Norwegian Directorate of Fisheries Electronic Reporting System database (ERS). Our method was efficient in reproducing capelin spawning regions and approximate time windows for commencement of spawning. The results showed that spawning occurred mainly over the eastern part of historical spawning areas during the whole period. A westward extension of the preferred spawning areas occurred in several years regardless of the rising Barents Sea water temperatures, especially during the second half of the period. The ERS data can be used to identify the arrival times and migration fronts of pre-spawning capelin along the coast.

This paper uses a two-parameter logistic function to model the dynamics of length-at-maturation for the Barents Sea capelin over the past 47 years. We estimate the function parameters using a combination of length-age data from scientific surveys, and commercial catch statistics. Using temporal variability in the function parameters, we demonstrate that the time series of stock biomass defines a three-state Markov process, that qualitatively represent high, moderate, and collapse states of the stock biomass. We make inference about transition times between the states by calculating the mean passage times for the Markov process. Our analyses also show that maturation intensity is higher at low stock size (leading to shorter lengths at maturation), compared to when biomass levels are either high or moderately high. Our results are central to management of this stock, as uncertainty in estimating the proportion of maturing biomass affects harvest decisions and ultimately, the sustainability of the stock.

Recent warming in the Barents Sea has led to changes in the spatial distribution of both zooplankton and fish, with boreal communities expanding northwards. A similar northward expansion has been observed in several rorqual species that migrate into northern waters to take advantage of high summer productivity, hence feeding opportunities. Based on ecosystem surveys conducted during August–September in 2014–2017, we investigated the spatial associations among the three rorqual species of blue, fin, and common minke whales, the predatory fish Atlantic cod, and their main prey groups (zooplankton, 0-group fish, Atlantic cod, and capelin) in Arctic Ocean waters to the west and north of Svalbard. During the surveys, whale sightings were recorded by dedicated whale observers on the bridge of the vessel, whereas the distribution and abundance of cod and prey species were assessed using trawling and acoustic methods. Based on existing knowledge on the dive habits of these rorquals, we divided our analyses into two depth regions: the upper 200 m of the water column and waters below 200 m. Since humpback whales were absent in the area in 2016 and 2017, they were not included in the subsequent analyses of spatial association. No association or spatial overlap between fin and blue whales and any of the prey species investigated was found, while associations and overlaps were found between minke whales and zooplankton/0-group fish in the upper 200 m and between minke whales and Atlantic cod at depths below 200 m. A prey detection range of more than 10 km was suggested for minke whales in the upper water layers.

Climate change has large effects on population dynamics of fish species in high latitude ecosystems. Arctic fish stocks experience multiple pressures with changing abiotic living conditions and increased competition and predation from boreal species. However, there are many unknowns regarding how environmental change influences the dynamics of those populations. Here, we focused on the Barents Sea polar cod, a pan-Arctic zooplanktivorous key fish species physiologically and ecologically adapted to the presence of sea ice. We developed an age-resolved Bayesian state-space model of the dynamics of polar cod based on 30 yr of survey data (1986-2015). Using this model, we quantified how inter-annual changes in abundance were associated with abiotic variables (temperature and sea ice cover) and biotic variables (prey biomasses and a predation index). Using the model output, we used a hindcast scenario approach to investigate to which degree the observed variations in total population size were related to the abiotic or biotic variables. Our results showed that variation in abundance of young polar cod (ages 0 and 1) was best explained by abiotic variables while variation in the older age groups (ages 3 and 4) was best explained by predation. Hindcast scenarios showed that the abiotic variables had a more evident effect than predation on population dynamics, but none of the variables we considered could explain the drastic population decline observed in recent years. Our work shows the advantages of studying age-specific responses as a stepping stone to understand changes at the population level.