The vertical transport and storage of atmospheric carbon through the biological carbon pump (BCP) follows multiple pathways. The lipid pump is a recently recognized BCP process driven by the seasonal vertical migration of copepods utilizing carbon-rich lipid reserves during extended overwintering periods at depth where they release carbon through respiration and mortality. Previous estimates for the Arctic suggest that the lipid pump carbon (LPC) flux exports similar amounts of carbon as the sinking particulate organic carbon (POC) in the gravity pump. However, concurrent sampling is necessary for accurate comparisons between the BCP pathways. Here, we used in situ imaging to simultaneously observe the distributions of Calanus copepods and marine snow within seven Arctic subregions. The calculated LPC flux ranged from 0.77-4.15 g C m -2 y -1 which was comparable to the range of POC flux estimates, confirming the importance of the lipid pump in Arctic vertical carbon flux. We found that average LPC flux is double that of POC flux in deep basin subregions, implying that deep habitats where diapausing copepods reside are critical for Arctic carbon storage. This study provides a foundation for improved understanding of the BCP pathways in a changing Arctic and offers methodological advancements for expanding the monitoring of the lipid pump.

During foraging activity, baleen whales are generally dependent on the occurrence of dense prey patches. These are largely influenced by the interaction between ocean currents and the vertical distribution of individual prey. Identifying the spatio-temporal dynamics of the areas where prey form dense aggregations, at submeso scale (similar to 10 km) and over a wide area, is particularly informative for predicting the use of the busy coastal environment by some endangered marine mammal species. In this study, we simulate the distribution of advection-induced copepod aggregations that are the main prey for the North Atlantic right whale (Eubalaena glacialis), which is particularly threatened by conflicts with human activities. The distribution of aggregations results from simulated copepods' trajectories, inferred from a high-resolution 3D hydrodynamic model coupled with a model simulating variable copepods' depth distribution. The results show that occurrences of densely aggregated simulated prey are positively correlated to the distribution of right whales and efficiently describe their main feeding areas. Moreover, daily aggregation simulations better predict right whales' distributions than longer-term averaged simulations. This tool could be used effectively in combination with other models to predict potential foraging hotspots and aid the conservation of the North Atlantic right whale and other baleen whale species as well.

The Canadian Arctic and sub-Arctic are under­going rapid ecological, social, and political transformations that profoundly affect local food systems. This study examines how households in Nunavik, Canada, navigate their complex, mixed subsistence and market-based food systems and identifies the factors shaping decision-making when selecting food sources. Specifically, this study asks: (1) What are the primary food acquisition sources for households, how are they accessed, and what challenges are faced? and (2) What key factors influence household decision-making in selecting food acquisition sources? Using a co-production of knowledge approach, participatory workshops were conducted with 17 community members from five Nunavik communities. Data included co-created conceptual food system maps, participant-placed preference cards, and facilitator notes from scenario-based discussions. Findings were subsequently validated in a follow-up community engagement session. Thematic qualitative analysis revealed that Nunavimmiut households draw on over 13 distinct food acquisition sources, ranging from direct harvesting and community freezers to grocery stores, family and friends, and increasingly, social media platforms that facilitate both sharing and informal markets. High costs of food and fuel, transportation barriers, time constraints, and other social factors emerged as persistent challenges. Income level and the presence of a skilled hunter strongly influenced decision strategies: low-income households without hunters relied heavily on community freezers and kin networks, while higher-income households accessed additional options such as buying country food from Inuit harvesters and southern grocery stores. These results underscore the adaptive capacity of Nunavik’s food systems alongside their sensitivity to economic and environmental disruptions. Further, these results point to the importance of addressing both structural barriers and digital access. Policies that enhance community freezers, transportation, food assistance programs, and culturally grounded digital tools can strengthen food security and resilience. This study highlights the value of interdisciplinary, community-based approaches that integrate cultural preferences, economic dimensions, and social networks in supporting Arctic food systems under conditions of rapid change.

In the Hudson Bay system, changes in the diets of seabirds and seals over the past four decades suggested a gradual displacement of polar cod (Boreogadus saida) by capelin (Mallotus villosus) and to a lesser extent by sand lance (Ammodytes spp.). However, this hypothetical borealization can only be inferred due to the lack of recent data on abundance and distribution of fish in the region. In this study, we combined age-0 fish samples from various expeditions, forming the most extensive fish dataset in the region to date, and laying down baseline information about distribution and abundances of fish species within the Hudson Bay system. The fish larvae assemblages indicate that there are three key age-0 fish larvae species in the Hudson Bay system; polar cod, capelin, and sand lance. Capelin larvae were by far the most abundant overall (maximum of 5841 individuals 1000 m−3), but assemblages containing polar cod were more widespread across the region and had little geographical overlap with capelin. Polar cod larvae were on average 20 mm larger than capelin larvae, suggesting different prey size spectra and thus little food competition. However, polar cod larvae overlapped temporally, geographically, and in size with northern sand lance (Ammodytes dubius) larvae. We conclude that the co-distribution of ichthyoplankton species in the Hudson Bay system shows no conclusive evidence that capelin have completely displaced polar cod in this region yet.

Faecal pellets of marine zooplankton play a key role in the biological carbon pump, i.e. all biologically mediated processes by which organic carbon produced by photosynthesis is stored in the ocean's interior. Numerous factors (size and biomass of faecal pellets, composition and abundance of zooplankton, etc.) can affect the sinking rate of zooplankton faecal pellets and thus the efficiency of their export at depth. A number of quantitative studies of the role of zooplankton faecal pellets in the biological carbon pump have been conducted, focusing either on a region or a type of faecal pellets. These studies highlighted the large variability in the contribution of faecal pellets to carbon fluxes, ranging from 0% to 100%. Here, we used a meta-analysis approach to extract quantitative data on the size, biomass, and role of marine zooplankton faecal pellets in ocean carbon export from 197 scientific articles. Our study focused on the six most studied faecal pellet types (mixed, cylindrical, ellipsoidal, tabular, spherical, and drop-shaped). We showed that abundance and biomass fluxes of faecal pellets, as well as their contribution to particulate organic carbon fluxes, increased with ecosystem productivity, here approximated by surface chlorophyll-a concentration. Furthermore, the fluxes of marine zooplankton faecal pellets (both by abundance and biomass) were positively correlated, and the sampling location, rather than the type of faecal pellet, better explained this correlation. Additionally, sinking rates were strongly correlated with volume, length, and width of faecal pellets, for all faecal pellet types. Sinking rates did not vary with depth, although measurements become scarcer with depth. Our literature review highlights the crucial role of faecal pellets in the biological carbon pump and the need to study less known types of faecal pellets, such as ellipsoidal faecal pellets, and to measure multiple variables on the same samples. Finally, we recommend that modellers wishing to represent faecal pellets in global biogeochemical models choose a constant sinking rate with depth within the range of the quantitative values reported here.

Oxygen Minimum Zone (OMZ) expansion is a major challenge to marine ecosystems and associated zooplankton. Calanoid copepods include lineages that tolerate hypoxia and exhibit functional traits such as diel vertical migrations to, or dormancy within, hypoxic mesopelagic zones. However, the evolutionary origins and molecular drivers of these traits remain unclear. Herein, we integrate a time-calibrated phylotranscriptomic tree of 50 copepod species with ancestral trait reconstruction, gene family copy number variation, and palaeoceanographic data to infer the evolutionary timing and ecological drivers of hypoxia adaptation. Our results support that post-embryonic dormancy originated in calanoid ancestors, accompanied by widespread gene expansions primarily involving hypoxia-response pathways as well as lipid and amino acid metabolism. Mesopelagic colonisation by calanoid lineages likely occurred during the Ordovician deep-sea oxygenation event. This was followed, during the Carboniferous deep-sea deoxygenation, by a secondary habitat shift toward shallower waters and embryonic dormancy, and gene contractions in the superfamily Diaptomoidea. We further analysed the hypoxia-induced transcriptomic response of Eucalanus hyalinus from the Benguela upwelling OMZ, and identified a coordinated response involving extracellular matrix remodelling, amino acid recycling for anaerobic energy and antioxidant production as well as triglycerides to wax ester conversion. Gene family expansions upstream (proteolysis, transport) and downstream (antioxidant biosynthesis) of core metabolic pathways suggest purifying selection on dosage-sensitive nodes. Together, these results link palaeoclimate change to lineage-specific genome evolution patterns supporting copepod adaptation to oxygen limitation.

Abstract Zooplankton play a crucial role in the biological carbon pump by producing sinking particles including sloppy feeding by-products, fecal pellets, molts and carcasses. However, quantifying their impact of these particles on the carbon cycle remains difficult. The contribution of fecal pellets to particulate organic carbon export is usually assessed using fecal pellets collected from sediment traps and laboratory studies. Here, we identified 50 771 fecal pellet-like particles distributed across three morphological clusters. These were extracted from 987 236 in situ images of non-living particles collected from Baffin Bay (Arctic Ocean) using the Underwater Vision Profiler (UVP). We associated which taxonomic groups produced the fecal pellets by comparing the UVP images with observations of fecal pellet morphology and length. Our results emphasize the feasibility of quantifying fecal pellets from in situ images and the importance of developing the resolution of imaging tools that would simultaneously identify smaller fecal pellet-like particles and capture images of large crustacean zooplankton. Using in situ images in identifying fecal pellets will facilitate a better understanding of their dynamics, a more accurate calculation of carbon fluxes, and the representation of fecal pellets in biogeochemical models.

Mesozooplankton have a pivotal role in marine food webs, linking primary producers to higher trophic levels. Their abundance and traits serve as key indicators of ecosystem structure and function, making them essential components of long-term ocean monitoring. However, the need to monitor biodiversity and functional traits, combined with their pronounced spatial and temporal variability, requires extensive sampling and presents significant laboratory bottlenecks and cost-related challenges. Imaging instruments, combined with automated image classifiers such as Ecotaxa, offer a promising solution by enabling high-throughput, cost-effective processing of large numbers of samples, while also providing highly precise trait measurements previously unattainable with traditional methods. In this study, we compare the performance of human-sorted microscopy, human-sorted images and computer-sorted images across three contrasting coastal ecosystems on Canada's Pacific and Atlantic coasts. First, we demonstrated that upfront investment in identifying a larger number of images contributed to the development of robust regional image libraries, which significantly enhanced the performance of automated classifiers (e.g., mean F1 score = 0.54 with up to 200 images per taxon and 0.68 with up to 5000 images per taxon). Results showed that automated image classification performance varies with specimen characteristics such as symmetry, geodesic thickness, and taxa richness. We then assessed how each method captures local mesozooplankton diversity and altered key ecological indicators. Based on observed ecosystem-specific differences, we provide recommendations for optimizing classification workflows in relation to local diversity patterns. This study provides large-scale empirical evidence that investing in the development of regional image libraries enhances the scalability and accuracy of coastal ecological assessments. These emerging digital assets have the potential to significantly advance ecosystem monitoring and management.

Calanus hyperboreus is a large-bodied, biomass dominant species that performs a crucial ecosystem energy transfer by converting the spring phytoplankton bloom into lipid reserves that fuel the higher trophic levels of the Gulf of St. Lawrence (GSL) pelagic ecosystem, including the critically endangered North Atlantic right whale (Eubalena glacialis). Given that the GSL, the southernmost core habitat of C. hyperboreus, is undergoing rapid warming, developing a population model allows us to synthesize existing knowledge of the species, and to examine the species response to environmental conditions. To simulate the multi-year life cycle in the northwest GSL, model equations are implemented for ingestion, assimilation, respiration, egg production, stage development, mortality, and vertical migration behaviors including dormancy entry and exit. The 1-D particle-based model predicts the evolution of individual stage, structural mass, lipid, age, sex, abundance, and egg production, as well as the seasonal evolution of the population structure in the northwest GSL. Individual lipid-based thresholds inform the timing of ontogenetic vertical migration. Life cycle targets defined from a literature review are used to guide model parameterization and assess its performance. The simulated population structure, phenology, and size at stage are generally consistent with observations. Under 10 years of repeat year forcing, the model simulates a quasi-stable overwintering population composed of late stages CIV, CV and CVI. Observations suggest that stage CIV is the first overwintering stage in the GSL, and point to the occurrence of iteroparous females. Using the model, the relative success of diverse dormancy and reproductive phenotypes are explored. Second reproduction females reproduce earlier in winter than first reproduction females, with implications for the ability of the new generation to match the spring bloom and accumulate sufficient lipid to overwinter as stage CIV. Without iteroparity, the time window of reproduction contracts and the population is reduced, underscoring the role of a flexible multi-year life cycle in population success.

In the Arctic Ocean the peak of the phytoplankton bloom occurs around the period of sea ice break-up. Climate change is likely to impact the bloom phenology and its crucial contribution to the production dynamics of Arctic marine ecosystems. Here we explore and quantify controls on the timing of the spring bloom using a one-dimensional biogeochemical/ecosystem model configured for coastal western Baffin Bay. The model reproduces the observations made on the phenology and the assemblage of the phytoplankton community from an ice camp in the region. Using sensitivity experiments, we found that two essential controls on the timing of the spring bloom were the biomass of phytoplankton before bloom initiation and the light under sea ice before sea ice break-up. The level of nitrate before bloom initiation was less important. The bloom peak was delayed up to 20 days if the overwintering phytoplankton biomass was too low. This result highlights the importance of phytoplankton survival mechanisms during polar winter to the pelagic ecosystem of the Arctic Ocean and the spring bloom dynamics.

In Arctic marine ecosystems, large planktonic copepods form a crucial hub of matter and energy. Their energy-rich lipid stores play a central role in marine trophic networks and the biological carbon pump. Since the past similar to 15 years, in situ imaging devices provide images whose resolution allows us to estimate an individual copepod's lipid sac volume, and this reveals many ecological information inaccessible otherwise. One such device is the Lightframe On-sight Keyspecies Investigation. However, when done manually, weeks of work are needed by trained personnel to obtain such information for only a handful of sampled images. We removed this hurdle by training a machine learning algorithm (a convolutional neural network) to estimate the lipid content of individual Arctic copepods from the in situ images. This algorithm obtains such information at a speed (a few minutes) and a resolution (individuals, over half a meter on the vertical), allowing us to revisit historical datasets of in situ images to better understand the dynamics of lipid production and distribution and to develop efficient monitoring protocols at a moment when marine ecosystems are facing rapid upheavals and increasing threats.

This article is intended as an introduction to discuss the development of a modelling framework to examine simulated climate change and river discharge regulation and their combined impact on marine conditions in the Hudson Bay Complex as a contribution to BaySys, a collaborative project between Manitoba Hydro, Hydro-Quebec, the University of Manitoba, the University of Alberta, Université Laval and Ouranos. In support of this work, a sea ice and oceanographic model was improved and then used to further study the effects of freshwater loading and ice cover on the circulation of Hudson Bay. This modelling perspective is based on the Nucleus for European Modelling of the Ocean (NEMO) ocean general circulation model coupled to version 2 of the Louvain-la-Neuve sea ice model (LIM2). The goal of the modelling was to provide a framework and tool for simulating projected changes in marine state and dynamic variables, while also enabling an integration of observations and numerical analyses. A key aspect of this work was the climate-hydrologic-ocean model integration aspect. The inclusion of a biogeochemical model and explicit tidal forcing to examine the evolution of a Canadian marginal sea with century-long integrations was also a novel aspect of the work. Overall, this work examines the NEMO modelling configuration used in BaySys, how it is set up and the experiments carried out. A broader picture evaluation of the model output is made including the BaySys mooring observations, showing that the modelling framework is suitable to examine the posed questions on the role of climate change and river regulation.

The strong seasonality of sub-Arctic seas needs to be considered to understand their ecosystems. The Hudson Bay system undergoes strong seasonal changes in 1) sea ice conditions, alternating between complete ice cover in winter and open water in summer; 2) river discharge, which peaks in the spring and influences the stratification of the bay; and 3) surface circulation that consists of a weak double gyre system in spring and summer and a cyclonic system in the autumn. Recent studies that included data collected during spring have shown that the annual primary productivity in the Hudson Bay system is higher than previously reported. Similarly, the regional zooplankton assemblages have been studied mostly in late summer, possibly leading to an underestimation of the annual secondary production. Here, we use data collected during five one to six week-long expeditions of the CCGS Amundsen in the Hudson Bay system between 2005 and 2018 to describe the seasonality in mesozooplankton assemblages and investigate how it depends on environmental variables. In general, small pan-Arctic and boreal copepods such as Microcalanus spp., Oithona similis and Pseudocalanus spp. dominated the assemblages. From spring to summer, the relative abundance of the Arctic-adapted Calanus hyperboreus and Calanus glacialis decreased, while the proportion of the boreal Pseudocalanus spp. and Acartia spp. increased. The day of the year and the ice break-up date explained most of the variation in mesozooplankton assemblages. Physical processes explained most of the species distribution in spring, while the lack of lipid-rich zooplankton species in late summer and autumn, especially in coastal regions, suggests some top-down control. This lack of lipid-rich zooplankton late in the season contrasts with other seasonally ice-covered seas. More data are needed to fully understand the implications of these dynamics under climate change, but this study establishes a baseline against which future changes can be compared.

Multiple factors influence the spatial and temporal chlorophyll‐ a concentration of marine systems. The Hudson Bay Complex has historically been seen as a large, low‐production inland sea situated in the north of Canada. However, recent field campaigns, for the BaySys project, have provided new data on primary production in the bay. Due to the Hudson Bay complex's positioning, it experiences seasonal sea‐ice cover and has many rivers draining into it, resulting in a unique estuarine‐like environment. We use the biogeochemical model BLINGv0 + DIC, coupled to the online regional physical oceanographic and sea‐ice models, NEMOv3.6 and LIM2, respectively, forced with two bias‐corrected Coupled Model Intercomparison Project 5 climate forcings (MIROC5 and MRI) to simulate the base of the ecosystem. The simulations were evaluated with chlorophyll‐ a satellite imagery and observations collected in 2018 and analyzed with Empirical Orthogonal Functions to understand the underlying physical forcings and key areas of chlorophyll‐ a concentration distribution. The evaluation showed that both simulations successfully reproduced the sea‐ice melt, from west to east and formation, from north to south and correlated well with spatial bloom patterns. The main drivers of phytoplankton growth are the seasonal light and nutrient levels (48% and 54%), the mixed layer depth dynamics (18% and 14%), nutrient supply from rivers (13% and 8%), and sea ice production (7%) for the MIROC5 and MRI simulations, respectively. The sea‐ice dynamics and river runoff played a significant role in the system's productivity. Therefore, with future climate change and increased river regulation projects, up to 20% of overall chlorophyll‐ a may be negatively impacted.

Arctic marine species, from benthos to fish and mammals, are essential for food security and sovereignty of Inuit people. Inuit food security is dependent on the availability, accessibility, quality, and sustainability of country food resources. However, climate change effects are threatening Inuit food systems through changes in abundance and nutritional quality of locally harvested species, while foundational knowledge of Arctic food webs remains elusive. Here, we summarized scientific knowledge available for the western Baffin Bay coastal and shelf ecosystem by building a food web model using the Ecopath with Ecosim modeling framework. Based on this model, we calculated ecological network analysis indices to describe structure and function of the system. We used Linear Inverse Modeling and Monte Carlo analysis to assess parameter uncertainty, generating plausible parameterizations of this ecosystem from which a probability density distribution for each index was generated. Our findings suggest that the system is controlled by intermediate trophic levels, highlighting the key role of Arctic cod (Boreogadus saida) as prey fish, as well as the importance of other less studied groups like cephalopods in controlling energy flows. Most of the ecosystem biomass is retained in the system, with very little lost to subsistence harvest and commercial fisheries, indicating that these activities were within a sustainable range during the modeling period. Our model also highlights the scientific knowledge gaps that still exist (e.g., species abundances), including valued harvest species like Arctic char (Salvelinus alpinus), walrus (Odobenus rosmarus), and seals, and importantly our poor understanding of the system in winter. Moving forward, we will collaborate with Inuit partners in Qikiqtarjuaq, NU, Canada, to improve this modeling tool by including Inuit knowledge. This tool thus serves as a starting point for collaborative discussions with Inuit partners and how its use can better inform local and regional decision-making regarding food security.

Pigmentation is often overlooked in zooplankton, since these organisms are mostly colorless to fit the translucid water medium. However, one of the dominant zooplankton taxa in aquatic ecosystems-copepods-often show a bright red-orange or blue coloration owing to the accumulation of carotenoid pigments in some parts of their bodies. Even though there are many functional traits describing copepod's performance (e.g., size, feeding, and reproductive modes), it is surprising that the role of such a simple and visible trait as coloration has not been studied in a coherent manner yet. Here, by reviewing 95 studies, we demonstrate that carotenoid-based pigmentation (mainly caused by astaxanthin molecules) is a widespread functional trait in freshwater and marine copepods. We propose a way to disentangle the complex and thus intriguing patterns of pigment expression along latitudinal and altitudinal gradients, addressing its relationship to diet quality and quantity, temperature, ultraviolet radiation stress, predation pressure, lipid metabolism, and reproduction. We show that large-scale variations in pigmentation are difficult to tackle because of the fundamental plasticity of this trait at short time scales (i.e., hours, days), and the most recent information about carotenoid bioconversion are addressed (genes and enzyme identification, and influence of microbiota). From this literature review, we hypothesize that pigments play a "Swiss-army knife" role for copepod's fitness, useful in various ecosystem conditions owing to the strong antioxidant power and the finely-tuned metabolism of astaxanthin. With larger antioxidant capacities (survival), higher metabolisms (growth), and more offspring in better condition (reproduction), red morphs appear more successful than their uncolored siblings. Also, the potential camouflage strategies allowed by red and blue pigmentation are discussed. We finally formulate new directions and future research fields from molecular to ecosystem scales. Routine quantifications of copepod's pigmentation through trait-based approaches could be useful (1) to obtain an accurate copepod fitness indicator and (2) to better estimate the transfer of antioxidant to higher trophic levels in ecosystems, including humans.

The trophic relationships interconnecting marine organisms together into a dynamic trophic network drive the structure and the functioning of the entire ecosystem. Since the flow of carbon within trophic networks is controlled by a variety of functional traits related to food acquisition and individual survival, it is crucial to understand how functional diversity relates to marine ecosystems properties such as the resistance and resilience against perturbations. In the Arctic, marine ecosystems are facing stronger and faster environmental changes than anywhere on Earth, leading to profound perturbations in the planktonic assemblages at the base of the trophic networks. While it is known that mesozooplankton plays a crucial role of matter and energy hub within marine Arctic food web, the precise role of the diverse mesozooplankton functional groups in carbon circulation and in marine ecosystems functioning remains poorly known. We coupled a trait-based approach of mesozooplankton diversity to an ecological network analysis approach to test whether similar mesozooplankton functional groups played similar ecological roles in three Arctic ecosystems during the summer period. We formed nine mesozooplankton functional groups by gathering different species according to their feeding strategies. Then we implemented those into inverse food web models (linear inverse modelling) describing three contrasted Arctic ecosystems. In each ecosystem, we performed sensitivity analysis experiments where each mesozooplankton functional group was removed one at a time. Our results showed that, although the same main functional groups composed the three ecosystems, the few outstanding changes observed in the carbon circulation within the food web were strongly controlled by both the initial whole-network properties and productivity of the ecosystem. The various roles played by a given mesozooplankton functional group in the ecosystem depend on its impact on carbon flows through the food web it belongs to. As a result, identifying which functional groups could be threatened, and which carbon flows could be altered by climate change is critical information to predict future ecosystems functioning. Read the free Plain Language Summary for this article on the Journal blog.

Plankton imaging systems supported by automated classification and analysis have improved ecologists' ability to observe aquatic ecosystems. Today, we are on the cusp of reliably tracking plankton populations with a suite of lab-based and in situ tools, collecting imaging data at unprecedentedly fine spatial and temporal scales. But these data have potential well beyond examining the abundances of different taxa; the individual images themselves contain a wealth of information on functional traits. Here, we outline traits that could be measured from image data, suggest machine learning and computer vision approaches to extract functional trait information from the images, and discuss promising avenues for novel studies. The approaches we discuss are data agnostic and are broadly applicable to imagery of other aquatic or terrestrial organisms.