A unique feature of the polar oceans is formed by sea ice. The annual cycle of freezing and melting causes tremendous seasonal variation in ice-cover. Other factors, such as wind and ocean currents, cause a continuous reshaping of the sea ice resulting in assemblages of ice of different sizes and structures. This highly dynamic feature forms a habitat for life in the polar ocean. Sea ice and its seasonal changes influence the physical features of the environment, such as light availability and water column properties, resulting in fluctuating rates of primary production both within the sea ice and in the water column. This has a large impact on the food availability for higher trophic levels, which also use the sea ice for e.g. reproduction and as a shelter for predation. Zooplankton and nekton species living underneath the sea-ice of the polar oceans have in different ways adapted to the fluctuation in food availability and their seasonally changing habitat. Life cycle events, such as reproduction, are often timed to coincide with peaks in primary production. Furthermore, several species developed different overwintering strategies to cope with the food scarcity during this season, such as relying on reserves, lowering metabolism or shifting diet. Antarctic krill (Euphausia superba) and polar cod (Boreogadus saida) are considered to be key species of the under-ice surface water in the Southern and Arctic Oceans, respectively. Information on how organisms utilize the sea-ice habitat remains incomplete. The understanding of sea-ice ecology has been hampered by the difficulty to collect sufficient samples that allow the identification of large-scale spatial trends. The Surface and Under Ice Trawl (SUIT) was developed to overcome this limitation and has already been used to gain insight in the differences in zooplankton distribution and community structure between open and ice-covered waters in the Southern Ocean. Further knowledge on how polar species utilize the sea ice and how the sea ice affects their distribution is necessary to predict the consequences of ongoing climate change. Particularly in the Arctic a marked decrease in sea-ice cover and thickness has been recorded over the last decades. But also in the Antarctic regional changes in air temperature and sea-ice extent have been observed. Knowledge on the sea-ice associated food web is, furthermore, important for aiding management directed to improve sustainability of ongoing and future fisheries. Antarctic krill and several fish species are harvested in the Southern Ocean, and a northward expansion of commercially harvested fish stock from the sub-Arctic region has been observed. The aims of this thesis were to gain knowledge on the association of marine species with the seaice habitat and the functioning of polar marine food webs by assessing the abundance and distribution of species in ice-covered oceans, the important of sea-ice derived carbon sources, and the variability in energy density of marine species. In the Southern Ocean, the winter season has been regarded as a critical period for larval and juvenile Antarctic krill that hatched in the previous summer (age class 0 or AC0 krill). In Chapter 2, the population structure of AC0 Antarctic krill was studied to investigate their population structure at the ice-water interface layer and look at habitat partitioning of different developmental stages between different depth layers. The SUIT was used to sample the upper two meters of the water column underneath sea ice and a Rectangular Midwater Trawl (RMT) was used to sample deeper water layers. Results showed that the population of AC0 krill in the ice-water interface could be divided in geographically distinct cohorts based on size and developmental stage composition and that the size of the same developmental stages differed between regions. The differences between cohorts could be a result of a different time of spawning and/or different growth rates caused by variability in environmental conditions encountered during advection. The behaviour and physiological differences associated with developmental stages likely cause a change in distribution in the water column. In general, the volumetric density of the AC0 krill was significantly higher in the under-ice surface layer compared to the 0-500m depth stratum, indicating that the composition, abundance and distribution may not be represented well if only conventional pelagic sampling gear is used. To further look at the utilization of the sea-ice habitat, the diet of the AC0 krill during winter was investigated (Chapter 3). Multiple methods were used to study spatial and temporal variability in the diet. Stomach content gives information on the most recent feeding of a consumer, while both fatty acid and stable isotope analyses provide information on trophic interaction over a larger temporal scale. The stomach contents of AC0 krill contained mainly centric and pennate diatoms in terms of abundance, and centric diatoms and copepods in terms of biomass. Identifiable food items mainly consisted of species that are known to reside within or close to the sea-ice. Variation in the stomach content of AC0 krill between stations was mirrored in variation in the zooplankton community assemblage, which was investigated in an earlier study. Differences in fatty acids profiles and stable isotope values were found between cohorts. The fatty acids profile of the cohort with the smallest krill showed the largest difference with the other cohorts. In addition, this cohort had the lowest δ15N and δ13C values which are used as proxies for heterotrophy and the contribution of sea-ice algal produced carbon to the diet, respectively. This cohort also had the lowest C/N ratio, which is used an indicator for lipid storage and body condition. Results suggest that sea-ice resources are the main food source for AC0 krill residing in ice-covered waters during late winter. A lack over sea-ice resources over a longer period may negatively affect the AC0 krill’s condition in ice-covered waters. Another method for studying trophic interactions is studying the energy density of species, which can be used in food web models and for understanding the energy flux in an ecosystem. Prey quality can influence the distribution, behaviour and physiology of predators. The energetic density of a species can be influenced by their diet, physiology and by life cycle events. To investigate the variability in energy density between zooplankton and nekton species and causes of variability within species, the current knowledge on energy density of Southern Ocean species was reviewed (Chapter 4). Previously unpublished data was included. Fish were the most studied organisms regarding energy density measurements. For crustacean species, most measurements were conducted on Antarctic krill, which showed varying energy densities between sexes and developmental stages. For the myctophid fish Electrona antarctica a relationship between size and energy density was found. In addition, relationships between water content and energy density were shown for Electrona antarctica, Gymnoscopelus braueri and Bathylagus antarcticus, the latter showing seasonal variation. For most Southern Ocean marine species, little data was available and a proper regional and seasonal coverage was lacking. Several methods used to measure the energy density of marine species are described and discussed. In the Arctic Ocean, one- and two-year old polar cod are known to reside in the under-ice surface waters. The diet of polar cod was investigated to assess if the fish relies on food provided by sea ice in this habitat (Chapter 5). Similar to Chapter 2, multiple methods are used to study diet composition and carbon sources, including stomach content, fatty acid and stable isotope analyses. In addition, the proportional contribution of ice algal-produced carbon was quantified. Different polar cod tissues were investigated. Stomach contents consisted mainly of the copepods Tisbe spp. and Calanus spp., and the amphipod Apherusa glacialis in terms of numbers, and of A. glacialis in terms of biomass. Feeding rates were sufficient to sustain a good body condition. The lipid content was highest in the liver, suggesting that this is the main lipid depot of polar cod. Ice algal-produced carbon contributed over 50% to the carbon in the tissues of the fish. Results indicate that the sea ice provides sufficient resources for polar cod in the central Arctic Ocean. A loss of sea ice, resulting from climate change, can weaken the advantage of being able to exploit the sea-ice habitat that polar cod has over potential competitors. Amphipods are an important part of the Arctic marine community, with A. glacialis often being the most abundant in the ice-water interface. The influence of sea-ice properties on the distribution of A. glacialis has previously been investigated on a small-sale in research mainly conducted by divers. Different studies report a variety of results, including high numbers of the amphipod found underneath both highly-structured and smooth ice. In Chapter 6, the spring relationship between A. glacialis abundance and distribution, and environmental properties was investigated on a large scale, using data on sea-ice topography, water column properties and oceanographic features. Although, the sample size was too small to unravel likely interacting effects that explain the overall variation, some evidence for environmental drivers of the large-scale abundance of A. glacialis were found. The variable that best explained the abundance of the amphipod during spring was chlorophyll a concentration. When data collected during a summer expedition was added, the variables explaining the abundance best were temperature, salinity and sea-ice structures. This indicated that the trade-off between food availability and predation pressure, and therefore the main driver of A. glacialis distribution, changes between seasons. The findings of this thesis contribute to the understanding of the biology and ecology of key species in the sea-ice food web and life in ice-covered waters. The sea ice habitat provides many functions for organisms in the polar oceans. The dominant role of sea ice for a certain species can shift with, for example, season or age. Regardless of food provisioning being the main reason for species to make use of this substrate or not, species residing underneath it largely feed of sea-ice associated food sources. The unique habitat that sea ice provides deserves careful management.