As sea ice melts, areas of once permanent sea ice are now shifting to seasonal sea ice zones, with concomitant changes in light availability, upper-ocean mixing, and community structure. In this issue, Matthes et al. document the spatial variability of UV and PAR transmission through melting sea ice and conclude that spatial averages in transmission are more representative than single point irradiance measurements used for estimating nutrient availability and, by inference, primary production. Marmillot et al. exploit spatial variations in physiochemical seawater properties to explore its relationship with lipid and fatty acid distributions, and highlight the importance of long-lived subsurface chlorophyll maximum layers in supplying PUFA-rich POM to the food web. Using biogeochemical modelling, Benkort et al. include a sea-ice component for the Barents Sea, which links the dynamics of the sea-ice, pelagic and benthic environments. Their findings indicate the important role of sea-ice algae in influencing the timing and amplitude of pelagic primary and secondary production, and in seeding pelagic diatoms.

The Arctic Ocean is surrounded by the northernmost regions of the American and Eurasian continents where glaciers and permafrost are decreasing in areal extent (Chadburn et al., 2017). The Arctic basin is unique in that it holds less than 2% of the ocean’s volume but receives 10% of global riverine discharge and, therefore, processes occurring in Arctic rivers can have disproportionate consequences for biogeochemical cycling across the wider Arctic Ocean (Holmes et al., 2012). As climate forcing progresses, increases in riverine discharge associated with increases in precipitation and permafrost thaw emphasize land–ocean connectivity. McGovern et al. describe the effects of increased terrestrial riverine input on a fjordal system in Svalbard, and highlight the need for detailed and high resolution sampling to explore biogeochemical and ecological responses over time. Similarly, Juhls et al. describe the dynamics of the Lena River biogeochemistry. Using an unprecedented high temporal frequency of samples, they reveal seasonal changes in the composition and sources of dissolved organic matter (DOM). These new data indicate a shift in subsurface DOM properties towards older sources which are mobilized from within deeper soil horizons and permafrost deposits, raising concerns about positive climatic feedbacks.

Changes to biogeochemical properties in the Arctic can also influence the broader ecosystem, although the mechanistic basis of many of these linkages are poorly understood, and do not always follow expected patterns. O’Daly et al. measure sinking particulate organic carbon during a particularly ice free and warm summer in the Pacific-influenced Arctic. Contrary to expectations, they find high carbon fluxes which suggest the potential for high productivity in a warming Arctic ocean. Meanwhile, using a mesocosm study, Reed et al. investigate the reproductive response of abundant benthic invertebrates to projected temperature and CO₂ concentrations. They show no change in oocyte size frequency, and suggest that the quantity and quality of food, often available ad libitum in laboratory experiments, is likely to be an important determinant of physiological responses to projected environmental change.

The characteristics and degradability of organic matter has important implications for climate. Jongejans et al. describe the organic matter characteristics in a thermokarst lake in Siberia and postulate on the ecological landscape through time while exploring the degradation of organic matter through permafrost thaw. They find that the organic carbon inventory of thawed permafrost reflects poor deposition, partial mobilization and release as methane from the lake, while the frozen elements of the permafrost indicate that the input signal of the organic matter still exceeds the degradation signal from thaw underneath. Their study indicates that changes in environmental circumstance can have substantive effects on organic matter retention and release. Indeed, Saidi-Mehrabad et al. considers whether or not climatic-induced transitions in state could occur in Arctic regions by examining how soil chemical and microbial parameters operate across the Pleistocene–Holocene boundary. As modern cold-adapted systems near a climatic threshold they conclude that cold soils could transition in ways similar to those seen across the Pleistocene–Holocene boundary with unknown ecosystem consequences.