To investigate CO2 sink capacity in the western Arctic Ocean (north of 65°N), we conducted underway, ship-based observations of partial pressures of CO2 (pCO2) and total dissolved inorganic carbon (TCO2) in surface seawater in early winter (November 2018). From these two properties of the seawater inorganic carbon system, we calculated total alkalinity (TA). In the early winter, surface seawater pCO2 in most places was lower than atmospheric pCO2. The weighted mean of the air-sea fluxes of CO2 were calculated to be -7.5 ± 1.6 mmol m−2 d−1. The calculated fluxes implied that the area acted as a moderate sink for atmospheric CO2 in early winter, and its rate of CO2 uptake was comparable to that (-8.0 ± 1.7 mmol m−2 d−1) in summer (late August-September 2017). Spatial variations of surface seawater pCO2 in the early winter could be attributed mostly to conservative changes of TCO2 and TA, which together accounted for more than 70% of the pCO2 variations. In the marginal ice zone, however, there was a drawdown of surface seawater pCO2 by 70-90 matm due to horizontal advection of water with an anomalously high temperature from the Pacific Ocean and its subsequent cooling. We found that TA was an important determinant of the spatial variations of pCO2 in the western Arctic Ocean because of the conservative nature of the changes of TA and TCO2 during mixing of water masses. This conservative behavior was observed in both the early winter and summer.

Long-distance migrations by marine fish have long fascinated scientists, but are difficult to track by visual surveys. Here, we propose a new method to easily and precisely track such migrations using stable nitrogen isotopic composition at the base of the food web (δ15NBase), which can be estimated by using compound-specific isotope analysis. δ15NBase exclusively reflects the δ15N of nitrate in the ocean at a regional scale and is not affected by the trophic position of sampled organisms. We initially constructed a δ15NBase isoscape in the northern North Pacific, and determined retrospective δ15NBase values of chum salmon (Oncorhynchus keta) from their vertebral centra. Then, we estimated the migration routes of chum salmon during their skeletal growth by using a state-space model. Our isotope tracking method successfully reproduced a known chum salmon migration route between the Okhotsk and Bering seas, and indicates the presence of a novel migration route to the eastern Bering Sea Shelf during a later growth stage.