The zooplankton community composition in the Pacific Arctic Ocean depends heavily on the sea ice and hydrodynamic conditions. The calanoid copepod Calanus hyperboreus is a dominant Arctic zooplankton, but its diet sources in the Pacific Arctic are unclear. C. hyperboreus individuals were collected in the northern Chukchi Sea (NCS) and the northern‐East Siberian Sea (NESS), which display contrasting chlorophyll abundance and sea ice concentration in summer 2018. Nitrogen isotopes of individual amino acid and fatty acid compositions were measured to determine its trophic level (TL) and dominant diet. C. hyperboreus collected in the NCS had higher TL values (3.0 ± 0.2) and a relatively low proportion of C20:5(n‐3). In the NESS, C. hyperboreus had lower TL values (2.7 ± 0.2) and a larger percentage of C20:5(n‐3), suggesting a greater proportion of diatoms in its diet. Spatial TL variations of C. hyperboreus between the NCS and NESS may be caused by variations in phytoplankton composition (micro‐phytoplankton vs. nano‐ and picophytoplankton) as a result of surface seawater stratification triggered by the inflow of sea ice meltwater as well as summer Bering Sea water through the Bering Strait. These results suggest that the filter‐feeding copepod C. hyperboreus could be a useful indicator to understand trophic dynamics in zooplankton food web.

Despite their crucial roles in transporting primary productions in marine food webs, the trophic dynamics of zooplankton throughout the seasons have rarely been studied. In this study, four dominant zooplankton taxa with phytoplankton size composition and productivity were collected over four seasons in the East Sea, which is known to change more rapidly than global trends. We then analyzed the δ¹³C and δ¹⁵N values and fatty acid composition of zooplankton. The heavy δ¹³C values in February and August 2021 were observed with high concentrations of total chlorophyll-a, and the δ¹³C differences among the four zooplankton taxa in the coastal region (site 105-05) were most pronounced in February 2021. The relative amounts of eicosapentaenoic acid (C20:5(n-3)) and docosahexaenoic acid (C22:6(n-3)), indicators of phytoplankton nutritional quality, were also highest in February 2021. Non-metric multivariate analyses showed dissimilarity among zooplankton taxa during the high productivity period based on chlorophyll-a concentrations (51.6%), which may be due to an increase in available foods during the highly productive season. In conclusion, the dietary intake of zooplankton can be reduced by the transition of phytoplankton, which has important implications for the impact of climate change on planktonic ecosystems in the East Sea.

Nitrogen isotope ratios of amino acids (δ¹⁵N_AA) have widely been employed as a powerful tool for estimating the trophic position of organisms in food webs. This estimation is based on an elevation pattern in the δ¹⁵N_AA value, a large elevation for the trophic amino acids (e.g., 8.0‰ for glutamic acid) while a little elevation for the source amino acids (e.g., 0.4‰ for phenylalanine), associated with the deamination of amino acids in consumers. However, several elevation patterns can be found in natural environments, one of which may be caused by the hydrolysis of specific protein under nutritional stress. In the present study, we identified ‘protein-specific elevation pattern’ for collagen, which connected to muscle tissues of fish and gastropods under nutritional stress. Time-series starvation for 45 days results in that collagen is consumed by 86% for the fish Girella punctata and by 50% for the gastropod Turbo sazae. Moreover, during the starvation, although a little change in the δ¹⁵N_AA value is found in muscle fibers, a gradual elevation in the δ¹⁵N_AA value is found in collagen for both trophic and source amino acids (e.g., by up to 11‰ for glutamic acid and up to 3‰ for phenylalanine, respectively, which can be explained by the Rayleigh fractionation model). We thus demonstrate that the consumption of collagen in organisms under starvation shows a unique elevation pattern in the δ¹⁵N_AA value, which is consistent with the observation that collagen is degraded non-quantitatively by the collagenase reaction, whereas muscle is degraded quantitatively by the chaperone-mediated autophagy. The effect of δ¹⁵N elevation in all amino acids of collagen is negligible for G. punctata and diluted for T. sazae to the elevation in whole muscle tissue (i.e., = muscle fibers + collagen) even under long-term starvation, because the collagen proportion in the whole muscle is relatively small and is considerably decreased under the nutritional stress. Based on these results, we predict that the difference in the δ¹⁵N_AA value between collagen and muscle fibers can be useful for evaluating the nutritional stress of fishes. However, great care will be required if studying food webs where collagen-rich organisms (i.e., gelatinous zooplankton) are abundant or if using collagen-rich materials in our studies.