We present the first nitrogen isotope (δ15N) measurements of planktic foraminifera, paleoceanographically important zooplankton, from the nutrient‐rich waters of the modern Southern Ocean. Foraminifera were collected from net tows in the Subantarctic and Polar Frontal Zones (SAZ and PFZ, respectively) south of Africa during winter 2015 and late summer 2016. In late summer, consistent with preferential uptake of 14N‐nitrate and the progressive, northward depletion of nitrate by phytoplankton across the Southern Ocean, foraminifer tissue and shell‐bound δ15N rise equatorward along with nitrate δ15N. However, foraminifer δ15N is ~3‰ lower than expected for heterotrophs relying on photosynthetic biomass generated directly from nitrate. This discrepancy appears to originate with the particulate organic N (PON) in late‐summer surface waters, the δ15N of which is lowered by ammonium recycling. In winter, when overall productivity and foraminifer production are reduced, foraminifer δ15N is higher (by 4.6 ± 0.8‰ for tissue and by 4.0 ± 1.5‰ for shell‐bound N compared to late summer) and exhibits no clear north‐south trend. These characteristics can also be explained by the feeding‐driven connection of foraminifera to PON, which is elevated in δ15N by net degradation in winter. Therefore, foraminifer δ15N is more closely tied to PON δ15N than to nitrate δ15N in the Southern Ocean mixed layer. Combining our isotope data with previously reported sediment trap fluxes from the western Pacific SAZ/PFZ suggests that, under modern conditions, the late‐summer ammonium recycling signal outweighs that of wintertime decomposition on the annually integrated δ15N of sinking foraminifera.

The Subantarctic Ocean is a sink for atmospheric CO2, largely due to its biological pump, which is enhanced by the influence of the Subantarctic islands on the plankton ecosystem (the so‐called “island mass effect”). The influence of the Prince Edward Islands archipelago in the Indian Subantarctic on the surrounding hydrography and benthos has been well studied; however, over the last two decades, little attention has been paid to the functioning and productivity of its plankton ecosystem. Here, we present the first measurements of primary production at the archipelago in over 20 years and the first‐ever rates of secondary production, interpreted in the context of hydrographic, biogeochemical, and plankton community composition data. In autumn 2017, after the late‐summer bloom, nanophytoplankton dominated the near‐island waters and regenerated nutrients fuelled 76% of phytoplankton growth. Primary production and carbon export potential (inferred from nitrate uptake) reached a local maximum in the inter‐island region, which we attribute to water‐mass retention and stratification over the inter‐island plateau (a manifestation of the island mass effect). We observed a diverse mesozooplankton community, likely a remnant of the late‐summer bloom rather than representing the consumers feeding on the in situ (nano)phytoplankton. We estimate that even after the decay of the late‐summer bloom, roughly a quarter of the planktonic carbon was potentially exportable. This finding implies that the archipelago's upper‐ocean ecosystem sequesters atmospheric CO2 at least through autumn despite high rates of recycling and inefficient trophic transfer, thereby contributing to the island mass effect‐associated strengthening of the Subantarctic Ocean's biological pump.

While marine populations change all the time, sometimes regime shifts involve an entire ecosystem, resulting in crucial and sometimes permanent alterations in the ecosystem trophic web and services. A commonly used method to detect shifts in marine systems is the Sequential t-test Analysis of Regime Shifts (STARS). In this work, we chose to analyse the limits and performance of STARS because of its free open-source software and wide use. For the first time, we tested the STARS algorithm using synthetic time series and autoregressive integrated moving average time series, designed to resemble natural observations. We then applied the information obtained from these tests to investigate the STARS detections on an observed time series, that of Calanus finmarchicus in the North Sea. Our tests indicated that in no experiments did STARS detect 100% of the artificial change points at the exact time of the shift. In most cases, STARS tended to anticipate the shift by a few time units. Overall, we determined STARS to be a good method to detect shifts in observed natural time series, so long as the exact time of the shift is not necessary and the possibility of false positives is taken into account.