Arctic mid-winter phytoplankton growth revealed by autonomous profilers

Abstract

It is widely believed that during winter and spring, Arctic marine phytoplankton cannot grow until sea ice and snow cover start melting and transmit sufficient irradiance, but there is little observational evidence for that paradigm. To explore the life of phytoplankton during and after the polar night, we used robotic ice-avoiding profiling floats to measure ocean optics and phytoplankton characteristics continuously through two annual cycles in Baffin Bay, an Arctic sea that is covered by ice for 7 months a year. We demonstrate that net phytoplankton growth occurred even under 100% ice cover as early as February and that it resulted at least partly from photosynthesis. This highlights the adaptation of Arctic phytoplankton to extreme low-light conditions, which may be key to their survival before seeding the spring bloom.

Fig. 1. Baffin Bay, an Arctic Sea.

(A) Location of Baffin Bay between the Arctic and Atlantic Oceans [inset marking the location of (B)]. (B) Float trajectories from 2017 to 2019, with winter trajectories interpolated (dotted lines). (C and D) Temperature-salinity plots with marginal histograms for 2 months at a time show a strong seasonality in the upper 50 m (C) and consistent water masses throughout the year in subsurface layers (50 to 250 m) (D).

Fig. 2. Annual cycles of phytoplankton biomass.

(A) Mixed layer depth (dashed line, individual) and isolume depths for two different irradiance levels. The “surface layer” is defined as the maximum of mixed layer depth and the depth of the 0.4 mol photons m⁻² d⁻¹ isolume for any given profile. X axis labels indicate the start of each month. (B and D) Vertical profiles of two proxies of phytoplankton biomass (B: chlorophyll a fluorescence, chl-a, D: particle backscattering at 700 nm, bbp) show a strong seasonality with rising biomass in winter and spring and a subsurface maximum in summer. (C) Surface-layer integrated chl-a and bbp as observed by four autonomous floats (different markers). (E and F) Surface-layer averaged chl-a and bbp. Horizontal bars indicate the 750- to 800-m depth averaged background values and their variability (±1 SD) (E) Full time series 2017–2019. (F) Time series as in (E) but collapsed into one annual cycle and smoothed using a generalized additive model (solid lines). (G) Net specific growth rates r as calculated from measured chl-a (green tint) and bbp (violet tint), treating each float as a separate time series and afterward averaged in 28-day bins. Growth rates based on bbp turned positive later in winter because increases are masked by environmental noise, shown in (E) and (F). Phytoplankton cell division rates μ (black dots) calculated from measured photosynthetic parameters and each profile’s in situ light field. White dots, μ calculated from PAR values at the noise level. (H) Net specific growth rates r calculated from the smoothed time series of surface layer biomass, mixed layer depths, and isolume depths. Black curve, Smoothed phytoplankton cell division rates μ.

Fig. 3. Environmental constraints of Arctic phytoplankton throughout the year.

(A) Ocean-atmosphere heat flux. Positive values indicate energy leaving the ocean, leading to loss of buoyancy and vigorous mixing. (B) Sea ice concentration, averaged across the study area. (C) Sun elevation angle. During the polar night, the sun does not rise for 2 months.