Western Arctic primary productivity regulated by shelf-break warm eddies

Abstract

The response of phytoplankton to the Beaufort shelf-break eddies in the western Arctic Ocean is examined using the eddy-resolving coupled sea ice–ocean model including a lower-trophic marine ecosystem formulation. The regional model driven by the reanalysis 2003 atmospheric forcing from March to November captures the major spatial and temporal features of phytoplankton bloom following summertime sea ice retreat in the shallow Chukchi shelf and Barrow Canyon. The shelf-break warm eddies spawned north of the Barrow Canyon initially transport the Chukchi shelf water with high primary productivity toward the Canada Basin interior. In the eddy-developing period, the anti-cyclonic rotational flow along the outer edge of each eddy moving offshore occasionally traps the shelf water. The primary production inside the warm eddies is maintained by internal dynamics in the eddy-maturity period. In particular, the surface central area of an anti-cyclonic eddy acquires adequate light, nutrient, and warm environment for photosynthetic activity partly attributed to turbulent mixing with underlying nutrient-rich water. The simulated biogeochemical properties with the dominance of small-size phytoplankton inside the warm eddies are consistent with the observational findings in the western Arctic Ocean. It is also suggested that the light limitation before autumn sea ice freezing shuts down the primary production in the shelf-break eddies in spite of nutrient recovery. These results indicate that the time lag between the phytoplankton bloom in the shelf region following the summertime sea ice retreat and the eddy generation along the Beaufort shelf break is an important index to determine biological regimes in the Canada Basin.

Appendix: Photosynthesis formulation in NEMURO

The NEMURO model configuration assumes that the photosynthesis of PS and PL is a function of nitrate NO₃, ammonium NH₄, and silicate Si(OH)₄ concentration, water temperature T, light intensity I (Kishi et al. 2007). The formulation of gross primary production rate of PS (GppPS) and PL (GppPL) consists of nitrate, ammonium, silicate uptake terms (Nit, Amn, Sil, respectively), temperature-dependent term (Tmp), and light availability term (LA), and the biomass itself PS and PL as follows:

$$ \begin{aligned} \rm{GppPS} &= V_{\rm maxs} \left( \underbrace{ \frac{{\hbox{NO}}_3}{{\hbox{NO}}_3+K_{\hbox{NO}_3}} \exp(-\Psi {\hbox{NH}}_4) }_{[\rm{Nit}]} + \underbrace{ \frac{{\hbox{NH}}_4}{{\hbox{NH}}_4+K_{\hbox{NH}_4}} }_{[\rm{Amn}]} \right) \\ &\qquad \times\underbrace{ \exp(\kappa T) }_{[\rm{Tmp}]} \underbrace{ \frac{I}{I_{\rm opt}} \exp\left( 1-\frac{I}{I_{\rm opt}} \right) }_{[\rm{LA}]} \underbrace{ {\hbox{PS}}}_{[\rm{PS}]} , \\ \rm{GppPL} &= V_{\rm maxl} \min \left( \frac{{\hbox{NO}}_3}{{\hbox{NO}}_3+K_{\hbox{NO}_3}} \exp(-\Psi {\hbox{NH}}_4) + \frac{{\hbox{NH}}_4}{{\hbox{NH}}_4+K_{\hbox{NH}_4}}, \underbrace{ \frac{{\hbox{Si(OH)}}_4}{{\hbox{Si(OH)}}_4+K_{\hbox{Si(OH)}_4}} / \rm{RSiN} }_{[\rm{Sil}]} \right) \\ & \qquad \times\exp(\kappa T) \frac{I}{I_{\rm opt}} \exp\left( 1-\frac{I}{I_{\rm opt}} \right) \underbrace{ {\hbox{PL}} }_{[\rm{PL}]} , \end{aligned} $$

where the parameter values of maximum photosynthetic rate at 0 °C V _maxs (0.7 day⁻¹), V _maxl (2.0 day⁻¹), half-saturation constant for nitrate $K_{\rm{{NO}_3}}$ (0.7 μM), ammonium $K_{\rm{{NH}_4}}$ (0.2 μM), silicate $K_{\rm{{Si(OH)}_4}}$ (1.15 μM), ammonium inhibition coefficient $\Psi$ (1 $\upmu \text{M}^{-1}$), temperature coefficient for photosynthetic rate κ (0.0693 °C⁻¹), optimum light intensity constant I _opt (104.7 $\text{W\,m}^{-2}$), and Si:N ratio RSiN of 2 are all given based on Zhang et al. (2010). The nutrient uptake, Tmp, and LA terms are represented by traditional Michaelis–Menten, Q ₁₀ relationship, and P–E curve formulation, respectively.