Nitrogen flux in the northern Bering Sea

Abstract

Much of the primary production occurring over the Bering Sea continental shelf is thought to be associated with both ice edge and spring blooms. The nature of summer production over the shelf is now being addressed. A general model is presented for summer phytoplankton production along the Bering Sea shelf break front and subsequent transport of phytodetritus into the northern Bering Sea. Production associated with the shelf break front is estimated to be 200 g $\cdot$ C $\cdot$ m$\sp{-2}$ over a 120-day growing season and is supported by nutrients from the Bering Slope Current. A portion of the biomass accumulating over the front is advected into the Chirikov Basin, supplying 26% of the daily carbon demand of the benthos. The Bering Slope Current bifurcates at Cape Navarin and one branch, referred to as the Anadyr Current, flows north through Anadyr and Bering Straits. Nutrients in the Anadyr Current support an intense surface bloom over the western Chirikov Basin where total nitrogen uptake rates are $>$6.0 mg-at N $\cdot$ m$\sp{-2} \cdot$ h$\sp{-1}$ and nitrate contributes up to 50% of the total nitrogen uptake. Modified Bering Shelf water contains phytoplankton at two depths: both a surface accumulation and a deep layer. Nitrate contributes $<$35% to total nitrogen uptake rates of 1.80 mg-at N $\cdot$ m$\sp{-2} \cdot$ h$\sp{-1}$ in this water. Nitrogen productivity is lowest in Alaskan Coastal water (1.0 mg-at N $\cdot$ m$\sp{-2} \cdot$ h$\sp{-1}$) where nitrate uptake averages only 15% of the total. A simple nitrogen budget suggests that 29% and 62% of the annual nitrogen productivity in modified Bering Shelf and Anadyr waters, respectively, is exported through Bering Strait into the southern Chukchi Sea for deposition. Improved estimates of the rates of urea production and uptake by phytoplankton in the northern Bering Sea were made after determining the change in $\sp{15}$N-atom % enrichment of urea during incubations. Estimates of uptake rates increased by up to 83% using a $\sp{15}$N accumulation model and by $>$210% using a $\sp{15}$N disappearance model. However, a discrepancy exists between the $\sp{15}$N-urea removed from the aqueous phase and the $\sp{15}$N accumulated in the particulate phase. The ability to find in the particulate fraction the $\sp{15}$N removed from solution as $\sp{15}$N-urea was improved by 72% following removal of the $>$20-$\mu$m particulate fraction.