Redfield stoichiometry has proved a robust paradigm for the understanding of biological production and export in the ocean on a long-term and a large-scale basis. However, deviations of carbon and nitrogen uptake ratios from the Redfield ratio have been reported. A comprehensive data set including all carbon and nitrogen pools relevant to biological production in the surface ocean (DIC, DIN, DOC, DON, POC, PON) was used to calculate seasonal new production based on carbon and nitrogen uptake in summer along 20°W in the northeast Atlantic Ocean. The 20°W transect between 30 and 60°N covers different trophic states and seasonal stages of the productive surface layer, including early bloom, bloom, post-bloom and non-bloom situations. The spatial pattern has elements of a seasonal progression. We also calculated exported production, i.e., that part of seasonal new production not accumulated in particulate and dissolved pools, again separately for carbon and nitrogen. The pairs of estimates of 'seasonal new production' and 'exported production' allowed us to calculate the C : N ratios of these quantities. While suspended particulate matter in the mixed layer largely conforms to Redfield stoichiometry, marked deviations were observed in carbon and nitrogen uptake and export with progressing season or nutrient depletion. The spring system was characterized by nitrogen overconsumption and the oligotrophic summer system by a marked carbon overconsumption. The C : N ratios of seasonal new as well as exported production increase from early bloom values of 5-6 to values of 10-16 in the post-bloom/oligotrophic system. The summertime accumulation of nitrogen-poor dissolved organic matter can explain only part of this shift.

Unfiltered water samples were taken in 10 ml ampoules (sealed immediately after being acidified with phosphoric acid to pH<2, sampling), or 40 ml screw-lid vials, and measured onboard at sampling or next days, or frozen (-20°C) until being acidified and measured in home labs.Carbon measurement was by high-temperature catalytic oxydation in a 10 cm column packed with 5% Pt on aluminum oxide beads at 900°C in a stream of oxygen, and CO2 detection by infrared extinction after the removal of moisture and SO2 by appropriate traps (cold trap, Mg-percarbonate, Na-pyrophosphate, tin, bronze or Sulfix). The apparatus was the dual channel Dimatek 2000 equipped with a Binos 200 detector.Nitrogen was measured by chemoluminescence detection of NO2 in the combustion gases after leaving the Binos detector in one of the two channels of the setup. Most measurements of samples containing high nitrate were discarded, when data were inconsistent. The nature of nitrate interference is not clear.

Supplement to: Körtzinger, Arne; Koeve, Wolfgang; Kähler, W; Mintrop, Ludger J (2001): C:N ratios in the mixed layer during the productive season in the northeast Atlantic Ocean. Deep Sea Research Part I: Oceanographic Research Papers, 48(3), 661-688