The contribution of the marine biota to air-sea fluxes of CO2 and O2 is often described in terms of biological production concepts, such as new production, export production, and net community production. We evaluate these three quantities using a basin-scale ecosystem-circulation model of the North Atlantic Ocean based on Redfield stoichiometry into which we introduce an artificial tracer which records the biotic contribution to air-sea exchange of gases like O2 and CO2. It is found that on average the biological production rates overestimate the biotically effected air-sea flux by some 20% and, in some regions, even predict the wrong direction. With primary production restricted to the euphotic zone, but respiration extending to farther below, the discrepancy can largely be attributed to the different integration depths used in the different concepts (euphotic zone, surface mixed layer), and on annual and longer timescales, all rates converge when using the base of the winter mixed layer rather than that of the euphotic zone as the reference depth. For the surface carbon budget, which ultimately controls air-sea exchange of CO2, it is irrelevant whether carbon atoms cross this boundary in organic or inorganic speciation. Hence the transports of biotically generated surpluses or deficits of dissolved inorganic matter must also be accounted for. While their contribution amounts to only a few percent on the basin scale, the subduction of newly remineralized inorganic matter can locally account for about half of the biotically effected air-sea flux, for example, in regions of mode-water formation.