Estimate of global yearly N assimilation by photolithotrophs are 417 Tmol N in the oceans and 167 Tmol on land and in freshwater, of which diazotrophy contributes 1 (sea) and 10 (land plus freshwater) Tmol N. More than half of the combined N assimilated (416 and 157 Tmol N year(-1) in the sea and on land plus freshwater, respectively) is due to reduced N, i.e. organic N and, mainly, NH3/NH4+. Assimilation of reduced N amounts to up to 334 Tmol N year(-1) in the oceans and at least 79 Tmol N year(-1) in freshwater and on land. Reassimilation of NH3/NH4+ within the plant which is related to photorespiration is at least as great as primary NH3/NK4+ assimilation in the sea, and 8 times greater on land. The less frequently considered reassimilation of NH(3)MH(4)(+) that is related to phenylpropanoid (mainly lignin) synthesis in land plants is similar(lll Tmol N) to the primary assimilation of NH3/NH4+ on land each year. Shoots of terrestrial plants have higher NH3 compensation partial pressures than most natural soils, and especially than have ocean-surface biota. However, gaseous transfer of NH3/NH4+ from land to the oceans is a negligible component of the global N cycle. Consideration of area-based N assimilation rates. diffusion distances and diffusion coefficients can rationalise why steady-state NH3/NH4+ concentrations in the sea are lower than in the soil solution. The possibility that photolithotrophs can catalyse the oxidation of NH3/NH4+, or organic N at the same redox level, to N-2, N2O, NO, -NO2, NO2-, NO2 or NO3-, is critically assessed. The tentative conclusions are that such oxidation probably occurs, but is not a major component of the global conversion of reduced N to N-2 and more oxidized N species. More work is needed, especially to determine if NO generated from reduced N (conversion of arginine to citrulline plus NO) has a regulatory role in plants analogous to that established in metazoa. Relative to NO3- (or N-2) as N sources, growth using NH3/NH4+ as N source has a number of potential advantages in terms of cost of other resources. Mechanistically predicted economies for NH4+ as N source are: (1) lower cost of photons used and, in transpiring plants, (2) less water lost per unit C assimilated, and (3) lower costs of catalytic Fe, Mn and Mo (unit C assimilated)(-1) s(-1), as well as (4) a higher maximum growth rate. The lower photon costs are frequently borne out by experimentation and the predicted higher maximum growth rates sometimes occur, while the predicted lower water costs are invariably contradicted. Few data are available for the cost of Fe, Mn or Mo as a function N source.
