Karl et al. (1997):  

Recently, nitrogen fixation and its biogeochemical consequences have been extensively studied (Gruber and Sarmiento, 1997) and quantified in the Arabian Sea (Capone et al., 1998), the open Atlantic (Carpenter et al., 1999) and Pacific oceans (Letelier and Karl, 1996; Karl et al., 1998) and coastal waters (Bell et al., 1999) N₂ fixation is now considered as a significant input in the marine and global nitrogen cycle (Capone et al., 1997), and has been introduced into biogeochemical models (Bisset et al., 1999; Tyrrell, 1999; Walsh et al., 1999). Tyrrell (1999) elaborated a simple one-dimentional, two-box model of the global ocean, which shows how N₂ fixation, although low relative to nitrate input on a global basis, would control on a long term the global amount of available reactive nitrogen and therefore the whole ocean productivity.  

The major N₂ fixing organisms in Tropical Ocean

In marine phytoplankton, only some species affiliated to the phylogenic group of cyanobacteria have the capacity to use dinitrogen to satisfy their N metabolic requirements. Most cyanobacteria in tropical and sub-tropical waters belong to 4 main genus, forming the most fascinating group of marine phytoplankton. (1) Prochlorococcus are the smallest and the most numerous oxygenic phototrophic bacteria  (Chisholm et al., 1992) (2) Synechococcus are fewer and a little larger (Johnson and Sieburth, 1979; Waterbury et al., 1979) (3) Richelia lives in symbiosis in some species of diatoms such as Rhizosolenia (Burford et al., 1995) or Hemiaulus (Villareal, 1991) Richelia form filamentous colonies called trichomes and have heterocysts which are special cells fixing nitrogen. (4) Trichodesmium also form trichomes but have no heterocyst. Trichomes are often associated as spherical aggregates called puffs or as bundles called tufts. Trichodesmium, though lacking heterocysts, are known to fix nitrogen (Dugdale et al., 1961) They are probably the most quantitatively important cyanobacteria in terms of oceanic N input. By contrast, experiments intended to demonstrate N₂ fixation in natural Prochlorococcus and Synechococcus are not yet conclusive to date, even if genetic analyses do not exclude such a possibility (Zehr et al., 1998).  

 N₂ fixation in the oceans: magnitude underestimated

Quantitative studies devoted to N₂ fixation have proved that at the basin scale, the rate of N₂ fixation in the nitrogen balance has been largely underestimated (Gruber and Sarmiento, 1997), in part owing to spatial and temporal undersampling of the marine environment. Most historical estimates have shown low rates of N₂ fixation relative to the total N requirement for primary production in the sea (Karl et al., 1997) By contrast, several recent studies demonstrate the contrary. For example, Karl et al. (1997) have reported that at station ALOHA north of Hawaii, N₂ fixation could satisfy up to 50% of nitrogen requirements needed to sustain the export of particulate matter out of the upper layer. In other words, input of N by diazotrophic activity would represent an important source of « new » nitrogen (sensu Dugdale and Goering, 1967) at ALOHA station but probably also in the whole north and south central gyres, especially during El Niño periods (Karl et al., 1997) .

Furthermore, recent geochemical studies indicate that N₂ fixation rates in the North Atlantic are at least twice as high as rates derived from direct measurements (Hood et al., 1999).