Why are some areas of the world's oceans rich in plant growth, whereas others are biological deserts? The answer is that more than half the ocean has too little nitrogen to support the growth of microscopic plants, or phytoplankton. But where does the missing nitrogen go?

For decades most oceanographers, including Bess Ward at Princeton University in New Jersey, thought 'denitrification' was responsible for virtually all nitrogen loss in the ocean. In oxygen-poor areas of the ocean, anaerobic bacteria, which do not rely on oxygen for their growth, use biologically available forms of nitrogen, such as nitrates and nitrites, for energy generation, producing, through several intermediate steps, dinitrogen (N2) — a form of nitrogen that plants cannot use.

But starting in 1995, investigators in the Netherlands, Germany and Denmark uncovered an additional biochemical reaction, called anaerobic ammonia oxidation, or anammox, that produces N2 in ocean waters. Here, the bacteria responsible use up nitrite (NO2−) and ammonia (NH4+) to directly generate N2.

The Europeans showed that anammox, in the absence of denitrification, accounted for most, if not all, N2 generation. “But given the chemistry of the two processes it was impossible for us to understand how that could be,” says Ward.

The problem, she thought, was that nitrite and ammonium are intermediate products in the denitrification process. And in oxygen-poor waters, denitrification is the only substantial source of these chemicals. “We couldn't understand how anammox could be occurring in the absence of denitrification,” says Ward, “so that's what we set out to resolve.”

This involved sending postdoc Jeremy Rich, now a researcher at Brown University in Providence, Rhode Island, on a trip to Europe. The European investigators had devised new methods for measuring anammox and denitrification in small (10 millilitre) samples of ocean water. “The methods were so new and so different that a lot of people were suspicious there might be some artefact involved,” says Ward. “So Jeremy visited some of our European colleagues and learned the methods, so we could go out and try to reproduce their results.” Ward and her colleagues also developed their own methods for measuring denitrification and anammox in larger (10 litre) samples.

The next step was to apply the techniques to different parts of the ocean. First stop was Chesapeake Bay, on the east coast of the United States, where Ward's team found “exactly what we expected in our own backyard, which was both anammox and denitrification” occurring together, she says. In the Eastern Tropical South Pacific, however, off the coast of Peru, “we got exactly what the Europeans had, which was all anammox and no denitrification”, she says. Convinced that their methods were working, Ward and her colleagues made an additional stop in the Arabian Sea, “where it turned out that it was all denitrification and very little anammox” (page 78).

So in different parts of the ocean, and at different times, a variety of mechanisms may be responsible for marine nitrogen loss. “We used to think we understood denitrification and now we know we don't,” says Ward. The current study “does not increase clarity”, she says. “But it makes it more compelling that we find the answer.”