Low nitrate: phosphate ocean ratios corrected⃛

Abstract

It has become apparent, since publication of our Letter on low nitrate: phosphate ratios in the global ocean (T. Tyrrell and C. S. Law, Nature 387, 793–796; 1997; correction, Nature 393, 396; 1998), that the World Ocean Atlas 1994 (WOA94) database we used contains transcription errors for nutrient measurements. Discrepancies between the original data and the corresponding data sets in WOA94 have been confirmed for the eastern tropical Pacific (identified by A. Longhurst and S. Sutherland), the western North Pacific (M. Aoyama, K. Hirose and K. Ishikawa) and the Agulhas retroflection area south of South Africa (R. Schlitzer and A. Longhurst).

Main

Removal of falsely identified low nitrate:phosphate points (LNP, ([NO₃⁻] ÷ [PO₄^3-]) < 3.0 and [PO₄^3-] > 1.5 μmol kg⁻¹) in these areas shows that: (1) LNP points do not occur in the open ocean away from the coast; and (2) the western and northern North Pacific does not represent a hitherto-unquantified significant site for denitrification, as we originally suggested.

However, by examining other independent and more rigidly controlled data sets and re-evaluating WOA94 data using stricter criteria, we have confirmed that LNP is a feature of the Baltic and Black Seas, the Peruvian upwelling system, the Cariaco trench, the Bering Straits, and Tomales Bay. Although this confirms our original premise that the Redfield ratio is not universally applicable, it is clear that LNP points are restricted to coastal and enclosed shelf sea regions. These observations accord with current understanding of the extent of denitrification.

We thank those named above for detecting data discrepancies, and Steve Smith, Louis Codispoti, Geoff Bailey, Sergey Konovalov, Lee Cooper, Peter McRoy, Chirk Chu, Jean Garside, Chris Garside, Gwo-Ching Gong, Alexander Bychkov and Mikael Krysell for sending data sets.