Comparison between chemical and isotopic measurements of biological nitrate utilization: further evidence of low new-production levels in the equatorial Pacific

New-production (nitrate uptake) rates in the equatorial Pacific were estimated by parallel measurements of nitrate disappearance from sea water using a colorimetric method and of ¹⁵N-labelled nitrate (¹⁵NO₃ ⁻) incorporation into particulate organic nitrogen (PON) collected on GF/F filters (net nitrate uptake, conventional ¹⁵N-tracer method) and Anopore (0.2 μm) membranes. Regression analyses of 74 sample pairs gathered during 12 and 24 h productivity experiments revealed a significant positive relationship between decreasing nitrate level and ¹⁵NO₃ ⁻ accumulation into PON retained on GF/F filters, but the slopes of Model I and Model II regression lines were 1.18 and 1.29, respectively, suggesting that 15 to 22% of ¹⁵NO₃ ⁻ removed from the dissolved fraction were lost to another N-pool. Two possible avenues for the missing ¹⁵NO₃ ⁻ have been examined: uptake by submicron particles passed through the GF/F filters, and loss as dissolved organic nitrogen (DON). Nitrate uptake by small cells not recovered on GF/F filters, could be safely eliminated as a cause of loss, since ¹⁵NO₃ ⁻ uptake rates obtained from ¹⁵N entering PON collected on GF/F filters agreed well with those obtained from ¹⁵N entering PON collected on Anopore membranes (32 sample pairs). Inspection of the DON pool of 0.2 μm filtrates for excess-¹⁵N enrichment (20 samples) revealed that in nitrate-rich waters (equatorial upwelling between 1°N and 10°S), loss of ¹⁵NO₃ ⁻ as DO¹⁵N accounted for <5% of net nitrate uptake. In samples from subtropical oligotrophic waters (from 11°S southward), however, ¹⁵NO₃ loss as DO¹⁵N represented up to 20% of net NO₃ ⁻ uptake. These results, as well as experimental considerations concerning the use of colorimetric and isotopic methods to measure new production show that: (1) earlier reported high discrepancies between nitrate decreases (ΔNO₃ ⁻) and ¹⁵NO₃ ⁻ incorporation into filterable particles (ΔNO₃ ⁻/¹⁵NO₃ ⁻ incorporation >2) were probably erroneous; (2) the use of GF/F filters does not result in an underestimation of new production, although it was found to underestimate PON concentrations by up to 60%; (3) in the equatorial upwelling area (1°N to 10°S), which has high ambient nitrate levels (>2000 nmol l⁻¹) but only slight changes in concentration (0 to 80 nmol l⁻¹ d⁻¹), new production is more accurately estimated by the isotopic method than by the chemical method; (4) in subtropical oligotrophic waters (from 11°S southward) with low ambient nitrate levels (0 to 100 nmol l⁻¹), both procedures are appropriate as long as nitrate removal per incubation period is >3 nmol l⁻¹ (lower rates are only detectable with the isotopic method); (5) the traditional ¹⁵N-tracer technique does not substantially underestimate net new-production in the equatorial Pacific, and failure to account for the loss of ¹⁵NO₃ ⁻ as DON, i.e. to estimate gross nitrate uptake (gross uptake = net uptake + ¹⁵N loss) tends to underestimate new production on an average by only 10%. Overall, the apparent low level of new production in the nitrate-rich area of the central equatorial Pacific seems to be a fact, and may be ascribable to other nutrient (macro and micro) deficiencies and/or to intense in situ recycling of ammonium and nitrate (regenerated production) rather than to inaccurate nitrate uptake rates measured with the classical ¹⁵N-tracer technique. Received: 24 November 1998 / Accepted 10 March 2000