Abstract
The process of nitrogen fixation in the subtropical North Atlantic has received considerable study over the last few decades. The findings have highlighted a large discrepancy in estimates for the locations and rates of nitrogen fixation when results from biological techniques are compared to geochemical techniques. Here, we evaluated the distribution and rates of excess nitrate development in the North Atlantic using World Ocean Circulation Experiment (WOCE) nutrient data. These data indicate that excess nitrate development is largely confined to depths of ∼150–400 m in the region of 15–25°N by 25–75°W, an area considerably smaller than that employed by Gruber and Sarmiento [Glob. Biogeochem. Cycles 11 (1997) 235] (10–50°N by 10–90°W) to estimate rates of nitrogen fixation in the North Atlantic. The areally integrated nitrogen fixation rate for the subtropical North Atlantic was 0.045 mol N m
−2 year
−1, or 62% of the geochemical estimate by Gruber and Sarmiento [Glob. Biogeochem. Cycles 11 (1997) 235]. The regional rate of fixation was 3.1×10
11 mol N year
−1 (1.5–4.6×10
11 mol N year
−1 given a 50% uncertainty), similar to rates expected from biological measures of fixation, but only 15% of the areal rate estimated by Gruber and Sarmiento [Glob. Biogeochem. Cycles 11 (1997) 235]. An accurate assessment of the region over which excess nitrate accumulates is critical to the estimate of nitrogen fixation, but remains prone to large uncertainty because of the gaps in spatial coverage. Additional survey work in the North Atlantic must be done to lessen the uncertainty. With this work, we reduce the differences between the biological and geochemical rate estimates, and describe a conceptual model for the location and dynamics of nitrate excess development in the North Atlantic.
