Iron, nitrogen and phosphorus in the ocean

Abstract

It has been proposed that widespread deficits of nitrate in the ocean, like those observed today, are caused by iron limitation of marine nitrogen fixation¹. That is, only when iron is sufficiently abundant to satiate nitrogen fixers will the ratio of nitrate to phosphate in the ocean increase to 16, the average for phytoplankton. Tyrrell² developed a simple two-box model of oceanic nitrogen and phosphorus cycles to describe the regulation of both nitrate and phosphate concentrations in the global ocean. His criterion for nitrate deficit in the ocean, a molar ratio of N:P in surface waters (R_s) of less than 16, is satisfied without recourse to iron limitation, calling into question Falkowski's proposal¹ about the biogeochemical significance of iron limitation as it relates to nitrogen fixation and oceanic levels of nitrogen and phosphorus. Here I show that small changes in the assumptions of Tyrrell's model, well within acknowledged uncertainty, can lead to values of R_s greater than 16. Consequently, the consistency of the model with the observed distributions of nutrients in the ocean is uncertain, and the influence of iron may still be considered important.

Main

In Tyrrell's model², competition between nitrogen-fixing and other phytoplankton controls the level of nitrate, continually pushing molar concentrations to slightly less than 16 times those of phosphate. Results of the model show that phosphate is the ultimate limiting nutrient because extra phosphate in the system supports the proliferation of nitrogen fixers that can add new nitrogen to the ocean. Even when subjected to extensive sensitivity analysis², the model consistently predicts a deficit of nitrate in the surface layer, defined by Tyrrell as R_s<16.

Further calculations (Fig. 1) indicate that the model's prediction of a nitrate deficit in surface waters of the ocean is uncertain. R_s is very sensitive to chosen values for the Michaelis–Menten half-saturation constants for the growth of phytoplankton on nitrate (K_S[NO₃]) and phosphate (K_S[PO₄]) (N_H and P_H, respectively, in Tyrrell's notation). A 20% increase of K_S[NO₃] to 0.6 mmol m⁻³ from the assumed 0.5 mmol m⁻³ obliterates the predicted nitrate deficit, bringing R_S to 16. The reported² uncertainty in K_S[NO₃] is 0.1 to 4.2 mmol m⁻³ (ref. 3), corresponding to an R_S value between 2.7 and 112 mol mol⁻¹.

Figure 1: Influence of assumed Michaelis–Menten half-saturation constants and phytoplankton growth rates on steady-state solutions for R_S, the molar ratio of nitrate to phosphate in the surface layer of the ocean (model of ref. 2).

Nitrate deficit at the surface is indicated by R_S<16 (shaded). The half-saturation constant for growth versus nitrate, K_S[NO₃], was varied from the specified 0.5 mmol N m⁻³ to obtain a range of K_S[NO₃]:K_S[PO₄], keeping K_S[PO₄] constant at the specified 0.03 mmol P m⁻³. The solution for R_S against K_S[NO₃]:K_S[PO₄] is independent of K_S[PO₄] (equation 13 in ref. 2). The maximum growth rate of nitrogen-fixing phytoplankton, μ′_NF, was maintained at 0.24 d⁻¹, as specified in the model. The maximum growth rate of other phytoplankton, μ′_O, was varied from the original 0.25 d⁻¹, as indicated. The original model solution is identified with the open circle.

Small changes in the maximum growth rate for other phytoplankton, μ′_O (d⁻¹), compared with 0.24 d⁻¹ for nitrogen fixers, also strongly influence R_S (Fig. 1). There is little experimental basis for excluding assumed growth rates that lead to an R_S value of 16.

This simple analysis, based directly on Tyrrell's model, suggests that regulation of oceanic nitrogen fixation by iron cannot be excluded as a potentially important influence on cycles of nutrients and primary productivity in the ocean.

Reply - Toby Tyrrell