The ability to model ''new'' (nitrate-based) production is crucial to predicting the ocean's role in the global carbon cycle. The ability to model the distribution of high-nutrient/low-chlorophyll (HNLC) areas is particularly important in this regard. Here I draw together three elements that appear to be necessary for constructing the required model: (1) iron limitation of algal growth rates as an ultimate cause of the HNLC condition; (2) ammonium inhibition of nitrate uptake and utilization as a proximate mechanism that leads to reduced nitrate use; and (3) the dependence of both processes on algal cell size. In the model, cells are postulated to maximize their growth rates by partitioning scarce iron between nitrogen-and carbon-related demands. The effect of iron limitation is postulated to depend on cell size through surface/volume effects on uptake efficiency; this dependence on cell size in turn affects phytoplankton community structure and community-level uptake of nitrate. The efficacy of the model is demonstrated by its ability to reproduce community-level curves of nitrate uptake versus ammonium concentration from both HNLC and non-HNLC areas. The partition formalism can be incorporated directly into ecosystem models; when implemented in an ecosystem model with multiple size classes, the model should produce HNLC versus non-HNLC conditions in appropriate locations and for appropriate reasons. In particular, the model's ability to produce iron-light and iron-light-nitrogen colimitation should be useful in understanding and predicting the HNLC condition in parts of the Southern Ocean. With suitable changes to parameter values, the postulated mechanism for iron partitioning may also prove useful in modeling iron and energy partitioning in nitrogen fixers. The ability to model the distribution of high-nutrient/lowchlorophyll (HNLC) areas, where concentrations of nitrate and phosphate remain high throughout the year, is critical to prognostic studies of the global carbon cycle. Empirical evidence suggests that the ability to model iron limitation of growth and nitrate uptake will be a key element in predicting HNLC conditions. As the prime example, a scarcity of iron has recently been shown to limit phytoplankton growth, biomass accumulation, and size and taxonomic structure in the equatorial Pacific, a major HNLC region (Coale et al. 1996). Addition of iron to surface waters of this region caused a massive phytoplankton bloom and drawdown of nitrate, accompanied by a marked reduction in the fugacity of CO 2 (Cooper et al. 1996). The possibility of similar conversion of the subarctic Pacific (La Roche et al. 1996) and the Southern Ocean (de Baar et al. 1995;Kumar et al. 1995) from HNLC to non-HNLC by the addition of iron is supported by more indirect evidence.
AcknowledgmentsConversations with Richard Geider helped greatly in the development of these ideas, particularly in the realization that it could be extremely useful scientifically to have a single model of iron limitation that both biogeochem...