A physical-biological model was developed for the equatorial Pacific Ocean. It was used to investigate the physical and biological causes for high nutrient condition in the equatorial Pacific. At 0 degree 140 degrees West, nitrate fluxes due to the mixing were small compared to the advective fluxes; vertical transport by the equatorial upwelling and zonal transport in the equatorial undercurrent were the major fluxes. In the equatorial cold tongue, the net physical flux of nitrate to the euphotic layer in carbon (C) units was 15.09 millimole C per square meter per day.

Model experiments showed that iron limitation of phytoplankton is not necessary for maintenance of a high-nutrient plume; that is, the plume of high nitrate water can be generated solely by physics. However, model results indicated that iron limitation determined the concentration level of nitrate in the nutrient-rich plume and created north-south asymmetry in the nutrient fields. Model results also suggested that if the equatorial Pacific Ocean were micronutrient replete, nitrate concentration would be reduced by half from its present value. The zooplankton grazing hypothesis was tested by reducing or enhancing the maximum zooplankton grazing rate. Model results suggested that the ratio of the maximum zooplankton grazing rate to the maximum phytoplankton growth rate should be between 0.5 and 0.75.

Interannual variations in the equatorial Pacific Ocean were simulated. In the model as in nature strong trade winds create the west-east asymmetry of the upper ocean heat and nitrate content which set up the salient zonal characteristics of the basinwide ecosystem in the equatorial Pacific Ocean. A strong El Nino event completely eliminated the basinwide west-east asymmetry, while a La Nina reinforced such asymmetry. The modeling results support the interpretation of Barber (1988) concerning the role of zonal asymmetry in determining the equatorial productivity gradient. The shift-up phenomenon described by MacIsaac et al. (1985) may constrain nitrate concentrations during the overshoot period, which would explain why the model without "shifted-up" rates predicts nitrate concentrations higher than the observed concentrations.