IRON Iron concentrations were determined using a modification of the FIA system described by Obata et al. (1993). The most significant of these modifications was the use of a 10-port injection valve. The 10-port valve allows the sample to be preconcentrated onto an 8-hydroxyquinoline column (8-HQ, Landing et al., 1986) and then injected into the analytical stream without placing the 8-HQ column in the stream of reagents flowing directly to the detector. This avoids the pressure pulse and detector baseline shift that invariably occurs with a 6-port injection valve when reagent flow to the detector is diverted through the column to elute sample iron. A single 6-port stream selection valve was used to draw a strong acid wash (1.0 N double distilled HCl), sample (or standards and blank), a rinse with ultrapure MQ water (Millipore MilliQ, 18 M?/cm), and weak acid (0.3 N HCl for Fe) through the column. The column is eluted when the 6-port stream selection valve is switched to the position containing the same weak acid carrier solution that flows to the detector. The eluted sample stream then flows through the injection loop of the 10-port valve when it is in the load position. The 10-port injection valve is switched to the inject position when the slug of acid carrying the eluted metal fills the injection loop. The carrier stream flowing to the detector never passes through the column, as a result of this arrangement. Accurate timing was essential to ensure that all of the MQ wash was pushed out of the sample loop without loosing the sample slug. The valve switching and data collection was implemented with a Computer Boards, Inc. PCM DAS 16 PC card that was controlled by software (Microsoft Visual Basic) written in our laboratory. The reaction pH for the luminol and Fe mixture was adjusted by “fine tuning” the flow rate of the ammonium hydroxide solution by using a separate pump for this reagent. This allowed us to optimize the reaction pH for each new batch of reagents, which further minimized the remaining blank peaks that resulted from the slight dilution of the eluent acid in the sample loop any of the MQ wash solution that had remained in the loop. The 8-HQ columns for iron analyses were prepared by the method of Landing et al. (1986) except that a larger grained support material (TosoHaas, ToyoPearl HW-40C) was used to reduce backpressure and the tendency of the column to compact during extended sample load times. Preconcentration times varied depending on the time of year and sensitivity required. The flow manifolds were heated to a constant temperature of 30?C. All reagents were used as received. Any contamination was accounted for in the baseline, which remained steady as long as the composition of the flow was unchanged. The 10-port injection valve system ensured that the composition changes were minimal. Surface water was sampled at each station with a pumping system that consisted of an epoxy coated aluminum weight, polyethylene tubing and a ship-board, air-driven Teflon? diaphragm pump. The weight was lowered over the bow and was equipped with a vane to keep the attached tubing pointed into the current and away from the ship. This system delivered water into the ship’s laboratory at a rate of 5 L/min. Part of the sample stream was diverted into a class 100 laminar flow hood where samples were collected into acid washed LDPE bottles and refrigerated. All analyses were carried out within 24 hours of sample collection. The samples were acidified to pH 3.3 just prior to analysis (Obata et al., 1993). References Landing, W.M., C. Haraldsson, and N. Paxeus (1986) Vinyl polymer agglomerate based transition metal cation chelating ion-exchange resin containing the 8-hydoxyquinoline functional group. Analytical Chemistry, 28, 3031-3035. Obata, H., Karatani, H. & E. Nakayama. (1993) Automated determination of iron in seawater by chelating resin concentration and chemiluminescence detection. Analytical Chemistry, 5, 1524-1528. CLEAN WATER FOR PRODUCTIVITY Water for the productivity experiments was collected at seven fixed depths, representing 100, 50, 30, 15, 5,1 and 0.1% of the light penetration depths (LPD's), which were estimated by secchi disk. The type of sampling system and cleaning of components, as well as bottle handling and filtration, was modeled after the recommendations of FITZWATER et al. (1982). Measurements of chlorophyll and particulate carbon and nitrogen were made on samples collected in the upper 200m with the rosette sampler on the CTD. PRIMARY PRODUCTIVITY The radioactive isotope, 14C, was used to measure primary production. Samples were drawn into 280ml polycarbonate bottles which had been washed using the FITZWATER et al. (1982) technique for cleaning Go-Flo bottles. The bottles were then encased in nickel-cadmium screens (Perforated Products) that acted as neutral density filters to reduce the light intensity to the same level as that occurring at the depth from which the sample was collected. The screens were calibrated using a Biospherical QLS-100 to 100, 50, 30, 15, 5, 1, and 0.1% light levels. Approximately 10µCi of 14C were added to each sample bottle. An initial sample was inoculated with the tracer and filtered immediately, with no incubation, to determine abiotic particulate incorporation. The remaining samples were incubated for 24 hours in on-deck, seawater-cooled, Plexiglas incubators utilizing the natural sunlight as the light source. For determination of particulate carbon fixation, the samples were filtered onto Whatman GF/F filters at <200 mm mercury and the filters were soaked overnight with 0.5 N HCl to purge the filters of inorganic carbon isotope. The 14C filters were placed in 10 ml of Cytoscint ES scintillation cocktail and counted in a Beckman LS-3801 liquid scintillation counter. CHLOROPHYLL Chlorophyll a and phaeopigments were determined by the fluorometric technique using a Turner Designs Model 10-005 R fluorometer that was calibrated with commercial chlorophyll a (Sigma). Samples for determination of plant pigments were filtered onto 25-mm Whatman GF/F glass fiber filters and extracted in 90% acetone in a freezer for between 24 and 30 hours (Venrick and Hayward, 1984). Other than the modification of the extraction procedure, the method used is the conventional fluorometric procedure of Holm-Hansen et al. (1965) and Lorenzen (1966). Additional samples were also filtered onto 1.0 and 5.0 micron pore size Nuclepore membrane filters. PROTISTAN BIOMASS Phytoplankton and small heterotrophs were sized and counted with epifluorescence microscopy (Chavez et al. 1990, 1991). NUTRIENTS Nutrient samples were drawn, from the M1 station for all depths, into seasoned polyethylene scintillation vials and frozen aboard ship for later processing with an AlpChem autoanalyzer. Surface samples were collected at all other sites. The samples were analyzed for NO3, NO2, PO4 and SiO4 concentrations. REFERENCES Fitzwater, S. E., G. A. Knauer and J. H. Martin. 1982. Metal contamination and its effects on primary production. Limnology and Oceanography, 27: 544-551. Holm-Hansen, O., C. J. Lorenzen, R. W. Holmes and J. D. Strickland. 1965. Fluorometric determination of chlorophyll. Journal Cons. Perm. Int. Explor. Mer, 30: 3-15. Lorenzen, C. J. 1966. A method for the continuous measurement of in vivo chlorophyll concentration. Deep-Sea Research, 13: 223-227. Sakamoto, C. M., Friederich, G. E., Codispoti, L. A. 1995. MBARI Procedure for Automated Nutrient Analyses Using a Modified Alpkem Series 300 Rapid Flow Analyzer. MBARI Technical Report No. 90-2 Venrick, E. L., and T. L. Hayward. 1984. Determining chlorophyll on the 1984 CalCOFI surveys. California Coop. Oceanic Fish. Invest. Report 25:74-79.