Total Alkalinity:  Alkalinity was determined using an open cell titration with HCl (Dickson et al., 2007).  The pH meter was a Corning model 109 which had been adapted so that the millivolt out was logged by computer through a 14 bit A to D converter.  The electrode was an Orion Ross ultra semi-micro glass electrode model 8103-BNUWP.  The pH electrode was standardized with accurate pH buffers.  The tris buffer ( 2-amino-2-hydroxymethyl-1,3-propanediol) had a pH of about 8.09 depending on temperature.  The AMP buffer (2-aminopyridine) had a pH of about 6.79 depending on temperature.  Both were dissolved in artificial seawater at S = 35 (Dickson et al., 2007, SOP-6a).  Samples and standards were titrated with a 0.15 M HCl solution in 0.45 M NaCl and the temperature was measured to the nearest 0.01C using a NIST calibrated platinum thermometer.  The alkalinity standards generally were precise alkalinity/total carbon dioxide seawater standards from Scripps Institute of Oceanography (SIO), but early experiments also used a phosphate buffer standard comprised of an equal molar mixture of KH2PO4 and Na2HPO4 -7 H2O in 0.70 M NaCl.  This phosphate standard was cross calibrated with the SIO standards.  The procedure used generally gave the precision of several replicate standard titrations of 0.06% (standard error of the mean as percent of the mean value).  Alkalinities were determined by the fitting procedure described in Dickson et al., 2007 (SOP-3b).

Total Carbon Dioxide:  These concentrations were determined by acid stripping a 1.113 ml volume of water sample or TCO2 standard, trapping the expelled CO2, and then injecting it into a Shimadzu Model GC-8A gas chromatograph with a thermal conductivity detector (Christensen, 2008).  Two standards were employed, ones made from prebaked and freshly made Na2CO3, and the previously mentioned SIO total carbon dioxide seawater standards.  This analytical system obtained a precision of about 0.06% (standard error of the mean as percent of the mean).  However, in the results listed in this report, precision was less, averaging about 0.25% (standard error of the mean as percent of the mean) because sample analysis time was speeded up causing slightly less efficient trapping of the sample's CO2. 

Salinity and Nutrients:  Salinity was determined using an Autosal 8400A conductivity salinometer with IAPSO standard seawater standards.  Replicate determinations of a single sample were made until two consecutive readings of conductivity matched within +/- 0.002 ppt.  Nutrients were determined by autoanalyzer using the methods for nitrate and nitrite of Armstrong et al. (1967) and Pavlou (1972), for ammonium of Koroleff (1970) and Slawyk and MacIsaac (1972), for dissolved inorganic phosphate (Drummond and Maher, 1995), and dissolved silicate (Armstrong et al., 1967).  Concentrations were measured in mol L-1 and converted to mol kg-1 based on the sample's sigma-t value computed from the sample's salinity and the laboratory temperature during analysis.

Calculation of Carbonate System Parameters:  Carbonate system parameters, include total pH, were calculated from the measured chemistry of the water samples using the carbonate equilibrium model, CO2SYS (DOE, 1994; Lewis and Wallace, 1995).  This program employs the equilibrium coefficients of Roy et al. (1993) for carbonate coefficients, K1 and K2, of Weiss (1974) for carbon dioxide, K0, of Dickson (1990a) for borate, of Dickson and Riley (1979) for fluoride, of Dickson (1990b) for sulfate, and of Millero (1995) for phosphate (kp1, kp2, kp3) and silicate.  Seawater density at atmospheric pressure was that of UNESCO (1981).