, 2007). The depth of this barrier layer may also vary with evaporation and precipitation changes. The presence of the barrier layer in the WPWP inhibits the mixing of TCO2 rich

waters into the surface mixed layer and leads to only a small seasonal range in TCO2 (Feely et al., 2002 and Ishii et al., 2009). Outside the WPWP and the NECC regions, barrier layers are rarely detected (De Boyer Montégut et al., 2007) and deeper mixing could result in a greater seasonal change in TCO2. Our results show that surface NTA variations are small in time and space for the Pacific study area (NTA = 2300 ± 6 μmol kg− 1; Fig. 4). This implies PD-1 inhibitor that the residence time of surface waters in the region is small enough for net CaCO3 production in reefs and pelagic waters to only have a small effect on TA variability at regional scales. The TCO2 change generated by net CaCO3 production can be estimated from half the normalized alkalinity and nitrate PLX3397 price (NNO3) change

(Chen, 1978) such that ΔNTCO2(CaCO3) = − 0.5 × (ΔNTA + ΔNNO3). The annual mean NO3 concentration along the equator increases eastward from 1 to 5 μmol kg− 1 and the rest of the region has an annual mean of 0.25 μmol kg− 1 (Garcia et al., 2010). Hence, the annual maximum estimated ΔNTCO2(CaCO3) is 2.5 μmol kg− 1. Based on this analysis, net calcification does not appear to have a significant impact on the large seasonal or regional changes in TCO2. However, localized calcification and production could influence TCO2 and TA variability at the scale of coral reefs (Shaw et al., 2012). The averaged aragonite saturation state, Ωar, for the Pacific region is 3.8 (Fig. 6). Values of Ωar below the mean occur in the subtropical waters at the northern and southern boundaries of the study area, and in the equatorial Pacific and North Pacific to the east of 180°E (NECC and CEP). Values above 3.8 occur in the WPWP, SECC, and SEC waters between about 5°S and

25°S that are away from the influence of the equatorial upwelling in the CEP. Feely et al. (2012) calculated the aragonite saturation states using TCO2 and TA measured Gefitinib cost on repeat hydrography sections, P06W 2003 and P16N 2006, which are within our study area. Using a 0.01/yr decrease in the aragonite saturation state (Feely et al., 2012), we can compare saturation states of these sections with the year 2000 mean values of Ωar. For example, along 160°W, surface Ωar during P16N 2006 was 3.4 ± 0.4. At a rate of − 0.01/yr, Ωar would have been 3.5 ± 0.4 in 2000. This calculated value agrees with our 2000 Ωar value of 3.8 ± 0.2 within the errors of the calculations. Similarly, along 30°S, surface Ωar during P06W 2003 was 3.2 ± 0.2 and would have been about the same value in 2000, agreeing with our 2000 Ωar value of 3.7 ± 0.3.