Ocean metabolism observed with oxygen sensors on profiling floats in the South Pacific

We estimated rates of production and export in the South Pacific (80uW to 180uW in a zonal band between 35uS and 50uS) using 1.5 yr of oxygen measurements from profiling floats. Export production, calculated from oxygen utilization rates below the compensation depth from December to April, was 10.7 6 2 mmol C m22 d21 (n 5 36, 95% CI). The corresponding satellite net primary production was 46 6 4 mmol C m22 d21, yielding a regional e-ratio of 0.23 6 0.05. Averaging oxygen utilization rates resulted in a net cancellation of most water mass changes related to advection and float migration. The composite vertical profile of remineralization rates, obtained by binning 36 rate profiles, agreed with published measurements based on oxygen utilization rates in hydrographic surveys and fits the classic form of a particulate organic carbon (POC) attenuation function. However, the disagreement between oxygen-based remineralization rates and those obtained by sediment traps suggests fundamental differences between these two methods. Using float data to constrain a one-dimensional mixed-layer model, the annual net community production at 45uS, 144uW was ,2.5 mol C m22 yr21. Spatial trends in export production coinciding with the New Zealand shelf and Subtropical Front are identified. The quantification of basin-scale magnitudes and trends in net primary production (NPP), net community production (NCP), and export production (EP) are required for a first-order estimate of biologically driven transport of carbon to depth (i.e., the strength of the ‘biological pump’; Sarmiento and Gruber 2006). A great deal of work has focused on the links between depth-integrated rates of NPP (NPP 5 gross carbon fixation 2 autotrophic respiration), NCP (NCP 5 NPP 2 heterotrophic respiration), and EP (EP 5 vertical organic carbon flux). On the annual mesoscale the ocean is near steady-state and NCP , EP (Brix et al. 2006). The e-ratio, EP : NPP, is a fundamental term in the carbon budget because it determines what fraction of NPP is sequestered by the ocean interior for decades or longer. A primary metric of the validity of ocean ecosystem models is their ability to accurately depict export flux or e-ratios (Lima and Doney 2004). Presently, estimates based on remote sensing provide near-comprehensive coverage of ocean NPP on large geographic scales (Behrenfeld and Falkowski 1997; Behrenfeld et al. 2005). Satellites constrain global ocean NPP to ,610% (Falkowski et al. 1998) and local NPP to no better than 635% (Hooker and McClain 2000). Export flux and remineralization rates have been measured throughout the ocean using a variety of techniques, including sediment traps, oxygen utilization rates, and 234Th disequilibrium. Because measurements of NCP and EP are typically made at point locations on subannual timescales (often subdaily for NCP), episodic, seasonal, or annual decoupling of NCP from EP may result in a break from the steady-state paradigm at the sampling location (Plattner et al. 2005). Locally, NCP and EP are measured with accuracy of 650% (Emerson et al. 1997). There is, however, no direct estimate of EP or NCP with spatial and temporal coverage complementary to satellite-based NPP. Dunne et al. (2005) compiled 122 e-ratio observations spanning the past 30 yr and derived empirical relationships for EP : NPP as a function of both chlorophyll a (Chl a)– temperature and NPP–temperature. Limited agreement between these relationships and e-ratios obtained through biogeochemical Ocean General Circulation Models (OGCMs; Gehlen et al. 2006) underscores both the sparseness of data and the simplifications which are inherent, but necessary, in models. Presumably, inverse techniques provide the most reliable estimates of global EP, but there is not necessarily a consensus between models constrained by satellite vs. those constrained by hydrographic measurements. For example, Schlitzer (2002) showed that EP estimated through an inversion of the World Ocean Circulation Experiment (WOCE) data set only moderately agreed with EP calculated by transforming satellite NPP into EP using an empirical (temperatureand NPP-dependent) e-ratio (e.g., Laws et al. 2000). Clearly, modeling efforts will become more reliable as the density of EP : NPP measurements increases. Because photosynthesis is restricted to the uppermost strata of the ocean, NCP has often been inferred from changes in the inventory of nutrients, carbon dioxide (CO2) species, or oxygen levels from the surface to the base of the Acknowledgments We thank two anonymous reviewers for insightful comments that greatly improved the manuscript. Work at MBARI was supported by a grant from the David and Lucile Packard Foundation and by the National Science Foundation through the Biocomplexity in the Environment Program Grant ECS0308070. Research at the University of Washington was supported by the National Oceanic and Atmospheric Administration Grant NA17RJ1232 Task 2 and by Office of Naval Research Contract ONR NOOO14-03-1-0446. Limnol. Oceanogr., 53(5, part 2), 2008, 2094–2111 E 2008, by the American Society of Limnology and Oceanography, Inc.