9 Perspectives on DOC in the Carbon Cycle of the North Atlantic Ocean

Dennis A. Hansell

Bermuda Biological Station for Research

Ferry Reach GE-01, Bermuda

When studying the important processes controlling the size of the carbon sink in the North Atlantic we must include for consideration all forms in which the carbon can be sequestered and exported. The commonly constrained terms are inorganic carbon export, which in the North Atlantic can occur with deep water formation, and vertical export of particulate organic carbon resulting from new production. In this essay I evaluate possible mechanisms by which DOC can play a role to increase the magnitude of the sink now represented by POC and CO2 export alone.

The role of DOC needs to be evaluated for its potential contribution as a carbon sink on both local and basin scales. On the local scale it is necessary to first form a hypothesis as to how the concentrations of DOC vary seasonally as a function of latitude. Hypothesis is required because few reliable data for DOC concentrations exist - but by considering the physical nature of the systems rigorous hypotheses are possible.

The first hypothesis is that the potential for DOC export will be a direct function of the turnover time of the DOC pool. That is, the export potential of long lived DOC will be high, and directly related to the degree of vertical advection and mixing in a particular region. Short lived compounds comprising DOC, on the other hand, are likely mineralized in the euphotic zone. These compounds would not contribute significantly to export flux.

In permanently stratified, low latitude water masses, DOC in the upper mixed layer might be viewed as a temporary reservoir for carbon, the size of which is a few ten's of microM above deep water values for DOC. Vertical export of the carbon from surface waters as DOC is likely to occur only slowly by diffusional exchange with deeper water. On the other hand, where there is seasonal stratification at mid-latitudes, then the bulk of the long lived DOC that accumulated during stratification can be mixed to depth rapidly (and exported) during the winter mixing period. This period of winter mixing would represent the strongest pulse of export of DOC on an annual basis. Such a phenomenon is represented by the export that may occur at the site of the Bermuda Atlantic Time-series Study (BATS). At the BATS site, stratification of the surface water develops with warming through the spring months and stability remains until November or December. During this period of stability DOC that is produced in the euphotic zone is to a large extent mineralized rapidly (D. Hansell, unpubl. data). A portion of the DOC produced is refractory, however, giving rise to an accumulation of DOC to concentrations a few ten's of microM C above those found during maximum deep mixing (C. Carlson, unpubl. data). It is this accumulated carbon that can be exported during the next period of destratification.

The length of the period of stratification decreases with increasing latitude. At low latitudes the surface waters can be viewed as permanently stratified while at high latitudes periods of stratification may last only a few days or weeks at a time. What then is the potential for export of DOC in the high latitude systems? We know that the higher latitude systems are characterized by strong spring blooms and that there can be a large accumulation of highly labile DOC in the surface waters (Kirchman et al., 1991). We do not know if the refractory DOC that likely builds up during these periods is of a large enough magnitude to greatly contribute to the export flux dominated by sinking particles in these blooms. This is a question that needs to be answered.

Another interesting hypothesis that relates latitude to DOC export is as follows: DOC that is exported in high latitude systems, and which is generated during blooms, is a product of primary production driven by nitrate uptake (new production) while DOC accumulating at lower latitudes, as described above, is largely the product of primary production driven by regenerated production. At the BATS site there is very little new production and only low nanomolar levels of nitrate (Lipschultz, pers. comm.), yet there appears to be an accumulation of DOC after the spring stabilization of the water column. So in low latitude systems the export realized via downward mixing of DOC may be based on regenerated production. This concept may not fit current modeling efforts to quantify carbon export, particularly those based on nitrogen availability. Resolution of this question is required.

A final scenario for DOC export, one that must be viewed on a basin scale, is as follows: Large quantities of DOC produced in the surface waters at low latitudes in the North Atlantic are transported north with the Gulf Stream. Once at high latitude, North Atlantic deep water formation during winter transports the particularly refractory DOC to depth. This scenario requires a portion of the Gulf Stream to transit into the Norwegian Sea for eventual deep water formation. The strongest support for the hypothesis of DOC export with deep water formation is the difference in DOC concentrations in the deep water of the North Atlantic and in deep water of the equatorial Pacific. The concentration of DOC at >2000 m at the BATS site and elsewhere in the North Atlantic is 42-48 microM (Sharp, 1993; D. Hansell, unpublished data; Carlson and Ducklow, 1994; Peltzer, 1994) while in the equatorial Pacific the concentration is 38+/- 2 microM (Peltzer, 1994). During transit of the deep water from the North Atlantic to the Pacific there may then be a 4-10 microM reduction in DOC concentration. This reduction suggests that DOC is exported with deep water formation in the North Atlantic and mineralized slowly at depth. The source of DOC may be local (the Norwegian Sea) rather than the Gulf Stream but a local source during the winter period of deep water formation is difficult to reconcile.

Dissolved organic carbon likely plays a role in exporting carbon from surface waters in the North Atlantic Ocean. We do not at this time know the important mechanisms nor the quantities involved but an assessment of each in a JGOFS North Atlantic process study is vital. A few plausible mechanisms have been listed here but they should not be viewed as exhaustive.