Report 18 -- Q3

What controls primary production in the Southern Ocean?

Low phytoplankton biomass and primary productivity are widespread in the waters south of the APFZ in spite of persistently high nutrient concentrations. Thus the Southern Ocean is the largest of the three well documented high-nutrient/low chlorophyll (HNLC) regions of the ocean (the other two being the subarctic and equatorial Pacific; e.g., Cullen, 1991). As is the case in other HNLC areas, hypotheses that seek to explain low productivity in the Southern Ocean center around the irradiance/mixing regime (Mitchell et al., 1991; Nelson and Smith, 1991), iron limitation (Martin et al., 1990) and grazing (e.g., Miller et al., 1991; Frost, 1991). It is unlikely that any single hypothesis can explain the HNLC condition throughout the Southern Ocean because of the spatial heterogeneity and strongly seasonal character of many of the processes (irradiance, vertical mixing, atmospheric dust inputs, formation and melting of pack ice, krill swarms etc.) Moreover, the hypotheses that have been put forward are not mutually exclusive; it is likely for example, that phytoplankton blooms develop in marginal ice zones in response to the combined effects of water-column stabilization and iron enrichment (Martin et al., 1990; Nelson and Smith, 1991). Similar conditions may enhance productivity at the APFZ and other open-water frontal systems, although the direct observational evidence from those systems is less developed at this point.

It is hoped that field studies during AESOPS can achieve a reasonably mechanistic understanding of how primary production is controlled in the Southern Ocean. If that effort is successful the biological understanding obtained can provide a basis for modeling the predicted changes in the ecosystem in response to any climate-change scenario, and can indicate which properties of the system can most usefully be monitored from satellites and moored instruments (e.g., mixed-layer depths, light attenuation coefficients, atmospheric dust inputs) to track the control of primary productivity through time.

Implications for field measurements: Field experiments should be directed toward testing specific control hypotheses in several subsystems of the Southern Ocean (e.g., the APFZ, the low-productivity open waters to the south, the coastal waters surrounding Antarctica). A program-level effort should be made to coordinate proposed studies of individual control mechanisms in a way that permits competing and/or complementary hypotheses to be tested in the same areas of the ocean at the same times. It is likely that much of this coordination will have to take place after the individual hypothesis-testing proposals have been submitted, as investigators will normally be unable to know ahead of time what other studies will be proposed by other groups. As is the case for all biological studies in the Southern Ocean, studies of the control of primary production should address the dominant seasonal time scale, and should specifically address the question of whether there are major temporal changes in the controlling mechanisms during the spring and summer.

Implications for modeling: Primary production models, and primary production components of larger models, should be based upon control mechanisms that can be demonstrated at sea. Nutrient-based models, for example, are inapplicable to the Southern Ocean because of the almost universally nutrient-rich conditions that prevail. Models should also be comprehensive enough to reflect the fact that high-productivity events do occur in the Southern Ocean, and that they cannot occur unless both the physical environment (e.g., light and vertical mixing) and the chemical environment (e.g., Fe and other micronutrients) are favorable. Models invoking grazer control should take into account the fact that some bloom-forming phytoplankton groups (e.g., diatoms) are readily grazed by particle-selective herbivores and other bloom-forming groups (e.g., Phaeocystis) are assiduously avoided by many grazers.

It would also be helpful if algorithms can be developed by which those properties that are shown to be important in controlling primary production can be estimated from remote sensing data (e.g., mixed-layer depths from local wind-stress, solar irradiance from cloud cover, light attenuation coefficients from ocean color, atmospheric dust inputs from large-scale wind patterns). Such algorithms could permit satellite data to be used in monitoring processes known or believed to be of major importance in controlling primary production.