Pelagic ecosystems are driven by the interaction of organisms that fall into four functional components: the autotrophic phytoplankton and the heterotrophic bacteria, protozoa and metazoa. The systematic study of the protozooplankton has traditionally lagged behind that of the other components because of their great heterogeneity, despite recognition of their important role in structuring pelgic ecosystems. The organisms representing this component are phylogenetically and functionally more diverse than either of the other components. Their study requires specialisation and dedication to specific size classes, taxonomic groupings or mineral-bearing taxa with the result that the component as a whole tends not to be regarded on an equal footing with the others of the plankton. Yet, despite their heterogeneity they share common features that identify them as a distinct component: they tend to be of the same size class as their prey and, being unicellular and heterotrophic, can potentially have higher growth rates than the phytoplankton.

It is now well established that grazing pressure of the hetrotrophic nanoflagellates (HNF) restricts the biomass of bacteria but also the pico- and nanophytoplankton of the microbial network and prevent them from reaching bloom proportions in the presence of adequate resources. So only larger-celled phytoplankton, in particular diatoms and the colonial Phaeocystis contribute to the blooms that fuel the higher trophic levels of pelagic food chains and drive the biogeochemical cycles of elements. But this is a paradox because there are various large-celled protozoa capable of feeding on bloom-forming phytoplankton using various feeding modes. Why are these organisms not capable of checking biomass build-up as are their smaller counterparts the HNF? One explanation supported by field observations is that biomass of this component is kept in check by the selective grazing pressure of metazooplankton, in particular copepods, which feed preferentially on protozoa rather than phytoplankton. This begs the question why protozoa are more palatable than bloom-forming phytoplanton species. Another additional explanation is that the bloom-forming phytoplankton species have developed adequate defences against many of the protozoan grazers implying that the evolutionary arms race is an important factor driving speciation. This would mean that co protozoa and their prey have co-evolved implying that phylogenetic groups comprising the protozooplankton are selective rather than generalist feeders. In the former explanation biomass is controlled by proximate factors and in the latter ultimate, or evolutionary factors pertaining to species-specific adaptations are at work.

These explanations are obviously not mutually exclusive but determining their relative roles can only be carried out in the field. Our approach to studying the complex relationships underlying the structure of pelagic food webs with particular reference to the protozoan component is to follow the populations of the various dominant species of protozoa in relation to that of their potential prey and their predators under experimental conditions. In iron fertilization experiments, the pelagic system is relieved from its limiting resource, which in the Southern Ocean is iron. Following the responses of the various groups and their interrelationships enables keeping track of proximate factors controlling population sizes of the dominant species and sheds light on possible intrinsic controls.

The major groups under study are dinoflagellates and ciliates of which several species are cosmopolitan. The former are able to graze on a large variety of different size classes depending on feeding behaviour.

The thecate (armoured) dinoflagellates of the cosmopolitan genera Protoperidinium, Oblea and Dinophysis can feed on prey much larger than themselves by either extruding a feeding membrane that can engulf entire diatom chains (pallium feeding) or by using an umbilical cord-like pseudopod (peduncle) to pierce the prey (peduncle feeding). The latter is apparently deterred by the diatom frustule and colony skin of Phaeocystis. Athecate dinoflagellates (e.g. Gyrodinium) can engulf prey of similar size meaning that chains and colonies are out of their "reach". Unarmoured ciliates of the cosmopolitan genera Strobilidium, Strombidium, Laboea and Tontonia ingest prey smaller than themselves and are deterred by bristles and spines. Tintinnid ciliates are even more restricted by the aperture of their lorica (a shell-like structure in which they live).

Apart from these dominant protozoan groups attention is also focussed on large sarcodines (e.g. foraminifera, radiolaria, acantharia and heliozoa) that are less abundant but, in contrast to most other plankton species. preserve well in the fossil record. Knowledge of the ecological predilections of these species can provide valuable information for paleoeanographic reconstruction.