The background to change
The flagella by which "flagellates" move or feed could be seen easily by the early light microscopists. It was inevitable that organisms bearing this organelle would be groups together. This is evident from the first comprehensive attempts at systematic classification of the protists (Butschli 1880-1889, Copeland 1956). That the "flagellates" should have survived so long after the advent of ultrastructure insights is not. Electron microscopy has a higher resolution than light microscopy and so is more informative. Information not previously available (on flagellar systems, cytoskeletal structures, nuclear division patterns, appearance of mitocondria, chloroplasts, the endomembrane system, and so on) have acted as test of the high level taxa bequeathed by light microscopists. Groups with more than one pattern of ultrastructural organization are deemed to be polyphlyetic (Smith and Patterson 1986). Usually, these conclusions have been corroborated by molecular studies (Patterson and Sogin 1992, Sogin 1991). Taxa that have failed to satisfy requirements to be monophyletic and ideally holophyletic (Ashlock 1971) have been replaced. Continuing ultrastructural studies are providing revised circumscriptions of groups. The descriptions are largely refractory to the acquisition of new information. The resilience of the new concepts to change is taken as evidence of the monophyly of these groups. Ultrastructural studies are effective descriptors of monophyletic lineages becauses protists evolved by invention and modification of organelles. The electron-microscope is so far the only tool which can provide access to the differing complements and patterns of ultrastructural organization. Groups described with substantial reference to ultrastructural features have an "ultrastructural identity" (Patterson and Brugerolle 1988). Closely related organisms (such as all ciliates or all cryptomonads or all Apicomplexa) share a common ultrastructural identity. This concept has an ecological dimension. Each distinctive complement of organelles provides a group with a restricted range of adaptive competences. As a result, species sharing a common ultrastructual identity have ecological characteristics in common. Although an extreme generalization, this concept has some value (all collar-flagellates occupy largely similar niches which differ from those of all pelobionts which differ from that of all kathablepharids). This generalization needs to be amended in the case of lineages in which chloroplasts were acquired (dinoflagellates, euglenids. etc.). The flagellates include groups with plastidic species (euglenids, dinoflagellates, green algae, prasinophytes, cryptomonads, "heterokonts," haptophytes) and groups containing only heterotrophic species. There has long been an understanding that the algal (plastidic) flagellates include organisms with differing ultrastructural identities (Cox 1980, Dodge 1973). Although less extensively reviewed, the ultrastructural identities among the heterotrophic flagellates are very diverse (Patterson and Larsen 1991, Taylor 1976) alerting us to the possibility that they may form a polyphyletic group.
Incompatibility of traditional and phylogenetic concepts of flagellates
If defined phylogenetically, the "flagellates" would be "those eukaryotes with flagella and all taxa derived from them" (flagellum being interpreted as the synapomorphy of a monophyletic and holophyletic group). Flagella are motility organelles with a peripheral cylinder of paired microtubules surrounding a core of single microtubules plus organelles which evolved from them. The term embraces the "n + n" structures encountered in pelobionts (Brugerolle 1991, Griffin 1988) because the arrangment of a cylinder of doublet microtubules around a core of single microtubules is held to be homologous with the "9 + 2" structures in most eukaryotes. The "n + n" structures in pelobionts are held to be ancestral. Flagella evolved within the primitively amitochondriate Hypochondria (Patterson and Sogin 1992). Only the primitively amitochondriate Microspora may have had origins before the appearance of flagella. The phylogenetic concept of "flagellates" corresponds closely to "all eukaryotes excepting Microspora". A definition based on an evolutionary innovation is precise. The composition of the group defined in relation to the origins of flagella is very much at variance with traditional concept of flagellates. There are about 100 well-described types of eukaryotes (Patterson and Sogin 1992). Although virtually all fall into the domain of phylogenetically defined "flagellates", only 40 would satisfy the traditional concepts of flagellates. The traditional concept could be rendered phylogenetically meaningful by specific reference to the plants, animals, fungi, slime moulds, amoebae, ciliates, algae, and sporozoa which fall into the phylogenetic concept of "flagellates" but lie outside the traditional concept. This is impracticable, and the phylogenetic and traditional concepts should be considered irreconciable. The word flagellate should not be used for both the traditional and phylogenetic groups, as this will confuse (Patterson & Larsen 1992). Flagellate is here used only in traditional sense.
Reclassifying heterotrophic flagellates
The free-living heterotrophic flagellates embrace about 45 well-characterized types of organisms, each with its own ultrastructural identity (Table 1A). Over 60 known genera of free-living flagellates have yet to be studied by ultrastructural techniques (Table 1B).
The flagellates as ecological entity
Flagellates play a significant role in aquatic ecosystems (Azam et al. 1983, Patterson and Larsen 1991). The position assigned to flagellates is usually that of organisms between 2 and 20 um and consuming suspended bacteria. This is simplistic because it leads to the view that all flagellates are ecologically similar. In this concept is simplistic, then models of how ecosystems operate and which include this concept of flagellates are also simplistic. Some ecologists (e.g. Mitchell et al. 1988, Sherr and Sherr 1991) have tried to understand the ecological significance of the diversity of flagellates. However, others (Fenchel 1986, 1991 inter alia) have developed arguments which suggest that the traditional concepts of "flagellate" needs little elaboration to be ecologically meaningful. The argument that the "traditional flagellates" have a common ecological identity derives from the assumption that they have few flagella, and that they use the flagella for propelling water relative to the cell body for feeding or for cell locomotion. Cells with one or a few flagella are adaptively constrained by the characteristics and limitations of the flagellar axoneme. The stiffness of microtubules determines bending patterns and the properties of dynein determine the rate of intermicrotubule sliding and the amount of power which can be generated. Hydrodynamic considerations lead to the conclusion that flagella can create only a limited amount of thrust (Fenchel 1991). This indicates an upper size limit of about 50 um for flagellates. A lower size limit of 2 or 3 um is introduced by the need to fit a basic complement of eukaryotic organelles into one cell. Similar considerations dictate swimming speed, the volume of water from which food particles can be removed, the minimum concentration of food particles which can support the growth of flagellates, and the growth rate of the flagellates.