Tintinnoinea
Taxonomy

Taxonomic studies of South Atlantic tintinnids include a few that covered large areas (Brandt, 1906-1907; Laackmann, 1910; Souto, 1979, 1981; Alder, 1995), and several focusing on geographically restricted, usually coastal, areas (Da Cunha and Da Fonseca, 1917; Faria and Da Cunha, 1917; Balech 1942, 1944, 1945, 1947, 1948, 1951, 1958a, b, 1962a, 1971a, 1973, 1976; Souto, 1970a, b, 1972, 1974; Barría de Cao, 1981, 1986, 1992). Brandt (1906-1907) and Souto (1979, 1981) provided comprehensive information on tintinnid species in tropical and temperate waters, while studies by Laackmann (1906, 1907, 1910) and Alder (1995) dealt mainly with Antarctic species. Laackmann (1910) also described the fauna off South Africa from the "Gauss" expedition. Finally, Kofoid and Campbell (1929, 1939) described tintinnids from the eastern Pacific Ocean (North Temperate, Tropical, and South Temperate Pacific, from the Arctic Circle to the Tropic of Capricorn), re-arranging all genera, species, forms, and varieties chiefly from Brandt's (1906, 1907) and Laackmann's (1910) monographs. Kofoid and Campbell also introduced the concept of "series" within some genera in order to group species of similar morphology, and indiscriminately erected hundreds of new species on the basis of previous subspecific categories. Besides Kofoid and Campbell's (1929, 1939), the only other comprehensive studies of tintinnids are those by Daday (1887b) and Brandt (1906, 1907).

Two very different criteria are currently applied to the systematics of Tintinnoinea. One of them uses the ratio of lorica total length to oral diameter as the chief taxonomic character, and assigns the same diagnostic value to all other size parameters and features of the lorica. This was Kofoid and Campbell's (1929, 1939) basis for describing over 1000 species. As opposed to this rather simplistic approach, Brandt (1906-1907), Laackmann (1910), Alder (1995), among others, stressed the high intraspecific polymorphism of tintinnids, recognizing far fewer species and proposing the use of various infraspecific categories, including "Formenkreis", form, and variety. However, a major problem of Brandt's and Laackmann's studies is lumping species of very dissimilar morphology in the same genus.

In an attempt to correct this oversimplification, Kofoid and Campbell (1929, 1939) erected a very large number of new genera and species, thus falling into the opposite error. The great majority of these new taxa (especially the species) are based on minor and most probably intraspecifically variable morphological details. This classification erroneously suggests that tintinnid loricae exhibit sharp morphologic discontinuities, thus contradicting the increasing body of evidence pointing to smooth transitions between dissimilar morphotypes. Because Kofoid and Campbell's work has long been a primary reference for subsequent studies, subsequent literature is plagued with spurious species names.

Although later studies on South Atlantic tintinnids (Balech, 1942, 1945, 1948, 1951, 1958a, b, 1971a; Souto, 1979, 1981) placed some of Kofoid and Campbell's species into synonymy, they followed the general guidelines of these authors without major changes (mainly regarding the arrangement of species into families and genera). Alder's (1995) work on Antarctic tintinnids also followed the overall scheme proposed by Kofoid and Campbell but, having noticed extremely high intraspecific variability, proposed a drastic reduction in the number of species from 70 to 11, and invalidated several genera.

The neglect of intraspecific variability may be attributable to the scarcity of surveys based on abundant materials from geographically large areas. Indeed, most previous taxonomic studies have covered only restricted, geographically isolated locations and/or short time scales. Such spatial and temporal sample discontinuities, often combined with low sample retrieval and/or examination, prevented researchers from tracing subtle morphological transitions between the variability extremes within one species. Thus, closely spaced samples of neritic to oceanic waters at different times of year show that a single species may encompass several highly dissimilar morphotypes (e.g., Boltovskoy et al., 1989; Alder and Boltovskoy, 1991a, b; Alder 1995), as shown in morphological variat. C. gaussi. These results strongly support Brandt's (1906, 1907) emphasis on infraspecific ranks like "Formenkreis" and variety. Laackmann (1910), who also had very ample collections at his disposal, closely followed the criteria proposed by Brandt.

The methodology for classifying tintinnid species adopted in the present chapter does not differ from that applied by most specialists, based largely on attributes of the lorica. Its general morphology, presence of a collar, structure of the oral end, presence and type of striae and teeth, number of spiral turns and fenestrae, and ability to agglutinate particles are the most useful features for differentiating species. Other commonly used attributes, such as presence of an aboral horn or pedicel, and the type and density of agglutinated particles, were recently found to be highly variable at the intraspecific level (Laval-Peuto and Brownlee, 1986; Alder, 1995). In cases where the structure of the lorica is obscured by a dense agglomeration of particles (e.g., Tintinnopsis complex; see Identification of the Taxa), cytological and/or ultrastructural analyses are recommended. Most of the morphometric parameters regularly used for taxonomic purposes (total length, collar length, bowl length, length of aboral horn and pedicel, external oral diameter, maximum, minimum and middle diameter of bowl, aboral diameter) were also found to be highly variable intraspecifically. In agreement with several previous observations (e.g. Balech, 1959, Souto, 1979, Laval-Peuto, 1981, Laval-Peuto and Brownlee, 1986; Alder, 1995), the internal oral diameter is considered the only trait with sufficient intraspecific stability to be useful for taxonomic identifications. At any rate, because of the very high specific and supraspecific morphological variability of loricae, it is generally acknowledged that cytological studies are needed for a sound classification system.

Despite efforts to sort out the maze of conflicting names and morphotypes, the system presented herein contains undeniable ambiguities, especially where subtle morphological gradations between species and even genera are concerned. Keeping these taxa separate or lumping them into a single taxon is, at this point, largely a question of preference. Whenever possible these details are stressed in order to call the reader's attention.