Family Amoebidae Ehrenberg, 1838 (= Chaidae Poche, 1913)
Diagnosis; Locomotive form with one or, commonly, more broad, subcylindrical granuloplasmic pseudopodia with hyaline caps. Floating form with pseudopodia like those of locomotive form but thinner and usually more numerous, projecting in several directions. Uni- or multinucleate. In majority, nucleolar material in many small pieces. Mitosis, where known, acentric and semi-open or open. Numerous cytoplasmic crystals which usually have a bipyramidal or plate-like form. Glycocalyx normally developed into a thick coat of filaments often radiating perpendicularly from plasma membrane. Distinct inner fibrous lamina in nuclei of most but not all genera.
A. Taxonomic characters for the Amoebidae
1) Locomotive form
The characteristics of the locomotive form used in the past (e.g., Schaeffer 1926, Page 1976) to distinguish amongst genera include polypodial v. monopodial habit, dominance or not of one pseudopodium, and differentiation of posterior end, i.e. uroid. The first of these was recently discussed by Siemensma and Page (1986). The occurrence of several rather than one main pseudopodium (i.e., dominance or not of a single pseudopodium in a polypodial form), which Schaeffer (1926) made the distinguishing character of Polychaos, is less striking in P. fasciculatum than in the larger, P. dubium. The posterior bundle of pseudopodial remnants on these two species seems at least as useful (Page and Baldock 1980). However, the common palmate or polypodial form of P. dubium is very similar in the Russian strain, in other laboratory strains, in Schaeffer's (1916) American amoebae, and in those which I have observed 25 years ago in the USA.
Although Shah (1971) described a large Amoeba without crystals, all well-studied laboratory strains of Amoebidae contain crystals, the most common form of which is bipyramid. The paired inclusions of D. algonquinensis are therefore especially remarkable as one indication amongst others of a difference from other species classified as Amoeba, though the observations of P. dubium demonstrate that D. algonquinensis is not unique in this family in containing paired inclusions. That this difference is not due to diet is supported by the observations that crystals in Amoeba sp. (Bor) and in A. proteus do not differ noticeably in form whether the food organism is Tetrahymena or Chilomonas. Furthermore, both P. fasciculatum and T. sinuosa, feeding on Chilomonas, contained bipyramids but no paired bodies.
3) Surface structure
The coat of hair-like filaments, usually individually distinguishable, is characteristic of this family, and amongst all Gymnamoebia with their diverse surface structures (Page 1983) it takes this form only in the Amoebidae and in Thecamoeba proteoides, which was considered intermediate between Thecamoebidae and Amoebidae (Page 1978). The lack of such a covering on A. leningradensis can be interpreted as a reduction of the glycocalyx found in related species and may be associated with the greater force required to detach A. leningradensis from a substratum, compared with A. proteus (Opas and Kalinina 1980: strain Sh = A. leningradensis).
4) Nuclear number
The finding of a honeycomb-like organisation in the nuclear laminate of C. carolinense and C. nobile reduces even further the known differences between the multinucleate Chaos and the uninucleate Amoeba. If other multinucleate Amoebidae are found to differ unltrastructurally from these two species, a genus based almost entirely on nuclear number may be difficult to uphold. A re-examination of C. illinoisense, in which Daniels and Roth (1964) found no honeycomb lamina, would be important for determining relationships of multinucleate Amoebidae, if that species could be re-isolated.
5) Nuclear envelope
The five largest nuclei had an inner lamina organised in the honeycomb pattern, a finding which lends some support to the idea that the principal function of the lamina is to support larger nuclei in maintaining their shapes (Page and Robson 1983). On the oter hand, the smallest nucleus, that of D. algonquinensis, had a distinct, thick lamina of fibres running parallel to the nuclear membrane, but the larger nuclei of T. sinuosa and H. hydroxena apparently had none and there was only a suggestion of such a layer in P. fasciculatum, whose mean nuclear diameter is almost twice that of D. algonquinensis. The differing degrees of irregularity in nuclear contour do, however, support the suggestion of this function.
6) Nucleolar organisation
The typical nucleus of the Amoebidae is the granular or ovular one (Raikov 1982) with many nucleoli, often in a layer just inside the nuclear envelope. However, P. fasciculatum, undoubtedly a member of the family, demonstrates that the nuclear material can occur in large bodies, even a single, lobed piece (Page and Baldock 1980). The ovular nucleus is therefore not the only one possible. In the Thecamoebidae, even leaving the borderline Thecamoeba proteoides out of consideration, nucleolar organisation ranges from a single central nucleolus as in T. quadrilineata to ovular nuclei as in T. similis (Page 1976, 1977). The occurrence of a similar diversity in the Amoebidae cannot be ruled out.
7) Mitotic pattern
Gromov (1985) has demonstrated that mitosis in A. proteus is semi-open, an interpretation that can be placed also on the earlier findings of Roth et al. (1960). Whether mitosis in Chaos is open, as suggested by the findings of Roth and Daniels (1962) and Daniels and Roth (1964) is a question that needs re-examination in view of Gromov's report.
The constancy of these and other taxonomic characters is suggested by the similarity even ultastructure amongst strains of A. proteus from the USA, Scotland, and the USSR.
B. Nomenclatural problems
The older taxonomic names have provided material for much controversy, but those arguments will not be revived here. In general, usage in this publication follows that current amongst workers studying the organisms. The type genus of the family is therefore spelled Amoeba rather than Amiba. The argument about the validity of Chaos and identity of the organism seen by Rosel von Rosenhof will never be settled, and it must be acknowledged that the type species of Chaos as well as those of several other genera of gymnamoebae cannot be identified. There is little question that the organism described by Wilson (1900) as Pelomyxa carolinensis is the largest Chaos known today. The application to it of the name Chaos carolinensis rather than Chaos chaos, now widely accepted, seems to date from Jahn and Jahn (1949). This usage was accepted by Page (1976), with correction of the adjectival ending, as Chaos carolinense. It must be emphasised that geographic names often used as specific epithets (e.g., carolinense, leningradensis) do not indicate limited geographic distribution. The widespread occurrence of species of Amoebidae is becoming evident. Authorships are another matter for controversy. Ehrenberg (1838), defining the genus which he wrote Amoeba, described it as "Animal Amoebaeorum familiae characteribus instructum" and is therefore recognised here as author of the family, according to an interpretation of Article 11(f) of the zoological code (International Commission of Zoological Nomenclature, 1985). If this authorship is not acceptable because the world "Amoebaeorum" is in the genitive, it might be noted that Dujardin (1841) also designated the "Amibiens" as a family. The common attribution of the authorship of "Diesing, 1848" (Page 1976) should therefore be dropped. Although we should not recognise A. proteus by any description before Leidy's (most accessible in Leidy, 1879), the accepted attribution of the species is Amoeba proteus (Pallas, 1766) Leidy, 1878.
C. Possible relationships
1) Intrafamilial relationships
The utrastructural findings leave little doubt that the Amoebidae are a natural family. Likewise, they suggest some relationships within the family, according in general with those indicated by Friz (1974, 1979, 1984). For example, Friz (1974) doubted that A. proteus and A. dubia (= P. dubium) could be classified in the same genus if A. proteus and C. carolinense were classified in different genera, Sopina et al. (1979) also found such a separation. The ultrastructural findings not only support these conclusions but emphasise once more that nuclear number seems the only clear difference between Amoeba and Chaos. It may seen that more diversity is tolerated within the genus Polychaos than the genus Amoeba. The retention of both P. dubium and P. fasciculatum in one genus contrasts with the separation of D. algonquinensis from Amoeba. However, D. algonquinensis is in clear opposition to a group of species within the genus Amoeba, and it seems likely that Deuteramoeba will prove to include several species of smaller polypodial Amoebidae lacking the distinctly hair-like coating found on Amoeba and Polychaos but with a more amorphous surface coat. The eventual status of crystalline form as a generic character is uncertain. Perhaps future examination of more species of Polychaos will indicate similar division of the genus. However, the locomotive form of P. fasciculatum has some common characters (e.g., multiple trailing pseudopodial remnants) with those often found in P. dubium. The nuclear structures are the point of greatest difference between the two species. Some differences may be related to the difference in cell size, especially the sparser development of the long, thin, hair-like surface filaments of P. fasciculatum. On the other hand, maintenance of a separate genus for the parasitic Hydramoeba hydroxena over against the free-living Trichamoeba sinuosa and other species of Trichamoeba appears even more difficult in the face of their ultrastructural similarities. The possibility that H. hydroxena may not be obligately parasitic was considered once more by Page and Robson (1983), but a decision on the generic question should probably be deferred until more rigorous tests of Hydamoeba's nutritional needs have been conducted.
2) Possible relationships with other families
Information which suggests but does not conclusively demonstrate relationships between the Amoebidae and several other families has been accumulating. In general, similarities between the Amoebidae and the Thecamoebidae appear greatest, particularly if attention is directed to the larger members of the latter family, Thecamoeba (Page 1977, Page and Blakey 1979) and Thecochaos (Page 1981), though no ultrastructural information is available for the latter. These are mostly medium-sized to large organisms. Ovular nuclei occur in the Thecamoebidae. Houssay and Prenant (1970) and Haberey (1973) found an inner nuclear lamina, which appeared honeycomb-like, in an amoeba identified as Thecamoeba sphaeronucleolus. All members of the genus Thecamoeba have a thick glycocalyx (Faure-Fremiet and Andre 1968, Page and Blakey 1979), but this is more amorphous than the distinct radiating filaments of most Amoebidae. The one exception to the latter statement is Thecamoeba proteoides, which appears intermediate between the two families (Page 1978), so that one must now wonder whether it should be transferred to the Amoebidae. Apart from the light microscopic appearance which originally suggested this present taxonomic position T. proteoides lacks crystals or birefringent inclusions, a character in which it agrees with the Thecamoebidae. Thus the Amoebidae and the Thecamoebidae appear related though distinct families. Another family for which a possible relationship with the Amoebidae can be argued is the Hartmannellidae. A comparative ultrastructural study (Page 1985) has presented evidence that this is a natural family. However, the genus Saccamoeba, whose position in the Hartmannellidae is also supported by that study, shows marked similarities to Trichamoeba. The best-known species, Saccamoeba limax, usually contains bipyramidal crystals, though in smaller numbers than the Dutch Trichamoeba sp. and other Amoebidae. Saccamoeba and Trichamoeba have similar locomotive habits, though the floating form of Saccamoeba (like that of other hartmannellids) lacks the pseudopodia found on floating forms of Trichamoeba and other Amoebidae. The nucleus of known Saccamoeba species has a single, central nucleolus, but the advisability of caution about the possible diversity of nucleolar organisation in Amoebidae has been pointed out. The surface structures of Saccamoeba and Trichamoeba differ somewhat. Less obvious are the indications of a relationship between the Amoebidae and larger amoebae of the family Paramoebidae, particularly Mayorella. This genus has subpseudopodia produced usually from its main lobopodium, as do many other gymnamoebae. However, close observations of formation of Mayorella subpseudopodia (Page 1981, 1983) suggest strongly that these subpseudopodia may be produced by a hydrostatic mechanism compatible with the generalised cortical contraction theory of amoeboid movement in A. proteus (Grebecki 1982, Grebecka 1982). Finer subpseudopodia of some other genera contain filamentous cores (Bowers and Korn 1968, Page 1981, Page and Willumsen 1983), indicating a different method of production in those genera. This apparent functional similarity of Amoebidae and Mayorella may or may not have taxonomic significance. The paired inclusions so far known only in Deuteramoeba, P. dubium, and Mayorella (Page 1983) may indicate another physiological similarity. Speculatively, then, similarities if not definite relationships have begun to appear amongst members of these four families. The positions in this scheme of smaller amoebae classified in the Thecamoebidae and the Paramoebidae are not obvious, and the Paramoebidae may prove to be heterogeneous. The possibility of three-dimensional rather than exclusively linear relationships amongst the families merits consideration.