Garbary, Hansen, & Scagel (1980), after comparing various schemes of classifying Rhodophyceae subclass Bangiophycidae at the ranks of order and family, beginning with Skuja (1939), proposed one of their own that incorporates certain notable features. In their scheme, they give more importance to cytology than to gross morphology, so that all unicellular and pseudofilamentous forms in which the cell has a single stellate (or modified stellate) chloroplast with a central pyrenoid are placed in one family. This treatment contrasts with previous classifications in which the two morphological groups were assigned to separate orders or at least to separate families. I agree that the blurring of the boundary between the two groups by the capacity of certain genera to form palmelloid colonies favors the recognition of a single family.
In choosing the correct name for this family, three names come into consideration: Goniotrichaceae, Stylonemataceae, and Porphyridiaceae. The oldest of the three names is Goniotrichaceae, which, although usually accredited to Skuja (1939: 31), was first proposed by G.M. Smith (1933: 120, 121). This name, however, must remain attached to its type, Goniotrichum Kützing, which, after many years of indecision, has turned out to be referable to the Erythrotrichiaceae rather than to the family traditionally called Goniotrichaceae.
The problem was initiated by Kützing (1843b: 244), who, in establishing Goniotrichum, erroneously associated the material at hand (upon which he based the generic diagnosis) with Conferva ceramicola Lyngbye (1819: 144, pl. 48D), the type of Erythrotrichia Areschoug (1850b: 435) (nom. cons.). Kützing's error was widely recognized, but its perceived effect on nomenclature varied in accordance with different interpretations of the ICBN. Most modern workers tacitly accepted Kützing's material as the type of the generic name, considering his citation of Conferva ceramicola an admissible error. K. Drew (1956), however, counseled by Mr. Robert Ross (British Museum [Natural History]), took the view that C. ceramicola must be accepted as the type of Goniotrichum and that this generic name, therefore, must be considered an earlier nomenclatural synonym of Erythrotrichia. The unfortunate consequence of replacing Erythrotrichia with Goniotrichum would be forestalled by the fact that Erythrotrichia has already been conserved (against Ceramicola Oersted 1844). A replacement for Goniotrichum as applied to pseudofilamentous forms was sought, and the generic name Stylonema Reinsch (1875: 40) was uncovered. This genus was established for a new species, S. cornu-cervi, which is clearly referable to the genus described by Kützing as Goniotrichum. Realizing that the name Goniotrichaceae G.M. Smith (erroneously accredited to Skuja) and Goniotrichales Skuja (1939: 31) were no longer applicable to pseudofilamentous forms, K. Drew (1956: 73) proposed the names Stylonemataceae and Stylonematales, both validated by reference to Skuja's diagnosis.
The uncertainty as to whether a generic name should be typified by material at hand (upon which the diagnosis was based) or by a cited species was ended by the Nomenclature Section of the Thirteenth International Botanical Congress, meeting at Sydney in 1981, which decided in favor of a cited species (Art. 10.2). Only by conservation can the type be a specimen or illustration that is not the type of a cited species (Art. 10.4). The fact that most literature dealing with Goniotrichum has used this name would seem to argue for conservation, and an appropriate proposal was made by Demoulin & Hoffmann (1992), but the availability and ever-increasing use of an unencumbered synonym (Stylonema) and its derivatives (Stylonemataceae and Stylonematales) makes conservation unnecessary.
The third family name to be considered, Porphyridiaceae, is usually accredited to Kylin (1937b: 122), who, however, listed the included genera but not its characters. The name was validated by Skuja (1939: 30), who provided a single diagnosis that applied both to the order Porphyridiales and the family Porphyridiaceae. Stylonemataceae K. Drew (1956) is available for classifications in which the pseudofilamentous forms are placed in a family separate from the unicellular forms, but Porphyridiaceae Kylin ex Skuja (1939) has priority when the two families are merged, as in the present catalogue. Similarly, Porphyridiales Kylin ex Skuja (1939) has priority over Stylonematales K. Drew (1956) when the two orders are merged, as in the present catalogue. Porphyridiales is usually accredited to Kylin (1937b: 122), who, however, did not provide a validating description.
Garbary, Hansen, & Scagel (1980) distributed the multicellular Bangiophycidae among three orders: Erythropeltidales, Bangiales, and Rhodochaetales. This treatment differs from previous classifications in emphasizing the importance of the pattern of reproduction, in which a monosporangium is cut off by a curved wall from an undifferentiated cell. As a consequence, Compsopogonaceae was joined with Erythropeltidaceae in one order, Erythropeltidales, leaving the Bangiaceae as the sole component of the Bangiales. The establishment of the Erythropeltidales as a new order, however, was unnecessary because of the availability of the ordinal name Compsopogonales Skuja (1939: 34). Considering that the morphology of Compsopogon and Compsopogonopsis is complex compared to that of other members of the Erythropeltidales and that there are several reports of variant structural and reproductive features (admittedly unconfirmed), I prefer to retain the Compsopogonaceae in its own order. The order comprising the Erythropeltidaceae exclusive of the Compsopogonaceae is correctly called Erythropeltidales.
The family usually called Erythropeltidaceae Skuja (1939: 33) should be called Erythrotrichiaceae G.M. Smith (1933: 120, 122), which has priority but which was inadvertently omitted from my compilation of family names of living algae (P. Silva, 1980b). The name Erythrotrichiaceae was initially incorrect because the family to which it applied included Compsopogon, for which the family name Compsopogonaceae Schmitz (1896c: 318) is available. It is correct, however, to use Erythrotrichiaceae in a sense that excludes Compsopogon, as in this catalogue.
Kylin (1956: 51, 55--56) recognized four genera in the Erythrotrichiaceae (Erythropeltidaceae):
Erythrotrichia Areschoug (1850b: 435). Thallus erect, filamentous or ribbonlike, attached by rhizoids. (Type = Conferva ceramicola Lyngbye.)
Erythropeltis Schmitz (1896b: 313). Thallus composed of erect filaments arising from a monostromatic disc. (Type = Erythrotrichia discigera Berthold.)
Erythrocladia Rosenvinge (1909: 71). Thallus a monostromatic (seldom distromatic) disc formed by branched prostrate filaments that tend to coalesce. (Lectotype = Erythrocladia irregularis Rosenvinge.)
Porphyropsis Rosenvinge (1909: 68). Thallus a monostromatic membrane resulting from the rupture of a sac that develops from a polystromatic crust. (Type = Porphyra coccinea J. Agardh ex Areschoug.)
On the basis of culture studies, Kornmann (1984) and Kornmann & Sahling (1985) proposed significant changes in this generic scheme. Kornmann (1984) found that an alga that he identified as Erythropeltis discigera (Berthold) Schmitz (Erythrotrichia discigera Berthold, 1882a: 511; 1882b: 25, pl. 1: figs. 15--18) had a life history comprising two facultative heteromorphic phases, one filamentous and the other discoid, both reproducing by monospores. (In addition, spermatia were found in abundance on the filaments.) Earlier, Batters (1900: 374--375) and Rosenvinge (1909: 72, footnote) had pointed out that when Schmitz (1896b: 313) established the genus Erythropeltis on the basis of Erythrotrichia discigera, he characterized it as having only a disc, either overlooking or ignoring the fact that Berthold described erect filaments arising from a disc.
Rejecting the name Erythropeltis because of Schmitz's erroneous interpretation of the type species and restricting Erythrotrichia to those algae that have erect filaments attached by rhizoids, Kornmann (1984) established a new genus to receive E. discigera, which he named Erythrotrichopeltis. Considering E. discigera conspecific with Bangia ciliaris Carmichael (in W. Hooker, 1833: 316), Kornmann made the combination Erythrotrichopeltis ciliaris. A second species of the genus was also recognized, Erythrotrichopeltis boryana (Montagne) Kornmann, based on Porphyra boryana Montagne (1846b: 150, pl. 3: fig. 2). As Wynne (1986a) has pointed out, Porphyra boryana is the type species of Porphyrostromium Trevisan (1848: 100), a name that Kornmann should have used for his new genus. Wynne made the appropriate combination Porphyrostromium ciliare. The second species should be called Porphyrostromium boryanum (Montagne) P. Silva, comb. nov. When establishing the genus, Trevisan illegitimately changed the epithet of Porphyra boryana to boryi. Wynne (1986a: 329), considering Trevisan's change of epithet to be a correctable error, wrote ``Porphyrostromium boryana (Montagne) Trevisan'', but because boryi and boryana are distinct epithets rather than orthographic variants, a new combination is necessary. The binomial as used by Wynne is not valid because the page of publication of the basionym was not cited.
Erythrocladia was established by Rosenvinge to receive two previously undescribed species, E. irregularis and E. subintegra. Both form a prostrate, essentially monostromatic thallus, but in E. irregularis, as the epithet implies, the filaments are irregularly branched and coalesce only basally, while in E. subintegra the filaments coalesce throughout except at the margin of the disc, where the apices are forked. Heerebout (1968: 141) merged E. subintegra with E. irregularis, the latter having been chosen as lectotype of the genus by Kylin (1956: 55). Garbary, Hansen, & Scagel (1981: 254) treated E. subintegra as a forma of E. irregularis. Kornmann & Sahling (1985), by contrast, considered the difference between the two species to be of generic value, basing their opinion on a comparison of the development of the thallus from monospores. In E. subintegra a disc is apparent as early as the 4-celled stage, continuing to expand by marginal growth. Heerebout's conclusion was said to have been based on mixed cultures (Kornmann & Sahling, 1985: 221, 223; Kornmann, 1989: 223, 225).
Retaining the name Erythrocladia for E. irregularis, Kornmann & Sahling applied the name Erythropeltis to E. subintegra. They believed that Schmitz's definition of Erythropeltis, although erroneous when based on E. discigera, was correct when based on E. subintegra. Such nomenclatural legerdemain is not tenable, however, and the name Erythropeltis must remain with its type species, E. discigera. Like Erythrotrichopeltis, it is a taxonomic synonym of Porphyrostromium. Upon realizing that the genus conceived by Kornmann & Sahling as typified by E. subintegra was without a valid name, Kornmann (1989) rectified the situation by describing it under the name Sahlingia.
The order Bangiales is usually accredited to Schmitz (1892: 15), but was first proposed by Nägeli (1847: 136). (The name used by Nägeli, Bangiaceae, is correctable to Bangiales in accordance with Art. 17.3). The family Bangiaceae should be accredited to Engler (1892: 16). As traditionally circumscribed, Bangiaceae includes the genus Porphyra, for which the family name Porphyraceae Kützing (1843b: 382, as Porphyreae) was established. A proposal to conserve Bangiaceae against Porphyraceae (P. Silva, 1980b: 106) was approved by the Committee for Algae and approved by the Fifteenth International Botanical Congress (see P. Silva, 1993b: 707).
The order Rhodochaetales is usually accredited to Skuja (1939: 34), but was first proposed by Bessey (1907: 291).