Journ. Hattori Bot. Lab. No. 64 : 87-95 (June /988) A PROPOSAL FOR A NEW SYSTEM OF THE ANTHOCEROTAE, WITH A REVISION OF THE GENERA lIRO HASEGAWA1 In the Anthocerotae, so far, seven genera have generally been recognized - they are Anthoceros, Dendroceros, Folioceros, Leiosporoceros, Megaceros, Notothylas and Phaeoceros. The relationships among these taxa are not well understood and several classifications of the Anthocerotae are standing together. To improve such a confused condition I tried to re-evaluate many important char­ acters and came to the conclusion that the Anthocerotae could be classified in three families, namely the Anthocerotaceae, the Notothyladaceae and the Dendrocerotaceae (a new family). The present paper deals with some main points of this new system. I. TAXONOMIC STATUS OF THE LEIOSPOROCEROTACEAE Leiosporoceros dussii originally described under the genus Anthoceros from Mar­ tinique Island by Stephani (1893) may receive a special attention, be-cause it has several unusual features including oviod spores and isobilateral spore-tetrads. Hasegawa (1986) reported this species from Mexico showing the peculiarity of spores, spore-tetrads and elaters, and transferred the species to the genus Phaeoceros. Hassel de Menendez (1986) revised the taxonomy of Anthoceros dussii and proposed a new genus Leiosporoceros and a new family Leiosporocerotaceae for this species based on the character of spores and spore-tetrads. I) Characteristics of Leiosporoceros dussii a) Spores and spore-tetrads. The spores are ovoid or kidney-shaped, small (mostly 17-22 p. long), and nearly smooth (Fig. I-a). The spore-tetrads are of the isobilateral type (Fig. I-b). In addition to this tetragonal tetrads, a few hexahedral tetrads and various kinds of intermediate forms between tetragonal and hexahedral tetrads were found, but t~trahedral tetrads were never observed. As has been reported by Hasegawa (I.c.) and HasseI de Menendez (I.c.), the isobilateral spore-tetrads and the ovoid spores of this species are unique in the Anthocerotae. However, it is well-known that in the fern both the isobilateral and tetrahedral spore-tetrads occur in a single genus, a single species or even in a single sporangium. Moreover, also in hepatics some Riccia species and Geothallus tuberosus are known to have both types of spore-tetrads. So, I don't consider that the existence of isobilateral 1 Laboratory of Applied Botany, Faculty of Agriculture, Kyoto University, Kyoto 606, Japan. 88 Journ. Hattori Bot. Lab. No. 64 198 8 FIG. I. Leiosporoceros dussii (a-c) and Phaeoceros hirticalyx (d-f). a. Spore. b. Spore-tetrad. c. Elaters and spores. d. Proximal side of spore. e. D istal side of spore. f. Elater. a-c taken from Eggers & Frahm s.n. (herb. Eggers 34/15) in KYO; d- f from Akiyama 2366 in KYO. Scale line = O.O I mm. J. HASEGAwA: New system of the Anthocerotae 89 spore-tetrads in Leiosporoceros dussii is a character important enough to support a new family and a new genus. b) Elaters and thalli. Elaters of this species are also peculiar; they are narrow­ elongated and thick-walled (Fig. I-c), and somewhat look like those of the genus Folioceros. Thalli are solid and have dorsal surfaces densely covered with numerous small outgrowths (Fig. 2-a, c), which are also characteristic of this species. 2) Characteristics of Phaeoceros hirticalyx2 Phaeoceros hirticalyx shares some important characteristic features with Leio­ sporoceros dussii. In P. hirticalyx spores are small (mostly 22-27 ft in diam.) and nearly smooth except for narrow areas along the triradiate ridge (Fig. I-d, e), elaters are narrow-elongated and thick-walled (Fig. I-f), and dorsal surfaces of thalli are densely covered with numerous outgrowths (Fig. 2-e), just like those of Leiosporoceros dussii. 3) Relationship between Leiosporoceros dussii and Phaeoceros hirticalyx As mentioned above, Leiosporoceros dussii is different from Phaeoceros hirticalyx only in ovoid spores and isobilateral spore-tetrads, and both species have many con­ spicuous characteristics in common. Judging from the similarities of characters between Leiosporoceros dussii and Phaeoceros hirticalyx, I consider that they are vicarious species differentiated in the Neotropic and Paleotropic regions, respectively. So, Leiosporoceros dussii and Phaeoceros hirticalyx should be placed in the same genus. 4) Status of Leiosporoceros dussii Leiosporoceros dussii and Phaeoceros hirticalyx have yellow spores and solid thalli, both of which are important to define the genus Phaeoceros, but they can be distinguish­ ed from other species of Phaeoceros in the small, smooth spores and the narrow­ elongated, thick-walled elaters. Thus, I consider that it is reasonable to separate them 2 Phaeoceros hirticalyx (Steph .) Hasegawa, comb. novo Basionym: Aspriomitus hirticalyx Steph., Spec. Hep. 5 : 966 (1916). Syn. : Anthoceros polyandrus Steph., Spec. Hep. 5 : 987 (1916). Anthoceros tjibodensis Meijer, Reinwardtia 2 : 415 (1954). Anthoceros hirticalyx (Steph.) Meijer, J. Hattori Bot. Lab. 18 : 2 (1957). Phaeoceros polyandrus (Steph.) Haseg., Acta Phytotax. Geobot. 37 : 13 (1986). I (1986) recorded Anthoceros polyandrus from Ambon, transferring it to the genus Phaeoceros, and gave a detailed description with some drawings based on the material from Ambon. Thereafter, I noticed that Phaeoceros polyandrus had an earlier synonym: Meijer (1 957) treated Anthoceros poly­ andrus as a synonym of Anthoceros hirticalyx. After the investigation of the type material of Aspiro­ mitus hirticalyx (Schiffner 66 in G), I concluded that it was reasonable to follow Meijer (I.e.) as to the relationship between Aspiromitus hirticalyx and Anthoceros polyandrus. Meijer (I.e.) refrained from regarding Phaeoceros as an independent genus and assigned this species to the genus Anthoceros. However, I transfer Anthoceros hirticalyx to the genus Phaeoceros after Proskauer's (1951) classification of the genus Anthoceros S. lat. to which I agree. Although Phaeoceros hirticalyx has been so far recorded only from Java, Sumatra and Ambon, I presume that it might be widely distributed in tropical Asia. 90 Journ. Hattori Bot. Lab. No. 64 1 988 FIG. 2, Leiosporoceros dussii (a-c) and Phaeoceros hirticalyx Cd-e). a. A part of cross-section of thallus, showing dorsal surface with numerous small outgrowths, x 315. b. Thallus, x 4.3. c. Cross-section of thallus, showing large Nostoc colony, x 75. d. Thallus, x 4.2. e. A part of cross-section of thallus, x 75. a-c drawn from Eggers & Frahm s.n. (herb. Eggers 34/15) in KYO; d-e from Akiyama 2366 in KYO. from other species of Phaeoceros at a rank of subgenus. n. RELATIONSHIP BETWEEN DENDROCEROS AND MEGACEROS The genus Dendroceros has been considered to be related to the genus Megaceros, because they share some characters (for example unispiral elaters and sporophytes without stomata), and because there exist some intermediate forms between Dendro­ ceros and M egaceros. Dendroceros giganteus is generally regarded as a member of the genus Dendroceros, but it is said that the species is also related to the genus Megaceros and sometimes said that it is situated in an intermediate position between Dendroceros and Megaceros. J. HASEGAWA: New system of the Anthocerotae 91 Indeed the species was regarded as a member of M egaceros by Stephani (1916) and Khanna (1944). 1) Characteristics of Dendroceros giganteus a) Thallus form. Compared with other species of Dendroceros, D. giganteus has fairly large thalli (Fig. 3-a), but the structure of thalli consisting of costae and unistra­ tose laminae is similar to those of other species of Dendroceros. Chiefly because of this similarity in thallus-structure, D. giganteus might be placed in the genus Dendroceros rather than in the genus Megaceros. However, in Dendroceros giganteus laminae are usually absent in young parts (Fig. 3-b). According to E. O. Campbell (1986), the laminae develop late from a mar­ ginal meristem in this species. So, the manner of lamina-formation in D. giganteus seems to be different from that of other species of Dendroceros. In Dendroceros the laminae develop directly from a segment of the apical cell (Renzaglia 1978). b) Laminal cells. As is well-known, laminal cells of Dendroceros species usually have trigone-like, triangular or quadrate to rectangular, thick walls at their corners, often perforated (Fig. 3-c). Such collenchymatous cells in Dendroceros are unique in the Anthocerotae, and I consider that this character is important enough to define the genus Dendroceros. In Dendroceros giganteus laminae consist of thin-walled cells throgh­ out and never have perforations (Fig. 3-d). c) Chloroplasts. One of the most conspicuous features of the genus Megaceros is the nature of chloroplasts. In Megaceros species most epidermal cells of thalli have two or more chloroplasts in a cell and inner cells have usually many small chloroplasts - according to Burr (1969) the number of chloroplast attains to 14. Dendroceros giganteus also has tow or more chloroplasts in a cell of thallus­ laminae (Fig. 3-d), in contrast to other species of Dendroceros which have a single chloro­ plast in a cell (Fig. 3-c). d) Spores. In Dendroceros spores germinate within the spore coat and are multi­ cellular at maturity. The shape is mostly globose to rounded ovoid and the surfaces are finely vermiculate to minutely papillate throughout with indistinct triradiate marks (Fig. 3-e). However, in Dendroceros giganteus spores are unicellular and rounded tetrahedral with distinct triradiate marks, and surfaces are minutely papillate in the proximal side and roughened with scattered truncate outgrowths in distal side (Fig. 3-f, g). The shape and surface ornamentation of D. giganteus are quite similar to Megaceros species (Fig. 3-h). 2) Taxonomic position of Dendroceros giganteus As discussed above, Dendroceros giganteus has little relationship to the genus Dendroceros but a close relationship to the genus Megaceros. Now, D. giganteus should be classified in the genus M egaceros. I consider, however, that it is reasonable to dis­ tinguish the species from other species of Megaceros at a rank of subgenus, because of its peculiar thallus form.