J. Hattori Bot. Lab. No. 76: 21-34 (Oct. 1994)

NEW CLASSIFICATION OF ANTHOCEROTAE

JIRO HASEGAWA 1

ABSTRACT. The classification of the Anthocerotae is revised based on a cladistic analysis. In the analysis the phylogenetic relationships among ten anthocerote genera or subgenera are based on the available morphological, anatomical and chemical data, using the Hepaticae and the Musci as the outgroup. Six most parsimonious cladograms resulted. The strict consensus tree calculated from these six trees suggests that Notothylas is the sister group to all the remaining taxa, which again form two major branches: one contains five taxa including Anthoceros and Phaeoceros, and another has four taxa including Dendroceros and Megaceros. In the former branch, the clade uniting Anthoceros and Folioceros is strongly supported, but the relationships of Phaeoceros are poorly resolved. In the latter branch, three paraphyletic terminal clades are found. On the basis of the cladistic analysis pre­ sented here, the two family and two subfamily classification of the Anthocerotae is proposed. The taxonomic status of Folioceros, Leiosporoceros, Notoceros, Apoceros and Hattorioceros is discussed, and the genus Notoceros (Schust.) Haseg. stat. nov. and the genus Hattorioceros (Haseg.) Haseg., stat. nov. are also proposed.

INTRODUCTION The Anthocerotae are a monophyletic small group with a rather small number of fam­ ilies, genera and species. Although special attention has been paid to the Anthocerotae from a phylogenetic point of view, little has been mentioned about the classification of higher categories of the Anthocerotae until recently when Schuster ( 1987), Hassel de Menendez ( 1988) and Hasegawa ( 1988) almost simultaneously proposed new classifica­ tions of the Anthocerotae (Table l ). As shown in Table I, the classifications of Schuster and Hasegawa are very similar in all essential points. They classified the Anthocerotae into three families or subfamilies (in­ cluding the autonomous family or subfamily, Notothyladaceae or Notothyladoideae), and regarded Folioceros and Leiosporoceros as subgena of Anthoceros (Aspiromitus sensu Schuster) and Phaeoceros (Anthoceros sensu Schuster), respectively. On the contrary, the classification of Hassel de Menendez is quite different from those of Schuster and Hasegawa. In her classification, five families (including the monotypic families, Foliocero­ taceae and Leiosporocerotaceae) were recognized, and Notothylas and Phaeoceros were classified in the same family. Schuster (I 992) reviewed these three classifications, and sug­ gested that the classification of Hassel de Menendez could not reconcile with that of the other two. Thus, at present two basically different classifications of the Anthocerotae exist. To improve this confusion, I attempted to re-evaluate characters on which these classi­ fications were based. Moreover, as the classification of Hassel de Menendez (I 988) was based on a cladistic analysis, I also attemped to examine the reality of these classifications in the light of the cladistic analysis.

1 Laboratory of Applied Botany, Faculty of Agriculture, Kyoto University, Kyoto 606, Japan. 22 J. Hattori Bot. Lab. No. 76 I 9 9 4

Table 1. Classification of the Anthocerotae.

Hassel de Menendez ( 1988) Hasegawa (1988) Schuster ( 1992)

Ord. Anthocerotales Ord. Anthocerotales Ord. Anthocerotales Fam. Anthocerotaceae Fam. Anthocerotaceae Fam. Anthocerotaceae Gen. Anthoceros Gen. Anthoceros Subfam. Anthocerotoideae Sphaerosporoceros Subgen. Anthoceros Gen. Aspiromitus Fam. Notothyladaceae Folioceros Subgen. Aspiromitus Subfam. Notothyladoideae Gen. Phaeoceros Folioceros Gen. Notothylas Subgen. Phaeoceros Gen. Anthoceros Subfam. Phaeocerotoideae Leiosporoceros Subgen. Anthoceros Gen. Phaeoceros Gen. Megaceros Leiosporoceros Ord. Foliocerotales Subgen.Megaceros Subfam. Dendrocerotoideae Fam. Foliocerotaceae Australoceros Gen. Megaceros Gen. Folioceros Fam. Dendrocerotaceae Subgen.Megaceros Ord. Leiosporocerotales Gen. Dendroceros Notoceros Fam. Leiosporocerotaceae Fam. Notothyladaceae Gen. Dendroceros Gen. Leiosporoceros Gen. Notothylas Subgen. Dendroceros Ord. Dendrocerotales Apoceros Fam. Dendrocerotaceae Subfam. Notothyladoideae Gen. Dendroceros Gen. Notothylas Megaceros

A cladistic analysis of the Anthocerotae was, for the first time, given by Mishler and Churchill (1985). Although only five genera of the Anthocerotae were included in their analysis, it suggested some critical points involved in the classification of the Anthocero­ tae; they included problems relating to ( l) their origin and ancestry, (2) the naturalness of the family Anthocerotaceae (sensu lat.), (3) the relationships between Anthoceros and Phaeoceros, and (4) the relationships between Notothylas and the remaining anthocerote genera. Hassel de Menendez (1988) also presented a cladistic analysis of the Anthocerotae, but the cladogram given by her was quite different from that by Mishler and Churchill (1985). One of the most serious problems invloved in a cladistic analysis of the Anthocerotae is that the relationships of the Anthocerotae to other plant groups are so ambiguous that ap­ parent homoplasy occurs everywhere. Mishler and Churchill (1985), in their cladistic ap­ proach to the phylogeny of bryophytes, described the situation as follows; "The relation­ ships of the Anthocerotae are not so straightforward, which is curious since this group has fewer species and is less diverse than the other groups in the analysis. Almost all the appar­ ent homoplasy we detected in the analysis involved the hornworts one way or the other." Thus, it seemed very difficult to designate the outgroup of the Anthocerotae. Recently, Mishler et al. (I 992) and Waters et al. ( 1992), however, showed the mono­ phyletic origin of land plants including the Anthocerotae and the paraphyletic relationships of bryophytes based on their molecular approach to the phylogeny of bryophytes, and gave support to the view of Mishler and Churchill ( 1985), i.e., that the Anthocerotae are the sis- J. HASEGAWA : New classification of Anthocerotae 23 ter group of the lineage including the Musci and the Tracheophyta, and the Hepaticae is the sister group of the remaining land plants including the Anthocerotae. The present study is intended to revise the classification of the Anthocerotae through developing the ideas on which the analysis of Mishler and Churchill (1985, 1984) was based.

CLADISTIC ANALYSIS OF THE ANTHOCEROTAE

Taxa included in the cladistic analysis In this analysis the following ten genera or subgenera were recognized as operational taxonomic units (OTU); Anthoceros, Apoceros, Dendroceros, Folioceros, Hattorioceros, Leiosporoceros, Megaceros, Notoceros, Notothylas [in Schuster's (1992) nomenclature the genus Aspiromitus was accepted, and consequently Phaeoceros was reduced into a syn­ onym of Anthoceros, but to avoid confusion here I follow the traditional Proskauer (1951) nomenclature]. Hattorioceros was a taxon recently described as a subgenus of Phaeoceros for a species with unusual spore features which had not been known from the Anthocerotae (Hasegawa 1994). In the classification of Hassel de Menendez (1988) another genus Sphaerosporoceros was recognized, but it was not included in this analysis, because as sug­ gested by Schuster (1992) the naturalness of the genus Sphaerosporoceros as defined by Hassel de Menendez was debatable. On the contrary, Notoceros (Australoceros sensu Hasegawa) and Apoceros which were not treated by Hassel de Menendez (I 988), were included in this analysis, as I considered them as well-defined.

Methods of the analysis Forty-two characters listed below were analysed using the branch and bound algo­ rithum in PAUP (version 3.1 for the Macintosh; Swofford 1993); no transformation series were hypothesized for the multistate characters and all character states were weighted equally, and the Hepaticae and the Musci were designated as outgroups based on the phy­ logeny for the bryophytes presented by Mishler and Churchill (1985), Mishler et al. ( 1992) and Waters et al. (1992). The distribution of characaters among ten taxa and the outgroup is given in Table 2.

Characters included in this analysis I. Archegonium stalked (with free archegonial wall) (O); embedded in thallus (without free archegonial wall) (1). 2. Antheridium originated exogenously (O); endogenously ( l ). 3. First division of zygote transverse (O); vertical ( l ). 4. Slime papillae or slime hairs present on gametophyte (O); absent. (1). 5. Stoma-like clefts absent on gametophyte (O); present (1). 6. Asymetric spermatozoid (O); bilaterally symetric spermatozoid (I). 7. Channel thylakoid absent (O); present (1) 8. Ability to distinguish D-methionin absent (O); present (1). Table 2. Character matrix used for this analysis. "'.j>. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21)

Anthoceros 1 1 l 1 1 1 1 1 0 0 0 0 I 0 011 0 1 0 0 Folioceros I I I 1 1 1 1 I 0 0 0 0 1 0 Oil 0 I 0 0 Phaeoceros 1 1 1 I I l 1 I 0 0 0 0 0 Oi l 0 0 0 0 0 Leiosporoceros 1 1 1 l 1 1 I 1 0 0 0 0 0 0 0 ? ? 0 0 Hattorioceros 1 1 1 1 I 1 I 1 0 0 0 0 0 0 0 ? ? 0 0 Megaceros I I 1 1 1 1 1 1 0 0 0 0 0 0 0 1 0 0 0 Notoceros 1 I I 1 1 1 1 1 ? 1 0 0 1 0 0 1 0 0 0 Dendroceros 1 1 1 1 1 1 1 1 1 I 1 I 0 0 0 1 0 0 0 Apoceros 1 I 1 I 1 1 1 I 1 I I I 1 0 0 I 0 0 0 1 Notothylas 1 1 I I l 1 1 I 0 0 0 0 0 0 0 0 0 1 1 011 0 ~ ::i:: ::::: "'0 Hepaticae 0 0 0 0 0 0 0 0 ? Oi l I ? 011 Oi l Oi l 011 ? 0 0 ? 0 :l. Musci 0 0 0 ? 0 0 0 I ? ? 0 ? ? 011 Oi l ? ? 0 0 ? 0 tti ?. r "'~ (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) z ?.._, a, Anthoceros I 1 1 0 0 0 0 0 0 0 0 2 I 0 1 I 0 1 0 I 1 Folioceros I 1 I 0 0 0 0 0 0 0 Oi l 2 1 1 1 Oil 0 1 0 I 1 Phaeoceros 1 1 I 0 0 I 0 0 0 0 0 1 1 0 0 1 0 1 1 0 0 Leiosporoceros I I 1 1 I 0 1 0 1 0 0 1 I I ? I 0 1 ? ? 0 Hattorioceros I I 1 ? 2 0 0 1 1 0 0 1 1 0 ? 1 0 1 ? ? 0 Megaceros 1 1 1 0 0 0 0 0 0 0 1 0 I 2 0 0 1 1 1 0 0 Notoceros 1 1 1 0 0 0 0 0 0 0 1 0 1 2 0 0 1 1 1 0 0 Dendroceros 1 1 1 0 0 0 0 0 0 1 1 0 1 2 0 0 0 1 I 0 0 Apoceros 1 1 1 0 0 0 0 0 0 1 1 0 1 2 0 0 0 1 I 0 0 Notothylas 0 011 0 0 0 1 0 0 0 0 0 112 0 0 0 0 0 0 1 0 0 - Hepaticae 0 0 0 Oi l 011 0 0 0 Oi l 011 011 ? ? ? 0 0 1 ? 0 0 ? '° Musci 011 011 0 0 0 0 0 0 011 011 011 ? ? ? 0 011 1 ? 0 0 ? '°.j>. J. HASEGAWA : New classification of Anthocerotae 25

Characters 1-8, which were referable to Mishler and Churchill ( 1985) and Vaughn et al. (1992), were involved as defining features of the Anthocerotae. 9. Apical cell of thallus cuneate with four cutting faces (O); hemidiscoid with three cutting faces (1). Gametophytes of the Anthocerotae, except for Dendroceros and Apoceros, have a wedge-shaped or cuneate apical cell. Dendroceros and Apoceros whose gametophytes are differentiated into distinct costae and laminae have a hemi­ discoid apical cell (Renzaglia 1978). In Notoceros the gametophyte also has distinct costae and laminae, but the form of the apical cell is unknown. 10. Thallus ecostate (O); costate with unistratose laminae (1). As mentioned under char­ acter 9, some taxa have the thallus differentiated into costae and laminae. In other taxa, thalli are more or Jess thick at the middle and thin out marginally. 11 . Cells of thallus without trigones (O); with triangular or quadrate trigones (thallus often perforated there)(!). In Dendroceros andApoceros cells ofthallus-laminae usu­ ally have triangular or quadrate trigones, but in other taxa no collenchymatous cells are present in the thallus. 12. Thallus with stoma-like clefts only on ventral side (O); both on ventral and dorsal sides (I). In the Anthocerotae, the so-called stoma-like clefts universally occur on the ventral surface of thalli. Among the taxa whose thalli are costate, Dendroceros and Apoceros have the stoma-like clefts also on the dorsal side but Notoceros has them only on the ventral side (Campbell 1986). 13 . Thallus solid (O); cavernous (1). In some taxa thalli have schizogenous internal cavi­ ties, but others have simply organized, solid thalli. 14. Thallus without tubers (O); with pedicellated or not-pedicellated tubers on vental sur­ faces or margins (1). The so-called tubers, which are sessil or stalked subspherical structures (or hump- or hill-like outgrowths) consisting of compact tissues, occur on the vental surface and the margin of thalli in Phaeoceros. In other taxa no tubers are found. 15 . Thallus without gemmae (0); with detachable spherical to ovoid gemmae (1). Some species of Anthoceros and Folioceros bear numerous, spherical to ovoid, multicellular gemmae on the dorsal surface and the margin of thalli. No gemmae are known from other taxa. 16. More than one antheridium in each cavity (O); a single antheridium in each cavity (1). In Anthoceros and Folioceros the number of antheridia often exceed twenty on ac­ count of the proliferation of secondary antheridia from the bases of the primary an­ theridia. In Phaeoceros the number of antheridia in a cavity is 2 to 8, and in Dendro­ ceros, Apoceros, Megaceros, Notoceros a single antheridium is developed in a cavity. In Leiosporoceros and Hattorioceros no androecia have been known. 17. Jacket of antheridial body composed of irregularly arranged small cells (O); com­ posed of four-tiered cells ( l ). In Anthoceros and Fo/ioceros jackets of antheridial bod­ ies are composed of four-tiered cells which are uniformly narrow-rectangular and regularly arranged side by side, and open by separation of cells of the uppermost tier. In other taxa except for Leiosporoceros and Hattorioceros in which no antheridia have been found, antheridial bodies are composed of irregularly arranged various- 26 J. Hattori Bot. Lab. No. 76 I 9 9 4

shaped small cells, and open irregularly. 18. Capsule erect (O); horizontal in position (I). In Notothylas, capsules are usually situ­ ated at the thallus-margin, and are parallel to or slightly ascending from dorsal sur­ faces ofthalli. In all other taxa capsules are erect on the thallus surface. 19. Capsule projecting from involucre (O); enclosed by involucre (I). In Notothylas, cap­ sules are wholly enclosed by an involucre even at maturity, and are exposed by decay­ ing of the involucre. In other taxa capsules project far from the involucre at the matu­ rity [even in some species of Anthoceros and Phaeoceros which have very short cap­ sules (less than 1 cm long), a considerable part of capsules protrudes over the involu­ cre ]. 20. Capsule ruptured irregularly or indehiscent (O); ruptured into two valves (l ). As a rule, capsules of the Anthocerotae have special suture lines on the wall and are dehis­ cent along these lines, but in some species of Notothylas capsules have no device for dehiscence and are irregularly ruptured or indehiscent. 21. Meristematic activity in sporophyte limited (O); persisting (I). In the Anthocerotae an intercalary meristem is present at the basal part of capsule, and the activity of the meristem is persisting in all taxa except for Notothylas in which the meristem acts for a limited short period. 22. Columella poorly developed or absent (O); well-developed (1). With the exception of Notothy/as the Anthocerotae always have well-developed columellae (normally 16 cells in cross-section but sometimes more massive) in the capsule. In Notothylas, some species have well developed columellae (N indica), but in most species col­ umellae are poorly developed (N orbiculareis, N temperata, N breutelii) or absent (N javanica, N levierii) (Pande 1934, Hassel de Menendez 1976, Hasegawa 1979). 23 . Sporogenous cells derived from endothecium (O); from amphithecium (1). As a rule sporogenous tissues of the Anthocerotae are derived from the inner layer of am­ phithecium, but according to Lang (1907) and Pande (1934) in some species of No­ tothylas the sporogenous tissue arises from the endothecium. 24. Synchronous spore production (O); continuous spore production (I). In Notothylas all spores in a capsule mature simultanously, but in other taxa spores are produced con­ tinuously and maturation of spores develops acropetally. 25. Spore-tetrad tetrahedral (O); bilateral (1). Leiosporoceros is the only taxon that has the isobilateral type of spore-tetrads. Those of the remaining Anthocerotae are of the tetrahedral type (Hasegawa 1988). In Hattorioceros details of spore-tetrads are un­ known. 26. Spore globose to tetrahedral (O); ovoid to kidney-shaped (1); amorphous (2). Spores derived from the tetrahedral type of spore-tetrads are globose to tetrahedral. Spores of Leiosporoceros which has the isobilateral type of spore-tetrads are ovoid to kidney­ shaped. Spores of Hattorioceros are basically spheroidal to ovoid, but exhibit a wide variation (1994). 27. Spore without equatorial girdle (O); with it (1). In spores of Notothylas and Phaeo­ ceros the so-called equatorial girdle is seen with optical microscope, but it is not seen in other taxa (Hassel de Menendez 1988). J. HASEGAWA: New classification of Anthocerotae 27

28. Spore surface ornamented with various formed outgrowths (O); smooth (1). In Leiosporoceros spores are almost completely smooth (no descernible indication of surface ornamentation is observed even under the highest magnification of optical microscope), but in other taxa spores have more or less rugged surfaces which ehxibit various ornamentations (Hasegawa 1988). 29. Spore surface not canaliculate-striate (O); canaliculate-striate (I). In Hattorioceros spore surfaces are canaliculate with straight to sinuous muri which merge at the apex into a swollen or auriculate protuberance. This type of surface-ornamentation of spores is known from a few species of the pteridophytes, but not from the bryophytes (Hasegawa 1994). 30. Spore medium-sized to large (O); extremely small (less than 22 JLm) (I). In the An­ thocerotae the size of spores is usually more than 25 JLm in diam. (exceptionally, Megaceros flagellaris has smaller spores which are only 20 JLm in diam., but the size of spores in this species varies greatly and often exceeds 30 JLm in diam.). In Leiosporoceros and Hattorioceros spores are mostly 19- 22 JLm long and 14-18 JLm long, respectively, and rarely exceed 22 JLm long (Hasegawa 1988). 31. Spore unicellular (exosporic germination) (O); multicellular (endosporic germination) (1 ). In Dendroceros and Apoceros spores germinate within the spore coat and are multicellular at the time of release, but in other taxa spores are unicellular at maturity. 32. Spore without chloroplast (O); with chloroplast (1). Dendroceros, Apoceros, Megaceros and Notoceros have chlorophyllous spores, but in other taxa mature spores have no chloroplast within (Schuster 1992). Exceptionally, in Folioceros fuciformis spores have a chloroplast. 33. Spore colourless or grayish (O); yellow (I); dark brown to black (2). Three different types of spore wall colouration are recognized in the Anthocerotae. Spores of Megaceros and Dendroceros often have been described as greenish because of the presence of chloroplasts within, but the wall is actually colourless or dull gray. 34. Pseudoelater unicellular, more or less rounded, often poorly developed or absent (O); multicellular, rectangular to elongated rectangular or fusiform, always well-developed (1). In Notothylas pseudoelaters are often poorly developed and decomposed in a sin­ gle cell, and sometimes totally lacking (Hasegawa 1979). In other taxa they are al­ ways well-developed and multicellular (Leiosporoceros has unicellular pseudoelaters which are always well-developed and fusiform) (Hasegawa 1986). 35 . Pseudoelater thin-walled (O); strongly thick-walled (1); unispirally thick-walled (2). Folioceros and Leiosporoceros have strongly thick-walled pseudoelaters which have no spiral thickening bands (Hassel de Menendez 1986, Hasegawa 1986). Four taxa in­ cluding Dendroceros and Megaceros have unispirally thick-walled pseudoelaters, and in the remaining taxa they are thin-walled and sometimes have irregular bands of thickenings on their wall. 36. Foot of sporophyte without palisade layer of haustorial cells (O); with it (1). The structure of the foot in Anthoceros and Folioceros is different from that in other taxa in that the foot consists of fewer cells and has a distinct palisade cell layer at the bot­ tom (Vaughn and Hasegawa 1993). 28 J. Hattori Bot. Lab. No. 76 I 9 9 4

37. Sporophyte without stomata (O); with them (1). Anthoceros, Folioceros, Phaeoceros, Leiosporoceros and Hattorioceros have stomata on their capsule walls, but the re­ maining tax.a have no stomata. Exceptionally Folioceros incurvus has no stomata (Bharadwaj 1972). 38. Chloroplast one per cell, with pyrenoid (O); more than one per cell, without pyrenoid (1). In most anthocerote genera, a single large chloroplast with a pyrenoid is present in each cell of the thallus. These features are characteristic of green algae but are not known from land plants except for the Anthocerotae (Vaughn et al. 1992). The single large chloroplast with a pyrenoid is, consequently, considered plesiomorphic. 39. Pyrenoid separated only by stroma lamella (O); widely separated by both granal and storoma lamellae as well as stroma (1). The pyrenoid of the Anthocerotae is different from that of green algae in that the pyrenoid region is traversed by thylakoids (both stacked and unstacked) and areas of non-pyrenoid stroma (most green algae have the "unit" pyrenoid). In Notothylas the pyrenoid is separated only by stroma lamellae like that of Coleochaete which is considered to be the extant green algal genus most like the progenitor of land plants (Vaughn et al. 1992). Judging from these facts, such a pyrenoid structure as found in Notothylas is considered plesiomorphic. 40. Ability to biosynthesize megacerotonic acid absent(O); present (1). The Anthocerotae biosynthesize several unusual phenolic compounds belonging to the so-called lignans which, besides the Anthocerotae, are known only from seed plants (Takeda et al. 1990). Among them megacerotonic acid is biosynthesized by most taxa except An­ thoceros and Folioceros (Hasegawa et al 1992). 41. Ability to biosynthesize rosmarinic acid absent (O); present (1 ). Rosmarinic acid, a phenolic compound widely distributed in phanerogams, occurs in the Anthocerotae (Takeda et al. 1990). Anthoceros and Folioceros, which biosynthesize no megacero­ tonic acid, contain a large amount of rosmarinic acid (Hasegawa et al. 1992). 42. Pseudoelater with yellowish autofluoresent substance on the wall (O); without it (1). Under the fluorescence microscope (under ultra-violet illumination with blue exciter filter), pseudoelaters fluoresce yellowish-brown due to the presence of an autofluore­ cent substance on their wall (in Anthoceros and Folioceros no fluorescence is detected on their pseudoelaters) (Hasegawa 1990).

Results and Discussion The analysis resulted in six most parsimonious trees of 51 steps in length (consistency index =0.824). In all these six trees, the first branch contains only a single genus Notothy­ /as, and the other major branch contains all the remaining taxa, which dichotomizes into the Anthoceros - Phaeoceros clade with five taxa and the Dendroceros - Megaceros clade with four taxa. The differences among the six equal length trees are in the final resolution of these two clades. One of these six trees, which fits most closely to the classification arbi­ trarily postulated as most comprehensive by the author, is shown in Fig. 1, and the strict consensus tree calculated from these six trees is in Fig. 2. As shown in Fig. 1 and Fig. 2, the most parsimonious cladogram for the Anthocerotae has Notothylas as the sister group to the remaining anthocerote taxa. These cladograms in- J. HASEGAWA: New classification of Anthocerotae 29

30 Leiosporoceros ....._ ____ Hattorioceros 27' 37 ...___,.__ ___~ Phaeoceros ---. Anthoceros 35'

16 32 33•352 1-1 23 40

13•

18 19 21·

Fig. l. One of the most parsimonious cladograms of the Anthocerotae. Solid bars in­ dicate both unique and homoplasious apomorphies (homoplasious characters are indicated with an asterisk). Open bars indicate apomorphies with the reversal. In the characters which have three character states, each state is indicated by adding 0, l or 2 to the characer number. 1 2 1 In character 33, 33 and 33 are put in parentheses, because if the yellow spore-wall (33 ) is 2 assumed to be plesiomorphic, the brown to black spore-wall (33 ) would be apomorphic, and vice versa. In both cases the colourless spore-wall (33°) is apomorphic.

Anthoceros

---i Folioceros

Lelosporoceros

-----1 Hattorioceros

Phaeoceros

Dendroceros

r--4 Apoceros

Megaceros

Notoceros

Notothylas

Fig. 2. The strict consensus tree of the Anthocerotae calculated from six most parsi­ monious cladograms. 30 J. Hattori Bot. Lab. No. 76 I 9 9 4 volve characters 21 - 22, 24, 34 and 39 as synapomorphies for the clade containing all the taxa except for Notothylas. In accordance with Mishler and Churchill (1985), most charac­ ter states concerning the sporophyte found in Notothylas are considered as plesiomorphic, but characters 18 and 19 (the horizontally located capsule which is wholly enclosed by the involucre) are regarded as derived features of Notothylas. Thus, the cladistic relationships of Notothylas to other taxa is apparently well supported by the characters mentioned above. Hassel de Menendez ( 1988), however, presented a cladograrn in which Notothylas and Phaeoceros were combined in the same clade supported by a single character 27 (the pres­ ence of an equatorial girdle in the spores). Such a cladograrn is required to assume charac­ ters 21 - 22, 24, 34 and 39 as being reversed in Notothylas, and is surely less parsimonious than that presented here. Character 27 should be considered to have evolved independently in Notothylas and Phaeoceros. Of the two clades branching at the next node, the one including Dendroceros, Apo­ ceros, Megaceros and Notoceros is supported by four synapomorphies (characters 16, 32, 2 33° and 35 : the single antheridium in each antheridial cavity, the spores with a chloroplast, the colourless spore-wall and the pseudoelaters with unispiral thickening bands), all of which were used for defining the Dendroceros - Megaceros clade by Hassel de Menendez (1988). This clade was also presented by Mishler and Curchill (1985 ~, and considered to be well-defined. On the contrary, the other clade including Anthoceros, Folioceros, Leiosporoceros, Hattorioceros and Phaeoceros is supported only by a single synapomorphy (character 37: the presence of stomata in the sporophyte). Mishler and Churchill (1984, 1985) considered the presence of stomata in the sporophyte as a synapomorphic character shared by the An­ thocerotae, the Musci and the Tracheophyta, and the absence of stomata in some antho­ cerote taxa as a secondary reversal. In their cladistic analysis, consequently, Anthoceros and Phaeoceros had no synapomorphies, which resulted in the trichotomy at this node in their cladogram. In the present study, the stomata in the sporophyte are considered to have evolved independently in the Anthocerotae and the Musci-Tracheophyta clade, and thus Anthoceros, Phaeoceros and other taxa with stomata in the sporophyte form a branch which is supported by this character. Hassel de Menendez (1988) also regarded the absence of stomata as a derived feature. However, I prefer the hypothesis of the independent evolu­ tion of stomata to that of the secondary reduction (in the former case one step is required for this character to construct the most parsimonious tree, but in the latter, two steps are needed). As shown in the strict consensus tree (Fig 2), each branch of the primary dichotomy in the clade containing all taxa excluding Notothylas, is umesolved yet and each has three pa­ raphyletic clades. Among these six terminal clades, the Anthoceros - Folioceros clade and the Dendroceros - Megaceros clade seem to be well-defined; they are supported by seven and five apomorphic characters, respectively. Hassel de Menendez (1988), however, treated Anthoceros and Folioceros as paraphyletic. In her analysis, they shared two apomorphic characters (the jacket of antheridial body composed of four-tiered cells; the cavernous thal­ lus), but these two characters were assumed to have evolved independently in each genus. In the present study, five more characters besides the two characters mentioned by Hassel J. HASEGAWA: New classification of Anthocerotae 31 de Menendez (1988) are incorporated in the synapomorphy for Anthoceros and Fo/ioceros. They include some chemical and ultrastructural characters recently revealed (characters 36, 40, 41, 42), among which characters 36, 41 and 42 (the palisade cell layer of sporophyte foot; the ability to biosynthesize rosmarinic acid; the presence of an autofluorescent sub­ stance on the pseudoelater) are unique apomorphies for Anthoceros and Fo/ioceros. The ac­ ceptance of the paraphyletic relationship between Anthoceros and Folioceros would force us to assume the parallel evolution of all these apomorphic characters, and consequently would result in far less parsimonious cladograms. Anthoceros and Folioceros should be considered to be united by these synapomorphies. As mentioned above, the terminal branch with Dendroceros and Apoceros is well-de­ fined by five apomorphic characters, among which characters 9, 12, (the hemidiscoid apical cell; the stoma-like clefts on the dorsal side of thalli) are unique to them. Apoceros is a tax on which was established as a subgenus of Dendroceros by Schuster ( 1987). According to Schuster (1987), it is distinguished from Dendroceros by the cavernous thallus-costae and the reticulated thallus-laminae with large perforations. However, the thallus-laminae, which form the reticulated structure due to the presence of large perforations, also occurs in some species of Dendroceros whose costae have no cavities within. Accordingly, the only character to distinguish Apoceros from Dendroceros is the cavernous thallus-costae (this character is assumed to be the apomorphy with parallelism). Furthermore, Dendro­ ceros s. str. has no convincing autapomorphies, and thus the distinction between Dendro­ ceros and Apoceros does not seem so clear. Here I treat Apoceros as an autonomous taxon defined by character 13* (the cavernous thallus-costae), but the naturalness of this taxon re­ quirs reinforcement with additional apomorphic characters. Leiosporoceros and Hattorioceros are united by a single apomorphic character (30, the extremely small spore), but I consider the validity of this character debatable, and should be confirmed by further comparative studies, because the small spore character seems to occur sporadically in various groups. I hesitate to accept the clade supported only by such a poorly-defined character. I prefer to regard Leiosporoceros and Hattorioceros as paraphyletic It should be mentioned that Leiosporoceros and Hattorioceros were regarded as a sub­ genus of Phaeoceros (Hasegawa 1988, 1994; Schuster 1992). In this classifiction the genus Phaeoceros was defined by the following characters: (1) the solid thallus, (2) the jacket of the antheridial body composed of irregularly arranged small cells, (3) the stomata in the sporophyte, (4) the yellow spores and (5) the gametophytic cells with a single chloroplast. In the cladistic analysis presented here, however, the only character that is shared by these three taxa is the yellow spores, and the other three characters except for the stomata in the sporophyte which is the synapomorphy for the Anthoceros - Phaeoceros clade, are regard­ ed as plesiomorphic characters. Furthermore, if the hypothesis that the brown to black 2 1 spores (33 ) evolved from the yellow spores (33 ) is accepted, the genus Phaeoceros as de­ fined by Hasegawa (1988, 1944) and Schuster (1992) would have no synapomorpies, and consequently Phaeoceros and Leiosporoceros would become paraphyletic. Taking the re­ sult of the cladistic analysis and the problems discussed above into consideration, I consid­ er it reasonable to treat Leiosporoceros and Hattorioceros each as a taxon coordinate in 32 J. Hattori Bot. Lab. No. 76 I 9 9 4 rank with Phaeoceros. Megaceros and Notoceros are united into the terminal clade in some of the six most parsimonious cladograms. This clade is supported only by a single apomorphic character 38 (the plural chloroplasts without the pyrenoid in a cell). As suggested by the strict con­ sensus tree (Fig. 2), however, Notoceros is also combined with the Dendroceros and Apoc­ eros clade in some of the most parsimonious cladograms; they share a single apomorphic character 10 (the costate thallus), which is not known from other anthocerote taxa. In order to resolve the relationships of Notoceros to Megaceros and Dendroceros, additional infor­ mative characters are required. At present I consider that Notoceros should be classified as a taxon coordinate in rank with Megaceros and Dendroceros.

CLASSIFICATION OF THE ANTHOCEROTAE On the basis of the cladistic analysis presented here, the classification of the Antho­ cerotae is revised as follows:

Anthocerotales Family Notothyladaceae (Milde) K. Muell. ex Prosk., Phytomorphology 10 : 10 (1960). Genus Notothylas Sull., Amer. J. Sci. Arts: 74 (1846). Family Anthocerotaceae Dum., Annal. des fam. des pl. 69 (1829). Subfamily. Anthocerotoideae Genus Anthoceros L. emend. Prosk., Bull. Torrey Bot. Club 78: 346 (1951 ). Subgenus Anthoceros Subgenus Folioceros (Bharadw.) Schust., Phytologia 63(3): 200 (1987). Genus Leiosporoceros Hassel, J. Bryol. 14: 255 (1986). Genus Hattorioceros (Haseg.) Haseg. stat. nov. [Basionym: Phaeoceros subgen. Hattorioceros Haseg., J. Hattori Bot. Lab. 75 : 272 (1994)). Genus Phaeoceros Prosk., Bull. Torrey Bot. Club 78: 346 (1951). Subfamily Dendrocerotoideae Schust., Phytologia 63(3): 200 (1987). Genus Dendroceros Nees in Gott. et al., Syn. Hep.: 579 (1846). Subgenus Dendroceros Subgenus Apoceros Schust., Phytologia 63(3): 200 ( 1987). Genus Megaceros Campbell, Ann. Bot. 21: 484 ( 1907). Genus Notoceros (Schust.) Haseg. stat. nov. [Basionym: Megaceros subgen. No­ toceros Schust., Phytologia 63(3): 200 (1987)).

ACKNOWLEDGEMENTS I am much indebted to Mr. T. Fukuhara (Kyoto University) for his kind discussion and aid in performing the cladistic analysis. I also thank Dr. N. Kitagawa (Nara University of Education) for his valuable suggestions, and Dr. J. J. Engel (Field Museum of Natural His­ tory, Chicago) for correcting the English.

LITERATURE CITED Bharadwaj, D. C. 1972. On some Asian and African species of Folioceros Bharadwaj. Geophytology J. HASEGAWA: New classification of Anthocerotae 33

2(1): 74-89. Campbell, E. 0. 1986. Notes on some Anthocerotae of New Zealand (5). Tuatara 28(2): 83- 94. Hasegawa, J. 1979. Taxonomical studies on Asian Anthocerotae I. Acta Phytotax. Geobot. 30(1- 3): 15- 30. Hasegawa, J. 1986. Anthoceros dussii Steph. (Anthocerotae) and its isobilateral spore tetrads. Hikobia 9: 357- 360. Hasegawa, J. 1988. A proposal for a new system of the Anthocerotae, with a revision of the genera. J. Hattori Bot. Lab. 64: 87- 95. Hasegawa, J. 1990. Autofluorescence of elaters in the Anthocerotae and its taxonomic significance. J. Hiraoka Envir. Sci. Lab. 3: 1-9. Hasegawa, J. 1994. A remarkable new species of Phaeoceros (Anthocerotae) with canaliculate-striate spore surface. J. Hattori Bot. Lab. 75: 267- 273. Hasegawa, J. , M. Shinozaki and R. Takeda. 1992. Phenolic constituent of the Anthocerotae and its taxonomic significance. Amer. J. Bot. 79(6), Suppl. (Abstract): 42. Hassel de Menendez, G. G. 1976. Taxonomic problems and progress in the study of the Hepaticae. J. Hattori Bot. Lab. 41: 19- 36. Hassel de Menendez, G. G. 1986. Leiosporoceros Hassel n. gen. and Leiosporocerotaceae Hassel n. Fam. of Anthocerotopsida. J. Bryol. 14: 255- 259. Hassel de Menendez, G. G. 1988. A proposal for a new classification of the genera within the Antho­ cerotophyta. J. Hattori Bot. Lab. 64: 71- 86. Lang, W H. 1907. On the sporogonium of Notothylas. Ann. Bot. 12: 251 - 256. Mishler, B. D. and S. P. Churchill. 1984. A cladistic approach to the phylogeny of the bryophytes. Brittonia 36: 406-424. Mishler, B D. and S. P. Churchill. 1985. Transition to land flora: phylogenetic relationships of the green algae and bryophytes. Cladistics I: 305- 328. Mishler, B. D., P.H. Thrall, J. S. Hopple, E. De Luna and R. Vilgalys. 1992. A molecular approach to the phylogeny of bryophytes: cladistic analysis of chloroplast-encoded 16S and 23S ribosomal RNA genes. Bryologist 95 : 172- 180. Pande, S. K. 1934. On the morphology of Notothylas /evieri Schiff., Ms. Proc. Ind. Acad. Sci., B, 1(5): 205- 217, pls.23- 32. Proskauer, J. 1951. Studies on Anthocerotales. III. Bull. Torrey Bot. Club 78: 331 - 349. Renzaglia, K. S. 1978. A comparative morphology and developmental anatomy of the Anthoceroto­ phyta. J. Hattori Bot. Lab. 44: 31 - 90. Schuster, R. M. 1987. Preliminary studies on Anthocerotae. Phytologia 63(3); 193- 201. Schuster, R. M. 1992. The Hepaticae and Anthocerotae of North America. Vol. VI. Field Museum of Natural History. Chicago. Swofford, D. L. 1993 . PAUP: phylogenetic analysis using parsimony, version 3.1. Computer program distributed by the Illinois Natural History survey, Champaign. Takeda, R., J. Hasegawa and M. Shinozaki. 1990. The first isolation of lignans, megacerotonic acid and anthocerotonic acid, from non-vascular plants, Anthocerotae (Hornworts). Tetrahedron Lett. 31 (29): 4159-4162. Vaughn, K. C., R. Ligrone, H. A. Owen, J. Hasegawa, E. 0. Campbell, K. S. Renzaglia and J. Monge­ Najera. 1992. The anthocerote chloroplasts: a review. New Phytol. 120: 169-190. Vaughn, K. C. and J. Hasegawa. 1993. Ultrastructural characteristics of the placental region of Fo/io­ ceros and their taxonomic significance. Bryologist 96(1): 112- 121. Waters, D. A., M. A. Buchheim, R. A. Dewey and R. L. Chapman. 1992. Preliminary inferences of 34 J. Hattori Bot. Lab. No. 76 I 9 9 4

the phylogeny of bryophytes from nuclear-encoded ribosomal RNA sequences. Amer. J. Bot. 79(4); 459-466.

ADDENDUM When I was preparing a manuscript for this paper, a paper about the cladistic analysis of the An­ thocerotae was published by Hyvonen and Piippo (1993)"', in which they analysed the phylogenetic relationships among nine anthocerote genera in a very rigorous manner and revised the classification of the Anthocerotae based on their cladistic analysis. I basically agree with their ideas on which their analysis is based, and consequently the most parsimonious cladogram presented here is similar to a great extent to that of theirs. Some notable diferences are, however, present between them; they in­ clude the relationships of Folioceros and Leiosporoceros to the other taxa. rn Hyvonen and Piippo (1993) Folioceros and Leiosporoceros are united in the same clade, which is the sister group of the lineage with Anthoceros. According to them, the synapomorphies supporting the Folioceros - Leiosporoceros clade are (1) the crenulate or lamellate surface of the involucre, (2) the thick-walled pseudoelater, (3) the pseudoelater with acute apices and (4) the yellow or light brown pseudoelater. Among these four characters, (I) and (4) are not included in the analysis presented here, because the involucre with crenulate-lamellate surfaces sporadically occurs in almost all anthocerote taxa, and is considered to have little taxonomic value at this level of classification. As to the colour of the pseu­ doelaters, I consider that it may be informative, but ifthe character is accepted, the opposite character state should be given to Leiosporoceros, because the pseudoelaters of Leiosporoceros are light brown and rather close to those of Phaeoceros which show a wide range of variation in their colour (from pale yellow to brownish yellow), while Folioceros are rather similar to Anthoceros in having dark brown pseudoelaters. Thus, I consider it problematic to regard this character as a synapomorphy (sec­ ondary reverse from the derived state in their analysis) uniting Folioceros and Leiosporoceros. The only character that is informative to analyse the relaltionships of these taxa seems to be the thick­ walled pseudoelaters with acute apices. Although Hyvonen and Piippo (1993) treated this character as comprising two independent characters i.e. the thick-walled pseudoelater and its acute apices, in the present study they are treated as a single character, because these two characters are considered to represent two different faces of the essentially same nature of the pseudoelaters consisiting of the strongly thick-walled, narrow-elongated cells. If these two features are treated as independent charac­ ters, one more homoplasy would be required to constract the most parsimonious tree presented here. Even in such a case, however, the cladogram in which Folioceros is united with Anthoceros, is more parsimonious than the one with a terminal branch consisiting of Folioceros and Leiosporoceros. Fur­ thermore, I consider that the condition of pseudoelaters found in Leiosporoceros might have the dif­ ferent origin from that of Folioceros, as suggested by the fact that in Leiosporoceros pseudoelaters are unicelluar and the wall of the pseudoelater cells has smooth inner surfaces, whereas in Folioceros they are multicellular and their wall has uneven inner surfaces. All these facts seem to suggest that Folioceros is more closely related to Anthoceros than to Leiosporoceros.

• Hyvonen, J. and S. Piippo. 1993. Cladistic analysis of the homworts (Anthocerotophyta). J. Hat­ tori Bot. Lab. 74: 105- 119.