Notes on Early Land Plants Today. 36. Generic treatment of Lophocoleaceae (Marchantiophyta)

Recent molecular phylogenetic studies on Lophocoleaceae have recovered some well-supported monophyletic lineages within Chiloscyphus s. lat. that merit recognition at the generic level. As a consequence, Chiloscyphus is herein circumscribed to include only Chiloscyphus subgen. Chiloscyphus, Chiloscyphus subgen. Connati is elevated to generic rank to which the new name Cryptolophocolea is applied, and Lophocolea, Pachyglossa and Clasmatocolea are recognized as distinct genera. Tetracymbaliella is transferred to the Brevianthaceae, and a synopsis of the genera currently recognized in Lophocoleaceae is provided. Current knowledge of the family Lophocoleaceae During the last 50 years Lophocoleaceae Vanden Berghen (1956: 208) have (mostly as Geocalycaceae Klinggräff 1858: 34) mainly been circumscribed in a very broad sense. However, Hentschel et al. (2006a) showed that Lophocoleaceae is well separated from Geocalycaceae and Harpanthaceae Arnell (1928: 147), but remains a large and diverse family. Lophocoleaceae is a family with a turbulent history and many taxa have been moved back and forth among genera. Engel & Schuster (1985) introduced a very broad concept of the genus Chiloscyphus including in it all species generally referred to Chiloscyphus and Lophocolea at that time. This concept was not adopted by several authors, including Grolle (1995), Paton (1999), Gradstein & Costa (2003), and Crandall-Stotler et al. (2009), but has been more widely accepted since molecular studies showed that Chiloscyphus s. str. nests within the traditionally circumscribed Lophocolea (He-Nygrén & Piippo 2003, Hentschel et al. 2006a). However, several later studies (Hentschel et al. 2007, Glenny et al. 2009, Engel et al. 2010, Engel & He 2010) have shown that unless the genus Chiloscyphus s. lat. is split into additional genera, Pachyglossa, Leptoscyphus and Clasmatocolea would also have to be included within it. But the same studies also show that there are several well-supported monophyletic lineages within the Chiloscyphus s. lat. clade that merit recognition at the generic level. All molecular analyses resolve the taxa assigned to Chiloscyphus subg. Connati in a strongly supported clade that is sister to the rest of Chiloscyphus, or even to a combined Chiloscyphus-Leptoscyphus clade (Hentschel et al. 2007). Recognizing Chiloscyphus subg. Connati, as circumscribed by Engel (2010), at a genus level would allow the robustly supported clade of Leptoscyphus to be kept following Vanderpoorten & Long (2006) and Vanderpoorten et al. (2010), as discussed by Hentschel et al. (2007) and Crandall-Stotler et al. (2009). 36 Accepted by Jon Shaw: 10 Apr. 2013; published online in PDF: 1 May 2013 Licensed under a Creative Commons Attribution License http://creativecommons.org/licenses/by/3.0 Likewise, Lophocolea, which is resolved as the lineage labelled Chiloscyphus subg. Lophocolea (Dumortier 1831: 59) J.J.Engel et R.M.Schust. in Engel (1998: 42) by Hentschel et al. (2006b), can also be recognized as a distinct genus with moderate to strong support, especially since this fairly large clade includes the generitype of Lophocolea, but excludes that of Chiloscyphus (Crandall-Stotler et al. 2009). According to current molecular evidence (Hentschel et al. 2006a, 2006b, Glenny et al. 2009), the genus should comprise taxa previously assigned to Chiloscyphus subg. Lophocolea (excluding Chiloscyphus subg. Lophocolea sect. Bicornuti (Spruce 1885:424) Engel & Schuster (1985: 409), which was transferred to Chiloscyphus subg. Connati by Engel in 2010), Chiloscyphus subg. Microlophocolea (Spruce 1885: 426) Engel (1999: 22), Chiloscyphus subg. Fragilifoliae (Schuster 1978: 245) Engel & Schuster (1985: 409) and Chiloscyphus subg. Spinoscyphus Engel (2010: 1999). Of the remaining taxa included by Hentschel et al. (2007) in the “Chiloscyphus I” clade, Chiloscyphus s.str., has been consistently resolved as a robustly supported, monophyletic lineage, comprising Chiloscyphus polyanthos (Linnaeus 1753: 1131) Corda (1829: 651) (the generitype of Chiloscyphus) and Chiloscyphus pallescens (Hoffmann 1796: 87) Dumortier (1831: 67). Recognizing Chiloscyphus to include only Chiloscyphus subg. Chiloscyphus is consistent with Grolle's (1995) contention that Chiloscyphus is a small, strictly Laurasian genus that is morphologically isolated from the more speciose, broadly distributed elements of the Lophocoleaceae, like Lophocolea and Heteroscyphus. Disposition of the remaining taxa in the "Chiloscyphus I" clade is problematic. In Henschel et al. (2006b, 2007), Chiloscyphus subg. Notholophocolea (Schuster 1980: 183) Engel & Schuster (1985: 410) is robustly supported as a monophyletic lineage, with Pachyglossa tenacifolia (Hook.f. et Taylor in Hooker 1845: 152) Herzog & Grolle (1959: 153) moderately supported sister to it. In the analysis of Glenny et al. (2009), on the other hand, Chiloscyphus subg. Notholophocolea is nested in Lophocolea. These contrasting results may be due to the use of different loci and/or different taxa by the two groups. The analyses of Hentschel et al. (2006b, 2007) include two accessions of Chiloscyphus (Notholophocolea) gottscheoides (Bescherelle & Massalongo 1886: 631) Engel & Schuster (1985: 415) and a single accession of Chiloscyphus (Notholophocolea) austrigenus (Hooker & Taylor 1844: 466) Engel & Schuster (1985: 411) subsp. austrigenus from southern South America in an nrITS dataset, while that of Glenny et al. (2009) is based on 3 plastid loci and a single accession of Chiloscyphus (Notholophocolea) austrigenus subsp. okaritanus (Stephani 1906: 785) Engel (1992: 113) from New Zealand. Obviously, more extensive molecular studies of this complex are needed to resolve unambiguously the relationships of these geographically separated taxa. However, on the basis of morphological similarities (Schuster 1980) and the more extensive sampling of Hentschel et al. (2007) Chiloscyphus subg. Notholophocolea should be transferred to Pachyglossa, and Pachyglossa should continue to be recognized as a genus. A polyphyletic Clasmatocolea is also resolved within the "Chiloscyphus I" clade with most species currently included in molecular studies widely separated from Clasmatocolea vermicularis (Lehmann 1829: 361) Grolle (1960: 28), which is morphologically close to or conspecific with the generitype, Clasmatocolea heterostipa Spruce (1885: 441). These results are not surprising given the morphological heterogeneity of the six subgenera recognized to comprise the genus (Engel 1980). For example, Clasmatocolea rigens (Hooker & Taylor 1844: 461) Engel (1973: 156) (Clasmatocolea subg. Protoclasmatocolea Engel (1980:44)), not included in molecular studies, is morphologically very similar to species of the genus Lophocolea (s. str.), and it has been suggested that Clasmatocolea cucullistipula (Stephani 1900) Grolle (1960: 71) (Clasmatocolea subg. Plicaticalyx Engel (1980: 183)) should perhaps be separated at a generic level (Engel 1980). Since the molecular data of Hentschel et al. (2007), however, do not provide unambiguous resolution of the lineages within Clasmatocolea, it should be kept in its current form until the genus is more broadly sampled. On the basis of molecular evidence, three monotypic genera have been recently reduced to older taxa. Specifically, Amphilophocolea sciophila Schuster (2001: 98) of the monotypic genus Amphilophocolea Schuster (2001: 98) has been shown by Engel et al. (2010: 47) to be conspecific with Heteroscyphus knightii (Stephani 1908: 129) Grolle (1987: 251), Cyanolophocolea echinella (Gottsche et al. 1847: 703) Schuster (2001: 102) has been transferred to Heteroscyphus (Engel & He 2010: 155), and Physotheca autoica Engel & Phytotaxa 97 (2) © 2013 Magnolia Press • 37 GENERIC TREATMENT OF LOPHOCOLEACEAE Gradstein (2003:764) has been transferred to Leptoscyphus (Vanderpoorten et al. 2012: 252). Until representatives of the genera Bragginsella, Conoscyphus, Evansianthus, Hepatostolonophora, Leptoscyphopsis, Otoscyphus, Perdusenia, Pigafettoa, Stolonivector and Xenocephalozia are included in molecular studies, and their relationships to other genera of the Lophocoleaceae are resolved, they should remain as currently defined.. The generic status of Xenocephalozia is especially problematic. The genus was initially reduced by Grolle (1966) to Clasmatocolea, with affinities suggested to Clasmatocolea trachyopa (Hooker & Taylor 1844: 471) Grolle (1960: 73) and Clasmatocolea obvoluta (Hooker & Taylor 1845: 80) Grolle (1960: 72), taxa recognized by Engel (1980) to comprise Clasmatocolea subg. Lacerifolia Engel (1980: 50). Characters supporting the relationship of Xenocephalozia navicularis (Stephani 1915: 835) Schuster (1965: 25) to this subgenus of Clasmatocolea include the slightly armed and strongly adaxially concave leaves and relatively small underleaves, as also noted by Engel (1980). In Hentschel et al. (2007), Clasmatocolea trachyopa and Clasmatocolea obvoluta form a well-supported lineage within the "Chiloscyphus I" clade. The relationship of this lineage to Xenocephalozia and its status as a genus, however, cannot be determined without more extensive sampling. Many taxa that have been referred to Chiloscyphus in the past have not been included in molecular studies, so their phylogenetic affinities are unknown. None of them have been morphologically aligned with Chiloscyphus subg. Chiloscyphus, and should be excluded from Chiloscyphus s. str. as herein defined on the basis of morphology. Taxa included in Chiloscyphus subg. Parachiloscyphus Hässel (2000: 452) and Chiloscyphus subg. Eurychiloscyphus Hässel (2001: 38) show features in common with Leptoscyphus, where they were placed by Grolle (1962). Chiloscyphus subg. Septati Engel (2010: 125) is morphologically similar to Stolonivector and the two species in Chiloscyphus subg. Phaeochiloscyphus Engel & Schuster (1985: 409) have been previously aligned, respectively, with Leptoscyphus and Heteroscyphus (Hässel de Menéndez 1999). Recent studies indicate that some elements included in Lophocoleaceae should be removed from the family. Engel and Schuster (1985) reduced Tetracymbaliella to a subgenus of Heteroscyphus, but in several molecular studies it is clearly resolved as a genus distinct from Heteroscyphus (e.g., Glenny et al. 2009; Engel & He, 2010). The genus is placed as sister to Brevianthus in He-Nygrén et al. (2006) and the two genera are sister to Plagiochilaceae Müll.Frib. et Herzog in Müller (1956: 877) instead of Lophocoleaceae in the same study, but sister to Lophocoleaceae in He & Glenny (2010). Brevianthus is monotypic but of the 3-5 species of Tetracymbaliella, only Tetracymbaliella cymbalifera (Hook.f. et Taylor in Hooker 1846: 151) Grolle (1961: 50) (the type of the genus) has been studied using molecular methods. Although morphologically quite different from each other, these two genera are tentatively placed in Brevianthaceae, pending additional studies. The genus Deceptifrons is provisionally placed in Lophocoleaceae pending further studies. The genus Pseudolophocolea Schuster & Inoue in Schuster & Engel (1982: 71) was originally described as being closely related to Pedinophyllopsis Schuster & Inoue (1981: 311). The authors even speculated that the two genera might be regarded as subgenera of a single genus. Although both genera were placed in Geocalycaceae subfam. Leptoscyphoideae Schuster (1980: 267), an affinity with the plagiochiloid Pedinophyllum (Lindberg 1874: 366) Lindberg (1875: 504) was also noted. He-Nygrén & Piippo (2003) showed that Pedinophyllopsis is best placed in Plagiochilaceae, a hypothesis further supported by the molecular studies of Groth (2005). Based on the morphological similarities of Pseudolophocolea with Pedinophyllopsis and Pedinophyllum detailed in Schuster & Engel (1982), Pseudolophocolea is also placed in Plagiochilaceae.