2. Hymenasplenium Hayata, Bot. Mag. (Tokyo). 41: 712. 1927.
膜叶铁角蕨属 mo ye tie jiao jue shu
Boniniella Hayata.
Herbs, epilithic, epiphytic, or terrestrial. Rhizome dorsiventral, thin, up to ca. 5 mm in diam., widely creeping, with clathrate scales. Fronds herbaceous, remote; stipe usually shiny and castaneous to dark purplish or black, rarely grayish green, semiterete abaxially, sulcate adaxially; lamina 1-pinnate, rarely simple; rachis sulcate adaxially, basiscopic margin of pinnae often decurrent and forming narrow abaxial wings on rachis; pinnae asymmetrical with basal basiscopic part cut away and becoming dimidiate; basiscopic margin entire, acroscopic margin crenate, undulate, or serrate, sometimes with retuse teeth. Veins free, rarely anastomosing, anadromously branching, becoming simple toward pinna apex, not reaching margin, 1 to several basal basiscopic veins lacking. Sori solitary, rarely double, linear to subelliptic, indusiate; indusium thinly membranous to papery, free margin entire to erose; stalks of sporangia long uniseriate, annuli of 20-28 hardened cells. Spores bilateral, elliptic, perispore elaborate, exospore smooth; in sexual plants 64 spores per sporangium. Plants sexual or agamosporous. x = (36), 38, 39.
More than 30 species: pantropical; 18 species complexes (eight endemic) in China.
Hymenasplenium was described by Hayata for Asplenium unilaterale Lamarck and based on the peculiar vascular system of its dorsiventral rhizome. It was later reduced to a section of Asplenium by Iwatsuki (Acta Phytotax. Geobot. 27: 44. 1975). Recent molecular phylogenetic analyses show that Hymenasplenium is the most basally diverged genus in Aspleniaceae, not closely related to any of the other members (Murakami, J. Plant Res. 108: 257-268. 1995).
Chromosome numbers of Hymenasplenium were first reported, by Manton and Sledge (Philos. Trans., Ser. B, 238: 138, 167. 1954) in H. unilaterale (Lamarck) Hayata s.l. from Sri Lanka, to be n = 80 and 2n = ca. 158. However, Mitui et al. (Amer. J. Bot. 76: 1691. 1989) showed that the basic chromosome number of Hymenasplenium is x = 39, with one exception being x = 38 in H. subnormale. Apomictic reproduction was discovered in three unrelated Asian species, indicating that such reproductive mode evolved at least three times in the Asian group.
In a cytological study of Asplenium cardiophyllum s.l. (Hymenasplenium ikenoi, see p. 309), Kato et al. (Bot. Mag. (Tokyo) 103: 461-468. 1990) found that its basic number is also x = 39 (n = 78) and concluded that "A. cardiophyllum" should be included in Hymenasplenium, although this species has cordate simple leaves quite different from those of the other members of Hymenasplenium. The close relation between "A. cardiophyllum" and Hymenasplenium is supported by molecular studies (Murakami, loc. cit.: 267).
SW China (Yunnan) is the center of diversity of Hymenasplenium. The reports of morphologically intermediate or "transitional" forms as well as the existence of different cytotypes (ancestral sexual diploids and tetraploids next to agamosporous taxa) in this area suggest that Hymenasplenium is still in an active state of diversification. Reticulate evolution is not yet reported in this genus but may explain some of the intermediate forms and the difficulty in describing them as clearly and morphologically separate taxa. Because the distribution and relationships of the Asian "unilaterale" group are not well understood, the present treatment probably underestimates the number of species and reflects our limited knowledge of this group.
The name (Hymen)asplenium unilaterale has been used erroneously in SE Asia for several other taxa, such as H. apogamum, H. hondoense, and H. murakami-hatanakae. True H. unilaterale differs in its more strongly dimidiate pinnae, lacking (5-)7 or 8 basal basiscopic veins, its teeth that are not retuse, and the dark blackish rachis color that distinctly continues onto the costa abaxially. We have not so far found true H. unilaterale occurring in China. All Asian-Himalayan taxa require further study to discover their relationships with Chinese taxa. In particular, Hymenasplenium rivulare (Fraser-Jenkins) Viane & S. Y. Dong, comb. nov. (Basionym: Asplenium rivulare Fraser-Jenkins, Taiwania 53: 190. 2008, based on A. unilaterale var. rivale Beddome, Handb. Ferns Brit. India, 153. 1883, not A. rivale Baker (1867); "A. hindusthanensis" [sic!] Bir, Fern Gaz. 14: 309. 1994, not validly published, Melbourne Code, Art. 41.5) from S India needs attention, as Beddomes drawing (Suppl. Ferns S. Ind. t. 356. 1876, under A. resectum var. rivale) shows both the venation pattern and the marginal retuse teeth with veins ending just below the minute notch, which are characteristic for the E Asian H. retusulum-H. latidens group.