Sematophyllaceae Broth.

Plants slender to robust, forming dense mats or tufts, often glossy. Stems creeping or ascending, irregularly to densely and pinnately branched, without a central strand, usually ± red. Stem and branch leaves similar or occasionally differentiated (as in Wijkia), crowded in many rows but sometimes complanate, often homomallous or secund, mostly symmetric, variously shaped but often ± ovate and often acuminate (rarely broad or rounded) at apex, mostly variably toothed, mostly unbordered. Laminal cells linear to rhombic, smooth or papillose; alar cells nearly always clearly differentiated, often inflated, vesiculose at extreme angles, less often ± quadrate. Costae short and double or lacking. Paraphyllia lacking. Pseudoparaphyllia often present and foliose.

Autoicous or dioicous. Perichaetia lateral and with the leaves differentiated. Setae elongate, mostly smooth but sometimes papillose; capsules mostly inclined to pendulous and asymmetric, rarely erect and symmetric (in non-N.Z. taxa), ovoid- or oblong-cylindric, smooth, mostly pale brown; exothecial cells often collenchymatous or subcollenchymatous; stomata restricted to neck; annulus often not differentiated; operculum conic at base, short- or long-rostrate. Peristome mostly double and hypnoid, pale. Calyptra cucullate or rarely mitrate, smooth. Spores spherical.

The Sematophyllaceae are a large and predominantly tropical and subtropical family with limited diversity in temperate regions. Brotherus (1925) recognised 36 genera distributed worldwide and placed them in four subfamilies.

The family is taxonomically difficult in most parts of its range and even the modest number of taxa occurring regionally pose some practical identification difficulties.

The generic limits in the Sematophyllaceae remain highly controversial and fluid. They are likely to become more settled only after large-scale monographic study, including molecular analysis, of this family. Hedenäs & Buck. W.R. (1999) performed a cladistic analysis of the genera traditionally placed in the family using morphological characters and recognised 35 genera (their Table 7), three of which they placed in a subfamily Wijkioideae. They suggested that these three genera (Wijkia, Acanthorrhynchium, and Trismegistia) are the "genera of Sematophyllaceae that have the least specialised perichaetia and sporophytes, and … likely to be the most primitive ones in the family." This result largely mirrors the conclusions of Tan & Jia (1998) in their cladistic study of Chinese members of the family. Ten years later, Goffinet et al. (2009) proposed a radically different classification, recognising (within the Hypnales) a Sematophyllaceae of 28 genera while simultaneously placing several genera traditionally assigned here in a newly described family, the Pylaisiadelphaceae. Among the genera they placed in the latter family were Taxithelium, Trismegistia, and Wijkia.

Even the eponymous genus Sematophyllum is poorly circumscribed, with Buck (1998, p. 368) describing it as "scarcely more than the dregs of the Sematophyllaceae… characterized by the lack of various specialised features."

The Australian flora, with it tropical components, is far richer than the N.Z. flora for this family. Ramsay et al. (2002; 2004) provided an Australian treatment recognising 14 genera that helps to place our more limited flora in context. Their discussion of large-scale systematic modifications proposed to the family prior to 2002 also provided some clarity to a confusing set of proposals. Because of the lack of consensus concerning the limits of the Sematophyllaceae, and generic boundaries within it, it is expedient to adopt the generic assignments proposed by Ramsay et al. (2002; 2004). For their regional revision Ramsay et al. employed a largely Brotherean classification, albeit with several exceptions, including notably: (1) they did not recognise the four subfamilies employed by Brotherus (1925) and (2) they excluded from the family the large and well-known genus Taxithelium (which does not occur in N.Z.). Tan et al. (1996) presented a useful key to the Australian genera.

Four genera, nine species, and one non-typical variety are accepted as part of the N.Z. flora. While generic placements largely follow Ramsay et al. (2002; 2004), not all of their species-level taxonomy is accepted here. For reasons alluded to above, no attempt is made to present a key to the N.Z. genera.

Hypnum cyparioides Brid. According to Dixon, the type of this obscure name was collected by de Labillardière in "Nova Hollandiae". Both Dixon (1929, p. 308) and Ramsay et al. (2004) placed Hypnum cyparioides Brid. in the synonymy of Rhaphidorrhynchium amoenum (Hedw.) M.Fleisch. (or its nomenclatural equivalent). Mitten (1859) applied the name H. cyparioides to a Tasmanian collection by Archer. Dixon discussed the confusion between H. cyparioides Brid. and H. amoenum Hedw. at some length but could locate types of neither name. Hypnum cyparioides Brid. (or combinations based on it) have received little or no application to N.Z. collections and this name is not considered further.

Hypnum leptorhynchum Brid. This name seems to have been first discussed in relation to H. leucocytus Müll.Hal. and H. cerviculatum Hook.f. & Wilson by Dixon (1929, p. 309). The name H. leptorhynchum Brid. is used by Wilson & Hooker in both Flora Antarctica (1845, p. 141) and in Flora Novae-Zelandiae (Wilson 1854, p. 112) for material from N.Z. Confusingly, Wilson (1854, p. 113) described the setae in such material as smooth ("laevi"). Many of the older collections in BM named as H. leptorhynchum Brid. are referable to S. amoenum. According to Tropicos (accessed 2 Nov. 2015) H. leptorhynchum Brid. is an invalid name and it is discussed no further here.

Rhaphidostegium dallii Broth. & Geh. This species was described from a single collection of unlocalised N.Z. provenance. It is discussed by Dixon (1929, p. 309) who saw only vegetative material. He tentatively accepted it while stating that "vegetatively it agrees exactly with R. leucocytus." Sainsbury (1955, p. 466) listed it as a doubtful species. Ramsay et al. (2004, p. 58) examined type material and considered it to be a synonym of Warburgiella macrospora, which is considered here to be inseparable from W. leucocytus. It is not considered further here.

Sematophyllum subcylindricum sensu Sainsbury and sensu Allison. Some N.Z. specimens named (mostly by K.W. Allison) as S. subcylindricum (Broth.) Sainsbury in N.Z. herbaria are perplexing due to their poor quality and for having ± "subcollenchymatous" exothecial cells and smooth setae (e.g., G.O.K. Sainsbury 4740 from Hopuruahine Hill near Lake Waikaremoana, Gisborne L.D. (CHR 570159) and K.W. Allison 3207 from Pukerimu Bush, near Taupō, S Auckland L.D. (CHR 570161). Most material so-named is aberrant and poor Rhaphidorrhynchium amoenum. Sematophyllum subcylindricum is considered here to be worthy neither of taxonomic recognition nor of further consideration. The relevant collections remain filed under S. subcylindricum in CHR.