Drepanocladus (Müll.Hal.) G.Roth

Plants medium-sized to robust, forming interwoven mats, mostly growing in wet places and often submerged or emergent. Stems mostly ascendant, subpinnately or irregularly branched and with branches in one plane (± distichous), in cross-section with small, incrassate outer cells and a central strand usually present; rhizoids sparse, inserted at or just below the leaf insertion, red-brown, smooth. Stem leaves falcate or straight, often strongly falcate-secund at stem apices, ovate-lanceolate or ovate, smooth or striolate (not plicate) when dry, plane at margins, entire in N.Z. taxa, not decurrent in N.Z. taxa; mid laminal cells unistratose in N.Z. taxa, smooth, vermicular, not or weakly porose, not differentiated at margins or at leaf apex and lacking rhizoid initials; basal laminal cells usually wider and shorter, sometimes porose near costa; alar group distinct and usually large, triangular or quadrate, the cells inflated, often hyaline and thin-walled. Costa single, unbranched or occasionally branched, mostly extending to mid leaf or longer. Branch leaves smaller but otherwise similar, often strongly falcate-secund at branch tips. Paraphyllia absent.

Autoicous or dioicous. Perichaetia scattered on stems, the leaves plicate, elongate, and clasping. Perigonia scattered on stems, often conspicuous below perichaetia in autoicous species, the bracts spreading. Setae elongate, flexuose; capsules inclined to horizontal, short-cylindric or ovoid from a short neck, strongly curved, often constricted below the mouth when dry; mouth slightly oblique; exothecial cells firm-walled; stomata numerous in neck, with elliptic pores; annulus differentiated, usually falling with the operculum; operculum conic. Peristome hypnoid, as per family; exostome teeth golden, lanceolate and ± shouldered, bordered, on the outer surface with a zig-zag median line, finely cross-striolate below, papillose above, on the inner surface with numerous projecting trabeculae and lacking papillae; endostome from a high, pale basal membrane, the segments elongate, keeled, narrowly perforate and alternating with 1–4 nodose cilia. Calyptra cucullate, naked. Spores spherical, finely papillose.

With their myriad subaquatic and aquatic forms, both species of Drepanocladus can be confused with Warnstorfia fluitans. Our Drepanocladus spp. are more robust plants with ovate-lanceolate stem leaves compared to W. fluitans which has more narrowly acuminate leaves. In Drepanocladus the branches lie in a single plane (i.e., ± distichous) while in Warnstorfia the branches are distributed around the stem, but this feature cannot always be observed. The perichaetial leaves of both Drepanocladus species are distinctly plicate while those of W. fluitans are not; however the plicate nature of the perichaetial leaves here seems to be less conspicuous in female material lacking capsules. Both N.Z. species of Drepanocladus usually grow in relatively cation-rich sites, whereas Warnstorfia fluitans grows in nutrient-poor and usually acidic habitats.

Hedenäs (2003, p. 16) used axillary hairs (cited as 1–3-celled in their upper part in Drepanocladus vs 1–7-celled in Warnstorfia) to distinguish these two genera. I have been unable to usefully apply this character to N.Z. material.

A narrow concept of Drepanocladus, conforming to that of Hedenäs (2003) is presented here; this concept also agrees with that employed by Fife (1995), although one additional species is recognised here. According to Hedenäs’ concept, Drepanocladus is a genus of nine species. It is best developed in temperate regions, but also occurs at high elevations in the tropics. The distinctions between the two recognised N.Z. species are often not clear in sterile material and perhaps as many as half of N.Z. Drepanocladus specimens cannot be confidently assigned to either species using morphological characters alone.

The variability of D. aduncus and D. brachiatus in response to environmental conditions further blurs the species distinctions. The recognition by some authors of infra-specific taxa in D. aduncus in many parts of its extra-N.Z. range highlights this variability. Straight-leaved forms have been recognised as the variety kneiffii (Schimp.) Mönk. by numerous authors (e.g., Crum & Anderson 1981, p. 965) and these are discussed briefly under D. aduncus.

When well-developed, both D. aduncus and D. brachiatus are distinctive and distinguishable by the characters used in the key. Fruiting material of D. brachiatus can nearly always be recognised by its autoicous sexuality. However, many collections cannot be readily sexed. Depending on environmental regimes the degree of costal development, degree of alar cell differentiation, leaf dimensions, and the degree of leaf curvature are all subject to variation. The size of the alar group varies, even on single stems, and leaves with well-developed alar groups should be examined in making species distinctions.

Hedenäs (1996) has quantified the inter-dependent variability of some leaf characters in European D. aduncus using culture under controlled conditions.

Because of the difficulty in naming sterile material, the distributions for both recognised species are probably understated. Given the wide distribution of both species in N.Z., provenance is not useful in determining material. Unnameable sterile material of Drepanocladus has been seen from N Auckland (including offshore islands), S Auckland, Gisborne, Hawke’s Bay, Wellington, Nelson, Marlborough, Canterbury, Westland, Otago, Southland, Ch, and A. The decision to recognise two species, while acknowledging the difficulties in distinguishing them, is presented here as a pragmatic solution to an otherwise intractable taxonomic problem.

It is worth noting that under Hypnum sect. Adunca Hooker (1867, p. 472) did not report Hypnum aduncum from N.Z., but recorded only the newly described Hypnum brachiatum and H. kneiffii.

Drepanocladus longifolius R.S.Williams was cited in the synonymy of D. brachiatus (Mitt.) Dixon by Dixon (1929, p. 316. It is not considered here.

Drepanocladus polycarpos (Voit) Warnst. was cited from N.Z. by Żarnowiec (2001) and from Macquarie I. by Seppelt (2004); it is not recognised here for reasons discussed above.