Asplenium decurrens Willd., an earlier name for A. northlandicum (Brownsey) Ogle

ABSTRACT

The nomenclatural history of the Asplenium obtusatum complex is reviewed. It is shown that when treated at subspecific rank, A. obtusatum subsp. northlandicum Brownsey, based on a New Zealand type, is the correct name for the octoploid taxon that occurs in both New Zealand and Australia. However, when raised to species rank there are two earlier validly published names, A. decurrens Willd. and A. sarmentosum Willd., based on Australian types, that take priority over Asplenium northlandicum (Brownsey) Ogle. Both names relate to the same taxon and were published simultaneously. The name A. decurrens is preferred because the holotype is based on a herbarium specimen, whereas the type of A. sarmentosum is an illustration. This short communication is a contribution towards clarifying the taxonomic and nomenclatural status of New Zealand plants for the plant names database (Ngā Tipu Aotearoa) and the electronic Flora of New Zealand.

KEYWORDS:

• Aspleniaceae
• Asplenium decurrens
• Asplenium northlandicum
• Asplenium obtusatum subsp. northlandicum
• Asplenium sarmentosum
• lectotypes
• New Zealand flora
• taxonomy

Introduction

The nomenclature for New Zealand’s pinnate spleenworts in the Asplenium obtusatum clade (Perrie & Brownsey 2005, figure 2; Ohlsen et al. 2014, figure 3) has had a very chequered history. Forster (1786) described three species from New Zealand—Asplenium obliquum G.Forst. and A. obtusatum G.Forst.—apparently differing only in their acuminate or obtuse pinna apices, respectively, and both poorly distinguished from the third species, A. lucidum G.Forst. (now A. oblongifolium Colenso; see Brownsey 1979), which was said to have oblong-ovate rather than oblong pinnae. Subsequent authors struggled to interpret these three species. Richard (1832) reduced A. lucidum to synonymy with A. obliquum but made no mention of A. obtusatum. Cunningham (1837) recognised all three species, citing specimens for each of them from the Far North. Hooker (1844, p. 108) was the first to recognise the similarity between A. obtusatum and A. obliquum in the New Zealand flora, reducing A. obliquum to a variety of A. obtusatum and stating that ‘all the various stages between [them] exist in Lord Auckland’s group [Auckland Islands], and probably in other islands of which this plant is an inhabitant’. He also stated that he had retained the name of A. obtusatum for the species, because it was the more frequent of the two plants described by Forster. Later, Hooker (1855, 1867) continued to recognise A. lucidum, A. obtusatum and A. obtusatum var. obliquum, but noted that all three taxa were abundant throughout the islands of New Zealand. By the turn of the century there was some greater clarity around A. obtusatum which Cheeseman (1906) characterised as a littoral plant occurring around the coast from the Three Kings Islands to the subantarctic. However, his understanding of A. lucidum was much less clear, adopting the five varieties (var. lucidum, var. obliquum, var. scleroprium, var. lyallii and var. anomodum) that had been proposed by Moore (1857–1862). Cheeseman considered that the typical form was widely distributed in lowland districts as far south as Stewart Island, and var. obliquum from North Cape to Campbell Island. In the second edition of his Manual, Cheeseman (1925) restored A. anomodum Colenso and A. scleroprium Hombr. to species rank, but still retained var. obliquum and var. lyallii under A. lucidum. Allan (1961) reversed some of this by recognising var. obliquum under A. obtusatum, and once again returned A. scleroprium to a variety of A. lucidum. The distributions of both A. lucidum and A. obtusatum were said to extend from the Kermadec Islands to Campbell Island, but on the other hand var. obliquum was restricted to the southern South Island and the subantarctic islands.

Meanwhile, in Australia, Labillardière (1807) had recognised A. obtusatum and A. obliquum, citing specimens of both from Tasmania. However, Brown (1810) and Bentham (1878) reduced A. obliquum to synonymy with A. obtusatum.

Brownsey (1977a) revised Asplenium in New Zealand and showed that A. obliquum was simply a morphological variant within A. obtusatum. He also clarified the distinction between A. obtusatum and A. oblongifolium (as A. lucidum), and recognised A. scleroprium and A. lyallii (Hook.f.) T.Moore (syn. A. anomodum) as distinct species. The distribution of A. oblongifolium was shown to extend from the Kermadec Islands to the middle of the South Island, whereas A. obtusatum occurred from the Kermadec Islands to the subantarctic. Most importantly, A. obtusatum was shown to consist of two cytotypes—a tetraploid referred to subsp. obtusatum that occurred from the subantarctic to Cook Strait, and an octoploid referred to subsp. northlandicum Brownsey that occurred from Auckland to the Kermadec Islands. The rank of subspecies was chosen because the two cytotypes were morphologically very difficult to distinguish except by chromosome number and spore size.

Ogle (1988, p. 592) challenged this decision and raised A. northlandicum to species rank on the basis that the two taxa ‘are morphologically distinct, probably reproductively isolated, and appear to differ ecologically … ’. However, Brownsey (1998) continued to accept the octoploid cytotype as a subspecies of A. obtusatum in his treatment of Asplenium in Australia where only the octoploid is thought to occur. He included the names A. decurrens Willd. and A. sarmentosum Willd., based on Australian types, in the synonymy of subsp. northlandicum.

Brownsey (1977b) suggested that subsp. northlandicum might be an autopolyploid derivative of the morphologically very similar, but tetraploid, subsp. obtusatum. However, Shepherd et al. (2008) used nuclear DNA sequence data to show that this origin was unlikely. Rather, they suggested that A. northlandicum was an allopolyploid, with either A. obtusatum or A. oblongifolium as the chloroplast parent and an unknown tetraploid as the other parent.

That conclusion supported the recognition of the octoploid and tetraploid cytotypes as distinct species, rather than as subspecies of A. obtusatum. However, although Ogle (1988) had previously made the necessary combination at the species level for New Zealand material, when New Zealand and Australian octoploid plants are united at the rank of species, there are actually two earlier available names based on Australian types—A. decurrens Willd. and A. sarmentosum Willd. When the octoploid is treated at subspecific rank, A. decurrens and A. sarmentosum have no priority and are correctly treated as synonyms of A. obtusatum subsp. northlandicum, but when the octoploid is raised to specific rank, one of A. decurrens and A. sarmentosum must take priority (Art. 11.5). This short communication addresses that choice.

Taxonomy

Forster (1786) described A. obtusatum and A. obliquum from New Zealand. Today these two names are regarded as synonymous (Brownsey 1977a), but for a long time they were thought to be different species. Even Allan (1961) still treated A. obliquum as a variety of A. obtusatum.

Labillardière (1807) accepted Forster’s two species. His Novae Hollandiae Plantarum Specimen contains descriptions and illustrations of both of them from Tasmania. He stated that the descriptions by Swartz (1801) and Forster (1786) were included within his concept of the two species. His concepts therefore extended the distribution of the two species from being New Zealand endemics to being present in both Australia and New Zealand.

Willdenow (1810) had a very different notion. He accepted four species—A. obtusatum and A. obliquum in New Zealand, and A. decurrens and A. sarmentosum in Australia. Asplenium decurrens and A. sarmentosum were described by Willdenow in the same publication (1810; see http://bibdigital.rjb.csic.es/ing/Libro.php?Libro=3417) from Tasmania (“Habitat in Novae Hollandiae capite Van Diemen”), and distinguished from A. obtusatum G.Forst. and A. obliquum G.Forst. Willdenow treated A. decurrens as an exclusively Australian plant, citing Labillardière’s description of A. obliquum and his illustration of Australian material (Tab. 242, Figure 1), but specifically excluding (“exclusis synonymis”) “Swartz , fil. p.78” and “Forst. prodr. No. 429” both of which are descriptions based on Forster’s type from New Zealand. This exclusion meant that A. decurrens was a distinct Australian species, and not a synonym of the New Zealand A. obliquum. Willdenow provided a description of his own and stated “v.s.” (vidi siccam—I have seen a dried specimen). There is a specimen in Willdenow’s herbarium at B, collected by Labillardière from Nova Hollandia, which is labelled Asplenium decurrens and bears a description in Willdenow’s hand-writing identical to that in the protologue. This specimen (B-W 19885-01 0) is the only one labelled Asplenium decurrens in Willdenow’s herbarium and is almost certainly the holotype.

Willdenow (1810) also treated Asplenium sarmentosum as an Australian species, citing Labillardière’s description of A. obtusatum and his illustration (Tab. 242, Figure 2). Again, he specifically excluded “Swartz , fil. p.78” and “Forst. prodr. No. 430” both of which are descriptions based on Forster’s type from New Zealand. He provided a description but, in this case, did not indicate that he had seen a dried specimen. There is no specimen labelled Asplenium sarmentosum in Willdenow’s herbarium. Hence the only extant element known to have been available to Willdenow when drawing up his description is Labillardière’s illustration that he cited. This illustration can therefore be regarded as the holotype (Art. 9.1).

Current knowledge (Brownsey 1998; Ohlsen et al. 2014; Perrie & Brownsey 2016) suggests that the tetraploid cytotype (A. obtusatum s.s.) is present in New Zealand but not in Australia, whereas the octoploid cytotype occurs in both New Zealand and Australia, and when treated at species level either of Willdenow’s simultaneously published names could be applied to it (Art. 11.5). Asplenium decurrens is preferred because Willdenow indicated that he had seen a herbarium specimen of it, and the holotype collected by Labillardière is extant in the Willdenow herbarium (B-W 19885-01 0). By contrast, Willdenow gave no indication of having seen a herbarium specimen of A. sarmentosum, none are now extant in the Willdenow herbarium, and the holotype of A. sarmentosum must therefore be the illustration in Novae Hollandiae Plantarum Specimen (Labillardière 1807) cited by Willdenow (1810).

When treated at subspecific rank A. obtusatum subsp. northlandicum is the correct name for this taxon, but when raised to species rank the correct name is A. decurrens with a full synonymy and typification as follows. Following the conclusions drawn by Shepherd et al. (2008) we accept this taxon at the species level.

Asplenium decurrens Willd., Sp. Pl. 5(1): 316 (1810)

Holotype: Nova Hollandia, Labillardière, Herb. Willdenow, B-W 19885-01 0 (!online at http://ww2.bgbm.org/Herbarium/default.cfm)

= Asplenium obtusatum subsp. northlandicum Brownsey, New Zealand J. Bot. 15: 49 (1977)

≡ Asplenium northlandicum (Brownsey) Ogle, New Zealand J. Bot. 25: 592 (1988)

Holotype: Karekare, Auckland, P.J. Brownsey NZ 667, 19 Mar. 1974, CHR 308924!

= Asplenium sarmentosum Willd., Sp. Pl. 5(1): 316 (1810)

Holotype: Tab. 242, Figure 2 in Labillardière, Novae Hollandiae Plantarum Specimen, Vol. 2 (1807)—see http://www.biodiversitylibrary.org/item/127076#page/230/mode/1up

Distribution

Asplenium decurrens occurs in Australia and New Zealand. Morphologically similar plants from some Pacific islands (Tonga, French Polynesia, Pitcairn and Easter Island) are probably referable to this species. By contrast, A. obtusatum occurs in New Zealand and Tristan da Cunha, from which tetraploid chromosome counts have been obtained (Manton & Vida 1968; Brownsey 1977a), with morphologically similar plants from southern South America and other islands in the Southern Ocean probably also referable to this species.

Acknowledgements

Associate Editor: Dr Matt Renner.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was supported by Core funding for Crown Research Institutes from the Ministry of Business, Innovation and Employment’s Science and Innovation Group.