Senecio australis Willd. (Asteraceae: Senecioneae) – a new and uncommon addition to the indigenous vascular flora of New Zealand

Abstract

Senecio australis is confirmed as present in New Zealand. A full description based on New Zealand specimens is also provided. Senecio australis was discovered in New Zealand in 1994. Prior to this date, the absence of supporting specimens had caused this species to be rejected from the New Zealand flora and treated as endemic to Norfolk Island. In this paper, the biostatus of S. australis in New Zealand is reviewed and it is concluded that the species is most likely a recent natural, bird-assisted arrival from Norfolk Island. The most probable seed vector of the Senecio is the grey ternlet (Procelsterna cerulea albivitta), a small seabird common on Norfolk and Kermadec Islands, which periodically visits northern New Zealand waters. The conservation status of S. australis is briefly discussed and the current status of ‘Non-Resident Native/Vagrant’ awarded to the species by the New Zealand Threatened Indigenous Vascular Plant panel is considered to be still appropriate.

Keywords:

• Asteraceae
• biostatus
• ITS DNA sequences
• lautusoid Senecio
• Norfolk Island group
• New Zealand flora
• S. australis
• S. marotiri
• Senecio

Introduction

During 1994 and 1997 specimens of Senecio marotiri were collected from Motukino (Fanal Island) (35°56′31.4″S, 175°8′59.9″E) (Fig. 1), the largest of the Mokohinau Island group (Fig. 1). These specimens formed the basis of an article discussing the species' morphological variability, distribution and conservation status (de Lange 1998). In that article, de Lange (hereafter PdL) noted that some Motukino gatherings of plants attributed to S. marotiri possessed leaf characters comparable with the Norfolk Island (29°2′32.3″S, 167°57′13.4″E; Fig. 1) endemic S. australis Willd., as described by Belcher (1992) and Green (1994), an observation prompting de Lange (1998) to suggest that a critical comparison of both these species was warranted.

Figure 1 A, Location of Norfolk and Kermadec Islands in relation to New Zealand and also of other island groups discussed in the text; B, North Island showing location of key island groups utilised by grey ternlets; C, Mokohinau Islands Group and Great Barrier Island (Aotea Island) showing location of Harataonga Bay [Great Barrier Island (Aotea Island)]; D, Fanal Island (Motukino Island) showing location of Gut Rock and Maori Rocks.

Subsequently, a visit by PdL to Norfolk Island in 1998 enabled fresh specimens of S. australis to be critically compared in the field with notes and measurements obtained from Motukino herbarium specimens of suspected S. australis lodged at AK (see representative specimens listed below). These field observations were later supplemented by studying live plants of S. australis from Norfolk Island that were in cultivation at the Auckland Botanic Gardens and additional New Zealand gatherings of senecios made from Motukino in 1999 and 2000, and from Harataonga Bay, Great Barrier Island (Aotea Island) (36°9′47.8″S, 175°28′55.7″E) (Fig. 1) in 2002. From the morphology of these plants, herbarium specimens and field observations it was concluded that S. australis is present in northern New Zealand. Nevertheless, despite accepting the species for New Zealand (de Lange et al. 2004a, 2005, 2009, 2013; de Lange & Rolfe 2010), no formal admission of the species into the New Zealand flora has been made.

In this paper, we provide further evidence from internal transcribed spacer (ITS) DNA sequence data for the presence of S. australis in New Zealand and a morphological description of S. australis based on New Zealand specimens. We also review the biostatus of the species, concluding that that this species is indigenous and that it probably has colonised New Zealand from Norfolk Island through seabird movement. Further, based on our survey of New Zealand herbarium holdings of lautusoid (sensu Belcher 1992) Senecio taxa and our field surveys we think it likely that S. australis colonised New Zealand ‘recently’ (i.e. within the last 50 years or less).

Materials and methods

Field observations, herbarium specimens and live plants of Motukino Senecio australis and S. marotiri were made by PdL, Ewen Cameron and Ian McFadden in 1994, by Ian McFadden in 1995, by PdL and David Norton in December 1997, and by Phil Todd during September 2000. During April 2002, PdL and David Norton also collected S. australis from Aotea (Great Barrier Island). Plants of S. australis and S. marotiri raised from Motukino seed were cultivated at the School of Biological Sciences experimental growth rooms, University of Auckland during 2000–2004, and later at Oratia Native Plant Nursery (2004–2008). These New Zealand S. australis plants were then compared with live plants of Norfolk Island S. australis grown on from seed sent from Norfolk Island by M. Christian and cultivated by the Auckland Regional Council Botanic Gardens (Accession AUCK 981356). In addition, herbarium specimens, type material and cibochromes of type material of S. marotiri and S. australis held at AK, B, BM, CHR, K, MEL, NSW and WAIK were examined.

Total genomic DNA was extracted from one Norfolk Island S. australis specimen (P. J. de Lange 4304, AK), two S. australis specimens from Motukino (P. J. de Lange 5463 and 5514, AK) and a S. marotiri specimen (P. J. de Lange CHC585 with P. B. Heenan, CHR) using the Qiagen DNeasy Plant Mini Kit. PCR amplification of the ITS region followed Pelser et al. (2002). PCR products were purified with the Promega Wizard SV Gel and PCR Clean-Up System. Cycle sequencing was carried out with Applied Biosystems BigDye Terminator v. 3.1. The sequenced samples were run on ABI 3130xL Genetic Analyzer at the University of Canterbury. Geneious 6.1.7 (Biomatters Ltd) was used for trace file editing. Edited sequences are deposited in GenBank (accession numbers KM365024–KM365027).

ITS sequences obtained from the Norfolk Island and New Zealand S. australis specimens and the S. marotiri specimen were included in an ITS alignment containing all Senecio accessions used in previous Senecioneae studies by Pelser et al. (2002, 2003, 2004, 2007, 2010a,b, 2012) as well as a large selection of unpublished ITS sequences of Australian and New Zealand lautusoid Senecio species (including all New Zealand representatives) that is being compiled as part an ongoing study into the diversification of lautusoid Senecio. This data set was used to determine genetic similarities among Norfolk Island and New Zealand S. australis and other Senecio species, particularly S. marotiri.

DNA sequencing results

The three S. australis accessions included in our study have identical DNA sequences of the ITS region (ITS1, 5.8S, ITS2). These sequences are unique among all Senecio species included in the global ITS alignment. At two nucleotide positions (one in ITS1 and one in ITS2), they contain nucleotides that are not present in any of the other lautusoid Senecio accessions that were studied. The ITS sequence obtained from S. marotiri (KM365027) is different from the S. australis sequences (KM365024 – KM365026) in three insertion/deletions and 26 nucleotide substitutions.

Taxonomy

Senecio L. Sp. Pl. 2: 866 (1753)

Lectotype species. S. vulgaris L. (Fide Greuter et al. 1993).

Senecio australis Willd. Sp. Pl., ed. 4, 3: 1981 (1803)

Holotypus (vide Belcher 1992). ‘New Zealand’ [Norfolk Island], J. R. & G. Forster; B-W, IDC microfiche 7740/1.1130/19.

Isotypi. K!, P!.

Description (of New Zealand specimens only). Erect to semi-erect, heavily branched and leafy, annual or short-lived perennial herb, (0.2−) 0.8–1.0 m tall, arising from a stout woody rootstock. Exposed portion of rootstock and stems frequently supporting adventitious roots. Stems green, maturing dark purple to purple–black, at first sparingly covered in fine arachnoid trichomes, becoming glabrous with age. Foliage dark green, glabrescent and fleshy. Cotyledons, ovate, 7–8 × 4–6 mm, dark green; lower leaves 55–75 × 25–30 mm, dark green, shortly petiolate, obovate to oblanceolate, entire, very rarely lobed 1–2×, apex obtuse; mid-cauline leaves 50–140 × 20–45 mm, dark green, gradually diminishing in size up the stem, at first petiolate or amplexicaul, if the former then with petioles up to 10–15 mm long, often sparingly covered in fine arachnoid trichomes; lamina ovate, obovate to elliptic, rarely rhomboidal, entire, coarsely dentate, or lobed 2–3×, leaf apices obtuse; uppermost leaves smaller, usually lanceolate to linear or linear–spathulate, lobulate or remotely, dentate, with obtuse apices. Inflorescences usually partially obscured within foliage; sparingly branched in a lax cyme of 3–8 capitula. Capitula urceolate. Peduncles terete, sparsely covered in white arachnoid trichomes; supplementary bracts 4–5, widely spaced, not or scarcely imbricate, 2.0–2.5 mm long, all linear–lanceolate, green with margins sparsely covered in arachnoid trichomes; calycular bracteoles 2–3, loosely subtending the involucre (never imbricate), 1.5–2.0 mm long, linear–lanceolate, apex maroon, crested with fine white, arachnoid trichomes. Involucral bracts (9−) 13 (−15), uniformly dark green, lanceolate, 5–6 mm long, shorter than ray and disc florets, thereby exposing a distinct band, 0.5–0.8 mm wide of pappus trichomes; bracts ± glabrous, sometimes with the margins sparsely invested in white, arachnoid trichomes, apex green (very rarely tinged maroon), crested with fine, white, arachnoid trichomes. Ray florets (0−) 9 (−12); ligules, ± involute or plane, dark sulfur-yellow, 1–5 mm × 1 mm; blunt-ended or with apex incised 3 times. Disc 4–5 mm diameter, disc florets, dark sulfur-yellow, (10−) 15 (−20). Cypselas narrowly elliptic to narrowly oblong–elliptic, slightly narrowed at apex, 2.5–2.8 × 0.3–0.5 mm; base cuneate; ribs 5–8, glabrous; grooves between ribs invested with 4–5 rows of short, white, somewhat flexuous retrorse trichomes; surface yellow–brown. Pappus caducous, scabridulous. Chromosome number 2n = 80 (P. J. de Lange 5514, AK 259121; P. J. de Lange 5463, CHR 200641, AK 283447 [spread over two sheets)] see de Lange et al. (2004b).

Representative specimens. Norfolk Island: J D. McCormish s.n., 20 Aug 1937, AK 90803/WELT SP073617; Anson Bay, W. R. B. Oliver s.n., 15 Nov 1956, WELT SP005424; Anson Bay, W. R. Sykes Norfolk 210, 20 Oct 1971, CHR 224353; Cemetery Area, Emily Bay, W. R. Sykes 785/87, 25 Nov 1987, CHR 455914; Kingston, Cemetery, R. O. Gardner 5782, 30 Sep 1989, AK 191986; Rocky Point Reserve, Cascade Bay, R. O. Gardner 5829, 4 Oct 1989, AK 190157; R. O. Gardner 5858, 6 Oct 1989, AK 190139; Cascades Beach, P. J. de Lange NF1 & G. M. Crowcroft, 5 Nov 1998, AK 237818; Rocky Point (The Hundred Acres), P. J. de Lange NF69 & G. M. Crowcroft, 7 Nov 1998, AK 237898; Mt Pitt National Park, Bird Rock Track, P. J. de Lange NF78 & G. M. Crowcroft, 8 Nov 1998, AK 237890. New Zealand: Mokohinau Islands Group: Motukino (Fanal Island): I. McFadden s.n., 21 Mar 1995, Gut Rock, AK 284669; P. J. de Lange 5410 & D. A. Norton, 7 Dec 1997, AK 256390; P. Todd s.n., 26 Sep 2000, AK 253008. Aotea (Great Barrier Island): Harataonga Beach, P. J. de Lange 5459 & D. A. Norton, 19 Apr 2002, AK 256401. Cultivated: ex Norfolk Island: W. R. Sykes s.n., Oct 1971, CHR 200641 (Chromosome Voucher); W. R. Sykes s.n., 3 Feb 1972, CHR 231768A; P. J. de Lange 4304, 16 Mar 2000, AK 251840 (Chromosome and DNA Voucher). New Zealand, ex Motukino (Fanal Island): P. J. de Lange 5463, Apr 2003, AK 283447 (Chromosome and DNA Voucher).

Discussion

ITS DNA sequences

The ITS region is the most variable DNA region that is routinely sequenced in molecular phylogenetic studies of Senecioneae (e.g. Pelser et al. 2002, 2003, 2004, 2007, 2010a,b, 2012; Calvo et al. 2013; Roda et al. 2013) at the species level. The finding of ITS sequences that are identical among Norfolk Island and putative New Zealand S. australis populations, but unique among all Senecio species that have been sequenced to date (including S. marotiri), in combination with morphological similarities between these populations, therefore suggest that S. australis is not endemic to Norfolk Island, but is also present in New Zealand and should be formally recognised as a component of the New Zealand flora.

Distribution and habitats

Within New Zealand, S. australis is so far known from the Mokohinau Island group, where it has only been collected from cliff ledges on Motukino (Fanal Island) and nearby Gut Rock (35°56′3″S, 175°9′7.9″E; Fig. 1; de Lange et al. 1995), and from rocks at the northern end of Harataonga Bay, Great Barrier Island (Aotea Island). At all of these locations, S. australis grows sympatrically with S. lautus. However, on Motukino (Fanal Island) and Gut Rock it also grows in association with S. marotiri and S. repangae subsp. pokohinuensis. On Motukino (Fanal Island) and Gut Rock other associated species include those typical of northern New Zealand seabird islands, for example, Lepidium oleraceum, Stellaria aff. parviflora, Wahlenbergia vernicosa, Chenopodium trigonum subsp. trigonum, Chionochloa bromoides and Disphyma australe subsp. australe. In these habitats, the species is closely associated with the roosts of white-fronted tern (Sterna striata), red-billed gulls (Larus novaehollandiae scopulinus) and the uncommon grey ternlet (Procelsterna cerulea albivitta), as well as the nesting sites of diving petrels (Pelecanoides urinatrix urinatrix), little shearwater (Puffinus assimilis haurakiensis), fluttering shearwater (Puffinus gavia) and white-faced storm petrel (Pelagodroma marina marina). On Great Barrier Island (Aotea Island), at Harataonga Bay, S. australis was found on cliff faces at the northern end of the bay in sites used as roosts by red-billed gulls. Here, the list of associated species reflects the more heavily modified nature of that island, for example, Paspalum dilatatum, Pseudognaphalium luteoalbum, Anthosachne multiflorum subsp. multiflorum, Sporobolus africanus, Anagallis arvensis, Hebe pubescens subsp. rehuarum and occasional Scandia rosifolia.

On Norfolk Island, S. australis occupies similar habitats to those seen in New Zealand, often flourishing in rank grassland and African olive (Olea europaea subsp. cuspidata) treeland dominated by naturalised species. It is also known to grow in open Norfolk pine (Araucaria heterophylla) forest, sometimes in association with another lautusoid species Senecio evansianus. However, it reaches its greatest abundance in active seabird nesting grounds where it commonly grows, often as the seasonal dominant with Achyranthes velutina, Canavalia rosea, Oxalis rubens and Tetragonia implexicoma amongst wedge-tailed shearwater (Puffinus pacificus) burrows, and also on cliff faces frequented as roosts and seasonal nesting grounds by white terns (Gygis alba), grey ternlets and noddies (Anous spp.).

Recognition

Of the lautusoid New Zealand Senecio, S. australis (Fig. 2) is superficially similar to the New Zealand endemic S. marotiri (Fig. 3) with which it was initially confused (de Lange 1998). Senecio australis (Fig. 2) is more heavily branched, bushy and leafy, has a somewhat spreading growth habit, oval rather than lanceolate cotyledons, characteristically broader, lyrate–pinnatifid leaves, silky hairy stems and leaf bases, and non-imbricating calyculus. In addition, the inflorescences of S. australis, which are often partially obscured within foliage, are in lax cymes of 3–8 capitula. Furthermore, the capitula are usually urceolate, and typically possess longer, blunt-ended dark yellow ray-florets (Fig. 2). In comparison with S. marotiri (Fig. 3) the cypselas of S. australis are also distinctive (Fig. 2), they are yellow–brown in colour, smaller and broader, and are scarcely if ever narrowed below the apex.

Figure 2 Senecio australis. A, Growth habit of S. australis [cultivated plants ex Motukino (Fanal Island), Mokohinau Islands]; B, mid-cauline leaves; C, basal cauline leaves; D, inflorescence; E, close-up of captiulum; F, cypsela (scale bar = 1 mm).

Figure 3 Senecio marotiri. A, Growth habit of S. marotiri [cultivated plants ex Motukino (Fanal Island), Mokohinau Islands]; B, mid-cauline leaves; C, basal cauline leaves; D, inflorescence; E, close-up of captiulum; F, cypsela (scale bar = 1 mm).

Aside from S. marotiri, S. australis is unlikely to be confused with the other New Zealand lautusoid Senecio taxa: S. carnosulus, S. lautus subsp. esperensis, S. lautus subsp. lautus, S. repangae subsp. pokohinuensis, S. repangae subsp. repangae and S. sterquilinus. From all these species, S. australis is amply distinguished in the field by the growth habit, leaf shape, inflorescence characters and cypsela morphology. Cytologically, it is also distinct from all New Zealand lautusoid taxa except S. carnosulus and S. marotiri with which it shares the same chromosome number (2n = 80).

New Zealand plants of S. australis are morphologically indistinguishable from Norfolk Island examples of this species (including those of Norfolk Island origin that had been cultivated in New Zealand), for example, P.J. de Lange NF1 & G.M. Crowcroft (AK 237818), P.J. de Lange NF78 & G.M. Crowcroft (AK 237890), P.J. de Lange NF69 & G.M. Crowcroft (AK 237898) and P.J. de Lange 4304 (AK 251840).

Biostatus

Although the presence of S. australis in New Zealand is confirmed, the question remains as to what its biostatus in the flora should be. Currently, the species is accepted as indigenous (see de Lange et al. 2005, 2013; de Lange & Rolfe 2010), however, the basis for that decision has not been properly documented. Here, we review the evidence for and against admitting S. australis into the indigenous flora of New Zealand.

On Norfolk Island, S. australis is a widespread and common species of open coastal habitats. It is especially common within seabird roosts and nesting sites, on rubble slopes and cliff faces. The species is probably also indigenous to Philip Island (29°7′36.6″S, 167°56′25.7″E; Fig. 1), although seed from Norfolk Island was deliberately sown there (M. Christian pers. comm.). Nevertheless, despite its abundance in these habitats, S. australis is not a ‘weedy’ species, nor is it seen further inland, or known to colonise roadsides, urban areas or farmland.

Given its frequent association with seabird roosts and nesting sites, and the scabrid nature of the cypsela, it seems very likely that S. australis is dispersed by seabirds. On Norfolk Island, it is most commonly found on exposed coastal promontories in close association with wedge-tailed shearwater nesting grounds, while on Philip Island it often associates with Kermadec petrel (Pterodroma neglecta neglecta). On the cliff faces of both islands it is commonly found in places used by grey ternlets and noddies as roosts and, at appropriate times, as nesting grounds. Although all of these birds have been recorded from northern New Zealand waters (Miskelly 2013), of them, grey ternlets deriving from both Norfolk and Kermadec Island groups (Fig. 1) are known to visit the Hauraki Gulf waters during warm sea temperatures in late summer (Szabo 2013). During their periodic visits to the Hauraki Gulf, grey ternlets have been recorded from the Poor Knights (35°28′6.3″S, 174°44′16.3″E) and Mokohinau Island groups (Fig. 1), which they mostly use as post-breeding roosts (Szabo 2013). Within the Mokohinau Islands they are most commonly reported from the Cathedral Rocks/Maori Rocks (35°55′24.2″S, 175°9′46.5″E; a series of rock stacks within close proximity to Motukino – for flora see Cameron & Wright 1990a; Szabo 2013) but they have also been seen on the eastern side of Motukino and on Gut Rock (P.J. de Lange unpubl. data) at sites where S. australis has been found.

Of those seabirds known to associate with S. australis on Norfolk Island, grey ternlets are considered the most likely seed vector because of their feeding habitats which usually consist of hovering close to water but rarely entering it, preferring instead to dip their heads in the water to catch their prey (Szabo 2013). Although grey ternlets are usually restricted to feeding grounds within a 20 km radius from their roosts and nesting sites, the fact that Norfolk and Kermadec Islands populations are known to occasionally extend their range to northern New Zealand waters, during which time they utilise outer Hauraki Gulf islands and rock stacks as temporary roosts, makes their postulated role as seed vector for S. australis even more likely. If so, then one would expect that S. australis would also be present on the Kermadec Islands, where other Norfolk Island seabird plant associates have been found, for example, Achryanthes velutina (reported as A. aspera by de Lange et al. 2004c), Canavalia rosea and Ipomoea pes-caprae subsp. brasiliensis (Sykes 1977; de Lange 2011a,b,c, 2012). However, the species has yet to be recorded from there, suggesting that it has either been missed by recent botanical surveys of the islands (de Lange 2011a,b,c, 2012), or, more likely, considering that long-distance dispersal is a chancy affair, the species has yet to successfully colonise those islands. The same appears to be the case for northern New Zealand, where despite a periodic presence of grey ternlets on the Poor Knights Islands (Szabo 2013), S. australis has yet to be seen (de Lange & Cameron 1999), nor has the species been reported from the Three Kings (34°10′42.9″S, 172°4′48.5″E) and Aldermen Islands (36°57′38.9″S, 176°5′17.9″E) or the Volkner Rocks (37°28′38.7″S, 177°8′2.8″E) (P.J. de Lange unpubl. data), (Fig. 1). This pattern suggests that either S. australis is a recent arrival in the New Zealand flora, or that it has always been in New Zealand but that it has been overlooked.

Because S. australis has a superficial similarity to S. marotiri and depauperate and/or narrow-leaved variants of S. lautus it is possible that it might have gone unrecognised in collections of that species from New Zealand. However, we think this unlikely because the S. lautus complex has been so well collected and intensively studied (Ornduff 1960; Sykes 1971; Webb 1988a,b; de Lange & Murray 1998). Therefore, had S. australis been any more widespread in New Zealand than is now known, we think it unlikely that it would have been overlooked. Indeed, Belcher (1992) made a special search for New Zealand specimens of S. australis, when in the process of resolving the origin of the type specimen, which purportedly came from New Zealand. Part of Belcher's argument for S. australis being a Norfolk Island endemic was the then absence of S. australis from New Zealand herbaria (and other world herbaria holding New Zealand collections of S. lautus; Belcher 1992).

It could be argued that Belcher's lack of success in locating New Zealand S. australis specimens merely reflects the remote locations at which S. australis has been found in New Zealand, and so the possibility that it had yet to be collected. However, this too seems unlikely because the Mokohinau Islands and Great Barrier Island (Aotea Island) S. australis specimens were mostly collected at sites that have been very well explored (Wright 1980a,b; Cameron & Wright 1982, 1990a,b; Bartlett & Gardner 1983; Cameron 1990, 2001), including by people such as A.E. Wright (pers. comm.) who had a particular interest in collecting lautusoid Senecio. Considering the timing of these visits (i.e. they were within the growing season of this plant), the distinctive nature of S. australis and its superficial similarity to S. marotiri which was collected from these islands during these surveys (see de Lange 1998), it is doubtful that S. australis would have been missed had it already been present there.

A final possibility is that S. australis is a naturalised species in New Zealand. This too seems rather unlikely, for despite the detailed botanical knowledge of the islands from which it has been collected, these places are remote from the main points of entry into New Zealand. Further, the two Mokohinau Islands locations, Motukino and Gut Rock are part of a remote island nature reserve with permit-only access. Motukino is a cliff girt island with a difficult landing, such that it is scarcely visited outside scientific investigations. Gut Rock, by contrast, can be accessed only by helicopter, and was first explored in 1995 (I. McFadden pers. comm.). Even on Great Barrier Island (Aotea Island), despite the heavily modified nature of Harataonga Bay, S. australis still occurs in a remote place far from the main ports of entry into New Zealand, and there it occupied similar habitats to those seen on the Mokohinau Islands, as well as on Norfolk Island. Further, S. australis is, as already observed, not a ‘weedy’ species, nor is it easy to maintain in cultivation. Its absence from habitats frequented by tourists on Norfolk Island and its apparent inability to colonise roadsides and urban areas on that island also argue against its accidental introduction from there to New Zealand by human agency. Finally, assuming that it were introduced by human agency, one would expect to see it appear in the vicinity of the main ports of entry, or within urban areas first, not on remote island groups in the outer Hauraki Gulf.

Therefore, based on the available evidence we believe that Senecio australis is indigenous to New Zealand, and that it probably arrived here from Norfolk Island via seabirds, of which the grey ternlet seems the most plausible candidate, and that this happened recently.

Conservation status

Using the New Zealand Threat Classification System (Townsend et al. 2008) Senecio australis is currently assessed as ‘Non Resident Native/Vagrant’ and qualified ‘SO’ [Secure Overseas] by de Lange et al. (2013). This reflects accurately that the species is extremely uncommon in New Zealand, being known only from a handful of plants in three locations, and it has yet to be seen if it will successfully expand its range. This conservation assessment assumes that the species is a recent, natural arrival – that is, it reached New Zealand sometime after 1950, and that its occurrences here, so far, appear to be sporadic and possibly of limited duration. The present-day status of S. australis also needs to be established, as the last collection of it, that from Harataonga Bay, Great Barrier Island (Aotea Island), was made in 2000, and there is no update on its presence on the Mokohinau Islands since its discovery there in 1994. Further survey is needed to determine whether S. australis has expanded its range on these islands, or whether it has declined or died out. Until this is done, the current conservation assessment is considered appropriate.

Acknowledgements

We thank Margaret Christian (formerly Australian National Parks & Wildlife Service, Norfolk Island) for assistance in the field in 1998 and comments. Ewen Cameron, Curator, (AK) provided us with access to Senecio specimens in his care, and as referee his comments helped improve the paper. We also wish to thank Ian McFadden, David Norton and Phil Todd for their company in the field and/or collections of Senecio australis specimens, and both Anthony Wright and Graeme Taylor for helpful discussion. Geoff Davidson (Oratia Native Plant Nursery, West Auckland) kindly cultivated Mokohinau Island specimens of S. australis so enabling this paper, and Rhys Gardner commented on a draft of this article.