Accepted Article Received Date : 13-Oct-2016 Revised Date : 19-Dec-2016 Accepted Date : 22-Dec-2016 Article type : Original Article Running head: Puccinia otagensis comb. nov. The intriguing and convoluted life of a heteroecious rust fungus in New Zealand M. Padamsee* & E. H. C. McKenzie padamseem@landcareresearch.co.nz mckenziee@landcareresearch.co.nz Landcare Research Private Bag 92170 Auckland New Zealand * Author for correspondence Abstract: Molecular phylogenetic analyses of New Zealand rust fungi suggested that four taxa, Aecidium otagense on Clematis spp., Puccinia tiritea on Muehlenbeckia spp., P. rhei-undulati sensu auct. NZ on Rheum ×rhabarbarum, and an unidentified Puccinia sp. on Rumex sagittatus, are a single species. Morphological studies and multi-locus molecular data, together with inoculation studies, confirmed this finding. This species is only the third heteroecious rust fungus known to be native to New Zealand. This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/ppa.12672 This article is protected by copyright. All rights reserved. Keywords: Accepted Article Host jumps, Mikronegeria fuchsiae, One fungus one name, Pucciniales, Systematics, Uredinales. Introduction Rust fungi are well known for their complex life cycles, with up to five spore stages, which are completed on two unrelated hosts in a heteroecious species, or on one host in an autoecious species. Most of the native rust species in New Zealand are assumed to be autoecious and may lack one or more spore stages (microcyclic). New Zealand is considered to have a depauperate rust mycota that consists of approximately 125 native species, of which about one third are in the form genera Aecidium and Uredo (McKenzie, 1998). There are only two known native heteroecious rust species in New Zealand; Puccinia caricina DC. on native and introduced Carex species with native Urtica species as the alternate host (McKenzie, 1998) and Mikronegeria fuchsiae P.E. Crane & R.S. Peterson on Fuschia spp. with Phyllocladus spp. as the alternate host (Crane & Peterson, 2007). Rust was first recorded on rhubarb (Rheum ×rhabarbarum, Polygonaceae) in New Zealand by Cunningham (1945). He stated that it was first found in 1932, in a private garden in Wellington, and appeared to be spreading slowly throughout the country. On the basis of its host and urediniospores Cunningham identified the rust as Puccinia rhei-undulati (Diet.) Hirats. f., a species described from Japan on Rheum spp. (Dietel, 1906; Hiratsuka et al., 1992). Only urediniospores have been found on New Zealand material. The rust has been also widely recorded in Australia (e.g., Sampson & Walker, 1982). It has been assumed that the rust was introduced to New Zealand from Japan. The Japanese rust was described from plants This article is protected by copyright. All rights reserved. of Rheum rhaponticum and R. undulatum (cult.), Prov. Musashi, Tokyo (N. Nambu), Accepted Article cultivated in the University of Tokyo botanical gardens (Ito, 1950). Unfortunately, there is no voucher specimen and no living plants in the botanical garden (B.G. Koishikawa, University of Tokyo Botanical Gardens, pers. comm.) Rhubarb, which probably originated in China where it is used for medicinal purposes, is now widely cultivated throughout temperate areas of the world including Europe and North America. Interestingly, Puccinia rhei-undulati has been recorded only in Australia, Japan and New Zealand. The only other rust known on Rheum spp. is the aecial stage of the widespread, heteroecious P. phragmitis (Schumach.) Tul. Both Rheum and Rumex species are the alternate host for this rust, while the uredinia and telia occur on Phragmites communis (Wilson & Henderson, 1966). Puccinia phragmitis has been recorded on Rheum ×rhabarbarum in Australia (Shivas, 1989), but it is not known in New Zealand. In New Zealand, several native species of Clematis (Ranunculaceae) are infected by the iconic endemic rust, Aecidium otagense Linds. (McKenzie, 1998). Infection usually leads to extensive and spectacular swelling and distortion of stems and flowers, which can give the impression that the plant is the resting place of a snake or lizard. This rust was described by the Scottish lichenologist-physician W. Lauder Lindsay who visited New Zealand in 1861–62 (Lindsay, 1867). Lindsay collected for approximately three months near Dunedin on the South Island, and among his specimens were several rust fungi including A. otagense, which he described and reported on three different host plants—Clematis hexasepala, Epilobium junceum, and Microseris forsteri (Lindsay, 1867). He gave an extensive description of the symptoms of infection on Clematis as “…filiform or slender petioles, in particular, not only become twisted and curled variously, but are the seat of irregular, succulent, gouty swellings—of cucumber or cactus-like growths, whose nature is rendered apparent by the beautiful buff-coloured peridia, by which they are covered”. Shortly after, Lindsay (1868) This article is protected by copyright. All rights reserved. described the fungus as “…producing monstrosities of the flowers and flower-petioles”. At Accepted Article this time he designated the rust on each of the three host plants as varieties a, b, and c, respectively. Two of the three specimens were examined by McKenzie (1981). The rust on Clematis hexasepala (host redetermined as C. paniculata) was designated the lectotype of Aecidium otagense, while that on Epilobium junceum (nomen confusum) was found to be the aecial stage of Puccinia pulverulenta Grev. The third specimen on Microseris forsteri could not be located, but it is perhaps the aecial stage of Puccinia hieracii (Röhl.) H. Mart., later determined to occur on a second species of Microseris, M. scapigera (McKenzie, 1981). Aecidium otagense has been illustrated by several workers. The colourful symptoms of infection on Clematis were illustrated by Lindsay (1867). John Buchanan also illustrated the rust (as Aecidium tragopogonis) in an unpublished manuscript (MS-41) held in Auckland Museum Library. Although not dated the illustration (Fig. 1a) is likely to date from about 1875. A black and white photograph of the rust also featured as the frontispiece for the New Zealand rust mycota (Cunningham, 1931). Aecidium otagense has been recorded on six endemic Clematis species in New Zealand: C. afoliata (Cunningham, 1945), C. foetida (Cunningham, 1926), C. forsteri (Lindsay, 1867, as C. hexasepala), C. hookeriana (Kirk, 1908, as C. colensoi), C. marata (Kirk, 1908, as C. hexasepala) and C. paniculata (Kirk, 1906, as C. indivisa). Another rust Puccinia tiritea G. Cunn. is found throughout New Zealand on three endemic hosts, Muehlenbeckia australis (Polygonaceae), M. axillaris, and M. complexa (Cunningham, 1923; Cunningham, 1924). The rust is known only from New Zealand, although all three hosts are also found in Australia. Puccinia tiritea produces uredinia and telia and, according to Cunningham (1923; 1931), also spermagonia. The coarsely warted teliospores and lack of apical thickening separate this rust from an introduced rust, Puccinia muehlenbeckiae (Cooke) P. Syd. & Syd. that occurs on Homalocladium platycladum This article is protected by copyright. All rights reserved. (Polygonaceae) in New Zealand (Cunningham, 1931) and on Muehlenbeckia adpressa Accepted Article (Shivas, 1989), M. gracillima (McAlpine, 1906), and M. gunnii (Sampson & Walker, 1982) in Australia. Another rust, Uromyces politus (Berk.) McAlpine occurs on M. cunninghamii in Australia (McAlpine, 1906), but this species produces only aecia and telia, the latter producing mainly one-celled teliospores. Although Cunningham (1923; 1931) described spermagonia for P. tiritea on Muehlenbeckia, he did not indicate on either the labels or packets of specimens housed in Fungarium PDD that this spore stage was present on any of the specimens that he examined (despite usually doing so), and they do not appear to be present on any of the specimens held in PDD. In June 1996, a rust fungus was noticed on climbing dock (Rumex sagittatus, Polygonaceae) in Auckland. The host originates from South Africa, but there appear to be no records of rust on this host plant in extensive literature searches. Both uredinia and telia were present and the rust was thought to be a new species by the authors of this paper. The three species of rust on other Rumex species in New Zealand—Puccinia kirkii G. Cunn., P. ludwigii Tepper, and Uromyces rumicis (Schumach.) G. Winter—are morphologically distinct from the rust on R. sagittatus. In a molecular phylogenetic study of New Zealand rust fungi, it was observed that four of the above mentioned taxa were all recovered in the same clade. We therefore examined them more closely to assess whether Aecidium otagense, Puccinia tiritea, and the rusts on Rheum ×rhabarbarum (hitherto known as P. rhei-undulati in New Zealand) and Rumex sagittatus are the same species. This article is protected by copyright. All rights reserved. Materials and methods Accepted Article Isolates and morphology For examination of morphological features the rusts were mounted in lactophenol and examined using an Olympus BH-2 microscope. Voucher materials are deposited in Fungarium PDD. Phylogenetic analyses The rust fungi were also examined by molecular methods. Sori were excised and DNA was extracted using an X-tractor Gene System (Corbett Life Science, NSW, Australia). Part of the nuclear ribosomal internal transcribed spacer region (ITS) and the nuclear ribosomal large subunit (LSU) locus was amplified with a rust-specific primer Rust2inv (Aime, 2006) and LR6 (Vilgalys & Hester, 1990) using the protocol for PCR conditions in Aime (2006). The LSU locus was sequenced with the following primers: Rust2inv, LR6, LR3 (Vilgalys & Hester, 1990), and LROR (Moncalvo et al., 1995). Partial sequences of the ß-tubulin gene were amplified using a semi-nested protocol with rust-specific primers ß-tub 1317F and ß-tub 2262bR (CTCCATCTCGTCCATTCTA) for the first PCR reaction and ß-tub 1442F and ß-tub 2262bR for the second PCR reaction (Van der Merwe et al., 2007). The latter two primers were also used to sequence the partial ß-tubulin gene. The thermal cycling conditions for the first PCR reaction were as follows: 5 minutes at 94 ºC, followed by 35 cycles of 30 seconds at 94 ºC, 45 seconds at 45 ºC, 70 seconds at 72 ºC, and finally 10 minutes at 72 ºC. The thermal cycling conditions for the second PCR reaction were as follows: 5 minutes at 94 ºC, followed by 32 cycles of 40 seconds at 94 ºC, 40 seconds at 58 ºC, 100 seconds at 72 ºC, and finally 15 minutes at 72 ºC. Additionally, we tested various combinations of ITS primers to This article is protected by copyright. All rights reserved. amplify the ITS locus. The following primer combinations were tested: ITS 1F (Gardes & Accepted Article Bruns, 1993)/ ITS 4BR (Vialle et al., 2009), ITS 1F/ ITS 4B (Gardes & Bruns, 1993), ITS 1F/ ITS 4 (White et al., 1990), ITS 1rustF10d (Barnes & Szabo, 2007)/ ITS 4, and Rust ITS 1F/ Rust ITS 2 R (Toome & Aime, 2015), using the specified PCR protocols. PCR products were diluted (1:5) before sequencing with BigDye Terminator sequencing enzyme v.3.1 (Applied Biosystems, Life Technologies New Zealand Limited, Auckland, NZ) in a 10 µl reaction. Sequencing reactions were cleaned by BigDye XTerminator® Purification Kit (Applied Biosystems), and sequenced on an Applied Biosystems 3100-Avant Genetic Analyzer. Resulting ITS 2, LSU and ß-tubulin sequences were aligned with 41 sequences from this study and GenBank (Table 1) representing major clades identified in Aime (2006) using MAFFT v7 (Katoh & Standley, 2013). The sequence alignment is available at: http://doi.org/10.7931/J2JM27JR. The concatenated alignment was run as a partitioned dataset with maximum likelihood as the phylogenetic criterion. The dataset was analysed using RAxML 3.0 (Stamatakis, 2006) as implemented in Geneious Pro v7.0.5 (Biomatters, http://www.geneious.com/) to find the best-scoring likelihood tree. The model of evolution specified was GTR+Γ+I. The bootstrap analyses were run with a random starting tree and rapid hill climbing and with 1000 maximum likelihood bootstrap replicates (MLBS). The supermatrix consisted of 2025 characters, including ITS 2 for 49 taxa, LSU for 51 taxa, and B-tub for 18 taxa (Table 1). Inoculation of Rheum rhabarbarum, Muehlenbeckia complexa, and M. axillaris. Aeciospores of Aecidium otagense collected on 1 December 2014 were used to inoculate two rust-free Rheum ×rhabarbarum plants and two rust-free Muehlenbeckia complexa plants on 9 This article is protected by copyright. All rights reserved. December 2014. The plants were dusted with spores and placed in a mist chamber with 90% Accepted Article humidity at 23 ºC for 24 hours. Plants were then moved to a greenhouse with natural light and watered regularly. Urediniospores from an infected Rheum ×rhabarbarum plant collected on 20 January 2015 were used to inoculate two uninfected Muehlenbeckia australis and two M. complexa plants on 21 January as outlined above. Results Morphological characters Aeciospores on Clematis were subglobose, ellipsoidal to somewhat angular, 24.6 × 21.2 µm, with pale golden-yellow contents, and with readily detached plugs present on the spore wall (Fig 2e, h). The urediniospores from collections on Muehlenbeckia, Rumex, and Rheum were all obovoid to subglobose or ellipsoid in shape (Fig. 2b–d, g). The average urediniospore measurements were 25.9 × 20.6 µm on Rumex, 27.7 × 20.2 µm on Rheum, and 27.5 × 22.4 µm on Muehlenbeckia, with 3–5 germ pores. The teliospores from collections on Rumex and Muehlenbeckia averaged 26.8 × 20.7 µm and 28.6 × 19.4 µm, respectively, and were at least apically echinulate (Fig. 2a, f). Data matrix and phylogenetic analysis The two isolates each of the rust on Clematis (Aecidium otagense), Rumex (Puccinia sp.), Rheum (Puccinia rhei-undulati sensu auct. NZ), and three isolates on Muehlenbeckia (Puccinia tiritea) all had 99–100% identical sequences in the ITS2-LSU region. We amplified only the LSU sequence for an additional isolate of P. tiritea (PDD 101544) that This article is protected by copyright. All rights reserved. was 99.2% identical to the other nine LSU sequences. Partial ß-tubulin sequences were 99.6– Accepted Article 100% identical for nine isolates of the rust that were successfully amplified (Table 1). Phylogenetic analyses of the combined ITS2, LSU, and B-tub data using maximum likelihood recovered a highly supported Pucciniaceae (100% MLBS) (Fig. 3). Melampsoraceae (100% MLBS) was recovered in a highly supported clade (100% MLBS) with Pucciniastraceae (77% MLBS) and Coleosporiaceae (93% MLBS), which is similar to previously recovered topologies (e.g., McTaggart et al., 2015). The 10 isolates mentioned above were recovered in a highly supported clade (100% MLBS). Puccinia clavata P. Syd. & Syd., which occurs on Clematis foetida, was recovered in a well-supported clade with P. myrsiphylli (Thüm.) G. Winter and P. coronata Corda (80% MLBS). Uredo puawhananga G.T.S. Baylis, which occurs on Clematis paniculata, was recovered sister to Coleosporium tussilaginis (Pers.) Lév. (100% MLBS). The ITS primer pairs had variable success with amplifying the ITS region from the rust species (Suppl. Table 1). The primer pairs, ITS 1rustF10d/ ITS 4 and Rust ITS 1F/ Rust ITS 2R, had the best success at amplifying the correct product at 30% and 60%, respectively. The three other primer pairs, ITS 1F/ ITS 4, ITS 1F/ ITS 4B, and ITS 1F/ ITS 4BR, either did not amplify any products, produced low quality sequences, or preferentially amplified contaminants (Suppl. Table 1). As a result of these variable amplifications, the ITS2 region that was amplified with the Rust2inv and LR6 primers was utilised for the phylogenetic analyses. This article is protected by copyright. All rights reserved. Accepted Article Inoculation studies Two weeks after the inoculation with Aecidium otagense, a few uredinia with typical urediniospores were observed on Rheum plants. Additional uredinia were observed on the same Rheum plants five weeks after inoculation. No pustules were observed on the Muehlenbeckia plants inoculated with A. otagense or with urediniospores from R. ×rhabarbarum. Taxonomy Puccinia otagensis (Linds.) McKenzie & Padamsee, comb. nov. MycoBank MB818525 ≡ Aecidium otagense Linds., Transactions of the Royal Society of Edinburgh 24: 430, 1867. Type: Dunedin, East Taieri Bush, on Clematis paniculata, 5 Nov 1861, W.L. Lindsay (lectotype of Aecidium otagense, designated by McKenzie, 1981); Rangitikei, Pohangina Valley, on Muehlenbeckia australis, 21 May 2015, E.H.C. McKenzie (PDD 107783—II, III; epitype designated here of Aecidium otagense). Note: An epitype is designated here that has urediniospores and teliospores of Puccinia otagensis and from which molecular sequence data was obtained. = Puccinia tiritea G. Cunn., Transactions and Proceedings of the New Zealand Institute 54: 654 [Latin 698], 1923. Type: Wellington, Palmerston North, Tiritea, 300 m, on Muehlenbeckia australis, 3 Mar 1921, G.H. Cunningham (PDD 382—II, III, holotype of Puccinia tiritea). Misdetermined: “Puccinia rhei-undulati” sensu auct. NZ (non P. rhei-undulati (Diet.) Hirats. 1935). This article is protected by copyright. All rights reserved. On Clematis spp.: Accepted Article Spermagonia associated with aecia, immersed, scattered. Aecia amphigenous and on stems, petioles, and sepals, aggregated on inflated distorted areas up to 16 cm long and 0.5–1 cm diam., yellow; peridia white-cream, cupulate, lacerated and strongly revolute (Fig. 1a, b). Aeciospores (19.5–)22–28(–31) × (17.5–)19–23.5(–26.5) μm (mean 24.6 × 21.2 μm), subglobose, ellipsoidal to somewhat angular, contents pale golden-yellow; wall (0.8–)1–1.5(– 1.7) μm thick, hyaline, verruculose, readily detached plugs present on the wall (Fig. 2e, h). On Muehlenbeckia spp.: [Spermagonia (not observed, although described by Cunningham, 1923) as epiphyllous, in circular groups seated on yellow spots, minute, immersed, honey-brown.] Uredinia hypophyllous, scattered or sometimes aggregated on brown spots up to 2 mm diam., pale cinnamon, circular or elliptical, mainly 0.25–0.4 × 0.25 mm, pulverulent, surrounded by the epidermis (Fig. 1e). Urediniospores 24–33.5 × (19–)20–25.5 μm (mean 27.5 × 22.4 μm), obovoid, subglobose, or ellipsoidal, contents pale cream; wall 1–1.5 μm thick, cream coloured, coarsely and sparsely echinulate, germ pores (3–)4(–5), equatorial or slightly superior, with conspicuous caps (Fig. 2b, g). Telia hypophyllous, scattered, cinnamon, circular or elliptical, mainly 0.25–0.4 × 0.25 mm, pulverulent. Teliospores (23–)25.5–31(–33) × (16–)17–21(–23) μm (mean 28.6 × 19.4 μm), ellipsoidal, not or slightly constricted at septum, straight or slightly curved, wall 1.25–2.5 μm thick, not thickened at apex, coarsely warted or echinulate, more prominent apically, pale cinnamon, contents pale cinnamon, apex rounded, pore in upper cell immediately above septum, pore in lower cell immediately below septum; pedicel usually very short (ca. 5–10 × 4.5–6 μm) but sometimes up to 50 μm long, hyaline (Fig. 2a, f). This article is protected by copyright. All rights reserved. On Rheum ×rhabarbarum: Accepted Article Uredinia amphigenous, mainly hypophyllous, scattered, amber, circular, up to 2 mm diam., surrounded by ruptured epidermis, pulverulent, often seated on discoloured spot, which extends to opposite leaf surface and associated with leaf chlorosis (Fig. 1d). Urediniospores (22.5–)25–31(–35) × (17–)19–22(–23.5) μm (mean 27.7 × 20.2 μm), obovoid, broadly ellipsoidal or subglobose, contents dark cream; wall 1–2 μm thick, pale brown, echinulate, germ pores 3–4, conspicuous, equatorial (Fig. 2c). Teliospores not observed. On Rumex sagittatus: Uredinia amphigenous, mainly hypophyllous, golden-brown, pulverulent, up to 0.25 mm diam., circular, on red spots, with a small yellow centre, spot also visible on the upper surface (Fig. 1c). Urediniospores 22.5–29(–33) × (16.5–)18.5–22(–25) μm (av. 25.9 × 20.6 μm), broadly ellipsoid, obovoid, or subglobose; wall 1–1.5 μm thick, golden yellow, echinulate, germ pores 4(–5), equatorial or in upper half of spore, with inconspicuous caps, spore contents pale (Fig. 2d). Telia similar to the uredinia, dark brown. Teliospores 25–27.5(–29.5) × 19–22(–23) μm (av. 26.8 × 20.7 μm), ellipsoid, slightly constricted at septum, apex rounded; wall 1.5–2.5 μm thick, golden-yellow, upper cell warted, lower cell smooth, contents golden-yellow; pedicels up to 13 × 7 μm, persistent, hyaline, thin-walled. Habitat and Distribution. Known from Clematis afoliata, C. foetida, C. forsteri, C. hookeriana, C. marata, C. paniculata, Muehlenbeckia australis, M. axillaris, M. axillaris × ephedrioides, M. complexa, Rheum ×rhabarbarum, and Rumex sagittatus. This rust species occurs on the North and South Island of New Zealand, Stewart Island, and the Chatham Islands. This article is protected by copyright. All rights reserved. Additional specimens examined (I = aeciospores, II = urediniospores, III = teliospores). Accepted Article On Clematis foetida, Southland, North of Tuatapere, Clitden, Lonekers Scenic Reserve, 8 Jan 1997, K. Holyoake (PDD 66912—I). On C. forsteri, Taupo, 23 Apr 1965, J. McLean (PDD 24568—I). Wellington, SW coast between Makara and Opau Bay, 16 Oct 2011, P.J. Garnock-Jones, S.B. Malcolm, (PDD 102314—I). On C. paniculata, Dunedin, Dunedin Botanic Gardens, 21 Oct 2014, K. Caldwell (PDD 104485—I). On Muehlenbeckia australis, Waikato, Torehape, 22 May 1998, E.H.C. McKenzie, P.R. Johnston & R.E. Beever (PDD 69385—II, III). Mid Canterbury, Glentui Conference Centre, 5 May 2010, E.H.C. McKenzie (PDD 99223—II, III). Chatham Islands, Te One, 31 Mar 1993, E.H.C. McKenzie, P.R. Johnston (PDD 62197—II). On M. complexa, Northland, Waipoua State Forest, 22 Apr 1964, R.F.R. McNabb (PDD 23276—III). Wellington, Seatoun, seashore, 28 Mar 1922, G.H. Cunningham (PDD 10070— II, III). On Rheum ×rhabarbarum, Auckland, Mt Albert, Stilwell Road, 1 Dec 2010, P. Wilkie (PDD 101523—II). Mid Canterbury, Christchurch, Riccarton, 22 Apr 1997, A.J. Healy (PDD 68241—II). Chatham Islands, Waitangi, 6 Apr 1993, E.H.C. McKenzie & P.R. Johnston (PDD 62725—II). On Rumex sagittatus, Auckland, Mt Albert, Lloyd Ave, 23 Jun 1996, E.H.C. McKenzie (PDD 66405—II, III); Mt Albert, Summit Drive, 28 Aug 1996, J.P. Wilkie (PDD 66388—II); Mt Albert, Lloyd Ave, 30 Aug 2015, E.H.C. McKenzie (PDD 105389—II). Wellington, Waikanae Beach, 11 May 2009, E.H.C. McKenzie (PDD 97462—II). Notes— This article is protected by copyright. All rights reserved. The detachable plugs on the wall of the aeciospores of Puccinia otagensis on Clematis Accepted Article spp. are similar to those seen in aeciospores of Uromyces scaevolae G. Cunn. and Puccinia lagenophorae Cooke. Cunningham (1931) called them ‘cellulose plugs’. Similar plugs were also described for aeciospores of Allodus podophylli (Schwein.) Arthur in the USA and termed ‘pore plugs’ (Dodge, 1924). Their function and composition remains unknown although Dodge suggested that the pore plugs may act as a fulcrum when a spore breaks loose from its neighbours. In an examination of thin sections of aeciospores by transmission electron microscopy, von Hofsten & Holm (1968) found similar refractive bodies in two species, Gymnosporangium libocedri (Henn.) F. Kern and P. caricina. The spore wall under the plugs was considerably thinner than elsewhere, but had no relation to the spore pores. Scanning electron micrographs of these deciduous plugs were illustrated in aeciospores of several species of Puccinia and Uromyces by Zwetko & Blanz (2012). The urediniospores of Puccinia rhei-undulati, as described by Dietel (1906) and Hiratsuka et al. (1992), are similar in size to those of P. otagensis, and the urediniospores of both fungi have the same number of equatorial germ pores. The teliospores of P. rhei-undulati are, however, slightly longer and narrower than those of P. otagensis. In addition, the teliospores of P. rhei-undulati are smooth-walled with the apical wall 5–10 µm thick (Dietel, 1906; Hiratsuka et al., 1992), while those of P. otagensis are warted, especially towards the apex, and the apical wall is not thickened being only 1.25–2.5 µm thick. Discussion The lifecycle of Puccinia otagensis is unusual in that it has all eudicot hosts. Commonly heteroecious rust species, such as Puccinia graminis Pers., have monocots as primary (i.e., telial) hosts (e.g., Triticum) and eudicots (dicots) as alternate hosts (e.g., Berberis), or in the case of Cronartium ribicola J.C. Fisch. have eudicots as primary hosts (e.g., Ribes) and This article is protected by copyright. All rights reserved. gymnosperms as alternate hosts (e.g, Pinus). Interestingly there is another heteroecious Accepted Article species, Puccinia septentrionalis Juel, which is circumboreal in distribution that has teliospores on Polygonum spp. (Polygonaceae) and aeciospores on Thalictrum spp. (Ranunculaceae) (Wilson & Henderson, 1966), the same host families that are parasitized by P. otagensis. Inoculation studies demonstrated that infection was possible with aeciospores from Clematis paniculata to Rheum ×rhabarbarum; however, we were unable to demonstrate infection on Muehlenbeckia complexa plants possibly due to the small (5–20 × 2–15 mm) and coriaceous leaves, among other reasons. Additionally we were unable to infect Muehlenbeckia with urediniospores from R. ×rhabarbarum, which may suggest that infection of the primary host may only occur via aeciospores from Clematis spp. or reinfection via urediniospores from infected Muehlenbeckia. It may be possible that infection on Rumex sagittatus plants can only be caused by infection by aeciospores from Clematis spp. or urediniospores from other infected R. sagittatus or Rheum ×rhabarbarum, but we were unable to investigate this. We also did not investigate whether Clematis spp. could be infected via basidiospores produced by germinating teliospores from infected Rumex sagittatus or Muehlenbeckia spp. Although it would be intriguing to further explore the infection biology of Puccinia otagensis, DNA sequences from three different loci effectively demonstrate that the rusts on all these hosts are the same species. Collections of P. otagensis had similarly identical ITS2LSU sequences as did Mikronegeria fuchsiae, i.e., >99%. In a report on the alternate host of Puccinia striiformis Westend. (Jin et al., 2010) DNA sequences were utilised after each successful inoculation to confirm that the infections observed were caused by P. striiformis. It can be time-consuming and certainly impractical in the case of rare plant hosts to conduct inoculations. Using DNA sequence data or a phylogenetic species concept are robust methods This article is protected by copyright. All rights reserved. for taxonomic purposes, which may assist in the speedy resolution of whether rust spore Accepted Article stages on different hosts represent the same species. This would expedite giving rust fungi one name according to the current International Code of Nomenclature for algae, fungi, and plants (Melbourne Code) adopted in 2011 (McNeill et al., 2012). The nuclear ribosomal ITS locus has been selected as the universal barcode region for fungi (Schoch et al., 2012); however, this region is difficult to amplify correctly for the rust fungi due to nucleotide repeats or introns (Alaei et al., 2009) and there may be multiple types present (Virtudazo et al., 2001). As a result, there are multiple ITS primer combinations developed for the rust fungi (e.g., Vialle et al., 2009; Toome & Aime, 2015). Our study demonstrated that direct amplification (without cloning), even using rust-specific ITS primers, resulted in the preferential amplification of non-rust DNA (Suppl. Table 1). If the ITS region is required, it may be necessary to alter the DNA extraction and PCR protocols (e.g., Alaei et al., 2009), to ensure that the correct product is amplified. Otherwise the ITS 2 region plus the 5’ end of the LSU should be sufficient for species separation at present. Historically, the divergence of rust fungi was thought to be tightly linked to the evolution of their plant hosts (e.g., Savile, 1971). This raises an interesting question as to the origin of Mikronegeria fuchsiae because the aecial host, Phyllocladus, evolved 6.3 million years ago and the telial host, Fuchsia, evolved 30 million years ago (Wallis & Trewick, 2009). It also remains unclear whether P. otagensis evolved firstly on Clematis and then on Muehlenbeckia or vice versa. The presence of two other, unrelated rust species on Clematis in New Zealand may suggest that Clematis is the ancestral host. Recent studies (e.g., McTaggart et al., 2015) have demonstrated that host jumping appears to have a dominant influence in diversification of rust species. In the case of Puccinia otagensis, we can be reasonably confident that the species spread to Rheum ×rhabarbarum This article is protected by copyright. All rights reserved. and Rumex sagittatus in New Zealand within the last 60–150 years, which was when these Accepted Article plants were putatively introduced into the country. It is also intriguing that both of these host jumps have occurred on Polygonaceae species, possibly through ecological opportunities (Savile, 1971), as Rheum and Rumex sagittatus are associated with human activities and Muehlenbeckia spp. are widely cultivated as hedges. Telial stages appear to have wider host ranges in certain species, including P. otagensis, which may be the result of their plasticity (McTaggart et al., 2015). The host jump onto Rheum is peculiar in that only urediniospores have been observed on this host despite rhubarb being of economic importance as an edible plant and it having a longer history within New Zealand than Rumex sagittatus. Even though rust fungi are mostly host-specific, there are at least two other rust species on Clematis spp. in New Zealand and in particular, C. forsteri is host to Puccinia otagensis, P. clavata, and Uredo puawhananga G.T.S. Baylis. Uredo puawhananga on C. paniculata was recovered in a clade with Coleosporium and P. clavata on Clematis foetida was recovered in a clade with P. myrsiphylli in this study. The three rust species on Clematis spp. in New Zealand can be separated morphologically, but it certainly cautions one to rely solely on host identity for the identification of the rust fungus. Figure Legends Figure 1. Macroscopic images of Puccinia otagensis. a) Aecidium otagense (as A. tragopogonis) by John Buchanan (ca. 1875) illustrating distortion of Clematis stem, aecia, spermagonia, and aeciospores (protospores). Courtesy of Auckland War Memorial Museum Tāmaki Paenga Hira, MS-41); b) Aecidium otagense on Clematis paniculata; c) Puccinia sp. on Rumex sagittatus; d) Puccinia rhei-undulati on Rheum ×rhabarbarum; e) Puccinia tiritea on Muehlenbeckia australis. This article is protected by copyright. All rights reserved. Figure 2. Light and scanning electron microscopic images of Puccinia otagensis. a) Accepted Article Teliospores on Muehlenbeckia australis; b) urediniospores on M. australis; c) urediniospores on Rheum ×rhabarbarum; d) urediniospores on Rumex sagittatus; e) aeciospores on Clematis paniculata; f) teliospores on M. australis; g) urediniospores on M. australis; h) aeciospores with spore plugs on C. forsteri. Scale bars = 10 µm. Figure 3. Phylogram obtained from maximum likelihood analysis of nuclear rDNA loci, LSU and ITS2, and ß-tubulin sequences. The topology was rooted with Mikronegeria fuchsiae. Bootstrap support values (> 60%) from a maximum likelihood search with 1000 replicates shown. Acknowledgements This research was supported through the Landcare Research Systematics Portfolio, with funding from the Science and Innovation Group of the New Zealand Ministry of Business, Innovation and Employment. We thank Cissy Pan for help with the figures and Chris Winks for taking care of the plants. We are grateful to Zoe Richardson at Auckland War Memorial Museum for her assistance with the John Buchanan plate. The authors confirm that there are no conflicts of interest. References Aime MC, 2006. Toward resolving family-level relationships in rust fungi (Uredinales). Mycoscience 47, 112–22. This article is protected by copyright. All rights reserved. Alaei H, De Backer M, Nuytinck J, Maes M, Höfte M, Heungens K, 2009. Phylogenetic Accepted Article relationships of Puccinia horiana and other rust pathogens of Chrysanthemum × morifolium based on rDNA ITS sequence analysis. Mycological Research 113, 668–83. Barnes C, Szabo L, 2007. Detection and identification of four common rust pathogens of cereals and grasses using real-time polymerase chain reaction. Phytopathology 97, 717–27. Crane PE, Peterson RS, 2007. Mikronegeria fuchsiae sp. nov., a rust fungus on Fuchsia and Phyllocladus in New Zealand. New Zealand Journal of Botany 45, 707–13. Cunningham GH, 1923. The Uredinales, or rust-fungi, of New Zealand: Part I— Pucciniaceae, tribe Puccineae (containing descriptions and illustrations of seventy-five species). Transactions and Proceedings of the New Zealand Institute 54, 619–704. Cunningham GH, 1924. Second supplement to the Uredinales of New Zealand. Transactions and Proceedings of the New Zealand Institute 45, 392–6. Cunningham GH, 1926. Fourth supplement to the Uredinales and Ustilaginales of New Zealand. Transactions and Proceedings of the New Zealand Institute 57, 186. Cunningham GH, 1931. The rust fungi of New Zealand. John McIndoe, Dunedin. Cunningham GH, 1945. Additions to the rust fungi of New Zealand, I. Transactions of the Royal Society of New Zealand 75, 324–7. Dietel PV, 1906. Beschreibungen einiger neur Uredineen. Annales Mycologici 4, 303–8. Dodge BO, 1924. Expulsion of aecidiospores by the may-apple rust, Puccinia podophylli Schw. Journal of Agricultural Research 28, 923–6. Gardes M, Bruns TD, 1993. ITS primers with enhanced specificity for basidiomycetes‐application to the identification of mycorrhizae and rusts. Molecular Ecology 2, 113–8. Hiratsuka N, Sato S, Katsuya K, et al., 1992. The rust flora of Japan. Tsukuba Shuppankai. This article is protected by copyright. All rights reserved. von Hofsten A, Holm L, 1968. Studies on the fine structure of aeciospores I. Grana 8, 235– Accepted Article 51. Ito S, 1950. Basidiomycetes. No. 3. Uredinales-Pucciniaceae. Uredinales Imperfecti. Tokyo: Yokendo. Jin Y, Szabo LJ, Carson M, 2010. Century-old mystery of Puccinia striiformis life history solved with the identification of Berberis as an alternate host. Phytopathology 100, 432–5. Katoh K, Standley DM, 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution 30, 772–80. Kirk TW, 1906. Divisions of Biology and Horticulture, and Publications. Report of T.W. Kirk, Biologist, Chief of Divisions. New Zealand Department of Agriculture, Annual Report 14, 337–462. Kirk TW, 1908. Division of Biology and Horticulture. Report of T.W. Kirk, Biologist, Chief of Division. New Zealand Department of Agriculture, Annual Report 16, 97–162. Lindsay WL, 1867. Observations on new lichens and fungi collected in Otago, New Zealand. Transactions of the Royal Society of Edinburgh 24, 407–56. Lindsay WL, 1868. II.—List of Fungi collected in Otago, New Zealand. Transactions and Proceedings of the Botanical Society of Edinburgh 9, 13–22. McAlpine D, 1906. The Rusts of Australia. Melbourne, Australia: Robt. S. Brain, Government Printer. McKenzie EHC, 1981. New Zealand rust fungi: additions and corrections. New Zealand Journal of Botany 19, 227–32. McKenzie EHC, 1998. Rust fungi of New Zealand—an introduction, and list of recorded species. New Zealand Journal of Botany 36, 233–71. This article is protected by copyright. All rights reserved. McNeill J, Barrie FR, Buck WR, et al., 2012. International Code of Nomenclature for algae, Accepted Article fungi, and plants (Melbourne Code). Regnum Vegetabile 154. Oberreifenberg, Germany: Koeltz Scientific Books. McTaggart AR, Shivas RG, Nest MA, Roux J, Wingfield BD, Wingfield MJ, 2015. Host jumps shaped the diversity of extant rust fungi (Pucciniales). New Phytologist 209, 1149– 58. van der Merwe M, Ericson L, Walker J, Thrall PH, Burdon JJ, 2007. Evolutionary relationships among species of Puccinia and Uromyces (Pucciniaceae, Uredinales) inferred from partial protein coding gene phylogenies. Mycological Research 111, 163–75. Moncalvo J-M, Wang H-H, Hseu R-S, 1995. Phylogenetic relationships in Ganoderma inferred from the internal transcribed spacers and 25S ribosomal DNA sequences. Mycologia 87, 223–38. Sampson PJ, Walker J, 1982. An annotated list of plant diseases in Tasmania. Department of Agriculture, Tasmania. Savile DBO, 1971. Coevolution of the rust fungi and their hosts. Quarterly Review of Biology 46, 211–8. Schoch CL, Seifert KA, Huhndorf S, et al., 2012. Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi. Proceedings of the National Academy of Sciences 109, 6241–6. Shivas RG, 1989. Fungal and bacterial diseases of plants in Western Australia. Journal of the Royal Society of Western Australia 72, 1–62. Stamatakis A, 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688–90. Toome M, Aime M, 2015. Reassessment of rust fungi on weeping willows in the Americas and description of Melampsora ferrinii sp. nov. Plant Pathology 64, 216–24. This article is protected by copyright. All rights reserved. Vialle A, Feau N, Allaire M, et al., 2009. Evaluation of mitochondrial genes as DNA barcode Accepted Article for Basidiomycota. Molecular Ecology Resources 9, 99–113. Vilgalys R, Hester M, 1990. Rapid genetic identification and mapping of enzymatically amplified ribosomal DNA from several Cryptococcus species. Journal of Bacteriology 172, 4238–46. Virtudazo EV, Nakamura H, Kakishima M, 2001. Ribosomal DNA-ITS sequence polymorphism in the sugarcane rust, Puccinia kuehnii. Mycoscience 42, 447–53. Wallis GP, Trewick SA, 2009. New Zealand phylogeography: evolution on a small continent. Molecular Ecology 18, 3548–80. White TJ, Bruns T, Lee S, Taylor J, 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis, MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR Protocols: A Guide to Methods and Applications. New York, NY: Academic Press, 315–22. Wilson M, Henderson DM, 1966. British rust fungi. Cambridge: Cambridge University Press. Zwetko P, Blanz P, 2012. Aeciospore types in rusts on Ranunculus and allied genera. Stapfia 96, 105–121. Table 1. Species of rust, accession number, and GenBank numbers of taxa included in the analyses. Accession Taxon name number ITS2 + LSU B-Tub Aecidium myopori PDD 93248 KX985769a Aecidium otagense PDD 104485 KX985737b KX985778b Aecidium otagense PDD 102314 KX985741b KX985780b insignis PDD 101515 KX985751 KX985788 Chrysomyxa reticulata PDD 92535 KX985767 Chrysomyxa sp. PDD 94468 KJ716349c Aecidium ranunculi- This article is protected by copyright. All rights reserved. PDD 98309 KJ716348 Coleosporium tussilaginis PDD 93250 KX985766 Kuehneola uredinis PDD 101520 KX985770 Melampsora hypericorum PDD 97325 KJ716353 Melampsora ricini PDD 98363 KJ716352 Mikronegeria fuchsiae PDD 94465 KX985771 Mikronegeria fuchsiae PDD 101517 KJ716350 Mikronegeria fuchsiae PDD 101516 KX985772 Mikronegeria fuchsiae PDD 97448 KX985773 Miyagia pseudosphaeria PDD 97677 KX985753 Phragmidium mexicanum PDD 99249 KX985774 Phragmidium violaceum PDD 99246 KJ716351 Puccinia caricina PDD 98313 KX985750 Puccinia clavata PDD 74903 KX985761 Puccinia coronata PDD 101513 KX985764 Puccinia hieracii PDD 98711 KX985752 Puccinia hordei PDD 101656 KX985762 Puccinia junci PDD 99243 KX985745 KX985787 Puccinia malvacearum PDD 101511 KX985756 KX985792 Puccinia menthae PDD 99229 KJ716344 KX985789 Puccinia myrsiphylli PDD 99278 KX985760 PDD 94556 KX985754 PDD 104489 KX985743b KX985783b PDD 101523 KX985739b KX985784b PDD 66404 KX985740b sagittatus PDD 97462 KX985744b KX985785b Puccinia tiritea PDD 101546 KX985738b KX985782b Puccinia tiritea PDD 97495 KX985742b KX985779b Accepted Article Coleosporium tussilaginis KX985793 KX985794 Puccinia pelargoniizonalis Puccinia rhei-undulati sensu auct. NZ Puccinia rhei-undulati sensu auct. NZ Puccinia sp. on Rumex sagittatus Puccinia sp. on Rumex This article is protected by copyright. All rights reserved. PDD 101544 KX985775b,d KX985781b Puccinia tiritea PDD 107783 KX9857362 KX985777b Puccinia unciniarum PDD 99245 KX985746 KX985786 Puccinia urticata PDD 95256 KX985748 Puccinia urticata PDD 95255 KX985749 lapsanae PDD 98713 KX985747 Puccinia wahlenbergiae PDD 99238 KX985758 Pucciniastrum guttatum PDD 91889 KJ716345 Pucciniastrum myosotidii PDD 92567 KJ716346 Tranzschelia discolor PDD 92021 KX985768 Uredo puawhananga PDD 101549 KX985776d Uredo puawhananga PDD 93531 KX985765 Uredo toetoe PDD 99175 KX985759 Uromyces beticola PDD 101534 KX985757 Uromyces magnusii PDD 94487 KX985755 Uromyces rumicis PDD 93529 KX985763 Uromyces viciae-fabae PDD 101522 KJ716343 Accepted Article Puccinia tiritea Puccinia variabilis var. KX985791 KX985790 a GenBank numbers in bold generated for this study. b Sequences of Puccinia otagensis deposited to GenBank under listed names. c GenBank numbers not in bold from Padamsee & McKenzie. 2014. Phytotaxa 174: 223–230. d LSU sequence only. This article is protected by copyright. All rights reserved. Accepted Article This article is protected by copyright. All rights reserved. Accepted Article This article is protected by copyright. All rights reserved. Accepted Article This article is protected by copyright. All rights reserved.