A co-evolutionary relationship exists between Endoraecium (Pucciniales) and its Acacia hosts in Australia.

Endoraecium is a genus of rust fungi that infects several species of Acacia in Australia, South-East Asia and Hawaii. This study investigated the systematics of Endoraecium from 55 specimens in Australia based on a combined morphological and molecular approach. Phylogenetic analyses were conducted on partitioned datasets of loci from ribosomal and mitochondrial DNA. The recovered molecular phylogeny supported a recently published taxonomy based on morphology and host range that divided Endoraecium digitatum into five species. Spore morphology is synapomorphic and there is evidence Endoraecium co-evolved with its Acacia hosts. The broad host ranges of E. digitatum, E. parvum, E. phyllodiorum and E. violae-faustiae are revised in light of this study, and nine new species of Endoraecium are described from Australia based on host taxonomy, morphology and phylogenetic concordance.

INTRODUCTION

Endoraecium (Pucciniales, Raveneliaceae) was established for two endocyclic species of rust fungi with pedicellate spores, E. acaciae (the type species) and E. hawaiiense, which were found on Acacia koa (Fabaceae, subfamily Mimosoideae) in Hawaii (Hodges & Gardner 1984). There are over 1 000 species of Acacia s.str. (hereafter referred to as Acacia) known in Australia (Murphy et al. 2010, Maslin 2013). The classification of Acacia has moved from a traditional morphological one based on sections, to a molecular phylogenetic one based on clades (Murphy et al. 2010, Miller et al. 2013). At the commencement of this study, 20 species of Acacia were recorded as hosts of seven species of Endoraecium in Australia (Berndt 2011). These 20 host species fall within three clades of the plurinerved, uninerved and Botrycephalae (p.u.b.) group of Acacia (Murphy et al. 2010), namely the i) Botrycephalae subclade (Murphy et al. 2010); ii) Juliflorae p.p. clade (Miller et al. 2013); and iii) Plurinerves p.p. clade (Miller et al. 2013), hereafter referred to as Botrycephalae, Juliflorae and Plurinerves, respectively.

Species of Endoraecium produce a range of symptoms on Acacia that include bullate swellings or pulvinate sori on phyllodes, or galls and witches’ brooms on stems. These rust fungi produce spores of three types: i) spermogonia; ii) pedicellate spores, which have reticulate to foveolate ornamentations and were variously called aecial urediniospores (Walker 2001), uredo-like aeciospores (Berndt 2011) or teliospores in the endocyclic species (Hodges & Gardner 1984); iii) teliospores, which are single celled, subhyaline and smooth walled. The teliospores are morphologically similar to those of Uromyces s.l. and several species of Endoraecium were at one time classified in that genus, including U. digitatus (Winter 1886), U. bicintus (McAlpine 1906), U. phyllodiorum (McAlpine 1906), U. hyalosporus (Sawada 1913) and U. koae (Stevens 1925). On the basis of morphology, some species currently considered as Endoraecium were previously placed in other genera, including Pileolaria (Dietel 1921), Maravalia (Dietel 1924), Poliotelium (Mains 1939), Atelocauda (Cummins & Hiratsuka 1983, Ono 1984, Gardner 1991), and Racospermyces (Walker 2001).

Gardner (1991) proposed that Endoraecium may be closely related to Uromycladium (Pileolariaceae), another genus of rust that occurs on Acacia in Australia. However, in combined analyses of the Large Subunit (LSU) and Small Subunit (SSU) regions of ribosomal DNA (rDNA), the systematic position of Endoraecium was shown by Scholler & Aime (2006) and Aime (2006) to have an affinity with mimosoid rusts in the Raveneliaceae rather than with the Pileolariaceae.

In this study, the systematic relationships of Endoraecium on Acacia in Australia were investigated with freshly collected field material and herbarium specimens. Currently, Endoraecium contains 13 species, all on Acacia, with seven endemic to Australia, one to South-East Asia and five to Hawaii (McAlpine 1906, Sawada 1913, Hodges & Gardner 1984, Berndt 2011). The species are E. acaciae (type on A. koa, Hawaii), E. angustiphyllodium (type on A. koa, Hawaii), E. bicinctum (type on A. fasciculifera, Australia), E. digitatum (type on A. no-tabilis, Australia), E. hawaiiense (type on A. koa, Hawaii), E. hyalo-sporum (type on A. confusa, Taiwan), E. kauaianum (type on A. koa, Hawaii), E. koae (type on A. koa, Hawaii), E. parvum (type on A. leiocalyx, Australia), E. phyllodiorum (type on Acacia sp., Australia), E. tierneyi (type on A. harpophylla, Australia), E. violae-faustiae (type on Acacia sp., Australia) and E. walkerianum (type on A. penninervis, Australia). The SSU, Internal Transcribed Spacer (ITS) and LSU regions of rDNA, and cytochrome c oxidase subunit 3 (CO3) of mitochondrial DNA, were used for phylogenetic reconstruction to determine whether the current taxonomy of Australian species was supported by molecular data and if further diversity existed within Endoraecium.

MATERIALS AND METHODS

Taxon selection

Fresh specimens were collected from New South Wales, the Northern Territory, Queensland, Victoria and Western Australia, representing a broad diversity of Acacia (Table 1). Species from representative genera of Raveneliaceae, namely Diorchidium, Kernkampella, Ravenelia and Sphaerophragmium, were selected as outgroup taxa for the phylogenetic analyses. Uromycladium simplex (Pileolariaceae) was included as an outgroup to the Raveneliaceae, as it represented another Australian endemic rust on Acacia.

Table 1

Specimens of Endoraecium included in this study.

Taxon	Accession number	Subclade1 of Acacia s.str.	Host	GenBank accession
				LSU	ITS	SSU	CO3
E. acaciae	BPI 871098	Plurinerves	A. koa	DQ3239162	N/A	DQ3239172	N/A
E. auriculiforme	BRIP 55609	Juliflorae	A. auriculiformis	KJ862296	KJ862353	N/A	KJ862430
	BRIP 56550	Juliflorae	A. auriculiformis	KJ862297	KJ862354	N/A	KJ862431
	BRIP 56548*	Juliflorae	A. auriculiformis	KJ862298	KJ862355	N/A	KJ862432
	BRIP 56549	Juliflorae	A. auriculiformis	KJ862299	KJ862356	N/A	KJ862433
E. carnegiei	BRIP 57926	Botrycephalae	A. dealbata	KJ862300	KJ862357	N/A	KJ862434
	BRIP 57924*	Botrycephalae	A. dealbata	KJ862301	N/A	N/A	KJ862435
	BRIP 59218	Botrycephalae	A. dealbata	KJ862302	N/A	N/A	N/A
E. disparrimum	BRIP 55659	Juliflorae	A. disparrima	KJ862303	N/A	KJ862402	KJ862436
	BRIP 55626*	Juliflorae	A. disparrima	KJ862304	KJ862358	KJ862403	KJ862437
	BRIP 55632	Juliflorae	A. disparrima	KJ862305	KJ862359	KJ862404	KJ862438
E. falciforme	BRIP 57583*	Botrycephalae	A. falciformis	KJ862306	KJ862360	KJ862405	KJ862439
	BRIP 57643	Botrycephalae	A. falciformis	KJ862307	N/A	N/A	N/A
E. hawaiiense	BPI 871064	Plurinerves	A. koa	DQ3239202	N/A	N/A	N/A
E. irroratum	BRIP 55671	Botrycephalae	A. irrorata	KJ862310	KJ862364	KJ862406	KJ862442
	BRIP 57279	Botrycephalae	A. irrorata	KJ862311	KJ862365	N/A	KJ862443
	BRIP 57286*	Botrycephalae	A. irrorata	KJ862312	KJ862366	KJ862407	N/A
	BRIP 58054	Botrycephalae	A. irrorata	KJ862313	N/A	N/A	N/A
E. koae	BPI 871071	Plurinerves	A. koa	DQ3239182	N/A	DQ3239192	N/A
E. maslinii	BRIP 57872*	Botrycephalae	A. daphnifolia	KJ862314	KJ862367	KJ862408	KJ862444
E. parvum	BRIP 57514	Juliflorae	A. leiocalyx	KJ862315	KJ862368	N/A	N/A
	BRIP 57524	Juliflorae	A. leiocalyx	KJ862316	KJ862369	KJ862409	KJ862445
	BRIP 57527	Juliflorae	A. leiocalyx	KJ862317	KJ862370	N/A	N/A
	BRIP 57534	Juliflorae	A. leiocalyx	KJ862318	KJ862371	N/A	N/A
	BRIP 57568	Juliflorae	A. leiocalyx	KJ862319	KJ862372	N/A	N/A
	BRIP 57631	Juliflorae	A. leiocalyx	KJ862320	KJ862373	KJ862410	KJ862446
	BRIP 57512	Juliflorae	A. leiocalyx	KJ862321	KJ862374	N/A	N/A
	BRIP 53616	Juliflorae	A. leiocalyx	KJ862322	KJ862375	N/A	N/A
E. peggii	BRIP 55602*	Juliflorae	A. holosericia	KJ862308	KJ862361	N/A	KJ862440
	BRIP 55631	Juliflorae	A. holosericia	KJ862309	KJ862362	N/A	KJ862441
	BRIP 58324	Juliflorae	A. holosericia	N/A	KJ862363	N/A	N/A
E. phyllodiorum	BRIP 57310	Juliflorae	A. aulacocarpa	KJ862323	KJ862377	N/A	N/A
	BRIP 57516	Juliflorae	A. aulacocarpa	KJ862324	KJ862378	KJ862411	KJ862447
	BRIP 57569	Juliflorae	A. aulacocarpa	KJ862325	KJ862379	N/A	N/A
	BRIP 57570	Juliflorae	A. aulacocarpa	KJ862326	KJ862380	N/A	N/A
	BRIP 57578	Juliflorae	A. aulacocarpa	KJ862327	KJ862381	KJ862412	KJ862448
	BRIP 57579	Juliflorae	A. aulacocarpa	KJ862328	N/A	KJ862413	N/A
	BRIP 57580	Juliflorae	A. aulacocarpa	KJ862329	KJ862382	N/A	N/A
	BRIP 57588	Juliflorae	A. aulacocarpa	KJ862330	KJ862383	N/A	N/A
	BRIP 57589	Juliflorae	A. aulacocarpa	KJ862331	KJ862384	N/A	N/A
	BRIP 57590	Juliflorae	A. aulacocarpa	KJ862332	KJ862385	N/A	N/A
E. podalyriifolium	BRIP 57294	Botrycephalae	A. podalyriifolia	KJ862333	KJ862386	N/A	N/A
	BRIP 57576*	Botrycephalae	A. podalyriifolia	KJ862334	KJ862387	KJ862414	KJ862449
E. tierneyi	BRIP 27071*	Plurinerves	A. harpophylla	KJ862335	N/A	KJ862415	KJ862450
	BRIP 27880	Plurinerves	A. harpophylla	N/A	KJ862388	N/A	N/A
	BRIP 27887	Plurinerves	A. harpophylla	N/A	KJ862389	N/A	N/A
E. tropicum	BRIP 56555	Juliflorae	A. tropica	KJ862336	KJ862390	KJ862416	KJ862451
	BRIP 56557*	Juliflorae	A. tropica	KJ862337	KJ862391	KJ862417	KJ862452
E. violae-faustiae	BRIP 55601	Juliflorae	A. aulacocarpa	KJ862338	KJ862392	KJ862418	KJ862453
	BRIP 55629	Juliflorae	A. aulacocarpa	KJ862339	KJ862393	KJ862419	KJ862454
	BRIP 55660	Juliflorae	A. aulacocarpa	KJ862340	N/A	KJ862420	KJ862455
	BRIP 56544	Juliflorae	A. difficillis	KJ862341	KJ862395	KJ862421	N/A
	BRIP 55616	Juliflorae	A. difficillis	N/A	KJ862396	KJ862422	KJ862456
	BRIP 55611	Juliflorae	A. difficillis	N/A	KJ862397	KJ862423	N/A
	BRIP 56539	Juliflorae	A. difficillis	KJ862342	KJ862398	KJ862424	KJ862457
	BRIP 56540	Juliflorae	A. difficillis	KJ862343	KJ862399	KJ862425	N/A
	BRIP 56545	Juliflorae	A. difficillis	KJ862344	KJ862400	KJ862426	N/A
	BRIP 56547	Juliflorae	A. difficillis	KJ862345	KJ862401	KJ862427	KJ862458
Kernkampella breyniae	BRIP 56909	NA	Breynia cernua (Euphorbiaceae)	KJ862346	N/A	KJ862428	KJ862459
Ravenelia neocaledoniensis	BRIP 56907	NA	Vachellia farnesiana	KJ862347	N/A	N/A	N/A
	BRIP 56908	NA	V. farnesiana	KJ862348	N/A	N/A	KJ862460
Ravenelia sp.	BRIP 56904	NA	Cassia sp.	KJ862349	N/A	N/A	KJ862461
Sphaerophragmium sp.	BRIP 56910	NA	Albizia sp.	KJ862350	N/A	KJ862429	KJ862462
Uromycladium simplex	BRIP 57571	Botrycephalae	A. pycnantha	KJ6329903	N/A	N/A	N/A

* Holotype

1 Botrycephalae = Botrycephalae subcade (Murphy et al. 2010); Juliflorae = Juliflorae p.p. clade (Miller et al. 2013); Plurinerves = Plurinerves p.p. clade (Miller et al. 2013).

2 Scholler & Aime (2006).

3 Doungsa-ard et al. (2014).

Morphology

Spores were scraped from leaf material, mounted in lactic acid and gently heated to boiling. Preparations were examined with a Leica DMLB microscope and images were taken with a Leica DFC500 camera. Measurements were made digitally from photographed spores. All previous taxonomic studies of Endoraecium have adopted an ontogenic approach to spore classification (Hodges & Gardner 1984, Walker 2001, Scholler & Aime 2006, Berndt 2011). Aeciospores, urediniospores and in some cases teliospores, are often indistinguishable with this approach. For this reason, we used a morphological approach to classify spores (Laundon 1967), which recognises urediniospores of Endoraecium as spores with reticulate ornamentation, pedicels and germ pores.

DNA extraction, PCR and sequencing

DNA was extracted from fresh or archived material. Spores were obtained from leaf material according to the protocol in McTaggart et al. (2014), and then extracted with the UltraClean Microbial DNA Isolation Kit (MoBio Laboratories Inc., Solana Beach, CA, USA).

The ITS region was amplified with ITS1F/ITS4B (Gardes & Bruns 1993). The ITS2-LSU region was amplified with Rust2inv (Aime 2006) / LR7 (Vilgalys & Hester 1990) and nested with LROR/LR6 (Vilgalys & Hester 1990). The SSU region was amplified with NS1 (White et al. 1990) / Rust 18S-R (Aime 2006) according to the protocol by Aime (2006). CO3 was amplified with the primers CO3_F1/CO3_R1 (Vialle et al. 2009). All PCRs were performed with high fidelity Phusion enzyme (New England Biolabs Inc.) according to the manufacturer’s instructions. The PCRs were performed with the following annealing temperatures: SSU, ITS and nested LSU at 62 °C, the initial LSU at 60 °C, and CO3 at 55 °C. PCR products were sent to Macrogen Korea for direct sequencing. Sequences were uploaded to GenBank under the accession numbers listed in Table 1.

Phylogenetic analyses

The SSU (19), ITS (47), LSU (51) and CO3 (29) sequences were aligned separately in MAFFT (Katoh et al. 2009) (available: http://www.ebi.ac.uk/Tools/msa/mafft/). Non-homologous regions in the ITS alignment were removed using GBlocks (Castresana 2000). The sequences from each locus were concatenated in a supermatrix and run as partitioned datasets with maximum likelihood and Bayesian inference as phylogenetic criteria. CO3 was translated and included as amino acid coding sequence. JTT was selected as a protein coding model of evolution in MEGA5 (Tamura et al. 2011). Maximum likelihood was implemented as a search criterion in RAxML (Stamatakis 2014). GTRGAMMA was specified as the model of evolution for nucleotide sequence data. The RAxML analyses were run with a rapid Bootstrap analysis (command -f a) using a random starting tree and 1 000 maximum likelihood bootstrap replicates. A Markov Chain Monte Carlo (MCMC) search in a Bayesian analysis was conducted with MrBayes (Ronquist & Huelsenbeck 2003). Four runs, each consisting of four chains, were implemented for 5 million generations. The cold chain was heated at a temperature of 0.25. Substitution model parameters were sampled every 500 generations and trees were saved every 1 000 generations. Convergence of the Bayesian analysis was confirmed using AWTY (Nylander et al. 2008) (available at: ceb.csit.fsu.edu/awty/) and a burn-in of 18 000 generations was calculated. The ML and Bayesian analyses were run three times to test accuracy. Alignments and trees were uploaded to TreeBASE (http://purl.org/phylo/treebase/phylows/study/TB2:S15410).

RESULTS

The topologies recovered by Bayesian inference and maximum likelihood were identical (Fig. 1). No differences were recovered in the tree topologies between nucleotide and amino acid coding sequences of CO3. Endoraecium was recovered as a monophyletic group sister to other members of the Raveneliaceae.

Fig. 1

Phylogram obtained in a maximum likelihood search in RAxML of the SSU, ITS, LSU and CO3 gene regions. Bootstrap support (≥ 70 %) values from 1 000 replicates above nodes. Posterior probabilities (≥ 0.95) from 4 200 trees in a Bayesian search are shown below nodes. Synapomorphic characters, namely host subclade and morphology of urediniospores, are mapped on to the topology.

Three major clades of Endoraecium corresponded to three clades in Acacia, namely Plurinerves, Juliflorae and Botrycepha-lae. The clade of Endoraecium that diversified on Plurinerves contained the Hawaiian species of Endoraecium and the Australian species, E. tierneyi. Rust taxa formed well-supported clades corresponding to host species. This is contrary to the current broad host ranges proposed for E. digitatum, E. parvum, E. phylliodorum and E. violae-faustiae (Berndt 2011).

Morphology of the urediniospores was a synapomorphic character for the two large clades on Juliflorae and Botrycephalae (Fig. 1). Urediniospores of species on Juliflorae were golden brown with reticulate to foveolate ornamentation. Urediniospores of those on Botrycephalae were subhyaline to pale yellow with reticulate ornamentation.

TAXONOMY

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808980; Fig. 2

Fig. 2

Endoraecium auriculiforme. a. Bullate sori on leaf (BRIP 56550); b, c. urediniospores (BRIP 55609); d. teliospores (BRIP 56550); e. germinating teliospore (BRIP 56549). — Scale bars: a = 1 cm; b–e = 10 μm.

Etymology. Name refers to the host, Acacia auriculiformis, on which it was found.

Type. AUSTRALIA, Northern Territory, Darwin, Howard Springs (-12.4564, 131.0522), on A. auriculiformis, 15 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas (holotype BRIP 56548).

Uredinia on both phyllode surfaces, subepidermal, erumpent, on bullate swellings up to 1 cm long, or pulverulent on phyllode surface, linear to round, yellowish brown; paraphyses intrasoral, cylindrical, apex 6–9-digitate, subhyaline to yellow, 35–70 × 10–14 μm, thickened at apex, pedicel up to 40 μm. Urediniospores obovoid, oval to fusiform, apex acute, yellowish brown, 31–46 × 16–20 μm; wall 3–5 μm thick at sides and 5–9 μm thick at apex, foveolate, with 3–4 equatorial germ pores; pedicel 3–5 μm, or absent. Telia formed from uredinia, erumpent, reddish brown. Teliospores cylindrical to fusiform, apex rounded, 2–4-digitate, hyaline to subhyaline, 36–54 × 13–25 μm; wall 1.5–2.5 μm thick at sides, 6–8 μm thick at apex; pedicel persistent, up to 38 μm. Basidia cylindrical, 3-septate, up to 35–40 × 8–12 μm. Basidiospores globose, hyaline, 4–6 μm, smooth-walled.

On phyllodes of Acacia auriculiformis (Juliflorae).

Additional specimens examined. AUSTRALIA, Northern Territory, Howard Springs, on A. auriculiformis, 9 May 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 55609; 15 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, R.G. Shivas, BRIP 56550; Northern Territory, Darwin, Bronzewing Ave, Howard Springs, 15 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt & R.G. Shivas, BRIP 56549.

Notes — Endoraecium auriculiforme has foveolate uredinio-spores with an acute and thickened apex, and digitate paraphyses that resemble narrow teliospores. Endoraecium violae-faustiae and E. peggii also produce similar paraphyses. These three species all occur on hosts in the Juliflorae, with E. auriculiforme restricted to A. auriculiformis; E. violae-faustiae to A. aula-cocarpa and A. crassicarpa, and E. peggii to A. holosericea. In the phylogenetic analyses, E. auriculiforme was recovered as sister to E. peggii (Fig. 1).

(McAlpine) M. Scholler & Aime, Myco-science 47: 163. 2006

Basionym. Uromyces bicinctus McAlpine, Rusts of Australia: 93. 1906.

≡ Pileolaria bicincta (McAlpine) Dietel, Ann. Mycol. 19: 302. 1921.

≡ Atelocauda bicincta (McAlpine) Cummins & Y. Hirats., Illustr. Gen. Rust Fungi, rev. edn (St. Paul): 147. 1983.

≡ Racospermyces bicinctus (McAlpine) J. Walker, Australas. Mycol. 20: 16. 2001.

Type. AUSTRALIA, Queensland, Rockhampton, on A. fasciculifera, 30 Nov. 1866, G.H. Robinson (holotype VPRI 5751).

Notes — Endoraecium bicinctum is known from four collections, which were examined by Walker (2001). It is only known to occur on A. fasciculifera in Botrycephalae (Murphy et al. 2010, Maslin 2013).

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808981; Fig. 3

Fig. 3

Endoraecium carnegiei. a. Telia on leaves (BRIP 59218); b. gall (BRIP 57924); c. urediniospores (BRIP 57924); d, e. teliospores (BRIP 57924). — Scale bars: b = 1 cm; c–e = 10 μm.

Etymology. Named after the Australian forest pathologist and fungal taxonomist, Dr Angus J. Carnegie, who generously assisted this study by collecting many specimens of rust on Acacia, including the type specimen on which this taxon is based.

Type. AUSTRALIA, New South Wales, Ando, Glen Allen State Forest, Poddy Hut Road (-36.7150, 149.3486), on A. dealbata, 6 Oct. 2012, A.J. Carnegie (holotype BRIP 57924).

Uredinia forming galls on stems, irregular, up to 3 cm, yellowish brown. Urediniospores ellipsoidal to ovoid, subhyaline to yellow, 34–40 × 21–25 μm; wall 4.0–4.5 μm thick, uniform or slightly thicker at sides, with 6–8 equatorial germ pores. Telia on leaves, pulverulent, erumpent, elliptical, less than 1 mm, yellowish brown. Teliospores cylindrical to fusiform, apex rounded, 1–5-digitate, hyaline to subhyaline, 41–76 × 10–17 μm; wall 1.0–1.5 μm thick at sides, 6–19(–32) μm thick at apex; pedicel persistent, up to 51 μm.

On stems and leaves of A. dealbata (Botrycephalae).

Additional specimens examined. AUSTRALIA, New South Wales, Mila, Bondi State Forest (-37.0847, 149.1078), on A. dealbata, 4 Oct. 2012, A.J. Carnegie, BRIP 57926; Victoria, Kergunyah, Murramurrangbong Range, Simpson Road (-36.3142, 146.9819), on A. dealbata, 13 May 2013, C. Doungsa-ard, W. Khem-muk & A.D.W. Geering, BRIP 59218.

Notes — Endoraecium carnegiei is one of three species that forms galls on the stems of its host. The other two species, E. digitatum and E. irroratum, also infect species of Acacia in the Botrycephalae. Endoraecium carnegiei is only known to occur on A. dealbata in south-eastern Australia.

(G. Winter) M. Scholler & Aime, Mycoscience 47: 163. 2006.

Basionym. Uromyces digitatus G. Winter, Rev. Mycol. (Toulouse) 8: 209. 1886.

≡ Coeomurus digitatus (G. Winter) Kuntze (as ‘Caeomurus’), Revis. Gen. Pl. 3: 450. 1898.

≡ Atelocauda digitata (G. Winter) Cummins & Y. Hirats., Illustr. Gen. Rust Fungi, rev. edn (St. Paul): 147. 1983.

≡ Racospermyces digitatus (G. Winter) J. Walker, Australas. Mycol. 20: 13. 2001.

Type. AUSTRALIA, South Australia, near Gawler, on A. notabilis, 1 July 1885, J.G.O. Tepper (holotype SF35352); South Australia, west of Gawler, along road to Mallala (-34.5665, 138.7184), on A. notabilis, 16 Oct. 2009, V. Faust-Berndt & R. Berndt (epitype SF35352).

Notes — Our study shows that the known host range of E. digitatum s.str. is restricted to A. notabilis in Botrycephalae. Walker (2001) adopted a broad host range for E. digitatum, but suspected it was a complex of closely related taxa, confined to one or a small group of hosts. Berndt (2011) divided E. digitatum into five species in Australia. He accepted E. phyllodiorum as distinct from E. digitatum, and further described three new taxa, E. parvum, E. violae-faustiae and E. walkerianum. Berndt (2011) designated an epitype for E. digitatum on A. notabilis, and listed six species of Acacia as hosts. We found four of these hosts, A. dealbata, A. falciformis, A. irrorata and A. podalyriifolia, were infected by novel species of Endoraecium. The remaining two hosts, A. deanei and A. oshanesii, most likely represent two new hosts of Endoraecium. The urediniospores of E. digitatum were not examined in this study, however, the description of the epitype (Berndt 2011) is similar to other species of Endoraecium with subhyaline to golden yellow urediniospores and reticulate ornamentation on species of Acacia in Botrycephalae.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808982; Fig. 4

Fig. 4

Endoraecium disparrimum. a. Host (BRIP 55659); b, c. urediniospores (BRIP 55659); d. teliospores (BRIP 55626). — Scale bars: a = 1 cm; b–d = 10 μm.

Etymology. Name refers to the host, Acacia disparrima, on which it was found.

Type. AUSTRALIA, Queensland, Babinda (-17.3397, 145.8675), on A. dis-parrima, 3 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas (holotype BRIP 55626).

Uredinia on both phyllode surfaces, forming bullate swellings up to 1 cm long, or pulverulent, linear to round, up to 1 mm diam, yellowish brown. Urediniospores subglobose to ovoid, apex obtuse, yellowish brown, 29–41 × 20–26 μm; wall 2–4 μm thick at sides and 5.5–9.0 μm thick at apex, reticulate, with 2–5 equatorial germ pores; pedicel 2 μm or absent. Telia formed from uredinia, yellowish brown. Teliospores cylindrical to fusiform, apex acuminate to rounded, hyaline to subhyaline, 31–50 × 20–26 μm; wall 1–2 μm thick at sides, 4–14 μm thick at apex; pedicel persistent, up to 33 μm.

On phyllodes of A. disparrima (Juliflorae).

Additional specimens examined. AUSTRALIA, Queensland, Herberton, Silver Valley Drive, on A. disparrima, 10 Apr. 2012, R. Berndt & V. Faust-Berndt, BRIP 55659; Queensland, near Yungaburra, Danbulla State Forest, at the Chimney’s Rest Area, on A. disparrima, 9 Apr. 2012, R. Berndt & V. Faust-Berndt, BRIP 55632.

Notes — Endoraecium disparrimum occurs on A. disparrima in Juliflorae. It was recovered as sister to E. violae-faustiae, which occurs on closely related species of Acacia. It differs from E. violae-faustiae in that the urediniospores have an obtuse apex with reticulate rather than foveolate ornamentation, and paraphyses are absent.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808983; Fig. 5

Fig. 5

Endoraecium falciforme (BRIP 57583). a. Uredinia and telia surrounding spermogonia; b. uredinium; c. teliospores; d, e. urediniospores. — Scale bars: b = 1 cm; c–e = 10 μm.

Etymology. Name refers to the host, Acacia falciformis, on which it was found.

Type. AUSTRALIA, Queensland, Mt Tibrogargan (-26.9283, 152.9494), on A. falciformis, 15 Aug. 2012, C. Doungsa-ard & A.R. McTaggart (holotype BRIP 57583).

Spermogonia on fruit or phyllodes, amphigenous, subepidermal, erumpent, c. 100 μm. Uredinia on fruit or on both phyllode surfaces, causing distortion, surrounding spermogonia, subepidermal, erumpent, pulverulent, linear, round to ellipsoidal, greater than 10 mm, reddish brown. Urediniospores ellipsoidal to ovoid, subhyaline to yellow, 34–43 × 21–25 μm; wall unevenly thickened at equator, 3.0–6.5 μm and 3–5 μm thick at apex, reticulate, with 4–9 equatorial germ pores; pedicel 2 μm. Telia on both phyllode surfaces, formed separately from uredinia, erumpent, c. 1 mm, orange to yellowish brown. Teliospores cylindrical to clavate, apex rounded, 2–5-digitate, subhyaline, 47–63 × (13–)15–23 μm; wall 1.5–2.0 μm thick at sides, 12–24 μm thick at apex; pedicel persistent, up to 64 μm long × 6–8 μm thick.

On phyllodes and fruit of A. falciformis (Botrycephalae).

Additional specimen examined. AUSTRALIA, Queensland, Mt Tibrogargan, 4 Sept. 2012, K.M. Thomson, BRIP 57643.

Notes — Endoraecium falciforme is morphologically similar to E. walkerianum, which both occur on species of Acacia species in Botrycephalae. Endoraecium falciforme has distinctively thicker walls at the equator of the urediniospores (3.0–6.5 μm) than E. walkerianum (3.0–4.5 μm). Endoraecium falciforme is recorded from south-east Queensland, whereas E. walkerianum is known from southern Australia. Urediniospores from a paratype specimen of E. walkerianum (BRIP 14205) were found to have thickened apices, whereas those of E. falciforme were thickened at the equator, or the wall was of uniform thickness.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808984; Fig. 6

Fig. 6

Endoraecium irroratum. a. Stem gall (BRIP 55671); b, c. urediniospores (BRIP 55671); d. teliospore (BRIP 57286). — Scale bars: a = 1 cm; b–d = 10 μm.

Etymology. Name refers to the host, Acacia irrorata, on which it was found.

Type. AUSTRALIA, Queensland, Main Range National Park, on A. irrorata, 11 June 2012, A.D.W. Geering (holotype BRIP 57286).

Uredin ia forming galls on stems, up to 2 cm long, yellowish brown. Urediniospores ovoid, apex obtuse, subhyaline, yellow to yellowish brown, 32–42 × 17–24 μm; wall 2.0–3.5 μm thick at sides, apex sometimes thickened (up to 4 μm), reticulate, with 4–10 equatorial germ pores. Telia formed from uredinia, yellowish brown. Teliospores cylindrical to subfusiform, apex acuminate to rounded, 2–7-digitate, subhyaline to yellow, 47–73 × 15–22 μm; wall 1.0–1.5 μm thick at sides, 11–21 μm thick at apex; pedicel persistent, up to 46 μm.

On stems and phyllodes of A. irrorata (Botrycephalae).

Additional specimens examined. AUSTRALIA, New South Wales, Warrumbungle National Park, on A. irrorata, 20 Mar. 2012, R. Berndt & V. Faust-Berndt, BRIP 55671; Queensland, Mt Mee, Mt Mee State Forest, on A. irrorata, 3 June 2012, C. Doungsa-ard & R.G. Shivas, BRIP 57279; Queensland, Mt Glorious, Wivenhoe Dam lookout area (-27.3077, 152.7136), A. irrorata, 30 Nov. 2012, A.D.W. Geering, BRIP 58054.

Notes — Endoraecium irroratum occurs on A. irrorata in Botrycephalae. Berndt (2011) considered that the rust on A. irrorata was E. digitatum based on morphology. There are no clear morphological differences that separate E. irroratum from the epitype of E. digitatum (Berndt 2011). However, the results from the molecular phylogenetic analysis in this study indicate the rusts on Botrycephalae are each restricted to a single host species.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808985; Fig. 7

Fig. 7

Endoraecium maslinii (BRIP 57872). a–c. Bullate swellings on leaves; d, e. urediniospores. — Scale bars: c = 1 cm; d, e = 10 μm.

Etymology. Named after the Australian botanist Bruce R. Maslin, an expert on Acacia taxonomy, who has published approximately 250 Acacia taxa and kindly identified many specimens in this study.

Type. AUSTRALIA, Western Australia, Midlands Rd, 10 km SE of Mingenew (-29.2847, 115.5319), on A. daphnifolia, 28 Sept. 2012, C. Doungsa-ard & A.R. McTaggart (holotype BRIP 57872).

Uredinia forming bullate swellings on fruit and both phyllode surfaces, round, up to 1 cm, yellowish brown. Urediniospores ellipsoidal, ovoid to fusiform, apex obtuse, subhyaline to yellowish brown, 37–46 × 20–26 μm; wall 2.5–5.0 μm thick at side, apex slightly thickened, 3.5–6.0 μm, reticulate, with 6–8 equatorial germ pores; pedicel absent or up to 2.5 μm.

On phyllodes of A. daphnifolia (Botrycephalae).

Notes — Endoraecium maslinii occurs on A. daphnifolia in Botrycephalae and is the only species of Endoraecium known from Western Australia. Only urediniospores have been found, which are yellowish brown and reticulate, and morphologically similar to other species on Acacia in Botrycephalae.

Berndt, Mycol. Progr. 10: 510. 2011

Type. AUSTRALIA, Queensland, Caloundra, on A. leiocalyx, 25 Aug. 1932, S.T. Blake (holotype BRIP 7543!).

Uredinia on both phyllode surfaces, subepidermal, erumpent, pulverulent, linear to round, yellowish brown. Urediniospores globose, subglobose to obovoid, apex obtuse, yellowish brown to reddish brown, 25–38 × 16–22 μm; wall 3–4 μm thick at equator, hardly thickened at apex, reticulate, with 2–4 equatorial germ pores, pedicel 1.5–3.0 μm. Telia formed from uredinia, erumpent, linear to dome-shaped, reddish brown. Teliospores oval, apex rounded, 1–3-digitate, hyaline to subhyaline, 31–60 × 13–27 μm; wall 1–2 μm thick, mostly thickened at apex (5–14 μm); pedicel persistent, up to 66 μm.

Additional specimens examined. AUSTRALIA, Queensland, Caloundra, on A. leiocalyx, 25 Aug. 1932, S.T. Blake, BRIP 7543; Queensland, Mt Coolum (-26.5656, 153.0969), on A. leiocalyx, 28 July 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57512; Queensland, Mt Coolum (-26.5622, 153.0942), on A. leiocalyx, 28 July 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57514; Queensland, Cunningham Highway (-28.0317, 152.4697), on A. leiocalyx, 1 Aug. 2012, C. Doungsa-ard, A.R. McTaggart, A.D.W. Geering & R.G. Shivas, BRIP 57524; Queensland, Lake Moogerah Road (-28.0247, 152.5189), on A. leiocalyx, 1 Aug. 2012, C. Doungsa-ard, A.R. McTaggart, A.D.W. Geering & R.G. Shivas, BRIP 57527; Queensland, West Haldon (-27.7722, 152.0903), on A. leiocalyx, 1 Aug. 2012, C. Doungsa-ard, A.R. McTaggart, A.D.W. Geering & R.G. Shivas, BRIP 57534; Queensland, Brisbane, Geebung (-27.3375, 153.0511), on A. leiocalyx, 2 Sept. 2012, R.G. & M.D.E. Shivas, BRIP 57631; New South Wales, Shark Creek (-29.5664, 153.2003), on A. leiocalyx, 13 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57568.

Notes — Endoraceium parvum is distinguished from other species by the comparatively short (25–38 μm) urediniospores (Berndt 2011). Berndt (2011) described E. parvum from A. leiocalyx (as A. concurrens) and A. mangium. Specimens of rust on A. mangium were not included in this study and these warrant further study.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808986; Fig. 8

Fig. 8

Endoraecium peggii a. Uredinia (BRIP 55631); b. teliospores (BRIP 55602); c, d. urediniospores (BRIP 55631). — Scale bars: a = 1 cm; b–d = 10 μm.

Etymology. Named after the Australian forest pathologist and rust specialist, Dr Geoff S. Pegg, who has discovered several new plant pathogens in Australia.

Type. AUSTRALIA, Northern Territory, Darwin, 8 km from airport, on A. holosericea, 7 May 2012, R.G. Shivas (holotype BRIP 55602).

Uredinia on both phyllode surfaces, pulverulent, subepidermal, erumpent, linear to round, up to 2 mm, yellowish brown; paraphyses intrasoral, cylindrical, yellow, with pedicel, 37–84 × 9–10 μm, thickened at apex, digitate. Urediniospores ovoid, apex obtuse, yellowish brown to reddish brown, 31–45 × 20–25 μm; wall 3–5 μm thick at sides, apex 4–10 μm thick, foveolate to reticulate, with 2–3 equatorial germ pores; pedicel 3–6 μm. Telia rare, on both surfaces of phyllode, darker than uredinia, erumpent, linear, brown to black. Teliospores oval, apex rounded, no digitations, hyaline to subhyaline, 45–60 × (15–)19–27 μm; wall 1.5–2.0 μm thick at sides, sometimes thickened at apex, 7–17 μm; pedicel persistent, up to 52 μm.

On phyllodes of A. holosericea (Juliflorae).

Additional specimens examined. AUSTRALIA, Queensland, Dimbulah, Bourke Developmental Road, on A. holosericea, 6 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 55631; Queensland, Brisbane, Mt Coot-tha Botanic Gardens (-27.477655, 152.972270), on A. holosericea, Nov. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 58324.

Notes — Endoraecium peggii is known from A. holosericea in Juliflorae. This species of Acacia is native to northern Australia (Maslin 2013). The rust was also collected in south-east Queensland on a plant in the Mt Coot-tha Botanic Gardens. The teliospores of E. peggii lack the apical digitations found in all other species of Endoraecium, although this ornamentation is present on the paraphyses.

(Berk. & Broome) Berndt, Mycol. Progr. 10: 503. 2011

Basionym. Melampsora phyllodiorum Berk. & Broome, Trans. Linn. Soc. London 2, ser. 2: 67. 1883.

≡ Uromyces phyllodiorum (Berk. & Broome) McAlpine, The Rusts of Australia: 95. 1906. (McAlpine (1906) described teliospores from the type specimen, which were overlooked by Berkeley & Broome (1883). Under previous rules of nomenclature, U. phyllodiorum was considered the name of a new species with a teleomorphic type (Berndt 2011). Under the current system of nomenclature, the name is cited as originally published, and is typified by the type of the basionym.)

≡ Pileolaria phyllodiorum (Berk. & Broome) Dietel, Ann. Mycol. 19: 302. 1921.

= Uromyces phyllodii Cooke & Massee (as ‘phyllodiae’), in Cooke, Grevillea 17: 70. 1889.

≡ Coeomurus phyllodii (Cooke & Massee) Kuntze (as ‘Caeomurus’), Revis. Gen. Pl. 3: 450. 1898.

Type. AUSTRALIA, Queensland, Brisbane, on Acacia sp., F.M. Bailey no. 269, K(M) 146703.

Uredinia on both phyllode surfaces, subepidermal, erumpent, pulverulent or on bullate swellings, linear, round to irregular, yellowish brown. Urediniospores subglobose to oval, apex obtuse, yellowish brown, (28–)30–52 × 17–26 μm; wall 2.5–4.0 μm thick at sides, 5–10 μm thick at apex, reticulate, with 3–7 equatorial germ pores; pedicel absent or 1.5–3.5 μm. Telia form-ed from uredinia, erumpent, linear, reddish brown. Teliospores cylindrical to oval, apex rounded, 1–4-digitate, hyaline to subhyaline, 38–58 × 17–28 μm; wall 1–2 μm thick at sides, 6–17 μm thick at apex; pedicel persistent, up to 30 μm.

Additional specimens examined. AUSTRALIA, Queensland, Mount Coolum (-26.5611, 153.0839), on A. aulacocarpa, 28 July 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57310; Queensland, Mt Coolum (-26.5611, 153.0839), on A. aulacocarpa, 28 July 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57516; New South Wales, Woombah, at the beginning of Iluka Road (-29.3544, 153.2492), on A. aulacocarpa, 11 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57578; New South Wales, Woombah, at the beginning of Iluka Road (-29.3544, 153.2492), on A. aulacocarpa, 11 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57580; New South Wales, Woombah, Iluka Bluff (-29.3967, 153.3722), on A. aulacocarpa, 11 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57579; New South Wales, Shark Creek (-29.5664, 153.2003), on A. aulacocarpa, 13 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57569; New South Wales, Shark Creek (-29.5664, 153.2003), on A. aulacocarpa, 13 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57570; Queensland, Mt Tibrogargan (-26.9264, 152.9417), on A. aulacocarpa, 15 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57588; Queens-land, Mt Tibrogargan (-26.9258, 152.9508), on A. aulacocarpa, 15 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57589; Queensland, Mt Tibrogargan (-26.9278, 152.9378), on A. aulacocarpa, 15 Aug. 2012, C. Doungsa-ard & A.R. McTaggart, BRIP 57590.

Notes — Berndt (2011) discussed the taxonomy of E. phyllodiorum, and listed the hosts as A. aulacocarpa, A. crassicarpa, A. holosericea, A. mangium and tentatively A. auriculiformis. The type specimens of Melampsora phyllodiorum and Uromyces phyllodii were collected on unidentified species of Acacia in Brisbane, Australia. Acacia aulacocarpa is the only host species listed by Berndt (2011) that occurs naturally in or near to Brisbane. The other four species of Acacia are restricted to northern Queensland and the Northern Territory (Maslin 2013). Herbarium records and field observations by the authors show that Endoraecium spp. occur on four species, A. aulacocarpa, A. irrorata, A. leiocalyx and A. podalyriifolia, in the Brisbane region. The rust on A. aulacocarpa forms bullate swellings on phyllodes, very similar to those seen in the type specimen of E. phyllodiorum and illustrated by Berndt (2011). Consequently, we suggest that A. aulacocarpa is the host of the type of E. phyllodiorum. Berndt (2011) proposed that E. phyllodiorum had a wide host range, including A. aulacocarpa, A. auriculiformis, A. holosericea and A. mangium. The rusts on A. auriculiformis and A. holosericea are now identified as E. auriculiforme and E. peggii, respectively. The rust on A. mangium is unknown. Acacia aulacocarpa is the only species of Acacia known to host two species of Endoraecium, E. phyllodiorum and E. violae-faustiae.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808987; Fig. 9

Fig. 9

Endoraecium podalyriifolium (BRIP 57576). a, b. Pulverulent sori on leaves and stems; c. teliospores; d, e. urediniospores. — Scale bars: b = 1 cm; c–e = 10 μm.

Etymology. Name refers to the host, Acacia podalyriifolia, on which it was found.

Type. AUSTRALIA, New South Wales, Maclean, Wharf Street (-29.4589, 153.2111), on A. podalyriifolia, 12 Aug. 2012, C. Doungsa-ard & A.R. McTaggart (holotype BRIP 57576).

Uredinia on both phyllode surfaces, subepidermal, erumpent, pulverulent, round, 0.5–10.0 mm, yellow to yellowish brown. Urediniospores cylindrical to ovoid, apex obtuse, subhyaline to yellow, (30–)33–44 × 23–27 μm; wall 3–4 μm thick, reticulate, with 4–6 germ pores. Telia formed from uredinia, subepidermal, erumpent, yellow. Teliospores cylindrical to fusiform, apex rounded, 2–5-digitate, subhyaline to yellow, 44–60 × 12–19 μm; wall 1 μm thick at sides, 6–12 μm thick at apex; pedicel persistent, over 60 μm.

On phyllodes of A. podalyriifolia (Botrycephalae).

Additional specimen examined. AUSTRALIA, Queensland, South Ripley, Ripley Road, on A. podalyriifolia, 17 July 2012, A.D.W. Geering, BRIP 57294.

Notes — Berndt (2011) considered the rust on A. podalyriifolia was E. digitatum based on morphology and the close relationships of host species within this group. However, the results from the molecular phylogenetic analysis in this study indicate the rusts on Botrycephalae are each restricted to a single host species.

(J. Walker & R.G. Shivas) M. Scholler & Aime, Mycoscience 47: 163. 2006

Basionym. Racospermyces tierneyi J. Walker & R.G. Shivas, in Walker, Australas. Mycol. 20: 23. 2001.

Type. AUSTRALIA, Queensland, Tambo, near Castlevale, on A. harpophylla, 6 June 2000, G.S. Pegg (holotype BRIP 27071!).

Uredinia on both phyllode surfaces, subepidermal, erumpent, linear, up to 500 μm, yellowish brown. Urediniospores cylindrical, oval to broadly fusiform, apex obtuse, reddish brown, 26–36 × 13–18 μm; wall 2.0–2.5 μm thick at sides, 4.0–4.5 μm thick at apex, reticulate, with 4–8 equatorial germ pores. Telia on both phyllode surfaces, erumpent. Teliospores subglobose, ovoid to obovoid, apex rounded, reddish brown, darker at the apex, 27–41 × 20–26 μm; wall 1.0 μm thick at sides, 7–13 μm thick at apex, smooth; pedicel persistent, 50–55 μm long.

Additional specimens examined. AUSTRALIA, Queensland, Springsure, on A. harpophylla, 27 Feb. 2001, G.S. Pegg, BRIP 27887; Queensland, Caldervale, on A. harpophylla, 1 Mar. 2001, G.S. Pegg, BRIP 27880.

Notes — Endoraecium tierneyi occurs on A. harpophylla in Plurinerves. It was recovered in the phylogenetic analyses as sister to three species of Endoraecium in Hawaii that occur on A. koa, also in Plurinerves. The teliospores are subglobose to obovoid, while other species of Endoraecium in Australia have clavate teliospores.

McTaggart & R.G. Shivas, sp. nov. — MycoBank MB808988; Fig. 10

Fig. 10

Endoraecium tropicum. a. Pulverulent sori (BRIP 56557); b, c. urediniospores (BRIP 56557); d. teliospores (BRIP 56555). — Scale bars: a = 1 cm; b–d = 10 μm.

Etymology. Name refers to the host, Acacia tropica, on which it was found.

Type. AUSTRALIA, Northern Territory, Gregory, Victoria Highway (-15.6003, 131.2136), on A. tropica, 20 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas (holotype BRIP 56557).

Uredinia on both phyllode surfaces, subepidermal, pulverulent, up to 2 mm, brown. Urediniospores ovoid to fusiform, apex obtuse, yellowish brown, 30–40 × 17–22 μm; wall 3–4 μm thick at sides, apex usually 3–4 μm thick, sometimes slightly thickened (4–5 μm), reticulate, with 3–4, equatorial germ pores; pedicel 3–4 μm. Telia formed from uredinia, erumpent. Teliospores clavate, apex rounded, papillate rather than digitate, 40–53 × 16–23 μm; wall 1.5–2.0 μm thick at sides, 8–11 μm thick at apex; pedicel persistent, 5 μm.

On phyllodes of A. tropica (Juliflorae).

Additional specimen examined. AUSTRALIA, Northern Territory, Gregory, Victoria Highway (-15.6014, 131.2136), on A. tropica, 20 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 56555.

Notes — Endoraecium tropicum was closely related to E. par-vum in the phylogenetic analysis. Both species have urediniospores with uniformly thickened spore walls. The distribution of A. tropica is restricted to the Northern Territory and north-west Queensland. Acacia leiocalyx, the host of E. parvum, occurs in eastern Australia (Maslin 2013).

Berndt, (as ‘violae-faustae’) Mycol. Progr. 10: 513. 2011

Type. AUSTRALIA, Queensland, Cairns, Barron Gorge, MacDonald’s trail, on A. crassicarpa (as Acacia sp.), 18 Aug. 2006, V. Faust-Berndt & R. Berndt (holotype BRIP 53388!).

Uredinia on both phyllode surfaces, subepidermal, erumpent, pulverulent, linear to round, yellowish brown, not forming bullate sori; paraphyses cylindrical with digitate heads, subhyaline to yellow, 35–88 × 6–15 μm thick at sides, thickened at apex. Urediniospores globose, subglobose, obovoid to oval, apex acute, yellowish brown, 28–53 × 17–26 μm; wall 2.5–4.5 μm thick at sides, apex mostly thickened, 5.0–10.5 μm thick, foveolate, with 3–5 equatorial germ pores; pedicel 3–5 μm. Telia formed from uredinia, erumpent, reddish brown. Teliospores obovoid to oval, apex rounded, 1–4-digitate, hyaline to subhyaline, 41–53 × 20–28 μm; wall 1.5–3.0 μm thick at sides, 5–14 μm thick at apex; pedicel persistent, up to 43 μm.

Additional specimens examined. AUSTRALIA, Queensland, Coen, Peach No. 7, on A. crassicarpa 18 July 1999, R.G. Shivas & M. Gunther, paratype BRIP 25816; Queensland, Julatten, Abbatoir Swamp (-16.6074, 145.3428), on A. aulacocarpa, 5 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 55601; Queensland, Mareeba (-16.9414, 145.5411), on A. difficilis, 6 Apr. 2012, C. Doungsa-ard, A.R. Mc-Taggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 55611; Queensland, Innisfail (-17.5003, 146.0756), on A. difficilis, 7 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 55616; Queensland, Dimbulah (-17.2967, 144.9736), on A. aula-cocarpa, 6 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt, V. Faust-Berndt, M.D.E. & R.G. Shivas, BRIP 55629; Northern Territory, Humpty Doo (-12.5964, 131.2083), on A. difficilis, 22 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt & R.G. Shivas, BRIP 56539; Northern Territory, Humpty Doo (-12.5964, 131.2083), on A. difficilis, 22 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt & R.G. Shivas, BRIP 56540; Northern Territory, Humpty Doo (-12.5964, 131.2083), on A. difficilis, 22 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt & R.G. Shivas, BRIP 56545; Northern Territory, Humpty Doo (-12.5964, 131.2083), on A. difficilis, 22 Apr. 2012, C. Doungsa-ard, A.R. McTaggart, R. Berndt & R.G. Shivas, BRIP 56547.

Notes — Berndt (2011) described the urediniospores of E. violae-faustiae as foveolate, with a thickened and acute apex. The most distinctive character was the variable paraphyses, which resembled thin teliospores, similar to those found in E. auriculiforme and E. peggii. Endoraecium violae-faustiae was reported on A. aulacocarpa and A. crassicarpa, also in Juli-florae, as well as A. flavescens in Plurinerves (Berndt 2011). The host of the holotype (BRIP 53388), was re-identified as A. crassicarpa, which is closely related to A. aulacocarpa (Maslin 2013). The host plant of the specimen on A. flavescens (BRIP 25816) was re-identified as A. crassicarpa. The distribution of E. violae-faustiae is restricted to the northern parts of Australia on A. aulacocarpa, A. crassicarpa and A. difficilis.

Berndt, Mycol. Progr. 10: 509. 2011

Type. AUSTRALIA, New South Wales, Cavan Gap near Yass, on A. penninervis, 20 Nov. 1951, E. Gauba (holotype SF35354).

Additional specimen examined. AUSTRALIA, New South Wales, Bald Rock, on A. penninervis, 26 Feb. 1984, J.W. Tierney, paratype BRIP 14205.

Notes — Endoraecium walkerianum was reported from A. penninervis and A. obliquinervia (Berndt 2011) in Botrycephalae (Maslin 2013).

A KEY TO THE KNOWN SPECIES OF ENDORAECIUM IN AUSTRALIA

1. Urediniospores hyaline to pale yellow (on Botrycephalae) .................................................. 2

1. Urediniospores golden or reddish brown ....................... 7

2. Urediniospore wall up to 5 μm or thicker at sides ................ 3

2. Urediniospore wall 2.0–4.5 μm thick at sides ......................... 4

3. Forming galls on stems of A. notabilis ......................... E. digitatum

3. Forming hypertrophied lesions on A. falciformis ................................................. E. falciforme

3. Forming hypertrophied lesions on A. daphnifolia ..........................................E. maslinii

4. Forming hypertrophied lesions on phyllodes ......................... 5

4. Forming galls on stem ......................... 6

5. On A. fasciculifera ......................... E. bicinctum

5. On A. podalyriifolia ......................... E. podalyriifolium

5. On A. penninervis ......................... E. walkerianum

6. On A. irrorata ......................... E. irroratum

6. On A. dealbata ......................... E. carnegiei

7. Teliospores globose to ellipsoid on A. harpophylla .................................................. E. tierneyi

7. Teliospores clavate to obovoid or absent (on Juliflorae) 8

8. Urediniospore wall uniform ......................... 9

8. Urediniospore with thickened apex ......................... 10

9. Teliospores digitate, on A. leiocalyx ......................... E. parvum

9. Teliospores not digitate, on A. tropica .........................E. tropicum

10. Urediniospores foveolate, paraphyses present ......................... 11

10. Urediniospores reticulate, paraphyses absent ......................... 13

11. Urediniospores with an acute apex ......................... 12

11. Urediniospores not acute, on A. holosericea ......................... E. peggii

12. On A. auriculiformis ......................... E. auriculiforme

12. On A. aulacocarpa, A. crassicarpa or A. difficilis ............................................... E. violae-faustiae

13. On A. disparrima ......................... E. disparrimum

13. On A. aulacocarpa ......................... E. phyllodiorum

DISCUSSION

Phylogenetic analyses of loci from nuclear rDNA and mitochondrial DNA showed that species of Endoraecium in Australia have narrow host ranges. Savile (1971) hypothesised that rusts speciated either by divergence with their hosts or by host jumps. The close relationship between Endoraecium and species/subclades of Acacia observed in this study indicates that Endoraecium diversified by co-evolution with its hosts. For example, the five species of Endoraecium on Botrycephalae, namely, E. carnegiei, E. falciforme, E. irroratum, E. maslinii and E. podalyriifolium, were closely related with few differences in the studied genes and short branch lengths between species in the recovered phylogenetic trees. Species of Acacia in Botrycephalae diversified c. 3.4–3.8 million years ago (Miller et al. 2013), which would represent the maximum age of the corresponding Endoraecium species.

Endoraecium tierneyi on A. harpophylla in Australia was recovered as sister to the Hawaiian species, E. acaciae, E. hawaiiense and E. koae, on A. koa. These Australian and Hawaiian rusts all occur on species of Acacia in Plurinerves. Hodges & Gardner (1984) and Walker (2001) proposed that the Australian species of Endoraecium were the ancestor of the Hawaiian rusts. Our study supports this view by showing the rusts on Plurinerves had an Australian ancestor (plesiomorphic state). Further, the three rusts on Acacia in Hawaii have differences in life cycle and morphology, but are identical in the LSU region of rDNA. These rusts have diversified on A. koae after its relatively recent split from Australian species of Acacia (Brown et al. 2012). Endocyclic species occur in Hawaii but not in Australia, which shows that a reduced life cycle is a derived character in Endoraecium.

Endoraecium hyalosporum from South-East Asia infects A. confusa, which is the only example of a species of Endoraecium on a host outside of the p.u.b. clade of Acacia (Brown et al. 2012). This may represent an example of a host jump. However the systematic position of E. hyalosporum is unknown, and it is possible that it does not have a close phylogenetic relationship to other species of Endoraecium. It is noteworthy that E. hyalosporum has been assigned to other genera at various times, namely, Maravalia (Dietel 1924), Poliotelium (Mains 1939) and Atelocauda (Ono 1984).

The morphology of urediniospores may be synapomorphic for the three clades of Endoraecium. For example, the rusts on Juliflorae have reticulate to foveolate ornamentation on golden-brown urediniospores. The rusts on Botrycephalae have a raised reticulate ornamentation on subhyaline to pale yellow urediniospores. Endoraecium tierneyi was the only rust examined on Plurinerves, and it had reticulate, reddish brown urediniospores. Berndt (2011) noted the urediniospores of species of Endoraecium on A. koa have similar morphology to the rusts on Juliflorae, such as E. phyllodiorum. We cannot propose any apomorphies for the rusts on Plurinerves at this stage. The rusts on Juliflorae, namely E. auriculiforme, E. disparrimum, E. peggii, E. phyllodiorum and E. violae-faustiae, have urediniospores with thickened apices, whereas their sister rusts E. parvum and E. tropicum, also in Juliflorae, have uniformly thickened spore walls.

Other morphological characters were less useful for species delimitation in Endoraecium. In general, the teliospores of rusts on species of Acacia in Botrycephalae had numerous (up to 15) apical digitations and were clavate. The teliospores of rusts on species of Acacia in Juliflorae had fewer (up to four) apical digitations and were obovoid. The presence of a gall or bullate swelling is not a useful character for species delimitation. For example, the sori of E. auriculiforme and E. phyllodiorum may occur on bullate swellings or on undistorted phyllodes (Berndt 2011).

Endoraecium is likely to show further diversity. Only 14 species of Acacia were sampled in this study. Other species such as A. glaucocarpa (Botrycephalae), A. juliflorae (Juliflorae) and A. viscidula (Plurinerves) were observed by the authors as hosts for potentially new species of Endoraecium. New taxa will need to be classified by a combined morphological, ecological (host range) and molecular approach. We predict that many new species of Endoraecium, which have diversified by coevolution with their Acacia hosts, remain to be found in Australia.