Morphological and molecular characterization of Langdonia walkerae sp. nov. infecting Aristida stricta and A. beyrichiana in longleaf pine-grassland ecosystems in the southeastern USA.

A smut fungus that hinders wiregrass restoration efforts in longleaf pine-grassland ecosystems was collected from Aristida stricta and A. beyrichiana (Poaceae) in three states in the southeastern USA. Morphological and phylogenetic characteristics of this fungus were examined. These data show that the specimens from both plant species were infected by the same fungus and represent a new species of Langdonia. The new species differs morphologically from other species of Langdonia by teliospores being solitary and not compacted into spore balls. Spore wall ornamentation and teliospore size also differ from other Langdonia species. Phylogenetic analyses of DNA sequences of the ITS, LSU, and EF-1α supported separation of the species from A. stricta and A. beyrichiana from other Langdonia species. Based on these results, a new species, Langdonia walkerae, is proposed.

INTRODUCTION

Aristida, a member of the Poaceae and the Aristidoideae (Barkworth 2003, Cerros-Tlatilpa ), is distributed worldwide with approximately 300 species (Barkworth 2003) primarily located in Central, North and South America, Australia, and Africa (Cerros-Tlatilpa ). This genus is distinct from other grass genera by its three-awned lemma (Clewell 1989, Barkworth 2003). In the southeastern United States, two Aristida species, A. stricta and A. beyrichiana, are common grasses found in the ground layer of longleaf pine, Pinus palustris, forests (Peet 1993), which extend from the border of Texas and Louisiana eastward to the Atlantic Ocean and from the middle of Florida northward into Virginia (Brockway , Frost 2006). Aristida stricta, called Carolina wiregrass or the pineland threeawn, occurs in North Carolina and the northern part of South Carolina, and is distinguished morphologically by having hairs along the length of the blades, wider culms, and shorter ligules (Peet 1993, Van Eerden 1997). Aristida beyrichiana, called southern wiregrass or Beyrich threeawn, grows in the southern part of South Carolina into Florida and west to Mississippi, and is distinguished morphologically by its soft, short hairs at the base of the blades and the glumes are more unequal in length (Peet 1993, Van Eerden 1997).

These perennial bunchgrasses are adapted to fire, enabling them to persist and reproduce despite frequently recurring fires (Platt 1999, Means 2007) as well as to survive for decades without burning (Shearman ). Flowers and seed production are most abundant within a year of burning; without fire, few, if any, are produced each subsequent year (Clewell 1989, Van Eerden 1997, Fill ). These native wiregrasses are considered keystone species in the longleaf pine-grass ecosystem as they produce the fine fuels needed for frequent burning, a requirement for sustained ecosystem structure and function (Clewell 1989, Duever 1989, Noss 1989). Thus, the seed viability of wiregrasses is important for ecosystem restoration efforts (Van Eerden 1997) in projects that focus on wiregrass as a primary understory species.

The production of A. stricta and A. beyrichiana seeds has been observed to be affected by a smut fungus that replaces the developing ovaries with teliospores (Van Eerden 1997). Smut fungi are pathogens that primarily infect grasses and occur world-wide. They produce sori, sporocarps in which teliospores are produced, in different organs of their host (Vánky 2013). Previously, many grass-infecting smut fungi, including Sporisorium, were united into the genus Ustilago that has proved to be polyphyletic after the introduction of molecular phylogenetic studies (Begerow , 2006). In the taxonomy of smut fungi, the host plant is one of the criteria usually used to classify groups (Bauer , Begerow ). Consequently, McTaggart emended the description of the genus Sporisorium and separated all species that occur on Aristida into a new genus, Langdonia, based on the host plant and morphological and phylogenetic differences. Langdonia is a monophyletic group within the Ustilaginaceae (McTaggart ). Currently, eight species have been described on other Aristida species from various countries - including Argentina, Australia, Bolivia, Madagascar, Mexico, Thailand, and the USA (Durán 1987, McTaggart , Vánky 2012, Denchev ). The defining morphological characteristics of the genus Langdonia are sori that can be found in some or all ovaries of the panicle, lack of columella and sterile cells, teliospores compacted into spore balls, and Ustilago-type germination (McTaggart , Vánky 2013). Based on these criteria, the tentative identification of the fungus causing smut on A. stricta and A. beyrichiana was a species of Langdonia. Therefore, the objectives of this study were to identify the fungus that infects A. stricta and A. beyrichiana and to determine if the same fungus species infects both A. stricta and A. beyrichiana.

MATERIALS AND METHODS

Sample collection and documentation

Naturally infected plants were collected from seven locations in North Carolina, South Carolina, and Florida during the seed production stage (Table 1). In 2017, samples were collected from two locations - one in South Carolina and one in Florida, and in 2018 samples were collected from six locations in all three states, with only one location being common to both years. All sampled sites had been burned by prescription within the year. The locations and the wiregrass species from which each sample was obtained are included in the “specimens examined” section and in Table 1. For collection, samples of infected wiregrass culms, erect stems bearing inflorescences, were cut at ground level, placed in paper bags, and returned to the laboratory where the specimens were kept at room temperature for subsequent use. Plant voucher specimens were deposited in the Clemson University Herbarium (CLEMS), and fungus specimens were deposited in the Washington State University Mycological Herbarium (WSP) and the U.S. National Fungus Collections - Herbarium (BPI) and kept as voucher specimens.

Table 1

Smut samples on Aristida beyrichiana and A. stricta collected and examined for this study.

Site	Voucher	Isolate	State	City/County	Aristida spp.	Date
Silver Bluff Audubon Center	CLEMS0080381	JK 5017	SC	Jackson/Aiken	A. beyrichiana	10 Nov. 2017
	Holotype: WSP74240
	Isotype: BPI911222
	CLEMS0080382	JK 6017
	CLEMS0080383CLEMS0080384	SB 1218				6 Dec. 2018
	Aiken Gopher Tortoise Heritage Preserve	CLEMS0080385	AGT 1218	SC	Aiken/Aiken	A. beyrichiana	6 Dec. 2018
Savannah River Site	CLEMS0080388WSP74241BPI 911220	SRS 1218	SC	Barnwell/Barnwell	A. beyrichiana	4 Dec. 2018
Apalachicola Bluffs and Ravines Preserve	CLEMS0080387	FLW 1017	FL	Bristol/Liberty	A. beyrichiana	31 Oct. 2017
Austin Cary Forest	CLEMS0080386	FL 1218	FL	Gainesville/Alachua	A. beyrichiana	3 Dec. 2018
Carolina Sandhills National Wildlife Refuge	CLEMS0080389WSP74242BPI 911221	CSN 1118	SC	Hoffman/Richmond	A. stricta	27 Nov. 2018
North Carolina Sandhills Game Land	CLEMS0080390WSP74243BPI 911219	NC 1018NC 2018NC 3018NC 5018	NC	Chesterfield/Chesterfield	A. stricta	29 Nov. 2018

Morphological examination

The characteristics of the sori and teliospores were examined on infected plants. Pictures of sori were taken using a Nikon D5100 camera. Teliospore characteristics were studied using a compound microscope (LM; BX60F; Olympus optical Co. Ltd., Japan) equipped with ProgRes C5 camera (JENOPTIK, Germany) and CapturePro software; and a scanning electron microscope (SEM; Hitachi SU6600) at 5.0 kV at the Clemson University Electron Microscopy Facility. For LM, teliospores were mounted in lactic acid (85–90 %, VWR, International, LLC) (Savchenko ) and examined at 1 000 × magnification. The diameters of 30 teliospores, oriented in plane view so that they appeared globose, were measured from each sample collection. The colors of the sori and the teliospores were described according to Rayner (1970). For SEM examination, teliospores were dusted on double-sided adhesive carbon tape, mounted on aluminum stubs, and sputter-coated with platinum using a Cressington sputter-coater (ca. 30 nm in 6 min).

Growth on artificial medium

Sori were carefully removed from infected plants and surface-sterilized by immersing in 70 % ethanol for 30 s, rinsing with sterile distilled water, immersing in 3 % sodium hypochlorite for 10 s, and rinsing twice with sterile distilled water. Each sorus was handled individually using aseptic techniques; one sorus was placed in a 2.0 mL centrifuge tube and gently ground with a tissue grinder pestle to release the teliospores. To make a spore suspension, 500 μL sterile distilled water was added to a tube, the contents were thoroughly mixed, 100 μL was pipetted onto a petri dish containing Difco™ malt extract agar (MEA; Becton, Dickinson and Co., Sparks, MD) and the suspension was spread with a L-shaped cell spreader over the agar surface. The MEA was made according to the manufacturer’s instruction and amended with streptomycin (MEA+S; 200 ppm = 200 mg/L). Plates were placed at 28 °C for 2–3 d; and then individual colonies were aseptically subcultured onto MEA+S and incubated for 10 d at 28 °C. The morphological characteristics of shape, color, and diameter of the colony were recorded. The color of the colony was defined according to Rayner (1970).

DNA Extraction, Polymerase chain reaction (PCR), and sequencing

DNA was extracted from 10-d-old cultures using a DNeasy Plant Mini Kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer’s instructions. DNA concentration was determined using a Thermo Scientific Nanodrop 2000 /2000c spectrophotometer. All DNA samples were stored at -20 °C until used for amplification. Genomic DNA was amplified using an Eppendorf Mastercycler Gradient thermocycler with Thermo Scientific Phusion High-Fidelity DNA Polymerase following the manufacturer’s cycling and reaction conditions. Two nuclear ribosomal DNA regions and one gene locus were amplified and sequenced: the internal transcribed spacer (ITS) region, the large subunit region (LSU), and translation elongation factor-1α locus (EF-1α). The ITS region was amplified with primers M-ITS1 (Stoll ) and ITS4 (White ) and annealed at 58 °C (McTaggart ); the LSU region was amplified with primers LROR and LR5 (Vilgalys & Hester 1990) at 60 °C; and the EF-1α locus was amplified with primers EF-1αF and EF-1αR (McTaggart ) at 64 °C. The PCR products were analyzed using gel electrophoresis on a 1.0 % agarose gel and sent to Arizona State University Core Laboratories for purification and sequencing using the same forward and reverse PCR primers. AB1 sequence trace files were assembled using Geneious Prime® v. 2020.0.5 software, and a BLAST search was conducted for all resulting sequences to confirm accurate species identification. The sequences were deposited in GenBank and their accession numbers are shown in Table 2.

Table 2

Species of fungi used in phylogenetic analyses for this study, including host plant and country of origin.

Species	Voucher/Isolate	Host	Country	GenBank accession no.
				ITS	LSU	EF-1α
Langdonia aristidae	H.U.V.19145	Aristida urugayensis	Germany	AY740048	AY740101	n/a
	HUV19145	Aristida urugayensis	Argentina	JN367292	JN367317	JN367369
Langdonia aristidicola	BRIP 26930	Aristida jerichoensis	Australia	HQ013091	n/a	HQ013032
Langdonia confusa	BRIP 42670	Aristida queenslandica	Australia	HQ013095	HQ013132	n/a
	BRIP 52755	Aristida sp.	Australia	HQ013096	n/a	n/a
Langdonia fraseriana	BRIP 49668	Aristida nitidula	Australia	HQ013100	n/a	n/a
Langdonia inopinata	M-0215944	Aristida adscensionis	Zambia	KY929612	MF668632	n/a
Langdonia walkerae	JK 5017	Aristida beyrichiana	USA, SC	MT429291	MT429279	MT577017
	JK 6017	Aristida beyrichiana	USA, SC	MT429292	MT429280	MT577018
	SB 1218	Aristida beyrichiana	USA, SC	MT429293	MT429281	MT577019
	AGT 1218	Aristida beyrichiana	USA, SC	MT429294	MT429282	MT577020
	SRS 1218	Aristida beyrichiana	USA, SC	MT429295	MT429283	MT577021
	FLW 1017	Aristida beyrichiana	USA, FL	MT429296	MT429284	MT577022
	FL 1218	Aristida beyrichiana	USA, FL	MT429297	MT429285	MT577023
	CSN 1118	Aristida stricta	USA, SC	MT429298	MT429286	MT577024
	NC 1018	Aristida stricta	USA, NC	MT429299	MT429287	MT577025
	NC 2018	Aristida stricta	USA, NC	MT429300	MT429288	MT577026
	NC 3018	Aristida stricta	USA, NC	MT429301	MT429289	MT577027
	NC 5018	Aristida stricta	USA, NC	MT429302	MT429290	MT577028
Macalpinomyces eragrostiellae	Ust. Exs. 960	Eragrostiella bifaria	India	AY740036	AY740089	n/a
Melanotaenium euphorbiae	HUV17733	Euphorbia heterophylla	Papua New Guinea	JN367289	JN367314	JN367365
Sporisorium andropogonis	56588 (M)	Bothriochloa saccharoides	Ecuador	AY740042	AY740095	n/a
Sporisorium exsertum	KVU965	Themeda triandra	Australia	JN367293	JN367318	JN367370
Sporisorium manilense	Ust.Exs.854 (M)	Sacciolepis indica	India	AY740059	AY740112	n/a
Sporisorium moniliferum	BRIP 52504	Heteropogon contortus	Australia	HQ013104	n/a	HQ01302
Sporisorium ophiuri	HB 20	Rottboellia cochinchinensis	Unknown	AY740019	AJ236136	n/a
Sporisorium sorghi	MP 2036a	Sorghum bicolor	Nicaragua	AY740021	AF009872	n/a
Sporisorium vermiculum	BRIP 49748	Sorghum plumosum	Australia	HQ013114	HQ013134	HQ01307
Sporisorium wynaadense	BRIP 27640	Sarga leiocladum	Australia	HQ013116	HQ013124	HQ01309
Ustilago bromivora	H.U.V.19322	Bromus catharticus	Argentina	AY740064	AY740118	n/a
Ustilago hordei	Ust.exs.784	Hordeum vulgare	Iran	AY345003	AF453934	n/a
Ustilago nuda	H.U.V.17782	Hordeum leporinum	Greece	AY740069	AJ236139	n/a
Ustilago nunavutica	DAOM 91211	Puccinellia angustata	Canada	KF381025	KF381049	n/a
Ustilago sparsa	Ust.exs.892	Dactyloctenium aegyptium	India	AY345008	DQ864974	n/a
Ustilago striiformis	HAI 4610	Milium effusum	Ukraine	KF381021	KF381046	n/a
Ustilago xerochloae	Ust. Exs. 1000	Xerochloa imberbis	Australia	AY345012	AY740150	n/a

Note: Species and accession numbers in bold are new sequences from this study.

Phylogenetic analysis

The phylogenetic analysis was conducted using Geneious Prime® v. 2020.0.5 software. The relationships among the isolates of Langdonia species sequenced in this study and other taxa in the Ustilaginaceae were inferred from a phylogenetic tree based on the ITS, LSU, and EF-1α data sets. The sequences for all taxa were first concatenated for each taxon and then aligned using the MUSCLE algorithm. The final data set contained sequences from 12 isolates collected in this study and 43 reference sequences obtained from GenBank (Table 2). Alignments were uploaded to NGphylogeny.fr (https://ngphylogeny.fr; Lemoine ) and curated using BMGE (Criscuolo & Gribaldo 2010) to remove poorly aligned positions. The final super matrix contained 1 465 characters, including gaps.

The phylogenetic analysis was conducted using Bayesian inference and maximum likelihood; a GTR model with GAMMA distribution (Nylander 2004) was selected for both analyses. For Bayesian inference, MrBayes v. 3.2.6 was used to conduct a Markov Chain Monte Carlo (MCMC) (Huelsenbeck & Ronquist 2001). Four runs, each consisting of four chains, were conducted for 1 000 000 generations, and the cold chain was heated at a temperature of 0.25. Trees were sampled every 1 000 generations. The standard deviation of split frequencies was 0.01. RAxML v. 8 (Stamatakis 2014) was used for the maximum likelihood using a bootstrapping analysis. RAxML analyses were run with a rapid bootstrap analysis to search for the best-scoring likelihood tree using a random starting tree and 1 000 maximum likelihood bootstrap replicates. Trees were rooted using Melanotaenium euphorbiae (HUV17733). The final alignment and trees were deposited in TreeBASE 23819.

RESULTS

When teliospores were cultured on MEA+S, germ-tube formation occurred after 6 h incubation at 28 °C, and the germination rate increased steadily over a period of 2–3 d. Spore germination was the Ustilago-type, making a phragmobasisium on which ovoid sporidia (i.e., basidiospores) were produced. Colonies grew slowly; most colonies were visible after 3–5 d, and when subcultured, a colony reached 3–4 cm diam after 14 d. However, more than 50 % of the teliospores remained dormant throughout the incubation period. Young colonies first appeared pale luteous but later became ochreous with age. Colony shape was convex and irregular with undulate edges. At a later stage of growth, the colony surface in some isolates became wrinkled and formed cerebriform sectors (Ulloa & Hanlin 2000).

A total of 12 isolates from the culture were sequenced for two nuclear ribosomal DNA regions, ITS and LSU and one gene locus, EF-1α, and included in the final data set with 43 reference sequences obtained from GenBank. The list of taxa, their host plants and geographic sources, and the GenBank accession numbers for these sequences are shown in Table 2. The phylogenetic relationships between the isolates from A. stricta and A. beyrichiana exhibited full support to recognize them as one species based on the MrBayes and RAxML analyses, with values of 1.0 and 100, respectively. The topology of the MrBayes and RAxML trees (Figs 1, 2) indicate that the new species described below forms a highly supported sister group to other species of Langdonia as found from ITS+LSU+EF-1α analyses. Estimates for posteriori probabilities are indicated on branches of the tree (Fig. 1). In both analyses, isolates collected in this study clustered with values of 1.0 and 100 for MrBayes and RAxML, respectively (Figs 1, 2). The ITS+LSU+EF-1α dataset showed that the monophyly of these isolates within the Ustilaginaceae was fully supported. All taxa of Langdonia on Aristida spp. formed a well-supported monophyletic clade in the Ustilaginaceae and were well separated from the representatives of species in other genera of this family. Within this clade of Langdonia species, all isolates from both Aristida species clustered with strong support. In MrBayes and RAxML analyses of the alignment, Macalpinomyces eragrostiellae was sister to the clade of Langdonia species (Fig. 2).

Fig. 1.

Bayesian inference of phylogenetic relationships resulting from the analysis of ITS, LSU and EF-1α sequences. Numbers on the branches are estimates for Posterior Probability from Bayesian inference. The tree was rooted using Melanotaenium euphorbiae.

Fig. 2.

Maximum likelihood of phylogenetic relationships resulting from the analysis of ITS, LSU and EF-1α sequences. Numbers on the branches are estimates of bootstrap support values. The tree was rooted using Melanotaenium euphorbiae.

Taxonomy

Alqurashi, J. Kerrigan & K. G. Savchenko, MycoBank MB 839451. Fig. 3A–F.

Fig. 3.

Sori and teliospores of Langdonia walkerae sp. nov. on Aristida beyrichiana (A–C) and on Aristida stricta (D–F). A. Sori (WSU 74240). B. Teliospores viewed with transmitted light. C. Scanning electron micrograph (SEM) of teliospores. D. Sori. E. Teliospores viewed with transmitted light. F. SEM of teliospores. Scale bars: A, D = 2 mm; B, E = 10 μm; C, F = 5 μm.

Etymology: Named after Dr. Joan L. Walker (USDA-Forest Service), a plant ecologist and conservationist who observed and contemplated the ecological significance of the sori in A. stricta and A. beyrichiana inflorescences as early as 1979 and brought that to our attention for study.

Typus: USA, South Carolina, Silver Bluff Audubon Center, on Aristida beyrichiana, 10 Nov. 2017, J. Kerrigan & A.S. Alqurashi, CLEMS 0080381, CLEMS 0080382 (holotype WSP 74240; isotype BPI 911222).

Sori in some ovaries, covered at first by a dark herbage green and later by a dark brick peridium. Sori are conspicuous, ovoid or symmetrically fusiform with an acute tip and sized between 1–1.5 × 2.5–7 mm. Teliospores are solitary, not compacted into spore balls, dark sienna in color. In the plane view, teliospores are globose, subglobose, or ellipsoidal and sized 8–13 μm diam. In the side view, teliospores are slightly flattened tangentially. Spore walls are uniformly thick, aculeate, and have dense conical spines (Ulloa & Hanlin 2000). Columella, spore balls, and sterile cells are lacking.

Host: Aristida beyrichiana (Poaceae).

Distribution: USA, southern part of South Carolina to southern Florida.

Additional specimens examined: On Aristida beyrichiana. USA, South Carolina, Silver Bluff Audubon Center, 6 Dec. 2018, A.S. Alqurashi & A.H. Alqurashi, CLEMS 0080383 & CLEMS 0080384; South Carolina, Aiken Gopher Tortoise Heritage Preserve, 6 Dec. 2018, A.S. Alqurashi & A.H. Alqurashi, CLEMS 0080385; South Carolina, Savannah River Site, 4 Dec. 2018, L. Lee, CLEMS 0080388, WSP 74241, BPI 911220; Florida, Apalachicola Bluffs and Ravines Preserve, 31 Oct. 2017, J. Walker, CLEMS 0080387; Florida, Austin Cary Forest, 3 Dec. 2018, J. Hong, CLEMS 0080386. On Aristida stricta. USA, North Carolina, North Carolina Sandhills Game Land, 27 Nov. 2018, A.S. Alqurashi, A.H. Alqurashi & B. Beck, CLEMS 0080389, WSP 74242, BPI 911221; South Carolina, Carolina Sandhills National Wildlife Refuge, 29 Nov. 2018, A.S. Alqurashi & A.H. Alqurashi, CLEMS 0080390, WSP 74243, BPI 911219.

Notes: Morphologically, Langdonia walkerae is distinguished from other species of Langdonia by teliospores that are not compacted into spore balls (McTaggart , b). Phylogenetically, there is a strong support to recognize L. walkerae as a separate species.

DISCUSSION

Smut specimens from two species of wiregrass, A. stricta and A. beyrichiana, collected from longleaf pine ecosystems in three southeastern states — South Carolina, North Carolina, and Florida — were examined and compared in terms of morphology and DNA sequence data. Our findings determined both Aristida species were infected by the same pathogen, which was identified as a new species. Langdonia walkerae is designated as a new species based on its morphological and phylogenetic differences from the other described species of Langdonia. Only eight other species of Langdonia have been erected, including L. aristidae, L. aristidaria, L. aristidicola, L. clandestina, L. confusa, L. fraseriana, L. goniospora, and L. inopinata. All of these species have been reported from two or more Aristida species. For example, L. confusa, originally named Sporisorium confusum, was reported from several species of Aristida including A. dichotoma, A. fendleriana, A. spiciformis, and A. wrightii from several US states (Farr & Rossman 2020). In contrast, one host has been reported to be infected by different Langdonia species. For example, Aristida adscensionis is infected by Sporisorium aristidicola ≡ Langdonia aristidicola, S. consanguineum ≡ L. aristidae, and S. inopinatum ≡ L. inopinata in Zambia, India, and Zimbabwe, respectively (Farr & Rossman 2020). In the present study, sori were found in a proportion of the ovaries of A. stricta and A. beyrichiana. This is similar to four species of Langdonia, namely L. aristidaria (Durán 1987), L. aristidicola, L. clandestina and L. inopinata (Vánky 2012). In the other four species of Langdonia, namely L. confusa, L. aristidae, L. fraseriana and L. goniospora, the sori are produced in all ovaries of an inflorescence on a culm of the host plants (Vánky 2012).

Langdonia walkerae teliospores were solitary in all specimens examined on A. stricta and A. beyrichiana, and no spore balls were observed. This finding conflicts with that from McTaggart , who considered this an apomorphic characteristic among the species of Langdonia. Based on our data, the genus Langdonia seems to be sharing the recent common ancestor with Macalpinomyces eragrostiellae, another species with well-defined spore balls. It is possible that the ancestors of L. walkerae have had spore balls, but they were lost in the process of species evolution.

The absence of spore balls leads us to infer that this fungus is a new species of Langdonia, and the genus description, which states that teliospores are compacted into spore balls (McTaggart ), needs to be emended. The size of teliospores is distinct from L. confusa and L. aristidae, both of which have larger teliospore diameters, and from L. aristidaria, L. aristidicola, L. clandestina, L. fraseriana, L. goniospora, and L. inopinata, all of which have smaller teliospore diameters (Durán 1987, Vánky 2012, Denchev ). Furthermore, the teliospore wall ornamentation of L. walkerae is distinguishable from other described species by having dense conical spines whereas the teliospore wall ornamentation in other species of Langdonia varies from smooth to verruculose (Vánky 2012). For example, the wall ornamentation in Sporisorium confusum ≡ L. confusa on A. dichotoma was densely verrucose-echinulate and in S. consanguineum ≡ L. aristidae on A. arizonica, the teliospores wall were almost smooth to finely verruculose (Vánky 2012). Teliospore germination of L. walkerae is Ustilago-type, and columella and sterile cells are absent, fitting the description of the genus Langdonia. No data are available from previous studies to compare the culture growth of L. walkerae with other species of Langdonia.

Phylogenetic analysis demonstrated that all 12 isolates included in this study, five on A. stricta and seven on A. beyrichiana, form a highly supported monophylum, clustering in one separate phylogenetic clade within the Ustilaginaceae. These isolates were sister to other species of Langdonia and together formed a monophyletic group supporting the results of McTaggart .

Although many species of Langdonia were reported on different species of Aristida in different regions around the world, only limited molecular data are available in GenBank for some of these species. Thus, additional sequences are needed for a more complete understanding of the phylogenetic relationships within this genus. The current study shows that L. walkerae on A. stricta and A. beyrichiana is clearly distinguishable morphologically and phylogenetically from other species of Langdonia on Aristida spp. More material from additional locations in longleaf pine ecosystems should be collected and analyzed to determine if L. walkerae occurs throughout the natural range of longleaf pine.