Introduction

Fraxinus L. (ash) is a tree genus native to the temperate and subtropical regions of the Northern Hemisphere. It belongs to the family Oleaceae and includes 43 species, most of which are large or medium-sized trees, with some shrub species widespread in dry areas (Wallander 2008). Three species, F. angustifolia Vahl (narrow leaved ash), F. excelsior L. (European ash) and F. ornus L. (manna ash) are widely planted as ornamentals.

Since the early 1990s severe branch dieback of F. excelsior was observed in different countries of Central, Eastern and Northern Europe (Przybyl 2002; Lygis et al. 2005; Pukacki and Przybyl 2005; Bakys et al. 2009). All of these studies demonstrated the occurrence of several pathogenic fungi from necrotic shoots of ash. Among these, Diplodia mutila (Fr.) Fr. was one of the most consistently detected species. This pathogen was also reported associated with cankers and branch dieback of F. ornus in Sicily (Italy) (Sidoti and Granata 2004).

In recent years, during surveys carried out in Portugal and Sardinia (Italy) aimed at clarifying the causes of a decline affecting Fraxinus spp. in urban and natural areas, a large collection of Diplodia mutila-like strains were isolated from trees showing V-shaped cankers and a progressive dieback of shoots and branches. Although morphologically similar to D. mutila, some of these strains differed in their colony morphology, larger conidia and DNA sequence data (ITS and EF1-α) from other known strains of D. mutila. This species has been reported from a wide range of hosts of agricultural and forestry importance, where it has been associated with canker, dieback and fruit rot symptoms (Farr and Rossman 2013). However, there are conflicting reports regarding its pathogenicity, and in particular, on Fraxinus spp. (Przybyl 2002; Sidoti and Granata 2004; Bakys et al. 2009), which may be a result of differences in strain virulence but may also be due to the existence of cryptic species. Cryptic speciation is common in the family Botryosphaeriaceae and in the genus Diplodia, which renders species identification difficult when based solely on morphological characters (Phillips et al. 2012, 2013). Fries (1823) described the species Sphaeria fraxini Fr. on Fraxinus sp., and later (Fries 1849) he transferred it to Diplodia as Diplodia fraxini (Fr. : Fr.) Fr. Subsequently, this fungus was placed by Saccardo (1884) in the genus Botryodiplodia (Sacc.) Sacc. as B. fraxini (Fr. : Fr.) Sacc. Unfortunately, no ex-type cultures are available for this species.

Currently, 17 Diplodia species are known from culture (Phillips et al. 2013). These species have been recognised mainly on the basis of DNA sequence data (single or multi-locus) and minor differences in conidial morphology (de Wet et al. 2003; Alves et al. 2004, 2006; Gure et al. 2005; Damm et al. 2007; Lazzizera et al. 2008; Pérez et al. 2010; Jami et al. 2012; Phillips et al. 2012, 2013; Linaldeddu et al. 2013; Lynch et al. 2013). For the majority of species there are ex-type or ex-epitype cultures deposited in publicly available culture collections that can serve as standards for the morphological and molecular characterisation of the species. The only exception is D. mutila for which several cultures are available but none has been linked to the type of the species. The lack of an ex-type culture, or other cultures linked to the holotype of D. mutila has hampered taxonomic studies on the genus Diplodia (which is typified by D. mutila), especially those based on DNA sequence data.

Therefore, the main aims of this work were: 1) to characterise collections of D. mutila-like and other Diplodia spp. isolates in terms of morphological and phylogenetic relationships to known Diplodia species; 2) to select a suitable epitype specimen for D. mutila and a neotype specimen for D. fraxini with related cultures that can be made available for future studies.

Materials and methods

Isolates and morphology

Diplodia isolates used in this study were obtained from branches of F. angustifolia and F. ornus showing sunken cankers and dieback. Diplodia species isolated from other symptomatic trees including Populus alba L. (white poplar), Cupressus sempervirens L. (Italian cypress) and Quercus coccifera L. (kermes oak) were also included in this study. Isolations were made from chips of inner bark and wood tissues approx. 5 mm2 cut aseptically from the margin of necrotic lesions or directly from pycnidia. All samples were cultured on potato dextrose agar (PDA, Oxoid Ltd.) in Petri dishes. After incubation at 25 °C for 1 week, colonies were sub-cultured onto half-strength PDA (1/2 PDA) or on water agar supplemented with autoclaved poplar twigs to enhance sporulation. All colonies were kept on the laboratory bench at about 20–25 °C where they received diffused daylight.

Monoconidial cultures were obtained by spreading conidia on the surface of PDA and incubating overnight at 25 °C. Single germinating spores were transferred to fresh plates of PDA. Cardinal temperatures for growth were determined on PDA plates (90 mm) incubated at 5, 10, 15, 20, 25, 30, 35 and 40 °C (±0.5 °C) in the dark. Five replicate plates for each isolate were made and colony diameters were measured after 4 days.

For microscopy, the contents of conidiomata were dissected out and mounted in 100 % lactic acid. For observations of conidiogenesis, the conidiogenous layer was dissected out and mounted in 100 % lactic acid. Measurements of conidia were made with the Leica IM 500 measurement module from images recorded on a Leica DFC 320 digital camera. From measurements of 50 conidia the mean, standard deviation and 95 % confidence intervals were calculated. Conidial dimensions are given as the range of dimensions with extremes in parentheses. Dimensions of other structures are given as the range of at least 20 measurements.

Representative isolates were deposited at the Centraalbureau voor Schimmelcultures (CBS), Utrecht, the Netherlands and nomenclatural data in MycoBank (Crous et al. 2004). Specimens were lodged with the herbarium of Estação Agronómica Nacional, Oeiras, Portugal (LISE). Isolates used for phylogenetic analyses in this study are provided (Table 1).

Table 1 Isolates used in this study
Full size table

DNA extraction, PCR amplification and sequencing

DNA was isolated from fungal mycelium by the method of Santos and Phillips (2009). Procedures and protocols for DNA sequencing were as described by Alves et al. (2004). PCR reactions were carried out with Taq polymerase, nucleotides and buffers supplied by MBI Fermentas (Vilnius, Lithuania) and PCR reaction mixtures were prepared according to Alves et al. (2004), with the addition of 5 % DMSO to improve the amplification of some difficult DNA templates. All primers were synthesised by MWG Biotech AG (Elbersberg, Germany). The ITS region was amplified using the primers ITS1 and ITS4 (White et al. 1990) as described by Alves et al. (2004). The primers EF1-728F and EF1-986R (Carbone and Kohn 1999) were used to amplify part of the translation elongation factor 1-α gene (EF1-α) as described by Alves et al. (2006). The amplified PCR products were purified with the JETQUICK PCR Purification Spin Kit (GENOMED, Löhne, Germany). The PCR products were sequenced by STAB Vida Lda (Portugal).

Phylogenetic analysis

The ITS and EF1-α sequences were combined and the dataset, including sequences of other Diplodia species downloaded from GenBank, was compiled with the outgroup Lasiodiplodia theobromae (Pat.) Griffon & Maubl. (Table 1). Sequences were aligned with ClustalX v. 1.83 (Thompson et al. 1997), using the following parameters: pairwise alignment parameters (gap opening = 10, gap extension = 0.1) and multiple alignment parameters (gap opening = 10, gap extension = 0.2, transition weight = 0.5, delay divergent sequences = 25 %). Alignments were checked and manual adjustments were made where necessary.

Phylogenetic analyses of sequence data were done using PAUP* v.4.0b10 (Swofford 2003) for Maximum-parsimony (MP) analyses and Mr Bayes v.3.0b4 (Ronquist and Huelsenbeck 2003) for Bayesian Inference (BI) analyses. The general time-reversible model of evolution (Rodriguez et al. 1990), including estimation of invariable sites and assuming a discrete gamma distribution with six rate categories (GTR+Γ+G) was used for BI analyses. Trees were visualized with TreeView (Page 1996).

Maximum-parsimony analyses were performed using the heuristic search option with 1,000 random taxon additions and tree bisection and reconnection (TBR) as the branch-swapping algorithm. All characters were unordered and of equal weight and gaps were treated as fifth character. Maxtrees were set to 500, branches of zero length were collapsed, and all multiple equally parsimonious trees were saved. The robustness of the most parsimonious trees was evaluated from 1,000 bootstrap replications (Hillis and Bull 1993). Other measures used were consistency index (CI), retention index (RI) and homoplasy index (HI).

Bayesian analyses employing a Markov Chain Monte Carlo method were performed. Four MCMC chains were run simultaneously, starting from random trees for 1,000,000 generations. Trees were sampled every 100th generation for a total of 10,000 trees. The first 1,000 trees were discarded as the burn-in phase of each analysis. Posterior probabilities (Rannala and Yang 1996) were determined from a majority-rule consensus tree generated with the remaining 9,000 trees. This analysis was repeated three times starting from different random trees to ensure trees from the same tree space were sampled during each analysis.

A comparison of highly supported clades (bootstrap support values ≥ 70 %) among trees generated from MP analyses of individual data sets was performed in order to detect conflict between individual phylogenies (Alves et al. 2008).

Results

DNA phylogeny

Approximately 550 and 300 bases were determined for ITS and EF1-α respectively. New sequences were deposited in GenBank (Table 1) and the alignment in TreeBase (14460). Individual gene phylogenies revealed no major conflicts thus indicating that the two loci could be combined. The combined ITS and EF1-α dataset consisted of 853 characters (including alignment gaps) for 87 ingroup and 2 outgroup taxa. Of the 853 characters 165 were excluded due to ambiguous alignment, 538 were constant and 9 were variable and parsimony-uninformative. Maximum parsimony analysis of the remaining 141 parsimony-informative characters resulted in 279 most parsimonious trees of 302 steps (CI = 0.717, RI = 0.945, HI = 0.283) and one is shown in Fig. 1.

Fig. 1
figure 1

One of the 279 most parsimonious trees resulting from the combined analysis of ITS and EF1-α sequence data. Bootstrap support and posterior probability values are given at the nodes. The tree was rooted to Lasiodiplodia theobromae. Ex-type isolates are in bold. The bar shows five changes

Full size image

In the phylogenetic analysis four main clades (labeled 1 to 4) were resolved in the ingroup (Fig. 1). These clades are characterized by distinct conidial morphological features as explained in detail by Phillips et al. (2012, 2013). Within these four main clades, 19 sub-clades corresponding to species were recognized; of which 17 represent known Diplodia species (Fig. 1). The D. mutila-like isolates studied were distributed into 5 sub-clades within clade 1. Thus, two isolates from P. alba were identified as Diplodia malorum Fuckel; one isolate from Q. coccifera was identified as Diplodia olivarum A.J.L. Phillips, Frisullo & Lazzizera and a set of isolates from P. alba, F. ornus and Vitis vinifera L. (grapevine) clustered in a larger group identified as D. mutila. For one of the remaining two clades the name Diplodia fraxini (Fr. : Fr.) Fr. was deemed to be appropriate and this name is reinstated for a set of isolates obtained from F. excelsior and F. angustifolia. The second clade containing isolates from F. excelsior, F. ornus and Lonicera nigra L. (black-berried honeysuckle) represents a previously unrecognized species, which is described here as Diplodia subglobosa sp. nov.

The remaining isolates, morphologically distinct from D. mutila, were distributed amongst clades 2 and 3 and were identified as Diplodia cupressi A.J.L. Phillips & A. Alves (one isolate from C. sempervirens), Diplodia seriata De Not. (one isolate from F. ornus and two isolates from F. excelsior) and Diplodia pseudoseriata C.A. Pérez, Blanchette, Slippers & M.J. Wingf. (two isolates from F. angustifolia).

Taxonomy

Diplodia Fr., in Montagne, Ann. Sci. Nat., Bot., 2e Sér., 1: 302. 1834.

MycoBank: MB8047

Diplodia fraxini (Fr. : Fr.) Fr., Summa Veg. Scand. 2: 417. 1849.

MycoBank: MB247549 (Fig. 2)

Fig. 2
figure 2

Diplodia fraxini. a Colony of typical D. fraxini after 7 days growth at 25 °C on PDA. b colony morphology of D. fraxini morphotype A growing on PDA. c conidia oozing from picnidia. d–e conidia developing on conidiogenous cells. f–k conidiogenous cells with periclinal thickenings (arrowed). l hyaline conidia of D. fraxini morphotype A. m hyaline conidia of D. fraxini. n hyaline and one pale brown aseptate conidia. o pale brown one septate conidium. p mature, brown conidium showing two septa. Bars: c = 1 mm; d–p = 10 μm

Full size image

≡ Sphaeria fraxini Fr. : Fr., Syst. Mycol. 2: 493. 1823.

≡ Botryodiplodia fraxini (Fr. : Fr.) Sacc., Syll. Fung. 3: 378. 1884.

Ascomata not seen. Conidiomata stromatic, pycnidial, solitary to aggregated, immersed, partially erumpent when mature, dark brown to black, globose, up to 600 μm diam., wall composed of three layers, an outer of dark brown, thick-walled textura angularis, a middle layer of dark brown thin-walled cells, an inner layer of thin-walled hyaline cells. Ostiole central, circular, papillate. Conidiophores absent. Conidiogenous cells (11–)12–15(–16.5) × (2.5–)3.5–4.5(–5) μm, holoblastic, discrete, cylindrical, hyaline, smooth, indeterminate, proliferating at the same level giving rise to periclinal thickenings, or proliferating percurrently to form one or two indistinct annellations. Conidia hyaline, aseptate, smooth, thick-walled, oblong to ovoid, straight, both ends broadly rounded, becoming pale brown to brown and one-two septate with age.

Significant differences in conidial dimensions were observed among the D. fraxini isolates considered in this work. In most of the isolates the conidia were (23.5−)25–27(–30) × (11−)12.5–13.5(−15) μm, 95 % confidence limits = 25.72–26.45 × 12.87–13.33 μm (mean ± S.D. of 50 conidia = 26.08 ± 1.31 × 13.10 ± 0.83 μm, L/W ratio = 2.00 ± 0.14), which correspond to the conidial dimensions reported by Saccardo (1884) and for this reason these isolates are here considered as typical of the species. Two isolates, one from Portugal and one from Sardinia, consistently produced longer conidia measuring (26.5−)29–31.5(–33) × (11−)12.5–14(−17.5) μm, 95 % confidence limits = 29.28–30.23 × 13.12–13.83 μm (mean ± S.D. of 50 conidia = 29.76 ± 1.72 × 13.47 ± 1.28 μm, L/W ratio = 2.22 ± 0.19). Since these isolates are phylogenetically indistinguishable from the typical ones we regard these as a morphological variant and report them as Diplodia fraxini morphotype A.

Cultural characteristics: Colonies on PDA grew moderately, reaching 90 mm diameter or less in 7 days at 25 °C, the mycelium was sparsely to moderately aerial, surface white at first and later turned pale to dark grey and greyish to brown in reverse (Fig. 3). Colonies of morphotype A on PDA grey-brown with dense aerial mycelium.

Fig. 3
figure 3

Diplodia mutila. a Sectioned ascoma. b Immature asci and pseudoparaphyses. c, d Asci with ascospores. e, f Ascospores. g Conidiomata partially erumpent through host. h Sectioned conidioma. i–l Conidiogenous cells. m–p Conidia. m Hyaline, aseptate conidia of CBS 112553. n Hyaline, aseptate conidia of CBS 136014. o Hyaline, aseptate conidia of BPI 599153. p Hyaline, aseptate conidia of K(M) 99664. Scale bars a = 100 μm, b = 10 μm, e, f = 10 μm, g = 500 μm, h = 100 μm, i, l = 10 μm, m = 10 μm. Scale bar in b applies to c, d. Scale bar in i applies to j, k. Scale bar in m applies to n, o, p

Full size image

Cardinal temperatures for growth: minimum 5 °C, maximum <35 °C and optimum 25 °C.

Habitat: On branches of F. angustifolia and F. excelsior.

Known distribution: Italy, Portugal and the Netherlands (this paper), Sweden, Germany, Italy, France (Saccardo 1884).

Specimens examined: PORTUGAL, Monte da Caparica, on dead twigs of Fraxinus angustifolia, 14 March 2013, Antonio Deidda (LISE 96134, neotype designated herein), MBT176183, culture ex-neotype CBS 136010 = CAD001. Cascais, on dead twigs of F. angustifolia, 13 April 2013, Antonio Deidda, designated morphotype A, culture CBS 136012 = CAD010. ITALY, Bortigiadas, isolated from a branch canker of F. angustifolia, 03 June 2011, Benedetto T. Linaldeddu, culture CBS 136011 = BL70. ITALY, Siliqua, isolated from a branch canker of F. angustifolia, 11 November 2009, Benedetto T. Linaldeddu, CBS 136013 = BL16 (morphotype A). Additional isolates are given in Table 1.

Notes: Fries (1849) did not refer specifically to any previous description and gave only a brief Latin comment “Vidi triloc.” (= “I have seen trilocular [pycnidia]”). Nevertheless, the binomial Diplodia fraxini should be interpreted as a recombination based on Sphaeria fraxini as suggested by Saccardo (1884). The holotype of neither S. fraxini nor D. fraxini (on a branch of Fraxinus sp. collected in Sweden by Fries) could be located and are presumed lost. For this reason a neotype (LISE 96134) is designated here. All except two of the isolates studied here conformed morphologically to Saccardo’s (1884) description of the species. The morphotype A isolates, with large conidia, were phylogenetically indistinguishable from typical D. fraxini isolates and thus were considered to be a morphological variant of D. fraxini.

Diplodia mutila (Fr. : Fr.) Fr., Summa Veg. Scand. 2: 417. 1849.

MycoBank: MB201741 (Fig. 3)

≡ Sphaeria mutila Fr. : Fr., Syst. Mycol. 2: 424. 1823.

≡ Physalospora mutila (Fr. : Fr.) N.E. Stevens, Mycologia 28: 333. 1936.

≡ Botryosphaeria stevensii Shoemaker, Can. J. Bot. 42: 1299. 1964.

Ascomata unilocular, solitary or clustered, immersed, partially erumpent when mature, globose, up to 300 μm diam., dark brown to black, thick-walled, wall composed of outer layers of thick-walled, dark-brown textura angularis, inner layers of thin-walled, hyaline, textura angularis. Ostiole central, circular, papillate, periphysate. Pseudoparaphyses hyaline, branched, septate, 2–3 μm wide, constricted at septa. Asci clavate, stipitate, bitunicate with a thick endotunica and well-developed apical chamber, 100–160 × 14–22 μm (including stipe), containing eight, biseriate ascospores. Ascospores (25–)28–35(–36) × (9.5–)10–12.5(–13.5) μm; 95 % confidence limits = 30.8–32.1 × 11.2–11.7 μm (mean ± S.D. of 50 ascospores = 31.5 ± 2.3 × 11.4 ± 0.9 μm) with L/W of 2.8 ± 0.3, fusiform to oval, widest in the middle, both ends obtuse, hyaline, thin-walled, smooth, aseptate, rarely becoming light brown with age. Conidiomata solitary or aggregated in clusters of up to five or more, immersed, partially erumpent when mature, dark brown to black, more or less globose, up to 600 μm diam., wall composed of three layers, an outer of dark brown, thick-walled textura angularis, a middle layer of dark brown thin-walled cells, an inner layer of thin-walled hyaline cells. Ostiole central, circular, papillate. Conidiophores absent. Conidiogenous cells (11−)12–15 × 4–5 μm, holoblastic, discrete, cylindrical, hyaline, smooth, indeterminate, proliferating at the same level giving rise to periclinal thickenings, or proliferating percurrently to form one or two indistinct annellations. Conidia hyaline, aseptate, smooth, thick-walled, oblong to ovoid, straight, both ends broadly rounded, (20−)21.5–25.5(–27.5) × (9.5−)12–14(−15.5) μm, 95 % confidence limits = 24.69–25.73 × 13.26–13.78 μm (mean ± S.D. of 50 conidia = 25.4 ± 1.0 × 13.4 ± 0.5 μm, L/W ratio = 1.9 ± 0.1), rarely becoming pale brown and one-septate with age. Table 2 illustrate conidial sizes for all species of Diplodia belonging to clade 1.

Table 2 Conidial sizes and L/W ratios
Full size table

Habitat: While Farr and Rossman (2013) list 55 hosts for D. mutila it is now clear that many of the earlier reports of this fungus could be misidentifications (Alves et al. 2004, 2006; Lazzizera et al. 2008; Phillips et al. 2012). The following are confirmed hosts: Chamaecyparis lawsoniana, Fraxinus spp., Malus spp., Populus spp., Taxus baccata, Vitis vinifera (Phillips et al. 2013).

Known distribution: England, France, Italy, Portugal, South Africa, USA (California) (Phillips et al. 2013), New Zealand (Dingley 1969; Laundon 1973).

Specimens examined: Diplodia mutila: FRANCE, Ardenne, Sedan, on bark of Populus nigra, date unknown, Montagne, (K 99664, isotype). PORTUGAL, Alentejo, Montemor-o-Novo, Vitis vinifera, 1996, A.J.L. Phillips (CBS H-20187), living culture CBS 112553. PORTUGAL, Beira Litoral, Aveiro, Populus alba, 2012, A. Alves, (LISE 96136, epitype of Diplodia mutila designated herein), MBT176182, culture ex-epitype CBS 136014. Sphaeria mutila: Scler. Suec. 164 (STR); Scler. Suec. 385 (STR). Physalospora mutila: ENGLAND, Cornwall, Saltash, on bark of Malus sp., 22 Aug. 1935, N.E. Stevens (BPI 599153, lectotype); Surrey, Ranmore Common, on Fraxinus sp., 19 Apr. 1957, C. Booth (BPI 599150 ex IMI 69064).

Notes: Although Montagne (1834) indicated Sphaeria mutila as the type of the new genus Diplodia, the 1834 protologue did not make any definite association of “mutila” with “Diplodia”, as required for a valid comb. nov. Therefore, the frequently cited date of 1834 for publication of the combination Diplodia mutila is incorrect. Fries (1823) described Sphaeria mutila and distributed two exsiccati under that name as Scler. Suec. 164 and 385. Stevens (1933) and Sutton (1980) reported that these two exsiccati in BPI and K had no spores. Alves et al. (2004) examined material of the same two exsiccati in STR and also found no spores. Montagne sent Fries a fungus that was identified as S. mutila. The record was listed under S. mutila Fr. by Montagne (1834) with the note that this would become the type of a new genus, Diplodia, later characterized by Fries (1849). Thus, the binomial Diplodia mutila was first introduced by Fries (1849). Montagne distributed this fungus in his exsiccatus No. 498. According to Alves et al. (2004) no material of this exsiccatus could be found in STR. Alves et al. (2004) examined Montagne’s specimen of D. mutila in Kew, K(M) 99664 (presumed to be an isotype) and found it to agree in all aspects with Stevens’ (1933) account of Montagne’s exs. 498. Stevens (1933) described Physalospora mutila as the teleomorph of D. mutila referring to BPI 599151, but this name was invalid because it lacked a Latin description. Alves et al. (2004) examined this specimen and could find no teleomorph, but they did find ample material of the telomorph on BPI 599153, which is a specimen on apple collected by Stevens from the same locality at same time he collected BPI 599151. Shoemaker (1964) considered the teleomorph to be a species of Botryosphaeria and since the name B. mutila was already taken, he proposed the name Botryosphaeria stevensii. After Crous et al. (2006) revised Botryosphaeria reducing it to B. dothidea (Moug.) Ces. & De Not. and B. corticis (Demaree & Wilcox) Arx & E. Müll., the fungus known as B. stevensii was referred to only by its anamorphic name D. mutila. The epitype designated here conforms in all ways with the isotype of D. mutila and with the asexual morph on BPI 599153 as described by Alves et al. (2004).

Diplodia subglobosa A.J.L. Phillips, Deidda & Linaldeddu sp. nov.

MycoBank: MB806049 (Fig. 4)

Fig. 4
figure 4

Diplodia subglobosa. a Colony on PDA after 7 days at 25 °C. b, c Conidiomata formed in culture on pine needles. d–h Conidiogenous cells with developing conidia. i Hyaline, aseptate conidia. j Hyaline, aseptate and coloured, 1-septate conidia. Bars: b, c = 500 μm; d–j = 10 μm

Full size image

Etymolog y: Named for the sub-globose conidia.

Ascomata not seen. Conidiomata solitary, immersed in the host, dark brown to black, globose to ovoid, up to 560 μm diam and 400 μm high, wall composed of three layers, an outer of dark brown, thick-walled textura angularis, a middle layer of dark brown thin-walled cells, an inner layer of thin-walled hyaline cells. Ostiole central, circular, papillate. Conidiogenous cells 11–25 × 3–9 μm, holoblastic, discrete, cylindrical, hyaline, smooth, indeterminate, proliferating at the same level giving rise to periclinal thickenings, or proliferating percurrently to form one or two indistinct annellations. Conidia hyaline, aseptate, smooth, thick-walled, oblong to ovoid, straight, both ends broadly rounded, (24.0−)24.5–27.0(−32) × (15.5−)16.5–19.0(−22) μm, 95 % confidence limits = 27.7–28.0 × 18.1–18.5 μm (mean ± S.D. of 150 conidia = 27.7 ± 1.8 × 18.3 ± 1.2 μm, L/W ratio = 1.5 ± 0.1), becoming pale brown and septate when aged.

Habitat: Twigs and branches of Fraxinus spp. and Lonicera nigra.

Known geographic distribution: Italy and Spain.

Specimens examined: ITALY: Sicily, Fraxinus ornus, 2006, A. Sidoti (CBS 124131). SPAIN: Cataluña, Fraxinus excelsior, (no date), J. Luque (CBS 124132); Lonicera nigra, (no date), J. Luque Holotype LISE 96135 (culture ex-type CBS 124133). ITALY, Fraxinus excelsior, (no date) B. Slippers (CMW7776).

Notes: The relatively wide conidia and L/W ratio of 1.5 of this species are distinctive amongst Diplodia species with hyaline conidia belonging to clade 1.

Discussion

Results of phylogenetic analyses provide robust evidence that the D. mutila-like isolates from Fraxinus spp. in Italy, the Netherlands, Portugal and Spain belong to three separate sub-clades corresponding to three distinct species within Diplodia clade 1. For one clade, which includes several isolates from F. angustifolia in Italy and Portugal, and an isolate from F. excelsior in the Netherlands (CBS 431.82), the name D. fraxini was considered appropriate. The type of neither S. fraxini nor D. fraxini could be located, but morphologically the isolates we studied agree in all ways with Saccardo’s (1884) description of the species and therefore the name is re-instated and a neotype designated. Two isolates, one from Italy and one from Portugal, differed in cultural characteristics and conidia dimensions from typical isolates and since they were phylogenetically indistinguishable from the typical isolates they were considered to be morphological variants and are designated as morphotype A. The descriptions of distinct morphological forms has been reported for another species in Diplodia, namely D. sapinea (morphotype A and C) and the C morphotype is considered to be the most virulent (de Wet et al. 2002).

The second clade includes isolates from F. ornus in Italy, F. excelsior in Italy and Spain and one isolate from Lonicera nigra in Spain and represents a previously unrecognized Diplodia species, which we describe here as D. subglobosa sp. nov. Conidia of this new species are sub-globose with an average L/W ratio of 1.5. In this respect it resembles the anamorph of “Botryosphaeria” quercuum (Shoemaker 1964) but it can be distinguished on account of its larger conidia (24.0−)24.5–27.0(−32) × (15.5−)16.5–19.0(−22) μm. In “Botryosphaeria” quercuum the conidia are (18–)21–24(–25) × (12–)15–16(–17).

The third group of isolates obtained from F. ornus in Portugal were considered to be D. mutila, the type species of the genus Diplodia (Montagne 1834; Fries 1849). Unfortunately, no live cultures linked to the holotype of D. mutila are extant and this has severely hampered studies on the taxonomy and phylogeny of Diplodia. Alves et al. (2004) provided a detailed description of this species based on one isolate from grapevines in Portugal (CBS 112553), an isotype of D. mutila (K99664) and one of Stevens’ (1936) specimens of Physalospora mutila (BPI 99153). They showed that CBS 112553 correlated closely with the morphology of D. mutila and this was confirmed in the present study. This culture has subsequently been cited as typical of D. mutila and has been regarded as a standard isolate for this species (Alves et al. 2006; Damm et al. 2007; Lazzizera et al. 2008; Phillips et al. 2012; Lynch et al. 2013). Although it is possible to use as epitype a specimen collected from a different host it is preferable, whenever possible, to obtain the epitype from the same host as the type specimen.

In an effort to obtain a suitable specimen that can be used as epitype for D. mutila we collected samples from Populus (type host of D. mutila) and Fraxinus (type host of the sexual morph). No sexual morph was found on either of the hosts, but several isolates with typical morphological features of D. mutila were obtained. Isolates obtained from P. alba and F. ornus in Portugal clustered with isolates from several other hosts including V. vinifera, Persea americana Mill., Taxus baccata L. and Chamaecyparis lawsoniana (A. Murray bis) Parl. CBS 302.36, deposited by N. E. Stevens as Physalospora mutila N.E. Stevens also clustered in this group. Since the culture is no longer sporulating, most likely due to the fact that it has been in culture for many years now, we could not study its morphology. This culture is not linked in any way to the type specimen but given that it was obtained by Stevens it can be regarded as representative of his concept of the sexual morph of D. mutila. Thus, this clade is regarded as representing true D. mutila. The isolates from P. alba and F. ornus conformed in all ways with the morphological characters of the isotype of D. mutila (K99664), the asexual morph of P. mutila on BPI 599153 (Alves et al. 2004) and the isolate from grapevines in Portugal (CBS 112553) that has been used as a standard. Furthermore, the type host of D. mutila is a Populus species. Therefore, the specimen on P. alba is herein designated as epitype for D. mutila. The data presented here supports the plurivorous nature of this pathogen since isolates from several different hosts clustered together in the clade corresponding to this species.

Diplodia mutila clustered in the Diplodia clade 1, which includes species with hyaline conidia that become brown and one-septate some time after discharge from the pycnidia. This clade comprises eight species that are morphologically similar and can be difficult to separate on morphology alone. Nonetheless, they can be differentiated on slight differences in conidial dimensions (Table 2). Thus, conidia of Diplodia africana Damm & Crous are the largest in the clade 1, while D. olivarum has the smallest conidia. Although conidia of D. malorum are morphologically similar to those of Diplodia rosulata Gure, Slippers & Stenlid this last species forms distinctive rosulate colonies (Gure et al. 2005). Diplodia agrifolia S.C. Lynch & A. Eskalen differs from D. mutila by its longer and wider conidia. Moreover, conidia of D. agrifolia are hyaline and aseptate, but most become dark brown and one-septate before discharge from pycnidia (Lynch et al. 2013).

Another group of isolates obtained in this study from cankered branches of declining P. alba trees in Italy clustered in another clade together with D. malorum. The name D. malorum was reinstated by Phillips et al. (2012) for isolates obtained from Malus spp. Although this species is morphologically similar to D. mutila it can be distinguished by its larger conidia and from DNA sequence data (Phillips et al. 2012, 2013). Until now D. malorum has been reported only from Malus spp. and other Rosaceae and this represents the first report of the species in a different host (Populus) and a different family. In addition, it is reported for the first time in Italy.

A single isolate obtained from a cankered branch of Quercus coccifera in Tunisia clustered within the D. olivarum clade. This species was initially described from Olea europaea L. (olive tree) in Italy where it was associated with diseased olive drupes (Lazzizera et al. 2008). Since then, it has been found associated with cankers on Ceratonia siliqua L. (carob tree) in Italy (Granata et al. 2011) and trunk disease of Prunus dulcis (Mill.) D.A. Webb (almond) in Spain (Gramaje et al. 2012). This represents the first report of the species in Tunisia and first report on Q. coccifera.

Additional Diplodia isolates morphologically different from D. mutila were included in this study. One isolate from F. ornus in Portugal and two isolates from F. angustifolia in Italy clustered within D. seriata. This is the first report of the species on Fraxinus spp. in Europe. The only other known report is on Fraxinus americana L. (green ash) in the USA under the name Physalospora obtusa (=Botryosphaeria obtusa) (Farr et al. 2013) which is the sexual morph of D. seriata (Phillips et al. 2007). This species is well known by its cosmopolitan and plurivorous nature (Punithalingam and Walker 1973; Phillips et al. 2007).

Another two isolates from F. angustifolia in Italy clustered with D. pseudoseriata and D. alatafructa. Diplodia pseudoseriata was described from several species of native Myrtaceae trees in Uruguay (Pérez et al. 2010) while Diplodia alatafructa Mehl & Slippers was first reported on Pterocarpus angolensis DC. (wild teak) in South Africa (Mehl et al. 2011). Interestingly, the two Italian isolates form a distinct sub-clade within D. pseudoseriata/D. alatafructa clade with 75 % bootstrap support. Further morphological and phylogenetic analysis are currently in progress to clarify the status of these two isolates.

In this study we have shown that the name D. mutila has been applied to a number of cryptic species. In order to stabilize the name and allow its unambiguous application an epitype specimen with associated ex-epitype cultures was selected. At the same time the name D. fraxini is re-instated and a neotype designated. In the future more studies should be done in order to verify the pathogenicity of D. mutila, D. fraxini and D. subglobosa on Fraxinus spp. and establish their role in the etiology of ash dieback.