European species of Dendrostoma (Diaporthales)

Abstract European species of the genus Dendrostoma (Erythrogloeaceae, Diaporthales) occurring on Castanea sativa and Quercus spp. based on freshly collected material are presented. Using a matrix of sequences from ITS, LSU, rpb2, and tef1, five species are recognized, and their phylogenetic positions are determined. Four species are added to the 14 described species of Dendrostoma. Dendrostoma atlanticum on Castanea sativa, D. creticum on Quercus coccifera and D. istriacum on Q. ilex are described as new species, Valsa castanea is combined in Dendrostoma, and D. leiphaemia is redescribed and illustrated. A key to the European species of Dendrostoma is provided.


Introduction
The genus Cryptodiaporthe, based on Cryptospora aesculi, is one of several segregates from the large genus Diaporthe (Diaporthales), characterized by the lack of stromatic zones and with asexual morphs recognized by Petrak (1921) as Septomyxa. In 1933, Wehmeyer (1933 recognized the relatively large number of species with a simple type of stroma development having various asexual morphs as a heterogeneous grouping. Petrak (1971) removed C. tiliae (as C. hranicensis) to his new genus Amphiporthe, mainly due to its Amphicytostroma asexual morph, where subsequently several species were added. Using the phylogenetic markers ITS, LSU, and rpb2, Mejía et al. (2008) detected that C. aesculi is congeneric with the generic type of Plagiostoma, P. euphorbiae. Thus, Cryptodiaporthe became a synonym of Plagiostoma. Subsequently (Mejía et al. 2011), several other species of Cryptodiaporthe were combined in Plagiostoma. Since the first phylogenetic treatment of the Diaporthales using DNA data (Castlebury et al. 2002), many old genera have been split and new ones described, and the proliferation of family names has forwarded a current number of 28, more than a half of which having been erected during the last three to four years (compare Jaklitsch et al. (2016), who listed 11 families). One of these families is the Erythrogloeaceae, whose members are based on phytopathogenic coelomycetous fungi (Chrysocrypta, Disculoides, Erythrogloeum). The only genus of this family for which sexual morphs are known is Dendrostoma (Fan et al. 2018). This genus is characterized by features common to many other diaporthalean genera forming pseudostromata lacking black stromatic margins, including Amphiporthe and Plagiostoma (Cryptodiaporthe). Rossman et al. (2015) already noted that Amphiporthe castanea and A. leiphaemia are not congeneric with A. tiliae (syn. A. hranicensis) and would need a new generic name. Amphiporthe leiphaemia was combined in Dendrostoma by Senanayake et al. (2018), based on ITS and LSU sequences of a CBS strain without giving any further information, whereas A. castanea has not been treated recently, although Jiang et al. (2019), who substantially enlarged the scope of the genus by describing 10 new species from Castanea and Quercus in China, recognized seven species on Castanea mollissima. Here we report on recently collected species of Dendrostoma occurring on Castanea sativa and Quercus spp. in Europe.

Sample sources
All isolates included in this study originated from ascospores of freshly collected specimens derived from recently dead branches or twigs. Details of the strains including NCBI GenBank accession numbers of gene sequences used to compute the phylogenetic trees are listed in Table 1. Strain acronyms other than those of official culture collections are used here primarily as strain identifiers throughout the work. Representative isolates have been deposited at the Westerdijk Fungal Biodiversity Centre (CBS-KNAW), Utrecht, The Netherlands. Details of the specimens used for morphological investigations are listed in the Taxonomy section under the respective descriptions. Freshly collected specimens have been deposited in the Fungarium of the Department of Botany and Biodiversity Research, University of Vienna (WU).

Morphology
Microscopic observations were made in tap water except where noted. Morphological analyses of microscopic characters were carried out as described by Jaklitsch (2009). Methods of microscopy included stereomicroscopy using a Nikon SMZ 1500 and Nomarski differential interference contrast (DIC) using the compound microscopes Nikon Eclipse E600 or Zeiss Axio Imager.A1 equipped with a Zeiss Axiocam 506 colour digital camera. Images and data were gathered using a Nikon Coolpix 4500 or a Nikon DS-U2 digital camera and measured by using the NIS-Elements D v. 3.0 or 3.22.15 or Zeiss ZEN Blue Edition software packages. For certain images of ascomata the stacking software Zerene Stacker v. 1.04 (Zerene Systems LLC, Richland, WA, USA) was used. Measurements are reported as maxima and minima in parentheses and the range representing the mean plus and minus the standard deviation of the number of measurements given in parentheses.

Phylogenetic analyses
The newly generated sequences were aligned with the sequences of Jiang et al. (2019), and a combined matrix of the three loci (partial SSU-ITS-LSU rDNA, rpb2, and tef1) was produced for phylogenetic analyses, with three species (Chrysocrypta corymbiae, Disculoides eucalypti, and Disculoides eucalyptorum) added as the outgroup according to Jiang et al. (2019). The GenBank accession numbers of sequences used in the analyses are given in Table 1. Sequence alignments were produced with the server version of MAFFT (http://mafft.cbrc.jp/alignment/server/), checked and refined using BioEdit v. 7.2.6 (Hall 1999). The combined data matrix contained 4194 characters, viz. 1637 nucleotides of SSU-ITS-LSU, 1075 nucleotides of rpb2, and 1482 nucleotides of tef1. Maximum parsimony (MP) analyses were performed with PAUP v. 4.0a165 (Swofford 2002). All molecular characters were unordered and given equal weight; analyses were performed with gaps treated as missing data; the COLLAPSE command was set to MINBRLEN. MP analysis of the combined multilocus matrix was done using 1000 replicates of heuristic search with random addition of sequences and subsequent TBR branch swapping (MULTREES option in effect, steepest descent option not in effect). Bootstrap analyses with 1000 replicates were performed in the same way but using 10 rounds of random sequence addition and subsequent branch swapping during each bootstrap replicate.
Maximum likelihood (ML) analyses were performed with RAxML (Stamatakis 2006) as implemented in raxmlGUI 1.3 (Silvestro and Michalak 2012), using the ML + rapid bootstrap setting and the GTRGAMMA substitution model with 1000 bootstrap replicates. The matrix was partitioned for the different gene regions. In the Results and Discussion, bootstrap values below 70% are considered low, between 70-90% medium, and above 90% high.

Phylogenetic analyses
Of the 4194 characters included in the phylogenetic analyses, 703 were parsimony informative (133 from the SSU-ITS-LSU, 247 from rpb2, 323 from tef1). MP analyses revealed eight MP trees 1552 steps long, one of which is shown as Figure 1. The tree backbone was identical in all MP trees, except for the position of Dendrostoma castaneicola, which was embedded within D. castaneae in some of the MP trees (not shown). The best ML tree (lnL = −13985.7598) revealed by RAxML was compatible with the MP strict consensus tree, except for an interchanged position of D. atlanticum and D. shaanxiense (not shown). The genus Dendrostoma received maximum and medium support in the MP and ML analyses, respectively, and most of the tree backbone received significant support as well ( Fig. 1). Although Dendrostoma accessions from Quercus and Castanea were interspersed, host-related patterns were obvious in the various Dendrostoma subclades ( Description, emended here. Sexual morph: pseudostromata immersed in bark and erumpent, causing a pustulate bark surface, consisting of an ectostromatic disc and entostroma with embedded ascomata. Ectostromatic disc flat or convex, surrounded by bark flaps. Entostroma light-coloured, prosenchymatous to nearly pseudoparenchymatous, mixed with bark cells, sometimes forming a more-or-less conical central column beneath the disc. Stromatic zones lacking or sometimes bark dorsally darkened. Ascomata perithecial, subglobose. Ostioles flat in the disc or slightly projecting, cylindrical, often with conical apical part. Paraphyses deliquescent. Asci oblong, fusoid, narrowly clavate or subellipsoid, with a refractive apical ring, containing (4-)6-8 ascospores in various arrangements, becoming detached at maturity. Ascospores hyaline, ellipsoid, fusoid, oblong to subacicular, often inequilateral, straight to curved,  Figure 1. Phylogram showing one of 8 MP trees 1552 steps long revealed by PAUP from an analysis of the combined ITS-LSU-rpb2-tef1 matrix of Dendrostoma, with Chrysocrypta corymbiae, Disculoides eucalypti and D. eucalyptorum added as outgroup taxa. MP and ML bootstrap support above 50% are given above or below the branches. The asterisk (*) denotes the node collapsed in the strict consensus of the eight MP trees. Accessions in bold were sequenced in the present study; accessions in blue were isolated from Castanea, those in green from Quercus, in orange from Malus and in red from Osmanthus.
Culture characteristics. On CMD at 16 °C in the dark colony more-or-less circular, of loose mycelium, first white, variably covered by white aerial hyphae, becoming dense, forming white and apricot to orange zones, darkening and turning black from the centre, sometimes forming reddish brown dots, spots or tubercles.
Asexual morph co-occurring with the sexual morph, acervular, pulvinate, scattered to aggregated, 0.5-2.7 mm in diameter, appearing as superficial discs 0.3-2 mm in diameter, with undulate surface, cream to pale brown and becoming brittle in the centre and nearly black at the periphery and often also indicated as dark zone on the bark surface around the disc; inside consisting of a pale or yellowish brown, loose and brittle central column consisting of pale brown t. prismatica and a lateral ring-like, dense, white to distinctly yellow fertile part with even or undulating margin, the latter also raising above the column, outside surrounded by a partly undulating, ca 20-25 µm thick black wall consisting of dark brown textura angularis of cells 4-10 µm in diameter at apical and upper peripheral regions, becoming paler downward and being absent at the base and lower sides. Interior of the fertile chambers consisting of isodiametric to elongate hyaline supporting cells and richly and irregularly branched hyphal conidiophores bearing phialides and conidia. Wall, supporting cells and phialides turn- ing dilute violaceous in 3% KOH. Phialides arranged on supporting cells in palisades along the walls and on conidiophores, (6-)8.2-12(-15.3) × (1.7-)2.5-3.5(-5) µm (n = 80), repetitive, mostly lageniform, often with long necks; conidia also formed on cylindrical pegs and denticles. Conidia (6-)6.7-8(-8.8) × (2.5-)3-3.5(-3.7) µm, l/w (1.7-)2.1-2.6(-3.1) (n = 85), subfusiform, subclavate or ellipsoid, scar often distinct, smooth, with few minute drops. Notes. Sizes of pseudostromata and acervuli strongly depend on twig thickness. Remarkably, red colour of the ostiolar tip, when present, turns purple in 3% KOH and yellow in lactic acid, a feature, which is typical of the Hypocreales and within the Diaporthales otherwise only found in the Cryphonectriaceae.
So far, confirmed records of D. castaneum are only known from Europe where the species is widely co-occurring with its host, Castanea sativa. Kobayashi (1970) reported and illustrated D. castaneum (as Cryptodiaporthe castanea) from Castanea crenata and C. mollissima in Japan. However, it is unlikely that these collections are conspecific with the European D. castaneum, considering their different spore shape and hosts. The 1 or 2 large guttules per ascospore cell and the ascospore appendages illustrated in Kobayashi (1970: fig. 32) are similar to D. atlanticum rather than to D. castaneum. Remarkably, he also reported and illustrated dimorphic conidia for the Japanese collections, which we also observed in D. atlanticum. Considering hosts and distribution, the Japanese collections likely represent one of the species described by Jiang et al. (2019) or an undescribed species.
Asexual morph unknown. Culture characteristics. On CMD at 16 °C in the dark colony circular to irregular, dense, white, partly covered by short, white aerial hyphae, zonate, soon turning dark brown to black with pale apricot spots and margin and apricot to orange pigment diffusing into agar, reverse dark brown with orange margin.
Notes. Dendrostoma creticum is similar to the closely related D. istriacum but differs by distinctly longer ascospores, darker ectostromatic discs and a different host species.
Culture characteristics. On CMD at 16 °C in the dark, colony circular to irregular, dense, white, partly covered by short white aerial hyphae, zonate, soon turning dark brown to black with pale apricot to reddish brown spots and margin and some pale apricot pigment diffusing into agar, reverse dark brown with pale apricot margin.
Notes. Dendrostoma istriacum is closely related to D. creticum but differs from that species by distinctly shorter ascospores and a different host species.  Diagnosis. Dendrostoma leiphaemia is recognized by conspicuous ectostromatic discs, broad conical ostiolar necks, and broad multiguttulate ascospores.
Culture characteristics. On CMD at 16 °C in the dark, colony irregular or dimorphic, dense, white, partly covered by short white aerial hyphae, zonate, soon turning dark brown to black with red or reddish brown spots, reverse dark brown, reddish brown with white, pale apricot or reddish brown spots and margins. Notes. Asexual fructifications of this species are reported to have dimorphic conidia (Butin 1980;Wehmeyer 1933). However, for the description above only overmature material with a single type of conidia was available, the measurements of which agree with the cylindrical form given as 7-12 × 1.5-2 µm by Wehmeyer (1933), but their shape is more variable, possibly due to their age. As Butin (1980) observed, the asexual morph precedes the sexual morph and may still be present as separate acervuli among sexual pseudostromata or as locules within the periphery of the latter.

Discussion
Our phylogenetic analyses are largely congruent with those of Jiang et al. (2019), and different topological positions of, e.g., D. aurorae and D. parasiticum concern backbone nodes with low to medium support. A notable difference concerns the position of the generic type, D. leiphaemia, which in Jiang et al. (2019) is contained within the D. osmanthi -D. qinlingense -D. quercus clade with medium (80% MP) to high (90% ML) support, while in our analyses it is placed basal to the D. dispersum -D. mali -D. quercinum clade with medium support (76% MP, 88% ML). These differences may be due to different taxon and marker sampling, as in the analyses of Jiang et al. (2019) only the ITS and LSU rDNA were available for D. leiphaemia. Previous authors recorded phytopathogenic potential in all species of Dendrostoma studied by them (Fan et al. 2018;Jiang et al. 2019). As an example, Dendrostoma castaneicola, D. castaneae, and D. shaanxiense were reported to cause chestnut canker (termed "Dendrostoma canker") on Castanea mollissima in China (Jiang et al. 2019). It is remarkable that almost all Chinese Dendrostoma species recorded as canker pathogens by Jiang et al. (2019) were only found as asexual morphs, which were abundantly produced on the dead twigs. This may, at least partly, be linked to the fact that Jiang et al. (2019) mainly investigated chestnut plantations, in which asexual reproduction of virulent pathotypes may be particularly favoured by genetically uniform host cultivars. However, pathogenicity of these species has not been confirmed by inoculation experiments. Défago (1937) observed canker disease symptoms of Castanea sativa after artificial inoculation with Dendrostoma castaneum, and Kobayashi (1970) mentioned unpublished inoculation experiments showing pathogenicity of Dendrostoma sp. (as Cryptodiaporthe castanea) on cultivated Japanese chestnut varieties. Phillips and Burdekin (1982) considered D. castaneum to be a weak wound pathogen. In our studies, we have not seen any obvious disease symptoms exhibited by Castanea and Quercus species infected by species of Dendrostoma. The typical habitat of species like D. castaneum or D. leiphaemia are cut branches piled up on the ground. Species on evergreen Quercus spp. may occur on dead branchlets attached to trees, but their appearance is rather inconspicuous, and specific searches are necessary to spot them. However, as our observations have not been conducted to specifically study disease symptoms, it is premature to make predictions about potential pathogenicity, which thus cannot be excluded. Frequent association of Dendrostoma spp. with Cytospora spp. may suggest weak or facultative parasitism, but inoculation experiments are required to prove pathogenicity by fulfilling Koch's postulates.
Although other genera of the Erythrogloeaceae produce acervuli, asexual morphs of Dendrostoma have been termed pycnidia (Fan et al. 2018;Jiang et al. 2019). This may be due to studies in culture, as asexual fructifications on agar may easily be interpreted as pycnidia, even when no true ostioles are present. However, none of the asexual morphs of the European species we have seen on natural substrates have preformed openings that may be termed ostioles. Therefore, we recognize asexual fructifications of Dendrostoma on natural substrates generally as acervuli. Jiang et al. (2019) found dimorphic conidia in a single species of Dendrostoma, D. quercus. Here we add another such species, D. atlanticum. These forms occur at the same time in the same asexual fructifications. However, to gain a complete picture of asexual morphs and elucidate entire life cycles of Dendrostoma species, long-term studies may be required, as certain asexual fungi have two different morphs, which may not occur at the same time (Butin 1980).
Most species of Dendrostoma are only known as asexual morphs. Only one of the 10 species described by Jiang et al. (2019), D. quercus, has a sexual morph. However, it is unclear whether in these species sexual morphs are absent, only rarely produced or have not yet been recorded, e.g., due to unfavourable weather conditions for development, unsuitable substrates or an untimely sampling season. Other species, for which sexual morphs are known are D. mali on Malus spectabilis, D. osmanthi on Osmanthus fragrans, and D. quercinum on Quercus acutissima (Fan et al. 2018). All five species of Dendrostoma we describe or redescribe from Europe, two from Castanea sativa and three from Quercus spp., have sexual morphs and in all but one (D. creticum) we found also an asexual morph on the natural hosts.
The high species biodiversity of Dendrostoma recorded from Eastern Asia as well as the phylogenetic patterns indicate that the group may have originated in this area. This is also supported by the fact that the European species do not form a monophyletic group, but are embedded within Eastern Asian lineages, indicating that Europe has been colonised from Asia several times independently. In addition, evolutionary radiation may have started on Castanea as the basal subclade A exclusively contains accessions from that host (Fig. 1). However, detailed additional studies including other areas as well as hosts are necessary to vigorously test these hypotheses.