Fenestelloid clades of the Cucurbitariaceae

Fresh collections and their ascospore and conidial isolates backed up by type studies and molecular phylogenetic analyses of a multigene matrix of partial nuSSU-, complete ITS, partial LSU rDNA, rpb2, tef1 and tub2 sequences were used to evaluate the boundaries and species composition of Fenestella and related genera of the Cucurbitariaceae. Eight species, of which five are new, are recognised in Fenestella s.str., 13 in Parafenestella with eight new species and two in the new genus Synfenestella with one new species. Cucurbitaria crataegi is combined in Fenestella, C. sorbi in Synfenestella, Fenestella faberi and Thyridium salicis in Parafenestella. Cucurbitaria subcaespitosa is distinct from C. sorbi and combined in Neocucurbitaria. Fenestella minor is a synonym of Valsa tetratrupha, which is combined in Parafenestella. Cucurbitaria marchica is synonymous with Parafenestella salicis, Fenestella bavarica with S. sorbi, F. macrospora with F. media, and P. mackenziei is synonymous with P. faberi, and the latter is lectotypified. Cucurbitaria sorbi, C. subcaespitosa and Fenestella macrospora are lecto- and epitypified, Cucurbitaria crataegi, Fenestella media, F. minor and Valsa tetratrupha are epitypified in order to stabilise the names in their phylogenetic positions. A neotype is proposed for Thyridium salicis. A determinative key to species is given. Asexual morphs of fenestelloid fungi are phoma-like and do not differ from those of other representatives of the Cucurbitariaceae. The phylogenetic structure of the fenestelloid clades is complex and can only be resolved at the species level by protein-coding genes, such as rpb2, tef1 and tub2. All fungal species studied here occur, as far as has been possible to determine, on members of Diaporthales, most frequently on asexual and sexual morphs of Cytospora.

After the original publication of Fenestella by Tulasne & Tulasne (1863), who recognised three species in the genus including F. princeps, a synonym of F. fenestrata (see Jaklitsch et al. 2018), 52 additional species names were created in the genus. Eleven names including Fenestella bipapillata (Jaklitsch & Barr 1997) and Fenestella frit (see Jaklitsch et al. 2018) have been removed to other genera or they, among others, are not interpretable, because no type material exists (for more data see notes to species and Discussion). Barr (1990) recognised eight species in Fenestella occurring in North America, which she keyed out and described morphologically. She also gave a detailed diagnosis of the genus Fenestella recognising its fungicolous habit. However, she subsumed American fungi under European Fenestella names without having seen type material of most of them. As a result, several of her taxonomic interpretations and conclusions are either erratic or too broad.
A definition of what fenestelloid fungi are is difficult, particularly when compared to other members of the Cucurbitariaceae. The main character apart from a more marked tendency to form valsoid groups or pseudostromatic pustules, are the ascospores, whose septa are variable in number and often difficult to count due to incompleteness, dense insertion and apparent oblique or shifted superposition in sectional view. This character is  Table 1 Isolates and accession numbers used in the phylogenetic analyses. Isolates/sequences in bold were isolated / sequenced in the present study.  shared with the morphologically rather pleomassariaceous genus Seltsamia (Jaklitsch et al. 2018), whose ascospores have an indefinitely swelling, bipartite sheath. A similar situation is found in Fenestella as shown below for F. granatensis, where the ascospore sheath swells however in a limited manner. Other unrelated, non-lichenised pyrenocarpous fungi on or in wood and bark having ascospores with many transverse and longitudinal eusepta in more or less cylindrical, fissitunicate asci are Aigialus, differing from fenestelloid fungi, e.g., in different ecology, as ascomata are immersed in submerged wood of mangroves in marine environments (Kohlmeyer & Schatz 1985), Decaisnella and Karstenula in the very wide concept of Barr (1990), which, e.g., lack a subiculum and are not associated with Diaporthales, or Ostreichnion, which produces conchate, superficial ascomata on wood (Boehm et al. 2009).
Here we take a detailed look into the taxonomy and phylogenetic structure of fenestelloid fungi described from Europe on woody hosts, from which fresh material was available for study. These fungi include several species originally described in Fenestella, Cucurbitaria or Thyridium, and cluster in three clades representing the three genera Fenestella, Parafenestella and Synfenestella.

Isolates and specimens
All isolates used in this study originated from ascospores or conidia (where noted) of fresh specimens. Numbers of 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, Utrecht, The Netherlands (CBS culture collection). Details of the specimens used for morphological investigations are listed in the Taxonomy section under the respective descriptions. Herbarium acronyms are according to Thiers (2018). Freshly collected specimens have been deposited in the Fungarium of the Department of Botany and Biodiversity Research, University of Vienna (WU).

Culture preparation and phenotype analysis
Cultures were prepared and maintained as described previously (Jaklitsch 2009) except that CMD (CMA; Sigma, St Louis, Missouri; supplemented with 2 % (w/v) D(+)-glucose-monohydrate) or 2 % malt extract agar (MEA; 2 % w/v malt extract, 2 % w/v agar-agar; Merck, Darmstadt, Germany) was used as the isolation medium. Cultures used for the study of asexual morph micro-morphology were grown on CMD or 2 % MEA at 22 ± 3 °C in darkness. 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. Some images obtained by using the Nikon interference contrast may be slightly too dark. 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 mean plus and minus the standard deviation of a number of measurements given in parentheses.

Analysis of sequence data
For the phylogenetic analyses, a combined matrix of nSSU-ITS-LSU rDNA, rpb2, tef1 and tub2 sequences was produced. The newly generated sequences were complemented with GenBank sequences of Cucurbitariaceae from Jaklitsch et al. (2018), and sequences of six taxa of Pyrenochaetopsis (Pyrenochaetopsi daceae) were added as outgroup according the results of the phylogenetic analyses of Jaklitsch et al. (2018). All alignments were produced with the server version of MAFFT (www.ebi. ac.uk/Tools/mafft), checked and refined using BioEdit v. 7.2.6 (Hall 1999). Large insertions sometimes present in the SSU and at the terminal 3' end of the SSU of the partial SSU-ITS-LSU fragment were removed from the alignments.
Maximum parsimony (MP) analysis was performed using a parsimony ratchet approach. For this, a nexus file was prepared using PRAP v. 2.0b3 (Müller 2004), implementing 1 000 ratchet replicates with 25 % of randomly chosen positions upweighted to 2, which was then run with PAUP v. 4.0a164 (Swofford 2002). The resulting best trees were then loaded in PAUP and subjected to heuristic search with TBR branch swapping (MUL-TREES option in effect, steepest descent option not in effect). Bootstrap analysis with 1 000 replicates was performed using 5 rounds of replicates of heuristic search with random addition of sequences and subsequent TBR branch swapping (MULTREES option in effect, steepest descent option not in effect) during each bootstrap replicate. In all MP analyses molecular characters were unordered and given equal weight; analyses were performed with gaps treated as missing data; the COLLAPSE   (SSU-ITS-LSU , rpb2, tef1, tub2) of Cucurbitariaceae and Pyrenochaetopsidaceae, with the latter selected as outgroup. MP and ML bootstrap support above 50 % are given at the first and second position, respectively, above or below the branches. Strains formatted in bold were isolated and sequenced in the current study; ex-type strains are indicated by a superscript T. Nodes collapsed in the strict consensus of the 33 MP trees are marked by an asterisk (*).
command was set to minbrlen. Maximum likelihood (ML) analyses were performed with RAxML (Stamatakis 2006) as implemented in raxmlGUI 1.5 (Silvestro & Michalak 2012), using the ML + rapid bootstrap setting and the GTRGAMMA substitution model with 1 000 bootstrap replicates. The matrix was partitioned for the individual gene regions, and substitution model parameters were calculated separately for them. For evaluation and discussion of bootstrap support, values below 70 % were considered low, between 70 and 90 % medium/ moderate and above 90 % high.

Phylogenetic analyses
Of the 5 707 nucleotide characters of the combined matrix, 1 266 are parsimony informative (283 of ITS-LSU, 5 of SSU, 423 of rpb2, 302 of tef1 and 253 of tub2). Maximum parsimony analyses revealed 33 MP trees 6 241 steps long, one of which is shown as Fig. 1. Topologies of the MP trees were identical except for one backbone node each within the Fenestella and Parafenestella clades, and two backbone nodes within the Neocucurbitaria clade (marked by asterisks in Fig. 1).
Like in the previous phylogenetic analyses of Jaklitsch et al. (2018), many of the deeper nodes within Cucurbitariaceae were unsupported or received only low support, while the genera Cucurbitaria and Neocucurbitaria were highly supported. The fenestelloid clade received maximum support and contained three highly supported subclades here recognised as three distinct genera: Fenestella, Parafenestella (both with maximum support in MP and ML analyses) and Synfenestella (99 % MP and ML bootstrap support). The sister group relationship of Fenestella and Parafenestella received only low support (54 % MP and 51 % ML), while most backbone nodes within the three genera received high to maximum support ( Fig. 1). Notes -See Jaklitsch et al. (2018) for description and typification of the genus and its type species. In that work it was stated that all materials of F. fenestrata available for study were more or less overmature, which made identification of the fungal host difficult. The black encasement of ascomata was interpreted as belonging to a Diaporthe sp. Considering that all other species of the genus occur on Cytospora spp. (see below), it appears probable that the stromatic encasement belonged to a Cytospora sp. having a Leucostoma sexual morph.
Habitat -On Cytospora sp. (sexual and asexual morphs) on branches and twigs of Crataegus monogyna.
Distribution -Europe (France), only known from the type locality; possibly occurring also in North America.
Notes -Fenestella gardiennetii may have travelled on its host from North America and may thus occur also there. It is closely related to F. granatensis, which occurs on Acer gra natense and differs from the former by more distinctly clavate ascospores that are surrounded by a mucous sheath.
Distribution -Europe (Spain), only known from the type locality.
Habitat -On Cytospora spp. (sexual and asexual morphs) on various deciduous trees and shrubs, particularly common on Corylus avellana.

Fenestella subsymmetrica
Culture characteristics and asexual morph in culture -Ascospores germinating simultaneously from many cells. Colony radius on CMD at 22 °C in the dark 6 mm after 1 wk, c. 20 mm after 3 -4 wk; colony white, centre turning black by pycnidia after 4 d, soon entire colony turning grey, brownish grey to olivaceous, margin often hyaline to white, covered by a white to pale grey mat of aerial hyphae; odour indistinct; no diffusing pigment formed. Pycnidia 120-240 µm diam, more or less globose, first hyaline to greenish, turning green to black, numerous, often concentrically and very densely arranged, spreading over entire colony or remaining in the centre; often covered by mats of aerial hyphae; conidia amassing in whitish to greenish turbid drops. Phialides 4.5-8 × 2-4 µm, lageniform to subglobose with a long neck. Conidia (3.2-) 3.5-4.2(-4.5 Habitat -On Cytospora spp. (sexual and asexual morphs) on various deciduous trees and shrubs.
Distribution -Europe, possibly North America; locally common in winter; sometimes co-occurring with F. media. Other  Notes -Fenestella subsymmetrica is hardly distinguishable from F. media by morphology alone. Ascospores of F. subsym metrica often tend to appear broader, with more distinct septa and a more median primary septum. However, individual specimens pose serious problems in morphological identification. For example, culture C286x derived from distinctly asymmetric asco spores of WU 36977 yielded ITS and LSU sequences, which are identical with those derived from symmetric ascopores. In cultures on CMD no pigment is formed. Mature asci are very unstable in water, therefore they were mostly measured and illustrated in 3 % KOH. Jaklitsch & Voglmayr, sp. nov. -MycoBank MB829746;Fig. 8 Etymology.
Habitat -On Cytospora spp. (both morphs; sexual morph of the Leucostoma type) on Viburnum spp.
Distribution -Europe. Notes -Fenestella viburni is one of three cryptic species, morphologically most closely related to F. subsymmetrica, but difficult to differentiate. In individual specimens ascospores tend to be distinctly pointed terminally. Formation of pseudostromatic pustules is less pronounced and asci are more stable in water than with F. media and F. subsymmetrica. Pustules are difficult to assess, as they are usually produced basically by its Leucostoma host. Older, not sequenced specimens from Viburnum spp. are added tentatively to the list above.
Habitat -On dead partly corticated twigs of Sorbus aria.
Habitat -Associated with perithecial fungi on Rosa canina and possibly Crataegus monogyna.
Habitat -On Diaporthe decedens on Corylus avellana. Distribution -Central Europe (Germany), only known from the type locality.
Habitat -On both morphs of a Cytospora (Valsa) sp. on Salix cinerea.
Distribution -Central Europe (Austria), only known from the type locality.
Culture characteristics and asexual morph in culture -Co lony radius on CMD at 22 °C in the dark 6 -7 mm after 1 wk, 15 -16 mm after 2 wk; colony first hyaline to whitish, thick, dense, turning olivaceous from the centre, becoming grey to greyish brown due to a dense whitish to greyish villose mat of aerial hyphae; reverse dark grey to black; odour indistinct. No asexual morph detected.
Distribution -Europe, only known from the type locality in Ukraine.
Habitat -Associated with Cytospora spp. and other perithecial fungi on various species of Rosaceae, recorded from Crataegus, Prunus, Pyracantha, Pyrus, Rosa and Sorbus aria.
Distribution -Central Europe (Austria).  *prunorum. However, we have not seen type material and a varietal name is not binding. Parafenestella rosacearum is a complex species. In spite of splitting into two or three groups in multigene analyses (see Fig. 1), we recognise a single species, because there are no morphological differences among those groups and particularly due to the following observations: tef1 sequences of C203 and C283 are identical and tef1 of C309 is nearly identical with them, whereas rpb2 sequences of C203, C315, FM1 and FP11 are virtually identical, while those of C269 and C283, which are identical, differ from the first group by c. 20 nucleotides. This finding was verified by repetition of DNA amplifications and sequencing. Species of Parafenestella on Rosaceae, particularly on Rosa spp., are difficult to distinguish morphologically; ascospores of P. faberi are characteristic due to their often truncate hyaline terminus of the end cells, and P. austriaca differs from P. rosacearum by a rather invariable ascospore shape. In both of the latter species ascospore end cells are rounded, slightly projecting and may be only diffusely paler to hyaline.
Distribution -Europe (Austria, Germany, UK), Russia, etc. Notes -Judging on the specimens we have seen, P. salicis is the most common species of the genus on Salix spp. and occurs often together with other Parafenestella species. Counting of ascospore septa in sectional view is particularly difficult in this species, due to conspicuous superposition; they differ from those of P. salicum and P. parasalicum in less and more widely spaced septa and from the latter also by size. Ascospores are often inequilateral and often slightly constricted at other than the primary septum, traits not seen with P. pseudosalicis. Like with other species, mature ascospores are often dark reddish brown in herbarium material.
Habitat -On thin branches of Salix alba, presumably on both morphs of a Cytospora (Valsa) sp.
Other material examined. AustRiA, Vienna, 22nd district, Lobau, Panozzalacke, grid square 7865/1, on thin branches of Salix alba, soc. effete Cytospora sp., Keissleriella holmiorum, 8 Feb. 1997, W. Jaklitsch W.J. 1015 Notes -Parafenestella salicum differs from P. salicis by larger asci, larger, particularly broader ascospores with a larger number of septa. While phylogenetically closely related to P. parasalicum, P. salicum differs from that species by smaller ascospores, although some old aberrant ascospores may approach those of P. parasalicum in size. The fungal host of P. salicum is not unequivocally clear. As is usual in the Cucur bitariaceae, ascospores are more reddish brown in herbarium material.
Habitat -On or in conidiomata of Melanconis alni on Alnus glutinosa.
Distribution -Central Europe (Austria), only known from the type locality in Vienna.
Notes -This species is well characterised by its narrow ascospores. Ascomata 300 -1000 µm diam, globose, subglobose or pyriform, ostiolate, dark brown to black, immersed below bark epidermis, scattered, forming inconspicuous valsoid groups or conspicuous pseudostromatic pustules on pseudostromata or conidiomata of Diaporthales, surrounded and connected by thick-walled subicular hyphae, the latter sometimes short and subsetose in the ostiolar region. Peridium c. 15 -130 µm thick, thicker and whitish inside in the apical region, pseudoparenchymatous, dark brown and thick-walled outside and pale to hyaline and thin-walled inside. Hamathecium consisting of numerous 1-3.5 µm wide, branched and anastomosing paraphyses with free ends in a matrix. Asci cylindrical to oblong, bitunicate, fissitunicate, with an ocular chamber, a short or long undulating stipe and simple or knob-like base, containing 4 -8 ascospores in (obliquely) uniseriate arrangement. Ascospores ellipsoid, oblong to fusoid, sometimes subglobose, symmetric to slightly curved, with upper part often wider, initially 2-celled, hyaline and surrounded by a swelling sheath, developing additional transverse and longitudinal septa, turning yellow to golden brown (when fresh), finally dark brown with concolorous rounded ends, ends sometimes paler or apiculate upon germination, usually strongly constricted at the primary septum; in 3 % KOH surface smooth; turning dark olivaceous to blackish brown.

Synfenestella
Culture characteristics and asexual morph in culture -Co lony on CMD at 22 °C in the dark typically producing a diffusing yellow pigment. Pycnidia more or less globose, green to black, densely aggregating and fusing. Phialides lageniform, ampulliform, subglobose or subconical, sessile or produced on short simple conidiophores. Conidia cylindrical, oblong to allantoid, less commonly ellipsoid, 1-celled, hyaline, with 2 drops, smooth; pro duced on phialides and pegs.
Habitat -On or in pseudostromata or conidiomata of Dia porthales or in loose association with them on Rosaceae.
Habitat -On Cytospora sp. (both morphs; sexual morph of the Leucostoma type) on Pyrus communis.
Distribution -Europe (Austria), only known from the type locality.
Habitat -On both morphs of Cytospora sp. of the Leuco stoma type, in pseudostromata on effete ascomata of Diaporthe impulsa or in loose association with it on Sorbus aucuparia.
Distribution -Europe, North America fide Barr (1990;under Fenestella subcaespitosa) and Mirza (1968;under Cucurbitaria sorbi Notes -In H two syntypes of Cucurbitaria sorbi are extant; No. 3687 (Finland, Vaasa, on Sorbus, 12 Aug. 1867, P.A. Karsten) contains mostly Dothiora pyrenophora, a fungus with 3-septate ascospores (possibly Nigrograna sp.), an immature nectriaceous fungus and only little and old S. sorbi (ascospore measurements given on the label: 25-30 × 12-14). In contrast, No. 3686 contains good material and is thus selected as lectotype. The latter was also examined by M.E. Barr according to the annotation slip. Synfenestella sorbi has been found in direct association with both Cytospora sp. (Leucostoma morph) and Diaporthe impulsa. Also other ascomycetes are associated with the fungus, thus parasitism on other fungi may be possible. Fenestella bavarica is clearly a synonym of S. sorbi. In its holotype, S. sorbi grew on Cytospora (Leucostoma morph), has typical long-pedicellate asci with often aberrantly developed ascospores, many aborted or distorted or even globose. At the pustule surface cylindrical Leucostoma ostioles peek through brown compacted subiculum of the Synfenestella; also a Tym panis is present on thin twigs.
Synfenestella sorbi is extremely variable in appearance, ranging from solitary ascomata (on Diaporthe) to large conspicuous pseudostromata. Hyphal appendages sensu Barr (1990;sub Fenestella subcaespitosa) are subicular hyphae and occur variably in the entire family. In S. sorbi they are often but not always differentiated from other hyphae around the ascomatal apices by having a nearly setose appearance, darker colour and slightly thicker walls. The long-pedicellate asci were already described by Mirza (1968;under Cucurbitaria sorbi). Cucurbi taria subcaespitosa from S. aria is clearly a different species and here combined in Neocucurbitaria.
There are some sequence differences between isolates derived from specimens containing the fungal host Cytospora (Leuco stoma) and those growing on Diaporthe impulsa. However, as these differences are not convincing and the specimens are morphologically indistinguishable, we do not recognise two separate species. In the lectotype H 3686 S. sorbi is associated with Cytospora, therefore we epitypify it with material containing the same fungal host.

KEY TO FENESTELLOID SPECIES
Fungal hosts are only given in the key when different from or additional to Cytospora spp. Parafenestella ostryae from Ostrya carpinifolia (Wanasinghe et al. 2017b) is not included, as we have not seen material of this species. Some species, particularly Fenestella media vs F. subsymmetrica, Parafenestella salicis vs P. pseudosalicis, or P. austriaca vs P. rosacearum cannot be safely distinguished morphologically.

Other species of Fenestella (and Cucurbitaria)
See notes under the species described in the Taxonomy section above and in Jaklitsch et al. (2018), where many names in Cucurbitaria are commented on. In this work we studied particularly European species of Fenestella s.lat. with available fresh material. We did not study, e.g., F. betulae nor F. parvula, because no fresh material from Betula was available. No type material of Fenestella cydoniae (basionym Pleospora cydoniae) could be located. The protologue may suggest affiliation with F. media or a similar species. Fenestella microspora was described from Corylus in France with ascospores 10 -11 × 6 -7 µm, thus it is apparently not a member of Cucurbitariaceae.
Fenestella phaeospora is not a member of the Cucurbitaria ceae phylogenetically (unpublished results). Fenestella pru nastri from Prunus spinosa in Luxemburg was described with asco spores having 3 -4 transverse and 1 longitudinal septa, 20 -22 × 8 -9 µm, pale yellow, as similar to F. lycii, i.e., not a species of Fenestella, but possibly rather belonging to Camaro sporidiellaceae or Coniothyriaceae. Species with muriform ascospores on Lycium spp. (Cucurbitaria varians, Fenestella lycii) are referable to Camarosporium. Fenestella subvestita from Alnus glutinosa in Lyngby, Denmark, was described with ascospores 15 -19 × 7.5 -8.5 µm having 3 transverse and 1 longitudinal septa, i.e., not a species of Cucurbitariaceae. The genera Teichospora and Thyridium may house additional cucurbitariaceous species, but the task to examine all available type materials of these genera is beyond this work.

Molecular phylogeny
As shown in Fig.1, there is significant infraspecific variation in taxa of the genera Parafenestella and Synfenestella. On the other hand, in those Fenestella species, where several isolates were available, F. media, F. subsymmetrica and F. viburni, there is no or very little variation in DNA sequences of all markers. This is particularly interesting, as most species of all genera grow on Cytospora spp., i.e., they share the same ecology. The reason for the infraspecific variability is unclear but may indicate that these lineages are in the process of ongoing speciation, probably triggered by host specificity on different Cytospora hosts. This, however, is difficult to assess, as the identification of the effete Cytospora hosts is usually impossible when ascomata are produced.

Ecology, hosts and development
Fenestelloid fungi produce their ascomata and conidiomata on or in tight association with effete ascomata and conidiomata of mostly if not always Diaporthales, in temperate zones mostly in the cold season (late autumn to early spring). This means that climatic conditions have first to be favourable enough to promote development of the fungal host and secondly, the following period must be long and favourable enough to support ascomatal development and maturation. In the cold season such periods are often short and irregular, as temperatures may variably decrease below 0 °C often terminating fungal development, and this may influence development of ascospores dramatically. Moisture is usually not a criterion limiting fungal development in the cold season, but in recent years and decades, even this has changed in many temperate regions, where little precipitation occurs even in winter. At high elevations (montane to alpine levels) impact on development by climatic variation is even more drastic. One example is Synfenestella sorbi, whose plant host Sorbus aucuparia occurs from lowlands to subalpine levels. As a result, ascospores often vary tremendously with regard to size and shape, even within a single ascus, in this but also other species. Often only one to four ascospores in an ascus are mature and others are aborted. The resulting drastic infraspecific variation of ascospore characters makes species recognition and identification using morphological traits alone extremely difficult.
Also, size and development of ascomatal groups or pseudostromata vary considerably, as they depend on the size and development of the Cytospora (and other) hosts, thickness of twigs or branches, and again on (micro-)climatic conditions on twigs and branches determining the magnitude of infection of the present host pseudostromata or conidiomata. As the latter are mostly effete when colonised by the Fenestella, we have no information on host identity on the species level, which is difficult also due to the unsettled taxonomy of Cytospora and the fact that several Cytospora species may occur on a single host plant. Specificity toward the fungal host is unknown, as, e.g., Fenestella media occurs on several plant hosts, which may be substrates of different Cytospora spp. Thus, F. media either has a wide specificity regarding its Cytospora host species or alternatively, it grows on a single Cytospora sp., which occurs on many different plants. As species of Cytospora are numerous and common, we expect that many more species of fenestelloid fungi will be described in future, at least from temperate zones worldwide.

Morphological characters
Ascomata of fenestelloid fungi are always immersed in to erumpent from bark and have a marked tendency to form groups, besides forming solitary ascomata in a colony, too. Numbers of ascomata in a group is often less than 10 in Parafenestella, but may be much higher in Fenestella and Synfenestella. A few species of the latter genera may produce conspicuous pseudostromatic pustules. Morphological variation in Synfenestella is remarkable, as fructifications of S. pyri are very inconspicuous and consist of only few ascomata in a group, and asci are mostly short-stipitate, while in contrast S. sorbi often forms large conspicuous pseudostromata and asci are long-stipitate. However, the swelling sheath in young ascospores within asci unites the two species. It should be borne in mind that elongated ascus stipes and partly biseriate arrangement of ascospores in microscopic mounts may also be consequences of exerting pressure on the cover slip in order to free asci from the hymenium. The peridium of Fenestella was characterised as 3-layered by Barr (1990;sub Fenestellaceae). However, the innermost narrow layer of brownish compressed and elongated cells is not always present and the outer two layers (dark brown, of thick-walled cells outside, followed by thinner-walled and paler brown to hyaline cells) are, after consideration of all species, rather a single layer with internal variation. The outermost region can be narrow and then more easily be interpreted as a layer of its own, particularly when the transition to lighter colour and thin walls, which basically is always gradual, takes place at a short distance. This is especially obvious in F. fenestrata and F. parafenestrata and in the three sibling species F. media, F. subsymmetrica and F. viburni. The hamathecium has been identified as apically free paraphyses present among immature asci in several species, thus this may be a character common to the entire family Cucurbitariaceae (see also Jaklitsch et al. 2018). Morphology and illustration of several species in this work was done using dried material; thus images of fresh vital ascospores are missing. However, ascospores of fenestelloid fungi are mostly yellow-brown or golden-brown when fresh, but turn dark brown upon drying and may often be dark reddish brown in herbarium material. They turn greenish olivaceous when immature to blackish brown when mature in 3 % KOH. Verruculose ornamentation has been seen in many species. This might however be an effect of drying of the perispore.

Asexual morphs in the Cucurbitariaceae
As pointed out by Jaklitsch et al. (2018), asexual morphs of the Cucurbitariaceae do not offer sufficient traits for reliable distinction at the species or generic level. As we have seen here, asexual morphs of all fenestelloid fungi are morphologically similar and may be termed phoma-like.