Linosporopsis, a new leaf-inhabiting scolecosporous genus in Xylariaceae

Based on molecular phylogenetic and morphological evidence, the new genus Linosporopsis (Xylariales) is established for several species previously classified within Linospora (Diaporthales). Fresh collections of Linospora ischnotheca from dead overwintered leaves of Fagus sylvatica and of L. ochracea from dead overwintered leaves of Malus domestica, Pyrus communis, and Sorbus intermedia were isolated in pure culture, and molecular phylogenetic analyses of a multi-locus matrix of partial nuITS-LSU rDNA, RPB2 and TUB2 sequences as well as morphological investigations revealed that both species are unrelated to the diaporthalean genus Linospora, but belong to Xylariaceae sensu stricto. The new combinations Linosporopsis ischnotheca and L. ochracea are proposed, the species are described and illustrated, and their basionyms lecto- and epitypified. Linospora faginea is synonymized with L. ischnotheca. Based on similar morphology and ecology, Linospora carpini and Linospora magnagutiana from dead leaves of Carpinus betulus and Sorbus torminalis, respectively, are also combined in Linosporopsis. The four accepted species of Linosporopsis are illustrated, a key to species is provided and their ecology is discussed.


Introduction
The genus Linospora was established by Fuckel (1870) for five species growing on dead leaves of Salicaeae. He did not designate a generic type, but Clements and Shear (1931) selected Linospora capreae, which grows on Salix caprea, as lectotype. The genus is characterized by long, filiform ascospores arranged in a single fascicle within the ascus, and by reduced black stromata embedded in dead leaf tissue containing usually one (in L. ceuthocarpa up to six) perithecia with laterally inserted ostioles. The black stromata appear in spring and are noticeable as black dots of ca. 0.5-1 mm diam on both sides of the dead, usually bleached leaves. The characteristics of ascomata and asci are clearly diaporthalean, and its classification within Gnomoniaceae (Monod 1983;Barr 1990) has also been corroborated by molecular phylogenetic analyses . So far, the about eight accepted species of Linospora inhabit leaves of Salix or Populus spp. (Salicaceae), but morphological evidence suggests the presence of additional undescribed species on Salicaceae (Monod 1983).
Soon after its description, additional species with long filiform ascospores and black ascomata or stromata embedded in leaf tissues were added to Linospora. However, critical morphological re-investigations by Monod (1983) revealed that many of these are not diaporthalean and therefore unrelated to the generic type. Five of them, L. carpini from leaves of Carpinus betulus; L. faginea, and L. ischnotheca from leaves of Fagus sylvatica; L. magnagutiana from leaves of Sorbus torminalis and L. ochracea from leaves of various other rosaceous hosts from subtribe Pyrinae, were considered to be synonymous and to belong to the genus Ophiodothella Section Editor: Marc Stadler Isolates were prepared from ascospores as described in Jaklitsch (2009) and grown on MEA or on 2% corn meal agar plus 2% w/v dextrose (CMD). Growth of liquid culture and extraction of genomic DNA was performed as reported previously (Voglmayr and Jaklitsch 2011; using the DNeasy Plant Mini Kit (QIAgen GmbH, Hilden, Germany).

Data analysis
The newly generated sequences were aligned to the sequence alignments of , and GenBank sequences of four taxa of Diaporthales (Gnomonia gnomon, Juglanconis juglandina, Linospora capreae, and Melanconis stilbostoma) were added as the outgroup. Some taxa included in the matrix of  which contained poor or incomplete sequence data and which were not relevant for this study were removed from the matrices. The GenBank accession numbers of sequences used in these analyses are given in Table 1.
Sequence alignments for phylogenetic analyses 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 ITS-LSU rDNA, RPB2 and TUB2 matrices were combined for subsequent phylogenetic analyses. After exclusion of ambiguously aligned regions and long gaps, the final combined data matrix contained 4718 characters (622 nucleotides of ITS, 1355 nucleotides of LSU, 1169 nucleotides of RPB2 and 1572 nucleotides of TUB2). Familial   classification of Xylariaceae and pylogenetically related families follows  and Wendt et al. (2018). 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 5 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. For evaluation and discussion of bootstrap support, values below 70% were considered low, between 70 and 90% medium/moderate and above 90% high.

Molecular phylogeny
The combined multilocus matrix used for phylogenetic analyses comprised 4718 characters, of which 2129 were parsimony informative (360 from ITS, 273 from LSU, 658 from RPB2 and 838 from TUB2). Figure 1 shows a simplified phylogram of the best ML tree (lnL = − 131,936.737) obtained by RAxML. Maximum parsimony analyses revealed four MP trees 31,692 steps long, which were identical except for slightly different positions of Daldinia andina and Stilbohypoxylon quisquiliarum (not shown). The backbone of the MP trees was similar to the ML tree, except for a few minor topological differences of uns u p p o r t e d n o d e s w i t h i n t h e B a r r m a e l i a c e a e , Graphostromataceae, Hypoxylaceae and Xylariaceae (not shown). Linospora ischnotheca and L. ochracea were revealed as closely related but distinct species with maximum support (Fig. 1). They were placed remotely from Linospora capreae (Diaporthales) in a basal position within Xylariaceae sensu Mycelium in dead overwintered leaves, strongly bleaching the host tissue. Pseudostromata immersed in dead leaves, reduced, forming a distinct black clypeus-like structure on both sides of the leaf above and below the single perithecium, composed of dark brown, septate hyphae in dead host epidermis cells and forming a textura epidermoidea-intricata. Ascomata perithecial, scattered, solitary, immersed, (sub)globose, with a central apical papilla. Peridium thin, composed of hyaline, thin-walled, pseudoparenchymatous to prosenchymatous cells forming a textura angularis. Hamathecium of unbranched, thin-walled, hyaline, septate, apically tapering paraphyses. Asci unitunicate, long-cylindrical, with a short stipe, with an indistinct, inamyloid or slightly amyloid apical apparatus, containing 8 ascospores in a single fascicle. Ascospores long-filiform, hyaline, smooth, without visible septa, without sheath or appendages. Asexual morph unknown.
No Notes: Although no DNA data are yet available, morphology of ascomata, asci and ascospores leave no doubt that the species belongs to Linosporopsis, and considering the high host specificity of the genus, we recognize L. carpini as a distinct species. Apart from the type collection, this species is to our knowledge only known from an additional collection in northern Italy (Veneto, near Conegliano), which was collected in the same year as the type (Saccardo 1877). On the herbarium label of the type collection, it was stated to be common in the forests around Rastatt; however, we are not aware of any recent collections. The type collection has been edited and distributed in numerous copies in Rabenhorst, Fungi Eur. Exs. 2132, but we have investigated in detail only the copy deposited in W, that consists of a single leaf with a few perithecia. To save material, no sections were performed, and only a microscope preparation for documentation and measurements of asci, ascospores, paraphyses and clypeus hyphae was done. Our measurements revealed distinctly longer asci and ascospores than reported in the original description (118-165 μm vs. 70-80 μm in Rabenhorst 1876), which therefore is within the range of the other accepted Linosporopsis species.
Colonies on CMD and MEA white; aerial hyphae abundant. No asexual morph observed.
Habitat and host range: Dead overwintered leaves of Fagus sylvatica and F. orientalis; rarely also on Quercus sp.
Distribution: Europe; known from France, Germany, Italy, Spain, and Switzerland.
Typification: France, without place, date and collector, on dead leaves of Fagus sylvatica, in Desmazières, Pl. Crypt. N. France, Ed. 1, no. 2098 (K(M)  Notes: DNA sequence data and morphology place the species within Xylariaceae, as closest relative of L. ochracea. Desmazières (1851) first included specimens from leaves of Fagus sylvatica in his Sphaeria ochracea, but soon thereafter, he described them as a distinct species, S. ischnotheca (Desmazières 1852). In the protologue, he mentioned that the type collection contained only immature asci without spores, which was confirmed for all syntypes investigated in our study. The type collection was edited and distributed in two sets as Pl. Crypt. N. France, Ed. 1, nos. 1798 and 2098, which is also mentioned in the protologue. Neither locality nor collector are mentioned on the herbarium labels and in the original description of the species, and no original notes of Desmazières are attached to the two copies present in PC. However, the herbarium labels of a specimen in M, probably also a syntype, indicates that it was collected by M.R. Roberge in Caen, i.e. the same place and collector as the type of L. ochracea (see below), which appears plausible considering that material of Fagus was mentioned in the original description of L. ochracea. As the type collection of Sphaeria ischnotheca is immature, we here designate a recent mature collection, for which a culture and DNA sequences are available, as epitype to stabilize the species nomenclature.
Linospora faginea, which was also described from dead leaves of Fagus sylvatica, is obviously a synonym of L. ischnotheca; the protologue in Saccardo (1878) fully matches our material. As Saccardo material of PAD is not sent out on loan, we have not been able to investigate the type in detail, but the illustrations of the specimen and label kindly provided by the Erbario dell'Università di Padua show that it agrees with L. ischnotheca (see Fig. 3d).
Notes: Due to the lack of fresh specimens, no cultures and sequence data are available for L. magnagutiana, but its morphology clearly places it in Linosporopsis. Only few historic records from northern Italy, all collected in the 1870ies, are known. We have not been able to investigate the type from PAD, which is not sent out on loan, but two additional authentic collections from the same area were available for study. As the historic material is very brittle, no useable section of the peridium could be prepared. The rosaceous host, Sorbus torminalis, and similar morphology indicates that L. magnagutiana may be conspecific with L. ochracea. However, in one locality (Bayerisches Landesarboretum "Weltwald"), where leaves of Pyrus domestica and Sorbus latifolia were heavily infected by L. ochracea, no Linosporopsis could be found on leaves of directly close-by Sorbus torminalis, indicating that they are distinct. In addition, the asci and ascospores of L. magnagutiana are slightly shorter than those of L. ochracea ( (79- Notes: DNA sequence data and morphology place the species within Xylariaceae, as closest relative of L. ischnotheca. It was first described as Sphaeria ochracea by Desmazières (1851), but the name is illegitimate as it is a younger homonym of Sphaeria ochracea Pers. (1801). Therefore, Linospora ochracea Sacc., originally established as a new combination of Sphaeria ochracea Desm., is to be treated as a replacement name and represents the valid basionym.
In the protologue, Desmazières (1851) listed leaves of Crataegus, Cydonia, Mespilus, Sorbus and also Fagus as hosts; however, no collection or specimen data were given. For the specimens on Fagus, Desmazières (1852) subsequently described a distinct species, Sphaeria ischnotheca (see above). As concluded from the original material of Desmazières in PC and K, and from his notes attached to the specimen PC 0706581, the species was based on material collected by M.R. Roberge in Hérouville-Saint-Clair near Caen in May 1850, which Desmazières edited in his Pl. Crypt. N. France, Ed. 1, no. 2099. This exsiccatum contains material from Crataegus monogyna and Sorbus latifolia. From the same locality, Desmazières also distributed material from Pyrus argentea (as Pl. Crypt. N. France, Ed. 1, no. 2100 and Pl. Crypt. N. France, Ed. 2, Ser. 1, no. 1800), under the unpublished name Sphaeria ochracea f. pyrina, which, however, does not qualify for the type, as this host is not listed in the protologue; in addition, it was collected one year later (Apr. 1851) than the type, which may be a reason why this host was not cited in the protologue.
Unlike all other accessions of L. ochracea investigated by us, which had an indistinct, inamyloid apical apparatus, the Swiss collection WU 40028 from Malus domestica showed a tiny, wedge-shaped, slightly amyloid apical apparatus after KOH pre-treatment (see Fig. 5t). However, the sequences obtained from this accession fully matched the other collections, indicating a variable iodine reaction that probably depends on the maturity and preservation of the specimen.

Discussion
The results of our molecular phylogenetic investigations confirmed the conclusions of Monod (1983) that the species treated here are not congeneric with Linospora and do not belong to Diaporthales. However, while he assumed that they belong to Ophiodothella, currently classified within Phyllachoraceae (Phyllachorales), our phylogenetic analysis placed them in a basal clade of Xylariaceae sensu stricto (Xylariales). Based on the presence of an amyloid apical ascus ring, conidia resembling Diatrypaceae and a single nuSSU rNDA sequence, Hanlin et al. (2002) assumed xylarialean affinities of Ophiodothella; however, these conclusions were based on non-type species and need to be verified by re-investigation of the generic type. No type material of the generic type, O. atromaculans (Henn.) Höhn., is extant in B where the material of Hennings is kept (R. Lücking, personal communication). However, even if xylarialean, the following features do not support that Ophiodothella is congeneric with the species treated here: an obligate parasitic lifestyle in living leaves, a tropical to subtropical distribution almost exclusively in the New World, formation of pycnidial or acervular conidiomata, lack of distinct bleaching of the substrate and morphological differences of the ascomata (Hanlin et al. 1992(Hanlin et al. , 2002(Hanlin et al. , 2018. Particularly the generic type, O. atromaculans, deviates significantly from our species by an extended effuse, black stromatic crust (Hennings 1904;Hanlin et al. 1992). Additional genera with solitary clypeate ascomata and filiform ascospores that were previously attributed to Xylariales include Linocarpon and Neolinocarpon; however, these have been shown to belong to Chaetosphaeriales by sequence data (Konta et al. 2017). As no suitable described genus is available within Xylariaceae, we establish the new genus Linosporopsis for them.
Sister group relationship of Linosporopsis to the Clypeosphaeria mamillana-Anthostomelloides krabiensis clade is highly supported in the ML analyses, but receives only low support in the MP analyses. Linosporopsis is similar to the latter species in solitary ascomata of similar size that are
Ecologically, there is evidence that Linosporopsis occupies a niche as a leaf endophyte, and there is so far no indication of parasitism. Observations in Austrian and Swiss sites with abundant sporulation of Linosporopsis ochracea on dead overwintered Pyrus and Malus leaves revealed no obvious symptoms on living Pyrus and Malus leaves during the following summer. Evidently, the life cycle of Linosporopsis is connected with that of their hosts, as the short-lived ascospores are only produced briefly after their hosts unfold their new leaves in spring. These young leaves are then infected by the ascospores to complete the life cycle, with the living leaf tissue remaining asymptomatic during the growing season. After leaf abscission, the mycelium continues growth on the fallen leaves during the winter season, causing a distinctive bleaching of the decaying leaves, and finally ascomata and ascospores are produced again in the following spring.
The filiform, hyaline ascospores of Linosporopsis are very unusual for Xylariaceae, which mostly have more or less ellipsoid, brown ascospores, and therefore, the placement of Linosporopsis within Xylariaceae sensu stricto is somewhat surprising. However, ascospore morphology has proven not to be a good character for family segregation in the Xylariales, while the asexual morphs seem to agree better with the phylogeny Rogers 1996, 2002;Wendt et al. 2018). So far, no asexual morph is known for Linosporopsis. The hyaline, filiform spores are likely an adaptation to colonization and infection of living leaves of trees. While little understood and investigated in detail, there is strong evidence that long, curved spores are effective adaptations to facilitate attachment on vertical or otherwise challenging exposed surfaces and are therefore advantageous for successful germination and establishment on aerial plant parts (Calhim et al. 2018). It is therefore not surprising that filiform ascospores have independently evolved in leaf-inhabiting species of various ascomycete lineages. This also provides an explanation for the morphological similarities to the unrelated diaporthalean genus Linospora, which has a similar ecology.