Schroeteria decaisneana, S. poeltii, and Ciboria ploettneriana (Sclerotiniaceae, Helotiales, Ascomycota), three parasites on Veronica seeds: first report of teleomorphs in Schroeteria

Ciboria ploettneriana, Schroeteria decaisneana, and S. poeltii produce morphologically very similar apothecia emerging from fallen stromatized seeds of Veronica spp., the former two on V. hederifolia agg. in temperate central Europe and S. poeltii on V. cymbalaria in mediterranean southern Europe. They are described and illustrated in detail based on fresh collections or moist chamber cultures of infected seeds. A key is provided to differentiate the three species from their teleomorphs. For the first time, connections between two teleomorphs and two Schroeteria anamorphs are reported. Members of the anamorph-typified genus Schroeteria are known as host-specific plant parasites that infect seeds of different Veronica spp. In earlier times, they were classified in the Ustilaginales (Basidiomycota), but since more than 30 years, they are referred to as false smut fungi producing smut-like chlamydospores, based on light microscopic and ultrastructural studies which referred them to the Sclerotiniaceae (Helotiales). During the present study, rDNA sequences were obtained for the first time from chlamydospores of Schroeteria bornmuelleri (on V. rubrifolia), S. decaisneana (on V. hederifolia), S. delastrina (generic type, on V. arvensis), and S. poeltii (on V. cymbalaria) and from apothecia of C. ploettneriana, S. decaisneana, and S. poeltii. As a result, the anamorph-teleomorph connection could be established for S. decaisneana and S. poeltii by a 100% ITS similarity, whereas C. ploettneriana could not be connected to a smut-like anamorph. Ciboria ploettneriana in the here-redefined sense clustered in our combined phylogenetic analyses of ITS and LSU in relationship of Sclerotinia s.l., Botrytis, and Myriosclerotinia rather than Ciboria, but its placement was not supported. Its affiliation in Ciboria was retained until a better solution is found. Also Schroeteria poeltii clustered unresolved in this relationship but with a much higher molecular distance. The remaining three Schroeteria spp. formed a strongly supported monophyletic group, here referred to as “Schroeteria core clade”, which clustered with medium to high support as a sister clade of Monilinia jezoensis, a member of the Monilinia alpina group of section Disjunctoriae. We observed ITS distances of 5–6.3% among members of the Schroeteria core clade, but 13.8–14.7% between this clade and S. poeltii, which appears to be correlated with the deviating chlamydospore morphology of S. poeltii. Despite its apparent paraphyly, Schroeteria is accepted here in a wide sense as a genus distinct from Monilinia, particularly because of its very special anamorphs. A comparable heterogeneity in rDNA analyses was observed in Monilinia and other genera of Sclerotiniaceae. Such apparent heterogeneity should be met with skepticism, however, because the inclusion of protein-coding genes in phylogenetic analyses resulted in a monophyletic genus Monilinia. More sclerotiniaceous taxa should be analysed for protein-coding genes in the future, including Schroeteria. Four syntype specimens of Ciboria ploettneriana in B were reexamined in the present study, revealing a mixture of the two species growing on V. hederifolia agg. Based on its larger ascospores in comparison with S. decaisneana, a lectotype is proposed for C. ploettneriana.


Introduction
The family Sclerotiniaceae comprises about 150 mainly plant parasitic species in 28 accepted genera . Members of most of these genera are pleomorphic, producing apothecia (teleomorph, sexual) and a conidial state (anamorph, asexual), but some are still without a known teleomorph. Sclerotiniaceae are thought to be a relatively recently evolved lineage of primarily necrotrophic to biotrophic host generalists or specialists (Andrew et al. 2012). Species of the family are generally hygrobionts; i.e. they form their apothecia and conidial states on fallen, permanently moist remnants of various mono-and dicotyledonous plants: on herbaceous stems, leaves, flowers, fruits, and seeds, also on wood and bark, rarely on dung (Whetzel 1945;Schumacher & Kohn 1985;Spooner 1987;Palmer 1991;. Three xerobiotic, apparently necrotrophic species on air-exposed branches, previously included in the heterogeneous subfamily Encoelioideae of Helotiaceae, have recently been transferred to Sclerotiniaceae and placed in the new genus Sclerencoelia Pärtel & Baral (Pärtel et al. 2016), which raises the number of accepted genera to 29.
A main characteristic of the Sclerotiniaceae and the closely related, paraphyletic Rutstroemiaceae is the amyloid ascus apical ring of the Sclerotinia type (Baral 1987;Verkley 1993). Members of both families have usually brownish coloured apothecia, and their stipe base is often blackish. Apothecia of Sclerotiniaceae emerge from a black sclerotium (sclerotial stroma) or, similar as in Rutstroemiaceae, from stromatized host tissue (substratal stroma, pseudosclerotium). Species of Sclerencoelia deviate from the remaining genera of the family by persistent, drought-tolerant apothecia and in asci with a more or less reduced apical ring. Sclerotiniaceae generally have at their flanks of the receptacle an ectal excipulum of textura globulosa which often includes rhomboid crystals, whereas Rutstroemiaceae mostly have a textura prismatica without crystals .
Mononematous or sometimes sporodochial, acervular, or pycnidial anamorphs are typical of the family Sclerotiniaceae (for references, see . Various members form characteristic macroconidial anamorphs, the most familiar being Botrytis P. Micheli ex Pers. and Monilia Bonord., which are important plant pathogens and also known for their teleomorph-typified names Botryotinia Whetzel and Monilinia Honey. Most members of the family possess phialidic microconidial synanamorphs which are either formed directly from ascospores or on short germ tubes (Schumacher & Kohn 1985).
The genus Schroeteria G. Winter, a group of false smut fungi, which over a long time has been misplaced in Ustilaginales (now Ustilaginomycetes), is extraordinary in forming sori of pigmented mitosporic diaspores, which are classified as chlamydospores, in fruits of different Veronica spp. They form a powdery spore mass which at maturity often completely fills the capsules of their host. The chlamydospores are roundish, warted, somewhat thick walled, and show light yellowish to reddish or greyish brown colours under transmitted light, but appear macroscopically dark brown to blackish. They are either formed singly or cohere in pairs or larger groups. Schroeteria also possesses a phialidic microconidial synanamorph that develops on the chlamydospores, for which it was assumed to represent a member of Sclerotiniaceae (Brefeld 1883(Brefeld , 1912Vánky 1982;Nagler et al. 1989). The mycelium of Schroeteria spp. destroys the interior of the capsules (seeds and/or funiculi and placenta) of the living plants without forming a dark stroma.
During 1986-2019, two morphologically similar sclerotiniaceous discomycetes have been collected on fallen previous year's seeds of Veronica hederifolia agg. (ivyleaved speedwell) at different sites of central Europe. In the first collection made in 1986 by P. Blank near Schaffhausen (Switzerland), the substrate was misinterpreted as gall of a gall wasp; therefore, the species was compared with gallinhabiting Sclerotiniaceae (Baral 1986;Palmer 1991). Further collections were made in 2003-2004 by G. Hensel, M., W. and E. Huth, and P. Rönsch in Sachsen-Anhalt (Germany), during which the substrate was correctly identified as Veronica seeds, owing to the close similarity of whitish, uninfected seeds of V. hederifolia agg. with those being stromatized. Not only fallen blackened seeds with apothecia but also blackened seeds without apothecia were collected at two localities (Trebnitz, Werder canal; Freyburg, Alte Göhle) in spring 2003 and incubated in a garden inside a plastic box with some moist earth and moss, in which apothecia developed during next spring.
At that time, it became evident that one of the two discomycetes must be Ciboria ploettneriana (Kirschst.) N.F. Buchw. This rarely reported species was described by Kirschstein (1906) on seeds of V. hederifolia agg. from collections made in 1899 and 1905 near Brandenburg a. d. Havel, and distributed by him in Ascom. exsicc. (Rehm 1905). Only a few later reports under this name are known to us (all from Germany): Benkert (2005) presented a sample made in 1992 in Berlin (Baumschulenweg), Kreisel (2011) published another sample made in 2008 by D. Benkert in the park of castle Liebenberg in Brandenburg, and Huth (2009, p. 103, pl. 36 fig. 104) reported three samples collected by G. Hensel, M. and W. Huth, andP. Rönsch in April-May 2003-2004 near Merseburg and Freyburg in Sachsen-Anhalt. Apothecia of a third species were detected by F.J. Valencia in January 2017 on seeds of V. cymbalaria in a mediterranean area of southern Spain. Although it differs only slightly from S. decaisneana in its teleomorph morphology, it turned out to belong to a species of its own based on rDNA.
In 2019 we started our bibliographic and molecular investigations on the genus Schroeteria, based on M. Bemmann's suspicion that this genus could represent the anamorph of the two discomycetes. Chlamydospores of S. decaisneana (Boud.) De Toni could subsequently be detected in May 2019 by H. and U. Richter at the Zeuchfeld site on mature non-stromatized seeds in the capsules of living plants and offered the possibility for comparative rDNA studies. referred to the anamorph-typified genus Myrioconium, which is characterized by small, ± globose, hyaline, and smooth or sometimes warted conidia. The three sclerotiniaceous genera Coma, Microgloeum, and Mycopappus have been erected for their microconidial anamorphs, but are so far without a known teleomorph.
In their "Recommendations on generic names competing for use", Johnston et al. (2014b) proposed to use the anamorphtypified names Botrytis and Valdensia for the holomorph in replacement of Botryotinia and Valdensinia, respectively. In all the remaining pleomorphic genera, the authors proposed to maintain the teleomorph-typified name, with one exception: although they gave recommendations on three of the four abovementioned microconidial genera, they did not treat the oldest genus Myrioconium as a genus competing with Myriosclerotinia. According to Schumacher and Kohn (1985), the type species of Myrioconium, M. scirpi, is a synonym of Myrioconium scirpicola, the anamorph of Myriosclerotinia scirpicola, which in turn is the type species of Myriosclerotinia. Thus, Myrioconium is a synonym of the younger Myriosclerotinia which was introduced by Buchwald (1947) at a time when different names were required for different morphs. Today Myriosclerotinia should be protected as it appears in Internet search engines much more often than Myrioconium.
Schroeteria chlamydospores are formed in sori inside the capsules of the host plants, where they replace either more or less entirely the seeds (e.g. S. delastrina, S. poeltii) or only placenta, funiculus, and hilum by leaving the testa, endosperm, and embryo unaffected (S. decaisneana). When the capsules finally burst, the chlamydospores appear to be spread by wind and water. During germination on agar, they form hyaline germ tubes on which phialides arise that produce globose, smooth microconidia (earlier referred to as "sporidioles" or "sporidia") with an eccentric oil drop (Brefeld 1883;Vánky 1982;Nagler et al. 1989). Other taxa earlier placed in Schroeteria (S. annulata Ellis & Everh., S. arabica (Henn.) Henn., S. cissi (DC) De Toni) lack this kind of phialides and grow on other host genera.
Nomenclature. Schroeteria was erected by Winter (1881) as a replacement name (nom. nov.) for Schröter's illegitimate name Geminella. The hitherto-cited original publication by Schröter (in Rabenhorst 1870b: 137) represents a copy of his shortly before-issued herbarium label of Fungi europaei exsiccati (Rabenhorst 1870a(Rabenhorst : no. 1376), which includes a valid description of the genus including a diagnosis (Plate 2, U. Braun & K. Bensch pers. comm.). Here, Schröter recognized only one species in Geminella, G. delastrina. Geminella and consequently also Schroeteria are thus typified by G. delastrina. Vánky (1982) incorrectly cited "Geminella Schröter (1869, p. 5)", being unaware that 1869 was not the year of printing but refers to the reporting year of the publication to which Schröter's article was assigned.

History about the higher classification of the genus Schroeteria
Since its valid description by J. Schröter (in Rabenhorst 1870a, b) under the illegitimate name Geminella, the genus has been considered over more than a hundred years as belonging to the smut fungi (Ustilaginales, now Ustilaginomycetes) in the Basidiomycota. Therefore, its dark-coloured warted chlamydospores were often called "teliospores", whereas Brefeld (1912), who pointed out the true relationship with Ascomycota, called them chlamydospores because of their thallic ontogeny that represents a direct transformation of hyphal cells into spores. Vánky (1982Vánky ( , 1983 and Nagler et al. (1989) simply named them "spores", whereas Bauer et al. (2001) and Vánky (2008a, b) proposed to reinstate the term teliospores in a wide sense for thick-walled resting spores of plant parasites surviving unfavourable periods, mainly during winter, and not to restrict the term to dicaryotic probasidia of basidiomycetous rusts and smuts. They also redefined the term "smut fungus", i.e. from a taxonomic group to a life strategy and organization, to include non-ustilaginomycetous groups of plant parasites that develop teliospores as organs of dispersal and resistance. Other authors have used the term "false smuts" for such nonustilaginomycetous fungi, e.g. Tanaka et al. (2008). This usage traces back to Brefeld's opinion who applied the German version "Falsche Brandpilze" (Brefeld 1908, p. 221).
The later ignorance of its true relationship is astonishing, since already Brefeld (1883Brefeld ( , 1912 mentioned the strong similarity of germ tubes, phialides, and microconidia in G. delastrina and G. parvispora (= S. decaisneana) with those of the helotialean genus Sclerotinia, viz. in S. tuberosa (Hedw.) Fuckel, S. sclerotiorum (Lib.) de Bary, and perhaps S. trifoliorum Erikss. (all under the name Peziza, the latter as P. ciborioides), compared to a high dissimilarity to hemibasidia of Ustilaginomycetes. Brefeld was convinced about Schroeteria belonging in relationship of Sclerotinia, e.g. because the microconidia did not germinate, which is typical of Sclerotiniaceae and generally interpreted as an indication that they function as spermatia. He could even verify the formation of sclerotium-like bodies in pure culture after several weeks, but did not succeed in obtaining apothecia (see also Vánky 1982;Nagler et al. 1989). Brefeld also failed to find sclerotia or apothecia at the sites where the infected plants grew (Brefeld 1912: 79). Although Brefeld's observations and conclusions have later widely been recognized, they were only sometimes accepted, particularly by Schellenberg (1911) who listed Schroeteria delastrina among the genera and species to be excluded from the Ustilaginaceae. Also Lindau (1912) accepted Brefeld's opinion by expressing doubts that Schroeteria belongs in the "Ustilagineae", but he preferred to leave it there because of the custom at that time to associate the genus with this family.
Others doubted a relationship with Ascomycota, e.g. Ferdinandsen & Winge (1914, p. 4). Thirumalachar & Whitehead (1968) referred Schroeteria in synonymy with Schizonella J. Schröt. (Ustilaginales) by misinterpreting the observations of Schröter and Brefeld as an untypical case, believing that the microconidia never separate from each other; i.e. the germ tubes convert into "beaded cells". In their studies of S. delastrina (l.c.: fig. 8), these authors observed instead elongate fusoid "secondary sporidia" of 6 × 2.5 µm on septate germ tubes and they concluded that this type of germination is typical of Ustilago spp. Also Vánky (1982), who illustrated the characteristic microconidial synanamorph of Schroeteria in detail and stressed its possible relationship with Ascomycota, still retained the genus in Ustilaginales and referred to the chlamydospores as "teliospores". Nagler et al. (1989) studied S. delastrina and S. poeltii by cultural and ultrastructural methods. The authors could not obtain sclerotial structures, but they concluded that Schroeteria represents an anamorphic genus of Ascomycota, based on the absence of caryogamy and meiosis, the consistent presence of multinucleate germ tubes, the morphology of the spindle pole bodies, the presence of septal pores with a pore plug and Woronin bodies, and the absence of layering in the cell wall. This opinion then also Vánky (1994) accepted. Nagler et al. (l.c.) doubted Thirumalachar & Whitehead's (1968) findings of spore germination in S. delastrina as they could never see this type of germination in their studies. Their unusual observation of endogenous maturation of microconidia inside chlamydospores of S. poeltii and inside germ tubes of S. delastrina (Nagler et al. 1989: figs. 7, 9) requires further attention.

Other false smut fungi with ascomycetous relationship
A similar fungus with warted brown spores, though not formed in pairs, is Restilago Vánky with one species, R. capensis Vánky, growing in capsules of Ischyrolepis capensis (Restionaceae, Poales). This was shown to be "the second genus of smut fungi of ascomycetous origin" (Vánky 2008a) because of Woronin bodies at the septal pores. Another false smut fungus is Hapalosphaeria deformans (Syd. & P. Syd.) Syd. It causes stamen blight of blackberry (Rubus ?corylifolius agg., as Rubus dumetorum) and was already stated by Diedicke & Sydow (1908) as belonging to ascomycetes, but requires further, particularly molecular investigations. Contrary to Schroeteria and Restilago, it forms hyaline, smooth phialoconidia inside brown pycnidia in anthers of Rubus. No sequence data of these two genera are known to us. Another false smut is Ustilaginoidea Brefeld, which forms teliospore-like, olive-brown, subglobose, warted chlamydospores (Brefeld 1895, p. 194 f.). Earlier placed in Ustilago or Tilletia, the two economically important plant parasites of rice (Ustilaginoidea oryzae (Pat.) Bref.) and a bristle grass (Ustilaginoidea setaria Bref.) were shown by Brefeld to produce in culture Claviceps-like ascomata emerging from sclerotia (Brefeld 1896, p. 103;1912, pl. 3 figs. 1-15; see also Tanaka et al. 2008). Based on molecular methods, Bischoff et al. (2004) placed Ustilaginoidea in Clavicipitales. Finally, Tanaka et al. (2020) referred four species, which were previously placed in Ustilago and infect ovaries of monocot flowers of the family Commelinaceae, to a new genus Commelinaceomyces in Clavicipitaceae (Sordariomycetes).

Materials and methods
Morphology. All collections were examined in the living state in tap water (see Baral 1992), using a Zeiss Standard 14 microscope with 10 × Zeiss and 15 × Euromex oculars and Nikon Coolpix E4500 (H.B.), a Zeiss L 421 (1939) with 7 × and 10 × oculars (P.R.), a Zeiss Axioscope with Nikon Coolpix E4500 and Zeiss Stemi 2000C with Canon EOS 600D (V.K.), and an Optika B-353PLi microscope with E-Plan IOS objectives and a Canon EOS 1200D reflex camera, and for macrophotographs the same camera with a Tamron SP AF 90 mm macro lens (C.V.L.). The iodine reaction was tested with Lugol's solution (IKI = ~ 1% I 2 , 2% KI, in H 2 O), without and with potassium hydroxide (KOH ca. 3%) pretreatment. Brilliant Cresyl Blue (CRB, ~ 1% in H 2 O) added to a water mount was used for vital staining the vacuoles, also for staining spore wall surfaces for the detection of gel. All drawings were done free-hand. Kirschstein's syntype specimens were reexamined from B (Botanisches Museum Berlin-Dahlem, Germany) and Terrier's specimen on Veronica campylopoda from NEU (Université de Neuchâtel, Switzerland). Personal voucher specimens were deposited in the following herbaria: Used abbreviations: LBs = KOH-inert oil drops (lipid bodies); VBs = KOH-sensitive refractive vacuolar bodies; SCBs = KOH-sensitive cytoplasmic bodies; OCI = relative oil content index (0 = without oil drops, 5 = maximum lipid content); * = observation of living cells, † = observation of dead cells; MTB = Messtischblatt (German topographic map), IVV = www. in-vivo-verit as. de (link to drawings and photographs); sin. doc. = without macro-or microscopic documentation, ø = unpreserved, sq. = sequence in GenBank.
Pure cultures from ascospores were tried on MEA (Malt Extract Agar) and MMN (Modified Melin Norkrans Medium) (A.U.). In order to induce the formation of microconidia from ascospores, apothecia were placed for 3-4 days in a moist chamber at 10-20 °C. The formation of apothecia from seeds was achieved by picking up infected, blackened seeds in June and placing them at the same day on damp earth covered by moss inside a not completely water-tight plastic box that was deposited on the ground under a boxwood hedge in a garden.
Veronica species were identified using Jäger (2017) and Parolly & Rohwer (2019), for V. cymbalaria Jahn & Schönfelder (1995), and for V. campylopoda Hong & Fischer (1998). V. cymbalaria from Spain was confirmed by an ITS sequence. Current plant names were established using The Plant List (2020), and the names of fungi generally follow Index Fungorum.

Molecular methods.
Sequences of ITS and LSU rDNA were obtained by A. Urban from apothecia of two samples of Schroeteria decaisneana (Zeuchfeld near Freyburg in Sachsen-Anhalt, Germany) and one of Ciboria ploettneriana (Alte Göhle near Freyburg). Further sequences of ITS and LSU rDNA were obtained by J. Kruse and S. Ploch from anamorph sori of Schroeteria decaisneana (Zeuchfeld), S. delastrina (Kyffhäuser and Hainleite, Thüringen), S. bornmuelleri (Mashhad, Iran), and S. poeltii (Rhodos, Greece), and from apothecia of S. poeltii (Ronda, Spain). These sequences comprise the entire ITS region and the LSU D1-D2 domain. DNA from three sori extracts was 2-3 × sequenced. For verifying the macroscopically identified V. cymbalaria, the plant ITS was sequenced.
Methods used by A. Urban: About a quarter to one ascocarp was processed from dry fungarium samples, depending on ascocarp size. The material was placed in 2 mL plastic reaction tubes and pulverized in a grinding mill (MM2, Retsch) using a combination of five 2 mm and two 3 mm glass beads and about at 30 Hz (maximum speed) for 15 min. If pulverization was incomplete, about 0.4 mL of coarse quartz sand was added and the grinding repeated. A modified CTAB miniprep DNA extraction protocol (Schickmann et al. 2011) was used for DNA extraction. DNA fragments spanning the whole nuclear ribosomal internal transcribed spacer region (ITS1, 5.8S, ITS2) and about 600 basepairs from the 5′ end of the 28 s ribosomal RNA encoding DNA were amplified using the following primer combination: ITS1F (CTT GGT CAT TTA GAG GAA GTAA; Gardes & Bruns 1993) and TW13 (GGT CCG TGT TTC AAG ACG; http:// nature. berke ley. edu/ bruns lab/ tour/ prime rs. html). Sanger sequencing was performed using the PCR primers and the internal primers ITS3 (GCA TCG ATG AAG AAC GCA GC) and ITS4 (TCC TCC GCT TAT TGA TAT GC; White et al. 1990). DNA was amplified with the following thermocycling pattern: initial denaturation at 96 °C for 2 min (one cycle); denaturation at 95 °C for 30 s; annealing at 54 °C for 30 s; 72 °C for 90 s (40 cycles); and final elongation at 72 °C for 180 s; using a TGradient thermocycler (Biometra, Göttingen, Germany) and DreamTaq™ green PCR master mix (Thermo Fisher Scientific). Exo1/FastAP co-digestion (Thermo Fisher Scientific) was used to dephosphorylate unincorporated nucleotides and to digest excess primers. Sanger sequencing was performed using BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and was analysed on an automated DNA sequencer (ABI 3730xl Genetic analyser, Applied Biosystems) after purification of cycle sequencing products using Sephadex (GE Healthcare Life Sciences) filtration.
Methods used by J. Kruse and S. Ploch: About 2-10 mg of spore mass was taken from infected seeds of the fungarium samples J.K. S1346 (GLM-F129032), S1304, B2278, V. K. P1652-23, -26, -27, H.U.V. 750 ex TUB and C.V.L.040117 (GLM-F29000). The material was placed in 2 mL plastic reaction tubes and homogenized in a mixer mill (MM2, Retsch) using a combination of five to eight 1 mm and three 3 mm metal beads at 25 Hz for 5 min. Genomic DNA was extracted using the E.Z.N.A Plant DNA Mini Kit (VWR).The incubation time was extended to one hour. The complete nrITS of all DNA extracts was amplified using the primer pair ITS1F and ITS4 (White et al. 1990;Gardes & Bruns 1993) at 56 °C annealing temperature. The cycling reaction was performed in a thermocycler (Eppendorf Mastercycler 96 vapo protect; Eppendorf, Hamburg) with an initial denaturation at 95 °C for 4 min, 36 PCR cycles of denaturation at 95 °C for 40 s, annealing at 56 °C for 40 s and elongation at 72 °C for 60 s, followed by a final elongation at 72 °C for 4 min. The LSU rDNA was amplified using the primer pair LR0R and LR5 (Vilgalys 1988) with the condition mentioned in Vilgalys & Hester (1990). The resulting amplicons were sequenced at the Biodiversity and Climate Research Centre (BiK-F) laboratory using the abovementioned PCR primers. Sequences were deposited in GenBank (Table 1).
Alignments were done with MAFFT (https:// mafft. cbrc. jp/ align ment/ server/). They can be found in supplementary information S4-S6. Maximum likelihood phylogenetic analyses were carried out with MEGA7 and MEGAX with the settings "use all sites, nearest-neighbour-interchange, weak branch swap filter", also with IQ-tree. Branch support is given as maximum likelihood bootstrap percentages. P-distances were evaluated with MEGA6 or MEGA7 using individual alignments of species pairs, with the settings transitions + transversions and pairwise deletion.
Plate 4 Schroeteria decaisneana (teleomorph). 1a Fresh apothecia emerging from seeds of Veronica hederifolia agg.; 2a idem, rehydrated; 1b-d median section of seed (b-c with central hilum, c-d with embryo, d with insertion of apothecial stipe); 1e hyphae inside epidermal cells of seed; 2b median section of receptacle (with amyloid subhymenium and outer medullary excipulum); 2c ascus and paraphyses; 2d ascus bases with croziers; 2e apices of immature (above) and mature (below) asci; 3a paraphyses, terminal cell con-taining inconspicuous large vacuoles but no VBs; 1f, 2f, 3b mature ascospores containing one central nucleus and one large and some minute LBs at each end); 2 g, 3c overmature ascospores budding conidia; 3d conidia detached from ascospores, containing one large eccentrical LB. Living state (1 & 3), dead state (2) Taxonomic remarks: The first collection of the teleomorph of Schroeteria decaisneana was made in 1986 by P. Blank near Schaffhausen (Switzerland) and studied by the first author in the fresh state (Plate 4: 1). The fungus was briefly reported by Baral (1986) as "Ciboria" cf. gemminicola Rehm, because it was erroneously believed to grow on galls of Neuroterus albipes (Cynipidae, Hymenoptera), a gall wasp that inhabits leaves of Quercus in Europe. In fact, oak galls, especially those of the related N. numismalis, resemble seeds of Veronica. In his review of Sclerotiniaceae on oak galls, Palmer (1991) mentioned the above collection by reproducing the first author's drawing, though without personal study. Ciboria gemminicola was described in Wagner (1895) on galls of Cynips gemmae with distinctly smaller, especially narrower spores of 8-9 × 3.5 µm.
In the following years, the species was repeatedly collected in Sachsen-Anhalt (Plate 3). At this occasion, the fungus turned out to grow in fact on seeds of Veronica hederifolia agg. Therefore, the provisional name "Ciboria seminis-veronicae" was used for it, which appears also on the herbarium label of the specimen deposited in M as the intended holotype (H.B. 8687).
Variation: In all collections of Schroeteria decaisneana, from which detailed documentations were made, the paraphyses were cylindrical and apically scarcely inflated. Yet, in one of the large apothecia of a sample from Freyburg (Zeuchfeld, H.B. 8687), they were predominantly strongly inflated and variously shaped (Plates 3, 4: 3a, 7: 1j). The pale blue IKI-reaction of the subhymenium and outer medullary excipulum appears to be variable: it was present in the samples from Schweinfurt and Schaffhausen, but absent in those from Freyburg (no data available for those from Brandenburg and Merseburg). Variation was also noted in ectal excipular cell size at the lower flanks, being much larger in the samples from Freyburg (H.B. 8687, 8955) compared to Schweinfurt (H.B. 5698) (no data available for those from Brandenburg, Merseburg, and Schaffhausen). Also the orientation of the cells varied. Finally, crystals were present in the collections from Schaffhausen and Freyburg, but they were not seen in those from Schweinfurt and Merseburg (no data available for that from Brandenburg). In the Schweinfurt collection, the germinating ascospores were predominantly 1-septate but also non-septate, and often they budded at both ends (Plate 4: 2g), whereas in that from Freyburg, they were always non-septate and germinated only at one conidiogenous locus (Plate 4: 3c).
Cultural studies: Microconidia were abundantly produced from ascospores in senescent apothecia after incubation in a moist chamber. When shot on agar medium (MEA), the ascospores did not produce a mycelium even after several weeks of incubation at room temperature. According to studies by Brefeld (1912) and Vánky (1982), fresh chlamydospores germinated in tap water or strongly diluted nutrient solution at room temperature after 1-2 weeks and formed hyphae that abundantly produced phialides and microconidia. After adding concentrated nutrient solution, abundant mycelium developed instead. No attempts have been undertaken in the present study to obtain apothecia from stromatized seeds in a moist box.
Similar species: The study by the first author of a recent collection of Ciboria polygoni-vivipari Eckblad (Sweden, Lapland, Saxnäs, Marsfjället, 950 m, on bulbils of Bistorta vivipara (= Polygonum viviparum), 28.VII.2010, P. Perz, H.B. 9387, Baral ined., see IVV and https:// svampe. datab asen. org/ taxon/ 11632) revealed some resemblance with the S. decaisneana teleomorph, particularly regarding the ascospores containing mostly two polar, mediumsized LBs. However, the spores were distinctly larger (*15-19 × 7.2-7.8 µm) and contained at least two nuclei, the paraphyses contained many low-refractive VBs in their apex, the ectal excipulum was of textura globulosa, and no crystals were seen. Ciboria seminicola (Kienholz & E.K. Cash) Hechler (including the questionable C. betulae (Woronin) W.L. White) on fruits of Alnus and Betula differs in somewhat longer and narrower, fusoid, warted ascospores and in asci arising from simple septa without croziers (Baral ined.,H.B. 3677,3682,5136,9774,see IVV). Terrier (1958) believed to have found Schroeteria decaisneana on Veronica campylopoda based on a collection received from Yerevan (Armenia). We have some doubts about this, because V. campylopoda belongs to Veronica subgenus Pocilla with a clearly different seed morphology. The present reexamination of Terrier's specimen in NEU verified the host species but confirmed our doubts about the fungus, which severely deviated from S. decaisneana by chlamydospores which filled the entire capsules similar as in S. delastrina but, unlike those, never cohered in pairs. The chlamydospores had a size of †(9.5-)10-12 µm diam. and an ornamentation of 0.3-0.5 µm high ridges forming an incomplete network (see IVV). The low ornamentation and the absence of cohering spores are actually reminiscent of S. decaisneana, except for the slightly larger spore size.
Distribution: The distribution of the anamorph of S. decaisneana comprises various countries of Europe (Scholz & Scholz 1988 For a description see Vánky (1982).
Ecology: Schroeteria delastrina was reported by earlier workers on different Veronica species, though mainly on Veronica arvensis, the type substrate. Our investigated collections from Germany and Greece were all on this host plant which occurred at the collection sites in different biotops: in Germany in ruderal vegetation at the border of farmland (Kelbra), in mown meadows (Hannover, Sondershausen, Thale), and in a Mesobromion grassland (Bad Frankenhausen, Leistadt), and in Greece in an annual vegetation on a raised area on the border of a summer-dry stream bed. The geology was mainly basophilically influenced, the phenology is May-June, in southern Europe March.

Specimens included (all on Veronica cymbalaria).
Teleomorph ( Taxonomic remarks: In the teleomorph, Schroeteria poeltii deviates from S. decaisneana merely by slightly smaller asci and ascospores, the latter having their LBs a bit closer to the spore ends, perhaps also in the absence of glycogen in the ascospores and in wider marginal excipular cells (Table 2). Despite these minor differences, the anamorph sharply differs by smooth, yellowish brown chlamydospores cohering in strongly curved (U-shaped) chains of (2-)4-7 spores from the other Schroeteria spp. with their warted, greyish brown chlamydospores cohering in straight chaines of 2-4 spores or detaching as single spores. The exclusive occurrence of S. poeltii on V. cymbalaria and the remarkable anamorph make any confusion with other species very unlikely. Nagler et al. (1989) investigated the type of Schroeteria poeltii and a collection of S. delastrina on V. arvensis by light and electron microscopical methods. The authors observed microconidia formed endogenously in phialides which emerge either from germ tubes of the chlamydospores or directly from the chlamydospores. Besides, the authors illustrated an unusual case of endogenous maturation of microconidia inside of chlamydospores ( fig. 7), which they also observed inside hyphal cells of S. delastrina ( fig. 9). Problematic is that the scales are wrong in some of their illustrations. In order to achieve reasonable measurements, the scales in figs. 6 and 9 should be around 20 µm instead of 5 µm and that in fig. 7 around 3 µm instead of 2 µm, whereas the scales in figs. 12-17 appear to be correct.
In our collection from Filerimos, not all capsules of a given plant individuum were filled with chlamydospores but numerous were with seed formation. However, this varied even on a single shoot of a plant. Whether the seeds are able to germinate has not been tested. This is in contrast to our collections of Schroeteria delastrina on Veronica arvensis where all capsules of a shoot on a plant were either filled with chlamydospores or with seeds, suggesting a systemic infection.
Typification and etymology: Vánky (1983), who investigated the type of S. poeltii, stated that the species was only known from the type locality, where it was collected by H. Teppner in 1962. In a footnote, he explained the reason for naming the species S. poeltii: shortly after publication of Vánky's Schroeteria monograph which appeared in 1982, Josef Poelt had sent him a specimen of Teppner's collection. Part of this collection then remained in GZU, whereas the holotype was stated by Vánky to have been deposited at UPS. However, according to Å. Kruys (pers. comm.) the specimen could not be found at UPS. In 1987 Vánky collected a topotype, which Nagler et al. (1989) investigated. In 2013 Vánky gave his entire herbarium to BRIP (Brisbane) where the holotype and also the topotype are listed in the online database.
Ecology: The host plant of Schroeteria poeltii has a (sub) mediterranean distribution and forms urn-shaped seeds similar as in V. hederifolia (cyathiform fide Muñoz-Centeno et al. 2006). The few phenological data suggest that the anamorph occurs during spring (March-June) and the teleomorph during winter (January). ; 1d-f idem, ectal excipulum at lower flanks and junction with stipe; 1 g idem, ectal excipulum in stipe; 1 h idem, medullary excipulum and subhymenium; 1i mature asci; 1j paraphyses, containing large non-refractive vacuoles; 1 k ascus bases with croziers; 1 l-m, o, q apices of mature asci (l-m with amyloid ring); 1n, p free ascospores containing two large and some minute LBs, central nucleus faintly visible; 1r central nucleus more clearly visible (right spore with two glycogen regions); 1 s overmature ascospores budding conidia with large eccentrical LB. Living state (in water, 1r in IKI), except for 1 l-m (dead state in IKI), ascus in 1o, four spores in 1p.

◂
The French holotype collection was from a mesomediterranean semihumid site northeast of Monaco in dépt. Alpes-Maritimes. The collecting locality is not fully clear: the rough coordinates (43°47′N, 7°30′E) given by Vánky (1983) are in a region about 1 km north of the seaside town Menton and cover a region of about 50-200 m altitude. The given data about altitude (ca. 600 m) and site (tract Ste. Agnès, Umgebung von Ste. Agnès) in Vánky (l.c.) and on the handwritten isotype label in GZU (which lacks coordinates), suggest that the collection site was about 4 km northwest of Menton, i.e., about 3 km away from the published coordinates.
The Schroeteria poeltii anamorphs from thermomediterranean semihumid Greece were collected at the site Filerimos in an open grove with Quercus coccifera on top of a hill in the monastery park at 212 m altitude in the north of Rhodos, and near Charaki in a ruderal farmland area with Mercurialis annua etc., a former military area close to coastline at 7 m altitude in the middle east of Rhodos.
The teleomorph from supramediterranean semihumid Spain was from an acidic floodplain forest in Málaga (Andalucía). The apothecia were found on stromatized seeds of Veronica cymbalaria fallen to the ground. The plants grew just above the ground on the vertical area of a schist rock. The floodplain forest consisted of Populus alba and Salix alba, and mediterranean plants such as Quercus faginea and Rubia peregrina, also Lamium flexuosum, Rubus ulmifolius, Dorycnium rectum, Ficaria verna, and Vinca difformis. Outside of the flooded area are large Castanea sativa plantations, the main agricultural product of the local people.

Taxonomic remarks: Ciboria ploettneriana resembles
Schroeteria decaisneana in most of its morphological traits, including various microscopic details, such as size of asci and paraphyses, size and shape of medullary and ectal excipular cells, IKI-reaction of ascus apex and medullary excipulum, and presence of crystals. The species deviates macroscopically by light-coloured, often whitish, somewhat shorter apothecial stipes with a lighter colour at the base, and seeds with a more blackish surface, causing a strong optical contrast between stipe and seed, in contrast to greybrown seeds in S. decaisneana. Microscopically, C. ploettneriana deviates from S. decaisneana by distinctly larger, slightly more heteropolar ascospores with often pointed, subacute ends. Under vital study, C. ploettneriana differs in ascospores with a distinctly lower lipid content composed of much smaller LBs and (2-)4 nuclei instead of only one, and in ± strongly refractive vacuoles (VBs) in the terminal cells of paraphyses (Table 2). Less evident or possibly variable features of C. ploettneriana are slightly longer asci, slightly smaller conidia formed on longer phialides, and an ectal excipulum of vertically oriented cells in the receptacle (in S. decaisneana often horizontal) and wider cells in the stipe (textura prismatica vs. t. porrecta), also with much shorter cells at the margin (Table 2).
Variation: Ascus and spore size was rather consistent among the specimens studied. Also the size of LBs never exceeded 1 µm diam., with rare exceptions of single spores. The number of nuclei in the spores varied between two and four within all four samples in which living spores have been studied in detail (H.B. 9037 & 9271, M.R. 6806, 2.V.2019), but the tetranucleate spores were often distinctly more numerous than the binucleate spores when regarding only mature, freshly ejected spores.
Cultural studies: Microconidia were abundantly produced from the ascospores in senescent apothecia after incubation in a moist chamber. When shot on agar medium (MEA), the ascospores did not produce a mycelium even after several weeks of incubation at room temperature.
In two of the here studied samples, apothecia were obtained after incubation of blackened seeds placed in June directly on damp earth in an untight plastic box that was deposited on the ground at a shady place. In this plastic box the apothecia developed during April of the following year.
The first report of Sclerotinia ploettneriana by Rehm (no. 1603, in sched.) and also the publication of this exsiccata in Annales Mycologici (Rehm 1905) is invalid, because only the collection data was provided but no description. Valid publication with description was done in Kirschstein (1906). The unillustrated protologue reports apothecia 2-3 mm diam., stipes 1-10 × 0.5 mm, emerging singly or up to four from one seed, asci 160-180 × 10-12 µm, and oval spores of 15-18 × 6-7 µm with 1-2 small oil drops. Kirschstein emphasized the black stromatization of the seeds in contrast to the whitish uninfected seeds.
The Rathenow specimen (B 70 0100003) contains principally about ten light brown seeds of Veronica hederifolia agg. from which apothecia with ellipsoid spores of †9-11 × 5-6 µm emerge (Plate 9). The light seed colour and the small spores indicate that this collection represents Schroeteria decaisneana, although the spores often contained more than one comparatively small oil drop in each half. A single seed without adhering apothecium appears to The two samples from Groß Behnitz (B 70 0100005-6, one shown on Plate 20: 1) each contain numerous bright to dark black-brown seeds of Veronica hederifolia agg., and apothecia which much better fit the protologue of Sclerotinia ploettneriana regarding spore shape (ellipsoid to fusoid, rarely clavate) and spore size (B 70 0100005: †11-16 × 5-6 µm, B 70 0100006: †12-15 × 4.5-6 µm, examined in water). Because of the small spore size in the Rathenow collection, we conclude that the protologue derives from the large-spored species alone and not from a mixture with Schroeteria decaisneana. Therefore, we here designate one of the two specimens from Groß Behnitz, B 70 0100006, as lectotype of Sclerotinia ploettneriana.
The examined duplicate of Rehm's exsiccata (B 70 0100004) provided a surprise: it does not contain any seeds of Veronica, but half a dozen of sclerotia with apothecia of an undetermined Sclerotiniaceae, also a few seeds of a ?Stellaria. The original label indicates that this exsiccata was composed of a mixture of two collections: the vast majority of the material was from Groß Behnitz (IV.1905), whereas a sparse minority came from Rathenow (29.X.1899). Other duplicates of this exsiccata might therefore actually contain seeds of Veronica with apothecia of Ciboria ploettneriana.
Ecology: Ciboria ploettneriana was recorded near Freyburg at three sites, in a thermophilous deciduous forest with Ulmus minor bordered by Prunus spinosa (Zeuchfeld, Plate 8), in a thermophilous Querco-Carpinetum (Alte Göhle), and in a more shady Aceri-Fraxinetum (Hirschrodaer Graben). The site near Merseburg was a ?Pruno-Fraxinetum river bank forest with Prunus padus (Elster-Luppe-Aue), that near Naumburg a nitrophilous Quercus-Fraxinus forest edge (Alliarion, Plate 19), and that near Wismar a narrow forest strip of Acer, Aesculus and Tilia between road and farmland. Monilinia johnsonii on Crataegus fruits occured together with Ciboria ploettneriana at the site near Naumburg.
Distribution: Besides the collections from Mecklenburg-Vorpommern, Brandenburg (including Berlin), and Sachsen-Anhalt as listed under Specimens included, no further unequivocal records of Ciboria ploettneriana came to our notice.

Molecular results
The obtained sequences of the three Schroeteria decaisneana samples fully concur in the ITS region between the two apothecial isolates (teleomorph) and the sorus isolate (anamorph), except for 1 nt in the ITS1 in one of the apothecial isolates (Table 3). Also in the LSU D1-D2, no difference was observed between the two apothecial isolates and the sorus isolate (Table 3). Likewise, the two S. poeltii sequences fully concur in the ITS and LSU D1-D2 between apothecial and sorus isolate, except for 1 nt in the LSU. Considering the comparatively high interspecific distances among the Schroeteria species and the concordant host of each species, our data prove that teleomorph and anamorph belong together.
The Schroeteria core clade clustered particularly with two Ericaceae inhabiting species: it formed a medium (Plate 21, S1) or strongly (S3) supported sister clade of M. jezoensis, for which only ITS was available, though with a rather long branch, and together with this species a less supported sister clade of M. azaleae (Plate 21, S1). Both species grow on Rhododendron and are members of the Monilinia alpina group of section Disjunctoriae, to which also M. cassiopes belongs (Batra 1991, p. 102) and according to our analyses also Stromatinia pyrolae.
In contrast to the phylogenetic affiliation of the Schroeteria core clade with Monilinia section Disjunctoriae, Schroeteria poeltii clustered distant from Schroeteria s.str. It formed with long branch an unsupported (Plate 21) or strongly supported (S3) sister clade of the supported clade of Sclerotinia bulborum and S. "binucleata" (an undescribed species of Sclerotinia s.l. on Ficaria verna and Corydalis cava). Ciboria ploettneriana clustered unsupported and with comparatively short branch with these and other typical members of Botrytis, Grovesinia, Myriosclerotinia, Sclerotinia, Stromatinia etc.

Specific nucleotide positions in the rDNA
Within the core clade of Sclerotiniaceae, no specific nucleotide positions in the ITS region and LSU D1-D2 domain have been found that characterize any of the different genera, such as Botrytis, Elliottinia, Grovesinia, Monilinia s.str., Myriosclerotinia, Ovulinia, Pycnopeziza, Sclerotinia s.l., or Valdensia. Regarding placement of Ciboria ploettneriana, only one position in the middle of ITS1 gives a hint on its generic affiliation: the species shares the motif GGGG YCT (Y = C or T) with most species of Sclerotinia s.l., but also with Grovesinia moricola, whereas other species have mostly the motif HGGG CCT (H = A or C or T).
The Schroeteria core clade shows some characteristic motifs. In the ITS region, pos. 123 of the 5.8S region is C and pos. 4 of the ITS2 region is G, whereas S. poeltii and other Sclerotiniaceae have T+T, except for Monilinia jezoensis which has T+G. In the LSU D1-D2 domain, Schroeteria decaisneana and S. delastrina have several extraordinary positions. Some of them occur also in Ciborinia erythronii, but none was observed in Ciboria ploettneriana and any other Sclerotiniaceae in GenBank (Plate 21). Vánky (1982) treated five species in Schroeteria, which he distinguished by the mature chlamydospores occurring mostly in pairs or threes to fours (S. bremeri, S. delastrina) or mostly single (S. banatica, S. bornmuelleri, S. decaisneana). Species delimitation was further accomplished by spore wall thickness and the kind of spore ornamentation. Chlamydospore size lies in the five species within a similar range of about (7-)8-16(-20) µm diam., which makes spore measurements comparatively useless for species delimitation, considering the high infraspecific, particularly intrapopulational variability observed in each species. For instance, chlamydospore size of S. decaisneana varied considerably within a preparation made by us from a single sorus (Plate 10: 2e-k). Vánky (1982) used chlamydospore size in his key, but the given measurements strongly overlap. Chlamydospore size clearly refers to single cells, whereas in the older literature, it is not always clear whether authors mean single cells or twin spores. Vánky (1983) described the morphologically deviating S. poeltii as a sixth species within Schroeteria, characterized by up to 6(-7)-celled, strongly curved (horseshoe-shaped), almost smooth chlamydospores.

Species delimitation within Schroeteria, doubtful measurements of chlamydospores and microconidia
Boudier (1887) separated S. decaisneana from S. delastrina owing to slightly smaller, soon single, at first glaucous or bluish grey, finally grey-black or slate-grey spores born on narrower hyphae, and a different host plant on which it merely attacks the funiculus, leaving the seed and placenta intact. Brefeld (1912, p. 75) was unaware of S. decaisneana when he proposed the name "Geminella (Schroeteria) parvispora" (a taxon which we here consider as a synonym of S. decaisneana) for collections on Veronica hederifolia agg. The name Geminella parvispora, which was not listed in databases before November 2020, is mentioned several times in Brefeld's text and in his legend to tab. III figs. 16. When first mentioned on p. 75, he cites it as follows Plate 13 Schroeteria poeltii (teleomorph). 1a, e, j Median section of apothecium (1e in IKI, 1f, j in KOH); 1b idem, medullary excipulum; 1c, f, i idem, ectal excipulum; 1d surface view on stipe showing hairs (in KOH); 1 g-h crystals in stipe and medullary excipulum, respectively (in KOH); 1 k faintly amyloid subhymenium (in IKI); 2a, e living mature asci; 2b ascus base with croziers; 2c-d dead asci and spores (in IKI), with amyloid ring; 3a-c living ascospores; 4a-c living paraphyses. In water, if not otherwise stated. -1-4 4.I.2017 (C.V.L. 040117). Spain, Andalucía, Málaga, Pujerra. -Phot. F.J. Valencia ◂ (translated from German): "The fungus living on V. hederaefolia is the small, single-spored bluish form which produces an easily dispersed spore dust and is here named Geminella (Schroeteria) parvispora". In our opinion, the taxon is to be considered as validly described on p. 75 under the name Geminella parvispora.
Two years later, the valid combination Schroeteria parvispora was used by Ferdinandsen & Winge (1914, p. 4). In the same year, also Fischer (1914) published this combination, but that paper appeared later, apparently after September 1914, whereas Ferdinandsen & Winge's appeared on 17. July 1914. Later, Liro (1938) considered G. parvispora as a synonym of S. decaisneana. Brefeld (1912) characterized Geminella parvispora by small, mostly single, only slightly rough, blueviolet spores that easily get dispersed, in contrast to S. delastrina on V. triphyllos and V. arvensis etc. which has rough-warted, black, double-sized spores that are formed in pairs or sometimes threes. He did not indicate the origin of his samples, but it can be assumed that he collected them during his term in Münster (Nordrhein-Westfalen, Germany) where he worked as a botanist at the university and director of the botanical garden until 1898 (Brefeld 1912, p. 79). Thereafter, he was offered a chair in Breslau where he started to go blind in the same year due to a glaucoma.
Regrettably, Brefeld did not give any measurements of conidia or other elements. When looking at his illustrations (Brefeld 1912: pl. 3 figs. 16 & 18), it is evident that the individual cells of the chlamydospores of Geminella parvispora are only slightly smaller than those of G. delastrina and not "almost half the size" (Brefeld 1912, p. 75) or, a few lines later, vice versa those of G. delastrina not "more than double" the size of G. parvispora, provided that he compared the diameter of the cells and not their volume. Brefeld referred in this context to "single spores", and as he wrote that the "spores" often remain connected in pairs or threes, it appears that his remark on double-sized spores of G. delastrina cannot refer to the length of twin spores compared to single spores in G. parvispora. On the other hand, some authors used to measure chlamydospores of Schroeteria as an entity; e.g. in his key, Ciferri (1938) gave for S. delastrina a spore size of 15-23 × 8-12 µm (referring to twin spores) and for S. decaisneana 10-12 × 8-12 µm (referring to single spores), and also Bubák (1916) measured S. delastrina as 20-30 × 12-17 µm regarding twin or sometimes triple spores, and S. decaisneana as 7.5-15 µm regarding single spores (see Tables 4 and 5).
As a common usage in earlier times, Brefeld (1883: pl. 11 fig. 13, pl. 12 figs. 14-18;1912: pl. 3 figs. 16, 18) did not provide scales but only enlargement factors for his detailed illustrations, which were drawn at a 150 × up to 400 × magnification, according to his captions. Our reevaluation of spore size based on the printed books yields values much above the current data of the two species (Tables 2 and 3). Actually, a cell diameter above 17-18 µm appears to have never been reported for chlamydospores of Schroeteria; therefore, the real values of Brefeld's material were probably much lower. The actual average chlamydospore cell sizes of Schroeteria delastrina and S. decaisneana lie in the range of 8-12 µm, which is just half of what can be evaluated from Brefeld's sketches of S. delastrina (Table 4), whereas his drawing of G. parvispora yields a cell size of about 1.5 × larger than the current values (Table 5). On the other hand, Brefeld's (1883: pl. 6) illustration of Microbotryum cardui (as Ustilago cardui) yields teliospores of 16-19 µm diam., in good Brefeld's observations and drawings were always made from material he cultured in nutrient media rather than from freshly collected specimens. As an example, he described the enormous swelling of the endosporium of a twin spore (Brefeld 1883, p. 144, pl. 12 fig. 17a), which means that spore size evaluated from his drawings needs to be compared with caution with measurements of other authors who usually observed uncultured material. However, spores drawn by him without germ tubes or without emerging endosporium appear also oversized, although they should concur in size with uncultured spores because of an inelastic exosporium.
The presumed error in Brefeld's scale becomes evident when comparing microconidial size among reports of different authors. Surprizingly, also Cocconi's (1898) illustration of S. delastrina var. reticulata Cocc. yields doublesized values for chlamydospores (20-26 µm) as well as microconidia (7-8 µm), according to the stated magnification factor (Table 4). We thereby presume that the spore size of 16-20 µm given by Cocconi for twin spores of this variety refers to the diameter of single cells.
Without having examined the type, Ciferri (1931Ciferri ( , 1938) raised doubts about Cocconi's var. reticulata, which Cocconi (1898) distinguished by a reticulate epispore from the type variety which has a verrucose epispore. Because Cocconi's drawing shows spores with a dense spiny ornament in contrast with the description, and the host plant was seemingly Veronica praecox, on which also S. delastrina has been reported, Ciferri concluded that S. delastrina var. reticulata is a synonym of S. delastrina. Also Vánky (1982) considered Cocconi's description of reticulate spores as an inaccuracy, arguing that the spore surface in Schroeteria is generally verrucose and ribbed (but Vánky figured reticulate spores in S. decaisneana), and he also doubted Cocconi's large spore measurements.  Nagler et al. (1989)
Seed morphology and dispersal. The seeds of Veronica hederifolia agg. and V. cymbalaria are extraordinary in resembling a collapsed ball (Juan et al. 1994;Muñoz-Centeno et al. 2006). They are called cymbiform (boat-shaped) or cyathiform (urn-shaped) by showing a roundish ventral cavity. Also V. persica seeds have a ventral but more elongated cavity. The cavity is finally filled with air which aids in their transport by rain. The cavity includes also the elaiosome, a fleshy structure rich in nutrients. The elaiosome attracts ants which transport the seeds with their head. In the case of V. hederifolia agg., the elaiosome contains sugars, proteins, ricinoleic acid, and vitamins B1 and C (Bresinsky 1963: tabs 3, 6-7). Seeds of many other Veronica spp., e.g. V. arvensis, have more elongated, ellipsoid to flattened seeds without a cavity.
Phenology and life cycle. Host infection by Schroeteria spp. is systemic; i.e. the mycelium grows endophytically in the plant up to the flowers (Winter 1876;Brefeld 1912, p. 75). Like in smut fungi, it produces mitospores (here called chlamydospores) only in a special organ of the host (here Veronica fruits). According to Brefeld (1912) and Vánky (1982), plants infested by Schroeteria spp. do not differ in general appearance from healthy ones. The vegetative mycelium can be found in the intercellular space of the medullary parenchyma of the entire host plant (Winter 1876, p. 147, pl. 4 fig. 15), though sometimes one or more shoots or only some fruits may remain healthy. The mycelium grows through the floral pedicel, placenta, and funiculi into the young seeds where the chlamydospores are formed. In S. decaisneana, mycelium and chlamydospores replace placenta, funiculus, and hilum (Boudier 1887, p. 150) or only the funiculus (Bubák 1916, p. 60) by leaving the seed morphologically unaffected although this can no longer germinate, whereas in S. delastrina, the seeds are entirely absorbed by leaving placentae and funiculi unaffected (Winter 1876, p. 148, Boudier l.c.: 151, Bubák l.c.). Winter (1881, p. 118), who did not distinguish S. decaisneana from S. delastrina, wrote that the mycelium infects placentae, funiculi, and young seeds. The produced spores form a moldy-smelling, grey-brown, greyish blueviolet, or greyish black powdery spore mass which in S. decaisneana fills the ventral cavity of each seed and in S. delastrina the entire capsule (Winter 1881, see also Vánky 1982. The spore mass is generally called "sorus" following the custom with teliospores of ustilaginomycetous smut fungi. The capsule later usually tears open to release the spores passively. Kirschstein (1906) based his description of Ciboria ploettneriana on collections from October 1899 and April 1905. Under the assumption that the ascospores infect flowers of other individuals of this plant, he was astonished about the occurrence of apothecia in October as he could not find any evidence for a second flowering period of the annual host plant, which generally blooms in central Europe during (February-)March-May(-June) and fruits during April-June. However, Kirschstein's collections belonged to two different fungal species: the April collection we have selected as lectotype of Ciboria ploettneriana and the October collection we have reidentified as Schroeteria decaisneana. Indeed, the two species have a different phenology: apothecia of S. decaisneana were observed during end of October to first half of May and S. poeltii once in January, whereas those of C. ploettneriana only in April and first half of May.
Comparable to other taxa recognized in Ciboria, such as C. seminicola growing on Alnus seeds or C. amentacea on male Alnus catkins, the infection by C. ploettneriana ascospores could happen via the flowers during spring. After seed formation, the infected seeds would fall to the soil and later get stromatized. In the next spring, a new generation of apothecia will be being produced from these stromatized seeds. The apparent absence of a seed-born anamorph would exclude a pleiomorphic life cycle of this fungus (the phialoconidia of the microconidial anamorph probably function as spermatia).
The formation of apothecia of Schroeteria spp. not only in spring but also in late autumn and winter suggests that the ascospores do not or not only infect the host's flowers but start their life cycle in another way. Brefeld (1912, p. 79) was quite convinced that Schroeteria chlamydospores do not infect flowers of the host plants but germinate in the soil by producing a persistent mycelium that infects roots of young plants. Brefeld (1895Brefeld ( , p. 204, 1912) further correctly imagined that ovaries of the host plants might be transformed into "sclerotia", which in turn form apothecia during the first stages of seed germination. However, Brefeld and his coworker A. Kappenberg could not detect pseudosclerotia in the ovaries or apothecia associated with Veronica during their field work.
The hypothetical life cycle of Schroeteria decaisneana can thus be circumscribed as follows: Infection of young winter-annual seedlings of Veronica hederifolia agg. appears to take place during late autumn, warmer stages in winter, or early spring of the following year, either via chlamydospores that have germinated by forming a mycelium in the soil or via ascospores that have germinated there or perhaps on the host plant. During the next fruiting period in late spring (May-June), the mycelium inside the infected plants produces a new generation of chlamydospores in the capsules. According to our observations, virtually all capsules of infected plants contained chlamydospores in the seed cavities. Although such seeds superficially look healthy, Bubák (1916) found them to be incapable to germinate. We suspect, therefore, that the mycelium also enters the endosperm and/ or the embryo of the seeds. We further suspect that the capsules either open in late spring to distribute their chlamydospores, e.g. by wind or water, or the infected capsules or seeds fall down into the litter, where they successively lose their chlamydospores. In either case, the seeds may finally get stromatized in order to produce apothecia during the next autumn, winter, or early spring. In this way, the fungus has two possibilities to infect young seedlings: (1) via chlamydospores produced in spring and resting or producing a persistent mycelium in the soil during summer, and (2) via stromatized seeds resting in the soil during summer and producing apothecia and ascospores during the colder season.
In S. poeltii we observed that most capsules of an infected population of V. cymbalaria were filled with chlamydospores, but some shoots of a few plants had not only capsules with chlamydospores but also some with apparently normally developed seeds. It seems possible that these seeds are also infected and, after falling to the soil, finally get stromatized and produce apothecia during the germination time of the winter annual plant. The observation of S. poeltii apothecia in January suggests that also in this species the ascospores infect young seedlings after having germinated in the soil.

Hyperparasitism
The occurrence of two sclerotiniaceous species on the same organ of the same host plant could also mean that one species is a hyperparasite on the stroma of the other. Within Sachsen-Anhalt, apothecia of Ciboria ploettneriana and Schroeteria decaisneana were usually not collected at the same site; only at the site Zeuchfeld near Freyburg both species were found sympatric in the same habitat, though in different months. Comparable hypotheses have been proposed in other taxa of Sclerotiniaceae, in which two different species emerge from sclerotial structures formed on the same host plant. For instance, Spooner (1987, p. 251)  The hypothesis of S. decaisneana being a hyperparasite of C. ploettneriana would be in contradiction to the biology of its anamorph which, like anamorphs of other Schroeteria spp., is a direct parasite of the plant. C. ploettneriana as a hyperparasite of S. decaisneana would mean that during spring its ascospores attack plants that already have been invaded by S. decaisneana. Yet, any observation that supports this hypothesis is lacking.

Generic concepts within Sclerotiniaceae
Different generic concepts have been proposed within the Sclerotiniaceae in the past. The available phylogenetic analyses of rDNA (including our analyses) and rarely proteincoding genes suggest heterogeneity of some of the classical genera, such as Ciboria, Monilinia, Schroeteria, and Stromatinia, their species clustering in different clades with often unresolved phylogenetic position. These analyses also raise doubts about the current splitting into small genera; for instance, they question the distinction between Dumontinia, Grovesinia, Sclerotinia s.str., and Stromatinia s.str., which Plate 18 Ciboria ploettneriana (from Sachsen-Anhalt, Freyburg). 1a Median section of receptacle; 2a-c idem, ectal excipulum at flanks; 2e idem, at margin; 2d, 1b idem, medullary excipulum with crystals; 1i-j hair-like elements on ectal excipulum at flanks, containing refractive vacuoles (staining turquoise in CRB); 1c ascus; 1 l ascogenous hyphae with croziers; 1 h, k apices of immature and mature asci; 1d-f paraphyses, containing refractive vacuoles (in 1f turning purplish with age), 1 g stained turquoise in CRB; 2f-g mature ascospores containing a few minute LBs, 2-4 nuclei faintly visible; 2 h idem, in IKI, nuclei distinctly visible; 2i, k overmature ascospores budding conidia from phialides; 2j phialides formed on germ tube. Living state (in water; 2b-c, 2 h, 2 k in IKI; 1f in water, colour change in older apothecia; 1 g, 1i in CRB), dead state (1 k in IKI Based on anamorph morphology, heterogeneity of Monilinia was observed by Honey (1936), who subdivided the genus into two sections: Junctoriae (Monilinia s.str., growing on fleshy, edible fruits of domesticated Rosaceae, without intercalating disjunctors of the macroconidial chains) and Disjunctoriae (Monilinia s.l., growing on stromatized fruits of Rosaceae and Ericaceae-including the former Empetraceae and Pyrolaceae-with intercalating disjunctors). This subdivision was followed by Batra (1988Batra ( , 1991, who distinguished in 1991 five different groups within Disjunctoriae according to the inhabited host, and Schumacher & Holst-Jensen (1998), who raised the Disjunctoriae to a new though never validly published genus Franquinia Holst-Jensen & T. Schumach.
Heterogeneity of Monilinia was confirmed by molecular phylogenetic analyses by Holst-Jensen et al. (1997a: fig. 9, SSU+ITS+LSU; 2004, ITS), Takahashi et al. (2005, ITS), Masuya et al. (2009, ITS), and in the present study (Plate 21), with the conclusion that the genus represents two distinct evolutionary lineages. However, a phylogenetic analysis of three protein-coding genes (hsp60, g3pdh, cal) by Andrew et al. (2012) provided evidence for a supported monophyletic clade for Monilinia s.l., suggesting validity of the genus in the sense of Honey (1936) (1998), who followed a narrow concept of Sclerotinia. They found that the type species of Grovesinia, G. pyramidalis (= G. moricola), consistently contributed to the paraphyly of Sclerotinia in all of their ITS rDNA analyses, unless one ignores the multicellular diaspores of Grovesinia as a valuable morphological marker and includes Grovesinia in Sclerotinia.
Heterogeneity of Schroeteria in the current generic concept resulted from the distant position of S. poeltii (Plate 21). Although Veronica cymbalaria, the host plant of S. poeltii, is closely related to V. hederifolia agg. (Muñoz-Centeno et al. 2006), the chlamydospore chains of the S. poeltii anamorph are very different in shape compared to those of other Schroeteria spp. This and the deviating molecular result appear to justify separation of S. poeltii at some taxonomic level. In the morphology of the teleomorph, however, S. poeltii can hardly be distinguished from S. decaisneana, including the stromatized seeds. The specialized occurrence of Schroeteria s.str. and S. poeltii on seeds of the same host genus Veronica suggests a common ancestor which also grew as a biotrophic parasite of Veronica seeds, making a scenario of polyphyletic evolution quite improbable. In the absence of conclusive evidence, we refrain from postulating non-monophyly of Schroeteria and from implying highly convergent anamorph and teleomorph morphologies, host relationships, and life cycles. From these considerations, we think that a generic split of Schroeteria is premature at the moment. Multigene analyses based on protein-coding genes should be carried out to better understand the phylogenetic position of Schroeteria and in particular S. poeltii, comparable to the study by Andrew et al. (2012), which revealed for Monilinia a supported monophyletic relationship between the two distant subgroups.
The genetically most heterogeneous genus in our analyses was Ciboria. Its heterogeneity was also seen in analyses of, e.g. Galán et al. (2015: fig. 4, LSU, fig. 5, ITS), Pärtel et al. (2016: fig. 2, ITS), and Navaud et al. (2017, ITS). The genus is currently circumscribed by apothecia emerging from locally stromatized catkins, fruits, leaves, or wood and bark; the lack of both sclerotia and a macroconidial anamorph; an ectal excipulum of non-gelatinized textura globulosa; and comparatively small, hyaline ascospores with a low lipid content (Spooner 1987, Baral in Baral & Krieglsteiner 1985. However, generic concepts vary among authors. Based on the first author's personal observations, Ciboria can hardly be distinguished by teleomorph and microconidial anamorph morphology from various other genera of Sclerotiniaceae, such as Botrytis, Scleromitrula (= Ciborinia), and Sclerotinia s.l.
In most of our analyses (Plate 21, S2, S3), a supported core clade of Ciboria was formed, with two species growing on male catkins: C. caucus (Rebent.) Fuckel (type species, on Populus and Salix, Salicaceae) and C. amentacea (Balb.) With the present knowledge, the generic position of Ciboria ploettneriana could not satisfyingly be resolved, neither with morphological nor with molecular methods. In our combined analyses (Plate 21, S3), the species clustered unresolved within the Sclerotiniaceae though close to the type species of Dumontinia (D. tuberosa), Sclerotinia (S. sclerotiorum), and Stromatinia (S. rapulum), but also to those of Botrytis, Kohninia, and Myriosclerotinia. The distances to these taxa in the ITS region and LSU D1-D2 domain appear to be too low in order to resolve generic limits. Multigene analyses would probably better resolve phylogenetic lineages in this group.
Because of a high similarity in the teleomorphs, a taxonomically satisfying solution which does not strictly follow monophyletic principles but also includes morphological considerations is very difficult to achieve. Despite its low molecular distance to species of Sclerotinia s.l. (including Dumontinia, Stromatinia s.str., and perhaps Grovesinia) and a characteristic motif in the ITS1 region (see above), we here use the current combination Ciboria ploettneriana instead of Sclerotinia, where it was originally placed, because Sclerotinia and Dumontinia have been characterized by freely formed sclerotia which do not incorporate remnants of host tissue (Kohn 1979, pp. 377-378). Stromatinia forms an indefinite stroma comparable to C. ploettneriana and might be a suitable genus for this species. Nevertheless, BLAST search for the ITS region of C. ploettneriana yields species of Sclerotinia s.str. as closest match. Placement of C. ploettneriana in Sclerotinia s.l. is supported by very similar ascospores which contain 2-4 nuclei associated with comparatively small LBs, perhaps also by the lack of a macroconidial state which is only known in Grovesinia. However, VBs in the paraphyses, which are characteristic of Ciboria ploettneriana, have not been seen in other members of Sclerotinia s.l., but are typical of Botrytis (see IVV).
When Buchwald (1949, p. 165) proposed the combination Ciboria ploettneriana, he did not give arguments for doing so and also did not describe the fungus. Here and in his treatment of Danish Sclerotiniaceae (Buchwald 1947, pp. 240, 255), he distinguished two subgroups within Ciboria: subgenus "Euciboria Boud." for species on flowers and subgenus "Stromatinia Boud." for species on fruits. It must be noted that Buchwald (1947, p. 309) treated Stromatinia rapulum (type of Stromatinia) as a synonym of Sclerotinia tuberosa. Therefore, Buchwald (1949, p. 164) suggested to rename subgenus Stromatinia to subgenus Pseudociboria Buchw. (non Pseudociboria Kanouse). In his concept of subgenera, Buchwald followed Boudier (1885, p. 115) who considered the type of stromatization as taxonomically important and distinguished three subgenera within Ciboria (subgenus Sclerotinia with sclerotia, subgenus Stromatinia with a stroma, and subgenus Ciboria without stromatization). Buchwald accepted Sclerotinia as a distinct genus, and the stromatization in C. ploettneriana was obviously the reason why he included this species in Ciboria subgenus Stromatinia (in 1949 named Pseudociboria).
Because Ciboria ploettneriana clustered in our phylogenetic analysis near Sclerotinia spp. distant from all investigated Schroeteria spp. and because only one smutlike anamorph is known on Veronica hederifolia agg., it appears improbable that C. ploettneriana also possesses such an anamorph. Although S. poeltii did not group with the core clade of Schroeteria nor with any other clade in the family, we assume that all species with a smut-like anamorph parasitizing Veronica seeds evolved from a common ancestor on this host genus, whereas C. ploettneriana should have derived from another lineage of sclerotiniaceous fungi.