Rhagadodidymellopsis endocarpi gen. et sp. nov. and Arthopyrenia symbiotica (Dothideomyceta), two lichenicolous fungi growing on Endocarpon species

The lichenicolous fungus Rhagadodidymellopsis endocarpi (Dothideomyceta) growing on the thallus of the terricolous lichen Endocarpon pusillum is described from Spain and Australia as new to science. The new genus and species is compared with other taxa from the genera Didymellopsis and Zwackhiomyces (Xanthopyreniaceae, Collemopsidiales, Dothideomyceta), in particular with D. perigena, a species also having hyaline didymospores and also growing on Endocarpon. Rhagadodidymellopsis endocarpi is characterized by its almost completely superficial stromatic ascomata with a coarse and irregular surface, and an ascomatal wall of very irregular thickness, and ascospores smaller than those of D. perigena. We also compare the new species with other Endocarpon parasites, including Arthopyrenia symbiotica. This is a misunderstood species, originally described as Verrucaria symbiotica, which we also discuss in detail in this study.


Introduction
Lichenicolous fungi are a specialized group of fungi that develop on lichens in a relatively inconspicuous way, with lifestyles ranging from parasymbiotic to necrotrophic parasitism and generally showing high specificity for their host (Diederich 2000). There are about 2320 species of lichenicolous fungi, described mostly within the Ascomycota, with less than 5% of the species belonging to the Basidiomycota. Authors claim, however, that up to 3000-5000 lichenicolous species could eventually be described (Diederich et al. 2018).
During study of the lichen diversity of Cap de Creus Natural Park in north-eastern Spain (Fernández-Brime 2012), we collected a lichenicolous fungus that was relatively abundant on the squamules of Endocarpon pusillum. The fungus was characterized by having fissitunicate asci, a hamathecium formed by abundant, branched and anastomosing, thin interascal filaments and one-septate hyaline ascospores, similarly to species of Didymellopsis or Zwackhiomyces (Grube & Hafellner 1990). In these two genera, included in the family Xanthopyreniaceae (Collemopsidiales, Dothideomyceta; Pérez-Ortega et al. 2016), the ascomata are at least partially immersed in the host thallus, except for some Zwackhiomyces species that produce superficial ascomata (Calatayud et al. 2007) aggregated into a common stroma, and the ascomata wall thickness is uniform, sometimes becoming thicker towards the ostiole. Our specimens have consistently superficial ascomata with a characteristic coarse fissurate surface, grouped on a stromatic structure. In addition, the ascomatal wall is rather irregular and clearly thicker than the one in taxa from the above-mentioned genera. The latter macroscopic ascomatal features make this fungus resemble species of Rhagadostoma (Sordariomycetes), a genus distinguished by unitunicate asci and an evanescent hamathecium formed by thicker hyphae (Navarro-Rosinés & Hladun 1994;Navarro-Rosinés et al. 1999) and therefore most probably not related to our fungus.
Based on the unique ascomata wall characteristics of our fungus, and as these traits do not fit any known genus, we propose to describe it as a new species and accommodate it in a new genus, Rhagadodidymellopsis.
For comparison, we have included in this study other lichenicolous fungi with bitunicate asci and hyaline, uniseptate ascospores growing also on Endocarpon. One of these is Verrucaria symbiotica, described by Nylander (1885), which was later combined in Arthopyrenia (Zahlbruckner 1922). This species develops perithecia completely immersed in the thallus of the host and has a smooth ascomata wall; hence it cannot be confused with the new Rhagadodidymellopsis endocarpi.

Material and methods
The specimens were examined morphologically and anatomically and are preserved in BCN, GRZ, H and NY (Thiers 2017), and in the personal herbarium of Javier Etayo. Macroscopic characters were examined using an Olympus SZ60 dissecting microscope. Microscopic characters were studied using a Zeiss Axioskop compound microscope. Hand-cut sections were mounted in water and stained with lactophenol cotton blue to increase the contrast of structures. Lugol's iodine solution with or without pre-treatment with 10% potassium hydroxide (K) was used to detect amyloid structures (indicated as I and K/I, respectively). Measurements were made only on dead herbarium material. Measurements of ascoma wall cells, asci and ascospores were made on material mounted in water; the values are the extreme values after rejecting 10% of the highest and lowest values; the highest and the lowest values are given in parentheses. For ascospore measurements the average value is also provided, in italics. The total number of measurements (n) is also given. Drawings were prepared with the aid of a drawing tube fitted to the microscope. Photographs were taken with a Pixera PRO150ES and an Olympus SC30 camera.
Ecology and distribution. Up to date, only four collections from three localities of Rhagadodidymellopsis endocarpi are known. The type material was collected in north-eastern Catalonia (Spain) from several specimens of E. pusillum growing on siliceous clay soil in sun-exposed areas. Unlike other species living in the same locality, such as Epiphloea terrena and Gyalideopsis athalloides, Endocarpon pusillum is not ephemeral and does not become inconspicuous during dry periods. The other two samples were collected in Canberra (Australia), and also grow on typical E. pusillum.
Notes. In the genera Didymellopsis and Zwackhiomyces, the ascomata are usually produced solitarily, and the ascomatal wall thickness is uniform or slightly thicker towards the ostiole. Rhagadodidymellopsis endocarpi, in contrast, has the ascomata mostly grouped in stromata, and the ascomatal wall varies in thickness around the grouped pseudothecia, resulting in a markedly rugose excipular surface.
Didymellopsis perigena also grows on Verrucariaceae species with a squamulose thallus. This species differs from R. endocarpi by having solitary, non-stromatic perithecioid ascomata (150-240 µm in diam.; fide Grube & Hafellner 1990), an ascomal wall of rather constant thickness, 30-65 µm wide, longer ascospores, 18-25 × 6-8.5 µm, with an average length/width value of ~3, and by growing mostly on the squamule margins of Placidium squamulosum (Grube & Hafellner 1990;Khodosovtsev & Klymenko 2015; see Table 1). Grube & Hafellner (1990) mentioned that D. perigena could grow not only on Placidium but also on Endocarpon, based on a record of 'Didymella' perigena from Nice (France) cited in Vouaux (1913). In Vouaux's study, the presence of flattened ascomata and ascospores of 15-21 × 6-8 µm are mentioned. These features are fairly similar to those observed in R. endocarpi. However, it is also stated that the ascomata grow in the margin of the lichen squamules, which corresponds to the typical growth of D. perigena on the host thallus. Unfortunately, we were not able to locate Vouaux's specimen and therefore cannot determine whether the 'Didymella' perigena specimen described by Vouaux (1913) belongs to R. endocarpi or to D. perigena.  More recently, Yazıcı & Etayo (2015) reported D. perigena in Turkey, growing on Endocarpon cf. pusillum. We revised this Turkish specimen and it matches well the description of D. perigena. There is a further citation of D. perigena from Cabo Verde (van den Boom 2012), potentially growing on E. pusillum, but as the author did not include morphological data regarding the lichenicolous fungus we cannot determine whether it corresponds to the newly described R. endocarpi or to D. perigena.
An additional ecological observation is that the irregularities on the surface of the stromata of Rhagadodidymellopsis endocarpi are always colonized by cyanobacteria, while no cyanobacteria are observed on lichen squamules devoid of the lichenicolous fungus. Based on this observation, we speculate that R. endocarpi might have a habit similar to the one mentioned by Grube & Hafellner (1990) in D. perigena: these authors hypothesized that as D. perigena was the only species from the genus not growing on cyanolichens, it could establish symbiosis with cyanobacteria accumulated at the base of the ascomata and using the lichen only as a mere substrate.
Notes. Arthopyrenia symbiotica is a barely known and misunderstood species growing on Endocarpon pallidum, for which we studied here the type specimen collected close to Amélie-les-Bains in southern France (Nylander 1985(Nylander , 1891. Besides the type material, we studied a more recent collection of A. symbiotica from Ablitas (northern Spain) growing on Endocarpon loscosii. This collection was initially identified as Didymellopsis perigena by Etayo (2008).
Didymellopsis perigena and Arthopyrenia symbiotica are taxa with very similar asci and ascospores, but they can be distinguished, as the first has globose and smooth-walled ascomata that grow almost completely superficially, mainly located at the margin of the squamules of the lichen host, while the latter has subglobose to pyriform ascomata that grow completely immersed in the host squamules.
Didymella pulposi var. garovaglii (Vouaux 1913) was considered a synonym of A. symbiotica by Roux et coll. (2017), and we follow this criterion here. The type material of this variety was also collected in southern France, close to Béziers, but growing on E. pusillum (Roux et coll. 2017). The description by Vouaux (1913) seems to be compatible with the characteristics we observed in the type material of A. symbiotica: ascomata growing completely immersed in the host thallus, 4-spored asci and ascospores 15-25 × 7-10 µm, slightly wider than in our studied specimens. Due to the immersed position of the ascomata stated by Vouaux (1913), it seems feasible that Didymella pulposi var. garovaglii is in fact A. symbiotica. We cannot fully confirm this synonymy, however, as none of the Didymella pulposi var. garovaglii samples were found in the Vouaux Herbarium (MARSSJ).
Lichenochora hypanica (Kondratyuk et al. 2014) is another species growing on Endocarpon and with uniseptate, hyaline ascospores. Based on the characters provided in the original description, the morphology of L. hypanica does not correspond to that of the genus Lichenochora. Among other traits, Lichenochora has unitunicate asci and thick, irregular paraphyses that persist after the asci mature (Etayo & Navarro-Rosinés 2008). However, in Kondratyuk et al. (2014) there is no reference to the ascus type and it is stated that interascal filaments are visible only in young ascomata. Lichenochora hypanica is similar to A. symbiotica in having the ascomata completely immersed in the host thallus, although it differs in having an evanescent hamathecium and smaller spores (Table 1). We have not been able to see the type material of L. hypanica, so its revision in relation to the taxa treated in this study must wait.
Artopyrenia symbiotica is then provisionally maintained in its current genus until further material can be studied.