Citizen science project reveals high diversity in Didymellaceae (Pleosporales, Dothideomycetes)

Abstract Fungal communities play a crucial role in maintaining the health of managed and natural soil environments, which directly or indirectly affect the properties of plants and other soil inhabitants. As part of a Citizen Science Project initiated by the Westerdijk Fungal Biodiversity Institute and the Utrecht University Museum, which aimed to describe novel fungal species from Dutch garden soil, the diversity of Didymellaceae, which is one of the largest families in the Dothideomycetes was investigated. A preliminary analysis of the ITS and LSU sequences from the obtained isolates allowed the identification of 148 strains belonging to the family. Based on a multi-locus phylogeny of a combined ITS, LSU, rpb2 and tub2 alignment, and morphological characteristics, 20 different species were identified in nine genera, namely Ascochyta, Calophoma, Didymella, Juxtiphoma, Nothophoma, Paraboeremia, Phomatodes, Stagonosporopsis, and Xenodidymella. Several isolates confirmed to be ubiquitous plant pathogens or endophytes were for the first time identified from soil, such as Ascochyta syringae, Calophoma clematidis-rectae, and Paraboeremia litseae. Furthermore, one new genus and 12 novel species were described from soil: Ascochyta benningiorumsp. nov., Didymella degraaffiaesp. nov., D. kooimaniorumsp. nov., Juxtiphoma kolkmaniorumsp. nov., Nothophoma brennandiaesp. nov., Paraboeremia rekkerisp. nov., P. truiniorumsp. nov., Stagonosporopsis stuijvenbergiisp. nov., S. weymaniaesp. nov., Vandijckomycella joseaegen. nov. et sp. nov., V. snoekiaesp. nov., and Xenodidymella weymaniaesp. nov. From the results of this study, soil was revealed to be a rich substrate for members of Didymellaceae, several of which were previously known only from diseased or apparently healthy plant hosts.


Introduction
Due to high plasticity and the capacity to adapt and survive in adverse or unfavourable conditions, fungi are exceedingly successful soil inhabitants (Frąc et al. 2018). The majority of the fungal species presently known can survive in, or directly adapt to, the soil environment (Bridge and Spooner 2001;Botha 2011). Soil-borne fungi play essential roles in nutrient cycling in terrestrial ecosystems and are able to break down all kinds of organic matter, decompose soil components or act as effective biosorbents of toxic metals, thereby helping to maintain soil health (Anderson and Domsch 1973;Bender et al. 2013;Rudgers et al. 2014;Tedersoo et al. 2014;Yang et al. 2017;Frąc et al. 2018). Soil fungal communities also form symbiotic associations with plants, thereby improving nutrient absorption (Voøíšková and Baldrian 2012). Most fungal taxa found in the soil are continuously present in the environment as harmless saprobic organisms, but some also play a negative role. For instance, plant pathogenic fungi in soil could infect seedlings or other plant tissues when conditions are suitable, resulting in significant damage (van Agtmaal et al. 2017). In addition, some fungi reside in soil in the form of propagules to survive in an unsuitable environment, posing a long-term threat to other inhabitants (Maryani et al. 2019).
Didymellaceae is a ubiquitous fungal family including saprobic, endophytic and pathogenic species (Aveskamp et al. 2008(Aveskamp et al. , 2010Marin-Felix et al. 2017). More than 50% of the species in this family have been reported as plant pathogens, causing great losses to a wide range of economic crops (Aveskamp et al. 2008). Other species are found in different substrates, including soil, air, and water or cyst nematodes (Dorenbosch 1970;Chen et al. 1996;Boerema et al. 2004;Aveskamp et al. 2010;Porras-Alfaro et al. 2011;Chen et al. 2015Chen et al. , 2017Grishkan 2018;, and even in some extreme environments such as deep-sea sediments, or soils in Antarctica, deserts, and karst caves (Ruisi et al. 2007;Li et al. 2016;Zhang et al. 2016aZhang et al. , 2016bZhang et al. , 2017Chen et al. 2017;Nagano et al. 2017;Grishkan 2018). Although recent research has suggested that the soil environment represents an important niche for the discovery of novel phoma-like species , van Agtmaal et al. 2017), very few studies have investigated the diversity of Didymellaceae in soil, which is a massive reservoir for plant, animal and human pathogens.
The first paper systematically investigating Didymellaceae species from soil was published by Dorenbosch (1970), who provided diagnostic characteristics and a usable identification method (keys) for nine ubiquitous phoma-like fungi from soil, including Pyrenochaeta acicola, Phoma chrysanthemicola, Ph. eupyrena, Ph. exigua, Ph. fimeti, Ph. glomerata, Ph. herbarum, Ph. medicaginis var. pinodella, and Ph. prunicola (names used at that time). Later, Boerema et al. (2004) and Domsch et al. (2007) illustrated several Didymellaceae species from soil and provided their ecological distributions. Since then, a few species have been reported sporadically, along with the research of root and seed diseases, but studies of Didymellaceae from soil are still rare, with even fewer describing new taxa from soil. Most species in previous studies have been reallocated to other genera in this family based on their DNA phylogeny (Chen et al. 2015Valenzuela-Lopez et al. 2018). To date, only approximately 30 species from eight genera in Didymellaceae have been recorded from soil, namely Ascochyta, Phoma, Didymella, Neodidymelliopsis, Epicoccum, Cumuliphoma, Ectophoma and Juxtiphoma (Dorenbosch 1970;Boerema et al. 2004;Domsch et al. 2007;Chen et al. 2017;Valenzuela-Lopez et al. 2018). Although most of the species documented from soil are plant-associated (pathogens and endophytes), some species, such as Ph. herbarum and J. eupyrena, are characterised as soil-borne (Dorenbosch 1970;Boerema et al. 2004).
Didymellaceae species from soil always produce diverse metabolites, some of which can be cytotoxic, including cytochalasin A and B, deoxaphomin, proxiphomin and tenuazonic acid (Bennett et al. 2018). Currently, most Didymellaceae species thus far found in the soil environment were originally described from plant substrates, such as leaves, seedlings, wood, stem bases or roots, some of which are even capable of wood decay (Boerema et al. 2004;Aveskamp et al. 2008Aveskamp et al. , 2010Chen et al. 2015). On the contrary, crops that are grown in close proximity to infected soil appear to be more contaminated, given that soil is a known source of plant pathogenic fungi (Paterson and Lima 2017). Besides, some species have also been reported to be opportunistic pathogens in animals and humans, such as J. eupyrena (= Phoma eupyrena) and Phoma herbarum (Bakerspigel et al. 1981;Tullio et al. 2010). Considering the potential threat and great losses caused by soil-borne pathogens, and the application in the biotechnological or pharmaceutical industries, knowledge of the diversity of Didymellaceae in soil is urgently needed to better understand the functions, interactions and ecosystem feedback of fungi in the terrestrial environment.
The present Citizen Science Project was initiated by the Westerdijk Fungal Biodiversity Institute (WI) and the Utrecht University Museum, aiming to investigate the diversity of fungi in Dutch garden soil collected by children in their home gardens from different regions in the Netherlands . During the course of this project thousands of isolates were obtained from 293 soil samples. Of these, 148 isolates were found to belong to Didymellaceae, and subsequently selected for study. The aim of the present study was to investigate the diversity of Didymellaceae from Dutch garden soil, describe and illustrate novel species, and compare them with known and related species.

Sampling and isolation
Protocols for the collection and processing of soil samples are described in Groenewald et al. (2018) and Giraldo et al. (2019). Isolates are maintained in the Johanna Wester-dijk (JW) working collection housed at the WI in Utrecht, the Netherlands. New and interesting strains were also deposited in the CBS fungal collection and holotypes in the fungarium at the WI, respectively.

Selection of Didymellaceae strains
A preliminary species identification of the strains was carried-out by a BLASTn search performed with each ITS and/or LSU sequence against the NCBI (http://blast.ncbi. nlm.nih.gov) and WI (http://www.westerdijkinstitute.nl/Collections) databases. The ITS and/or LSU sequences generated in this study with more than 98 % similarity with reference sequences for Didymellaceae were selected for further study (Table 1).

Sequence alignment and molecular phylogenetic analysis
To further study the phylogenetic relationships, reference sequences of Didymellaceae were downloaded from GenBank (Table 1). Sequences of single loci were aligned with MAFFT v.7 using default settings (Katoh et al. 2017), and manually edited in MEGA v.6.0 when necessary (Tamura et al. 2013). MrModeltest v.2.3 (Nylander 2004) was used to select the best-fit models of evolution for the four data partitions according to the Akaike information criterion. Bayesian inference (BI), maximum-likelihood (ML) and maximum parsimony (MP) methods were implemented for phylogenetic analysis of individual gene regions and the combined dataset. The multi-locus gene dataset was generated using SequenceMatrix v.1.8 (Vaidya et al. 2011).
Bayesian analyses were performed using MrBayes v.3.2.6 (Ronquist et al. 2012) as described by Chen et al. (2015). The burn-in fraction was set to 0.25, after which the 50 % majority rule consensus trees and posterior probability (PP) values were calculated. The ML analyses including 1000 bootstrap replicates were conducted using RAxML v.7.2.6 (Stamatakis and Alachiotis et al. 2010) as described by Chen et al. (2015). Statistical support for the branches was evaluated using a bootstrap analysis (BS) of 1000 replicates. MP analyses were carried out using PAUP v.4.0b10 (Swofford 2003) as described by Braun et al. (2018). Statistical support for the branches was evaluated using a bootstrap analysis (PBS) of 1000 replicates. Trees were visualised in FigTree v.1.4.0 (Rambaut 2014) and the layout was created in Adobe Illustrator. Alignments and phylogenetic trees derived from this study were uploaded to TreeBASE (www.treebase.org) and sequences deposited in GenBank (Table 1).

Morphological characterisation
Isolates of Didymellaceae were transferred to fresh oatmeal agar (OA), 2 % malt extract agar (MEA) and potato dextrose agar (PDA)  plates and incubated at 25 °C under near-ultraviolet (UV) light (12 h light/12 h dark) to induce sporulation. Colony diameters were measured after 7 d of incubation (Boerema et al. 2004), and macroscopic characters and colony colours were described after 14 days of incubation and rated according to the colour charts of Rayner (1970). Preparations were mounted in distilled water to study the micro-morphological structures of mature conidiomata, conidiogenous cells and conidia from OA cultures (Aveskamp et al. 2010;Chen et al. 2015). Morphological observations included the general characteristics of the conidiomata, shape, presence of mycelium/setae on conidiomata, number of ostioles, thickness and texture of the pycnidial wall, length and width of the conidiogenous cells and conidia. To study the pycnidial wall, sections of mature conidiomata were generated using a Leica CM 1900 freezing microtome (Aveskamp et al. 2010;Chen et al. 2015). Observations of micro-morphological characteristics were processed with a Nikon Eclipse 80i compound microscope with differential interference contrast (DIC) optics and a Nikon AZ100 dissecting microscope, both equipped with a Nikon DS-Ri2 high-definition colour digital camera (Nikon, Tokyo, Japan) using NIS-elements imaging software v.4.3. The NaOH spot test was carried out using a drop of concentrated NaOH to determine the secretion of metabolite E on OA cultures (Boerema et al. 2004). Morphological descriptions and taxonomic information for the new taxa were deposited in MycoBank (Crous et al. 2004).

Results
A total of 293 soil samples were analysed, and nearly 3000 fungal strains were obtained. Among them, 148 Didymellaceae isolates were identified from 89 different garden soil samples, representing several locations in the Netherlands (Table 1).

Phylogenetic identification
A multi-locus phylogeny comprising 325 strains, including the JW soil isolates and reference strains from GenBank, was used to infer the relationships among species in Didymellaceae ( Figure 1, Table 1). Coniothyrium palmarum (CBS 400.71) and Leptosphaeria doliolum (CBS 505.75) were used as outgroups. The final combined ITS, LSU, rpb2 and tub2 alignment comprised 2317 characters including gaps (500 for ITS; 859 for LSU; 602 for rpb2; 356 for tub2), of which 1563 characters were constant, 106 parsimony-uninformative, and 618 were parsimony-informative. For the Bayesian analysis, SYM+I+G was selected as the best-fit model for the ITS dataset, and GTR+I+G was selected as the best model for the LSU, tub2 and rpb2 datasets. The phylogenetic trees obtained with three analyses showed a similar topology and were congruent with each other, and only the ML tree is presented herein with BS, PP, and PBS values plotted on the branches (Figure 1). In the phylogenetic analysis, the 148 isolates from Dutch soil were distributed in 10 clades (Figure 1). The majority of the isolates clustered in Juxtiphoma (n=63) which were recovered from 48 soil samples and 28 cities, followed by Paraboeremia (n=61) from 29 soil samples and 19 cities. Other isolates belonged to Didymella spp.  Table 1).
In the Juxtiphoma clade species clustered in two lineages, one corresponding to J. eupyrena (77/1/-) and the other representing a potentially new species (100/1/99). In the Paraboeremia clade, the soil isolates clustered in P. putaminum (86/0.99/67) and P. litseae (98/1/97). However, 21 isolates were distributed in two different lineages (with 6 and 15 isolates, respectively) that were phylogenetically distant from other species, representing two potentially new taxa. The soil isolates belonging to Stagonosporopsis clustered in a clade (100/1/99) that was phylogenetically distant from the other species, representing two potentially new species. In Didymella, the species were distributed in D. macrostoma (100/1/100) and D. pomorum (100/1/100), while isolates JW 195004 and JW 27006 were placed in two different branches, representing two putative new species. In Ascochyta one isolate grouped with A. syringae (93/1/86), whereas three isolates grouped in a different clade distant from previously known species, representing a potentially new species (100/1/100). The other three isolates grouped together at the bottom of the tree in a distant unknown lineage, which is introduced herein as a new genus with two species (100/1/90). All the new taxa are introduced in the taxonomy section based on the phylogenetic analysis and supported by morphological data. Descriptions and illustrations of the new taxa are provided in the taxonomy section below.
Species in Ascochyta are commonly regarded as plant pathogens, especially of cereal crops and legumes (Davidson and Kimber 2007;Tivoli and Banniza 2007), and only a few species were reported from soil, namely A. fabae, A. lentis, A. pisi, A. rabiei (Gossen and Morrall 1986; Tivoli and Banniza 2007) and A. syringae in the current study.
Nevertheless, A. benningiorum is phylogenetically and morphologically distinct from these species (Figure 1; Chen et al. 2015). Etymology. degraaffiae refers to Janne de Graaff who collected the soil sample from which the ex-type strain was isolated.
Culture characteristics. Colonies after 7 d at 25 °C, on OA reaching 65-70 mm diam, aerial mycelium floccose, orange to olivaceous, margin regular; reverse black near the centre, pale grey towards the periphery. On MEA reaching 55-60 mm diam, aerial mycelium floccose, buff to pale olivaceous, with white mycelium pellet and radially furrowed zones near the centre, margin regular; reverse buff near the centre, olivaceous to yellow towards the periphery. On PDA reaching 50-55 mm diam, aerial mycelium floccose, concentric circles pale brown, pale olivaceous grey, dark olivaceous, honey, margin irregular; reverse black with a pale olivaceous edge. NaOH spot test negative on OA.

Didymella kooimaniorum
Culture characteristics. Colonies after 7 d at 25 °C, on OA reaching 55-60 mm diam, aerial mycelium floccose, pale smoke grey, pale brown towards periphery, abundant production of confluent pycnidia, margin regular; reverse pale olivaceous, with some olivaceous black zones. On MEA reaching 50-55 mm diam, aerial mycelium woolly, pale olivaceous grey, margin irregular; reverse buff near the centre, dark brown with orange edge. On PDA reaching 50-55 mm diam, aerial mycelium floccose, pale mouse grey with olivaceous edge, margin irregular; reverse dark brown with pale brown edge. NaOH spot test negative on OA.
Culture characteristics. Colonies after 7 d at 25 °C, on OA reaching 45-60 mm diam, aerial mycelium cottony to floccose, isabelline to olivaceous, margin irregular; reverse concolorous. On MEA reaching 45-55 mm diam, aerial mycelium cottony to floccose, smoke grey to pale olivaceous grey with white edge, margin entire; reverse buff to smoke grey near the centre, olivaceous black with buff edge. On PDA reaching 45-50 mm diam, aerial mycelium cottony to floccose, olivaceous buff, dull green to buff, margin irregular; reverse smoke grey near the centre, olivaceous black with buff edge. NaOH spot test negative on OA.
Culture characteristics. Colonies after 7 d at 25 °C, on OA reaching 50-55 mm diam, aerial mycelium scarce, spore mass with grease-like appearance, dark brick to sepia, cinnamon to the edge, abundant production of confluent pycnidia, margin en- tire; reverse concentric rings umber to cinnamon. On MEA reaching 47-50 mm diam, aerial mycelium scarce, spore mass with grease-like appearance, dark brick to sepia, cinnamon to the edge, abundant production of confluent pycnidia, margin entire; reverse concentric rings umber to cinnamon. On PDA reaching 50-55 mm diam, aerial mycelium moderate to scarce, cottony, buff, spore mass with grease-like appearance, dark brick, ochreous to the edge, margin entire; reverse concentric rings dark brick to cinnamon. NaOH spot test negative on OA.
Additional specimens examined. The Netherlands, Gelderland province, Culemborg, isolated from garden soil, Mar. 2017, R. Notes. Based on the phylogenetic analyses, P. truiniorum is represented by six isolates, forming a distinct lineage (Figure 1). Paraboeremia truiniorum is charac- terised by the dense mycelial outgrowths on its pycnidia. Both P. truiniorum and P. rekkeri are phylogenetically close to the well-known soil-borne species, P. putaminum. However, P. putaminum is distinguished from these two new species by producing smaller conidia (3.2-4.2 × 2-2.6 µm) with greenish guttules (Boerema et al 2004). Etymology. stuijvenbergii refers to Simon van Stuijvenberg, who collected the soil sample from which the ex-type strain was isolated.
Additional specimens examined. The Netherlands, Gelderland province, Arnhem, from garden soil, Mar. 2017, D. Peters, JW 14003;Utrecht province, Utrecht, from garden soil, Mar. 2017, N. Francisca, JW 44014;Utrecht, from garden soil, Mar. 2017, P. de Koff, JW 33021. Notes. Phylogenetically, S. stuijvenbergii is most closely related to S. weymaniae, another novel species collected from Dutch soil in this study ( Figure 1). However, S. stuijvenbergii is distinguishable from S. weymaniae by the colour and the size of its pycnidia, being brown and measuring 200-1000 × 195-930 µm in S. stuijvenbergii, whereas S. weymaniae produces whitish pycnidia, measuring 330-650 × 250-550 µm. Furthermore, S. weymaniae produces microconidia and chlamydospores, which were not observed in S. stuijvenbergii. Although there are several reports that Stagonosporopsis spp. could survive in soil for a short time (Vaghefi et al. 2016), this is the first record of a Stagonosporopsis species only known from soil (Domsch et al. 2007). Stagonosporopsis stuijvenbergii is represented by four strains isolated from different samples collected in Utrecht and Gelderland provinces. Etymology. weymaniae refers to Anna Weyman, who collected the soil sample from which the ex-type strain was isolated.

Notes. The new genus
Vandijckomycella is introduced to accommodate two new species isolated from soil samples which form an independent lineage in Didymellaceae, being clearly separated from other genera (Figure 1). Based on the phylogenetic analysis, V. joseae forms a distinct lineage which is distant from the nearest species V. snoekiae, and chiefly differs on tub2 and rpb2 sequences. Morphological differences between V. joseae and V. snoekiae are discussed under the latter species. Vandijckomycella joseae is characterised by producing pycnidia with longer whitish hyphal outgrowths, and with elongated necks. Etymology. snoekiae refers to Rana Marit Ida Snoek who collected the soil sample from which the ex-type strain was isolated.
Culture characteristics. Colonies after 7 d at 25 °C, on OA reaching 50-55 mm diam after 7 d, covered by floccose aerial mycelium, pink to grey, darker grey near the centre, margin regular; reverse black near the centre, yellow towards the periphery. On MEA reaching 50-55 mm diam, aerial mycelium floccose to cottony, buff with some mouse grey zones, margin regular; reverse orange with some radial yellow lines and some black zones. On PDA, reaching 45-50 mm diam, covered by floccose aerial mycelium, vinaceous grey to pale olivaceous, olivaceous grey near the centre, margin irregular; reverse buff to orange, black near the centre. NaOH spot test on OA: pale reddish discolouration.
Culture characteristics. Colonies after 7 d at 25 °C, on OA reaching 55-60 mm diam, aerial mycelium floccose near the centre, flat towards the periphery, pale olivaceous to whitish, black pycnidia visible near the centre, margin regular; reverse buff to salmon, pale olivaceous towards the periphery. On MEA reaching 40-45 mm diam, aerial mycelium felty, sectors with cottony mycelium, white, buff to pale olivaceous, margin regular; reverse yellow to orange, dark brown and pale grey near the centre. On PDA reaching 45-60 mm, aerial mycelium floccose, whitish in the centre, honey towards the periphery, margin regular; reverse concentric circles dark brown in centre, orange, yellow, buff towards the periphery. NaOH spot test negative on OA.

Discussion
During the present Citizen Science project which focused on Dutch soil fungi, numerous unknown species of filamentous and yeast fungi were described Groenewald et al. 2018;Giraldo et al. 2019). As part of the project, we focused on investigating species diversity of Didymellaceae from soil samples obtained in the Netherlands.
As one of the largest families in the fungal kingdom, at least 26 genera are accepted in Didymellaceae (Chen et al. 2015Valenzuela-Lopez et al. 2018) and more than 5400 species names are recorded in MycoBank to date (Crous et al. 2004), representing 4.2 % of the 120000 accepted fungal species. However, only around 30 ubiquitous species have been found in the soil environment as saprophytes, mainly in Ascochyta, Boeremia, Didymella, Epicoccum and Phoma (Boerema et al. 2004;Aveskamp et al. 2008Aveskamp et al. , 2010Chen et al. 2015Chen et al. , 2017. In our set of samples, we found 20 species distributed in 10 genera i.e. Ascochyta, Calophoma, Didymella, Juxtiphoma, Nothophoma, Paraboeremia, Phomatodes, Stagonosporopsis, Vandijckomycella and Xenodidymella. However, we did not find any species of Phoma and Epicoccum, probably due to the media used for primary isolation, and also because of the taxonomical changes that have been suffered by many species of both genera in recent years (Aveskamp et al. 2010, Chen et al. 2015 Paraboeremia and Juxtiphoma were the most dominant genera. Species of Paraboeremia are more common on plants than in soil, except for P. putaminum, which is regarded as a widespread soil-borne fungus isolated from the subterranean parts of various herbaceous and woody plants ( de Gruyter and Noordeloos 1992;Boerema et al. 2004). In the present study this species was the most abundant species, being recovered from 29 soil samples from 19 cities. Besides P. putaminum, one isolate was identified as P. litseae, which was previously only known on diseased leaves of Litsea from China (Jiang et al. 2016). In addition, two new species were described, namely P. rekkeri found in Gelderland, North Brabant, North Holland and Utrecht provinces and P. truiniorum found in South Holland and Utrecht provinces.
The second most abundant species was Juxtiphoma eupyrena. The monotypic genus Juxtiphoma was recently introduced to accommodate Phoma eupyrena ), a cosmopolitan soil-inhabiting fungus, which may cause dampingoff of seedlings of herbaceous and woody plants ( de Gruyter and Noordeloos 1992;Boerema et al. 2004;Morgan-Jones and Burch 1988), but was also reported as an opportunistic human pathogen (Bakerspigel et al. 1981). Furthermore, a new species was introduced in this genus as Juxtiphoma kolkmaniorum which includes 12 of our soil isolates (JW) and one strain (CBS 527.66) isolated from soil in a wheat field in Germany.
Among our isolates we found Phomatodes nebulosa, Didymella macrostoma and D. pomorum which are plurivorous and cosmopolitan species often isolated from soil (Boerema 1993;de Gruyter et al. 1993;Farr and Rossman 2019). Interestingly, we found two species identified as plant pathogens that had not been previously reported from soil, including Ascochyta syringae and Calophoma clematidis-rectae. Ascochyta syringae causes ascochyta blight of Lilac (Syringa vulgaris) in America, Australia and Europe (Farr and Rossman 2019), while Calophoma clematidis-rectae is known on Clematidis spp. in the Netherlands (Aveskamp et al. 2010). In addition, Stagonosporopsis is recognised mainly as a phytopathogenic genus on different plant hosts (Marin-Felix et al. 2019). However, we found two new species from soil, namely S. stuijvenbergii and S. weymaniae. Other new species described include A. benningiorum, D. degraaffiae, D. kooimaniorum, N. brennandiae, V. joseae, V. snoekiae, and X. weymaniae. These findings suggest that species of Didymellaceae are also widely distributed in soil. Previous studies have revealed that many pathogens survive in soil by producing resting bodies (Dorenbosch 1970;Aveskamp et al. 2008), such as A. pinodes (currently: Didymella pinodes) and Phoma medicaginis var. pinodella (currently: Didymella pinodella) that produce chlamydospores or brown, thick-walled, swollen hyphae associated with sporocarps, which allow these species to survive in the soil for several years after the decay of their host tissues (Tivoli and Banniza 2007). On the other hand, some harmless saprobes in this family have also been observed to switch from an opportunistic to pathogenic lifestyle once in contact with the appropriate host (Aveskamp et al. 2008). Therefore, it is probable that the described new taxa are dormant in soil, remaining able to infect hosts under favourable conditions, especially species from phytopathogenic genera such as S. stuijvenbergii, S. weymaniae, N. brennandiae and X. weymaniae. However, considering that soil is a dynamic and multifunctional system and that the fungal community and its distribution are closely related to various living organisms such as plants, animals and insects, it was difficult to establish whether the species found in this study were true soil inhabitants or transferred to the soil via external vectors (such as worms, nematodes, etc.). Whether these new taxa originate from other habitats, or could change to pathogenic or endophytic lifestyles given the right conditions, remains to be determined. Furthermore, as the soil ecosystem is very complex and each type of soil and location may possess its own unique species diversity, the true diversity of Didymellaceae and their role in soil remains to be elucidated.
Recently, additional research based on cultivation-independent and cultivationdependent methods has revealed that Didymellaceae species present in various soil environments are more diverse than one might have expected (Bell et al. 2014;Nallanchakravarthula et al. 2014;Li et al. 2016;Miao et al. 2016;Zhang et al. 2016aZhang et al. , 2016bChen et al. 2017;Nagano et al. 2017). Although recent high-throughput methods have detected a higher diversity of soil fungi compared with those based on culturedependent methods, it is not possible to identify these taxa to species or even to genus level, as ITS sequence data alone are insufficient for species delimitation in most fungal families including Didymellaceae. Therefore, cultivation-dependent methods are still indispensable in the investigation of true species diversity of Didymellaceae based on additional loci such as rpb2 and tub2 obtained from cultivated isolates.
In summary, results of our study revealed the presence of a large number of unknown species and even a novel genus in soil, illustrating that this substrate is an important source for the discovery of novel taxa, and demonstrating that species diversity of Didymellaceae in soil is considerably greater than current estimates.