Proposal of a new family Pseudodiploösporeaceae fam. nov. (Hypocreales) based on phylogeny of Diploöspora longispora and Paecilomyces penicillatus

ABSTRACT During a field survey of cultivated Morchella mushroom diseases, Diploöspora longispora and Paecilomyces penicillatus, causal agents of pileus rot or white mould disease were detected, which resulted in up to 80% of yield losses. Multi-locus phylogenic analysis revealed that the fungi were affiliated in a distinct clade in Hypocreales. We further constructed a phylogenetic tree with broader sampling in Hypocreales and estimated the divergence times. The D. longispora and P. penicillatus clades were estimated to have diverged from Hypocreaceae around 129 MYA and Pseudodiploösporeaceae fam. nov is herein proposed to accommodate species in this clade. Two new genera, i.e. Pseudodiploöspora and Zelopaecilomyceswere, were introduced based on morphological characteristics and phylogenic relationships of Diploöspora longispora and Paecilomyces penicillatus, respectively. Five new combinations – Pseudodiploöspora cubensis, P. longispora, P. fungicola, P. zinniae, and Zelopaecilomyces penicillatus – were proposed.


Introduction
True morels (Morchella, Morchellaceae, Pezizales, Ascomycota) are one of the most popular edible mushrooms with a long history of consumption in Asia, Europe, and North America (Pilz et al. 2017;Liu et al. 2018). Because of their good taste, culinary qualities, and pharmacological performances in antitumor, anti-inflammatory, and antioxidant activities (Tietel and Masaphy 2018;Zhang et al. 2019), demands for morels have significantly increased in the market. In recent years, large-scale field cultivation of Morchella was successfully achieved in China . The morel cultivation area reached approximately 12,000 ha in the production season of 2021-2020 in China, with an economic value of over RMB 10 billion. However, with the rapid expansion of cultivation, diseases become a bottleneck for morel production, especially for diseases caused by fungi. Several common fungal diseases have been identified in the fruiting bodies of cultivated Morchella: stipe rot disease caused by the Fusarium incarnatum-F. equiseti species complex (Guo et al. 2016) and by Purpureocillium lilacinum (Masaphy, 2022), cobweb disease caused by Hypomyces/Cladobotryum species (Lan et al. 2020), white mould disease caused by Paecilomyces penicillatus (He et al. 2017) and pileus rot disease caused by Diploöspora longispora (He et al. 2018;Liu et al. 2018). Previous investigations showed that white mould diseases and pileus rot resulted in up to 80% of morel yield losses each year, which was attributed to a large number of conidia quickly spreading around the cultivation areas . However, the fungal pathogens were mainly identified based on sequence similarity of internal transcribed spacer gene region (ITS) but lacked convincing morphological evidence (Hyde et al. , 2018Liu et al. 2018). When did a blast of the ITS sequence of D. longispora, Tanney et al. (2015) also showed that D. longispora is most closely related to P. penicillatus including its ex-type (CBS 448.69).
The genus Diploöspora was established by Grove (1916) with Diploöspora rosea as the type species. This genus was characterised by producing chains of hyaline, cylindrical to fusiform, aseptate, or 1-3-septate conidia (Tanney et al. 2015). Currently, phylogenetic analysis of the partial sequences of small subunit (SSU) ribosomal RNA gene, internal transcribed spacers (ITS), and large subunit (SSU) ribosomal RNA gene reveals that D. rosea is an onygenalean fungus (Tanney et al. 2015). Diploöspora longispora was firstly isolated from a dead leaf of Colocasia esculenta var. antiquorum in Japan (Matsushima 1976). Two varieties of D. longispora are available, namely Diploöspora longispora var. longispora and Diploöspora longispora var. cubensis, and the latter was originally obtained from the fallen leaves of Leguminosae in Cuba (Castaneda 1987). Tanney et al. (2015) presented that D. longispora and its varieties were most closely related to P. penicillatus belonging to the order Hypocreales and reached affinity with Hypocreaceae. However, apart from conidial chains, there is little morphological similarity between P. penicillatus and D. longispora, namely the penicillate conidiophores of P. penicillatus with their basipetal conidiogenesis versus the branched conidiophores and acropetal conidiogenesis of D. longispora (Tanney et al. 2015).
The genus Paecilomyces was introduced by Bainier (1907) with Paecilomyces variotii as the type species (Samson 1974). This genus was featured by verticillate conidiophores with divergent whorls of phialides, having a cylindrical or inflated base tapering to a long and distinct neck and producing typically hyaline, one-celled conidia. Phylogenetic analysis based on 18S rDNA demonstrates that Paecilomyces is polyphyletic across two classes (Luangsa-Ard et al. 2004). The type species, P. variotii, and its thermophilic relatives were placed in Eurotiales (Eurotiomycetes), while mesophilic species are in Hypocreales (Sordariomycetes) (Luangsa-Ard et al. 2004, 2005. Paecilomyces penicillatus was introduced by Samson in 1974, which was first isolated from rotten mushrooms. Based on the morphological characteristics, it was placed in Eurotiales (Samson 1974). Based on the molecular phylogenetic analysis of 18S rRNA sequences and β-tubulin, P. penicillatus was transferred to the order Hypocreales (Sordariomycetes) and revealed an uncertain affinity with Hypocreaceae (Luangsa-ard et al. 2004(Luangsa-ard et al. , 2005; however, P. penicillatus is still placed in Paecilomyces sensu stricto. According to Hyde et al. (2020), this order comprises 14 families, including Bionectriaceae, Calcarisporiaceae, Clavicipitaceae, Cocoonihabitaceae, Cordycipitaceae, Flammocladiellaceae, Hypocreaceae, Myrotheciomycetaceae, Nectriaceae, Niessliaceae, Ophiocordycipitaceae, Sarocladiaceae, Stachybotryaceae, and Tilachlidiaceae. Recently, Polycephalomycetaceae was introduced for the accommodation of the fungicolous species from Ophiocordycipitaceae based on a concatenated matrix of six genetic markers (LSU, ITS, SSU, TEF, RPB1, and RPB2, personal communication). These hypocrealean fungi are mostly found as saprobes on decaying wood and in soil, pathogens or endophytes of plants, nematodes, and insects (Zhang et al. 2018, personal communication), as well as parasites on other fungi and lichens (Zhu and Zhuang 2013;Sun et al. 2019a). Generally, hypocrealean fungicolous taxa are the more serious, common pathogens of most cultivated mushrooms (Sun et al. 2019b). Diploöspora longispora and Paecilomyces penicillatus (Hypocreales) are recognised as the serious pathogens of cultivated Morchella, and previous studies indicated that their family ranks are uncertain (Luangsa-Ard et al. 2004;Luangsa-ard et al. 2005;Tanney et al. 2015). Additionally, Tanney et al. (2015) and our analyses presented that D. longispora and its variants were most closely related to several P. penicillatus strains (including CBS 448.69, the ex-type strain) based on an ITS BLAST query. Currently, multi-locus phylogenetic analyses and divergence time estimation have been used to clarify the higher ranking of fungal taxa (Hyde et al. , 2020. This study performed phylogenetic analysis and divergence time estimation using a concatenated matrix of five genetic markers (LSU, ITS, SSU, TEF, and RPB2). Based on the results, a new family, Pseudodiploösporeaceae, and two new genera Pseudodiploöspora and Zelopaecilomyces, are introduced to accommodate the fungal taxa misplaced under Diploöspora and Paecilomyces, respectively. Additionally, we introduce four combinations including the fungal pathogens causing pileus rot disease and white mould disease of cultivated Morchella.

Specimens, isolates, and morphological observation
Fresh specimens were collected along with the fruiting bodies of cultivated Morchella mushrooms in Kunming Yunnan province, Ankang, Shanxi province, Baoding, Hebei, province, Ningxia Hui Autonomous Region during the 2019 and 2021 cultivation seasons.
Fruiting bodies were examined from free-hand sections using a stereomicroscope. The conidia were picked and streaked on potato extract agar (PDA) and incubated at 25°C for 7 days. A Nikon Eclipse 80i light microscope, equipped with differential interference contrast (DIC) optics, was used to capture digital images. Tarosoft (R) v.0.9.7 Image Frame Work was used to measure the morphologic structures, and Adobe Photoshop CS6 Extended version 13.0.1 software (Adobe Systems, USA) to edit the photographic plates.
For observation by SEM, each patch (0.3 × 0.3 cm) of the fresh infected and un-infected M. sextelata was fixed in 2.5% glutaraldehyde in 0.05 M phosphatebuffered saline (BPS, pH 7.2) at 4°C. After 24 hours, the samples were washed with deionised 0.1 M PBS for 7 min three times, then dehydrated in graded ethanol (50%,70%, 80%, 95%) for 15 min, respectively. Subsequently, the samples were dehydrated in 100% ethanol for 15 min three times and dried in a fume hood using critical point dryers (Autosamdri® 931, Tousimis, MD, USA) with CO 2 . Finally, the samples were sputter-coated with gold by an ion sputter coater (ISC150, SuPro Instruments, Shenzhen, China) with a voltage of 110 V, a frequency of 50/60 Hz, and a current of 10 mA under vacuum of lower than 1-2 Pa for 60 s. The samples were loaded onto the SEM (SU8010, Hitachi, Tokyo, Japan) and observed.

Phylogenetic analysis
SeqMan Pro v. 7.1.0 (DNASTAR Lasergene) was used to trim the low-quality bases at both ends of the raw forward and reverse reads and to assemble them. The newly obtained sequences were queried against the nuclear database of NCBI. For species delimitation, the aligned ITS sequence matrix of 58 taxa including our isolates, Diploöspora, and available species of Paecilomyces and its allied fungi, as well as Alternaria species (outgroup taxa) were used to construct the phylogenetic tree. The SSU, ITS, LSU, RPB2, and TEF sequences of available generic type species and reprehensive of Hypocreales and representative species of all accepted Hypocaceae from recent studies (Sun et al. 2017) were employed for multi-locus phylogenetic analysis. Gelasinospora tetrasperma, Neurospora crassa and Sordaria fimicola were chosen as the outgroup taxa. The alignments were generated by using MAFFT version 7.03 with the Q-INS-I strategy (Katoh and Standley 2013). Conserved blocks were selected from the initial alignments with Gblocks 0.91 b (Castresana 2000). The best nucleotide substitution model for each gene was determined by using jModeltest2.1.1 (Darriba et al. 2012). GTR+G + I was estimated as the best-fit model for ITS; RPB2, TN93 + G was estimated as the best-fit model for SSU; and LSU, TN93 + G + I as the best-fit model for TEF-1α under the output strategy of BIC. The multi-locus phylogenetic analyses included 1403 characters for SSU, 607 characters for ITS, 893 characters for LSU, 1044 characters for RPB2, and 907 characters for TEF. All characters were weighted equally, and gaps were treated as missing characters.
Maximum likelihood (ML) analyses were performed by RAxML2.0 (Edler et al. 2021), using the GTR+GAMMA+I model. The maximum likelihood bootstrap proportions (MLBP) were determined using 1000 replicates. Bayesian inference (BI) analyses were conducted with MrBayes v3.2.7 (Ronquist et al. 2012). Metropolis-coupled Markov Chain Monte Carlo (MCMC) searches were calculated for 10,000,000 generations, sampling every 100th generation with the best-fit model for each gene. Two independent analyses with six chains each (one cold and five heated) were carried out until the average standard deviation of the split frequencies dropped below 0.01. The initial 25% of the generations of MCMC sampling were discarded as burn-in. The refinement of the phylogenetic tree was used for estimating Bayesian inference posterior probability (PP) values. The tree was viewed in FigTree v1.4 (Rambaut 2012), and values of maximum likelihood bootstrap proportion (MLBP) greater than 50% and Bayesian inference posterior probabilities (BIPP), greater than 95% at the nodes, are shown along branches.

Relative divergence time estimation
Molecular dating analysis was performed using BEAST v1.10.4 (Suchard et al. 2018). The aligned data were partitioned for each SSU, ITS, LSU, RPB2, and TEF1 dataset, and these were loaded to BEAUti v1.10.4. to prepare the XML file. The data partitions were set with unlinked substitution and clock models to independently estimate each gene partition. Taxa sets were developed for each calibration of the common ancestor nodes, associated with the most recent common ancestor (TMRCA). The Hypocreales crown with a normal distribution (mean = 216, SD = 27.5, with 97.5% of CI = 269 MYA). Calibration of the core Clavicipitaceae, using a normal distribution (mean = 133.7, SD = 20.8, with 97.5% of CI = 174.5 MYA). Calibration of the Ophiocordyceps crown, using an exponential distribution (offset = 100, mean = 27.5, with 97.5% CI of 200 MYA) (Samarakoon et al. 2016;Hyde et al. 2017). The Yule process tree prior was used to model the speciation of nodes in the topology with a randomly generated starting tree. The analyses were performed for 100 million generations, with sampling parameters every 1000 generations. The effective sample sizes were checked in Tracer v.1.7.2 and the acceptable values are higher than 200. The first 10,000 trees (10%) representing the burn-in phase were discarded based on Tracer v.1.7.2, and 90,000 trees were combined in LogCombiner v1.10.4. The maximum clade credibility (MCC) tree was given by summarised data and estimated in TreeAnnotator v1.10.4. The molecular dating tree was viewed in FigTree v1.4 (Rambaut 2012). In the MCC tree, node bars indicate 90% confidence intervals for the divergence time estimates.

Relative divergence time estimation
According to the divergence time estimates, the crown age of Hypocreales is around 206 (165-246) MYA (Figure 3)     Note: The genus Diploöspora was established by Grove (1916) with Diploöspora rosea as the type species. Phylogenetic evidence supported that D. rosea is an onygenalean fungus within Eurotiomycetes (Tanney et al. 2015). Diploöspora longispora and its two variants, D. longispora var. longispora and D. longispora var. cubensis, were isolated originally from the fallen leaves (Matsushima 1976;Castañeda 1987). Based on an ITS BLAST query, Tanney et al. (2015) proposed that D. longispora and its varieties belong to the order Hypocreales, and reached affinity with Hypocreaceae. While, our phylogenetic analysis based on the ITS sequence data also showed strains of D. longispora (UAMH 340, UAMH 6404, UAMH 6367, strain 60,319, and strain 60,320), and D. longispora var. cubensis (CBS 727.87, IMI 186962) grouped with strong support (MLBP/BIBP = 100%/1.00, Figure 1) in Sordariomyetes rather than in Eurotiomyetes. In our multi-locus phylogenetic tree, those taxa clustered in a distinct clade within Hypocreales but do not belong to Hypocreaceae (Figure 2), representing a new genus rank. Morphologically, despite those taxa and Diploöspora producing conidial chains, they are distinct from Diploöspora in acropetal conidiogenesis and the shape and size of conidia. Pseudodiploöspora is therefore introduced herein to accommodate those species misplaced in Diploöspora.  Figure 5 Synonym: Diploöspora longispora var. longispora Matsush., Icones Microfungorum a Matsushima lectorum: 61 (1975) Type Distribution: Japan, China Note: Diploöspora longispora was first isolated from a dead leaf of Colocasia esculenta var. antiquorum in Japan (Matsushima and Matsushima 1976). We introduce a new combination of Pseudodiploöspora longispora to accommodate D. longispora. Both analyses in Tanney et al. (2015) and this study suggested that Pseudodiploöspora longispora is most closely related to Paecilomyces penicillatus. However, the latter differs from Pseudodiploöspora longispora in the penicillate conidiophore, and basipetal conidiogenesis, as well as the shape and size of conidia (Figures 5 and 6). We did not treat Paecilomyces penicillatus as a synonym of Pseudodiploöspora longispora herein because of the great morphological differences. Pseudodiploöspora cubensis (R.F. Castañeda) Jing Z. Sun, X.Z. Liu & H.W. Liu, comb. nov  Substrate/Host: On fallen leaves of Leguminosae: Cuba (Castaneda 1987). On porcupine dung in a cave (Including UAMH 6367, UAMH 6404) (https:// www.uamh.ca/index.html) Distribution: Cuba Note: Pseudodiploöspora cubensis was originally obtained from the fallen leaves of Leguminosae in Cuba (Castaneda 1987). The ITS sequence of CBS 727.87 is 96% similar to Pseudodiploöspora longispora (identities, 514/537, gaps, 5/537). Additionally, regarding the ellipsoidal conidia of Pseudodiploöspora cubensis against cylindrical or ramoconidia of P. longispora, as well as the original isolation resource and location, we introduce a new combination of Pseudodiploöspora cubensis to accommodate D. longispora var. cubensis. There is no available sequence of Diploöspora zinnia, and we transfer this fungus to Pseudodiploöspora based on its sympodial, acropetal conidiogenesis, and the cylindrical-fusiform conidia (Matsushima 1981   submerged in the agar, single, ellipsoidal to pyriform, aseptate. Note: The genus Paecilomyces was introduced by Bainier (1907) with Paecilomyces variotii as the type species. The type species, P. variotii, and its thermophilic relatives were placed in Eurotiales (Eurotiomycetes), while entomopathogenic mesophilic species were placed in Hypocreales (Sordariomycetes) under the genus Isaria but did not include Paecilomyces penicillatus (Luangsa et al. 2004;Luangsa-Ard et al. 2005). Those taxa placed in Isaria were accepted in Samsoniella (Hypocreales, Cordycipitaceae) (Mongkolsamrit et al. 2018). Our phylogenetic analyses showed that Z. penicillatus (CBS 448.69, ex-type strain) was positioned in Pseudodiploösporeaceae (Figure 2). Despite a more than 99% similarity of the SSU (identities, 1589/1590, gap, 1/1590) and ITS sequence (identities, 502/505; gaps, 3/505 gaps) between Z. penicillatus (CBS 448.69) and Pseudodiploöspora loogispora, respectively. The ITS of Z. penicillatus is 4% different from Pseudodiploöspora cubensis (identities, 471/477; gaps, 6/477). Additionally, Z. penicillatus differs from both P. cubensis and P. loogispora in having penicillate conidiophores and basipetal conidiogenesis. Herein, we introduce Zelopaecilomyces for the accommodation of P. penicillatus based on its morphological distinctions.

Discussion
A combination of phylogenetic analyses and divergence time estimation has been widely used in solving the classification schemes and higher ranking of taxa ). According to this polyphasic approach, a large number of taxonomic positions of fungi have been refined (Hyde et al. 2020;He et al. 2022). Hyde et al. (2020) gave an update of Sordariomycetes based on phylogenetic analyses and divergence time estimation. According to their results, Hypocreales contained 14 families: Bionectriaceae, Calcarisporiaceae, Clavicipitaceae, Cocoonihabitaceae, Cordycipitaceae, Flammocladiellaceae, Hypocreaceae, Myrotheciomycetaceae, Nectriaceae, Niessliaceae, Ophiocordycipitaceae, Sarocladiaceae, Stachybotryaceae, and Tilachlidiaceae. Both our multilocus phylogeny and divergence time evidence reveal the proposed natural classification of Hypocreales. Multilocus phylogeny reals a family rank for Pseudodiploösporeaceae because its taxa formed a strongly supported and distinct clade sister to Hypocreaceae. Hyde et al. (2017) introduced a temporal banding for Ascomycota, and time ranges of 150-250 MYA and 50-150 MYA were recommended as the boundary for orders and families, respectively. Our MCC results presented that the crown age of Hypocreales is around 206 (165-246) MYA (Figure 3), which concurs with the previous results (Hyde et al. , 2020. Within Hypocreales, the divergence time for currently accepted families is within the range of 116-159 MYA suggesting that a family can at best be as young as 116 MYA in Hypocreales. Divergence time showed that the family Pseudodiploösporeaceae divorced from Hypocreaceae about 129 MYA, falling within the temporal band of families. Additionally, Polycephalomycetaceae was recently introduced as a new family based on a concatenated matrix of six genetic markers (SSU, ITS, LSU, RPB1, RPB2, and TEF) (personal communication), both our phylogenetic tree and MCC tree also support its family rank in Hypocreales herein. Vu et al. (Vu et al. 2019) proposed a taxonomic threshold predicted for filamentous fungal identification, and 88.5% similarity of ITS barcodes was suggested for family rank. A BLAST querying the ITS sequence of species from Pseudodiploösporeaceae presented less than 89% similarity against that species from Hypocreales, which also supported distinct family rank for Pseudodiploösporeaceae.
The taxonomic position of Diploöspora Grove was confirmed as a member of Eurotiomycetes by reexamination of its generic type species D. rosea (Tanney et al. 2015). Several species including Pseudodiploöspora longispora (previously known as Diploöspora longispora) and Pseudodiploöspora cubensis (previously known as Diploöspora longispora var. cubensis) placed previously in Diploöspora were shown an affinity for Hypocrealean fungi (Sordariomycetes) based on the phylogenetic analysis (Luangsa-Ard et al. 2004, 2005Tanney et al. 2015). Our phylogenetic analysis also supported that P. longispora and P. cubensis were more closely related to Hypocreaceae (Figure 1-2). In our multi-locus phylogenetic tree, those taxa clustered in a distinct clade within Hypocreales but were outside of the core Hypocreaceae (Figure 2), representing a new family and subsequent genera. Pseudodiploöspora is therefore introduced herein to accommodate those species misplaced in Diploöspora concerning the original nomenclature. Pseudodiploöspora is distinct from Diploöspora in having head-to-tail (acropetal) arrays of conidiogenesis against the latter of tail-to-head (basipetal) arrays of conidiogenesis. Additionally, the conidia of Pseudodiploöspora are longer but more slender than that of Diploöspora (Tanney et al. 2015). We introduce P. longispora and P. cubensis for accepting D. longispora and D. longispora var. cubensis regarding the 95.66% similarity of ITS sequence between P. cubensis (CSB 727.877) and other P. longispora isolates. Despite lacking molecular data on Diploöspora fungicola and Diploöspora zinnia, we enrolled them in Pseudodiploöspora according to the morphological features in the original description. Diploöspora coprophilia with phialides and producing subglobose conidia is unlikely to be related to Diploöspora rosea (Tanney et al. 2015) and Pseudodiploöspora longispora. Its taxonomic position needs to be further demonstrated. It was suggested that Diploöspora indica producing brown conidiophores may be better placed in Parapleurotheciopsis but not in Diploöspora (Tanney et al. 2015). We also excluded D. indica from Pseudodiploöspora in consideration of the brown conidiophore of the fungus.
Both analyses by Tanney et al. (2015) and this study presented that P. longispora is most closely related to Z. penicillatus (CBS 448.86) (Figure1-2). Vu et al. (2019) proposed a 99.6% similarity of ITS barcode for a species taxonomic threshold. When comparing the similarity of the ITS sequence, Z. penicillatus presented less than 98.63% similarity to that of P. longispora, and showed less than 94.3% similarity to that of P. cubensis (KT279809, CBS 727.87, ex-living type, previously known as Diploöspora longispora var. cubensis), respectively. There were no available EF1-α and RPB2 sequences in GenBank. We did not compare the similarity of EF1-α and RPB2 sequences. However, P. penicillatus differs from the latter in having penicillate conidiophores and basipetal conidiogenesis. Herein, we introduce a new genus, Zelopaecilomyces, for the accommodation of P. penicillatus.
The taxonomic position of Paecilomyces was revised and refined by phylogenetic analyses, habitats, host range, etc. (Luangsa-Ard et al. 2004, 2005. The entomopathogenic mesophilic species were placed in the class Sordariomycetes belonging to Hypocreales (Luangsa et al. 2004(Luangsa et al. , 2005. Generally, those taxa were placed in Isaria, which were accepted by a new genus Samsoniella (Hypocreales, Cordycipitaceae) currently (Mongkolsamrit et al. 2018). Our phylogenetic analyses presented that Z. penicillatus (CBS 448.69, ex-type strain) was positioned in Pseudodiploösporeaceae (Figure 2), which was owing to a higher similarity of the SSU and ITS sequence between P. penicillatus and D. longispora. However, previous phylogenetic studies have evidenced that the SSU and ITS sequence data alone are insufficient to provide good resolution in most of the groups in Sordariomycetes (Hyde et al. 2020). Morphologically, Z. penicillatus differs from P. longispora by penicillate conidiophore, basipetal conidiogenesis, and the shape and size of conidia ( Figure 5 and 6). Additionally, the divergence time revealed that Z. penicillatus diverged from P. longispora about 14 MYA (Figure 3). Tanney et al. (2015) thought that P. longispora and Z. penicillatus may be two extremes of a continuum, However, we treat P. penicillatus as a distinct species other than a synonym of D. longispora not only relying on morphological differences but also following the divergence time.
Both P. longispora and Z. penicillatus were originally isolated from decaying leaves and found on the rotten mushroom successively (Samson 1974;Matsushima 1976;Castaneda 1987). He et al. (2017) identified Z. penicillatus (as Paecilomyces penicillatus) as the causing agent of the white mould disease of cultivated Morchella only relying on ITS phylogenetic analysis but lacking morphological evidence. Liu et al. (2019) reported P. longispora (as D. longispora) infecting cultivated Morchella, resulting in pileus rot but not offered the typical morphological feature of P. longispora. The phylogenetic analyses in both Tanney et al. (2015) and this study revealed that ITS and SSU are unable to adequately distinguish D. longispora and Z. penicillatus, but our study offered robust morphological evidence on the taxonomy of P. longispora and Z. penicillatus. Since P. longispora has been reported as a serious fungal pathogen Liu et al. 2018), reliably taxonomic information will facilitate tracing the origin and understanding of pathogenesis.