Ophiostomatoid fungi associated with Ips subelongatus, including eight new species from northeastern China

Ips subelongatus is a major pest that infects larch plantations over large areas of northern and northeastern China. Ips species are closely associated with ophiostomatoid fungi that are morphologically well-adapted for dispersal by beetles. These associations result in important threat for coniferous forests worldwide. The aim of this study was to characterize the ophiostomatoid communities associated with I. subelongatus infesting Larix species and sympatric Pinus sylvestris var. mongolica in northeastern China forests. Morphological and multilocus phylogenetic approaches (based on six markers: ITS, LSU, 60S, β-tubulin, EF-1α, and CAL gene regions) allowed identifying 14 species of four genera (Ceratocystiopsis, Endoconidiophora, Leptographium and Ophiostoma). Eight species are showed to be new to science. Most strains resided in two Ophiostoma species complexes, viz. the O. clavatum and the O. ips complexes, all together accounting for 76.8% of all isolates. Ophiostoma hongxingense sp. nov., O. peniculi sp. nov., and O. subelongati sp. nov. (O. clavatum complex) and O. pseudobicolor sp. nov. (O. ips complex) were the four dominant species. The ophiostomatoid communities associated with larch bark beetles, I. cembrae and I. subelongatus, in Europe and Asia, China and Japan, also were compared. These comparisons showed distinct, specific assemblage patterns.


INTRODUCTION
Globalization has hastened the emergence of tree pests, prompting the urgent need for a global strategy to manage the vitally important issues of forest pests (Wingfield et al. 2015). Bark beetles (Curculionidae: Scolytinae) are phloem-boring insects, some of which are the primary pests responsible for considerable mortality of coniferous forests in the northern hemisphere (Raffa et al. 2015). In Eurasia, species of the bark beetle genera Ips, Tomicus, and Dendroctonus have received a great deal of attention because of the damage they cause to local forest ecosystems and/or to tree plantations (Miao et al. 2001, Grégoire & Evans 2004, Vega & Hofstetter 2014. Ips subelongatus is a widely distributed bark beetle species in east Asia, spanning over Japan, Korea, Northern China, Mongolia, and the Russian Far East. It infests numerous Larix species (Pinaceae) including L. gmelinii, L. olgensis, L. principis-rupprechtii, L. kaempferi, L. sibirica, and sometimes Pinus spp. In China, I. subelongatus mainly infects three allopatric larches (Yang et al. 2007); they are L. gmelinii in the Da Xing'an and Xiao Xing'an mountain ranges in the Inner Mongolia Autonomous Region and Heilongjiang Province, L. olgensis in southeastern Heilongjiang Province, the Chang Bai mountain range in Jilin and Liaoning Provinces, and L. principisrupprechtii in middle Inner Mongolia as well as Beijing, Hebei, and Shanxi Provinces. These larches constitute the main component of one of the largest forested area in northeastern China.
The Asian eight spined larch bark beetle has commonly been considered a secondary pest that mainly attacks dying trees or colonizes stock logs (Stauffer et al. 2001, Yamaoka et al. 1998). However, extensive infestations of larches with high insect density, high tree mortality rates, and subsequent forest decline have been noticed in these areas since the 1980s (Yin et al. 1984, Yu 1992, Zhang et al. 2007). Because of possible incidental introductions through the timber trade, I. subelongatus was presented in the EPPO alert list A2 as an important pest that is affecting coniferous trees in native regions and which represents a threat to non-native regions (EPPO 2005).
The association between beetles and fungi commonly plays an important role in the success of beetle colonization (Kirisits 2004). One of the most important beetle associated fungal groups are the ophiostomatoid fungi (Wingfield et al. 1993, Kirisits 2004. Ophiostomatoid fungi are an assemblage of species that share similar morphological and ecological traits. They belong mainly to the order Ophiostomatales (Sordariomycetidae, Sordariomycetes, Ascomycota), which includes the genera Ophiostoma, Leptographium, Sporothrix, Raffaelea, and Ceratocystiopsis, and to the order Microascales (Hypocreomycetidae, Sordariomycetes, Ascomycota), which includes Ceratocystis, Graphium and Endoconidiophora , De Beer et al. 2014, De Beer et al. 2016. These fungi are assumed to be closely associated with bark beetles because of their morphological and ecological characteristics (Kirisits 2004). Some of these fungi are known as trees pathogens [e.g. Ophiostoma ulmi and O. novo-ulmi causing the Dutch elm disease (De Hoog et al. 1974, Brasier 1991, Leptographium wageneri, responsible for the black root disease (Harrington & Cobb 1988), or Endoconidiophora fujiensis, which could kill mature larch trees (Yamaoka et al. 1998)], but the majority are blue stain agents of timber, causing economic and ecological losses to the forestry industry.
The primary goal of the present study is to fill this gap in the knowledge of the communities of ophiostomatoid fungi on larch species in northeastern China, based on an extensive field survey and using integrated morphological observations and multilocus DNA sequence data to characterize the species. The occurrences of the ophiostomatoid communities in larch forests are analyzed. In addition to species diversity, communities among European and Asian eight spined larch bark beetles are also compared.

Collection of samples and isolation of fungi
Fungi were isolated from adults of Ips subelongatus and their breeding galleries in Larix gmelinii, L. olgensis, L. principis-rupprechtii and, in some cases, in Pinus sylvestris var. mongolica during the beetle's second mass flight period, which is from July to August, at 20 locations in northeastern China, including the three provinces of Heilongjiang, Liaoning, and Jilin and the autonomous region of Inner Mongolia (Fig. 1), from year 2010 to 2017. At each sampling location, beetle infested bark areas were collected from three to five dying trees or stock logs. Adults of I. subelongatus and their galleries were placed individually in sterile Eppendorf tubes and envelope bags, respectively. These organisms were subsequently stored at 4°C until fungal isolation. Ten later-developed galleries in phloem and 10 adult beetles collected from these galleries at each location were used for fungal isolation. Galleries were disinfected for 1 min with 1.5% sodium hypochlorite, rinsed with sterile water three times, then cut into tissue pieces approximately 3 × 3 mm 2 in a laminar flow hood, and five pieces of each gallery were selected and transferred onto 2% malt extract agar (MEA, malt extract and agar: AoBoXing Company Ltd., Beijing, China; recipe: add 20 g malt extract and 20 g agar per 1000 mL water). Adult beetles were crushed on the surface of 2% MEA without superficial disinfection. After a period of incubation at 25°C in dark, all strains were purified by single-spore isolations and/or mycelium apex and routinely grown on 2% MEA. After an initial analysis of macro-and microscopical characteristics, representative strains of each morphotype were selected for further in-depth morphological, physiological, and molecular studies. All strains were deposited in the culture collection of the Chinese Academy of Forestry (CXY) ( Table 1). Representatives were also deposited at the China Forestry Culture Collection Centre (CFCC) and the Mycothèque of the Université Catholique de Louvain, Belgium (BCCM/MUCL) (Table 1).

Morphological and cultural studies
Morphological structures were observed and recorded using an Olympus BX51 microscope, Olympus SZX16 stereomicroscope, and Olympus DP70 digital camera (Olympus, Centre Valley, PA, USA). For the strains selected as holotypes, the lengths and widths of 30 reproductive structures per strain were measured. The average (mean), standard deviation (SD), minimum (min), and maximum (max) measurements are presented as the (min-) (mean − SD)-(mean + SD) (−max).
For growth rate studies, a 5 mm diameter agar plug was taken from an actively growing fungal colonies and placed in the centre of 90 mm diameter Petri plates containing 2% MEA. These cultures were then incubated in the dark at 5°C intervals from 5 to 40°C. There were five replicate plates of each strain at each temperature, and two orthogonal diameter measurements were recorded daily until the fastest-growing mycelium reached the edge of the MEA plate. Colony colors were described based on the color chart of Rayner (1970). All relevant data pertaining to type specimens were deposited into MycoBank (http://www.MycoBank.org/).

DNA extraction, amplification, and nucleotide sequencing
Prior to DNA extraction, the strains were grown on 2% MEA at 25°C for 5 to 7 days. The actively growing mycelium from one MEA plate per strain was scraped from the surface of the medium using a sterile scalpel and transferred to 1.5 μL Eppendorf tubes. DNA extractions and purification were conducted using an Invisorb Spin   (Stielow et al. 2015), Bt2a/Bt2b (Glass & Donaldson 1995), EF1F/EF2R (Jacobs et al. 2004) or EF2F (Marincowitz et al. 2015)/ EF2R, and CL2F/CL2R (Duong et al. 2012) or CL3F/ CL3R (Musvuugwa et al. 2015) were used for amplification of internal transcribed spacer regions 1 and 2 of the nuclear ribosomal DNA operon, including the 5.8S region (ITS), the nuclear ribosomal large subunit region (LSU), the partial 60S ribosomal protein RPL10 gene (60S), the β-tubulin gene region (βT), the transcription elongation factor-1α gene region (EF-1α), and the calmodulin gene region (CAL), respectively. The PCR assays were performed in 25 μL volumes (2.5 mM MgCl 2 , 1 × PCR buffer, 0.2 mM dNTP, 0.2 mM of each primer, and 2.5 U Taq polymerase enzyme). The PCR conditions for amplification of the ITS region were an initial denaturation step at 94°C for 3 min, followed by 35 cycles of 1 min at 94°C, 45 s at 55°C, and 1 min at 72°C, and then final chain elongation at 72°C for 8 min. The five other gene regions were amplified using a denaturation step at 95°C followed by 35 cycles under the same conditions as above, except that the annealing temperatures varied between 54 and 58°C depending on the primers used, and a final chain elongation at 72°C for 8 min. The PCR products were cleaned using a MSB Spin PCRapace Kit (250) (Invitek, Berlin, Germany) according to the manufacturer's instructions.
Sequencing reactions were performed using a CEQ DTCS Quick Start Kit (Beckman Coulter, Brea, CA, USA) according to the manufacturer's instructions with the same PCR primers as above. Nucleotide sequences were determined using a CEQ 2000 XL capillary automated sequencer (Beckman Coulter).

Phylogenetic analysis
Preliminary identifications of the strains were conducted using standard BLAST searches. Representative sequences with the highest similarity matching and type strain sequences of similar species were downloaded from GenBank. Alignments were constructed with the  (Katoh & Standley, 2013). The genus-level dataset was aligned using the FFT-NS-i strategy with a 200 PAM/k = 2 scoring matrix, a gap opening penalty of 1.53, and an offset value of 0.00 (Linnakoski et al. 2016). The species complex or group-level datasets consisted of closely-related DNA sequences and were thus aligned using the G-INS-i strategy with a 1 PAM/ k = 2 scoring matrix, a gap opening penalty of 1.53, and an offset value of 0.00 (Linnakoski et al. 2016). Datasets were compiled in Molecular Evolutionary Genetic Analyses (MEGA) 7.0 (Kumar et al. 2016). Phylogenetic analyses of the aligned sequences were conducted using the maximum parsimony (MP), maximum likelihood (ML), and Bayesian inference (BI) methods. PAUP* version 4.0b10 (Swofford 2003) was used for MP analysis, with gaps treated as a fifth base. One thousand bootstrap replicates were generated to estimate the branch node confidence, with max trees set to 200 and clades compatible with the 50% majority rule in the bootstrap consensus tree were retained. The analysis settings were as follows: tree bisection reconnection branch swapping, starting tree obtained via stepwise addition, steepest descent not in effect, and MulTrees effective.
ML phylogenetic analyses were conducted using RAxML-HPC v.8.2.3 (Stamatakis 2014) available in the CIPRES Science Gateway (Miller et al. 2010, http://www. phylo.org/); the GTR + G model of site substitution included estimation of Gamma-distributed rate heterogeneity and a proportion of invariant sites (Stamatakis 2006). ML bootstrap support values were estimated using 1000 bootstrap replicates.
For Bayesian analyses, the best substitution models for each data set were determined using the corrected Akaike Information Criterion (AICc) in jModelTest v. 2.1.7 (Darriba et al. 2012

Collection of samples and isolation of fungi
In total, 496 strains of ophiostomatoid fungi were obtained from the adult beetles and galleries. Growth rates, macro-and microscopical morphological features were used for preliminary identification. Standard nucleotide BLAST searches at GenBank were performed using the BT sequences of all strains for preliminary sorting and searching for affinities. Subsequently, 41 representative strains were selected for in depth morphological study and multi-locus phylogenetic analysis (Table 1).

Phylogenetic analysis
The three phylogenetic methods used resulted in similar topologies with slight variations of the statistical support for each of the individual sequence datasets. Phylograms obtained by ML are presented for all the individual datasets, with branch supports obtained from ML, MP, and BI analyses indicated. The best-fit evolutionary models selected by jModelTest v. The ITS sequences did not allow distinguishing closely related species in all cases but they enabled grouping strains into species complexes within Ophiostoma (Fig. 2). However, the partial DNA sequences of three protein-coding genes (βT, CAL, EF-1α, and combined) had sufficient internal information allowing identification of Ophiostoma to the species level (Figs. 3, 5, 6, 8, 10; Additional file 2: Figure S1, Additional file 3: Figure  S2, Additional file 4: Figure S3, Additional file 5: Figure  S4, Additional file 6: Figure S5, Additional file 7: Figure  S6). The LSU, 60S, or EF DNA sequences also were employed to identify the strains of other three genera (Ceratocystiopsis, Endoconidiophora, and Leptographium) ( Fig. 11; Additional file 8: Figure S7, Additional file 9: Figure S8).
Within Ophiostoma s.l., the ITS fragments were approximately 650 bp long. The ITS dataset included 88 entries representing 85 taxa and 663 positions (including gaps). Our strains nested in four species complexes in  (Fig. 2). Furthermore, one strains (representing one phylogenetic species) fall outside any currently recognized species complex but belong to the previously shown "Group A" (Chang et al. 2017 Our 11 representative strains within the O. piceae complex formed four independent well-supported terminal clades representing four phylogenetic species (Taxa 1, 5, 9, and 11) in combined datasets (βT + CAL + EF-1α) phylogenetic inferences (Fig. 3). These phylogenetic species were related to O. breviusculum, O. brunneum, and O. rufum (Jankowiak et al. 2019). Clades of taxa 1 and 5 are well-supported in phylogenetic analyses based on the βT, EF-1α, CAL, and combined datasets (Additional file 2: Figure S1, Additional file 3: Figure S2, Additional file 4: Figure Figure S3). Six representative strains of the O. minus complex were grouped into two independent well-supported clades of taxa 4 and 6 in the ITS based inferences (Fig. 4). The clade of taxon 4 was related to the O. minus Eurasian clade (Gorton & Webber 2000, Lu et al. 2009), and the clade of taxon 6 nested in the near vicinity of the O. olgensis and O. album clades. In the βT-based phylogram, the clade of taxon 4 separated into a well-supported subclade, which is a sister to a formerly defined Eurasian subclade within the O. minus clade (Fig. 5). Furthermore, the βT-based phylogram confirmed the strains of clade of taxon 6 represented O. olgensis (Fig. 5).
Ten representative strains belonging to the O. clavatum complex formed three independent terminal clades (Taxa 2, 7, and 10) with good support values in inferences based on the βT (Additional file 5: Figure S4), EF-1α (Additional file 6: Figure S5), CAL (Additional file 7: Figure S6), and combined dataset (βT + CAL + EF-1α) (Fig. 6). They were Six representative strains of taxon 8 in the O. ips complex were used in the analyses. Taxon 8 formed a distinct clade with good statistical support in ITS (Fig. 7) and βT (Fig. 8) analyses. It was closely related to O. bicolor.
Two strains of taxon 3 in Group A were used in the analyses. Taxon 3 formed an independent lineage with good support values in ITS and βT based phylograms (Figs. 9, 10) and was most closely related to O. saponiodorum and O. pallidulum.

Ceratocystiopsis
The LSU phylogenetic tree of the genus Ceratocystiopsis did not yield an independent clade for taxon 12 (Fig. 11), although its strains showed clear dissimilarity with C. pallidobrunnea in terms of the LSU sequence data. The lack of available reference data in GenBank for other genes for C. pallidobrunnea impeded any closer comparison. Therefore, this species is hitherto recorded as Ceratocystiopsis cf. pallidobrunnea.
Cultures: Colonies on 2% MEA at 25°C reaching 80 mm diam. in 10 d, initially hyaline, later becoming brown, mycelium superficial or sparsely aerial, the colonies edge thinning radially, synnemata and perithecia scattered in the centre. Optimal temperature for growth at 25°C, no growth observed at 5°C and 35°C.
Ecology: Isolated from Ips subelongatus infesting dying Larix gmelinii and stock log.
Distribution: Currently only known from the Inner Mongolia Autonomous Region, China.
Notes: Ophiostoma genhense and O. multisynnematum formed two distinct, well-supported clades within the O.
piceae complex (Fig. 3), in which they were closely related to O. breviusculum (Chung et al. 2006  Description: Sexual morph not observed.
Cultures: Colonies on 2% MEA at 25°C reaching 58 mm diam. in 5 d, initially hyaline, discoloring progressively to dark olivaceous from the centre of the colonies to the margin, the colonies edge thinning radially; mycelium superficial on the agar. Optimal temperature for growth at 30°C, no growth observed at 5°C and 40°C.
Cultures: Colonies on 2% MEA at 25°C reaching 78 mm in diameter in 10 d, initially hyaline, thinning radially toward the margin, later becoming dark olivaceous and massive synnemata arising in the centre; hyphae superficial, aerial mycelium sparse. Optimal temperature for growth at 25°C, no growth observed at 5°C and 35°C.
Ecology: Isolated from Ips subelongatus infesting dying Larix gmelinii and stock log.
Habitat: L. gmelinii pure plantation.  Cultures: Colonies on 2% MEA at 25°C reaching 75 mm diam. in 5 d, initially hyaline, the colonies edge thinning radially, becoming dark olivaceous in the centre; mycelium mostly superficial, sparsely aerial. Optimal temperature for growth at 30°C, no growth observed at 5°C and 40°C.
Cultures: Colonies on 2% MEA at 25°C, fast growing, reaching 80 mm diam. in 5 d, effuse, cottony, hyaline to white at first, becoming white gray or gray brown; hyphae submerged in agar with many aerial mycelium. Optimal temperature for growth at 30°C, no growth observed at 5°C or 40°C.   (Fig. 18) Etymology: The epithet subelongati (Latin) refers to the vector (Ips subelongatus) from which this fungus was isolated.
Diagnosis: Ophiostoma subelongati colonies gradually turned brownish grey from the centre, but the O. hongxingense colonies centre turned dark olivaceous.
Description: Sexual morph not observed.
Cultures: Colonies on 2% MEA at 25°C reaching 61 mm diam. in 5 d, initially hyaline, the colonies edge thins radially, then from the centre of the colonies to the periphery it becomes brownish grey and develops superficial mycelium on the agar. Optimal temperature for growth is 30°C; no growth observed at 5°C or 40°C.
Habitat: L. gmelinii or L. olgensis pure plantation.  (Fig. 19) Etymology: The epithet xinganense (Latin) refers to the Xing'an mountains from where this taxon was first isolated.
Cultures: Colonies on 2% MEA at 25°C reaching 75 mm diam. in 10 d, initially whitish gray, the colonies edge thinning radially; hyphae mostly superficial, sparsely aerial, synnemata developing abundantly in the colonies centre. Optimal temperature for growth at 25°C, no growth observed at 5°C and 40°C.
Ecology: Isolated from Ips subelongatus infesting dying Larix gmelinii and stock log.
Distribution: Currently only known from the Inner Mongolia Autonomous Region, China.
Notes: Ophiostoma xinganense is closely related to O. rufum and O. brunneum (Hausner et al. 2003, Jankowiak et al. 2019 (Fig. 3). Ophiostoma xinganense and O. Wang et al. IMA Fungus (2020) Table 2). Leptographium zhangii was previously reported to have been isolated only in Heilongjiang (L. gmelinii), but we also isolated it in Inner Mongolia (L. gmelinii). Meng et al. (2015) first determined that Endoconidiophora fujiensis extensively existed in three allopatric larch forests in northeast China, and we also isolated it from L. gmelinii in Heilongjiang (Table 2). In China, seven additional species of the O. clavatum complex have been recently described (Yin et al. 2016;Chang et al. 2017Chang et al. , 2019. Two of them are Ophiostoma shangrilae and O. poligraphi, which have been described based on isolates found in association with three bark beetles (viz. Ips shangrila and Dendroctonus micans infesting Picea purpurea; Polygraphus poligraphus and D. micans infesting P. crassifolia) from Qinghai province. The other four are Ophiostoma jiamusiensis, O. songshui, O. ainoae, and O. brunneolum, which have been described based on strains isolated from I. typographus infesting spruces in northeastern China (Yin et al. 2016;Chang et al. 2019). Ophiostoma brevipilosi was originally described from strains isolated from Tomicus brevipilosus infesting Pinus kesiya in Yunnan province (Chang et al. 2017). Ophiostoma hongxingense, O. peniculi, and O. subelongati are currently known from larch in two northeastern provinces in this study.
Ophiostoma rufum (Jankowiak et al. 2019), and three of the new species described here belong to the O. piceae complex (Harrington et al. 2001), which is mainly characterized by a synnematous, pesotum-like and sporothrix-like asexual state. Jankowiak et al. (2019) described O. rufum with a brownish orange to a rust brown colonies and a sporothrix-like asexual state. Our strains, however, differ in having whitish gray colonies and a hyalorhinocladiella-like asexual state (Additional file 10: Fig. S9). Whether these deviating characters are caused by intraspecific variation or different culture conditions remains unclear and needs to be further studied. To date, 10 species in the O. piceae complex have been recorded in China (Lu et al. 2009;Paciura et al. 2010a;Yin et al. 2016;Chang et al. 2017Chang et al. , 2019 (Yin et al. 2016, Chang et al. 2019. Previously, O. piceae and O. setosum were reported to associated with Larix, Pinus, and Tsuga in Jilin and Yunnan provinces (Lu et al. 2009, Paciura et al. 2010a, Chang et al. 2017). The four new species here described were all isolated from I subelongatus infesting L. gmelinii in Inner Mongolia.
Ophiostoma pseudobicolor forms part of the O. ips complex , in which it is related to O. bicolor, a species associated with various bark beetles in China (Chang et al. 2019), Japan (Yamaoka et al. 1997), and Europe (Upadhyay 1981, Linnakoski et al. 2010). These two species can be distinguished from each other by their genetic divergences, as evidenced by the phylogenetic analyses, but also by morphological data, such as the size of the ascocarps. Furthermore, their association with bark beetles and hosts affinities are differential too.
Two known species from the O. minus complex (Gorton et al. 2004), O. olgensis and O. minus, were also recorded in our study. Ophiostoma olgensis was first described from the northeastern China, associated with I. subelongatus (Wang et al. 2016); it was again observed, and the species might be common in these northeastern China larch ecosystems. Ophiostoma minus has a wide distribution in northern hemisphere pine forests (Gorton & Webber 2000, Lu et al. 2009, Wang et al. 2019. In a phylogenetic perspective, the O. minus lineage was subdivided into two clades, viz. a North American and an Eurasian clades, which are considered as two allopatric populations, each with a differential autecology as far as their host is concerned. The North American population is associated with Dendroctonus spp., whereas the Eurasian population is associated with various pine-infesting beetles (Gorton & Webber 2000, Lu et al. 2009). In a previous study of ophiostomatoid species associated with Tomicus species infesting pines in Yunnan, southwestern China, strains of O. minus formed a distinct, third clade, which was interpreted as a third allopatric population (Wang et al. 2019). In the current study, our strains of O. minus clustered together with Wang et al. IMA Fungus (2020) 11:3 the Eurasian population (Figs. 4,5) and not with the Yunnan population (Wang et al. 2019). The origin, worldwide dispersion, and insect relationships range of these populations still require further studies. Four Leptographium species have also been isolated from Ips subelongatus in northeast China to date (Paciura et al. 2010b, Liu et al. 2016. Leptographium zhangii, which was observed also in our study, has previously been collected from other parts of northeastern China (Liu et al. 2016), confirming its widespread occurrence in the region.
Ips subelongatus and I. cembrae have been long considered as a single species with a wide distribution range. Their fungal associates were also thought to be generally identical over the presumed beetle geographic distribution range (Wood & Bright 1992, Yamaoka et al. 1998, Stauffer et al. 2001. In particular, Endoconidiophora laricicola, a pioneer invader and the most virulent fungal associate, has also been considered as widespread fungus, following the distribution of the beetle (Yamaoka et al. 1998, Stauffer et al. 2001. However, accurate comparison of specimens of eight spined larch bark beetles from Europe and Asia showed two allopatric species, corresponding to I. cembrae and I. subelongatus (Stauffer et al. 2001). In parallel, Japanese strains of E. laricicola associated with I. subelongatus were shown by multigene phylogenetic inferences to also represent a distinct species, E. fujiensis (Marin et al. 2005). Another species in genus the Endoconidiophora, i.e., E. polonica, is associated with subspecies and/or distinct geographic populations of I. typographus was also shown to represent two distinct populations, that may have coevolved with the two allopatric populations of their beetle vector, I. typographus (Stauffer & Lakatos 2000, Marin et al. 2009). Beetle and fungus speciation seemed to occur concomitantly with dispersal.
In China, Meng et al. (2015) reported that E. fujiensis is widely distributed in three larch forests in northeastern China, forming a stable association with I. subelongatus under such ecological conditions. In the present study, E. fujiensis was again collected from this area, supporting the previous observations.
The pathogenicity of the Chinese strains of E. fujiensis was tested by inoculation on mature, both native and Fig. 20 Venn diagram showing overlaps of the ophiostomatoid fungal communities associated with Ips cembrae in Europe and I. subelongatus in China and Japan* species common to all three regions; ** species common to both China and Japan; *** species common to both China and Europe; **** species common to both Europe and Japan; # species identity confirmed by molecular data. Wang et al. IMA Fungus (2020)  introduced, larches in the field . In that study, Chinese strains of E. fujiensis caused limited necrotic areas (approx. 5 cm in length over 2 months) in the three native larches, whereas it caused necrosis of more than 70 cm in length (over 2 months) in Japanese larch (L. kaempferi) , results that are very similar to those of a previous report (Yamaoka et al. 1998). The Japanese larch was introduced and has been planted over very large areas of China, from northeastern to northwestern provinces (e.g., Gansu province) down to the southern provinces (e.g., Hubei) because of its rapid growth and stress resistance (China Flora Editorial Committee of Chinese Academy of Sciences 1978, Ma & Wang 1990, Zhu et al. 2015. Therefore, the introduction and extensive afforestation by Japanese larch in China needs careful consideration and reevaluation because of its high susceptibility to forest pathogens. The ophiostomatoid fungi associated with I. cembrae and I. subelongatus in Palaearctic larch forests have been investigated extensively and is well documented also in Europe and Japan (Additional file 1: Table S1 and references cited therein). Based on the data available, 54 species were identified as being associated with I. cembrae and I. subelongatus infesting larches. Hitherto, the highest species diversity was observed in European larch forests, followed by those in China and Japan, with 29, 21, and 12 species recorded locally, respectively (Fig. 20, Additional file 1: Table S1). However, this might still reflect incomplete survey in Eastern Asia, especially China. The direct comparison among the ophiostomatoid communities is difficult. Seven species have a known distribution range extending over two or three regions. Two species are shared by China and Japan, whereas four species are shared by China and Europe, and three by Europe and Japan (Fig. 20). The 47 other species are endemic to a single region. This high level of endemism might be explained by the endemism of both the beetle vector and larch species and by the wide geographical differences. A similar conclusion was drawn from comparisons among fungal assemblages associated with I. typographus (Chang et al. 2019).
Ophiostoma piceae is the only species shared among Europe, Japan, and China (Fig. 20). However, in Japan, the species associated with I. subelongatus were identified based solely on morphological characteristics, which is poorly informative in the O. piceae complex. This complex has recently been greatly enriched based on multilocus DNA sequence comparisons and further phylogenetic analyses (Linnakoski et al. 2010;Yin et al. 2016;Jankowiak et al. 2017;Chang et al. 2017Chang et al. , 2019. It seems clear that the identity of the Japanese O. piceae complex strains must be reevaluated with the aid of multiple gene sequences. Ophiostoma minus, O. rufum, and Graphium laricis occur both in Europe and northeastern China (Pashenova et al. 1995(Pashenova et al. , 2004Kirisits et al. 2000;Stauffer et al. 2001;Jacobs et al. 2003;Kirisits 2004;Jankowiak et al. 2007;Linnakoski et al. 2010;Liu et al. 2016;Jankowiak et al. 2019). Ophiostoma minus seems to be undergoing population differentiation or a speciation process (Figs. 4, 5, Wang et al. 2019). Ceratocystiopsis minuta and O. brunneo-ciliatum were reported as present in both Europe and Japan (Aoshima 1965;Redfern et al. 1987;Redfern 1989;Yamaoka et al. 1998Yamaoka et al. , 2009Kirisits et al. 2000;Stauffer et al. 2001;Kirisits 2004;Jankowiak et al. 2007;Jankowiak et al. 2017;Yamaoka 2017); however, the reports of their existence in Japan also relied only on morphological identification, and would require molecular confirmation. This is particularly pertinent for O. brunneo-ciliatum, which is one of more frequently reported species associated with I. subelongatus in Japan (Aoshima 1965;Yamaoka et al. 1998Yamaoka et al. , 2009Yamaoka 2017).
Endoconidiophora fujiensis is the only species that is extensively associated with I. subelongatus in northeastern Asian larch forests. A concatenate phylogenetic analysis showed a genetic differentiation within this species, much higher than the intraspecific various of two sibling species, E. polonica and E. laricicola . These findings are consistent with the possible differentiation of the beetle vector I. subelongatus (Zhang et al. 2007, Song et al. 2011, Chen et al. 2016).

CONCLUSIONS
The results of this study indicate a high diversity of ophiostomatoid species associated with I. subelongatus infestations of larch and pine forests in northeastern China. Fourteen species were identified, of which eight Ophiostoma species were new to science. The dominant species were O. peniculi, O. hongxingense, and O. subelongati in the O. clavatum complex and O. pseudobicolor in the O. ips complex. The comparisons among ophiostomatoid communities associated with I. subelongatus in China and Japan, and with I. cembrae in Europe showed distinct assemblage patterns. The difference between Asian and European communities might be reasonable due to huge geographical distance and quite different environments, but was unexpected for the difference between northeastern Chinese and Japanese communities. However, the conclusion still need to be confirmed though molecular identification on all species compositions. As a pioneer invader, E. fujiensis caused noticeable necrosis to Japanese larch (L. kaempferi) but seemed weakly virulent to the local larches . Therefore, the introduction and extensive afforestation of Japanese larch in China needs careful consideration and reevaluation because of its high susceptibility to this forest pathogen.