Ophiostomatoid fungi associated with pine bark beetles and infested pines in south-eastern Australia, including Graphilbum ipis-grandicollis sp. nov.

The ophiostomatoid fungi are an assemblage of ascomycetes which are arguably best-known for their associations with bark and ambrosia beetles (Curculonidae) and blue stain (sap stain) of many economically important tree species. These fungi are considered a significant threat to coniferous forests, which has resulted in numerous studies characterising the diversity of bark beetles and their ophiostomatoid associates globally. The diversity of ophiostomatoid fungi present in Australian pine plantations, however, remains largely undetermined. The aims of this study were therefore to reconsider the diversity of ophiostomatoid fungi associated with Pinus in Australia, and to establish the baseline of expected taxa found within these plantation ecosystems. To achieve this, we reviewed Australian plant pathogen reference collections, and analysed samples collected during forest health surveillance programs from the major pine growing regions in south-eastern Australia. In total, 135 ophiostomatoid isolates (15 from reference collections and 120 collected during the current study) were assessed using morphological identification and ITS screening which putatively distinguished 15 taxonomic groups. Whole genome sequencing (WGS) of representative isolates from each taxon was performed to obtain high-quality sequence data for multi-locus phylogenetic analysis. Our results revealed a greater than expected diversity, expanding the status of ophiostomatoid fungi associated with Pinus in Australia to include 14 species from six genera in the Ophiostomatales and a single species residing in the Microascales. While most of these were already known to science, our study includes seven first records for Australia and the description of one new species, Graphilbum ipis-grandicollis sp. nov.. This study also provides an early example of whole genome sequencing (WGS) approaches replacing traditional PCR-based methods for taxonomic surveys. This not only allowed for robust multi-locus sequence extraction during taxonomic assessment, but also permitted the rapid establishment of a curated genomic database for ophiostomatoid fungi which will continue to aid in the development of improved diagnostic resources and capabilities for Australian biosecurity. Supplementary Information The online version contains supplementary material available at 10.1186/s43008-021-00076-w.


Introduction
Fungi within Ophiostomatales and Microascales are best known for their associations with arthropod vectors and include examples of some of the most devastating fungalinsect symbioses known to plant pathologists over the past century (Fisher et al. 2012;Wingfield et al. 2017b;Brasier and Webber 2019). Notable examples include the Dutch elm disease pathogens, Ophiostoma ulmi and O. novo-ulmi (Santini and Faccoli 2015;Brasier and Webber 2019), the laurel wilt pathogen Raffaelea lauricola (Harrington et al. 2008) as well as the numerous pathogens belonging to Ceratocystis which cause tree mortality in natural and agricultural ecosystems (Roux et al. 2007;Wingfield et al. 2017b;Tsopelas et al. 2017). In a recent review on novel associations for members of Ophiostomatales and Microascales, Wingfield et al. (2017b) highlight the numerous biological and anthropogenic factors that influence the dispersal of these fungi and their vectors globally; a major feature of the ever-increasing threat these fungi pose to global biosecurity.
Despite being formally recognised as two distinct orders in the Sordariomycetes, species belonging to Ophiostomatales and Microascales share a long and complicated taxonomic history and are collectively referred to as the ophiostomatoid fungi (Wingfield et al. 1993;Seifert et al. 2013). This is due to similarities shared across their biology, particularly in key morphological characters, that is believed to have been driven by convergent evolution in adaptation to insect-mediated dispersal Wingfield et al. 2017b). Ophiostomatoid fungi are commonly associated with bark (Coleoptera: Scolytinae) and ambrosia (Curculonidae: Scolytinae, Platypodinae) beetles (Kirisits 2004;Hofstetter et al. 2015), where a greater dependency and specificity is apparent for Ophiostomatales compared to Microascales (Wingfield et al. 2017b). Ophiostomatoid genera that are most commonly associated with beetles include: Ambrosiella, Endoconidiophora and Graphium in Microascales; and Affroraffaelea, Aureovirgo, Ceratocystiopsis, Fragosphaeria, Graphilbum, Leptographium, Ophiostoma, Raffaelea, and Sporothrix of Ophiostomatales (Hyde et al. 2020).
While not all ophiostomatoid fungi are responsible for tree mortality, many are well recognized as the causal agents of blue stain (sap stain) in the wood of economically important tree hosts (Kirisits 2004;Seifert et al. 2013). This is particularly true for pine (Pinus) plantations globally de Errasti et al. 2018;Jankowiak et al. 2021). Systematic surveys of bark beetles and ophiostomatoid fungi associated with pine have been completed in North and Central America (Zhou et al. 2004a;Kim et al. 2011;Klepzig and Hofstetter 2011;Taerum et al. 2013;Marincowitz et al. 2020), Europe (Linnakoski et al. 2012;Romón et al. 2014;Jankowiak et al. 2012Jankowiak et al. , 2020, Asia (Zhou et al. 2013;Masuya et al. 2013;Kirisits et al. 2013), with a significant number of surveys conducted recently in China (Chang et al. 2017Wang et al. 2018Wang et al. , 2019Wang et al. , 2020, South America (Zhou et al. 2004b;de Errasti et al. 2018) and New Zealand (Thwaites et al. 2005(Thwaites et al. , 2013. The diversity of ophiostomatoid fungi present in exotic pine plantations in Australia, however, remains largely undetermined. Since its first detection in the 1960s (Vaartaja 1967), Ophiostoma ips has been regarded as the most common fungal species associated with blue stain and pine bark beetles (specifically Ips grandicollis) in Australia Simpson 1987, 1990;Hood and Ramsden 1997;Zhou et al. 2007;. Additionally, surveys of the fungal associates of Ips grandicollis on Pinus taeda and P. elliottii in New South Wales (NSW) in the late 1980s serve as the first reports of a Ceratocystiopsis and Graphilbum species detected in Australian pine plantations Simpson 1987, 1990), while Grosmannia huntii was first reported in NSW in 1998 (Jacobs et al. 1998). To date, these serve as the few detailed surveys of fungi associated with Australian pine bark beetles. Other, somewhat incidental records include the detection of Ophiostoma floccosum, O. quercus and an unknown species reported as a Pesotum aff. fragrans, all isolated from woodchips of P. radiata from the Tantanoola paper mill in South Australia (Harrington et al. 2001;Thwaites et al. 2005). Evidently, the historical record of ophiostomatoid fungi in Australian pine plantations has relied heavily on morphology, and/or the association of blue stain in the presence of the pine bark beetle, I. grandicollis (Carnegie and Nahrung 2019;. Interestingly, the introduction of I. grandicollis in 1943 coincides with the introduction of two other exotic pine bark beetles, namely Hylastes ater and Hylurgus ligniperda in 1936 and 1942, respectively (Nahrung et al. 2016). Both are known to also vector ophiostomatoid fungi (Kim et al. 2011;de Errasti et al. 2018). The above-mentioned pine bark beetles, along with their associated ophiostomatoid fungi, are considered as established exotics to Australia.
Recent efforts to improve on the capacity of forest biosecurity surveillance, through programs such as the forest health surveillance program, and the more targeted highrisk site surveillance program (Carnegie et al. 2018), has led to several detections of cryptic fungal species associated with pine bark beetles and blue stain in NSW (Carnegie and Nahrung 2019). This includes the recent pest detections of Graphilbum fragrans, O. angusticollis, O. pallidulum and Sporothrix cf. abietina, illustrating the value of targeted surveillance programs for the detection of novel pests . These findings also emphasize the need for an updated record of the diversity of established ophiostomatoid fungi associated with Australian pine and pine bark beetles.
The overall aim of this study was to reconsider the diversity of ophiostomatoid fungi associated with pine and pine bark beetles in south-eastern Australia. In order to achieve this, we looked to: (1) review all available ophiostomatoid reference material previously reported from pine and lodged in Australian plant pathogen reference collections; (2) survey the ophiostomatoid fungi found in pine plantations during the 2019-2020 forest health surveillance period; and (3) use whole genome sequencing (WGS) of representative taxa to establish a curated database for improved molecular diagnostics of ophiostomatoid fungi for Australian biosecurity.

Literature and Australian plant pathogen reference collection review
Ophiostomatoid fungi previously collected from Pinus spp. in Australia were included as references in this study. Living cultures were recovered from the Victorian Plant Pathology Herbarium (VPRI) and the New South Wales (NSW) Plant Pathology and Mycology Herbarium (DAR) following database searches using the currently accepted nomenclature ) and all putative synonyms (MycoBank Database, www. mycob ank. org; Species Fungorum, www. speci esfun gorum. org) of ophiostomatoid fungi that were recorded in the respective Australian collections and associated with Pinus. Additionally, a literature and GenBank database search (http:// www. ncbi. nlm. nih. gov) was performed for Australian specimens previously reported from Pinus in order to identify additional specimens that had publicly available DNA sequence data.

Sample collection during forest health surveillance
Annual forest health surveillance programs are conducted in pine plantations across Australia, including NSW (Carnegie et al. 2008), Victoria (Smith et al. 2008), Tasmania (Wotherspoon 2008), and South Australia (Phillips 2008). These surveillance programmes capture a broad overview of plantation health, achieved through aerial and ground surveys across the major growing regions for each state. Taking advantage of this routine surveillance, sampling was concentrated on pine trees showing typical symptoms of bark beetle infestation, which included any dead or dying trees, but also tree stumps in recently harvested sections. Samples were either collected and sent in by respective state agencies conducting the surveillance, or by the first author accompanying forest health surveillance. Samples collected from May 2019 to March 2020 originated from 40 locations, including collections from NSW (n = 34), Victoria (n = 2), Tasmania (n = 2), and South Australia (n = 2) (Additional file 1: Table S1). Samples of sapwood and/or pieces of bark containing beetle galleries were collected and individually placed into sampling bags to retain moisture. Where possible, pine bark beetles were collected into 50 ml collection vials on site using forceps and submitted along with their respective wood samples.
Finally, all wood submissions were screened upon arrival in the laboratory for any remaining beetles that may have been concealed within the galleries. Pine bark beetles present in each sample were sorted into morphospecies, pooled and then treated as a single submission (representative specimens were morphologically identified by Crop Health Services diagnostics unit, Agriculture Victoria). All samples were stored at 4 °C until they were processed for fungal isolations.

Fungal isolations
Fungal isolations from beetle galleries were performed by directly transferring aerial mycelia and/or spore masses found on sporing structures characteristic of ophiostomatoid fungi, such as ascomata or synnemata, onto malt extract agar (MEA; Oxoid MEA as per manufacturer instructions; Oxoid, Basingstoke, UK) amended with 0.1 g Tetracycline (Fluka Analytical, Sigma-Aldrich, MO, USA) per 1000 ml of media. When sporing structures were absent, samples were incubated in moistened plastic containers at room temperature for approximately 21 days to encourage sporulation. When blue stained sapwood was present in a sample, wood chips of approximately 5 × 5 mm were cut, surface sterilized with 1.5% sodium hypochlorite for 1 min, and plated onto MEA. Beetle isolations followed an amended protocol from Alamouti et al. (2006). Beetles from each sample were vortexed in 1 ml of 0.01% Tween80 solution (Nuplex Industries, South Australia, Australia) for 3 min. Thereafter, spore suspensions were spread onto MEA plates and incubated at 22 °C in the dark for 7 d during which all germinating single spores and hyphal tips were transferred onto individual MEA plates, producing axenic cultures which were maintained under the same growing conditions.

Preliminary identification and ITS screening
Isolates were preliminarily grouped based on culture morphology and growth on MEA. In addition to this, a Chelex-based internal transcribed spacer (ITS) region sequencing protocol was used to confirm the putative identification of all ophiostomatoid fungi. In order to achieve this, a small amount of mycelia was scraped from each isolate using a sterile needle tip and placed into individual 200 µl reaction tubes containing 100 µl of molecular biology grade Chelex 100 resin (Bio-Rad Laboratories, Hercules, CA, USA) following a modified protocol for Chelex DNA preparation (Walsh et al. 1991). The ITS region was PCR amplified using the ITS1F and ITS4 primers (White et al. 1990;Gardes and Bruns 1993). PCR reactions included 3 µl Chelex DNA template, 15 µl of MyTaq Red mix (Bioline, London, UK), 0.4 µM of each primer (forward and reverse) and were made up to a  (Chen et al. 2018). Following quality trimming, initial de novo genome assemblies were produced using SPAdes v3.14.1 (Nurk et al. 2013). Assemblies were performed on error-corrected reads with a kmer range of 33, 55, 77, 97 and 111. Assembled genomes provided a platform for sequence extraction of commonly used barcoding loci, including the ITS, the large subunit of ribosomal DNA (LSU), betatubulin (BT), translation elongation factor 1-α (TEF), and the calmodulin (CAL) regions. For each locus, reference sequences for type collections of ophiostomatoid fungi available in GenBank were used to create reference sets. Sequencing reads for each isolate were subsequently mapped against each reference set using BBMap (Bushnell (2014); sourceforge.net/projects/bbmap/). Locusspecific binned reads were generated for each isolate, and these reads were then mapped back to the respective de novo assembled genome in order to extract the assembled locus. This mapping step served as an additional check point to ensure cultures were axenic and only a single sequence was generated from the consensus of all mapped reads using a minimum of 10 × coverage. Extracted loci were then BLASTn searched to confirm taxonomic affinities and obtain similar sequences from GenBank to be included in phylogenetic analyses along with the sequences of type ophiostomatoid fungi.
For multi-locus phylogenetic analysis, the ITS and LSU datasets were used for initial placement of Australian isolates within well-defined species complexes of Ophiostomatales and Microascales. Subsequent phylogenetic analyses of the BT, TEF and CAL regions were performed within each species complex where loci were chosen based on availability of reference data from previous studies (e.g. BT and CAL for Sporothrix) which allowed for more accurate delineation of the Australian taxa. Sequence alignments were performed with MAFFT v7.388 using the E-INS-i algorithm and a gap open penalty of 1.53 (Katoh et al. 2019). The scoring matrix for alignments spanning across multiple genera was 200PAM/k = 2, while for within genus analyses the scoring matrix was set at 1PAM/k = 2 (Linnakoski et al. 2012;Katoh et al. 2019). All aligned sequence datasets were submitted to TreeBase (No. 27096). Maximum Likelihood (ML) analysis was performed with RAxML v8.2.11 (Stamatakis 2014), using the GTR model with optimization for substitution rates and the estimation of rate heterogeneity (GAMMA) specified, while the proportion of invariable sites (+ I) was selected based on results of model estimation using Smart Model Selection (SMS; Lefort et al. (2017); available at http:// www. atgc-montp ellier. fr/ sms/). Confidence support was estimated with bootstrapping of 1000 replicates. Bayesian Inference (BI) analyses were done using MrBayes 3.2.6 (Huelsenbeck and Ronquist 2001). The substitution models and estimated rate parameters, estimated with SMS, were then included manually in MrBayes. Four Markov chain Monte Carlo (MCMC) chains were run at the same time from a random starting tree for 5 000 000 iterations. Trees were sampled every 100 generations with a burnin length of 25%. Posterior probabilities were calculated from a majority rule consensus tree.

Taxonomy
Morphological studies were performed on selected isolates belonging to putative novel lineages identified following phylogenetic analysis. Cultures were grown at 22 °C on 2% MEA (33 g Oxoid MEA, 10 g Oxoid agar, 1 L deionized water), as well as water agar (WA; 15 g Oxoid agar; 1 L deionized water) amended with autoclaved pine needles in order to encourage sporulation. Subsequently, reproductive structures were mounted on glass slides with 85% lactic acid and examined using Leica DM6B and M205C microscopes (Leica, Heerbrugg, Switzerland). Measurements of taxonomically characteristic structures (approximately fifty measurements for each character wherever possible) were made using a mounted Leica camera operated using the Leica application suite software v 3.06. Measurements are presented as, (minimum-) (mean-standard deviation) -(mean + standard deviation) (-maximum).

Genomes of representative species of ophiostomatoid fungi from Australian pine plantations
Draft genomes of representative isolates for each ophiostomatoid taxon collected in this study were subjected to genome quality assessments using QUAST v5.0.2 (Mikheenko et al. 2018). In order to perform suitable comparisons, the QUAST analyses also included publicly available genomes of ophiostomatoid fungi that corresponded to the genera obtained during this study. This was done to update genome completeness assessments against the latest lineage-specific datasets available for BUSCO (Benchmarking Universal Single-Copy Orthologs tool, BUSCO; https:// busco. ezlab. org/), as well as to assess gene predictions using a single prediction tool (GenMark-ES run in fungal mode). BUSCO models were predicted using the Sordariomycetes_odb10 lineage coupled with the Augustus species parameter option set as Neurospora crassa. Draft genome data for the representative isolates sequenced in this study has been deposited at DDBJ/EMBL/GenBank under BioProject PRJNA667796. The accession numbers for each genome are presented in Table 3.

Sample collection and fungal isolation
A total of 135 ophiostomatoid isolates were collected during this study, 15 of which were obtained from Australian plant pathogen reference collections ( The remaining 120 isolates were obtained from samples received during the 2019-20 forest health surveillance period, which included isolations from beetles, beetle galleries and blue-stained wood chips (Additional file 1: Table S1). Samples were largely collected from Pinus radiata (62.5%), the most common Pinus species grown across temperate regions of south-eastern Australia, and P. caribaea x elliottii hybrids (22.5%), the most commonly planted species in the subtropical parts of northern NSW. The remainder were collected from P. taeda (7.5%), P. elliottii (5%), and included a single sample from an amenity planting of P. ponderosa. Three species of pine bark beetles, namely Ips grandicollis, Hylastes ater and Hylurgus ligniperda, and the ambrosia beetle Xyleborus nr. ferrugineus., were recovered from 22 of the samples collected (Additional file 1: Table S1). Ips grandicollis was the most abundant beetle species sampled during this study, comprising approximately 97% of the beetles included in our dataset. Samples containing H. ater and Hy. ligniperda came only from sites in South Australia and Tasmania respectively, while a single sample from northern NSW included the Xyleborus species.
Preliminary identification and ITS screening characterised the 120 ophiostomatoid isolates into 15 taxonomic groups, 14 of which resided in , and a single taxon belonged to Microascales (Table 1). Ophiostomatoid isolates were recovered evenly from the sampled pine tissue (56%) and bark beetles (44%), with about two thirds of all isolations associated with a P. radiata host (Additional file 2: Table S2). Ophiostoma ips (Taxon 9) and Sporothrix pseudoabietina (Taxon 14) were isolated most frequently, making up approximately 53% and 19% of the dataset, respectively (Additional file 2: Table S2). This trend was consistent for the abundantly sampled bark beetle vector, Ips grandicollis, where five additional taxa (taxa 1, 3, 4, 8 and 15) were represented by the 44 fungal isolates collected from this source. The remaining taxa were only recovered occasionally, with the host association and isolation frequencies recorded in Additional file 2: Table S2. Finally, 46 isolates representing all major taxonomic groups were selected for further phylogenetic analysis and taxonomic placement (Table 1).

Phylogenetic analysis
Phylogenetic analysis of the ITS (

Ophiostomatales
Taxon 1 comprised of four representative isolates grouping as a well-supported clade within Ceratocystiopsis (Figs. 2,3). Phylogenetic analysis of the LSU dataset revealed Taxon 1 grouped as an independent lineage, close to Ceratocystiopsis (Cop.) ranaculosa and Cop.   brevicomis (Fig. 2). Analysis of ITS and BT regions (Fig. 3) supports this placement and illustrates that the Australian isolates are most closely related to a previously undescribed taxon reported as Ceratocystiopsis species 1 (Cop. minuta-like) from Canada (Plattner et al. 2009).
Multi-locus analysis suggests the isolates of Ceratocystiopsis sp. (Taxon 1) represent a novel lineage. Three taxa (Taxa 2, 3 and 4) residing within Graphilbum were collected during this study (Fig. 1). Reference collection isolate DAR84707 and two representative isolates collected during this survey forming Taxon 2 (Table 1) were confirmed as Graphilbum fragrans (Figs. 1,  4). Taxon 3 and 4 (which comprised of four and two isolates, respectively;  G. kesiyae, and G. sexdentatum. This new species is described below. Taxon 4 formed a well-supported clade with two previously undescribed Graphilbum isolates reported from the USA (Fig. 4). Within Leptographium s.l., two taxa (Taxa 5 and 6) were collected (Table 1; Fig. 2). Taxon 5 comprised of two reference collection isolates, as well as a single isolate collected during this survey. Phylogenetic analysis confirmed Taxon 5 as Grosmannia huntii (Figs. 2, 5a). Taxon 6 comprised of three representative isolates which fell into a clade within the Gro. galeiformis species complex (Fig. 2). Analysis of BT and TEF confirmed the delineation of Taxon 6 as the species Gro. radiaticola (Fig. 5b).
Taxa 7 to 10 resided within Ophiostoma s. lat. with Taxon 9 the only one belonging to a well-recognised species complex (Fig. 1). Taxon 9 included the reference collection isolates DAR84692, DAR84817, VPRI42284, VPRI42255 and VPRI43316, along with five additional isolates collected during this study ( Table 1). Analysis of the ITS region identified Taxon 9 isolates as Ophiostoma ips (Fig. 1). Despite incongruence with regards to the delimitation of O. ips using ITS alone, BT analysis confirmed little variation between the Australian isolates, and established a clear grouping with several isolates recently confirmed as O. ips (Fig. 6a). Taxa 7, 8 and 10 all grouped peripherally to Ophiostoma s. str. and are regularly referred to as Group A/Lineage G (Chang et al. 2017;Wang et al. 2020). Taxon 7 included two reference collection isolates forming a lineage along with two species, namely O. angusticollis and O. denticulatum (Fig. 6b). The currently available molecular data for reference specimens within this lineage is lacking for appropriate taxonomic comparison, and thus clear differentiation between these species is limited. For now, Taxon 7 is referred to as O. angusticollis. Taxon 8 included a single isolate collected during this study (Table 1), with ITS and BT analyses identifying this taxon as O. fasciatum (Figs. 1, 6b). Taxon 10 included a single strain preliminarily identified as O. pallidulum (Fig. 1). BT analysis further confirmed this identification (Fig. 6b).
The single isolate of Taxon 11 grouped within Raffaelea s. str. (Fig. 2). Analysis of the LSU sequence showed that the Australian isolate forms a well-supported lineage with Raffaelea deltoideospora (Fig. 2). Analysis of the ITS region further validated Taxon 11's placement within Raffaelea s. str. and the delimitation of this species as R. deltoideospora (Additional file 5: Fig. S2).
Three Sporothrix taxa were obtained during this study (Taxa 12, 13, and 14; Table 1). Taxon 12 and Taxon 14 both grouped within the S. gossypina complex (Fig. 1). While analysis of the ITS region gave limited resolution within the S. gossypina complex (Fig. 1), analysis of BT and CAL was able to distinguish between the closely related species (Fig. 7a). The single isolate of Taxon 12 was identified as S. euskadiensis (Romón et al. 2014). Taxon 14 included all reference collection Sporothrix isolates as well as 5 additional representative isolates collected during this study (Table 1). ML analysis placed isolates of Taxon 14 in a well-supported clade alongside the type strain of S. pseudoabietina (Fig. 7a). Taxon 13 comprised of a single isolate collected in this study, with the ITS region placing the taxon among species belonging to a group within Sporothrix recently referred to as " Group G" (De Beer et al. 2016) (Fig. 1). Taxon 13 shared an almost identical ITS sequence with the type sequences for S. nigrograna and S. zhejiangensis (Fig. 7b). While analysis of the BT region was unable to clearly distinguish between S. nebularis and S. zhejiangensis (Fig. 7b), analysis of CAL did show good support for the distinction of Taxon 13 from S. nebularis (Additional file 6: Fig.  S3). A lack of available molecular data for these species limited further phylogenetic comparisons and thus, the placement of the Australian taxon.

Microascales
The single isolate (Taxon 15) residing in Microascales was identified as a Graphium species (Fig. 8). Analysis of ITS and TEF regions revealed that this isolate resides closely to the species of Gra. basitruncatum and Gra. carbonarium (Fig. 8). While this taxon may represent a novel lineage, we have chosen not to formally describe it until additional specimens and/or reference material can be examined. Taxon 15 is thus referred to as a Graphium species.

Species of Ophiostomatales and Microascales associated with Australian Pinus , verified by DNA sequence data
Revision of the literature, as well as the database search, allowed for the comparison of our results to the historical records of ophiostomatoid fungi associated with Pinus in Australia (Table 2)     current study represent first records for Australia, molecular sequence data has verified the previous morphological records of a Ceratocystiopsis sp. (Stone and Simpson 1987) and a Graphium sp. (Vaartaja 1967). Notably however, five previous records still require molecular confirmation and their current status should be treated with care due to the numerous taxonomic re-evaluations that have taken place since their initial identification (Table 2; footnotes).

Draft genomes of representative isolates of ophiostomatoid fungi from Australian pine plantations
Genome summary statistics of the representative draft genomes produced in this study are summarised in Table 3 (see Additional file 3: Table S3 for extended comparison). Genomes were assembled to an average size of 28 Mb and were represented by a mean scaffold number of 148. The N50 ranged from 208,570 to 1,285,428 bp, with the longest contig of 3,412,636 bp generated for the S. pseudoabietina strain, VPRI34531. The GC content had a mean of 57%, with a standard deviation of 3% from this mean. Gene predictions resulted in an average estimate of 7800 Open Reading Frames (ORFs), with a gene density ranging from 240 to 341 ORFs/Mb. All draft genomes had a high BUSCO completeness assessment score ranging between 93.48 and 98.24%. All representative draft genomes were made publicly available on Gen-Bank with Accession details summarised in Table 3.
Diagnosis: Graphilbum ipis-grandicollis is phylogenetically distinct from all morphologically similar species,  from which it can be readily distinguished using molecular sequence data for the ITS, beta-tubulin, elongation factor 1-alpha, and calmodulin regions (Fig. 4   Species identified in the current study are presented in bold, and their accession details can be found in Table 1 a We speculate that the taxon reported as Ceratocystiopsis sp. by Simpson (1987, 1990) is likely the same taxon recorded in the current study. Refer to discussion for more information b Stone and Simpson (1990) reported a Graphilbum sp. associated with Ips grandicollis in NSW. This taxon could refer to any of the four taxa currently confirmed using molecular data c Griggs, J.A. recorded Leptographium lundbergii in association with Hylastes ater infesting P. radiata in TAS. This ID has not been verified molecularly and should be treated with caution considering the taxonomic re-evaluation of L. lundbergii by Jacobs et al. (1998) d Vaartaja (1967 identified several species of Graphium. This descriptor is somewhat ambiguous and could refer to species in both the Ophiostomatales and Microascales

Genus Species/taxon recorded State/s Verified GenBank accession References
Ophiostomatales Ceratocystiopsis Ceratocystiopsis sp. a NSW + Table 1 Stone and Simpson (1987and Simpson ( , 1990  flat on MEA and WA, with colonies reaching approx. 90 mm diam after 14, and 21 d, respectively. Sporulation evident after one week of growth on pine needle amended WA. Initial formation of hyalorhinocladiellalike morphs submerged in agar, with sparse formation on the agar surface. This is followed by formation of the pesotum-like morphs, first forming on the pine needle between 7 and 14 d, and eventually observed sparsely on the agar surface after 4-6 wk. Aerial hyphae bearing conidiophores, mycelial balls, and white to yellow synnemata-like clusters were also randomly observed on the two media. Ecology: Isolated from beetles and beetle galleries found on various Pinus hosts. Host trees: Pinus radiata, P. elliottii and P. caribaea x elliottii hybrid (Additional file 2: Table S2). Insect vector: Ips grandicollis.
Distribution: Currently known only from New South Wales, Australia.
Notes: Graphilbum ipis-grandicollis forms part of an expanding species complex in Graphilbum, which includes G. crescericum, G. furuicola, G. interstitiale, G. kesiyae, and G. sexdentatum. Using morphology alone makes distinction between these closely related species difficult, as they share considerable similarities in the size and shapes of conidia, conidiogenous apparatus, and the asexual morphs recorded (Jankowiak et al. 2020

Discussion
This study was undertaken to review and update the status of ophiostomatoid fungi associated with pine and pine bark beetles in plantations in south-eastern Australia. This was achieved by reviewing reference isolates available from historic collections lodged in Australian collections, as well as including a total of 120 new isolates collected through routine forest health surveillance during the 2019-20 period. Multi-locus phylogenetic analysis using whole genome sequencing of 46 representative isolates revealed a greater than expected diversity of ophiostomatoid fungi, including 14 species from six genera in Ophiostomatales and a single species residing in . While most species reported in this study were already known, our study includes seven first reports and three verifications for Australia, including the identification of three previously undescribed lineages, viz. Graphilbum ipis-grandicollis sp. nov. (Taxon 3), Ceratocystiopsis sp. (Taxon 1) and a Graphium sp. (Taxon 15). Draft genomes of representative isolates for each taxon are also provided here to contribute to a curated reference database of ophiostomatoid fungi for Australian biosecurity.
Of the five ophiostomatoid genera previously recorded from pine in Australia, isolates were available for Ophiostoma, Graphilbum, Leptographium s. lat., and Sporothrix. Although historical records also included the morphological identification of taxa belonging to Ceratocystiopsis Simpson 1987, 1990), no reference material was available of this genus. Similarly, with the recent detection of O. pallidulum (Carnegie and Nahrung 2019;, no isolates were readily available for inclusion in this study. Detections made during the current survey included seven taxa not previously recorded in Australia. Four were identified as known species: specifically, Gro. radiaticola (Taxon 6), O. fasciatum (Taxon 8), R. deltoideospora (Taxon 11), and S. euskadiensis (Taxon 12). Taxon 4 and 13 are tentatively identified here as G. cf. Fig. 9 Graphilbum ipis-grandicollis (VPRI43762). a Fourteen-d culture on MEA. b-d Pesotum-like macronematal asexual morph formed on pine needle mounted in WA. e-f Conidiogenous cells (e) and conidia (f) of Pesotum-like macronematal asexual morph. g. Hyalorhinocladiella-like asexual morph. h, i Conidiogenous cells (h) and conidia (i) of Hyalorhinocladiella-like asexual morph. Bars: b = 500 µm; c, d = 50 µm; e-i = 10 µm of species complexes and/or groupings that are consistently associated with bark beetles or other insect vectors Wingfield et al. 2017b). While our goal was to assess the diversity of ophiostomatoid fungi associated with pine bark beetles and beetle galleries collected during routine forest health surveillance, I. grandicollis was the more commonly encountered beetle species during the current surveillance period. This was a somewhat expected observation as, historically, I. grandicollis is more commonly caught in NSW than either H. ater or Hy. ligniperda (Stone et al. 2010). While this could explain the dominance observed for some of the ophiostomatoid species isolated, such as Ophiostoma ips and Sporothrix pseudoabietina, our results highlight the potential phoresy of these associations with I. grandicollis being linked to seven of the taxa recovered during this study (Additional file 2: Table S2). While we were able to make a few general observations regarding the patterns of isolations, a more in-depth systematic review would be required for an improved understanding and description of these fungus-vector associations across Australia. More targeted surveys, particularly studies focused on the insect vectors present, are likely to reveal an even greater diversity, for example, the isolates of Gro. radiaticola and Gro. huntii were only recovered from samples that came from H. ater and Hy. ligniperda infestations.
Genome assemblies for the isolates chosen as representatives for each taxon collected in this study resulted in the addition of 12 draft genomes to the Ophiostomatales, and the release of the first draft genome publicly available for an isolate residing in the Graphiaceae (Microascales). The genome assembly statistics of the Ophiostomatales isolates collected during this study mirror those available for species residing in Ophiostoma, Sporothrix, Graphilbum, Leptographium s.l., Ceratocystiopsis and Raffaelea (DiGuistini et al. 2011;Forgetta et al. 2013;Haridas et al. 2013;Teixeira et al. 2014;van der Nest et al. 2014;Wingfield et al. 2015aWingfield et al. , b, 2016Wingfield et al. , 2017aWingfield et al. , 2018D' Alessandro et al. 2016;Huang et al. 2016;Shang et al. 2016;Jeon et al. 2017;Vanderpool et al. 2018;Liu et al. 2019). Comparisons of genome statistics, specifically estimated size, GC content and number of predicted genes, generally correlate with the taxonomic placement of each species (Additional file 3: Table S3). This is evident, for example, when comparing the genomes of Gro. galeiformis and Gro. radiaticola, or S. euskadiensis and S. pseudoabietina. In both cases the size, GC content and number of predicted open reading frames (ORFs) vary marginally. There are however slight deviations evident within some genera. For example, in Ceratocystiopsis the genome sizes range from 20.45 to 21.30 Mb, and the number of predicted ORFs are somewhat lower for Cop. brevicomis (6884 ORFs) and Ceratocystiopsis sp. VPRI43766 (Taxon 1; 6967 ORFs) when compared to that of Cop. minuta (7786 ORFs).
In the modern era, fungal taxonomy relies more heavily on an integrative approach where genealogical concordance is combined with morphological examination to recognise and delimit species (Lücking et al. 2020). For taxonomists and diagnosticians looking to delineate taxa of ophiostomatoid fungi, this could include analysing anything from two to ten different gene regions De Beer et al. 2014 all while comparing morphological characters that can prove extremely difficult to distinguish (Jankowiak et al. 2020). Another major challenge that was exemplified several times in this study is the inconsistency of recovering sequence data for specific taxa. As the cost of sequencing continues to decrease, the feasibility for future taxonomic surveys to include whole genome sequences should become more readily attainable. As shown in this study, future taxonomic surveys could strive to include whole genome sequence data published alongside their identifications and/or descriptions of novel taxa. Currently, in the Ophiostomatales the number of available genomes encompasses 38 species across 11 genera. Expanding these genomic resources provides a fundamental platform on which diagnostic and biosecurity capacity can be developed.

Conclusions
The results of this study have uncovered a higher than expected diversity for ophiostomatoid fungi associated with pine and pine bark beetles in south eastern Australia. The current status of ophiostomatoid fungi in Australian pine plantations confirmed using molecular data has been expanded from 7 previously confirmed taxa to now include 15 verified species across six genera in the Ophiostomatales, as well as a single taxon identified in the Graphiaceae (Microascales). As demonstrated several times in this study, a major challenge for accurate fungal diagnostics and species delimitation is the availability of multi-locus sequence data for reference specimens. With the ever-decreasing costs of sequencing, as well as the need for multi-locus sequence data, our study provides an early example of WGS replacing standard PCR-based approaches. Future taxonomic studies could begin to look in earnest at the opportunities of providing the full complement of DNA sequence data along with the results of a given taxonomic survey. This would ensure that taxonomic studies continue to improve upon the availability of molecular data while rapidly expanding on the numbers of sampled taxa. Results of the current survey, coupled with other recent detections in Australia, illustrates the need for continued surveillance of ophiostomatoid fungi. This not only provides an important platform for recognising the underlying diversity of these fungi but