Six new species of Sporothrix from hardwood trees in Poland

Abstract Sporothrix (Sordariales, Ascomycota) is a well-supported monophyletic lineage within the Ophiostomatales, species of which occur in a diverse range of habitats including on forest trees, in the soil, associated with bark beetles and mites as well as on the fruiting bodies of some Basidiomycota. Several species have also been reported as important human and animal pathogens. During surveys of insect- and wound-associated Ophiostomatales from hardwood trees in Poland, many isolates with affinity to Sporothrix were recovered. In the present study, six undescribed Sporothrix spp. collected during these surveys are characterized based on their morphological characteristics and multi-locus phylogenenetic inference. They are described as Sporothrixcavum, Sporothrixcracoviensis, S.cryptarchum, S.fraxini, S.resoviensis, and S.undulata. Two of the Sporothrix spp. reside in the S.gossypina-complex, while one forms part of the S.stenoceras-complex. One Sporothrix sp. is a member of lineage F, and two other species grouped outside any of the currently defined species complexes. All the newly described species were recovered from hardwood habitats in association with sub-cortical insects, wounds or woodpecker cavities. These species were morphologically similar, with predominantly asexual states having hyaline or lightly pigmented conidia, which produce holoblastically on denticulate conidiogenous cells. Five of the new taxa produce ascomata with necks terminating in long ostiolar hyphae and allantoid ascospores without sheaths. The results suggest that Sporothrix species are common members of the Ophiostomatales in hardwood ecosystems of Poland.

de Hoog et al. (1985) recognized that Sporothrix is not a homogenous group. As DNA sequencing technology was applied to resolve taxonomic relationships for fungi, evidence emerged that S. schenckii is phylogenetically related to species of Ophiostoma (Berbee and Taylor 1992;Hausner et al. 1993Hausner et al. , 2000. In these studies, species producing only sporothrix-like asexual states were treated as members of the S. schenckii-O. stenoceras complex in Ophiostoma sensu lato (De Beer et al. 2003;Villarreal et al. 2005;Roets et al. 2006;Zipfel et al. 2006;De Meyer et al. 2008;Linnakoski et al. 2010;Kamgan Nkuekam et al. 2012). The genus Sporothrix was recently redefined and emended based on the analysis of partial 18S and 28S rDNA sequences for species in the Ophiostomatales (De . Sporothrix was consequently separated from species of Ophiostoma and various complexes were defined within Sporothrix. Sporothrix is now defined as one of nine relatively clearly defined genera in the Ophiostomataceae (De Beer and Wingfield 2013;De Beer et al. 2013a. As currently recognized, Sporothrix includes 56 species Ngubane et al. 2018;Wang et al. 2019;Musvuugwa et al. 2020), which are characterized by their dark brown to black, globose ascomata with elongated necks up to 1600 μm, occasionally terminating in an ostiole, often surrounded by ostiolar hyphae. Ascospores are usually curved and lunate to reniform, without a sheath (De Beer and Wingfield 2013). The asexual states have conidiophores that proliferate sympodially and produce hyaline or occasionally pigmented conidia on denticulate conidiogenous cells (De Beer and Wingfield 2013).
Sporothrix includes a large assemblage of species that are widely distributed across various climatic zones of the world (De Beer and Wingfield 2013;. Species also occupy a wide range of habitats. The greatest numbers of species are found on bark, in the infructescences of Protea spp. and on the wood of different forest trees (e.g., Roets et al. 2008Roets et al. , 2009Roets et al. , 2013De Errasti et al. 2016). Other species have been described from soil, bark beetles, ambrosia beetles, mites, and from the fruiting bodies of basidiomycetes (e.g., Constantinescu and Ryman 1989;Marmolejo and Butin 1990;De Meyer et al. 2008;Roets et al. 2008;De Errasti et al. 2016). Several species are also well-known as human and animal pathogens (Travassos and Lloyd 1980;Summerbell et al. 1993;Barros et al. 2004;Lòpez-Romero et al. 2011;Zhang et al. 2015). Jankowiak et al. (2019a) conducted the first extensive survey of fungal associates of hardwood-infesting bark and ambrosia beetles in Poland. In the same year, Ophiostomatales associated with wounds on hardwood trees were also studied in Poland (Jankowiak et al. 2019b). These studies reported several Sporothrix species, which were apparently new to science, but names were not provided for them. In addition, one unknown Sporothrix species was isolated from cavities of woodpeckers in Poland (Jankowiak et al. 2019c). In this study, morphological characters and DNA sequence data for the ITS region (ITS1-5.8S-ITS2) and three protein coding genes (β-tubulin, calmodulin, translation elongation factor 1-α) were analyzed to characterize six new species of Sporothrix. These were compared with closely related known species and formal descriptions have been provided for them.

Fungal isolates
The collection details for the isolates included in the present study (Table 1) are provided in previous studies (Jankowiak et al. 2019a(Jankowiak et al. , 2019b(Jankowiak et al. , 2019c. The cultures are maintained in the culture collection of the Department of Forest Ecosystems Protection, University of Agriculture in Krakow, Poland, and in the culture collection of the Natural Resources Institute Finland (Luke), Helsinki, Finland. The ex-type isolates and representative isolates of the new species described were deposited in the culture collection (CBS) of the Westerdijk Fungal Biodiversity Institute, Utrecht, The Netherlands. Dried cultures were deposited as holotype specimens in the Mycological Herbarium (O), Natural History Museum, University of Oslo, Norway.

Microscopy and growth studies
Morphological characters were examined for selected isolates as well as for the herbarium specimens selected as types. Cultures were grown on 2% Malt Extrat Agar (MEA) made up of 20 g Bacto malt extract, 20 g agar Bacto agar powder (Becton Dickinson and Company, Franklin Lakes, USA) in 1 l deionized water. In attempts to induce the formation of ascomata, autoclaved twigs of host trees including the bark were placed at the centres of agar plates containing MEA. Fungal cultures were derived from single spores. To promote the production of ascomata, single conidial isolates were crossed in all possible combinations, following the technique described by Grobbelaar et al. (2009). These cultures were incubated at 25 °C and monitored regularly for the appearance of fruiting structures.  (Jankowiak et al. 2019a(Jankowiak et al. , 2019b(Jankowiak et al. , 2019c. Morphological features were examined by mounting fungal tissue in 80% lactic acid on glass slides, and fruiting structures were observed using a Nikon Eclipse 50i microscope (Nikon Corporation, Tokyo, Japan) with an Invenio 5S digital camera (DeltaPix, Maalov, Denmark) to capture photographic images. Microscopy followed the technique described by Kamgan Nkuekam et al. (2011). Colour designations were based on the colour charts of Kornerup and Wanscher (1978).
For each taxonomically relevant structure, fifty measurements were made, when possible, using the Coolview 1.6.0 software (Precoptic, Warsaw, Poland). Averages, ranges and standard deviations were calculated for the measurements, and these are presented in the format '(min-)(mean-SD)-(mean+SD)(-max)'.
Growth characteristics for the novel species were determined by analysing the radial growth for 12 isolates (two for each species) ( Table 1). Agar disks (5 mm diam.) were cut from the actively growing margins of fungal colonies and these disks were placed at the centres of plates containing 2% MEA. Four replicate plates for each of the six putative new species were incubated at temperatures between 5, and 35 °C at 5 °C intervals. The radial growth (two measurements perpendicular to each other per plate) was determined 14 d after inoculation, and growth rates were calculated as mm/d.

PCR, sequencing and phylogenetic analyses
DNA extractions were performed as described by Jankowiak et al. (2019d). For sequencing and phylogenetic analyses, four loci were amplified: the internal transcribed spacer region (ITS, consisting of ITS1, 5.8S, and ITS2), beta tubulin (βT), calmodulin (CAL), and the translation elongation factor 1-alpha (TEF1-α). The primers used for PCR and sequencing of the various gene regions were as follows: ITS1-F (Gardes and Bruns 1993) and ITS4 (White et al. 1990) for ITS; T10 (O'Donnell and Cigelnik 1997) or Bt2a together with Bt2b (Glass and Donaldson 1995) for βT; F-728F (Carbone and Kohn 1999) and EF2 (O'Donnell et al. 1998) were used for TEF1-α; CL1 and CL2a (O'Donnell et al. 2000) or CL3F and CL3R  were used for CAL. PCR and sequencing protocols were as described by Jankowiak et al. (2019d), other than the annealing temperature being optimised for some individual reactions. All analyses were run independently for each gene partition (Figs 1-4). Resulting trees were visually compared for topological incongruence. Gene partitions showing no topological incongruence (βT, CAL) were combined and presented as a concatenated construct (Fig. 5).
MP analyses were performed using PAUP* 4.0b10 (Swofford 2003). Gaps were treated as fifth state. Bootstrap analysis (1000 bootstrap replicates) was conducted to determine the levels of confidence for the nodes within the inferred tree topologies. Tree bisection and reconnection (TBR) was selected as the branch swapping option. The tree length (TL), Consistency Index (CI), Retention Index (RI), Homoplasy Index (HI) and Rescaled Consistency Index (RC) were recorded for each analysed dataset after the trees were generated.
BI analyses using Markov Chain Monte Carlo (MCMC) methods were carried out with MrBayes v3.1.2 (Ronquist and Huelsenbeck 2003). Four MCMC chains were run for 10 million generations applying the best-fit model for each dataset. Trees were sampled every 100 generations, resulting in 100,000 trees. Tracer v1.4.1 (Rambaut and Drummond 2007) was utilized to determine the burn-in value for each dataset. The remaining trees were utilised to generate a 50% majority rule consensus tree, which allowed for calculating posterior probability values for the nodes.

Phylogenetic Analyses
Alignments for the ITS dataset contained 575 characters; for the βT 303 characters; for CAL 543 characters; and for TEF1-α 812 characters; for the concatenated combined dataset 826 (including gaps), of which respectively 202, 123, 271, 439, 390 were parsimony-informative. The exon/intron arrangement of the βT data included exons 5 and 6, interrupted by intron 5. The exon/intron arrangement of the CAL data included exons 4 and 5, interrupted by intron 4. The aligned TEF1-α gene region consisted of intron 3 and exons 4 and 5, but lacked intron 4.
DNA sequence data were generated for 24 isolates considered in this study (Table 1). Blast analyses of the ribosomal DNA sequences placed all the isolates in Sporothrix. Based on phylogenetic analyses of the ITS (Fig. 1), the isolates emerged as six undescribed taxa. Phylogenetic analysis of the ITS indicated that the unknown species resided in two previously defined Sporothrix species complexes, including the S. gossypina-and S. stenoceras-species complexes, and lineage "F". Additionally, isolates representing two new species grouped outside any of the currently defined species   (Fig. 1). Based on the availability of sequence data for these complexes, different datasets were assembled and analysed separately for each species complex. Seven isolates from hardwood-infesting bark beetles identified as Sporothrix 7 and Sporothrix 8 by Jankowiak et al. (2019a) resided in the S. gossypina-complex (Fig. 1). All three gene regions (ITS, βt, CAL) separated Sporothrix sp. 8 from the other known species with strong statistical support (Figs 2-4). The ITS and βt gene regions grouped isolates of this species together with the ex-type isolate of S. variecibatus, while CAL gene region placed it with S. aurorae (Figs 1-3). Isolates representing Sporothrix sp. 7 had ITS sequences that were almost identical to the ITS sequences for S. fusiformis, S. lunata and S. prolifera (Fig. 1). In the βt and CAL trees (Figs 2, 3), Sporothrix sp. 7 formed lineages that clearly separated this species from the known species in the S. gossypina complex, and although there were differences in the βt sequence compared to other species, the node lacked statistical support (Fig. 2). The combined analyses of the βt and CAL datasets clearly distinguish Sporothrix sp. 7 and Sporothrix sp. 8 into separate lineages within the S. gossypina-complex (Fig. 5).
The single isolate from a wound on Betula pendula identified as Sporothrix sp. 10 by Jankowiak et al. (2019b), resided in S. stenoceras-complex and grouped closely with S. stenoceras sensu stricto based on analysis of ITS, βt, CAL, and TEF1-α gene regions (Figs 1-4). All three gene regions separated Sporothrix sp. 10 from S. stenoceras, although this separation was not statistically supported by the ITS gene region (Figs 1-4). The combined analyses of the βt and CAL datasets clearly distinguish Sporothrix sp. 10 into separate lineages within the S. stenoceras-complex (Fig. 5).
Two isolates from woodpecker cavities identified as Sporothrix sp. 18 by Jankowiak et al. (2019c), belonged to the lineage F defined by  based on the ITS tree. All the three gene regions (ITS, βt, CAL) separated Sporothrix sp. 18 from the other known species in lineage F with strong statistical support (Figs 1-4). The combined analyses of the βt and CAL datasets clearly distinguish Sporothrix sp. 18 into separate lineages within the Sporothrix spp. (Fig. 5).
Fourteen isolates from wounds on different species of hardwood trees and nitidulid beetles identified as Sporothrix sp. 11 and Sporothrix sp. 12 by Jankowiak et al. (2019b) did not group in any of the defined Sprothrix species complexes based on analysis of ITS gene region and formed a monophyletic lineage within Sporothrix (Fig. 1). Isolates of Sporothrix sp. 11 had ITS sequences that were identical with ITS sequences noted in Sporothrix sp. 12. In the βt, CAL, and TEF1-α trees (Figs 2-4), Sporothrix sp. 11 and Sporothrix sp. 12 formed well-supported lineages that clearly separated these two putative new species from each other. The combined analyses of the βt and CAL datasets also separated Sporothrix sp. 11 and Sporothrix sp. 12 from the other known species in Sporothrix spp. and also from each other (Fig. 5).
Sporothrix resoviensis was represented by one isolate collected from a wound on Betula pendula. It corresponds to Sporothrix sp. 10 in the study of Jankowiak et al. (2019b).
Notes. This species is phylogenetically distinct from the other Sporothrix species based on the ITS, βT, CAL and TEF1-α sequences. Sporothrix cryptarchum is phylogenetically closely related to S. undulata (Sporothrix sp. 12) described in the present study. This species also shares morphological similarities such as kidney-shaped ascospores and two morphological forms of conidia with S. undulata. However, S. cryptarchum has narrow ascospores (0.8-1.5 μm) compared to S. undulata (1.1-2 μm). It also has distinct ostiolar hyphae, with those in S. cryptarchum often dichotomously branching while in S. undulata these hyphae occur only sporadically and do not have dichotomous branching. Both species produce hyaline and pigmented conidia. However, S. cryptarchum cultures are predominantly hyaline whereas those in pure cultures of S. undulata are mostly pigmented. Their conidial shapes in these two species are similar but their dimensions are distinct. Sporothrix cryptarchum has conidia that are smaller than those of S. undulata. In addition, cultures of S. cryptarchum are white and grow in a circular pattern with smooth margins while those of S. undulata grow in a circular pattern with undulate margins and some have grey pigmentation.
Sporothrix cryptarchum was represented by four isolates collected from Poland. It corresponds to Sporothrix sp. 11 in the study of Jankowiak et al. (2019b). Sporothrix cryptarchum was isolated from wounds on hardwood trees and nitidulid beetles (Coleoptera: Nitidulidae), which visited fresh wounds on Quercus robur.
Notes. This species is phylogenetically distinct from the other Sporothrix species based on the ITS, βT, CAL and TEF1-α sequences. Sporothrix undulata is phylogenetically closely related to S. cryptarchum described in this study. The morphological differences between S. undulata and S. cryptarchum are described in the section above treating S. cryptarchum. Sporothrix undulata was represented by nine isolates collected from Poland. It corresponds to Sporothrix sp. 12 in the study of Jankowiak et al. (2019b). In this study S. undulata was isolated from wounds on hardwood trees and from adults of nitidulid beetles (Coleoptera: Nitidulidae), which visited wounds on Quercus robur. Sporothrix cavum R. Jankowiak sp. nov. Mycobank: 840479 Fig. 11 Etymology. From Latin, referring to the hollow cavities produced by woodpeckers and from which this fungus was collected.
Host tree. Notes. This species is phylogenetically distinct from the other Sporothrix species based sequences for the ITS, βT, CAL and TEF1-α regions. Sporothrix cavum is related to S. polyporicola based on analyses of the ITS sequences. However, S. cavum in contrast to S. polyporicola, does not produce a sexual morph (Constantinescu and Ryman 1989). In addition, S. cavum has obovoid and short conidia (3.1-7.8 μm), whereas S. polyporicola has clavate and longer conidia (6-14 μm) (Constantinescu and Ryman 1989).
Sporothrix cavum was represented by two isolates collected from the cavities produced by the woodpeckers Dendrocopos major on Salix fragilis and Dendrocopos medius on Malus domestica. It corresponds to Sporothrix sp. 18 in the study of Jankowiak et al. (2019c).

Discussion
Our work (Jankowiak et al. 2019a(Jankowiak et al. , 2019b(Jankowiak et al. , 2019c this study) has led to the discovery of six novel Sporothrix species associated with hardwood trees in Poland. Description of these new species brings the total number of species in this genus to 62, of which 16 occur in Poland. These include the six species described here as well as S. aurorae (Jankowiak et al. 2019b), S. cantabriensis (Jankowiak et al. 2017), S. dentifunda (Aghayeva et al. 2005, Jankowiak et al. 2019b), S. eucastaneae (Jankowiak et al. 2019a(Jankowiak et al. , 2019b(Jankowiak et al. , 2021, S. fusiformis (Jankowiak et al. 2019a(Jankowiak et al. , 2019b, S. inflata (Jankowiak et al. 2012;Bilański 2013a, 2013b), S. inflata '2' (Jankowiak et al. 2019a(Jankowiak et al. , 2019b, S. prolifera (Kowalski and Butin 1989;Jankowiak et al. 2019aJankowiak et al. , 2019b, S. stenoceras, (Kowalski and Butin 1989;Bilański 2013b, Jankowiak et al. 2019b) and S. variecibatus (Jankowiak and Bilański 2013b). All of the species described in this study are morphologically similar, having asexual states with hyaline or lightly pigmented conidia produced holoblastically on denticulate conidiogenous cells or directly from the hyphae. Where ascomata were present, these tended to have globose bases with elongated necks terminating in long ostiolar hyphae and allantoid or kidney-shaped ascospores not surrounded by hyaline sheaths. All of the newly described species grew optimally at 25 °C and all also grew well at 30 °C on MEA. Sporothrix undulata and S. cavum differed from the other four species in having pigmented as opposed to white cultures on MEA. All of the newly described species were recovered from hardwood ecosystems in Poland in association with bark and ambrosia beetles, nitidulid beetles, naturally occurring tree wounds or woodpecker cavities.
The six species described in this study can easily be distinguished from each other and from the other species of Sporothrix based on the DNA sequence comparisons. Analyses of the ITS sequence data were insufficient to distinguish between S. cryptarchum and S. undulata or between S. cracoviensis and S. fusiformis. However, analyses of sequence data for the protein-coding genes, including the βT, CAL and TEF1-α showed that S. cracoviensis, S. cryptarchum, and S. undulata represent distinct taxa. Furthermore, the two closely related species, S. cryptarchum and S. undulata formed a new and well-supported lineage in Sporothrix including species infecting wounds on a variety of hardwood trees. The species in this lineage are characterised by having both hyaline as well as pigmented conidia and kidney-shaped ascospores.
The asexual morphs of the Sporothrix species described in this study had variable morphology. All species had hyaline conidia produced holoblastically on denticulate conidiogenous cells that proliferate sympodially or arise directly from hyphae. Sporothrix cryptarchum and S. undulata also had pigmented globose conidia formed singly or in chains, either directly on the sides of the vegetative hyphae or on short lateral branches. The presence of two different conidial types has previously been found in other Sporothrix species, including Sporothrix dimorphospora and S. brunneoviolacea (Madrid et al. 2010) as well as S. brasiliensis, S. globose, and S. mexicana (Marimon et al. 2007).
Recently, Jankowiak et al. (2019b) provided evidence that fresh wounds on hardwood trees in Europe are preferred habitats for some Sporothrix species. These authors isolated 15 Sporothrix species from trees belonging to 12 species of angiosperms. Likewise, nine Sporothrix species have been described from fresh wounds on non-native Eucalyptus spp. and various genera of native trees in South Africa (Kamgan Nkuekam et al. 2012;Musvuugwa et al. 2016Musvuugwa et al. , 2020Osorio et al. 2016).
Three species of wound-associated Sporothrix spp. collected during a survey reported in the study of Jankowiak et al. (2019b) were included in the present study. The greatest number of isolates (194) obtained during that survey were those of S. undulata. This species was found as a common associate of bleeding wounds on Quercus robur and Salix fragilis, suggesting that they might have some level of pathogenicity. The other species inhabiting wounds on hardwood trees that was collected during the survey of Jankowiak et al. (2019a) was S. cryptarchum (34 isolates). Transfer of this species to the sampled tree wounds was most likely by nitidulid (Coleoptera, Nitidulidae) beetles as previously noted by Jankowiak et al. (2019b) who suggested that these insects commonly transmit Ophiostomatales, including Sporothrix species to tree wounds in Poland. Likewise, Kamgan Nkuekam et al. (2012) have demonstrated that the nitidulid beetles Brachypeplus depressus and Carpophilus spp. vector S. candida and S. fumea in the Eucalyptus plantations of South Africa. This association is also consistent with other studies providing compelling evidence that nitidulid beetles act as vectors of the well-known pathogens, such as Bretziella fagacearum (De Beer et al. 2017;Jagemann et al. 2018) and Ceratocystis albifundus (Heath et al. 2009).
The second largest number of isolates (81 in total) included in this study represented two species in the S. gossypina-complex, bringing the total number of species in that complex to 15 Wang et al. 2019). Sporothrix cracoviensis was represented by 45 isolates from the ambrosia beetles Trypodendron domesticum and T. signatum collected on Fagus sylvatica (Jankowiak et al. 2019a). This is not unusual given that an association between ambrosia beetles has recently been recorded by De Errasti et al.(2016) in a study on Nothofagus pumelo in Patagonia. The other species residing in this complex collected during the survey of Jankowiak et al. (2019a) is S. fraxini (36 isolates). This fungus was found on Fraxinus excelsior in association with the bark beetles Hylesinus crenatus and H. varius (Jankowiak et. al. 2019a).
The Polish study by Jankowiak et al. (2019a) revealed that, apart from S. cracoviensis and S. fraxini, five other Sporothrix species (S. fusiformis, S. prolifera, S. eucastanea, Sporothrix sp. 4, Sporothrix sp. 9) were associated with bark beetles. These findings confirm that most species in the S. gossypina complex are associated with galleries of conifer-infesting bark beetles worldwide . The other species in the S. gossypina-complex were isolated from the stained oak wood (Kowalski and Butin 1989;Aghayeva et al. 2004), cankers caused by Cryphonectria parasitica on chestnut (Davidson 1978), a hardwood tree native to South Africa (Musvuugwa et al. 2016), and from mites infesting the infructescences (flower heads) of Protea in South Africa (Roets et al. 2008).
Sporothrix cavum, the remaining taxon collected from hardwood trees during the surveys that formed the basis of the present study, resided in lineage F defined by De . This lineage includes three species, namely S. polyporicola, S. dimorphospora, and S. inflata '2'. Two of these species (S. dimorphospora, and S. inflata '2') are known from soil and S. polyporicola was isolated from basidiocarps of the polypores Fomitopsis pinicola and Amaropostia stiptica (Constantinescu and Ryman 1989;Madrid et al. 2010). The results of the present study show that species in this complex also accommodate wood-inhabiting Sporothrix species. Other than the fact that S. cavum was isolated from cavities on Salix fragilis and Malus domestica made by woodpeckers (Jankowiak et al. 2019c), nothing is known regarding the ecology or distribution of the fungus. It could, for example, be introduced into these cavities by arthropods or have some relationship with the woodpeckers themselves.
The results of this study have substantially expanded our knowledge of Sporothrix and the ecology of species in this genus. Broadly, the results suggest that Sporothrix species are common members of the Ophiostomatales in hardwood ecosystems in Poland. Furthermore, interesting questions have arisen that should shape future investigations regarding these fungi.