Toward a Natural Classification of Botryosphaeriaceae: A Study of the Type Specimens of Botryosphaeria sensu lato

The genus Botryosphaeria includes more than 200 epithets, but only the type species, Botryosphaeria dothidea and a dozen or more other species have been identified based on DNA sequence data. The taxonomic status of the other species remains unconfirmed because they lack either morphological information or DNA sequence data. In this study, types or authentic specimens of 16 “Botryosphaeria” species are reassessed to clarify their identity and phylogenetic position. nuDNA sequences of four regions, ITS, LSU, tef1-α and tub2, are analyzed and considered in combination with morphological characteristics. Based on the multigene phylogeny and morphological characters, Botryosphaeria cruenta and Botryosphaeria hamamelidis are transferred to Neofusicoccum. The generic status of Botryosphaeria aterrima and Botryosphaeria mirabile is confirmed in Botryosphaeria. Botryosphaeria berengeriana var. weigeliae and B. berengeriana var. acerina are treated synonyms of B. dothidea. Botryosphaeria mucosa is transferred to Neodeightonia as Neodeightonia mucosa, and Botryosphaeria ferruginea to Nothophoma as Nothophoma ferruginea. Botryosphaeria foliicola is reduced to synonymy with Phyllachorella micheliae. Botryosphaeria abuensis, Botryosphaeria aesculi, Botryosphaeria dasylirii, and Botryosphaeria wisteriae are tentatively kept in Botryosphaeria sensu stricto until further phylogenetic analysis is carried out on verified specimens. The ordinal status of Botryosphaeria apocyni, Botryosphaeria gaubae, and Botryosphaeria smilacinina cannot be determined, and tentatively accommodate these species in Dothideomycetes incertae sedis. The study demonstrates the significance of a polyphasic approach in characterizing type specimens, including the importance of using of DNA sequence data.

INTRODUCTION status of most taxa accommodated in Botryosphaeria sensu lato remains uncertain. As a fundamental element in the current Code of Nomenclature for algae, fungi, and plants, type studies play a critical role in epitypification, as well as in defining species or genera of Ascomycetes (Zhang et al., 2009). Specifically for taxa in the Botryosphaeriales, there are few studies based on DNA sequence data. Almost all the older names linked to Botryosphaeria lack cultures or DNA sequence data, and they can consequently not be classified to genus or even family rank with confidence. Thus, these names are unusable unless they are either epitypified or supplemented with DNA sequence data (Slippers et al., 2014). The aims of the present study were thus to verify the identity of 17 selected type or authentic specimens (representing 16 species) currently placed in Botryosphaeria, using morphological characteristics and nuDNA sequence data.

Type Study
Type specimens of 16 putative Botryosphaeria species were obtained on loan from the Conservatoire et Jardin botaniques de la Ville de Genève (G), Naturhistorisches Museum Wien (W), Field Museum of Natural History (F), Royal Botanic Gardens (K), University of Michigan (MICH) and New York State Museum (NYS) ( Table 1). The type specimens were described and illustrated following the protocol by Zhang et al. (2012). Sections made from specimens were studied at × 1,000 magnification using a Nikon E600 compound microscope. Ascomata were examined under a Leica M125 dissecting microscope. Sections of ascomata, hamathecia, asci, and ascospores were mounted in water or 10-100% lactic acid. Micrographs were made from tissues mounted in water with 10-100% lactic acid or cotton blue. Question marks (?) indicate possible type specimens.
nuDNA Extraction, PCR Amplification and Cloning After getting the DNA extraction permission, nuDNA was extracted from ascomata or conidiomata using a Forensic DNA kit (OMEGA Bio-tek). The internal transcribed spacer of regions (1 and 2) of the nuDNA (ITS) was amplified and sequenced with primers ITS-4 and ITS-5 (White et al., 1990). The 28S large subunit nuDNA (LSU) was amplified and sequenced with primers LROR and LR5 (Vilgalys and Hester, 1990). Sections of the translation elongation factor-1α (tef1-α) with primers EF1-688F and EF1-1251R  and the β-tubulin gene (tub2) with primers Bt2a and Bt2b (Glass and Donaldson, 1995). PCR amplification and sequencing was conducted following the protocol by Zhang et al. (2009). Some of the resulting sequences had ambiguous base calls, possibly due to the contamination of the other fungi occurring on the specimens. All PCR products exhibiting this phenomenon were cloned using the pGEM-T Vector System I cloning kit (Promega).

Sequence Alignment and Phylogenetic Analysis
For the sequences obtained, a search was conducted using BLAST (Basic Local Alignment Search Tool) in GenBank sequences 2 to confirm the generic status of the related specimens. Sequence data for each individual gene region, ITS, LSU, tef1-α and tub2, as well as the combined datasets were used to infer the phylogenetic relationships among all confirmed Botryosphaeria, Neofusicoccum, Nothophoma species for which sequence data were available from GenBank (see text footnote 2), together with the sequences generated in this study. Alignments were made in MEGA v. 6 (Tamura et al., 2013) and phylogenetic analyses performed in PAUP v. 4.0b10 (Swofford, 2002) and MrBayes v. 3.1.2 (Ronquist and Huelsenbeck, 2003). Prior to phylogenetic analyses, ambiguous sequences at the start and the end of sequences were deleted and gaps manually adjusted to optimize the alignments. Maximum Parsimony (MP) was used to conduct heuristic searches as implemented in PAUP with the default options method (Zhang et al., 2008). Analyses were made under different parameters of maximum parsimony criteria as outlined in Zhang et al. (2008). Clade stability was assessed in a bootstrap analysis with 1,000 replicates, random sequence additions with MaxTrees set to 1,000 and other default parameters as implemented in PAUP. Maximum likelihood (ML) was also conducted using heuristic searches with the default options method as implemented in PAUP. For the ML analysis, best-fit model of nucleotide evolution was selected by hierarchical likelihood ratio test (hLRT) in MrModeltest 2.3. A bootstrap analysis with 1,000 replicates was used to test the statistical support of the branches. For the MrBayes 2 https://www.ncbi.nlm.nih.gov/genbank/ analyses, the best-fit model of nucleotide evolution was selected by Akaike information criterion (AIC; Posada and Buckley, 2004) in MrModeltest v. 2.3. The metropolis-coupled Markov Chain Monte Carlo (MCMCMC) approach was used to calculate posterior probabilities (Huelsenbeck and Ronquist, 2005). Trees were viewed in TREEVIEW. Phylograms obtained based on combined loci or for a single locus were all deposited in TreeBASE. The nucleotide sequences reported in this study were deposited in GenBank (Supplementary Table 1).

Molecular Phylogenetic Analysis
Based on the results of BLAST in GenBank, Botryosphaeria aterrima (Fuckel) Sacc., B. berengeriana var. weigelae Rehm and Botryosphaeria mirabile (Fuckel) Cooke belong to Botryosphaeria, B. cruenta and B. hamamelidis to Neofusicoccum, Botryosphaeria ferruginea to Nothophoma Qian Chen and L. Cai. The phylogenetic analysis of the Botryosphaeria dataset included 14 ingroup taxa and two outgroup taxa (Supplementary Figure 4). The combined ITS, LSU, tef1-a, and tub2 matrix contained 2,268 characters, of which 1,940 were constant and 39 were variable and parsimony-uninformative. Maximum parsimony analysis of the remaining 289 parsimony-informative characters resulted in 2,275 equally most parsimonious trees (Supplementary Figure 4). The phylogenetic tree resulting from the Bayesian analysis using the general time reversible model of nuDNA evolution (Rodríguez et al., 1990), including estimation of non-variable sites and assuming a discrete gamma distribution with six rate categories (GTR+ +G), had a topology identical to the MP tree presented. In both analyses (MP and Bayesian), the clade of Botryosphaeria had a high bootstrap support (100% for MP) and high posterior probabilities (1.00 for MrBayes  The analysis for Neofusicoccum involved 38 taxa including two outgroup species, i.e., B. corticis and B. dothidea. The combined ITS, tef1-α and tub2 nuDNA sequence matrix included 920 characters, 136 were constant and 39 were variable and parsimony-uninformative. Maximum parsimony analysis for the remaining 181 parsimony-informative characters resulted in 5,000 equally most parsimonious trees (Supplementary Figure 5, TreeBASE number S21059). The phylogenetic tree resulting from the Bayesian analysis using the general time reversible model of nuDNA evolution (Rodríguez et al., 1990), including estimation of invariable sites and assuming a discrete gamma distribution with six rate categories (GTR+ +G), had a topology identical to the MP tree presented. In both analyses (MP and Bayesian) the clade accommodating Neofusicoccum had a high level of support (100% for MP and 1.00 PP for MrBayes). Isolates of N.  The analysis for Nothophoma spp. involved 22 taxa including one outgroup species, i.e., Didymella calidophila. The combined ITS and LSU sequence matrix included 1,821 characters, of which 1,751 were constant and 38 were variable and parsimonyuninformative. Maximum parsimony analysis for the remaining 32 parsimony-informative characters resulted in 1,000 equally most parsimonious trees. For the Bayesian analysis, TNe+I was selected as the best-fit model for the ITS and LSU dataset, had a topology identical to the MP tree and ML tree presented. Phylogenetically, species of Nothophoma formed a robust clade. Isolates of No. ferruginea formed sub-clade representing an individual species of Nothophoma spp. Only the Bayesian tree is presented herein with MP, PP, and ML values plotted on the branches (Supplementary Figure 7).   Ascostromata not observed. Conidiostromata forming dense botryose aggregate, 2-7 mm diam., pseudothecial, aggregated into botryose clusters, 220-420 µm diam., spherical to globose with a central ostiole, 1 /2 to 3 /4 emergent, rarely embedded, black. Peridium comprising 7-15 layers of textura angularis, outer region of dark brown cells, inner region of 3-7 layers of pale brown cells lining the locule. Paraphyses when present hyaline, septate, up to 70 µm long, 2-4 µm broad at the base, tapering to acutely rounded apices, 1.5-2 µm broad at the tip. Conidiogenous cells holoblastic, hyaline, sub-cylindrical, 8-20 × 3-5 µm. Conidia hyaline, narrowly fusiform, or irregularly fusiform, base subtruncate to bluntly rounded, (40-)42-60(-62) × (7-)9-11 µm (−x = 51.3 × 9.9 µm, n = 20), L/W = 5.2.
Notes -Only the sexual morph was observed on the type material, the morphology of which is consistent with members of Botryosphaeriaceae in having gregarious ascomata, broadly clavate asci and hyaline, aseptate ascospores as well as lacking pseudoparaphyses. Only LSU sequence was obtained for the type material of Melanops cruenta in this study, and a few other species of Neofusicoccum have LSU sequences available from GenBank as well. The phylogenetic analysis based on these LSU sequences suggested that M. cruenta resides in Neofusicoccum, being sibling to other species in the genus (Supplementary Figure 6). Thus, we have assigned M. cruenta to Neofusicoccum as a new combination, N. cruenta.
Notes -Botryosphaeria aesculi was introduced as Laestadia aesculi, which was subsequently assigned to Botryosphaeria as B. aesculi by Barr (1972). Morphologically, the scattered and erumpent ascomata, small-sized and ellipsoidal asci differ from the species of Botryosphaeria sensu stricto, while the hyaline, aseptate and ellipsoidal to fusiform ascospores are consistent with taxa in the Botryosphaeriales. Herein we tentatively retain it within Botryosphaeria with its taxonomical status remaining to be undetermined until further phylogenetic analysis is carried out on verified specimens.
Notes -The erumpent botryose ascostroma, cellular pseudoparaphyses, hyaline, aseptate, large-sized ascospores suggest an affiliation in the Botryosphaeriaceae, while the ovoid to narrowly ellipsoid ascospores fit both Botryosphaeria and Neofusicoccum. Since there are no obvious morphological differences between Neofusicoccum and Botryosphaeria, the generic status of B. dasylirii remains uncertain until DNA sequence data can be obtained for it. Thus B. dasylirii is tentatively kept in Botryosphaeria sensu stricto herein until further phylogenetic analysis is carried out on verified specimens.
Notes -The bitunicate asci, cellular pseudoparaphyses, the hyaline, broadly ellipsoid ascospores with 3-4 transversal septa and 1-2 longitudinal septa in central cells of Melanops ferruginea differ from members of Botryosphaeriales. Phylogeny based on ITS and LSU nuDNA sequences indicated that M. ferruginea closely related to Nothophoma Qian Chen and L. Cai (Didymellaceae, Pleosporales). Therefore, we assign it to Nothophoma as N. ferruginea.
Notes -Botryosphaeria foliicola was introduced by Sivanesan and Nair (1988) from the leaves of Michelia nilgarica in India, which distinguishes it from other species of Botryosphaeria by its obovoid and characteristically constricted ascospores. Phyllachorella micheliae, the generic type of Phyllachorella, was reported from the leaves of the same host in India. The strong morphological similarity of Botryosphaeria foliicola and Phyllachorella micheliae warranted their conspecific status (Liu et al., 2012). Based on priority, the later name Botryosphaeria foliicola is reduced to synonymy with Phyllachorella micheliae, which is retained in Botryosphaeriales genera incertae sedis (Wijayawardene et al., 2020).
Four of the 17 taxa of Botryosphaeria sensu lato considered in the present study have been confirmed as members of Botryosphaeria sensu stricto, which include B. berengeriana var. acerina, B. aterrima, B. berengeriana var. weigelae, and B. mirabile with both B. berengeriana var. acerina and B. berengeriana var. weigelae reduced to synonyms of B. dothidea. Two other species of Botryosphaeria sensu lato were assigned in Neofusicoccum, viz., N. cruenta and N. hamamelidis. Neofusicoccum was separated from Botryosphaeria and introduced as a new genus based on combined multigene phylogenetic analysis and subtle morphological differences, i.e., pycnidial paraphyses only exist in Botryosphaeria (Fusicoccum), which have never been reported in any Neofusicoccum species (Crous et al., 2006;Phillips et al., 2013). Because of their morphological similarities, it is possible or even probable that some other species of Botryosphaeria sensu lato may actually more appropriately reside in Neofusicoccum.
Based on the phylogenetic analyses of combined ITS, LSU, tef1-a and tub2 loci, B. laricina and N. hamamelidis ( = B. hamamelidis) form a conspecific clade. The morphological characteristics of their sexual morphs also support their conspecific status (see comments above). Neofusicoccum hamamelidis was originally reported from the dead twigs of Hamamelidis virginiana in Canada, while B. laricina causes shoot blight of larch, which is one of the most important quarantine diseases in China (Liu et al., 2009). The conspecific status of N. hamamelidis and B. laricina supports its broad host range and wide distribution, and these will help in making practical quarantine rules, as a comprehensive knowledge as well as accurate identification of pathogens are extremely important when formulating quarantine regulations (Kumar et al., 2008).
The sexual morph of Botryosphaeria sensu lato is morphologically conserved (for example, in the size of the asci and ascospores), while the morphology of asexual morph and host association varies considerably more, which contributes to a natural classification of this group of fungi (Shoemaker, 1964;Pennycook and Samuels, 1985;Phillips et al., 2013;Slippers et al., 2014). For example, Botryosphaeria mucosa was assigned to Neodeightonia (as N. mucosa, Botryosphaeriaceae) based on its bambusicolous host association, aggregated ascostroma, shape of asci, ascospore shape and septation, as well as conidial morphology. Based on the morphological characteristics or DNA sequences comparisons, many specimens considered in this study should be excluded from Botryosphaeriales. For example, B. ferruginea in Nothophoma as N. ferruginea (Pleosporales), Botryosphaeria apocyni (Basionym: Laestadia apocyni), B. gaubae and B. smilacinina (Basionym: Sphaeria smilacinina) in Dothideomycetes incertae sedis. Thus, a polyphasic taxonomic approach should be applied in type studies of Botryosphaeria sensu lato, including the use of host association, morphological characteristics of both sexual and asexual morphs, geographical distribution, DNA sequences as well as epitypification where possible.
Based on both morphological characters and results of nuDNA sequence analysis, Botryosphaeria sensu stricto now includes ten species, namely B. agaves, B. aterrima, B. corticis, B. dothidea, B. fabicerciana, B. kuwatsukai, B. mirabile, B. qingyuanensis, B. ramosa and B. scharifii, of which B. fabicerciana and B. qingyuanensis have previously been reported from China. The current study shows that further studies are necessary on other type specimens of Botryosphaeria sensu lato in order to clarify their taxonomic status. Fresh collections are also needed to facilitate their epitypification.
In summary, Botryosphaeria sensu lato is highly polyphyletic, and species belong to various genera or families of Botryosphaeriales or even other orders within Dothideomycetes. Studying the type material of Botryosphaeria sensu lato helps to understand the circumscription of genera or families within Botryosphaeriales. Redescribing and obtaining DNA sequences of the type specimens makes it possible to epitypify those species and clarify their taxonomic status (Zhang et al., 2008). Of the 286 epithets within Botryosphaeria sensu lato, less than 20% have DNA sequences available from the type materials (Denman et al., 1999;Smith et al., 2001;Slippers et al., 2004;Ariyawansa et al., 2016;Zhang et al., 2021). Thus, further study is required to obtain a more natural classification for species presently accommodated in Botryosphaeria sensu lato.

DATA AVAILABILITY STATEMENT
The data presented in the study are deposited in the TreeBASE and GenBank repository, accession numbers are S21054 for Botryosphaeria, S21059 and S21050 for Neofusicoccum, GenBank accession are listed in Supplementary Table 1.

AUTHOR CONTRIBUTIONS
YZ designed the experiments. YZ and YPZ prepared the samples, conducted the molecular experiments, and analyzed the data. WS, LZ, DP-Z, PC, BS, and YD revised the manuscript. All authors contributed to the article and approved the submitted version.

FUNDING
This study was supported by the National Natural Science Foundation of China (General Programs, 31971658, 31770015, and 31370063) and NSFC Projects of International Cooperation and Exchanges (3155461143028).

ACKNOWLEDGMENTS
The curators of the herbarium G, W, CUP, F, K, MICH, and NYS are thanked for providing herbarium specimens on loan and for their assistance in locating specimens. YZ and YD thank the Beijing Forestry University for research support. Supplementary Table 1 | Species, specimens and GenBank accession numbers of sequences used in this study (newly generated sequences are indicated in bold).