| Sordariomycetes,the second largest class within Ascomycota,exhibits remarkable species diversity.This class encompasses both agriculturally important pathogens such as Pyricularia oryzae,Fusarium spp.,and Colletotrichum spp.,as well as economically valuable fungi like Ophiocordyceps sinensis and Trichoderma atroviride,and those impacting human and animal health such as Sporothrix schenckii.The diverse topography and climate of Sichuan Province offer a myriad of potential habitats conducive to the diversity of Sordariomycetes.Nevertheless,the exploration and research into this area are still lacking.DNA barcoding technology is pivotal in the study of fungal taxonomy,yet the comprehension and application of this technology face limitations due to several factors,including the inconsistent use of barcodes,difficulties in evaluating the effectiveness of barcodes,and a lack of appropriate methods for quantitative assessment.In the context of higher-level taxonomic studies of Sordariomycetes,particularly at the family level and above,the resolving power of the quad-barcode approach(SSU-LSUTEF1-RPB2)proves insufficient,leading to an unstable taxonomic framework with numerous contentious groups.Moreover,there has yet to be an investigation into whether the phylogenetic trees generated from the quad-barcode system correspond with the natural evolution of Sordariomycetes.Grounded in the present state of taxonomic research of Sordariomycetes in Sichuan Province,the current understanding of DNA barcoding technology within the taxonomic community,and the extant scientific challenges within the higher-level taxonomic framework of Sordariomycetes,this study aims to undertake investigations focusing on resources,DNA barcoding,and the higherlevel taxonomic framework.This study conducted 33 field samplings in Sichuan Province,resulting in the collection and identification of 175 specimens related to Sordariomycetes.Through morphological studies,BLASTn analysis,and multi-gene phylogenetic analysis,the collected specimens were ultimately classified into 106 species,covering 5 subclasses,14 orders,27 families,and 58 genera.Among these species,there were 33 new species,19 new host records,and 10 new records for China.Additionally,this study established two new genera,Sporovalis and Cylindrocaputspora.Descriptions and morphological illustrations of sexual morphs were provided for 3 asexual species including Coniochaeta fibrosae,Clonostachys farinosa,and Apiospora jiangxiense;asexual morph for sexual species Neocosmospora tenuicristata.Based on the newly collected specimens and molecular data,13 known species were synonymized.From a methodological perspective,a successful quantitative and visual scheme for comparing trees’ differences was established using the genus Colletotrichum as a practical case study.This scheme includes two quantitative metrics: calculating Generalized Robinson-Foulds distances between evolutionary trees to measure the similarity of their topologies,and computing the proportion of high-support nodes within the trees to the total number of nodes to evaluate the stability.This process deepened the understanding of DNA barcoding technology in phylogenetic analysis,primarily encompassing the six points:(1)overall,most barcodes displayed stability in both the total number of sites and parsimony sites within the trimmed datasets,with minimal changes observed as the number of taxa increases.Conversely,the proportion of high-support nodes within the evolutionary trees decreases;(2)the molecular evolutionary models of barcodes generally remain stable,with little influence from the quantity of taxa;(3)in multilocus phylogenetic analysis,the order of barcode concatenations has minimal influence on the final results,comparable to the inherent algorithmic errors,which can be disregarded;(4)with an increase in the number of barcodes,the proportion of high-support nodes within trees increases,rendering the trees more stable;(5)barcode sequences are subject to varying degrees of incomplete lineage sorting,leading to potential conflicts in the provided evolutionary signals.These conflicts can be visualized through the proposed scheme;(6)due to incomplete lineage sorting,barcode schemes based on a small number of sequences,guided by genome evolution trees,show only modest improvements compared to existing barcode systems.This study completed the whole-genome sequencing,assembly,and annotation of15 Sordariomycetes strains,providing reference genomes for 15 families,with 5 being the first genomes within their respective families.Currently,there are a total of 4126 available genomes for Sordariomycetes.However,their distribution is uneven.At the class level,71.4% of the classes(5/7)have available genomes,while at the order level,this percentage drops to 49%(25/51),and at the family level,it decreases further to only37.7%(72/191).Based on the protein sequences of core BUSCO genes from these genomes,phylogenetic and time trees were constructed at the family level,resolving taxonomic or controversy issues for 7 groups:(1)transferring the genus Torrubiellomyces from the family Ophiocordycipitaceae to Clavicipitaceae;(2)assigning the genus Emericellopsis to the family Bionectriaceae;(3)supporting the placement of the genus Stanjemonium within the family Bionectriaceae;(4)recognizing the placement of the genus Hapsidospora within the family Bionectriaceae;(5)transferring the genus Valetoniellopsis from the family Niessliaceae to Sarocladiaceae;(6)assigning the order Amplistromatales to the subphylum Diaporthomycetidae;(7)supporting the placement of the genus Canariomyces within the family Chaetomiaceae.Additionally,this study systematically compared the phylogenetic trees generated from the four-barcode(SSULSU-TEF1-RPB2)approach with the genome-based phylogenetic trees.While the two were generally consistent at the subclass level,at the order and family levels,the fourbarcode phylogenetic trees failed to reproduce the evolutionary relationships between orders and families observed in the genome-based phylogenetic trees.The similarity coefficient of the topologies between the genome-based and multilocus phylogenetic trees was only 0.669.The proportion of high-support nodes in the genome-based phylogenetic tree was 95.6%,whereas in the multilocus phylogenetic tree,it was only 36.2%.Additionally,multiple branches in the four-barcode phylogenetic tree contained mixed and unrelated taxa.In summary,this study systematically investigated Sordariomycetes in Sichuan Province,discovering new species and records,thus significantly adding to the fungal resources of both Sichuan Province and China.Using the genus Colletotrichum as a practical case study,a successful quantitative scheme for comparing phylogenetic trees was established,deepening the understanding of DNA barcoding technology.This scheme provides a feasible evaluation method for the selection of DNA barcodes in future studies of various fungal groups.Furthermore,based on genome data,a stable phylogenetic tree at the family level was constructed for Sordariomycetes,resolving taxonomic or controversial issues within 7 groups.Additionally,the differences between the phylogenetic trees generated from the four-barcode system and the genome-based phylogenetic trees at the class,order,and family levels were compared,providing important references for future research on the higher-level taxonomy of Sordariomycetes. |
PDF Full Text Request