﻿Two new species of Samsoniella (Cordycipitaceae, Hypocreales) from the Mayao River Valley, Guizhou, China

﻿Abstract Samsoniella species have been often found in the forest habitat and rarely found in special karst eco-environments, such as Tiankeng, valleys and caves. In this research, eleven cordyceps specimens were collected from Mayao River Valley. A known species (S.haniana) and two new species (S.duyunensis and S.vallis) were established and described according to a multilocus phylogenetic analysis and morphological characteristics. Our results provide insight that the richness of Samsoniella species in karst eco-environments and further attention should be paid to entomopathogenic fungi in such habitats.


Introduction
The genus Samsoniella Mongkols., Noisrip., Thanakitp., Spatafora & Luangsa-ard was proposed based on the phylogenetic analysis of Isaria-like morphs in Cordycipitaceae and characterised by oval to fusiform conidia and bright red-orange teleomorphic stromata and anamorphic synnemata by Mongkolsamrit et al. (2018).One Isaria-like species, Penicillium alboaurantium G. Sm. was transferred to Samsoniella and two new species, S. aurantia and S. inthanonensis were described.
Subsequently, Chen et al. (2020a) reported three new species, S. coleopterorum W.H. Chen, Y.F.Han & Z.Q.Liang, S. hymenopterorum W.H. Chen, et al. and S. lepidopterorum W.H. Chen, et al. from the forestry of Xishui and Rongjiang County, Guizhou Province, China.Those species have mononematous conidiophores rather than synnemata and associated with hymenopteran larvae, coleopteran larvae and lepidopteran pupae, respectively.Wang et al. (2020) described nine new species and a new combination from the forest habitats of Yunnan Province, China.The other thirteen new species were reported by Chen et al. (2021aChen et al. ( , 2022)), Wang et al. (2022) and Crous et al. (2023).Currently, there are twenty-nine species described in the genus of Samsoniella.Additionally, it has been reported that Samsoniella species are found in the forest habitat.However, the other ecological habitats, especially the karst eco-environment which has special niches like Tiankeng, valleys and caves should have insects and entomopathogenic fungi.In this research, eleven cordyceps specimens were collected from Mayao River Valley, Guizhou, China.After detailed multiloci phylogenic analysis and morphological observation, two new species and one known species were identified.

Specimen collection and identification
Eleven cordyceps specimens were collected from Mayao River Valley (26°22'8.3748"N,107°23'16.96"E),Duyun City, Qiannan Buyei and Miao Autonomous Prefecture, Guizhou, on 4 September 2021 and 30 July 2022.The samples were placed in an ice box and brought to the laboratory and preserved in refrigerator at 4 °C before use.The surface of each arthropod body was rinsed with sterile water, followed by sterilisation with 75% ethanol for 3-5 s and rinsing again three times with sterilised water.After drying on sterilised filter paper, a piece of the synnemata, mycelium or the sclerotia was cut from the specimen and inoculated on agar plates of potato dextrose agar (PDA) or PDA modified by the addition of 1% w/v peptone containing 0.1 g/l streptomycin and 0.05 g/l tetracycline (Chen et al. 2019a).After fungal colonies emerged from the inoculated samples, a piece of mycelium from the colony edge was transferred onto new agar plated and cultured at 25 °C for 14 days under 12 h light/12 h dark conditions (Zou et al. 2010).The specimens and axenic cultures were deposited at the Institute of Fungus Resources, Guizhou University (formally Herbarium of Guizhou Agricultural College; code, GZAC), Guiyang City, Guizhou, China.
Colony morphology was determined on PDA cultures incubated at 25 °C for 14 days and the growth rate, the presence of octahedral crystals and the colony colours (surface and reverse) were observed.To investigate the microscopic characteristics, a little of the mycelia was picked up from the colony and mounted in lactophenol cotton blue or 20% lactate acid solution and the asexual morphological characteristics (e.g., conidiophores, phialides and conidia) were observed and measured under a Leica DM4 B microscope.

DNA extraction, polymerase chain reaction amplification and nucleotide sequencing
DNA extraction was carried out using a fungal genomic DNA extraction kit (DP2033, BioTeke Corporation) according to Liang et al. (2011).The extracted DNA was stored at −20 °C.Polymerase chain reaction (PCR) was used to amplify genetic markers using the following primer pairs: ITS4/ITS5 for the internal transcribed spacer (ITS) region (White et al. 1990), LR0R/LR5 for 28s large subunit ribosomal (LSU) (Vilgalys and Hester 1990), CRPB1/RPB1Cr for RNA polymerase II largest subunit (RPB1) (Castlebury et al. 2004), fRPB2-5F/fRPB2-7cR for RNA polymerase II second largest subunit (RPB2) (Liu et al. 1999) and 983F/2218R for translation elongation factor 1 alpha (TEF) (Castlebury et al. 2004).The thermal cycle of PCR amplification for these phylogenetic markers was set up following the procedure described by Chen et al. (2021).PCR products were purified and sequenced at Sangon Biotech (Shanghai) Co.The resulting sequences were submitted to GenBank (Table 1).
The new strains or species are in bold type.

Sequence alignment and phylogenetic analyses
DNASTAR Lasergene (version 6.0) was used to edit DNA sequences in this study.The ITS, LSU, RPB1, RPB2 and TEF sequences were downloaded from GenBank, based on Mongkolsamrit et al. (2018), Chen et al. (2020aChen et al. ( , 2021aChen et al. ( , 2022)), Wang et al. (2020Wang et al. ( , 2022) ) and Crous et al. (2023) and others selected on the basis of BLASTn searches in GenBank.ITS sequences and other loci were aligned and edited by MAFFT v.7.037b (Katoh and Standley 2013) and MEGA6 (Tamura et al. 2013).Combined sequences of ITS, LSU, RPB1, RPB2 and TEF were obtained using SequenceMatrix v.1.7.8 (Vaidya et al. 2011).The model was selected for Bayesian analysis by ModelFinder (Kalyaanamoorthy et al. 2017) in PhyloSuite software (Zhang et al. 2020).ITS sequences, other loci and the combined loci were analysed using Bayesian inference (BI) and maximum likelihood (ML) methods.For BI, a Markov chain Monte Carlo (MCMC) algorithm was used to generate phylogenetic trees with Bayesian probabilities using MrBayes v.3.2 (Ronquist et al. 2012) for the combined sequence datasets.The Bayesian analysis resulted in 20,001 trees after 10,000,000 generations.The first 4,000 trees, representing the burn-in phase of the analysis, were discarded, while the remaining 16,001 trees were used to calculate posterior probabilities in the majority rule consensus tree.After the analysis was finished, each run was examined using the programme Tracer v.1.5(Drummond and Rambaut 2007) to determine burn-in and confirm that both runs had converged.ML analyses were constructed with IQ-TREE (Trifinopoulos et al. 2016), using an automatic selection of the model.

Genealogical Concordance Phylogenetic Species Recognition (GCPSR) analysis
The Genealogical Concordance Phylogenetic Species Recognition model was applied to analyse the related species.The pairwise homoplasy index (PHI) (Bruen et al. 2006) is a model test based on the fact that multiple gene phylogenies will be concordant between species and discordant due to recombination and mutations within a species.The test was performed in SplitsTree4 (Huson and Bryant 2006) as described by Quaedvlieg et al. (2014) to determine the recombination level within phylogenetically closely-related species using a two-locus concatenated dataset.The new species and their closely-related species were analysed using this model.The relationships between closely-related species were visualised by constructing a split graph, using both the Log-Det transformation and splits decomposition options.
The final value of the highest scoring tree was -15,629.246,which was obtained from the ML analysis of the dataset (ITS+LSU+RPB1+RPB2+TEF).The parameters of the GTR model used to analyse the dataset were estimated, based on the following frequencies: A = 0.235, C = 0.273, G = 0.270, T = 0.222; substitution rates AC = 1.00000,AG = 1.93319,AT = 1.00000,CG = 1.00000,CT = 4.27255 and GT = 1.00000; as well as the gamma distribution shape parameter α = 0.509.The selected models for BI analysis were SYM+G4 (ITS+L-SU+RPB1+RPB2+TEF).The phylogenetic trees (Fig. 1), constructed using the ML and BI analyses were largely congruent and strongly supported in most branches.Strains DY091021, DY091022, DY091031, DY091032, DY091151, and DY091152 were clustered into an independent subclade and formed a subclade with Samsoniella haniana Hong Yu bis, Yao Wang & Z.Q.Wang with high statistical support (100% ML /1 PP).Strains DY09161, DY09162, DY07501 and DY07502 were clustered into an independent clade with high statistical support (100% ML/1 PP).Strains DY07241, DY07242, DY091091 and DY091092 were clustered with S. aurantia in a clade with high statistical support in ML analysis (94% ML).

Discussion
Samsoniella species are widely distributed and commonly isolated from soil, insects and spiders or as a fungicolous (Mongkolsamrit et al. 2018;Chen et al. 2020aChen et al. , 2021aChen et al. , 2022;;Wang et al. 2020Wang et al. , 2022;;Crous et al. 2023) Chen et al. and S. hymenopterorum W.H. Chen et al. belonged to Coleoptera and Hymenoptera, respectively.In addition, the substrates of S. alboaurantia, S. farinospora and S. hepiali were soil, spider and fungi, respectively.Here, we reported Samsoniella species with two different hosts from the valley habitat.More Samsoniella species with different hosts or substrates will be reported from diverse habitats.
The taxonomic delimitation of Samsoniella was originally based on morphological characteristics and a multi-locus phylogenetic analysis.In the present study, the phylogenetic analysis of a single locus of an individual gene or gene fragment of ITS, LSU, RPB1, RPB2 and TEF was tested for the new species (Suppl.materials 1-5) and only the TEF could distinguish the new species.However, the new species S. vallis could not form an independent clade and clustered with S. aurantia as a subclade.A PHI test was added and could solve the taxonomic delimitation of S. vallis and S. aurantia.Thus, we recommend that the TEF locus should be used to distinguish the cryptic Samsoniella species and multiple approaches should be used for the further confirmation of a cryptic species.

Figure 1 .
Figure 1.Phylogenetic relationships amongst the new strains and their allies based on multigene dataset (ITS, LSU, RPB1, RPB2 and TEF).Statistical support values (≥ 70%/0.70)are shown at the nodes for ML bootstrap support/BI posterior probabilities.The new strains or species are in bold type.

Figure 2 .
Figure 2. Results of the pairwise homoplasy index (PHI) test of closely-related species using both LogDet transformation and splits decomposition.PHI test results (Փw) < 0.05 indicate significant recombination within the dataset.The new strains or species are in bold type.

Table 1 .
List of strains and GenBank accession numbers of sequences used in this study.