Results

Phylogenetic analyses

Analyses 1: The phylogenetic tree (Fig. 1) was generated to determine the relationship among the new strains (SQ57131 and SQ57132) and their related species within the genus Arachnidicola. Kanoksria zaquensis (Y.Hui Wang et al.) Khons. et al. (HMAS 246915 and HMAS 246917) was used as the outgroup taxon in the analysis. The dataset included 19 taxa and consisted of 3,653 (ITS: 515 bp, LSU: 813 bp, rpb1: 573 bp, rpb2: 891 bp, and tef-1α: 861 bp) characters with gaps.

Figure 1.: Phylogram retrieved from IQTREE of the new species and related species in Arachnidicola using the combined dataset of ITS, LSU, rpb1, rpb2 and tef-1α gene regions. The statistical values are provided at nodes as ML/PP (ML value above 70% and BI value above 0.70). The tree is rooted with Kanoksria zaquensis (HMAS 246915 and HMAS 246917). Ex-types and new strains are indicated by the superscript “T” and in bold, respectively.

For the ML analysis, the optimal substitution model was selected using ModelFinder (Kalyaanamoorthy et al. 2017) under the BIC criterion, and the selected model was TN+F+G4. The final log-likelihood score of the best-scoring ML tree was –11,206.652. The parameters of the GTR model used to analyze the dataset were estimated based on the following frequencies: A = 0.239, C = 0.284, G = 0.271, T = 0.207; substitution rates AC = 1.00000, AG = 2.43038, AT = 1.00000, CG = 1.00000, CT = 5.62105 and GT = 1.00000, as well as the site proportion and rates: (0.907, 0.538) and (0.093, 5.488). For the BI analysis, the selected models were GTR+F+I+G4 for the ITS, LSU, rpb1 and tef-1α partitions, and K2P+G4 for the rpb2 partition. The MCMC analysis was run for 10,000,000 generations with two independent runs; the final average standard deviation of split frequencies was 0.007, well below the recommended threshold of 0.01, indicating satisfactory convergence. The phylogenetic tree (Fig. 1) reconstructed from both ML and BI analyses was largely congruent, with strong support for most branches.

The phylogenetic tree (Fig. 1) revealed a well-resolved topology within the genus Arachnidicola, with most interspecific relationships receiving robust support. Arachnidicola fodingshanensis (SQ57131 and SQ57132) was recovered as a distinct lineage, forming a sister relationship with the clade comprising A. kanyawimiae (Mongkols. et al.) Khons. et al. (TBRC 7244 and TBRC 7242) and A. sinensis Wan H. Chen et al. (ZY06511^T^ and ZY06512). These three species, together with A. carrolliae Y.P. Tan et al. (MST FP3895), A. hookerae Y.P. Tan et al. (MST FP3877), A. araneogena (Z.Q. Liang et al.) Khons. et al. (GZUIF SN1 and GZUIF DX1), A. zunyiensis Wan H. Chen et al. (ZY06061 and ZY06062), A. tiankengensis (Wan H. Chen et al.) Khons. et al. (KY11571), A. bashanensis (Wan H. Chen et al.) Khons. et al. (CQ05621 and CQ05622), and A. beibeiensis (Wan H. Chen et al.) Khons. et al. (CQ05921 and CQ05922), were recovered within a strongly supported major clade (100% ML/1.00 PP), indicating a close phylogenetic affinity among these taxa despite the poor support for the basal node uniting A. fodingshanensis, A. kanyawimiae, and A. sinensis.

Analyses 2: The phylogenetic tree (Fig. 2) was generated to determine the relationship among the new strains (SQ57101, SQ57102, SQ57111 and SQ57112) and their related species within the genus Simplicillium. Beauveria bassiana (Bals.-Criv.) Vuill. (ARSEF 1564) and B. brongniartii (Sacc.) Petch (ARSEF 617) were used as the outgroup taxa in the analyses. The dataset included 40 taxa, and consisted of 3,076 (ITS: 586 bp, LSU: 847 bp, rpb1: 711 bp, and tef-1α: 932 bp) characters with gaps.

Figure 2.: Phylogram retrieved from IQTREE of the new species and related species in Simplicillium using the combined dataset of ITS, LSU, rpb1 and tef-1α gene regions. The statistical values are provided at nodes as ML/PP (ML value above 70% and BI value above 0.70). The tree is rooted with Beauveria bassiana (ARSEF 1564) and Beauveria brongniartii (ARSEF 617). Ex-types and new strains are indicated by the superscript “T” and in bold, respectively.

The selected model for the ML analysis was TNe+G4. The final value of the highest scoring tree was –16,476.563, which was obtained from the ML analysis of the dataset. The parameters of the GTR model used to analyze the dataset were estimated based on the following frequencies: A = 0.233, C = 0.278, G = 0.265, T = 0.224; substitution rates AC = 1.00000, AG = 2.12145, AT = 1.00000, CG = 1.00000, CT = 4.73680 and GT = 1.00000, as well as the site proportion and rates are (0.756,0.277) and (0.244,3.242). The selected model of the dataset for BI analysis was GTR+F+I+G4 (ITS, rpb1 and tef-1α), and K2P+G4 (LSU). The MCMC analysis was run for 10,000,000 generations with two independent runs; the final average standard deviation of split frequencies was 0.005, well below the recommended threshold of 0.01, indicating satisfactory convergence. The phylogenetic tree (Fig. 2) revealed a largely congruent topology with strong support for most branches. Within the genus Simplicillium, all included species formed well-resolved, monophyletic lineages, each with high statistical support (≥ 90% ML/≥ 0.95 PP). Notably, Simplicillium fodingshanense (SQ57101 and SQ57102) and S. shiqianense (SQ57111 and SQ57112) were recovered as two distinct, well-supported independent lineages (100% ML/1.00 PP), clearly separated from each other and from all other known species. These two novel species, together with Simplicillium subtropicum Nonaka et al. (JCM 18180, JCM 18181, JCM 18182, and JCM18183), formed a strongly supported subclade (100% ML/1.00 PP), indicating a close phylogenetic affinity among them.

GCPSR analysis

A three-locus concatenated dataset (ITS: 515 bp, LSU: 813 bp, tef-1α: 861 bp, containing 12 parsimony-informative sites) was used to assess the recombination levels within the Arachnidicola species, including A. fodingshanensis (SQ57131), A. araneogena (GZUIF DX1), A. bashanensis (CQ05621), A. beibeiensis (CQ05921), A. carrolliae (MST FP3895), A. hookerae (MST FP3877), A. kanyawimiae (TBRC 7244), A. sinensis (ZY06511), Arachnidicola sp. (KY47341), A. tiankengensis (KY11571), and A. zunyiensis (ZY06061) (Fig. 3). This three-locus combination (ITS + LSU + tef-1α) was selected because these loci were consistently available across all closely related reference strains and the newly described taxa, whereas rpb1 and rpb2 sequences were incomplete for a substantial proportion of the taxa within the Simplicillium species, which would have led to reduced taxon sampling and potential biases in the recombination test if included. The use of three independent loci for PHI analysis follows the methodological recommendations of Bruen et al. (2006) and Quaedvlieg et al. (2014), who demonstrated that the pairwise homoplasy index is robust for detecting recombination with as few as three unlinked loci, provided that sufficient parsimony-informative sites are present. Another three-locus concatenated dataset (ITS: 576 bp, LSU: 770 bp, tef-1α: 924 bp, containing four parsimony-informative sites) was used to determine the recombination level within the Simplicillium species, including S. fodingshanense (SQ57101), S. bursae (ZY06121), S. shiqianense (SQ57111), and S. subtropicum (JCM 18180) (Fig. 4).

Figure 3.: Results of the pairwise homoplasy index (PHI) test of the new strains and their closely-related species using both LogDet transformation and splits decomposition. PHI test results (Φw) < 0.05 indicate significant recombination within the dataset. The new strains are in bold type.

Figure 4.: Results of the pairwise homoplasy index (PHI) test of the new strains and their closely-related species using both LogDet transformation and splits decomposition. PHI test results (Φw) < 0.05 indicate significant recombination within the dataset. The new strains are in bold type.

According to Chaiwan et al. (2022), a pairwise homoplasy index (PHI) below the 0.05 threshold (Φw < 0.05) indicates significant recombination in the dataset, suggesting that related species within a group do not differ in recombination levels. Conversely, a PHI above 0.05 (Φw > 0.05) indicates non-significant recombination, implying that related species at the group level are distinct. In this study, the PHI test result for the Arachnidicola species listed above was 0.6876, and for the Simplicillium species was 0.1587. Both results confirm that the new species are distinct from their respective relatives.