Plant seeds, leaves, and stem segments that were successfully used in previous studies were tested as baits to isolate Calonectria species from soil samples. These tested baits satisfied at least one of the following three criteria: (1) the bait had been used to successfully isolate Calonectria in a previous study; (2) Calonectria had been isolated from this plant species in a natural environment; and (3) the family to which the plant species belongs included a relatively high number of Calonectria host species (>10 species). Additionally, to isolate Calonectria within a short period of time, the baits were required to satisfy the following conditions: (1) successfully isolate Calonectria within no more than 15 days; and (2) in the case of seed baits, the seeds should be less than 5 mm in length to be accommodated in an isolation cup and germinate within 2 days. Based on these criteria, 24 baits (bait IDs 1–24) were selected (Table 1). These included seeds of nine plant species, leaves of 14 plant species, and soybean stem segments (Table 1). Furthermore, human hair, which was used to isolate Calonectria in a previous study (Griffin 1960), was included. Medicago sativa seeds, widely used to bait Calonectria spp. (Crous 2002; Lombard et al. 2015; Pham et al. 2022a; Liu et al. 2024), were included as a positive control. In all, a total of 26 baits were used in this study. Seeds were commercially sourced from Taobao Network Co., Ltd. (Yuhang District, Hangzhou City, Zhejiang Province, China). Plant leaves were collected from Fujian Agriculture and Forestry University (Jinshan Campus) (Cangshan District, Fuzhou City, Fujian Province, China). Soybean stem segments were collected from plants grown for 1–2 weeks after germination.

To obtain Calonectria with high species and genetic diversity, soil was collected from eight forests in five provinces and autonomous regions of China located in different climatic zones. These sites included forests at eight sites in Fujian (3), Hainan (2), Heilongjiang (1), Henan (1), and Xizang (1) (Table 2, Fig. 1). Details including the geographic region, climatic conditions, GPS coordinates, and altitude of these sites are provided in Table 2. Samples were collected from Xizang in October 2023, from Fujian and Hainan in January 2024, and from Heilongjiang and Henan from August to October 2024.

Soil samples were collected from natural forests in Henan, Heilongjiang, and Xizang and from E. urophylla × E. grandis plantations in Fujian and Hainan. The Eucalyptus trees were more than 5 years old, and both the natural forests and plantations had thick layers of leaf litter, which were removed before soil sample collection. Samples consisted of a mixture of soil from the surface to a depth of 50 cm. For each sample, approximately 0.02 m³ of soil was collected and placed in sturdy plastic bags. Samples were promptly transported to the laboratory and stored at 25 °C for baiting, isolation of Calonectria, and identification of isolates using DNA sequencing methods.

Prior to comparing different baits, soil samples were tested for the presence of Calonectria spp. To achieve this goal, a total of 40 soil samples were collected from the eight sites in five provinces (Table 2, Fig. 1). Each of these soil samples was mixed thoroughly. The mixed soil in each of the 40 bags was transferred to five plastic isolation cups (diameter = 4.4 cm, height = 5.5 cm, and volume = 80 mL; Chengdu Rich Science Industry Co., Ltd., Chengdu, China). The soil in these cups was moistened with sterile water and mixed well using a sterilized bamboo stick. Surface-disinfested M. sativa seeds were scattered onto the soil surface, and the samples were incubated at 25 °C. After 7 days of incubation, each isolation cup was examined under a dissection microscope for the presence of conidiophores with typical morphological characteristics of Calonectria (Crous 2002).

For each of the eight sampling sites, one soil sample shown to be positive for Calonectria was used to compare the isolation efficacy of 26 baits. In total, this provided eight soil samples from eight sites, including Fujian (3), Hainan (2), Heilongjiang (1), Henan (1), and Xizang (1), for bait testing (Table 2, Fig. 1).

For each of the eight soil samples, five replicate isolation cups were used for each of the 26 baits. This resulted in a total of 1,040 isolation cups (five isolation cups/bait/soil sample × 26 baits × eight soil samples = 1,040 isolation cups). Each of the 26 baits was dipped in 75% ethanol for 30 s and then washed three times with sterile water. The leaves were cut into 3–4 mm pieces using sterilized scissors. The soils placed in the isolation cups filled two-thirds of the cup volume and were moistened with sterile water and mixed well using a sterilized bamboo stick. Each of the 26 baits was placed on the soil surface in each isolation cup, and these were incubated at 25 °C.

All isolation cups were examined daily for 15 days for the presence of Calonectria conidiophores (Crous 2002; Li et al. 2022) using a dissection microscope. Where present, conidial masses were transferred using a sterilized needle onto 2% malt extract agar (MEA) containing 16 g malt extract (Qingdao Hope Bio-Technology Co., Ltd., Qingdao, China) and 16 g agar (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) per 800 mL of water. After incubation at 25 °C for 3–4 h, germinated conidia were transferred individually onto fresh MEA and incubated at 25 °C for 6–7 days to obtain single-conidium cultures. Two single-conidium cultures were obtained from each sampling cup, with each taken from a different position on the bait. All single-conidium cultures were deposited in the Culture Collection (CSFF) of the Forest Pathogen Center at the College of Forestry of Fujian Agriculture and Forestry University, Fuzhou, Fujian Province, China.

Single conidial isolates were grown on 2% MEA at 25 °C for 7 days; mycelium was scraped from the surface of the cultures using a sterilized bamboo skewer and transferred to a 2 mL Eppendorf tube. The total genomic DNA of each isolate was extracted using the CTAB method described by van Burik et al. (1998). Extracted DNA was dissolved by adding 30 μL of TE buffer (1 M Tris-HCl and 0.5 M EDTA, pH 8.0), and 2.5 μL of RNase (10 mg/mL) was added for RNA degradation. The mixture was incubated at 37 °C for 1 h. A NanoDrop Lite Plus spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) was used to measure the DNA concentrations. All DNA samples were diluted to approximately 100 ng/μL by adding DEPC-treated water (Sangon Biotech Co., Ltd., Shanghai, China) and stored at –4 °C for further use.

Based on previous research (Liu et al. 2020, 2021; Wu and Chen 2021; Liu and Chen 2023), four gene regions, namely, translation elongation factor 1-alpha (tef1), β-tubulin (tub2), histone H3 (his3), and calmodulin (cmdA), were chosen as reliable DNA barcodes to identify the Calonectria spp. These four gene regions were amplified using the primer pairs and PCR protocols described by Liu et al. (2020). All PCR products were sequenced in both directions using the same primers used for PCR amplification. The PCR products were sequenced by the Beijing Genomics Institute located in Guangzhou, China. All initial sequences were edited using MEGA v. 7.0 software (Kumar et al. 2016). All sequences obtained in this study were deposited in GenBank (http://www.ncbi.nlm.nih.gov).

The tef1 and tub2 gene regions were amplified and sequenced for all isolates. The tef1 and tub2 sequences were used in a standard nucleotide BLAST search in the NCBI database (https://blast.ncbi.nlm.nih.gov) for preliminary species identification and to determine the complexes in which the isolates reside. Based on preliminary identification, representative isolates for all the genotypes determined by tef1 and tub2 sequences were selected for PCR amplification and sequencing of the his3 and cmdA regions. Data for all isolates with tef1, tub2, cmdA, and his3 sequences were used for phylogenetic analyses. Sequences of the isolates from the type specimens of all published species in the relevant Calonectria species complexes were used in these analyses. The sequence datasets were aligned using the online version of MAFFT v. 7 (Katoh and Standley 2013) (http://mafft.cbrc.jp/alignment/server) with an FFT–NS–i strategy (slow; interactive refinement method). The aligned datasets were curated in MEGA v. 6.0.5 software (Tamura et al. 2013).

Sequence datasets for each of the four gene regions and a concatenated dataset for those regions were used for phylogenetic inference. PhyML v. 3.0 (Guindon and Gascuel 2003) was used to perform maximum likelihood (ML) analyses. Sequence data for two isolates of Curvicladiella cignea (CBS 109167 and CBS 109168) were used as outgroups (Liu et al. 2020). For phylogenetic analyses, the best substitution model for each of the five datasets was determined using JModeltest 2.1.7 (Posada 2008). Bootstrap support for ML analyses was estimated using 1,000 replicates to assess the reliability of the inferred phylogenies. Phylogenetic trees were viewed using MEGA v. 7.0. Based on phylogenetic analyses and sequence comparisons, the Calonectria isolates obtained in this study were identified.

The number of soil samples from which Calonectria spp. were obtained using the 26 baits was recorded. For each of the 26 baits, the percentage of sampling cups that successfully yielded Calonectria isolates was determined. The efficacy of obtaining Calonectria for both the genus as a whole and each species was computed and compared across the baits. At the genus level, the number of species in the genus (NSG) and number of isolates of the genus (NIG) isolated using each of the 26 baits were calculated. For each Calonectria species, the genotypes of the isolates were determined using the tef1 and tub2 sequences, and the number of genotypes of each species (NGS) and the number of isolates of each species (NIS) collected using each bait were calculated and compared. To determine whether the baits differed significantly and were equally influenced across climatic zones, analyses of variance (ANOVA) and multiple comparisons among means were conducted using SPSS Statistics 19 software (IBM Corp., Armonk, NY, USA) to compare the NSG and NIG obtained from each soil sample cup.

A total of 1,040 isolation cups were exposed to the 26 baits and examined for 15 days. No Calonectria fruiting bodies were found on the human hair samples. With the exception of Lolium perenne seeds and Buxus sinica leaves, which yielded conidiophores with morphological characteristics typical of Calonectria after 15 and 11 days, respectively, all other 23 baits displayed Calonectria sporulation within 10 days. Baits such as leaves of Dianthus caryophyllus, Pennisetum purpureum, and Rosa chinensis showed Calonectria sporulation within 4–5 days (Table 1). Calonectria spp. were found on baits in 641 of the 1,040 isolation cups tested. Other Nectriaceae, including Cylindrocladiella, Gliocladiopsis, and Neocosmospora, were also frequently observed on the baits. Taking two single-conidium isolates from each of these isolation cups resulted in a collection of 1,282 Calonectria cultures for further study (Suppl. material 1: table S1).

The tef1 and tub2 gene regions were sequenced for all 1,282 isolates (Suppl. material 1: table S1), and their genotypes were recorded. This resulted in 20 genotypes for tef1 sequences, 28 genotypes for tub2 sequences, and 41 genotypes for the combination of tef1 and tub2 sequences. Based on these 41 genotypes, one to four representative isolates of each genotype were further selected for sequencing of the his3 and cmdA gene regions.

The criteria for selecting isolates for his3 and cmdA gene sequencing were as follows: (1) four isolates were selected for each tef1-tub2 genotype in each province; (2) these isolates were obtained from different soil collection bags and isolated from different baits, as far as possible; and (3) for genotypes represented by fewer than four isolates, all isolates of that genotype were selected. Following this approach, a total of 142 isolates were selected for sequencing of the his3 and cmdA gene regions, and 49 genotypes were generated based on the tef1, tub2, cmdA, and his3 sequences, enabling their identification (Suppl. material 1: table S2). The remaining 1,140 isolates were identified based exclusively on the tef1 and tub2 sequences.

Phylogenetic analyses were performed using the 142 isolates representing 49 genotypes based on four gene sequences (Suppl. material 1: table S2). A standard nucleotide BLAST search using the tef1, tub2, cmdA, and his3 gene sequences indicated that the isolates belonged to three Calonectria species complexes. These were the Calonectria kyotensis species complex, Ca. colhounii species complex, and Ca. reteaudii species complex. The sequences of 101 isolates (ex-type and other strains) of 57 Calonectria spp. in these three species complexes were retrieved from GenBank (Suppl. material 1: table S3). These sequences were used in phylogenetic analyses of the four individual gene regions and a combination of these four gene regions.

The sequenced isolates yielded approximately 500 bp for the tef1, 565 bp for the tub2, 440 bp for the his3, and 685 bp for the cmdA gene regions. The model for each gene region dataset was selected based on JModeltest v. 2.1.5. The TIM2+G, HKY+I+G, TIM1+G, TIM3+I+G, and GTR+I+G models were selected for tef1, tub2, cmdA, his3, and the concatenated dataset, respectively. ML trees displaying the bootstrap values from ML analyses are shown in Suppl. material 1: figs S1–S5.

Based on the combined gene phylogenies (Suppl. material 1: fig. S1), 142 isolates clustered in eight groups (Groups 1–8). Of these, 128 isolates belonged to the Ca. kyotensis species complex and were identified as six species (Groups 1–6; Suppl. material 1: fig. S1). A total of 29 isolates (including CSFF27699) grouped with Ca. hongkongensis (Group 1), 46 (including CSFF27794) with Ca. kyotensis (Group 2), nine (including CSFF28233) with Ca. chinensis (Group 3), four (including CSFF27818) with Ca. pacifica (Group 4), 22 (including CSFF28421) with Ca. aconidialis (Group 5), and 18 (including CSFF31784) with Ca. canadiana (Group 6). Five isolates (including CSFF28154) were in the Ca. reteaudii species complex and clustered with Ca. pseudoreteaudii (Group 7). Nine isolates (including CSFF31597) belonged to the Ca. colhounii species complex and grouped with Ca. lichi (Group 8).

All 1,282 Calonectria isolates identified based on two to four gene regions were identified as Ca. hongkongensis (370 isolates; 28.9%), Ca. aconidialis (274 isolates; 21.4%), Ca. canadiana (233 isolates; 18.2%), Ca. pacifica (181 isolates; 14.1%), Ca. kyotensis (157 isolates; 12.2%), Ca. chinensis (23 isolates; 1.8%), Ca. lichi (35 isolates; 2.7%), and Ca. pseudoreteaudii (nine isolates; 0.7%) (Fig. 2). The first six species belonged to the Ca. kyotensis species complex. Calonectria lichi resides in the Ca. colhounii species complex, and Ca. pseudoreteaudii is part of the Ca. reteaudii species complex.

Of the 1,282 Calonectria isolates identified as eight species, 41 tef1-tub2 genotypes were generated based on the tef1 and tub2 gene sequences. These included Ca. hongkongensis (7), Ca. aconidialis (7), Ca. canadiana (4), Ca. pacifica (1), Ca. kyotensis (14), Ca. chinensis (3), Ca. lichi (3), and Ca. pseudoreteaudii (2).

The percentages of isolation cups that yielded Calonectria in Fujian, Hainan, Heilongjiang, Henan, and Xizang were 79.2%, 62.3%, 73.1%, 30.0%, and 27.7%, respectively (Table 3). The number of isolates and their identifications are provided in Table 3. Calonectria hongkongensis was isolated from Hainan and Fujian soil samples, Ca. aconidialis from Fujian and Xizang, and Ca. canadiana from Heilongjiang and Henan. All Ca. pacifica and Ca. kyotensis isolates were obtained from Fujian, and all Ca. chinensis and Ca. pseudoreteaudii isolates were from Hainan. Calonectria lichi was predominantly isolated from Henan, with a small number from Heilongjiang (Suppl. material 1: fig. S6). Other than Ca. canadiana and Ca. lichi, which were found in both mid-temperate (Heilongjiang) and warm-temperate zones (Henan), all other species were found in tropical (Hainan) and subtropical (Fujian and Xizang) regions (Table 2; Suppl. material 1: fig. S6).

Results showed that the efficacy of the baits was not significantly influenced across tropical, subtropical, warm temperate, and mid-temperate zones. They also showed that there was a proportional relationship between the number of soil samples and the number of different fungal species isolated (Table 3). Statistical analyses of multiple comparisons indicated that NSG and NIG values from each soil sample cup were not significantly influenced by the climatic regions where soil samples had been collected (Suppl. material 1: figs S7, S8). For example, both the NSGs and NIGs from Hainan (tropical zone) were significantly (P < 0.05) greater than those in Henan (warm temperate zone). In contrast, the NSGs and NIGs of Xizang (subtropical zone) were significantly (P < 0.05) smaller than those of Heilongjiang (mid-temperate zone) (Suppl. material 1: figs S7, S8).