Samples of pecan leaves and shuck displaying typical anthracnose symptoms were collected from pecan trees in Jiande, Zhejiang Province; Ji’an, Jiangxi Province; and Yuxi, Yunnan Province in August 2018 and 2019 ( Table 1 ). This period generally covers the early nut maturation and filling stages. Sampling followed the procedures described by Cai et al. (2009) and Prihastuti et al. (2009). Briefly, leaf and fruit surfaces were rinsed with distilled water, disinfected with 75% alcohol for 30 s, disinfected with 3% sodium hypochlorite for 2–3 min, rinsed with sterile water thrice, and dried with sterilized paper. Pieces were sliced from the junctions of diseased and healthy tissues, which were cut into squares with a side length of 3–4 mm. The excised tissues were placed onto potato dextrose agar (PDA, 20% diced potato, 2% glucose, and 1.5% agar in distilled water) plates and incubated at 28°C.Conidia were also collected, suspended in sterilized water, diluted to a concentration of 1 × 10⁴ conidia per mL, and spread onto the surface of water agar (WA, 2% agar in distilled water) to generate discrete colonies (Choi et al., 1999). Six single colonies of each isolate were picked up with a sterilized needle (insect pin, 0.5 mm diameter) and transferred onto PDA plates.

The isolates were purified using single spore isolation and stored at 4°C on PDA slants for further use. They were stored on filter paper at -80°C for long-term preservation. Isolates were transferred from PDA slants to PDA plates and cultivated at 28°C under a 12-h photoperiod for 14 days. The following morphological characteristics were recorded: conidia, appressoria, conidiomata, and conidiophores. The mean lengths and widths of 100 randomly selected conidia from each isolate were measured using 100× magnification in a microscope (Nikon Ti-S inverted microscope, Japan). Among the 36 isolates thus obtained, 11 representative isolates were selected for further multi-locus phylogenetic analyses based on geographical location, morphology (e.g., colony shape and color and other physical characteristics of aerial mycelia and conidia), and ITS sequences.

The genomic DNA of each isolate was isolated from 0.5 g of fresh hyphae using a DNA extraction kit (TaKaRa Bioengineering Co., Ltd, Dalian, China) and stored at −20°C. The ITS, GAPDH, ACT, TUB2, CHS-1, and CAL sequences were amplified and sequenced as previously described in Weir et al. (2012). The primers used, along with their respective sequences, are presented in Table 2 .

The PCR was performed according to the methods of Weir et al. (2012). All the PCR reactions were conducted in 25-μl volumes containing 12.5 μl PCR MasterMix (TIANGEN BIOTECH (BEIJING) CO., LTD., Beijing, China), 10 μM primers (both), 1 μl of the template DNA (20 ng μl^-1), and 9.5 μl of double-distilled H₂O. The PCR program to amplify ACT and GAPDH included a denaturation step at 94°C for 5 min, followed by 35 cycles of 94°C for 45 s, 59°C for 30 s, and 72°C for 2 min. The final cycle comprised 72°C for 10 min. The amplifications of ITS, TUB2, CHS-1, and CAL were also performed similarly with annealing temperatures of 56, 58, 58, and 57°C, respectively. The amplification products were analyzed on a 1.0% agarose gel in tris/borate/EDTA buffer. The PCR products were then purified and sequenced at Hangzhou Shangya Biotechnology Co., Ltd (Zhejiang, China).

The reference sequences for ITS, GAPDH, CHS-1, ACT, TUB2, and CAL were downloaded from NCBI ( Table 3 ). The forward and reverse sequences were edited and assembled using DNAMAN v. 8.0 (Lynnon Biosoft). According to the sequence of ITS-ACT-CAL-CHS-1-GAPDH-TUB2, six loci were combined and aligned using BioEdit v. 7.2.5 (Hall, 1999). Bayesian analyses were performed on concatenated alignments using MrBayes v. 3.2.1 (Ronquist et al., 2012). Maximum likelihood (ML) analyses were performed on the multilocus alignment using IQ-TREE (Nguyen et al., 2015), and the nucleotide substitution models were selected by Model Finder (Kalyaanamoorthy et al., 2017).

Same-sized healthy lesion-free leaves were collected from ‘Mahan’ subjects growing in a disease-free orchard located in Jiande, Zhejiang Province. Fruits were collected in the same manner as the leaves a month before harvest. Pathogenicity tests were conducted on pecan following the methods described by Kanchana-udomkan et al. (2004); Than et al. (2008), and Fu et al. (2019). Healthy pecan leaves and fruits were separately surface sterilized in 1% NaClO for 5 min, washed twice with sterile-distilled water, and air dried on sterile filter paper. Each fruit was inoculated with 20 μl of a conidial suspension (1 × 10⁷ conidia/ml). The suspension was injected into the surface of non-wounded fruit using a microsyringe (Eppendorf, Shanghai, China). Control leaves and fruits were treated with 20 μl of distilled water. Each isolate was inoculated into five replicate fruits. The inoculated fruits were incubated in a moist chamber at 28°C and were examined daily for symptoms for 9 d.

Virulence assay was performed in vivo on seven Colletotrichum spp. identified as representative isolates. These were selected according to the species, orchard province, and morphological characteristics. The virulence assay was carried out in a greenhouse of the Institute of Subtropical Forestry, Chinese Academy at Forestry Sciences (Zhejiang, China) from late June to August 2020. During the 5-year black spot resistance investigation on different cultivars preserved in the Pecan Resource Garden in Jiande City, Zhejiang, China, we found highly resistant cultivars, such as ‘Kanza’, and susceptible cultivars, such as ‘Mahan’. In this experiment, annually grafted container seedlings of the highly resistant variety, ‘Kanza’, and the susceptible variety, ‘Mahan’, served as plant materials. The fourth, fifth, and sixth pairs (from bottom to top) of pinnate compound leaves were selected, surface sterilized with 75% ethanol, and rinsed with sterile water. Each leaf was impaled at its middle near the midvein and wounded five times in a 5 mm region with a sterilized needle. The 1 × 10⁷ ml^-1 spore suspension was dropped on the wounds using a pipette. Some wounded leaves also served as controls and were inoculated with the same amount of sterile water instead of the spore suspension. Sterile damp absorbent cotton was placed on the pinnate compound leaves after air-drying, and sealed bags were used to retain the moisture. The experiment was set up with three biological replicates per isolate. The bags and cotton were removed after two days. The pathogenicity of each isolate was determined by evaluating the diameters of the disease lesions after 18 days. The one-way analysis of variance was performed with SPSS v. 16.0 software; means were compared using Duncan’s test at the significance level of 0.05.

From 2018 to 2019, pecan leaves and shuck displaying anthracnose symptoms were collected from three pecan orchards in three provinces of China. Colletotrichum isolates associated with pecan anthracnose were collected from 60 affected pecan samples. Leaves were collected for fungal isolation, resulting in a total of 36 Colletotrichum isolates identified based on morphology and ITS sequences. A total of 11 representative isolates were chosen for further analyses based on their morphology (colony shape, color, and conidial morphology), ITS sequences, types of symptoms, and origin ( Table 1 ). At least two isolates from each field were chosen for further analysis.

In total, 36 Colletotrichum isolates were obtained from three pecan orchards. Eleven single-spore representative isolates were used for the following morphological characterization and phylogenetic analyses ( Table 4 ). These comprised two C. fructicola isolates from leaves and six C. fioriniae and three C. fructicola isolates from the shuck.

Phylogenetic analyses on the six loci (ITS, ACT, CAL, GADPH, CHS-1, and TUB2) of the Colletotrichum spp. included 11 isolates. The sequences of ITS, GAPDH, ACT, TUB2, CHS-1, and CAL genes of Colletotrichum spp. isolates from pecan were deposited to Genbank ( Table 4 ). They were compared with reference sequences of Colletotrichum isolated from other countries and plant hosts available on GenBank (Table 3 ). The C. boninense (CBS 123755*) was used as the outgroup. For the maximum likelihood inference, the best fit model for the six loci was set at HKY+I+G4+F with UFbootstrap 20000. For the Bayesian inference, the best fit model was HKY+F+G4 with a gamma rate for six loci. The posterior probabilities of the Bayesian tree were consulted to confirm the topology of the maximum likelihood tree. Phylogenetic analysis provided enough information to distinguish two Colletotrichum species; six and five isolates belonged to C. fioriniae and C. fructicola, respectively ( Figure 2 ). Among the seven isolates collected from the orchards in Jiande, six were C. fioriniae and one was C. fructicola. The four isolates from the orchards in Yunnan and Jiangxi were all C. fructicola ( Figure 2 ).

The diameter of the colonies was measured daily for five days to calculate their mycelial growth rate (mm/d). The shape, color, and density of colonies were recorded after 14 days. Differences in colony morphology were observed between the two species identified when grown on PDA. Colonies generally showed dense, white to greyish or red growth ( Figure 3 ). The mycelia of C. fructicola appeared dark grey on PDA plates after 14 days; their orange conidial masses were yellow and cylindrical or oval with an area of 16.8 ± 1.5 × 5.4 ± 0.4 μm². Colletotrichum fioriniae was pinkish in color after 14 days of growth on PDA, and the conidial masses were orange. Its conidia were spindle-shaped with an area of 15.5 ± 1.5 × 5.2 ± 0.3 μm². The mycelium growth rate varied from 12.3 to 15.6 mm/day (average = 14.6 mm/day) for C. fructicola and 10.3 to 13.2 mm/day (average = 11.24 mm/day) for C. fioriniae.

All the Colletotrichum isolates were pathogenic to the detached pecan leaves and shucks. The inoculated leaves, pericarps, and nuts showed necrotic spots, while these tissues remained healthy in the controls. No lesions were induced in the control tissues that had been inoculated with sterile water. The morphology of fungal colonies re-isolated from the symptomatic shuck was the same as those produced by the original isolate used for inoculation, satisfying Koch’s postulate ( Figure 3 ).

In the virulence assay, the seven isolates caused symptoms on the leaves of 1-year-old pecan seedlings and showed different levels of virulence. The lesion diameters on the attached leaves start to differ 18 days after inoculation with various isolates ( Figures 4A, B ). No significant differences were observed in the lesion sizes produced on the highly resistant variety, ‘Kanza’, by the various isolates (p = 0.29), while there were significant differences on the susceptible variety, ‘Mahan’ (p = 0.001), indicating different virulence levels among the seven isolates. PCJD179 was noted as the strongest.

The virulence assay showed that the lesions caused by Colletotrichum spp. Were mainly restricted to the inoculated areas of the pecan variety, ‘Kanza’, and their diameters did not extend beyond 4 mm ( Figure 4A ). There were significant differences in the infection lesions on the ‘Mahan’ variety. The lesion diameters caused by PCJD179, PCJD32, and PCJD29 isolates of C. fioriniae and the PCJD7536 isolate of C. fructicola all exceeded 5.1 mm on ‘Mahan’. Among them, the diameter of the lesion caused by the isolate PCJD179 reached up to 10 mm, indicating that the isolate was relatively the higher virulent ( Figure 4B ).

Additionally, different isolates of Colletotrichum species displayed pathogenic differences. For example, the C. fioriniae isolates PCJD179, PCJD32, and PCJD29 produced large necrotic spots (5.1–10.0 mm). However, the spot diameters caused by PCJD12 in this species were 4.0 mm, and they were concentrated near the inoculation point. The spots produced by C. fructicola isolate, PCJD7536, were also significantly larger than those produced by isolates, PCJX073 and PCYN1751.