Sampling

In 2015 and 2016, five sites were sampled in peach and plum commercial orchards located at the upper, intermediate and lower part of the Zinapecuaro production area (Table 1). Trees were identified with different damage symptoms (leaf wilting, tree top yellowing, stunting, basal canker, neck exudate, lack of vigor and death) at each site. A destructive sampling was undertaken to collect samples of primary and secondary roots and trunks at a height of 15-30 cm. It was verified that pathogenic signs and symptoms were present in the root system of the tree, where in some cases rhizomorphs, typical vegetative structures of Armillaria, and whitish-brown mycelial fans between bark and wood were found. Soil samples were taken at a depth of 30-60 cm.

Table 1 Sites where samples of roots and trunks of Prunus sp. and soil were collected in the Zinapecuaro county, Michoacán, 2016. 

Processing plant and soil samples

Samples of plant material were thoroughly washed to remove soil; they were then cut into pieces 1 cm long, placed in 1.1 % sodium hypochlorite and gently shaken for 10 min. The pieces were then immersed in 70 % alcohol and washed with sterile distilled water between changes. The rhizomorphs were washed with 3 % hydrogen peroxide for 5 min, rinsed twice with sterile distilled water and dried with sterile paper towels, following the method of ^{Harrigton et al. (1992)}. Around 15-20 woodchips with mycelium and brown-to-black rhizomorphs were sown in Petri dishes containing BDS (Benomyl-Dicloran-Streptomycin) selective culture medium (^{Worrall, 1991}; ^{Aguín et al., 1998}). The dishes were covered with aluminum foil, placed in polypaper bags and incubated at 20 °C in darkness during 4-5 weeks.

To explore the mycoflora population present in soil at the collection sites, the soil dilution PDA-TS (potato-dextrose-agar-tergitol) technique was used (^{Steiner and Watson 1965}). The dishes were kept at a temperature of 22-24 °C for 7 to 10 days, period during which the fungi developed. Isolates were transferred to water-agar medium to be purified using the hyphae-point technique and sown individually in dishes containing water agar and PDA.

Molecular identification. DNA extraction, specific amplification and sequencing

After obtaining pure isolates of the fungi, they were molecularly characterized. DNA extraction was performed using the AxyPrep™ Multisource Genomic DNA Miniprep (Axygen) kit, according to the manufacturer’s instructions. PCR was performed using the ARMEF-3-A (CGT GAY TTY ATC AAG AAC ATG) and ARMEF-R (TAC CCG TTC GGC GAT CAA TCT) primers designed in the amplified region for the Tef-1α gen (translation elongation factor 1α) (^{Ross-Davis et al., 2012}). Amplified reactions were performed in a standard TProfessional thermocycler (Biometra, Germany) using a final volume of 25 µL with the following reagents: Taq DNA Buffer PCR 1X, MgCl2 1 mM, dNTP’s 0.2mM, Platinum Taq DNA polymerase 1U (Invitrogen, USA), 10 µM of each primer and 4 µL of total DNA. The amplification program was as follows: 96 °C for 3 min, 35 cycles at 95 °C for 30 s, 58 °C for 30 s, 72 °C for 45 s, and a final extension at 72 °C for 4 min. To verify the presence of the fragment electrophoresis in 2 % agarose gels stained with ethidium bromide (5 mg L^-1) was utilized; the gels were documented using the Gel Doc (Bio-Rad, USA) photo documentation system. The PCR products were purified and sequenced in both directions by the Macrogen Inc. (Seoul, South Korea).

Multiple alignment and phylogenetic analysis

Sequences of the TEF-1a gene in the isolates were edited and manually aligned using BioEdit 7.2.5 software (^{Hall, 1999}). The sequences were compared with those available in the National Center of Biotechnology Information (NCBI) database using the BLAST® algorithm. For the phylogenetic analysis, all the sequences of Armillaria spp. isolates from Michoacan were grouped in a single consensus sequence, and sequences of isolates of 34 Armillaria species characterized worldwide were used, according to the method of ^{Maphosa et al. (2006)}, ^{Hasegawa et al. (2010)}, ^{Ross-Davis et al. (2012)}, ^{Elías-Román et al. (2013)} and ^{Guo et al. (2016)}. Pleurotus ostreatus (NCBI accession number AY883432) was used as a root node outside the group. All the sequences were aligned with the CLUSTAL-W algorithm and analyzed by the Neighbor-joining (NJ) method (^{Saitou and Nei 1987}) using Molecular Evolutionary Genetics Analysis software (MEGA 7). One thousand support value replications (Bootstrap) were used to obtain a consensus tree with a 50 % majority rule (^{Tamura et al., 2004}). For this task, the Kimura 2 Replacement Parameter-2 Model (K2P) was used, according to the method of ^{Hasegawa et al. (2010)}.

Morphological identification

The strains collected in orchards 1, 2 and 5 (Table 1) from the bark of trees infected with fungal mycelium had a fan-like growth pattern and the rhizomorphs from roots showed a cottony creamy yellow growth in the culture medium. Rhizomorphs with dichotomic growth formed after 20 days and hyphae characteristic of the Armillaria genus formed at 28 days. No basidiocarps were found at the study sites before, during or after the rainy season.

Samples of plant material collected from orchards 3 and 4 did not include Armillaria spp. isolates. The genera Trichoderma sp., Pilatoporus sp., Mucor sp., Penicillium sp., Aspergillus sp. and Fomitopsis sp. were found in the soil isolates from all the sampling sites. Fomitopsis sp. is an associated pathogen that can cause trees to decay and die (^{Cibrián et al., 2007}). In orchard 5, the presence of Fusarium oxysporum was detected, a root pathogen that damages the vascular system.

Molecular identification. Multiple alignment and phylogenetic analysis

The DNA sequences obtained in this study do not have reliable coverage and have not been identified (≥ 99 %) as being one of the known species of Armillaria reported in NCBI; thus it appears to be a new species of the genus. However, the sequences were 99 % similar to those reported as being Armillaria spp. by ^{Elías-Román et al. (2013)}, based on isolates from the State of Mexico, concluded that root rot in peach trees is caused by a species of the Armillaria genus that has not yet been described. The sequences obtained in this study are kept in the GeneBank with accession numbers KY611443 to KY611450 and are available. These sequences and those reported by ^{Elías-Román et al. (2013)} are from the same organism, which belongs to a species not yet described in the literature.

Based on the alignment of 48 sequences of the EF-1α gene from at least 34 species of the Armillaria genus that have been characterized worldwide. Following the results reported by ^{Maphosa et al. (2006)}, ^{Hasegawa et al. (2010)}, ^{Ross-Davis et al. (2012)}, ^{Elías-Román et al. (2013)} and ^{Guo et al. (2016)}, a phylogenetic analysis was performed using the Neighbor-joining algorithm. A 50 % or higher probability consensus tree was obtained in which the Armillaria species were grouped into four main clades (Figure 1). The EF-1a gene was useful for separating all the species. The studied sequences were grouped in the same clade as the species Armillaria mellea, but in different branches (Figure 1). This confirmed that they are new species that appear to be phylogenetically related to A. mellea. However, the phylogenetic analysis, BLAST and the sequence analysis (single-nucleotide polymorphisms, insertions and deletions) all clearly distinguish both organisms as being different species, which agrees with the results reported by ^{Elías-Román et al. (2013)} using the isolates obtained in the State of Mexico.

Figure 1 Phylogenetics of Armillaria genus developed using the Neighbor-joining (NJ) algorithm inferred from sequences of the EF-1α gen, including isolates collected in Michoacan, Mexico. A similar topology was obtained using ML and MP. Pleurotus ostreatus was used as a root node outside the group. Bootstrap support values (1000 replications) for NJ (≥ 50%) are shown in the internodes with a blue circle. 

In the Armillaria genus, species grouped in the same clade as A. mellea and A. ostoyae are associated with or considered aggressive pathogens that usually cause diseases in various fruit and forest trees, compared with species that are saprophytes or that behave as weak pathogens and that are grouped in the other clades (^{Ross-Davis et al., 2012}; ^{Guo et al., 2016}). In this study, the phylogenetic analysis grouped the sequences in the same clade as A. mellea sequences that have been reported in North America, Asia and Europe.

In a parallel study, the cumulative incidence of dieback over three years in orchard 1 was calculated at 21.3 % (145/679 trees) (^{Rivas-Valencia and Fernández-Montes et al., 2017}). This could be interpreted as being an epidemic of great importance, caused by a species of Armillaria closely related to A. mellea.

The information about new Armillaria species in Mexico is based on reports from ^{Alvarado and Blanchette (1994)} on native species from forest regions of central Mexico showed that some of the isolates obtained in the state of Puebla were not compatible with reference isolates of species known in North America, which suggests that they belong to a different species. Due to the lack of sequences, it is not known if those isolates belong to the same biological species studied in this research. However, using biotechnological tools and, specifically, DNA sequence data, it is possible to identify the Armillaria species, which is essential to evaluate the threat of disease and its required management.