The collections were discovered with a cultivator, digging around trees up to a depth of 15 cm. All encountered fruiting bodies were photographed fresh and then dried at 50 °C. The chosen material was cut by hand and rehydrated with 5% KOH for morphological studies. Thirty spores were measured. Peridial slices were made and observed under optical microscopy (Castellano et al. 1989). For scanning electron microscopy pictures (JSM5600LV, JOEL, Tokyo, Japan), the spores were coated with gold and palladium. Voucher collections are deposited at José Castillo Tovar (ITCV) Herbarium, Instituto Tecnólogico de Ciudad Victoria, Mexico.

Genomic DNA was obtained with CTAB (Martínez-González et al. 2017)) or using Fungal DNA extraction Kit (Bio Teke Corporation, China) from 2–3 mg of dry tissue. DNA quantification was performed with Nanodrop (Thermo, USA). Each sample was diluted to 20 ng/uL for PCR amplification. LR0R and LR5 primers were used to amplify the LSU gene (Vilgalys and Hester 1990). The PCR reaction contained the following: enzyme buffer 1×, Taq DNA polymerase, 0.8 mM deoxynucleoside triphosphates (0.2 mM each), 100 ng DNA, 20 pmol of each primer, and 2 units of GoTaq DNA (Promega, USA), with a final volume of 15 µL. The amplification program was run as follows: denaturalization at 96 °C for 2 min, 35 cycles of denaturalization at 94 °C for 1 min, annealing at 57 °C for 1 min, polymerization at 72 °C for 1 min, and final elongation at 72 °C for 5 min. All PCR reactions were carried out in a Peltier Thermal Cycler PTC-200 (BIORAD, Mexico). The PCR products were verified by agarose gel electrophoresis run for 1 h at 95 V cm^-3 in 1.5% agarose and 1× TAE buffer (Tris Acetate-EDTA). The products were then dyed with GelRed (Biotium, USA) and viewed in a transilluminator (Infinity 3000 Vilber, Loumat, Germany). Finally, the products were purified using the ExoSap Kit (Affymetrix, USA) according to the manufacturer’s instructions and were prepared for the sequencing reaction using the BigDye Terminator Cycle Sequencing Kit v. 3.1 (Applied BioSystems).

Sequencing was carried out in a genetic analyzer (Model 3130XL, Applied BioSystems, USA) at the Biology Institute of the National Autonomous University of Mexico (UNAM). The sequences of both strains of each sample were analyzed, edited, and assembled using BioEdit v. 1.0.5 (Hall 1999) to create consensus sequences. The consensus sequences were compared with those in the GenBank database of the National Center for Biotechnology Information (NCBI) using the BLASTN 2.2.19 tool (Zhang et al. 2000). The LSU region was aligned using the online version of MAFFT v. 7 (Katoh et al. 2002, 2017; Katoh and Standley 2013). The alignment was revised in PhyDE (Müller et al. 2005), and small manual adjustments were then made to maximize the similarity between characters. The matrix was composed of 30 taxa (640 characters) (Table 1). Ramaria gelatinosa (access number AF213091) was used as the outgroup. The phylogeny was performed using Bayesian inference in MrBayes v. 3.2.6 64× (Huelsenbeck and Ronquist 2001). The information block matrix included two independent runs of the MC3 chains for ten million generations (standard deviation ≤ 0.01); the reversible-jump strategy was used (Huelsenbeck et al. 2004). An evolutionary model was used, so a proportion of invariable sites were designated, and the other proportion came from a gamma distribution (invgamma). The convergence of chains was visualized in Tracer v. 1 (Rambaut et al. 2014). The phylogram of maximum credibility for the clades was recovered in TreeAnotator v. 1.8 (Bouckaert et al. 2014) based on the burning of 2.5 million trees.

ITS and LSU sequences of 12 samples of A. guanajuatensis were obtained (Table 1). ITS and LSU sequences are respectively identical. Based on this, only four sequences were selected for phylogenetic analysis. Then after, alignment was performed with 6 sequences of Aroramyces and 22 sequences of Hysterangium (Table 2). Phylogenetic results were as follows: According to the Bayesian analysis, after 10 million generations, 25% trees were discarder as the burn-in. The standard deviation between the chains stabilized at 0.002, indicating that MC3 reached a stationary phase. To confirm that the sample size was enough, the “parameter” file was analyzed using Tracer v. 1.6 (Rambaut et al. 2014), verifying that all parameters had an estimated sample size above 1,500. The subsequent probabilities (SP) were estimated based on the strict consensus rule produced by MrBayes and indicated on the maximum credibility clade tree. The Bayesian inference analysis recovered A. guanajuatensis as a monophyletic group, with a posterior probability of 1. (Fig. 1). Ramaria gelatinosa was used to root the tree. Aroramyces balanosporus and A. guanajuatensis showed the closest relationship but were branched with a probability of 1 and a dissimilarity of 2.19, supporting the existence of a new taxon. The Hysterangium species segregated and formed two different branches.

Figure 1. 

Maximum probability phylogram of clades obtained with Bayesian inference. The posterior probabilities for each clade are shown on the branches. The accession numbers in the sequence labels indicate the GenBank accession numbers.