Two new species of Lactifluus (Fungi, Russulales) from tropical Quercus forest in eastern Mexico

Abstract Two new species of Lactifluus subgenus Lactifluus were discovered during a three-year monitoring of the ectomycorrhizal fungi in a tropical oak forest from central Veracruz, Mexico. Systematic sampling of basidiomes allowed recording of the morphological variation of fruit-bodies in different growth stages along with their fructification season. Both new species were distinguished, based on macro- and micromorphological features and on molecular data. A phylogenetic analysis of a concatenated nuc rDNA ITS, D1 and D2 domains of nuc 28S rDNA (LSU) and the 6–7 region of the second largest subunit of the RNA polymerase II (rpb2) sequence dataset of species of Lactifluus is provided. In the phylogeny inferred, one of the new species is sister to L. dissitus Van de Putte, K. Das & Verbeken and the other belongs to the group of species of L. piperatus (L.) Kuntze, sister to an unidentified species from U.S.A. The studied taxa grow under Quercus oleoides in the study site. The species are presented and illustrated here.

owned, at Alto Lucero Co. (450-500 m elevation). Sampling of the two Lactifluus species studied was developed in monodominant stands of Q. oleoides, surrounded by a coffee trees plantation or land used for livestock.
Macro-morphological features and colours were recorded from fresh samples in different growth stages. Alpha-numeric colour codes in descriptions follow Kornerup and Wanscher (1967) (e.g. 7C8) and Munsell (1994) (e.g. 10YR 8/6). Basidiomes were dried with a hot air dehydrator (45 °C) over a week. Measurements and colours of micromorphological structures were recorded in 3% potassium hydroxide (KOH) and Melzer´s solution. Methods to determine basidiospore ranges are those used by Montoya et al. (2019). Thirty five basidiospores per collection were measured (length and width of the spore in lateral view, excluding the ornamentation). These measurements are presented in taxonomic descriptions accompanied by the symbols: X representing the range of X (where X is the average of basidiospores length and width in each collection) and Q refers to the range of Q (where Q is the average of the ratio of basidiospore length/basidiospore width in each collection). The methods used to produce scanning electron microscope (FEI, Quanta 250 FEG.) images of their basidiospores are those used by Montoya and Bandala (2003). Twenty five basidia and cystidia per collection were measured. Line drawings were made with the aid of a drawing tube. Collections are part of the herbarium of the Institute of Ecology, A.C., Xalapa, Mexico (XAL) (Thiers B. [continuously updated] Index Herbariorum: a global directory of public herbaria and associate staff. New York Botanical Garden's Virtual Herbarium. http://sweetgum.nybg.org/science/ih/ accessed June 2019).

DNA extraction, PCR amplification and sequencing
Genomic DNA was extracted from fresh and dried basidiome tissue, according to Cesar et al. (2018). PCR was performed to amplify the nuc rDNA ITS (Internal Transcribed Spacer) and D1-D2 domains of nuc 28S rDNA (28S), using primers ITS1F and ITS5/ITS4 and LR0R/LR21 and LR7, respectively (Vilgalys and Hester 1990, White et al. 1990, Gardes and Bruns 1993. Regions 6 and 7 of the nuclear gene that encode the second largest subunit of RNA polymerase II (rpb2) were amplified with primers bRPB2 6f/fRPB2 7CR (Liu et al. 1999, Matheny 2005. The thermal cycler conditions for ITS and rpb2 markers were (i) initial denaturation at 95 °C for 5 min; (ii) 35 cycles of 30 sec at 95 °C, 30 sec at 55 °C and 40 sec at 72 °C (for LSU this condition was for 60 sec); and (iii) a 5 min final elongation at 72 °C. Amplified PCR products were sequenced using a Genetic Analyzer 3730XL (Applied Biosystems). Once sequences were assembled and edited, they were deposited at GenBank (Benson et al. 2017) and the accession numbers are indicated in Table 1.

Phylogenetic methods
Following preliminary analyses that placed the new species within Lactifluus subgenus Lactifluus, phylogenetic analyses were performed with the newly generated sequenc-  (Müller et al. 2010), aligned with MUSCLE algorithm (Edgar 2004) and corrected inconsistencies manually. Using the IQ-Tree (Nguyen et al. 2015) in an interface online (Trifinopoulos et al. 2016), we calculated the evolutionary model with a partitioning analysis (Kalyaanamoorthy et al. 2017;Chernomor et al. 2016) and Edge-unlinked partition model (Lopez et al. 2002), using the Bayesian Information Criterion (BIC), the Akaike Information Criterion (AIC) and corrected AIC to select the best-fit model. This later was used to generate a phylogenetic tree with the Maximum Likelihood (ML) method, with a Nearest Neighbour Interchange (NNI) heuristic, with TNe+I+G evolutionary model and Ascertainment Bias Correction (ASC). We also generated a consensus tree, calculating the Robinson-Foulds distance between the ML tree and the consensus tree, the branches being tested by means of Ultrafast Approach Bootstrap (UFBoot), SH-like approximate Likelihood Ratio Test (SH-aLRT), Approximate Bayes test (aBayes) and Bootstrap Standard (BS). A phylogenetic tree was generated also by Bayesian Inference (BI), using MrBayes v. 3.2.6 (Ronquist et al. 2012). The phylogenies from ML and BI analyses were displayed using FigTree v1.4.3 (Rambaut 2016).

Results
We generated 18 new sequences from Lactifluus species studied, six from each of ITS, nLSU regions of rDNA and rpb2 (Table 1 and alignment deposited in TreeBASE S23676). The dataset built included a total of 54 sequences and Auriscalpium vulgare, Bondarzewia montana and Stereum hirsutum as the outgroups. In the phylogenetic trees, inferred using both ML and BI, terminal clades were concordant amongst topologies and internal nodes that had significant BS score (≥ 70%), BI (≥ 0.90), UFBoot (≥ 95%), aBayes (≥ 0.90) and SH-aLRT (≥ 80%). The ML tree with the two former values for the nodes is presented here (Fig. 1). The generated sequences from the Mexican specimens clustered with strong support in two terminal clades. Based on morphological features and supported with the grouping displayed in the phylogenetic tree, we recognised two groups of the Mexican samples studied representing two distinct new species of Lactifluus. One of them, Lactifluus mexicanus, appears sister (with strong support) to L. dissitus from India and the other one, L. lorenae, clusters in a clade with L. piperatus (L.) Kuntze from Europe and related species from North America and Asia, sister (with strong support) to an undescribed species from U.S.A.
Etymology. In honour of Dr. Lorena E. Sánchez Higueredo because of her interest in the conservation of tropical oak forest relicts in Veracruz, Mexico.
Habitat Diagnosis. Recognised by the combination of pileus disc faintly rugose, margin rugose to strongly venous-rugose, lamellae close to very close, the stipe including pinkish tinges and by the size of lamprocystidia and pileipellis terminal elements.

Discussion
The results inferred in the multilocus phylogeny (Fig. 1), strongly support the recognition of the two new species, Lactifluus lorenae and L. mexicanus. Although we faced difficulties to amplify rpb2 region, fortunately, the Mexican collections processed allowed us to recover with success, this and also ITS and 28S regions. The resolution obtained in our phylogeny may be related to the vouchers selection, mostly having sequences of the three regions (ITS, 28S and rpb2). The strong support of the clades, especially of L. mexicanus and L. lorenae allow us to complement morphological results and, on this basis, we decided to describe them. Both species are members of subgenus Lactifluus, the first one falling in section Piperati and the second in section Lactifluus, according to the classification proposed by De Crop et al. (2017).
Lactifluus lorenae is a white milkcap, with basidiomes showing macromorphological similarities with L. piperatus, as narrowly cicumscribed by De . When comparing the macro-and micromorphological variation displayed in the Mexican samples and the information provided by De  about L. piperatus in the strict sense, significant differences between the two taxa are detected. Basidiomes of the Mexican species show a uniform tendency to develop an orange-brown colouration on the surfaces when handled and in the context when exposed. The latex can be somewhat serous, staining white paper yellow and becoming brownish after some minutes. When comparing micromorphological features between L. lorenae and L. piperatus (according to the later authors), in the former, the basidiospores are more globose (Q = 1.20-1.27 vs. Q = 1.28-1.40) and pleurocystidia are distinctly shorter (40-53 × 7-9 µm vs. 50-70(-90) × 8-11 µm). Another difference between the taxa is the pileipellis structure, which in the Mexican species presents a thicker hyphoid suprapellis (30-60 µm thick vs. 10-30 µm thick) and with abundant dermatocystidia in the suprapellis in L. piperatus, while scarce in the subpellis in the Mexican taxon. Organoleptic differences may be noted between both taxa too, because in L. lorenae, the odour is somewhat like chlorine, while in L. piperatus, it is slightly acidic, distinctly honey-or apple-like when drying. In the inferred phylogeny (Fig. 1), L. lorenae clusters sister to an unidentified species, L. aff. piperatus USA 3-North America 3, but unfortunately, there is no information available on its morphological features and habitat from the U.S.A. to compare with the Mexican species.
From the weekly sampling in tropical Quercus forest, during 2015-2017, we conclude that basidiomes of the studied species are produced in June-October, with those of Lactifluus mexicanus being more abundant. Although close to other edible species (Boa 2004, Borah et al. 2018, we have no records of edibility for L. mexicanus in the area. Considering the high diversity of Quercus and Pinus species in Mexico, they represent important ECM hosts, related with the milkcaps in the country. Quercus oleoides, with a wide distribution from Mexico to Costa Rica, especially represents a key ECM host for this group of fungi in its range. In Costa Rica, however, at an elevation around 215 m, associated with Q. oleoides, Desai et al. (2016) found 37 ECM species belonging to different genera, three of which were determined as Lactarius but no Lactifluus was recorded. Considering that the two Lactifluus species, here studied, were found in a monodominant area of Q. oleoides, we consider them as putative mycobionts of this tree species. However, this will need to be confirmed at root tip level with molecular evidence, as in other milkcaps, such as Lactarius trichodermoides Montoya, Bandala & M. Herrera and L. subplinthogalus Coker (Herrera et al. 2018b). The two latter species associate with Q. sapotifolia and Q. glaucescens, respectively, in the relicts of the tropical oak forests from central Veracruz, Mexico.