Two new species and a new record of yellow Cantharellus from tropical Quercus forests in eastern Mexico with the proposal of a new name for the replacement of Craterellus confluens

Abstract Two new species of yellow Cantharellus and a new record of Cantharellus tabernensis associated with tropical species of Quercus are presented, based on the taxonomic study of fresh specimens and in a phylogenetic analysis of transcription elongation factor 1-alpha (tef-1α) and the large subunit of the ribosome (nLSU) sequences. One of the new species proposed here, corresponds to a choice edible mushroom, which, in our molecular phylogeny, resulted in it being related to the group of species around C. lateritius and sister with Craterellus confluens type specimen. This latter is here formally transferred to Cantharellus and consequently a new name, Cantharellus furcatus, is proposed to replace the homonym Cantharellus confluens (Schwein.) Schwein. 1834 a later synonym of Byssomerulius corium. Detailed macroscopic and microscopic descriptions accompanied with illustrations and a taxonomic discussion are presented for each species.


Introduction
In the American continent, especially from USA, new species of Cantharellus had been proposed, several of them look-alikes of the commonly cited C. cibarius Fr., C. cinnabarinus Fr. and C. lateritius (Berk.) Singer (Arora and Dunham 2008;Buyck et al. 2010Buyck et al. , 2016aFoltz et al. 2013;Leacock et al. 2016; Thorn et al. 2017). Further explorations in tropical America are achieving also the discovery of undescribed species of the genus (e.g. Wartchow et al. 2012;Henkel et al. 2014;Nascimento et al. 2014;Buyck et al. 2016b;Herrera et al. 2018) as occurred also with Craterellus (Wilson et al. 2012).
Species delimitation in Cantharellus is often said to be hard to address, especially because of the overlap of phenotypic variation, including scarce microscopic morphoanatomic taxonomically informative features. In such a sense, Buyck et al. (2014) explicitly defined that basidiomes of Cantharellus species "…under the microscope … exhibit a discouraging monotony…". Studying Cantharellus specimens from Mexico, we have noted that the difficulty in revising early records is exacerbated by frequent incomplete data accompanying herbaria specimens. For or instance, there is poor or no information on features like hymenophore and color variations of basidiomes along their development or even by weathering effects. It is of primary importance then, to be able to count on accurate observations of specimens in fresh that lead to the characterization of their phenotypes and establish robust concepts for pertinent taxonomic conclusions. It is important even to count on data on the spatial/temporal distribution, and associated tree species.
In Cantharellus violaceovinosus (Herrera et al. 2018), for example, it was possible to document wide macromorphological and color information through a register of samples collected over more than five years, even in weekly explorations along three years sampling. Such a record allowed us to recognize its phenology between July-October in pure stands of Quercus oleoides, and found it less frequent in association with Q. glaucescens and Q. sapotifolia. In fact, such a record together with molecular information facilitated the distinction of C. violaceovinosus from other phenotypically similar species. Olariaga et al. (2015) informed about the identity of some taxa previously described solely based on colored or unpigmented variants, i.e., while in C. amethysteus (Quel.) Sacc., C. cibarius Fr., C. ferruginascens P.D. Orton, C. pallens Pilát and C. romagnesianus Eyssart. & Buyck, white specimens may occasionally occur, in C. cibarius and C. pallens orange forms can be found. Among other conclusions, these authors demonstrated that white forms of C. cibarius already described as varieties (var. inodorus Velen, f. pallidus R. Schulz) corresponded molecularly indeed to a single taxon, and C. gallaecicus (Blanco-Dios) Olariaga, lacking yellow-orange tones, is in fact the same as the orange-yellow to ochre-yellow C. romagnesianus (Olariaga et al. 2017).
During a systematic multiyear sampling of basidiomes, as part of a project focused to study ectomycorrhizal fungi in tropical Quercus forests in eastern Mexico (Montoya et al. 2019a, b), we found coexisting three species of yellow Cantharellus. Two of these taxa are distinctive by having short-sized basidiomes with veined to gill-like folded hymenophore, while a third one, is distinctive by its medium-sized, moderately robust basidiomes, with smooth or at times rugulose hymenophore, this latter apparently corresponding to what was earlier reported as "C. odoratus".
We report here the results of both, a morphological study of fresh specimens and a phylogenetic analysis of the transcription elongation factor 1-alpha (tef-1α) and the large subunit of the ribosome (nLSU) sequences obtained from our recent collections and those available in GeneBank. Three well-supported clades inferred in the phylogenetic tree, allowed us to recognize two new species and the record in Mexico of C. tabernensis, described from Southern Mississippi in USA (Feibelman et al. 1996). One of the new species here proposed, corresponds to the yellow Cantharellus with smooth hymenophore, which interestingly, in our phylogenetic analysis appears independent of Craterellus confluens (holotype), Cantharellus lateritius (holotype) and C. flavolateritius Buyck & V. Hofst. (paratype) sequences. Both macromorphological and color variation mentioned in the descriptions were recovered from fresh basidiomes through seven years of sampling. The monitoring of monodominant stands of three different species of tropical Quercus, allowed registering also, the putative ectomycorrhizal interaction of the studied species of Cantharellus.

Sampling and morphological study
Cantharellus basidiomes were collected through a weekly sampling during June-October 2015-2018, with sporadic collections among 2009-2014, in tropical oak forests from Municipalities of both Zentla (837-850 m a.s.l.) and Alto Lucero (400-500 m a.s.l.) in central Veracruz (eastern Mexico). In these oak forests, Quercus oleoides is dominant, and even forming pure stands. In the Zentla locality, Q. glaucescens and Q. sapotifolia are also present, and form monodominant small stands. Descriptions are derived from recording the morpho-anatomical features of fresh samples, the records of color follow Kornerup and Wanscher (1978) (e.g. 4A4-8) and Munsell (1994) (e.g. 2.5Y 7/8-8/8). Basidiomes were dried in a hot air dehydrator (45 °C) for their preser-vation. Microscopic features were examined from desiccated specimens, measured in 3% KOH and stained with 1% Congo red or analyzed in Melzer´s solution. Thirtyfive basidiospores per collection were measured in lateral view following Montoya et al. (2019b)

Phylogenetic analysis
We constructed a concatenated dataset, using PhyDE v.0.9971 (Müller et al. 2010), with 19 sequences obtained here (nLSU and tef-1α) (Table 1), together with sequences of related taxonomic groups, and additionally taking as reference works on chantarelles by An et al. (2017), Buyck et al. (2014Buyck et al. ( , 2016a, Herrera et al. (2018) and Olariaga et al. (2017). The dataset was aligned with MAFFT online service (Katoh et al. 2019), and the inconsistencies were corrected manually. Phylogenetic trees were generated according to Montoya et al. (2019a). The evolutionary model was calculated using the IQ-Tree 2.0-rc1 (Minh et al. 2020;Kalyaanamoorthy et al. 2017) and the best-fit model selected using the Bayesian Information Criterion (BIC), the Akaike Information Criterion (AIC) and corrected AIC. This later was used to generate a phylogenetic tree with the Maximum Likelihood (ML) method, with a Nearest Neighbour Interchange (NNI) heuristic, with the TIMe+I+G4 evolutionary model. A consensus tree was also generated 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). Another phylogenetic tree (not shown) was also generated by Bayesian Inference (
In our phylogenetic analysis, C. veraecrucis is related also with C. lateritius. This latter species exhibits pale to deep yellow or even apricot orange  or bright orange or slightly pinkish orange colors (Petersen 1979).  with their field experience also cited that C. lateritius "… has an often excentrical, sometimes laterally compressed, short to long, more or less yellow stipe that can remain white at the base but is concolorous with the cap higher up, and it has an almost smooth to clearly veined often slightly pinkish tinted hymenophore (the senior author has never seen absolutely smooth specimens)...". Based on our revision of the epitype of C. lateritius (Buyck 07.025 kept at PC, designated by Buyck and Hofstetter 2011), it microscopically differs from C. veraecrucis by the basidiospores shape (ellipsoid to slightly phaseoliform) and the terminal hyphae of the pileipellis, which are (19-) 21-60 (-70) × 5-11 µm, cylindrical to subclavate, tending to be wider than those of C. veraecrucis (Fig. 7).
Etymology. Referring to a small, yellow chanterelle; from parvus (Lat.): small and flavus (Lat.): yellow Description. Pileus 6-26 mm diam, subhemispheric in young, becoming convex to plane convex and centrally depressed, some finally irregularly infundibuliform; margin incurved when young, becoming inflexed to somewhat straight, undulate or irregular or more or less crenate, not or obscurely translucid striate; surface dry, with appressed fibrils at center when young, glabrous at remaining areas, with waxy appearance, bright yellow-orange (5A5-A8) with tiny white to light yellow scales in the center when young, paler at edge when young. Hymenophore decurrent or shortly decurrent, with gill-like folds up to 2 mm deep, subdistant to more frequently distant, at times forked, moderately thick with margin entire or often irregular or eroded, frequently intervenose, some specimens (especially towards the stipe) with irregular low and sinuous veins, often with lower irregular anastomosis among the folds, in some specimens the anastomosis occur practically in the whole hymenophore, while in others only at some areas, especially at pileus margin, with some short lamellulae-like folds, concolorous with the pileus. Stipe (10-) 15-42 × 2-6 mm, broadened towards the apex, somewhat fused, compressed at times or furrowed, solid but soon fistulous to hollow, glabrous, concolorous with pileus. Context fleshy, concolorous with pileus or somewhat paler, with waxy appearance, odor mild, agreeable; taste mild, agreeable.
Basidiospores 6.5-8.5 × 4.5-5 µm [X -= 7.32-7.34 × 4.8-4.9 µm, X -= 1. Remarks. In our phylogeny Mexican sequences of specimens Herrera 120 and 121 clustered (Fig. 1) with high values of Bootstrap and Bayesian posterior probabilities (96/0.99) with a sequence of the type specimen of Cantharellus tabernensis from U.S.A., produced by Buyck et al. (2014). The morphological description provided above includes both mentioned specimens, and in fact, in the most relevant characters, those specimens agree with the species. It should be mentioned, however, that the following features recorded in the description provided by Feibelman et al. (1996) were not observed in the Mexican material: pileus mat felted overall, often umbilicate, sometimes perforated, basidia 4-5-6 -spored and dark plasmatic pigment confined to clavate terminal cells of the surface hyphae at disc.
The record presented here of C. tabernensis, in its turn provides additional information on the species distribution. It is known from the mixed pine and hardwood forests, usually near Pinus elliotii Engelm., at the Gulf coastal plain in Texas, Mississippi, and Louisiana states in USA (Feibelman et al. 1996), and now C. tabernensis is known also in the tropical Quercus forest from Veracruz, in the coastal plain of Veracruz state in the Gulf of Mexico.

Proposal of a new name for the replacement of Craterellus confluens
Derived from the fact that Craterellus confluens was described by Berkeley (1867) from the Orizaba region in Veracruz (Mexico), later records of yellow chantherelles occurring in the Zentla region (north of Orizaba) were referred to in the past by Guzmán and Sampieri (1984) as "Cantharellus odoratus" following Corner (1966). This latter author introduced that C. lateritius and Cr. confluens were the same as Schweinitz (1822) described as Merulius odoratus. Burt (1914) mentioned particularly the macroscopic resemblance among the Cr. confluens isotype specimen in Schweinitz herbarium and the specimens of C. odoratus that he studied, thus he synonymized the former and pointed out that "…the type of Cr. confluens has the hymenium rugose-wrinkled, as is often the case in specimens of C. odoratus; its habit, dimensions, structure, coloration, and spores are quite those of C. odoratus…". In the molecular phylogeny here generated (Fig. 1) Cr. confluens holotype specimen (Botteri 6, kept at K) is supported with high values of bootstrap and Bayesian posterior probabilities sister to C. veraecrucis here described (above), and both are closely related with C. lateritius (including a sequence of the type) and C. flavolateritius. Petersen (1979) after type studies considered indeed, separately C. lateritius, Cr. odoratus (Schwein.) Fr. and Cr. confluens, being a combination of characters such as clamps (present or not), basidiome colors and the leathery, funnel-shaped basidiocarps (with a hollow stipe), among other features, considered in the distinction of such taxa. Molecular studies have also shown that Schweinitz's species belongs to Craterellus (i.e. Cr. odoratus) (Feibelman et al. 1997;Dahlman et al. 2000) and now, our analysis confirms (Fig. 1) that Cr. confluens holotype specimen belongs to Cantharellus, among the group of yellow species around C. lateritius.  suggested "…to refrain from using the name C. confluens any longer…", but rather a new specific name in Cantharellus is required for such taxon because in Cantharellus the specific name is preoccupied by C. confluens (Schwein.) Schwein. 1834, i.e. Merulius confluens Schwein. 1822, a meruliod species (Burt 1917) member of Byssomerulius (Ginns 1975;Zmitrovich et al. 2006). Possibly Cr. confluens exhibits a rare occurrence in the site that we explored in the Zentla region or it has a more restricted occurrence in some other ecosystem, near or around the city of Orizaba, Veracruz. Considering the features of the fruitbodies ("…stem divided…") mentioned by Berkeley (1867) in the diagnosis, we propose to replace the name as follows: Remarks. Presumably having been separated from the entire collection, the holotype specimen consists of a single unipileate basidiome but the diagnostic feature mentioned by Berkeley (1867) "… stem divided above into numerous pilei…", a feature practically not observed in close related species (C. flavolateritius, C. lateritius, C. veraecrucis) is present, as noted and depicted by Burt (1914), in the isotype collection at Farlow Herbarium (https://huh.harvard.edu/pages/farlow-herbarium-fh), and it is  well-depicted and described for collections from SE USA studied by Petersen (1979). The particularity of producing multipileate basidiomes and/or with fused stipes, in combination with the smooth pileus surface, pileus and hymenophore predominantly orange colored (aurantiacus in the diagnosis) hymenophore rugulose, irregularly forking and anastomosing, rarely smooth, with yellow stipe and lacking pinkish shades (Petersen 1979), are the distinctive macroscopic features of C. furcatus.