Six specimens were collected from two northern Thailand provinces, Lampang and Phrae, and one southern Thailand province, Narathiwat. The environment in northern Thailand has a tropical savanna climate supporting the semi-deciduous Dipterocarpus forests to evergreen forests. Both Lampang and Phrae experience tropical monsoon climates. Collections were conducted in dipterocarp forests dominated by Dipterocarpus obtusifolius Teijsm. ex Miq. and D. tuberculatus Roxb. in July and August. In Narathiwat province, specimens were collected from Princess Sirindhorn Wildlife Sanctuary (Pru To Daeng Wildlife Sanctuary) located near Malaysia’s southern border. This is the largest peat swamp forest in Thailand, a waterlogged tropical forest dominated by mangroves and trees belonging to Arecaceae. One of the specimens presented in this study was collected from this locality, inhabiting a submerged dead log in August.

Macromorphological characters from immature to mature basidiomata were recorded in the forest and laboratory. Images of the fresh basidiomata were captured with a Nikon DSLR D3400 in the forest. Color codes and terms follow the Methuen Handbook of Color (Kornerup and Wanscher 1978). Chemical reaction tests with 5% KOH (potassium hydroxide) were done on the context of the specimens to observe color changes. After completion of the macro-characterization, the specimens were completely dried in an air dryer at 55 °C for 48 hours and then kept in paper bags. All micromorphological characteristics (basidia, hymenial cystidia, pileipellis elements, and basidiospores) were observed with an OLYMPUS BX-53 compound microscope. Freehand sections from dried specimens were mounted in distilled water (H₂O), 5% KOH, or Melzer’s reagent to observe the gloeocystidial elements and determine the presence of pigments or color reactions. Sections from lamellae, pileus, and stipe were also mounted in 1% ammoniacal Congo Red after a short treatment with 5% KOH to observe the characters prominently. Micromorphological drawings were prepared with a drawing tube attached to the microscope at 1000× magnification. All measurements were taken with the help of CellSens Standard software dedicated to the OLYMPUS BX-53 microscope. The basidium length excludes sterigmata. Basidiospores were examined in Melzer’s reagent and 5% KOH and measured in side view. Thirty basidiospore measurements were recorded for each specimen. Basidiospore measurements and length/width ratios (Q) are recorded as minimum-mean-maximum. Thirty measurements for each of the other micromorphological characters were also recorded to have a range. Photomicrographs were taken with a camera attached to the compound microscope. After characterization, the holotype, paratype, and all the studied specimens are deposited in the Herbarium of Mae Fah Luang University (MFLU). MycoBank numbers were procured for the new species.

The genomic DNA was isolated from dry herbarium specimens (approximately 50 mg per specimen) using the High Pure PCR Template Preparation Kit (Roche) following its protocol. The NanoDrop One Microvolume UV-Vis spectrophotometer (Thermo Scientific, USA) was utilized to assess both the quality and the quantity of DNA by measuring absorbance readings. The polymerase chain reaction (PCR) amplification of the nrITS and nrLSU was performed using the primer pairs ITS1-F and ITS4, and LR0R and LR5, respectively (White et al. 1990). PCR was carried out on a PCR thermal cycler (Eppendorf AG 22331 Hamburg Mastercycler) programmed for three minutes at 94 °C, followed by 35 cycles of 30 seconds at 95 °C, one minute at 55 °C, one minute at 72 °C, and a final extension step of ten minutes at 72 °C for the amplification of the nrITS and nrLSU regions. The sequencing of both strands of the PCR products was performed by SolGent Co., Ltd., Yuseong-gu, Daejeon, South Korea. The quality of the raw sequences was checked and edited manually where needed in BioEdit v.7.0.9 (Hall 1999) and assembled using SeqMan (DNAstar, Madison, WI, USA) and Geneious 9.1.2 (Kearse et al. 2012). The final consensus sequences were deposited at GenBank (https://www.ncbi.nlm.nih.gov/genbank/) to procure the accession numbers for nrITS (L. variicystis: PQ530286 and PQ530287, L. polyhabitata: PQ530288 and PQ530289, L. cf. epia: PQ530290 and PQ533676) and nrLSU (L. variicystis: PQ601632 and PQ601633, L. polyhabitata: PQ601634 and PQ601635, L. cf. epia: PQ601636 and PQ601637).

To confirm the genus and identify the closest matches of the three species, Basic Local Alignment Search Tool (BLAST) analysis was performed for each sequence against the National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov/genbank/). Table 2 shows the details of the sequences used in the phylogenetic analysis. To begin with, the phylogenetic analysis based on nrITS and nrLSU sequence data, two datasets (nrITS and nrLSU) were prepared separately, including the newly generated sequences, close relatives of the new taxa, available sequences of other species of the genus, and the outgroup species [Mycenella minima Singer and Mycenella salicina (Velen.) Singer] acquired from the BLAST search (Altschul et al. 1997) in GenBank (Clark et al. 2016), the UNITE database (Abarenkov et al. 2024), and relevant published phylogenies (Hosen et al. 2016; Izhar et al. 2022). These two datasets were then aligned using the online version of the multiple sequence alignment program MAFFT v7 (Katoh et al. 2019) with the iterative refinement methods as L-INS-i. The alignment was edited with trimAl 1.4.1 (Capella-Gutiérrez et al. 2009) to eliminate ambiguously aligned positions. The nrITS alignments were separated into three distinct partitions: ITS1, 5.8S, and ITS2. The selection of substitution models was accomplished using jModelTest 2.0 (Guindon and Gascuel 2003; Darriba et al. 2012). The best-fit models were HKY+G for ITS1, ITS2, and LSU and K80+G for 5.8S. Species delimitation was first examined using single-locus phylogenies. When no significant conflict was observed among single-locus phylogenies, we concatenated the two single-locus datasets (nrITS and nrLSU) into one dataset using Mesquite v.3.81 (Maddison and Maddison 2008). The nrITS and the concatenated nrITS-nrLSU sequences were separately analyzed using maximum likelihood (ML) and Bayesian inference (BI). Since only a few species have nrLSU sequences, the nrITS-based phylogenetic tree has also been presented for a better understanding of species delimitation (Fig. 1).

Figure 1. 

Phylogram inferred from maximum likelihood (ML) analysis by raxmlGUI 2.0 and Bayesian inference by Mr.Bayes v.3.2.6 based on nrITS sequence data. Generated sequences in this study are presented in red bold for the novel species and blue bold for the new record. Sequences of L. cf. epia are presented in green bold. The clade representing L. variicystis, L. polyhabitata sp. nov., and L. cf. epia are demarcated by the yellow, blue, and purple boxes, respectively. Maximum likelihood bootstrap support values (MLB) ≥ 70% are shown on the left of “/,” and Bayesian posterior probabilities (BPP) ≥ 0.95 are shown on the right or below “/” at nodes.

The ML analyses were performed in RAxMLGUI 2.0 (Edler et al. 2021) with the GTRGAMMA substitution model. ML analysis was executed by applying the thorough bootstrap algorithm with 1000 replicates to obtain nodal support values. Maximum likelihood bootstrap percentages (MLB) of 70% and above are considered significant support for clades.

The BI analyses were executed by employing four Markov chain Monte Carlo (MCMC) chains for a total of 1,000,000 generations, with termination criteria set at a standard deviation of split frequencies falling below the 0.01 threshold. Trees were sampled at every 100^th generation, with the initial 25% of trees being discarded as burn-in. The convergence of chains was assessed using Tracer 1.5 (Rambaut et al. 2018), ensuring that effective sample size (ESS) values exceeded 200, which confirmed the reliability of the results. In our phylogenetic analyses, gaps within the alignment were treated as missing data. Bayesian posterior probabilities (BPP) values of 0.95 and exceeding are considered strong support. The phylogenetic trees were observed in FigTree v. 1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/) and were edited in Adobe Photoshop 2020 (Adobe Systems, USA).

The nrITS dataset (Fig. 1) for Lactocollybia includes 51 sequences representing nine species (with two outgroup species), six sequences of which are newly generated in this study. The nrITS alignment comprises 575 bases, including gaps. The final concatenated (nrITS-nrLSU) dataset (Fig. 2) consists of 24 sequences and 1460 characters, including gaps. The genus is monophyletic in both analyses (Figs 1, 2).

Figure 2. 

Phylogram inferred from maximum likelihood (ML) analysis by raxmlGUI 2.0 and Bayesian inference by Mr.Bayes v.3.2.6 based on nrITS-nrLSU sequence data. Generated sequences in this study are presented in red bold for the novel species and blue bold for the new record. Sequences of L. cf. epia are presented in green bold. Maximum likelihood bootstrap support values (MLB) ≥ 70% are shown on the left of “/,” and Bayesian posterior probabilities (BPP) ≥ 0.95 are shown on the right or below “/” at nodes.

In the nrITS phylogenetic tree (Fig. 1), the two sequences of our studied Lactocollybia (PQ530286 and PQ530287) are well nested within the “L. variicystis clade,” representing sequences of L. variicystis from São Tomé, Pakistan, India, and Iraq, showing the conspecificity of Thai collections. It also showed similarity with other L. variicystis specimens from the Philippines, India, and the USA.

The sequences PQ530288 and PQ530289 are representing a separate clade and strongly clustered with two sequences named L. angiospermarum (PP850289, PP850674) and three sequences from unidentified sequences of Lactocollybia: one from the USA (OR785928), one from Australia (KP012742), and one from Mexico (MH166807). However, this whole clade is recovered as a separate species.

The specimens PQ530290 and PQ533676 are grouped in one of L. epia clades (designated as L. cf. epia 1). The Thai specimens are grouped with the sequences designated as L. epia (KP840552) from Italy, China (PP622161, PP622162, and PP622163), and some unidentified sequences from the USA (ON175991), Laos (UDB034409), and Russia (PP277296) (Fig. 1), while two Indian sequences are nested in L. cf. epia 2.

The nrITS-nrLSU concatenated tree (Fig. 2) exhibited the same relationship and topology as the nrITS tree, with the clades remaining consistent despite fewer sequences being included.

Morphological characters are often used in species identification and delimitation. However, morphological variation within a single species can be broad and lead to incorrect taxonomic assessments. Recently, L. variicystis has been reported from various localities in tropical and subtropical regions. The species was first described from South Africa by Reid and Eicker (1998) and later has been reported in São Tomé (Desjardin and Perry 2017), Pakistan (Izhar et al. 2022), and Iraq (Al-Khesraji et al. 2022). Besides the differences with the type specimen (discussed under notes of L. variicystis), we also found variations in some morphological characters among the other specimens. Detailed comparisons have been given in Table 3. Desjardin and Perry (2017) described L. variicystis with larger ellipsoid to amygdaliform basidiospores (7.4–9.6 × 5.1–6.4 μm), hymenial gloeocystidia (88–136 × 11–18 μm), cheilocystidia (48–75 × 6.5–14.5 μm), and pileal gloeocystidia (>100 × 9.5–23 μm). The specimen has a yellowish-brown tint at the disc, a pellucid-striate pileus margin, close spacing of lamellae with 2–3 series of lamellulae, and is found in coastal forests. Pakistani and Iraqi specimens have ellipsoid to amygdaliform basidiospores, large hymenial gloeocystidia and cheilocystidia, and the absence of caulocystidia (Al-Khesraji et al. 2022; Izhar et al. 2022). Additionally, the light pink lamellae, plicate-sulcate striate margin, and lamellae spacing in Pakistani L. variicystis are quite unlike our described L. variicystis (Izhar et al. 2022). Overall, the major distinguishing features between the other L. variicystis specimens and Thai specimens include a striated pileus margin, close to subdistant lamellae spacing, lamellulae in 1–3 series, yellow or golden hymenial gloeocystidia, quite larger gloeocystidia in the hymenium and pileipellis, absence of any pigment in the pileipellis, larger cheilocystidia, prominent presence of pleurocystidia (except the type specimen), and caulocystidia (only in African specimens). These morphological variations from different geographical collections raise questions about intraspecific variability and potential environmental effects on phenotypic plasticity.

Before molecular studies, minor morphological differences often suggested species delimitation; in the type description of L. variicystis (Reid and Eicker 1998), it was noted to be very similar to L. globosa, mainly differing in its growth habit (scattered versus large troops). Such phenotypic differences debate whether the two specimens are distinct species or interbreeding populations within a species (Petersen and Hughes 1999). The geographical isolation or local adaptation can cause such phenotypic plasticity (West-Eberhard 1989), resulting in the phenotypic variation of the segregated population being broader while the genotype remains unified. The type specimen of Lactocollybia variicystis from South Africa (Reid and Eicker 1998), the São Tomé collection (Desjardin and Perry 2017), and the recent specimen from Pakistan (Izhar et al. 2022) showed similar tropical habits to the Thai specimens, whereas the Iraqi L. variicystis has a temperate host association (Al-Khesraji et al. 2022). Despite all specimens sharing the typical morphology, such as white to cream basidiomata, ellipsoid to amygdaliform spores, and the abundance of hymenial cystidia and gloeocystidia, variations, especially in the size of cystidia, can occur in individual collections. In this study, we found notable differences in the Thai specimens, such as a non-striate pileus margin, lamellae spacing with different lengths of lamellulae, a smaller size of gloeocystidia, and the absence of pleurocystidia and caulocystidia (Table 3).

In the nrITS phylogeny (Fig. 1), the lack of resolution of the branches gave it the appearance of a single cohesive group or a complex clade representing L. variicystis. However, the short branch lengths can also suggest the occurrence of rapid speciation events that resulted in little genetic divergence among lineages. Our specimens are nested in a large clade with São Tomé, Pakistani, and Iraqi specimens (90% bootstrap support) (Fig. 1). This concludes that these sequences belong to Lactocollybia and can be identified as L. variicystis. The inconsistent identification may be a result of the superficial similarity and the overlooking of some microscopic characters.

The morphological variations among different geographical samples could be a signal for the differentiation of cryptic species within a species complex. This incident may infer rapid diversification or limited genetic divergence during space and time. To ascertain these assumptions, molecular verification of the type material, multi-gene analyses, morphological study, and ecological data of undescribed sequences and additional samples are required. Considering the current data, our Thai collections are hence identified as L. variicystis.

Lactocollybia epia was originally described from Sri Lanka by Pegler (1986). Based on the morphological similarity, another species, L. angiospermarum Singer, first reported from Florida, USA (Singer 1948), was synonymized with L. epia (Pegler 1986), suggesting a pantropical distribution. The occurrence of L. epia in parts of Africa was also inferred from the herbarium specimens of L. angiospermarum by Reid (1975), Pegler (1977), and Reid and Eicker (1998). The protologue of L. angiospermarum (Singer 1948) and L. epia (Pegler 1986) indicates extensive morphological overlap between both species. Consequently, multiple authors have treated L. angiospermarum as a synonym of L. epia (Pegler 1986; Reid and Eicker 1998; Yang 2000). However, this conclusion has not been validated at the molecular level due to the absence of confirmed sequences for both species. Based on the currently available morphological data, we concur that L. angiospermarum may be considered a synonym of L. epia. Furthermore, in our phylogenetic analysis, the sequences designated as L. angiospermarum did not fall in L. cf. epia clades. The absence of any validated publication of these sequences further contributes to whether these sequences truly represent L. angiospermarum/L. epia.

The geographical distribution of mushroom species remains poorly understood due to insufficient data and sampling. The analyses of DNA sequences of mycorrhizal and saprotrophic fungi often provide evidence of regional geographic distribution rather than intercontinental distribution. The study of Bazzicalupo et al. (2019) suggested that the distribution range of the majority of species could be up to ~ 4,000 km, while only a few species could have a broad distribution, most of which have been associated with anthropogenic activity.

Intercontinental conspecificity refers to the occurrence of the same species across different continents in both the southern and northern hemispheres, highlighting the reasons behind such wide geographical distribution. Various factors contribute to assessing such distribution of any taxa, especially in species-rich tropical countries (Vasco-Palacios et al. 2022). The wide intercontinental distribution of several taxa has been reported in previous studies (Wicklow 1981; Lucking 2003; Feuerer and Hawksworth 2007; Galloway 2008; Werth 2011; Allen and Lendemer 2015; Yang et al. 2016; Vasco-Palacios et al. 2022). Further, molecular approaches, along with ecological data, have developed more discrete distribution patterns for many species (Peay et al. 2010; Summerell et al. 2010; Tedersoo et al. 2014; Song and Cui 2017). This has led to the proper identification of many taxa that previously relied solely on morphology or other taxonomic concepts (Vasco-Palacios et al. 2022).

Lactocollybia is a saprobic genus and is found on living tree parts (cortex, fallen fruits, etc.) or rotting log stumps or any plant debris (Singer 1986; Reid and Eicker 1998). The members of this genus predominantly inhabit tropical to subtropical climates, with some exceptions in temperate regions [occurrence of L. cycadicola (=L. epia) with cycads]. The currently accepted species of this genus have been documented worldwide, but the insufficient molecular data of most species hinder the inference of the actual distribution pattern of species in Lactocollybia. Based on our findings in this study, all studied species are clustered in clades with public sequences from different geographical localities, indicating that the intercontinental conspecificity may occur more than one time during space and time. In the case of L. variicystis, a species originally described from South Africa, its ubiquitous presence in Pakistan, Iraq, the Philippines, India, the USA, the United Kingdom, and more recently, Thailand, is determined by our phylogenetic analyses. Such distribution from tropical climates to temperate climatic zones is uncommon. The presence of conspecific populations across distant geographic regions suggests that L. variicystis has either evolved a remarkable adaptability to various ecological niches or has long-distance spore dispersals. However, the more obvious reasons seem to be its preferences for specific tropical to subtropical climates that have facilitated its widespread distribution.

In the nrITS phylogeny (Fig. 1), the closeness of L. polyhabitata to the sequences from the countries of both the southern and northern hemispheres refers to its wide geographical distribution. Two Thai samples of the species are conspecific with records from the USA, Australia, and Mexico, respectively. All these records were reported from tropical to subtropical climates. The three USA samples were found in Florida, where the climate varies from subtropical to tropical. The Mexican sample was found in Guadalajara, which has a humid subtropical climate relatively close to a tropical climate. The Australian specimen was reported from Howard Spring Nature Park, which is located in Darwin, Northern Territory of Australia. The region is close to Southeast Asia and has a tropical climate, featuring a wet and dry season.

Thai specimens of Lactocollybia cf. epia from Thailand showed relatedness to specimens from China, Laos, Italy, Russia, and the USA. The Chinese specimens were from Hainan, an island province of China with a tropical climate. The specimen from Italy is from Sicily, which has a typical Mediterranean subtropical climate with hot, dry summers and mild, wet winters. The other Lactocollybia sequences are from the USA, Laos, and Russia. Laos shares a comparable tropical climate. The USA specimen is from Ohio, which experiences a humid continental climate, whereas the sample from Russia is from Novosibirsk, which has a continental climate with extreme winters. The closeness with sequences from Italy, Russia, and the USA suggests the intercontinental conspecificity in this species.

These findings refer to these species’ global distribution. Another reason could be the poor knowledge of the ecology of all the Lactocollybia species and the species verification using molecular data. The incorporation of a thorough morphological characterization and phylogenetic analyses of all the specimens is essential to confirm the authenticity of all the sequences. We believe that understanding the pattern of intercontinental conspecificity in Lactocollybia may provide insight on the evolution and biogeography of saprobic mushroom-forming fungi.

The authors have declared that no competing interests exist.

No ethical statement was reported.

No funding was reported.

Ishika Bera conceptualized the research, designed the methodology, collected, analyzed, and interpreted the data, and drafted the manuscript. Komsit Wisitrassameewong’s contribution involves research design, data analysis and interpretation, and conclusion of the research. Naritsada Thongklang supervised the research, provided critical revisions, and helped refine the manuscript. All authors reviewed and approved the final version of the manuscript.

Ishika Bera https://orcid.org/0000-0003-0207-3644

Komsit Wisitrassameewong https://orcid.org/0000-0003-1195-0338

Naritsada Thongklang https://orcid.org/0000-0001-9337-5001

All of the data that support the findings of this study are available in the main text.