Taxonomy of the major rhizomorphic species of the “Melanopus group” within Polyporaceae in Yasuní National Park, Ecuador

Yasuní National Park in Ecuador is one of the most biodiverse places on earth. The fungi in this tropical rainforest are also diverse but have received little research attention. This research paper focuses on an important group of fungi in the family Polyporaceae and examines the genera Polyporus, Atroporus, and Neodictyopus that form aerial melanized cord-like structures called rhizomorphs. Phylogenetic analyses, macro and micromorphological descriptions of basidiomata and rhizomorphs, as well as cultural characterization were completed to better understand these ecologically important fungi. Here we describe four new species: Atroporus yasuniensis, Atroporus tagaeri, Neodictyopus sylvaticus, and Polyporus taromenane, and a new variety Polyporus leprieurii var. yasuniensis. The information presented in this study adds important new knowledge about the unusual rhizomorph producing fungi found in Yasuní National Park, Ecuador and other tropical rainforests.


Introduction
Rhizomorphs are fully independent organs formed of parallel hyphal aggregations with different levels of hyphal specialization [1][2][3]. They are often covered by a black cuticle of melanin for protection against microbial antagonism, UV light and other environmental hazards [2,4,5]. Rhizomorphs can transport air, water, and solutes by using a complex network system that responds to gravity, light, carbon dioxide and changes in nutrient availability [2,3,5]. The development of rhizomorphs can confer advantages for survival in harsh and highly colonized environments, allowing the rhizomorphic fungus to explore and colonize new substrates, extract and store resources, and promote infection of woody plants [3]. Additionally, they can be a survival mechanism allowing the fungus to survive much longer than spores and basidiomata [6,7].
These structures have been reported to be produced by fungi across 54 families in the Basidiomycota [2]. However, it is more likely this number is underestimated as these structures are

Morphology
Macro-morphological descriptions and ecological observations were made in the field. Substrates were identified as fine woody debris (FWD) when wood diameters were 0.5-9 cm and coarse woody debris (CWD) when wood diameters were � 10 cm. When growing on a living tree, the trees were identified from the Forest Dynamic Laboratory database, Pontificia Universidad Católica del Ecuador. Color terms used followed the "Mycological Colour Chart" and "Munsell soil color charts" [20,21]. Micro-morphological characters were observed under a compound light microscope (Nikon E6001) using standard methods for the describing polypores [22]. Freehand cross sections of dried samples mounted with Melzer's reagent, and/or, 3% NaOH, 5 and 10% KOH, sterile water, 1% Congo red, 1% phloxine, and 75% glycerol were prepared in order to observe spores. For hyphal systems and other individual structures, cross sections in 3% NaOH were incubated in a 60˚C water bath for four hours, then vigorously shaken for one min at 20000 speed in the MagNA Lyser Roche1, dissected under the stereomicroscope (Nikon SM28001) and mounted on water and/or the reagents listed above. Basidiospores and other structures were measured using a light microscope at 100x magnification with deionized water and/or reagents (n = 40 to 100 depending on the number of structures available within each sample). For basidiospores, shape terminology was that of the Stalpers database (https://russulales.mycobank.org/). Abbreviations are as follow: IKI+ = dextrinoid, IKI-= inamyloid and indextrinoid, CB+/-= cyanophilous/acyanophilous, ave = arithmetic mean, Q = the ratio of length/width of basidiospores.

Fungal isolations and cultural characterization
Using aseptic techniques, small pieces of wood substrates (~5 mm), from areas where rhizomorphs and basidiomata were attached to wood, were placed in 0.5% NaOCl solution for one minute, rinsed with running sterile water for ten seconds, and excess solution was removed with clean paper towels. The small pieces of wood were then placed in petri dishes containing malt agar with additives to prevent the growth of bacteria and contaminating Ascomycota. A semi-selective media (BSA) was used containing 15 g malt extract, 15 grams of agar, 2 g of yeast extract, 0.05 g benomyl, 0.5 g streptomycin and 1 L deionized water [23]. Isolates were monitored daily for growth and when mycelium extended 10mm, they were transferred to new plates. Fungal isolates were used for DNA extractions and to determine cultural characteristics.
For cultural characterization, two culturing media were used including malt extract agar (MEA) 15 g malt extract, 15 g of agar, 1 L deionized water and malt yeast agar (MYA) 15 g malt extract, 15 grams of agar, 2 g of yeast extract and 1 L deionized water [23]. Isolates were incubated at 24-26 C˚. Six plates were used for each type of media. After five days, growth (mm) was determined using a digital caliper. Hyphal density, margins of the colony, color, and hyphal melanization were noted. Morphological descriptions and terms followed Stalpers (1978) [24].

DNA extraction, PCR amplification and sequencing
DNA extractions were made from pure cultures grown on MYA and/or from rhizomorphs, stipe, and/or pileus samples preserved in 95% ethanol following the CTAB extraction protocol Of Blanchette et al. 2016 [23]. When DNA extractions were from pure cultures, ¼ of fungal tissue on the petri (10 mm diameter) dish was scraped. When isolations were from other tissues, a 5 mm piece of tissue was cut into smaller pieces. Fungal tissue was suspended in 500 μl of cetyltrimethylammonium bromide (CTAB) lysis buffer with glass beads and vortexed for 1 minute for hyphal tissue and 5 minutes for pileus/stipe/rhizomorphs, samples were centrifuged for 30 seconds to aggregate fungal material. Supernatant was transferred to a new microcentrifuge tube and placed in a hot water bath at 65˚C for 20 minutes. Subsequent, 500 μl of chloroform/phenol/isoamyl alcohol (25:24:1) was added to the tubes, shaken vigorously and centrifuged for 5 min at 15,000 rcf. The supernatant was then transferred to a clean microcentrifuge tube and isopropanol (stored at -20˚C) was added (2/3 the amount of supernatant), tubes were inverted a few times and incubated at room temperature for 5 min. Tubes were then centrifuged for 7 min at 20,000 rcf, and supernatant was carefully removed. To the remaining DNA pellet, 500 μl of 70% ethanol (stored at -20˚C) were added and centrifuged for 3 min at 20,000 rcf to wash the pellet. Ethanol was carefully removed, and tubes were left in a sterilized bio-safety cabinet to air dry. DNA was rehydrated with 100 μl of sterile water.
The partial regions of the nuclear ribosomal small subunit (SSU), nuclear ribosomal internal transcribed spacer (ITS), nuclear ribosomal large subunit (LSU), elongation factor 1-alpha (EF1-α) and the RNA polymerase II largest subunit (RPB1) were amplified by Polymerase Chain Reaction (PCR). The primers used for each region were: PNS1 with NS41 for SSU, ITS1F with ITS4 or ITS4B for ITS, LROR with LR7 or LR5 for LSU, EF1-983F with EF1-1567R for EF1-α and RPB1-Af with rpb1-Cr for RPB1 [25]. The PCR final volume was 25.5 μl, each tube containing 1 μl of DNA template, 1 μl of each primer (10 μM for RPB1 and 5 μM for the other regions), 0.5 μl of bovine serum albumin (BSA), 12.5 μl of GoTaq1 Green Master Mix and 9.5 μl of sterile water in a thermocycler. PCR program for ITS and SSU were: 94˚C for 5 min, 35 cycles of 94˚C for 1 min, and 72˚C for 1 minute, followed by a final extension step of 72˚C for 5 min. For LSU: 95˚C for 3 min, 94˚C for 1min, 50˚C for 45 sec, 74˚C for 2 min ramping up 0.3˚C/sec, 94˚C for 1 min for 29 cycles and a final step of 72˚C for 5 min. For EF1-α and PBB1 the same program was followed as: 94˚C for 2 min, 94˚C for 40 sec, 60˚C for 40 sec decreasing 1˚C per cycle, 72˚C for 2 min,94˚C for 40 sec for 8 cycles, 94˚C for 45 sec, 53˚C for 1:30 min, 72˚C for 2 min, 94˚C for 45 sec for 36 cycles and a final step of 72˚C for 10 min.
Sequencing was carried out with Big Dye V3.1 and run on ABI 3730xl instruments (Functional Biosciences, Madison, WI, USA). Consensus sequences were assembled using Geneious 7.0 [26]. For samples with multiple DNA origins (pure culture, basidiomata and/or rhizomorphs), individual consensus sequences were created and then aligned to verify the DNA came from the same sample. When alignments had more than 99% pairwise identity, a single consensus sequence was created with all multiple DNA sequences to represent a specific sample. All final consensus sequences were compared through BLAST searches to GenBank sequences, whose identification has been supported with taxonomic publications to prevent sequences matching those wrongly identified. All sequences generated in this study were submitted to GenBank (Table 1).

Phylogenetic analyses
All 135 newly generated sequences were deposited in GenBank, 28 SSU, 30 ITS, 28 LSU, 27 EF1, and 22 RPB1 (Table 1). However, not all sequences were used in the phylogenetic analysis but are available in GenBank for future taxonomic studies. SSU sequences were not included in the analyses as a multiple alignment of the region had 93% identical sites among all taxa, resulting in individual phylogenetic trees without clear delimitation of taxa. Ninety-six sequences from other studies available in GenBank were used in the analyses ( Table 1). Two datasets were constructed, the first based on two molecular markers (ITS and LSU), and the second on four molecular markers (ITS, LSU, RPB1, and EF1-α). Datasets included sequences generated in this study and reference sequences of Polyporus and allied genera downloaded from GenBank. Individual datasets for each region were aligned using Clustal Omega [27] sequence aligning tool and then manually examined and adjusted in Geneious 7.0 [25]. Due to the high variation of sequence length in the LSU alignment, sequences were trimmed to fit the length of most sequences, for shorter sequences, ends were filled with Ns. Alignments for the other regions were trimmed for an even length for all sequences. Annotations were made to identify polymorphisms (Indels and SNPs) on individual alignments of each molecular marker to identify variants frequency. The best-fit evolutionary model for each partition was calculated on JModelTest [28,29] following the Bayesian Information Criterion (BCI). The Partition Homogeneity Test (PHT) was run on PAUP to test congruence [30]. Maximum-likelihood trees were generated on MEGA 10.05 tool [31] using the General Time Reversible model to choose the tree with the highest log likelihood with 1000 replicates and Gamma Distributed with invariant sites (G+I) for rates among sites. Bayesian Inference was

Nomenclature acts
The electronic version of this article in Portable Document Format (PDF) in a work with an ISSS or ISBN will represent a published work according to the International Code of Nomenclature for algae, fungi, and plants, and hence the new name contained in the electronic publication of a PLOS article are effectively published under that code from electronic edition alone, there is no longer any need to provide printed copies. In addition, new names contained in this work have been submitted to MycoBank from which they will be made available to the Global Names Index. The unique MycoBank number can be resolved, and the associated information viewed through any standard web browser by appending the MycoBank number contained in this publication to the prefix http://www. mycobank.org/MB/. The online version of this work is archived and available from the following digital repositories: PubMed Central and LOCKSS.

Results
Observations of collections from Yasuní National Park indicated that several Polyporus-like species were associated with the aerial formation of rhizomorphs. These fungal structures are common on the forest floor and surveys found 109 collection sites with Polyporus-like basidiomata samples. Rhizomorphs were associated with these Polyporus species in 100 of the sample sites. Differences in morphological characteristics were found among samples, suggesting that several previously undescribed and/or misidentified species were present in this ecosystem. Phylogenetic studies were needed to elucidate the genetic relationships among the sampled taxa and their taxonomy. We report here the phylogenetic analyses and morphological descriptions for the new taxa.

Phylogenetic analysis
Eighty-eight of the 135 newly generated sequences were used in this study (25 ITS, 21 LSU, 22 EF1-α, and 20 RPB1). Another 96 related sequences (including 34 ITS, 32 LSU, 18 EF1-α, and 12 RPB1) used in the phylogenetic analyses were downloaded from GenBank and are listed in Table 1. Datasets were concatenated as follow: ITS-LSU resulted in an alignment with 1468 total characters (including 618 ITS + 850 LSU nucleotides) and ITS-LSU-RPB1-EF1α with 3267 total characters (including 618 ITS + 850 LSU + 1232 RPB1 + 567 EF1-α nucleotides). The partition homogeneity test indicated that sequences in both datasets were not congruent (P-value = 0.01). In the first dataset, 756 characters were constant (proportion = 0.5), 120 variable characters were parsimony-uninformative, and 592 characters were parsimony-informative. For the second dataset, 1635 characters were constant (proportion = 0.5), 344 variable characters were parsimony-uninformative, and 1288 characters were parsimony-informative. In the Bayesian inference trees, for both datasets, total chain length was set up as 1,100,000, subsample frequencies at 200, and total samples analyzed were 4951. While in the Maximumlikelihood analysis, GTR+I+G model was used, and partitions were established considering individual markets. A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories (+G, parameter = 0.5373)). The bootstrap consensus tree inferred from 1000 replicates where branches corresponding to partitions reproduced in less than 50% bootstrap replicates were collapsed. Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value [32]. One hundred twenty-one variants were found in the first dataset and four hundred seven variants in the second dataset. SNPs were the most common variants among datasets, being 74% and 85% respectively of the total variants found in each dataset ( Table 2).
The "squamosus" clade named after Polyporus squamosus (Huds.) Fr. was formed by sequences of P. squamosus, P. varius, P. hemicapnodes, P. leprieurii, and P. leprieurii var. yasuniensis variety novo. The last two formed an internal clade named "leprieurii" after Polyporus leprieurii Mont. This clade included samples from Ecuador (Yasuní), Costa Rica, and Belize. We have accepted as the "true" P. leprieurii the sequences from Costa Rica and Belize for the closer proximity to the type location, the French Guiana, as no sequence is available for the type collection. In this phylogeny, fifteen sequences from Yasuní were closely related to the "true" P. leprieurii. However, only six samples were designated to the newly proposed P. leprieurii var. yasuniensis, while the remaining sample, forming two well supported related clades within the "leprieurii" clade, need further taxonomical designation. P. hemicapnodes, an accepted synonym of P. leprieurii from China is closely related to the "leprieurii" clade, but forms its own well-defined clade (Fig 1).
The "guianensis" clade, named after P. guianensis Mont., was formed by sequences of P. guianensis from Argentina and Venezuela; three samples of P. taromenane species novo from Ecuador; and two samples from Ecuador plus one from Costa Rica previously misidentified as P. guianensis.
The taxa designated as Picipes clustered in two clades: "picipes" named after Picipes I.  The "neodictyopus" clade, named after the Neodictyopus Palacio, Robledo, Reck & Drechsler-Santos genus was formed by taxa from Brazil, Ecuador and Belize. N. sylvaticus species novo was closely related to one of the sequences of N. dictyopus from Brazil. The sequence of   Ryvarden (1973) based on swollen binding hyphae in the context of the type species, but this characteristic was considered too vaguely defined for a genus description [33]. Atroporus was later amended by Palacio, Robledo, Reck & Drechsler-Santos (2017), using basidiospore descriptions and Melzer reagent reaction of the skeletal-binding hyphae in the trama [17].

PLOS ONE
The genus shares macromorphological characteristics with the "Melanopus group", which includes annual to biannual basidiomata, coriaceous when fresh and tough when dry; centrally to eccentrically stipitate, with a black cuticle on the stipe. Spores medium to large, 6-12 x 2-4 μm [4].
Remarks: Basidiospores narrowly ellipsoid to subcylindrical, thin-walled, smooth, hyaline, IKI-. Hyphal system dimitic with generative and skeletal-binding hyphae. Generative hyphae present in trama with clamp connections; skeletal-binding hyphae in trama dextrinoid with a distinctive wide stalk with sharply pointed apex that aids in the identification of the genus. All species grow on dead wood, typically on dead fallen branches of relatively thin diameter trees [17]. Collection number: CTR-2-46, Toapanta-Alban, July 2019 Etymology: yasuniensis (Yasuní = the paradise/God's creation (from the Wao Terero language use by the Huaorani/Waorani Indigenous People living in the Amazon Rainforest between the Napo and Curaray Rivers in Ecuador)) + ensis (Latin suffix to denote origin). In reference to the site where samples were collected, Yasuní National Park in Ecuador.
Basidiomata annual, gregarious, centrally stipitate, up to and 6 cm tall. Pilei circular, depressed, infundibuliform with irregular wavy-striate margins that become more pronounced over time; up to 8 cm in diameter and 1.3 mm thick (Fig 2).
Rhizomorphs, circular to slightly flat, solid, robust, coriaceous, thick (up to 5mm diameter), long (up to 3 m), well-branched, bearing a black-reddish lacquered cuticle with white-yellowish tips when fresh (Fig 2). Rhizomorphs are perennial (found growing for over four years in the same area), well distributed along the forest's floor, emerging from decayed wood (coarse and fine woody debris). The formation of a rhizomorphic mat (brown-reddish tissue that becomes darker over time) facilitates the attachment to new substrates and other plant material on the forest floor. Rhizomorphs undergo anastomosis to form a complex rhizomorphic network. Rhizomorphs use trees and other surfaces to maintain upward growth. They occur on substrates accompanying basidiomata or alone.
Cultural characteristics: Cultures were easily obtained from rhizomorphs and colonized substrates. Initial growth is characterized by white, appressed, silky mycelium that can turn dark brown at different areas of the colony. Growth rate, characteristics of the mycelial mat, presence of aerial mycelium, and color of the colony are variable among colonies growing in the same media.
On MEA, the marginal zone is irregular, usually submerged. When mycelium is submerged, pseudosclerotial plates are formed giving the agar a brown coloration and chamoislike appearance. Aerial mycelium is white, floccose, and crustose becoming brown, thick, coriaceous over time. On MYA, margins are regular to irregular, generally appressed, and rarely raised. Mycelium can be submerged and aerial. Aerial mycelium silky, floccose, cottony, white with dark colorations that are not as common as on MEA. Rhizomorphs rarely form on culture media but may be formed after long periods (6-12 months) of growth when nutrients apparently have been utilized (on MYA only).

PLOS ONE
Taxonomy of the rhizomorphic species of Polyporaceae in Yasuní National Park, Ecuador with simple septa; weakly IKI+ reaction as hyphae turned brown in the reagent; KOH-. Generative hyphae on crustose sections of mycelium are yellowish-brown to brown-reddish, thickwalled (2-8 μm wide) with abundant round-like swellings and clamp connections.
Substrate: Basidiomata and rhizomorphs emerge from decayed fine and coarse woody debris of angiosperms.
Distribution: known from the Amazon rainforest of Ecuador along valleys (temporarily flooded forests).
Remarks: Atroporus yasuninensis was described from basidiomata and rhizomorphs collected over three years in an area of 20 x 10 m 2 in a temporarily flooded forest. It is characterized by gregarious, medium to big basidiomata (up to 10 cm), centrally stipitate. Pilei infundibuliform with irregular wavy-striate margins, up to 1.5 mm thick. Compared to A. yasuniensis, A. rufoatratus and A. diabolicus have relatively smaller basidiomata with thicker pilei and have been reported to have solitarily habit. Microscopically, A. yasuniensis has similarities with A. rufoatratus, with thick-walled distinctive stalks with pointing apex and 1-5 branches present in the trama. Cystidioles shape differs between A. rufoatratus and A. yasuniensis, with subulate cystidioles in A. rufoatratus and ventricose and oblong-like cystidioles in A. yasuniensis. Basidiospores are similar to other taxa in Atroporus; ellipsoid, narrowly ellipsoid, to slightly cylindrical, hyaline, smooth. Rhizomorphs cylindrical to slightly flat, solid, robust, flexible, coriaceous, well-branched, long, bearing a black-reddish cuticle. Rhizomorphs Etymology: tagaeri from the Huaorani language (Wao-Terero) in reference to the voluntarily isolated tribe, named after one of their members Tagae. Tribe living in and near Yasuní National Park, Ecuador.
Basidiomata annual, gregarious, centrally stipitate, up to and 4 cm tall. Pileus circular, convex to slightly umbonate with regular margins when young; circular, slightly depressed when old; up to 3 cm in diameter and 0.7 to 1 mm thick (Fig 5).

PLOS ONE
Taxonomy of the rhizomorphic species of Polyporaceae in Yasuní National Park, Ecuador Hyphal system dimitic with generative and skeletal-binding hyphae. In the trama, generative hyphae hyaline, highly branched with thin (2-3 μm) and thick-walled (2-5.5 μm) sections; simple septa and clamp connections. Thin-walled sections along branches and thick-walled sections on the main stem (Fig 6).
No rhizomorphs were associated with basidiomata. Cultural characteristics: No cultures were obtained. Substrate: Basidiomata emerge from decayed fine woody debris of angiosperms. Distribution: Known from the Amazon rainforest of Ecuador along ridges and slopes (tierra firme forest).
Remarks: Atroporus tagaeri was described from samples collected in an area of 1m 2 in terra firme forest. It is characterized by gregarious, small to medium basidiomata (up to 4 cm), centrally stipitate. Pilei circular, slightly depressed with regular margins, up to 3 cm diameter and 1 mm thick. In terms of shape and size, A. tagaeri is similar to A. rufoatratus, however, A. tagaeri has thinner pilei. Hyphal system dimitic, with generative and skeletal-binding hyphae, similar to other taxa in Atroporus. The distinctive thick-walled stalks with pointing apex, aid in the description of the genus were not found in the examined specimens. Cystidioles are less frequent in A. tagaeri than in A. yasuniensis 2017), recommended carrying out surveys and collections from the type locality in order to define the circumscription and distribution.
The genus is also part of the "Melanopus group" since it shares the macromorphological similarities previously described.
Remarks: Basidiomata lateral to eccentrically stipitate; pileus reniform to spatulate. Hyphal system dimitic, with clamped generative hyphae and skeletal-binding hyphae of the arboriform type, slightly dextrinoid (in mass only) in the trama. Basidiospores subcylindrical to bacilliform. Distributed along the Neotropics, however, might be pantropical. All species grow on dead wood producing white rot on the substrate. Neodictyopus is morphologically similar to taxa within the Polyporus and Atroporus genera. However, Neodictyopus differs from the type species of Polyporus (P. tuberaster) in the texture of the pileus and from Atroporus by the spore shape and the strong dextrinoid reaction in the skeletal-binding hyphae in the trama proper of Atroporus. Etymology: sylvaticus from Latin sylva ("forest"), in reference to the geographical region where specimens were found, Amazon rainforest of Ecuador.

Neodictyopus sylvaticus
Basidiomata annual, gregarious, centrally to eccentrically stipitate, up to and 2.5 cm tall. Pileus reniform to flabelliform with regular to irregular wavy margins; up to 2.3 cm in diameter and 0.83-1.6 mm thick (Fig 7).

PLOS ONE
Taxonomy of the rhizomorphic species of Polyporaceae in Yasuní National Park, Ecuador Hyphal system dimitic with generative and skeletal-binding hyphae. In the trama, generative hyphae, hyaline, highly branched with thick and thin-walled sections; thicker sections, slightly rounded, 3-5.5 μm along the main stem and thinner sections 2-3 μm on branches; simple septa and clamp connections (Fig 8).
No rhizomorphs were found associated with basidiomata. Cultural characteristics: No cultures were obtained. Substrate: Basidiomata emerge from decayed fine woody debris of angiosperms. Distribution: Known from the Amazon rainforest of Ecuador along ridges and slopes (tierra firme forest).
Remarks: Neodictyopus sylvaticus was described from samples collected in an area of 1m 2 in terra firme forest. It is characterized by gregarious, small basidiomata (up to 3 cm), centrally to eccentrically stipitate. Pilei reniform to flabelliform with regular to irregular wavy margins, up to 2.5 cm diameter and 1.5 mm thick. In terms of shape, N. sylvaticus is similar to N. dictyopus, however, N. sylvaticus has smaller basidiomata with thinner pilei. Stipe is short, flat, solid, robust, which differ from the other taxa in the genus, described to have circular stipes. Hyphal system dimitic, with generative and skeletal-binding hyphae. Generative hyphae are distinctive from other taxa in the genus. These are highly branched, thin and thick-walled, slightly rounded along the main stem. In trama, hyphal modification as plectenchymatous thickenings are common. Such hyphal modifications also found in the context of A. tagaeri, but not reported in any other taxa in Neodictyopus. Tramal setae embedded in tubes are very rare, reported to N. sylvaticus but not in other taxa in Neodictyopus or allied genera. Basidiospores narrowly ellipsoid to cylindrical. Rhizomorphs have been reported for P. dictyopus, however, no rhizomorphs were found associated with N. sylvaticus.  (Fig 9).

PLOS ONE
(ave = 3.2 μm) wide with a basal clamp connection. Skeletal-binding hyphae of the arboriform type found in trama, context and stipe, hyaline, thin and thick walled; thin-walled sections 1-3 μm along branches and thick-walled sections along the main stem 2-7 μm. In context, hyphal modifications are frequent with segments of plectenchymatous thickenings 40-145.5 μm (ave = 68 μm) long and 3.6-8 μm (ave = 6.4 μm) wide; clamp connections observe where ramifications start. Hymenium composed of paraphysiods structures (cystidioles, basidioles, and hyphidia). Hyphidia, smooth, tubular, thin-walled. Cystidioles of the ventricose type 8-18 μm x 3-6.5 μm. Basidioles, clavate, 7.5-12.7 μm x 3.6-5.8 μm. Basidia not found. Basidiospores not found, however, measurements found on collection label are: 5-7 x 2.5-3.5 μm, Q = 1.6-2.3 (ave = 1.9); ellipsoid to narrowly ellipsoid, slightly cylindrical. Remarks: The herbarium sample examined was collected in Cuyabeno Wildlife Reserve, Ecuador in 1993. It was previously determined as P. leprieurii, however, based on macromorphological characteristics and phylogenetic analysis we determined this sample belongs to Neodictyopus. This specimen is characterized by basidiome, eccentrically stipitate, solitary; pile reniform, dark vinaceous. Stipe, circular, short, bearing a black cuticle. Macromorphological characteristics in terms of shape and color are similar to N. dictyopus, however it has a smaller basidiome. Cystidioles of the ventricose type, which are different in shape to the ones present in N. atlanticae and N. dictyopus with subulate cystidioles. Thick-walled distinctive stalks with rounded pointing-edges and 1-3 branches are present in trama, similar structures present in most species of Atroporus. In context, plectenchymatous thickenings are abundant. The presence of rhizomorphs was not annotated on the label, however, P. dictyopus, which is closely related to this taxon, has been reported to develop rhizomorphs.
Polyporus P. Micheli ex Adans., Familles des plantes 2: 10 (1763) Type species: Polyporus tuberaster (Jacq. Ex Pers.) F., Systema Mycologicum 1:347 (1821) Type specimen: Plate 71, Fig 1 in Micheli (1729), was selected as lectotype in Ryvarden (1991) Polyporus accommodates many morphologically heterogeneous species. Taxa within the genus has been accommodated in six infrageneric groups. The genus has undergone important taxonomical changes since it first was proposed. Initial changes were based on macro and micromorphological observations [34]. The later changes were made with the development of molecular biological techniques that allowed to identify the phylogenetic relationships within the taxa under Polyporus [15][16][17][18]. More taxonomic changes are likely to occur as more sophisticated phylogenetic studies are available.
Description: The type concept followed in this study refers to species with stipitate basidiomata growing on wood or arising from a sclerotium. Basidiomata can be annual or biannual, centrally, eccentrically, or laterally stipitate. Pilei can be circular to dimidiate, convex to infundibuliform; pilear surface smooth to scaly, glabrous to finely tomentose, white to deep brown or black; pilear texture tough when fresh, leathery, or brittle when dry. Pore surface white to creamy, or dark brown when dry. Pores entire, circular to angular, small to large, decurrent or not on the stipe. Stipe creamy to black, glabrous to finely tomentose, with or without a cuticle, smooth to longitudinally wrinkled, in some. Hyphal system dimitic, with arboriform vegetative hyphae; generative hyphae hyaline with clamp connections and simple septa and skeletal-binding hyphae hyaline to brown, solid, or with a lumen. Cystidia none. Hyphal pegs present or absent. Basidia clavate, 4-sterigmate. Basidiospores cylindrical to subellipsoid, straight to slightly bent, thin-walled, smooth, hyaline, IKI-. Substrate, woody debris of angiosperms, conifers, or developing from sclerotia buried in the ground. Species are mostly saprophytes and rarely parasites, causing white rot on wood [4].
Hyphal system dimitic with generative and skeletal-binding hyphae. In the trama, generative hyphae, hyaline, highly branched, thin-walled 1-2.5 μm, simple septa and clamp connections. Skeletal-binding hyphae of the arboriform type, abundant in trama and context, thin and thick walled; thin-walled 1-2.5 μm along branches and thick-walled along the main stem 2.5-4.5 μm. Hymenium composed of paraphysiods structures (basidioles, and hyphidia) and basidia. Hyphidia tubular, smooth, thin-walled, 3.4-6 μm. Basidioles abundant, clavate, 7.5-12.7 x 3.6-5.8 μm. Basidia clavate, four-esterigmate, 20-30 (ave = 21) μm x 4.5-6 (ave = 5) μm. Basidiospores ellipsoid to narrowly ellipsoid, thin-walled, hyaline, smooth, IKI-, 4.5-6.6 (ave = 5.5) μm Remarks: The species was described by Camille Montagne in 1840 while describing other taxa from French Guiana. P. leprieurii share macromorphological characteristics with members of the "Melanopus group". The specimen examined was from the Center for Forest Mycology Research, Wisconsin, was determined as P. leprieurii by Leif Ryvarden in 2001. It is a partial specimen cut into three pieces; size cannot be determined. Pilear surface smooth, glabrous, ochraceous; rigid texture as dried. Pore surface light brown to pinkish when dry. Pores circular to angular. Stipe solid, glabrous, bearing a black cuticle. Basidiospores ellipsoid to narrowly ellipsoid. Hyphal system dimitic with generative and skeletal-binding hyphae. The production of rhizomorphs is well documented in this taxon; however, their presence was not established on the collection label. P. leprieurii is known to be distributed in tropical and subtropical areas in America and East Asia. According to "The Global Fungal List", most records are in Brazil while others are in the Americas (North, Central, and South America), the Caribbean region, Africa (Congo, Cameroon), Northern Asia (Russia), East Asia (China), South and Southeast Asia (India, Sri Lanka, Singapore), and Oceania (Australia, Samoa, Papa New

PLOS ONE
Taxonomy of the rhizomorphic species of Polyporaceae in Yasuní National Park, Ecuador Guinea). Obligated synonyms are P. hemicapnodes Berk. & Broome, P. pusillus Rostr., P. calyculus Pat., & Gaillard, P. savoyanus Pat., Revue, P. subelegans Murrill, P. tephromelas Mont., and P. atripes Rostr. In terms of shape and size, P. leprieurii is similar to Neodictyopus, however, the color of the pilei differs, with light orange to ochraceous hues in P. leprieurri and Neodictyopus with dark-brown to darker-reddish hues. Presumably, P. leprieurri could be a species complex that needs further revision to unveil the phylogenetic relationships among specimens in tropical and subtropical regions as macromorphological variations are not obvious between samples in tropical and subtropical regions. Etymology: yasuniensis (Yasuní = the paradise/God's creation (from the Wao Terero language use by the Huaorani/Waorani Indigenous People living in the Amazon Rainforest between the Napo and Curaray Rivers in Ecuador)) + ensis (Latin suffix to denote origin). In reference to the site where samples were collected, Yasuní National Park in Ecuador.
Emerging from the generative hyphae, thick-walled distinctive stalks (up to 7 μm thick and 25-60 μm length) with pointing apexes somewhat rounded to sharp and up to three branches emerging from the middle to the top sections of the stalks; branches thin-walled, 1-3 μm with simple septa. Skeletal-binding hyphae found in the trama, context, stipe, and rhizomorphs (Fig 13).
In the trama and context, skeletal-binding hyphae are of the arboriform type, hyaline, highly branched, thin and thick-walled (1.5-4 μm) with simple septa. Skeletal-binding hyphae in the stipe and rhizomorphs are also of the arboriform type, somewhat less branched, hyaline, thin and thick-walled (1-5 μm) with simple septa. In context and stipe, plectenchymatous thickenings are common, 4.5-11 μm (ave = 6 μm) wide and 17-60 μm (ave = 68 μm) long. In the stipe, generative hyphae can be found holding cuticle cells, which are yellowish to brown, well compacted forming a palisade layer of cells that emerge from thinner hyphae in groups of three or more rounded and lobed cells forming a crown-like shape structure. The hymenium is composed of paraphysoids structures (cystidioles, basidioles, and hyphidia) and basidia. Hyphidia can be tubular and branched (dichohyphidia), hyaline, smooth, 2.5-6 μm (ave = 3) wide, IKI-. Cystidioles can be thick-walled round-like cells (5x5 μm) and of the ventricose type

PLOS ONE
Rhizomorphs, cylindrical, solid, coriaceous, thin, up to 2 mm diameter and 0.3-1 m long; bearing a black-reddish lacquered cuticle with yellow and white tips when fresh; black and rigid when dried (Fig 12). Rhizomorphs in this taxon are abundant, found in all collection

PLOS ONE
Taxonomy of the rhizomorphic species of Polyporaceae in Yasuní National Park, Ecuador sites (40), distributed on valleys (temporarily flooded forest), slopes, and ridges (terra firme forest). Can be annual to perennial as they can be present for more than four years in colonized areas from 5 m 2 to 20 m 2 . Rhizomorphs emerge from fine and coarse woody debris, standing and downed dead trees. Often rhizomorphs are found attached to trees and other surfaces to maintain their upward growth. Rhizomorphs attach to new substrates and other rhizomorphs by forming a rhizomorphic mat of white mycelium that turns brown-reddish and becomes darker over time. When attached to other rhizomorphs, they form a complex rhizomorphic network covering living and dead plant material found on the forest floor. Rhizomorphs can accompany basidiomata or be alone. It is rare, but basidiomata can emerge from rhizomorphs.
Cultural characteristics: Cultures were easily obtained from rhizomorphs, basidiomata, and colonized substrate. Initial growth is characterized by white, appressed, silky to cottony mycelium that can turn dark brown to reddish. Colonies can develop crustose melanized tissue at sections or on the whole colony. Some colonies might not be melanized at all. Growth rate, characteristics of the mycelial mat, presence of aerial mycelium, and color of the colony are highly variable.
On MEA, the marginal zone is irregular, usually appressed or submerged in the agar giving the appearance that aerial mycelium is absent in the margin. Aerial mycelium white, floccose, silky, cottony; thick, crustose, coriaceous, brown tissue can be formed over time. The agar can have a glossy brown discoloration and chamois-like appearance when the mycelium is submerged; pseudosclerotial plates are often formed. On MYA, margins of colonies are irregular, generally appressed, and raised near the plate's wall. Mycelium can be submerged and aerial; when aerial, silky, floccose, cottony white-creamy with dark radial discolorations. Crustose mycelium is often formed on the melanized sections which become tough coriaceous and

PLOS ONE
Taxonomy of the rhizomorphic species of Polyporaceae in Yasuní National Park, Ecuador difficult to cut overtime. Rhizomorphs developed on culture media but only after long periods (6-12 months) of growth apparently after nutrients have been utilized (on MYA only).
Hyphal system is dimitic, with generative hyphae of the arboriform type with thin and thick-walled sections in the white-aerial mycelium and skeletal hyphae in the melanized-crustose sections of mycelium. Generative hyphae, thin-walled sections 1-2.5 μm wide and thickwalled sections 2-6 μm wide; simple septa and clamp connections; IKI-; KOH-. Skeletal hyphae from crustose sections of mycelium are yellowish-brown, similar to hyphae from the cuticle on stipe and rhizomorphs. Skeletal hyphae are arranged in clusters forming a palisade, individual cells 2-8 μm wide, thicker along branches near the melanized tissue, clamp connections abundant.
Substrate: Basidiomata and rhizomorphs emerge from decayed wood: fine and coarse woody debris, small pieces of lignified tissue from the veins of palm leaves, standing and downed dead trees. In laboratory assays, rhizomorphs can emerge from colonized wood of conifers (data not presented here).
Distribution: Known from the Amazon rainforest of Ecuador along valleys, slopes and ridges (temporary flooded and terra firme forests).
Remarks: The type specimen of P. leprieurii var. yasuniensis was collected in an area of 40 x 30 m 2 in terra firme forest in Yasuní National Park, Ecuador. This collection site was observed to produce rhizomorphs and basidiomata for over four years. This taxon is very common along the forest floor in terra firme and temporarily flooded forests with collection sites that vary from 0.5-40 m 2 . P. leprieurii var. yasuniensis is characterized by annual basidiomata, lateral or eccentrically stipitate, gregarious. Pilei dimidiate, flabelliform or reniform with regular to irregular margins; thin up to 1 mm thick; coriaceous, flexible when fresh and rigid when dried. Hyphae system dimitic with generative and skeletal-binding hyphae of the arboriform type. Basidiospores ellipsoid to narrowly ellipsoid, slightly cylindrical. Cystidia-like terminal thick-walled rounded structures and abundant cystidioles accompanying basidia and the sterile hyphae in the hymenium. Rhizomorphs common, short to long, thin, flexible, coriaceous, non to highly branched, bearing a black-reddish cuticle. Neodictyopus, can be confused with P. leprieurii (Mont.) and P. leprieurii var. yasuniensis, however they differ in the color of the pilei when basidiomata are aged with white-creamy to orange hues in P. leprieurii and P. leprieurii var. yasuniensis and reddish to purplish in Neodictyopus. Macromorphologically, P. leprieurii and P. leprieurii var. yasuniensis have non variations. Microscopically, P. leprieurii (Mont.) and P. leprieurii var. yasuniensis differ in the presence of cystidioles and thick-walled stalks with pointing apex present in the trama of P. leprieurii var. yasuniensis. Hyphal system dimitic with generative and skeletal-binding hyphae. In the trama, generative hyphae, hyaline, highly branched with thin and thick-walled sections (1-5 μm wide) with simple septa and clamp connections (Fig 16).
Cultural characteristics: Cultures were obtained from colonized substrate. Initial growth is characterized by white, appressed, silky to cottony mycelium that grows relatively slow compared to other fungi in the group. There are small variations regarding grow rate, characteristics of the mat, aerial mycelium and color among colonies growing on different types of media. Colonies can develop melanized tissue, however, it is spotted and appears as small dots on the colony which can be lost overtime.
Growth rate on MEA 8.23-13.13 mm (ave = 11.77) and MYA 7.1-21.31 mm (ave = 10.68) in a month of growth is relatively slow compared to other taxa examined here (Fig 17).
On MEA, the marginal zone is irregular, usually appressed or submerged in the agar giving the appearance that aerial mycelium is absent. Aerial mycelium, white to creamy, compacted, cottony at sections. On MYA, margins of colonies are irregular, generally submerged and slightly appressed. Aerial mycelium, silky, compacted, slightly floccose, white to creamy, but not melanized.
Substrate: Basidiome and rhizomorphs emerge from decay wood (coarse and fine woody debris).
Distribution: Known from the Amazon rainforest of Ecuador along ridges and slopes (terra firme and temporary flooded forests).
Remarks: The type specimen of P. taromenane was described from a single basidiome collected in October 2019 in an area of 0.5 m 2 in terra firme forest in Yasuní National Park. The collection site produced rhizomorphs over four years, but only one basidiome was found within that time period. This taxon is somewhat rare compared to P. leprieurii var. yasuniensis, with only three collections within the 50-ha study site. P. taromenane is characterized by annual basidiomata, central, stipitate, solitary. Pileus circular, infundibuliform, depressed with regular hirsute margins. Big angular to slightly elongated pores. Hyphal system dimitic with generative and skeletal-binding hyphae of the arboriform type. Skeletal-binding hyphae in context and stipe with plectenchymatous thickenings. Basidiospores narrowly ellipsoid to cylindrical. Sterile structures present in the hymenium: ventricose cystidioles are rare; hyphidia common; basidioles clavate, abundant. Rhizomorphs circular, solid, robust, long, somewhat not branched, flexible, coriaceous, bearing a black-reddish cuticle and small brown hairs. P. taromenane is closely related to P. guianensis Mont., with this, it shares macromorphological similarities in terms of, pilei color and texture, color and texture of the stipe, pore's shape and size. Micromorphological differences with P. guianensis, include the presence of sterile elements in the hymenium and presence of plectenchymatous thickenings in the skeletal-binding hyphae in the context and stipe. There are no reports of rhizomorph development in P. guianensis, which have most reports in Tropical Asia.

Discussion
In this work, we contribute to the knowledge of fungal diversity in Yasuní National Park by describing four new species and one new variety, three of them producing aerial rhizomorphs. Yasuní National Park, the biggest natural reserve in Ecuador, is highly biodiverse as well as highly threatened by illegal logging, hydrocarbon and mining projects, as well as land transformation due to oil palm plantations, and large-scale transportation projects [36]. Understanding the true biodiversity of poorly studied groups like fungi, and the role they play in such a biodiverse ecosystem is needed to unveil their ecological significance.
The newly described species Atroporus yasuniensis, Atroporus tagaeri, Neodictyopus sylvaticus, and Polyporus taromenane, and the new variety Polyporus leprieurii var. yasuniensis, were named after Yasuní National Park and the voluntarily isolated indigenous groups (Tagaeri and Taromenane) that live in the area. The results from this study strongly suggests additional taxonomic studies are needed to identify the other very diverse fungi present. A better understanding of the organisms found in the forest will help to enhance conservation efforts and protect this highly biodiverse ecosystem and the people living there.
This taxonomic study was aimed to identify the fungal species producing rhizomorphs that are present in Yasuní National Park. Rhizomorphs are versatile structures and the role they have in colonization, survival, pathogenicity, and ecological function in forest ecosystems has been well documented on subterranean temperate fungal species such as the Armillaria species complex [37][38][39][40]. However, little is known of the role aerial rhizomorphs have in the tropics or their ecological function in such a diverse forest ecosystem [3,10,12,41,42]. To contribute to this understanding, we identified and described fungi producing rhizomorphs found on the forest floor. The information presented here will help to better understand the diversity of rhizomorph-forming fungal species and contribute to future ecological studies looking at the role these species play in this ecosystem.
The work took place in one of the 72 permanent forest plots for the long-term study at the Forest Global Earth Observatory (ForestGEO). Permanent forest dynamic plots have become an important starting point for the development of research in tropical ecosystems across a wide range of fields [43]. The little anthropogenic disturbance that occurs in Yasuní Forest Dynamic Plot allowed us to observe and report growth patterns of fungi, including the colonization of new substrates and the production of basidiomata and rhizomorphs across four years of sampling.
Identifying the taxa producing these rhizomorphs can be challenging if basidiomata are not present on the substrate. In addition, for rare species, the likelihood of finding basidiomata and rhizomorphs associated with the species in the same substrate can be very difficult. In the research presented here for taxa associated with the P. leprieurii group, only 35 collections out of 120 had basidiomata on the substrate, or near the collection sites where rhizomorphs were collected. In the case of P. taromenane, only three collections among all the surveyed quadrants and transects were found, with only one basidiome present in the same site where rhizomorphs were collected in previous years. In the Atroporus group, only A. yasuniensis has been found to develop large complex systems of rhizomorphs that are well distributed along the forest floor. Basidiomata for this species were only found after the third year of sampling in the area. Basidiomata of other species in the Atroporus genus were collected in similar areas, but rhizomorphs associated with them were not found. This exemplifies the challenge mycologists face while linking rhizomorphs to specific taxa, especially when surveys are not repeated in the same site.
The type of rhizomorphs found associated with the taxa studied share some similar morphological features, such as upward growth, coriaceous texture, bearing a dark cuticle, formation of a rhizomorphic/mycelial mat to attach to substrates, and anastomosis of rhizomorphs to form a complex network. However, in the case of P. taromenane, rhizomorphs in addition to having a black cuticle also have small hairs that give rhizomorphs a velvety appearance and distinguish it from the other taxa in the "Melanopus group". Phylogenetically, P. taromenane is quite distance, and we hypothesize there may be other related species in this clade.
Within the Polyporus genus and allied genera, rhizomorphs have been reported to be formed as a response to environmental stress [4]. In our study, rhizomorphs were found growing along ridges, slopes (terra firme forest), and valleys (temporarily flooded forests), ecosystems that are exposed to seasonal droughts and exceeding wet conditions. In both ecosystems, the production of rhizomorphs in areas exceeding the 5 m 2 of the colonized substrate was vast and continuously developed during the four years of sampling. It appears that the establishment of rhizomorphs allows these fungi to survive in areas prone to environmental stresses linked to excessive or limited water availability, and/or to be formed as a strategy to find new substrates and obtain new sources of nutrients. This role of rhizomorph utility was supported by our observation during the cultural characterization assays when rhizomorph development occurred rarely on more than six-months old MYA cultures. However, the influence of water stress and scarce nutrients on the substrate needs further examination to establish the exact influence on the rhizomorph formation within the species in the "Melanopus group". In addition to the stress response, we believe rhizomorphs provide these species with mechanisms that allow them to prevail in harsh environmental conditions including floods, desiccation during dry periods, and damage from UV light. In addition, melanized rhizomorphs appear to restrict insect predation. This has also been observed in rhizomorphic fungi from temperate forests [37][38][39][40]44]. The complete biological significance of rhizomorphs in the "Melanopus group" should be studied in greater detail to better elucidate their important role in the ecosystem.
The taxa examined in this study are part of the artificially designated "Melanopus group", within the Polyporaceae family, many of them previously classified as Polyporus P. Micheli ex Adans [4]. The group has undergone important recategorization of its taxa since phylogenetics became a tool in taxonomic studies [14][15][16]. Such studies helped to identify the polyphyletic origin of taxa in the group. The new taxa describe in this study were proposed by phylogenetic differences using reference sequences of closely related taxa.
The "leprieurii" clade appears to be a species complex with taxa morphologically more similar in the neotropics, sharing macro and micromorphological similarities with the original description of the type specimen of P. leprieurii Mont. These include the position of the stipe, color of the pilei, size of the pores, shape and size of basidiospores, and the presence of rhizomorphs. In the absence of the type specimen, an herbarium sample (BZ-958) from Belize and a reference sequence from Costa Rica were used to compare the phylogenetic and morphological similarities between the samples from Ecuador and specimens from other locations. Samples from Belize and Costa Rica were phylogenetically similar to each other, but distant to the specimens from Ecuador. Micromorphologically, the specimen from Belize was more similar to the type description, as both lack cystidia-like cells, but other hyphal descriptions are similar. The new variety, P. leprieurii var. yasuniensis differs from the type specimen and herbarium sample since it had cystidioles of the ventricose type, rounded cystidia-like cells, and distinctive thick-walled stalks with pointed apex and 1-3 branches accompanying the hyphae in the trama. These distinctive hyphal characteristics have been described and used as diagnostic of Atroporus [17]. The specimens described as P. leprieurii followed the same trend as those in the former P. dictyopus species complex, having macromorphological similarities, but being genetically distant. The true phylogenetic relationships withing the "leprieurii" clade require a detailed study among specimens from different regions, especially new collections from the type location. For now, we have identified a new variety that differs from the type specimen on the presence of cystidioles and distinctive hyphae in the trama.
The "guianensis" clade was strongly supported by Maximum Likelihood and Bayesian inferences scores. Unfortunately, it was not possible to use the type specimen in this study. However, reference sequences were used to differentiate our samples. Sequences from Argentina and Venezuela remained as P. guianensis. Two other sequences from Ecuador (ECU 1 and ECU 2) and the Costa Rica sample seem to be different from P. guianensis Mont. and P. taromenane species novo. The genetic difference between our sequences and the reference sequences from Argentina and Venezuela are significant, as they share only 75% sequence identity. Sequences ECU1 and ECU2 and from Costa Rica are closely related to the newly proposed species, P. taromenane from Yasuní, however, they are still significantly different from P. taromenane. There are macro and micromorphological similarities between P. taromenane and P. guianensis, including stipe position, stipe surface, pilei shape, size of pores, hyphal system, and basidiospores narrowly ellipsoid to cylindrical. However, P. guianensis is known to be common in tropical regions of Asia, our phylogenetic analysis confirms that these are two distinct taxa.
Taxa belonging to Picipes Zmitr. et Kovalenko were included to help reveal the phylogenetic relationships among the taxa sampled in Yasuní and allied genera within the "Melanopus group". The "picipes" clade represented in this study includes six distinct species from China.
The "neodictyopus" clade includes sequences from Belize, Brazil, and Ecuador with five distinct taxa in the clade. Neodictyopus Palacio, Robledo, Reck & Drechsler-Santos, was proposed by Palacio et al. (2017), based on phylogenetic analysis and morphological descriptions of freshly collected and herbarium samples previously identified as P. dictyopus. Five of the eight Neodictyopus samples used in our analysis come from Brazil, two from Yasuní National Park in Ecuador, one from Cuyabeno-Ecuador, and one sample from Belize [45]. Neodictyopus sylvaticus species novo shares general morphological features with N. atlanticae, the type species for the genus. These features include the position of the stipitate, pileus shape and color, hyphal system, and shape of basidiospore. Although, these characteristics can be shared with other taxa within the genus and Picipes Zmitr. et Kovalenko (2016). Genetically, N. dictyopus is the closest taxon with 80% sequence identity, being highly distinctive along the ITS region with only 60% sequence identity. In addition, relevant macro and micromorphological differences were found between N. dictyopus and N. sylvaticus, including smaller basidiomata, smaller pores, stipe shape, the presence of ventricose tramal setal, and generative hyphae highly branched, slightly cylindrical along the main stem which are unique characteristics of N. sylvaticus. We believe these morphological differences and the phylogenetic results support the proposal of this new species. Additionally, in the "neodictyopus" clade, EC-97 herbarium sample previously identified as P. leprieurii by Laessoe (1995), is now identified as a taxon within the Neodictyopus genus. Pilear shape, color and texture; stipe position; and spore characteristics are macro and micromorphological features shared within several taxa in the "Melanopus group". Since most members have broad morphological plasticity, such characteristics might influence the misidentification of some taxa. As Palacio et al. (2017) pointed out, the color of the pilear surface could be a broad and obvious characteristic to distinguish taxa between the "leprieurii" and "neodictyopus" groups. In the "leprieurii" group, the pilear surface can have orange-ochraceous hues, while the "neodictyopus" group has darker colors of reddish-brown hues. Other microscopical features such as shape and size of spores are broadly shared in the Polyporaceae family, therefore the hyphal system and other structures found in the hymenophore should be considered while identifying a specific taxon in addition to spore descriptions. The "neodictyopus" clades presented by Palacio et al., (2017) and our study include only taxa from the neotropics, and supports the information presented by Palacio et al., (2017) that the genus is neotropical. In our study, the "neodictyopus" clade is closer to "picipes".
The "atroporus" clade presented in our study includes sequences from Brazil and Ecuador with strong Maximum likelihood bootstrap and Bayesian inference scores. Atroporus was proposed by Ryvarden (1973) with Polyporus diabolicus Berk. as the species type. It was recently amended by Palacio et al. (2017), using the descriptions of the basidiospores and strongly dextrinoid skeletal-binding hyphae in the trama with a distinctive thick-walled stalk with a pointed apex. In this clade, A. yasuniensis and A. tagaeri were two new species described in this study. A. yasuniensis was closely related to A. rufoatratus, but with clear genetic and morphological differences. Genetically, A. yasuniensis and A. rufoatratus had only 88% sequence identity. Morphologically, A. yasuniensis had a gregarious habit, big basidiomata, smaller pores, and developed of rhizomorphs compared to the solitary, small basidiome with relatively larger pores and absence of rhizomorphs of A. rufoatratus. Microscopically, A. rufoatratus had subulate cystidioles and A. yasuniensis had ventricose cystidioles.
The second newly described species, A. tagaeri was genetically different from A. diabolicus, A. rufoatratus, and A. yasuniensis with 83%, 80%, and 84% of sequence identity respectively. Morphologically, A. tagaeri shared macromorphological features with other taxa in the clade, including centrally stipitate basidiomata with a circular, slightly depressed pileus and the characteristic black cuticle bearing the stipe. Basidiomata of A. tagaeri were small to medium and pores were smaller and similar to A. yasuniensis. A remarkable difference between A. tagaeri and other taxa in the "atroporus" clade was that in A. tagaeri, the characteristic thick-walled stalk with pointed apex was not found. Considering this is an important characteristic for the description of the genus, it is necessary to carry out additional collections to verify this feature, which was also reported to be present in taxa outside Atroporus and has taxonomic value to differentiate those groups. Plectenchymatous thickening of hyphae were rare, but found in A. tagaeri. These segments of hyphal modifications appeared to be common on N. sylvaticus and P. taromenane, and its presence should be used for micromorphological descriptions, but not considered to have taxonomic value. These hyphal modifications have been previously described to be present in Polyporus s. str., Phellinus and Daedaleopsis [24] as well as on several species of lichens, ectomycorrhizal, and other taxa in the Agaricomycetes fungi [46][47][48][49][50][51][52][53][54]. These hyphal modifications seem to be associated with interlocking hyphae involved in the development of specific tissue accomplishing a specific function. In our results we observed that plectenchymatous thickenings can be common or rare among taxa and can be present along the skeletal-binding hyphae in trama, context, stipe and rhizomorphs. There is still not a clear function of this tissue in the taxa we described, however, it is important to recognize it for future comparisons.
Finally, sequences (ITS, LSU and EF1-α) corresponding to two samples of P. tuberaster from China were included, as P. tuberaster is the type species for the Polyporus genus, however, the type sequences were not use in this study. This last clade is apparently not related to any of the clades formed in our phylogenetic analysis. The use of genetic data accompanying morphological descriptions has been key to unveil the true taxonomy of several members in Polyporus s. l. We believe that more genera will be proposed as additional collections and analyses are carried out in the Amazonian regions of South America.
To accomplish the job of unveiling the true phylogenetic relationships in the "Melanopus group" and other artificial groups within the Polyporaceae family, researchers need to select molecular markers that match reference sequences, generate more sequences of taxa from neotropical, pantropical, and temperate regions and compare them with the taxa of interest.

Conclusions
We described four new species and one new variety of ecologically important fungi from the Amazon rainforest of Ecuador. Three of these fungi produce aerial rhizomorphs that provide these species with strategies to colonize new substrates and survive in environmentally challenging ecosystems that seasonally exhibit flooding and temporarily drought conditions.
Polyporus leprieurii var. yasuniensis and taxa within the "leprieurii" subclade are prolific species in the forest at Yasuní National Park where they produce rhizomorphs, often accompanied by basidiomata. While, Atroporus yasuniensis has been found producing abundant rhizomorphs and basidiomata growing on the same substrate, this species seems to be rare with only one collection site among all the sampled quadrants and transects we studied. Polyporus taromenane is also rare with only three collection sites, however the amount of rhizomorphs it produces is not as abundant as the other species.
Rhizomorphs have not been typically considered to be of taxonomical value. However, in this study we observed that rhizomorphs exhibit morphological characteristics similar to the stipes of basidiomata for each species. This feature helps to differentiate the taxa and can be important for identification when rhizomorphs are found but basidiomata are not present. Polyporus taromenane appears to have the most distinctive rhizomorphs and stipe, which has hirsute hairs giving it a velvety appearance. Also, basidiome production seems to be rare in the rainforest and the presence of its rhizomorphs appears to be sufficient to identify this taxon when found.
Macrocromorphological features regarding pileus and stipe shape, texture, color and position are useful to identify taxa within artificial groups and/or fungal families. Although, these features are broadly shared between the groups, some of them may be used to separate some genera, like the color of pileus between P. leprieurii and Neodyctiopus. However, these characteristics are not useful when identifying species complexes. Micromorphological structures should be also carefully considered while identifying and proposing new taxa. For example, spore shape, size, and texture are commonly shared in the examined taxa and in many members within the Polyporaceae family. Other characteristics regarding hyphal types need further revision among more representatives of a specific taxon, as some characteristics like the skeletal-binding hyphae with a distinctive "sharply pointed apex" used to propose Atroporus was also found in P. leprieurii var. yasuniensis.
Phylogenetic analyses have become an important tool in taxonomic studies, and we believe this should be used along with morphological descriptions. In this and similar studies, the lack of reference sequences available shows the ongoing need to collect and access genetic material of fungi to create better databases that allow us to understand the true diversity and distribution of these species, especially in highly diverse ecosystems where there is large fungal diversity. Finding enough collections to carry out phylogenetic and taxonomic studies can be challenging when species are rare and inhabit endangered ecosystems. Collaborations among Fungarium and mycological institutions, and researchers are key to access relevant reference sequences and taxa in aims to unveil the true fungal diversity of poorly studied ecosystems and regions such as the Amazon Rainforest of Ecuador. and valuable work during the field session. To Connor Lund, Leah Grim, Amanda Stear and other people in the Blanchette lab for their help during the laboratory work and the staff at Yasuní Research Station in Ecuador for providing support during the field investigations. To Eric C. Otto, Benjamin Held, and Robert Alvarez-Quinto for their valuable comments on and revisions of the manuscript. To the USDA Forest Products Laboratory, Center for Forest Mycology Research, Wisconsin, USA for loaning the herbarium specimens used in this study.