A new monotypic sequestrate genus, Longistriata is described based on collections from the Neotropical forest of Atlantic forest in Paraíba, Northeast Brazil – an area known for its high degree of endemism. The striking features of this new fungus are the hypogeous habit, the vivid yellow peridium in mature basidiomes, broadly ellipsoid basidiospores with a distinct wall that is ornamented with longitudinal striations and lageniform cystidia with rounded apices. Phylogenetic analysis, based on LSU and tef-1α regions, showed that the type species, Longistriata flava, is phylogenetically sister to the monotypic sequestrate African genus Mackintoshia in Boletaceae. Together these two species formed the earliest diverging lineage in the subfamily Zangioideae. Longistriata flava is found in nutrient-poor white sand habitats where plants in the genera Coccoloba (Polygonaceae) and Guapira (Nyctaginaceae) are the only potential ectomycorrhizal host symbionts.

Boletales, ITS, phylogeny, sequestrate fungi, taxonomy, tropical forest.

Fungi in the order Boletales (Agaricomycetes, Basidiomycota) comprise a morphological diverse group including agaricoid, boletoid, gasteroid, secotioid, corticioid, merulioid, hydroid and polyporoid forms (Binder and Hibbett 2006) with ectomycorrhizal (ECM), saprophytic or ligninolytic members (Kirk et al. 2008). The order is a globally distributed group of mushroom-forming fungi growing in most forest ecosystems (Chu-Chou and Grace 1983; Binder and Hibbett 2006). Despite thorough morphological (Rolland 1899; Høiland 1987; Pegler and Young 1989; Montecchi and Sarasini 2000) and phylogenetic coverage of the order Boletales (Kretzer and Bruns 1999; Binder and Bresinsky 2002; Binder and Hibbett 2006; Orihara et al. 2016a) new phylogenetically supported genera are still being discovered, particularly representatives with a sequestrate habitat (Nuhn et al. 2013; Wu et al. 2014, 2016). The sequestrate habitat has arisen in this order multiple times and a large number of sequestrate genera in Boletaeceae have been described: Carolinigaster M.E. Sm. & S. Cruz (Crous et al. 2018), Chamonixia Rolland (Binder and Bresinsky 2002; Orihara et al. 2016a), Heliogaster Orihara & Iwase (Orihara et al. 2010), Kombocles Castellano, T.W. Henkel & Dentinger (Castellano et al. 2016), Octaviania Vittad. (Vittadini 1831; Orihara et al. 2012), Mackintoshia Pacioni and Sharp (2000). Rhodactina Pegler and T.W.K. Young (Yang et al. 2006; Vadthanarat et al. 2018), Rossbeevera T. Lebel and Orihara (Lebel et al. 2012), Royoungia Castellano, Trappe and Malajczuk (Castellano et al. 1992), Solioccasus Trappe et al. (Trappe et al. 2013), Turmalinea Orihara and N. Maek. (Orihara et al. 2016b) and Afrocastellanoa M.E. Smith & Orihara (Orihara and Smith 2017).

Sequestrate Boletaceae have been described from across the globe with records from all continents except Antarctica but relatively little is known about sequestrate boletoid fungi in South America (Putzke 1994; Sulzbacher et al. 2017). Species of Rhizopogon Fr. & Nordholm (Fries and Nordholm 1817) and Scleroderma Pers. (Persoon 1801) are broadly distributed and most frequently recorded in forest plantations with introduced pines, eucalypts or pecan trees (Martín 1996; Giachini et al. 2000; Baseia and Milanez 2002; Nouhra et al. 2012; Sulzbacher et al. 2016a, 2018). However, there are relatively few citations of sequestrate taxa from native ectotrophic forests. Examples from temperate habitats include Alpova austroalnicola L.S. Domínguez in Alnus acuminata Kunth ssp. acuminata forests in the Yunga District of Argentina (Nouhra et al. 2005) and Scleroderma patagonicum Nouhra & Hern. Caff. in Patagonian Nothofagus forests (Nouhra et al. 2012). Recently, undescribed taxa of sequestrate Boletaceae were cited from tropical forests in Guyana (Henkel et al. 2012; Smith et al. 2013) and formally described as Jimtrappea T.W. Henkel, M.E. Smith & Aime, Castellanea T.W. Henkel & M.E. Sm. and Costatisporus T.W. Henkel & M.E. Sm. (Smith et al. 2015). These new records from the Guiana Shield suggest that other unexplored tropical forests in South America may host additional diversity of sequestrate Boletales, similar to recent reports from Asia and Africa (Castellano et al. 2016; Chai et al. 2019).

In Brazil, there are numerous surveys that have documented epigeous Boletales in exotic plantations and native forests (Rick 1961; Guzmán 1970; Putzke 1994; Watling and de Meijer 1997; Baseia and Milanez 2000; Giachini et al. 2000; Baseia and Milanez 2002; Sobestiansky 2005; de Meijer 2006; Gurgel et al. 2008; Cortez et al. 2011; Magnago and Neves 2014; Barbosa-Silva and Wartchow 2017; Barbosa-Silva et al. 2017; Magnago et al. 2017a, 2017b 2018, 2019). However, information related to sequestrate hypogeous fungi is scanty (Sulzbacher et al. 2016a).

As part of recent studies on ectomycorrhizal and sequestrate fungi in northeastern Brazil (Sulzbacher et al. 2013, 2017), we collected a sequestrate taxon that could not be assigned to any current species in the family Boletaceae. Here we describe and characterize the new sequestrate boletoid species in a newly erected genus Longistriata based on sequence analyses of the ITS, nLSU, and TEF1 molecular markers as well as detailed analysis of morphological features. From available collections and publicly available sequences we discuss how this new species differs from all currently described genera in Boletales and we discuss the trophic mode of this new species and genus.

The nLSU + TEF1 combined dataset consisted of 85 taxa and 2,014 aligned nucleotide positions. The Bayesian inference reached convergence after 4.6M generations. We therefore discarded the first 4,600 trees in each chain, and the remaining 5,401 trees in each chain were summarized to approximate Bayesian posterior probabilities (PPs). ESS of all the model parameters were sufficiently large (>200). The total arithmetic and harmonic means of Likelihoods (lnL) were -29,498.16 and -29,562.71, respectively. In RAxML analysis the log likelihood of the ML tree was -29,121.825209.

The nLSU + TEF1 combined tree of the Boletaceae supported our hypothesis that the sequestrate basidiomes of the vivid yellow fungus belong to an undescribed genus in the Boletaceae (Fig. 2). The species described here as Longistriata flava Sulzbacher, Orihara, Grebenc, M.P. Martín & Baseia, sp. nov. formed a sister lineage to the African monotypic sequestrate genus Mackintoshia (KC905034) with moderate to high statistical support (PP = 1.0, ML-BS = 59%). The phylogenetic analyses further suggested that the Longistriata-Mackintoshia clade is the earliest diverging lineage within the subfamily Zangioideae (PP = 1.0, ML-BS =60%). The epigeous yellowish bolete species (Tylopilus sp. Sulzbacher 454 in Suppl. material 1: Table S1) that sometimes occurred sympatrically with Longistriata flava was distantly related to L. flava and was instead more closely related to Tylopilus balloui. Other genera of Boletaceae that are closely related to Longistriata based on our phylogenetic analysis are species of Australopilus Halling & Fechner, Chiua Yan C. Li & Zhu L. Yang, Harrya Halling, Nuhn & Osmudson, Hymenoboletus Yan C. Li & Zhu L. Yang, Royoungia Castellano, Trappe & Malajczuk, and Zangia Yan C. Li & Zhu L. Yang. All sister clades have significant bootstrap support in phylogenetic analyses and show a range of morphological differences that support the erection of Longstriata as a separate genus. The ITS rDNA barcode sequences of L. flava specimens UFRN-fungos 1756 and UFRN-fungos 2110 were 751 bp in length (Suppl. material 1: Table S1). These sequences were less than 93% similar to all other ITS rDNA sequences in the INSD database. Below we describe this new genus and species and provide detailed morphological analysis and direct comparison with previously described sequestrate Boletaceae.

Figure 2. 

A nLSU + TEF1 combined Maximum likelihood (ML) phylogram showing the phylogenetic relationship of Longistriata (UFRN-Fungus 1756, holotype) in relation to representative taxa in the Boletaceae. Non-parametric bootstrap branch supports (MPbs / MLbs) are given for nodes with bs>50.

Longistriata is a striking new monotypic genus described from the Atlantic forest in the Northeastern part of Brazil. The only known representative of the genus is the newly described Longistriata flava. This species is characterized by the hypogeous habit, a smooth and bright yellow peridium (Fig. 3), presence of cystidia, and the absence of clamp connections in all tissues (Fig. 4). Based on a combined phylogenetic analysis of nLSU + TEF1 the closest relative is Mackintoshia persica (Fig. 2). However, L. flava is differentiated from M. persica (Pacioni and Sharp 2000) based on its well-developed, sterile base that forms a short stipe, lageniform cystidia with rounded apices, basidiospores with persistent sterigmal attachments that are covered by thin longitudinal striations and spores that are 15–19 × 13–16 µm. Mackintoshia has smooth and smaller elliptical basidiospores (8–12 × 5–7 μm) (Pacioni and Sharp 2000) and is known only from Africa (Castellano et al. 2000). The two genera also have different host plants; Mackintoshia is found in habitats dominated by ECM plants in the Fabaceae and Uapacaceae (Pacioni and Sharp 2000) whereas Longistriata is found with ECM plants in Nyctaginaceae and Polygonaceae. This combination of morphological features is unique, separating the sister clade Mackintoshia from Longistriata.

In addition to Longistriata several other genera of sequestrate Boletaceae, Chamonixia, Rossbeevera, Rhodactina, and Turmalinea, also have basidiospores with longitudinal ridges. However, members of these genera all differ in the shape and number of ridges. Rosbeevera has ellipsoid to fusiform basidiospores with 3–5 ridges (Lebel et al. 2012), Chamonixia has subglobose to broadly ellipsoid basidiospores with 6–10 ridges (Lebel et al. 2012), Rhodactina has broadly ellipsoid to subfusiform basidiospores with 8–10 ridges (Yang et al. 2006), and Turmalinea has ovoid to fusoid basidiospores with 5–10 longitudinal ridges that are often branched to irregularly broken and spores can be with or without a hilar appendage (Orihara et al. 2016b). The spore colors are also different in these other genera; Rossbeevera and Chamonixia have brown to brownish spores (Montecchi and Sarasini 2000; Lebel et al. 2012), Turmelinea has inamyloid, non-dextrinoid, spores that are brick red to dark brown at maturity (Orihara et al. 2016b) and Rhodactina species have spores that are deep purple (Yang et al. 2006). In Longistriata the number of ridges is greater than in any of the other genera. On average, spores of Longistriata flava have 10 complete ridges across the longitudinal axis of the spore with additional irregular, thin and low ridges (Figs 3F–G, 4D). Furthermore, the ridges in this species are thin, low and irregular as compared to the other genera listed above. In some spores the ridges of Longistriata can be fused or bifurcating (Fig. 5C). Species in the sequestrate genus Gautieria also has spores with longitudinal ridges but this genus is very different from Longistriata because the basidioma of Gautieria species often lack a peridium and they belong in the distantly related order Gomphales (Montecchi and Sarasini 2000; Giachini et al. 2010). The ridged basidiospores of Longistriata are also superficially similar to those of the epigeous bolete genus Boletellus because species in both genera typically have longitudinal ridges. However, our phylogenetic analyses indicate that these two genera are only distantly related within the Boletaceae (Fig. 2).

The hypogeous habit, shape of basidiomes (e.g. globose, subglobose, tuberiform) and the rudimentary sterile base in L. flava suggest a possible relationship with the sequestrate truffle-like genus Octaviania (Orihara et al. 2012). However, the basidiospores are very different in Octaviania (e.g. globose to ellipsoid spores with ornamentation of large, thick-walled, pyramidal to conical ornaments) and molecular data indicate that Octaviania is a distant relative of Longistriata. The bright yellow peridium of fresh basidiomata and the presence of a stipe in L. flava resembles members of the Boletus chromapes group (e.g. Zangia and Harrya) as well as the genus Royoungia where at least some taxa have similar bright yellow coloration (Li et al. 2011; Halling et al. 2012).

The phylogenetic analyses suggest that the new Brazilian genus is closely related to several genera in the subfamily Zangioideae that also have bright yellow colors at the base of the stipe (e.g. Chiua Yan C. Li & Zhu L. Yang, Harrya Halling, Nuhn & Osmudson, Royoungia Castellano, Trappe & Malajczuk, and Zangia Yan C. Li & Zhu L. Yang) (Wu et al. 2014, 2016) (Fig. 2). Within the Zangioideae only one hypogeous sequestrate taxon, Royoungia boletoides, was previously known (Wu et al. 2014, 2016). All of the other genera in Zangioideae are characterized by the epigeous habit, with a well-developed and central stipe and smooth basidiospores. The fresh appearance of Longistriata flava, with its bright yellow peridium, resembles the colors found in Chiua or Zangia (from Asia with Fagaceae and Pinaceae) or Royoungia (from Australia with Myrtaceae) (Li et al. 2011; Halling et al. 2012; Wu et al. 2016).

Several other sequestrate Boletaceae are similar to L. flava, either in their morphology or in their tropical distribution. Members of the sequestrate genus Mycoamaranthus Castellano, Trappe & Malajczuk also produce bright yellow basidiomata and belong to Boletaceae (Binder and Hibbett 2006) but GenBank BLASTn queries based on the ITS rDNA indicate that Longistriata is distantly related to Mycoamaranthus (e.g. the ITS is <85% similar to both Mycoamaranthus congolensis and M. cambodgensis). Another genus that shares several morphological similarities with Longistriata is the genus Solioccasus (Trappe et al. 2013). This genus differs from Longistriata by the large and copious rhizomorphs appressed to peridial surfaces, a dendroid and cartilaginous columella, smooth basidiospores, and basidiomes with bright orange and reddish coloration. The monotypic genus Afrocastellanoa from tropical Africa (Orihara and Smith 2017) is distinct from Longistriata because it is characterized by whitish basidiomata, globose to subglobose basidiospores with warty to spiny spore ornaments and its phylogenetic relationship with the epigeous genus Porphyrellus. Recently, Smith et al. (2015) discovered three new monotypic hypogeous sequestrate genera within Boletaceae, Jimtrappea guyanensis T.W. Henkel, M.E. Smith & Aime, Castellanea pakaraimophila T.W. Henkel & M.E. Smith and Costatisporus cyanescens T.W. Henkel & M.E. Smith (Smith et al. 2015). Like Longistriata, all three new genera are endemic to tropical South America but their macro- and microscopic characteristics are notably different. Jimtrappea guyanensis is characterized by the white peridium, unchanging tissues, short columella, smooth subfusiform, reddish brown basidiospores and prominent dextrinoid cystidia (Smith et al. 2015). The white peridium, the unchanging tissues, smooth subfusiform basidiospores, and prominent cystidia of J. guyanensis contrast with the yellow peridium and unique basidiospores morphology of L. flava. Phylogenetic analysis also confirms that J. guyanensis and L. flava are not closely related. Castellanea pakaraimophila is similar to L. flava because both have subglobose basidioma and a short stipe but L. flava has a bright yellow peridium. The two species can also be easily differentiated by their spores because C. pakaraimophila has smooth, subfusiform basidiospores whereas L. flava has ellipsoid spores with distinct longitudinal striations (Smith et al. 2015). Costatisporus cyanescens is easily differentiated from L. flava by its grayish yellow peridium and the dark blue staining reaction on the peridium. Microscopically, the longitudinally ridged basidiospore ornamentation of C. cyanescens (fig. 4c, in Smith et al. 2015) is similar to that in L. flava (Fig. 5A–D). However, the basidiospores are broadly ellipsoid and dextrinoid in L. flava and the ornamentation is formed by a series of thin longitudinal striations on all surfaces. In contrast, the spores of C. cyanescens are unreactive in Melzer’s reagent and are ovate to subfusiform with ridges that are somewhat spiraled. The two species are also distantly related based on our phylogenetic analysis (Fig. 2). The ecology of Jimtrappea, Castellanea and Costatisporus are also different from Longistriata. These three genera are apparently endemic to the Guiana Shield and are associated with the ECM tree genera Aldina, Dicymbe and Pakaraimaea (Smith et al. 2015).

Unfortunately, we have not yet confirmed the ECM status of Longistriata flava based on sequences from ECM root tips from native Brazilian trees. However, basidiomes of Longistriata have always been collected in the lowland semi-deciduous forest to savanna known as “tabuleiro” in close proximity to woody plants in the ECM genera Coccoloba (Polygonaceae) and Guapira (Nyctaginaceae). Given that ECM plants in these genera are known to host a wide array of ECM fungi from other sites in tropical South America (Tedersoo et al. 2010; Séne et al. 2015; Põlme et al. 2017) and that other taxa in the Zangioideae are known to be ECM (Tedersoo and Smith 2013), we hypothesize that L. flava is also ECM. The ECM nutritional mode is also likely to be favored in the nutrient-poor sandy soil ecosystem of Neotropical forest fragments of the Atlantic Forests.

The authors wish to thank Dr. Jomar G. Jardim for identifying tree species in the Guaribas Biological Reserve. This study is in part a result of the Ph.D. thesis of Marcelo A. Sulzbacher, with a scholarship provided by the Brazilian Government (CAPES). The work was partially supported by CNPq project PVE 407474/2013-7 and co-financed by the Brazil-Slovenia bilateral project (BI-BR/11-13-005(SRA) / 490648/2010-0 (CNPq)), research project J4-1766 “Methodology approaches in genome-based diversity and ecological plasticity study of truffles from their natural distribution areas” and the Research Program in Forest Biology, Ecology and Technology (P4-0107) of the Slovenian Research Agency.