Synonymy of Suillus imitatus , the imitator of two species within the S . caerulescens / ponderosus complex

Identification of three species of Suillus, S. caerulescens, S. ponderosus, and S. imitatus, has always been difficult because of overlapping and non-discrete morphological characters. To solidify the identification of these taxa, we compared the nucleotide sequences from the internal transcribed spacer region (ITS) of the type specimens of S. caerulescens, S. ponderosus, S. imitatus var. imitatus, and S. imitatus var. viridescens with field collected specimens which we identified as S. caerulescens, S. ponderosus, and S. imitatus in northern California. Based on ITS sequence identity and phylogenetic inference, specimens of S. caerulescens and S. ponderosus formed well-supported clades with the holotype of the respective species. However, S. imitatus var. imitatus fits within the S. caerulescens clade and S. imitatus var. viridescens fits within the S. ponderosus clade. Therefore, we synonymize S. imitatus var. imitatus with S. caerulescens and S. imitatus var. viridescens with S. ponderosus, and show that the species can indeed be recognized morphologically based on annulus characteristics. Key words—bolete, phylogenetics Introduction Suillus Gray is a large ectomycorrhizal genus in the Boletales with about 170 known species (Index Fungorum and Mycobank); about half of those species are found in North America (Bessette et al. 2000) and about 40 species in the western United States (Smith 1979). The four taxa in this study, S. caerulescens, S. ponderosus, S. imitatus var. imitatus, and S. imitatus var. viridescens (all within section Boletinus), are restricted to the North American west coast. Smith & Thiers (1964) described S. caerulescens, S. ponderosus, and S. imitatus as three distinct species based mainly on texture of the annulus/annulus remnant and pileus color. Later, Smith & Trappe (1972) described S. imitatus var. viridescens from western Oregon based on the strong greening reactions of cut tissue in the 390 ... Nguyen & al. Table 1. Morphological comparisons of four Suillus taxa. * S. caerulescens S. ponderosus S. imitatus var. imitatus S. imitatus var. viridescens Holotype MICH12297 (Smith 48733) MICH12305 (Smith 20204) MICH12300 (Smith 48732) MICH12301 (Smith 78754) Pileus size 6–14 cm 9–25 cm 4–12 cm 4–12(–15) cm Pileus color Vinaceous, ochraceous tawny to ochraceous buff Deep vinaceous brown to testaceous Orange cinnamon to dingy cinnamon Reddish orange, ferruginous, dull to cinnamon rufous to creambuff Tube mouths (mm) 1–2 1–3 2–3 ~ 1 Stipe size (cm) 2.5–8 × 2–3 9–14 × 3–6 3.5–6 × 1.5–2.5 3–9 × 1–3 Stipe staining Base slowly blue when cut Surface blue to greenish Base surface (1/2–1/3) quickly blue when cut Surface blue and/or green in moist (not dry) specimen Annulus Band-like, fibrillose, pallid to white, not gelatinous Membranous, reddishcinnamon, gelatinous Band-like, dingy pallid, staining brownish, feltytomentose (edges with some gluten) Wide band with buff-orange slimy edge, fragmentary in age Spores (as reported) 8–11 × 3–5 μm 8–10(–12) × 3.8–5 μm 7–9 × 4–4.5 μm 8–11 × 4–4.5 μm Spores (as measured) 7.6–10.2 × 4–4.4 μm 6–8.4 × 4–4.2 μm 7–10 × 4–4.5 μm 7–11 × 4–4.4 μm * Adapted from Smith (1979), Smith & Thiers (1964), and Smith & Trappe 1972. Spore dimensions are reported based on the literature and as measured in this study. stipe and the strong green colored specimens in the field. Table 1 provides an overview of the morphological characteristics of these four taxa. The four taxa are recognized in monographs and western guidebooks (e.g. Arora 1986, Bessette et al. 2000, Smith & Thiers 1964, Thiers 1967, 1975). Suillus caerulescens and S. ponderosus are very common mycorrhizal associates of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), while S. imitatus was described as a spruce associate (Smith & Thiers 1964) and is listed in modern field guides as occurring in mixed conifer forests in the Pacific Northwest (Bessette et al. 2000). In many years, fruit bodies belonging to the caerulescens/ ponderosus complex can be found in great abundance in coastal California forests. As a result, it is not unusual to come across these taxa and to be faced with the challenge of identifying them correctly. In the field, we rely heavily on annulus characters to differentiate between species (Fig. 1), but this is a shortlived, developmental character that can be altered or at least obscured by age Suillus imitatus & the S. caerulescens-ponderosus complex ... 391 Figure 1. Suillus caerulescens (A) and S. ponderosus (B): Comparison of veil and other morphological characters. and weather. Prof. Harry Thiers used to joke that in dry weather all specimens had a dry annulus and were called S. caerulescens and in wet weather the annuli were more gelatinous and the species was called S. ponderosus. In contrast, the name S. imitatus is rarely applied to California material, except to those collections that have a greenish tint and are referred to as S. imitatus var. viridescens (Arora 1986). The original description of S. imitatus var. imitatus compared it primarily with S. lakei (Murrill) A.H. Sm. & Thiers, but field guides usually compare it with S. caerulescens from which it is distinguished with difficulty, by its supposedly slightly shorter spores and slightly more orangish pileus (Thiers 1975, Arora 1986, Bessette et al. 2000). Suillus lakei, another Douglas-fir associate closely related to the caerulescens/ ponderosus complex (Kretzer et al. 1996), is easily distinguished by its reddish, fibrillose pileus. To help sort out the species present in our area, we combined existing GenBank ITS sequences with newly acquired ITS sequences from the holotypes and multiple collections of caerulescens/ponderosus complex. Materials & methods Our field collections were obtained from northern California (Table 2). We obtained distribution information from our field-collected specimens, those deposited in the University and Jepson Herbarium at the University of California, Berkeley (UC), photos deposited in the online database Mushroom Observer that have clear identification, and available checklists for Arizona (Bates 2006), Montana (C. Cripps pers. comm.), and the Intermountain Region (Piep 2003). These sources combined to provided an acceptable account for the overall natural distribution of the species studied here. We extracted DNA from freshly collected specimens (UC 1861375-UC 1861389) using a modified Sigma Extract-N-Amp kit (Sigma Aldrich, St. Louis, MO). For 392 ... Nguyen & al. Table 2. Specimens sequenced in this study. Species Collection date Collection location* Herbarium^ # GenBank # S. caerulescens 3 Dec. 2006 CA: San Francisco Bay area UC 1861383 JQ958308 3 Dec. 2006 CA: San Francisco Bay area UC 1861384 JQ958309 3 Dec. 2006 CA: San Francisco Bay area UC 1861385 JQ958310 3 Dec. 2006 CA: Marin Co., Mill Valley, Mt. Tamalpais UC 1861386 JQ958311 3 Dec. 2006 CA: Marin Co., Mill Valley, Mt. Tamalpais UC 1861387 JQ958312 3 Dec. 2006 CA: Marin Co., Mill Valley, Mt. Tamalpais UC 1861388 JQ958313 3 Dec. 2006 CA: Marin Co., Mill Valley, Mt. Tamalpais UC 1861389 JQ958314 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861378 JQ958315 S. ponderosus 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861375 JQ958320 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861376 JQ958321 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861377 JQ958322 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861379 JQ958323 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861380 JQ958324 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861381 JQ958325 19 Nov. 2006 CA: Mendocino Co., JDSF, Co. Rd 409 & Little Lake Rd., 39.3256 N 123.7372 W UC 1861382 JQ958326 S. caerulescens, holotype 9 Oct. 1954 WA: Pierce Co, Spanaway Park MICH 12297 JQ958316 S. imitatus var. imitatus, holotype 9 Oct. 1954 WA: Pierce Co., Mt. Rainier National Park, Power House Woods MICH 12300 JQ958317 S. imitatus var. viridescens, holotype 2 Oct. 1970 OR: Lincoln Co., near Otis, Van Duzer Corridor MICH 12301 JQ958319 S. ponderosus, holotype 24 Oct. 1944 OR: Clackamas Co., Rhododendron MICH 12305 JQ958318 *Location abbreviations: JDSF = Jackson Demonstration State Forest. ^Herbarium codes: UC = University and Jepson Herbaria, University of California, Berkeley; MICH = University of Michigan Herbarium. Suillus imitatus & the S. caerulescens-ponderosus complex ... 393 Figure. 2. ITS-sequence based maximum likelihood tree of the Suillus species in this study. Maximum likelihood followed by maximum parsimony bootstraps provide statistical support for branches with >50% support. Each sequence contains the isolate/specimen number followed by a GenBank accession number. Suillus spectabilis served as the outgroup. specimens that had been stored in a herbarium, it was necessary to obtain high quality DNA so we used a standard CTAB buffer lysis followed by chloroform extraction and ethanol precipitation. DNA was extracted from the following herbarium type specimens: MICH 12305, MICH 12297, MICH 12300 and MICH 12301. Herbarium abbreviations are according to Thiers (2012; continuously updated). For field/recently collected specimens, we used the standard fungal specific primers ITS1F (Gardes & Bruns 1993) and ITS4 (White et al. 1990). For herbarium specimens with degraded DNA we amplified the ITS gene in two segments when necessary. We used standard ITS and Suillus ITS specific primers (ITS-2S and ITS-3S) in various combinations: ITS1F and ITS2S or ITS2 and ITS3S or ITS3 and ITS4 (Bruns et al. 2010). Sequences were produced using the standard BigDye Terminator v3.1 Kit and ran on an ABI Prism 3700 Genetic Analyzer (Life Technologies). 394 ... Nguyen & al. We examined each automated sequence and manually interpreted and corrected ambiguous bases using Sequencher (Gene Codes Corporation). These sequences were deposited in GenBank under accession numbers (JQ958308-JQ958326). Using BLAST we gathered from GenBank all sequences similar to the species mentioned as of April 24, 2012: S. lakei (GenBank L54086, DQ365643, DQ367917, EF218818, AF166499, GU452528, JF792516, EU557325, JQ393127, JF792515), S. caerulescens (GenBank L54096, EU486453), and S. spectabilis (Peck) Kuntze (GenBank L54104) was used as an outgroup species. We aligned the sequences manually and analyzed the alignment using the maximum likelihood model GTRGAMMA in RAxML (Stamatakis 2006). Maximum likelihood bootstrap was performed using the RAxML online BlackBox interface (Stamatakis et al. 2008). Maximum parsimony bootstrap was performed using PAUP v.4.0b6 with 1000 replicates (Swofford 2001). Corrected distances between sequences were found using the Kimura 2-parameter in PAUP (Swofford 2001). Results & discussion We obtained full length ITS coverage for three of the four type specimens along with the 15 specimens that we collected for this study. We were able to obtain only the ITS1 and ITS2 regions of the S. caerulescens holotype MICH12297. We combined the sequences generated with GenBank sequences of S. lakei, and the outgroup sequence from S. spectabilis. The matrix contained 32 taxa, 613 characters, and 30 parsimony informative characters. The resulting maximum likelihood tree is shown in Fig. 2. To our surprise specimens that we had determined as S. caerulescens and S. ponderosus based on annulus differences (Fig. 1) separated into two clades with good statistical support (Fig. 2). Despite the morphological similarities between these two species, S. caerulescens is sister not to S. ponderosus, but to S. lakei. Together these three species form a clade with moderately high support (85%) based on an analysis of 40 Suillus taxa (data not shown). From these results, we conclude that annular characters, when present, are a reliable way to separate S. caerulescens from S. ponderosus (Table 1, Fig. 2). In contrast, S. imitatus sequences did not fall within their own clade but fit within the clades for the two other species: Suillus imitatus var. imitatus within the S. caerulescens clade, and S. imitatus var. viridescens within the S. ponderosus clade. At least so far, morphologically defined Suillus species separate fairly well by ITS, but we know from other taxa that not all species may be distinguished by differences within the ITS region (Manian et al. 2001, Bruns et al. 2010). This prompts the question of whether there are morphological reasons to suspect that one or both S. imitatus varieties might be distinct at the species level. Interestingly, in the discussion that followed separation of the S. imitatus varieties, Smith & Thiers (1964) and Smith & Trappe (1972) initially thought that S. imitatus var. imitatus might represent S. caerulescens and S. imitatus var. viridescens might be S. ponderosus. They decided instead to recognize two Suillus imitatus & the S. caerulescens-ponderosus complex ... 395 taxa separated by pileus color and spore size differences. The shorter spores for S. imitatus var. imitatus reported by Smith & Thiers (1964) and cited in many later references were given as 7–9 × 4–4.5 μm. Although this overlaps the lower end of the spore range given for S. caerulescens (8–11 × 3–5 μm), it could indicate a subtle distinction. However, we re-measured the spores of the type collection as 7–10 × 4–4.5 μm, which now puts the var. imitatus spores within the center of the range for S. caerulescens, S. ponderosus, and S. imitatus var. viridescens. The orangish pileus color of S. imitatus var. imitatus certainly overlaps the highly plastic pileus colors seen in S. caerulescens and S. ponderosus, and the greenish tint of S. imitatus var. viridescens could easily be environmentally induced, much as the aqua colors seen in the stipe of Leccinum species (den Bakker & Noordeloos 2005). For these reasons we see no compelling argument to maintain either variety of S. imitatus as a separate taxon and so synonymize var. imitatus with S. caerulescens and var. virescens with S. ponderosus. Suillus lithocarpi-sequoiae was described by Singer (1960) from Muir Woods, Marin County, California. The descriptions for this species fit well within our current concept of S. ponderosus. The type specimen (Singer N 1531) was deposited in Fundación Miguel Lillo, Argentina (LIL). Unfortunately, the type specimen could not be located and is assumed lost. Singer placed (1962) and maintained (1975, 1986) this species in Pulveroboletus, even though Smith & Thiers (1964) and Thiers (1975) thought it better placed in Suillus. The species was never encountered again at the type locality, despite repeated search efforts (Thiers 1975). For these reasons we consider the name S. lithocarpi-sequoiae doubtful. Below is an updated taxonomy, and emended diagnoses for S. caerulescens and S. ponderosus.