Introduction

Ectomycorrhizal (EcM) fungal symbioses involve ca. 7,000 plant species and 7,000–10,000 species of basidiomycetes and ascomycetes (Taylor and Alexander 2005). In both phyla, EcM fungi have repeatedly arisen from saprotrophic or plant parasitic ancestors as indicated in recent phylogenetic analyses (LoBuglio et al. 1996; Hibbett et al. 2000). Such a switch to EcM lifestyle has evolved between 15 and 18 times within basidiomycetes (Binder et al. 2005).

The hymenochaetoid clade of homobasidiomycetes (sensu Binder and Hibbett 2002) comprises taxa with contrasting lifestyles and fruiting habits. Most members of the hymenochaetoid clade form tough, annual or perennial, conk-shaped to resupinate fruit-bodies on wood (Wagner and Fischer 2002a, b; Larsson et al. 2004, 2006; Binder et al. 2005). These wood-inhabiting taxa are predominantly saprobes or weak plant parasites, causing white rot (Hibbett and Donoghue 2001). Some basal members of the hymenochaetoid clade (e.g. Rickenella and Cantharellopsis) form biotrophic associations with mosses and produce tiny agaricoid fruit-bodies. Stipitate, more-or-less terrestrial fruit-bodies occur also in the Coltricia-Coltriciella lineage and in Hymenochaete damicornis (Link) Lév (Wagner and Fischer 2002a). However, fruit-bodies of these taxa, particularly of Coltricia and Coltriciella, are often associated with woody debris (Cunningham 1948; Grand and Vernia 2005), living roots (Corner 1991) or living tree trunks (Aime et al. 2003). Contrary to the observed woody habitat of fruit-bodies, Pachlewski and Chrusciak (1980) reported successful EcM synthesis experiments with Coltricia perennis (L.: Fr.) Murrill and Pinus sylvestris L, but provided no details of the association. Later, Danielson (1984) synthesised EcM and arbutoid mycorrhiza of C. perennis with Pinus banksiana Lamb. and Arctostaphylos uva-ursi (L.) Spreng, respectively. Danielson described the EcM morphology as highly conspicuous but, despite considerable sampling effort in the field, he failed to demonstrate the EcM habit of C. perennis in situ. In addition, fertile basidia formed in pure culture, which raised further doubts as to the true EcM lifestyle of C. perennis (Danielson 1984). Following mycelial connections in situ, Thoen and Ba (1989) identified brown, bristly EcM of Coltricia cinnamomea (Pers.) Murr. on Uapaca guineensis Müll. Arg. in Senegal. The authors also noted the wide distribution of C. cinnamomea in many African vegetation types. In support of the potential EcM links of other members Coltricia and Coltriciella, Aime et al. (2003) observed their fruit bodies exclusively near EcM-forming Dicymbe sp. (Caesalpiniaceae) in Guayanan neotropical forests that are generally dominated by arbuscular mycorrhizal plants. Contrary to wood rotting members of the hymenochaetoid clade and other taxa, C. perennis lacks most of the oxidative decay enzymes (Pachlewski and Chrusciak 1980; Hibbett and Donoghue 2001).

In this study, we use sequence data from the rDNA internal transcribed spacer (ITS) and large subunit (LSU) regions of both EcM root tips and fruit bodies to demonstrate that four species of Coltricia and Coltriciella form EcM with both native and introduced trees in Seychelles. We provide morphological descriptions of EcMs to facilitate the morphology-based identification of these taxa in further studies.

Materials and methods

Sampling

Root samples were collected in the course of a more inclusive EcM community study in Seychelles between 28 February and 15 March 2006 (Tedersoo et al. 2007b). Coltricioid EcM morphotypes were retrieved from three host tree roots at five study sites. Only a single putative EcM host tree was present at each site. Briefly, Anse Major (Mahé Island; 4°37.3′S, 55°23.5′E; 40 m a.s.l.) is a steep coastal shrubland on exposed granite bedrock, dominated by the native Intsia bijuga (Colebr.) Kuntze (Caesalpiniaceae; 20–50 years old; six scattered individuals), Nephrosperma vanhoutteana (Wendl. ex van Houtt) Balf. and thickets of the invasive Chrysobalanus icaco L. Vallée de Mai UNESCO World Heritage Site (Praslin Island; 4°20.0′S, 55°44.0′E; 180 m a.s.l.) comprises a virgin forest of endemic palms with a few clumps of I. bijuga (25–80 years old; 14 individuals) on eroded, yellow-to-red lateritic soils. L’Abondance (Mahé Island; 4°41.3′S, 55°29.1′E; 480 m a.s.l.) is an undisturbed forest dominated by the endemic Vateriopsis seychellarum Heim (Dipterocarpaceae; 20–200 years old; ca. 15 individuals), Cinnamomum verum L. and Pandanus spp. on shallow humus (2–7 cm) overlying coarse quartz sand. Casse Dent (Mahé Island; 4°39.3′S, 55°26.4′E; 410 m a.s.l.) is a plantation of Camellia sinensis (L.) Kuntze with a single planted V. seychellarum tree (40 years old) and multiple current-year seedlings on ploughed soils. Le Niol (Mahé Island; 4°37.5′S, 55°25.9′E; 210 m a.s.l.) comprises a Eucalyptus robusta Sm. (Myrtaceae; 29 years old; ca. 200 individuals) plantation, dominated by C. verum and a thick herb layer of the invasive fern Dicranopteris linearis (Burm.) on strongly eroded red lateritic soils.

Root samples (15 × 15 cm to 5 cm depth) were taken from each site (n = 5–24) between 28 February and 15 March 2006. Within 24 h, EcM roots were separated manually from adhering debris and mineral soil, and cut into fragments of ca. 3-cm. Twenty randomly selected root fragments from each sample were mounted into 10 ml 1% CTAB lysis buffer (100 mM Tris-HCl, pH 8.0; 1.4 M NaCl; 20 mM EDTA; 1% CTAB). Within 6 weeks, EcM root tips were further morphotyped and anatomotyped following Agerer (1991), and photographed using Axiovision 4.0 software (Zeiss, München, Germany). Plan views were recorded at 32–50 x magnification and mantle structure was recorded at 1,000 x magnification with Normarski differential interference contrast. Longitudinal and cross sections (10–20 μm thickness) of EcM root tips were cut using a cryo-microtome and checked for the presence of a Hartig net at 400–1,000 x magnification. The mantle images were edited and sharpened using Adobe Photoshop 5.5 software (Adobe Systems, San Jose, CA).

Molecular typing

Mycorrhizal root tips and pieces from vouchered Coltricia and Coltriciella fruit-bodies (Table 1) were subjected to DNA extraction using a High Pure PCR Template Preparation Kit for Isolation of Nucleic Acids from Mammalian Tissue (Roche, Indianapolis, IN), following the manufacturer’s instructions. PCR was performed using primers ITS1F (5′ cttggtcatttagaggaagtaa 3′) and TW13 (5′ ggtccgtgtttcaagacg 3′). In case of insufficient product, both the ITS and LSU regions were separately amplified using primer pairs ITS1F/ITS4 (5′ tcctccgcttattgatatgc 3′) and ctb6 (5′ gcatatcaataagcggagg 3′)/LR5 (5′ tcctgagggaaacttcg 3′), respectively. Sequencing was performed using primers ITS4, ITS5 (5′ ggaagtaaaagtcgtaacaagg 3′), ctb6 and LR5 or TW13. Contigs were assembled using Sequencher 4.0 (GeneCodes, Ann Arbor, MI). Mycorrhizal root tips with sequences of Coltricia and Coltriciella are presented in this paper. EcM root tip and fruit-body sequences of these two genera were submitted to EMBL (accession nos. AM412240–AM412260; Tables 1, 2).

Table 1 Coltricia and Coltriciella fruit bodies sequenced for this study
Full size table
Table 2 List of ectomycorrhizal (EcM) root tips identified as Coltricia and Coltriciella species
Full size table

Phylogenetic analyses

Additional fruit body sequences of Coltricia, Coltriciella and potential outgroup taxa (based on phylogenetic analyses in Larsson et al. 2006) were downloaded from GenBank (http://www.ncbi.nlm.nih.gov). The sequences were aligned using MAFFT ver. 5.861 (Katoh et al. 2005). The initial alignment comprised 242 taxa, which was finally reduced to 42 taxa. The final alignment included 887 characters of which 248 were parsimony-informative.

Parsimony and Bayesian analyses were performed using PAUP* ver. 4.0d81 (Swofford 2002) and Mr. Bayes ver. 3.1.1 (Ronquist and Huelsenbeck 2003), respectively. Parsimony analyses were run with tree-bisection-reconnection (TBR) branch swapping, gaps as missing data, maxtrees set to 10,000 and up to 100 saved trees per replicate. Bootstrap support was calculated based on 1,000 bootstrap replicates. For Bayesian analyses, model selection was performed using a program Mr. Modeltest ver. 2.2 (Nylander 2004). Bayesian analyses were performed with 2,000,000 generations and a burn-in value of 2,500.

Results

Phylogenetic placement

Amplification of both root tips and fruit bodies of Coltricia and Coltriciella was difficult using any of the primers — only 6 out of 13 (46%) extracted EcM root tips produced sequencable material. Sequencing of mycorrhizal root tip DNA revealed four species of coltricioid fungi that possessed the ITS region between 886 and 932 bp, including 447–488 bp of ITS1. ITS sequences were highly divergent and poorly alignable among closely related species. All species of Coltricia and Coltriciella shared two hypervariable insertions in the LSU that were missing in all other members of the hymenochaetoid clade. With or without inclusion of these regions, Coltricia and Coltriciella formed a 100% supported lineage (Fig. 1). Phylogenetic analysis placed Coltriciella nested within the genus Coltricia, but with moderate bootstrap support (72%). Common taxa with large ranges (C. cinnamomea and C. dependens) were strongly polyphyletic or paraphyletic, suggesting the presence of several cryptic species. Sequence types L2367 and L2528 had ITS and LSU sequences identical with two Seychelles specimens of Coltriciella dependens. Phylogenetic analyses placed these and an additional sequence type, L2480, within Coltriciella spp. (Fig. 1). L2410 had ITS and LSU sequences identical with Coltricia aff. oblectans (AM412245). L2368 and L2450 had identical ITS sequences, but matched no fruit body sequences. Based on LSU sequences, these isolates and L2410 clustered with several species of Coltricia, particularly C. cinnamomea, C. aff. oblectans and C. montagnei.

Fig. 1
figure 1

Bayesian tree demonstrating the phylogenetic placement of six root tip isolates (bold) among Coltricia and Coltriciella species based on partial large subunit (LSU) sequences. Posterior probabilities (≥95%) and bootstrap support (≥70%) are shown above and below the branches, respectively

Full size image

Ectomycorrhiza descriptions

Coltricioid EcM occurred as mycorrhizal root tips or in small clusters of up to six mycorrhizal root tips among other morphotypes. EcM of Coltricia and Coltriciella spp. were highly similar in gross morphology, sharing a thin (3–5 hyphal layers), orange-brown to dark brown mantle with abundant dark brown, thick-walled emanating hyphae and/or cystidia (Fig. 2a, f, k,). None of the species possessed clamp connections. Hartig net was present (Fig. 2e), but sometimes poorly developed.

Fig. 2
figure 2

Morphological and anatomical characters of Coltricia and Coltriciella ectomycorrhiza (EcM). a–e Coltriciella dependens (L2367 and L2528) on Vateriopsis seychellarum; a plan view; b outer mantle layer; c inner mantle layer; d emanating hyphae; e Hartig net (cross section); f–j Coltricia aff. oblectans (L2410) on Intsia bijuga; f plan view; g outermost mantle layer with abundant cystidia; h inner mantle layer; i emanating hyphae; j cystidia; k–o Coltriciella sp. (L2480) on V. seychellarum; k plan view; l outer mantle layer; m middle mantle layer; n inner mantle layer; o cystidia; p–t Coltricia sp. (L2368 and L2450) on I. bijuga; p plan view; q outer mantle layer; r inner mantle layer; s, t emanating hyphae. Bars a, f, k, p 1 mm; b–e; g–j; l–o; q–t 10 μm

Full size image

Coltriciella dependens (L2367 and L2528; Fig. 2a–e)

Unramified ends straight or slightly bent, orange-brown to dark brown, 0.4–1.7 mm × 0.18–0.23 mm (length × diam.). Outer and middle mantle layers transitional between plectenchymatous and pseudoparenchymatous, with a unique labyrinthine pattern. Outer layer cells extremely thick-walled [(0.8) 1.2–1.7 (2.1) μm], 2–7 μm diam. Middle layer cells thick-walled (0.3–0.8 μm), 14–26 (34) × 3.5–6 μm. Inner cells oblong, arranged in a rosette-like pattern, moderately thick-walled [0.1–0.5 (0.7) μm], 15–35 (43) × 3.6–6 (10) μm. Emanating hyphae moderately abundant, dark brown, stellately branching, thick-walled (0.4–0.7 μm), without warts, 2.5–4 (5) μm diam. Cystidia abundant, dark brown, with a basal cell, thick-walled at the base (0.4–0.9 μm), long (60–160 μm) and narrow (2–3 μm diam.).

Coltricia aff. oblectans (L2410; Fig. 2f–j)

Unramified ends straight or slightly bent, orange-brown to dark brown, 0.4–3.6 × 0.25–0.4 mm. Outer and inner mantle layers transitional between pseudoparenchymatous and plectenchymatous. Outer layer cells oblong, thick-walled (0.2–0.7 μm), arranged in a stretched mosaic-like pattern, 5–7 μm diam. Inner layer cells thin-walled (<0.2 μm), 3.5–6 μm diam. Emanating hyphae abundant, dark brown, straight, moderately thick-walled (0.3–0.5 μm), without warts, 5–11 μm diam. Cystidia abundant, dark brown, covering the whole outer mantle layer, always unicellular, stout, cylindrical, moderately thick-walled (0.3–0.4 μm), 30–120 × 7–12 μm.

Coltriciella sp. (L2480; Fig. 2k–o)

Unramified ends straight or slightly bent, orange-brown to dark brown, 0.3–11 × 0.18–0.22 mm. Outer and inner mantle layers transitional between pseudoparenchymatous and plectenchymatous. Outer layer cells oblong, irregular, forming a net-like pattern, thick-walled (0.3–0.6 μm), 3–6 μm diam. Inner layer cells more elongated, forming a mosaic-like pattern, sometimes arranged in a rosette, moderately thick-walled [0.1–0.3 (0.8) μm], 2.5–6 μm diam. Emanating hyphae abundant, dark brown, thick-walled (0.3–0.6 μm), without warts, 4–6 μm diam. Cystidia scarce, dark brown, with a basal cell (4–5 μm diam.), multicellular, slender, moderately thick-walled (0.3–0.4 μm), 60–100 × 2.5–3.5 μm.

Coltricia sp. (L2368 and L2450; Fig. 2p–t)

Unramified ends straight, club-like, red-brown to dark brown, 0.3–1.1 × 0.19–0.27 mm. Outer and middle mantle layers distinctly pseudoparenchymatous. Cells rectangular to oval, extremely thick-walled [0.5–1 (1.3) μm], 15–25 (31) × 8–19 μm. Inner mantle layer intermediate between plectenchymatous and pseudoparenchymatous. Inner layer cells relatively thin-walled (<0.2 μm), 3.5–6 μm diam. Emanating hyphae moderately abundant, dark brown, covered with abundant warts, stellately branching, thick-walled (0.4–0.7 μm), 5–10 μm diam. Cystidia absent.

Discussion

Phylogeny and ecology

We provided morphological and molecular evidence for in situ EcM habit of Coltricia and Coltriciella spp. within the predominantly saprotrophic hymenochaetoid clade. Thus, our results complement earlier reports on EcM formation between C. perennis and Pinus spp. in axenic synthesis experiments (Pachlewski and Chrusciak 1980; Danielson 1984). In addition, C. perennis possessed ratios of stable 13C and 15N isotope concentrations similar to that of most EcM basidiomycetes and clearly different from saprobes (Tedersoo et al. 2007a). These results suggest that Coltricia and Coltriciella spp. may function as typical EcM fungi.

Based on nuclear LSU sequence data and Bayesian analyses, Hyphodontia-Schizopora lineage is a sister taxon to Coltricia-Coltriciella lineage (Larsson et al. 2006). Most members of the Hyphodontia-Schizopora lineage form resupinate fruit bodies on wood. Indeed, the majority of basidiomycetous EcM lineages have resupinate fruiting saprotrophic ancestors (Larsson et al. 2004), suggesting that habitat in close contact to soil and roots in dead wood provides preconditions for the development of EcM lifestyle in general.

The genera Coltricia and Coltriciella are distributed world-wide, especially in tropical ecosystems (Ryvarden and Johansen 1980; Corner 1991). The panglobal distribution suggests that Coltricia-Coltriciella is an ancient taxon and members of this lineage likely form EcM associations in all continents. So far, their presence in below-ground EcM communities may have been simply overlooked, for two reasons. First, coltricioid taxa form relatively small and thin EcMs that are low in abundance on a fine scale in Seychelles and Tasmania (Tedersoo 2007; Tedersoo et al. 2007b). Second, high mutation rates in the ITS and LSU regions and long ITS sequences may contribute to low success in DNA amplification and sequencing, especially from stored specimens (this study).

Despite additional evidence for EcM habit, the autecology of Coltricia and Coltriciella spp. remains poorly understood. Unlike saprotrophic and most EcM fungi, C. perennis lacks laccase and tyrosinase activities in spot tests with pure cultures, suggesting that at least the studied isolate is not adapted to decomposition (Pachlewski and Chrusciak 1980). This lends further evidence to the view that fruiting habit on decomposed wood does not necessarily indicate saprotrophic lifestyle. Similarly, wood-inhabiting Lactarius sect. Panuoidei were originally considered saprotrophic or root parasitic (Singer 1984). However, these taxa were later proven to be EcM by molecular analyses of associated EcM root tips (Henkel et al. 2000). Also, the ubiquitous ‘inconspicuous, wood-inhabiting’ resupinate thelephoroid, athelioid and sebacinoid fungi were long considered saprotrophic. But why do some EcM fungi fruit preferentially on dead wood or the butts of living trees? We hypothesise that tree trunks provide a more exposed place for releasing spores (Henkel et al. 2006) or a safe site against flooding. Contrary to Lactarius spp. and similar to resupinate fungi, fruit bodies of most Coltricia spp. are long-lived and may use dead wood as a relatively stable substrate for physically supporting fruit bodies. In addition, and perhaps more importantly, dead wood provides a substantial moisture reserve during short dry periods (Harvey et al. 1978, 1981). Thus, the mycelium of coltricioid taxa could potentially provide water from dead wood to fruit bodies shortly after the rainy season. Unlike the elevated abundance of resupinate thelephoroid, athelioid and sebacinoid taxa in dead wood in boreal forests (Tedersoo et al. 2003), we found no evidence for substrate preference of coltricioid or any other taxa in Seychelles (Tedersoo et al. 2007b).

Ectomycorrhiza morphology and anatomy

The morphological and anatomical characters of Coltricia and Coltriciella EcM presented here resemble brief descriptions of C. perennis and C. cinnamomea EcM (Danielson 1984; Thoen and Ba 1989). Thus, Coltricia and Coltriciella spp. studied so far share orange-brown to dark brown EcM mantle with moderately to extremely thick cell walls. Clampless emanating hyphae and/or cystidia (‘bristles’) are usually abundant, forming a dense brown weft around the mycorrhizal root tips. Both outer and inner mantle layers are transitional between plectenchyma and pseudoparenchyma, except in Coltricia sp. (L2368 and L2450), which has distinctly pseudoparenchymatous outer layer cells. The EcM of coltricioid taxa most strongly resemble members of the Genea-Humaria, Wilcoxina-Trichophaea hybrida and Geopora lineages of Pezizales with respect to plan morphology and mantle structure. Wilcoxina and Geopora spp. form mantles intermediate between pseusoparenchymatous and plectenchymatous, but emanating hyphae are scarce and only 2–4 layers of mantle cells are present (Tedersoo et al. 2006). Coltricia sp. (L2368 and L2450) was nearly identical to the EcM of Humaria hemisphaerica and Genea spp, differing only by the lack of anastomoses in the outer mantle cells and somewhat thinner cell walls (Brand 1991; Tedersoo et al. 2006). The EcM of coltricioid taxa also resemble certain Tomentella species that differ by a thicker mantle with more differentiated layers (Agerer 2006; L.T. personal observation). In addition, most Tomentella spp. possess clamp connections in emanating hyphae and moderately thick cell walls in the outer mantle layer.

In conclusion, accumulating evidence suggests that several species of Coltricia and Coltriciella form anatomically conspicuous EcM with several host plants. Further experimental studies are needed to establish the ecophysiological traits of individual species and their position in the mutualism-parasitism continuum (Egger and Hibbett 2004).