Genetic structure of Cenococcum geophilum populations in primary successional volcanic deserts on Mount Fuji as revealed by microsatellite markers


  • Bingyun Wu,

    Corresponding author
    1. Department of Plant Science and Resources, College of Bioresource Sciences, Nihon University, Kanagawa, Japan;
    2. Department of Forest Sciences, Graduate School of Agricultural and Life Science, the University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Kazuhide Nara,

    1. Asian Natural Environmental Science Center, the University of Tokyo, Tokyo, Japan;
    Search for more papers by this author
  • Taizo Hogetsu

    1. Department of Forest Sciences, Graduate School of Agricultural and Life Science, the University of Tokyo, Tokyo, Japan
    Search for more papers by this author

Author for correspondence: Bingyun Wu Tel: +81 3-5841-5222 Fax: +81 3-5841-5433 Email:


  • • Polymorphic simple sequence repeat (SSR) markers were used to investigate the genetic structure in a Cenococcum geophilum population associated with Salix reinii in an early successional volcanic desert at Gotenba, on the south-eastern slope of Mount Fuji in Japan, and in three other populations associated with the same host at more developed sites on the mountain, one at Fujinomiya and two at Subashiri.
  • • The genotype richness of Cgeophilum tended to be higher in more developed vegetation patches as well as in more developed sites, suggesting that genotype richness increased with advanced succession because new genotypes might have been introduced into these sites over time.
  • • High genotypic similarity was observed between the Gotenba and Fujinomiya populations but not between the Gotenba and Subashiri populations, suggesting that Cgeophilum genotypes in Gotenba were introduced from the direction of Fujinomiya.
  • • Genotypes in the Gotenba population were clearly distinguishable into two groups. The absence of any intermediate genotype suggests the absence of frequent recombination in this Cgeophilum population associated with early successional vegetation.


Ectomycorrhizal (ECM) fungi play a fundamental role in forest ecosystems by facilitating nutrient and water uptake by their host plants (Smith & Read, 1997). Cenococcum geophilum Fr. [Syn. Cenococcum graniforme (Sow.) Ferd & Winge] is an ECM fungus that is distributed worldwide, from arctic or alpine areas (Gardes & Dahlberg, 1996) to tropical rain forests (Lee et al., 1997), and colonizes a wide variety of host species. Moreover, this fungus is the most common and most readily recognized ectomycorrhizal fungus in the world (Trappe, 1964; LoBuglio et al., 1991; Massicotte et al., 1992; Shinohara et al., 1999), dominating the underground ECM communities of many forest ecosystems (Vogt et al., 1981, 1982). This fungus is more resilient to drought stress than are other ECM fungi (Mexal & Reid, 1973; Coleman et al., 1989) and also protects host plants from drought (Pigott, 1982a,b; Wu et al., 1999).

Cgeophilum does not form detectable sexual or asexual spores, which are important taxonomic criteria for fungi. While it forms abundant sclerotia that serve as dormant propagules during times of environmental stress (Massicotte et al., 1992), its sexual stage has yet to be observed. Worldwide molecular surveys of Cgeophilum isolates conducted to understand the genetic diversity of this common and functional ECM fungus have revealed its significant and unexpectedly high genetic diversity (LoBuglio et al., 1991, 1996; Shinohara et al., 1999; Panaccione et al., 2001; Jany et al., 2002; LoBuglio & Taylor, 2002). Using RAPD, PCR/RFLP of the rDNA internal transcribed spacer (ITS), and other molecular analyses, Jany et al. (2002) observed a high degree of genetic diversity between and within Cgeophilum populations in beech stands and suggested a high rate of mitotic or meiotic recombination. LoBuglio & Taylor (2002) also analyzed polymorphisms at nine loci in two C. geophilum populations by PCR-single strand conformation polymorphism (SSCP) techniques and found that genotypic variation occurred on a fine scale, in some cases even within one soil sample (equivalent to a volume of c. 500 ml). However, the reason for such high genotypic variation in this fungus is not currently understood, given that its sexual stage has never been observed. Analyses of the early developmental process of a Cgeophilum population may help to explain its high genetic variation.

On the south-eastern slope of Mount Fuji, succession is still in the early primary stages in a volcanic desert following the Hoei eruption of 1707; vegetation is recovering patchily, forming islands on the scoria desert. The vegetation islands are usually initiated by Polygonum cuspidatum L. establishment and thereafter are invaded by other herbaceous species and the pioneer woody shrub Salix reinii Franch. et Savat (Lian et al., 2003; Zhou et al., 2003). At this site, Nara et al. (2003b) used terminal restriction fragment length polymorphism (terminal-RFLP) and sequence analyses of the ITS region in nuclear r-DNA to identify 21 ECM fungal species in the underground ECM community associated with S. reinii. The underground ECM community in undeveloped patches (coverage < 0.5 m2) was dominated by mushroom-fungi like Laccaria and Inocybe, and no Cgeophilum was found there. In developed patches (coverage > 45 m2), however, Cgeophilum colonized S. reinii even as a minor constituent. The Cgeophilum population at this site seems to be at an early stage of establishment, and a genetic investigation of this population would be informative toward understanding the genotypic differentiation among Cgeophilum populations.

In this study, we analyzed the genetic structure of Cgeophilum populations in early successional volcanic deserts on Mount Fuji. First, we developed new polymorphic simple sequence repeat (SSR) markers for Cgeophilum, because SSR markers often show higher polymorphism than do other markers and have advantages in population genetics as a result of their codominance (Byrne et al., 1996; Zhou et al., 2001a,b; Lian et al., 2003). We also studied Cgeophilum populations in different areas on the mountain to compare the population structure between early and later developmental stages.

Materials and Methods

Study sites and soil sampling

We collected Cgeophilum samples from four sites in Salix reinii habitats on Mount Fuji: Gotenba, Fujinomiya, Subashiri A, and Subashiri B (Fig. 1).

Figure 1.

Locations of the research sites Gotenba, Fujinomiya, Subashiri A, and Subashiri B on Mount Fuji, Japan.

The Gotenba site was almost the same as the research quadrat (100 × 550 m) described by Zhou et al. (2003) and Nara et al. (2003a). The quadrat was positioned at altitudes of 1500–1600 m above sea level (asl; 35°20′18″ N, 138°47′23″ E) on the south-east slope of Mount Fuji. The vegetation on this slope was completely destroyed by the Hoei Eruption in 1707 and is still recovering. Vegetation is distributed patchily on the scoria substrate, and the total ground coverage is now around 5%. Polygonum cuspidatum Sieb. et Zucc. initially forms small patches by vegetative and sexual reproduction, thus providing subsequent plant species with stable habitats on the unstable scoria desert (Adachi et al., 1996; Zhou et al., 2003). Many pioneer herbaceous and woody species grow in these patches and subsequently enlarge them. Sreinii is the most dominant woody species at this site and the first-colonizing ECM host species (Nara et al., 2003b). Gotenba consists of 161 vegetation patches, including 159 patches described by Nara et al. (2003a,b). Sreinii was distributed in 39 of these patches. From July to August 2002, a total of 311 soil samples (10 × 10 × 10 cm) were collected every 1–2 m from the periphery of the Sreinii colonizing area (cf. Fig. 3) in these 39 vegetation patches.

Figure 3.

Distribution of SSR (simple sequence repeat) genotypes of Cenococcum geophilum in patch 89. Numerals 1–17 indicate the locations of soil sampling. The letters indicate SSR genotypes. Two genotypes were detected from soil sample 13. The dotted areas indicate the distribution of Polygonum cuspidatum, and shaded areas indicate the distribution of Salix reinii.

Fujinomiya was located at an altitude of 2400 m asl (35°20′14″ N, 138°44′2″ E) on the south slope of Mount Fuji, 5 km from Gotenba. Although this site was not affected as severely by the last eruption, vegetation was also patchily distributed owing to its location above the tree line. Many vegetation patches in this site included S. reinii. Other ECM host species, including Larix kaempferi and Alnus maximowiczii, were also found in patches.

Subashiri A and Subashiri B were located at an altitude of c. 2000 m asl (35°21′53″ N, 138°46′45″ E) on the east slope, 5 km from Fujinomiya and 3 km from Gotenba. Both Subashiri A and B were less affected by the last eruption than was Gotenba because they are a greater distance from the Hoei crater. These two sites were separated by a small valley. Although the vegetation in Subashiri A was still distributed patchily on the scoria substrate, the total vegetation coverage was much more extensive than at Gotenba. In addition to S. reinii, we also observed Betula ermanii and L. kaempferi, two tree species that follow S. reinii in the course of succession, in many patches. Such vegetation characteristics indicate that this site is at a later stage of primary succession. In Subashiri B (c. 300 m from Subashiri A), vegetation was fairly developed, and bushes, including Sreinii, were distributed along a forest edge. In early October 2002, 20 soil samples (10 × 10 × 10 cm) were collected randomly under Sreinii in an area (c. 0.5 ha including more than 10 vegetation patches at Fujinomiya and Subashiri A) of each of the three developed sites (Fujinomiya, Subashiri A, and Subashiri B).

Isolation of SSR loci

About 300 mg of cultured mycelial tissue of a Cgeophilum strain isolated from a pine forest was collected after a 3-month period of liquid, shaking incubation in MMN liquid medium (Marx, 1969) at 25°C. The mycelium was then washed in sterilized distilled water, frozen by liquid nitrogen, and crushed into powder in a mortar with a precooled pestle. DNA was extracted from the powdered sample by a modified cetyltrimethylammonium bromide (CTAB) method (Zhou et al., 1999). SSR regions of Cgeophilum were identified by a dual-suppression-PCR technique (Lian & Hogetsu, 2002; Wu et al., 2002).

DNA extraction from field samples of C. geophilum

All roots contained in each soil sample were carefully washed with tap water. Roots of nontarget plants were excluded as the S. reinii roots were clearly distinguishable based on morphology. Five mycorrhizal root tips of Cgeophilum in each soil sample were randomly sampled for DNA extraction. All sclerotia were also sampled from the soil samples. Each sample (single root tip or sclerotium) was pulverized in a 2.0-ml tube containing three zirconia balls using a homogenizer (FastPrep, Funakoshi Co., Tokyo, Japan) for 25 s. Five hundred µl of 2 × CTAB solution [2% CTAB, 0.1 m Tris-HCl (pH 8.0), 20 mm EDTA (pH 8.0) 1.4 m NaCl, and 0.5%β-mercaptoethanol] were added to the tube. Pellets in the tubes were homogenized again for 25 s and incubated in a block heater at 65°C for 1 h. After adding 500 µl chloroform:isoamyl alcohol mixture (24 : 1, v/v), each tube was vortexed and then centrifuged at 19100 g for 7 min; the supernatant was transferred to a 1.5-ml tube. DNA was precipitated from the supernatant by adding an equal volume of isopropyl alcohol and holding the tube at −30°C for 10 min. After centrifugation at 5400 g at 4°C for 10 min, the DNA pellet was washed once with 80% ethanol, dried, and resuspended in 20 µl TE buffer [10 mm Tris-HCl (pH 8.0) and 1 mm EDTA (pH 8.0)], and stored at −30°C until use.

PCR amplification

PCR was performed in a 10-µl reaction mixture containing 1 µl template DNA, 0.4 mm of each dNTP, 2.5 mm MgCl2, 0.5 mm SSR primer (Table 1), 1 µl of 10 × PCR buffer II (Mg2+ free), and 0.25 U of LA Taq DNA polymerase (Takara Shuzo Co., Tokyo, Japan), using a PCR thermal cycler (TP3000, Takara Shuzo Co., Tokyo, Japan). The reverse primer was 5′-labeled with Texas Red (Hitachi Instruments Service Co., Tokyo, Japan). The thermal cycling schedule was as follows: first one cycle at 94°C for 1 min; second 29 cycles at 94°C for 30 s, annealing temperature of the primer pairs shown in Table 1 for 1 min, and 72°C for 1 min; and third one cycle at 94°C for 30 s, annealing temperature for 1 min, and 72°C for 5 min.

Table 1.  Primer pairs for amplification of five SSR (simple sequence repeat) regions in Cenococcum geophilum and some characteristics of the SSR loci
LocusPrimer pair sequence (5′ to 3′)Ta (°C)Repeat motifSize range (bp)Allele numberGenBank Accession no.
  1. Ta indicates annealing temperature. The repeat motif was estimated from cloned sequences. The size range and the allele number were determined based on banding patterns of 35 genets of C. geophilum from four Salix reinii sites on Mount Fuji.

58(TG)11 91–993AB111864

Gel electrophoresis

One portion of the PCR product (5 µl) was mixed with 2 µl of loading buffer (50% glycerin, 1 mm EDTA, and 0.25% xylene cyanol FF; Wako Co. Osaka, Japan) and electrophoresed on a 1.2% agarose II gel including 0.33 ppm of ethidium bromide. Band patterns were visualized on a UV transilluminator (Toyobo, Tokyo, Japan).

Another portion of the PCR product was denatured at 95°C for 5 min and electrophoresed on a sequencing gel [6% Long Ranger (FMC BioProducts Co., ME, USA), 6.1 m urea, and 1.2 × TBE (0.1 m Tris, 3.0 mm EDTA, and 0.1 m boric acid)] in 0.6 × TBE, using a DNA sequencer (SQ-5500, Hitachi Electronics Engineering Co., Tokyo, Japan). The band sizes were estimated from DNA size standards using FRAGLYS 3.0 software (Hitachi Electronics Engineering Co.).

SSR genotypes and similarity between genotypes

ECM root tips of Cgeophilum with the same PCR profile patterns for all primer pairs used were regarded as the same SSR genotype in the following polymorphism analyses.

The similarity between genotypes was evaluated by the equation:

F = 2nAB/(nA + nB)

(Nei, 1987), where nA and nB represent the number of loci in genotypes A and B, respectively, and nAB represents the number of loci at which genotypes A and B have alleles in common. Dendrograms were produced by an unweighted pair-group method using arithmetic averages (upgma) analysis based on the data of similarity between SSR genotypes using the Molecular Evolutionary Genetics Analysis (MEGA, version 2.1) software (Kumar et al., 2001,


SSR markers

We first examined whether the five SSR markers were species-specific to Cgeophilum in PCR amplifications. Using individual primer pairs for these five SSR markers, PCR amplification was tested with the DNA templates from Cgeophilum sclerotia and from sporocarps of eight ECM fungal species (Boletus cf. rubellus Krombh., Cantharellus cibarius Fr., Cortinarius decipiens[pers. Fr.] Fr., Hebeloma leucosarx Orton, Inocybe lacera[Fr.] Kumm, Laccaria laccata[Scop. Fr.] Berk. & Br., Russula norvegica Reid, and Scleroderma bovista Fr.) reported to colonize S. reinii at Gotenba (Nara et al., 2003a). No fragments were amplified by any of the primer pairs from DNA templates of any species except C. geophilum. DNA fragments were amplified from Cgeophilum sclerotial DNA by all five primer pairs. In our study, another black morphotype that was very similar to C. geophilum ECM was also found in some soil samples. Therefore, an internal transcribed spacer (ITS) analysis (Zhou & Hogetsu, 2002) was performed before the microsatellite analysis so as to ascertain the typing of the ectomycorrhizas. The result of ITS analysis showed that it belonged to ascomycetes but not C. geophilum. Moreover, the specificity test also indicated that DNA fragments could not be amplified by any of the primer pairs from DNA templates of this black morphotype of ectomycorrhizas.

Five loci containing polymorphic SSR sequences were identified. Primer pairs were designed from sequences flanking the SSR loci. The characteristics of the five SSR loci (CG-1, CG-2, CG-3, CG-4, and CG-5) are summarized in Table 1.

The five SSR were single locus markers. DNA bands were always single in the electrophoresis of the C. geophilum PCR products from individual mycorrhizal root tips or sclerotia at the five SSR markers.

C. geophilum at Gotenba, the early successional volcanic desert on Mount Fuji

Roots of S. reinii were present in all of the 311 soil samples collected from 39 patches at the Gotenba site, in which 22 samples from eight patches contained C. geophilum ectomycorrhizas. In total, 52 C. geophilum root tips collected from these soil samples were grouped into seven genotypes by five SSR markers (Fig. 2/ Table 2). Four genotypes (A, C, D, and G) occurred only in single soil samples. Each of the other three genotypes (B, E, and F) was distributed over two or more patches. The most dominant genotype, B, was detected in 13 soil samples from five different patches, which were distributed over a maximum distance of 78 m along the slope. High similarity values (F) were found among genotypes A, B and C, and among D, E, F and G (cf. Fig. 4).

Figure 2.

Distribution of seven SSR (simple sequence repeat) genotypes of 22 samples from eight patches at Gotenba. Light grey areas indicate vegetation patches containing Salix reinii. Ectomycorrhizal root tips of Cenococcum geophilum were sampled only from the dark grey patches. Numerals indicate the patch number. Each letter indicates an SSR genotype from one soil sample. Two letters in brackets indicate that two genotypes were detected in one soil sample.

Table 2.  Numbers of root tips containing each SSR (simple sequence repeat) genotype of Cecococcum geophilum in four populations on Mount Fuji
  GotenbaFujinomiyaSubashiri ASubashiri B
  1. Values within parentheses indicate the frequency (%) of root tips of each genotype.

 A 1 (1.9) – –
B24 (46.2) 2 (7.4) –
C 8 (15.4) – –
D 5 (9.6) 2 (7.4) –
E 6 (11.5) 1 (3.7) –
F 4 (7.7) 3 (11.1) –
G 4 (7.7) – –
H – 5 (18.5) –
I – 3 (11.1) –
J – 1 (3.7) –
K – 5 (18.5) 1 (4.2)
L – 2 (7.4) –
M – 3 (11.1) –
N – –1 (6.3) –
GenotypeO – –4 (25.0) –
P – –2 (12.5) –
Q – –5 (31.3) –
R – –1 (6.3) –
S – –1 (6.3) –
T – –1 (6.3) –
U – –1 (6.3) –
V – – 2 (8.3)
W – – 5 (20.8)
X – – 1 (4.2)
Y – – 3 (12.5)
Z – – 2 (8.3)
a – – 4 (16.7)
b – – 1 (4.2)
c – – 1 (4.2)
d – – 4 (16.7)
Genotype number30 710810
Figure 4.

UPGMA dendrograms showing relationships among 30 SSR (simple sequence repeat) genotypes of Cenococcum geophilum on Mount Fuji, Japan. Clusters with > 60% similarity were assigned Arabic numerals as group numbers.

The number of Cgeophilum genotypes detected in a vegetation patch increased with increased patch size. Smaller patches (e.g. patches 101, 111, 157, and 161) were colonized by a single genotype of Cgeophilum, but larger patches (e.g. 89, 90, and 94) were colonized by multiple genotypes. Although two genotypes were detected in the same soil sample collected from patch 89, all other individual soil samples contained only one genotype (Fig. 3).

C. geophilum populations at other sites on Mount Fuji

Sclerotia of Cgeophilum were found in some soil samples. Genotype of every sclerotium was consistent with a root tip genotype found in the same soil sample. However, the number of sclerotia was not involved in Table 2 (i.e. the sclerotia were not used for diversity assessments of sites).

At Fujinomiya, Cgeophilum was detected in nine of 20 soil samples. Six soil samples contained a single genotype of C. geophilum, but multiple genotypes were found in three soil samples. In total, 27 C. geophilum root tips were collected from these soil samples and grouped into 10 genotypes (Table 2).

At Subashiri A, Cgeophilum was detected in five of 20 soil samples collected from seven vegetation patches on the scoria substrate. Sixteen C. geophilum root tips collected from these soil samples were grouped into eight genotypes (Table 2). Five of these genotypes were represented by single root tips. Only one genotype was detected in two soil samples which were collected in the same vegetation patch.

At Subashiri B, Cgeophilum was detected in seven out of 20 soil samples collected from the forest edge. In total, 24 C. geophilum root tips collected from these soil samples were grouped into 10 genotypes (Table 2). Each genotype was detected in only a single soil sample.

Overall, 30 genotypes of Cgeophilum were found among the four sites, including Gotenba (Table 2). Four of 10 genotypes detected at Fujinomiya were common to Gotenba. At Subashiri A and B, all of the identified genotypes were different from those found at Gotenba. Moreover, there was no common genotype between Subashiri A and B.

Genotypic structure of C. geophilum populations on Mount Fuji

Table 2 shows the frequency of each Cgeophilum genotype at the four sites. The genotypic structure of C. geophilum populations apparently changes with vegetation development. At the Gotenba site, genotype B was most abundant, prominently occupying just about half of the analyzed Cgeophilum root tips. However, at Fujinomiya and both the Subashiri sites where vegetation was more developed than Gotenba, all genotypes tended to show rather even abundance.

Dendrograms depicting the relationships among SSR genotypes of Cgeophilum at the four sites were drawn with the upgma algorithm based on the data of similarity values (F) between SSR genotypes (Fig. 4). Genotypes with > 60% similarity were grouped so that the 30 genotypes were classified into eight groups in total. All genotypes at Gotenba were grouped into Groups 1 and 3. In both groups, genotypes at Gotenba were mainly intermingled with those at Fujinomiya. Genotypes found at Fujinomiya were grouped into Group 4 and Group 8 in addition to Groups 1 and 3. Group 8 contained some genotypes at Subashiri A and B. Genotypes at Subashiri A were grouped into Groups 2, 5, 7 and 8. Except Group 2, genotypes at Subashiri A were intermingled with those at Subashiri B. Groups 2, 4 and 6 were composed of genotypes found only at Fujinomiya, Subashiri A and Subashiri B, respectively.


A phylogenetic analysis of C. geophilum by nuclear small subunit ribosomal RNA analysis revealed that this fungus is a lineage of ascomycetes (LoBuglio et al., 1996). As proliferating hyphae of ascomycetes are usually haploid, the C. geophilum mycelium may also consist of haploid cells. In this study, a single band was always amplified from every examined C. geophilum root tip by each of the five markers. This is in accordance with the haploidy in Cgeophilum.

This study provides information on the establishment and development of a C. geophilum population during primary succession. Cgeophilum colonization increased as succession proceeded. Cgeophilum root tips were quite rare in the early successional site and became more common at the other three sites where the vegetation was more developed. While Cgeophilum was detected in only 22 soil samples out of the 311 soil samples at Gotenba (7.1%), it was detected in nine, five, and eight soil samples out of the 20 soil samples at Fujinomiya, Subashiri A, and Subashiri B (45%, 25%, and 40%), respectively. Such low occurrence of this fungus at the early successional site might result in part from the smaller supply of inocula from outside sources.

The genotype richness of Cgeophilum was also lower at Gotenba than at the other three sites. Seven, 10, eight, and 10 genotypes of Cgeophilum were detected at Gotenba, Fujinomiya, Subashiri A, and Subashiri B, respectively; the numbers of Cgeophilum genotypes per Cgeophilum-colonized root tip were 0.13, 0.37, 0.50, and 0.42, respectively. The frequency distribution of genotypes shown in Table 2 indicates that dominance by the most frequent genotype was much more prominent at Gotenba. These results suggest that the genotype richness of Cgeophilum increased as succession proceeded.

Increased genotype richness of Cgeophilum accompanying advanced host development was also observed on a smaller scale. At Gotenba, small patches were generally colonized by a single genotype of Cgeophilum, but big patches were colonized by several genotypes. As patch size roughly represents the time since patch establishment (Zhou et al., 2003), this indicates that Cgeophilum genotype richness increased with patch development. At Gotenba, Nara et al. (2003a,b) demonstrated above-ground and underground ECM fungal succession; a few fungal species first colonize small Sreinii, and fungal species richness increases with host size development. The reason for this is species enrichment by the accumulation of species introduced from the surrounding communities. The genotype richness of Cgeophilum might follow a similar pattern to fungal species richness by a similar mechanism.

Allelic analysis of the SSR markers provides some insight into genetic interrelations between populations. High similarities among some genotypes were found over different populations (Fig. 4), including between Gotenba and Fujinomiya (in Groups 1 and 3), between Fujinomiya and both Subashiri populations (in Group 8), and between the two Subashiri populations (in Groups 5, 7, and 8). The south-eastern slope of Mount Fuji, including the Gotenba site, was newly covered by volcanic scoria produced by the Hoei eruption in 1707. Therefore, the C. geophilum population at Gotenba was initially established by propagules introduced from populations surrounding the scoria desert. The similarity analysis indicates that the genotypes at Gotenba might have been transferred after the eruption from the direction of Fujinomiya, but not from Subashiri. While none of the seven Cgeophilum genotypes at Gotenba was detected at Subashiri A and B, four genotypes (B, D, E, and F) out of the seven were common to Fujinomiya. This fact also supports the above inference.

This fungus is thought to form no meiospores or mitospores, and known propagules for C. geophilum are only sclerotia, that is aggregations of mycelium, that may be dispersed by wind, water, and animals, including insects and birds (Trappe, 1969; Massicotte et al., 1992). Therefore, genetic flow may be mostly caused by the dispersal of sclerotia. A sclerotium of Cgeophilum is relatively large, so its long-distance dispersal could be quite limited compared with that of far smaller and lighter propagules, like the basidiospores of many ECM fungi. Both the lower frequency and the reduced genotype richness of Cgeophilum at Gotenba might largely be owing to this limited propagule supply from the outside, although the long-distance dispersal of Cgeophilum propagules is prerequisite for the initial colonization of the fungus in this volcanic desert that might have occurred only occasionally.

A spreading mechanism of a C. geophilum genotype on a smaller scale can be seen in the Gotenba population that might be at the early colonization stage. Genotype B was associated with many patches and predominantly colonized the host willows, and the maximum distance between two patches that contained genotype B was 78 m. The distribution patterns of genotypes E and F widely overlapped genotype B along the slope, although their occurrence was much less than that of the latter. At this site, many Cgeophilum sclerotia were found in soil samples containing Cgeophilum root tips. Frequent avalanches scoop up scoria at a depth of several tens of centimeters and carry them to lower positions; these avalanches may contribute to the dispersal of sclerotia.

A short and inconspicuous sexual process or parasexuality is suggested to occur in the Cgeophilum life cycle based on its significant genetic diversity (Panaccione et al., 2001; Jany et al., 2002; LoBuglio & Taylor, 2002). However, this does not seem to be the case in this study. Although the genotypes were diverse even at the early stage of primary succession at Gotenba, all of them were highly similar to those at Fujinomiya. This suggests that the genotypic diversity at Gotenba only reflects the diversity of the surrounding populations supplying the propagules and that sexual events such as recombination did not contribute to genotypic diversity. Two discrete genotypic groups (Groups 1 and 3 in Fig. 4) have been maintained at Gotenba, and we were unable to find intermediate genotypes generated by the recombination of these two genotypes. This result also supports the absence of frequent recombination in the Gotenba population. High similarity within each genotype group at the Gotenba population might be an indication of mutation events. Higher genotypic variety was found at three other developed sites surrounding the Gotenba site. In these sites, periods after establishment of the C. geophilum populations should be longer than the Gotenba site. Thus, the higher variety might be caused by recombination or mutation events accumulated during the longer period as suggested in a study of mature forests (Jany et al., 2002). Some other features such as geographic factors might also contribute to the genotypic diversity among the sites. For example, the low similarity of genotypes between the closely situated sites Subashiri A and B might be caused by the separation of a valley.

In conclusion, this study demonstrates that a C. geophilum population under primary succession on Mount Fuji shows a significant genetic diversity, but that the diversity may not be directly caused by genetic recombination.


This study was supported in part by a Grant-in-Aid from the Ministry of Education, Sports, Culture, Science, and Technology of Japan.