- Top of page
- Materials and Methods
The majority of recent studies on the ecology of basidiomycetes are based on population structure analysis. Using the phenomenon of somatic incompatibility (Guillaumin et al., 1996; Worrall, 1997; Malik & Vilgalys, 1999), it is possible to identify genetically distinct secondary mycelia of the same species coexisting within a population. These mycelia are called genets and can be considered the units of a population. Knowledge about the size and spatial distribution of genets can provide important information on the processes of establishment and growth of a particular fungal population (Kirby et al., 1990; Worrall, 1994). The population structure of several basidiomycetes has been studied, including that of Heterobasidion annosum (e.g. Swedjemark & Stenlid, 1993), Phellinus weirii (e.g. Hansen & Goheen, 2000), Fomitopsis pinicola (Högberg et al., 1999), Resinicium bicolor (Kirby et al., 1990), Marasmius androsaceus (Holmer & Stenlid, 1991), Suillus bovinus (e.g. Dahlberg & Stenlid, 1994), and Armillaria spp. (e.g. Worrall, 1994).
The genus Armillaria comprises several species which are important components of the mycoflora in many forest ecosystems worldwide (Shaw & Kile, 1991). Armillaria spp. can behave as primary or secondary pathogens causing root and butt rot on numerous coniferous and broadleaved tree species in both naturally regenerated forests and plantations (Guillaumin et al., 1993; Morrison et al., 2000). As parasites, Armillaria spp. can cause significant economic losses (Morrison & Mallett, 1996) and influence the tree species composition of forests (Kile et al., 1991). In addition, all Armillaria species can survive saprotrophically in woody substrates such as roots, stumps and debris. To date, about 40 species of Armillaria are known, seven of which occur in Europe (Watling et al., 1991). The European species differ in geographical distribution, ecological behaviour, host range, and pathogenicity (Guillaumin et al., 1993).
Different species of Armillaria frequently coexist in the same forest stand (Rizzo & Harrington, 1993; Legrand et al., 1996; Bruhn et al., 2000; Marxmüller & Holdenrieder, 2000). In central Europe, the pathogenic Armillaria ostoyae (Romagnesi) Herink and the preferentially saprotrophic Armillaria cepistipes Velenovsky often occur sympatrically in mountainous forests (Legrand et al., 1996; Marxmüller & Holdenrieder, 2000). Armillaria species are probably significant competitors of the root and butt rot pathogen Heterobasidion annosum (Holdenrieder & Greig, 1998) and they also compete with each other for resources (S. Prospero et al. unpublished). Therefore, potential agents for the biological control of H. annosum and parasitic Armillaria species may be primarily saprotrophic Armillaria species and their interaction with these fungi deserves attention.
The purpose of the present study was to analyse the population structure of A. cepistipes and A. ostoyae coexisting in three managed Norway spruce (Picea abies (L.) Karst.) stands in the Alps. Specifically, we addressed the following questions: (i) What is the spatial distribution of the genets of the two species?; (ii) Are there overlaps between genets of the same or of different species?; and (iii) What are the spatial patterns of genets along interspecific boundaries? Considering these aspects, inferences are made about intra- and interspecific interactions.
- Top of page
- Materials and Methods
We conducted an intensive sampling in the soil and stumps in three comparable spruce stands to analyse the population structure of coexisting Armillaria species. A. cepistipes and A. ostoyae were shown to be the dominant species in all investigated spruce stands.
Based on somatic incompatibility, two to six genets of each Armillaria species were detected in each 1-ha plot. The resulting densities of seven to nine Armillaria genets per hectare are similar to those observed by Worrall (1994) in North America in plots with only A. ostoyae. In France, Legrand et al. (1996) found two study sites with A. cepistipes, A. ostoyae, and A. gallica, where the density was either five or six Armillaria genets per hectare. Most other species of wood-inhabiting basidiomycetes are generally characterised by more dense populations than those of Armillaria. For example, Kirby et al. (1990) found 19 genets of the wood-decay fungus Resinicium bicolor in a spruce stand of 1250 m2 (152 genets per ha). Populations of the pathogenic H. annosum can be composed of up to 4800 genets per hectare (Piri et al., 1990; Swedjemark & Stenlid, 1993). According to Hansen & Hamelin (1999), the density of a basidiomycete population, measured as the number of genets in a stand, is affected by substrate availability, frequency of opportunities for establishment, external disturbances (e.g. forest management), and intra- and interspecific competition.
The number of genets that can be detected probably depends on the intensity of sampling. The intensive main sampling conducted in our study has provided a representative view of the density of large and medium-sized Armillaria genets. However, the detection of additional very small genets in two subplots suggests that the actual number and density of genets is higher than estimated.
All Armillaria genets found on the stumps were also found in the soil. Thus, a systematic soil sampling seems to be sufficient to determine the occurrence and minimum density of Armillaria species and genets. Nevertheless, stump investigation can provide supplementary detailed information on the sizes and spatial distribution of genets. Delimiting the exact physical extent of genets, which is known to be difficult in studies of the population structure of soil-borne basidiomycetes (Anderson & Kohn, 1995), was not a primary objective of this study. Nevertheless, the estimated sizes of the larger genets (0.3–0.4 ha) are comparable with those reported for A. cepistipes and A. ostoyae in coniferous and hardwood forests in central Europe (Legrand et al., 1996; Marxmüller & Holdenrieder, 2000). Similar sizes of A. ostoyae genets were also observed by Smith et al. (1994) in a red pine (Pinus resinosa Aiton) seedling plantation in North America. The larger genets occurring in our plots probably further extend into the surrounding stands. Therefore, sampling on a larger scale would be necessary to determine the effective size of these genets. Some genets could also be smaller than estimated. Sampling in the five subplots demonstrated that, on a small scale, considerable gaps can exist within the territory occupied by Armillaria. These observations indicate that rhizomorphs of a genet are not evenly distributed in the soil, but vary considerably in density.
Several of the identified genets showed a noncontiguous, patchy distribution. The single patches can be considered as the physically and physiologically independent units of the genets, called ramets (Dahlberg & Stenlid, 1994; Dettman & van der Kamp, 2001). Kile (1983) suggested that the probability of fragmenting increases with the size and age of a genet. The scattered genets observed in our plots could therefore represent remnants of larger genets. They could also be the result of an irregular vegetative growth from an initial point (e.g. colonised stump) through the production of rhizomorphs or spread via root contacts. Another possible explanation is that noncontiguous genets are inbred siblings (Kile, 1983), which are not always distinguished by somatic incompatibility tests (Guillaumin et al., 1991). The highly scattered A. ostoyae genet Da C3 in Dalpe could have been created by past management operations. By dragging stems downhill it is possible that rhizomorphs or colonised woody debris were displaced and became established in another sector of the stand. Fragmented genets have also been observed in other rhizomorph- and cord-producing fungi, such as Marasmius androsaceus (Holmer & Stenlid, 1991), Tricholomopsis platyphylla (Thompson & Rayner, 1982), as well as in the ectomycorrhizal fungus Suillus bovinus (Dahlberg & Stenlid, 1990). Fragmentation could be beneficial for the survival of a genet by impeding the spread of deleterious cytoplasmic elements. Fragmented genets could be the result of a famine-induced thinning of the rhizomorph network or mycelium because of insufficient disposability of resources (Smith et al., 1994; Anderson & Kohn, 1995).
The assessment of genet fragmentation is strongly affected by the specific criteria employed to identify noncontiguity between neighbouring isolates. In our study, two isolates of the same genet, distant ≤ 20 m, were considered to represent a contiguous genet. Decreasing this arbitrary distance would certainly increase the number of fragmented genets identified in our three plots. The quantitative estimates of spatial overlaps between genets also depend on the criteria adopted to define the limits of genets.
In Ludiano, no spatial overlaps between A. cepistipes and A. ostoyae were observed. The two species occupied different sectors of the plot and there is probably little interspecific interaction. By contrast, a genet of A. cepistipes partially overlapped with the two genets of A. gallica. The detection of A. gallica in Ludiano was unexpected, because this species typically occurs in hardwood forests at low altitudes (Guillaumin et al., 1993). Both A. cepistipes and A. gallica are preferentially saprotrophs and spatial overlaps among genets of the two species have been rarely observed (Legrand et al., 1996). In Lurengo, the two small genets of A. cepistipes were located within the territory occupied by an A. ostoyae genet, which could suggest a recent establishment. Whether these two A. cepistipes genets will gain size in the future at the cost of the A. ostoyae genet needs to be determined.
The third plot (Dalpe) was characterised by large overlaps between A. cepistipes and A. ostoyae. About 38% of the total territory covered by Armillaria was occupied by both species. Dalpe was also the plot with the highest incidence of fragmented and irregularly distributed genets. Therefore estimates of overlapping areas in this plot will strongly depend on the criteria adopted to define the boundaries and contiguity of a genet. Detailed sampling in the subplots, however, showed that spatial overlaps along interspecific boundaries are common in all three plots. Considerable overlaps between preferentially saprotrophic and pathogenic Armillaria species have been observed in other studies and have been assumed to indicate different colonisation strategies and resource partitioning between the species (Rizzo & Harrington, 1993; Smith et al., 1994; Legrand et al., 1996; Baumgartner & Rizzo, 2001). In our plots, this hypothesis is supported by the finding that A. ostoyae is more efficient in primary stump capture than A. cepistipes (S. Prospero et al., unpublished).
Our study shows that, in comparable Norway spruce stands where A. cepistipes and A. ostoyae coexist, the incidence of each species, as well as the spatial patterns and sizes of genets, can vary considerably. The factors determining this variation are not known. It can be assumed that site characteristics (e.g. exposition, soil, and topographical heterogeneity) and stand characteristics (e.g. history, management practise, and species composition) could influence the population dynamics of Armillaria. A general type of interaction between A. cepistipes and A. ostoyae is difficult to deduce from our results. On parts of the investigated areas, both species seem to exclude each other indicating competition, probably because of a similar ecological strategy. However, on other parts (e.g. in Dalpe) there were considerable interspecific overlaps, which could be explained by a more neutralistic coexistence with both species ignoring each other. Our study only gives a snapshot of a dynamic system. Further studies, focusing on the dynamic and experimental manipulation of local Armillaria populations are needed to examine the degree of competition between A. cepistipes and A. ostoyae.