Draft Copy – Please do not distribute
I. Introduction
Habitat size and isolation are central to modern community ecology
Habitat size and isolation play a key role in our understanding of the processes that generate and maintain diversity in biological communities. The predictable rise in species richness with increasing area is among the earliest quantitative patterns in ecology (Watson 1859, Arrhenius 1929), and dispersal limitation played a central role in Gleason’s (1929) thinking about non-equilibrium (i.e. individualistic) plant communities. The effects of habitat size and isolation were extended and formalized in theories such as island biogeography (MacArthur and Wilson 1967), metapopulations (Levins 1969, Gilpin and Hanski 1991), the spatial competition hypothesis of diversity (Tilman 1994), and the neutral theory of biodiversity and biogeography (Hubbell 2001). While niche partitioning certainly plays some role in driving species area relations (Rosenzweig 1995), most modern theory is based on the principles of population ecology. Larger habitats are likely to contain large populations that are less vulnerable to stochastic extinction events. When extinctions do occur, less isolated habitats are more likely to be recolonized. Because many species differ in their dispersal abilities (particularly sessile organisms), it is often possible to predict which organisms are most vulnerable to extinction from habitat loss (Tilman et al.1994).
While these theories about habitat size and isolation have been successfully applied to macroorganisms (e.g. Simberloff and Wilson 1969, Rosenzweig 1995, Wardle et al. 2003 and many others), there is currently strong debate about whether the diversity patterns of microbial organisms are governed by fundamentally different processes (e.g. Fierer and Jackson 2006, Bell et al. 2005, Fenchel and Finlay 2005, Horner-Devine et al. 2004, Green et al. 2004, Finlay 2002 and many others). The often cited Baas-Becking (1934) hypothesis that, “everything is everywhere, but, the environment selects” (as translated in de Wit and Bouvier 2006) has been called into question as DNA based molecular tools provide increasing taxonomic resolution for microbial organisms. Recent molecular based studies have shown significant spatial turnover of microbial organisms (Green et al. 2004, Horner-Devine et al. 2004), but the magnitude of this effect is still low enough to support arguments that microbes diversity patterns are fundamentally different from macroorganisms (but see Bell et al. 2005). Further, skeptics argue that when environmental factors are taken into account spatial turnover has no significant effect on microbial diversity patterns (Fierer and Jackson 2006).
Why study the spatial scaling of ectomycorrhizal communities?
Ectomycorrhizal fungi are a major component of the soil microflora (Wardle 2002), contributing to the biodiversity, ecology and nutrient cycling of forest ecosystems. These fungi form an obligate symbiosis with the roots of many dominant trees species in boreal, temperate and tropical forests. In this symbiosis, the fungi provide nutrients to their hosts in exchange for photosynthetically derived carbon (Smith and Read 1997). The importance of the relationship is illustrated by the substantial proportion of host plant nitrogen (60-86%) derived from ectomycorrhizal fungi, and the large share of net primary productivity (15-20%) that is allocated by ectomycorrhizal plants to their fungi (Vogt et al. 1982, Smith and Read 1997, Hobbie and Hobbie 2006). While ectomycorrhizal fungi cannot grow in nature without access to a host, it is less well known that the ecological success of many plants has been shown to be dependent on the availability of ectomycorrhizal partners (Terwilliger and Pastor 1999, Horton et al. 1999, Nara and Hogetsu 2004, Nara 2005).
Despite their ecological importance, study of ectomycorrhizal ecology has lagged behind that of plants and animals, primarily because they are hard to observe, diverse, and produce diagnostic morphological structures infrequently. This knowledge gap has narrowed in recent years due to the popularization of molecular genetic tools and DNA sequence databases for fungal identification (Horton and Bruns 2001). Surprisingly, despite major differences in growth form and nutrition, studies on the effects of abiotic factors, competition, and succession, suggest that the ecology of ectomycorrhizal fungi may be very similar to plants ( ). However, while habitat size requirements and dispersal limitation are fundamental components of plant and animal ecology, very little is known about the spatial scale of population processes or dispersal by ectomycorrhizal fungi. The few published studies which could shed light on these processes in communities of soil fungi have provided mixed evidence. Nara et al. (2003) found that ectomycorrhizal sporocarp richness was higher in larger patches of willow, but patch size in this study was directly correlated with patch age. Mangan et al. (2004) found that forest fragmentation did not reduce richness of arbuscular mycorrhizal fungi but did affect community composition, while Green et al. (2004) found that ascomycetous soil fungi exhibited a weak, but significant, increase in richness with area. From these studies it is unclear whether the spatial scaling in fungal communities will be different from that of plants or animals.
The study will investigate the effects of habitat size and isolation on ectomycorrhizal communities and help close the knowledge gap for this group of organisms. The spatial scaling of communities is governed by two major factors: (1) the minimum habitat size necessary to maintain an individual or population, and (2) the ease with which organisms disperse between habitats.
III. The spatial scale of aerial spore dispersal by ectomycorrhizal fungi (H5, H6)
Hypothesis 5: Abundance of aerial ectomycorrhizal spores in the coastal scrub will decrease significantly with distance from P. muricata forest.
Hypothesis 6: Aerial spore abundances will differ between species, and will be greater for early stage (e.g. Suillus, Thelephora spp.) versus late stage (e.g. Russula, Inocybe spp.)
Background: The Baas-Becking (1934) hypothesis of microbial diversity states that “everything is everywhere, but, the environment selects” (as translated in de Wit and Bouvier 2006). While some recent studies suggest that microbes exhibit spatial turnover consistent with dispersal limitation (e.g. Whitaker et al. 2004, Horner-Devine 2004, Hughes Martiny et al. 2006) , unambiguous support is lacking and highly contested (e.g. Fenchel and Finlay 2004, Fierer and Jackson 2006). If the Baas-Becking hypothesis is true, the negative ecological effects of isolation would be irrelevant to the conservation of microbial diversity, and by extension to microbially mediated ecosystem processes. Relatively little is known about the dispersal of EM fungi. However, what can be inferred from population genetic and phylogenetic work suggests that, for at least some taxa, there are barriers to dispersal at landscape and continental scales (Bergemann and Miller 2002, Kretzer et al. 2004). However, rigorous studies on the spatial scale of dispersal limitation for ectomycorrhizal fungi are lacking. Preliminary results from Experiment 1 of this study suggest that early stage taxa (that we believe are good dispersers) dominate on small tree islands, while taxa that we believe are good competitors (but poor dispersers) only appear on large islands. A tradeoff between competition and dispersal has been hypothesized as a major force structuring the communities of other sessile organisms, such as plants (Tilman 1994). Tilman et al. (1994) also proved analytically that when this tradeoff exists, good dispersers will dominate on small, fragmented habitats at the expense of good competitors. This seems to fit the preliminary pattern from Experiment 1. However, ectomycorrhizal fungi are also known to respond differentially to host carbon availability, nutrient source and abundance, soil moisture, pH, soil type, soil horizon and disturbance (Erland and Taylor 2002) and the observed pattern could also be consistent with changes in niche types as island size increases. To test between these two alternatives and determine the spatial scale of ectomycorrhizal dispersal limitation we have designed an experiment to sample ectomycorrhizal spore loads at a range of distances from the tree islands that serve as sources of inoculum.
Experiment
To test this question we will need to sample and quantify the spores of EM fungi at varying distances from spore sources. Spores can be sampled directly in the field using a pump that draws air into a chamber equipped with a greased slide. Spores drawn into the chamber are embedded on the slide and can be brought back to the lab for analysis. In October 2006 we plan to establish an evenly spaced sampling matrix of 24 locations, covering the area of the study site from Question 1. We will visit each site once per month, for the entire mushroom fruiting season (Oct to May). The distance of each sampling site from existing P. muricata patches will be calculated using ArcGIS 8.1. In addition, if preliminary species occurrence patterns observed from Experiment 1 continue to hold, it is reasonable to assume that certain species will only be found on islands above a certain size. For these late stage species (e.g. Russula, Inocybe) we will also calculate species specific distances to spore sources.
Spore identification & quantification: Spores captured on the slides will be identified to taxonomic group using spore morphology. This can be done quite efficiently for some important groups, such as the Russulales, Boletales, and Thelephorales (see Figure 5). At our study sites these groups are represented by only a few taxa, and within each group share similar ecologies and island establishment patterns. Because the spore traps sample a known volume of air, it will be possible to calculate the number of spores per liter of air for these groups. For each sample we will use t-RFLP to determine the presence or absence of individual species. In addition, q-rtPCR will be used to quantify the DNA present from four target species (Suillus pungens, Thelephora terrestris, Laccaria amethysteo-occidentalis, Russula amenoelens) with differing distribution patterns. We will collect spores from fruit bodies of each of these species at Point Reyes and use a haemocytometer to create spore dilutions. DNA extracted from the spore dilutions will be used to construct standard curves for use in q-rtPCR. Using these spore dilutions as standards for q-rtPCR we will be able to estimate the original number of spores present in our samples (e.g. Schweigkofler et al. 2005).
Data analysis & Expected Results: To determine the effects of distance on dispersal, spore quantities for each taxonomic group will be analyzed using a generalized linear model including sampling date and distance (either to nearest P. muricata patch or to the nearest possible source islands). If the relationship between distance and spore quantity is negative, this will suggest that ectomycorrhizal fungi are dispersal limited even at the few kilometer scale of this study. If spore quantities are significantly greater for early stage species this would suggest that early stage species invest more resources into spore production. If the slope of the relationship between spore quantity and distance is more negative for late stage vs. early stage fungi, this would suggest that spores of early stage species disperse farther distances than late stage species. If either of these is true they would provide evidence that there is a strong-tradeoff between investment in dispersal and competitive abilities in ectomycorrhizal communities. Based on the ANCOVA equations we will produce landscape maps of spore dispersal patterns for individual species or species groups.
Draft Copy – Please do not distribute
Figure 3: Nested pattern of ectomycorrhizal species occurrence on study islands. Each column of the matrix represents an ectomycorrhizal species and each row represents an island. The number of islands a species occurs on and the number of species that occur on each island are listed at the end of each column and row, respectively. The matrix was maximally nested using the Nestedness Calculator (Atmar and Patterson 1995), and 5000 randomizations of the matrix showed that the level of nestedness was highly significant (p<0.0001). Note that assemblages are nested almost perfectly with respect to island size.