Competition-function tradeoffs in ectomycorrhizal fungi

Introduction

Ectomycorrhizal fungi (EMF), the belowground mutualistic partners of most of the world’s temperate tree species, are key regulators of primary production and nutrient remineralization in terrestrial ecosystems. EMF mediate the transfer of water and nutrients from the soil to their host trees (Brownlee et al., 1983; Leake et al., 2004) and serve as a pathway for photosynthetically fixed carbon into the soil community (Talbot, Allison & Treseder, 2008). These fungi also affect the extent of carbon storage (Averill, Turner & Finzi, 2014; Clemmensen et al., 2015) and the rate of nutrient cycling (Leake et al., 2004; Koide, Fernandez & Malcolm, 2014) belowground.

The species composition of EMF communities affects these functions. For example, some taxa utilize distinct foraging strategies, such as the formation of rhizomorphs which allow the transport of water and nutrients over long distances (Brownlee et al., 1983; Agerer, 2001; Agerer, 2006), accumulate relatively high amounts of mycelial biomass belowground (Hobbie, 2006), and are linked with efficient nitrogen mobilization and low carbon sequestration (Clemmensen et al., 2015). EMF that form melanized hyphae may be more tolerant of water stress (Fernandez & Koide, 2013) and resistant to decomposition (Fernandez & Koide, 2014; Koide, Fernandez & Malcolm, 2014) than other taxa. EMF also differ in the extent to which they produce extracellular enzymes that break down soil organic matter and release nitrogen and phosphorus (Courty et al., 2006; Courty, Franc & Garbaye, 2010; Jones et al., 2010; Tedersoo et al., 2012). Because fungal traits are linked through their function to ecosystem processes (Koide, Fernandez & Malcolm, 2014; Treseder & Lennon, 2015), understanding the factors controlling ectomycorrhizal community composition on host tree root systems is of major importance.

While both abiotic (e.g., edaphic environment (Moeller, Peay & Fukami, 2014)) and biotic (e.g., plant community context (Bogar & Kennedy, 2013; Moeller et al., 2015)) conditions may play a primary role in filtering members of the ectomycorrhizal community, recent work has highlighted the importance of fungal species interactions to ectomycorrhizal community assembly (Kennedy, 2010). Competitive interactions, in particular, may locally structure fungal communities within a host plant’s root system, leading to competitive exclusion (Villeneuve, Le Tacon & Bouchard, 1991; Kennedy & Bruns, 2005; Kennedy, Peay & Bruns, 2009) or spatial segregation (Taylor & Bruns, 1999; Pickles et al., 2012). Over longer time scales, such competitive interactions may contribute to the successional patterns observed within ectomycorrhizal communities (Mason et al., 1983; Visser, 1995; Nara et al., 2003), likely in part through a tradeoff between colonization and competitive abilities (Lilleskov & Bruns, 2003; Kennedy et al., 2011).

Some functional traits such as foraging type (Peay, Kennedy & Bruns, 2011; Clemmensen et al., 2015) and propagule persistence (Baar et al., 1999; Taylor & Bruns, 1999) are associated with successional stage; however, it is not clear whether these differences in functionality result in differing competitive abilities. Here, we tested for a tradeoff between competitiveness and nutrient acquisition ability (measured as enzymatic activity) across three different ectomycorrhizal fungal genera. By holding host tree age, inoculum potential, and environmental conditions constant, we experimentally tested three hypotheses.

First, we hypothesized that a dominance hierarchy exists among the three fungal taxa used in our study, Rhizopogon occidentalis, Suillus pungens, and Thelephora terrestris. Based on prior greenhouse experimental work, we expected R. occidentalis to be competitively dominant to S. pungens (Kennedy et al., 2007; Kennedy et al., 2011), and we expected T. terrestris, an aggressive seedling colonizer (Mason et al., 1983; Velmala et al., 2013), to be competitively dominant to both these species. Second, we expected this dominance hierarchy to be inversely related to fungal enzymatic function. We based this hypothesis on the rationale that fungi experience an energetic tradeoff: they can either produce metabolically costly extracellular enzymes, or they can invest in chemical defenses against their competitors. Third, we hypothesized that this tradeoff would impact host tree seedling growth: seedlings would accrue the highest biomass when associating with the least dominant, most highly enzymatically functional fungi.

Materials and Methods

Results

Fungal mycorrhization levels

In single-species treatments, all fungi established associations with P. muricata seedlings. Mycorrhization levels differed by fungal taxon (Fig. 1). Thelephora terrestris had the highest mycorrhization level of 68.2 ± 0.00159% (mean ± standard deviation) root length colonized. Rhizopogon occidentalis (23.5 ±0.0952%) and Suillus pungens (27.6 ± 15.6%) were similar to one another in abundance. No ectomycorrhizal fungi were detected on the control seedlings.

Hypothesis 2: dominance-function tradeoff

Enzyme profiles from single-species treatments supported our hypothesis that the least competitive fungus, S. pungens, would have the highest enzyme activity levels. S. pungens enzymatic activity was higher than non-mycorrhizal control root tips for all five extracellular enzymes assayed (Fig. 2, yellow bars). In contrast, T. terrestris, the competitively dominant fungus, had enzymatic activity levels indistinguishable from controls (Fig. 2, blue bars), and R. occidentalis, of intermediate competitive ability, had elevated enzymatic activity levels only for α- and β-glucosidase and N-acetyl-glucosaminidase (Fig. 2, red bars). Fungi varied significantly in their overall enzymatic profiles (Fig. 2F, Table 1; see Fig. S1 for additional PCA axes). Variance in enzymatic activity was greatest for S. pungens, intermediate for R. occidentalis, and lowest (equivalent to control tips) for T. terrestris (Fig. 3). In multi-species treatments, β-Xylosidase and Acid Phosphatase expression levels were elevated for T. terrestris-infected tips (Fig. 4).

Table 1:

Principal components analysis of enzymatic activity for single-species treatments.

Principal component axis	1	2	3	4	5
Percent explained	79.84	14.18	4.06	1.11	0.800
Loadings
α-glucosidase	−6.61	1.53	−2.10	0.0609	0.775
β-glucosidase	−6.36	2.60	1.67	−0.981	0.113
N-acetylglucosaminidase	−6.82	1.33	0.965	1.29	−0.289
β-xylosidase	−6.77	−1.60	−1.19	−0.437	−1.01
Acid phosphatase	−5.21	−4.77	0.91	−0.00634	0.568

DOI: 10.7717/peerj.2270/table-1

Hypothesis 3: seedling growth response

Seedling aboveground, belowground, and total biomass did not differ by treatment; thus our third hypothesis was not supported. Overall, we did observe a positive relationship between mycorrhization level and seedling biomass (Fig. 5).

Discussion

In this study, we present an experimental test for competition-function tradeoffs across three genera of ectomycorrhizal fungi (EMF). We observed a clear competitive dominance hierarchy among the EMF that was inversely related to their extracellular enzymatic activities. Specifically, the most competitively dominant EMF, Thelephora terrestris, had enzymatic activities indistinguishable from non-mycorrhized tree roots, whereas the competitively inferior Suillus pungens exhibited elevated enzymatic activity across all five enzymes assayed. A number of studies have previously documented differences in enzymatic activity profiles across (Courty et al., 2005; Courty et al., 2006; Buée et al., 2007; Courty, Franc & Garbaye, 2010; Kipfer et al., 2012; Velmala et al., 2013; Walker, Ward & Jones, 2016) and within (Jones et al., 2010) fungal genera. These differences, especially when observed in the field among members of assembled EMF communities, are suggestive of functional complementarity (Courty et al., 2005; Buée et al., 2007; Jones et al., 2010). In some cases, e.g., Rhizopogon species, enzymatic activities are disproportionately high relative to species abundance (Walker et al., 2014). However, to our knowledge, this study is the first to experimentally link these differences in enzymatic activity with competitive ability. Because our study, like others that quantify species-specific enzymatic activity, used root tips to obtain fungal tissues, our enzymatic data are representative only of root-tip associated exoenzyme activities, which may be different than expression levels in other parts of the soil (e.g., at the edges of hyphal extent furthest from the host tree where fungi are foraging for nutrients, or at local hotspots of resource availability in the heterogeneous soil environment) (Wright et al., 2005; Liao et al., 2014). Thus, much work remains to be done to determine the whole organism’s functional potential.

That T. terrestris was the most competitively dominant EMF in our study was not surprising given its proclivity for vigorous growth under greenhouse conditions (Mason et al., 1983; Velmala et al., 2013). However, S. pungens and R. occidentalis are key community members on Pinus muricata in the field, including at Point Reyes National Seashore where the seeds, soils, and fungal spores used in this study were collected (Peay et al., 2007). All three species are capable of colonizing seedlings through spore dispersal, with S. pungens and R. occidentalis more consistently observed on small tree islands than T. terrestris (Peay et al., 2007). R. occidentalis colonizes through generation of a long-lived spore bank (Bruns et al., 2009), while S. pungens and T. terrestris are the two most prolific aerial spore dispersers in the system (Peay et al., 2012). Our competitive dominance hierarchy, which is the inverse of this dispersal hierarchy, suggests a competition-colonization tradeoff among these taxa similar to that observed by Kennedy et al. (2011), at least in terms of root system abundance. While EMF may compete in other ways, such as by competing for nutrient and water resources in the soil or by competing for plant carbon resources (delivery of which may vary by root tip occupancy), in this case the observed complete exclusion of other EMF by T. terrestris suggests strong competitive dominance. However, in cases where exclusion is incomplete measurements of carbon and nitrogen acquisition may be necessary to determine competitive dominance. Although competitively excluded from seedling root systems by T. terrestris in our study, the persistence of S. pungens and R. occidentalis in the field (at least through the first decade of succession) could be the result of several mechanisms. First, less competitive fungi may be maintained through active partner maintenance. This could be the case if trees allocate carbon to their fungal partners in proportion to their partners’ provision of resources (Hoeksema & Kummel, 2003), which would likely be greater for highly enzymatically active EMF like S. pungens. Second, trees with larger root systems are likely to support a greater diversity of EMF than the seedlings in our study (Nara et al., 2003). Third, competition among EMF can be context dependent (Kennedy, Peay & Bruns, 2009), and while T. terrestris is dominant under greenhouse conditions, this is unlikely to be the case for all biotic and abiotic conditions.

Greenhouse growth conditions may also be responsible for the homogeneity of the seedling growth response. Although Kipfer et al. (2012) found that the most enzymatically active fungus in their study, Suillus granulatus, had a positive effect on seedling growth, we did not observe statistically significant differences in growth by treatment in our study. In part, this was likely due to the homogeneous, high-quality soil environment created by autoclaving the experimental soils, which can release substantial amounts of nutrients into plant-accessible pools, reducing the impact of EMF on seedling growth (Peay, Bruns & Garbelotto, 2010) and plant carbon allocation to EMF (Hobbie, 2006). Other studies have found no relationship between EMF competitiveness and seedling growth (Kennedy, Peay & Bruns, 2009). Perhaps, at least at the early seedling stage, growth effects are obscured by conflicting mechanisms. “High-quality” partners like S. pungens may be more expensive thanks to the energetic demands of producing extracellular enzymes; this additional carbon cost may negate any benefits to the host seedling, particularly in nutrient-rich soils.

Although T. terrestris was the only fungus observed by harvest time in all treatments that included it as a source of inoculum, we did observe reductions in its mycorrhization level and increases in some of its enzymatic activities in multi-species treatments relative to monoculture. These differences may be due to a time lag in competitive displacement of R. occidentalis similar to that observed by Lilleskov & Bruns (2003) when R. occidentalis was in competition with Tomentella sublilacina (like T. terrestris, a member of the Thelephoraceae), and/or shifts in T. terrestris enzymatic function induced by the presence of other EMF. Further experimental manipulation of community composition is likely to elucidate the roles of such mechanisms and clarify the role of fungal identity in mutualism function.

We also observed differences in absolute mycorrhization levels among the EMF in our single-species treatments. By our metric (percent of total root length colonized), T. terrestris had mycorrhization levels that were almost double those of the other EMF. In part, this may be due to different root growth forms associated with R. occidentalis and S. pungens, which tend to induce the production of tightly bunched clusters of root tips (HV Moeller & KG Peay, pers. obs., 2011; see also images in the DEtermination of EctoMYcorrhizae database, http://www.deemy.de/) whose prevalence would be underestimated by our grid-intersect sampling method. In contrast, T. terrestris exhibits greater spatial extent along the length of fine roots, and the formation of root tip clusters is not observed. While this difference in mycorrhization may also be due to interspecific differences in spore inoculum viability or rates of vegetative spread across root systems, prior studies have observed similar levels of mycorrhization for the three genera we studied (Browning & Whitney, 1993; Kennedy & Bruns, 2005). Indeed, Kennedy & Bruns (2005) observed ∼30% mycorrhization levels for Rhizopogon species within two months of inoculation. The evolutionary and ecological reasons for these differences in colonization strategy remain unclear, but a fuller understanding of EMF spatial extent and enzymatic function beyond the plant’s immediate root zone will likely shed light on these questions.

Supplemental Information