Is diversity of ectomycorrhizal fungi important for ecosystem function?

One of the most challenging questions faced by ecologists is whether there is a general relationship between biological diversity and ecosystem function. As rates of species extinction appear to be increasing – and evidence has emerged of decreasing biodiversity in plant and microbial communities as a result of air and soil pollution, intensification of agriculture and forestry (Pimm et al., 1995) – the importance of establishing whether biodiversity per se is important for ecosystem function has become a central issue in ecology. While the effects of higher plant species diversity on above-ground productivity has received considerable attention (Loreau & Hector, 2001), and there is evidence that arbuscular mycorrhizal (AM) diversity increases plant species diversity and productivity (van der Heijden et al., 1998; Klironomos et al., 2000) the role of ectomycorrhizal (ECM) diversity in ecosystem function, with one exception (Jonsson et al., 2001), has not been experimentally investigated. In this issue, Baxter & Dighton (2001) (pp. 139–149) report on a study designed specifically to answer the question 'does ectomycorrhiza diversity affect plant growth and nutrient acquisition?'. 'Establishing biodiversity–function relationships remains one of the most intractable challenges in ecological research' Increasing awareness of the wide range of mycelial structures produced by ECM in soil and on roots (Agerer, 1996) together with evidence of considerable interspecific and intraspecific variation in production of nutrient-mobilizing enzyme systems by these fungi (Leake & Read, 1997) provides strong theoretical support for ECM diversity increasing effectiveness of nutrient acquisition from different spatial locations and different substrates in the soil. Indeed, it is arguable that any case for biodiversity affecting ecosystem functioning is much stronger for ECM than for AM fungi because of the greater variation in structure and functioning of ECM. Furthermore, as attempts are now being made to derive generalized relationships between plant biodiversity and ecosystem function, using data derived exclusively from studies of herbaceous plant communities that have arbuscular mycorrhizas (Loreau & Hector, 2001) it is increasingly important that the effects of ECM biodiversity on forest ecosystems are investigated. Baxter & Dighton (2001) grew Betula populifolia seedlings for 10 wk in Petri dishes containing sterilized peat-vermiculite in which the plants were inoculated with 0–4 different species of ECM fungi drawn from a pool of six species. To establish mixed-species communities of ECM fungi in symbiotic association with a plant under axenic conditions is a significant achievement in itself, especially as the species chosen in this case are representative of some of the major different functional and taxonomic groups of the fungi. The effects of realized ECM diversity on root and shoot biomass, nitrogen and phosphorus contents of the plants were examined. The authors conclude that ECM diversity rather than species composition or rates of mycorrhizal colonization is the main factor affecting responses of the plants (changes in shoot/root biomass allocation and uptake of phosphorus) to mycorrhizal associations. The finding that ECM diversity was positively correlated with the total P uptake by the plants lends support to the hypothesis advanced by Roger Koide, in relation to AM fungi, that diverse mixtures of mycorrhizal fungi may provide functional complementarity in the uptake of P, enabling more effective and complete exploitation of soil P (Koide, 2000; Smith et al., 2000). While the most diverse communities of ECM assembled by the authors contained only four species, which is low compared to the typical diversity seen in established forests (Dahlberg, 2001), this level of diversity is appropriate for studies of tree seedlings. However, the pool of species drawn upon in this case (six species) is very low. Kranabetter & Wylie (1998), for example, found that 4-yr-old naturally regenerating seedlings of Tsuga heterophylla growing in forest gaps had a mean of 3.8 ECM species-morphotypes per seedling in the centre of the gaps rising to 6.1 species per seedling under the undisturbed forest canopy. However, in this natural situation the pool of species with which the seedlings were infected was much greater, ranging from 20 species in the clearings to 32 under the closed forest. Although Baxter and Dighton (2001) is a pioneering paper, it is premature to draw general conclusions about links between diversity and functioning of ectomycorrhizas from it. The authors acknowledge the implications of the study are constrained in part by the artificiality of the experimental system, but in addition, as in many previous studies of biodiversity–ecosystem function relationships, there are potentially confounding factors that make the interpretation of the results and their wider significance problematic. Nonetheless, this is a timely paper that has the potential to stimulate further work in this very difficult field of research and it usefully highlights some of the experimental problems that need to be overcome. In their study, Baxter & Dighton (2001) were careful to design an experiment to avoid two of the pit-falls encountered in previous biodiversity–function studies (Huston, 1997): they randomly allocated species to treatments and they tested the effects of each ECM species in symbiosis with the plant to enable the effects of species richness to be distinguished from those of species composition. Unfortunately, probably as a result of the experimental conditions in which the plants were grown in a small volume of peat and vermiculite to which relatively high concentrations of mineral nutrients were added at the start of the experiment, the effects of ECM on growth and nutrition of the plants were, with few exceptions, not significant. Consequently, virtually no species-specific effects can be discerned in this case. The final communities of ECM that were assembled were not strictly random, despite the care taken in the experimental design, since two species were responsible for most of the cases in which infections were not established. As the number of ECM fungi inoculated onto the Betula seedlings was increased from one to two and then to four there was a progressive decrease in effectiveness of establishment of all inoculated species from 100% to 86% and 52%, respectively. Thus as community diversity increases there appears to be increasing bias in the species combinations that successfully infect the plants simultaneously. This problem is hard to avoid. It may be possible to employ methods of inoculation that reduce the rates of failure to form mycorrhiza, but the decline in numbers of root tips infected by each species as diversity is increased is predictable, and clearly shown in this paper and in Jonsson et al. (2001). Assembly of 'random' communities of ECM fungi becomes increasingly difficult at higher levels of biodiversity and this confounds attempts to distinguish biodiversity and species composition effects. A second confounding factor that can be easily remedied in future studies is the method of inoculation, which differed across the diversity treatments. In the single species replicates, ECM inoculum was placed as a single patch in the centre of each Petri dish in which the tree roots were grown. In the two-species combinations, two spatially separate patches of inoculum were added and in the four-species combinations four spatially separate patches of inoculum were added. The dispersal of the inoculum therefore progressively increased in parallel with the increased diversity. The authors report increases in mycorrhizal colonization of the seedlings with increasing diversity of ECM. Although the strength of correlations between species diversity and plant responses were stronger than the effects of ECM colonization, we cannot exclude the possibility that these effects are due to the different methods of inoculation. The recent advances in knowledge of mycorrhizal community structures achieved through morphotype and molecular identification of ECM on roots (Dahlberg, 2001) provides a much firmer basis on which to select species for inclusion in future studies of biodiversity and function. The size of the species pool and the numbers of species combinations used (the level of diversity) can increasingly be based on knowledge of natural ECM communities. As the abundance of species that rarely or never produce visible fruiting structures has become apparent (Dahlberg, 2001) increased efforts will need to be made to include them in biodiversity and functional studies. Clearly too, it is important that experimental conditions are employed that provide the kinds of spatial heterogeneity and variation in forms of nutrients that occur in the field, since this is likely both to result in significant benefits to the plants from their mycorrhizal associations and to reveal any 'complementarity' of ECM species in accessing different nutrient pools. One possible way to achieve this may be to establish communities of ECM on tree seedlings in soil-based microcosms in which natural sources of organic and inorganic nutrients are supplied (Timonen & Sen, 1998). Establishing biodiversity-function relationships remains one of the most intractable challenges in ecological research. Ectomycorrhizal communities tend to be increasingly diverse as tree seedlings grow, become established and finally become part of mature forests (Kranabetter & Wylie, 1998). It is therefore likely that any general relationship between ECM diversity and ecosystem function will be most important in mature forests, but artificial reconstruction of these communities with different levels of ECM diversity does not appear to be a realistic experimental approach. Seedlings are clearly more amenable than trees to such studies, but there is increasing evidence that the range of ECM diversity that can be supported by individual seedlings is quite small (Kranabetter & Wylie, 1998), almost certainly as a result of the constraint provided by the limited carbon-fixing capacity of seedlings compared with established trees. Investigations of the cost-benefit relationships of ectomycorrhizal diversity are a high priority. How are the competing demands for carbon by different ECM met by the plants? Is more carbon allocated proportionally to the partners that provide the most nutrients or other functional benefits? Does ECM diversity have any consistent effect either on individual plants or on ecosystem functioning? These questions are easy to pose. The answers are difficult to obtain. A major challenge lies ahead.