An estimated 90% of terrestrial plants form symbiotic associations with soil fungi, and the majority of those plant species belong to families that characteristically form associations with arbuscular mycorrhizal (AM) fungi (Smith & Read, 1997). The function of these associations is largely based upon the transfer of carbon (C) from the plant to the fungus, and upon the transfer of mineral nutrients, mainly phosphorus (P), from the fungus to the host plant. Arbuscular mycorrhizal fungi may confer other benefits, including improved soil aggregate stability, plant–water relationships and resistance to plant pathogens, but these benefits have rarely been quantified in an ecosystem context (Newsham et al., 1995).
‘... , the costs and benefits ascribed to the association in microcosms are especially challenging to confirm in nature.’
In microcosms, the plant's benefit from mycorrhizal associations varies with species of AM fungi (Van der Heijden et al., 1998a). This variation in benefit is ecologically relevant (Bever et al., 2002) as it explains plant diversity responses to manipulations of fungal diversity (Van der Heijden et al., 1998b). However, the costs and benefits ascribed to the association in microcosms are especially challenging to confirm in nature (Johnson et al., 1997). Field experiments have focused on agricultural monocultures and prevalent species of AM fungi (e.g. Glomus intraradices) (Lekberg & Koide, 2005). Studies of mycorrhizal functioning in more complex natural systems are lacking (Read, 2002).
Quantification of the cost–benefit of individuals and consortia of AM fungi is pivotal for understanding C and nutrient cycling in the context of ecosystem adaptation to global change (Fitter et al., 2000). Although mycorrhizal research has advanced in the study of P and C exchanges at discrete scales of size and organization, studies of function at lower scales have not yet explained AM fungal interactions for individual genotype, phenotype and environment (Miller & Kling, 2000). New and creative in-field approaches for measuring AM fungus and host responses are crucial for assessing the temporal and spatial variabilities of mycorrhizal interactions.
Pringle & Bever's study, in this issue of New Phytologist (pp. 162–175), focuses on AM fungal species as a determinant of the benefit for plant species in a North Carolina grassland. Their experimental design substantiates the use of lower-scale evaluation of phenotype for AM fungi based on the finding that the mycorrhizal responsiveness of plant species under growth chamber conditions is correlated with that response in the highly divergent field environment. They conclude that AM fungus identity influences the survival and fitness of grassland plants by demonstrating that certain fungal species consistently promote the growth of a diverse set of plants. To define fungal phenotype in the field they confine the evaluation of plant response to the timescale over which the roots are still predominantly colonized by the introduced AM fungus. The consistency of the mycotrophic response across a range of plant species, as well as environmental conditions, validates estimation of a ‘functional phenotype’ in the field based on the growth response under controlled conditions.
Attempting to estimate a ‘functional phenotype’ might be questioned in light of recent evidence for AM fungus–host specificity (Klironomos, 2003) and wide variation in plant vs fungal pathways for P acquisition among host–fungal interactions (Smith et al., 2004). If root colonization is not a good predictor for P uptake or the host response, measurement of lower-level mechanisms for nutrient uptake in mycorrhizal plants may not readily scale-up to the field level. Nevertheless, definition of a growth response phenotype for AM fungi across different host, soil and fertility conditions may provide a practical approach for assessment of their functioning under natural conditions.
Mediation by AM fungi of vegetation responses to environmental change will require experiments at a minimum of two scales to develop and test models for mycorrhizal functioning (Huston, 1999). Experimental designs should either (1) integrate multiple mechanisms at the landscape scale and include such measures as mycorrhizal influences on net primary production, evapotranspiration and nutrient cycling or (2) integrate measures of AM fungal diversity into assessment of ecosystem function. Extension of lower-level processes that determine ecosystem dynamics must be accomplished with smaller-scale experiments such as implantation of plant species of varying life histories colonized by representative AM fungi from that ecosystem, as validated by Pringle & Bever. Lower-scale experiments need to be designed to test each component of fitness of the symbiotic partners under relevant environmental and spatial scales in the field. The design could involve the use of in situ root observation windows, mesh dividers, or bags and isotopic tracers and signature fatty acids (e.g. Fig. 1) to assess the dynamics of nutrient and C exchanges. Measurements of C and P exchanges under controlled conditions are well refined and have been successfully extended to the field (Olsson et al., 1999; Schweiger & Jakobsen, 1999; Johnson et al., 2001).
Even with novel approaches for scaling up the evaluation of functioning in field studies, the dynamics of C and P in mycorrhizal associations in the field will be difficult to interpret under complex conditions in plant communities, as illustrated by the recent study of Li et al. (2008). In their attempt to integrate the phenomenon of plant vs AM pathways for root P acquisition into the consideration of cost–benefit of wheat mycorrhizas, they concluded that mycorrhizal responses of plants grown singly may not apply at the population level. Thus, consideration of functional diversity under natural conditions of plant and fungus competition remains a challenging frontier in mycorrhizal biology.