An empirical investigation of the possibility of adaptability of arbuscular mycorrhizal fungi to new hosts.

Little is known about the adaptive capacity of arbuscular mycorrhizal (AM) fungi to novel hosts. Here we assessed the possibility of two heterospecific AM fungal isolates to adaptively change, in terms of host biomass response, as a function of host plant identity, over the course of a growing season. First, we produced pure inocula of Rhizophagus clarus and Rhizophagus intraradices, each starting from a single spore. Second, we "trained" each isolate individually in a community with two plants, sudangrass (Sorgum bicolour subsp. drummondii) and leek (Aliium ampeloprasum var. porrum), using a dual-compartment system to allow the establishment of a common mycorrhizal network between the two hosts. Third, we conducted a greenhouse experiment to reciprocally test each "trained" clone, obtained from each compartment, either with the same (home), or the other host (away) under two contrasting phosphorus levels. Overall, results did not support adaptive responses of the AM fungi to their hosts (i.e., greater host biomass under "home" relative to "away" conditions), but the opposite (i.e., greater host biomass under "away" relative to "home" conditions) was more frequently observed. These changes in AM fungal symbiotic functioning open the possibility for relatively rapid genetic change of arbuscular mycorrhizal fungi in response to new hosts, which represents one step forward from in vitro experiments.