Arbuscular mycorrhizal fungi (AMF) represent important players in the structure and function of many ecosystems. Yet, we learn about their roles mostly from greenhouse-based experiments, with results subjected to cultivation bias. This study explores multiple aspects of this bias and separates the effect of increased nutrient availability from other cultivation specifics. For 15 grassland plant species from two functional groups (C3 grasses vs dicotyledonous forbs), we compared AMF communities of adults collected from non-manipulated vegetation with those in plants grown in a greenhouse. Nutrient availability was comparable to field conditions or experimentally elevated. We evaluated changes in AMF community composition, diversity, root colonisation, and the averages of functional traits characterising hyphal soil exploration. Additionally, we use the data from the greenhouse experiment to propose a new plant functional trait-the change of AMF colonisation in response to nutrient surplus. The AMF community differed profoundly between field-collected and greenhouse-grown plants, with a larger change of its composition in grass species, and AMF community composition in grasses also responded more to fertilisation than in forbs. Taxonomic and phylogenetic diversity declined more in forbs under cultivation (particularly with elevated nutrients), because in their roots, the AMF taxa from families other than Glomeraceae largely disappeared. A decline in AMF colonisation was not caused by greenhouse cultivation itself but selectively by the elevation of nutrient availability, particularly in grass host species. We demonstrate that the extent of decrease in AMF colonisation with elevated nutrients is a useful plant functional trait explaining an observed response of the plant community to manipulation.
Keywords: Arbuscular mycorrhizal fungi; Cultivation bias; Microscopy; Next-generation sequencing; Nutrient availability; Rhizophagus irregularis.