N 2 O is a potent greenhouse gas involved in the destruction of the protective ozone layer in the stratosphere and contributing to global warming. The ecological processes regulating its emissions from soil are still poorly understood. Here, we show that the presence of arbuscular mycorrhizal fungi (AMF), a dominant group of soil fungi, which form symbiotic associations with the majority of land plants and which influence a range of important ecosystem functions, can induce a reduction in N 2 O emissions from soil. To test for a functional relationship between AMF and N 2 O emissions, we manipulated the abundance of AMF in two independent greenhouse experiments using two different approaches (sterilized and re-inoculated soil and non-mycorrhizal tomato mutants) and two different soils. N 2 O emissions were increased by 42 and 33% in microcosms with reduced AMF abundance compared to microcosms with a well-established AMF community, suggesting that AMF regulate N 2 O emissions. This could partly be explained by increased N immobilization into microbial or plant biomass, reduced concentrations of mineral soil N as a substrate for N 2 O emission and altered water relations. Moreover, the abundance of key genes responsible for N 2 O production (nirK) was negatively and for N 2 O consumption (nosZ) positively correlated to AMF abundance, indicating that the regulation of N 2 O emissions is transmitted by AMF-induced changes in the soil microbial community. Our results suggest that the disruption of the AMF symbiosis through intensification of agricultural practices may further contribute to increased N 2 O emissions.

Abstract. It is well established that agricultural practices alter the composition and diversity of soil microbial communities. However, the impact of changing soil microbial communities on the functioning of the agroecosystems is still poorly understood. Earlier work showed that soil tillage drastically altered microbial community composition. Here we tested, using an experimental grassland (Lolium, Trifolium, Plantago) as a model system, whether soil microbial communities from conventionally tilled (CT) and non-tilled (NT) soils have different influences on plant productivity and nutrient acquisition. We specifically focus on arbuscular mycorrhizal fungi (AMF), as they are a group of beneficial soil fungi that can promote plant productivity and ecosystem functioning and are also strongly affected by tillage management.Soil microbial communities from CT and NT soils varied greatly in their effects on the grassland communities. Communities from CT soil increased overall biomass production more than soil communities from NT soil. This effect was mainly due to a significant growth promotion of Trifolium by CT microorganisms. In contrast to CT soil inoculum, NT soil inoculum increased plant phosphorus concentration and total plant P content, demonstrating that the soil microbial communities from NT fields enhance P uptake. Differences in AM fungal community composition resulting, for instance, in twofold greater hyphal length in NT soil communities when compared to CT, are the most likely explanation for the different plant responses to CT and NT soil inocula.A range of field studies have shown that plant P uptake increases when farmers change to conservation tillage or direct seeding. Our results indicate that this enhanced P uptake results from enhanced hyphal length and an altered AM fungal community. Our results further demonstrate that agricultural management practices indirectly influence ecosystem services and plant community structure through effects on soil biota.

Arbuscular mycorrhizal fungi (AMF) are promoted as biofertilizers for sustainable agriculture. So far, most researchers have investigated the effects of AMF on plant growth under highly controlled conditions with sterilized soil, soil substrates or soils with low available P or low inoculum potential. However, it is still poorly documented whether inoculated AMF can successfully establish in field soils with native AMF communities and enhance plant growth. We inoculated grassland microcosms planted with a grass-clover mixture (Lolium multiflorum and Trifolium pratense) with the arbuscular mycorrhizal fungus Rhizoglomus irregulare. The microcosms were filled with eight different unsterilized field soils that varied greatly in soil type and chemical characteristics and indigenous AMF communities. We tested whether inoculation with AMF enhanced plant biomass and R. irregulare abundance using a species specific qPCR. Inoculation increased the abundance of R. irregulare in all soils, irrespective of soil P availability, the initial abundance of R. irregulare or the abundance of native AM fungal communities. AMF inoculation had no effect on the grass but significantly enhanced clover yield in five out of eight field soils. The results demonstrate that AMF inoculation can be successful, even when soil P availability is high and native AMF communities are abundant.