This doctoral thesis reports on the relevance of arbuscular mycorrhizal fungi (AMF) on denitrification potential activity and nitrous oxide (N2O) emissions from a native fertile agricultural soil. Agricultural soils are the main source of N2O, a powerful greenhouse gas contributing to on-going climate change and destruction of our stratospheric ozone layer. There is a growing recent scientific interest in understanding of the significance of AMF, widespread symbiotic soil fungi, on denitrification and N2O emissions. However, most of the existing studies have performed reductionist experiments with single plant host systems grown under artificial soil substrates and have highlighted that the composite effect of AMF reduces denitrification and N2O emissions. In Chapter 2, we reviewed the role of AMF on environmental controls of denitrification and N2O emissions. In Chapter 3, to address direct vs indirect effects of AMF on denitrification potential opposed to composite effects, we performed a realistic experiment where we combined a manipulation of two factors: AMF and soil aggregation (state which represents the native microhabitats of denitrifying microbes) through either crushing mechanically soil aggregates or using agricultural soil with intact soil aggregates. Furthermore, we added AMF inoculum or not in the form of Rhizophagus irregularis into the indigenous AMF community in both soil aggregate treatments. We then grew a single host plant Zea mays in compartmentalized mesocosms for 4 months. Our results were inconsistent with the existing literature, highlighting that AMF reduced denitrification and N2O emissions. In contrast, using a denitrification enzyme activity (DEA) assay, we found that potential N2O emission rates and the denitrification potential ratio of potential N2O emission rates over total potential denitrification activity were higher in the undisturbed soil and in the mesocosms with higher AM hyphal densities (direct effect of AMF). Further, AMF-promoted soil aggregation increased both denitrification potential parameters (indirect effect of AMF). In Chapter 4, we explored how AMF might behave in a typical European grassland plant community experimental setting with a realized gradient in plant diversity. A gradient in plant diversity emerged from stochastic differences in plant seedling establishment across the sixteen replicates, irrespective of AMF treatment. Furthermore, plant diversity mitigated potential emission rates of N2O while AMF reduced microbial denitrification potential activity and the denitrification potential ratio. Our results were incongruent to the existing literature, indicating that the interactive effects between AMF and plant diversity mask their independent contributions to the denitrification ecosystem process. We presented a novel picture of AMF functioning on denitrification potential in natural grassland systems. Finally, in Chapter 5 we tested whether AMF also mitigates current N2O emissions when there are no differences in nitrate (NO3-) availability between AMF treatments at harvest using a gas-flow-soil-core incubation system. We found that AMF reduced the convexity of the cumulative N2O emission curves in inoculated agricultural soil cores with AMF compared to noninoculated incubated in oxic and anoxic conditions. A reduction in the convexity meant a decrease in microbial denitrification activity rates and N2O emissions from soil cores by AMF. We found here the first compelling evidence that AMF hyphae abundance mitigates current cumulative N2O emissions originating probably from nitrification and denitrification under native fertile agricultural soils, irrespective of initial soil NO3- concentrations. By performing realistic experiments and using unsterilized arable soil with intact soil aggregates inoculated or not with a model AMF species, R. irregularis, we increased the ecological relevance of our results by demonstrating that AMF strongly influence N2O emission rates, but also denitrification potential activity and denitrification ratio compared to the existing studies. The results presented here point out that managing the AM plant symbiosis could have significant implications for better management of N2O emissions from agricultural soils.
