| Abstract: | We established a 13‐week greenhouse experiment based on replicated microcosms to test whether the effects of defoliation on grassland plants and soil organisms depend on plant species composition and the presence of arbuscular mycorrhizal (AM) fungi. The experiment constituted of three treatment factors – plant species composition, inoculation of an AM fungus and defoliation – in a fully factorial design. Plant species composition had three levels: (1) Trifolium repens monoculture (T), (2) Phleum pratense monoculture (P) and (3) mixture of T. repens and P. pratense (T+P), while the AM inoculation and the defoliation treatment had two levels: (1) no inoculation of AM fungi and (2) inoculation of the AM fungus Glomus claroideum BEG31, and (1) no trimming, and (2) trimming of all plant material to 6 cm above the soil surface three times during the experiment, respectively. At the final harvest, AM colonization rate of plant roots differed between the plant species compositions, being on average 45% in T, 33% in T+P and 4% in P. Defoliation did not affect the colonization rate in T but raised the rate from 1% to 7% in P and from 20% to 45% in T+P. Shoot production and standing shoot and root biomass were 48%, 85% and 68% lower, respectively, in defoliated than in non‐defoliated systems, while the AM fungus did not affect shoot production and root mass but reduced harvested shoot mass by 8% in non‐defoliated systems. Of the plant quality attributes, defoliation enhanced the N concentration of harvested shoot biomass by 129% and 96% in P and T+P, respectively, but had no effect in T, while the C concentration of shoot biomass was on average 2.7% lower in defoliated than in non‐defoliated systems. Moreover, defoliation reduced shoot C yield (the combined C content of defoliated and harvested shoot biomass) on average by 47% across all plant species compositions and shoot N yield by 37% in T only. In contrast to defoliation, the AM fungus did not affect shoot N and C concentrations or shoot N yield, but induced 10% lower C yield in non‐defoliated systems and 17% higher C yield in defoliated T. In roots, defoliation led to 56% and 21% higher N concentration in P and T+P, respectively, and 28% higher C concentration in P, while the mycorrhizal fungus lowered root N concentration by 9.7% in defoliated systems and had no effect on root C concentrations. In the soil, the nematode community was dominated by bacterivores and the other trophic groups were found in a few microcosms only. Bacterivores were 45% more abundant in defoliated than in non‐defoliated systems, but were not affected by plant species composition or the AM fungus. Soil inorganic N concentration was significantly increased by defoliation in T+P, while the mycorrhizal fungus reduced NH4–N concentration by 40% in T. The results show that defoliation had widespread effects in our experimental systems, and while the effects on plant growth were invariably negative and those on bacterivorous nematodes invariably positive, most effects on plant C and N content and soil inorganic N concentration varied depending on the plant species present. In contrast, the effects of defoliation did not depend on the presence of the AM fungus, which suggests that while the relative abundance of legumes and grasses is likely to have a significant role in the response of legume–grass communities to defoliation, the role of AM fungi may be less important. In line with this, the AM fungus had only a few significant effects on plant and soil attributes in our systems and each of them was modified by defoliation and/or plant species composition. This suggests that the effects of AM fungi in legume–grass communities may largely depend on the plant species present and whether the plants are grazed or not. |
