Grazing increases the positive feedback of legumes while decreasing the negative feedback of grass

Abstract Herbivore grazing affects plant growth and community structure in grasslands. This effect could be directly through foraging and dung/urine return or indirectly through plant–soil feedbacks (PSFs). Addressing the grazing effect on the feedback of plants can explicate the causes of community changes in the grazing system. However, how grazing and PSF interact to affect plant growth remains unclear. Here, we conducted a classic PSF experiment. In the conditioning stage, two native plant species (a grass Bromus inermis and a legume Medicago sativa ) were planted in the field with four simulated grazing treatments (ambient, mowing, dung/urine addition, and mowing + dung/urine addition) in a meadow grassland of northern China. In the feedback stage, B. inermis and M. sativa were planted in the soils (both unsterilized and sterilized) from each treatment in the field experiment. Plant biomass of M. sativa showed positive feedback while B. inermis showed negative feedback across all the simulated grazing treatments. Simulated grazing (mowing and dung/urine addition) increased the positive feedback of M. sativa , while decreasing the negative feedback of B. inermis . The addition of dung/urine to the soil was found to have a significantly stronger impact on plant growth feedback compared to the effect of mowing. Dung/urine addition enriches the soil with higher levels of available nitrogen and phosphorus. Our results suggested that legume plants should have positive PSFs while grass should have negative feedback, which might be amplified by grazing because of the dung/urine fertilization effect. Our study improves the understanding of PSF effects on plant growth and community change in grazed grassland.