Abstract
Arbuscular mycorrhizal fungi (AMF) are the predominant type of mycorrhizal fungi in roots and rhizosphere soil of grass species worldwide. Grasslands are currently experiencing increasing grazing pressure, but it is not yet clear how grazing intensity and host plant grazing preference by large herbivores interact with soil- and root-associated AMF communities. Here, we tested whether the diversity and community composition of AMF in the roots and rhizosphere soil of two dominant perennial grasses grazed differently by livestock change in response to grazing intensity. We conducted a study in a long-term field experiment in which seven levels of field-manipulated grazing intensities were maintained for 13 years in a typical steppe grassland in northern China. We extracted DNA from the roots and rhizosphere soil of two dominant grasses, Leymus chinense (Trin.) Tzvel. and Stipa grandis P. Smirn, with contrasting grazing preference by sheep. AMF DNA from root and soil samples were then subjected to molecular analysis. Our results showed that AMF α-diversity (richness) at the virtual taxa (VT) level varied as a function of grazing intensity. Different VTs showed completely different responses along the gradient, one increasing, one decreasing and others showing no response. Glomeraceae was the most abundant AMF family along the grazing gradient, which fits well with the theory of disturbance tolerance of this group. In addition, sheep grazing preference for host plants did not explain a considerable variation in AMF α-diversity. However, the two grass species exhibited different community composition in their roots and rhizosphere soils. Roots exhibited a lower α-diversity and higher β-diversity within the AMF community than soils. Overall, our results suggest that long-term grazing intensity might have changed the abundance of functionally-diverse AMF taxa in favor of those with disturbance-tolerant traits. We suggest our results would be useful in informing the choice of mycorrhizal fungi indicator variables when assessing the impacts of grassland management choices on grassland ecosystem functioning.