Grazing exclusion promotes grasses functional group dominance via increasing of bud banks in steppe community

To understand the bud banks response to grazing exclusion, we conducted a demographic experiment in long-term grazing exclusion (20 year and 30 year) typical steppe. Results showed that grass functional group constituted the vast majority of the aboveground vegetation and belowground bud bank in all treatments. Long-term grazing exclusion significantly increased total aboveground biomass (2.5 and 2.6 times in 20y and 30y grazing exclusion grasslands, respectively), and decreased total stem density (31% and 37% in 20y and 30y grazing exclusion grasslands, respectively). Grazing exclusion for 20 and 30 years increased grass aboveground biomass respectively by 6.0 and 8.0 times, and decreased grass stem density by 38% and 33%. Grazing exclusion had different effects on belowground bud density of grass and forb functional group. Long-term grazing exclusion significantly increased plant buds and bud bank size (25% and 37% in 20y and 30y grazing exclusion grasslands, respectively), especially for grass functional group (49% and 95% in 20y and 30y grazing exclusion grasslands, respectively), but had no significant effects on forb bud density. Changes of aboveground community were significantly related to changes of belowground bud bank under both grazing and grazing exclusion grasslands. The bud bank density of grass functional group was significantly positive with total (R² = 0.33, P < 0.05) and grass aboveground biomass (R² = 0.36, P < 0.01), while negative related with total (R² = -0.27, P < 0.05) and grass stem density (R² = -0.22, P < 0.05). Grazed grasslands, 20y and 30y grazing exclusion grasslands all were not meristem limited and had large reserve bud banks, which would completely replace the aboveground stem population during the growing season. These findings indicate that grazing exclusion could not only improve a large bud bank for grassland restoration but also improve the dominance of grass functional group by increasing grass belowground bud banks in typical steppe community. We propose that the belowground bud bank might be a good approach to indicating potential succession direction of aboveground community.