Climate change is expected to increase the magnitude and frequency of extreme drought in most grassland ecosystems. Exploring the responses of below‐ground bud banks and their relationships with above‐ground plant structure and drought is need to explain how climate change will impact grassland ecosystems. However, studies on the response of community‐scale bud and shoot densities to experimental drought along an aridity gradient are rare. We experimentally removed 66% of growing season precipitation for 4 years in three temperate grasslands that spanned an aridity gradient in northern China. We quantified the legacy effects of drought on grass, forb and total community below‐ground bud density, above‐ground shoot density and the ratio of bud to shoot density 1 year following treatment. Below‐ground bud density was lowest at the highest aridity site for the entire community, while above‐ground shoot density was highest at the medium aridity site. Below‐ground bud and above‐ground shoot densities were the lowest at the high aridity site for grasses but the highest for forbs at this site. Bud:shoot ratios decreased with increasing aridity for grasses, yet remained constant for forbs along the aridity gradient. Below‐ground bud density in drought plots remained lower than controls a year following drought at each site. Experimental drought did not alter the below‐ground bud bank for grasses but decreased forb bud banks across sites. Experimental drought had little legacy effects on above‐ground shoot density and bud:shoot ratios for grasses, forbs and the total community at each site. Our results suggest that grass and forb bud banks can differ in their responses to both multi‐year drought along an aridity gradient, and that bud limitation for shoot generation may increase as grasslands get drier. Bud bank responses to climate will impact plant community functioning and resilience. Thus, incorporating bud bank dynamics will improve projections of grassland ecosystems under future climate change. Read the free Plain Language Summary for this article on the Journal blog.

Asexual reproduction plays a fundamental role in the structure, dynamics and persistence of perennial grasslands. Thus, assessing how asexual reproductive traits of plant communities respond to drought may be key for understanding grassland resistance to drought and recovery following drought. Here, we quantified three asexual reproductive traits (i.e. above‐ground tiller abundance, below‐ground bud abundance and the ratio of tillers to buds) during a 4‐year severe drought and a 2‐year drought recovery period in four grasslands that spanned an aridity gradient in northern China. We also assessed the relationship between these traits and the resistance and recovery of above‐ground net primary productivity (ANPP). We found that drought had limited and largely inconsistent effects on asexual reproduction among drought and recovery years and grasslands overall. Drought increased tiller abundance in the first treatment year and reduced bud banks by the fourth treatment year across grasslands. However, neither of the three asexual reproductive traits were correlated with drought resistance of ANPP. Drought legacies differed among the four grasslands with positive, negative and no legacies evident for the three asexual reproductive traits, and no clear relationship with aridity. Bud banks and tiller to bud ratio decreased and increased, respectively, in the first recovery year, but not in the second recovery year. In contrast to drought resistance, community bud abundance was strongly related to recovery, such that communities with higher bud abundance had greater ANPP recovery following drought. Synthesis. These results suggest that asexual reproductive traits may be important drivers of ecosystem recovery after drought, but that variable responses of these asexual reproduction traits during drought complicates predictions of overall grassland responses.

No abstract