Nutrients and herbivores have independent effects on the temporal stability of aboveground biomass in grasslands; however, their joint effects may not be additive and may also depend on spatial scales. In an experiment adding nutrients and excluding herbivores in 34 globally distributed grasslands, we found that nutrients and herbivores mainly had additive effects. Nutrient addition consistently reduced stability at the local and larger spatial scales (aggregated local communities), while herbivore exclusion weakly reduced stability at these scales. Moreover, nutrient addition reduced stability primarily by causing changes in local community composition over time and by reducing local species richness and evenness. In contrast, herbivore exclusion weakly reduced stability at the larger scale mainly by decreasing asynchronous dynamics among local communities, but also by weakly decreasing local species richness. Our findings indicate disentangling the influences of processes operating at different spatial scales may improve conservation and management in stabilizing grassland biomass.
To predict plant responses under climate change, we need to understand how thermal conditions and herbivory contribute to shoot growth. Here, we used empirical dynamic modelling (EDM) to analyse an 18-year time series from an experimental system at the forest-tundra ecotone to identify relationships between growth, climate, insect herbivores, browsers and ramet age. We found that negative effects of insect herbivory on willow shoot growth are intensified during warmer years. Moreover, the negative effect of insect herbivores was moderated by ramet age, but this moderation was only realized in the absence of vertebrate herbivores – under browsing by both ptarmigans and reindeer, the positive effects of ramet age were eliminated. Jointly, these results demonstrate the context-dependent and dynamic effects of climate and multiple herbivores on shoot growth, and improve our ability to predict effects of climatic warming in arctic environments.