Plant population responses to environmental variability are primarily driven by survival-reproduction trade-offs and mediated by aridity.

Natural populations may suffer negatively from increased environmental variability due to climate change; however, several mechanisms can mitigate those effects by changing the vital rates of a population (e.g., survival, reproduction). Despite important analytical and theoretical advances, we still do not know how and to what extent environmental regimes, life history traits, and evolutionary history determine the buffering capacity of natural populations. To address these questions, we parameterise a Bayesian generalised linear mixed model with high-resolution vital rate data from 121 natural populations across 78 plant species. We show that population responses to environmental variability vary four orders of magnitude along a 'demographic buffering continuum'. Furthermore, the position of a given population along said continuum is predicted by a survival-reproduction trade-off and by the degree of aridity the population experiences. Our findings open a promising avenue of research to improve ecological forecasts and management of natural populations in the Anthropocene.