Demographic buffering in natural populations: A multi-level perspective

Environmental stochasticity poses significant challenges to population persistence. A key mechanism thought to buffer populations against such variability is demographic buffering-the ability of a population to stabilise growth despite temporal fluctuations in survival, development or reproduction. However, empirical tests of demographic buffering remain limited and often yield conflicting results. Here, we propose an integrative demographic framework that combines two complementary approaches to identify demographic buffering: (1) stochastic elasticities, which quantify the sensitivity of long-term stochastic growth rates (λ_s) to variance in demographic processes, and (2) second-order derivatives of deterministic growth (λ₁), which indicate whether selection acts to reduce or amplify variance in vital rates. Applying this framework to 43 natural populations across 37 mammalian species, we position each species along a variance continuum and assess whether those with low stochastic elasticities-suggestive of buffering-also exhibit signs of concave selection on key demographic processes. While most primates and a few other long-lived mammals occupy the buffered end of the continuum, only one species-the Columbian ground squirrel-exhibits strong support for our hypothesis, with key vital rates both critical for λ₁ and under concave selection. In contrast, primates, despite showing low stochastic elasticities, often show convex or absent second-order effects on their most influential vital rates, indicating a mismatch between ecological buffering and evolutionary constraint. Our findings suggest that demographic buffering is more dynamic and context dependent than previously recognised. Selection does not consistently act to reduce variance in key demographic processes, even in species where population growth appears robust to environmental variability. This decoupling implies that evolutionary and ecological signals of buffering may not always align. Our framework offers a new lens to dissect the demographic and selective processes underpinning resilience, providing a scalable tool for exploring demographic strategies across taxa. Future work integrating phylogenetic context, trait covariation and environmental drivers will be essential to understand the adaptive value of demographic buffering under global change.