ABSTRACT Aims Total above‐ground biomass is a key indicator of grassland productivity and ecosystem functioning. While responses of total biomass to environmental gradients are well studied, it remains unclear which abundance groups, dominant, subordinate or rare species, drive these responses and whether aggregate patterns reflect species‐level sensitivities or emerge from compensatory dynamics. We aimed to disentangle the contributions of different abundance groups to community‐level biomass–environment relationships. Location Poland and Czechia (Sudetes Mountains). Methods We used vegetation‐plot data from species‐rich mesic hay meadows, integrating species‐specific biomass measurements. We partitioned community biomass into three groups based on relative abundance: dominant, subordinate and rare species. We used random forest and linear mixed‐effects models to quantify environmental sensitivity at group and species levels. Spatial biomass stability and compensatory dynamics were assessed using coefficients of variation and species synchrony metrics along soil, climatic and topography‐related gradients. Results Random forest models for non‐dominant groups (subordinate and rare) closely matched the performance and predictor importance of the total biomass model, despite weak and heterogeneous responses of individual species to abiotic gradients. In contrast, dominant species, when considered as an aggregate, showed no detectable environmental signal, whereas they exhibited strong species‐specific responses, exceeding those of subordinate and rare species. This apparent mismatch arose from strong asynchrony and compensatory turnover among dominant species along environmental gradients. Non‐dominant species responded more synchronously to environmental variation, increasing positive covariance among species. Conclusions Our results reveal a decoupling between species‐level environmental sensitivities and abundance‐group biomass responses, demonstrating that community‐level patterns emerge from the dominance structure and covariance of species responses rather than from mean species effects alone. Biomass stability in the dominant group emerges from compensatory dynamics among species, whereas non‐dominants respond synchronously to environmental gradients and drive community‐level biomass responses. Changes within the non‐dominant biomass thus provide sensitive indicators of environmental variation.
Swacha et al. (Fri,) studied this question.