• aquatic insects;
  • community composition;
  • forest canopy;
  • microcosm;
  • mixed-effects model;
  • resource concentration;
  • structural equation modelling


1. Species richness is typically positively correlated with ecosystem size, yet there is no general consensus on the proximate mechanisms (resource concentration, disturbance, colonization–extinction dynamics or habitat heterogeneity) driving this relationship. Ecosystem-size effects are often attributed to increasing resource concentration, but the inherent intercorrelation of resource concentration with other potential proximate mechanisms has led to strong debate over its significance as a mediator of ecosystem-size effects.

2. We disentangle the proximate mechanisms underlying ecosystem-size effects on species richness by experimentally reversing resource concentration and enhancing drought disturbance, while holding colonization–extinction dynamics and habitat heterogeneity constant, in field microcosms.

3. Contrary to theory and much empirical evidence, species richness decreased with increasing ecosystem size, due explicitly to experimental manipulation of the resource-concentration gradient. Structural equation modelling revealed that resource concentration was the principal driver of ecosystem-size effects on species richness, while drought disturbance and habitat quality strongly determined the identity and composition of colonizing species.

4. Our results support the logical contention that ‘ecosystem size’ is not a mechanism of effect in its own right, and that with appropriate experimental manipulation it is possible to tease apart the multiple underlying proximate drivers of ecosystem-size effects on species richness.

5. Our results imply that the universally accepted relationship between ecosystem size and biodiversity can be reversed by nutrient enrichment, an increasingly observed human-induced driver of global environmental change.