How Many Species Are Enough? Long-Term Tree Diversity Effects on Biomass and Demography in a Restored Tropical Forest

Understanding how biodiversity shapes the recovery of ecosystem functioning is central to applying the biodiversity–ecosystem functioning (BEF) framework to forest restoration. A key question is how many tree species are required to restore biomass accumulation and carbon storage while reestablishing natural demographic dynamics in species-rich tropical systems. Here, we analysed an 18-year restoration experiment in the Atlantic Forest of Brazil with tree diversity treatments of 20, 59, and 112 native species planted under identical conditions. We assessed how richness influenced demography and stand-level aboveground biomass, and how functional traits mediated these responses. Aboveground biomass increased over time but peaked at intermediate diversity, suggesting saturation at the highest richness level. Mortality rose sharply after 11 years and was greatest at the highest diversity level, which also showed the lowest growth rates. Treatments also differed in the timing of the knee point—the onset of self-thinning, marking the transition from density-independent to density-dependent mortality—with the 112-species treatment reaching this transition 3 years earlier than the lower-diversity treatments. A trait PCA revealed two main axes: a drought-tolerance trade-off between wood density and turgor loss point, and an SLA gradient. These axes were reflected in demographic patterns, with conservative species exhibiting lower mortality and growth increasing with community-weighted wood density while declining with SLA. Functional diversity was particularly important at the highest diversity level, where greater multi-trait functional dispersion significantly reduced mortality. Our results provide new empirical evidence that increasing species richness enhances biomass accumulation and carbon storage but may reach a functional plateau at high diversity, where higher mortality and lower growth were observed. They also point to the role of functional strategies in shaping demographic performance. This long-term experiment helps clarify biodiversity thresholds for restoring ecosystem functioning in tropical forests and emphasises the importance of trait-based selection in restoration design.