Summary

1.Effective assessment of the success of ecological restoration projects is critical in justifying the use of restoration in natural resource management as well as improving best practice. One of the main goals of ecological restoration is the recovery of ecosystem function, yet most researchers assume that increasing species and or functional diversity equates with restoration of ecosystem function, rather than empirically demonstrating these mechanistic relationships.

2.In this study we assess how dung beetle species diversity, community composition, functional diversity and ecological functions vary along a restoration chronosequence and compare restored areas with reference (rainforest) and degraded (pasture) systems. We also directly investigate the dung beetle diversity – ecosystem functioning relationship in the context of ecological rainforest restoration by testing the predictive power of traditional taxonomic indices and functional diversity metrics for functionality.

3.Species richness, abundance, biomass and functional richness all increased with restoration age, with the oldest restoration sites being most similar to rainforest; whereas functional evenness and functional divergence decreased with restoration age. Community composition in the restored areas was clearly progressing towards the rainforest sites and deviating from the pasture sites with increasing restoration age.

4.Secondary seed dispersal rates increased with restoration age, but there was only a weak positive relationship between dung removal and soil excavation and restoration age. Biodiversity metrics explained 47–74% of the variation in functions mediated by dung beetles; however, functional trait-based indices provided greater explanatory power of functionality than traditional species-based metrics.

5.Synthesis and applications. Our results provide empirical evidence on the potential of tropical forest restoration to mitigate biodiversity losses, recovering not only faunal species diversity, but also functional diversity and ecosystem functions in a relatively short period of time. We also demonstrate that functional trait-based metrics are better predictors of functionality than traditional species-based metrics but that the relationship between restoration age, diversity and ecosystem functioning is not straightforward and depends on the functions, traits and metrics used.