Quantifying the Impacts of Defaunation on Natural Forest Regeneration in a Global Meta-Analysis

Background

Intact forests provide critical ecosystem services, such as carbon storage and climate regulation, which are heavily dependent on interactions between woody vegetation and animal species. However, defaunation is altering these interactions at a global scale. While previous studies have explored the effects of defaunation on individual ecosystems, a comprehensive global analysis quantifying its impact on natural forest regeneration has been lacking. This study conducts a meta-analysis to assess how vertebrate loss influences forest regeneration across multiple regions and ecosystems, identifying key taxonomic groups and ecological processes most affected by defaunation.

Goals and Methods

The researchers conducted a systematic review and meta-analysis, synthesizing data from 43 studies covering 41 forest landscapes in 27 countries. The study compared forests with varying levels of vertebrate abundance, assessing the effects of defaunation on forest regeneration by examining changes in tree recruitment, seed dispersal, seed predation, and overall biomass accumulation. The authors categorized the impact of defaunation based on the type of animal lost (e.g., primates, birds, ungulates) and the disrupted ecological process (e.g., seed dispersal, herbivory). 

Conclusion

The meta-analysis revealed that defaunation significantly reduces forest regeneration in real-world contexts, particularly due to the loss of seed dispersers such as primates and large birds. In observed studies, vertebrate loss led to declines in tree recruitment and shifts in species composition, with large-seeded, high-carbon tree species being replaced by smaller-seeded or abiotically dispersed species. Interestingly, manipulated experiments that excluded specific animal groups showed mixed results—while removing herbivores sometimes increased seedling survival, the overall effect of defaunation remained negative for long-term forest health.