Delayed recovery of seed-dispersal interactions after deforestation.

Deforestation has reduced the area of old-growth tropical forests, making restoration essential to recover biodiversity. Because most tropical plants rely on seed dispersal by animals, plant-animal interactions are crucial for natural forest recovery. Yet, the recovery time of seed-dispersal interactions is still unknown and likely depends on both local and landscape-scale factors, such as habitat suitability and connectivity to the surrounding forest. We conducted an empirical study in 62 plots along a 40-year chronosequence of forest recovery in the Ecuadorian Chocó, covering plots in active agricultural land, recovering forests, and old-growth forests, differing in their connectivity to the surrounding forest. In each plot, we recorded interactions between fleshy-fruited plants and frugivorous animals and measured key functional traits related to seed-dispersal interactions. Using a Bayesian model, we estimated the recovery times for the functional diversity of plants, animals, and their interactions contingent on patch connectivity. Due to the presence of remnant trees, plant functional diversity was similarly high as in old-growth forests shortly after deforestation. At mean levels of patch connectivity, animal functional diversity required 40 years to recover, and seed-dispersal interactions recovered after 19 years. The functional diversity of interactions needed more than 30 years to recover at low levels of patch connectivity. Our results show that seed-dispersal interactions can require several decades to functionally recover, depending on the connectivity to the surrounding forest. This finding provides a new benchmark for the timing of tropical forest restoration that critically depends on re-establishing functionally diverse seed-dispersal interactions.