Tropical forests increasingly suffer from anthropogenic disturbances and about 82% of these ecosystems are degraded to some extent (Watson et al., 2018). In the meantime, these ecosystems constitute biodiversity hotspots. They account for nearly half of the world’s forest ecosystems and play a crucial role in the provision of many ecosystem services. In a global context of biodiversity erosion, investigating the response of these hyperdiverse ecological systems to anthropogenic disturbances is a crucial issue to predict their fate.

Our work focuses on the understanding of tree diversity trajectories in disturbed forests. We define biodiversity trajectories as the potential dynamics of tree diversity indices over time in permanent sampling plots (PSPs). We aim at modelling trajectories of tree diversity recovery in logged and secondary forests. We considered both of them as disturbed ecosystems along a disturbance intensity gradient, with various disturbance types. To assess the impact of anthropogenic disturbances on the studied PSPs, we developed a new conceptual framework characterizing the effect of disturbance intensity and type on diversity relative recovery rate following disturbances.

We worked in Costa Rica, characterized by a highly fragmented landscape, and in French Guiana, part of the Amazon basin. We used 55*PSPs with an area ≥1ha (25 in old-growth forests, 22 in logged forests, 8 in secondary forests) within the rainforest ecoregion, where trees with a DBH≥10cm were inventoried at least 5 times over 20 years. We computed both basal area values and Shannon diversity indices for each census, as data to infer a hierarchical bayesian model, retrieving both metrics trajectories. We defined disturbance intensity as the difference between the asymptotic and initial basal area value, and looked at its effect with disturbance type on diversity relative recovery rate.

We modelled tree diversity trajectories for each PSP and found that:


Diversity recovers more slowly as timber logging intensity increases
Diversity recovers more quickly as sylvicultural treatment intensity increases in Costa Rican Atlantic rainforests dominated by a single highly dynamic species
Diversity trajectories are predictable but also uncertain
Disturbance types have a different impact on diversity and biomass relative recovery rates


Bridging the gap between various disturbance types and intensities in terms of tree diversity recovery with the kind of conceptual modelling framework we developed would be the first step to integrate disturbed forest in a single landscape conservation framework, while being aware of each forest specificities.