Interactive persistent effects of past land-cover and its trajectory on tropical freshwater biodiversity

Abstract

Evidence indicating that ecological communities show delayed responses to environmental change has raised the need to better understand the effects of landscape history on biodiversity. We investigated how freshwater biodiversity is related to both recent and past land use change in tropical river catchments. More specifically, we analysed the relationship between biodiversity and change in forest cover that occurred across five decades, including landscape trajectories of forest gain and loss. The abundance of stream insects showed a more delayed response to landscape change than rarefied richness, whereas the Tsallis diversity index was not related to past forest cover. However, both alpha diversity and local abundance were related to mean forest cover across the five decades when conditioned by the trajectory of change. We found a negative relationship between the historical mean of forest cover and community descriptors in streams embedded in landscapes on a trajectory of forest loss, but a positive relationship in landscapes on a trajectory of forest gain. We provide the first evidence of delayed responses of tropical freshwater insect communities to landscape history and suggest that the magnitude of change in forest cover, mediated by its trajectory, is a major driver of delayed responses. We show that a trajectory of forest loss might result in both immediate loss of species and a set of species doomed for future extinction. Synthesis and applications. Delayed responses of multispecies communities to past landscape changes depend on their extent, frequency and intensity. Our findings indicate that, beyond the relative amount of forest cover within river catchments, the magnitude of change in forest cover mediated by its historical trajectory can be a major driver of delayed responses in stream communities. We suggest that, if biodiversity conservation is the main aim, stream restoration should be prioritized in catchments with higher forest cover, as even under a forest gain trajectory, stream communities continue responding to past landscape changes depending on how widespread and intensive changes were.