Increased habitat availability as revealed by LiDAR contributes to the tropicalization of a subtropical butterfly community

Abstract

Species redistribution through climate change remains a global problem. However, factors such as habitat availability can complicate the attribution of species distribution shifts. We used habitat metrics derived from repeated airborne LiDAR surveys in 2010 to 2020 to examine the underlying causes for the establishment of new butterfly species in Hong Kong. For six species newly arrived since 2000, we built species distribution models using the Random Forest algorithm based on LiDAR data from 2020 to characterize species' preferred habitats across the region. Through hindcasting, we applied the model to LiDAR data from 2010 to observe any changes in the availability of preferred habitat. We found widespread vertical forest growth across Hong Kong and increased probability of occurrence based on increased habitat availability for all six species. The underlying habitat drivers, however, varied significantly across species; two species (Lethe chandica, Notocrypta paralysos) were associated with closed forest while two other species (Prosotas dubiosa, Prosotas nora) were associated with urbanicity. Our results highlight how changes in habitat can occur concurrently with climatic change and together drive the redistribution of biodiversity. Particularly for vertically complex tropical forests, airborne LiDAR data can be leveraged to observe changes in habitat complexity and how these relate to shifts in species distributions.