Applying species distribution modelling to improving conservation based decisions: a gap analysis of Trinidad and Tobago’s endemic vascular plants

Abstract

For the successful conservation of a species, habitat loss and fragmentation must be controlled through a protected area network that adequately covers its habitat. Here the suitable habitats of all of Trinidad and Tobago’s endemic plant species are determined and used to perform a gap analysis of a proposed protected area network. Data from a recently completed botanical survey, the WorldClim 2 environmental parameters, and a range of other sources were used to determine the habitat of each species using the species distribution model MaxEnt. Modelled habitat suitability for each species was combined and used to create maps showing endemic richness, weighted endemism and corrected weighted endemism, and to rank areas by conservation value using Zonation. The coverage of the proposed protected area network and a land use map were overlaid on these modelled distributions. We identified data limitations which meant that more than half of the 66 endemic species could not be modelled with confidence. For the remaining species, we found that the proposed protected area network contains just 13 ± 7% of the total modelled habitat of the endemic species. For eight endemic species > 25% of the suitable habitat is degraded. Model analysis indicated that elevation and temperature seasonality are the most important drivers of endemism. Based on a gap analysis the inclusion of high elevation areas of Trinidad’s Northern Range in the proposed protected area network would expand the coverage to include > 25% of the total modelled habitat of the endemic species, thus greatly increasing the long-term sustainability of the endemic species populations.