We examined differences in species richness between reference and impacted sites to illustrate the extent to which estimated impacts of land-use change on biodiversity can be affected by the degree of baseline suitability (intactness of reference sites) and the species assemblage under consideration. We mist-netted birds at five continuous Amazonian forest sites and 33 land-bridge forest islands (0.63–1699 ha) within a large hydroelectric reservoir. We then produced a gradient of baseline suitability based on forest area of five sets of reference sites, namely continuous forest, 1000 ha, 500 ha, 250 ha and 100 ha, and contrasted these with all smaller islands combined considering two types of species assemblages. The first comprised only species captured at reference sites (baseline species assemblage), whereas the second comprised all species captured at all sites (overall species assemblage). We also examined biodiversity complementarity to define the minimum set of forest islands retaining the most number of species occurring both at continuous forest sites and across all sites. A focus on the baseline species assemblage from the most suitable baseline (continuous forest) resulted in an estimated decrease of 67% in species richness at impacted sites. In contrast, a focus on the overall species assemblage along with the use of the least suitable baseline (100 ha) as a reference condition reversed this trend, resulting in an estimated increase of 43% in species richness at impacted sites. We therefore underline the imperative of considering the intactness of reference sites to accurately assess the impacts of land-use change on biodiversity and establish conservation strategies.
- Environmental impact assessment
- Habitat fragmentation
- Habitat loss
- Intact forest landscapes
- Shifting baseline