Limitations in the use of species distribution models for environmental impact assessments in the Amazon

Abstract

Species distribution models have been recognized as a potential tool for guiding public decisions, including in the environmental licensing process of large ventures. In this paper, we present a case study using data on frogs that were obtained from the impact assessment of a hydroelectric project located in Amazon Basin to evaluate the use of SDMs in the decisionmaking process in this context. Because conservation strategies must prioritize targets, we defined the priority species and analyzed their respective sampling situations. Based on the expectations of environmental legislation in Brazil and using the tools recommended in the literature, we discussed the limitations and potential use of SDM techniques for guiding mitigation and compensation actions. The results suggest that the lack of knowledge regarding the distribution of species poses a risk to biodiversity when potentially damaging enterprises are licensed in the Amazon. However, there were insufficient data for most of the target species to be included in the distribution models because most of those species have not yet been described. The mandatory surveys are typically conducted in areas adjacent to the affected areas, and so models must extrapolate beyond the sampled data to guide decisions, such as defining areas to be used to offset the negatives effects. The results of geographical extrapolation simulations, were corroborated by several authors who suggested that predictions can be varied when it is use different arrangements of data for calibration of the tools, since random data within the range of many Amazonian species are not available. We conclude that the use of SDMs for supporting decisions to license projects in the Amazon requires expanded sampling areas for impact studies or an investment in integrated and comparative survey strategies to improve biodiversity sampling and to fulfill the prerequisites for the use of such techniques.