Mapping the invasion risk of alien fishes in the Upper Paraguay River Basin

Abstract

Understand biological invasion patterns of invasive alien species (IAS) and their effects on the ecosystem changes is of great importance faced to global conservation goals. Unfortunately, research efforts have not been enough to evaluate the potential risks of complex ecosystems to the IAS. Here, I generate species distribution models to predict invasion risk of the main IAS fishes introduced in the Upper Paraguay River Basin (The Pantanal wetland and their upstream tributaries). I collected occurrences data through the Citizen Science approach, by using semistructured interviews with local fishermen and geotagged photos of introduced fishes from social media Instagram. With the records obtained and using a set of freshwater hydroclimatic variables for the basin, I modeled the potential invasion area for the alien fishes using an ensemble forecasting approach. I evaluate the models using the True Skills Statistic metric (TSS) and select the models that showed better performance for mapping the habitat-suitability for each IAS fishes (TSS≥ 0.5). Next, I compare the habitat-suitability among several areas of the basin - between the highlands and the lowlands regions, among the different sub-basins, among rivers of different stream-order, and amid in the different reservoir’s sizes, to predict accurately the areas under highest invasion risk. I found that Apa, Miranda and Aquidauana sub-basins in the highland region, such as rivers of stream-order 4, 5 and all reservoirs have optimal Habitat-Suitability conditions, i.e. these areas have a higher invasion risk of IAS fishes than the other areas modeled in the basin. The predictive models developed here can be useful to support effective preventative actions and planning the management of the IAS fish species in the Upper Paraguay basin.