The best way to reduce problems related to invasive species is by preventing introductions into potentially susceptible areas. The purpose of this study was to create distribution models for the invasive gastropod Achatina fulica Bowdich, 1822 in South America in order to evaluate its potential geographic distribution and identify areas at potential risk. This mollusc, considered one of the 100 world's worst invasive alien species, is the focus of intense concern due to its impact on agriculture, human health, and native fauna. We tested two commonly used ecological niche modeling methods: Genetic Algorithm for Rule-Set Prediction (GARP) and Maximum Entropy (MaxEnt). Models were run with occurrence points obtained from several sources, including the scientifi c literature, international databases, governmental reports and newspapers, WorldClim bioclimatic variables, and altitude. Models were evaluated with the threshold-independent Receiver Operating Characteristic (ROC) and Area Under the Curve (AUC). Both models had consistent performances with similar areas predicted as susceptible, including areas already affected and new potentially susceptible areas in both tropical and temperate regions of South America.

Dinoflagellates of the genus Ceratium are predominantly found in marine environments, with a few species in inland waters. Over the last decades, the freshwater species Ceratium hirundinella and Ceratium furcoides have colonized and invaded several South American basins. The purpose of this study was to create a distribution model for the invasive dinoflagellate C. furcoides in South America in order to further investigate the basins at potential risk, as well as the environmental conditions that influence its expansion. This species is known to develop blooms due to its mobility, resistance to sedimentation, and optimized use of resources. Although nontoxic, blooms of the species cause many problems to both the natural ecosystems and water users. Potential distribution was predicted by using a maximum entropy algorithm (MaxEnt). Model was run with 101 occurrences obtained from the scientific literature, and climatic, hydrological and topographic variables. The developed model had a very good performance for the study area. The most susceptible areas identified were mainly concentrated in the basins between southeastern Brazil and northeastern Argentina. Besides already affected regions, new potentially suitable areas were identified in temperate regions of South America. The information generated here will be useful for authorities responsible for water and watershed management to monitor the spread of this species and address problems related to its establishment in new environments.

No abstract