Patterns of Genetic Variability in Island Populations of the Cane Toad (Rhinella marina) from the Mouth of the Amazon.

The Amazonian coast has several unique geological characteristics resulting from the interaction between drainage pattern of the Amazon River and the Atlantic Ocean. It is one of the most extensive and sedimentologically dynamic regions of the world, with a large number of continental islands mostly formed less than 10,000 years ago. The natural distribution of the cane toad (Rhinella marina), one of the world's most successful invasive species, in this complex Amazonian system provides an intriguing model for the investigation of the effects of isolation or the combined effects of isolation and habitat dynamic changes on patterns of genetic variability and population differentiation. We used nine fast-evolving microsatellite loci to contrast patterns of genetic variability in six coastal (three mainlands and three islands) populations of the cane toad near the mouth of the Amazon River. Results from Bayesian multilocus clustering approach and Discriminant Analyses of Principal Component were congruent in showing that each island population was genetically differentiated from the mainland populations. All FST values obtained from all pairwise comparisons were significant, ranging from 0.048 to 0.186. Estimates of both recent and historical gene flow were not significantly different from zero across all population pairs, except the two mainland populations inhabiting continuous habitats. Patterns of population differentiation, with a high level of population substructure and absence/restricted gene flow, suggested that island populations of R. marina are likely isolated since the Holocene sea-level rise. However, considering the similar levels of genetic variability found in both island and mainland populations, it is reliable to assume that they were also isolated for longer periods. Given the genetic uniqueness of each cane toad population, together with the high natural vulnerability of the coastal regions and intense human pressures, we suggest that these populations should be treated as discrete units for conservation management purposes