How environmental factors affect the abundance and distribution of two congeneric species of Amazonian frogs

In this study, we test the hypothesis that, at a fine scale, environmental variables influence differently sister species that live in sympatry and are phylogenetically closely related. We sampled two Amazonian anuran species, Phyzelaphryne miriamae and Phyzelaphryne sp., in 11 permanent sampling modules distributed across ~600 km in the Purus-Madeira Interfluve between 2013 and 2014. Using mixed generalized linear models, we found that the species have distinct environmental associations, which may facilitate their coexistence in sympatry. Phyzelaphryne miriamae was more frequent in environments with low precipitation and low water tables, suggesting this species is better adapted to live in drier places. In contrast, Phyzelaphryne sp. appeared to be a generalist regarding to habitat and resource use. These patterns are in accordance with the hypothesis that environmental variables influence sister species differently on a fine scale. Phyzelaphryne miriamae is larger than Phyzelaphryne sp., which may make it more resistant to dehydration, allowing it to explore drier environments. In conclusion, our results are in concordance with the hypothesis that the evolution of characteristics resulting from selection may have reduced competition for resources between closely related species, thus facilitating coexistence in sympatry.

Methods
Data were collected in 11 RAPELD permanent sampling modules (Magnusson et al., 2013) distributed along approximately 600 km in the MPI. Modules are located approximately every ~60 km along the BR-319, a federal highway that runs through the interfluve from north to south. They are positioned perpendicular to the highway and have two parallel 5-km tracks of separated by 1 km. A 250 m long plot is located every 1 km along each trail, totaling 5 plots per track and 10 plots per module. Sampling of the focal species was carried out in the rainy seasons (October–March) of 2013 and 2014 through auditory and visual sampling during their afternoon activity period (16–19h). Plots were sampled once in each of two field campaigns. Frogs were detected by visual and auditory surveys, but almost all detections of the focal species in this study were by auditory cues, which are much less affected by vegetation clutter than visual records. Each 250-m plot was divided into 10-m segments, and the presence or absence of the species was recorded in each segment, giving an abundance index varying from zero to 25.