Background
The flea toad Brachycephalus sulfuratus was recently described from southeastern and southern Brazil. In its description, the authors overlooked previous records of flea toads that had been identified as “Brachycephalus sp. nov.” and B. hermogenesi occurring in the same regions, which could suggest the possibility of up to three flea toads coexisting in southern Brazil. In addition, B. sulfuratus is characterized by substantial phenotypic variability, to an extent that compromises its current diagnosis with respect to its congener B. hermogenesi. Therefore, the current state-of-affairs regarding the geographical distribution of these two species and the identification of previously known populations is hitherto uncertain. Our goals are to reassess previous records of flea toads attributable to B. hermogenesi, B. sulfuratus and “Brachycephalus sp. nov.”, considering the description of B. sulfuratus, and to review the diagnosis of B. sulfuratus.
Methods
A critical analysis of the species identity of flea toad specimens attributable to B. hermogenesi, B. sulfuratus, or to a potentially undescribed species from southeastern and southern Brazil was based either on the analysis of morphology or on their advertisement calls. These analyses include our independent examinations of specimens and, when not possible, examinations of published descriptions. To allow for a consistent comparison of advertisement calls between B. hermogenesi and B. sulfuratus, we made recordings of both species, including in the type locality of the former.
Results
We found that morphological and call characters originally proposed as diagnostic for B. sulfuratus in relation to B. hermogenesi vary intraspecifically. Live individuals with ventral yellow spots correspond to B. sulfuratus; individuals without yellow spots can be either B. sulfuratus or B. hermogenesi. In preservative, they are indistinguishable. Previous records of Brachycephalus sp. nov. correspond to B. sulfuratus. We propose that the reduced number of notes per call and the presence of only isolated notes in the call of B. sulfuratus, as opposed to a high number of notes per call with isolated notes and note groups in the call of B. hermogenesi, as the only diagnostic characters between them. Regarding their distributions and based in our assessment, only B. sulfuratus occurs in southern Brazil, without any overlap with B. hermogenesi. There is a narrow gap between the distributions of these species around the southeast of the city of São Paulo. Our revision also revealed that some records previously attributed to B. hermogenesi in Rio de Janeiro and north São Paulo represent a distinct, unidentified flea toad that is not B. sulfuratus. Both species occur side by side in Corcovado, São Paulo, a locality from where five paratypes of B. hermogenesi were obtained. Biogeographic events that might have led to vicariance between B. hermogenesi and B. sulfuratus are discussed.