Redescription of the advertisement call of Brachycephalus tridactylus (Anura: Brachycephalidae)
Background. Brachycephalus includes miniaturized frogs with restricted geographical distributions throughout the Brazilian Atlantic Forest. Ecological data for most species are still scarce. For instance, advertisement calls have only been described for 12 of the 36 known species, including B. tridactylus, a recently described species from southern Brazil. Posteriorly, features of the advertisement call of B. tridactylus were compared with congeners and the unique characteristics of its call were highlighted. To confirm these potentially divergent characteristics, we reanalysed an original recording of B. tridactylus and analysed our own recordings and verified that the original description of its advertisement call is inaccurate. Thus, we redescribe its advertisement calls. Methods. We asked the descriptors of B. tridactylus the original recordings that they made and requested access to the only original recording deposited in a collection of sounds. We received from André Lima a copy of one recording, the same as the one that had deposited, and obtained permission to re-analyze it. We studied this recording and compared it with our own recordings, made at the type locality of the species on March, 2016. Sound samples were analysed with Raven Pro 1.5.0 and call analyses were made under a note-centered approach. Results. The original recording was amplified somehow by at least 6 dB and was also clearly low-pass filtered with a cutoff frequency of 10 kHz. Our analyses did not allow us to recognize several of the acoustic parameters normally described in Brachycephalus. The sound we heard from the notes overlapped with other signals (noise?), which prevented us from clearly determining the end of the note and other important features, such as the presence of pulses. According to our recordings (n = 15 individuals), B. tridactylus emitted a relatively long advertisement call (50.8 s, on average), composed by 10–13 notes emitted in a note rate of 3.7–8.3 notes per minute. Only isolated notes were present. The notes were composed by 1–3 pulses and the note duration varied from 0.002–0.021 s. Discussion. The original description of the call of B. tridactylus is incorrect because it included background noise and amplification artefacts as part of the call parameters. However, we recognize that the original recording and our recordings have captured the same type of call. In our measurements of the species calls, note duration was nearly an order of magnitude shorter as the original description. The existence of notes with 1–3 pulses was not acknowledged in the original description. With few pulses per notes, the advertisement call of B. tridactylus is distinct from the notes with several pulses of B. ephippium, B. pitanga, B. crispus, B. sulfuratus, and B. darkside. The advertisement calls of B. tridactylus is also distinct from that of B. albolineatus and B. mirissimus by having only isolates notes, instead of isolated notes and note groups.