A new species of Brachycephalus (Anura: Brachycephalidae) from southern Brazil

Introduction

Although the first Brachycephalus was discovered in the early 19th century, half of its 35 currently recognized species have been described since 2011 (Frost, 2018). One of the main reasons for this late burst of new species descriptions is probably the relative inaccessibility of many of its populations, which tend to be restricted to high elevation regions of the Brazilian Atlantic Forest (Pie et al., 2013; Bornschein et al., 2016a). Many of these species are microendemic, being found only in one or a few adjacent mountaintops (Bornschein et al., 2016a), a factor that, together with the destruction of their habitat and their relatively low reproductive rate, makes them particularly vulnerable to extinction. Therefore, there is an urgent need to advance our understanding of the taxonomy and distribution of Brachycephalus species to ensure their long-term conservation.

Brachycephalus has been recently divided into three phenetic groups, namely the Brachycephalus ephippium, B. didactylus, and B. pernix species groups (Ribeiro et al., 2015). There are intriguing differences in how species of each group respond to altitude. For instance, species from the B. didactylus group occur from sea level to higher altitudes (0–1,110 m a.s.l.) and are more environmentally tolerant (Pie et al., 2013; Bornschein et al., 2016a). On the other hand, species from the remaining groups tend to be found at higher elevations, but can also occur locally at lower altitudes if particular microclimatic conditions are met (B. ephippium group occurs between 200 and 1,900 m a.s.l. and B. pernix group between 455 and 1,640 m a.s.l.; Bornschein et al., 2016a). Therefore, it is not surprising that extensive field work in the highlands of the southern Brazilian Atlantic Forest led to the discovery of several new species of Brachycephalus of the B. pernix group (Pie & Ribeiro, 2015; Ribeiro et al., 2015; Bornschein et al., 2016b; Ribeiro et al., 2017). As part of this continued effort, in the present study, we describe a new species of Brachycephalus from the state of Santa Catarina, southern Brazil.

Methods

Results

• Brachycephalus mirissimus sp. nov.

• Urn:lsid:zoobank.org:act:urn:lsid:zoobank.org:act: urn:lsid:zoobank.org:act:B1C3F142-E68C-4B4F-9647-F69CA6C25EB9

• Figures 1–4

Holotype. MHNCI 10793 (Figs. 1–3) male, collected at Morro Santo Anjo (26°37′41″S, 48°55′50″W; DATUM WGS84; 535 m a.s.l.), municipality of Massaranduba, state of Santa Catarina, southern Brazil, on January 15, 2018 by Marcos R. Bornschein and Larissa Teixeira.

Paratopotypes. MHNCI 10794 male, collected on January 15, 2018 by Marcos R. Bornschein and Larissa Teixeira; MHNCI 10795–9 and MHNCI 10802–3 males, collected on January 18, 2018 by Marcos R. Bornschein, Marcio R. Pie, Luiz F. Ribeiro, André Confetti, and Mário J. Nadaline; MHNCI 10800–1 females, collected on January 18, 2018 by Marcos R. Bornschein, Marcio R. Pie, Luiz F. Ribeiro, André Confetti, and Mário J. Nadaline.

Diagnosis. B. mirissimus is a member of the genus Brachycephalus based on its position in a phylogenetic tree (Fig. 5). B. mirissimus is a member of the B. pernix group, as defined by Ribeiro et al. (2015) and modified above, by having a bufoniform body shape and linea masculina (Figs. 3C and 4H). B. mirissimus is distinguished from all of the species in the genus by the following combination of characters: (1) body robust and bufoniform; (2) size SVL 9.9–11.7 mm for males and 10.0–12.9 mm for females (Table 1); (3) smooth dorsum (Figs. 3 and 4); (4) general color (in life) orange with white dots and stripe in the middle of the head and along its vertebral column (Figs. 3 and 4); (5) iris completely black (Fig. 3A); (6) advertisement call composed of note groups; (7) isolated notes with 1–3 pulses; and (8) short isolated notes (0.002–0.027 s).

Comparison between Brachycephalus. B. mirissimus is unique among the species of its genus by its distinctive white dorsal pattern in contrast with an orange body. However, other species present a dorsal stripe, as follows. B. albolineatus have a very similar white dorsal pattern, but not in all specimens and, when present, they also show a green to dark green dorsum. B. boticario, B. auroguttatus, B. mariaeterezae, B. quiririensis, B. guarani (Clemente-Carvalho et al., 2012), some B. ferruginus, and some B. verrucosus also have mid-dorsal stripes but with different coloration: yellow in B. boticario, B. quiririensis, B. auroguttatus, and B. verrucosus, light blue in B. mariaeterezae, brown in B. guarani (Clemente-Carvalho et al., 2012), and reddish-brown in B. ferruginus. The dorsum of some of these species is also different: orange in the new species, as opposed to yellow in B. mariaeterezae, brown in B. quiririensis, light brown in B. boticario, light green in B. verrucosus, brown in B. quiririensis, and pure yellow anteriorly and increasingly mixed with brown instead of yellow toward the posterior region in B. auroguttatus. Two of those species with relatively similar pattern with a contrastingly colored stripe on dorsum have a similar orange dorsum as found in B. mirissimus, that is, B. ferruginus and B. guarani (Clemente-Carvalho et al., 2012). Specimens of B. ferruginus without a reddish-brown stripe on dorsum have dispersed reddish-brown patches on middle dorsum or an entirely orange dorsum. B. mirissimus is also distinguished from some of those species by its coloration on ventral surface, being orange instead of orange with brownish-green regions in B. albolineatus, yellow in B. verrucosus, yellow with small brown spots in B. mariaeterezae, orange with small green spots in B. ferruginus, orange anteriorly and brown with orange spots posteriorly in B. quiririensis, and orange mixed with brown in B. auroguttatus. B. mirissimus is also reminiscent of some specimens of B. fuscolineatus, but nevertheless can be safely distinguished from that species by its orange body and white stripe in contrast with the yellow flanks and dark brown to black middle of the dorsum of B. fuscolineatus. The remaining species of the genus present only one uniform dorsal coloration or two dorsal colorations, but in this case without any of these colors restricted to a narrow stripe, being representatively distributed along dorsum. Beside this, other features can also be used to diagnose the species. From the remaining species of the B. pernix group, B. mirissimus is easily distinguished by the orange body with white dorsal stripe coloration instead of (1) general dorsal body color dark green with dark brown regions in B. actaeus, (2) dark brown in B. brunneus and B. curupira, (3) dark green in B. olivaceus, (4) orange (or orange and yellow), normally with dark spots on the flanks, in B. izecksohni, B. leopardus, B. pombali, and B. tridactylus, (5) orange on head and on central dorsum with black remaining parts in B. pernix, and (6) pale red from the head to the pelvic region and yellowish green on the lateral body and dorsal thighs in B. coloratus. Also, within species of the B. pernix group, the new species share the similar rough dorsum of B. actaeus, B. auroguttatus, B. boticario, B. fuscolineatus, B. leopardus, B. mariaeterezae, B. olivaceus, B. quiririensis, and B. verrucosus, but is distinct in this feature from B. fuscolineatus, that have smooth dorsum. From the remaining species of the B. ephippium group, B. mirissimus is easily distinguished by the orange body with white dorsal stripe coloration instead of (1) general color orange in B. alipioi, B. crispus (Condez et al., 2014), B. darkside (Guimarães et al., 2017), B. ephippium (Pombal, 1999), B. garbeanus (Pombal & Izecksohn, 2011), B. margaritatus (Pombal & Izecksohn, 2011), B. vertebralis, (2) general color orange with dorsal greenish spots in B. toby, (3) general color orange with reddish spots in B. pitanga, and (4) general color greenish yellow to dark gray-green with yellow warts in B. nodoterga (Clemente-Carvalho et al., 2016). B. mirissimus is also distinguishable from species of the B. ephippium group in body size of males, which is larger in some species, such as B. alipioi (SVL = 12.5–16.2 mm; Pombal & Gasparini, 2006), B. darkside (SVL = 14.8–18.5 mm; Guimarães et al., 2017), and B. margaritatus (SVL = 15.0–18.9 mm; Pombal & Izecksohn, 2011). From B. bufonoides, the new species is diagnosable by is body size (SVL 10.0–12.9 mm against 13.5–16.4 in B. bufonoides, considering both sexes; Pombal, 2010). The bufoniform body shape and light color of the body also distinguish the new species from all species of the B. didactylus group, namely B. didactylus (Izecksohn, 1971), B. hermogenesi, B. pulex, and B. sulfuratus, which have leptodactyliform body shapes and homogeneous dorsal coloration, at times with an “X”-shaped darker mark on their dorsum (Izecksohn, 1971; Giaretta & Sawaya, 1998; Napoli et al., 2011; Condez et al., 2016). We were unable to compare the new species with B. atelopoide, from Piquete, São Paulo, because this species is unknown in the wild and its type is apparently missing (Pombal, 2010).

The advertisement calls of B. mirissimus resemble those of species of the B. pernix group, as well as B. hermogenesi, from the B. didactylus group, in that they are composed of relatively short notes, as opposed to the “buzz” structure found in species of the B. ephippium group (see below; see Table S2). With isolated notes having 1–3 pulses, B. mirissimus is easily distinguishable from B. crispus (7–12 pulses per note; Condez et al. (2014)), B. darkside (5–8 pulses per note; Guimarães et al. (2017)), B. ephippium (5–15 pulses per note; Pombal, Sazima & Haddad (1994)), and from B. pitanga (6.9–13.3 pulses per note; Tandel et al. (2014); see also Araújo et al. (2012)), of the B. ephippium group, and is also easily distinguishable from B. sulfuratus (6.90–13.30 pulses per note; Condez et al. (2016)), from the B. didactylus group. The distinctiveness of the new species from those species above can also be expressed by the duration of the isolated notes, short in B. mirissimus (0.002–0.027 s; mean of 0.010 ± 0.007 s) and long in B. crispus (mean of 0.28 ± 0.02 s; Condez et al. (2014)), B. darkside (0.083–0.163 s; Guimarães et al. (2017)), B. ephippium (0.093–0.125 s; Pombal, Sazima & Haddad (1994)), B. pitanga (0.15–0.25 s; Tandel et al. (2014); see also Araújo et al. (2012)), and B. sulfuratus (0.131–0.233 s; Condez et al. (2016)). Within species of the B. pernix group, the advertisement calls of B. mirissimus are easily distinguished from those of B. tridactylus by having note groups, whereas the latter presents only isolated notes. On the other hand, we cannot distinguish the advertisement call of the new species from that of B. albolineatus (Bornschein et al., 2018), due to the general similarity, as least considering the features described to date. Finally, we also cannot distinguish the advertisement call B. mirissimus from that of B. hermogenesi (Verdade et al., 2008) and B. actaeus (Monteiro et al., 2018), but in these cases this is likely due to the simplicity of the description of the call of B. hermogenesi and the use of the call-centered approach in B. actaeus (Table S2) rather than by its supposed similarities.

Description of the holotype. Male with robust bufoniform body; head slightly wider than long; HL 40% of snout-vent length; snout short: its shape semicircular in dorsal view, and rounded in lateral view (Fig. 2); nostrils protuberant, directed anterolaterally; canthus rostralis not distinct; lips nearly sigmoid; loreal region weakly concave; eye slightly protruding in dorsal and lateral views; ED 33% of HL; tympanum indistinct; vocal sac not expanded externally; tongue longer than wide, with posterior half not adherent to floor of mouth; choanae relatively small, rounded; vomerine teeth absent. Upper arm and forearm relatively slender, upper arm approximately as long as forearm; tips of Fingers I, II and II rounded, Finger IV greatly reduced; relative lengths of fingers IV<I<II<III; subarticular tubercles and inner and outer metacarpal tubercles absent; legs short, thigh robust; THL 39% of SVL, crus length 86% of THL; toe II short but distinct, toe III distinct and toe IV long; toes I and V not visible externally; relative length of toes II<III<IV; subarticular tubercles and inner metatarsal tubercles absent; outer metatarsal tubercle distinct, large and ovoid. Skin rough on dorsum of head and central body; skin granular on dorsolateral surfaces of body, flanks, and dorsal surface of thighs, with juxtaposed, large glandular warts; sides of the body granular; large, round juxtaposed glandular warts on the sides of the body, belly and thighs; chin, arms, and legs smooth.

Coloration of the holotype. In life, almost completely orange, except for a patch on the dorsum of the head, a line in the middle of the dorsum and a pair of white patches in the distal dorsum, partly surrounded by a thin yellow margin; dorsal part of leg articulations light orange; iris black (Fig. 3). In preservative, orange regions turned yellowish pale cream and white region (1) remain white or (2) became pale cream, or (3) became pale cream with white edges (Fig. 1).

Measurements of holotype (in mm). SVL = 10.9, HL = 3.7, HW = 4.4, ED = 1.3, IOD = 2.4, IND = 1.3, EN = 0.7, SL = 0.8, UEW = 0.8, FLL = 2.8, HAL = 1.8, THL = 4.3, TL = 4.0, TSL = 2.8, FL = 3.1.

Variation in the type series. Morphometric variation is given in Table 1. There are slight differences in coloration among specimens (Fig. 4). The orange coloration has a yellowish hue in some specimens, both in the dorsal and in ventral regions. In addition, the width of the white stripe and the extent of the white spots on the dorsum of the head and on the distal dorsum also vary (Fig. 1). The latter is light yellow instead of white in some individuals.

Phylogenetic relationships. The phylogenetic analysis of species of the B. pernix species group places B. mirissimus as part a clade which includes B. fuscolineatus and B. boticario (Fig. 5), which are the southernmost species of the genus and are distributed in the region of the new species (Fig. 6).

Advertisement call. We analyzed 31 advertisement calls from 12 individuals, six of which were collected (MHNCI 10793–8). We recorded eight individuals two to five times ($\overline{x}=3.37$ times per individual). The calls we deposited resulted in 31 separate recordings (MHNCI 052–82). All the described features are shown in Table 2 and the distribution of the number of pulses per note are provided in Table 3. Some advertisement calls were not recorded from the beginning, with some notes being heard before the recordings started (Table 3). To measure the duration of the call, we also considered the advertisement calls with up to two initial notes missing in the recordings. Below, we describe the call features reporting the mean ± SD, with the range in parentheses.

Brachycephalus mirissimus emitted an advertisement call of 111.83 ± 46.60 s (37.70–255.20; Fig. 7A). An individual can remain silent for several minutes after emitting an advertisement call (occasionally for more than 20 min), when it emits a new call. Advertisement calls included 23.55 ± 10.29 notes (6–52), with a note rate of 11.69 ± 2.12 notes per minute (7.48–15.93). The advertisement calls included both isolated notes and note groups (in this case, with two notes involved in each particular note group; Table 3; Figs. 7C and 7E). Advertisement calls could be composed only by isolated notes (26.7% of the advertisement calls), but usually included both isolated notes and note groups (Table 3). Each advertisement call with note groups began with isolated notes and then changed to note groups (Table 2). Some advertisement calls began with warming notes, at least with up to three of this attenuated note (see example of one warming note in the oscillogram of the Fig. 7A—a small peak of energy just above 0). The part of the advertisement call composed of isolated notes has a duration of 63.05 ± 22.83 s (16.26–100.30) and include 13.85 ± 5.34 notes (6–25), emitted in a rate of 10.54 ± 1.59 note per minute (7.48–14.28). The part of the advertisement call composed by note groups have a duration of 47.06 ± 39.31 s (6.71–182.40) and include 14.00 ± 7.58 notes (6–38), emitted in a rate of 17.77 ± 4.66 note per minute (12.20–27.43). There are 1.69 ± 0.47 pulses per isolated notes (1–3; Figs. 7B and 7D) and 1.96 ± 0.32 pulse per note in note groups (1–3; Figs. 7C and 7E). In each note groups (two notes counting in a single value of pulses), there are 3.92 ± 0.60 pulses (2–6). A total of seven combinations of number of pulses in each note in note groups were recorded, that is, 1–1 (n = 10 note groups; n = 2 individuals), 2–2 (n = 135 note groups; n = 10 individuals), 3–3 (n = 3 note groups; n = 2 individuals), 1–3 (n = 1 note group), 2–1 (n = 2 note groups; n = 2 individuals), 2–3 (n = 1 note group), and 3–2 (n = 2 note groups; n = 2 individuals). Note duration of isolated notes is 0.02 ± 0.01 s (0.00–0.03) and note duration of note groups is 0.43 ± 0.04 s (0.36–0.59). The inter-note interval in isolated notes is 5.83 ± 1.35 s (3.92–10.62) and the inter-note group interval is 7.04 ± 1.18 s (5.32–10.93). The inter-note interval within note groups is 0.39 ± 0.03 s (0.35–0.49). The note dominant frequency is 6.66 ± 0.28 kHz (6.00–7.23). Finally, the highest frequency is 8.42 ± 043 kHz (7.26–10.06) while the lowest frequency is 4.26 ± 0.71 kHz (2.67–5.62).

Etymology. The specific epithet mirissimus is a superlative of the Latin adjective mirus, which means wonderful, marvelous.

Habitat, abundance, and distribution. We recorded B. mirissimus calling throughout the day under the leaf litter, but with more intense vocal activity in the morning and later in the day. We did not hear the species throughout the study area, as it showed a patchy distribution, and it is not homogeneously abundant in these patches. In the patch where it appeared particularly abundant, we heard 14 males in 202 m², resulting in one calling male per 14.5 m².

The species is known from the type locality (Fig. 6), where it was found in a patchy distribution between 470 and 540 m a.s.l. in montane forest (Floresta Ombrófila Densa Montana; Fig. 8) that reaches about 18–28 m in height. We did not find the species in montane forest with a lower canopy (<10 m), which was in a very steep terrain. We estimate its “actual” extent of occurrence to be 56.8 ha (calculated excluding forested areas in very steep terrains). That estimate resulted in three in-line polygons, distant from each other by 190 and 60 m, which became isolated due to replacement of the original vegetation by Eucalyptus sp., Pinus sp. and palm plantations of Archontophoenix alexandrae H. Wendl. & Drude (Fig. 9; areas encompassed by polygons are 28.3, 23.1, and 5.4 ha). The historical extent of occurrence (i.e., before the deforestation) taken the above criteria, resulted in a continuous polygon of 111.8 ha.

We recorded some species at the type locality that are typical of high altitudes, such as Quelusia regia Vell., in the case of plants, and Piculus aurulentus, Attila phoenicurus, Carpornis cucullata, and Scytalopus speluncae (taxonomy according to Maurício et al. (2010)), in the case of birds. On the other hand, we also recorded some plants and birds at the type locality that are typical of lowland habitats, for example Bathysa australis (A.St.-Hil.) K.Schum., Cecropia sp., Bactris setosa Mart, and Euterpe edulis Mart., in the case of plants, and Myrmotherula unicolor, Conopophaga melanops, Hemitriccus orbitatus, and Phylloscartes kronei, in the case of birds, showing a mixed flora and fauna from both high and low elevations.

Remarks. The type locality of B. mirissimus is 17.4 km distant in a straight line from the type locality of B. albolineatus, 18.9 km distant from the type locality of B. fuscolineatus, and 19.5 km distant from the type locality of B. boticario.

Discussion

Our phylogenetic analysis provided clear support for B. mirissimus as part of the southernmost clade of Brachycephalus, which includes B. albolineatus, B. fuscolineatus, and B. boticario (Fig. 6). In particular, B. mirissimus was placed in a clade with B. albolineatus and B. fuscolineatus with moderate support, but there was not enough signal in the analysis to uncover the relationships within this clade. Despite the close phylogenetic and geographical proximity between these four species indicated (Fig. 6), they differ considerably in coloration. On the other hand, the calls of B. mirissimus are nearly indistinguishable from the only closely related Brachycephalus species whose call has been described to date—B. albolineatus (Bornschein et al., 2018). Such low rate of evolution in prezygotic isolation mechanisms is not unexpected given that these species are allopatric and the risk of hybridization is minimized.

Brachycephalus mirissimus is the fourth species of the B. pernix group whose advertisement call has been described to date (see Garey et al., 2012; Bornschein et al., 2018; Monteiro et al., 2018). The advertisement calls of species of B. pernix group share a similar overall resemblance, including structural, temporal, and spectral patterns (see Table S2). This similarity is most apparent when comparing calls of B. albolineatus and B. mirissimus by the presence of both isolated notes and note groups. On the other hand, note groups are absent from the calls of B. tridactylus, according to our analysis (see list of examined recordings in Appendix 2, which also includes the recording used in the original description by Garey et al. (2012)). The description of the call of B. actaeus under the call-centered approach (sensu Köhler et al., 2017; Table S2) prevents us from recognizing if this species presents or not note groups.

Although the number of pulses per note in the advertisement call of B. mirissimus and B. albolineatus is not a diagnostic character, given that both species present 1–3 pulses per note, some particularities are striking. In the new species, 20.6% of the notes showed one pulse (n = 138 notes), 77.5% showed two pulses (n = 521 notes), and only 1.9% showed three pulses (n = 13 notes), while in B. albolineatus there was a similar low percentage of notes with one pulse (4.7%; n = 26 notes), a relatively reduced percentage of notes with two pulses (46.4%; n = 257 notes), but a comparatively very high percentage of notes with three pulses (48.9%; n = 271 notes). The presence of three pulses in these species presents diametrically opposite results, which may reflect a tendency of the new species to be losing the emission of notes with three pulses.

The altitudinal range of occurrence of B. mirissimus confirms the tendency of species of B. pernix group to occur at lower altitudes when at higher latitudes (Bornschein et al., 2016b). The abundance of Brachycephalus, estimated using the same methodology, revealed that B. mirissimus is much less abundant than B. curupira and B. albolineatus. We obtained the estimates of one calling individual of B. mirissimus per 14.5 m² (this study), one calling individual of B. curupira per two to three m² (Ribeiro et al., 2017), and one calling individual of B. albolineatus per 3–4 m² (Bornschein et al., 2016b). These results raise concern for the conservation of B. mirissimus because all of those estimates were made where the species appeared to be most abundant, yet its abundance is about four to six times lower than its congeners (see below).

The present extent of occurrence of B. mirissimus (56.8 ha) is among the smallest ranges of vascular plants and fishes around the world (<100 ha; Brown, Stevens & Kaufman, 1996) and is similar to other microendemic Brachycephalus from southern Brazil (Bornschein et al., 2016a). According to the criteria for classification of endangered species of the world (IUCN, 2012), B. mirissimus corresponds to Critically Endangered based on the following criteria: B2a, b(i). Evidence for adherence to this criterion includes selective harvest of trees, edge effects, and erosion around two roads that go across one of the forest fragments that constitute the extent of occurrence of the new species. We estimate that the present extent of occurrence was reduced by 53% by deforestation (Fig. 9).

Appendix 1. Examined Specimens

Brachycephalus actaeus. SANTA CATARINA: Morro do Cantagalo, municipality of São Francisco do Sul MHNCI 10829–31, and seven additional uncatalogued specimens.

Brachycephalus albolineatus. SANTA CATARINA: Morro Boa Vista, boundary of the municipalities of Jaraguá do Sul and Massaranduba MHNCI 10290 (holotype), MHNCI 10295–10300, MNRJ 90349 (all paratypes).

Brachycephalus alipioi. ESPÍRITO SANTO: Alto Castelinho, municipality of Vargem Alta MHNCI 10804–6.

Brachycephalus auroguttatus. SANTA CATARINA: Pedra da Tartaruga, municipality of Garuva DZUP 375 (holotype), DZUP 373–4, 376–85, 387–89 (all paratypes).

Brachycephalus boticario. SANTA CATARINA: Morro do Cachorro, boundary of the municipalities of Blumenau, Gaspar, and Luiz Alves DZUP 440 (holotype), DZUP 414–5, 438–9, 444–5, 459 (all paratypes).

Brachycephalus brunneus. PARANÁ: Camapuã, Serra dos Órgãos, boundary of the municipalities of Campina Grande do Sul and Antonina MHNCI 10165–74; Caratuva, Serra dos Órgãos, municipality of Campina Grande do Sul MHNCI 1919–20, MNRJ 40289–91 (paratypes), MHNCI 10175–84.

Brachycephalus coloratus. PARANÁ: Estância Hidroclimática Recreio da Serra, Serra da Baitaca, municipality of Piraquara MHNCI 10273 (holotype), MHNCI 10274–79, MNRJ 89949–50 (all paratypes).

Brachycephalus curupira. PARANÁ: Serra do Salto, Malhada District, municipality of São José dos Pinhais MHNCI 10280 (holotype), MHNCI 10281–87, 10292 (all paratypes).

Brachycephalus didactylus. RIO DE JANEIRO: municipality of Engenheiro Paulo de Frontin ZUEC 10825; Sacra Família do Tinguá, municipality of Engenheiro Paulo de Frontin ZUEC 1132–3, MZUSP 13613–20, 64810–1, 94621 (topotypes).

Brachycephalus ephippium. RIO DE JANEIRO: Parque Nacional Serra dos Órgãos MZUSP 104140–7; Vale de Revolta MCZ A–108655. SÃO PAULO: municipality of Cotia MHNCI 2611–16.

Brachycephalus ferruginus. PARANÁ: Olimpo, Serra do Marumbi, municipality of Morretes MHNCI 125, 128 (topotypes), MHNCI five specimens uncatalogued.

Brachycephalus fuscolineatus. SANTA CATARINA: Morro do Baú, municipality of Ilhota DZUP 159 (holotype), DZUP 158, 160, 401–5 (all paratypes), MHNCI two specimens uncatalogued.

Brachycephalus hermogenesi. SÃO PAULO: Picinguaba, Parque Estadual da Serra do Mar, municipality of Ubatuba ZUEC 9715 (holotype), ZUEC 9716–25 (paratypes), MHNCI 10823–25.

Brachycephalus izecksohni. PARANÁ: Torre da Prata, Serra da Prata, boundary of the municipalities of Morretes, Paranaguá, and Guaratuba CFBH 7381–2, 7384 (all paratypes), MHNCI 10835.

Brachycephalus leopardus. PARANÁ: Morro dos Perdidos, municipality of Guaratuba DZUP 274–83, MHNCI three specimens uncatalogued; Serra do Araçatuba, municipality of Tijucas do Sul DZUP 490 (holotype), DZUP 478–89, 491–2 (all paratypes).

Brachycephalus mariaeterezae. SANTA CATARINA: Reserva Particular do Patrimônio Natural Caetezal, top of the Serra Queimada, municipality of Joinville MHNCI 9811 (holotype), DZUP 372, 393–9 (all paratypes).

Brachycephalus nodoterga. SÃO PAULO: Reserva Biológica Tamboré, municipality of Santana de Parnaíba MZUSP 147711–6.

Brachycephalus olivaceus. SANTA CATARINA: base of the Serra Queimada, municipality of Joinville MHNCI 9813 (holotype), DZUP 371 (paratype); Castelo dos Bugres, municipality of Joinville MHNCI 9814–8 (paratypes); Morro do Boi, municipality of Corupá MHNCI 10288–9.

Brachycephalus pernix. PARANÁ: Anhangava, Serra da Baitaca, municipality of Quatro Barras MNRJ 17349 (holotype), CFBH 2597–8 (paratypes), MHNCI 1818–9, 3000–4 (all paratypes), MHNCI 1820, ZUEC 9433–7 (paratypes), DZUP 539–55.

Brachycephalus pitanga. SÃO PAULO: SP 125, municipality of São Luiz do Paraitinga MHNCI 10733–34; Trilha do Ipiranga 50 m from the Rio Ipiranga, Núcleo Santa Virgínia, Parque Estadual da Serra do Mar, municipality of São Luiz do Paraitinga MHNCI 10821–22.

Brachycephalus pombali. PARANÁ: Morro dos Padres, Serra da Igreja, municipality of Morretes CFBH 8042 (holotype), 8043–53 (paratypes), MHNCI 11 specimens uncatalogued.

Brachycephalus quiririensis. SANTA CATARINA: Serra do Quiriri, municipality of Campo Alegre DZUP 172 (holotype), DZUP 171, 173–6, 524–30 (all paratypes), MHNCI five specimens uncatalogued.

Brachycephalus sulfuratus. SÃO PAULO: base of the Serra Água Limpa, municipality of Apiaí DZUP 362; near Jurupará dam, municipality of Piedade MHNCI 10829–31. PARANÁ: Caratuval, near the Parque Estadual das Lauráceas, municipality of Adrianópolis DZUP 139; Corvo, municipality of Quatro Barras DZUP 150–7; Fazenda Thalia, municipality of Balsa Nova DZUP 221–4; Mananciais da Serra, municipality of Piraquara MHNCI 10302; Recanto das Hortências, municipality of São José dos Pinhais DZUP 463; Salto do Inferno, Rio Capivari, municipality of Bocaiúva do Sul MHNCI 9800. SANTA CATARINA: Morro do Garrafão, municipality of Corupá MHNCI 10826–28.

Brachycephalus toby. SÃO PAULO: Morro do Corcovado, Parque Estadual da Serra do Mar, municipality of Ubatuba MHNCI 10807–09.

Brachycephalus verrucosus. SANTA CATARINA: Morro da Tromba, municipality of Joinville MHNCI 9819 (holotype), MHNCI 9820, DZUP 464–78 (all paratypes).

Brachycephalus tridactylus. PARANÁ: Serra do Morato, Reserva Natural Salto Morato, municipality of Guaraqueçaba MHNCI 10185–89, MHNCI 10294, 10729–30 (topotypes).

Brachycephalus vertebralis. RIO DE JANEIRO/SÃO PAULO: Morro Cuscuzeiro, Núcleo Picinguaba of the Parque Estadual da Serra do Mar and Parque Nacional da Serra da Bocaina, boundary of the municipalities of Parati, Rio de Janeiro state, and Ubatuba, São Paulo state MHNCI 10810–20.

Appendix 2. Examined Recordings

Brachycephalus albolineatus. SANTA CATARINA: Morro Boa Vista, boundary of the municipalities of Jaraguá do Sul and Massaranduba MHNCI 001–34.

Brachycephalus tridactylus. PARANÁ: Serra do Morato, Reserva Natural Salto Morato, municipality of Guaraqueçaba MHNCI 035–51, FNJV 0032950.

Supplemental Information