Allobates flaviventris, Melo-Sampaio, Paulo Roberto, Souza, Moisés Barbosa De & Peloso, Pedro Luiz Vieira, 2013
publication ID |
https://doi.org/ 10.11646/zootaxa.3716.3.2 |
publication LSID |
lsid:zoobank.org:pub:FAEA2F7D-2517-4B8F-B977-009C9FAD9AE6 |
DOI |
https://doi.org/10.5281/zenodo.6154323 |
persistent identifier |
https://treatment.plazi.org/id/EE3587BC-FF8C-F147-19E9-814FE81EBC49 |
treatment provided by |
Plazi |
scientific name |
Allobates flaviventris |
status |
sp. nov. |
Allobates flaviventris View in CoL sp. nov.
Figures 1–2 View FIGURE 1 View FIGURE 2
Colostethus marchesianus (Cardoso & Souza, 1996 part) Colostethus sp. A—(Köhler & Lötters, 1999 part p. 263) Colostethus sp. 4—(Amézquita et al. 2006 p. 1878, 1881) Allobates sp.—(Souza et al. 2008 part p. 53)
Portuguese name: rãzinha ripária de barriga amarela English name: yellow-bellied stream frog
Spanish name: rana de quebrada de barriga amarilla
Holotype. UFAC-RB 4650 an adult male, collected on May 16, 2010 by Paulo R. Melo-Sampaio and Auristo da Conceição Melo, at Colocação Olho D´água, Fazenda Bonal, Senador Guiomard, state of Acre, Brazil (09º53’45.1”S; 67º18’14.8”W), aproximately 150 m. a.s.l.
Paratypes. UFAC-RB 4599–4601, females (FEC on January 3, 2010); UFAC-RB 4603, female (type –locality on April 2, 2010); UFAC-RB 4633–4635 (type locality on May 0 8, 2010); UFAC-RB 4640–4641 (FEC on January 23, 2010); UFAC-RB 4649 (same data as holotype); UFAC-RB 4657, 4659–4660, 4670 (all from type locality on June 4, 2010); UFAC-RB 4675-4677; UFAC 4602, 4604, 4631–4632, males, (type locality on May 8, 2010); UFAC-RB 4658, 4661–4666, 4669 (all from type locality on June 4, 2010); UFAC-RB 4671 (FEC on June 6, 2010); UFAC-RB 4678 (FEC on June 12, 2010). All specimens were collected by the senior author and field assistants (see acknowledgements).
Diagnosis. The new species is diagnosed by the following combination of characteristics: median lingual process absent; canthus rostralis rounded in both lateral and dorsal profile. Medium-sized species, mean SVL of males 18.8 mm (range 16.7–19.7); mean SVL of females 20.4 mm (range 19.3–21.1). Dorsolateral stripe absent in both sexes; ventrolateral stripe represented by irregular spots in males, but present in females; short, diffuse oblique lateral stripe present only in inguinal region in both sexes. Males with gray to violet-gray throat and golden-yellow belly; females with yellow throat and golden-yellow belly. Finger III of males not swollen in both sexes ( Fig. 1 View FIGURE 1 and 2 View FIGURE 2 ), disc on finger III wider than diameter of finger; basal webbing absent on hands. Webbing absent between toes III and IV; relative toe length IV> III> V> II> I. Advertisement call consists of groups of 2–10 notes separated by irregular intervals; dominant frequency 3617.6–4651.2 Hz.
Comparison with other species. The new species is compared here with all Allobates species. The new species differs from putatively aposematic species in the genus, i.e., with red, yellow or orange color in thighs or inguinal blotches or white stripes by lacking such color patterns (vs. presence in A. femoralis , A. hodli , A. myersi , A. rubriventris and A. zaparo ). From non-Amazonian species, the species differs, by presence of yellow venter (vs. absence) from A. alessandroi , A. algorei , A. bromelicola , A. caribe , A. chalcopis , A. craspedoceps , A. fratisenescus , A. goianus , A. granti , A. humilis , A. ignotus , A. juanii , A. mandelorum , A. mcdiarmidi , A. picachos , A. pittieri , A. olfersioides , A. sanmartini and A. spumaponens ; from Allobates insperatus and A. niputidea , two non-aposematic species, by larger body size (13.6–18.4 and 15.7–18.0 mm vs. 16.7–21.1 mm respectively); from A. kingsburyi by smaller size (16.7–21.1 mm vs. 22.3–26.2 mm); from the remaining non-aposematic species, by third finger of males swollen in A. humilis , niputidea , mandelorum , ornatus, picachos , and wayuu (non-swollen in the new species). Because most species of Allobates are known to be microendemic and their ranges do not normally appear to extend beyond biogeographic boundaries (Simões et al. 2013), a more detailed comparison is here presented only for non-aposematic species from the Amazon region. Allobates flaviventris differs from A. brunneus by its larger SVL in females, the absence of a gray to violet-gray throat (light color in A. brunneus ) and advertisement call (long trains of single notes emmited in bouts in A. flaviventris ); A. caeruleodactylus in not having sky blue digits on males, and larger size (SVL 14.9–17.4 mm; uniform dorsum and sky blue digits on A. caeruleodactylus ). A. cepedai in having finger I longer than finger II, by not having swollen third fingers in males and dorsal pattern (same size of fingers I and II, and swollen third finger in males A. cepedai ); A. conspicuus by its larger SVL in both sexes (16.5 mm in males and 17.2 mm in females of A. conspicuus vs. mean 18.8 mm in males and 20.4 mm in females of A. flaviventris ), and by the absence of dorsolateral stripes (present in A. conspicuus ); from A. crombiei its coloration in life (venter, including thorat, of males entirely yellow in A. crombiei Lima et al. 2012), and by its advertisement call (a long series of notes—25–59—in Allobates crombiei with frequency modulation, dominant frequency 4522–5383Hz, emitted at a high and regular rate: Lima et al. 2012); A. fuscellus by the absence of a swollen third finger (swollen in A. fuscellus ); Allobates gasconi by the absence of a swollen third finger (swollen in A. gasconi ); Allobates grilissimilis by its dorsal pattern (uniform in A. grilissimilis ), and by its advertisement call (trills of short pulses emitted in a variable number in A. grilisimilis ); Allobates marchesianus by its advertisement call (number of notes per calls,dominant frequency andnote duration consisting in thrill of notes in A. marchesianus ); A. masniger by the absence of dorsolateral stripes (present in A. masniger ) and throat color (black in A. masniger ); A. melanolaemus by its smaller SVL and advertisement call (1–6 notes with frequency ranging 3840–4560 Hz in A. melanolaemus Grant & Rodriguez 2001 ); A. nidicola by absence of dorsolateral stripes (present in A. nidicola ), throat color (black in A. nidicola ), in their advertisement call (long trains of single notes emmited in bouts, dominant frequency of 3440–5010Hz) and by its free-swimming tadpole (nidicolous tadpole in A. nidicola ); A. paleovarzensis by absence of swollen third finger (swollen third finger in A. paleovarzensis ) and transparent egg capsules in clutches (opaque gelatinous capsules in Allobates paleovarzensis ); A. subfolionidificans by the presence of hour-glass on dorsum, larger SVL, throat coloration (dorsolateral and ventrolateral stripes absent, males uniform white venter and females cream uniform venter), and calls (call structure with continuous and similar notes in A. subfolionidificans ); Allobates sumtuosus by not having swollen in third finger (swollen in A. sumtuosus ), ventral coloration dark throat in males A. flaviventris (light in A. sumtuosus ); A. trilineatus in not having swollen fingers (third and sometimes second finger swollen in adult males of A. trilineatus ; see Figure 4 View FIGURE 4 ) and by its advertisement call (multipulsed call, 9–13 pulses/call, dominant frequency 4920–6040Hz); A. vanzolinius by smaller SVL and by absence of swollen third finger (swollen in A. vanzolinius ). At the type locality another unnamed species ( Allobates aff. conspicuus ) is distinguished from Allobates flaviventris by the absence of an hourglass pattern on the dorsum, light yellow throat, uniform belly, and a smaller SVL.
Character Holotype Females (n = 16) Males (n = 14) SVL 18.1 20.4 (±0.51) 18.8 (±0.88) 19.3–21.1 16.7–19.7 HL 5.7 6.1 (±0.32) 5.7 (±0.20) 5.7–6.5 5.4–6.1
HW 5.7 6.4 (±0.18) 5.8 (±0.15) 6.1–6.7 5.5–6.1
IOD 3.9 4.1(±0.07) 3.9 (±0.18) 4.0–4.2 3.6–4.2
ESD 3.0 3.40 (±0.23) 3.1 (±0.22) 2.9–3.7 2.7–3.5
IND 2.9 2.89 (±0.16) 2.6 (±0.15) 2.5–3.1 2.4–2.9
ED 2.4 2.80 (±0.09) 2.5 (±0.18) 2.6–2.9 2.2–2.8
TYM 1.1 1.23 (±0.17) 1.1 (±0.08) 1.0–1.5 1.0–1.3
ARM 3.9 4.64 (±0.33) 4.4 (±0.38) 4.0–5.0 3.6–4.9
HAND 4.9 4.85 (±0.22) 4.8 (±0.14) 4.3–5.1 4.6–5.1
WDFIII 0.4 0.53 (±0.03) 0.4 (±0.04) 0.5–0.6 0.4–0.5
TL 9.0 9.89 (±0.14) 9.4 (±0.28) 9.6–10.0 8.7–9.7
FL 8.7 9.30 (±0.45) 9.0 (±0.35) 8.4–9.7 8.4–9.5
WDTIV 0.6 0.65 (±0.07) 0.6 (±0.04) 0.5–0.8 0.5–0.7 Description of holotype. An adult male, 18.1 mm SVL (additional measurements in Table 1), slender body, head width equal to head length, head width 31.4% of SVL; snout blunt, broadly rounded to nearly truncate in dorsal view ( Figs. 1 View FIGURE 1 C) and acutely rounded in lateral view ( Fig. 1 View FIGURE 1 A); snout 21.5% of HL; internarial distance 50% of head width, END 1.25 times greater than ED. Nostrils lateral, slightly protuberant, opening posterolateral. Iris tan with black reticulations. Tympanum round, posterolaterally directed; tympanic membrane inconspicuous, tympanic annuli 46.9% of ED. Tongue nearly twice as long as wide, attached anteriorly, fringed along posterior margin; median lingual process absent; vocal sac and vocal slits present; choanae small placed anteriorly; maxillary teeth present.
Skin on dorsum granular, granulation weaker on head than dorsum; skin granular on dorsal surface of legs; skin smooth on ventral surfaces. Dorsolateral stripe absent; ventrolateral stripe diffuse consisting of small spots near upper lip and becoming more intense posterior to tympanum and reaching the groin. Forearm slender, 88.3% of upper arm; ulnar fold absent; finger I longer than finger II when appressed; relative finger length III>I>II>IV ( Fig. 3 View FIGURE 3 ); basal webbing between fingers absent; subarticular tubercle in finger III absent; finger III not swollen; disc width of finger III 0.5 mm ( Fig. 3 View FIGURE 3 ).
Hindlimbs relatively robust; tibia length (49.7%) of SVL; foot length (47.9%) of SVL length. Webbing between toes absent. Relative toe length IV> III> V> II> I ( Fig. 2 View FIGURE 2 ). Toes without fringes; small fold between the first subarticular tubercle of 5th toe and outer metatarsal tubercle. Subarticular tubercles of 3rd toe almost imperceptible ( Fig. 3 View FIGURE 3 ). Outer tarsal tubercle round. Inner tarsal tubercle ovoid; discs of all toes expanded; Discs of all toes expanded except Toe V (slightly enlarged). One subarticular tubercle on Toes I and II; two on Toes III and V; three on Toe IV. Basal subarticular tubercle on Toe IV absent ( Figure 3 View FIGURE 3 ).
Color of adults in life. Color of dorsum usually varies from light grayish brown to brown with pale cream ( Fig. 1 View FIGURE 1 A, 1C and 1D). Hourglass markings brown, sometimes diffuse, diamond, or triangular, extending from between orbits to sacral region. Surface of upper arm pale cream-tan; upper surfaces of legs light brown with dark brown cross bands on thigh, legs, and feet in many specimens. Adult males with throat violet-gray with evenly dispersed melanophores on vocal sac; chest and belly golden yellow; dark line sometimes present on lower lip; adult female with center part of throat and chest pale yellow; belly cream, yellow, or golden yellow. Lateral and ventral surface of thigh golden yellow; posterior surface of legs and feet light gray, internal surface of the arm golden yellow. Oblique lateral stripe pale gray to pale brown, forming several diffuse areas extending from groin to middle of body or sometimes insertion of arm; ventrolateral band consists of series of white spots, irregular, elongated, extending from anterior corner of eye to groin. Iris bronze centrally with metallic gold edges with black reticulations; pupil uniform black.
Specimen UFAC-RB 4671 has bad-defined hourglass pattern on the dorsum. The dorsum had a greenish color that formed a continuous design that began in the interorbital region and reached the sacral region, where it became faded. All males have dark spots in the axilla. All specimens have a “W” between the orbits formed by the anterior edge of the hourglass pattern. In some individuals the hourglass pattern is almost imperceptible, but have dorsolateral stripes that bifurcate on the head toward laterally to center forming reentrance that continues as stripes and end in the sacrum. The hourglass shape is observed only with colors in life in these specimens.
Color of adults in preservative. Dorsum light brown with darker brown hourglass pattern that ends at the level of sacrum; lateral portion of body dark chocolate. Dorsal surfaces of the forearm ranging from cream to brown. Throat and chest gray in males and cream in females. The limbs are lighter than body. Forelimbs lighter than hindlimbs. Small dark spots on the knee and elbow. Belly and forelimbs light cream. Dark brown transverse stripe present on the tibia and thighs in all specimens. Lower lip with dark spots in males. Cloacal area dark chocolate in all males, and sometimes restricted to inferior portion in any females (UFAC-RB 4649, 4675). White ventrolateral stripe diffuse in males, absent in females. Oblique lateral stripe present in all females, most perceptible in lighter specimens.
Variation in the type series: In preservative, specimens UFAC-RB 4601, 4671, 4664 and 4650 have a nearly uniform brown dorsum except for a slight darkening in the central region of the back where the hourglass pattern is present in live specimens. Specimens UFAC-RB 4665, 4632, 4659, 4675, 4678, 4669, 4657, 4670, 4635, 4649, 4631 and 4663 have light cream stripes on the body. All examined males of new species show third finger not swollen (see Fig. 2 View FIGURE 2 ). Specimens UFAC-RB 4666, 4600, 4634, 4677, 4632, 4658 and 4670 have the typical hourglass pattern. Sexual dimorphism is present in many characters with females slightly larger than males [Table 3 (t = 5,22, df = 28, P = 0,000015)]. Average female SVL is 1.4 mm greater than male SVL, and females have relatively wider heads [head wider than long; male HW 31.1% of SVL in males, 31.4% of SVL in females (t = 5.09, df = 28, P = 0.000021) see Table 2]. Snout blunt, broadly rounded to nearly truncate in dorsal view and acutely rounded in lateral view, extending past lower jaw ( Figs. 1 View FIGURE 1 B and 1E). Male SL 54.2% and female END 55.3% of HL; IND in males 45.5% and females 43.7% of HW; END in males 53.1% and in females 73% of HW; tympanic membrane inconspicuous, round, TYM 45% of ED in both sexes.
Advertisement call. Call description is based on three recordings (two of the holotype UFAC-RB 4650, recorded 7 minutes apart from each other; and one of a paratype UFAC-RB 4678). The call of Allobates flaviventris consists of groups of 2–10 notes separated by irregular intervals, both between notes in the same call (0.034– 0.250 s), and between calls (0.2– 11.4 s). Call rate varied among the three recordings from 18 to 37 calls / min. Although our observations in the field suggest a difference between warm-up calls and “regular” calls, at least in some individuals, it is difficult to ascertain differences in these calls. Two recordings of the same individual (UFAC-RB 4650; taken 7 minutes apart) show conflicting evidence on whether the calls are consistently emitted in bouts of if they are emitted sporadically. The first recording (UFAC-RB 4650-A) show calls were emitted with irregular intervals, while the second recording (UFAC-RB 4650-B) shows well structured bouts ( Fig. 5 View FIGURE 5 ) with two bouts consisting of warm-up calls with fewer notes (2–4 notes) and several “regular” calls (4–6 notes). Duration of bouts was respectively 19.4 s and 17.0 s for the first and second recorded bouts. UFAC-RB 4678 does not show evident bouts (although it is a very short recording, 22 s total).
The call of a specimen from Catuaba (not collected) also showed a clear bout structure, with two bouts, 14.8 and 13.2 s of duration. Different from what was observed in the other recordings, most calls (96.3%) have only two notes, whereas the other calls have either three or four notes.
Since most calls are apparently unstructured regarding bouts, summary statistics ( Fig. 5 View FIGURE 5 ) for the calls are all calculated as if there is no bout structure. This certainly affected temporal parameters such as interval between calls, but was judged to be the best fit for the data in hand. Additional calls recorded for longer intervals will help clarify this issue. Spectral parameters show some variation among and within individuals, with dominant frequency values ranging from 3617.6–4651.2 Hz. A large number of harmonics are detectable in the two recordings of the holotype (UFAC-RB 4650; up to 9 harmonics) but fewer are detectable in UFAC-RB 4678 (up to 6 harmonics). Dominant frequency varied from 3617.6–4651.2 Hz. In the two recordings of UFAC-RB 4650 the dominant frequency is not the fundamental frequency, whereas the same pattern was not detected in UFAC-RB 4678. Summary of acoustic parameters are given in Fig. 5 View FIGURE 5 . See also additional detail in Amézquita et al. (2006) [ Fig. 3 View FIGURE 3 p. 1878] who comments on the call of the species, labeled there as Colostethus sp. 4.
Etymology. From the Latin flavo meaning yellow and ventris meaning ventral surface or belly. The specific name refers to the golden-yellow bellies in both sexes.
Natural history. Allobates flaviventris inhabits open forest with bamboo near streams. Similar to other Allobates in the region, individuals are most active early in the morning and late afternoon. Males usually call between 0630–0930 h, and again from 1630 h until dusk (GMT-5h). On cloudy days, the males can be heard calling throughout the day. The males usually call from leaf litter on the forest floor ( Fig. 6 View FIGURE 6 ), or sometimes elevated a few centimeters from the ground. Vocalization and reproductive activity occur throughout the rainy season. At the type locality and vicinities, the rainy season extends from November to May with annual average precipitation of 2000 mm.
During sampling at the type locality a single clutch containing 13 embryos surrounded by transparent gelatinous capsules was found on the forest floor in high humidity conditions ( Fig. 6 View FIGURE 6 ) into ferns patches. A calling male inspected the clutch, suggesting multiple mating during the breeding season. At FEC we found a male with seven tadpoles on its back (also in a moist place). This species uses small ponds on the forest floor for tadpole deposition, sometimes these ponds has connectivity with streams One of us (PRMS) checked husks of Brazil nuts for tadpoles, but they were not found. In the FEC site, husks of Brazil nuts are used by Allobates hodli, Ameerega hahneli, and A. trivittata for tadpole deposition.
Distribution. Allobates flaviventris is known from southwestern Amazonia (see Fig. 7 View FIGURE 7 ), in the states of Acre and Amazonas, Brazil. The northern limit reach the basin of Purus River near Boca do Acre city, state of Amazonas. The site of the westernmost occurrence of A. flaviventris is located near the city of Rio Branco, Acre, ca. 5 km on BR-364 highway (10º 00’ 58’’ S, 67º 44’ 05’’ W). The species is known to occur on the both banks of the Iquiri River. Thus, this river does not represent a barrier to the distribution of this species. Therefore, the distribution of this species may extend south to Departamento Pando in Bolivia. Köhler & Lötters (1999 p. 264) noted the occurrence of this unnamed species, (= Colostethus sp. A) in the region of Cobija, Bolivia, near the border with Brazil. The morphological features of the species in that area are similar to those found for the populations described here. The habitat used and call parameters reported by Köhler and Lötters (4450 Hz, 36.7 ms intervals) are similar to those found by us (see call analysis) and their photograph is assignable to Allobates flaviventris by lack ventrolateral stripes, having white marks instead, and by having dorsolateral stripes in hourglass pattern.
No known copyright restrictions apply. See Agosti, D., Egloff, W., 2009. Taxonomic information exchange and copyright: the Plazi approach. BMC Research Notes 2009, 2:53 for further explanation.