Allobates tapajos, Lima, Albertina P., Simões, Pedro Ivo & Kaefer, Igor Luis, 2015

Lima, Albertina P., Simões, Pedro Ivo & Kaefer, Igor Luis, 2015, A new species of Allobates (Anura: Aromobatidae) from Parque Nacional da Amazônia, Pará State, Brazil, Zootaxa 3980 (4), pp. 501-525 : 504-521

Allobates tapajos sp. nov.

Figures 1–7

Colostethus sp. ( C. trilineatus complex) Caldwell & Araújo 2005 p. 10–11.

Allobates “Curuá-Una” Grant et al. 2006 pg. 128, Fig. 73, Curuá-Una terminals; Santos et al. 2009, supplementary material, Fig. S3A, “ Allobates marchesianus BR” terminals.

Holotype. INPA-H 34425 (original field number APL12978), an adult male, collected on 21 February 2008 by Albertina P. Lima in Parque Nacional da Amazônia, Municipality of Itaituba, on the western margin of the Tapajós River, a southern tributary of the Amazon River (04°28’13.76” S; 56°17’04.60” W; elevation 132 m a.s.l.), in Pará State, Brazil. The type locality is 52 km southwest of the city of Itaituba on BR-230 road (Rodovia Transamazônica). The specimen was collected near a stream, 100 m from the station named “Base 1”, near the park’s main gateway.

Paratypes. Thirteen males (INPA-H 34402–34406, 34410–34412, 34414–34416, 34418, 34423), original field numbers APL12100–12102, 12965, 12974–12977, 12979–12983) and ten females (INPA-H 34407–34409, 34413, 34417, 34419–34422, 34424), original field numbers APL 12966–12973,12984 and 12985), all collected by A. P. Lima, William E. Magnusson and Laudelino S. Vasconcelos between 20–24 February 2008 at the species type-locality. We designated INPA-H 34424 (original field number APL 12966), an adult female collected by A. P. Lima on 21 February 2008, as the species allotype ( Fig. 2 D).

Etymology. The specific epithet makes reference to the Tapajós River, a major clearwater tributary of the Amazon River. The Tapajós River Basin has paramount and countrywide relevance in natural resources and biodiversity, and also in cultural and scenic terms. The Tapajós limits the eastern edge of Parque Nacional da Amazônia.

Description of holotype. An adult male (INPA-H 34425), SVL 15.6 mm, head slightly wider than long, head length 94% of head width, head width 34% of SVL; head length 33% of SVL; snout rounded to nearly truncate in dorsal view ( Figs. 1 A and 2A) and bluntly rounded in lateral view ( Figs. 1 C and 2C), extending past lower jaw ( Fig. 2 B); snout length 45% of head length; internarial distance 37% of head width; eye-naris distance 71% of eye length; nares opening posterolaterally; tympanum round, 43% of eye length; tympanic annulus visible ( Fig. 1 C); tongue nearly twice as long as wide, attached anteriorly, rounded along posterior margin; median lingual process absent; vocal sac present ( Fig. 1 B). Teeth on premaxilla and maxilla imperceptible under light microscope at 50X magnification. Cloacal tubercles absent.

Skin texture smooth on ventral and dorsal surfaces, with small scattered tubercles on dorsum from level of eyes to end of urostyle, visible in life ( Fig. 1 A) and in the preserved specimen ( Fig. 2 A); forearm slightly thicker than upper arm, length of forearm 80% of upper arm length; Finger III length 24% of SVL; Finger I 0.1 mm longer than Finger II; Finger II, 0.2 mm longer than Finger IV; Finger IV not reaching distal subarticular tubercle of Finger III when fingers are apressed; Finger II reaching the mid level of the distal subarticular tubercle of Finger III; relative length of fingers III>I>II>IV; Finger III not swollen ( Fig. 3 B); basal webbing and lateral fringes absent on fingers; palmar tubercle nearly round, 0.40 mm in maximum diameter, approximately 10% of Finger III length; width of thenar tubercle half the diameter of palmar tubercle ( Fig. 3 B); a single subarticular tubercle present on Fingers I, II, and IV, two subarticular tubercles present on Finger III ( Fig. 3 B); Discs on Fingers I, II, III and IV moderately expanded, all with distinct dorsal scutes; width of disc on Finger III 0.6 mm; width of distal phalange of Finger III at its mid length about 63% of the width of Finger III’s disc ( Fig. 3 B).

Hindlimb robust; leg and tibia with the same length, foot slightly shorter. Tibia length 48% of SVL; relative length of toes IV>III>V>II>I; basal webbing present between Toes III–IV; rudimentary basal webbing present between Toes II–III ( Fig. 3 A); tip of Toe I reaching mid level of of subarticular tubercle of Toe II when toes are appressed; tip of Toe III reaching past the subarticular tubercle of Toe IV; tip of Toe V reaching past one third of the second phalange of Toe IV ( Fig. 3 A); disc on Toe IV moderately expanded, width of disc 0.6 mm; inner metatarsal tubercle oval and outer metatarsal tubercle round; median metatarsal tubercle round present on both feet, but poorly defined ( Fig. 3 A). Metatarsal fold absent; tarsal keel tuberclelike and curved, located 1.4 mm from proximal edge of inner metatarsal tubercle and not connected by a fold ( Fig. 3 A). A single subarticular tubercle present on Toes I and II; two subarticular tubercles present on Toes III and V; three subarticular tubercles present on Toe IV, proximal and distal tubercles poorly defined ( Fig. 3 A).

Diagnosis. A small species of cryptically-colored Allobates . (1) Mean SVL of males 15.5 mm (range 14.9– 16.2 mm); mean SVL of females 16.3 mm (range 15.5–16.8 mm); (2) Dorsal surface of snout uniformly light brown, same color extending posteriorly until eye level; dorsum with irregular dark brown blotches scattered from eye level to urostyle region. Dorsum laterally light brown from eye level to urostyle region, width of light brown area variable; background color of dorsal surfaces of legs light brown; diffuse, weakly conspicuous transverse bands of a darker brown shade sometimes present across dorsal surfaces of thigh (on its proximal region), shank (at mid level) and foot (at mid level); posterior surface of thigh and groin dark brown; skin texture of dorsum smooth, with small granules uniformly scattered from the level of the eyes to posterior edge of dorsum; (3) lateral dark brown band with a diffuse lower edge (4) pale dorsolateral stripe present in preserved specimens, inconspicuous in life; (5) oblique lateral stripe absent, pale cream spots scattered on inguinal region of the lateral dark brown band; (6) pale ventrolateral stripe absent; (7) pale paracloacal mark present, half-moon shaped; (8) throat and chest of males cream in preserved specimens, golden yellow in life; abdominal surfaces white centrally, yellow towards the flanks and ventral surface of legs in live males; (9) throat, chest and abdomen uniformly white to translucent in preserved and in live females; chin light yellow in life, cream in preserved female specimens; (10) dark throatcollar absent; (11) iris is metallic gold, without a pupil ring; (12) median lingual process absent; (13) vocal sac distinct, subgular; (14) maxillary teeth present; (15) third phalange of Finger III not swollen in males, morphology of Finger III similar between males and females; (16) distal tubercle absent on Finger IV; (17) tip of Finger IV not reaching the distal subarticular tubercle of Finger III; (18) tip of Finger II reaching the middle of the distal subarticular tubercle of Finger III; (19) basal webbing present between Toes III and IV, rudimentary basal webbing present between Toes II and III; (20) fringes absent on toes; (21) testes not pigmented, extending one-half the length of kidney; (22) mature oocytes pigmented; (23) eggs laid on opaque jelly nests, on dead leaves on the forest floor; (24) caudal musculature of tadpoles bifurcated dorsally on body region; (25) anterior labium of tadpoles with short round papillae only on lateral margins, 4–5 on each side; posterior labium with a single row of marginal papillae variable in length, short and round medially and on the outer lateral folds, elongate between these regions; (26) diurnal habits, males calling in daytime; (27) advertisement calls characterized by the continuous emission of note pairs; notes with ascending frequency modulation with peak frequencies between 5.3–5.9 kHz, split by approximately regular silent intervals of 0.30 to 0.49 s; additional call arrangements (random emission of single notes and note trios, and emission of short trills) produced less frequently.

The new species is unambiguously assigned to genus Allobates by the combination of the above characters 10, 12, 19, 21, 26.

Comparisons with other species. Because the new species is distributed in the Tapajós River basin (see “Distribution and natural history” below), comparisons with other Allobates species are limited to all species occurring in Brazil, and to two additional species distributed in the Guiana Shield ( Allobates granti and A. amissibilis— Appendix II).

Allobates tapajos differs from Allobates femoralis (Boulenger 1883) , A. hodli Simões, Lima & Farias 2010 , and A. myersi (Pyburn 1981) by the absence of a red or yellow flash mark on dorsal surface of thigh, and by the absence of black and white marbling on the abdomen.

Allobates tapajos is distinguished from A. crombiei (Morales 2000 “2002”), A. masniger (Morales 2000 “2002”), A. nidicola ( Caldwell & Lima 2003), A. paleovarzensi s Lima, Caldwell, Biavati & Montanarin 2010 and A. vanzolinius (Morales 2000 “2002”) by it smaller size (maximum SVL <16.9 mm in A. tapajos ; minimum SVL> 17.9 mm in A. crombiei ,> 17.9 mm in A. masniger ,> 18.5 mm in A. nidicola ,> 18.3 mm in A. paleovarzensis ,> 19.6 mm in A. vanzolinius ), by the presence of irregular dark brown blotches scattered on dorsum from eye level to urostyle region (dorsum uniformly brown in A. masniger , A. nidicola , A. paleovarzensis and A. vanzolinius ), by the absence of a pale ventrolateral stripe (pale ventrolateral stripe present and unbroken in A. crombiei , A. masniger , A. nidicola , A. paleovarzensis and A. vanzolinius ), and by the color of throat and chest in adult males (cream in preserved specimens, golden yellow in live specimens; throat and chest dark gray or dark brown in A. masniger , A. nidicola , A. paleovarzensis and A. vanzolinius ).

Allobates tapajos is similar in size to A. marchesianus (Melin 1941) , A. caeruleodactylus ( Lima & Caldwell 2001), A. conspicuus (Morales 2000 “2002”), A. fuscellus (Morales 2000 “2002”), A. sumtuosus (Morales 2000 “2002”), A. granti Kok et al. 2006 , A. subfolionidificans ( Lima, Sanchez & Souza 2007), A. grillisimilis Simões, Saturaro, Peloso & Lima 2013 and A. amissibilis Kok, Hölting & Ernst 2013 . It can be easily distinguished from these species by the presence of diffuse irregular dark brown blotches scattered on dorsum from eye level to urostyle region (dorsum uniformly brown in A. amissibilis , A. caeruleodactylus , A. conspicuus , A. fuscellus , A. granti , A. grillisimilis , A. marchesianus , A. subfolionidificans , A. sumtuosus ). Throat and chest of male A. tapajos are golden yellow in life, uniformly cream with no dark pigmentation in preserved specimens (throat and chest white to translucent in male A. caeruleodactylus , A. grillisimilis , A. subfolionidificans and A. sumtuosus ; throat and chest with gray pigmentation in A. amissibilis , A. fuscellus and A. marchesianus ; throat and chest whitish, stippled with melanophores in live and preserved specimens of A. granti ). Allobates tapajos can be also distinguished from A. amissibilis , A. conspicuus , A. fuscellus , A. granti , A. grillisimilis , A. marchesianus , A. subfolinidificans and A. sumtuosus in having a dark brow lateral band with diffuse lower edge (lateral band solid dark brown, with a continuous lower edge in the other species). Allobates tapajos differs from A. caeruleodactylus in live males lacking sky-blue fingers (fingers brown to orange-brown in dorsal view in live A. tapajos males) and from A. conspicuus by the presence of a pale ventrolateral stripe (dark brown lateral band has a diffuse lower edge, not delimited by a pale ventrolateral stripe in A. tapajos ).

Allobates tapajos can be distinguished from A. olfersioides (Lutz 1981) and A. goianus (Bokermann 1975) by the lack of a cross-like pattern on dorsum (a dark brown, cross-like mark is present on dorsum of A. goianus and A. olfersioides ).

Allobates tapajos differs from A. flaviventris Melo-Sampaio, Souza & Peloso 2013 and from A. magnussoni Lima, Simões & Kaefer 2014 by its smaller size (maximum SVL <16.9 mm in A. tapajos ; minimum SVL> 17.4 mm in A. flaviventris and A. magnussoni ), and by the absence of a white ventrolateral stripe in live or preserved specimens (interrupted or continuous pale ventrolateral stripe present in A. flaviventris ; white iridescent blotches form a diffuse ventrolateral line in A. magnussoni ). Throat and chest of male A. tapajos is golden yellow in life, cream in preserved specimens, with no dark pigmentation (dark pigments present on throat and chest surfaces of A. flaviventris and A. magnussoni ).

Allobates tapajos is distinguished from Allobates brunneus (Cope 1887) by the color of ventral surfaces of live males (live males of A. brunneus have greenish-yellow throats, with melanophores evenly dispersed on vocal sac and chest; abdomen is greenish-yellow). A white ventrolateral stripe is visible in life in A. brunneus (absent in A. tapajos ). Allobates tapajos differs from A. gasconi (Morales 2000 “2002”) by males not having a swollen Finger III (Finger III swollen in A. gasconi males), by the absence of a pale ventrolateral stripe (present in A. gasconi ), and by the absence of dark pigment on throat (throat with dark pigmentation in A. gasconi ). Additionally, dark brown areas on light brown background form an hourglass-shaped figure on dorsum of A. brunneus , A. crombiei , A. flaviventris , A. gasconi and A. magnussoni (dark brown blotches are irregularly scattered, forming variable shapes in A. tapajos ).

Variation within type series. SVL of females significantly larger than that of males (T-test: t = 5.98, df = 22, P <0.0001; Table 1 ). Internarial distance in relation to head width significantly shorter in males than in females (mean IN/HW 0.39 ± 0.02 in males, 0.42 ± 0.02 in females; t = 3.81, df = 22, P = 0.001). Remaining morphometric ratios on head not statistically different between sexes. The following ratios are estimated based on all specimens pooled: Head width 0.35 ± 0.01 of SVL (range 0.33–0.37); SL 0.44 ± 0.07 of HL (range 0.31–0.53); HL 0.94 ± 0.05 of HW (range 0.80–1.02); EN 0.25 ± 0.02 of HW (range 0.19–0.28); EN 0.30 ± 0.03 of IO (range 0.24–0.36); tympanic membrane round, 0.43 ± 0.05 of EL (range 0.33–0.53).

Variation of limb ratios among 24 specimens: length of forearm 0.78 ± 0.06 of upper arm length (range 0.65– 0.87); subarticular tubercles on fingers similar in size and number to those of the holotype in all specimens ( Fig. 3 B). Length of Finger III in relation to SVL is statistically longer in males than in females (mean HANDIII/SVL 0.24 ± 0.01 in males, 0.23 ± 0.001 in females; t = -2.20, df = 22, P = 0.047). Width of the third phalange on Finger III not significantly different between sexes (t = -0.39, df = 22, P = 0.695); Fingers I, II and IV with nearly the same length.

Ratio of leg length 0.96 ± 0.04 (85–100%) of tibia length; foot length 0.85 ± 0.04 (0.79–0.93) of tibia; Ratio of tibia length to SVL in males (mean 0.48 ± 0.02) significantly (t = -2.07, df = 22, P = 0.05) greater than in females (mean 0.47 ± 0.02); foot length 0.41 ± 0.02 (0.37–0.44) of SVL; basal webbing present between Toes III-IV and a rudimentary webbing present between Toes II-III in all specimens (as Fig. 3 A); disc width on Toe IV 0.64 ± 0.08; metatarsal fold absent; tarsal keel and subarticular tubercles on toes similar as in holotype in all specimens ( Fig. 3 A).

Color in life of adults. Color in life of the holotype is shown in Fig. 1 . All specimens have a light brown background on dorsum. Dorsal surface of snout light brown, extending posteriorly until eye level. Central region of dorsum with irregular dark brown blotches from eye level to urostyle region. Dorsum light brown laterally from eye level to urostyle region. Width of the light brown area variable among specimens ( Fig. 4 A–D). Dorsal surface of arm orange-brown; upper surfaces of legs light brown with one or two diffuse dark brown bands on tibia and foot; inner surface of thigh and groin dark brown ( Fig.1 A, Fig. 4 A–D). Dorsolateral stripe absent or diffused with the background color of dorsum ( Fig. 1 A, Fig. 4 A, C, D). Throat and chest of calling males golden yellow; abdomen whitish centrally, gradually merging into yellow towards the flanks and ventral surface of legs ( Fig. 1 B, Fig. 4 H). Adult females have a light yellow chin; throat, chest, abdomen, and ventral surface of leg white to translucent, without melanophores ( Fig. 4 G, I). Paracloacal marks cream-colored, forming a half-moon shape; palmar surface of hand light brown; plantar surface of foot dark brown ( Fig. 1 A, Fig. 4 C, H, I). Color photographs and a video of a calling male can be obtained from Project Sapoteca’s website (http://ppbio.inpa.gov.br/en/ sapoteca/home).

Color of adults in preservative. The color in preservative of the holotype and allotype are shown in Fig. 2 . Considering the type series, background color of dorsum is cream. Dorsal surface of snout cream, extending posteriorly until eye level. Central region of dorsum with irregular dark brown blotches from eye level to urostyle region ( Fig. 4 A). Pale dorsolateral stripe present in all specimens. Ventrolateral stripe absent. Lateral surface below the lateral brown band cream with white spots in some specimens (as in the holotype and allotype, Fig. 2 ). Diffuse, pale cream spots scattered posteriorly on the dark brown lateral band. Tympanum dark brown dorsally (up to about one quarter of its area), same color as the brown lateral band at the supratympanic region. Remaining area of tympanum cream.

Dorsal surfaces of hind limb cream with one or two diffuse transversal brown bands on tibia and foot. Paracloacal region and internal surface of thigh brown ( Fig. 2 A–D, Fig. 5 A). Half-moon shaped paracloacal mark cream ( Fig. 5 A). Ventral surfaces of throat, chest, belly, leg and upper arm in males and females pale cream, without melanophores ( Fig. 2 B, Fig. 5 B). Ventral surfaces of forearm and foot brown in all specimens ( Fig. 5 B).

Description of tadpoles. Descriptive statistics for 17 morphometric characters were based on 35 tadpoles at developmental stages 25, 33, 35 and 39 ( Table 2 ). The following detailed description is based on a sample of 8 tadpoles from the same lot (INPA-H 34427), collected from a single clutch and reared to stage 35 before preservation.

Body ellipsoid, round anteriorly and truncate posteriorly in dorsal view, slightly flattened in lateral view ( Fig. 6 ). Body length 38% and tail length 75% of TL; body wider than deep, BH 45% of BW at the level of spiracle; HWLE at level of eyes 99% of BW at level of spiracle; snout rounded in dorsal and lateral views; END 72 % of ED; eyes dorsal and directed laterally; mean ED 0.8 ± 0.05 mm; IOD 24% of HWLE at level of eyes. Small naris located dorsolaterally and directed anterolaterally; mean internarial distance 1.08 ± 0.08 mm. Fleshy ring present on the inner margin of nostrils, round and straight, not ornamented. Sinistral spiracle a free tube of 1.12 ± 0.19 mm on average, slightly anterior to mid-body ( Fig. 6 ). Average length of vent tube 1.04 ± 0.16 mm, dextral. Dorsal fin begins at body-tail insertion, its dorsal edge shallow and straight anteriorly (along approximately 20% of its length), deeper and arched posteriorly, reaching its maximum depth about half its length. Dorsal fin slightly deeper than lower fin. Tail tip acuminate. Caudal musculature of tadpoles bifurcated dorsally on body region, reaching up to two thirds of body length from body-tail insertion to past the level of spiracle ( Fig. 6 ).

Oral apparatus located anteroventrally, emarginated laterally, transversely elliptical, 1.96 ± 0.21 mm wide in average ( Fig. 7 ). Anterior labium with groups of four or five short and round papillae distributed in a single row on each side of the labium’s lateral margins, and split by a medial gap corresponding to approximately 75% of the apparatus’ width. Posterior labium with a single row of marginal papillae variable in length. Papillae short and round medially (4–5 papillae) and on the outer lateral folds of labium (4–5 papillae on each side). Elongate papillae (4-5 papillae on each side) distributed between the lateral and medial groups of shorter papillae. Elongate papillae reach up to three or four times the length of short papillae ( Fig. 7 ).

Upper jaw sheath arch-shaped, longer than lower jaw sheath. Lower jaw sheath V-shaped, deeper than upper jaw sheath. Cutting edge of upper and lower jaw sheaths serrated, serrations extending along the entire length of both sheaths. Labial tooth row formula 2(2)/3(1); tooth row A-1 complete, 1.5 mm long in average; tooth row A-2 interrupted medially, consisting of two widely separated rows at the level of upper jaw sheath, segments 0.50 mm in length, with gap of 0.50 mm, in average ( Fig. 7 A). Posterior tooth rows P-1 and P-2 the same length as row A-1 (1.5 mm, in average), both longer than P-3 (1.2 mm, in average).

Morphology of oral disc is constant throughout all developmental stages examined. Elongate papillae on posterior labium are evident from stage 25 to stage 39 ( Fig. 7 A–D).

Color of preserved tadpoles. Background color of dorsum, anteroventral and lateral surfaces of the body of tadpoles pale cream, but all of these surfaces are densely covered with dark brown melanophores. Background color appears as irregular pale blotches or spots, more frequent in lateral view ( Fig. 6 ). Posteroventral surface of body transparent with intestines visible through the skin. Tail musculature cream. Tail fins transparent, with scattered irregular brown melanophores forming irregular blotches ( Fig. 6 ).

Comparison of the tadpoles with other species. The tadpoles of Allobates tapajos differ from those of A. granti , A. subfolionidificans , A. brunneus , A. paleovarzensis , A. sumtuosus and A. magnussoni ( Kok et al. 2006; Lima et al. 2007, 2009, 2010, 2014; Simões & Lima 2013) by having a posterior labium partially surrounded by long marginal papillae, variable in length (posterior labium surrounded by short papillae of similar size in tadpoles in the other six species). Tadpoles of A. nidicola and A. masniger are endotrophic and develop entirely in terrestrial nests. Those highly specialized tadpoles lack an oral disc and a spiracle.

Tadpoles of A. tapajos are more similar to those of A. caeruleodactylus , A. marchesianus , and A. grillisimilis (Caldwell et al. 2002a, 2002b; Simões et al. 2013a) in having fewer and longer papillae on posterior labium. Tadpoles of A. caeruleodactylus and A. marchesianus have distinct transverse dark bars on tail (absent in A. tapajos ) and two short papillae on each side of the lateral margin of anterior labium (four to five in A. tapajos ). The tadpoles of A. tapajos differ from those of A. marchesianus by having labial tooth rows A-2 and P-3 (rows A-2 and P-3 absent in tadpoles of A. marchesianus ). Row A-2 is longer than A- 1 in tadpoles of A. tapajos (row A-2 shorter than A- 1 in A. caeruleodactylus ). The tadpoles of A. tapajos differ from those of A. grillisimilis by having labial tooth row P-3 distinctively shorter than rows P-1 and P-2 (all rows equal to sub-equal in length among tadpoles of A. grillisimilis ); row A-2 is longer than A- 1 in tadpoles of A. tapajos (A-2 and A-1 about the same length in A. grillisimilis ).

Call description and call variation. Advertisement calls of Allobates tapajos consist of the continuous emission of short tonal notes, with ascending frequency modulation. Notes predominantly arranged in pairs or couplets ( Fig 8 A, 9), with silent intervals between notes in each pair varying from 0.112 to 0.162 s (0.138 ± 0.015 s in average, n = 135 intervals, measured from calls of nine males). Longer silent intervals split consecutive note pairs. These vary from 0.305 to 0.490 s (0.403 ± 0.058 s in average, n = 135 intervals, measured from calls of nine males). The arrangement in note pairs is not regular along calling bouts. Single notes and note trios emitted irregularly along some call sections ( Fig. 8 C, 9). Note pairs are, however, the prevalent arrangement, corresponding to 311 of 424 (73%) arrangements inspected from call sections of nine males. Single notes emitted between note pairs was the second most common arrangement, corresponding to 23% of the arrangements inspected from calls of nine males. Note trios were the rarest arrangement, corresponding to 18 of 424 (4%) arrangements inspected, and being detected only along call sections of five males. Despite the existence of variation in note arrangement, spectral and temporal characteristics of notes are constant along advertisement calls of each individual ( Fig. 8 B, D).

Notes 0.042 ± 0.004 s long in average (n = 250 notes, analyzed from calls of ten males—Table 3). Notes have an average lower frequency of 5093.4 ± 217.9 Hz and an average upper frequency of 6004.0 ± 215.6 Hz. Average peak frequency of notes 5618.9 ± 209.4 Hz. Silent intervals between adjacent notes are more variable, ranging from 0.041 to 0.806 s considering raw individual measurements (0.293 ± 0.037 s, in average), reflecting variations in note arrangement. Notes emitted at a rate of 2.9 ± 1.4 notes per second in average, ranging from 2.73 to 3.23 notes per second among call sections of nine males.

A single male (INPA-H 34411) was recorded while producing a second type of call before starting to emit regular advertisement calls as described above. From a total recording length of four minutes, this alternative call type was emitted during the first two and a half minutes of recording, after which the focal male shifted to regular advertisement calls. This call type consists of short trills of single notes, split by long and irregular silent intervals, ranging from 10.73 up to 39.16 s ( Fig. 10 ). A total of six complete trills were captured on the recorded section and used for description of call arrangement.

Trills formed by 10–14 notes (mode = 13 notes). Duration of trills ranged from 2.46 to 3.37 s (3.04 ± 0.33 s, in average), and the rate of note emission ranged from 3.6 to 4.2 notes per second (4.0 ± 0.3 notes per second, in average). The rate of note emission during trills is measurably faster than that estimated from regular advertisement calls of the same individual (2.73 notes per second), reflecting shorter and less variable silent intervals between notes along trills (0.19 ± 0.02 s in average; silent intervals between notes along regular advertisement calls of the same individual 0.30 ± 0.18 s long, in average; n = 25 intervals, uniformly distributed along each call section).

Mean values of note duration and lower, upper and peak frequencies of notes on trills were 0.04 s, 4972.20 Hz, 5883.19 Hz and 5644.28 Hz, respectively. There were no statistically significant differences between the peak frequencies of notes emitted in regular advertisement calls or alternative trills by this male (t = -0.790, P =0.434, 46.70 d.f.). Note duration, however, was shorter, on average, when produced in trills (t = 4.228, P <0.001, 43.18 d.f.), and notes on trills were emitted at a slightly higher frequency bandwidth than those emitted in regular calls (lower frequency: t = -2.669, P = 0.011, 37.29 d.f.; upper frequency: t = -2.188, P = 0.034, 44.73 d.f.).

TABLE 4. Uncorrected pairwise (lower left) and Kimura-2-Parameter (upper right) genetic distances between Allobates tapajos and clades representing cryptically colored Allobates species distributed in the cis-Andean region. Distances were based in a 420 bp fragment of the 16S rDNA mitochondrial gene.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 tapajos 0.09 0.10 0.12 0.11 0.09 0.09 0.07 0.12 0.13 0.10 0.12 0.01 0.09 0.21 0.10 0.08 0.12 0.10 0.10 0.14 algorei 0.08 0.08 0.09 0.10 0.09 0.08 0.10 0.10 0.12 0.09 0.12 0.09 0.12 0.20 0.09 0.08 0.10 0.09 0.08 0.16 Genetic distances and mtDNA relationships. Phylogenetic relationships based on a 420 bp fragment of the 16S rDNA gene indicate that Allobates tapajos is monophyletic in relation to the 20 congeneric taxa analyzed, except in relation to the two samples tentatively identified as Allobates marchesianus taken from a locality near Curuá-Una River on the right bank of the Tapajós River ( Fig. 11 ). Mean uncorrected pairwise and Kimura-2- parameter (K2P) genetic distances between sequences of A. tapajos and A. aff. marchesianus do not exceed 1% (Table 4). Average uncorrected pairwise genetic distances between sequences of the new species and those of other congeneric species occurring in the Amazon region range from 6% ( A. gasconi from Eirunepé, in the middle course of the Juruá River) to 13% (typical Allobates undulatus from Cerro Yutajé, in Amazonian Venezuela). Average K2P genetic distances ranged from 7% (also between A. tapajos and A. gasconi from Eirunepé) to 14% (between A. tapajos and A. undulatus ). The greatest estimates of average genetic distances are observed between A. tapajos and A. olfersioides , a species distributed in the Brazilian Atlantic Forest (18% uncorrected pairwise, 21% K2P).

Parque Nacional da Amazônia is the type locality of two additional species of cryptically colored Allobates : A. masniger and A. magnussoni , which are generally syntopic with A. tapajos . Average uncorrected pairwise and K2P genetic distances between typical A. tapajos and A masniger are 11% and 12%, respectively. Mean genetic distance between typical A. tapajos and A. magnussoni is 10%, considering both pairwise and K2P estimates ( Fig. 11 , Table 4).

Distribution and natural history. At present, Allobates tapajos is known from the Tapajós River basin, in terra-firme rainforests on both banks of the river´s middle and lower course. Known occurrences include the type locality in Parque Nacional of Amazônia, a federal conservation unit and forest reserve located on the western bank of the Tapajós , about its middle course, in the south of Pará State, Brazil ( Fig. 12 ). Within the park, the species inhabits forested areas along the margins of clearwater streams. The species is also distributed across the Tapajós River, in forests located near the city of Santarém, near Curuá-Una River, about Tapajós River’s lower course. Preliminary observations of the latter populations (from which “ aff. marchesianus ” in Fig. 11 proceed) indicate that they markedly differ from typical A. marchesianus (which have uniformly brown dorsum, solid lateral dark brown band with a continuous lower edge and dark grey vocal sac pigmentation, among other distinctive characters —Caldwell et al. 2002), rendering earlier identifications equivocal and supporting their assignment to the new species.

At the type locality, A. tapajos is most active early in the morning and late in the afternoon. Most calling males are found from 0 530 to 0 930 h, and again from 1630 h until dusk. Two jelly nests were found on the leaf litter, inside folded and rolled leaves. One clutch had 27 eggs and the second 23 tadpoles at developmental stage 25. The jelly protecting both clutches was opaque ( Fig. 13 A and 13B).

A single calling male was observed entering a folded leaf on the leaf litter and being followed by a conspecific female. It was not possible to observe courtship movements inside the leaf. However, a recently laid clutch was found after the couple left the nest. When we played a recorded advertisement call back to 10 male individuals engaged in calling activity, they all advanced towards the speakers. Thus, it appears that males of this species are territorial.

Figures

FIGURE 1. (A) Dorsal, (B) ventral and (C) lateral views in life of the holotype of Allobates tapajos (INPA-H 34425). (C). Photographs by A. P. Lima.

FIGURE 2. (A) Dorsolateral, (B) ventral and (C) lateral views in preserved of the holotype of Allobates tapajos (INPA-H 34425). (D) Dorsolateral view of the allotype (INPA-H 34424) from the species type locality in Parque Nacional da Amazônia (PNA). Photographs by A. P. Lima.

FIGURE 3. Foot and hand of Allobates tapajos holotype (INPA-H 34425). Scale bars near the foot correspond to 2.0 mm and hand 1.0 mm. Black arrows indicate basal webbing between Toes II and III and between Toes III and IV.

FIGURE 4. Color in life of Allobates tapajos. (A) Dorsolateral and (G) ventral view of allotype (INPA-H 34424) from the species type locality in Parque Nacional da Amazônia (PNA). (A), (B), (C) and (D) Variation in dorsal color pattern. Note absence of dorsolateral lines. Ventral view of (I) a female (INPA-H 34407) and of (H) a male (INPA-H 34404). Note whitish ventral surfaces of female specimens and the yellow throat and chest on males. Lateral views of females (A, D and F) and male (E) specimens. Note that the ventrolateral stripe is absent or is restricted to round or elongate iridescent-white spots. Photographs by A. P. Lima.

FIGURE 5. Preserved females (A — lower line) and males (A—upper line) of Allobates tapajos showing variation in dorsal pattern. Note the central region of dorsum with irregular dark brown blotches from eye level to urostyle region (A). Color of ventral surface (B) is similar among preserved females and males (F = female, M = male, numbers correspond to original field numbers).

FIGURE 6. Tadpoles of Allobates tapajos from Parque Nacional da Amazônia, near Itaituba, Pará State, Brazil, proceeding from egg clutches collected in February 2008. (A) From top to bottom: dorsal, ventral and lateral views of a tadpole in developmental stage 35 (INPA-H 34426). (B) From top to bottom: dorsal, ventral and lateral views of a tadpole in developmental stage 39 (INPA-H 34427). A tadpole from lot INPA-H 34427 was used in the detailed description (see text). Scale bar corresponds to 5.0 mm.

FIGURE 7. Oral discs of Allobates tapajos from Parque Nacional da Amazônia, Itaituba, Pará State, Brazil. (A) Ventral and (B) oblique views of the oral disc of a tadpole at developmental stage 39. (C) Oral disc of tadpole at stage 35. Papillae on the right portion of posterior labium are turned anteriorly to emphasize their length. (D) Oral disc of a second tadpole at developmental stage 35, emphasizing relative length of tooth rows. Scale bar corresponds to 0.5 mm.

FIGURE 8. Oscillograms (upper graphs) and sonograms (lower graphs) of advertisement calls of Allobates tapajos recorded at the species type locality in Parque Nacional da Amazônia, Itatituba, Pará State, Brazil. (A) Ten seconds of continuous advertisement calls of a male (INPA-H 34402, SVL = 15.02 mm) recorded at 23.9 ° C. Notes are predominantly arranged in consecutive note pairs split by slightly longer silent intervals. (B) Detailed view of two note pairs from the same call session, ascending frequency modulation is evident in the sonogram. (C) Advertisement calls of a second male (INPA-H 34423, SVL = 15.32 mm), also recorded at type locality at 23.9 ° C, exemplifying variation in note arrangement (note several single notes scattered along call length). (D) Detailed view of the same call session demonstrating that spectral parameters of notes do not vary across distinct note arrangements.

FIGURE 10. An alternative call type of Allobates tapajos recorded from a single male (INPA-H 34411, SVL = 16.08 mm) at the species type locality. Recording was conducted at air temperature of 25 ° C. (A) This alternative call type is formed by trills of single notes split by relatively long silent intervals. (B) Detailed view of a trill of notes. Notes in trills were slightly shorter and emitted in a higher bandwidth when compared to notes in regular advertisement calls of the same individual.

FIGURE 11. Evolutionary relationships between Allobates tapajos and other species of cryptically colored Allobates occurring in Brazil and across cis-Andean South America, as inferred from an unrooted maximum likelihood phylogenetic tree based on a 420 bp fragment of the 16 S rDNA mitochondrial gene. All sequences proceed from specimens collected at the type localities of each species, except when indicated between parentheses. Live male frogs depicted on the right represent typical specimens of A. tapajos, A. magnussoni and A. masniger, which occur in sympatry in Parque Nacional da Amazônia, and their respective clades. Clade labels indicate support values obtained from 5000 bootstrap replicates (only support values> 80 % are shown). Basal clades received weak support and should not be considered to reflect the actual evolutionary relationships between the taxa analyzed.

FIGURE 12. Location of some large Amazonian rivers and relevant Allobates type localities in northern South America. Yellow star: Type locality of Allobates tapajos in Parque Nacional da Amazônia (PNA), on the west bank of Tapajós River, Pará State, Brazil. PNA is also the type locality of Allobates masniger and A. magnussoni. Yellow circle: Additional occurrence location of A. tapajos near Curuá-Una River on the east bank of the Tapajós River, southeast of the city of Santarém. Yellow square: Type locality of Allobates marchesianus. Allobates tapajos has been erroneously identified as A. marchesianus in past literature and molecular databases (see text). Yellow triangle: Origin of A. gasconi samples near Eirunepé, State of Amazonas, Brazil. Black dots: Type localities of Allobates species from the Amazon basin which are similar in size with A. tapajos and can be more easily misidentified with the latter by the examination of preserved specimens: 1 — A. conspicuus; 2 — A. fuscellus; 3 — A. gasconi; 4 — A. subfolionidificans; 5 — A. grillisimilis; 6 — A. caeruleodactylus; 7 — A. amissibilis; 8 — A. sumtuosus; 9 — A. granti.

FIGURE 13. Calling male (A), egg clutch (B) and tadpole nest (C) of Allobates tapajos from Parque Nacional da Amazônia, Itaituba, Pará State, Brazil.

Tables

TABLE 1. Morphometric measurements (in mm) of adult male and female Allobates tapajos collected from the species type locality in Parque Nacional da Amazônia, Pará State, Brazil. Abbreviations are defined in the text. Values are as: mean ± standard deviation (range of variation).

Character	Holotype	Males(n = 13)	Females (n = 10)
1-SVL	15.64	15.3 ± 0.4 (14.9–16.1)	16.2 ± 0.3 (15.6–16.5)
2-HL	5.10	5.02 ± 0.35 (4.40–5.60)	5.36 ± 0.24 (4.80–5.70)
3-IO	4.50	4.51 ± 0.19 (4.20–4.90)	4.72 ± 0.18 (4.40–5.00)
4-HW	5.40	5.38 ± 0.21 (5.00–5.80)	5.63 ± 0.09 (5.50–5.80)
5-SL	2.30	2.13 ± 0.29 (1.75–2.50)	2.44 ± 0.26 (2.00–2.75)
6-EN	1.50	1.35 ± 0.10 (1.15–1.50)	1.44 ± 0.16 (1.10–1.60)
7-IN	2.00	2.12 ± 0.06 (2.00–2.25)	2.38 ± 0.08 (2.20–2.50)
8-EL	2.10	2.15 ± 0.08 (2.00–2.25)	2.24 ± 0.09 (2.05–2.35)
9-TYM	0.90	0.92 ± 0.09 (0.75–1.10)	0.94 ± 0.12 (0.75–1.15)
10-FAL	3.20	3.15 ± 0.23 (2.70–3.50)	3.19 ± 0.42 (2.60–3.90)
11-UAL	4.00	4.10 ± 0.26 (3.70–4.50)	4.07 ± 0.24 (3.80–4.50)
12-LL	7.50	7.06 ± 0.35 (6.30–7.50)	7.33 ± 0.22 (7.10–7.70)
13-TL	7.50	7.40 ± 0.29 (6.70–7.90)	7.58 ± 0.22 (7.30–8.00)
14- FL	6.60	6.29 ± 0.23 (5.90–6.70)	6.50 ± 0.21 (6.20–6.80)
15- HAND I	3.00	2.68 ± 0.13 (2.50–3.00)	2.84 ± 0.21 (2.50–3.10)
16-HAND II	2.90	2.60 ± 0.10 (2.40–2.80)	2.67 ± 0.09 (2.50–2.80)
17-HAND III	3.80	3.60 ± 0.12 (3.40–3.80)	3.70 ± 0.09 (3.50–3.80)
18-HAND IV	2.70	2.33 ± 0.19 (2.00–2.70)	2.44 ± 0.16 (2.20–2.80)
19-WFD	0.55	0.53 ± 0.04 (0.50–0.65)	0.56 ± 0.07 (0.45–0.65)
20-DPT	0.40	0.46 ± 0.06 (0.35–0.55)	0.49 ± 0.03 (0.45–0.55)
21-WTT	0.20	0.25 ± 0.02 (0.20–0.30)	0.26 ± 0.03 (0.20–0.30)
22-WPF	0.40	0.35 ± 0.04 (0.30–0.45)	0.34 ± 0.06 (0.25–0.45)
23-WTD	0.60	0.61 ± 0.09 (0.45–0.75)	0.66 ± 0.07 (0.55–0.80)

TABLE 2. Measurements (in mm) of 35 tadpoles of Allobates tapajos in four different developmental stages (according to Gosner 1960). Values are means ± standard deviation; ranges are in parentheses.

Character Stage 25 (n=2)	Stage 33 (n=13)	Stage 35 (n=8)	Stage 39 (n=12)
1 -TL 10.58 ± 0.11 (10.50–10.66)	20.03 ± 0.55 (18.46–20.83)	20.15 ± 0.89 (18.46–21.50)	20.46 ± 0.04 (20.41–20.53)
2 - BL 3.25 ± 0.07 (3.20–3.30)	6.96 ± 0.20 (6.60–7.40)	7.03 ± 0.28 (6.70–7.60)	7.80 ± 0.17 (7.50–8.10)
3 -TAL 6.93 ± 0.08 (6.87–7.00)	12.60 ± 0.73 (11.36–13.76)	13.16 ± 0.64 (12.52–14.00)	15.26 ± 0.07 (15.16–15.37)
4 - BW 1.77 ± 0.03 (1.75–1.80)	5.14 ± 0.20 (4.85–5.65)	5.33 ± 0.26 (4.95–5.60)	5.88 ± 0.16 (5.65–6.20)
5 - BH 0.85 ± 0.07 (0.80–0.90)	2.00 ± 0.18 (1.83–2.33)	2.06 ± 0.29 (1.66–2.63)	2.63 ± 0.23 (2.16–2.96)
6 - HWLE 2.00 ± 0.00 (2.00–2.00)	4.96 ± 0.19 (4.50-5.30)	5.18 ± 0.14 (5.00–5.50)	5.87 ± 0.20 (5.50–6.10)
7 - TMW 0.83 ± 0.00 (0.83–0.83)	1.77 ± 0.14 (1.60–2.16)	1.95 ± 0.14 (1.66–2.13)	2.44 ± 0.11 (2.30–2.63)
8 - MTH 1.20 ± 0.14 (1.10–1.30)	3.11 ± 0.26 (2.60–3.56)	3.32 ± 0.21 (3.00–3.63)	3.61 ± 0.20 (3.23–3.90)
9 - TMH 0.91 ± 0.11(0.83–1.00)	1.72 ± 0.09 (1.60–1.93)	1.77 ± 0.15 (1.53–2.03)	2.05 ± 0.10 (1.83–2.16)
10 - IOD 0.61 ± 0.02 (0.60–0.63)	1.20 ± 0.11 (1.00–1.36)	1.22 ± 0.10 (1.06–1.33)	1.37 ± 0.06 (1.33–1.53)
11 - IND 0.46 ± 0.04 (0.43–0.50)	1.06 ± 0.09 (1.00–1.30)	1.07 ± 0.08 (1.00–1.16)	1.19 ± 0.06 (1.06–1.26)
12 - END 0.20 ± 0.04 (0.16–0.23)	0.59 ± 0.06 (0.50–0.66)	0.62 ± 0.06 (0.53–0.73)	0.69 ± 0.04 (0.66–0.80)
13 - NSD 0.33 ± 0.00 (0.33–0.33)	0.60 ± 0.05 (0.50–0.66)	0.59 ± 0.07 (0.50–0.66)	0.72 ± 0.11 (0.50–0.86)
14 - ED 0.46 ± 0.04 (0.43–0.50)	0.77 ± 0.04 (0.66–0.83)	0.80 ± 0.05 (0.70–0.86)	0.97 ± 0.03 (0.90–1.03)
15 - VTL -	1.04 ± 0.19 (0.75–1.40)	1.03 ± 0.16 (0.75–1.25)	1.28 ± 0.19 (0.75–1.50)
16 - STL 0.50 ± 0.00 (0.50–0.50)	1.00 ± 0.14 (0.75–1.25)	1.11 ± 0.18 (0.75–1.30)	1.27 ± 0.14 (1.00–1.65)
17 - ODW 0.47 ± 0.03 (0.45–0.50)	1.86 ± 0.09 (1.75–2.05)	1.95 ± 0.21 (1.75–2.40)	2.05 ± 0.06 (1.95–2.15)