Allobates magnussoni

Lima, Albertina P., Simões, Pedro Ivo & Kaefer, Igor Luis, 2014, A new species of Allobates (Anura: Aromobatidae) from the Tapajós River basin, Pará State, Brazil, Zootaxa 3889 (3), pp. 355-387 : 359-378

Allobates magnussoni

Figures 1–6 , 8 .

Colostethus brunneus Caldwell & Araújo 2005 p . 10

Allobates brunneus Grant et al. 2006 pg. 123, Fig. 72, brunneus terminals; Santos et al. 2009, supplementary material, Fig. S3A, “ Allobates brunneus BR” terminals.

Holotype. INPA-H 32960 (original field number APL12814), an adult male collected on 18 February 2008, in Parque Nacional da Amazônia, municipality of Itaituba, on the western margin of the Tapajós River (04°33’12.3” S; 56°18’00.3” W, elevation 132 m a.s.l.), Pará State, Brazil. The type locality is 65 km southwest of Itaituba’s urban perimeter on BR-230 (Transamazônica) road. The specimen was found near a water spring located ~ 300 m from the park’s research base, towards the margin of the Tapajós River.

Paratopotypes. Thirteen males (INPA-H 32960–32973, original field numbers APL12986–12990, 13000–13003, 13005, 13006, 13009, 13015), eight females (INPA-H 32974–32976, 32978–32982, original field numbers APL 12998, 13004, 13010, 13012–13014, 13016, 13017), all collected by APL, William E. Magnusson and Laudelino S. Vasconcelos between 18–24 February 2008 at the species type-locality. We designate INPA-H 32974 (original field number APL 12998), an adult female collected by APL on 22 February 2008, as the species allotype ( Fig. 3 A).

Paratypes. All from Pará State, Brazil. Treviso: INPA-H 10105–10109 (field number APL 200–207), INPA-H 33930–33932 (APL12076–12078) and INPA-H 33933–33934 (APL 13505 and 13508) 10 males, 3 females, Treviso, municipality of Belterra (03°08’44” S, W 54°50’33” W), collected in January 2002, February 2006 and March 2009 by APL and E.V. Farias. Jamanxim: INPA-H 33935 (APL 20537), municipality of Itaituba (06°18’22” S, 55°42’38” W).

Etymology. The specific epithet is dedicated to Dr. William (Bill) E. Magnusson, a professor from Instituto Nacional de Pesquisas da Amazônia who has trained many students to understand and love the Amazon forest and its frogs.

Diagnosis. (1) a relatively large species of cryptically colored Allobates , mean SVL of males 17.78 ± 0.76 mm (range 16.09–19.59 mm); mean SVL of females 19.50 ± 0.74 mm (range 17.97–20.84 mm). (2) dorsum has a darkbrown hourglass-shaped marking from the anterior to mid dorsal surface, consisting of three contiguous convex sections: the first between the eyes, approximately bat-shaped, the second approximately diamond-shaped, and the third similar to the first or triangular ( Fig. 1 A, C; Fig. 3 A, B, C; Fig. 4 ); dorsal surfaces of leg and arms brown in life and when preserved; thigh and tibia with one dark brown transverse bar each, which line up when sitting at rest; foot with two transverse bars in dorsal view, bars sometimes diffuse, one of them aligning with the bars on tight and tibia when specimens are at rest ( Fig. 1 A, C; Fig. 3 A, C, G; Fig. 4 ); (3) pale dorsolateral stripe, when present, diffuse; (4) a distinct pale brown oblique lateral bar crosses the dark-brown lateral stripe about mid body ( Fig. 1 A, C; Fig. 3 C, D); (5) ventrolateral stripe absent, or diffuse as small irregular iridescent white spots (as in the holotype in Fig. 1 A, C; see also Fig. 3 A, G, H) or as a series of wavy, elongate, interconnected spots ( Fig. 2 B, I); (6) pale paracloacal mark present, small, not forming a half-moon shape; (7) throat and chest of males pigmented, grayish-violet, gray in preserved specimens; belly yellowish, cream in preserved specimens ( Fig.1 B; Fig. 3 E, F; Fig. 4 ); (8) females with throat, chest and belly uniformly yellow (cream in preserved specimens), with a few melanophores on chin ( Fig. 3 D, E; Fig. 4 ); (9) dark throat collar absent; (10) iris metallic gold with tiny black flecks and a pupil ring; (11) paired dorsal digital scutes, (12) median lingual process absent; (13) tympanum inconspicuous, tympanic annulus inconspicuous; (14) vocal sac distinct, subgular; (15) maxillary teeth present; (16) third phalange of Finger III on males weakly wider in males than in females; (17) rudimentary webbing present only basally between Toes II-III and III-IV ( Fig. 2 ); (18) distal tubercle absent on Finger IV; (19) tip of Finger IV surpasses distal subarticular tubercle of Finger III when fingers are pressed against each other; (20) Finger II about 6% shorter than Finger I; (21) finger discs moderately expanded, middle section of third phalange of Finger III about 69% of finger disc’s width; (22) lateral fringes absent on fingers;(23) overall morphology of Finger III similar between males and females, only the third phalange of Finger III weakly wider on males.; (24) carpal pad absent; (25) thenar tubercle conspicuous, oval, moderately protuberant; (26) tarsal keel present, strongly curved; (27) a weak metatarsal fold is present; (28) disc of Toe IV weakly expanded; (29) fringes absent on toes; (30) mature oocytes pigmented; (31) Adult testis unpigmented, about one-third the length of kidney; (32) caudal musculature of tadpoles is bifurcated dorsally on body region, dorsal fin extending shortly between them ( Fig. 8 ); (33) oral disc of tadpoles emarginate, not umbelliform, anterior labium with short round papillae only on lateral margins, 12–13 on each side; posterior labium projected anteriorly, covering posterior tooth rows, and entirely surrounded by 32 to 35 short, pyramidal to slightly elongate papillae ( Fig. 6 A, B, C); (34) eggs are laid on transparent jelly nests, on dead leaves on the forest floor ( Fig. 11 A, B); (35) diurnal activity, males calling during daytime; (36) advertisement calls characterized by the continuous emission of single notes split by regular silent intervals, shifting sporadically to note-pairs; notes have peak frequencies of 4.2–4.8 kHz ( Fig. 6 ).

The new species is unambiguously assigned to Allobates by the combination of the above characters 1, 9, 12, 17, 20, 33, 34, and 35.

Comparisons with other species. Because of its geographic range is restricted to the Tapajós River basin, in central Brazilian Amazonia, and because of the presence of a distinct hourglass-shaped mark on dorsum, comparisons of Allobates magnussoni with other Allobates are limited to species occurring in Brazil, and to four additional species distributed in the Guiana Shield ( Allobates granti , A. amissibilis , A. sumtuosus and A. undulatus ). These comprise all species that could be potentially confounded with A. magnussoni or potentially overlap in geographic distribution.

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

Allobates magnussoni is distinguished from A. marchesianus (Melin 1941) , A. caeruleodactylus ( Lima & Caldwell 2001), A. conspicuus (Morales 2000 “2002”), A. fuscellus (Morales 2000 “2002”), A. masniger (Morales 2000 “2002”), A. sumtuosus (Morales 2000 “2002”). A. vanzolinius (Morales 2000 “2002”), A. nidicola ( Caldwell & Lima 2003), A. subfolionidificans ( Lima, Sanchez & Souza 2007), A. paleovarzensis Lima, Caldwell, Biavati & Montanarin 2010, A. grillisimilis Simões, Saturaro, Peloso & Lima 2013 and A. amissibilis Kok, Hölting & Ernst 2013 by the presence of a distinct dark bat-shaped mark between the orbits followed by a diamond-shaped and a triangular or bat-shaped form; Allobates magnussoni is distinguished from A. marchesianus , A. caeruleodactylus , A. conspicuus , A. fuscellus , A . sumtuosus, A. subfolionidificans , A. grillisimilis and A. amissibilis by its larger size (minimum SVL> 16.0 mm in males and> 17.9 mm in females). Allobates magnussoni is distinguished from A. vanzolinius and A. paleovarzensis by the absence of solid and continuous pale ventrolateral and dorsolateral stripes (present in A. vanzolinius and A. paleovarzensis ) and by the presence of a distinct oblique lateral bar (absent in A. vanzolinius and A. paleovarzensis ). Allobates magnussoni differs from A. olfersioides (Lutz 1981) and A. goianus (Bokermann 1975) in lacking a cross-like pattern on dorsum.

Preserved specimens of Allobates magnussoni are most easily confounded with A. brunneus ( Cope 1887) , A. crombiei (Morales 2000 “2002”), A. gasconi (Morales 2000 “2002”), A. undulatus (Myers & Donnelly 2001) and A. flaviventris Melo-Sampaio, Souza & Peloso 2013 , because of the presence of a distinct, dark-brown, hourglassshaped pattern on dorsum. Therefore, more detailed comparisons between these and the new species are warranted. Allobates magnussoni differs from Allobates brunneus ( Cope 1887) by the color of throat in males, which is gray in preserved specimens (cream to yellowish in A. brunneus ) and grayish-violet in life (yellow in live A. brunneus ); by females with yellow throat in life (white in A. brunneus ); by males and females with heads of similar width (males have relatively wider heads than females in A. brunneus ); by internarinal distance 39–48% of head width (internarinal distance 38% of head width in A. brunneus ); by a Finger I visibly longer than Finger II (slightly longer in A. brunneus ), Finger II slightly longer than Finger IV (visibly longer in A. brunneus ) ( Lima et al. 2009).

Allobates magnussoni differs from A. crombiei (Morales 2000 “2002”) by the color of throat in males, which is gray in preserved and grayish-violet in life (translucent-white in A. crombiei ; by a ventrolateral stripe absent or diffuse in live or preserved specimens (evident and continuous in A. crombiei ); and by the presence of a distinct oblique lateral bar (absent A. crombiei ) (Lima et al. 2012). Allobates magnussoni differs from A. gasconi (Morales 2000 “2002”) by its larger size (minimum SVL> 16.0 mm in males and> 17.9 mm in females in A. magnussoni ; maximum SVL <17.3 mm in A. gasconi ); by a Finger III not swollen in males (Finger III swollen in male A. gasconi ); by the absence or presence of diffuse ventrolateral and dorsolateral stripes (both present and conspicuous in A. gasconi ), and by the presence of a distinct oblique lateral bar (absent in A. gasconi ).

Allobates undulatus is only known from Venezuelan Amazonia, and is distinguished from A. magnussoni by a larger body-size (minimum SVL of male A. undulatus > 19.6 mm, maximum SVL of male A. magnussoni <19.6 mm) and by a solid black lateral stripe on flank, not limited ventrally by iridescent-white spots or by a diffuse iridescent ventrolateral line (lateral stripe on flank is dark-brown in A. magnussoni , interrupted by a distinct palebrown oblique lateral bar at mid body and limited ventrally by the presence of a diffuse, iridescent-white ventrolateral line, that may appear as a series of irisdescent spots of irregular shape in most specimens).

Allobates magnussoni differs from A. flaviventris by having absent or diffuse ventrolateral stripes (present and continuous in live and preserved males of A. flaviventris ); by the presence of a distinct oblique pale brown lateral bar (short, diffuse, not surpassing the inguinal region in A. flaviventris ); by the presence of basal webbing between Toes II-III and III-IV (basal webbing absent between toes in A. flaviventris ); and by advertisement calls characterized by the continuous emission of single notes, and eventually of note-pairs, always split by regular silent intervals (advertisement calls of A. flaviventris consist predominantly of the emission of note-pairs or groups of 2–10 notes separated by irregular intervals). Morphometric ratios of Allobates magnussoni ’s holotype differ from those of the holotype of A. flaviventris by head width 90% of head length (head width equal to head length in A. flaviventris ); snout 48 % of head length (snout 21.5 % of head length in A. flaviventris ); eye-to-nostril distance ( END) 22% smaller then eye-diameter ( END 25% greater than ED in A. flaviventris ); forearm thicker than upper arm, 75% of upper arm length (forearm slender, 88.3% of upper arm in A. flaviventris ); tibia length 50% and foot length 49.5% of SVL (tibia length 49.7% and foot length 47.9% of SVL in A. flaviventris ).

Description of holotype. An adult male, with SVL of 18.21 mm (holotype measurements presented in Table 1 ). Head slightly wider than long, head length 90% of head width, head width 33% of SVL; snout round to nearly truncate in dorsal view and bluntly rounded in lateral view ( Fig. 1 A, C), tip of snout extending past lower jaw anteriorly ( Fig. 1 D); snout length 48% of head length; internarial distance 44% of head width; distance from eye to nostril 78% of eye length; nares opening posterolaterally; tympanum round, 46% of eye length; tympanic membrane inconspicuous; tongue nearly twice as long as wide, attached anteriorly, rounded along posterior margin; median lingual process absent; vocal sac and vocal slits present ( Fig. 1 B, D). Teeth on premaxilla and maxilla are imperceptible under light microscope at 50X magnification. Dorsum has a distinct dark brown hourglass marking on the anterior to mid dorsal surface, consisting of three contiguous sections: the first between the eyes in the approximate shape of a bat, the second approximately diamond-shaped, and the third similar to the first, but with opposite orientation ( Fig. 1 A, C); the skin is smooth on dorsal and ventral surfaces, with small scattered tubercles on the urostyle region; forearm thicker than upper arm, 75% of upper arm length; HAND III length 26% of SVL; Finger I 0.2 mm longer than Finger II; Finger II slightly shorter than Finger IV (HAND II = 3.5 mm; HAND IV = 3.6 mm); tip of Finger IV reaching past distal subarticular tubercle of Finger III when fingers are juxtaposed; Relative length of fingers: III> I> II = IV; Fingers I and III swollen only basally ( Fig. 2 ); fingers not webbed, with no lateral fringes; palmar tubercle nearly round, 0.65 mm in maximum diameter, 14% of Hand III length; width of thenar tubercle about 53% the diameter of palmar tubercle ( Fig. 2 ); one subarticular tubercle present on Fingers I, II and IV; two subarticular tubercles present on Finger III; distal tubercles on Finger III small ( Fig. 1 F); discs on all fingers moderately expanded, with distinct dorsal scutes; width of disc on Finger III 0.75 mm, width of finger 60% the disc width ( Fig. 2 ).

Hindlimb is robust; leg length 97% of tibia length and 49% of SVL; tibia length 50% of SVL; foot length 49% of SVL; relative length of toes IV> III> V> II> I; basal webbing present only between Toes II-III and III-IV; lateral fringes absent on all toes, Toe I reaching middle of subarticular tubercle of Toe II when appressed; discs on Toes II, III, IV and V larger than width of toes; disc on Toe I slightly wider than width of toe; disc width of Toe IV 0.95 mm; inner metatarsal tubercle oval; outer metatarsal tubercle round; median metatarsal tubercle absent. Metatarsal fold present and weak; tarsal keel tubercle-like and strongly curved at proximal end, located 1.0 mm from proximal edge of inner metatarsal tubercle, connected by a fold ( Fig. 2 ). One subarticular tubercle on Toes I and II; two on Toes III and V; three on Toe IV. Basal subarticular tubercle on Toe IV poorly defined ( Fig. 2 ).

Variation within type series. The first and second dimensions of a Principal Component Analysis (PCA) generated from 15 morphometric ratios do not show differences between groups of specimens collected on opposite sides of the Tapajós River ( Fig. 5 ; T-test: t = -0.47, d.f. = 34, P = 0.60, first component, and t = -0.47, d.f. = 34, P = 0.64, second component). Eight females had SVL significantly larger than 14 males (T-test: t = 5.9, d.f. = 34, P <0.0001). However, there were no statistically significant differences in most of the morphometric ratios between sexes (measurements in Table 1 ). Therefore we describe variation in mean of 36 specimens together, regardless of sex unless otherwise noted. HW 34 ± 0.01 (31–36)% of SVL; SL 48 ± 0.04 (40–56)% of HL; IN 44 ± 0.02 (39–48)% of HW; EN distance 29 ± 0.02 (25–33)% of HW; tympanic membrane inconspicuous, round, 42 ± 0.05 (36–48)% of EL; skin smooth on dorsal and ventral surfaces, with scattered tubercles more concentrated on the urostyle region and present in all specimens ( Fig. 4 ); dorsolateral stripe absent in 15 and diffuse in 21 preserved specimens (as in Fig. 4 ), diffuse or absent in live specimens ( Fig. 1 A, Fig. 3 A, B, C); ventrolateral stripe is absent or diffuse as small irregular iridescent-white spots present on flank in specimens from PNA ( Fig. 1 A,C; Fig. 3 A, G, H). In live and preserved specimens from Treviso and Jamanxim the ventrolateral line is diffuse, appearing as a wavy series of elongate, interconnected iridescent-white spots ( Fig. 3 B, I); an oblique lateral bar is present in all specimens ( Figs. 1 A, C; Fig. 3 A, C, G, H).

Variation of limb proportions among 36 specimens: forearm 79 ± 5 (67–93)% of upper arm; HAND III 26 ± 1.0 (22–28)% of SVL; palmar tubercle nearly round, diameter 14 ± 2.0 (10–18)% of HAND III; subarticular tubercles on fingers similar in size and number to those of the holotype in 27 specimens ( Fig. 2 ), distal tubercles on Finger III in 9 specimens almost inconspicuous. Finger III weakly larger in males; width at middle section of the third phalange on Finger III in males 0.69 ± 0.05 in average, significantly greater than in females (mean 0.63 ± 0.04 mm; T-test, t = -3.2, d.f. = 34, P = 0.003). Finger IV with nearly the same length (99.7 ± 3.8 [91–100]%) of Finger II; length of Finger IV 73 ± 3 (69–83)% the length of Finger I.

Legs slightly longer in males than in females, TL/SVL ratio in males 0.49 ± 0.02, significantly greater than in females (0.47 ± 0.02, T-test t = -3.93, d.f. = 34, P <0.0001); FL/SVL ratio in males 0.46 ± 0.03, significantly greater than in females (0.45 ± 0.02, T-test, t = -2.97, d.f. = 34, P = 0.005); FL 47.8 ± 3.0 (42–54)% of SVL in males and 44.5 ± 2.5 (41–49)% of SVL in females.

Color of adults in preservative. Color of the holotype in preservative is shown in Fig. 1 C, D. The following description of color in preservative was based on 36 specimens (25 males and 11 females). In 30 specimens, the background surface of dorsum is generally brown, pale brown in five specimens (as in Fig. 4 A–F) and dark brown in one specimen from Jamanxim. In 31 specimens, a distinct dark-brown hourglass-shaped marking appears on dorsum, extending from between the orbits to mid-body ( Fig. 1 C, Fig. 3 A–F). In five specimens, this mark is lighter ( Fig. 4 C). In 21 specimens, a dorsolateral stripe is diffuse and extends from above the eye to mid-body ( Fig. 4 C, E, F). In 15 specimens, it is inconspicuous (similar to the holotype, Fig. 1 C, and a female, Fig. 4 A). The upper surface of the arm is cream in 21 specimens and brown in 15 specimens. The dorsal surface of leg and foot same color of dorsum, with a dark-brown transversal bar at the middle section of thigh and shank. Transverse bars line up when thighs and shanks are juxtaposed ( Fig. 4 A–F). In all specimens, the dorsum of foot is the same color of dorsum with two or three diffuse dark brown stripes ( Fig. 4 A–F). Ventrolateral stripe is absent or diffuse as small irregular white spots on flank in 30 specimens (as in the holotype, Fig. 1 C). In six specimens from Treviso and Jamanxim it appears as a wavy series of elongate interconnected white spot. The lateral surface below the darkbrown lateral band is cream with whitish spots in some specimens (as in the hototype, Fig. 1 C). The dark-brown lateral band is interrupted by a distinct oblique light-brown bar at mid-body. Tympanum is dark-brown at its dorsal half and cream ventrally in all specimens. Reproductive females have a cream throat, chest, and belly without melanophores; males have light to dark vocal sac and upper chest ( Fig. 3 G–I). A pale paracloacal mark is present, cream colored, small, not moon-shaped, in all specimens ( Fig. 4 A–F).

Color in life. Color in life of the holotype is shown in Fig. 1 (A, B). Background color of dorsum is uniform light-brown. All specimens have a distinct dark-brown hourglass-like marking on the anterior dorsal surface, consisting of three contiguous sections: the first between the eyes, in the form of a bat, the second diamond-shaped, and the third triangular or similar to the first section ( Fig.1 A; Fig. 3 A–C). A dorsolateral stripe is absent or diffused with the background color of dorsum ( Fig. 1 A; Fig. 3 A–C). Upper surface of arm is light-brown. Dorsal surface of leg is light-brown with a dark-brown transversal bars about the middle section of thigh and shank, which line up when individuals are sitting at rest ( Fig. 1 A; Fig. 3 A–C, G); foot has two transverse dark-brown bars, sometimes diffuse. The throat of adult calling males is dark grayish-violet (as in Fig. 2 F), acquiring a lighter grayish-violet shade in males captured while not emitting calls (as in Fig. 1 A and 2E). Melanophores are densely scattered on the vocal sac, throat and chest regions. In males the abdominal region is white to translucent anteriorly, bright yellow posteriorly and on flanks. Ventral surface of limbs also bright yellow among specimens from PNA and dull yellow among specimens observed in Treviso and Jamanxim ( Fig. 1 B; Fig. 3 E, F). Adult females have a uniformly bright yellow throat, chest and abdomen, lacking melanophores on ventral surface of limbs ( Fig. 3 D, E). Both sexes (30 specimens) have small, iridescent white irregular spots on the ventrolateral region, on pale yellow background ( Fig.1 A; Fig. 3 G, H). In six specimens from Treviso and Jamanxim it is diffuse as a wavy series of elongate interconnected iridescent white spots (as in the Fig. 3 B, I). The base of the thigh has a small cream paracloacal mark, not moon-shaped, in all specimens. Palmar and plantar surfaces are heavily pigmented, dark-brown ( Fig. 3 D, F).

Color of metamorphs in life resembles that of adults, with similar drawing patterns on dorsum ( Fig. 12 ), indicating that this character is inherited, does not vary ontogenetically and could be used with confidence as a diagnostic trait.

Description of vocalization. The most frequent arrangement of advertisement calls of Allobates magnussoni consist of the continuous emission of short tonal notes, separated by regular silent intervals ( Fig 6 A). One additional arrangement was recorded (see below). The prevalent arrangement of advertisement calls is formed by the emission of short tonal notes, with slightly ascendant frequency modulation (difference between lowest and highest frequency of notes = 640.3 ± 90.8 Hz, in average—Fig. 6B) at an average rate of 1.9 ± 0.3 notes per second ( Table 2). At the species type locality, notes are 0.08 ± 0.01 s long in average (n = 325 notes, analyzed from calls of 13 males—Table 2). Notes have an average lowest frequency of 4089.8 ± 113.4 Hz and an average highest frequency of 4733.1 ± 141.2. Hz. Average peak frequency of notes is 4462.8 ± 141.6 Hz. Silent intervals between notes range from 0.35 to 0.53 s (0.45 ± 0.06 s, in average).

Holotype PNA Treviso Allobates brunneus

Nº of recorded males 1 13 8 5

Notes analyzed 25 325 200 125

Temperature (°C) 25.5 23.9–27.4 23.0–26.5 25.5–26.7

SVL (mm) 18.2 16.4–18.8 17.4–18.7 14.8–18.3

ND (s) 0.065 0.080 ± 0.011 0.064 ± 0.012 0.041 ± 0.011

(0.065–0.104) (0.047–0.083) (0.027–0.056) NPF (Hz) 4366.9 4462.8 ± 141.6 4635.8 ± 104.6 5161.3 ± 216.8

(4273.0–4704.6) (4497.8–4867.3) (4940.5–5447.9) NLF (Hz) 4122.2 4089.8 ± 133.4 4182.2 ± 127.7 4753.1 ± 166.1

(3809.1–4303.7) (4025.5–4430.8) (4601.7–4960.1) NHF(Hz) 4680.6 4733.1 ± 141.1 4916.6 ± 145.7 5461.2 ± 203.1

(4484.9–4961.4) (4721.9–5132.8) (5188.6–5662.9) NSI (s) 0.347 0.451 ± 0.062 0.509 ± 0.161 0.390 ± 0.029

(0.347–0.535) (0.369–0.811) (0.357–0.422)

RNE (notes/s) 2.4 1.90 ± 0.28 1.89 ± 0.49 2.29 ± 0.11

(1.63–2.40) (1.30–2.56) (2.13–2.43) The acoustic arrangement described above is Allobates magnussoni ’s main advertisement call, constituting most of the recording length registered for all males. However, one additional arrangement was recorded among calls of fewer individuals.

Males emit advertisement calls formed by note-pairs ( Fig. 6 C,D), shifting to regular, single-note calls after approximately 10–20 s (INPA-H 32962, 32963, 10108, 10110 and two additional males, not collected). Detailed characterization of these calls is presented in Table 3. Isolated note-pairs appeared in recordings of two additional males, scattered among continuous single note calls, but were not considered in the characterization due to their rarity along recording length (less than one note pair per 50 single notes). Among calls of males recorded in PNA, silent intervals between notes in each pair vary from 0.07 to 0.12 s (0.10 ± 0.02 s in average, n = 40 intervals, measured from calls of four males), whereas intervals splitting consecutive note-pairs vary from 0.43 to 0.63 s (0.52 ± 0.09 s in average, n = 40 intervals, measured from calls of four males). The rate of note emission is faster during production of note-pairs than along production of regular calls (average note emission rate is 2.6 ± 0.3 notes per second during the emission of note-pairs—estimated from 10 s long sections of recordings of each of the four males that produced this type of arrangement).

First and second notes in each pair are distinct in relation to temporal and spectral parameters. First notes are 0.09 ± 0.01 s long in average (n = 40 notes, analyzed from calls of four males). Average peak, lowest and highest frequencies of the first notes were 4461.9 ± 126.1 Hz, 4131.5 ± 57.3 Hz and 4788.6 ± 86.7 Hz, respectively. Second notes are 0.05 ± 0.01 s long in average. Average peak, lowest and highest frequencies of the second notes were 4526.0 ± 61.3 Hz, 4231.3 ± 53.1 Hz and 4737.6 ± 83.9 Hz, respectively.

Comparisons with calls of specimens morphologically similar to topotypical A. magnussoni revealed that temporal and spectral acoustic traits of specimens proceeding from Treviso largely overlap with those registered at PNA, considering continuous advertisement calls ( Table 2) and calls formed by note-pairs ( Table 3). Call arrangement is also similar between recordings obtained in both locations ( Fig. 7 A–C).

Continuous advertisement calls of typical Allobates brunneus are characterized by the emission of tonal notes with ascending frequency modulation (708.0 ± 109.7 Hz in average—Fig. 7E). Notes are shorter when compared with those of A. magnussoni , and are emitted at a higher frequency bandwidth, not overlapping in terms of lowest, highest or peak frequencies ( Table 2). Along some recordings, continuous calls are interrupted by the emission of note trills ( Fig. 7 F), consisting in the second most common arrangement among A. brunneus calls.

PNA Fazenda Treviso Allobates flaviventris

Nº of recorded males 4 3 4

Nº of note-pairs analyzed 40 30 40

T (°C) 23.9–25.4 23.0–26.5 25.0

SVL (mm) 16.3–17.0 15.5–16.9 17.8–19.0

1ND (s) 0.085 ± 0.008 0.088 ± 0.011 0.052 ± 0.002 (0.075–0.093) (0.075–0.096) (0.050–0.054)

2ND (s) 0.052 ± 0.003 0.063 ± 0.007 0.039 ± 0.002 (0.049–0.055) (0.056–0.071) (0.037–0.042)

1NPF(Hz) 4461.9 ± 126.1 4691.3 ± 113.5 4369.8 ± 114.0

(4310.9–4577.9) (4584.4–4810.5) (4291.6–4539.2)

1NLF (Hz) 4131.5 ± 57.3 4279.6 ± 94.4 3843.4 ± 127.3

(4086.1–4209.2) (4179.6–4367.2) (3731.9–4016.8)

1NHF (Hz) 4788.6 ± 86.6 5064.7 ± 16.19 4735.7 ± 223.9

(4720.9–4913.1) (5049.4–5081.7) (4569.3–5062.3)

2NPF (Hz) 4526.0 ± 61.3 4676.3 ± 124.9 4412.0 ± 137.0

(4447.8–4586.6) (4552.1–4801.9) (4323.9–4614.5)

2NLF (Hz) 4231.3 ± 53.1 4365.1 ± 115.7 4026.4 ± 166.9

(4172.8–4299.2) (4231.7–4438.3) (3894.3–4270.3)

2NHF (Hz) 4737.6 ± 83.9 5100.8 ± 21.6 4809.9 ± 225.1

(4641.6–4846.2) (5078.5–5121.6) (4660.9–5142.7)

NSIL (s) 0.100 ± 0.021 0.078 ± 0.015 0.033 ± 0.003 (0.077–0.127) (0.066–0.095) (0.029–0.037)

SIL (s) 0.523 ± 0.087 0.576 ± 0.237 0.369 ± 0.043 (0.425–0.630) (0.359–0.829) (0.313–0.416)

RNE (notes/s) 2.62 ± 0.33 2.52 ± 0.65 4.14 ± 0.68

(2.30–3.00) (1.90–3.20) (3.36–4.93) Calls of Allobates flaviventris proceeding from Rio Branco consist of a multiple series of note-pairs split by irregular silent intervals ( Fig. 7 D). In relation to note-pairs emitted by A. magnussoni , those produced by A. flaviventris are shorter in length, reflecting shorter note durations and inter-note silent intervals. Notes are also emitted at faster call rates (over four notes per second, in average—Table 3). Notes are emitted by A. flaviventris at a lower frequency bandwidth, but overall ranges of peak, lowest and highest frequencies overlap with those of A. magnussoni ( Table 3).

Summary statistics of acoustic parameters derived from individual male mean values are provided in Appendix IV (for A. magnussoni and A. brunneus males that emitted continuous calls) and Appendix V (for A. magnussoni and A. flaviventris males that emitted note-pairs).

Description of tadpoles. Descriptive statistics for 17 morphological characters were based on 66 tadpoles at developmental stages 25, 27, 36 and 38 ( Table 4). The following detailed description is based on a sample of 5 tadpoles from the same lot (INPA-H 32983, original field number APL13031), collected from a single clutch and reared to stage 37 before preservation.

Body is ellipsoid in dorsal view, slightly flattened in lateral view ( Fig. 8 ). Body 30% and tail 70% of total length; body wider than deep, BH 34% of BW at level of spiracle; HWLE at level of eyes 92.6% of BW at the level of spiracle; snout bluntly rounded in dorsal and lateral views; END 74% of ED; eyes dorsal and directed laterally; ED 1.01 ± 0.02 mm; IOD 24% of HWLE at the level of the eyes. Small nostrils located dorsolaterally and directed anterolaterally; internarial distance 1.12 mm. A fleshy ring is present on the inner margin of nostrils, round and straight, not ornamented. Spiracle is sinistral, forming a free tube of 1.04 mm, slightly anterior to mid-section of body ( Fig. 8 ; Fig. 9 B). Vent tube is 1.24 ± 0.02 mm long, opening dextrally. Gut long coiled, concealing other organs in ventral view, with coiling axis directed to the left side, located near the spiracle. Fins have a shallow arch. Dorsal fin begins 3.87 mm up onto body ( Fig. 5 A). Caudal musculature is bifurcated over body, with dorsal fin extending shortly between them. Dorsal and ventral fins have similar height at mid-tail ( Fig. 8 ).

Oral apparatus is emarginate, located anteroventrally, 1.97 ± 0.07 mm in width ( Fig. 9 A, B, C); anterior labium with short round papillae only on lateral margin, 12 or 13 on each side; posterior labium projected anteriorly, obstructing view of posterior tooth rows in all specimens if not folded manually ( Fig. 9 B); posterior labium entirely surrounded by 32 to 35 short, pyramidal to slightly elongate papillae; papillae on posterolateral margin same size as the posteromedial papillae ( Fig. 9 A, C); submarginal papillae absent; lower jaw sheath V-shaped, deeper than upper jaw sheath; upper and lower sheaths serrated; serrations extend entire length of sheaths; labial tooth row formula 2(2)/3(1); tooth row A-1 complete, 1.25; tooth row A-2 interrupted medially, consisting of two widely separated rows at level of upper jaw sheath, segment 0.65 mm in length, with a gap of 0.38 mm ( Fig. 9 A, C). Lower tooth rows longer than A-1; P-1 and P-2 same length (1.50 mm) and longer than P-3 (0.75 mm).

Character Stage 25 (n=11) Stage 27 (n=19) Stage 36 (n=25) Stage 38 (n=11)

(N = 11) (N = 3) (N = 10) (N = 5)

1 -TL 15.77 ± 1.45 17.91 ± 2.47 20.66 ± 0.28 20.32 ± 0.14 (14.33–18.50) (12.33–21.17) (20.17–21.17) (20.00–20.48) 2 - BL 5.66 ± 0.59 6.43 ± 1.02 8.00 ± 0.63 7.52 ± 0.63 (4.90–6.50) (3.90–7.50) (6.00–8.70) (6.50–8.50) 3 -TAL 10.16 ± 1.06 11.36 ± 1.85 15.04 ± 0.57 14.69 ± 1.27 (9.13–12.13) (7.88–14.38) (13.38–15.44) (12.13–15.63) 4 - BW 3.89 ± 0.52 4.31 ± 0.81 5.93 ± 0.46 5.58 ± 0.40 (3.35–4.95) (3.00–5.85) (4.00–6.25) (4.85–6.00) 5 - BH 1.65 ± 0.16 1.82 ± 0.32 2.05 ± 0.24 1.94 ± 0.32 (1.43–1.97) (1.33–2.50) (1.47–2.67) (1.50–2.50) 6 - HWLE 3.81 ± 0.47 4.14 ± 0.74 5.61 ± 0.26 5.38 ± 0.64 (3.10–4.60) (2.50–5.20) (5.00–6.00) (4.00–6.10) 7 - TMW 1.64 ± 0.25 1.66 ± 0.39 2.58 ± 0.21 2.56 ± 0.21 (1.30–2.07) (0.83–2.33) (2.03–2.97) (2.13–2.80) 8 - MTH 2.89 ± 0.39 2.73 ± 0.62 3.86 ± 0.39 3.54 ± 0.48 (2.43–3.47) (1.40–3.60) (2.67–4.17) (2.67–4.00) 9 - TMH 1.58 ± 0.23 1.53 ± 0.33 2.32 ± 0.18 2.20 ± 0.25 (1.37–2.00) (0.73–1.90) (1.83–2.53) (1.70–2.50) 10 - IOD 0.93 ± 0.08 1.04 ± 0.17 1.33 ± 0.05 1.32 ± 0.08 (0.83–1.03) (0.67–1.23) (1.17–1.47) (1.10–1.40) 11 - IND 0.70 ± 0.10 0.73 ± 0.16 0.98 ± 0.15 0.94 ± 0.12 (0.60–0.83) (0.33–1.00) (0.67–1.27) (0.73–1.17) 12 - END 0.54 ± 0.14 0.55 ± 0.13 0.74 ± 0.10 0.72 ± 0.11 (0.33–0.87) (0.27–0.73) (0.50–0.97) (0.60–0.97) 13 - NSD 0.62 ± 0.09 0.53 ± 0.19 0.67 ± 0.13 0.55 ± 0.14 (0.50–0.77) (0.20–0.83) (0.37–0.83) (0.33–0.77) 14 - ED 0.62 ± 0.09 0.74 ± 0.26 0.80 ± 0.10 0.95 ± 0.10 (0.50–0.77) (0.43–1.73) (0.67–0.93) (0.83–1.07) 15 - VTL 0.72 ± 0.20 0.65 ± 0.27 1.12 ± 0.26 0.96 ± 0.29 (0.50–1.10) (0.30–1.25) (0.50–1.45) (0.50–1.25) 16 - STL 0.83 ± 0.34 0.79 ± 0.22 1.05 ± 0.15 0.95 ± 0.11 (0.50–1.75) (0.30–1.25) (0.75–1.30) (0.75–1.10) 17 - ODW 1.44 ± 0.19 1.34 ± 0.46 1.90 ± 0.11 1.74 ± 0.20 (1.15–1.75) (0.25–1.85) (1.65–2.05) (1.35–2.05) Comparison with tadpoles of other species. The tadpoles of Allobates caeruleodactylus , A. marchesianus , A. grillisimilis , A. brunneus differ from those of A. magnussoni in having fewer papillae on lateral labium and distinctively longer papillae on posterior labium (short and numerous in A. magnussoni ). Allobates caeruleodactylus and A. marchesianus have distinct dark transversal bars on tail (absent in A. magnussoni ) ( Caldwell et al. 2002a). Labial tooth rows P-1, P-2 and P-3 are sub-equal in tadpoles of A. sumtuosus (Simões & Lima 2012); row P-3 is lacking in tadpoles of A. granti (Kok et al. 2006) (all rows present in A. magnussoni , and P- 3 shorter than P-1 and P-2); row A-2 shorter than A- 1 in tadpoles of A. brunneus ( Lima et al. 2009) (A-2 longer than A- 1 in A. magnussoni ). Tadpoles of A. paleovarzensis have a distinct dark brown longitudinal bar extending from the snout to the eye, and towards mid-body (Lima et al. 2010), which is lacking in A. magnussoni ( Fig. 8 ; Fig. 9 ). The tadpoles of Allobates nidicola and Allobates masniger are endotrophic and develop entirely in a terrestrial nest; these highly modified tadpoles lack oral discs and spiracles ( Caldwell & Lima 2003; APL, unpublished data).

Color of tadpoles. Background color of dorsal, lateral and anteroventral surfaces of body cream, with small brown spots formed by the aggregation of melanophores ( Fig. 8 ). In all specimens, a region of darker brown appears on dorsum between the diverging sections of caudal musculature. Posteroventral surface of body is transparent and immaculate; the intestines are visible through the skin. Tail muscle is cream; tail fins are transparent with scattered irregular brown melanophores ( Fig. 8 ). In life, pale iridescent blotches are present on dorsal and lateral surfaces of body. Pale iridescent spots also appear scattered along tail ( Fig. 9 D–F).

Genetic distances and mtDNA relationships. Based on evolutionary analyses using a 481 bp fragment of the 16S RNA mitochondrial gene, typical Allobates magnussoni and samples proceeding from Treviso form a wellsupported monophyletic clade in relation to the remaining cryptically colored Allobates species occurring in Brazil, and across the Guiana Shield and eastern Peruvian Amazon ( Fig. 10 ). Mean pairwise uncorrected genetic distances or K2P distances between samples proceeding from PNA and Treviso do not exceed 0.2% (Table 5). Samples proceeding from the Madeira River and Guajará-Mirim, tentatively identified as A. flavientris , form the sister clade to A. magnussoni . Mean pairwise genetic distances between these clades and typical A. magnussoni range from 4 to 6% (5–6% when considering K2P distances).

Genetic distances between typical A. magnussoni and remaining Allobates species range from 7 to 15% (7 to 16% considering K2P distances), minimum genetic distances being estimated between A. magnussoni and A. crombiei and maximum distances estimated between A. magnussoni and A. olfersioides . Allobates crombiei sampled on both margins of the Xingu River form a well-supported monophyletic clade, mean pairwise genetic distances between samples from opposite banks not exceeding 1%. However, haplotype sharing does not occur between riverbanks. Mean pairwise genetic distances between A. magnussoni and A. crombiei vary from 7 to 8% considering both uncorrected and K2P-corrected estimates.

Mean genetic distance between clades formed by sympatric Allobates magnussoni and A. masniger , both collected at the same region of Parque Nacional da Amazônia, surpass 9% (and over 10% when considering K2P distances). In Treviso, A. magnussoni co-occurs with a smaller undescribed species (“ Allobates aff. marchesianus ”, which also occurs in PNA. See Introduction). In average, genetic distances between A. magnussoni and the latter species exceed 9% in this location.

TABLE 5. Uncorrected pairwise (lower left) and Kimura-2-Parameter (upper right) genetic distances between Allobates magnussoni and clades representing Allobates species distributed in Brazil and across the Guiana Shield and Peruvian Amazon regions. Distances were based in a 481 pb fragment of the 16S rDNA mitochondrial gene. Total 403 positions were used in distance calculations after elimination of positions presenting gaps or missing data in over 10% of sequences in the final alignment.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 magnussoni 0.00 0.06 0.05 0.06 0.08 0.07 0.10 0.09 0.11 0.08 0.08 0.12 0.09 0.10 0.08 0.16 0.09 0.09 0.11 0.09 0.09 0.12 magnussoni (Treviso) 0.00 0.06 0.05 0.06 0.08 0.08 0.10 0.09 0.11 0.08 0.08 0.12 0.09 0.10 0.08 0.16 0.09 0.09 0.11 0.09 0.09 0.12 cf. flaviventris 0.06 0.06 0.01 0.00 0.09 0.10 0.10 0.09 0.10 0.09 0.08 0.10 0.10 0.10 0.09 0.18 0.09 0.09 0.10 0.09 0.08 0.14 Guajará-Mirim)

cf. flaviventris (Jaci- 0.04 0.05 0.01 0.02 0.09 0.09 0.10 0.08 0.10 0.07 0.08 0.10 0.09 0.09 0.08 0.17 0.09 0.08 0.10 0.08 0.08 0.13 Paraná)

cf. flaviventris 0.06 0.06 0.00 0.02 0.10 0.10 0.09 0.08 0.10 0.09 0.08 0.09 0.10 0.10 0.09 0.18 0.09 0.09 0.10 0.09 0.08 0.13 Cachoeira do Jirau)

crombiei 0.08 0.08 0.09 0.08 0.09 0.01 0.10 0.09 0.11 0.09 0.08 0.09 0.10 0.11 0.08 0.16 0.09 0.08 0.10 0.07 0.08 0.11 crombiei (Fazenda 0.07 0.07 0.09 0.08 0.09 0.01 0.11 0.10 0.11 0.09 0.08 0.09 0.10 0.11 0.09 0.17 0.09 0.08 0.11 0.07 0.08 0.12

de Sol)

algorei 0.09 0.09 0.09 0.09 0.09 0.09 0.10 0.10 0.11 0.10 0.07 0.12 0.10 0.12 0.12 0.19 0.10 0.09 0.11 0.09 0.07 0.14 caeruleodactylus 0.08 0.08 0.08 0.08 0.08 0.09 0.09 0.09 0.11 0.10 0.10 0.11 0.11 0.10 0.10 0.17 0.10 0.08 0.10 0.10 0.10 0.13. conspicuus (Rio 0.10 0.10 0.10 0.09 0.09 0.10 0.10 0.10 0.10 0.11 0.09 0.13 0.12 0.13 0.10 0.17 0.10 0.09 0.02 0.10 0.09 0.16 Juruá)

. gasconi (Eirunepé) 0.08 0.08 0.08 0.07 0.08 0.08 0.08 0.09 0.09 0.10 0.08 0.11 0.06 0.10 0.09 0.18 0.08 0.09 0.10 0.09 0.08 0.13. gasconi (Porto 0.08 0.08 0.07 0.08 0.07 0.07 0.07 0.06 0.10 0.08 0.07 0.10 0.08 0.10 0.09 0.17 0.08 0.07 0.09 0.06 0.01 0.13 Walter)

. grillisimilis 0.11 0.11 0.09 0.09 0.09 0.09 0.08 0.11 0.10 0.11 0.10 0.09 0.11 0.12 0.11 0.18 0.11 0.11 0.13 0.11 0.10 0.13. aff. marchesianus 0.09 0.09 0.09 0.08 0.09 0.09 0.09 0.09 0.10 0.11 0.05 0.07 0.10 0.12 0.09 0.19 0.10 0.08 0.12 0.09 0.08 0.14 Treviso)

. masniger 0.09 0.09 0.09 0.09 0.09 0.10 0.10 0.11 0.10 0.12 0.09 0.09 0.11 0.11 0.07 0.21 0.13 0.12 0.13 0.11 0.10 0.14. nidicola 0.07 0.08 0.08 0.07 0.08 0.08 0.08 0.11 0.09 0.09 0.09 0.08 0.10 0.09 0.06 0.17 0.09 0.09 0.10 0.08 0.09 0.12. olfersioides (Bahia) 0.14 0.15 0.16 0.15 0.16 0.14 0.15 0.16 0.15 0.15 0.16 0.15 0.16 0.17 0.18 0.15 0.18 0.17 0.17 0.17 0.17 0.16. ornatus 0.09 0.09 0.08 0.08 0.08 0.08 0.08 0.09 0.09 0.09 0.08 0.07 0.10 0.09 0.11 0.08 0.16 0.09 0.10 0.09 0.09 0.15. paleovarzensis 0.08 0.08 0.08 0.08 0.08 0.07 0.08 0.09 0.08 0.09 0.08 0.06 0.10 0.08 0.11 0.08 0.15 0.08 0.08 0.05 0.07 0.12. subfolionidificans 0.10 0.10 0.09 0.09 0.09 0.10 0.10 0.10 0.10 0.02 0.09 0.08 0.12 0.11 0.12 0.09 0.15 0.09 0.08 0.09 0.09 0.15. sumtuosus 0.08 0.08 0.08 0.08 0.08 0.07 0.07 0.08 0.10 0.09 0.08 0.06 0.10 0.08 0.10 0.08 0.15 0.08 0.05 0.09 0.07 0.13. trilineatus 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.07 0.10 0.08 0.08 0.01 0.09 0.08 0.10 0.09 0.15 0.08 0.07 0.08 0.06 0.12 Departamento

Huanuco)

. undulatus 0.11 0.11 0.12 0.12 0.12 0.11 0.11 0.13 0.12 0.14 0.12 0.12 0.12 0.12 0.13 0.11 0.15 0.14 0.11 0.13 0.12 0.11 Samples of Allobates gasconi proceeding from the upper (Porto Walter, Acre State, Brazil) and middle (Eirunepé, Amazonas State, Brazil) courses of the Juruá River are not monophyletic ( Fig. 10 ), mean genetic distance between clades exceeding 7%. The most inclusive clades had extremely low support from bootstrap analyses and should not be regarded to reflect the true evolutionary relationships among the Allobates samples used.

Distribution and natural history. Allobates magnussoni is known from the Tapajós River basin, in forested areas on both riverbanks along the river’s central and lower course. The type locality consists of forests not subject to seasonal flooding (terra-firme forests) near headwater streams on the western bank of the Tapajós River within the limits of Parque Nacional da Amazônia, south of the city of Itaituba, Pará State, Brazil ( Fig. 11 ). The species is also known to occur on the east bank of the Tapajós River in forest areas managed for timber logging, south of Santarém (Treviso). The new species was not detected in areas surveyed by us or by other research groups along the interfluve between the Madeira and Purus rivers, or along the eastern bank of the Madeira River ( Fig. 11 ). Therefore, A. magnussoni appears to be restricted to the Tapajós River basin, south of the Amazon River.

Like other species of Allobates in the region, individuals are generally active early in the morning and late in the afternoon. Most calling males can be found from 05:30 to 09:30h and again from 16:30h until dusk. Males call partially hidden beneath dead leaves in the forest floor, and are frequently found close to egg clutches.

At type locality two nests were found in the leaf litter on folded leaves. One clutch had 18 embryos and the second only 3 tadpoles in Gosner developmental stage 25. In both clutches, the jelly was transparent ( Fig. 12 A, B). The clutch with 18 embryos is thought to belong to A. magnussoni because a male was observed fleeing the nest. Tadpoles were also found in small headwater streams, with margins inhabited by a large number of adult A. magnussoni . Tentative broadcasts of a recorded advertisement call towards ten male specimens elicited territorial behavior, with males advancing toward the sound speakers.

Courtship behavior of A. magnussoni was registered in the morning of 24 February 2002 at Treviso. A female approached a calling male, which left from below a folded leaf and moved towards her. The male reentered the leaf followed by the female at 07:50h. The male was seen leaving the leaf 45 min later (08:35h), while the female left at 08:53h. A clutch of 13 eggs was found within the leaf. At Treviso, 10 additional nests were found in the leaf litter, eight of which were empty (only jelly). The remaining clutches had 5 and 7 tadpoles, respectively. Both male (n = 4) and female (n = 1) tadpole transport behavior were observed. Males transported from two to seven tadpoles ( Fig. 12 C), while the female was seen transporting six tadpoles on its back.

Figures

FIGURE 1. (A) Dorsolateral and (B) ventral views in life of the holotype of Allobates magnussoni (INPA-H 32960). (C) Dorsolateral and (D) ventral views of the same specimen after preservation in 70 % ethanol. Photographs by A. P. Lima and W. E. Magnusson.

FIGURE 2. Hand and foot of Allobates magnussoni ’ s holotype (INPA-H 32960). Scale bars correspond to 1.0 mm. Black arrows indicate basal webbing between Toes II and III and between Toes III and IV.

FIGURE 3. Color in life of Allobates magnussoni. (A) Dorsolateral view of a female (INPA-H 32974) from the species type locality in Parque Nacional da Amazônia (PNA), on the west bank of Tapajós River; (B) dorsolateral view of a female from Treviso and (C) of a male specimen from Jamanxim (INPA-H 33935), both on the east bank of the Tapajós River. Note variation in dorsal color pattern and diffuse dorsolateral lines. (D), (E), (F): Ventral view of a female specimen from PNA, of a male and a female specimen from Treviso, and of a male specimen from Jamanxim, respectively. Note uniformly yellow ventral surfaces of female specimens and the brownish-violet throat and chest on males. (G) and (H): Lateral views of female and male specimens (INPA-H 32974 and 32976) from PNA. (I) Lateral view of a male (INPA-H 33935) from Jamanxim. Note variation in patterns of the diffuse ventrolateral stripe, formed by round or elongate iridescent-white spots. Photographs by A. P. Lima.

FIGURE 4. Preserved females (A – C) and males (D – F) of Allobates magnussoni showing variation in dorsal pattern. Note brown background color with a distinct dark-brown hourglass-shaped marking on the anterior dorsal surface consisting of three contiguous sections: the first between the eyes in the form of a bat (all specimens), the second diamond shaped (all specimens), and the third triangular (A – C) or trapezoidal (D – F). Note diffuse dorsolateral stripe extending from behind the eye (A – F). Color of ventral surface is similar among preserved females (G – I) but chest and vocal sacs of males preserved males vary from very light to dark-gray (F = female, M = male, numbers correspond to original field numbers).

FIGURE 5. Distribution of 36 specimens of Allobates magnussoni (males and females pooled) along the first and second dimensions of a Principal Component Analysis (PCA) generated from 15 morphometric ratios commonly used in Allobates species descriptions (see text). Note that samples from opposite sides of the Tapajós River do not form distinct clusters (P = Parque Nacional da Amazônia, the species type locality on west riverbank; T = Treviso and J = Jamanxim, both located on the east riverbank).

FIGURE 6. Oscillograms (upper graphs) and sonograms (lower graphs) of advertisement calls of Allobates magnussoni recorded at the species type locality in Parque Nacional da Amazônia, Itaituba, Pará State, Brazil. (A) Ten seconds of regular continuous advertisement calls of a male (INPA-H 32968, SVL = 18.4 mm) recorded at 24.4 ° C. Notes are split by short and regular silent intervals. (B) Detailed view of two notes from the same call session, illustrating the slight ascending frequency modulation. (C) Advertisement calls of a second male (INPA-H 32964, SVL = 17.0 mm), recorded at type locality at 23.9 ° C, while emitting calls with an alternative arrangement, with notes grouped in pairs. (D) Detailed view of the same call session demonstrating that the first note of each pair is longer and has a wider frequency bandwidth than the second note.

FIGURE 7. Comparisons of advertisement calls of Allobates magnussoni with those of similar species by means of 5 s long oscillograms (upper graphs) and sonograms (lower graphs). (A) Continuous emission of single notes split by regular silent intervals, the most frequent call arrangement sampled for A. magnussoni at the species type locality in Parque Nacional da Amazônia (PNA), Pará State, Brazil (INPA-H 32962, 16.8 mm SVL, recorded at 24.4 ° C). (B) Emission of note-pairs split by regular silent intervals, an alternative call arrangement produced sporadically by a few males sampled in PNA and in Treviso, an additional location on the east bank of the Tapajós River (unvouchered male recorded in PNA at 24.5 ° C). (C) Call formed by the continuous emission of a single note, produced by a male recorded in Treviso (INPA-H 10108, 15.5 mm SVL, recorded at 23.0 ° C). Note similarity with calls presented in (A). (D) Advertisement calls of one Allobates flaviventris male recorded at Fazenda Experimental Catuaba, in Acre State (18.2 mm SVL, recorded at 25.0 ° C). Note shorter intervals between consecutive note-pairs and between notes in the same pair, when compared to those emitted by A. magnussoni (B). (E) and (F) Advertisement calls of two Allobates brunneus males recorded at the species type locality in Chapada dos Guimarães, Mato Grosso State. Notes produced continuously by A. brunneus (E) are shorter and emitted at higher frequency bandwidth when compared to those produced by A. magnussoni (A, B); the former also emits calls in short trills formed by 6 – 11 notes at air temperatures above 28 ° C (F). (G) Advertisement call of Allobates crombiei recorded at the species type locality at Cachoeira do Espelho, on the east bank of the Xingu River (17.6 mm SVL, recorded at 25.5 ° C). (H) Advertisement call of Allobates masniger recorded in PNA (18.2 mm SVL, recorded at 24.3 ° C) where it occurs syntopically with A. magnussoni.

FIGURE 8. From top to bottom: dorsal, ventral and lateral views of a tadpole of Allobates magnussoni from Parque Nacional da Amazônia, proceeding from egg clutches collected in February 2008 and reared to stage 37 before preservation (INPA-H 32983) used in the detailed description. Scale bar corresponds to 5.0 mm.

FIGURE 9. (A) Ventral and (B) oblique views of the oral disc of a tadpole of Allobates magnussoni at developmental stage 37 proceeding from the species type-locality in Parque Nacional da Amazônia, Pará State (INPA-H 32983). Note orientation of posterior labium, projected onwards and partially hiding posterior tooth rows from sight in ventral view. (C) Ventral view of the oral disc of an additional tadpole of the same lot, emphasizing that posterior labium orientation is constant among tadpoles. Scale bars correspond to 0.5 mm. (D) Dorsal, (E) ventral and (F) lateral views of a live tadpole of A. magnussoni.

FIGURE 10. Evolutionary relationships among Allobates magnussoni and other species of Allobates occurring in Brazil and across the Guiana Shield region, in northern Amazonia as inferred from an unrooted maximum likelihood phylogenetic tree based on a 481 bp fragment of the 16 S rDNA mitochondrial gene. Clades highlighted in different colors correspond to species morphologically similar to A. magnussoni (A. crombiei, A. cf. flaviventris, A. gasconi), all of which present a dark-brown hourglass-shaped mark on dorsum. Clade labels indicate support values obtained from 5000 bootstrap replicates (only support values> 70 % are shown).

FIGURE 11. Distribution of Allobates magnussoni and position of type localities of morphologically similar species in genus Allobates in northern South America. Yellow stars: (1) Allobates magnussoni ’ s type locality, defined as Parque Nacional da Amazônia, on the left bank of the Tapajós River, south of the city of Itaituba, Pará State, Brazil. (2) and (3): Additional sampling records of A. magnussoni in Treviso and Jamanxim, both located on the right bank of the Tapajós River. Yellow circles, triangles and square indicate approximate geographic locations of precedence of voucher specimens used in the descriptions and re-descriptions of Allobates brunneus, A. flaviventris and A. crombiei, respectively. These three species share with A. magnussoni the presence of hourglass-shaped marks on dorsum. White dots correspond to permanent sampling sites subject to anuran surveys both by our research group and by Programa de Pesquisa em Biodiversidade (detailed information on location of sampling sites and biological surveys available at http: // ppbio. inpa. gov. br) where the new species has not been detected.

FIGURE 12. Allobates magnussoni’s natural history. (A) Egg-clutch photographed immediately after oviposition. Note transparent jelly. (B) Clutch with 12 embryos. (C) Male captured while transporting seven tadpoles. (D) Dorsolateral view of an early metamorph. Note that the dorsal pattern of coloration is similar to that of adult specimens.

Tables

TABLE 1. Morphometric measurements (in mm) of adult male and female Allobates magnussoni collected from its type locality and from Fazenda Treviso, both in Pará State, Brazil. Abbreviations are defined in the text. Values are: mean ± standard deviation (range of variation).

		PARNA da Amazônia (Type locality)	Treviso
Character	Holotype	Males (n = 14) Females (n = 8)	Males (n = 11)	Females (n = 3)
1-SVL	18.21	17.77 ± 0.96 (16.09–19.59) 19.48 ± 0.82 (17.97–20.84)	18.01 ± 0.38 (17.36–18.71)	19.53 ± 0.62 (18.81–19.90)
2-HL	5.50	5.40 ± 0.30 (4.90–6.00) 5.68 ± 0.15 (5.40–5.80)	5.54 ± 0.15 (5.20–5.80)	5.80 ± 0.10 (5.70–5.90)
3-IO	5.40	5.45 ± 0.26 (5.00–5.90) 5.55 ± 0.21 (5.30–5.90)	5.33 ± 0.15 (5.10–5.60)	5.53 ± 0.25 (5.30–5.80)
4-HW	6.10	6.08 ± 0.22 (5.70–6.40) 6.45 ± 0.12 (6.20–6.60)	6.14 ± 0.13 (6.00–6.40)	6.23 ± 0.21 (6.00–6.40)
5-SL	2.60	2.48 ± 0.23 (2.00–2.75) 2.72 ± 0.28 (2.30–3.10)	2.75 ± 0.13 (2.50–2.95)	2.82 ± 0.08 (2.75–2.90)
6-END	1.95	1.77 ± 0.12 (1.50–1.95) 1.85 ± 0.14 (1.65–2.15)	1.76 ± 0.09 (1.60–1.90)	1.82 ± 0.03 (1.80–1.85)
7-IN	2.70	2.64 ± 0.10 (2.50–2.80) 2.83 ± 0.16 (2.60–3.15)	2.69 ± 0.13 (2.45–2.90)	2.83 ± 0.08 (2.75–2.90)
8-EL	2.50	2.46 ± 0.12 (2.25–2.70) 2.67 ± 0.09 (2.50–2.80)	2.85 ± 0.20 (2.60–3.25)	3.02 ± 0.23 (2.80–3.25)
9-TYM	1.15	1.02 ± 0.15 (0.75–1.25) 1.14 ± 0.14 (1.00–1.35)	1.20 ± 0.13 (1.00–1.45)	1.13 ± 0.06 (1.10–1.20)
10-FAL	3.80	3.76 ± 0.15 (3.50–4.00) 3.81 ± 0.20 (3.50–4.00)	3.95 ± 0.17 (3.50–4.10)	4.10 ± 0.10 (4.00–4.20)
11-AL	5.10	4.79 ± 0.38 (4.20–5.50) 4.93 ± 0.46 (4.40–5.50)	4.90 ± 0.20 (4.50–5.00)	5.00 ± 0.00 (5.00–5.00)
12-THL	8.90	8.55 ± 0.44 (7.50–9.00) 9.03 ± 0.20 (8.70–9.30)	8.60 ± 0.35 (8.10–9.20)	8.73 ± 0.49 (8.40–9.30)
13-TL	9.20	8.78 ± 0.46 (8.00–9.60) 9.14 ± 0.33 (8.70–9.60)	8.84 ± 0.38 (8.50–9.50)	8.90 ± 0.44 (8.60–9.40)
14- FL	9.00	8.24 ± 0.52 (7.50–9.30) 8.53 ± 0.35 (8.00–9.00)	8.91 ± 0.50 (8.10–9.70)	9.07 ± 0.67 (8.50–9.80)
15- HAND I	3.70	3.56 ± 0.30 (2.90–4.00) 3.77 ± 0.24 (3.50–4.20)	3.96 ± 0.09 (3.80–4.10)	3.87 ± 0.23 (3.60–4.00)
16-HAND II	3.50	3.35 ± 0.18 (2.90-3.50) 3.55 ± 0.14 (3.40–3.80)	3.47 ± 0.13 (3.30–3.70)	3.40 ± 0.10 (3.30–3.50)
17-HAND III	4.70	4.63 ± 0.26 (4.00–5.00) 4.85 ± 0.28 (4.60–5.40)	4.75 ± 0.24 (4.40–5.10)	4.73 ± 0.25 (4.50–5.00)
18-HAND IV	3.60	3.36 ± 0.18 (3.00–3.70) 3.45 ± 0.21 (3.20–3.80)	3.48 ± 0.06 (3.40–3.60)	3.40 ± 0.10 (3.30–3.50)
19-WFD	0.75	0.63 ± 0.07 (0.50–0.75) 0.63 ± 0.40 (0.60–0.70)	0.68 ± 0.04 (0.65–0.75)	0.70 ± 0.05 (0.65–0.75)
20-DPT	0.65	0.59 ± 0.09 (0.45–0.80) 0.64 ± 0.08 (0.55–0.75)	0.67 ± 0.10 (0.50–0.90)	0.72 ± 0.03 (0.70–0.75)
21-WTT	0.35	0.33 ± 0.06 (0.25–0.50) 0.32 ± 0.06 (0.25–0.40)	0.40 ± 0.04 (0.35–0.45)	0.42 ± 0.03 (0.40–0.45)
22-WPF	0.45	0.44 ± 0.05 (0.35–0.50) 0.40 ± 0.04 (0.35–0.45)	0.45 ± 0.02 (0.45–0.50)	0.43 ± 0.06 (0.40–0.50)
23-WTD	0.95	0.77 ± 0.08 (0.60–0.95) 0.80 ± 0.04 (0.75–0.85)	0.88 ± 0.06 (0.75–0.95)	0.87 ± 0.03 (0.85–0.90)