Calling Site Fidelity and Call Structure of a Neotropical Toad, Rhinella Ocellata (Anura: Bufonidae)

Calling Site Fidelity and Call Structure of a Neotropical Toad, Rhinella ocellata (Anura: Bufonidae)

1 JANALEE P. CALDWELL AND DONALD B. SHEPARD

Sam Noble Oklahoma Museum of Natural History and Department of Zoology, University of Oklahoma, Norman, Oklahoma 73072, USA

ABSTRACT.—Rhinella ocellata is a relatively unknown species of toad that occurs in Brazilian cerrado, a savanna-like biome. We studied home range and calling behavior of a population of this species in the state of Tocantins, Brazil, in 2004. Unlike most other species of bufonids, male R. ocellata did not migrate to ponds or other aquatic sites and form choruses for reproduction. Instead, R. ocellata had a prolonged breeding period during which males called from terrestrial sites typically on bare sandy soil near grass clumps or other shrubby vegetation. Calling sites were within open gallery forest in cerrado and ranged from 10–64 m from the shoreline of a river; similar calling sites were observed at another site in Tocantins in 2005. Home-range size was small compared to other toads that have been studied. Most individual R. ocellata remained within a small area during the study, typically calling nightly except during hot, dry periods. Calling bouts of R. ocellata consisted of an average of 36.9 short calls with fairly long intervals between bouts. Calling bout length averaged 31 sec. Neighboring individuals frequently alternated calls. One clutch of eggs found in a backwater pool of the river was tentatively identified as that of R. ocellata. The derived breeding behavior of R. ocellata is consistent with the historical difficulty in placing this species phylogenetically.

RESUMO.—Rhinella ocellata e´ uma espećie relativamente desconhecida que ocorre no Cerrado brasileiro. No´s estudamos a a´rea de vida e o comportamento de vocalizaçaõ de uma populaçaõ dessa espećie no estado do Tocantins, norte do Brasil, no ano de 2004. Diferentemente de outras espećies de bufonı´deos, os machos de R. ocellata naõ migraram para poças ou outros ambientes aqua´ticos formando agregacoes reprodutivas. Ao inve´s, R. ocellata teve um perıódo de reproduçaõ prolongado durante o qual machos vocalizaram em ambientes terrestres, tipicamente sobre solo arenoso pro´ximo a gramıńeas ou moitas de vegetaçaõ. Os sı´tios de vocalizaçaõ se encontraram dentro de matas de galeria e estiveram localizados de 10 a 64 m de distaˆncia da margem de um rio. Sı´tios de vocalizaçaõ com caracterı´sticas similares foram observados em outra localidade em Tocantins em 2005. O tamanho me´dio da a´rea de vida foi de 163.3 m2, valor considerado pequeno quando comparado ao de outras espećies de bufonı´deos ja´ estudadas. A maioria dos indivı´duos de R. ocellata permaneceu limitada a uma a´rea pequena durante todo o estudo, vocalizando todas as noites, exceto em dias muito quentes e secos. A sequ¨ eˆncia de canto de R. ocellata consistiu em me´dia de 36.9 cantos curtos, com intervalos relativamente longos entre as sequ¨ eˆncias. Cada sequ¨ eˆncia durou em me´dia 31 segundos. Indivı´duos pro´ximos alternavam o canto frequentemente. Uma possı´vel desova de R. ocellata foi encontrada em uma poça pro´xima ao rio. O comportamento reprodutivo derivado de R. ocellata e´ consistente com a dificuldade em posicionar essa espećie filogeneticamente.

Anurans are known for their often conspicu- direct development (Lynn and Lutz, 1946; ous and complex acoustic communication, Ortega et al., 2005), and most dendrobatids which may function to attract mates, defend transport tadpoles on their backs to water a territory, signal an unwillingness to mate, or bodies after tadpoles develop from eggs de- startle a potential predator (Gerhardt and posited on land (Grant et al., 2006). Exceptions Huber, 2002). In many species, calling occurs to the ancestral reproductive pattern within primarily after rains initiate migration to some a clade usually point to evolutionary novelties, type of water body for reproduction. Although which subsequently lead to new insights about anurans exhibit a diversity of reproductive natural selection and how organisms respond to modes (Duellman and Trueb, 1986), species their biotic and abiotic environments. within a given clade typically display similar Among anurans, species within Bufonidae reproductive patterns (Duellman, 1985; Haddad typically migrate to ponds or other water bodies and Prado, 2005). For example, most species of after heavy rains and form large choruses. Most Eleutherodactylus deposit eggs on land and have bufonids are explosive breeders, remaining at breeding sites for only a few days; however, some may remain up to several weeks. With 1 Corresponding Author. E-mail: [email protected] few exceptions, clutch size is large, varying 612 J. P. CALDWELL AND D. B. SHEPARD from several thousand to more than 50,000 eggs (Duellman and Trueb, 1986; Crump, 1989; Hag- man and Shine, 2006), and small tadpoles undergo rapid development and metamorphose in a few weeks. At least one species, however, deviates from this general pattern; males of the Neotropical Rhinella (5 Bufo) castaneotica call sporadically throughout the rainy season from Brazil nut capsules scattered on the ground throughout their rain-forest habitat. Mean clutch size of this species at two localities was about 130–230 eggs, which were deposited in the water-filled capsules (Caldwell, 1993; Cald- well and Arau´jo, 2004). We recently discovered that another species, Rhinella ocellata, does not migrate in large FIG. 1. Map of Brazil with expanded view showing numbers to ponds or other water bodies to locations of the primary study area at Rio Novo, and reproduce. Rhinella ocellata, originally described where toads were observed calling in 2005 at Parque as Bufo ocellatus by Gu¨ nther (1858) and rede- Estadual do Cantaõ. States of Brazil are: TO 5 scribed by Cochran (1954), is a relatively poorly Tocantins; GO 5 Goia´s; MT 5 Mato Grosso; PA 5 known species that occurs in the Brazilian states Para´;MA5 Maranhaõ; PI 5 Piauı´;BA5 Bahia. of Tocantins, Para´, Mato Grosso, Minas Gerais, and Goia´s. It has been variously placed in the history of this species to show that differences Bufo typhonius (5 margaritifer) group by Bou- in social behavior between R. ocellata and most lenger (1882) and in the Bufo granulosus group by Leaõ and Cochran (1952). Duellman and other bufonid species are consistent with the Schulte (1992) placed 45 of the 51 species of inability to clearly place this toad in a species South American Bufo known at that time into group based on morphological characters. eight phenetic groups. Bufo ocellatus, known only from two specimens other than the MATERIALS AND METHODS holotype at that time, was one of six species that could not be placed in any group. Two Study Area.—The study area was located recent phylogenetic studies (Pramuk, 2006; within Parque Estadual do Jalapaõ in the state Frost et al., 2006) mentioned B. ocellatus; the of Tocantins in central Brazil, 37.2 km west of 9 0 3 former placed B. ocellatus as sister to the Bufo Mateiros on the Rio Novo (10u32 59.0 S margaritifer group but not in it or in any other 46u45938.70W), hereafter referred to as the Rio group, and the latter extensively revised Bufo- Novo site (Fig. 1). The area was a relatively nidae but retained ‘‘Bufo’’ ocellatus in the narrow strip (approximately 54 3 220 m) of nonmonophyletic ‘‘Bufo’’ (sensu lato; nonmono- disturbed gallery forest along the Rio Novo. phyly designated by placing the name in Gallery forest vegetation consisted of sparsely quotes) and did not assign it to any species scattered small trees, which were denser along group. Chaparro (2007) used the generic name the shoreline of the river, and scattered small Rhinella for most species groups of South shrubs. Vegetation adjacent to the gallery forest American toads; herein, we follow this arrange- was cerrado, a savanna-like habitat. The rainy ment. season in this area extends from October into While working in a cerrado/gallery forest March (www.inmet.gov.br), although periods area in Brazil, we discovered that R. ocellata is of hot, dry weather may occur during this time. highly unusual among species of toads in that Toads were encountered primarily in gallery they call from terrestrial sites away from water. forest and rarely in cerrado, usually in open Further, we discovered that individual toads areas of light-colored sand, the prevalent soil call from nearly the same location from night to type in the area. night. Here, we examine calling site fidelity, Locating and marking individuals.—Toads were movements, and home-range size of R. ocellata. heard calling nearly every night of the 36 In addition, we describe the call of this species consecutive days of the study (30 November for the first time and examine call variation 2004 to 4 January 2005). Because our open-air within and among individuals at our study site. camp was located in the center of the study We describe some aspects of calling behavior, area, no calling nights were missed. We spent particularly the propensity of neighboring approximately two hours before midnight each males to alternate calls. We combine these data night searching the area for toads. Individuals with additional information on the natural were located by tracing their calls or by their CALLING BEHAVIOR OF RHINELLA OCELLATA 613 weak eyeshine. In addition, we carefully searched the entire study area several times each night to locate noncalling individuals. Toads were recognized individually by unique dorsal spotting patterns (Fig. 2). In addition, some toads had distinctive scars or crooked middorsal stripes that were also noted. Unique spots or patterns have been used in other studies to recognize individuals (Hyla versicolor, Bertram et al., 1996; Dendrobates, Caldwell and de Oliveira, 1999; Pseudis, Miran- da et al., 2005). The first time a toad was encountered, we sketched its pattern and distinctive marks, determined its mass with a portable electronic balance, and recorded its snout–vent length (SVL) to the nearest millime- ter with a ruler. Toads were handled for 10 min or less and then returned to the exact site where they were captured; most resumed calling after a few minutes. Fourteen individual toads were located and processed between 30 November and 26 December; toads were numbered B4– B17. Each time a toad was captured or resighted, its location was recorded on a scaled map of the area. Toads were not handled when resighted. If the distance moved between successive captures was small, we measured the actual distance and direction. In cases of long distance movement (.50 m), the distance moved was determined with a Garmin GPS 76CS receiver (66 m). We used the program XYit 3.1 (Geomatrix, Ltd., Barrow on Humber, North Lincolnshire, U.K.) to digitize maps and convert toad locations to X–Y coordinates for analyses. Voucher specimens of R. ocellata collected outside the study area but in the vicinity were deposited in the Herpetology Collection at the Sam Noble Oklahoma Museum of Natural FIG. 2. Two individuals of Rhinella ocellata showing History and the Coleçaõ Herpetolo´gica da differences in dorsal spot patterns from (A) Parque Universidade de Brası´lia. Estadual do Cantaõ, and (B) the Rio Novo study site. Photos by JPC and DBS, respectively. Home range and movement.—For toads with more than three observations, we calculated the Minimum Convex Polygon (MCP) home-range alternating calls with the individual being estimate and movement statistics in ArcView recorded) and three duetting pairs were re- 3.2 (ESRI, Redlands, California) using the corded in the primary study area. Sound Animal Movement Extension (Hooge and Ei- analysis was conducted using Raven 1.2.1 for chenlaub, 1997). We used linear regression to Mac OS X or Raven 1.3, beta version (Cornell determine whether home-range estimates and Laboratory of Ornithology, Ithaca, New York). total distance moved were related to the Call variables analyzed from recordings in- number of observations or duration between cluded length of calling bout, number of calls the first and last captures. per calling bout, call length, intercall interval, Analysis of calls and calling behavior.—We number of pulses per call, pulse length, pulse collected call data in two ways: (1) by recording rate, dominant frequency, and frequency mod- calls of vocalizing males; and (2) by noting ulation of the call. Measurements of these calling behavior during focal observations. variables were the same as defined in Cocroft Vocalizations were recorded with a Sony and Ryan (1995) or Tarano (2001). We averaged WMD-6C tape recorder with a Sennheiser call variables within each individual, then ME67 directional microphone. Three individu- across individuals to obtain means for each call als calling alone (i.e., no individual within 4 m variable unless otherwise noted. We used JMP 614 J. P. CALDWELL AND D. B. SHEPARD version 5.0.1a for statistical analyses; values are means 6 SE. We conducted 30-min focal observations on six individual males to determine rate of calling bouts, number of calls per calling bout, length of calling bouts, and length of intercalling bout intervals. The observer (JPC) sat quietly at least 10 m from a toad, recorded beginning and ending times of each calling bout using a stop- watch, and counted the number of calls com- prising each calling bout. Focal observations were made throughout the study period after dark between 1916 and 2333 h. Focal observations were made only on nights when toads were calling consistently. Neighboring male toads were frequently observed alternating calls with focal males. We defined a duetting pair as two males that were calling relatively close to each other, typically ,4 m, with no other individuals calling nearby. To determine whether duetting males were actually interacting with each other, we used linear regressions to examine timing of calls in duetting toads; this method was used by Bosch and Ma´rquez (2001), based on methods of Wickler and Seibt (1974). If male B is calling in response to male A’s calls, then the delay in the call of individual B should be related to the call duration of duetting partner A. For each duetting pair, we analyzed runs of 20 alternating calls within a calling bout. Because toads typically required one or two calls to avoid overlap at the beginning of a duetting sequence, we began the analysis after toad B had called three times. Of the three duetting pairs, we analyzed two bouts of two pairs, and four bouts of one pair.

RESULTS Fourteen male R. ocellata, largely clumped in two parts of the study area, were found during the study. Snout–vent length of 13 of these FIG. 3. Map of study area showing locations of the individuals was 49.6 6 1.9 mm (range 47–53), 11 individuals of Rhinella ocellata that were captured and mass was 11.7 6 1.5 g (range 9.5–13.9). One more than once. The roads indicated in the study area gravid female captured in a drift fence on 30 are small one-vehicle wide tracks in the sandy soil; December was 53.5 mm SVL, and two juveniles, two parallel tracks were present south of the camp. found on 1 December and 25 December, The initial point of capture for B11 is not shown; this measured 15.3 and 16.6 mm SVL, respectively. toad was initially found by the road 12 m from the Of the 14 male toads, three were captured only north end of the study area. Numbers above tightly once. The remaining 11 toads were captured an grouped symbols indicate number of points in the average of 7 6 3.4 times (range 3–15; Fig. 3). grouping. Jagged lines in the center of the figure indicate a section of the study area where no toads Home range and movement.—Home-range size were found. and total distance moved varied considerably among individuals, averaging 81.5 6 12.7 m2 and 163.3 6 35.9 m, respectively (Table 1). The and the duration between the first and last number and duration of observations differed captures (F1, 9 5 0.13, P 5 0.72). For eight of 11 among individual toads (Table 1), but home- individuals, the average distance between loca- range size estimates were unrelated to the tions was 8.8 6 1.1 m, and movement rate number of observations (F1, 9 5 0.21, P 5 0.66) averaged 2.9 6 0.4 m per day. Individual toads CALLING BEHAVIOR OF RHINELLA OCELLATA 615

TABLE 1. Number of times 11 toads were observed, the time interval between the first and last observations, the total distance moved, and the minimum convex polygon (MCP) estimate of home-range size for each individual.

Toad no. No. obs. Time interval (days) Tot. dist. moved (m) MCP (m2) B4 11 34 77.6 117.3 B6 15 29 121.1 364.1 B7 7 23 41.6 165.0 B8 7 23 77.7 26.0 B9 6 18 20.5 21.5 B10 7 23 69.1 61.3 B11 5 15 145.7 316.6 B12 4 13 135.9 222.0 B13 3 9 114.4 208.4 B14 8 15 45.9 45.3 B16 5 11 47.5 248.9 Mean 6 SE 81.5 6 12.7 163.3 6 35.9 Range (20.5–145.7) (21.5–364.1) were frequently found ,1.5 m from their pre- area. During exceptionally hot and prolonged vious location, but three individuals moved 109, dry periods, toads ceased nighttime calling or 130, and 135 m between successive captures; the called only sporadically. Calling was more largest move was made over a two-day period. intense from sunset until dark and at first light Total distance moved was unrelated to the until sunrise, although consistent calling oc- duration between the first and last captures curred on many nights. (F1, 9 5 0.09, P 5 0.77), indicating that toads Call structure.—Rhinella ocellata produces generally did not move greater distances over a moderately long train of short calls, here longer time periods. Toads appeared to prefer referred to as a calling bout, with fairly long specific areas of the study site (Fig. 3), and we intervals between calling bouts (Fig. 4). In often observed different individuals calling general, individual calls within a calling bout from the same location on different nights. In are about 0.3 sec in length with an interval of the northern section of the study area, five toads about 0.8 sec between calls (Table 2). Calls are 2 were initially sighted within a 150-m area, and pulsed (Fig. 5), and pulses can be detected by in the southern section of the study area, the human ear. A typical call has about seven original locations of six toads were within pulses with interior amplitude modulation 2 a 640-m area. (appearing as subpulses) and then a short series Calling behavior.—We observed only adver- of single pulses with decreasing amplitude. The tisement calls during the study. Male toads dominant frequency appears to equal the were not observed aggressively interacting with fundamental frequency. After the first 2–5 calls each other, and we never heard release calls. in the bout, all calls have the same frequency. Toads called throughout the study period, The first few calls in a calling bout may show although they called more intensely after some frequency modulation (from call to call); rainfall. Sporadic calling occurred during day- for example, in one individual (B4), dominant time, particularly if a storm approached the frequency of all calls in three of five bouts was 1378 Hz, but dominant frequency began at 1205 Hz and increased to 1378 Hz by the sixth call and began at 1292 Hz and increased to 1378 Hz by the third call in two other bouts. The first few calls in a calling bout may have some single pulses interspersed among the calls with subpulses. The initial call in each bout is shorter than the subsequent calls (Fig. 5A). For example, initial calls in five bouts in one toad (B4) averaged 0.220 6 0.01 sec (range 0.199–0.254) compared to the overall average of 0.294 sec for all calls of this toad (Table 2). Pulses in initial calls in a bout FIG. 4. Waveform and spectrogram of the adver- typically lack subpulses. tisement call of Rhinella ocellata. This calling bout Call timing.—Neighboring individuals fre- consists of 31 calls. quently alternated calls (Fig. 6). Linear regres- 616 J. P. CALDWELL AND D. B. SHEPARD 9.2 8.5 13.1 6.8 6 6 6 6 (0–86) (0–79) (0–158) (Hz) Frequency 36.9 15.8 modulation 8.5 13.1 54.7 23.3 15 6 6 6 5 1378 17.2 N (1185–1264) (1343–1501) Dominant 1248.2 frequency (Hz) 1432.5 0.00 0.00 0.00 0.00 1352.9 15 6 6 6 6 5 Pulse N (0.019–0.023) (0.014–0.021) (0.013–0.017) length (ms) 0.021 0.017 0.015 0.4 0.2 0.2 0.9 0.017 15 6 6 6 6 5 N (26.3–32.3) (24.2–27.3) (28.3–30.6) 28.2 26.1 29.3 Pulse rate (/sec) 0.3 0.3 0.3 1.0 27.9 15 6 6 6 6 5 (9–12) (7–12) (12–16) N Pulses/call 13.3 10.3 10.5 lternating calls with a nearby individual). 15 0.02 0.01 0.02 0.04 11.4 5 6 6 6 6 N (0.692–0.894) (0.800–0.966) (0.533–0.900) (ms) 0.783 0.906 0.795 Intercall interval 0.00 0.00 0.00 0.01 0.828 15 6 6 6 6 5 N (0.275–0.308) (0.254–0.302) (0.232–0.298) 0.294 0.281 0.261 Call length (ms) 2.1 3.8 4.8 1.1 0.279 6 6 6 6 (31–44) (30–47) (23–51) 38.8 36.8 35.0

FIG. 5. (A) Expanded waveform and spectrogram showing single pulses of an introductory call. (B) 1.9 3.6 4.3 1.4 36.9 Expanded waveform and spectrogram of a call in 6 6 6 6 which most pulses show interior amplitude modula- (sec) No. calls/bout (25.3–36.9) (27.2–42.0) (18.0–42.1) tion (i.e., appear double-pulsed). Call bout length 3) 31.0 SE; range in parentheses. sion analysis revealed that the alternation 6 5 pattern was highly synchronized in eight bouts N of three duetting pairs of toads (Table 3). The proportion of variation in the call delay of caller

2. Call variables of three individuals recorded while calling alone (not a B explained by the call duration of caller A ranged from 0.34–0.86 (Table 3). ABLE Toad no./no. bouts Values are mean T Focal observations.—Focal observations on six B4/5 31.2 B9/5 33.3 B11/5 28.5 Grand means ( individuals revealed that male R. ocellata pro- CALLING BEHAVIOR OF RHINELLA OCELLATA 617

FIG. 6. Waveform and spectrogram of calling bouts of two males of Rhinella ocellata in which calls are alternated. Note that calls of Male B do not overlap calls of Male A. duced an average of 11.7 6 3.3 calling bouts that use streams breed explosively during dry (range 8–17) per 30-min focal period. Calling periods when eggs can be deposited in still bout length averaged 27.7 6 6.4 sec (range 20.8– pools (Jofre´ et al., 2005). With few exceptions, 39.2), and mean interval between calling bouts relatively large clutches of eggs are deposited in was 167.1 6 41.6 sec (range 111.4–230.7). The gelatinous strings; two exceptions include Ana- number of calls per bout averaged 29.4 6 8.5 xyrus (5 Bufo) debilis and Ollotis (5 Bufo) calls (range 21.4–45.9). One individual pro- marmorea, which deposit single eggs (Blair, duced nearly twice as many calls per bout 1972; Tandy and Keith, 1972). Toad tadpoles (45.9 calls) as other toads; without this in- are small, and the larval period is typically 1– dividual, the number of calls per bout averaged 2 months in length. 26.1 6 3.1 calls (range 21.4–29.0). Rhinella ocellata differs from typical bufonids in that its reproductive behavior does not involve explosive migrations to water and formation of large choruses for breeding. Unlike DISCUSSION most toads, which breed for only short time Most species of bufonids migrate to ponds or periods at aquatic sites (Bertoluci and Rodri- other aquatic sites to reproduce (Muths, 2003; gues, 2002; Prado et al., 2005), R. ocellata is Prado et al., 2005). Many breed explosively with a prolonged breeder that uses terrestrial calling the onset of heavy rains, although some species sites.

TABLE 3. Regression analysis showing that nonoverlapping calls of duetting Rhinella ocellata are highly related (i.e., the delay in the call of Male B was related to the call duration of duetting partner A). Unm 5 unmarked male.

Individuals NF-value df b Adj. r2 P B9/B5, duet 1 20 33.0 1, 18 0.740 0.63 ,0.0001 B9/B5, duet 2 20 95.2 1, 18 1.185 0.83 ,0.0001 B7/Unm, duet 1 20 42.7 1, 18 0.455 0.69 ,0.0001 B7/Unm, duet 2 20 115.3 1, 18 0.759 0.86 ,0.0001 B4/B10, duet 1 20 10.8 1, 18 0.460 0.34 0.004 B4/B10, duet 2 20 49.8 1, 18 1.116 0.72 ,0.0001 B4/B10, duet 4 20 50.5 1, 18 0.952 0.72 ,0.0001 B4/B10, duet 5 20 40.5 1, 18 0.895 0.68 ,0.0001 618 J. P. CALDWELL AND D. B. SHEPARD

Calling sites of R. ocellata were typically on activity to certain areas and what was special bare sandy soil surrounded by grass clumps or about specific calling sites was not obvious and other shrubby vegetation. Although calling sites warrants further investigation. were within open gallery forests along rivers in Call structure.—The call of R. ocellata is similar cerrado (a savanna-like tropical biome), indi- in structure to some other species of bufonids viduals do not call immediately adjacent to or in (Cocroft and Ryan, 1995). Phonetically, each call water. Calling sites ranged from 10–64 m from in the calling bout sounds like a low, slow trill the river. In both our study site and a site where in which individual pulses can be distin- we observed R. ocellata in 2005 (a state park guished. Like other species of bufonids dis- called ‘‘Parque Estadual do Cantaõ’’ located cussed by Cocroft and Ryan (1995), dominant near Caseara, Tocantins; 9u18932.90S 3 frequency and fundamental frequency are the 49u57935.50W; Fig. 1), toads called from sites same in R. ocellata. In some calls, three harmo- within gallery forests along rivers, the Rio Novo nics are present, but they are diffuse and not (this study) and the Rio do Cocos (a small obvious in most calls. tributary of the Rio Araguaia) at the 2005 site Pulses in individual calls of R. ocellata show (JPC and DBS, pers. obs.). interior amplitude modulation (i.e., they appear Home range and movement.—Home-range sizes as ‘‘double-pulsed’’). Martin (1972) showed that of male R. ocellata were small compared to those this feature of pulses in some bufonids is related of other toads that have been studied. For to a morphological character, the presence of example, home-range size of male Anaxyrus (5 arytenoid valves in the larynx. Cocroft and Bufo) boreas was 58,299 m2 (Muths, 2003), well Ryan (1995), in a study on evolution of above the average of 163.3 m2 found by us for R. advertisement calls in some species of North ocellata. Home ranges in other studies have American bufonids, showed that this feature typically included the area encompassed by (pulse subdivision) was present in Ollotis (5 migrations to water bodies for reproduction, Bufo) valliceps and the Anaxyrus (5 Bufo) accounting partly for these differences. Further, cognatus group but was absent in the Anaxyrus the duration of study and the methods em- (5 Bufo) americanus clade. Although we have no ployed (mark-recapture vs. radio-telemetry, or information on arytenoids in R. ocellata, a mor- MCP vs. kernel density home-range estimates) phological study of this species may reveal their could result in different estimates. Our study presence. was limited to a 36-day period during the wet Calling behavior.—Calling behavior may vary season. Toad home ranges may have been larger dramatically among male anurans depending had our study extended over a longer period on density of individuals at the breeding site (e.g., into the dry season), but home-range size (Brown and Pierce, 1967; Wells, 1988). Some of R. ocellata is still likely to be small because it clades (especially tropical ones) include many does not exhibit the long-distance mass migra- species in which males call from positions fairly tions to aquatic breeding sites like many other isolated from other males (e.g., Brachycephali- bufonid species. dae), whereas other clades include species that High site fidelity is characterized by little primarily migrate in large numbers to ponds or movement and displacement over an extended other water bodies to breed, usually in response time period (Wells, 1977; Shepard, 2004). Most to rainfall. In large breeding congregations, R. ocellata remained within a small area during synchronized calling may make individuals less our study, although three individuals moved obvious to predators that cue in on calls to relatively long distances. High site fidelity is locate prey (Tuttle and Ryan, 1982). In other a component of territoriality, but territoriality situations, calls may be alternated to avoid also involves defense of a limited resource overlapping calls with neighboring males. Grafe (Wells, 1977; Mathis et al., 1995; Shepard, (1996) suggested several hypotheses about why 2004). Male R. ocellata were never observed males would alternate calls: (1) call alternation aggressively interacting with each other, either may play a role in intermale spacing in the physically or through encounter calls, and no habitat; (2) call overlap may disrupt a female’s resources needed by females (e.g., oviposition ability to access temporal information about sites) were apparent near the calling sites. We a male’s call and, thus, should be avoided; and observed individuals calling consistently night (3) call overlap makes it difficult for females to after night from the same areas throughout the locate males in the environment. duration of our study, and different individuals Call alternation is clearly a feature of neigh- were often observed calling from the same boring male R. ocellata. Klump and Gerhardt location on different nights. This behavior was (1992) noted that goals of measuring patterns of identical to our observations of male R. ocellata acoustic interactions include establishing at our 2005 site in Tocantins (JPC and DBS, pers. whether frogs are actually interacting vocally, obs.; see above). Why males restricted their examining the range of sound pressure levels CALLING BEHAVIOR OF RHINELLA OCELLATA 619 over which frogs interact, and determining how sion to conduct research in Parque Estadual do these interactions translate into reproductive Jalapaõ was granted by SEPLAN and NATUR- success for individuals. Our results showed that ATINS within the Governo do Estado do male R. ocellata are interacting vocally, but we Tocantins (Auth. 033/2004). The study followed have no information on how results of these protocols approved by the University of Okla- interactions affected reproductive success. Rhi- homa IACUC, Assurance A3240-01. L. J. Vitt nella ocellata would be an ideal species for provided comments on a draft of the paper. A. further investigation of call alternation and R. Kuhns helped with home-range and move- female choice. Males can be easily located and ment analyses. G. C. Costa provided the recognized individually by unique dorsal spot Portuguese translation of the abstract. This patterns; they call in open habitats and can be research was supported by National Science easily observed and recorded; and presentation Foundation grant DEB-0415430 to L. J. Vitt and of females or playback calls could be readily JPC; additional funds were supplied by the Sam accomplished. Noble Oklahoma Museum of Natural History. Natural history and phylogenetic implications.— Any opinions, findings, and conclusions or Because R. ocellata has a prolonged breeding recommendations expressed in this material season, females are predicted to become re- are those of the authors and do not necessarily productive and arrive at breeding sites asyn- reflect the views of the National Science chronously (Wells, 1977). We found one clutch Foundation. of eggs (two separate strands with eggs in a single row) in a backwater pool along the river near our study area that we tentatively identi- LITERATURE CITED fied as eggs of R. ocellata. The eggs were black BERTOLUCI, J., AND M. T. RODRIGUES. 2002. Seasonal and had a mean size of 1.74 6 0.02 mm (range patterns of breeding activity of Atlantic rainforest 1.70–1.79; N 5 5). A single strand contained anurans at Boraceía, southeastern Brazil. Amphibia- 1,120 eggs; thus, clutch size was approximately Reptilia 23:161–167. 2,240 eggs. Of the three other bufonids in the BERTRAM, S., M. BERRILL, AND E. NOL. 1996. Male mating area (Vitt et al., 2005), Rhinella (5 Bufo) schneideri success and variation in chorus attendance within has a much larger clutch in which eggs likely and among breeding seasons in the Gray Treefrog occur in a double row as in the closely related (Hyla versicolor). Copeia 1996:729–734. LAIR Bufo Rhinella (5 Bufo) marinus (Kennedy, 1982); B , W. F. 1972. of North and Central America. In W. F. Blair (ed.), Evolution in the Genus Bufo, 5 Rhinella ( Bufo) granulosus breeds in large pp. 93–101. University of Texas Press, Austin. congregations in temporary ponds; and Rhaebo BOSCH, J., AND R. MA´ RQUEZ. 2001. Call timing in male- (5 Bufo) guttatus has a larger clutch of eggs that male acoustical interactions and female choice in occur in a double row within a strand (Caldwell the Midwife Toad Alytes obstetricans.Copeia and Arau´jo, 2005). Additional work will be 2001:169–177. required to fully understand the natural history BOULENGER, G. A. 1882. Catalogue of the Batrachia and reproductive biology of this species and to Salientia in the collection of the British Museum. obtain the undescribed tadpoles. Historically, 2nd ed. British Museum (Natural History), Lon- the phylogenetic position of R. ocellata has been don. difficult to determine, and the behavioral BROWN,L.E.,AND J. R. PIERCE. 1967. Male-male differences between R. ocellata and other bufo- interactions and chorus intensities of the Great Plains Toad, Bufo cognatus. Copeia 1967:149–154. nid species are consistent with this problem. CALDWELL, J. P. 1993. Brazil nut fruit capsules as More information on the ecology of this species phytotelmata: interactions among anuran and in- and other bufonids may help to clarify phylo- sect larvae. Canadian Journal of Zoology 71: genetic relationships and would provide insight 1193–1201. into the selective pressures that have led to the CALDWELL, J. P., AND M. C. ARAU´ JO. 2004. Historical and evolution of its derived behavioral traits. ecological factors influence survivorship in two clades of phytotelm-breeding frogs (Anura: Bufo- Acknowledgments.—Field logistics for work in nidae, Dendrobatidae). In R. M. Lehtinen (ed.), Tocantins, Brazil, were arranged by G. R. Colli. Ecology and Evolution of Phytotelm-Breeding Permits to conduct research and collect speci- Anurans. Miscellaneous Publications, Museum of mens in Brazil were issued by Conselho Zoology, University of Michigan 193:11–21. Nacional de Desenvolvimento Cientı´fico e Tec- ———. 2005. Amphibian faunas of two eastern Amazonian rainforest sites in Para´, Brazil. Occa- nolo´gico (CNPq, Portaria MCT 649, de 2004, sional Papers of the Sam Noble Oklahoma Muse- renewed 3 January 2005) and Instituto Brasileiro um of Natural History 16:1–41. do Meio Ambiente e dos Recursos Naturais CALDWELL, J. P., AND V. L. DE OLIVEIRA. 1999. Determi- Renova´veis (IBAMA, 0188/2004-CGFAU), re- nants of biparental care in the Spotted Poison Frog, spectively, under a research convenio between Dendrobates vanzolinii (Anura: Dendrobatidae). OMNH and the University of Brasilia. Permis- Copeia 1999:565–575. 620 J. P. CALDWELL AND D. B. SHEPARD

CHAPARRO, J. C., J. B. PRAMUK, AND A. G. GLUESENKAMP. KLUMP, G., AND H. C. GERHARDT. 1992. Mechanisms and 2007. A new species of arboreal Rhinella (Anura: function of call-timing in male-male interactions in Bufonidae) from cloud forest of southeastern Peru. frogs. In P. K. McGregor (ed.), Playback and Herpetologica 63:203–212. Studies of Animal Communication, pp. 153–174. COCHRAN, D. M. 1995 ‘‘1954.’’ Frogs of southeastern Plenum Press, New York. Brazil. Bulletin of the U.S. National Museum LEA˜ O, A. T., AND D. M. COCHRAN. 1952. Revalidation 206:1–423. and re-description of Bufo ocellatus Gunther, 1858 COCROFT, R. B., AND M. J. RYAN. 1995. Patterns of (Anura: Bufonidae). Memo´rias do Instituto de advertisement call evolution in toads and chorus Butantan, Saõ Paulo 24:271–280. frogs. Animal Behaviour 49:283–303. LYNN, W. G., AND B. LUTZ. 1946. The development of CRUMP, M. L. 1989. Life history consequences of Eleutherodactylus guentheri Stdnr. 1846 (Salientia). feeding versus non-feeding in a facultatively non- Boletim do Museu Nacional Brasil (Nova Ser.) feeding toad larva. Oecologia 78:486–489. 71:1–46. DUELLMAN, W. E. 1985. Reproductive modes in anuran MARTIN, W. F. 1972. Mechanics of sound production in amphibians: phylogenetic significance of adaptive toads of the genus Bufo: passive elements. Journal strategies. South African Journal of Science of Experimental Zoology 176:273–294. 81:174–178. MATHIS, A., R. G. JAEGER,W.H.KEEN,P.K.DUCEY,S.C. DUELLMAN,W.E.,AND R. SCHULTE. 1992. Description of WALLS, AND B. W. BUCHANAN. 1995. Aggression and a new species of Bufo from northern Peru with territoriality by salamanders and a comparison comments on phenetic groups of South American with the territorial behavior of frogs. In H. toads (Anura: Bufonidae). Copeia 1992:162–172. Heatwole and B. K. Sullivan (eds.), Amphibian DUELLMAN, W. E., AND L. TRUEB. 1986. Biology of Biology. Vol. 2. Social Behavior, pp. 633–676. Sur- Amphibians. McGraw-Hill, New York. rey Beatty and Sons, Chipping Norton, New South FROST, D. R., T. GRANT,J.FAIVOVICH,R.H.BAIN,A. Wales, Australia. ´ HAAS,C.F.B.HADDAD,R.O.DE SA,A.CHANNING, MIRANDA, T., M. EBNER,M.SOLE, AND A. KWET. 2005. M. WILKINSON,S.C.DONNELLAN,C.J.RAXWORTHY, Estimation of population size in Pseudis cardosoi J. A. CAMPBELL,B.L.BLOTTO,P.MOLER,R.C.DREWES, (Anura, Pseudidae), using photoidentification as R. A. NUSSBAUM,J.D.LYNCH,D.M.GREEN, AND a method for individual recognition. Biociencias W. C. WHEELER. 2006. The amphibian tree of life. 13:49–54. Bulletin of the American Museum of Natural MUTHS, E. 2003. Home range and movements of boreal History 297:1–370. toads in undisturbed habitat. Copeia 2003:160–165. GERHARDT,H.C.,AND F. HUBER. 2002. Acoustic ORTEGA, J. E., V. H. SERRANO, AND M. P. RAMI´REZ-PINILLA. Communication in Insects and Anurans. Univer- 2005. Reproduction of an introduced population of sity of Chicago Press, Chicago. Eleutherodactylus johnstonei at Bucaramanga, Co- GRAFE, T. U. 1996. The function of call alternation in lombia. Copeia 2005:642–648. the African Reed Frog (Hyperolius marmoratus): PRADO, C. P. A., M. UETANABARO, AND C. F. B. HADDAD. precise call timing prevents auditory masking. 2005. Breeding activity patterns, reproductive Behavioral Ecology and Sociobiology 38:149–158. modes, and habitat use by anurans (Amphibia) in GRANT, T., D. R. FROST,J.P.CALDWELL,R.GAGLIARDO, a seasonal environment in the Pantanal, Brazil. C. F. B. HADDAD,P.J.R.KOK,D.B.MEANS,B.P. Amphibia-Reptilia 26:211–221. NOONAN,W.E.SCHARGEL, AND W. C. WHEELER. 2006. PRAMUK, J. B. 2006. Phylogeny of South American Bufo Phylogenetic systematics of Dart-Poison Frogs and their relatives (Amphibia: Athesphatanura: Den- (Anura: Bufonidae) inferred from combined evi- drobatidae). Bulletin of the American Museum of dence. Zoological Journal of the Linnean Society Natural History 299:1–262. 146:407–452. SHEPARD, D. B. 2004. Seasonal differences in aggression GU¨ NTHER, A. 1858. Neue Batrachier in der Sammlung des britischen Museums. Archiv fu¨ rNatur- and site tenacity in male green frogs, Rana geschichte 34:319–328. clamitans. Copeia 2004:159–164. TANDY, M., AND R. KEITH. 1972. Bufo of Africa. In W. F. HADDAD, C. F. B., AND C. P. A. PRADO. 2005. Re- productive modes in frogs and their unexpected Blair (ed.), Evolution in the Genus Bufo, diversity in the Atlantic Forest of Brazil. BioScience pp. 119–170. University of Texas Press, Austin. 55:207–217. TARANO, Z. 2001. Variation in male advertisement calls HAGMAN,M.,AND R. SHINE. 2006. Spawning site in the Neotropical frog Physalaemus enesefae. selection by feral Cane Toads (Bufo marinus)atan Copeia 2001:1064–1072. invasion front in tropical Australia. Austral Ecol- TUTTLE, M. D., AND M. J. RYAN. 1982. The role of ogy 31:551–558. synchronized calling, ambient light, and ambient HOOGE,P.N.,AND B. EICHENLAUB.1997.Animal noise, in anti-bat-predator behavior of a treefrog. movement extension to ArcView, version 1.1. Behavorial Ecology and Sociobiology 11:125–131. U.S. Geological Survey, Anchorage, AK. VITT, L. J., J. P. CALDWELL,G.R.COLLI,A.A.GARDA,D. JOFRE´ , G. M., C. J. READING, AND I. E. DI TADA. 2005. O. MESQUITA,F.G.R.FRANÇA,D.B.SHEPARD,G.C. Breeding behavior and reproduction in the Pampa COSTA,M.M.VASCONCELLOS, AND V. DE NOVAES E de Achala Toad, Bufo achalensis. Amphibia-Reptilia SILVA. 2005. Uma atualizaçaõ do guia fotogra´fico 26:451–458. dos re´pteis e anfı´bios da regiaõ do Jalapaõno KENNEDY, A. B. 1982. Ecce Bufo: the toad in nature and Cerrado Brasileiro. Special Publications in Herpe- in Olmec iconography. Current Anthropology 23: tology, Sam Noble Oklahoma Museum Natural 273–290. History 2:1–24. CALLING BEHAVIOR OF RHINELLA OCELLATA 621

WELLS, K. D. 1977. The social behaviour of anuran WICKLER, W., AND U. SEIBT. 1974. Rufen und Antworten amphibians. Animal Behaviour 25:666–693. bei Kassina senegalensis, Bufo regularis und anderen ———. 1988. The effect of social interactions on Anuren. Zeitschrift fuer Tierpsychologie 34:524– anuran vocal behavior. In B. Fritzsch, M. J. Ryan, 537. W. Wilczynski, T. E. Hetherington, and W. Walk- owiak (eds.), The Evolution of the Amphibian Auditory System, pp. 433–454. John Wiley, New York. Accepted: 4 June 2007.