SHORT COMMUNICATIONS 343

Speckled Crakes usually inhabit dense vegetation and This specieshas been thought to be rare, occurring lo- are glimpsedonly when burstinginto flight, aswhen fiight- cally over a large region (Blake 1977). Although it is not ened by a person or a harvesting machine. On such oc- common at our study area, our experience indicates that casions they fly as high as 6 m and drop back into the it is not so much scarceas difficult to find. The same is vegetation after traveling about 70 m. Despite their con- true of other small, timid, and supposedlyrare crakessuch trastingly patterned plumage, they are difficult to see in as Laterallus xenopterusand Micropygia schomburgkii, the field. being obscuredin the dim light of their habitat. which inhabit dense vegetation. Two specimenshave been obtained in our study area. An adult male was collectedon 14 May 1976 by Herculano We thank Herculano Alvarenga for information and Alvarenga and is now in his private collection. Its testes specimens,and J. Nacinovic, C. Tribe, and E. Willis for werewell-developed and its stomachcontained small (about suggestionson the manuscript. Our researchwas partially 1 mm) seedsof Graminae (800/& pieces of small arthro- supportedby the ConselhoNational de Desenvolvimento pods (15%), and fine gravel (l-3 mm; 5%). The second Cientifico e Tecnolgico. , capturedon 2 1 August 1982, is still alive in captivity with us as of this writing, and we have learned some details LITERATURE CITED of behavior by watching it. When alarmed, many rails flick the erect tail, in some BLAKE,E. R. 1977. A manual of neotropical . Uni- casesshowing contrastingspots on the &sum and under versity of Chicago Press, Chicago. tail coverts. Our captive SpeckledCrake, however, seems MEYERDE SCHAUENSEE,R. 1962. Notes on Venezuelan to show alarm by remaining motionless in a horizontal birds with a history of the rail Coturnicopsnotata. position, with the tail held downward and fully opened. Not. Nat. (Phila.) 357:1-8. The wings are folded but held obliquely upward, showing PINTO,0. M. 0. 1938. Catalogodas aves do Brasil. Rev. the whitish tips of some secondariesand exposing the Mus. Paulista 22. profusely white-spotted underparts. PINTO,0. M. 0. 1964. Omitologia Brasiliense. Depto. The SpeckledCrake is diurnal, retiring to roost at dusk. Zoologia, Secretariade Agricultura, S&oPaulo. As do some other rails (e.g., Aramides cajanea and A. PINTO,0. M. 0. 1978. Novo catalog0das aves do Brasil. ypecaha,Teixeira 198l), our captive roosts on a perch up Conselho National de Desenvolvimento Cientifico e to 2 m above the ground. It sleepswith its head under its Tecnologico (CNPq), Sao Paulo. wings and with one foot retracted into the belly plumage. RIPLEY,S. D. 1977. Rails ofthe world. David R. Godine, The bird also calls frequently at night from itsperch,at Boston. least in cantivitv. Unlike other rails. its vocalizations are TEIXEIRA,D. M. 1981. Notas sobrea saracuratres-potes, inconspicuousand easily confusedwith or masked by oth- Aramides cajanea (Muller, 1776): a ocorrtncia do er marsh sounds.When giving its distinct two-syllable call, ninho criadeira. Bol. Mus. Paraense Emilio Goeldi the crake stands upright with neck erect. In this call, (Zool.) 110:l-24. “kooweee-tack,” the “kooweee” is high and brief, and the “tack” is louder and drier, ending abruptly. We have also Se@0 de Omitologia, Museu National, Rio de Janeiro heard a whistling “keeee” in an alarm situation and a single (RJ), CEP 20942, Brazil. Received 13 July 1983. Final high “kyu.” acceptance6 February 1984.

The Condor86:343-344 helps explain certain records of the speciesdistant from 0 The Cooper Ornithological Society 1984 previously known populations. The Serra dos Caraias. a reaion of about 18.000 km* THE between the Araguaia/To&t& and Xingu rivers, is cen- tered at approximately 6”OO’S, 50”3O’W. The serra is of (PROCNIAS ALBA) IN Cretaceousage and reachesover 800 m at its highestpoints. THE SERRA DOS CARAJAS, The region is exceptionally rich in minerals, including SOUTHEASTERN PARA, BRAZIL iron, manganese,copper and gold. The vegetation of the slopesis high Amazonian rain forest in most places,while a rupicolousscrub known as “canga” or “camp0 rupestre” PAUL ROTH covers exposed iron ore deposits on some of the high plateaus.The observationsreported here were made dur- DAVID C. OREN ing six visits to the Carajh region in 1983:January (FCN), July (PR), August (DCO), and May, June and September AND (Goeldi Museum technicians).The specimensreported were FERNANDO C. NOVAES collected in August and September 1983. We thank Hel- mut Sick for sharing information he obtained during a visit to Carajas in May 1969. In the Carajasregion, White Bellbirds occurin montane The known geographicaldistribution of the White Bellbird forest at altitudes of 500 to 750 m. Much of the high Am- (Procnias alba) encompasses eastern Venezuela, the azonian rain forest on the slopes is especially luxuriant Guianas, and Brazil north of the Amazon, including the with a canopy over 35 m high, while in other places it is lower and middle Rio Negro (Novaes, Ann. Sot. Sul-Rio- more open with many emergents.White Bellbirds are lo- grandse Omitol. 1:5-7, 1980; Snow, D., The . cally common, at least seasonally,but with a patchy dis- Cornell Univ. Press.Ithaca. NY, 1982). Here we report a tribution. The males’ loud calls were conspicuousfrom population of White Bellbirds south ofthe Amazon River June to September, but they were not noted during visits in the Serra dos Carajas, southeasternPara, Brazil. This in January and May. It is unclear whether this indicates extendsthe range of the speciessome 1,000 km south and migration away from Carajasoutside the breeding season, 344 SHORT COMMUNICATIONS or only the inconspicuousnature of the birds when they in millimeters are as follows: wing 153-159 (K= 155.5); are not calling. During July and August, White Bellbirds tail 93-97 (95); tarsus29-30 (29.7); billculmen 18.2-19.9 were common in most valleys along a 60-km west-east (19.2); bill, nostril to tip 9-12 (10.7); billgape 33-36 (34.7). transectthrough the highest parts of the serra. They were Compared to Snow’s (1973, 1982) data for northern absent at the two ends of the transect (“Pojuca” in the birds, the Carajasmales are somewhatsmaller in wing and west and near the airport in the east), and at several in- tail measurements,while the tarsusand, especially,the bill termediate placesthat appearedto have appropriate hab- are larger. The size of the bill appears to be of adaptive itat. Strong dependenceon fruits from speciesof arbores- importance becauseit determines the maximum diameter cent is thought to influence local distribution of exploitable fruits. The White Bellbirds of Carajas have ofbellbirds (Snow, D. W., Bull. Br. Mus. Nat. Hist. (Zool.) the largestbills of any Procnius,and presumablyeat larger 25367-391, 1973), but we have no data on tree species fruits than those of any other population. It may be sig- composition in the valleys. The observedpopulation den- nificant that the Carajas specimensin life had black upper sity ranged from 2 calling males within hearing range in mandibles and medium gray lower mandibles tipped with some valleys, to at least 10 calling males in others. black. The lower mandibles turned completely black with- The males called from exposedperches 25 to 40 m high in a few days after death. Bill color in other populations in forest at sites ranging from the bottom of the valleys of bellbirds is described as all black (Snow 1982), so the (approximately 500 m) to the highest forested ridges. We true color of the lower mandible in life may be another heard calls at all hours of the day from 06:30 to 17:30. distinguishingfeature of the Carajas population. They conformed well with those described for the White This southernpopulation of P. albu occursnearer to the Bellbird in Guyanaby B. K. Snow (Auk 90:743-75 1,1973). closestknown populationsof the Bearded Bellbird (P. av- The most frequently uttered call was the normal sharp eruno)in southernMaranhHo (about 600 km east) than to bell-like “ding-ding” or “kong-kay” call. Sometimes the the closest known conspecific populations (about 1,000 vocalization was audibly harsher and louder, which is km north). Followina Snow’s (1973) interoretation. the probably this call accompanied by a rapid swing of the Carajas population may have ‘been ‘established by ‘long head and body. The birds also emitted the musical “doing- distancedispersal from the Guiana highlands, sufficiently doing” or “dor-ong” call, albeit more rarely than the data long ago to have allowed for some differentiation. White indicatefor .The birds’ visitingbehavior also agreed Bellbirds frequent lower altitudes than any of their con- with that describedby B. K. Snow (1973). One of us (PR) geners(Snow 1973, 1982), and extensive tracts of lowland once observed two males on the same perch above the forest would not represent an insurmountable barrier to canopy.The secondmale to arrive evidently was the owner dispersal.Wallace ’s (A narrative of travels on the Amazon of the perch. He performed a “ritualized jumping move- and Rio Negro. Ward, Lock and Co., London, 1853) rec- ment,” but without leaving the perch, and then uttered a ord of White Bellbirds near the citv of BelCmin Julv 1849 bell call at the moment when the other male flew OK now seemsless puzzling; those birds may have been wan- Breedingprobably takesplace in Augustand September, derers from the southern population. The appearanceof but may extend into other months as well. Females were vagrants far from their breeding grounds is well-estab- observed in associationwith adult males in August. The lished among the bellbirds. The records for the middle testes of two adult males collected in August were ap- and lower Rio Negro (Novaes 1980, Snow 1982) may be proximately the same size (10 x 4 mm and 10 x 5 mm) seenin new light. Given the incomplete stateof knowledge as those of a singlemale taken in September (9 x 5 mm). about Amazonian bird distributions, it seems probable According to D. W. Snow (Ibis 118:366-401, 1976), male that unidentified populations of P. ulbu exist in montane bellbirds take no part in nest duties and are free to molt forest in that region. during the reproductive season.The onset of molt in male We thank the following institutions for collaboration Procniascoincides with egg-layingby females. In all three with this work: Companhia Vale do Rio Dote, the Bra- male specimens collected at Carajas (Museu Paraense zilian National ResearchCouncil, the Federal University Emilio Goeldi collection 35042 [25.VIII.83], 35043; of MaranhHo, and the Museu ParaenseEmilio Goeldi. Part [26.VIII.83], and 35302 [10.1X.83]), the first primary was of Roth’s technical equipment was generouslyprovided new and the secondand/or third were in molt. The middle by the Zoological Museum of the University of Zurich and remigeswere the most worn while the outermost appeared the Janggen-PbhnFund, St. Gallen, Switzerland. practically new. The tail showed no molt, but the inner rectriceswere much more worn than the outer ones. This Depurtamentode Biologia, UniversidadeFederal do Maru- pattern of wear is what might be expected in birds that nhdo, Largo dos Amores, 21, 65.000 Sio L&s, Mara- take 160 days to go through a molt cycle (Snow 1976). nhrlo,Brazil. Addressof secondand third authors:Depur- One of the specimenshad a few dark patches on three tamento de Zoologia,- Museu Paruense Emilio Goeldi, of the innermost secondaries,indicating that it was prob- Cuixa Postal 399, 66.000 BelPm, Par&, Brazil. Received ably a third-year individual (Snow 1976), while the other 17 August 1983. Final acceptance13 April 1984. two males were completely white. The standardmeasures