COLORPL^FE. ChapadaFlycatcher, a new speciesof tyrant-flycatcher(Tyrannidae: Suiriri) from the cer- rado regionof centralSouth America. The male is pictured in profile on the left; the femaleis displaying with wings raised,on the right. From a watercolorpainting by Daniel E Lane. The Auk 118(1):56-78, 2001

A CRYPTIC NEW OF FLYCATCHER (TYRANNIDAE: SUIRIRI) FROM THE REGION OF CENTRAL

KEVIN J. ZIMMER,1'2'5 ANDREW WHITTAKER, 3 AND DAVID C. OREN4 tLosAngeles County Museum of NaturalHistory, 900 ExpositionBoulevard, Los Angeles, California 90007, USA; 21665Garcia Road, Atascadero, California 93422, USA; 3Estradodo Aleixo, Conjunto Acariquara Sul, Rua da Samaumas214, Manaus,Amazonas, 69085; and 4Departamentode Zoologia, Museu Paraense Em•'lio Goeldi, Caixa Postal 399, 66.017-970 Beldm,Pard, Brazil

ABSTRACT.--Anew speciesof (Suiriri islerorum) is describedfrom the cer- radoregion of Braziland adjacenteastern . The species previously had been confused with Suiriri suiriri affinis,with which it is syntopicat multiple sites.The new specieswas first identifiedby voice.Although cryptically similar to S.s.affinis in manyrespects, the new speciesis readilyidentified by all vocalizations,bill size,color pattern of the tail, and shape of the centralrectrices. Most distinctive are the male-femaleduets, which are accompanied by dramaticwing-lifting displaysnot performedby any congeners.Reciprocal playback ex- perimentsof tape-recordedvocalizations demonstrated that the new speciesand S.s.affinis do not respondto oneanother's vocalizations. We provide information on thenatural history of the new flycatcher,along with spectrogramsof its variousvocalizations. We alsoprovide vocalanalysis of all othernamed taxa in Suiriri,and discussthe variousintrageneric rela- tionships.In particular,S.s. affinisand S.s. bahiae,although distinct morphologically, are vocallyand behaviorallysimilar, and respondto oneanother's vocalizations in playbackex- periments.Received 27 December1999, accepted5 September2000.

THE SUIRIRI has been the focus of tax- bahiaeas a subspeciesof S. affinis,from which onomicdebate for decades.Two speciestradi- he foundit to differ from typical formsby "up- tionallywere recognized, a short-billed,white- pertail covertsdark hair brown like the tail; bellied nominate form Suiriri suiriri; and a rectriceswithout any yellowishat the baseand longer-billed, yellow-bellied form, S. affinis. without the pale brownishapical band." The apparentintermediacy of a large seriesof Zimmer (1955) pointed out an additional Suiririfrom northeasternParaguay led Zimmer complication:five AMNH specimensof S.affin- (1955)to suggestthat the two taxa were con- is from Mato Grosso and Goias, Brazil, all of specific.Meyer de Schauensee(1966) and Tray- which were anomalousin having a short, lor (1979)followed Zimmer in recognizingonly broad-basedbill and with the "pale terminal onespecies of Suiriri,whereas Short (1975) did band on the rectricesunusually wide and dis- not agreethat Paraguayanbirds showedsigns tinct, exceedingany other specimenat hand of intergradation and continued to recognize whether from the same or other localities and two species.Traylor (1982) reexamined the Par- whether suiriri, affinis,or bahiae."Zimmer fur- aguayanmaterial in the AmericanMuseum of ther noted that "the significanceof this com- Natural History (AMNH), and found it as de- binationof charactersis completelypuzzling, scribedby Zimmer, concurringthat S. suiriri sincethere is no allied group toward which and S. affinisshould be treatedas conspecific. these features,singly or together,suggest a More recent authors have remained divided: trend." In his reexaminationof Suiriri, Traylor Sibley and Monroe (1990) recognizedtwo spe- (1982)commented on the five aberrentAMNH cies of Suiriri, whereas Ridgely and Tudor specimensand an identical specimen from (1994) recognizedonly one. As currentlytreat- Maranhao, Brazil (from the Field Museum of ed, the genusalso includes one additional tax- Natural History,Chicago; hereafter FMNH). In on, S.s. bahiae,a yellow-bellied taxon from additionto the shortbill and broad,pale tips to northeasternBrazil. Hellmayr (1927) treated the rectrices,the six specimensfurther couldbe distinguishedfrom typical affinisby their dis- E-mail: [email protected] tinctly broader central rectricesand a lack of

57 58 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118 any contrastingpale edging to those feathers specimensand tissue sampleswere deposited (Traylor 1982). Traylor noted that "ordinarily with the MPEG. We also located additional the close correlation of two such discrete and sitesof sympatrybetween the two vocaltypes. unrelated characters as the short bill and the colorationand shapeof the rectriceswould be METHODS strongevidence that we have two siblingspe- cies.I think that eventuallythis will prove to We made observations of Suiriri at various sites in be true,but the possibilitythat the short-billed Mato Grosso,Brazil (August and October1991, and may somehowbe relatedto the intergra- every Septemberfrom 1993 to 1999); Pernambuco, Brazil (eachJanuary from 1996to 1999,and February dation betweenaffinis and suiriri cannotbe ig- 1996);Goias, Brazil (January1996); and Amapa,Bra- nored at this time." He concludedby stating zil (February1998 and 1999).All measurementsused "most important,without field studiesof the in behavioraldata (e.g. distances,heights) are esti- various taxa, it is uselessto speculate." mates. Sexual identification of birds in the field was During the courseof severalyears of field basedon sex-specific,stereotyped displays or vocal- work at various sites in Brazil Zimmer and izations,the sexual specificityof which was con- Whittaker found that there were two markedly firmed by postmorteminspection of specimensthat different vocal types of yellow-bellied Suiriri, we collected.Mapped distributions(as they appear oneof whichwas relatively widespread, occur- in this paper) are basedentirely on label data from ring from Amapa to Bahia to Goias;and the specimensthat the seniorauthor examined, on more recentrecords documented by tape recordings,and other which we knew only from the Chapada on our own collectingsites. Those localities were en- dosGuimaraes region of Mato Grosso.The first tered into a sector-basedGeographic Information includedbirds that, by both distribution System(Isler 1997)and mappedby Morton Isler. and morphologicalcharacters, clearly were as- We assume that vocalizations of Suiriri, like those signableto S.s. bahiae,as well asbirds that ap- of other suboscines,are mostly or entirelyinherited peared to be typical of S.s. affinis.The second (Kroodsma1984, 1989; Kroodsma and Konishi1991), type was distinctive not only for its vocaliza- and as such,provide potentiallyinformative char- tions, but also for its dramatic wing-flapping acters for systematicstudy as in other suboscines displaysthat invariablyaccompanied all terri- (Lanyon 1978, Isler et al. 1997, Krabbeand Schulen- torial duetsby mated pairs (seebelow). Recip- berg 1997).To analyzevocalizations, we assembled recordingsof all namedtaxa of Suiririfor compari- rocal playback experimentsconducted at mul- sonwith our own recordings.Locations and record- tiple sitesrevealed that the two vocaltypes did istsfor all recordingsexamined are listedbelow. For not respondto one anothers'vocalizations. In comparison,vocalizations were categorizedas loud- September1998, we found the two vocaltypes songs,duets, or calls."Loudsongs" were consistent- occurring syntopically near Cuiaba, Mato ly patternedmultinote vocalizations (Isler et al. 1997) Grosso,Brazil. Close comparison revealed that givenby an individualbird, seeminglyin thecontext onevocal type wasnoticeably short-billed com- of territorialadvertisement. Those were givenindi- pared to the otherand had a distinctbuffy ter- vidually by both sexes,but most frequentlywere in- minal fringe to the tail, and that the birdswere corporatedinto duets. "Duets" involved simulta- alwaysassortatively paired with their own vo- neoussinging by a pair of Suiriri, and consistedof typical loudsongsof both sexes,combined with oth- cal and morphologicaltypes. er calls.Vocalizations categorized as callsmost often Suspectingthat the short-billedbirds repre- were structurally simple (typically involving only sentedthe samebirds as the anomalousspeci- oneor a few notes;exceptions are notedbelow) and mens reported by Zimmer and Traylor, we re- usually were given in the contextof contactvocali- turned to the Cuiabaregion in September1999 zationsbetween mates or family members.Our re- with Dionisio Pimentel of the Museu Paraense cordingswere made with SonyTCM-5000 tape re- Emilio Goeldi, Belem (hereafter MPEG) and corders and Sennheiser ME-80, ME-66, and MKH-70 collected10 specimensof eachvocal type. Prior shotgunmicrophones. Spectrograms were madeby to collecting,we madevoucher tape-recordings PhyllisIsler on a MacintoshG4 computerusing Ca- nary version1.2.1 (BioacousticsResearch Program, of eachbird (all recordingsto be archivedat the Cornell Laboratory of Ornithology, Ithaca, New Library of Natural Sounds,Cornell Laboratory York).Canary parameterswere standarddefault ex- of Ornithology, Ithaca, New York; hereafter ceptthat frame overlapused was >-96.88%. LNS). Tissue samplesalso were preservedfor Playback experimentswere conductedto deter- molecular analysis of genetic differences.All minereactions of Suiririof eachtype to vocalizations January2001] NewFlycatcher from Cerrado Region 59 of the others.In eachcase, an individual or pair of Museum of Natural History, Yale University (YPM). flycatchersfirst was presentedwith a prerecorded A wing rule with a perpendicularstop at zero was tape of anothertaxon. Each prerecorded tape includ- used to measureflattened wing chord (wing), tail ed a combined10 to 15 songsand duets(interspersed length(tail), breadthof the pale terminaltip to the with calls)from two pairs of birds of a given taxon rectrices(pale tip), and width of the centralrectrix and geographiclocality. For the S.s. suiriritape, we (centralrectrix), and dial caliperswere usedto mea- usedonly songsand calls because we had no record- sure tarsuslength (tarsus),culmen length from the ings of duets. For all trials, each taxon was repre- anterior end of the nares to the tip (culmen), bill sentedby a singleplayback tape consistingof vocal- depth at the anterior end of the nares (bill depth), izations of two pairs (or individuals in the caseof and bill width at the anterior end of the nares (bill nominateSuiriri) of birds.During eachtrial, thetape width). All calipermeasurements were madeto the of 10 to 15 songsand duetswas playedto conclusion. nearest0.01 mm, those taken with the wing rule To avoid confounding effects, a buffer period of were made to the nearest 0.5 mm. General linear about2 min was used to separateplayback of vocal- models were used to investigategender-corrected izationsof differenttaxa. Flycatchers that did not re- differencesamong taxa for each of the characters spondto playbackof othertaxa then were presented measured.Residual diagnosticsindicated no viola- with a playbacktape of their own vocal type. That tionsof the generallinear modelassumptions of er- wasdone to accountfor potentialseasonal influences ror normality and constantvariance. Plumage was by resolvingthe questionof whetherbirds that were describedfrom specimensand comparedto a stan- unresponsivein the first trials were discriminating dard color reference(Smithe 1975). betweenthe different vocalizations,or were gener- ally nonterritorialduring the trial periodand there- RESULTS fore unresponsiveto playbackof any kind. Failureto account for seasonal differences in responsiveness hasbeen cited as a potentialdesign weakness in oth- Examinationof our specimensand tape re- er playbackexperiments (Kroodsma 1986). Respons- cordings,study of a large selectionof museum es to playback were characterizedas "strong," specimens,extensive field observationsthat in- "moderate,.... weak," or "none." A strong response cluded reciprocal playback experiments,and involved immediate and repeated vocalizations discoveryof multiple sitesof sympatrywithout (from previouslynonvocalizing birds) as well as ap- evidenceof interbreedingconvinced us that the proach toward the sound source.A moderatere- two vocal types of yellow-belliedSuiriri rep- sponseinvolved either a cautiousapproach without resentdistinct biological species. Furthermore, vocalizing, or sustained vocalizing without ap- it is clear that the short-billed form that we col- proach.A weak responseinvolved single or unsus- lected from Mato Grosso is the same unnamed tainedvocalizations (from previouslynonvocalizing birds)without approach.The nonecategory includes form that was first pointed out by Zimmer instancesin whic•hbirds remainedsilent (in the case (1955) and later hypothesizedto be a cryptic of previouslynonvocalizing birds) and did not ap- speciesby Traylor (1982).We proposeto name proach the sound source,as well as instancesin this flycatcher which alreadyvocalizing birds neitherchanged the delivery or rate of their vocalizations,nor ap- proachedthe sound source. A potentialfifth category Suiriri islerorum sp. nov. wouldbe a "negativeresponse," in whichthe subject ChapadaFlycatcher respondedto the soundstimulus by movingfar- ther away.No negativeresponses to playbackwere Holotype.--MPEG No. 54867, adult female observedfrom any taxa. from kilometer 2 on the road to Agua Fria, Zimmer examined representative specimens of 15ø25'S,55ø46'W, west of the town of Chapada Suiriri (n = 170) for comparisonwith our 20 speci- dos Guimaraes, approximately 720 m eleva- mens from Mato Grosso. Those specimensare tion, state of Mato Grosso, Brazil, collected 21 housed at the Museu Paraense Emilio Goeldi, Belem, September1999 by DionisioPimentel, tape-re- Brazil(MPEG); Academy of Natural Sciencesof Phil- cordedby Kevin J. Zimmer. Voice specimento adelphia,Philadelphia (ANSP); CarnegieMuseum, be archived at Library of Natural Sounds Pittsburgh(CM); Field Museumof Natural History, (LNS), CornellLaboratory of Ornithology,Ith- Chicago(FMNH); Los AngelesCounty Museum of aca, New York. Natural History,Los Angeles (LACM); the Louisiana State University Museum of Natural Science,Baton Diagnosis.--Similarin plumage and size to Rouge(LSUMZ); the Museumof VertebrateZoology, Suiriri s. affinis(Tables 1 and 2), but differs as Berkeley(MVZ); the National Museum of Natural follows:bill distinctlyshorter, relatively broad- History,Washington, D.C. (USNM); and the Peabody er, and more nearly uniform blackish (baseof 60 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118

TABLE1. Measurements(mm) of Suiriri islerorumand congeners.Values are means _+standard deviation followedby ranges,with samplesize in parentheses.

Characters S. islerorurn S. affinis S. bahiae S. suiriri Culmen 8.0 _+ 0.7 9.6 -+ 0.6 8.5 ñ 0.7 7.7 _+ 0.4 7.5-8.6 (23) 8.2-11.1 (61) 7.9-9.2 (3) 6.7-8.7 (108) Bill depth 4.2 _+0.3 4.6 _+0.3 4.5 _+0.9 4.2 _+0.3 3.7-4.9 (23) 3.8-5.4 (61) 4.5-4.6 (2) 3.7-4.9 (106) Bill width 5.7 ñ 0.3 5.4 _+ 0.4 5.5 ñ 0.1 5.0 ñ 0.3 5.1-6.3 (23) 4.7-6.1 (61) 5.4-5.6 (2) 4.4-5.6 (107) Wing 82.6 _+3.3 82.0 _+4.4 77.2 ñ 4.0 73.0 ñ 3.3 72-87 (21) 70-88.5 (38) 73.5-81.5 (3) 65.5-81 (108) Tail 71.5 ñ 2.7 69.4 + 3.8 66.7 ñ 4.2 69.0 ñ 3.6 66-78 (21) 59-76 (38) 62-70 (3) 58-78.5 (108) Tarsus 20.3 + 0.7 20.5 + 1.0 20.3 ñ 0.7 19.8 _+ 0.7 19.1-21.7 (20) 17.2-22.5 (37) 19.8-21.1 (3) 17.9-22.1 (106) Pale tip 6.4 + 1.5 2.1 ñ 1.5 1.2 ñ 2.0 0.0 - 0.0 3-11 (23) 0.0-4.8 (61) 0.0-3.5 (3) 0.0 (108) Central rectrix 10.5 _+ 1.1 8.6 ñ 1.3 7.0 + 1.0 -- 8-12 (21) 6-11 (38) 6-8 (3) --

mandiblefleshy-pink in many specimensof af- S. islerorurn;only to the bottomof thebill in all finis, particularly in females);dorsal surfaceof S.s. affinisexamined). Suiriri islerorurndiffers most of tail darker, more blackish-brown further from specimensof S.s. bahiaein having (browner, between Fuscous,color #21, and Van- the grayish crown and nape less sharply de- dyke Brown, color#121, in affinis);pale termi- marcatedfrom the greenerback; back darker nal fringe to tail broaderand morehighly con- and less distinctly green; breast less whitish; trasting (either absent,or, if present,narrow rump and uppertail covertsmore contrastingly and weaklycontrasting in affinis);both websof pale; and central rectrices broader and not outer rectricesmore extensivelypale, and dark pale-edged. Suiriri islerorurndiffers from S.s. subterminalpatch on inner web of outerrectri- suiriri as follows:distinctly larger overallwith cesdistinctly smaller;central rectrices broader proportionately shorter, broader tail; belly, and without paler edges;white of throat ex- flanks,thighs, and crissumyellow ratherthan tendsfarther up on sidesof face(to the gapein white; undertailwith a broad,contrasting pale

TABLE2. Meansñ SD (samplesizes in parentheses)of measurements(mm) of Suiriri islerorumand S.s. affinisby sex,followed by a statisticalcomparison of the meansof thepooled samples (both sexes) for each species.Values of P (= probabilityof a greatervalue of F) derived from Analysisof Variance(ANOVA).

islerorum affinis Characters Males Females Males Females P

Culmen 8.1 ñ 0.4 8.0 ñ 0.3 9.8 -+ 0.5 9.3 ñ 0.5 <0.0001 (6) (16) (31) (26) Bill depth 4.2 ñ 0.4 4.2 ñ 0.3 4.7 ñ 0.3 4.6 ñ 0.4 <0.01 (6) (15) (31) (26) Bill width 5.8 ñ 0.4 5.6 ñ 0.2 5.5 ñ 0.3 5.3 + 0.4 <0.01 (6) (15) (31) (27) Wing chord 85.1 ñ 0.7 81.6 ñ 3.6 85.0 _+2.3 78.6 ñ 3.8 <0.05 (5) (14) (lS) (17) Tail 71.2 ñ 1.6 71.0 _+ 2.7 70.9 ñ 2.9 67.8 _+ 4.1 <0.01 (5) (14) (lS) (17) Tarsus 21.2 ñ 0.6 20.0 ___0.06 21.2 ñ 0.6 19.7 + 0.8 NS (5) (14) (17) (17) Pale tip 6.9 ñ 0.8 6.2 ñ 1.7 2.2 ñ 1.5 2.2 + 1.6 <0.0001 (6) (16) (31) (27) Central rectrix 9.6 ñ 1.1 10.9 ñ 0.9 8.8 _+ 1.2 8.3 + 1.3 <0.0001 (5) (16) (lS) (17) January2001] NewFlycatcher from Cerrado Region 61

fusinginto a grayerhindcrown and nape, clos- est to Glaucous(color #79). The grayishnape gradesinto an Olive (color#30) back and scap- ulars. Uppertail covertsand rump contrasting- ly paler than back,between Light Drab (color #119C) and SmokeGray (color#45). Auriculars concolorwith hindcrown, perhaps slightly duskier Malar region between whitish and Pale Neutral Gray (color #86). Lores slaty. A narrow, pale, supraorbitalbrow extendsto the forehead.Four long rictal bristleson both sides of gape,the longest8.0 mm. Wings generally dusky, closestto Vandyke Brown (color #221), but secondariesnarrowly pale-edged (closest to Light Drab, color#119C) on outerweb. Basal portion of inner web of all remiges broadly pale-edged,between Pale Horn (color#92) and FIc. 1. Displaying pair of Suiriri islerorum:female in foregroundwith wingsraised. Key morphological Drab-Gray (color #119D). Greater and median featuresthat distinguishthis species from S.s.affinis secondarycoverts broadly tipped Light Drab can be seenin this photo:the short, thick bill, and (#119C),forming two distinct wingbars.Axil- the broad, pale, terminal fringe to the tail. Photo- lariesand underwinglinings pale SulphurYel- graphednear Cuiab•, Mato Grosso, Brazil on 22 Sep- low (color #57). tember 1999 by Kevin J. Zimmer. Chin and throatwhitish, grading posterior- ally into Pale Neutral Gray (color#86). Center and sidesof breastGlaucous (color #80), sepa- base (undersideof tail in S.s. suiriri entirely ratingwhiter throatfrom SulphurYellow (color dark exceptfor contrastinglypale narrow outer #57)belly, flanks, thighs, and undertail coverts. web of outerrectrix on eachside and pale shaft Most of exposeddorsal surfaceof tail dark, be- of outer rectrices);central rectricesdistinctly tween Dusky Brown (color #19) and Dark broaderand not narrowlypale-edged on either GrayishBrown (color#20). Contrastingpaler web; inner web of outer rectrix pale-basedand band at tips of rectricesbetween Drab (color pale-tipped,not entirelyblackish; broader gray #27) and Dark Drab (color#119B); that paler tip band acrossthe breast; less extensivelydark 7 mm in lengthon centralrectrix. Base of both auriculars;and lessdistinct white supraorbital webs of outer rectrices, and base of inner web brow. Suiriri islerorum differs from both Suble- of all otherrectrices contrastingly pale, closest gatusmodestus and Sublegatusarenarum as fol- to Cream color (#54). That paler base to tail lows: markedly larger; greener upperparts mostly concealedfrom aboveby uppertail co- showingmore contrastbetween nape and back verts, but visible from below as a wide (35 mm) (brownerwith little or no contrastin Sublega- pale base that contrastswith a dark central tus);rump and uppertail covertscontrastingly paler than back (no contrastin );up- panel (closestto Vandyke Brown, color #221), persideof tail blacker(less brown) with a pale whichin turn, contrastswith the pale terminal baseto therectrices; throat contrastingly white. fringe.Basal portion of the shaftsof mostrec- The new speciesdiffers from mostmembers of trices are closestto Pale Horn (color #92); those of the central rectrices are closer to Salmon Elaeniain lacking elongatecrown feathersand a contrastingly colored, concealed coronal (color #6). Outermost rectrix on each side has patch,as well asin havingan entirelyblackish the outer web almostentirely pale, with a dark bill. The contrastinglypale rump and baseof subterminalpatch (26 mm in length)on the in- the tail are characters not found in . ner web.The pale outerweb of the outerrectrix, Descriptionofholotype.--See Figure 1 and col- in combinationwith the pale baseto the rectri- or plate. Capitalized colornames and numbers ces and the pale terminal fringe, isolateand are from Smithe (1975). Forecrownbrownish outlinethe dark centralpanel on the underside gray,closest to Hair Brown (color#119A), suf- of the tail. 62 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118

Tocantins, and Maranhao. The westernmost recordsare from Campodo Estanho,Amazon- as, Brazil (M. Cohn-Haft tape recording),and ParqueNacional Noel Kempff Mercado,depto. Santa Cruz, Bolivia (LSUMZ 150871). Un- doubtedlyoccurs more extensivelywithin the area boundedby thoserecords, but probably overlookeddue to morphologicalsimilarity to S.s. affinis. Elevationaldistribution.•The only specimens of Suiriri islerorum with elevational data on the labels are the ones that we collected in Mato Grosso,Brazil. Thoserange from 250 m on the Coxipodo Ouro Roadnear Cuiaba, to 720m on the Agua Fria Road near Chapada dos Gui- maraes. That elevational range also encom- passesall of our recordsthat havebeen docu- mented by tape recordings. The Bolivian specimen(LSUMZ) was from 450 m (J. Bates FIG. 2. Distribution of Suiriri islerorum and its pers. comm.).Paynter and Traylor (1991)pro- congenersas confirmed by examinationof specimens vide elevations for several of the sites from or tape recordings:black circles= S. islerorum;dia- which S. islerorumwere collected by other monds= S.s. affinis;circles touching diamonds = workers.Those should be viewedas pertaining sites where islerorumand affinis are sympatric; squares= S.s.suiriri; plus signs = siteswhere affinis to generalcollecting localities, and may not ap- and suiriri are sympatric;and triangles= S.s. bahiae. ply specificallyto siteswhere Suiriri were col- "T" next to a symbolindicates the type localityfor lected.Including those data would extendthe that taxon. "T" next to a circle-diamond locates Cha- probableelevational range of S. islerorumfrom pada dos Guimarfies,Mato Grosso,Brazil, the type 150 m at Tocantinia, Tocantins, Brazil (LACM localityfor S. islerorum.Note that no type localityis 46004) to 1200 m at Sao Joaoda Allanna, Goias, located for S.s. bahiae,which was described from a Brazil (LACM 40087). trade specimenwhose only locality was "Bahia." Specimensexamined: Skins.•nly specimens Map by Morton Isler. measuredare listed. Plumagesof all sympatric speciesof Elaeniaand Sublegatuswere com- Soft parts in life: iris, medium brown; max- pared superficially. Suiriri islerorum:Brazil, ilia, black; mandible, basal one-quarter slaty Mato Grosso, six males and nine females gray,remainder black; lining of the mouth,yel- (MPEG 54865 to 54874, and 19299; AMNH low-orange;tarsus and toes,black; soles of feet 127882,127883, 33101, and 33135);Goi•s, three and toes,paler gray; nails, black. females (LACM 460087 and 40089, MPEG Measurementsof holotype.--Culmenfrom an- 19299);Maranhao, one female (FMNH 63435); terior edge of nares,7.8 mm; bill depth at an- Para, one female (MPEG 49036); Tocantins,one terior edgeof nares,4.4 mm; bill width at an- female (LACM 46004); Mato Grossodo Sul, one terior edge of nares, 5.8 mm; wing chord, unsexedspecimen (ANS 159808).Bolivia, dep- 85 mm; tail, 73 mm; tarsus, 20.1 mm; body to. Santa Cruz, one female (LSUMZ 150871).Of mass22.8 g. Ovary, 6 x 3 mm; oviduct,3 mm those,four males and six females(MPEG 54865 wide; largest ovum, 4 mm. to 54874) were collectedby us in Mato Grosso, Geographicdistribution.--Locally distributed Brazil. in cerrado and camposhabitats throughout Suiriri suiriri affinis:Brazil, Amap•, 10 males much of central Brazil, south of the Amazon and 9 females (MPEG 46607 to 46619, AMNH and eastof the Rio Madeira,and at leastspar- 33100, USNM 515199);Goias, five males, six fe- ingly to depto.Santa Cruz, easternBolivia (Fig. males and two unsexedspecimens (LSUMZ 2). Most confirmed records are from Mato 31774, 67522, 67523; ANS 186661; USNM Grosso,Brazil, with scatteredspecimens east 516085; MPEG 22013; FMNH 344593 to 344595 and north to the Brazilian states of Par•, Goi•s, and 75099; LACM 46003, 40090, and 40088); January2001] NewFlycatcher from Cerrado Region 63

Mato Grosso, seven males and six females providedin the correspondingfigure legends. (MPEG 54875 to 54884, AMNH 33100, 33131, All recordingsby Kevin J. Zimmer unlessoth- and 33133); Maranh•o, one male and five fe- erwise indicated. Suiriri islerorum: Brazil, Mato males (FMNH 63431, 63432, 63434; MPEG Grosso,west of Chapadados Guimar•es along 43526 to 43528); Minas Gerais, one male and the road to Agua Fria, 20 specimens,and Cox- two females (MPEG 36210, FMNH 56657, ipo do Ouro road near Cuiab•, 8 specimens; LACM 40086); Tocantins, one male (LACM Amazonas,160 km E-SE of Humaita, 1 speci- 46005);Cear•, one male (USNM 264644);PiauL men (M. Cohn-Haft).Suiriri suiriri affinis: Bra- one unsexedspecimen (MPEG 50917).Bolivia, zil, Amap•, 12 specimens;Goi•s, 2 specimens depto. Santa Cruz, three males (LSUMZ (A. Whittaker);Mato Grosso,west of Chapada 124634, 124642, 150872). Of these, six males dosGuimaraes along the road to Agua Fria, 4 and four females(MPEG 54875 to 54884) were specimens,and Coxipo do Ouro road near Cu- collectedby us in Mato Grosso,Brazil. iab•, 24 specimens.Bolivia, depto.Santa Cruz, Suiriri suiriri bahiae:Brazil Bahia, one male 1 specimen(T. A. Parker,LNS 52044). Suririri and two females (LACM 37067 and 37068, suiriri bahiae:Brazil, Pernambuco,near Lagoa FMNH 64119). Grande,16 specimens.Suiriri suiriri suiriri:Ar- Suiriri suiriri suiriri: , Entre Rios, gentina,Salta, three specimens(S. Hilty). Bo- seven males, five females, and one unsexed livia, depto. SantaCruz, one specimen(T. A. specimen(YPM 66068 to 66073 and 83326, CM Parker,LNS 33629).Brazil, Rio Grande do Sul, 140482, 140489, 140490, 140670, 140806, and one specimen (W. Belton, LNS 19531). Para- 140867); Misiones, five males and five females guay,Chaco Province, one specimen (D. Finch, (LSUMZ 56714 to 56723); Tucuman, three LNS 57891). males,four females,and two unsexedspeci- Etymology.--Wetake great pleasurein nam- mens (FMNH 58012, 58019, 58369, and 58377; ing thisspecies after our goodfriends and col- USNM 284898, 285001, 285002, and 285109; leagues,Morton and PhyllisIsler, in recogni- YPM 24195);La Pampa,one male and two fe- tion of their numerous contributions to males (USNM 284329, 284332, and 284335); Neotropicalornithology. Mort and Phyllisau- Chaco, one male and one female (USNM thored a landmark monographon tanagers 284330 and 284334); Buenos Aires, two males (Isler and Isler 1987),and they have madepi- (USNM 55681 and 55683);Salta, one male (CM oneeringcontributions in developingmethod- 45839); and Santiago del Estero, one male ologiesfor the analysisof vocal charactersto (MVZ 108323). Bolivia, Santa Cruz, 18 males, assessspecies limits in antbirdsand othersub- 14 females, and 1 unsexed specimen (CM oscines (Isler et al. 1997, 1998, 1999). Their 119820, 120212, 120213, 32786, 32787, 32831, work on antbirdshas served to heightenaware- 32889, 32891, 32952, 51402, 78970, 79244, nessof the importanceof vocalizationsas tax- 80363, and 80364; FMNH 181493, 181502, onomiccharacters in all suboscinegroups, and 181506, 294367, 295413, 296295, 296298, as such,has far-reachingimplications for the 296301, 335151, and 335152; LSUMZ 124633 to accurateassessment of regional endemismand 124637, 153757, and 153758; YPM 31816 and biodiversity,and, ultimately, for conservation. 31817); Tarija, five males and three females The Islershave contributed greatly to our own (ANS 138951, 138952, 138954, 138955, 138958, studiesof Neotropicalbirds, and are constant and 138959; FMNH 294369 and 294370); Co- sourcesof advice, encouragement,and help. chabamba,one male and two females(FMNH TheEnglish name calls attention to thetype lo- 181498 and 181500;ANS 136168). , cality,the Chapadados Guimar•es, Mato Gros- depto. Presidente Hayes, two males (MVZ so, Brazil, which is an importantreservoir of 168120and 168121);depto. Chaco, four males threatened cerrado flora and fauna. (FMNH 153007, 153010to 153012).Brazil Mato Grosso do Sul, one male and two females REMARKS (FMNH 64121 to 64123); Parana, one male and onefemale (USNM 16367and 16368)., Sexualdimorphism and variation in thetype se- RepresaPalmar, one male (ANS 187738). ries.--The type seriesof Suiriri islerorumcon- Specimensexamined: Tape recordings.--Data sistsof 23 adult specimens(6 males,16 females, for recordings reproduced as sonogramsare and 1 unsexedspecimen). Males resemblefe- 64 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118 males, and plumage variation among para- wear. We believe that all of those birds were in types appearsto be neithersexually nor geo- breeding condition.We never have observed graphically influenced. There is some family groupsof S.islerorum during our August individual variationin color of the uppertail and Septembervisits to Mato Grosso.That, covertsand in the extentof the contrastingly combinedwith the apparentbreeding readi- palebase of the rectrices.Compared to the ho- ness and territorial nature of the birds we col- lotype, someindividuals (MPEG 54872, LACM lectedin lateSeptember, suggests that the main 46004, LSUMZ 150871, AMNH 127882) are hatchingand fledging period is probablyin Oc- more greenish-yellow(closest to Citrine, color tober.That would fit a commonpattern for in- #51) on the uppertail coverts,another (MPEG sectivorousbirds in regionswith pronounced 54870) was buffier, and two (MPEG 54871 and dry seasons,in which hatchingcoincides with FMNH 63435)had the uppertailcoverts nearly onsetof the rainy seasonand the concomitant the same color as the back. Two specimens flush of insect populations(Walsberg 1977, (MPEG 54865 and LACM 40087) are more ex- Marr 1981, Zimmer 1993). tensivelygray dorsally,with the mantle and The four AMNH specimensfrom Mato Gros- lowerback more nearly concolor with thenape. so were collectedbetween 7 February and 26 Similar variation in rump-lower-back contrast May. All had remigesand rectricesthat were was apparent amongthe 61 specimensof S.s. less worn and more broadly pale-eged or affinisexamined. fringed than in our specimens.Similarly, the Most observed variation among paratypes three LACM specimensfrom Goi•s and Tocan- was related to feather wear and molt condition. tins were collectedbetween 14 April and 3 May, Most specimensare browneron the forecrown and all werein fresherplumage than our series and grayer on the hindcrown,but a few are from Mato Grosso.The ANSP specimen(ANSP moreuniformly gray throughout the crown.In- 159808)was collected30 Juneand was in fresh dividual feathers of the forecrown have a condition. The LSUMZ specimen (LSUMZ brownishlongitudinal streak along the shaft, 150871) was collected29 August, and was in and are gray around the fringe. When those fresh condition with a trace of molt on the head feathersare worn,the extentof thegray relative and with an ovary that measured6 x 7 mm. to the brownish central streak is reduced, and The conditionof thosespecimens suggests that the forecrownassumes a morebrownish tinge. S. islerorurnundergoes one molt sometimeafter Freshly molted birds had grayer forecrowns. breedingin Septemberto October,and then Similarly,variation in the breadth of the pale probably undergoes at least a partial molt terminal fringe to the rectrices,the buffy tips again in spring or summerprior to the next of the secondarycoverts, and the pale edgesof breedingseason. the remigesall wasinfluenced by wear. Vocalizations.--Male and female Suiriri isle- One male specimen(MPEG 54871) had an rorumhave different loudsongs, which are typ- unusuallylong tip (hook)to the maxilla. The ically delivered as sex-specificelements of si- culmen measurement for that bird was 8.52 multaneous duets, but are occasionally mm, but without the tip the culmen would deliveredseparately. The male loudsong(Fig. measure 7.64 mm. In general, males were 3A) is a loud seriesof paired couplets,all notes slightlylarger than females, although bill mea- having a distinctlytwangy quality.The song surementswere nearly identical(Table 2). hastwo discerniblephrases that are repeated in Breedingand molt.--Ten of the 23 specimens predictablepatterns (where where, whooz it whooz of Suiriri islerorum(MPEG 54865 to 54874) were it, where where whooz it, whooz it, whoozit, whooz collected from Mato Grosso, Brazil between it). Sometimes males will sing a series of 20-22 September1999. All had enlarged go- "whooz it" notes without the "where where" nads, and two females (MPEG 54866 and notes. The number of consecutive "whooz it" 54872)had well-developedbrood patches.All couplets may vary from two to six (rarely were vocal and territorial. Most birds showed more), but no more than one "wherewhere" cou- signsof moderateto heavywear of the second- plet seemsto be givenwithout interjecting the ary covertsand rectrices.Similarly, the FMNH "whooz it" notes. In response to playback, specimenfrom Maranhfiowas collected2 Sep- malesfrequently sing longersongs. The most tember 1925, and showedmoderate to heavy commonlyheard male calls are isolatednotes January2001] NewFlycatcher from Cerrado Region 65

Suiriri islerorum MALE 6 A loudsong B calls

4

3

2 •' 1

•o i i i i i i i i i • 0 0.5 I .0 I .5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

FEMALE "' 7 C loudsong D "wheer-deer" E "zhuwheep""zhuwheep-oo"

..•

i i i i i i i i 0 0.5 I .0 1.5 2.0 2,5 3.0 3.5 4.0 4.5 5.0 Time (seconds)

FIG. 3. Vocalizationsof Suiriri islerorumrecorded from Mato Grosso(MT), Brazil. MPEG numbersin pa- renthesesmatch tape recordingswith specimensin the MuseuParaense Emilio Goeldi,Be16m, Par J, Brazil. (A) Male loudsong(MT, near Cuiab& 22 September1999). (B) Male (MPEG no. 54874)calls (MT, Chapada dos GuimarSes,21 September1999). (C) Femaleloudsong (MT, near CuiabJ, 22 September1999). (D) Female (MPEG no. 54872) "wheer-deer"call (MT, Chapadados GuimarSes,21 September1999). (E) Female"zhu- wheep"and "zhuwheep-oo"calls (MT, near CuiabJ, 22 September1999). All recordingsby Kevin J. Zimmer. or coupletssimilar to thosegiven in loudsongs ble of notes, with the female rattle and the (Fig. 3B). The female loudsong(Fig. 3C) is a twangy male notesequally distinct. loud,bubbly rattle of variablelength that is typ- As with S. islerorum, the male and female ically precededby one or two well-differenti- loudsongsof S.s. affinisare most commonly ated "whur" or "wheer"notes. In responseto heard as sex-specificelements of simultaneous playback,females typically lengthenthe rattle duets,but are sometimesdelivered separately. elementof their songs.Females commonly give The male loudsongof S.s. affinisis a seriesof two types of calls. The first is a series of sneezy"pi-chew" notes, variable in length,that "wheerDEER"calls (Fig. 3D) that aresomewhat typically begins with severalsqueaky intro- similarto the male couplets,but that are more ductory notes and ends with severalone-syl- emphatic(particularly on the secondsyllable), labled"chew" notes (Fig. 5A). The femaleloud- with a lesstwangy, less strident quality, which song(Fig. 5B) is a seriesof loud, squeakynotes are frequentlyinterjected into duets.Outside of that typicallydecelerates noticeably toward the duets,the mostcommonly heard femalecall is end. Thosenotes are similar in quality to the a queruloussounding, two-or-three-syllabled introductorynotes in malesongs, but arehigh- "zhuwheep"or "zhuwheep-oo,"in which the long er frequency,higher amplitude, and more first note is diphthongaland distinctlyup- widely spaced.The terminal notesin a female ward-inflected at the end, and the second, loudsong often drop in frequency. Females shorternote, when present, drops in frequency have a variety of single-notecalls, most of (Fig.3E). That call is typicallygiven in the con- which are nasal and have a distinctlywhiny text of a contactcall when a femaleis visually quality(Fig. 5C, D). Thoseare routinely uttered separatedfrom her mate.In duets(Fig. 4), the ascontact calls, but whena femalegives several male and femalesing their loudsongssimulta- differentiatedwhiny callsin rapid succession neously,creating a somewhatdiscordant jum- ("suiree,swee, swiroo, chew") it is usually the 66 ZIMMER,WHITTAKER, AND OREIN [Auk, Vol. 118

Suiriri islerorum lO A male/female duet 8

6

4

2

o 0

B male/female duet

10.0

8 C male/femaleduet ! ß ..;, '...',.:..., ......

i i i i i i i i i 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Time (seconds)

FIG. 4. Male-female duets of Suiriri islerorum recorded from Mato Grosso (MT), Brazil. MPEG numbers in parenthesesmatch tape recordingswith specimensin the Museu ParaenseEmilio Goeldi, Be16m,Parfi, Brazil. (A) male-female (MPEG nos. 54865 and 54866) duet (MT, near Cuiabfi, 20 September1999). (B) Male- female (MPEG nos. 54874 and 54872) duet (MT, Chapada dos Guimarfies,21 September1999). (C) Male- female(MPEG nos. 54869and 54867)duet (MT, Chapadados Guimarfies,21 September1999). The female from this pair (MPEG no. 54867)is the holotypefor the species.All recordingsby KevinJ. Zimmer. synchronizingcue that begins a duet.A female, The male loudsongof S.s. bahiaebegins with a whosemate we had collected,repeated those rapid seriesof squeaky,somewhat run-together single-notenasal calls for more than an hour notesthat drop in frequencybefore yielding to afterwards. The most frequently heard male a distinctive series of "pitty-chew" elements call outsideof a duet contextis a single-note, (Fig. 7A). The introductoryseries recalls the somewhatguttural "werk" (Fig. 5E). Lesscom- soundsthat would be madeby squeezinga toy monly heard is a low-frequencyrattle that may rubber duck severaltimes in rapid succession. functionas an aggressioncall (Fig. 5E G). We The female loudsong(Fig. 7B) is a seriesof have heard that call given severaltimes from squeaky "choop" notes (peak frequency of birds of unknown sex within foragingfamily about 8 kHz) delivered at about 8 to 10 notes groups,as well as from a female that was at- per second.Those notes are similar in tonality tempting to aggressively displace a Lesser to thosethat precedethe "pitty-chew"elements Elaenia (Elaeniachiriquensis) from a fruiting in the male song,but are higher in frequency, tree.In duets(Fig. 6), the higheramplitude fe- more widely separated,and do not drop in fre- male loudsongstend to mask the male loud- quencythrough the series. Both sexes give their songs.The similarityof the femalenotes to the respective songs during simultaneousduets introductory male notes makes it even more (Fig. 6D). In duets,the femaleloudsong is typ- difficult to distinguishthe full male song. ically louder than that of the male. A single All vocalizationsof S.s. bahiaeappear to be duet bout might include two or more songs homologousto the vocalizationsof S.s. affinis. from eachbird, with the final female songin January2001] NewFlycatcher from Cerrado Region 67 the bout often deceleratingnoticeably toward sumeda prominentperch in the crownof a tree the end of the song.The terminalnotes in such or on a baresnag, usually within 0.5 m of one a deceleratedseries are lower in frequencythan another Duets were always accompaniedby the introductorynotes. Males frequentlycon- dramatic, sex-specific,stereotyped motions. cludeduets with a few sputtering"pitty-chew" The male flickedboth wings high abovehis calls. Outside of duets, the most frequently back (the wings nearly touching)at a rate of heard male call is a guttural, singlenote (Fig. about one per secondfor the duration of the 7C). Thiscall apparentlyfunctions as a contact duet (6-10 s). With each downstroke of his call.The seriesof "pitty-chew"calls also is fre- wings,he flippedhis tail upwards,at abouta quentlygiven independent of the squeakypre- 45ø angle. When the wings were once again lude (Fig. 7D), particularlyin duets.The most flickedupwards, the tail was flipped down 30 commonlyheard femalecall is a nasal"nyew" to 45ø below the horizontal.During the display, with a distinctlywhiny quality.Those calls are the maleleaned forward, so that his rump was typicallydelivered singly in the contextof con- held at aboutthe sameplane as his head. The tactcalls, but theyalso are oftenstrung togeth- tail wasfanned throughout much of the display, er in rapid succession(Fig. 7E), particularly by emphasizingthe contrastbetween the yellow- birds agitated by tape playback. Femalesre- ish-buff base and the blackish distal half. Some sponding to playback typically begin giving males rotated at approximatelya quarter turn thosewhiny notesin flight as they approach with eachup-and-down cycle of the wingsand the soundsource. As in S.s.affinis, a rapidlyde- tail, suchthat they were constantlyturning in livered seriesof thosefemale notes following circleswhile displaying.Simultaneous with the playbackappears to be the synchronizingcue male'sdisplay, the femalealso raised her wings thatprecipitates the responsorialduet from the high aboveher back, but heldthem there, fully- pair The only vocalizationof S.s. affinisfor fanned, for about 2-3 s at a time before lower- which we couldfind no homologin S.s. bahiae ing them gradually.Females repeated that mo- wasthe rattlecall (Fig.5EG). The apparentab- tion two or threetimes during each duet. At the senceof sucha vocalizationin S.s. bahiaemay sametime that a femalewas raising her wings, be an artifact of sampling,because rattle calls shewould alsofan and dip her tail downward were only occasionallyrecorded from S.s. at a 20 to 45ø angle, flipping the tail backup to affinis. the horizontalposition as the wingswere low- Our inventory of tape recordingsof nomi- ered.The fanningof the wingsby femalesac- nate S.s. suiriri is small.What appearsto be a centuatedthe largely pale inner webs of the loudsongin Argentina (Fig. 8A) is a long (3 to remiges such that backlit birds appeared to 5 s), frenetic,sputtering series of nasal notes havetranslucent wings. Females also frequent- that rise and fall in frequency,with occasional ly rotatedwhile displaying,but wereless likely differentiated querulous notes scattered to do so than the frenetic males. Females also through the middle. We do not know if this maintaineda moreupright posture throughout loudsongis specificto one sex, and, if so, to their displays.Those displays of S.islerorum ac- which one. The most common vocalization is a companiedeach of the more than 200 duet singlenasal "row?' or "jyow"note that is re- boutsthat we witnessed,and are remarkably peated incessantlyfor prolongedperiods. Mi- similar to thoseof the Streamer-tailedTyrant nor variations of this call l•ave been recorded (Gubernetesyetapa) of southernSouth America, from Argentina, Brazil, Bolivia, and Paraguay and, to a lesserextent, of White-beardedFly- (Fig. 8B-E). Anothercall recordedfrom Para- catcher(Phelpsia inornata) of Venezuela(K. Zim- guay is a 2-3 s rattle at about5-6 kHz (Fig. 8F). mer pers. obs.). Duetdisplay behavior.--The most striking be- We never have witnessedsimilar displays haviors of S. islerorum were those associated from eitherS.s. af,finis or S.s. bahiae,in spiteof with territorialduets of matedpairs (Fig. 1 and having witnessedmore than 200 duetsof the colorplate). Pairs were tightly synchronizedin former and more than 100 duets of the latter. their duets. The first few notes of either a male Sick(1993) described the songof S.s. affinisas or female loudsongwere enoughto initiate "repeatedwhile lifting bothwings vertically." mostduets, with the secondbird joiningin im- That descriptionalmost certainly is basedupon mediately.Prior to singing, pairs typically as- observationsof S. islerorum.Suiriri s. af,finis and 68 ZIMMER,WHITTAKER, AND OREIN [Auk, Vol 118

Suiriri suiriri affinis A maleloudsong 5- ... •. intro "pl-chew"elements "chew" notes 4-

3-

2-

1

0 I i i i i i i i 0 I .0 I ,5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

10 B femaleloudsong 9-

8-

7-

6

5

4

3

I I I I I i i i i 0 0.5 I .0 I .5 2.0 2.5 3,0 3.5 4.0 4.5 5.0

C female"whiny" calls D female"whiny" calls

5

4

3

2

1

0 i i i i I I 0 0.5 I .0 I .5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

male"guttural" call F female rattle call G rattle call 6- ......

5-

4

3

2

1

0 I I I I i i i • 0 0.5 I .0 I .5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Time (seconds)

FIC. 5. Vocalizationsof Suiriri s. affinisfrom various sitesin Amap• (AP) and Mato Grosso(MT), Brazil. MPEG numbersin parenthesesmatch tape recordingswith specimensin the MuseuParaense Emilio Goeldi, Be16m,Par•, Brazil. (A) Male (MPEG no. 54881)loudsong (MT, near Cuiab•, 22 September1999). (B) Female loudsong(MT, near Cuiab•, 21 September1998). (C) Female"whiny" calls(MT, near Cuiab•, 22 September 1999). (D) Female"whiny" calls (AP, north of MacapP, 12 February 1999). (E) Male guttural call (MT, January2001] NewFlycatcher from Cerrado Regwn 69

10 Suiririsuiriri qf. finis A male/female duet male/femaleloudsongs ... female"whiny" calls

i i i 0 1.0 2.•0 3;0 4.0 5.0 6JO 7:0 8'.0 9'.0 10.0

10 Sui• suid• affinb B male/femaleduet male/femaleloudsongs 8 female"whiny" calls • , } 6

....

1.0i 2.0i 3.0i 4.0i 5.0i 6. i0 7.0i 8.0i 9.0i 10.0 • Suiririsuiriri afJ•nis male/femaleduet male/femaleloudsongs female'lwhiny" calls

I i i i i i i i i 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0

1 0 Suiririsuiriri bahiae D male/femaleduet female"whiny" call femaleloudsong maleloudsong

I I I I I I I I l o I .0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 lO.O Time (seconds)

FIG.6. Male-femaleduets of Suiriris. affinisand S.s. bahiaefrom variouslocalities in Amap• (AP), Mato Grosso(MT), and Pernambuco(PE) Brazil.(A) S.s. affinisfrom Chapadados Guimarfies, MT (21 September 1999).(B) S.s.affinis from Chapadados Guimarfies, MT (21 September1999). (C) S.s.affinis from nearPorto Grande, AP (7 February1999). (D) S.s. bahiaefrom Lagoa Grande, PE (15 January1997). All recordingsby Kevin J. Zimmer.

S.s. bahiae exhibited identical behaviors when male,land next to her, and thenthe pair would displaying.Duets almostalways were initiated simultaneouslysing their respective loud- by severalnasal, single-notefemale calls (Fig. songs.The postureof both birds during duets 6A-D) given in rapid succession.If the male wasnearly horizontal with thewings typically was in a differenttree or anotherpart of the drooped slightly at the sides. Birds usually sametree, he would rapidly toward the fe- quiveredtheir wings slightly while keeping

Chapadados Guimarfies,21 September1999). (F) Rattlecall from female(MT, near Cuiab•, 22 September 1999).(G) Rattle call of bird of unknownsex (AP, north of Macap& 12 February1999). All recordingsby Kevin J. Zimmer 70 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118

Suiriri suiriri bahiae

A maleloudsong

õ-

4-

2

0 o 0.5 1. 1. 2.0 2.5 3.0 315 410 4.•5 5 lO B femaleloudsong 9-

8-

7-

6-

5

4

3

2

1

0 0 0.5I 1.0i 1.5'1 2.0i 2.5i 3.0i 3.5i 4.0i 4,5i 5.0

C male"guttural" call D male"pitty-chew"

0 0.5' 1. '0 I .5' 2.0' 2. •5 3. '0 3.5' 4.0' 4.5' 5.0

E female"whiny" calls 7-

6-

5-

4-

3

2

1

0 0 0.5 1. 1.5 2. 2.5 3.0 3.5 4.0 4 5 5.0 Time (seconds)

FIC. 7. Vocalizationsof Suiriri s. bahiaerecorded near LagoaGrande, Pernambuco, Brazil. (A) Male loud- song (25 January1999). (B) Femaleloudsong (25 January1999). (C) Male guttural call (25 January1999). (D) Male "pitty-chew"calls (25 January 1999). (E) female"whiny" calls(25 January 1999). All recordingsby Kevin J. Zimmer January2001] NewFlycatcher from CerradoRegion 71

Suiriri suiriri suiriri

7- A presumedloudsong .....

6-

5

4

3

2 1 ..... '"...... 0

0.5i 1.0i 1.5i 2.0i 2. 15 3.0i 3.5i 4.0i 4.5i 5.0

B single-notecall C single-notecall D single-notecall E single-notecall .;;•'. •-...

•...• j'•..•'.

I i I 0 0.5 1 0 1.5 2 0 2.5

F rattle call

......

0 i ) i i i i i i i 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Time (seconds) FIG.8. Vocalizationsof Suiriris. suiriri. (A) loudsong(Argentina, Salta, 6 October1987; S. Hilty record- ing). (B) Single-notecall (Argentina,Salta, 6 October1987; S. Hilty recording).(C) Single-notecall (Brazil, RioGrande do Sul, 50 km southwest Santo Antonio das Miss6es, 4 November 1975; W. Belton recording, LNS 19531).(D) Single-notecall (Bolivia,depto. Santa Cruz, west of SanIsidro, 6 October1983; T. Parkerrecord- ing, LNS 33629).(E) Single-notecall (Paraguay,Chaco Province, Filadelfia, 21 August1990; D. Finchrecord- ing, LNS 57891).(F) Rattlecall (Paraguay, Chaco Province, Filadelfia, 21 August1990; D. Finchrecording, LNS 57891). themdrooped, and sometimessimultaneously back by flyingin and perchingnear us without shiveredthe tail up-and-downin a shallowarc vocalizing. Even after leaving and being of lessthan 5 ø. Unlike S. isIerorum, neither affinis broughtback by tape playbacktwo or three norbahiae habitually selected prominent perch- times,those males would not sing on their own. es from which to display,and often duetted In eachcase, a femaleeventually flew in giving fromwithin thecanopy of trees.Birds respond- severalnasal calls and the pairs immediately ing to tape playbackoften sangas they were beganto duet. flying towardthe soundsource, a behaviorthat Tapeplayback experiments.--Playback experi- we did not observe in S. isIerorum. On several mentswith Suiririrecordings offer further ev- occasionswe had affinismales that were sepa- idenceof the significanceof vocal differences rated from their matesrespond to tape play- amongthe taxa,and of their role as potential 72 ZIMMER,WHITTAKER, AND OREN [Auk,Vol. 118

TABLE3. Summaryof playbacktrials (asdetailed in the text) involvingSuiriri flycatchers (S. islerorurn, S.s. affinis,S.s. bahiaeand S.s.suiriri). Numbers in parenthesesindicate the numberof pairsor family groups of the subjecttaxon that were testedfor eachstimulus taxon.

Responseto playback of Locality islerorum affinis bahiae suiriri S. islerorum Mato Grosso,Brazil strong(11) none (11) none (11) none (6) S.s. affinis Amapa, Brazil none (3) strong(3) strong(3) none (2), moderate(1) Mato Grosso,Brazil none (12) strong(12) strong(12) -- Totals none (15) strong(15) strong(15) none (2), moderate(1) S.s. bahiae Pernambuco,Brazil none (5) strong(5) strong(5) none (3), strong(2) isolating mechanisms.We performed tape to whichnone of themresponded. We then pre- playbackexperiments on 11 pairs of Suiriri is- sentedthem with playbackof nominateS.s. lerorumnear Cuiaba and Chapadados Guimar- suiririfrom Argentina.The adult pair from the aes,Mato Grosso,Brazil (six pairs in Septem- first family group respondedby approaching ber 1998and five pairsin September1999). The the soundsource and peeringabout on two oc- six pairsin the 1998trials werepresented with casions, but did not vocalize. None of the four recordingsof S.s.bahiae from Pernambuco,Bra- birdsfrom the otherfamily groupshowed any zil, S.s. affinisfrom Cuiab•, Mato Grosso,Bra- interestin playbackof nominateS.s. suiriri, nor zil and S.s. suiriri from Argentina. The five did the lonepair. When presented with tapeof pairsin the 1999trials were presented with the S.s. affinisfrom Mato Grossoand S.s. bahiae samerecordings of S.s. bahiaeand S.s. affinis. from Pernambuco,all three pairs of adult S.s. None of the 11 pairs tested showedany re- affinisfrom Amapa respondedstrongly by ap- sponseto tape playbackof suiriri,affinis, or bah- proachingand duetting. iae vocalizations.When presentedwith play- In January1999, we performedtape playback back of prerecordedS. islerorumvocalizations, experimentson five pairs of S.s. bahiaenear La- all 11 pairs responded by immediately ap- goa Grande,Pernambuco, Brazil. Thosebirds proachingthe soundsource and duetting mul- were first presentedwith taped vocalizations tiple times.In severalcases, those birds flew in of S. islerorum from Mato Grosso, Brazil. None from >100 m away when respondingto play- of the five pairsshowed any responseto those back of their own vocalizations. playbacks.We thenpresented them with play- We performedtape playbackexperiments on back of S.s. affinisfrom Mato Grosso.All five 12 pairs or family groupsof S.s.affinis near Cu- pairsresponded strongly to playbackof S.s.af- iaba and Chapadados Guimaraes, Mato Gros- finis,always approaching the soundsource and so,Brazil (six pairs in September1998 and four duettingby the fourthplayback. We alsopre- pairs and two family groupsof four in Septem- sentedthe five pairsof S.s.bahiae with playback ber 1999).Birds in all 12 trials first were pre- of nominateS.s. suiriri from Argentina.Three sentedwith taped vocalizationsof S. islerorum of the pairs did not respondto the voice of from Chapada dos Guimaraes, then with re- nominatebirds. The othertwo pairsresponded cordingsof S.s. bahiaefrom Pernambuco,Bra- strongly (approachand vocalizations),but it zil. None of the 12 pairs or family groupsre- shouldbe pointedout that they wereprevious- sponded to playbacks of vocalizationsof S. ly stimulatedby tape of affinisfrom Mato Gros- islerorum,but all 12 pairs of adults responded so.The resultsof thoseplayback trials are sum- strongly to playback of S.s. bahiaevocaliza- marized in Table 3. tions.We alsoperformed tape playbackexper- Foragingbehavior and socialstructure.--Suiriri iementson two family groupsand one pair of islerorumtypically was encounteredin closely S.s. affinisnorth of MacapP, Amap•, Brazil associatedpairs, apart from otherspecies. They (February1999). Those birds first were pre- foragedmostly in trees (rarely descendingto sentedwith playbackof S.islerorum recordings, low shrubsor the ground),moving energeti- January2001] NewFlycatcher from Cerrado Region 73 cally with shortpauses between hops and fre- ilar to foragingtechniques routinely employed quent and erratic changesof direction.They by bluebirds(Sialia spp.; K. Zimmer pers.obs.) gleanedvarious small arthropod prey from leaf in .Suiriri s. affinis also made oc- and branchsurfaces. Once prey was spotted, casionalshort, looping sallies of 1-2 m out they frequently used wing-assistedhops to from crowns of trees in pursuit of flying in- closedistance rapidly, then lunged forward for sects,but that behaviorwas lessfrequent than a horizontal capture,or jumped upward to in S. islerorum.Like S. islerorum,affinis periodi- gleana preyitem from overhead.They also en- cally wagged its tail downward in a relaxed gagedin frequentdarting aerialsallies upward mannerwhen foraging, particularly after perch or outward to glean prey from undersidesof changes.Foraging behavior of S.s. affinisob- leaves or branches. Those sallies varied from 15 servedin Amap•, Brazil was identicalto that cm to 1 m and typicallyinvolved little truehov- described above from Mato Grosso. ering. Thosebehaviors were similar to that of The social structure of S. islerorumrequires membersof the generaTolmomyias and Platyr- further investigation.We have never encoun- inchus,although S. islerorumwere more con- tered more than two birds together.However, stantly in motion and were more agile and ac- in September1999, we had two experiences robatic in pursuing prey (K. Zimmer pers. that suggestthat S. islerorumis at least occa- observ.).Suiriri islerorumalso made occasional sionally polygynous. On 21 September 1999, longer(up to 2 m), loopingsallies out from the we taped a pair of S. islerorumalong the Agua crown or lateral branches of trees and back to Fria Road, Mato Grosso,Brazil. After obtaining pursueflying insects.Perch changes were usu- vouchertapes, we collectedthe displayingfe- ally followedby an obvious,but relaxeddown- male (MPEG 54867).The male flew a short dis- ward dip of the tail, similar to tail movements tanceout of sight.Immediately after retrieving of VermilionFlycatcher ( rubinus). the female,we broadcastmore tape playback Suiriri s. affinisin the sameareas frequently from the samespot, and a male returned.As perch-gleanedor hover-gleanedsmall fruits soon as the male perchedin front of us, a sec- (particularlythose of Curatellaamericana), al- ondbird flew into thesame tree and sang a typ- thoughthey spentmore time in pursuit of ar- ical female loudsong.The male immediately thropods.Suiriri s. affinis foraged mostly in rel- chased that second female, pursuing her ativelyopen trees, progressing by hopsof 5 to aroundtrees and shrubs.This wasrepeated on 15 cm, primarily alongmain branches.Their five occasions;each time we presentedtape movementsthrough trees were generallymore playbackof duets the male would approach, leisurelyin pace (almost Vireo-like)and more followedclosely by thesecond female, which he predictablein progressionthan in S. islerorum. would then aggressivelydrive off. Only once Birds pausedfor 2-15 s betweenhops to scan during those encountersdid the two birds live foliage, fruit clusters, or bark surfaces. duet. On our last playback attempt, the male They generallyworked from the interior of a drove the female to the ground and appeared tree outward, or from the subcanopyup, but to mounther. They then flew to a nearbytree without frequentchanges of directionas in S. and performedmore duetsbefore we collected islerorum.Arthropod prey usually(-75% of all both birds (male MPEG 54869, secondfemale captures)were gleaned directly from leaf or MPEG 54868). Gonadsof all three birds were branchsurfaces by reachingout, up, or down. enlarged. Foragingbirds oftenpicked directly at branch Later on 21 September,we were collecting surfaces,particularly on trees with furrowed farther along the Agua Fria Road when we bark. They alsofrequently hover-gleaned prey taped in another pair of S. islerorum.Once by hovering1-2 s beneathgreen leaves in the again, we collectedthe female (MPEG 54872) outerportion of the canopy.That was different first, and the male flew off. After a few minutes from the darting sallies(with little or no true we presentedmore playbackof duet vocaliza- hovering)made by S. islerorum.Suiriri s. affinis tions. Immediately,a pair of S. islerorumflew often dropped 1-3 m from branchesto the into the same tree from which we had collected groundto pounceon prey spottedfrom above. the female, and proceeded to duet multiple That behaviorwas particularlycommon in re- times. After obtainingvoucher tapes, we col- centlyburned or very openareas, and wassim- lectedboth birds (secondfemale MPEG 54873, 74 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118 male MPEG 54874).The first femalehad a con- fying the sexof a territorialtresspasser (Fara- spicuousbrood patch and the gonadsof all baugh1982). A likely byproductof sexualspec- threebirds were enlarged. ificity of duet roles is the strengtheningof In each of those instances, what we assumed reproductiveisolation. In an habitually duet- to be the originalmale appearedto return with ting genussuch as Suiriri, vocal differences be- a secondadult female in breeding condition tween taxa would seemto presentmajor obsta- within 5 min of the time that we had collected cles to reproductive compatability.Further his previousmate, which suggeststhat those barrierswould be presentedby significantdif- maleshad beenmated to two birds all along.It ferencesin stereotypedphysical displays as- seemshighly unlikely that morethan onemale sociated with the vocalizations. As detailed was involved in either case. There was no evi- above,pairs of S. islerorumperform distinctive denceof a secondterritorial male or pair in the displayswith eachduet. Nothing that we have vicinity prior to collectionof the first females. heard or seenin the vocal and display reper- Had we been at a boundary between two ter- toire of affinisor bahiaesuggests an approachto ritories,it seemslikely that our tape playback islerorum. wouldhave elicited responses from both neigh- The biologicaland taxonomicsignificance of boring pairs.Instead, playback brought an im- thosedifferences is suggestedby the resultsof mediate approachand duet responsefrom a our playbackexperiments. Although we were singlepair. Followingour collectionof the fe- unableto perform all pairwise tests,the ones male of eachpair, the respectivemales flew off performedwere unequivocal. None of the affin- out of sight,after which a male flew backfrom is or bahiaetested showedany responseto re- roughly the same area minutes later in re- peated playbacksof islerorum,nor did any of sponseto further playback.Birds responding the islerorumrespond to playbacksof the other to the secondround of tape playbackperched taxa. in the sameor adjacentshrubs as those used by Morphologicaldifferences from affinisand the birds respondingto the first tape play- bahiaeare subtle yet distinct, even when ob- backs.We haveonly rarely foundpairs of S.is- servedin the field.Among the eightcharacters lerorumoccurring in closeenough proximity to that were measured, Suiriri islerorum differed oneanother that we couldhear (evendistantly) significantlyfrom affinisin culmenlength, bill two pairs from a single spot, making it even less likely that two independent territorial width, bill depth, wing chord, tail length, width of the central rectrix, and width of the pairswere involvedin eachinstance. Systematicrelationships.--Some generaliza- pale tip to the rectrices(Table 2). The two spe- tions regardingrelationships among the yel- ciesalso differed in the colorpattern of the out- low-belliedgroup of Suirirican be made.Suiriri er rectrices, and in whether or not the central islerorumis sufficientlydistinct to be consid- rectriceswere pale-edged. Our sampleof affinis ered a separatespecies under any of the widely indicatedsubstantial sexual dimorphism in bill acceptedspecies concepts (McKitrick and Zink size and color of the mandible, whereas no such 1988). Vocal distinctions are particularly dimorphismwas apparentin islerorum.It is not strong.Suiriri islerorumdiffers strikingly from known whetherjuveniles of S. islerorumshare nominatesuiriri, affinis,and bahiaein its male the unique spottedpattern to the upperparts and femaleloudsongs, duets, and in all known found in juvenilesof other Suiriri. male and female calls. Duets serve multiple Most important, Suiriri islerorumand S.s. af- functions, including coordinateddefense of finis are syntopicat multiplesites, and show no territory and pair-bonding.In Suiriri, asin the evidenceof interbreeding.At the two localities majorityof duettingspecies, the sametypes of in Mato Grosso, Brazil, where we conducted duets are used during bouts of territorial de- field studies,both forms were relatively com- fense, as a prelude to copulation,and when mon and all pairs of Suiriri were assortatively pairs reunite followingvisual separation(Far- mated by vocal type and morphologicaltype abaugh1982, K. Zimmer pers.observ.). In sex- (as confirmedby tape recordings,specimens, ually monochromaticspecies (Suiriri), sexual and tissuesamples). Thus, they meet the pri- specificity of duet contributions probably mary criteria for being consideredspecies un- serves in mate attraction as well as in identi- der the biologicalspecies concept. January2001] NewFlycatcher from Cerrado Region 75

Finally, preliminary molecular analyses of responded to suiriri vocalizations,whereas tissue samples indicates that S. islerorumis most pairs showedno response.More compre- stronglydifferentiated genetically from both S. hensiveinventories of tape recorded vocaliza- s. affinisand S.s. bahiae(J. Batespers. comm.). tionsof nominatebirds are neededfor a proper Those results will be presentedelsewhere as vocal analysis. Also needed are field studies part of an overview of genetic diversity in from the contactzone between suiriri and affin- Suiriri. is, as well as more extensiveplayback experi- Basedon vocal similarities,S.s. affinisand S. mentsfocused on this pair of taxa. Molecular s. bahiaeclearly are closelyrelated. The latter analyses in progress (J. Bates et al. unpubl. formresembles the formerin havinga longbill data) also shouldclarify thoserelationships. and yellow belly, but differs in being shorter- Ecologyand conservation.--Suiririislerorum is winged, showing little back-rump contrast, a bird of the cerradoregion, a biogeographic and in lackingthe contrastingpale baseto the region that encompasses1.5 to 2 million km2 rectrices (Traylor 1982). However, our exami- centered on the Brazilian shield, with exten- nation of 61 specimensof affinissuggests that sions into eastern Bolivia and northeastern Par- the degreeof back-rumpcontrast and the ex- aguay (Eiten 1972). The term "cerrado" has tent of the contrastingpale baseof the rectrices been applied generallyto virtually all of the are individually variable characters,and ex- nonforesthabitats of thatregion, although geo- tremeexamples of affinismay at leastapproach graphicallyit variesgreatly in speciescompo- typical examplesof bahiae.Hayes (2001) also sition and physiognomy.Five major subtypes found significantplumage variability in speci- of cerradohave been recognizedby botanists mens of bahiae.All vocalizationsof bahiaeap- (Eiten 1972).Of those,we have found Suiriri is- pear to be homologousto thoseof affinis,and lerorumto occurin three: (1) cerradosensu stric- displaybehaviors of duettingpairs are identi- to, woodland with closed scrub and scattered cal. Furthermore,in reciprocal tape playback trees;(2) campocerrado, more openscrub with experiments, all pairs of bahiaeresponded a few trees;and (3) camposujo, with stronglyto playbackof affinisvocalizations, as scatteredshrubs and few trees.In the Chapada did all pairs of affinisto vocalizationsof bahiae. dos Guimar•es and Cuiab• regions of Mato The relationshipof nominateS.s. suiriri to Grosso,Suiriri islerorumis locally fairly com- the yellow-bellied forms remains uncertain. mon in someareas but seeminglyabsent from Nominate birds are strongly differentiated others.Moderately closed, shrubby areas with morphologicallyand vocally from islerorum, a significant grass component and scattered and in tape playbackexperiments, the latter trees2-5 m tall (asexemplified by the first sev- speciesdid not respondto playbackof suiriri eral kilometersof the Agua Fria road nearCha- vocalizations. There is no reason to consider pada dos Guimar•es)have the highestdensi- them as anythingother than separatespecies. ties of S. islerorum.Our limited surveyshave In the absence of contact, nominate birds also averaged1 to 2 pairs per kilometerof road, but would appearto be sufficientlydistinct from af- more extensivedawn surveysusing systematic finis and bahiaein both morphologicaland vocal tapeplayback undoubtedly would yield higher charactersas to warrant separatespecies sta- densities.The speciesis somewhatless com- tus. However,the statusof intergradesor hy- mon along the Coxipo do Ouro road near Cu- brids from the contact zone in northern Para- iab•, where the is more wooded with guay and southwesternBrazil remains an pocketsof regularlyspaced trees (closer to cer- issue.Hayes (2001) found that presumedhy- rado sensustricto; Eiten 1972). Suiriri s. affinis bridsfrom the contactzone had plumagechar- occursalongside islerorum in both areas,but is actersintermediate between suiriri and affinis, more common in the more open woodlands and exhibitedsignificant plumage variability. and less common in the shrubbier,grassier The apparentscarcity of parentalphenotypes cerrado. from within the contactzone suggestedthat The areas in which we have found Suiriri is- the two forms were freely interbreeding,and lerorumare subjectto regular human pertur- therefore,conspecific (Hayes 2001). Further un- bation.The biggestdisturbance during the ten- certaintyis introducedby our playbackexper- ure of our work has been fire. Accidental or iments,in which somepairs of affinisand bahiae deliberatelyset fires have burned significant 76 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118 portionsof cerradoalong the Agua Fria roadin incompleteknowlege of distributionsmay re- three of the last five years, and have also af- sult in the seriousunderestimation of regional fectedthe nearbynational park. Localpopula- and local biodiversity and endemism.Such tions of both Suiriri islerorumand S.s. affinis miscalculationscould have seriousrepercus- havepersisted in heavilyburned areas, but the sionswhen important conservationpriorities repeatedburning of the same areas without are beingset. In an analysisof the intensityof sufficienttime for recoveryof the shrub com- collectingefforts in the cerradoregion, Silva munity likely will havea negativeeffect. Hous- (1995) found that roughly70% of the cerrado ing developmentshave spread into the cerrado regiondid not meetthe criteriaof havingbeen along the Agua Fria road, and continuedex- "minimally sampled"for birds. It is instructive pansionsoon will threatenthe type localityof that Suiriri islerorum occurs in at least seven of islerorum. the sitesthat met Silva'srequirements for being On a broaderscale, the dry forestand grass- satisfactorallysampled, including three re- land of central South America are con- gions (Goiania, Cuiab•, and ParqueNacional sideredboth amongthe mostendangered and Noel KempffMercado) with a high densityof unique biomesin the Neotropics(Stotz et al. minimallysampled localities, and yet, hases- 1996). The cerrado is especiallythreatened, caped detectionby field workers.If a cryptic with more than 75% of the 41 endemic birds speciesof bird could be missedin the most considered at risk, and more than 45% consid- heavily sampled30% of the cerradoregion, ered eitherthreatened or endangered(Stotz et then the potentialfor overlookedbiodiversity al. 1996).The primary threatderives from the in the remainingunder-sampled areas must be wholesale mechanized conversion of native great. Our recognitionof Suiriri islerorumas a cerradoand campocommunities to large-scale crypticspecies was initially basedon an appre- agriculture,particularly soybeans.The dry- ciation of vocal differences within Suiriri over seasonpractice of annuallyburning open coun- a broadgeographic range. It is criticalthat field try habitatsfor small-scaleagriculture, deeply personnelconducting surveys of the cerrado ingrained among most rural populations,is a and other endangeredhabitats have accessto secondarythreat whose long term environmen- adequate tape recording equipment and be tal consequenceshave only recently raised proficientin the recognitionof bird voices,or widespread concern. more taxa inevitably will be undersampledor Before the true conservation status of Suiriri overlooked (Parker 1991). islerorum can be assessed, we first need a more ACKNOWLEDGMENTS completeunderstanding of its distribution.We expectthat it will eventuallyprove to be broad- We are particularly grateful to Dionisio Pimentel ly sympatricwith S.s. affinisthroughout much (MPEG), who skillfully handled the collectingand of the cerradoregion. At present,Suiriri islero- specimen-preppingduties on our collectingtrip. T. rum is known to occur in two large national Schulenberg was a constant source of advice parks,Chapada dos Guimaraes National Park, throughoutthis project, and alsoprovided assistance Mato Grosso,Brazil, and Noel KempffMercado in obtaining somekey references.J. Batesalso pro- vided much useful advice, and facilitated the iden- NationalPark, depto. Santa Cruz, Bolivia.Both tification of plants from our study areas. Special parks contain extensive areas of seemingly thanks must go to M. and E Isler, who generously suitablecerrado habitat, and shouldbe system- donatedtheir time and talentsto providingthe map atically surveyedto assessdensities of Suiriri and spectrograms(respectively) used in this paper islerorum and other cerrado endemics. Another K. Garrett (LACM) provided criticalassistance in co- protectedarea that shouldbe surveyedfor is- ordinatingspecimen loans from severalinstitutions. lerorumis the ReservaEco16gica Serra das Ar- We also thank J. V. Remsenand S. Cardiff (LSUMZ); aras in western Mato Grosso, Brazil. T. Schulenberg,D. Willard, and S. Hackett(FMNH); The need for adequate surveys of endan- R. K. Panza (CM); S. L. Olson and P. Angle (USNM); E C. Sibley(YPM); C. Cicero (MVZ); and R. Ridgely geredcerrado habitats has been pointed out by and D. Agro (ANSP) for the loan of specimensfrom others (Bateset al. 1992, Silva 1995, Parker and their respectiveinstitutions. A. Schaffnerassisted Willis 1997),but its importancecannot be over- with the statisticalanalysis of morphologicaldata. G. stated.Lack of understandingof specieslimits Budney,T. Bridgeford,A. Priori, and the restof the in widespread,polytypic groups, along with staff at the Library of Natural Sounds,Cornell Uni- January2001] NewFlycatcher,from Cerrado Region 77 versity were of great help in allowing accessto the los (Rhinocryptidae),with descriptionsof the LNS collectionof recordings.Thanks also to S.Hilty, Ecuadoriantaxa, including three new species. R. Ridgely,and M. Cohn-Haftfor sharingadditional Pages47-88 in Studiesin NeotropicalOrnithol- recordingswith us. D. Lane painted the attractive ogy Honoring Ted Parker (J. V. Remsen,Jr., Ed.). colorplate. B. Carlosprovided transportationon our OrnithologicalMonographs no. 48. collectingtrip. We thank E. Goodale,D. Kroodsma, KROODSMA,D. E. 1984.Songs of the T. Schulenberg,and an anonymousreviewer for their (Empidonaxalnorum) and (Em- helpful commentson earlier drafts of this manu- pidonaxtraillii) are innate.Auk 101:13-24. script.Finally, we gratefullyacknowledge V. Eman- KROODSMA,D. E. 1986.Design of songplayback ex- uel for providing us with travel opportunitiesthat periments. Auk 103:640-642. havemade much of our work possible. KROODSMA,D. E. 1989.Male EasternPhoebes (Say- ornis ,Tyrannidae, Passeriformes)fail to LITERATURE CITED imitate songs.Journal of ComparativePsychol- ogy 103:227-232. BATES,J. M., T. A. PARKERIII, A. E CAPPARELLA,AND KROODSMA, D. E., AND M. KONISHI. 1991. A subos- T. J. DAVIS. 1992. Observationson the campo, cine bird (EasternPhoebe, Sayornis phoebe) de- cerrado and forest avifaunasof easternDpto. velopsnormal songwithout auditoryfeedback. SantaCruz, Bolivia, including21 speciesnew to Behaviour 42:477-487. the country.Bulletin of the British Ornitholo- LANYON,W. E. 1978.Revision of the Myiarchusfly- gists' Club 112:86-98. catchers. Bulletin of the American Museum of EITEN,G. 1972.The cerradovegetation of Brazil. Bo- Natural History 161:429-627. tanical Review 38:201-341. MARR,T. G. 1981.Breeding and foragingecology of FARABAUGH,S. M. 1982. The ecologicaland social the Cactus Wren in a variable environment. significanceof duetting. Pages85-124 in Acous- Ph.D.dissertation, New MexicoState University, tic Communication in Birds (D. E. Kroodsma Las Cruces. and E. H. Miller, Eds.) vol. 2. Academic Press, New York. MCKITRICK,M. C., AND R. M. ZINK. 1988. Species conceptsin ornithology.Condor 90:1-14. HAYES,E E. 2001.Geographical variation, hybridiza- MEYERDE SCHAUENSEE, R. 1966. The Speciesof Birds tion and the leapfrog pattern of evolutionin the of South America. Livingston, Wynnewood, (Suiriri suiriri) complex of Pennsylvania. SouthAmerica. Auk 118:in press. PARKER,T. A. III. 1991.On the use of tape recorders HELLMAYR,C. E. 1927. Catalogue of birds of the in avifaunalsurveys. Auk 108:443-444. , part 5. Tyrannidae. Field Museum of PARKER, t. A. III, AND E. O. WILLIS. 1997. Notes on Natural History Zoological Series13. three tiny grasslandflycatchers, with comments ISLER,M. L. 1997.A sector-basedornithological geo- graphicinformation system for the Neotropics. on the disappearanceof SouthAmerican fire-di- Pages345-354 in Studiesin NeotropicalOrni- versifiedsavannas. Pages 549-555 in Studiesin thologyHonoring Ted Parker(J. V. Remsen,Jr., NeotropicalOrnithology Honoring Ted Parker Ed.). OrnithologicalMonographs no. 48. (J. V. Remsen,Jr., Ed.). OrnithologicalMono- ISLER,M. L., AND P. R. ISLER.1987. The Tanagers: graphsno. 48. Natural History, Distribution and Identification. PAYNTER,R. A., JR., AND M. A. TRAYLOR,JR. 1991. Smithsonian Institution Press, Washington, OrnithologicalGazetteer of Brazil. Museum of D.C. Comparative Zoology, Cambridge, Massachu- ISLER, M. L., P. R. ISLER, AND B. M. WHITNEY. 1997. setts. Biogeographyand systematicsof the Thamno- RIDGELY, R. S., AND G. TUDOR. 1994. The Birds of philus punctatus (Thamnophilidae) complex. SouthAmerica, vol. 2. Universityof TexasPress, Pages355-381 in Studies in Neotropical Orni- Austin. thology Honoring Ted Parker (J. V. Remsen,Jr., SHORT,L. L. 1975. A zoogeographicanalysis of the Ed.). OrnithologicalMonographs no. 48. South American Chaco avifauna. Bulletin of the ISLER, M. L., P. R. ISLER, AND B. M. WHITNEY. 1998. AmericanMuseum of Natural History 154:163- Useof vocalizationsto establishspecies limits in 352. antbirds(Passeriformes: Thamnophilidae). Auk SIBLEY,C. G., AND B. L. MONROE, JR. 1990. Distri- 115:577-590. bution and Taxonomyof Birds of the World. Yale ISLER, M. L., P. R. ISLER, AND B. M. WHITNEY. 1999. University Press,New Haven, Connecticut. Specieslimits in antbirds (Passeriformes:Tham- SICK,H. 1993. Birds in Brazil: A Natural History. nophilidae): The Myrmotherulasurinamensis Princeton University Press, Princeton, New complex.Auk 116:83-96. Jersey. KRABBE,N., AND T. S. SCHULENBERG.1997. Species SILVA,J. M. C. 1995. Avian inventory of the cerrado limits and naturalhistory of Scytalopustapacu- region,South America: Implications for biolog- 78 ZIMMER,WHITTAKER, AND OREN [Auk, Vol. 118

ical conservation. Bird Conservation Interna- WALSBERG,G. E. 1977.Ecology and energeticsof con- tional. 5:291-304. trasting social systems in Phainopeplanitens SMITHE, F. B. 1975. Naturalist's Color Guide. Ameri- (Aves: Ptilognatidae).University of California can Museum of Natural History, New York. Publicationsin Zoology, no. 108. STOTZ,D. F., J. w. FITZPATRICK,t. A. PARKERIII, AND ZIMMER,J. T. 1955.Further notes on tyrant flycatch- D. K. MOSKOVITS.1996. Neotropical Birds: Ecol- ers (Tyrannidae).American Museum Novitates, ogy and Conservation.University of Chicago no. 1749. Press,Chicago. ZIMMER,K. J. 1993.Spatial and temporal variation in TRAYLOR,M. A., JR.1979. Tyrannidae. Pages 1-229 in the breeding and foraging ecologyof Black- Checklist of the Birds of the World, vol. 8 (M. A. throated Sparrows. Ph.D. dissertation,New Traylor, Jr., Ed.). Museum of ComparativeZo- Mexico StateUniversity, Las Cruces. ology,Cambridge, Massachusetts. TRAYLOR,M. A., JR.1982. Notes on tyrant flycatchers (Aves:Tyrannidae). Fieldiana Zoology 13:1-22. Associate Editor: D. Kroodsma