The Auk 118(2):457-471, 2001

GEOGRAPHIC VARIATION, HYBRIDIZATION, AND THE LEAPFROG PATTERN OF EVOLUTION IN THE SUIRIRI FLYCATCHER (SUIRIRI SUIRIRI) COMPLEX

FLOYD E. HAYES 1 Departmentof Biology,Caribbean Union College, P.O. Box175, Portof Spain,Trinidad and Tobago

ABSTRACT.--TheSuiriri Flycatcher(Suiriri suiriri) of SouthAmerica is representedby three distinctforms occurring in parapatry:(1) S.s.suiriri to the southwestin the Chaco/Pampas; (2) S.s. affinislocated centrally in the /southernAmazonia; and (3) S.s. bahiaeto the northeastin the Caatinga.On the basisof an analysisof 366 specimensof S.suiriri, I found meager evidencefor long-distancemigration, little support for Bergmann'srule, and more supportfor Gloger'srule. I postulatethat an ancestralpopulation of S.suiriri split into three isolatedpopulations, with the centralpopulation differentiating most rapidly into affinis, thusexplaining the leapfrogpattern of greatersimilarity between peripheral suiriri and bah- iae.After secondarycontact, affinis freely hybridizedwith nominatesuiriri in a zone to the southwest,where specimensdemonstrate morphometric intermediacy and increased plumage variability; the rarity of parental phenotypeswithin the hybrid zone suggeststhat the two forms are conspecificaccording to the biologicalspecies concept. In the northeast, affinismay havehybridized with a remnantpopulation of suiriri, possiblyrepresented by poorly known bahiae.The intermediatesize and increasedplumage variability of bahiaere- semblethat of suiriri x affinishybrids, supporting a hypothesisof hybrid origin for bahiae, but alternative hypothesescannot be ruled out. This hypothesisof differentiationis sup- ported by the concordantpatterns of disjunctionamong severalpairs of sistertaxa of speciesthat occur in the Chacoand Caatinga,with no interveningpopulations in the Cer- rado, implying a sharedhistorical process of vicarlance.Genetic and behavioralstudies are neededto elucidatefurther the statusand history of differentiationwithin S.suiriri. Received 31 January2000, accepted13 December2000.

EVOLUTIONARYBIOLOGISTS have long recog- terbreedingbetween populations in secondary nized the importance of documenting geo- contact" (Sibley and Short 1964:148),is a ge- graphicvariation in organisms(e.g. Mayr 1963, netic phenomenonthat is widespreadin 1970; Gould and Johnston1972, Zink and Rem- (Grant and Grant 1992) and can be inferred sen 1986). The data derived from such studies phenotypically.Schueler and Rising (1976:284) are crucial for testing hypothesesof phyloge- provided evidence that "an increase in vari- netic relationships and speciation processes ability and intermediacy in concert constitute among sister taxa. Analyses of the nature of the only pheneticevidence for hybridization." contactzones between sistertaxa are important The Suiriri Flycatcher(Suiriri suiriri), a me- for evaluatingthe extent of gene flow and the dium-sized speciesof the avian family Tyran- developmentof isolatingmechanisms between nidae, is widely distributed to the east of the differentiating populations (e.g. Remington Andes in (Short 1975, Ridgely 1968, Short 1969, Woodruff 1973, Schueler and and Tudor 1994). As with many Neotropical Rising 1976, Moore 1977, Barton and Hewitt birds, little is known about its natural history 1985a, b, Harrison 1990). Where zones of phe- (e.g. Sick 1984,Belton 1985,Ridgely and Tudor notypic intermediacy occur between sister 1994, Zimmer et al. 2001). It is an uncommon taxa, it is particularly important to distinguish residentof relatively dry, openforest and forest betweenprimary intergradationand secondary edges,usually occurringalone or in pairs. As intergradation.Hybridization, defined as "in- with most flycatchers,the sexesare monochro- matic. Although altitudinal migration appar- • Presentaddress: Unit of Zoology,Department of ently occurs(Chesser 1997), it is not known to Life Sciences,University of the West Indies, St. Au- be a long-distancelatitudinal migrant (Zimmer gustine,Trinidad and TobagoE-mail: floyd_hayes@ 1955). The speciesis representedby three dis- hotmail.com tinct races: (1) nominate S.s. suiriri of southern

457 458 FLOYDE. HAYES [Auk, Vol. 118

Bolivia, southern , , , Campanario,and reported them to be as de- and northernArgentina; (2) S.s.affinis of south- scribedby Zimmer (1955). Traylor (1979,1982) ern Surinam, central and eastern Brazil, and concurredwith Zimmer (1955) in recognizing eastern ; and (3) S.s. bahiaein a small suiriri, affinis, and bahiaeas three races of a area of the Caatingain easternBrazil. Thus far polytypic . Nevertheless,subsequent there have been no attempts at rigorously ex- authorities differed in their taxonomic treat- amining morphologicaland geographicvaria- ment of S. suiriri. Remsenand Ridgely (1980) tion within the species. and Sibley and Monroe (1990) recognizedtwo The purposesof this studyare to evaluatethe species.Remsen and Traylor (1983,1989), Rid- relationshipsbetween the three formsof S.sui- gely and Tudor (1994), and Hayes(1995) rec- riri by examining (1) geographicvariaton of ognized a single species. morphometricand plumagecharacters, (2) sex- The southernrace, nominate suiriri, has up- ual dimorphism,(3) the characterand extentof perpartsranging from dark gray to dark olive, hybridization between forms, and (4) the a whitish belly, a dark, pale-tippedtail, and a strengthof correlationbetween mensural and relatively short and black bill. The northern environmental variables. The results of these race, affinis,differs markedly from suiriri by analysesare further used to (1) evaluatequan- having a bright yellowbelly, paler gray or olive titatively Bergmann'srule, which predictsthat upperparts contrastingwith a whitish or yel- body size is inverselycorrelated with temper- lowish rump extending onto the rectrices,a ature and humidity (James 1970, Zink and longerbill that is often partially pale,and over- Remsen 1986); (2) evaluate qualitatively Glo- all larger size. The two racesare illustrated in ger's rule, which predicts that populationsin Ridgely and Tudor (1994:plate 31). Unfortu- more humid areasare moreheavily pigmented nately the behavior and ecologyof those two than those in drier areas (Zink and Remsen raceshave not beenstudied in detail. Although 1986); and (3) infer the evolutionaryhistory of Sick (1984) described their songs as being the three forms. markedly different and suggestedthat they were separatespecies, Zimmer et al. (2001)pro- TAXONOMIC BACKGROUND vided evidence that their vocalizations were similar and that Sick'sdescription of affiniswas In an early study of the taxonomicstatus of based on observationsof an unrecognized S.suiriri, Cory and Hellmayr (1927)recognized cryptic species(see below). The poorly known nominatesuiriri and affinisas distinct species, northeastern race, bahiae, restricted to the Caa- and consideredbahiae as a subspeciesof affinis. tinga of Brazil, has grayishupperparts includ- The two "species"appeared to be allopatric ing a dark rump (as in suiriri), a belly coloral- until Laubmann(1940) reported five presumed legedlyvarying from whitish to bright yellow, hybrid specimensof suiriri x affinisfrom De- and is intermediate in size between suiriri and partamento (hereafter Dpto.) Concepci6n, affinis.Its vocalizationsare apparently similar northeastern Paraguay, and two specimens to those of suiriri and affinis (Zimmer et al. from the same area that appearedto represent 2001). typical affinis.Zimmer (1955)subsequently ex- Among a series of affinisspecimens at the amined a large seriesof suiriri at the American AMNH, Zimmer (1955) noted that five were re- Museum of Natural History (AMNH) and re- markably short-billedwith an unusuallywide, ported 17 suiriri x affinis hybrids from the pale terminal band on the rectrices.Enigmati- same area in northeasternParaguay, plus an- cally, thosespecimens were obtainedfrom lo- other from Campanario,Mato Grossodo Sul, calities in central Brazil where typical affinis Brazil. Because of the variable and intermediate were also taken. Traylor (1982) further noted characteristicsof the populationin the zone of that those five specimensand a sixth in the intergradation,Zimmer (1955) consideredthe Field Museum of Natural History (FMNH), ini- two forms conspecific.Short (1975), however, tially describedby Hellmayr (1929),possessed failed to find the AMNH hybrid specimensand distinctlybroader central rectrices lacking the recognizedsuiriri and affinisas distinct species. pale edgesof bothsuiriri and affinis, and a more Traylor (1982) relocated the AMNH hybrid gradualtransition from the pale-graycrown to specimensfrom northeasternParaguay and the pale-oliveback than in typicalaffinis. Tray- April 2001] Evolutionin theSuiriri Flycatcher 459

lor (1982) and Ridgely and Tudor (1994) spec- Morphometricvariation.--Using digital calipers, I ulated that the short-billed specimensrepre- measured(nearest 0.1 mm) the following mensural sentedan undescribedsibling species. Zimmer variables:bill length (BL), from nostril to tip of max- et al. (2001) provided morphological,vocal, ilia; wing chord length (WL), from bend of folded and behavioral evidence that the short-billed wing to tip of longest primary feather; and tarsus length (TS), from junction of tibiotarsusand tarso- form wassufficiently distinct from otherSuiriri metatarsusto distal junction of hind toe and tarso- taxa under any of the widely acceptedspecies metatarsus.Tail length (TL), from baseof tail to tip concepts(McKitrick and Zink 1988) to warrant of longestrectrix, was measured(nearest millimeter) recognitionof a new species,the ChapadaFly- with a blunt wooden ruler to minimize damage to catcher (S. islerorum). the specimens.I did not attempt to evaluate mea- Although Zimmer (1955) gave the range of surement error, which undoubtedly contributed to measurementsfor bill length(from base)in sui- the variation in the data. riri and affinis(including short-billed affinis), no I used those data to compute four dimensionless measurements of bahiae and the intermediate (scale-free) shape variables. These included bill length/wing length (BL/WL), tarsus length/wing populationshave been published.Because of length (TS/WL), and tail length/wing length (TL/ the larger size of northernaffinis, Short (1975) WL). Those variables were used to evaluate relative consideredthe genus to be an exception to shapes,but becauseratios are nonlinear,often intro- Bergmann'srule. However, there have been no ducing spurious correlations(Atchley et al. 1976), quantitative analyses of morphometric and they were not subjectedto statisticalanalyses. plumage variation within the species. For eachspecimen, I alsorecorded the sex,locality, date, and mass (MS; nearest 0.1 g) when available, from the attachedspecimen labels. The sex was not METHODS identified in 15 specimensof suiriri, 6 of affinis,and 2 of suiriri x affinis;none of those specimenswas I examinedand measured366 study skinsof adult usedin the morphometricanalyses described below. S. suiriri (specimenswith evidenceof juvenal plum- Two-sample (Student's)t-tests (Sokal and Rohlf age-speckled feathers on upperparts--were ex- 1981) were used to comparethe means of eachmen- cluded; Wetmore 1926) in American museums (see suralvariable between the sexes for suirir•,affinis, Acknowledgments).These included 298 specimens and Paraguayansuiriri x affinishybrids. One-way of suiriri, 44 specimensof affinis,4 specimensof bah- analysesof variance (Sokal and Rohlf 1981) were iae,and 20 specimensof presumedhybrid suiriri x used to comparethe means of each mensural vari- affinis(all with intermediate and variablecharacters, and taken from the zone of contact between the two able between suiriri, affinis,and the Paraguayansui- riri x affinishybrids; a posteriorimultiple comparison parentalphenotypes). Additional data on bahiaewere tests were computed using t-tests (t statistic)with obtained from Brazilian museums (E. Willis pers. Bonferroniadjustments of the alpha level (Lentner comm. and published literature) and from field ob- and Bishop 1986). I did not attempt to compare sta- servationsby Zimmer et al. (2001) and B. Whitney tistically variablesfor specimensof bahiaeand hy- (pers.comm.). brids from Bolivia and Brazil due to the small sample Geographicdistribution.--The locality for eachspec- sizes. imen was taken from the specimenlabel and its lon- A principal components(PC) analysis(Sokal and gitude, latitude, and elevationwere obtained(when Rohlf 1981)was computedusing the correlationma- available) from ornithological gazetteers for each trix of raw data for the four mensural variables (BL, country (Paynter1985, 1989, 1992;Paynter and Tray- WL, TS, TL) to produce individual PC loadingsfor lor 1991) or from the specimenlabel. Each locality all 366 specimensof S. suiriri and 11 specimensof S. was plotted on a map. isleforum.PC1 and PC2, loosely interpreted as size Seasonaldistribution.--Evidence for long-distance and shapeaxes (e.g. Rising and Somers1989), were migration,particularly for the southernrace suiriri, plotted to comparemultivariate trends among Suiriri would complicatean analysisof geographicvaria- taxa. All statistical analyses were computed with tion. To examine the seasonal distribution of S. sui- Statistix 3.1 software (Anonymous 1990). riri, the locality for each specimen (as described Plumagevariation.--I used a simple"hybrid index" above) and date of collection were taken from the to scorethree plumage charactersfor all specimens specimen label. A bivariate plot of latitude and of S. suiriri. I initially used a five- scorehybrid index, month of collectionwas made for all specimens.An but found it easier to place specimensinto a three- absenceof specimensfrom the southernlatitudes scorehybrid index. Although somecriticize the use and a concentration of specimensat northern lati- of a hybrid index as crude and subjective,indepen- tudes during the austral winter (June-August) dent studiesobtain essentiallyidentical results(Cor- would provide evidencefor long-distancemigration. bin and Barrowclough 1977). Back coloration was 460 FLOYDE. HAYES [Auk, Vol. 118

\ o suiriri affinis _ ß affinis white-belliedbahiae o bahiae ' yellow-belliedbahiae . hybrds' ß . Paraguayhybrids Brazilhybrid Boliviahybrids om 40 ß ß •white-bellied ß I : e * * .

o o oO ß o oOo 20 • • o • o • • o o 0 • o

30 o o • o o o o o o o o o o o o

40 _ o J FMAM J J ASOND MONTH

FIG. 2. Seasonal variation in latitudinal distribu- tion of Suiriri suiriri specimens,including S.s. suiriri FIG. 1. Distribution of Suiriri suiriri specimens, x S.s. affinishybrids, examined in this study. including S.s. suiriri x S.s. affinishybrids, examined in this study. For S. suiriri bahiae,localities indicated with numeral 1 refer to specimensexamined during Chaco and Pampas,affinis in the Cerrado and this study; localitiesindicated with numeral 2 refer southernAmazonia, and bahiaein the Caatinga to published specimen records and recent sight records. (Fig. 1). The rangesof suiriri and affinismeet between the Chaco and Cerrado in eastern Bo- livia, northeasternParaguay and southwestern scoredas 0 if dark grayish or olive as in typical sui- Brazil, where hybridization apparently occurs riri, 2 if light grayishor olive as in typical affinis,and but is well documentedonly in Paraguay.The 1 if intermediatebetween the two taxa. Rump color ranges of affinisand bahiaemeet betweenthe was scoredas 0 if dark as in typical suiriri, 2 if light Cerrado and Caatinga, where hybridization as in typical affinis,and 1 if intermediatebetween the two taxa. Belly color was scoredas 0 if whitish as in may occur,but is poorly documented. typical suiriri, 1 if light yellow,and 2 if bright yellow Seasonaldistribution.--Although short-dis- as in typical affinis.The three scoreswere summedas tancealtitudinal migration apparently occurs in a measureof plumage variability (0 for typical sui- nominate suiriri, at least in Bolivia (Chesser riri, 6 for typical affinis,and 1-5 for intermediatebe- 1997),there is little evidencefor long-distancelat- tween the two taxa). itudinal migration (Fig. 2). Specimenrecords Geographicvariation.--To assess geographic varia- from Argentinian and Uruguayanmuseums in- tion in morphologicaltraits, I computedPearson cor- dicatethat the southernmostpopulations of nom- relation coefficients(r; Sokal and Rohlf 1981)for each inate suiriri remain presentthroughout the year mensuralvariable with latitude, longitude,and ele- vation. This was done for each sex of all S. suiriri (L. Josephunpubl. data). combined and separately for suiriri and affinis.Be- Morphometricvariation.--In nominate suiriri, causetemperature is correlatedwith both latitude sexual dimorphism occurred in all size vari- and elevation,a significantlypositive correlation be- ables,with malesaveraging significantly larger tween mensural variables and either latitude or ele- (Table1). In the considerablysmaller sample of vation would provide support for Bergmann'srule. affinisspecimens, females also averagedsmall- er for eachmeasure of body size (exceptmass, RESULTS for which there was insufficient data), but did not differ significantlyfor BL (Table1). In both Geographicdistribution.--The three racesof S. suiriri and affinis, TL/WL averaged slightly suiriri are parapatrically distributed, with the higher for femalesthan for males, suggesting range of eachspecies roughly coincidingwith that females had relatively shorter WL com- well known biogeographicregions (e.g. Short pared to TL, but no other shapevariables dif- 1975):nominate suiriri occursprimarily in the fered appreciablybetween sexes for eithertax- April 2001] Evolutionin theSuiriri Flycatcher 461 462 FLOYDE. HAYES [Auk, Vol. 118

TABLE2. Descriptivemeasures of size and shapevariables (see Table i for definitions)for threehybrid pop- ulationsof Suiriri s. suiriri x affinis.Statistical comparisons are given betweenthe Paraguayhybrids and suiriri and affinisfor eachsex. No significantsexual dimorphism was found in the Paraguayhybrid pop- ulation.

Paraguay Bolivia Brazil Variable • SD Min,Max n • SD Min, Max n R n BL c• 8.77 •,c 0.24 8.4,9.1 9 7.80 0.28 7.6,8.0 2 9.40 1 • 8.82 • 0.32 8.4,9.2 5 ------0 -- 0 WL c• 80.01 •,c 0.32 76.0,84.7 9 76.15 2.19 74.6,77.7 2 85.10 1 • 77.95 • 0.18 74.7,79.8 6 ------0 -- 0 TS c• 20.21b 0.72 18.9,21.2 9 18.2 0.14 18.1,18.3 2 19.00 1 • 19.83 b 0.65 19.2,21.0 6 ------0 -- 0 TL c• 70.78 d 3.07 65.0,75.0 9 68.0 1.41 67.0,69.0 2 78.00 1 • 72.83 b 4.79 69.0,79.0 6 ------0 -- 0 BL/WL c• 0.11 0.004 0.10,0.12 9 0.10 0.00 0.10,0.10 2 0.11 1 • 0.11 0.004 0.11,0.12 5 ------0 -- 0 TS/WL c• 0.25 0.01 0.24,0.26 9 0.24 0.01 0.24,0.24 2 0.22 1 • 0.25 0.01 0.25,0.27 6 ------0 -- 0 TL/WL c• 0.89 0.05 0.81,0.96 9 0.89 0.04 0.86,0.92 2 0.92 1 • 0.93 0.06 0.88,1.02 6 ------0 -- 0 PC1 C• -1.95 • 0.95 -3.30,-0.66 9 0.59 0.39 0.31,0.87 2 -3.33 1 9 -1.81 • 0.82 -2.96,-0.94 5 ------0 -- 0 PC2 c• -0.60 0.54 -1.67,0.14 9 0.47 0.47 0.14,0.80 2 1.32 1 9 0.05 1.13 -1.07,1.44 5 ------0 -- 0

Differs from suiriri of same sex, P < 0.001. Differs from suiriri of same sex, P < 0.01. Differs from affinisof same sex,P < 0.001. Differs from affinisof same sex,P < 0.05. on (Table 1). Females of bahiae likewise affinisand approachedthe mean for suiriri (Ta- averagedsmaller than males for each size var- ble 2). The femalehybrids differed significantly iable (exceptmass, for whichthere was no data) from suiriri,but not from affinisfor all four size and had a largervalue for TL/WL (Table1), but variables;WL was intermediate,though closer the data were insufficient for statistical to affinis,whereas BL, TS, and TL averagedthe comparisons. same or even higher than affinis(Table 2). For Both sexesof affinis averaged significantly the three shape variables, males appeared to higher for each size variable (exceptmass, for differ from suiriri for TS/WL and TL/WL, which there was insufficient data) than suiriri whereasfemales appeared to differ from suiriri (Table 1). The higher TS/WL and TL/WL for only for BL/WL; neither sex differed from af- suiriri suggestedthat WL is relatively shorter finis for any shape variable. The two Bolivian for suiriri than for affinis.In the small sampleof hybrids resembledsuiriri in size variablesand bahiae,the means of BL were similar to those of affinisin shape variables;the Brazilian hybrid affinis, WL and TS averaged slightly longer resembledaffinis in BL, WL, and TL, and suiriri than suiriri, TL averagedslightly shorterthan in TS and TS/WL. suiriri, and TL/WL was intermediate (Table 1). A bivariateplot (Fig. 3) of PC1 (59.7%of var- Hybrid suiriri x affinis from Paraguay iance) and PC2 (19.4% of variance) separated showedno significantsexual dimorphismfor most suiriri and affinis specimens.The bahiae any size or shapevariable, but the samplesizes specimensappeared to be intermediate be- were small (Table 2). Although the hybrids tween suiriri and affinis.The male suiriri x af- tended to be intermediate in size between sui- finis hybrids from Paraguay and Brazil ap- riri and affinis,the nature of variationwas com- peared intermediatebetween suiriri and affinis, plex. For males, BL and WL were intermediate but the two males from Bolivia clustered with- between suiriri and affinis,differing signifi- in the range of variation of suiriri. In contrast, cantly from both; TS averaged significantly the female suiriri x affinis,all of which were longer than suiriri and approachedthe mean from Paraguay,clustered within the range of for affinis;and TL was significantlyshorter than variation of affinis.The widely scatteredspeci- April 2001] Evolutionin theSuiriri Flycatcher 463

3 - (AMNH 499741), another specimenof suiriri • MALES 2 [] • Bolivia from Yuto, Jujuy, (Museumof Com- - Brazil•--_•o ¸ o•0•oo o/• parative Zoology; MCZ 262238), and in a fe- 1 o ø- •oø% oO/1øo o _ wh•te-belhedß . ¸ • '•0• 0 0 _• • o male specimenof affinisfrom Chapada, Mato ca 0 o Ooo Grosso,Brazil (AMNH 33101); a specimenof O oo unknown sex from Miranda, Mato Grosso do o_ -1 _ . Sul, Brazil (MCZ 154623),had a nearly-white o o - .o o •o belly within the normal range of variation of -2 ß o _ _• suiriri. None of thosespecimens showed other "o S.s.suiriri • S.s.suiriri xaffinis yellow-bellied -3 indications of hybridization. Wetmore (1924, u S.s. bahiae 1926) described an additional yellow-bellied -4 _eS. s.affinis •S. islerorum •o specimen (not examined by me) from Tucu- -6 -5 • -3 -2 -1 0 1 2 man, Argentina, as a new species(S. improvisa), PC1 but Cory and Hellmayr (1927:443)pointed out

- [] [] o FEMALES that otherspecimens from the locality(e.g. 2 [] o AMNH 499741 described above) varied from o o - # O Oo O the typical "white" of nominate suiriri to the 1 ' % • •.o øø yellow of improvisaand regardedthe type spec- _ ..-[] o imen of the latter as a mutant suiriri. ß ? •oCJ•c• o• The four bahiaespecimens examined resem- bled suiriri in back and rump color,but had Ooo highly variable belly coloration:white in two - white_bellied//• o specimensand yellow in the other two (Table3, -2 ß yellow:bellied Appendix).The two white-belliedbahiae spec- o imens,both taken from the samelocality, were -3 ' I ' I ' i ' i ' I ' i , • , i ' -5 -4 -3 -2 -1 0 1 2 3 4 indistinguishablefrom suiriri on the basis of PC1 plumage,but BL of both specimens(male, 9.5 mm; female, 9.0 mm) exceeded the maximum FiG.3. •fi•cipa] compo•e•[s a•a]ysis o• BL of suiriri (9.0 and 8.7 mm for males and fe- males, respectively;see Table 1). Furthermore, specimensexamined i• [his study. Zimmer (1955) pointed out that the outer web of the outerrectrix was dull asin affinisand bah- mens of S. islerorum tended to cluster more with iae,in contrastwith the lighter outerweb of sui- affinisthan suiriri. riri. Anotherwhite-bellied specimen at the Mu- Plumagevariation.--The total plumage score seu de Zoologia da Universidadede Sao Paulo was0 for all but two specimensof nominatesui- (MZUSP 8418), a male collectedat Pirapora, riri, and 6 for all but two specimensof affinis Minas Gerais,in August1912, was regardedas (Table3). Intermediatebelly color occurredin indistinguishablefrom suiriri and considereda a specimenof suiriri from Tucuman,Argentina migrant (Cory and Hellmayr 1927, Pinto and

TABLE3. Number of individualsfor eachplumage score for eachtaxon or populationof Suiriri suiriri.

Back Rump Belly Total score Taxa/population 0 I 2 0 I 2 0 I 2 0 I 2 3 4 5 6 S.s. suiriri 298 0 0 298 0 0 296 2 0 296 2 0 0 0 0 0 S.s. affinis 0 0 48 0 0 48 I I 46 0 0 0 0 I I 46 S.s. suiriri x affinis Bolivia 0 2 0 0 2 0 2 0 0 0 0 2 0 0 0 0 Paraguaycombined 0 16 1 17 0 0 3 5 9 0 3 4 9 1 0 0 San Luis de la Sierra 0 9 1 9 1 0 2 1 7 0 2 0 7 1 0 0 ZanjaMorotf 0 7 0 7 0 0 1 4 2 0 1 4 2 0 0 0 Brazil 0 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0 S. s. bahiae 4 0 0 4 0 0 2 1 1 2 1 1 0 0 0 0 464 FLOYDE. HAYES [Auk, Vol. 118

Camargo 1961, Ridgely and Tudor 1994).Zim- Twenty specimenstaken from the zone of mer (1955) and Pinto and Camargo (1961) contactbetween suiriri and affinisshowed in- pointed out that a female (MZUSP 8415) col- termediate and variable plumage characters. lected from the same locality was assignedto The two hybrids from easternBolivia were in- affinis(identity reconfirmedby E. Willis, pers. termediatein back and rump color,but belly comm.). Zimmer (1955), who did not examine color resembled suiriri (Table 3, Appendix). the specimens,queried the identity of the The 17 hybrids from two localities(in north- white-bellied bird as suiriri and suggestedit easternParaguay) were consistentlydark-rum- was a white-bellied affinis;however, the dark ped (as in suiriri), intermediatein back color- rump and back describedby Pinto and Camar- ation, and highly variable in belly coloration go (1961)are more characteristicof bahiaethan (Table3, Appendix). The distributionof total affinis.The specimenwas recentlyreexamined plumagescores for the two Paraguayanlocali- by E. Willis (pers.comm.), who reportedBL to ties did not differ significantly(X 2 = 8.02,df = be 9.0 mm, the maximum for male suiriri (Table 4, P = 0.09). Traylor (1982) reported two ad- 1); other measurements(WL, 81.5 mm; TS, 20.4 ditional hybrid specimens(AMNH) from Be- mm; TL, 74.3 mm) were within the range of 16n,Dpto. Concepci6n,which is south of the variation of male suiriri and affinis. Further- other Paraguayanhybrid localities;however, I more, Willis describedthe specimen'sback as found the Be16nspecimens to be indistinguish- "rather olivebut with rusty or brownishwash, able from suiriri (Hayes 1995).The singleBra- ... more like affinisthan like the few suiriri in zilian hybrid specimenwas intermediate in the collection." This specimenappears to rep- backcoloration, but otherwiseresembled affinis resent a white-bellied bahiae. (Table3, Appendix). Other bahiaespecimens reported in the liter- Geographicvariation.--Significant geographic variation occurred for several mensural vari- ature (but not examined by me) were appar- ables.Among all S.suiriri specimens combined, ently yellow-bellied,including the type speci- latitudewas negatively correlated with BL,WL, men from "Bahia" in the Berlepschcollection TS, and TL, and positivelycorrelated with MS (Cory and Hellmayr 1927).However, E. Willis (but only one affinismale in sample)and PC1 (pers. comm.) reexaminedthe MZUSP speci- (Table4). Longitudewas negativelycorrelated mensassigned to bahiaeby Cory and Hellmayr with BL, WL, TS (males only), PC1, and PC2 (1927) and Pinto and Camargo (1961), and (males only), and positively correlatedwith found that all specimens,including two from PC2 (femalesonly; Table4). Elevationwas pos- Juazeiro, Bahia (09ø25'S, 40ø30'W), two from itively correlatedwith WL (femalesonly), TL, Coremas, Paraiba (07ø01'S, 37ø58'W), and two and PC2, and negativelycorrelated with WL from Ibipetuba(= Santa-Rita-de-Cassia),Bahia (males only, in contrast with females), MS (11ø00'S,44ø32'W), were all pale-rumped, ap- (malesonly), and PC1 (Table4). parently representingaffinis. I found that two Within nominate suiriri, latitude was posi- Los Angeles County Museum (LACM) speci- tively correlatedwith MS and negativelycor- mens assignedto bahiae,also from Ibipetuba, related with TL (femalesonly) and PC2 (fe- Bahia, were also pale-rumped, representing malesonly; Table4). Longitudewas positively affinis. correlatedwith TL and PC2 (malesonly), and All recent recordsof bahiaewere apparently negativelycorrelated with MS (malesonly) and of yellow-belliedbirds. Those include obser- PC1 (femalesonly; Table4). Elevationwas pos- vationsby R. Ridgely(pers. comm.), Zimmer et itively correlatedwith WL, TL and PC2, and al. (2001) near Lagoa Grande, Pernambuco, negativelycorrelated with MS (malesonly) and from 1996-1999,and B. Whitney (pers.comm.) PC1 (Table 4). at the followinglocalities: 20 km southof Car- Samplesizes were limited within affinis,es- inhanha, Minas Gerais (14ø17'S, 43ø47'W), on peciallyfor males(Table 4). Formales, elevation 11 November 1994 (one bird); 25 km east of was positivelycorrelated with TS; for females, Guanamb•, Bahia (14ø07'S,42ø37'W), on 12 No- latitudewas positivelycorrelated with WL and vember 1994 (four); and 15 km north of Petro- TL and negatively correlatedwith PC1, and lina, Pernambuco (09ø24'S,40ø30'W), on 21 No- longitude was positively correlatedwith TL vember 1994 (two). (Table 4). April 2001] Evolutionin theSuiriri Flycatcher 465

TABLE4. Geographiccorrelates (r) of body size and shapevariables (see Table 1 for definitions)in S.suiriri. Note that S. suiriri refers to all forms combined,including hybrids;sample sizesfor S.s. bahiaeand S.s. suiriri x affiniswere inadequatefor analysis.See Methods for definitionsof variables.

Latitude and longitude Elevation Variable Sex/taxon n Lat. Long. n Elev. BL c• S. suiriri 186 -0.47 a -0.59 a 183 -0.05 c•S. s. suiriri 158 -0.03 -0.01 155 0.06 c•S. s. affinis 15 -0.11 -0.38 14 -0.31 9S. su•rirl 150 - 0.59 • - 0.63 • 147 - 0.08 9S. s. suirlri 122 -0.02 -0.09 121 0.02 9S. s. affinis 22 -0.25 0.25 20 -0.04 WL c• S. 190 -0.42 • -0.45 • 187 -0.20 b c•S. s. suirlrl 161 - 0.05 0.11 158 0.44 a c•S. s. affinis 16 0.17 -0.10 15 0.05 9S. 151 - 0.46 a - 0.41 • 148 0.20 c 9S. s. suirm 122 -0.14 0.27 b 121 0.45 • 9S. s. affinis 22 0.60 b 0.27 20 0.38 TS c• S. sulrin 189 -0.17 c -0.31 • 186 0.00 c•S. s. suinrl 161 0.05 - 0.02 158 0.04 c•S. s. affinis 15 0.36 -0.08 14 0.65 c 9S. sulrirl 151 -0.16 c -0.11 148 0.07 9S. s. suir•r• 122 -0.04 0.17 121 0.13 9S. s. affinis 22 0.24 0.03 20 - 0.03 TL c• S. su•rir• 188 -0.18 c 0.00 186 0.42 • c•S. s. suinrl 160 -0.06 0.22 b 157 0.49 • c•S. s. affinis 15 0.11 -0.20 15 0.17 2S. su•rir• 149 -0.23 a -0.01 146 0.42 • 9S. s. suir•n 120 -0.25 b 0.23 c 119 0.53 • 9S. s. affinis 22 0.50 c 0.48 c 20 0.28 MS c• S. sulrin 21 0.46 c - 0.23 21 -0.60 b c•S. s. suirlrl 20 0.76 • - 0.47 c 20 - 0.72 • c•S. s. affinis 1 -- -- 0 -- 9S. sulrirl 14 0.75 b -0.45 14 -0.44 9S. s. suinn 14 0.75 b -0.45 14 -0.44 9S. s. affinis 0 -- -- 0 -- PC1 c• S. sulrin 183 0.40 • 0.43 a 181 -0.18 c c•S. s. suinn 157 0.03 -0.13 154 -0.42 • c•S. s. affims 13 -0.23 0.15 13 -0.37 9S. sulrir• 148 0.49 • 0.42 • 145 -0.19 c 9S. s. suirlrl 120 0.17 -0.24 b 119 -0.43 • 9S. s. affinls 22 -0.47 • -0.38 20 -0.22 PC2 c• S. su•rirl 183 0.08 0.36 • 181 0.39 • c•S. s. suirlrl 157 -0.07 0.20 c 154 0.41 * c•S. s. affinls 13 -0.31 -0.32 13 -0.45 9S. sulrirl 148 0.12 0.32 a 145 0.40 • 2S. s. suirlrl 120 -0.21 c 0.16 119 0.43 • 2S. s. affinis 22 0.35 0.28 20 0.27

0.001. 0.01. 0.05.

DISCUSSION known within the range of S. suiriri (e.g.Hoff- man 1975, Prohaska 1976), and an assessment This study documents considerablegeo- of their potential effectson morphologicalvar- graphic variation in morphometricand plum- iability is warranted (Zink and Remsen1986). age traits within S.suiriri. Althoughit remains Temperature is inversely correlated with unclear whether such variation is attributable bothlatitude and elevation.Because tropical af- to natural selection, environmental induction, finis is clearly larger than more temperatesui- or stochasticity,the geographicpatterns of var- riri, Short (1975) pointed out that S. suiriri vi- iability for temperatureand humidity are well olated Bergmann'srule. When all specimens 466 FLOYDE. HAYES [Auk, Vol. 118 are combined,the negativecorrelation of most coincidingwith the boundariesbetween major morphometric traits with latitude supports biogeographicregions (Chaco-Cerrado-Caatin- Short's (1975) conclusion.However, within the ga), which may be characterizedby relatively more southerlyrace suiriri, the positive corre- abrupt environmentalchanges. However, the lation of MS with latitude supportsBergmann's boundariesbetween those regions may alsorep- rule, though the negative correlation of MS resent locations of previous vicariance events with elevationis puzzling. Although WL is of- (e.g.Cracraft and Prum 1988).Furthermore, the ten used as an index of body size for studiesof morphological intermediacy and increased geographicvariation (Zink and Remsen1986), plumagevariability of specimensfrom the con- MS is often considered to be a better measure tact zone between suiriri and affinisin north- of body size (e.g.Rising and Somers1989, Free- easternParaguay provide convincingevidence man and Jackson1990). In the case of suiriri, for secondary intergradation (hybridization) WL was significantlycorrelated with elevation rather than primary intergradation(Schueler as noted in Bolivian specimensby Traylor and Rising 1976).The occurrenceof hybridiza- (1950), but not latitude. PC1, often used as a tion effectivelyfalsities the parapatric model multivariatemeasure of body size (Risingand and is consistentwith both the dispersaland vi- Somers 1989), was not correlated with latitude cariancemodels of allopatricdifferentiation. and was negativelycorrelated with elevation. Long-distancedispersal is unlikely to occur The effects of temperatureon morphological among relatively sedentarytaxa, including S. variation remain unclear, as it is with many suiriri, for which there is little evidenceof long- North American speciesof birds (Zink and distance migration. This suggests that the Remsen 1986). northeastern race bahiae, which resembles The peripheralraces suiriri and bahiaeoccur southernsuiriri more than affinison the basisof in drier regionsthan doesaffinis. Although af- plumage scores,was unlikely to have been es- finis is generally paler dorsally (primarily on tablished by long-distancemigrants of suiriri rump) than eithersuiriri or bahiae,the latterrac- from southern South America. es are paler on the underparts and grayer on Becausethe geographicrange of affinisoc- the upperparts, in general conformity with curs between morphologically similar suiriri Gloger'srule, which is well supportedamong and bahiae,S. suiriri provides yet another ex- North American birds (Zink and Remsen 1986). ample of the leapfrog pattern of geographic Severalhistorical processesmay be postulat- variation (Remsen 1984) in which two similar ed to explain the geographicpattern of differ- populationsare separatedfrom eachother by entiationdemonstrated by S. suiriri. First, dif- one or more different, interveningpopulations ferentiation may have evolved in parapatry, of the same species.Remsen (1984) suggested with current contactzones between adjacent severalhypotheses that may explainsuch a pat- races coincidingwith relatively abrupt envi- tern, including (1) convergentevolution in the ronmental changes(Endler 1982a, b). In that phenotypicallysimilar but geographicallysep- scenario,each contactzone would representa aratedtaxa; (2) more rapid, divergentevolution "step cline" of primary intergradationwhere in the central,intervening taxa; (3) centrifugal geneflow is reducedbetween contiguous pop- speciation;(4) ancientcorridors connecting the ulations that have never been isolated. Second, currently separatedbut phenotypicallysimilar differentiationmay have evolved in allopatry taxa; and (5) random phenotypic changesin followinglong-distance dispersal of "founder" geographicallyisolated populations of a taxon, individuals acrosspre-existing barriers. And with the changesoccurring first--by chance-- third, differentiation may have evolved in al- in the intervening population(s).The lack of lopatry as a contiguousancestral population concordancein the rangesof Andeanbird taxa becamefragmented by vicariance.For both the exhibiting the leapfrog pattern supports the secondand third hypotheses,the current con- fifth hypothesis,suggesting that phenotypic tact zoneswould representsecondary intergra- changesappeared randomly with respect to dation (i.e. hybridization) occurringwhere the geography,and were not necessarilyinduced expandingranges of adjacentpopulations met. by the environment(Remsen 1984). Can this The parapatric differentiationmodel is sup- hypothesisexplain the geographicpattern of ported by the positionof currentcontact zones differentiation in S. suiriri? April 2001] Evolutionin theSuiriri Flycatcher 467

Many speciesof South American birds are freely hybridized in the absenceof isolating representedby pairs of closely related taxa mechanisms,thus forming a hybrid swarm of with disjunct populations in the Chaco and individuals exhibiting intermediacy and in- Caatinga,with essentiallyno interveningpop- creased variability. Because the population ulation in the Cerrado (exceptperhaps in the from the hybrid zone in northeasternParaguay periphery;e.g. seedistributional maps of Short appears to be composedentirely of hybrids 1975, Ridgely and Tudor 1989, 1994). Essen- with few parentalphenotypes (22 hybrid spec- tially monotypic specieswith little differenti- imens vs. 2 purportedly typical affinis;Laub- ation between populations include the White- mann 1940, Zimmer 1955), little or no assorta- breasted Tinamou (Nothura boraquira) and tive mating occurred. Thus, the two forms Chotoy Spinetail (Schoeniophylaxphryganophi- appear to be conspecificwhen the criterion of la). Other specieswith more distinct racial dif- reproductiveisolation--a hallmark of the bio- ferencesinclude (racesgiven in Chaco/Caatin- logical species concept (e.g. Mayr 1970)--is ga sequence): the Blue-crowned Parakeet applied. (Aratingaacuticaudata acuticaudata /hemmorhous), In the Caatinga,hybridization may alsohave RufousCachalote (Pseudoseisura cristata unirufa / occurredwhen affinisand the isolatedremnant cristata),Stripe-backed Antbird (Myrmorchilus suiriri population cameinto contact,but the sit- strigilatussuspicax/strigilatus), White Monjita uation there remains less clear. Poorly known (Xolmisirupero irupero / nivea),Greater Wagtail- bahiaemay have shared a most recentcommon Tyrant (Stigmaturabudytoides inzonata/gracilis) ancestorwith (1) an isolated population of af- and Bay-winged Cowbird (Molothrusbadius finis (Fig. 4A), (2) a remnantpopulation of sui- badius/fringillarius).Other examples occur at riri (Fig. 4B), or (3) may instead constitutea hy- the specieslevel. The concordanceof this dis- brid populationof remnantsuiriri x affinis(Fig. junct distribution pattern among speciesbe- 4C). The proximity of affinis and the distant longing to unrelated taxonomicgroups implies current distribution of suiriri support the first a shared historical processof vicariance (e.g. hypothesis;if that is true, the shared traits of Cracraft and Prum 1988). Although plausible, bahiaeand suiriri would have evolvedindepen- long-distancedispersal of a taxon acrossun- dently. The secondhypothesis is supportedby favorable Cerrado habitat from one side to the the greaterresemblance between bahiae and sui- other is unlikely to occur in relatively seden- riri and the concordantpattern of disjunctCha- tary taxa not known for long-distancemigra- co-Caatinga taxa. The intermediatesize and in- tion (e.g. White-breasted Tinamou). Rather, a creasedvariability in plumageof the few bahiae relatively recent vicariance event or series of specimensavailable, combined with descrip- eventsprobably occurred,which subdivided a tions of seeminglyintermediate affinis x bahiae contiguous ancestral population into isolated specimensin the region, support the third hy- Chaco and Caatinga units. Suchan event likely pothesisof hybrid origin for bahiae.Indeed, the occurred with S. suiriri. The most plausible bahiaespecimens exhibit a remarkablesimilar- agent for vicariancewas the formation of hu- ity with the suiriri x affinishybrids from Par- mid forestbarriers during previousinterglacial aguay; that was first noted by Zimmer (1955: periods (e.g. Haffer 1985, 1987;Willis 1992). 19), who statedthat "some of the specimensof In conformitywith the vicariancemodel of the intermediate population of northeastern differentiation,I postulate the following sce- Paraguayalmost meet the requirements[of bah- nario for differentiationin S.suiriri. Initially, an iae], but show, in most cases, too much of the ancestralpopulation of S. suiriri resembling yellowish basal area of the rectricesto agree nominatesuiriri occupiedthe entire arid diag- fully." If bahiaeis indeed a hybrid taxon,as the onal from the Chacoto the Caatinga.After one morphological data suggest, it should not be or more vicariance events isolated Chaco, Cer- considereda valid .Furthermore, it rado, and Caatinga populations,the Cerrado would appear to represent a disjunct hybrid population differentiated more rapidly than population situated >1,000 km from the near- the two peripheral isolates,eventually acquir- est current population of suiriri, a presumed ing the traits of affinis.As the southernpopu- parental phenotype, unless the white-bellied lation of nominate suiriri spread northward specimenstentatively assignedto bahiaeactu- and affinis spread southward, they met and ally representa remnant suiriri populationbe- 468 F]•oY]::)E. HAYES [Auk, Vol. 118

S. islerorum S. s. suiriri S. s. affinis S. g. bahiae S. islerorum S. s. affinis S. s. suiriri S. s. bahiae

S. islerorum S. s. affinis S. s. bahiae S. s. suirid S. islerorum S. s. affinis S. s. suiriri S. s. bahiae

FIG. 4. Hypotheticalrelationships among Suiriri taxa illustratedby four cladograms.Note that S. suiriri is monophyleticin cladogramsA-C and paraphyleticin cladogramD.

coming extinct through competitionor intro- gelesCounty Museum; R. Browning and G. Graves gression from affinis (e.g. Rhymer and at the National Museum of Natural History; and P. Simberloff 1996). If indeed the white-bellied Chu, J. Hinshaw, R. Payne,and R. Storerat the Uni- bahiaespecimens represent a remnant suiriri versity of Michigan Museum of'Zoology.I thank the followingindividuals for the loan of specimensfrom population,then bahiaemay be a distincttaxon their institutions:E Gill of the Academy of Natural lessvariable than it appears,represented exclu- Sciencesof Philadelphia;K. Parkesof the Carnegie sivelyby yellow-belliedbirds. If that hypothe- Museum;and R. Paynterof the Museum of Compar- sisis correct,bahiae may haveshared a mostre- ative Zoology. For supplying literature, specimen cent common ancestor with either an isolated data, recentsight recordsor insightfulconversation, populationof affinisor remnantsuiriri. Finally, I thank M. Berres,N. Collie, T. Goodwin, L. Joseph, thesehypotheses are basedon the assumption R. Prum, J. da Silva, B. Whitney, and E. Willis. of monophylywithin S. suiriri (Fig. 4A-C). The superficialsimilarity betweenS. islerorumand LITERATURE CITED S.s.affinis suggests that they may be sistertaxa, in which caseS. suiriri would be paraphyletic ANONYMOUS.1990. Statistix Manual. AnalyticalSoft- ware, St. Paul, Minnesota. (Fig. 4D; McKitrick and Zink 1988 discussthe ATCHLEY, W. R., C. t. GASKINS, AND D. ANDERSON. implicationsfor speciesconcepts), but S. isle- 1976.Statistical properties of ratios.I. Empirical rorum appears behaviorally and vocally dis- results.Systematic Zoology 25:137-148. tinct (Zimmer et al. 2001). Genetic studies are BARTON,N.H., AND G. H. HEWITT. 1985a. Hybrid clearlyneeded to elucidatefurther the relation- zones and speciation.Pages 109-145 in Essays shipsand history of differentiationamong the on Evolution and Speciationin Honor of M. J.D. forms of S. suiriri and S. islerorum. White (W. R. Atchley and D. S. Woodruff,Eds.). CambridgeUniversity Press,Cambridge, United ACKNOWLEDGMENTS Kingdom. BARTON,N.H., AND G. H. HEWITT.1985b. Analysis I thank J.Bates and D. Stotzfor reviewingthe man- of hybrid zones.Annual Review of Ecologyand uscript.This study was funded by grants from the Systematics16:113-148. American Museum of Natural History, Field Muse- BELTON,W. 1985. Birds of Rio Grande do Sul, Brazil, um of Natural History, Loma Linda University, and part 2. Formicariidaethrough Corvidae. Bulletin the University of the West Indies. Assistancewas of the American Museum of Natural History provided by C. Griffiths, M. LeCroy,L. Short,and E 180:1-241. Vuilleumier at AMNH; D. Maurer, T. Schulenberg, CHESSER,R. t. 1997. Patterns of seasonal and geo- and D. Willard at FMNH; K. Garrett at the Los An- graphical distribution of austral migrant fly- April 2001] Evolutionin theSuiriri Flycatcher 469

catchers(Tyrannidae) in Bolivia. Ornithological LAUBMANN,A. 1940. WissenschaftlicheErgebnisse Monographs48:171-204. der Deutschen Gran Chaco-Expedition. Die Vo- CORBIN, K. W., AND G. E BARROWCLOUGH. 1977. Re- gel von Paraguay,vol. 2. Streckerund Schroder, producibility of hybrid index scores.Condor 79: Stuttgart, Germany. 497-498. LENTNER,M., ANDT. BISHOP.1986. Experimental De- CORY,C. B., AND C. E. HELLMAYR.1927. Catalogue sign and Analysis. Valley Book Co., Blacksburg, of birds of the Americas, part 5. Field Museum Virginia. of Natural History Zoological Series13. MARINI,M. •., J.c. MOTTA-JUNIOR,L. A. S.VASCON- CRACRAFT,J., AND R. O. PRUM. 1988. Patterns and CELLOS, AND R. B. CAVALCANTI. 1997. Avian processesof diversification:Speciation and his- body massesfrom the Cerradoregion of central torical congruencein some Neotropical birds. Brazil. Ornitologfa Neotropical 8:93-99. Evolution 42:603-620. MAYR,E. 1963. Speciesand Evolution.Harvard ENDLER,J. A. 1982a.Problems in distinguishinghis- UniversityPress, Cambridge, Massachusetts. torical from ecologicalfactors in biogeography. MAYR, E. 1970. Populations, Species and Evolu- American Zoologist 22:441-452. tion. Harvard University Press, Cambridge, ENDLER,J. A. 1982b.Pleistocene forest refuges:Fact Massachusetts. or fancy?Pages 179-200 in BiologicalDiversifi- MCKITRICK,M. C., AND R. M. ZINK. 1988. Species cation in the Tropics (G. Prance,Ed.). Columbia conceptsin ornithology. Condor 90:1-14. University Press,New York. MOORE,W. S. 1977. An evaluationof narrow hybrid FREEMAN, S., AND W. M. JACKSON.1990. Univariate zones in vertebrates.Quarterly Review of Biol- metrics are not adequateto measurebody size. ogy 52:263-277. Auk 107:69-74. PAYNTER,R. A., JR.1985. Ornithological Gazetteer of FRY,C. H. 1970. Ecologicaldistribution of birds in Argentina. Museum of Comparative Zoology, north-eastern Mato Grosso State, Brazil. Anales Cambridge, Massachusetts. de la Academia Brasilefia de Ci•ncias 42:275- PAYNTER,R. A., JR.1989. Ornithological Gazetteer of 318. Paraguay, 2nd ed. Museum of Comparative Zo- GOULD,S. J., AND R. E JOHNSTON.1972. Geographic ology, Cambridge, Massachusetts. variation. Annual Review of Ecology and Sys- PAYNTER,R. A., JR.1992. Ornithological Gazetteer of tematics 3:457-498. Bolivia, 2nd ed. Museum of Comparative Zool- GRANT,P. R., AND B. R. GRANT.1992. Hybridization ogy, Cambridge,Massachusetts. of bird species.Science 256:193-197. PAYNTER,R. A., JR., AND M. A. TRAYLOR.1991. Orni- HAFFER,J. 1985. Avian zoogeography of the Neo- thologicalGazetteer of Brazil, 2 vols. Museum of tropical lowlands. OrnithologicalMonographs 36:113-146. ComparativeZoology, Cambridge, Massachusetts. PINTO, O. M. DE O., AND E. A. DE CAMARGO. 1961. HAFFER,J. 1987. Biogeographyof Neotropical birds. Resultadosornitol6gicos de quatro recentesex- Pages105-150 in Biogeographyand Quaternary pedi•6es do Departamentode Zoologia ao nord- History in Tropical America (T. C. Whitmore and G. T. Prance, Eds.). Clarendon Press, este do Brasil, coma descri•ao de seis novas su- Oxford. besp6cies.Arquivos de Zoologia de Sao Paulo 11:193-284. HARRISON,R. G. 1990. Hybrid zones: Windows on evolutionary process.Pages 69-128 in Oxford PROHASKA,F. 1976. The climate of Argentina, Para- Surveys in Evolutionary Biology,vol. 7 (D. Fu- guay and Uruguay. Pages13-112 in World Series tuyma and J. Antonovics,Eds.). Oxford Univer- of Climatology,vol. 12. Climates of Central and sity Press,Oxford. SouthAmerica (W. Schwerdtfeger,Ed.). Elsevier HAYES,E E. 1995. Status,distribution and biogeog- ScientificPublishing Company, Amsterdam. raphy of the birds of Paraguay.American Bird- REMINGTON,C. L. 1968. Suture-zonesof hybrid in- ing AssociationMonographs in Field Ornithol- teraction between recently joined biotas. Pages ogy 1:1-230. 321-428 in Evolutionary Biology,vol. 2 (T. Dob- HELLMAYR, C. E. 1929. A contribution to the orni- zhansky, M. K. Hecht, and W. C. Steere,Eds.). thologyof northeasternBrazil. Field Museumof Appleton-Century-Crofts, New York. Natural History Publication255, ZoologicalSe- REMSEN,J. V., JR.1984. High incidenceof "leapfrog" ries vol. 12, no. 18. pattern of geographicvariation in Andean birds: HOFFMAN, J. A. J. 1975. Climatic atlas of South Amer- Implications for the speciationprocess. Science ica, vol. 1. Maps of mean temperature and pre- 224:171-173. cipitation. WMO (Geneva),UNESCO (Paris) and REMSEN,J. V., JR.,AND R. S. RIDGELY.1980. Additions Cartographica(Budapest). to the avifauna of Bolivia. Condor 82:69-75. JAMES,E C. 1970. Geographicsize variation in birds REMSEN,J. V., JR., AND M. A. TRAYLOR,JR. 1983. Ad- and its relationshipto climate.Ecology 51:365- ditions to the avifaunaof Bolivia, part 2. Condor 390. 85:95-98. 470 FLOYDE. HAYES [Auk, Vol. 118

REMSEN,J. V., JR., AND M. A. TRAYLOR,JR. 1989. An SOKAL,R. R., ANDF. J. ROHLF. 1981. Biometry, 2nd ed. Annotated List of the Birds of Bolivia. Buteo W. H. Freemanand Company,New York. Books, Vermillion, South Dakota. TRAYLOR M. A. 1950. Altitudinal variation in Boli- RHYMER,J. M., AND D. SIMBERLOFF.1996. Extinction vian birds. Condor 52:123-126. by hybridizationand introgression.Annual Re- TRAYLOR,M. A. JR.1979. Subfamily Elaeniinae. Pag- view of Ecologyand Systematics27:83-109. es 3-112 in Check-list of Birds of the World (M. RIDGELY, R. S., AND G. TUDOR. 1989. The Birds of A. Traylor,Jr., Ed.). Museum of ComparativeZo- South America, vol. I. The Oscine . ology,Cambridge, Massachusetts. University of TexasPress, Austin. TRAYLOR,M. A., JR.1982. Notes on tyrant flycatchers RIDGELY, R. S., AND G. TUDOR. 1994. The Birds of (Aves:Tyrannidae). Fieldiana Zoology,new se- South America, vol. II. The Suboscine Passer- ries 13:1-22. ines. University of TexasPress, Austin. WETMORE,A. 1924. Descriptionof a new flycatcher RISING,J. D., AND K. M. SOMERS.1989. The measure- from Argentina. Auk 41:595-596. ment of overallbody size in birds. Auk 106:666- WETMORE,A. 1926. Birds of Argentina, Paraguay, 674. Uruguay,and Chile. Bulletinof the United States SCHUELER,E W., AND J. D. RISING. 1976. Phenetic ev- National Museum 133:1-448. idence of natural hybridization. SystematicZo- WILLIS,E. O. 1992.Zoogeographical origins of east- ology 25:283-289. ern Brazilianbirds. OrnitologfaNeotropical 3:1- SHORT,L. L. 1969. Taxonomicaspects of avian hy- 15. bridization. Auk 86:84-105. WOODRUFF,D. S. 1973.Natural hybridizationand hy- SHORT,k. L. 1975. A zoogeographicanalysis of the brid zones.Systematic Zoology 22:213-218. South American Chaco avifauna. Bulletin of the ZIMMER, K. J., A. WHITTAKER,AND D.C. OREN. 2001. American Museum of Natural History 154:167- A crypticnew speciesof flycatcher(Tyrannidae: 352. Suiriri)from the Cerradoregion of centralSouth SIBLEY,C. G., AND B. L. MONROE, JR. 1990. Distri- America. Auk 118:56-78. bution and Taxonomyof Birdsof the World.Yale ZIMMER,J. T. 1955. Further notes on tyrant flycatch- University Press,New Haven, Connecticut. ers (Tyrannidae). American Museum Novitates SIBLEY,C. G., AND L. L. SHORT,JR. 1964. Hybridiza- 1749:1-24. tion in the orioles of the Great Plains. Condor 66: ZINK, R. M., ANDJ. V. REMSEN,JR. 1986. Evolutionary 130-150. processesand patterns of geographicvariation SICK,H. 1984. Ornitologia Brasileira, uma introdu- in birds. Current Ornithology 4:1-69. cao, vol. 2. Editora Universidade de Brasilia, Brasilia. Associate Editor: R. Zink April 2001] Evolutionin theSuiriri Flycatcher 471

APPENDIX.Specimen number, sex, locality, date (1990s),and hybrid index scorefor three plumage color variablesfor all Suiriri suiriri bahiaeand hybrid S.s. suiriri x affinisspecimens examined in this study. Date Plumagescore Specimen Sex Locality (D-M-Y) Back Rump Belly S.s. bahiae AMNH 243914 ct Belo Jardim, Pernambuco, Brazil 26-02-27 0 0 1 AMNH 243916 c• Gilbfies/Pindal'ba, PiauL Brazil 11-06-27 0 0 0 AMNH 243917 9 Gilbfies/Pindal'ba, PiauL Brazil 11-06-27 0 0 0 FMNH 64119 9 Queimadas, Bahia, Brazil 12-08-13 0 0 2 S.s. suiriri x affinis AMNH 319815 Zanja Moroti, Concepc16n,Paraguay 03-09-30 1 0 2 AMNH 319812 Zanja Moroif, Concepc•6n,Paraguay 31-08-30 1 0 2 AMNH 319814 Zanja MorotL Concepc16n,Paraguay 03-09-30 1 0 1 AMNH 319811 Zanja Moroti, Concepc16n,Paraguay 31-08-30 1 0 1 AMNH 319813 Zanja Moroti, Concepc16n,Paraguay 03-09-30 1 0 1 AMNH 319808 Zanja Moroif, Concepc16n,Paraguay 03-09-30 1 0 0 AMNH 319816 Zanja MorotL Concepc16n,Paraguay 05-09-30 1 0 1 AMNH 321129 San Luis de la Sierra,Concepci6n, Paraguay 12-05-31 1 0 2 AMNH 321128 San Luis de la Sierra,Concepc16n, Paraguay 22-05-31 1 0 2 AMNH 321127 San Luis de la Sierra,Concepc16n, Paraguay 16-05-31 1 0 2 AMNH 321123 San Luis de la Sierra,Concepc16n, Paraguay 23-05-31 1 0 0 AMNH 321126 San Luis de la Sierra, Concepcl6n,Paraguay 15-05-31 1 0 2 AMNH 321125 San Luis de la Sierra, Concepc16n,Paraguay 15-05-31 1 0 2 AMNH 321130 San Luis de la Sierra,Concepc16n, Paraguay 15-05-31 1 0 2 AMNH 321124 San Luis de la Sierra,Concepc16n, Paraguay 12-05-31 2 1 1 AMNH 321122 San Luis de la Sierra,Concepc16n, Paraguay 18-05-31 1 0 0 AMNH 321131 San Luis de la Sierra,Concepc16n, Paraguay 21-05-31 1 0 2 AMNH 319489 Campan•rio, Mato Grossodo Sul, Brazil 07-07-30 1 2 2 CM 80363 R•o Quizer, Santa Cruz, Bolivia 07-06-18 1 1 0 CM 80364 Rio Quizer, Santa Cruz, Bolivia 07-06-18 1 1 0