The Auk 109(2):346-357, 1992

BIRD SPECIATION IN SUBTROPICAL IN RELATION TO FOREST EXPANSION AND RETRACTION

MANUEL NORES ConsejoNacional de InvestigacionesCientfficas y Tgcnicas,Centro de ZoologfaAplicada, Casillade Correo122, 5000 Cdrdoba,

ABS?RACT.--Ianalyzed two patternsof distribution in subtropicalSouth America. The first was the disjunctdistribution of pairs of speciesand subspeciesbetween the southern Yungas and the Paranenseforests, which are separatedby 700 .kin of xerophytic Chaco woodland.Forest penetrateup to 200 km into the Chacoalong gallery forests,but do not crossthe rest of the Chaco, which constitutesan effective barrier. The secondpattern involved several zones of secondarycontact located in the Chaco lowland, where several woodlandand grasslandspecies and subspeciesinteract. I concludethat both bird-distribution patternswere producedby forest expansionsalong the Bermejoand Pilcomayorivers that connectedthe southernYungas to the Paranenseregion and interrupted the arid vegetation in the center of the Chaco.Vegetational fluctuations probably occurredseveral times during the Quaternary and producedphenotypic differentiationof sistertaxa, both in the now- disjunctforests and in the currently continuousChaco. Received 26 June1990, accepted10 January 1992.

THE FORESTand woodland area located around cluded that the Paranense forest of Misiones the Tropic of Capricorn in South America is was once continuous, through , with biogeographicallycomplex. The region shows the Yungasof Tucum•n-.However, O1- two distinct features: (1) the southernmost ex- rog (1984) reversedthis idea when he presented tension of tropical South American forests(Pa- a present-day ornithogeographic map of Ar- ranenseand southernYungas), with a high de- gentina in which he connectedboth regionsby gree of avian endemism (Cracraft 1985, Nores a continuousgallery forestalong the Pilcomayo 1989); and (2) the central Chaco woodland, with River. With regard to the secondpattern, Short low avian endemism (Short 1975). Relatively (1975) identified the area of the Bermejo and little is known about the climatic history of the Pilcomayorivers as a zone that marks a disjuc- area and the effects of Quaternary climatic tion in bird distribution in South America. changeson the dispersaland differentiation of Research carried out in South America over bird .Understanding shifts in the re- the last 30 years has shown that, during the gion's vegetation during the Quaternary could Quaternary,semiarid conditions alternated with help to provide clues about events in the rest humid phases.These cyclescaused shrinkages of South America (B. Simpson,pers. comm.). of formerly widespread forests into humid Two separatepatterns in the distribution of pockets(refuges), followed by their subsequent the avifauna probably resultedfrom large-scale reexpansion (Garner 1959, Bigarella and An- vegetational dynamicsduring the Quaternary. drade 1965, Haffer 1969, 1974,Mayr 1969,Mayr First, there are several species and subspecies and O'Hara 1986, Vanzolini and Williams 1970, pairs that have disjunct populations in the Vuilleumier 1971, Prance 1974, Van der Ham- southern Yungas and Paranenseforests, sepa- men 1974, 1982,Tricart 1974, 1975, Simpsonand rated by 700 km of xerophyticChaco woodland Haffer 1978, Absy 1979, 1982, Schubert 1988). (Fig. 1). Second,a zone of disjunction occursin Paleoecologicaldata and present distribution the Chaco, along the Bermejo and Pilcomayo patterns of and plants suggest that rivers, where the distribution of several species during interglacial phasesthere were periods and subspeciespairs of woodland and grassland when temperature and humidity were higher birds meet (Fig. 1). and forestsmore widely distributedthan today The first pattern was previously pointed out (Groeber 1936, Smith 1962, Vanzolini 1968, 1974, by Olrog (1963), who noted that several bird 1981, Haffer 1974, 1985, Van der Hammen 1974, speciesof the southern Yungasalso occur in the 1982,1983, Fitzpatrick 1980,Markgraf 1985,No- Paranenseforest, or are representedthere by res and Cerana 1990). Some authors consider anothersubspecies. Later, he (Olrog 1979a)con- the evidence of aridity during the glacial pe- 346 April1992] BirdSpeciation inSouth America 347

Jill • IllIll[[ I• • ß IIIIIIll I1•_ ß ß JillIfil ß IIIIIlll ß • IIIIflll

IIIIIIII

b

• Amazonian

• Paranense

L• Serra do Mar

•1Yungas and eastern slopes

O 1OOO 2000

Km

Fig. ! Biogeographicprovinces of South America, excluding the Pacific coast,most of the Andes, and southernand centralArgentina (adaptedfrom Cabreraand Willink 1973).Three regionsthat form the "arid diagonal" (Vanzolini 1974, Fitzpatrick 1980): (a) Chaco, showing the area of interaction of woodland and grasslandbirds (box with oblique lines); (b) Cerrado;and (c) Caatinga. From south to north, the breaks in the black band (representingYungas and Andes slopes)mark the limit of the southernYungas, northern Yungas,and easternslopes of the northern Andes.

riods to be insufficient to indicate that the fauna are relatedto Quaternaryclimatic changes wet tropical lowlands changed substantially and if they have common histories, I studied (Connors 1986, Colinvaux 1989). Endlet (1977, the distributionof forestbirds and plantsin the 1982a,b) pointed out that current patterns of Chaco lowland. I then comparedthe levels of distribution are consistentwith geographical phenotypic differentiation of the birds that in- divergenceand adaptationsto present-daycon- habit the southernYungas with thoseoccurring ditions.Consequently, it is not necessaryto pos- in the Paranense region, and examined the tulate refuges to explain the observed distri- ranges of woodland and grasslandbirds that butions. currently interact in the center of the Chaco. To determine if the two patterns of the avi- Finally, I relatedthe extentof phenotypicdif- 348 MANUEL NORES [Auk, Vol. 109

-[] Yungasforest ] Paranenseforest ..,[]Galleryforest

I • I I I 0 100 200 Km

Fig. 2. Vegetationdistribution in studyarea. Solid dots indicate location of dry riverbedswith forest patches.Triangles indicate presence of foresttrees related to present-daywatercourses. (a, b and c) The most extremelocations of galleryforests in Chacolowlands. (d) Paranenseforest wedge reaching Paraguay River. Galleryforests located east of Paraguayand Paran•rivers, in floodedarea, are not represented(from Notes 1989).

ferentiation of these birds with vegetation proximately700 km wide,this area slopes gently from changesthat apparently occurredin South west to east,and extendsinto Argentina,Bolivia and America during the Pleistoceneand Holocene. Paraguay.Rainfall increasesmarkedly from 500 mm in the west to 1,300 mm in the east. The Chaco, to- STUDYAREA getherwith the "Cerrado"and "Caatinga"in , formsthe "arid diagonal"(Vanzolini 1974, Fitzpatrick The studyarea is subtropicalSouth America,spe- 1980; Fig. 1). cificallythe region drainedby the Bermejo,Pilco- The third componentis the forest and "campos" mayo,Paraguay, and Paran•rivers (Fig. 2). Fromwest (Paranenseregion) of southeasternBrazil, eastern to east it includes three components. Paraguayand northeasternArgentina, around the up- The first component is the forests on the eastern per coursesof the Paran•and Uruguayrivers. Rainfall slopesand foothills of the Andes,and on the Sierras decreases towards the west from 1,700 mm to 1,300 Pampeanassouth of 18øS(southern Yungas). These mm. As a result,the vegetationbecomes increasingly forests stretch from Santa Cruz and Cochabamba de- impoverishedand it ends in a narrow wedge on the partmentsin Bolivia to CatamarcaProvince in Ar- ParaguayRiver (Fig. 2, point d). Detailed descriptions gentina.I definedthree altitudinal belts: (a) basalfor- of the study area are given in Hauman et al. (1947), est between 350 m and 600 m; (b) cloud forest between Ad•rnoli et al. (1972),Short (1975), and Cabrera(1976). 600 and 1,500 m; and (c) montane deciduousforest between 1,400 and 2,500 m. MATERIALS AND METHODS The secondcomponent includes the Chacolowland drained by the Bermejoand Pilcomayorivers and Avifauna.--I carried out field work from 1973 to other small tributaries of the ParaguayRiver. Ap- 1988 in the forests and Chaco woodland of northern April1992] BirdSpeciation inSouth America 349

Argentina, southern Bolivia, eastern Paraguay, and river and stream banks. Outside the flooded ar- southern Brazil (Nores 1989). From 1986 to 1988 I eas,the gallery forestswiden considerably.On concentratedon areas of particular biogeographic the Pilcomayoand Bermejorivers, the gallery complexity. These included river and stream banks, forests extend about 200 km to the west, reach- forest patcheson dry riverbeds, Paranenseprairies, and the interaction zone in the Chaco lowland. I cov- ing Fortln SargentoPrimero Leyes (24ø30'S, ered Argentine National Highway 81, which crosses 59ø23'W)and Puerto Bermejo(25ø39'S, 60ø08'W; the Chaco lowland from east to west, and most of the see Fig. 2, points a and b). On the streams,the secondaryroads that give accessto the Bermejoand gallery forestsare somewhatshorter. Pilcomayorivers and other small water courses(Fig. In the central and western Chaco rainfall is 2). At each study site, I identified all bird species lower and forest trees occuras small patches(of presentthrough a combinationof visualobservations, one to five species)on dry riverbeds,or in small songidentification, song playback, and capturewith groups(of one to three species)on river banks. mist nets. The most common species are: Phyllostylon I examined skin collections at the Fundaci(•n Mi- rhamnoides(Ulmaceae); Chlorophora tinctoria guel Lillo in Tucum•n and at the Museo Argentino de Ciencias Naturales in Buenos Aires. These insti- (Moraceae);Ruprechtia laxifiora (Polygonaceae); tutions housegood bird collectionsfrom the study Pisoniazapallo (Nyctaginaceae); Acacia albocorti- area, especiallyfrom the forestsof Misiones, Tucu- cata,Anadenanthera colubrina, Cathormion polyan- m•n, and Jujuy.To estimatethe levelsof phenotypic thum, Enterolobiumcontortisiliquum, Gleditsia differentiation reached by the populations of forest amorphoides,Pithecellobium scalare and Pterogyne and Chaco birds, I analyzed plumage pattern, color, nitens(Fabaceae); Astronium balansaeand A. urun- and body measurements(wing chord, tail and ex- deuva(Anacardiaceae); Patagonulaamericana posedculmen). After applying generalmethods used (Boraginaceae);Tabebuia ipe (Bignoniaceae); and in comparativesystematics, I classifieddisjunct pop- Calycophyllummultifiorum (Rubiaceae). Near the ulationsin termsof presenceor absenceof geograph- Andes, a short gallery forest extends eastward ical variation. In thosecases showing variation, I ap- along the BermejoRiver for about 45 km, up to plied the terminologyof Mayr (1963),Amadon (1966), and Amadon and Short (1976), but I used their cate- Batea(23ø22'S, 63ø43'W; see Fig. 2, point c). There goriesin a more general way. I recognizedallospecies is no gallery forest along the PilcomayoRiver asallopatric species with similarmorphological char- in this zone. Consequently,the shortest dis- acteristics;megasubspecies as subspecies that are con- tance between the Paranense and southern spicuouslydistinct in color or size, and subspeciesas Yungas forests,about 450 km, occursbetween populationswith somemorphological differences, and the gallery forestsalong the upper and lower generally with partial overlap of diagnosticcharac- Bermejo River (Fig. 2). ters. Polytopicsubspecies are composedof geographi- Distributionof forest birds.--The distribution of cally separatedand indistinguishablepopulations. forestbirds in the Chacolowland (Fig. 3) shows Vegetation.--I determined the presenceor absence that the easterngallery forestsrepresent major of gallery forestand the foresttree speciespresent at each sampling site on the Bermejo and Pilcomayo avenues by which the Paranensebirds pene- trate westward into the Chaco lowland. Of 236 rivers, and on small rivers and streams.By visual observation, I surveyed dry riverbeds from the base forest bird species recorded between 54ø and of the Andes up to the ParaguayRiver. At eachdry 55øW,62 (26%) reach the ParaguayRiver, and riverbed, I determined the presence or absence of 54 (23%) occurin the gallery forestsof the Ber- forest patches and recorded the forest tree species mejo and Pilcomayo rivers and other water present (Nores 1989). courses. In the west, the gallery forestof the Bermejo RESULTS River also allows dispersalof forest birds from the southern Yungas into the Chaco, but this Distributionof forest vegetation.--Someforest encroachmentis not as conspicuousas in the trees occur throughout the entire Chaco (Fig. east. Of 115 forest bird speciesrecorded in the 2). This contradictsstatements by Morello and southern Yungas, only 10 (11%) occur in the Ad•moli (1974), who indicated that a large Bermejogallery forest.In the remainder of the semiarid gap preventsany plant speciesfrom central and western Chaco lowland, including the southern Yungas from reaching the Para- the forest patches,there are virtually no forest nense region. birds. Only Todirostrumplumbeiceps (recorded on In the flooded areas of the eastern Chaco, the PilcomayoRiver [21ø09'S,63ø12'W], on the forestvegetation occurs as narrow gallerieson Bermejo River [24ø38'S,61ø25'W], and in Bru- 350 MANUELNOKV. S [Auk,Vol. 109

Tinamus solitarius Claravispretiosa ,4ra maracana a/is.. icillata

, ruficapillu.v

•stanotis albicollis -- Batara cinerea

Leptopogonatnauroc½ nauroct•halus _ = Pacbyratnplua• viridis Tityra. •da Cyanocorm:c) Nonosiapileata

• melanops

•eciosum 5' 2• o River

•gas forest

8 15 17, ermejoRiver•-••..._..• 1210 18 "• Paranenseforest' Mcsetnbrinibis • c•nnctuds• Leptodoncaj anet•is SpizaetusO coll• , Micrastur s itorquams Falco 'ris Pcncloln:obscura Cratzfasciolata l lcliomisfulica Colutnbaca), Forlmsxanth• •,tc•us iliani.

itorquat• Nyctidromusalbicollis•lbicoll• Ratnphastosdicolorusicoloms Ratttl•hastostoco Vcniliornist DO•copusl, • plaO•os•h •)vtallttvisscutata Philydorrufosu u•rciliatust Myiopag ]icata lum•icel• Knipoleguscy Colonia colonus •egatusIcUcol: ,hai• Pachyratni Pachyratnp : valid• Tio,rainqutfftor Turdus leuco•nelelas tida I lonithraupis?a Euphonia•nusica #a• ost•s ' $altator aimills Basileuteru• qicivo•s Basilcuten•. •coblephams•- Psarocc Cacicus haetnorr•ho•.• Scaphidura April 1992] BirdSpeciation in South America 351 chard [25ø09'S,59ø58'W]) crossed the western DISCUSSION Chacofrom the southernYungas along the riv- ers. Some Paranensebirds occasionallyfol- At present,faunistic corridors (gallery forests lowed the rivers or marshylands westward be- and one forestwedge; Fig. 2, point d) allow the yond the limit of the gallery forests,but did dispersal of the Paranense and Yungas birds from the east and west into the Chaco. How- not penetrate the xerophytic Chaco. Examples are Columbacayennensis, Batara cinerea, Pachyram- ever, the barrier of xerophytic vegetation, cur- phusviridis, P. castaneus,and Arremonfiavirostris rently 450 km wide, precludesany direct con- nection between the birds of these two forest (Fig. 3). Distributionof nonforestbirds.--About 180 spe- regions. cies of land birds occur in the Chaco zones In the recent past (possiblyduring the Qua- (woodland, grasslandand savanna)of northern ternaryinterglacial periods), the Chacoarea may Argentina, western Paraguay,and southeastern have been more humid than today (Vuilleumier Bolivia (Short 1975,my study).A commonpat- 1965,Short 1975,Simpson 1979). The presence tern in the distributions of these birds is a wide of dry riverbedswith relict forest patches,and range from north to south without differenti- the occurrenceof foresttrees in the upper part ation. of the channels of the Bermejoand Pilcomayo rivers indicate that the area was more humid in In contrast to this pattern, a group of Chaco birds is representedby pairs of speciesor sub- former times. Consequently,forests advanced species that have distributional limits in the from the southern Yungas and the Paranense area of the Bermejoand Pilcomayorivers (Figs. region along the Bermejoand Pilcomayorivers 4-6). Six of these pairs have contact zones in deep enoughto form a continuousforest bridge this region (Figs. 4A, 4B, 4C, 5A, 5B, and 6B), between both regions. Birds presumably ex- two others show narrow overlap (Figs. 5C and panded to form a continuous distribution and 5D), and the remaining species(Figs. 4D and becameisolated again in the subsequentdrier 6A) form hybrid zones.With the exceptionof period. This connection-interruption process Thamnophiluscaerulescens and Phacellodomusruff- most likely occurredseveral times during the frons,which frequent Chacowoodland as well Pleistocene and Holocene (Vuilleumier 1971, as forest,all of thesespecies inhabit only non- Van der Hammen 1974, 1983,Fitzpatrick 1980). forest habitats. Different levels of bird speciationcurrently ob- Levelsof speciation.--Mystudy of museum servable in both forests(Table 1), therefore, can specimensfrom the southern Yungas and the be tentatively related to the time elapsedsince Paranense forest showed that the species com- the speciescrossed. Thus, speciesthat crossed mon to both regions exhibited various degrees during the last connectionwould not yet have of phenotypic differentiation (Table 1). One had time to differentiate.Another group of spe- group (6 taxa)included polytopic subspecies and cies that show differences at subspecieslevel monotypicspecies, a second(22 taxa) included presumablycrossed during an earlier connec- subspecies,a third (6 taxa) included megasub- tion. Slightly differentiated subspeciesof this species,and a fourth (8 taxa) was composedof group (marked with footnote b in Table 1) prob- allospecies.Similarly, the study of specimens ably belong to the first group. A third group of Chaco birds from the interaction zone showed could have crossed even earlier and differen- different levels of speciation similar to those tiated to megasubspecieslevel. Finally, pairs of observed in forest birds. allopatricspecies probably representdifferen-

Fig.3. Distributionof forestbirds in Chacolowland. Numbers indicate most extreme records: (1) Asunci6n City, Bernalcueand Lainbar6;(2) Monte $ociedad;(3) Rio PilcomayoNational Park;(4) Fortin Sgto.Primero Leyes;(5) Montelindo River, near Fortin Mcal. L6pez; (6) EscalanteLagoon; (7) He He Stream;(8) Monte Lindo Stream;(9) PilagfiStream; (10) Gran Guardia; (11) El Colorado;(12) 15 km NW of Piran6;(13) Presidencia Roca;(14) GuaycurfiStream, (15) E1Resguardo; (16) PuertoBermejo; (17) Bruchard;(18) ColoniaKm 503; (19) Pozo de los Suris; (20) Batea and Padre Lozano; (21) El Quimilar; and (22) Las Varas (from Notes 1989). 352 MANUELNORES [Auk,Vol. 109

A B

Rhea americana [] R. a. araneipes ::[] N. boraquira •[] R. a. albescens [] N. maculosapallida [] N. chacoensis

G D

Colaptes formosa [] C. campestris ß E.f. mira [] Hybridzone •1ß E. f. formosa [] C. campestroides

Fig. 4. (A) Rangesof two subspeciesof Greater Rhea (Rhea americana;modified from Short 1975). The PilcomayoRiver marksdistribution limit of thesesubspecies. (B) Rangesof three speciesof Nothura (Nothura). For N, maculosa,only pallidarace representedand, for N. boraquira,only its southernpopulation shown. The three speciesoverlap in center of Chaco.(C) Range of QuebrachoCrested (Eudromiaformosa). The PilcomayoRiver marks distribution limit of its two races.(D) Rangesof two allospeciesof Campo Flicker (Colaprescampestris), showing zone of hybridization in Paraguay(from Short 1975). April 1992] BirdSpeciation in South America 353

A B

Asthenes baeri Carnpylorhamphustrochilirostris C. t. lafresnayanus • A. b. baeri(and neifi) ::• C.t. hellmayri ß A. b. chacoensis

c D

Phacellodomus Pseudoseisura ,'[] P. rufifronssincipitalis [] P. cristataunirufa [] Zoneof overlap •:D P. Iophotesargentina ;[] P. sibilatrix

Fig. 5. (A) Rangesof two selectedsubspecies of Red-billed Scythebill (Campylorhamphustrochilirostris; modifiedfrom Short 1975).Their rangesare limited by PilcomayoRiver. (B) Rangeof Short-billedCanastero (Asthenesbaeri), showing distributionlimit of two of its racesnear PilcomayoRiver. (C) Rangesof southern subspeciesof Rufous-frontedThornbird (Phacellodomusrufifrons) and its allospecies,the Little Thornbird (P. sbilatrix). Overlap occurs in center of Chaco. (D) Ranges of southern race for Rufous-crestedCacholote (Pseudoseisuracristata; from Short 1975)and of Brown Cacholote(P. lophotes)showing narrow overlapin center of Chaco. 354 MANUELNORES [Auk, Vol. 109

A B

Thamnophiluscaerulescens :.[q•lT. c. paraguayensis Rhinocryptalanceolata • T, c. dinellii ß R. I. saturata [] R. I. lanceolata

Fig. 6. (A) Rangesof two subspecies(possibly species) of Variable Antshrike (Thamnophiluscaerulescens), showingsmall zone of hybridizationin centerof Chaco.(B) Rangeof CrestedGallito (Rhinocrypta lanceolata). Its two subspecieshave their distributionlimit near PilcomayoRiver.

tiation that began very early in the wet-dry cy- and subspecieslevels, I believe that this barrier cles, although it is possiblethat isolation and interrupted the arid diagonalseveral times dur- speciation began before the Quaternary. Al- ing the Quaternary. though this classificationassumes that all pop- Forestbirds may have reached the disjunct ulationsdiverge at approximatelythe samerate, area by crossingthe arid diagonalwithout the ! think that other studies, such as biochemical need of a forest bridge. Although this alterna- analyses,may slightlymodify this classification, tive may apply to somecanopy or middle-story and provide better data concerningrelative ages species,such as Pachyramphus validus and Turdus of populations. nigriceps(which colonizeisolated forested areas; The distribution patterns of nonforest birds Nores 1989, Nores and Cerana 1990) or to To- that interact in the center of the Chaco (Figs. dirostrumplumbeiceps, it is not likely to apply to 4-6) also are consistent with the former exis- terrestrialor forestinterior species(Haffer 1974: tence of a forest belt along the Bermejo and 163, Nores 1989).Mayr (1969) pointed out that Pilcomayo rivers. Mayr (1963), Simpson and savannasof all typesare formidable barriersfor Haffer (1978), and other authors have inter- birds of the tropical rain forest. Vuilleumier preted zones of hybridization, characterdis- (1965) suggestedthat the lack of marked dif- placement,and narrow overlap as indicatorsof ferentiation between two subspeciesof Penelope secondarycontact between formerly isolated obscurainhabiting these forests indicates that populations.At present,there are no geograph- gene flow was continuousbetween them. How- ic or ecologicalfeatures in the area that suggest ever, a more likely explanationis that a popu- a plausiblealternative hypothesis. Because Cha- lation of this speciesrecently crossed(probably co birds are capable of crossingextensive un- before the last connection) and, therefore, has favorableareas (unpubl. data), the Bermejoand not yet acquiredmarked differences. Pilcomayorivers, with a maximumwidth of 500 A second alternative is that forest birds have m, cannot constitute effective barriers to dis- colonized the southern Yungas and the Para- persal.Because the degreesof speciationillus- nenseregion from Amazonia. In the southern trated by these birds include various species Yungas, this may have occurredthrough the April 1992] BirdSpeciation in South America 355

TABLE1. Speciationlevels of forestbirds inhabiting northern Yungas (Fig. 1). In the Paranensere- southernYungas and Paranenseregions. a gion, this could have occurred only in former

Paranense geological times, when Amazonia may have SouthernYungas region beenconnected with the Paranenseregion and the Serra do Mar (Smith 1962, Vanzolini 1968, Polytopic subspeciesand monotypic species 1974, 1981, Haffer 1974, 1985). This hypothesis Corythopisdelalandi Phyllomyiasburmeisteri burmeisteri could accountfor the presenceof only 30% of Myiopagiscaniceps caniceps the speciesinvolved, becausethe remainder is Leptopogonamaurocephalus amaurocephalus not present in Amazonia. Moreover, neither of Trichothraupismelanops the two alternative hypothesiscould account Euphoniamusica aureata for the interaction zone in the center of the Trichothraupismelanops Chaco. Euphoniamusica aureata Subspecies Crypturellusobsoletus punensis C. o. obsoletus Penelopeobscura bridgesi P.o. obscura ACKNOWLEDGMENTS Pionusmaximiliani siy P.m. melanoble- pharus I thank D. Yzurieta and S. Salvador for their par- Piaya cayanamogenseni ticipation and cooperationduring field work. Very Pulsatrixperspicillata boliviana P. p. pulsatrix helpful commentswere made on the manuscriptby Aegoliusharrisii dabbenei A. h. iheringet• P. Feinsinger, J. V. Remsen, A. H. Brush, E. Bucher, Nystaluschacuru uncirostris N. c. chacuru J. Haffer, B. B. Simpson,A. Balmford, J. Faaborg,K. Dendrocolaptespicumnus casaresi D. p. extimus Silvius, J. Navarro, W. Norcore, and H. Capurro. I am P. r. acritus Philydorrufosuperciliatus oleagi- grateful to: J. Navas (Museo Argentino de Ciencias neus Naturales)and J. B. de Herrera (Fundaci6nMiguel Philydorrufus bolivianus P. r. rufus Xenopsrutilans connectens X. r. rutilansb Lillo, Tucum•n) for permitting me to work on the Bataracinerea argentina B. c. cinerea Collectionsin their care; M. M. Cerana for the iden- Thamnophilusruficapillus cocha- T. r. ruficapillus tification of plant species;P. Canevari for providing bambae me a list of birdsof Rio PilcomayoNational Park;and Chamaezacampanisona boliviana C. c. tshororo M. A. Gon for providing me with recent data on the Elaenia obscura obscura E. o. sordida Pilcomayo River drainage. J. Warde drew the illus- Phylloscartesventralis tucumanus P. v. ventralis • trations. The project was supportedby the National Todirostrumplumbeiceps viridi- T. p. plumbeiceps GeographicSociety (Grant No. 3253) and CONICOR. ceps Turdusnigriceps nigriceps T. n. subalaris Hemithraupisguira boliviana H. g. fosteri Pipraeideamelanonota venezue- P.m. melanonota LITERATURE CITED lensis Basileuterus culicivorus virides- B. c. azarae • certs ABS¾,M.L. 1979. A palynologicalstudy of Holocene Phaeothlypisrivularis boliviana P. r. rivularis sediments in the . Thesis, Univ. Megasubspecies Amsterdam, Amsterdam. Tigrisomafasciaturn pallescens T. f. fasciatum ABSY,M. 1982. Quaternary palynologicalstudies in Picumnuscirratus thamnophi- P. c. temminckii the Amazon basin.Pages 67-73 in Biologicaldi- loides versification in the tropics (G. T. Prance, Ed.). Lochmias nematura obscurata L. n. nematura Columbia Univ. Press, New York. Pachyramphusvalidus audax P. v. validus AMADON,D. 1966. The superspeciesconcept. Syst. Turdusalbicollis contemptus T. a. paraguayensis Zool. 15:245-249. Arremonfiavirostris dorbignii A. f. polionotus AMADON, D., AND L. L. SHORT. 1976. Treatment of Allospecies subspeciesapproaching species status. Syst. Zool. Leptotilamegalura L. rufaxilla 25:161-167. Amazona tucumana A. pretrei AD.•MOLI, J., R. NEUMANN, A.D. RATIER, AND J. Veniliornisfrontalis V. passerinus MORELLO. 1972. E1 Chaco aluvional saltefio. Rev. Synallaxissuperciliosa S. ruficapilla Inv. Agr. Inta 9:165-237. P. virescens Phyllomyiassclateri BIGARELLA,J. J., AND G. O. DE ANDRADE. 1965. Con- Elaeniastrepera E. mesoleuca tribution to the study of the Brazilian Quater- Knipolegussignatus K. cyanirostris Cyanocoraxcyanomelas C. caeruleus nary. Geol. Soc. Am. Spec.Pap. 84:433-451. CABRERA,A. L. 1976. Regiones fitogeogr•ficasar- aDistribution data from Peters(1934-1951, 1960-1979) and Olrog (1979b). follows Peters (1960-1979), Meyer de Schauensee gentinas. Encic. Arg. Agr. Jard. Ed. Acme. B. Ai- (1966), and Short (1982). res. bSlightly differentiatedsubspecies. CABRr•A,A. L., ANDA. WILL•NK.1973. Biogeografia 356 MANUELNORES [Auk, Vol. 109

de AmericaLatina. Progr. Reg. Des. Cognate. Tec- ic evidencesupporting the Pleistoceneforest ref- nol. Ser. Biol. No. 13. OEA. uge hypothesis.Evolution 40:55-67. COLIt,lVAIJX, P.A. 1989. Pasadoy futuro de la Ama- MEYERDE $CHAUENSEE, g. 1966. The speciesof birds zon{a.Invest. y Cienc. 154:8-15. of SouthAmerica and their distribution.Living- CONNORS, E. F. 1986. The role of Pleistocene forest ston Publishing Co., Narberth, Pennsylvania. refugia in the evolution and biogeography of MORELLO,J., AND J. AD•MOLI. 1974. Lasgrandes uni- tropical biotas. Trends Ecol. & Evol. 1:165-168. dadesde vegetaci6ny ambientedel Chacoar- CRACRAET,J. 1985. Historical biogeographyand pat- gentino. Segundaparte: Vegetaci6ny ambiente terns of differentiation within the South America de la Provinciadel Chaco.Serie Fitogeog. No. 13. avifauna: Areas of endemism. Pages 49-84 in INTA, Buenos Aires. Neotropical ornithology (P. A. Buckley, M. S. NORES,M. 1989. Patronesde distribuci6ny causas Foster,E.S. Morton, R. S. Ridgely,and F. G. Buck- de especiaci6nen avesargentinas. Thesis, Univ. ley, Eds.).Ornithol. Monogr. No. 36. C6rdoba,C6rdoba, Argentina. Et,lr•I•ER, J.A. 1977. Geographicvariation, speciation, NORES,M., ANDM. M. CERANA.1990. Biogeography and clines. Princeton Univ. Press,Princeton, New of forest relics in the mountains of northwestern Jersey. Argentina. Rev. ChiL Hist. Nat. 63:37-48. ENDLER,J.A. 1982a. Problemsof distinguishinghis- OLROC;,C. C. 1963. Lista y distribuci6nde las aves torical from ecologicalfactors in biogeography. .Opera Lilloana 9:1-377. Am. Zool. 22:441-452. OI•ROG,C.C. 1979a. Notasornitol6gicas del noroeste ENDLER,J.A. 1982b. Pleistoceneforest refuges:Fact argentino (Aves Passeriformes).Neotr6pica 25: or fancy. Pages641-657 in Biologicaldiversifi- 125-126. cation in the tropics (G. T. Prance,Ed.). Columbia OLROC;, C. C. 1979b. Nueva lista de la avifauna ar- Univ. Press, New York. gentina. Opera Lilloana 27:1-324. FITZPATRICK,J.W. 1980. Some aspectsof speciation OLROG,C.C. 1984. Las aves argentinas.Una nueva in SouthAmerican flycatchers. Pages 1273-1279 gula de campo.Adm. Parq. Nac. Arg. Ind. Graf. in Symposiumon speciationin South American Alvi., Madrid. birds (L. L. Short,Convenor). Actis 17 Congressus PETERS,J. L. 1934-1951. Check-list of birds of the Internationalis Ornithologici. Deutsche Orni- world, 6 vols. Museum of ComparativeZoology, thologen-gesellschaft,Berlin. Cambridge,Massachusetts. GARNER,H.F. 1959. Stratigraphic-sedimentarysig- PETERS,J. L. 1960-1979. Check-list of birds of the nificanceof contemporaryclimate and relief in world. A continuation of the work of J. L. Peters, four regionsof the Andesmountains. Bull. Geol. 8 vols. Museum of ComparativeZoology, Cam- Soc. Am. 70:1327-1368. bridge, Massachusetts. GROEBER,P. 1936. Oscilaciones de clima en la Ar- PRANCE,G. T. 1974. Phytogeographicsupport for gentina desdeel Plioceno.Holmbergia 1:71-84. the theory of Pleistoceneforest refuges in the HAFFER,J. 1969. Speciationin Amazonianforest birds. Amazon Basin, based on evidence from distri- Science 165:131-137. bution patternsin Caryocaraceae,Chrysobalana- HAFFER,J. 1974. Avian speciation in tropical South ceae, Dichapetalaceaeand Lecythidaceae.Acta America. Nuttall Ornithol. Club, Cambridge, Amazonica 3:5-28. Massachusetts. SCHUBERT,C. 1988. Climaticchanges during the last HAFFER,J. 1985. Avian zoogeographyof the Neo- glacial maximumin northern SouthAmerica and tropical lowlands. Pages 113-146 in Neotropical the Carribean: A review. Interciencia 13:128-137. ornithology (P. A. Buckley, M. S. Foster, E.S. SHORT,L. L. 1975. A zoogeographicanalysis of the Morton, R. S. Ridgely, and F. G. Buckley, Eds.). South American Chaco avifauna. Bull. Am. Mus. Ornithol. Monogr. No. 36. Nat. Hist. 154:163-352. HAUMAN, L., A. BURKART, L. R. PARODI, AND A. L. SHORT,L. L. 1982. Woodpeckersof the world. Del- CABRERA.1947. La vegetaci6nde la Argentina. aware Mus. Nat. Hist., Monogr. Ser. 4. Geogr. Rep. Argent. 8:1-346. SIMPSON,B. B. 1979. Quaternary biogeographyof MARKGRAI•,V. 1985. Paleoenvironmentalhistory of the high montaneregions of SouthAmerica. Pages the last 10,000years in northwesternArgentina. 157-188 in The South American herpetofauna:Its Pages1739-1749 in SudamerikaGeomorphologie origin, evolution, and dispersal(W. E. Duellman, und Palaookologiedes jungeren Quartars (K. Ed.). Univ. Kansas,Mus. Nat. Hist. Monogr. 7. Garleffand H. Stingl, Eds.).E. Schweizerbart'sche Verlagsbuchhandlung,Stuttgart. SIMPSON,B. B., AND J. HAFFER. 1978. Speciation pat- terns in the Amazonian forest biota. Annu. Rev. MAYR,E. 1963. speciesand evolution. Har- vard Univ. Press,Cambridge, Massachusetts. Ecol. Syst.9:497-518. MAYR, E. 1969. Bird speciation in the tropics. Biol. SMITH,L. B. 1962. Origins of the flora of southern J. Linn. Soc. 1:1-17. Brazil. Bull. U.S. Natl. Mus. 35:215-249. MAYR,E., ANDR. J. O'HARA. 1986. The biogeograph- TRICART,J. 1974. Existencede p•riodes s•ches au April1992] BirdSpeciation inSouth America 357

Quaternaire en Amazonie et dans les r•gions VANZOLINI,P. E. 1974. Ecologicaland geographical voisines.Rev. Geomorph.Dynam. 4:145-158. distribution of lizards in Pernambuco, north- TRICART,J. 1975. Influence des oscillations clima- eastern Brasil (Sauria). Pap. Avulsos Zool. (Sao tiques r•centes sur le rnodel• en Amazonie Ori- Paulo) 28:61-90. entale(Region de Santar•m)d'apr•s les images V^tqzoLItqI,P.E. 1981. A quasi-historicalapproach radar lateral. Z. Geomorph. N.F. 19:140-163. to the natural history of the differentiation of VAI•IDER HAMMEI•I, T. 1974. The Pleistocenechanges reptiles in tropical geographic isolates. Pap. of vegetationand climatein tropicalSouth Amer- Avulsos Zool. (Sao Paulo) 34:189-204. ica. J. Biogeogr. 1:3-26. VANZOLINII,P. E., AND E. E. WILLIAMS. 1970. South VA•qr•R H^MMEtq,T. 1982. Paleoecologyof tropical American anoles:The geographicdifferentiation South America. Pages60-66 in Biologicaldiver- and evolution of the Anolis chrysolepisspecies sification in the tropics (G. T. Prance, Ed.). Co- group(Sauria, Iguanidae). Arq. Zool. (SaoPaulo) lumbia Univ. Press, New York. 19:1-298. VAN DERHAMMEN, T. 1983. The palaeoecologyand VUILLEUMIER,B.B. 1971. Pleistocene changes in the palaeogeographyof savannas. Pages 19-35 in fauna and flora of South America. Science 173: Tropical savannas (F. Bourliere, Ed.). Elsevier, 771-780. Amsterdam. VUILLEUMIER,F. 1965. Relationshipsand evolution VAhlZOl,It,n, P. E. 1968. Geography of the South within the Cracidae (Aves, Galliformes). Bull. American Gekkonidae (Sauria). Arq. Zool. (Sao Mus. Comp. Zool. 134:1-27. Paulo) 17:85-111.