MIXED-SPECIES FLOCKS OF BIRDS IN THE CERRADO, SOUTH AMERICA: A REVIEW
Dárius Pukenis Tubelis
Departamento de Ecologia, Instituto de Biociências, Universidade de São Paulo, São Paulo-SP, 05508-900, Brazil. E-mail: [email protected]
Resumo. – Bandos mistos de aves no Cerrado, América do Sul: uma revisão. – Revisões sobrebandos mistos de aves geralmente examinam informações obtidas em várias regiões do mundo, nãofocando avifaunas de ecossistemas particulares. Este trabalho teve como objetivo revisar informaçõessobre bandos mistos encontrados no Cerrado. Publicações (n = 28) mencionaram 172 espéciesparticipando de bandos mistos, em várias fitofisionomias e regiões. Estudos mais detalhados foramconduzidos somente em cerrado sensu stricto e campo cerrado. Membros de bandos foram vistosconsumindo frutos, néctar, insetos e sementes. Bandos com até 16 espécies e 40 indivíduos foramregistrados. Doze espécies de seis famílias já foram apontadas como espécies nucleares. Em geral, são avesde coloração contrastante, que também ocorrem em grupos mono-específicos quando ausentes de bandosmistos. Há evidência de que o risco de predação é um fator levando à formação de bandos mistos emcampo cerrado e cerrado sensu stricto. Também, ataques predatórios por espécies de Falco foram flagrados.Bandos mistos são menos freqüentes e têm menor número de indivíduos e de espécies durante a épocareprodutiva do que no período não-reprodutivo das espécies. Avaliações da avifauna conduzidas em váriasescalas espaciais revelaram que a participação em bandos mistos é uma estratégia adotada por uma porçãoconsiderável da avifauna encontrada no Cerrado. Entre as sugestões para pesquisas futuras está oaproveitamento do ambiente do Cerrado (seus mosaicos de vegetação, sua grande extensão e a fortesazonalidade de seu clima) para pesquisar aspectos da biologia de bandos mistos pouco estudados ao redordo mundo.
Abstract. – Reviews on mixed-species flocks usually examine information obtained in several regionsaround the world, not focusing the avifauna of particular ecosystems. This study aimed to review informa-tion on mixed-species flocks found in Cerrado. Publications (n = 28) mentioned 172 species participatingin mixed-species flocks in several vegetation physiognomies and regions. More detailed studies were con-ducted only in cerrado sensu stricto and “campo cerrado” vegetation. Flock members were found consum-ing fruits, nectar, seeds and insects. Flocks with up to 16 species and 40 individuals were recorded. Twelvespecies of six families were considered as nuclear species. In general, these are birds with contrasting color-ation that also occur in mono-specific groups in absence of mixed-species flocks. There is evidence thatpredation risk is a factor leading to the formation of mixed-species flocks in savanna vegetation. Also,predatory attacks by Falco species were noted. Mixed-species flocks are less frequent and have lower num-ber of species and individuals during the breeding season than in the non-breeding season. Assessmentsconducted at several spatial scales revealed that the participation in mixed-species flocks is a strategyadopted by a large part of the bird species richness found in Cerrado. Among suggestions for futureresearch is the consideration of the Cerrado environment (its mosaics of vegetation, its great extensionand its strongly seasonal climate) for studying aspects of the biology of mixed-species flocks poorly inves-tigated world-wide. Accepted 1 October 2006.
Mixed-species flocks of birds are related totwo major advantages gained by flock mem-bers – reduced predation risks and increasedforaging efficiency (Morse 1970, 1977; Dia-mond 1981, Powell 1985, Terborgh 1990,Thiollay 1999). According to these reviews,birds enhance these two benefits in severalways through the acquisition of informationand protection from other birds participatingin these inter-specific associations.
Mixed-species flocks have been found onall continents (Greig-Smith 1978, Bell 1980,Hutto 1987, Latta & Wunderle 1996, Thiollay1999, Hino 2000, Tellería et al. 2001). In theNeotropical region, they have been recordedin a diverse range of ecosystems or regions,such as Amazonia (Powell 1979, Gradwohl &Greenberg 1980, Munn 1985, Stotz 1993, Jul-lien & Thiollay 1998, 2001), the Atlantic For-est (Machado 1999, Develey & Peres 2000),Patagonia (Vuilleumier 1967) and Andes(Poulsen 1996, Bohórquez 2003).
In the Cerrado province, detailed investi-gations on mixed-species flocks have beenconducted since the 1980s (Silva 1980, Silva& Oniki 1988, Alves & Cavalcanti 1996,Ragusa-Netto 1997, 1999, 2000, 2002; Tubelis2004, Tubelis et al. 2006). Additionally, severalpublications concerning the avifauna of theCerrado provide brief information on theseflocks (Alves 1990, Willis & Oniki 1990,1991; Cavalcanti 1992, Marini 1992, Ridgely1994, Parker & Willis 1997, Sick 1997, Vas-concelos et al. 1999, Olmos & Boulhosa 2000,Pearce-Higgins 2000, Ragusa-Netto 2001, Sil-veira et al. 2001, Willis 2003, Lopes 2004).
Some aspects of its avifauna and land-scapes make Cerrado an interesting region forthe study of mixed-species flocks. First, adiverse range of habitat requirements shownby the numerous species found in Cerrado(Willis & Oniki 1990, Silva 1995, Macedo2002) favor the study of mixed-species flocks
in diverse situations. For example, these asso-ciations can be examined within vegetationpatches (Alves & Cavalcanti 1996, Ragusa-Netto 2000), as well as across boundariesformed by the juxtaposition of distinct habi-tats (Tubelis et al. 2006). Second, Cerrado ismarked by a strongly seasonal climate (Eiten1993), with consequent seasonal changes inresource availability (Oliveira 1998, Pinheiroet al. 2002). Thus, Cerrado might be an inter-esting region to examine seasonal formationof flocking – a question often investigatedworldwide (Morse 1970, Powell 1985). Third,Cerrado harbors a high species richness ofavian predators (Silva 1995), birds that mightlead to the formation of mixed-species flocksaround the world (Terborgh 1990, Thiollay1999).
Reviews on mixed-species flocks tend toconsider information obtained world-wide,not focusing particular ecosystems (Morse1970, 1977; Diamond 1981, Powell 1985, Ter-borgh 1990). On the other hand, two reviewson the biology of mixed-species flocks in theNeotropical region (Develey 2001, Jullien &Thiollay 2001) have emphasised the Amazonand the Atlantic Forest. Recently, socialaspects of the Cerrado’s avifauna have beenreviewed (Macedo 2002), but mono-specificgroups were the major focus. Thus, studies ofmixed-species flocks in Cerrado remain unre-viewed in details.
This study reviews the information onmixed-species flocks in the Cerrado province.First, I review data related to the Cerradoenvironment and its resources: 1) the geo-graphic distribution of records within thisprovince, 2) the use by mixed-species flocksof different vegetation physiognomies, 3) thefood items consumed by flock members.Additionally, I use the compiled records toinvestigate patterns of habitat use, socialityand habit (forest or non-forest species)among the nuclear species. Also, the compiledinformation was used to examine the propor-
77
MIXED-SPECIES FLOCKS IN CERRADO
tion of species participating in mixed-speciesflocks at three spatial scales in Cerrado. Thenext topics reviewed in this paper includethree major aspects of mixed-species flocksoften investigated around the world: the num-ber of birds and species found in mixed-spe-cies flocks, evidence of advantages gained byparticipation in mixed-species flocks, and theseasonal occurrence of these flocks. Also, Ipresent a general overview and provide sug-gestions for future research of mixed-speciesflocks in Cerrado.
METHODS
Cerrado. This South American vegetationprovince occupies about 2,000,000 km2 inBrazil, Bolivia and Paraguay (Fig. 1). Its vege-tation covers most of the highlands of centralBrazil and extends through peninsulas anddisjunct patches to Caatinga, Amazonia,Chaco and the Atlantic Forest (Eiten 1972,1993; Cavalcanti 1999a, Oliveira & Marquis2002).
Landscapes in the Cerrado are usually
FIG. 1. Geographical distribution of localities (black spots) in which mixed-species flocks of birds havebeen found in the Cerrado, South America.
78
TUBELIS
dominated by cerrado sensu lato, whichencompasses a range of vegetation physiog-nomies varying from open grasslands towoodlands and forests (Eiten 1972, 1993;Ribeiro & Walter 1998, Oliveira-Filho & Rat-ter 2002). Grasslands without shrubs or treesare called “campos limpos”, while “campossujos” are those with scattered shrubs andfew trees. Savanna vegetation with intermedi-ate and higher densities of shrubs and treesare called “campo cerrado” and cerrado sensustricto, respectively. “Cerradão” is forest, withhigher and denser trees and a much reducedground layer (Eiten 1972, 1993; Ribeiro &Walter 1998, Oliveira-Filho & Ratter 2002).
Besides cerrado sensu lato, other major veg-etation physiognomies are found in uplands,such as semi-deciduous forests, deciduousforests and rocky grasslands. Gallery forests,marshes, floodplain grasslands and “veredas”(wet grasslands with scattered shrubs andpalm trees) occur in valleys (Eiten 1972, 1993;Ribeiro & Walter 1998, Oliveira-Filho & Rat-ter 2002). Additional information on vegeta-tion and landscapes can be found in Sano &Almeida (1998), Cavalcanti (1999a) andOliveira & Marquis (2002).
Mixed-species flocks. They are inter-specific birdassociations that forage within a giving area,keeping the group cohesion even whenchanging directions; they are guided by spe-cies that often display alarm calls and sentinelbehavior – the nuclear species (Morse 1977,Powell 1985, Terborgh 1990, Stotz 1993, Jul-lien & Thiollay 1998). Although these flocksare often called mixed-species flocks (e.g.,Alves & Cavalcanti 1996, Parker & Willis1997, Silva et al. 1997, Tubelis et al. 2006), theyalso receive other denominations, such asmixed species bird flocks (e.g., Develey &Peres 2000), mixed flocks (e.g., Cavalcanti1992), mixed bands (e.g., Dubs 1992), birdmixed flocks (e.g., Ragusa-Netto 2000, 2002)and “bandos mistos” (e.g., Silva & Oniki
1988, Willis & Oniki 1990). All publicationsmentioning the occurrence of such flocks inCerrado (Fig. 1) were included in this review.Mixed-species flocks are very different fromant-following bird groups (Willis & Oniki1978) and bird aggregations in a given area asa result of localized food resources (e.g.,water, fruiting trees, swarming insects), whoserecords were not included in this review. Also,records in Cerrado of several species foragingtogether but with no comment on interac-tions in mixed-species flocks (e.g., Bagno &Rodrigues 1998, D’Angelo 2000) were notincluded. The nomenclature of bird speciesfollows Sigrist (2006).
Literature review. The literature review wasbased on the bibliography of Oniki & Willis(2002), the “Bibliografia Recente da Ornitolo-gia Brasileira” published regularly in “Arara-juba” (the former journal of the BrazilianSociety of Ornithology) until 2002, and theZoological Records.
RESULTS AND DISCUSSION
Regional distribution of records of mixed-species flocksin the Cerrado. A total of 28 publications hasreported the occurrence of mixed-speciesflocks in Cerrado. These associations havebeen found through most of the Cerradoregion (Fig. 1; Appendix 1). In central Brazil,published records come from the DistritoFederal (Silva 1980, Alves 1990, Cavalcanti1992, Marini 1992, Ridgely 1994, Alves &Cavalcanti 1996, Lopes 2004, Tubelis 2004,Tubelis et al. 2006) and Goiás (Ridgely 1994,Sick 1997, Tubelis 2004). Investigations inwest Cerrado areas have been conducted inBolivia (Pearce-Higgins 1996, 2000) and inBrazilian Mato Grosso (Silva & Oniki 1988,Willis & Oniki 1990, Dubs 1992, Parker &Willis 1997). Records in northern Cerrado arerestricted to the states Amapá (Silva et al.1997) and Piauí (Silveira et al. 2001). Those in
79
MIXED-SPECIES FLOCKS IN CERRADO
south-eastern Cerrado are from São Paulo(Willis & Oniki 1993, Ragusa-Netto 1997,1999; Olmos & Boulhosa 2000, Ragusa-Netto2000, 2001, 2002; Willis 2003) and MinasGerais (Willis & Oniki 1991, Silveira 1998,Vasconcelos et al. 1999). Further research isprobably going to increase the number oflocalities with records of these associations, asparticipation in mixed-species flocks appearsto be a strategy widespread in Cerrado.
Despite numerous records throughoutmost of its extension, studies comparingmixed-species flocks in different regions orlocalities are rare in Cerrado. Silva et al. (1997)reported that the species composition offlocks found in cerrado sensu stricto vegetationin northern Cerrado (Amapá) is similar tothose flocks recorded in central Brazil (Dis-trito Federal). Tubelis (2004) comparedmixed-species flocks of forest species foundin savannas adjacent to forests in CaldasNovas (Goiás) and the Distrito Federal.Flocks were guided by the same nuclear spe-cies in both localities. It was suggested thatdifferences in the number of species found inmixed-species flocks of both localities couldresult from the size of gallery forests and thevegetation structure of adjacent savannas(Tubelis 2004).
Use of vegetation physiognomies. Mixed-speciesflocks were recorded in a diverse range ofnative physiognomies (Appendix 1). Forestvegetation included gallery forest (Silva &Oniki 1988, Willis & Oniki 1991, Cavalcanti1992, Dubs 1992, Marini 1992, Tubelis 2004),dry forest (Willis & Oniki 1990, 1991) and“cerradão” (Olmos & Boulhosa 2000). Non-forest vegetation comprised cerrado sensustricto (Silva 1980, Silva & Oniki 1988, Alves1990, Willis & Oniki 1991, Ridgely 1994,Alves & Cavalcanti 1996, Silva et al. 1997, Sil-veira et al. 2001, Lopes 2004, Tubelis 2004,Tubelis et al. 2006), rocky “cerrado” (Tubelis2004), “campo cerrado” (Parker & Willis
1997, Vasconcelos et al. 1999, Ragusa-Netto1999, 2000, 2001, 2002), grasslands (Silva &Oniki 1988, Willis & Oniki 1990, 1993;Ridgely 1994, Silveira 1998, Vasconcelos et al.1999), marshes (Sick 1997) and wet grasslands(Pierce-Higgins 1996, 2000). Thus, morestudies have been done in open vegetationthan in forests. Mixed-species flocks have notyet been studied in rocky grasslands and“veredas” (Appendix 1).
Records in managed or exotic vegetationwere in orchards and plantations (Silva &Oniki 1988), pastures with scattered trees(Ragusa-Netto 1997) and in eucalypt planta-tions with recovering native understory (Willis2003) (Appendix 1). It was not determined ifflocks could establish whole territories withinpatches of such vegetation or if they wereusing them as additional foraging areas.
Despite the recording of mixed-speciesflocks in a wide range of native and exoticvegetation (Appendix 1), no studies madecomparisons between different habitats inCerrado. Thus, the biology of mixed-speciesflocks in different native physiognomies andthe response of mixed-species flocks to man-induced changes in native vegetation remainunknown in Cerrado.
Food items consumed by flock members. Studiesrecorded a diverse range of resources eaten bybirds joining mixed-species flocks. Fruitsincluded those of Melastomataceae (Silva &Oniki 1988), Ficus sp. (Willis & Oniki 1990),Miconia albicans, M. fallax, M. ferruginata, M.rubiginosa, Miconia sp., Byrsonima lancifolia andQualea parviflora (Tubelis 2004), and unidenti-fied species (Cavalcanti 1992, Alves & Caval-canti 1996). Inflorescences of Mabea sp.(Willis & Oniki 1990), Mabea fistulifera (Olmos& Boulhosa 2000), Caryocar brasiliense, Phora-dendron crassifolium, Qualea grandiflora andRoupala montana (Tubelis 2004) also weremajor resources used by flock members. Theconsumption of seeds by mixed-flock mem-
80
TUBELIS
bers was reported but the plant species werenot identified (Silva & Oniki 1988, Willis &Oniki 1990, Alves & Cavalcanti 1996, Silveira1998). Similarly, some studies mentionedflock members preying on unidentifiedinsects (Alves & Cavalcanti 1996, Silveira et al.2001, Tubelis 2004). In general, studies didnot mention the bird species feeding on theresources, nor the species (especially seedsand arthropods) consumed. Further, therewas no quantification of the food resourcesconsumed by mixed-flock members.
No studies examined the consumption of
food items by particular bird species when inand out mixed-species flocks, as examined inCosta Rican forests (Valburg 1992). Similarly,relationships between flock formation andfood availability (e.g., Poulsen 1996) also werenot examined in Cerrado. Investigations likethese could contribute to the understandingof the advantages of participation in mixed-species flocks in Cerrado.
Despite the apparent dominance of stud-ies reporting mixed-species flocks feeding onarthopods in the Neotropical region (e.g.,Powell 1979, Gradwohl & Greenberg 1980,
TABLE 1. Taxa considered as nuclear species of mixed-species flocks in Cerrado, with information ontheir habits, sociality when not in mixed-species flocks, use of vegetation physiognomies, and the source ofinformation. An asterisk (*) indicates species endemic to Cerrado (according to Cavalcanti (1999a) andSilva (1995)), while the sign (**) indicates that the species was considered as a probable nuclear species bythe respective authors.
Munn 1985, Hutto 1987, Machado 1999, Tell-ería et al. 2001, Bohórquez 2003), a consider-able proportion (5/8) of the studiesconducted in Cerrado mentioned the con-sumption of plant resources by flock mem-bers (see above). Thus, investigations aimingto examine the influence of food availabilityon flock formation and structure in Cerradoshould take into account plant and animalcomponents of the biota. This suggestion isreinforced by the recording of the consump-tion of insects, flowers and fruits by membersof particular mixed-species flocks (Willis &Oniki 1990, Tubelis 2004).
The nuclear species. According to major reviews,the nuclear species are those that guide andkeep the cohesion of mixed-species flocks(Morse 1970, Diamond 1981, Powell 1985,Terborgh 1990). Eleven out of 28 publica-tions mentioning the occurrence of mixed-species flocks in Cerrado identified theirnuclear species (Table 1). Twelve speciesbelonging to six families of the order Passeri-formes have been considered as nuclear spe-cies of mixed-species flocks in Cerrado. Mostof them are members of the sub-order Pas-seri. Among them, there was a predominanceof tanagers (Thraupidae) as nuclear species(Table 1).
The nuclear species in Cerrado are socialbird species (Table 1) as they occur in groupsof three or more individuals when not partici-pating in mixed-species flocks (Alves 1990,Ridgely 1994, Sick 1997, Ragusa-Netto 1997,2001; Lopes 2004). The only exception isLepidocolaptes angustirostris which occurs aloneor in pairs during most of the year (pers.observ.). However, the role of this wood-creeper as a nuclear species is secondary asNeothraupis fasciata and Suiriri suiriri are themajor nuclear species in cerrado sensu strictopatches (Silva 1980, Silva & Oniki 1988, Alves& Cavalcanti 1996). Similarly to these findingsof my review, Greig-Smith (1978) had noted
in savanna woodlands in Ghana that the mostfrequent species in mixed-species flocks occurin mono-specific groups when out of theseinter-specific associations.
Land birds found in Cerrado can be classi-fied as forest species and non-forest species(Lins 1994, Cavalcanti 1999b). Only five(42%) of the nuclear species are forest birds(Table 1). Two of them (Lanio versicolor andRamphocelus carbo) were detected by an investi-gation conducted in forests (Silva & Oniki1988). However, the other three species(Hemithraupis guira, Tangara cayana and Dacniscayana) were pointed out by a study conductedin savannas (Tubelis 2004) which examinedforest-savanna movements by forest birds.Thus, this relatively low number of forestnuclear species found until now in Cerradoresults, in part, from the low number ofdetailed studies in forests.
Ten of the 12 nuclear species have beenfound leading flocks in open vegetation(Table 1). This prevalence of nuclear speciesin savanna vegetation is result of a series ofdetailed observations in savannas (Silva 1980,Alves & Cavalcanti 1996, Ragusa-Netto 1999,2000, 2001, 2002; Tubelis 2004, Tubelis et al.2006). On the other hand, only five nuclearspecies were found in forests (Table 1).Although Tubelis (2004) and Tubelis et al.(2006) sampled savannas, mixed-speciesflocks also were seen in adjacent gallery for-ests. These numbers of nuclear species inboth savanna and forest vegetation stronglyreflect the numbers of detailed investigationsconducted in these physiognomies, as identifi-cation of nuclear species requires intensiveobservations or sampling.
Future detailed research of mixed-speciesflocks in Cerrado will likely reveal furthernuclear species, especially if conducted in for-est, grassland and other vegetation still poorlysampled. Several species are expected to berevealed as nuclear species. This becausenuclear species in Amazonia and the Atlantic
82
TUBELIS
Forest, such as Philydor rufus, Thamnomanes cae-sius, Habia rubica, Hemithraupis ruficapilla, Tricho-thraupis melanops, Basileuterus culicivorous andBasileuterus hypoleucus (see review in Develey2001), also might lead mixed-species flocks inCerrado. Further, some studies reported theoccurrence of flocks formed uniquely bySporophila species in Cerrado (Willis & Oniki1993, Ridgely 1994, Pearce-Higgins 1996,Sick 1997, Silveira 1998, Pearce-Higgins2000). Also, research in regions not yet stud-ied in detail might reveal other nuclear spe-cies.
Co-occurrence of forest and savanna birds in mixed-species flocks. Bird movement between adjacentvegetation physiognomies is a major phenom-enon in Cerrado landscapes (e.g., Cavalcanti1992, Lins 1994, Tubelis et al. 2004). As a con-sequence, a particular vegetation patch mightbe used by species typical of other patches.For example, forest bird species might useadjacent savanna vegetation, and savannabirds might use adjacent gallery forests. Thus,forest and non-forest (savanna) bird species(Lins 1994, Cavalcanti 1999b) might occur ina given mixed-species flock. Records ofmixed-species flocks with birds typical of dis-tinct landscape units (forest and savanna birdspecies) are shown below.
A forest woodpecker (Picumnus sp.) wasseen flocking with savanna birds in cerradosensu stricto (Silva 1980). Willis & Oniki (1990)recorded the forest species Turdus amaurochali-nus foraging in canopies with species of forestedges (Thraupis palmarum and T. sayaca). Also,T. sayaca, T. palmarum and Coryphospingus cucul-latus joined forest species in dry forests (Willis& Oniki 1990). The forest species Myiarchusswainsoni was found with savanna birds in cer-rado sensu stricto (Alves & Cavalcanti 1996).Olmos & Boulhosa (2000) mentioned Thrau-pis sayaca with the forest species Dacnis cayanaand Tangara cayana as common flock membersin “cerradão”. The forest species Veniliornis
passerinus was recorded with bird species ofopen physiognomies in “campo cerrado”(Ragusa-Netto 2000, 2002). Silveira et al.(2001) mentioned flocks formed by savannabirds and forest species (Piranga flava andHemithraupis guira) in cerrado sensu stricto. Inthis same habitat, six forest bird species(Veniliornis passerinus, Picumnus albosquamatus,Cyclarhis gujanensis, Serpophaga subcristata, Piayacayana and Colaptes melanochloros) were amongthe less frequent participants of savanna birdflocks (Ragusa-Netto 2002). Further, Elaeniaflavogaster, Elaenia chiriquensis, Camptostoma obso-letum, Thraupis palmarum, Lepidocolaptesangustirostris and Coryphospingus cucullatus joinedforest bird flocks in adjacent cerrado sensustricto patches at Distrito Federal (Tubelis2004). Thraupis sayaca also was found flockingwith forest species in similar patch-matrixmovements at Caldas Novas (Tubelis 2004).
With the exception of Tubelis (2004),these studies did not mention co-occurrenceof species typical of distinct landscape units(e.g., forest and savanna species) in the sameflock. Although these events have beenrecorded occasionally, they illustrate the influ-ence of habitat proximity on the species com-position of mixed-species flocks in Cerrado.Such flocks with species of distinct landscapeunits will likely be found more often if obser-vations are conducted close to boundariesbetween forest, savanna and/or vegetationassociated with aquatic environment.
Despite the recording of forest and non-forest (savanna) bird species in a given flock,similar records did not occur for nuclear spe-cies. In cerrado sensu stricto, detailed studiesidentified three forest nuclear species (Hemi-thraupis guira, Tangara cayana and Dacnis cayana)and three savanna nuclear species (Lepidoco-laptes angustirostris, Neothraupis fasciata andSuiriri suiriri) (Table 1). However, taxa of thesetwo groups of nuclear species were alwaysfound in distinct flocks. Thus, the guidance ofparticular mixed-species flocks by both forest
83
MIXED-SPECIES FLOCKS IN CERRADO
and savanna nuclear species has not beenreported. The reasons for this fact could beseen as venues of future research.
Participation of the bird species in mixed-speciesflocks. A total of 172 bird species has beenobserved in mixed-species flocks in Cerrado(Appendix 2). Species of the order Passeri-formes were dominant, as only 18 (10.5%) ofthe species are non-passerines. Families withhigher numbers of species in these bird asso-ciations were Tyrannidae, Thraupidae,Emberizidae, Thamnophilidae and Furnari-idae, with 32, 27, 24, 13 and 11 species,respectively (Appendix 2). The participationof the bird species richness in mixed-speciesflocks could be evaluated at three spatialscales.
An assessment at a macro scale could bedone, after discarding records outside theCerrado core area (Willis & Oniki 1990, 1991;Dubs 1992, Willis & Oniki 1993, Pearce-Hig-gins 1996, Ragusa-Netto 1997, Silva et al.1997, Vasconcelos et al. 1999, Olmos & Boul-hosa 2000, Pearce-Higgins 2000, Ragusa-Netto 2000, 2001, 2002; Willis 2003). A par-ticipation of 18% was observed, as 153 of the837 species recorded in the Cerrado core area(Silva 1995) have been found in mixed-speciesflocks (Appendix 2). When families (e.g.,Ardeidae, Columbidae, Caprimulgidae andApodidae) whose species have not beenrecorded in mixed-species flocks in Cerrado(Appendix 2) are not considered, the partici-pation of the avifauna in mixed-species flocksraises to 28% (153/550).
Two regions characterised by both exten-sive inventories and considerable knowledgeon mixed-species flocks permitted the assess-ment of participation of the avifauna at aregional level. In the Estação Ecológica Serradas Araras, 100 (43%) of the 233 speciesrecorded were found in mixed-species flocks(Silva & Oniki 1988). Also, of 355 bird speciesrecorded in four protected reserves in the
Distrito Federal (Braz & Cavalcanti 2001), 71(20%) have been found in mixed-speciesflocks (Silva 1980, Alves 1990, Cavalcanti1992, Marini 1992, Ridgely 1994, Alves &Cavalcanti 1996, Lopes 2004, Tubelis 2004,Tubelis et al. 2006). These proportions raisesto 60% (100/167) and 33% (71/216), respec-tively, after discarding families whose specieshave not yet been found in mixed-speciesflocks in Cerrado (Appendix 2).
Two studies examined the participation ofthe avifauna in mixed-species flocks at a locallevel. Of 38 species recorded in a cerrado sensustricto patch in the Distrito Federal, 14 (37%)were recorded in mixed-species flocks (Silva1980). Tubelis et al. (2006) showed that 50%of 66 forest bird species using savannas adja-cent to gallery forests at Distrito Federaljoined mixed-species flocks. These percent-ages raise to 64% and 82%, respectively, whenconsidering only those taxa whose familieshad species found in mixed-species flocks inCerrado (Appendix 2).
Although these proportions probably willchange with further inventories and studies ofmixed-species flocks, it can be noted that par-ticipation in mixed-species flocks is a majorstrategy adopted by a considerable part of theavifauna in the Cerrado. The high tendency ofthe Cerrado avifauna to flock in such inter-specific associations became more evidentwhen assessments considered only familieswhose species were noted in mixed-speciesflocks.
Also, participation of the avifauna inmixed-species flocks tended to be lower atgreater than at smaller spatial scales. Thismight result mainly of three factors. First,assessments conducted at greater scalesincluded avifaunas of a diverse range of habi-tats, while those conducted at local scalesincluded only birds found in cerrado sensustricto – a vegetation where birds often formmixed-species flocks (Appendix 1 and 2). Sec-ond, estimates at the Cerrado core area and
84
TUBELIS
regional scales involved detailed and non-detailed observations on mixed-speciesflocks, while estimates at local scales involvedonly detailed observations on these flocks.Third, characteristics of the avifauna might beinvolved, as tendencies to participate inmixed-species flocks vary among bird species.For example, participation of savanna birdspecies in mixed-species flocks was lower(64%) than that of forest bird species (82%)in cerrado sensu stricto patches at Distrito Fed-eral (Silva 1980, Tubelis et al. 2006). This dif-ference observed in savanna vegetation mightresult from the fact that savanna birds are intheir major habitat, while forest birds are in ahabitat that provide less protection than theirmajor (forest) habitat.
Flock size, flock species richness and flock formation.Information on the number of species foundper flock (flock species richness) was presentin more studies than on the number of birdsper flock (flock size) (Appendix 1), probablydue to difficulties in censusing individuals.Mixed-species flocks with higher number ofspecies were recorded in “campo cerrado”(Ragusa-Netto 2002), cerrado sensu stricto(Tubelis 2004) and gallery forests (Marini1992). On the other hand, flocks with lowspecies richness (e.g., those with only two orthree species) were found in a wide range ofvegetation physiognomies, such as grasslands,“campo cerrado”, cerrado sensu stricto, galleryforests, marshes and managed vegetation(Appendix 1). Larger flocks also wererecorded in several habitats, such as grass-lands, “campo cerrado”, cerrado sensu strictoand gallery forests (Appendix 1).
Besides vegetation and the richness ofbird communities, the level of detail of eachstudy might influence flock size and speciesrichness. For example, numerous studiesobserved flocks just briefly, thus likely con-tributing to reports of flocks with few (two orthree) species. On the other hand, flocks with
only two or three species can be frequentlyformed in Cerrado, as reported in detailedstudies (Silva 1980, Ragusa-Netto 1997, Tube-lis 2004).
Four studies in patches of cerrado sensustricto (Silva 1980, Alves & Cavalcanti 1996)and “campo cerrado” (Ragusa-Netto 2000,2002) dealt with the relationships betweenspecies richness and bird abundance inmixed-species flocks. All of them found apositive correlation between flock size andflock species richness. It was suggested thatflock size increases due to the entry of newspecies, as usually only a group of each spe-cies is present in each flock. Interestingly, thenumber of species and the number of birds ina mixed flock can vary according to thenuclear species of these associations (Ragusa-Netto 2002).
Another factor causing variation in mixed-species flocks is seasonality. Detailed investi-gations that examined the seasonal occur-rence of mixed-species flocks were conductedin the Distrito Federal. In a cerrado sensu strictopatch, Silva (1980) made observations duringa 12-month period and recorded flocks ofsavanna bird species only between Januaryand August. Contrastingly, similar mixed-spe-cies flocks were recorded in all months of ayear period in this same vegetation (Alves1990, Alves & Cavalcanti 1996). An investiga-tion carried out in gallery forests recordedmixed-species flocks of forest bird species inbasically all months of a year (Marini 1992).Further, forest flocks left gallery forests toforage in adjacent savannas throughout theyear (Tubelis 2004, Tubelis et al. 2006). Thesestudies sampling flocks year round suggestthat mixed-species flocks are formed duringboth the breeding and non-breeding seasonsof forest and savanna bird species in centralCerrado. Probably, their absence during a 4-month period (Silva 1980) might haveresulted from a relatively lower samplingeffort done in that period.
85
MIXED-SPECIES FLOCKS IN CERRADO
Some of these studies also examined sea-sonal variation in the frequency of occurrenceof mixed-species flocks. Considering compa-rable sampling efforts, Silva (1980) contrastedthe recording of 14 flocks in March (non-breeding period) with that of only three inAugust (early breeding period). In the samearea, the frequency of occurrence of flockswas higher in the early dry season, when birdswere not reproducing (Alves 1990, Alves &Cavalcanti 1996). Similarly, forest flocks wererecorded more frequently in adjacent savan-nas during the non-breeding than the breed-ing period of species (Tubelis 2004, Tubelis etal. 2006). Further investigations were consid-ered necessary to verify the roles of the higherabundance of food resources and of thebreeding activities in the lower frequency ofoccurrence of mixed-species flocks during thebreeding period (Alves & Cavalcanti 1996,Tubelis 2004).
Studies that have not conducted samplesyear round could be divided in two major cat-egories. First, were investigations that con-ducted detailed observations for severalmonths, during the nonbreeding period ofspecies. These studies examined flocksbetween June and September (Ragusa-Netto1997) and from March to September (Ragusa-Netto 1999, 2000, 2002), probably to achieveinformation when flocks are more frequent.The second category included numerousstudies whose major objectives focuseddiverse aspects of the avifauna, other than theseasonal formation of mixed-species flocks.Thus, they provided occasional findings ofthese mixed-species flocks, usually reportingthe month in which they were found (Appen-dix 1). Considering all the information on sea-sonal occurrence revealed by theseinvestigations, it can be noted that mixed-spe-cies flocks have been found in all months ofthe year in Cerrado (Appendix 1). This is truefor flocks of forest species as well as for thoseof open-habitat birds. Thus, the compilation
of these scattered records confirms the factthat mixed-species flocks can be found yearround in Cerrado, as has been reported forNeotropical forests (e.g., Powell 1979, Jullien& Thiollay 1998, Develey & Peres 2000).
Evidence of advantages of participation in mixed-spe-cies flocks. The hypothesis of increased forag-ing efficiency as an advantage gained bymembers of mixed-species flocks (Morse1977, Diamond 1981, Powell 1985, Terborgh1990) has been tested by only one study inCerrado. In pastures with scattered trees, thepecking rate of a species (Furnarius rufus) wassignificantly higher in mixed-species flocksthan when foraging alone or in mono-specificgroups (Ragusa-Netto 1997). It was suggestedthat the vigilance provided by sentinels of thenuclear species (Mimus saturninus) allowed F.rufus to spend more time with feeding activi-ties.
The anti-predatory hypothesis in the for-mation of mixed-species flocks (Morse 1977,Diamond 1981, Powell 1985, Terborgh 1990)has been tested by more studies. In a cerradosensu stricto patch of the central Cerrado, theintensity of sentinel behavior by a nuclear spe-cies (Neothraupis fasciata) was significantlylower in mixed-species flocks than in mono-specific groups (Alves & Cavalcanti 1996).This pattern was consistent with their predic-tion that enhanced protection gained with thepresence of other species leads to lower vigi-lance by this nuclear species. In a “campo cer-rado” patch in eastern Cerrado, the time spentin sentinel activities by a nuclear species(Cypsnagra hirundinacea) was positively relatedto the rate of encounters between mixed-spe-cies flocks and avian predators (Ragusa-Netto2000). In central Cerrado, the proportion offorest bird species participating in mixed-spe-cies flocks when foraging in adjacent savannas(cerrado sensu stricto) was higher at greater dis-tances from gallery forests (Tubelis et al.2006). This tendency of being in mixed-spe-
86
TUBELIS
cies flocks at more distant savanna vegetationwas interpreted as a reluctance to forage aloneor in mono-specific groups at greater dis-tances from cover (forest). Thus, participationin mixed-species flocks was considered astrategy adopted by forest species to reducepredation risk in less protective savanna vege-tation (Tubelis et al. 2006).
Further evidence of predation risk toflock members has been obtained by observa-tions of attacks by four species of avian pred-ators in open habitats. Fourteen unsuccessfulpredatory attacks (ten by Falco sparverius andfour by F. femoralis) on mixed-species flockswere recorded in a “campo cerrado” patch ineastern Cerrado (Ragusa-Netto 1999, 2000,2001, 2002). Near this same study area, I.Sazima (pers. com.) observed a successfulattack by F. femoralis (Ragusa-Netto 1997).Two studies provided some data on the fre-quency of occurrence of such predatoryattacks in Cerrado. Silva (1980) informed thatno predator attacks on mixed-species flockswere recorded during 444 h of observation incerrado sensu stricto. One predatory attack wasrecorded every 30 h, on average, in “campocerrado” patches (Ragusa-Netto 2002).
Additionally, there are reports of flockmembers escaping to cover in response to thepresence of avian predators in three openvegetation physiognomies. These eventsinvolved Falco rufigularis and F. femoralis ingrasslands (Willis & Oniki 1990), F. femoralisand F. sparverius in “campo cerrado” (Ragusa-Netto 1999, 2000, 2001, 2002) and Cyanocoraxcristatellus in cerrado sensu stricto (Silva 1980).These movements into cover in response topredator approaches were induced by alarmcalls displayed by sentinels of nuclear species,which included Mimus saturninus, Neothraupisfasciata, Cypsnagra hirundinacea and Saltator atri-collis in “campo cerrado” (Ragusa-Netto 1999,2000, 2001, 2002). In this habitat, the timespent with sentinel activities by nuclear spe-cies was proportional to the rate of encoun-
ters with avian predators (Ragusa-Netto2002). This study recorded eight species ofavian predators threatening mixed-speciesflocks: Buteo albicaudatus, Rupornis magnirostris,Elanus leucurus, Herpetotheres cachinnans, Milvagochimachima, Falco femoralis, F. sparverius and Rhi-noptynx clamator (Ragusa-Netto 2002). Despiteevidence of threat by birds, no studies men-tioned non-avian predators threateningmixed-species flocks in Cerrado.
The compilation of information broughtby studies involving advantages of participa-tion in mixed-species flocks leads to threemajor conclusions. First, predation risk is animportant factor influencing bird communi-ties in Cerrado’s woodland savannas. Indirectevidence for this fact was brought by detailedstudies involving distinct approaches (e.g.,Alves & Cavalcanti 1996, Ragusa-Netto 2002,Tubelis et al. 2006). Additionally, ten bird spe-cies of four families (Accipitridae, Falconidae,Strigidae and Corvidae) have been notedthreatening or attacking members of mixed-species flocks in “campo cerrado” and cer-rado sensu stricto. Second, sentinel behaviorand alarm calls appear to be frequent anti-predatory mechanisms shown by the nuclearspecies of mixed-species flocks inhabitingsavanna woodlands in Cerrado (e.g., Alves& Cavalcanti 1996, Ragusa-Netto 2000,2002). Third, advantages gained with the for-mation of mixed-species flocks around theworld – increased foraging efficiency andreduced predation risks – were noted in Cer-rado. Also, a study at forest-savanna bound-aries argued that an additional advantagemight benefit members of mixed-speciesflocks. Tubelis et al. (2006) have shown thatthe formation of mixed-species flocks playedan important role in promoting the use ofadjacent savannas by forest birds. Based onthis fact, they pointed out a novel advantagegained with participation in these inter-spe-cific associations – greater use of adjacentvegetation patches.
87
MIXED-SPECIES FLOCKS IN CERRADO
Overview. Of 28 publications reporting theoccurrence of mixed-species flocks in Cer-rado, only nine (32%) stated in the objectivestheir aim of studying these bird associations(Appendix 1). Of these, Silva & Oniki (1988)elaborated a list of species participating or notin mixed-species flocks in Serra das Araras.The other eight studies involved detailed sam-pling schemes and were carried out at DistritoFederal (Silva 1980, Alves 1990, Alves & Cav-alcanti 1996, Tubelis 2004, Tubelis et al. 2006)and in Brotas-SP (Ragusa-Netto 1997, 2000,2002). Therefore, detailed investigations onmixed-species flocks were conducted exclu-sively in the central and southeastern regionsof Cerrado, despite the recording of theseflocks through most of its extension.
Of these eight detailed investigations, fiveexamined mixed-species flocks inhabiting cer-rado sensu stricto vegetation in the Distrito Fed-eral, while three investigated flocks in “campocerrado” vegetation in Brotas (Appendix 1).This fact leads to two major conclusions.First, potential effects of regional variationand habitat on flock structure and biologywould be confounded in eventual compari-sons of results obtained by these two groupsof detailed studies. Second, aspects of thebiology of mixed-species flocks have beenexamined in details only in savanna wood-lands, thus remaining relatively poorly investi-gated in forests, grasslands (“campo limpo”and “campo sujo”) and other native vegeta-tion.
Aspects of the biology of mixed-speciesflocks more often investigated in details wereadvantages of flock formation (Silva 1980,Alves 1990, Alves & Cavalcanti 1996, Ragusa-Netto 1997, 2000, 2002; Tubelis et al. 2006)and seasonal variation in flock structure andformation (Silva 1980, Alves & Cavalcanti1996, Tubelis 2004, Tubelis et al. 2006). Theyare aspects usually examined in major reviewsof mixed-species flocks (Morse 1970, 1977;Diamond 1981, Powell 1985, Terborgh 1990).
On the other hand, the diet and feeding habitsof flock members, regional variation in mixed-species flocks, and the use of distinct vegeta-tion patches by flocks are among aspects oftheir biology not yet examined in details inCerrado. Information on these poorly studiedaspects results of eventual records reported instudies focusing aspects of the avifauna otherthan the biology of mixed-species flocks.
All publications reviewed in this studyprovided information on the locality/region,and habitat where mixed-species flocks werefound (Appendix 1). Information on theperiod of the year in which mixed-speciesflocks were recorded was provided less often,only in 24 (86%) of the 28 publicationsassessed. Only four studies (Dubs 1992,Ridgely 1994, Ragusa-Netto 2001, Willis2003) have not indicated the period of theyear in which flocks were found (Appendix 1).Information on the number of species foundper flock was provided in only 15 (54%) ofthe 28 publications included in this review.Only six publications (21%) provided infor-mation on the number of birds found perflock. Information on the period of flockoccurrence, and on the species richness andbird numbers found per flock was publishedsince the earliest to the most recent publica-tions involving records of mixed-speciesflocks in Cerrado (Appendix 1). Thus, no ten-dencies for the study of particular aspect ofthe biology of mixed-species flocks appear tohave occurred along the history of investiga-tions on these bird associations in Cerrado.The same is true for studies with detailedsampling schemes and those involving onlybrief information (Appendix 1).
Several records of mixed-species flocks inCerrado have been presented only briefly in19 (68%) non-detailed studies on these associ-ations. These studies focused, for example,inventories of species (e.g., Willis & Oniki1990, Pearce-Higgins 2000), bird-plant inter-actions (e.g., Olmos & Boulhosa 2000) or
88
TUBELIS
aspects of the biology of particular taxa (e.g.,Marini 1992, Parker & Willis 1997, Vasconce-los et al. 1999, Lopes 2004). Despite notanswering complex questions on mixed-spe-cies flocks, these studies contributed to theunderstanding of aspects such as the occur-rence of mixed-species flocks within Cerrado,the use of vegetation physiognomies, theperiod of their occurrence, and the speciesrichness and bird abundance found per flock(Appendix 1). Further, 41 (24%) of the birdspecies found in mixed-species flocks in Cer-rado have been recorded exclusively by these19 studies that have not shown detailed dataon these flocks (Appendix 2). Thus, briefcomments on the occurrence and biology ofmixed-species flocks would be welcome infuture studies on the Cerrado avifauna.
Of the 28 publications, only eight (Silva1980, Alves 1990, Alves & Cavalcanti 1996,Ragusa-Netto 1997, 2000, 2002; Tubelis 2004,Tubelis et al. 2006) provided information onthe criteria used to identify nuclear species.Also, these eight publications were the onlystudies to define when a bird was participat-ing or not in a mixed-species flock. There-fore, near three quarters of the 28 studies didnot make clear statements on the criteria usedto identify mixed-species flocks and theirnuclear species. Despite this fact, I consideredtheir data because I assumed that authors ofthese 20 publications were aware of defini-tions of nuclear species and mixed-speciesflocks. However, I suggest that future publi-cations cite references (e.g., Powell 1985,Stotz 1993) or mention criteria regarding def-initions of mixed-species flocks and nuclearspecies.
Considering the evidence of predationthreat to bird species brought by detailedinvestigations in savannas, the importance ofnuclear species for their survival becamemore evident. Thus, bird species functioningas nuclear species of flocks in native habitatsshould receive more research and conserva-
tion attention because of their role in assistingnumerous bird species.
Some interesting aspects of the biology ofmixed-species flocks were not included in thisreview due to the low number of studies.Among them were the use of vegetation strataby species (Silva 1980, Alves & Cavalcanti1996), flock movement (Silva 1980), territoryof flocks (Alves & Cavalcanti 1996), inter-spe-cific aggressions (Silva 1980) and the role ofdifferent species in the vigilance towardsavian raptors (Ragusa-Netto 2002). Theirachievements also should receive the atten-tion of investigators of mixed-species flocks.
Suggestions for future research. Numerous inter-esting suggestions for future investigations ofthe biology of mixed-species flocks have beenproposed recently by Greenberg (2000). Toavoid being repetitive, my suggestions focuson the Cerrado environment and knowledgeof its avifauna. My suggestions for futureresearch on mixed-species flocks in Cerradoare: (1) Identification of nuclear species lead-ing mixed-species flocks in different habitatsand regions; (2) improvement of knowledgeof major food items consumed by flock mem-bers, especially for the nuclear species; (3)patch-matrix and inter-patch movements bymixed-species flocks examined in a diverserange of boundaries and situations in thediverse range of land mosaics found in Cer-rado (this kind of research will likely bringnew insights on the biology of mixed-speciesflocks, as most research conducted world-wide has involved flocks foraging within asingle vegetation patch); (4) examination offlock responses to habitat and landscapechanges essential for their appropriate conser-vation in protected reserves and in human-modified landscapes; (5) co-occurrence ofbirds typical of distinct landscape units in agiven mixed-species flock as indicator ofaspects of flock cohesion still poorly investi-gated world-wide; (6) basic questions regard-
89
MIXED-SPECIES FLOCKS IN CERRADO
ing the biology of mixed-species flocks, suchas their seasonal occurrence, their size, theirspecies composition, their home range, andthe advantages gained with flock formationaddressed in Cerrado, especially in forests,grasslands, marshes and other less sampledvegetation; (7) given the huge dimensions ofCerrado, studies comparing regional variationin flock structure and guidance providing rele-vant information on factors involved in theirorganisation; (8) during surveys, observerscould try to distinguish two situations: birdsin mixed-species flocks and birds outsidethese associations. Acquisition of these twotypes of information in a given sample orstudy site would permit the examination ofthe propensity of species to join mixed-spe-cies flocks. This aspect is still poorly investi-gated world-wide, probably due to difficultiesin obtaining data, but might be possible instructurally less complex vegetation (e.g.,grassland and savanna).
Overall, researchers could consideraspects of the Cerrado environment (e.g., itsgreat extension, its patchy environment, andits strongly seasonal climate) to study aspectsof the biology of mixed-species flocks poorlyinvestigated world-wide.
ACKNOWLEDGMENTS
I thank many people that provided commentsand suggestions since the early phases of thisreview: R. B. Cavalcanti, A. Cowling, P. F.Develey, C. Donnelly, R. Gabon-Lima, S.Gilmore, P. MacDonald, R. H. F. Macedo, P.Olsen and M. N. Sato. I also thank ClaudinhaMelo, M. Â. Marini, Y. Oniki, J. W. Pearce-Higgins and A. Tubelis for helping with refer-ences. T. C. Anacleto, R. McNeil, Y. Oniki, J.Ragusa-Netto, P. Warden-Hutton, E. O. Willisand two anonymous reviewers greatlyimproved this manuscript. I was granted fel-lowships from CNPQ (process no. 200102-00/1) and FAPESP (process no. 05/00773-3)
during the elaboration of this manuscript.
REFERENCES
Alves, M. A. S. 1990. Social system and helpingbehavior in the White-banded Tanager (Neo-thraupis fasciata). Condor 92: 470–474.
Alves, M. A. S., & R. B. Cavalcanti. 1996. Sentinelbehavior, seasonality, and structure of birdflocks in a Brazilian savanna. Ornitol. Neotrop.7: 43–51.
Bagno, M. A., & F. H. G. Rodrigues. 1998. Novosregistros de aves para o Estado de Goiás, Brasil.Ararajuba 6: 64–65.
Bell, H. L. 1980. Composition and seasonality ofmixed-species feeding flocks of insectivores inthe Australian Capital Territory. Emu 80: 227–232.
Bohórquez, C. I. 2003. Mixed-species bird flocks ina montane cloud forest of Colombia. Ornitol.Neotrop. 14: 67–78.
Braz, V. S., & R. B. Cavalcanti. 2001. A representa-tividade de áreas protegidas do Distrito Federalna conservação da avifauna do Cerrado. Arara-juba 9: 61–69.
Cavalcanti, R. B. 1992. The importance of forestedges in the ecology of open country cerradobirds. Pp. 513–518 in Furley, P. A., J. Proctor, &J. A. Ratter (eds.). Nature and dynamics of for-est-savanna boundaries. Chapman & Hall, Lon-don, UK.
Cavalcanti, R. B. 1999a. Ações prioritárias para aconservação da biodiversidade do Cerrado ePantanal. Available on: http://www.bdt.fat.org.br/workshop/cerrado/br
Cavalcanti, R. B. 1999b. Bird species richness andconservation in the Cerrado region of centralBrazil. Stud. Avian Biol. 19: 244–249.
D’Angelo Neto, S. 2000. Ocorrência de Molothrusrufoaxillaris (Passeriformes: Emberizidae) naregião de Francisco Sá, Norte de Minas Gerais.Melopsittacus 3: 134–136.
Develey, P. F. 2001. Os bandos mistos de aves nasflorestas neotropicais. Pp. 39–47 in Albuquer-que, J. L. B., J. F. Candido Jr, F. C. Straube, & A.L. Roos(eds.). Ornitologia e conservação. Daciência às estratégias. Editora Unisul, Tubarão,Brazil.
Develey, P. F., & C. A. Peres. 2000. Resource sea-
90
TUBELIS
sonality and the structure of mixed species birdflocks in a coastal Atlantic forest of southeast-ern Brazil. J. Trop. Ecol. 16: 33–53.
Develey, P. F., & P. C. Stouffer. 2001. Effects ofroads on movements by understory birds inmixed-species flocks in central AmazonianBrazil. Conserv. Biol. 15: 1416–1422.
Diamond, J. M. 1981. Mixed-species foraginggroups. Nature 292: 408–409.
Dubs, B. 1992. Birds of southwestern Brazil. Cata-logue and guide to the birds of the Pantanal ofMato Grosso and its border areas. Bretona,Küsnacht, Switzerland.
Eiten, G. 1972. The cerrado vegetation of Brazil.Bot. Rev. 38: 205–341.
Eiten, G. 1993. Vegetação do Cerrado. Pp. 17–73in Pinto, M. N. (ed.). Cerrado: caracterização,ocupação e perspectivas. Editora Univ. Brasília,Brasília, Brazil.
Greig-Smith, P. W. 1978. The formation, structureand function of mixed-species insectivorousbird flocks in west African savanna woodland.Ibis 120: 284–297.
Gradwohl, J. R., & R. Greenberg. 1980. The for-mation of antwren flocks on Barro ColoradoIsland, Panama. Auk 97: 385–395.
Greenberg, R. 2000. Birds of many feathers: theformation and structure of mixed speciesflocks of forest birds. Pp. 521–558 in Boinski,S., & P. A. Garber (eds.). Birds on the move.Univ. Chicago Press, Chicago, Illinois.
Hino, T. 2000. Intraspecific differences in benefitsfrom feeding in mixed-species flocks. J. AvianBiol. 31: 441–446.
Hutto, R. L. 1987. A description of mixed-speciesinsectivorous bird flocks in western Mexico.Condor 89: 282–292.
Jullien, M., & J. M. Thiollay. 1998. Multi-speciesterritoriality and dynamic of Neotropical forestunderstorey bird flocks. J. Anim. Ecol. 67, 227–252.
Jullien, M., & J. M. Thiollay. 2001. The adaptive sig-nificance of flocking in tropical understoreyforest birds: the field evidence. Pp. 142–159 inBongers, F., P. Charles-Dominique, P. Forget, &M. Théry (eds.). Dynamics and plant-animalinteractions in a Neotropical rainforest. KluwerAcademic Publishers, Dordrecht, Germany.
Latta, S. C., & J. M. Wunderle Jr. 1996. Ecological
relationships of two todies in Hispaniola:effects of habitat and flocking. Condor 98:769–799.
Lins, L. V. 1994. O papel da mata ciliar na estrutu-ração de uma comunidade de aves do cerrado(Brasília, DF). M.Sc. thesis, Univ. of Brasília,Brasília, Brazil.
Lopes, L. E. 2004. Biologia reprodutiva de Suiririaffinis e Suiriri islerorum (Aves: Tyrannidae) noCerrado do Brasil central. M.Sc. Thesis, Univ.of Brasília, Brasília, Brazil.
Macedo, R. H. F. 2002. The avifauna: ecology, bio-geography, and behavior. Pp. 242–263 inOliveira, P. S., & R. J. Marquis (eds.). The Cer-rados of Brazil. Columbia Univ. Press, NewYork, New York.
Machado, C. G. 1999. A composição dos bandosmistos de aves na Mata Atlântica da Serra deParanapiacaba, no sudeste Brasileiro. Rev. Bra-sil. Biol. 59: 75–85.
Maldonado-Coelho, M., & M. Â. Marini. 2004.Mixed-species bird flocks from Brazilian Atlan-tic forest: the effects of forest fragmentationand seasonality on their size, richness and sta-bility. Biol. Conserv. 116: 19–26.
Marini, M. Â. 1992. Foraging behavior and diet ofthe Helmeted Manakin. Condor 94: 151–158.
Morse, D. H. 1970. Ecological aspects of somemixed-species foraging flocks of birds. Ecol.Monogr. 40: 119–168.
Morse, D. H. 1977. Feeding behavior and predatoravoidance in heterospecific groups. BioScience27: 332–339.
Munn, C. A. 1985. Permanent canopy and under-story flocks in Amazonia: species compositionand population density. Ornith. Monogr. 36:683–711.
Oliveira-Filho, A. T., & J. A. Ratter. 2002. Vegeta-tion physiognomies and woody flora of theCerrado biome. Pp. 91–120 in Oliveira, P. S., &R. J. Marquis (eds.). The Cerrados of Brazil.Ecology and natural history of a Neotropicalsavanna. Columbia Univ. Press, New York,New York.
Oliveira, P. E. 1998. Fenologia e biologia reprodu-tiva das espécies de Cerrado. Pp. 169–192 inSano, S. M., & S. P. Almeida (eds.). Cerrado.Ambiente e flora. Embrapa, Planaltina, Brazil.
Oliveira, P. S., & R. J. Marquis. 2002. The Cerrados
91
MIXED-SPECIES FLOCKS IN CERRADO
of Brazil. Ecology and natural history of a Neo-tropical savanna. Columbia Univ. Press, NewYork, New York.
Olmos, F., & R. L. P. Boulhosa. 2000. A meeting ofopportunists: birds and other visitors to Mabeafistulifera (Euphorbiaceae) inflorescences. Arara-juba 8: 93–98.
Oniki, Y., & E. O. Willis. 2002. Bibliography ofBrazilian birds: 1500–2002. Divisa, Rio Claro,Brazil.
Parker, T. A., & E. O. Willis. 1997. Notes on threetiny grassland flycatchers, with comments onthe disappearance of South American fire-diversified savannas. Ornith. Monogr. 48: 549–555.
Pearce-Higgins, J. W. 1996. Seedeaters in the NoelKempff Mercado National Park, Bolivia. Cotin-ga 5: 69–71.
Pearce-Higgins, J. W. 2000. The avian communitystructure of a Bolivian savanna on the edge ofthe Cerrado system. Hornero 15: 77–84.
Pinheiro, F., I. R. Diniz, D. Coelho, & M. P. S. Ban-deira. 2002. Seasonal pattern of insect abun-dance in the Brazilian Cerrado. Austral Ecol.27: 132–136.
Poulsen, B. O. 1996. Relationships between fre-quency of mixed-species flocks, weather andinsect activity in a montane cloud forest inEcuador. Ibis 138: 466–470.
Powell, G. V. N. 1979. Structure and dynamics ofinterspecific flocks in a Neotropical mid-eleva-tion forest. Auk 96: 375–390.
Powell, G. V. N. 1985. Sociobiology and adaptativesignificance of interspecific foraging flocks inthe Neotropics. Neotrop. Ornithol. 36: 713–732.
Ragusa-Netto, J. 1997. Evidence for the possibleadvantage of heterospecific social foraging inFurnarius rufus (Passeriformes: Furnariidae).Ararajuba 5: 233–235.
Ragusa-Netto, J. 1999. Sócio-ecologia dos bandosmistos de aves em campo-cerrado (Brotas, SP).Ph.D. thesis, Univ. Estadual Paulista, Rio Claro,Brazil.
Ragusa-Netto, J. 2000. Raptors and ”campo-cer-rado” bird mixed flocks led by Cypsnagra hirundi-nacea (Emberizidae: Thraupinae). Rev. Brasil.Biol. 60: 461–467.
Ragusa-Netto, J. 2001. Sentinels in Saltator atricollis
(Passeriformes: Emberizidae). Rev. Brasil. Biol.61: 317–322.
Ragusa-Netto, J. 2002. Vigilance towards raptors bynuclear species in bird mixed flocks in a Brazil-ian savannah. Stud. Neotrop. Fauna E. 37: 219–226.
Ribeiro, J. F., & B. M. T. Walter. 1998. Fitofisiono-mias do bioma Cerrado. Pp. 89–166 in Sano, S.M., & S. P. Almeida (eds.). Cerrado. Ambiente eflora. Embrapa, Planaltina, Brazil.
Ridgely, R. S. 1994. The birds of South America.The Suboscine Passerines. Univ. of Texas Press,Austin, Texas.
Sano, S. M., & S. P. Almeida. 1998. Cerrado. Ambi-ente e flora. Embrapa, Planaltina, Brazil.
Sick, H. 1997. Ornitologia brasileira. Editora NovaFronteira, Rio de Janeiro, Brazil.
Sigrist, T. 2006. Aves do Brasil. Uma visão artística.Birds of Brazil. An artistic view. Fosfertil, SãoPaulo, Brazil.
Silva, E. M. D. 1980. Composição e comporta-mento de grupos heteroespecíficos de aves emárea de cerrado, no Distrito Federal. M.Sc. the-sis, Univ. of Brasília, Brasília, Brazil.
Silva, J. M. C. 1995. Birds of the Cerrado region –South America. Steenstrupia 21: 69–92
Silva, J. M. C., & Y. Oniki. 1988. Lista preliminar daavifauna da Estação Ecológica Serra das Araras,Mato Grosso, Brasil. Bol. Mus. Para. E. GoeldiSer. Zool. 4: 123–143.
Silva, J. M. C., D. C. Oren, J. C. Roma, & L. M. P.Henriques. 1997. Composition and distributionpatterns of the avifauna of an Amazonianupland savanna, Amapá, Brazil. Ornithol.Monogr. 48: 743–762.
Silveira, L. F. 1998. The birds of Serra da CanastraNational Park and adjacent areas, Minas Gerais,Brazil. Cotinga 10: 55–63.
Silveira, L. F., A. Calonge-Méndez, & G. R. R.Brito. 2001. Range extensions and new recordsfor birds in Piaui state, Brazil. Int. J. Ornithol.4: 219–224.
Stotz, D. F. 1993. Geographic variation in speciescomposition of mixed species flocks in lowlandhumid forests in Brazil. Pap. Avul. Zool. 38:61–75.
Tellería, J. L., E. Virgós, R. Carbonell, J. Pérez-Tris,& T. Santos. 2001. Behavioural responses tochanging landscapes: flock structure and anti-
92
TUBELIS
predator strategies of tits wintering in frag-mented forests. Oikos 95: 253–264.
Terborgh, J. 1990. Mixed species and polyspecificassociations: costs and benefits of mixedgroups to birds and monkeys. Am. J. Primat.21: 87–100.
Thiollay, J. M. 1999. Frequency of mixed speciesflocking in tropical forest birds and correlatesof predation risk: an intertropical comparison.J. Avian Biol. 30: 282–294.
Tubelis, D. P. 2004. Species composition and sea-sonal occurrence of mixed-species flocks offorest birds in savannas in central Cerrado, Bra-zil. Ararajuba 12: 105–111.
Tubelis, D. P., A. Cowling, & C. Donnelly. 2004.Landscape supplementation in adjacent savan-nas and its implications for the design of corri-dors for forest birds in the central Cerrado,Brazil. Biol. Conserv. 118: 353–364.
Tubelis, D. P., A. Cowling, & C. Donnelly. 2006.The role of mixed-species flocks in the use ofadjacent savannas by forest birds in centralCerrado, Brazil. Austral Ecol. 31: 38–45.
Valburg, L. K. 1992. Flocking and frugivory – theeffect of social groupings on resource use inthe common bush-tanager. Condor 94: 358–
363.Vasconcelos, M. F., M. Maldonado-Coelho, & R.
Durães. 1999. Notas sobre algumas espécies deaves ameaçadas e pouco conhecidas da porçãomeridional da Cadeia do Espinhaço, MinasGerais. Melopsittacus 2: 44–50.
Vuilleumier, F. 1967. Mixed-species flocks in Pat-agonian forests with remarks on interspeciesflock formation. Condor 69: 400–404.
Willis, E. O. 2003. Birds of a Eucalyptus woodlotin interior São Paulo. Brazil. J. Biol. 63: 141–158.
Willis, E. O., & Y. Oniki. 1978. Birds and armyants. Ann. Rev. Ecol. Syst. 9: 243–263.
Willis, E. O., & Y. Oniki. 1990. Levantamento pre-liminar das aves de inverno em dez áreas dosudoeste de Mato Grosso, Brasil. Ararajuba 1:19–38.
Willis, E. O., & Y. Oniki. 1991. Avifaunal transectsacross the open zones of northern MinasGerais, Brazil. Ararajuba 2: 41–58.
Willis, E. O., & Y. Oniki. 1993. New and recon-firmed birds from the State of São Paulo, Bra-zil, with notes on disappearing species. Bull. Br.Ornithol. Club 113: 23–34.
93
MIX
ED
-SPECIE
S FLOCK
S IN CE
RRAD
OAPPENDIX 1. Studies that recorded mixed-species flocks in Cerrado, with information on their study area, on the vegetation physiognomiesused by flocks, on their seasonal occurrence, and on the species richness and bird abundance found per flock. Studies were grouped according to the coun-try where they were conducted and then listed in chronological order. Questions examined by studies involving detailed sampling schemes also were pro-vided.
Country/Sources
Study area Vegetation Seasonal occurrence Numbers of species Number of birds Question
Brazil1
2
34
5
67
8910
11
121314
Faz. Água Limpa, DF
E. E. Serra das Araras, MT
Faz.Água Limpa, DFPorto Limão, MT
Serras das Araras, MTPontes e Lacerda, MT
Januária, MG
Faz. Água Limpa, DFCerrado bordering the
Pantanal, MTFaz. Água Limpa, DF
Itirapina, SPBrasília NP, Brasília, DF
Emas NP, GOFaz. Água Limpa, DF
Serra das Araras, MTBrotas, SP
Emas NP, GO
Cerrado ss
Gallery forestForest
Cerrado ssCampo sujo and limpo
Cerrado ssDry forestForest (?)
Grassland and forest Cerrado ss
ForestGallery forest
ForestGallery forest
GrasslandsCerrado ssGrasslandCerrado ss
Campo cerradoPasture with trees
Marshes
Jan to Aug
Jan, Feb and/or MarJan, Feb and/or MarJan, Feb and/or MarJan, Feb and/or MarAll months of year
JulJun, Jul, Sep or Oct
AugSepSep
Dry season
–Near year round
Nov–
OctYear round
JanJun to Sep
Oct
2 to 6, mean = 3.8
–4 to 8
–––––––––
–3 to 11, mean: 5.0
3––
Mean ± s.d. = 3.9 ± 1.6
323
4 to 27, mean = 11.5
Up to 30
10 to 15Up to 40
–––––––
––––––
–––
Seasonal variationAdvantage
Advantage
AdvantageSeasonal variation
Advantage
94 TUBE
LISAPPENDIX 1. Continued.
Country/Sources
Study area Vegetation Seasonal occurrence Numbers of species Number of birds Question
APPENDIX 2. Bird species recorded in mixed-species flocks in Cerrado, with the source of information.The nomenclature and sequence of species follows Sigrist (2006).
