An extraordinary feeding assemblage of birds at a termite swarm in the 
Serra da Mantiqueira, S?o Paulo, Brazil  
Storrs L. Olson1 and Herculano M. F. Alvarenga2
1 Division of Birds MRC 116, National Museum of Natural History, P. O. Box 37012, Smithsonian Institution, Washington, D. C. 20013-
7012.  E.-mail: olsons@si.edu
2 Museu de Hist?ria Natural de Taubat?, Rua Juvenal Dias de Carvalho, 111, CEP 12070-640, Taubat?, SP Brasil.  E.mail:halvarenga@uol.
com.br
Recebido em 25 de abril de 2006; aceito em 28 de julho de 2006
RESUMO. Uma not?vel concentra??o de aves se alimentando de cupins alados na Serra da Mantiqueira, S?o Paulo, Brasil. Em 29 de abril de 
2001, durante 40 minutos em um trecho de cerca de 100 m ao longo de uma estrada na Serra da Mantiqueira, Estado de S?o Paulo, Brasil, coberta por 
fl oresta montana, observamos 23 esp?cies de aves atra?das por uma revoada de cupins (Isoptera). V?rias esp?cies dessa agrega??o n?o s?o tidas como 
importantes ca?adoras de insetos, mas estavam ativamente apanhando e devorando os cupins em pleno v?o. A esp?cie predominante, pela abund?ncia e 
atividade era o Quete Poospiza lateralis, que normalmente forrageia no interior de arbustos.
PALAVRAS-CHAVE: Isoptera, revoadas de cupins, aleluias, aves montanas. 
KEY WORDS: Isoptera, montane forest birds, termite swarm.
Swarms of emerging winged termites provide an ephemeral 
but temporarily abundant source of very rich food that is 
highly prized by birds, which may aggregate in numbers to 
take advantage of the abundance in both temperate (Blake 
1941) and tropical environments in both the Old (Thiollay 
1970, Dial and Vaughan 1987) and New World (Eisenmann 
1961, Sazima 1989, Vasconcelos 1999, Gussoni and Campos 
2003).  Such aggregations are sporadic and short-lived and 
provoke much aberrant behavior in birds so that any obser-
vations of such phenomena deserve careful documentation. 
We report here a feeding assemblage that included species 
that may well never have been seen together before in such 
a manner.  This took place in the mountains of the Serra da 
Mantiqueira that rise up on the northwestern side of the Tau-
bat? Basin in the northeastern extension of the state of S?o 
Paulo, Brazil.  
The site of our observations (22? 41? 09? S; 45? 35? 37? 
W, elevation 1,550 m) was on a narrow paved  road running 
from the town of Campos do Jord?o to Pedra do Ba?, c. 4.9 
km almost due north of Campos do Jord?o, municipality of 
S?o Bento do Sapuca?, SP, Brazil.  Vegetation is montane for-
est that constitutes part of the interior Atlantic Forest, with 
a predominance of gymnospermous trees such as Araucaria 
and Podocarpus.  The period of observation was from 13:50 
to 14:30 h on 29 April 2001.  The weather was cool (c. 17? C) 
and completely overcast, sometimes with mist.  Observations 
ended with a brief rain shower that coincided with the end of 
feeding activity.  
We had been searching along the same road for most of the 
morning, but because of inclement weather we noted relative-
ly little avian activity.  We then encountered a large fl ock of 
Streptoprocne swifts that were feeding on swarming insects 
and began to note other birds feeding on the same insects. 
The intensity of activity built up over a period of 40 minutes 
and then ended abruptly so that we consider that we must have 
arrived very near the beginning of the feeding swarm, which 
took place along no more than 100 m of the road.  
The insects were winged termites (Insecta: Isoptera) with 
orangish bodies and black wings that appeared to be emerging 
from the ground in great numbers along one side of the road. 
The air was essentially still and the termites quickly rose to 
considerable heights.  They were also quite fat to judge from 
the grease stains left on a paper envelope in which we saved a 
few specimens that were given to a specialist for identifi cation 
but were subsequently lost.  
Observations of the species of birds that were actively 
feeding at the swarm of alate termites are presented below:
White-collared Swift Streptoprocne zonaris.  These swifts 
were among the fi rst to be attracted to the termites and their 
numbers built up until they fi lled the sky, actively gather-
ing termites.  There were between 100 and 150 individuals 
that eventually moved off, circled in the distance, and then 
moved on.
Surucua Trogon Trogon surrucura.  At least two were actively 
fl ying out and catching termites over the road, being so intent 
that they allowed close approach.
Chicli Spinetail Synallaxis spixi.  At least two birds were part 
of the assemblage but remained close to the ground catching 
insects.
Streaked Xenops Xenops rutilans.  A single individual forag-
ing in the normal manner was later joined by two others, at 
least one of which was seen actively fl ycatching.
Revista Brasileira de Ornitologia 14 (3) 297-299
setembro de 2006
298
Buff-fronted Foliage-gleaner Philydor rufus.  Three to six in-
dividuals were present and actively fl ycatching.
Olivaceous Woodcreeper Sittasomus griseicapillus.  Two or 
three were actively fl ycatching.  
Black-capped Piprites Piprites pileatus.  Three individuals of 
this rare and poorly known species were very actively engaged 
in the aerial capture of termites.  
Gray-hooded Flycatcher Mionectes rufi ventris.  One was 
around for most of the time and was not particularly active, 
although it was defi nitely seen fl ycatching termites.
Mottle-cheeked Tyrannulet Phylloscartes ventralis.  At least 
three spent most of their time in trees but were also seen 
hawking insects.
Shear-tailed Gray Tyrant Muscipipra vetula.  A pair arrived 
rather late and actively engaged in aerial feeding on termites.
Blue-and-white Swallow Notiochelidon cyanoleuca.  Perhaps 
as many as 20 or 30 passed by taking very high-fl ying ter-
mites, although we could not be sure if these numbers includ-
ed individuals that visited more than once.  
Eastern Slaty-Thrush Turdus subalaris.  One or two individu-
als were seen to engage in fl ycatching.  One or two individuals 
of Rufous-bellied Thrush T. rufi ventris were also present and 
probably fed on termites as well although we did not specifi -
cally note it.
Rufous-crowned Greenlet Hylophilus poicilotis.  Two birds 
stayed mainly in the vegetation but occasionally sallied out 
for fl ying termites.
Rufous-browed Peppershrike Cyclarhis gujanensis.  One or 
two were very actively fl ycatching.
Golden-crowned Warbler Basileuterus culicivorus.  At least 
two were most active in shrubbery close to the ground but 
were also seen actively fl ycatching.
Chestnut-headed Tanager Pyrrhocoma rufi ceps.  At least 
two males were present and very actively catching termites 
in the air.
Diademed Tanager Stephanophorus diadematus.  Despite its 
being very common in the area we saw only three individu-
als of this species and those were not present the whole time. 
When they were, they were very actively hawking termites.
Fawn-breasted Tanager Pipraeidea melanonota.  One was 
present for a short period only, but was actively fl ycatching.
Burnished-buff Tanager Tangara cayana.  A single bird stayed 
for only a short period but was actively fl ycatching.
Red-rumped Warbling-Finch Poospiza lateralis.  These were 
the most avid feeders on termites of all the birds present. 
They must have come from far and wide as there were 50 
to 70 individuals present.  The loud whirring of their wings 
was the dominant sound during the feeding frenzy and dozens 
were constantly in the air over our heads for the whole period 
of activity.  
Rufous-collared Sparrow Zonotrichia capensis.  Two or three 
were very active on the ground in the vicinity of the emerging 
termites, especially where these had become entangled in spi-
der webs.  Vasconcelos (1999) and Gussoni and Campos (2003) 
also reported this species feeding on winged termites.  
Green-winged Saltator Saltator similis.  One or two were ac-
tive and fl ycatching in the early part of the feeding activity.
Golden-winged Cacique Cacicus chrysopterus.  Only a single 
individual arrived even though it called almost constantly. 
This bird was extremely vigorously engaged in catching ter-
mites out of the air.
The following birds appeared to be part of the feeding 
assemblage but were not observed catching termites.  Their 
presence, usually for an extended period, may have been due 
to fl ocking behavior stimulated by the large number of other 
species actively engaged in feeding.
Red-breasted Toucan Ramphastos dicolorus.  A pair stayed in 
evidence for about half the period of observation.  We later 
saw one cross the road and hesitate in fl ight as though catch-
ing an insect, but this was not repeated.  
White-spotted Woodpecker Veniliornis spilogaster.  A single fe-
male remained in the area the whole time we were present.
Scaled Woodcreeper Lepidocolaptes squamatus.  First ob-
served singly but then up to three were seen together in the 
area of main feeding activity of which they appeared to be 
a part.  
Variable Antshrike Thamnophilus caerulescens.  One pair was 
active in the area for most of the period of observation but 
were not seen to take termites, despite the fact that the Slen-
der Antbird Rhopornis ardesiaca and several other species of 
Thamnophilidae have been seen preying on winged termites 
(Teixeira 1987).
Rufous-bellied Thrush Turdus rufi ventris. See T. subalaris, 
above.
Tropical Parula Parula pitiayumi.  One bird seemed to be part 
of the general activity.
The presence of single individuals of three other species 
appears to have been entirely incidental to the termite swarm 
and its predatory assemblage: Yellow-headed Caracara Milva-
go chimachima, Brazilian Ruby Clytolaema rubricauda, and 
White-throated Woodcreeper Xiphocolaptes albicollis.  
It is most interesting to compare our observations with 
those made by Gussoni and Campos at a termite swarm also 
in the state of S?o Paulo, only about 100 km to the SW but at 
a lower elevation (c. 750 m).  They observed a similar number 
of species (26) but only two of these Notiochelidon cyano-
leuca and Zonotrichia capensis were shared between the two 
assemblages.  Over half of the species (16) observed by Gus-
soni and Campos belong to families that ordinarily capture 
fl ying insects (Tyrannidae and Hirundinidae), in contrast with 
the assemblage we observed.  
Thiollay (1970:273) reported that in West Africa each of 
some 150 species of birds seen feeding at termite swarms 
?maintains in the swarms his particular hunting habits, strata 
and even preferred feeding hours, so that any interspecifi c 
competition is avoided.?  This is certainly different from the 
behavior reported by Dial and Vaughan (1987) in East Africa 
and what we saw in Brazil.  The feeding aggregation we ob-
served took place at a time of day when avian feeding activ-
ity is usually minimal, as was also observed by Eisenmann 
Storrs L. Olson and Herculano M. F. Alvarenga
299
(1961) and many of the species observed departed markedly 
from their normal feeding habits and ecological strata.  
As with the aggregations seen by Eisenmann (1961), Dial 
and Vaughan (1987), and Gussoni and Campos (2003), aerial 
insectivores (swifts, swallows, fl ycatchers) were present, but 
they were joined by scansorial, frugivorous, and granivorous 
species, as well as terrestrial and arboreal insectivores.  Un-
like African assemblages, raptors were not present.  Interspe-
cifi c competition would probably not have been a factor due 
to the abundance of the emerging termites.  
As noted by Dial and Vaughan (1987), termite emergences 
are unpredictable and predators must respond immediately 
in order to benefi t from a hatch.  Wide-ranging aerial species 
such as swifts and swallows would be expected to be the fi rst 
to detect a swarm, as appeared to be the case in our instance. 
These might then attract other aerial feeders such as fl ycatch-
ers.  The most assiduous, indeed almost frenetic, predator at 
the swarm we observed was Poospiza lateralis, and this may 
have been the ?nucleus? species whose large numbers and 
constantly whirring wing beats may have drawn the attention 
of understory and canopy dwelling species.  
ACKNOWLEDGMENTS
We thank Gary R. Graves and an anonymous reviewer for 
calling pertinent references to our attention.
REFERENCES
Blake, C. H.  (1941) Termites taken by birds.  Auk 58:104.
Dial, K. P. and T. A. Vaughan  (1987)  Opportunistic predation 
on alate termites in Kenya.  Biotropica 19:185-187.
Eisenmann, E.  (1961)  Favorite foods of Neotropical birds: 
fl ying termites and Cecropia catkins.  Auk 78:636-638.
Gussoni, C. O. A., and R. P. de Campos (2003) Registro de 
uma grande concentra??o de aves se alimentando de 
?aleluias? Isoptera.  Lundiana 4:71.
Sazima, I.  (1989)  Peach-fronted Parakeet feeding on winged 
termites.  Wilson Bull. 101:656-657.
Teixeira, D. M. (1987) Notas sobre o ?Gravatazeiro?, Rhopor-
nis ardesiaca (Wied, 1831) (Aves, Formicariidae).  Rev. 
Brasil. Biol. 47:409-414.
 Thiollay, J.-M.  (1970)  L?exploitation par les oiseaux des es-
saimages de fourmis et termites dans une zone de contact 
savane-for?t en C?te-d?Ivoire.  Alauda 38:255-273.
Vasconcelos, M. F. (1999) Cupins alados (Neocapritermes 
sp.) como alimento do tico-tico, Zonotrichia capensis. 
Atualidades Ornitol. 87:12.
An extraordinary feeding assemblage of birds at a termite swarm in the Serra da Mantiqueira, S?o Paulo, Brazil