Rediscovery of the Selva Cacique (Cacicus koepckeae) in Southeastern Peru with Notes on Habitat, Voice, and Nest

2004·6.1 MB·English

by  Nathaniel G Gerhart

Wilson Bulletin, 1 16(1), 2004, pp. 74-82
REDISCOVERY OE THE SELVA CACIQUE {CACICUS KOEPCKEAE)
IN SOUTHEASTERN PERU WITH NOTES ON HABITAT, VOICE,
AND NEST
NATHANIEL G. GERHART' ^
—
ABSTRACT. Two new locations, vocalizations, and the nest of the little known Selva Cacique (Cacicus
koepckeae, Icteridae) are described from southeastern Peru. Similarities between the vocalizations ofSelva and
Ecuadorian caciques (C. sciateri) indicate that the two species may be closely related. The Selva Cacique may
be ecologically restricted to narrow rivers and headwater regions, where found in river margin habitats and
nearby transitional forest. Its occurrence from an elevation of 300 m at the type locality to <575 m at the
headwaters ofthe Rio Manu Chico, 240 km from the type locality, indicates that it may occur in small numbers
over a much larger area than was previously known. Received 29 July 2002, accepted 11 March 2004.
Until recently, the Selva Cacique {Cacicus Sony TCM-5000EV tape recorder and a Senn-
koepckeae) was known from only two speci- heiser short-shotgun microphone (model
mens collected at the type locality: Balta (10° #ME66 with K6 powering module) to record
08' S, 71° 13' W; elevation 300 m), on the Rio vocalizations.
Curanja, Depto. Ucayali, Peru (Lowery and
O’Neill 1965). Here, I report two new local- RESULTS AND DISCUSSION
ities for the species in southeastern Peru, and —
provide the first description of its voice and Field identification. C. koepckeae is—dis-
tinctive, and its pattern—of coloration all
nest.
black with a yellow rump is striking. In this
STUDY AREA AND METHODS region ofPeru, no other birds resemble it. The
obsBeertvweeden1-427C.Mkaorecphckaenade s2e0veAnprtiilmes19i9n8,theI ssyhmopwastrmicucYhelmloorwe-ryuemlpleodw Coancitqheueru(mC.p,cetalial),
and wings. Like C. cela, C. koepckeae has a
vicinity ofthe Nanti village ofMontetoni (11°
54' S, 72° 21' W; elevation 550 m) on the up- light-colored bill and a bluish eye. The bill of
per Rfo Camisea, Depto. Cusco, and on the C. koepckeae is grayish white and small for
nearby Rfo Manu Chico, Depto. Madre de an icterid, imparting a small-headed look rel-
Dios, Peru (Fig. 1). On 1 October 1998, I ob- ative to the long, strong-headed appearance of
served a pair ofbirds near the Rfo Shihuaniro, C. cela (especially the male). C. koepckeae
a tributary of the Rfo Timpfa, in the Matsi- (23 cm) is slightly smaller than C. cela (23—
genka (Machiguenga) community of Timpfa 29.5 cm; Ridgely—and Tudor 1989).
(12° 04' S, 72° 49' W; elevation 410 m) on the Vocalizations. On 11 April 1998, 09:00
olbosweerrveRdfogrUoruupbsaomfba2,-5Debpirtdos.sCeuvesrcaol,tPiemreus.atI EkSoTe,pcIkemaaedevotchaelifziartstioknnsonwenarretchoerdRifnogsMoafnCu.
this and other nearby sites on the Rfo Shihu- Chico, elevation 575 m, Manu National Park,
aniro in Timpfa on 23 October 1999, 16-17 just across the Isthmus ofFitzcarrald from the
May 2000, and 27-28 July 2001. All of these village of Montetoni on the upper Rfo Cami-
sightings were along narrow tributaries, but on sea (LNS #96000; see track 93 ofSchulenberg
24 and 29 July 2001 I observed a pairofSelva et al. 2000). One bird vocalized while 3-4 in-
Caciques visiting flowering Erythrina (Legu- dividuals moved through the canopy of a pre-
minosae) trees along the main course of the dominantly Cecropia (Moraceae) riverine for-
lower Rfo Urubamba across from the eco- est grove along an old, overgrown, dry river
tourism lodge of the Machiguenga Center for course. I recorded 20 sec of vocalizations be-
Tropical Studies in Timpfa. I used a modified fore playback and 35 sec after playback.
The vocalization I recorded was a rapid se-
^'Ccu/orrSeenltvaadSdurre,ssC:as2i3l1la5 1M2c0G0,eeCuAsvceo.,,PBeerruk.eley, CA rciheiscko-fpoluowu.d,Iqnuimcky, eaxnapllyossiisv,e,eapcaihrebdurnsotteso:f
94703, USA; e-mail: [email protected] sound constituted a note. Isler et al. (1998:
74
Gerhart • SELVA CACIQUE IN PERU 75
74°00' 73°30' 73°00' 72°30' 72°00' 71°30' 71°00' 70°30' 70°00'
FIG. I. Mapofsoutheastern Peru showing the location ofthe type locality (Balta. solid triangle), theauthor's
two new locations for Cacicus koepckecie (Montetoni and Timpfa, solid circles), and other ohservers' recent
encounters (Parotori, Cocha Cashu, and Shanibuyacu, solid sc|uares). Inset at upper left shows approximate
j location within Peru of the center of the enlarged map.
580) defined a note as “an unbroken trace on before the first noticeable pause—. Follouing
a spectrogram, including associated over- playback, the pace was slower abotit one
tones.” Paired notes were separated from each chick-pouw phrase every 2 sec for 35 sec. In
other by brief patises and formed musical both cases, the series began v\ith 2-3 lone
phrases, which may be repeated one or more chick notes before the first two-note phrase.
times (F"ig. 2). Before playback, the two-note Before playback, .several three-note chick-
phrases were given at about I/sec for 20 sec jyouw'-poKW' phrases were interspersetl in the
76 THE WILSON BULLETIN • Vol. JJ6, No. 1, March 2004
s
EIG. 2. Sound spectrograms of vocalizations of (A) Cacicus koepckeae (from the Rio Manu Chico, south-
eastern Peru; recorded before playback by N. Gerhart; LNS 96000) and (B) Cacicus sclateri (from the Kapawi
Ecological Reserve, Rio Pastaza, southeastern Ecuador; recorded before playback by J. Moore).
series of two-note phrases, but this only oc- Moore (1997) during September 1996 at Ka-
curred once after playback. pawi Lodge, Rfo Pastaza, in eastern Ecuador.
On 1 October 1998, I recorded a pair on the Seven two-note chick-kuh phrases of C. scla-
Rio Shihuaniro at about 09:00, when the birds teri were similar in pitch and quality to the
vocalized from the canopy of river margin chick-pouw phrase given by C. koepckeae.
vegetation at a height of 5-8 m (LNS The first note of this C. sclateri vocalization
#96001). The rhythm of the vocalizations in is almost identical to the first note of C. koep-
this encounter was slightly different. Phrases ckeae’s vocalization, with the second note
were slower, more widely spaced, and not giv- only slightly shorter and sounding more
en in a long series: chih-chih, pouw-pouw. clipped. The explosive quality and speed of
The sharp chih notes were given in quick suc- the vocalizations ofC. sclateri are also similar
cession, followed by a brief pause; the two
to those of C. koepckeae. Servat and Pearson
pouw notes were given more slowly and less
(1991:94) described the vocalizations of C.
forcefully. One 14-sec sequence of the pair
sclateri as a “ringing song, pee-chur, pee-
contained six phrases, with the birds loosely
chur, pee-chur, chur-chur-chur.”
alternating phrases as they moved away. Only
two of the six phrases were complete four- Comparison ofthe last 2.5 sec ofthe sound
note phrases. Three shorter chih-chih, pouw spectrograms in Fig. 2 shows that C. koep-
phrases and one phrase with only the two ckeae gives 1 1 notes, whereas C. sclateri
loud, introductory notes were interspersed gives 6. C. koepckeae's pace is 4.4 notes/sec,
among the complete phrases. Long recordings whereas that of C. sclateri is 2.4 notes/sec. C.
(2 min 46 sec before playback and 29 sec after koepckeae changes the rhythm of its vocali-
playback) of a single bird at a different site zation from a two- to a three-note phrase once
on the Rio Shihuaniro on 17 May 2000 during the series, whereas C. sclateridoes not.
showed similar variations; at times, the ca- Nevertheless, the basic two-note structure of
dence was similar to that in each of the orig- both vocalizations is similar. Both vocaliza-
inal two recordings made in 1998. tions contain a steady and narrow range of
These vocalizations of C. koepckeae are frequencies. The F-shaped first note of both
similar to one of the vocalizations of C. scla- species is similar in shape and almost identical
teri, the Ecuadorian Cacique, recorded by in frequency. The second note ofthe C. koep-
Gerhart • SELVA CACIQUE IN PERU 77
ckeae phrase is heavily down-slurred, giving was never heard or seen inside continuous
it a longer, fuller, less clipped sound. high ground, or terrafirme, forest.
These vocal similarities suggest a close tax- C. koepekeae's preferred habitat, at foothill
onomic relationship between C. sclateh and elevations close to the eastern slope ofthe Pe-
C
koepckeae (see below). The principal vo- ruvian Andes, is the transitional forest that
calizations of other caciques that could be lines narrow, high-gradient rivers such as the
closely related are substantially different from Rios Shihuaniro, Manu Chico, and upper
those of C. koepckeae and C. sclateri. For ex- Camisea. In addition to the 17 sightings of C.
ample, when not mimicking other species, C. koepckeae, I heard it 10 times without seeing
cela typically gives a loud, variable four-syl- it, for a total of 27 encounters. All 27 en-
lable song beginning with a strong screech. C. counters were in river margin habitats, includ-
cela's frequent call note is a loud tchak. The ing Gynerium (Graminae) canebrakes, or in
Solitary Cacique (C. solitahus) gives a varied transitional forest near rivers. Nine of the 17
assortment of loud, whiney, squealing calls, sightings of C. koepckeae (including almost
but two ofits most often-heard songs in south- all of the earliest sightings near Montetoni)
i, efaosltleorwnedPerbuyaareh(i1g)haTseErEi-esO,ofalnodw (w2h)upa nlootueds, hinagvethbreoeungho,ftbhiercdsaneoipthyerofperrivcehredmairng,ionrfmoorevs-t
I piercing series of high TEW notes, sometimes in areas where braided channels or dry river
preceded by a softer, low growl. The Southern courses create a patchy distribution of river
i Mountain-Cacique (C. chrysonotus; Ridgely margin vegetation types. Six other sightings
I and Greenfield 2001) gives a variety of loud, have been of pairs foraging along or flying
jay-like calls. Often one bird repeats these across unbraided sections of the upper Rios
Camisea and Shihuaniro, plus two sightings of
harsh notes while another duets with a de-
scending whistle. The principal song of the birds foraging in transitional forest along an
unbraided (wider) portion of the lower Rio
Golden-wingedWHCEacEiOque (C. chrysopterus) is Urubamba in Timpia.
a high, loud note, sometimes dou-
The dry river courses, side branches, and
bled, and preceded by several lower glo-glo
braids (referred to as otsegoa by local resi-
notes. Its call is a catbird-like wreyur(Ridgely
dents) of these narrow, high-gradient rivers
and Tudor 1989).
carve up the river margin forests into a patch-
The vocalizations of C. koepckeae appear
work of successional habitats of varying ages
to lack the variability of those of C. cela, C.
and structures, forming the preferred habitat
solitarius, and C. chrysonotus. The primary
for C. koepckeae. The structure and distribu-
chick-pouw paired notes constituted all C. tion of otsegoa habitat are spatially and tem-
koepckeae vocalizations heard and recorded
porally variable. Although these otsegoa often
dbiefffoerreent27v-o2c8aliJzualtyio2n00f1r,omwohneenorI broetchormdeedm-a sfitllandduirnigngwathteerradiunryisnegastohne adnrdy rseetaaisnonp,oomlasnoyf
bers of a pair disturbed by tape playback of are partially or fully overgrown with succes-
the primary vocalization recorded along the sional species such as Cecropia, Balsa (Bom-
upper Rfo Shihuaniro in Timpfa. The new vo- bacaceae), and Gynerium. In some cases, es-
calization, reminiscent of some sounds made tablished groves of riverine forest shade an
by C. solitarius, consisted of two consecutive open understory scraped clean by rainy-sea-
mournful, downslurrcd, high-pitched whines, son floods, but the understory can also be
together lasting about 1.5—sec. overgrown with dense, viney, rank vegetation.
Habitat and behavior. J. P. O’Neill (Rid- In places, otsegoa forests arc similar in
gely and Tudor 1989:372) posited that C’. structure to the well-known primary succes-
koepckeae was likely “an arboreal bird oc- sional habitats described from the Manu Bio-
curring mostly at forest borders.” \. A. Parker, sphere Reserve area ('rerborgh ct al. 1984).
III, reported that it “appears to be restricted Nevertheless, they are more heterogeneous
to riverine habitats in the region” (C’ollar et and patchily distributcti than the Gyneriutn
al. 1992:961). My recent observations support canebrakes and transitional forest habitat
those ideas. None of my encounters was far types familiar to observers along the witlei
from river margin habitats, and the species Manu and Mailrc ile Dios rivers. Otsegoa for-
1
78 THE WILSON BULLETIN • Vol. 116, No. 1. March 2004
ests seem to predominate at slightly higher el- in the otsegoa forest. They moved to the open
evations closer to the Andes, where Ficus crown ofa 15- to 20-m-tall En'thrina tree that
(Moraceae), Cedrela (Meliaceae), and Ery- was just starting to leaf out, and appeared to
thrina species are less prominent members of probe along the bare branches from which
the canopy tree community than in the tran- hung some dry seed pods. They hopped be-
sitional forests along lowland, meandering tween perches, staying mostly in the crown’s
rivers, such as the middle and lower Manu. outer branches. They then moved to a female
Gynerium cane is a common element in the (red) Triplaris (Polygonaceae) tree with dry-
understory ofotsegoa forests, and often forms ing, brown seed pods, about 15 m above
isolated patches between rocky river beds; it ground. Five birds actively foraged in the
does not tend to occur in the broader, more clumps ofdry pods by hanging in, and climb-
homogeneous stands referred to as the zaholo ing through, the clumps. Because I made these
habitat type by Terborgh et al. (1984). In some observations with lOX binoculars from 50 to
places, patches of Giiadiia (Graminae) bam- 150 m away, I could not determine whether
boo are also found in the understory of otse- the birds were taking insects or nectar, or
goa forests. C. koepckeae has been observed probing in open seed pods. Nevertheless, they
using these areas, even perching in bamboo, did not take fruit from an adjacent tree.
but it ranges throughout otsegoa forests rather During the early morning hours of 28 July
than being restricted to bamboo patches with- 2001, on the upper Rio Shihuaniro, I observed
in them. repeatedly a pair of C. koepckeae visiting a
Four of the 17 sightings occurred 13-20 flowering Erythrina tree to take nectar. In each
April 1998 at a Gynerium-doxmndiiQd island in of several visits, the birds would arrive to-
the middle of a dry river course {otsegoa) on gether, visit clumps of flowers mostly in the
the upper Rio Camisea, downstream of the outer branches, and never stay more than a
village ofMontetoni. In each encounter, a pair few minutes. On 24 and 29 July 2001, at 16:
of birds flew to the island between 17:30 and 15 and 15:00, respectively, I observed a pair
18:00 from the transitional forest that lined the of C. koepckeae foraging in a group of Ery-
dry river course and perched at the tip of the thrina trees in full bloom, alongside the main
tallest Balsa sapling on the island before hop- course of the lower Urubamba River. Heard
ping down out of sight to roost. Playback of and seen on both occasions on the bank op-
the vocalizations recorded on 11 April 1998 posite from the ecotourism lodge of the Ma-
would sometimes draw one or both individu- chiguenga Center for Tropical Studies, the
als across one arm of the dry river course, pair took nectar from several trees before
providing me with excellent views. Once they moving on. They did not linger in any one
reached the island for the night, the pair did tree as C. cela and other large icterids often
not vocalize more than once or twice, even in do in Erythrina trees in full bloom. Although
response to playback. the species is largely restricted to narrower
Between 06:50 and 07:04 on 23 October rivers and headwater regions, my two sight-
1999 at the Rio Shihuaniro in Timpia, I ob- ings along the main course of the lower Rio
served 1—5 individuals in four situations in the Urubamba indicate that C. koepckeae may vis-
crowns of transitional river margin forest. it other transitional and riverine forest habitats
This spot, where I also saw and recorded a opportunistically to take advantage of season-
pair of birds on 1 October 1998, is about 1 ally available nectar resources, especially rich
km up the Rio Shihuaniro from the main pop- patches of flowering Erythrina trees.
ulation centerofthe Timpia community on the My observations from the Rio Shihuaniro
lower Rio Urubamba. A single bird sunned suggest that pairs or family groups ofC. koep-
m
and preened about 10 above ground on top ckeae start their morning foraging slowly,
of a broken trunk that emerged from the rank sunning and preening, and then become more
vegetation ofa small, partially overgrown side active, and perhaps more vocal, as they move
channel {otsegoa) along the main river. Three to denser vegetation. My experiences with
more birds emerged, presumably from the birds heard along the Rio Shihuaniro indicate
dense undergrowth, and the four birds gath- that response to playback is variable at short
ered 10 m above ground in the top of a tree range and minimal at long range. A pair re-
Gerhart • SELVA CACIQUE IN PERU 79
spending to close playback on the Rio Shi- side the chamber for 5-70 sec. Several Crest-
huaniro on 16 May 2000 remained silent, but ed Oropendolas were building nests on the
li
readily moved across the river twice, perching other side of the same tree.
in the open 4-15 m above ground in the On 16 May 2000, at separate sites along the
crowns of riverside trees and bamboo. At the Rio Shihuaniro, I observed two old, inactive
same spot on 27 July 2001, what was presum- nests of similar shape and dimensions that
ably the same pair also drew nearer and then were reported by a local resident to be nests
I
crossed the river several times in response to of C. koepekeae. Each nest was situated 5-6
I playback. The pair had been actively vocal- m above ground in 12- to 15-m-tall trees,
i
izing when encountered, and responded to hanging near the end ofa branchlet offa main
j: playback with a series of chick-poiiw vocali- lateral branch. The trees were within 5 m of
I zations, eventually giving the disturbed vo- the river bank, and one nest almost hung over
j| calization and a rapid series of chih-chih- the water. In each case, as with the active nest,
![ pouw-pouw phrases after repeated playback. the tree’s crown stood alone, away from other
It appears that foraging pairs or small crowns. At all three sites, transitional forest
groups of birds range through the canopy of with a broken canopy lined the Rio Shihuani-
transitional forest throughout the middle of ro. At two of the sites, the understory was a
I the day, with foraging concentrated in the ear- dense, young stand of Giiadna bamboo. A lo-
ly morning and late afternoon. In my experi- cal resident said that C. koepekeae never nest-
enee, C. koepekeae moves rather steadily ed in colonies with the oropendolas or with
through its preferred habitat, vocalizing only C. eela, but that it nested at around the same
occasionally. In one case, it roosted with co- time ofyear (during the June to Novemberdry
lonial icterids, such as Olive Oropendola season).
—
(Psarocolius yurcicares). Russet-backed Oro- Taxononiie relationships. Although pre-
pendola (P. emgustifrons), and Crested Oro- cise relationships within this genus are not
pendola (P. decLimanus). A Matsigenka resi- known (Lanyon and Omland 1999), similari-
dent of Timpia noted that C. koepekeae sea- ties in vocalizations, habitat, and behavior
sonally fed at the white flowers of an Inga sp. support the hypothesis that C. koepekeae and
(Leguminosae) growing near the Rio Shihu- C. selateri are closely related, considered by
aniro. C. koepekeae should be looked for some to be “sister species’’ (T. A. Parker in
around the patchily concentrated nectar re- Cardiff and Remsen 1994). In addition to its
C
sources such as Erythrina and Inga trees, and vocal similarities, the all-black selateri is
possibly Miieuna sp. (Leguminosae) vines, also almost identical to C. koepekeae in size
which a—re visited by other icterids. and morphological structure, and it has a sim-
Nest. On 28 July 2001, along a braided ilar lowland Amazonian distribution (Cardiff
section of the upper Rio Shihuaniro, ob- and Remsen 1994). Based primarily on plum-
I
served an active nest of C. koepekeae for age characteristics, Lowery and O'Neill
about an hour. The solitary nest was hanging (1965) originally proposed that C. ehrysopte-
from the tip of one of the lowest branches, riis was C. koepekeae's closest relative; struc-
15-18 m above ground in a mature, 28- to 30- tural differences in the tail, feet, and bill were
m-tall Erythrina tree growing within 10 m of mentioned but not considered diagnostic. J. P.
the river bank. 1 estimated the nest to be 50- O’Neill (Jaramillo and Burke 1999) now con-
70 cm in length. It resembled the pendant nest siders this early idea to be in error. Other
of P. angustifrons, but appeared shorter and black and yellow caciques that could be close-
slimmer. It was constructed of woven plant ly related inclutle the similarly plumaged
libers that formed a tear-drop shape, had a top ehrysonotns. which occurs in temperate An-
entrance, and hung from a long, thin “tail” dean forest in southeastern Peru (jirincipally
like the nest of P. angnstifrons. A (presum- at 1,800-3,()()() m); C\ ehrysonotns. howexer,
ably) adult koepekeae came to the nest is much larger, and, like ('. ehrysopterns. tlif-
about every 5 min, but in one instance it made (ers in habits and voice (Ridgely and fudor
as many as three visits in a 6-min period. I 1989). C\ eela tliffers in size, plumage, \oice,
could not tell what it was bringing to the nest, and has a colonial social system.
but each time it entered the nest it stayed in- Other factors such as foraging behavior and
80 THE WILSON BULLETIN • Vol. 116, No. 1, March 2004
habitat preference also can be considered in vation in southeastern Peru. The range or hab-
determining species’ relatedness (Remsen and itat requirements of C. koepckeae must be
Schulenberg 1997). Similarities in habitat use somewhat restricted because, until recent
and foraging behavior of C. sclateri and C. sightings (see below), it had not been ob-
koepckeae support their close relationship. served at the several heavily studied sites in
Both are primarily arboreal foragers and occur Depto. Madre de Dios (although there were
in similar transitional forest or forest edges several possible sightings during the 1980s in
near water. Like C. koepckeae, C. sclateri for- the Manu National Park area; Collar et al.
ages in middle and upper levels offlood-plain 1992). Other narro—w rivers at similar eleva-
forest (Ridgely and Tudor 1989, Servat and tions in the region for example, the head-
Pearson 1991). In one case, C. sclateri for- waters of the Rfos Alto Manu, Serjali, Paqui-
aged actively for nectar or insects at flowers rfa, Mishagua, Cashpajali, Sepahua, Las—Pie-
4-10 m above ground in a leguminous tree dras, Cujar, Alto Purus, and Curanja and
alongside the Samiria River (Servat and Pear- possibly those in western Brazil, could pro-
son 1991). J. P. O’Neill (pers. comm.) ob- vide habitat for C. koepckeae.
served a pair of birds foraging “4-5 m up in Based on its habitat requirements, one
Gynerium and Cecropia at the edge of an would expect C. koepckeae to occur in the
Aguaruna-Jivaro garden plot near the edge of eastern part of the Manu Biosphere Reserve
the [Cenepa] river.’’ Ridgely and Tudor (1989: in Depto. Madre de Dios, where appropriate
368) classify C. sclateri as “apparently rare habitat may exist away from the main river. It
in canopy and middle levels of forest borders likely occurs in small numbers at least as far
and woodland,’’ where birds foraged in pairs, southeast as Boca Manu (12° 15' S, 70° 50'
“well up in trees, often probing into epiphytes W), where a possible sighting was reported in
and the bases of leaves.” Despite these simi- 1983 (Collar et al. 1992; M. Kessler pers.
larities, vocal differences (especially the struc- comm.). Although not particularly well doc-
tural difference ofthe second note oftheirpri- umented, this sighting, along an old branch of
mary vocalizations) do not support the sug- the Rfo Manu, would represent a significant
gestion that C. koepckeae and C. sclateri range extension to the east and south of the
could be subspecies of one species (cf. Jar- three confirmed sites discussed in this paper.
amillo and Burke 1999). Several other recent encounters add to our
—
Distribution and conser\'ation. Based on knowledge of the distribution and ecology of
my observations, C. koepckeae may be eco- C. koepckeae. A specimen of C. koepckeae,
logically restricted to narrow rivers, such as only the third of its kind, was collected along
those found in headwater regions at foothill the pipeline route of the Camisea gas project
elevations close to the eastern slope ofthe Pe- in Parotori, Depto. Cusco, about 25 km west
ruvian Andes. This seems especially evident of Timpfa in the upper Rfo Picha drainage (I.
in the Timpfa sightings. Before the sightings Franke pers. comm.). During September2001,
of24 and 29 July 2001, 8 months offieldwork near the Cocha Cashu Biological Station (11°
(spread over 3 years) along the main stem of 51' S, 71° 19' W; elevation —380 m) in Manu
the Rio Urubamba had resulted in no records National Park, J. Tobias (pers. comm.) re-
of C. koepckeae. Over this same time period, peatedly observed and tape recorded foraging
less than 15 km away, I recorded C. koep- groups of C. koepckeae along a bamboo-lined
ckeae on 8 of the 11 days I spent up the nar- stream near a site known as Playa Bonita,
row Rio Shihuaniro. In Timpfa, the conven- about 8 km inland from the main stem of the
tional wisdom among community members Manu River. During June 2000, P. Hocking
about C. koepckeae (known aspichocho in the (pers. comm.) reported seeing a C. koepckeae
lower Urubamba dialect of the Matsigenka fly from a “very long” nest along the Rfo
language) is that it is rarely, ifever, seen down Shambuyacu (a tributary of the Rfo Purus),
as far as the main course of the lower Rfo Depto. Ucayali, near the Brazilian border (i.e.,
Urubamba. relatively nearthe type locality at Balta). Most
Between the type locality and the sites of significantly, in 1996 D. Lane (pers. comm.)
my recent observations, C. koepckeae has now reported having seen a possible C. koepckeae
been recorded between 300 and 575 m ele- near Cushabatay in Depto. Ucayali, several
Gerhart • SELVA CACIQUE IN PERU 81
hundred km from the sites mentioned in this enous reserve to the west of the park may af-
paper. Indeed, if C. koepckeae and C. sclateri ford the species some de facto protection.
form a superspecies (as suggested by J. P. Nevertheless, current and proposed petro-
O’Neill in Jaramillo and Burke 1999), then C. chemical development in the Camisea and Rfo
koepckeae could occur well north into east- de Las Piedras areas may threaten the long-
central Peru. term integrity of C. koepckeae habitat in those
The September 2001 sightings documented areas. The conservation measures proposed by
with tape recordings from Manu National Collar et al. (1992:961) for this little-known
Park (J. Tobias pers. comm.) offer hope that species state that the first requirement is the
more will soon be discovered about the ecol- identification of “general areas in which some
ogy of this little-known species. J. Tobias reasonable population exists.” My observa-
1
(pers. comm.) reported that three different tions have begun to do that, but further explo-
family groups of C. koepckeae (of four, five, ration is necessary to document more fully the
and six individuals, respectively) were readily distribution and ecological requirements of C.
observable along a stream 4-8 m wide in low- koepckeae.
land rainforest. During eight visits along a 5-
I km stretch ofthe stream over a4-week period, ACKNOWLEDGMENTS
Tobias recorded the species 17 times, with I thank C. Munn, the Wildlife Conservation Society,
I each family group tending to be found re- and the Conservation Association for the Southern
peatedly along the same stretch ofstream. To- Rainforest (Selva Sur) for supporting my field work.
bias reported that during 2 months ofintensive R H. English supplied the sound spectrograms and
I made importantcomments on several drafts. L. Blanco
fieldwork in the surrounding mature, flood- Zamalloa produced the original map, M. P. Gerhart
plain forest, he never encountered C. koep- digitized and improved it, and D. Blanco and A. Trep-
I ckeae away from the stream: “They were al- tow assisted with the map and figure. W. Mayer, C.
ways along the stream itself, never in adjacent Richart, D. Huaman, and S. Russo made commentson
forest, and appeared to be restricted to dis- the manuscript. W. Schelsky, P. Donahue, and T.
! turbed riverine regrowth containing cane, Wood’s comments on structure were particularly help-
i bamboo, dense creepers, and some Heliconia ful. I thank I. Eranke, M. Kessler, P. Hocking, D. Lane,
\I;i l[Hoewliucnodneirascteoarey].t.o.t.heTuhpepyerrcaanngoepyd,fbrutomspetnhet talhenendibreersogp,wencSi.aelnLlc.youJH.nitlTteoyrb,siaJw.sitVf.horRCs.ehmaksroeiennpg,ckiJenra.f,eo.rJ.maT.Pt.iSoO.n’NSaecbiholuul-t,
most time 1-5 m above the ground in dense and C. Munn provided articles, encouragement, and
|) tangles of bamboo or vines. The species was important comments on various drafts of this manu-
J
^ almost always located by its loud and distinc- script. Special thanks go to J. Gerhart, C. Beier, L.
j tive vocalizations. . . . The species is not so Michael, and the people of Montetoni and Timpfa.
much skulking as inconspicuous and highly LITERATURE CITED
local.” These observations support my hy-
pothesis that C. koepckeae is ecologically re- Cardiifmfe,nsS.oWf.biarndsdiJn. tV.heReMmussf:enu,mJro.f1N9a9t4u.raTlypSeciesnpcece-.
stricted to narrower rivers and the patchy, suc- Louisiana State University. Occasional Papers of
' cessional habitats along their margins. the Museum ofNatural Science, no. 68. Louisiana
Further research is needed on the foraging State University, Baton Rouge.
I efocroeloigtsycaonndserdivsattriiobnutistoantuosfcCa.nkboeepacskseesaseedb.e-I Coi-i.rAoRn.oN.NtJ:.i,roL,.LP. CGi.ON/N.aArraiAn,joN,. TKraAb.bf:P,arAk.frM.aIdll-,
hmaavne hsaebeintaatinodn,hewahredreit asemveorsaalictiomfescunletairvahtue-d AAmNeDriIc).asC:.tWhfeCiIFC.dfP1/9I92U.C’TNhrreedatdeanteadbboiorkd,spoafrtth2e.
International ('ouncil lor Bird Preser\ation. (’am-
areas dominated the transitional forest near bridge. United Kingtlom.
the river edge. The pattern of low-impact, IsiiR, M. L., P. R. IsiFR. AND B. M. Whmma. 1998.
small-scale agriculture practiced by the indig- Use of vocalizations to establish species limits in
I! enous peoples of the lower Rfo Urubamba re- antbirds (I’asseriformes: rhamnophiliilae). Auk
gion does not seem to be incompatible with I LSi^STV-.SdO.
I the ecological needs of koepckeae. Most Jaramii H), a. and P. Bi rki . 1999. New World black-
I birds: the icteriils. ('hristophei Helm. Loiulon,
riverine habitats in the aforementioned head- Uniteil Kingdom.
! waters regions are still intact, and the Maim I.XNVON. S. M. AND K. I:. 0\ii \ND. 1999. A molecular
Biosphere Reserve and the contiguous iiulig- phylogeny of the blackbinls (Icteriilae): five line-
i
82 THE WILSON BULLETIN • Vol. 116, No. I, March 2004
ages revealed by cytochrome-fi sequence data. South America, vol. 1: the oscine passerines. Uni-
Auk 116:629-639. versity ofTexas Press, Austin.
Lowery, G. H., Jr., and J. P. O’Neill. 1965. A new Schulenberg, T. S., C. A. Marantz, and—P. H. En-
species of Cacicus (Aves: Icteridae) from Peru. glish. 2000. VoicesofAmazonian birds birdsof
Occasional Papers ofthe Museum ofZoology, no. the rain forest of southern Peru and northern Bo-
33. Louisiana State University, Baton Rouge. livia, vol. 3: ground antbirds (Eormicariidae)
Moore, J. V. 1997. Ecuador: more bird vocalizations through jays (Corvidae) [compact disc]. Library
fV.roMmoothreelNoawtluarnedRreacionfrodriensgt,s,voSla.n3Jo[scaes,seCtatlei]f.orJnoihan. othfolNoagtyu,raIlthaScoau,ndNse,wCoYronrekl.l Laboratory of Orni-
Remsen, j. V., Jr., and T. S. Schulenberg. 1997. The Servanott,esG.anadndreDc.orLd.sPefoarrssoenv.en19p9o1o.rlNyatukrnaolwhnistboirryd
pervasive influence ofTed Parker on Neotropical species from Amazonian Peru. Bulletin of the
f7i-e1l9d.ornithology. Ornithological Monographs 48: TerbBorrigthis,hjO.rnWi.t,hoj.loWg.icaElitCzplautbri1c1k,1:a92n-d95L.. Emmons.
Ridgely, R. S. and P. j. Greeneield. 2001. The birds 1984. Annotated checklist of bird and mammal
of Ecuador, vol. 1: status, distribution, and tax- species ofCocha Cashu Biological Station, Manu
onomy. Cornell University Press, Ithaca, New National Park, Peru. Eieldiana (Zoology) New Se-
York. ries, no. 21. Eield Museum of Natural History,
Ridgely, R. S. and G. Tudor. 1989. The birds of Chicago, Illinois.