Content uploaded by Alexandre Aleixo
Author content
All content in this area was uploaded by Alexandre Aleixo
Content may be subject to copyright.
38
The Wilson Journal of Ornithology 120(1):38–49, 2008
DISTRIBUTION, BEHAVIOR, AND CONSERVATION STATUS OF
THE RUFOUS TWISTWING (CNIPODECTES SUPERRUFUS)
JOSEPH A. TOBIAS,
1,7
DANIEL J. LEBBIN,
2
ALEXANDRE ALEIXO,
3
MICHAEL J. ANDERSEN,
4
EDSON GUILHERME,
5
PETER A. HOSNER,
6
AND
NATHALIE SEDDON
1
ABSTRACT.—The Rufous Twistwing (Cnipodectes superrufus), a newly described Amazonian tyrant-fly-
catcher, is known from five specimens and five localities in Cuzco and western Madre de Dios departments,
Peru. We report three additional specimens and eight new localities extending the known range of the species
east across Dpto. Madre de Dios, Peru, into Dpto. Pando, Bolivia, and Acre State, Brazil. The new localities
increase the distribution from
⬃
3,400 to
⬃
89,000 km
2
. We collected biometric data from five individuals, made
behavioral observations in the field, and recorded three separate types of vocalizations, two of which (including
the song) were previously unknown. We provide quantitative description of these vocalizations, consider their
function, and compare them with vocalizations of the only known congener, the Brownish Twistwing (Cnipo-
dectes subbrunneus). Unique vocal repertoires support the classification of these two forms as sister species.
The Rufous Twistwing resembles the Brownish Twistwing in producing loud vocalizations from regular song
posts and both species appear to have a polygamous mating system. We provide further evidence consistent
with the hypothesis the Rufous Twistwing is a Guadua bamboo specialist and recommend that it be listed as
Vulnerable on the IUCN Red List. Received 5 July 2006. Accepted 23 March 2007.
The Rufous Twistwing (Cnipodectes super-
rufus) is a large, distinctive tyrant-flycatcher
from southwest Amazonia. It occurs in a re-
gion relatively well studied by ornithologists,
but was only recently described and little is
known about its behavior, distribution, and
population size (Lane et al. 2007).
The only previous publication considered
the Rufous Twistwing likely to be a Guadua
bamboo specialist, but noted that it was al-
most certainly heard in ‘‘river-edge thicket
vegetation with no bamboo growth visible’’ at
Manu Lodge (12
⬚
06
⬘
S, 71
⬚
06
⬘
W; elevation
340 m) in 1991 and 2000 (Lane et al. 2007:
768). The five confirmed localities (Lane et al.
1
Edward Grey Institute, Department of Zoology,
University of Oxford, South Parks Road, Oxford, OX1
3PS, United Kingdom.
2
Department of Ecology and Evolutionary Biology,
Cornell University, E148 Corson Hall, Ithaca, NY
14853, USA.
3
Museu Paraense Emı´lio Goeldi, Coordenac¸a˜o de
Zoologia, Caixa Postal 399, 66040-170, Bele´m PA,
Brazil.
4
Macaulay Library, 159 Sapsucker Woods Road,
Ithaca, NY 14850, USA.
5
Universidade Federal do Acre, Departmento de
Cieˆncias da Natureza, BR-364, km 04, Campus,
69915-900, Rio Branco AC, Brazil.
6
Cornell Museum of Vertebrates, 159 Sapsucker
Woods Road, Ithaca, NY 14850, USA.
7
Corresponding author;
e-mail: joseph.tobias@zoo.ox.ac.uk
2007) encompass a small area (
⬃
3,400 km
2
)
in western Dpto. Madre de Dios and Dpto.
Cuzco, Peru. However, the existence of large
blocks of Guadua bamboo-dominated terra
firme forest in southwestern Amazonian Bra-
zil, southeastern Peru, and northwestern Bo-
livia (Silman et al. 2003) suggest the global
distribution of the species may be extensive
(Lane et al. 2007).
The only published vocalization is an agi-
tation call (Lane et al. 2007), whereas the full
vocal repertoire is likely to be similar to that
of the Brownish Twistwing (C. subbrunneus),
a presumed sister species with at least three
distinct vocalizations (Parker and Remsen
1987, Schulenberg et al. 2000, Ridgely and
Greenfield 2001). Male Brownish Twistwings
are notably disinterested in playbacks of their
songs, which are given from small, stable, and
often adjacent territories, suggesting a polyg-
ynous mating system (Parker and Remsen
1987, Ridgely and Tudor 1994, Ridgely and
Greenfield 2001, Fitzpatrick 2004). Lane et al.
(2007) reported no information regarding the
breeding behavior of Rufous Twistwing.
In this paper we report: (1) new localities,
(2) vocalizations, (3) habitat preferences, and
(4) behavior of the Rufous Twistwing to add
to the details presented by Lane et al. (2007).
We also assess the conservation status of the
species, based on our results, following IUCN
Red List criteria (IUCN 2001).
39
Tobias et al.
•
RUFOUS TWISTWING BEHAVIOR AND DISTRIBUTION
FIG. 1. Sites mentioned in the text mapped in relation to major cities (closed circles), rivers, and international
boundaries. Crosses indicate the previously known range of Rufous Twistwing in Peru: (1) Rı´o Urubamba
centered on Camisea, and (2) Rı´o Manu centered on Pakitza, Dpto. Madre de Dios (type locality). Stars indicate
new localities for the species: (3) Oceanı´a, Dpto. Madre de Dios, Peru; (4) Extrema, Dpto. Pando, Bolivia; (5)
Estac¸a˜o Ecolo´ gica do Rio Acre, Acre State, Brazil; (6) Ramal Jarinal, Acre State, Brazil; (7) Rio Branco, Acre
State, Brazil; (8) Timpı´a, Dpto. Cuzco, Peru; (9) Montetoni, Dpto. Cuzco, Peru; and (10) Tambopata, Dpto.
Madre de Dios, Peru. Open circles indicate study sites where C. superrufus was absent: (11) Camino Mucden,
Dpto. Pando, Bolivia; (12) Palmera, Dpto. Pando, Bolivia; (13) San Sebastia´ n, Dpto. Pando, Bolivia; and (14)
CICRA, Dpto. Madre de Dios, Peru.
METHODS
Study Sites and Survey Methods.—We sur-
veyed six Guadua-dominated forest sites in
southwestern Amazonia, including one site in
Bolivia and three in Brazil. Avifaunal surveys
were undertaken by three independent teams
at field sites using different methods at each
site except CICRA and Oceanı´a where the
same methods were used. Playback of pre-re-
corded Rufous Twistwing calls was used op-
portunistically at sites 1–3; in all cases we
used the agitated calls recorded by Thomas
Valqui H. at Kirigueti, Dpto. Cuzco, (i.e., Fig.
3A in Lane et al. 2007). Additional records
were sought from experienced observers. Lo-
calities were mapped (Fig. 1) and range sizes
measured using equal-area projections with
ArcView 3.2.
40
THE WILSON JOURNAL OF ORNITHOLOGY
•
Vol. 120, No. 1, March 2008
1. Centro de Investigacio´n y Capacitacio´n
Rı´o Los Amigos (CICRA), 270 m, Dpto. Ma-
dre de Dios, Peru (12
⬚
35
⬘
S, 69
⬚
12
⬘
W),
2001–2007. This research station is at the con-
fluence of the Rı´o Los Amigos and the Rı´o
Madre de Dios in the Los Amigos Conser-
vation Concession, a 135,000-ha private con-
cession managed by Amazon Conservation
Association. DJL, MJA, PAH, JAT, and NS
surveyed multiple patches of Guadua weber-
baueri and G. sarcocarpa bamboo for multi-
ple bird species over a 7-year period using
random transects, field observations, tape-re-
cording, and intensive mist netting. This site
includes camp CM1 on the south bank of the
Rı´o Los Amigos.
2. Oceanı´a, 250 m, Dpto. Madre de Dios,
Peru (11
⬚
23
⬘
S, 69
⬚
32
⬘
W), 11–18 October
2004. This locality is 6 km west of Iberia, a
town
⬃
130 km north of Puerto Maldonado on
the In˜apari road. DJL, MJA, and PAH sur-
veyed a large patch of Guadua weberbaueri
and G. sarcocarpa bamboo, in places collaps-
ing to form impenetrable thickets with a can-
opy at
⬃
4 m. The bamboo patch was on an
upland terrace, bordered by land recently
cleared and burned for agriculture, extending
to the nearby Rı´o Tahuamanu (to the east and
south), as well as successional forest and ma-
ture forest (to the north, south, and west). The
presence of burned and chained logs, and a
logging road revealed a recent history of hu-
man disturbance. We conducted daily census-
es using random transects between 0445 and
1000 hrs EST (12–16 Oct), and operated mist
nets daily from dawn until midday (11–17
Oct). A combination of 6- and 12-m nets were
rotated over 30 net locations covering a total
length of 700 m of trail in 7 days. Total effort
was
⬃
400 net hrs (for 12-m nets).
3. Extrema, 250 m, Dpto. Pando, Bolivia
(11
⬚
28
⬘
S, 69
⬚
15
⬘
W), 6–9 November 2004.
This military outpost is
⬃
45 km east-south-
east of the Oceanı´a site on the banks of the
Rı´o Tahuamanu where this river flows from
Peru into Bolivia. The natural habitat appears
to be forest with an open canopy and an un-
derstory largely composed of Guadua bam-
boo. JAT and NS surveyed habitat here and
at the adjacent guard-post of Alto Peru, 1 km
north in Dpto. Madre de Dios, Peru. We sur-
veyed habitat with and without bamboo on
random transects between 0445 and 1800 hrs
daily (6–9 Nov), spending 18 hrs in bamboo
at Extrema and 2 hrs in bamboo at Alto Peru,
Madre de Dios, Peru (2 km from Extrema).
4. Estac¸a˜o Ecolo´gica do Rio Acre, 250–
350 m, Acre State, Brazil (11
⬚
03
⬘
S, 70
⬚
12
⬘
W) on the border with Peru, 13–24 August
2005 and 3–15 February 2006. Site A was
surveyed in 2005 by AA, including riparian
forest along Rio Acre and a small tributary
(Igarape´ do Tombo); AA and EG surveyed
site B,
⬃
7 km west of site A, in 2006. Exten-
sive patches of Guadua bamboo were present
at both sites. Field effort included
⬃
250 hrs
of random transects through the main vege-
tation types, observing, mist netting, tape-re-
cording, and collecting birds.
5. Rio Branco, 250 m, Acre State, Brazil
(09
⬚
57
⬘
S, 67
⬚
57
⬘
W), 1998–2006. Long-term
fieldwork is being conducted by EG at Parque
Zoobotaˆnico, a field station managed by Univ-
ersidade Federal do Acre. The site is within
the suburban limits of Rio Branco and the
habitat (
⬃
100 ha of humid forest with patches
of Guadua sarcocarpa and G. weberbaueri
bamboo) is becoming increasingly disturbed
and isolated by peripheral clearance (Guilher-
me 2001). The main survey technique was
mist netting: 10 mist nets (12 m) were oper-
ated for 6 hrs/day (usually dawn until midday)
on
⬃
35 days/year (evenly split between dry
and wet seasons). Annual mist netting effort
was
⬃
2,000 net hrs.
6. Ramal Jarinal, 320 m,
⬃
100 km WNW
of Rio Branco, Acre State, Brazil (09
⬚
54
⬘
S,
68
⬚
28.5
⬘
W), 12–22 November 2006. EG and
Marcos Pe´rsio D. Santos (Universidade Fed-
eral do Piauı´) surveyed this site and collected
specimens. The habitat was terra firme forest
dominated by bamboo (Guadua sarcocarpa
and G. weberbaueri) and palms (e.g., Attalea
sp.) with high human disturbance including
recently cleared cattle pastures and selective
logging.
Sound Recording, Analysis, and Playback
Experiments.—Recordings of vocalizations
were made at Oceanı´a and Extrema using a
Sennheiser ME 67 shotgun microphone with
a Sony PCM-M1 MiniDAT or a Sony TC-D5
Pro II Cassette Recorder. All recordings are
deposited at the Macaulay Library (ML) at
Cornell University.
Recordings were automatically filtered and
digitized at a sampling frequency of 44.1 kHz
41
Tobias et al.
•
RUFOUS TWISTWING BEHAVIOR AND DISTRIBUTION
using Avisoft SASLabPro Version 4.15 (
2002 Raimund Specht, Berlin, Germany) with
a 16-bit acquisition sound card (0 VIA [Wave]
5.10). Sonograms were generated using stan-
dard broadband (216 Hz) filter settings (FFT
⫽
1,024, Frame
⫽
75%, Window
⫽
FlatTop,
Overlap
⫽
87.5%). Sonograms were inspected
visually, assigned to three types, and de-
scribed quantitatively using on-screen cursors
to measure standard temporal (sec) and fre-
quency (kHz) parameters: overall duration,
duration of first note, duration of middle note,
duration of final note, number of notes, pace
of notes (notes/min within a vocalization),
pace of calls (calls/min within a recording),
maximum overall frequency, minimum over-
all frequency, and bandwidth.
Playback experiments were conducted at
Extrema to investigate the function of vocal-
izations. Recordings were played through a
Sony SRS-58 speaker from a range of 10–40
m; the subject of most playbacks (7/10) was
the individual from which the tape was re-
corded. Samples were too small to permit sta-
tistical analysis.
RESULTS
Field Observations.—Intensive field work
at CICRA has not yet produced records of the
Rufous Twistwing, despite extensive bamboo
and a broad community of bamboo-specialist
birds. The species was more readily found at
other sites, including Oceanı´a, where it was
encountered 19 times (1–4 times daily) during
an 8-day survey. Fifteen of these encounters
were aural and four were visual; two obser-
vations lasted
⬍
1 min, one lasted
⬃
20 min,
and one lasted
⬃
45 min. At least three indi-
viduals were involved (2 were color-banded
and 1 unbanded bird was later seen).
The first record for Bolivia involved at least
four individuals at Extrema where the species
was encountered on 11 occasions during a
4-day survey. Eight observations were aural
and three were visual. Ten encounters were
within 2 km of the Extrema guard-post and
one encounter was within Peru near the Alto
Peru guard-post. One bird was photographed
and tape-recorded on 7 November 2004; at
least three other individuals were heard or
seen daily along a 2-km transect.
The first record for Brazil involved a single
individual mist netted and photographed at
Parque Zoobotaˆnico, Rio Branco on 22 May
1998, although it was only identified to spe-
cies in February 2006. The species was en-
countered repeatedly at Parque Zoobotaˆnico
from 1998 onwards: nine individuals were
mist netted, one specimen was collected, and
another window-killed individual was found
and prepared as a specimen. Both specimens
are currently housed at Museu Paraense Em-
ı´lio Goeldi (MPEG), Bele´m, Brazil (Table 1).
The Rufous Twistwing was encountered at
two further Brazilian sites. It was found once
at Estac¸a˜o Ecolo´gica do Rio Acre during
⬃
250 hrs of field work: AA heard and tape-
recorded a single individual on 15 August
2005 in the dense understory of an extensive
thicket of Guadua bamboo along a narrow
creek (Igarape´ do Tombo). Another individual
was found by EG and Marcos Pe´rsio D. San-
tos on 20 November 2006 at Ramal Jarinal
and a male was heard calling consistently
throughout the morning from a Guadua bam-
boo thicket in terra firme forest. It was attract-
ed with playback of its own calls and was col-
lected; the skin is stored in MPEG (Table 1).
Three other localities have recently come to
light (Fig. 1). The Rufous Twistwing was
found three times near the Machiguenga com-
munity of Timpı´a (12
⬚
04
⬘
S, 72
⬚
49
⬘
W; ele-
vation 410 m) on the Rı´o Urubamba, Dpto.
Cuzco, in 1998–2001 (N. G. Gerhart, pers.
comm.). It was also recorded once nearby at
Rı´o Shihuaniro, a tributary of the Rı´o Timpı´a,
in October 1998 and once near Montetoni
(11
⬚
54
⬘
S, 72
⬚
21
⬘
W; elevation 550 m) on the
upper Rı´o Camisea, Dpto. Cuzco in April
1998 (N. G. Gerhart, pers. comm.). A retro-
spective report involves a single individual in
Guadua bamboo at Explorer’s Inn, Tambopata
(13
⬚
08
⬘
S, 69
⬚
36
⬘
W; elevation 250 m) in
March 1984 (Michael Kessler, pers. comm.).
The Rı´o Shihuaniro locality is documented
with a sound recording (ML 125894), as are
the Timpı´a and Montetoni localities (N. G.
Gerhart, pers. comm.); the Tambopata record
is supported by a convincing description by
an experienced observer. Adding the eight
new localities to the original five produces a
global range of
⬃
89,000 km
2
(Fig. 1).
Mist Netting and Specimen Data.—Two
birds were mist netted at Oceanı´a on 12 Oc-
tober 2004, a capture rate of one bird per
⬃
200 net hrs (for 12-m nets). Both birds had
42
THE WILSON JOURNAL OF ORNITHOLOGY
•
Vol. 120, No. 1, March 2008
TABLE 1. Measurements of Rufous Twistwings (n
⫽
5).
Locality Date Age/Gender
Wing length
(mm)
a
Tail (mm)
Bill length
(mm)
b
Bill width
(mm)
c
Tarsus
(mm)
d
Mass (g)
Catalog
number
Oceanı´a
e
12 Oct 2004 Adult/Male 117 108 18.6 6.4 25.3 42 —
Oceanı´a
e
12 Oct 2004 Subadult?/Male 114 111 17.2 6.1 21.9 42 —
Rio Branco
f
01 Oct 2003 Subadult?/Female? 88.3 87 16.27 6.6 19.1 22 60001
Rio Branco
f
25 Feb 2006 Subadult/Male 100.3 95.4 17.73 6.9 20.3 35 60000
Ramal Jarinal
f
20 Nov 2006 Adult/Male 118.7 113.5 20.08 7.3 22.5 41 61351
a
Flattened wing chord.
b
Exposed culmen.
c
Measured at mid-nares.
d
Measured from tibiotarsus/tarsometatarsus joint to ventral base of halux.
e
From live birds by PAH.
f
From MPEG specimens by AA.
modified primaries suggesting they were
males. One individual had 100% skull ossifi-
cation and was molting secondaries, upper tail
coverts, and extensively over the body; the
other individual (possibly subadult) had
roughly 50% skull ossification, shorter crest
feathers, and no molt.
Two of nine birds mist netted at Rio Branco
were banded with color bands (1998) and five
with metal rings (2 in 1999, 1 in 2000, 1 in
2004, and 1 in 2005). Only one of these nine
birds was recaptured (once) and only one bird
was considered to be juvenile (in June) on the
basis of downy underwing plumage and traces
of a commisure. The trapping rate was
⬃
one
bird/1,800 net hrs (for 12-m nets). Unfortu-
nately, no biometric data were collected from
these mist net captures. An immature bird
found dead at Rio Branco in October was un-
dergoing molt of remiges, rectrices, and con-
tour feathers. The male collected at Ramal Jar-
inal on 20 November 2006 was an adult with
enlarged testes.
Biometric data for two mist-netted birds
and three specimens varied (Table 1). Mea-
surements at Oceanı´a were from live birds and
cannot be compared directly with other bio-
metrics as specimens tend to shrink slightly
during drying. Our sample is consistent with
earlier data as males were considerably larger
than females in mass and wing-length, but
more similar in terms of bill and tarsus size.
Total length of three fresh specimens was
18.4–24 cm.
Habitat.—We report 44 encounters with the
Rufous Twistwing (combining field observa-
tions with mist net data) of which 43 were in
or immediately adjacent to patches of Guadua
bamboo. It was recorded at Oceanı´a only
within one large patch of Guadua bamboo.
Eighteen of 19 encounters were in dense
⬃
5-m tall bamboo and one was in taller,
sparser bamboo under mature forest. It oc-
curred at Extrema in large stands of Guadua
bamboo, apparently preferring areas with a
highly discontinuous tree canopy, occurring
either in low treeless growth where the bam-
boo had collapsed (canopy height 2–4 m), or
in taller bamboo supported by scattered trees
(canopy height 10–25 m); two of 11 encoun-
ters were in the former habitat and seven in
the latter habitat. The species was also heard
and tape-recorded on two occasions at Extre-
43
Tobias et al.
•
RUFOUS TWISTWING BEHAVIOR AND DISTRIBUTION
ma in an extensive patch of young secondary
forest with a dense understory (canopy height
⬃
10–15 m). This habitat contained only small
amounts of bamboo but bordered an extensive
stand of Guadua. The record at Montetoni
was in scrubby secondary forest with a broken
canopy, a dense understory of vines, and
patches of Guadua bamboo; two records at
Timpı´a were in similar broken secondary for-
est with dense bamboo dominating the under-
story (N. G. Gerhart, pers. comm.). We failed
to detect the species at all sites in forest types
lacking bamboo, although one recent obser-
vation by N. G. Gerhart (pers. comm.) was in
dense secondary forest near Timpı´a with no
bamboo component.
The Rufous Twistwing appears to be re-
stricted to the understory. Birds at Oceanı´a
foraged 1–3 m above ground, usually at the
lower end of that scale. Similarly, individuals
watched at Extrema did not ascend higher
than 3 m above ground and spent most of their
time at 1–2 m.
Movement and Flight.—DJL and PAH ob-
served one Rufous Twistwing continuously
for 45 min at Oceanı´a as it performed 32 wing
raises in the following sequence: RRRRRR
LLLLLLR (R/L) LLRRRRRRRRRRLLLR
LL (bracketed codes denote right and left
wing simultaneous raises). The impression
given during this bout, and throughout all oth-
er field observations, was a strong preference
for single-wing raises, with both wings fa-
vored in roughly equal proportion. This indi-
vidual made significantly more wing raises by
raising the same wing repeatedly (23) than by
switching between wings (7):
2
⫽
8.53, P
⫽
0.003 (ignoring the simultaneous raise). Wing
raising did not seem to correlate with any par-
ticular context, usually occurring randomly
during bouts of vocalizing and foraging. There
was no discernable increase in rate for birds
accompanied by a second individual or in re-
sponse to playback of calls.
The flight action of the Rufous Twistwing
when moving between perches through the
understory was erratic with slightly floppy
and irregular downbeats. It was seen flying
low across roads on three occasions at Extre-
ma, at which time its flight was stronger and
more direct.
Foraging Behavior.—The Rufous Twist-
wing foraged for prey from fairly open, often
horizontal perches, slowly rotating the head to
search at different angles. Individuals perched
with an upright posture and flew to a new
perch after 30–300 sec. They made abrupt sal-
lies to small branches or leaves (bamboo and
non-bamboo) to snatch invertebrates. One in-
dividual at Oceanı´a made three direct, upward,
1-m sallies from a horizontal bamboo stem to
glean prey from bamboo leaves. It flew after
each strike to a new perch at the same level
as the original perch (once), or to a lower
perch (twice). Another individual performed
at least two upward sallies, two downward
sallies, and one horizontal sally towards prey
on small bamboo branches and leaves during
a 45-min period. In one typical sally the bird
traveled
⬃
30 cm upward, plucked a prey item
from a dead bamboo branch (with dry leaves),
and then dropped to a lower perch. It was ob-
served in the same 45-min period beating one
⬃
20-mm orthopteran against a branch. Other
prey items taken were small, unidentifiable,
and swallowed immediately.
Vocalizations.—We grouped all vocaliza-
tions into three categories from a total of
⬃
150 min of recorded vocalizations. These
categories may not encompass the species’ en-
tire repertoire, but it is likely they cover all
regularly given vocalizations.
Vocalization #1 is a loud, non-stereotyped,
scolding series. It was heard during 25 en-
counters at Oceanı´a and Extrema. The full call
is a loud series of 2–14 notes each containing
at least three, usually four, harmonics (Fig.
2A) and is, at times, preceded by one or two
loud pyew calls (Vocalization #3). The phrase
(without the pyew call) is fairly even in pitch
and pace, and is usually repeated during pro-
longed bouts at a rate (x¯
⫾
SD) of 13.8
⫾
8.9
calls/min (n
⫽
30 calls, from 3 bouts of call-
ing by 2 individuals; Table 2).
Vocalization #1 is deposited at the Macau-
lay Library, Cornell University (catalog num-
bers ML 124448, 124449, 124450). It was re-
corded at Oceanı´a and heard irregularly at Ex-
trema. A similar call was also recorded at Ex-
trema; it was more subdued with fewer notes
per series (ML 124455, Fig. 3B). The loud
version was heard throughout the day, often
in late morning and mid-afternoon when most
other species were silent; the quiet call was
given at 1630 hrs on 6 November 2004. Both
loud and quiet versions are similar in struc-
44
THE WILSON JOURNAL OF ORNITHOLOGY
•
Vol. 120, No. 1, March 2008
FIG. 2. Three call types of the Rufous Twistwing: (A) Vocalization #1 (ML 124448), an agitation call, (B)
Vocalization #2 (ML 124453), apparently the song, and (C) Vocalization #3 (ML 124454), the pyew call.
TABLE 2. Acoustic properties (mean
⫾
SD) for known vocalizations of the Rufous Twistwing.
Parameter
Vocalizations
d
#1
a
#2
b
#3
c
Overall duration, sec 0.97
⫾
0.57 1.12
⫾
0.13 0.33
⫾
0.05
Duration of first note, sec 0.07
⫾
0.01 0.15
⫾
0.02
Duration of mid-note, sec 0.07
⫾
0.01 0.09
⫾
0.01
Duration of final note, sec 0.07
⫾
0.01 0.11
⫾
0.01
Number of notes 6.5
⫾
4.3 6.4
⫾
0.8 1.0
Note rate, notes/sec 6.5
⫾
1.01 5.9
⫾
0.11
Call rate, calls/min 13.8
⫾
8.9 5.7
⫾
2.3 7.1
⫾
7.8
Max frequency, kHz 5.0
⫾
0.3 2.20
⫾
0.19 5.19
⫾
0.13
Min frequency, kHz 1.01
⫾
0.10 1.14
⫾
0.06 1.55
⫾
0.10
Bandwidth, kHz 4.0
⫾
0.29 1.01
⫾
0.25 3.64
⫾
0.16
Number of calls measured 30 11 13
a
Data from ML 124448 (locality: Oceanı´a, Peru; recordist: MJA), ML 124449 (locality: Oceanı´a, Peru; recordist: MJA), and ML 124455 (locality:
Extrema, Bolivia; recordist: JAT).
b
Data from ML 124452 (locality: Oceanı´a, Peru; recordist: PAH) and ML 124453 (locality: Extrema, Bolivia; recordist: JAT).
c
Data from ML 124454 (locality: Extrema, Bolivia; recordist: NS).
d
Harmonics were included in measurements for Vocalizations # 1 and # 3, but ignored for Vocalization # 2 where they were much weaker.
ture. Vocalization #1 at times is given by ag-
itated birds as a single note (ML 128954).
Vocalization #2 was recorded once at Oce-
anı´a (ML 124452) and heard four times at Ex-
trema (ML 124453). It consists of a loud
chiming series of 5–7 evenly paced notes, de-
scending in pitch, with the first note more di-
syllabic, ringing and prolonged than subse-
quent notes (n
⫽
11 calls, from 3 bouts of
calling by 2 individuals; Table 2, Fig. 2B).
45
Tobias et al.
•
RUFOUS TWISTWING BEHAVIOR AND DISTRIBUTION
FIG. 3. Sonograms, showing delivery pattern over longer time-frames, of (A) Vocalization #3, followed by
a ‘long call’, (B) a subdued version of Vocalization #1 (ML 124455), and (C) Vocalization #2, all given by a
single individual at Extrema, Dpto. Pando, Bolivia.
The individual notes have only two, rarely
three, harmonics. The typical phrase is more
stereotypic than Vocalization #1 and, at a rate
of 5.7
⫾
2.3 calls/min, is distinctly slower
(Fig. 3B, C). It is at times reminiscent in em-
phasis and pattern to the song of the Buff-
throated Foliage-gleaner (Automolus ochro-
laemus).
Vocalization #3 was a loud whistle (pyew).
It was heard during two encounters at Oceanı´a
and Extrema. Analysis of a sample (n
⫽
13
calls, from three bouts by one individual at
Extrema) indicates this call consists of a fun-
damental note with a single strong harmonic,
given alone or in couplets (Table 2, Fig. 2C).
Occasionally, Vocalization #3 was given in a
triplet followed by a series of trills lasting
over 10 sec (ML 124454, Fig. 3A). These
trills are, at times, given without an introduc-
tion by Vocalization #3 (e.g., ML 128955) and
seem to be an accelerated version of Vocali-
zation #1.
The functional significance of these vocal-
izations is not known. Vocalization #1 appears
to be an agitation call, Vocalization #2 is
probably the song, and Vocalization #3 can be
described as a ‘short call’. Vocalizations #2
and #3 have not previously been described
and appear most likely to function in court-
ship. The most stereotyped call type is Vocal-
ization #2 (Fig. 2B), which was given at reg-
ular intervals (Fig. 3C); it differs in structure
and regularity from non-stereotyped agitation
calls (Vocalization #1: Figs. 2A, 3B).
The Rufous Twistwing vocalized frequently
at Oceanı´a and Extrema with Vocalization #1
heard throughout the day, especially between
dawn (
⬃
0530 hrs) and
⬃
0800 hrs. Calls were
generally given in short bursts interspersed
with long periods of silence but, at times, in-
dividuals called more or less consistently for
periods up to 1 hr. At least two birds were
heard counter-calling on one occasion at Oce-
anı´a between
⬃
1610 and 1700 hrs. One in-
dividual at Extrema called intermittently in re-
sponse to playback for several hours in the
same area. All calls were given from low
perches, 1–3 m above ground.
Playback of Vocalization #1 at times (2/7
trials) resulted in noisy and aggressive behav-
ior, including flights around the source of the
playback, and loud vocalizing. More often,
however, playback of Vocalization #1 elicited
no response or slight response (5/7 trials).
Three playbacks of Vocalization #2 elicited a
stronger response than Vocalization #1; all re-
cipients approached slowly and vocalized
loudly with the same call type. On one occa-
sion, playback of Vocalization #2 resulted in
repeated production of Vocalization #3.
Sociality and Mating System.—We ob-
served two individual Rufous Twistwings to-
46
THE WILSON JOURNAL OF ORNITHOLOGY
•
Vol. 120, No. 1, March 2008
gether on three occasions, twice at Oceanı´a
and once at Extrema. At least two birds were
calling simultaneously within
⬃
25 m of each
other in the afternoon of 11 October 2004 at
Oceanı´a and, the following day, two individ-
uals were captured nearby in mist nets. The
alarm calls of one bird drew the second into
the net suggesting some cohesion between the
two birds, both of which appeared to be males
(with modified primaries).
Playbacks at Extrema resulted in two indi-
viduals giving Vocalization #1 within 15 m of
the observer and within 10 m of each other.
The birds were not seen side-by-side and one
individual (a male) contributed virtually all
vocalizations. It was unclear whether these
observations related to pairs, neighbors, or co-
operative males near a calling post. It is pos-
sible two males were involved, or that repeat-
ed playback and response to playback attract-
ed a female.
One (or two) Rufous Twistwings at Extre-
ma were vocal in a small (30 m
2
) area of tall
bamboo over a period of 4 days; at least one
individual spent much of each day in the vi-
cinity. Three or four vocal individuals at Oce-
anı´a appeared to call regularly from the cen-
tral dense portion of the site with calling birds
appearing to be relatively clumped in distri-
bution.
DISCUSSION
Our results increase the known range of the
Rufous Twistwing and confirm its presence in
Bolivia and Brazil. The five sites from which
the species was known prior to our study were
distributed in two clusters in Peru (Lane et al.
2007) producing a total range of 3,400 km
2
.
Adding the new localities presented here in-
creases the known global range to
⬃
89,000
km
2
(Fig. 1). Our findings are consistent with
the hypothesis that the species is a Guadua
bamboo specialist distributed throughout
southwest Amazonian bamboo forests, as pre-
dicted by Lane et al. (2007). Dense young sec-
ondary forest is also used by the species, usu-
ally in the vicinity of bamboo.
A sighting from Tambopata, Peru (Michael
Kessler, pers. comm.) in 1984, 6 years before
the type specimen was collected at Pakitza,
becomes the first report of the species. The
first record for Brazil is also reported retro-
spectively. EG took a series of in-hand pho-
tographs of a puzzling bird mist netted at Rio
Branco in 1998 and distributed them at the
Brazilian Ornithology Congress (Rio de Ja-
neiro) in the same year. Participants of the
congress reached a consensus that it was a
Brownish Twistwing and it was so listed in a
paper about the avifauna of Parque Zoobotaˆn-
ico (Guilherme 2001). AA and EG re-exam-
ined the evidence in February 2006 including
another specimen collected at the same site in
2003. They exchanged digital images with D.
F. Lane, who confirmed that both records in-
volved the Rufous Twistwing.
The two species of Cnipodectes presumably
overlap in distribution with the Brownish
Twistwing occurring north and south of the
range of the Rufous Twistwing. To the north,
the Brownish Twistwing occurs from Central
America to lowland Amazonia south of the
Amazon River (Meyer de Schauensee 1966,
Lane et al. 2007); it is also known from the
humid forests of Dpto. Pando, Bolivia (Mon-
tambault 2002, Moskovits et al. 2003) includ-
ing Camino Mucden (Parker and Remsen
1987),
⬃
50 km from Extrema. This wide-
spread form inhabits the tangled lower growth
of terra firme forest throughout much of its
range, the only exception being in Bolivia
where it was repeatedly encountered in Guad-
ua bamboo at Camino Mucden (Parker and
Remsen 1987). Bamboo, some of it extensive,
was surveyed in October 1999 at San Sebas-
tia´n (11
⬚
24
⬘
S, 69
⬚
01
⬘
W) on the Rı´o Tahua-
manu and Palmera (11
⬚
30
⬘
S, 69
⬚
03
⬘
W) on
the Rı´o Muyumanu, Dpto. Pando, Bolivia (Al-
verson et al. 2000). Both sites are
⬃
30 km
southeast of Extrema, but no species of Cni-
podectes was found. The apparent conver-
gence in habitat requirements of Brownish
Twistwing and Rufous Twistwing near the
point of contact deserves further study.
The Rufous Twistwing is generally scarce
and has been difficult to relocate at some sites
(Lane et al. 2007). It has been reported only
once at the intensively studied locality of
Tambopata, and once during
⬎
250 hrs of field
work at Estac¸a˜o Ecolo´gica do Rio Acre. In
addition, it is the only regional bamboo spe-
cialist that has not been recorded during 7
years of intensive fieldwork at CICRA. This
absence is particularly of interest as the site
contains extensive bamboo patches supporting
all 19 of Kratter’s (1997) bamboo specialist
47
Tobias et al.
•
RUFOUS TWISTWING BEHAVIOR AND DISTRIBUTION
birds (obligate, near-obligate, and facultative),
along with six other bird species strongly or
seasonally associated with bamboo in this re-
gion. These results suggest the species is rel-
atively specialized, or confined to larger or
older blocks of habitat, as suggested by Lane
et al. (2007).
Results from three sites indicate the Rufous
Twistwing can be relatively common and eas-
ily encountered, especially when vocalizing.
We recorded it 19 times in an 8-day survey at
Oceanı´a and 11 times in a 4-day survey at
Extrema suggesting that, at least in optimum
habitat, encounter rates can be fairly high.
Consistent captures at Rio Branco also indi-
cate that some populations can be sedentary
for periods of several years in suitable habitat.
Overall, our data suggest that populations are
patchy and unpredictable, and vary widely in
density.
The Rufous Twistwing is known to sally
from perches to catch insects and to lift one
or other of its wings repeatedly (Lane et al.
2007). Our results affirm that wing raising is
a common behavior in this species and sug-
gest that wing raises tend to be given in one-
sided bouts. The function of wing raises in
this species and several other genera of tyrant-
flycatcher remains unclear. They may serve as
signals in intersexual contexts, although they
are often given by solitary birds.
The function of modified primaries in both
Cnipodectes species has not been fully ex-
plained, but primaries presumably are used in
sound production (Zimmer 1939, Ridgely and
Greenfield 2001, Lane et al. 2007). The wings
of the male Brownish Twistwing make an au-
dible mechanical pr’r’r’r’r’ wing-rattle in
flight (Hilty and Brown 1986) and a similar
sound has been heard from the Rufous
Twistwing (Lane et al. 2007). It is likely that
sound production in both species is prominent
during courtship display, as in other understo-
ry birds with modified wing feathers, i.e., Pi-
pra manakins and Smithornis broadbills, for
example.
The Rufous Twistwing is more often heard
than seen. The only vocalization previously
reported is an agitation call which closely re-
sembles an aggressive call of the Brownish
Twistwing recorded in Ecuador (Krabbe and
Nilsson 2003). Our vocal data enlarge the
known repertoire of the Rufous Twistwing to
at least three call types. In addition to the ag-
itation call (Vocalization #1) noted by Lane et
al. (2007), we describe a 5–7 note song (Vo-
calization #2), and a 1–3 note pyew call (Vo-
calization #3). One Rufous Twistwing at Ex-
trema, Bolivia responded strongly to playback
of Vocalization #2 and gave this call-type
from a small area over 3 consecutive days,
suggesting that it functions in advertising and
courtship.
The song of the Brownish Twistwing con-
tains 1–3 notes (Fig. 3C in Lane et al. 2007).
It is perhaps analogous to the Rufous Twist-
wing’s pyew call (Vocalization #3), which is
also given 1–3 times in succession (Fig. 2C,
double version and Fig. 3A, triple version). A
relationship between the pyew note of the Ru-
fous Twistwing (Fig. 2C) and the song of the
Brownish Twistwing (Fig. 2D in Lane et al.
2007) is suggested by similarities in their
structure; both have a 2 kHz fundamental with
one major harmonic and a ridge-shaped intro-
duction which leads into a level section. How-
ever, we believe it is unlikely the pyew note
is the main song of the Rufous Twistwing as
it is given rarely and from random points.
Vocalization #2 is given at shorter intervals
from fixed points and may be analogous to the
stereotyped song of the Brownish Twistwing,
although it is a longer, faster series of notes.
The Rufous Twistwing does not appear to give
an upwardly inflected cueet or kuuuu-wit note,
the unsolicited single-note call of the Brown-
ish Twistwing (Fig. 3D in Lane et al. 2007).
It is possible the pyew call (Vocalization #3)
is homologous with the cueet call. This con-
clusion is supported by our observation that
both calls are given at fairly long and irregular
intervals.
The vocal repertoire of the Rufous Twist-
wing is similar to the Brownish Twistwing.
However, presumed homologous vocalizations
are diagnosably different between the two
forms supporting their treatment as separate
species. They may produce different mechan-
ical noise; songs of the Brownish Twistwing
are often introduced by bill snapping sounds
(Hilty and Brown 1986), which have not yet
been reported for the Rufous Twistwing. The
wing-rattles of the two species have not been
compared.
Male Brownish Twistwings sing from ha-
bitual song posts and do not appear to defend
48
THE WILSON JOURNAL OF ORNITHOLOGY
•
Vol. 120, No. 1, March 2008
pair-territories (Parker and Remsen 1987).
They therefore differ from males of most ty-
rant-flycatchers and resemble those of certain
manakin genera (e.g., Neopelma and Tyran-
neutes) with polygamous breeding systems.
Our observations suggest the behavior of male
Rufous Twistwing is similar. The sexual size
dimorphism in both Cnipodectes species is ex-
treme for tyrant-flycatchers (Lane et al. 2007),
again suggesting polygamy. Further research
is required to clarify breeding biology and
ranging behavior in Cnipodectes.
CONSERVATION STATUS
We estimate a global range of
⬃
89,000 km
2
on the basis of all 13 known localities for the
Rufous Twistwing (Fig. 1). However, our re-
sults fit the prediction of Lane et al. (2007)
that the global range coincides with the dis-
continuous block of Guadua–dominated forest
that extends over much of southeastern Peru,
northern Bolivia, and western Amazonian
Brazil. The total area of this habitat is un-
known, but has been estimated at
⬃
121,000
km
2
(Nelson 1994) to
⬃
180,000 km
2
(Saatchi
et al. 2000).
The Rufous Twistwing is probably the least
abundant and perhaps the most threatened of
all bamboo specialists in Amazonia. The risk
of extinction in the short term is low but re-
cent development projects, including the paved
Trans-Oceanica Highway, will increase human
settlement and habitat destruction in the re-
gion (Nepstad et al. 2001, Conover 2003, To-
bias and Brightsmith 2007). This may not be
detrimental to the Rufous Twistwing because
Guadua bamboo at times proliferates on de-
forested land and abandoned chacras (Gris-
com and Ashton 2003; D. F. Lane, pers.
comm.), thereby increasing suitable habitat for
the species. We note, however, that it is usu-
ally recorded in large, mature patches of
Guadua bamboo and appears to be absent
from stretches of more recent Guadua re-
growth along the proposed route of the Trans-
Oceanica Highway. We also note the socio-
economic value of large bamboos and the in-
creasing tendency to harvest them (Bystria-
kova et al. 2004). These factors suggest the
extent of suitable habitat for the species may
decline in the future.
The Rufous Twistwing is more or less re-
stricted to patchy bamboo forests. It is gen-
erally scarce, and we have shown it to be ab-
sent from some apparently suitable sites. More
field data are needed before population size
can be calculated but, for the purpose of clas-
sifying its conservation status according to
current data and IUCN guidelines (IUCN
2001), we estimate this species may have a
declining global population of
⬍
10,000 indi-
viduals, and that all subpopulations may con-
tain fewer than 1,000 individuals. We there-
fore recommend that it be listed on the IUCN
Red List as Vulnerable under criterion C2a (i)
(IUCN 2001) with a caveat that data quality
is poor and that further surveys may lead to
future changes in status. Only a small part of
the proposed distribution of the Rufous
Twistwing has been surveyed and the threat
from habitat disturbance remains unclear.
Field surveys are required throughout the
bamboo forests of southwest Amazonia and
the extent of bamboo should be assessed at
intervals, perhaps through remote sensing
techniques.
ACKNOWLEDGMENTS
Permission for field work was granted in Peru by
Instituto Nacional de Recursos Naturales (INRENA)
and in Brazil by Instituto Brasileiro do Meio Ambiente
e dos Recursos Naturais Renova´ veis (IBAMA). Logis-
tical support and/or funding was provided by SOS
Amazoˆnia and WWF-Brazil (to Alexandre Aleixo),
and Centro Nacional de Pesquisa para Conservac¸a˜o
das Aves Silvestres (CEMAVE) (to Edson Guilherme).
We are indebted to Macaulay Library for loan of sound
recording equipment, and to N. C. A. Pitman and R.
B. Huayllapuma for logistical support. We thank J. M.
Barnett, S. H. M. Butchart, C. D. Cadena, N. G. Ger-
hart (deceased), Michael Kessler, D. F. Lane, P. C. Pul-
garı´n R., Marcos Pe´rsio D. Santos, G. P. Servat, and
Thomas Valqui for useful data, comments, and correc-
tions. Fieldwork was funded in part by a Fulbright
grant (to D. J. Lebbin), a CNPq/SECTAM joint Re-
gional Development Research Program research grant
#35.0415/2004-8 (to Alexandre Aleixo), and a Junior
Research Fellowship from Newnham College, Univer-
sity of Cambridge (to Nathalie Seddon).
LITERATURE CITED
A
LVERSON
, W. S., D. K. M
OSKOVITS
,
AND
J. M. S
HOP
-
LAND
(Editors). 2000. Bolivia: Pando, Rı´o Tahua-
manu. Rapid Biological Inventories Report 1. The
Field Museum, Chicago, Illinois, USA.
B
YSTRIAKOVA
, N., V. K
APOS
,
AND
I. L
YSENKO
. 2004.
Bamboo biodiversity. UNEP-WCMC/INBAR,
Cambridge, United Kingdom.
C
ONOVER
, T. 2003. Peru’s long haul: highway to riches
or ruin. National Geographic June: 80–111.
49
Tobias et al.
•
RUFOUS TWISTWING BEHAVIOR AND DISTRIBUTION
F
ITZPATRICK
, J. W. 2004. Family Tyrannidae (tyrant
flycatchers). Pages 170– 462 in Handbook of the
birds of the world (J. del Hoyo, A. Elliott, and D.
A. Christie, Editors). Volume 9. Cotingas to pipits
and wagtails. Lynx Edicions, Barcelona, Spain.
G
RISCOM
,B.W.
AND
P. M. S. A
SHTON
. 2003. Bamboo
control of forest succession: Guadua sarcocarpa
in southeastern Peru. Forest Ecology and Man-
agement 175:445–454.
G
UILHERME
, E. 2001. Comunidade de aves do campus
e Parque Zoobotaˆnico da Universidade Federal do
Acre, Brasil. Tangara 1:57–73.
H
ILTY
,S.L.
AND
W. L. B
ROWN
. 1986. A guide to the
birds of Colombia. Princeton University Press,
Princeton, New Jersey, USA.
IUCN. 2001. IUCN Red List categories and criteria.
Version 3.1. IUCN Species Survival Commission,
Gland, Switzerland and Cambridge, United King-
dom.
K
RABBE
,N.
AND
J. N
ILSSON
. 2003. Birds of Ecuador,
DVD-ROM. Bird Songs International, Westernie-
land, The Netherlands.
K
RATTER
, A. W. 1997. Bamboo specialization by Am-
azonian birds. Biotropica 29:100–110.
L
ANE
,D.F.,G.P.S
ERVAT
,T.V
ALQUI
H.,
AND
F. R .
L
AMBERT
. 2007. A distinctive new species of ty-
rant flycatcher (Passeriformes: Tyrannidae: Cni-
podectes) from southeastern Peru. Auk 124:762–
772.
M
EYER DE
S
CHAUENSEE
, R. 1966. The species of birds
of South America with their distribution. Acade-
my of Natural Sciences, Philadelphia, Pennsylva-
nia, USA.
M
ONTAMBAULT
, J. R. 2002. Informes de las evalua-
ciones biolo´gicas de Pampas del Heath, Alto Mad-
idi, Bolivia, y Pando, Bolivia. Conservation Inter-
national, Washington, D.C., USA.
M
OSKOVITS
, D. K., W. S. A
LVERSON
,
AND
I. H
ALM
(Ed-
itors). 2003. Bolivia: Pando, Federico Roman,
Rapid Biological Inventories 06. The Field Mu-
seum, Chicago, Illinois, USA.
N
ELSON
, B. W. 1994. Natural forest disturbance and
change in the Brazilian Amazon. Remote Sensing
Reviews 10:105–125.
N
EPSTAD
, D., G. C
ARVALHO
,A.B
ARROS
,A.A
LENCAR
,
J. C
APOBIANCA
,J.B
ISHOP
,P.M
OUTINHO
,P.L
EFEB
-
VRE
,S.U.L
OPES
,
AND
E. P
RINS
. 2001. Road pav-
ing, fire regime feedbacks, and the future of Am-
azon forests. Forest Ecology and Management
154:395–407.
P
ARKER
,T.A.
AND
J. V. R
EMSEN
. 1987. Fifty-two Am-
azonian bird species new to Bolivia. Bulletin of
the British Ornithologists’ Club 107:94–107.
R
IDGELY
,R.S.R.
AND
P. J. G
REENFIELD
. 2001. The
birds of Ecuador: a field guide. Volume 2. Prince-
ton University Press, Princeton, New Jersey, USA.
R
IDGELY
,R.S.R.
AND
G. T
UDOR
. 1994. The birds of
South America: the suboscine passerines. Volume
2. University of Texas Press, Austin, USA.
S
AATCHI
, S. S., B. N
ELSON
,E.P
ODEST
,
AND
J. H
OLT
.
2000. Mapping land cover types in the Amazon
Basin using 1 km JERS-1 mosaic. International
Journal of Remote Sensing 21:1201–1234.
S
CHULENBERG
, T. S., C. A. M
ARANTZ
,
AND
P. H . E
N
-
GLISH
. 2000. Voices of Amazonian birds: birds of
the rainforest of southeastern Peru and northern
Bolivia. Cornell Laboratory of Ornithology, Itha-
ca, New York, USA.
S
ILMAN
, M. R., E. J. A
NCAYA
,
AND
J. B
RINSON
. 2003.
Los bosques de bambu´ en la Amazonı´a occidental.
Pages 63–73 in Alto Puru´s: biodiversidad, con-
servacio´ n y manejo (R. Leite Pitman, N. Pitman,
and P. A
´lvarez, Editors). Center for Tropical Con-
servation, Lima, Peru.
T
OBIAS
,J.A.
AND
D. J. B
RIGHTSMITH
. 2007. Distribu-
tion, ecology and conservation status of the Blue-
headed Macaw Primolius couloni. Biological
Conservation 139:126–138.
Z
IMMER
, J. T. 1939. Studies of Peruvian birds: notes
on the genera Myiotriccus,Pyrrhomyias,Myi-
ophobus,Onychorhynchus,Platyrinchus,Cnipo-
dectes,Sayornis, and Nuttallornis. American Mu-
seum Novitates 1043:1–15.
