Discovered on the brink of extinction: A new species of Pygmy-Owl (Strigidae: Glaucidium) from Atlantic Forest of northeastern Brazil

Abstract

Glaucidium mooreorum sp. nov. (Pernambuco Pygmy-Owl) is described based on two study skins (Ornithological Collection of the Universidade Federal de Pernambuco No. 1030 [holotype] and No. 1017) from the Reserva Biológica de Saltinho (8o44'S, 35o11'W, elevation 140 m), Rio Formoso, Pernambuco, Brazil. This new species is only known from tall second-growth forests at two lowland ( < 150 m) localities in the state of Pernambuco, being possibly related to the Least Pygmy-Owl (G. minutissimum) species complex. Glaucidium mooreorum sp. nov. differs from the two geographically closest species (G. hardyi, from Amazonia, and G. minutissimum from southeastern Brazil) of this complex by plumage coloration, morphometric data, and vocalization. The region in which G. mooreorum sp. nov. was found is a quite distinctive area of endemism (Pernambuco Center) for birds and other groups of organisms in South America. From a historical viewpoint, the avifauna of this area is composed of two distinct elements, as its endemic taxa have their closest relatives either in Amazonia or in the Atlantic Forest south of the São Francisco River. The new species is apparently on the brink of extinction and we present evidence to rank it as "critically endangered". A bioregional planning involving forest restoration in critical areas as well as the establishment of ecological corridors to connect the few remaining large forest fragments is suggested as the best strategy to rescue the unique biota of the Pernambuco Center from extinction.

Full text

Frontispiece. Pernambuco Pygmy-Owl, a new species of the Glaucidium minutissimum complex from the vanishing Atlantic Forest of northeastern
Brazil. From a watercolor painting by Carl Christian Tofte.

123
Ararajuba 10 (2): 123-130
dezembro de 2002
Birds are generally considered as a particularly well-
studied taxon, as most species are thought to have been
described (Diamond 1985). New bird species are currently
described at a rate of three per year, a rate possibly lower
than any other class of vertebrates (LeCroy and Vuilleu-
mier 1992). However, it does not mean that avian taxonomy
at the species level is stable, because new studies on
morphological, vocal and genetic variation have revealed
that numerous taxa described as subspecies are in fact
distinct biological species. A good example is the genus
Glaucidium that comprises a set of small owls widespread
in the Americas, Eurasia and Africa (Peters 1940, Sibley
and Monroe 1990). The number of species recognized in
this genus has increased dramatically in the last 50 years,
going from 13 species listed by Peters (1940) to 29 (Sibley
and Monroe 1990, Howell and Robbins 1995, Robbins
and Stiles 1999).
The species of Glaucidium in the Americas are
separated into four major groups (Howell and Robbins
1995, Robbins and Stiles 1999): The G. gnoma complex
(western Canada to Colombia), the G. brasilianum
complex (southwestern United States to southern South
America), the G. jardinii complex (Costa Rica to South
America), and the G. minutissimum complex (Mexico to
South America). The G. minutissimum (Least Pygmy-Owl)
complex currently comprises six monotypic species that
inhabit tropical and subtropical forests from Mexico to
southern Brazil and Paraguay (Howell and Robbins 1995).
Discovered on the brink of extinction: a new species of Pygmy-Owl
(Strigidae: Glaucidium) from Atlantic Forest of northeastern Brazil
José Maria Cardoso da Silva
1, 2
, Galileu Coelho
2
and Luiz Pedreira Gonzaga
3
1
Conservation International do Brasil. Av. Nazaré 541/310, 66035-170, Belém, PA, Brazil. E-mail: j.silva@conservation.org.br
2
Departamento de Zoologia, Universidade Federal de Pernambuco, Recife, PE, Brazil. E-mail: argamico@aol.com.br
3
Departamento de Zoologia, Universidade Federal do Rio de Janeiro, Caixa Postal 68033, 21944-970, Rio de Janeiro, RJ,
Brazil. E-mail: lpg@biologia.ufrj.br
Recebido em 15 de dezembro de 2002; aceito em 15 de março de 2003.
RESUMO. Descoberta à beira da extinção: uma nova espécie de caburé (Strigidae: Glaucidium) da Floresta Atlântica do nordeste do Brasil.
Glaucidium mooreorum sp. nov. é descrita com base em dois exemplares taxidermizados (Coleção Ornitológica da Universidade Federal de Pernambuco
nº
.
1030 [holótipo] e nº
.
1017), procedentes da Reserva Biológica de Saltinho (8º44´S, 35º11´W, elevation 140 m), Rio Formoso, Pernambuco, Brasil.
A nova espécie é conhecida somente de florestas secundárias altas de duas localidades de baixa altitude ( < 150 m) em Pernambuco, sendo possivel-
mente associada ao complexo de espécies G. minutissimum. As duas espécies desse complexo geograficamente mais próximas são G. hardyi (da
Amazônia) e G. minutissimum (do sudeste do Brasil), das quais G. mooreorum sp. nov. difere pela coloração da plumagem, dados morfométricos e
vocalização. A região onde esta espécie ocorre é uma área de endemismo (Centro Pernambuco) bastante distinta para aves e outros grupos de
organismos sul-americanos. Do ponto de vista histórico, a avifauna endêmica dessa área pode ser classificada como composta, pois ela possui
espécies ou subespécies que têm os seus táxons mais próximos ou na Amazônia ou na Floresta Atlântica ao sul do rio São Francisco. A nova espécie
está aparentemente à beira da extinção e evidências são apresentadas para classificá-la como “criticamente ameaçada”. Um planejamento biorregional
envolvendo a restauração de floresta em áreas críticas e o estabelecimento de corredores ecológicos para conectar os poucos fragmentos grandes de
floresta remanescentes é sugerido como a melhor estratégia para resgatar a biota singular do Centro Pernambuco da extinção.
PALAVRAS-CHAVE: biogeografia, conservação, Glaucidium, Mata Atlântica, Strigidae, taxonomia.
ABSTRACT. Glaucidium mooreorum sp. nov. (Pernambuco Pygmy-Owl) is described based on two study skins (Ornithological Collection of the
Universidade Federal de Pernambuco No. 1030 [holotype] and No. 1017) from the Reserva Biológica de Saltinho (8º44’S, 35º11’W, elevation 140
m), Rio Formoso, Pernambuco, Brazil. This new species is only known from tall second-growth forests at two lowland ( < 150 m) localities in the
state of Pernambuco, being possibly related to the Least Pygmy-Owl (G. minutissimum) species complex. Glaucidium mooreorum sp. nov. differs
from the two geographically closest species (G. hardyi, from Amazonia, and G. minutissimum from southeastern Brazil) of this complex by plumage
coloration, morphometric data, and vocalization. The region in which G. mooreorum sp. nov. was found is a quite distinctive area of endemism
(Pernambuco Center) for birds and other groups of organisms in South America. From a historical viewpoint, the avifauna of this area is composed
of two distinct elements, as its endemic taxa have their closest relatives either in Amazonia or in the Atlantic Forest south of the São Francisco River.
The new species is apparently on the brink of extinction and we present evidence to rank it as “critically endangered”. A bioregional planning
involving forest restoration in critical areas as well as the establishment of ecological corridors to connect the few remaining large forest fragments
is suggested as the best strategy to rescue the unique biota of the Pernambuco Center from extinction.
KEY WORDS: Atlantic forest, biogeography, conservation, Glaucidium, Strigidae, taxonomy.

124 Ararajuba 10 (2): 121-130 J. M. C. da Silva et al.
With the exception of G. palmarum, which inhabits tropical
deciduous forests, all species of the Least Pygmy-Owl
complex inhabit the upperstorey of tall, humid forests,
helping to explain why they are often overlooked and why
there are so few specimens in museums (Howell and
Robbins 1995).
Species of the G. minutissimum complex generally
have small and nearly linear ranges, the only exception
being Amazonian Pygmy-Owl (Glaucidium hardyi), whose
known range includes most of Amazonia (Brazil, Bolivia,
Peru and Guyana) and a single record from Pernambuco
(Reserva Biológica de Saltinho) in the Atlantic Forest of
northeastern Brazil (Vielliard 1989, Howell and Robbins
1995, M. Robbins pers. comm.). Although this record in
the Atlantic Forest is generally referred to in all taxonomic
publications about this species complex, the specimens
upon which it was based have never been examined and
compared with other specimens of G. hardyi.
We examined the only two specimens of pygmy-owls
collected in Pernambuco, and compared them with
specimens of Amazonian Pygmy-Owl from Amazonia as
well as with specimens of Brazilian Pygmy-Owl (G.
minutissimum) from southern Atlantic Forest, the
geographically closest species. In addition, we compared
the only vocal record available for this population with
the voices of other species of the G. minutissimum
complex. We conclude that specimens from northeastern
Brazil represent an undescribed species (see frontispiece).
We propose to call this new owl:
Glaucidium mooreorum sp. nov.
Pernambuco Pygmy-Owl
Holotype. Ornithological Collection of the Univer-
sidade Federal de Pernambuco, No. 1030; adult male (testes
7 x 4 mm). Brazil: Pernambuco, Rio Formoso, Reserva
Biológica de Saltinho, 08º44’S, 35º11’W, elevation 140 m,
collected 6 November 1980 by Galileu Coelho.
Diagnosis. A typical Neotropical pygmy-owl of the
Glaucidium minutissimum species complex (see Howell
and Robbins 1995). The two geographically closest species
(G. hardyi, from Amazonia, and G. minutissimum from
southern Atlantic Forest) are similar in plumage coloration
(figure 1), but the new species differs from both in the
coloration of the following body parts (capitalized color
descriptions and numbers follow Smithe 1975, 1981): (a)
crown between Raw Umber (223) and Raw Umber (123)
rather then between Dark Grayish Brown (20) and Fuscous
(21); (b) back between Raw Umber (223) and Burnt Umber
(22) rather than Dark Grayish Brown; and (c) flanks Raw
Sienna (136) rather than between Raw Umber (123) and
Cinnamon (123a). In addition, there are differences in body
proportions and vocalizations (see below).
Description of holotype. Crown between Raw Umber
(223) and Raw Umber (123). Entire pileum and sides of
face with conspicuous white spots bordered with Sepia
(119). Concealed white nuchal collar, followed by a narrow
belt of Raw Sienna (136). Back, scapulars, upper wing
coverts and rump between Raw Umber (223) and Burnt
Umber (22). Remiges are darker than back, with
conspicuous, irregularly shaped Cinnamon (39) spots on
the outer web and concealed large buffy white (between
Chamois 123D and Buff 124) spots on the inner web. Tail
dark brown (closest to Dusky Brown 19) with 12 rectrices
that have five incomplete white bands that appear as
irregularly shaped dots. The two central rectrices have all
five white bands in both sides. The outer retrices have the
inner side presenting five white bands while the outer side
presents none or only one white band. Center of under-
parts, from chin to crissum, white. Sides of chest Raw
Sienna (136) with a few inconspicuous white spots. Sides
and flanks Raw Sienna with a few white streaks. Abdomen
white with Raw Sienna streaks. Soft part colors: irides
yellow, bill greenish-yellow, tarsi and toes orange-yellow.
Measurements of the holotype. Wing chord 87 mm,
tail length 51.2 mm, bill length (from cere to tip) 11.2 mm,
body mass 51 g.
Paratype. Ornithological Collection of the Universi-
dade Federal de Pernambuco, No. 1017; adult male (testes
7 x 4.5 mm) from the same locality as the holotype,
collected 27 November 1980 by Galileu Coelho.
Measurements of the paratype. Wing chord 87 mm,
tail length 50.6 mm, bill length (from cere to bill) 10.2 mm.
Etymology. We take great pleasure in naming this owl
in honor of the Dr. Gordon Moore and his wife Betty Moore
who have greatly contributed to biodiversity conservation
worldwide and in particular Brazil through a clear vision
and understanding that conservation should be driven by
science and deliver concrete results in the field.
Specimens examined (see acknowledgments for
institutional acronyms). Glaucidium mooreorum sp. nov.:
type locality (2 males, UFPE). Glaucidium hardyi: Brazil:
Rondônia, Cachoeira Nazaré, Rio Jiparaná (3 males,
MPEG; 1 male, FMNH); Parque Estadual de Guajará
Mirim, Guajará Mirim (1 male, MPEG); Pará, Rio
Trombetas (1 male, MPEG), Capim (1 male, MZUSP).
Peru: Madre de Dios, Rio Tambopata (1 male, FMNH).
Glaucidium minutissimum: Mato Grosso do Sul, Santana
do Parnaíba (1 male, MZUSP). São Paulo, Campos do
Jordão (1 male, MZUSP), Ipiranga (1 female, MZUSP),
Presidente Venceslau (1 female, MZUSP).
REMARKS
Body size. The new species has, on average, longer
tail and smaller wing chord than G. hardyi and longer tail
than G. minutissimum (figure 2). There is overlap in tail
length and no overlap in wing chord with G. hardyi but

Ararajuba 10 (2): 121-130 125A new species of Pygmy-Owl (Glaucidium)
Figure 1. Comparisons of Glaucidium hardyi (left), the type of G. mooreorum sp. nov. (middle),
and G. minutissimum (right). (a) dorsal aspect; (b) ventral aspect.

126 Ararajuba 10 (2): 121-130 J. M. C. da Silva et al.
there are overlaps in both body measurements with G.
minutissimum. The ratio tail/wing is useful to separate
these three species, as it ranges between 0.47 and 0.55 in
G. hardyi, between 0.56-0.60 in G. minutissimum and is
above 0.6 in G. mooreorum sp. nov. (figure 2). The validity
of these comparisons needs to be evaluated when more
specimens become available.
songs due to strong reverberation that could not be filtered
out. Frequency measurements were made in the spectro-
gram; given values correspond to the “peak frequency”,
that is the frequency at which the highest amplitude in a
given time interval occurs (Charif et al. 1995: 88). Only
seven songs in the available recording were suitable for
accurate frequency measurements. Comparisons with vocal
characters of other species of New World Glaucidium were
based on the values and spectrograms presented in Vielliard
(1989), Howell and Robbins (1995), and Robbins and
Stiles (1999), which are not repeated here.
The primary song of Glaucidium mooreorum sp. nov.
is a short phrase composed of 5-7 notes, most often (89%,
n = 46 songs) six relatively short notes (87-170 ms, mean
= 133.6 ms, SD = 16.32; n = 56 notes, 11 songs) with a
mean dominant frequency of 1.28 kHz (SD = 0.05; n = 42
notes, 7 songs). Duration of six-note songs varied from
1.37 s to 1.52 s (mean = 1.44 s, SD = 0.04; n = 41 songs),
resulting in a mean pace of 4.2 notes given per second.
The interval between phrases (inter-song interval) is more
variable (4.5-15.3 s, mean 6.5 s; SD = 2.31, n = 26). Each
note shows slightly upward frequency modulation, starting
at an average 1.24 kHz (SD = 0.01) and increasing to an
average 1.33 kHz (SD = 0.03) at approximately the middle
of the note (n = 42 notes, 7 songs). The point where this
shift in frequency occurs corresponds to that of peak
amplitude of the note (figure 3). The mean frequency of
each note decreases slightly from the beginning (1.31 kHz,
SD = 0.007) to the end (1.25 kHz, SD = 0.009) of the
phrase (n = 7 songs). Notes with the greatest peak
amplitude are most often the forth (33%) or the fifth (41%)
80
84
88
92
96
40 44 48 52 56 60
Tail
(
mm
)
Wing (mm)
Figure 2. The relationship between wing length (chord) and tail length
in Glaucidium hardyi (squares), G. minutissimum (circles) and G.
mooreorum sp. nov. (triangles). Body measurements of G. hardyi and
G. minutissimum include, in addition to the specimens examined for
this study, other specimens listed in Robbins and Howell (1995).
Song. The only available tape recording of the natural
vocalization (i.e. no playback involved) of G. mooreorum
sp. nov. was made by G.C. on 24 October 1990 at the type
locality, using a National RQ 5015 open-reel tape recorder
with a National WM 2095N dynamic microphone at a tape
speed of 9.5 cm/s, and is housed at the Arquivo Sonoro
Prof. Elias Coelho (ASEC No. 11204), Departamento de
Zoologia, Universidade Federal do Rio de Janeiro. This
recording was digitized at 44.1 kHz, 16-bit and sound
measurements and spectrograms were made in a Power
Macintosh 7200/120 microcomputer with Canary software
(v. 1.2.4) from Cornell Laboratory of Ornithology
(Bioacoustics Research Program).
The song characters examined were number of notes
and duration of each note in a song, song duration
(measured from the beginning of the first note to the end
of the last one), pace (derived from number of notes and
song duration), inter-song intervals, dominant frequency
of notes and relative amplitude of notes in a song. Time
measurements were made in the waveform after the sound
has been filtered above 1.5 kHz and below 1.0 kHz to
remove most of background noise. In spite of this, accurate
measurements of note duration were not possible in most
Figure 3. Oscillogram (top panel) and narrow-band (filter bandwidth
266 Hz, Blackman window) spectrogram (bottom panel) of a note of
the song of Glaucidium mooreorum sp. nov., showing the amplitude
peak at the time when upward frequency modulation occurs.

Ararajuba 10 (2): 121-130 127A new species of Pygmy-Owl (Glaucidium)
(n = 42 songs), so that these phrases are slightly accented
just before the end (figure 4).
Both the upward shift in frequency of individual notes
and the decrease in mean frequency of notes along the
song may prove to be unique among New World
Glaucidium. These characters have not been reported for
any other species in the G. minutissimum species complex
(Vielliard 1989, Howell and Robbins 1995), nor for G.
brasilianum, whose song has been described as having
“arched” (i.e. with an upward followed by a downward
modulation) notes (Howell and Robbins 1995), or species
in the G. gnoma and G. jardinii complexes (Robbins and
Stiles 1999). However, the shift in frequency of short notes
cannot be easily distinguished in wide-band, small-scaled
spectrograms (L. P. G. pers. obs.) such as those used by
all these authors, so that further analyses are warranted
for this character.
In addition, the song of G. mooreorum sp. nov. differs
from that of G. hardyi (12-36 notes, 30-40 ms long) by
having fewer and longer notes, delivered at a slower pace
and at much shorter inter-song intervals. From that of G.
minutissimum (2-4, typically two notes, 260 ms long), the
song of G. mooreorum sp. nov. differs by having more,
shorter notes. The song of G. mooreorum sp. nov. also has
more notes than that of Subtropical Pygmy-Owl (G.
parkeri), which has 2-4, typically three, notes, and that of
Tamaulipas Pygmy-Owl (G. sanchezi), which has 1-3 notes
(Howell and Robbins 1995, Robbins and Howell 1995).
Distribution and habitat. Glaucidium mooreorum
sp.nov. is known from only two localities in Pernambuco
(figure 5): Reserva Biológica de Saltinho, a protected area
of 4.8 km
2
composed primarily of old secondary forests;
and Usina Trapiche, Sirinhaém (08º35’S, 35º07’W), where
L. P. Gonzaga, A. Long and L. C. Marigo observed one
individual of this species, eating a large cicada high up in
the canopy near the edge of a forest belt, with about 100
hectares, on 19 November 2001. The elevation of these
forests varies between 0 and 150 m a.s.l. The new species
has not been recorded in other well-surveyed forest sites
of the region, such as Murici and Usina Serra Grande,
whose elevation varies from 400-600 m a.s.l..
Biogeography. The range of G. mooreorum sp. nov. is
separated from the nearest records of G. hardyi and G.
minutissimum by a straight line distance of 1,580 km and
820 km, respectively (figure 5). The gap between the
ranges of G. hardyi and that of the new species is occupied
by a tropical dry vegetation known as caatinga (Andrade-
Lima 1981). Because both pygmy-owl species are
associated with tall, humid forests at low elevations rather
than other habitat types, they possibly do not occur in
caatinga environments. However, within the Caatinga
region there are small-to-medium size patches of humid
forests located on the slopes of some old isolated plateaus
(figure 5). These patches of humid forests are regionally
known as “brejos”. Some of these forest enclaves have
been already sampled for birds, but only G. brasilianum
has been recorded (Pinto and Camargo 1961; G. C. and
J.M.C.S., unpubl. data). The ranges of G. minutissimum
and G. mooreorum sp. nov. are apparently separated by
the São Francisco River, although G. minutissimum has
not been recorded yet in the few forest fragments between
its northernmost record and this river (figure 5).
The region in which G. mooreorum sp. nov. was found
is regarded as an important area of endemism in eastern
South America for terrestrial vertebrates (Müller 1973),
vascular plants (Prance 1987), forest butterflies (Tyler et
al. 1994) and birds (Stattersfield et al. 1998). This region,
named Pernambuco Center or the Pernambuco area of
Figure 4. Oscillogram (top panel) and narrow-band (filter bandwidth 266 Hz, Blackman window) spectrogram (bottom panel) of a six-note song
of Glaucidium mooreorum sp. nov., showing slight decrease in mean frequency of notes and amplitude peak at the fifth note.

128 Ararajuba 10 (2): 121-130 J. M. C. da Silva et al.
endemism, includes the entire block of coastal Atlantic
Forests north of the São Francisco River (figure 5). From
a biogeographic viewpoint, the avifauna of the Pernambuco
Center is composed of two distinct historical elements, as
its endemic taxa have their closest relatives either in
Amazonia or in Atlantic Forests south of the São Francisco
River. Examples of endemic taxa with Amazonian sister
relatives are Mitu mitu, Pyriglena leuconota pernambu-
censis, and Cercomacra laeta sabinoi. Examples of
endemic taxa with southern Atlantic Forest closest relatives
are Philydor novaesi, Myrmotherula snowi, Terenura sicki,
and Phylloscartes ceciliae. Teixeira and Gonzaga (1983)
suggested that these two biogeographic elements present
an altitudinal segregation along the region, with the
Atlantic Forest elements occurring, on average, at higher
altitudes than the Amazonian ones. Based on this
biogeographic pattern and in its known altitudinal
distribution, G. mooreorum sp. nov. may be more closely
related to G. hardyi than to G. minutissimum. However,
this prediction needs to be evaluated with a phylogenetic
survey of the entire G. minutissimum species complex.
Conservation. The Pernambuco Center is by far the
most modified region of Atlantic Forest. From 39,567 km
2
of original forest, only 1,907 km
2
are left. The remaining
forest does not form a single and large block, but it is
distributed in more than 1,400 fragments that are mostly
small ( < 100 hectares) and isolated in a matrix dominated
by agriculture fields and urban environments. There are
52 reserves in this region, summing up 238.9 km
2
.
However, this protected area system is far from adequate
to protect regional biodiversity. There are two main reasons
for this statement. First, reserves are small (78% are smaller
than 5 km
2
), isolated and do not encompass most of the
environmental heterogeneity of the region. Second,
although the creation of these reserves helped to halt
deforestation of the targeted forest patches, most of them
Figure 5. Ranges of Glaucidium hardyi (squares), G. mooreorum sp. nov. (circles), and G. minutissimum (triangles). These records are based on
specimens examined, literature and a new record provided by M. Robbins.

Ararajuba 10 (2): 121-130 129A new species of Pygmy-Owl (Glaucidium)
are not fully functional, as they have not any structure and
personal (Uchôa-Neto 2002).
Forest patches outside reserves, in spite of all legal
restrictions approved in the last decades, continue to be
logged and burned. Large vertebrates, such as birds and
mammals, that play an important role as seed dispersers
of key canopy tree species have been extirpated because
of intensive illegal hunting (Silva and Tabarelli 2000,
2001). These factors plus edge effect contribute to that
these fragments undergo dramatic changes in both
composition and vegetation structure, leading to local
extinctions of several species of plants and animals that
are totally dependent of tall, mature forests (Tabarelli et
al. 2002).
Glaucidium mooreorum sp. nov. is described already
on the edge of extinction. Although it is known to occur
in a biological reserve, this reserve seems to be small (4.8
km
2
) to maintain viable populations of this species. Based
on IUCN’s criteria for classifying species at high risk of
global extinction (BirdLife International 2000), the new
species should be ranked as “critically endangered”. Three
reasons support this classification. First, the species’
estimated extent of occurrence is less than 100 km
2
.
Second, forests within the species’ range are severely
fragmented. Third, these forests continue to decline in
quality as a result of the ever-growing expansion of human
activities in the region.
In addition to G. mooreorum sp. nov., the Pernambuco
Center harbors 17 globally threatened species (BirdLife
International 2000): one is already extinct in the wild (Mitu
mitu), five are classified as critically endangered
(Synallaxis infuscata, Philydor novaesi, Myrmotherula
snowi, Phylloscartes ceciliae, and Curaeus forbesi), six
are considered as endangered (Amazona rhodocorytha,
Touit surda, Terenura sicki, Myrmeciza ruficauda,
Iodopleura pipra, Xipholena atropurpurea, and Tangara
fastuosa), and four are classified as vulnerable (Leucop-
ternis lacernulata, Herpsilochmus pectoralis, Hemitriccus
mirandae, and Carpornis melanocephalus). Because most
of these threatened species are endemic to the Pernambuco
Center, this region ranks as one of the South America’s
most important regions for urgent conservation action to
prevent extinction of a number of bird species. Unfor-
tunately, efforts for protecting forests in the Pernambuco
Center have always been directed towards establishing
protected areas that are small-to-medium size forest
remnants selected based on the presence of one or more
threatened species as well as of some critical environmental
feature, such as headwaters. We suggest that a regional
strategy, based on the guidelines proposed by Soulé and
Terborgh (1999), rather than site-specific strategies, is
required if the biodiversity of Pernambuco Center is to be
rescued from the brink of extinction. Silva and Tabarelli
(2000) pointed out three major strategies for the creation
of an effective conservation network in this region: (a)
protection of landscapes composed of archipelagos of
fragments, including the present-day reserves, connected
by corridors of original or restored vegetation which must
represent several thousand hectares of forest; (b) in
addition to connect local landscapes, it is necessary also
to connect landscapes at a regional level to guarantee
maximum representation of all important habitat types that
exist in this region; (c) large fruit-eating birds and
mammals should be elevated to the category of “umbrella
species” to indicate the minimum area required in this
regional network to maintain the key ecological processes
responsible for forest maintenance and regeneration.
Implementation of this conservation network will require
visionary leaders, long-term funding schemes, good basic
scientific data, new economic mechanisms to promote a
new biodiversity-friendly regional economy, decision-
support systems that take in account risks for biodiversity
in major infra-structure and economic programs, and,
mostly, permanent support from both local and interna-
tional communities.
ACKNOWLEDGMENTS
We are grateful to Mark Robbins for providing
information on body measurements of specimens of
Amazonian Pygmy-Owl and Brazilian Pygmy-Owl housed
in US collections. J. Bates and S. Hackett afforded access
to the collection of the Field Museum of Natural History
(FMNH). M. de Vivo and M. de Pinna provided timely
loan of specimens of the Least Pygmy-Owl species
complex housed at Museu de Zoologia da Universidade
de São Paulo (MZUSP). Maria Luiza V. Marceliano gave
J.M.C.S. permit to examine specimens housed at Museu
Paraense Emílio Goeldi (MPEG). Carl Christian Tofte
painted the attractive color plate, and Luís Barbosa
produced the figures. Studies by G.C. and J.M.C.S.
received support from Universidade Federal de Pernam-
buco (UFPE), FACEPE, CNPq, and Center of Biodiversity
Conservation (CBC) of the Conservation International-
Brazil. Support to the Arquivo Sonoro Prof. Elias Coelho
of UFRJ came from the Fundação Universitária José
Bonifácio. The visit of L. P. G. to the Usina Trapiche was
sponsored by BirdLife International and had the local
support of Clodoaldo José Bakker Faria and Mário Jorge
P. Seixas Aguiar.
REFERENCES
Andrade-Lima, D. (1981) The caatingas dominium. Re-
vista Brasileira de Botânica 4:149-153.
BirdLife International (2000) Threatened birds of the
world. Barcelona and Cambridge, UK: Lynx Editions
and BirdLife International.

130 Ararajuba 10 (2): 121-130 J. M. C. da Silva et al.
Charif, R. A., S. Mitchell and C. W. Clark (1995) Canary
1.2 user’s manual. Ithaca, New York: Cornell
Laboratory of Ornithology.
Diamond, J. (1985) How many unknown species are yet
to be discovered? Nature 315:538-539.
Howell, S. N. G. and M. B. Robbins (1995) Species limits
of the Least Pygmy-Owl (Glaucidium minutissimum)
complex. Wilson Bull. 107:7-25.
Lecroy, M. and F. Vuilleumier (1992) Guidelines for the
description of new species in ornithology. Bull. Brit.
Orn. Cl., Centenary Supplement 112A:191-198.
Müller, P. (1973) The dispersal centers of terrestrial
vertebrates in the neotropical realm: a study in the
evolution of the neotropical biota and its native
landscapes. Haag: Dr. W. Junk.
Peters, J. L. (1940) Check-list of birds of the world, v. 4.
Cambridge, Massachusetts: Harvard Univ. Press.
Pinto, O. M. O. and E. A. de Camargo (1961) Resultados
ornitológicos de quatro recentes expedições do Depar-
tamento de Zoologia ao Nordeste do Brasil, com a des-
crição de seis novas subespécies. Arqs. Zool., São Paulo
11:193-284.
Prance, G. T. (1987) Biogeography of neotropical plants,
pp. 175-196. In: T. C. Whitmore and G. T. Prance
(eds.).Biogeography and quaternary history in tropi-
cal America. Oxford, UK: Clarendon Press.
Robbins, M. B. and S. N. G. Howell (1995) A new species
of Pygmy-Owl (Strigidae: Glaucidium) from the eastern
Andes. Wilson Bulletin 107:1-6.
_______ and F. G. Stiles (1999) A new species of Pygmy-
Owl (Strigidae: Glaucidium) from the Pacific slope of
the Northern Andes. Auk 116:305-315.
Sibley, C. G. and B. L. Monroe Jr. (1990) Distribution
and taxonomy of birds of the world. New Haven,
Connecticut: Yale Univ. Press.
Silva, J. M. C. and M. Tabarelli (2000) Tree species
impoverishment and the future flora of the Atlantic
forest of northeast Brazil. Nature 404:72-73.
_______ and _______ (2001) The future of the Atlantic
forest in Northeastern Brazil. Cons. Biol. 15:819-820.
Smithe, F. B. (1975) Naturalist’s Color Guide. New York,
New York: American Museum of Natural History.
_______ (1981) Naturalist’s Color Guide, part III. New
York, New York: American Museum of Natural History.
Soulé, M. E. and J. Terborgh (1999) Continental
conservation: scientific foundations of regional reser-
ve networks. Washington, D.C.: Island Press.
Stattersfield, A. J., M. J. Crosby, A. J. Long and D. C.
Wege (1998) Endemic bird areas of the world:
priorities for bird conservation. Cambridge, UK:
BirdLife International.
Tabarelli, M., J. F. Marins and J. M. C. Silva (2002) La
biodiversidad brasileña amenazada. Investigacion y
Ciencia 308:42-49.
Teixeira, D. M. and L. P. Gonzaga (1983) A new antwren
from northeastern Brazil. Bull. Brit. Orn. Cl. 103:133-
135.
Tyler, H., K. S. Brown Jr. and K. Wilson (1994)
Swallowtail butterflies of the Americas: a study in
biological dynamics, ecological diversity, biosyste-
matics and conservation. Gainesville: Scientific
Publishers.
Uchôa-Neto, C. A. (2002) Integridade, grau de implemen-
tação e viabilidade das unidades de conservação de
proteção integral na Floresta Atlântica de Pernambuco.
M. Sc. Dissertation. Recife: Universidade Federal de
Pernambuco.
Vielliard, J. (1989) Uma nova espécie de Glaucidium
(Aves, Strigidae) da Amazônia. Revta bras. Zool. 6:685-
693.