ArticlePDF Available

Phylogeography of the Robsonius Ground-Warblers (Passeriformes: Locustellidae) Reveals an Undescribed Species from Northeastern Luzon, Philippines (La Filogeografía de Robsonius (Passeriformes: Locustellidae) Revela una Especie No Descripta del Noreste de Luzón, Filipinas)

Authors:

Abstract and Figures

The Robsonius ground-warblers are forest birds endemic to the Luzon Island complex in the Philippine archipelago. Their systematic relationships have long remained ambiguous; until recently they were included in the timaliid genus Napothera. Two Robsonius species are currently recognized on the basis of plumage differences: R. rabori from northern Luzon in the Cordillera Central and the northern Sierra Madre, and R. sorsogonensis from southern Luzon and Catanduanes Island. Recent specimen collections, including the first adult specimen from the Cordillera Central, establish plumage differences between populations of R. rabori in the Cordillera Central and Sierra Madre and reveal a third diagnosable population within Luzon. These differences have gone unnoticed because R. rabori (sensu stricto) had been known only from the juvenile holotype. Molecular phylogenetic data further support the hypothesis that three highly divergent taxa occur across the Luzon Island complex: Robsonius rabori is known only from the northern Cordillera Central in Ilocos Norte; an undescribed taxon (formerly included in R. rabori) occurs in the northern Sierra Madre in Cagayan, Isabela, Aurora, and Nueva Vizcaya provinces; and R. sorsogonensis occurs in southern Luzon (Bulacan and Laguna provinces), the Bicol Peninsula, and on Catanduanes Island. The existence of three putatively allopatric species within the Luzon island complex highlights the role of in situ diversification in island systems, and brings attention to the need for forest conservation to protect geographically restricted populations throughout the Luzon Island complex.
Content may be subject to copyright.
VOL U M E
115
NU M BE R
3
AUG UST
201 3
PU B L I S H E D BY T H E CO O P E R OR N I T HOL O G I C A L SO C I E T Y
COVER: Bicol (Robsonius sorsogonensis, top), Sierra Madre (R. thompsoni, center; newly described in this issue, Hosner et al., pp. 630–639) and Cordilleran
(R. rabori, bottom) Ground-Warblers; three allopatric species of the forest understory of Luzon and Catanduanes islands, Philippines. Original watercolor
by Marco A. Pineda M.
630
6E-mail: hosner@ku.edu
Manuscript received 25 July 2012; accepted 11 February 2013.
The Condor, Vol. 115, Number 3, pages 630–639. ISSN 0010-5422, electronic ISSN 1938-5422. 2013 by The Cooper Ornithological Society. All rights reserved. Please direct all
requests for permission to phot ocopy or reproduce article content t hrough the Un iversity of Californ ia Press’s Rights and Per missions website, htt p://www.ucpressjour nals.com/
reprintInfo.asp. DOI: 10.1525/cond.2013.120124
PHYLOGEOGRAPHY OF THE ROBSONIUS GROUND-WARBLERS
(PASSERIFORMES: LOCUSTELLIDAE) REVEALS AN UNDESCRIBED SPECIES
FROM NORTHEASTERN LUZON, PHILIPPINES
PETER A. HOSNER1,2,6, NIKKI C. BOGGESS1,2, PHILLIP ALV IO L A3, LUIS A. SÁNCHEZ-GONZÁLEZ2,4,
CARL H. OLIVEROS1,2, ROLLY URRIZA5, AND ROBERT G. MOY LE1,2
1Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence KS 66045
2Biodiversity Institute, Universit y of Kansas, Lawrence KS 66045
3Museum of Natural History, University of the Philippines Los Baños, Los Baños, Laguna, Philippines
4Museo de Zoología “Alfonso L. Herrera,” Departamento de Biología Evolutiva, Facultad de Ciencias, Universidad Nacional
Autónoma de México, Apartado Postal 70-399, México, D.F. 04510, México
5Ornithology Section, Zoology Division, Philippine National Museum, Rizal Park, Burgos St., Manila, Philippines
Abstract. The Robsonius ground-warblers are forest birds endemic to the Luzon Island complex in the Philip-
pine archipelago. Their systematic relationships have long remained ambiguous; until recently they were included
in the timaliid genus Napothera. Two Robsonius species are currently recognized on the basis of plumage differ-
ences: R. rabori from northern Luzon in the Cordillera Central and the northern Sierra Madre, and R. sorsogonensis
from southern Luzon and Catanduanes Island. Recent specimen collections, including the first adult specimen from
the Cordillera Central, establish plumage differences between populations of R. rabori in the Cordillera Central
and Sierra Madre and reveal a third diagnosable population within Luzon. These differences have gone unnoticed
because R. rabori (sensu stricto) had been known only from the juvenile holotype. Molecular phylogenetic data
further support the hypothesis that three highly divergent taxa occur across the Luzon Island complex: Robsonius
rabori is known only from the northern Cordillera Central in Ilocos Norte; an undescribed taxon (formerly included
in R. rabori) occurs in the northern Sierra Madre in Cagayan, Isabela, Aurora, and Nueva Vizcaya provinces; and
R. sorsogonensis occurs in southern Luzon (Bulacan and Laguna provinces), the Bicol Peninsula, and on Catan-
duanes Island. The existence of three putatively allopatric species within the Luzon island complex highlights the
role of in situ diversification in island systems, and brings attention to the need for forest conservation to protect
geographically restricted populations throughout the Luzon Island complex.
Key words: endemism, Napothera, Philippines, phylogeography, diversification.
La Filogeografía de Robsonius (Passeriformes: Locustellidae) Revela una Especie No Descripta
del Noreste de Luzón, Filipinas
Resumen. Las aves del genero Robsonius son endémicas de los bosques de la geológicamente compleja isla
Luzon en el archipiélago de Filipinas. Sus relaciones sistemáticas han permanecido ambiguas por mucho tiempo;
hasta hace poco, estuvieron incluidas en el género Napothera. Actualmente se reconocen dos especies de Robso-
nius con base en diferencias de plumaje: R. rabori del norte de Luzon en la Cordillera Central y el norte de Sierra
Madre, y R. sorsogonensis del sur de Luzón y la isla Catanduanes. Colecciones recientes de especímenes, incluy-
endo el primer espécimen adulto de la Cordillera Central, permitieron diferencias de plumaje entre las poblaciones
de R. rabori en la Cordillera Central y Sierra Madre y revelaron una tercera población diagnosticable. Estas dife-
rencias han pasado inadvertidas debido a que R. rabori (sensu stricto) ha sido conocida solo por el holotipo juvenil.
Datos moleculares filogenéticos refuerzan el apoyo a la hipótesis de que tres taxones altamente divergentes se en-
cuentran a lo largo del complejo de la isla Luzón: R. rabori es conocida sólo del norte de la Cordillera Central en
Ilocos Norte; un taxón no descrito (anteriormente incluido en R. rabori) se encuentra en el norte de Sierra Madre
en las provincias de Cagayan, Isabela, Aurora y Nueva Vizcaya; y R. sorsogonensis, que se encuentra en el sur
de Luzón (provincias de Bulacan y Laguna), la península Bicol y en la isla Catanduanes. La existencia de tres es-
pecies alopátricas putativas dentro Luzón destaca del papel de la diversificación in situ en los sistemas de islas y
resalta la necesidad de conservar el bosque para proteger poblaciones geográficamente restringidas a través de la
compleja isla de Luzón.
The Cond or 115(3):630 –639
The Cooper Ornithological Society 2013
NEW PHILIPPINE GROUND-WARBLER 631
INTRODUCTION
In 1959, D. S. Rabor led an expedition to Ilocos Norte Prov-
ince in northwestern Luzon Island, Philippines. Inland from the
small coastal village of Pagudpud, the team collected a single
juvenile of a long-legged, long-tailed, medium-sized passerine
that differed strikingly from any known species. Rand (1960)
described the species as Napothera rabori, believing it was
related to southeast Asian Napothera babblers (Timaliidae).
Rabor secured two more juveniles at Mt. Cagua in Cagayan
Province, northeastern Luzon, the following year, and four
more specimens (adults and juveniles) from Sorsogon Province
at the southern end of Luzon in 1961. The southern birds dif-
fered from the northern birds in lacking the rusty tinge on the
head present in the juvenile northern specimens. Rand and
Rabor (1967) named the southern birds N. sorsogonensis, after
the type locality. Thus, within two years, two species of a group
of unique and enigmatic passerines were discovered from the
opposite ends of Luzon Island.
Approximately a decade after the initial discoveries, DuPont
(1971a, b) received an adult specimen from Laguna, in central
Luzon, and declared that it was intermediate between N. rabori
and N. sorsogonensis, notwithstanding that he was comparing
adults and juveniles. He lumped N. rabori and N. sorsogonensis
into a single species and named the Laguna specimen as a new
subspecies, N. r. mesoluzonica.
Ornithological exploration in the 1980s and 1990s began
to clarify the natural history and distribution of the rabori
complex, until then virtually unknown in life, which led to
questions about its phylogenetic relationships. Goodman and
Gonzales (1990) observed an individual flipping leaf litter and
wood debris while walking on the ground, presumably for-
aging for insects. De Roever (1990) observed an individual
walking and running with its tail cocked and likened it to a
small rail or a neotropical antthrush. Lambert (1993) observed
a pair walking and foraging on the forest floor, noted that this
behavior would be unusual for Napothera (other species of
Napothera hop), and suggested that the species may belong
in another genus. Harrap and Mitchell (1994) described the
song—high-pitched phrases given from a horizontal branch
or log—as similar to those of warblers of the genera Bradyp-
terus or Urosphena, and, on the basis of song and behavior,
advocated placing the complex in its own genus.
Collar (2006) synthesized the natural history observations
and morphological evidence and proposed for the rabori com-
plex a new genus, Robsonius, within the Timaliidae. On the
basis of four distinctive plumage differences between the two
taxa, he also returned to Rand and Rabor’s (1967) treatment of
two species, R. rabori and R. sorsogonensis (the later including
R. s. mesoluzonica). In a comprehensive molecular phylogeny
of the babblers, Moyle et al. (2012) found that Robsonius fell
far outside of the main babbler lineages, further justifying its
removal from Napothera. Most recently, Oliveros et al. (2012)
reported that Robsonius is sister to the grassbirds and allies
(Locustellidae), and they coined a new English name for the
genus, the ground-warblers.
In June 2011, a field team from the University of Kansas
Biodiversity Institute, Philippine National Museum, and
University of Utah visited the forests of Ilocos Norte to
survey terrestrial vertebrates and their parasites. The team
surveyed two localities south of the small village of Adams,
only 5–10 km from where Rabor and his team collected the
unique juvenile holotype of R. rabori. We collected an adult,
salvaged from a mammal trap, which differed in several
plumage characters from all other adult specimens of Rob-
sonius (Appendix 1, available at http://dx.doi.org/10.1525/
cond.2013.120124). Because of these plumage differences,
we investigated the molecular phylogeographic structure
within the genus to assess whether the differences reflect
individual variation within a poorly known species or a suite
of distinct, diagnosable taxa. Analysis of the data revealed
(Fig. 1) that populations from the southern, northeastern, and
northwestern sectors of Luzon are genetically (on the basis
of mitochondrial DNA and nuclear introns) and morphologi-
cally distinct, and that an unnamed lineage of ground- warbler
is present in the northern Sierra Madre of northeastern Luzon
Island (Fig. 2). Because the name R. rabori applies to the
species occurring in the Cordillera Central in Ilocos Norte
(northwestern Luzon), the northern Sierra Madre birds, long
attributed to R. rabori, remain undescribed, long hidden from
taxonomic recognition because the adult plumage of R. rabori
sensu stricto was unknown. We proudly name this species
Robsonius thompsoni sp. nov.
Sierra Madre Ground-Warbler
Holotype. Philippine National Museum (PNM) 20006; origi-
nally catalogued as University of Kansas Biodiversity Insti-
tute (KU) 114678, adult female (skull 100% pneumatized, no
bursa), KU tissue number 19632, collected on 18 June 2009 in
the Philippines, Luzon Island, Aurora Province, San Luis Mu-
nicipality, 12 km SW Baler (15.680° N, 121.529° E, elevation
525 m). This individual was net-captured in secondary lowland
forest and prepared as a study skin by Jameson B. Reynon.
Description of holotype. Adult female; ovary 6 × 3 mm;
light fat; mass 63 g; molt on wing, breast, and nape; stom-
ach contents insect parts; maxilla dark brown, mandible
pale gray; iris dark brown; legs and feet light brown. Col-
ors in the following description of the plumage are those of
Smithe (1975). Crown and nape amber, the feathers tipped
with dusky brown; auriculars amber. Lores white, the feath-
ers tipped with black. Thin eye ring whitish; small area of
bare gray skin behind eye. Throat white, the feathers tipped
with black; malar stripe black, formed of feathers with white
bases; submoustachial stripe white with black feather edging.
Black feather tips on the lower throat and upper breast form
a necklace of spots, which separates the primarily white
632 PETER A. HOSNER ET A L.
throat from the gray breast. Breast medium neutral gray,
with feather shafts slightly paler; belly whitish; flanks
dusky brown, with cinnamon-brown to chestnut tinge. Back
cinnamon-brown with dusky brown tips to individual feath-
ers. Long fluffy rump feathers form a thick mat and are
dusky brown with a chestnut tinge; white feather tips form a
concealed white rump band; uppertail and undertail coverts
and tail dusky brown with chestnut tinge. Wings chestnut to
dusky brown, each feather dusky brown with a broad chest-
nut edge, so that the wing appears mostly chestnut when
folded. Alula and wing coverts broadly tipped with white,
so that the folded wing has several bars of white spots. Outer
three primaries also tipped with white, although not visible
in the folded wing.
Diagnosis. Adult plumage: Robsonius thompsoni (cover;
Fig. 3C) is most similar to R. rabori (Fig. 3A) but differs in
three plumage characters: presence of a necklace of black spots
(lacking in R. rabori), black feather tips on the throat and sub-
moustachial area (pale gray in R. rabori), and a uniform gray
breast with pale feather shafts and faint darker edging (in
R. rabori, breast feathers have broad white bases and centers and
gray edges, lending the breast a scaled appearance). Robsonius
FIGURE 1. Bayesian consensus tree of the concatenated dataset of six genes (cytochrome b, ND2, ND3, TGFb2-5, Fib-5, MUSK) and
networks of phased haplotypes of three nuclear introns. Each shade in the haplotype networks corresponds to one of the three clades of
Robsonius; black squares represent unsampled haplotypes. Values at nodes indicate support by Bayesian posterior probability/maximum-
likelihood bootstrap; scale bar indicates 0.03 substitutions per site.
NEW PHILIPPINE GROUND-WARBLER 633
thompsoni d iffers fr om R. sorsogonensis (Fig 3E) in four plumage
characters: amber crown, nape, and auriculars (uniform dark gray
with white feather shafts on the auriculars in R. sorsogonensis),
dark gray to blackish tips on throat feathers (unmarked white in
R. sorsogonensis), and pale feather shafts in the gray breast band
(uniform gray without pale feather shafts or feather bases in R. sorso-
gonensis). Juvenile plumage: from a limited number of specimens,
R. thompsoni (n = 4) and R. rabori (n = 1) are not distinguishable
(Fig. 3B and 3D). Overall the juvenile plumage is similar to
the adult plumage, but the throat, back, and underparts are
variably cinnamon brown to olive brown with paler feather
bases; crown, nape, and auriculars similar to those of the
adult but lores and eye ring uniform amber. The juvenile of
R. sorsogonensis (Fig 3F; n = 4) is similar to that of R. rabori and
R. thompsoni except that the underparts (especially flanks) are
richer chestnut brown and the crown, nape, auriculars, lores, and
eye ring are cinnamon brown.
Vocalizations. All three species of Robsonius sing similar
extremely high-pitched (7.5–10.0 kHz) songs, from the ground
or an elevated perch. Each song bout is approximately 1.6–2.2 sec
in duration and generally consists of three or four variable
phrases with ascending and descending notes. Each phrase is
separated by a brief (0.1–0.2 sec) pause. Individuals give song
bouts approximately every 5–10 sec when singing regularly.
Because of small sample sizes (rabori, n = 1; thompsoni, n = 10;
sorsogonensis, n = 2), it is currently unclear whether slight
differences in songs represent geographic or individual varia-
tion. More recordings are needed from additional localities
to assess whether or not each species may be identified solely
by vocalizations. In addition to the song, a rapid trill, thought
to be an alarm call, has been recorded from one of a group of
R. thompsoni (P. Noakes, Xeno-canto [XC] 40990).
Designation of paratypes. Field Museum of Natural History
(FMNH) 472602; adult male (skull pneumatized, no bursa)
captured 10 April 2010, Philippines, Luzon Island, Aurora Prov-
ince, Dunalungan Municipality, 1.9 km S, 4.0 km E Mt. Anacuao
(16.237° N, 121.927° E; 1300 m), in primary lower montane forest.
This specimen was originally prepared as a fluid specimen in for-
malin (DSB 7110) but re-prepared as a skin by PAH: mass 57 g;
fat moderate; stomach with sclerotized insect fragments; molt on
wing and body; testes 5 × 4 mm. KU 119893; juvenile male (skull
not pneumatized, bursa 10 × 8 mm), tissue number KU 25788. This
individual was net-captured on 7 July 2011 in the Philippines, Lu-
zon Island, Cagayan Province, Gonzaga Municipality, Mt. Cagua
crater (13. 219° N, 122.11E; 780 m) and prepared as a study skin
by PAH: fat light; mass 52.5 g; stomach empty; iris dark brown;
legs dusky; maxilla dusky with yellow tomium; mandible yellow
with dusky tomium; molt on body; mouth lining yellow.
Variation within Robsonius. We found no apparent signif-
icant differences in size (Baldwin 1931, Winker 1998) among
the three species of Robsonius (Table 1; ANOVA, P > 0.01)
or between the sexes (t-test, P > 0.01). Juveniles tend to have
bills shorter (mean 12.3 mm) than those of adults (mean 13.8
mm; t-test P = 0.0007); no other differences in measurements
were significant (t-test, P > 0.01). Most plumage variation in
adults of Robsonius results from variation in the amount of
dark feather edging on the back, throat, and breast, which
is strongly affected by feather wear. Overall, R. sorsogonen-
sis has the least dusky feather edging; some specimens show
faint dusky scalloping on the back, but none has scalloping on
the throat and upper breast as in R. rabori and R. thompsoni.
The malar stripe, for med from grayish or blackish feather
tips, is reduced in R. sorsogonensis and R. rabori i n compa ri-
son to R. thompsoni. The distinctive necklace of black spots in
R. thompsoni is variable and influenced by feather wear; in
some specimens, the throat is clean white and only a few
black spots on the upper breast form the necklace, whereas
other specimens have spotted or scalloped throats and prom-
inent black necklaces. The size of the white throat patch,
thickness of the breast band, and amount of white on the belly
are also variable within species, apparently mostly as a func-
tion of feather wear and style of specimen preparation, so the
biological significance of this variation remains unclear.
In our examination of specimens of Robsonius, we found
no diagnosable differences between R. s. sorsogonensis and
FIGURE 2. Distribution of Robsonius in the Luzon Island com-
plex, which includes Luzon, Catanduanes, Polillo, and Marinduque.
Areas of environmental conditions suitable for Robsonius (as in-
ferred from ecological niche models) are shown in dark gray. All lo-
calities known for Robsonius (specimens and observations), used to
train models, are displayed on top of the modeled distribution.
634 PETER A. HOSNER ET AL.
FIGURE 3. Representative specimens of adult and juvenile plumages of all three species of Robsonius, viewed ventrally (first column)
and laterally (second column). (A) adult R. rabori KU 119500); (B) juvenile R. rabori (FMNH 253557, holotype); (C) adult R. thompsoni
(PNM 20144, holotype); (D) juvenile R. thompsoni (KU 119893); (E) adult R. sorsogonensis (DMNH 37276); (F) juvenile R. sorsogonensis
(CM 153961).
NEW PHILIPPINE GROUND-WARBLER 635
TABLE 1. Mean measurements of specimens of the three species of Robsonius, with sample size (in parenthe-
ses) and ranges of measurements. We found no significant differences between the species or sexes, although
juveniles had significantly shorter bills than adults. Sample sizes varied because some measurements were not
possible on some specimens.
Bill length Bill depth Bill width Wing chord Tai l Tars us
Robsonius rabori 13.5 (2) 5.3 (1) 4.5 (2) 78.6 (2) 75.8 (2) 29.9 (2)
12.7–14.3 5.3 4.1–4.8 76.6–80.5 70.7–80.8 29.8–30.0
R. sorsogonensis 13.7 (21) 5.5 (17) 5.0 (19) 85.3 (21) 74.6 (17) 30.2 (21)
10.0 –15.6 4.7–6.2 4.1– 6.2 71.5 –9 6.0 67.0–87.7 28.0–31.8
R. thompsoni 12.9 (9) 5.1 (8) 5.0 (10) 79.8 (10) 73.3 (9) 29.7 (10)
12.1–14.5 4.3–5.8 4.6 –5.7 71.0– 87.0 64.3–81.7 27.4–32.3
R. s. mesoluzonica. Each of Dupont’s (1971a) characters
varies from individual to individual and are influenced
strongly by preparation style and feather wear. The holotype
of sorsogonensis is heavily worn, in contrast to the holotype of
mesoluzonica. Topotypes (Pakil, Laguna) of mesoluzonica are
variable, with fresh-plumaged birds similar to the holotype of
mesoluzonica, worn birds more similar to that of sorsogonensis.
Our DNA-sequence data, which include two samples from central
Luzon within the geographic range of mesoluzonica, also suggest
no population structure within R. sorsogonensis. As a conse-
quence, we suggest that R. s. mesoluzonica is not a diagnosable
subspecies and recommend treating R. sorsogonensis as mono-
typic, with the name mesoluzonica DuPont as a ju nior synonym .
Etymology. We name this species in honor of Max C.
Thompson, for his decades of contributions to natural history
collections and ornithology in particular. Long employed as
a professor of biology at Southwestern College, in Winfield,
Kansas, his involvement in diverse initiatives has produced
scientific insights and extensive specimen collections not only
from the Philippines, but also from Africa, Asia, Australia, the
southwestern Pacific, and numerous sites in the New World. His
collections are deposited at the University of Kansas, Smith-
sonian Institution, Bernice P. Bishop Museum, and American
Museum of Natural History (AMNH), and have provided an
invaluable resource for the ornithological community.
The English names currently and recently used for Robso-
nius are misleading because they refer to previous taxonomic
treatments when Robsonius was con side re d a babbler ( Rabor’s
Wren-Babbler/Luzon Wren-Babbler), or because they refer to
plumage characters that, in light of new specimen evidence, do
not diagnose the species (Rusty-headed Babbler/Gray-banded
Babbler; two species have a rusty head, all three have a gray
breast band, albeit with slight differences between each species).
We suggest new English names that highlight the restricted
distributions and geographic areas of endemism occupied
by each species within Luzon: Robsonius rabori, Cordillera
Ground-Warbler; R. thompsoni, Sierra Madre Ground-Warbler;
and R. sorsogonensis, Bicol Ground-Warbler.
Specimen material examined. Robsonius rabori: FMNH
253557 (holotype); KU 119500. Robsonius sorsogonensis:
AMNH 807095 (photos only); British Museum of Natural
History (BMNH) 1977.16.65–6 (photos only); Carnegie
Museum (CM) 151227, 153961; Delaware Museum of Natural
History (DMNH) 10800 (holotype of mesoluzonica, photos
only) 17443, 21812, 37275–6, 37928–33, 43771, 55857; FMNH
275745 (holotype), 399710, 462013, 472703; PNM 16656,
16795, 17532, 20144; Rijksmuseum van Natuurlijke Histoire
(RMNH) 99810 (photos only, from Collar 2006), University
of the Philippines Los Baños (UPLB) 3554; United States
National Museum (USNM) 608086 (photos only). Robsonius
thompsoni: Cincinnati Museum of Natural History (CMNH)
37710–1; FMNH 259385, 449800, 454990, 472601–3; KU
114634, 119893; PNM 16801, 19167, 20006 (holotype), Uni-
versity of Michigan Museum of Zoology (UMMZ) 226770
(photos only), USNM 607458, Yale Peabody Museum (YPM)
39989.
Audio records examined. Robsonius rabori: Macaulay
Library, Cornell University (ML) 166395. Robsonius thomp-
soni: Xeno-Canto (XC) 23080, 35259–61, 40988–92, 57572–3.
Robsonius sorsogonensis: 2 recordings (Scharringa 2005).
REMARKS
SYSTEMATICS
We used an initial molecular phylogenetic framework from
recent higher-level systematic studies that included Robso-
nius (Moyle et al. 2012, Oliveros et al. 2012) to clarify the
phylogenetic relationships among the populations. Character
sampling (4092 bp) included three mitochondrial genes
(1143 bp cytochrome b, 1041 bp ND2, 351 bp ND3) and three
nuclear introns (544 bp TGFb2-5, 570 bp Fib-5, 443 bp MUSK).
Sequences of 13 individuals (Appendix 1; GenBank accession
numbers KC603622–KC6033686) were derived from fresh
tissue samples, whereas the sequence of the juvenile type speci-
men of R. rabori was derived from DNA extracted from a toepad
clip. Outgroups included Donacobius, Thamnornis, Megalurus,
Locustella, and Bradypterus ( Oliveros et al. 2012). We inferred
trees by Bayesian (MrBayes 3.1, Ronquist and Huelsenbeck
2003; 20 million generations) and maximum-likelihood
(RAxML, Stamatakis 2006; 1000 bootstrap replicates) analysis
of the concatenated dataset. Preliminary analyses of individual
loci indicated no strongly supported conflicts in phylogenetic
636 PETER A. HOSNER ET AL.
signal between loci, justifying concatenation. See Moyle et al.
(2012) and Oliveros et al. (2012) for descriptions of gene regions
sequenced, laboratory protocols, and details of analysis. In ad-
dition to analyses described by Moyle et al. (2012), we recon-
structed phased ( Stephens et al. 2001) haplotype networks for
each nuclear locus with TCS (Clement et al. 2000).
Bayesian and maximum-likelihood analyses recovered
three strongly supported geographic clades within Robsonius
(Fig. 1). One clade comprised samples from southern Luzon,
including the Bicol Peninsula and Bulacan Province (1.0 Bayes-
ian posterior probability [BP], 87% maximum-likelihood
bootstraps [BS]); a second clade comprised samples from the
northern Sierra Madre in northeastern Luzon, including Aurora,
Nueva Vizcaya, and Cagayan provinces (1.0 BP, 100% BS); and
a third clade comprised samples from the Cordillera Central in
Ilocos Norte Province, northwestern Luzon (1.0 BP, 100% BS).
The northwestern clade (Cordillera Central, including the type
specimen of R. rabori) and the northeastern clade (northern
Sierra Madre, R. thompsoni) were strongly supported as sister
taxa (1.0 BP, 97% BS), which together were sister to the southern
Luzon clade (R. sorsogonensis).
Mitochondrial haplotypes of the three clades were highly
divergent; in ND2, uncorrected pairwise distances between the
three populations ranged from 7.4 to 8.7 % (raborithompsoni,
8.3–8.7 %; raborisorsogonensis, 7.9–8.2 %; thompsoni
sorsogonensis, 7.4–8.1 %). In cytochrome b, uncorrected pair-
wise distances between populations ranged from 5.2 to 7.4 %.
For comparison, d ivergences i n mtDNA of most recently
described species have been less than 5% (Voelker et al. 2010,
Pyle et al. 2011, Lara et al. 2012), and many are less than 2%
(O’Neill et al. 2011, Carneiro et al. 2012, Seeholzer et al.
2012). Haplotype networks of nuclear genes (Fig. 1) showed
no haplotype sharing among species, except in Fib-5, of which
R. sorsogonensis and R. thompsoni shared two haplotypes.
The ND3 sequence amplified from a toepad of the type speci-
men of R. rabori was the same haplotype as that from sample
KU 25708 and included no stop codons or heterozygous sites,
lending additional confidence that the DNA amplified is of
true mitochondrial origin.
Molecular data and plumage independently indicate three
diagnosable lineages of Robsonius within Luzon, consistent with
treatment of three species under the phylogenetic, evolutionary,
and general-lineage species concepts (de Queiroz 2007). Dis-
tributions of the three taxa are currently not known to overlap,
precluding direct evaluation of reproductive isolation and deter-
mination of species status under the biological species concept.
However, indirect evidence strongly suggests reproductive iso-
lation between the three species. First, deep genetic divergences
between lineages support a long independent history with no
evidence of hybridization. For example, specimens of R. thomp-
soni and R. sorsogonensis collected from southern Aurora and
northern Bulacan provinces show no evidence of phenotypic or
genetic intergradation, despite a separation of only 60 km and
no intervening break in continuous lowland forest. Second, all
plumage differences between the lineages are 100% diagnosable,
discrete, and fixed, no specimens being intermediate.
DISTRIBUTION
We used ecological niche modeling to produce a model
of environmental requirements of Robsonius as a clade, a
model with which we could assess distributional patterns
and whether phylogenetic breaks coincide with zones of low
environmental suitability. Occurrence data (Appendix 1)
consisted of specimen records (data accessed via the Global
Biodiversity Information Facility or museum collection man-
agers), supplemented with observations from the literature
(de Roever 1990, Lambert 1993, Harrap and Mitchell 1994,
Poulsen 1995, Collar 2006), web-reported sightings (eBird;
Wood et al. 2011), and data associated with vocal archives
(XC, ML). The occurrence data thus included 26 unique
localities with voucher specimens and 12 additional unique
localities based on observations and audio recordings, for a
total of 38 occurrence points, adequate for producing robust
models (Pearson et al. 2007). We chose to model Robsonius at
the genus level in light of the small sample size and because
ecological niches of allopatric replacement species tend to
be similar (Peterson et al. 1999, Peterson 2011). Climate data
(30 spatial resolution, or about 1 km) were drawn from the
WorldClim climate archive (Hijmans et al. 2005); we used the
following data layers: annual mean temperature, mean diurnal
temperature range, maximum temperature of warmest month,
minimum temperature of coldest month, annual precipitation,
and precipitation of the wettest and driest months. We devel-
oped niche models with GARP (Stockwell and Peters 1999)
and Maxent (Phillips et al. 2006).
Suitable areas identified by the niche models (Fig. 2) sug-
gest that the distribution of Robsonius is limited by environ-
mental factors within Luzon. Analysis of the contributions of
variables in Maxent indicates that precipitation in the driest
month (62.5%), minimum temperature in the coldest month
(18.8%), and annual mean temperature (11.9%) explain the
most variance; all other variables contributed less than 5%
of the variance. Thus niche models suggest that Robsonius is
confined to wetter rainforest and submontane forest and ab-
sent from seasonally dry monsoon forest in western Luzon
and the Cagayan Valley and from montane forest above ~1500
m. The environmentally unsuitable Cagayan Valley may iso-
late R. rabori and R. thompsoni, whereas R. thompsoni and
R. sorsogonensis do not appear to be isolated currently by
gaps in suitable environmental conditions.
DISCUSSION
SYSTEMATIC RELATIONSHIPS, BIOGEOGRA PHY,
AND DISTRIBUTION
Two sc e na r io s m ay ex pl ai n t h e d iv er si fic at io n of Robsonius
within the Luzon island complex. First, the three species may
NEW PHILIPPINE GROUND-WARBLER 637
constitute a case of intra-island diversification in a lowland
forest bird. Generally, birds are not thought to speciate
readily within the confines of islands (Diamond 1977, Coyne
and Price 2000), although recent molecular genetic studies
suggest that this phenomenon may be more widespread than
previously appreciated (e.g., Ryan et al. 2007, Hosner et al.
2013). Alternatively, three ancient taxa could have evolved on
separate proto-islands that constitute the current Luzon (Hall
2002, Sly et al. 2011).
Robsonius is limited to the Luzon Pleistocene aggre-
gate island complex (Heaney 1986, Brown and Diesmos
2002; Appendix 1), with records from Catanduanes Island
as well as Luzon itself. Ecological niche models suggest that
climatically suitable areas also exist on the smaller satellite
islands Polillo and Marinduque (Fig. 2), to which the com-
plex evidently had access during the low sea levels of the
Pleistocene. Recent surveys on Polillo have not encountered
Robsonius; however, Marinduque has received little ornitho-
logical attention since the voice of Robsonius was described
(Har rap and Mitchell 1994), so its presence there could have
been overlooked.
Robsonius thompsoni replaces R. rabori east of the Ca-
gayan Valley in forests associated with the northern Sierra
Madre and has been recorded in Cagayan, Isabela, Aurora,
and Nueva Vizcaya provinces; niche models indicate poten-
tial for occurrence in Quirino Province also (Fig. 2). Robsonius
sorsogonensis replaces R. thompsoni south of the Mid-Sierra
Madre Filter Zone, a region that has been hypothesized as an
important isolating barrier in birds (e.g., Sterrhoptilus nigro-
capitatus and S. dennistouni; Kennedy et al. 2000) and other
vertebrates (Welton et al. 2010). Robsonius sorsogonensis has
been recorded in Bulacan, Laguna, Quezon, Camarines Norte,
Camarines Sur, and Sorsogon provinces; ecological niche mod-
els suggest it may also occur in Rizal, Marinduque, and Albay
provinces.
Robsonius rabori has been recorded at only three locali-
ties (the type locality and two localities explored near Adams
in 2011) in northernmost Ilocos Norte Province, hence the first
species of this genus to be described is by far the least well
known. Ecological niche models identify broader environmen-
tal suitability in northwestern Luzon (Fig. 2), including parts of
Apayao, Kalinga, Mountain, and Ifugao provinces. This block
of lowland forest in the northern Cordillera Central is presum-
ably occupied by R. rabori but is poorly known by biologists
and requires further surveys. Alternatively, the distribution
of R. thompsoni may extend into the southern Cordillera
Central. It is known from Mt. Palali, an outlying peak of the
Sierra Madre just east of the Magat River (the largest tributary
of the Cagayan River, which separates the Sierra Madre from
the Cordillera Central). We encourage researchers working in
Apayao, Kalinga, Mountain, and Ifugao provinces to search for
Robsonius to clarify the range limits of each species and deter-
mine whether contact zones exist.
HABITAT
Limited data indicate that the three species of Robsonius have
similar habitat requirements. They have been collected and
observed in broad-leaved lowland and lower montane forest,
including primary forest, secondary forest, forest edge, logged
second growth, and forest on karst, from sea level to at least
1300 m. I n these habitats, Robsonius seems to prefer areas of
dark, thick undergrowth, including level areas with limestone
rocks, outcrops, and fallen logs; steep slopes with bamboo and
moss-covered boulders (de Roever 1990, Lambert 1993, Harrap
and Mitchell 1994, Poulsen 1995, Kennedy et al. 2000, Col-
lar and Robson 2007); tree-fall gaps; and steep, shrub-filled
ravines (PAH, pers. obs.). Occasionally, it has been found in
tall, thick grass near the edge of secondary forests (Poulsen
1995; PAH, pers. obs. at Mt. Cagua, Cagayan Prov.). In areas
of mixed primary and secondary habitats, our limited observa-
tions suggest that Robsonius may be more frequent in younger
second growth. For example, at Adams, Ilocos Norte, we heard
R. rabori only twice in 10 days of survey effort at the tall- forest
site on Mt. Pao but heard up to five birds in a single day in
secondary forest on nearby Mt. Cabacan. Greater abundance
in secondary forest may be a function of suitable dense under-
growth rather than preference for secondary habitats per se; we
are unaware of records of Robsonius populations in isolated
patches of secondary forest away from large tracts of tall forest.
ECOLOGY AND BEHAV IOR
Because of the birds’ secretive habits and occurrence in dense
undergrowth, the ecology and behavior of Robsonius remain
poorly understood. Most sightings are of individuals or family
groups (adults with juveniles) walking slowly on the ground,
flipping over leaves and woody debris in search of invertebrates
(Goodman and Gonzales 1990, de Roever 1990, Lambert 1993,
Harrap and Mitchell 1994, Poulsen 1995, Collar and Robson
2007). Stomach contents (n = 4) included primarily sclerotized
insect parts. Robsonius walks or runs across the forest floor,
with the tail held straight out or cocked at a 30–60° angle, oc-
casionally to 90° when startled or agitated (de Roever 1990),
In addition to its typical ground-walking habits, in response to
playback, an agitated R. sorsogonensis made short wing-as-
sisted jumps between several small vertical stems and perched
vertically in a posture similar to that of many wrens (Troglo-
dytidae) and neotropical antbirds (i.e., Pithys or Gymnopithys;
PAH, pers. obs. at Mt. Labo, Camarines Norte Prov.).
Collar and Robson (2007) and Sánchez-González et al.
(2010) described the nest of Robsonius sorsogonensis and
thompsoni as a large ball structure with a side entrance and
placed in understory vegetation, reminiscent of nests of Pitta
and some species of Bradypterus and Megalurus. Sánchez-
González et al. (2010) referred to the northern Sierra Madre
populations as R. rabori, consistent with past taxonomic
treatments, but the nest actually belonged to R. thompsoni,
so the nest was described before the species had a name. The
638 PETER A. HOSNER ET AL.
nest of R. rabori remains undescribed. The clutch size in each
described nest was two, and the eggs were white with reddish-
brown speckles.
CONSERVATION
In the most recent conservation assessments, R. “rabori” (in-
cluding both R. rabori a nd R. thompsoni) and R. sorsogonensis
have each been treated as vulnerable on the basis of limited
range (<6000 km2), small number of known localities, and
suspected population declines from forest fragmentation
(BirdLife International 2012). The species of Robsonius are
now known from more localities and a larger area than pre-
viously reported (Appendix 1): 3, 21, and 14 localities for
R. rabori, thompsoni, and R. sorsogonensis, respectively.
Recognizing R. thompsoni as a species results in smaller
distributions; on the basis of this information and revised tax-
onomy, we recommend elevating R. rabori to endangered, but
treating R. thompsoni and R. sorsogonensis as vulnerable, in
each case following the criteria of the IUCN (2013).
In recent years, the forests of the northern Sierra Madre have
rightfully received a great deal of attention from the conserva-
tion community. They are the largest within Luzon and protect
large numbers of endemic, endangered, and threatened species
(Mallari and Jensen 1993, Poulsen 1995). However, the discovery
we report here illustrates that conserving small portions of spe-
ciesranges may leave differentiated populations unprotected,
particularly in a landscape whose beta diversity is as great as the
Philippines’ (Peterson 2006, Welton et al. 2010). We hope that
R. rabori, now the only bird known to be endemic to the lowland
forests around the Cordillera Central, can become a flagship
species for forest conservation in the region. Tracts of lowland
forest persist in Ilocos Norte, Apayao, Kalinga, and Mountain
provinces. Compared with the northern Sierra Madre region, the
lowland avifauna of the Cordillera Central is poorly known, and
renewed interest in the area will likely result in discovery of other
bird populations important for conservation.
ACK NOW LEDGME NTS
We thank the collectors and field biologists who facilitated field work
or collected the specimens used in this study: Jameson Reynon,
Nevong Puna, Robin Jones, and Arvin Diesmos. We thank in par-
ticular Rafe Brown, for making possible our recent field program
in the Philippines, and Lawrence R. Heaney, whose field program
contributed numerous important specimens and tissue samples. We
thank the Depar tment of Environment and Natural Resources and
the Protected Areas and Wildlife Bureau of the Philippines for facil-
itating collecting and export permits. The Field Museum of Natu-
ral Histor y generously provided tissue samples, including a toe clip
from the holotype of R. rabori. The Cincinnati Museum of Natural
History, Carnegie Museum of Natural History, Delaware Museum
of Natural History, and Field Museum of Natural History lent speci-
mens, and the American Museum of Natural History, University of
Michigan Museum of Zoology, and the Delaware Museum of Natu-
ral History and the Natural History Museum (Tring) provided photos
of specimens. Town Peterson, Mark Robbins, and three anonymous
reviewers provided helpful comments improving the quality of the
manuscript. The National Science Foundation supported field work
(DEB-0743491) and lab work (DEB-0743576, DEB-1110619). The
Consejo Nacional de Ciencia y Tecnología (CONACyT) supported a
postdoctoral fellowship (Exp. 93730) to LASG.
LITERATURE CITED
BALDWIN, S. P., H. C. OBERHOLSER, AND L. G. WORLEY. 1931. Mea-
surements of birds. Scientific Publications of the Cleveland
Museum of Natural Histor y 2:1–165.
BIRDLIFE INTERNATIONAL [ONLINE]. 2012. IUCN Red List for birds.
<http://www.birdlife.org> (7 March 2012).
BROWN, R. M., AND A. C. DIESMOS. 2002. Application of lineage-
based species concepts to oceanic island f rog populations: the
effects of differing taxonomic philosophies on the estimation of
Philippine biodiversity. Silliman Journal 42:133–162.
CARNEIRO, L. S., L. P. GONZAGA, P.S. REGO, I. SAMPAIO, H. SCHNEIDER,
AND A. ALEIXO. 2012. Systematic revision of the Spotted Antpitta
(Grallariidae: Hylopezus macularius) with the description of a
cryptic new species from Brazilian Amazonia. Auk 129:338–351.
CLEMENT, M., D. POSADA, AND K. A. CRANDALL. 2000. TCS: a com-
puter program to estimate gene genealogies. Molecular Ecology
9:1657–1659.
COLLAR, N. J. 2006. A partial revision of the Asian babblers, Timali-
idae. Forktail 22:85–112.
COLLAR, N. J., AND C. ROBSON. 2007. Family Timaliidae (babblers),
p. 70–291. In J. del Hoyo, J., A. Elliott, and D. A. Christie [EDS.],
Handbook of the birds of the world, volume 12 : Pic ath ar tes to tits
and chickadees. Lynx Edicions. Barcelona, Spain.
COYN E, J. A., AND T. D. PRICE. 2000. Little evidence for sympatric
speciation in island birds. Evolution 54:2166 –2171.
DE QUEIROZ, K. 2007. Species concepts and species delimitation.
Systematic Biology 56: 879–886.
DE ROEVER, J. 1990. Notes on the Luzon Wren-Babbler. Bulletin of
the Oriental Bird Club 12:33 –34.
DIAMOND, J. M. 1977. Continental and insular speciation in Pacific
landbirds. Systematic Zoology 26:263–268.
DU PONT, J. E. 1971a. Note s on Ph ilippi ne bi rds ( No. 1) . Nemour ia 3:1–6 .
DU PONT, J. E. 1971b. Philippine birds. Monograph Series 2. Dela-
ware Museum of Natural Histor y, Greenville, DE.
GOODMAN, S. M., AND P. C. GONZALES. 1990. The birds of Mt. Isa-
rog National Park, southern Luzon, Philippines, with par ticular
reference to altitudinal distribution. Fieldiana Zoology 60:1–39.
HALL, R. 2002. Cenozoic geological and plate tectonic evolution
of SE Asia and the SW Pacific: computer-based reconstr uc-
tions, model and animations. Journal of Asian Earth Sciences
20:353–431.
HARRAP, S., AND K. MITCHELL. 1994. More notes on Rabor’s Wren-
Babbler Napothera rabori. Oriental Bird Club Bulletin 20:50–51.
HEANEY, L. R. 1986. Biogeography of mammals in SE Asia:
estimates of rates of colonization, extinction and speciation.
Biological Journal of the Linnaean Society 28: 127–165.
HIJMANS, R. J., S. E. CAMERON, J. L. PARRA, P. G. JONES, AND A.
JARVIS . 2005. Very high resolution interpolated climate sur-
faces for global land areas. International Journal of Climatology
25:1965–1978.
HOSNER, P. A., Á. S. NYÁR I, AND R. G. MOYLE . 2013. Water barriers
and intra-island isolation contribute to diversification in the insu-
lar Aethopyga sunbirds (Aves: Nectariniidae). Journal of Bioge-
ography 40:1094–1106.
IUCN [ONLINE]. 2013. IUCN red list of threatened species, version
2013.1. <www.iucnredlist.org> (22 July 2013).
NEW PHILIPPINE GROUND-WARBLER 639
KENNEDY, R. S., P. C. GONZALES, E. C. DICKINSON, H. C. MIRANDA,
AND T. H. FISHER. 2000. A guide to the birds of the Philippines.
Oxford University Press, Oxford, England.
LAMBERT, F. L. 1993. Some key sights and significant records of bird
from the Philippines and Sabah. Bird Conservation International
3:281–297.
LARA, C. E., A. M. CUERVO, S. V. VALDERRAMA, D. CALDERÓN-F.,
AND C. D. CADENA. 2012. A new species of wren (Troglodytidae:
Thryophilus) from the d ry Cauca R iver canyon, northwester n
Colombia. Auk 129: 537–550.
MALLARI, N. A. S., AND A. JENSEN. 1993. Biological diversity in
northern Sierra Madre, Philippines: its implication for conserva-
tion and management. Asia Life Sciences 2:1–12.
MOYLE , R. G., M. J ANDERSEN, C. H. OLIVEROS, F. STIENHEIMER, AND
S. REDDY. 2012. Phylogeny and biogeography of the core babblers
(Aves: Timaliidae). Systematic Biology 61:631–651.
OLIVEROS, C. H., S. REDDY, AND R. G. MOYLE . 2012. The phylogenetic
position of some Philippine “babblers” spans the muscicapoid and syl-
vioid radiations. Molecular Phylogenetics and Evolution 65:799–804.
O’NEILL, J. P., D. F. LANE, AND L. N. NAKA. 2011. A cryptic new spe-
cies of thrush (Turdidae: Turd us) from western Amazonia. Condor
113 :86988 0.
PEARSON, R. G., C. J. RAXWORTHY, M. NAKAMURA, AND A . T. PETERSON.
2007. Predicting species distributions from small numbers of
occurrence records: a test case using cryptic geckos in Madagas-
car. Journal of Biogeography 34:102–117.
PETERSON, A. T. 2006. Taxonomy is important in conservation: a
preliminary reassessment of Philippine species-level bird taxon-
omy. Bird Conservation International 16:155–173.
PETERSON, A. T. 2011. Ecological niche conservatism: a time-struc-
tured review of evidence. Journal of Biogeography 38:817–827.
PETERSON, A. T., J. SOBERÓN, AND V. SÁNCHEZ-COR DERO. 1999. Con-
servatism of ecological niches in evolutionary time. Science
285:1265–1267.
PHILLIPS, S. J., R. P. ANDERSON, AND R. E. SCHAPIRE. 2006. Maxi-
mum entropy modeling of species geographic distributions. Eco-
logical Modelling 190:231–259.
POULSEN, M. K. 1995. The threatened and near-threatened birds of
Luzon, Philippines, and the role of the Sierra Madre mountains
in their conservation. Bird Conservation International 5:70–115.
PYLE, P., A. J. WELCH AND R. C. FLEISCHER. 2011. A new species of
shearwater (Puffinus) recorded from Midway Atoll, northwestern
Hawaiian Islands. Condor 113:518–527.
RAND, A. L. 1960. A n ew sp ecies of babbl ing thr ush fr om the Philip -
pines. Fieldiana Zoology 39:377–379.
RAND, A. L., AND D. S. RABOR. 1967. New birds from Luzon, Philippine
Islands. Fieldiana Zoology 51:85–99.
RONQUIST, F., AND J. P. HUELSENBECK. 2003. MrBayes 3: Bayesian
phylogenetic inference under mixed models. Bioinformatics 19:
1572–1574.
RYAN, P. G., P. BLOOMER, C. L. MOLON EY, T. J. GRANT, AND W. DEL-
PORT. 2007. Ecological speciation in south Atlantic island finches.
Science 315:1420–1423.
SÁNCHEZ-GONZÁLEZ, L. A., C. OLIVEROS, N. PUNA, AND R. G.
MOYLE. 2010. Nests, nest placement, and eggs of three Philip-
pine endemic birds. Wilson Journal of Ornithology 122:587–591.
SCHARRINGA, J. 2005. Birds of tropical Asia 3. Bird Songs Interna-
tional. Enschede, Netherlands.
SEEHOLZER, G. F., B. M. WINGER, M. G. HARVEY, D. CÁCERES A., AND
J. D. WECKSTIEN. 2012. A new species of barbet (Capitonidae:
Capito) from the Cerros del Sira, Ucayali, Peru. Auk 129:551–559.
SLY, N., A. K. TOWNSEN D, C. C. RIMMER, J. M. TOWNSEN D, S. C.
LATTA, AND I. J. LOV ETTE. 2011. Ancient islands modern inva-
sions: disparate phylogeographic histories among Hispaniola’s
endemic birds. Molecular Ecology 20:5012–5024.
SMITHE, F. B. 1975. Naturalist’s color guide. American Museum of
Natural History, New York.
STAMATA KIS, A. 2006. RAxML-VI-HPC: maximum likelihood-
based phylogenetic analyses with thousands of taxa and mixed
models. Bioinformatics 22:2688–2690.
STEPHENS, M., N. SMITH, AND P. DONNELLY. 2001. A new statisti-
cal method for haplotype reconstruction from population data.
American Journal of Human Genetics 68: 978 –989.
STOCKWELL, D. R. B. AND D. P. PETERS. 1999. The GAR P modell ing sys-
tem: problems and solutions to automated spatial prediction. Inter-
national Journal of Geographical Information Science 13:143-158.
VOELKER, G., R. K. OUTLAW, S. REDDY, M. TOBLER, J. M. BATES, S. J.
HACKETT, C. KAHINDO, B. D. MARKS, J. C. K. PETERHANS, AND T.
GNOSKE. 2010. A new species of boubou (Malaconotidae: Lania-
rius) from the Albertine Rift. Auk 127:678–689.
WELTON, L. J., C. D SILER, D. BENNETT, A. DIESMOS, M. R. DUYA,
R. DUGAY, E. L. B. RICO, M. VAN WEERD, AND R. M. BROWN.
2010. A spectacular new Philippine monitor lizard reveals a hid-
den biogeographic boundary and a novel flagship species for con-
servation. Biology Letters 6:654–658.
WINKER, K. 1998. Suggestions for measuring external characters of
birds. Ornitología Neotropical 9: 23–30.
WOOD, C., B. SULLIVAN, M. ILIFF, D. FINK, AND S. KELLING. 2011.
eBird: Engaging birders in science and conservation. PLoS
Biology 9:e1001220.
... elevational ranges (Manne et al., 1999;Şekercioglu et al., 2008;White and Bennett, 2015), large-bodied (Bennett and Owens, 1997;Boyer, 2008;Wang et al., 2018), have limited dispersal ability (Moore et al., 2008;Lees and Peres, 2009;Sheard et al., 2020), be ecologically specialized (Weerd et al., 2003;Norris and Harper, 2004;Şekercioglu, 2011), have exposed nests or nest on the ground (Terborgh, 1974;Wilcove, 1985;Boyer, 2008), or have low fecundity (Bennett and Owens, 1997;Kruger and Radford, 2008). While we are assessing all of the resident birdlife in the Philippines, we especially focused on species that have either been split within the last decade or have been proposed to consist of cryptic populations that may warrant species-level status (Lohman et al., 2010;Collar, 2011;Rasmussen et al., 2012;Hosner et al., 2013;Campbell et al., 2016). We also aim to identify Philippine bird species which are not currently recognized as threatened by the IUCN but may be at risk of extinction in the near future. ...
... To examine the threat status of recently split species, we identified the species in the literature that have been associated with taxonomic updates and proposed revisions within the last decade (Supplementary Table 2). This includes the 35 species recognized by BirdLife International (2021) that have recently been described as a result of species splits (Lohman et al., 2010;Collar, 2011;Rasmussen et al., 2012;Hosner et al., 2013;Campbell et al., 2016;Arndt et al., 2019) and, if applicable, the endemic Philippine parent species from which these birds were split (Supplementary Table 2). We also included the 19 species that consist of one or more cryptic populations that have been recommended to be split but have yet to be split and recognized by BirdLife International (2021) as separate species (Supplementary Table 2; Lohman et al., 2010;Collar, 2011;Campbell et al., 2016). ...
... For our thresholds for fitted threat status, we chose a range between 0.5 and 1.5 to indicate a species was NT, 1.5-2.5 for VU, 2.5-3.5 for EN, and 3.5 and higher for CR; a range of 0.0-0.5 indicated LC. We also used the same approach to extract and assess fitted threat values for 35 species (and their parent species, if applicable) recognized by BirdLife International (2021) that were recently split (Lohman et al., 2010;Collar, 2011;Rasmussen et al., 2012;Hosner et al., 2013;Campbell et al., 2016;Arndt et al., 2019), as well as the 19 species which consist of populations of one or more cryptic species (Lohman et al., 2010;Collar, 2011;Campbell et al., 2016) that have not yet been recognized by BirdLife International (2021). Finally, we used our model, which had excluded DD species, to predict the threat status of the four DD species. ...
Article
Full-text available
The majority of the world’s biodiversity occurs in the tropics, but human actions in these regions have precipitated an extinction crisis due to habitat degradation, overexploitation, and climate change. Understanding which ecological, biogeographical, and life-history traits predict extinction risk is critical for conserving species. The Philippines is a hotspot of biodiversity and endemism, but it is a region that also suffers from an extremely high level of deforestation, habitat degradation, and wildlife exploitation. We investigated the biological correlates of extinction risk based on the IUCN Red List threat status among resident Philippine birds using a broad range of ecological, biogeographical, and life history traits previously identified as correlates of extinction risk in birds. We found strong support across competing models for endemism, narrower elevational ranges, high forest dependency, and larger body size as correlates significantly associated with extinction risk. Additionally, we compared observed threat status with threat status fitted by our model, finding fourteen species that are not currently recognized by the IUCN Red List as threatened that may be more threatened than currently believed and therefore warrant heightened conservation focus, and predicted threat statuses for the four Philippine Data Deficient bird species. We also assessed species described in recent taxonomic splits that are recognized by BirdLife International, finding 12 species that have a fitted threat status more severe than their IUCN-designated ones. Our findings provide a framework for avian conservation efforts to identify birds with specific biological correlates that increase a species’ vulnerability to extinction both in the Philippine Archipelago and elsewhere on other tropical islands.
... By contrast, fewer studies have examined the importance of intra-island diversification, which could have played a significant role in generating diversity on the larger and more topographically complex islands in the Philippines (Heaney, 2000;Kisel & Barraclough, 2010;Whittaker et al., 2008). Evidence indicating the potential importance of intra-island speciation in the Philippines has emerged from research on birds (Hosner, Nyari, & Moyle, 2013;Hosner, Boggess, et al., 2013), skinks (Linkem, Diesmos, & Brown, 2011), treehoppers (Su et al., 2014), shrews , and murid rodents (Balete, Rickart, Heaney, & Jansa, 2015;Balete et al., 2012;Heaney, Balete, Rickart, & Jansa, 2014;Heaney et al., 2011;Justiniano et al., 2015;Steppan, Zawadski, & Heaney, 2003), with many of these studies implicating the numerous 'sky islands' on Luzon and Mindanao as important drivers of diversification. The extent to which such intra-island processes represent the rule or exceptions in the Philippines, however, remains inadequately assessed. ...
... The importance of Luzon's sky islands was perhaps most evident in our AMOVA, which found that nearly 70% of mtDNA variation was explained by these biogeographical regions (Table 1). This result was also corroborated by high pairwise F ST estimates between sky islands (Table 2) (Welton et al., 2010), ground warblers (Hosner, Boggess, et al., 2013), and several murid rodent lineages Justiniano et al., 2015). In addition, we draw parallels between our results and those documented for the Old Endemic murid rodents on Luzon, which also highlighted the importance of sky islands as drivers of diversification (Balete et al., , 2012Heaney et al., 2011Heaney et al., , 2014Heaney et al., , 2016Justiniano et al., 2015). ...
Article
To investigate patterns of genetic and morphological differentiation in a ‘New Endemic’ Philippine rodent (Bullimus) and examine the dynamics on intra-island speciation on oceanic islands. The Philippine archipelago, focusing on Luzon Island. We examined patterns of genetic differentiation by sequencing one mitochondrial and three nuclear loci for an archipelago-wide sampling of Bullimus, which we analysed using phylogenetic, population genetic, and coalescent-based methods. We also measured 16 craniodental features for Bullimus luzonicus to examine morphological differentiation on Luzon. Time-calibrated species tree analysis indicated that Bullimus colonized the southern Philippines during the late Pliocene or Pleistocene and diversified on Luzon during the last c. 540 kyr. We found substantial phylogeographical structure in B. luzonicus, most completely developed in the southern part of the island, where colonization initially occurred. Coalescent species delimitation analysis indicated strong support for genetic isolation among eight B. luzonicus populations, however, morphometric analyses suggested that only two or three of these populations were morphologically distinctive. Phylogeographical structure in B. luzonicus largely corresponded to Luzon's numerous isolated mountain ranges, which likely played a major role in driving diversification in central/northern Luzon over the last c. 230 kyr. The majority of B. luzonicus populations appear to be in the process of speciation, with only the furthest south population from Camarines Sur warranting elevation to species status. We found little evidence for diversification in B. bagobus on Mindanao, suggesting that Luzon may be an exceptional setting for murid rodent diversification.
... The copyright holder for this preprint this version posted February 10, 2022. ; https://doi.org/10.1101/2022.02.07.479408 doi: bioRxiv preprint collection of the first adult specimen of true R. rabori (Hosner et al., 2013). All four standardized world lists currently recognize all three species: Robsonius rabori, R. sorsogonensis, and R. ...
Preprint
Full-text available
Biodiversity research has advanced by testing expectations of ecological and evolutionary hypotheses through the linking of large-scale genetic, distributional, and trait datasets. The rise of molecular systematics over the past 30 years has resulted in a wealth of DNA sequence data from around the globe, facilitating biodiversity research. However, advances in molecular systematics also have created taxonomic instability, as new estimates of evolutionary relationships and interpretations of species limits have led to widespread scientific name changes. Taxonomic instability, or “splits, lumps, and shuffles”, present logistical challenges to large-scale biodiversity research because species or populations may be listed under different names in different data sources, or because different species or populations may be listed under previous names. Consequently, distributional and trait data are often difficult to link directly to DNA sequence data without extensive and time consuming curation. Here, we present RANT: Reconciliation of Avian NCBI Taxonomy. RANT applies taxonomic reconciliation to standardize all avian names in use in NCBI GenBank, a primary source of genetic data, to a widely-used and regularly-updated avian taxonomy: eBird/Clements. Of 14,341 avian species or subspecies names used by GenBank, 11,031 names directly matched an eBird/Clements name, which were linked to over 6 million nucleotide sequences. For the remaining unique avian names in GenBank, we used Avibase’s taxonomic concepts, taxonomic descriptions in Cornell’s Birds of the World, and DNA sequence metadata to identify corresponding eBird/Clements names. Reconciled names were linked to over 600,000 nucleotide sequences, approximately 9% of all avian sequences on GenBank. Nearly 10% of eBird/Clements names had nucleotide sequences listed under two or more GenBank names. Our avian GenBank naming reconciliation is open source and available at GitHub, where it can be updated to correspond with future annual eBird/Clements taxonomic updates. LAY SUMMARY – 23% of avian names on GenBank do not match eBird/Clements, a widely-used standardized avian taxonomy – 600,000 nucleotide sequences on GenBank are associated with names that do not match eBird/Clements – 10% of eBird/Clements names have nucleotide sequences listed under multiple GenBank names – We provide an open source taxonomic reconciliation to mitigate difficulties associated with non-standardized name use for GenBank data
... However, despite the Philippines' importance as a natural laboratory and status as a 'biodiversity hotspot' (Myers et al., 2000), knowledge of the pattern and process of diversification across the archipelago remains largely incomplete. Even within well-studied vertebrate groups, new species are regularly described from the Philippines, largely resulting from both increased sampling efforts and the widespread application of molecular phylogenetic tools in taxonomic research (Posa et al., 2008;Welton et al., 2010;Heaney et al., 2009;Heaney et al., 2011;Heaney et al., 2016;Hosner et al., 2013a). Consequently, estimates of endemism for Philippine taxa have risen substantially, and currently range from 56 to 80% for terrestrial vertebrate taxa . ...
Article
Full-text available
The traditional view of species’ distributions is that they are less abundant near the edges of their ranges and more abundant towards the centre. Testing this pattern is difficult because of the complexity of distributions across wide geographical areas. An alternative strategy, however, is to measure species’ distributional patterns along elevational gradients. We applied this strategy to examine whether lowland forest birds are indeed less common near their upper range limits on a Bornean mountain, and tested co‐occurrence patterns among species for potential causes of attenuation, including signatures of habitat selection and competition at the periphery of their ranges. Mt. Mulu, Borneo. Rain forest birds. We surveyed lowland forest birds on Mt. Mulu (2,376 m), classified their elevation‐occupancy distributions using Huisman–Olff–Fresco (HOF) models, and examined co‐occurrence patterns of species pairs for signatures of shared habitat patches and interspecific competition. For 39 of 50 common species, occupancy was highest at sea level then gradually declined near their upper range edges, in keeping with a “rare periphery” hypothesis. With respect to habitat selection, lowland species do not appear to cluster together at sites of patchy similar habitat near their upper range limits; neither are most lowland species segregated from potential montane competitors where ranges overlap. High relative abundance at sea level implies that species inhabit “truncated niches” and are not currently near the limits of their fundamental niche, unless unknown critical response thresholds exist. However, indirect effects of increasing temperature predicted under climate change scenarios could still influence lower range limits of lowland species indirectly by altering habitat, precipitation regimes and competitive interactions. The lack of non‐random co‐occurrence patterns implies that patchy habitat and simple pairwise species interactions are unlikely to be responsible for upper range limits in most species; diffuse competition across diverse rain forest bird communities could still play a role.
... Based on numbers of distinctive subspecies once recognized as species (McGregor 1909), several other areas in the Philippines likely contain substantial unrecognized diversity. Luzon Island contains many distinctive north/south replacement subspecies, some of which have recently been re-evaluated and elevated to full species (Sheldon et al. 2012;Hosner et al. 2013a). Similarly, Palawan Island has many diagnosable subspecies that are distinct from those of Borneo, some of which have recently been elevated to species (Oliveros and Moyle 2010;Moltesen et al. 2012). ...
Article
Full-text available
The Philippine archipelago is recognized as a biodiversity hotspot because of its high levels of endemism and numerous threatened species. Avian lineages in the Philippines feature morphologically distinct allopatric taxa, which have been variably treated either as species or subspecies depending on species concepts and recognition criteria. To understand how alternative species limits would alter diversity metrics and patterns of endemism in the Philippines, we selected 19 focal lineages of birds, each containing multiple described taxa within the Mindanao Island Group. We delimited species in an integrative, lineage-based framework using three operational criteria: species must (1) form well-supported, geographically circumscribed clades, (2) be monophyletic with significant genetic differentiation identified by a coalescent model, and (3) feature fixed differences in phenotypic characters. Our criteria identified 40 species from the original 19 focal lineages, a 50–74% increase over recent comprehensive taxonomic treatments. Genetic criteria in isolation identified an additional 10 populations that could be cryptic species in need of further study. We identified fine-scale endemism within the Mindanao Island Group, with multiple unrecognized avian endemics restricted to Samar/Leyte, Bohol Island, and the Zamboanga Peninsula. Genetic and phenotypic information support the hypothesis that polytypic bird species in the Philippines tend to be composed of evolutionarily distinct, range-restricted, allopatric replacements rather than widespread and variable “superspecies”. We conclude that lack of species recognition has resulted in underestimates of species diversity and overlooked fine-scale endemism in the Philippines. Recognizing this diversity would alter conservation priorities, shifting efforts to protect microendemics on smaller islands and finer scale endemic areas within larger islands.
... Bialowieza in Poland gets 15 times more income from visiting birders than from logging (Czeszczewik and Walankiewicz, 2017). Even the most cryptic, obscure, hard-to-identify bird species receive considerable public attention if they are in need of conservation (see, for example, Hirschfeld et al., 2013;Hosner et al., 2013). In addition, birders form a sizeable pool of volunteers for various scientific projects, often contributing their money as well as manpower. ...
Article
Full-text available
During the 20th century, birding evolved from a little-known hobby into a global phenomenon important for ornithology and bird conservation. More recently a similar change has begun for mammalwatching, which is rapidly gaining popularity and is already providing financial support, observational data, diagnostic information, and a volunteer base for mammalogy and mammalian conservation. The study data suggest that mammalwatching has the potential to end decades of neglect of small mammals in dire need of conservation, to improve our knowledge of mammalian status and distribution, and to increase public support for conservation measures, especially for species not seen as particularly charismatic by the general public. Professional mammologists and conservation workers can benefit from this new trend, but they can also help it. We offer a number of suggestions as to how professionals mammalogists and the amateur community can better work together to promote conservation and science.
... However, despite the Philippines' importance as a natural laboratory and status as a 'biodiversity hotspot' (Myers et al., 2000), knowledge of the pattern and process of diversification across the archipelago remains largely incomplete. Even within well-studied vertebrate groups, new species are regularly described from the Philippines, largely resulting from both increased sampling efforts and the widespread application of molecular phylogenetic tools in taxonomic research (Posa et al., 2008;Welton et al., 2010;Heaney et al., 2009;Heaney et al., 2011;Heaney et al., 2016;Hosner et al., 2013a). Consequently, estimates of endemism for Philippine taxa have risen substantially, and currently range from 56 to 80% for terrestrial vertebrate taxa . ...
Article
Molecular phylogenetic approaches have greatly improved our knowledge of the pattern and process of biological diversification across the globe; however, many regions remain poorly documented, even for well-studied vertebrate groups. The Philippine archipelago, one of the least-studied 'biodiversity hotspots', is an ideal natural laboratory for investigating the mechanisms driving diversification in an insular and geologically dynamic setting. We investigated the history and geography of diversification of the Philippine populations of a widespread montane bird, the White-browed Shortwing (Brachypteryx montana). Leveraging dense archipelago-wide sampling, we generated a multi-locus genetic dataset (one nuclear and two mtDNA markers), which we analyzed using phylogenetic, population genetic, and coalescent-based methods. Our results demonstrate that Philippine shortwings 1) likely colonized the Philippines from the Sunda Shelf to Mindanao in the late Miocene or Pliocene, 2) diversified across inter-island barriers into three divergent lineages during the Pleistocene, 3) have not diversified within the largest island, Luzon, contrary to patterns observed in other montane taxa, and 4) colonized Palawan from the oceanic Philippines rather than from Borneo, challenging the assumption of Palawan functioning exclusively as a biogeographic extension of the Sunda Shelf. Additionally, our finding that divergent (c. 2.1 mya) lineages are coexisting in secondary sympatry on Mindanao without apparent gene flow suggests that the speciation process is likely complete for these shortwing lineages. Overall, these investigations provide insight into how topography and island boundaries influence diversification within remote oceanic archipelagos and echo the results of many other studies in demonstrating that taxonomic diversity continues to be underestimated in the Philippines.
... Luzon PAIC populations lacked genetic structure, with the exception of H. ardens, for which we recovered slight mitochondrial differentiation between northern and southern Luzon populations (Fig. S3). Genetic differentiation within the Luzon PAIC has been reported in other bird (Sánchez-González and Moyle 2011;Sheldon et al. 2012;Hosner et al. 2013b) and other vertebrate groups (Welton et al. 2010), but the potential role of paleoclimate in causing this differentiation has not been assessed. Model projections in these eight taxa suggest that suitable environmental conditions for forest species were widespread and continuous within the Luzon PAIC throughout Pleistocene/late Pliocene climate oscillations. ...
Article
Biodiversity research has advanced by testing expectations of ecological and evolutionary hypotheses through the linking of large-scale genetic, distributional, and trait datasets. The rise of molecular systematics over the past 30 years has resulted in a wealth of DNA sequences from around the globe. Yet, advances in molecular systematics also have created taxonomic instability, as new estimates of evolutionary relationships and interpretations of species limits have required widespread scientific name changes. Taxonomic instability, colloquially “splits, lumps, and shuffles'', presents logistical challenges to large-scale biodiversity research because (1) the same species or sets of populations may be listed under different names in different data sources, or (2) the same name may apply to different sets of populations representing different taxonomic concepts. Consequently, distributional and trait data are often difficult to link directly to primary DNA sequence data without extensive and time-consuming curation. Here, we present RANT: Reconciliation of Avian NCBI Taxonomy. RANT applies taxonomic reconciliation to standardize avian taxon names in use in NCBI GenBank, a primary source of genetic data, to a widely used and regularly updated avian taxonomy: eBird/Clements. Of 14,341 avian species/ subspecies names in GenBank, 11,031 directly matched an eBird/Clements; these link to over 6 million nucleotide sequences. For the remaining unmatched avian names in GenBank, we used Avibase’s system of taxonomic concepts, taxonomic descriptions in Cornell’s Birds of the World, and DNA sequence metadata to identify corresponding eBird/Clements names. Reconciled names linked to over 600,000 nucleotide sequences, ~9% of all avian sequences on GenBank. Nearly 10% of eBird/Clements names had nucleotide sequences listed under 2 or more GenBank names. Our taxonomic reconciliation is a first step towards rigorous and open-source curation of avian GenBank sequences and is available at GitHub, where it can be updated to correspond to future annual eBird/Clements taxonomic updates.
Article
Full-text available
Ensuring that conservation decisions are informed by the best available data is a fundamental challenge in the face of rapid global environmental change. Too often, new science is not easily or quickly translated into conservation action. Traditional approaches to data collection and science delivery may be both inefficient and insufficient, as conservation practitioners need access to salient, credible, and legitimate data to take action. Open access data could serve as a tool to help bridge the gap between science and action, by providing conservation practitioners with access to relevant data in near real time. Broad-scale citizen-science data represent a fast-growing resource for open access databases, providing relevant and appropriately scaled data on organisms, much in the way autonomous sensors do so on the environment. Several such datasets are now broadly available, yet documentation of their application to conservation is rare. Here we use eBird, a project where individuals around the world submit data on bird distribution and abundance, as an example of how citizen-science data can be used to achieve tangible conservation science and action at local, regional, and global scales. Our examination illustrates how these data can be strategically applied to improve our understanding of spatial and temporal distributions of birds, the impacts of anthropogenic change on ecological systems, and creative conservation solutions to complex problems. We raise awareness of the types of conservation action now happening with citizen-science data, and discuss the benefits, limitations, and caveats of this approach.
Article
Full-text available
Theory predicts low niche differentiation between species over evolutionary time scales, but little empirical evidence is available. Reciprocal geographic predictions based on ecological niche models of sister taxon pairs of birds, mammals, and butterflies in southern Mexico indicate niche conservatism over several million years of independent evolution (between putative sister taxon pairs) but little conservatism at the level of families. Niche conservatism over such time scales indicates that speciation takes place in geographic, not ecological, dimensions and that ecological differences evolve later.
Article
Full-text available
We describe a new species of wren in the genus Thryophilus (Troglodytidae) based on analysis of morphological, vocal, and genetic variation. Individuals of the new species are readily separated in the field or the museum from those of any other wren species, including its closest relatives T. rufalbus and T. nicefori, by a combination of traits including, but not limited to, plumage coloration of the upperparts, the pattern of barring on the wings and tail, overall smaller body size, a richer repertoire of syllable types, shorter trills, and distinctive terminal syllables. The new species is allopatrically distributed in relation to its congeners, being restricted to the dry Cauca River Canyon, a narrow inter-Andean valley enclosed by the Nechí Refuge rainforests and the northern sectors of the Western and Central Andes of Colombia. Individuals or pairs have been found only in remnant patches of dry forest and scrub at 250– 850 m elevation. This newly discovered species is uncommon and threatened because of ongoing transformation of natural habitats in the Cauca River Canyon, and especially because of the planned construction of a major dam in the region; immediate conservation actions are thus imperative. Received 21 February 2012, accepted 19 May 2012.
Article
Diamond, J. M. (Physiology Department, University of California Medical Center, Los Angeles, California 90024) 1977. Continental and insular speciation in Pacific land birds. Syst. Zool. 26:263-268. —Three modes of allopatric speciation can be distinguished, depending on whether the isolating geographic barrier is within a single land mass (“continental speciation”), between islands of the same archipelago, or between different archipelagoes (“insular speciation”). The contributions of these three modes to speciation in Pacific land birds are analyzed. Continental speciation in birds has occurred in no Pacific land mass smaller than Australia, New Guinea, and possibly New Zealand; intraarchipelagal speciation has occurred only on six of the most remote archipelagoes; and inter-archipelagal speciation has produced most of the sympatric bird species pairs from the Bismarcks to Samoa. The frequency of each mode depends on area and isolation of the island, and on mobility and perhaps population density of the taxa involved. What is an “island” to some taxa may be a “continent” to others. For example, New Caledonia behaves as a continent to higher plants, insects, and lizards, but not to birds or ferns. [Speciation; Pacific land birds.]
Book
This work is in press and copyright 2000. The authors do not have digital copies to make available. Requests for copies have bee individually replied to up to now. This comment advises that such requests will henceforth just be declined.
Article
Colonization and subsequent isolation across deep-water barriers is thought to be the primary driver of diversification in insular birds. Shallow-water barriers and intra-island isolation are less well-documented drivers of avian diversification. We examined the relative roles of different geographical barriers in the diversification of Aethopyga sunbirds, a widespread Southeast Asian genus that has its greatest diversity in the Philippine Archipelago. Philippines, Southeast Asia, Wallacea. We reconstructed the phylogenetic relationships among Aethopyga sunbirds with mitochondrial and nuclear DNA sequences. Phylogeny was inferred using concatenated and coalescent frameworks, implemented in maximum likelihood and Bayesian analyses. We used maximum likelihood ancestral state reconstructions to examine the ancestral distribution and colonization history of Aethopyga. To determine whether the diversification at each node occurred within a continent, across a shallow-water barrier, across a deep-water barrier or within an island, we used a series of statements based on the phylogeny, current distribution of species and bathymetric reconstructions. Ancestral state reconstructions inferred that the core Aethopyga ancestor was continental, and that the diversity of Aethopyga on oceanic islands is the result of three or four independent colonization events. Dispersal and subsequent isolation across deep-water barriers was the most common mode of diversification in insular Aethopyga, although intra-island isolation contributed to diversity, producing a small montane radiation within Mindanao. Analyses inferred only a single unequivocal event of diversification across a shallow-water barrier. Deep molecular divergences between phenotypically distinct subspecies suggested that the taxonomy of Aethopyga is overly conservative and obscures biogeographical patterns. We recommend elevating five subspecies, all of which are endemic to the Philippines, to full species. In addition to corroborating the importance of isolation across deep-water barriers, these data also underscore a potential role for additional isolating mechanisms in the generation of biodiversity on oceanic islands. A clade of four montane Aethopyga taxa evolved in situ within the Mindanao sky-island system, an exception to the paradigm that birds do not diversify within the geographical confines of oceanic islands.