A new Myzomela honeyeater (Meliphagidae) from the highlands of Alor Island, Indonesia

Abstract

We here describe a new species of Myzomela honeyeater from the Lesser Sunda island of Alor (southeast Indonesia). The new species is phylogenetically most closely related to Myzomela kuehni from the adjacent island of Wetar and most closely resembles that species in plumage. However, it differs in important morphological, bioacoustic and ecological characteristics. The discovery of a new bird species on Alor is of great biogeographic importance and elevates this island to the status of an Endemic Bird Area. The new Alor Myzomela is restricted to montane eucalypt woodland mostly above 900 m elevation and is currently known from few sites across the island. Based on its occurrence records and human population trends in the highlands of Alor Island, we recommend classification under the IUCN threat status endangered.

Full text

Vol.:(0123456789)
1 3
Journal of Ornithology
https://doi.org/10.1007/s10336-019-01722-2
ORIGINAL ARTICLE
A new Myzomela honeyeater (Meliphagidae) fromthehighlands
ofAlor Island, Indonesia
MohammadIrham1· HidayatAshari1· Suparno1· ColinR.Trainor2· PhilippeVerbelen3· MengYueWu4·
FrankE.Rheindt4
Received: 3 June 2019 / Revised: 9 September 2019 / Accepted: 24 September 2019
© Deutsche Ornithologen-Gesellschaft e.V. 2019
Abstract
We here describe a new species of Myzomela honeyeater from the Lesser Sunda island of Alor (southeast Indonesia). The
new species is phylogenetically most closely related to Myzomela kuehni from the adjacent island of Wetar and most closely
resembles that species in plumage. However, it differs in important morphological, bioacoustic and ecological characteristics.
The discovery of a new bird species on Alor is of great biogeographic importance and elevates this island to the status of
an Endemic Bird Area. The new Alor Myzomela is restricted to montane eucalypt woodland mostly above 900m elevation
and is currently known from few sites across the island. Based on its occurrence records and human population trends in the
highlands of Alor Island, we recommend classification under the IUCN threat status endangered.
Keywords Myzomela· Honeyeater· Alor· Wetar· Plumage· Bioacoustics· Ecological characteristics· Novel species
Zusammenfassung
Eine neue Art von Honigfresser der Gattung Myzomela (Meliphagidae) aus dem Hochland der Alor Insel, Indonesien
In diesem Beitrag beschreiben wir eine neue Art von Honigfresser der Gattung Myzomela von der Kleinen Sunda-Insel Alor in
Südost-Indonesien. Die neue Art ist phylogenetisch am engsten mit Myzomela kuehni von der danebengelegenen Insel Wetar
verwandt, und ähnelt jener Art am meisten in Gefiederfärbung, unterscheidet sich jedoch auch in wichtigen morphologischen,
bioakustischen und ökologischen Eigenschaften. Die Entdeckung einer neuen Vogelart auf Alor ist biogeografisch äußerst
wichtig und erhebt den Status dieser Insel zu einem „Gebiet von Vogel-Endemismus“. Der neue Alor-Honigfresser ist auf eine
Berg-Eukalyptus-Waldlandschaft oberhalb von 900m Höhe beschränkt und ist derzeit von nur wenigen Lokalitäten auf der
Insel bekannt. Aufgrund seines spärlichen Vorkommens und menschlicher Bevölkerungsentwicklungen in den Hochländern
der Insel Alor empfehlen wir seine Klassifizierung unter dem IUCN-Bedrohungsstatus „stark gefährdet“.
Communicated by J. T. Lield.
Electronic supplementary material The online version of this
article (https ://doi.org/10.1007/s1033 6-019-01722 -2) contains
supplementary material, which is available to authorized users.
* Frank E. Rheindt
dbsrfe@nus.edu.sg
1 Research Center forBiology, Indonesian Institute
ofSciences, Jl. Raya Jakarta Bogor KM. 46, Cibinong,
Indonesia
2 Department ofEnvironment andEnergy, Research Institute
fortheEnvironment andLivelihoods, Charles Darwin
University, Darwin, Australia
3 Krijgsgasthuisstraat 89, 9000Ghent, Belgium
4 Department ofBiological Sciences, National University
ofSingapore, 14 Science Drive 4, Singapore,
RepublicofSingapore

Journal of Ornithology
1 3
Introduction
Honeyeaters of the genus Myzomela constitute an important
Australasian bird radiation that has colonized many islands
throughout the Indonesian and Melanesian archipelagos
(Higgins etal. 2018; Eaton etal. 2016). These colorful nec-
tarivores are an important component of insular ecosystems
and often make up some of the most abundant canopy birds
of local avifaunas across eastern Indonesia. Their taxonomy
has gone through upheaval as classifications have differed in
the number of species that have been recognized. There has
been a considerable increase in the number of recognized
species in more recent taxonomic classifications that affords
importance to vocal characters and critical color traits (e.g.,
Eaton etal. 2016; Higgins etal. 2018; Prawiradilaga etal.
2017).
Despite the general conspicuousness of this genus, there
have been recent discoveries of novel species diversity
within Myzomela. Prawiradilaga etal. (2017) described a
new species (M. irianawidodoae) from the island of Rote, a
Timor satellite. This new species bears a surprising resem-
blance in coloration with M. dammermani from distant
Sumba while showing a very different coloration from M.
vulnerata on adjacent Timor. Bioacoustic characters reliably
separate this new Rote Myzomela from all of its regional
congeners, including the similarly plumaged M. dammerm-
ani. In addition, what is thought to be a new species of
Myzomela has been discovered on but not yet described from
Taliabu (Davidson etal. 1991; Rheindt 2010; Rheindt etal.
2014), between Sulawesi and the Moluccan archipelago.
In the present contribution, we formally describe a new
species of Myzomela from the dry Lesser Sundaic island
of Alor (Fig.1), henceforth referred to by the vernacular
name Alor Myzomela. This new species, which superficially
resembles the Wetar Myzomela (M. kuehni) from the adja-
cent island of Wetar (Figs.1, 2), was first documented by
Trainor etal. (2012a) from high-elevation eucalypt forests
on Alor above 900m. In 2015, we collected a single male
specimen from the highlands of Alor that constitutes the
holotype of our description.
Methods
Field visits
We surveyed Alor from sea level to c. 1250m between 2002
and 2016 and Wetar from sea level to 1150m between 2009
and 2010 (Table1 ESM). The first Myzomela observation on
Alor was made in May 2002 on the slopes of Mount Koya
Koya. However, the observation was dismissed at the time
as a likely misidentification (Trainor etal. 2012a). A visit
to Alor in June–July 2009 by PV and Veerle Dossche (VD)
and a visit to Alor in September 2009 by PV resulted in
Fig. 1 Study sites on Alor and Wetar: 1 = Pintumas; 2 = Mutiara Bay;
3 = Sika Island, Mali and Hutan Nostalgia; 4 = Mainang; 5 = Man-
mas-Subo; 6 = Bukapiting; 7 Mt Koya Koya; 8 = Tanglapui Timur-
Naumang; 9 = Ili Moso, Mt Botak, and, 10 = Lerokis. Type specimen
was collected at Manmas (site no. 5). The inset map shows the loca-
tion of Alor and Wetar islands in the Lesser Sundas. The black star
in the inset is placed in the vicinity of Alor Island to facilitate easy
orientation

Journal of Ornithology
1 3
confirmed Myzomela observations above Apui along with
the discovery of a new population of Locustella bush war-
bler (Trainor etal. 2012b). Details of subsequent visits up
to 2012 are described by Trainor etal. (2012a). We also
consider observations by other visitors to Alor and Wetar.
The late John Hornbuckle visited Alor between 14 and 17
October 2012, obtaining photographs of the Alor Myzomela
(Hornbuckle 2012). Stijn Cooleman (SC) visited Alor and
Wetar in November–December 2013 specifically to obtain
additional recordings and photographs of Myzomela hon-
eyeaters on Alor and Wetar. CRT visited Alor in August
2012, discovering the Myzomela at a new site. PV visited
Alor for 7days in April 2014, adding more sound record-
ings and photographs. And CRT visited Alor again in 2016,
including 3days near the mountain of Koya Koya in eastern
Alor. This series of visits to Alor and Wetar was important
in discovering localities of the new Myzomela, obtaining
sound recordings, photographing the birds and character-
izing habitat use on Alor and Wetar. Details about the trip
to collect the type material are outlined below.
Sound recordings andanalysis ofvocalizations
Sound recordings were made with an Olympus LS-10
recorder (CRT), Olympus LS-12 recorder (FER), and an
Edirol R09HR (PV) in combination with Sennheiser ME-66
directional microphones. All recordings are deposited with
the Avian Vocalization Center (AVoCet; www.avoce t.zoolo
gy.msu.edu) and are included in the supplementary online
materials. Sound parameters were measured on sono-
grams in RavenLite (www.birds .corne ll.edu), including the
duration [in seconds (s)], pace [number of notes per second
(n/s)], maximum frequency (Hz), minimum frequency (Hz),
and bandwidth (Hz) measured for a total of 34 homologous
songs across 16 independent recordings of Alor Myzomela
and for five homologous songs across four independent
recordings of Wetar Myzomela. For contact calls, the dura-
tion (s), pace (number of elements per second), maximum
frequency (Hz), minimum frequency (Hz) and bandwidth
(Hz) were measured for 48 homologous elements across 12
independent recordings of Alor Myzomela and 76 homolo-
gous elements across 13 independent recordings of Wetar
Myzomela. A non-parametric Mann–Whitney U test was
used to examine the mean ranks of differences in vocal
parameters between Alor Myzomela and Wetar Myzomela
(online at: https ://www.socsc istat istic s.com/tests /mannw
hitne y/).
Collecting trip andcollecting methods
FER, Suparno and HA spent several days from 19 to 24
November 2015 in the surroundings of the hamlet of Man-
mas (hereafter Manmas-Subo, centered around S8° 17′
51.9″, E124° 43′ 56.0″; Fig.1) in the highlands of Alor
at ~ 1100–1300m to observe, study and collect the new spe-
cies of Myzomela. Birds were mist-netted with 20 ground
nets of 6–12m length. The live captured holotype was sub-
jected to the following measurements, observations and
processing steps: weight, wing length, tarsus length, bill
length, tail length, total length, wing spread, presence of
molt, and bare parts coloration. The holotype was collected
under permit by Indonesian authority in the presence of
Fig. 2 Two myzomelas that resemble each other but significantly dif-
fer in certain characters, especially the extent of the hood. Left: two
individuals of the new Alor Myzomela perched in Eucalyptus uro-
phylla tree at 1250m near Manmas-Subo in April 2014 (photo: PV).
Right: Wetar Myzomela (photo: CRT)

Journal of Ornithology
1 3
two officers from the Indonesian Institute of Science (HA,
Suparno), prepared as a skin by Suparno and deposited in
the Museum Zoologicum Bogoriense (MZB) in Cibinong
(West Java), while liver and muscle tissue samples were
taken for potential future DNA analysis. Further details of
the MZB expedition are provided by Ashari etal. (2018).
We do not believe that our collection of a single Myzomela
specimen would have any negative impacts on the survival
of this species because our collection occurred in relatively
good habitat in which the species remains locally common.
Mitochondrial sequence generation andanalysis
We extracted DNA from the breast muscle tissue of the holo-
type of the new species (n = 1), from Myzomela irianawido-
doae (n = 5) and from an unstudied population of Myzomela
from Taliabu (n = 4) (Davidson etal. 1991; Rheindt 2010;
Rheindt etal. 2014). Then we sequenced the mitochondrial
coding gene NADH dehydrogenase subunit 2 (ND2) fol-
lowing Lim etal. (2019). Using CodonCode Aligner ver-
sion 8.0.2. (CodonCode), our ND2 sequences were trimmed
and assembled. The sequences can be found on Genbank
under accession numbers MN431646 to MN431655. These
sequences of 991bp length were then aligned in MEGA7
(Kumar etal. 2016) with the ND2 sequences of 23 other spe-
cies of Myzomela generated by Marki etal. (2017), including
a sequence of Sugomel nigrum from Driskell and Christidis
(2004) as the outgroup (Table2 ESM) based on its rela-
tive completeness in comparison to other potential outgroup
sequences. We ran two types of phylogenetic analyses with
MEGA7, maximum parsimony and neighbor-joining, and
computed uncorrected raw p-divergences between Wetar and
Alor myzomelas. Maximum likelihood analysis was carried
out in raxmlGUI (Silvestro and Michalak 2012) under the
GTR + Inverse + Gamma model as identified by jModelTest
2.1.7 (Guindon and Gascuel 2003; Darriba etal. 2012).
Results
We here describe the new species of Myzomela honeyeater
from Alor as:
Myzomela prawiradilagae, species nova.
English name: Alor Myzomela.
Indonesian name: Myzomela Alor.
Holotype
Museum Zoologicum Bogoriense (MZB) No. 34.605, pre-
pared as a dry skin by Suparno (Fig.3), field number 018.
Adult male. This is the only known specimen, collected on
21 November 2015 by FER, HA and Suparno at S8° 17′
51.9″ E124° 43′ 56.0″, at 1270m above sea level in the
vicinity of the hamlet of Manmas, south Alor district, Alor
Island, province of Nusa Tenggara Timur, Indonesia. Meas-
urements for this specimen are given in Table1.
Etymology
The new Alor Myzomela is named after Dr. Dewi Malia
Prawiradilaga. One of the first leading female Indonesian
ornithologists, “Ibu Dewi” is the head of the bird division
at the Museum Zoologicum Bogoriense, Research Center
for Biology, Indonesian Institute of Science. She has made
substantial contributions to the development of research on
the ecology and conservation of Indonesian birds.
Description ofholotype
Color nomenclature and definitions follow Smithe (1975),
with color designations capitalized. Head with thick jet
black (89) loral stripe from the base of the bill, extend-
ing as a thin eye ring and slightly beyond the back of the
eye. Forehead and crown between spectrum red (11) and
crimson (108). Auriculars to chin and throat lighter-colored
between spectrum red (11) and geranium (12), extending
down the throat in the form of a bib. Neck, nape, mantle and
back brownish olive (29), separated from crown by discrete
dividing line. Rump and uppertail coverts geranium (12).
breast smoke-grey (44–45), grading into cream color (54)
on belly to undertail coverts, thighs, and flanks. Uppertail
dusky brown (19) to dark greyish brown (20). The outer
rectrices have white margins and tips. Undertail brighter,
approaching sepia (219). Alula white. Upper primaries and
secondaries dusky brown (19) with thin, paler olive-greyish
edges. Upperwing coverts dusky brown (19) with lighter
edges. Thin wing bar formed by pale tips to greater coverts.
Underwing paler in color than upperwing. Tarsus dark grey-
ish horn. Mandibles greyish black. Iris black. Measurements
for this specimen are given in Table1.
Diagnosis
Biometrically, Myzomela prawiradilagae is comparable to
a range of other regional congeners (Table1). In plumage,
the new species most closely resembles the Wetar Myzomela
M. kuehni from Wetar. Our diagnosis from kuehni is based
on comparison of the type specimen of M. prawiradilagae
with a kuehni specimen from the Natural History Museum
at Tring, UK, no. 1904.6.10.21/6611755 (Natural His-
tory Museum 2014) examined by CRT, the description of
M. kuehni by Higgins etal. (2018), the original descrip-
tion by Rothschild (1903) and pictures of a specimen in

Journal of Ornithology
1 3
the American Museum of Natural History, New York, no.
692693, kindly provided by Matthew Shandley (Fig.3). Our
diagnosis from taxa other than kuehni is based on specimen
comparison within the Museum Zoologicum Bogoriense
(MZB, Cibinong, West Java). In the following comparisons,
we follow Eaton etal.’s (2016) taxonomic treatment.
Fig. 3 Visual juxtaposition of the holotype of Alor Myzomela with
specimens of Myzomela kuehni. Left: holotype of the new Alor
Myzomela at Museum Zoologicum Bogoriense, MZB No. 34.605
(photo by MI). Center: specimen of M. kuehni at the Natural His-
tory Museum in Tring, UK, NHM no. 1904.6.10.21/6611755 (photo
by CRT courtesy of NHM). Right: specimen of M. kuehni from
the American Museum of Natural History, New York, AMNH No.
692692 (photo courtesy of AMNH and Matthew Shandley)
Table 1 Measurements of
regional Myzomela taxa taken
from Museum Zoologicum
Bogoriense specimens
All specimens measured by MI, but measurement for M. kuehni extracted from Rothschild (1903). Units
are in millimeters and rounded
n sample size, n/a not available
Species nSex Bill length Wing Tail Tarsus
Myzomela prawiradilagae
(holotype)
1 M 17 62 45 17
Myzomela kuehni 1 M 15 n/a 42 16
Myzomela wakoloensis 2 M 14–16 54 31–32 14
Myzomela chloroptera 3 M 17 54–55 33 14–15
Myzomela eques 1 M 22 69 54 18
Myzomela obscura 2 M 19–20 57–67 40–42 17–19
Myzomela boiei 2 13–15 55–56 33–35 16
Myzomela rosenbergii 2 M 19–20 63 41–42 16
Myzomela dammermani 2 M 17–18 57–58 37 18

Journal of Ornithology
1 3
The adult male M. prawiradilagae is distinguished from
M. kuehni by having a more intensely red-colored head
not extending as far onto the nape and breast (Figs.2, 3).
The red bib on the throat is narrower, shorter and does not
reach the upper breast in M. prawiradilagae. A smoke-grey
upper breast encloses the bib and grades into the cream-
colored belly in M. prawiradilagae, whereas the underparts
of kuehni lack a strong greyish tone to the breast and are
more uniformly whitish with a slight pale yellow tint. M.
prawiradilagae has a stronger olive tint on the upperparts as
compared to kuehni’s more greyish coloration. On the wings,
M. prawiradilagae has obvious pale olive-greyish primary
edging, which is much less conspicuous or absent in kuehni.
Myzomela prawiradilagae is easily distinguished from
Myzomela eques (Papua) and M. obscura (Halmahera) by
its bright-red color especially on the head and its compara-
tively paler upperparts and underparts, while M. eques and
M. obscura are almost uniformly brown to white-brownish
except for their red narrow strip on the throat.
M. prawiradilagae is easily distinguished from all
the members of a large Indonesian radiation of color-
ful Myzomela species, namely M. boiei, M. chloroptera,
M. wakoloensis, M. dammermani, M. irianawidodoae
(Prawiradilaga etal. 2017), M. batjanensis, M. elisabethae,
M. rosenbergii and M. vulnerata, by its dusky brown upper-
wings that are much paler than the black upperwings of all
the other species. Numerous additional plumage marks dis-
tinguish each of those species from M. prawiradilagae.
Vocal analysis
Song comparison betweenAlor andWetar myzomelas
The song of the new Alor Myzomela comprises two or three
introductory notes followed by a main part characterized
by a mean of 11.4 notes over 2s (Fig.4a). The pitch of the
introductory notes (phonetically ‘wick’) rises from 4,800
to 6,430Hz and then falls to 4,960Hz, which is defined as
an overslur; these introductory notes have a different struc-
ture and are slightly lower-pitched than typical contact notes
(Fig.4a, c). The main part’s elements are high-pitched at
7,640Hz with 0.13s between notes (Table2).
The song of the Wetar Myzomela comprises a mean of
21.8 rapidly produced high-pitched twittering notes without
introductory notes (Table2; Fig.4b). The few song record-
ings available had a similar structure but were highly vari-
able in the number of notes. The small sample size available
for songs of Wetar Myzomela made statistical comparisons
problematic, but there were significant differences between
Alor and Wetar Myzomela in maximum frequency and
bandwidth (Table2).
Call comparison betweenAlor andWetar myzomelas
There were substantial statistically significant differences
between calls of the Alor and Wetar myzomelas (Table2).
The Alor Myzomela produces single, high-pitched, smooth,
overslurred (with falling then rising pitch) contact notes
which can be phonetically transcribed as ‘tssip’ or ‘vick’.
Note duration is ~ 0.1s, beginning at 4220Hz, ascending
to 6620Hz and then descending to 4990Hz, with 2–2.5s
breaks between notes (Fig.4c; Table2). We detected some
minor variation from this typical note structure which
involved higher-pitched overslurs (producing a sharp
inverted ‘v’ shape in sonograms) given as double notes at
5940–7370Hz, and ‘N-shaped’ downslurs which initially
rise to 6320Hz, then descend to 4660Hz and then rise
slightly to 5000Hz.
The Wetar Myzomela produces single, high-pitched,
downslurred contact notes (first falling and then rising in
pitch) which can be transcribed phonetically as ‘tsieuw’ or
‘teeyu’. Note duration is ~ 0.15s, at a frequency beginning
at 8870Hz, descending to 4350Hz, with 2.5–3.0s breaks
between notes (Fig.4d).
Phylogenetic analysis
Phylogenetic comparison with a genus-wide dataset of mito-
chondrial ND2 sequences (Marki etal. 2017) confirmed the
morphological assessment that Wetar and Alor Myzomela
are closely related sister species (Fig.5). Uncorrected mito-
chondrial pairwise divergence between the two—at 1.7%—
turned out to be lower than is typical for sister species com-
parisons in many barcoding studies (e.g., Hebert etal. 2004;
Kerr etal. 2007).
Discussion
Taxonomic species rank forthenew Alor Myzomela
Species delimitation has been a perennial problem in biol-
ogy in general, and in ornithology in particular (Wheeler and
Meier 2000). For the longest time throughout the twentieth
century, the ‘Biological Species Concept’ (BSC), which
is based on the principle of reproductive isolation, was
the guiding framework in delineating species (Mayr 1942,
1969). When two or more bird taxa overlap across the same
region, the BSC performs well in species delimitation by
simply observing whether they withstand the ‘test of sympa-
try’, i.e., whether their populations are essentially one unit or
not. However, in allopatry, when two species do not overlap
in space, species delimitation becomes problematic because

Journal of Ornithology
1 3
the question of whether they would recognize each other as
members of the same unit is essentially unknowable. This
allopatry conundrum led to a proliferation of new species
concepts in the 1990s and early 2000s (Wheeler and Meier
2000), with a substantial destabilization in the field of tax-
onomy. By now, most ornithological practitioners have come
around to accept Mayr’s (1996) revised ‘multi-dimensional’
version of the BSC, which explicitly addresses allopatry.
Fig. 4 Sonograms of the
songs of Alor Myzomela and
Wetar Myzomela; a song of
Alor Myzomela with three
introductory notes (Manmas-
Subo, Alor, PV, AV#19827);
b song of Wetar Myzomela
without introductory notes
(Lerokis, Wetar, Stijn Coole-
man, AV#19600); c call of Alor
Myzomela (Manmas-Subo,
Alor, PV, AV#19810); and, d
call of Wetar Myzomela (Ler-
okis, Wetar, Stijn Cooleman,
AV#19600). The y-axis reflects
frequency (0–12kHz) and the
x-axis reflects time (in seconds)

Journal of Ornithology
1 3
Specifically, the unknowability of whether two separate
populations would or would not behave as one species can
be dealt with by comparing their differences to the differ-
ences between other, preferably closely related, species pairs
whose taxonomic status is not in doubt. This is the so-called
yardstick approach (Mayr and Ashlock 1991).
The new Alor Myzomela is the only member of its
genus on Alor. Its closest relative, the Wetar Myzomela,
is geographically disjunct (Fig.1). Hence, our taxonomic
Table 2 Summary statistics (mean ± standard deviation [range]) for measurements of Myzomela notes from Alor and Wetar; Mann–Whitney U
test results comparing mean ranks in measurements between islands
ns not significant, n sample size, n/a not applicable, s seconds, Hz Hertz
* − 0.01; ** − 0.001; *** − 0.0001
Songs Mann–Whitney
U test Z scores
Contact notes Mann–Whitney
U test Z scores
Alor (n = 34) Wetar (n = 5) Alor (n = 48) Wetar (n = 76)
Number of notes 11.4 ± 4.4 (5–24) 21.8 ± 13.3 (7–42) − 1.87ns 1 1 n/a
Duration (s) 1.52 ± 0.85 (0.58–
4.44)
2.21 ± 1.1 (0.90–
3.72)
− 1.44ns 0.11 ± 0.04
(0.04–0.25)
0.15 ± 0.04
(0.07–0.28)
5.74***
Pace (notes per
second [in songs]
or elements per
second [in contact
notes])
0.13 ± 0.05 (0.08–
0.28)
0.11 ± 0.05 (0.06–
0.19)
1.24ns 2.55 ± 1.54
(0.79–7.72)
3.20 ± 1.47
(1.70–7.25)
3.61*
Maximum fre-
quency (Hz)
7637 ± 263 (6973–
8173)
8716 ± 656 (7788–
9576)
− 3.15* 6243 ± 268
(5827–6899)
7956 ± 746
(5774–9047)
8.29***
Minimum frequency
(Hz)
5728 ± 475 (4845–
6384)
5800 ± 155 (5614–
6000)
0.10ns 5023 ± 224
(4514–5416)
4635 ± 330
(4030–5501)
− 6.06***
Bandwidth (Hz) 1909 ± 514 (657–
2807)
2911 ± 580 (2154–
3690)
− 3.04* 1219 ± 284
(634–1841)
3321 ± 766
(1478–4618)
9.25***
Fig. 5 Maximum likelihood topology for NADH dehydrogenase
subunit 2 (ND2). Branch support is depicted at each branch that is
backed by significant bootstrap support ≥ 90 in the following order:
maximum likelihood, maximum parsimony, neighbor-joining. Insig-
nificant branch support values not provided. Maximum branch sup-
port in all three tree-building methods indicated by bold 100

Journal of Ornithology
1 3
assessment must be based on a yardstick comparison of
differences with other Myzomela species pairs. In the fol-
lowing, we perform such yardstick comparisons across four
data types (morphology; vocalizations; ecology; genetics),
leading us to propose species status for the Alor Myzomela.
1. Morphologically, disregarding additional minor plum-
age differences, the most salient distinguishing mark
between Alor and Wetar Myzomela is the extent of their
red bibs, in Wetar reaching well onto the mid-portion of
the breast, whereas in Alor barely covering the throat
(Figs.2, 3). This plumage mark is conspicuous and goes
well beyond the characters that have in the past been
considered sufficient for species-level differences within
the genus, such as slight variations in iridescence in the
all-black plumages of Bismarck Black Myzomela (M.
pammelaena) and Sooty Myzomela (M. tristrami) (Hig-
gins etal. 2018). The geographically most appropriate
comparison is probably with another Lesser Sundaic
species pair: Sumba Myzomela (M. dammermani) and
Rote Myzomela (M. irianawidodoae) display plumage
differences in the extent of their black breast band that
are arguably similar or lesser in magnitude compared
to the differences in bib size between Alor and Wetar
myzomelas, yet extensive vocal differentiation has
led to their recognition as different allopatric species
(Prawiradilaga etal. 2017).
2. The strongest line of evidence for species status is our
vocal dataset. In many songbirds, vocalizations have
been shown to be better indicators of taxonomy than
plumage characters (Rheindt etal. 2008), and the use
of vocalizations in passerine species delimitation has
become routine (Alström and Ranft 2003), including in
Indonesian birds (Rheindt and Hutchinson 2007; Van
Balen etal. 2013; Ng etal. 2016; Gwee etal. 2019). Our
analyses showed that the song of the Alor Myzomela is
a slower, lower-pitched series of notes than that of the
Wetar taxon (Fig.4; Table2). Alor songs are invariably
preceded by introductory elements that are absent on
Wetar. Similarly, contact call notes differ conspicuously
in many aspects including frequency, duration and sound
modulation (Fig.4; Table2). Differences are consistent
across many recordings. This level of vocal differentia-
tion equals that of the only other Myzomela sister spe-
cies pair in Indonesia for which rigorous vocal analyses
have been published (Prawiradilaga etal. 2017). The
Rote (M. irianawidodoae) and Sumba myzomelas (M.
dammermani) have a greater vocal repertoire than the
species on Wetar and Alor, including multiple call types
only found in one but not the other species. However,
in the two main types of vocalization, which are shared
between Rote and Sumba, substantial differences are
detected in one measured parameter each and are argua-
bly not as deep in magnitude as between Alor and Wetar
(see Prawiradilaga etal. 2017; Tables1, 2).
3. Our mitochondrial phylogeny, which includes DNA
material from a wide variety of Myzomela species from
a previous study (Marki etal. 2017) in addition to novel
sequence derived from the holotype of the new species,
confirms plumage-based assumptions that Wetar and
Alor myzomelas are close sister species within the Indo-
nesian Myzomela radiation (Fig.5). However, ND2 dis-
tance between the two species—at ~ 1.7%—is on the low
side of mitochondrial divergence typically associated
with avian species-level differences. Barcoding studies
using the mitochondrial gene COI typically document
divergences around 3% or higher between sister species
(e.g., Hebert etal. 2004; Kerr etal. 2007), as did the
study by Marki etal. (2017) for the genus Myzomela.
However, caution is required with such comparisons, as
the same barcoding studies regularly unearth examples
of biologically distinct pairs of species whose mito-
chondrial divergence can be much lower than the 3%
threshold (see numerous examples in Table1 in Kerr
etal. 2007). Even exceptional hybridization events suf-
fice for genetic introgression to set divergences between
well-established species back to zero (for an overview,
see Rheindt and Edwards 2011), which is conceivable
in the case of these Myzomela species through spo-
radic overwater dispersal. When available, biological
data should trump mitochondrial divergence values in
assigning species rank. We believe that the pronounced
biological differences here documented between Alor
and Wetar myzomelas are more important than a mito-
chondrial divergence that is slightly below some average
threshold.
4. Ecological differences between the two species are
pronounced (see sections below): Wetar Myzomela
occurs at all elevations on its home island, whereas Alor
Myzomela is generally restricted to Eucalyptus wood-
land above 1000m, arguing in favor of much more pro-
nounced differences in ecological preferences and life-
style than is typical among Myzomela sister lineages.
For context, Sumba Myzomela (M. dammermani) and
Rote Myzomela (M. irianawidodoae) do not appreciably
differ in habitat and lifestyle (Prawiradilaga etal. 2017).
Habitat use andavailability
The Alor Myzomela is restricted to montane habitat and is
strongly associated with Eucalyptus woodland on ridge tops
(Fig.1 ESM). The myzomela mainly forages above 1000m,
but was observed feeding at E. alba flowers at ~ 900m on the
slopes of Mt Koya Koya in far eastern Alor and at ~ 900m
in the Manmas area (April 2014). A total of ~ 20 individu-
als were observed at Manmas-Subo at 1000–1270m as

Journal of Ornithology
1 3
singles, pairs and small groups in E. oropsylla woodland,
and inside the village (Fig.2). The holotype was collected
here at ~ 1270m. Up to three birds were heard, recorded
and photographed while feeding in the sub canopy (4–12m
above ground) of E. oropsylla trees, Acacia sp. and Casu-
arina junghuhniana in and around Manmas village. They
fed at flowers, fruits, around leaves and under and around
the bark, gleaning and picking insects and taking nectar.
They were also observed feeding low at about 1m in a
fruiting montane shrub Photinia integrifolia together with
ashy-bellied white-eyes Zosterops citrinellus. Groups of 2–5
birds were also seen feeding with Indonesian honeyeaters
Lichmera limbata in the canopy of flowering trees (J. Horn-
buckle, pers. comm. 2013). At Mainang there was a single
record of three birds (either females or juvenile plumaged
birds) chasing each other in the sub canopy of E. alba wood-
land on a ridge at c. 1180m.
In contrast to Alor Myzomela, the similar-looking Wetar
Myzomela is one of the most widespread birds on its island,
with regular records of singles and pairs from sea level up
to at least 1100m in extensive Eucalyptus woodlands and
tropical forest. It has been found in the sub canopy (5–15m
above ground) and rarely down to 2m above ground. Habitat
is different, but no distinctive behaviors were noted. Like the
Alor Myzomela, the Wetar Myzomela feeds mostly around
flowering trees and shrubs.
The presence of Alor Myzomela broadly corresponds to
the change from Eucalyptus alba to E. oropsylla woodland
at c. 1000m. Elevation itself is unlikely to be the key envi-
ronmental factor underlying distribution patterns, but is a
proxy for spatial variation in the environment (Graham etal.
2014). The Alor Myzomela has probably either been pulled
into the mountains by the low bird densities there and lack
of competition for resources (cf. Mayr and Diamond 1976),
or less likely has been pushed into the mountains by lowland
competitors.
Conservation status andbiogeography
The presence of an endemic species of Myzomela honeyeater
on Alor is of great biogeographic significance. Despite the
proximity of Alor, Wetar and Timor (Fig.1), the three
islands are separated by deep sea (> 4600m deep) and there-
fore would not have been appreciably closer during recent
glacial periods (Voris 2000). Alor’s avifauna has long been
considered to be dominated by a ‘Flores ornis with but little
alteration’ (Hartert 1898) or ‘evidently derived from Flo-
res’ (White and Bruce 1986), but our recent field observa-
tions show that its avifaunal origins are complex and include
several additional species typical of Timor Island (Trainor
etal. 2012a). Alor’s endemic avifaunal component has long
been overlooked: only recently, the distinct endemic Alor
Boobook (or Alor Hawk Owl) Ninox plesseni was upgraded
to species status on the basis of bioacoustic research (Gwee
etal. 2017). A regional field guide additionally elevates an
endemic taxon of cuckooshrike to species level (Alor Cuck-
ooshrike Coracina alfrediana; Eaton etal. 2016), a split that
is not universally followed. Hence, by differing taxonomic
standards, the Alor Myzomela becomes the second or third
endemic bird species of Alor, elevating the island to the
status of an Endemic Bird Area (Stattersfield etal. 1998).
The relative plumage similarity between Alor and Wetar
myzomelas notwithstanding, the avifaunas of Alor and
Wetar show surprisingly little commonality (Stattersfield
etal. 1998; Carstensen and Olesen 2009; Trainor 2010).
Only five restricted-range birds are shared between Alor and
Wetar, mostly wide-ranging and unspecialized small-island
species (e.g., Flame-breasted Sunbird Cinnyris solaris and
Olive-headed Lorikeet Trichoglossus euteles). An additional
two restricted-range bird species are shared between Alor
and Timor, but the taxonomy of one of these (Timor Bush
Warbler Locustella timorensis) on Alor remains unclear
(Trainor etal. 2012b; Verbelen and Trainor 2012; Alström
etal. 2015). In contrast, recent observations from Wetar
have not added any additional avian representatives typical
of Alor or the Flores island group (Trainor etal. 2009a, b).
There is also no precedent for a close taxonomic relationship
(no shared endemic species or endemic subspecies) among
any of the other 89 resident landbirds known from Wetar, or
the 88 resident landbirds known from Alor (Holmes 1995;
Trainor 2005; Trainor etal. 2012a).
The conservation status of the Alor Myzomela is dif-
ficult to assess. Our observations were limited to two sites
separated by 14km in the mountainous center of the island
of Alor and one further observation on the slopes of Mount
Koya Koya (Fig.1). The new species was locally common
at Manmas-Subo, recorded by all visitors in a ~ 10ha area
during dry and wet seasons between 2009 and 2015. The
Manmas-Subo site is located in the Tuti Adagae Nature
Recreation Park, but there appears to be no on-ground
management of natural resources for conservation. On
Alor, habitat above 1000m is extensive (254km2) but
only 120km2 is woodland habitat. We found that not all
of this habitat was occupied by Alor Myzomela, with the
extent of preferred Eucalyptus woodland on ridges prob-
ably covering less than c. 50km2. For example, none
were recorded during a 17km walk between Mainang
and Apui at 850–1271m through Eucalyptus woodland
on 11 January 2010. There is also a tendency for vil-
lages on Alor to be located on ridgetops, which clearly
causes disturbance (e.g., tracks; collection of timber for
firewood and construction). Around Manmas hamlet, the
Alor Myzomela had adapted to feeding in planted Acacia
trees. We expect that the Alor Myzomela may be more
widespread and locally common in suitable habitat on
Alor but further field searches are now required to test this

Journal of Ornithology
1 3
hypothesis. The extensive montane habitat on Mt Koya
Koya (c. 30km2 above 900m) in the far east should be
a priority for further searches, although one of us (CRT)
failed to find Alor Myzomela during 3days in 2016. Given
the great avifaunistic overlap between the two islands, it
may also be worthwhile to explore adjacent Pantar island,
where there has been no ornithological exploration above
900m (Trainor etal. 2012a, b) and no specific searches
for Myzomela honeyeaters.
We propose the IUCN status Endangered for the new
Myzomela prawiradilagae (IUCN 2012) on the basis of cri-
teria B1 and B2 (i.e., extent of occurrence estimated to be
less than 5000km2 and area of occupancy estimated to be
less than 500km2), in combination with our observations
that the range of this new species is severely fragmented, is
currently known to exist at no more than five localities, and
we infer a continuing decline in its area of occupancy given
the increase of habitat degradation in Alor’s highland areas
due to a burgeoning human population.
Acknowledgements Thanks to Stijn Cooleman, James Eaton and Rob-
ert Hutchinson for providing recordings, and the late John Hornbuckle
for providing field observations; and Matthew Shanley of the American
Museum of Natural History for taking photographs of Wetar Myzomela
specimens. PV wishes to thank Veerle Dossche for companionship in
the field during his first visit in July 2009 and Hermolinda S. Kameglet
and Charles J. Mallaka for their hospitality, allowing him to stay in
their house at Apui (April 2014). FER, Suparno and HA are indebted
to Pak Simeon from Manmas village and his family for their hospital-
ity during their stay at Manmas. We wish to thank Hans Matheve for
editing and preparing sonograms and Rohan Fisher for providing data
on the extent of habitat on Alor. We thank the Indonesian Institute of
Science (Lembaga Ilmu Pengetahuan Indonesia-LIPI) and the local for-
estry authorities (BKSDA) for providing permits to conduct field work
and collection. The Ministry of Research and Higher Education of the
Republic of Indonesia issued the research permit No: 9/TKPIPA/E5/
Dit.KI/IX/2015 on 9 September 2015 to FER for the main field work.
The Research Centre for Biology-LIPI was the official sponsor of this
research. FER acknowledges funding from a Singaporean Ministry of
Education Tier 2 grant (R-154-000-A59-112).
References
Alström P, Ranft R (2003) The use of sounds in avian systematics,
and the importance of bird sound archives. Bull Br Orn Club
123A:114–135
Alström P, Xia C, Rasmussen PC, Olsson U, Dai B, Zhao J, Leader
PJ, Carey GJ, Dong L, Cai T, Holt PI, Le Manh Hung, Song G,
Liu Y, Zhang Y, Lei F (2015) Integrative taxonomy of the Russet
Bush Warbler Locustella mandelli complex reveals a new species
from central China. Avian Res 6:9. https ://doi.org/10.1186/s4065
7-015-0016-z
Ashari H, Prawiradilaga DM, Eaton JA, Suparno Rheindt FE (2018)
New records and range extension of birds from Timor, Alor and
Rote. Treubia 45:47–64
Carstensen DW, Olesen JM (2009) Wallacea and its nectarivorous
birds: nestedness and modules. J Biogeogr 36:1540–1550. https
://doi.org/10.1111/j.1365-2699.2009.02098 .x
Darriba D, Taboada GL, Doallo R, Posada D (2012) jModelTest 2:
more models, new heuristics and parallel computing. Nat Methods
9(8):772
Davidson PJ, Lucking RS, Stones AJ, Bean NJ, Raharjaningtrah W,
Banjaransari H (1991) Report on an ornithological survey of
Taliabu. University of East Anglia, Indonesia
Driskell AC, Christidis L (2004) Phylogeny and evolution of the Aus-
tralo–Papuan honeyeaters (Passeriformes, Meliphagidae). Mol
Phylogenet Evol 31:943–960
Eaton JA, van Balen B, Brickle NW, Rheindt FE (2016) Birds of the
Indonesian Archipelago. Greater Sundas and Wallacea. Lynx Edi-
cions, Barcelona
Graham CH, Carnaval AC, Cadena CD, Zamudio KR, Roberts TE,
Parra JL, McCain CM, Bowie RCK, Moritz C, Baines SB, Schnei-
der CJ, VanDerWal J, Rahbek C, Kozak KH, Sanders NJ (2014)
The origin and maintenance of montane biodiversity: integrating
evolutionary and ecological processes. Ecography 37:711–719.
https ://doi.org/10.1111/ecog.00578
Guindon S, Gascuel O (2003) A simple, fast and accurate method to
estimate large phylogenies by maximum-likelihood. Syst Biol
52:696–704
Gwee CY, Christidis L, Eaton JA, Norman JA, Trainor CR, Verbelen
P, Rheindt FE (2017) Bioacoustic and multi-locus DNA data of
Ninox owls support high incidence of extinction and recolonisa-
tion on small, low-lying islands across Wallacea. Mol Phylogenet
Evol 109:246–258
Gwee CY, Eaton JA, Garg KM, Alström P, van Balen S, Hutchin-
son RO, Prawiradilaga DM, Le Manh Hung, Rheindt FE (2019)
Cryptic diversity in Cyornis (Aves: Muscicapidae) jungle-fly-
catchers flagged by simple bioacoustic approaches. Zool J Linn
Soc 186:725–741
Hartert E (1898) On the birds of Lomblen, Pantar, and Alor. Novit
Zool 5:455–465
Hebert PDN, Stoeckle MY, Zemlak TS, Francis CM (2004) Identifica-
tion of birds through DNA barcodes. PLoS Biol 2:e312
Higgins P, Christidis L, Ford H, Bonan A (2018) Honeyeaters
(Meliphagidae). In: del Hoyo J, Elliott A, Sargatal J, Christie DA,
de Juana E (eds) Handbook of the birds of the world alive. Lynx
Edicions, Barcelona
Holmes DA (1995) Additions to the avifauna of Pulau Alor, Nusateng-
gara. Kukila 7:155–156
Hornbuckle J (2012) Eastern Indonesia: 26th September–25th October.
http://www.surfb irds.com/trip_repor t.php?id=2274. Accessed 10
Sept 2014
IUCN (2012) IUCN red list categories and criteria: version 3.1, 2nd
edn. IUCN, UK, p 32
Kerr KCR, Stoeckle MY, Dove CJ, Weigt LA, Francis CM, Hebert
PDN (2007) Comprehensive DNA barcode coverage of North
American birds. Mol Ecol Notes 7:535–543
Kumar S, Stecher G, Tamura K (2016) MEGA7: molecular evolution-
ary genetics analysis version 7.0 for bigger datasets. Mol Biol
Evol 33:1870–1874
Lim BT, Sadanandan KR, Dingle C, Leung YY, Prawiradilaga DM,
Irham M, Ashari H, Lee JG, Rheindt FE (2019) Molecular evi-
dence suggests radical revision of species limits in the great spe-
ciator white-eye genus Zosterops. J Ornithol 160:1–6
Marki PZ, Jønsson KA, Irestedt M, Nguyen JMT, Rahbek C, Fjeldså J
(2017) Supermatrix phylogeny and biogeography of the Australa-
sian Meliphagides radiation (Aves: Passeriformes). Mol Phylo-
genet Evol 107:516–529
Mayr E (1942) Systematics and the origin of species, from the view-
point of a zoologist. Harvard University Press, Cambridge
Mayr E (1969) Principles of systematic zoology. McGraw-Hill, New
Yor k
Mayr E (1996) What is a species and what is not? Philos Sci
63:262–277

Journal of Ornithology
1 3
Mayr E, Ashlock PD (1991) Principles of systematic zoology, 2nd edn.
McGraw-Hill, New York
Mayr E, Diamond J (1976) Birds on islands in the sky: origin of the
montane avifauna of northern Melanesia. Proc Natl Acad Sci USA
73(5):1765–1769. https ://doi.org/10.1073/pnas.73.5.1765
Natural History Museum (2014) Dataset: collection specimens.
Resource: specimens. Natural history museum data portal (data.
nhm.ac.uk). https ://doi.org/10.5519/00029 65
Ng EYX, Eaton JA, Verbelen P, Hutchinson RO, Rheindt FE (2016)
Using bioacoustic data to test species limits in an Indo–Pacific
island radiation of Macropygia cuckoo doves. Biol J Linn Soc
118:786–812
Prawiradilaga DM, Baveja P, Suparno Ashari H, Ng NSR, Gwee
CY, Verbelen P, Rheindt FE (2017) A colourful new species of
Myzomela honeyeater from Rote island in eastern Indonesia. Treu-
bia 44:77–100
Rheindt FE (2010) New biogeographic records for the avifauna of Tali-
abu (Sula islands, Indonesia), with preliminary documentation of
two previously undiscovered taxa. Bull Br Orn Club 130:33–51
Rheindt FE, Edwards SV (2011) Genetic introgression: an integral but
neglected component of speciation in birds. Auk 128:620–632
Rheindt FE, Hutchinson RO (2007) A photoshot odyssey through the
confused avian taxonomy of Seram and Buru (southern Moluc-
cas). Bird Asia 7:18–38
Rheindt FE, Norman JA, Christidis L (2008) DNA evidence shows
vocalizations to be a better indicator of taxonomic limits than
plumage patterns in Zimmerius tyrant-flycatchers. Mol Phylogenet
Evol 48:150–156
Rheindt FE, Prawiradilaga DM, Suparno Ashari H, Wilton PR (2014)
New and significant island records, range extensions and eleva-
tional extensions of birds in eastern Sulawesi, its nearby satellites,
and Ternate. Treubia 41:61–91
Rothschild WMP (1903) Myzomela kuehni n. sp. Bull Br Orn Club
95(13):24
Silvestro D, Michalak I (2012) raxmlGUI: a graphical front-end for
RAxML. Org Divers Evol 12:335–337
Smithe FB (1975) Naturalist’s color guide. Part 111. Color guide.
American Museum of Natural History, New York
Stattersfield AJ, Crosby MJ, Long AJ, Wege DC (1998) Endemic Bird
Areas of the world: priorities for biodiversity conservation, 7th
edn. BirdLife International, Cambridge
Trainor CR (2005) Species richness, habitat use and conservation of
birds of Alor island, Lesser Sundas, Indonesia. Emu 105(2):127–
135. https ://doi.org/10.1071/MU030 30
Trainor CR (2010) Timor’s fauna influence of scale, history and land-
use on faunal patterning. PhD Thesis, Charles Darwin University
Trainor CR, Imanuddin Aldy F, Verbelen P, Walker JS (2009a) The
birds of Wetar, Banda sea: one of Indonesia’s forgotten islands.
Bird Asia 10:78–93
Trainor CR, Imanuddin, Aldy F, Walker JS (2009b) The status and con-
servation of the Endangered Wetar ground-dove (Gallicolumba
hoedtii) and other wildlife on Wetar island, Indonesia, 2008. Final
technical report, columbidae conservation
Trainor CR, Verbelen P, Johnstone RE (2012a) The avifauna of Alor
and Pantar, Lesser Sundas, Indonesia. Forktail 28:77–92
Trainor CR, Verbelen P, Hoste S (2012b) Rediscovery of the Timor
Bush-warbler Locustella (Bradypterus) timorensis on Alor and
Timor, Wallacea: clarifying taxonomic affinities, defining habitat
and survey recommendations. Bird Conserv Int 22(3):354–369.
https ://doi.org/10.1017/S0959 27091 10005 30
Van Balen B, Eaton JA, Rheindt FE (2013) Biology, taxonomy and
conservation status of the short-tailed Green Magpie Cissa thalas-
sina from Java. Bird Conserv Int 23:91–109
Verbelen P, Trainor CR (2012) Rediscovery of the Timor Bush Warbler
on Alor and Timor, Wallacea, Indonesia. Bird Asia 17:47–48
Voris HK (2000) Maps of Pleistocene sea levels in south-
east Asia: shorelines, river systems and time durations.
J Biogeogr 27(5):1153–1167. https ://doi.org/10.104
6/j.1365-2699.2000.00489 .x
Wheeler QD, Meier R (2000) Species concepts and phylogenetic the-
ory. Columbia University Press, New York
White CMN, Bruce MD (1986) The birds of Wallacea (Sulawesi, the
Moluccas & Lesser Sunda islands Indonesia): an annotated check-
list. British Ornithologists’ Union, London (Check-list 7)
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.