ArticlePDF Available

Speciation in Indo-Pacific swiftlets (Aves: Apodidae): Integrating molecular and phenotypic data for a new provisional taxonomy of the Collocalia esculenta complex

Authors:

Abstract and Figures

White-bellied swiftlets of the Collocalia esculenta complex constitute a radiation of colony-breeding swifts distributed throughout the tropical Indo-Pacific region. Resolution of their taxonomy is challenging due to their morphological uniformity. To analyze the evolutionary history of this complex, we combine new biometric measurements and results from plumage assessment of museum specimens with novel as well as previously published molecular data. Together, this body of information constitutes the largest systematic dataset for white-bellied swiftlets yet compiled, drawn from 809 individuals belonging to 32 taxa for which new molecular, biometric, and/or plumage data are presented. We propose changing the classification of white-bellied swiftlets, for which two species are currently recognized, to elevate eight regional forms to species level, and we also describe two new subspecies. The ten taxa we recommend recognizing at the species level are: Collocalia linchi (Java to Lombok, Sumatran hills), C. dodgei (montane Borneo), C. natalis (Christmas Island), C. affinis (Greater Sundas, including the Thai-Malay Peninsula and Andaman–Nicobar Islands), C. marginata (Philippines), C. isonota (Philippines), C. sumbawae (west Lesser Sundas), C. neglecta (east Lesser Sundas), C. esculenta (Sulawesi, Moluccas, New Guinea, Bismarck Archipelago, Solomon Islands), and C. uropygialis (Vanuatu, New Caledonia). Future molecular and morphological work is needed to resolve questions of speciation and population affinities in the Philippines, Christmas Island, Wallacea and central Melanesia, and to shed light on historic diversification and patterns of gene flow in the complex.
No caption available
… 
Content may be subject to copyright.
Accepted by P. Rasmussen: 1 Mar. 2017; published: 7 Apr. 2017
ZOOTAXA
ISSN 1175-5326 (print edition)
ISSN
1175-5334
(online edition)
Copyright © 2017 Magnolia Press
Zootaxa 4250 (5): 401
433
http://www.mapress.com/j/zt/
Article
401
https://doi.org/10.11646/zootaxa.4250.5.1
http://zoobank.org/urn:lsid:zoobank.org:pub:BB14617A-A05F-4C0D-82E7-0EA575A120B3
Speciation in Indo-Pacific swiftlets (Aves: Apodidae): integrating molecular
and phenotypic data for a new provisional taxonomy of the Collocalia esculenta
complex
FRANK E. RHEINDT
1
, LES CHRISTIDIS
2,3
, JANETTE A. NORMAN
3
, JAMES A. EATON
4
,
KEREN R. SADANANDAN
1
& RICHARD SCHODDE
1
Department of Biological Sciences, National University of Singapore, 14 Science Drive 4, Singapore 117543
E-mail: dbsrfe@nus.edu.sg
2
Department of Genetics, University of Melbourne, Parkville, Victoria 3052, Australia
3
National Marine Science Centre, Southern Cross University, Coffs Harbour, NSW 2054, Australia
4
A-3A-5, Casa Indah 1, Persiaran Surian, Petaling Jaya, Selangor, Malaysia
?Australian National Wildlife Collection, National Research Collections Australia, CSIRO, Australian Capital Territory 2601,
Australia
Abstract
White-bellied swiftlets of the Collocalia esculenta complex constitute a radiation of colony-breeding swifts distributed
throughout the tropical Indo-Pacific region. Resolution of their taxonomy is challenging due to their morphological uni-
formity. To analyze the evolutionary history of this complex, we combine new biometric measurements and results from
plumage assessment of museum specimens with novel as well as previously published molecular data. Together, this body
of information constitutes the largest systematic dataset for white-bellied swiftlets yet compiled, drawn from 809 individ-
uals belonging to 32 taxa for which new molecular, biometric, and/or plumage data are presented. We propose changing
the classification of white-bellied swiftlets, for which two species are currently recognized, to elevate eight regional forms
to species level, and we also describe two new subspecies. The ten taxa we recommend recognizing at the species level
are: Collocalia linchi (Java to Lombok, Sumatran hills), C. dodgei (montane Borneo), C. natalis (Christmas Island), C.
affinis (Greater Sundas, including the Thai-Malay Peninsula and Andaman–Nicobar Islands), C. marginata (Philippines),
C. isonota (Philippines), C. sumbawae (west Lesser Sundas), C. neglecta (east Lesser Sundas), C. esculenta (Sulawesi,
Moluccas, New Guinea, Bismarck Archipelago, Solomon Islands), and C. uropygialis (Vanuatu, New Caledonia). Future
molecular and morphological work is needed to resolve questions of speciation and population affinities in the Philippines,
Christmas Island, Wallacea and central Melanesia, and to shed light on historic diversification and patterns of gene flow
in the complex.
Key words: speciation, DNA sequencing, morphology, Wallacea, Indo-Pacific, zoogeography
Introduction
Swiftlets of the Collocalia esculenta (Linnaeus, 1758) complex, known collectively as white-bellied swiftlets and
characterized by blue-green glossed backs and white bellies, range widely across the archipelagos of the tropical
Indian and west Pacific Oceans (Figure 1). The complex presents a geographic mosaic of phenotypic variation that
has been difficult to interpret evolutionarily. Most previous classifications have relied on morphological characters
for taxon delimitation and identification. Regional forms differ only in subtle plumage characters, most of which
are difficult or impossible to discern in free-flying birds. Consequently, grouping of taxa into species has differed
widely, with species boundaries often drawn along different geographic lines (Oberholser 1906; Stresemann 1912,
1940; Salomonsen 1983; Somadikarta 1986; Chantler 1999).
Wallace (1864) first separated the complex into three species—C. linchi Horsfield & Moore, 1854, C.
esculenta (Linnaeus, 1758) and a third species then unnamed but later referred to as C. uropygialis G.R. Gray,
1866—based on the presence or absence of a white rump band and large white spots normally hidden on the inner
RHEINDT ET AL.
402
·
Zootaxa 4250 (5) © 2017 Magnolia Press
rectrix vanes. According to Wallace (1864), dark-rumped, plain-tailed linchi occurs from Java west to the
Andaman Islands; dark-rumped, spot-tailed esculenta ranges from Sulawesi east to Papua; and the then-unnamed
white-rumped, spot-tailed uropygialis, is endemic to Vanuatu. To these, the following subsequently described taxa
were added: dull, spot-tailed C. neglecta G. R. Gray, 1866 from Timor, east Lesser Sundas; pale-rumped, plain-
tailed C. marginata Salvadori, 1882 from the Philippines; greyish, spot-tailed C. natalis Lister, 1889 from
Christmas Island, Indian Ocean; and small, plain-tailed C. dodgei Richmond, 1905 from Mt. Kinabalu, Borneo.
Stresemann (1912) accepted linchi as separate from esculenta, but (1921, 1925) subsequently combined all other
taxa in a single polytypic species, esculenta, an arrangement followed by many authorities (e.g. Salomonsen 1983;
White & Bruce 1986; Inskipp et al. 1996; Wells 1999).
FIGURE 1. Distribution of the white-bellied swiftlet (Collocalia esculenta) complex across the Indo-Pacific region, with
Stresemann’s Line separating eastern and western taxa.
Later, Stresemann (1940) accepted Wallace’s (1864) rationale for partitioning taxa by presence of white tail
spotting, and provided further detail on the geographic course of the boundary. “Stresemann’s Line" (Figure 1), as
Somadikarta (1986) called it, separated western, plain-tailed linchi from eastern, spot-tailed esculenta along a line
between the Philippines and Sulawesi, then down the Makassar Strait between Borneo and Sulawesi, and between
Sumbawa and Lombok , and finally curving west to include Christmas Island within eastern C. esculenta. Western
and eastern groups have since been widely recognized, even if not always as species (e.g. Stresemann 1912;
Salomonsen 1983; White & Bruce 1986); Peters (1940) and Sibley & Monroe (1990) separated Philippine
populations (marginata) as a species as well. Somadikarta (1986) then demonstrated syntopy in the western plain-
tailed group, between (1) linchi sensu stricto, with greenish-glossed plumage and a naked hind toe on Java, Bali,
Lombok, hills of Sumatra and montane Borneo; and (2) the other Sundaic plain-tailed form with bluish-glossed
plumage and feathered hallux, throughout Sumatra, Borneo, peninsular Malaysia and surrounding islands. Despite
showing that the two overlap in Sumatran hills and Mt. Kinabalu, indicative of speciation and providing grounds
for recognizing the bluish-glossed, plain-tailed form as another species, Somadikarta (1986) retained the latter in C.
esculenta.
DNA sequencing, although limited in taxon sampling, has supported Somadikarta’s (1986) findings for the
western, plain-tailed group. Analyses of mitochondrial DNA (mtDNA; Lee et al. 1996; Clayton & Johnson 2000;
Thomassen et al. 2003, 2005; Price et al. 2004, 2005) have showed deep divergences between C. linchi sensu
Zootaxa 4250 (5) © 2017 Magnolia Press
·
403
NEW COLLOCALIA ESCULENTA COMPLEX
Somadikarta (1986) and various populations of C. esculenta, indicating that bluish-glossed, plain-tailed western
white-bellied swiftlets from Borneo and peninsular Malaysia (C. e. cyanoptila Oberholser, 1906) are more closely
related to Philippine and eastern spot-tailed taxa than to C. linchi from its terra typica in Java. Some of these
studies suggested additional species-level divisions within both C. esculenta and C. linchi sensu Somadikarta
(1986). Based on ~2.2% and 4.0% pairwise mtDNA divergence, respectively, Clayton & Johnson (2000) argued for
specific status for both C. e. bagobo Hachisuka, 1930 (southernmost Philippines) and C. e. marginata (remaining
Philippines). Moyle et al. (2008) found that Mt. Kinabalu dodgei, which Somadikarta (1986) had included in
linchi, is rather deeply diverged in DNA from its sister, Javan linchi, and still further from and paraphyletic with
the syntopic C. esculenta cyanoptila. Moyle et al. (2008) raised C. dodgei to species status, and recovered western
plain-tailed and eastern spot-tailed esculenta as paraphyletic with respect to linchi. Price et al. (2004, 2005) also
found deep division on either side of Stresemann’s Line.
Most recent sources have followed Somadikarta’s (1986) two-species arrangement, recognizing as species
only his “greenish-glossed” linchi and “bluish-glossed” esculenta, the latter with c. 30 plain- and spot-tailed forms
west and east of Stresemann’s Line, respectively (Chantler 1999; Dickinson 2003; Clements 2007; Clements et al.
2014; Dickinson & Remsen 2013; del Hoyo & Collar 2014). By documenting sympatry in plain-tailed western
populations, Somadikarta (1986) demonstrated that reliance on a single morphological character (tail spots), as
done by Wallace (1864) and Stresemann (1912 1940), was insufficient for delineating species. Indeed, Gill &
Donsker (2016), who followed the split of C. dodgei from C. linchi on Mt. Kinabalu, nevertheless kept Sundaic
plain-tailed western “esculenta” (cyanoptila) in spot-tailed eastern esculenta, despite the considerably greater
molecular distance between them (Moyle et al. 2008). The circumscription of esculenta as an inclusive species
ranging from the Bay of Bengal (Andaman Islands) to New Caledonia was not necessarily intended by
Somadikarta (1986), who noted that he had “…not yet decided upon the relationship between the esculenta taxa
without tail spots to the west … of Stresemann’s line and those with tail spots to the east”. Nevertheless, east of
Somadikarta’s focal area, all spot-tailed taxa have remained lumped in one broad species, C. esculenta.
Our purpose here is to add resolution to the systematics and biogeography of the white-bellied swiftlet
complex by integrating new and previously published mtDNA sequence data with a morphological review of
white-bellied swiftlet populations across the entire range of the complex.
Material and methods
For mtDNA analysis, a 406-bp fragment of cytochrome-b (cyt b) was sequenced for 15 individuals of the C.
esculenta complex plus, as outgroups, two individuals from other swift genera, Hemiprocne mystacea and
Hirundapus caudacutus (Table 1). We used DNA extraction, PCR amplification and sequencing procedures as
described by Norman et al. (1998) with PCR annealing temperatures of 52–58°C. The primers CeCb-L
(CCAAATATCMTTCTGAGGYG) and CeCb-H (TTCTGGTTTGATRTGGGGG) were designed specifically for
amplification of Collocalia esculenta, using sequences presented by Lee et al. (1996). We checked DNA sequences
for stop codons and reading frame. A lack of indels rendered sequence alignment straightforward. Using the
Akaike information criterion as implemented in the program jModelTest (Posada 2008), we found that the best fit
among 88 evolutionary models was a TrN+I+G model. Homologous GenBank sequences were available for
incorporation into the dataset from another 28 Collocalia individuals sequenced in four other studies (Lee et al.
1996; Clayton & Johnson 2000; Price et al. 2004; Päckert et al. 2012). Identity, source, collection locality, voucher
information and GenBank accession numbers of all sequences are listed in Table 1. The study by Moyle et al.
(2008) did not employ cyt b sequences, so its sequence information could not be included in our dataset. We did not
include sequences from Thomassen et al. (2003) in our study: these sequences are all derived from blood samples
and hence have a high likelihood of containing ‘numts’, which are nuclear paralogs of mtDNA genes (Sorenson &
Quinn 1998; Bensasson et al. 2001; Rheindt et al. 2014). We utilized maximum parsimony (MP), maximum
likelihood (ML) and Bayesian inference to construct phylogenetic relationships using the programs MEGA5
(Tamura et al. 2011) and BEAST 1.6.1 (Drummond & Rambaut 2007), respectively. Support for individual nodes
was estimated through heuristic bootstrap resampling using 1000 replicates. Following jModelTest results, we set
evolutionary parameters to a Tamura-Nei model with a gamma shape parameter and a proportion of invariable sites
in ML analysis but let the program estimate other parameters. In BEAST analysis, we carried out three independent
RHEINDT ET AL.
404
·
Zootaxa 4250 (5) © 2017 Magnolia Press
runs over 10,000,000 generations, sampling trees every 1000 generations for the evaluation of posterior
probabilities and using a TrN+I+G model for each run. Before actual analysis, we conducted a test run of 500,000
generations to tune operator settings for the main runs. Likelihood versus generation plots generated in Tracer
Version v1.4.1 (Rambaut & Drummond 2008) showed that a burn-in of 10% was appropriate for all runs. We
evaluated convergence using Tracer, making sure that Bayesian runs reached an effective sample size far greater
than 200 at burn-in. The three runs were then merged and a further burn-in of 10% was applied to the final tree.
Raw genetic p-divergences among taxa were computed with the program MEGA5 (Tamura et al. 2011) using 1000
bootstrap replicates. Genetic divergences were used to compute an approximate scenario for the timing of
diversification in the C. esculenta complex based on Weir & Schluter’s (2008) clock rate for avian mtDNA coding
genes, although any such estimate must be considered provisional without further supporting evidence.
For morphological analysis, RS and JAE took biometric measurements from 531 specimens and assessed
plumage traits in 809 specimens representing 29 of the 35 species-group taxa recognized by Dickinson & Remsen
(2013) and Gill & Donsker (2016) from 68 localities across the entire range of the complex (Table 2). This material
was made available by 11 museums listed in the Acknowledgements. Geographic place names used are
biogeographic and follow the Times Comprehensive Atlas of the World (2011). The formal names for the
provinces of Nusa Tenggara (formerly Lesser Sundas) were not adopted because they are incongruent with
Collocalia zoogeogeography. In the text below, we use the term “west Nusa Tenggara” collectively for the islands
of Lombok, Sumbawa, Flores and Sumba, and “east Nusa Tenggara” for all islands of the chain east from and
including Pantar and Alor, and Savu and Timor. “Wallacea” is used here in its traditional sense for the Indonesian
archipelago east of the Asian Sunda Shelf and west of the Australasian Sahul Shelf (Darlington 1957).
Wing chord measurements were taken from the carpal joint to the tip of the longest primary whilst the tail was
measured from the base of the central tail feathers to the tips of the rectrices T1 and T5. Primaries were numbered
descendently, from inner to outer. Plumage traits noted were those found to be taxonomic markers in previous
studies (e.g. Stresemann 1940; Salomonsen 1983; Somadikarta 1986) and were scored across specimens laid out in
series. The traits documented comprised: dorsal tone and gloss; ventral pattern; brightness of pale lore spot;
presence of spicate (needle-like) white aftershafts in mantle down; presence, form and distribution of large white
spots on inner vanes of rectrices; whiteness of rump; and hind toe and tarsal feathering. Glossed dorsal colour,
which varies under different angles of incident light, was assessed from specimens laid out in regional series at the
same angle, c. 90°. Descriptions of plumage pattern and colour tone in the text and Table 2 are provided at the level
of fine shades to ensure the documentation of normal variation within populations and to compare differentiation
among taxa. Sexes were pooled in measurements because females are as large as males, or even sometimes
marginally larger (Mayr 1931; Salomonsen 1983), and specimens molting primary 9 or rectrices were excluded.
The effects of wear, which dulls and greens dorsal gloss and darkens the breast through wear of white feather
fringing, were taken into account. First-year immatures, which have duller dorsal gloss, were identified by their
pale pink feet and then excluded from adult comparisons.
To delimit taxa, we first produced a phylogenetic frame of regional relationships from our mtDNA sequences
of 45 individuals (Figure 2) as well as pairwise mtDNA divergence values, and then contrasted molecular results
with our morphological data from all regional populations. Because divergence values from our partial cyt b
sequences may be imprecise and prone to higher stochastic fluctuations than those from longer DNA fragments, we
only use divergence values in combination with morphological characters or strongly supported topological
features of our mtDNA tree as a basis for taxonomic recommendations.
The morphometric measurement data and sample sizes for each taxon are summarized in Appendix I. The
results of pairwise comparisons of wing and wing-tail ratios are displayed in Appendix II, where a Bonferroni-
corrected p-value of 0.0001 was applied to test for significance.
Results
Results from mtDNA analysis are presented in Figure 2 and Table 3. MP, ML and Bayesian analyses were
congruent across strongly-supported nodes, and the resultant tree recovered each taxon as a monophyletic lineage,
albeit with varying nodal support (Figure 2). No stop codons were discovered in any of the sequences. Inter- and
intra-taxon cyt b divergences are given in Tables 3 and S1.
Zootaxa 4250 (5) © 2017 Magnolia Press
·
405
NEW COLLOCALIA ESCULENTA COMPLEX
7$%/(                   
    
cyanoptila

7D[RQ


*HQEDQN$FFHVVLRQ
1XPEHU
/RFDOLW\ 6RXUFH
natalis   
natalis   
natalis   
sumbae   
sumbae   
sumbae   
sumbawae   
neglecta   
nitens  


nitens   
nitens   
nitens   
nitens   et alb
becki A S Pet al

cyanoptila   et al
cyanoptila   et al
cyanoptila   et al
cyanoptila   et al
cyanoptila   et al
cyanoptila   et al
cyanoptila   et al.

cyanoptila   et al.
cyanoptila   et al.
 ......continued on the next page
RHEINDT ET AL.
406
·
Zootaxa 4250 (5) © 2017 Magnolia Press
7$%/( 
7D[RQ


*HQEDQN$FFHVVLRQ
1XPEHU
/RFDOLW\ 6RXUFH
cyanoptila   et al.

cyanoptila   et al.

cyanoptila   et al.
bagobo   et al.
bagobo   et al.
bagobo   et al
bagobo   et al
marginata   et al.
marginata   et al
marginata   et al
linchi   et al.
linchi   et al.
linchi   et al.
linchi   et al
linchi   et al.
linchi   et al.
troglodytes   et al
troglodytes   et al.
troglodytes   et al.
esculenta   
esculenta   
esculenta   
Hemiprocne mystacea   
Hirundapus caudacutus   
Zootaxa 4250 (5) © 2017 Magnolia Press
·
407
NEW COLLOCALIA ESCULENTA COMPLEX
7$%/(


,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
$QGDPDQDQG
1LFREDU
,VODQGV
affinis affinis










cyanoptila



affinis%HDYDQ

0HUJXL,VODQGV
affinis
elachyptera






±









affinis

affinis%HDYDQ

1LDVZHVWRI
6XPDWUD
affinis
vanderbilti






±










±



cyanoptila

affinis%HDYDQ

%DWXDQG
0HQWDZDL
,VODQGV
affinis
oberholseri






±






±


linchi

C. esculenta


affinis%HDYDQ

6XPDWUD
ORZODQG
affinis
cyanoptila




   



 



affinis%HDYDQ

  ……continued on the next page
RHEINDT ET AL.
408
·
Zootaxa 4250 (5) © 2017 Magnolia Press
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
%DOL/RPERN
linchi dedii











linchi

linchi
+RUVILHOG
0RRUH
&KULVWPDV
,VODQG
natalis




















esculenta

esculenta

C.linchi

natalis
/LVWHU
%RUQHRXSSHU
0W.LQDEDOX
dodgei













C. affinis cyanoptila
C.linchi

et al.


dodgei
5LFKPRQG
3KLOLSSLQHV±
%DEX\DQ
,VODQGV
marginata
septentrionalis























marginata




marginata
6DOYDGRUL
  ……continued on the next page
Zootaxa 4250 (5) © 2017 Magnolia Press
·
409
NEW COLLOCALIA ESCULENTA COMPLEX
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
3KLOLSSLQHV±
QRUWK/X]RQ
0RXQWDLQV
isonota isonata









c
bagobo



marginata




isonota
2EHUKROVHU
3KLOLSSLQHV
PLG/X]RQWR
PLG
3KLOLSSLQHV
"3DODZDQ
marginata
marginata



















marginata




marginata
6DOYDGRUL
3KLOLSSLQHV
0LQGRUR
0LQGDQDR6XOX
$UFKLSHODJR
isonota bagobo














marginata



et al.
bagobomarginata

isonota
2EHUKROVHU
1XVD7HQJJDUD
6XPEDZD
sumbawae
sumbawae























linchi
neglecta


C.
neglectaq.v
sumbawae
6WUHVHPDQQ
  ……continued on the next page
RHEINDT ET AL.
410
·
Zootaxa 4250 (5) © 2017 Magnolia Press
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
1XVD7HQJJDUD
)ORUHV
sumbawae
sumbawae



















sumbawae



sumbawae
6WUHVHPDQQ
1XVD7HQJJDUD
6XPED
sumbawae
sumbae


















 sumbawae


sumbawae
6WUHVHPDQQ
1XVD7HQJJDUD
6DYX5RWL
7LPRU
neglecta
neglecta





















esculenta 

natalis
natalis




neglecta*5
*UD\
1XVD7HQJJDUD
$ORU:HWDU
.LVDU
neglecta
perneglecta























neglecta
sumbawae




neglecta*5
*UD\
  ……continued on the next page
Zootaxa 4250 (5) © 2017 Magnolia Press
·
411
NEW COLLOCALIA ESCULENTA COMPLEX
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
6RXWK%DQGD
6HD,VODQGV
5RPDQJ
'DPDU
7DQLPEDU
neglecta
perneglecta

























neglecta
esculenta


neglecta
LQWHUJUDGLHQW
neglecta-esculenta"
1RUWK6XODZHVL
6DQJLKHDQG
7DODXG,VODQGV
esculenta
manadensis













esculenta
nitens

esculenta /LQQDHXV

&HQWUDOVRXWK
6XODZHVL
esculenta
esculenta
















esculenta /LQQDHXV

)ORUHV6HD
,VODQGV
7DQDKMDPSHD
.DODR
esculenta
minuta
















esculenta /LQQDHXV

%DQJJDL6XOD
,VODQGV
esculenta
esculenta




















esculenta /LQQDHXV

  ……continued on the next page
RHEINDT ET AL.
412
·
Zootaxa 4250 (5) © 2017 Magnolia Press
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
6RXWK0DOXNX
WR.DLDQG$UX
,VODQGV
esculenta
esculenta


















esculenta



esculenta /LQQDHXV

1RUWK0DOXNX
esculenta
spilura
















esculenta 




esculenta /LQQDHXV

:HVWHUQ
3DSXDQ,VODQGV
:DLJHR
esculenta
amethystina















esculenta /LQQDHXV

&HQGHUDZDVLK
%D\,VODQGV
1XPIRRU
"%LDN
esculenta
numforensis












nitens




spilura
esculenta /LQQDHXV

1HZ*XLQHD
VDWHOOLWHLVODQGV
DQG&DSH<RUN
3HQLQVXOD±c
PDVO
esculenta nitens


















manadenis





esculenta /LQQDHXV

  ……continued on the next page
Zootaxa 4250 (5) © 2017 Magnolia Press
·
413
NEW COLLOCALIA ESCULENTA COMPLEX
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
$OSLQHZHVW
1HZ*XLQHD!
c.PDVO
esculenta nitens
erwini
V\QRQ\PL]HG
nitens


nitens


nitens


nitens


±±nitens








esculenta /LQQDHXV

/RXLVLDGHDQG
7UREULDQG
$UFKLSHODJRHV
esculenta
misimae









 






nitens



tametamele

esculenta /LQQDHXV

%LVPDUFN
$UFKLSHODJR
$GPLUDOW\
,VODQGV
esculenta
stresemanni

















nitens 




esculenta /LQQDHXV

%LVPDUFN
$UFKLSHODJR
1HZ,UHODQG
1HZ+DQRYHU
esculenta
heinrothi


















nitens





stresemanni nitens



heinrothi

stresemannikalili

esculenta /LQQDHXV

  ……continued on the next page
RHEINDT ET AL.
414
·
Zootaxa 4250 (5) © 2017 Magnolia Press
Zootaxa 4250 (5) © 2017 Magnolia Press
·
415
NEW COLLOCALIA ESCULENTA COMPLEX
7$%/(   
,VODQG
RFFXUUHQFHDQG
WD[RQLGHQWLW\
'RUVDOWRQH
DQGJORVV
9HQWUDO
SDWWHUQ
&RQWUDVWLQJ
SDOHUXPS
:KLWHVSRWV
LQWDLOLQQHU
ZHEV
1DNHG
1RU
IHDWKHUH
G)
WDUVXV
0LFUR
IHDWKHUWXIW
RQKDOOX[

7D[RQRPLFDQGQRPHQFODWXUDO
FRPPHQWV
6SHFLHVDFFRUGLQJ
WRWKLVVWXG\
&HQWUDO
6RORPRQ
,VODQGV
&KRLVHXOWR
0DODLWDDQG
*XDGDOFDQDO
esculenta becki














nitens




esculenta /LQQDHXV

(DVW6RORPRQ
,VODQGV6DQ
&ULVWREDO
JURXS
esculenta
makirensis




















esculenta /LQQDHXV

(DVW6RORPRQ
,VODQGV
5HQQHOO
esculenta
desiderata

















esculenta
uropygialis


esculenta

esculenta /LQQDHXV

7HPRWX
9DQXDWX
uropygialis
uropygialis














esculenta




uropygialis *5
*UD\
1HZ&DOHGRQLD
*UDQGH7HUUH
ÌOHV/R\DXWp
uropygialis
albidior















uropygialis



uropygialis *5
*UD\
RHEINDT ET AL.
416
·
Zootaxa 4250 (5) © 2017 Magnolia Press
7$%/(  b nitens
et al.

troglodytescyanoptilalinchinatalissumbawae
6XPED
sumbawae
6XPEDZD
neglectamarginatabagobonitensbeckiesculenta
troglodytes ±   
cyanoptila  ±  
linchi   ±  
natalis      
sumbae
6XPED
      
sumbawae
6XPEDZD
     QD 
neglecta       QD 
marginata         
bagobo         ± 
nitens          ± 
becki           QD
esculenta            
Zootaxa 4250 (5) © 2017 Magnolia Press
·
417
NEW COLLOCALIA ESCULENTA COMPLEX
FIGURE 2. Phylogenetic tree topology based on Bayesian analysis of cyt b sequence data. Taxon names and localities are
given for each population. High branch support is given in Maximum Parsimony bootstrap, Maximum Likelihood bootstrap
and Bayesian posterior probability (multiplied by 100 for ease of reference) in the same order with a cut-off of ≥85 for
bootstrap and ≥99 for posterior probability.
Plumage states and biometrics from all regional and subspecific populations (Chantler 1999; Dickinson &
Remsen 2013) are summarized in Table 2. Information on the seven regional subspecies not examined has been
added from cited literature.
Intensity and tone of dorsal gloss and colour vary taxonomically, circumscribing two main groups west of
Stresemann’s Line and two in the east, as noted by previous authors (e.g. Gray 1866; Oberholser 1906; Stresemann
1925; Salomonsen 1983; Somadikarta 1986). The two in the west are a duller, greener linchi group and a richer
more brightly greenish-blue-glossed Sundaic cyanoptila and Philippine marginata group. The two in the east are
bluer, one variably dull-glossed through Nusa Tenggara and the other intensely dark blue-glossed from Sulawesi
east to the Solomon Islands. Differentiation in both gloss and tone is nuanced at different angles of incident light,
perhaps explaining discrepancies between our characterizations and those of Salomonsen (1983) and Somadikarta
(1986). Limitations in geographic coverage are another issue; whereas Salomonsen dealt mainly with populations
east of Stresemann’s Line, and Somadikarta only with those west of the Line, we compared dorsal differentiation
among populations on both sides using carefully formulated characterizations. Somadikarta (1986) also restricted
characterization exclusively to colour, whereas we found that variation in gloss was involved as well: there is a
general positive correlation between intensity of gloss and blue colour tone. Thus in Sundaic linchi-dodgei and
cyanoptila, dorsal differentiation is as much or more due to intensity of gloss, linchi-dodgei being distinctly duller
(Table 2).
In the text below, the entire ventral plumage is referred to as the “venter”. Lettered symbols for ventral pattern
in the table denote the following states:
A—dark: throat and upper breast plain dark, almost sooty grey, grading into coarse, deep greenish-grey chevrons
over lower breast and flanks, and spreading onto white belly, sometimes almost covering it;
B—bold: throat and upper breast deep metallic-grey with fine white scalloping, grading into coarse, medium
metallic-grey scallop-chevroning over lower breast and flanks and spreading onto sides of white belly;
C—light bicolored: throat and upper breast mid-deep grey with clear (if fine) white scalloping, rather sharply
demarcated from extensively white lower breast, flanks and belly;
RHEINDT ET AL.
418
·
Zootaxa 4250 (5) © 2017 Magnolia Press
D—light: throat and upper breast plain mid-grey or with fine whitish scalloping, grading in grey chevrons and
spots over lower breast and flanks, but leaving most of belly white;
E—muted: throat and upper breast diffuse mid-grey, feathers often edged whitish, grading into extensively white
lower breast, flanks and belly;
F—dark bicolored: throat and upper breast plain dark grey, rather sharply demarcated from extensively white lower
breast, flanks and belly.
Two further traits in Table 2—presence/absence of white tail spots and feathered hallux—were as
geographically disjunct and taxonomically informative as dorsal colour and gloss; but all traits showed degrees of
intrinsic variation and none were consistently correlated. Two other traits used conventionally to distinguish taxa
were not included in Table 2 because of intrinsic variation within local populations and distortion in specimen
preparation. They are, respectively, spicate white aftershafts in mantle down and pale whitish lore spot; the
development of white aftershafts in dorsal down could be age- or season-related. These characters, nevertheless,
consistently distinguish some adjacent island forms of eastern esculenta (see Discussion). We also found
considerable irregularity in the expression of a feather tuft on the hind toe, only some of which, from residual
microscopic shafts and epidermal marks, appeared to be due to damage in specimen preparation or preservation, or
moult. Since discovery by Somadikarta (1986), the presence of hind toe feather tufts in “esculenta” west of
Stresemann’s Line has been assumed to be a general condition for all populations of “esculenta” that separates
them from “‘linchi”’, in which it is absent. Allowing for moult and post-mortem loss, we found that these feather
tufts were normally present in western Sundaic “esculenta” (affinis-cyanoptila group) and Philippine populations,
but normally absent in “eastern esculenta” except for Moluccan and southeast Solomon Island populations and the
neglecta group in the east Lesser Sundas (Table 2).
Discussion
Species delimitation. The taxonomic classification generated here is provisional insofar as many more taxa—
some of them geographically remote and rarely collected—await molecular analysis, particularly those of the
Moluccas and Bismarck Archipelago to Vanuatu and New Caledonia. Our assessment is laid out in three main
groups that have long been apparent from both morphological (Stresemann 1912, 1940; Salomonsen 1983;
Somadikarta 1986) and molecular (Lee et al. 1996; Clayton & Johnson 2000; Price et al. 2004, 2005; Thomassen et
al. 2005; Moyle et al. 2008) data. They are the “western linchi”, “western white-bellied swiftlet (former
esculenta)” and “eastern esculenta” groups.
The “linchi” group. Identified morphologically by Somadikarta (1986) and corroborated by all previous
molecular studies (Lee et al. 1996; Clayton & Johnson 2000; Thomassen et al. 2003; Price et al. 2004, 2005;
Moyle et al. 2008), this group is confirmed by the present study, which adds Christmas Island natalis to it (Figure
2; see also below). It is characterized by dull greenish dorsal gloss, a light to muted ventral pattern, plain tail except
for a pale wash in the base of the rectrices (natalis excepted), a consistently naked hind toe, and small to medium
size. Divergence from other groups in the complex is 3.03–7.20% mtDNA distance (Table 3; Clayton & Johnson
2000). Geographically, this group ranges through the south Sundaic island arc from hilly Sumatra east to Lombok,
with outliers on Mt. Kinabalu (Borneo) and Christmas Island. On Sumatra and Borneo it is sympatric with western
white-bellied swiftlets (see below), justifying species separation (Somadikarta 1986) that is supported by mtDNA
sequence data in Moyle et al. (2008). Where the two overlap, western white-bellied swiftlets, as the form
cyanoptila, are widespread through both lowland and montane altitudes, and linchi and its outlier dodgei are
apparently centred in hilly and montane zones.
Collocalia dodgei. This is the outlier of the linchi group on Mt Kinabalu, Borneo, and neighbouring mountain
ranges. Moyle et al. (2008) raised it to species rank because of its small size, contrary to Bergmann’s Rule, and
depth of molecular divergence from linchi at 4.5% mtDNA distance. We can confirm its small size and that it is
more deeply satin sooty dorsally than linchi, supporting its separation as a monotypic species. Unfortunately, there
was no cyt b sequence based on non-blood tissue available for our study, as Moyle et al.'s (2008) sequences are
based on another gene.
Collocalia linchi ripleyi. There are no DNA sequences available for this form, which is currently confirmed to
occur only on the Pusat Ranges of north Sumatra and in hills of Talangpadang in south Sumatra (Somadikarta
Zootaxa 4250 (5) © 2017 Magnolia Press
·
419
NEW COLLOCALIA ESCULENTA COMPLEX
1986); nor have we seen specimens of it. Where Somadikarta (1986) found it in a mixed colony with C.
“esculenta” cyanoptila at Talangpadang, the material sampled had morphological traits uniformly consistent with
those of linchi without evidence of intergradation with “esculenta” cyanoptila. We retain ripleyi as a subspecies of
Collocalia linchi because its differentiation from the nominate form in shorter tail and heavier breast chevroning is
recorded as slight but constant (Somadikarta 1986).
Interaction, if any, between linchi and “esculenta” cyanoptila on adjacent peninsular Malaysia is less clear.
Wells (1999) reported “mixed colonies” of blue-glossed, green-glossed and intermediate birds in Selangor
(Malaysia) that were said to be uniform in allozymes, leading him to lump linchi and esculenta in one species. Hind
toe feathering was not recorded and the allozyme data are unavailable, leaving open the possibility, indeed
likelihood, that the “green” and intermediate birds at Selangor were age variants or worn specimens of “esculenta”
cyanoptila, not C. linchi. Although we have no access to the allozyme results, allozymes are much cruder in
detecting molecular differentiation than mtDNA sequences, and the allozyme results pointed out by Wells (1999)
stand in stark contrast to the deep mtDNA divergence recorded here and by Price et al. (2004, 2005) and Moyle et
al. (2008) between Sundaic C. “esculenta” cyanoptila and C. linchi/C. dodgei. Therefore we support treating linchi
and esculenta as different species.
The form ripleyi has not yet been included in any published DNA sequence-based analysis. Its close affinity
with linchi, as advocated here, rests entirely on morphological grounds. Future molecular analysis is required to
analyze its level of differentiation from nominate linchi from Java, or to confirm whether it is embedded within
linchi at all. Whether C. linchi (probably as the subspecies ripleyi) even extends to the Malaysian mainland is
moot. A single trade skin from “Malacca” has been identified as such by Somadikarta (1986) but its exact
provenance is questionable, given that C. l. ripleyi occurs in nearby Aceh.
Collocalia linchi linchi. This form, the nominate subspecies, is of medium size. It is recorded only from Java,
Bawean, Madura and Penida islands, with presumed sight records from the Kangean Islands (Somadikarta 1986).
That distribution is anomalous insofar as Penida lies in the Lombok Strait between Bali and Lombok which are
occupied by C. linchi dedii Somadikarka, 1986, effectively making the distributions of nominate linchi and dedii
parageographic with respect to one another. Nomenclature has obfuscated zoogeographic issues here, because
Meise (1941) named the Penida population as a third subspecies, plesseni, which has priority over dedii, if, as
geography suggests, the Penida population is found to belong to the Bali-Lombok form. Dickinson & Remsen
(2013: 99, footnote 5) responded to this by deleting the Penida population from range statements for the linchi
group. We have not seen material from Penida, whereas Somadikarta (1986) examined a series of 13, measuring
12, and found that “the colour of the plumage and the measurements of wing and tail are not sufficiently different
to separate plesseni from (nominate) linchi”. In these circumstances, we retain plesseni in synonymy under
nominate linchi pending better evidence for its relationships. For validation of the name linchi for this species, see
Somadikarta (1982).
Collocalia linchi dedii. This form is endemic to Bali and Lombok immediately east of Java, and is
distinguished from nominate linchi by darker throat and breast and larger size (Somadikarta 1986: Tables 1 and 2).
Unfortunately, no cyt b sequence derived from non-blood tissue was available for our study. Our morphological
review of dedii populations (n=18) found them to differ little from Javan linchi in ventral tone and pattern, and only
marginally in size (Table 2), indicating not only that linchi and dedii are conspecific but also perhaps
consubspecific. If so, the issue of the identity of plesseni on Penida Island (see above) becomes irrelevant. Because
of the more comprehensive series of dedii (n=45) examined by Somadikarta (1986), however, we refrain from
synonymizing it at this point.
Collocalia natalis. In morphology, natalis is markedly different from all other members of the white-bellied
swiftlet complex (Table 2), not only in patterns and tones of plumage but also body shape, traits that appear to be
genetically independent, adapted to a broadened ecological niche indicating deep biological differentiation. C.
natalis shares with linchi a naked hind toe and green cast to what dorsal gloss there is, but it is another step duller
than linchi and much the dullest member of Collocalia, with a mid-grey-green dorsum that is uniquely mottled
dull-grey over the back. Its rump is scalloped with white as well, as in Philippine marginata, and its tail has the
white spotting of eastern esculenta but of a distinctively elongate and diffuse form. However, the mitochondrial
sequence data suggests that natalis should be treated as a Christmas Island endemic subspecies of linchi. Its cyt b
sequence divergence from nominate linchi is shallow at 1.10–1.45 % (Table 3), and it is sister to linchi in our
phylogenetic tree (Figure 2). Taken at face value, the mtDNA data indicate that natalis is more closely related to
RHEINDT ET AL.
420
·
Zootaxa 4250 (5) © 2017 Magnolia Press
linchi, as suggested by Christidis & Boles (2008), than to esculenta, with which it was linked by all other revisers
since Oberholser (1906) and Stresemann (1940).
Overall, natalis is closest in appearance to the neglecta group in the east Lesser Sundas, probably due to
convergent plumage evolution, as our mtDNA phylogeny indicates that neglecta and natalis are quite distantly
related within Collocalia. Also significantly, natalis is slenderer than any other species of Collocalia, larger than
linchi, and has a uniquely long and more furcate tail (Table 2) and more pointed wing with primary 10 consistently
2–4 mm longer than primary 8. Such differences may seem small but they are critically important in hawking
insectivores, and may well reflect a wider and higher foraging zone in natalis, extending to open spaces outside and
above the forest canopy (Gibson-Hill 1947; Stokes 1988) in the absence of competitive species of Aerodramus
swiftlets. Two explanations may account for the mismatch in mtDNA similarity and morphological difference
between linchi and natalis. Either natalis has been subject to mitochondrial sweeps from occasionally invading
individuals of the nearest colonizing source, Javan linchi, in its recent history; or, through founder effects and
competitive release, it has diverged and adapted morphologically at an accelerated rate. Based on mtDNA
divergence, natalis may have separated from C. l. linchi ~0.5–0.7 million years ago (Weir & Schluter 2008; Figure
2), but short sequence length prevents us from placing too much confidence in this estimate. Because of its marked
morphological divergence from all other members of the complex, we treat natalis as an allospecies of linchi.
Classification of the “linchi” group. In summary, we propose the following taxonomic arrangement for the
linchi group, the species of which form a superspecies:
Collocalia dodgei Richmond, 1905—Mt. Kinabalu
Collocalia linchi
Collocalia linchi ripleyi Somadikarta, 1986—hilly Sumatra, perhaps peninsular Malaysia
Collocalia linchi linchi Horsfield & Moore, 1854—Java, Bawean, Madura, Kangean Islands; Nusa Penida
(subspecies?)
Collocalia linchi dedii Somadikarta, 1986—Bali, Lombok
Collocalia natalis Lister, 1889—Christmas Island
The “western white-bellied swiftlet” group
The significance of Stresemann’s Line. Except in the Lesser Sundas, Stresemann’s Line (Somadikarta 1986),
which separates the white-bellied swiftlets of the Greater Sundas and Philippines from those of Wallacea and
Papuasia (Figure 1), abruptly separates western plain-tailed forms from eastern spot-tailed forms that are often
combined in a single species, C. esculenta (Chantler 1999; Dickinson & Remsen 2013; del Hoyo & Collar 2014;
Gill & Donsker 2016). Previous molecular studies comparing populations on either side found the separation to be
phylogenetically deep, at ~5.8% mtDNA distance and greater (Price et al. 2004; Moyle et al. 2008), but sampling
was limited. Our study, with a larger representation of eastern populations, including samples from the Sula
Archipelago, Sumbawa, Sumba and Timor much closer to the Line, confirms the division at 3.29–8.77% mtDNA
distance (Figure 2; Table 3).
Within western white-bellied swiftlets, we recovered a particularly deep split between Sundaic and Philippine
populations, in concordance with other studies (Price et al. 2004; Moyle et al. 2008). Although the exact branching
pattern among the Philippine, western white-bellied and eastern esculenta groups is contested among publications,
there is overwhelming agreement that all three lineages are deeply diverged and on different evolutionary
trajectories.
Populations of white-bellied swiftlets on Borneo (in the west) and Maluku, Papua and Solomons (in the east)
straddle both sides of Stresemann's Line and were formerly classified as members of C. esculenta. Our study has
corroborated previous results (Price et al. 2004; Moyle et al. 2008) in showing a deep separation at 4.66–8.59%
mtDNA distance across the Line (Table 3). These western and eastern groups form distantly related clades,
separated by deep divergences (Figure 2; Table 3).
Morphologically, white-bellied swiftlets from east and west of Stresemann’s Line (Figure 1) also differ
markedly where they abut on the Line. Those in the west (Borneo, south Philippines) are moderately glossed
greenish-blue, dark-chested with heavy dark chevroning, usually feathered on the hind toes and plain-tailed with
just a pale wash in the base. Those immediately to the east (Sulawesi) are, in addition to their clearly spotted tails,
intensely glossed dark (purplish) blue, paler-chested and lack feathering on the hind toe.
Zootaxa 4250 (5) © 2017 Magnolia Press
·
421
NEW COLLOCALIA ESCULENTA COMPLEX
Unpublished findings from four individuals recently photographed in the hand on Ternate (North Maluku)
suggest that the local population there consists of both plain-tailed and spot-tailed individuals, a finding in stark
disagreement with previous specimen work (Lord Cranbrook, pers. comm.). This new insight may suggest the
presence of a cryptic form on Ternate or inflated variability in the amount of tail spotting in the local taxon spilura
(Gray, 1866). Specifically, it may indicate undocumented age- or sex-related variation, which is, however, unlikely
given that the caught individuals looked otherwise adult and that our specimen series from North Maluku numbers
24 spot-tailed individuals, making it improbable that all of them could be of one sex. While the new, unpublished
findings from Ternate suggest that the morphological division across Stresemann's Line may not be as clear-cut as
previously assumed, total evidence indicates that C. esculenta as currently construed comprises separate species
complexes west and east of Stresemann’s Line.
Species groups in western white-bellied swiftlets. Western white-bellied swiftlets range over the Greater
Sundas, from the Andaman Islands to Borneo, and extend through Palawan and the Sulu Archipelago to the rest of
the Philippines. Our cyt b sequences confirm two major divergences found in the group previously (Lee et al. 1996;
Clayton & Johnson 2000; Price et al. 2004, 2005; Moyle et al. 2008). The primary break is between Greater Sunda
(cyanoptila) and Philippine (marginata, bagobo) populations at 5.21–6.37% mtDNA distance, with a secondary
shallower split between marginata and bagobo at 2.74–3.06% mtDNA distance (Figure 2; Table 3; see below).
Apart from greyer rumps in some Philippine populations (marginata), morphological differences between Sundaic
and Philippine taxa are slight but consistent, involving size, dorsal tone (bluer in Greater Sundas, greener in
Philippines), breast pattern and frequency of feathering of the hind toe (Table 2). MtDNA distance and relative
morphological differences indicate completed speciation in which morphological divergence has been slow.
Differentiation in the Greater Sundas. Molecular and morphological variation among populations on the
main lands of the Sunda shelf—Sumatra, peninsular Malaysia and Borneo—is minimal (Figure 2; Tables 1, 3; also
Price et al. 2004; Moyle et al. 2008). Like earlier revisers (Stresemann 1912, 1925; Chantler 1999), we treat them
as consubspecific under the name cyanoptila. Their sympatry with C. linchi has been discussed above. Current
checklists (Chantler 1999; Dickinson & Remsen 2013; del Hoyo & Collar 2014; Gill & Donsker 2016) recognize
four other subspecies among Sunda populations, on island systems off the west coasts of the Thai-Malay Peninsula
and Sumatra. None has been sampled molecularly, and we have examined specimens of only one, affinis Beavan,
1867 from the Andaman–Nicobar chain (Table 2). The latter has the feather tuft on the hind toe akin to cyanoptila,
not linchi (Table 2), but is smaller, greener-backed and more fuscous-breasted, and is evidently at least
subspecifically differentiated from cyanoptila.
The status of the other three subspecies—elachyptera, vanderbilti and oberholseri—is less certain. Oberholser
(1906) described Mergui Archipelago elachyptera (n = 5) in the species linchi, which was then thought to comprise
the plain-tailed swiftlets west of Stresemann’s Line; but he placed it between similarly small-sized affinis and
Philippine isonota because of its greener-glossed back. From its original description, vanderbilti of Nias Island
resembles cyanoptila, and it is unclear if its describers, Meyer de Schauensee & Ripley (1940), compared material
of the two directly. Stresemann (1912) did not even describe neighbouring Mentawai Islands oberholseri from
specimens, using the notes of Oberholser (1906: 204–205) instead. Oberholser (1906) had confused his series of
large, cyanoptila-like but greenish-glossed specimens (n = 5) from North Pagai Island in the Mentawai group with
small, dull-backed nominate C. linchi from Java. He also considered a specimen from Singapore as “identical” with
the North Pagai series, raising further the likelihood that oberholseri falls within the range of variation in
cyanoptila. Although the status of these questionable subspecies should not be changed without due specimen
comparison, their geographic range and descriptions of dorsal gloss allow us, with reasonable confidence, to assign
them all to the species of western white-bellied swiftlets in Sundaland whose senior name is affinis Beavan, 1867.
Differentiation in the Philippines. Based on Dickinson (1989), populations in the Philippines have been
separated since into two allopatric and parapatric pairs of subspecies, one with clear white edgings to rump
feathers, the other uniformly dark-rumped. Several authors (Chantler 1999; Dickinson & Remsen 2013; del Hoyo
& Collar 2014; Gill & Donsker 2016) followed Dickinson (1989) in treating all four as subspecies of a broadly
defined C. esculenta. The forms of the two groups replace one another alternately north to south through the
Philippines, as follows: pale-rumped septentrionalisBabuyan Islands; dark-rumped isonota—northern, higher
altitude Luzon; pale-rumped marginata—central and southern Luzon, lowland Mindoro and all central and eastern
(Visayan) islands south to Negros, Cebu, Bohol and Leyte; and dark-rumped bagobo—Mindanao, the Sulu
Archipelago and montane Mindoro. The cyt b sequence data of Price et al. (2004, 2005) and Moyle et al. (2008)
RHEINDT ET AL.
422
·
Zootaxa 4250 (5) © 2017 Magnolia Press
confirm the relatively deep divergence between whitish-rumped marginata and dark-rumped bagobo that led
Clayton & Johnson (2000) to separate the two as species (Figure 2). This divergence, at 2.74–3.06% mtDNA
distance (Table 3; Clayton & Johnson 2000), is borderline for species delimitation. It is correlated, nevertheless, by
evidence of sympatry with limited intergradation on Mindoro and altitudinal replacement where the two co-occur
there and on Luzon: whitish-rumped forms occur at lower altitudes and dark-rumped forms in montane zones
(Delacour & Mayr 1945; Ripley & Rabor 1958; Dickinson 1989). The two taxa are disjunct in rump color in a
species complex in which morphological markers of speciation are few and cryptic. Thus the burden of proof is
shifted to those who would keep marginata and bagobo conspecific.
Despite this, we treat species separation of whitish- and dark-rumped forms as very much a working
hypothesis, for two reasons. First, there is evidence of more geographically widespread and intra-island variation in
rump color than has so far been reported (Table 2). In specimens seen by us from Sibuyan and Mindanao, which are
the only sources of all mtDNA material so far sequenced of marginata and bagobo, respectively, rumps were
extensively white in five but faint in two from Sibuyan, and dark in seven but with clear traces of white in three
from Mindanao. Dickinson (1989) also recorded intermediate states in birds from Bataan, Camiguin and Dinagat. It
suggests local swarms of polymorphic intergradients that challenge either speciation or rump patterning as a
speciation marker. Secondly, available mtDNA sequences are so far limited not just to two of the four forms
recognized in the Philippines, but to one island for each of those forms: Sibuyan and Mindanao. Much wider
sampling is needed to track the distribution of taxa in the Philippines and to test whether dark-rumped (isonota) and
whitish–rumped (septentrionalis) forms in the north are actually conspecific with bagobo and marginata,
respectively. That bagobo on Mindanao in the tropical south would be represented in the cooler north by
populations restricted to still cooler montane altitudes immediately raises questions. In the interim, nevertheless,
we accept isonota as a more extensively white-bellied isolate of bagobo in Luzon (this study; Dickinson 1989), and
septentrionalis as a more extensively white-ventered isolate of marginata in the Babuyan Islands and possibly
lowland north Luzon. The latter appears to be separated from nominate marginata in central and southern Luzon
today by the connecting spurs of the central cordillera and Sierra Madre between Lingayen Gulf and Baler Bay
which are occupied by isonota.
Classification of the western white-bellied swiftlet group. In summary, we propose the following taxonomic
arrangement for the western “esculenta” group. The species marginata and isonota appear to form a superspecies.
Collocalia affinis
Collocalia affinis affinis Beavan, 1867—Andaman, Nicobar Islands
Collocalia affinis elachyptera Oberholser, 1906—Mergui Archipelago, off Myanmar
Collocalia affinis vanderbilti Meyer de Schauensee & Ripley, 1940—Nias
Collocalia affinis oberholseri Stresemann, 1912—Batu and Mentawai Islands
Collocalia affinis cyanoptila Oberholser, 1906—Sumatra, Borneo, peninsular Malaysia, Natuna
Collocalia marginata
Collocalia marginata septentrionalis Mayr, 1945—Babuyan Islands, Philippines
Collocalia marginata marginata Salvadori, 1882—Central Luzon to Mindoro, Negros, Bohol and Leyte,
Philippines
Collocalia isonota
Collocalia isonota isonota Oberholser, 1906—Northern Luzon (montane only?), Philippines
Collocalia isonota bagobo Hachisuka, 1930—Mindanao, Sulu Archipelago, montane Mindoro,
Philippines
The hitherto unsequenced population on Palawan, with variable whitish to dark rump and long tail (Delacour
& Mayr 1945) and dark venter (Parkes 1960), is of unknown affinity. Possibly based on its conspicuous rump, it
has been associated by Dickinson et al. (1991) with Philippine bagobo rather than with Sundaic cyanoptila. At any
rate, it may represent a fifth, undescribed, form in the marginata-isonota group.
The “eastern esculenta group
Geographic range and species differentiation. Members of “eastern esculenta” are consistently spot-tailed
(except as detailed below), although the extent of spotting varies, and range from the Lesser Sundas and Sulawesi
Zootaxa 4250 (5) © 2017 Magnolia Press
·
423
NEW COLLOCALIA ESCULENTA COMPLEX
east through Wallacea and Papuasia to central Melanesia. Recent unpublished photographs of plain-tailed
individuals from Ternate Island (Lord Cranbrook, pers. comm.) are the only evidence of individuals lacking tail
spots within the range of the eastern esculenta group (see above). The eastern esculenta group was revised by
Salomonsen (1983) based on morphological analysis of c. 440 specimens from across its entire range. He
recognized 22 subspecies, eight of these newly described by him, in a single species, esculenta. Although our
mtDNA sampling lacks key taxa east of New Guinea, the combined mtDNA and morphological dataset identifies
three separate clusters of taxa within the group: (1) a dull-plumaged Lesser Sundas complex; (2) a complex of
richly glossed forms through the north Wallacean-Papuasian arc, from Sulawesi and the Moluccas to the Bismarck
Archipelago and Solomon Islands; and (3) a distinctively white-rumped cluster in central Melanesia (Santa Cruz,
Vanuatu and New Caledonia). The most significant result from our study is the discovery that the Lesser Sundas
complex comprises two primary lineages that are not only deeply diverged from one another at 4.13–4.40%
mtDNA distance, but also from all other white-bellied swiftlets at 2.91–7.71% (Table 3). Furthermore, the two
lineages, one (sumbawae) in the western Lesser Sundas and the other (neglecta) in the east, may respectively be
closer to the “linchi” and Philippine “western esculenta” groups at 3.03–3.86% and 3.29–4.93% mtDNA distance
than either is to New Guinean and Solomon Islands “eastern esculenta” at 4.93–7.42% distance (Table 3). In
morphology, sumbawae and neglecta differ from other members of “eastern esculenta” in generally duller tones
and from one another in degrees of dullness (Table 2); both also have traces of white feather edging on the rump,
often extending up onto the back. C. sumbawae is relatively dark, blue-glossed dorsally much as in Sundaic
western white-bellied swiftlets (C. affinis), but is paler and more finely chevron-flecked over the breast and usually
has a naked hind toe as in C. linchi (Table 2). Collocalia neglecta is a step duller and paler both dorsally and
ventrally, intermediate between C. linchi and C. natalis in dorsal tone but bluer than both (Table 2). Unlike
sumbawae, linchi and natalis, neglecta has a dull lore spot and usually a feather tuft on the hind toe. Correlated
mtDNA and morphological data thus indicate that both sumbawae and neglecta have speciated.
White-bellied swiftlets of the vast north Wallacean-Papuasian arc, from Sulawesi to the Solomon Islands
(Figure 1), stand out from all other members of the complex in the uniformity and brilliance of their deeply blue
glossed dorsa (Table 2). There is regional differentiation in the intensity and blue tone of the gloss, but it is never
quite matched in other species groups. Ventrally, most populations are also rather pale with type C and D breast
patterns (Table 2), with extensively white bellies and finer grey spotting over lower breast and flanks than the
coarser freckling in western white-bellied swiftlets. Although mtDNA data are limited, samples span almost the
entire geographic range of the group: the Sula Archipelago, New Guinea and the central Solomon Islands. They
show that the group is monophyletic with respect to Lesser Sundas, linchi and the western white-bellied swiftlet
groups, and deeply diverged from them at 4.93–8.59% mtDNA distance (Table 3). MtDNA distances among the
three regional samples are much shallower, but at 2.74–3.29% still approach those among other species (Rheindt et
al. 2014). Where the geographic boundaries for these prospective species lie in the island arc system and what their
morphological markers are, however, is unclear. The most distinctive morphological groups in the arc are the
variably-dark ventered, hallux-feathered populations in the North Moluccas, the white-rumped populations in the
north Bismarck Archipelago, and the populations in the Solomons with dark dorsal down, and on Makira Island
with unique foot-feathering. Yet hallux feathering is also present in duller-backed and more consistently whiter-
bellied South Moluccan populations, and there are gradients in the expression of white rump (Table 2) and dorsal
down feathering through the Bismarck Archipelago. None of these regional groups has yet been DNA-sampled. In
these circumstances, we provisionally retain all populations from Sulawesi to the Solomons in a single species,
pending molecular clarification of inter-island relationships and with the caveat that the group may yet be found to
comprise several species. The senior specific name for the group is esculenta Linnaeus, 1758.
No mtDNA data are available for the isolated group uropygialis of Santa Cruz, Vanuatu and New Caledonia. It
falls so far outside morphological limits among most other species (Table 2) that we separate it at species rank. Its
pale lore spot is obsolete in nearly all local populations, and its dorsum is deeply bluish-black, with a muted satin
sheen, not a gloss, indicating differences in keratin microstructure overlying melanin in the feather barbules.
Ventrally, sharp demarcation between the dark grey of the breast and white of the belly imparts a distinctively pied
appearance in Vanuatu and Santa Cruz populations, those closest to Solomon Islands esculenta. Across the rump,
furthermore, is a broad band of immaculate creamy-white; and tail spotting is reduced and dull. Because the
esculenta isolate on Rennell Island in the Solomons close to Santa Cruz and Vanuatu has broken white scalloping
in a rump that is otherwise dark as in the rest of the Solomons, Mayr (1931) named it desiderata: “the long-desired
RHEINDT ET AL.
424
·
Zootaxa 4250 (5) © 2017 Magnolia Press
missing link between esculenta and uropygialis”. It is on that basis that uropygialis has been combined
intraspecifically in esculenta ever since (Peters 1940; Sibley & Monroe 1990; Chantler 1999; Mayr & Diamond
2001; Dickinson & Remsen 2013; del Hoyo & Collar 2014; Gill & Donsker 2016). Although its white rump
scalloping could reflect historic connections and gene flow with uropygialis, the Rennell population otherwise
resembles Solomon Islands esculenta in its clear white lore spot, glossy dark blue dorsum and ventral pattern;
populations on nearby Makira, which is just as close to Santa Cruz and Vanuatu as Rennell, are even more
Solomons-like: white dorsal scalloping is confined to the sides of the rump, and often missing altogether.
Interaction between east Solomons esculenta and uropygialis today thus appears to be limited at best, as might be
expected from the wide marine barrier between them (Mayr & Diamond 2001: 221); molecular testing is needed.
Differentiation in the Lesser Sundas. Populations referred to the re-ranked western Lesser Sunda species
sumbawae occur on Sumbawa, Sumba and Flores (Rensch 1931; Mayr 1944a; Salomonsen 1983; Mees 2006).
They resemble one another closely (Table 2) except for occasional feathering of the hallux on Flores, and more
extensively white bellies and more extensive white edging to the feathering of rump and lower back in Sumba. The
population on Flores, of which we have seen no material, may be differentiated, if judged by its described greenish
gloss to the dorsum (Mees 2006). We could not confirm this, however, and Rensch (1931) downplayed the
difference, implicating juvenility. No mtDNA data are available from Flores, but populations on Sumbawa and
Sumba are rather deeply diverged at 2.9% mtDNA distance (Table 3). Sumba is a young, <2–3 mya emergent
outlying island of the Sunda shelf, separated from older Sumbawa on the Nusa Tenggara volcanic spur by a deep
water trench never closed during Pleistocene cycles of lowered sea-level (Hall 2002, 2009; van Oosterzee 2006).
Although the avifauna of Sumba is evidently derived from the Nusa Tenggara spur, a third of its terrestrial avifauna
has speciated or subspeciated endemically (Mayr 1944a). MtDNA and palaeogeographic data together indicate that
the Sumba enclave of sumbawae has been isolated since founding events around the early Pleistocene, and has at
least subspeciated. Accordingly we formally describe it as:
Collocalia sumbawae sumbae Schodde, Rheindt & Christidis, new subspecies
Holotype: AMNH 346710, adult male, collected by G. Stein at Langgabroe, Sumba, 29 May 1932.
Diagnosis: Differs from nominate C. s. sumbawae in more extensive off-white edging to feathers of rump and
lower back, more extensive white belly with reduced and more rounded dark grey freckling on lower breast and
flanks, and proportionally slightly shorter tail; also differs from other subspecies of the complex in the Lesser
Sundas in its richer and darker blue-glossed dorsum and consistently naked hind-toe. Size small: wing 88–95 mm.
Distribution: Sumba, Lesser Sunda Islands (Nusa Tenggara).
Etymology: The subspecies is named for its terra typica, Sumba.
Between C. sumbawae on Flores and C. neglecta on Timor and Roti in the eastern Lesser Sundas are
populations on Sawu, Alor, Wetar, Kisar, Romang and Damar that link the two species in morphology (Mayr
1944a; Salomonsen 1983; Mees 2006). These central populations have been presumed to link sumbawae and
neglecta phylogenetically (Mayr 1944a; Salomonsen 1983; White & Bruce 1986); they have also been grouped
together in their own subspecies perneglecta Mayr, 1944a, type locality Wetar. No mtDNA data is available, but
our morphological review queries their taxonomic integrity and relationships. Despite intermediacy between
sumbawae and neglecta in plumage tone and gloss, populations from Sawu through Damar are not uniform among
themselves (see also White & Bruce 1986). Those on Sawu, on the younger outer rim of the Banda arc between
Sumba and Timor (Hall 2002, 2009), are barely darker than nominate neglecta on Timor; we associate them with
that species. The series from the inner island arc, from Alor to Damar and including Wetar, is darker and glossier
dorsally, to the extent that it approaches nominate C. esculenta further east in the arc as much or more than
sumbawae in the west. This is clearest in the most easterly populations on Romang, Damar and Tanimbar which are
intergradient with nominate esculenta in appearance (Table 2). Evidence of trait flow between neglecta,
perneglecta and nominate esculenta rather than sumbawae may also be signalled by the presence of hallux
feathering, which is commonly present in the first three but absent in sumbawae, the Flores population excepted
(Table 2). Regional zoogeography also seems indicative. Most avifaunal affinities of the Alor to Damar section of
the inner Banda island arc, and of Wetar in particular, lie with Timor on the outer arc, not the Nusa Tenggara spur
of Flores, Sumbawa and Lombok to the west (Mayr 1944a). Yet mtDNA divergence between neglecta and
nominate esculenta (Sula Archipelago) is deep (Table 3), raising the possibility that intermediate morphological
states in geographically interposed perneglecta may not indicate gene flow but parallelism in ancestral states or
Zootaxa 4250 (5) © 2017 Magnolia Press
·
425
NEW COLLOCALIA ESCULENTA COMPLEX
convergence. Settling white-bellied swiftlet relationships and taxonomy in the Lesser Sundas and south Moluccas
will require a thorough molecular enquiry of populations throughout that region. Meanwhile, we conservatively
retain perneglecta, and treat it as a subspecies of neglecta.
Differentiation in the north Wallacean-Papuasian arc. We corroborate Salomonsen’s (1983) exhaustive
review of C. esculenta in this region, elaborating only on trends and discrepancies. Morphological traits for the
regional forms are given in Table 2. Populations on the Minahassa Peninsula of north Sulawesi (manadensis
Salomonsen, 1983) are disjunctly darker and more heavily marked ventrally and more deeply blue-glossed dorsally
than those from the remainder of Sulawesi. North Moluccan populations (spilura G.R. Gray, 1866) are similar,
differing in their reduced white tail spotting and usually feathered hind toes (Table 2), but not more extensively
dark grey-mottled ventra or bluer-glossed backs (pace Salomonsen 1983). Abutting in the presently colliding
Sangihe and Halmahera island arcs (Hall 2009), manadensis and spilura seem to be part of a chain of more
intensely dark-glossed forms that extends east through New Guinea and its northern offshore islands (nitens,
numforensis). Populations from Sulawesi (except the Minahassa Peninsula) through the Sula Archipelago and
south Moluccas to the Kai Islands and Tanimbar are not as extensively dark over the breast or as intensely glossed
on the dorsum, which has a distinct, if faint, green cast. They are currently combined in the nominate subspecies,
esculenta, but there are intrinsic differences: south Moluccan to Kai-Tanimbar populations have faint white edging
at the sides of the rump and variable feathering on the hind toe, as do those in the north Moluccas, whereas those on
Sulawesi and its fringing islands consistently lack both. It suggests more complex differentiation in north Wallacea
than has so far been appreciated but which cannot be resolved without molecular analysis. Trends from intensely
glossed, dark-ventered populations in north Sulawesi (manadensis) and north Moluccas (spilura) through the south
Moluccas (esculenta) and inner Banda arc (perneglecta) to dull-backed, pale-ventered populations on Timor and
Sawu (neglecta) match increasingly drier habitats north to south.
Outlying populations in the Flores Sea resemble nominate esculenta but are smaller (Table 2), justifying the
subspecies minuta. Salomonsen (1983) assigned populations from Obi in the central Moluccas to nominate
esculenta but our small series (n=3) matches darker and glossier spilura from Halmahera, consistent with the
tectonic origin of Obi in the Halmahera arc (Hall 2002, 2009). Salomonsen (1983) also referred Aru Islands
populations on the Australo-New Guinea continental platform to nominate esculenta, with the comment that, like
specimens from the nearby Kai Islands off the platform, they “are not quite typical esculenta, but tend towards the
New Guinea form nitens”. We did not record gloss or ventral pattern in Aru material but because avian taxa on the
Aru Islands are invariably New Guinean in affinity, not Moluccan, Salomonsen’s identification needs reevaluation.
Throughout mainland New Guinea there is limited variation in plumage: all populations have intensely blue-
glossed dorsa and moderately dark breasts contrasting with extensive white bellies, clearly marked tail spots and
usually no feather tuft on the hind toe. There is, nevertheless, marked variation in size with altitude (Table 2). This
led Mayr (1941) and Salomonsen (1983) to separate populations above c. 1600–2400 m altitude as the large
subspecies erwini from the smaller nitens below 1600 m. Salomonsen’s (1983) detailed analysis, nevertheless,
shows with clarity that the shift in size is smoothly clinal on both altitudinal and longitudinal gradients, with the
largest individuals occurring on the higher western end of the New Guinean cordillera, in accord with Bergmann’s
ecogeographical rule. It suggests adaptation to roost temperatures, not genetic differentiation, so we do not
recognize erwini here. Enclaves on several peripheral island groups have evidently differentiated: brilliantly
glossed, dark-ventered amethystina on Waigeo, more heavily mottle-breasted and greener-backed numforensis on
islands in Cenderawasih Bay, and slightly duller, greener-backed misima with faint white feather edging on the
sides of the rump on Trobriand and Woodlark Islands and in the Louisiade Archipelago. Whitening of feather
edging at the sides of the rump, clearest in Woodlark material, may reflect trait flow from New Britain. All island
populations are clearly spot-tailed, and none appear to have feathering on the hind toe. The taxonomic status of
populations on the D'Entrecasteaux Archipelago, which Salomonsen (1983) omitted from the range of this species
complex, is controversial.
Like New Guinean populations, those in the Bismarck Archipelago usually have extensive spicate white
aftershafts in the down of the mantle and back. Apart from inter-island differences in degrees of dark ventral
freckling (Table 2), they also show a stepped, southeast to northwest trend in whitening of the rump. It begins in
New Britain and Witu tametamele with whitened feather edging at rump sides and ends in a patchy band of white
across the rump in Admiralty Islands stresemanni. In between on New Ireland and New Hanover, local populations
are variably intermediate. Salomonsen (1983) interpreted dark-rumped individuals as resident and white-rumped
RHEINDT ET AL.
426
·
Zootaxa 4250 (5) © 2017 Magnolia Press
individuals (including the type of heinrothi) as nomads from the Admiralty Islands. This led him to introduce the
new name kalili for “resident” dark-rumped individuals. We consider, as implied by Hartert (1924), that New
Ireland and New Hanover populations are not discrete but intrinsically variable, some more whitish-rumped, others
less, across a narrow range more suggestive of local variation than active introgression with or invasion from
stresemanni. Accordingly, we use the prior name heinrothi for these ventrally rather heavily freckled populations.
In the Hibernian Islands immediately east of New Ireland are entirely dark-rumped populations with slightly duller
and more greenish-glossed dorsa; spicate aftershafts in the dorsal down are also often dulled to greyish.
Salomonsen (1983) split Hibernian enclaves into two subspecies, spilogaster on Lihir and Tabar and
hypogrammica on Nissan, but Mayr & Diamond (2001) lumped them. As the latter gave no reason for their
decision, and the dorsum in the Nissan series (Table 2) is marginally duller and greener, we retain hypogrammica
until evidence is produced to the contrary. The identity of enclaves on intervening Tanga and Feni islands has not
been convincingly established.
In the Solomons, there are at least four disjunct groups (also Mayr 1931; Salomonsen 1983; Mayr & Diamond
2001): (1) Buka, Bougainville and probably Shortland Islands (described below as a new taxon); (2) most of the
Solomon Islands between Bougainville and Makira (becki); (3) Makira and offshore islets (makirensis); and (4)
Rennell and Bellona Islands (desiderata). All resemble New Guinean nitens in dorsal gloss and ventral pattern, but
together differ in much-reduced to absent white aftershafts in mantle down; except in makirensis, hallux feathering
is all but absent (Table 2). Salomonsen (1983) recognized that Bougainville-Buka populations differed from those
in the remaining Solomons in having white patches on the sides of the rump, which led him to place them with
New Britain tametamele. Not only are the white side-rump patches in Bougainville and Buka larger than in
tametamele, but their mantle down is dark, as in other Solomons populations, and they are separated from
tametamele by hypogrammica on Nissan Island which has no white side-rump patches and is duller dorsally and
more extensively freckled ventrally. Zoogeographically, it is exceptional for Bougainville-Buka populations to be
allied with sister forms in the Bismarck Archipelago instead of the Solomons, yet not uncommon for them to be
differentiated from sister taxa within the Solomons (Mayr & Diamond 2001: 223, 240–242). Accordingly, we
interpret Bougainville-Buka populations as convergent on tametamele in white patching on rump sides, and
distinguish them as:
Collocalia esculenta lagonoleucos Schodde, Rheindt & Christidis, new subspecies
Holotype: ANWC 2993, adult female, collected by R. Schodde and Yaga Kanz at Konone, south foot of Mt.
Takuan massif, c. 760 m altitude, south Bougainville Island, 16 August 1964.
Diagnosis: Differs from geographically flanking C. e. tametamele (New Britain) in its usually all-dusky dorsal
under-down, finer and more rounded grey freckling over the flanks, more extensive white sides to the rump, and
larger size; from C. e. hypogrammica (Nissan Island) in its more vividly dark blue-glossed dorsum, with extensive
white sides to the rump, and larger size; and from C. e. becki (central Solomon Islands) in its more extensively
white sides to the rump, more extensively white belly and large size. Wing (100–)102–107(–109) mm (both sexes).
Distribution: Buka and Bougainville Islands, Solomon Islands. The population on adjacent Shortland Island
may belong to this subspecies, although Salomonsen (1983) and Mayr & Diamond (2001: 383) have referred it to
C. e. becki (see below).
Etymology: the epithet lagonoleucos is an arbitrarily formed word derived from the Greek noun λαγον,
meaning “flanks”, and the Greek adjective λευκος, meaning “white”, in reference to the distinctive white sides to
the rump of this subspecies. It is to be treated as a noun in apposition [Article 31.2.3 of the Code (ICZN 1999)].
Uniformly dark-rumped populations in the Solomons (becki) are found from Choiseul and Vella Lavella in the
northwest to Malaita and Guadalcanal in the southeast, but they are rather variable in the extent of dark ventral
freckling (Mayr 1931) and may comprise more than one taxon. On Makira and satellite islets, makirensis has been
distinguished by a darker back, more extensively white belly and patchy white feather edging at the sides of the
rump, all traits accentuated in the form desiderata on Rennell and Bellona in the extreme southeast of the
Solomons. Apparently unique to makirensis, however, and previously unreported, is a finely dusky-feathered tarsus
(n=9 of 9) and feathered hind toe (n= 7 of 9), which may reflect deep genetic divergence and speciation. C. e.
desiderata has a naked tarsus and hind toe (n=10).
Differentiation in central Melanesia. New Caledonian-Loyalty Island populations are differentiated from
those in Santa Cruz-Vanuatu in their much paler, freckled breasts and clearer tail spots, as described by Salomonsen
Zootaxa 4250 (5) © 2017 Magnolia Press
·
427
NEW COLLOCALIA ESCULENTA COMPLEX
(1983). Enclaves in the Banks Islands, northern Vanuatu (n=4), have rather clear white lore spots, approaching
those of C. esculenta, and may also prove to have differentiated.
Classification of the eastern esculenta group. In summary, we propose the following taxonomic arrangement
for the eastern esculenta group. C. esculenta itself may comprise several allospecies. We have already followed this
newly proposed treatment in a recent treatise on the birds of the Indonesian Archipelago (Eaton et al. 2016)
C. sumbawae
C. sumbawae sumbawae Stresemann, 1925—Sumbawa; Flores (subspecies?)
C. sumbawae sumbae Schodde, Rheindt & Christidis, this workSumba
C. neglecta
C. neglecta neglecta G.R. Gray, 1866—Sawu, Roti, Semau, Timor
C. neglecta perneglecta Mayr, 1944a—Alor, Wetar, Kisar, ?intergradient with nominate C. esculenta on
Romang, Damar and Tanimbar
C. esculenta
C. esculenta minuta Stresemann, 1925—Tanahjampea, Kalao
C. esculenta esculenta (Linnaeus, 1758)—S & C Sulawesi, Banggai and Sula Archipelagos, S Moluccas
(Buru, Seram, Ambon, and Seram Laut, Watubella, Kai Islands), Aru Islands (subspecies?)
C. esculenta manadensis Salomonsen, 1983—N Sulawesi (Minahassa Peninsula), Sangihe and Talaud
Archipelagos
C. esculenta spilura G.R. Gray, 1866—N Moluccas (Morotai, Halmahera and offshore islands, Bacan,
Obi)
C. esculenta amethystina Salomonsen, 1983—Waigeo
C. esculenta numforensis Salomonsen, 1983—Numfor, ?Biak Islands (Cenderawasih Bay, New Guinea)
C. esculenta nitens Ogilvie-Grant, 1914—Western Papuan Islands, New Guinea, Yapen and Karkar Islands
(includes C. e. erwini Collin & Hartert, 1927)
C. esculenta misimae Salomonsen, 1983—Trobriand and Woodlark Islands, Louisiade Archipelago,
D’Entrecasteaux Archipelago (subspecies?)
C. esculenta tametamele Stresemann, 1921—Long, Witu and Tolokiwa Islands, New Britain (Bismarck
Archipelago)
C. esculenta stresemanni Rothschild & Hartert, 1914—Admiralty Islands (Bismarck Archipelago)
C. esculenta heinrothi Neumann, 1919—New Hanover, Nusa, New Ireland, Djaul (Bismarck Archipelago)
C. esculenta spilogaster Salomonsen, 1983—Tabar, Lihir, Tanga, Feni (Hibernian Islands)
C. esculenta hypogrammica Salomonsen, 1983—Nissan (Hibernian Islands)
C. esculenta lagonoleucos Schodde, Rheindt, & Christidis, this work—Buka, Bougainville, Shortland
(?subspecies) (NW Solomon Islands)
C. esculenta becki Mayr, 1931—all Solomon Islands between, but excluding, Bougainville and Makira
C. esculenta makirensis Mayr, 1931—Makira Group (SE Solomon Islands)
C. esculenta desiderata Mayr, 1931—Rennell and Bellona Islands (SE Solomon Islands)
C. uropygialis
C. uropygialis uropygialis G.R. Gray, 1866—Utupua, Vanikoro (Santa Cruz Islands), Vanuatu
C. uropygialis albidior Salomonsen, 1983—New Caledonia, Loyalty Islands
English names for newly defined species taxa. We suggest the following English names for the species-level
groups newly distinguished here: Collocalia affinis—Plume-toed Swiftlet (named for its distinctive feathered
hallux setting it apart from C. linchi, with which it is sympatric); Collocalia natalis—Christmas Island Swiftlet
(named for the island to which it is restricted); Collocalia isonota—Ridgetop Swiftlet (named for its preference for
mountainous and hilly settings in areas where it co-occurs with sympatric C. marginata); Collocalia sumbawae
Tenggara Swiftlet (named for its occurrence in comparatively large areas of Nusa Tenggara, literally “the
southeastern archipelago”); Collocalia neglecta—Drab Swiftlet (named for its comparative lack of iridescence and
general drabness); Collocalia uropygialis—Satin Swiftlet (named for its exceptionally glossy plumage suggestive
of satin texture and hue). Other species of white-bellied swiftlets for which English names are already in common
use are Collocalia linchi (Linchi or Cave Swiftlet), C. dodgei (Kinabalu or Bornean Swiftlet), C. marginata (Gray-
rumped Swiftlet) and C. esculenta (Glossy Swiftlet) (Sibley & Monroe 1990; Moyle et al. 2008; Gill & Donsker
2016).
RHEINDT ET AL.
428
·
Zootaxa 4250 (5) © 2017 Magnolia Press
Correlates of differentiation. Morphological differences among species and subspecies in the white-bellied
swiftlet complex are slight and frequently cryptic, and similarities such as hallux feathering probably often
convergent given that they occur in such widely disjunct forms as C. affinis, C. neglecta and C. esculenta
makirensis. Morphological trends, such as north-south dulling and paling of plumage from the Moluccas to Timor
and stepped whitening of the rump from the Solomon Islands to Vanuatu, should be treated with caution as
indicators of gene flow until confirmed by molecular data.
Despite the great geographic range of the complex and its aerial capabilities, its individual taxa are not
consistent over-sea colonisers. That is obvious from the separation of species through the Philippines and Lesser
Sundas, and splitting of subspecies along the Wallacean-Papuasian arc. Neighbouring species in the Philippines
may be separated by as little as 20–50 km-wide stretches of water, with so far only localized evidence of
intergradation. In the Lesser Sundas, sea straits apparently separating C. sumbawae and C. neglecta on the inner
Banda arc are barely 10 km wide. Subspecies between Sulawesi and the Solomons are commonly separated by
water gaps about 80 km wide. Processes of allopatric localization appear to offer the simplest explanation for
resultant patterns of differentiation: small founder populations, roost- and nest-site philopatry, competitive
exclusion and periodic genetic bottlenecks during Pleistocene fluctuations have combined to generate divergence
between local populations and genetic integrity within them.
Speciation in the complex is centred zoogeographically in the region between the Philippines, Greater Sundas
and Wallacea, on either side of Wallace’s Line (Wallace 1860; Mayr 1944b; van Oosterzee 1997, 2006). Five of the
most deeply diverged species groups in the complex reach their eastern (linchi, affinis, and isonota) or western
(sumbawae, esculenta) limits on or near Wallace’s Line. Although the Sunda shelf fringing southeast Asia has been
a relatively stable land form since the early Mesozoic, the region to the north and east, from the Philippines through
Wallacea to the west rim of the Australo-Papuan plate, has been perhaps the most tectonically volatile on earth
through the Cenozoic. Rough and approximate dating of the divergence process in white-bellied swiftlets based on
Weir & Schluter’s (2008) mitochondrial clock rate for avian mtDNA coding genes indicates that diversification in
the complex commenced ~5–4 mya (Figure 2), coinciding with major tectonic rearrangements in the region that
have shaped the present-day geographic position of most regional landmasses (Hall 2002, 2009), underlining the
importance of earth-historic processes for biotic differentiation in the region.
Acknowledgments
FER acknowledges a Ministry of Education (Republic of Singapore) Tier I grant WBS R-154-000-658-112. LC
and JAN acknowledge funding from Parks Australia Northern Branch for DNA analysis. RS acknowledges several
Collections Grants from the American Museum of Natural History (AMNH) for work on its whole specimen
collections. RS and JAE thank the curatorial officers of the following institutions for making their specimens
available for study: AMNH, Australian Museum (AMS), Australian National Wildlife Collection (ANWC), Field
Museum of Natural History (FMNH), Museum of Natural Science at Louisiana State University (LSUMNS),
Museum Victoria (NMVM), Natural History Museum at University of Kansas (KUNHM), Natural History
Museum (BMNH), Naturalis Biodiversity Center (NBC), Lee Kong Chian Museum of Natural History at National
University of Singapore (LKCNHM) and Western Australian Museum (WAM). Walter Bock assisted RS in his
work at the AMNH and Arjan Brenkman helped JAE during his visit to the NBC. Pamela Rasmussen deserves
warm thanks for ample input on the text that helped improve the manuscript. We also acknowledge the helpful
feedback by two anonymous reviewers.
References
Beavan, R.C. (1867) The avifauna of the Andaman islands. Ibis, 9, 314–334.
https://doi.org/10.1111/j.1474-919X.1867.tb06434.x
Bensasson, D., Zhang, D., Hartl, D.L. & Hewitt, G.M. (2001) Mitochondrial pseudogenes: evolution's misplaced witnesses.
Trends in Ecology and Evolution, 16, 314–321.
https://doi.org/10.1016/S0169-5347(01)02151-6
Chantler, P. (1999) Family Apodidae (Swifts). In: del Hoyo, J., Elliott, A. & Sargatal, J. (Eds.), Handbook of the Birds of the
Zootaxa 4250 (5) © 2017 Magnolia Press
·
429
NEW COLLOCALIA ESCULENTA COMPLEX
World. Vol. 5. Lynx Edicions, Barcelona, pp. 338–457.
Christidis, L. & Boles, W. (2008) Systematics and taxonomy of Australian birds. Csiro Publishing, Melbourne, 288 pp.
Clayton, D.H. & Johnson, K.P. (2000) Testing species limits of non-echolocating Philippine swiftlets (Collocalia spp.) using
molecular genetic data. Sylvatrop, 10, 70–77.
Clements, J.F. (2007) The Clements Checklist of Birds of the World. 6
th
Edition. Cornell Laboratory of Ornithology, Ithaca, xix
+ 843 pp.
Clements, J.F., Schulenberg, T.S., Iliff, M.J., Roberson, D, Fredericks, T.A., Sullivan, B.L. & Wood, C.L. (2014) The eBird /
Clements Checklist of Birds of the World: Version 6.9. Available from: http://www.birds.cornell.edu/clementschecklist/
download/ (accessed 14 March 2017)
Collin, J.A. & Hartert, E.J.O. (1927) Collocalia esculenta erwini. Novitates Zoologicae, 34, 50.
Darlington, P.J. Jr. (1957) Zoogeography: The Geographical Distribution of Animals. Wiley, New York, xiv + 675 pp.
Delacour, J. & Mayr, E. (1945) Notes on the taxonomy of the birds of the Philippines. Zoologica, 30, 105–117.
del Hoyo, J. & Collar, N.J. (2014) HBW and BirdLife International Illustrated Checklist of the Birds of the World. Vol.1. Non-
passerines. Lynx Edicions, Barcelona, 903 pp.
Dickinson, E.C. (1989) A review of smaller Philippine swiftlets of the genus Collocalia. Forktail, 5, 23–34.
Dickinson, E.C. (Ed) (2003) The Howard and Moore Complete Checklist of the Birds of the World. 3rd Edition. Christopher
Helm, London, 1039 pp.
Dickinson, E.C., Kennedy, R.S. & Parkes, K.C. (1991) The Birds of the Philippines. British Ornithologists’ Union, Tring, 369
pp.
Dickinson, E.C. & Remsen, J.V. (Eds.) (2013) The Howard & Moore Complete Checklist of the Birds of the World. 4
th
Edition.
Vol. 1. Aves Press, Eastbourne, 461 pp.
Drummond, A.J. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary
Biology, 7, 214.
https://doi.org/10.1186/1471-2148-7-214
Eaton, J.A., van Balen, B., Brickle, N.W. & Rheindt, F.E. (2016) Birds of the Indonesian Archipelago. Lynx Edicions,
Barcelona, 496 pp.
Gibson-Hill, C.A. (1947) Field notes on the birds of Christmas Island. Bulletin of the Raffles Museum, 18, 87–165.
Gill, F. & Donsker, D. (2016) IOC World Bird List. Version 6.1. Available from: http://www.worldbirdnames.org/ioc-lists/
crossref/ (accessed 14 March 2017)
Gray, G.R. (1866) A synopsis of the species of the genus Collocalia, with descriptions of new species. The Annals and
Magazine of Natural History, Series 3, 17 (98), 118–128.
Hachisuka, M. (1930) Contributions to the birds of the Philippines. Ornithological Society Japan Supplement, 14, 141–222.
Hall, R. (2002) Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based
reconstructions, model and animations. Journal of Asian Earth Sciences, 20, 353–431.
https://doi.org/10.1016/S1367-9120(01)00069-4
Hall, R. (2009) Southeast Asia’s changing palaeogeography. Blumea, 54, 148–161.
https://doi.org/10.3767/000651909X475941
Hartert, E. (1924) The birds of New Hanover. Novitates Zoologicae, 31, 194–213.
Horsfield, T. & Moore, F. (1854) A Catalogue of the Birds in the Museum of the East-India Company. Vol. I. WM. H. Allen and
Co., London, 807pp.
Inskipp, T., Lindsey, N. & Duckworth, W. (1996) An Annotated Checklist of the Birds of the Oriental Region. Oriental Bird
Club, Sandy, Bedfordshire, 294 pp.
ICZN (International Commission on Zoological Nomenclature) (1999) International Code of Zoological Nomenclature. 4
th
Edition. The International Trust for Zoological Nomenclature, The Natural History Museum, London, xxix + 306 pp.
Lee, P.L.M., Clayton, D.H., Griffiths, R. & Page, R.D.M. (1996) Does behavior reflect phylogeny in swiftlets (Aves:
Apodidae)? A test using cytochrome b mitochondrial DNA sequences. Proceedings of the National Academy of Sciences,
USA, 93, 7091–7096.
https://doi.org/10.1073/pnas.93.14.7091
Linnaeus, C. (1758) Systema Naturae. Vol. 1. 10
th
Edition. Impensis Direct. Laurentii Salvii, Holmiae, 824 pp.
Lister, J.J. (1889) On the natural history of Christmas Island, in the Indian Ocean. Proceedings of the Zoological Society of
London, 1888, 512–531.
Mayr, E. (1931) Birds collected during the Whitney South Sea Expedition. XIII. A systematic list of the birds of Rennell Island
with descriptions of new species and subspecies. American Museum Novitates, 486, 1–29.
Mayr, E. (1941) List of New Guinea Birds. American Museum of Natural History, New York, xi + 260 pp.
Mayr, E. (1944a) The birds of Timor and Sumba. Bulletin of the American Museum of Natural History, 83, 127–195.
Mayr, E. (1944b) Wallace’s Line in the light of recent zoogeographic studies. The Quarterly Review of Biology, 19 (1), 1–14.
https://doi.org/10.1086/394684
Mayr, E. & Diamond, J. (2001) The Birds of Northern Melanesia: Speciation, Ecology, and Biogeography. Oxford University
Press, Oxford, xxiv + 492 pp.
Mees, G.F. (2006) The avifauna of Flores (Lesser Sunda Islands). Zoologische Mededelingen, Leiden, 80, 1–261.
Meise, W. (1941) Ueber die Vogelwelt von Noesa Penida bei Bali nach einer Sammlung von Baron Viktor von Plessen. Journal
RHEINDT ET AL.
430
·
Zootaxa 4250 (5) © 2017 Magnolia Press
für Ornithologie, 89, 345–376.
https://doi.org/10.1007/BF01905704
Meyer de Schauensee, R. & Ripley, S.D. (1940) Zoological results of the George Vanderbilt Sumatran expedition, 1936–1939.
Part I—Birds from Atjeh. Proceedings of the Academy of Natural Sciences of Philadelphia, 91, 311–398.
Moyle, R.G., Hosner, P.A., Nais, J., Lakim, M. & Sheldon, F.H. (2008) Taxonomic status of the Kinabalu linchi’ swiftlet.
Bulletin of the British Ornithologists' Club, 12 8 , 9 4 100.
Neumann, O. (1919) Über einige arten des genus Collocalia. Ornithologische Monatsberichte, 27, 108–110.
Norman, J.A., Christidis, L., Westerman, M. & Hill, F.A.R. (1998) Molecular data confirms the species status of the Christmas
Island Hawk-Owl Ninox natalis. Emu, 98, 197–208.
https://doi.org/10.1071/MU98028
Oberholser, H.C. (1906) A monograph of the genus Collocalia. Proceedings of the Academy of Natural Sciences of
Philadelphia, 58, 177–212.
Ogilvie-Grant, W.M. (1914) Collocalia nitens. Bulletin British Ornithological Club, 35, 35.
Päckert, M., Martens, J., Wink, M., Feigl, A. & Tietze, D.T. (2012) Molecular phylogeny of Old World swifts (Aves:
Apodiformes, Apodidae, Apus and Tachymarptis) based on mitochondrial and nuclear markers. Molecular Phylogenetics
and Evolution, 63, 606–616.
https://doi.org/10.1016/j.ympev.2012.02.002
Parkes, K.C. (1960) Notes on some non-passerine birds from the Philippines. Annals of the Carnegie Museum, 35, 331–340.
Peters, J.L. (1940) Check-list of Birds of the World. Vol. IV. Harvard University Press, Cambridge, Massachusetts, xii + 291 pp.
Posada, D. (2008) jModelTest: phylogenetic model averaging. Molecular Biology and Evolution, 25, 1253–1256.
https://doi.org/10.1093/molbev/msn083
Price, J.J., Johnson, K.P. & Clayton, D.H. (2004) The evolution of echolocation in swiftlets. Journal of Avian Biology, 35, 135–
143.
https://doi.org/10.1111/j.0908-8857.2004.03182.x
Price, J.J., Johnson, K.P., Bush, S.E. & Clayton, D.H. (2005) Phylogenetic relationships of the Papuan Swiftlet Aerodramus
papuensis and implications for the evolution of echolocation. Ibis, 147, 790–796.
https://doi.org/10.1111/j.1474-919X.2005.00467.x
Rambaut, A. & Drummond, A.J. (2008) Tracer; MCMC Trace Analysis Tool.Version 1.4.1. Available from: http://
beast.bio.ed.ac.uk/Tracer (accessed 14 March 2017)
Rensch, B. (1931) Die Vogelwelt von Lombok, Sumbawa und Flores. Mitteilungen aus dem Zoologischen Museum, Berlin, 17,
451–637.
Rheindt, F.E., Norman, J.A. & Christidis, L. (2014) Extensive diversification across islands in the echolocating Aerodramus
swiftlets. Raffles Bulletin of Zoology, 62, 89–99.
Richmond, C.W. (1905) Description of a new swiftlet from Mount Kina Balu, Borneo. Smithsonian Miscellaneous Collections,
47, 431–432.
Ripley, S.D. & Rabor, D.S. (1958) Notes on a collection of birds from Mindoro Island, Philippines. Peabody Museum of
Natural History Yale University Bulletin, 13, 1–13.
Rothschild, W. & Hartert, E.J.O. (1914) Collocalia esculenta stresemanni. Novitates Zoologicae, 21, 239.
https://doi.org/10.5962/bhl.part.16328
Salomonsen, F. (1983) Revision of the Melanesian swiftlets (Apodes, Aves) and their conspecific forms in the Indo-Australian
and Polynesian region. Kongelige Danske Videnskabernes Selskab Biologiske Skrifter, 23 (5), 1–112.
Salvadori, T. (1882) Ornitologia della Papuasia e delle Molucche. Stamperia reale G.B. Paravia e co. di I. Vigliardi, Torino,
595 pp.
Sibley, C.G. & Monroe, B.L. (1990) Distribution and Taxonomy of Birds of the World. Yale University Press, New Haven, 1111
pp.
Somadikarta, S. (1982) The White-bellied Swiftlet Collocalia esculenta from Java. Bulletin of the British Ornithologists’ Club,
102, 18–20.
Somadikarta, S. (1986) Collocalia linchi Horsfield and Moore—a revision. Bulletin of the British Ornithologists' Club, 106,
32–40.
Sorenson, M.D. & Quinn, T.W. (1998) Numts: a challenge for avian systematics and population biology. Auk, 115, 214–221.
https://doi.org/10.2307/4089130
Stokes, T. (1988) A review of the birds of Christmas Island, Indian Ocean. Australian National Parks and Wildlife Service
Occasional Paper, 16, i–viii + 1–39.
Stresemann, E. (1912) Ornithologische Miszellen aus dem indo-australischen Gebiet. XVII. Die Formen von Collocalia linchi
Horsf. & Moore. Novitates Zoologicae, 19, 347–349.
https://doi.org/10.5962/bhl.part.1551
Stresemann, E. (1921) Collocalia esculenta tametamele. Anzeiger der Ornithologischen Gesellschaft in Bayern, 1 (5), 37.
Stresemann, E. (1925) Beiträge zur Ornithologie der indo-australischen Region, II. Bruchstücke einer Revision der Salanganen
(Collocalia). Mitteilungen aus dem Zoologischen Museum, Berlin, 12, 349–354.
Stresemann, E. (1940) Die Vögel von Celebes. Teil III. Systematik und Biologie. Journal für Ornithologie, 88, 389–487.
https://doi.org/10.1007/BF01670259
Zootaxa 4250 (5) © 2017 Magnolia Press
·
431
NEW COLLOCALIA ESCULENTA COMPLEX
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. & Kumar, S. (2011) MEGA5: Molecular evolutionary genetics
analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and
Evolution, 28, 2731–2739.
https://doi.org/10.1093/molbev/msr121
Thomassen, H.A., Wiersema, A.T., de Bakker, M.A.G., de Knijff, P., Hetebrij, E. & Povel, G.D.E. (2003) A new phylogeny of
swiftlets (Aves: Apodidae) based on cytochrome-b DNA. Molecular Phylogenetics and Evolution, 29, 86–93.
https://doi.org/10.1016/S1055-7903(03)00066-6
Thomassen, H.A., den Tex, R.-J., de Bakker, M.A.G. & Povel, G.D.E. (2005) Phylogenetic relationships amongst swifts and
swiftlets: A multi locus approach. Molecular Phylogenetics and Evolution, 37, 264–277.
https://doi.org/10.1016/j.ympev.2005.05.010
The Times (2011) Comprehensive Atlas of the World. 13
th
Edition. HarperCollins, London, 544 pp.
van Oosterzee, P. (1997) Where Worlds Collide: The Wallace Line. Reed Books, Sydney, 234 pp.
van Oosterzee, P. (2006) Drawing the Wallace Line. In: Merrick, J.R., Archer, M., Hickey, G.M. & Lee, M.S.Y. (Eds.),
Evolution and Biogeography of Australasian Vertebrates. Auscipub, Oatlands, New South Wales, pp. 95–110.
Wallace, A.R. (1860) On the zoological geography of the Malay Archipelago. Journal of the Proceedings of the Linnean
Society, London, Zoology, 4, 172–174.
https://doi.org/10.1111/j.1096-3642.1860.tb00090.x
Wallace, A.R. (1864) On the identification of the Hirundo esculenta of Linnaeus, with a synopsis of the described species of
Collocalia. Proceedings of the Zoological Society of London, 1863, 382–385.
Weir, J.T. & Schluter, D. (2008) Calibrating the avian molecular clock. Molecular Ecology, 17, 2321–2328.
https://doi.org/10.1111/j.1365-294X.2008.03742.x
Wells, D.R. (1999) The Birds of the Thai-Malay Peninsula. Academic Press, London, 648 pp.
White, C.T & Bruce, M.D. (1986) The Birds of Wallacea (Sulawesi, The Moluccas & Lesser Sunda Islands, Indonesia). An
Annotated Check-list. B.O.U. Check-list No. 7. British Ornithologists’ Union, London, 524 pp.
APPENDIX I. Morphological data (mean, standard deviation and sample size) of the forms of the white-bellied swiftlet
complex across its range, west to east, north to south, in zoogeographic groupings. In the column for tail furcation, t1 =
innermost pair of rectrices, and t5 = outermost.
Species and locality Wing size Proportional tail length
(tail/wing ratio)
Tail furcation (t1/t5)
affinis (Andaman-Nicobar Is.) 96.00 ± 2.83 (n=4) - -
affinis (Sumatra) 98.27 ± 2.93 (n=15) 0.38 ± 0.017 (n=11) 0.92 ± 0.013 (n=5)
affinis (Peninsula Malaysia) 103.67 ± 2.05 (n=6) 0.37 ± 0.010 (n=6) -
affinis (Borneo) 99.53 ± 5.06 (n=19 0.37 ± 0.014 (n=19) 0.93 ± 0.030 (n=14)
linchi (Java, Penida I.) 96.08 ± 1.73 (n=13) 0.39 ± 0.020 (n=13) 0.93 ± 0.022 (n=11)
linchi (Bali, Lombok) 97.33 ± 1.94 (n=9) 0.43 ± 0.010 (n=8) 0.92 ± 0.019 (n=5)
natalis (Christmas I.) 100.10 ± 2.40 (n=42) 0.44 ± 0.015 (n=41) 0.91 ± 0.020 (n=15)
dodgei (Borneo; Mt. Kinabalu) 72.00 ± 1.00 (n=2) 0.37 ± 0.015 (n=2) 0.94 ± 0.020 (n=2)
isonota (Philippines; N. Luzon) 100.80 ± 3.87 (n=5) 0.36 ± 0.000 (n=2) 0.95 ± 0.025 (n=2)
marginata (Philippines; mid-Luzon to
mid-Philippines, ?Palawan)
97.82 ± 5.55 (n=33) 0.38 ± 0.019 (n=29) 0.93 ± 0.035 (n=19)
isonota (Philippines; Mindanao, Mindoro,
Sulu Arch.)
98.58 ± 3.20 (n=12) 0.38 ± 0.020 (n=12) -
sumbawae (W. Lesser Sundas; Sumbawa,
Flores)
92.00 ± 3.26 (n=10) 0.43 ± 0.017 (n=10) 0.94 ± 0.016 (n=10)
sumbawae (Sumba) 91.38 ± 2.90 (n=13) 0.42 ± 0.015 (n=13) 0.93 ± 0.019 (n=8)
neglecta (E. Lesser Sundas; Sawu, Roti,
Timor)
95.82 ± 3.82 (n=17) 0.41 ± 0.011 (n=10) 0.92 ± 0.000 (n=1)
......continued on the next page
RHEINDT ET AL.
432
·
Zootaxa 4250 (5) © 2017 Magnolia Press
APPENDIX 1. (Continued)
Species and locality Wing size Proportional tail length
(tail/wing ratio)
Tail furcation (t1/t5)
neglecta (E. Lesser Sundas; Alor, Wetar,
Kisar)
91.36 ± 2.64 (n=11) 0.42 ± 0.014 (n=8) -
neglecta-esculenta? (S. Banda Sea Is.;
Romang, Damar, Tanimbar)
96.00 ± 4.20 (n=11) 0.41 ± 0.019 (n=6) -
esculenta (N. Sulawesi, Talaud, Sangihe
Is.)
98.42 ± 2.87 (n=12) 0.41 ± 0.013 (n=5) -
esculenta (C. and S. Sulawesi) 99.00 ± 2.05 (n=9) 0.41 ± 0.012 (n=9) -
esculenta (Flores Sea Is.; Tanahjampea,
Kalao)
93.33 ± 0.47 (n=3) 0.43 ± 0.014 (n=3) -
esculenta (Banggai, Sula Is.) 90.00 ± 2.00 (n=2) 0.42 ± 0.005 (n=2) -
esculenta (South Moluccas to Kai & Aru
Is.)
99.78 ± 2.90 (n=9) 0.38 ± 0.016 (n=8) -
esculenta (North Moluccas) 95.00 ± 1.56 (n=9) 0.41 ± 0.027 (n=4) -
esculenta (W. Papuan Is.; Waigeo) 93.50 ± 2.50 (n=2) - -
esculenta (Cenderawasih Bay Is.) 97.50 ± 1.71 (n=6) - -
esculenta (New Guinea, satellite is., Cape
York, altitudes below 2000 m)
103.40 ± 3.21 (n=108) 0.39 ± 0.017 (n=107) 0.92 ± 0.023 (n=75)
esculenta (Louisiade, Trobriand Arch.) 97.10 ± 3.14 (n=10) 0.42 ± 0.015 (n=9) -
esculenta (Bismarck Arch.; Admiralty Is.) 99.89 ± 1.66 (n=9) 0.4 ± 0.014 (n=9) -
esculenta (E. Solomon Is.; Makira) 93.89 ± 1.85 (n=9) 0.41 ± 0.009 (n=7) -
esculenta (E. Solomon Is.; Rennell) 98.40 ± 3.16 (n=15) 0.41 ± 0.018 (n=10) -
uropygialis (Santa Cruz; Vanuatu) 96.87 ± 3.22 (n=15) 0.42 ± 0.017 (n=15) 0.94 ± 0.010 (n=2)
uropygialis (New Caledonia, Loyalty Is.) 97.50 ± 3.04 (n=10) 0.4 ± 0.011 (n=10) -
Zootaxa 4250 (5) © 2017 Magnolia Press
·
433
NEW COLLOCALIA ESCULENTA COMPLEX
... 'Spot-tailed' populations, from Sulawesi east through and beyond New Guinea, were believed by Wallace (1864), and thereafter by Stresemann (1940) and Somadikarta (1982Somadikarta ( , 1986, to be differentiated by a white spot on the concealed inner vane of all but the central pair of rectrices. In a review of speciation in the C. esculenta complex integrating phenotypic and molecular data, Rheindt et al. (2017) broadly confirmed Wallace's ideas, with some revision of species boundaries within these two groups. Among the plain-tailed group of white-bellied swiftlets a second indicator is variation in the green or blue tone to the glossy upperparts. ...
... Within the western plaintailed white-bellied swiftlets, excluding those populations in the Philippines, two species groups have been separated by this mix of characters (Somadikarta 1982(Somadikarta , 1986. One consists of the species C. a. affinis (sensu Rheindt et al. 2017) on the Andamans, Nicobar and other islands, and C. a. cyanoptila from Sumatra, Peninsular Malaysia and Borneo. Within this group, members of three geographically separated colonies of C. a. cyanoptila in Selangor, Peninsular Malaysia, exhibited individual age-and wear-related change, from the greenish gloss of fresh plumage to blue gloss of old, worn plumage. ...
... There is more than one instance of sympatry involving the C. linchi superspecies and C. affinis. In Sabah, Malaysia, the endemic C. dodgei, a member of the linchi superspecies (Rheindt et al. 2017), overlaps in daily activity range with the more numerous local population of C. affinis cyanoptila (Cranbrook et al. 2005, Moyle et al. 2008). On Sumatra, Somadikarta (1986) reported overlapping ranges of C. affinis cyanoptila and C. linchi ripleyi, with a mixed colony in a cave at Talangpadang, South Lampung. ...
Article
Full-text available
Summary.—Among white-bellied glossy swiftlets of the Collocalia group, A. R. Wallace was first to recognise the Makassar Strait, separating Borneo and Sulawesi, as a geographical barrier between different phenotypes: plain-tailed to the west and spot-tailed to the east. Other morphological characters used to define species within the group have been blue or green gloss to the dorsal plumage, and the presence or absence of a single minute tufted feather on the hallux. The value of these characters as taxonomic markers is now known to be unreliable due to the discovery of phenotypically mixed populations east of the Makassar Strait, from North Maluku province, Indonesia, through Papua New Guinea to New Ireland. We combine field observations of plumage characters with genetic evidence to establish taxonomy of Collocalia group swiftlets. Sequencing specific mitochondrial genes (Cytb and ND2), the nuclear-encoded Fib gene, and a subset of mitochondrial genomes provided data for phylogenetic analysis. Genetic divergence of c.4.7% is observed between two Collocalia clades either side of the Makassar Strait: the plain-tailed C. affinis cyanoptila sampled at Fraser’s Hill, Peninsular Malaysia, and a phenotypically mixed population of C. esculenta spilura from North Maluku, Indonesia. Each population formed high-affinity genetic clades, within which divergence was <0.5%. These findings are consistent with geographic but not phenotypic separation between populations. We therefore conclude taxonomy based on these plumage features in glossy swiftlets of the Collocaliini is unreliable.
... Taxonomic insights on Asian birds published in 2020 Collocalia swiftlets In a paper summarised in one of our previous reviews (BirdingASIA 30: 33), Rheindt et al. (2017) took on the seemingly impossible task of making sense of the wide variation in the white-bellied swiftlets in the C. esculenta complex, using extensive genetic, morphometric and plumage character data. They concluded that the complex comprised no fewer than ten different species between the Andamans and Nicobars in the west to New Caledonia and Vanuatu in the east. ...
... Within this broad division, different populations are characterised by having either green-or blue-glossed plumage and by the presence or absence of a single small plumelet at the base of the hind toe. Using both mitochondrial and nuclear markers, Davies et al. (2020) confirm several of the splits proposed by Rheindt et al. (2017) but show that there is a high degree of phenotypic variation within populations. For example, a minority of birds in populations east of Stresemann's Line have plain tails, and others have very reduced tail-spotting. ...
... In the meantime, the results of a detailed analysis of the taxonomy of Collocalia swiftlets in the Indo-Pacific region (Rheindt et al. 2017) led to the proposal of radical changes to the established taxonomy; this was a big challenge due to the morphological uniformity of these taxa. Rheindt et al. (2017) studied the evolutionary history of the complex, combining new biometric measurements and plumage assessment of museum specimens with novel as well as previously published molecular data, with a total of 809 individuals representative of 32 taxa being assessed. ...
... In the meantime, the results of a detailed analysis of the taxonomy of Collocalia swiftlets in the Indo-Pacific region (Rheindt et al. 2017) led to the proposal of radical changes to the established taxonomy; this was a big challenge due to the morphological uniformity of these taxa. Rheindt et al. (2017) studied the evolutionary history of the complex, combining new biometric measurements and plumage assessment of museum specimens with novel as well as previously published molecular data, with a total of 809 individuals representative of 32 taxa being assessed. The authors propose changing the classification of white-bellied swiftlets, for which just two species are currently recognised, to elevate eight regional forms to species level, as well as describing two new subspecies. ...
... one or more species, some taxonomy committees refuse to acknowledge their validity until ad hoc studies assembling different sources of supporting information and/or producing pertinent genetic evidence are carried out (e.g. Saitoh et al. 2015;Ng et al. 2017;Rheindt et al. 2017;Lim et al. 2019;Tyler et al. 2020). ...
... one or more species, some taxonomy committees refuse to acknowledge their validity until ad hoc studies assembling different sources of supporting information and/or producing pertinent genetic evidence are carried out (e.g. Saitoh et al. 2015;Ng et al. 2017;Rheindt et al. 2017;Lim et al. 2019;Tyler et al. 2020). ...
Article
Over the last two decades, vocal and genetic data have been extensively used in avian studies addressing taxonomic and systematic issues. However, even when multiple lines of evidence lean toward lumping or splitting of species, some taxonomic committees refuse to acknowledge their validity until convincing genetic evidence is produced and integrated with other sources of data. As a case in point, the genus Pseudocolaptes (Furnariidae) consists of three mostly allopatric and overtly distinct neotropical taxa differing in plumage and vocal features: the Streaked Tuftedcheek (P. boissonneautii), the Buffy Tuftedcheek (P. lawrencii), and the contentious Pacific Tuftedcheek (P. johnsoni). In this study, we assemble already available mitochondrial and nuclear DNA sequences to assess their taxonomy and to provide appointed committees with specific proof to ascertain the number of Pseudocolaptes species. Phylogenetic inference and species delimitation analysis indicate three species equally divergent from each other. We hope that this work will contribute to resolve the long-lasting taxonomical confusion regarding this genus, and that proper future conservation plans will be adopted to preserve each species within the inhabited biodiversity hotspot.
... Third, there are instances of clear contradiction between his publications. Fourth, some people who (1913), Mertens (1930), Rensch (1930), Raven (1935), Stresemann (1939), Scrivenor et al. (1943) and Parenti & Ebach (2010). have reviewed his writings have represented him incorrectly and this may have impacted the subsequent literature. ...
Article
Due to its position between the highly distinct Oriental and Australasian biogeographical realms, much effort has been spent demarcating associated separations and transitions in the faunal assemblages of the Indo‐Australian Archipelago. Initially, sharp boundary lines were proposed, with the earliest dating from the mid‐1800s. Notably, the one published by Alfred R. Wallace in 1863, based upon land‐mammal and land‐bird distributions, has since achieved iconic status and today its significance is recognized well beyond the confines of the biogeography community. Over the next four decades many such divides were engraved onto plates and inked onto charts of SE Asia using additional information, different organisms or other criteria. However, it became apparent that, as Wallace had noted, all such lines were to some degree permeable, and by the 1880s transition zones were being put forward instead; the label ‘Wallacea’ was introduced in 1924. Interestingly, the last decade has seen new divides and sub‐regions being proposed, some departing markedly from earlier offerings. Although currently there is general agreement regarding much of the terminology associated with both the lines and the areas, the record of publication indicates that this consensus has emerged obliquely, and in some cases is weakly founded. This review does not present new data nor new analyses; rather it summarizes the development of ideas and reflects upon attendant issues that have emerged. After reviewing the key proposals, recommendations are presented that should in future alleviate perceived difficulties or inadequacies. Reference to specific divides must be true to their original definitions; there are many instances where the secondary literature has portrayed them incorrectly and with some this has rippled through into later publications. Moreover, Wallace's 1863 line is not the one that he finally settled upon (in 1910); its path around Sulawesi was transferred from the west to the east of the Island. Ideally, Huxley's divide (1868) should carry his name rather than Wallace's; the latter never accepted the proposition. Lydekker's Line (1896) ought to be labelled the Heilprin–Lydekker Line in recognition of Angelo Heilprin's 1887 contribution. Concerning transition zones, ideally Wallacea should correspond to its original 1924 description, which incorporated the Philippine islands bar the Palawan group. Notably, though, a smaller form (introduced by Darlington in 1957, used frequently from 1998 onwards) in which all of the Philippine islands are excluded is entrenched within the recent literature, but this is often without evident justification. It should also be recognized that the ‘reduced’ (=southern) Wallacea area was effectively defined by Heilprin in 1887, but was then labelled the ‘Austro‐Malaysian Transition Zone’. Finally, the application in recent years of modern analytical techniques has not led to a consensus view on where the lines/areas should run/be placed; with such a large, diverse set of organisms, each with differing histories, this is perhaps not surprising.
... Concerning the endemic species and where they dispersed from relative to Wallace's Line, from the west there is one, the swiftlet Collocalia natalis Lister, 1889. According to Rheindt et al. (2017), it is sister to Collocalia linchi linchi Horsfield & Moore, 1854 (Fig. 3), which occupies Java and some islands in the eastern Java Sea (IUCN Red List, 2020). Norman et al. (1998) indicates that the hawk-owl Ninox natalis Lister, 1888 came from the east, because it is most closely related to Ninox squamipila (Bonaparte, 1850) on Seram and nearby islands, and to Ninox hypogramma (Gray, 1861) on Halmahera. ...
Article
Based on a comprehensive literature survey, we determined the sources of the terrestrial vertebrate species on Christmas Island, asking where they originated relative to Wallace’s Line (the southern end of the divide lies 1100 km to the east, where the Lombok Strait adjoins the eastern Indian Ocean). The two bats, Pipistrellus murrayi and Pteropus natalis, are from the west. Concerning the endemic and ‘resident’ bird species, one is from the west (Collocalia natalis), four are from the east (Accipiter fasciatus, Egretta novaehollandiae, Falco cenchroides and Ninox natalis) and the other 15 are ambiguous or indeterminate. Most of the land-locked species are also from the east: rodents Rattus macleari and Rattus nativitatis, and squamates Cryptoblepharus egeriae, Emoia nativitatis and Lepidodactylus listeria. Additionally, two have westerly origins (Crocidura trichura and Cyrtodactylus sadleiri), one is ambiguous (Emoia atrocostata) and another is unknown (Ramphotyphlops exocoeti). West-directed surface currents that flow across the eastern Indian Ocean towards Christmas Island would have facilitated most of the land-animal colonizations. We therefore suggest that Wallace’s Line be redrawn such that the landmass is placed on the Australasian side of this fundamental biogeographical boundary.
... Plume-toed Swiftlet Collocalia affinis vanderbilti: First recorded on Nias by von Rosenberg (1878). We follow DNA evidence by Rheindt et al. (2017) in separating Plume-toed Swiftlet (as here circumscribed) from the more easterly Glossy Swiftlet Collocalia esculenta. Nias hosts an endemic subspecies of Plume-toed Swiftlet, vanderbilti, whose mantle gloss is characterised as being blue rather than green or greenish-blue (Meyer de Schauensee & Ripley, 1940). ...
Article
Full-text available
Nias is the largest of a chain of islands off the west coast of Sumatra. Historically, it received extensive attention from ornithologists, leading to the description of numerous endemic subspecies of birds. However, serious attention ceased before the end of World War II, and there has been an almost complete lack of modern-day work. At the same time, Nias now has the most degraded and fragmented natural environment of all the larger West Sumatran islands. Here, we report the results of recently renewed fieldwork coupled with an exhaustive perusal of the island’s ornithological literature, and present a fresh taxonomic and conservation appraisal of the Nias avifauna. We furnish detailed information on all 165 bird species known from the island to date, including five newly recorded species, and recommend that seven species be removed from Nias’s list. We flag a number of taxa, foremost the local Red-backed Dwarf Kingfisher Ceyx [rufidorsa] captus and Brown Wood Owl Strix [leptogrammica] niasensis, whose taxonomic distinctness may have been overlooked or underestimated in past accounts, while equally providing information to help synonymise numerous dubiously endemic taxa. We discovered the largest surviving population of the globally Critically Endangered Silvery Woodpigeon Columba argentina on Nias and its offshore islands. Based on the results of this study, we point to the possible local extinction of a number of species on Nias, including endemic subspecies, that can largely be attributed to wholesale loss of original forest cover on the island, and recommend comprehensive conservation efforts to ensure the survival of the remaining avifauna.
... Examples of such taxonomic neglect are abundant in the Indonesian Archipelago (e.g. Rheindt et al. 2017Rheindt et al. , 2020Ng et al., 2016) ...
Article
Full-text available
Southeast Asian avifauna is under threat from both habitat loss and illegal poaching, yet the region’s rich biodiversity remains understudied. Here, we uncover cryptic species-level diversity in the Sunda Blue Robin (Myiomela diana), a songbird complex endemic to Javan (subspecies diana) and Sumatran (subspecies sumatrana) mountains. Taxonomic inquiry into these populations has previously been hampered by a lack of DNA material and the birds’ general scarcity, especially sumatrana which is only known from few localities. We demonstrate fundamental bioacoustic differences in courtship song paired with important distinctions in plumage saturation and tail length that combine to suggest species-level treatment for the two taxa. Treated separately, both taxa are independently threatened by illegal poaching and habitat loss, and demand conservation action. Our study highlights a case of underestimated avifaunal diversity that is in urgent need of revision in the face of imminent threats to species survival.
... Concerning the endemic species and where they dispersed from relative to Wallace's Line, from the west there is one, the swiftlet Collocalia natalis Lister, 1889. According to Rheindt et al. (2017), it is sister to Collocalia linchi linchi Horsfield & Moore, 1854 (Fig. 3), which occupies Java and some islands in the eastern Java Sea (IUCN Red List, 2020). Norman et al. (1998) indicates that the hawk-owl Ninox natalis Lister, 1888 came from the east, because it is most closely related to Ninox squamipila (Bonaparte, 1850) on Seram and nearby islands, and to Ninox hypogramma (Gray, 1861) on Halmahera. ...
Article
Full-text available
Based on a comprehensive literature survey, we determined the sources of the terrestrial vertebrate species on Christmas Island, asking where they originated relative to Wallace’s Line (the southern end of the divide lies 1100 km to the east, where the Lombok Strait adjoins the eastern Indian Ocean). The two bats, Pipistrellus murrayi and Pteropus natalis, are from the west. Concerning the endemic and ‘resident’ bird species, one is from the west (Collocalia natalis), four are from the east (Accipiter fasciatus, Egretta novaehollandiae, Falco cenchroides and Ninox natalis) and the other 15 are ambiguous or indeterminate. Most of the land-locked species are also from the east: rodents Rattus macleari and Rattus nativitatis, and squamates Cryptoblepharus egeriae, Emoia nativitatis and Lepidodactylus listeria. Additionally, two have westerly origins (Crocidura trichura and Cyrtodactylus sadleiri), one is ambiguous (Emoia atrocostata) and another is unknown (Ramphotyphlops exocoeti). West-directed surface currents that flow across the eastern Indian Ocean towards Christmas Island would have facilitated most of the land-animal colonizations. We therefore suggest that Wallace’s Line be redrawn such that the landmass is placed on the Australasian side of this fundamental biogeographical boundary.