Molecular Phylogenetics and Evolution 53 (2009) 948–960Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier .com/locate /ympevA molecular phylogenetic survey of caprimulgiform nightbirds illustrates the utility
of non-coding sequences
Michael J. Braun a,b,*, Christopher J. Huddleston a
aDepartment of Vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, 4210 Silver Hill Road, Suitland, MD 20746, USA
bBehavior, Ecology, Evolution, and Systematics Program, University of Maryland, College Park, MD 20742, USA
a r t i c l e i n f oArticle history:
Received 10 February 2009
Revised 19 August 2009
Accepted 25 August 2009
Available online 29 August 2009
Keywords:
Phylogeny
Evolution
Caprimulgiformes
Aegothelidae
Apodiformes
Bird
Systematics
Insertions
Deletions
Rare genomic changes
Non-coding sequences1055-7903/$ - see front matter Published by Elsevier
doi:10.1016/j.ympev.2009.08.025
* Corresponding author. Address: Department of
Museum of Natural History, Smithsonian Institution, 4
Maryland 20746, USA. Fax: +1 301 238 3059.
E-mail address: braunm@si.edu (M.J. Braun).a b s t r a c t
The order Caprimulgiformes comprises five bird families adapted to nocturnal activity. The order has been
regarded as monophyletic, but recent evidence suggests that swifts and hummingbirds (Apodiformes)
belong within it. To explore the group’s phylogeny, we obtained more than 2000 bp of DNA sequence
from the cytochrome b and c-myc genes for 35 taxa, representing all major lineages and outgroups.
Non-coding sequences of the c-myc gene were unsaturated, readily alignable and contained numerous
informative insertions and deletions (indels), signalling broad utility for higher level phylogenetics. A
12 bp insertion in c-myc links Apodiformes with owlet-nightjars, confirming paraphyly of the traditional
Caprimulgiformes. However, even this rare genomic change is homoplasious when all birds are consid-
ered. Monophyly of each of the five traditional families was strongly confirmed, but relationships among
families were poorly resolved. The tree structure argues against family status for Eurostopodus and
Batrachostomus, which should be retained in Caprimulgidae and Podargidae, respectively. The genus
Caprimulgus and both subfamilies of Caprimulgidae appear to be polyphyletic. The phylogeny elucidates
the evolution of adaptive traits such as nocturnality and hypothermia, but whether nocturnality evolved
once or multiple times is an open question.
Published by Elsevier Inc.1. Introduction
The impact of molecular data on systematics has been dramatic.
Insights from molecular phylogenies have revolutionized our
thinking about evolutionary relationships, especially in poorly
known groups. At the same time, it has become clear that short
DNA sequences from a single gene will not be sufficient to obtain
stable phylogenies for many groups. Single-gene trees may differ
from each other and from the species tree under a variety of cir-
cumstances (Maddison, 1997; Kubatko and Degnan, 2007), and
the resolving power of a particular gene will depend on how well
its rate of evolution is matched to the structure of the tree in ques-
tion (Braun and Kimball, 2002). Obtaining reliable, well-resolved
phylogenies for many groups will require data frommultiple genes
evolving at a variety of rates (Rokas and Carroll, 2005). To this end,
we here present mitochondrial and nuclear DNA sequence data for
a broad sampling of a ubiquitous but little known order of birds.Inc.
Vertebrate Zoology, National
210 Silver Hill Road, Suitland,The traditional avian order Caprimulgiformes comprises about
120 species intricately adapted to nocturnal, mostly insectivorous
lifestyles. It includes some familiar birds, such as the nightjars
(Caprimulgus) and nighthawks (Chordeiles), as well as many exotic
and bizarre forms, such as the oilbird (Steatornis), a frugivorous,
echo-locating troglodyte of South American jungles. Their noctur-
nal nature, cryptic plumage, secretive behavior, and generally trop-
ical distributions make most caprimulgiforms difficult to study in
life. They are consequently among the most poorly known groups
of birds. Information concerning their taxonomy, distribution and
life history has been spotty and scattered, but several recent trea-
tises summarize available information (Cleere, 1998, 1999; Cohn-
Haft, 1999; Holyoak, 1999a,b; Thomas, 1999; Holyoak, 2001).
The generally primitive state of caprimulgiform systematics is
illustrated by the continuing discovery of new genera (Cleere
et al., 2007) and new species (Louette, 1990; Lencioni-Neto, 1994;
Safford et al., 1995; Sangster and Rozendaal, 2004). However, con-
tinued research on vocalizations and morphology further improves
our understanding of nightbird species and species groups (Davis,
1962, 1978; Schulenberg et al., 1984; Fry, 1988; Cohn-Haft, 1993;
Robbins et al., 1994; Robbins and Parker, 1997; Pratt, 2000; Cleere,
2002) and a number of pertinentmolecular andmorphological phy-
logenetic studies have now been published (Sibley and Ahlquist,
M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960 9491990; Mariaux and Braun, 1996; Brumfield et al., 1997; Johansson
et al., 2001; Mayr, 2002; Mayr et al., 2003; Dumbacher et al.,
2003; Barrowclough et al., 2006; Ericson et al., 2006; Larsen et al.,
2007; Livezey and Zusi, 2007; Hackett et al., 2008).
Seminal works on caprimulgiform systematics were written by
Sclater (1866a,b), Beddard (1886), Wetmore (1918), Peters (1940),
and Tripepi et al. (2006). In addition, a detailed historical review
was given by Sibley and Ahlquist (1990) and updated by later
authors. Key features and questions are summarized here. The or-
der includes five distinctive lineages, usually accorded family rank.
These are the nightjars and nighthawks (Caprimulgidae; 89–90
species), frogmouths (Podargidae; 13 species), owlet-nightjars
(Aegothelidae; 9 species), potoos (Nyctibiidae; 7 species), and oil-
bird (Steatornithidae; 1 species). Although the monophyly of each
of these lineages has never been seriously questioned, Sibley and
Ahlquist (1990) suggested erecting two new families based on
deep divergences in their DNA hybridization data: Eurostopodidae
for the nightjar genus Eurostopodus, and Batrachostomidae for the
frogmouth genus Batrachostomus. This finding is consistent with a
fossil record extending back to Eocene times for several groups
(Mourer-Chauviré, 1982, 1987; Olson, 1987; Peters, 1991), which
is as early as any modern bird families appear (James, 2005). Other
molecular data are also consistent with these deep splits (Mariaux
and Braun, 1996; Brumfield et al., 1997; Barrowclough et al., 2006;
Cleere et al., 2007; Larsen et al., 2007).
Morphological diversity (Hoff, 1966) and genetic divergences
(Sibley and Ahlquist, 1990; Mariaux and Braun, 1996) among the
five lineages are even greater than those within them. In fact, accu-
mulated divergence is so great that several analyses of molecular
andmorphological data failed to recover a close relationship among
the five lineages (Johansson et al., 2001; Mayr, 2002; Chubb, 2004;
Cracraft et al., 2004; Fain and Houde, 2004; Barrowclough et al.,
2006; Ericson et al., 2006; Livezey and Zusi, 2007), leading to suspi-
cions that the order might be polyphyletic. These suspicions were
exacerbated by strong indications of a sister group relationship of
owlet-nightjars with swifts and hummingbirds (Mayr, 2002; Mayr
et al., 2003; Cracraft et al., 2004; Barrowclough et al., 2006; Ericson
et al., 2006). However, studies sampling multiple nuclear genes and
a larger number of taxa have now found strong support for mono-
phyly of an expanded Caprimulgiformes with Apodiformes nested
within it (Ericson et al., 2006; Hackett et al., 2008). A higher-order
morphological phylogeny also recovered a clade consisting of these
groups, but with Apodiformes sister to the traditional Caprimulgi-
formes (Livezey and Zusi, 2007). However, two of the five synapo-
morphies proposed by Livezey and Zusi for the traditional clade
Caprimulgiformes were criticized by Mayr (2008; see also
Discussion).
Generic and subfamily demarcations are in need of revision. The
large genus Caprimulgus (>50 species), for example, appears to be a
catch-all of distantly related, morphologically conservative lin-
eages, but no satisfactory arrangement has yet been proposed
(Sibley and Ahlquist, 1990; Cleere, 1998, 1999; Holyoak, 2001;
Barrowclough et al., 2006; Larsen et al., 2007). Conversely, mor-
phologically divergent lineages are represented by five small cap-
rimulgid genera (Macrodipteryx, Macropsalis, Uropsalis,
Hydropsalis, Eleothreptus) based on sexually selected characters
that may or may not warrant generic distinction. Still other genera
(e.g., Nyctibius, Batrachostomus) contain species that are older than
many avian families, if even an approximate molecular clock can
be applied (Mariaux and Braun, 1996; Cleere et al., 2007). Finally,
although the standard division of Caprimulgidae into nightjar
(Caprimulginae) and nighthawk (Chordeilinae) subfamilies has a
morphological basis in the structure of the palate (desmognathous
in Chordeilinae vs. schizognathous in Caprimulginae; Cleere, 1999)
and the presence or absence of rictal bristles (Oberholser, 1914;
Holyoak, 2001), the character states have not been polarized so itis not clear whether they support monophyly of either group. Both
DNA hybridization and DNA sequence data strongly suggest that
Eurostopodus species are sister to a clade including all other capri-
mulgids examined, and that relationships within the remaining
caprimulgids are inconsistent with the standard caprimulgine/
chordeiline split (Sibley and Ahlquist, 1990; Mariaux and Braun,
1996; Barrowclough et al., 2006; Larsen et al., 2007).
A well-resolved and well-supported phylogenetic tree for
Caprimulgiformes would provide an organizing framework with
which to understand their biology and evolutionary history.
Toward that end, we here extend the cytochrome b sequence data
set of Mariaux and Braun (1996) to include new ingroup and out-
group taxa, and add a second gene sequence from the nuclear locus
c-myc. The c-myc gene is more slowly evolving than cytochrome b,
providing better resolution deep in the tree (Harshman et al.,
2003). Because it includes non-coding sequence elements that
accommodate length variation more readily than protein-coding
sequences, c-myc also provides a relatively large number of phylo-
genetically informative insertion/deletion (indel) characters. The
expanded data set provides clear evidence that the traditional or-
der Caprimulgiformes is paraphyletic, and allows us to address
several of the systematic issues mentioned above.2. Materials and methods
2.1. Tissue samples and data collection
Tissue samples from 34 nightbirds and putative relatives were
obtained from our own fieldwork and from genetic resource collec-
tions of major museums (Table 1). The partial cytochrome b se-
quences of Mariaux and Braun (1996) were confirmed and
extended to encompass the entire coding sequence. We used
chicken (Gallus gallus) cytochrome b and c-myc sequences from
GenBank (NC_001323 and J00889) as an unambiguous outgroup
for phylogenetic analyses. Genomic DNA was extracted from tis-
sues using methods described in Mariaux and Braun (1996) and
Robbins et al. (2005). Cytochrome b and a portion of c-myc were
amplified via PCR using the primers listed in Table 2 and methods
similar to those described in Robbins et al. (2005) and Harshman
et al. (2003), respectively. Some c-myc sequences required cloning,
due to length heterozygosity. Sequences were generated on auto-
matic sequencers using ABI Big Dye Ready-Reaction kits and man-
ufacturer standard protocols, except the Big Dye was diluted 1:16.
DNA sequences were deposited in GenBank under accession num-
bers shown in Table 1.2.2. Sequence alignment and phylogenetic analyses
Cytochrome b sequences were aligned using ClustalW
(Thompson et al., 1994). No internal gaps were needed. C-myc
sequences were aligned using Clustal W as well, followed by
manual adjustment by eye. A 39 bp segment near the 30 end of
the c-myc intron was excluded from further analyses, due to
ambiguous alignment of a hypervariable poly-T element.
Because of the very different rates of evolution and mutational
biases between nuclear and mitochondrial (mtDNA) genes, we ex-
plored the possibility of phylogenetic incongruence among loci be-
fore combining data (Bull et al., 1993; Swofford et al., 1996). We
used the incongruence length difference test (Farris et al., 1995)
as implemented in PAUP* (Swofford, 2002), where it is known as
the partition homogeneity test.
Maximum parsimony (MP) tree searches were conducted for
each gene separately and the two genes combined using PAUP* v.
4.0b10 (Swofford, 2002). We ran heuristic searches with equal
weighting, 10 random sequence addition replicates, and tree
Table 1
Specimens examined. Taxonomy follows Dickinson (2003).
Common name Scientific name Voucher and/or tissue no. Collector GenBank Accession Nos.
Cytochrome b c-myc
Mountain owlet-nightjar Aegotheles albertisi MVICa E044 Schodde, R. X95764 FJ588484
Barred owlet-nightjar Aegotheles bennettii MVICa E636 Christidis, L. X95774 FJ588482
Australian owlet-nightjar Aegotheles cristatus MVICa W191 Baverstock, P.R. X95775 FJ588483
Feline owlet-nightjar Euaegotheles insignis KUNHMb 95997 Mack, A. FJ588456 EU738243
Philippine frogmouth Batrachostomus septimus CMNHc 36767 (B499) Gonzales, D. EF100673 EU738255
Marbled frogmouth Podargus ocellatus MVICa C332 Christidis, L. X95771 FJ588474
Papuan frogmouth Podargus papuensis MVICa C876 Wombey, J. X95772 FJ588475
Philippine nightjar Caprimulgus manillensis USNMd B06090 Ross, C.A. FJ588443 FJ588463
Chuck-will’s-widow Caprimulgus carolinensis USNMd 622565 (B16652) Schmidt, B.K. FJ588442 FJ588461
White-tailed nightjar Caprimulgus cayennensis USNMd 625369 (B11295) Schmidt, B.K. FJ588444 FJ588464
Blackish nightjar Caprimulgus nigrescens USNMd 586047 (B04478) Robbins, M.B. FJ588446 FJ588466
Great eared nightjar Eurostopodus macrotis USNMd 607328 (B03732) Dickerman, R.W., and party FJ588447 EU738292
White-throated nightjar Eurostopodus mystacalis MVICa JWC129 Wombey, J. X95779 FJ588468
Papuan nightjar Eurostopodus papuensis MVICa E660 Schodde, R. X95780 FJ588467
Common poorwill Phalaenoptilus nuttallii USNMd B00084 Braun, M.J. X95770 FJ588462
Common nighthawk Chordeiles minor USNMd 586281 (B07837) unknown FJ588441 FJ588459
Sand-colored nighthawk Chordeiles rupestris ANSPe T2755 Sornoza, F. X95778 FJ588460
Rufous-bellied nighthawk Lurocalis rufiventris ANSPe 4467 Davis, T.J. FJ588445 FJ588465
Long-tailed potoo Nyctibius aethereus LSUMZf B10877 Meyer, A.S. X95781 FJ588473
Rufous potoo Nyctibius bracteatus LSUMZf B4509 Cardiff, S.W. X95765 EU738327
Great potoo Nyctibius grandis LSUMZf B15415 Bates, J.M. X95766 EU738328
Common potoo Nyctibius griseus USNMd 612299 (B00493) 1989 Bocas Expedition X95767 FJ588472
Northern potoo Nyctibius jamaicensis KUNHMb B2116 netted FJ588449 FJ588471
White-winged potoo Nyctibius leucopterus LSUMZf B20315 Cohn-Haft, M. X95768 FJ588469
Andean potoo Nyctibius maculosus LSUMZf B271 Braun, M.J. FJ588448 FJ588470
Oilbird Steatornis caripensis LSUMZf B7474 Willard, D.E. EF100675 FJ588476
Crimson topaz Topaza pella USNMd 625397 (B11959) Schmidt, B.K. FJ588450 FJ588477
Escudo hummingbird Amazilia tzacatl handleyi USNMd 607712 (B01021) Parsons, T.J. FJ588452 FJ588479
Bronzy hermit Glaucis aeneus USNMd B00319 1989 Bocas Expedition FJ588451 FJ588478
Uniform swiftlet Aerodramus vanikorensis USNMd 608672 (B04039) Angle, J.P. and Beehler, B.M. FJ588453 EU738244
Chapman’s swift Chaetura chapmani USNMd 609124 (B05266) Braun, M.J. FJ588454 FJ588480
Whiskered treeswift Hemiprocne comata USNMd 607338 (B03790) Ross, C.A. FJ588455 FJ588481
Barn owl Tyto alba USNMd 612697 (B06443) James, R.C. FJ588458 EU738388
Eastern screech owl Otus asio USNMd B27714 Braun, M.J. FJ588457 FJ588485
a Museum of Victoria, Melbourne, Australia.
b University of Kansas Natural History Museum, Lawrence, KS.
c Cincinnati Museum of Natural History, Cincinnati, OH.
d National Museum of Natural History, Smithsonian Institution, Washington, DC.
e Philadelphia Academy of Natural Sciences, Philadelphia, PA.
f Museum of Zoology, Louisiana State University, Baton Rouge, LA.
950 M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960bisection-reconnection (TBR) branch swapping. Bootstrap analyses
were performedwith equal weighting, 1000 bootstrap pseudorepli-
cates, and10 randomsequenceaddition replicates for everypseudo-
replicate. Weighted parsimony analyses (WMP) were performed in
the same way with the following weights: c-myc—1st positions =
10.3, 2nd positions = 15.5, all other sites = 1; cytochrome b—1st
positions = 3, 2nd positions = 10, and 3rd positions = 1.
For maximum likelihood (ML) analysis, models of sequence
evolution for cytochrome b, c-myc, and a combination of both
genes were evaluated using ModelTest 3.06 (Posada and Crandall,
1998). First, a neighbor-joining (NJ) tree was produced via PAUP*
under the Jukes–Cantor model of evolution (Jukes and Cantor,
1969). Parameters were then calculated for fifty-six nested models
of evolution on the NJ tree and the best model of sequence evolu-
tion was selected using the Akaike Information Criterion (AIC). This
model and parameter values were fixed in a heuristic maximum
likelihood search in PAUP* with 10 rounds of random sequence
addition and TBR branch swapping. Model estimation was then re-
peated on the resulting tree. This process was repeated in a succes-
sive approximation approach (Sullivan et al., 1995; Swofford et al.,
1996) until tree topology and model parameter values converged.
The model and parameter values were then fixed for bootstrap
analyses in PAUP*. In the cases of cytochrome b alone and c-myc
alone, 100 pseudoreplicates with one random sequence addition
at each pseudoreplicate were performed. For the combined analy-
sis, we performed bootstrap analyses using 650 pseudoreplicates
with one random addition at each pseudoreplicate.Bayesian analyses were performed using MrBayes 3.1 (Huelsen-
beck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003). A par-
tition was set for each gene with separate 6-parameter models.
Four Markov chains were run with heating and swapping for 20
million generations each. Topology and model parameters were
sampled every 100 generations and used to determine the poster-
ior probabilities of clades and estimates of model parameters. The
first 500 samples were discarded to allow for burn-in to the target
distributions; default settings were used for all other options.
Hypotheses of monophyly derived from traditional classifica-
tions were examined by comparing globally optimal trees to the
best trees that could be found compatible with each hypothesis. In
each analysis, a traditional group was constrained to be monophy-
letic in PAUP*, and a new model of sequence evolution estimated
by the method above. Then, a heuristic ML search with 10 random
sequence addition replicates was conducted to find the optimal tree
under that constraint. The difference in log likelihood of the con-
strained tree vs. the unconstrained tree was taken as a measure of
support for a particular hypothesis of relationship or its alternative.3. Results
3.1. Sequence analysis
The aligned data set comprises 2274 nucleotide sites from 35
taxa, about equally divided between cytochrome b and c-myc
Table 2
Main oligonucleotide primers used for PCR and sequencing.a
Primer Sequence (50–30) Reference
c-myc
MYC-FOR-01b TAATTAAGGGCAGCTTGAGTC Harshman et al.
(2003)
MYC-FOR-02b TGAGTCTGGGAGCTTTATTG Harshman et al.
(2003)
MYC-FOR-03 AGAAGAAGAACAAGAGGAAG Harshman et al.
(2003)
MYC-FOR-04 AAAAGGCTAAAGTTGGAC Harshman et al.
(2003)
MYC-FOR-05 CACAAACTYGAGCAGCTAAG Harshman et al.
(2003)
MYC-REV-01b CCAAAGTATCAATTATGAGGCA Harshman et al.
(2003)
MYC-REV-02b TGAGGCAGTTTTGAGGTTCT Harshman et al.
(2003)
MYC-REV-03 CATTTTCGGTTGTTGCTG This study
MYC-REV-04 GGCTTACTGTGCTCTTCT Harshman et al.
(2003)
MYC-REV-06 TTAGCTGCTCAAGTTTGTG Harshman et al.
(2003)
Cytochrome bc
L14703b GGMCAAAAMATTGCMTCYCAC This study
L14764b TGRTACAAAAAAATAGGMCCMGAAGG Sorenson et al.
(1999)
L14990 CCATCCAACATCTCAGCATGATGAAA Kocher et al.
(1989)
L15323 CCATGAGGACAAATATCATTCTGAGGTGC Mariaux and
Braun (1996)
H15298 CCCCTCAGAATGATATTTGTCCTCA This study
H15730 GGGATTGAGCGTAGGATGGC This study
H16060b TTTGGYTTACAAGACCAATG Robbins et al.
(2005)
a Other ancillary primers were used to complete sequences of some taxa; their
sequences are available from the authors.
b Amplification primers.
c Numbering of cytochrome b primers based on the Gallus mtDNA sequence
(Desjardins and Morais, 1990).
M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960 951(Table 3). All sites in cytochrome b code for protein, while only
about half of c-myc sites do. The rest of c-myc is split between a
50 intron and the 30 untranslated region (UTR) of the gene. More
than half of cytochrome b sites (51.3%) are variable, compared to
one-third of c-myc sites (33.4%). Most variable cytochrome b sites
are also phylogenetically informative (506 of 582 or 87%), while
only 58% (221 of 378) of variable c-myc sites are.
Pairwise uncorrected sequence divergence is quite high in cyto-
chrome b, ranging from 3.0 to 22.7% among ingroups (i.e., tradi-
tional caprimulgiforms) and from 13.6 to 25.1% in ingroup to
outgroup comparisons. The c-myc gene is more slowly evolving,Table 3
Summary of sequence data and tree statistics.
Cytochrome b c-myc
Intron b
Aligned length 1143 365
Variable sites 586 192
Informative sites
Coding (1/2/3)a 121/36/354 —
Non-Coding — 113
Total 511 113
Indels
Informative 0 18
Total 0 40
MP tree length 3304 390
Consistency index (CI)b 0.31 0.64
Retention indexb 0.37 0.74
Rescaled CIb 0.11 0.47
a Coding (1/2/3) refers to first, second, and third codon positions.
b Consistency and retention indices for the various data partitions were calculated onwith ingroup distances ranging from 0.0 to 8.7% and ingroup–out-
group distances ranging from 4.9 to 13.8% (distances greater than
10.1% all involve Gallus). The relatively high ratios for both genes
of informative to variable sites and ingroup to outgroup distances
suggest that there are deep divergences within the ingroup, and
that some ingroup lineages may rival outgroups in age. Normalized
plots of pairwise distances suggest that cytochrome b is heavily
saturated for substitutions among the deeper lineages (Fig. 1a).
The numbers of transitions and transversions plateau quickly and
eventually converge at larger distances. The intron and UTR ele-
ments of c-myc are evolving more rapidly than the exon, but none
appear to be saturated within the ingroup (Fig. 1b and c; distances
below 0.15). For the UTR, a cloud of points at higher distances
gives a semblance of plateauing, but this may be unduly influenced
by a single taxon, because all these points involve the outgroup
Gallus.
Base composition varied between the two genes and among
functional partitions within each gene (Table 4). Overall, base fre-
quencies of c-myc were relatively uniform. The exon sequence was
relatively low in T, and the 30 UTR was low in G. However, no par-
tition of the c-myc data differed significantly from homogeneity of
base frequencies across taxa (Table 4). On the other hand, the cyto-
chrome b sequences were high in C and low in G, and third codon
positions were especially low in G and T. Both variable sites and
third positions, which comprise the majority of variable sites for
cytochrome b, were significantly heterogeneous in base composi-
tion across taxa. Most but not all of this heterogeneity appeared
to be due to the inclusion of swifts and tree swifts, which had a
very low proportion of T (mean = 0.072) at third positions. When
those three taxa were deleted, third position heterogeneity was
only marginally significant after multiple test correction (Table 4).
3.2. Length variation
Indels were absent from our sequences of cytochrome b, a
mtDNA gene whose length is usually highly conserved. The aligned
c-myc data, on the other hand, includes at least 54 inferred indels,
ranging in length from 1 to 16 bp (Table 3). The high degree of se-
quence conservation in c-myc makes the alignment unambiguous
for most indels. In the few cases for which there is more than
one equally parsimonious placement, the alternative alignment
generally would have little or no effect on phylogenetic inference.
There is, however, one poly-T element near the intron–exon junc-
tion that is highly variable in length. This region (39 sites) was ex-
cluded from the dataset for phylogenetic analysis.
Fifty-one of the 53 alignable indels occur in non-coding regions
and 28 of 53 are phylogenetically informative (Table 3). Of the twoTotal
Exon 3 30 UTR Indels
579 187 — 2274
123 66 — 967
6/4/62 — — 127/40/416
— 39 — 152
72 39 — 735
2 9 — 28
2 12 — 54
271 93 29 —
0.50 0.77 0.97 —
0.66 0.84 0.99 —
0.33 0.65 0.96 —
the ML tree for the combined data.
c-myc (Intron b & 3'UTR)
0
50
100
150
200
250
300
0 0.05 0.1 0.15 0.2 0.25 0.3
ML Distances (c-myc)
Intron b Substitutions
3'UTR Substitutions
c-myc (Exon 3)
0
50
100
150
200
250
300
0 0.05 0.1 0.15 0.2 0.25 0.3
N
o.
 o
f S
ub
st
itu
tio
ns
 p
er
 k
b
Exon 3 Substitutions
Cytochrome b
0
50
100
150
200
250
300
0 0.05 0.1 0.15 0.2 0.25 0.3
Transitions
Transversions
Fig. 1. Saturation plots of observed substitutions per kb in various data partitions vs. divergence of c-myc sequences estimated from the ML model for all pairwise
comparisons among taxa.
952 M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960non-coding regions, more indels occur per unit length in the intron
than the 30 UTR (11 vs. 6.4 per 100 bp), but the UTR has a much
higher proportion of informative indels, so the number of informa-
tive indels per unit length is similar (4.7 per 100 bp of intron, 4.8
per 100 bp of UTR).
The direction of mutation can be inferred for 47 indels from
their distribution among taxa on the optimal trees derived fromsubstitutional variation. Twenty are insertions and 27 are dele-
tions. Short indels of 1–3 bp are more common than long ones,
and this is especially true for insertions (Fig. 2). Only one insertion
is longer than 4 bp, while there are eight such deletions. The great-
er number and length of deletions results in a marked directional-
ity in total length change: across the data set, a total of 99 bp have
been deleted and only 44 bp have been inserted.
Table 4
Base composition of sequence data.
Data partition Mean %A Mean %C Mean %G Mean %T Chi-squarea P-valueb
c-myc
All sites 29.9 22.9 23.9 23.3 18.6812 1.0000
Variable sites 27.5 23.5 25.7 23.3 45.5399 1.0000
Intron b 22.8 20.6 24.8 31.9 27.0078 1.0000
Exon 3 33.6 24.2 25.1 17.1 9.7654 1.0000
30 UTR 31.0 22.6 18.5 27.9 8.5670 1.0000
Exon 3rd positions 25.0 28.7 28.8 17.5 24.6905 1.0000
Cytochrome b
All sites 28.4 33.9 12.5 25.2 62.7774 0.9992
Variable sites 32.1 42.7 6.8 18.4 145.4238 0.0031
1st positions 26.8 29.5 20.9 22.7 19.2206 1.0000
2nd positions 20.7 26.9 13.0 39.4 4.8254 1.0000
3rd positions 37.7 45.2 3.6 13.5 207.7503 0.0000
3rd positions (minus swifts) 37.9 44.6 3.5 14.1 134.0671 0.0034
a The homogeneity of base frequencies across taxa was tested by chi-square.
b All P-values are shown prior to multiple test correction. Using the sequential Bonferroni procedure (Hochberg, 1988), the adjusted critical value over all tests was
a < 0.005.
0
2
4
6
8
10
12
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Length of indel (bp)
N
o.
 o
f i
nd
el
s
Deletions
Insertions
Fig. 2. Number of insertions and deletions of various lengths inferred in c-myc
sequences. All indels that could be polarized are included (47 of 54).
M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960 9533.3. Phylogenetic analyses—deeper nodes
When c-myc and cytochrome b datasets are analysed sepa-
rately, both genes provide some phylogenetic resolution at various
depths in the tree down to the family level (Fig. 3). In general, cyto-
chrome b provided better resolution at the tips of the tree, while
c-myc provided better resolution of deeper nodes, as might be ex-
pected from their relative evolutionary rates (Fig. 1). The phyloge-
netic signals from the two genes appear to be in good agreement.
The topologies of the optimal trees from separate analyses of
c-myc and cytochrome b are largely congruent, and there is low
support for those nodes that are in conflict. A partition homogene-
ity test found no significant incongruence between the two genes
(P = 0.189). We therefore combined the data for further analysis.
Combining data from the two loci dictates a hybrid model of se-
quence evolution for ML analyses in PAUP* (Table 5), but separate
models were applied to the two gene partitions in Bayesian
analyses.
The resulting phylogeny has strong support for many traditional
and some novel groupings (Fig. 4). Among the deeper nodes, all the
traditional families of caprimulgiforms and outgroups were recov-
ered with ML bootstrap support greater than 95% and Bayesian
posterior probabilities (PP) of 1.0 (Fig. 4). All traditional families
were further supported by one to four indel characters in c-myc.
The traditional orders Apodiformes and Strigiformes received MLbootstrap support of greater than 90% and Bayesian PP of 1.0
(Fig. 4). A 1 bp indel in c-myc also supported apodiform mono-
phyly. Most of these traditional groupings were also found in single
gene analyses, albeit generally with lower support values (Fig. 3).
The oilbird, Steatornis, grouped consistently with potoos (Nyctibius)
in analyses of c-myc (Fig. 3) or combined analyses (Fig. 4), but with
low support.
Conspicuously absent was support for the traditional order Cap-
rimulgiformes. This group was not monophyletic in any analysis of
single genes or combined data. ‘‘Outgroups” (swifts, humming-
birds, owls) were interspersed with ‘‘ingroups” in many analyses,
but with inconsistent and weak associations (Figs. 3 and 4). Many
basal branches were very short, indicating that there is little signal
supporting them (e.g., Fig. S1). The one grouping that appears con-
sistently is a clade linking Apodiformes (swifts and hummingbirds)
with Aegothelidae (owlet-nightjars). This clade was found in all
analyses of c-myc, and received relatively strong support in
WMP, ML and Bayesian analyses (Fig. 3). It was also found in MP,
WMP and Bayesian analyses of cytochrome b (not shown), but with
weaker support. This nontraditional clade appears again in all anal-
yses of the combined data, with 79% ML bootstrap support and
Bayesian PP of 1.0 (Fig. 4). All apodiforms and aegothelids also
shared a 12 bp insertion in the coding sequence of c-myc not found
elsewhere in the data set, providing an apparent synapomorphy for
the clade (see Phylogenetic Utility of Introns, UTRs and Indels below).3.4. Phylogenetic analyses—shallow nodes
Within families, a number of interesting results emerge (Fig. 4).
In Caprimulgidae, the large genus Caprimulgus is non-monophy-
letic. Lurocalis and Phalaenoptilus are found in well-supported
clades with one or more Caprimulgus species. Lurocalis and
Chordeiles are found in separate clades, with strong support values
for one and a 3 bp insertion in the c-myc intron supporting the
other. This topology renders both subfamilies, Caprimulginae and
Chordeilinae, non-monophyletic. Support for each of these results
comes from both cytochrome b and c-myc (Fig 3).
Basal to all other caprimulgids, we find three exemplars of the
genus Eurostopodus on two successive branches in both single gene
and combined analyses (Figs. 3 and 4). However, the branching or-
der of the two groups is reversed in single gene analyses. The cyto-
chrome b divergence of Eurostopodus macrotis from the other
Eurostopodus averages 17.8%, much higher than that found be-
tween most avian congeners. The three Eurostopodus are clearly
associated with other caprimulgids by strong support values and
Fig. 3. Phylogenetic trees based on separate analyses of cytochrome b and c-myc sequences. ML bootstrap values are shown above the line and Bayesian posterior
probabilities are shown below. Topologies are majority rule consensus trees from the ML bootstrap analyses (i.e., nodes with support below 50% are collapsed).
Table 5
Models of sequence evolution for various data partitions in ML analyses.
Cytochrome b
(GTR+I+C)a
c-myc
(HKY+I+C)b
Combined
(GTR+I+C)
Base frequencies
A 0.3381 0.2839 0.3035
C 0.4493 0.2374 0.3587
G 0.0546 0.2412 0.1667
T 0.1580 0.2375 0.1711
Substitution rates
AM C 0.2092 1.0000 1.1592
AM G 7.6057 3.4269 4.1531
AM T 0.6518 1.0000 1.331
CM G 0.3125 1.0000 0.291
CM T 8.0419 3.4269 13.0189
GM T 1.0000 1.0000 1.0000
C shape parameter (a) 0.4761 0.7106 0.4466
Proportion of invariant sites (I) 0.4118 0.3236 0.3785
a GTR—general time reversible (substitution rates symmetrical within types, free
to vary between types).
b HKY—(Hasegawa et al., 1985).
954 M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960three c-myc indels on the node that defines the family, but they are
also clearly separated from other caprimulgids by strong supportand a 7 bp deletion on the node that groups those taxa together.
Most of the support linking Eurostopodus with other caprimulgids
comes from c-myc; single gene analyses of cytochrome b showed
weak signal for this clade (Fig. 3).
In the Nyctibiidae, we find two strongly supported pairs of
small gray potoos, griseus + jamaicensis and leucopterus +maculosus
(Fig. 4). These two pairs form a clade for which nodal support is
less strong, but signal for that clade includes a 1 bp indel in c-
myc. Basal to the small gray potoos, we find the two large potoos,
aethereus and grandis, and the small rufous bracteatus on three suc-
cessive branches with moderate support. Within the remaining
caprimulgiform families, Podargidae and Aegothelidae, relation-
ships of the sampled taxa are well-resolved, with some support
coming from each gene.
In the Apodiformes, the tree swift, Hemiprocne, is sister to two
apodid swifts, and three hummingbirds are sister to the swift clade
(Fig. 4). Each of these nodes conforms to traditional relationships
and receives support from both cytochrome b and c-myc (Fig. 3).
Within hummingbirds, we find the strongest case of apparent con-
flict between genes in the current dataset. Cytochrome b supports
a Topaza + Glaucis clade with 95% ML bootstrap, while c-myc sup-
ports a Glaucis + Amazilia clade with 77% ML bootstrap. Interest-
ingly, the c-myc topology appears in combined analyses (Fig. 4),
overriding the apparently strong signal from cytochrome b.
Fig. 4. Best estimate phylogeny based on combined analysis of cytochrome b and c-myc. Topology is the majority rule consensus tree from the combined ML bootstrap
analyses (i.e., nodes with support below 50% are collapsed). Numbers above branches are ML bootstrap values (left) and Bayesian posterior probabilities (right). All
phylogenetically informative indels were mapped onto the tree by parsimony. Open circles indicate deletions and filled circles indicate insertions, with indel length in bp
given below. Open and filled stars denote two homoplasious indels discussed in the text.
Table 6
Comparison of alternative tree topologies.a
Constrained groupb D ln Lc
Cytochrome b c-myc Combined
Caprimulgiformes monophyletic 13.04 17.69 28.40
Chordeilinae monophyletic 19.38 18.49 34.84
Trochilinae monophyletic 4.03 2.33 1.93
Caprimulgus monophyletic 44.78 36.98 78.41
Eurostopodus monophyletic 7.70 1.24 6.67
a Constrained and unconstrained ML heuristic tree searches were conducted in
PAUP* (Swofford, 2002) as described in Section 2.
b Each constrained tree was compared to the respective optimal unconstrained
ML tree topology for cytochrome b (ln L = 14144.65509), c-myc (ln
L = 5540.89994) and combined datasets (ln L = 20342.74842).
c Numerical values of D ln L are the decreases in likelihood of the constrained
versus the unconstrained trees.
M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960 9553.5. Comparisons of alternative hypotheses
The traditional order Caprimulgiformes was non-monophyletic
in all our optimal trees, as were the subfamilies Chordeilinae and
Trochilinae, and the genera Caprimulgus and Eurostopodus. To as-
sess whether the sequence data strongly contradicted monophyly
of these traditional taxa, we compared the likelihoods of the data-
sets on the optimal tree to their likelihoods on the best alternative
topologies found when these traditional groups were constrained
to be monophyletic (Table 6). Monophyly of Caprimulgiformes re-
quires a decrease of 28 log likelihood units for the combined data.
Similarly, monophyly of either Chordeilinae or Caprimulgus would
imply substantially reduced likelihood for all data partitions. The
decrease in likelihood required to make either Eurostopodus or
Trochilinae monophyletic is less dramatic, due in part to conflict
between the two genes, but both genes support non-monophyly
of these taxa.
956 M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–9603.6. Phylogenetic utility of introns, UTRs, and indels
The nuclear partition of our dataset (c-myc) has less homoplasy
than the mitochondrial partition (cytochrome b), as indicated by
consistency indices (Table 3). This is true for both the coding and
non-coding elements of c-myc, with the non-coding elements
showing the highest consistency indices.
The indel characters in the data set show very little homoplasy
(Table 3). Of the 28 phylogenetically informative indels, 26 map to
a single node on the optimal trees (Fig. 4), resulting in a high re-
scaled consistency index (RC = 0.96). The few indels that are homo-
plasious are either associated with repetitive elements, show
sequence variation indicating independent origins, or both. Exam-
ples of homoplasious repetitive elements include the highly vari-
able, 39 bp poly-T region that was excluded from phylogenetic
analysis, and a GAA insertion near the beginning of the c-myc exon,
which becomes the fourth in a string of directly repeated glutamic
acid codons. This GAA insertion is inferred to have occurred twice
on the optimal trees, once in a swift, Chaetura chapmani, and once
in an owlet-nightjar, Aegotheles bennetti (Fig. 4). The only other in-
del that is homoplasious on the tree is a 2 bp insertion at intron po-
sition 181–182 that occurs in hummingbirds and in chicken.
However, the inserted sequence is different (GT in hummingbirds,
AT in chicken), suggesting an independent origin.
The 12 bp c-myc insertion that supports the clade of swifts,
hummingbirds and owlet-nightjars is not homoplasious within
the current dataset. However, when we examine all available
c-myc sequences, it is found again in some barbets (Fig. 5). This
insertion is associated with a mildly repetitive element; it forms
a third copy of a tandem duplication found in all groups of birds
and crocodilians. The evidence suggests that the barbet third copy
arose independently of the swift/hummingbird/owlet-nightjar
insertion. The barbet insertion is not found in other piciform birds
(Fig. 5). Thus, it is phylogenetically nested in a way that suggests
independent origin.Fig. 5. Twelve bp insertion in c-myc coding sequences of apodiform, aegothelid,4. Discussion
4.1. Phylogenetic utility of introns, UTRs, and indels
The potential advantages of non-coding nuclear sequences for
deeper level phylogenetics have been noted previously (e.g.,
Prychitko and Moore, 2003; Harshman et al., 2003, 2008; Mathee
et al., 2007; Chojnowski et al., 2008; Hackett et al., 2008), and they
certainly applied with the current dataset. The lower homoplasy
found in c-myc vs. cytochrome b is due in part to the higher rate
of evolution of mitochondrial genes (e.g., Table 5) and the rela-
tively great time depth of the tree. In addition, cytochrome b se-
quence evolution is limited by strict constraints of protein
function, such that relatively few sites are free to vary (Table 5).
The combined result of these factors is that the variable sites in
cytochrome b accumulate homoplasious substitutions on the dee-
per branches and are saturated with change (Fig. 1a). The c-myc
exon shares with cytochrome b the functional constraints of all
protein-coding sequences, but its slower evolutionary rate results
in less saturation (Fig. 1) and less homoplasy (Table 3) in our data-
set. The non-coding elements of c-myc (intron and 30 UTR) have
less functional constraint and intermediate evolutionary rates
(Fig. 1). Consequently, they have less homoplasy than either of
the coding elements (Table 3), and, in that sense, produce ‘‘better”
phylogenetic characters.
The other great advantage of non-coding nuclear sequences is
that they are much more likely to retain indel mutations. Indels
arise through molecular processes distinct from substitutions.
They are therefore expected to provide a largely independent
source of phylogenetic information, and be relatively free of many
of the caveats that apply to substitutions (Rokas and Holland,
2000). The indels in the present c-myc data fulfilled these expec-
tations. They provided confirmation for many nodes in the
tree, including some with marginal support from substitutions.
They had very low homoplasy overall, and the ones that wereand some piciform birds. Piciform sequences are from Hackett et al. (2008).
M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960 957homoplasious were all associated with repetitive elements, a clear
signal that those regions could be expected to be hypermutable.
Patterns of indel evolution, with short indels more common than
long ones and deletions more common than insertions, were sim-
ilar to those found in other recent studies of birds and mammals
(Fain and Houde, 2004; Johnson, 2004; Mathee et al., 2007),
although the ratio of deletions to insertions found here (1.35/1)
was lower than the same ratio for beta-fibrinogen intron 7 in doves
(6/1; Johnson, 2004) or three introns in mammals (3/1; Mathee
et al., 2007). This ratio may vary from gene to gene, as crocodilian
c-myc sequences also had a lower ratio (1/1; Harshman et al.,
2003).
Phylogeneticists have often avoided non-coding sequences for
fear of uncertainty or subjectivity in alignment. It is clear from
our dataset and others (Prychitko and Moore, 2003; Harshman
et al., 2003; Fain and Houde, 2004; Ericson et al., 2006; Mathee
et al., 2007; Hackett et al., 2008) that such sequences can be useful
over larger phylogenetic distances than previously supposed. Algo-
rithmic advances promise to make alignment of non-coding se-
quences more straightforward and objective (e.g., Katoh et al.,
2005; Loytynoja and Goldman, 2008; Liu et al., 2009). Needed
now are methods to model the evolution of indels and allow their
frequency in relation to substitutional changes to be estimated. It
seems clear from our data (Fig. 2) and other studies (Johnson,
2004; Mathee et al., 2007) that useful models with relatively few
parameters can be devised to estimate indel evolution on fixed
alignments of moderately sized datasets. This would allow indels
to be incorporated directly in maximum likelihood or Bayesian tree
searches, fully utilizing the information content of non-coding
sequences.
4.2. Phylogeny and systematics—deep nodes
Among the deeper nodes in the phylogeny, all traditional fami-
lies and orders, with the exception of Caprimulgiformes, were
recovered with strong support in the combined analyses. Most
were also supported by one or more indel synapomorphies. Thus,
all of these taxa are likely to represent monophyletic clades,
reflecting traditional notions of avian relationships.
The novel association of Aegothelidae with Apodiformes re-
ported here has been found by the majority of recent relevant stud-
ies. With regard to morphological evidence, Mayr (2002) and Mayr
et al. (2003) found this clade based on morphological data sets of
up to 89 characters and 29 taxa, and highlighted the cruciform ori-
gin of the splenius capitis muscle as an ‘‘unambiguous synapomor-
phy” for the group (Mayr et al., 2003). Burton (1971) first reported
the shared cruciform origin of the splenius capitis muscle, but did
not consider it evidence of relationship, doubtlessly because that
would have seemed a radical proposal at the time. The extinct
Aegialornithidae may well represent a link between Apodiformes
and Caprimulgiformes (Collins, 1976), particularly Aegothelidae
(Mayr, 2002). Indeed, Collins’ proposition that the Aegialornithidae
are ‘‘representatives of a caprimulgiform lineage that later gave
rise to the swifts” now seems prescient.
On the other hand, Livezey and Zusi (2007), analysing a mor-
phological dataset of 2954 characters scored for 150 bird lineages
in a parsimony framework, found the Apodiformes to be sister to
the traditional order Caprimulgiformes, including Aegothelidae.
The discrepancy may be a due to a rooting issue; because Aegothe-
lidae is basal in Livezey and Zusi’s Caprimulgiformes, shifting the
root by one node will produce an aegothelid–apodiform clade.
However, Mayr (2008) also criticized two of Livezey and Zusi’s five
‘‘diagnostic apomorphies” for Caprimulgiformes, namely beak
morphology and presence of the tapetum lucidum. Mayr
(2008:66) found it ‘‘incomprehensible” that Livezey and Zusi
assigned the same character state for beak morphology toPodargidae, Steatornithidae, Caprimulgidae and Aegothelidae, but
a different one to Apodidae and Hemiprocnidae, whose beaks are
‘‘extremely similar” to those of Aegothelidae. Mayr also found
Livezey and Zusi’s coding of the tapetum lucidum as present for
all members of Caprimulgiformes an ‘‘unacceptable generaliza-
tion”, citing the lack of firm evidence. In fact, after detailed study,
Martin et al. (2004) found ‘‘no evidence of a tapetum in either the
retina or choroid” of oilbird eyes. The tapetum is a reflective struc-
ture responsible for brilliant eyeshine and increased sensitivity to
light in caprimulgids and many other nocturnal vertebrates (Nicol
and Arnott, 1974; Ollivier et al., 2004). We have not found a clear
statement in the literature on whether podargids and aegothelids
have eyeshine or a tapetum of any kind, but potoos must certainly
have a reflective structure responsible for their brilliant orange
eyeshine, which can be seen at great distances (van Rossem,
1927; Braun, pers. obs.).
The molecular evidence for an aegothelid–apodiform clade has
been consistent, and can now be considered nearly unassailable,
having been found in numerous nuclear genes, a mitochondrial
gene and indels. While the relationship was not detected in
DNA hybridization studies (Sibley and Ahlquist, 1990), it is clear
in retrospect that it was beyond the resolving power of that
technique. Support for this clade has been reported in substitu-
tional variation of at least 16 nuclear genes (Mayr et al., 2003;
Cracraft et al., 2004; Ericson et al., 2006; Barrowclough et al.,
2006; Hackett et al., 2008) and one mitochondrial gene (cyto-
chrome b; this study). In general, these studies have found weak
to moderate support in analyses of individual genes. Support val-
ues become truly convincing when multiple genes are analysed
in an ML framework (Hackett et al., 2008). The possibility that
base compositional artifacts could produce this grouping was
specifically examined and rejected by Barrowclough et al.
(2006) and in the present study.
Finally, indel support for an aegothelid–apodiform clade is
found in the 4 codon c-myc insertion discussed here and previously
(Braun and Huddleston, 2001; Mayr et al., 2003; Cracraft et al.,
2004), and a 5 codon deletion in RAG-1 (Barrowclough et al.,
2006). The concordance of signal for this clade from nuclear se-
quence data, mitochondrial sequence data and indels is reassuring,
as these are essentially independent lines of evidence.
An aegothelid–apodiform clade renders the traditional order
Caprimulgiformes non-monophyletic. But would a broader group-
ing, including all the traditional caprimulgiform and apodiform
families, be monophyletic? Livezey and Zusi (2007) found morpho-
logical evidence for such a clade, although Mayr et al. (2003) did
not. Most molecular studies to date have not produced convincing
support for or against this grouping (Sibley and Ahlquist, 1990;
Mayr et al., 2003; Cracraft et al., 2004; Fain and Houde, 2004;
Barrowclough et al., 2006; this study). Ericson et al. (2006) did re-
cover this clade, but did not address the possibility of inflation of
its posterior probability, a common problem in Bayesian analyses
(Alfaro and Holder, 2006). Hackett et al. (2008) found strong sup-
port for a caprimulgiform + apodiform clade in combined ML anal-
yses of 19 genes, but that clade was present in only 3 of 19 single
gene analyses. Thus, while these results are promising, a rigorous
examination of this node to exclude potential artifacts, such as
long branch attraction, base compositional bias, and gene tree
problems is in order (e.g., Harshman et al., 2008).
The taxonomic implications of the aegothelid + apodiform clade
were discussed by Sangster (2005) and Barrowclough et al. (2006).
Two possibilities are to (1) transfer Aegothelidae to Apodiformes,
or (2) recognize a supraordinal taxon that includes swifts, hum-
mingbirds and owlet-nightjars. However, the probable monophyly
of a grouping of all five traditional caprimulgiform families plus
Apodiformes is pertinent to this discussion. If that larger clade is
indeed monophyletic, expanding Caprimulgiformes to include
958 M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960swifts and hummingbirds is a third possibility. More molecular
data relevant to this question are likely to be forthcoming, and it
seems prudent to delay taxonomic changes until sufficient data
are available to make stable ones.
The association of Steatornis with Nyctibiidae seen here was
also found by Hackett et al. (2008). Although the signal for it is
not strong, this grouping deserves further scrutiny. It is interesting
biogeographically, given that the current ranges of both taxa are
restricted to the Neotropics. However, both taxa apparently
occurred more widely in the past (Mourer-Chauviré, 1982, 1987;
Olson, 1987).
Livezey and Zusi (2007) proposed the new suborder Hemiprocni
on the basis of their morphological analysis suggesting that tree
swifts are sister to a swift + hummingbird clade. Both our c-myc
and cytochrome b sequence data reject that notion, and a 1 bp in-
del in c-myc also conflicts with it. In fact, all molecular evidence to
date strongly supports the traditional idea that swifts are sister to
treeswifts to the exclusion of hummingbirds (Sibley and Ahlquist,
1990; Chubb, 2004; Barrowclough et al., 2006; Ericson et al.,
2006; Hackett et al., 2008).
4.3. Phylogeny and systematics—shallow nodes
At the family level, Sibley and Ahlquist (1990) recommended
recognition of new families for Eurostopodus (=Eurostopodidae)
and Batrachostomus (=Batrachostomidae) based on large DNA
hybridization distances to other genera in Caprimulgidae and
Podargidae, respectively. While that suggestion received support
from mtDNA distance data (Mariaux and Braun, 1996), it is now
complicated by the discovery of additional deep nodes in these
families. The deep divergence within Eurostopodus reported here
would require recognition of another new family if Sibley and
Ahlquist’s (1990) suggestion were followed. It would seem more
reasonable to recognize the genus Lyncornis (to include at least
the eared taxa macrotis and temmincki) as suggested by Cleere
(1998, p.174), and retain both Eurostopodus and Lyncornis in
Caprimulgidae. These taxa surely form a monophyletic group with
other caprimulgids (Barrowclough et al., 2006; this study) despite
the relatively deep molecular divergences among them. Similarly,
the description of a new frogmouth genus, Rigidapenna, with large
molecular divergences from both Podargus and Batrachostomus
(Cleere et al., 2007), would require erection of another new family
if Batrachostomidae were recognized. Because all these taxa
are frogmouth-like in body form and habits, it seems more useful
to retain them in Podargidae, which is surely monophyletic
(Barrowclough et al., 2006; Cleere et al., 2007; this study).
Within Caprimulgidae, our data are consistent with previous
evidence that several long-standing taxa are probably not mono-
phyletic. These include the genera Caprimulgus and Eurostopodus,
and the subfamilies Caprimulginae and Chordeilinae (Sibley and
Ahlquist, 1990; Mariaux and Braun, 1996; Barrowclough et al.,
2006; Larsen et al., 2007). However, it is not yet possible to propose
a comprehensive revision of the family; this will require more
extensive taxon sampling and probably more data, preferably from
other genes (Barrowclough et al., 2006).
Within Nyctibiidae, our data confirm and extend the conclusions
of Mariaux and Braun (1996) based on 656 bp of cytochrome b and
Brumfield et al. (1997) based on isozymes. Two pairs of small gray
potoos (griseus + jamaicensis, leucopterus +maculosus) belong to a
well-supported clade. The larger aethereus and grandis, and the
small rufous bracteatus are all more divergent, but their relation-
ships are not well-resolved by the present data. The large genetic
divergences among these taxa suggest that one or more new
generic designations may be in order, but we prefer to wait until
the phylogeny is well-resolved, so as to propose a classification that
reflects their evolutionary history as accurately as possible.The conflict between cytochrome b and c-myc on hummingbird
relationships is intriguing. Each gene provides fairly strong support
for one of two mutually exclusive groupings. However, neither of
these groupings conforms to traditional notions of hummingbird
relationships, as Glaucis is usually placed in a separate subfamily,
Phaethorninae, with a few genera of hermits, while all other hum-
mingbirds are placed in Trochilinae. Interestingly, the topology
supported by c-myc is the one found in all combined analyses, even
though the single gene support values were higher for the cyto-
chrome b topology. We suspect that the cytochrome b support val-
ues may be inflated by the base compositional heterogeneity found
in apodiform cytochrome b. The c-myc topology, with Topaza and
relatives basal to all other hummingbirds, was also found in a re-
cent analysis involving two mitochondrial genes, ND2 and ND4,
and two nuclear genes, adenylate kinase 1 and beta-fibrinogen
(McGuire et al., 2007).
4.4. Evolution of adaptive traits—nocturnality
Was the common ancestor of Caprimulgiformes + Apodiformes
nocturnal? The exact topology of an expanded clade consisting of
Caprimulgiformes + Apodiformes has important implications for
the evolutionary history of nocturnal adaptation. While nocturnal
activity is found in many groups of birds, only a few major lineages
have adopted a largely nocturnal lifestyle—caprimulgiforms, owls
and kiwis are the principal examples. The finding that Aegotheli-
dae is the sister group of Apodiformes suggests that the diurnal
swifts and hummingbirds, with their remarkable powers of flight,
may be derived from a nocturnal ancestor. This suggestion is rein-
forced by the apparent monophyly of Caprimulgiformes + Apodi-
formes (Ericson et al., 2006; Hackett et al., 2008). This scenario
implies that many adaptations to nocturnality were gained in their
common ancestor and subsequently lost in Apodiformes. However,
it is important to note that the basal branching structure of the
Caprimulgiformes + Apodiformes is not yet well-resolved. If the
root of this clade actually falls between Aegothelidae + Apodifor-
mes and the rest of the caprimulgiform families, then it is equally
parsimonious to suppose that there were two independent transi-
tions to nocturnality rather than a loss of nocturnality by Apodifor-
mes. More sequence data are needed to resolve the basal branching
structure and clarify this issue.
Whether the transition to nocturnality occurred once or multi-
ple times in an expanded Caprimulgiformes is particularly relevant
in light of frequent observations of their morphological and genetic
diversity, especially in traits that may be adaptive for nocturnal
activity. For example, the tapetum lucidum, as mentioned above,
is apparently absent in oilbirds (Martin et al., 2004). Instead, in-
creased photosensitivity is achieved by a unique banked structure
of the retina, in which rod photoreceptors are stacked three layers
deep. Reflective structures (tapeta) in the eyes of nocturnal verte-
brates are morphologically diverse and probably evolved indepen-
dently in a number of lineages (Ollivier et al., 2004). Thus, whether
the eyeshine of potoos is produced by a structure homologous to
that of caprimulgids (Nicol and Arnott, 1974), and whether frog-
mouths and owlet-nightjars even have tapeta are open questions.
There is also considerable variation in the morphology of the
cerebellum, telencephalon and Wulst, brain structures related to
behavior and vision (Iwaniuk and Hurd, 2005; Iwaniuk et al.,
2006; Iwaniuk and Wylie, 2006). The owlet-nightjar sequence for
arylalkylamine N-acetyltransferase, a gene potentially associated
with nocturnal activity, fails to form a clade with other caprimulg-
iform sequences in phylogenetic analyses (Fidler et al., 2004).
Although found in no other bird, echolocation has arisen twice in
this group, in oilbirds and Aerodramus swiftlets, as an adaptation
to nesting in caves (Lee et al., 1996). Considerable variation in
the structure of the main arteries near the heart has also been
M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960 959noted (Glenny, 1953). Understanding of the evolutionary history of
all these traits will be improved by a well-resolved phylogeny.
4.5. Evolution of adaptive traits—hypothermia
Hypothermic responses are well-known in both caprimulgiform
and apodiform birds. The literature on facultative hypothermia in
birds was reviewed by McKechnie and Lovegrove (2002). Data ex-
isted for 29 families in 11 orders at the time. While the capacity for
shallow to moderate hypothermia is widespread, extreme hypo-
thermia, during which body temperature may be reduced by
20 C or more, was reported in only three families: Apodidae, Troc-
hilidae, and Caprimulgidae. Current evidence indicates that these
families all belong to a single clade, Caprimulgiformes + Apodifor-
mes (Ericson et al., 2006; Hackett et al., 2008). Thus, extreme hypo-
thermia, which is associated with daily torpor and/or hibernation,
seems to be a derived trait of this group. Most families of this clade
are known to have the capacity for at least moderate hypothermia
(reduction in body temperature of >10 C). Telemetric studies of
free ranging birds indicate that facultative hypothermia is used
routinely in Aegothelidae, Podargidae, and some Caprimulgidae
(Brigham et al., 2006).
Acknowledgments
We thank T. Glenn for assistance in the laboratory, R. Brumfield,
J. Harshman, K. Karol, D. Swofford, and J. Wilgenbusch for advice on
phylogenetic analyses, and G. Barrowclough, N. Cleere, J. Cracraft,
and M. Robbins for useful discussion of caprimulgiform systemat-
ics. R. Zusi alerted us to Burton’s work on the splenius capitus mus-
cle, and C. Collins pointed us to the literature on Aegialornithidae.
The institutions and collectors listed in Table 1 provided tissue
samples, and L. Christidis, J. Dean, D. Dittmann, L. Joseph,
R. Kennedy, J. Norman, M. Robbins, and F. Sheldon facilitated their
transfer. This work was supported in part by the US National
Science Foundation Assembling the Tree of Life Program Grant
DEB-0228675.Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ympev.2009.08.025.
References
Alfaro, M.E., Holder, M.T., 2006. The posterior and the prior in Bayesian
phylogenetics. Annu. Rev. Ecol. Syst. 37, 19–42.
Barrowclough, G.F., Groth, J.G., Mertz, L.A., 2006. The RAG-1 exon in the avian order
Caprimulgiformes: phylogeny, heterozygosity, and base composition. Mol.
Phylogenet. Evol. 41, 238–248.
Beddard, F.E., 1886. On the syrinx and other points in the anatomy of the
Caprimulgidae. Proc. Zool. Soc. London, XI, 147–153.
Braun, E.L., Kimball, R.T., 2002. Examining basal avian divergences with
mitochondrial sequences: model complexity, taxon sampling, and sequence
length. Syst. Biol. 51, 614–625.
Braun, M.J., Huddleston, C.J., 2001. Molecular phylogenetics of caprimulgiform
nightbirds. In: Abstracts of 119th Stated Meeting of the American
Ornithologists’ Union, Seattle, WA, p. 47.
Brigham, R.M., Woods, C.P., Lane, J.E., Fletcher, Q.E., Geiser, F., 2006. Ecological
correlates of torpor use among five caprimulgiform birds. Acta Zoolog. Sin. 52
(Suppl.), 401–404.
Brumfield, R.T., Swofford, D.L., Braun, M.J., 1997. Evolutionary relationships among
the potoos (Nyctibiidae) based on isozymes. Ornithol. Monogr. 48, 129–145.
Bull, J.J., Huelsenbeck, J.P., Cunningham, C.W., Swofford, D.L., Waddell, P.J., 1993.
Partitioning and combining data in phylogenetic analysis. Syst. Biol. 42, 384–
397.
Burton, P.J.K., 1971. Some observations on the splenius capitis muscle of birds. Ibis
113, 19–28.
Chojnowski, J.L., Kimball, R.T., Braun, E.L., 2008. Introns outperform exons in
analyses of basal avian phylogeny using clathrin heavy chain genes. Gene 410,
89–96.Chubb, A.L., 2004. Nuclear corroboration of DNA–DNA hybridization in deep
phylogenies of hummingbirds, swifts, and passerines: the phylogenetic utility
of ZENK (ii). Mol. Phylogenet. Evol. 30, 128–139.
Cleere, N., 2002. A review of the taxonomy and systematics of the Sickle-winged
and White-winged Nightjars (Caprimulgidae). Bull. Br. Ornithol. Club 122, 168–
179.
Cleere, N., 1998. Nightjars: A Guide to Nightjars, Nighthawks and their Relatives.
Yale University Press, New Haven.
Cleere, N., 1999. Family Caprimulgidae (Nightjars). In: del Hoyo, J., Elliott, A.,
Sargatal, J. (Eds.), Handbook of the Birds of the World. Lynx Edicions, Barcelona,
pp. 302–386.
Cleere, N., Kratter, A.W., Steadman, D.W., Braun, M.J., Huddleston, C.J., Filardi, C.E.,
2007. A new genus of frogmouth (Podargidae) from the Solomon Islands—
results from a taxonomic review of Podargus ocellatus inexpectatus Hartert 1901.
Ibis 149, 271–286.
Collins, C.T., 1976. Two new species of Aegialornis from France, with comments on
the ordinal affinities of the Aegialornithidae. Smithson. Contrib. Paleobiol. 27,
121–127.
Cohn-Haft, M., 1993. Rediscovery of the white-winged potoo (Nyctibius leucopterus).
Auk 110, 391–394.
Cohn-Haft, M., 1999. Family Nyctibiidae (potoos). In: del Hoyo, J., Elliott, A.,
Sargatal, J. (Eds.), Handbook of the Birds of the World. Lynx Edicions, Barcelona,
pp. 288–301.
Cracraft, J., Barker, F.K., Braun, M.J., Harshman, J., Dyke, G., Feinstein, J., Stanley, S.,
Cibois, A., Schikler, P., Beresford, P., García-Moreno, J., Sorenson, M.D., Yuri, T.,
Mindell, D.P., 2004. Phylogenetic relationships among modern birds
(Neornithes): toward an avian tree of life. In: Cracraft, J., Donoghue, M.J.
(Eds.), Assembling the Tree of Life. Oxford University Press, New York, pp. 468–
489.
Davis, L.I., 1978. Acoustic evidence of relationship in Caprimulginae. Pan Am.
Studies 14, 22–57.
Davis, L.I., 1962. Acoustic evidence of relationship in Caprimulgus. Tex. J. Sci. 14, 72–
106.
Desjardins, P., Morais, R., 1990. Sequence and gene organization of the chicken
mitochondrial genome—a novel gene order in higher vertebrates. J. Mol. Biol.
212, 599–634.
Dickinson, E.C. (Ed.), 2003. The Howard and Moore Complete Checklist of the Birds
of the World, third ed. Princeton University Press, Princeton, NJ.
Dumbacher, J.P., Pratt, T.K., Fleischer, R.C., 2003. Phylogeny of the owlet-nightjars
(Aves: Aegothelidae) based on mitochondrial DNA sequence. Mol. Phylogenet.
Evol. 29, 540–549.
Ericson, P.G.P., Anderson, C.L., Britton, T., Elzanowski, A., Johansson, U.S., Källersjö,
M., Ohlson, J.I., Parsons, T.J., Zuccon, D., Mayr, G., 2006. Diversification of
Neoaves: integration of molecular sequence data and fossils. Biol. Lett. 2, 543–
547.
Fain, M.G., Houde, P., 2004. Parallel radiations in the primary clades of birds.
Evolution 58, 2558–2573.
Farris, J.S., Källersjö, M., Kluge, A.G., Bult, C., 1995. Testing significance of
incongruence. Cladistics 10, 315–319.
Fidler, A.E., Kuhn, S., Gwinner, E., 2004. Convergent evolution of Strigiform and
Caprimulgiform dark-activity is supported by phylogenetic analysis using the
arylalkylamine N-acetyltransferase (Aanat) gene. Mol. Phylogenet. Evol. 33,
908–921.
Fry, C.H., 1988. Skulls, songs and systematics of African nightjars. Pan-African
Ornithol. Congr. 6, 105–131.
Glenny, F.H., 1953. A systematic study of the main arteries in the region of the heart
Aves XX. Caprimulgiformes. Pt. 1. Ohio J. Sci. 53, 356–357.
Hackett, S.J., Kimball, R.T., Reddy, S., Bowie, R.C.K., Braun, E.L., Braun, M.J.,
Chojnowski, J.L., Cox, W.A., Han, K.-L., Harshman, J., Huddleston, C.J., Marks,
B.D., Miglia, K.J., Moore, W.S., Sheldon, F.H., Steadman, D.W., Witt, C.C., Yuri, T.,
2008. A phylogenomic study of birds reveals their evolutionary history. Science
320, 1763–1768.
Harshman, J., Huddleston, C.J., Bollback, J.P., Parsons, T.J., Braun, M.J., 2003. True and
false gharials: a nuclear gene phylogeny of Crocodylia. Syst. Biol. 52, 386–402.
Harshman, J., Braun, E.L., Braun, M.J., Huddleston, C.J., Bowie, R.C.K., Chojnowski, J.L.,
Hackett, S.J., Han, K.L., Kimball, R.T., Marks, B.D., Miglia, K.J., Moore, W.S., Reddy,
S., Sheldon, F.H., Steadman, D.W., Steppan, S.J., Witt, C.C., Yuri, T., 2008.
Phylogenomic evidence for multiple losses of flight in ratite birds. Proc. Natl.
Acad. Sci. USA 105, 13462–13467.
Hasegawa, M., Kishino, H., Yano, T.A., 1985. Dating of the human-ape splitting by a
molecular clock of mitochondrial DNA. J. Mol. Evol. 22, 160–174.
Hochberg, Y., 1988. A sharper Bonferroni procedure for multiple tests of
significance. Biometrika 75, 800–802.
Hoff, K., 1966. A Comparative Study of the Appendicular Muscles of Strigiformes
and Caprimulgiformes. Unpubl. PhD. Thesis, Washington State University,
Pullman, WA.
Holyoak, D.T., 2001. Nightjars and their Allies: The Caprimulgiformes. Oxford
University Press, New York.
Holyoak, D.T., 1999a. Family Aegothelidae (owlet-nightjars). In: del Hoyo, J., Elliott,
A., Sargatal, J. (Eds.), Handbook of the Birds of the World. Lynx Edicions,
Barcelona, pp. 252–265.
Holyoak, D.T., 1999b. Family Podargidae (frogmouths). In: del Hoyo, J., Elliott, A.,
Sargatal, J. (Eds.), Handbook of the Birds of the World. Lynx Edicions, Barcelona,
pp. 266–287.
Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: Bayesian inference of phylogeny.
Bioinformatics 17, 754–755.
960 M.J. Braun, C.J. Huddleston /Molecular Phylogenetics and Evolution 53 (2009) 948–960Iwaniuk, A.N., Hurd, P.L., 2005. A multivariate analysis of cerebrotypes in birds.
Brain Behav. Evol. 65, 215–230.
Iwaniuk, A.N., Hurd, P.L., Wylie, D.R.W., 2006. The comparative morphology of the
cerebellum in Caprimulgiform birds: evolutionary and functional implications.
Brain Behav. Evol. 67, 53–68.
Iwaniuk, A.N., Wylie, D.R.W., 2006. The evolution of stereopsis and the Wulst in
Caprimulgiform birds: a comparative analysis. J. Comp. Physiol. A Neuroethol.
Sens. Neural Behav. Physiol. 192, 1313–1326.
James, H.F., 2005. Paleogene fossils and the radiation of modern birds. Auk 122,
1049–1054.
Johansson, U.S., Parsons, T.J., Irestedt, M., Ericson, P.G.P., 2001. Clades within the
‘higher land birds’, evaluated by nuclear DNA sequences. J. Zoolog. Syst. Evol.
Res. 39, 37–51.
Johnson, K.P., 2004. Deletion bias in avian introns over evolutionary timescales.
Mol. Biol. Evol. 21, 599–602.
Jukes, T.H., Cantor, C.R., 1969. Evolution of protein molecules. In: Munro, H.N. (Ed.),
Mammalian Protein Metabolism. Academic Press, New York, pp. 21–132.
Katoh, K., Kuma, K.-I., Toh, H., Miyata, T., 2005. MAFFT version 5: improvement in
accuracy of multiple sequence alignment. Nucleic Acids Res. 33, 511–518.
Kocher, T.D., Thomas, W.K., Meyer, A., Edwards, S.V., Pååbo, S., Villablanca, F.X.,
Wilson, A.C., 1989. Dynamics of mitochondrial DNA evolution in animals:
amplification and sequencing with conserved primers. Proc. Natl. Acad. Sci. USA
86, 6196–6200.
Kubatko, L.S., Degnan, J.H., 2007. Inconsistency of phylogenetic estimates from
concatenated data under coalescence. Syst. Biol. 56, 17–24.
Larsen, C., Speed, M., Harvey, N., Noyes, H.A., 2007. A molecular phylogeny of the
Nightjars (Aves: Caprimulgidae) suggests extensive conservation of primitive
morphological traits across multiple lineages. Mol. Phylogenet. Evol. 42, 789–
796.
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. Proc. Natl. Acad. Sci. USA 93, 7091–7096.
Lencioni-Neto, F., 1994. Une nouvelle espéce de Chordeiles (Aves, Caprimulgidae)
de Bahia (Brésil). Alauda 62, 242–245.
Liu, K., Raghavan, S., Nelesen, S., Linder, C.R., Warnow, T., 2009. Rapid and accurate
large-scale coestimation of sequence alignments and phylogenetic trees.
Science 324, 1561–1564.
Livezey, B.C., Zusi, R.L., 2007. Higher-order phylogeny of modern birds (Theropoda,
Aves: Neornithes) based on comparative anatomy. II. Analysis and discussion.
Zoolog. J. Linn. Soc. 149, 1–95.
Loytynoja, A., Goldman, N., 2008. Phylogeny-aware gap placement prevents errors
in sequence alignment and evolutionary analysis. Science 320, 1632–1635.
Louette, M., 1990. A new species of nightjar from Zaire. Ibis 132, 349–353.
Maddison, W.P., 1997. Gene trees in species trees. Syst. Biol. 46, 523–536.
Mariaux, J., Braun, M.J., 1996. A molecular phylogenetic survey of the Nightjars and
Allies (Caprimulgiformes) with special emphasis on the potoos (Nyctibiidae).
Mol. Phylogenet. Evol. 6, 228–244.
Martin, G., Rojas, L.M., Ramirez, Y., McNeil, R., 2004. The eyes of oilbirds (Steatornis
caripensis): pushing at the limits of sensitivity. Naturwissenschaften 91, 26–29.
Mathee, C.A., Eick, G., Willows-Munro, S., Montgelard, C., Pardini, A.T., Robinson, T.J.,
2007. Indel evolution of mammalian introns and the utility of non-coding
nuclear markers in eutherian phylogenetics. Mol. Phylogenet. Evol. 42, 827–
837.
Mayr, G., 2002. Osteological evidence for paraphyly of the avian order
Caprimulgiformes (Nightjars and Allies). J. Fur Ornithol. 143, 82–97.
Mayr, G., 2008. Avian higher level phylogeny: well supported clades and what we
can learn from a phylogenetic analysis of 2954 morphological characters. J.
Zool. Syst. Evol. Res. 46, 63–72.
Mayr, G., Manegold, A., Johansson, U.S., 2003. Monophyletic groups within ‘higher
land birds’—comparison of morphological and molecular data. J. Zoolog. Syst.
Evol. Res. 41, 233–248.
McGuire, J.A., Witt, C.C., Altshuler, D.L., Remsen Jr., J.V., 2007. Phylogenetic
systematics and biogeography of hummingbirds: Bayesian and maximum
likelihood analyses of partitioned data and selection of an appropriate
partitioning strategy. Syst. Biol. 56, 837–856.
McKechnie, A.E., Lovegrove, B.G., 2002. Avian facultative hypothermic responses: a
review. Condor 104, 705–724.
Mourer-Chauviré, C., 1982. Les oiesaux fossiles des Phosphorites du Quercy (Éocène
supérieur è Oligocène supérieur): implications paléobiogéographiques. Geobios
Mem. Special 6, 413–426.
Mourer-Chauviré, C., 1987. Les Strigiformes (Aves) des Phosphorites du Quercy
(France): systematique, biostratigraphie et paleobiogeographie. Docum. Lab.
Géol. Lyon 99, 89–127.
Nicol, J.A.C., Arnott, H.J., 1974. Tapeta lucida in the eyes of goatsuckers
(Caprimulgidae). Proc. R. Soc. Lond. Ser. B 187, 349–352.
Oberholser, H.C., 1914. A monograph of the genus Chordeiles Swainson, type of a
new family of goatsuckers. Bull. US Nat. Mus. 86, 1–123.Ollivier, F.J., Samuelson, D.A., Brooks, D.E., Lewis, P.A., Kallberg, M.E., Komáromy,
A.M., 2004. Comparative morphology of the tapetum lucidum (among selected
species). Vet. Ophthalmol. 7, 11–22.
Olson, S.L., 1987. An early Eocene oilbird from the Green River formation of
Wyoming (Caprimulgiformes: Steatornithidae). Docum. Lab. Géol. Lyon 99, 57–
69.
Peters, D.S., 1991. Zoogeographical relationships of the Eocene avifauna from
Messel (Germany). In: Bell, B.D., Cossee, R.O., Flux, J.E.C., Heather, B.D.,
Hitchmough, R.A., Robertson, C.J.R., Williams, M.J. (Eds.), Acta XX Congressus
Internationalis Ornithologici. New Zealand Ornithological Congress Trust Board,
Christchurch, pp. 572–577.
Peters, J.L., 1940. Check-list of Birds of the World. Museum of Comparative Zoology,
vol. 4. Harvard University Press, Cambridge, Massachusetts.
Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNA substitution.
Bioinformatics 14, 817–818.
Pratt, T.K., 2000. Evidence for a previously unrecognized species of owlet-nightjar.
Auk 117, 1–11.
Prychitko, T.M., Moore, W.S., 2003. Alignment and phylogenetic analysis of beta
fibrinogen intron 7 sequences among avian orders reveal conserved regions
within the intron. Mol. Biol. Evol. 20, 762–771.
Robbins, M.B., Braun, M.J., Huddleston, C.J., Finch, D.W., Milensky, C.W., 2005. First
Guyana records, natural history, and systematics of the White-naped Seedeater
(Dolospingus fringilloides). Ibis 147, 334–341.
Robbins, M.B., Parker, T.A. 1997. Voice and taxonomy of Caprimulgus (rufus) otiosus
(Caprumulgidae), with a reevaluation of Caprimulgus rufus subspecies. In:
Remsen, J.V. (Ed.), Studies in Neotropical Ornithology honoring Ted Parker.
Ornithol. Monogr. pp. 601–608.
Robbins, M.B., Ridgely, R.S., Cardiff, S.W., 1994. Voice, plumage and natural-history
of Anthony’s Nightjar (Caprimulgus anthonyi). Condor 96, 224–228.
Rokas, A., Carroll, S.B., 2005. More genes or more taxa? The relative contribution of
gene number and taxon number to phylogenetic accuracy. Mol. Biol. Evol. 22,
1337–1344.
Rokas, A., Holland, P.W.H., 2000. Rare genomic changes as a tool for phylogenetics.
Trends Ecol. Evol. 15, 454–459.
Ronquist, F., Huelsenbeck, J.P., 2003. MRBAYES 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19, 1572–1574.
Safford, R.J., Ash, J.S., Duckworth, J.W., Telfer, M.G., Zewdie, C., 1995. A new species
of Nightjar from Ethiopia. Ibis 137, 301–307.
Sangster, G., 2005. A name for the clade formed by owlet-nightjars, swifts and
hummingbirds (Aves). Zootaxa, 1–6.
Sangster, G., Rozendaal, F.G., 2004. Territorial songs and species-level taxonomy of
nightjars of the Caprimulgus macrurus complex, with description of a new
species. In: Dekker, R.W.R.J., Dickinson, R.J. (Eds.), Systematic Notes on Asian
Birds. Zool. Verhand., Leiden, pp. 7–45.
Schulenberg, T.S., Allen, S.E., Stotz, D.F., Wiedenfield, D.A., 1984. Distributional
records from the Cordillera Yanachaga, Central Peru. Le Gerfaut 74, 57–70.
Sclater, P.L., 1866a. Notes upon the American Caprimulgidae. Proc. Zool. Soc.
London, 123–145.
Sclater, P.L., 1866b. Additional notes on the Caprimulgidae. Proc. Zool. Soc. London,
581–590.
Sibley, C.G., Ahlquist, J.E., 1990. Phylogeny and Classification of Birds. A Study in
Molecular Evolution. Yale University Press, New Haven, CT.
Sorenson, M.D., Ast, J.C., Dimcheff, D.E., Yuri, T., Mindell, D.P., 1999. Primers for a
PCR-based approach to mitochondrial genome sequencing in birds and other
vertebrates. Mol. Phylogenet. Evol. 12, 105–114.
Sullivan, J., Abdo, Z., Joyce, P., Swofford, D.L., 1995. Evaluating the performance of a
successive- approximations approach to parameter optimization in maximum-
likelihood phylogeny estimation. Mol. Biol. Evol. 22, 1386–1392.
Swofford, D.L., 2002. PAUP*: Phylogenetic Analysis Using Parsimony ( and Other
Methods). Sinauer Associates, Sunderland, MA.
Swofford, D.L., Olsen, G.J., Waddell, P.J., Hillis, D.M., 1996. Phylogenetic inference.
In: Hillis, D.M., Moritz, C., Mable, B. (Eds.), Molecular Systematics. Sinauer
Associates, Sunderland, MA, pp. 407–514.
Thomas, B.T., 1999. Family Steatornithidae (Oilbird). In: del Hoyo, J., Elliott, A.,
Sargatal, J. (Eds.), Handbook of the Birds of the World. Lynx Edicions, Barcelona,
pp. 244–251.
Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL-W—improving the
sensitivity of progressive multiple sequence alignment through sequence
weighting, position-specific gap penalties and weight matrix choice. Nucleic
Acids Res. 22, 4673–4680.
Tripepi, S., Jamieson, B.G.M., Brunelli, E., 2006. Ultrastructure of the spermatid of
Caprimulgus europaeus Linnaeus 1758, the European nightjar (Aves;
Caprimulgidae), with phylogenetic implications. J. Morph. 267, 1157–1164.
van Rossem, A.J., 1927. Eyeshine in birds, with notes on the feeding habits of some
goatsuckers. Condor 29, 25–28.
Wetmore, A., 1918. On the anatomy of Nyctibius with notes on allied birds. Proc. US
Nat. Mus. 54, 577–586.