ArticlePDF Available

Convergence and divergence in the evolution of aquatic birds

Authors:

Abstract and Figures

Aquatic birds exceed other terrestrial vertebrates in the diversity of their adaptations to aquatic niches. For many species this has created difficulty in understanding their evolutionary origin and, in particular, for the flamingos, hamerkop, shoebill and pelecaniforms. Here, new evidence from nuclear and mitochondrial DNA sequences and DNA-DNA hybridization data indicates extensive morphological convergence and divergence in aquatic birds. Among the unexpected findings is a grouping of flamingos and grebes, species which otherwise show no resemblance. These results suggest that the traditional characters used to unite certain aquatic groups, such as totipalmate feet, foot-propelled diving and long legs, evolved more than once and that organismal change in aquatic birds has proceeded at a faster pace than previously recognized.
Content may be subject to copyright.
Convergence and divergence in the evolution
of aquatic birds
Marcel Van Tuinen1{, Dave Brian Butvill2,JohnA.W.Kirsch
2and S. Blair Hedges1*
1Department of Biology, Institute of Molecular Evolutionary Genetics and Astrobiology Research Center, 208 Mueller Laboratory,
The Pennsylvania State University, University Park, PA 16802, USA
2The University of Wisconsin Zoological Museum, 250 North Mills Street, Madison, WI 5370 6, USA
Aquatic birds exceed other terrestrial vertebrates in the diversity of their adaptations to aquatic niches.
For many species this has created di¤culty in understanding their evolutionary origin and, in particular,
for the £amingos, hamerkop, shoebill and pelecaniforms. Here, new evidence from nuclear and
mitochondrial DNA sequences and DNA^DNA hybridization data indicates extensive morphological
convergence and divergence in aquatic birds. Among the unexpected ¢ndings is a grouping of £amingos
and grebes, species which otherwise show no resemblance. These results suggest that the traditional
characters used to unite certain aquatic groups, such as totipalmate feet, foot-propelled diving and long
legs, evolved more than once and that organismal change in aquatic birds has proceeded at a faster pace
than previously recognized.
Keywords: phylogeny; £amingo; grebe; avian; DNA sequence; DNA^DNA hybridization
1. INTRODUCTION
Many adaptations in birds to life around water are
related to feeding style (Storer 1971). Examples include
the traditional pelecaniforms with their webbing around
all four toes (totipalmate feet), loons and grebes with
their posteriorly positioned legs for diving and storks and
allies with their long legs adapted for wading. Tradition-
ally, such morphologically distinct groups have been
given taxonomic and evolutionary status (Cracraft 1981,
1988; Feduccia 1996) (¢gure 1). Although £amingos have
most often been placed with other long-legged waders
(Cracraft 1981, 1988; Sibley & Ahlquist 1990), some
characters have suggested an a¤nity with ducks (Sibley
& Ahlquist 1990; Feduccia 1996) or with shorebirds
(Olson & Feduccia 1980). Furthermore, the ¢ve living
species of £amingos show many unique characters related
to their unusual ¢lter-feeding lifestyle in tropical saline
waters (Olson & Feduccia 1980; Zweers et al. 1994). The
shoebill and hamerkop are two additional enigmatic
species, which show a blend of morphological characters
shared with either waders or non-waders (Sibley &
Ahlquist 1990; Feduccia 1996). For these reasons, close
relationships of these enigmatic birds to other long-legged
wading birds have remained tenuous at best.
Among molecular phylogenetic investigations of aquatic
birds, one DNA^DNA hybridization study, which was
performed by Sibley & Ahlquist (1990), is unique in both
the number of their species comparisons and the nature
of their ¢ndings. Except for ducks, cranes and rails, the
aquatic birds formed a single evolutionary group (Cico-
niiformes). The relationships within this large aquatic
group were non-traditional and the `Pelecaniformes'
appeared as polyphyletic. In particular, some birds with
divergent morphologies formed sister groups (e.g. storks
with condors and pelicans with shoebill), while other
birds with similar morphologies seemed more distantly
related (e.g. loons with grebes and pelicans with cormor-
ants). In order to account for such relationships, rapid
rates of morphological evolution were implied. Such
scenarios have generally received little support (Feduccia
1996). However, both morphological convergence and
divergence have been described before in aquatic organ-
isms (Storer 1971; Nikaido et al. 1999). It is therefore
surprising that we still lack subsequent molecular investi-
gations with complete familial representations of these
aquatic birds. In order to address this issue, we examined
the phylogenetic position of the enigmatic £amingos,
hamerkop and shoebill among the other major aquatic
bird families by obtaining new DNA sequences and
DNA^DNA hybridization data.
2. METHODS
(a) DNA sequence analyses
Mitochondrial gene sequences were obtained from the 12S
rRNA, tRNAVa l and 16S rRNA genes, yielding ca. 3kb per
sequence. Twenty-six representatives of the major families of
Ciconiiformes and Gruiformes (Sibley & Ahlquist 1990) were
included, with the domestic duck and fowl serving as
outgroups. With Ciconiiformes as part of a neoavian clade,
galloanserine birds have been shown to provide appropriate
outgroup taxa to this clade (Sibley & Ahlquist 1990; Van
Tuinen et al. 2000). Most mitochondrial sequences have been
obtained previously (Van Tuinen et al. 2000): Balaeniceps rex
(AF173569), Charadrius semipalmatus (AF173565), Ciconia nigra
(AF173571), Fregata magni¢cens (AF173576), Gavia immer (AF173577),
Gavia stellata (AF173578), Grus canadensis (AF173564), Gymnogyps
californianus (AF173574), Larus glaucoides (AF173566), Neophron
percnopterus (AF173581), Pelecanus occidentalis (AF173570), Phaethon
aethereus (AF173592), Phalacrocorax brasilianus (A F173580),
Phoenicopterus ruber (AF173568), Podiceps auritus (AF173567),
Pu¤nus gravis (AF173572), Pygoscelis adeliae (AF173573), Sula
nebouxii (AF173579) and Vultur gryphus (AF173575). A heron
(Nycticorax nycticorax), spoonbill (Platalea alba), hamerkop (Scopus
umbretta) and a second grebe (Aechmophorus occidentalis) were
Proc. R. Soc. Lond. B(2001)268, 1345^1350 1345 &2001 The Royal Society
Received 29 January 2001 Accepted 22 March 2001
doi 10.1098/rspb.2001.1679
*Author for correspondence (sbh1@psu.edu).
{Present address: Department of Biological Sciences, Stanford
University, Stanford, CA 94305, USA.
added to this data set using standard primers and a sequenc-
ing protocol (Hedges & Sibley 1994; Van Tuinen et al.2000).
Other available representative sequences were obtained from
GenBank (Ciconia ciconia (AB026818), Ciconia boyciana (AB026193),
Falco peregrinus (AF0 9 0338), Gallus gallus (X52392) and Anas
platyrhynchos (L16770)). A 750-base pair (bp) fragment of a
grebe (Podiceps auritus) was sequenced, aligned and analysed for
the mitochondrial cytochrome bgene, with more than 180
other aquatic bird sequences available for this gene. Accession
numbers for those sequences and alignments and other supple-
mentary data are available at http://www.evogenomics.org/
publications/data/£amingo/index.htm.
Unless otherwise noted, sequences from two nuclear genes
were obtained with representatives identical to the mito-
chondrial rRNA data set and using the same methods as for
the mitochondrial sequences. A 600-bp exon fragment of the
c-mos proto-oncogene was sequenced in 16 of the earlier
mentioned aquatic species using previously described primers
(Cooper & Penny 1997) and added to available sequences from
a loon (Gavia arctica (U88423)), penguin (Eudyptes pachyrhynchus
(U88420)), shearwater (Pu¤nus griseus (U88421)), tropic bird
(Phaethon rubricauda (U88418)),gull(Larus heermanni ( U88419) ),
guinea fowl (Numida meleagris (U88425)) and domestic fowl
(G. gallus (M19412)). This exon fragment was only partially
obtained (sequenced) in the domestic duck (391bp). A 370-bp frag-
ment of intron 11 (with respect to G. gallus (M11213)) of the
glyceraldehyde-3-phosphodehydrogenase (G3PDH) gene was
obtained for 18 birds (with one loon and grebe representative
and Porphyrio porphyrio as a gruiform representative but excluding
frigate bird and vulture representatives due to ampli¢cation
problems) using the primer pairs G3P13/G3P14 and G3P13A/
G3p14 (G3P13 TCCAYCTTTGATGCGGRTGCTGGMAT,
G3P13A GGCATTGCACTGARYGAYCATTT and G3P14
ARRTCCACAACACGGTTGCTGTA). The combined nuclear
and mitochondrial data set included a shoebill, hamerkop and
one £amingo as well as one heron, stork, spoonbill, pelican,
cormorant, booby, tropic bird, penguin, shearwater, plover,
gull, gruiform, grebe, loon, domestic duck and fowl. New
sequences have been deposited in GenBank under accession
numbers AF339322^AF339361.
Sequences were aligned by individual gene using the multiple
alignment option in CLUSTAL W (Thompson et al.1994). Phylo-
genetic analyses and estimation of most data parameters were
performed in PAUP*(Swo¡ord 1998) and MEGA (Kumar et al.
1993). The shape parameter of the gamma distribution for vari-
able evolutionary rates among sites was estimated for each gene
and the combined data set using PAML (Yang 1997). The
domestic duck and fowl were used as outgroup species in all
analyses and a pairwise deletion option (Nei & Kumar 2000)
was used whenever nuclear genes were involved due to the nature
of these data sets (randomly distributed indels in the G3PDH
intron and a shorter c-mos sequence size of the domestic duck).
The stability of the topologies based on the combined data set
(n19) was tested using di¡erent tree-building methods
(maximum likelihood, maximum parsimony and neighbour
joining). Combined and gene-speci¢c analyses (using individual
genes) in conjunction with assessment of the e¡ect of distance
correction (P,gamma,Jukes
^Cantor and Kimura two-parameter
methods), di¡ering substitution rates (transversion weighted
analyses), biased base composition (Tamura^Nei distances),
possible heterogeneous base composition among sequences (loga-
rithmic determined-transformed distances), di¡ering frequency
of invariant sites (Swo¡ord 1998) and di¡ering order of sequence
addition into initial alignment were performed with the neigh-
bour-joining method. Unless stated otherwise, the Tamura^Nei
distance was used because it takes into account the biased base
composition of the respective data sets. The signal for alternative
topologies was explored with Kishino^Hasegawa tests in
conjunction with maximum likelihood (Swo¡ord 1998) and
through spectral analyses (Hendy & Penny 1993) on maximally
2020 distance matrices from the separate and combined gene
data. Spectral analyses were performed with SPECTRUM
(Charleston & Page 1997). In order to assess the signi¢cance
of resulting nodes, the bootstrap method was applied with
2000 (neighbour joining), 500 (maximum parsimony) and 100
(maximum likelihood) iterations using MEGA (Kumar et al.
1993) and PAUP*(Swo¡ord 1998) and values of 95% were
considered to be signi¢cant. Standard error tests were employed
with neighbour joining only. Relative apparent synapomorphy
analysis (RASA) was employed for assessing the overall signal of
the separate and combined genes and the topological e¡ect of
subsequent noise reduction (Lyons-Weiler et al. 1996), performing
optimal outgroup analyses (Lyons-Weiler et al. 1998) and identi-
fying signi¢cant topology-altering long branches using taxon-
variance plots. These analyses identi¢ed the F. peregrinus RNA
sequence as a signi¢cant long branch and we excluded this
sequence from the ¢nal analyses.
(b) DNA^DNA hybridization analyses
DNA hybridization experiments were performed using a
protocol modi¢ed from that of Sibley & Ahlquist (1990) and
described previously (Kirsch et al. 1990; Bleiweiss et al. 1994)
and included 21 mostly aquatic species. Representative birds
were as in the DNA sequence study except for Accipiter
melanoleucus,Amaurornis phoenicurus,Anhinga rufa,Bubo virginianus,
Diomedeabulleria,Egrettanovaehollandiae,Mycteria americana,Phaethon
rubricauda,Phalacrocorax carunculatus,Plegadis falcinellus,Pygoscelis
papua and Sula dactylatra. Every species was labelled and the
¢nal matrix was complete except for 14 heterologous
comparisons, with an average of 3.2 replicates per cell. Missing
measurements were estimated by re£ection from known
reciprocals after symmetrization of the matrix (Sarich &
Cronin 1976). The G, S and P 0 options were employed in
FITCH tree calculation, the input order of birds was randomly
varied 100 times and the tree was validated by bootstrapping
and jackkni¢ng procedures (Krajewski & Dickerman 1990;
Felsenstein 1993; Lapointe et al. 1994). See http://www.
evogenomics.org/publications/data/£amingo/index.htm for the
hybridization matrices and a nearly congeneric Sibley &
Ahlquist (1990) data set (n18) used for comparison.
3. RESULTS AND DISCUSSION
We obtained sequences from four mitochondrial and
two nuclear genes from representative aquatic birds as well
as other possible £amingo relatives (the crane, rail and
domestic duck) and outgroup species. The RASA analyses
(Lyons-Weiler et al. 1996) showed that each gene contained
a signi¢cant (p50.01), non-random phylogenetic signal
and that, for all of these genes, the domestic duck and
fowl sequences were not long branches and, thus, pro-
vided valid outgroup species (tRASArooted 4tRASAunrooted)
(Lyons-Weiler et al. 1998). Because the combined data set,
which comprised 19 birds and 4062 sites, displayed the
largest signal (tRASA 7.2) among sets with equal
numbers of birds, we performed phylogenetic analyses on
1346 M.VanTuinen and others The evolution of aquatic birds
Proc. R. Soc. Lond. B(2001)
this larger data set. Although this data set was unable to
resolve the entire aquatic bird phylogeny, signi¢cant
resolution was found for the three enigmatic waders in
question. Together with other widely accepted clusters
(booby^cormorant, gull^plover and penguin ^shearwater),
the bootstrap consensus topology showed signi¢cance for
an assemblage of the hamerkop with the shoebill and
pelican and an unexpected cluster formed by the £amingo
and grebe (¢gure 2). These ¢ve clusters appeared re-
gardless of tree building method, distance correction,
frequency of invariant sites, use of noise reduction
(excluding 1070 sites) or when varying the order of
sequence input during the aligning process.
Because the latter two clusters (hamerkop plus shoe-
bill plus pelican and £amingo plus grebe) con£ict with
other available molecular data (Sibley & Ahlquist 1990),
we subsequently investigated phylogenetic consistency
between genes regarding these birds. The separate gene
trees all supported the £amingo^grebe cluster (72, 84 and
96%, respectively, for c-mos (n25), G3PDH (n19)
and combined mitochondrial RNA (n27 excluding
Falco) genes using a Tamura^Nei distance on transversions
only). In the species-dense (n181 sequences) conditions
of the mitochondrial cytochrome bdata set, where few
groupings were supported by signi¢cant bootstrap values,
the £amingo and grebe clustered together either with
(56%) or without (40%) transitions. The phylogenetic
placement of the shoebill with the pelican was also
supported by all (individual and combined) genes, as well
as an alliance of this group with the hamerkop in the four
individual non-protein coding genes. Although not in
con£ict with such a placement, the protein-coding c-mos
fragment alone did not resolve the position of the
hamerkop.
The evolution of aquatic birds M. VanTuinen and others 1347
Proc. R. Soc. Lond. B(2001)
penguins
loons
shorebirds, gulls, auks
NW vultures
hamerkop
shoebill
storks
flamingos
herons
other birds
pelecaniforms*
rails, cranes
pigeons
waterfowl
gamefowl
'wading' 'foot-propelled diving' 'totipalmate'
grebes
other raptors, owls
ibises, spoonbills
shearwaters, albatrosses, petrels
Figure 1. Traditional relationships of aquatic birds based on morphological studies (Cracraft 1981). More recently, herons were
instead placed near the shorebirds, cranes and pigeons and the New World vultures were placed closest to the storks (Cracraft
1988). Bold-faced birds form a single order Ciconiiformes based on DNA^DNA hybridization data (Sibley & Ahlquist 1990).
Pelecaniforms include the pelicans, frigate birds, tropic birds, darters, boobies and cormorants.
1348 M.VanTuinen and others Th e evolution of aquatic birds
Proc. R. Soc. Lond. B(2001)
ibis
heron
shoebill
pelican
hamerkop
stork
loon
penguin
shearwater
grebe
flamingo
gruiform
gull
plover
tropic bird
booby
darter
albatross
frigate bird
cormorant
owl
goshawk
domestic duck
domestic fowl
56
0
5
10
15
20
6460 68 76 84 9272
temperature (°C)
total counts (%)
80 88
*
*
*
*
**
*
*
*
*
*
*
*
96
WG
GF
WH
DF
Figure 2. Molecular evidence bearing on the positions of £amingos, hamerkop and shoebill. Colour coding as in ¢gure 1. Left:
DNA^DNA hybridization phylogeny based on a FITCH (Felsenstein 1993) computation on Tms (median melting point
di¡erences) using the G, S and P 0 options and varying the input order of taxa 100 times. The unexplained tree sum of
squares was 0.9% of the total matrix sum of squares (n1431) (average repeated measurements per ce ll 3.2, average standard
deviation (s.d.) 0.39 and correlation of s.d.s with distance 70.01). The tree was tested by the bootstrap for distances
(Krajewski & Dickerman 1990) and jackknife for weighted trees (Lapointe et al. 1994). Asterisks indicate nodes supported by
100% of the 1000 bootstrap replicates, except for that uniting all taxa including and above the booby^darter^cormorant clade,
which received 98% support. All other nodes were supported by 450% of the replicates. The shoebill^pelican^hamerkop and
In addition, we tested for the presence of an alternative
signal in the combined sequence data set (n19) by
comparing the maximum-likelihood values for trees based
on the same observed data parameters (invariable
proportion 0.502, alpha 0.27 and transition 1.6) but
under di¡erent topological constraints. These Kishino^
Hasegawa tests (Swo¡ord 1998; but see Goldman et al.
2000) signi¢cantly (p50.0001) supported the tree that
yielded a grebe^£amingo cluster (lnL730001.37)
(shown in condensed form in ¢gure 2), as opposed to a
duck^£amingo relationship (lnL730120.882) (Sibley
& Ahlquist 1990), a plover^£amingo relationship
(lnL730084.66) (Olson & Feduccia 1980; Feduccia
1996), a topology based on previous hybridization data
(lnL730151.45) (Sibley & Ahlquist 1990) and the
traditional phylogeny (lnL730320.34) (Cracraft 1988).
Spectral analyses also supported the ¢ve groupings with
high support:low con£ict while showing low or no support
for other alternative associations. Speci¢c to the position of
the enigmatic aquatic birds in this study, the support:con-
£ict ratio for a groupingof the £amingo with the grebe was
45.5, which was higher than those ratios for a grouping of
the £amingo with either the duck (1.4 10 73) or shorebird
(3.1 1073). The support for the hamerkop as sister to a
shoebill^pelican grouping yielded a support:con£ict
ratio of 114.9, as opposed to 4.11074for a monophyletic
Pelecaniformes.
In order to investigate the apparent discrepancy
between the sequence data and the available hybrid-
ization data further (Sibley & Ahlquist 1990), we
constructed a new, nearly complete hybridization distance
matrix comprising 21 representative birds. Although
agreeing with earlier hybridization data (Sibley &
Ahlquist 1990) in showing the polyphyletic associations of
various `pelecaniforms', these data also showed a
signi¢cant £amingo^grebe relationship as well as a
hamerkop^shoebill^pelican cluster (¢gure 2). Because of
the latter di¡erence from earlier hybridization results, we
re-examined Sibley & Ahlquist's (1990) original data for
birds, which approximately corresponded to those in
¢gure 2. Among those data were two distances from the
western grebe (which was radioactively labelled with
iodine-125) to the £amingo that were on average shorter
than those to other birds in the data set. We calculated
bootstrapped trees with a 70% complete Sibley &
Ahlquist (1990) data set for 18 species using an additive-
estimation program in order to complete each pseudo-
replicate matrix (Landry et al. 1996). The results, like
those for our own matrix, also united the western grebe
and £amingo, with 86% (T50Hs-expressed: median
melting points corrected for per cent hybridization) or
69% (Tmodes-expressed: temperatures at which the largest
number of sequences melt) support in 1000 replicates on
the Sibley & Ahlquist (1990) data. Other associations of
`pelecaniforms' such as the booby^darter^cormorant
grouping and that of the pelican with the shoebill and
hamerkop received 100% bootstrap support. The `peleca-
niform' frigate bird and tropic bird were separated from
these trios and from each other as suggested previously
(Sibley & Ahlquist 1990). Presumably, Sibley & Ahlquist
(1990) did not report the £amingo^grebe and hamerkop^
shoebill pairing because their comparisons were so few or
because both members of a pair were not included in any
tree calculated by them.
Thus, the hybridization and sequence data consistently
united the £amingos with grebes and the hamerkop with
the pelican and shoebill, thereby further disassociating
the `pelecaniforms'. Some of these ¢ndings have support
from other data as well, including a smaller molecular
data set (Hedges & Sibley 1994), fossil evidence (Olson
1985), middle ear morphology (Sai¡ 1978) and jaw
articulation (Cottam 1957). However, to our knowledge,
the £amingo^grebe grouping is novel. We further elimi-
nated the possibility of contamination of either £amingos
or grebes by cross-checking our sequences with other
available sequences in public databases as well as using
independent DNA sources in the sequence and hybridiza-
tion experiments in this study. We propose that this
unusual alliance of birds has been overlooked because the
exceptional adaptations to their respective aquatic niches
have obscured evolutionary history.
The distant relationship of grebes and loons within
aquatic birds implies convergent evolution of morphology
imposed by the aquatic niche. Their hind leg muscula-
ture, bill shape and streamlined body are clear adapta-
tions for catching ¢sh by means of foot-propelled diving
(Storer 1971) and fossils have provided evidence for the
antiquity of this lifestyle (Chiappe 1995; Feduccia 1996).
Flamingos are divergent from both this body plan and
lifestyle. Our data also indicated morphological conver-
gence among the totipalmate birds (`pelecaniforms') and
among the wading birds (¢gures 1 and 2). Speci¢cally, the
characters employed in uniting Pelecaniformes, e.g. the
totipalmate condition and presence of a gular pouch in
wading birds (sensu Cracraft 1981), e.g. related to long leg
size, have probably converged in di¡erent aquatic bird
lineages. In addition, these data suggest that the shoebill
is not an aberrant pelecaniform, as proposed before
(Cottam 1957; Sai¡ 1978), but instead that the pelicans
are aberrant long-legged waders in which leg size has
been secondarily reduced. Denser species sampling is
needed in order to determine the closest relative of the
shoebill^hamerkop^pelican assemblage. These new ¢nd-
ings add to previous evidence for convergence among
The evolution of aquatic birds M. VanTuinen and others 1349
Proc. R. Soc. Lond. B(2001)
Figure 2. (Contd)grebe
^£amingo groups were also fully supported by the jackknife, which was based on all single and 5000
random multiple deletions of taxa. Nodes that were not supported by jackkni¢ng are shown with dotted lines. Inset: representa-
tive stepwise thermal-elution curves for four taxa, corrected for percentage hybridization (from right to left, the homologous
western grebe (WG), greater £amingo (GF), white-faced heron (WH) and domestic fowl (DF)). Vertical marks indicate Tms.
Note that the £amingo curve is some 48closer to the homologue than are the other two heterologues. Right: sequence-based
phylogeny, as shown by a 50% condensed bootstrap consensus tree (Nei & Kumar 2000) for the combined genes (c-mos proto-
oncogene exon, G3PDH intron 11 and complete 12S rRNA, tRNAVal and 16S rRNA genes). The combined data parameters
are as follows: 4062 total, 1855 variable and 1196 parsimony informative sites, alpha0.268 and transition/transversion 1.6
and T 21 .2%, C 26.1%, A 30.0% and G 22.7%. Asterisks indicate nodes that were signi¢cantly (495%) supported
based on bootstrapping using maximum likelihood or a standard error test using neighbour joining.
auks and penguins and gulls and albatrosses (Storer 1971;
Feduccia 1996).
Some insight into the early evolution of £amingos and
grebes can be gleaned from their fossil record. Although
some fossil £amingos (Palaelodus) might be interpreted as
being grebe-like in appearance and behaviour (Feduccia
1996), the earliest £amingo fossils resemble a more
typical shorebird (Olson & Feduccia 1980). The earliest
extant shorebirds and gruiforms consist of small-bodied
rail-like water birds (Sibley & Ahlquist 1990). The
£amingo style of ¢lter feeding may have evolved through
accidental water intake during pecking at food in water
(Zweers et al. 1994). Likewise, the grebe style of diving
would appear to be equally derived. Thus, the £amingos
and grebes probably each represent morphological diver-
gence from a typical shorebird habitus and lifestyle.
We thank T. Barkman and J. Lyons-Weiler for assistance with
the RASA analyses, P. De Benedictis for supplying the compiled
Sibley & Ahlquist (1990) data and W. Feeny for producing the
inset to ¢gure 2. We owe special thanks to the late Dr C. G.
Sibley for his advice and support during the initial stage of this
project. S.B.H. was supported by National Aeronautics and
Space Administration and the National Science Foundation.
REFERENCES
Bleiweiss, R. E., Kirsch, J. A. W. & Matheus, J. C. 1994 DNA
hybridization evidence for subfamily structure among
hummingbirds. Auk 111, 8^19.
Charleston, M. A. & Page, R. D. M. 1997 SPECTRUM, version
2.0. Glasgow, UK: University of Glasgow.
Chiappe, L. M. 1995 The ¢rst 85 million years of avian evolu-
tion. Na t u r e 378, 349^355.
Cooper, A. & Penny, D. 1997 Mass survival of birds across the
K^T boundary: molecular evidence. Science 275, 1109^1113.
Cottam, P. A. 1957 The pelecaniform characters of the skeleton
of the shoebill stork Balaeniceps rex.Bull. Br. Mus. Nat. Hist. 5,
51^ 72.
Cracraft, J. 1981 Toward a phylogenetic classi¢cation of the
recent birds of the world (class Aves). Au k 98, 681^ 714.
Cracraft, J. 1988 The major clades of birds. In The phylogeny and
classi¢cation of the tetrapods (ed. M. J. Benton), pp. 339^361.
Oxford, UK: Clarendon Press.
Feduccia, A. 1996 The origin and evolution of birds. New Haven,
CT: Yale University Press.
Felsenstein, J. 1993 PHYLIP, phylogenetic inference package, program
and document ation, v. 3.5c. Seattle,WA: University of Washington.
Goldman, N., Anderson, J. P. & Rodrigo, A. G. 2000
Likelihood-based tests of topologies in phylogenetics. Syst.
Biol. 49, 652^670.
Hedges, S. B. & Sibley, C. G. 1994 Molecules vs. morphology in
avian evolution: the case of the `pelecaniform' birds. Proc. Natl
Acad.Sci.USA91, 9861^9865.
Hendy, M. D. & Penny, D. 1993 Spectral analysis of phyloge-
netic data. J. Class. 10, 5^24.
Kirsch, J. A. W., Springer, M. S., Krajewski, C., Archer, M.,
Aplin, K. & Dickerman, A. W. 1990 DNA^DNA hybridiza-
tion studies of carnivorous marsupials. I. The intergeneric
relationships of bandicoots (Marsupialia: Perameloidea). J.
Mol. Evol. 30, 434^448.
Krajewski, C. & Dickerman, A. W. 1990 Bootstrap analysis of
phylogenetic trees derived from DNA hybridization distances.
Syst. Zool. 39, 383^390.
Kumar, S., Tamura, K. & Nei, M. 1993 MEGA: molecular evolu-
tionary genetic analysis, version 1.01. University Park, PA: Institute
of Molecular Evolutionary Genetics, The Pennsylvania State
University.
Landry, P.-A., Lapointe, F.-J. & Kirsch, J. A. W. 1996
Estimating phylogenies from lacunose distance matrices:
additive is superior to ultrametric estimation. Mol. B iol. Evol.
13, 818^823.
Lapointe, F.-J., Kirsch, J. A. W. & Bleiweiss, R. E. 1994
Jackkni¢ng of weighted trees: validation of phylogenies
reconstructed from distance matrices. Mol. Phylogen et. Ev ol. 3,
256^267.
Lyons-Weiler, J., Hoelzer, G. A. & Tausch, R. J. 1996 Rela-
tive apparent synapomorphy analysis (RASA) I. The statis-
tical measurement of phylogenetic signal. Mol. Biol. Evol. 13,
749^757.
Lyons-Weiler, J., Hoelzer, G. A. & Tausch, R. J. 1998 Optimal
outgroup analysis. Biol. J. Linn. Soc. Lond. 64, 493^511.
Nei, M. & Kumar, S. 2000 Molecular evolution and phylogenetics.
Oxford University Press.
Nikaido, M., Rooney, A. P. & Okada, N. 1999 Phylogenetic
relationships among cetartiodactyls based on insertions of
short and long interpersed elements: hippopotamuses are the
closest extant relatives of whales. Proc. Natl Acad. Sci. USA 96,
10261^10 266.
Olson, S. L. 1985 The fossil record of birds. In Avian biology (ed.
D. S. Farner, J. R. King & K. C. Parkes), pp. 79^252. New
York: Academic Press.
Olson, S. L. & Feduccia, A. 1980 Relationships and evolution of
£amingos (Aves: Phoenicopteridae). Smithson. Contr. Zool. 316,
1^73.
Sai¡, E. I. 1978 The middle ear of the skull of birds: the
Pelecaniformes and Ciconiiformes. Zool. J. Linn. Soc. Lond. 63,
315^ 370.
Sarich, V. M. & Cronin, J. E. 1976 Molecular systematics of the
primates. In Molecular anthropology, genes, and proteins in the
evolutionary ascent of primates (ed. M. Goodman & R. E.
Tashian), pp. 141^170. NewYork: Plenum Press.
Sibley,C.G.&Ahlquist,J.E.1990Phylogeny and classi¢cation of
birds: a study in molecular evolution. New Haven, CT: Yale
University Press.
Storer, R. W. 1971 Adaptive radiation in birds. In Avian biology
(ed. D. S. Farner & J. R. King), pp. 149^188. New York:
Academic Press.
Swo¡ord, D. L. 1998 PA U P *:phylogenetic analysis using parsimony
(*and other methods), v. 4.0. Sunderland, MA: Sinauer Associates.
Thompson, J. D., Higgins, D. G. & Gibson, T. J. 1994 Clustal
W: improving the sensitivity of progressive multiple sequence
alignment through sequence weighting, positions-speci¢c gap
penalties and weight matrix choice. Nucl. Acids Res. 22,3488^
3494.
Van Tuinen, M., Sibley, C. G. & Hedges, S. B. 2000 The early
history of modern birds inferred from DNA sequence of mito-
chondrial and nuclear ribosomal genes. Mol. Biol. E vo l. 17,
451^457.
Yang, Z. 1997 PAML: a program package for phylogenetic
analysis by maximum likelihood. CABIOS 13, 555^556.
Zweers, G. A., DeJong, F., Berkhoudt, H. & Vandenberge, J. C.
1994 Filter-feeding in £amingos (Phoenicopterus ruber ruber).
Condor 97,297^324.
1350 M. VanTuinen and others The evolution of aquatic birds
Proc. R. Soc. Lond. B(2001)
... Both mitochondrial and nuclear genetic information were well studied to know the systematics status, evolutionary relationships, and population structures of Charadriiformes birds (Liebers et al. 2001;Crochet et al. 2003;Ericson et al. 2003;Paton et al. 2003;Thomas et al. 2004;van Tuinen et al. 2004;Pons et al. 2005;Baker et al. 2007). The comparative genomics is also employed for estimating the genetic variability, substitution pattern, phylogenetic relationship, and evolution of avifauna (van Tuinen et al. 2001;Paton et al. 2002;Backstr€ om et al. 2008;Hackett et al. 2008;K€ unstner et al. 2010). As of now, 47 complete mitochondrial genomes of Charadriiformes birds are available in the GenBank database. ...
Article
Full-text available
The complete mitogenome sequence of the brown-headed gull, Chroicocephalus brunnicephalus was determined in this study. The 16,771 bp genome consists of 13 protein-coding genes (PCGs), two ribosomal RNA (rRNA) genes, and 22 transfer RNA (tRNA) genes, and a control region (CR). The decoded mitogenome was AT-rich (54.77%) with nine overlapping and 17 intergenic spacer regions. Most of the PCGs were started by a typical ATG initiation codon except for cox1 and nad3. Further, the usual termination codons (AGG, TAG, TAA, and AGA) were used by 11 PCGs except for cox3 and nad4. The concatenated PCGs based Bayesian phylogeny clearly discriminates all the Laridae species and reflects the sister relationship of C. brunnicephalus with C. ridibundus. The present mitogenome-based phylogeny was congruent with the earlier hypothesis and confirmed the evolutionary position of the brown-headed gull as masked species. The generated mitogenome of C. brunnicephalus is almost identical to the previously generated mitogenome from China except for two base pairs in CR. To visualize the population structure of this migratory species, we propose more sampling from different geographical locations and the generation of additional molecular data to clarify the reality.
... Such cryptic taxa are often discovered through DNA sequencing and open a new window to diversity and conservation [16]. In birds, taxonomy has traditionally been based on morphology, including plumage patterns, and convergent as well as divergent evolution has often led to misclassifications at every taxonomic level, e.g., [17][18][19][20][21][22]. ...
Article
Full-text available
Larks constitute an avian family of exceptional cryptic diversity and striking examples of convergent evolution. Therefore, traditional morphology-based taxonomy has recurrently failed to reflect evolutionary relationships. While taxonomy ideally should integrate morphology, vocalizations, behaviour, ecology, and genetics, this can be challenging for groups that span several continents including areas that are difficult to access. Here, we combine morphometrics and mitochondrial DNA to evaluate the taxonomy of Calandrella larks, with particular focus on the African C. cinerea and the Asian C. acutirostris complexes. We describe a new range-restricted West African taxon, Calandrella cinerea rufipecta ssp. nov. (type locality: Jos, Plateau State, Nigeria), with an isolated relic population 3000 km from its closest relative in the Rift Valley. We performed molecular species delimitation, employing coalescence-based multi-rate Poisson Tree Processes (mPTP) on cytochrome b sequences across 52 currently recognized lark species, including multiple taxa currently treated as subspecies. Three species-level splits were inferred within the genus Calandrella and another 13 across other genera, primarily among fragmented sub-Saharan taxa and taxa distributed from Northwest Africa to Arabia or East Africa. Previously unknown divergences date back as far as to the Miocene, indicating the presence of currently unrecognized species. However, we stress that taxonomic decisions should not be based on single datasets, such as mitochondrial DNA, although analyses of mitochondrial DNA can be a good indicator of taxa in need of further integrative taxonomic assessment.
... Clearly, the relationships of certain higher-level clades in the molecular phylogenies, such as the relationship of tropicbirds, kagus, and sunbitterns and their proximity to the Aquatic & Semiaquatic Birds, are unexpected and ask for new ontogenetic and morphological studies in order to assess the anatomical plausibility of these findings. In some cases, other surprising clades derived from analyses of molecular data have already been confirmed by morphological data, such as the sister group relationship between the morphologically and behaviorally very disparate grebes and flamingos (Van Tuinen et al. 2001;Mayr 2004). More recently, it was also hypothesized that the plesiomorphic presence of a large lacrimal bone may support the basal position of the Caprimulgidae within Caprimulgiformes, with this bone being reduced in other Caprimulgiformes (Chen et al. 2019;contra Mayr 2010). ...
Article
Full-text available
Presumably, due to a rapid early diversification, major parts of the higher-level phylogeny of birds are still resolved controversially in different analyses or are considered unresolvable. To address this problem, we produced an avian tree of life, which includes molecular sequences of one or several species of ∼ 90% of the currently recognized family-level taxa (429 species, 379 genera) including all 106 for the non-passerines and 115 for the passerines (Passeriformes). The unconstrained analyses of noncoding 3-prime untranslated region (3'UTR) sequences and those of coding sequences yielded different trees. In contrast to the coding sequences, the 3'UTR sequences resulted in a well-resolved and stable tree topology. The 3'UTR contained, unexpectedly, transcription factor binding motifs that were specific for different higher-level taxa. In this tree, grebes and flamingos are the sister clade of all other Neoaves, which are subdivided into five major clades. All non-passerine taxa were placed with robust statistical support including the long-time enigmatic hoatzin (Opisthocomiformes), which was found being the sister taxon of the Caprimulgiformes. The comparatively late radiation of family-level clades of the songbirds (oscine Passeriformes) contrasts with the attenuated diversification of non-passeriform taxa since the early Miocene. This correlates with the evolution of vocal production learning, an important speciation factor, which is ancestral for songbirds and evolved convergent only in hummingbirds and parrots. Since 3'UTR-based phylotranscriptomics resolved the avian family-level tree of life, we suggest that this procedure will also resolve the all-species avian tree of life.
... Therefore, it is essential to esti-mate the loss of genetic diversity from populations to carry out a genetic management for conservation of rare species. However, there are few reports on the analysis of genetic diversity in penguins [6][7][8][9][10][11][12][13][14][15] , and reports on the use of genetic approach for penguin conservation are extremely limited. In this study, we focused on the major histocompatibility complex (MHC) gene locus which contains highly polymorphic DNA markers for estimating the biodiversity in conservation genetics 16,17) . ...
Article
Full-text available
Endangered penguins are in need of the genetic management to estimate the risk for loss of genetic diversity from populations resulting from habitat fragmentation or inbreeding. However, as for genome diversity in penguin species, there are limited reports and it is insufficient to give useful and appropriate information. In the present study, we obtained the full length sequence of the MHC class II gene, which is expected to be a useful genetic marker for biodiversity in conservation genetics. The 4.4 kb genome region containing two novel genes was determined for nucleotide sequences using genomic DNA extracted from Humboldt penguin, by using the inverse PCR method. Homology analysis of MHC class II genes with those from other birds suggested that the novel two genes were alpha and beta genes. In addition, phylogenetic analysis suggested that the beta gene of penguins was clustered with the beta genes from waterfowl. These observations provide basic information on the structure of MHC class II locus to relieve the genetic diversity of penguin species.
... A volant distant outgroup taxon, Phaethon rubricauda (Phaethontiformes), was selected to better identify the evolutionary trajectory of character evolution as per the well-supported and commonly cited (e.g., Baker and Manwell, 1975;Raikow et al., 1988;Thomas et al., 2011; hypothesis that penguins are descended from aerially flighted ancestors . Studies of higher avian phylogenies have long recovered Phaethontiformes among aquatic birds (Cracraft, 1981(Cracraft, , 1988, part of a lineage sister to core waterbird clades (e.g., van Tuinen et al., 2001;Jarvis et al., 2014;Prum et al., 2015;Thomas, 2015;Houde et al., 2019). Phaethon rubricauda (Phaethontiformes) was scored for all osteological characters. ...
Article
Full-text available
Numerous skeletal remains recovered in situ from the late early to middle Paleocene Takatika Grit of Chatham Island, New Zealand, are among the oldest known fossils attributed to the penguin clade (Aves, Sphenisciformes). They represent a new medium-sized taxon, for which we erect a new genus and species, and a second, notably larger form. These new penguins are analysed in a parsimony and Bayesian framework using an updated and revised phylogenetic matrix, based on morphological and molecular characters, and interpreted as among the most basal of known sphenisciforms, closely related to Waimanu. While sharing numerous characteristics with the earliest wing-propelled divers, the novel taxon records the oldest occurrence of the characteristic penguin tarsometatarsus morphology. These ancient Chatham Island representatives add to a growing number and increased morphological diversity of Paleocene penguins in the New Zealand region, suggesting an origin for the group there. With their addition to other Paleocene penguins, these taxa reveal that sphenisciforms rapidly diversified as non-volant piscivores in the southern oceans following the end-Cretaceous mass extinction. They also provide further evidence for the hypothesis that their origin predates the Paleocene. This implies that stem Sphenisciformes and their sister group, the Procellariiformes, both originated in, and so may be expected to occur in, the Late Cretaceous.
... Unlike the present study, most of the conflicting reports explicitly or implicitly place more faith in mtDNA than in non-molecular data (e.g. Kusmierski et al. 1997, McCracken & Sheldon 1998, McCracken et al. 1999, Brumfield & Braun 2001, Omland & Lanyon 2000, van Tuinen et al. 2001. A notable exception is Lee et al. (1997) who state that morphological data probably reflect the species phylogeny of ratite birds better than mtDNA. ...
Article
Full-text available
We used the mitochondrial control region and ND2 and ATP8+6 genes (c. 2900 bp), an intron in the CHD1Z gene on the Z-chromosome (c. 550 bp) and non-molecular (mainly morphological) data to reconstruct the phylogeny of nearly all of the least-inclusive nominal taxa in the avian genus Motacilla (wagtails). All three data sets are incongruent, although the mitochondrial DNA tree deviates most. We propose that, in wagtails, the CHD1Z tree conforms better than the mitochondrial DNA tree with the taxon phylogeny. We argue that the mitochondrial DNA tree has been affected by introgressive hybridisation, possibly also by lineage sorting. We suggest that introgression and successive fixation of foreign mitochondrial DNA may be a common phenomenon in recently diverged taxa, where reproductive isolating barriers are generally incomplete, and stress that this will lead to permanent incongruence between gene trees and species trees. Our results highlight the importance of using multiple independent data sets in phylogeny reconstructions. However, a review of the literature published in the past seven years reveals that virtually all avian phylogenies in that time period are based only on mitochondrial sequences. We argue against indiscriminately combining independent data sets, especially if the number of informative sites differs markedly between the data sets, as is the case in the present study.
Article
Full-text available
Knowledge of the higher-level phylogenetic relationships of birds has grown substantially during the past two decades due to the application of genomic data. However, the nomenclature of higher-level taxa has not become more stable, due to the lack of regulation of taxon names above the level of superfamily by the ICZN, and the usage of rank-based nomenclature, which is not tied to clades in a phylogeny. Lack of regulation and the instability of rank-based nomenclature impede effective communication among systematists. Here we review support for higher-level avian clades using a set of 10 phylogenomic data sets, and identify clades that are supported by congruency of at least four of these. We provide formal definitions of the names of these clades based on the rules of the recently published PhyloCode. The names of 25 clades are here defined using minimum-crown-clade (n = 23), minimum-clade (n = 1) and maximum-crown-clade (n = 1) definitions. Five new names are introduced here: Dinocrypturi, Pteroclimesites, Musophagotides, Phaethoquornithes and Pelecanes. We also review diagnostic apomorphies of the relevant clades, and identify known synonyms and homonyms. By establishing a formal link between higher-level taxon names and well-supported phylogenetic hypotheses, our phylogenetic definitions will provide a solid basis for the stabilization of avian higher-level nomenclature.
Chapter
Many readers will be acquainted with phylogenetic terminology and avian osteology, and it is beyond the scope of the present work to provide an in-depth overview of these topics, each of which could fill a book on its own. For those less familiar with essential terms and definitions, these are outlined in the present chapter, which also introduces major features of the skull and some of the limb and pectoral girdle bones. Current hypotheses on the interrelationships of extant birds are reviewed, which constitute a phylogenetic framework for the study of fossil taxa. In order to set the following chapters on Paleogene birds into a full context, the Mesozoic fossil record of neornithine birds is furthermore discussed and an overview is given of major Paleogene fossil localities.
Article
Full-text available
The webbed feet of waterbirds are morphologically diverse and classified into four types: the palmate foot, semipalmate foot, totipalmate foot, and lobate foot. To understand the developmental mechanisms underlying this morphological diversity, we conducted a series of comparative analyses. Ancestral state reconstruction based on phylogeny assumed that the lobate feet possessed by the common coot and little grebe arose independently, perhaps through distinct developmental mechanisms. Gremlin1, which encodes a bone morphogenetic protein (BMP) antagonist and inhibits interdigital cell death (ICD) in the foot plate of avian embryos, remained expressed in the interdigital tissues of webbed feet in the duck, common coot, little grebe, and great cormorant. Differences in Gremlin1 expression pattern and proliferating cell distribution pattern in the toe tissues of the common coot and little grebe support the convergent evolution of lobate feet. In the totipalmate-footed great cormorant, Gremlin1 was expressed in all interdigital tissues at St. 31, but its expression disappeared except along the toes by St. 33. The webbing of the cormorant’s totipalmate foot and duck’s palmate foot may have risen from distinct developmental mechanisms.
Article
A large body of comparative macromolecular data bearing on primate systematics is now available. There does not exist, however, any consensus as to just what that bearing is or what it should be. We hope that this chapter can contribute toward such a consensus.
Article
We used DNA-DNA hybridization to determine large-scale phylogenetic structure among hummingbirds (Trochilidae). Analyses of complete matrices of $\Delta T_{{\rm m}}$ and $\Delta T_{{\rm mode}}$ statistics among eight hummingbird genera and a swift generated the same fully resolved topology, which bootstrapping and jackknifing analyses supported at the 100% level. The data are consistent with monophyly for the traditional hermit (Phaethornithinae) and nonhermit (Trochilinae) subfamilies, and with placement of the hermitlike Tooth-billed Hummingbird (Androdon aequatorialis) and Green-fronted Lancebill (Doryfera ludoviciae) among trochilines. Among the trochilines examined, D. ludoviciae is more closely related to the Sparkling Violet-ear (Colibri coruscans) than to A. aequatorialis, and the Collared Inca (Coeligena torquata) is the sister group to these three. Among the hermits examined, the White-tipped Sicklebill (Eutoxeres aquila) represents the first branch, followed by the White-whiskered Hermit (Phaethornis yaruqui), and the closely related Bronzy Hermit (Glaucis aenea) and Bandtailed Barbthroat (Threnetes ruckeri). Evolutionary rate estimates from $\Delta T_{{\rm m}}$ trees corrected for nonadditivity indicate significant rate variation among lineages. Calibration of divergence times with the earliest-known fossil swift suggests that the diverse Andean radiation of trochilines is comprised of at least two lineages (C. torquata, D. ludoviciae/C. coruscans) whose origins date to a period of uplift during the mid-Miocene.
Article
Likelihood-based statistical tests of competing evolutionary hypotheses (tree topologies) have been available for approximately a decade. By far the most commonly used is the Kishino-Hasegawa test. However, the assumptions that have to be made to ensure the validity of the Kishino–Hasegawa test place important restrictions on its applicability. In particular, it is only valid when the topologies being compared are specified a priori. Unfortunately, this means that the Kishino-Hasegawa test may be severely biased in many cases in which it is now commonly used: for example, in any case in which one of the competing topologies has been selected for testing because it is the maximum likelihood topology for the data set at hand. We review the theory of the Kishino-Hasegawa test and contend that for the majority of popular applications this test should not be used. Previously published results from invalid applications of the Kishino–Hasegawa test should be treated extremely cautiously, and future applications should use appropriate alternative tests instead. We review such alternative tests, both nonparametric and parametric, and give two examples which illustrate the importance of our contentions.