Content uploaded by Dietmar Lieckfeldt
Author content
All content in this area was uploaded by Dietmar Lieckfeldt on Sep 15, 2015
Content may be subject to copyright.
Conservation Genetics 5: 205–215, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands. 205
Genetic variation across the current range of the Asian houbara bustard
(Chlamydotis undulata macqueenii)
Christian Pitra1∗, Marie-Ann D’Aloia2, Dietmar Lieckfeldt1& Olivier Combreau2
1Department of Evolutionary Genetics, Institute for Zoo- and Wildlife Research, PF 1103, D-10252 Berlin,
Germany; 2National Avian Research Center, Environmental Research and Wildlife Development Agency, P.O. Box
45553 Abu Dhabi, United Arab Emirates (∗Corresponding author: E-mail: Pitra@IZW-Berlin.de)
Received 31 March 2003; accepted 16 July 2003
Key words: Chlamydotis, microsatellites, mitochondrial DNA, phylogeography, population expansion
Abstract
The houbara bustard (Chlamydotis undulata), the main quarry for Arab falconry, is currently threatened by
excessive hunting and poaching as well as by habitat loss and fragmentation. We have investigated the genetic
diversity, population structure and demographic history of houbara bustards across their geographical range by
analyzing mitochondrial (mt)DNA sequences (a 370 bp fragment of control region I and a 264 bp fragment of the
cytochrome bgene) and 4 microsatellite loci in 74 individuals sampled from the Canary Islands to China. Both
markers revealed low to moderate diversity that could be partitioned into two monophyletic groups or evolutionary
significant units belongingto the North African (C. u. undulata)andAsian(C. u. macqueenii) subspecies. A history
of relatively recent population growth (∼35,000 years ago) accompanied by range expansion is the most likely
demographic scenario for the Asian subspecies. In addition, Macqueen’s bustards are able to disperse efficiently
over broad areas, which is consistent with our inference of weak phylogeographicstructure (global FST =0.20and
0.04 for mtDNA and microsatellites, respectively) and high levels of homogenizing gene flow on wide geographical
scales. We therefore suggest that management actions should focus on maintaining migratory connectivity between
breeding and non-breeding areas.
Introduction
The houbara bustard (Chlamydotis undulata)is
a desert-adapted avian species patchily distributed
across the arid zones of the Northern Hemisphere.
Traditionally, the houbara bustard has been classified
into three subspecies groups based on size, colour,
distribution (Cramp and Simmons 1980), courtship
display (Launay and Paillat 1990; Gaucher et al.
1996), and genetic divergence (Granjon et al. 1994;
Gaucher et al. 1996; D’Aloia 2001; Pitra et al. 2002,
Broders et al. 2003). These subspecies were desig-
nated C. u. fuertaventurae occurring in the eastern
Canary Islands, C. u. undulata occurring from North
Africa and Mauritania to the western part of the
Nile river valley in Egypt, and C. u. macqueenii,
the Asian houbara bustard or Macqueen’s bustard,
extending from the Sinai, across Arabia and the
Caspian Sea, further east to Middle Asia, Mongolia
and China (Figure 1). The North African subspecies
is considered sedentary or locally nomadic, whereas
the northern breeding populations of the Asian subspe-
cies are strongly migratory, wintering mainly in
India, Pakistan, Iran and the Arabian Peninsula.
Given the broad range and the tremendous mobility
of C. u. macqueenii (distances covered range from
1,000 to 14,000 km per year), it remains difficult
to design recovery strategies until demographic inter-
connections between breeding and wintering sites
or along migration routes can be characterised, and
to elucidate factors/processes with the most impact
both on historical and recent population structuring.
Recent tracking studies via satellite have provided
much data on individual movements and migration
routes (Osborne et al. 1997; Launay et al. 1999;
Combreau et al. 1999). However, it is unknown
206
Figure 1. Macqueen’s bustard sample locations and generalized distribution. Larger circles refer to breeding areas of migratory birds. Smaller
circles indicate localities which harbour residents and/or wintering migrants. Frequency of the numerically prominent haplotype M1 (black) in
each population is indicated in pie charts.
whether these movements led to gene flow. We there-
fore conducted a survey on the variability at mitochon-
drial and microsatellite loci among and within eleven
populations of the Macqueen’s bustard to address
issues of phylogeography and population history of
this species. Results from population genetic analyses
of these data are used to draw inferences on past
demographic processes, and to provide conservation
relevant insights.
Material and methods
Population samples
66 Macqueen’s bustard samples were collected from
11 localities over most of the subspecies distribu-
tion range (Figure 1 and Table 1). All samples were
collected either from recently dead animals (tissue
biopsies) or from birds captured as part of ongoing
radio-telemetry projects. In order to obtain intra-
specific divergence estimates, we analysed three and
five specimens of the North African C. u. undu-
lata and Canary Islands C. u. fuertaventurae subspe-
cies, respectively. Two great bustards, representing
different subspecies (O. t. tarda and O. t. dybow-
skii), were also sampled. The specimens are currently
held in the vertebrate collection in the National Avian
Research Center, Abu Dhabi, United Arab Emirates
under the accession numbers HUBA.1 to HUBA.74.
Data collection
Total genomic DNA was extracted using the
method of Müllenbach et al. (1989) or by use of
standard commercial kits (E.Z.N.A.Tissue DNA kit,
PEQLAB). A 264 bp fragment of the mitochon-
drial cytochrome bgene (cyt b) was amplified using
the primers Cb1L and Cb2H (Kocher et al. 1989).
The entire cyt bgene (1143 bp) was amplified with
primers Nd5L1 5GCCTACYTAGGATCITTYGC 3
and Cb6ThrH (Palumbi et al. 1991). All of
these primers and additional Cb3RL (Palumbi et
al. 1991), aTrCb1L, aTrCb3L, and aTrCb8H
(Pitra et al. 2002) were used for the sequen-
cing procedure. For the hypervariable part I
(370 bp) of the mitochondrial control region (HV1),
PCR and sequencing reactions were carried out
207
Table 1. Geographic distribution of composite mtDNA haplotypes, average number of nucleotide differences (k), haplotype diversity (h), and nucleotide diversity (π) within houbara bustard
populations (M = Mori, China; G = Gansu, China; A = Afghanistan; P = Pakistan; I = Iran; O = Oman, SA = Saudi Arabia; UAE = United Arabian Emirates; S = Sinai; EK = East Kazakhstan;
WK = West Kazakhstan; AL = Algeria; CI = Canary Islands)
Haplotype∗Variable sites∗∗ M G A P I O SA UAE S EK WK AL CI Number of
Control region cyt b 45.50 N 39.00 N 31.20 N 28.30 N 31.00 N 18.50 N 31.00 N 23.75 N 31.00 N 45.50 N 52.00 N 33.50 N 28.42 N individuals
11111111111122333 4455 75.00 E 102.70 E 64.00 E 64.85 E 56.00 E 56.00 E 40.50 E 53.30 E 34.50 E 75.00 E 45.00 E 05.00 E 13.98 W
311222334578948256 4726
378348035275565632 8299
M1 CCC-TCTAGCCAGTACTA CCCT 53 15412131 94 48
M2 ...-..........G... .... 22 4
M3 ...-.....T........ .... 11
M4 ...-.........C.... .... 31 1 5
M5 ...-C.......A...C. T... 1 1
M6 T..-.............. .... 11
M7 ...-.............G .... 1 1
M8 ..T-.............. ..T. 33
M9 ...-......T....... .... 22
U1 ...ACT.GA..G.CGTC. .T.C 11
U2 T..ACTCGA..G.CGTC. .T.C 11
U3 ...ACTCGA..G.CGTC. .T.C 11
F1 T..-CTCGA..G.CGTC. .T.C 44
F2 TT.-CTCGA..G.CGTC. .T.C 11
Numberof 94 1851216410635 =74
individuals
k1.39 0.50 — 0.68 0.40 — — 0.36 1.00 0.20 0.53 1.33 0.40
h0.64 0.50 — 0.61 0.40 — — 0.34 0.50 0.20 0.53 1.00 0.40
π0.22 0.08 — 0.11 0.06 — — 0.06 0.16 0.03 0.08 0.21 0.06
∗Lables of haplotypes encountered in each classical subspecies (M = C. u. macqueenii;U=C. u. undulata;F=C. u. fuertaventurae).
∗∗The sequenced fragments correspond to positions 298-669 in control region and 15574-15837 in cytochrome bin the complete Ciconia ciconia mitochondrial genome (AB026818). Dots
indicate identity with the reference sequence (haplotype M1), and letters designate base substitutions. Nucleotide position numbers of the houbara bustard mtDNA correspond to positions in
the alignment.
208
with primers CtrIaL (Martin et al. 2002) and
TetCtrIH 5ATGGCCCTGAAATAGGAACC 3.PCR
and cycling conditions are described in detail in
Martin et al. (2002). All PCR products were
purified (E.Z.N.A. Cycle-Pure kit, PEQLAB), and
directly sequenced using the fluorescent Prism
BigDye Terminator Cycle Sequencing kit (Applied
Biosystems) according to the manufacturers instruc-
tions followed by product separation on automated
310/3100 Genetic Analyzers (Applied Biosystems).
The observation that the entire cyt bsequences
contained no stop codons, and that the nucleotide
diversity in our sequences was 2.5 times higher in the
noncoding HV1 than in the coding cyt bgene render
it unlikely that nuclear copies of mtDNA genes were
sequenced accidentally (Arctander 1995). Sequences
were deposited in the GenBank database (accession
numbers AY078586-AY078659 for the 370 bp frag-
ment of the HV1, AY078581-AY078585 for the
complete cyt bgene, and AY078511-AY078543 for
the 264 bp fragment of the cyt b).
Four microsatellite loci (Otmic26, Otmic33,
Otmic27, and Otmic38) developed originally for the
great bustard and also polymorphic in houbara
bustards were amplified by PCR from houbara bustard
genomic DNA using fluorescently labelled primers.
Loci Otmic26 and Otmic33, corresponding primers
and the standard reaction conditions for all loci
are described in Lieckfeldt et al. (2001). Primers
designed for locus Otmic27 (cloned fragment length
217 bp, repeat (AG)10, GenBank acc.nr. AY173113)
were Otmic27F 5AACCAACATTAAGAGCTTACG
3; Otmic27R 5CAGAGTGAACTTTGCTACAG 3,
and primers for locus Otmic38 (cloned fragment
length 264 bp, repeat (AC)5AA(AC)5,GenBank
acc.nr. AY173114) were Otmic38F 5CTAGTTATC-
CTAAGTAATG 3; Otmic38R 5AAGGACTATGT-
TGACTTTTC 3. The PCR products were sepa-
rated on automated 310/3100 Genetic Analyzers
(Applied Biosystems), scored and analysed using the
computer program GENESCAN 3.1 (Applied Biosys-
tems). The alignments of the combined mtDNA haplo-
types and the microsatellite data are available at
the Population Genetics Database (PGDB) website
(http://seahorse.louisiana.edu/PGDB/).
Data analysis
MtDNA sequences were aligned using CLUSTALX
(Thompson et al. 1997). Initial sequence comparisons
and measures of variability were performed using
MEGA 2.0 (Kumar et al. 1993). Several population
genetic parameters such as nucleotide diversity and
haplotype diversity were estimated from the mtDNA
data set using DNASP 3.0 (Rozas and Rozas 1999).
To assess the extent of differentiation within and
among populations an Analysis of Molecular Vari-
ance (AMOVA) (Excoffier et al. 1992) was used to
estimate FST and M (absolute number of migrants
exchanged between populations) values, whose statis-
tical significance was tested using 10000 permutations
as implemented in ARLEQUIN 2.0 (Schneider et al.
2000). The genealogical relationship between haplo-
types was examined by statistical parsimony (TCS
program, Clement et al. 2000) to depict phylogen-
etic, geographical, and potential ancestor-descendant
relationships among the identified mtDNA haplo-
types. This network construction method defines first
the uncorrected distance above which the parsimony
criterion is violated with more than 5% probability
(parsimony limit). Then, all connections are estab-
lished among haplotypes starting with the smallest
distances and ending either when all haplotypes are
connected or the distance corresponding to the parsi-
mony limit has been reached (Crandall and Templeton
1993).
Inference of past population expansion events was
performed using Fu’s (1997) Fs test of selective
neutrality as implemented in ARLEQUIN. The Fs
statistic is particularly suited to detect departures from
neutrality in nonrecombining sequences characterized
by a high frequency of rare haplotypes, which gener-
ally lead to large negative Fs values (Fu 1997). The
significance of the Fs statistic was tested by gener-
ating 5000 random samples under the hypothesis of
selective neutrality and population equilibrium. Fu’s
Fs statistic should be considered significant if the
P-value is below 0.02 (Fu 1997). The ARLEQUIN
program was also used to calculate values of Tajima’s
(1989) Dstatistic. Negative Dvalues indicate an
excess of rare variants, as can result from a recent
population expansion or processes such as background
selection, whereas positive values indicate an excess
of intermediate-frequency variants. The significance
of the Dstatistic was tested by generating 5000
random samples under the hypothesis of selective
neutrality and population equilibrium.
To accurately estimate the mutation rate and diver-
gence times of bustard mtDNAs, the cyt bgene
of one animal from each lineage was completely
sequenced and a neighbour-joining (NJ) tree from
the Kimura (1980)-corrected distances was recon-
209
structed using MEGA. Divergence times and their
corresponding 95% confidence intervals (CI95%)in
conjunction with a conservative fossil record calibra-
tion were computed using the method of Haubold and
Wiehe (2001) implemented in CITE (downloadable at
http://www.soft.ice.mpg.de/cite).
The software FSTAT version 2.9.2 (updated by
Goudet 1995) was used to estimate microsatel-
lite allele frequencies, observed heterozygosities
(Ho) and expected heterozygosities (Hs) within
samples as well as gene diversity and allelic rich-
ness, r(n), the expected number of alleles in a
standard sample size of n. ARLEQUIN was used
to test for deviations from Hardy–Weinberg and
linkage equilibrium, and to derive estimates of
population subdivision (Michalakis and Excoffier
1996). The statistical significance of FST values
was tested using 10000 permutations. A principal
component analysis (PCA) was performed on the
covariance matrix of untransformed microsatellite
frequency data to investigate spatial patterns of
genetic variation. We used the computer package
PCAGEN written by J. Goudet (downloadable at
http://www.unil.ch/izea/softwares/pcagen.html). An
important characteristic of this program is that it tests
the significance of the total inertia as well as indi-
vidual PCA axes inertia by using a randomization
procedure (Manly 1997). Therefore, it allows us to
avoid the interpretation of nonsignificant axes. We
performed 1000 randomizations of genotypes to test
for significance of individual axes inertia. Genotypes
are permuted among samples and a PCA is realized on
each permuted data set. The observed value (propor-
tion of inertia per axis) is compared to the distribu-
tion of the values obtained from the randomized data
sets. The proportion of values larger or equal to the
observed one is an unbiased estimate of the P-value of
the test (J. Goudet, pers. comm.).
Results
mtDNA sequence variation
The two combined mtDNA regions (a total of 634 bp)
of 74 houbara bustards revealed 22 variable sites (17
transitions and one indel in HV1 and four transitions
in cyt b), which defined 14 haplotypes (Table 1).
Using statistical parsimony network analysis, North
African and Asian houbara bustards assorted into two
separate haplogroups with a division of eleven steps
between the C. u. undulata and C. u. macqueenii
group (Figure 2A). The only uncertainty found in the
cladogram was the connection between these groups,
where two equidistant alternatives exist. The haplo-
types representative of C. u. fuertaventurae clustered
closely to C. u. undulata, differing by only a single site
in the combined mtDNA sequence. AMOVA results
confirmed the distinctive undulata –macqueenii split
with 94.3% of the total variation resulting from differ-
ences between the subspecies and 4.6% of the vari-
ation from within the subspecies (p= 0.005). Within
the C. u. macqueenii haplogroup, a star-like pattern
was identified in the haplotype network constructed
using the program TCS (Figure 2A). Using criteria
outlined in Crandall and Templeton (1993), haplotype
M1 was designated as ancestral. All sequences rooted
back to haplotype M1 that was also the most common
and geographically widespread haplotype depicted in
the centre of the network surrounded by rare haplo-
types differing by a small number of mutational
steps. Both the numerical (pooled frequency of 72%)
and topological (out group weight of 87%) predomi-
nance of haplotype M1 were indicative for its interior
(presumably ancestral) status. Furthermore, the occur-
rence of haplotype M1 in all populations sampled
(Figure 1) and the low divergence between haplotypes
resulted in relatively small nucleotide diversity esti-
mates (Table 1) and a weak phylogeographicstructure.
AMOVA revealed that a large percentage of the total
mtDNA variation in C. u. macqueenii was distrib-
uted within populations (79.87%) and a small but
significant percentage was found among populations
(20.13%; P<0.01). Grouping locations in different
ways in hierarchical AMOVAs did not increase the
global FST . This degree of subdivision is expected
if there are on average >9 female migrants (in the
genetic sense) entering each population each genera-
tion, assuming an island model of migration (Whitlock
and McCauley 1999). On the basis of pairwise FST
estimates there was no significant genetic differen-
tiation between most of the populations (Table 2).
The Sinai population was genetically most divergent
from other populations, with a mean pairwise FST of
60.7%. Because this amount of subdivision represents
approximately 0.35 migrants entering each popula-
tion per generation, the Sinai Peninsula population
at the western end of the current range, is likely
to be demographically more isolated from the other
populations.
210
Figure 2. (A) Statistical parsimony network of houbara bustard mtDNA haplotypes (numbered as in Table 1). The size of circle is proportional
to the number of individuals found with that haplotype. Small open circles indicate missing haplotypes. The broken lines represent possible
evolutionary pathways. (B) Time-corrected NJ tree on the entire cytochrome bgene sequence obtained from different bustard taxa. The nodes
are arranged by the age estimated from the branch length. Confidence intervals (95%) for dates are indicated by grey bars. The reference node
() has a minimum age given by the fossil record (see text).
Table 2. Genetic differentiation among Macqueen’s bustard populations. Population labels are the same as listed in Table 1
MG WKEK P I UAES
M 0.000 0.105 0.106∗0.028 0.033 0.141∗0.451∗
G 0.000 0.111 0.081 0.000 0.006 0.055 0.571
WK 0.000 0.013 0.198 0.138∗0.119 0.174∗0.606∗
EK 0.000 0.033 0.008 0.093 0.040 0.017 0.727∗
P 0.079∗0.118∗0.124∗0.037 0.040 0.000 0.577∗
I 0.000 0.021 0.000 0.000 0.062 0.608
UAE 0.018 0.086∗0.024 0.000 0.033 0.000 0.697∗
S 0.041 0.052 0.000 0.000 0.115∗0.008 0.073
Upper triangular matrix = FST estimates from mtDNA data; lower triangular matrix = FST estimates from microsatellite data. ∗Significant
differentiation at the 95% level.
211
Table 3. Measures of diversity at four microsatellite loci in Houbara bustard populations (= 3 individuals). Population labels
are the same as listed in Table 1
Loci Population
EKPI M&GUAEWKSALCI
No. individuals 10 8 5 13 16 6 4 3 4
No. of alleles 26 4 3 4 4 5 3 3 2 2
33433 4 44312
27221 1 31223
38222 3 34321
Allelic richness 26 2.84 2.35 3.07 2.90 3.12 2.76 2.71 2.00 2.00
33 2.93 2.57 2.47 2.92 2.67 3.05 2.75 1.00 1.96
27 1.79 1.97 1.00 1.00 1.66 1.00 1.96 2.00 2.75
38 1.87 1.63 1.87 1.86 1.86 2.99 2.75 2.00 1.00
Gene diversity 26 0.61 0.57 0.68 0.63 0.71 0.67 0.58 0.67 0.50
33 0.69 0.61 0.53 0.67 0.59 0.65 0.63 0.00 0.50
27 0.34 0.50 0.00 0.00 0.23 0.00 0.42 0.50 0.71
38 0.39 0.23 0.35 0.31 0.33 0.73 0.71 0.50 0.00
Hobserved all 0.50 0.50 0.30 0.52 0.47 0.47 0.69 0.33 0.29
Hexpected all 0.48 0.45 0.34 0.49 0.45 0.46 0.52 0.33 0.33
Microsatellite variation
The four microsatellites were at least marginally poly-
morphic in all populations, and 6, 4, 4 and 5 alleles
were detected at the loci Otmic26, 33, 27, and
38, respectively (Table 3). The sizes of alleles in
the most polymorphic locus, Otmic26, ranged from
273 to 291 bp with alleles differing by 2–18 bp.
No locus showed significant departure from Hardy–
Weinberg and linkage equilibrium in populations
(≥3 individuals) although the most localities prob-
ably contained not enough individuals. The levels of
genetic polymorphism estimated from the microsatel-
lite data were relatively similar for all nine populations
(Table 3). To examine the overall pattern of population
differentiation, PCA was conducted with the first two
axes (PC1 and PC2), which cumulatively explained
64.3% of the total inertia contained in the data set
(Figure 3). The portions of inertia associated with
the first two axes were significant (P= 0.001). From
this plot it can be seen that all the local macqueenii
populations clustered together, well separated from
the undulata and fuertaventurae populations. Thus,
the pattern of large-scale geographical differentiation
in microsatellites was generally congruent with that
revealed by mtDNA.
There were eight alleles found exclusively within
C. u. macqueenii, but none of them was specific for
any of the seven macqueenii populations listed in
Table 3. As for mtDNA, a series of AMOVAs was
performed for pairwise comparisons (Table 2) and
different hypothesized geographical partitions. Aver-
aged over all four microsatellite loci, there was a
low yet significant genetic differentiation among C. u.
macqueenii populations (FST = 0.0419; P<0.001),
ranging from 0.0142 for locus Otmic26 to 0.1607 for
locus Otmic27. Several possible scenarios of popula-
tion subdivision were considered, but could not be
thoroughly assessed at the present time due to limit-
ations in sample size for particular areas.
Population expansion
The star-like pattern of derived variants surrounding
the most common haplotype M1 (Figure 2A) indicated
a history of recent ancestral monomorphism followed
by a population expansion (Slatkin and Hudson 1991;
Troy et al. 2001). Such a demographic event was
supported by the results of two other treatments of
the data. First, Fu’s FSstatistic, which is particularly
sensitive to population growth (Fu 1997) yielded a
significantly large negative value (FS= –5.42; P=
0.0025) for the entire Macqueen’s bustard popula-
tion, indicating an excess of low frequency haplotypes
(8) as compared with the expected number (3.81) in
a stationary population and thus providing evidence
212
Figure 3. Principal component analysis on four autosomal microsatellites allele frequency data. The proportion of inertia of the first two axes
is significant (P= 0.001). Plot names are as follows: East Kazakhstan (1), Pakistan (2), Iran (3), China (4), UAE (5), West Kazakhstan (6),
Sinai (7), Algeria (8), and Canary Islands (9). The ellipse (dotted) was drawn by hand to include the distributions of the Macqueen’s bustards
sampled in different localities. Subspecies designations are indicated.
for recent population expansion. However, when the
Chinese population was analysed separately, the FS
value became positive (0.93; P= 0.687), while that for
all other populations remained negative (–5.36; P=
0.001). Second, Tajima’s test of selective neutrality
also yielded significantly large negative Dvalues
for the entire Macqueen’s bustard population (–1.91;
P= 0.008) and for the population outside of China
(–2.12; P= 0.002), but not for the Chinese population
(–1.18; P= 0.12). Hence, the Chinese population itself
seems to be relatively stable, whereas the populations
outside of China show evidence of recent demographic
expansion.
Estimating date of expansion
We determined the nucleotide differences between
mtDNAs of the genus Chlamydotis and the previously
identified sister-genus Otis (Pitra et al. 2002) in an
attempt to estimate most accurately the mutation rate
and divergence times of bustard mtDNAs. This was
done by inferring the required information from the
NJ tree shown in Figure 2B. Under the assumptions
that Chlamydotis and Otis bustards diverged about
4 million years ago (Sanchez Marco 1990; Kretzoi
1960/61; Bochenski and Kurochkin 1987) and that
the mutation rate remained relatively constant, we
obtained conventional mutation rates of 0.89 ×10−8
and 2.25 ×10−8/ site per year for the cyt bgene
and the HV1, respectively. Using both mutation rates,
we inferred the age of the last common ancestor of
the mtDNAs obtained from the Macqueen’s bustards
at 32.8 thousand years (32.8 kyr) (CI95% =5.01–
87.30 kyr) for the cyt bgene and at 36.1 kyr (CI95% =
13.8 – 70.2 kyr) for the HV1 region. Although a wide
error is associated with these estimates, any error in
the clock calibration would only affect the absolute
time estimates, but not the chronological order of
evolutionary processes.
Discussion
Observed levels of genetic diversity in Macqueen’s
bustard were low for the mtDNA data, and medium
in terms of microsatellite variability when compared
to many other avian taxa (Moore 1995; Primmer and
Ellegren 1998). However, if only the 370 bp long
HV1 fragment is considered also used by Martín et al.
(2002) in the great bustard (Otis tarda), Macqueen’s
bustards show similar values in several measures of
diversity (e.g., 0.6% versus 0.8% sequence divergence
among haplotypes, 0.0015 versus 0.0050 nucleotide
diversity) relative to the great bustard. The microsatel-
lite variation observed in Macqueen’s bustard is also
comparable to that of great bustards, for which these
213
markers were originally designed (Lieckfeldt et al.
2001). Taken together, these results suggest a fairly
recent divergence of the present-day Macqueen’s
bustard lineages or a recent origin of all extant
Macqueen’s bustards from an ancestral population
which in itself was limited in diversity, perhaps
as a result of climatically mediated range restric-
tion. The pattern observed in the mtDNA network
of Macqueen’s bustard is consistent with a recent
population expansion, with little divergence among
haplotypes (Figure 2A) and insufficient time and/or
isolation to generate regional differentiation. The
inference of a population expansion in the rela-
tively recent past is also supported by significantly
negative Fu’s Fs and Tajima’s Destimates. The
overall coalescence date of all sampled Macqueen’s
bustard mtDNA sequences around the central haplo-
type M1 was estimated at 32.8–36.1 kyr ago, indi-
cating a population expansion around or before this
time. These molecular dates should also be confirmed
in the future using other dating methods, such as
for example the Bayesian approach of Thorne et al.
(1998).
Analyses of the genetic structure of Macqueen’s
bustard populations showed no evidence of major
geographical partitions, old subdivision events, or
complete barriers to historical gene flow (Table 2).
This is consistent with findings of D’Aloia (2001)
that were based on RAPD analysis using ten random
decamer primers. The only evidence for statistically
significant phylogeographical differentiation seems to
support the traditional division of houbara bustard
subspecies (Cramp and Simmons 1980; Collar 1996).
The historical separation and geographical distribution
of the African and Asian houbara bustards revealed by
both marker types appears concordant.
The observed patterns of low genetic variability
and weak population differentiation can be interpreted
as being caused by a combination of several demo-
graphic processes, starting with a relatively recent
expansion of the modern Macqueen’s bustard as a
whole, so that sufficient time for extensive genetic
differentiation has not yet accumulated. In addi-
tion, Macqueen’s bustards are able to disperse effi-
ciently over broad areas, which is consistent with
our inference of high levels of homogenizing gene
flow on wide geographical scales. Recent migra-
tion studies have shown that Macqueen’s bustards
from different breeding populations share the same
wintering grounds (Launay et al. 1999), and birds
from the same breeding populations are wintering in
different places (Combreau et al. 1999). Therefore, the
observed distribution of haplotypes may be attributed
to populations in panmixia or to long distance gene
flow between populations over the recent past. With
a larger sample size and a higher level of polymorph-
isms it might be possible to discriminate between these
two possibilities.
Conservation implications
The houbara bustard (Chlamydotis undulata)isthe
prime quarry for traditional Arab falconry with
diametrical consequences: steady decline in numbers
during the 20th century, mainly due to over-hunting
and poaching (Collar 1996; Goriup 1997; Combreau
et al. 2001) on one hand, and extensive conserva-
tion efforts, including ecological research, captive-
breeding and reintroduction/restocking programmes
as well as habitat protection (Combreau et al. 1995;
Seddon et al. 1995) on the other. The current status
of the houbara bustard is lower risk – near vulner-
able, and the population trend is listed as deteri-
orating (Hilton-Taylor 2000), based on observed
declines in many populations. In addition, our data
provide support for the existence of at least two
major geographical partitions defining old and isolated
groups which could be viewed as subspecies or
Evolutionary Significant Units (sensu Moritz 1994).
Our results are compatible with currently favoured
strategies for Macqueen’s bustard conservation on a
broad regional basis (Fox et al. 2000), with large-
scale biomes or ecosystems as operational manage-
ment units. This strategy is advisable both in terms
of maintaining Macqueen’s bustards as an important
component of functional desert ecosystems and as an
intricate part of the Arab heritage. To preserve the
evolutionary potential of the migratory Macqueen’s
bustard and to increase its chances of long-term
survival, it will be important to manage breeding
populations separately and to raise caution when
issuing harvest quotas in these areas. Finally, because
of continued reduction and isolation of remaining
Macqueen’s bustard populations, it is important to
consider the inference presented here for high histor-
ical levels of gene flow over broad areas. Thus,
our results encourage management alternatives that
maintain these original and likely adaptive popula-
tion dynamics, possibly including joint international
efforts and intervention to ensure an intermediate level
of gene flow by annual long-distance migration among
remnant populations.
214
The authors are grateful to the Environmental
Research and Wildlife Development Agency
(ERWDA) for the financial support of the project and
ERWDA’s Management for their support and interest
in this study. For providing tissue samples we thank
J.C. Alonso, E.C. Vidal, A. Martin, G. Diaz, J.L.
Rodriguez and J.C. Cillera from Spain, H. Litzbarski
from Germany, M. Yiqing and T. Xiuhua from China,
S. Chan from Japan, S. Hemon from Saudi Arabia,
and F. Launay and X. Eichaker from the United
Arab Emirates. We thank A. Schmidt for technical
assistance. We also thank J. Fickel, the Associate
Editor, and two anonymous reviewers for enhancing
the manuscript.
References
Arctander P (1995) Comparison of a mitochondrial gene and a
corresponding nuclear pseudogene. Proc. Roy. Soc. Lond. B,262,
13–19.
Bochenski Z, Kurochkin EN (1987) Pliocene bustards (Aves:
Otididae and Gryzajidae) of Moldavia and S. Ukraine. In:
L’Evolution des oiseaux dþapres le temoignage des fossiles
(ed Mourer-Chauvire C), pp. 173–186. Docum. Lab. Géol. 99,
Lyon.
Broders O, Osborne T, Wink M (2003) A mtDNA phylogeny of
bustards (family Otididae) based on nucleotide sequences of the
cytochrome b-gene. J. Ornithol.,144, 176–185.
Clement M, Posada D, Crandall KA (2000) TCS: a computer
program to estimate gene genealogies. Mol. Ecol.,9, 1657–1659.
Collar NJ (1996) Family Otididae (bustards). In: Handbook of the
Birds of the World, Vol. 3. (eds. Del Hoyo J, Elliot A, Sargatal J),
pp. 240–273. Lynx Edicions, Barcelona.
Combreau O, Saint-Jalme M, Seddon P et al. (1995) A program
for houbara bustard restoration in Saudi Arabia. In: Integrating
People and Wildlife for a Sustainable Future (eds. Bissonette
JA, Kraussman PR), pp. 520–524. The Wildlife Society USA,
Bethesda.
Combreau O, Launay F, Al Bowardi M, Gubin B (1999) Outward
migration of houbara bustards Chlamydotis undulata macqueenii
from two breeding areas in Kazakhstan. Condor,101, 159–164.
Combreau O, Launay F, Lawrence P (2001) An assessment
of annual mortality rates in adult Asian houbara bustards
(Chlamydotis [undulata] macqueenii). Anim. Conserv.,4, 133–
141.
Cramp S, Simmons K (1980) Handbook of the Birds of Europe, the
Middle East and North Africa, Vol. 2. Oxford University Press,
Oxford.
Crandall KA, Templeton AR (1993) Empirical tests of some predic-
tions from coalescent theory with applications to intraspecific
phylogeny reconstruction. Genetics,134, 959–969.
D’Aloia M-A (2001) Studies on the population structure of the
houbara bustard Chlamydotis undulata in the Middle East with
DNA analysis techniques. Zool. Middle East,22, 25–35.
Excoffier L, Smouse P, Quattro J (1992) Analysis of molecular
variance inferred from metric distances among DNA haplo-
types: application to human mitochondrial DNA restriction data.
Genetics,131, 479–491.
Fox N, Al Bowardi M, Macdonald H, Launay F (2000) A Global
Strategy for the Conservation of Falcons and Houbara. Environ-
mental Research and Wildlife Development Agency, Abu Dhabi.
Fu Y-X (1997) Statistical tests of neutrality of mutations against
population growth, hitchhiking and background selection.
Genetics,147, 915–925.
Gaucher P, Paillat P, Chappuis C, Saint Jalme M, Lotfikhah F,
Wink M (1996) Taxonomy of the houbara bustard Chlamydotis
undulata subspecies considered on the basis of sexual display
and genetic divergence. Ibis,138, 273–282.
Goudet J (1995) fstat Version 1.2.: a computer program to calculate
F-statistics. J. Hered. ,4, 485–486.
Goriup PD (1997) The world status of the houbara bustard
Chlamydotis undulata.Bird Conserv. Inter.,7, 373–397.
Granjon L, Gaucher P, Greth A, Paillat P, Vassart M (1994)
Allozyme study of two subspecies of houbara bustard
(Chlamydotis undulata undulata and C. u. macqueenii).
Biochem. Syst. Ecol.,22, 775–779.
Haubold B, Wiehe T (2001) Statistics of divergence times. Mol.
Biol. Evol.,18, 1157–1160.
Hilton-Taylor C (2000) IUCN Red List of Threatened Species.
IUCN, Gland, Switzerland & Cambridge, UK.
Kocher TD, Thomas WK, Meyer A et al. (1989) Dynamics of
mitochondrial DNA evolution in mammals: amplification and
sequencing with conserved primers. Proc. Nat. Acad. Sci. USA,
86, 6196–6200.
Kimura M (1980) A simple method for estimating evolutionary rates
of base substitutions through comparative studies of nucleotide
sequences. J. Mol . Evol.,16, 111–120.
Kretzoi M (1960/61) Prähistorischer Großtrappenfund und die
Geschichte der Trappen. Aquila,67–68, 189–190.
Kumar S, Tamura K, Nei M (1993) MEGA: Molecular Evolutionary
Genetics Analysis, Version 1.02. Pennsylvania State University,
University Park.
Launay F, Paillat P (1990) A behavioural repertoire of the adult
houbara bustard (Chlamydotis undulata macqueenii). Rev. Ecol.
(Terre Vie), 45, 65–88.
Launay F, Combreau O, Al Bowardi M (1999) Annual migration
of houbara bustard Chlamydotis undulata macqueenii from the
United Arab Emirates. Bird Conserv. Inter.,9, 155–161.
Lieckfeldt D, Schmidt A, Pitra C (2001) Isolation and characteriz-
ation of microsatellite loci in the great bustard, Otis tarda.Mol.
Ecol. Notes,3, 133–134.
Manly BFJ (1997) Randomization, Bootstrap and Monte Carlo
Methods in Biology, 2nd edn. Chapman & Hall, London.
Martín CA, Alonso JC, Alonso J, Pitra C, Lieckfeldt D (2002) Great
bustard population structure in central Spain: concordant results
from genetic analysis and dispersal study. Proc. Roy. Soc. Lond.
B,269, 119–125.
Michalakis Y, Excoffier L (1996) A generic estimation of popula-
tion subdivision using distances between alleles with special
reference to microsatellite loci. Genetics,142, 1061–1064.
Moore WS (1995) Inferring phylogenies from mtDNA variation:
mitochondrial-gene trees versus nuclear-gene trees. Evolution,
49, 718–726.
Moritz C (1994) Defining ‘Evolutionarily Significant Units’ for
conservation. Trends Ecol. Evol.,9, 373–375.
Müllenbach R, Lagoda PJL, Welter C (1989) An efficient salt-
chloroform extraction of DNA from blood and tissues. Trends
Genet.,5, 391.
Osborne PE, Al Bowardi M, Bailey TA (1997). Migration of the
houbara bustard Chlamydotis undulata from Abu Dhabi to Turk-
menistan: the first results from satellite tracking studies. Ibis,
139, 192–196.
215
Palumbi SR, Martin A, Romano S et al. (1991) The Simple Fool’s
Guide to PCR. University of Hawaii, Department of Zoology,
Honolulu.
Pitra C, Lieckfeldt D, Frahnert S, Fickel J (2002) Phylogenetic
relationships and ancestral areas of the bustards (Gruiformes:
Otididae), inferred from mitochondrial DNA and nuclear intron
sequences. Mol. Phyl. Evol.,23, 63–74.
Primmer CR, Ellegren H (1998) Patterns of molecular evolution in
avian microsatellites. Mol. Biol. Evol.,15, 997–1008.
Rozas J, Rozas R (1999) DnaSPversion 3: an integrated program for
molecular population genetics and molecular evolution analysis.
Bioinformatics,15, 174–175.
Sanchez Marco A (1990) A new bustard (Otididae; Aves) from the
early Pliocene of Layna (Soria, Spain). Paleontol. Evol.,23, 223–
229.
Seddon PJ, Saint-Jalme M, Van Heezik Y, Paillat P, Gaucher P,
Combreau O (1995) Restoration of Houbara bustard populations
in Saudi Arabia: developments and future directions. Oryx,29,
136–142.
Schneider S, Kueffer JM, Roessli D, Excoffier L (2000) ARLE-
QUIN, Version 2.01: A Software for Population Genetic Data
Analysis. University of Geneva, Geneva.
Slatkin M, Hudson RR (1991) Pairwise comparisons of mito-
chondrial DNA sequences in stable and exponentially growing
populations. Genetics,129, 555–562.
Tajima F (1989) Statistical method for testing the neutral mutation
hypothesis by DNA polymorphism. Genetics,123, 585–595.
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG
(1997) The ClustalX windows interface: flexible strategies for
multiple sequence alignment aided by quality analysis tools.
Nucleic Acids Res.,25, 4876–4882.
Thorne JL, Kishino H, Painter IS (1998) Estimating the rate of
evolution of the rate of molecular evolution Mol. Biol. Evol., 15,
1647–1657.
Troy CS, MacHugh DE, Bailey JF et al. (2001) Genetic evidence
for Near-Eastern origins of European cattle. Nature,410, 1088–
1091.
Whitlock M, McCauley D (1999) Indirect measures of gene flow
and migration: Fst not equal to 1/(4Nm-1). Heredity,82, 117–
125.
