Content uploaded by Carlos Yamashita
Author content
All content in this area was uploaded by Carlos Yamashita on Apr 14, 2014
Content may be subject to copyright.
ORIGINAL PAPER
Genetic variation and population structure
of the endangered Hyacinth Macaw
(Anodorhynchus hyacinthinus): implications
for conservation
Patrı
´
cia J. Faria Æ Neiva M. R. Guedes Æ Carlos Yamashita Æ
Paulo Martuscelli Æ Cristina Y. Miyaki
Received: 7 June 2007 / Accepted: 10 December 2007 / Published online: 1 January 2008
Ó Springer Science+Business Media B.V. 2007
Abstract The Hyacinth Macaw (Anodorhynchus hyacinthinus) is one of 14 endangered
species in the family Psittacidae occurring in Brazil, with an estimated total population of
6,500 specimens. We used nuclear molecular markers (single locus minisatellites and
microsatellites) and 472 bp of the mitochondrial DNA control region to characterize levels
of genetic variability in this species and to assess the degree of gene flow among three
nesting sites in Brazil (Pantanal do Abobral, Pantanal de Miranda and Piauı
´
). The origin of
five apprehended specimens was also investigated. The results suggest that, in comparison
to other species of parrots, Hyacinth Macaws possess relatively lower genetic variation and
that individuals from two different localities within the Pantanal (Abobral and Miranda)
belong to a unique interbreeding population and are genetically distinct at nuclear level
from birds from the state of Piauı
´
. The analyses of the five apprehended birds suggest that
the Pantanal is not the source of birds for illegal trade, but their precise origin could not be
assigned. The low genetic variability detected in the Hyacinth Macaw does not seem to
pose a threat to the survival of this species. Nevertheless, habitat destruction and nest
poaching are the most important factors negatively affecting their populations in the wild.
The observed genetic structure emphasizes the need of protection of Hyacinth Macaws
from different regions in order to maintain the genetic diversity of this species.
P. J. Faria C. Y. Miyaki
Departamento de Gene
´
tica e Biologia Evolutiva, Instituto de Biocie
ˆ
ncias, Universidade de Sa
˜
o Paulo,
Rua do Mata
˜
o 277, CEP 05508-090 Sa
˜
o Paulo, SP, Brasil
P. J. Faria (&)
School of Biosciences, Cardiff University, Cardiff CF10 3US, UK
e-mail: patriciajfaria@yahoo.com.br; fariap@cf.ac.uk
N. M. R. Guedes
Projeto Arara Azul/UNIDERP, Rua Klaus Sturhk 178, Campo Grande, MS, Brasil
C. Yamashita
IBAMA – Gere
ˆ
ncia de Sa
˜
o Paulo, Nu
´
cleo de Fauna, Alameda Tiete
ˆ
637, Sa
˜
o Paulo, SP, Brasil
P. Martuscelli
Instituto Insularis, Rua Gravata
´
387, Condomı
´
nio Quinta da Boa Vista, Mairipora
˜
, SP, Brasil
123
Biodivers Conserv (2008) 17:765–779
DOI 10.1007/s10531-007-9312-1
Keywords Anodorhynchus hyacinthinus Hyacinth Macaws Macaws Parrots
Psittacidae Genetic variation
Introduction
The Hyacinth Macaw (Anodorhynchus hyacinthinus) is the largest macaw in the world
(Forshaw 1989). It is listed in the Appendices I and II of The Convention on International
Trade in Endangered Species of Wild Fauna and Flora (CITES) and it is considered
endangered by the International Union for the Conservation of Nature and Natural
Resources (IUCN; BirdLife International 2004). The genus Anodorhynchus is currently
represented by three species: Glaucous macaw (A. glaucus), Lear’s macaw (A. leari) (both
critically endangered, BirdLife International 2004), and Hyacinth Macaw (A. hyacinthinus,
endangered, BirdLife International 2004).
Hyacinth macaws are highly specialized in their diet and choice of nest sites (Guedes and
Harper 1995). In the Pantanal, they mainly feed on nuts from two palm trees: acuri (Schelea
phalerata) and bocaiu
´
va (Acrocomia totai) and about 95% of their nests are found in
manduvi trees (Sterculia apetala) (Guedes 1993). Nevertheless, in other regions of Brazil,
such as in the northeast, it also breeds on cliffs (Collar 1997). These macaws have low
reproductive rates, they breed every two years and on average, only one chick per pair
survives as a high percentage of eggs and chicks are predated (Munn et al. 1989; Guedes
1993; Pinho and Nogueira 2003). These characteristics contribute to a slow potential for
population recovery in the wild. Populations are mostly threatened by human-related factors,
such as habitat destruction, cattle overgrazing, hardwood logging, African grass cropping,
poaching and trapping of wild birds; in addition, in northern Brazil, they are hunted by
indigenous people for food and handcrafts (Munn et al. 1989; Guedes and Harper 1995).
The Hyacinth Macaw occurs in Bolivia, Paraguay and Brazil (which has the largest
population) (Collar 1997). Currently, there are about 6,500 Hyacinth Macaws living in
three separate areas in Brazil (Guedes 2004): approximately 5,000 individuals in the
Pantanal Matogrossense, 1,000 in Gerais (states of Maranha
˜
o, Piauı
´
, Bahia, Tocantins,
Goia
´
s and Minas Gerais) and 500 in the Amazon Region (Para
´
and Amazonas states)
(Munn et al. 1989; Collar et al. 1992; Sick 1997; Guedes 2004).
Effective species conservation programs require the evaluation of levels of intra-specific
genetic diversity (Haig 1998
). To examine the genetic variability in parrot and macaw
species and consequently help their conservation programs, several molecular markers for
population genetic analyses of this group of birds were selected in a previous study (Faria
and Miyaki 2006). So far, no information is available about the distribution of the genetic
variation in Hyacinth Macaws and our work aimed to fill this gap.
In this study, for the first time, we analyze genetic diversity in Hyacinth macaws from
three different areas in Brazil (Abobral and Miranda—both in the Pantanal—and Piauı
´
)
using three kinds of molecular markers (two single locus minisatellite probes, two
microsatellite loci and 472 bp of mitochondrial control region). We aim: (1) to characterize
the levels of genetic variability of three Hyacinth Macaw populations in Brazil, (2) to
assess the level of genetic structure among the two regions in Pantanal (Abobral and
Miranda) and Piauı
´
, (3) to verify the utility of our data set to identify the origin of five
seized birds, and (4) to use our results for conservation programs. We predict that the
genetic variation found in Hyacinth Macaws will be lower than that observed in other
species of birds that are not threatened and/or have larger effective population sizes. We
also expect to find no evidence of gene flow between the Pantanal and Piauı
´
due to habitat
766 Biodivers Conserv (2008) 17:765–779
123
fragmentation and philopatric behavior. We discuss the conservation implications of our
results and their forensic use.
Material and methods
Forty-three individuals were sampled in the wild and in captivity: 16 individuals from
Abobral (19°22
0
S, 57°02
0
W), 16 from Miranda (19°56
0
S, 57°20
0
W), both these regions
are within the Pantanal Matogrossense in Mato Grosso do Sul state, six from Sa
˜
o Gonc¸alo
da Gurgue
´
ia (10°01
0
S, 45°18
0
W) in Piauı
´
state and five specimens from a seizure by the
Brazilian wildlife authority (IBAMA) in January 2000 in Bahia state. The geographical
distribution of Hyacinth Macaws in Brazil and the sampling sites are presented in Fig. 1.
Blood samples (100 ll from the brachial wing vein) were obtained from nestlings and
only one chick per nest was used for the analysis. The samples were kept at -20°Cin
absolute ethanol and total DNA was extracted using the method described in detail by
Bruford et al. (1992).
Single locus minisatellites
For each bird, 6 lg of genomic DNA were digested with MboI. The DNA fragments were
transferred onto a nylon membrane by Southern blotting (Sambrook et al. 1989). The
Fig. 1 Geographical distribution of Hyacinth Macaw in three regions in Brazil: Amazon, Gerais and
Pantanal. Black triangles show the sampling sites: (1) Piauı
´
, (2) Pantanal do Abobral and (3) Pantanal de
Miranda. Map modified from BirdLife International (2006) (available at: http://www.birdlife.org) according
to suggestions of Paulo T. Z. Antas
Biodivers Conserv (2008) 17:765–779 767
123
membranes were hybridized with two single locus minisatellite probes separately, selected
in a previous work (for further details see Faria and Miyaki 2006). The probes were labeled
with [a P
32
] dCTP by random priming and hybridization was carried out at 65°C for 24 h
(in a solution of Na
2
HPO
4
0.263 M; EDTA 1 mM; SDS 7% and BSA 1%). After the
washes, the filters were X-rayed at -70°C using X-ray films (Kodak RX) with intensifying
screens for 2–7 days.
Microsatellites
Two polymorphic microsatellite loci previously tested in this species (for details see Faria
and Miyaki 2006) were amplified with the following primer pairs: HYA1172F (GAT
CCTTTGCTTAAGACAGATGTC) and HYA1172R (GAGTGAAATACACATTCAGC
TTCTG); and MAC436F (GCACCAAACACAACATCTTATTC) and MAC436R
(TTGGGACACCAATGTAATTTG) (Scott Davis, pers. comm.). The amplifications were
carried out using 35 cycles of 95°C for 60 s, 51–52°C for 30 s and 72°C for 40 s and the
PCR products were separated in 6.5% acrylamide gels. Genotyping was performed through
two different procedures: direct incorporation of [a P
32
] dCTP during PCR and silver
staining of polyacrylamide gel.
Mitochondrial DNA
Amplification of the first domain of the control region was performed in 10 ll volumes of:
19 amplification buffer (Pharmacia); 0.2 mM of each dNTP, 0.75 units of Taq DNA
polymerase (Pharmacia), 10 pmol of each primer (CR522Rb: TGGCCCTGACYTAG
GAACCAG and LGlu: GCCCTGAAARCCATCGTTG; Eberhard et al. 2001) and 20 ng
of genomic DNA. A cycle of 95°C for 40 s, 53°C for 20 s and 72°C for 30 s was repeated
35 times.
The amplification products were checked using 1% agarose gels. Afterwards, the
products were purified with 10 units of exonuclease I and 1 unit of shrimp alkaline
phosphatase (Amersham/Pharmacia). Approximately 60 ng of the purified PCR product
was sequenced using a Big Dye Terminator sequencing kit (Perkin Elmer) following the
manufacturer’s instructions and were loaded in an automated sequencer ABI 377 (Applied
Biosystems). The sequences were edited in the program Sequence Navigator version 1.0
(Applied Biosystems).
Population analysis
ARLEQUIN (version 2.0; Schneider et al. 2000) was utilized to estimate mean observed
and expected heterozygosities and allele frequencies, and to detect linkage disequilibrium
and deviations from the Hardy–Weinberg equilibrium in the nuclear DNA data (mini and
microsatellites). Population differentiation was assessed through pairwise comparisons of
F
ST
values. A Bayesian maximum likelihood approach to infer the number of populations
without prior information of the sampling location was used through the software
STRUCTURE (version 2.2; Pritchard et al. 2000; Falush et al. 2003). We assumed the
admixture model with correlated allele frequencies, a burn-in of 50,000 interactions and
50,000 MCMC replications after burn-in. The program was run 10 independent times for
768 Biodivers Conserv (2008) 17:765–779
123
each number of populations (K) between 1 and 4. The significance of the results was
estimated as described by Evanno et al. (2005) through the plot of the mean likelihood
values (Ln Pr(X/K)) over 10 runs for each K and the rate of change of the likelihood
function with respect to K (DK). Assignment tests were also conducted using STRUC-
TURE and the proportion of individuals assigned to the genetic clusters was estimated.
We combined mini and microsatellite results into a single dataset, and confined our
analysis to allele frequency-based methods since the statistical power gained by combining
markers is expected to outweigh the disadvantage of assuming a single evolutionary model
for the nuclear markers (Nichols and Beaumont 1996).
The comparisons of heterozygosity values were carried out according to Nei and Kumar
(2000). First, we calculated the difference in gene diversity for the ith locus between
populations X and Y (d
i
= h
xi
- h
yi
) for all polymorphic loci. The mean (d) and the
variance [V(d)] of d
i
were estimated by the following formulae:
d ¼
X
L
0
i¼1
d
i
=L
0
VðdÞ¼
X
L
0
i¼1
ðd
i
dÞ
2
=½L
0
ðL
0
1Þ
where L
´
is the number of polymorphic loci.
Further, the significance of any difference of the values of heterozygosity in different
populations was tested using a t-test:
t
L1
= d/s(d)
where L - 1 is the degrees of freedom, s(d) is the square root of V(d).
ARLEQUIN was also used to estimate the haplotype (H) and nucleotide diversity (p)
(Nei 1987) of the mitochondrial DNA data. Population differentiation was assessed through
pairwise comparisons of U
ST
values and hierarchical analyses of molecular variance
(AMOVA; Excoffier et al. 1992). Median joining networks were estimated using NET-
WORK 4.1.1.1 (Bandelt et al. 1999) (available at http://www.fluxus-engineering.com/).
Additionally, the demographic history, evidence for past population growth, decline or
stability, were tested using two neutrality tests (Fu’s Fs tests by Fu 1997 and Tajima’s D by
Tajima 1989) and by generating mismatch distributions. These tests were conducted in
DNAsp (version 4.1; Rozas et al. 2003). ARLEQUIN was used to test the significance of
the mismatch distribution (Rogers and Harpending 1992), through a parametric bootstrap
test between the observed and expected mismatches, the sum of square deviations (SSD)
(Schneider and Excoffier 1999).
Results
Single-locus minisatellite and microsatellite variation
Minisatellite loci GguA
6
and GguB
10
produced six and four alleles, respectively. Micro-
satellite loci presented two (HYA1172) to four (MAC436) alleles (Table 1). The values of
H
e
ranged from 0.06 to 0.84, whilst the values of H
o
ranged from 0 to 0.80 (Table 1).
Departures from the Hardy–Weinberg equilibrium were found only for the locus MAC436
Biodivers Conserv (2008) 17:765–779 769
123
in the apprehended samples (P = 0.011). Linkage disequilibrium was found in Miranda
samples between loci HYA1172 and MAC436 (P = 0), in Piauı
´
for loci GguA
6
and
GguB
10
(P = 0) and between loci MAC436 and GguA
6
for the seizure (P = 0.034) (data
not shown).
The pairwise comparisons of the heterozygosity values did not reveal any significant
value (0.0015 \ t \ 2.047) among all pairs of populations analyzed (data not shown).
Consequently, the values of genetic diversity obtained in each locality are similar to each
other.
Population genetic analysis showed no evidence for genetic differentiation between
Miranda and Abobral (F
ST
= 0.05; P = 0.12) and also between Piauı
´
and the seized
samples (F
ST
=-0.03; P = 0.39) (Table 2). However, significant population differ-
entiation was found between the Piauı
´
and Pantanal populations (F
ST
= 0.33; P \ 0.001
and F
ST
= 0.25; P \ 0.001; Abobral and Miranda, respectively) and between the
seized samples and the Pantanal (F
ST
= 0.25; P = 0.005 and F
ST
= 0.17; P = 0.019;
Abobral and Miranda, respectively) (Table 2). Similar results were obtained using the
Bayesian approach on STRUCTURE, where the most probable number of populations
inferred was two (Fig. 2): genetic population 1 (red) was mainly constituted by indi-
viduals from Abobral and Miranda, while genetic population 2 (green) by Piauı
´
and
seizure. The mean likelihood value for 10 independent runs was great at K = 2
(-241.15). Most Hyacinth macaws from Abobral and Miranda were assigned to genetic
population 1 (62.5% and 61.4%, respectively), while individuals from Piauı
´
and seizure
had higher probabilities of belonging to genetic population 2 (60.6% and 72.6%,
respectively).
Table 1 Number of alleles (A),
expected (H
e
) and observed (H
o
)
heterozigosities obtained with the
minisatellite loci A
6
and B
10
and
with microsatellites loci
HYA1172 and MAC436 in each
locality
Locality A6 B10 HYA1172 MAC436
Abobral (n = 16) A 3 2 1 2
H
e
0.53 0.42 0.27 0.06
H
o
0.33 0.43 0.31 0.06
Miranda (n = 16) A 4 2 2 2
H
e
0.59 0.48 0.48 0.06
H
o
0.69 0.56 0.50 0.06
Piauı
´
(n = 6) A 4 2 2 2
H
e
0.71 0.20 0.16 0.53
H
o
0.40 0.20 0.16 0.16
Seizure (n = 5) A 5 3 2 3
H
e
0.84 0.60 0.20 0.71
H
o
0.80 0.80 0.20 0
Table 2 Nuclear pairwise F
ST
(below the diagonal) and mito-
chondrial DNA U
ST
(above the
diagonal)
Significant values at P = 0.05
are indicated in bold
Locations Abobral Miranda Piauı
´
Seizure
Abobral – -0.04 0.18 0.18
Miranda 0.05 – 0.14 0.13
Piauı
´
0.33 0.25 – 0.16
Seizure 0.25 0.17 -0.03 –
770 Biodivers Conserv (2008) 17:765–779
123
Mitochondrial DNA
Control region sequences of 472 bp revealed thirteen variable sites that defined eight
haplotypes, their frequencies in each population are shown in Table 3. One haplotype was
shared by all populations (haplotype 4), while haplotypes 1, 2 and 3 were found exclu-
sively in Pantanal. Haplotypes 6, 7 and 8 were detected only in Piauı
´
, whilst haplotype 5
was only found in seized samples. The values of nucleotide diversity (p) per site ranged
from 0.002 to 0.008 (Table 3).
Heteroplasmic sites were observed, nevertheless there was strong evidence that these
sequences are mitochondrial as opposed to nuclear copies. Long PCRs were carried out in a
previous study (Tavares et al. 2004) to verify the authenticity of the sequences amplified
by these primers (CR522Rb/LGlu) in the Hyacinth Macaw. Further, both strands were
independently sequenced for all individuals and several individuals were sequenced more
than once, in both situations the same result was obtained, without any ambiguities.
Non-significant pairwise U
ST
values were obtained through all comparisons (-0.04 \
U
ST
\ 0.18; 0.08 \ P \ 0.69) (Table 2), indicating a lack of genetic structure at mito-
chondrial level. Initially only one comparison was conducted in AMOVA (Abobral and
Miranda versus Piauı
´
and Seizure), as based on our assumptions, these two main groups
Fig. 2 Bar plot of population assignment proportions for K = 2 generated by STRUCTURE. Each
individual is represented by a single vertical line divided into the two inferred clusters: genetic population 1
in red and genetic population 2 in green
Table 3 Haplotypes and variable sites found in the control region sequences of the mitochondrial DNA of
Hyacinth Macaw from three locations in Brazil and individuals from seizure
Haplotypes Variable sites Haplotype frequency
112234 Abobral Miranda Piauı
´
Seizure
1256667230606
0954897316841
1 TACTTGRTTCTWG 0.375 0.250 – –
2 .G.CC.R.C...C – 0.125 – –
3 .GTCC.R.C.CWC 0.062 0.125 – –
4 .GTCCAR.C.CWC 0.563 0.5 0.333 0.6
5 .G.CCARCC.CWC – – – 0.4
6 .GTCCAR.C..W. – – 0.167 –
7 .GTCCAR.CTCWC – – 0.333 –
8 GGTCCAR.C.CWC – – 0.167 –
p 0.008 0.007 0.003 0.002
H 0.575 0.70 0.86 0.60
Biodivers Conserv (2008) 17:765–779 771
123
would be more plausible. Non-significant results were obtained in AMOVA and the
genetic variation was distributed as follow: 18.54% between groups (P = 0.33), -2.84%
among populations within groups (P = 0.44) and 84.31% within populations (P = 0.10).
Afterwards, all possible combinations were tested, but once more non-significant results
were obtained (data not shown).
The star shape of the network generated by control regions sequences (Fig. 3) indicates
a population expansion. For the neutrality tests, Tajima’s D and Fu’s Fs were non-sig-
nificant (P [ 0.30) for all populations (Abobral: D = 2.41, Fu’s F = 5.63; Miranda:
D = 1.68, Fu’s F = 3.25). Nevertheless, negative values for both neutrality tests were
obtained for Piauı
´
population (D =-0.67, P = 0.30; Fs =-0.99, P = 0.13). Mismatch
distributions of Abobral and Miranda were ragged (Fig. 4), as expected in a constant
population size model, and differed significantly from population expansion model
(SSD = 0.25, P = 0.10; SSD = 0.08, P = 0.25, Abobral and Miranda, respectively).
Despite the unimodal shape of Piauı
´
’s mismatch distribution, similar to the pattern
observed in expanding populations, once more no statistical significance from stationary
populations models was detected (SSD = 0.01, P = 0.70).
Fig. 3 Medium spanning
network of Hyacinth Macaw
control region (472 bp). Each
circle represents a haplotype and
the diameter scales to the
haplotype frequency. Mutational
steps are represented by black
bars on lines connecting the
haplotypes. Blank circles
represents haplotypes exclusive
from Piauı
´
populations
772 Biodivers Conserv (2008) 17:765–779
123
Discussion
Genetic variation
The comparison of genetic diversity values among Hyacinth Macaws from three locations
in Brazil did not result in significant differentiation. We predicted that, based on the larger
population size of the Pantanal, genetic variability would be higher in this region when
Fig. 4 Mismatch distributions
for Hyacinth Macaw control
region sequences. Expected
(solid lines) according to the
sudden expansion model and
observed (dotted lines) showing
the frequencies of pairwise
differentes within (a) Abobral,
(b) Miranda and (c) Piauı
´
Biodivers Conserv (2008) 17:765–779 773
123
compared to Piauı
´
. Nevertheless, considering that Hyacinth Macaws are strong flyers, there
could conceivably be gene flow between Piauı
´
and the nearest population in the Amazon
(that could not be studied here). This might increase the effective size of the Piauı
´
pop-
ulation and its levels of genetic variability would be higher.
Despite the fact that drawing comparison of genetic diversity indices obtained for
different species and by different molecular markers needs to be done with caution, this
information can still help to understand the degree of genetic threat of a certain species.
Our data show that, in comparison with other population genetic analyses of Psittacines
(Nader et al. 1999; Wright and Wilkinson 2001; Caparroz 2003; Wright et al. 2005), the
Hyacinth Macaw possesses relatively lower levels of genetic variability.
Analysis of 1,290 bp of the control region of 50 Blue and yellow Macaws (Ara ar-
arauna) revealed 38 variable sites that defined 28 haplotypes, while the analysis of
microsatellite loci produced three to 11 alleles with average heterozygosity of 0.604
(Caparroz 2003). Analysis of 680 bp of the control region of the Yellow-naped Amazon
(Amazona auropalliata, n = 75) showed 59 variable sites that defined 19 haplotypes
(Wright and Wilkinson 2001) and the microsatellite analysis detected three to 25 alleles
with average heterozygosity of 0.59 (Wright et al. 2005). Faria and Miyaki (2006) eval-
uated the levels of genetic variation in five species of parrots and macaws, using different
nuclear molecular markers, and found that among the three species of macaws analyzed
(Ara ararauna, A. chloroptera and Anodorhynchus hyacinthinus), the Hyacinth Macaw
possessed the lowest values of genetic variability.
One of the main predicted consequences of a loss of genetic variability is a reduction of
the ability of a species to react to environmental changes, which operates over long-time
scales (Frankham et al. 2002). Nevertheless, planning for the conservation of biodiversity
is usually focused over short time-scales (Haig 1998; Amos and Balmford 2001). It has
been pointed out that demographic factors, environmental fluctuations and natural catas-
trophes can drive populations to extinction before genetic factors impact them (Lande
1988; Amos and Balmford 2001; Frankham et al. 2002; Spielman et al. 2004).
Despite the low values of genetic variation detected in our study (H
average
= 0.35;
2 \ number of alleles \ 6 for nuclear data; 13 variable sites and eight haplotypes for
mtDNA), they seem to be higher than those obtained in other endangered bird species such
as the Andean condor (Vultur gryphus; Hendrickson et al. 2003), the whooping crane
(Grus americana; Glenn et al. 1999), the great prairie-chicken (Tympanuchus cupido;
Westemeier et al. 1998; Johnson et al. 2004) and the Spotted owl (Strix occidentalis;
Barrowclough et al. 1999). Consequently, the levels of genetic variation presented by
Hyacinth Macaws in our study do not seem to pose a current concern in terms of con-
servation (see Conservation Implications).
Genetic structure
A lack of population differentiation was found between the two populations from Pantanal
(Abobral and Miranda) with nuclear and mitochondrial DNA markers (F
ST
= 0.05 and
F
ST
= -0.04, respectively) and STRUCTURE results suggest the existence of a single
population in Pantanal. Both locations are about 100 km distant from each other and there
are no physical barriers among them. The monitoring of Hyacinth Macaws with radio
tracking in Pantanal showed that juveniles are able to fly about 35 km after fledging
(Seixas et al. 2002). Also, several field observations have suggested that adults were
observed up to 100 km away from their natal areas (Yamashita, pers. comm.). These
774 Biodivers Conserv (2008) 17:765–779
123
results show, as expected, that Hyacinth Macaws are able to fly long distances; however, it
does not mean that effective dispersal occurs. Banding and recapture data would be nec-
essary in order to evaluate the philopatric behavior of Hyacinth Macaws. Nevertheless, to
date these data are scarce, mainly due to the high level of disturbance involving the capture
of adults. Guedes and Harper (1995) suggest that there is nest site fidelity for Hyacinth
Macaws in Pantanal, and the same pair could be using the same tree cavity for breeding in
consecutive years, but so far there is not enough evidence of philopatric behavior in
Hyacinth Macaw.
Even though the two locations from Pantanal seem to constitute a single breeding
population, as shown by our data, other locations from Pantanal need to be evaluated
before conclusions about the population structure of this region can be drawn. Thus, it
would be desirable to sample individuals from the other regions from Pantanal (mainly
those from the north) and use additional molecular markers.
Restrictions to gene flow can occur due to geographical, ecological or behavioral iso-
lation (i.e., philopatry) or population history. Parrots and macaws are capable of flying long
distances; therefore their dispersal seems to be more constrained by the selective use of the
habitat and the availability of natural forest and resources than mobility (Myers and
Vaughan 2004; Evans et al. 2005). Their movements are mainly determined by spatial and
temporal patterns of fruiting, seeding and water availability (Collar 1997; Renton 2002).
Consequently, habitat fragmentation seems to be an important factor driving their distri-
bution, choice of breeding sites and the level of gene flow among populations.
Based on this, we predicted the occurrence of strong genetic structure between Pantanal
and Piauı
´
. Our results show significant genetic differentiation between Pantanal and Piauı
´
(F
ST
[ 0.25; P \ 0.05) for nuclear DNA. These findings are strongly supported by the
results from STRUCTURE that show two genetic populations: one constituted by Pantanal
(Abobral and Miranda) and another by Piauı
´
and seizure. Also, private alleles and hapl-
otypes were found in Piauı
´
for nuclear and mtDNA, respectively, and according to Slatkin
(1985) they are good indicators of population structure and limited gene flow between
populations.
Despite the detection of significant genetic structure at nuclear level, a lack of genetic
differentiation was obtained using mtDNA control region sequences. Two potential
explanations may arise from this data: female-biased gene flow and/or the nature of the
molecular markers used. Birds usually present female-biased gene flow (Greenwood 1980),
while males show higher levels of philopatry which result in population genetic structure at
nuclear markers. Among the Psittacinae, sex-biased gene flow has been described for
Yellow-Naped Parrots (Amazona auropalliata; Wright and Wilkinson 2001; Wright et al.
2005) and for Blue-and-Yellow Macaws (Ara ararauna; Caparroz 2003). On the first
study, despite the lack of genetic structure among nine locations in Costa Rica and no
correlation with dialects, the authors described the occurrence of female-biased gene
dispersal (Wright et al. 2005). Contrastingly, Blue-and-Yellow Macaws presented male-
biased gene flow (Caparroz 2003).
In our study, the migration rates (N
e
m) obtained from the analysis of mtDNA control
region sequences (3.01 B N
e
m \ ?) was around six times higher than the estimative
using nuclear DNA (0.70 B N
e
m \ 6.46), suggesting female-biased gene flow in Hyacinth
Macaws. Nevertheless, our results need to be seen with caution and need to be further
investigated with the use of more accurate methodology to estimate migration, the
inclusion of demographic parameters and the estimative of direct dispersal. Besides, there
are some additional reasons described below.
Biodivers Conserv (2008) 17:765–779 775
123
The differences in the level of genetic structure between nuclear and mtDNA detected
in our study may also be the effect of the effective population size and mutation rates
(Chesser and Baker 1996). Populations that diverged recently and have not experienced
enough time to achieve the equilibrium between genetic drift, mutation and gene flow can
retain ancestral variation that can produce misleading conclusions about the current levels
of gene flow (Birky et al. 1989). Microsatellites present higher mutation rates than
mtDNA, therefore they are more effective in revealing recent genetic structure than
mtDNA. Nevertheless, since the effective population size of mtDNA is one-fourth of
nuclear markers, mtDNA markers would achieve the equilibrium faster than nuclear
markers and could potentially reflect more adequately the contemporary pattern of genetic
structure. On the other hand, the assumption of smaller effective size of mtDNA is usually
violated as it depends strictly of equal sex ratio, random mating and male reproductive
success (Prugnolle and Meeus 2002).
Testing for mutation, drift and migration is a difficult procedure (Birky et al. 1989) and
it could not be conducted here. Consequently, in our study it was not possible to dis-
criminate among the alternative hypotheses that would better explain the population
structure pattern obtained in Hyacinth Macaws. Besides, Prugnolle and Meeus (2002)
pointed that, in order to discriminate the effects of sex-biased gene flow and molecular
markers nature it is necessary to have demographic data.
The historical demography of Hyacinth Macaw populations is unknown. According to
Yamashita (1997), the areas occupied by the macaws from the genus Anodorhynchus have
changed considerably over the years due to its strict dependence on few species of palm
seeds. Also, their range might have been drastically reduced after the megafauna extinc-
tions in South America, around 10,000 years ago, as these large vertebrates were the main
seed-dispersers of Neotropical trees (Yamashita and Valle 1993; Yamashita 1997). Fur-
thermore, the introduction of domestic livestock in the XVIth century, themselves potential
seed-dispersers, might have modified the scenario once more, and subsequently changed
the areas occupied by Hyacinth Macaws (Yamashita 1997; Yamashita, pers. comm.).
Thus, it seems that the effective population size of Hyacinth Macaws has not been stable
over the years. Mismatch distribution and neutrality tests were conducted in order to gather
some information about this issue; however, only non-significant results were obtained.
Despite that, the star shape of the network (Fig. 3) indicates a rapid and recent population
expansion. The haplotypes at the tips of the network show few mutations and occur in
individuals from Piauı
´
. This fact, together with the negative values of neutrality tests suggest
that Piauı
´
’s population might have been colonized by Pantanal and it is still expanding.
In sum, our study detected a significant genetic differentiation at the nuclear level
between Pantanal and Piauı
´
. Possible causes driving this differentiation were raised,
including female biased-gene flow, current habitat fragmentation and recent colonization
followed by population expansion in Piauı
´
. In order to identify the most important feature
and clarify this scenario, the demographic history of Hyacinth Macaws needs to be further
examined with broader sampling and additional molecular markers. The effect of the current
habitat fragmentation on the precise gene flow patterns of Hyacinth macaws could not be
determined in this study and this information is necessary to plan conservation actions.
Origin of seized birds
The identification of the origin of seized specimens is potentially important to help
authorities to identify the areas where those birds have been poached. According to our
776 Biodivers Conserv (2008) 17:765–779
123
results, the individuals from the seizure do not belong to the Pantanal. This result is
corroborated by the presence of alleles exclusive to individuals from Piauı
´
, significant F
ST
values when compared to both regions of Pantanal (Abobral and Miranda) and the
assignment of seizure individuals to genetic population 2. However, due to the absence of
samples from other locations, mainly from the Amazon region, it is not possibly to identify
the origin of these birds, neither prove that they belong to the Piauı
´
.
Conservation implications
The low genetic variation detected does not seem to pose a current concern to the Hyacinth
Macaw conservation. On the other hand, the main threats seem to be the high specialization
in diet and in nest site preference and low reproductive rates of the three species of the
genus Anodorhynchus, and also human related factors. The decline of the Glaucous Macaw
(A. glaucus) was possibly caused by the decline of palm groves (Butia yatay), its exclusive
food source, as a consequence of the introduction of domestic herbivores (Yamashita and
Valle 1993). The Lear’s macaw (A. leari), an endemic species from the caatinga ecosystem
in Bahia state, northeastern Brazil, is highly threatened because of wild animal traffic
(BirdLife International 2006). Currently, there are only around 500 individuals in the wild
(Amaral et al. 2005).
The results concerning the genetic structure of Hyacinth Macaw populations have
implications for the conservation of this species. If the different locations represent discrete
breeding populations, as suggested by nuclear DNA data, then effective protection is
needed in all regions that shelter Hyacinth Macaws in order to preserve the overall genetic
variability of this species. Thus, it is urgent to protect the populations that are not being
permanently monitored (populations from North and Gerais of Brazil), since they are the
possible sources of individuals for the illegal trade. Regarding the current captive indi-
viduals of Hyacinth Macaw, it would be desirable that they were managed according to
their geographical origin.
Considering the results obtained here, it seems that the survival of Hyacinth Macaws,
such as other species of parrots (Gonza
´
lez 2003; Myers and Vaughan 2004; Vaughan et al.
2005; Wright et al. 2005) depends mainly on the protection of natural habitat and work
with local communities to reduce nest poaching. These measures can help guarantee the
maintenance of the Hyacinth Macaw populations in the wild and their survival.
Acknowledgments We thank Anita Wajntal and Mike W. Bruford for encouragement and invaluable
suggestions on the manuscript. Se
´
rgio L. Pereira, Se
´
rgio Matioli, Elisa
ˆ
ngela P. Quedas, Anto
ˆ
nia M. P.
Cerqueira, Andre
´
a Bernardino and Adriana R. Oliveira-Marques for support during the lab work and
statistical analysis. Scott K. Davis generously provided unpublished microsatellites primers sequences. Luı
´
s
G. Maluf and (IBAMA) helped with the blood samples collection. This study was supported by FAPESP,
CNPq, CAPES and Fundac¸a
˜
o BioBrasil. CYM has a CNPq research productivity fellowship.
References
Amaral ACA, Herna
´
ndez MIM, Xavier BF, Bella SD (2005) Dina
ˆ
mica de ninho de Arara-azul-de-lear
(Anodorhynchus leari Bonaparte, 1856) em Jeremoabo, Bahia. Ornithologia 1:59–64
Amos W, Balmford A (2001) When does conservation genetics matter? Heredity 87:257–265
Bandelt HJ, Forster P, Rohl A (1999) Median-joining networks for inferring intraspecific phylogenies. Mol
Biol Evol 16:37–48
Biodivers Conserv (2008) 17:765–779 777
123
Barrowclough GF, Gutie
´
rrez RJ, Groth JG (1999) Phylogeography of spotted owl (Strix occidentalis)
populations based on mitochondrial DNA sequences: gene flow, genetic structure, and a novel bio-
geographic pattern. Evolution 53:919–931
Birdlife International (2004) Anodorhynchus hyacinthinus. In: IUCN 2006. 2006 IUCN Red List of
Threatened Species. http://www.iucnredlist.org. Cited 10 July 2006
Birdlife International (2006) Anodorhynchus leari. http://www.birdlife.org. Cited 25 July 2006
Birky CW, Fuerst P, Maruyama T (1989) Organelle gene diversity under migration, mutation, and drift:
equilibrium expectations, approach to equilibrium, effects of heteroplasmic cells, and comparison to
nuclear genes. Genetics 121:613–627
Bruford MW, Hanotte O, Brookfield JFY, Burke T (1992) Single locus and multilocus DNA fingerprinting.
In: Hoezel CAR (ed) Molecular genetic analysis of populations: a practical approach. Oxford Uni-
versity Press, New York
Caparroz R (2003) Filogeografia, estrutura e variabilidade gene
´
tica da Arara-caninde
´
(Ara ararauna,
Psittaciformes: Aves) no Brasil baseadas na ana
´
lise de DNA mitocondrial e de DNA nuclear. Ph. D.
Dissertation, Universidade de Sa
˜
o Paulo, Sa
˜
o Paulo, Brazil
Chesser RK, Baker RJ (1996) Effectives sizes and dynamics of uniparentally and biparentally inherited
genes. Genetics 144:1225–1235
Collar NJ (1997) Family Psittacidae (parrots). In: del Hoyo J, Elliott A, Sargatal J (eds) Handbook of the
birds of the world: sandgrouse to cuckoos. Lynx Edicions, Barcelona
Collar NJ, Gonzaga LP, Krabbe N, Nieto AM, Naranjo LG, Parker TA, Wege DC (1992) Threatened birds
of the Americas: The ICBP/IUCN red data book. Smithsonian, Cambridge
Eberhard JR, Wright TF, Bermingham E (2001) Duplication and concerted evolution of the mitochondrial
control region in the parrot genus Amazona. Mol Biol Evol 18:1330–1342
Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software
STRUCTURE: a simulation study. Mol Ecol 14:2611–2620
Evans BEI, Ashley J, Marsden SJ (2005) Abundance, habitat use, and movements of Blue-winged Macaws
(Primolius maracana) and other parrots in and around an Atlantic forest reserve. Wilson Bull 117:154–
164
Excoffier L, Smouse PE, Quattro JM (1992) Analysis of molecular variance inferred from metric distances
among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131:479–
491
Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype
data: linked loci and correlated allele frequencies. Genetics 164:1567–1587
Faria PJ, Miyaki CY (2006) Molecular markers for population genetic analyses in the family Psittacidae
(Psittaciformes, Aves). Genet Mol Biol 29:231–240
Forshaw JM (1989) Parrots of the world. Landsdowne editions, Melbourne
Frankham R, Ballou JD, Briscoe DA (2002) Introduction to conservation genetics. Cambridge Press, New
York
Fu Y (1997) Statistical tests of neutrality of mutations against population growth, hitchhiking and back-
ground selection. Genetics 147:915–925
Glenn TC, Stephan W, Braun MJ (1999) Effects of a population bottleneck on whooping crane mito-
chondrial DNA variation. Conserv Biol 13:1097–1107
Gonza
´
lez JA (2003) Harvesting, local trade and conservation of parrots in the Northeastern Peruvian
Amazon. Biol Conserv 114:437–446
Greenwood PJ (1980) Mating systems, philopatry and dispersal in birds and mammals. Anim Behav
28:1140–1162
Guedes NMR (1993) Biologia reprodutiva da Arara azul (Anodorhynchus hyacinthinus) no Pantanal-MS,
Brasil. MSc Dissertation, Universidade de Sa
˜
o Paulo, Sa
˜
o Paulo, Brasil
Guedes NMR (2004) Araras azuis: 15 anos de estudos no Pantanal. In: IV Simpo
´
sio sobre Recursos Naturais
eSo
´
cio-econo
ˆ
micos do Pantanal, Corumba
´
,MS
Guedes NMR, Harper LH (1995) Hyacinth macaws in the Pantanal. In: Abramson J, Speer L, Thomsen JB
(eds) The large macaws: their care, breeding and conservation. Raintree Publications, Hong Kong
Haig SM (1998) Molecular contributions to conservation. Ecology 79:413–425
Hendrickson S, Bleiweiss R, Matheus JC, Matheus LS, Ja
´
come NL, Pavez E (2003) Low genetic variability
in the geographically widespread Andean condor. Condor 105:1–12
Johnson JA, Bellinger MR, Toepfer JE, Dunn P (2004) Temporal changes in allele frequencies and low
effective population size in greater prairie-chickens. Mol Ecol 13:2617–2630
Lande R (1988) Genetics and demography in biological conservation. Science 241:1455–1460
Munn CA, Thomsen JB, Yamashita C (1989) The Hyacinth macaw. In: Chadler WD (ed) Audubon wildlife
report. Academic Press, New York
778 Biodivers Conserv (2008) 17:765–779
123
Myers MC, Vaughan C (2004) Movements and behavior of scarlet macaws (Ara macao) during the post-
fledging dependence period: implications for in situ versus ex situ management. Biol Conserv
118:411–420
Nader W, Werner D, Wink M (1999) Genetic diversity of scarlet macaws Ara macao in reintroduction
studies for threatened populations in Costa Rica. Biol Conserv 87:269–272
Nei M (1987) Molecular evolutionary genetics. Columbia University Press, New York
Nei M, Kumar S (2000) Molecular evolution and phylogenetics. Oxford University Press, New York
Nichols RA, Beaumont MA (1996) Is it ancient or modern history that we can read in the genes? In:
Hochberg M, Clobert J, Barbault R (eds) Aspects of the genesis and maintenance of biological
diversity. Oxford University Press, Oxford
Pinho JB, Nogueira FMB (2003) Hyacinth Macaw (Anodorhynchus hyacinthinus) reproduction in the
northern Pantanal, Mato Grosso, Brazil. Ornitol Neotrop 14:29–38
Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype
data. Genetics 155:945–959
Prugnolle F, Meeus T (2002) Inferring sex-biased dispersal from population genetic tools: a review.
Heredity 88:165–169
Renton K (2002) Seasonal variation in occurrence of macaws along a rainforest river. J Field Ornithol
73:15–19
Rogers AR, Harpending H (1992) Population growth makes waves in the distribution of pairwise genetic
differences. Mol Biol Evol 9:552–569
Rozas J, Sanchez-Delbarrio JC, Messeguer X, Rozas R (2003) DNASP, DNA polymorphism analyses by the
coalescent and other methods. Bioinformatics 19:2496–2497
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual. Cold Spring Harbor
Laboratory Press, New York
Schneider S, Excoffier L (1999) Estimation of demographic parameters from the distribution of pairwise
differences when the mutation rates vary among sites: application to human mitochondrial DNA.
Genetics 152:1079–1089
Schneider S, Roessli D, Excoffier L (2000) Arlequin 2.0: a software for population genetic data analysis.
Genetics and Biometry Laboratory, University of Geneva, Switzerland
Seixas GHF, Guedes NMR, Martinez J, Prestes NP (2002) Uso de radiotelemetria no estudo de psitacı
´
deos.
In: Galetti M, Pizo MA (eds) Ecologia e Conservac¸a
˜
o de Psitacı
´
deos no Brasil. Melopsittacus Pub-
licac¸o
˜
es Cientı
´
ficas, Belo Horizonte
Sick H (1997) Ornitologia Brasileira. Editora Nova Fronteira, Rio de Janeiro
Slatkin M (1985) Rare alleles as indicators of gene flow. Evolution 39:53–65
Spielman DB, Brook W, Frankham R (2004) Most species are not driven to extinction before genetic factors
impact them. Proc Natl Acad Sci USA 101:15621–15624
Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism.
Genetics 158:1147–1155
Tavares ES, Yamashita C, Miyaki CY (2004) Phylogenetic relationships among some neotropical parrot
genera (Psittacidae, Aves) based on mitochondrial sequences. Auk 121:230–249
Vaughan C, Nemeth NM, Cary J, Temple S (2005) Response of a Scarlet Macaw Ara macao population to
conservation practices in Costa Rica. Bird Conserv Int 15:119–130
Westemeier RL, Brawn JD, Simpson SA, Esker TL, Jansen RW, Walk JW, Kershner EL, Bouzat JL, Paige
KN (1998) Tracking the long-term decline and recovery of an isolated population. Science 282:1695–
1697
Wright TF, Wilkinson GS (2001) Population genetic structure and vocal dialects in an amazon parrot. Proc
R Soc Lond B 268:609–616
Wright TF, Rodriguez AM, Fleischer RC (2005) Vocal dialects, sex-biased dispersal, and microsatellite
population structure in the parrot Amazona auropalliata. Mol Ecol 14:1197–1205
Yamashita C (1997) Anodorhynchus macaws as followers of extinct megafauna: a hypothesis. Ararajuba
5:176–182
Yamashita C, Valle MP (1993) On the linkage between Anodorhynchus macaws and palm nuts, and the
extinction of the Glaucous macaws. Bull B O C 113:53–60
Biodivers Conserv (2008) 17:765–779 779
123
