Journal of Heredity 2004:95(1):11–18
DOI: 10.1093/jhered/esh001
Ó 2004 The American Genetic Association
Microsatellite Analysis of Genetic
Diversity of the Vietnamese Sika Deer
(Cervus nippon pseudaxis)
S. THÉVENON, L. T. THUY, L. V. LY, F. MAUDET, A. BONNET, P. JARNE,
AND
J.-C. MAILLARD
From CIRAD-EMVT, Rangeland and Wildlife Management Program, TA 30/F, Campus International de Baillarguet, 34398
Montpellier cedex 5, France (Thévenon and Maudet), National Muséum of Natural History of Paris, Parc Zoologique de
Vincennes, 53 avenue de St Maurice, Paris 75012, France (Thévenon and Bonnet), National Institute for Animal Husbandry,
Thuy Phuong, Tu Lien, Hanoi, Vietnam (Thuy and Ly), Centre d’Ecologie Fonctionnelle et Evolutive, 1919 route de Mende,
34293 Montpellier cedex 5, France (Jarne), and CIRAD-EMVT, Health Animal Program, TA 30/G, Campus International
de Baillarguet, 34398 Montpellier cedex 5, France (Maillard). We would like to thank Mr. Binh, Mr. Zhung, the Nghe An
and Ha Tinh People Committees, and numerous farmers for providing samples, and the National Institute for Animal
Husbandry of Hanoi (NIAH) for its support and efficient organization. We are thankful to F. Claro (National Museum of
Natural History, Paris), E. Camus, P. Chardonnet, J. Domenech, G. Mandret, and F. Monicat (Centre International en
Recherche Agronomique pour le Développement, CIRAD) for establishing the collaborative project with the NIAH. We are
grateful to C. Dutech (Centre d’Ecologie Fonctionnelle et Evolutive), G. Luikart, C. Maudet, and P. Taberlet (Grenoble
University) for fruitful discussions. L. Guerrini (CIRAD) drew the sampling map. This work was funded by the CIRAD,
the Ministère Français des Affaires Etrangères, the Muséum National d’Histoire Naturelle, and the International Game
Foundation. Sophie Thévenon was supported by a grant from the Ministère de l’Enseignement Supérieur et de
la Recherche (MESR).
Address correspondence to Sophie Thévenon, CIRDES BP454, 01 Bobo Dioulasso, Burkina Faso,
or e-mail: [email protected].
Abstract
The Vietnamese sika deer (Cervus nippon pseudaxis) is an endangered subspecies of economic and traditional value in Vietnam.
Most living individuals are held in traditional farms in central Vietnam, others being found in zoos around the world. Here
we study the neutral genetic diversity and population structure of this subspecies using nine microsatellite loci in order to
evaluate the consequences of the limited number of individuals from which this population was initiated and of the breeding
practices (i.e., possible inbreeding). Two hundred individuals were sampled from several villages. Our data show both
evidence for limited local inbreeding and isolation by distance with a mean FST value of 0.02 between villages. This suggests
that exchange of animals occurs at a local scale, at a rate such that highly inbred mating is avoided. However, the genetic
diversity, with an expected heterozygosity (He) of 0.60 and mean number of alleles (k) of 5.7, was not significantly larger
than that estimated from zoo populations of much smaller census size (17 animals sampled; He ¼ 0.65, k ¼ 4.11). Our
results also suggest that the Vietnamese population might have experienced a slight bottleneck. However, this population is
sufficiently variable to constitute a source of individuals for reintroduction in the wild in Vietnam.
Due to various anthropic pressures, numerous animal
species have gone extinct or are critically endangered or
threatened (Diamond 1989; Hilton-Taylor 2000; Pimm and
Raven 2000; Pimm et al. 1995). Beyond the ethical and
ecological aspects, the loss of animal species or subspecies
may represent a social or economic loss for human
populations, especially in developing countries. Wildlife
may provide food, income, and employment to human
populations at a local level (Chardonnet 1995). For long-term
conservation and development purposes, it therefore
appears compulsory to manage wildlife to maintain both
species survival and within-species genetic diversity (Franklin
1980). The loss of genetic variability, as a consequence of
sustained small census size (e.g., because of breeding or
hunting practices), is indeed expected to threaten the longterm population viability (see, e.g., Coulson et al. )1998] for
theoretical work; Eldridge et al. [1999] for field studies;
Lynch et al. 1995; Saccheri et al. 1998) and to prevent adaptation to novel environmental conditions (Frankham et al.
1999; Franklin 1980; Lande and Arnold 1983; Vida 1994).
11
Journal of Heredity 2004:95(1)
An interesting example in this context is the Vietnamese
sika deer (Cervus nippon pseudaxis). Thirteen subspecies of
the sika deer (C. nippon) have been described, and their
distribution ranges from North Korea to Vietnam, including
Taiwan and Japan (Whitehead 1993). The Vietnamese
subspecies is the most tropical one, and might, as such,
present some local adaptations. Its natural populations had
a striking decline in the wild at the beginning of the 20th
century (Thuy LT, personal communication). At about the
same time, breeding was initiated in traditional farms for
velvet production. Captive breeding has possibly been
associated with a genetic bottleneck, since breeding programs were probably initiated with no more than 200 wild
individuals in the Ha Tinh province, in central Vietnam
(Thuy LT, personal communication). The entire current
population from farms (about 8000–10,000 individuals)
originates from these individuals. Velvet production is
a major source of income for farmers (approximately
US$200–US$300/year/male) and is therefore of huge social
and economic value. This also has some consequences in
population structure. The sex ratio of the sika deer in farm
populations is skewed toward males (about 65% are males in
the Nghe An and Ha Tinh provinces studied here), because
males produce velvet, which should lower the effective
population size (Crow and Kimura 1970). Individuals are
maintained in small individual lodges, and some never
reproduce. Animals are exchanged or sold locally, within or
among villages. Variation in supply and demand for velvet
has recently caused substantial variation in population size
with a reduction of 50% within a couple of years.
Captive and semicaptive populations are also maintained
in both zoological parks (around 225 individuals in the
world; Rüdloff 1999) and Vietnamese national parks (70–80
individuals in Cuc Phuong [Wemmer 1998] and an unknown
number in Cat Ba), where poaching is a serious threat
(Wemmer 1998). On the other hand, the Vietnamese sika
deer is considered extinct in the wild; the last wild living
specimens were reported in 1990 in the Nghe Tinh
Mountains (Huynh et al. 1990). As a consequence, it is listed
as critically endangered by the World Conservation Union.
Given that the majority of sika deer are maintained on
traditional farms, they should play a focal role in any
conservation and development programs.
The goal of our study was to evaluate the consequences
of recent history and breeding practices on the genetic
variability of the Vietnamese sika deer as part of a larger
study on the management of genetic variability in this
subspecies. We took advantage of recent advances in
molecular and statistical techniques used for analyzing
neutral variability, especially microsatellite markers (Goldstein and Schlötterer 1999; Luikart and England 1999;
Rousset 2001). The study was conducted on 200 sika deer
sampled from traditional farms of several villages in central
Vietnam and on individuals sampled from two zoo
populations. The population genetic structure was analyzed
using nine microsatellite markers, in order to address the
following questions: What is the current level of genetic
diversity and how is it structured? Is there any evidence of
12
inbreeding? Is the genetic diversity maintained on traditional
farms higher or lower than the one preserved in Vietnamese
sika deer from European zoological parks? Should conservation programs focus on this population from a genetic
perspective?
Materials and Methods
Subspecies Studied and Sampling Area
Most Vietnamese sika deer are currently located in the Nghe
An and Ha Tinh provinces, with roughly 4000–5000 animals
in each province. Sampling was performed in these two
provinces in October 1999 and 2000. One hundred seventysix and 24 individuals were sampled in the Nghe An and Ha
Tinh provinces, respectively (Figure 1). The geographical
position of each individual was determined using a global
positioning system. Sampling of related animals within farms
was avoided by asking farmers about animal relatedness. The
overall sample size was relatively low due to sampling
difficulties (local sociological constraints and no animal
catching, which could have resulted in stress and/or injuries).
Blood samples were also collected from 17 Vietnamese sika
deer from two European zoological parks, Tierpark Berlin
and Mulhouse (Thévenon et al., 2003), in order to compare
their variability to that of the Vietnamese population.
DNA Extraction and Microsatellite Analysis
Ear tissue (1 cm2) was sampled from each individual. It was
stored in 70% ethanol at room temperature in the field and
then at 48C until DNA extraction. DNA was extracted from
approximately 25 mg of ear tissue using the DNeasy tissue kit
(Qiagen). Nine bovine microsatellite markers were amplified
(Table 1). They were chosen on the basis of substantial
variability in other deer species (Maudet F, personal
communication; Slate et al. 1998; Talbot et al. 1996).
Amplification of individual DNA was performed using
a multiplex procedure: three groups of three loci were
amplified in single polymerase chain reactions (PCRs)
following Bonnet et al. (2002) (Table 1). Briefly, PCRs were
carried out on 10–100 ng of genomic DNA in a 15 ll mix
volume containing 200 lM each dNTP, 1.5 ll of GeneAmp
103 buffer (Perkin-Elmer), 2.5 mM MgCl2, 2.5 U Taq
polymerase (Ampli Taq gold, Perkin-Elmer), and the primers
at various concentrations (Table 1). Prior to this step,
a fluorescent dye was incorporated at the 59 end of one
primer per primer pair (a single dye per group of loci). The
following PCR conditions were used: 12 min of Taq
polymerase activation at 938C (Perkin-Elmer specification),
35 cycles of 1 min at 948C (denaturation), 1 min at 588C
(annealing), 1 min at 728C (elongation), followed by an
elongation step of 60 min at 728C and 2 h at 258C in a PerkinElmer thermocycler (2400 or 9600). The three PCR products
were mixed together and completed to a final volume of 105
ll by adding sterile water, as described in the Stockmarks
cattle paternity test protocols (Applied Biosystems PE).
Then 0.4 ll were mixed with 2 ll of loading buffer,
Thévenon et al. Vietnamese Sika Deer Genetic Diversity
Figure 1. Locations of the 200 individuals sampled in central Vietnam and of the villages.
denatured at 958C during 5 min, and loaded on 5% Long
Ranger gels. Migration took place in an ABI 377 sequencer.
Gels were scanned using Genescan 2.1 and the genotypes
were determined using Genotyper 2.5.
Statistical Analysis
The per locus variability was quantified by calculating the
allelic range, number of alleles, and gene diversity over the
200 individuals (Genepop 3.1c; Raymond and Rousset 1995).
The sample size per village was sometimes as low as one
individual and the population structure analysis was
therefore conducted using a subset of individuals (144) of
the Nghe An province. This corresponds to those nine
villages in which more than five individuals each were
sampled (Table 2). Allelic frequencies were estimated per
locus and subpopulation, a subpopulation corresponding to
a village. The number of alleles, observed heterozygosity, and
gene diversity were estimated both per locus and subpopulation, and per subpopulation over all loci (see Nei
1987). Departures from Hardy-Weinberg equilibrium at each
locus were computed for each subpopulation using exact
tests (Rousset and Raymond 1995). Departures over all loci
were evaluated in each subpopulation using Fisher’s method.
Genotypic disequilibria were tested using exact tests and the
sequential Bonferroni correction (Rice 1989). The unbiased
estimator f^ of Wright’s inbreeding coefficient FIS was calculated according to Weir and Cockerham (1984). All
calculations and tests above were performed using Genepop
3.1c (Raymond and Rousset 1995).
Departure from drift-mutation equilibrium was tested
using the Bottleneck program (Cornuet and Luikart 1996;
Luikart and Cornuet 1998). A transient excess of expected
heterozygosity relative to that expected under mutation-drift
equilibrium may indicate a bottleneck. The statistical analysis
was carried out using the Wilcoxon test under the stepwise
mutation model (SMM; Ohta and Kimura 1973), the infinite
allele model (IAM; Kimura and Crow 1964), and the twophase model (TPM; DiRienzo et al. 1994). A shift away from
an L-shaped distribution of allelic frequencies was also tested
(Luikart et al. 1998). The tests were applied to the entire
population (n ¼ 200 individuals).
Genotypic differentiation was tested between pairs of
subpopulations using both Fisher’s exact test and a loglikelihood exact test (Goudet et al. 1996). The corresponding
pairwise FST values were estimated following Weir and
Cockerham (1984) using Genepop 3.1c. Isolation by distance
between subpopulations was tested using the Mantel-like test
13
Journal of Heredity 2004:95(1)
Table 1.
Microsatellite loci used with their GenBank and Bovmap references
Locus
GenBank
Bovmap
PCR group
C
Range
k
He
INRA121
IDVGA55
BMC1009
VH110
BM757
BL42
TGLA53
BM203
CSSM43
All loci
X71545
X85071
NR
NR
G18473
NR
NR
G18500
U03824
BTA000813
BTA000802
BTA001204
BTA002786
BTA001319
BTA000661
BTA000755
BTA001040
BTA000704
1
1
1
2
2
2
3
3
3
0.4
0.2
0.2
0.2
0.1
0.1
0.1
0.2
0.4
134–166
196–210
281–303
96–130
163–187
247–277
150–172
218–234
277–233
7
5
5
5
5
9
4
4
7
5.7
0.81
0.68
0.61
0.62
0.46
0.51
0.66
0.57
0.51
0.60
C is the PCR primer concentration (lmol/l). The allelic size range, number of alleles (k), and gene diversity (He) are given for the entire sample (n ¼ 200
individuals). NR: not referenced.
Bovmap references and primers sequences are available at http://locus.jouy.inra.fr/cgi-bin/bovmap/intro2.pl.
(Mantel 1967) implemented in Genepop, and FST/(1 FST)
was regressed on the logarithm of geographical distance
(Rousset 1997). Population structure was further analyzed
using the kinship coefficient between pairs of individuals i
and j (rij), which was estimated for each pair of individuals
and each allele k following Loiselle et al. (1995) as rij ¼ ( pi –
pk )( pj – pk )/ pk (1 pk ) þ 1/(2n 1), where pk is the mean
frequency of allele k in the population sampled, pi and pj are
the frequency of k in individuals i and j (taking the values 0,
0.5, and 1), and n is the sample size. When estimating the
multiallelic coefficient, rij was weighted by its respective
polymorphism, pk (1 pk ). The multilocus rij was the sum of
multiallelic coefficients divided by the sum over all loci and
alleles of pk (1 pk ). The standard error over loci was
estimated using a jackknife method. The average kinship
coefficient over pairs of individuals was computed for 10
distance classes, with an average pairwise distance between
individuals within distance classes of 2.6 km. The hypothesis
of absence of spatial genetic structure was tested within each
class using a bootstrap method (10,000 replicates): spatial
distances were randomly permuted among pairs of individuals, and the estimated value of the kinship coefficient was
compared to the distribution after permutations. The
hypothesis of isolation by distance was tested by plotting
the multilocus kinship coefficient of each pair against the
logarithm of geographical distance, since a linear decrease is
expected in a two-dimensional habitat (Rousset 2000). The
significance of the regression was tested using a Mantel-like
test. The product of (inbreeding coefficient 1) and the
inverse of the regression slope gives an estimate of 4pDr2,
where r2 is the second moment of the dispersal distance
between parents and progeny and D is the density of
individuals (Hardy and Vekemans 1999; Rousset 1997, 2000).
All calculations and tests were performed using AUTOCORG 3.0 (Hardy and Vekemans; available at ohardy@
ulb.ac.be).
The variability of the Vietnamese (n ¼ 200 individuals)
and zoo populations was compared using a Wilcoxon signed
rank test on the mean number of alleles and gene diversity
across loci. The estimates of these two parameters
14
were corrected for sample size using a rarefaction method
implemented in the program Contrib (Petit et al. 1998;
available at www.pierroton.inra.fr/genetics/labo/Software).
Results
Genetic Variability
The nine microsatellites studied were polymorphic, with
a mean number of alleles per locus of 5.7 (range 4–9; Table 1)
over the 200 individuals studied. The gene diversity over all
individuals and loci was 0.60 (Table 1). The subsequent
analyses of genetic variability focused on the nine villages
from the Nghe An province (see Materials and Methods).
Gene diversity across loci varied very little among villages,
from 0.53 in Nghia Loc to 0.64 in Quynh Xuan (Table 2).
Only slightly more variation was observed for the number of
alleles, which may be surprising given the large variation in
sample size (Table 2). Departure from Hardy-Weinberg
proportions was tested in 80 locus-village combinations, and
only six tests were significant at the 0.05 level. This is about
what is expected by chance. However, three tests were
significant in Duc Thanh, and the test over all loci was highly
significant in this subpopulation (Table 2). Tests over all loci
were also significant in two other villages (Nghia Loc and
Quynh Luong). The associated estimates of FIS ranged
between 0.09 and 0.27 (Table 2). As the few significant
heterozygote deficiencies were not systematically associated
with the same locus, we tentatively conclude that there is
limited evidence for null alleles.
Seventeen pairs of loci, out of 322, exhibited genotypic
disequilibrium. Once again, this is about what is expected by
chance. However, the VH110-BM203 pair gave a significant
test in three villages (Duc Thanh, P ¼ .0004; Quynh Giang,
P ¼ .009; and Quynh Luong, P ¼ .042), and VH110-CSSM43
in two villages (Duc Thanh, P ¼ .004; and Quynh Luong,
P ¼ .047). This could result from several processes, including physical or statistical linkage, selection on multilocus
genotypes, or population structure. Since the FIS estimates
differed significantly from zero in Duc Thanh and Quynh
Thévenon et al. Vietnamese Sika Deer Genetic Diversity
Table 2. Sample size and geographic position of the villages from the Nghe An province in which the population structure
analysis was conducted (see also Figure 1)
Village
Sample size
Latitude (north)
Longitude (east)
k
He
Duc Thanh
Tan Son
Nghia Loc
Quynh Giang
Quynh Tan
Quynh Yen
Quynh Minh
Quynh Luong
Quynh Xuan
13
17
5
11
14
19
25
34
6
19.15962
19.19235
19.23257
19.15325
19.24413
19.13218
19.13815
19.15634
19.20762
105.47221
105.49128
105.50440
105.62554
105.64534
105.67514
105.70611
105.71867
105.70308
4.33
3.89
3.11
4.11
4.22
4.00
4.11
4.56
3.33
(1.00)
(0.78)
(1.17)
(0.93)
(0.97)
(1.00)
(1.05)
(1.33)
(0.71)
0.63
0.61
0.53
0.61
0.57
0.57
0.57
0.60
0.64
Ho
(0.11)
(0.15)
(0.28)
(0.13)
(0.18)
(0.13)
(0.17)
(0.14)
(0.08)
0.51
0.67
0.40
0.61
0.60
0.57
0.54
0.57
0.70
FIS
(0.10)
(0.17)
(0.24)
(0.14)
(0.24)
(0.13)
(0.13)
(0.18)
(0.26)
0.19
0.09
0.27
0.00
0.04
0.01
0.07
0.04
0.11
(,103)
(.934)
(.005)
(.500)
(.585)
(.583)
(.060)
(.012)
(.876)
Estimates of the number of alleles (k), gene diversity (He), observed heterozygosity (Ho), and FIS.
The numbers in parentheses are the standard errors (k, He, and Ho) or the P value of the associated exact test (FIS).
Luong, subpopulation structure is the most likely explanation
for these genotypic disequilibria.
There was a significant excess of heterozygosity in the
Vietnamese population relative to mutation-drift expectation
under the IAM (P ¼ .01), but not under both the SMM and
TPM (P ¼ .54 and .82, respectively). The distribution of allelic
frequencies did not show any significant departure from
a standard L-shape in the mode shift test (results not shown).
Population Structure
The overall test of subpopulation differentiation was highly
significant (P , 104), though the associated FST estimate
was only 0.02. Fifteen exact tests of differentiation between
subpopulation pairs, out of 36, produced probability values
below 0.05. The corresponding pairwise FST estimates were
never higher than 0.06 (Table 3). The Mantel-like test
showed that genetic differentiation increased with geographical distance (slope ¼ 0.009, P ¼ .016). The per locus
estimate of the average kinship coefficient decreased with
geographical distance, although there was variation among
loci (Figure 2). In the first distance class (,1 km), the per
locus estimate of the average kinship coefficient was 0.008
(SE ¼ 0.006), and individuals in the first distance classes are
genetically closer than expected by chance (P ¼ .033). The
correlogram including all loci has a significantly negative
slope (0.003; P , 104), suggesting a slight isolation by
distance. The intraindividual kinship coefficient was significant (mean value over loci ¼ 0.041, SE ¼ 0.028, P ¼ .027).
Comparing Vietnamese and Zoo Populations
The European zoo populations had a mean gene diversity of
0.65 (SE ¼ 0.09) and a mean number of alleles of 4.11 (SE ¼
1.17) over all loci. Gene diversity was not significantly
different from that of the Vietnamese population (Wilcoxon
signed rank test, P ¼ .25), but the number of alleles was
significantly less in the zoo population (P ¼ .008). However,
this difference was not significant (P ¼ .43) when the sample
size is corrected using a rarefaction method.
The estimate of FST between the population of European
zoological parks and the Vietnamese individuals was 0.10,
and significantly differed from zero (P , 103). Thirteen
alleles, found in the Vietnamese population, were not
detected in the zoo population, while three alleles from the
zoo population were not identified in the Vietnamese
population.
Discussion
Variability within the Vietnamese Sika Deer Population
Our study indicates that the mean number of alleles and gene
diversity per locus in the Vietnamese sika deer are similar to
the values reported in various populations of the Japanese
sika deer (C. nippon ssp.) (Goodman et al. 2001; Tamate et al.
2000) and in the North American elk (Talbot et al. 1996).
The Vietnamese individuals also do not display significantly
more genetic diversity than the zoo population. That no
Table 3. Exact probability value of differentiation (above diagonal) and associated estimates of FST (below diagonal) per
subpopulation pairs
Duc Thanh Tan Son Nhia Loc Quynh Giang Quynh Tan Quynh Yen Quynh Minh Quynh Luong Quynh Xuan
Duc Thanh
Tan Son
0.01
Nhia Loc
0.05
Quynh Giang
0.02
Quynh Tan
0.06
Quynh Yen
0.03
Quynh Minh
0.03
Quynh Luong 0.02
Quynh Xuan 0.03
0.219
0.05
0.003
0.04
0.01
0.01
0.03
0.02
0.053
0.112
0.03
0.05
0.06
0.04
0.06
0.05
0.970
0.295
0.181
0.02
0.003
0.01
0.004
0.03
,103
0.007
0.128
0.112
0.01
0.02
0.01
0.05
0.033
0.092
0.027
0.249
0.146
0.01
0.01
0.01
0.003
0.041
0.055
0.542
0.012
0.053
0.004
0.02
,103
,103
0.017
0.043
0.136
0.108
0.341
0.870
0.178
0.089
0.019
0.047
0.162
0.009
0.021
0.02
15
Journal of Heredity 2004:95(1)
Figure 2. Estimate of the Loiselle kinship coefficient as
a function of geographic distance. Abscissa: the average
pairwise distance for the pairs of individuals belonging to the
interval in which the kinship coefficient was calculated (in
kilometers). Vertical lines refer to standard errors. Stars show
kinship coefficients differing significantly from zero (P ¼ .033,
.008, and ,103 for the first, second, and ninth class,
respectively). Slope of the regression: 0.0035 (P , 104).
more microsatellite variability is maintained in the Vietnamese population than in the zoo population, despite a higher
population census size in Vietnam than in Europe (about 225
individuals in the world zoo populations, including 180 in
Europe; Rüdloff 1999), might be surprising. The relatively
low level of genetic variability in the Vietnamese population
could be due to ongoing genetic drift and/or a single
bottleneck event that occurred when the population was
founded. There are several sources of genetic drift in the
farm population: the census size has suffered large variation,
the sex ratio is skewed toward males, and a large proportion
of individuals do not reproduce (since some farmers are only
interested in velvet production and not in reproduction). On
the other hand, the zoo populations were founded with
animals originating from several, significantly differentiated
populations, which should explain the relatively high level of
variation found here (Thévenon et al. 2003).
The bottleneck test showed a significant heterozygosity
excess under the IAM, but not under the SMM and TPM.
However, this test is more powerful under the IAM than
under other models of mutation. It is also more prone to
produce a ‘‘wrong’’ bottleneck signal when microsatellite
data are analyzed (Luikart and Cornuet 1998). More
statistical power could have been achieved under the TPM
and SMM if more microsatellite markers were analyzed (see
Luikart and Cornuet 1998). A cautious conclusion is that the
Vietnamese population has not gone through a marked
bottleneck. Note that it is difficult to distinguish the effect of
a past bottleneck from that of a current strong genetic drift.
It would be interesting to assess the current dynamics of
genetic variability; for instance, through temporal sampling
at a short time scale (Berthier et al. 2002; Luikart et al. 1998).
Founding effect and genetic drift might also explain the
large genetic differentiation between the zoo and Vietnamese
populations. Some rare alleles occur exclusively in one
16
population (private alleles) and the zoo population retains
alleles that have been lost in the Vietnamese population.
However, large differentiation and private alleles could also
be explained by hybridization with other sika subspecies.
Hybridization is possible in the zoo population, since the
geographic origin and taxonomic status of founding animals
were not well recorded. In some zoos, different subspecies of
C. nippon are maintained in neighbor parks and hybridization
is therefore not impossible (see Thévenon et al. 2003, for
more details). Concerning the Vietnamese population,
individuals of Dybowksi sika deer were introduced in the
Hanoi zoo and some hybrid animals have possibly been
dispatched among farms (Ratajszczak et al. 1993). However,
more information (e.g., about the number of individuals
introduced) is not available.
Breeding System and Genetic Differentiation
Among Villages
Significant estimates of FIS were detected within some
villages. However, the FIS estimates, as well as the mean
intraindividual kinship values, were quite low. This suggests
that animals are exchanged at a rate that for most villages
prevents extensive inbreeding. A further argument is that
almost as much genetic variability is maintained within
villages as at the scale of the whole study. Discussions with
the farmers show that they are indeed conscious of the
deleterious consequences of inbreeding and avoid mating
related animals.
Our results also show that, at the scale of the Nghe An
province, subpopulations were significantly differentiated,
although this is associated with a relatively low estimate of
FST (0.02). This is consistent with results from factorial
correspondence analyses (not shown) indicating limited
differentiation among populations. Of greater interest is
that our analysis shows a pattern of isolation by distance.
Moreover, the regression slope of the kinship coefficient
estimates on geographical distance was indeed significantly
different from zero. Its small value indicates either a high
local effective population size (D) or a high variance of
dispersal distance (r2) (Rousset 1997; see Materials and
Methods). This agrees with the fact that farmers exchange
animals more among geographically close villages than
among remote villages, mainly as a consequence of difficulties associated with animal transportation. Of course, rare
events of migration at great distances remain possible. The
observed pattern of isolation by distance can be expected
under a model of subdivided population with colonization in
progress (Rousset 2001). It is thus compatible with the fact
that the Vietnamese sika deer population kept on traditional
farms was probably founded from about 200 wild individuals
with subsequent spread toward the neighboring districts
from Ha Tinh and Nghe An provinces. However, a note of
caution should be introduced when interpreting the result of
isolation by distance analyses. First, isolation by distance is
expected over a specific range of dispersal distance (Rousset
1997, 2001). Second, the population studied might have
experienced variation in size, as mentioned above. The
Thévenon et al. Vietnamese Sika Deer Genetic Diversity
consequences of such events on analyses of isolation by
distance are poorly known (Hardy and Vekemans 1999;
Rousset 2001).
Diamond JM, 1989. The present, past and future of human-caused
extinctions. Philos Trans R Soc Lond B Biol Sci 325:469–476.
Conclusion
Eldridge MDB, King JM, Loupis AK, Spencer PBS, Taylor AC, Pope LC,
and Hall GP, 1999. Unprecedented low level of genetic variation and
inbreeding depression in an island population of the black-footed rockwallaby. Conserv Biol 13:531–541.
Our study suggests that the population of Vietnamese sika
deer held on traditional farms retains enough neutral
variability for serious consideration as a source of individuals
in reintroduction programs. Zoo populations might be
slightly more interesting in this perspective. However, zoo
and Vietnamese populations are experiencing markedly
different environmental constraints in terms of climate,
food, and pathogens. Vietnamese individuals would therefore not experience as marked a change in their environment
as zoo individuals if they were to be reintroduced in the wild
in Vietnam. Promoting genetic exchange among populations
using artificial insemination to reduce disease transmission
could therefore be proposed. Further genetic analyses are
required, however, in order to assess the probability of past
hybridization with other sika deer subspecies. These studies
could be based both on mtDNA sequencing (Randi et al.
2001) and microsatellite markers and should include
individuals from as many sika deer subspecies as possible.
Reintroduction in the wild is likely to take place in various
protected areas, since projects for conservation and development of wildlife in Vietnam are ongoing (CIRAD and
NIAH, personal communication). From a genetic point of
view, reintroducing a few individuals would be sufficient,
provided they are adapted to local environmental conditions
and might reach carrying capacity fast enough (Thévenon
and Couvet 2002). Consequently assessment of environment
suitability (e.g., food availability, density of predators) and
population viability should be verified (Sarrazin and
Legendre 2000). A few generations after reintroduction,
genetic analyses could be conducted for assessing the magnitude of genetic drift (Berthier et al. 2002; Luikart et al. 1998)
and the contribution of founding animals to the current
population.
DiRienzo A, Peterson AC, Garza JC, Valdes AM, Slatkin M, and Freimer
MB, 1994. Mutational process of simple-sequence repeat loci in human
populations. Proc Natl Acad Sci USA 91:3166–3170.
Frankham R, Lees K, Montgomery M, England PR, Lowe EH, and Briscoe
DA, 1999. Do population size bottlenecks reduce evolutionary potential?
Anim Conserv 2:255–260.
Franklin IR, 1980. Evolutionary change in small populations. In:
Conservation, biology: an evolutionary-ecological perspective (Soulé ME
and Wilcox BA, eds). Sunderland, MA: Sinauer Associates; 135–149.
Goldstein DB and Schlötterer C, 1999. Microsatellites: evolution and
applications (Goldstein DB and Schlötterer C, eds). New York: Oxford
University Press.
Goodman SJ, Tamate HB, Wilson RA, Nagata J, Tatsuzawa S, Swanson
GM, Pemberton JM, and McCullough DR, 2001. Bottlenecks, drift and
differentiation: the population structure and demographic history of sika
deer (Cervus nippon) in the Japanese archipelago. Mol Ecol 10:1357–1370.
Goudet J, Raymond M, de Meeus T, and Rousset F, 1996. Testing
differentiation in diploid populations. Genetics 144:1933–1940.
Hardy OJ and Vekemans X, 1999. Isolation by distance in a continuous
population: reconciliation between spatial autocorrelation analysis and
population genetics models. Heredity 83:145–154.
Hilton-Taylor C, 2000. IUCN red list of threatened species Gland,
Switzerland: IUCN.
Huynh DH, Duc TV, and Khien HM, 1990. The status of endangered
species of deer in Vietnam Hanoi, Vietnam: National Centre for Scientific
Research.
Kimura M and Crow JF, 1964. The number of alleles that can be
maintained in a finite population. Genetics 49:725–738.
Lande R and Arnold SJ, 1983. The measurement of selection on correlated
characters. Evolution 37:1210–1226.
Loiselle BA, Sork VL, Nason J, and Graham C, 1995. Spatial genetic
structure of a tropical shrub, Psychotoria officinalis (Rubiaceae). Am J Bot
82:1420–1425.
Luikart G and Cornuet JM, 1998. Empirical evaluation of a test for
identifying recently bottlenecked populations from alleles frequencies data.
Conserv Biol 12:228–237.
References
Luikart G and England PR, 1999. Statistical analysis of microsatellite DNA
data. Trends Ecol Evol 14:253–256.
Berthier P, Beaumont MA, Cornuet JM, and Luikart G, 2002. Likelihoodbased estimation of the effective population size using temporal changes in
allele frequencies: a genealogical approach. Genetics 160:741–751.
Luikart G, Sherwin W, Steele BM, and Allendorf FW, 1998. Usefulness of
molecular markers for detecting population bottlenecks via monitoring
genetic change. Mol Ecol 7:963–974.
Bonnet A, Thévenon S, Maudet F, and Maillard JC, 2002. Efficiency of
semi-automated fluorescent multiplex PCRs with eleven microsatellite
markers for genetic studies of deer populations. Anim Genet 33:343–350.
Lynch M, Conery J, and Burger R, 1995. Mutation accumulation and the
extinction of small populations. Am Nat 146:489–518.
Chardonnet P, 1995. Faune sauvage africaine: la ressource oubliée. Tome 1
Luxembourg: European Commission.
Cornuet JM and Luikart G, 1996. Description and power analysis of two
tests for detecting recent population bottlenecks from allele frequency data.
Genetics 144:2001–2014.
Coulson TN, Pemberton JM, Albon SD, Beaumont M, Marshall TC, Slate J,
Guiness FE, and Clutton-Brock TH, 1998. Microsatellites reveal heterosis
in red deer. Proc R Soc Sci Lond 265:489–495.
Crow J and Kimura M, 1970. An introduction to population genetics theory
Minneapolis: Burgess.
Mantel N, 1967. The detection of disease clustering and a generalized
regression approach. Cancer Res 27:209–220.
Nei M, 1987. Molecular evolutionary genetics New York: Columbia
University Press.
Ohta T and Kimura M, 1973. A model of mutation appropriate to estimate
the number of electrophoretically detectable alleles in a finite population.
Genet Res 22:201–204.
Petit RJ, El Mousadik A, and Pons O, 1998. Identifying populations for
conservation on the basis of genetic markers. Conserv Biol 12:844–855.
Pimm SL and Raven P, 2000. Biodiversity. Extinction by numbers. Nature
403:843–845.
17
Journal of Heredity 2004:95(1)
Pimm SL, Russel GJ, Gittleman JL, and Brooks TM, 1995. The future of
biodiversity. Science 269:347–350.
Randi E, Mucci N, Claro-Hergueta F, Bonnet A, and Douzery E, 2001. A
mitochondrial DNA control region phylogeny of Cervidae: speciation in
Cervus and implications for conservation. Anim Conserv 4:1–11.
Ratajszczak R, Adler J, and Smielowski A, 1993. The Vietnamese sika deer
Cervus nippon pseudaxis conservation project. Int Zool Yearbook 32:56–60.
Raymond M and Rousset F, 1995. An exact test for population
differentiation. Evolution 49:1280–1283.
Rice WR, 1989. Analyzing tables of statistical tests. Evolution 43:223–225.
Rousset F and Raymond M, 1995. Testing heterozygote excess and
deficiency. Genetics 140:1413–1419.
Rousset F, 1997. Genetic differentiation and estimation of gene flow from
F-statistics under isolation by distance. Genetics 145:1219–1228.
Rousset F, 2000. Genetic differentiation between individuals. J Evol Biol
13:58–62.
Rousset F, 2001. Inference from spatial population genetics. In: Handbook
of statistical genetics (Balding D, Bishop M, and Cannings C, eds). New
York: Wiley; 239–269.
Talbot J, Haigh J, and Plante Y, 1996. A parentage evaluation test in North
American elk (wapiti) using microsatellites of ovine and bovine region.
Anim Genet 27:117–119.
Tamate H, Okada A, Minami M, Ohnishi N, Higuchi H, and Takatsuki S,
2000. Genetic variation revealed by microsatellite markers in small
population of the sika deer (Cervus nippon) on Kinkazan island, northern
Japan. Zool Sci 17:47–53.
Templeton AR, 1986. Coadaptation and outbreeding depression. In:
Conservation biology, the science of scarcity and diversity (Soulé ME and
Wilcox BA, eds). Sunderland, MA: Sinauer Associates; 105–116.
Thévenon S, Bonnet A, Claro F, and Maillard JC., 2003. Genetic diversity
analysis for captive propagation: the Vietnamese sika deer (Cervus nippon
pseudaxis) in zoological parks. Zoo Biol 22:465–475.
Thévenon S and Couvet D, 2002. The impact of inbreeding depression on
population survival depending on demographic parameters. Anim Conserv
5:53–60.
Vida G, 1994. Global issues of genetic diversity. In: Conservation genetics
(Loeschcke V, Tomiuk J, and Jain SK, eds). Basel, Switzerland: Birkhäuser
Verlag; 9–19.
Rüdloff K, 1999. International studbook of the Vietnamese sika deer.
Berlin: Tierpark Berlin.
Weir BS and Cockerham CC, 1984. Estimating F-statistics for the analysis
of population structure. Evolution 38:1358–1370.
Saccheri I, Kuussaari M, Kankare M, Vilkam P, Fortelius W, and Hanski F,
1998. Inbreeding and extinction in a butterfly metapopulation. Nature
392:491–494.
Wemmer C, 1998. Deer Status survey and conservation action plan (IUCN/
SSC Deer Specialist Group, ed). Gland, Switzerland: IUCN.
Sarrazin F and Legendre S, 2000. Demographic approach to releasing adults
versus young in reintroductions. Conserv Biol 14:488–500.
Slate J, Coltman DW, Goodman SJ, MaLean I, Pemberton JM, and
Williams JL, 1998. Bovine microsatellite loci are highly conserved in red
deer (Cervus elaphus), sika deer (Cervus nippon) and Soay sheep (Ovis aries).
Anim Genet 29:307–315.
18
Whitehead GK, 1993. The Whitehead encyclopedia of deer. United
Kingdom: Swan-Hill Press.
Received November 8, 2002
Accepted August 30, 2003
Corresponding Editor: C. Scott Baker