Comparison of levels of genetic diversity detected with AFLP and ...

clusteriranianΒιοτεχνολογία

23 Οκτ 2013 (πριν από 4 χρόνια και 2 μήνες)

148 εμφανίσεις



1

Comparison of levels of genetic diversity detected with AFLP and microsatellite
1

markers within and among mixed
Q.

petraea

(Matt.) Liebl. and
Q
.

robur

L. stands

2

Envoyé à Silvae Genetica le 30/05/01

3

Stéphanie Mariette
1
*
, Joan Cottrell
2
, Ulrike M. Csaikl
3
, Pa
blo Goikoechea
4
, Armin König
5
,
4

Andrew J. Lowe
6
, Barbara C. Van Dam
7
, Teresa Barreneche
1
, Catherine Bodénès
1
, Réjane
5

Streiff
1
,

Kornel Burg
3
,
Katrin Groppe
5
,

Robert C. Munro
6
, Helen

Tabbener
2

and Antoine
6

Kremer
1

7


8

1
INRA, Equipe de Génétique et Amélioration de
s Arbres Forestiers, BP45, F
-
33610
9

Cestas, France
.

10

2
Forest Research, Northern Research Station, Roslin, Midlothian, Scotland, EH25 9SY,
11

United Kingdom
.

12

3
Ö
sterreichisches Forschungszentrum, Seibersdorf Ges. m. b. H., A
-
2444 Seiberdorf,
13

Austria
.

14

4
NEIKER, Dpt
o Producción y Protección Vegetal, Granja Modelo
-
Arkaute, Apdo 46,
15

01080 Vitoria Gasteiz, Spain
.

16

5
BFH, Institut für Forstgenetik und Forstpflanzenzuechtung, Sieker Landstrasse 2,
17

22927 Grosshandorf, Germany.

18

6
Centre for Ecology and Hydrology, CEH, Edinburg
h Bush Estate, Penicuik,
19

Midlothian, EH26 0QB, Scotland, United Kingdom.

20

7
Alterra Green World Research, POB 47, Droevendaalsesteeg 3, 6700 AA Wageningen,
21

The Netherlands.

22

23



2

*
Corresponding author

:

1

Stéphanie Mariette

2

INRA, Laboratoire de Génétique et Amélior
ation des Arbres
3

Forestiers, BP45, 33610 Cestas, France.

4

Tel: +33
-
5
-
57
-
97
-
90
-
83

5

Fax: +33
-
5
-
57
-
97
-
90
-
88

6

e
-
mail: stephanie.mariette@pierroton.inra.fr

7


8


9

Running title: AFLP markers, microsatellites and
Quercus

spp.

10

11



3

Summary

1

In this study, we compare the genet
ic diversity within and among
Quercus

spp. populations
2

assessed with two contrasting types of molecular markers: a limited number of highly
3

polymorphic microsatellite markers and
numerous less informative AFLP markers.
Seven
4

mixed stands of
Quercus petraea

and
Quercus robur

were analysed with six microsatellite
5

markers and 155 AFLP loci. Genetic differentiation and genetic diversity within each
6

population was assessed. The intra
-

and inter
-
locus variances were calculated and the results
7

were used to compare

the genetic diversity between populations. The rankings of populations
8

provided by the two types of markers were compared. The results obtained with the two
9

types of markers revealed the same general trend. The genetic diversity within population and
10

the
genetic differentiation among populations were greater in
Q.

petraea

than in
Q. robur
. The
11

genetic differentiation was generally higher when AFLP markers were used as compared to
12

microsatellites, and it was also the case when only polymorphic AFLP fragment
s were used.
13

For AFLP, the inter
-
locus variance was always much higher than the intra
-
locus variance, and
14

explains why it was not possible to distinguish populations based on the level of diversity for
15

this marker system. Finally, no significant positive c
orrelation was found between the level of
16

within
-
population assessed with the two markers.

17

Keywords: microsatellite, AFLP, genetic diversity, genetic differentiation,

Quercus robur
,
18

Quercus petraea
.

19

20



4

Introduction

1

The assessment of genetic diversity with mo
lecular markers in natural populations follows a
2

two
-
stage sampling: (
i)

sampling of populations and individuals and (
ii
) sampling of loci within
3

the genome. The associated components of sampling variance have been termed “intra
-
locus
4

variance” and “inter
-
locus variance” respectively and in theory the inter
-
locus variance should
5

be much higher than the intra
-
locus component (Nei 1987). On the basis of allozymic data,
6

Nei concluded that “a large number of loci should be examined even if the number of
7

individ
uals per locus is small”. The larger inter
-
locus variance is likely to be the result of large
8

differences in the mutation rates across the loci within the genome. There have been major
9

advances in molecular techniques in recent years and as a result there
is currently a wide range
10

of markers available (Karp et al. 1997). Markers such as microsatellites are based on
11

sequence data and this makes their development expensive. However, they have the
12

advantage that they are codominant markers. In contrast markers

such as Random Amplified
13

Polymorphic DNA (RAPD) and Amplified Fragment Length Polymorphism (AFLP) are
14

cheap to develop as no knowledge of DNA sequence is required for their development. They
15

provide information on many loci which are randomly distributed
throughout the genome;
16

however they are usually dominant markers (Breyne et al. 1997). Consequently, for a given
17

investment of time and money they provide information on a wider range of loci than
18

microsatellites, but the information at a given locus is le
ss specific. As a result, for a given
19

amount of resources, two contrasting sampling strategies can be adopted to assess genetic
20

diversity using molecular markers: (
i
) selection of highly informative markers at a few loci
21

(microsatellites), (
ii
) sampling of

numerous less informative markers randomly distributed
22

within the genome (RAPD or AFLP). It is not yet clear that these two extreme strategies will
23



5

produce similar results when used to measure within
-

and among
-
populations diversity. Most
1

of studies that
compare different types of markers focused on allozymes and RAPD markers:
2

Baruffi et al.
(1995), Cagigas et al. (1999), Isabel et al. (1995), Lannér
-
Herrera
et al.

(1996)
3

or Le Corre et al.
(1997) all gave allozyme and RAPD diversity data sets that do not
provide
4

congruent results. Still few comparative studies involve AFLP though they provide a high
5

number of markers. We report here on the comparison of the levels of genetic diversity within
6

and among populations of two closely related white oak species
Q.

petraea

(Matt.) Liebl. and
7

Q. robur

L. using these two contrasted types of nuclear molecular markers: microsatellite and
8

AFLP.

9

The two species are sympatric and generally occupy different but proximal ecological niches.
10

However, leaf and fruit interspecif
ic differences are clearly recognized (Dupouey & Badeau
11

1993). Genetic variation in
Q. petraea

and
Q. robur

populations has previously been
12

analysed in several studies using allozymes (Müller
-
Starck & Ziehe 1991, Kremer et al. 1991,
13

Müller
-
Starck et al. 19
93; Kremer & Petit 1993). Nevertheless, due to different population
14

sampling strategies, results were not congruent and could not be directly compared.
Zanetto et
15

al.
(1994) showed that the level of diversity within
Q. petraea

populations is slightly highe
r
16

than within
Q
.
robur

and that the proportion of variation partitioned among populations within
17

species is low for both species. In addition,
Q. petraea

populations are more differentiated
18

than those of
Q. robur
. Studies with allozymes also showed that bo
th species share the same
19

alleles and exhibit only small differences in allele frequencies, the two species exhibit extremely
20

low genetic intraspecific differentiation. Analysis of total proteins confirmed results found with
21

allozymes that is a low level o
f genetic differentiation between
Q. petraea

and
Q. robur

22

(Barreneche et al. 1996).
Bodénès et al.
(1997b) investigated the geographic variation of the
23



6

species differentiation throughout their natural range. After screening 2800 PCR amplification
1

products
using random primers, they found only two per cent of the amplified fragments that
2

exhibited significant frequency differences between the two species and none of them was
3

species specific. Finally, Bacilieri et al. (1994) and Streiff et al. (1998) studied

the spatial
4

genetic structure of the two species in the same oak mixed stand (“La Petite Charnie”)
5

respectively with allozymes and microsatellites: only slight differences in the levels of genetic
6

diversity were found between the two species.

7

Within the f
ramework of a research project supported by the European Union, seven mixed
8

Q. petraea

and
Q. robur

stands were selected in six different countries. Within each stand,
9

every tree was analysed with six microsatellite markers by each laboratory within the pr
oject.
10

In addition, approximately 45 samples of each species from each stand were screened with
11

155 AFLP loci by one of the laboratory (INRA). These data sets were used to compare the
12

levels of diversity within and between the two species. The objectives o
f this study were
13

twofold: (
i
) to compare contrasting marker systems for the assessment of gene diversity in
14

oaks. The main purpose was to verify whether these markers would provide the same ranking
15

when populations (within each species) were ordered by th
eir level of diversity and (
ii
) to
16

compare the level of diversity among populations (within each species) by considering the
17

whole genome. A method was developed here to assess the inter
-
locus sampling variance and
18

the intra
-
locus sampling variance of gene

diversity and was used to compare the level of
19

diversity between populations.

20


21



7

Material and Methods

1

Sampling of stands

2

Seven stands from six countries were selected. The criteria for selection were as follows: (
i
)
3

the stand should be mixed and comprise
Q.

petraea

and
Q
.
robur

in approximately equal
4

proportions. The stands should consist of three zones: two monospecific zones and one zone
5

where the two species were mixed tree by tree, (
ii
) the stand should be of natural origin, (
iii
)
6

the stand should consis
t of adult trees (more than 120 years old), (
iv
) the population size for
7

each species should be close to 200. Although these were the properties of an ideal site, it
8

was not possible to fulfill all these criteria at every site. The sampling within the stan
d was
9

exhaustive, since all trees on a given area were
analysed
. A standard protocol based on leaf
10

morphology characters was established so that the distinction between the two species was
11

based on the same criteria in the different countries (Dupouey & Ba
deau 1993). A principal
12

component analysis (PCA) of 14 leaf characters enabled each tree to be assigned to a
13

species. Trees exhibiting intermediate morphology were excluded from the analysis.

14

The location of stands is given on Figure 1. The list of stands
and their composition are given in

15

Table 1.

16

Microsatellite scoring

17

Beforehand, a technical workshop was organized to standardize the methods. All trees in the
18

study were genotyped using six microsatellite loci: ssrQpZAG9, ssrQpZAG36,
19

ssrQpZAG104 and ssrQpZ
AG1/5 (Steinkellner et al. 1997), MSQ4 and MSQ13 (Dow et al.

20

1995) by each laboratory. The extraction, amplification and detection protocols for these
21

microsatellites are described in Streiff et al. (1998). A test cross was further implemented in
22

order to
compare the scoring procedure between laboratories. Each participant sent 10 DNA
23



8

extracts to the INRA laboratory who performed the comparison between his own scoring and
1

the scoring procedure of the different participants. The test cross indicated five dif
ferent
2

scoring discrepancies among laboratories resulting from:

3

(Error 1) a systematic shift in allele size. For example, the allele that was scored as 202 by the
4

INRA laboratory was actually scored as 205 by the Austrian laboratory. The difference in
5

alle
le size was constant across the range of allele sizes.

6

(Error 2) a systematic shift in allele size. However the difference in allele size was
not

constant
7

across the range of allele sizes. The difference amounted to a certain value when the allele size
8

was

lower than a given threshold, and then changed above this threshold.

9

(Error 3) a random variation of allele sizes. There were occasionally discrepancies between
10

allele sizes. For example, one allele was for example scored as 203 by one of laboratory and
11

2
04 by the INRA one.

12

(Error 4) differences in genotype identification, especially inconsistencies in differentiating
13

heterozygotes and homozygotes. In some cases a tree bearing for example the alleles 203 and
14

205 was scored as heterozygote by the INRA labor
atory and as homozygote (205
-
205) by
15

another laboratory. This mainly occurred when the two alleles exhibited a small difference in
16

size.

17

(Error 5) a miscoring of rare, high molecular weight alleles. For a few loci, there were alleles of

18

unusually extreme s
ize corresponding to either a deletion or insertion in flanking regions. In
19

general, there were small discrepancies in the assessment of the size of these alleles scored
20

across the laboratories.

21



9

AFLP scoring

1

Within

each stand, approximately 45 randomly sel
ected trees of each species were screened
2

in the INRA laboratory with four AFLP Primer
-
Enzyme Combinations following the protocol
3

described in Gerber et al. (2000): P
st
I+CAG / M
se
I+CAA, P
st
I+CAG / M
se
I+GCA,
4

P
st
I+CAG / M
se
I+GGA and P
st
I+CCA / M
se
I+CAA. The
RFLPscan version 3.0
5

(Scanalytics) software was used to score the AFLP fragments. The STR marker (purchased
6

by LI
-
COR, Biotechnology Division) was used to determine accurately the sizes of individual
7

fragments.

8

The four AFLP Primer
-
Enzyme combinations prov
ided 155 scorable loci of which seventy per

9

cent were polymorphic in at least one population (Table 2).

10

Data analysis of microsatellite markers

11

The following standard genetic parameters were estimated for each species and each stand
12

(Brown & Weir 1983):

al
lelic richness (
A
), effective number of alleles (
A
E
=1/(1
-
H
E
)),
13

observed heterozygosity (
H
O
), expected heterozygosity (
H
E
) and fixation index (
F
IS
). In
14

addition, the within
-
population gene diversity (
H
i
), the mean within
-
population gene diversity
15

(
H
S
), the
total diversity (
H
T
) and the genetic differentiation (
G
ST
) were calculated following
16

Nei’s procedure (1987). Parameter estimates were made as the mean value across the
17

different loci. The coefficient of gene differentiation among populations (
G
ST
) was comp
uted
18

between both species in each stand and among populations of each species.
Those parameters

19

were computed using the DIPLOIDE program (Antoine Kremer, Equipe de Génétique et
20

Amélioration des Arbres Forestiers, Cestas, France).

21

Because of the discrepanci
es with the scoring of microsatellites between the different
22

laboratories the estimation of genetic diversity was performed in two different ways:

23



10

(
1
st

analysis
)


the comparison of diversity between
Q. petraea

and
Q. robur

was
1

done separately within each s
tand, by using the scoring procedure developed by the
2

laboratory in charge of the given stand.

3

(
2
nd

analysis
)


the comparison of diversity of
Q. petraea
/
Q. robur

populations across
4

sites was performed after transforming the original data by taking into acc
ount the
5

discrepancies. The transformation of data could be done when systematic discrepancies were
6

identified (systematic shift of allele size, constant or not across the range size of alleles).
7

Corrections were made according to the results obtained by t
he test cross and allele sizes
8

were shifted accordingly. Furthermore, alleles differing by one base pair and present in low
9

frequencies were merged into common allelic classes in order to correct for random variation
10

of allele sizes. In this case, the comp
arison of diversity was restricted only to the expected
11

heterozygosity (
H
E
) and to the within
-
population gene diversity (
H
i
) which are known to be
12

less sensitive to small changes in allele frequencies than the other gene diversity statistics as the
13

allelic

richness
A
.

14

Data analysis of AFLP markers

15

The analysis of the AFLP markers was based on the assumption that each AFLP amplification
16

product, regardless of its relative intensity, corresponded to a dominant allele at a unique
17

locus. Polymorphic amplified l
oci were scored as “1” for the presence and “0” for the absence
18

of a locus. Only major well
-
resolved amplified loci were used for data analysis.

19

Both phenotypic and genotypic types of analysis were performed on the AFLP data set.
20

Phenotypic analysis consid
ered two types of variants: the individuals exhibiting a band
21

(frequency
P
) and those without the band (frequency
Q
).
P

and
Q

were deduced directly from
22



11

the DNA electrophoretic profiles and used to compute
H
i
,

H
S
,
H
T
, and
G
ST

at the phenotypic
1

level (param
eters with the same definition as described above).

2

Genotypic analysis considered the frequencies,
p

and
q
, of alleles responsible for the presence
3

or the absence of bands respectively. A hypothesis of genetic structure allows
p

and
q

to be
4

deduced from
Q
:

if the deficiency of heterozygotes (estimated by
F
IS
) is known, then
5

. Assuming that the true value
F
IS

is known, an asymptotically unbiased
6

estimate of
q

is obtained by the use of a second order Taylor expression (Kendall & Stuart
7

197
7):

8





(1)

9

with

and
N

the number of trees sampled per population.

10

Genotypic analysis was performed using the
F
IS

value that was estimated with the average of
11

the six microsatellite loci. We performed the genotypic
analysis over all loci and we also
12

restricted the analysis to loci that showed an observed frequency smaller than (1
-
(3/
N
)),
13

where N is the population sample size, as recommended by Lynch & Milligan (1994). Any
14

fragment that exhibited a higher frequency th
an (1
-
(3/
N
)) in a single population was removed
15

from the whole data set. Lynch & Milligan (1994) showed that the bias introduced to the
16

estimation of
q

due to a small sample size was substantial when the null allele was rare. Gene
17

diversity statistics were

computed by using the allelic frequencies as estimated by formula (1).
18

The two analyses are respectively denoted G1 and G2 analysis in the following text.

19

The phenotypic and genotypic analysis were performed by respectively using the HAPLOID
20

and the HAPDO
M programs (Antoine Kremer, Equipe de Génétique et Amélioration des
21

Arbres Forestiers, Cestas, France).

22



12

Intra
-

and inter
-
locus sampling variances

1

The total sampling variance of gene diversity statistics (
A, A
E
,
H
O
,
H
E
,
H
i
,
H
S
,
H
T
,
F
IS

and
2

G
ST
) is due to a
two step sampling procedure: sampling of individuals within populations (
V
intra
-
3

locus
) and sampling of loci within the genome (
V
inter
-
locus
). The total sampling variance is
V
total

=
4

V
intra
-
locus

+
V
inter
-
locus
.

5

We estimated these two components by using re
sampling methods (bootstrap). All resampling
6

procedures were done with replacement. One thousand bootstrap samples were made each
7

time for estimating the sampling variances.

8

V
intra
-
locus

was estimated by resampling individuals within populations. For
G
ST
,
V
intra
-
locus

was
9

estimated by resampling populations as suggested by Petit & Pons (1998).

10

V
inter
-
locus

was estimated by resampling loci across individuals.

11

Statistical test of differences between populations

12

The distributions of the diversity statistics es
timated by bootstrapping
were used to test for the
13

difference between two populations a and b (a and b being the two species populations from
14

the same stand or being populations of the same species from two different stands). For
15

example, in the case of
H
i
,
values of
H
i
a

and
H
i
b

were calculated for each bootstrap sample in
16

each population as well as the difference of
H
i

between the two populations (
H
i
a



H
i
b
)
. The
17

distribution of (
H
i
a



H
i
b
)
was then compared with the null hypothesis (
H
i
a



H
i
b

=
0) and the

18

associated probability
p

was calculated.

19

Statistical test of AFLP allelic frequencies differences between species

20

At the species level and within each stand, the frequencies of AFLP markers were compared
21

between the two species by performing Fisher’s exac
t tests.

22



13

Comparison of diversity statistics between markers

1

The
Q. petraea

and
Q. robur

populations were ranked according to the parameters
A
,
A
E
,
2

H
O

and
H
i

for microsatellites and according to
H
i

for AFLPs. The value of these parameters
3

were then compared

by computing Spearman’s rank coefficient correlation:
r
S

(Sokal & Rohlf
4

1995).

5


6


7


8

Results

9

1
-
Difference of gene diversity between Q. petraea and Q. robur

10

1a
-
Microsatellites

11

For every stand, the genetic diversity as measured by microsatellites was higher wi
thin the
12

Q.

petraea

population than within
Q.

robur
. Results shown in Table 3 were obtained by
13

keeping
the scoring procedure developed by the laboratory responsible for a given stand (
1
st

14

analysis
). Since the two species within a stand were scored the same

way, the comparison of
15

species was not affected by between laboratory discrepancies of scoring.
There was at least
16

one measure of diversity (among
A
,
A
E
,
H
E
,
H
i

and
H
O
) per stand that showed higher values in
17

Q.

petraea

than in
Q
.
robur
. The only site in w
hich
Q. robur

ever demonstrated a higher
18

measure of genetic diversity was Dalkeith Old Wood but this was probably due to the very
19

low numbers of
Q. petraea

at this site (Table 1). The differences in the levels of diversity were
20

usually small but significan
t when the intra
-
locus standard deviation was used. In three stands,
21

the fixation index was significantly higher in
Q. robur

than in
Q. petraea
, indicating a higher
22

excess of homozygotes in
Q. robur

than in
Q. petraea

in those stands.

23



14

1b
-
AFLPs

1

The phenotyp
ic analysis and the G1 analysis (with all bands) based on AFLP data indicated a
2

higher genetic diversity within
Q
.
petraea

for five out of the seven stands (Table 4). None of
3

the results was significant when the total standard deviation was used to compare

between the
4

species within each stand. However, when the comparison was based only on the intra
-
locus
5

standard deviation, the phenotypic analysis indicated significantly higher genetic diversity for
Q.
6

petraea

in Petite Charnie, Escherode and Meinweg (dat
a not shown). The G2 analysis
7

(reducing the number of analysed bands) inverted results for Petite Charnie, Escherode,
8

Dalkeith Old Wood and Meinweg but results were not significant. The standard deviation due
9

to sampling different loci within the genome wa
s always higher than the standard deviation due
10

to the sampling of individuals, whatever the adopted analysis. There was also a higher sampling

11

variance associated with the G2 analysis for both components (intra and inter
-
locus).

12

1c
-
Microsatellite and AFLP

analysis at the species level

13

When the data were pooled from the seven populations on a species basis, both microsatellite
14

and AFLP markers indicated a higher genetic diversity within
Q
.
petraea

for all parameters
15

and analyses (Table 5). The only exceptio
n was the G2 analysis of AFLPs that gave similar
16

levels of diversity in the two species (0.252 and 0.256 for
Q. petraea
and
Q. robur
17

respectively). The genetic diversity in
Q. petraea

was only significantly higher when the intra
-
18

locus variance was consider
ed and was never found to be significantly greater when the total
19

sampling variance was used.

20



15

2
-
Genetic differentiation between Q. petraea and Q. robur

1

The genetic differentiation (
G
ST
) between
Q. petraea

and
Q. robur

as measured by
2

microsatellites was low

(Table 6a), ranging from 0.005 (Roudsea Wood) to 0.024
3

(Sigmundsherberg).

4

The genetic differentiation detected using AFLP markers was higher when the phenotypic
5

analysis was applied as compared to the genotypic analyses (Table 6a). There was also an
6

impor
tant difference between the G1 and G2 methods. In comparison to microsatellites, the
7

genetic differentiation found by AFLP markers was higher, even using the G2 method.
8

Furthermore, there was no correlation between genetic differentiation among markers, ex
cept
9

between the phenotypic and the G1 analysis (
r
S
=0.883,
p
=0.034).

10

3
-
Genetic differentiation among populations within species

11

The genetic differentiation among
Q
.
petraea

populations markers (0.023) was higher than
12

among
Q
.
robur

populations (0.020) when

microsatellites were used but the difference was
13

not statistically significant (Table 6b).

14

AFLP markers did not show a significantly higher genetic differentiation among
Q
.
petraea

15

populations. The genetic differentiation assessed using microsatellites am
ong
Quercus

robur

16

populations or among
Quercus

petraea

were not significantly different from the differentiation
17

found with AFLP with the G2 method. However, when the phenotypic and G1 methods were
18

applied to the AFLP data, the differentiation was found to

be significantly higher than that
19

found using microsatellites.

20

4
-
Distribution curves of genetic differentiation for AFLPs

21

As a higher differentiation was found using AFLP analysis compared to microsatellites, the
22

variation of
G
ST

values among loci for the
se markers was further analysed. The distribution of
23



16

G
ST

values is given in Figure 2, 3 and 4. The three curves resemble an L
-
shaped distribution
1

where more extreme values occurred among populations rather than between species. For
2

example, only four loci
exhibited a differentiation greater than 10% between
Q. petraea

and
3

Q. robur

populations whereas 36 loci among
Q
.
petraea

populations and 23 loci among
Q
.
4

robur

populations demonstrated this level of differentiation. At the species level, the null
5

hypothes
is was rejected for independence between the observed frequencies and the species
6

for 39% of the loci. This percentage of loci ranged between 9% (Dalkeith Old Wood) and
7

22% (Escherode).

8

5
-
Comparison of levels of diversity across stands

9

The comparison of
H
E

or
H
i

values across stands for the two species and the two markers is
10

given in Table 7. For microsatellites, differences between stands were statistically significant
11

and the overall ranking of stands was very similar when the
Q
.
petraea

and
Q
.
robur

port
ions
12

of the stands were considered independently (
r
S
=0.769,
p
=0.043) and the two extreme values

13

were the same for both species. Roudsea Wood exhibited the highest value of diversity and
14

Sigmundsherberg the lowest.

15

For AFLP markers, no significant differenc
e was detected among stands when the total
16

standard deviation was used. For each analysis, the correlation between the ranking for
Q.
17

petraea

and the ranking for
Q. robur

was positive but never significant.

18

6
-
Comparison of levels of diversity assessed with

different measures or with different
19

markers

20

No significant correlation was found when rankings given by
A

and
H
i

for the microsatellite
21

data were compared (Table 8). For AFLP, rankings given by the G2 analysis for
Q
.
petraea

22

tended to be different from t
he rankings given by the two other methods of analysis but the
23



17

result was not significant. For
Q
.
robur
, the three analyses were congruent and the correlations

1

were significant.

2

When
H
i

for the microsatellite data was compared with
H
i1
,
H
i2

or
H
i3

for the
AFLP data, no
3

significant correlation was found.

4

Discussion

5

Intra
-

and inter locus variances of gene diversities

6

Our experimental results confirm Nei’s prediction that a larger variance is attributable to the
7

effect of sampling different loci within a geno
me than to the sampling of individuals within a
8

population. The AFLP data demonstrate that the former source of variance can be up to 15
9

times greater than the latter. Interestingly, both microsatellites and AFLPs provided similar
10

estimates of intra
-
locus
variance, which suggests that the sampling variance is independent of
11

the number of alleles (Table 3 and Table 4). If differences in the levels of diversity have to be
12

assessed on the whole genome, the number of loci rather than the number of individuals s
hould
13

be as great as possible. This is also likely to be the case when monitoring of gene diversity is
14

done for conservation purposes. When the attributes for which diversity is assessed are
15

unknown, diversity should be measured at the whole genome level,
by using a random set of
16

markers distributed throughout the genome. However, a larger sampling variance would be
17

expected, which would lead to a reduction in the power of any statistical test applied to
18

measure diversity differences among populations. This

is clearly reflected in our results. There
19

is a trend towards higher genetic diversity in
Q. petraea
, although the difference between the
20

two species is not significant. More AFLP markers would have been necessary to reveal
21

significant differences between

the two species.

22



18

Comparison of AFLPs and microsatellites for measuring gene diversity and
1

differentiation

2

When examining genetic diversity within the two species, the two markers show similar trends
3

and indicate that
Q. petraea

is more variable than
Q. ro
bur

(Table 3 and Table 4).
4

Comparable results are only obtained as long as the analysis is done at the phenotypic level
5

and at the genotypic level with all the markers for AFLP data. When the analysis is restricted
6

to the subset of polymorphic markers only
, following the recommendation by Lynch &
7

Milligan (1994), there is an increase in the sampling variance at the intra
-

and inter
-
locus level.
8

This increase in variance is likely to be due to the reduction in the number of loci (from 155
9

loci to 61 loci). F
or comparative analysis of levels of diversity between populations, it would
10

therefore be preferable to use all markers, polymorphic and monomorphic.

11

For genetic differentiation, contrasting results are obtained between microsatellite and AFLP
12

markers. In
general, AFLP markers exhibit higher levels of differentiation than microsatellites.
13

There may be two explanations for these observations. First, mutation rates are higher in
14

microsatellites and cannot be ignored when compared to migration rates. Both muta
tion and
15

migration tend to decrease population differentiation (Jin & Chakraborty 1995, Rousset 1996,
16

Slatkin 1995). Second, there are more AFLP loci than microsatellites and the likelihood that
17

some of the AFLP markers are linked to adaptative traits cann
ot be excluded. Oak
18

populations are known to be highly differentiated for growth and phenological traits (Ducousso
19

1996). As a result, we might expect a high heterogeneity of
G
ST

values for different AFLP
20

fragments. This is found to be the case and is show
n by Figures 1, 2 and 3 where the
21

distribution of
G
ST

values follows a L
-
shaped curve. A few markers exhibit unusually high
G
ST

22

values. As a result the overall
G
ST

value for AFLP markers is higher than for microsatellites,
23



19

most probably because there is a
higher likelihood that some loci are linked to an adaptive trait
1

than for microsatellites.

2

There is an important discrepancy between the
G
ST

values obtained from the two methods of
3

genotypic analyses for the AFLP data. Differentiation is much lower when th
e analysis is
4

restricted to polymorphic markers. This result was also demonstrated by Isabel et al. (1999)
5

who used RAPD markers and several differentiation parameters. Again, unexpected effects
6

can be induced by restricting the analysis to polymorphic mar
kers. For example, an AFLP
7

fragment that is present and fixed in population A, but absent from population B and fixed in
8

population B, would be excluded by this method of analysis. However, this fragment would
9

have a
G
ST

value of 1.

10

Other comparative studi
es of different marker systems conducted in oaks provide more
11

congruent results. For example, in a genetic study on 21 populations of
Q. petraea
, Le Corre
12

et al. (1997) compared the level of differentiation between 31 RAPD markers and 8 allozyme
13

loci and f
ound that the levels were similar (2.7% for allozymes and 2.4% for RAPDs) and
14

comparable with the results we obtained here with microsatellites (2.3%, Table 6b). The
15

congruence between these results can again be interpreted by the sampling effect within th
e
16

genome. The low number of loci that are sampled by each of this method results in preferential
17

selection of loci that are neutral and located within the tail of the L
-
shaped curve of
G
ST

18

values.

19

Comparison of levels of genetic diversity and differentiati
on among species

20

The results obtained here using microsatellites and AFLP markers confirm earlier studies
21

based on other markers, characters and
Quercus

populations. Gene diversity surveys based
22

on isozymes (Müller
-
Starck et al. 1993, Kremer et al. 1991, Z
anetto et al. 1994) indicated
23



20

that heterozygosity values are higher in
Q. petraea

than in
Q. robur
. Data from DNA analyses
1

conducted on both species (Moreau et al. 1994, Bodénès et al. 1997a) also reached similar
2

conclusions. These differences in diversity

may be related to the social status of the two
3

species. Stands of
Q.

petraea

are usually pure and of larger size than those of
Q.

robur

that
4

are more commonly found inter
-
mixed with other species. Furthermore the co
-
evolution of the
5

two species, the so ca
lled “regeneration” of
Q.

petraea

from successive unidirectional
6

hybridisation with
Q. robur

(Petit et al. 1998) can also be considered as a mechanism
7

contributing to the enrichment of genetic diversity within
Q.

petraea
. This enrichment results
8

from the a
dditional diversity arising following hybridisation with
Q. robur

and augments the
9

diversity already existing in
Q. petraea per se
. Lastly, results from mating system studies
10

(Bacilieri et al. 1996) have shown that the outcrossing rate is higher in
Q. petr
aea

than in
Q.
11

robur
. The difference in outcrossing rates is also likely to contribute as well to the observed
12

difference in fixation index (
F
IS
).

13

Microsatellites and AFLPs show a slightly higher level of genetic differentiation among
14

populations of
Q. pet
raea

and
Q. robur

than between the two species. In addition, a higher
15

genetic differentiation is observed among
Q. petraea

populations than among
Q. robur

16

populations but results are not significant. This observation was also true when allozyme
17

markers wer
e used in earlier studies (Zanetto et al. 1994).

18

Comparison of levels of genetic diversity among populations of the same species

19

Despite the fact that microsatellites and AFLPs provide congruent results, although not
20

significant, in levels of diversity amo
ng the two species, they do not agree in levels of intra
-
21

population genetic diversity within each species. As shown by the correlation matrix of
22

diversity statistics (Table 8), there is a positive trend among
H

values, especially in
Q. robur
,
23



21

but the corre
lation is never significant
. This lack of congruence may be due to a contribution of
1

different factors. Firstly, the level of diversity may be of similar magnitude in the different
2

populations, as indicated in Table 7. Oaks live in large populations and ex
hibit high migration
3

rates (Streiff et al. 1999). As a result, seed and pollen flow may contribute to the high
4

homogeneity of diversity between populations. Second, the diversity statistics are estimated
5

with an important sampling variance (Table 4). Again

a larger number of loci would be
6

necessary to increase the power of the statistical test to compare the level of diversity among
7

populations.

8


9

Overall, microsatellite and AFLP markers analysed in this study confirm earlier results
10

obtained for
Q
.
petraea

and
Q.

robur
. A higher level of genetic diversity was found within
Q.

11

petraea

species and population in each stand. In addition,
Q.

petraea

exhibited a higher
12

genetic differentiation than did
Q. robur
. However, the high inter
-
locus variance for AFLP
13

marker
s did not allow us to significantly distinguish populations. It also appeared from our
14

analysis that a restriction of loci analysed, as recommended by Lynch & Milligan (1994), leads
15

to different rankings of populations. Even if the evolutionary forces are
the same for the two
16

types of markers, the lack of information for AFLP markers and the limited number of
17

microsatellite loci that were analysed could explain the absence of significant positive
18

correlation.

19


20

Acknowledgements

21

This project was supported by
the European Union (FAIR1 PL95
-
0297) and by a grant by
22

BRG (Bureau des Ressources Génétiques). Ian Forrest was responsible for the morphological
23



22

assessment of the samples from Dalkeith Old Wood. Jan Bovenschen was in charge of
1

analysing samples from De Mei
nweg stand.

2


3

References

4

B
ACILIERI
,

R.,

L
ABBÉ
,

T.

and

K
REMER
,

A.
: Intraspecific genetic structure in a mixed
5

population of
Quercus petraea

(Matt.) Liebl. and
Q. robur

L. Heredity
73
: 130
-
141 (1994).

6

B
ACILIERI
,

R.,

D
UCOUSSO
,

A.,

P
ETIT
,

R.J.

and

K
REMER
,

A.
: M
ating system and asymmetric
7

hybridization in a mixed stand of European oaks. Evolution
50
: 900
-
908 (1996).

8

B
ARRENECHE
,

T.,

B
AHRMAN
,

N.

AND
K
REMER
,

A.
: Two dimensional gel electrophoresis
9

confirms the low level of genetic differentiation between
Quercus rob
ur

L. and
Quercus
10

petraea

(Matt.) Liebl. Forest Genetics
3
: 89
-
92 (1996).

11

B
ARUFFI
,

L.,

D
AMIANI
,

G.,

G
UGLIELMINO
,

C.R.,

B
ANDI
,

C.,

M
ALACRIDA
,

A.R.

and
12

G
ASPERI
,

G
.: Polymorphism within and between populations of
Ceratitis capitata
:
13

comparison between RAPD an
d multilocus enzyme electrophoresis data.
Heredity

74
: 425
-
14

437 (1995).

15

B
ODÉNÈS
,

C.,

J
OANDET
,

S.,

L
AIGRET
,

F.

and
K
REMER
,

A.: Detection of genomic regions
16

differentiating two closely related oak species
Quercus

petraea

(Matt.) Liebl. and
Quercus

17

robur

L. He
redity
78
: 433
-
444 (1997a).

18

B
ODÉNÈS
,

C.,

L
ABBÉ
,

T.,

P
RADÈRE
,

S
. and
K
REMER
,

A.: General
vs

local differentiation
19

between two closely related white oak species. Molecular Ecology
6
: 713
-
724 (1997b).

20

B
REYNE
,

P.,

B
OERJAN
,

W.,

G
ERATS
,

T.,

V
AN
M
ONTAGU
,

M
. and
V
AN
G
YSEL
, A.:
21

Applications of AFLP in plant breeding, molecular biology and gene tics. Belgian Journal of
22

Botany
129
: 107
-
117 (1997).

23



23

B
ROWN
,

A.H.D
. and
W
EIR
,

B.S
. Measuring genetic variability in plant populations. In:
1

T
ANKSLEY
,

S.D
. and
O
RTON
,

T.J.

(Eds.)
: Isozymes in Plant Genetics and Breeding, Part A.,
2

Elsevier Science Publishers BV, Amsterdam. pp. 219
-
239 (1983).

3

C
AGIGAS
,

M.E.,

V
AZQUEZ
,

E.,

B
LANCO
,

G.

and
S
ÁNCHEZ
,

J.A.:

Combined assessment of
4

genetic variability in populations of Brown trout (
Salmo tru
tta

L.) based on allozymes,
5

microsatellites and RAPD markers. Marine Biotechnology
1
: 286
-
296 (1999).

6

D
OW
,

B.D.,

A
SHLEY
,

M.V
. and
H
OWE
,

H.F
.: Characterization of highly variable (GA/CT)n
7

microsatellites in the bur oak,
Quercus macrocarpa
. Theoretical and A
pplied Genetics
91
:
8

137
-
141 (1995).

9

D
UCOUSSO
,

A.,

G
UYON
,

J.P
., and
K
REMER
,

A
.: Latitudinal and altitudinal variation of bud
10

burst in western populations of sessile oak [
Quercus petraea

(Matt.) Liebl.]. Annals of Forest
11

Sciences
53
: 775
-
782 (1996).

12

D
UPOUEY
,

J.L
. and
B
ADEAU
,

V
.: Morphological variability of oaks (
Quercus robur

L.,
13

Quercus petraea

(Matt.) Liebl.,
Quercus pubescens
) in northeastern France: preliminary
14

results. Annals of Forest Sciences
50:

35s
-
40s (1993).

15

G
ERBER
,

S.,

M
ARIETTE
,

S.,

S
TREIFF
,

R.,

B
ODÉNÈS
,

C.

and
K
REMER
A:

Comparison of
16

microsatellites and AFLP markers for parentage analysis. Molecular Ecology
9
: 1037
-
1048
17

(2000)
.

18

I
SABEL
,

N.,

B
EAULIEU
,

J.

and
B
OUSQUET
,

J
.: Complete congruence between gene diversity
19

estimates derived from genotypic d
ata at enzyme and random amplified polymorphic DNA
20

loci in black spruce.
Proceedings of the National Academy of Sciences of the USA

92:

6369
-
21

6373 (1995).

22



24

I
SABEL
,

N.,

B
EAULIEU
,

J.,

T
HÉRIAULT
,

P.

and
B
OUSQUET
,

J.:

Direct evidence for biased gene
1

diversity es
timates from dominant random amplified polymorphic DNA (RAPD) fingerprints.
2

Molecular Ecology
8
, 477
-
483 (1999).

3

J
IN
,

L
. and
C
HAKRABORTY
,

R
.: Population structure, stepwise mutations, heterozygote
4

deficiency and their implications in DNA forensics. Heredit
y
74
: 274
-
285 (1995).

5

K
ARP
,

A.,

K
RESOVICH
,

S.,

B
HAT
,

K.V.,

A
YAD
,

W.G.

and

H
ODGKIN
,

T
. Molecular tools in
6

plant genetic resources conservation: a guide to the technologies. IPGRI Technical Bulletin
2
:
7

36
-
42 (1997).

8

K
ENDALL
,

M.G
. and
S
TUART
,

A
. The Advanced
Theory of Statistics 1, 4
th

edition, Griffin,
9

London (1977).

10

K
REMER
,

A.

AND
P
ETIT
,

R.J
. Gene diversity in natural populations of oaks species. Annals of
11

Forest Sciences
50
: 186s
-
202s (1993).

12

K
REMER
,

A.,

P
ETIT
,

R.J.,

Z
ANETTO
,

A.,

F
OUGÈRE
,

V.,

D
UCOUSSO
,

A.,

W
AGNER
,

D.

and
13

C
HAUVIN
C.:

Nuclear and organelle gene diversity in
Quercus robur

and
Q. petraea
. In:
14

M
ÜLLER
-
S
TARCK
,

G.

and

Z
IEHE
,

M
. (Eds.): Genetic Variation in European Populations of
15

Forest Trees, Sauerländer’s Verlag, Frankfurt am Main. pp 141
-
166 (199
1).

16

L
ANNÉR
-
H
ERRERA
,

C.,

G
USTAFSSON
,

M.,

F
ÄLT
,

A.S.

and

B
RYNGELSSON
,

T.:

Diversity in
17

natural populations of wild
Brassica oleracea

as estimated by isozyme and RAPD analysis.
18

Genetic Resources and Crop Evolution

43:

13
-
23 (1996).

19

L
E
C
ORRE
,

V.,

D
UMOLIN
-
L
APÈG
UE
,

S
. and
K
REMER
A.:

Genetic variation at allozyme and
20

RAPD loci in sessile oak
Quercus petraea

(Matt.) Liebl.: the role of history and geography.
21

Molecular Ecology
6
, 519
-
529 (1997).

22



25

L
YNCH
,

M
. and
M
ILLIGAN
,

B.G
.: Analysis of population genetic structure
with RAPD
1

markers. Molecular Ecology
3
, 91
-
99 (1994).

2

M
OREAU
,

F.,

K
LEINSCHMIT
,

J
. and
K
REMER
,

A
. Molecular differentiation between
Q.
3

petraea

and
Q. robur

assessed by random amplified DNA fragments. Forest Genetics
1
: 51
-
4

64 (1994).

5

M
ÜLLER
-
S
TARCK
,

G.,

H
ERZO
G
,

S
. and
H
ATTEMER
,

H.H
.: Intra
-

and interpopulation genetic
6

variation in juvenile populations of
Quercus robur

L. and
Quercus petraea

Liebl. Annals of
7

Forest Sciences
50
: 233s
-
244s (1993).

8

M
ÜLLER
-
S
TARCK
,

G
. and
Z
IEHE
,

M
.: Genetic variation in populations
of
Fagus sylvatica

L.,
9

Quercus robur

L. and
Q. petraea
.
Liebl. in Germany. In:
M
ÜLLER
-
S
TARCK
,

G.

and

Z
IEHE
,

10

M
. (Eds.): Genetic Variation in European Populations of Forest Trees, Sauerländer’s Verlag,
11

Frankfurt am Main. pp 125
-
140 (1991).

12

N
EI
M.

Molecular E
volutionary Genetics. Columbia University Press, New
-
York. (1987).

13

P
ETIT
,

R.J
. and
P
ONS
,

O
.: Bootstrap variance of diversity and differentiation estimators in a
14

subdivided population. Heredity
80
: 56
-
61 (1998).

15

P
ETIT
,

R.J.,

P
INEAU
,

B.,

D
ÉMESURE
,

B.,

B
ACILI
ERI
,

R.,

D
UCOUSSO
,

A.

and
K
REMER
A
:
16

Chloroplast DNA footprints of postglacial recolonization by oaks.
Proceedings of the National
17

Academy of Sciences of the USA

94
: 9996
-
10001 (1998).

18

R
OUSSET
,

F
. Equilibrium values of measures of population subdivision for

stepwise mutation
19

processes. Genetics
14:

1357
-
1362 (1996).

20

S
LATKIN
,

M
. A measure of population subdivision based on microsatellite allele frequencies.
21

Genetics
139
: 457
-
462 (1995).

22

S
OKAL
,

R
. and
R
OHLF
,

F.J
. Biometry. Freeman WH and Company, New
-
York (199
5).

23



26

S
TEINKELLNER
,

H.,

F
LUCH
,

S.,

T
URETSCHEK
,

E.,

L
EXER
,

C.,

S
TREIFF
,

R.,

K
REMER
,

A.,

B
URG
,

1

K
. and
G
LOESSL
,

J
. Identification and characterization of (GA/CT)n microsatellite loci from
2

Quercus petraea
. Plant Molecular Biology
33:

1093
-
1096 (1997).

3

S
TREIFF
,

R
.,

L
ABBÉ
,

T.,

B
ACILIERI
,

R.,

S
TEINKELLNER
,

H.,

G
LÖSSL
,

J
. and
K
REMER
A:

4

Within
-
population genetic structure in
Quercus robur

L. and
Quercus petraea

(Matt.) Liebl.
5

assessed with isozymes and microsatellites. Molecular Ecology
7:

317
-
328 (1998).

6

S
TREIFF
,

R.,

D
UCOUSSO
,

A.,

L
EXER
,

C.,

S
TEINKELLNER
,

H.,

G
LÖSSL
,

J.

and

K
REMER
,

A.:

7

Pollen dispersal inferred from paternity analysis in a mixed oak stand of
Quercus robur

L. and
8

Quercus petraea

(Matt.)
Liebl.
Molecular Ecology
8
: 831
-
841 (1999).

9

Z
ANETTO
,

A.,

R
OUSSEL
,

G.

and
K
REMER
,

A.:

Geographic variation of inter
-
specific
10

differentiation between
Quercus robur

L. and
Quercus petraea

(Matt.) Liebl. Forest
11

Genetics
1
: 111
-
123 (1994).

12

13



27

FIGURE LEGENDS

1

Figure 1
:
Location of stands.

2

Figure 2
: Distribution curve of
G
ST

per l
ocus between
Q
.
petraea

and
Q
.
robur

populations

3

Figure 3
: Distribution curve of
G
ST

per locus among
Q
.
petraea

populations

4

Figure 3
: Distribution curve of
G
ST

per locus among
Q
.
robur

populations

5

In each figure, loci were numbered in order of their
G
ST

va
lues. The dotted line separated the
6

loci showing a differentiation superior to 10% from the other loci.

7

8



28

Figure 1

1

2



29

Figure 2

1

2



30

Figure 3

1

2



31

Figure 4

1

2



32

Table 1
: List of the stands and their composition

1

Country

Name of the location

Number of trees
Q
.
petraea

N
umber of trees
Q
.
robur

Number of trees of
intermediate morphology

France

Petite Charnie

199

215

8

Germany

Escherode

110

206

5

United Kingdom

Dalkeith Old Wood

21

351

27

The Netherlands

Meinweg

181

184

15

Austria

Sigmundsherberg

228

159

8

United King
dom

Roudsea Wood

205

56

11

Spain

Salinasco Mendia

233

45

-

2



33

Table 2
: Polymorphism of AFLP markers

1

Primer
-
Enzyme Combination

P
st
I+CAG

M
se
I+CAA

P
st
I+CAG

M
se
I+GCA

P
st
I+CAG

M
se
I+GGA

P
st
I+CCA

M
se
I+CAA

4 PECs

Number of analysed loci

48

34

30

43

155

Number of

polymorphic loci*

33

19

23

33

108

% of polymorphic loci

69

59

77

77

70

A locus is polymorphic as soon as the two phenotypes (presence and absence of the
2

fragment) existed in at least one population.

3

4



34

Table 3
: Microsatellites diversity statistics in
Q
.
p
etraea

and
Q
.
robur

populations

1

Petite Charnie

Q. petraea

sd 1

Q. robur

sd 1

p

H
i

0,877

0,004

0,866

0,004

0,036

H
O

0,819

0,014

0,804

0,013

0,209

A

18,67

0,39

18.00

0,44

0,077

A
E

8,14

0,26

7,46

0,22

0,046

F
IS

0,063

0,016

0,068

0,015

0,616

Escherode

Q.

petraea

sd 1

Q. robur

sd 1

p

H
i

0,878

0,005

0,835

0,004

0,001

H
O

0,884

0,011

0,832

0,010

0,001

A

19,50

0,41

18.00

0,39

0,024

A
E

8,23

0,29

6,04

0,16

0,001

F
IS

-
0,011

0,013

0.000

0,012

0,671

Dalkeith Old Wood

Q. petraea

sd 1

Q. robur

sd 1

p

H
i

0,869

NC

0,868

0,002

NC

H
O

0,881

NC

0,815

0,009

NC

A

12,67

NC

19,50

0,36

NC

A
E

7,63

NC

7,57

0,12

NC

F
IS

-
0,038

NC

0,058

0,010

NC

Meinweg

Q. petraea

sd 1

Q. robur

sd 1

p

H
i

0,867

0,003

0,860

0,004

0,079

H
O

0,790

0,011

0,748

0,014

0,007

A

18,17

0,37

17,83

0,42

0,278

A
E

7,50

0,18

7,12

0,20

0,108

F
IS

0,086

0,012

0,128

0,015

0,016

Sigmundsherberg

Q. petraea

sd 1

Q. robur

sd 1

p

H
i

0,883

0,004

0,882

0,004

0,306

H
O

0,819

0,010

0,757

0,014

0,001

A

26,67

0,44

24,67

0,51

0,005

A
E

8,54

0,17

8,45

0,14

0,351

F
IS

0,071

0,012

0,142

0,017

0,000

Roudsea Wood

Q. petraea

sd 1

Q. robur

sd 1

p

H
i

0,908

0,002

0,899

0,005

0,003

H
O

0,781

0,013

0,775

0,022

0,434

A

26,50

0,45

19,50

0,56

0,004

A
E

10,85

0,24

9,90

0,43

0,003

F
IS

0,136

0,014

0,131

0,026

0,501

Salinasco M
endia

Q. petraea

sd 1

Q. robur

sd 1

p

H
i

0,862

0,004

0,866

0,007

0,354

H
O

0,840

0,009

0,812

0,020

0,107

A

19,33

0,34

14,50

0,55

0,003

A
E

7,22

0,20

7,44

0,33

0,349

F
IS

0,023

0,010

0,052

0,023

0,928



35

sd 1

is the standard deviation associated to the intr
a
-
locus variance;

p

values are the
1

associated probabilities; significant values at 5% are in bold numbers; NC: not calculated,
2

because bootstrap mean values differed markedly from the observed values, indicating that the
3

bootstrap procedure is not adequate

here since the sample size is low.

4

Table 4
: AFLPs diversity statistics in
Q
.
petraea

and
Q
.
robur

populations

5

Petite Charnie

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.194

0.006

0.018

0.019

0.179

0.005

0.018

0.019

0.575

H
i
(G1)

0
.191

0.006

0.018

0.019

0.172

0.004

0.018

0.018

0.465

H
i
(G2)

0.234

0.012

0.022

0.025

0.252

0.010

0.022

0.024

0.610

Escherode

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.207

0.006

0.016

0.017

0.192

0.005

0.015

0.016

0.522

H
i
(G1)

0
.189

0.005

0.016

0.017

0.185

0.006

0.015

0.016

0.865

H
i
(G2)

0.211

0.007

0.017

0.018

0.233

0.008

0.020

0.022

0.441

Dalkeith Old Wood

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.214

NC

NC

NC

0.189

0.004

0.015

0.016

NC

H
i
(G1)

0.204

NC

NC

NC

0.180

0.005

0.014

0.015

NC

H
i
(G2)

0.234

NC

NC

NC

0.236

0.006

0.017

0.018

NC

Meinweg

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.200

0.006

0.016

0.017

0.186

0.005

0.015

0.016

0.549

H
i
(G1)

0.194

0.008

0.015

0.017

0.189

0
.005

0.015

0.016

0.834

H
i
(G2)

0.247

0.008

0.019

0.021

0.261

0.008

0.020

0.022

0.646

Sigmundsherberg

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.189

0.006

0.016

0.017

0.196

0.006

0.015

0.016

0.764

H
i
(G1)

0.189

0.006

0.016

0.017

0
.199

0.008

0.015

0.017

0.682

H
i
(G2)

0.262

0.010

0.020

0.022

0.262

0.008

0.020

0.022

1.000

Roudsea Wood

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.214

0.007

0.016

0.017

0.227

0.005

0.015

0.016

0.582

H
i
(G1)

0.202

0.007

0.015

0.01
7

0.217

0.005

0.016

0.017

0.535

H
i
(G2)

0.241

0.010

0.018

0.021

0.279

0.010

0.018

0.021

0.184

Salinasco Mendia

Q. petraea

sd 1

sd 2

total sd

Q. robur

sd 1

sd 2

total sd

p

H
i
(P)

0.195

0.005

0.016

0.017

0.194

0.006

0.015

0.016

0.968

H
i
(G1)

0.197

0.006

0.0
15

0.016

0.189

0.005

0.015

0.016

0.734

H
i
(G2)

0.264

0.008

0.019

0.021

0.260

0.008

0.018

0.020

0.881

H
i
(P)

is the phenotypic diversity;
H
i
(G1)

is the G1 gene diversity;
H
i
(G2
)

is the G2 gene diversity;
6

sd 1

is the standard deviation associated to the intr
a
-
locus variance;
sd 2

is the standard
7

deviation associated to the inter
-
locus variance;
total sd

is the total standard deviation
8

associated to the total variance; p is the associated probability; NC: not calculated, because
9

bootstrap mean values differed
markedly from the observed values, indicating that the
10

bootstrap procedure is not adequate here since the sample size is low.

11



36

1



37

Table 5
: Microsatellites and AFLPs diversity statistics at the species level

1


Quercus petraea

Quercus robur


Microsatellites



sd 1


sd 1

p
1

H
i

0.896

0,001

0.878

0,002

0,000

H
O

0.820

0,005

0.799

0,005

0,000

A

28.83

0,461

26.83

0,492

0,000

A
E

9.59

0,113

8.20

0,088

0,000

F
IS

0.084

0,005

0.089

0,006

0,249

AFLPs



sd 1

sd 2

total sd


sd 1

sd 2

total sd

p
1

p
2

p

H
i
(P)

0.228

0.00
2

0.015

0.015

0.220

0.002

0.014

0.014

0.001*

0.370

0.697

H
i
(G1)

0.225

0.002

0.015

0.015

0.219

0.002

0.014

0.014

0.008*

0.385

0.772

H
i
(G2)

0.252

0.004

0.015

0.016

0.256

0.003

0.016

0.016

0.202

0.380

0.857

H
i
(P)

is the phenotypic diversity;
H
i
(G1)

is the
G1 gene diversity;
H
i
(G2)

is the G2 gene diversity;
2

sd 1

is the standard deviation associated to the intra
-
locus variance and
p
1

the associated
3

probability;
sd 2

is the standard deviation associated to the inter
-
locus variance and
p
2

the
4

associated probabi
lity;
total sd

is the total standard deviation associated to the total variance
5

and
p

is the associated probability;
* indicates a significant difference between
Q. petraea

and
6

Q. robur
.

7

8



38

Table 6a
: Genetic differentiation between species

1


G
ST

(microsatelli
tes)

G
ST
(P)

(AFLPs)

G
ST
(G1)

(AFLPs)

G
ST
(G2)

(AFLPs)

Q
.
petraea /

Q
.
robur

0.013
2

0.037

0.030

0.016

Q
.
petraea

/
Q
.
robur

Petite Charnie

0.018
1

0.068

0.053

0.038

Q
.
petraea

/
Q
.
robur

Escherode

0.019
1

0.093

0.096

0.028

Q
.
petraea

/
Q
.
robur

Dalkeith Old

Wood

0.010
1

0.031

0.031

-
0.0003

Q
.
petraea

/
Q
.
robur

Meinweg

0.018
1

0.076

0.071

0.038

Q
.
petraea

/
Q
.
robur

Sigmundsherberg

0.024
1

0.060

0.056

0.023

Q
.
petraea

/
Q
.
robur

Roudsea Wood

0.005
1

0.063

0.053

0.034

Q
.
petraea

/
Q
.
robur

Salinasco Mendia

0.
015
1

0.051

0.040

0.021


2

Table 6b
: Genetic differentiation among populations within species

3


G
ST

(microsatellites)

G
ST
(P)

(AFLPs)

G
ST
(G1)

(AFLPs)

G
ST
(G2)

(AFLPs)

Q
.
petraea

populations

0.023
2

0.118

0.111

0.044

sd

0.007
2

0.016

0.015

0.009

Q
.
robur


popul
ations

0.020
2

0.114

0.111

0.030

sd

0.005
2

0.022

0.019

0.005

p
value

0.364
2

0.482

0.592

0.126


4

In Tables 6a and 6b,

p

values were obtained with bootstrap samples;
1

1
st

analysis of
5

microsatellites;
2

2
nd

analysis of microsatellites;
G
ST
(P
)

is the phenoty
pic differentiation;
G
ST
(G1)

6

is the G1 genetic differentiation;
G
ST
(G2)

is the G2 genetic differentiation.

7


8



39

Table 7
: Comparisons of levels of diversity across stands

1

Microsatellites (
H
i
)

AFLPs (
H
i
(P)
)

AFLPs (
H
i
(G1)
)

AFLPs (
H
i
(G2)
)

Quercus petraea

Roudse
a Wood

0.905

A





Dalkeith Old Wood

0.214

NC

Dalkeith Old Wood

0,204

NC

Salinasco Mendia

0,264

A

Escherode

0.881


B




Roudsea Wood

0.214

A

Roudsea Wood

0,202

A

Sigmundsherberg

0,262

A

Petite Charnie

0.877


B

C



Escherode

0.207

A

Salinasco Mendia

0,197

A

Meinweg

0,247

A

Meinweg

0.868



C



Meinweg

0.200

A

Meinweg

0,194

A

Roudsea Wood

0,241

A

Dalkeith Old Wood

0.866




NC


Salinasco Mendia

0.195

A

Petite Charnie

0,191

A

Petite Charnie

0,234

A

Salinasco Mendia

0.859




D


Petite Charnie

0.194

A

Eschero
de

0,189

A

Dalkeith Old Wood

0,234

NC

Sigmundsherberg

0.853




D


Sigmundsherberg

0.189

A

Sigmundsherberg

0,189

A

Escherode

0,211

A

Quercus robur

Roudsea Wood

0.898

A





Roudsea Wood

0.229

A

Roudsea Wood

0,217

A

Roudsea Wood

0,279

A

Dalkeith Old Wood

0.867


B




Sigmundsherberg

0.196

A

Sigmundsherberg

0,199

A

Sigmundsherberg

0,262

A

Salinasco Mendia

0.867


B

C



Salinasco Mendia

0.194

A

Meinweg

0,189

A

Meinweg

0,261

A

Petite Charnie

0.866


B

C



Escherode

0.192

A

Salinasco Mendia

0,189

A

Salinasco Me
ndia

0,260

A

Meinweg

0.862



C



Dalkeith Old Wood

0.189

A

Escherode

0,185

A

Petite Charnie

0,252

A

Escherode

0.851




D


Meinweg

0.186

A

Dalkeith Old Wood

0,180

A

Dalkeith Old Wood

0,236

A

Sigmundsherberg

0.832





E

Petite Charnie

0.179

A

Petite Charn
ie

0,172

A

Escherode

0,233

A

H
i
(P)

is the phenotypic diversity;
H
i
(G1)

is the G1 gene diversity;
H
i
(G2)

is the G2 gene diversity.
For microsatellites, rankings were performed with the intra
-
2

locus standard deviation. For AFLPs, rankings were based on the t
otal standard deviation. Populations having the same letter do not show a significant
3

difference in their level of diversity.

4


5



40

Table 8:
Spearman’s rank correlation analysis among diversity values obtained with different
1

markers

2

Quercus

petraea

H
i

(microsa
tellites)

H
i
(P)

H
i
(G1)

A

(microsatellites)

0.414
2



H
i
(P)

0.577



H
i
(G1)

0.090

0.673


H
i
(G2)

-
0.649

-
0.464

0.091

Quercus

robur

H
i

(microsatellites)

H
i
(P)

H
i
(G1)

A

(microsatellites)

0.429
2



H
i
(P)

0.180



H
i
(G1)

0.082

0.847*


H
i
(G2)

0.180

0.571

0.8
47*

Each Spearman’s rank correlation is followed by the associated probability; * indicates a significant
3

positive correlation at the 5% level;
H
i
: microsatellites within
-
population diversity;
2

2
nd

analysis of
4

microsatellites;

H
i
(P)

is the phenotypic div
ersity;
H
i
(G1)

is the G1 gene diversity;
H
i
(G2)

is the G2 gene
5

diversity.

6