Biological Journal of the Linnean Society (2000), 69: 411–430. With 4 figures
doi:10:1006/bijl.1999.0389, available online at http://www.idealibrary.com on
Genetic variation in three species of Epipactis
(Orchidaceae): geographic scale and evolutionary
inferences
BODIL K. EHLERS∗
Department of Genetics and Ecology, Ny Munkegade building 540, University of Aarhus,
DK-8000 Aarhus C, Denmark
HENRIK Æ. PEDERSEN
The Botanical Museum, University of Copenhagen, Gothersgade 130,
DK-1123 Copenhagen K, Denmark
Received 19 November 1998; accepted for publication 6 July 1999
The breeding system is expected to strongly influence the genetic structure of plant populations.
In the present study, isozyme variation is documented in Danish populations of three species
of Epipactis, varying in breeding system from allogamy to obligate autogamy. The allogamous
and widespread E. helleborine subsp. helleborine shows high levels of polymorphism. Most of
the genetic variation is found within local populations. A hierarchical analysis indicates
significant among-population differentiation, but no regional differentiation in E. helleborine
is apparent. This may be due to higher levels of gene flow in the past, before forest was
fragmented. The ecotype from coastal dunes, E. helleborine subsp. neerlandica, does not differ
from E. helleborine subsp. helleborine in any of the examined loci, but it has a significant population
inbreeding coefficient that can probably be explained by higher levels of geitonogamy and
the possibility of spontaneous autogamy. The entomophilous E. purpurata and the obligately
autogamous E. phyllanthes are monomorphic at all loci examined. Several factors, including
a founder effect at the time of colonization, high levels of geitonogamy, as well as habitat
specialization combined with erratic flowering may have contributed to the lack of variation
in E. purpurata. The lack of variation in the autogamous E. phyllanthes is probably due to
inbreeding. It is proposed that autogamy in Epipactis may in some cases have evolved through
paedomorphosis of allogamous flowers and that the occurrence of local breeding groups may
have facilitated the speciation process.
 2000 The Linnean Society of London
ADDITIONAL KEY WORDS:—breeding system – isozymes – population genetic structure
– congeners – orchids – Epipactis.
CONTENTS
Introduction . . . . . . . . . . . . . . . . . . . . . . .
412
∗ Corresponding author. Email: [email protected]
0024–4066/00/030411+20 $35.00/0
411
 2000 The Linnean Society of London
412
B. K. EHLERS & H. Æ. PEDERSEN
Material and methods
. . . . . .
Study species . . . . . . . .
Sampling . . . . . . . . .
Electrophoresis . . . . . . .
Data analysis . . . . . . . .
Results . . . . . . . . . . .
Discussion . . . . . . . . . .
Variation within species . . . .
Variation between species . . . .
Scenario: inferences on the evolution
Epipactis. . . . . . . . . .
Acknowledgements
. . . . . . .
References . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
towards autogamy
. . . . . .
. . . . . .
. . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
and speciation
. . . . .
. . . . .
. . . . .
.
.
.
.
.
.
.
.
.
in
.
.
.
.
.
.
.
.
.
.
.
.
413
413
413
413
416
416
417
417
423
.
.
.
424
425
425
INTRODUCTION
Plants are never uniformly distributed in the landscape, but occur as assemblages
of individuals. The clustering of individuals into local populations, the sizes of which
vary in time and space, influences the genetic structure of a species. Genetic
differences between local populations will evolve over time when there is little or
no gene flow between them (Wright, 1943, 1946).
Differentiation can occur on several levels. Local populations may be grouped
within regions and regions within some larger geographic area (Chakraborty &
Leimar, 1987). It is thus relevant to sample information on the genetic structure at
various geographic scales to understand at what distances differentiation takes place
in a given species.
Loveless & Hamrick (1984), Hamrick & Godt (1990) and Hamrick et al. (1991)
listed various ecological factors influencing the structuring of genetic variation. They
found that plant breeding system was a principal factor, but also that characteristics
such as life form, seed dispersal, and successional status could be important determinants of genetic structure.
The breeding system influences the genetic structure of a species by means of its
mode of gene flow. On the other hand, the occurrence (or lack) of gene flow may
influence the evolution of plant breeding systems. Karron (1991), among others,
suggested that evolution of autogamy (self-fertilization) from allogamy (outcrossing)
may take place in self-compatible species when pollinating agents are scarce or
unreliable, or when available mates are few. If plant species are adapted to live in
environments with little chance of cross pollination, the advantages of autogamy
(e.g. assurance of seed production) may amply offset its disadvantages (e.g. loss of
heterozygosity). It is believed that living autogamous species have, generally, evolved
from allogamous ancestors (Richards, 1986), and it is recognized that the factors
influencing evolution towards autogamy are many-fold (Uyenoyama, Holsinger &
Waller, 1993). The transition from allogamy to autogamy is well documented (e.g.
Wyatt, 1988) and has taken place independently in numerous taxa.
In the present study of Epipactis Sw. (Orchidaceae) we attempt to determine the
main geographic level at which intraspecific reproductive units occur. Furthermore,
we examine the genetic variation between species in order to make inferences on
the effect of the breeding system and the geographic distribution on the genetic
structure. Finally, we discuss the evolution of autogamy in Epipactis. The entomophilous E. helleborine (L.) Crantz subsp. helleborine is studied in detail, while data
GENETIC VARIATION IN EPIPACTIS
413
on the entomophilous E. purpurata Sm., the facultatively autogamous E. helleborine
subsp. neerlandica (Verm.) Buttler, and the obligately autogamous E. phyllanthes G. E.
Sm. are provided mainly for interspecific comparisons.
MATERIAL AND METHODS
Study species
The orchid genus Epipactis consists of long-lived, polycarpic, clonal plants and
belongs to the subtribe Limodorinae (Epidendroideae: Neottieae; cf. Dressler, 1993).
As in other members of this subtribe, the anther is dorsal, titled slightly downwards,
and extending beyond the rostellum (the modified non-receptive portion of the
median stigma lobe, cf. Dressler, 1989). The pollinia are two in number, soft to
mealy, and lack caudicles (pollinium stalks).
In obligately entomophilous members of Epipactis (e.g. E. helleborine subsp. helleborine,
E. purpurata) the rostellum is well developed. The glue at the top of the rostellum is
covered by a membrane (Schick, 1989) and should be referred to as a ‘diffuse’
viscidium, as the pollinia are not attached to the glue before the intervention of a
pollinator (Rasmussen, 1982). In obligately autogamous species (e.g. E. phyllanthes)
the rostellum is rudimentary or even lacking (see, e.g. Claessens & Kleynen, 1991;
van der Cingel, 1995).
The geographical ranges of the taxa studied are indicated in Figure 1. It should
be noted that the populations of E. phyllanthes in Sweden, Denmark, and Germany
have by many authors been treated as a separate species, E. confusa D. P. Young.
However, the latter is evidently conspecific with E. phyllanthes (Lundqvist, 1966;
Faurholdt, Pedersen & Christiansen, 1998).
Sampling
Flower buds from 12 Danish populations of E. helleborine subsp. helleborine, 1 of E.
helleborine subsp. neerlandica, 5 of E. purpurata, and 6 of E. phyllanthes were collected in
late July and early August 1997. The populations were numbered as their site of
origin (Table 1, Fig. 2). Samples were kept cool in an insulated bag until they were
stored at −70°C in the laboratory. Thirty to forty individuals were sampled from
each population except from those small populations in which this number could
not be reached. In such cases flower buds were sampled from all individuals found.
One flower bud was collected from each individual, and care was taken not to
sample from more than one shoot from each vegetatively coherent clone.
Electrophoresis
Flower buds, while kept on ice, were ground in 100 ll buffer at pH 7.5 containing
polyvinylpolypyrrolidone (PVP) and mercaptoethanol (formula available from the
laboratory at Department of Genetics & Ecology, AAU). The material was screened
for a total of 14 enzymes: aconitase hydratase (ACO, E.C. 4.2.1.3), aspartate
414
B. K. EHLERS & H. Æ. PEDERSEN
Figure 1. The geographic ranges of Epipactis helleborine subsp. helleborine (——), E. helleborine subsp.
neerlandica (diagonal hatching), E. phyllanthes (– – –), and E. purpurata (. . .). E. helleborine subsp. helleborine
is distributed eastwards through Central Asia to Japan. Additionally, it is naturalized in the northeastern parts of the USA and adjacent parts of Canada (e.g. Soper & Garay, 1954; Luer, 1975).
T 1. List of study sites. Abbreviations: hel, Epipactis helleborine subsp. helleborine; nee, E. helleborine
subsp. neerlandica; phy, E. phyllanthes; pur, E. purpurata
Number
Site
taxa
1
2
3
4a
4b
5
6
7a
7b
7c
8
9
10
11
12
13
14
Hjørring Kommunes Klitplantage
Mariager Vesterskov
Haslund Skov
Borum Møllebæk (N)
Borum Møllebæk (S)
Storring Præstegårdsskov
Århus Å
Århusskovene (N)
Århusskovene (C)
Århusskovene (S)
Hørret Skov
Stenhus Vænge
Boserup Skov
Dalby Skov
Skovhus Vænge
Folehave
Vintersbølle Skov
nee
phy
hel, pur
hel
pur
hel, pur
phy
hel, pur, phy
hel, phy
hel, phy
hel
pur
hel
hel
hel, phy
hel
hel
GENETIC VARIATION IN EPIPACTIS
415
N
DENMARK
1
50 km
2
3
6
4
Jutland
5
8
7
Zealand
9
10
11
Funen
12
13
14
Figure 2. Geographic situation of study sites 1–14 (numbers according to Table 1).
aminotransferase (AAT, E.C. 2.6.1.1), esterase (EST, E.C. 3.1.1.-), glucose-6-phosphate isomerase (GPI, E.C. 5.3.1.9), glucose-6-phosphate dehydrogenase (G6PDH,
E.C. 1.1.1.49), hexokinase (HEX, E.C. 2.7.1.1), isocitrate dehydrogenase (IDH, E.C.
1.1.1.42), lactate dehydrogenase (LDH, E.C. 1.1.1.27), malate dehydrogenase (MDH,
E.C. 1.1.1.37), malic enzyme (ME, E.C. 1.1.1.40), phosphogluconate dehydrogenase
(PGD, E.C. 1.1.1.44), phosphoglucomutase (PGM, E.C. 5.4.2.2), shikimate dehydrogenase (SKD, E.C. 1.1.1.25) and triosephosphate isomerase (TPI, E.C. 5.3.1.1).
Seven enzyme systems representing nine loci revealed a clear banding pattern for
all species, and the polymorphic loci showed banding patterns in accordance with
the quarternary structure expected. Idh-1, Idh-2, Gpi, Mdh-1, and Skd were resolved
416
B. K. EHLERS & H. Æ. PEDERSEN
on a 12% morpholine citrate gel at pH 6.1 (Clayton & Tretiak 1972), while Mdh2, Pgd, Pgm and Tpi were resolved on a 12% discontinuous citrate/histidine buffer
(Fildes & Harris, 1966) at pH 7.1. Different species were run on the same gel to
verify any differences in banding patterns. Both gel types were run for 5.0 h at
600 V, 80 mA, and 50 W. Gels were cut into slices 2 mm thick and stained according
to Wendel & Weeden (1989) with some modifications. To improve the resolution,
10 mg histidine was added to the staining solution of PGM, and twice the amount
of reactant was added to the staining solution of SKD and IDH (100 mg shikimic
acid and 200 mg isocitric acid, respectively).
Data analysis
The programme GENEPOP version 3.1 (Raymond & Rousset 1995) was used
to determine allele frequencies, number of alleles, and observed heterozygosity.
Estimates of gene diversity (He) were calculated as 1−Rp2i for each locus, and
averaged across loci in each population. Population inbreeding coefficients (Fis) were
calculated according to Weir & Cockerham (1984) by using the programme FSTAT
version 1.2 (Goudet, 1995) and their degree of significance was examined after
sequential Bonferoni corrections were applied (Rice, 1989). Fis estimates the inbreeding in individuals relative to the subpopulation to which they belong, and Fis=
0 is equivalent to the assumption of random mating within the subpopulation (Hartl
& Clark, 1989).
Nei’s genetic distance (Nei, 1972) was calculated from allele frequencies of all
study populations. This resulted in a distance matrix which was used to construct
a tree describing the genetic relationship among populations. Nei’s genetic distance
was used, as it assumes that gene frequencies change not only by genetic drift but
also by mutation, which we found most likely when comparing different species.
The resulting tree was constructed using the UPGMA (unweighted pair-group
arithmetic average) method of clustering (Legendre & Legendre, 1983). This was
done using the programmes GENDIST and NEIGHBOR from the PHYLIP package
(Felsenstein, 1993).
In order to determine the amount of variance attributable to geographic substructure, a hierarchical Fst analysis was performed using the programme Arlequin
version 1.1 (Schneider et al., 1997). This programme uses the AMOVA (Analysis of
Molecular Variance) method to partition the variance components into different
hierarchical levels, and significance of the variance components are tested using a
non-parametric permutation approach as described by Excoffier, Smouse & Quattro
(1992).
RESULTS
None of the examined loci from E. purpurata and E. phyllanthes show any intraspecific
variation, but the two species are homozygous for different alleles (Appendix 1).
Only 44% of the alleles found in this study are shared between the two species,
whereas they share 89% and 78% of their alleles with E. helleborine, respectively.
Two of the alleles found in E. phyllanthes, viz. Pgm1 and Mdh-12, are not found in
GENETIC VARIATION IN EPIPACTIS
417
any of the other taxa examined. Similarly, E. purpurata has one unique allele at the
Idh-2 locus. All other alleles found in E. phyllanthes and E. purpurata are also found
in E. helleborine (Appendices 1, 2).
In E. helleborine variation is found in eight of the nine loci examined (Appendix
2). All alleles detected in subsp. neerlandica were also found in populations of subsp.
helleborine. Subsp. neerlandica had fewer alleles, but those found were the same as the
ones most commonly found in subsp. helleborine. Estimates of Nei’s genetic distance
(Table 2) show no larger distances between subsp. neerlandica and subsp. helleborine,
than those found among subsp. helleborine populations. These genetic distance
estimates were used in an UPGMA cluster analysis to construct a dendrogram of
all populations studied (Fig. 3). In this dendrogram subsp. neerlandica (Hjorring NJ)
clusters within subsp. helleborine populations, and therefore, the two entities were
pooled together in the hierarchical analysis.
Genetic variability measures and inbreeding coefficients of the 13 populations
which showed variation in the examined loci (i.e. all the populations of E. helleborine
subsp. helleborine and subsp. neerlandica) are presented in Table 3. Except for the
populations at site no. 1 (subsp. neerlandica) and site no. 3 (subsp. helleborine), there is
no evidence of non-random mating in any of the local populations, as the Fis values
are not statistically significant. The mean gene diversity for all E. helleborine populations
examined is 0.27 and the mean number of alleles per locus, including the monomorphic ones, is 2.2.
The overall genetic differentiation between all local populations of E. helleborine
was found to be 0.09. The hierarchical F-statistic of all local populations sampled
shows that regions had no effect on the variance component; less than 1% of the
genetic variance among local populations can be ascribed to regional subdivision,
whereas more than 90% of the genetic variance was found within local populations
(Table 4). The hierarchical analysis of the local populations in Eastern Jutland only
(Table 5) was done by classifying each wood as a region. In the largest wood,
‘Århusskovene’ sensu Pedersen & Hansen (1998), three local populations were
recognized. All individual local populations fulfill the assumptions of random mating,
i.e. their Fis values do not differ from zero. A significant amount of differentiation
was found between these local populations (FSC =0.065), thus demonstrating the
existence of distinct reproductive units within the same wood. In Århusskovene 23
different alleles were found at the nine loci examined. Only one more allele was
detected when all the local populations from Eastern Jutland were included, and
no additional alleles were found when including local populations from Zealand
and Northern Jutland (Appendix 2).
DISCUSSION
Epipactis helleborine shows high levels of variation at the allozyme loci examined.
The hierarchical analysis indicates a significant among-population differentiation,
whereas no evidence for regional differences has been found. The alleles found in
Epipactis helleborine subsp. neerlandica do not differ from those in E. helleborine subsp.
helleborine, but species-specific alleles have been found in both E. phyllanthes and E.
purpurata. The latter two are monomorphic at all loci examined.
EN 1
EH 3
EH 4a
EH 5
EH 7a
EH 7b
EH 7c
EH 8
EH 10
EH 11
EH 12
EH 13
EH 14
EP
EPh
0
0.167
0.159
0.067
0.061
0.073
0.141
0.047
0.058
0.116
0.063
0.051
0.102
0.376
0.524
EN 1
0
0.031
0.112
0.110
0.104
0.124
0.092
0.103
0.042
0.048
0.083
0.072
0.512
0.469
EH 3
0
0.098
0.108
0.087
0.118
0.086
0.071
0.043
0.029
0.078
0.051
0.437
0.553
EH 4a
0
0.066
0.077
0.176
0.052
0.031
0.110
0.053
0.036
0.029
0.323
0.588
EH 5
0
0.022
0.110
0.056
0.108
0.092
0.044
0.040
0.107
0.496
0.675
EH 7a
0
0.057
0.047
0.114
0.110
0.043
0.038
0.109
0.539
0.673
EH 7b
0
0.067
0.186
0.122
0.083
0.134
0.209
0.613
0.593
EH 7c
0
0.050
0.072
0.026
0.067
0.082
0.380
0.570
EH 8
0
0.072
0.038
0.058
0.021
0.264
0.509
EH 10
0
0.022
0.097
0.081
0.408
0.506
EH 11
0
0.046
0.048
0.387
0.557
EH 12
0
0.040
0.442
0.539
EH 13
0
0.320
0.540
EH 14
0
0.981
EP
0
EPh
T 2. Coefficients of Nei’s genetic distance between Danish populations of four Epipactis taxa. EH, E. helleborine subsp. helleborine; EN, E. helleborine subsp.
neerlandica; EP; E. purpurata; EPh, E. phyllanthes. (Numbers after EH and EN indicate study sites in accordance with Table 1, Fig. 2). Note that since all
populations of E. purpurata and E. phyllanthes were identical for the loci examined, they are only represented once
418
B. K. EHLERS & H. Æ. PEDERSEN
GENETIC VARIATION IN EPIPACTIS
B
or
u
H mE
as
lu J
D nd
al
by EJ
Sk Z
ov
h
St us
or
Z
r
Vi ing
nt
E
J
e
B rZ
os
er
A up
rh
u Z
A sN
rh
E
J
u
Fo sC
le EJ
h
H ave
jo
rr Z
H ing
or
r NJ
A et E
rh
us J
pu S E
rp
J
u
ph ra
t
a
yl
la
nt
419
1
0.96
0.79
0.7
Figure 3. Dendrogram illustrating the genetic relationships among Danish Epipactis taxa and populations,
based on estimates of Nei’s genetic distances (Table 2) and constructed using the UPGMA method of
clustering. EJ, Eastern Jutland; NJ, Northern Jutland, Z, Zealand. Note that since all populations of
E. purpurata and E. phyllanthes were identical for the loci examined, they are only represented once.
T 3. Genetic variables estimated from 9 loci in 13 populations of E. helleborine. The numbering of
study sites is in accordance with Table 1. n, number of individuals examined, A, mean number of
alleles per polymorphic locus, P, percentage of polymorphic loci, He, gene diversity calculated as
1−Rp2i, Ho, observed frequency of heterozygotes, Fis, inbreeding coefficient, ∗P < 0.05 (significance
level corrected according to the sequential Bonferroni procedure)
Population
n
A
P
He
Ho
1
3
4a
5
7a
7b
7c
8
10
11
12
13
14
30
40
19
36
33
35
34
35
15
32
32
31
31
2.17
2.57
2.29
2.86
2.57
3.00
2.57
2.83
2.33
2.83
2.71
2.83
2.57
66.7
77.8
77.8
77.8
77.8
77.8
77.8
66.7
66.7
66.7
77.8
66.7
77.8
0.309
0.262
0.191
0.285
0.271
0.308
0.266
0.322
0.264
0.248
0.282
0.295
0.256
0.257
0.222
0.175
0.265
0.263
0.317
0.248
0.302
0.216
0.264
0.243
0.287
0.265
Mean
SD
31
2.625
0.250
73.6
0.274
0.034
0.256
0.037
Fis
0.193
0.166
0.107
0.061
0.049
−0.016
0.099
0.077
0.212
−0.048
0.153
0.045
−0.022
∗
∗
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
420
B. K. EHLERS & H. Æ. PEDERSEN
T 4. Hierarchical F-statistics for E. helleborine populations with two levels of subdivision (S population;
C region). Three regions are defined: Northern Jutland (study site 1) Eastern Jutland (sites 3–5, 7–8),
and Zealand (sites 10–14). Significance tests are based on 1000 permutations. ∗∗∗ P < 0.001
df
Sum of squares
Percentage of
variation
Among regions
Among populations within regions
Within populations
2
10
799
20.15
80.33
995.97
0.65
8.01
91.34
Total
811
1096.45
Source of variation
F 
Among populations within regions
Among populations among regions
Among regions
Fsc: 0.081∗∗∗
Fst: 0.087∗∗∗
Fct: 0.006 n.s.
T 5. Hierarchical F-statistics for E. helleborine populations in Eastern Jutland with two levels of
subdivision: S population; C region. Five regions (i.e. woods) are defined; Haslund Skov (site 3), Borum
Møllebæk (N) (site 4a), Storring Præstegårdsskov (site 5), Århusskovene (site 7a, 7b, 7c), and Hørret
Skov (site 8). Significance tests are based on 1000 permutations. ∗∗∗ P < 0.001
df
Sum of squares
Percentage of
variation
Among regions
Among populations within regions
Within populations
4
2
457
44.51
14.40
577.95
3.57
6.23
90.20
Total
463
636.85
Source of variation
F 
Among populations within regions
Among populations among regions
Among regions
Fsc: 0.065∗∗∗
Fst: 0.098∗∗∗
Fct: 0.036 n.s.
Variation within species
Schacchi, Lanzara & De Angelis (1987) studied four populations of E. helleborine
in Italy and found correspondingly high estimates of gene diversity (0.23) and mean
numbers of alleles per locus (1.8). In Britain, Hollingsworth & Dickson (1997)
estimated the number of alleles per locus and the gene diversity in both rural and
urban populations. Considering only estimates from the seven rural populations, an
unweighted mean gene diversity and number of alleles per locus were 0.18 and 1.5,
respectively. Thus, the level of genetic diversity in E. helleborine is of similar magnitude
west and east of the North Sea (Britain vs. Denmark) as well as north and south of
the Alps (Britain and Denmark vs. Italy).
The indication that local populations do not become more differentiated on a
larger geographic scale suggests that no ‘isolation by distance’ genetic structure exists
at the geographic levels sampled. This is also illustrated in the UPGMA cluster
analysis, where no geographic grouping is apparent (Fig. 3). In the above study of
GENETIC VARIATION IN EPIPACTIS
421
Hollingsworth & Dickson (1997), in England and Scotland, distinctly less genetic
variation was found to be partitioned between countries than between local populations in the same country. As in our study, an UPGMA dendrogram revealed no
geographic grouping of populations. Considering only rural populations, the mean
overall Fst value is 0.13, which is similar to the one found between the Danish rural
populations.
The significant differentiation found among local populations within regions and
even among local populations within the same wood, suggests that local populations
are sufficiently isolated to allow allele frequencies to drift apart. This demonstrates
that gene flow within E. helleborine occurs on a very local scale. Pollen flow is
restricted, partly because the pollen is packed in pollinia, partly because the flowers
are mainly pollinated by social wasps (family Vespidae) operating on a local scale
(e.g. Darwin, 1869; Wiefelspütz, 1970b; Judd, 1971; Vöth, 1982; Petit, 1986; Müller,
1988; Light & MacConaill, 1998). The large pollinarium adheres to the forehead
of the wasp, in a position which may affect its ability to fly. Furthermore, the wasps
are reported to be intoxicated by the nectar (Løjtnant, 1974; van der Cingel, 1995;
Ehlers & Olesen, 1997). The result is a slow and sluggish behaviour which probably
causes a certain amount of geitonogamy and reduces the probability of long-distance
pollen dispersal. Although E. helleborine subsp. helleborine is mainly allogamous (Waite,
Hopkins & Hitchings, 1991), gene flow through pollen is evidently very restricted,
occurring mainly within the local populations.
The wind dispersed seeds are numerous, minute and probably dominate the gene
flow among local populations of E. helleborine. Seed flow among local populations
within regions is likely higher than seed flow among regions. Occasional long
distance dispersal may have considerable effect on the genetic structure, retarding
the development of regional differences.
Most of the genetic variation in this species was found within local populations,
indicating an extensive (pre)historical or recent gene flow between populations.
Forest was the dominating biotope in Denmark, until it was seriously fragmented
by agriculture. By around 1800, less than 5% of Denmark was covered by forest, a
figure which has now increased to more than 12% (Møller & Staun, 1995) due to
legal protection and reforestation. The (pre)historical level of gene flow was probably
higher than the present one due to a much more continuous distribution of Epipactis
helleborine subsp. helleborine. Gene flow may still be high, even post-fragmented, but
a higher previous gene flow among populations may be considered an alternative
reason for the lack of regional differences. If fragmentation reduced gene flow, any
equilibrium in genetic differentiation among regions may not yet have been established. Assuming a generation time of one to a few decades, equillibrium cannot
have been reached since fragmentation happened only a moderate number of
generations ago.
Although the population of E. helleborine subsp. neerlandica is found to be polymorphic
and with a high level of gene diversity, the number of alleles per locus is lower than the
numbers found in populations of subsp. helleborine. This may be due to modifications in
pollination biology or geographical isolation.
Pollination by social wasps (Dolichovespula spp.) has been observed (Kapteyn den
Boumeester, 1989), but the generally more dense-flowered racemes of E. helleborine
subsp. neerlandica may increase the frequency of geitonogamy as compared with the
frequency in subsp. helleborine. In addition, the fact that the pollinia are less coherent
in subsp. neerlandica suggests the possibility of facultative autogamy (Claessens &
422
B. K. EHLERS & H. Æ. PEDERSEN
Kleynen, 1991). Indeed, in bagging experiments with E. helleborine in the dune areas
of Ameland, Weijer (1952) recognized a reasonably efficient seed set resulting from
spontaneous autogamy. Contrary to Weijer’s opinion, we find it evident from his
description of morphology and habitat that the population in question belongs to
subsp. neerlandica. Thus, both increased geitonogamy and facultative autogamy may
account for extensive inbreeding in E. helleborine subsp. neerlandica, resulting in the
lowered level of allozyme diversity and the significant Fis value found in our study.
The geographic range of E. helleborine subsp. neerlandica in Denmark is somewhat
separated from that of subsp. helleborine (Hansen, 1964) and distinctly isolated from
the range of subsp. neerlandica along the south coast of the North Sea (Fig. 1). This
tendency to geographic isolation must be considered an alternative reason for the
lower number of alleles in the sole study population of subsp. neerlandica.
The monomorphic condition found in E. purpurata may be explained by aspects
of migration, pollination biology, and/or habitat specialization.
The occurrences of E. purpurata in Denmark and the northernmost parts of
Germany constitute a northern ‘satellite range’ of this species (Fig. 1). It is therefore
possible that the monomorphic pattern observed can be explained by a population
bottleneck effect due to founder events at the time when this species migrated into
Denmark. It would be highly interesting to obtain data on the genetic variation of
E. purpurata from its main geographic range.
The pollination biology of E. purpurata is basically similar to that of E. helleborine
(e.g. Darwin, 1869; Wiefelspütz, 1970b; Løjtnant, 1974; Vöth, 1982; Müller, 1988;
Ehlers & Olesen, 1997). It should be noted, however, that the former has more
dense-flowered racemes, generally more flowering stems per genet, and a vertical
rhizome which causes the flowering shoots to form dense clusters (E. helleborine
usually has a horizontal rhizome). These features may well bring about an increased
geitonogamy/allogamy ratio as compared with the one in E. helleborine subsp. helleborine
(see also Wiefelspütz, 1970b). This may have contributed to the monomorphic pattern
observed.
While E. helleborine is a generalist occuring in an extensive range of forest habitats,
in scrubs, and even in urban sites (see, e.g. Hollingsworth & Dickson, 1997), E.
purpurata has much more specific habitat preferences. In Denmark it is restricted to
deciduous forest (mainly beech woods) of long continuity and situated in areas where
plastic clay from the eocene reaches the surface (Løjtnant, 1985; Wind & Faurholdt,
1992; Pedersen & Løjtnant, 1992). It seems that this species has slightly less specific
demands in its main range, although it clearly prefers old hardwood forest located
in areas with mineral soil close to the surface (e.g. Ziegenspeck, 1936; Summerhayes,
1951; Reichling, 1970; Füller, 1986; Eccarius, 1997). Thus, even in its main range,
it has distinctly more specific habitat demands than E. helleborine. The dependency
of E. purpurata on particular habitats limits the number of suitable patches. Additionally, the number of flowering shoots in any population fluctuates markedly
from one year to another, and one or more years may pass without the emergence
of any flowering shoots. Such patterns of flowering have been documented through
monitoring in permanent plots in Denmark since 1987 (Wind, 1998). Habitat
specialization combined with erratic flowering may cause the level of polymorphism
to decrease when founders are few and gene flow between patches is low. The
monomorphic condition of Danish E. purpurata may represent the ultimate genetic
outcome of a decrease caused by such factors.
The monomorphic condition observed in E. phyllanthes has probably originated
GENETIC VARIATION IN EPIPACTIS
423
through inbreeding caused by obligate autogamy. This species is spontaneously
autogamous (Young, 1962; Wiefelspütz, 1970b). The rostellum is strongly reduced,
and the fragile pollinia break apart and fall to the stigma only short time after
anthesis (more rarely, cleistogamous plants are found). Harris & Abbott (1997)
studied two populations of E. phyllanthes in Scotland; one was monomorphic at all
17 loci examined, but the other one varied at a single locus. Reviews of plant
allozyme data (Loveless & Hamrick, 1984; Hamrick et al., 1991) have found that
selfing, narrowly distributed species, in general, exhibit the lowest level of genetic
variation. A number of studies have even reported monomorphism (or very low
polymorphism) in highly selfing species (e.g. Gottlieb, 1984; Kesseli & Jain, 1986;
Ågren & Schemske, 1993). The Danish occurrences of E. phyllanthes are situated
within a larger range which also includes Southern Sweden and the northernmost
parts of Germany (Fig. 1), and the species also has much less specific habitat
preferences than E. purpurata (e.g. Summerhayes, 1951; Young, 1962; Lundqvist,
1965; Richards & Swan, 1976; Faurholdt et al., 1998). Thus, neither aspects of
migration nor habitat specialization seem to be important for interpretation of the
data on E. phyllanthes.
Variation between species
With regard to the nine loci examined, E. phyllanthes and E. purpurata share more
alleles with E. helleborine than with each other. At the same time, the occurrence of
unique alleles clearly demonstrates their justification as separate species.
Neither the allozyme data obtained in the present study nor the morphological
data presented by Kapteyn den Boumeester (1989) support the view of Delforge,
Devillers-Terschuren & Devillers (1991) that E. helleborine subsp. neerlandica should
be treated as a separate species, E. neerlandica (Verm.) J. Devillers-Terschuren & P.
Devillers.
Additionally four Epipactis taxa have previously been examined electrophoretically
by other authors. In Italy, Schacchi et al. (1987) sampled one population of E.
palustris (L.) Crantz, an entomophilous inhabitant of calcarous fens and meadows in
Europe and temperate parts of Asia (Wiefelspütz, 1970a; Nilsson, 1978). They
concluded that the gene diversity found was similar to figures found in indexes for
outbreeding plants. E. microphylla (Ehrh.) Sw., a facultatively autogamous species
growing mainly in dark beech woods in Central and Southern Europe eastwards to
the Caucasus (Wiefelspütz 1970a, b; Füller, 1986; Burns-Balogh, Szlachetko & Dafni,
1987), was studied in two sites approximately 7 km apart. No variation was found
either within or between the populations, and it was suggested that they might
originate from one genetically impoverished ancestral population. Epipactis leptochila
(Godfery) and E. youngiana A. J. Richards & A. F. Porter, both believed to be
obligately autogamous species, were by Harris & Abbott (1997) found to contain
surprisingly high levels of genetic diversity, which may question the status (or age)
of their current breeding system.
It generally seems that the widespread and obligately entomophilous taxon, E.
helleborine subsp. helleborine shows the highest level of genetic variation. The sole
population of the entomophilous E. palustris studied had lower levels of genetic
diversity than the sole population of the facultatively autogamous E. helleborine subsp.
neerlandica. However, based on only one population of each, no inferences should
424
B. K. EHLERS & H. Æ. PEDERSEN
be made on the effect of breeding system on the genetic variation in these taxa. It
should be noted, though, that also E. palustris shows higher levels of genetic diversity
than the obligately autogamous and more narrowly distributed E. phyllanthes. E.
purpurata has, to our knowledge, only been studied by means of electrophoresis in
Denmark, where it is isolated from its main range. Further genetic data from Central
Europe are needed before reliable conclusions on the influence of breeding system
and distribution on the amount of genetic variation can be drawn. With regard to
E. leptochila and E. youngiana, the effect of breeding system on the genetic structure
must be considered when further information on the mode of reproduction in these
taxa has been provided. The breeding systems of Epipactis taxa studied so far may
broadly predict the genetic structure of the taxa concerned. The predominantly
allogamous taxa are more genetically variable than the autogamous ones. However,
further knowledge of ecology and life history are needed for understanding results
that deviate from general expectations.
Scenario: inferences on the evolution towards autogamy and speciation in Epipactis
Estimates of genetic variation in the present study, in Harris & Abbott (1997), in
Hollingsworth & Dickson (1997), and in Schacchi et al. (1987) strongly support that
E. helleborine subsp. helleborine is an outcrossing taxon. However, the hierarchical
analysis in our study suggests that reproductive units occur on a very local scale,
thus facilitating the possibility of local evolution. Morphological changes influencing
the breeding system are more likely to become established in such local breeding
groups.
The genus Epipactis exhibits a wide variation in breeding system ranging from
allogamy to obligate autogamy (cf. Nilsson, 1981; Richards, 1982; van der Cingel,
1995). Most of the obligately autogamous species produce nectar, indicating that
these species have evolved from entomophilous ancestors (Richards, 1986; BurnsBalogh et al., 1987).
The morphological changes needed from allogamous to autogamous flowers in
Epipactis are very small, the main barrier preventing self-pollination being the
rostellum. In E. helleborine the rostellum is much shorter than the anther in an
early ontogenetic stage of the gynostemium (Rasmussen, 1982). This condition is
reminiscent of the condition in fully developed buds and flowers of certain obligately
autogamous Epipactis taxa, suggesting that some of the autogamous taxa may have
evolved from allogamous ancestors through paedomorphosis.
Neotenic evolution as the mechanism causing shift in pollination biology, subsequently leading to speciation, was demonstrated by Guerrant (1982) in Delphinium
nudicaule. If simple paedomorphosis is sometimes responsible for shifts from allogamy
to obligate autogamy in Epipactis, then such occasional events, combined with the
occurrence of local breeding groups recognized in the present study, would be
expected to result in the establishment of populations of new autogamous entities.
Especially when new autogamous entities evolve locally in extreme environments,
unfavourable to entomophilous plants, they have a chance of long-term survival,
possibly leading to speciation (Baker, 1955; Lloyd, 1965; de Arroyo, 1975; Schoen
1982; Robatsch 1983).
Epipactis helleborine and E. purpurata flower late in season, leaving only wasps as
pollinators. If harsh (e.g. sandy, low nutrient) environment retards the developmental
GENETIC VARIATION IN EPIPACTIS
425
A
a
p
v
1 mm
r
s
B
a
p
s
Figure 4. Gynostemia of (A) Epipactis helleborine subsp. neerlandica and (B) E. renzii in lateral view.
a, anther; p, pollinia; v, diffuse viscidium; r, rostellum; s, stigma.
process of individuals, there will be a selective pressure on those plants to shorten
the timespan of ontogenetic development. Such environmental conditions may well
trigger off the evolution of local, obligately autogamous entities which are sometimes
regarded as separate species. Epipactis renzii K. Robatsch, growing in sand dunes by
the north-western coast of Jutland, and E. pseudopurpurata P. Mered’a, growing in
the western part of the Strázovské vrchy Mountains of Slovakia, are both autogamous
and narrow endemics (Robatsch, 1988; Mered’a, 1996). These taxa may have
evolved very recently from E. helleborine subsp. neerlandica and E. purpurata, respectively.
Morphologically, they are only separated from their presumed ancestors through
modifications in column structure (Fig. 4).
ACKNOWLEDGEMENTS
We wish to express our grateful thanks to Niels Faurholdt for helping us locating
suitable populations for study, to Jens Mogens Olesen and Mikkel H. Schierup for
fruitful discussions and comments on the manuscript, to Jørgen E. Andersen for
technical assistance, and to Volker Loeschcke, Jon Ågren, and two anonymous
reviewers for useful comments on the manuscript. Additionally, we are indebted to
the Forestry Commission, under the Danish Ministry of Environment and Energy,
for exemption from the national preservation of orchids.
REFERENCES
Ågren J, Schemske DW. 1993. Outcrossing rate and inbreeding depression in two annual monoecious
herbs, Begonia hirsuta and B. semiovata. Evolution 47: 125–135.
426
B. K. EHLERS & H. Æ. PEDERSEN
Arroyo de MTK. 1975. Electrophoretic studies of genetic variation in natural populations of
allogamous Limnanthes alba and autogamous Limnanthes floccosa (Limnanthaceae). Heredity 35: 153–164.
Baker HG. 1955. Self-compatibility and establishment after ‘long-distance’ dispersal. Evolution 9:
347–349.
Burns-Balogh P, Szlachetko DL, Dafni A. 1987. Evolution, pollination, and systematics in the
tribe Neottieae (Orchidaceae). Plant Systematics and Evolution 156: 91–115.
Chakraborty R, Leimar O. 1987. Genetic variation within a subdivided population. In: Ryman,
R, Utter, F, eds. Population genetics & fisheries management. Seattle: Washington Sea Grant Program,
89–120.
Cingel van der NA. 1995. An atlas of orchid pollination. European orchids. Rotterdam: AA Balkema.
Claessens J, Kleynen J. 1991. Het geslacht Epipactis in de Benelux: bloemenbiologische beschrijvingen
en soorttypische kenmerken. Eurorchis 3: 5–38.
Clayton JW, Tretiak DN. 1972. Amine-citrate buffers for pH control in starch gel electrophoresis.
Journal of the Fisheries Research Board of Canada 29: 1169–1172.
Darwin C. 1869. Notes on the fertilization of orchids. Annals and Magazine of Natural History 4IV:
141–159.
Delforge P, Devillers-Terschuren J, Devillers P. 1991. Contributions taxonomiques et nomenclaturales aux orchidées d’Europe (Orchidaceae). Les Naturalistes belges 72: 99–101.
Dressler RL. 1989. Rostellum and viscidium: divergent definitions. Lindleyana 4: 48–49.
Dressler RL. 1993. Phylogeny and classification of the orchid family. Cambridge: Cambridge University
Press.
Eccarius W, ed. 1997. Orchideen in Thüringen. Uhlstädt: Arbeitskreis Heimische Orchideen Thüringen
e. V.
Ehlers BK, Olesen JM. 1997. The fruit-wasp route to toxic nectar in Epipactis orchids? Flora 192:
223–229.
Excoffier L, Smouse P, Quattro J. 1992. Analysis of molecular variance inferred from metric
distances among DNA haplotypes: Application to human mitochondrial DNA restriction data.
Genetics 131: 479–491.
Faurholdt N, Pedersen HÆ, Christiansen SG. 1998. Nikkende Hullæbe (Epipactis phyllanthes) –
en miskendt dansk orkidé. Urt 22: 52–59.
Felsenstein J. 1993. PHYLIP (Phylogenetic Inference Package, Version 3.5c) Seattle: Department of Genetics,
University of Washington.
Fildes RA, Harris H. 1966. Genetically determined variation of adenylate kinase in man. Nature
209: 262–263.
Füller F. 1986. Epipactis und Cephalanthera. Orchideen Mitteleuropas, 5. Teil. 3rd ed. Wittenberg Lutherstadt:
A. Ziemsen Verlag.
Gottlieb LD. 1984. Electrophoretic analysis of the phylogeny of the self-pollinating populations of
Clarkia xanthiana. Plant Systematics and Evolution 147: 91–102.
Goudet, J. 1995. FSTAT (version 1.2): A computer program to calculate F-statistics. Journal of Heredity
86: 485–486.
Guerrant EO. 1982. Neotenic evolution of Delphinium nudicaule (Ranunculaceae): A hummingbirdpollinated Lackspur. Evolution 36: 699–712.
Hamrick JL, Godt MJ. 1990. Allozyme diversity in plant species. In: Brown, AHD, Clegg, MT,
Kahler, AL, Weir, BS, eds. Plant population genetics, breeding, and genetic resources. Sunderland, Mass:
Sinauer Associates Inc., 43–63.
Hamrick JL, Godt MJW, Murawski DA, Loveless MD. 1991. Correlations between species traits
and allozyme diversity: Implications for conservation biology. In: Falk DA, Holsinger KE, eds.
Genetics and Conservation of rare plants New York: Oxford University Press, 75–86.
Hansen A. 1964. Epipactis helleborine var. neerlandica i Danmark. Botanisk Tidsskrift 59: 334–337.
Harris SA, Abbott RJ. 1997. Isozyme analysis of the reported origin of a new hybrid orchid species,
Epipactis youngiana (Young’s helleborine), in the British Isles. Heredity 79: 402–407.
Hartl DL, Clark AG. 1989. Principles of Population Genetics. Sunderland, Mass: Sinauer Associates, Inc.
Hollingsworth PM, Dickson JH. 1997. Genetic variation in rural and urban populations of Epipactis
helleborine (L.) Crantz. (Orchidaceae) in Britain. Botanical Journal of the Linnean Society 123: 321–331.
Judd WW. 1971. Wasps (Vespidae) pollinating helleborine, Epipactis helleborine (L.) Crantz, at Owen
Sound, Ontario. Proceedings of the Entomological Society of Ontario 102: 115–118.
Kapteyn den Boumeester DW. 1989. Epipactis helleborine subsp. neerlandica Vermeulen – problematiek,
veldwaarnemingen, bestuivers. Eurorchis 1: 93–111.
GENETIC VARIATION IN EPIPACTIS
427
Karron JD. 1991. Patterns of genetic variation and breeding systems in rare plants. In: Falk DA,
Holsinger KE, eds. Genetics and conservation of rare plants New York: Oxford University Press, 87–98.
Kesseli RV, Jain SK. 1986. Breeding system and population structure in Limnanthes. Theoretical and
Applied Genetics 71: 292–299.
Legendre L, Legendre P. 1983. Numerical Ecology Amsterdam: Elsevier Scientific Publishing Company.
Light MHS, MacConaill M. 1998. Factors affecting germinable seed yield in Cypripedium calceolus
var. pubescens (Willd.) Correll and Epipactis helleborine (L.) Crantz (Orchidaceae). In: Waite S, ed.
Orchid population biology: conservation and challenges. Botanical Journal of the Linnean Society 126:
3–26.
Lloyd DG. 1965. Evolution of self-compatibility and racial differentiation in Leavenworthia (Cruciferae).
Contributions from the Gray Herbarium of Harvard University 195: 3–134.
Løjtnant B. 1974. Giftig nektar, ‘fulde’ hvepse og orchideer. Kaskelot 15: 3–7.
Løjtnant B. 1985. Orchideerne – status og fremtid i Århus Amt. In: Højager S, Janniche A, Laursen
JT, eds. Status for planter, dyr og naturbeskyttelse i Århus Amt. Århus: Danmarks Naturfredningsforening,
Dansk Ornitologisk Forening, Østjysk Biologisk Forening, 100–109.
Loveless MD, Hamrick JL. 1984. Ecological determinants of genetic structure in plant populations.
Annual Review of Ecology and Systematics 15: 65–95.
Luer CA. 1975. The native orchids of the United States and Canada excluding Florida. New York: The New
York Botanical Garden.
Lundqvist Å. 1965. En ölandsk Epipactis phyllanthes G.E. Sm. Svensk Botanisk Tidskrift 59: 216–224.
Lundqvist Å. 1966. What is Epipactis confusa Young? Svensk Botanisk Tidskrift 60: 33–48.
Mered’a P. 1996. Epipactis pseudopurpurata Mered’a, spec. nova (Orchidaceae) – eine neue autogame
Sitter-Art aus der Slowakei. Preslia 68: 23–29.
Møller PF, Staun S. 1995. Danmarks skove. Aalborg: Politikens Forlag A/S.
Müller I. 1988. Vergleichende blütenökologische Untersuchungen an der Orchideengattung Epipactis.
Mitteilungsblatt Arbeitskreis Heimische Orchideen Baden-Württemberg 20: 701–803.
Nei M. 1972. Genetic distances between populations. American Naturalist 106: 283–292.
Nilsson LA. 1978. Pollination ecology of Epipactis palustris (Orchidaceae). Botaniska Notiser 131: 355–368.
Nilsson LA. 1981. Pollination ecology and evolutionary processes in six species of orchids. Abstracts
of Uppsala Dissertations from the Faculty of Science 593: 1–40 & enclosures I–VI.
Pedersen HÆ, Hansen AM. 1998. Århusskovenes flora. En undersøgelse af karplantefloraens
diversitet, autenticitet og beskyttelsesbehov. Gejrfuglen 34: 2–79.
Pedersen HÆ, Løjtnant B. 1992. Praktisk anvendelse af naturskovs-indikatorer – vist med karplanterne som eksempel. Gejrfuglen 28: 132–144.
Petit J. 1986. Quelques observations sur les insectes butineurs d’Epipactis helleborine (L.) Crantz.
Dumortiera 34–35: 112–116.
Rasmussen FN. 1982. The gynostemium of the neottioid orchids. Opera Botanica 65: 1–96.
Raymond M, Rousset F. 1995. GENEPOP (version 1.2): A population genetic software for exact
tests and ecumenicism. Journal of Heredity 86: 248–249.
Reichling L. 1970. Die Gattung Epipactis in Luxemburg. In: Senghas K, Sundermann H, eds.
Probleme der Orchideengattung Epipactis. Jahresberichte des Naturwissenschaftlichen Vereins Wuppertal 23:
88–97 (1–132 & 3 pls).
Rice WR. 1989. Analyzing tables of statistical tests. Evolution 43: 223–225.
Richards AJ. 1982. The influence of minor structural changes in the flower on breeding systems and
speciation in Epipactis Zinn (Orchidaceae). In: Armstrong JA, Powell JM, Richards AJ, eds. Pollination
and Evolution. Sydney: Royal Botanic Gardens, 47–53.
Richards AJ. 1986. Plant breeding systems. London: Chapman and Hall.
Richards AJ, Swan GA. 1976. Epipactis leptochila (Godfery) Godfery and E. phyllanthes G. E. Sm.
occurring in South Northumberland on lead and zinc soils. Watsonia 11: 1–5.
Robatsch K. 1983. Beiträge zur Blütenbiologie und Autogamie der Gattung Epipactis. In: Senghas,
K, Sundermann, H, eds. Probleme der Taxonomie, Verbreitung und Vermehrung europäischer
und mediterraner Orchideen. Jahresberichte des Naturwissenschaftlichen Vereins Wuppertal 36: 25–32, Pl.
3(14–16) (1–113, Pls 1–6).
Robatsch K. 1988. Beiträge zur Kenntnis der europäischen Epipactis-Arten (Orchidaceae). Linzer
Biologische Beiträge 20: 161–172.
Schacchi R, Lanzara P, De Angelis G. 1987. Study of electrophoretic variability in Epipactis
helleborine (L.) Crantz, E. palustris (L.) Crantz and E. microphylla (Ehrh.) Swartz (fam. Orchidaceae).
Genetica 72: 217–224.
428
B. K. EHLERS & H. Æ. PEDERSEN
Schick B. 1989. Zur anatomie und Biotechnik des Bestäubungsapparates der Orchideen. II: Epipactis
palustris (L.) Crantz und Listera ovata (L.) R. Br. Botanische Jahrbücher für Systematik, Pflanzengeschichte und
Pflanzengeographie 110: 289–323.
Schneider S, Kueffer J-M, Roessli D, Excoffier L. 1997. Arlequin ver. 1.1: A software for population
genetic data analysis. Geneva: Genetics and Biometry Laboratory, University of Geneva.
Schoen DJ. 1982. The breeding system of Gilia achilleifolia: Variation in floral characteristics and
outcrossing rate. Evolution 36: 352–360.
Soper JH, Garay LA. 1954. The Helleborine and its recent spread in Ontario. Bulletin, Federation of
Ontario Naturalists 65: 4–7.
Summerhayes VS. 1951. Wild orchids of Britain, with a key to the species. London: Collins.
Uyenoyama MK, Holsinger KE, Waller DM. 1993. Ecological and genetic factors directing the
evolution of self-fertilization. Oxford Surveys in Evolutionary Biology 9: 327–381.
Vöth W. 1982. Blütenökologische Untersuchungen an Epipactis atrorubens, E. helleborine und E. purpurata
in Niederösterreich. Mitteilungsblatt Arbeitskreis Heimische Orchideen Baden-Württemberg 14: 393–437.
Waite S, Hopkins N, Hitchings S. 1991. Levels of pollinia export, import and fruit set among
plants of Anacamptis pyramidalis, Dactylorhiza fuchsii, and Epipactis helleborine. In Wells, TCE, Willems,
JH, eds. Population ecology of terrestrial orchids. The Hague: SPB Academic Publishing bv, 103–110.
Weijer J. 1952. The colour-differences in Epipactis helleborine (L.) Cr. Wats. & Coult., and the selection
of the genetical varieties by the environment. Genetica 26: 1–32.
Weir BS, Cockerham CC. 1984. Estimating F-statistics for the analysis of population structure.
Evolution 38: 1358–1370.
Wendel JF, Weeden NF. 1989. Visualization and interpretation of plant isozymes. In: Soltis DE,
Soltis PS, eds. Isozymes in plant biology. Portland, Oregon: Dioscorides Press, 5–45.
Wiefelspütz W. 1970a. Zur Verbreitung der europäischen allogamen Epipactis-Arten. In: Senghas, K,
Sundermann, H, eds. Probleme der Orchideengattung Epipactis. Jahresberichte des Naturwissenschaftlichen
Vereins Wuppertal 23: 38–42 (1–132 & 3 pls).
Wiefelspütz W. 1970b. Über die Blütenbiologie der Gattung Epipactis. In: Senghas K, Sundermann
H, eds. Probleme der Orchideengattung Epipactis. Jahresberichte des Naturwissenschaftlichen Vereins
Wuppertal 23: 53–69 (1–132 & 3 pls).
Wind P. 1998. Naturovervågning. Overvågning af orkidéer 1997, Danmark. Arbejdsrapport fra DMU
74: 1–91.
Wind P, Faurholdt N. 1992. Fund af Tætblomstret Hullæbe (Epipactis purpurata) på Sjælland. Urt
1992: 9–16.
Wright S. 1943. Isolation by distance. Genetics 28: 114–128.
Wright S. 1946. Isolation by distance under diverse systems of mating. Genetics 31: 139–156.
Wyatt R. 1988. Phylogenetic aspects of the evolution of self-pollination. In: Gottlieb LD, Jain SK,
eds. Plant evolutionary biology. London: Chapman and Hall, 109–131.
Young DP. 1962. Studies in the British Epipactis VI. Some further notes on E. phyllanthes. Watsonia 5:
136–139.
Ziegenspeck H. 1936. Orchidaceae. In: von Kirchner O, Loew E, Schröter C, Wangerin W, eds.
Lebensgeschichte der Blütenpflanzen Mitteleuropas. Spezielle Ökologie der Blütenpflanzen Deutschlands, Österreichs
und der Schweiz, 1(4). Stuttgart: Verlagsbuchhandlung Eugen Ulmer, I–VII, 1–840.
GENETIC VARIATION IN EPIPACTIS
429
APPENDIX 1
Allele frequencies at nine loci in five populations of E. purpurata and six populations of E. phyllanthes.
The numbering of populations reflects the numbering of study sites in Table 1. (N=number of plants
examined). Designation of alleles indicate their migration on the gels, 1 being the fastest etc.
E. purpurata
Population:
N:
E. phyllanthes
3
29
4b
7
5
13
7a
34
9
18
2
19
6
34
7a
14
7b
14
7c
29
12
15
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
0
1
0
1
0
1
0
1
1
0
1
0
1
0
1
0
1
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
0
1
0
1
0
1
0
1
0
0
1
0
1
0
1
0
1
0
1
0
1
1
0
1
0
1
0
1
0
1
0
0
1
0
1
0
1
0
1
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
Idh-2
1
2
3
0
1
0
1
0
1
0
1
0
1
1
0
1
0
1
0
1
0
1
0
1
0
Pgd
1
1
1
1
1
1
1
1
1
1
1
1
Locus
Gpi
1
2
3
4
Tpi
1
2
3
4
5
Pgm
1
2
3
Skd
1
2
Mdh-1
1
2
3
Mdh-2
1
2
3
4
Idh-1
1
2
3
430
B. K. EHLERS & H. Æ. PEDERSEN
APPENDIX 2
Allele frequencies at nine loci in 12 populations of E. helleborine subsp. helleborine (nos 3–14) and one
of E. helleborine subsp. neerlandica (no. 1). The numbering of populations is in accordance with the
numbering of study sites in Table 1. (N=number of plants examined). Designation of alleles indicate
their migration on the gels, 1 being the fastest etc.
Population:
N:
1
28
3
40
4a
19
5
36
7a
33
7b
35
7c
34
8
35
10
15
11
32
12
32
13
31
14
31
Gpi
1
2
3
4
0.45
0
0.04
0.52
0.1
0
0.538
0.363
0.184
0
0.263
0.552
0.236
0.028
0.194
0.542
0.303
0
0.045
0.652
0.186
0.014
0.043
0.757
0.176
0.074
0.162
0.588
0.414
0
0.2
0.386
0.433
0
0.167
0.4
0.484
0.016
0.219
0.281
0.394
0
0.136
0.47
0.113
0
0.161
0.726
0.21
0
0.258
0.532
Tpi
1
2
3
4
5
0
0
0.55
0.45
0
0.075
0.113
0.55
0.263
0
0
0
0.816
0.184
0
0.045
0.076
0.47
0.288
0.121
0
0.091
0.273
0.591
0.045
0.086
0.071
0.414
0.4
0.029
0
0.337
0.652
0.121
0
0.014
0.086
0.571
0.3
0.029
0
0
0.767
0.233
0
0.047
0.188
0.578
0.188
0
0.052
0.052
0.603
0.276
0.017
0
0.078
0.484
0.422
0.016
0.016
0.065
0.661
0.258
0
Pgm
1
2
3
0.5
0.5
0.825
0.175
0.921
0.079
0.75
0.25
0.591
0.409
0.543
0.457
0.309
0.691
0.529
0.471
0.821
0.179
0.797
0.203
0.719
0.281
0.726
0.274
0.984
0.016
Skd
1
2
0.16
0.84
0.138
0.863
0.105
0.895
0.083
0.917
0.076
0.924
0.243
0.757
0.279
0.721
0.257
0.743
0.1
0.9
0.031
0.969
0.121
0.879
0.129
0.871
0.129
0.871
0.5
0.063
0.026
0.542
0.227
0.2
0.029
0.329
0.5
0.063
0.182
0.403
0.451
Locus
Mdh-1
1
2
3
0.5
0.938
0.974
0.458
0.773
0.8
0.971
0.671
0.5
0.938
0.818
0.597
0.548
Mdh-2
1
2
3
4
0
0.6
0.4
0
0.1
0.638
0.263
0
0.105
0.763
0.132
0
0.028
0.875
0.097
0
0.076
0.758
0.167
0
0.1
0.714
0.129
0.057
0.118
0.721
0.162
0
0.114
0.786
0.1
0
0.067
0.733
0.2
0
0.047
0.625
0.328
0
0.061
0.742
0.2
0
0.078
0.609
0.297
0.015
0.016
0.774
0.21
0
Idh-1
1
2
3
0
1
0
0.013
0.988
0
0
0.974
0.026
0
0.986
0.014
0
0.985
0.015
0
0.956
0.044
0
1
0
0
1
0
0
1
0
0
1
0
0.032
0.968
0
0
1
0
0.048
0.951
0
0
1
0
1
0
1
0
1
0
1
0
1
0.015
0.985
0
1
0
1
0
1
0
1
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Idh-2
1
2
3
Pgd
1