Himantoglossum hircinum (Lizard Orchid) reviewed in the light of new morphological and molecular observations

Abstract

Himantoglossum hircinum is one of the rarer and more charismatic orchids in the British flora. Morphometric comparison of the two largest and best-known populations in southern England – the coastal dune population at Sandwich and the chalk grassland population at Newmarket – using 46 characters showed that they differ only subtly, the Sandwich plants being on average more vegetatively robust and slightly more darkly pigmented, but possessing less extensive lip-spots and substantially longer 'arms'. A comparatively morphologically divergent semi-desert population from Ifrane, Morocco differs from the English populations in having broader stems, less recurved 'arms', a more strongly down-curved spur and in lacking near-circular spots within the sepals. Molecular comparison of 46 plants, representing 13 English populations and 18 populations from Continental Europe and Morocco, revealed only subtle distinctions in the high-copy nuclear region ITS, and smaller-scale comparisons of the low-copy nuclear (LEAFY) and plastid (four intron) regions proved to be even less discriminatory. These results reinforce prior morphological inferences that H. hircinum is a cohesive species. Scanning electron microscopy elucidated the ontogeny of these remarkable flowers, suggesting that the exceptionally elongate central labellar lobe originated by accelerated heterochronic growth and showing that the characteristic spiral torsion always runs counter-clockwise. Lateral fusion of the paired viscidia is convergent with several other lineages of subtribe Orchidinae. Review of pollination and life-history features of H. hircinum suggest that they are typical of food-deceptive species within Orchidinae. The Lizard Orchid is infamous for geographic mobility; its cycles of expansion and contraction through the last century have been interpreted as reflecting a net northward migration in response to recent climate change. Our data tentatively suggest relatively recent colonisation of Morocco at high altitudes and an overall northwestward direction of migration into the UK. ITS ribotypes indicate multiple immigration events leading to levels of genetic diversity in England comparable with those on the Continent. A non-recent origin is inferred for H. hircinum which, despite recent systematic revisions, may harbour further cryptic species; the taxonomic status of supposed outlying populations in southern Italy in particular is questioned by the present genetic data.

Full text

Himantoglossum hircinum (Lizard Orchid)
reviewed in the light of new morphological
and molecular observations
R. M. Bateman*
1
, P. J. Rudall
1
, J. A. Hawkins
2
, G. Sramko
´
3
1
Jodrell Laboratory, Royal Botanic Gardens Kew, Richmond, Surrey, TW9 0ED, UK,
2
School of Biological
Sciences, Plant Science Laboratories, University of Reading, Reading, RG6 6UR, UK,
3
MTA-ELTE-MTM Ecology
Research Group, Budapest, Hungary
Himantoglossum hircinum is one of the rarer and more charismatic orchids in the British flora. Morphometric
comparison of the two largest and best-known populations in southern England – the coastal dune
population at Sandwich and the chalk grassland population at Newmarket – using 46 characters showed
that they differ only subtly, the Sandwich plants being on average more vegetatively robust and slightly
more darkly pigmented, but possessing less extensive lip-spots and substantially longer ‘arms’. A
comparatively morphologically divergent semi-desert population from Ifrane, Morocco differs from the
English populations in having broader stems, less recurved ‘arms’, a more strongly down-curved spur and
in lacking near-circular spots within the sepals. Molecular comparison of 46 plants, representing 13 English
populations and 18 populations from Continental Europe and Morocco, revealed only subtle distinctions in
the high-copy nuclear region ITS, and smaller-scale comparisons of the low-copy nuclear (LEAFY) and
plastid (four intron) regions proved to be even less discriminatory. These results reinforce prior
morphological inferences that H. hircinum is a cohesive species. Scanning electron microscopy elucidated
the ontogeny of these remarkable flowers, suggesting that the exceptionally elongate central labellar lobe
originated by accelerated heterochronic growth and showing that the characteristic spiral torsion always
runs counter-clockwise. Lateral fusion of the paired viscidia is convergent with several other lineages of
subtribe Orchidinae. Review of pollination and life-history features of H. hircinum suggest that they are
typical of food-deceptive species within Orchidinae. The Lizard Orchid is infamous for geographic mobility;
its cycles of expansion and contraction through the last century have been interpreted as reflecting a net
northward migration in response to recent climate change. Our data tentatively suggest relatively recent
colonisation of Morocco at high altitudes and an overall northwestward direction of migration into the UK.
ITS ribotypes indicate multiple immigration events leading to levels of genetic diversity in England
comparable with those on the Continent. A non-recent origin is inferred for H. hircinum which, despite
recent systematic revisions, may harbour further cryptic species; the taxonomic status of supposed
outlying populations in southern Italy in particular is questioned by the present genetic data.
Keywords: climate change, floral ontogeny, geographic distribution, Internal Transcribed Spacer, LEAFY, morphometrics
Introduction
Despite being widespread and locally common across
much of Western Europe, the robust and visually
striking Lizard Orchid (Himantoglossum hircinum
(L.) Spreng.) is rare in the UK and was among the
first tranche of 21 vascular plant species to be placed
on Schedule 8 of the 1981 Wildlife and Countryside
Act. It still appears in the British Red Data Book,
where its national conservation status was down-
graded from Vulnerable to Near-Threatened between
1999 and 2005 (Farrell, 1999; Cheffings & Farrell,
2005). Brief scrutiny of the British Plant Atlas
(Preston et al., 2002) revealed 19 post-1987 hectad
records (excluding Jersey), scattered across a triangle
of south-eastern England that extends westward to
Berrow, Somerset and northward to Lakenheath,
Suffolk; however, localities are concentrated particu-
larly strongly in Kent (Fig. 1).
In England, the habitat preferences of the Lizard
Orchid mirror those of the Pyramidal Orchid
(Anacamptis pyramidalis): it prefers a shallow calcar-
eous soil – either limestone or blown coastal sand rich
in comminuted invertebrate shells – but is more
tolerant than some grassland orchids of competition
*Corresponding author: R.Bateman@kew.org
122
ßBotanical Society of Britain & Ireland 2013
DOI 10.1179/2042349713Y.0000000025 New Journal of Botany 2013 VOL.3 NO.2

from longer grasses and scrubby bushes, and of
periodic drought (Fig. 3). A further tolerance for
soils disturbed within antiquity means than a
significant proportion of its localities lie outside
nature reserves; sites include roadside verges and
even lawns (Fig. 3a and c). The predilection of
H. hircinum for golf courses and horse-racing courses
in England is legendary, but in truth, such sites
encompass only a minority of recent localities (Carey,
1999; Carey & Farrell, 2002).
The Lizard Orchid shows broadly similar habitat
preferences in mainland Europe, where its core
distribution almost precisely coincides with present-
day France (Pfeifer et al., 2009; contra Carey &
Farrell, 2002). Tongues protrude northwards into
southern England and the Low Countries and
northeastward into south-central Germany. To the
south, outliers occur to the southwest in northern
Spain, southern Iberia, and northern Morocco. In
northern Italy, H. hircinum is considered to be
replaced by its sister-species, H. adriaticum.
However, H. hircinum is widely considered to
reappear in southern Italy and Sicily, and reputedly
also occurs in the Tunisian hills opposite Sicily.
There have been comparatively few studies of
pollination in H. hircinum but these include one that
used this species as a model system for the study of
geitonogamy – transfer of pollinaria between geneti-
cally identical flowers occupying the same inflores-
cence (Kropf & Renner, 2008). A few populations
have also been subjected to unusually detailed studies
of annual demographics (Carey, 1998, 1999; Heinrich,
2003; Pfeifer, 2004; Pfeifer et al., 2006a).
Perhaps the most impactful scientific investigations
of H. hircinum have focused on its broader geographic
distribution, fuelled by the pioneering arguments of
Good (1936) that this species is particularly sensitive
to, and thus is a credible model indicator of, climate
change (Hartley et al., 2004). This assertion has been
reinforced by recent autecological/demographic stu-
dies by Carey and colleagues (Carey & Brown, 1994;
Carey, 1998, 1999; Farrell & Carey, 1999; Carey &
Farrell, 2002; Carey et al., 2002) in the UK and Pfeifer
and colleagues (Pfeifer, 2004; Pfeifer & Jetschke, 2006;
Pfeifer et al., 2006a, b, 2009, 2010) in south-central
Germany – a country where H. hircinum was desig-
nated the ‘Orchid of the Year’ in 1999 (Heinrich &
Voelckel, 1999). Most of these studies have paid
particular attention to the more northerly populations
of H. hircinum, whereas no previous author has
considered the southern-most occurrences of this
species – these are found in northwest Africa and were
sampled for the present study.
Although species circumscription in the genus
Himantoglossum has occasionally been addressed
(Delforge, 1999; Sramko´ & Molna´r, 2012; Sramko´
et al., 2013), the process has been less rigorous than in
several other genera of European Orchideae (cf.
Bateman, 2012). Interest has centred on H. adriaticum,
which reputedly replaces H. hircinum in northern Italy
and the Balkans, and on an eastern Mediterranean
complex that is viewed as a single species by some
observers but as many as six subtly distinct species by
others. In addition, the evolutionary origin and
ontogenetic underpinning of the remarkable floral
morphology of H. hircinum have received little atten-
tion relative to those of other genera of Orchideae
(Kurzweil, 1987; Box et al., 2008; Rudall et al., 2013).
Here, we offer a more rounded account of
H. hircinum, utilising morphometric, molecular and
microscopic techniques in order to better understand
its species circumscription, infraspecific variation,
and floral evolution. We set these new observations
in the context of previous interpretations of the
biology and ecology of this charismatic orchid.
Materials and methods
Morphometrics
Field sampling
The modest sampling for morphometric analysis
targeted the two largest, longest-established, best-
known and most carefully monitored English popula-
tions of H. hircinum, located at Sandwich, Kent
(maximum recorded number of plants5c.6000)
and Newmarket, Cambridgeshire (maximum5c.250).
These localities provided a useful contrast of the
species’ two preferred habitats: stabilised sand dunes
at Sandwich and chalk grassland at Newmarket
(Figs. 2 and 3). Ten plants were measured in each of
the two populations in June 2010. The site selected
within the Sandwich population was located in short
exposed grassland on a shallow west-facing slope
Figure 1 Distribution of Himantoglossum hircinum in the
British Isles, showing the localities sampled for DNA (red
spots) and for morphometric study (letters: S5Sandwich,
N5Newmarket). Base map reproduced from Preston et al.
(2002, p. 854).
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 123

along the line of dunes adjacent to the beach at c.1 m
a.s.l. (Fig. 3d). The smaller Newmarket population
was concentrated in longer, quinquennially burned
grassland clothing the steep south-facing bank of an
ancient earthwork at c.30 m a.s.l. (Fig. 3e). The even
smaller (five flowering plants) comparative population
Figure 2 Representative plants of Himantoglossum hircinum from the Sandwich (c, g), Newmarket (f), and Ifrane (b, e)
populations, plus the transient Headley plant (a, d). (e)–(g) are reproduced at the same scale (vertical dimension565 mm).
Images: (f)5D.M.T. Ettlinger, (g)5B.G. Tattersall, remainder5R.M. Bateman.
Bateman et al. Observations on the Lizard Orchid
124 New Journal of Botany 2013 VOL.3 NO.2

from Morocco, examined in May 2012, was located
3 km southwest of Ifrane along the northern slopes of
the Middle Atlas; it contrasted strongly in altitude
with the other study sites, occurring at 1650 m a.s.l.
The semi-arid site was a subdued rocky hillock
surrounded by sparse, goat-grazed dwarf scrub near
the margin of extensive cedar forests (Fig. 3b).
Data collection
Our within-site sampling strategy was designed to
minimise disturbance to individual plants. Within
each population, plants for study were chosen to
proportionately reflect the range of variation evident
in both morphology and habitat. Vegetative char-
acters were measured non-destructively from in situ
Figure 3 Habitats of Himantoglossum hircinum at Sandwich (d), Newmarket (e), and Ifrane (b), together with (c) plants
invading a garden lawn at Sandwich and (a) a transient ‘rogue’ plant (centre foreground) that flowered on an urban roadside
verge in Isleworth in 2000 and 2011. Images: (a)5B.G. Tattersall, (e)5I. Denholm, remainder5R.M. Bateman.
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 125

plants, and only two flowers from each plant were
removed for further study: the first was permanently
mounted and measured (Fig. 4), whereas the second
was placed in fine-grained dried silica gel to act as a
DNA-friendly voucher. Wherever possible, the floret
chosen to provide morphometric data on the flower,
ovary and bract was located 30–40% of the distance
from the base to the apex of the inflorescence, in
order to minimise the effect of any diminution in
flower size toward the apex.
The 46 characters scored (Table 1) described the
stem and inflorescence (9), leaves (4), labellum (17),
spur (3), lateral petals (2), lateral sepals (9), and
gynostemium (2). They can alternatively be cate-
gorised as metric (31), meristic (5), multistate-scalar
(5), and bistate (5). The colours of the ‘limbs’ of the
labellum, and of the reverse of the sepals, were
matched to the closest colour block(s) of the Royal
Horticultural Society Colour Chart (Anonymous,
1966) and later converted to three quantified vari-
ables recognised by the Commission Internationale
de I’Eclairage. A more detailed account of the chosen
characters and methods of measurement applied to
the genus Himantoglossum s.l. will be given elsewhere
(Sramko´ et al., in prep.).
Data analysis
Data for individual plants were summarised in an Excel
v14.2 spreadsheet. Means, sample standard deviations
and coefficients of variation were calculated for every
character in each of the three populations. Univariate
analyses were summarised and presented using
Deltagraph v5.6 (SPSS/Red Rock software, 2005).
The morphometric matrix contained 23 indivi-
duals646 characters. After combining the numbers
of basal and bracteoidal leaves, and omitting four
characters that varied among other species of
Himantoglossum but were invariant within H. hircinum,
the assembled data were analysed by multivariate
methods using Genstat v11 (Payne et al.,2008).All
calculated ratios were also omitted from the multi-
variate analyses as, by definition, they duplicated their
constituent characters.
The remaining 41 characters were used to compute
a symmetrical matrix that quantified the similarities
of pairs of data sets (i.e. plants) using the Gower
Similarity Coefficient (Gower, 1971) on unweighted
data sets scaled to unit variance. The matrix was in
turn used to construct a minimum spanning tree
(Gower & Ross, 1969) and subsequently to calculate
principal coordinates (Gower, 1966, 1985) – com-
pound vectors that incorporate positively or negatively
correlated characters that are most variable and
therefore potentially diagnostic. Principal coordinates
are especially effective for simultaneously analysing
heterogeneous suites of morphological characters and
can comfortably accommodate missing values; they
have proven invaluable for assessing relationships
among orchid species and populations throughout the
last three decades (reviewed by Bateman, 2001).
Molecular analyses
In addition to the two populations studied morpho-
metrically, a further 11 English localities were
sampled for molecular analysis, encompassing the
entire present geographic range of the species in the
British Isles (Fig. 1); at these additional sites,
observed population size ranged from c.30 flowering
plants to single transient individuals. A further 16
localities were sampled from Continental Europe in
Spain (2 populations), southern France (5), northern
France (3), southwest Germany (4), southern Italy (1),
and Sicily (1); also, two localities were sampled in the
Middle Atlas Mountains of Morocco (Fig. 6). Together,
these samples spanned the full geographical range of
H. hircinum, seeking inter- and infra-population varia-
bility that could be of phylogeographic as well as
taxonomic value (Kay et al., 2006).
DNA was extracted from a portion of each
desiccated flower following the 26CTAB (cetyltri-
methyl ammonium bromide) procedure (Doyle &
Doyle, 1990). The main molecular survey was
conducted by direct-sequencing the complete Internal
Transcribed Spacer assembly (ITS1–5.8S–ITS2) of the
nuclear ribosomal DNA. ITS was amplified by the
plant-specific ITS1A (59-GACGTCGCGAGAAGT-
CCA-3’) primer and the universal primer ITS4 (White
et al., 1990). The PCR mixture contained 0.1 volume
106Taqbufferwith(NH
4
)
2
SO
4
(Fermentas),
200 mM each of dNTPs (Fermentas), 2 mM MgCl
2
,
0.2 mM of each primer, 1.25 U Taq DNA polymerase
(Fermentas), and approximately 5 ng/ml genomic
Figure 4 Silhouettes of mounted flowers of plants measured
for the present morphometric survey from the three study
populations of Himantoglossum hircinum. (a) Sandwich, (b)
Newmarket, (c) Ifrane. Scale bar520 mm.
Bateman et al. Observations on the Lizard Orchid
126 New Journal of Botany 2013 VOL.3 NO.2

DNA extract. Amplifications were performed on a
GeneAmp PCR System 2400 (Perkin Elmer Corp.),
programmed for a denaturation step at 94uC for
4.30 minutes, followed by 33 cycles of denaturation
for 30 seconds at 94uC, annealing for 30 seconds at
51uC, and extension for 30 seconds at 72uC, the
extension time being increased by one second in each
successive cycle; thermal cycling was ended by a final
extension for 7 minutes at 72uC.
Subsampling allowed preparation of between one
and four individuals of H. hircinum from each of 13
localities in the UK and 17 non-UK sites. Direct
bidirectional sequencing of ITS (performed in Korea
by Macrogen) on a total of 46 plants revealed frequent
individuals that showed additive polymorphic sites in
ITS. Hence, one sample from three selected popula-
tions of H. hircinum (Winterbourne, near Bristol in the
UK, Fayence in the Var region of southeast France,
Capizzi in north-central Sicily) was subsequently
cloned, yielding a minimum of six clones (Sramko´ et
al., 2013). Most variable sites were single nucleotide
polymorphisms, but ITS ‘populations’ within five
plants were length polymorphic (i.e. reflected indels
that could be categorised but could not be assigned to
specific ribotypes using Collapse v1.2). The remaining
ribotypes were then inserted into an unrooted phyletic
network generated using TCS v1.21 (Clement et al.,
2000).
These ITS data were then combined with ITS
sequences obtained from all other taxa within
Table 1 Population means, sample standard deviations and coefficients of variation for morphometric characters
measured in the three study populations of Himantoglossum hircinum
Newmarket Sandwich Ifrane
Character Mean SSD CV (%) Mean SSD CV (%) Mean SSD CV (%)
Stem height 36.7 6.6 18 31.7 7.0 22 38.0 6.1 16
Stem diameter 3.95 0.72 18 5.18 0.79 15 7.97 0.55 7
Stem pigmentation 0.5 1.0 0
Inflorescence length 12.0 2.7 23 14.0 3.2 23 14.3 1.2 8
Flower number 27.2 7.2 27 50.9 18.3 36 68.5 NA NA
Bract length lowest 26.0 4.9 19 NA NA NA 45.0 6.2 14
Bract length median 21.0 3.9 19 30.3 6.3 21 27.3 0.6 2
Ovary length 12.7 0.8 6 13.0 1.4 11 12.7 0.6 5
Leaf total number 8.6 1.6 19 10.6 2.1 20 12.7 0.6 5
[Leaf basal number] 4.8 0.8 17 7.9 1.4 18 6.0 1.0 17
[Leaf bract-like number] 3.8 1.0 26 2.7 1.2 44 6.7 0.6 9
Leaf length longest NA NA NA 96 22 22 88 13 14
Leaf width longest NA NA NA 31 6 20 43 14 32
Lip shoulder width 6.6 0.9 13 7.3 0.8 12 7.5 1.6 22
Lip torso width 1.67 0.19 11 1.71 0.13 8 1.83 0.32 18
Lip maximum length 39.4 5.7 15 44.6 5.2 12 49.7 10.2 21
Lip torso length 38.1 5.9 16 43.9 5.3 12 47.3 12.5 26
Lip length to armpit 5.1 0.8 16 4.8 1.0 21 6.8 1.2 17
Lip crenulae number 4.7 1.1 23 4.9 1.4 29 4.7 0.6 13
Lip arm minimum length 6.8 1.6 23 12.9 1.9 15 9.7 2.2 23
Lip arm median width 1.04 0.20 19 1.04 0.16 15 0.97 0.12 12
Lip leg length 1.21 0.70 58 1.47 1.05 71 2.87 2.20 77
Lip leg median width 0.72 0.30 42 0.85 0.15 18 0.60 0.26 43
Lip tail presence 0.1 0 0
Lip margin colour x 374 3 1 400 4 1 415 0 0
Lip margin colour y 364 10 3 364 22 6 348 0 0
Lip margin colour Y 12.8 1.7 13 9.2 2.5 27 13.0 0 0
Lip spot number 14.7 3.6 25 9.0 1.9 21 12.3 1.5 12
Lip spot distribution 1.9 1.3 1.0
Torso versus stem position 2.0 2.0 2.0
Arm versus torso position 4.0 3.8 3.0
Spur length 2.20 0.44 20 2.62 0.20 8 3.07 0.49 16
Spur median width 1.95 0.30 15 2.26 0.22 10 2.37 0.38 16
Spur curvature 4.3 4.0 5.0
Lateral petal length 6.7 0.6 9 7.6 0.4 5 8.8 1.3 15
Lateral petal width 1.39 0.17 12 1.19 0.28 24 1.63 0.12 7
Lateral sepal length 8.9 0.9 11 10.3 0.6 6 11.9 1.5 13
Lateral sepal width 4.50 0.31 7 5.24 0.36 7 4.87 0.57 12
Sepal external colour x 376 3 9 373 0 0 357 0 0
Sepal external colour y 437 5 11 454 0 0 421 0 0
Sepal external colour Y 63.4 1.3 2 33.0 0 0 58.0 0 0
Sepal external margin 1.0 1.0 1.0
Sepal internal lines 1.0 1.0 1.0
Sepal internal dots 0.9 1.0 0
Lateral sepal position 1.0 1.0 1.0
Column length 4.02 0.29 7 3.80 0.18 5 4.23 0.55 13
Column width 2.42 0.19 8 2.62 0.17 7 3.33 0.31 9
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 127

Himantoglossum s.l., plus Steveniella satyrioides as
outgroup, to provide a broad phylogenetic context
for the present study. Phylogenetic tree reconstruc-
tion based on maximum parsimony was performed
using PAUP v4.0b10 (Swofford, 2003), employing
default settings for heuristic search and subsequent
bootstrap analysis with 1000 pseudo-replicates.
Scanning electron microscopy
An entire inflorescence of H. hircinum, presenting
both opening flowers and unopened buds, was
sampled from the seed-derived stock of a private
orchid grower in Gloucestershire, England. Material
was preserved in 70% ethanol before being processed
for scanning electron microscopy (SEM). Buds were
removed from the inflorescence and dehydrated
through an ethanol–water series to 100% ethanol.
Samples were dried in a Tousimis Autosamdri 815B
critical-point dryer (CPD) using carbon dioxide as
the carrier gas. Flowers were mounted onto stubs
using double-sided adhesive discs and dissected under
a Wild Heerbrugg M7A microscope. Partially dis-
sected samples were coated in platinum using an
Emitech K550 sputter coater and imaged using a
Hitachi S-4700 II cold-field emission SEM. For each
bud, multiple images were captured and, where
appropriate, later aggregated into colourised compo-
site reconstructions using Adobe Photoshop.
Results
ITS ribotypes
The ITS phylogeny (Fig. 5) confirmed the topology
for Himantoglossum s.l. (i.e. including the former
genera Comperia and Barlia) derived by Bateman
et al. (2003), and placed the Caucasian endemic
H. formosum in the predicted position as sister to the
remaining taxa of Himantoglossum s.s. (Sramko´ et al.,
2013). However, no reliable phylogenetic structure
was identified within this residual clade, which
identified only single autapomorphic states in ITS
supporting the putative eastern Mediterranean ende-
mics H. montis-tauri and H. galilaeum (Fig. 5).
We therefore chose to view the data as an unrooted
network of closely similar ribotype groups (Fig. 6).
Two ribotype groups were shown by Sramko´ et al.
(2013) to dominate the H. hircinum-adriaticum and
H. jankae-caprinum groups, respectively: RGa1 to
the west of the Bosphorus and RGb1 to the east
(note that, before 2012, H. jankae and H. caprinum
were widely known as H. caprinum and H. affine,
respectively: Molna´r et al., 2012a; Sramko´ et al., 2012).
Of 51 positions variable within the H. hircinum-
caprinum aggregate, 20 varied within H. hircinum;
indeed, H. hircinum yielded more ribotype variants
that its sister species, H. adriaticum,whichwas
similarly dominated by the core ribotype RGa1
(Fig. 7). Most of these variants, including RGa8
found at Masa in the mountains of northern Spain,
deviated from the core ribotype RGa1 by just one
single-nucleotide polymorphism or indel (Fig. 6).
In addition, the La Chapelle population in the
Vercors region of southeast France contained a
length-polymorphic variant (L05), and that from the
nearby Grasse locality in the Var yielded both a
further length-polymorphic variant (L02) and a puta-
tive pseudogene.
Levels of variation similar to those on the
Continent are evident among the UK populations
of H. hircinum (Fig. 7). At least 11 of the 13
sequenced populations contained the dominant wes-
tern European ribotype RGa1, but those from
Winterbourne and Newmarket also contained a
single-base-pair deviant RGa6 and the isolated
plant from the Kent coast at St Margaret’s yielded
another ribotype group, RGa5, which also appears
to be unique to England. In addition, the length-
polymorphic L05 ribotype previously observed in the
Vercors was also found in Royal St George’s golf
course at Sandwich, while sporadic occurrences of
ribotype group RGa3 stretched from Sicily to Hythe
via Calais (Fig. 7a–c).
Figure 6 Parsimony network for ribotype groups identified
in Himantoglossum hircinum. Each group number is pre-
fixed by ‘RGa’. Ribotype RGa1 is the core group; the two
black dots indicate hypothetical (undiscovered) ribotypes.
Figure 5 Parsimony tree of ITS sequences for 40 acces-
sions of Himantoglossum s.l. (numbers per taxon are given
in parentheses; these are identical within each species
except H. robertianum) plus three accessions of Steveniella
as outgroups. All internal branches received strong statis-
tical support (bootstrap support599–100%, posterior prob-
ability51.0) except that arrowed (BS575%, PP#0.5).
Bateman et al. Observations on the Lizard Orchid
128 New Journal of Botany 2013 VOL.3 NO.2

The most substantial ITS divergence was found in
southern Italy and especially Sicily, where the
supposed outlying populations of H. hircinum yielded
two more strongly divergent ribotypes that are
atypical not only of the core distribution of
H. hircinum but also of H. adriaticum, its sister
species that characterises central and northern Italy
and parts of the Balkans (Sramko´ et al., 2013).
Specifically, the Capizzi population from north-
central Sicily was dominated by a more divergent
ribotype RGa4 that differed in three nucleotides, also
incorporating the length-polymorphic group L04 and
group RGa3, which similarly characterised the other
southern Italian population sampled, Calvello.
Surveying the data at the lower hierarchical level of
specific ribotypes rather than ribotype groups showed
some intriguing similarities within the frequent RGa1
category. Ribotype H30 is shared by plants from
Slepe (Dorset, UK), Newmarket (Cambridgeshire,
UK), and Camber (East Sussex, UK), and ribotype
H29 by plants from Berrow (Somerset, UK),
Broadstairs (north-eastern Kent, UK) and Fayence
(the Var region of SE France). Most intriguingly,
ribotype H54 occurs in Guildford (Surrey, UK),
Isleworth (London, UK), Sandwich (eastern Kent,
UK), and Reviers (Normandy, France). However,
these ribotypes are so closely similar that it would be
an error to over-interpret their patterns.
Morphometric comparison
The morphometric matrix of 23 plants646 charac-
ters contained 27 (2.5%) missing values. Of these
absences, 14 represented leaf dimensions of the
Newmarket populations, which were severely desic-
cated by the time of measurement (in contrast, leaves
often persist into the flowering period at Sandwich:
Carey & Farrell, 2002). A further 10 missing values
represented the length of lower bracts at the
Sandwich population – a character that was intro-
duced into the study only after this pioneering dataset
had been acquired. Moreover, within the context of
this particular analysis, four characters proved to be
invariant: the reliable c.45uangle of the labellar
‘torso’ relative to the stem, together with the lateral
Figure 8 Plot of the first two principal coordinates for 23
individuals of Himantoglossum hircinum from three study
populations.
Figure 7 Distribution across western Europe of ITS ribo-
types observed in Himantoglossum hircinum through direct
sequencing; eight groups are delimited by single-nucleotide
polymorphisms and three through length-variable poly-
morphisms (L), together with a putative pseudogene (PG)
(see also Sramko´ et al., 2013). Each spot represents a single
plant; overlapping spots indicate multiple accessions
derived from the same locality. Scale bar is 18 km for Kent
and East Sussex (a), 100 km for England (b), and 500 km for
Western Europe (c). Base maps courtesy of Google Earth.
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 129

sepals being connivent into a hood and bearing
brownish-purple lines as both marginal strips exter-
nally (abaxially) and dashes internally. In addition,
only one study plant (from Newmarket) exhibited a
‘tail’ mid-way between the ‘legs’, and it was less than
1 mm long. Population means, sample standard
deviations and coefficients of variation are given in
Table 1 for each of the study populations.
The resulting plot of the first two principal
coordinates (Fig. 8) encompasses a respectable 56%
of the total variance and readily distinguishes all
individuals of each of the three study populations; the
superimposed minimum spanning tree (MST) in
particular identifies the two English populations as
being more similar to each other than either is to the
Moroccan population. The three populations appear
roughly equally cohesive on the plot, the weaker
MST links within the Ifrane population being
attributable to the smaller number of plants available
for statistical comparison.
When individual characters are considered, several
serve to distinguish Ifrane from the English popula-
tions (Table 1). Vegetatively, the stem is more robust
relative to its height (Figs. 2c and 9) and there is a
stronger disparity in length between the bracts of the
lowermost flowers and bracts occurring higher up the
inflorescence. Turning to labellum shape and size,
Ifrane superficially appears to have longer ‘legs’, but
its higher mean value actually reflects the presence of
just one plant possessing an exceptionally deeply
divided central lobe (Fig. 10). Also, the ‘arms’ of
Ifrane labella tend to be less strongly recurved
relative to the plane of the ‘torso’ (Fig. 2). The
provision of a marginally longer spur appears to
encourage greater downward curvature and the
column is somewhat wider. Most strikingly, the
Ifrane population lacks purple spots within the sepals
(a feature also absent from one sampled plant at
Newmarket), though the Ifrane plants share with the
English populations the presence of purple dashes
(Fig. 2). The Moroccan population generally showed
greater variation in floral dimensions but less varia-
tion in vegetative dimensions compared with the
English populations (Table 1).
Morphological differences distinguishing the two
English study populations are more subtle (Table 1).
The Newmarket population has the narrowest stem
relative to its height (Fig. 8) and is less floriferous. Its
flowers have somewhat narrower lateral sepals but
somewhat broader lateral petals. More strikingly, the
‘arms’ of Sandwich plants are on average longer than
those of Ifrane and double the length of those from
Newmarket (Figs. 2 and 10), a fact that is particularly
evident from representative labellar silhouettes
(Fig. 4) – indeed, the mean ‘arm’ length of Sandwich
plants of H. hircinum matches that of H. jankae from
eastern Europe (Sramko´ et al., 2012). The remain-
ing differences reflect anthocyanin pigmentation. The
upper portions of the stems of plants from Sandwich
are more reliably stained purple than those from
Newmarket (c.40% versus c.90%). Careful colour-
matching showed that the labellar margin of Sandwich
plants is a slightly deeper purplish-brown, and the
outer surfaces of the sepals are a somewhat deeper and
bluer green (Fig. 2). In contrast, the discrete purple
papillate spots tend to be slightly fewer and more
localised in the centres of Sandwich labella.
Floral ontogeny
All of the structures that will eventually constitute a
mature H. hircinum flower c.50 mm long are already
evident in a bud a mere 2 mm long (Fig. 11a). Early
growth favours the gynostemium, whereas later
expansion of the remaining structures (ovary, auricles,
labellar lobes, spur) occurs at approximately equal
rates (cf. Fig. 11a–c). Although there is equivocal
evidence that expansion of the labellar lobes continues
beyond those of other structures, labellum elongation
Figure 10 Plot of ‘arm’ length versus ‘leg’ length for 23
individuals of Himantoglossum hircinum from three study
populations.
Figure 9 Plot of stem diameter versus stem height for 23
individuals of Himantoglossum hircinum from three study
populations. Regression lines and r
2
values are also shown.
Bateman et al. Observations on the Lizard Orchid
130 New Journal of Botany 2013 VOL.3 NO.2

appears to be consistently in advance of that shown by
the equivalent developmental stages earlier divergent
species of Himantoglossum s.l. The famed tightly-
packaged ‘watch-spring’ morphology of the central
labellar lobe is clearly visible even in early stages of
growth, and is later mirrored (albeit in less tight spirals)
by coiling in the lateral lobes. Lastly, the crenulations
that reliably adorn the ‘shoulders’ of the lateral lobes
become evident only comparatively late in ontogeny
(Fig. 11c), presumably resulting from differential
expansion of the labellum margins. Measurements of
individual cells in contrasting developmental stages
suggestthattheenlargementofcontrastingfloral
structures primarily reflects cell division rather than
cell expansion.
Comparison of immature (Fig. 12a) with mature
(Fig. 12b) labella reveals that the perfectly aligned
longitudinal ranks of pavement-style epidermal cells
overlying the mid-vein in early developmental stages
later form a seemingly chaotic melange of highly
papillate cells, also demonstrating that the purple spots
on the labellum (Fig. 2e–g) are localised concentrations
of anthocyanins within a single extensive papillate
region rather than isolated clusters of papillae. The
early stages of circinnation of the labellum are well-
illustrated in Figure 12a, but only the later-stage
Figure 12b features the plicate ‘shoulders’ of the lateral
lobes that characterise the mature flower.
Discussion
Phylogenetic context of H. hircinum
Some authors have in recent years continued to
implicitly or explicitly treat all members of the
H. hircinum-adriaticum and H. jankae-caprinum aggre-
gates as just one or, at most, two species (Sundermann,
1973, 1980; Moore, 1980; Carey & Farrell, 2002; Foley
Figure 11 Artificially coloured, composite scanning electron micrographs depicting a series of three developmental stages (a–c) of
buds excised from a plant of Himantoglossum hircinum cultivated by Richard Manuel. The two lateral petals and all three sepals
have been removed. Mauve5ovary, grey5base of gynostemium, pale yellow5bursicles and connective, dark yellow5auricles,
blue5labellar spur, green5lateral labellar lobes, red5central labellar lob e. Scale bar5500 mm. Images: P.J. Rudall.
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 131

& Clarke, 2005). However, other authors recognised
larger numbers of species (Nelson, 1968; Landwehr,
1977; Teschner, 1980; Delforge, 1999, 2006); current
evidence strongly supports the recognition of up to 12
species within the expanded genus Himantoglossum s.l.
(Sramko´ et al., 2013, in prep.).
Molecular investigations conducted as part of a
broader phylogenetic study of the genus Himanto-
glossum s.l. by Sramko´ et al. (2013) included
five samples of H. hircinum that were selected to
span the full geographic distribution of the species:
Newmarket (England) to the northwest, Habkirchen
(SW Germany) to the northeast, Grasse (SE France)
representing the core distribution, Pacios (NW Spain)
representing the northern Iberian outlier, and Calvello (S
Italy) representing the southern Italian outlier. These
samples were sequenced for the first intron of the low-
copy (and developmentally crucial) nuclear gene LEAFY
and for three rapidly mutating and phylogenetically
informative plastid regions: the accD–psaI intron, the
psbA–trnH intron, and two introns of the trnL–ndhF
region, specifically trnL–rpl32 and rpl32–ndhF.
The tree generated from the four-intron plastid
matrix confidently (98% bootstrap value) placed the
six analysed accessions of H. hircinum as sister to three
accessions of H. adriaticum. Four of the hircinum
accessions (UK, Spain, France, Germany) yielded
identical sequences. Those from southern Italy
(Calvello) and Morocco (Ifrane W) deviated by a
single base-pair in trnL-ndhF, though the Moroccan
accession also exhibited three autapomorphic indels in
the accD–psaI intron. LEAFY sequences also reliably
distinguished H. hircinum from H. adriaticum as a
well-supported clade, but they too showed remarkably
little within-species variation. The Spanish accession
was tentatively placed as sister to the remaining
samples on the basis of just one base-pair difference,
and the accession from Grasse (south-eastern France)
apparently possessed two autapomorphic states.
Unfortunately, we were unable to obtain LEAFY
from the southern Italian populations, as the relevant
DNAs were of insufficient quality.
Thus, the main relevance of the results of the study of
Sramko´et al. (2013) to the present study is to
demonstrate a degree of genetic cohesion to H. hircinum
s.s. as a species – more so than the ITS phylogeny
shown here as Figure 5. Fortunately, ITS revealed
intriguing ribotype variation within the species.
Detailed morphology of H. hircinum in the UK
Given such taxonomic interest, and the appreciable
levels of morphological and genetic diversity demon-
strated within H. hircinum (Figs. 2, 5 and 7), sur-
prisingly few previous authors commented on mor-
phological variation among British populations. The
main exception was Ettlinger (1997), who correctly
observed variations in labellum shape (especially the
length of the ‘arms’ and the depth of the sinus
separating the ‘legs’) and pigmentation, notably
amount of purple spotting in the centre of the labellum
and the depth of the brownish-purple staining around
its periphery. In addition, an anthocyanin-poor
individual from Newmarket illustrated by Harrap &
Harrap (2009, p. 353) vaguely echoes the character-
istically unmarked and/or pale-flowered species of
Figure 12 Scanning electron micrographs of (a) an entire immature labellum and (b) the strongly papillate central region of a
mature labellum of a bud and flower respectively excised from a plant of Himantoglossum hircinum cultivated by Richard
Manuel. Scale bars5100 mm. Images: P.J. Rudall.
Bateman et al. Observations on the Lizard Orchid
132 New Journal of Botany 2013 VOL.3 NO.2

Himantoglossum s.s. that occur around the eastern
Mediterranean (Sramko´ et al., 2012, 2013).
When measured against our detailed morpho-
metric data (albeit representing only three popula-
tions: Table 1), descriptions of H. hircinum in the
literature have generally proven to be more accurate
than those of most other species of Orchidinae
(Davies et al., 1983; Sell & Murrell, 1996; Carey &
Farrell, 2002; Delforge, 2006; Harrap & Harrap,
2009). Nonetheless, inaccuracies are evident. An
overly short range of 7–10 mm attributed to the
sepals by Moore (1980) was reproduced in several
subsequent descriptions (Sell & Murrell, 1996; Carey
& Farrell, 2002; Foley & Clarke, 2005). A typographic
error in Sell & Murrell (1996) permitted labella to be as
little as 3 mm long! Some authors gave ranges of
overall labellum length that fail to encompass the
shorter end of the observed spectrum (Lang, 2004;
Stace, 2010) whereas others neglect the longer end of
the spectrum (Foley & Clarke, 2005). Similarly, some
authors attempted to restrict spur length to less than
2.5 mm (Davies et al., 1983; Foley & Clarke, 2005)
when it actually averages 2.7 mm and can reach
3.5 mm. Other authors under-estimated either the
length of the ovary (Sell & Murrell, 1996; Carey &
Farrell, 2002) or the width of the bracts (Carey &
Farrell, 2002); yet others over-estimated the propor-
tion of total plant height occupied by the inflorescence
(Harrap & Harrap, 2009). Some authors offered as a
diagnostic character the presence of a notch at the
apex of the central lobe (Delforge, 2006; Stace, 2010)
but ignored the c.10% of plants that lack this notch.
Lastly, Stace (2010, p. 880) noted that the leaves may
be ‘purple-mottled’, a statement that presumably was
intended to instead refer to the upper portion of the
stem; purple anthocyanins suffuse the stem within and
immediately below the inflorescence in c.40% of
individuals, but we have never seen these vegetative
pigments extend into the leaves of H. hircinum.
It is, of course, likely that our own limited
morphometric measurements have failed to capture
the full spectrum of morphology presented by
H. hircinum, as well as having inevitably been modified
by ontogenetic and ecophenotypic factors (cf.
Bateman & Denholm, 1989). For example, plants
measured by us previously at Sandwich in dune-slacks
located further from the beach were taller than the
3267cm (n510) reported here, averaging 47 cm in
1979 (n539) and 41 cm in 1981 (n527); similar
fluctuations in plant height between years were
reported in a Hungarian population of H. adriaticum
by Bo´ dis & Molna´ r (2009). The largest plant of
H. hircinum that we have encountered in the UK – an
isolated individual that appeared briefly at Headley
Warren, Surrey – measured 100 cm and bore 66
flowers in 1979 (Fig. 2b). Even larger plants, carrying
up to 200 flowers, occur occasionally in both the UK
and the Mediterranean region (Carey & Farrell, 2002;
Pfeifer et al., 2009).
Contribution of H. hircinum to examples of
phenotypic convergence
The pollinaria of Himantoglossum s.s. are distinguished
by unusually robust columnar caudicles terminating in
more-or-less hexagonal viscidia that are enclosed in a
desiccation-resistant hemispherical bursicle and are
fused laterally, such that both pollinia can only be
transferred by pollinators as a single pollinarium;
nonetheless, deposition on the stigma of pollinium
fragments is sufficient to ensure adequate pollination
(Carey & Farrell, 2002). Such lateral fusion of viscidia
shows considerable parallelism within the phylogeny of
Orchidinae (we are intrigued to know whether it is
congenital or postgenital in Himantoglossum, but
answering this question would require access to even
earlier stages of floral ontogeny). As well as character-
ising all species of Himantoglossum s.l. other than the
basally divergent H. (formerly Comperia)comperianum
(cf. Delforge, 1999; Claessens & Kleynen, 2011), fused
viscidia are also evident in Orchis (formerly Aceras)
anthropophora,Anacamptis pyramidalis, and all species
of Serapias. Together, these taxa constitute a remark-
able case of morphological parallelism.
In addition, the typical examples of flowers from our
study populations illustrated in Figure 4 show that the
labellum of H. hircinum is intermediate in outline shape
between those of the mid-Mediterranean H. adriaticum
and eastern European H. caprinum (formerly H. affine)
(Nelson, 1968; Delforge, 1999; silhouettes reproduced
in Davies et al., 1983, p. 137; Delforge, 2006, p. 351).
However, molecular data suggest that an origin of
H. hircinum through hybridisation of H. adriaticum
and H. caprinum is highly unlikely, thus indicating that
labellum outline is also subject to morphological
convergence within the genus.
A chromatographic (HPLC) analysis by Strack et al.
(1989) of the floral anthocyanins of H. adriaticum,sister
to H. hircinum, revealed a significant percentage of
unknown compounds but a spectrum of identifiable
pigments that was dominated by Serapianin and Seranin.
These pigments also characterise Himantoglossum sub-
genus Barlia,Serapias and Anacamptis papilionacea (a
species that is to some degree also convergent on
Serapias in floral morphology). Moreover, the somewhat
denser floral pigments of the Sandwich population of H.
hircinum relative to the Newmarket population (cf.
Fig. 2f and g) may simply reflect the epigenetic influence
of the substrate in which they are rooted. Species of other
genera of Orchideae, notably marsh-orchids of the genus
Dactylorhiza, also routinely generate richer colours
(presumably reflecting greater concentrations of antho-
cyanins) when growing in dune sands (Bateman &
Denholm, 1985).
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 133

Evolutionary-developmental aspects of floral
ontogeny
The early stages of floral development in H. hircinum
(Figs. 11 and 12) broadly follow those documented
for other genera of Orchideae (Kurzweil, 1987; Box
et al., 2008; Rudall et al., 2013). The greater size of
the labellum of the H. hircinum-jankae clade sensu
Sramko´et al. (2013) does not simply represent
giantism, as changes in its dimensions are non-
allometric; the central lobe increases greatly in length,
whereas its width decreases relative to those of sister-
species H. formosum and H. robertianum. Thus, a
shape change accompanies the size increase, showing
that these changes can be attributed to the category
of heterochrony known as peramorphy. We cannot
entirely rule out some modest effects from precocious
onset of labellum extension (pre-displacement sensu
Alberch et al., 1979; see also Box et al., 2008) or
delayed offset of growth (hypermorphosis), but the
buds illustrated in Figure 10 suggest that the main
contributor to the evolutionary elongation of the
labellum is greatly increased growth rate (accelera-
tion). We assume that growth rate is especially great
on the ‘shoulders’ of the labellum, thereby generating
the distinctive crenulations (Fig. 12b).
The final event in the ontogeny of the flowers is the
uncoiling at anthesis of the elongate labellum and the
immediate replacement of the planar coiling of the
central lobe with a spiral twist (Fig. 2d). The direction
of spiral (chirality s.l.) was coded as a morphometric
character in the present study but ultimately failed to
contribute to the multivariate analyses, as it rapidly
became clear that all individuals of all species of
Himantoglossum s.s. spiral counter-clockwise as viewed
from the apex (Fig. 2). This consistent result presents a
stark contrast to the helical inflorescences of orchids
such as the Autumn Ladies-tresses (Spiranthes spiralis);
large populations of this species maintain approxi-
mately equal numbers of plants with inflorescences that
spiral in clockwise and counter-clockwise directions,
reportedly encouraging allogamy via pollinating bees
(Schilthuizen & Gravendeel, 2012; also B. Gravendeel,
pers. comm., 2011; M. Mehrink, pers. comm., 2011).
However, the helix of the Himantoglossum labellum
does mirror the consistent direction of spiral in the
lateral sepals of Cypripedium species reported by Welch
(1998).
Pollination biology
In their comprehensive summary of pollinators of
European orchid species, Claessens & Kleynen (2011)
reported that 13 bee species (seven belonging to the
genus Andrena: see also Teschner, 1980; Vo¨th, 1990;
Kropf & Renner, 2008) and one species of Oedemera
beetle have been observed removing pollinaria from
plants of H. hircinum. Davies et al. (1983) and Carey
& Farrell (2002) suggested that flies or hoverflies may
also act as pollinators; certainly, taxonomically
broader spectra of pollinators have been recorded
in other species of Himantoglossum (Delforge, 2006;
Claessens & Kleynen, 2011).
The pungent goat-like scent of the flowers, pre-
sumed to be a pollinator attractant, was initially
ascribed to capronic acid (Kerner, 1891; Schmid,
1912), but a subsequent study combining gas chroma-
tography with mass spectrometry revealed the scent to
consist of two forms of decenoic acid plus lauric acid
(Kaiser, 1993). The prominent, dark purple papillate
spots adorning the ‘body’ of the labellum below the
stigma have been said to operate as scent-secreting
osmophores (Vogel, 1990). Certainly, our observa-
tions that the papillae are large, extensive and late-
formed during ontogeny (Fig. 11) are all consistent
with a likely function as osmophores.
A debate begun by Darwin (1877) regarding
whether H. hircinum offers a nectar reward or
alternatively operates entirely by food deceit is
ongoing even today (Bateman, 2012). Evidently
desperate to rebut the food-deceit theories of Fritz
Mu¨ ller, Charles Darwin speculated that a modest
nectar reward could be obtained by insects by
penetrating the tissues of the spur, an interpretation
later discussed by Teschner (1980). In addition, Vo¨th
(1990) argued that the dense papillae that partially
obscure the spur entrance may secrete tiny drops of
nectar, and several subsequent authors have also at
least tentatively inferred the presence of a small
nectar reward in H. hircinum (Neiland et al., 2001;
Bourne´rias & Prat, 2005; Inda et al., 2012). In truth,
such allegations could be levelled against any food-
deceitful species of Orchidinae; there is little doubt
that pollinators of Himantoglossum species receive no
meaningful reward (Carey & Farrell, 2002; Kropf &
Renner, 2008; Claessens & Kleynen, 2011; Bateman,
2012).
Similarly, there is little evidence of autogamy in
H. hircinum, either through direct observation or
through exploration of population-genetic structure.
However, the occurrence of at least some geitono-
gamy (cross-pollination of genetically identical flow-
ers occupying the same inflorescence) was confirmed
long ago by the fact that even solitary flowering
plants often set seed (Summerhayes, 1951). More
recently, detailed observations by Kropf & Renner
(2008) estimated a geitonogamy frequency of 36% in
three populations of Lizard Orchid in southern
Germany. As concluded by Kropf & Renner (2006,
p. 506), ‘foraging behaviour results in a mix of
geitonogamy, near-neighbour pollination, and occa-
sional long-distance outcrossing not fundamentally
different from the situation in many other insect-
pollinated perfect-flowered rewarding angiosperms’.
We suspect that geitonogamy frequencies of about
Bateman et al. Observations on the Lizard Orchid
134 New Journal of Botany 2013 VOL.3 NO.2

one-third are typical of most allogamous species of
Orchideae.
Life history
Harrap & Harrap (2009) noted an average seed-set of
c.30% in UK populations, though figures differ
considerably between years (Carey, 1999; Neiland
et al., 2001; Carey & Farrell, 2002; Carey et al., 2002;
see also data on H. adriaticum by Bo´ dis & Molna´r,
2009). Claessens & Kleynen (2011) reported that 18–
62% of capsules set seed in 12 German populations,
but fruit set was lower (6–7%) in the German
population studied in detail by Pfeifer (2004). No
seeds formed in a Dutch population examined by R.
Wielinga, whereas seed-set reached 95% in one Italian
population studied by Pfeifer et al. (2009). Individual
capsules have been estimated to average just 1200
seeds (Summerhayes, 1951; Carey, 1999). This is a
small number relative to most other species of
Orchidinae; given the comparatively large size of
the capsules and the comparatively small size of the
seed (testa5c.3406120 mm), Carey & Farrell (2002)
may have been right to suggest that typical seed
numbers are higher.
Seed matures within 6–8 weeks of pollination
(Farrell & Carey, 1999) and often germinates immedi-
ately, though periods of seed dormancy of up to
three years have also been documented (Carey &
Farrell, 2002). Germinated seeds typically generate a
first leaf three years later and reach flowering size in a
further three to five years (Rasmussen, 1995; Pfeifer,
2004; Pfeifer et al., 2006b; Harrap & Harrap, 2009). The
resulting leaf rosettes are wintergreen (Carey, 1999;
Carey & Farrell, 2002), allowing the plant to grow
rapidly in the spring but often leading to senescence of
the leaves by the time that the flowers open – a common
feature of winter-green species of Orchideae. Pfeifer
(2004) reported a half-life of approximately five years (a
figure typical of tuberous European orchids), the life
expectancies of individuals being increased by periods
of flowering and/or dormancy. Individual plants are
known to have survived for at least 19 years (Harrap &
Harrap, 2009).
Seedling mortality is high and sexual reproduction
is sporadic. The likelihood of a particular plant
flowering is reportedly increased by a preceding wet
autumn and warm, wet winter free of sharp frosts
(Pfeifer, 2004; Carey, 1996, 1998, 1999; Carey &
Farrell, 2002; Carey et al., 2002; Pfeifer et al., 2006a,
b, 2010), whereas spring droughts and/or frosts can
result in the abortion of inflorescences. Putative
climatic influences are discussed at greater length
under ‘Evidence for climatic drivers’.
Genetic diversity and putative migration routes
Himantoglossum hircinum supposedly colonised wes-
tern Europe only after the last glaciation (Delforge,
2006) – in other words, within the last 11 500 years.
The separation of the eastern H. jankae-caprinum
group (sensu Molna´r et al., 2012a) from the western
H. hircinum-adriaticum group is tentatively estimated
via the molecular clock rationale to have occurred
much longer ago, at about 600 kaBP (Sramko´ et al.,
2013) – that is, in the middle of the Quaternary
period. This date carries huge error bars and so
should be treated with great caution. If interpreted
literally, however, it would place the time of species
divergence within the Cromerian interglacial period,
which preceded three further glacial–interglacial
cycles. Presumably, the separation of H. adriaticum
from H. hircinum occurred appreciably more recently.
Such observations suggest that, at least in theory,
H. hircinum presents a good opportunity to use
population genetic data to study phylogeography –
specifically, migration routes that reflect contractions
to, and/or expansions from, glacial refugia.
A pioneering AFLP study of three isolated popula-
tions no more than 10 km apart in east-central
Germany, marking the northeastern limit of the
species’ range (Heinrich & Voelckel, 1999; Heinrich,
2003), revealed greater variation within populations
than between them (Pfeifer, 2004; Pfeifer & Jetschke,
2006). The results suggest that, at least in this region,
H. hircinum is freely interbreeding at a local scale.
When analysed at the level of individual plants, their
data reliably distinguished one population from the
other two, but the most localised of the two remaining
populations was nested within the third, indicating a
source–founder relationship between them.
Further developing this AFLP study, Pfeifer et al.
(2009) later used both AFLP markers and plastid
microsatellites to compare 20 populations (c.200
plants) of H. hircinum distributed across much of
Western Europe. Little pattern was evident in the
AFLP data, beyond greater divergence among
western populations relative to those located further
east. However, although plastid sequencing yielded
only five haplotypes, they nonetheless tentatively
suggested intriguing phylogeographic patterns.
Having identified (albeit with limited justification)
southern France as the species’ centre of post-glacial
migration, the authors argued that the ‘fitness’ of
populations (as crudely measured primarily by the
approximate proportion of individuals in flower in
the summer of 2007) decreased outwards in all
directions from this region towards the margins of
its geographic distribution. They also inferred the
existence of two additional glacial refugia, in south-
ern Italy and southern Spain, and three migratory
pathways: westward into Iberia, northwestward into
England, and northeastward into southern Germany.
Pfeifer et al. (2009) also suggested that English
populations of H. hircinum may reflect northward
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 135

migration from Iberia, as they share a predominance of
their plastid haplotype 1. However, this assertion was
based on analysis of only two English populations:
‘Burnham’ (5our Berrow) and Sandwich. Our ITS
ribotype data (Fig. 6) from 13 English populations
better support Pfeifer et al.’s alternative hypothesis that
the English lineage(s) migrated northwestward, possi-
bly even originating from Pfeifer et al.’s ‘core’ area in
southeast France. The presence of category RGa3
ribotypes in southern Italy, Calais and 55 km across
the English Channel in Hythe, Kent (Fig. 7a) is
intriguing, as is the greater diversity of ribotypes within
England along the Kent coast – a strip of land
especially well-placed to receive seed from the frequent
populations of H. hircinum distributed along the North
French coast. Statistically, the diversity of ribotypes in
general and ‘endemic’ ribotypes in particular is equal in
England and Continental Europe, arguing against very
recent colonisation of England (and mirroring similar
patterns of diversity documented in the biogeographi-
cally coincident orchid Ophrys fuciflora by Devey et al.,
2009).
Lastly, we found no evidence to support the
suggestion of Pfeifer et al. (2009, p. 2362) that
H. hircinum occupied a fourth glacial refugium in the
Balkans, but neither did we find evidence in support
of our a priori theory that the Atlas Mountains of
North Africa could have operated as a refugium;
migration to the Atlas Mountains from Iberia
appears more likely.
Evidence for climatic drivers of northward
migration
The recent decrease in the official conservation status
of the Lizard Orchid in the UK from Vulnerable to
Near-Threatened seemingly flies in the face of its
listing by Kull & Hutchings (2006) as one of the four
orchid species to have declined most rapidly in the
British Isles between the national botanical survey
conducted in 1930–1969 and the subsequent survey
conducted in 1987–1999 (yielding an estimated 83%
decline in hectad occurrences). However, much of the
range expansion that was observed in the 1920s and
1930s involved sporadic and transient occurrences of
single plants. Moreover, this expansionist phenom-
enon has recurred since the late 1990s (Fig. 1), the
most recent errant plant emerging in 2009 on a
remarkably unassuming urban roadside verge in west
London convenient for the H28 bus (Fig. 3a). When
happy with its immediate environment, H. hircinum
shows invasive tendencies throughout its range,
sometimes vigorously occupying roadside verges
and garden lawns in the UK (Fig. 3b).
In England, the expansionist pulse that occurred
during the 1920s increased population numbers from
five to 30z(Good, 1936; Carey, 1999), and that in
the 1990s doubled population numbers from a
previous medium-term average of about ten (Carey,
1999; Farrell & Carey, 1999). Although Foley &
Clarke (2005) argued that about 90% of occurrences
in the UK have been single flowering plants, the
historical data accumulated by Carey (1999) suggest
that this figure is exaggerated. The now sizeable
Sandwich population is inferred to have spawned
several satellite populations within 15 km radius
during the late 1990s (Carey et al., 2002). Such
‘embryonic’ populations are obviously very vulner-
able to rapid extirpation, not least through botanical
collecting; Carey (1999) estimated that 20% of UK
populations had been lost to collectors.
Himantoglossum hircinum has benefited from two
detailed programmes designed to appraise in detail its
responses to perceived climate change, the first
spearheaded by Peter Carey and the second by
Marion Pfeifer. To summarise briefly a vast panoply
of data, the probability of flowering is reportedly
increased by a preceding wet autumn and warm, wet,
frost-free winter – in other words, by the imposition
of a broadly Mediterranean climate (Carey, 1996,
1998, 1999; Carey & Farrell, 2002; Carey et al., 2002;
Pfeifer, 2004; Pfeifer et al., 2006a, b, 2010; see also
Kropf & Erz, 1996), whereas spring droughts and/or
frosts are detrimental to flowering – conclusions
largely congruent with those drawn from largely
anecdotal evidence much earlier by Good (1936). It is
worth noting that conditions optimal for the pro-
gressive growth of seedlings to reach flowering size (a
particular challenge to H. hircinum according to the
above authors) may not equate with those that are
optimal for prompting flowering per se. For example,
one author argued that unusually high summer
rainfall aids seedlings but represents a threat to more
mature tubers. Locally, the interaction of microcli-
mate with soil type appears to be of greatest
consequence to the health of populations.
In south-central German populations, the combina-
tion of plant size and weather conditions has been
estimated to explain 82% of the observed major
fluctuations in numbers of flowering plants within
particular populations (Pfeifer, 2004). The proportion
of flowering plants in one substantial population
averaged 5% and rarely exceeded 10% in any one year.
Moreover, c.70% of individuals flowered only once
during the 25-year survey, suggesting that producing
such large inflorescences is resource intensive, and
helping to explain the transient occurrences of single
plants close to the range margin of the species. It
may be pertinent that three of the ten plants measured
by us at Newmarket possessed only the seven leaves
shown by Pfeifer (2004) to be the minimum number
required to provide a probability exceeding 30% of
flowering being initiated during the forthcoming
season.
Bateman et al. Observations on the Lizard Orchid
136 New Journal of Botany 2013 VOL.3 NO.2

Expanding the geographical coverage of their
population sampling, Pfeifer et al. (2010) suggested
that front-edge and rear-edge populations of a
sweeping climatically-driven migration experience
significantly different demographic influences and
consequently require significantly different conserva-
tion strategies. These insights also have a broader
significance, because they could permit the develop-
ment of demographic models that have the potential
to become predictive if they can be rendered
sufficiently sophisticated (Carey & Brown, 1994;
Carey & Farrell, 2002; Pfeifer et al., 2006a).
Even if the distribution of H. hircinum and its
relatives is indeed particularly responsive to climate
change, their phenology appears to us to be more
robust. In Hungary at least, H. adriaticum – the sister
species of H. hircinum – has proven to be less
susceptible than most species of Orchidinae to
phenological shifts in presumed responses to climate
change (Molna´r et al., 2012b). In this context, it is
interesting to note that the typical altitude of
H. hircinum populations appears to increase with
decreasing latitude. No population in the UK exceeds
200 m a.s.l. (Carey & Farrell, 2002), whereas the
populations studied by Pfeifer et al. (2009) in France
and Germany occurred at 200–800 m a.s.l. and those
in the southern Italian and Spanish outliers at (700–
)900–1150 m a.s.l. Occurring as high as 1650 m a.s.l.,
the Moroccan population studied by us approached
the maximum altitude attained by H. hircinum
(1800 m a.s.l. according to Delforge, 2006); moreover,
two further small populations of H. hircinum exam-
ined by the authors in the Ifrane–Azrou region of
Morocco similarly occurred at 1620 and 1680 m a.s.l.
Such altitudinal responsiveness could be taken as
further (albeit circumstantial) evidence of the sensitiv-
ity of the species to its environmental conditions; even
at a latitude as low as 33uN, at least some populations
of H. hircinum do not reach peak flowering until early
June.
Lastly, the suggestion of Carey (1999) and Pfeifer
(2004) that the fact that individuals of H. hircinum
occur in clumped aggregates with metapopulations is
due to seeds typically falling close to the seed-parents
could instead reflect a need for the seeds to share their
parents’ mycorrhizal partners (Carey & Farrell,
2002). Under this hypothesis, intensive filtration as
a result of mycorrhizal specificity would prevent most
attempts at medium- and long-distance dispersal by
airborne seeds (Bateman, 2006). This hypothesis is
seemingly rendered less likely by a comparatively
early survey of mycorrhizae in H. hircinum that
reported 15 strains of endomycorrhizae in the roots
(Ga¨umann et al., 1961), thus suggesting that this
species may be a generalist when forming subterra-
nean symbioses. However, Carey & Farrell (2002)
later questioned whether the isolated fungi played a
genuinely mutualistic role with the orchid. Further
study of the mycorrhizae of H. hircinum using
more modern approaches is urgently required as an
essential contribution to any interpretation of cli-
matic influences on its distribution and migration.
Possibility that cryptic species exist within H.
hircinum
Returning to the accumulated molecular data, recent
analyses suggest that there is an approximately equal
probability that H. hircinum and H. adriaticum diverged
in western or central Europe (Sramko´et al., 2013). The
reliable genetic cohesion of H. hircinum evident from
plastid and low-copy nuclear LEAFY sequences suggests
that the inability to confidently distinguish H. hircinum
from other species of Himantoglossum s.s. using ITS
sequences is more likely to reflect incomplete lineage
sorting than more recent hybridisation (Sramko´et al.,
2013). Considered together, these phylogenetic patterns
suggest that H. hircinum hasexistedforlongerthanat
least some of the other species of Himantoglossum s.s.
that occur further east. These results encourage the belief
that contentious division through the last c.35 years of
the former H. hircinum s.l. into several species is largely
justified, but they also raise an additional question; do
multiple cryptic species (Bateman, 2011; Bernardo, 2011)
still exist within H. hircinum s.s.?
The chromosome number that is by far the most
frequently reported in Himantoglossum s.l. is 2n536
(reviewed by Pridgeon et al., 1997; Neiland et al., 2001;
Bateman et al., 2003) – the diploid number that typifies
the multi-genus ‘2n536’ clade of which Himantoglossum
s.l. is the earliest divergent member (Pridgeon et al.,
1997; Bateman et al., 2003). However, reports of
chromosome numbers in H. hircinum initially appeared
conflicting; most authors stated that 2n536 (Bourne´rias
& Prat, 2005; Delforge, 2006; Claessens & Kleynen,
2011), whereas others stated that 2n524 (Sell &
Murrell, 1996; Foley & Clarke, 2005), perpetuating a
count first published by Heusser (1915). Yet others
hedged their bets by offering 2n524, 36 or n512, 18
(Moore, 1980; Bianco et al., 1987; Neiland et al.,
2001; Carey & Farrell, 2002; Stace, 2010). A more
detailed examination of 12 Italian karyotypes of H.
adriaticum by D’Emerico et al. (1993) suggested the
additional presence of a B chromosome in one of the
plants and revealed the majority of the chromosomes
to be comparatively symmetrical, heterochromatin-
poor metacentrics.
More recent chromosomal studies have provided
greater detail and insight. Six assessments presently
listed on the IPCN chromosome database give values
of 2n536 (cf. D’Emerico et al., 1990) but include an
additional assessment from Salamanca, Spain of
2n518 (Bernardos & Amich, 2002). Interestingly,
genome size (rather than chromosome number)
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 137

determinations of two plants of H. hircinum yielded
values of C56.1 and 12.2 pg (Leitch et al., 2009), thus
similarly suggesting the existence within H. hircinum
s.s. of two ploidy levels. In contrast, no further
evidence has come to light of plants with the less
credible complement of 2n524.
Particular uncertainty hangs over the geographical
outlier in southern Italy and Sicily of populations that
today are widely attributed to H. hircinum despite
being separated from its distributional centre in France
by a swathe of populations attributed to its sister-
species, H. adriaticum. Firstly, our ITS ribotype data
show the presence of a unique and comparatively
divergent ribotype group in Sicily (Ga4: Fig. 7),
whereas ribotypes found in adjacent populations of
H. adriaticum correspond with the Ga1 ribotypes
that characterise the majority of populations of H.
hircinum elsewhere in Europe. Moreover, the sole
southern Italian locality studied by Pfeifer et al.
(2009), approximating our Calvello locality, was the
only population to be dominated by their plastid
haplotype 4. This insight reinforces the unique ITS
ribotypes observed by us in the same geographical
outlier, rather than supporting the suggestion of
Pfeifer et al. (2009) that haplotype 4 may have been
acquired through hybridisation with H. adriaticum
populations now contiguous with the northern
margin of this outlier. Thus, at present, we regard
as questionable the attribution of southern Italian
populations to H. hircinum s.s.
Denser sampling of populations of both putative
species in southern Italy is clearly warranted,
combined with acquisition of morphometric data
from the region (sadly, no such data are available at
present). It seems likely that microsatellites currently
being developed by Sramko´ and colleagues will help
to solve this and other riddles still posed by
Himantoglossum s.l.
Conclusions
Data from studies of population demographics (Carey,
1999; Pfeifer et al., 2006a) and pollination (summarised
by Claessens & Kleynen, 2011) are in accord with our
genetic data in suggesting that the Lizard Orchid is
typical of tuberous-rooted, non-pseudocopulatory
species of European Orchideae. In short, the ontoge-
netic pattern and mature size of the hyper-elongated
central labellar lobe and crenulated lateral lobes of
Himantoglossum s.s. appear in retrospect to be its most
remarkable feature, though detailed surveys of ploidy
levels via flow cytometry (e.g. compared with the study
of Gymnadenia by Tra´vnı´c
ˇek et al., 2012) and of
associated mycorrhizae (e.g. compared with the study
of Dactylorhiza by Jacquemyn et al., 2012) might yet
prove instructive. Irrespective of particular theories
of plant migration in response to climate change,
research conducted on H. hircinum during the last
two decades has clearly demonstrated the exceptional
value of conducting relatively long-term monitor-
ing experiments (Pfeifer, 2004; Bateman, 2011, 2012).
Nonetheless, achieving less ambiguous interpretations
will evidently require in-depth investigations of many
more plant species, adding to the pioneering work
already conducted on H. hircinum.
Acknowledgements
We thank Richard Manuel for providing an ontoge-
netic series of flowers from non-native, seed-grown
plants in his garden; Mark Spencer for drawing the
author’s attention to the presence of the rogue plant
in Isleworth; and Diana Hall and John Winterbottom
for field advice. We also thank Barry Tattersall, Ian
Denholm, and the late Derek Turner Ettlinger for the
loan of images.
References
Alberch, P., Gould, S. J., Oster, G. F. & Wake, D. B. 1979. Size
and shape in ontogeny and phylogeny. Paleobiology, 5: 296–
317.
Anonymous. 1966. Royal Horticultural Society Colour Chart.
London.
Bateman, R. M. 2001. Evolution and classification of European
orchids: insights from molecular and morphological charac-
ters. Journal Europa¨ ischer Orchideen, 33: 33–119.
Bateman, R. M. 2006. How many orchid species are currently
native to the British Isles? In: Bailey, J. P. & Ellis, R. G., eds.
Contributions to taxonomic research on the British and European
flora, pp. 89–110zPlate 1. London: Botanical Society of the
British Isles.
Bateman, R. M. 2011. The perils of addressing long-term
challenges in a short-term world: making descriptive taxonomy
predictive. In: Hodkinson, T. R., Jones, M. B., Waldren, S. &
Parnell, J. A. N., eds. Climate change, ecology and systematics,
pp. 67–95. Cambridge: Cambridge University Press.
Bateman, R. M. 2012. Circumscribing species in the European
orchid flora: multiple datasets interpreted in the context of
speciation mechanisms. Berichte aus den Arbeitskreisen
Heimische Orchideen, 28: 160–212.
Bateman, R. M. & Denholm, I. 1985. A reappraisal of the British
and Irish dactylorchids, 2. The diploid marsh-orchids.
Watsonia, 15: 321–355.
Bateman, R. M. & Denholm, I. 1989. A reappraisal of the British
and Irish dactylorchids, 3. The spotted-orchids. Watsonia, 17:
319–349.
Bateman, R. M., Hollingsworth, P. M., Preston, J., Luo, Y.-B.,
Pridgeon, A. M. & Chase, M. W. 2003. Molecular phyloge-
netics and evolution of Orchidinae and selected Habenariinae
(Orchidaceae). Botanical Journal of the Linnean Society, 142:
1–40.
Bernardo, J. 2011. A critical appraisal of the meaning and
diagnosability of cryptic evolutionary diversity, and its
implications for conservation in the face of climate change.
In: Hodkinson, T. R., Jones, M. B., Waldren, S. & Parnell, J.
A. N., eds. Climate change, ecology and systematics, pp. 380–
438. Cambridge: Cambridge University Press.
Bernardos, S. & Amich, F. 2002. Karyological, taxonomic and
chorological notes on the Orchidaceae of the central-western
Iberian Peninsula. Belgian Journal of Botany, 135: 76–87.
Bianco, P., Medgali, A. M., D’Emerico, S. & Ruggiero, L. 1987.
Numerici chromosomici per la Flora Italiana. Informatore
Botanico Italiano, 19: 322–332.
Bo´dis,J.&Molna´ r, E. 2009. Long-term monitoring of
Himantoglossum adriaticum H. Baumann population in
Keszthely Hills, Hungary. Natura Somogyiensis, 15: 27–40.
Bourne´rias, M. & Prat, D., eds. 2005. Les Orchide´es de France,
Belgique et Luxembourg, 2nd ed. Biotope: Meze´.
Box, M. S., Bateman, R. M., Glover, B. J. & Rudall, P. J. 2008.
Floral ontogenetic evidence of repeated speciation via paedo-
Bateman et al. Observations on the Lizard Orchid
138 New Journal of Botany 2013 VOL.3 NO.2

morphosis in subtribe Orchidinae (Orchidaceae). Botanical
Journal of the Linnean Society, 157: 429–454.
Carey, P. D. 1996. DISPERSE: a cellular automaton for predicting
the distribution of species in a changed climate. Global Ecology
and Biogeography Letters, 5: 217–226.
Carey, P. D. 1998. Modelling the spread of Himantoglossum
hircinum (L.) Spreng. at a site in the south of England.
Botanical Journal of the Linnean Society, 126: 159–172.
Carey, P. D. 1999. Changes in the distribution and abundance of
Himantoglossum hircinum (L.) Sprengel (Orchidaceae) over the
last 100 years. Watsonia, 22: 353–364.
Carey, P. D. & Brown, N. J. 1994. The use of GIS to identify sites
that will become suitable for a rare orchid, Himantoglossum
hircinum L., in a future changed climate. Biodiversity Letters,2:
117–123.
Carey, P. D. & Farrell, L. 2002. Himantoglossum hircinum (L.)
Sprengel (Biological flora of the British Isles, 641.1). Journal of
Ecology, 90: 206–218.
Carey, P. D., Farrell, L. & Stewart, N. F. 2002. The sudden
increase in the abundance of Himantoglossum hircinum in
England in the past decade and what has caused it. In:
Kindlmann, P., Willems, J. H. & Whigham, D., eds. Trends and
fluctuations and underlying mechanisms in terrestrial orchid
populations, pp. 187–208.
Cheffings, C. M. & Farrell, L. 2005. The vascular plant Red Data
List for Great Britain. Species Status, 7: 1–116. Peterborough:
JNCC.
Claessens, J. & Kleynen, J. 2011. The flower of the European orchid:
form and function. Published by the authors.
Clement, M., Posada, D. & Crandall, K. A. 2000. TCS: a computer
program to estimate gene genealogies. Molecular Ecology,9:
487–494.
Darwin, C. 1877. On the various contrivances by which British and
foreign orchids are fertilised by insects, and on the good effects of
intercrossing, 2nd ed. London: A. & C. Black.
Davies, P., Davies, J. & Huxley, A. 1983. Wild orchids of Britain
and Europe. London: Chatto & Windus.
Delforge, P. 1999. Contribution taxonomique et nomenclaturale au
genere Himantoglossum (Orchidaceae). Naturalistes Belges, 80:
387–408.
Delforge, P. 2006. Orchids of Europe, North Africa and the Middle
East. London: A. & C. Black.
D’Emerico, S. Bianco, P. & Medagli, P. 1993. Cytological and
karyological studies on Orchidaceae. Caryologia, 46: 309–319.
D’Emerico, S. Bianco, P., Medagli, P. & Ruggiero, L. 1990.
Karyological studies of some taxa of the genera Himantoglossum,
Orchis,Serapias and Spiranthes (Orchidaceae) from Apulia, Italy.
Caryologia, 43: 267–276.
Devey, D. S., Bateman, R. M., Fay, M. F. & Hawkins, J. A. 2009.
Genetic structure and systematic relationships within the
Ophrys fuciflora aggregate (Orchidinae: Orchidaceae): high
diversity in Kent and a wind-induced discontinuity bisecting
the Adriatic. Annals of Botany, 104: 483–495.
Doyle, J. J. & Doyle, J. L. 1987. A rapid DNA isolation procedure
for small quantities of fresh leaf tissue. Phytochemistry Bulletin,
Botanical Society of America, 19: 11–15.
Ettlinger, D. M. T. 1997. Notes on British and Irish orchids.
Dorking: Published by the author.
Farrell, L. & Carey, P. D. 1999. Himantoglossum hircinum (L.)
Sprengel (Orchidaceae). In: Wiggington, M. J., ed. British Red
Data Book 1: vascular plants, 3rd ed., pp. 193–194.
Peterborough: JNCC.
Foley, M. J. Y. & Clarke, S. 2005. Orchids of the British Isles.
Cheltenham: Griffin Press.
Ga¨ umann, E., Mu¨ller, E., Nu¨esch, J. & Rimpau, R. H. 1961. U
¨ber
die Wurzelpilze von Loroglossum hircinum (L.) Rich.
Phytopathologische Zeitschrift, 41: 89–96.
Good, R. 1936. On the distribution of lizard orchid
(Himantoglossum hircinum Koch). New Phytologist, 35: 142–
170.
Gower, J. C. 1966. Some distance properties of latent root and
vector methods used in multivariate analysis. Biometrika, 52:
325–338.
Gower, J. C. 1971. A general coefficient of similarity and some of
its properties. Biometrics, 27: 857–872.
Gower, J. C. 1985. Measures of similarity, dissimilarity and
distance. Encyclopedia of Statistical Sciences 5: 397–405. New
York: Wiley.
Gower, J. C. & Ross, G. J. S. 1969. Minimum spanning trees and
single linkage cluster analysis. Journal of the Royal Statistical
Society C, 18: 54–64.
Harrap, A. & Harrap, S. 2009. Orchids of Britain and Ireland, 2nd
ed. London: A. & C. Black.
Hartley, S., Kunin, W. E., Lennon, J. L. & Pocock, M. J. O.
2004. Coherence and discontinuity in the scaling of species’
distribution patterns. Proceedings of the Royal Society of
London, 271: 81–88.
Heinrich, W. 2003. Zur Ansiedlung und Wiedereinbu¨rgerung
heimischer Orchideen die Entwicklung einer neubegru¨ ndeten
Population der Bocks-Riemenzunge (Himantoglosssum hirci-
num). Journal Europa¨ ischer Orchideen, 35: 455–538.
Heinrich, W. & Voelckel, H. 1999. Die Bocks-Riemenzunge
[Himantoglossum hircinum (L.) Spreng.] – Orchidee des Jahres
1999. Berichte aus den Arbeitskreisen Heimische Orchideen, 16:
83–123.
Heusser, K. 1915. Die Entwicklung der generativen Organe von
Himantoglossum hircinum Spr. (5Loroglossum hircinum Rich.).
Beihefte zum Botanischen Zentralblatt, 32: 218–277.
Inda, L. A., Pimental, M. & Chase, M. W. 2012. Phylogenetics of
tribe Orchideae (Orchidaceae: Orchidoideae): based on com-
bined DNA matrices: inferences regarding timing of diversifi-
cation and evolution of pollination syndromes. Annals of
Botany, 110: 71–90.
Jacquemyn, H., Deja, A., de Hert, K., Cachapa Bailarote, B. &
Lievens, B. 2012. Variation in mycorrhizal associations with
tulasnelloid fungi among populations of five Dactylorhiza
species. PLoS ONE, 7(8): e42212 (10 pp).
Kaiser, R. 1993. The scent of orchids: olfactory and chemical
investigations. Basel: Editiones Roches.
Kay, J. M., Whittall, J. B. & Hodges, S. A. 2006. A survey of
nuclear ribosomal internal transcribed spacer substitution rates
across angiosperms: an approximate molecular clock with life
history effects. BMC Evolutionary Biology, 6: 36 (9 pp.).
Kerner, A. v. M. 1891. Pflanzenleben 2. Leipzig.
Kropf, M. & Erz, S. 1996. Die Bocksriemenzunge (Himantoglossum
hircinum (L.) Sprengel) – eine Charakteritische Orchideenart
der Weinbergsbrachen im Nahegebiet in Ausbreitung. Berichte
aus den Arbeitskreisen Heimische Orchideen, 12: 17–33.
Kropf, M. & Renner, S. S. 2008. Pollinator-mediated selfing in two
deceptive orchids and a review of pollinium tracking studies
addressing geitonogamy. Oecologia, 155: 497–508.
Kull, T. & Hutchings, M. J. 2006. A comparative analysis of the
decline in the distribution ranges of orchid species in Estonia
and the United Kingdom. Biological Conservation, 129: 31–39.
Kurzweil, H. 1987. Developmental studies in orchid flowers II:
Orchidoid species. Nordic Journal of Botany, 7: 443–451.
Landwehr, J. 1977. De wilde Orchideee¨n van Europa. Amsterdam:
VBNN.
Lang, D. L. 2004. Britain’s orchids. Hampshire: Wildguides.
Leitch, I. J., Kahandawala, I., Suda, J., Hanson, L., Ingrouille,
M. J., Chase, M. W. & Fay, M. F. 2009. Genome size diversity
in orchids: consequences and evolution. Annals of Botany, 104:
469–481.
Molna´ r, A. V., Kreutz, C. A. J., Ovari, M., Sennikov, A. N.,
Bateman, R. M., Takacs, A., Somlyay, L. & Sramko´ , G. 2012a.
Himantoglossum jankae (Orchidaceae: Orchideae), a new name
for a long-misnamed lizard orchid. Phytotaxa, 73: 8–12.
Molna´ r, A. V., To¨ko¨lyi, J., Ve´gva´ ri, Z., Sramko´ , G., Sulyok, J. &
Barta, Z. 2012b. Pollination mode predicts phenological
response to climate change in terrestrial orchids: a case study
from central Europe. Journal of Ecology, 100: 1141–1152.
Moore, D. M. 1980. Himantoglossum Koch. In: Tutin, T. G.,
Heywood, V. H., Burgess, N. A., Moore, D. M., Valentine,
D. H. Walters, S. M. & Webb, D. A., eds. Flora Europaea,5:
342. Cambridge: Cambridge University Press.
Neiland, R. M., Wood, J., D’Emerico, S., Vogelpoel, N., Greyer,
R. J., Widmer, A., Cozzolino, S. & Dafni, A. 2001.
Himantoglossum. In: Pridgeon, A. M., Cribb, P. J., Chase,
M. W. & Rasmussen, F. eds. Genera Orchidacearum 2,
Orchidoideae (Part 1), pp. 309–313. Oxford: Oxford
University Press.
Nelson, E. 1968. Monographie und Ikonographie der Orchidaceen-
Gattungen Serapias, Aceras, Loroglossum, Barlia. Chernex-
Montreux: Published by the author.
Payne, R. W., Murray, D. A., Harding, S. A., Baird, D. B. &
Souter, D. M., eds. 2008. Genstat v11. Hemel Hempstead: VSN
International.
Pfeifer, M. 2004. Variation of life history characteristics, long-term
population dynamics and genetic differentiation of
Himantoglossum hircinum (Orchidaceae). Doctoral thesis,
Friedrich-Schiller Universita¨ t, Jena. 94z19 pp.
Bateman et al. Observations on the Lizard Orchid
New Journal of Botany 2013 VOL.3 NO.2 139

Pfeifer, M., Heinrich, W. & Jetschke, G. 2006b. Climate, size and
flowering history determine flowering pattern of an orchid.
Botanical Journal of the Linnean Society, 151: 511–526.
Pfeifer, M. & Jetschke, G. 2006. Influence of geographical isolation
on genetic diversity of Himantoglossum hircinum (Orchidaceae).
Folia Geobotanica, 41: 3–20.
Pfeifer, M., Passalacqua, N. G., Bartram, S., Schatz, B., Croce, A.,
Carey, P. D., Kraudeit, H. & Jeltsch, F. 2010. Conservation
priorities differ at opposing species borders of a European
orchid. Biological Conservation, 143: 2207–2220.
Pfeifer, M., Schatz, B., Pico´ , F. X., Passalacqua, N. G., Fay, M. F.,
Carey, P. D. & Jeltsch, F. 2009. Phylogeography and genetic
structure of the orchid Himantoglossum hircinum (L.) Spreng.
across its European central–marginal gradient. Journal of
Biogeography, 36: 2353–2365.
Pfeifer, M., Wiegand, K., Heinrich, W. & Jetschke, G. 2006a.
Long-term demographic fluctuations in an orchid species
driven by weather: implications for conservation planning.
Journal of Applied Ecology, 43: 313–324.
Preston, C. D., Pearman, D. A. & Dines, T. D. 2002. New atlas of
the British and Irish flora. Oxford: Oxford University Press.
Pridgeon, A. M., Bateman, R. M., Cox, A. V., Hapeman, J. R. &
Chase, M. W. 1997. Phylogenetics of the subtribe Orchidinae
(Orchidoideae, Orchidaceae) based on nuclear ITS sequences.
1. Intergeneric relationships and polyphyly of Orchis sensu lato.
Lindleyana, 12: 89–109.
Rasmussen, H. N. 1995. Terrestrial orchids: from seed to
mycotrophic plant. Cambridge: Cambridge University Press.
Rudall, P. J., Perl, C. D. & Bateman, R. M. 2013. Organ
homologies in orchid flowers re-interpreted using the Musk
Orchid as a model. PeerJ, 1: e26. (23 pp.)
Schilthuizen, M. & Gravendeel, B. 2012. Left-right asymmetry in
plants and animals: a gold mine for research. Contributions to
Zoology, 81: 75–78.
Schmid, G. 1912. Zur O
¨kologie der Blu¨ te von Himantoglossum.
Berichte der Deutschen Botanischen Gesellschaft, 30: 463–469.
Sell, P. D. & Murrell, G. 1993. Flora of Great Britain, vol. 5.
Cambridge: Cambridge University Press.
Sramko´ , G. & Molna´r, A. V. 2012. Phylogenetics of the Eurasiatic
genus Himantoglossum (Orchideae, Orchidoideae). EOC
(2011) Proceedings (7 pp.).
Sramko´ , G., Molna´r, A. V., Hawkins, J. A. & Bateman, R. M.
2013. Molecular phylogenetics and evolution of the Eurasiatic
orchid genus Himantoglossum L. Molecular Phylogenetics and
Evolution [in review].
Sramko´ , G., Ovari, M., Yena, A. V., Sennikov, A. N., Somlyay, L.,
Bateman, R. M. & Molna´ r, A. V. 2012. Unravelling a century
of misuse: typification of the name Himantoglossum caprinum
(Orchidaceae: Orchideae). Phytotaxa, 66: 21–26.
Stace, C. A. 2010. New flora of the British Isles, 3rd ed. Cambridge:
Cambridge University Press.
Strack, D., Busch, E. & Klein, E. 1989. Anthocyanin patterns in
European orchids and their taxonomic and phylogenetic
relevance. Phytochemistry, 28: 2127–2139.
Summerhayes, V. S. 1951. Wild orchids of Britain. London: Collins.
Sundermann, H. 1973. Himantoglossum (Loroglossum) hircinum–
caprinum–calcaratum–affine. Acta Botanica Academiae
Scientiarum Hungaricae, 19: 367–374.
Sundermann, H. 1980. Europa¨ ische und mediteraane Orchideen –
eine Bestimmungsflora, 3rd ed. Hildesheim: Schmersow.
Swofford,D.L.2003.PAUP*. Phylogenetic Analysis Using
Parsimony (*and Other Methods). Version 4. Sunderland,
MA: Sinauer.
Teschner, W. 1980. Sippendifferenzierung und Besta¨ ubung bei
Himantoglossum Koch, in: Senghas, K. & Sundermann, H.,
eds. Probleme der Evolution bei europa¨ischen und mediterranen
Orchideen, pp. 104–115. Sonderheft: Die Orchidee.
Tra´ vnı´c
ˇek, P., Jersa´kova´ , J., Kuba´tova´ , B., Krejc
ˇı´kova´, J., Bateman,
R. M., Luc
ˇanova´ , M., Krajnı´kova´, E., Malinova´ , T., S
ˇtı´pkova´,
Z., Amardeilh, J.-P., Brzosko, E., Cabanne, O., Durka, W.,
Dworschak, W., Efimov, P., Hedre´n, M., Hermosilla, C.E.,
Kreutz, C. A. J., Kull, T., Marchand, O., Mohrmann, W., Rey,
M., Schiestl, F. P., Tali, K., Wasilewska, E., C
ˇurn, V. & Suda, J.
2012. Minority cytotypes in European populations of the
Gymnadenia conopsea complex (Orchidaceae) greatly increase
intraspecific and intrapopulation diversity. Annals of Botany,
110: 977–986.
Vogel, S. 1990. The role of scent glands in pollination. Washington,
DC: Smithsonian Institution.
Vo¨ th, W. 1990. Effektive und potentielle Besta¨uber von
Himantoglossum Spr. Mitteilungsblatt Arbeitskreis Heimische
Orchideen Baden-Wurttemburg, 22: 337–351.
Welch, C. J. 1998. Some observations concerning the asymmetry of
orchid flowers. Malayan Orchid Review, 32: 86–92.
White, T. J., Bruns, T. D., Lee, S. & Taylor, J. W. 1990.
Amplification and direct sequencing of fungal ribosomal RNA
genes for phylogenetics. In: Innis, M. A., Gelfand, D. H.,
Sninsky, J. J. & White, T. J. eds. PCR protocols: a guide to
methods and applications, pp. 315–322. San Diego, CA:
Academic Press.
Bateman et al. Observations on the Lizard Orchid
140 New Journal of Botany 2013 VOL.3 NO.2