A perennial glabrous herb with two entire, ovoid or subglobose tubers. Roots short and relatively thick. Stem (10–) 20–70 (–115) cm, pale green, sometimes faintly marbled with purple, solid, cylindrical, smooth, and obscurely angled above. Leaves pale to dark green; the lower 4–10, 5–20 × 3.5–6 cm, elliptical-oblong to oblong-lanceolate, obtuse to acute at apex, entire, keeled and erect or spreading, many-veined; the upper smaller, tapering to an acute apex, clasping the stem. Inflorescence a spike 1–25 (–50) × 4–12 cm, rather lax, and cylindrical. Bracts 0.5–5 × 0.1–0.2 cm, pale green or white, often tinged with rose, narrow linear, tapering to an acute apex, more or less membranous, often with in-rolled edges, with a green central vein and about three others on either side. Flowers greenish-white with a very long, ribbon-like and white, green and purple or brownish twisted labellum hanging obliquely downwards (on small spikes with few scattered flowers the labella often appear horizontal), and a strong smell of billy goats. Outer perianth segments 0.7–1 cm, forming a helmet or hood, cohering at the base, free or not at the tip, whitish-green often flushed with purple spots or streaks inside and out, ovate, rounded at apex, concave, 3- or 4-veined (often green), the upper arched forward and rather boat-shaped. Inner perianth segments slightly shorter than the outer, very narrow, linear spotted, and 1-veined. Labellum 2–6 cm, linear, wedge-shaped and broader at the base, with (approximately) 12 purple spots around the junction where the three lobes meet; the edges are crimped. Side lobes 0.5–2 cm, purple brownish, curled and ribbon-like. Middle lobe white on top near its base, becomes purple light brownish for the 80% nearest the tip, and is curled. Spur short, conical, sock-like, rounded at the apex and directed downwards. Column rather short and erect. Single anther greenish-white and pyriform. There are two olive-green, pyriform pollinia. Caudicles thick, yellow, longer than the pollinia and bent at the apex. Viscidium elliptical or quadrangular, enclosed in a bursicle and very sticky. Stigma obtusely 4-cornered or ovate-cordate and bordered by a dark line. Rostellum beak-like, projecting beyond the stigma. Ovary about 1 cm, pale green, spindle-shaped, twisted, pedicellate, with 3 slight, longitudinal ridges. Capsule long, tapering at the base, with marked ridges, 0.5–1.5 cm long and about 0.25–0.75 cm in diameter and contains 0–2000 seeds. Seeds 340 µm in length by 120 µm in diameter with the embryo measuring 100 µm in length and 70 µm in diameter. Seed mass > 0.01 µg.
A species of calcareous grasslands, sand dunes and occasionally woodland edges. It often appears sporadically as individuals or small colonies in new sites where it may last 20 years. In England larger colonies appeared only in the latter part of the 20th century and their persistence is not known.
I. Geographical and altitudinal distribution
The occurrence of Himantoglossum hircinum in the British Isles has been well documented (Good 1936; Carey 1999). It has been found only in England and has been restricted mostly to the south and east (Fig. 1). Populations have been found at about sea-level in Kent, Sussex, Somerset and Devon to approximately 200 m in Gloucestershire and on the South and North Downs.
On the continent, H. hircinum is spread over much of southern Europe (Fig. 2). The species is also reported in a few localities in North Africa, in Morocco and Algeria. The distribution shown by Meusel et al. (Vergl. Chor.) gives two subspecies: H. hircinum ssp. hircinum is found from Spain eastwards to the Balkans; with ssp. calcaratum in Yugoslavia, Romania and northern Greece. The European distribution of the different subspecies is somewhat blurred by the uncertainty caused by problems of identification. Pridgeon et al. (2001) state that there are four species: H. hircinum; H. caprinum; H. calcaratum and H. affine. Conversely, three subspecies of H. hircinum are identified by Tutin et al. (Fl. Eur. 5): ssp. hircinum; ssp. calcaratum; and ssp. caprinum. Arguably, there is a fourth subspecies or fifth species, ssp. samariensis, found on Crete (Antonis 1998). In the rest of this text H. hircinum refers to ssp. hircinum as it is the only subspecies found in the British Isles and western Europe.
Himantoglossum hircinum is especially common in the wine growing regions of northern, central and western France. The species is common in the Picos de Europa but appears only as scattered populations in the rest of northern Spain. Scattered populations are also found in Germany and Austria. In France and Switzerland, it is found up to an altitude of 850 m. The distribution in France is given by Bournérias et al. (1998) whilst the distribution and biology of the species in Germany are described in great detail by Heinrich & Voelckel (1999). Preston & Hill (1997) give H. hircinum as Submediterranean–Subatlantic.
(A) climatic and topographical limitations
Himantoglossum hircinum is likely to be limited to the north of its distribution in Europe by low temperatures in both winter and summer (Fig. 3) and to the east by low winter temperature. It has been suggested that H. hircinum will spread northwards and populations become more common as the climate warms (Good 1936; Carey & Brown 1994; Carey 1996). Rainfall during the vegetative growing season was related to a significant increase in flowering and seed production (Carey 1999).
The plant in England grows on flat or sloping ground with no particular orientation favoured. It most often occurs in tall sparse grassland and is associated with a regular (annual or biennial) mowing or cutting regime, especially in rough areas adjoining intensively managed grassland (golf courses and a horse race-course) or adjoining linear features such as roads (Table 1).
Table 1. The substrate and land-use on which H. hircinum has been recorded. The percentage of records from those where substrate and/or habitat were listed is given, as is the percentage of the total of 201 sites recorded in England from 1641 onwards. Land-use is separated into man-made linear features, pits, managed grassland and ‘unmanaged’ grassland
Percentage of sites where feature noted
Percentage of total
Sand and stones
Other managed grassland
Edge of wood
Edge of scrub/hedgerow
The records of all the English sites held at Monks Wood were consulted and those for which substrate was listed (29%) showed that Himantoglossum hircinum is almost always found on shallow calcareous soils or sand/gravel and in the past was predominantly restricted to chalk (Table 1). In recent years, two populations have been noted on neutral soils and one part of a population has been found on broken asphalt (this may have calcareous soil underneath or lime in the foundations). In France, where the species is most common, it is noted as a plant of roadside verges and has been seen growing through broken concrete. Both in France and England, the species is not found on neutral/acid substrates, e.g. Brittany, the Massif Centrale, Devon and Cornwall, even though the climate appears to be favourable. In Devon, the species has been found growing on calcareous sand dunes.
Himantoglossum hircinum is most often associated with established open chalk or calcareous grassland (Table 1). It has been noted in a number of chalk pits and on man-made structures such as railway embankments and roadside verges, indicating that the grassland does not have to be ancient (Table 1). It is frequently found in association with other orchid species, notably Anacamptis pyramidalis, but also Gymnadenia conopsea, Orchis morio, Ophrys apifera and Aceras anthropophorum. This may be linked to the presence of suitable fungi for the crucial mycorrhizal associations that the orchid species require (see VIe) as well as the presence of suitable grassland communities (Table 2).
Table 2. Plant species associated with Himantoglossum hircinum from 14 sites in England. Only those species found at more than one site are included
Number of Sites
Number of Sites
During a study of 27 populations of H. hircinum in France (Crompton & Farrell 1985), it was found most often on gently sloping, south-west facing, Bromopsis erecta or Brachypodium pinnatum dominated grassland on calcareous soils. These two grasses were dominant in 25 of the 27 sites. Frequent associates in France, found in 25–45% of sites, were Achillea millefolium, Arrhenatherum elatius, Carex flacca, Centaurea nigra, Clinopodium vulgare, Eryngium campestre, Euphorbia cyparissias, Festuca rubra, Helianthemum nummularium, Hippocrepis comosa, Linum catharticum, Medicago lupulina, Ononis repens, Pilosella officinarum, Poa pratensis, Sanguisorba minor, Seseli libanotis, Teucrium chamaedrys and Thymus polytrichus.
Himantoglossum hircinum never appeared in the quadrats that were used to create the National Vegetation Classification and so does not feature in the volumes so far produced. However, the communities in which it is found equate to: CG3 (Bromus erectus grassland), CG4 (Brachypodium pinnatum grassland), CG5 (Bromus erectus and Brachypodium pinnatum grassland) (Rodwell 1992); the edges of W21 (Crataegus monogyna–Hedera helix scrub) (Rodwell 1991); and SD8a (Festuca rubra–Galium verum fixed dune grassland, subcommunity Astragalo-Festucetum arenariae) (Rodwell 2000). In Germany it is listed as a species of the Mesobrometum and Geranion sanguinei (Pfl. Exk.).
IV. Response to biotic factors
As livestock are absent on all the English sites for Himantoglossum hircinum it seldom suffers from grazing by vertebrates. It does not appear in areas heavily grazed by rabbits, sheep or cattle. There was an exception in 2000 when the Dorset population was heavily grazed by rabbits over the winter. By May 2000 the rabbits had gone, or had been eradicated, and the plants flowered normally. In the Picos de Europa of northern Spain, plants were often observed growing in small meadows that had just been grazed by cattle and in one case 150 plants were found in an area approximately 30 × 50 m (P. D. Carey, personal observation). The grass was grazed short (< 5 cm) but the H. hircinum inflorescences had been totally avoided by the grazing cows and must be assumed to be unpalatable to them.
Himantoglossum hircinum grows to a larger size with a higher probability of flowering in relatively tall (30 cm) sparse grassland than in coarse vegetation or short turf. Summerhayes (1951) suggested that the species was associated with scrub and indeed flowering plants have been observed in dense vegetation under a hedge (P. D. Carey, personal observation) and adjacent to hawthorn and birch bushes (L. Farrell & G. Crompton, personal observation) but we assume that scrub would eventually shade out plants.
Individuals and populations have been wiped out by scrub encroachment (Carey 1999; L. Farrell, personal observation) and Rubus fruticosus (bramble) invasion (P. D. Carey, personal observation). Very young plants (single leaf, 2 mm × 10 mm) do not readily grow through mosses, e.g. Hypnum cupresssiforme, in autumns following a wet summer and are out-competed and swamped by rosette plants such as Hypochaeris spp., Pilosella spp. and Plantago spp. (P. D. Carey, personal observation & N. F. Stewart, personal communication).
Plants of H. hircinum often grow together in groups with one or two large specimens surrounded by seedlings and small ones. In extreme years (1990–92) up to 120 seedlings were found within a circle of radius 12.5 cm and that is approximately 5 cm greater than the radius of a large plant. Small plants have often been found underneath the leaves of larger ones. At these densities there must be some intraspecific competition.
The most serious damage due to trampling of plants of H. hircinum occurred at a site that was almost destroyed by horses during one winter. As already noted, in the British Isles H. hircinum tends to be found in localities where grazing animals are absent and so trampling by animals is slight. In the meadows of the Picos de Europa mentioned in Section IVa, some inflorescences were trampled by cattle. Trampling by humans seems to have relatively little effect on mature populations in areas where the plants are monitored and visited by photographers and wildlife enthusiasts. Individual plants, however, have been crushed by photographers and ecologists but apparently not fatally.
V. Response to environment
If conditions are favourable, H. hircinum is a long-lived polycarpic perennial. At most sites populations consist of only one or two flowering individuals. At two extant sites numbers have become much greater in recent years (Fig. 4). At one site in Cambridgeshire more than 250 flowering individuals were seen in 1995 and 2000, and as 30% flowered in monitored plots the population was estimated at approximately 850 plants. At the largest site in Kent, over 5000 flowers were observed in 2000. With 14.5% flowering in three permanent plots at that site the population is estimated at 27 500 plants covering an area of 1 × 0.5 km.
As mentioned previously, H. hircinum tends to form clumps around established individuals. This may be the result of the density of dispersed seeds and/or be related to the density of suitable mycorrhizal fungi.
Survivorship from three different closely monitored sites normally show a Deevey type II curve, but through exceptionally favourable runs of years a Deevey type I curve has been noted (P. D. Carey, unpublished data).
(b) performance in different habitats
Leaves are quite commonly present during the flowering stage at the largest population (P. D. Carey, personal observation), whereas this is rarer for other populations where leaves have normally senesced before flowering (Summerhayes 1951; P. D. Carey, personal observation).
In dense vegetation the plant can become larger (up to 12 leaves) with longer narrower leaves (4 × 20 cm) than normally seen in open vegetation where the leaves are broader and shorter (6 × 15 cm).
(c) effects of frost, drought and waterlogging
Himantoglossum hircinum can survive hard frosts in winter. This can be inferred by the presence of the species near Vienna. There is also no evidence of frost killing plants in any winter between 1977 and 2000, although Good believed that very hard December frosts could kill plants (D. Coombe, personal communication). Late frosts in May or June can cause flower buds to abort or inflorescences to reach only a minimum size (P. D. Carey & L. Farrell, personal observation; G. Crompton, personal communication).
Rainfall during the vegetative growing season (September to April) has a marked effect on individuals (Carey 1999). Plants at all stages of the life-cycle are killed by severe drought. Severe winter drought reduces the number of flowering plants. The winter rainfall in the years 1995–96 and 1996–97 was exceptionally low and this caused a noticeable decrease in the number of flowers at the largest sites in 1997–98 in Cambridgeshire (Fig. 4a) and 1996–97 in Kent (Fig. 4b). The probability of a plant producing pods and therefore seeds was positively related to rainfall in the growing season (t) immediately before the flowers were produced and the probability of flowering was positively related to the rainfall in the growing season before flowering (t–1). This suggested that 2 years with average or above average rainfall are required for a plant to produce a large number of seeds (Carey 1999). Since that work was published, the model has been subsequently tested with a further 5 years of data and proved not to be robust. Flowering is determined by plant size in the year of flowering at all sites, and at the Kent population by the size of the plant in the previous year as well (P. D. Carey, unpublished data). In a model, plant size was related highly significantly to an aridity index incorporating rainfall and temperature for September in year t and year t–1 using data from the Kent population (P. D. Carey, unpublished data). September is a critical month as this is the time of year that the leaves begin to emerge from the tuber and the growth of leaves and roots occurs. Unfortunately, the model failed when tested against the Cambridgeshire data. This may have been because the model was poor or because of differences between the sites or the phenotypes living there. The production of seeds is related to plant size because seeds will be produced only if there are flowers, and flowering is related to plant size. The success of pollination is obviously important for the number of seeds produced but this too could be related to the weather in the year of seed production with cool, damp summers being less favourable than hot, dry ones. The flowering ‘initial’ is produced 1 year before the plant flowers and produces seeds. This is a similar finding to the work of Wells et al. (1998) for other orchid species.
Drought in August may benefit H. hircinum, because during this part of the year it remains underground as a tuber and mosses and rosette plants with which it competes will be detrimentally affected or killed.
The species always occurs on light well-drained soils and no effect of waterlogging has been noted (by the authors or in any other source).
VI. Structure and physiology
The life-cycle of Himantoglossum hircinum is still uncertain despite 24 years of research and monitoring by the present authors and colleagues. The following account describes the life-cycle to the best of our knowledge.
Seeds germinate sometime between the autumn in which they are produced and the following autumn or the two after that (P. D. Carey & H. S. Scott, unpublished data) depending on the rain in the growing season. See Section VIII for a description of the next stage of the life-cycle.
Anecdotal evidence (Good 1936) suggests that 3 years after seeds are produced, young ‘seedlings’ appear above ground. Demographic data from permanent plots at the largest population showed that the plant may stay at this seedling stage for 1 year but commonly will remain at this stage for 2 or more years. The next recognizable stage is a two-leaf plant. Again, plants may stay at this stage for some years before getting larger or dying. The following stage is a three-leaf plant where it is bordering on being sexually mature. In extremely good years, two- and three-leaf plants have a low probability of flowering. Plants can stay at this three-leaf stage for many years or alternate with the two-leaf category or the four-or-more-leaf category. The four-or-more-leaf category can be considered sexually mature. In some years it will flower and in others not. The proportion of plants flowering is at least partly related to rainfall (Section V). Plants that emerge first tend to develop into the strongest individuals with the most leaves, and it is these individuals that often flower as they have built up more resources (L. Farrell, personal observation).
The drawings of R. Ward (Fig. 5) illustrate the growth of a large H. hircinum plant transplanted from Kent to the Botanic Gardens in Cambridge. The drawings follow the plant from the tuber stage in June to the flowering stage the following summer and into the next annual cycle. In June the plant existed as a tuber (Fig. 5a) and by mid-August a leaf shoot appeared (Fig. 5b). In mid-September roots were developing and the leaves forming (Fig. 5c). In March two new tubers were forming and the leaf tips of the plant were browning (Fig. 5f). By April the tubers had increased in size and were firm (Fig. 5g). In late May the plant was approximately two weeks from flowering, the leaves were browning and the roots were dying back (Fig. 5h). Only the larger of the two tubers survived the summer to produce leaves and roots the following autumn.
Leaves of larger plants have a stomatal density of 44.0 ± 3.4 mm−2 on the abaxial surface and no stomata on the adaxial surface.
A number of fungi were isolated from roots of H. hircinum by Gäumann et al. (1961) but there is no clear evidence that any were mycorrhizal. Of 15 strains, 14 were attributed to Ascomycetes and one to Rhizoctonia versicolor. As all these fungi induced defence reactions in the host tuber tissue, it is unlikely that they were mutualistic, but the association with Rhizoctonia was possibly mycorrhizal. Fungal isolate F403 has been used for symbiotic germination of H. hircinum and H. adriaticum at the Royal Botanic Gardens, Kew (G. Prendergast, personal communication).
The role of mycorrhiza in the germination of seeds is discussed in Section VIII below.
(c) perennation: reproduction
Himantoglossum hircinum is a perennial that survives from one year to the next via a tuber. Reproduction is by seed with between 0 and approximately 30 000 seeds produced per plant per year. It has been reported that H. hircinum can reproduce vegetatively by forming new tubers on the root system; evidence of this was shown at the Cambridge Botanic Gardens (D. Donald, personal communciation and Fig. 5). However, when a group of plants was dug up in Kent it was found that all plants were individuals and not attached by a communal root system (R. Fitzgerald, personal communication).
Himantoglossum hircinum has 36 chromosomes in the somatic cells (2n) and 18 or 12 in the gametes (Heuser 1915; Bianco et al. 1987). Cases of aneuploidy with 2n = 36 + 1B are known (Capineri & Rossi 1987; D’Emerico et al. 1993). The chromosome size in individuals collected in Italy varied from 4.05 to 1.89 µm. The karyotype was 24m + 2mss, + 2ms + 2sm + 4sms + 2smss. Pair 1 had a satellite on the long arm; pairs 2 and 5 have a satellite on the short arm and a microsatellite on the long arm; pairs 13 and 16 have a satellite on the short arm (D’Emerico et al. 1990, 1993).
(e) physiological data
Very little work has been carried out on the growth of H. hircinum. It has rarely been grown for research purposes in laboratory conditions or in planted field experiments. Orchid growers consider H. hircinum difficult to grow in garden or glasshouse conditions. However seedlings have been cultured in vitro by germinating seeds asymbiotically to produce protocorms which when transferred to a fresh substrate, containing coconut milk, 4 months after sowing developed rapidly (Lucke 1976). The Sainsbury orchid project at Kew has also produced seedlings in vitro (G. Prendergast, personal communication).
(f) biochemical data
The text for this section comes from Pridgeon et al. (2001), and is information collated by Nigel Veitch and Renée Grayer.
Infection of the tubers of H. hircinum with a strain of Rhizoctonia repens from Orchis militaris results in the synthesis of the phytoalexin orchinol (2,4-dimethoxy-7-hydroxy-9, 10-dihydrophenanthrene) and p-hydroxybenzylalcohol (Gäumann et al. 1960; Nüesch 1963). An isomer of orchinol, 2,4-dimethoxy-5-hydroxy-9, 10-dihydrophenanthrene (loroglossol) was reported at levels of approximately 150 mg kg−1 in H. hircinum tubers infected with Rhizoctonia versicolor (Hardegger et al. 1963; Urech et al. 1963). A related phytoalexin, 4-methoxy-2,5-dihydroxy-9,10-dihydrophenanthrene (hircinol), also occurs in the same source at higher concentrations (250–400 mg kg−1) than those found for loroglossol. The antifungal activity of orchinol and hircinol against Candida lipolytica (BY 17) has been assessed (Fisch et al. 1973). Orchinol was found to be more active than hircinol at both 50 p.p.m. and 100 p.p.m. and caused complete inhibition of growth of C. lipolytica for the first 6 days when applied at the latter concentration. Loroglossol was stated to be inactive in these assays. In a later report, the activity of loroglossol against spore germination of Monilinia fruticola (Wint.) Honey and Phytophthora infestans (Mont.) de Bary was described (Ward et al. 1975). The failure of this compound to show activity in earlier investigations was linked to its low solubility in water. Chemical structures of the three 9,10-dihydrophenanthrenes found in Himantoglossum and other Orchidinae are shown by Pridgeon et al. (2001).
The chemical composition of the scent of H. hircinum flowers was obtained by gas chromatography coupled with mass spectroscopy (Kaiser 1993). The principal components are (E)-3-methyl-4-decenoic acid (Z)-4-decenoic acid and lauric acid. The anthocyanin content of the flowers of H. adriaticum was investigated by Strack et al. (1989).
Leaves appear from late August (depending on rainfall) to April. Single leaf plants may continue to appear above ground after November. Mature plants flower from early June to late July depending on weather but typically appear 1–7 July and last for two to three weeks. Seed-pods appear mid to late July and are mature 4–6 weeks later but typically in the third or fourth week of August. The exact timing of flowering and seed production is dependent on the position of the flower on the inflorescence. The lower flowers open first and form pods first. The uppermost flowers can open up to 14 days after the lowest ones although this varies from plant to plant and from year to year.
The leaves of vegetative plants blacken or brown and disappear from March until July depending on the size of the plant and rainfall. Seedlings disappear first, then small and medium plants and finally large plants. In high rainfall years seedlings are apparent until at least the end of April and mature plants last until July. In dry years seedlings have browned by the end of March and small to large plants are browning/blackening from March onwards. Flowering plants often retain their leaves but in some dry years the leaves have disappeared by the time the flowers open.
VIII. Floral and seed characters
(a) floral biology
The floral biology of Himantoglossum hircinum was studied and described in detail by John Leonard while he was a warden looking after the population in Cambridgeshire in 1988. The following description is summarized from his unpublished report with further additions and corrections.
The morphology of H. hircinum flowers is shown in Fig. 6. Part of the rostellum forms a purple flap, the bursicle, that covers the viscidium. For the pollinia to be removed the bursicle has to be pushed forcibly down in order to reveal the viscidium, as a white and sticky pad underneath.
When removed in hand-pollination trials, the exposed viscidium becomes dry and loses its stickiness in less than three minutes which is the same phenomenon reported by Darwin (1890) for Orchis mascula. The presence of the bursicle therefore appears to have the function to keep the viscidium moist and sticky. This view was supported when it was noted that in flowers where the bursicle had been triggered but the viscidium had not been removed the viscidium was dry and not sticky. Once attached to something, the pollinia remain attached unless mechanically removed. Seven months after being collected pollinia were still attached to the glass tubes into which they were collected.
Hand-pollination studies showed pollinia that were dabbed onto the stigmatic surface led to 85% pollination whereas the addition of whole pollinia to the stigmatic surface led to 81% pollination. This suggests that an insect with the viscidium firmly stuck to its proboscis does not have to remove the pollinia to achieve pollination in a second flower but has only to brush the pollinia against the stigmatic surface. Using a needle to pollinate the flowers was very successful and it could be said that this method mimics the action of the proboscis. After pollination the stigmatic surface often blackens and this may be used as an indicator of successful pollination.
There is no evidence of self-pollination but pollinia were seen hanging out of the anthers and some were touching the stigmatic surface. Self-pollination is therefore a physical possibility. Drawings of the pollinia are shown by Dressler (1990).
Darwin (1890) noted that when removed the two pollinia ‘do not diverge, but become depressed sweeping through an angle of 90 degrees in about 30 seconds. They are then in a proper position for striking the single large stigma ...’. In the Orchideae most of the pollen occurs in tetrad formation, with no exine development and with the tetrads loosely united by elastic threads from the tapetum (Dressler & Dodson 1960). In three species of Himantoglossum examined, Schill & Pfeiffer (1977) reported verrucose–hamulate sculpturing. In the case of H. hircinum, there are frontal portions of the same massula, distinctly verrucose–hamulate on the sides but coalescing into gemmae of different sizes on the front (Pridgeon et al. 2001).
The presence of nectar in the flowers of H. hircinum is not certain although Teschner (1980) found evidence for the presence of glucose in the spur of most flowers. However, the closely related species Barlia robertiana is rewardless (Smithson & Gigord 2001) and tests using microcapillary tubes and filter papers have failed to find any evidence of nectar in H. hircinum flowers (A. Smithson, personal communication). The smell of the flowers is unmistakable. Some people say the smell is of male goats and in languages other than English the common name is (billy-) goat orchid. The strong smell may be an adaptation to attract flies or night-flying insects. The wall mason wasp Odyneris parietum was seen pollinating flowers in Sussex in the 1920s (notebook of Guerman Prey – from correspondence between D. C. Lang and G. Crompton). Bumble bees, hoverflies, and butterflies have all been seen visiting flowers in Cambridgeshire (L. Farrell, personal observation). Solitary bees (Megachile maritima) have been observed visiting flowers and carrying pollinia on sand-dunes (Willis 1990). The solitary bee Colletes cunicularis is listed as a visiting insect by Bournérias et al. (1998) but this species is not likely to be the pollinator in the British Isles because it has a contrary distribution to H. hircinum and flies at a time when the plants are not flowering (Edwards 1997). In addition to these likely pollinators, ants (L. Farrell, personal observation) and small carabid beetles (P. D. Carey, personal observation) have been seen crawling around in the flowers. As seed pods are produced on individual plants many miles from the nearest neighbouring flowers, cross-pollination from other flowers on the same plant can be said to occur.
There are no recorded hybrids of H. hircinum in the British Isles. In continental Europe a hybrid with Orchis simia has been reported (Hyb. Br. Isl.).
(c) seed production and dispersal
Seed production is 0–1200 seeds per pod (Summerhayes 1951), although in some years pods appear to have many more than 1200 (P. D. Carey, personal observation). The number of pods produced per year per plant is very variable. In some years there are no pods and in others there can be up to 50 on an individual plant (P. D. Carey, L. Farrell & N. F. Stewart, unpublished data). Seeds are approximately 340 × 120 × 120 µm with most of the seed being taken up by the ‘wing’; the nucleus is approximately 100 × 70 × 70 µm (Remes 1995). Assuming that the seed is the same density as water (it is probably lighter) it will weigh in the order of 5 × 10−8 g. Carey (1998) has modelled the dispersal of H. hircinum and predicted a mean dispersal distance of approximately 1 metre. It is assumed that orchid seeds can fly very long distances if they are picked up by an updraft of wind. The apparent random distribution of populations around the south of England suggests that long-distance dispersal is by rare, wind-blown events or another vector is involved. The most likely vector is man (Carey 1999). Of the 19 extant English populations, 6 are on golf courses. This suggests that golf courses represent large areas of suitable habitat thereby increasing the chances that a seed will land in a favourable place and/or golfers transport the somewhat adhesive seeds from one course to another.
(D) viability of seeds: germination
A long-term experiment is underway to establish the conditions that favour germination of seeds of H. hircinum and how this occurs (P. D. Carey & H. S. Scott, unpublished data). Seeds have been placed in packets in the field using the method of Rasmussen & Whigham (1993) so that they can be retrieved periodically and examined for mycorrhizal infection and state of germination. In the first year of this experiment (1996–97), only 0–20% of seeds showed any signs of infection and many seeds were desiccated. The growing season of the first year was exceptionally dry. In the second year, the seeds from the first year showed 1–5% germination whereas, for seeds produced and sown in the second year, 15–40% were in the early stages of germination with most of the seeds infected. The autumn and winter of 1997–98 were considerably wetter than those of 1996–97. In the third year, percentage infection was high in January but decreased by May. In May (1999) protocorms were discovered in one of the 3-year-old seed packets that was placed in the field in Year 1 (1996). These protocorms (approximately 0.3 cm and spherical) appear the same as those photographed growing on asymbiotic culture (Lucke 1976). In May 2000, more of the spherical protocorms were found and also a larger protocorm (approximately 1 cm by 0.5 cm) was discovered in one 2-year-old packet that had been placed in the field in Year 3 (1998). In April 2001, the small spherical protocorms were again found but not any of the larger size found in 2000. In 2001 many seeds from packets of different ages had rotted following the exceptionally wet year April 2000–April 2001. Early conclusions from this experiment are that the percentage of seeds becoming infected and germinating varies from year to year, and that the development from germinating seed to protocorm does not occur in a defined space of time.
The developmental biology of orchid species was described by Wirth & Withner (1959). Development in orchid embryos can be classified by reference to the development of suspensor cells. Five classes are recognized and H. hircinum is in class II where a filament of 5 to 10 cells appears (Heuser 1915).
(d) protocorm development and seedling morphology
The first tuber is formed during the first spring after germination from a bud in the axil of a scale leaf at the tip of the protocorm; the protocorm itself disappears in summer. In the second autumn, a short mycotrophic rhizome develops from the tip of the tuber, which produces a mycotrophic root and from the axil of a scale leaf the second tuber begins to develop, reaching full size in early summer. The rest of the plant then withers leaving the tuber as the only part to survive through the summer. In the next autumn, the tuber produces a short rhizome with a new mycotrophic root and eventually, when the seedlings are more than 2-years-old, a leafy shoot.
IX. Herbivory and disease
(a) animal feeders or parasites
Very few vertebrates have been noted as grazing Himantoglossum hircinum although horses did have a major effect on one population and in Cambridgeshire muntjac deer or a hare has eaten one inflorescence. Rabbits grazed one site during the winter of 1999–2000 but were removed by late spring and the plants flowered normally. Although several large plants did not reappear in 2001, these losses were more than balanced by the number of recruits. Slugs and snails often eat leaves or parts of leaves and flowers but rarely does this significantly affect the plant. On one occasion an unseasonal caterpillar of Phragmatobia fuliginosa (ruby tiger moth) ate an inflorescence (G. Crompton, personal communication).
There are no records for H. hircinum in the Phytophagous Insect Data Base.
(b) plant parasites and diseases
No known diseases or parasites are known for H. hircinum. A group of plants at one site have variegated leaves which is assumed either to be phenotypic variation or caused by an unidentified virus. It is also suspected that leaf miners may infect plants as features that resemble mines have been noted on leaves.
The earliest mention of Himantoglossum hircinum appears to be in the herbal of Dodoens (1583) that was cited but not referenced by Schweinfurth (1959). Schweinfurth states that Dodoens’ work shows ‘an unmistakeable likeness of the orchids now known as Habenaria (Platanthera) bifolia, Himantoglossum hircinum, Listera ovata, and Orchis maculata’.
The historical records for H. hircinum in the British Isles are probably more complete than for almost any other species over the last 300 years. A full account of the history of this species is given for the years from 1641 to 1934 by Good (1936) and for the years 1895–1998 by Carey (1999). The abundance of this species has varied considerably in England over time. The plant was exceptionally rare in England until a group of populations became established around Dartford in Kent in the 19th century. This group of populations declined so that by the end of the 19th century only four populations (including isolated individuals) were left in England. In the years during the First World War and again in the 1920s, the abundance of H. hircinum became much greater and it was more widespread so that there were 36 populations by 1927. After 1934, a dramatic decline took place so that from the years 1945–94 there were only 9–11 populations in any one year and the turnover in populations was high. From 1994, several new sites were recorded and in 1996 there were 16 populations of H. hircinum in the British Isles. In 2000, there were 19 extant populations, but four of those found in 1996 had disappeared and there were seven new ones at different sites.
(a) status and protection
Himantoglossum hircinum is a Schedule 8 species protected under the Wildlife & Countryside Act 1981 and is listed as ‘vulnerable’ in the British Red Data Book for Vascular Plants (Wigginton 1999). It is considered ‘not threatened’ on a European scale.
In the past, many plants were picked for Herbaria or dug-up to be transplanted to gardens (Carey 1999). This threat was considered so great in the 1970s and 1980s that a 24-hour guard was posted at the Cambridgeshire population which at the time consisted of only a handful of plants. This protection may have helped the population survive long enough to produce enough seed for the population to grow.
Public awareness of the damage that plant collecting has on natural populations has increased in recent years and it is unlikely that populations will be lost in future due to over-collection.
Active management has been carried out at most British sites to remove encroaching scrub and brambles. At some sites grass has been cut or burned to prevent the build up of rank vegetation and is especially important following wet summers. Burning reduced flowering in the year it was carried out but the plants survived to re-emerge the following autumn.
Where H. hircinum grows on golf courses, plants close to the areas of play are surrounded by ‘ground under repair’ hoops so that any ball landing near the plants can be moved by the player without penalty. This simple measure protects the plants very adequately. At the largest British population on a golf course in Kent, a long-term management plan was developed in the 1970s to protect H. hircinum and other nationally rare species on the site. The most important feature of the plan is that nutrient enrichment of the dune system is kept to a minimum, and this is achieved by making sure that grass cuttings are placed only in specially dug pits or are carried away from the dunes altogether.
(c) archive material
Seeds have been collected each year since the early 1990s for the Sainsbury Orchid Project and sent to the Royal Botanic Gardens at Kew to be stored in the seedbank held there. Some of the seeds are subsequently used for propagation purposes.
We thank Mrs G. Crompton and N. F. Stewart whose recording of this species and their collation of records over the years has been invaluable. Mrs Jane Croft produced the distribution map for the British Isles at the Centre for Ecology and Hydrology and David Roy interrogated the Phytophagous Insect Data Base. Prof. Arthur Willis, Dr Michael Proctor and Dr Tony Davy provided useful advice on earlier drafts of this text.