Devillers, P. & Devillers-Terschuren, J.A. 2000. Observations of the orchids of the Ophrys subfusca group in Tunisia. Extended summary. Royal Belgian Institute of Natural Sciences website, www.naturalsciences.net/cb. [Original published in French in Natural. Belges 81 (Orchid. 13)].
Mediterranean Africa appears to have been a centre of radiation for orchids of section Pseudophrys of genus Ophrys. They comprise there a large number of forms, sometimes very original, probably belonging to several lineages of which some are little or not represented in the rest of the Mesogean basin. It is also the only region where the Pseudophrys generally exceed the Euophrys in frequency, abundance and diversity. North Africa harbours forms that seem to be intermediate between well individualised groups in Europe and Asia Minor, a particularity to be expected in a centre of differentiation. This richness in intermediate forms is particularly striking in the Ophrys fusca and O. lutea groups and has long attracted attention. While these two entities have always been recognised in Europe as two distinct species, even by orchidologists espousing the broadest species concept, the African intermediates have often been considered as hybrids between them. However, botanists having an extended field experience in North Africa, like Murbeck (1897, 1899), Ferlan (1956), Maire (1959), Baumann (1975), Vallès & Vallès-Lombard (1988), have generally disagreed with this approach. In our systematic analysis of genus Ophrys, we proposed to make a distinct group within the ensemble O. iricolor-O. fusca-O. lutea of the North African plants showing characters intermediate between Ophrys fusca and O. lutea, a group comprising at least two species, O. subfusca and O. battandieri. Observations made during a recent visit to Tunisia in the spring of 1999, a favourable year for orchids, have comforted us in this approach while suggesting that the North African group was probably more diversified than we had indicated. This note summarises these observations.
Characters of the Ophrys subfusca group. In the 1994 analysis, we defined the Ophrys subfusca group by lip proportions and pilosity arrangement. Within the O. lutea-O. fusca-O. iricolor complex, the width-vs.-length ratio of the lip and the characters that accompany this parameter, such as the angle of divergence of the sides of the lip at its base, are fairly evenly distributed along a continuous gradient. The O. iricolor group has the most lanceolated lip, with an average ratio of the lip width to length of the order of 0.64 to 0.70. The O. lutea group, on the contrary, has the most orbicular lip with a width to length ratio on the order of 0.95 to 1.20. The different lineages that have been confused under the name of O. fusca s.l. fit between these two extremes with ratios of width to length situated mostly between 0.72 and 0.86. The plants constituting the O. subfusca group fit, in their turn, between these two swarms with width to length ratios varying from 0.83 to 0.93 according to species (Table 1). The arrangement of lip pilosity is also constant in the lineages. In the Ophrys iricolor group, the brown pilosity covers, except for the macula and the base, the entire surface of the lip, including the extreme edge. In the O. fusca, O. funerea, O. attaviria and O. migoutiana groups, it is similar but the edge of the lip has a glabrous border, generally fairly narrow and of a constant width along the entire edge, sometimes very well defined. In the O. lutea group, the brown pilosity is generally surrounded by a wreath of long yellow or golden hairs, then by a wide glabrous border enlarged near the extremity of the lobes. In the O. subfusca group the wreath of yellow hairs is present and extends nearly to the edge of the lip, leaving only the glabrous, usually narrow, edge separated by an often progressive transition zone with an irregular contour. The system of lip coloration and the fairly large mammosities located in the proximal part of the macula place the Ophrys subfusca group near that of O. lutea. The form and the arrangement of colours of this macula tend either toward those of O. lutea, or towards those of O. fusca. The carriage of the lobes of the lip is, on the contrary, very peculiar to the O. subfusca group. The lateral lobes tend to curve inwards, passing under the plane of the central lobe which is carried forward and upwards. This arrangement can show up in other pseudophrys, notably in O. sicula, but never as consistently as in the species of the O. subfusca group.
Ophrys battandieri. Within the group, Ophrys battandieri is the species that comes closest, visually, to the Ophrys lutea group (Fig. 1). It is a robust plant, often very multiflorous, fairly often growing in multi-stem clumps. The flowers are middle to large sized. The lip is generally strongly arched and convex, the central lobe curved, the lateral lobes somewhat rolled under laterally. The main plane of the lip is held at about a 45° angle to the vertical. The lip is habitually very invaded by yellow, with a quite variable central brown zone, most often fairly reduced, and a generally small, rather pale macula. (Devillers & Devillers-Terschuren 1994). We observed Ophrys battandieri in April 1993 in the isolated hills of the north-west periphery of Tunis (Jebel Amar, Jebel Nahali), where it had also been seen in March by Vallès & Vallès-Lombard (1988) and where Gölz & Reinhard (2000) also documented it between 6 and 26 March. Vallès & Vallès-Lombard (1988) record it from the end of February to mid-March in the Tunis coastal area and Hervouet & Hervouet (1998) show an excellent illustration made at the beginning of March on the north coast of Cap Bon. Gölz & Reinhard (2000) photographed it in March near the south coast of the cape. For the interior mountains, information exists for both the chains that prolong the Saharan Atlas in Tunisia, the pre-Dorsal, centred on the Téboursouk Mountains, and the Dorsale, stretching from Jebel Chambi to Jebel Zaghouan, and particularly the parts nearest to the coast. We saw O. battandieri in April 1999 on Jebel Lanzarine, a massif very near to the hills of the Tunis periphery but which constitutes the north-eastern extremity of the Téboursouk Mountains. Gölz & Reinhard (2000) found the species farther south, near Téboursouk. In the Dorsale, the species was seen by Gölz & Reinhard (2000) in the beginning of March in Jebel Zaghouan, the northernmost massif. Its Algerian distribution is relatively difficult to circumscribe because of uncertainties surrounding the identification of species of the O. subfusca group. Battandier's material is from the Tellian mountains south-west of Algiers (Maire 1959). It is also most likely to O. battandieri that should be attached the plants seen by Murbeck at the Constantine coast, between Bougie and Cape Carbon and illustrated by him under the name of O. subfusca (Murbeck 1897, 1899; Maire 1931). The descriptions and illustrations of plants from the Nador de Médéa, in the Tellian mountains of Algiers, on which rest the description of O. gauthieri and of the plants of the Algiers area that furnished the type of O. pouyannei seem to be O. battandieri (Lièvre 1922; Maire 1931). Finally, plants from coastal stations of the Algiers area included in the sample of O. murbeckii studied by Ferlan (1956) and Stebbins & Ferlan (1956) also seem to refer to O. battandieri, to judge from the lip proportions and the drawings and photographs. The type of O. fenarolii is part of that ensemble.
The flattened lip (Fig. 2) shows a fairly rounded shape. At the base, the sides of the lip diverge from the axis at an angle of about 45°, then, after reaching the greatest width, curve immediately inwards. The extremities of the lateral lobes touch the central lobe or are separated from it by a very closed sinus, as in the Ophrys lutea group. The proportions of the lip are the nearest to the O. lutea group within the O. subfusca group, with an average ratio of width to length on the order of 0.93 (Table 1), situated almost exactly midway between those of the other species of the O. subfusca group (0.83 to 0.88) on the one hand, and those of O. sicula and O. lutea (1 to 1.1), on the other.
The yellow lip fringe is usually very large. The brown zone is often reduced and irregularly saturated, separated from the yellow edge by an often diluted or fragmented boundary. It is limited to a narrow central zone or, when it penetrates onto the lateral lobes, it almost always avoids the sinuses, leaving between it and the edge a large yellow hem. Within these limits, the lip design is extremely variable. This variation is probably largely responsible for descriptions of variable and chaotic hybrid swarms between Ophrys fusca and O. lutea in North Africa. These are often written by naturalists whose main experience of the Ophrys subfusca group is based on O. battandieri, a fairly widespread species, particularly in the coastal, most often visited, areas. They contrast with the accounts of botanists who encountered the species discussed below, mainly in the interior, and who were struck by their constancy.
Ophrys subfusca. Ophrys subfusca and O. murbeckii are the names most used in North Africa to designate plants intermediate between O. fusca and O. lutea, particularly by orchidologists who recognised their independance from these two taxa. Their precise applicability has nevertheless been the object of diverse interpretations. The basionym of Ophrys subfusca (Reichenbach fil.) Haussknecht, which dates from 1899, is Ophrys lutea var. subfusca Reichenbach fil. dating from 1851, of which the type exists only in the form of a 1:1 scale illustration of two lips, published with the description and representing flowers collected in Algeria (Baumann & Künkele 1986). The lips are extremely small (Fig. 3). Murbeck (1899), in the remarkable report of his journey in North Africa, adopted the combination Ophrys lutea subsp. subfusca, basing it explicitly on Ophrys lutea var. subfusca Reichenbach fil., and thus on the same type. He applies it to populations that he had seen himself and that probably comprise several entities. In part, he illustrates a plant that he collected in Algeria, between Bougie and Cape Carbon, between 19 and 25 February (Murbeck 1897); it has relatively large flowers and seems to represent O. battandieri. It is this illustration that is reproduced by Maire (1931). For the rest, he describes populations that he observed on the summit of the Kalaa el Harrat, at about 1200 metres, between 25 May and 5 June, and on the summit of Jebel Serj, at 1300 metres, on 9 June (Murbeck 1897). These populations are certainly small flowered since, among the arguments he enumerates to not call them Ophrys funerea Viviani, he does not mention the size of the flowers. Ophrys murbeckii is a name proposed by Fleischmann (1925) to replace Ophrys lutea subsp. subfusca Murbeck. Ophrys lutea subsp. murbeckii (Fleischmann) Soo uses it as a basionym. These are most likely illegitimate names (Gölz & Reinhard 2000) or, at best, objective synonyms of O. subfusca, based on the same type. They were nevertheless used very often, notably by Ferlan (1956), Stebbins & Ferlan (1956) and Baumann & Künkele (1986), to designate plants whose flowers have about the size of those of O. sicula, and are thus clearly larger than Reichenbach's type.
In the absence of recent observations of plants with small flowers of the Ophrys subfusca group, we thought it prudent to preserve the usage of Ophrys subfusca for all of the North African plants with a narrow lip and a large central brown zone (Devillers & Devillers-Terschuren 1994). However, in 1999, we found in the Tunisian Dorsale very homogenous populations corresponding perfectly to Reichenbach's type. It now seems that it is to these, and to the homologues they certainly have in the mountains of Algeria, that the name O. subfusca must be reserved. We examined two colonies on 14 and 15 April 1999. The first is located about 13 km west-south-west of Makthar, at about 1100 metres altitude, above a pass on Jebel Skarna, in an Erinacea anthyllis hedgehog-heath. It comprised about 50 individuals, at the onset of flowering to full flower. O. subfusca was by far the most abundant orchid; the other species present were about ten plants of the groups of O. fusca and O. obaesa, in full flower to end of flowering, and three plants of the O. subfusca group with larger flowers than those of O. subfusca, discussed below. The second station is in the Téboursouk Mountains, 5 km west of Téboursouk, at about 500 metres altitude, on grassy terraces planted with eucalyptus. There were about 600 individuals in full flower, the last flower usually open, accompanied by about 200 individuals of the species with larger flowers, some plants of O. vallesiana, 30 O. speculum and 10 O. bombyliflora.
At first glance, the flowers of Ophrys subfusca evoke the O. lutea group, in particular O. sicula, by the reduction of brown colour on a central zone leaving a large yellow periphery, especially on the lateral lobes. They show well, however, the characteristics of the group that O. subfusca typifies. The spread lip is narrow, with an average ratio of lip width to length of the order of 0.83. This ratio is the smallest in the O. subfusca group (Table 1). The arrangement of the pilosity is also that of the group, with a brown pilosity surrounded by a belt of white pilosity, then by a yellow pilosity that advances far onto the lateral lobes.
Within the group, Ophrys subfusca seems very well individualised (Fig. 1). It is a small, low plant with a thin stem and relatively large leaves. It is generally few flowered, carrying from one to four flowers, most often two or three. The flowers are very small, with a lip length on average 15% less than that of O. sicula (Table 1, Fig. 2). The main plane of the lip is usually held at around 15° below the horizontal, an angle a little more marked than in the species with somewhat larger flowers that shares the stations with O. subfusca, but much smaller than in the large-flowered species of the group. The lip is not very convex, much less than in the large species of the group or the species of the group of O. fusca s.l. The lateral lobes are flat, neither rolled inwards as in O.battandieri nor habitually lifted to form gutters as in the case of the species with somewhat larger flowers. The proportions of the spread lip make it the species with the most lanceolated lip of the group, the nearest, from this point of view, to O. fusca s.l. (Table 1). The sinuses are also the most open of the group. The lip design is very constant. The macula is relatively large, dark blue, usually undivided, nearly always terminating in a paler, little contrasting, blue-violet W. The brown pilosity forms a well defined brown zone advancing usually to the centre of the divisions of the central lobe and sometimes a little onto the lateral lobes. The sinuses remain generally narrowly surrounded by yellow. The demarcation between the brown zone and the yellow periphery appears sharp at a distance. On closer inspection it proves to be nearly always accompanied by a zone of reddish dilution, generally narrow. In rare individuals this zone of dilution extends farther onto the lateral lobes. The demarcation zone is not accompanied by a change in the abundance or the height of the pilosity. It is luxuriant in the brown zone and weakens progressively in the yellow periphery, often leaving a fairly narrow glabrous or subglabrous zone along the edge of the lip. The pilosity is always well marked, reddish or white, at the contact with the apex of the sinus. The underside of the lip is whitish in the centre, greenish on the periphery. The sepals are pale green, the petals, medium to smll, are about the same colour but with a slightly more yellow nuance and yellow papillae along the edge.
The similarity between Ophrys subfusca and the type illustrated by Reichenbach (1851) is striking. Fig. 3 visualises it. The comparison leaves little doubt as to identity between the plant described here and Ophrys subfusca (Reichenbach fil.) Haussknecht. Apparently the species has not been illustrated since Reichenbach (1851). In any case, we have found no photos nor drawings that might correspond to it. It is probable, nevertheless, that it is O. subfusca that Murbeck observed at high altitude on Kalaa el Harrat and Jebel Serj in May and June (Murbeck 1897, 1899). His stations are very close to our station at Makhtar, town where he stayed. However, some or all of his observations could also have applied to the following species.
Ophrys numida. In the two stations where we studied Ophrys subfusca, the very small flowered plant was accompanied by a plant obviously very similar, but with flowers a little larger, a more robust carriage, and a later flowering time. Above the pass on Jebel Skarna we found two individuals that had opened their first flower and one beginning to open. Several plants in bud were present. At the Téboursouk station, 200 plants were in full flower, the last flowers often in bud. The taxon concerned has flowers remarkably close to the size of those of O. sicula, with an average lip length of the order of 8.5 to 9.0 mm, but with more elongated o lip proportions (Table 1). It is certainly this species that was studied in the Hodna mountains of Algeria by Stebbins & Ferlan (1956), under the name of O. murbecki, the homogeneity of which they noted. Because of the avatars of Reichenbach's name and its synonyms, the species does not seem to have a valid name. We propose to call it Ophrys numida (Annexe 1), in reference to its interior distribution, distinct from that of coastal species, with presently the known stations framing the Roman province of Numidia and, for Tunisia, at least, adjacent to the Numidian fortifications of Makthar and Dougga.
Ophrys numida is very near to O. subfusca (Fig. 1). It is a much more robust plant, taller, with a thicker stem, more multiflorous, carrying from 3 to 8 flowers, most often 6 or 5. Its flowers are a little larger than those of O. subfusca, with a lip length 15% greater on average, like that of O. sicula (Table 1). The main plane of the lip is generally held horizontal, not about 15° below the horizontal as in O. subfusca. The lip is larger (Table 1) and less narrow, with an average ratio of width to length of 0.88 for 0.83 in O. subfusca (Table 1). The lip is little convex, as in O. subfusca. The lateral lobes are broadened and lift up at the edge, often forming a marginal gutter, as in many individuals of species of the O. lutea group. It is this character, combined with the tall carriage of the plant and its spaced-out flowers held out horizontally, that confers on O. numida its most characteristic appearance. The sinuses that separate the lateral lobes from the central lobe are narrower than in O. subfusca. They are nevertheless much more open than in the large species of the group, especially O. battandieri. The lip design is very similar to that of O. subfusca and just as constant. The macula is similar. The brown pilosity forms the same well defined zone advancing to the centre of the divisions of the central lobe. The sinuses are surrounded by yellow, in general a little more widely than in O. subfusca. The demarcation between the brown zone and the yellow periphery is accompanied by the same often narrow zone of reddish dilution. This zone of dilution extends much more often onto the yellow periphery, particularly onto the lateral lobes, sometimes giving the flower a look evocative of O melaena. The lip pilosity is shorter and less luxuriant than in O. subfusca. It penetrates widely into the yellow edge, as in O. subfusca, but it is a little shorter there and leaves a larger glabrous or subglabrous, papillose, marginal zone. Contrary to that of O. subfusca, the pilosity is not very evident along the edges and the apex of the sinuses. The underside of the lip often has a pale pink spot in the centre. The sepals are the same colour as those of O. subfusca, the petals, on the other hand, are of a franker green and a little darker than the sepals, without yellow. They are very small.
There are apparently few illustrations of Ophrys numida published. The representations and measurements of Stebbins & Ferlan (1956) correspond well to those of the plants that we have seen and indicate the presence of the species in the Algerian interior massifs, in particular in the Hodna Mountains, where they studied three stations. It is probable that other data from Algeria included by Maire (1931) under Ophrys subfusca pertain to it. For Tunisia, the species is perhaps included in Murbeck's (1897, 1899) observations from Kalaa el Harrat and Jebel Serj. It was illustrated by Gölz & Reinhard (2000), who clearly encountered it in the Téboursouk region (their station 2, perhaps their station 4). Its known distribution in the country is thus limited to the Téboursouk Mountains and to the central region of the Dorsale, around Makthar.
Ophrys aspea. There exists in Tunisia at least one other taxon of the Ophrys subfusca group, that differs, like the two preceding species, from O. battandieri by an extensive brown zone and a macula more similar to that of O. fusca s.l. then to that of O. lutea, but whose flowers are much larger than those of O. subfusca and O. numida. We observed it in 1993 on the north coast of Cap Bon, in the Korbous area, and provisionally attached it to O. subfusca, the name under which we illustrated it (Devillers & Devillers-Terschuren 1994). The discovery in 1999 of the true O. subfusca demonstrated this approach to be erroneous. The ophrys of Cap Bon, that we saw again in two stations around Korbous in 1999, is a very stable and very constant form well differentiated from O. subfusca. It does not seem to have a valid name and we propose to call it Ophrys aspea (Annexe 1). The adjective is derived from Aspis, Greek name of Clypea, the present Kelibia, town on the east side of Cap Bon, but whose name also seems to have served to designate the surrounding regions of the Cap Bon peninsula, while other classical names that refer to this localisation concern only the promontory (Hazlitt 1851; Smith & Theil 1884). All the known observations are in deed from the peninsula.
Ophrys aspea (Fig. 1) is a low but robust plant, few flowered, carrying two to four flowers, most often three. The flowers are large for the group, with an average lip length a little larger than that of O. battandieri (Table 1, Fig. 2). The main plane of the lip is generally held at more than 45° below the horizontal, often nearly vertical. The lip is very convex, the lobes lowering strongly before, often, lifting again at the margin; the plane of extension of the lateral lobes is situated very much below that of the central lobe. The proportions of the flattened lip are intermediate between those of O. battandieri and O. subfusca, near to those of O. numida (Table 1). The sinuses are relatively closed, but less so than those of O. battandieri. As in O. subfusca and O. numida, the lip design is very constant. The macula is usually large but not always undivided, sometimes grooved at the base by the beginning of a ciliated furrow, as in the species of the O. funerea group, dark blue, fairly rarely terminated by a paler W. The brown pilosity forms a well defined zone whose extremity follows, at some distance, the shape of the edge of the central lobe, and whose sides follow also, usually at a greater distance, the shape of the lateral lobes. This brown zone encompasses or touches the apex of the sinuses, a character specific to the species. The demarcation between the brown zone and the yellow periphery is usually very sharp, without a zone of reddish dilution. The demarcation zone is accompanied by a corresponding change in the colour of the pilosity, but not in its abundance or height. Strong in the brown zone, it weakens progressively in the yellow periphery, but usually without leaving a glabrous region at the edge of the lip. The underside of the lip is yellow to yellowish, bright yellow near the edges. The sepals are pale green, the petals, rather large, generally strongly coloured fairly bright yellow, above as well as below, with a fine, even brighter, yellow edge.
We examined 30 plants of Ophrys aspea in three stations on Jebel Korbous, one along the sea south of Korbous on 5 April 1999, the two others north-west of Korbous, near the crest line of the jebel on 5 April 1993 and 7 April 1999. It is clearly the same species that was found by Baumann (1975) at 5 km east of Hammamet, on the south coast of the Cap Bon peninsula, near its base, on 26 March 1973. He gives an excellent illustration of it (Baumann 1975: 713, Fig. 2), accompanied by a series of flower analyses (op. cit.: 135, Fig. 1) and notes the specific originality of it. The species was again documented by C. Blanchon (in Vallès & Bournérias 1990: 17, sub nomen Ophrys subfusca) for Aïn Oktor, near Korbous. Among the plants of the O. fusca-O. lutea complex illustrated by Gölz & Reinhard (2000), the species is represented in a very characteristic way by plants of station 14, located near Somaa, on the southern slope of Cap Bon, not far from Baumann's station. All the stations in which the species has been found so far are thus located in the piedmont or on the slopes of the domes at the base of the peninsula of Cap Bon, probably on sandy calcareous tertiary deposits (Haack & Lautensach 1931; Krenkel 1938: 1546-1549).
Limits of the Ophrys subfusca group. It is now clear that the Ophrys subfusca group is, as undoubtedly are most of the lineages that form the O. fusca-O. lutea-O. iricolor complex, a diversified group. In Tunisia it comprises at least four species, divided between three rather distinct branches. O. battandieri evokes by its macula and overall appearance O. lutea. O. aspea is rather similar in shape, although more massive and bulbous, but extremely distinct in pattern. O. numida and O. subfusca are closely related to each other and combine an appearance evocative of O. sicula with traits recalling the small species of the O. fusca group. It is probable that other species of the group exist in North Africa, particularly in the vast complex of Algerian massifs, and undoubtedly elsewhere.
Allocation of species to the Ophrys subfusca group is not always straightforward. Indeed, the characters that differentially define the O. subfusca group against the O. lutea and O. fusca groups, are of a fundamentally intermediate nature between the attributes of these two groups. The best criteria to orient the analysis towards a species of the O. subfusca group are undoubtedly the proportions of the lip, comprised between those of Ophrys lutea and those of O. fusca, a pilosity that, as in O. fusca, extends to the edge of the lip or nearly, encompassing a yellow margin whose internal limit does not correspond to a discontinuity of the pilosity, large mammosities at the base of the macula, which, as in O. lutea, end distally by an abrupt change in the curve that often gives the impression of a break, and a tendency also shared with the O. lutea group for the petals and the underside of the lip to be brightly coloured.
Among European species, some or all of the characters of the Ophrys subfusca group are encountered in various Sicilian species, as shown in the photos published by, notably, Danesch & Danesch (1972), Paulus & Gack (1990), Delforge (1994), Melki (in Mark 1996). We have seen large populations of two of these species in the Iblei Mountains in April 1986. One has relatively small flowers and large sinuses, the other has rather large flowers, massive proportions and large sinuses. One of them is probably the species discussed by Paulus & Gack (1990) under the name of Ophrys (florentina-)fusca. We had already called attention to these Sicilian populations previously (Devillers & Devillers-Terschuren 1994), leaving, however, their relationship to one or another African species unresolved. A better understanding of the African group now establishes that the two Sicilian forms are quite distinct from the four African forms found to date. Their allocation to the group remains probable, although their lips seem somewhat narrow.
The external group whose affinities with the Ophrys subfusca group are the most obvious is the O. lutea group. Among the species that compose it, it is perhaps O. melaena that shows the greatest resemblance, in spite of the very different proportions of the flowers. In particular, the small-flowered populations on Monte Gargano, perhaps specifically distinct from those of Greece, show an obvious superficial resemblance with O. numida. In the remainder of section Pseudophrys ties with O. subfusca are somewhat less evident. However, the presence of very marked basal mammosities in the O. obaesa group, in particular in O. pallida, is striking. This group seems the closest to the O. lutea-O. subfusca ensemble, considered collectively, the allies of O. subfusca on the one hand, those of O. pallida on the other, constituting the link.
Key-words: Orchidaceae, genus Ophrys, Ophrys lutea, Ophrys fusca, Ophrys subfusca, Ophrys numida, Ophrys aspea, Ophrys battandieri, Ophrys murbeckii, Ophrys gauthieri, Ophrys pouyannei, Ophrys lievreae, Tunisia, Dorsale.