Devillers, P. & Devillers-Terschuren, J. 2000. The type of Ophrys eleonorae. Extended summary. Royal Belgian Institute of Natural Sciences website, www.naturalsciences.net/cb. [article published in French in Natural. Belges 81 (Orchid. 13)].
The considerable progress that has taken place over the past 20 years in the understanding of the taxonomy, the biogeography and the ecology of European orchids is largely the result of methodological developments in the collecting and recording of field data. Indeed, orchids lend themselves much less well than other groups to the analysis of herbarium specimens, which often poorly display the characters of floral structure and colour indispensable to diagnosis. This reality leads most orchid students to omit this material in their discussions. Even nomenclatural debates are waged without examination of the type material or of valid representations of it. If this tendency is understandable in the case of old, relatively inaccessible types for which the conservation condition and the quality of the accompanying notes could only lead to relatively uncertain conclusions, it appears somewhat summary in the case of recent material. A relatively extreme case of offhand treatment of type material has brought us to prepare this note. In 1991 we described (in Delforge et al. 1991), under the name of Ophrys eleonorae, a plant of the O. iricolor group whose presence had long been known in Sardinia (e.g. Sundermann 1980), that we had observed since April 1985 and presented before the Section "Orchidées d'Europe" of the Naturalistes Belges on 30 November 1985 (Coulon 1988). This treatment seemed necessary in the light of a definitive article on Sardinia (Gölz & Reinhard 1990) in which the authors, after a detailed biometric analysis, negated the existence of an independent species in Sardinia and recognised only a complex intermediate between O. iricolor and O. fusca. Written in the optic of Delforge's (1994) forthcoming field guide, the description was published in a succinct form, the more detailed characterisation and supporting photos being left for the guide and for the synthesis that we were ourselves preparing (Devillers & Devillers-Terschuren 1994). In two recent publications, Paulus & Gack (1995, 1999) have, while confirming the existence of a Sardinian entity belonging to the Ophrys iricolor group, marked their preference for the subspecific treatment of this taxon. It is a taxonomic choice that we do not share but which is of course perfectly legitimate. However, they also decided, without consultation of the type material, that the binomial O. eleonorae applies to hybrids between the species of the O. iricolor group and a species of the O. fusca group rather than to the species of the O. iricolor group itself. It is this abusive interpretation that we wish to dispel.
The type of Ophrys eleonorae (Fig. 1) was collected on 29 April 1986 from a homogeneous population of 15 individuals. As in all the plants of the colony, the type presented in a pronounced way the raised plateau of the base of the lip, expanded into crests sloping outwards, diagnostic of the Ophrys iricolor group (Devillers & Devillers-Terschuren 1994), as well as the red underside of the lip. The flowers were large, with the narrow shape typical of the O. iricolor group (Fig. 1). All the plants of the type locality of Ophrys eleonorae presented the characters illustrated by Paulus & Gack (1995) as an exemple of what they consider "pure" O. iricolor in Sardinia (e.g. plate 1, E, plant on the left) and none of the characters of their hybrid (e.g. plate 1, E, plant in the middle). The station was similar to a larger colony (100 individuals) observed in 1985 at Capo Testa, near Santa Teresa Gallura, of which various aspects are illustrated (Fig. 1; cf. also Devillers & Devillers-Terschuren 1994: 307 and Delforge 1994: 299). We have, as have Paulus & Gack (1995), encountered a number of plants likely to be hybrids, in particular, in the Sinis and in the hinterland of the Golfo di Orosei near Dorgali and near Oliena, but not at Capo Testa, nor on the coasts of the Golfo dell'Asinara, nor on Monte Albo, nor at the type locality. There is thus no doubt that Ophrys eleonorae is, at the species rank, the correct name of the Sardinian member of the O. iricolor group and cannot be applied to its hybrids with other taxa.
When compared to other Sardinian taxa of the section Pseudophrys, Ophrys eleonorae obviously manifests the characters of the group of O. iricolor, in particular the extremely characteristic basal plateau with its lateral escarpments that are reddish at the base, the pilosity extending uniformly to the edge of the lip, the long and narrow proportions of the labellum, the red coloration of its underside, the macula habitually of an intense blue. It is when O. eleonorae is compared to Aegean O. iricolor that appear the differential characters that we had mentioned in the original description: a taller carriage combined with a larger number of flowers, a lesser saturation of the red pigmentation of the labellum, giving a more orange, less carmine coloration to the basal crests, a more blue-grey, less ultramarine tint to the macula. These last characters are well seen on the flower of Fig. 1, as well as on the one Paulus & Gack (1995) have chosen to illustrate "pure" O. iricolor (plate 1, E, plant on the left).
The most obvious and most constant character of Ophrys eleonorae is the restriction of the red zone to the central part of the underside of the lip, leaving a border, sometimes narrow but always very well delimited, between it and the edge of the lip. This character can be seen in Fig. 1 as well as on the flowers illustrated by Gölz & Reinhard (1990, plate 1, b), particularly the one that shows in the most extreme way the characters of O. iricolor. We have found this character in all the plants of O. eleonorae that we have examined. It is made all the more striking by the fact that the limit between red centre and green border is always very abrupt, without attenuation, and that the existence of the yellow-green border is manifested by the presence of a fine yellow line observable on the lip seen from above (Fig. 1; Gölz & Reinhard 1990, plate 1, c). This condition contrasts with that characteristic of O. iricolor, very well illustrated, for example, by Delforge (1994: 298), which combines an entirely red underside with a fine red line observable from above.
Gölz & Reinhard (1990) analysed 43 quantitative characters of Sardinian populations of large-flowered Ophrys resembling O. iricolor, on the basis of 15 specimens, and showed a range of values situated between those characteristic of Aegean O. iricolor on the one hand, of French and Spanish populations of Ophrys fusca s.l., on the other. Although they find an intermediate swarm between the two entities, their statistical results (their Table 3), well illustrated by their Figs. 3 and 4 show a variability of the sample that is, for most of the characters, not greater than that of O. iricolor, and does not seem to be accompanied by bimodality, so that it is possible that a large part of their "hybrids" represent the interval of variability of O. eleonorae. It is clear, however, that true hybrids do exist between O. eleonorae and one or another taxon of the O. fusca or O. funerea complexes. We have seen several plants of this type and the illustration offered by Paulus & Gack (1995: plate 1, E, plant in the middle) is clearly correct, as shown by the proportions of the lip and the reduction of the basal structures.
Ophrys eleonorae was first included in O. iricolor, often within a wide concept, which clearly encompassed O. fusca s.s. or some of its allies (Danesch & Danesch 1969; Sundermann 1980; Kohlhaupt 1981; Pignatti 1982; Davies et al. 1983). Later, it became habitual to exclude Sardinia from the range of O. iricolor, while the latter was more narrowly defined (Baumann & Künkele 1982, 1988; Kurze & Kurze 1984; Delforge & Tyteca 1984; Buttler 1986, 1991). The two approaches coincided at the beginning of the 90's. Scrugli (1990) and Liverani (1991) included O. iricolor in the Sardinian flora; Giotta & Piccitto (1990) recognized only O. fusca and Gölz & Reinhard (1990) described entirely hybrid populations which they merged in O. fusca. In 1991 we proposed (in Delforge et al. 1991) to treat the Sardinian entity as a species allied to O. iricolor but distinct, a treatment adopted by Delforge (1994) and explained in our synthesis of genus Ophrys (Devillers & Devillers-Terschuren 1994). Paulus & Gack (1995, 1999), while confirming the individuality of a Sardinian taxon, suggested to treat it at subspecific rank.
The notion of subspecies differs fundamentally between zoological and botanical usage. As presently understood by zoologists, the subspecies, to which trinominal nomenclature applies, is based on a biological species definition. Its use supposes an absence, observed or presumed, of reproductive isolation. The botanical subspecies, to which trinominal nomenclature cannot be applied, is based on a morphological species definition. It covers taxa having the characteristics of species, but presenting a degree of morphological divergence less than that which characterises these (e.g. Davis 1965). The decision of treating as a species or as a subspecies a taxon having a discrete geographical range and a degree of observable discontinuous morphological differenciation is necessarily subjective. It depends on a taxonomist's evaluation of the threshold of divergence necessary to recognise two distinct species. The rank of subspecies, as that of subgenus, is only offered by the Code of Botanical Nomenclature as a supplemental category, to be employed when a need for additional subdivisions between species and variety ranks (Article 4.2) is perceived. As such it is not very useful in the framework of analysing evolutive biological or ecological relationships of orchids, and it hardly has had the favour of recent monographists (e. g. Cribb 1997, 1998).
If the decision to separate two taxa at the subspecies rather than the species rank is thus arbitrary and subjective, it seems nevertheless useful to maintain some comparative coherence in the treatment of a group. It is also essential to evolutive comprehension that formal subspecific subordinations do not mask real phylogenetic relationships. The group of Ophrys iricolor as it is presently known, comprises at least five taxa, Ophrys iricolor Desfontaines, O. mesaritica Paulus, C. & A. Alibertis, O. eleonorae J. & P. Devillers-Terschuren, O. vallesiana J. & P. Devillers-Terschuren, O. astypalaeica Delforge. These five taxa differ from each other in a more or less equal manner. Flower size groups large-flowered Ophrys iricolor and O. eleonorae, on the one hand, O. mesaritica, O. vallesiana and O. astypalaeica, small-flowered for the group (Paulus et al. 1990; Devillers & Devillers-Terschuren 1994; Delforge 1997a), on the other. Lip pattern, probably phylogenetically more important, regroups, on the contrary, O. iricolor and O. mesaritica who have the lip entirely or partially invaded by red, but without a distinct, sharp concentric zonation (cf plate 1, 2A in Paulus et al. 1990), from O. eleonorae, O. vallesiana and O. astypalaeica which present a clear cut zonation, with a red centre and an abrupt, well delimited, yellowish border (Devillers & Devillers-Terschuren 1994; Delforge 1997a). This under-lip pattern explains the thin red line along the edge of the lip of O. iricolor, the thin yellow edge of O. eleonorae, O. vallesiana and O. astypalaeica, and the variably coloured edge of O. mesaritica, corresponding to a pattern akin to that of O. iricolor but less saturated and more irregular (Alibertis and Alibertis 1989; Paulus et al.1990). O. mesaritica and O. eleonorae bloom very early or early in the local flowering waves (Paulus et al.1990; Devillers & Devillers-Terschuren 1994), O. astypalaeica just before O. iricolor (Delforge 1997a), the latter and O. vallesiana relatively late in these waves (Devillers & Devillers-Terschuren 1994). Some of these taxa are sympatric, sometimes separated by flowering time and perhaps by pollinator. Thus, O. iricolor and O. mesaritica seeem to coexist in Crete (Paulus et al.1990) and perhaps on Lesbos (Biel 1999), O. astypalaeica and O. iricolor on Astypalea (Delforge 1997a, 1997b), O. vallesiana and O. eleonorae in Tunisia (Hervouet & Hervouet 1998).
The cladistic relationships within the ensemble are not yet known but most likely unite sympatric rather than allopatric species. Ophrys mesaritica seems in fact morphologically closer to O. iricolor than one or another of them is to O. eleonorae or O. vallesiana. The two Cretan species share a system of coloration of the underside of the lip very distinct from that of the North African and the Cyrno-Sardian taxa. It is probable that they are issued from a local speciation event by adaptation to the emergence of different temporal waves of pollinators. In the same way, O. eleonorae and O. vallesiana seem relatively near, differing mostly by flower size. Besides, it is possible that some species, and in particular O. eleonorae, have incorporated in their gene pool an important part of elements coming from another species of the ensemble of O. fusca s.l., as suggested by Gölz & Reinhard (1990), which would make their attachment at the subspecific rank to O. iricolor s.s. all the more unhappy
In summary, it is not coherent, on the basis of morphological divergences, to treat some of the taxa of the Ophrys iricolor group at the species rank and others at the subspecies rank. They differ by few but well-marked characters. From a phylogenetic point of view, possible regroupings should await a better understanding of the succession of separation events but it is unlikely that these would suggest a link between O. eleonorae and O. iricolor, excluding, for example, O. mesaritica. Finally, a generalisation of the subspecies rank to all the taxa of the group has little utility, carries no information, and is undesirable in the presence notably of sympatric forms. We prefer thus to treat the five taxa of the O. iricolor group as distinct species. Among them, O. eleanorae differs from O. iricolor by an abruptly marked character of the coloration system, it is geographically very separated and and phylogenetically perhaps partly independant. It is particularly inappropriate for it to be subordinated to it as a subspecies.
Key-words: Orchidaceae, genus Ophrys, Ophrys eleonorae, Italy, Sardinia.