Pitcairnia feliciana (A.Chev.) Harms & Mildbr.
Literature references:
*move your mouse pointer over the page numbers to see comment
Comments:
- Within the Bromeliaceae hummingbirds form the most important group of pollinators (Martinelli, 1994). Despite the lack of detailed field observations it can be hypothesized that Pitcairnia feliciana is an ornithophilous species which would indicate that the ancestor likewise was bird-pollinated. If the ancestor had been visited by other pollinators it is hardly conceivable that P. feliciana switched towards bird pollination in Africa. Several floral traits of P. ,feliciana are in support of the bird pollination hypothesis, e.g. orange-red colour, no fragrance and copious nectar. Birds as pollinators are rare in West Africa, however, several
sunbirds (the genus Nectarinia) occur. Field observations are highly recommended to solve the question whether in P. feliciana sunbirds have replaced the hummingbirds as pollinators. Alternatively it could be possible that P. feliciana is pollinated by butterflies, but here likewise no observations have been made to date.
The Bromeliaceae are one of the most characteristic plant families of the New World. They comprise more than 2600 species in 56 genera (Smith and Till, 1998) frequently forming perennial rosette herbs which grow terrestrial, saxicolous or epiphytic. However, with Pitcairnia feliciana (A. Chev.) Harms & Mildbr. there exists one remarkable geographic outlier in West Africa more than 3000 km separated from the normal distributional range of bromeliads. Despite its fascinating biogeographic uniqueness remarkably little is known about this African bromeliad. This fact is due to the remote occurrence of P. feliciana in the Guinean Fouta Djalon highlands (Fig. 1) where the species grows in localities that are not easily accessed. Another reason is its relatively late discovery in the 1930s and the often disputed nature of this species as of being introduced by man. Before the discovery of P. feliciana by Henri Jacques-Felix nobody would have expected that a member of the neotropical Bromeliaceae occurs in tropical Africa. It was therefore only consequent that Auguste Chevalier (1937) to whom material was sent for the description of "...la plante tres remarquable..." placed it as a new genus (Willrussellia) within the Liliaceae. However, only one year later Harms and Mildbraed (1938) recognized the true nature of the material and corrected Chevalier's misinterpretation and placed the only African bromeliad in the genus Pitcairnia.
MATERIALS AND METHODS
Study material of Pitcairnia feliciana was collected in Guinea (near Kindia, on slopes of Mt. Gangan, c. 600 m a.s.l., Porembski No. 119) during the rainy season in 1992. Moreover, plants cultivated in the Botanical Garden of Bonn were investigated (Welz No. 12804). For light microscopy, living material was fixed in FAA, dehydrated by standard methods, and embedded in Paraplast for serial sections with a rotating microtome. Staining was implemented with safranine, auramine and astra blue. For scanning electron microscope studies, plant material was dried using the critical point method. The material was coated with gold (Balzer SCD 040 sputter) and observed in a SEM (Cambridge Stereoscan 200). Herbarium material of P. feliciana was studied at P. Determination of stable carbon isotope composition (d 13 C-values) was by mass spectroscopy (following Ziegler et al., 1976).
ECOPHYSIOLOGY
According to data obtained from a 13 C-analysis (conducted by H. Ziegler) it is obvious that P. feliciana is a C3 plant. Judging from the value of PC (PDB)-31.88%-it seems unlikely this species is even theoretically capable of conducting the CAM-mechanism of carbon fixation. This result agrees with other measurements of photosynthesis in Pitcairnia that have shown this genus is generally characterized by the C3 pathway of carbon fixation (e.g., Smith, 1989).
PHYTOGEOGRAPHICAL ASPECTS
The Bromeliaceae form one of the most characteristic families of the neotropics. According to Smith and Downs (1974, 1977, 1979) their area ranges from Virginia (c. 38°N) to Argentina (c. 44°S). With the exception of P. feliciana all species of Bromeliaceae occur in the tropical and subtropical regions of the neotropics where this family constitutes one of the most important components among vascular epiphytes. A similar distribution pattern (i.e. a disjunct occurrence between the New World and Africa) is demonstrated by the Rapateaceae with Maschatocephalus dinklagei (distributed in wet forest over poor soils in Sierra Leone, Liberia, Cote d'Ivoire) being the only Old World member of this family. Another well known example of a large and typical neotropical family being represent in the Old World by only a few disjuncts are the Cactaceae (i.e. Rhipsalis baccifera, see Barthlott, 1983).
Pitcairnia feliciana was discovered in 1937 by the French botanist Henri Jacques-Felix who collected extensively in Guinea where this species grows in the Fouta Djalon massive. his discovery of a bromeliad in tropical Africa was completely unexpected and not only resulted in its misidentification as a strange member of the Liliaceae (Chevalier, 1937) but also raised questions concerning the origin of this species. The statement of Jacques-Felix (pers. com. 1996) "...plusieurs auteurs qui pensent que ce Pitcairnia a ete recolte pres du littoral! et serait une banale introduction recente et accidentelle..." vividly illuminates the difficulties within the botanic community to accept the fact of the natural occurrence of a bromeliad in the Old World. Today there can be no doubt that P. feliciana is a truly indigenous African species since it occurs far inland at localities relatively uninfluenced by man. Therefore a recent introduction by man can certainly be ruled out.
It is still a matter of speculation from which neotropical region the precursor of P. feliciana has originated and by which means of dispersal the long way over the Atlantic was conquered. The most plausible assumption is that a chance long-distance dispersal event of seeds perhaps by means of a tropical storm resulted in the transport to the Guinean Fouta Djalon highlands. The palaeozaic sandstones of the Fouta Djalon massive form flat-topped mountains which are dominating landscape elements on the high plateaus which are typical for this region. Together with Mt. Nimba, Pic de Fon and the Ziama massif, the Fouta Djalon highlands are part of the "dorsale guineenne". Due to the occurrence of phytogeographically and taxonomically distinct taxa the sandstone outcrops of the Fouta Djalon form a center of endemism (Porembski et al., 1994). Remarkable are taxa (e.g. the tree-let like Cyperaceae Microdracoides squamosus) occurring disjunct in the Fouta Djalon region (also in adjacent Sierra Leone) and more than 1500 km to the southeast in Nigeria and Cameroon. One can assume that the area of these species is of relictual character and has been more continuous prior to climatic fluctuation during the Pleistocene. Thus the Fouta Djalon highlands may have acted as refuges during times of drastic global change. This fact may explain why P. feliciana which is restricted to rock outcrops is to be found there but nowhere else in West Africa which is otherwise largely devoid of rock massives attaining a similar altitude and surface extension like the Fouta Djalon.
The classic locations of P. feliciana are found around the small town of Kindia on the slopes of Mt. Gangan {Fig. 1). Here the plants typically grow in narrow, soil-filled crevices together with a number of geophytic species. The climate at this locality is highly seasonal with a dry season of five months, frequent occurrence of mist and annual precipitation around 1600 mm.
CONCLUSIONS
According to Winkler (1986) lithophytic bromeliads are intermediate between truely terrestrial and epiphytic species. On the other hand already Tietze (1906) proposed that the ancestral bromeliads were probably terrestrial, pitcairnioid-like plants which grew in exposed and possibly elevated regions of South America. This characterization holds remarkably true for the West African P. feliciana too. One can only speculate whether the fact of considerable phylogenetic antiquity has increased the chance (i.e. with increasing age of a taxon the probability of chance transports raises) of a putative P. feliciana ancestor of being transported by long distance dispersal across the Atlantic Ocean.
Within the genus Pitcairnia its only African representative forms an isolated and clearly distinct species that bears a number of characters not known up to now to occur in other Pitcairnias. For example the Pitcairnioideae are characterized by possessing a relatively large, solitary tongue-like ligula (Brown and Terry, 1992). In this respect P. feliciana is clearly diverging from all neotropical species in possessing instead two tooth-like appendages. A first survey on ligula-types in the genus Pitcairnia has revealed a large morphological diversity (own unpubl. data), however, the ligula-type of P. feliciana is absolutely unique. Brown and Gilmartin (1989) and Schill et al. (1988) provided overviews on stigma types in Bromeliaceae. The stigma lobes of P. feliciana bear papillae that do not occur in other species of Pitcairnia (Fig. 10) nor in any other bromeliad further emphasizing the isolated position of this species. To illustrate the distinctiveness of P. feliciana the morphology of the seeds and pollen is mentioned as last examples. In having testa cells with perforated outer periclinal walls, P. feliciana is differentiated from all other Pitcairnias_ Among the Bromeliaceae only one similar case has been reported within the genus Puya (Gross, 1988). Despite the fact that comparative analyses of Bromeliaceae pollen covered only a small number of Pitcairnia species (Ehler and Schill, 1973; Halbritter, 1988) it can be stated that the pollen of P. feliciana is deviating with respect to exine sculpture. The distinctiveness of P. feliciana indicates a long history of separation. However, the species is certainly not a Gondwanian relict surviving today in a relatively small refuge in the Guinean Fouta Djalon. It rather can be assumed that a precursor reached Africa via chance long-distance dispersal after the separation of South America and Africa. It is still not possible to draw a decisive conclusion as to which neotropical Pitcairnias might be the closest relatives of P. feliciana. Already Harms and Mildbraed (1938) suggested that the Caribbean P. fuertesii or P. pungens (Ecuador, Peru) could be closely related, however, without mentioning any arguments in support of this assumption. In contrast to this we speculate (based on unpubl. data about morphology and anatomy of various Pitcairnia species) that it seems to be more probable that P. feliciana dates back to a saxicolous eastern Brazilian species. More comparative data on, e.g. seed and pollen morphology and additional molecular characters are needed to solve this problem. —See Harvard Pap. Bot.
