genus Glomeropitcairnia Harms
Literature references:
Comments:
- Glomeropitcairnia, an Enigmatic Tillandsioid Genus by A. J. Gilmartin and G. K. Brown in J. Brom. Soc. 36(3):104-6. 1986
Key to the Species of Glomeropitcairnia
1.Inflorescence amply tripinnate; spikes recurved. Lesser Antilles. penduliflora.
1 Inflorescence bipinnate; spikes erect. Northeastern Venezuela, Trinidad. erectiflora
Among the largest, though by no means the most beautiful, members of subfamily Tillandsioideae are Glomeropitcaimia erectiflora Mez and G. penduliflora (Griesebach) Mez. These two are the only known species in this genus, a relative of the better known genera of the subfamily which includes Catopsis (19 species), Guzmania (126 species), Mezobromelia (4 species), Tillandsia (402 species), and Vriesea (249 species).
Glomeropitcaimia remains an enigma and is of interest to phylogeneticist: because at least three of its features, recognized by early workers such as Mez (1935), are not found in other members of the subfamily: seeds appendaged at both ends, capsules partially indehiscent, and ovary partly inferior. On the other hand, certain characteristics of the two "glomeros," including floral and leaf morphologies and foliar scale anatomy are typical of Tillandsioideae.
The sheer size of these plants and their appearance merit some pretty strong adjectives. Astonishing, came to mind in August, 1985 when I (AJG) first saw a plant of G. penduliflora on soggy Chances Peak on the island of Montserrat. Later, I was able to photograph them on Dominica where this species has been collected thus far four times by various people (1888, 1959, 1965, and 1967). In the back cover photograph, Dominican workers are shown displaying an individual attached to the severed stem of a tree fern on which it is growing epiphytically.
The two species are each known from a very few localities. Glomeropitcairnia erectiflora is known from Trinidad and Isla Margarita, Venezuela. G. penduliflora, the largest of the two is known from Dominica, Guadeloupe, Martinique, and Montserrat. The first and last islands are part of the British West Indies, while Guadeloupe and Martinique are in the French West Indies. Because of their limited distribution, if for no other reason, these appear to merit their status of endangered or threatened species. They are difficult to propagate either from seed or from field-collected plants (Luther). They do not seem to set pups, at least among hundreds of G. penduliflora no offshoots were found.
A good place to examine the astonishing G. penduliflora is on the moist slopes near Castle Bruce Road, Dominica. Frequently, individual glomeros are perched high on host trees (fig. 3). The enormous size was noted by Picado (1913) who experimentally determined that the plants could retain up to 20 liters of water within the leaf rosette. Their rosettes are unlike those of any other members of subfamily Tillandsioideae that we have seen in terms of size and number of leaves ( at least 60) and the longevity of the old leaves. The inner leaves of the rosette are green surrounded by several layers of apparently dead leaves. There was no indication that these leaves were unhealthy, they simply retain the old leaves upright around the periphery of the rosette.
According to Mez, and Smith and Downs (1977), seed of tillandsioids other than Glomeropitcairnia all have a plumose appendage that is either apical (Catopsis) or basal (Guzmania, Mezobromelia, Tillandsia, and Vriesea). Seed appendages of Glomeropitcairnia are not particularly plumose and extend both basally and apically on the seed. The appendages of G. penduliflora give the appearance of being immature even though the seed itself seems to be fully developed. They resemble seeds from immature capsules of Tillandsia in that appendages are little divided, i.e. not plumose. This feature, in combination with the incomplete dehiscence typical of the Glomeropitcairnia capsule (another possible immature feature), suggests seed neoteny. Neoteny, the expression of juvenile traits in conjunction with an otherwise mature plant, can happen when something bio-chemically or genetically suppresses the normal course of development. In this case, perhaps, seed (which is the first stage of the new generation of the plant) is able to mature while the surrounding appendage and capsule tissue retain juvenile traits. More important, neotenous development may be maintained in the lineage generation after generation, that is, it may be incorporated into the heredity of the species.
We have no way of knowing if this is an accurate description of the evolution of the development processes that resulted in some of the unusual features of the capsule and seed in Glomeropitcairnia. The answers will require comparative studies of developmental morphology of ovules in Glomeropitcairnia and in other bromeliads, in particular, Tillandsioideae and Pitcairnioideae.
Even if this explanation should turn out to be accurate, it does not explain the adaptive significance, if any, or phylogeny of the unusual appendages at both ends of the seeds, nor does it explain the partially inferior ovary of Glomeropitcairnia, a character almost unique in the subfamily. A waiting further data are the answers to questions of phylogenetic reconstructions and correct placement of this unusual genus. Is it at the end of an evolutionary branch, or is it close to the ancestor of the . subfamily?
Some N otes on the Remarkable Bromeliad Genus Glomeropitcairnia by
Walter Till, Heidemarie Halbritter, and Gunter Gortan in J Brom. Soc. 47(2):65-72. 1997
Grisebach (1864) described the earliest member of this genus as Tillandsia penduliflora, Baker (1889) transferred it to Caraguata, and Wittmack (1890) followed him. Mez (1896) placed it in its own subgenus of Pitcairnia which he named Glomeropitcairnia. Not until 1905, when Mez described the second species, did he recognize Glomeropitcairnia as a distinct genus within subfamily Tillandsioideae. These few examples demonstrate how much the leading Bromeliad experts of those days had been puzzled by the half-inferior ovary, the imperfectly dehiscent capsule, and the pappiformously bi-appendaged seeds, all being features characterizing this genus.
Glomeropitcairnia penduliflora (Grisebach) Mez is distributed on the Lesser Antillean islands of Montserrat, Guadeloupe, Dominica, and Martinique, while G. erectijlora Mez is restricted to the Paria peninsula in northeastern Venezuela and adjacent Isla Margarita and Trinidad. The most recent taxonomic treatment is that of Smith and Downs ( 1977). Foster (1956) first published a photograph of the tank rosette of G. penduliflora, Hodge ( 1962) illustrated a flowering specimen of this giant species. Field impressions accompanied by color photographs as well as a brief phylogenetic discussion have been published by Gilmartin and Brown ( 1986).
Both species grow in montane forests where temperatures are quite moderate, precipitation is high and fog is frequent (Foster 1956, Hodge 1962, Gilmartin & Brown 1986, Rase 1996). In Dominica W. Till has found G. penduliflora as a frequent epiphyte on the western slopes of Morne Trois Pitons and less frequent as a terrestrial in disturbed areas. At this locality the plants were in their final stage of flowering in mid February 1996. The species was further seen quite frequently on the southeastern slopes of Morne Macaque (= Micotrin) and on a mountain ridge south of Morne Diablotins northeast of Salisbury. The mentioned sites are located between (350- ) 500-800 m s.m. At least in Dominica G. penduliflora seems not to be threatened by collecting if not by habitat destruction. F. J. Rase has found G. erectiflora on the Serbatana ridge above Carupano (figure 5) at an elevation of 1100 m s.m. in a very foggy and humid environment. G. erectiflora is obviously restricted to very local populations and seems to be threatened by forest clearing and burning.
The large branched inflorescence of G. penduliflora is inconspicuously yellowish green, the floral bracts are green, the sepals are of similar color, and the petals are pastel orange (figure 6). In contrast, the slenderly cylindric inflorescence of G. erectiflora as well as the floral bracts are bright carmine red (figure 7), the sepals are greenish orange, and the petals are orange (figure 8).
Thanks to recently collected fresh material it is possible to present morphological data of stigmas and pollen of both species (plates 1 and 2). Anthetic flowers and flower buds have been fixed in FAA (formaldehyde acetic alcohol), chemically dehydrated with 2,2 dimethoxypropane (DMP) for pollen study, or flowers have been stored in Farmers fixative for stigma study. In the latter case, for floral dissections the samples have been transferred into 70% ethanol and have been dehydrated in graded ethanol series. In both cases the samples have been critical point-dryed with CO2, have been mounted on stubs, have been sputter coated with gold, and have been investigated with a JEOL T-300 scanning microscope at 10 kV: Voucher specimens are deposited in WU (both species: W Til1 12021, F: J Base s.n.) and in B, HB, SEL (G. penduliflora, W Til1 12021, only).
The only data on pollen of Glomeropitcairnia, those of Erdtman and Praglowski ( 1974 ) but are based on acetolysed material and therefore hardly useful. According to the present study the hydrated pollen is sulcate, spheroidal to elongated and about 40 to 50 µm large (long axis; plate 2, figs. 7,8,11,12). The pollen surface exhibits a reticulate exine pattern. The muri of the tectum are of equal diameter and are connected to the foot layer with short columellae (pl. 2, fig. 8). Lumina width is greatest at the proximal pole and becomes reduced towards the pollen equator and towards the aperture (germination area) respectively. The sulcus margins are not distinct but a gradual transition to the actual sulcus area is evident. This "diffuse sulcus type" (Halbritter 1992) is quite frequent in Bromeliaceae pollen, especially in that of members of subfamily Tillandsioideae. There are no significant differences in size and morphology of the pollen of the two species. In G. penduliflora the proximal lumina width and their reduction towards the sulcus is perhaps more distinct than in G. erectiflora. The sulcus area of the former species seems to be more covered by exine elements than it is in the latter species (plate 2, figs. 9, 10) but this difference is caused merely by different states of hydration and extention respectively.
Brown and Gilmartin (1989) described the stigma of G. penduliflora as "distinctive and thus far unique". They find "the three style-lobes fully connated at their margins to form an inverted-umbrella-like membranaceous stylar blade which lacks papillae and is undulate and irregularly folded". However, they stress that this conformation may not be free from artifacts. In a subsequent paper (Gilmartin et al. 1989) this description is repeated nearly identically and is accompanied by a SEM picture.
The present results demonstrate that the stigma morphology of both Glomeropitcairnia species is very similar to each other. The three stigma lobes are not twisted but are cup-shaped, are in horizontal position and their undulated margins are intensively folded. Stigmas of Glomeropitcairnia belong to the "convolute-blade type" (Brown and Gilmartin 1989) which obviously is autapomorphic for (= restricted to) subfamily Tillandsioideae and is met there also in Guzmania and Vriesea. In both species of Glomeropitcairnia stigma papillae are well developed. In comparison, the papillae of G. penduliflora (plate I, figs. 1-3) are shorter than in G. erectiflora (pl. I, figs. 4-6). However, the illustrated stigma of G. penduliflora is from a young bud while that of G. erectiflora is from a flower just before anthesis which might explain the differences. All papillae exhibit an intensively folded cuticle. Similarly folded cuticles of stigma papillae occur among certain Vriesea spp. and Tillandsia spp. Stigmas of anthetic flowers are completely covered with germinated pollen grains.
The data of pollen and stigma morphology support the retention of Glomeropitcairnia within subfamily Tillandsioideae as proposed by Gilmartin et. al. (1989). The diffuse aperture type of the pollen and the convolute-blade type of the stigma are very similar to that found in Guzmania patula Mez & Werckle or in Vriesea oligantha (Baker Mez ( Vriesea sect. Xiphion ), indicating a closer relationship of these genera or their segregates respectively. The taxonomic value of the folded cuticles of the stigma papillae remains unclear. However, folded cuticles are rare within Bromeliaceae. Gilmartin et al. ( 1989) resolved a cladogram with a Glomeropitcairnia/Mezobromelia/Guzmania lineage which is partly corroborated by pollen and stigma featyres. However, Terry and Brown (1996), which have resolved Glomeropitcairnia, Vriesea sect. Xiphion, and Mezobromelia as to be basal to the rest of the Tillandsioideae ( Catopsis excluded), have proposed Guzmania as to be terminal branches of their cladograms in a position far remote from Glomeropitcairnia. These latter findings are not supported by the present morphological data.
The seeds with a pappose appendage on both ends and the imperfectly dehiscent capsules remain as unique features and exhibit Glomeropitcairnia as a genus which obviously has branched off early during the evolution of subfamily Tillandsioideae. This view is supported both by Gilmartin et al. (1989) and by Terry and Brown (1996). Consequently, Vriesea sect. Xiphion must be considered as similarly basal to the Tillandsioideae clade. The position of Guzmania should be re-considered.