Canistrum pickelii (A.Lima & L.B.Sm.) Leme & J.A.Siqueira

Literature references:

FCBS (Web) J. Bromeliad Soc.: 52:115cdh Phytologia: 20:183 Siqueira & Leme 2006: 266,409h Smith & Downs 1979: 2041* U.S. Nat. Herb.

*move your mouse pointer over the page numbers to see comment

Comments:

• Discussion
  In historical terms, Canistrum pickelii was initially identified in October 1928 as Quesnelia aff. arvensis by Harms. Later, Andrade Lima identified the species as a member of Wittrockia, as mentioned before. Finally in 1970, Andrade Lima and Lyman B. Smith published the taxon as Portea pickelii. However, the original inclusion of this species in the genus Portea was probably because of the obscurely pedicellate flowers. Flowers with pedicels were excessively valued at the time, as having absolute diagnostic importance, which distanced the taxon from the other phylogenetically related taxa, especially the syntopic Canistrum aurantiacum.
  
  In the context of an interpretational analysis of the whole morphological traits of this species, facing a multiple correlation of its characteristics, as proposed by Leme (1997) to group the species of the nidularioid complex in a more natural manner, the taxon in question fits the genus Canistrum. This is because it possesses corymbose, involucral and multi-utriculate inflorescences, with large brightly colored rosulate and imbricate primary bracts, which impound some amounts of water for several days. In fact, the reasonable ability of the inflorescence of this species to store water is greater than in all the other species of Canistrum. Furthermore, its floral fascicles are densely subflabellate, complanate to subpulvinate and shortly pedunculate, bearing carinate floral bracts. In Portea, the inflorescence does not form a cup even when it is dense (e. g., P. kermesina) and it does not possess any ability to store rain water or even moisture.
  
  The sepals of Portea are ecarinate and usually connate at base for about half of their length, while in Canistrum the sepals are shorter connate and the posterior ones are almost always carinate, with the keels decurrent on the ovary, which is therefore trigonous. In contrast, the ovary in Portea is terete. Also, the sepals of C. pickelii, despite being distinctly asymmetrical, have a lateral wing shorter than their apex ( defined by the mucronulate termination of the midrib). This characteristic draws C. pickelii to C. aurantiacum, and distances it from Portea, whose sepals, in the typical species, present a lateral wing that will be at least equal in length to the apex and usually much longer than it.
  
  Although C. pickelii has pedicellate flowers, these pedicels are inconspicuous and not long, slender and conspicuous like those of Portea. The ovules, although caudate, do not contrast significantly with the ovules reported for Canistrum that can be obtuse to distinctly apiculate, as in C. aurantiacum.
  
  Preliminary results of recent cladistic studies of the nidularioid complex, based on Brown & Leme (2000) and using revised or new morphological figures, have positioned C. pickelii and the species described below in the Canistrum clade. These two species presented an intermediate position on the cladogram between the basal group formed by C. aurantiacum and C. camacaense (typical subgenus) and the group with more terminal topology, composed by species of the subgenus Cucullatanthus (Brown & Leme, in prep.). Despite its intermediate positioning, the general traits of C. pickelii permit its inclusion in the typical subgenus, which is: funnelform leaf rosette (not ellipsoid or tubular), petals suberect-recurved to slightly reflexed near the apex at anthesis (not erect), and caudate ovules (not obtuse). On the other hand, a certain proximity of C. pickelii to the Cucullatanthus subgenus derives from its bipinnate inflorescence (not tripinnate) and sublinear-lanceolate petals, with a narrowly obtuse apex (neither lanceolate, acuminate or acute as in subgenus Canistrum, nor sublinear or spatulate and obtuse-cucullate as in subgenus Cucullatanthus ).
  Canistrum piclelii occurs in areas of mountainous Atlantic Forest in the south of Pernambuco and north of Alagoas States. There is also reference that indicates its presence further into the interior of Pernambuco, at the Serra do Oruruba, in Pesqueira, in the so called "Brejos de Altitude" - highland wetlands (sensu Vasconcelos Sobrinho, 1970). It can also be found in rocky areas bordered by forests, like at the Pedra do Cruzeiro, in Jaqueira, where the plants are exposed to the full sun and are generally more compact
  In forests, the species occurs preferably as an epiphyte in the upper parts of trees such as Eriotheca crenulaticalyx (Bombacaceae ), Sloanea obtusifolia (Elaeocarpaceae ), Pterocarpus violaceus (Leguminosae) and Ficus cf. maxima (Moraceae ). It can occur together with C. aurantiacum, that is pollinated by hummingbirds. In contrast, C. pickelii depends on another group of pollinators, since it flowers at night, probably sphingids. Nocturnal anthesis is in fact one of the most remarkable characteristics of C. pickelii, being an uncommon phenomenon in Bromelioideae, only reported for Aechmea kleinii (Reitz 1983 ), which curiously presents a bright yellow, tubular corolla, like some of the other species of the same subgenus Ortgiesia.
  
  
  Conclusion
  
  The rediscovery of C. pickelii in nature after 25 years and the discovery of C. alagoanum reveals four basic aspects of the situation in which the Atlantic Forest of the Northeast, North of the Sao Francisco River, finds itself, and of the level of taxonomical knowledge of Bromeliaceae:
  1) It has been verified that an accelerated process of destruction of the few fragments of the Atlantic Forest biome that have survived in the Northeastern sector, North of the Sao Francisco River, is still underway despite the fact that only 2% of the original forest still exists (Teixeira, 1986).
  2) It can be verified that very little is yet known about the floristic composition and biology of the species of these fragments. In fact, these fragments of Atlantic Forest, that form one of the five most threatened hotspots of the planet (Mittermeier et al. , 2000) and now give us a vague picture of the original vegetation of yesteryear, still hold a great variety of organisms that are little known or completely unknown to science. A good example is the discovery of five new species (Aechmea frassyi Leme & J. A. Siqueira, A. gustavoi J. A. Siqueira & Leme, A. marginalis Leme & J. A. Siqueira, Cryptanthus alagoanus Leme & J. A. Siqueira and Neoregelia pernambucana Leme & J. A. Siqueira) during the taxonomical research already reported (Siqueira & Leme, 2000; Leme & Siqueira, 2001 ).
  3) It has also been verified that the perfection of taxonomical knowledge and the reconstruction of the phylogeny of Bromeliaceae are closely related to the increase of information provided especially in the field or with the maintenance in cultivation of live specimens. Research based only on herbaria samples is insufficient to fulfill the purpose of taxonomical and phylogenetic improvement of the Bromeliaceae, as exemplified by the recognition of C. alagoanum.
  4) Finally, it can be concluded that an intensified and immediate effort to research and preserve the fragments of Atlantic Forest of Northeastern Brazil should be carried out, emphasizing groups of organisms such as Bromeliaceae that are known to be bio-indicators of biodiversity.
  
  General notes
  Canistrum is a genus of the subfamily Bromelioideae, typical of the Brazilian Atlantic Forest. It is distinguished by flashy and ornamental capitate and involucral inflorescences, with upper scape bracts and primary bracts disposed in a cup-shaped format. Its flowers are bird-pollinated and it has distinctly asymmetrical mucronate to spinescent sepals, (Leme, 1997; Siqueira, 1998). There are currently ten species grouped in two subgenera, Canistrum and Cucullatanthus (Leme, 2000; Luther, 2001). The overwhelming majority of species occur in the northeastern sector of the country, throughout the States of Pernambuco, Alagoas and Bahia, with just one species found in Espirito Santo in the Southeast (Leme, 1997; 2000).
  
  In 1996, during the beginnings of the survey of Bromeliaceae in Pernambuco conducted by the first author in Brazilian herbaria (Siqueira, 2001 ), a specimen from the Empresa Pernambucana de Pesquisa Agropecuaria herbarium (IPA) soon attracted attention. It was the holotype of Portea pickelii, presenting a compact inflorescence and an unusual bract disposition that distinguished it considerably from all the species of Portea known at the time. Through comparison, it was possible to perceive the structural resemblance of its inflorescence to Canistrum aurantiacum, a typical species of Pernambuco.
  
  Portea pickelii was described in 1970 by the late botanists Dardano de Andrade Lima and Lyman B. Smith, on the basis on a specimen collected in 1963 by the former, in the Mata do Camocim, in Pernambuco (Smith, 1970). In the protologue, the authors characterized the species as a plant that lived in the shaded areas of low open woods, differing from P. kermesina - that was according to them, the species with most morphological similarities by smaller leaf spines and short connate sepals with short mucronate apex.
  
  After the conclusion of the revision of Canistrum (Leme, 1997), it became even more evident that P. pickelii was uncomfortably positioned in Portea. Curiously, the label of the holotype and also of the isotype (unclear on the latter), contains an identification by Andrade Lima himself that indicated the taxon as a new species of the genus Wittrockia. However, this initial conception by Andrade Lima, certainly intuitive and more natural from a phylogenetic point of view, was unrelated to the concepts of this genus that prevailed at the time.
  
  In order to respond to the numerous questions regarding the taxonomical positioning of P. pickelii, field investigations aimed at finding its remaining populations were intensified. The search was initiated at the type locality, known as Mata do Camocim, belonging to the Ecological Station of Tapacurci, in the municipality of Sao Lourenco da Mata, that shelters stretches of seasonal semideciduous forest (sensu Vasconcelos Sobrinho, 1970). However, after several expeditions, no populations of P. pickelii were found. These probably became locally extinct due to the flooding caused by the Tapacurci dam. Finally, in January 1997, dense rupicolous and epiphytic populations of P. pickelii were found in Pernambuco in the forests of the Frei Caneca sugar-mill, in the municipality of Jaqueira, whose morphological traits were revealed to be perfectly adequate to the protologue of the species.
  
  The study of these new populations has allowed deeper understanding of the taxonomical traits of the species, specially the floral morphology, but also of the phenological and floral biology. Also, the resumption of its studies and an expansion in field investigation, covering the neighboring state of Alagoas, gave rise to the recognition of another taxon, very similar to P. pickelii, but still unknown to science. So, taking the morphological, palynological, phenological and ecological information into account, it became clear that both should be placed within the genus Canistrum.
  
  Materials and Methods
  Besides several field expeditions between 1996 and 2001, the herbaria of the Empresa Pernambucana de Pesquisa Agropecuaria, Professor Dardano de Andrade Lima (IPA), Universidade Federal Rural de Pernambuco, Professor Vasconcelos Sobrinho (PEUFR) and Professor Sergio Tavares (HST), Universidade Federal de Pernambuco, Professor Geraldo Mariz (UFP), all in Pernambuco, Universidade Federal da Paraiba, Areia (EAN) in Paralba, Instituto do Meio-ambiente de Alagoas (MAC) and Universidade Federal de Alagoas (MUFAL) in Alagoas, Herbarium Alexandre Leal Costa (EAL) in Bahia, and the National Herbarium of the Botanical Department of the Smithsonian Institution in Washington, D.C. (US), were visited. (Three of the herbaria mentioned above (HST, MUFAL, and EAL) are not presently listed in the Index Herbariorum or on the electronic version available at http://www.nybg.org/bsci/ih/.)
  
  For scanning electron microscope studies, plant material was dried using the critical point method. The material was coated with gold and observed in a SEM.
  
  Herbarium material of the studied species collected by the authors were deposited in Herbarium Professor Geraldo Mariz (UFP) and Herbarium Bradeanum (HB), and living specimens were maintained in the author's collection in Recife and Rio de Janeiro. —See J. Bromeliad Soc.