Transfering Canistrum flavipetalum Wand. to Wittrockia.
Article Type: Report
Subject: Bromeliaceae (Identification and classification)
Bromeliaceae (Physiological aspects)
Phenetics (Research)
Biology (Identification and classification)
Biology (Research)
Botany (Morphology)
Botany (Research)
Authors: Leme, Elton M.C.
Luther, Harry E.
Pub Date: 05/01/2010
Publication: Name: Journal of the Bromeliad Society Publisher: Bromeliad Society International Audience: Academic Format: Magazine/Journal Subject: Biological sciences Copyright: COPYRIGHT 2010 Bromeliad Society International ISSN: 0090-8738
Issue: Date: May-June, 2010 Source Volume: 60 Source Issue: 3
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: Singapore; Brazil Geographic Code: 9SING Singapore; 3BRAZ Brazil
Accession Number: 248734310
Full Text: [FIGURE 1 OMITTED]

Taxonomical history

The genus Wittrockia was established by Lindman (1891) on the basis of a single specimen, W. superba Lindm. The concept of the genus has undergone a considerable change to arrive at the parameters proposed by Leme (1997; 1998). Lindman (1891) originally associated it to the genus Nidularium Lem., s. lato, and with the genus Bromelia L., using the petals appendages as a distinguishing character. The author also considered Wittrockia as intermediate to Nidularium and Neoregelia L. B. Sm. [then Regelia (Lem) Lindm.], because it didn't have the branched inflorescence with the showy primary bracts of the former nor the simple inflorescence of the latter.

Mez (1894) reduced Wittrockia to the rank of subgenus in Canistrum E. Morren (= subgenus Nidulariopsis Mez). As a subgenus, the circumscription of Wittrockia was expanded to include Canistrum ama%onicum (Baker) Mez [= Nidularium ama%onicum (Baker) Linden & E. Morren ex Lindm.]. According to Mez (1896; 1934), the subgenus Wittrockia was then distinguished from the subgenus Canistrum (= subgenus Eucanistrum Mez) by its high-connate petals versus the free petals in the latter.

Smith (1945) restored the generic status of Wittrockia, arguing that the petals partially connate at their base were enough to distinguish it from Canistrum and Aechmea subgen. Ortgiesia (Regel) Mez. The petal appendages were used in his key to separate Wittrockia from Nidularium and Neoregelia. The concept of the genus was once more expanded by Smith (1952; 1955; 1969) by means of the inclusion of some species currently included in Nidularium and Neoregelia [e. g., Nidularium minutum Mez, N. azureum (L. B. Sm.) Leme, Neoregelia bragarum (E. Pereira & L. B. Sm) Leme]. In their monograph, Smith & Downs (1979) did not modify the conceptual limits of the genus.

The concept of Wittrockia was expanded further by Leme (1989; 1991a; 1991b; 1992; 1995) by the inclusion of species nowadays positioned in Aechmea, Neoregelia and Nidularium. Based on such disparate taxonomic elements, its circumscription became very confusing, based as it was solely on the presence of petal appendages and basal petal concrescence. Such a chaotic conglomeration of elements from Aechmea, Neoregelia and Nidularium could only be compared to Baker's (1889) concept of Karatas Plumier.

When revising the genera of the so called "nidularioid complex", composed by Canistrum, Canistropsis (Mez) Leme, Edmundoa Leme, Neoregelia, Nidularium and Wittrockia), Leme (1997) introduced arguments in favor of the correlation of multiple traits, based on morphological, anatomical, palynological, biogeographic and ecological data, to retrieve the concept of Wittrockia in order to decrease its unduly artificial nature. The isolated individual use of some diagnostic elements (e.g., nidular inflorescence, flower pedicels, petal concrescence, petal appendages) was assigned a relative value, both in weight and importance (Leme, 1998; 2000). This allowed the recognition of the W. superba Lindm., W. cyathiformis (Vell.) Leme and W. gigantea (Baker) Leme as the "core" taxa of the genus (the last two historically assigned to Canistrum), as well as two other more peripheral taxa, W. spiralipetala and W. tenuisepala (Leme) Leme. Species of Wittrockia clearly related to Aechmea, Neoregelia and Nidularium were excluded from the genus and recombined. Later on, W. paulistana Leme was added to the "core" taxa of the genus (Leme, 2000).

Wittrockia, in a strict sense, as conceived by Leme (1997) became a genus comprising medium- to large-sized species, being anatomically and morphologically close related to Edmundoa, Canistropsis and Neoregelia. It is interesting to mention that in the phylogeny of Bromelioideae inferred from DNA data, the "nidularioid clade", based on Edmundoa lindenii (Regel) Leme, Neoregelia spp., Nidulariumprocerum Lindm., and W superba, is statisticaly highly supported (Schulte & Zizka, 2008; Schulte et al., 2009). A similar result for the nidularioid lineage (i.e. Edmundoa, Neoregelia, Nidularium, and Wittrockia) was obtained by Sass & Specht (2010) on the basis of chloroplast, nuclear ribosomal and low copy nuclear DNA. Additionally, the results obtained by these authors segregated Canistrum aurantiacum E. Morren and C. fosterianum L. B. Sm. from the nidularioid lineage and associated them to Aechmea s.l.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

In relation to leaf blade anatomy, Wittrockia superba, W. cyathiformis and W. gigantea are closely related to Edmundoa lindenii and E.perplexa (L. B. Sm.) Leme, Neoregelia brownii Leme and N. rubrovittata Leme (i.e., not related to Aechmea) because of the presence of sheath extensions formed by thick-walled cells, associated with vascular bundles. In the studied species of Canistrum (C. aurantiacum, C. camacaense Martinelli & Leme, C. montanum Leme and C. triangulare L. B. Sm. & Reitz), which are anatomically related to Aechmea, the leaf blades have a predominantly spongy mesophyll, and fibers are not associated with the vascular bundle (Sajo et al. 1998).

According to Halbritter & Till (1998), the pollen of the "core" Wittrockia is diporate with small porous and reticulate exines, are remarkable in that the lumina are filled in part with a very persistent kind of pollen kit in W. cyathiformis and W. gigantea. On the other hand, these authors highlighted the heterogenous pollen in Canistrum, mainly in the type species C. aurantiacum, which is polyporate, and in C. auratum Leme (subgenus Cucullatanthus Leme) having large pores. The palynological features of the first species are in fact more closely allied to the "Gravisia complex" of Aechmea (unpl. data), while the latter species belongs to the complex headed by Aechmeafosteriana, the "Cucullatanthus complex", which species are nowadays spreaded in three genera (i.e., Aechmea, Canistrum and Quesnelia; unpl. data) and were listed by Leme & Siqueira-Filho (2006). Recently, preliminary results based on DNA analysis gave a high support to the Cucullatanthus clade, taking into consideration C. fosterianum and Quesnelia edmundoi L. B. Sm. (Schulte & Zizka, 2008; Schulte et al., 2009) and C. fosterianum and Aechmea correia-araujoi E. Pereira & Moutinho (Sass & Specht, 2010).

Besides leaf blades anatomy and pollen differences, Leme (1997; 1998) distinguished Wittrockia s. str. from Canistrum by its comparatively larger flowers (45-84 mm vs. 25-50 mm long), the longer sepals [(12-) 25-45 mm vs. 10-20 (-24) mm long] which are subsymmetrical, lacking a pronounced lateral wing, with apex often attenuate-caudate and not pungent (vs. strongly asymmetrical with a well developed lateral wing equaling to exceeding the distal portion of the midnerve, with mucronulate to spinescent apex) and usually ecarinate (vs. the posterior ones usually alate-carinate with keels decurrent on the ovary). According to the inferred evolution of sepals symmetry based on DNA analysis, symmetric sepals are ancestral within subfamily, and in the case of the Edmundoa, Nidularium and Wittrockia of the Nidularioid clade, its symmetrical sepals are reversals stemed from common ancestry (Schulte et al., 2009).

In addition to sepal characters, Wittrockia s. str. also differs from Canistrum by the narrowly ovate to obovate or lanceolate petals (vs. sublinear), which are comparatively longer and broader [25-46 x (6-) 8-15 mm vs. 15-30 x 3-6 mm], anthers usually dorsifixed near the base (vs. dorsifixed near the middle), epigynous tube often lacking (vs. distinct, 1-4 mm long), and by the white, orange-yellow or red fruits (vs. with shades of purple, which is common in Aechmea-related taxa).

The geographical distribution of Wittrokia s. str. is concentrated in the southernsoutheastern Brazilian states of Santa Catarina to Rio de Janeiro, and further inland it occurs in the state of Minas Gerais, and a single species in the northeastern state of Bahia. In contrast, Canistrum predominates in the northeastern states, from Bahia to Pernambuco, with the exception of the state of Espirito Santo, in southeastern Brazil, where a single species can be found. That distinct pattern of geographical distribution is important because, according to Sass & Specht (2010), the apparently homoplastic morphological characters used to assign species to genera or subgenera may be useful taxonomically when geography is also taken into account.

Finally, the segregation of Wittrockia from Canistrum in the circumscription adopted by Leme (1997;1998) was well supported by cladistic analysis based mostly on morphological selected characters (Brown & Leme, 2000).

Recently, the Bromeliaceae of the state of Sao Paulo was revised in volume 5 of Flora Fanerogamica do Estado de Sao Paulo (Wanderley & Martins, 2007) adopting broad circumscription concepts for the genera of the Nidularioid complex in contrary to the mainstream views of the Bromeliaceae research community (Brown et al., 2008). However, this monograph did not presented any tangible new data and explicit scientific data based justification to explain the maintenance of older arguably artificial generic concepts (e.g., Smith & Downs, 1979) over more recent, data supported monographic revisions. One of the examples is the rejection of Edmundoa and Wittrockia as distinct genera from Canistrum on the basis of the vague justification that variability of diagnostic characteristics of the involved taxa is continuous (Wanderley et al., 2007), which is not rooted in any newly produced scientific data nor inspired in any more reasonable phylogenetic hypothesis, and just contributed to nomenclatural instability by means of proposed superfluous new binomials.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

A more recent consequence of this ideological-based position observed in part in the "Flora de Sao Paulo" treatment was the publication of a new species, Canistrum flavipetalum Wand., which is clearly a typical member of the genus Wittrockia according to Leme (1997; 1998), being closely morphologically related to W. cyathiformis, as correctly indicated in the protologue. It is interesting to mention that in the protologue, Wanderley (2008) stressed that the adopted concept of Canistrum, including Edmundoa and Wittrockia as synonyms, was "detailed discussed" in Wanderley et al. (2007). However, consulting the reference cited, the only justification found is the vague statement that the the variability of diagnostic characteristics is continuous.

Based on all arguments presented above, this newly described taxon does not have any relationship with Canistrum, neither with the typical subgenus, nor with the subgenus Cucullatanthus. So, a new combination is necessary and implemented here.

Taxonomic Treatment

Wittrockia flavipetala (Wand.) Leme & H. Luther, comb. nov.

Basionym: Canistrum flavipetalum Wand., Hoehnea 35 (4): 537, fig. 1, 2. 2008. Type: Brazil, Bahia, Abaira, Tijuquinho, 13[degrees]16'S 41[degrees]54'W, 1,700-1,800 m elev., 26 Febr. 1992, P. T. Sano & L. Laessoe H-52351 (Holotype, SPF, n.v.; Isotype SP, n.v.).

Plant ephiphyte or terrestrial, propagating by short basal shoots. Leaves ca. 20 in number, subcoriacous, suberect-arcuate, forming a broad funnilform rosette; sheaths broadly elliptic, 13-14 x 10-11 cm, greenish toward the apex, subdensely brown lepidote on both sides, strongly nerved with conspicuously castaneous nerves, coarsely spinose at extreme apex; blades sublinear-lanceolate, narrowed toward the base, 25-60 x 5-6 cm, sparsely and inconspicuously white-lepidote to glabrescent, green to reddish at apex and along the apical margins, with sparse and irregular darker green spots, apex acute and distinctly apiculate, apiculus 3-4 mm long, rigid, margins densely to subdensely spinose, spines subspreading, triangular, castaneous toward the apex, the basal ones ca. 3 mm long, the apical ones 1.5-2 mm long. Scape 23-30 cm long, ca. 0.8 cm in diameter, green, glabrescent, rugulose, mostly naked; scape bracts one near the base and few involucrate ones at apex, suboblong, acuminate, suberect to spreading toward the apex, green or sometimes slightly reddish at extreme apex and along the apical margins, inconspicuously and sparsely white-lepidote toward the apex, the upper ones exceeding the petals, ca. 50-80 x 30 mm, densely and coarsely spinose, spines pale castaneous, 1-2 mm long. Inflorescence compound, subumbellate, with a narrow obconic base and a stellate apex of spreading involucral and primary bracts, subtripinnate except for the simple central portion, distinctly elevated above the rosette, ca. 5 cm long, 8-10 cm in diameter at apex (including the primary bracts); primary bracts resembling the involucral bracts, but smaller, the outer ones oblong, apex broadly acute and distinctly apiculate, erect at base to spreading-recurved toward the apex, ca. 40 x 10 mm (the inner ones much reduced), about equalling the sepals length, inconspicuously and sparsely white-lanate at apex to glabrous, green, densely spinose, spines irregularly curved, ca. 1 mm long; fascicles ca. 5 in number, ca. 40 x 17 mm (excluding the petals), subflabellate, subcomplanate, ca. 3-flowered, inconspicuously stipitate, with few sterile floral bracts; floral bracts narrowly sublinear-lanceolate, apex acuminate, 27-32 x 3-5 mm, irregularly spinulose to subentire, obtusely carinate, finely nerved, greenish toward the apex, inconspicuously and sparsely white-lanate at apex to glabrous, about equalling 1/2 to 2/3 of the sepals length. Flowers ca. 45 mm long, subsessile, odorless; sepals oblong-lanceolate, subsymmetrical, apex acuminate to apiculate-caudate, not rigid, 12-15 x 5 mm, free, inconspicuously and sparsely white-lanate to glabrous, green, obtusely if at all carinate toward the base; petals narrowly elliptic-subobovate, ca. 25 x 8 mm long, free, membranous, apex emmarginate, the apical 1/3 yellow, the basal 2/3 whitish, erect except for the spreading-recurved apex at anthesis, bearing at base 2 appendages ca. 3.5 x 1.5 mm, suboblong-lanceolate, subacute, entire or nearly so, without callosities; filaments ca. 18 mm long, free, subterete; anthers ca. 5 mm long, base sagittate, apex obtuse, fixed near the base; stigma conduplicate-spiral, globose, ca. 2.5 in diameter, lobes entire or nearly so; ovary subclavate, trigonous, 17-18 mm long, 6-7 mm in diameter, greenish-white, glabrous; epigynous tube ca. 1 mm long; placentation from median to apical; ovules many, subcylindrical, obtuse. Fruits enlarged from the ovary, orange-yellow.

Material examined: Bahia: Chapada Diamantina, Palmeiras, Caete-Acu, riparian forest, 14 Mar. 1997, J. C. Garcia s. n., cult. E. Leme 3824 (HB); fl. cult. Sept. 1999, E. Leme 3824 (HB).

Additional remarks

Wittrockia flavipetala was first reported by Leme (2000) as an unpublished species in the supplement study of Wittrockia. It was not formally described because R. C. Forzza (pers. comm.) stated it was under study by her and M. G. L. Wanderley. So, only a full-page color photo of it and commentaries on its distribution and morphological affinities were then presented, as well as a expanded identification key for the genus.

According to the identification key presented by Leme (2000), Wittrockia flavipetala is closely related to W. cyathiformis, being distinguished by the shorter scape (23-30 cm vs. 30-55 cm long), involucral bracts and primary bracts green or sometimes reddish at their distal portion (vs. reddish-pink to purplish throughout), shorter primary bracts (ca. 4 cm vs. 7-12 cm long), shorter flowers (ca. 45 mm vs. 55-60 mm long), distinctly shorter sepals (12-15 mm vs. ca. 30 mm long) and by the shorter petals with entire appendages or nearly so (ca. 25 mm vs. ca. 43 mm long; petal appendages denticulate).

photos by E.M.C. Leme

References

Baker. J. G. 1889. Handbook of the Bromeliaceae. G. Bell & Sons, London. Pp.-243.

Brown, G. K. & E. M. C. Leme. 2000. Cladistic Analysis in the Nidularioid Complex. Pp. 240247, in E. M. C. Leme, Nidularium--Bromelias da Mata Atlantica, Sextante, Rio de Janeiro.

Brown, G. K., W Till, E. M. C. Leme & J. R. Grant. 2008. A review: Bromeliaceae--Flora Fanerogamica do Estado de Sao Paulo. J. BromeliadSoc. 58 (4): 159-162.

Halbritter, H. & W Till. 1998. Pollen Morphology of Nidularium and Related Genera. Pp. 114-121, in E. M. C. Leme, Canistropsis--Bromeliads of the Atlantic Forest, Salamandra, Rio de Janeiro.

Leme, E. M. C. 1989. Novas bromeliaceas nativas do Brasil--VI. Bradea 5 (19): 166-186.

--. 1991a. An addition to the genus Wittrockia. J. Bromeliad Soc. 41 (4): 147-154.

--. 1991b. Novas bromeliaceas nativas do Brasil--VIII. Bradea 5 (42): 391-396.

--. 1992. A new Brazilian Wittrockia species. J. Bromeliad Soc. 42 (2): 51-53.

--. 1995. Miscellaneous new species of Brazilian Bromeliaceae. Selbyana 16 (1): 110-122.

--. 1997. Canistrum--Bromeliads of the Atlantic Forest, Salamandra, Rio de Janeiro. Pp.-107.

--. 1998. Canistropsis--Bromeliads of the Atlantic Forest, Salamandra, Rio de Janeiro. Pp.-143.

--. 2000. Nidularium--Bromeliads of the Atlantic Forest, Sextante, Rio de Janeiro. Pp.-263.

Leme, E. M. C. & J. A. Siqueira-Filho. 2006. Consideracoes taxonomicas correlacionadas as bromelias de Pernambuco e Alagora. Pp. 383-407, in J. A. Siqueira-Filho & E. M. C. Leme (eds.), Fragmentos de Mata Atlantica do Nordeste, Biodiversidade, Conserva^ao e suas Bromelias. Andrea Jakobsson Estudio, Rio de Janeiro.

Lindman, C. A. M. 1981. Bromeliaceae Herbarii Regnelliani, I. Bromeliae. Ofvers. Forh. Kongl. Svenska Vetensk.-Akad. III 24 (8). Stockholm.

Mez, C. 1894. Bromeliaceae, in C. F. P. Martius Fl. bras. 3 (3). F. Fleischer, Lipsiae. Pp. 425-634.

--. 1896. Bromeliaceae, in C. DC., Monogr.phan. 9. Masson & Cie. Paris. Pp.102-105.

--. 1934. Bromeliaceae, in Engler Pflanzenr. IV. 32. W Engelmann, Leipzig. Pp.-667.

Sajo, M. G., S. R. Machado & S. M. Carmello-Guerreiro. 1998. Bromeliad Leaf Structure and the Implication for Species Groups. Pp. 102-111, in E. M. C. Leme, Canistropsis--Bromeliads of the Atlantic Forest, Salamandra, Rio de Janeiro.

Sass. C. & C. D. Specht. 2010. Phylogenetic estimation of the core Bromelioids with an emphasis on the genus Aechmea (Bromeliaceae). Molecular Phylogenetics and Evolution 55: 559-571.

Schulte, K., & G. Zizka. 2008. Multi locus plastid phylogeny of Bromelioideae (Bromeliaceae) and the taxonomic utility of petal appendages and pollen characters. Candollea 63: 209-255.

Schulte, K., M. H. J. Barfuss & G. Zizka. 2009. Phylogeny of Bromelioideae (Bromeliaceae) infered from nuclear and plastid DNA loci reveals the evolution of the tank habit within the subfamily. Molecular Phylogenetics and Evolution 51: 327-339.

Smith, L. B. 1945. The subfamilies and genera of the Bromeliaceae. Plant Life 1 (2-3): 40-44.

--. 1952. Bromeliaceas novas ou interessantes do Brasil--V Arq. Bot. Estado Sao Paulo II. 2: 195-198.

--. 1955. The Bromeliaceae of Brazil. Smithsonian Misc. Collect. 126 (1):-290.

--. 1969. Notes on Bromeliaceae, XXIX. Phytologia 18 (3): 137-142.

Smith, L. B. & R. J. Downs. 1979. Bromelioideae (Bromeliaceae). Fl. Neotrop. Monogr. 14. The New York Botanical Garden, New York. Pp. 1.493-2.142.

Wanderley, M. G. L. & S. E. Martins. 2007. Bromeliaceae. Pp. 39-43, in T. S. Melhem, M. G. L. Wanderley, S. E. Martins, S. L. Jung-Mendacolli, G. J. Shepherd & M. Kirizawa (eds). Flora Fanerogamica do Estado de Sao Paulo. Bromeliaceae 5. Fundacao de Amparo a Pesquisa do Estado de Sao Paulo, Sao Paulo.

Wanderley, M. G. L., S. E. Martins, S. L. Proenca & B. A. Moreira. 2007. Canistrum E. Morren. Pp. 73-79, in T. S. Melhem, M. G. L. Wanderley, S. E. Martins, S. L. Jung-Mendacolli, G. J. Shepherd & M. Kirizawa (eds). Flora Fanerogamica do Estado de Sao Paulo. Bromeliaceae 5. Fundacao de Amparo a Pesquisa do Estado de Sao Paulo, Sao Paulo.

Elton M. C. Leme (1) & Harry E. Luther (2).

(1) Herbarium Bradeanum, Rio de Janeiro, Brazil; leme@tjrj.jus.br.

(2) Gardens by the Bay, National Parks Board, Singapure; harry_luther@nparks.gov sg.
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.