DISCUSSION
Within the genus Canistropsis, C. correia-araujoi is unquestionably the most disparate taxon, in terms of morphology, when compared to the other species. The absence of additional collections of this bromeliad (see next topic) makes our analysis all the more difficult, besides the fact that several morphological traits shown by this species do not follow the normal pattern of this genus, nor is there close affinity to Nidularium or Neoregelia. Though compatible with the basic conceptual lines of Canistropsis, important dissonant elements are as follows: a) leaves stiffer than usual, but not coriaceous; b) central rosette leaves becoming slightly reddish purple at anthesis in contrast to the outer leaves; c) petals connate for over half their length.
Both the stoloniferous habit and the inflorescence structure conform to that of Canistropsis, but leaf texture and color are more closely associated with Neoregelia. Conversely, the corolla with its well-developed basal tube is typical of Nidularium. It is therefore possible that this taxon is a product of natural bigeneric hybridization between sympatric species of the nidularioid complex - and this is not an isolated phenomenon (Leme, unpubl. data). At least three species that might have played a role in this hybridization process are found in the area where C. correia-araujoi was collected: C. microps (see description below), Nidularium procerum (nonstoloniferous, leaves subrigid and blue corolla, distinctly gamopetalous) and Neoregelia indecora (stoloniferous, leaves subrigid, central leaves turning red at anthesis, and blue petals).
Stigma morphology studies have also raised the question as to the hybrid origin of this taxon as is demonstrated by the disruption of stigma blade development (Chapter 7). However, before deciding whether the hypothesis is correct or not, and which species would have contributed to hybrid formation, I think it wise to do more laboratory and field studies aimed at resolving this very issue, a course of action which was not possible during the course of this revision.
This species is known only from the type collection from remnant forests in the Jacarepagua area, Rio de Janeiro. There have been no new collections and therefore very little is known about the geographic distribution of the taxon. Canistropsis correia-araujoi flowers in cultivation during December and January in Rio de Janeiro; it is represented in Brazilian horticulture and overseas by the clone that produced the holotype. In theory, it may occur in the Tijuca National Park, the Grajau Forest Reserve or the Pedra Branca State Park, public conservation units near the site where the taxon was found for the first and only time. —SeeLeme 1998
