DISCUSSION
The overly artificial concept of the nidularioid-complex genera caused Canistropsis exigua to be placed originally in the genus Canistrum, based solely on the fact that this species has free petals with well-developed appendages. Today, however, it has been verified that the morphological characteristics of this taxon are totally incompatible with those of Canistrum which are, in fact, closely related to those of Aechmea s.lato (Leme, 1997).
Habit, texture of the strongly canaliculate leaves and leaf anatomy (Chapter 5), structure of the bracts and the inflorescence itself, subsymmetrical, ecarinate sepals, plus other traits, warrant the inclusion of this taxon in Canistropsis. The presence of an absolute character like the free, appendiculate petals is no longer sufficient to separate this species from the group where it was once so aptly located, based on an isolated assessment.
Canistropsis exigua reminds one of C. microps both in habit and in the delicate leaf structure, but leaf anatomy studies revealed a certain similarity to C. selloana (Chapter 5). This species resembles C. biIlbergioides in scape development and inflorescence position. It is easily distinguished from all other species in this genus by the traits mentioned above, plus bract and sepal shape and unique petal characteristics
Finally, it is interesting to note that inflorescence development in this species passes through a series of unusual stages. Contrary to development observed in the other species, where scape growth is concurrently proportional to that of the inflorescence (e. g. C. biIlbergioides, C. seIloana), the inflorescence of C. exigua becomes almost fully developed while still nestled in the leaf rosette. It is then thrust quickly upwards as the floral scape triples its length in the space of a few days. So a specimen collected in the pre-flowering stage may lead to false conclusions concerning the typical conformation of the inflorescence, which at anthesis is clearly elevated above the leaf rosette.
This species lives only in the Atlantic forest of the Serra do Mar in Sao Paulo state, at altitudes that vary from 50 m to at least 900 m. It grows mostly on rocks in the shade of the forest, and flowers from November to February. There is one clone in cultivation and the species has been recorded in the Caraguatatuba Forest Reserve where it apparently is protected. —SeeLeme 1998
