KRAMERIA (KRAMERIACEAE) FLOWERS: ORIENTATION AND ELAIOPHORE MORPHOLOGY

Abstract

SummaryThe flowers of Krameria have been misunderstood since they were first described by Loefling in 1758. Some of the misconceptions about the flowers stemmed from the failure of authors to recognize the organography of the two fleshy structures which flank the ovary. Other misinterpretations resulted from confusion about the orientation of the flowers. Twice the controversy as to whether or not the flowers of Krameria species are resupinate was raised and laid to rest, but doubts still linger. Recent morphological and anatomical studies have unequivocably shown that the flowers are not resupinate and that the three small petals are adaxial. The two abaxial petals are highly modified into the two fleshy structures that have only recently been explained. Within the last five years it has been shown that these petals are glandular and play an intricate role in the pollination biology of Krameria by secreting lipids that are collected by female bees of the genus Centris (Anthophoridae). A survey of the anatomy and morphology of the glandular petals, or elaiophores, has yielded a suite of characters that are useful in showing branching patterns within the genus because they show a sequence of specializations of the outer elaiophore surface from elongate, irregular patches of secretory cells over the entire surface covered by a cuticle that separates smoothly from the cells as oils accumulate beneath it, to either small, round or polygonal patches of secretory cells over the entire surface, or to restrictions of the secretory tissue to the distal portion of the fleshy petal. The trends in specializations of the secretory surfaces of the glandular petals show correlations with trends based on sepal arrangement, degree of fusion of the superior petals, and pollen morphology.