Types of organ fusion in angiosperm flowers (with examples from Chloranthaceae, Araliaceae and monocots)

Abstract

Fusion between floral organs or their parts is believed to have played key roles in the origin and subsequent diversification of angiosperms. Two types of fusion can be recognized: postgenital and congenital. Postgenital fusion is readily observable during flower development: primary morphological surfaces of contacting structures meet and join during this process. After perfect postgenital fusion, no trace of the original epidermal layers can be recognized, but these remain visible, often in modified form, after imperfect postgenital fusion. Congenital fusion cannot be directly observed and takes place due to differential growth. In the case of complete congenital fusion, free parts of fused organs cannot be seen at any developmental stages. Incomplete congenital fusion implies the presence of free organ parts on the common (united) base; it can be divided into early and late congenital fusion depending on whether the common base precedes or follows the initiation of free parts during development. Phenomena related to congenital fusion are the development of free organs from common primordia, hybridization of developmental pathways, loss of organ individuality, heterotopies and fasciation. Differences between congenital and postgenital fusion are much more unequivocal than those between the presence and absence of fusion. There is no abrupt boundary between imperfect postgenital fusion and transient contact between organs during development. Structures assumed to be congenitally fused clearly develop as a unit, but it is necessary to demonstrate that these structures indeed belong to different merged organs (instead of being parts of the same organ or two distinct organs on a common base). This only can be done in the framework of comparative morphology. Analyses of both types of fusion involve arbitrary decisions, so it is not appropriate to discard the existence of any type. Conventional interpretations of morphological concepts lie at the base of analyses of character evolution, even if they are performed using maximum parsimony or model based methods and molecular phylogenetic data. Patterns of organ fusion are discussed here using three case studies.