The Actinodium flowering head is a novel pseudanthium type
Reinvestigation of Actinodium inflorescence structure reveals that while Actinodium and true daisies correspond in their outward appearance, they represent an exciting example of analogy.
The typical head of a daisy develops from an expanded convex or flattened meristem that produces flowers in a centripetal sequence . Further medullar growth gives rise to the receptacle of the head. The outer flower primordia develop into monosymmetric ray florets while the tubular central florets are polysymmetric. The inflorescence meristem is completely consumed by flower production permitting no further growth, as known for “mantle-core” or “open II-type” inflorescence meristems [35, 36].
Actinodium differs in at least five basic characters from a true daisy. First, the meristem tip is not involved in the formation of the receptacle. Second, it can proliferate after flowering. Third, it is an inflorescence of the open I-type with a central-zone meristem . Fourth, the rays are not outer flowers, but branched short shoots, which, fifth, originate below the inflorescence and develop in a basipetal order.
In true daisies, the receptacle of the inflorescence originates by the enlargement of the reproductive meristem, which produces sessile flowers in a centripetal order. Meristem expansion is accompanied by a thickening of the medullar tissue that arranges the flowers typically on a flat plane. In Actinodium, however, the medullar tissue thickens after flower production, that is, below the active meristem, dislocating the already segregated flowers towards a horizontal plane.
The inflorescence shoot of Actinodium is able to continue vegetative growth after flowering, a phenomenon which has been termed by Troll  “inflorescence proliferation” (see ). This fact led Briggs and Johnson  to infer a “flexible condition at the meristem tip” of the Actinodium inflorescence (“conflorescence” after their terminology). Nevertheless, there are two possible explanations for the existence of proliferating inflorescences: either they actually rely on an inflorescence meristem that can be reverted to resume vegetative growth after flowering [39–43], or the supposed inflorescence is rather a vegetative shoot bearing lateral reproductive units, thus masking the appearance of a true inflorescence. This can be found in the same family, Myrtaceae (for example, Callistemon, Melaleuca, but also in other ones throughout the angiosperms (Drimys winteri, Mahonia aquifolium, and Lysimachia nummularia). Our observation of the sequential transformation of the inflorescence meristem in Actinodium definitively fits with the first interpretation.
The capacity of Actinodium to resume vegetative growth after flowering surely relies on the maintenance of the central zone in the inflorescence meristem throughout. Comparative developmental studies in open inflorescences have termed these inflorescences as open I [35, 46], in contrast to the meristematic organization of daisy heads, which do not show either central zone or proliferation capacity in the wild type at least.
In some Asteraceae, the dense cluster of flowers is surrounded by a circle of ray flowers. These outer flowers differ from the actinomorphic bisexual flowers in the center by their increased corolla size, monosymmetry, and female or sterile nature . Troll  termed the flower-like inflorescences pseudanthia (although this term was used in a different context before, see ), indicating their overt similarity with flowers to be a classical example of analogy. The remarkable similarity of the swamp daisy Actinodium cunninghamii to a true daisy almost certainly prompted Bentham  to interpret the rays as sterile flowers. Briggs and Johnson , Holm  and Claßen-Bockhoff  followed his interpretation, and only N. Marchant (unpubl. data), while preparing a revision of the Chamelaucium group, found the rare flowers in the ray structures and consequently concluded these to be short-shoots. Ray florets in true daisies are the outermost flowers of the inflorescence. They may originate with some delay compared to the disc florets but always arise from the same head meristem [34, 49]. In Actinodium, however, such a head meristem does not exist. Experiments in Arabidopsis thaliana illustrated vegetative buds below the main inflorescence to be stimulated by light and auxin flow to develop late lateral inflorescences in a basipetal sequence. The ray shoots in Actinodium develop in a basipetal order as well, showing concordance with the lateral inflorescences found in Arabidopsis. Basipetally flowering shoots separate from the terminal inflorescence have been termed “paraclades”  and may well represent the ray shoots in Actinodium.
The unique organization of the flower-like inflorescence of Actinodium, not known from any other plant family, requires the recognition of a novel pseudanthium type [26, 48]. This floral mimic is characterized by an inflorescence meristem with a persistent central zone, able to proliferate, a receptacle originating from medullar thickening, and showy paraclades composed of branched short-shoots.
The inflorescence of Actinodium is influenced by CYCLOIDEA-like gene activity
In angiosperms, shoot branching patterns are strictly controlled in order to achieve proper architecture. For example, in Arabidopsis and tomato, TCP transcription factors belonging to CYC1 clade play a key role in arresting axillary bud growth [14–16]. BRANCHED1 (BRC1) in Arabidopsis[14, 15], and two BRC1 paralogs in tomato  are all expressed in arrested axillary buds and down-regulated upon bud outgrowth. In Arabidopsis, the outgrowth of axillary buds typically occurs when the plant transforms to reproductive stage. Similar to tomato, we found two CYC1-like genes in Actinodium (Figure 5), indicating that gene duplication may have occurred during molecular evolution of both species. The two Actinodium CYC1-like genes share an expression pattern that correlates with the branching pattern of short shoots in the pseudanthium (Figure 6). In the outermost elements where the expression level is highest, activity of these genes may contribute to reproductive repression by preventing outgrowth of tiny buds located in the axils of the short shoots. In the inner, non-branched fertile units, the expression levels of CYC1-like genes were extremely low. In situ analyses of expression patterns in short shoots and flowers were not successful enough to provide tissue-specific expression patterning. Nevertheless, our qRT-PCR data provide strong correlative evidence, although functional studies would be required to confirm a role of AcCYC1-like genes in reproductive repression in Actinodium. For a non-model organism such as Actinodium, use of heterologous systems would be needed, in which case interpretation of results might be challenging. If the Actinodium pseudanthium can be compared to an individual Arabidopsis plant, with the showy sterile short shoots and their suppressed bud outgrowth analogous to an Arabidopsis rosette in its vegetative form, AcCYC1a/b and AtBRC1 might share a function in controlling reproductive development via shoot branch suppression.
While AtBRC1 seems to have a distinct function in controlling Arabidopsis shoot branching, the role of AtTCP1, a CYC2-clade gene, has been less clear. Recently, however, AtTCP1 has been shown to affect shoot development in terms of elongation of leaves, petioles and inflorescence stems , possibly in concert with hormonal regulation . In Actinodium the slight decrease in expression levels of AcCYC2 towards the inside of pseudanthium (significantly lower, however, in the innermost short-shoots) correlates with decreased length of bracteoles, and decreased elongation of hypopodes, which contributes to the flat shape of pseudanthium. Thus, in the case of Actinodium, a CYC2-clade gene may have been recruited to enhance showiness of the inflorescence by bracteole elongation, instead of floral symmetry changes as in the case of Asteraceae. Both strategies may serve in pollinator attraction rather than reproduction, as the ray (or ray-like) elements are often sterile.
In summary, CYC-like genes may be involved in providing the Actinodium pseudanthium with its unique structure: AcCYC1a/b via short-shoot branching and AcCYC2 via bracteole and hypopode elongation, thereby contributing to the showiness and reproductive success of the inflorescence. Future attempts to clone and characterize the expression of Actinodium CYC3-like genes may be similarly illuminating, as at least one such gene has been implicated in the control of flower type in the sunflower .