770 
Telopea Vol. 6(4): 1996 
the next hierarchical level (cells > meristems > metamers > modules > organ 
system > plant). Metamers are composed of four elements (node, internode, leaf and 
a^llary meristem) and the sum of metamers from a single meristem make up a module. 
Considering that selection can differently effect individual organs or elements within 
organs (Guerrant 1988), it is plausible to hypothesize that variation and selection as 
well as developmental and phylogenetic canalisation can occur at different levels 
within the metamer or even the subunit in inflorescences. In Carnarvonia it can be 
hypothesized that there is a level of canalisation of a subunit starting with the first 
three metamers. The first metamer does not have an elongate internode, the second 
metamer does have an elongate internode and the third bract primordium, based on 
developmental evidence, is positionally out of line with the predicted phyllotactic 
position of the bract primordium (160° as opposed to 100°-120"). The deviation in 
position of the third primordium of the third metamer is most likely a result of a 
timing (heterochronic) and/or spatial (heterotopic) shift between the initiation of the 
first and third primordium or possibly as a result of precoccious differentiation of the 
internode of the second metamer. 
Diversity of inflorescence architecture is the product of developmental shifts in timing 
and positions of various ontogenetic processes starting with the products and patterns 
of apical meristems (Grimes 1992). It has been hypothesized that physiological 
mechanisms inherent in a plant's architecture are affected at the level of axillary and 
apical meristems (White 1979, 1984; Grimes 1992) and that the duration and extent 
of growth of the meristem (that produces metamers or plant-units) is taxonomically 
constrained as well as ecologically influenced; conservation of form among taxa is 
the result. Under such a metameric concept, the inflorescence architecture of a plant 
is a product of conserved genetic and physiological mechanisms inherent in the 
plant or more specifically, shoot, as well as the product of historically imposed 
phylogenetic constraints (Cheverud et al. 1983, Janson 1992). 
There is a progressive acropetal reduction of flower numbers and branching patterns 
along the inflorescence. The variation of branching and/or flower numbers at 
levels along the inflorescence could be a function of the ontogenetic contingency of 
axillary meristems or in other words, the developmental fate of a primordium 
depends upon where and when it is produced within the architecture of the 
organism and what events have preceded it during ontogeny (Diggle 1995). In 
such a case, there is a shift in physiological and morphological products of the 
axillary meristems along the developing principal axes. An ontogenetic contingency 
as outlined above would necessarily involve a physiological or possibly an 
epigenetic feedback system (Sundberg et al. 1995). The variation in products of 
axillary meristems could be a result of resource availability or be based on a 
physiological cue to commit the axillary meristems to flower production given 
their position and size at a specific time. Both have been hypothesized in various 
plant systems. Mullins (1980) demonstrated that the inception of inflorescence 
meristems in Vitis can be induced with the presence of specific plant growth 
substances at specific times in development. In Carnarvonia, evidence for a 
physiological cue comes from the fact that subunits and inflorescences tend to 
develop synchronously on an individual and that there is a distinct difference in 
the shape of a meristem that will give rise to 'inflorescence' subunits/metamers 
(flat, less concave) compared to a meristem that will become a flower (higher and 
more concave). Physiological hypotheses have been proposed for the amplification 
of row numbers in Maize and putative relatives (Sundberg & Doebley 1990; 
Sundberg et al. 1995) and in the origin of the flower pairs in the proteaceous 
subfamily Grevilleoideae (Douglas 1994, Douglas & Tucker 1996a). 
