Robert Ornduff, Villarsia exaltata 
809 
Discussion 
Villarsia exaltata possesses an incompatibility system linked to its floral dimorphism. 
The strength of this incompatibility system can be expressed by comparing mean 
seed-sets following intramorph pollinations with mean seed-sets obtained from 
intermorph pollinations. In the Coffs Harbour progeny, the mean seed-set of thrum 
plants was 0.5 seeds per intramorph pollination and 16.7 seeds per intermorph 
pollination. The mean seed-set of tlirum plants after intermorph pollinations thus was 
33.4 times greater than that after intramorph pollinations. The mean seed-set of pin 
plants was 3.8 seeds per intramorph pollination and 21.1 seeds per intermorph 
pollination. The mean seed-set of pin plants after intermorph pollinations thus was 5.6 
times greater than that after intramorph pollinations. These figures suggest that the 
strength of incompatibility differs between the two morphs, with that of thrum plants 
about six times greater than that of pin plants. It is possible, however, that this may 
instead reflect a difference in mean ovule numbers of flowers of the two morphs, a 
feature that unfortunately was not measured. Only thrum plants were present in the 
Wilsons Promontory progeny; these also produced low seed-sets after self- and 
intramorph pollinations. Pollen stainabilities of both morphs in the Coffs Harbour 
progeny were high, and pollen sizes of the two morphs were not consistently different. 
The preceding figures offer a summary of the strength of the incompatibility system 
at the population level in Villarsia exaltata, but they do not convey the level of variability 
among plants in the functioning of this system. In the Coffs Harbour progeny, mean 
seed-sets of individual thrum plants after self-pollinations ranged from 0 to 0.37 times 
(in S-7) that following intermorph pollinations. Mean seed-sets of individual pin plants 
after self-pollinations ranged from 0 to 0.5 times (in L-1) that following intermorph 
pollinations. Mean seed-sets of thrum plants after xenogamous intramorph pollinations 
were uniformly very low, but in pin plants they ranged from 0 to 1.2 times (in L-2) 
that following intermorph pollinations of the same plants. One pin plant (L-2) produced 
a mean of 25.6 seeds per pollination with L-1 as the pollen parent, and a mean seed- 
set of 25.8 seeds following all pollinations with thrum plants. One intramorph 
pollination combination of this pin plant thus produced a mean seed-set equal to that 
of the same plant following intermorph pollinations. Used as a pollen parent in pin 
plant x pin plant pollinations, L-1 produced seed-sets in three pin plants (L-2, L-4, 
L-5) that exceeded the mean seed-sets of these same pins in some intermorph 
pollinations. The strength of self-incompatibility and of intramorph cross¬ 
incompatibility in pin plants of V. exaltata clearly varies depending on the pair of 
parent plants used. The explanation for this variability is unclear, since on physiological 
and genetic bases the seed-set from pin plant x pin plant pollinations should be 
equivalent to that from self-pollinations of pin plants. 
Mean seed-sets of individual pin plants following all intermorph pollinations of 
them ranged from 14.4 and 12.2 in L-4 and L-5, respectively, to 28.6 and 29.8 in 
L-6 and L-3, respectively, thus differing by factors of L9-2.4 for the same category 
of pollination in the progeny. Mean seed-sets of individual thrum plants following 
all intermorph pollinations of them ranged from 3.8 in S-10 to 30.5 in S-7, thus 
differing by a factor of 8.0. While L-4, L-5, and S-10 functioned poorly as maternal 
parents, they functioned well as male parents in intermorph pollinations. The 
effectiveness of individual plants in their male and female functions thus varied 
depending on the combinations of parental plants involved in pollinations. 
Since the pioneering studies of Darwin (1877) it has been usual to report seed-set 
data obtained from pollination programs of heterostylous species as pooled data, 
usually collected from a single flowering season (e.g. Ganders 1975; Philipp and 
Schou 1981; Ornduff 1982, 1988b; Goldblatt and Bernhardt 1990). This practice 
