Summary. The overall aim of the research project funded by the Australian Flora Foundation was to examine the factors limiting seed production by Telopea speciosissima (Waratah) in a natural population at Barren Grounds Nature Reserve, near Wollongong, NSW. This project was proposed following a pilot study conducted by Whelan and Goldingay in 1985, in which it was apparent that the low levels of fruit set observed in the field could be increased by experimental outcrossing of flowers using hand pollination. This result contrasted with previous studies on Waratah, in different study sites, by Pyke (1982) and Pyke & Paton (1983).
In order to understand the factors which limit fruit set in nature and to resolve the apparent contradiction between different studies on Waratah, we felt that is was necessary (i) to gain more information about the basic biology of the species: its breeding system and natural pollinators, and (ii) to devise a set of experiments to reveal the factors limiting fruit set and the mechanisms by which they operate.
Three specific questions were addressed:
(i) What is the underlying breeding system of the species? (i.e. is it an obligate out-crosser? If self-compatible, can it set fruit by autogamy without a pollinator?).
(ii) How do pollinators influence fruit set?
(iii) Is a plant capable of allocating resources to "favoured" pollinations? (i.e. if the plant is self-compatible, does it nevertheless favour the development of fruits on the flowers that receive outcrossing?).
In short, the contributions made by our research to these questions are as follows:
(i) Waratahs are obligate outcrossers; no fruit set by autogamy or self pollination.
(ii) Pollination in the field appears to be less that maximal in some years. In 1985, plants were capable of producing more fruits when inflorescences were hand pollinated with outcross pollen. However, a repeat of this experiment in 1988, as part of the present study, found no such increase in fruit set, indicating that natural levels of pollination were adequate to achieve maximum fruit production.
(iii) Lack of self-compatibility removes the significance of this question. However, maturation of fruits on inflorescences appears quite plastic. If 2/3 of the flowers are removed from an inflorescence, it can still produce just as many fruits from the remaining flowers. Hand pollination of a portion of the inflorescence (top 1/3 or bottom 1/3, but especially the latter) produces greater initiation of fruits in the treated portion and a lower rate of abortion than for the same portion in control inflorescences.
Taken together, the results of this study reveal that Waratah is an obligate outcrossing species, at least in the Barren Grounds Nature Reserve population. Honeyeaters are abundant in the site and are frequent visitors to Waratah inflorescences (especially New Holland Honeyeaters) marsupials were also recorded visiting inflorescences. Despite their abundance, the birds carry small pollen loads of Waratah and visit inflorescences in such a way that pick-up and deposition of pollen is minimized. This finding alone may explain the fact that fruit set in the population is sometimes pollen limited. The first flowers to open on an inflorescence (i.e. the bottom flowers) have very low levels of fruit set which can be increased substantially by hand pollination. Hand pollination of flowers either on the top third or bottom third of the inflorescence leads to more fruit initiation in the treated portion and, interestingly, in lower levels of fruit abortion than in equivalent portions of control inflorescences. Active selection for favoured matings, through non-random fruit abortion, appears a real possibility in this species. This is worthy of further investigation.