Reproductive biology of Melaleuca alternifolia (Maiden & Betche) Cheel

Abstract

Melaleuca alternifolia (Maiden & Betche) Chee! has a natural distribution in coastal and hinterland eastern Australia, between latitudes 28.43 - 29.55°S. The species is commercially important as the source of essential oil for the Australian tea tree oil industry. This study investigates and describes the reproductive biology of M. altern.ifolia: specifically, its floral biology and phenology; its breeding system, including self-incompatibility mechanisms; and its pollination biology. This study is the first comprehensive investigation of the reproductive biology of M. altern(folia; the information is important to the Australian breeding programme directed at improving the quality and quantity of tea tree oil. Components of the study were conducted in a managed natural population of M. alternifolia at Coombell, NSW; in two seed orchard populations, one of which (Wollongbar, NSW) borders on the spec ies' natural range in northern NSW, while the other (West Wyalong, southern NSW) is well outside it; in a plantation stand at West Wyalong; and in glasshouses at Canberra, ACT. The majority of the work was conducted from 2004 to 2007, although the study also drew on some prior observations. The floral morphology and development of M. alternifolia were monitored on sample trees in the glasshouse and at West Wyalong. Flowers are compounded in an inflorescence which develops acropetally. The compound inflorescence develops over a period of 6 days. No strong separation of male and female phases was found in any individual flower; pollen was shed by 1.4 days after anthesis and the stigma reached peak receptivity from 3 to 5 days after anthesis. Pollen of M. alternifolia remained viable after 52 weeks storage at -18"C. Dichogamy and acropetal floral development may lead to geitonogamy. Flowering phenology was observed within the Coombell natural stand and in the two seed orchards al West Wyalong and Wollongbar. Flowering intensity was recorded in all populations; the timing and duration of flowering were recorded in the seed orchards. M. alternifolia flowers synchronously at all three study sites, including the orchard well outside its natural range, with a peak period of three to four weeks. Flowering occurs during the months of October to November with the peak in November. Flowering intensity and success in producing capsules appears to be associated with total spring rainfall, as demonstrated by the poor flowering and loss of most resultant capsules during the severe drought at West Wyalong during 2006. Data describing environmental parameters which may influence flowering - daylength, temperature, and rainfall - were acquired for the natural population and seed orchard sites, and an ex pe riment investigating the effects of temperature on flowering was conducted in the glasshouse. Initiation of flowering in M. alternifolia appears to be con-elated with day length, or an environmental parameter closely con-elated with daylength. Flowering intensity varied considerably between years surveyed, sites and families, but appears to be promoted by a period of winter minimum temperatures below 5"C. Trees at West Wyalong flowered at a younger age and more heavily than trees at the other sites when spring rainfall allowed, and this appears to be a consequence of the lower winter temperatures at this site. Self-incompatibility was investigated by assessing seed yield, and microscopic examination of the development of pollen tubes, from controlled self- and cross-pollination with 4 unrelated families in the two seed orchards. No capsules were produced from controlled selfpollination; however, bagging of un-opened flowers produced several capsules, suggesting pollen contamination. There were significant differences in the mean number of pollen tubes found in the style between self-, cross- and open-pollination; with the greatest number of pollen tubes resulting from cross-pollination. Barriers to self-pollination appear to be operating in both the style and ovule of M. alternifolia, but there was some evidence that the effectiveness of these mechanisms varies between families. A small half-diallel mating design was used to investigate the mode of genetic control of seed yield. Whilst the results can only be considered as preliminary, it appears that SCA effects are stronger than GCA effects for seed yield; further investigation is required to confirm this. A preliminary investigation of hybridisation between M alternifolia, and the closely related M linariifolia, and M. dissitiflora revealed both possibilities for and constraints to hybridisation as a breeding strategy for tea tree oil production. Observation of floral visitors at the West Wyalong orchard, and exclusion experiments undertaken in both orchards, revealed that M. alternifolia flowers are visited by a variety of both large and small insects. There was no significant difference in the mean number of capsules set between open pollination and the finest mesh exclusion treatment (2 mm), suggesting that very small insects are the principal pollinators of M. alternifolia. Large numbers of Thrips imaginis and Thrips tabacci were identified as frequent floral visitors, and are the most likely pollinators of M. alternifolia. The results of this study are informative for M. alternifolia breeders. They suggest that breeding populations and seed orchards should be sited where winter temperatures are low, with minima below 5"C, and that the resultant earlier and more intense flowering at such sites will shorten generation times and increase seed yields. However, irrigation is likely to be necessary at such sites to ensure a high proportion of capsules mature when rainfall is low. Knowledge of stigma maturity means that pollen needs only to be applied once to achieve good results, and a count of developing buds 1 month after pollination is a good predictor of capsule numbers at maturity 16 months later. Pollen can be cold-stored for at least a year, and still retain a reasonable level of viability, but use of fresh pollen is preferable where possible. Whilst there are strong mechanisms favouring outcrossing in M. alternifolia, there are also characteristics of the species' reproductive biology that allow a level of natural self-pollination. Consequently, there are constraints to the effectiveness of conventional methods for isolation of flowers for controlled pollination, which need to be addressed if specific crosses are to be made for research or breeding purposes. The synchrony of flowering across sites makes it difficult for one research or controlled pollination team to work over multiple orchards located at any distance. Show more