Anjanette Marwick – The University of Melbourne, Burnley College


In the past, the improved selection of Australian native trees by means of vegetative propagation has been limited to high value forestry trees, with little of this knowledge applied to important amenity trees. One tree worthy of such an approach is Eucalyptus leucoxylon (Yellow Gum, South Australian Blue Gum), which has been planted extensively throughout metropolitan Victoria and South-Australia. The use of Eucalyptus leucoxylon within urban areas has for the most part been of the variety ‘Rosea’. Despite this, the natural origins of this cultivar are unknown. Consequently, as progeny have been derived from a tree in cultivation there is the distinct possibility there has been a loss of vigour due to inbreeding. This paper reports preliminary work which should lead to the identification of natural populations of Eucalyptus leucoxylon with qualities to equal those of ‘Rosea’. The subsequent development of vegetative propagation methods will ensure that regardless of the extent of use, desirable characteristics will be retained with no loss of vigour.


The variation within Eucalyptus leucoxylon is significant, with seven subspecies (Boland, 1979; Rule, 1989 -1992; Rule, 1998) described. One of these has been regarded as different enough to be elevated to species status, Eucalyptus petiolaris (Rule, 1989 – 1992). To understand the growth, development and variability within Eucalyptus leucoxylon, several of these subspecies were monitored for a period of nine months from germination.


Seed of Eucalyptus leucoxylon was sourced from eight locations (selected purely on the commercial availability of the seed) representing four subspecies (Boland, 1979), Eucalyptus leucoxylon ssp. leucoxylon, ssp. megalocarpa, ssp. pruinosa and ssp petiolaris (still used in this trial despite its current species status). For ease of labelling and identification, each population has been designated by a code letter, A-H (see Table 1). There is uncertainty as to the origins of the West Bendigo seed (D and E), but for the purposes of this paper they will be referred to as the West Bendigo provenance or population.

Table 1 Eight collection sites of Eucalyptus leucoxylon seed, including subspecies and identification code.


Figure 1 Eucalyptus leucoxylon populations used in the trial from South-eastern Australia. 

Eucalyptus leucoxylon ssp. leucoxylon, A; Adelaide Hills, B; Bendigo. Eucalyptus leucoxylon ssp. megalocarpa, C; Monarto, D&E; West Bendigo Eucalyptus leucoxylon ssp. petiolaris, F; UngarraCockaleechie, G; WarundaKoppio. Eucalyptus leucoxylon ssp. pruinosa, H; Horsham area.

Experimental Design and Measurements

Seed Characteristics

All seed was separated from the chaff, to ensure there was no bias towards size. 30 replicates of 30 seeds were selected and weighed for each seed lot. From these weights the number of seeds per gram was derived (Figure 2).

Germination, Growth and Development

A glasshouse germination trial was established at Burnley College, in September 2001. Seedling trays were set up in a randomised block design with 12 blocks, each representing the eight populations. To satisfy the light requirement for germination (Turnbull & Doran, 1987), the seed was sown on top of Burnley seedling raising mix, and sprinkled with vermiculite. Initially there were three seeds per cell for a total of 720 seeds per provenance, however after recording daily germinants for six weeks these were randomly culled to only one plant per cell (240 per provenance). At four months of age these were transplanted to 14cm olive pots, and placed outside under 50% shade. Regular measurements were taken to assess the form and variability between and within all provenances (Table 2).

Table 2           Summary of the attributes recorded for Eucalyptus leucoxylon and the regularity of measurements taken.

Table 2 Summary of the attributes recorded for Eucalyptus leucoxylon and the regularity of measurements taken.



All of the seed groups had significantly different (p<0.05) weights, except for sample F (E. leucoxylon ssp. petiolaris) and H (ssp. pruinosa). In general (Figure 2) the seed of Eucalyptus leucoxylon ssp. megalocarpa was heavier than for the other subspecies, with the exception of the population from Monarto (C), which had the lightest seed, however the proportion of those with no embryo may have contributed to this result (Table 3).

Figure 2 Number of seeds per gram for eight provenances. Eucalyptus leucoxylon ssp. leucoxylon (A & B), ssp. megalocarpa (C & D & E), ssp. petiolaris, (F & G), and ssp. pruinosa (H)


The modified final germination percentage (Table 3) reflects the number that actually germinated as a percent of those that would be expected to be viable given that they have an embryo (although it is acknowledged that the presence of an embryo is not necessarily an accurate indication of viability). For all provenances, this is greater than 80%. The two provenances of Eucalyptus leucoxylon ssp. leucoxylon from the Adelaide Hills and Bendigo, and E. leucoxylon ssp petiolaris from Ungarra-Cockaleechie both had final germination percents, which were higher than those that were expected to be viable.

Table 3 The percentage of seeds germinated, and the modified final germination percent, given the number of seeds without an embryo.


Table 4 outlines some of the characteristics that vary between the provenances. Many differences are evident, however the most prominent is probably the petiolate leaves in the two populations of ssp. petiolaris (Eucalyptus petiolaris) from the Eyre Peninsula. Generally these petioles are between 3 – 12mm in length, compared to the other subspecies which rarely reach 5mm. This and the early development of alternate leaves – as early as the fifth leaf pair in this provenance – are among the main criteria for its elevation to species status (Rule, 1989-1992). For most of the other subspecies, alternating leaves occur rarely (although most plants still have less than 20 leaf pairs), and if so, it is generally not until at least the 11th leaf pair.

The shape (as described by Brooker & Kleinig (1999)) and size of the leaves vary considerably with most displaying ovate leaves for the first seven or eight leaf pairs. The Eyre Peninsula provenances have elliptic leaves, and the leaves from both populations of E. leucoxylon ssp. megalocarpa from West Bendigo are considerably larger (see Table 4) and generally cordate and amplexicaul. Also worth noting is the tendency to develop lateral stems. The Victorian megalocarpa subspecies very rarely did, however it is strongly pronounced in ssp. megalocarpa from Monarto, and E. leucoxylon ssp. leucoxylon from the Adelaide Hills. The leaves in these latter subspecies are becoming distinctly lanceolate.

The glaucous covering that is unique to Eucalyptus leucoxylon ssp. pruinosa (Boland, 1979) within this group is evident in only half of those from the Horsham area, and even these only have a very fine waxy covering, with usually only two or three leaf pairs affected.


The differences evident within the selected subspecies of Eucalyptus leucoxylon have so far been limited to the juvenile characteristics. Differences have been observed in the growth rate, leaf size, shape, and arrangement; the presence or absence of petioles and lignotubers; seed weights, and the rates of germination. The effect that these differences will have on future growth and vigour is yet to be determined, however the close monitoring of these trees for a further 18 months should facilitate a greater understanding of how these traits relate to the development of mature trees.


Table 4 Growth characteristics of selected attributes in eight provenances of Eucalyptus leucoxylon



The inability to predict the final height and habit of Eucalyptus when growing in streetscapes is often the result of raising plants from seed. There is a need for the development of propagation systems to ensure offspring can be produced with desirable qualities, and a predictable mature height and form.

Budding and grafting of plants onto suitable rootstocks will ensure that subsequent scion growth will be representative of the material from which the plant was sourced. The success of this approach in Eucalyptus leucoxylon has not been well documented, so this preliminary trial was done to determine whether it is possible to graft six month old Eucalyptus leucoxylon ssp. megalocarpa plants onto other ssp. megalocarpa plants, of the same age and from the same seed source. It was also set up to determine whether a successful union could take place if a plant is grafted back to itself. Part of this trial will include the prevention of the cut surface from drying out, by keeping all surfaces wet during the grafting procedure.

Experimental Design

Eucalyptus leucoxylon ssp. megalocarpa plants were purchased from Mildura Native Nursery in October 2001. These were grown on until January 2002, 6 months of age. A total of 144 plants were used.

The plants were randomly allocated to groups and treatments, to ensure there was no bias as the budding and grafting techniques improved. Plants within the group were chip budded and grafted using the splice (whip) and whip and tongue graft (Hartmann et al, 2002).

To prevent the cut surface from drying out, half of the plants were fully submerged and the whole process was carried out under water, including the tying. For others, the time that the cut surfaces were exposed to air was kept to an absolute minimum, however in some cases this may have exceeded ten seconds. The four treatments for each graft type were as follows:

  1. The scion taken from a plant and grafted back to itself EXPOSED
  2. Plants are paired and the scions swapped between them EXPOSED
  3. The scion taken from a plant and grafted back to itself SUBMERGED
  4. Plants are paired and the scions swapped between them SUBMERGED

 This trial was undertaken in January 2002 at Burnley College Nursery. It was carried out undercover, however it was still subjected to air drafts and temperature changes. Air temperature for this trial was between 16oC and 22oC (Bureau of Meteorology). Water temperature was between 20oC and 22oC, and the relative humidity was quite high (Bureau of Meteorology).

Grafted plants were placed in a fog house maintaining 90% humidity for two weeks. Following this they were placed in a misting house for a further three weeks until the tape was removed, and then taken outside.



Overall results for this experiment were disappointing, with around a 10% success rate. This further fell to 4.17% in the following weeks, as the successful chip budded plants failed a week or so after tape removal.

Mortality rate was quite high with rootstock death occurring in 12 of 144 (8.33%) of the plants.

There was some sprouting of the scion buds on those plants that were whip and tongue, or splice grafted within the first week following the graft procedure. Some of these continued to develop; however in many trees the scion, and consequently the buds, died after a few days.

To date, the only successful grafts have been the whip and tongue, and the splice graft. These are still alive after seven months with even growth and a strong union.

Even though the numbers remaining are too small to get a clear indication as to the effect of submerging the cut surfaces, it can be seen that the plants are able to be grafted onto themselves, or others from the same seed stock, in either wet or dry conditions.


Although not actually measured, it appeared that the successful grafts were generally on plants that had slightly thicker stems.  Most of the failed grafts showed early necrosis beginning at the tip of the scion. The young age of these plants combined with the inexperience of the grafter may explain this low success rate. Further trials are planned for the future.


The preliminary results obtained, have emphasised the considerable variation that is present within Eucalyptus leucoxylon. The appearance and habit of the juvenile plants sourced from eight different sites throughout Victoria and South Australia have shown differences in both the pattern of development, and juvenile features. Whether these differences will be evident or significant in the mature plant, remains to be determined.

Despite the low success rate, the ability to vegetatively propagate Eucalyptus leucoxylon ssp. megalocarpa appears to be promising. The rate of failure could be attributed to the young age of the plants, and the small surface area over which the union was to occur.  Further work has been planned which will involve the use of older plant material, with the effect of stock pre-treatment and season also investigated.

At the conclusion of this project – June 2004, it should be possible to recommend provenances of Eucalyptus leucoxylon that are suitable for planting in urban areas. A method of vegetative propagation, which will ensure that a sustainable number of plants can be produced, reliably exhibiting the various features the provenance was selected for, should also have been determined.


  • Boland, D. J. 1979. A taxonomic revision of Eucalyptus leucoxylon F. Muell. Australian Forest Research 9. 65-72.
  • Brooker, M. I. H & Kleinig, D.A. 1999. Field Guide to Eucalypts. Volume 1. South-eastern Australia. Bloomings Books, Hawthorn, Victoria. 353 pages.
  • Bureau of Meteorology, 2002. Hourly Observations – Melbourne.
  • Hartmann, H. T., Kester, D. E., Davies Jr, F. T. & Geneve, R. L. 2002. Plant Propagation: Principles and             Practices. 7th Edition. Prentice Hall, New Jersey. 880 pages.
  • Rule, K. 1989 – 1992. Two new species of Eucalyptus (Myrtaceae) in South-Eastern Australia. Muelleria 7. 497-505.
  • Rule, K. 1989 – 1992. Two new subspecies within Eucalyptus leucoxylon F.Muell. and notes on that species. Muelleria 7. 389-403.
  • Rule, K. 1998. A new, rare Victorian subspecies of Eucalyptus leucoxylon F. Muell. Muelleria 11. 133-136.
  • Turnbull, J. & Doran, J. 1987. Species of Myrtaceae requiring light for germination. In Langkamp, P.J. Germination of Australian native plant seed. Melbourne, Inkata Press. page 198.

Leave a Comment