The effects of tree isolation on the genetic diversity and seed production of Camden White Gum (Eucalyptus benthamii Maiden et Cambage).

Alison Skinner

Supervised by Craig Gardiner and Penny Butcher

Eucalyptus benthamii Maiden et Cambage is a tall, smoothed-barked gum found in limited areas south-west of Sydney. It is placed within section Maidenaria of the Symphomyrtus sub-genus of Eucalyptus and is closely related to Eucalyptus viminalis. It is most easily distinguished from E. viminalis by its 7-flowered inflorescence, broadly ovate juvenile leaves and smaller capsules. This species has attracted interest for its potential use in pulp plantations, after performing well in field trials in Australia, South America and South Africa.

Clearing of land for agriculture and the flooding of Warragamba dam has severely restricted the distribution of E. benthamii and it now exists in only three populations south-west of Sydney; found at the Kedumba Valley, Bents Basin and Camden. These have been estimated as containing ten thousand, 400 and 18 individual trees respectively.

Numerous E. benthamii seed collections have been made to ensure the conservation of genetic material into the future, but until now, very little was known about the genetic makeup of the three populations or the details of its seed production. This study aimed to determine the genetic diversity within, and genetic differentiation among populations, and identify whether the small size and isolation of the Camden and Bents Basin populations was affecting seed set, seed size or seed viability.

Leaf and capsule collections from up to 30 trees per population were made using a ‘Big Shot’ slingshot or .308 calibre rifle. DNA was extracted from mature leaves and analysed for 9 microsatellite loci. Microsatellites (or SSRs – Simple Sequence Repeats) are regions of DNA which contain a variable number of short sequence repeats. This variation, scored as alleles at each locus, was used to calculate; allelic richness of each population, their observed and expected heterozygosity, the presence of private alleles (those which occur in only one population) and the genetic differentiation among the populations. A seed stand (located at Yarralumla) containing progeny of Camden and Bents Basin populations was also sampled.

Comparisons of seed set, seed size and viability between the different populations followed the methods outlined in Burrows, G.E. (2000). This involved the collection and drying of capsules, and measurement of empty capsule weight, weight of capsule contents, number of seeds per capsule, and seed weight. The viability of seeds was tested by sowing them into moist vermiculite, and recording the number of germinants after fourteen days.

Table 1: Genetic diversity parameters for the three populations and seed stand.

• There was no significant difference in allelic richness or observed heterozygosity between the populations. However, the expected heterozygosity (mean=0.67) was less than the value reported for the widespread eucalypt, E. sieberi from eight microsatellite loci (H_E=0.87) (Glaubitz et al., 2001).
• Genetic differences among the three natural populations were significant and accounted for 7% of the total genetic variation for the species.
• A cluster analysis of the genetic distances among populations showed that the stands at Camden and Bents Basin were more closely related to each other than either were to the Kedumba Valley population. Genetic distance was correlated with the geographic distance among populations (r=0.97; P<0.05; Mantel test).
• Seed set and viability were significantly lower in the Camden and Bents Basin populations (Table 2).

Table 2: Population means for measures of seed production.

Although allelic richness did not differ significantly between the populations, the higher number of private alleles in Kedumba provides evidence for a loss of rare alleles from the smaller populations.

The degree of genetic differentiation found among the populations in this study suggests some restriction to gene flow. As most of the trees sampled are old enough to pre-date the land clearing events of the last century, fragmentation of the original population may have occurred much earlier than expected, or else this population was characterised by a strong genetic gradient. If gene flow barriers are present, inbreeding in the smaller populations is likely to become more prevalent in the future. This would be apparent as an increased heterozygosity deficit in seedlings of the first generation onwards.

It is possible that the current seed crop sampled from Camden and Bents Basin is exhibiting the effects of inbreeding, shown by their reduced seed set and viability. However the degree to which pollinator abundance and behaviour are responsible for these findings is not known. The decrease in seed size with increased seed set indicates a trade-off for resources at the individual capsule level, and is not thought to be an important factor in population dynamics.

Barriers to gene flow are present between the three remaining E. benthamii populations, as shown by their high level of genetic differentiation. Therefore, the conservation of the entire E. benthamii gene pool relies on the future availability of genetic material from all three populations. Although several seed orchards already exist, in situ conservation of genetic resources is preferable, and presently, both of the smaller populations are in danger of extinction. Should this happen, the total genetic diversity of this species would be permanently reduced. The findings of this study also support the importance of careful layout in seed orchards, to minimise the crossing of closely related individuals.

Although the populations currently have relatively high allelic richness and heterozygosity levels, the reduction in seed set and seed viability seen in Camden and Bents Basin may be an indication of inbreeding effects due to their isolation. If so, the value of seed from these populations for use in commercial plantings or revegetation would be severely reduced.

Further research into the genetic diversity of progeny of the sampled trees is necessary to provide conclusive evidence of inbreeding in the smaller populations.

Many thanks to Penny Butcher and Craig Gardiner (co-supervisors) for their guidance, encouragement and considerable input to the project. Thanks also to Chris Harwood for assistance with statistical analyses, the staff of the Australian Tree Seed Centre who provided technical support and advice, and the National Parks and Wildlife Service for access to the Kedumba Valley site.

This project was funded by the Australian Tree Seed Centre within CSIRO Forestry and Forest Products, and the NSW Environmental Trust, and was administered through the Centre for Plant Biodiversity Research summer scholarship program.

Useful references

Benson, D.H. (1985) Aspects of the ecology of a rare tree species, Eucalyptus benthamii, at Bents Basin, Wallacia. Cunninghamia. 1(3): 371-383.

Burrows, G.E. (2000) Seed production in woodland and isolated trees of Eucalyptus melliodora (Yellow Box, Myrtaceae) in the South Western Slopes of New South Wales. Australian Journal of Botany, 48: 681-685.

Glaubitz, J.C., Emebiri, L.C. and Moran, G.F. (2001) Dinucleotide microsatellites from Eucalyptus sieberi: inheritance, diversity, and improved scoring of single-base differences. Genome. 44: 1041-1045.

Prober, S.M. and Brown, A.H.D. (1994) Conservation of Grassy White Box Woodlands: Population Genetics and Fragmentation of Eucalyptus albens. Conservation Biology. 8(4): 1003-1013.