All Content © CRC for Forestry 2007

Environmental selection key to stringy bark diversity

­­Dot Steane and Justin Bloomfield­

Justin Bloomfield (right) and supervisor Dr Dorothy Steane select mature stringy barks for the study.

Justin Bloomfield
University of Tasmania

 In February, Justin Bloomfield (left) was awarded Honours in Biotechnology at the University of Tasmania for his work on genetic diversity in Eucalyptus obliqua in Tasmania.  Justin is presently “filling some gaps” in the sampling, and when the complete data set has been reanalysed the results will be prepared for publication.  Below is a brief summary of Justin’s work.

Eucalyptus obliqua (Figure 1), commonly known in Tasmania as “stringy bark” - and on mainland Australia as “messmate” - is widespread in south-eastern Australia.  In Tasmania it is actively harvested from native production forests by a variety of means.  Most harvesting of E. obliqua is done using "partial" methods: aggregated retention in wet forests and typically a combination of seed tree retention, potential sawlog retention and/or advanced growth retention in dry forests.  Less than half of the harvesting of E. obliqua involves a clearfell, burn and sow silviculture regime.  To maintain species patterns and to conserve local gene pools whilst continuing timber harvesting, Forestry Tasmania has developed and adopted a protocol for regenerating areas that have been clear-felled, using aerial sowing of seed after the coupes have been burnt.  Tasmania has been divided into “seed zones” on the basis of environmental variables such as geology and rainfall.  ­ ­

­­ Lune River Tree

Figure 1. A fine specimen of a mature stringy bark at the Lune River "plain" population.

The regeneration guidelines require that, wherever possible, seed should be collected on-site prior to harvesting and used to resow the site.  When there is insufficient seed on site, transfer of seed should be within a seed zone and, if no seed is available from a particular zone, there are guidelines for the transfer of seed between similar seed zones.  When the seed zones were defined and last reviewed, information on E. obliqua regarding genetic diversity within seed zones and relationships of populations across seed zones was not available.  Forestry Tasmania was interested to know whether genetic information on native E. obliqua populations would have implications for changing the guidelines in order to maintain the spatial integrity of the genetic variation in the species.

A study done in the 1990s by Graham Wilkinson (Wilkinson 2008; see article from Biobuzz 5) found genetic variation in quantitative traits between E. obliqua trees from very different environments in close proximity to one another.  For example, Wilkinson collected seed from two sites at Lune River in Tasmania’s south.  One population was on a well-drained slope in wet sclerophyll forest; the other, just 180 – 620 m away, was on an exposed, wet plain at the bottom of the slope.  Another two sites were selected at Forestier Peninsula, on Tasmania’s south east coast.  Wilkinson planted all seed lots in a common garden at each of the four sites.  Over 45 months, he measured a range of quantitative traits and found that there were significant phenotypic differences between the progeny from the four localities.  For one part of his study, Justin resampled, as closely as possible, Wilkinson’s parent trees (some adjustments were made because of harvesting in some areas).  Justin wanted to test whether the variation in quantitative traits that Wilkinson observed between populations in close proximity to one another was the direct result of natural selection and adaptation in the face of ongoing gene flow between localities, or whether random genetic drift due to genetic isolation of populations was involved.

Figure 2

Figure 2. Tasmanian E. obliqua distribution map overlayed with the locations of the populations sampled in this study.  Lune River, Forestier and Mount Lofty had two populations sampled at each location. The distribution map indicates whether E. obliqua is present (black dots) or absent (no dots) within a 10 km square grid (Williams and Potts 1996).

­ In his study, Justin sampled 300 trees from 14 populations across the species’ range in Tasmania (see Figure 2).  The populations included Wilkinson’s four sites at Lune River and Forestier Peninsula.  Justin fingerprinted all the trees using seven microsatellite loci.  Microsatellites are short regions of repetitive DNA - eg, ACACACACAC – that tend to be highly variable because, during replication, DNA synthesising enzymes tend to get a bit muddled with all the repetition and sometimes accidentally add in or cut out one or more repeats, thereby producing different sized microsatellites in different lineages.  Microsatellite markers are generally assumed to be selectively neutral, so when we analyse these data we are examining the underlying neutral genetic diversity rather than diversity that results from selection.

From the fingerprint data, Justin was able to work out population-level diversity statistics.  He found that all the populations had similar and reasonably high levels of genetic diversity (expected heterozygosity, He = 0.79; observed heterozygosity, Ho = 0.78).  There was very little differentiation ­
­Figure 3

Figure 3.  Relationship between geographic distance and Nei’s (1972) genetic distance among 14 Tasmanian E. obliqua populations. r2 indicates strength of relationship between the two distance measures.

between populations, suggesting that there is gene flow (eg, through pollen and/or seed movement) between the populations that were sampled.  With respect to Wilkinson’s populations, there was no significant differentiation in microsatellite profiles between the paired sites.  This means that the quantitative variation observed by Wilkinson is likely to be a result of environmental selection and adaptation in the face of ongoing gene flow between phenotypically different populations.  There are no obvious genetic isolation mechanisms operating between the adjacent populations.  This is in stark contrast to the results of Foster et al. (2007) who found that dwarf populations of E. globulus growing adjacent to tall populations of the same species are highly differentiated in their microsatellite profiles and genetically isolated by differences in flowering time.  

On a larger scale, over the whole of Tasmania, there was a subtle - but significant (r2 = 0.41) - pattern of differentiation of populations that formed a classic “isolation by distance” trend (Figure 3).  This means that the further apart two populations are, the less related they tend to be (and vice versa).  This sort of trend is often seen in species with large continuous distributions.  There was also higher “allelic richness” in the east than in the west (Figure 4). 
Figure 4

Figure 4.  Allelic richness across seven microsatellite loci in 14 Tasmanian E. obliqua populations. Yellow circles indicate higher than average allelic richness in a population; red triangles represent below average allelic richness in a population. Average allelic richness across all populations is 10.3.

­ Allelic richness is a sensitive measure of genetic diversity, so in E. obliqua there tends to be more genetic diversity in the eastern regions than further towards the west.  Hence, inland areas of Tasmania may have been colonised from the east (eg, during the last glacial maximum inland areas were too cold to support forest trees) and areas on the west coast may have experienced some loss of genetic diversity due to isolation or because they were colonised last.  Eastern Tasmania is, thus, an important zone for the conservation of genetic diversity in the species.

Returning to the issue of seed zones and seed transfer guidelines, can we comment on whether Forestry Tasmania has got them right?  From our data so far it would appear that in E. obliqua some of the phenotypic variation that we observe between populations is a result of environmental selection and adaptation in the face of gene flow.  We observed that there is more genetic diversity in eastern populations and we are hoping to confirm this with more evidence from increased sampling.  Seed transfer guidelines recommend minimising the distance and environmental gradients across which seed is transferred; at this stage we can not recommend any improvements to this­ strategy.

Biobuzz issue eight, March 2009