You are here: Home» Membership» NZFFA Member Blogs» Ian Brown's Blog» Blackwoods - selecting the best trees

Blackwoods - selecting the best trees

Thursday, July 04, 2019, Ian Brown's Blog

In theory, there are two simple steps in tree improvement. Select a provenance that matches your site, and from it identify trees with superior traits to breed from. However in practice, it isn't quite that simple.

For radiata pine, provenance selection offers limited choices. There are just three locations to choose from on the Central Coast of California. (Most of our original land race appears to have been sourced from a scruffy patch of coastal forest at Ano Nuevo, near Santa Cruz. Obviously, appearances can be deceptive.)

However like most of the other big conifers that like to assemble themselves in monocultures on their home territory, pines tend to breed true, the kids generally resembling their parents. Although this has made life easier for the tree breeders, the present generation of quality pines that we have still needed many decades of world class research.

In contrast, with blackwoods we have an abundance of provenances to choose from, each of them adapted to its location. But for potential tree breeders there is a problem with individual tree selection. Blackwoods are at home in mixed communities, where survival demands flexible growth responses when jostling for space among some mean competitors, and this has shifted the balance between nature and nurture.
Like human societies, the best behaved blackwood parents often produce wayward offspring, easily led astray by environmental influences. On the other hand, the parent trees can deliver some unfortunate traits to their kids. You might consult the opening lines in Philip Larkin's poem, a favourite among small children, "This be the verse".

Provenance selection.

The natural range of blackwood is remarkable, even among Australian species, extending from the cool damp forests in Tasmania to the warm tropics in Queensland. So which provenance is best suited to New Zealand?

Blackwood genetics.

In 1993 J. Playford and colleagues published a genetic study, based on allozyme variation. They tested 27 blackwood provenances, covering the natural range. I will summarise some points of interest.

  • there is considerable genetic variation between provenances, linked to latitude of origin.
  • there is an even higher degree of variation among trees within each provenance.
  • there is a major genetic disjunction among populations, corresponding with the Hunter River district in NSW. Blackwoods north of that zone are genetically distinct from those further south, and may be in the process of separation from them further, to form a distinct species. Blackwoods are a temperate species, and in the north they find themselves outside their comfort zone. They are more at home there on the hills and ranges, where they form isolated communities. Geographical separation is one of the factors linked to speciation.
  • further south, there are some interesting anomalies between populations that are geographically separate but genetically linked. You might expect that the provenances in NE and NW Tasmania, that perform equally well in trials and which are close geographically, would have a strong genetic linkage. In fact each of them has a closer genetic linkage to mainland populations on the other side of the Bass Straight, NW Tasmania to the Otways, and NE Tasmania to South Australia and Victoria. It is as if the Bass Straight, a formidable stretch of water, did not exist. Of course until recently, in geological time, it did not. The former Bassian Plain, which linked Tasmania to the mainland during the recent ice age, with woodland on the western flank and dry steppe to the east, would have allowed a flow-path for gene transfer until il was flooded by rising seas 14000 years ago, in the process isolating mainland Australia from civilisation in the south.
  • the study tells us nothing about the specific attributes that are of interest to us, i.e. vigour and form. Some growth traits can be very localised. For example, we could take phyllode structure. This has an association with latitude of origin. In the northern populations the phyllodes are longer and thinner than those in the south. However phyllode morphology has a stronger genetic linkage to local influences, in particular the distance from the coast. Among inland populations the phyllodes are long and thin, tough and waxy, sacrificing photosynthetic output for survival in a hot dry climate. Among coastal blackwoods where soils and climate are more favourable, the challenge for survival is competition for light. Here the phyllodes are large and fleshy, giving them a high photosynthetic output, and as a consequence a greater growth potential.

The genetic evidence from studies of this type may be interesting, but from the perspective of tree growers it is of limited value. To obtain the information we need, we have to rely on field trials.

Provenance trials

Offshore trials

Approximately 60 provenance trials have been run internationally over many years. From these we might expect an abundance of useful information. Gordon Bradbury in Tasmania has looked at them critically and has found major flaws in most of them:

  • provenance numbers have often been limited, and focussed on local sources.
  • the trial sites have often been of poor quality, dry and exposed, and with impoverished soils. I have seen one or two of them, and would not choose to plant blackwoods there.
  • poor management. This includes lack of weed and predator control.
  • lack of documentation, and few published results.

A common theme is that local provenances perform best, but there are exceptions. In trials in SW Tasmania and in SW Victoria, trees from NW Tasmania outperformed the local provenance. In NE Victoria, NSW, and Queensland, the local provenances scored well.

In trials In South Africa, where blackwoods have been grown from the 19th century, local selections outperformed those from offshore.

The data from China are interesting. The Chinese have been trialling blackood from the 1990s in extensive experimental plantations in the Guangdong province, in the south east. They are keen to establish blackwood plantations because it matches the traditional timber used in their most valued furniture. They have tested 35 provenances, and early data have shown the best results from Queensland. (Perhaps this should not be surprising. Anyone who has spent time in Hong Kong will know that the climate in SE China matches Queensland more closely than cool Tasmania.)

In a trial site in Lanco, Chile, the Tasmanian provenances performed best. The NE Tasmanian provenance headed off NW Tasmania.

Meunna Trial

Of the Australian trials, The one that I think is most useful to us was set up in Meunna, in NW Tasmania in 1988 and 1989. The site has good quality soils, and high rainfall, so it matches the sites that we would use. It tested 56 Tasmanian families, and 7 from the mainland. The trial was carefully planned and managed. The trees were protected by tubes from predation by marsupials , and were tested in plots with and without a eucalypt nurse.

Gordon Bradbury assessed the outcome at age 17 and 18. The best performing provenance was from NE Tasmania (Scottsdale), followed by NW Tasmania. SE Tasmania lagged behind. The provenances differed in volume, but not in form, which was consistently bad. Gordon made some interesting observations on wood colour, which I will refer to later.

New Zealand Trials

There have been a few limited trials carried out in the past, and two more substantial trials which I will describe.

1984 trial.

This was established by FRI on 10 locations, 3 in the North Island and 7 in the South, ranging from Tairua in the Coromandel to Longwood in Southland. Of the 31 seedlots that were tested, there were 9 from Australia (6 Tasmania, 2 Victoria, and one NSW). 14 were from South Africa, and the rest were from New Zealand and Chile.

The North Island trial sites were assessed at age 8 in 1992. The best provenances were Smithton (NW Tasmania), Scottsdale (NE Tasmania), Jeeralang (Victoria), and Waipoua Northland. The Waipoua trees had been sourced from a Smithton seedlot. The South African and Chilean provenances performed poorly. The Jeeralang trees had good form but lesser diameter. Dudley Franklin examined the trees in the south for frost tolerance, and found that the Tasmanian provenances were most resistant.

A further detailed assessment was carried out on two of the North Island sites in 2002, at age 18. The beat provenances were Scottsdale ( marginally on top), Smithton, and Waipoua.

2004 trial

Ian Nicholas and colleagues from FRI established two trial sites, one on my property at Pirongia in the Waikato, and the other in Westland. They contained 65 seedlots from 13 districts. 12 of the districts were in Australia and one in New Zealand. Ian visited the Westland site several years later, and he told me that it had been overwhelmed by weeds, and was unlikely to provide any useful data, so it appears to have been abandoned. This leaves us with the Pirongia trial.

The Pirongia trial is on what I would regard as a typical good North Island site for blackwood. It is located on the lower slopes of Mt Pirongia, the soil is volcanic in origin , free draining, and of moderate fertility, gently sloping and reasonably sheltered from the prevailing wind. Rainfall is high, at over 2000 mm per year.
The seedlings were planted in compartments, each representing a district, and each contained 8 copies of 5 seedlots. There were 4 replicates of each, giving a total of 2080 trees.

Ian Nicholas made an early assessment, at which time the NE Tasmania provenance was heading the pack.

At age 7, Toby Stovold and colleagues made a detailed assessment, measuring height and diameter, and grading form for each tree on a nine point scale. The results were as follows, the best trees at the top.

  1. North Inland Tasmania
  2. North Eastern Tasmania
  3. New Zealand. (seed origin probably NW Tasmania)
  4. North West Tasmania
  5. North Tasmania
  6. Otways Victoria
  7. South Tasmania.
  8. Queensland and Chile
  9. Gippsland Victoria
  10. South West Victoria
  11. NSW high
  12. NSW
  13. Central Victoria

The message for us is obvious. The top five provenances are all in Northern Tasmania. Interestingly, our traditional seed source, NW Tasmania, came fourth on the list.

When the trees were at age 10, I could not resist the temptation to carry out some crude measurements of my own, that could be described as quick and dirty. What follows is not science, and I would suggest that anyone with a background in forest science should read no further.

I took the perspective of a guy with a chainsaw heading into a neglected woodlot for some selective culling. I made a subjective assessment of each tree, scoring them on a three point scale:

0 - a hopeless case. You could have done nothing with a tree like this. Pass me the chainsaw.

1 - pretty average. You might have made something decent with it, but it would have needed serious work.

2 - definitely a keeper. With some light form pruning this could have been a classy tree.

Well, you cannot complain you were not warned.

Bearing in mind that I perpetrated this raid on science a few years down the track, the pecking order followed quite closely the previous, and more legitimate 7 year assessment. Again, he top five provenances were all from Northern Tasmania. Top performer was NE Tasmania. I gave a higher ranking for NW Tasmania, now in second place, and a slightly higher place for Gippsland. Best performers in NE Tasmania were Scottsdale ( this name seems to crop up), Fingal, and Swansea, but the numbers are too small to mean much.

To be honest, when I walk through the plots from NE and NW Tasmania and simply glance at them they look very similar. I would be happy to plant either of them. At age 15 all the Northern Tasmanian groups now tower over the others. Considering that they also have lower mortality, it really is no contest.
Some of the provenances have distinctive features. At least two have a flowering time that is different from the rest, suggesting that they may be on the way to forming separate subspecies. One of the 65 provenances was resistant to attack by psyllids, which have been incriminated as a cause of malformation in blackwood. In theory it should have had better form than the others, but in fact it was very branchy, and subsequently had high mortality.

In compliance with the protocol the trees have not been thinned or pruned, so are now showing significant mortality. So the 7 year data will be definitive.

In summary

  • based on Tasmanian and New Zealand trials, the best source for blackwoods for New Zealand is Northern Tasmania.
  • several trials suggest that provenances in NE Tasmania might head off NW Tasmania, but any differences between them are likely to be minor.
  • we can be confident with our traditional sources, around Smithton in NW Tasmania. However someone might like to try some from the NE (perhaps Scottsdale)
  • on the mainland, sources in Southern Victoria (Otways, Gippsland) might be considered, but why bother when Northern Tasmania is better?
  • forget about Northern Victoria, South Australia, NSW, Queensland, and South Africa.

Individual Tree Selection

Within any woodlot or natural forest of blackwoods you can find occasional "plus trees" of superior form and vigour. These are obvious targets for selection for propagation or for setting up a seed orchard. If it works for pines, why not for blackwoods? The answer is that pines tend to breed true, but a mix of anecdote and science suggests that blackwoods do not.

For example, back in 1957 Ib Thulin from FRI went to Smithton in NW Tasmania, and selected from the natural forests 10 trees of superior growth and form, and collected seeds from them. He planted seedlings grown from them on several North Island sites. They did not meet expectations. In fact their performance was average, and no better than unselected seedlings.

There is clearly an environmental effect, but how strong is it? An obvious approach would be to eliminate the genetic influence by taking clones from selected plus trees, and compare them with unselected seedlings. So, what do the clonal studies tell us?

It is surprising, but as far as I can determine, they have not been done, or at least have not been published. Admittedly it is not easy to clone blackwoods. I know, I have tried it. Unless you have access to a tissue culture lab, you have to take root cuttings, and it requires specialised nursery techniques.

But there is a gap in our understanding of blackwood genetics that should be closed before making assumptions derived from different species, and filling seed orchards with trees that might be no better than average. We have some limited experience in New Zealand with blackwood clones, and that is not reassuring..

In 1998 Ian Nicholas and Ian Barton took root cuttings from74 superior trees in the Hunua forest near Auckland. 35 were successfully propagated, and they were planted out north of Auckland. The objective was to establish an orchard of superior trees for later propagation. This was abandoned when it was found that the trees showed wide variation in form and vigour, and had no resemblance to their parents.

In 1997 I planted 90 blackwood clones on my property at Pirongia. They had been taken from a single tree selected for superior form and vigour, and which had high quality colour in core samples. They were evenly spaced, and interplanted with random seedlings. None of the trees were pruned or thinned.

At five years, the clones showed wide variation in form and vigour, and seemed no better than the control seedlings. I introduced Ian Nicholas to the trees, and Ian, Ham Gifford and I took some measurements. We chose at random 32 clones and 40 control trees, and measured their height and diameter. Form was assessed using a nine point scale.

The data gave some surprising figures. Average values were as follows:

  DBH (mm) Height (m) Bole break (m) Form (1-9)
Clones 104.69 6.17 2.27 4.47
Controls 100.49 5.93 2.83 4.75

In summary, there was very little difference between the two groups, the clones showing slightly more vigour. On the other hand, they had slightly poorer form.

Even more surprising, when the Standard Deviation (which measures the range of variation), was calculated, this proved to be less for the controls than the clones for all attributes:


DBH 38 - 175 mm Height 3.3 - 8.6 metres
Bole break 0.5 - 5.2 metres
Form 2-7

If we put all this together, and bearing in mind the obvious limitations of the two clonal trials, I think we should have some reservations about the heritability of vigour and form in blackwood. The non-genetic influence seems overwhelming.

Within a population of blackwoods there are likely to be a few individuals with a better genetic endowment for vigour than the others. However they would be difficult to identify. To test this properly it would require more robust trials, using different clones, and across several sites. They would also need to be tested over several generations. I doubt that this would be justified:

  • apart from the cost and time required, any genetic gains are likely to be modest, and limited to vigour, rather than form.
  • any benefit that might be achieved would be minor compared with the gains that can be made by good silviculture. We know that with seedlings from a good provenance and grown on a suitable site, and with proper management we can consistently produce high quality trees with a rotation of 35 to 40 years.


Vigour in blackwood depends on the provision of adequate light to the growing crown. The key to this is early thinning. A delay in thinning , even by a few years, forces the crowns to compete. This leads to slow diameter growth, and a long rotation.


Malformation in blackwood can be reduced, although not eliminated, by interplanting with a nurse or trainer species. But there is a cost: reduced diameter, slow growth, and a possible adverse effect on wood colour. However you can eliminate malformation and produce consistently straight trees, even when open grown, if they are regularly form pruned.

Malformation in unpruned blackwoods is inescapable. It can result from any random influence that damages the leading shoot tip, such as attack by psyllids. However, there is a deeper influence.
Malformation is embedded in the growth habit of blackwoods, in which periods of extension growth are repeatedly terminated by abortion of the shoot tips and their subsequent replacement. This serves the interests of the tree, allowing it to develop and expand its branches when exposed to light. There is no way we could breed this out of the tree without converting blackwood into a different species.

Does that mean we should abandon hope of genetic gain?

Not necessarily. There is a third feature that has, with a few exceptions , been surprisingly neglected in efforts to breed improved blackwoods. That is the quality of its timber. After all, that is what makes blackwood so distinctive, and ensures its place among the great decorative timbers.
It could be even better if we could resolve a problem encountered by the end users, and that is the variability in colour and density of the timber. The question is, to what extent are colour and density genetically determined?


Environmental influence

In 1975 C.M. Harrison , in South Africa, published a study based on 196 trees, which examined the relationship of colour in blackwood to site. He concluded that there is a linkage, and that the best colour is associated with good soils, the number of rain days during the growing season, and sites that are cold and dry in winter. This study has been frequently referenced since then, but if true in South Africa, does it also apply here?.

More recently, Ian Nicholas in New Zealand, and Gordon Bradbury in Tasmania have independently studied a possible linkage, and both have found no evidence of a relationship between colour and either soil or rainfall.

For example, in a study in NW Tasmania where trees from the same seedlots and established on the same protocol, had been planted at Meunna (wet), and at Virginstow (dry) , there was no difference in colour between them.
Ian Nicholas examined thinnings from 18 year old trees planted on 6 North Island sites, and found no significant difference in colour between them.

If colour is not linked to site, what about a linkage to growth rate? It has been argued that fast growing plantation blackwoods could have inferior colour than slower growing trees in natural forests. Again, both Ian and Gordon have independently examined this, and have come to the same conclusion: there is no relationship between colour and growth rate in blackwood. We can therefore confidently employ silvicultural systems that support early crop rotation.

In a separate study, Ian Nicholas examined wood samples from New Zealand and Tasmania, and found no significant difference in colour between them.

Gordon Bradbury made an unexpected observation in the Meunna trial. The trees that had been interplanted with a nurse species (E. globulus) had not only smaller diameter, as you would expect, but they also had a paler and less interesting colour than those that had been open planted. I have not seen this reported previously, but if could be confirmed on other sites it would have obvious implications.

There have been anecdotal suggestions that the colour becomes darker as the trees age. This does not appear to have been tested in formal studies, but if it were confirmed, it would hardly justify a delay in milling.

Provenance studies have shown that colour can vary between provenances, and have confirmed that there is wide variation between trees in the same provenance. Colour can vary between locations within the same tree, and tends to be darker in tension wood.

Genetic influence.

The limited data that we have suggests that colour in blackwood is more strongly linked to genes than environment. I am aware that anecdotal evidence is not respectable, but anyone who has thinned out a large number of blackwoods will have been impressed by the wide range of colour in trees that are a few metres apart. It is hard to imagine that subtle differences in what appears very uniform soil could have that effect.

The trial data on colour that I have referred to also suggests that environmental influences are not significant, which would by default incriminate genetics as the main influence. The obvious way to sort this out would be through clonal studies that compare clones with random seedlings. As you might expect, these have not been done, or as far as I know published.

With some apprehension, I will refer to a small test, which could be described as a trial run, on my blackwood clones. I cut a large branch off 24 of the clones, and 24 of the seedlings, cut discs, and lined them up. All the branches had heartwood. From previous observations on thinnings , and a published report on A. Koa I am confident that the branch colour matches stem colour.

As expected, colour among the seedlings showed wide variation. Most were in the middle range. Some were pale and unexciting, and one or two were darker and more interesting. In contrast, the colour in the clones was darker. More significantly, they showed a high degree of uniformity.
This can hardly be described as science, but suggests a genetic link to colour that would be worth a formal study. Ian Nicholas and I had planned to set this up on the clones at Pirongia , and we had some funds approved from AMIGO, but we were overtaken by events. This could be activated, and I think would be worth doing.

A full study would need several different clones, tested against seedlings, on different sites. This would provide a definitive answer, and after removing the

control seedlngs could leave a valuable resource for milling, or breeding. Of course that would require you to choose your colour.

Choosing your colour.

Not all blackood timbers are of equal quality. The best of them have a rich colour and texture and could be mistaken for mahogany. Some have attractive reddish elements, others dark, more like black walnut. And there are some logs that are pale and uninteresting. Faced with this multicultural assemblage the miller has two options: to keep an inventory large enough to allow colour matching. Or what is usually done: to make the best of it, lump them together, and present the variations to the customer as a special feature of blackwood. However I am told that if given a choice they would welcome a uniform product , and pay a premium for it.
If you decide to go clonal, what colour should you choose? There is no point in asking the architects. They follow trends, which is currently for a deathly white. You can be sure that whatever is trendy today will be out of fashion in a few years. Don't ask the scientists. They are likely to drag a colorimeter into a furniture store. Ask the customer, who of course is always right.

Also ask the Chinese, who are likely to our main customers for blackwood timber. I am assured they would welcome a regular supply of blackwood timber. I have looked about in one or two furniture stores in China. They have some classy furniture, and the wood looks very much like rain forest mahogany. It probably is rain forest mahogany. However it could easily be replaced by certain grades of blackwood. In the Shanghai Museum there is some fine furniture, simple and modern in style and also resembling mahogany, that dates from the Ming Dynasty. My guess is that something in that colour range would be a safe bet. Or you might aim for a palette with two or three different options.


I will summarise the evidence from studies in New Zealand and Tasmania.

Blackwood has a wide range in density. At the lower level it should be sufficient to protect the furniture from a sustained attack by the toddler armed with her toy hammer. However with an increasing interest in blackwood for flooring, higher density values could be important. And if we could provide blackwood with uniform density, it would be welcomed by the furniture makers, who find that mixing boards with different densities can be a problem.

The evidence suggests a strong genetic influence on density. There is a weaker environmental effect. Whatever that might be, it does not appear to be linked to rainfall.

Fast growth does not compromise density. In fact, in a New Zealand study, faster growing blackwoods had slightly higher density. This links blackwood with the ring-porous species ( in ring porous species, such as oak, fast growth is linked to high density. In diffuse porous species, such as poplar, the opposite is true).

Conveniently, in blackwood there is some evidence of a possible linkage between higher density, and darker and reddish timber. If true, that might provide an opportunity to kill two birds with one stone.


Both growth and form in blackwoods are powerfully influenced by environmental factors.

Some trees within a provenance are likely to be better endowed than others in vigour, although this does not appear to apply to form. However they would be difficult to identify. This would need several generations of testing, and would be likely to produce only modest gains. These would be dwarfed by the benefit of good silvicultural practice.

On the other hand, the wood properties in blackwood, in particular colour, appear to have a strong genetic link. I think this should be the priority in any genetic improvement program, and could offer opportunities in a niche market.


By now, the reader will know that I have pillaged ideas and data from two leading researchers, the late Ian Nicholas in New Zealand, and Dr. Gordon Bradbury in Tasmania.

Over many years I have exchanged ideas with Ian, argued, collaborated , and learnt from him. Sorely missed.

Gordon has done excellent blackwood research in Tasmania. His website, is provocative and well informed on a range of topics, and certainly worth a visit.

6 posts.

Post from Gordon Bradbury on July 8, 2019 at 4:43PM

Hi Ian,

Thanks for a very good update and summary of blackwood research, including some new insights, especially about cloning blackwood.

Have you taken any wood samples from your provenance trial, before time and mortality reduce the trees to compost?


Dr Gordon Bradbury

Post from Ian Brown on July 14, 2019 at 11:00AM

Hi Gordon, there is no current plan to examine wood samples from the provenance trial trees. It would be interesting, but in view of natural mortality the sample numbers would be small. However we will retain representative trees from all districts. Ian

Post from Stuart Swanson on July 17, 2019 at 4:23PM

G'day Ian

I recall an early research paper of Ian Nicholas' that suggested some correlation between bark colour and timber colour (the darker the bark the darker the wood).

I'm sure I have a copy in storage, will have to dig it out to check. Have you noticed any such correlation? From my experience of milling (both hands on and watching others) I suspect a link between dark / rough bark and darker timber colour, how strong a link I'm not sure.

Anyone else with milling experience care to comment?


Stu Swanson

Post from Ian Brown on July 26, 2019 at 11:25AM

Hi Stu, good to hear from you. As far as I know, this has not been formally studied. From memory, Ian Nicholas thought there might be something in it. I am not so sure. If there is  a link it could be useful during a final thinning. It would therefore need to be valid for relatively young trees, I.e. less than 10 years, and from the right provenance. There is a lot of variation between provenances in their bark. For you guys in Tasmania, and for us, that means from the north of Tasmania. Certainly worth looking at during a thinning operation.


Post from Ian Brown on July 26, 2019 at 11:25AM

Hi Stu, good to hear from you. As far as I know, this has not been formally studied. From memory, Ian Nicholas thought there might be something in it. I am not so sure. If there is  a link it could be useful during a final thinning. It would therefore need to be valid for relatively young trees, I.e. less than 10 years, and from the right provenance. There is a lot of variation between provenances in their bark. For you guys in Tasmania, and for us, that means from the north of Tasmania. Certainly worth looking at during a thinning operation.


Post from Stuart Swanson on January 7, 2020 at 5:45PM

Can anyone shed any light on the Blackwood regime trials established by Ian Nicholas? They should be about 36 years old (from memory they were 18 when I inspected them in 2002 with AMIGO).

Disclaimer: Personal views expressed in this blog are those of the writers and do not necessarily represent those of the NZ Farm Forestry Association.

Farm Forestry - Headlines

Article archive »