Breeding Better Radiata Pine Trees
Leon Pederick (bio)
This article is one of many of Leon's contributions to this site.
The Forests Commission commenced a breeding program with radiata pine in 1958, but before describing the work it is useful to consider why it was started.
In the earlier years of pine plantation establishment in Victoria, seed for new plantings was obtained from cones collected from trees on the ground after final fellings. There was no selection for the type of tree from which the cones were taken. Sometimes, even cones were collected from trees felled as thinnings. Gradually, from around the world, reports were received that the characteristics of trees could be improved by more careful selection of the seed source. Indeed, there seemed no reason why tree crops could not be improved by using the plant breeding techniques so successful in many agricultural crops.
However, this appeared difficult for several reasons. Firstly, there was the problem of controlling the pollination on very tall and large plants. Then there was the longevity of trees, particularly the generation interval, the time taken from planting to seed production. The amount of seed needed for forestry operations could be very large.
By the mid 1950s techniques had been developed to overcome some of these problems. Firstly, the height of tall trees was no longer a limitation when branch shoots taken from the upper crown were grafted on to seedlings. The growth of the grafted plant retained the physiological age of the scion, and flowering could occur within a year or two. It was even possible for a worker to pollinate pine flowers while on a short ladder, or sometimes even when standing on the ground. The idea of a seed orchard led to production of seed in large quantities. An orchard was composed of grafts of selected trees, in an area remote from other stands of the same species, to ensure that only the pollen from the grafted trees would fertilise the flowers. Hence, all seeds produced in the orchard would be the product of a cross between two selected trees.
In the late 1950s when reports of good initial results were received from New Zealand about progress with breeding radiata pine there, the Commission's Chairman, Alf Lawrence, decided that his department should also undertake breeding. A staff member in the Research Branch, the writer of this paper, was given responsibility.
The first step was to find a number of the most outstanding trees to use as the initial breeding stock. A search through the Victorian State plantations was undertaken. The desired traits, in order of importance, were fast growth rate, straightness of stem, slenderness of branching (relative to growth rate), and a preference for above average corewood density was added in the later stages of the program. (Fig. 1)
The objective was to increase the yield of stronger sawn timber. The twenty best trees found, termed 'plus trees', were selected to form the initial breeding stock. An aluminium tree climbing ladder was obtained, composed of six interlocking ten-foot sections, to enable a worker to reach a height where the branches were sound enough to support him and allow him to climb higher. (Fig. 2) The plus trees were climbed during the winters of 1959 and 1960, and branch tips (scions) were collected from each. A glasshouse had been constructed in the grounds of the School of Forestry at Creswick to be used for the propagation work. The scions were then grafted on to the tips of potted seedlings held in the glasshouse. Growth of the scions commenced several weeks later. (Fig. 3) In subsequent years seedling stocks were grown in large pine veneer tubes, to prevent root coiling and to hold more plants on the same bench space.
Ten grafts of each plus tree were later planted in a designated area nearby, called a 'scion garden', because in the following years shoots could be taken from them and used as scions to increase the number of grafts of each plus tree. There was no need to revisit and climb the plus trees, most of which were located in distant plantations. Several years later when the planted grafts had grown and were producing both pollen and female flowers, controlled pollination was undertaken to produce seeds with full pedigree for planting in growth trials. The name of the area was then changed to 'breeding arboretum'.
When sufficient grafts of the required number of plus trees had been produced, the first unit of seed orchard was planted in 1961. It was located on a former cleared farm within an extensive native forest at Korweinguboora. (Fig. 4) Initially 16 clones (a clone being the propagules from a single plant) were used. This number was increased in later plantings to 20 and eventually 50, as more plus trees were added to the program, in order to ensure that additional gains could be made in the future by selecting the best subset of these for future breeding, while still retaining a suitably sized gene pool.
Later plantings of seed orchards were established at East Trentham (Campaspe) and Lal Lal. Initially the orchard at Korweinguboora flowered well and several good annual crops of cones were harvested, but later the yields declined, perhaps due to coldness of the site at flowering time. Although less fertile and drier, the Lal Lal site proved much more reliable for cone production and the Korweinguboora site was eventually phased out.
After the first few years, as the trees grew taller, the workers needed elevating equipment to harvest the cones. Later, the trees grew too tall to harvest, and the practice of pollarding was introduced, whereby the upper trunk was cut off in order to promote a bushier growth of lower side branches which then flowered and produced more cones. (Fig. 5) Pollarded trees would become temporarily out of production for several years, so only a small proportion of an orchard would be treated in any one year to ensure that enough trees were in production at all times.
The first general flowering at Korweinguboora occurred in about 1965 and a small crop of cones was collected two years later. Following extraction from the cones the first orchard seed was available for sowing in the nursery in 1968, with the seedlings ready for planting in 1969. Seed yields increased annually thereafter until the Commission's full requirement was achieved by the late 1970s.
The orchards had been established with plants selected on the basis of their superior appearance, but there was no certainty that they were also genetically superior. Trial plantings were needed to evaluate the relative genetic value of seed orchard seed, and eventually the genetic ranking of each of the clones in the orchards. When a trial with plants of Lal Lal orchard seed origin and those of routine unimproved origin were grown in replicated plots in the Warrenbayne plantation, and assessed at 10 years of age, the trees of orchard origin were on average 9% larger than the routine ones. They were also a bit straighter but it was not possible to put a financial value on this.
Trials (termed progeny trials) to test the genetic value of each individual clone as a parent were planted as seed became available. The first progeny trials were planted with seedlings grown from open-pollinated seed (collected from the plus trees at the time of the original climb). These were only partially successful because some of the trees yielded few cones and insufficient seed for satisfactory comparisons. Much better trials were established later using seed collected from the individual clones in the first unit of the seed orchard at Korweinguboora. Each seedlot from this source were all pollinated by the same pollen cloud generated in the orchard, so the differences in average growth, form, etc., among the families could be attributed to the maternal genes of the parent. An equivalent set of trials was undertaken later with seed derived from controlled-pollination with a mixture of pollens, though the comparisons were only valid between families derived from the same pollen mix.
From a number of such trials the genetic value of the 50 clones in the Lal Lal orchard was rated. The 10 clones judged to yield the poorest offspring were removed, and as a result it was estimated that the average growth rate of plants from the remaining 40 clones would be about 2-3% greater and additional to the 9% mentioned above.
There was some uncertainty about how adaptable individual radiata pine trees would be to different sites. Indeed, there was some opinion that certain trees would grow relatively better than others on some sites. This was regarded to be like “horses for courses” by some observers. Evidence to test this possibility was obtained from a large set of trials planted at four locations with plants derived from separate clones in the Korweinguboora orchard mentioned above. The trial sites ranged from deep sandy soils at Rennick (near Mt Gambier), granitic soil at Warrenbayne (Strathbogie Ranges), a red clay soil at Narbethong, and clay loam over sandstone near Myrtleford, with climatic differences as well. These sites were distributed from the far west to the north-east of the State. A measurement of these trials at age 10, and partially at 15 years, indicated that any preference of families for relatively better growth on certain sites was of minor importance. What mattered was that some families were best on all sites and others less so. This was an important result because it confirmed that one breeding program could be applied over the whole State. Separate programs for local areas would have been unworkable. It also meant that superior genetic material could be exchanged between breeders in different States and countries. The Victorian program had already benefitted from obtaining some superior material from the New Zealand program. Exchanges also occurred with programs in other Australian States. The reason behind these exchanges was to enlarge the genetic base of improved material so that a second round of selection could be undertaken, to produce further gains without reducing the size of the gene pool.
Having achieved productive orchards in what could be termed the first generation of breeding, attention turned to the identification of improved clones to be used in a second generation which would yield still higher gains. This work was started by selecting the most outstanding tree in each of the best families in the progeny trials. Such trees could be relied upon confidently to be genetically superior, and were termed 'second generation selections'. (Fig. 6) They were grafted and planted in the breeding arboretum. Some were even planted in new orchard extensions, though only when they were unrelated to clones already in the orchard, so as to avoid introducing inbreeding.
It became important to undertake a large program of controlled pollination in order to be able to plant out a large number of families of known bi-parental origin. Second generation selections could be made in the best of these families for inclusion in a second generation seed orchard, using the full-pedigree information to achieve a minimum of relatedness, and thus inbreeding depression. (It should be noted that some inbreeding cannot be avoided because of the chance that some of a tree's windborne pollen may fertilise its own flowers). Accordingly, a continuing program of controlled-pollination was undertaken each spring in the breeding arboretum at Creswick. (Fig.7)
There were two objectives. Firstly, to produce a large number of unrelated families in which to make future selections in the best of them, and secondly to make several crosses with each clone in order to estimate an average breeding value for each, as well as to provide more selection opportunities. The work continued over many years because the number of crosses required was large, some clones did not have flowers at a suitable height when desired, and resources to do the work were limited. However,when the cones matured two years later and the seed extracted, groups of seedlots were sown in the nursery and planted in field trials in replicated designs.
Progress in the breeding plan was sustained but slow, limited by the small number of staff available. It had reached a level of about one and a half generations of breeding when an entirely new development in radiata pine nursery practice occurred. This was the ability to propagate cuttings en masse. Such a technique had been developed in New Zealand. It involved the repeated trimming of young seedlings to produce a small rounded bush (stool plant), the cuttings from which retained their physiological juvenility, and growing them in a peaty nursery soil. Trials to grow cuttings in existing Victorian nurseries failed until, in the mid-1980s, a patch of peaty soil was found at Gelliondale, near Yarram in South Gippsland.
A nursery was soon established at Gelliondale. In the first years the great majority of plants produced were seedlings grown from seed orchard seed. Stool plants were established from seeds of superior genetic value produced in the breeding program. Controlled pollination using pollen and female flowers on trees, in a small orchard established next to the nursery, were undertaken to provide a continuing source of superior seed to use for replacement of the stool plants when needed. The annual production of cuttings gradually increased and that of seedlings diminished. Cuttings have some advantages over seedlings. They have stronger root systems and establish better when planted in the field. They grow straighter in the first few years, with finer branches. They grow just as fast.
With greater emphasis on the use of cuttings a smaller area of seed orchard became necessary to provide the seed for seedlings. Instead, a continuing program of controlled-pollination between the best parents was needed to provide high performing seeds with which to grow stool plants for the cuttings program.
When undertaking long term forestry programs it must be expected that unforeseen events may occur which affect the outcome or cause setbacks. It certainly happened in this breeding program. A large seed orchard development in the East Trentham (Campaspe) area was destroyed by the Ash Wednesday bushfires of 1983. The orchard had just reached the stage of producing its first cone crop, and had to be completely replanted, although the opportunity was taken to include a better selection of clones. In 1977, a fire from the north moved rapidly through the outskirts of Creswick and destroyed the breeding arboretum. Two years of controlled-pollination work was lost in the form of developing cones on the trees. By the time the replanted arboretum had grown to flowering age at least five years of breeding potential had been lost. Elsewhere, a 2ha extension of the Lal Lal orchard had just been planted when the local supervising foreman let sheep in to the fenced area in order to eat the grass between the planted grafts as a fire control measure. Little did he realise that the sheep would prefer to eat the fresh pine growth above the graft union, so further grafts had to be propagated to replace the grazed ones, involving a loss of 2 years and much effort directed towards the development of this unit of orchard.
The Commission's breeding program gradually ended when the Victorian Government decided to exit pine growing, firstly by transferring management to a Victorian Plantations Corporation (VPC) in 1992, and their complete privatisation by sale to HVP Plantations in 1998.
Pine breeding had been undertaken in some other States, even since the late 1950s. However, at Mount Gambier a group called the Southern Tree Breeding Association (STBA) formed as a co-operative between the South Australian Woods and Forests Department and private forestry companies in that State, and later added some interstate members. Victoria was not a member during the years of work described above, although it co-operated extensively with exchange of technical information. STBA employed highly qualified staff and became the specialised breeding centre for radiata pine in Australia. The VPC joined STBA as a member in 1993 as did HVP in 1998. The STBA was then able to take over the Victorian breeding assets and integrate them into its own program. The writer of this report retired in 1993 and is unable to comment on progress since that date.
How much improvement has been obtained by breeding? A gain of 9% in growth rate at age 10 years was reported above, with the expectation of a further 2% when some poorer clones were removed from the orchard. However, the results of breeding only become available many years after the actions are made, so it is likely that much higher gains have come from more recent work. Improvements in stem straightness are difficult to evaluate in monetary terms. Straighter stems yield more sawn timber, with less degrade due to compression wood. The selection for trees with more slender branches, relative to tree vigour, would result in smaller knots in sawn timber, hence higher grades and value. Cuttings of the best genetic material would certainly yield higher gains.
Some writers have estimated that gains in value of plantations may now attain 30%. When considered to apply over the national pine estate of half a million hectares, improvement of this magnitude becomes very significant. The Forest Commission's program described above made a useful contribution to the development of this improved national resource.