The first generation of selection and breeding is nearly complete for Sitka spruce, Scots pine, Corsican pine, and European and Japanese larch to give hybrid larch. Breeding and production populations either exist already or will be created over the next two to three years. Only Douglas-fir, a small proportion of Sitka spruce, and a very small number of hybrid larch progeny tests remain to be assessed in order to estimate the breeding values of selected plus trees for a number of different traits. By 2003, it is expected that breeding and production populations will exist for all the major conifer species. The table below gives the existing and predicted genetic gains by species from production populations created by selection within the first generation.
Currently, Sitka spruce is the only species for which there is a strategy to move into a second generation of breeding through selection and testing. A breeding population of 240 clones has already been sub-divided into 4 sub-lines of 60 clones of equal mean genetic value in terms of overall multi-trait index score. It is expected that two further sub-lines (an additional 120 clones) will be added to the breeding population following re-selection of original plus trees represented in the latter half of progeny testing.
The breeding population is divided into the sub-line structure to restrict the extent of inbreeding. As a degree of inbreeding occurs, equal numbers of clones will be taken from each sub-line to create a production population of totally unrelated trees. Once inbreeding within a sub-line is considered acute, sub-lines can be combined to increase the variation within each sub-line. Naturally, the extent of the inbreeding problem is dependent on the number of generations completed and the balance exercised by the breeder between short-term genetic gain, and long-term genetic variation. Although it is necessary to plan for all the possibilities it is impossible to say how many generations will take place in the breeding of Sitka spruce and therefore impossible to say when, if ever, inbreeding depression will become a problem.
Predicted genetic gain available now and in the near future for species subjected to the cycle of selection, testing and re-selection required to form breeding and production populations
|Family mixtures – general||15-22%||5-15%||0%|
|Family mixtures – high density||10-15%||15%||9%|
|Family mixtures – high straightness||15%||20%||0%|
Note: Improved seed are collected from clonal seed orchards containing tested parent clones unless otherwise stated. Gains are relative to:
Full-sib crosses will be made within each sub-line according to a pattern which is referred to as assortative mating. This means that clones with higher breeding values will be involved in a greater number of crosses. In sub-line 1 the 60 were divided into a top, middle and bottom third. Each clone within each of those thirds was used in a full-sib cross either 5, 4 or 3 times respectively. For the remaining sub-lines division into upper and lower halves with 4 and 3 full-sib crosses per clone respectively, are to be employed. This follows the experience of practical restrictions when completing the crossing in sub-line 1 and a concern that error attached to breeding value estimates may be greater than at first thought.
The crossing programme for sub-line 1 has recently been completed. One hundred full-sib families will be planted over 3 sites in single plant plots in spring 1999. Meanwhile controlled pollination work continues within sub-lines 2, 3 and 4 and will commence in sub-lines 5 and 6 once component clones have been identified. It is intended to establish all the resulting full-sib crosses to the field by 2008.
For the first time in the breeding of Sitka spruce, there will be the opportunity of exploiting an element of non-additive genetic variance when selecting outstanding full-sib families. It remains to be seen how observed family performances will vary relative to predictions based on breeding values estimated in the first generation. Once outstanding families have been identified, they will be incorporated into the family mixture production populations for multiplication by vegetative propagation.
One or two of the best individuals (in terms of multiple traits selection indices) will be selected from possibly the top 50% of families to form the next breeding population, and so the cycle will re-commence once forward selections have been brought to flower. Pedigree information of forward selections will become imperative to prevent the mating of related individuals for as long as possible.
The field tests from which the second generation will be selected will involve approximately 500 full-sib families or 30 hectares of field test sites. The best families for height will be determined 5 years from planting, and final selections for all traits will be carried out with the aid of statistical techniques developed by animal breeders (eg Best Limited Unbiased Predictor; BLUP), around 9 or 10 years from planting. The restriction to faster progress is the age at which forward selections can be made to flower. At the moment this is at least 15-20 years from seed. The rate of genetic gain could be increased if this was reduced to just nine years.
Forest managers want trees which are straighter, finer branched, faster growing and produce stronger timber. Since the objective of breeding is to produce material for the construction market, we need to get the trees to a suitable size and quality as quickly as possible so that returns on investment are maximised. It is now clear that the negative genetic correlation between growth rate (height or diameter) and wood density in Sitka spruce is probably the worst reported amongst conifers for which there are active breeding programmes. If wood density is to be maintained, growth rate cannot be maximised. Tree Breeders have been trying to balance this conundrum and, in order to answer critics and provide forest managers with a choice of improved seed sources, it is proposed that by careful selection within the General Breeding Population, a number of alternative production populations can be created to which differing weights for gain in diameter, density and stem straightness are given.
Most forest managers will be content with the General Production Population (GPP) which is intended to maximise growth rate and stem straightness whilst preventing a fall in wood density. The High Density (HDPP) and High Straightness (HSPP) Production Populations have greater weighting on straightness and wood density respectively with a possible fall in diameter (Lee, 1999).
It remains to be seen what interest will be generated by these novel production populations. If the uptake is poor, they will be discontinued; if the uptake is good then separate breeding populations may develop.
The amount of effort that will be devoted in future to species other than Sitka spruce, is uncertain. An investigation is pending into the cost and benefits of moving into a second generation of Corsican and Scots pine breeding. It is also possible that demand will increase for a separate production or breeding population of Scots pine clones well adapted to the higher rainfall areas of West Scotland.
Progeny tests have ably demonstrated the potential gains of re-selecting outstanding hybrid larch families relative to material from registered seed stands or directly imported Japanese larch. If rooting techniques could be improved for hybrid larch in general, or alternatively good rooting hybrid larch families can be selected, then there is likely to be increased activity in the crossing of selected Japanese and European larch trees.
Some clonal tests of a research nature have been established. The most interesting and recent of the limited number of Sitka spruce clonal tests was planted over 3 sites in 1990. The clones were random selections within full-sib families between parents which had then been thought suitable for the breeding population. More data have yet to be collected, but 4-year height and stem straightness at just one of these sites, Newcastleton in the Scottish Borders, found one full-sib family to be growing 50% faster than the QCI control and the best clone was growing 77% faster. It also seemed that the fastest growing clones were the straightest (Lee, 1997c).
However, clonal forestry is not yet being routinely practised within any of the breeding programmes. There remain technical problems regarding the physiological ageing which prevents the production of commercial quantities of selected clones either by tissue culture or cuttings, once they have been identified in the field (John, 1992). When the tree is in the juvenile phase, cuttings are easy to root and they grow very similarly to seedlings. However, they are untested. When the trees are tested to about 8 years old, they have entered the more mature phase of growth, cuttings are more difficult to root and the rooted plants tend to grow plagiotropically.
If tested clones are to be used for afforestation, techniques must be developed that will keep a sample of the genotype in the juvenile form whilst the genotype is tested. A number of possibilities exist to accomplish this. The genotype might be kept as a hedge whilst the clone is under field test. This technique might not be suitable for Sitka spruce as it tends to undergo the maturation process even when hedged. Techniques might be developed that reverse all the changes that occur when the plant matures, turning it from the mature to the juvenile phase. In vitro techniques are being developed which hopefully will accomplish this. Perhaps the most promising technique is cryopreservation. A sample of the clone would be kept in the juvenile form, frozen to the temperature of liquid nitrogen whilst the clone is field tested. The tested and superior clones could then be regenerated in the juvenile form from the cryopreserved tissues.
The traditional tree breeders are working more and more with biochemists and molecular biologists. It is clear that great advances are being made in understanding the working of the conifer genome at the molecular level. Some of the knowledge gained can be applied to the Sitka spruce breeding programme.
The use of biochemical markers to indicate the location of genes which contribute to the expression of economic traits under selection (Quantitative Trait Loci; QTL) potentially holds a great deal of promise, but a lot of further effort and resources need to be invested to determine whether the theoretical approach can be translated into practice. If markers can be found that indicate the location of the main genes which contribute to wood density or height then it will be possible to select the best individuals within each family by extracting embryos from seed in the laboratory thereby removing the need for costly progeny testing. However, due to the high negative genetic correlation between height (and diameter) and wood density, there seems little point in finding QTL for one trait without the other. Commercial exploitation of such selections would also depend on a solution to some of the problems of clonal forestry and progress to a further generation of breeding would also still have to wait until the individuals selected by such methods were old enough to flower.
Another possible area of future activity is that of genetically modified organisms – the introduction of genes from unrelated plants or animals into the genome of Sitka spruce. Ideas of the potential of this technology abound, for example, the introduction of the nitrogen fixing genes from alder, frost resistance genes from arctic fish, timber durability genes from red or yellow cedar etc. There are not only huge technological problems to be overcome before this is possible, but ethical questions also have to be faced. A recent meeting on this topic held in 1998 at NRS between interested parties concluded that only if the future of Sitka spruce was under threat would the public allow genetically transformed plants to be established in British Forests.
Tree Breeding in Britain has now reached the point at which most of the objectives set down by the Research Advisory Committee 50 years ago have been realised. However, science has progressed over this time and technologies now exist which were not even dreamed of in the early pioneering days.
In this record of the first half-century, an attempt has been made to give a history of the methodology used and the achievements gained. The initial process was one of learning techniques which, whilst already established in other countries, remained in the early stages of development. As competence grew, high and rigorous standards in the application of breeding methods were established. In addition there was a constant process of re-evaluation of procedures with the clear aim of modifying them in the interests of minimising both the time and cost of operations whilst maximising genetic gain. There is no doubt that this evolution will continue into the future and it is relevant to consider what the main changes are likely to be.
Whilst in the future the aims of tree breeding may remain the same, some tools will be different and inevitably more sophisticated. Tree breeders will also have to accept changes in what is demanded of them through the evolving needs of the industry and from other factors such as climate change. But tree breeders will have to convince the forestry industry to invest in the means of realising the continuing and varied gains which are possible. These are the challenges which Forest Research is already taking up as it enters its second fifty years.
These pages review the work performed by the Forestry Commission and Forest Research on tree improvement following the 50th anniversary of its establishment which passed in 1998. The genetic background describes the scientific procedures of tree breeding and the technical terms used in the remaining pages. All species are referred to by their common name in English.