Forestry Commission logo

Management of the sitka spruce breeding and production populations

Commissioned Report by Steve J Lee
August 2001


There are currently 240 plus trees in the Sitka spruce breeding population reselected following the first cycle of selection and testing. These reselections will be involved in a programme of controlled crosses to advance the Sitka spruce breeding strategy into the next generation. The very best of the breeding population in any given generation form the production populations which generate the future planting stock. Grafted replicates are taken of all reselected trees to ensure they are not lost from either the breeding or production populations. Grafts are stored in two field-based clone banks where they are managed for the production of flowers.

Successful management of breeding and production populations is measured by the frequency of flowering, the rate at which the crossing programme is completed and continuing increases in predicted gains from family mixtures. Major improvements in efficiency are envisaged as management of grafted replicates transfers from field-based clone banks to high facility polyhouses.


Following the first cycle of Sitka spruce plus tree selection and testing, 240 of the best plus trees thought to be of Queen Charlotte Islands origin demonstrating superior growth rate, stem straightness and wood density have been reselected to form the Sitka spruce breeding population. This number will be increased slightly as the last progeny tests to be planted are finally analysed, leading to the identification of further superior parental plus trees. Reselected trees will be involved in a series of controlled crosses designed to advance the genetic quality in economic terms of Sitka spruce in Britain (Lee, 2001 a).

The very best of the breeding population can be brought together to inter-mate to produce the seed which constitutes the future planting stock. This further sub-population of plus trees is referred to as a production population. A production population represents the best available genetic gain within a given generation for the traits under selection at the time of composition.

Managing a breeding population

Trees in the forest have a finite life which can be extended by taking shoots known as scions from the top of the tree and grafting them onto juvenile rootstocks, raised from seed to around 80 cm height in just one year. In this way several grafted replicates of a selected tree can be made and stored in a central area known as a clone bank. A clone bank therefore is a collection or archive of selected material; it ensures that genetic material is not lost. Tree breeders retain this genetic material primarily in order to carry out controlled pollinations, but also as a source of further scions to create more grafted replicates which can be planted elsewhere.

When the Sitka spruce breeding programme commenced in 1963 it was common practise to make grafted replicates of all plus trees regardless of whether they were flowering at the time of selection. These grafted replicates were planted in the first Sitka spruce clone bank located in Wauchope Forest (Scottish Borders) which quickly filled up with representatives of the early plus tree selection.

To accommodate all original selections, planting commenced in a second clone bank in 1971 at Ledmore, north of Perth (Tayside). At this time, grafting tended to be restricted to those plus tree which were showing some promise based on early progeny test results although additional grafting would be made of selections felled in the forest prior to the completion of genetic evaluation. More importantly, all clones re-selected for the breeding population based on genetic test data were established at Ledmore which is now the main centre for controlled pollination work each spring.

Why graft?

Shoots taken from the top of a mature tree retain their physiological maturity when grafted onto a juvenile rootstock. About 3 years after grafting, with good management, grafted replicates can be encouraged to flower. The advantage is that the flowers are on trees between 1 and 2 m tall planted on even ground, adjacent to copies of many other similar selections rather than at the top of a lone 20 m plus tree in a remote forest. Controlled pollination, the essence of progressive tree breeding, becomes a cost-efficient possibility.

Site selection

Flowering, and not vegetative growth, is required of grafted replicates growing in a clone bank. Site selection is therefore important. The trees need to be placed under sufficient drought stress to encourage regular flowering, but not so much stress that they die. Ledmore is a well-selected site and regularly gives acceptable numbers of flowers. Wauchope however, is too wet and frosty. Grafts planted there have too much vegetative growth and lose too many flowers each year to late spring frosts. Wauchope is being phased out in favour of Ledmore.

In case of disaster, it is wise to have grafted replicates of each selected tree present at more than one site. Ledmore will be retained in future, but a new site is needed to replace Wauchope and this has yet to be identified.

Location within the site

A number of grafted replicates of a clone are planted adjacent to each other in the clone bank. Spacing is 2m within the rows and 6m between the rows. Rows run in an east to west direction enabling development of a large crown on the warmer, south-facing side where the flowers tend to develop. Detailed maps are kept of the locations of grafted replicates of each selected tree within the clone bank which in turn is stored on a central database.

Tending the graft

Grafted replicates grow in the clone banks just like seedlings would, although they often appear to have a narrower crown consistent with the location and physiological age of the original scion.

Male and female flowers tend to occur towards the top of the tree. To keep the flowers within the manageable reach of short ladders, the tops of grafts are cut out at around 1.5 to 2.5 m on a three-year cycle. Grafts which have escaped topping operations have to be visited on mobile elevated platforms.

The life of a grafted replicate is around 30 years. After this time its fecundity tends to fall off and the complete tree is felled to make way for a new grafted replicate consisting, as before, of a physiological mature scion on a juvenile rootstock.

Encouraging flowering

Even with good site selection, grafted replicates do not flower every year. Good flowering years occur about once every 4 or 5 years. Good flowering tends to follow a warm, dry July and August when the tree forms more flower buds in preference to vegetative shoots for the following growing season. Flowers appear on the tree in late April to early May.

When flowering does take place there is great variation in the number of flowers (male and female) which a clone may produce; some selections are very precocious while others never give flowers. There is further variation between selections in the number of seeds per cone from close to zero up to 80 with an average of around 30. If the controlled crossing objective is to involve a particular selection in 5 different crosses requiring 400 seed per cross, the number of flowers required could vary from 25 to 2,000. As the precocity of a selection is unknown prior to commencing pollination work, breeders pollinate 50 flowers per cross, and make up any seed shortfall in subsequent years.

In order to advance the Sitka spruce breeding strategy (Lee, 2001a) as quickly as possible, trees need to be encouraged to flower. Work by Philipson (1983) and Philipson and Fletcher (1990) demonstrated that induction of drought stress, stem and shoot girdling and stem injection with the growth hormones gibberellin A4/7 all influence flowering to a certain extent. These techniques are regularly applied by breeders in the early summer preceeding the spring in which pollination will be carried out.

The effects of the weather

As well as the weather having an impact on the number of flower buds laid down by the tree in summer, it can also influence the success of a pollination season the following spring. Male and female flowers are very tender and are easily killed by late spring frosts. In addition, the clear-plastic isolation tubes placed around a developing female flower to prevent entry by background pollen are often blown off in spring gales damaging the female flowers beyond further use (Lee, 2001c).

Managing production populations

Sitka spruce production populations take the form of tested clonal seed orchards or family mixtures.

Seed Orchard management

A tested clonal seed orchard is a semi-permanent feature lasting for around 30 years. It consists of 20 to 40 grafted replicates from 40 reselected plus trees; 800 to 1600 grafts in all. There are 9 such seed orchards in Britain. See Lee (1999) for further information on orchard location and Lee (2001b) for details of revised genetic gains from these orchards. Grafted-ramets are allowed to inter-mate in a random fashion. The open-pollinated seed may go into storage or alternatively directly to commercial nurseries.

The management of a seed orchard is similar to the management of a clone bank. Flowers, not vegetative growth, are required as frequently as possible and on as many of the constituent clones as possible. Flower-induction treatment is essential to increase the frequency of good flowering years. Good management also involves periodic removal of the tops of grafts to ease harvesting, and trimming back of root-stock growth to prevent any flowers developing on rootstock shoots becoming confused with flowers which have developed from the grafted material.

Family mixture management

Family mixtures consist of seed mixed together following controlled pollination in a clone bank. Female flowers from 20 out of the top 40 reselections within the breeding population are pollinated with a mixture (polymix) of equally represented pollens from the other 20 out of the top 40 reselections. Equal numbers of the seed harvested from each of the reselections acting as female parents are mixed together and the resulting seed is sold to commerce where it is raised as stock plants for use in vegetative propagation.

Three types of family mixture are currently available i. General Production Population (GPP) which aims to maximise growth rate and stem straightness without a fall in wood density; ii. High Density Production Population (HDPP) which aims to increase wood density gains whilst accepting falls in growth rate relative to GPP; and iii. High Straightness Production Population which aims to maximise stem straightness again with possible falls in growth rate (Lee 1999).

Family mixtures are not permanent; different seed lots from store can be mixed together in different years. The annual objective is to improve the average genetic quality of successive family mixtures. One of the major advantages of family mixtures over seed orchards is that whereas it may take 10 to 14 years from planning a seed orchard to first harvesting seed it only takes 1 year with a family mixture provided flowering grafts are available. The genetic gains from a family mixture are therefore 9 to 13 years in advance of that available from the most recent seed orchard.

In order to advance genetic gains from a continuing family mixture programme, best-practise clone bank management is required. There needs to be the right number of healthy grafts from the best reselected plus tree, at an optimum height for controlled pollinations and with a large number of female flowers for isolation or male flowers from which pollen can be collected. There also needs to be good records of past seeds per cone, pollen viability, seed in store by mother trees and pollen composition of past polymixes to make future programmes as efficient as possible. See Lee (2001c) for details of the work required for a successful pollination season.

Moving the breeding and production populations inside

By establishing potted grafted replicates of reselected plus trees inside large high-facility polytunnels, breeders can remove many of the environmental variables effecting field-based pollination. Drought can be more easily induced in the summer, temperatures can be controlled throughout the year, and late spring frosts and high winds can be avoided. It is envisaged that eventually all pollinations will take place in such controlled environments and field-based clone banks will merely be a source of scion material. It is anticipated that the more intensive management which breeders will be able to impose may give substantially more flowers on selections which are otherwise reluctant to produce male or female flowers. Plus trees with high breeding-values hitherto omitted from family mixtures because of their poor levels of flowering may be included for the first time.

Two high-facility polytunnels have been constructed and are now being managed at the Northern Research Station. Each contains around 100 grafted-ramets of representatives from the production populations. It is intended that controlled pollinations will take place within these polytunnels in spring 2002/03. Management will be intensive. Grafts will be subject to flower induction techniques such as drought stress over an eight year life after which they are replaced with freshly grafted replicates.


The Sitka spruce breeding and production populations are highly selected populations which are preserved by planting grafted replicates in clone banks at two locations.

Successful management of the breeding and production populations involves the maintenance of sufficient numbers of healthy grafts to complete the proposed crossing programme, frequent harvests of seed from seed orchards and increasing genetic gains through family mixtures.

Good site selection for clone banks and seed orchards is important in inducing regular flowering. Flower induction treatments involving drought stress, girdling and stem injection-of gibberellin can also induce flowering.

The weather has an important impact on management of grafted replicates. The transfer of grafted replicates from field-based clone banks to high facility flowering halls is currently underway and should improve management efficiency.


LEE, S.J. (1999). Predicted genetic gains from Sitka spruce production populations. Forestry Commission Information Note 26. Forestry Commission, Edinburgh.

LEE, S.J. (2001a). Selection of parents for the Sitka spruce breeding population in Britain and the strategy for the next cycle of selection and breeding. Forestry 2, 129-143.

LEE, S.J. (2001b). Revised predicted gains from Sitka spruce seed orchards. Forestry and British Timber, March 2001, p6.

LEE, S.J. (2001c). Marriages made in heaven? Forestry and British Timber (in press).

PHILIPSON, J.J. (1983). The role of gibberellin A4/7, heat and drought in the induction of flowering in Sitka spruce. Journal of Experimental Biology 34, 291-302.

PHILIPSON, J.J. and FLECHER, A.M. (1990). Implications of cone induction techniques for breeding strategies and the production of improved seed. Forest Tree Improvement 23, 69-80.