A.E. SQUILLACEA.E. SQUILLACE is principal plant geneticist, Southeastern Forest Experiment Station, Naval Stores and Timber Production Laboratory, Forest Service, U.S. Department of Agriculture, Olustee, Florida, United States.
ALTHOUGH FOREST tree improvement is relatively new, development of techniques has advanced to the point where almost any forestry organization can produce improved seed or vegetative propagules for forest planting. Many methods are now available for obtaining or producing improved stock. Some of these are costly while others are not. A forestry enterprise can choose a procedure that is suited to its resources and needs, with reasonable assurance of obtaining a worthwhile degree of genetic gain.
This paper discusses:
1. considerations for choosing a tree improvement approach;
2. techniques and guidelines for developing a tree improvement programme;
3. means for starting action programmes.
The discussion is directed mainly toward organizations that are not yet involved in tree improvement work and that have limited facilities for a practical tree improvement programme. For this reason emphasis will be on simpler, less costly methods and on short-term aspects.
Before starting a tree improvement programme, the planner should obtain answers to a number of questions which will affect his decisions. Some of these are:
1. Needs of the organization. What are the major forest products to be marketed now or in the future?2. Species available. Are native species suited to the needs of the organization? Would exotics be more suitable ?
3. Biological characters of the species chosen. Does appreciable racial variation exist ? Are its native habitats highly variable? Does the species contain enough individual tree variation to warrant selection? Is vegetative propagation easy ?
4. Nature of the traits of interest. What traits are of major importance? Can they be evaluated at early ages? Are they strongly or weakly inherited? Is there a strong correlation between traits?
5. Magnitude of gains needed. Is a high genetic gain needed in order to obtain a market for the product?
6. Urgency of the need for seed and plants. How much planting is done? Is the need for seed and plants urgent ?
7. Availability of research information and techniques. Are results of species trials and provenance tests available ? Are vegetative propagation and controlled pollination techniques available ?
8. Availability of funds, facilities and personnel. How much money and what facilities are available for the programme? Can commitments be made for long-term programmes? Are personnel available for the work? If they need training, where and when can it be obtained ?
A specific tree improvement programme cannot be recommended for all situations. Special papers by Burley (1969), Steenackers (1969), and Thulin (1969) presented at the Second World Consultation on Tree Breeding make recommendations for certain species or groups of species, and one by Greathouse (1969) discusses some general principles. As indicated earlier, the factors affecting the choice of a programme are numerous. A planner must therefore develop his own specific programme, seeking competent advice to the extent possible.
Some general guidelines will be applicable in many situations. General considerations for each phase of activity are given first, followed by guidelines.
PLANTING ZONES
A single strain of trees is not likely to be superior in all portions of a region where climate, topography, and other environmental factors vary appreciably. That is, genotype X environment interaction is likely to occur. For this reason, the breeder should delineate planting zones. These are areas within which no appreciable genotype planting site interaction is expected to occur. Separate strains will be procured or developed for each zone. In mountainous regions several zones may be required. In flatter terrain. one zone may be sufficient. Generally, the more intensive the zonation, the greater will be the genetic gains. But the greater the number of zones, the greater will be the total cost of seed procurement or development, especially if superior strains are bred.
The term " planting zone " as used here should not be confused with " seed collection zone," which is generally used to indicate areas within which seed can safely be moved. Unfortunately, it can give the implication that local seed is always best and may discourage attempts to improve upon local seed. For example, tests have shown that seed of Pinus elliottii Engelm. can safely be moved within the whole northern portion of the species range (from northern Florida northward). But seed from southwest Georgia will usually outperform other sources when planted anywhere in the whole area noted. Namkoong (1969) gives other examples of this situation. The tree breeder should therefore select the most suitable seed source for each planting zone or breed a strain of trees for each as the case may be. The area of seed collection or the base populations from which superior strains will be bred may often, but not necessarily, coincide with the planting zones.
Guidelines
Early in a tree improvement programme the planner should delineate tentative planting zones within the region of his responsibility. In doing so he should use information from seed source trial plantings or progeny tests replicated over several sites if these are available. Data on climate, topography and other environmental factors believed to affect performance should also be used. Good judgement will be required to decide upon the number of zones in order to maximize gain per unit cost and to stay within available facilities. If the cost of seed procurement is relatively low, such as in using selected wild seed, the intensity of zonation can be rather high. But if costs of seed procurement will be high, such as in breeding superior strains, relatively few zones would be preferred at least to start with. As work progresses and planting trial information becomes available, changes can be made. Such changes may indeed be a fruitful avenue for future genetic gains.
SPECIES
Selection of species has very definite possibilities for obtaining better forest planting stock. This is evidenced by numerous successful plantings of exotics (MacDonald et al., 1957; Fielding, 1960; Streets, 1962; Morandini, 1964). These opportunities can easily be overlooked by tree improvement workers. It is often felt that nature has already done the job of species selection and, therefore, that exotics have only a very small chance of success. This is usually true, but in many cases exotics never had the opportunity of being " tested by nature " in the new location, because of isolation.
In fact there is reason to believe that, if environmental factors are similar, the more remote a species is geographically, the better is the chance for adaptation to a particular locality. Exotics adjacent to the new locality probably had an opportunity to invade the new locality, but they ['ailed to do so because of unsuitability. On the other hand, a distant exotic growing in a similar environment possibly never had the opportunity; therefore it has a chance of success.
It must also be kept in mind that particular conditions under which man grows trees may be appreciably different, from nature's method. For example, conditions for the natural germination of seed and the establishment of seedlings from direct seeding of an exotic may not be suitable in the new region. If seedlings are grown in a nursery and subsequently outplanted, this stumbling block may be overcome. Thus, a particular exotic may fail under natural conditions of establishment but may be successful under artificial conditions.
Guidelines
If native species are not satisfactory, consider the possibility of finding exotic species that would be suitable. Existing trials may have revealed a suitable exotic species. If' trials of exotics are lacking, they should be started along with efforts to improve native species.
SEED SOURCES
Appreciable racial variation occurs in practically all species. This variation offers opportunities for obtaining better seeds for forest planting (Callaham, 1964). Wettstein-Westersheim and Hufnagl (1960) stated that, by paying full attention to racial variation, Europeans can expect an average annual volume yield increase of 1 cubic metre per hectare. These opportunities occur whether native species or exotics are used.
Patterns of racial variation can take many forms. The genetic variation can be gradual (clinal) or discontinuous (ecotypic) in nature. However, combinations of both types usually occur. The nature of the variation usually depends upon the existence or absence within the species range of isolation and of abrupt changes in environment. If the species range includes isolated islands, definite ecotypes may occur. The same would be true for remotely isolated stands, which occur in some species. In both of these cases, however, clinal variation may occur within the patches or islands, associated with gradual changes in latitude, temperature, rainfall or other factors. In mountainous regions clines associated with elevation will usually occur, often with abrupt changes at mountain ridges or streams, associated with changes in aspect.
Squillace (1966) suggested that in most species inherent growth rate will be highest where environmental conditions are most suited to rapid growth. There the trees will be both phenotypically and genetically superior in growth rate. In regions where the environment is severe (often toward the fringes of the species range), the trees gradually become genetically inferior in growth rate.
Interactions between seed source and planting zone will often occur when seed is moved too far from its native environment. This may often be true for economically important traits, especially survival. Therefore, the problem is to determine the most suitable seed source for the planting zone in question, and this is best determined by provenance testing.
Guidelines
In making commercial seed collections and in choosing a population base for tree improvement work, an attempt should be made to determine the best source for each planting zone. Available information should be utilized on racial patterns, interactions with planting site, and environmental factors. If too much guesswork is involved, local provenance tests should be established. In such tests, seed should be included from those regions showing the best phenotypic performance (usually the optimum environment) along with other promising sources. Until reliable information becomes available, local stands are to be preferred if a native species is used. Seed should be collected from local land races (de Vecchi Pellati, 1969) and from native regions which most nearly match the environment of the planting zone if an exotic is used.
SEED PRODUCTION AREAS
Seed production areas (seed stands) offer a relatively cheap means of obtaining seed of known origin almost immediately. They also offer the possibility of modest genetic improvement. Good stands of trees are selected and thinned to favour the best trees (Andersson, 1963; Matthews, 1964). Cultural practices to stimulate seed production may be employed.
Genetic gains are likely to be small, especially if the seed, production areas are established in natural stands where age, spacing and microsite differ appreciably among trees. Gains are likely to be greater if they are established in plantations, providing the seed used in establishing the plantations came from a number of parent trees and from localities known to be suitable from the standpoint of adaptability. Convenience in obtaining seed of known origin is usually the main advantage of seed production areas.
Guidelines
If the need for seed is urgent, seed production areas may be established to provide an immediate source of seed of suitable origin, using preferably suitable plantations or managed natural stands for this purpose.
A suitable plantation is one which has proved itself to be, or originated from a seed source known to be suited to the area, and which was established from a mixture of seed from many parent trees. If suitable plantations or managed stands are not available, the seed production areas should be established in the best unmanaged native stands.
At least one seed production area should be established for each planting zone. Seed production areas should be large enough to minimize the possibility of inbreeding depression (Langner, 1960). If an exotic is used, seed production areas may be set up in several of the best sources in the planting trials. However, in this event the possibility of hybridization must be guarded against; isolation from undesirable pollen may be required.
SELECTING TREES
Phenotypic selection of superior individuals within a species is a basic part of most present forest tree improvement programmes. Through propagation of selected individuals, the breeder expects to make a genetic gain. The extent of genetic gain from mass selection depends strongly upon the magnitude of genetic variation in the population, the uniformity of environmental conditions in the population, and the intensity of selection. There is not much the breeder can do about genetic variance, but this usually is not a problem in forest trees. However, much can be done to increase the effectiveness of phenotypic selection by considering or manipulating selection intensity and environmental uniformity (Squillace, 1967).
Guidelines
If suitable plantations (as defined earlier) or managed natural stands are available, these should be preferred as the base population for selection rather than unmanaged natural stands. However, selection should not be restricted to a single stand but should be made in several stands with all possible adjustments for environmental effects. If provenance plantings are used, and if wind-pollinated seeds are collected for progeny testing, the possibility of hybridization should be remembered.
In developing plans for selection, proper weight should be given to the traits desired. If this is not done, there is a tendency to select for many traits and to give each equal weight. A selection index may be prepared if the required data are available (van Buijtenen, 1960). Otherwise, some sort of scoring index can be developed (Coordination Committee for Forest Tree Breeding and Genetics, 1964). If the main goal is high yield of wood per unit area per year, weight should be given to rapid growth and resistance to pests and diseases. Straight stems and small, horizontal branches are usually next in order of importance in most programmes, especially if the major products are sawlogs and veneer logs.
As a rule, it is not wise to select strongly abundant fruit production (Libby et al., 1969). Admittedly, selections must be capable of flowering, and abundant cone production is desired in seed orchards. But trees in commercial plantings should not be extremely fruitful because this may reduce wood production. Cultural techniques are becoming available for stimulating flowering in seed orchards (Matthews, 1964). Also, total fruit production can be increased by increasing orchard size and protecting from pests and diseases.
Generally, if heritability is low, more selections should be made than if heritability is high; even though this usually means decreasing the intensity of selection. Also, when heritability is low, greater emphasis should be placed on progeny testing than when heritability is high.
SEED ORCHARDS
In order to use superior selections in progeny tests and for seed production, most tree breeders employ seed orchards (Andersson, 1963; Matthews, 1964). Mass propagation by asexual means will be discussed later. The selections for seed orchards can be propagated either vegetatively to produce clonal seed orchards or by seed to produce seedling seed orchards. The relative merits of clonal versus seedling seed orchards have been discussed by many authors, for example Barber and Dorman (1964), Johnson (1964), Libby (1964), Stern and Hattemer (1964), Toda (1964), Wright (1964a), Zobel and McElwee (1964), and Namkoong et al. (1966). If seed orchards are to be established primarily to simplify progeny testing, Sweet (1969) has described a time- and cost-saving procedure for grafting shoots bearing flower buds in order to speed and simplify the job of producing pedigreed progenies.
Guidelines
Seed orchards should only be established for breeding if saving in time and other costs can be realized over controlled pollinations of the original selections. Seedling seed orchards should be considered if:
1. progenies can be adequately evaluated early in life;
2. the orchards can be established on more or less normal planting sites;
3. adequate supplies of seed can be obtained from trees of seedling origin;
4. vegetative propagation is difficult.
These conditions may permit effective roguing of families and of individuals within families on the basis of performance in the seed orchard. If controlled pollination of the original selections is feasible, it is better to produce control-pollinated seed. By a polycross mating scheme, a mixture of pollen from many of the selections may be prepared, and used in controlled pollinations on the selections. If controlled pollination is not feasible, wind-pollinated seed for the orchards may be used. Procedures, for seedling seed orchards are discussed by Goddard and Brown (1961), Wright (1961), and Goddard (1964).
Clonal seed orchards al e preferable if the special conditions required for early evaluation and roguing cannot be met and if there are no appreciable difficulties with vegetative propagation.
If at all possible, the orchards should be designed to permit roguing after progeny testing. An urgent need for seed or other considerations may prevent this. Orchards established before progeny testing will not produce seed having maximum genetic improvement. After progeny testing and roguing to remove genetically inferior parents, such orchards may not be suitable for maximum production of genetically superior seed. New orchards containing only proven elite parents will probably be needed.
PROGENY TESTING AND ROGUING
Progeny testing permits the breeder to make a genetic gain over and above that obtained by mass selection. Numerous variations of the technique are in use. Each variation has advantages or disadvantages peculiar to species, traits of interest and other considerations (Langner, 1960; Andersson, 1963; Johnson, 1964; Namkoong et al., 1966). Progeny tests may be established prior to, concurrent with, or after establishment of seed orchards. The tests may employ wind- or control pollinated seeds with one of several possible mating schemes. Progeny tests may be used as a basis for determining which selections to use in establishing an orchard. They provide the basis for roguing orchards already established.
Guidelines
If heritability is relatively low, strong emphasis should be placed on progeny testing in order to ensure appreciable gains; in this event, many selections are better than few. Tests should be started as soon as possible to permit a rapid turnover of selection and testing cycles. If heritability is relatively high, the offspring will closely resemble their selected parents and progeny testing may be minimized or omitted.
With limited resources progeny testing can be started concurrently with establishment of seed orchards, using wind-pollinated seed from the selections. As soon as the traits to be improved can be evaluated reliably, inferior parents can be rogued from existing orchards, and elite parents put into new seed orchards. In planning, second generation selection should be provided for. Although the progeny tests can be used for second generation selection, it is better to use wind pollinated seed from the orchard, after roguing, to establish special plantings.
Where resources permit, controlled pollination is preferable. In this event, there are the alternatives of making individual tree matings (such as a tester scheme or a diallel or partial diallel scheme) or of employing polycrossing. Individual tree mating schemes are suitable for measuring general combining ability. They allow exploitation of specific combining ability. Inbreeding can be held to a minimum. However, in order to provide for maximum recombination, many matings would be required, and this is costly. Considering all factors, the polycross scheme is-probably the most desirable. With this scheme there should be many selections (say, at least 50). Using a pollen mix from all or most of the selections will provide a good measure of general combining ability and maximum recombination. Under these conditions inbreeding in second generation orchards can be kept low.
Simple designs for progeny tests are preferable. Small plots (1-10 trees) and many replicates (10 or more) rather than large plots with few replicates should be used.
VEGETATIVELY PROPAGATED SPECIES
For species that can be vegetatively propagated easily, special procedures are available. These may include phenotypic selection, establishment of clonal tests, reselection on the basis of clonal performance, and mass vegetative propagation of the final selections for use in commercial afforestation Supplementary breeding and the establishment of seed orchards may be used to enhance the programme in various ways.
A great advantage over sexual propagation methods is the likelihood of maximum genetic gain. This is because the resemblance between ramets and ortets is apt to be greater than that between progenies and parents. The technique also offers possibilities for mass propagation of promising hybrids, which ordinarily may be difficult to produce by sexual methods (Schreiner, 1963).
Such techniques have been used successfully in improvement of Populus species (Giordano, 1960; Schreiner, 1963; Steenackers, 1969) and of Pinus radiata D. Don (Fielding, 1963; Thulin, 1969). Application to most pines is limited at present by the relative difficulty and cost of vegetative propagation. The possibility of loss of vigour resulting from physiological aging must be recognized (Sweet, 1964; Franklin, 1969).
Guidelines
If the species being used can be vegetatively propagated with ease, permitting clonal afforestation, highest priority should be given to clonal testing, reselection, and mass propagation procedures. Supplementary breeding and establishment of progenies should be employed to provide bases for future genetic gains.
CREATING NEW VARIATIONS
Hybridization between species and races offers definite possibilities for tree improvement in some situations (Wright, 1964b). An outstanding example of success is the hybrid between Pinus taeda L. and P. rigida Mill. (Hyun, 1960). Righter (1960) discusses the potentialities of inbreeding followed by hybridization as has been done with a number of agricultural crop plants.
Inducing mutations and polyploids likewise offers opportunities for improvement (Gustafsson, 1960). The development of superior Populus tremula L. polyploids in Sweden is an example of success with the polyploid technique.
Undoubtedly, these techniques will play an important role in long-term aspects of tree improvement. But opportunities for most practical-minded organizations to take advantage of these techniques are limited at present.
Guidelines
Techniques capitalizing on natural variation are preferable. Advantage should be taken of hybrids within and between species, of mutants, or of polyploids, as they are developed and become available from research.
Possibly the most important problem facing the instigator of a tree improvement programme is to convince himself and those who control his funds that a project would be practical and profitable. At first, a brief general plan of work should be drawn. The plan should encompass an economic analysis, even if data are weak or lacking and many assumptions are required. It should show estimated costs and gains, to the extent possible. Economic analyses or discussions by Perry and Wang (1958), Barber (1963), Cole (1963), Marler (1963), Davis (1967a and 1967b), Laffers (1967), and Swofford (1968) would be of help in making an appropriate plan and analysis.
The next step is to develop a more detailed plan. In doing so one should seek advice from experts. Visits to vigorous and successful action programmes would be helpful. The work plan should contain detailed procedures for the first few years of work, but it should provide great flexibility beyond that. It should allow for the possibility of changes in financing of the programme, new research results and changes in wood product markets.
The planner of a tree improvement programme should take advantage of any technical assistance available through international organizations. Several working groups in Section 22 of the International Union of Forestry Research Organizations may be of help. The Group on Provenance Research and Testing has recently developed a standardization of methods for provenance research and testing (Lines, 1967). This publication includes instructions for collection of seeds, design of field experiments, and evaluation of results. The Group on Procurement of Seed for Provenance Research is currently collecting seed of Pseudotsuga menziesii (Mirb.) Franco, Pinus contorta Dougl. and other conifers from western North America. The Group on International Testing of Clones is preparing a standardization of methods for clonal research and testing. The Group on International Provenance Trials will provide for exchange of information among cooperators and for coordination of tests.
The Organisation for Economic Co-operation and Development has recently prepared a scheme for the control of forest reproductive material moving, in international trade. This should help ensure that breeders receive the quality of material they request.
FAO offers technical assistance in a variety of ways. It serves as a clearinghouse for information on sources of seed. It sponsors training courses and study tours and helps provide expert consultation on the science and practice of forest tree improvement.
Opportunities should be explored for forming a local cooperative action programme or joining one already in existence. These cooperative programmes bring together forestry organizations, closely associated geographically and with similar goals, into union with a guiding organization which provides expert leadership by means of consulting and statistical services. Excellent examples in the United States are the North Carolina State-Industry Cooperative Forest Tree Improvement Program and the University of Florida Cooperative Forest Genetics Research Program. Such cooperatives also provide the opportunity for exchange of breeding stock. Sindelar and Zavadil (1969) discuss cooperation between research and practice in more detail.
ANDERSSON, ENAR. 1963. Seed stands and seed orchards en the breeding of conifers. FAO/FORGEN-63, Vol. II, 8/1: 1-18.
BARBER, JOHN C. 1963. How much is forest genetics helping the forester by increasing growth, form and yield? Proc. 7th South. Conf. Forest Tree Improvement, p. 16-20.
BARBER, JOHN C. &; DORMAN, KEITH W. 1964. Clonal or seedling seed orchards. Silvae Genet., 13: 11-17.
BURLEY, J. 1969. Breeding tropical pines. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-9/2. 10 p.
CALLAHAM, R. Z. 1964. Provenance research: investigation of genetic diversity associated with geography. Unasylva, 18(2-3): 40-50.
COLE, DONALD E. 1963. Management of pine seed production areas. Proc. 7th South. Conf. Forest Tree Improvement, p. 44-49.
CO-ORDINATION COMMITTEE FOR FOREST TREE BREEDING AND GENETICS. Directions for evaluating and selecting plus trees. Unasylva, 18(2-3): 128-129.
DAVIS, LAWRENCE S. 1967a . Investments in loblolly pine clonal seed orchards: production costs and economic potential. J. For., 65: 882-887.
DAVIS, LAWRENCE S. 1967b . Cost-return relationships of tree improvement programs. Proc. 9th South. Conf. Forest Tree Improvement, p. 20-26.
DE VECCHI PELLATI, E. 1969. Evolution and importance of land races in breeding. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-10/5.
FIELDING, JOHN M. 1960. The role of exotic species in forest tree improvement. Proc. 5th World Forestry Congress, 2: 742-746.
FIELDING, JOHN M. 1963. The possibility of using cuttings for the establishment of commercial plantations of Monterey pine. FAO/FORGEN-63, Vol. II, 5/10: 1-5.
FRANKLIN, E. C. 1969. Ortet age has strong influence on growth of vegetative propagules of Pinus elliottii Engelm. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-7/8.
GIORDANO, ERVEDO. 1960. Recent work on Populus deltoides in Italy. Proc. 5th World Forestry Congress, 2: 811-812.
GODDARD, R. E. 1964. Tree distribution in a seedling seed orchard following between and within family selection. Silvae Genet., 13: 17-21.
GODDARD, RAY E. & BROWN, CLAUD L. 1961. An examination of seed orchard concepts. J. For., 59: 252-256.
GREATHOUSE, T. E. 1969. Translating breeding principles into practice. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-9/3. 5 p.
GUSTAFSSON AKE. 1960. Polyploidy and mutagenesis in forest tree breeding. Proc. 5th World Forestry Congress, 2: 793805.
HYUN, SIN KYU. 1960. Mass production of control-pollinated seed of conifers. Proc. 5th World Forestry Congress, 2: 787-792.
JOHNSON, HELGE. 1964. Forest tree breeding by selection. Silvae Genet., 13: 41-49.
LAFFERS, A. 1967. Considerations on the profitability of seed orchards of conifers. Les (Bratislava), 23: 552-555.
LANGNER, WOLFGANG. 1960. Improvement through individual tree selection and testing seed stand, and clonal orchards. Proc. 5th World Forestry Congress, 2: 778-782.
LIBBY, W. J. 1964. Clonal selection, and an alternative seed orchard scheme. Silvae Genet., 13: 32-40.
LIBBY, W. J., STETTLER, R. F. & SEITZ, F. W. 1969. Forest genetics and forest tree breeding. Ann. Rev. Genetics., 3: 469-494.
LINES, ROGER. 1967. Standardization of methods for provenance research and testing. Proc. 14th IUFRO Congress, 3: 673-719.
MACDONALD, J. et al. 1957. Exotic forest trees in Great Britain. London, Forestry Commission. Bull. 30. 167 p.
MARLER, R. L. 1963. Economic considerations of the genetic approach. Proc. 7th South. Conf. Forest Tree Improvement, p. 59-63.
MATTHEWS, J. D. 1964. Seed productions and seed certification. Unasylva, 18(2-3): 104-118.
MORANDINI, RICCARDO. 1964. Genetics and improvement of exotic trees. Unasylva, 18(2-3): 51-60.
NAMKOONG, GENE SNYDER, E. B. & STONECYPHER, R. W. 1966. Heritability and gain concepts for evaluating breeding systems such as seedling seed orchards. Silvae Genet., 15: 76-84.
NAMKOONG, GENE. 1969. The non-optimality of local races. Proc. 10th South. Conf. forest Tree Improvement.
PERRY, THOMAS O. & WANG, CHI WU. 1958. The value of genetically superior seed. J. For., 56: 843-845.
RIGHTER, FRANCIS I. 1960. Forest tree improvement through inbreeding and intraspecific and interspecific hybridization. Proc. 5th World Forestry Congress, 2: 783-787.
SCHREINER, ERNST J. 1963. Improvement of disease resistance in Populus. FAO/FORGEN-63, Vol. II, 6a/2: 1-21.
SINDELAR, J. & ZAVADIL, Z. 1969. Cooperation to get results into action programs. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-9/6. 14 p.
SQUILLACE, A. E. 1966. Racial variation in slash pine as affected by climatic factors. Southeast. Forest Exp. Sta., U.S. Forest Serv. Res. Pap. SE-21. 10 p.
SQUILLACE, A. E. 1967. Selection of superior trees. Proc. 9th South, Conf. Forest Tree Improvement, p. 7-9.
STEENACKERS, V. 1969. L'état actuel de la sélection et de la création des clones de peupliers résistants aux diverses maladies. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-9/4. 17 p.
STERN, K. & HATTEMER, H. H. 1964. Problems involved in some models of selection in forest tree breeding, Silvae Genet., 13: 27-32.
STREETS, R. J. 1962. Exotic forest trees in the British Commonwealth. Oxford, Clarendon Press.
SWEET, G. B. 1964. The effect of physiological age of scion on growth of grafts in Pinus radiata. New Zealand Forestry Res. Notes No 37. 7 p.
SWEET, G. B. 1969. A method for the rapid production of controlled-pollinated seed in Douglas-fir. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-11/11.
SWOFFORD, THOMAS F. 1968. Profit and quality gain through tree improvement. Southern Lumberman, 217 (2704): 170-172.
THULIN, I. J. 1969. Breeding of Pinus radiata through seed improvement and clonal afforestation. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-9/5. 6 p.
TODA, ROYOOKITI. 1964. A brief review and conclusions of the discussion on seed orchards. Silvae Genet., 13: 1-4.
VAN BUIJTENEN, J. P. 1960. A selection index for aspen based on genetic principles. Progress Report. N° 6 to Lake States Aspen Genetics and Tree Improvement Group. Appleton, Wis., Inst. of Paper Chemistry.
WETTSTEIN-WESTERSHEIM, WOLFGANG & HUFNAGL, HANS. 1960. Improvement through racial selection and testing. Proc. 5th World Forestry Congress, 2: 775-777.
WRIGHT, JONATHAN W. 1961. Progeny tests or seed orchards. En Recent Advances in Botany, 2: 1681-1687.
WRIGHT, JONATHAN W. 1964a . Flowering age of clonal and seedling trees as a factor in choice of breeding system. Silvae Genet., 13: 21-27.
WRIGHT, JONATHAN W. 1964b . Hybridization between species and races. Unasylva, 18(2-3): 30-39.
ZOBEL, BRUCE, & MCELWEE, R. L. 1964. Seed orchards for the production of genetically improved seed. Silvae Genet., 13: 4-11.
