ERNST J. SCHREINERERNST. J. SCHREINER is principal geneticist, forest genetics research, at the Northeastern Forest Experiment Station, Forest Service, United States Department of Agriculture, Durham, New Hampshire. This forest genetics research is carried on in cooperation with the University of New Hampshire. This paper stems from Mr. Schreiner's attendance as a consultant to the FAO Forestry and Forest industries Division at the FAO Technical Conference on Exploration, Utilization and Conservation of Plant Gene Resources held at Rome, 18-26 September 1967, in cooperation with the International Biological Programme of the International Council of Scientific Unions.
A though man's direct or indirect activities have affected the population structures of many forest tree species, forest trees are still essentially wild populations. The poplar clones selected and cultivated in Europe and the Near East have been called primitive cultivars, but it must be remembered that, in contrast to the genetic variability of agricultural cultivars that are sexually propagated, a clone is a single genotype.
GEOGRAPHIC VARIATION IN FOREST TREES
There is an extensive literature on provenance tests during the past 75 years; Wright (FAO 1962, chapter 6) has reviewed this research. In his report to the FAO Technical Conference on Exploration, Utilization and Conservation of Plant Gene Resources,1 Callaham has abstracted the general conclusions derived from these provenance studies.
1 The report is in process of publication in a handbook on the exploration, conservation and utilization of plant gene resources based on the conference papers and discussions.
Wide intraspecific variation has been reported in many forest tree species, particularly in such important timber species as Pinus sylvestris, Pinus ponderosa, Pseudotsuga menziesii, and Picea abies. But the amount of intraspecific variation is not necessarily associated with an extensive natural range. Pinus resinosa, with a natural east-west range from Nova Scotia to Manitoba, and south to Pennsylvania, Michigan, and Minnesota, is a very uniform species. On the basis of a 25-year-old test planting of 50 different geographic-climatic seed sources of Pinus resinosa, Hough (1967) concluded that seed sources within the natural range of this species were relatively uniform in development. Fowler's (1965) research led him to conclude that "... red pine both as individual trees and as a species, is homozygous for a large number of alleles, self-fertile, self-compatible and that seedlings resulting from self-pollination exhibit little or no inbreeding depression."
NEED FOR BIOSYSTEMATIC STUDIES
The report of the International Union of Forest Research Organizations (IUFRO) working group on the standardization of provenance research (Lines, 1967a) recommends that provenance trials should be preceded by biosystematic studies. That the sampling of forest tree species for the conservation of gene resources should be preceded by a study of the variation of the species within its natural range was strongly stated by Zobel in his report to the conference2 on tree collection trips in Mexico. He emphasized that: "Thus far we have not been able to exploit the variation present in Mexican pines. The reason for failure appears to be simple: ignorance of the biological facts."
2 In preparation (see footnote 1).
INTROGRESSION AND HYBRID SWARMS
There is now sufficient evidence for introgression in our important timber trees to justify the statement that it must be taken into consideration in all but a few of our forest species. It is essential to differentiate between introgression that over a long period of time has incorporated the genes of two species into one or both of the parent populations, and hybrid swarms of relatively recent origin.
Introgression has been reported in Quercus, Tilia, American sugar maples, Betula, Eucalyptus, Picea, and Pinus. Wells and Wakeley (1966) have presented an interesting case for introgression with more resistant pine species (such as shortleaf pine and pitch pine) as the probable basis for resistance to fusiform rust in loblolly pine provenances.
POLYPLOIDY
Natural polyploidy is extremely rare in conifers; polyploid races are unknown but there are records of occasional polyploid individuals. Biosystematic studies and adequate sampling in some angiosperm genera (for example, Salix, Betula, Ulmus, Prunus, Acer, Fraxinus) will be complicated by intraspecific polyploidy. Matskevich (1959) has discussed the role of polyploidy in forest selection. The reader is referred to Wright (FAO 1962, chapter 9) for a review of the occurrence of natural polyploidy in forest trees and the importance of polyploidy in forest tree breeding.
POPULATION
Population, subpopulation, neighbourhood, provenance, race, biotic unit and stand have all been used in reports on forest trees; but too seldom have these terms been clearly defined by the authors. Definitions are not merely an exercise in semantics; they are essential for clarity of thought and precision of communication.
Tigerstedt (1967) notes that it is exceedingly difficult to form an unambiguous concept of what shall be considered a population of forest trees; and that the usual definition of a population, mainly established by zoologists, is quite useless to forest geneticists. He bases his opinion on the following particular features of long-lived forest trees:
1. Generations are exceptionally long.2. The abundance of seed production directs natural selection pressure on zygote fitness ability, and since the seed is shed mainly within the same stand, there is very little zygote interchange between stands.3
3. A particular tree may be pollinated by its neighbour, or by pollen masses of unknown origin, and pollination will be basically different from year to year.
4. The dominant forest tree populations covering large continuous areas (thousands of kilometres) will eventually give rise to adaptation of a clinal type (which is basically different from the "genetic cohesion" of localized populations) but with local adaptations along a cline giving rise to "linkage" clines.
3 The latter part of this statement would apply to most conifers and heavy-seeded hardwoods but not to such genera as Populus, Salix, and even Betula, with light seed that can be wind-distributed over long distances.
The term "deme" was proposed by Gilmour and Gregor (1939) to mean any assemblage of individuals closely related taxonomically. In the opinion of Langlet (1967): "The only true 'basic unity' resulting from ecological adaptation is the deme, in sensu Carter-Mayr... i.e., the local population within which there are no discernible ecological groups or clines, the sole individual differentiation being random and causeless." To avoid increasing technical jargon, the term deme can be accepted as the basic population unit - the interbreeding community within which there are no discernible ecological groups or clines. Deme size, with reference to area and number of individuals, will differ not only with species but also with local environment.
GENE EXCHANGE
The extent of gene exchange between demes (of forest tree species) has been the subject of some research and considerable speculation. Studies on the e effective pollination distance (and the possibilities for self-pollination) have been limited to anemophilous species; such studies do not appear to have been made with entomophilous forest trees. Wright (FAO 1962) has presented an excellent summary of the (then) available information and hypotheses on population size and gene migration rates. On the basis of this early information we may have been underestimating the areal extent of such exchange.
Sarvas (1967) has estimated, from recent measurements, that approximately one half of the pollen catch recorded at crown height in Pinus sylvestris subpopulations has its source outside the subpopulation. He concludes that: "Pollen exchange between subpopulations is very efficient, maybe more efficient than was formerly surmised. This finding somewhat lessens the interest in long-distance pollen transfer, because efficient gene exchange between subpopulations also serves gene exchange between populations."
MARGINAL POPULATIONS
It is generally assumed that marginal populations exhibit relatively little genetic variation. But the importance of semi-isolated populations of forest trees is illustrated by the so-called lost pines of Texas, outlier populations of Pinus taeda, but not out of the range of possible long-distance pollination. According to van Buijtenen and Stern (1967) the lost pines probably represent fairly recent introgression with Pinus echinata. They have shown considerable variability, particularly in adaptability to dry sites and resistance to fusiform rust. These authors note that marginal populations offer "some rather outstanding opportunities for utilizing a genetic potential towards special adaptations."
Because of the long-term investment in establishment and maintenance, and the requirement for diverse and extensive outplanting sites, it is inevitable that in most cases gene pools of forest trees (for the conservation of gene resources) will have to be incorporated in "practical" trials and collections. This will require compromises in sampling, evaluation and conservation.
The following types of collection meet present practical forestry needs:
1. exotic species trials;
2. provenance collections;
3. clonal and seedling seed orchards;
4. clone banks, also called clone archives.
But the objectives, and therefore the methodologies, for the establishment of these practical collections differ considerably from those required for the sampling, evaluation, conservation, and utilization of forest gene resources for future improvement breeding.
Provenance trials offer the best possibilities for minimum compromise between both provenance and gene pool objectives. The two objectives of provenance research have been defined in the IUFRO report (Lines, 1967a) as "a practical objective: to locate populations of trees, or provenances, whose seed will produce well-adapted, productive forests in a given region" and "a scientific objective: to define the genetic and environmental components of phenotypic variability between trees from different geographic sources."
Establishment of gene pools for future breeding should be with the object of providing the broadest possible genetic base, on the sound assumption that many of the future needs for genetic improvement are not known at present. This requires the conservation of both presently identifiable and cryptic variation. It can be predicted that resistance to currently unimportant or unknown diseases, insects, and nematodes will be among the most sought-for cryptic characteristics in gene pools of forest trees.
SAMPLING
The excessive concentration on ad hoc criteria in the selection of superior phenotypes of forest trees was criticized during the conference. There was emphasis on the need for a clear distinction between sampling for variability and sampling for superiority - to collect and preserve the fullest possible genetic diversity rather than limit collection to preconceived ideas of superior trees.
Sampling for variability
The intensity of sampling for maximum variability should be based on available biosystematic information; the selection of individual trees for seed collection should be at random.
The IUFRO recommendations (Lines, 1967a) specify that the natural unit for a provenance sample is a stand, defined as "a sufficiently large population of trees possessing sufficient uniformity in composition, constitution and arrangement to be distinguishable from adjacent crops." The recommendation that the stand should be isolated from minus stands and from related species or varieties would limit variability, and would eliminate collections from hybrid swarms and areas of introgression. A further limitation on random broad-base sampling is the restriction on selection of parent trees for seed collection: "Collection should be made from dominant and co-dominant trees."
There is reason to question the soundness of these restrictions even for provenance trials, particularly on the basis of the following quotation from the IUFRO report (Lines, 1967a):
"Biosystematic studies provide knowledge essential in planning traditional provenance experiments. But it cannot be too highly stressed that the important thing is the growth of the trees in the new environment and not the appearance of the parent stands from which they were collected. In western Europe flourishing plantations of Pinus contorta are growing from seed collected from scrubby looking trees growing on the coast of Washington; whereas, in the maritime climate of Great Britain poor plantations result from the phenotypically superior stands of Pinus contorta from the interior of Oregon or the Rocky Mountains."
It should be possible to reach a reasonable sampling compromise acceptable for both provenance and gene pool objectives.
Individual tree selection
There is an immediate need for phenotypic selection particularly of phenotypes that may be resistant to diseases or insects now of major or minor importance. The Marssonina leaf disease of poplars is an example of a disease, known for many years, that only recently has become an extremely serious threat to poplar culture in Europe. Resistance for any disease or insect will require intensive screening; but the judgement of a competent pathologist, entomologist, geneticist, or silviculturist who is familiar with the tree species and the pest would be sufficient to justify the selection of phenotypes for eventual screening.
The importance of conserving mutants was emphasized during the conference, particularly by Dr. Ake Gustafsson.
EVALUATION
The preservation of genetic resources could be centralized in a very few localities, but evaluation for variation and adaptability under different environments, including resistance to pests, will require replication under a wide range of environments. The IUFRO recommendations (Lines, 1967a) for short-term and medium-term provenance trials offer the best possibilities for minimum compromise of both provenance and gene pool objectives.
The objective of provenance studies is to determine the growth and development of populations under various environmental conditions; this requires periodic observations and measurements to determine average provenance values or ratings. For the evaluation of genetic resources in gene pools, the objective is to determine the genetic qualities and breeding value of individual trees; this will require progeny tests.
Short-term provenance trials
Short-term provenance trials, recommended in the IUFRO report, would be particularly suitable for combination provenance-gene pool plantations because they are designed for valid provenance results until the time when major competition sets in; they should be terminated before the first thinning. Plots can contain up to 16 trees arranged in squares, rectangles or lines. These trials are suitable for investigating ease of establishment, phonology, early growth rate, morphology, and early resistance to insects, diseases and climatic factors like frost.
Grading-out small nursery stock before outplanting, and thinning after the provenance trials are terminated would constitute a loss of individuals that may carry valuable genes. The gene pool objective would require:
1. outplanting all seedlings;
2. plantation spacing that would eliminate excessive competition.
With reference to grading, the IUFRO report recognizes that: "In special circumstances it may be the object of the experiment to compare the whole population of each provenance, in which case no grading is allowed." Since short-term provenance trials are not intended to evaluate the effect of stand competition, these plantations could be established at sufficiently wide spacing to meet the conservation objective for the establishment of gene pools.
Medium-term provenance trials
These are expected to yield valid results from the time of planting until about 1/3 to ½ the rotation age. The recommended minimum plot size is 6 X 6 plants, and sizes frequently used are 7 X 7 up to 12 X 12 plants. Such plantations would provide opportunity for phenotypic evaluation and for half and full-sib progeny tests up to 1/3 to ½ rotation age. Gradually increasing competition would permit evaluation and conservation of a large proportion of the best phenotypes beyond this age.
Early evaluation
Early diagnoses have included laboratory tests under closely controlled environments, and nursery and field tests of young trees. It is seldom safe to judge growth rate to maturity on the basis of juvenile growth; however, some characters such as tree form, drought resistance, and resistance to some diseases and insects can be determined at seedling or sapling ages.
Slow juvenile growth with gradually accelerating growth rate is an inherent characteristic of some individuals and probably of some provenances. The effect of growth conditions in the nursery may carry over for several to many years after outplanting; Lines (1967b) cites a report by Edwards that Larix decidua trees of the same provenance raised in different nurseries continued to show large differences in height up to 20 years after planting.
Resistance of hybrid poplar clones to the Japanese beetle was demonstrated under natural epidemic conditions in a replicated nursery-stool planting in Maryland (Schreiner, 1949). Lines (1967b) reported that height of Larix decidua provenances in Perthshire at 2 and 5 years was not a good indicator of height at 16 years, but incidence of canker at 5 years showed a clear relation with survival at 20 years after a period of severe dieback in the pole stage.
CONSERVATION
Bouvarel's report to the conference4 covered the possibilities and limitations for the conservation of gene resources of forest trees in situ, in arboreta, provenance test plantations, provenance conservation plantations, seed orchards, clone archives, and conservation by storage of seed and pollen.
4 In preparation (see footnote 1).
The Michaux Quercetum Project (Schramm and Schreiner, 1954) is an example of the possibilities for the evaluation and conservation of forest tree species in arboreta. In 1953, a total of 150 collections of approximately 100 seeds from individual trees of 37 oak species were received from foresters and botanists in 23 states of the United States. The collaborators were asked to make these collections from two randomly selected trees and to submit an herbarium specimen from each tree.
Irgens-Moller (1963) has described a Douglas fir archive collection, begun in 1957, that contained approximately 600 progenies in 1963. The objective is to bring together, in one locality, single-tree progenies of 10 trees from as many areas as possible within the natural range of the species, the progeny of each tree to be represented by approximately 10 seedlings in a completely randomized plantation.
Bannister (1963) made a proposal involving open-pollinated seed of Pinus radiata collected from 50 trees in each of four wild populations and two cultivated populations. Five trees of each of 300 progenies would be established in each of six blocks. He chose complete randomization "because of its simplicity in design, execution, and interpretation, and because it should effect massive cross-breeding between genotypes of different populations and families, leading rapidly to the synthesis of a new gene pool of exceptionally high heterozygosity."
It is obvious that all of the genetic variability cannot be preserved in a wild species; and the question, how much? can only be answered by, as much as practicable. It is felt that combination provenance-gene pool plantations can be designed to conserve an ample range of genetic variability.
All the genotypes in the original sample will not be preserved in provenance-gene pool plantations (or in any other collection). Natural selection over many years inevitably will eliminate some genotypes. In provenance-gene pool plantations established under a wide environmental range, different genotypes will be lost under the various environmental conditions. Thinnings, in medium-term trials, will further reduce the genetic base; the rate and amount of reduction will depend upon the original plantation spacing.
Seed storage
The seeds of many forest trees can be stored for 10 to 20 or more years; others, such as Populus, are short-lived. There is considerable information on the basic physiological processes of long-term seed storage, but little information on possible genetic changes due to method and length of storage.
Storage of forest tree seed is essential as a precaution against partial (or complete) failure of the first conservation plantings, and to provide seed for later establishment of additional plantations. But long-term conservation of forest tree gene resources by seed storage is of less importance than in the case of short-lived plants, mainly because of the years required for genotype evaluation, but also because the living collections may be expected to outlive the safe storage period of the seed collections from which they were derived.
Tissue culture would have a reasonably early possibility for practical use in the international exchange of vegetative propagules. Since tissue cultures can be grown free of disease, including virus, they could be exchanged between countries without the involvement of quarantine regulations. Such exchange would necessitate only short-term culture; this would mitigate the possibility of genetic change.
UTILIZATION
Genetic integrity versus panmixis
The maintenance of genetic integrity versus panmixis in the conservation of plants was discussed during the conference. Genetic integrity, through mass reproduction by seed, could be maintained in provenance conservation plantations as suggested in Bouvarel's report;5 but it is debatable whether the isolation of large numbers of populations to maintain genetic integrity in gene pools of long-lived forest trees is essential, and will justify the cost.
5 In preparation (see footnote 1).
The individual trees in the sample populations will maintain their genetic integrity during their life span. During this long period (50 to 100 years or more) individual trees can be progeny tested and used for breeding, and superior phenotypic variants can be vegetatively propagated and established in clonal seed orchards or clone banks for additional progeny tests, clonal screening and conservation. For clonal screening, they should be grown on their own roots; there is ample evidence from horticulture that the rootstock may effect not only the growth of the scion, but also such characters as disease resistance and frost-hardiness (Schreiner, 1966).
Maximum utilization
It is impossible to predict for what particular characters and breeding methods forest tree gene pools will be used. But it is certain that maximum utilization will require maximization of genetic diversity.
Since 1935, the author has advocated the importance of intra- and interspecific hybridization for the genetic improvement of forest trees (Schreiner, 1935, 1950, 1953, 1966). Forest tree improvement programmes have been too often narrowly oriented toward selection and breeding within a single native species. For forest trees, it is not the species but the genus that, through selective interspecific hybridization, will provide maximum diversity of genotypes needed for maximum genetic improvement.
For maximum future usefulness, forest tree gene pools should provide for association of the widest possible diversity of genotypes in plantations designed to permit extensive recombination of genes through free pollination, or, if necessary, artificial open pollination. Such mass intra- and interspecific hybridization in gene pools could, provide a practically inexhaustible source of gene combinations for the creation of synthetic varieties over a long span of years; this could save many years of controlled breeding.
Provenance-gene pool plantations would meet the requirement for mass intraspecific hybridization. But in addition, small-plot provenance-species gene pools should be established to provide for species hybridization.
Examples of the exploration, collection, and conservation of the genetic resources of forest trees were presented in reports to the conference by Lamb and Cooling (Low altitude tropical pines and fast-growing hardwoods), Larsen and Cromer (eucalypts), Pryor (Poplar and some tropical species), Zobel (Mexican pines).6 The FAO Forestry and Forest Products Division's report to the conference7 also summarized current projects.
6 These reports are in preparation (see footnote 1).7 PL/FO:PGR/67-P/35, 8 p. mimeo.
Lamb and Cooling (1967) have presented in more detail the present status, the future prospects, the reputation of provenances in exotic plantations, and the amount of seed exported, for each of the following species and varieties of low level tropical pines:
GROUP I. Central America, including the Caribbean and Bahamas. Pinus caribaea Mor, var, caribaea Barr. &; Golf.; var hondurensis Barr & Golf.; var. bahamensis Barr. & Golf.; Pinus cubensis Griseb Pinus occidentalis Sw.; Pinus oocarpa var. ochoterenai Martinez; Pinus strobus var. chiapensis Martinez; Pinus tenuifolia Benth.; Pinus tropicalis Mor.GROUP II. Southeast Asia and Australia. Pinus kesiya Royle ex Gordon; Pinus merkusii Jungh and de Vriese.
These authors conclude that firm action is needed by the countries concerned to preserve representative stocks of each major provenance and the best phenotypes within each provenance. They suggest that international forestry aid programmes should give high priority to the collection and dissemination to other tropical countries of representative samples of the seed of the best provenances for the establishment of gene pools.
The Forest Research Station, Wageningen, Netherlands, reported to the conference8 that it has taken measures for the preservation of clones and populations of the indigenous Populus nigra because this species is rapidly being replaced in the Netherlands by the planting of superior hybrid clones. Since it is anticipated that Fraxinus excelsior will become increasingly important in the future, the preservation of clones of this species also is well under way. Old cultivars of elm that became obsolete with the outbreak of the Dutch elm disease, to which they are generally susceptible, are also disappearing. Since these primitive cultivars represent stocks of otherwise desirable genes, the Station has a three-phase elm programme:
"1. inventory;
2. evaluation, because not all of the 200 cultivars are worth conserving;
3. conservation."8 PL/FO:PGR/67-M/15, 3 p. mimeo.
The elm project also included a three months' collection tour in 1960 to sample three elm species in the Himalayan foothills; the main species (Ulmus wallichiana) appears to be in danger of extinction in much of its area due to lopping the leaves for cattle fodder.
In the spring of 1967, the Poplar Council of America cooperated with the International Poplar Commission of FAO on the collection of P. deltoides seed. No attempt was made to sample the species systematically; as far as possible collections were made in the localities requested by the consignees who had submitted their request through the International Poplar Commission.
Financial support is generally the major difficulty encountered in exploration and seed collection. The financing of the IUFRO Working Group on Procurement of Seed for Provenance Research is unique; it is a reimbursable (self-liquidating) operation that may be applicable to other seed-procurement programmes. The coordinating centre of this working group is located at the Tree Improvement Station, Humlebaek, Denmark, and is under the immediate supervision of Mr. Helmuth garner. Through the joint effort of the managing director of the Danish State Forestry Department and Mr. Barner,9 a grant of $1500 was obtained from the I.C. Jacobsen's Memorial Foundation and Den Danske Landmandsbank granted a cash credit of about $22000. This cash credit was made on the condition "that the coordinating centre does not deliver any seed unless it has been paid for beforehand, and also that the yearly withdrawn amount of money should be fully paid, before other withdrawals are being made for the purpose of new seed collections." Twenty-two countries agreed to support this project. Douglas fir seed collections were made in 1966 under the field supervision of Mr. H. Stubgaard in cooperation with Canadian and United States foresters, 103 samples were collected at a total cost of approximately $22000. The price per kilogramme of seed was based on the total cost and the net yield of clean seed.
9 Personal communication from Mr. Barner.
Action needed
In a report to the conference10 the FAO Forestry and Forest Industries Division recommended that, in planning for the exploration and collection of genetic resources, it would be a rational policy to build on existing research organizations rather than to attempt to set up centres de novo or in isolation; eight such institutions were listed. The desirability of some form of reserve finance was suggested, at regional or international level, to cover both the exceptional seed year and any emergency operation necessitated by the imminent but unforeseen destruction of an important seed source. The report also discussed possible priorities of species currently in demand. For example, in considering Tropical High Forest species: " Priority should be given to the fast-growing pioneer species (e.g. Maesopsis, Triplochiton, Terminalia, Anthocephalus) which are in current demand for their useful general purpose timber and the comparatively short life cycle, which is likely to induce higher genetic variability than in the long-lived climax species." Teak was also given special mention because of its importance in the world timber trade.
10 PL/FO:PGR/67-P/35, 8 p., mimeo.
In addition to the collection and conservation of forest tree species to provide a broad genetic base for future breeding, there is also an immediate need for special collections. For example, gingham (1967) has stressed the importance of instituting botanical surveys and seed collections within the botanical range of Pinus griffithii, P. armandii, P. koraiensis, P. siberica, and P. pumila for blister rust resistance breeding.
The conference reports and discussions on the urgent need for action on an international basis resulted in the following requests to the Director-General of FAO (quoted from the report of the Fourteenth Session of the FAO Conference):
"244. Forest tree genetic resources. The Conference requested the Director-General to take into account Recommendation No. 62 of document C 67/AG/FO/1, in formulating the Programme of Work and Budget for 1970/71. It recognized that as development proceeds in the less as in the more advanced areas of the world, the reserves of genetic variation stored in the natural forests have been or are being displaced on an increasing scale. Moreover, efforts to explore and collect forest genetic sources were, on a world scale, inadequate and inadequately concerted.""245. The Conference requested the Director-General to establish a panel of experts on forest gene resources to help plan and co-ordinate FAO's efforts to explore, utilize and conserve the gene resource of forest trees and, in particular, help prepare a detailed short-term programme and draft a long-term programme for FAO's action in this field and to provide information to Member Governments. The Director-General should convene at least one session of the members of the panel in the biennium 1968/69."
The mills of international cooperative action, "like the mills of the gods," grind slowly. Active international coordination is essential to assure that they grind inexorably and "exceeding fine."
BANNISTER, M. H. 1963, Planning a genetical survey of Pinus radiata populations. FAO/FORGEN-63, 1, 4/1, 1-10.
BINGHAM, R. T. 1967, International aspects of blister rust resistance in white pines. Proc. XIV IUFRO Congress, Munich, 3, 832-841.
D'AMATO F. 1965, Endopolyploidy as a factor in plant tissue development. In Plant tissue culture (ed. P. R. White & A. R. Grove), 449-462. McCutchan Pub. Corp., Berkeley, Calif.
FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS. 1962, Genetics of forest tree improvement, by J. W. Wright. FAO Forestry and Forest Products Studies, No. 16, 1-399. FAO, Rome.
FOWLER, D. P. 1965, Effects of inbreeding in red pine. Pinus resinosa Ait. II. Pollination studies. Silvae Genet., 14, 12-23.
GILMOUR, J.S.L. and J.W. GREGOR. 1939, Demes: a suggested new terminology. Nature, 144, 333.
GUFSTAFSSON, A. Apomixis in higher plants. I. The 1946 mechanism of apomixis. Lunds. Univ. Aarsskr. N.F. Avd. 2, 42 (3), 1-66.
HOUGH, A.F. 1967, Twenty-five-year results of a red pine provenance study. Forest Sci., 13, 156-166.
IRGENS-MOLLER, H. 1963, Douglas-fir archives and their use in tree improvement. FAO/FORGEN-63, 1, 3/8, 1-4.
LAMB, A.F.A. and E.N.G. COOLING. 1967, Exploration, utilization and conservation of low altitude tropical pine gene resources. Cyclostyled report (limited distribution) 1-30.
LANGLET, O. 1967, Regional intra-specific variousness. Proc. XIV IUFRO Congress, Munich, 3, 435-458.
LINES, R. 1967a, Standardization of methods for provenance research and testing. (Report of IUFRO Sec. 22 Working Group, compiled by Roger Lines) Proc. XIV IUFRO Congress, Munich, 3, 672-718.
LINES, R. 1967b, The international larch provenance experiment in Scotland. Proc. XIV IUFRO Congress, Munich, 3, 755-781.
MATSKEVICH, N.V. 1959, Rol' poliploidii v lesnoi selektsii. Akademiya Nauk S.S.S.R. Soobschheniya Laboratorii Lesovedeniya, No. 1, 70-81. (The role of polyploidy in forest selection. English trans., United States Department of Commerce, IPST Cat. No. 1327, 1965, 1-13).
MUIR, W.H. 1965, Influence of variation in chromosome number on differentiation in plant tissue cultures. In Plant Tissue Culture (ed. P.R. White &; A.R. Grove), 485-492. McCutchan Pub. Corp., Berkeley, Calif.
SARVAS, R. 1967, Pollen dispersal within and between subpopulations; role of isolation and migration in micro-evolution of forest tree species. Proc. XIV IUFRO Congress, Munich, 3, 332-345.
SCHRAMM, J.R. and E.J. SCHREINER. 1954, The Michaux quercetum. Morris Arboretum Bull., 5, 54-57.
SCHREINER, E.J. 1935, Possibilities of improving pulping characteristics of pulpwoods by controlled hybridization of forest trees. Paper Trade Jour., Tech. Sec., 100 (8), 105-109.
SCHREINER, E.J. 1949, Poplars can be bred to order. In Trees, Yearbook of Agr. 1949 (ed. A. Stefferud), 153-157. United States Government Printing Office, Washington.
SCHREINER, E.J. 1950, Genetics in relation to forestry. Jour. Forestry, 48, 33-38.
SCHREINER, E.J. 1953, Possibilities of inbreeding, selective intraspecific breeding, racial and species hybridization, and polyploidy. Northeast. Forest Tree Improve. Conf. Proc., 1, 64-68.
SCHREINER, E.J. 1966, Future needs for maximum progress in genetic improvement of disease resistance in forest trees. In Breeding pest-resistant trees (Proc. N.A.T.O.-N.S.F. Advanced Study Institute University Park, Pa., 1964), 455-466. Pergamon Press, London.
SCHREINER, E.J. 1967, Physiological and biochemical research on asexual reproduction in forest trees; an essential approach to early maximum genetic improvement. Proc. XIV IUFRO Congress, Munich, 3, 224-247.
TIGERSTEDT, P.M.A. 1967, Quantitative genetics in provenance research of forest trees. Proc. XIV IUFRO Congress, Munich, 3, 395-412.
TORREY, J.G. 1965, Cytological evidence of cell selection by plant tissue culture media. Plant Tissue Culture (ed. P.R. White & A.R. Grove), 473-484. McCutchan Pub. Corp., Berkeley, Calif.
VAN BUIJTENEN, J.P. and K. STERN. 1967, Marginal populations and provenance research. Proc. XIV IUFRO Congress, Munich, 3, 319-331.
WELLS, O.O. and P.C. WAKELEY. 1966, Geographic variation in survival, growth, and fusiform-rust infection of planted loblolly pine. Forest Sci. Monograph 11, 1-40.
