Assistant Professor of Forest Genetics, Yale University School of Forestry, New Haven, Connecticut, U.S.A.
In preparation for the World Seed Year, FAO is seeking to assist national campaigns aimed at improving seed quality and promoting wider use of such improved seeds. An important element in the production of better seed and superior trees is research relating to genetics, and particularly to forest tree improvement. The subject of forest tree breeding research is being treated by Professor Mergen in two parts. In the previous issue of Unasylva, he dealt with individual tree selection, progeny and clonal tests, and vegetative propagation. In this article he deals with hybridization, controlled pollination, and induction of flowers and fruits. Forest tree breeding research is potentially of the greatest significance to foresters and forest products industries. Interested readers are invited to send their comments to the Director, Forestry and Forest Products Division, FAO, Rome.
WHEN plant geneticists use the term hybridization, they mean all crosses between parents that are genetically unlike. Included are grosses within species (intraspecific) as well as crosses between species (interspecific). Foresters, on the other hand, are generally referring to interspecies hybrids when they discuss the place of hybridization in forest tree improvement programs. The term will be so restricted in this paper.
Hybridization, per se, has probably always occurred amongst forest trees, but has not been seriously studied until fairly recently. Natural hybridization between compatible species brings about a convergence or blending between the traits of the parent species and, if fertile F1 and F2 progenies are produced, it allows genes to flow from one species into the other. These hybrids occasionally also form a bridge or by-pass between two species that are not compatible.
Natural hybridization takes place when two compatible species grow under sympatric conditions (occupy same area) and when the flowering season of the two species coincides. Because the F1 hybrids differ from the parent species in morphological and physiological qualities, their site requirements may also differ. Consequently, although fertile seed and seedlings might be produced from the cross between two species, these new trees are able to survive and perpetuate themselves only if they can find a suitable ecological niche. If they are able to establish themselves, they either cross-pollinate between themselves, or back-cross to either or both of their parents, or to a possible third species. There is one natural hybrid pine in Mississippi, of which some of the open-pollinated progeny display the characteristics of Pinus palustris, P. elliottii, and P. taeda.
This type of cross-pollination is essentially at random, and in natural hybridization no man-directed selection takes place. It is purely a chance phenomenon with limited practical application, although it is of extreme interest to the students of evolution and taxonomy. The frequency of natural hybridization varies considerably with the genera, and is more prevalent in some than in others. A good example of natural hybridization can be found in the genus Quercus and, as an example may be cited the conditions in the State of Pennsylvania, U.S.A., where 17 species of oak occur naturally. Some 18 different natural hybrids between oaks that occur in this State have been described in the literature (Chisman, 1955). These crosses were all between parents belonging to the same group - white oak (Lepidobalanus) or red oak (Erythrobalanus), - for each group has a different post-fertilization cycle. Of interest also is a note by Syrach Larsen (1956) that the famous 800- to 900- year-old "twisted oak" (snoegen) in Denmark is apparently not a true Quercus robur, but a hybrid between Q. robur and Q. petraea.
Despite the large number of natural hybrids that have been identified and described, there are few instances where natural hybrids have appeared superior to either of their parents or where they have found their way into artificial planting projects. The best known instance of a promising natural hybrid is the cross between the Japanese larch (Larix leptolepis) and European larch (L. decidua). This natural cross occurred after Japanese larch trees had been introduced into Scotland where they were wind-pollinated by European larches. The resulting hybrid, Larix × eurolepis, as it was named by Dr. A. Henry, has rapid growth and good form, but above all it shows resistance to the larch canker, and is comparatively free from attacks by Chermes (Larsen, 1956). The natural hybrid Salix × coerulea (Cricket-bat willow) received its name by virtue of its wood that is particularly well suited to the manufacture of cricket bats (Schreiner, 1937).
Of interest is the feet that the recently described Picea × lutzii from the cross P. sitchensis × P. glauca from the Kenai Peninsula in Alaska, is being used for afforestation in Iceland. Although the hybrid seedlings do not possess any advantages over their parents in their native habitat, seed was collected from this hybrid in Alaska and the seedlings were raised in a forest tree nursery in Iceland (Baxter, 1956).
The outstanding natural tree hybrid is the red-flowered horse chestnut' Aesculus × carnea. This is a true-breeding natural hybrid between A. hippocastanum and A. parvia, which has twice the number of chromosomes of the parental species (Skovsted, 1929). This double set of chromosomes from each parent allows this amphidiploid hybrid to perpetuate itself, since crossing-over of random assortments will not produce any new combinations. Another hybrid that has been used in large numbers for ornamental purposes, although it has little importance as a forest tree, is the cross between Platanus occidentalis and P. orientalis (Rehder, 1937). This natural hybrid is known as the London plane tree (Platanus × acerifolia), and clones are used in both Europe and on the American continent.
In natural hybridization, man, as has been said, does not exert any guiding influence, except in those instances where non-native trees are introduced on their own merits and then cross-pollinate with compatible native trees, or when man's interference with nature has created ecological conditions favorable to either cross-pollination or to the establishment of hybrid seedlings. In controlled hybridization, on the other hand, trees with desirable characteristics are crossed in an effort to obtain offspring with the desirable traits of two or more parents or parental species.
Under these circumstances many of the conditions required for natural hybridization are of no consequence. The ranges of the two species do not have to overlap, the trees need not grow within wind-pollination distance of each other, the receptive period of one species does not have to coincide with the pollen shedding of the other, the seed is germinated under optimum conditions, a favorable environment is artificially created for the seedlings or trees, and even some inherent incompatibilities have been bypassed by embryo-culture.
In nature, most of the forest-tree genera have extensive ranges and some encompass the most diverse climatic and edaphic site conditions. A good example of this is found in the genus Populus whose members occur naturally throughout Europe, North America, North Africa, and in Asia, with its northern boundary extending to the timber line beyond the Arctic Circle. Another genus with which extremes of site conditions are encountered is Pinus. Pinus serotina grows in swamps, while the forests of P. clausa are located on dry, sandy soils. The variation in altitude within some species is just as great, ranging from sea-level to the altitudinal timber line.
Besides an ability to grow under the most severe site conditions, there are other great genetic differences in physiological and morphological conditions encompassed within a genus, such as resistance to a particular insect or disease, types of root systems, bole and crown shapes, or ability to perpetuate themselves by sprouting, just to name a few features. Theoretically, this natural genetic diversity of desirable traits presents good opportunities to breed almost any type of tree suited for a particular environment. The apparently unlimited possibilities, along with the outstanding performance of hybrids in agriculture, has caused some foresters to become overoptimistic and talk about super-hybrids that would solve most of their silvicultural problems.
The actual situation insofar as the place of hybridization in forest tree improvement work is concerned, is as follows. The results of past hybridization experiments have shown firstly what hybrid crosses can be made or are possible and, secondly, how these hybrids compare with their parents. In view of the large number of species and their world-wide importance, members of the genus Pinus have received the greatest attention amongst the gymnosperms, and members of the genus Populus, for equally apparent reasons, have been used most extensively in hybridization work amongst angiosperms.
Since 1845, when Klotzsch first attempted to produce artificial hybrids between forest trees (Klotzsch, 1854), sporadic reports of other attempts appeared in the forestry literature. It was not until 1924, however, that the first sustained efforts were made by the Oxford Paper Company, in Maine, U.S.A., to produce hybrids of members within the genus Populus (Stout, et al., 1927). During the existence of this program some 13,000 hybrid seedlings were obtained from crosses involving members of 34 species, varieties or hybrids of poplars (Schreiner, 1935). The early interest in poplar breeding was possibly also based on the spontaneous occurrence of two cultivated fast-growing poplar strains in France. Both of the cultivated poplar clones Populus × euramericana cv. eugenei and Populus × euramericana1 cv. robusta originated in the same nursery (Anonymous, 1968). The interest in poplar breeding has in no way slackened, as evidenced by the reports of the International Poplar Commission.
1 P. × euramericana is referred to as P. × canadensis in the USA.
In a recent publication by Schreiner (1959), The Production of Poplar Timber in Europe, the history and importance of poplars in Europe is reviewed. France, although leading Europe in poplar acreage (247,000 acres or 99,000 hectares), is surpassed by other countries in the productivity and intensity of its poplar culture. In the Po Valley of northern Italy, e.g., when planted on fertile, agricultural land and grown on a 12- to 25-year rotation, poplars can produce a greater cash return than other agricultural crops. Many of the poplars planted in this area are selected clones of natural hybrids between Populus deltoides which was introduced from North America around 1790 and the native black poplar (Populus nigra). These natural hybrids have been subjected to vigorous selection and progeny testing, and the outstanding work by Dr. Piccarolo and his associates of the Institute for Poplar Research at Casale Monferrato has led to the isolation of several clones that are especially well suited for planting in Italy. It is of interest that the majority of the clones that are planted in Europe are also cultivars of natural hybrids, e.g. Populus × berolinensis (Denmark, Germany, Norway, Sweden), and Populus × euramericana (Holland, France, Belgium, Germany, United Kingdom).
Poplars lend themselves well to genetic studies for they flower at a fairly early age, the flowers can be forced to maturity in the greenhouse, and all poplars can be propagated vegetatively in different ways. Currently, there are no less than 25 different institutes throughout the world actively engaged in research on the genetics and breeding of poplars. Hybrid poplars (Populus tremuloides × P. tremula and P. deltoides × P. trichocarpa) of improved growth rate over the native European poplar are available commercially in Europe and are being used widely in artificial regeneration, especially in Sweden, Denmark, and also in Finland. The results from poplar breeding have nicely demonstrated the necessity for adequate clone and progeny testing before claims can be made as to the superiority of a particular arose. Often the hybrids showed improved growth in the nursery beds, but in the out-plantings they failed to maintain this rapid juvenile growth and some formed stagnated stands. In addition, many of the hybrids are susceptible to various diseases, possibly as a result of an unbalanced physiological condition in the F1 hybrids. An example of this is found in the F1 progeny of Populus deltoides × Populus trichocarpa (P. × generosa) which is characterized by rapid juvenile growth but becomes susceptible, as the trees grow older, to the weeping canker. Research efforts are being directed towards finding resistant clones of this cross. The progeny tests also provided good statistical evidence that the growth performance of a species cross is highly variable and greatly depends on the individual parents used in a particular cross (Anonymous, 1958).
In 1925, the Eddy Tree Breeding Station, later known as the Institute of Forest Genetics, was organized at Placerville, California, and started work on the genetics of pines. The first task was to accumulate as many species of pines as was physically possible, so that the genetic potential or variability within this genus could be fully utilized in hybridization work. More than 40 F1 hybrids have since been produced at the institute (Righter and Duffield, 1951), many of which have been highly fertile, and some of these crosses have produced seed in large quantities. An annotated list of 23 of their most promising two-species and three-species hybrids was prepared by Duffield and Righter (1953), in which a digest of their performance is given.
The results demonstrate that the majority of hybrid progenies are not superior in growth rate to their parents, although some crosses have yielded trees that showed great promise. On the other hand, it must be conceded that a hybrid can have great economic importance without being superior to either of its parents, by virtue of its ability to grow under a wider range of climatic conditions or by possessing the desirable traits of the parent species. For example, the tri-species hybrid resulting from the cross (Pinus ponderosa × P. latifolia) × P. montezumae has, when grown at Placerville, California, the exceptional height growth of Montezuma pine, the early diameter growth of Apache pine and the local adaptability of ponderosa pine (Righter, 1955). Another interesting hybrid is the backcross of P. jeffreyi × P. coulteri to P. jeffreyi. The results from cage-tests indicate that this hybrid shows resistance to the pine reproduction weevil, a pest that has caused high mortality in plantations of ponderosa and Jeffrey pines in northern California (Miller, 1950).
The progeny from the cross between P. palustris × P. elliottii var. elliottii which was the first made by Philip C. Wakeley of the U.S. Forest Service in 1931, was outplanted for testing outside the natural range of both parent species in Placerville, California. The juvenile growth of this hybrid and its reciprocal which has been on many occasions since, is intermediate between both parents and does not have a "grass-stage" which, in longleaf pine, can last up to 20 years. Although not adequately tested as to progeny, this hybrid may be suitable for planting on dry sandy soils and on other sites (Figure 1).
Figure 1. - Hybrid between Pinus elliottii and Pinus palustris in a test plantation at Olustee, Florida. The tree was photographed at the beginning of the ninth growing season, its height being 7 meters, and the d.b.h. 13.5 centimeters. - Courtesy, U.S. Forest Service
In the Haploxylon pine group (white pines) most of the hybridizing, along with selection, is concentrated towards producing a suitable timber tree that is resistant to the white pine blister rust and the white pine weevil. The early performance of some of the crosses has been most promising and warrants greater efforts.
Controlled hybridization work with members of the genus Picea is of fairly recent origin, and in most instances data are available only on the nursery performance and on early survival in plantations. The results from the crosses give, however, some idea of the cross-ability pattern of this genus.
A total of 70 different species crosses have been attempted by tree breeders throughout the world. Of these 16 yielded viable seed, 12 were apparently successful, and 42 were complete failures. Of interest is the fact that of the 16 successful crosses, 14 of the species used in these crosses had neighboring natural ranges or were morphologically similar, or both. In the 42 crosses that failed, 37 of the species involved had ranges that were widely separate (Wright, 1955).
As mentioned before, the performance of the "Dunkeld" larches caused several tree breeders to collect open-pollinated cones from these natural hybrids or to repeat the cross. The progeny from some of these are superior in height growth to either or both Japanese and European larch, with no apparent change in form (Larsen, 1968). However, further work is needed with this cross using parents of known origin, along with adequate progeny testing.
After the chestnut blight had killed vast numbers of Castanea dentata trees in the United States, large-scale attempts were made to eradicate the disease by mechanical and silvicultural means. These costly efforts proved worthless, but new hopes were generated by the fact that the Asiatic chestnuts, C. crenata and C. mollissima, are resistant to this blight and can be crossed successfully with the American species. Despite the considerable amount of work that was carried out in both the United States and Europe, tree breeders have not produced a tested chestnut hybrid that is a suitable forest tree and is fully resistant to the blight (Figure 2). The work with chestnuts has been a particularly difficult one because the American chestnut did not possess a blight resistance factor in any one of its members, the Asiatic chestnut trees have a poor form, and the seed set from controlled pollination has been low. This breeding work has produced, however, some promising hybrids from the cross between C. mollissima and selected trees from the F. progeny of C. crenata × C. mollissima. Some of the resulting hybrids, although not suitable for timber trees might find their way into commercial nut orchards (Nienstaedt and Graves, 1966).
FIGURE 2. - Control pollinating Chinese chestnuts in a plantation at Hamden, Connecticut. The trees with the drooping catkins are hybrids between Castanea cremata and Castanea dentata. - Courtesy, B.W.McFarland, Connecticut Agricultural Experiment Station, New Haven, Conn., U.S.A.
Besides the genera mentioned, controlled hybridizing is being done with trees of the following important forest tree genera: Abies, Araucaria, Tsuga, and Acer, Betula, Eucalyptus, Fagus, Quercus, Tilia, Fraxinus, Ulmus.
Once an outstanding hybrid has been produced on an experimental basis and has shown its superiority in progeny or clonal tests, it needs to be mass-produced for use in commercial plantings. This can be done by either vegetative propagation of the desired individuals, by repeating the cross, or by producing amphidiploid hybrids. If biologically feasible, the quickest way will be to repeat the cross. If vegetative propagation has to be used, clonal tests are necessary before a particular clone can be recommended. There might be some instances, however, where only a few individuals in an F. population are of the desirable type, in which case clonal propagation is advised.
The main shortcoming of hybridization attempts with forest trees so far is that the resulting hybrids are too young to permit drawing any reliable conclusions. Before one can state that a particular cross has hybrid vigor, or heterosis, adequate replicated progeny or clonal testing should be done within the natural ranges of both parents. The policy of the Institute of Forest Genetics at Placerville has been to send out quite large quantities of hybrid seed to reliable cooperators throughout the world, so that these hybrid trees might be tested under varied conditions.
It is also often difficult to locate trees for hybridization experiments that have an authenticated history. The result from several crosses have shown that the genotype of the trees used greatly affects the offspring. Such factors as geographic and altitudinal ecotypes must therefore be taken into account in these trials.
However, the results from hybridization work to date has laid a good foundation for future programs. It has shown that hybridization on a fairly large scale is feasible with forest trees and that, in some instances, hybrid seed can be produced at economically feasible prices. There are reliable performance records in regard to a number of hybrids, and in many instances they have shown great promise: some were superior to either or both of their parents in growth rate, and resistance factors to certain insects and diseases were also present in some of them. Also the cross-ability patterns have yielded much valuable scientific knowledge and allowed the revision of the systematics of some forest tree genera, especially within the hard pine groups (Duffield, 1952).
Controlling the pollen source during the pollination of forest trees is now an accepted practice throughout the world. Its purpose is to present undesirable wind-borne, or insect-borne, pollen from reaching the receptive female flowers of certain selected individual trees, or groups of trees. The techniques are quite variable and depend on the intensity and goals of the various programs and on the personal preference of the tree breeder. They can be divided into two main categories: cross-pollination under natural but controlled conditions, or pollination by hand.
There is currently much discussion on the establishment of seed production areas and seed orchards to supply forest tree seed of good quality. The seed production areas have been discussed under the section on tree selection (Unasylva, Volume 13, Number 2, pages 81 to 84), and need no further mention at this point. Seed orchards, on the other hand, are, by definition, plantations of trees that are genetically superior, and that are managed for the sole purpose of producing high (genetic) quality seed. These orchards resemble apple orchards to a greater degree than a forest stand, for the trees are planted at a wide spacing and they will be cultured so that they have large cone-bearing crowns. They will be established either with clonal material such as cuttings or grafts, or with seedlings.
Because the operation of seed orchards is new to forestry, many problems need to be worked out, such as: minimum number of clones per orchard to prevent inbreeding depressions in the resulting progeny; width of an isolation strip to prevent stray pollen from reaching the flowers; type of fertilizer to use; arrangement of clones to obtain a maximum amount of cross-pollination; protection against insects, disease, rodents and birds; and many other problems more common to agriculture than to silviculture. Since the inception of seed orchards, sizeable areas have been planted for this purpose. The first seed orchard in the United States, with grafts from proven genetically superior trees was established with Pinus elliottii for the production of trees for high yields of naval stores products (Merger, et al., 1955). It has, however, become practice to establish seed orchards with clones from superior phenotypes before the genetic superiority of these trees has been proven, and while the progeny tests are being made concurrently. In these instances, individuals that do not show any superiority in the progeny tests will be cut out.
Many of the ideas and concepts on seed orchards originated in the Scandinavian countries. The Society for Practical Forest Tree Improvement in Sweden has been a leader in the establishment large-scale orchards, for seed of Pinus sylvestris and Picea abies. Arnborg (1956) estimated that the Scots pine orchards will start to produce seed in sizeable quantities when the orchard is 10 years old, and that each acre will yield enough seed at age 20 for about 300,000 seedlings.
The estimate by Dr. Larsen (1956) for larch, based on the performance of grafts, is that each acre will produce seed for 360,000 trees. In the United States, seed orchards of Pinus elliottii, P. taeda, Pseudotsuga menziesii, and Tsuga heterophylla. among others, are being established at an almost alarming rate. In the State of Georgia, for instance, a program has been initiated to establish 400 acres of seed orchards to supply part of the annual seed requirements for 350 million seedlings (Figure 3). The orchards are established by grafting the scions either directly on stock in the fields or the forest owners collect the scions from their selected trees and ship them to a central grafting place.
FIGURE 3. - Overall view of a newly established seed orchard of Pinus taeda of the Georgia Forestry Commission. The scions from selected trees were grafted directly onto seedlings in the field. The spacing between trees is 16 × 16 feet (4.87 × 4.87 meters). All the selected trees are being progeny tested while the orchard develops, and the poorest genotypes will be removed in future thinnings. - Courtesy, Georgia Forest Research Council, Georgia Forestry Commission, and U.S. Forest Service
After unions are formed, the grafts are outplanted in the field. Besides providing seed of superior genetic quality, some forest companies are establishing seed orchards solely to assure adequate seed supplies of average genetic quality for their accelerated reforestation programs.
Another type of seed orchard was proposed by Righter (1946) of the Institute of Forest Genetics for the mass production of pine hybrids. He argued that, because genetic uniformity is not required or even desirable in silviculture, a variable F2 hybrid progeny might be used advantageously. Its seed could be produced on a mass-scale in seed orchards planted with F1 hybrids that are allowed to interbreed amongst themselves.
The controlled pollination practices described so far have dealt with entire trees or groups of trees, and are used mainly in the mass-production phase of tree improvement programs. Before this step can be undertaken, however, a great deal of hand pollination by the forest geneticists is required. When trees are hand-pollinated, the female flowers are isolated with pollination bags, and the selected pollen is then placed on these flowers when they become receptive (Figure 4).
FIGURE 4. - Process of control pollinating an F1 generation of Pinus elliottii in an 8-year-old progeny testing plantation at Olustee, Florida. The ladder equipped truck allows rapid movement from tree to tree without damage to the tree crowns. - Courtesy, U.S. Forest Service
Various types of bags have been used, e.g., glassine and polyvinyl bags, viscose sausage casings, Kraft paper and canvas bags, or plastic tubes (Figure 5). Plastic sausage casings have been widely accepted by tree breeders because they allow continuous observation of the development and condition of the flowers, air and moisture can pass through the walls, and the material is strong enough to withstand rough treatment. Recent tests, however, have shown that the temperature inside bags that are exposed to full sunlight can rise to a point where it may damage the pollen. This undesirable effect can be overcome quite easily by spraying the exposed side with white or aluminum paint.
FIGURE 5. - Four types of isolation bags in a test on Pinus elliottii at Olustee, Florida. They are, in a clockwise direction: Kraft paper, canvas bag with plastic window, clear plastic, and viscose sausage casing. The black tape on the bags closes the puncture made by the needle of the pollen injector. - Courtesy, U.S. Forest Service
Sometimes, when abundantly flowering trees are being used, it is expedient to isolate entire trees with tents. At Ekebo, Sweden, tents made with seven sections of a parachute are used. One of the advantages of this is the fact that the same tent can be used on several species of trees having successive periods of flowering (Johnsson, 1953). At the Petawawa Forest Experiment Station, at Chalk River, Ontario, a tent mounted on a pole has given good results with Picea glauca. The tent is manufactured of canvas, and is held open by hoops, and several hundred female flowers can be isolated in this manner.
When female flowers are isolated it is necessary that the male components be removed beforehand to prevent contamination from the tree's own pollen. Isolated trees are occasionally found that have only female flowers, and Wright (1955) used some of these trees without isolation bags in his spruce hybridization program. If a particular hybrid seedling shows enough variation from the nonhybrids, this technique is possible, and apparent contaminants can then be weeded out in the nursery beds.
When the female flowers become receptive, pollen is either dusted or blown on the flowers. The pollen is collected beforehand, and considerable quantities are sometimes necessary, especially when the bowers of an entire tree crown are isolated. For the conifers, in general, it is customary to collect the pollen from tree-ripened microsporangiate strobili (male flowers), while, for some of the hardwoods, forcing of male flowers in the greenhouse gives good results. With such species as aspen or poplars that ripen their seed some four to seven weeks after the flowers are pollinated, the entire process can be carried out with good results in the greenhouse. Flower-bearing branches are cut during late winter and forced into flowering by placing their cut bases in water and keeping them in a heated greenhouse. Viable pollen will shed from the male flowers and is dusted on the receptive female flowers; after a period of several weeks, the ripe seed capsules can be harvested.
The largest hand pollination program ever attempted was at the Institute of Forest Genetics at Suwon, Korea. Under the direction of Dr. Hyun about 60,000 ovulate strobili of Pinus rigida were isolated, using some 30,000 isolation bags. Approximately 1,200,000 hybrid seedlings of P. taeda × P. rigida resulted from this operation, at an average cost of $6.00 per 1,000 hybrid seedlings. This program is being continued and expanded.
Pollen from some species, such as the spruces, will remain viable during a period of one to several years if stored under the proper conditions of temperature and humidity. This long period of viability makes it possible to cross species that flower at different periods of the year. If stored pollen is used, it is advisable to make germination tests beforehand to avoid the use of dead pollen. Currently, there is a wide exchange of tree pollen between tree breeders throughout the world, for air transportation provides a speedy vehicle of exchange without any loss in viability during transit.
The time interval between pollination and seed shedding presents ample opportunities for various insects, fungi, rodents, and climatic factors to decimate the potential seed crop. This is of especial concern for members of the genus Pinus where a period of about three years elapses from the time the flower primordia are laid down until the seed ripens in the cones. Insects have been combatted by spraying with D.D.T. or B.H.C., and during this past year spraying with fungicides from an airplane was used to protect the flower crop in a seed production area of P. elliottii in the State of Georgia (Cole, 1958).
It has become an accepted procedure with some species either to leave the isolation bags over the flowers or to place a special bag over the developing cones or nuts (chestnuts) to protect them from dropping to the ground: this also minimizes the damage from rodents and birds.
The rate at which the science of forest genetics will advance is dependent on the time it takes to evaluate progeny tests, and on the interval between successive generations. To make controlled pollinations, flowers of both sexes are a prerequisite. But the general field of flower induction and fruit formation in forest trees has been neglected until recently, and considerable basic research is needed.
Until foresters started to breed forest trees, few people had closely observed their flowering and fruiting behavior. Most tree flowers, although quite interesting when examined closely, are not showy when viewed from a distance and, being mostly in the upper portion of the crown, they do not attract much attention. As a result, little or no precise information is available for the various species on the minimum age when flowers appear, on the ratio of male to female flowers, and on the factors that control flowering in general. A record of early flowering among 57 species or varieties of pines growing in the pinetum at Placerville, California, was published by Righter in 1939. The average minimum age at which staminate flowers appeared was 4.4 years, and 6.2 years for the ovulate flowers. One species, Pinus tabuleformis bore male flowers during the first year. The writer has collected viable pollen, as tested by actual germination, from three Mugo pine seedlings (P. mugo var. mughus) that were less than one year old. The reason for the precocious flowering in some trees is not known.
Numerous attempts to induce flowers experimentally in one- or two-year-old trees have been only moderately successful. In somewhat older trees that were closer to the normal flowering age, however, various techniques have yielded good results, for instance the addition of N.P.K. fertilizers to the soil, either singly or in mixture. These experiments have not so far covered a large enough variety of soils and species to allow any general recommendation to be made but one can deduce from past examples that almost any addition of a balanced fertilizer will improve the vigor of the tree and stimulate flower production. Besides this stimulation with nutrient elements, physical injury to the roots or stem has been successful in some cases. Such treatments as root pruning, stem strangulation, and various degrees of girdling have increased the flower crop. These treatments, although of a physical nature, affect the normal translocation of organic assimilates and have a tendency to accumulate the photosynthates above the injury.
Sax (1957) obtained significant flowering effects by inverting the bark in apple trees; this technique, however, has failed so far to yield results with forest tree species tested. Mirov (1951) was able to induce flowering during the third growing season, in five species of pine by grafting young seedlings into the crowns of a P. ponderosa tree that was flowering heavily. This technique has since been tried repeatedly without, however, giving consistent success. The external application of hormones or auxins has as yet failed to bring about flowering in a replicated experiment. Fruit induction, without viable seed, as a result of treatment with auxins is economically sound in the field of agriculture where the fruit is the principal crop, but this technique has no place in forestry where the only concern is viable seed. It might be pointed out, however, that berries will form from unpollinated female flowers on holly trees if they are sprayed at the proper time with indoleacetic acid.
If the sex of flowers could be altered at will, it would also help the progress of the hybridization experiments. Little work has been done along these lines, but Dr. Saito (1957) was successful in controlling artificially the sex differentiation of P. densiflora and P. thunbergii by injury to the shoot or by spraying with aqueous solutions of either alphanaphtaleneacetic acid or 4-dichlorphenoxyacetic acid. In each instance sex transition was from a rudimentary male strobilus to a female strobilus.
Controlled hybridization of forest trees has helped to advance considerably scientific forestry knowledge. Motivated usually by practical considerations, it has stimulated much basic work on the physiology, cytology, and general behavior of forest trees. The evidence obtained so far on interspecies hybridization should be most encouraging to the practicing forester, for it has shown that the desirable traits of two or more species can be combined in one individual. This approach to obtaining trees better adapted to silvicultural treatments will, if combined with other methods of improvement, such as selection or treatment with mutagenic agents, always deserve a prominent place.
ARNBORG, T. 1956. Pine seed orchards of SCA. Medd. nr. 82 fran Sällskapet för praktisk skogsförädling. Uppsala.
BAXTER, D. V. 1956. Skogar saga. Pacific Discovery 9 (3): 9-14.
CHISMAN, H. H. 1955. The Natural Hybrid Oaks of Pennsylvania. Penn. State For. School, Res. Paper 22, 5 pp.
COLE, D. E. 1958. Aerial application of benzene hexachloride for the control of cone insects on a slash pine seed production area. Journal of Forestry 56. 768.
DUFFIELD, J. W. 1955. Relationships and species hybridization in the genus Pinus. Zeitschrift f. Forstg. u. Forstpflztg. 1: 93-100.
DUFFIELD, J. W. and RIGHTER, F. I. 1953. Annotated List of Pine Hybrids Made at the Institute of Forest Genetics. U.S.D.A. Cal. For. Range Exp. Sta.; For. Res. Note 86.
JOHNSON, H. 1953. Methods of isolation - Parachute tent. Tree Genetics News Letter, Ekebo, Sweden. 2 (2): 3-4.
KLOTZSCH. 1854. Die Nutzanwendung der Pflanzenbastarde und Mischlinge. Ber. Bekanntm. Verhandl. K. Preuss. Akad. Wiss. Berlin. 535-562.
LARSEN, C. S. 1956. Genetics in silviculture. Oliver and Boyd, London.
MERGEN, F., HOEKSTRA, P. E., and ECHOLS, R. M. 1955. Genetic control of oleoresin yield and viscosity in slash pine. Forest Science 1: 19-30.
MILLER, J. M. 1950. Resistance of Pine Hybrids to the Pine Reproduction Weevil. U.S.D.A. Cal. For. Range. Exp. Sta.; For. Res. Note 68.
Mirov, N. T. 1951. Inducing Early Production of Pine Pollen. U.S.D.A. Cal. For. Range Exp. Sta.; For. Res. Note 80.
NIENSTAEDT, H. and GRAVES, A. H. 1955. Blight Resistant Chestnuts. Conn. Agric. Exp. Station; Circular 192.
REHDER, A. Manual of Cultivated Trees and Shrubs. 1937. Macmillan Co., New York.
RIGHTER, F. I. 1939. Early flower production among the pines. Journal of Forestry 37: 935-938.
RIGHTER, F. I. 1946. New perspectives in forest tree breeding. Science 104 (2688): 1-3.
RIGHTER, F. I. 1955. Possibilities and limitations of hybridization in Pinus. U.S.D.A.; Southeastern For. Exp. Sta.; Proc. 3rd Southern Conf. on For. Tree Improvement. 54-63.
RIGHTER, F. I. and DUFFIELD, J. W. 1951. Interspecies hybrids in pines. Jour. Hered. 42: 75-80.
SAITO, Y. 1957. Artificial control of sex differentiation in Japanese red pine and black pine strobiles. J. Faculty Agr., Tottori Univ., Japan 3 (1).
SAX, K. 1957. The control of vegetative growth and the induction of early fruiting of apple trees. Amer. Soc. Hort. Sci. Proc. 69: 68-74.
SCHREINER, E. J. 1937. Improvement of forest trees. U.S.D.A. Yearbook. 1242-1279.
SCHREINER, E. J. 1959. Production of poplar timber in Europe and its significance and application in the United States U.S.D.A. Agr. Handbook 150. 124 pp.
SKOVSTED, A. 1929. Cytological investigations of the genus Aesculus L., with some observations on Aesculus carnea Willd., a tetraploid species arisen by hybridization. Hereditas 12: 64-70.
STOUT, A. B., MCKEE, R. H. and SCHREINER, E. J. 1927. The breeding of forest trees for pulpwood. Jour. N.Y. Bot. Gard. 28: 49-63.
WRIGHT, J. W. 1955. Species cross-ability in spruce in relation to distribution and taxonomy. For. Science 1: 319-349.
The author wishes to express sincere thanks for the cooperation received from the various agencies that supplied illustrations for this article. The manuscript woo critically reviewed by Professor N. J. Lutz of the Yale School of Forestry, New Haven, Conn.; Professor J. W. Wright of the Department of Forestry, Michigan State University, East Lansing, Mich.; and Dr. E. B. Snyder of the Southern Institute of Forest Genetics, U.S. Forest Service, Gulfport, Miss. Their suggestions were most helpful and are hereby acknowledged with thanks.