1. All trees to be cut must be individually marked.
2. Only mature and dead timber may be removed.
3. All slash remaining after cutting must be removed from the forest.
   

"Conservationists" hailed the court decision as a landmark victory against clearcutting, but it is not difficult to visualize that if the order were eventually extended beyond the Monongahela boundaries it could set sound forest management back several decades and reduce the forester's role to one of a caretaker rather than the professional trained in managing and improving the forest resource. In situations where preservation of existing forest growing stock, often of inferior composition, becomes the guiding principle, the timber supply levels will not quickly improve.


INCREASING TIMBER SUPPLIES- A VITAL ROLE FOR TREE IMPROVEMENT

ABOVE: Loblolly pine is an ideal species to work with in tree improvement. It is fast growing and has desirable wood qualities. It is from stands such as this of Georgia-Pacific Corporation in South Carolina that select trees are chosen for use in seed orchards.

As implied earlier, improved utilization can augment and extend timber supplies both as an interim measure and in the long pull. However, the major long-term reliance must rest obviously and squarely on increased productivity of all available commercial forest land. This objective is not readily nor easily achieved. The problems encompass several major factors which are complex and separate in a sense, yet so closely interrelated that rendering solutions to them must not overlook possible vital interactions. Thus, they must be both broadly and specifically understood, and solutions to them cannot be indefinitely deferred.

The broader aspects of these problems have been analyzed in a penetrating way by Vaux (1973), with particular emphasis on land productivity in California. My purpose here is to discuss one major method for increasing the yield of wood from forest land, My thesis is simple; and it is not new. It involves the use of better trees, stressing the gains and economic impact of tree improvement programs on forestry operations in the Southeastern United States.² While stressing the role of tree improvement, I want to make it crystal clear that no single method of increasing yields can be successful without integration of other major factors. For example, without a "marriage" or proper balance of site preparation, fertilization, and genetically improved trees, efforts are likely to fall far below expectations, or may indeed be doomed to failure. The best genetic stock can never produce its full potential if inadequate site preparation is used, or if the soils have critical nutrient deficiencies. It makes no sense to spend large sums on thorough site preparation if scrawny planting stock is used, as has often been done in the past. Our studies show some pines to

ABOVE: Select trees used as parents in the seed orchards must have many good attributes. Shown is one of the best labially pine parents of Georgia Kraft Corporation. Such trees are selected and bred to have a wide genetic base with wide adaptability, a broad resistance to pests, and a narrowed base for faster growth, straight boles, small limbs and good wood. There are 3500 trees like this established in operational seed orchards. Trees having such a combination of desired characteristics are hard to find.

be unresponsive to fertilization or even to be inhibited in growth with fertilizer application (Zobel and Roberds, 1970). Obviously it is utter nonsense to use expensive fertilization if the planting stock is unresponsive or grows more slowly. These few examples are used to emphasize that the following discussion of improved trees includes results obtained from genetically good stock, performing in response to a combination of good management practices.

The stage for my discussion of increasing productivity of forest land was well set by the first S. J. Hall Lecturer, Harry Morgan, in 1969, who stated that the biggest asset of a forest industry is not the mill but rather the timberland and the existing inventory of timber on that land. Mr. Morgan emphasized the absolute necessity of the long-term outlook coupled with intensified management practices. Most impressively, be emphasized the immediate need of such improvement; this urgency has been difficult to project to administrators in the forest industry, even in the South, where intensive management holds such high promise and is so urgently needed.

The objective of our Cooperative Tree Improvement Program is to get as much improvement as possible as quickly as possible. Each unit of improvement, even if individually small, is magnified many times when gains are reflected in a planting program which comprises 400,000 acres a year. We are interested in gain per unit of time; the maximum gain for a characteristic may require a long time to develop, so we often use lesser but more quickly achieved gains to optimize total improvement per unit time on a given land area.

The genetic improvement of forest trees is a long-term, expensive undertaking. It can be done best through Cooperative efforts because most organizations cannot afford a team of highly trained specialists. In a Cooperative, one trained man can oversee a great deal of research. The need to keep the genetic base sufficiently broad is almost impossible for a single organization but is easily achieved in a cooperative effort. The funds and manpower required for a tree-improvement program for each member is minimal in a cooperative, yet enables maximum genetic gains. Plant materials, methods, equipment and even manpower are exchanged amongst members to the benefit of all. In my view, naturally biased, it would seem wasteful and even foolish for each organization to strike out on its own with an expensive, inadequate, and inefficient program when faster and greater gains are assured through joint action. The Florida, Texas, and North Carolina Cooperatives are convincing demonstrations of how well combined action has succeeded.

Tree improvement programs are suitable for both conifers and hardwoods. Although all examples used in this paper will be from the pine program, we also have a hardwood cooperative with 16 industries, one state forest service and one Hardwood Research Council. Less is known about hardwoods than about pines, but our preliminary studies and experience indicate that good genetic gains will be possible for the hardwoods. Because of the large number of species with different and complex breeding systems, hardwoods are more difficult to manipulate; nevertheless, substantial improvement, especially for quality hardwoods, appears promising.


Objectives of a Breeding Program

Time permits only a quick survey of characteristics to be emphasized in genetic improvement of pines; major attention is directed broadly toward (1) adaptability, (2) resistance to pests, (3) growth rate, (4) tree-form and quality, (5) wood qualities. Not all are equally emphasized by every individual organization and they vary depending on management objectives, land base, and end products; generally every well-balanced tree breeding program includes something of all five.


ADAPTABILITY

A first requirement of any successful tree breeding program is the development of planting stock adapted to the environment in which it is to be grown to marketable dimensions. Lacking proper adaptability, improvement in all other characteristics will be diminished in effectiveness and value.

If one is working with exotics or with native species of wide geographic range, a primary consideration of adaptability is provenance or geographic source, including elevational, soil, and climatic influences. It is often assumed that there are few provenances within the southern pines, but in fact wide-ranging species such as loblolly pine (Pinus taeda) exhibit large and important differences within the species. Florida sources freeze and easily succumb to drought while the northern sources suffer from the high temperatures, especially high night temperatures, in the Deep South. There are clear-cut differences between Coastal and Piedmont loblolly pine sources; for example, the latter flower very heavily and early, no matter Where planted. We know that southern and coastal loblolly pine sources grow faster than those from inland or from the north; in contrast, Piedmont sources are more tolerant to drought and cold. Our basic rule is to use local sources initially while concurrently conducting tests to find how far the inland and coastal sources can be safely moved.

Improving adaptability to adverse environments is a prime objective of our breeding program. As pressures for land me become greater, forestry is being "pushed" from the more fertile soils to marginal and submarginal sites, many of which now lack capability of producing merchantable timber crops. If forestry in the South is to attain its full potential, the poor sites must be made economically productive. Results from breeding for better adaptability are showing excellent promise; we now have seed orchards with strains of loblolly pine already in production specially suited for extra wet sites as well as for very droughty sites. Also, we have cold-resistant loblolly and one strain has proven especially fast growing on the basic marl soils. Hybrids with pitch pine have shown promise in tests north and west of the species range, enabling the afforestation of poor upland hardwood sites with this conifer.

ABOVE: Seed orchards are established from the outstanding parents to produce genetically improved seed on a commercial scale. Members of the Cooperative obtained enough seed orchard seed in 1973 to plant 150,000,000 trees (enough to plant 200,000 acres). Photo on left shows a young (5-year-old) loblolly pine orchard of Continental Can Company just coming into commercial seed production. On the right is a mature (12-year-old) loblolly orchard of Union Camp Corporation in Virginia which is producing large amounts of seed.

A most important result of the genetic programs on loblolly pine has been, the widening of the genetic base for adaptability while at the same time narrowing the base for desired qualities such as straightness, last growth, and wood properties. Broad adaptability has been achieved by rigorously adhering to a policy of establishing only nonrelated individuals in our seed orchards. Each parent tree used in a seed orchard comes from stands sufficiently spatially separated to preclude the possibility of "kinship". This practice is in sharp contrast with many commercial seed collections which are usually from few stands containing many closely related individuals. Because extensive portions of the current loblolly pine forest area were one time farmed for cotton and other crops, naturally regenerated loblolly stands in most localities originated predominantly from a very few parent trees growing along roads, field and fence lines, streams or around farm buildings. The resulting stands contain many closely related individuals which now are in a second or third generation of matings among themselves.

ABOVE: Not all parent trees produce good progeny, so orchards need to be rogued of poor parents. Shown is the removal of such a parent in the loblolly orchard of the Kimberly-Clark Corporation in Alabama. The clone being removed produced average progeny that were very disease susceptible.

This simple historic fact of a high probability of a narrow genetic base underlying a large number of our existing naturally regenerated loblolly pine stands has been generally overlooked, particularly by ardent preservationists and neo-conservationists who envision dire hazards and doom from pine "monocultures" generated from tree improvement efforts. To those of us in the thick of forest genetics programs, it is all the more surprising when even professors in biology departments of universities join in singling out tree improvement as the culprit in current epidemic insect infestations. Illustrative of this sort of nonsense are the allegations made by a biology professor in a university in Alabama given multicolumn coverage in an Alabama daily newspaper. Basically, the allegations deplored the narrowing of the genetic base through forest tree improvement, totally ignoring the fact that the insect attacks were occurring in older natural stands, not in the limited acreage of new plantings from seed orchard sources. The same article went on to the ridiculous extremes of blaming the littleleaf disease of shortleaf on forest genetics, again failing to observe that this species has not yet reached the stage of any planting of improved sources of seedlings. Finally, the same article alleged that forest geneticists were compounding the errors of cotton breeders and heading toward a boll weevil-type catastrophe, again totally failing to distinguish between the vastly different breeding systems employed in pine improvement in contrast to those in cotton. Suffice it to say that from the very beginning of the Tree Improvement Program, a major objective has been to develop broadbased, adaptable, vigorous strains of trees; this we have attempted to achieve through wide exchange of material, use of wide crosses within species, and strict avoidance of related individuals among seed orchard parent sources. Our work will be made easier when concerned critics develop greater understanding not only of forest genetics but of forest history as well.

The greatest over-all gains from the genetics programs in southern pines will result from a broader adaptability obtained by a wider and more variable gene base than could ever occur with natural regeneration or from seed collected from wild stands.


RESISTANCE TO PESTS

The worst enemy of slash and loblolly pines is fusiform rust (Cronartium, fusiforme) which causes hundreds of millions of dollars of loss. Luckily, rust-resistant trees are being found among seed orchard parent trees selected from wild populations of these species, and resistant strains are being propagated.

One of the most intensive research efforts on fusiform rust in forestry is now underway and involves a multiple attack by private, state, and federal agencies working in a closely coordinated program. A number of resistant pine families have been found, but variability in the rust itself makes the breeding program difficult. Even so, results to date on breeding for rust resistance have been spectacular, though much still remains to be learned.

Disease resistance information and techniques have been put to immediate use with over 200 acres in a dozen specialty rust resistance seed orchards having been established, producing seed in commercial quantities. Every economic analysis of tree improvement has shown that resistance to rust is of key importance if optimal gains are to be achieved in areas where rust is prevalent, and the gains can be considerable (Blair and Zobel, 1971; Zobel et al., 1971a) (Table 1). Special short-term methods of quick testing are being developed by the U.S. Forest Service through relating field infection to greenhouse test results. If acceptable gains are to be achieved from selection and breeding for disease resistance, thorough progeny testing is mandatory.

* From Blair, R. L. and Zobel, B. 1. 1971.
** Based upon an index which reflects the potential economic and biological impact as well as the incidence of the disease.

GROWTH RATE

ABOVE: Growth of seed orchard seedlings has been good as is shown by this 11-year old plantation of Westvaco Corporation in South Carolina. The planting was thinned once at 7.5 years of age and again at 11 years. Growth rate has averaged over 3 cords per acre per year and is increasing.

A common misunderstanding is that growth rate must be sacrificed to obtain better formed, disease-resistant trees possessing desired wood properties. This stems from overlooking the simple fact that all improvement we obtain is from selections within the fastest growth segment of the population. It is extremely fortunate that most desired characteristics are not related closely to growth rate, thus enabling gains to be achieved by breeding for each independently without loss of quantity production. Even the often stated negative relationship between growth rate and wood specific gravity is not evident for older trees. In fact, as the trees emerge from the juvenile stage and competition sets in (age 8-10 years), the faster-growing genotypes are often found to have the most dense wood (Matziris and Zobel, 1973; Zobel, et al., 1972; Stonecypher, et al., 1973), even though at younger ages the faster-growing trees did have less dense wood (Stonecypher and Zobel, 1966).

All genetic programs have improvement in yield as a major objective, and the most important component of yield is growth rate. We find, conservatively, that plantations from seed orchards give improved yields of 10 to 20 per cent over normal planting stock. As the orchards are rogued of the poorest parents, yields will be increased an additional 5 per cent. When only the very best two or three parents from each seed orchard are brought together in new seed orchards (commonly referred to as 1.5-generation orchards) at least another 5 per cent improvement is obtained. Gain predictions obtained from basic studies in unselected populations of loblolly pine showed a 25 per cent improvement in volume growth and 26 per cent in dry weight (Stonecypher, et al., 1973). There are many reports in the literature of gains in volume growth of 50 per cent or more, but at this stage we prefer to remain conservative and continue to cite 10 to 20 per cent as a readily defensible expectation (Table 3).

Because of the difficulty of selecting for volume in wild stands, our program has emphasized quality characteristics in first-generation seed orchards. It has been a most pleasant surprise to obtain the 10 to 20 per cent volume gains noted earlier because selected parent trees had only to be as large as the largest trees in the surrounding stand, and intensive volume superiority was not required.

* Roguing consists of removing from the seed orchard those parent trees or clones that produce undesirable progeny.
** A good general combiner is a parent that produces good progeny no matter how good or poor the other parent.

Although the values for young stands can be misleading, Union Camp Corporation obtained 14 per cent greater volume growth from seed orchard stock compared to commercial stock for 12 paired 5- and 6-year-old plantations. Kimberly-Clark Corporation found good superiority from 9-year-old plantations (Table 2). No one should guarantee this magnitude of gain to be maintained as stands get older, but present indications suggest that gains in older stands will be large.

In the second-generation selections we expect an addiitional 25 per cent gain, based on known heritabilities and selection intensities.³ Primary emphasis in selection for second-generation orchard parents is for volume, an efficient operation since all trees are the same age, planted at the same spacing on relatively uniform sites. Good initial gains in volume growth with selections from wild stands were not expected because growth is such a complex chat. acteristic. Our estimated volume gains for different kinds of seed orchards are shown in table 3; gains for orchard types 1-4 have already been fairly well confirmed by field tests.


TREE FORM AND QUALITY

We have been criticized for breeding for quality because "you get as much wood from a crooked as from a straight tree, from one with large limbs compared to one with small limbs." Even though the volume of bole wood may be the same, the quality and yields of desired paper or boards per unit volume of timber are reduced in a crooked or a large-limbed tree. When this loss is combined with added costs of harvesting and manufacturing crooked and large-limbed trees, breeding for improvement in tree form proves to be well worthwhile. An additional benefit from trees of better form throughout much of our area is their better resistance to ice and wet snows which can often be catastrophic.

It was expected that tree and bole form would respond well to genetic manipulation; in this achievement we have not been disappointed. Most responsive has been bole straightness (Shelbourne, 1969): we usually find that desired straightness can be obtained in one generation of breeding, thus freeing us for concentration on other characteristics in future generations.

Response to limb characteristics is less dramatic but yet worthwhile (Von Wedel, et al., 1967). Evenness of limbs and limb size are improved and the second generation selections are very outstanding in having beautiful small-limbed crowns, even though they are of very fast growth. Certainly, limb and crown form have been changed enough through genetic manipulation to have a significant effect on harvesting costs and quality and yield of the final product. For example, in a special pulping study from the heritability project in which trees of various combinations of bole straightness and limb size were selected, it was found that the straight small-limbed trees gave better yields and the paper produced had better resistance to tear than did paper from wood of crooked large-limbed trees. The interesting relationship that straighter trees have greater wood density, whether the trees are fast or slow grown, will be of considerable value if additional studies confirm initial indications (Blair, et al., 1974). It is evident that breeding for better tree form will have a strong economic impact.


WOOD QUALITIES

The most consistent genetically responsive characteristic has been wood specific gravity. This complex characteristic often called wood density, is of key importance in determining the quality of end product, whether it be for paper or solid wood products (Barefoot, et al., 1970). Wood density is controlled by a combination of per cent of summerwood, cell size, and wall thickness, yet despite its complex origin, it responds consistently and well to genetic manipulation. Heritability of wood specific gravity in southern pines is consistently high, varying from 0.5 to 0.8, and is little affected by differences in environment (Barker, 1972).⁴ There is great tree-to-tree variability, which results in good gains from a selection program.

Intensity of inheritance of wood specific gravity has been documented many times (Harris, 1969; Zobel, et al., 1972). The change possible in young trees from choosing high and low specific gravity parents is shown in Table 4, the results of a study undertaken with International Paper Company to determine if it is possible to develop trees with high specific gravity juvenile wood.

Although increased wood density is secondary to growth rate in maximizing weight of cellulose per acre, specific gravity has always been found to be of prime importance. For example, when five important characteristics were combined within a selection index, the two most important for dry weight production were tree height and wood specific gravity (Stonecypher, et al., 1973). Although most studies on the importance of wood specific gravity have been made in conjunction with cellulose production for the manufacture of paper, wood density is also a key for quality and utilization of solid wood products.

* Nearly 1,000 parent trees were sampled. Cones were collected from those having specific gravity juvenile wood similar to the mature wood; it is the 5-year-old trees from these seed that were sampled. Comparison was made between the specific gravity of the 5-year-old trees and the central, juvenile core (10 rings from the pith) of the parent trees. Study was in cooperation with International Paper Company, Georgetown, South Carolina.

A number of other wood properties such as tracheid length, moisture content, wall thickness, resin content, and wood color show inheritance patterns that would be useful in a selection program, while spiral grain and holocellulose yields have a less useful pattern. The time may come when several wood characteristics may be emphasized, but because of its overriding importance for both yield and quality (Barefoot, et al., 1970), specific gravity is the one now being emphasized: for some products, high specific gravity is desired; for others, low specific gravity produces the most salable product. Unfortunately, wood properties have been excluded from many breeding programs because of the uncertainty of future utilization and wood requirements. No matter what the ultimate objective, however, a wood-breeding program will result in greater uniformity, a characteristic desired for all products. Changes in wood specific gravity can be achieved with little sacrifice of other desired growth and form characteristics, and require little additional effort if correctly incorporated into the tree improvement program; gravity has a major effect on yields and quality (Zobel, et al., 1971b).

ECONOMICS OF A TREE IMPROVEMENT PROGRAM

All discussions on heritabilities and percentage improvements are academic unless they result in meaningful improvement under operable forest conditions. There has always been interest in the economics of tree improvement but early efforts to estimate the benefits of using a breeding program were unsatisfactory at best (Zobel, 1966). The economics of tree improvement cannot be accurately determined until sound data are available on the costs of developing improved trees so they can be equated with the resultant improvements. We now have very good estimates of costs - even though they vary widely, depending upon organization and method of accounting used, the data are reasonably accurate.

The most difficult problem is to determine the magnitude of quality improvements. Gains in such things as volume, tons of dry fiber, or board feet per acre can easily be translated into dollar improvements. The need is to put dollar

values on straighter trees, trees with smaller limbs, or trees with more uniform and desirable wood qualities. Everyone agrees such qualitative gains are useful but it is difficult to put actual values on them. In an early paper by Davis (1967), some theoretical calculations were made of the value of quality improvement which indicated that the profit picture for a mill could be greatly improved by using trees with better form and better wood. Davis summarized: "The main inference to draw from the example is that relatively small qualitative improvements can have profound effects on mill profits." To obtain some answers as to the question of the value of quality improvements, pulping studies on crooked vs. straight trees, large-limbed vs. small-limbed trees, and diseased vs. disease-free trees were made by International Paper Company (Blair, et al., 1974). Table 5 shows some of the results.

ABOVE: Emphasis in the North Carolina State Program has been on breeding for desired wood qualities. Shown is a cross section of the trunk of a very rapidly growing loblolly pine that had high wood specific gravity (wood density). Specific gravity is the most important of wood characteristics and is strongly inherited, making changes toward lighter or more dense wood relatively easy.

Similar studies have yet to be made as to the value of bole straightness and limb size on solid wood products such as lumber and plywood.

A major problem of any estimates of gain from a tree improvement program is that they are based upon young trees often just coming into merchantability. Will gains and values change as the stands mature? No one knows, but we have been most pleasantly surprised that up through one-third rotation age the genetic gains from breeding seem to be increasing rather than decreasing.

I have in my files a number of economic studies related to the value of tree improvement made by several members of the Cooperative. Although these reports are confidential in nature, and exact data cannot be reported, a summarization of the results makes it obvious that the economics of tree improvement are highly favorable. This is indicated by the scope of activities and continuing expansion of their individual tree improvement programs by members of the Cooperative. It has been of special interest to note that the scope of tree improvement activities within an industry always increases after an economic study has been undertaken. The value that the forest industry in the South attaches to improving forest trees through the application of genetic principles is indicated in that they spend several millions of dollars annually. Even during the slowest periods of economic growth of the industry, such as occurred in 1966 and 1969, they actually expanded.

The above generalizations sound good but we can now cite results from recent economic studies that provide sound data on the value of using genetics in forestry. The first published study, which largely dealt with costs, was by Davis (1967). His analysis showed that at 5 per cent interest and with stumpage prices of only $5 per cord, a genetic gain of 2.5 to 4 per cent would pay for tree improvement activities; compare this with the 10 to 20 percent gains we are actually achieving. Table 6 summarizes some of Davis's results.

* Twenty trees of each category were pulped.
** Resinous by-product obtained during the pulping process.
*** Measured at 500 ml Canadian Standard Freeness.

*When more current $10/cord and 10% rate of interest is used, values become

Davis, who admittedly did not think tree improvement activities were a sound economic investment when he started his study, summarized with this sentence, "... it certainly appears that current investment in loblolly pine seed orchards are well within the 'ball park' with respect to financial justification."

Another early study by Bergman (1968) dealt primarily with the economic impact of seed production from seed orchards. He found that only a moderate percentage increase in wood production is necessary to justify rather high seed costs (or conversely, small seed crops) in the orchards. Seed yield from individual clones has a major effect on the per cent improvement required; for example, clones producing only 7 pounds of seed per acre require 6.5 per cent increase in volume production of resulting plantations of seed in a tree improvement program while clones producing 50 pounds of seed per acre require only a I per cent increase in productivity.⁵ Bergman stressed that the value of a seed orchard program strongly hinges on the seed production capacity of the clones involved, results confirmed by Porterfields (1973) in-depth study of the economics of tree improvement. He, like Davis, emphasized that a seed orchard is a good investment even when genetic improvement is only moderate.

* From Smith, H. D. and Zobel, B. J., 1974. Tree Improvement Short Course Handbook.

* After Smith, H. D. and Zabel, B. J., 1974. Tree Improvement Short Course Handbook.
** Useful life of orchard is 30 years with major seed production after year 10. Tax estimation made on The basis of costs being expensed against other taxable income in the year incurred. This is possible if the orchard is initially charged against research and later transferred to operations, as is usually done.

To illustrate the value of loblolly pine seed, Table 7 was constructed to show optimistic and more pessimistic current conditions in the Southeast.

Numerous studies and tables could be presented on seed value related to the value of tree improvement activities. Table 8 shows actual costs of seed when all activities including tree selection, establishment, and management and testing of a seed orchard program are taken into account.

Based on generalized seed orchard yields in the Cooperative, the cost of improved seed ranges from $10 to $15 per pound; these are similar to the costs estimated by Davis in 1965.

The bulk of the total cost (89 per cent) of a seed orchard program is in site preparation, land costs, supervision, fertilization, mowing, insecticides, harvesting and seed extraction. These are all costs that do not vary with the genetic quality of the orchard trees, so it is clear that extreme care must be taken in selecting parent stock.

ABOVE: A genetics program does not stop with initial seed orchards. Better advanced-generation orchards are now being established from the best progeny from the seed orchards. Shown is one of the best 5-year-old second-generation selections of Weyerhaeuser Company that has all the attributes to be used as a parent in second-generation seed orchards. Advanced and specialty seed orchards (such as disease- or drought-resistant) are in full production throughout the Cooperative.

The question is always raised as to alternatives to genetic improvement since obtaining improved strains is a long-term, difficult, and expensive task. A recent very large study sponsored by TAPPI has been completed on the relative value of changing wood through silvicultural manipulation, genetic manipulation, or changing wood through mill processes.⁶ The first part of the study dealt with linerboard from southern pines; it included data from approximately 20 pulp and paper industries. Results, which were obtained from a complex linear programming approach, were generally determined on a discounted cash flow after tax basis, with the value of the objective function being the total profits from processing and selling linerboard plus the profit from growing wood.

Specific findings from this study cannot he released until the results are published by TAPPI in late 1974 but we can note some general points:

1. All economic analyses to date of pine genetics programs have shown tree improvement to be an extremely good investment. The evidence indicates that rates of return from 17 to 21 per cent are feasible, and this agrees well with the study by Porterfield when similar stumpage prices are considered.
   
   
2. The optimum rotation length will be relatively short for the production of many fiber products.
   
   
3. There is some trade-off between higher specific gravity and mill processing costs, and the most profitable strategy depends greatly on the interest rate selected.
   
   
4. An important point to consider when evaluating a tree-improvement program is that an increase in wood production per acre reduces the acres necessary to support a mill.
   
   
5. To date, two important studies have shown that volume, wood specific gravity, and disease resistance are the most important traits to manipulate in a genetics program. A major consideration is that once a genetic change has been made it is permanent, whereas costs of altering silvicultural and mill processes must be repeated each generation. The TAPPI study has strongly bolstered the tree improvement approach and has shown a high rate of return from genetic improvement.⁷
   

The recent summary of the long-term detailed Heritability Study (Stonecypher et al., 1973) gives a good but conservative indication of the magnitude of improvement that can be obtained through use of genetics on southern pine. Even though the reported heritabilities; obtained from natural, unselected stands are lower than many published papers indicate, gains were still relatively large (Table 9).

Improvements of volume (25 per cent) or dry weight (26 per cent) are considerably more than any of us had hoped, especially when based upon the very low heritabilities obtained from the unselected stands used in the Heritability Study. In an operation such as our Cooperative's, for example, in which 400,000 acres are planted each year, the value of a 25 per cent improvement in volume is worth many millions of dollars.

The gains shown in Table 9 can be achieved when breeding for individual traits. Improvement possible from multiple-trait breeding through use of selection indices gives a more accurate measure of the total improvement possible. When such calculations are made for the combined values of individual traits, height and wood specific gravity were found to contribute the bulk of the improvement when the five criteria of height, dbh, volume, specific gravity, and dry weight were assessed together.

* From Stonecypher, Zabel and Blair (1973).
** When a selection of 1 in 2 was made.
*** When a selection of 1 in 10 was made.

In a recent study by van Buijtenen (1972), selection appears as the most important phase of a tree-breeding program in terms of return on investment in first generation orchards. He points out that progeny testing is less efficient and its main value is in providing material for second generation selections, not for information necessary to rogue an orchard. His conclusion on the minimal value of roguing are in sharp contrast to those of Porterfield (1973), who reports that roguing and more intensive selection are both necessary for maximum economic returns from the tree-improvement effort.

In order to quantify potential gains, Porterfield (1973) made an intensive analysis of the economics of tree improvement. His studies included one industrial organization, one state forest service, and a generalized program based upon the 20 industries contributing to the extensive TAPPI study. The heritabilities which Porterfield used to determine gain were conservative, being based upon parent-progeny relationships obtained in the Heritability Study which have been distorted by heavy fusiform rust and tipmoth attacks. His findings were based on a model of a tree-improvement program that could be varied at will. The results were:

1. Total volume gains from seed orchard seed over unimproved plantations varied from 12 to 14 per cent for unrogued seed orchards in the tree-improvement programs assessed. Additionally, there was a gain of 5 per cent for specific gravity, while bole straightness and crown improvement were in excess of 5 per cent. Volume gains of more than 20 per cent are quite possible by increasing roguing intensity and intensifying wild-stand selection intensities.
   
   
2. Progeny testing is essential if an organization is to meet its goals for improvement in volume and disease resistance. Minimum rates of return can be earned without progeny testing, but in every case profitability is considerably increased by progeny testing and subsequent roguing of the seed orchard. Volume improvement can be up to five times greater with progeny testing.
   
   These findings indicate that while good gains are obtained from wild-tree selection, considerable additional gains will result from progeny testing. Without progeny testing for fusiform rust it would be necessary to greatly increase the selection intensity for wild trees if only minimum improvement goals are to be achieved. Two short-term methods could be used: (a) select trees only from stands heavily infected with fusiform rust, with up to 90 per cent infected stems; (b) screen potential parents by artificial inoculation.
   
   
3. An increased expenditure on wild-stand selection results in higher genetic gains and greater economic returns; current selection intensity could be more than tripled and still be economically justified as shown in table 10, which indicates marginal revenues and costs when selection expenditures are multiplied two to five times.
   
   
4. The profitability of a tree improvement program is closely related to seed yields from the orchard. The best genetic stock is of no value until sufficient seeds are collected and planted-the more seeds, the more acres that can be planted with superior seedlings. Porterfield's study illustrates the extreme importance of maximizing seed yields from orchards by use of the best parent trees, fertilization, irrigation and pest control. Only 8 pounds of seed per acre per year in the seed orchard (after age 10) are necessary to break even for seed which produces seedlings 10 per cent genetically superior in volume, at an eight per cent rate of return; however, each pound of this kind of seed has a present value of $116 and every effort should be made to obtain maximum seed yields.
   
   
5. On the average, a seed orchard costs between $4,500 and $6,500 per acre to establish and test the genetic worth of the parents.
   
   
6. At 1971 prices (87 to $8 per cord), the rates of return (R.O.R.) for the tree-improvement programs ranged from 10 to 14 per cent. When modestly increased future stumpage prices were assumed, the R.O.R. ranges from 12 to 16 per cent. The internal rate of return would be 16 per cent if stumpage prices increased at an annual rate of 3.2 per cent. Even though economic returns were calculated on volume alone, considerable value will accrue from improvement in quality traits, so the quoted R.O.R. values are conservative for several reasons,
   
   
7. The profitability of seed orchards in the Coastal Plain or Piedmont are about the same. Volume gains and stumpage prices are higher in the Coastal Plain but seed yields are lower; Piedmont orchards are able to match the Coastal Plain profitability with higher seed yields.
   
   
8. Volume improvement can vary from nine to 20 per cent under normal circumstances, depending on additional goals that must be satisfied. If a volume gain is 12 per cent when all other goals are satisfied, either heavy orchard roguing or higher heritability estimates will increase potential volume gains to highs of 23 to 25 per cent (table 11).
   

* Adapted from Porterfield, R. L., 1973.

* From Porterfield, R. L., 1973.

Porterfield concludes that the management of the tree improvement programs he studied is consistent with the goals of the organizations. His recommendations for bettering current programs are on degree of emphasis rather than on major changes. Porterfield ends his study with the following: "There is little doubt about the economic justification of tree improvement work with loblolly pine. Even when using very conservative genetic gain estimates and seed yields 25 per cent less than normal, the internal rate of return was 12 per cent for the 'representative' program. Progeny testing and subsequent roguing of the seed orchard increases profitability." A new study now in progress by Matziris backs Porterfield's findings, with an indication that usually about half the clones should be rogued from the first-generation orchards.

---

ENDNOTES

1 As reported in the Sierra Club FOOTNOTES, Vol. VI, p. 4, 1973.

2 Examples will be drawn from the N.C. State University-Industry Cooperative Tree Improvement Program supported by 26 pulp and paper industries and 3 state forest services in 13 southeastern states. Members of the Cooperative control over 20 million acres of forest each year. A description of certain aspects of the Cooperative is given in the Appendix.

3 The best trees of the best families are selected from the progeny tests and placed in new, improved second-generation seed orchards.

4 Heritability is a ratio indicating the relative genetic control of a characteristic.

5 Members of a clone all come from one parent tree; the parent is generally referred to as an ortet, the grafts from a parent tree are individually referred to as a ramet.

6 Technical Association of the Pulp and Paper Industry

7 A summary of the TAPPI studies will be presented at the Forest Biology Committee Meeting in Seattle in September, 1974, and will be published by TAPPI.

---

SUMMARY

The primary emphasis in this paper was on economic information related to tree improvement. The timber supply situation in the Southeast and factors affecting it are discussed in some detail to emphasize the need to increase productivity of forest land. Methods available to increase the timber supply, with emphasis on genetic improvement of planting stock, were discussed in some detail; these include improving adaptability, resistance to pests, growth rate, tree form and quality and wood qualities. The methods used and the gains achieved are outlined, based upon work of the N. C. State Cooperative-the Cooperative and extent of industrial participation is outlined in the Appendix (at end of paper).

The economics section summarized all studies made to date, with tables and data supporting the conclusions made. Results can be generalized as follows: The gains in volume and quality are real and significant, and make treeimprovement expenditures a good financial investment. When we consider the joint judgments of all assessments it appears that tree improvement efforts are in a very real sense exceeding our early and fondest hopes.


ACKNOWLEDGMENTS

I wish to especially acknowledge the help of Dr. T. E. Maki, Dr. C. B. Davey, and Mr. Don Smith of North Carolina State University for their critical review of the manuscript and their help in making it readable.

---

LITERATURE CITED

Barefoot, A., Hitchings, R., Ellwood, E. and Wilson, E. 1970. The relationship between loblolly pine fiber morphology and kraft paper properties. Tech. Bul. 202, N. C. Agri, Expt. Sta., Raleigh, N. C. 89 pp.

Barker, J. 1972. Location effects on heritability estimates and gain predictions for ten-year-old loblolly pine. Ph.D. Thesis, School of Forest Resources, N. C. State Univ., Raleigh, N. C. 107 pp.

Bergman, A. 1968. Variation in flowering and its effect on seed cost - a study of seed orchards of loblolly pine. Tech Rept. No. 38, School of Forest Resources, N, C. State Univ., Raleigh, N. C. 63 pp.

Blair, R. and Zobel, B. 1971. Predictions of expected gains in resistance to fusiform rust in loblolly pine. Proc., 11th South. For. Tree Inapt. Conf., Atlanta, Ga. pp. 52-66.

Blair, R., Zobel, B., Franklin, E., Djerf, A. and Mendel, J. 1974. The effect of tree form and rust infection on wood and pulp properties of loblolly pine. Proc., Ann. Meet., TAPPI, Miami, Fla. pp. 11-19. Tappi (In press).

Davis, L. 1967, Investments in loblolly pine clonal orchards, Jour. For. 65(12) :882-887.

Davis, L. 1969. Economic models for program evaluation. Proc., Second World Consultation on Forest Tree Breeding, Washington, D. C. pp. 1.13.

Harris, J. M. 1969. Breeding trees to improve wood quality. Proc., Second World Consul. on Forest Tree Breeding, Washington, D. C.

Holley, L. 1973. Economics of hardwoods. Handbook, Hardwood Short Course, School of Forest Resources, N. C. tate Univ, Raleigh.

Lowe, K. E. 1973. The complete tree. Pulp and Paper 47(13):42-47.

Matziris, D. and Zobel, B. 1973. Inheritance and correlations of juvenile characteristics in loblolly pine (Pinus taeda L.). Sil. Gen. 22 (1-2):38-44.

Porterfield, R. 1973. Predicted and potential gains from tree improvement programs - a goal programming analysis of program efficiency. Ph.D. Thesis, Yale Univ., New Haven, Conn. 240 pp.

Shelbourne, C. 1969. Breeding for stem straightness in conifers. Proc., Second World Consul. on For. Tree Breeding, 1:293-302. FAO.

Shirley, R. 1973. Georgia survey shows healthy forest resource gains. Forest Farmer 32(11):6-7.

Stonecypher, R. and Zobel, B. 1966. Inheritance of specific gravity in five-year-old seedlings of loblolly pine. Tappi 49(7):303-305.

Stonecypher, R., Zobel, B. and Blair, R. 1973. Inheritance patterns of loblolly pines from a nonselected natural population. Tech. Bul. No. 220, N. C. Agri. Expt. Sta., N. C. State Univ., Raleigh, N. C. 60 pp.

van Buijtenen, J. and Saitta, W. W. 1972. Linear programming applied to the economic analysis of forest tree improvement. Jour. For. 70 (3):164-167.

von Wedel, K., Zobel, B. and Shelbourne, C. 1968. Prevalence and effects of knots in young loblolly pine, For. Prod. Jour. 18(9):97-103.

Vaux, H. J. 1973. How much land do we need for timber growing? Jour. For. 71(7):399-403.

Zobel, B. 1966. Tree improvement and economics; a neglected interrelationship, Sixth World Forestry Gong., Madrid, Spain. 15 pp.

Zobel, B. J. and Roberds, 1. 1970. Differential genetic response to fertilizers within tree species. For. Biol. Work-shop, Sec. Artier. For., Michigan State University. 19 pp.

Zobel, B., Blair, R., and Zoerb, M. 1971. Using research data-disease resistance, Jour. For. 68(8):486.

Zobel, B. 1971a. Challenges of the seventies - wood for forest industries. Jury, For. 69(4):212-215.

Zobel, B. 1971b. Wood for forest industries. Amer. Paper Industry 53(1):26-36,

Zobel, B., Kellison, R. and Kirk, D. 1971b. Wood properties of young loblolly and slash pine. Symp. on "Effect of Growth Acceleration on Wood Properties," Madison, Wis. p. M1-M22.

Zobel, B., Kellison, R., Matthias, M, and Hatcher, A. V. 1972. Wood density of southern pines. Tech. Bul. No. 203, Agri. Expt, Sta., N. C. State Univ., Raleigh. 56 pp.

---

APPENDIX

To enable the reader to appreciate the discussions, results, and conclusions in the body of the paper, a description is given below of the cooperative programs frequently referred to in the text.

The Cooperative was initiated in 1956 with 12 charter members. It continued to grow, arid in 1966 was divided into a Pine Tree Improvement Cooperative and a Hard- wood Management and Utilization Cooperative. Both of these have expanded so that they now consist of the following (membership is shown on the last page):

Pine 26 industries 6 second working units* 3 state forestry divisions

Hardwood 16 industries 5 second working units* 1 state forestry division 1 Hardwood Research Council

The Cooperative is unique in that it has no contracts or written agreements. Any organization can withdraw at any time if it feels the value of the Cooperative does not warrant continued participation; conversely, any organization can be asked to withdraw if it is not making the contribution necessary to keep the Cooperative viable.** Routine business is handled by an Advisory Committee, composed of decision-makers on the administrative level; this group meets once a year at which time budgets, plans, and achievements of the Cooperative are reviewed. A Contact Men's Meeting, attended by those persons actively engaged in field operations, is held once yearly, each time at the field center of a different cooperator. At the Contact Men's Meeting, ideas, findings, successes and failures are reviewed and discussed. It is essentially a led discussion in which the contact men make most of the input, followed by a field trip to see and discuss activities of the host organization. In addition, an Executive Committee is designated that handles any political problems or monetary problems that come up between Advisory Committee Meetings.

Many organizations have asked to become members of the Cooperatives but have been refused, either because of their geographic location or because their scale of operation or their organization made it appear that they would not greatly benefit from, or contribute to, the Cooperative. All but the Hardwood Research Council of the Hardwood Cooperative aye also members of the Pine Cooperative. The area of operation of the Cooperatives is in the southeastern United States, delimited by the Mississippi River to the west and the Mason-Dixon Line to the north.

Members of the Cooperatives control approximately 20 million acres of forest land. It is the objective of the Pine Cooperative to help develop genetically improved stock for the requisite annual planting of 400,000 acres (300,000,000 000 trees) of its members. Progress is good, and in 1973 enough seed were obtained from seed orchards (App. Table 1) to produce 150,000,000 genetically improved trees (App. Table 2). Within 5 years all planting needs of members of the Cooperative will be met with genetically improved planting stock. The three state organizations are making the improved seedlings available to small landowners; all their nursery production should be from seed orchard seed within 6 or 7 years. Testing the quality of the 3,500 parent trees established in 177 separate seed orchards (App, table 1) has developed so well that advanced-generation and specialty seed orchards are now in operation, and by next year we should have nearly 200 acres of second-generation orchards established.

The staff of the Cooperative consists of five faculty members, eight technicians and secretaries, and about 35 professional foresters active in field work who are employees of members of the Cooperatives. In essence, staff members of the Cooperative handle the design, analysis and interpretation of field tests while members of the Cooperative do the actual field studies under guidance of the Cooperative staff. Thus, activities such as tree selection, orchard establishment, crossing for progeny testing, test establishment, and test measurement are done by members on their own lands, but tinder guidance of the Cooperative. Tree grading, test and crossing designs, computer analyses, wood analyses, and interpretation of the data are done by the staff of the Cooperative at Raleigh.

The main objective in the pine program is to produce genetically improved seed, while that in the Hardwood Cooperative is to help generate information necessary for a hardwood management program. Activities have expanded, however, and the Cooperatives are involved in an advisory capacity on wood properties, economics, and general forest management and investment problems faced by the members, occasionally including forest management of conifers and eucalyptus in tropical and subtropical regions.

Considerable basic research is undertaken, primarily by graduate students; the program has between 15 and 20 graduate students each year, most of whom are studying for the Ph.D. degree or are active as postdoctoral students. Basic studies such as the very large Heritability Study undertaken in cooperation with International Paper Company have produced information badly needed for the success of the applied program. Funds for basic research have been obtained from numerous sources, including the National Science Foundation, Rockefeller, National Institute of Health, Ford Foundation, Kellogg Foundation, N. Board of Science and Technology, the Southern Industrial Disease and Insect Research Council, and others. One important function is the summarization and dispensing forestry information deemed important to members of the Cooperative.

Perhaps the greatest strength of the Cooperative has been its multidisciplinary character. Faculty members of many specialties at the University contribute help, research and advice. For example, we have close ties with Pulp and Paper Science, Genetics, Botany, Statistics, Soils, Plant Breeding, and Pathology. Because of the complexity of needs of the two Cooperatives, such an interdisciplinary approach is mandatory.

Work within the Cooperative is operational and no longer considered research. The scope of activities in the seed orchard operation, which in 1973 produced enough seed to plant 200,000 acres with genetically improved stock, are indicated in Appendix Tables 1, 2 and 3. Progeny testing, so vital to maximum genetic improvement and development of advanced-generation breeding programs, has given us a wealth of genetic material necessary to keep the adaptability base broad and yet enable maximum gains for characteristics of economic importance. Perhaps most indictative of the status of tree improvement is that the industries have removed it from a research to an operational status.


* A second working unit is an operation within a member company but physically separated so that it requires services similar to a separate unit. As an example, a company with divisions in Georgia and Virginia has one considered the parent unit, the other a second working unit.

** No organizations have withdrawn, but one was dropped from the Hardwood Cooperative because of inaction.

Approximate annual regeneration by Cooperative members - 400,000

* In 1972 we obtained 8,491 bushels of cones from loblolly and slash pines. These averaged 1.06 pounds/bushel for coastal loblolly, .84 pounds/bushel for Piedmont loblolly, and .60 pounds/bushel for slash pine.

A total of 15,140 crosses and their checks have been planted.

3m-8'74(R9393L)VL

---

Introducing: Bruce Zobel


Bruce J. Zobel enjoys a special role in forestry as a university professor who works for thirty-one major forest products corporations. As the E. F. Conger Distinguished Professor of Forestry at North Carolina State University, his principal assignment is as Director, Industrial Forest Tree Improvement Program. Both his function and his accomplishments lie in the bringing of better forest management to the pine and hardwood forests of thirteen southeastern states, with better trees as the keystone of his approach.

A native Californian, Zobel earned the B.S. in forestry in 1943 at the University of California. He then headed for the redwoods, taking a position as assistant logging engineer for the Pacific Lumber Company. In 1945 and 1946 he had his first southern forestry experience as Forestry Officer for the Marine Corps at Camp LeJeune, North Carolina. Following this, he returned to the University of California where he worked as a senior laboratory assistant while completing his Master of Forestry degree in 1949 and the Ph.D. with specialization in forest genetics in 1951.

From 1951 to 1957 he served as silviculturist at Texas A & M University, developing a program in the genetics of southern pine. As the potentials of this work became increasingly clear, the interest of the forest industries mounted and in 1956 the Industrial Forest Tree Improvement Cooperative was established with Zobel as director. In 1957 the headquarters for this program was established at North Carolina University and he shifted there as associate professor. Promoted to professor in 1960, he was named as the E. F. Conger Distinguished Professor of Forestry in 1962.

His accomplishments both in research and in the development of a fully operational program for providing better trees for industrial forestry in the South have been widely recognized. In 1965 he received the Governor's Award for Conservation in North Carolina. In 1968 he received the Barrington Moore Memorial Award for biological research from the Society of American Foresters. In the same year he was elected a Fellow of the International Academy of Wood Sciences. In 1969 he was elected a Fellow of the Society of American Foresters. In 1973 the Technical Association of the Pulp and Paper Industry (TAPPI) honored him by election as Fellow and by the TAPPI Research and Development Award.

A creative scientist and gifted teacher of a large cadre of graduate students, Bruce Zobel is above all a professional forester who has contributed and will continue to contribute greatly to improving the economic efficiency of forestry as a mean of serving his fellow Americans.