A. BERGMANAXEL BERGMAN is in charge of forest seed and plant production at the National Board of Forestry, Stockholm.
IT IS NECESSARY to make many simplifications when approaching the huge topic of this paper. To start with, as a matter of principle forestry should be defined as an economic enterprise, an industry based upon trying to obtain the maximum profit, to get the greatest possible surplus of income above expenditure. To that extent, there is no difference between forestry and other enterprises.
However, forestry on a sustained yield basis has special characteristics which make calculations difficult. Among these, the long rotation period may be most important. The term " long " in this context should include rotation periods of 100 years as well as of 20 years, although the calculations required in each case may be more or less troublesome. The fact that the growing forest is a production factor as well as the end product is another characteristic causing trouble when calculations are made. Further, different categories of landowners certainly have different objects in view; therefore, economic calculations may vary considerably among the forest-owning farmer, the pulp and paper industry and the State. Calculations certainly vary even among the industries themselves, depending on the degree of self-sufficiency in the supply of roundwood and the possibilities of buying it.
It would be easy to list still other circumstances that make economic calculations difficult. However, the essential point here is that those circumstances also pertain to the more limited topic of forest tree improvement. It is impossible here to discuss all the possible situations that may occur. It is thoroughly justifiable to limit analysis to the more urgent question of the economic contribution of tree improvement programmes to forestry in general.
Forest tree breeding should be regarded as one method among others of increasing volume production and of improving quality properties. Even if it could be proved that forest tree breeding is lucrative comparisons would still have to be made with results that might be attained in other ways. The economic resources available will determine how much can be done. As long as the resources are limited and they always will be - it is necessary to establish priorities. However, rather than choose one method, the optimum economic result can generally be attained with the help of a combination of different methods. Thus forest tree breeding may need to be combined with, for instance, forest fertilization and silvicultural methods. This does not make calculations any easier. For it is scarcely possible to sum up the percentile improvement accruing from different measures or methods; these interact with one another in an often complicated interplay.
If it is decided to conduct a tree breeding programme, either in preference to or in combination with other methods, there are a number of alternatives. There is no single universally valid breeding method. On the contrary, different species, even if closely akin, may have greater or smaller chances of giving successful results with the same breeding programme in one and the same region. Condition may vary so markedly between different regions that they call for completely different programmes.
Therefore, the task is not to take up a position for or against a certain breeding programme or a certain type of seed orchard; rather is it to emphasize the necessity for detailed analysis leading to a number of alternative solutions.
Needless to say, high genetic gain is desirable in a tree improvement programme. Summaries in tabular form of genetic gain to be expected from different breeding and selection methods are to be found in papers and publications by Wright (1962), Namkoong et al. (1966), Shelbourne (1969) and Nikles (1969). Genetic gain is determined by the intensity of selection and by the heritability, as well as by the genetic variance. Economic value is determined by a great many characters. Therefore, the tree breeder has to deal with a very difficult biological and economic complex to which there is no valid general solution.
Selection indexes are formulated as a necessary help for the breeder. Here close cooperation between and forest economists is necessary at the very start of the breeding programme. As markets vary, economic values of characters vary also. The need to restrict the number of selected characters makes it important to obtain an accurate determination of their relative economic weights. The long-term character of breeding programmes requires appraisals of future markets. Possibilities of getting early results by progeny tests for selected characters vary according to species and conditions, a fact that influences the primary evaluations. Illy (1969) used linear programming and game theory as aids for decision-making in breeding programmes for Pinus pinaster. Papers by Le Roy (1960), van Buijtenen and van Horn (1960), and Stern (1964) give valuable guidance to the principles of constructing selection indexes.
Here it is important to remember that the genetic gain finally realized will very probably not be the same as the gain theoretically expected. Furthermore, the aim is not maximum genetic gain but an optimum economic result from the breeding programme. Since the aim is to maximize the economic result from forestry as an enterprise, costs to establish and manage different programmes and expected seed yields must be considered. Although from a biological viewpoint it may be interesting to know that a 5 or 50 percent increase in volume production can be obtained through tree breeding, even a large percentage improvement may in itself have little economic significance. Until the time and costs to produce one cord or cubic metre are known, we cannot estimate whether expected benefits will exceed costs.
PRESENT KNOWLEDGE
It should first be recognized that most of the data needed to make economic calculations are lacking. Yield differences obtained through breeding programmes are, in general, not yet known. An idea of possible figures is just becoming known, and too few data are available on the improvement of quality characters. Here the big difference between forest tree breeding and agricultural breeding should be fully recognized. Forestry does not benefit from the very short rotations and homogeneous growing conditions that make early results possible in agriculture. In forestry, the best knowledge seems to be of the costs and first benefits of some main breeding programmes Much broader knowledge about several kinds of programmes is necessary.
DIFFERENT APPROACHES
Several approaches can. be made when calculating the profitability of tree breeding programmes.
Two time scales can be used:
1. long-term, many generations;
2. short-term, one-generation.
Increases can be expressed as changes in:
1. total volume yields;
2. site index;
3. stumpage price reflecting improved quality.
As criteria the following may be chosen:
1. soil value;
2. earning rate of the investment.
Calculations may be carried out in two directions:
1. from known yield figures from breeding programmes;
2. from known cost figures of the improved seed, seedlings and plants.
It is very common in forest economics to calculate soil value. If the long-term approach mentioned above is used, future rotations should be included. Then " soil value " would mean the net discounted present value per unit area. This is the difference between the present values of all future anticipated cash receipts and of all required outlays. Values are obtained from models presuming repeated similar action in the future.
From this classical viewpoint, Streyffert (1948) discussed the maximum regeneration cost when genettically improved material is used. If the calculated soil value when ordinary material is used is Bu1, according to existing yield tables, the soil value when improved material is used is indicated by Bu2. C represents the capital value of all future regeneration costs with the improved material. Thus, the following formula is obtained:
Cmax = Bu2 - Bu1.
From that expression a formula can be obtained for the extra cost, cmax, that could be invested in planting with improved material:
in which also different rotations, u2 and u1 are presumed for improved and ordinary material respectively' and p is the discount rate. Lacking data, Streyffert chose to demonstrate the method by presuming that improved and ordinary stands would develop as stands on different site classes and thus derived tabulated soil. values. Libby (1965) also evaluated investments in tree improvement programmes through all future commercial harvests and presumed different increases in future economic yields. The sound justification for taking the long-term approach cannot be doubted. However, the scanty knowledge of the increased yields to be obtained through tree breeding programmes has to be recognized. Furthermore, economic development in general has caused very rapid changes. Forestry is evolving in a striking degree toward establishment of stands with fewer seedlings or plants, toward fewer thinnings, and, consequently, toward clear cutting in heavily stocked stands. Yield tables generally do not reflect these new programmes, even when ordinary seed and plants are used. So. a very poor knowledge of production figures in modern forestry programmes considerably restricts ability to make calculations.
As a rule, calculations of the profitability of tree improvement are based on a situation in which a decision to reforest has already been made. What occurs is that seed of different quality and price is offered on the market. Taking the uncertainty factor in forestry into consideration, it might be possible to evaluate the cost of the genetically improved seed while waiting for yield figures. From those cost figures it might be possible to draw conclusions that will influence further decisions.
IMPROVED SEED: AS A RULE AN EXTRA COST
The normal way of procuring forest seed is to make use of the cone setting of forest trees grown for timber production. Thus, forest seed has been produced " free of charge." Payment has only been made for collecting and extracting cones and seed. In comparing the cost of seed from seed orchards, it is observed that only expenditures for cone collecting and seed extraction are common to both methods. All other orchard costs are extra.
Not only is money spent to establish orchards especially intended for seed production. Having invested in such projects, it is also necessary to protect them from insects, rodents or fungi. Thus, the first extra cost entails further extra expenditures, plus additional costs for continuous management in the orchard. In natural stands, on the other hand, damaged cones are not collected. Cones are gathered only in years in which such attacks do not occur.
SEED PRODUCTION IN ORCHARDS
To calculate seed production costs in orchards, it is necessary to make simplifications and assumptions. The normal way is to divide the rotation age of the orchard (Figure 26) into an establishment period and a commercial production period. As seed crops will always be more or less periodical, average yield over the commercial time is calculated.
To begin an estimate of seed costs, the capital value per acre of the orchard at the start of commercial seed production is evaluated. Costs for establishment and management are estimated. The orchard should be amortized over the remaining, commercial period. After obtaining an annual depreciation charge, the cost per kilogramme of seed will be computed by adding annual operating expenses and harvesting costs per hectare and then dividing by the average yield per hectare. The seed cost obtained must then be reduced by the market price for " ordinary " seed to obtain the extra cost for improved seed. Davis (1967) used this method to analyse clonal seed orchards of Pinus taeda in the southeastern United States. Bergman (1967, 1968) used a similar approach, but excluded costs of harvesting and extraction in the calculation, assuming those costs to be the same in both stands and orchards (Figure 27). Thus, the estimated cost of improved seed is the net or extra cost required.
Using known figures of plantable seedlings per kilogramme of seed, the net investment or the extra cost per hectare at different spacings can then be estimated (Figure 28). Further, using known yield figures for different spacings and site indexes and known stumpage prices, it is possible to calculate what increase would justify the extra investment for genetically improved material (Figures 29 and 30). Investment return can be measured by the internal rate of return, the interest rate at which discounted costs equal discounted incomes (Figure 31).
FIGURE 26. - Seed orchard yield over time. - FROM DAVIS 1967
COSTS OF SEED ORCHARDS
The main costs in orchard programmes are for establishment, management, capital expenditure and progeny tests. Regardless of what has been said here in support of taking a one-generation approach when calculating the profitability of tree breeding programmes, the many generation approach of those programmes themselves should also be recognized. Progeny tests are an essential and urgent part of the whole breeding idea. This testing work, as well as the original selection work, is the kind of effort that will benefit future forest generations.
FIGURE 27. - Costs of seed production, excluding harvesting and extraction, in clonal seed orchards of Pinus taeda under assumed average conditions. In this and the following graphs the data were originally presented in the English system.
FIGURE 29. - The percent improvement required to justify extra costs for different planting programmes, stumpage prices and rotations, when genetically improved seed of different production costs is used.
FIGURE 30. - Detail of a similar graph, enlarged to show the 0 to 3 percent range of improvement.
If the approach described above is used, it would not be justified to charge both the costs of selection and progeny tests to the first seed orchard generation. It would be logical to include only the selection costs and to charge the progeny tests to the second generation of seed orchards. Those costs are in a way the same necessary start for that orchard generation as the selection of superior trees is for the first one. Charging the costs of progeny tests to the next seed orchard generation would recognize the long-term view of tree breeding. At the same time it would be possible to evaluate the benefits of the next forest generation on the basis of more up to date production figures.
It should further be recognized that simplifications are being made in both cases, whether the progeny tests in the short-term approach are included or excluded. In reality, early results from progeny tests may influence thinning of the first generation seed orchards. Thinning in turn affects both the costs of management and seed yield and thus finally the seed cost, but it also results in a second increment of genetic improvement. This would lead once more to the conclusion that progeny tests should be charged against the next seed orchard generation.
SEED PRODUCTION AREAS
It is necessary to include in this discussion the cases in which the improved seed is produced in combination with roundwood production. This is the case in the so-called seed production areas. It is true that genetic gain, even if selective thinning is made, must in almost every case be expected to be considerably less than in more advanced breeding methods. The low cost of obtaining seed should be taken into consideration and also the income from the wood production. In general, the method involves no extra cost compared with ordinary seed, although the stands may sometimes be fertilized.
USE OF FOREIGN PROVENANCES
The use of foreign provenances cannot be excluded from the present discussion. In many cases, provenance research work has given information about foreign provenances that are superior to local ones. This way of improvement contains the cost of trial plots and measurements; however, once the results are available, the seed is generally obtained without any extra cost as compared with the cost of local seed. Provenance research should be considered as basic research and the cost of research should not be added to these calculations.
In addition to providing rewarding biological results, provenance research clearly provides practical opportunities for applying results on a large scale at low cost due to the fact that seed is procured in the ordinary way. Provenance research is a very important part of tree improvement; it should be given highest priority at the outset of any programme (Callaham, 1964).
Even greater difficulties are encountered when trying to evaluate quality improvements. The main problem is the fact that quality has different meanings to different enterprises. The plywood industry, the sawmill and the pulp and paper industry undoubtedly have different demands. Also, very little is known about the magnitude or value of quality changes obtained through breeding.
It may perhaps be said, however, that the present trend is very roughly toward production of poorer raw material than formerly, especially with respect to external quality characters and also to some wood properties. Stands are being formed with fewer, more rapidly growing trees per hectare. This is an inevitable trend, resulting from the economic development in forestry, with its demands for reducing the high costs of capital, labour and transport. No doubt in these new and less dense stands there w ill be an impairment of quality as compared with the dense, slow-growing, high-quality natural forests of former times
FIGURE 31. - Rates of investment return from different red and jack pine tree improvement programmes for specified increases in site index, where improved trees are to be planted on site index 55 land. The site index increase was originally given in feet. - FROM LUNDGREN AND KING (1966)
Instead of speaking of the improvement in quality brought about by forest tree breeding it would perhaps not be unreasonable to see forest tree breeding as helping to make the deterioration of quality less than it otherwise would have been. This may apply at least in regions where the species dealt with are characterized throughout by good quality. If a tree breeding programme starts with intensive selection in populations exhibiting poor quality, then future quality improvement may be more probable, despite changed forms of forest management. However, minimizing deterioration of quality is a worthy contribution.
Davis (1969) shows the great effect on the cost of manufacturing at the mill that may be obtained from only a small percentile quality improvement. He points out, however, that procurement schedules and sorting methods at the mill will probably have a bigger effect than genetic and biological measures. Fenton (1969) also stresses the reduction in the cost of industrial processing brought about by improvement in quality through breeding programmes.
It is also necessary to consider the changing demands of industry for quality over the course of time, together with the complexity of the quality concept. Obviously calculations must be made on lines suited to the character of each individual enterprise.
FEW ALTERNATIVES TO RESEARCH ON RESISTANCE
When quality improvements are considered it is justified to emphasize the gains that should accrue from advances in the field of research on disease resistance. Losses in value due to damage by insects and disease may amount to huge sums. There are few alternatives to breeding for resistance; therefore, it is possible that the most valuable contribution to forestry by forest genetics will come from resistance research. Papers by Björkman (1964) and Gerhold (1969) review the considerable world-wide work that is being done in this field.
In Italy, outstanding results have been obtained in breeding poplars resistant to different diseases (Castellani and Prevosto, 1969). There, as in some other countries, poplars are grown under intensive culture on agricultural land. In this case a different kind of economic analysis is required. Alternative use of land is possible and profitability of wood production should be compared with that of agricultural crops. This example from Italy illustrates that the national/economic aspects of balancing wood production and food production on good sites should also be taken into account. As stated in the introduction, that situation is beyond the scope of this paper. However, it must be stressed here that progress in the field of genetic research on poplars makes the alternative use of agricultural land for wood production economically feasible.
Reliance on natural seed production can have substantial indirect, costs resulting from deferred forest pro auction when seed is not produced in sufficient quantities from desired provenances. The seed orchard idea in its broad sense means reliability in control of seed origin. If the growing difficulties of getting the required seed quantities in the ordinary way are considered (Kellison, 1969), the seed orchard solution appears to have considerable advantages.
Further it is evident that seed from orchards is improved physiologically as well. This fact means much both in nursery work and later in reforestation. Improved seedling and plant material improves the possibility to establish new stands with fewer trees per hectare. Mechanized logging systems of the future will demand as little variation in diameter among stems as possible. There is reason to believe that, through its increased uniformity, improved material will be of importance here too. Giordano (1969) also mentions reducing costs of cultural practices by selection of planting stock as another important goal for forest breeders.
At the World Forestry Congress in Madrid, Bouvarel (1966) and Zobel (1966) emphasized the importance of paying more attention to economic considerations in tree breeding. Obviously, with an increasing world demand for wood, many measures must be combined in efforts to increase wood production economically.
Resources available will determine the extent of the measures and methods. Because a great part of the cost of raw wood is due to harvesting and transport, technical developments in these fields will very likely have a great influence on the character of future forests. The role of forest tree improvement will be significant, but it must be analysed in context with all other opportunities for increasing benefits for forestry.
ARNAUD, C., BARADAT, P., CASTAING, J. P., ILLY, G. & MAINIE, P. 1969. Choice of a research and development strategy in maritime pine breeding. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-8/19.
BERGMAN, A. 1967. Synpunkter på skogsbrukets fröförsörjning. Skogsstyrelsens S-information No 26, 48 p. See also Sveriges Skogsvårdsförbunds Tidskrift, h. 3. Stockholm.
BERGMAN, A. 1968. Variation in flowering and its effect on seed cost. School of Forestry, North Carolina State University. Tech. Rep. 38. 63 p.
BJÖRKMAN, E. 1964. Breeding for resistance to disease in forest trees. Unasylva, 18(2-3): 7181.
BOUVAREL, P. 1966. Les facteurs économiques dans le choix d'une méthode d'amélioration. 6e Congrès forestier mondial. 6 CFM/G/C.T.I./4.1.15.
CALLAHAM, R. Z. 1964. Provenance research: investigation of genetic diversity associated with geography. Unasylva, 18(2-3): 40-50.
CASTELLANI, E. & PREVOSTO, M. 1969. Effects of research on the economy of poplar culture. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-6913/4.
DAVIS, L. S. 1967. Investments in loblolly pine clonal seed orchards. J. For., 65(12): 882-887.
DAVIS, L. S. 1969. Economic models for program evaluation. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-13/2.
FENTON, R. 1969. Economics of Monterey pine. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-13/3.
GERHOLD, H. D. 1969. A decade of progress in breeding disease resistant forest trees. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-5/1.
GIORDANO, E. 1969. Interaction between breeding and intensive culture. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-6912/1.
ILLY, G. 1969. Les indices de sélection, example du pin maritime. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-7/3.
KELLISON, R. C. 1969. Establishment and management of clonal seed orchards of pine. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-11/7.
LIBBY, W. J. 1965. Comparison of seven tree improvement schemes. Forestry 229 Course report, University of California. 22 p.
LUNDGREN, L. & KING, S. P. 1966. Estimating financial returns from forest tree improvement programs. Proc. Soc. Am. Foresters, 1965, p. 45-50.
NAMKOONG, G., SNYDER, E.B. & STONECYPHER, R. W. 1966. Heritability and gain concepts for evaluating breeding systems such as seedling seed orchards. Silvae Genet., 15: 76-84.
LE ROY, H. L. 1960. Statistische Methoden der Populationsgenetik. Basel, Birkhäuser.
NIKLES, D. G. 1969. Breeding for high yielding characters, growth and yield. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-2/1.
SHELBOURNE, C. J. A. 1969. Predicted genetic improvement from different breeding methods. Second World Consult. Forest Tree Breeding. FAO-FO-FTB-69-8/16.
STERN, K. 1964. Population genetics as a basic for selection. Unasylva, 18: 21-29.
STREYFFERT, T. 1948. Skogskulturåtgärdernas ekonomi. Sweden, Royal School of Forestry. Bull. I. 54 p. (English summary)
VAN BUIJTENEN, J. P., & VAN HORN, W. M. 1960. A selection index for aspen based on genetic principles. Lake States Aspen Gen. and Tree Impr. Group. Prog. Rep. 6. Appleton, Wisc.
WRIGHT, J. W. 1962. Genetics of forest tree improvement. Rome, FAO. 399 p. FAO Forestry and Forest Products Studies No. 16.
ZOBEL, B. J. 1966. Tree improvement and economics: a neglected relationship. 6th World Forestry Congr. 6 CFM/G/C.T.I./4.1.19.
