MSU LIBRARIES “ RETURNING MATERIALS: K Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. "on. 1"" .I. “(I 1"! 9,33, 1.1.5 24’) u; fiLPF’l 9' 45-331}! COST EFFECTIVE TIMBER MANAGEMENT PLANNING: A PROBLEM ANALYSIS FOR THE PROVINCE OF NOVA SCOTIA BY Kenneth Lee Runyon A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1982 c mot-8'7 ABSTRACT COST EFFECTIVE TIMBER MANAGEMENT PLANNING: A PROBLEM ANALYSIS FOR THE PROVINCE OF NOVA SCOTIA By Kenneth Lee Runyon The Province of Nova Scotia is currently faced with a softwood timber shortage. Past timber management practices and expansion of manufacturing facilities have resulted in a resource base that cannot continue to sustain current and project requirements. There are a number of alternatives including more intensive management and closer utilization. The problem is one of choice, given alternatives, and constraints. The purpose of this study is to identify the problem, that is, the decision-maker, objectives, alternatives and context; to determine the information needs to solve the problem; and finally to indicate how that information might be obtained. The decision-maker of concern is the Govern- ment of Nova Scotia. Its objective is interpreted as being to maintain employment and incomes in the forestry sector at the current level in the most cost-effective way. Alternatives are classified into three major categories. Kenneth Lee Runyon These are general administration and support, forest management, and utilization. Concern of this study is with a subset of forest management alternatives, namely, silviculture. Five silvicultural treatments are identi- fied. These are planting nonsatisfactory restocked lands; planting recent cutover, burned or insect damaged areas; cleaning; commercial thinning; and the combination of cleaning and commercial thinning. A review of past and current work shows that much effort has been put into development of a timber demand/ supply simulation model for the Province. This model was designed to provide a tool for management planning, but it has not been used. Structure and weaknesses of the model are identified and discussed. In addition, other relevant work both in Canada and the United States are examined. Information needs to solve this problem are categoriz- ed under six components: identifying silvicultural oppor- tunities - application of treatments and area available; biological effects; effects on allowable annual cut; costs and direct employment; roundwood consumption; and impact on employment. Within each component, variables, interre- lationships, and data availability are identified. Finally, five research studies and their priority are delineated. These follow the problem components listed above. For each study, the problem, objective, approach, and requirement are specified. ACKNOWLEDGMENTS The author is deeply grateful to the Canadian Forestry Service, Maritimes Forestry Research Centre for providing leave of absence and financial support for this study. Special appreciation is given to Dr. Robert Manthy, princi- pal advisor at the start of this study, and to Dr. Lee James who later accepted responsibility. Thanks are also expressed to Drs. Robert Marty, Allan Schmid, and Victor Rudolph for their effort in reviewing the study and provid- ing comments. Several individuals with the Nova Scotia Department of Lands and Forests and the Maritimes Forest Research Centre provided data and suggestions throughout the course of this work. Their input is sincerely appreciated. Thanks are ex- tended to Margaret Cameron for her assistance in editing and to Nancy Hay and Louise Yerxa for typing the manu- script. Finally, the author extends a special appreciation to his wife Cheryl who provided continuing encouragement and who shared so much in the effort. ii TABLE OF CONTENTS LIST OF FIGURES CHAPTER 1 INTRODUCTION Statement of Problem Objective and Scope of Study Justification Approach CHAPTER 2 DEVELOPMENT AND CURRENT STATUS OF FORESTRY SECTOR Forestry Resources Land Area Ownership Cover Type and Species Age Class Distribution Resource Administration Resource Use Primary Forest Production The Logging Sector Timber Harvest Other Removals Nonconsumptive Uses Allowable Annual Cut/Harvest Balance AAC Model AAC Compared with Harvest Forest Products Manufacturing and Markets Sawmills Pulp and Paper Mills Other Forest Products CHAPTER 3 DEFINITION OF THE PROBLEM The Decision Maker Objectives and Criterion for Choice Alternatives Governmental Mandate in the Forestry Sector Management of Provincial Crown Land Legislation Affecting Other Lands and Products Administration of Funds Forestry Sector Employment Activities Alternatives for Increasing Employment and Incomes Scope of Study iii vi co O‘kflUlv-‘l-l \OCDWQQ 10 11 11 11 12 14 15 16 16 18 19 19 21 22 23 23 23 24 28 28 28 29 30 31 34 38 Context of Problem Factors Affecting Silvicultural Opportunities Factors Affecting Cost of Alternatives Planting Cleaning Commercial Thinning Cleaning and Commercial Thinning Factors Affecting Impact of Silviculture CHAPTER 4 PAST AND CURRENT WORK Direct Assessment of Silvicultural Opportunities Timber Supply and Demand Analyses Nova Scotia Demand/Supply Model Model Description Evaluation of the Nova Scotia Model Forest Management Planning in Quebec Demand and Supply Analyses in the United States Contribution From Timber Sector Activities CHAPTER 5 PROBLEM COMPONENTS AND GENERAL RESEARCH APPROACH Identifying Silvicultural Opportunities Types of Treatments Application of Treatments Area Available Biological Effects of Treatments Effects on Allowable Annual Cut Costs and Direct Employment of Alternatives Silvicultural Costs Direct Employment From Alternatives Roundwood Consumption Planning Horizon Roundwood Supply Roundwood Demand Roundwood Consumption and Price Approaches to Forecasting Roundwood Consumption Naive Methods Barometric Technique Opinion Polling Econometrics Impact on Employment Harvesting Employment Manufacturing Employment Sawmills and Planing Mills Pulp and Paper Mills Other Forest Products Manufacturing iv 39 40 42 42 44 45 45 45 49 49 49 58 60 60 73 78 83 86 89 89 89 90 91 95 97 100 103 103 106 107 109 110 112 115 116 118 119 120 120 123 124 127 127 130 131 CHAPTER 6 RESEARCH STUDIES AND PRIORITIES Study Study Study Study Study Study 1: 2 3: 4: 5 6. Silvicultural Opportunities Biological Effects of Treatments Costs and Direct Employment of Silviculture Alternatives Effects on Allowable Annual Cut Roundwood Consumption Impact of Silvicultural Alternatives on Employment Discussion of Research Studies & Priorities CHAPTER 7 SUMMARY AND CONCLUSIONS LIST OF REFERENCES References 136 136 136 139 141 143 145 147 148 150 150 156 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10. 11. 12. 13. 14. 15. 16. 17. LIST OF FIGURES Forestry sector employment activities. Forest management alternatives for indus- trial roundwood production. Conceptual framework of Nova Scotia demand/supply study. Theoretical determination of derived demand for wood. Determination of derived demand for wood in Nova Scotia model. Determination of derived demand for a mill producing more than one end-product. Determination of derived demand for peeler logs for plywood. Determination of derived demand for sawlogs. Forest management planning in Quebec. Supply and in Quebec. demand analysis allocation plan Comparison in Quebec. and choice among allocation plans Factors and relationships affecting bio- logical yield of silvicultural treatments. Hypothetical short-, medium-, roundwood supply curves. and long-term Hypothetical short-, medium-, demand curves for roundwood. and long-term Effects of changes in roundwood supply and demand on consumption and price. Roundwood production, man-hours paid, and production per man-hour in Nova Scotia, 1965-1979. Lumber production, man-hours paid, duction per man-hour in Nova Scotia, 1979. and pro- 1965- vi 33 37 61 63 65 66 67 68 80 81 82 99 112 114 115 125 129 Figure 18. Figure 19. Wood inputs in pulp and paper mills, man- hours paid, and man-hours paid/unit of wood input for Nova Scotia, 1965-1979. 132 Man-hours paid per m3 of wood input in sawmills compared to pulp and paper mills in Nova Scotia, 1965-1979. 134 vii CHAPTER 1 INTRODUCTION Statement of Problem Growth of manufacturing requirements and past timber management practices in Nova Scotia have resulted in a softwood resource base that cannot continue to sustain current and projected timber requirements. Much of the land harvested has not regenerated adequately. Stocking levels are low on some areas, while on others, regeneration consists of poor quality hardwoods that are uneconomical to harvest because of size or that cannot be used because of a lack of manufacturing facilities. Over the past five years spruce budworm has been a serious problem. During this period, it is estimated that some 9 million 1113 of soft- wood timber has been killed by budworm and an additional 9 to 10 million m3 is at high risk. Some of this volume will be salvaged but a considerable amount will not. Fur- ther, there is evidence that the proportion of merchantable timber volume available from small private woodlots is decreasing. Forest land is being taken out of timber pro- duction for other uses or the potential for timber produc- tion has become restricted for a variety of environmental reasons. Recent estimates for the Province show that without increased management inputs (or without seriously depleting the growing stock volume), the annual allowable cut (AAC) for softwoods would have to be reduced from the present level of 3.3 million m3 to 2.4 million m3, a reduction of one-third. Since actual harvest is now 3.3 million m3, some level of management inputs must be provided to sustain the yield and industrial output and employment. The problem and potential impacts vary for different geo- graphical areas and industry sectors. Spruce budworm damage is heaviest in the eastern region of the Province. Also, a reduction of harvest would likely affect sawmill operators most because they control only a small proportion of their timber requirements. There are several options available for increasing AAC and employment. Areas not regenerating naturally or regen- erating with undesirable species can be planted with genet- ically superior seedlings of desirable species. Dense young stands can be spaced to increase growth rates. Pro- tection against fire, insects, and disease can be expanded. Closer utilization can be practiced on traditional softwood stands, and manufacturing facilities and products can be altered to utilize low quality hardwoods. In recent years, increased effort has been put into these types of activities. Although individuals and forest products companies have provided limited funds for this work, the major source of financing has been the Provincial and Federal governments through a series of five-year regional development agreements. The problem is that funds are scarce relative to alternatives. It is questioned whether the current choice of activities is the best one - whether the objectives are being met most effectively. There are many options and each has varying biologi- cal, economic, social, and institutional impacts and limi- tations. Silvicultural activities such as planting will not yield merchantable timber for 30 or 40 years. Physical yield on different sites is likely to vary because of soil fertility or moisture. Availability and distribution of output will depend on land tenure. The impact of output will be influenced by future market conditions for end products, location and characteristics of manufacturing facilities, availability of alternative resources, and technology. In addition, environmentally acceptable alter- natives for effective insect protection are limited. Land tenure may restrict options on some areas. Transportation costs can prohibit distribution of timber from areas with a surplus to areas with a shortage. Considerable effort has gone into timber management planning in an attempt to provide guidelines as to the optimal choice of alternatives. In 1976, a timber demand/- supply simulation model was developed for the Province (N.S. Research Foundation Corporation, 1976). One of the purposes of this study was to provide current and projected demand and supply estimates for various geographical centers within the Province. It was anticipated that these estimates could then be used as a basis for identifying and evaluating alternative timber management intensification programs. Although this study required three years to complete and cost more than $300,000 to develop, it has not been used. There appears to be a number of reasons for this, including lack of background data, model complexity, cost of operating, and conceptual weaknesses. At present, the basis for timber management planning is an allowable annual cut model. This model tracks inven- tory over time, given various assumptions about harvesting and management inputs including silviculture and protec- tion. Future timber requirements are based on the current industry structure and inputs. There is no attempt to forecast demand and supply in an economic sense nor is there an explicit analysis of the cost of alternatives. Evaluation of the current planning approach shows that emphasis is largely on physical timber supply, that is, allowable annual cut. Little consideration is given to resource quantity and price at different processing sites over time. In addition, no attempt is made to forecast timber demand in the medium or long term. How are timber requirements likely to change over time and place, and what prices will processors be willing and able to pay for these inputs? There are a number of alternative planning approaches. One is the demand/supply simulation model discussed above. However, as indicated, this can be very expensive and might not provide any real advantages over current efforts. What is required is an examination of the specific problem, information needs to solve that problem, and direction as to how that information might be obtained. Objective and Scope of Study The main purpose of this study is to conduct a problem analysis for cost-effective timber management planning for the Province of Nova Scotia. This will include identifica- tion of the problem; information needs to solve the prob- lem; and general approach for obtaining the information. Although there are many individuals and groups in the Prov- ince which make timber management decisions, concern here is with the Provincial government. Also, while there are a variety of management alternatives, emphasis of this study is on silviculture and its effects on industrial roundwood production. Justification The Province of Nova Scotia is currently facing a serious softwood timber shortage. Estimates show that with no management inputs, the sustainable supply is about 2.4 million m3. Present harvest is 3.3 million m3. Any significant reduction of industrial production and asso- ciated employment is politically unacceptable. Over the past ten years, the Provincial and Federal governments have spent, on average, $7-8 million per year directly on forest management. Private nonindustrial land- owners and forest products companies are also contributing. This study is expected to provide information which will assist in more effective timber management planning. In addition, considerable effort is being put into research and technical studies. Approach The plan of this study is first to provide some per- spective on the development and current status of the forestry sector in Nova Scotia. Included in this is a description of the resource characteristics and ownership, their administration, resource use and change, and finally the type of manufacturing operations and product markets. This background section is followed by a chapter on problem identification. Who is the decision maker, what are his objectives and choice criteria? What is the decision maker's mandate and what are the alternatives available for meeting these objectives? Finally, what is the context of the problem - the factors, activities, and constraints over which the decision maker has no control? The fourth chapter provides a summary of past and current work relating to the problem including methods, experiences, and results. This is followed in Chapter 5 by an identification of problem components and general re- search approach. This includes the variables, their rela- tionships, and factors that influence them - the informa- tion needed to solve the problem. In addition, this chapter includes a discussion of how the information might be obtained. The sixth chapter outlines specific research studies required to obtain the information identified and their relative priority. In addition, an attempt is made to identify specific study objectives and alternative approaches. A summary and conclusions are presented in the final chapter. CHAPTER 2 DEVELOPMENT AND CURRENT STATUS OF FORESTRY SECTOR Forestry Resources Land Area Land and water area in Nova Scotia is estimated to be a little over 5.5 million ha. About 75% of this or 4.1 million ha is classified as forest land most of which is considered productive for forestry purposes. Although there have been two resource surveys in the past 20 years, there is little definitive information on changes in the forest land base over time. It is known that a consider- able amount of agricultural land has reverted to forestry, however, forest land has also been taken out of production for a variety of other uses including highways, transmis- sion lines, and housing. Ownership Ownership of forest land is distributed among three broad classes of owners as follows: Provincial Crown, 24%; Federal Crown, 3%; large freeholds with more than 400 ha, 21%; and small freeholds with less than 400 ha, 52%. This present tenure resulted primarily from opening up of the Province to refugees, disbanded troops, and Loyalists during and immediately following the American 8 Revolution. As a result, Nova Scotia has the second high- est proportion (73%) of private forest land ownership of all provinces in Canada. Only Prince Edward Island ranks higher (90%). For Canada as a whole, less than 10% of forest land is privately owned. As might be expected with this settlement pattern, there is a large number of small woodland owners. Recent figures show that for the small freehold group (owners with less than 400 ha with 52% of total forest land area) there are an estimated 30,000 owners. The mean size of ownership for this group is 100 ha (MacQuarrie, 1981). Except for this information, there are little data on land ownership change. Cover Type and Species Slightly over half (53%) of the forest land in the Province is classified as softwood (less than 25% hardwood by volume); mixedwood cover (26-74% hardwood volume) makes up 31% of the area and, the hardwood cover (more than 75% hardwood) makes up the remaining 16% of area. The forests of Nova Scotia contain about 280 million m3 of gross merchantable volume. Just over two-thirds of this is in softwood species, mainly spruces and balsam fir. Principal hardwoods are red and sugar maple, and yellow and white birch. Softwood volume distribution by size class is as follows: 24% in the 10-14 cm DBH class; 33% in the class 16-20 cm; and 43% in the class 22 cm and larger. 10 In terms of product potential this means that about 50% of total volume is suitable for sawlogs (30% conven- tional sawlog quality and 20% in studwood-sized logs), and 50% is suitable for pulpwood. For hardwoods only 16% of the growing stock volume is classified as suitable for sawlogs, 80% is classified as boltwood and pulpwood, and the remainder is estimated to be cull (N.S. Dep. Lands and Forests 1981b). Age Class Distribution Analysis of the age class distribution shows that the majority of forest stands is mature or overmature. More than 40% of the area is occupied by stands over 60 years old and 35% has stands 40-60 years old. In essence, the age class distribution is unbalanced and a high proportion is mature or overmature. The older balsam fir is particu- larly susceptible to insect, disease, and wind damage. The age class distribution for hardwoods is about the same. Resource Administration Primary responsibility for administration of the forestry resources in Nova Scotia rests with the owners - the private nonindustrial owners, companies, and Provincial and Federal governments. Within the Provincial Government this responsibility is held by the Department of Lands and Forests. Administration of Federal lands is divided among 11 the Departments of National Defense, Indian and Northern Affairs, and Environment. With the exception of the Department of Lands and Forests there is little information on size, organization, and operations of administrative staff. A rough estimate of employment in general administration, protection, silvi- culture, and so on for 1969 was 900 man-years or about 12% of the total forestry sector employment (Runyon, et al., 1973). Resource Use Primary Forest Production The Logging Sector The logging sector includes those businesses primarily engaged in felling and bucking, bunch- ing, yarding, forwarding, decking, and loading of round- wood. In Nova Scotia, like many other areas, this is a very heterogeneous group. It includes individuals such as farmers and fishermen who work in the off-season at log- ging. It includes sawmill crews who work part-time in the mill and part-time in the woods. It also includes full- time professional cutters who contract their services to landowners or stumpage buyers. Finally, it includes full- time loggers who are employees of large manufacturing firms such as pulp and paper mills. Logging methods and equipment vary considerably. These range from selection logging with horses or farm 12 tractors, to large-scale clearcutting with the use of skid- ders, forwarders, and whole tree processors. The most common method by far, however, is clearcutting with chain saws and movement to landing with rubber-tired skidders. Despite the fact that roundwood production has in- creased appreciably in the past 15 years, the number of logging establishments and employment have changed very little. In 1965, there were estimated to be 128 logging establishments in the Province employing 1200 persons. By 1978 the number of establishments had increased to 149 but employment dropped to a little over 1000 (Statistics Canada data). Due to the heterogenous nature of this group, how- ever, data are not felt to be very reliable. Timber Harvest Timber harvest in Nova Scotia has increased from 1.9 million m3 in 1935 to 4.3 million m3 in 1980. From 1935 to the late 1950's, harvest did not vary substantially except for the short-term cyclical changes (5-7 years). Since 1958, total roundwood harvest has been rising steadily (downturns in 1966, 1968, and 1974) from 1.9 million 1113 to 4.3 million m3. This increase resulted mainly because two new pulp and paper mills were built during the period. Softwood species have traditionally made up the larg- est proportion of total volume. In 1935, softwoods accounted for 87% of the total. This proportion increased to almost 98% in 1964 and then declined to 88% in 1980. 13 The increased use of hardwoods in recent years is due main- ly to pulp production. In 1967, slightly more hardwood volume went into sawlogs than into pulpwood. The use of hardwoods as sawlogs has not changed much since 1967 but their use as pulpwood increased 13 times. While data on the use of roundwood are not very definitive, there is little question that the proportion of total volume going into sawn products (lumber, ties, pit props, staves, and headings) as compared to pulp products has declined sub- stantially. Roundwood volume going into sawlogs has declined from 37% of the total in 1967 to 23% in 1980. This is substantiated by product output figures. Lumber production has stayed about the same over this period - about 470 000 1113 while pulp and paper producton has prob- ably doubled. Analysis of timber harvest by landowner group in the past 15 years shows the proportion of total volume produced from small woodlots has declined significantly. Proportion of production from large freehold land over this period increased, then decreased, and now remains at about the same proportion as it was at the beginning of the period. Production from Provincial Crown lands has increased more than four times during the period and the percent of total has more than doubled. In 1967, small woodlot owners (52% of total forest land in parcels of less than 400 ha) pro- duced 60% of total roundwood. This dropped to a low of 34% in 1979 and rose to 44% in 1980. Output from large 14 freehold properties (21% of forest land area in parcels over 400 ha) increased from 30% in 1967 to 44% in 1973, and has subsequently declined to 30% in 1980. Roundwood production from Provincial Crown land increased steadily from 11% of the total in the beginning of the period to 29% in 1979. By 1980 this proportion dropped to 26%. Other Removals Other removals include noncommercial harvest, logging residue, and losses due to fire, insects, disease, and wind. Data on these removals are limited but a number of studies are now underway to determine volumes. Noncommercial harvest includes cutting for fuelwood, fence- posts, and a variety of other products principally for farm use. Fuelwood is by far the most important of these. The estimated annual removal for fuelwood is 442 000 m3, most of which is hardwood species. An additional 100 000 m3 of softwood are thought to be used as fuelwood, but this is material already accounted for through sawlog removal (N.S. Dep. Lands and Forests, 1981b). Logging residue is not known. However, in calculating AAC, the Province assumes a loss of 15% as a result of cull and logging waste. Fire losses vary considerably on a year-to-year basis. In 1979, it was estimated that fire losses amounted to 269 000 m3 of softwood and about 80 000 m3 of hardwood. This represents about 6% of total volume removals. It is not known how much of this material was salvaged. 15 Spruce budworm is at present the most destructive forest pest in eastern North America. Although there has been a periodic problem with budworm in Nova Scotia in the past, its potential damage now is more acute because of the large proportion of mature and overmature spruce and fir and the industry's dependence on that timber. Since the inception of the current outbreak in 1974, it is estimated that the budworm has killed 9 million 1113 in the eastern portion of the Province alone. This figure is expected to double over the next 5 to 7 years. The estimated loss of 18 million m3 represents roughly five year's requirements of the forest products industry in the Province or 10% of the Province's softwood gross merchantable volume. There are expected losses in other areas of the Province as well (N.S. Dep. Lands and Forests, 1981b). It is difficult to estimate how much of this material is salvaged and how much is lost. Landowners are concen- trating harvest on these areas but due to the large volume involved relative to industrial needs it is unlikely that any significant proportion can be utilized. Nonconsumptive Uses The forests of Nova Scotia as elsewhere provide a variety of outputs - wildlife, recreation, and water qual- ity and quantity control. These outputs are important, however, because of their availability relative to 16 requirements, there is no problem. Some constraints have been imposed on consumptive uses to safeguard areas for aesthetics, for example, along highways and streams, and to protect municipal water supply areas. In the latter case, about 240 000 ha representing 63 surface water supply areas, have been identified for controlled use. Forestry is compatible within these areas, but there are constraints (N.S. Dept. Lands and Forests, 1981a). There are two national parks in the Province consist- ing of 83 250 ha. Forest management-silviculture and pro- tection (except fire) and utilization is not practiced in these areas. In addition, the Province has about 14 000 ha under park reserve, but harvesting is not specifically excluded from this area. Allowable Annual Cut/Harvest Balance AAC Model In the mid-1970's, an AAC model was developed by the Provincial Department of Lands and Forests. However, because of limited documentation and modifications over time it is difficult to get a clear picture of the model structure, assumptions, and data sources. To the best that can be determined (MacQuarrie, undated) the information base for the model is the continuous forest inventory. The land base is divided into 21 major categories (three ownership classes in each of seven subdivisions). 17 For each major category, the forest is further classified according to cover type, site, age, density, and stand nature (plantation, first or second rotation natural stand). The result is a total of 2,340 inventory types. For each type, area, volume per acre by species group and size, and basal area are determined from cruise line data.1 There are basically three types of model operations. These include changes in the data base - deleting or adding of land, harvesting, and management inputs. Changes in the data base can be made as new inventory data become avail- ablezo Area is removed (land set aside for a park), or added (agriculture land reverts to forestry). Harvested lands can be controlled through specification of the con- version period or period over which the forest is to be regulated, by specifying area to harvest in each of 10-year simulation periods, and by specifying volume to be harvest- ed in each simulation period. In addition, harvesting routines must be specified. For example, harvesting prior- ity might be to cut oldest age class first, softwood cover, 1Except for plantation data which were estimated from plantation surveys and second rotation natural stands volumes which were estimated from Nova Scotia normal yield tables. 2The basic inventory unit is the subdivision - there are seven in the Province, each with about 0.5 to 0.6 million ha. The inventory cycle is basically seven years. The last complete cycle was 1965-1971. At present, all but two subdivisions have been reinventoried. The latest cycle has been delayed because of the changeover from imperial to metric units. 18 or best site. Management options include clearing, plant- ing, cleaning, and thinning. AAC Compared with Harvest At present it is estimated that with no management inputs, the softwood AAC for the Province is about 2.4 million m3, and for hardwoods, 1.4 million m3. Annual softwood harvest for the period 1976-1980 averaged slightly less than 3.5 million m3 and was 3.7 million m3 in 1980.3 Average annual harvest of hardwoods (for commer- cial use) for the same period was 350 000 m3 and for 1980 was 523 000 m3- If it is assumed that firewood removal is in the order of 450 000 m3 there is still a sizable potential for hardwood utilization. As stated in the introduction, the Provincial and Federal governments and private landowners have been spend- ing on average about $7-8 million per year on forest man- agement. With this level of input or slightly more, about 12 000 ha per year can be treated (6000 ha site prepared and planted, 3000 ha salvaged, 4000 ha cleaned, and so on). The allowable annual cut of softwoods given this level of input is estimated to be 3.3 million m3 during the 1981- 85 period, rising to 4.1 million 1113 by the year 2020- 3Softwood harvest for the years 1977 through 1980 include wood salvaged from budworm-damaged stands and put into storage. The Department of Lands and Forests estimates that 3.3 million m3 of softwood and 400 000 m3 of hard- wood are required annually to enable industry to operate at capacity (N.S. Dep. Lands and Forests, 1981b). 19 Hardwood AAC is expected to remain the same since no treat- ments are scheduled to increase hardwood output. Forest Products Manufacturing and Markets Sawmills4 The sawmilling industry in Nova Scotia dates back to the arrival of European settlers in the early 1600's. Initially, planks and squared timbers were produced and shipped to Europe on ships making return voyages. With the expansion in shipbuilding, local construction needs, and export demand, the industry grew to a peak level in the mid 1800's. By 1861, there were over 1400 mills in the Prov- ince employing over 3000 men. A century later, the number of mills had declined to slightly over 500. This downward trend continued until 1971 when the number of mills dropped to 279. Over the past 10 years, the number of mills has been increasing and by 1980 totalled 347. The reduction in number of mills was accompanied by a decrease in lumber production. The highest production level in recent years was 1.4 million 1113 in 1951. By 1961 this figure had drOpped to less than half, or 0.6 million m3 and by 1971 was down to 0.4 million. Over the 4Sawmills are defined as mills that produce lumber only, boxwood only, or lumber in combination with boxwood, staves, headings, shingles and so forth. 20 past 10 years, production has levelled off at about 0.5 million m3- A look at the industry structure shows that it is characterized by a few large mills which produce most of the output. In 1980 only five mills (of a Provincial total of 347) produced more than 14 000 m3 and only 15 produced more than 7500 m3. These 15 mills (42 of the total) produced about 50% of total output in 1980. By contrast, the smallest size classes (those producing less than 2350 1113 per year) numbered 300 but produced only 20% of total output. While the percentage of mills by size class has not changed appreciably over the past 20 years, the larger mills have been producing a higher proportion of total volume. Horizontal and vertical integration are limited. A few of the larger mills are affiliated with pulp and paper companies but the extent of this is probably much lower than in the rest of Canada or the United States. There is likewise little vertical integration mainly be- cause the principal product is construction lumber. Hard- wood lumber makes up less than 5% of the total, primarily because of the low quality of the resource. The principal market for lumber is within the Province (traditionally about 70%) with small quantities going to adjacent provinces, the United Kingdom, and the United States. This has not changed much over the past 20 years. At present, the sawmilling industry is a major employ- er in the Province and most of the jobs are provided in 21 rural areas where unemployment is high. In addition, the industry provides a local outlet for woodlot products as well as a local source of building materials. Because of the substantial impact of the industry, problems and solutions are a real concern. The principal problem in the sawmilling industry (which probably results in large part from the structure) is in obtaining an adequate supply of quality sawlogs. Owners can meet only 20-30% of requirements from their own lands; there is a high dependence on small freehold and pulp and paper company freehold lands for the remainder. In most cases there are no long-term contracts. Without an assured supply of logs, companies are reluctant to modern- ize or expand. This limits productivity and consequently, incomes (Runyon et al., 1972). Pulp and Paper Mills The pulp and paper industry in the Province consists of four mills. Two of these produce newsprint (one of these produces market pulp as well), one produces ground- wood pulp and paperboard, and one produces bleached kraft pulp. One of the newsprint mills and the bleached kraft pulp mill are relatively new - the newsprint mill was built in 1957 and expanded in 1970, and the bleached kraft mill was built in 1964. This industry uses about 80% of total roundwood pro- duction in the Province - 3.3 million m3 in 1980. About 22 15% of this was hardwoods. In addition, about 300 000 m3 of sawmill chips were used. Roundwood source is split about equally among company freehold, small freehold, and Crown lease. An ongoing problem has been a lack of depend- able supply from small freehold property. In addition, concentration of spruce budworm damage in the eastern part of the Province has created a severe problem for one of the mills. Virtually all pulp and paper products are exported mainly to the United States eastern seaboard area, the United Kingdom, and other countries within the European Economic Community. Other Forest Products The only other significant manufacturer (in terms of wood inputs) is a hardboard plant. This plant uses about 225 000 m3 of wood, about 75% of which is hardwood. One- third of the hardwood output is sold within Canada while the remainder goes to the United States and the United Kingdom. CHAPTER 3 DEFINITION OF THE PROBLEM The Decision Maker Although there are many individuals and groups in Nova Scotia who plan and make decisions on timber manage- ment, the concern here is with the Provincial government - specifically the Department of Lands and Forests. This is not to say that others do not have similar problems and choices. Decisions are influenced at the Provincial level by other departments such as Development and Environment. Also, since a large amount of forest resource development funding comes from the Federal government, a number of these departments share in the administration of funds. Finally, since some management activities take place on private land or involve forest products manufacturers, they also influence the opportunities and choices. The Department of Lands and Forests is not the only group with timber management problems. Most other provin- cial governments in Canada are concerned about present and projected timber shortages and potential effects on employ- ment and incomes. Forest products opportunities may differ from those of the Department of Lands and Forests but 23 24 the problem of choice among a number of management alternatives given varying constraints is the same. Objectives and Criterion for Choice It is difficult to find explicit governmental or de- partmental statements of objectives, however, there have been several joint federal/provincial forestry development agreements that state very clearly the objectives of the agreements. These appear to be the best indication of the Provincial objectives. In 1972, the A.R.D.A.1 III forestry program was signed. This was a five-year program funded at $16 mil- lion. Objectives were ”(a) To improve the land ownership pattern, now extremely fractured into numerous small hold- ings, by increasing Crown holdings and con- solidating others into more viable units. (b) To improve forest management practices on Crown and freehold lands in the province with sub-objectives as follows: (i) To produce wood in perpetuity for the forest industries. (ii) To increase yields per acre on all capable forest lands. (iii) To Optimize the inputs of capital and labor so as to achieve the greatest socio-economic benefits to the prov- ince. (iv) To increase efficiency in production activities. 1Funding was authorized under the Agriculture and Rural Development Act (A.R.D.A.) administered by the Federal Department of Regional Economic Expansion. 25 (v) To obtain higher income per capita for those engaged in use of this primary resource, and consequently improved living standards and working condi- tions. (c) To assess the supply/demand relationships of the forest resource. (d) To supply the public and the news media with a greatly increased flow of information on wise utilization of the forest resources in the province, and to provide for technical train- ing" (N.S. Dep. Lands and Forests, undated). This program was followed by a Canada-Nova Scotia Subsidiary Agreement for Forestry. The term for this agreement is 1977-1982, and although initially funded for $25.6 million, was subsequently increased to about $60 million. Objectives for this present program are ”(a) to maintain the existing aggregate employment based mainly on the softwood species; and (b) to increase employment opportunites and incomes in timber harvesting and processing based mainly on the hardwood species" (Can. Dep. Regional Expansion and N.S. Dep. Lands and Forests, 1977). Elaboration of these objectives is provided under Scope of the Agreement with the following statements: ”Within the framework of an overall development strat- egy for Nova Scotia, the policy of the Province is to maintain and increase employment and income opportu- nities for that segment of the population living in small urban centres and rural areas. The forest indus- try of Nova Scotia has long been important to the economy of the Province and to rural areas, in partic- ular. The aim of this Agreement is to stabilize and increase employment and incomes derived from harvest- ing and processing this important renewable resource.’ A further agreement is currently being planned for the period 1982-87. The initial draft has been prepared and the tentative objectives are 26 "To increase the supply of wood fiber to maintain or increase aggregate employment directly dependent on the forest resource” (N.S. Dep. of Lands and Forest, 1981b). It is further stated that an improvement in the soft- wood supply should provide greater opportunities for using low quality species and, ”Therefore, a compatible secondary objective of this Agreement is to increase employment and incomes through the realization of identified income-gene- rating opportunities.' This second objective is aimed at increased hardwood utilization and Christmas tree production. Since this initial draft was written, there have been indications of changes in objectives and programs to put greater emphasis on the provision of employment and incomes in rural areas. The specific implications of this are not known but it is conceivable that programs and projects (forest management on small private woodlands) will be identified for particular individuals or groups in selected geographic locations (areas of high unemployment). Although these objectives appear to be straightfor- ward, there is some question as to whether they accurately reflect Provincial views. The forestry development agree- ments are Federal/Provincial cost-shared with the Federal share traditionally making up about 80% of total funds. Federal expenditures are administered by the Department of Regional Economic Expansion whose principal concern is employment and incomes. The Provincial Department of Lands 27 and Forests' focus is on wood supply - specifically, to increase AAC to meet present industrial requirements. This combined with increased utilization of low-quality species, mainly hardwoods, is assumed to be sufficient to maintain employment and incomes in the Province's forestry sector. This difference in focus is not surprising. The Department of Lands and Forests is a resource management agency. Its principal concern is resource development rather than economic or social development. The Federal department, on the other hand, sees forest resource devel- opment as one of a number of means to achieve employment and income goals. It might be questioned whether there is any substantive difference in the two approaches. The only apparent difference appears to be in where the emphasis is put. In this study, it is assumed that the Province's objective is to maintain employment and incomes at the current level in the forestry sector in the most cost- effective way. There are changes taking place in end- product demand, labor productivity, and so on, that will affect employment and incomes. It is not sufficient to assume that desired targets will be met simply by ensuring that softwood AAC will be large enough to meet present requirements. At this point in time, it is not possible to identify specific employment and income targets by geographical region or sector but these are obviously real political 28 constraints. These are discussed further under the problem concept. The criterion for choice among alternatives is minimum cost to the Provincial government or where programs are federal/provincial cost-shared to minimize costs to both governments. Alternatives Governmental Mandate in the Forestry Sector There is a variety of laws and regulations that give the government authority, responsibility, and the means to pursue the above objectives. Basically, the governmental mandate can be separated into three areas based on type of responsibility. These are management of Provincial Crown lands; making and enforcing legislation affecting manage- ment of private lands, sale of timber, transportation, and manufacturing and marketing of products; and administration of funds for forest management and manufacturing. These are described in more detail below. Management of Provincial Crown Land Authority for management of Provincial Crown land is provided specifical- ly through the Lands and Forest Act. This states, ”The Minister of Lands and Forests shall have under his supervision the management of Crown Lands, the conservation of all forests and timber lands and the protection of game and game fish" (MacQuarrie, 1974). Under this Act, the Minister has the authority to set aside that land for reserve, protection of water quality; lease the land if, for example, the licensee (leasee) undertakes 29 to spend money in building a mill for wood products manufacturing; or grant licenses to cut timber. At present, there are about 985 000 ha of Provincial Crown land (24% of total land area). Roughly 70% of this is under long-term licence to pulp and paper companies. In addition, there are a number of timber cutting licenses granted to other types of manufacturers. Legislation Affecting Other Lands and Products The principal authority for control of private lands by the government comes from the Forest Improvement Act passed in 1968. The purpose of this Act is to "...promote forest growth and cultivation.' (MacQuarrie, 1974.). Various sections of the Act include provision for registration of commercial Operators - loggers and buyers; designation of green belts along highways and streams; restrictions on cutting of immature trees; regeneration of cutover lands; establishment of a Timber Loan Board to provide funds for forest land purchase; and establishment of a Forest Prac- tices Improvement Board, which would have broad responsi- bility for implementation of the Act. Although the Act was passed in 1968, most sections were not proclaimed into law until 1978. Regulations for some sections have not yet been written. The Department of Lands and Forests Minister has responsibility for the Pulpwood Marketing Act. This Act is administered by a five-member Board appointed by the Minister and it may 30 ...from time to time make regulations, (a) regulating and controlling the marketing of pulpwood, including the agency or agencies through which such pulpwood may be marketed in Nova Scotia; (b) providing for the licensing of persons regularly engaged in the marketing of pulpwood" (MacQuarrie, 1974). In addition to the above, the Lands and Forests Minister has direct responsibility for administering acts that deal with parks development, game, land taxes, sur- veys, and scaling of wood. Other legislation affects forests and forest products through environmental, trans- portation, manufacturing regulations, export of roundwood, and other products and labor laws. Administration of Funds The means by which the above responsibilities are carried out include direct financial appropriations of the Department of Lands and Forests and other provincial departments, and through special sector agreements with the Federal government. Historically, the Lands and Forests' budget has been directed mainly toward management of Crown lands (with the exception of fire pro- tection), tree nursery operations, and general administra- tion of acts under the Minister's authority. In addition, the Provincial Department of Development provides reduced interest loans for forest products firms. In recent years, the main source of funding for forest management in the Province has been through joint Federal/ Provincial agreements. As indicated earlier, the Federal share of funding is about 80%. These agreements are plan- ned and administered by a Management Committee consisting 31 of Federal and Provincial representatives. The Provincial Department of Lands and Forests is responsible for imple- mentation. Forestry Sector Employment Activities At present there are slightly more than 8,000 people directly employed in Nova Scotia's forestry sector. The Department of Lands and Forests estimates that there are 5,600 people directly employed in pulp and paper mills and related activities. Sawmills and dependent sectors employ an estimated 2,400 people (N.S. Dep. Lands and Forests, 1981b). There is a problem in obtaining reliable employment figures by activity. Statistics Canada collects annual employment data through a census of manufacturers. This includes logging establishments, sawmills, pulp and paper mills, and other wood-using industries. Data do not include government employees (for example, Lands and Forests staff), or other service groups such as consul- tants, truck drivers who haul wood, and silviculture staff unless these people are full-time employees of a logging company, sawmill, or other manufacturer. Statistics Canada does not distinguish as to type of employment. For example, figures are presented for pulp and paper mills but they are not broken down as to how many work in the mill as opposed to in the woods. Many sawmills are operated on a seasonal basis and employees may work in the mill part-time 32 and in logging operations part-time. Some pulp and paper company logging crews also do silvicultural work. The problem is compounded because many of the logging and sawmill operations are small, they are not surveyed. Individually, they are not significant, but collectively, they can be quite important. To identify alternatives for generating employment and incomes, it is important first to identify types of employ- ment activities and interrelationships. There are several different ways to classify activities. The one shown in Figure I appeared to be the most appropriate, given con- cerns of this study. In this Figure, three broad categor- ies are delineated. These are general administration and support, forest management, and utilization. Utilization if further subdivided into consumptive and non-consumptive use. General administration and support includes planning, inventory, mapping, education and extension, and research. There is little information on current employment in this group, but the largest employer would undoubtedly be the Department of Lands and Forests. Most of the larger com- panies, however, do have some staff responsible for this work. Forest management is used here to include basically the operational aspects (as opposed to support) of forest production. This category can be further subdivided on the basis of principal products desired (wildlife, timber), or 33 General administration and support l Forest management Utilization non— consumptive use Maple lsyrup| Outdoor recreation Wildlife 1 l r l consumptive use Industrial products Non-industrial products Harvesting l Transporting Processing Figure 1. Forestry sector employment activities. 34 types of operations (silviculture, protection). In this study concern is with management Operations for industrial timber production. Activities include nursery production, site preparation, planting, weeding, cleaning, thinning, protection, and roadbuilding for stand establishment and tending purposes. On Crown lands, management is generally done by staff or contractors. Private land operations are carried out by small woodlot owners, company crews, or contractors. Some data are available on employment and incomes from these activities but they are not considered to be accurate. Utilization is separated into non-consumptive and consumptive activities. The former includes maple syrup production, wildlife management, and the provision Of out- door recreation. The three major categories of consumptive use activities are harvesting, transporting, and processing roundwood for industrial use, and other non-industrial products such as Christmass trees and firewood. This study focuses on roundwood for industrial use. Alternatives For Increasing Employment and Incomes There are a number of opportunities for expanding employment and incomes in the forestry sector as shown in Figure 1. Increased funding (or changes in existing polic- ies and programs) can have both a direct and indirect impact. For example, increased spending for extension staff will directly result in additional jobs and incomes 35 but it may also improve forest productivity on private lands, which raises the potential for harvesting, trans- porting, and manufacturing. The potential from increased inputs into forest management is well known and is discussed further in the next section. In the area of utilization, funding for research and development in logging methods can reduce costs of harvesting marginal stands and, in effect, increase AAC. Upgrading of roads and highways might permit logging in more remote areas. Finally, processing facili- ties can be modernized to permit use of species, which are presently not considered economic, or to improve recovery in pulping or sawmilling. Each of these alternatives can result in increased employment and incomes directly or indirectly through their effects on AAC. There is the possibility, however, that some inputs will actually decrease the number of jobs. Cleaning Of dense young stands might permit more mechanized logging which reduces labor requirements. Modernization of mills (automation of carriages, edgers, trimmers, sorters in sawmills) will in most cases reduce employment. This latter example points to a potential conflict in Objectiv- es. Increased labor productivity in any of the employment activities identified means that fewer people can produce the same quantity/quality of output. But productivity increases may result in higher wages and salaries. The question arises as to what the trade-off is with respect to 36 employment and income. Is the Province better-Off with 5,000 people employed at $10 per hour or 10,000 people at $5 per hour? Since both employment and incomes are impor- tant, both have to be identified. Concern of this study is with a subset of forest management, namely silvicultural opportunities for increas- ing employment and incomes through increased industrial roundwood production and utilization. These Opportunities are emphasized because most current funding is being directed here and because these appear to Offer the most potential, given the Province's Objectives and constraints. The specific relationship Of silviculture with other management activities is shown in Figure 2. Increases in industrial roundwood production can be obtained by alternatives which improve productivity of a given land base or by expanding the land base. The dichotomy here is not entirely distinct. For example, some type of drainage may bring additional land into production while in other cases, it simply improves productivity. There are several limitations with regard to access. There may be problems of a physical nature - there are no roads into an area. It may be an institutional constraint. Many small woodlot owners do not practice any management. Land tenure change may result in more land being brought into production. 37 Forest management for industrial roundwood f | Forest land Forest land productivity change base change Afforestation Silviculture [ Drainage J [ Protection J I Access J Figure 2. Forest management alternatives for industrial roundwood production. 38 Scope of Study The foregoing discussion only partially indicates the level of decision making Of concern in this study. It is recognized that there are employment opportunities outside the forestry sector and the Provincial government must ultimately decide on whether to allocate expenditures to forestry or some other sector such as agriculture, fisher- ies, or manufacturing. This is an intersectoral decision level. Within the forestry sector there are various alter- natives as shown in Figure 1. There is a decision level within major activities, for example, within forest manage- ment. This is the level of concern in the present study. Finally, there is a more detailed level of alternatives within principal treatments. For example, in cleaning, what is the Optimum spacing, or for planting, what is the best species/site selection or optimum number of trees to plant per hectare? In effect there is a hierarchy of decisions and ideal- ly the "best" a1ternative(s) at each level would be deter- mined and compared. This would be done until an optimum was reached for all government alternatives. In this study, it is assumed that the optimum number of trees to plant per hectare, rotation age, thinning regime, and so on are known. Concern here is with the optimal treatment with respect to geographical area, owner- ship, site, and stand condition class. Decisions with respect to Optimal rotation, spacing intensity, and so on 39 are important but data are very limited. There are basically three types of silvicultural treatments of concern in this study. These are planting, cleaning, or juvenile spacing and thinning. It is helpful for evaluation, to further delineate these alternatives into five treatments. These are 1) planting non- satisfactory restocked (NSR) lands; 2) planting recent cutover, burned, or insect-damaged areas; 3) cleaning; 4) commercial thinning; and 5) the combination of cleaning and thinning. These have been identified on the basis of what is currently practiced or contemplated and what is thought to be feasible given past experience within the Province and elsewhere. Context of the Problem The silvicultural opportunities identified above, their costs, and impacts on employment are affected by several biological, economic, technological, and institu- tional factors. The Province, as decision maker, has no control or little control over many of these, yet they cannot be ignored because they constrain, modify, and directly determine the Opportunities and impacts on the objective. These are categorized below on the basis of whether they affect opportunities or impacts. 40 Factors Affecting Silvicultural Opportunities The Department of Lands and Forests has estimated that there are some 400 000 ha of mature and overmature forests in the Province and that these stands should be planted after cutover. Although it is known where these are, who owns them, the site type, and so on, it is not known which ones are available for treatment each year. It is not known which ones are harvested. Harvesting decisions are made for the most part by small woodlot owners and compan- ies. At the present time, even the annual area of harvest for the Province is not known. Method of harvest affects planting Opportunities. While mature and overmature stands are generally clearcut, some are probably cut on a selec- tion or shelterwood system. Areas that should be cleaned or thinned are known. Opportunities for silvicultural treatments are con- strained by the landowner and his Objectives. It is esti- mated that there are about 30,000 small woodlot owners in the Province (each owns less than 400 ha; as a group small owners hold 52% of forest land area). A recent survey of these owners (MacQuarrie, 1981) showed that 72% had not sold any forest products in the last 2 years. Also, over 53% of owners surveyed indicated that they had not con- sidered a management plan - simply because they had no interest or felt the property was too small. About 60% of the forest land area (and 60% of production) is considered to be available for commercial production and use. This is 41 the land area controlled by those who stated that they intended to harvest, sell products, or produce Christmas trees in the future. Availability Of skilled labor can limit silvicultural alternatives, but the Province has some control over this factor. In the current forestry subagreement there is a project designed to organize forest management crews. Alternatives may be limited by political considerations. At the present time, it is doubtful if silvicultural pro- grams could be concentrated on a specific geographical area and landowner class. For example, it is unlikely that the Province could designate funds for planting on small free- hold lands in, say, the eastern region of the Province only, or on large freehold properties only. Environmental constraints may limit areas that can be planted. If speci- fic areas require a certain type of site preparation or use of herbicide to control vegetative competition, it might not be feasible to plant these areas because of environmen- tal regulations. The production of seedlings requires that there be a seed source. At present, some of these sources are severely threatened by spruce budworm and the spruce coneworm. Use of chemicals to protect against these insects is not permissible and the effectiveness of biological control measures is questionable. With economic and technological change, it is probable that new silvicultural Opportunities will be created in the future. Most Of the planting done now is with bareroot and 42 container seedlings. These are expensive to produce and plant. Direct seeding problems and solutions are being examined. Fertilization is a potential alternative but preliminary cost and response figures do not look favor- able. Future alternatives will be affected strongly by harvesting and manufacturing technology. Most of the cur- rent treatments are aimed at producing higher quality or larger softwoods. Alternatives to increase production of biomass for both softwoods and hardwoods will undoubtedly become more important as costs of traditional forms of fiber go up and as harvesting and manufacturing methods improve. Factors Affecting Cost Of Alternatives The major cost components for each of these treatments are labor, energy, capital, and materials and supplies. The degree of importance of these inputs by treatment, the factors that affect them, and therefore treatment costs are discussed below. Planting Planting involves the collection of seed, production of seedlings, transportation, site preparation, setting in the ground, and possibly follow-up treatment with herbicide to control vegetative competition. Labor is probably the major cost component followed by energy, capital, and supplies. Methods of seed collection include gathering cones from commercial harvesting operations, from plus trees or stands, or from seed tree orchards. The 43 latter method is the most expensive but it offers the high- est potential genetic improvement. At present, insect damage can be serious and, therefore, affects the cost of seed collections. Seedling production is most commonly done in green- houses for container stock, or greenhouses and outside transplant beds for bareroot stock. Major costs include labor for seeding and weeding containers or flats, energy costs for heating greenhouses, capital and maintenance costs of the greenhouse, and supplies such as containers. The decision maker has little control over unit costs (labor rates, fuel oil price) but can affect overall costs through the use of more automated equipment, installation of heat blankets in the greenhouses, and so on. Insects and diseases can be a major problem. Seedling transportation costs are not too significant. The Province has control over where production facilities are located but there are economies of scale. These have to be balanced against cost of transportation. Site prepa- ration methods and costs depend on the harvesting system, time since harvest, type of species to be planted, and environmental considerations. Methods include cutover clearing, prescribed burning, deep plowing, and crushing. Planting has been separated into two treatments, nonsatis- factory restocked lands and recent cutover, because in the former, cleaning the site involves considerably higher costs. Setting of seedlings in the ground is the most 44 expensive component of this alternative. Most seedlings are planted by hand. Topography, logging slash, and rocky sites limit the use of the more capital intensive equip- ment. Herbicide treatment is Often required after planting to control hardwood competition. At the present time, methods and chemicals used are being closely examined because of effects on environment. If these alternatives are seriously restricted, it would not be feasible to plant on many sites. Control of competition is required to ensure an acceptable survival and growth rate, and mechani- cal control Of competition is prohibitively expensive at the present time. Cleaning This operation is recommended for stands that regenerate too thickly. Areas with a high balsam fir component (or hardwood) normally regenerate well after cutting. Left alone, the stands result in reasonably high volumes per hectare but volume is distributed over a large number of small diameter trees. Harvesting, handling, and processing are expensive due to tree size. Cleaning or spacing of stands is recommended at 10 to 15 years of age to remove a high proportion of the stems as well as unwant- ed species. This is generally done by hand with a brush saw. It is labor intensive and costly. Various types of mechanical methods are being tried but to date have not been too successful because of the variable nature of terrain and difficulty of being selective. Patch and strip methods with mechanical equipment are being examined. 45 Commercial Thinning Thinning is proposed for stands where sawlog production is desired, or as a means of sal- vaging dead and dying trees caused by insects and disease. The method is straight forward - generally felling with chainsaw and use of tractor or skidder for removal to road- side. Cleaning and Commercial Thinning This treatment is thought to be feasible for stands that regenerate well naturally but with too many stems per unit area. Where sawlogs are desired, stands can be cleaned or spaced at 10 to 15 years of age and thinned at 30 to 40 years of age. In addition to the factors identified above, costs of treatments will vary by size of operation and landowner. Unit area costs are generally cheaper for larger areas because of reduced time and cost of moving equipment and personnel. Costs by landowner group will differ because of skill and experience of operators. Larger companies gener- ally have specialized silvicultural crews and equipment whereas the woodlot owner with 100 ha generally does not. Factors Affecting Impact of Silviculture The ultimate reason for undertaking silvicultural work is to generate employment. There are two potential sources of employment resulting from these treatments. Jobs are created directly in the Operations - greenhouses, cleaning, and so on, - and as a result of the outputs of these 46 operations, through changes in quality, quantity, and cost of timber produced. Employment from Operations depends on the alternative, method, and amount Of area treated. Decisions as to choice of alternative and area are largely controlled by the Department of Lands and Forests through funding alloca- tions. The method, that is, whether cleaning is done with a brush saw or specialized equipment, in most cases is decided by the landowner. Employment resulting from outputs of these operations is more difficult to estimate. Basically, the need is to determine the increased quality and quantity yield from the treatments in time and space; changes in logging, harvest- ing, and manufacturing, and resulting employment. At present, estimates of these effects are being made but they are based mainly on current relationships, that is, employ- ment per unit of wood harvested, transported, and manufac- tured now. No attempt is made to estimate possible changes in industrial roundwood demand, manufacturing processes, industry structure, and employment impact over time. There is some question about how critical this is to the decision maker - the Province. Present concern is basically with resource supply, which appears to be typical of most resource management agencies. Part of this is due to the fact that other agencies have primary responsibility for processing. Also, the lack of definitive information on biological results of treatments including the impact of 47 insects, responses from small private woodlot owners, and uncertainty about timber allocation from large private owners, appears to diminish or mask the importance of information on future timber demand. Despite this prevailing attitude, there is growing recognition within the Province of the need to consider demand aspects more closely. The assumption of this study therefore is that while the current approach to estimating impacts of treatments might appear to be adequate now, there is a need to provide a more comprehensive framework for both timber supply and demand analysis. The Province has and is putting considerable effort into AAC projec- tions. Very little is being done with respect to identify- ing medium- or long-term roundwood demand. Technological changes are taking place in harvesting equipment such as increasing use of whole-tree processors, and in manufacturing processes such as thermomechanical pulping and increased use of hardwoods. In sawmilling, size of logs is becoming less critical with more automated equipment. There is increasing concern over biomass as opposed to conventional merchantable volumes. Future developments that reduce limitations as to what is merchan- table are particularly critical to the cleaning alterna- tives where biomass with and without treatment may not differ significantly. Although it is important to recognize the need for more comprehensive supply and demand analysis, there are 48 real limitations regarding data, staff, and funding avail- able for research and planning. Most forest products companies probably have a larger staff and budget for this than does the Department of Lands and Forests. CHAPTER 4 PAST AND CURRENT WORK The purpose of this chapter is to review pertinent work with respect to data availability, methods, and results. This is categorized into three major sections as follows: direct assessment of silvicultural Opportunities; timber supply and demand analyses; and contribution Of timber sector activities. In the second section, special emphasis is put on a demand/supply model developed for Nova Scotia. This model was designed, in part, to answer questions identified in the present study. Direct Assessment of Silvicultural Opportunities Probably the most relevant work in this area with respect to Nova Scotia is a recent study titled, "The Atlantic Region Silviculture Program: An Evaluation" (Can. Dep. Reg. Econ. Expansion, 1981).1 The objectives of this study were to evaluate economic benefits of public expenditure on silviculture, to assist in the evaluation Of the General Development Agreement, and to improve 1Subsequently referred to as DREE study. 49 50 methodology in determining impacts of silviculture. Three distinct Objectives of silviculture programs were 2 These are l) to increase employment and identified. incomes; 2) to maintain employment and incomes in forest industries; and 3) to provide temporary employment. The first Objective is interpreted to be economic growth measured as GNP (gross national product) or GPP (gross provincial product). Since silviculture is only one of several alternatives for increasing growth, the evaluation should be based on "...... whether investing in silvicul- ture is at least as good as investing in an 'average' sector in the economy." The appropriate criterion is thought to be net benefits which are gross profits, returns to labor, foreign exchange adjustment, and the negative benefit (cost) for energy subsidies resulting from har- vesting and processing timber outputs from silviculture. Costs include capital (greenhouse cost capitalized over its life and apportioned to seedling production), site prepar- ation, planting, road construction, stand improvement, and research and development. From these figures, a weighted average cost for reforestation and stand improvement is determined. Cost per unit of timber output is calculated and compared to benefit per unit output. 2These objectives are stated in the forestry subagree- ments which are Federal/Provincial government cost-shared. Presumably the objectives accurately reflect not only the view of the Federal government but also those of the Province. 51 The basic assumption of this approach is that "current harvesting rates would be maintained in the absence of a silviculture program. Thus no benefits arise until treated stands mature." In other words, no allowable cut effect is attributed to silviculture. This is considered a pure investment approach. Results show that net present bene- fits from both reforestation and stand improvement are negative (-$3.18 and -$1.41/m3, respectively). In the analysis using the second objective, that is, to maintain employment and incomes in forest industries, it is assumed that in the absence of a silviculture program, the Province would cut back on harvesting to the sustained yield level.3 It is also assumed that any increased volume growth will have an allowable cut effect (ACE) now.4 Therefore, benefits accrue now. Using this approach the net benefit is $9.20/m3. This alternative is seen as a community stability perspective as opposed to an economic growth perspective. The conclusion drawn from this study is "... that silviculture is preferable to cutting back on current harvest rates if the province is following a policy of 3Current AAC for softwoods at the Provincial level is less than the annual harvest, assuming no silvicultural inputs. 4Allowable annual cut (AAC) is the sustainable harvest of timber based on current inventory and anticipated changes. The allowable cut effect (ACE) is the additional harvest permitted now resulting from an anticipated increase in growth and yield due to some silvicultural or management treatment undertaken now. 52 restraining the AAC to the sustainable yield. However, silviculture is not a productive investment and should not be used as a means to expand incomes and employment." It is further stated, "The key question then is whether the provinces are engaged in silviculture in order to maintain social stability (the ACE perspective) or to increase income levels (the investment perspective). If the former is their goal, then silviculture is indeed worthwhile. If the latter, then other investments will show a greater return" (Can. Dep. Reg. Econ. Expansion, 1981). What are the implications of this work for the present study? It is Obvious that one's interpretation of stated objectives is critical. Is the concern with timber supply or employment and incomes? If the latter, is it with employment or income or both? Is it realistic given biolo- gical, institutional, political, and financial constraints to increase employment and incomes? The appropriate Objec- tive (and assumptions) is critical to the evaluation criteria. Earlier, in the problem definition it was stated that the objective as interpreted here was to maintain current employment and incomes. Given the gap between AAC without management inputs (2.4 million m3) and the current harvest level (3.3 million m3), it is felt that an Objective of increasing employment and incomes over the current level would be unrealistic. At present, the Province is relying heavily on federal funds simply to maintain an AAC equal to harvest. 53 Whether the ACE perspective is valid under either Objective identified by the DREE study depends first on the structure and composition of the forest. Apparently, since there is a large proportion of softwood timber in the older age classes, AAC and harvest could be increased immediate- ly. But it does not seem reasonable to assume the ACE perspective for one objective and not the other. If AAC is below industrial requirements it would be expected that any increase in AAC would be utilized. However, even if the AAC (through ACE) became higher than present industrial requirements, there is no reason to believe that industries would not expand (either new mills or expansion of existing ones). Are the criteria of net benefits as calculated approp- riate? Incomes as measured by gross profits, wages, and salaries, and foreign exchange are important but the Province's concern seems to be with employment. Nova Scotia's unemployment rate has traditionally been higher than the Canadian average. However, as noted earlier, it is not realistic to look at employment numbers without also considering incomes. The focus of the DREE study is at the Provincial level and there is no real attempt to evaluate alternatives within the silviculture program. As a result, there are limited data presented which are useful in distinguishing impacts of Opportunities by geographical area, landowner group, site, or specific silvicultural treatment. This is 54 not criticism of the DREE study, but simply a recognition Of its scope and limitations. A second significant study relating to forest manage- ment alternatives in Nova Scotia is an evaluation of the forestry subsidiary agreement (MacLaren Plansearch, 1981). This Federal/Provincial agreement consists of four programs (forest management, industry development, support services, and education, information, and evaluation). Within the four programs there are 13 projects (including forest management on private lands, salvage of spruce budworm damaged timber, and new uses for hardwoods). Objectives of the evaluation were to measure effectiveness and efficiency of programs and projects, identify reasons for performance, and to make recommendations for change and form of subse- quent agreements. The objective of the Agreement was identified to be, to "facilitate joint federal-provincial cooperation in initiatives for the economic and socio-economic development of Nova Scotia." Goals were "a) to maintain existing aggregate employment based mainly on the softwood species; b) to increase employment Opportunities and incomes in timber harvesting and processing, based mainly on hardwood species." The forest management program is of direct concern here. It includes five projects. These are forest manage- ment of private lands; forest management of Crown lands; salvage of spruce budworm damaged timber; group management 55 of private lands; and wood salvage/storage. The program objective is to maintain existing aggregate employment based mainly on the softwood species. The goal is to contribute towards the long-term supply of timber to forest industries in Nova Scotia. Conclusions of the evaluation study are that the forest management program should increase AAC and therefore contribute to long-term wood supply. On the basis of this, the program should contribute "... towards the long-term stabilization of employment." However, no attempt is made to identify how much AAC is likely to increase. This is considered to be beyond the mandate of the evaluation. This evaluation study therefore provides only limited information with respect to selection of optional silvicul- tural activities. However, it does provide a source of data regarding costs of those activities. For example, cost of site preparation on small freehold properties averaged $178/ha cleaning cost was almost $400/ha; and thinning averaged $470/ha.5 Work by Marty and Newman (1969) in the United States provides an approach to identifying and ranking timber management opportunities on national forests. Management activities included reforestation, precommercial thinning, and commercial thinning. Sixty timberland classes were 5Cost as used here is actually the level of assistance paid by government. For small woodlots, however the assistance was intended to cover 100% of costs. 56 delineated on the basis of site, timber type, and geo- graphic location. The relative efficiency of management intensification was then determined on the basis of rate of return and the timberland classes were ranked in terms of priority. Management costs were based on existing data and adjusted for changed conditions. Stumpage values were projected to the year 2000 on the basis of national forest timber sales, expected changes in timber quality, and demand. Results of the study show effects of various levels of expenditure or budget in terms of average available expec- ted impact on AAC. It is recommended that intensification be designated for management units (compartments, drain- ages) rather than for timberland classes. Classes can be used to establish priorities within the management units. A subsequent study by Marty (1973) evaluates alternative silviculture investments for softwood timber production. In this study, two "single practice" silvicultural alternatives (forestation, stand improvement) and a general management intensification alternative are analysed for a number of timber types and site classes. Forestation and stand improvement are seen as one-time treatments for specific stands or compartments. The general management alternative is looked at as an ongoing, or continuing treatment of a timber production unit. The criterion for evaluating effectiveness is internal rate Of return. Physical yields are estimated for various 57 treatments by stand type and site. However, there is considerable difference in assumptions for singe-practice versus general management intensification. For the forest- ation and stand improvement treatments it is assumed that there is no allowable cut effect, while an effect is assumed for general management intensification. Costs of treatments are the direct marginal costs. In the case of forestation, the average is estimated to be $124/ha and for sapling stand improvement it is $62/ha. General management intensification is estimated to cost $1/acre per year "... for all acres in the production unit." Benefits are determined from stumpage price projec- tions which are seen to be related to lumber prices. Rates of return are then calculated for each treatment, stand type, and site for various lumber price indices. In addition to the above general assessments, there have been a variety of evaluations of specific silvicul- tural treatments. In Canada, these include work by Nadeau (1970) on reforestation, Dobie and Wright (1978) on thin- ning and pruning, and Wiksten (1970) on several alterna- tives. In the United States, studies of specific alterna- tives include those by Beuter and Handy (1974) on reforest- ation, and Marty et al. (1966) on reforestation and stand improvement. The majority of these studies are financial analyses using internal rate of return or net present values as criteria. 58 Timber Supply and Demand Analyses Development of computer technology and increasing concern with a systems approach over the past 10 years has resulted in a variety of models designed to estimate timber supply and demand over time. It is difficult to categorize and evaluate this work with respect to the Nova Scotia problem because much of it is developed to answer other questions (policy analysis). However, identification of the impacts of timber management intensification is generally a principal concern. Adams and Haynes (1980) provide a comprehensive review of previous work designed to model and project forest products demand and supply. Briefly, they state that timber assessments are developed to meet the following objectives: "1) to describe the current state Of the resource. This is essentially an inventory problem ---. 2) to project the levels of future resource use. For timber assessments, this amounts to projecting future product and stumpage market activity ---. 3) to project changes in resource states resulting from (2) above. Again this is primarily an inventory or biometrics problem ---. 4) to evaluate the potential impacts Of alternative private investment and/or public policy programs on resource use and states ---." Although there have been considerable improvements in inventory and resource use projections, Adams and Haynes point out several weaknesses in past work. With respect to 59 private stumpage supply, there has been a failure to ... consider the price responsive nature of private timber supplies." In the short-run this leads to conservative estimates. In the long-run, it is generally assumed that management inputs on private lands will remain constant. In the determination of stumpage demand, spatial considerations Of forest products markets have been ignored, or the effects of manufacturing costs have not been adequately handled. Finally, it is concluded that, "Projection methods do no readily allow evaluation of impacts of alternative policies ---. Projection processes employed are often so cumbersome or insensitive that examination of alternative policies is extremely limited or even precluded." Previous modelling studies are categorized by Adams and Haynes as gap, nonspatial market, quasi-spatial market, and spatial market models, "--- based on whether or not they provide estimates of equilibrium prices and quantities ('gap' versus 'market' models) and the extent to which they recognize the spatial attributes of forest products markets." Two Canadian and several United States models are reviewed here. Special emphasis is given to the Nova Scotia study because it was designed to answer many of the questions posed earlier. 60 Nova Scotia Demand/Supply Model In 1972, the Province of Nova Scotia and the Federal Department of Regional Economic Expansion (DREE) entered into a forestry development agreement. The agreement involved six major programs including land acquisition, forest management intensification, supply/demand analysis, and education. While the proposal provided a basis for general forest management programs and funding there was no indication of where these programs should be carried out, at what levels, during what periods, and in what sequence. Nor was there a detailed analysis of expected costs and benefits. In an attempt to provide guidelines for timber manage- ment by the Provincial Government, a demand/supply study was begun in 1973.6 The study was completed in 1976. In the final report (Nova Scotia Research Foundation Corpora- tion, 1976), it is stated that the purpose Of the study was to ---estimate current and expected supply and demand functions for forest products in Nova Scotia. This information will be used to evaluate alternative forest management investment programs." Model Description The conceptual framework for the Nova Scotia model is illustrated in Figure 3. It is a 6This study was funded jointly by DREE and the Province of Nova Scotia. Supply analysis was conducted by the Nova Scotia Research Foundation and the demand analysis by Woods Gordon Consultants Ltd. of Toronto. 61 .hwsum aaamsm\vcmaov mHuoom m>oz mo xuo3wamum Hanummoaoo .m madman mmoaum owmQBSum ‘ huouao>aa ‘ - . i puma c3090 _ _ _ w _ manmsoaa< iii - I uwoo _ - Shouao>cfi umn3o wnwumm>pmm — mum>fiua owuma ovwmvmou 9 _ _ huouam>dfi uma3o , @003 Mo kaaasm A oum>fium Hamam as _ - _ _ coaumunoamcmum_ _ _ i . i n. - iifl _ Hawazeod :Ofiumooaa< i @003 mo a am @003 _ n - wcfiwsaoxo umou H _ wcfiuauommscma HHHE ® @003 Mom parade wo>fiuoo mumoo uuommcmuu uosvoua ram Haas s moans, uosvoua cam _ OOHHQ - _ panache cam 62 simulation model consisting of four submodels, which include derived demand, harvesting cost, transportation cost, and the master simulation program. Derived demand for wood is calculated by first estimating present and future prices for end products such as sulphate pulp, paper, lumber, plywood, particle board, and so on, over three different market areas - United Kingdom, United States, and Canada. These price projec- tions are then modified by transportation costs, tariffs, and discounts from Nova Scotia to arrive at a net mill price or product demand curve. End-product demand curves are assumed to be perfectly elastic because Nova Scotia's current and potential production is not likely to be a significant proportion of total demand (and only a small amount is marketed locally). Manufacturing costs (excluding wood) are determined for several sizes and types of mills (e.g., 500, 750, and 850 tons/day, sulphate pulp mill). This product demand curve and manufacturing cost curve form the basis for calculating derived demand for wood at the mill site (either existing mill locations or potential sites). In theory, derived demand in the short-run can be illustrated by the output curves in Figure 4. If we assume that output represents tons Of pulp per day or year, price represents net mill price for the product, and the firm's objective is the maximum profit it will produce at that level of output where marginal cost (MC) = marginal revenue 63 (MR). The average variable cost (AVC) and MC curves here exclude wood costs (wood = $0/cunit). Assuming no wood cost, the mill would theoretically operate at 600 tons/day. 3 of wood, and If we now assume 1 ton of pulp requires 1 m each cunit costs $20, this raises AVC and MC by $20. Optimum level of output is now lower (assuming that product price and costs of other inputs do not change). As wood cost goes up, the optimum level of output and consequently wood demanded decreases to the point where AVC (including wood) is just equal to average revenue (AR). At higher wood costs, AVC is greater than AR and the mill should close down. An important point here is that in the short' run, fixed costs are not relevant to Optimum level of output nor to derived demand for wood. P MCwood=$40 P $/ton wood=$20 $/cunit Cwood=$0 ‘Vcwood=$40 100 AR=MR=P DDwood AVCwood=$20 40 r ' Cwood=$0 : : ' I I U I I ' A Q l I Q 400 500 600 200 400 600 Output Cunits (a) (b) Figure 4. Theoretical determination of derived demand for wood. 64 In the model, derived demand for wood is determined somewhat differently. First, possible mill sizes are discrete. Because of design and equipment technology, a pulp mill might be 500, 750, or 850 tons/day capacity. A 500 ton/day mill may operate at 90 or 110% capacity (slowing down or speeding up equipment temporarily) but is not technically designed for those levels of output. Revenue and cost functions, therefore, apply to a specific size of mill. Secondly, average variable costs include chemicals, water, and power. These are treated as constant per unit of output (marginal cost constant). This is probably not an unreasonable assumption since they apply to a specific mill size (for example, 500 ton/day pulpmill). Average variable costs do not include labor. This is grouped with maintenance and administrative personnel under quasi-fixed costs. They are quasi~fixed in the sense that a mill cannot generally lay them off for short periods to reduce production. These costs can only be eliminated if the mill is shut down for more than 1 year. True fixed costs include property taxes, depreciation, interest, and so on which are incurred whether or not the mill operates. It is important to note that mill sizes, technology, and costs reflect current conditions. Costs are based on engineering data for current technology and are not necessarily representative of mills currently operating in the Province. 65 Determination of the derived demand for wood in the model can be shown by Figure 5. Graph (a) indicates end- product revenue and average variable costs. Part (b) shows that derived demand for wood is simply the difference between AR and AVC (excluding wood). If we assume that a mill would always Operate at capacity, derived demand for wood is simply a point. P P AR S/ E : cunit l c ! ANLVC V 1 l ' Pw DD | I I : l I | Q 0 Output 500 Q o Cunits 500 (a) (b) Figure 5. Determination of derived demand for wood in Nova Scotia model. In the long-run, a mill cannot stay in business and pay the maximum derived demand for wood indicated here (OPW = OPE - OCV) because fixed and quasi-fixed costs are not covered. In theory, derived demand for wood in the long-run would be the difference between average total or aggregate cost and price or average revenue. This is handled in the model by keeping account of net cash flow. If wood costs are too high, net cash flow is negative and if cumulative net cash flow is negative after 5 years, the 66 mill goes out of business. This is discussed further in the section on wood supply. The above situation for computing derived demand is complicated further by mills that produce more than one product (pulpmill which produces softwood and hardwood pulp, sawmill produces lumber and chips) and mills that use several types of wood. Possible combinations are illustrated below. In Figure 6, the two products are softwood and hardwood pulp. Revenue & Cost P $/ ARS cunit A Rn DDc DDS ANLVC DDH ANLVC1 Q Q Output Cunits (a) (b) Figure 6. Determination Of derived demand for a mill producing more than one end-product. Derived demand for softwood is equal to average revenue for softwood pulp (ARS) minus average non-log variable cost (ANLVC), and is similar for hardwood. If this pulpmill uses softwood chips from sawmill residue, 67 derived demand for chips equals ARs - (ANLVC + ANLVCI) where ANLVCl is simply the cost of chipping roundwood. Derived demand for chips is higher because there is no chipping cost. Plywood and sawmills may operate on a one or two shift basis. Average non-log variable costs may differ for the second shift (generally lower). Plywood mills may sell chips as well as boards. Derived demand for wood is calculated in Figure 7. In this case, average revenue equals ARply + ARc and derived demand equals ARply + ARC - ANLVC. P Revenue 1 5‘ AR Cost : ply I l I r DD ' ANLVC ‘, ' I I I I I I AR I I c I l I Output Q Cunits Q (a) (b) Figure 7. Determination of derived demand for peeler logs for plywood. Derived demand for sawlogs presents the problem of having a variety of products and the possibility of two shifts. This situation is shown in Figure 8. 68 Revenue & AR (3 ' D ost ' AR 3 VD g_, ANLVC . D I ?j ANLVCVD 44 ' AR I C I I ARS . Q Output Figure 8. Determination of derived demand for sawlogs. Sawmills have the Option Of producing dressed (D) or undressed (UD) lumber, shavings (S), and chips (C). Average non-log variable costs are shown for both dressed and undressed lumber. Derived demand is simply average revenue minus non-log variable costs. As indicated earlier, the derived demand submodel identifies end-product prices and from these calculates mill net prices in the Province. Manufacturing costs (excluding wood) are deducted to arrive at derived demand for wood, that is, the maximum price a mill could afford to pay for wood in the short-run. The main sequence of this submodel is specification of industry structure (types, sizes, and locations of mills); initializing industry cost data (transportation, labor); updating end product prices, mill net prices, costs; calling derived demand (what mill could afford to pay for 69 wood); and printing. End-product prices and simulation runs are currently possible for a 25-year period. Having determined derived demand for wood at various geographical locations and mills, it was necessary to identify how much wood could be made available at these locations at various prices. This involved identifying forest resource characteristics and ownership, harvesting, and transportation costs. Description of the forest resource basically involved identifying area and volume by ownership, cover type, and size class for each subdivision. The inventory data do not permit identification by stand type. Since the subdivision is a relatively large area and since transportation costs would vary considerably within the area, each subdivision was further delineated into a number of harvest zones. These zones were identified primarily on the basis of land capability and ranged in size from about 9000 to 30 000 ha. For the Province, there are 167 harvest zones or about 25 for each subdivision. The centroid of each zone is used to calculate transportation cost. To estimate inventory data and ownership by harvest zone, the subdivision totals were apportioned to each zone on the basis of hectareage. In other words, if 50% of the subdivision forested area is Crown owned, then 50% of each harvest zone is considered to be Crown owned. Supply on small private ownerships (less than 400 ha) is treated differently than that from Crown and large 70 privately owned land. In the case of small owners, a survey7 was conducted throughout the Province to deter- mine a reservation price8 (essentially a stumpage price) and cost of availability (harvest cost, transportation cost, and rate Of return). These two components make up reservation cost or small private supply curve at roadside. Wood supply at roadside from Crown and large private ownerships consists Of a stumpage price and harvest cost. These ownerships were separated from the small private group first of all because quantity of wood available is restricted by allowable cut (there is no restriction on small private land except total wood available) and because harvesting methods on Crown and large private land are assumed to be the cheapest possible (the harvesting submodel is described below). Supply curves at roadside for small private owners and Crown and large private owners form the composite supply curve at roadside (or for each harvest zone). Transporta- tion costs (described below) are then determined for each combination Of harvest zone and existing or potential mill locations to form millyard supply curves. The harvest cost submodel is designed to determine the cost of harvesting ($/m3) roundwood on Crown and large 7The survey was done by personal interview. Sample size was about 400 owners or about 1% of the estimated population size. 8Gregory, G.R. Forest Resource Economics - The Ronald Press Co., New York, 1972. 71 private land. Productivity (cubic metres per available machine hour) is defined for 10 different harvesting machines (manual felling with forwarder) for various cover types and size classes of timber. In the submodel, it is then assumed that the least cost harvesting system is used.9 Cost of access roads and overhead are added. These costs plus stumpage make up roadside price. The purpose of the transportation cost submodel is to determine the cost of transporting various roundwood products and chips from harvest zone centroids to mill locations. Basically, this consists of identifying primary road networks, minimum time distances between geographical centers, and transportation rate schedules (truck or rail). The purpose of the master simulation program is to generate harvest zone supply curves for each of three roundwood products. These include hardwood, softwood pulpwood, and softwood logs. Allocation of roundwood proceeds in an iterative way. A mill's requirement is first checked to see if only hardwoods, softwoods, or both are acceptable. A check is then made to determine wood diameter limitations (e.g., plywood mill can only use roundwood equal to or larger than 25 cm DBH). For each point on the harvest zone supply curve, the mill that can Offer the highest price is allocated the wood. Some mills, 9Constraints can be put on the type of machine or system that can be used within a subdivision or inventory type. 72 such as pulp and paper, are residue users and others, such as sawmills, are residue producers. Since residues are a source of wood, it must first be determined how much residue is produced before it can be decided what the supply is. Therefore, residue producers are allocated wood first. A residue supply function then enters in the over- all supply to residue users. After an initial allocation is made, a check is made to see if the allowable cut constraint is met. Another search is made of the mill's cash flow. If negative, that mill is dropped for the current year and a reallocation is made. A mill showing a negative cumulative cash flow after 5 years is considered nonviable. The master simulation program subroutines include provision for determining allowable cut and making land tenure changes; allocation of wood; grocut which computes yearly cut and growth data and updates; and various report writers. The grocut subroutine, as stated, keeps a record Of how much is cut by harvest zone and subdivision. In the case of Crown and large private land, the growth rate is assumed to be 2.5 cm DBH per 10 years. This, of course, updates volumes but also affects harvesting costs since costs are influenced by diameter. In the case of small private lands, growth in cubic metres per hectare per year is applied to volume by harvest zone and product class. 73 This description of the model explains the basic structure and approach. As stated earlier, the maximum length Of run, currently, is 25 years. Not much explana- tion has been made on updating of costs (manufacturing, harvesting, and transportation). These are modified primarily by economic drivers or indexes based on wholesale prices and unemployment rates. Evaluation of the Nova Scotia Model More than three years were required in the design and development of this model and while the exact cost is not known, it was in excess of $300,000. Since completion of the model or study in 1976, there has been no apparent use to date. There appears to be several reasons for this, including failure to address a specific problem or objective; inadequate user participation; inappropriate level of detail; and questionable validity of results.10 Failure to identify a specific problem or objective stems largely from poor project management. As indicated earlier, development of the model was contracted to two different consulting firms. A steering committee made up of Provincial and Federal government employees and an outside scientific advisor was responsible for writing terms of reference, guidance, and monitoring progress. The lack of a clear concensus as to the specific problem among 10Annino and Russell (1979) list these among the ten most frequent causes of simulation analysis failure. 74 steering committee members is evident from statements made in both the progress and final reports. The firm responsible for modelling the demand side states, ”From the inception of this study, the nature of the work assigned --- required the consideration of a number of alternative design features for the model. Particularly highlighting the non-uniqueness of possible model design have been the many types Of questions that have been put forth by the Monitoring Committee as being representative of those to which the model should address itself. Many of these questions were such that the set of given initial conditions and the event or phenomenon being explained were completely reversed from one question to the next. While from a general level of discourse and conceptualization this is acceptable, it poses particular problems from the point of view of conflicting requirements for detailed model specification" (Woods Gordon & Co., 1974). In the final report of the study group (Nova Scotia Research Foundation Corp., 1976) it is stated that "by studying the supply and demand in conjunction with one another, alternative industry structures relative to the resource base could be investigated, areas Of wood surplus identified, investment alternatives in the forest resource studied, etc. During development of the model, the above were suggested as the types of questions that might be asked. However, specific questions were not posed and, therefore, the model was kept flexible as possible." There is little doubt that this approach resulted in considerable delay in completion of the model and also sub- stantially higher costs. This might be acceptable if the increased flexibility provided a useful tool for answering the principal questions of concern. But, one of the major concerns was the impact of timber management intensifica- tion policies and programs. However, since the model was designed to run for only 25 years, no consideration was 75 given to silvicultural inputs - planting, cleaning, and thinning. It was simply assumed that these practices would have no impact within the planning period. Areas cutover were in effect "forgotten." The model treats the resource as a stock which changes as a result of growth within existing merchantable stands and harvest. Inadequate user participation was a problem from the start of the project. The Nova Scotia Department of Lands and Forests simply did not have any staff experienced with this type of problem and methods Of analysis. Although a Departmental employee was nominally the project manager, his lack of expertise meant that there was no effective project leadership. Day-to-day problems encountered by the consultants had to be resolved by them or discussed at a later date with the steering committee. Also, because of the lack of experienced personnel within the Department there was no continuing interest in further refinement and testing of the model. Annino and Russell (1979) say that inappropriate level of detail is often a problem in simulation analysis. This should be dictated by goals or objectives but also must be "--- consistent with the availability of data and other resources." There is no question that the spatial detail in the Nova Scotia model resulted in some major data and computational problems. For wood supply analysis the Province was delineated into 167 harvest zones based loosely on land capability. This was to provide better 76 sensitivity with respect to transportation costs. But, inventory data were not available for these zones; it was prorated from the subdivision (basic inventory unit) on the basis of area. While this level of spatial detail provides more sensitive transportation cost figures, it also increases the model complexity and computational costs. For these reasons (including data availability) it is questioned whether the additional sensitivity is justified. It would appear to be more reasonable to aggregate at the sub- division level (there are seven in the Province). Validity of the model outputs are questionable but there is insufficient historical data or experience to test. In the timber allocation routine, it can be assumed that wood is allocated on the basis of derived demand - mill that has the largest residual excluding wood costs gets the wood, transportation cost, or some other criter- ion. Because of the land tenure and lack of data on timber distribution, it is difficult to determine what would be a reasonable assumption. At present, the assumption is that timber is allocated to the mill (pulp and paper or sawmill) which is closest, that is, minimum transportation cost. This is not to say that total wood transportation costs for all mills in the Province are minimized. It is questioned whether the assumptions and model results with respect to timber supply from small woodlot owners are realistic. The small woodlot owner response is 77 based on a questionnaire and sample size of 400. This is about 1% of the population. The real question, however, is not sample size, per se, but the approach. Would owners respond to prices (harvest volumes) the way they say they would? Determination of derived demand for sawlogs is based on export markets for lumber. But about 70% of lumber produced in the Province is sold locally. The demand curve for lumber in the Province is probably not perfectly elastic as is assumed in the model. Therefore, the demand for sawlogs is probably not elastic but fairly inelastic. Derived demand for wood is determined, in part, by end product manufacturing costs. However, manufacturing costs for processing plants are based on engineering data for new mills. Many of the existing mills (including a pulp and paper mill) are old. It is not likely that those cost data would be very realistic and because of confidentiality it would be difficult to obtain more accurate data. The computer costs for a 10-year run of the model in 1980 were over $300 and apparently this would increase proportionately for a 25-year period. Relative to expected benefits this cost is high. Aggregation over space would reduce this considerably. Have there been any benefits from the model and what is the potential? To date there have been no apparent benefits arising from the model except possibly as a learning experience for those directly involved in its 78 development. The Province has developed an allowable annual cut model but this can only be indirectly attributed to the demand/supply model. As to potential benefits, it is difficult to conceive of any with the model in its present form. Some components such as the transportation or harvesting cost submodels could be useful for specific studies. For example, how would harvesting costs be affected by environmental regulations? The transportation component might be used to reallocate Crown land timber. In this latter case, however, a model is probably not necessary. End-product quantity and price projections for principal markets areas are useful but more up-to-date data are available from other sources. Forest Management Planning in Quebec In 1975, the Quebec Department of Lands and Forests began development of a comprehensive plan for timber allocation and forest management.11 Objectives of the plan are to obtain: " A) An adequate subdivision of the territory into sustained yield units ---; B) An 'optimum' resource allocation plan for a given set of management, harvesting, and utilization alternatives; C) An 'optimum' choice of the alternatives previously simulated; D) A frame of reference for the preparation of management plan at the unit level." 11Information on the Quebec plan is taken from Lussier (1976). 79 Principal steps in the planning process are shown in Figure 9. The initial step is to divide public forest land12 into 44 management units each with an average of about 2 million ha (4.5 million acres). This subdivision is based primarily on administrative criteria. Sustained- yield units are identified on the basis of socio-economic criteria (stabilize commuter operations and not increase wood costs significantly). The wood supply analysis consists of identifying allowable cut by sustained yield unit for --- various sets of management and utilization hypotheses.’ Demand analysis starts with an identification of processing centers where each pulp and paper mill is a separate center. Sawmills or other wood users are grouped. Conversion factors and normal operating levels are used to determine wood demand by product and destination. For each sustained yield unit and consuming center, a cost matrix is developed. Costs include forest management, logging, and transportation. Wood supply, demand, and the cost matrix provide the basis for a theoretical optimum allocation by a linear programming model. This theoretical optimum allocation plan is then adjusted by considering actual allocation and by manipulation with a simulation model (Figure 10). Finally each alternative set of allocation and management hypotheses are evaluated by a 12Public forest land makes up about 85% of total productive forest land. 80 swam Sumuiuuonm L coaumsam>m 1 h- Oman WOOHumOOHHm Shoulwaoq All uHas unmaowmcma he Oman Bumulwcoq .oonmdo a“ wcaacmaa unmamwmcma ummuom .o ousmfim amucoo hp >woaocnoou wamamu wcfiuauommsamz uosvoum Omnmflcfim umuaoo hp *1 ,L Ai mammamcm i parade woo: Manama mmocmumfiw mafia: umoo woos coaumuuommcmue All, unmaowmcma Ouch mono mo . - . coama>ficnnm mass sash» b. wmcflmumsm 4 %n mammamcm u _ a hammsm woos mm>HanumuHm .- wcfiumm>umn woo? cmnu was nonuo mom: uaoamwmcmz pom panama 81 socio-economic model that includes criteria such as dis- counted cash flow, employment, cost- benefit ratio, and surplus resource availability. A long-term management and allocation plan for the Province is prepared (Figure 11) and long-and short-term plans for management units are then developed. Wood supply by L. sustained yield ‘ unit 5 Allocation by Theoretical by destination programming allocation I Cost matrix }-— I ‘ ’ Acceptable I Actual I J Allocation by |.__. allocation - simulation for given set allocation [—I l of hypothesis Figure 10. Supply and demand analysis allocation plan in Quebec The Quebec planning process obviously provides consid- erably more capability than just the determination of opti- mal silvicultural inputs. As indicated in the study ob- jectives, the aim was to Obtain, "An 'optimum' resource allocation plan for a given set of management, harvesting and utilization alternatives ---", and to obtain, "an " 'optimum' choice of alternatives previously simulated ---. 82 Allocation given hypothesis set 1 Decision criteria Allocation given hypothesis set 2 Long-term soc io- i management economic and allocation model plan Allocation given hypothesis set n Ti..-- --_‘:r‘ T J Figure 11. Comparison and choice among allocation plansin Quebec What is the relevance of this work to the Nova Scotia problem? It appears to be conceptually sound and a ration- al approach for Quebec but substantial resources are required in planning and implementation. It has taken more than five years to develop the basic approach. Also, it is estimated that there were more than 70 professionals and technicians involved full-time in the early stages of plan preparation. Nova Scotia does not have this capability. In addition, land tenure within the two provinces is considerably different. In Quebec, the Province controls 85% of the productive forest land. It therefore has much more control over allocation of timber and management of the land than in Nova Scotia where Crown land makes up only 25% of the total. 83 Demand and Supply Analyses in the United States Because of the extensive nature of modelling work in the States and elsewhere on demand and supply forecasting no attempt is made here to classify or categorize it. However, a few studies that are felt to be relevant to the Nova Scotia situation are examined and some references to methodologies are presented. A recent study for the State of Oregon (Beuter et al., 1976) attempts to ---project reasonable possible changes in future timber harvests and their contribution to the econo- mies of major economic areas within Oregon - under varying assumptions about land-use changes, timber growth rates, harvest regulation policies, and utili- zation efficiencies." Projections were made for each of 10 timbersheds in the State. "The key assumptions chosen to vary between projec- tions were management intensities and harvest control." Seven management intensity options were considered. The highest level includes reforestation, genetic improvement, regeneration harvest, precommercial thinning, commercial thinning, and fertilization. Current management intensi- ties are assigned to administrative units. Two target distributions are identified for future management. One is based on likely changes in management over the next 30 years and the other on desired intensity given no time or financial constraints. Harvest level can be specified with or without a sustained yield constraint. Assumptions are made regarding land-use shifts, regeneration (stocking), Species conversion, utilization, growth, and so on. 84 Projections show a range of reasonably possible occur- rences of timber harvest based on varying assumptions. Estimates were then made of timber-dependent employment, private timber taxes, and public in-lieu payments associ- ated with the projections. Methods for identifying these impacts are discussed in the next section. In summary, several points can be made about this study. First, concern is with possible harvest levels and impacts. No consideration is given to demand. Secondly, no attempt is made to determine economic viability for different management intensities. It is stated, "The assumptions regarding all these practices were developed by interviewing forest managers about their intentions for managing their lands in the future. Presumably, these intentions represent a rational assessment of technical and economic possibilities. Thus, we are assuming that the management practices of the projections are economically viable and will, in fact, be carried out." As mentioned earlier, Adams and Haynes (1980) provide a comprehensive review of forest products demand and supply studies. These are categorized as gap, nonspatial market, quasi-spatial market, and spatial market models. Gap models are distinguished from market models in that the former do not provide --- estimates of equilibrium prices and quantities.' Examples of gap models are the United States Forest Service's timber trend studies. In these studies, historic data are used to establish end-product consumption and production relationships. Assumptions of price behavior in the future are made on the basis of past trends. Future consumption of 85 end-products is based on income, population, construction activity, technology, and so on. Stumpage requirements are then estimated from the end-use estimates. Supply projec- tions are determined from timber inventory and growth, allowable cut for public lands, and private harvest. Adams and Haynes (1980) say that, "In the several Forest Service studies previously cited, it has been explicitly assumed that the intensity of management practices (thinning, fertili- zation, and so forth) would remain the same in the projection period as during the base period." Although projections are made on a regional basis, "gaps" between demand and supply are identified at the national level. Nonspatial market models involve the endogenous deter- mination of prices and quantities but only within a single region. Transportation costs are not considered. Quasi- spatial market models are similar except that a limited number of supply or demand regions may be identified. Transporation costs are generally ignored. In spatial market models, prices and quantities are endogenously determined simultaneously for a set of demand and supply regions. Transportation costs are explicity considered. Adams and Haynes provide an extensive list of examples and references for these different types of models. Objectives, methods, and problems in forecasting for- est products supply and demand are discussed in a variety of studies including those by Marty (1969); Gregory et al., (1971); Manthy (1977); and Chappelle (1979, 1980). 86 Contribution from Timber Sector Activities The principal objective of timber management intensi- fication in Nova Scotia is employment. The problem is to determine how employment is affected over time by different types and levels of silvicultural inputs. Impacts can be broken down into those resulting directly from silvicul- tural operations and those occurring from effects of operations. Direct employment impacts are relatively easy to estimate. Those occurring from effects of operations involve identifying timber yield changes and employment resulting from harvesting, transportation, and processing. A number of studies have been done to identify and project forest industry employment changes over time. The most extensive work is that done by Wall (1969, 1973) and Wall and Oswald (1975) for the United States Pacific North- west. Wall (1969) projected wood consumption, forest based employment, and payrolls from 1965 to 2020 for the Colum- bia-North Pacific Region. Data for the period 1950 - 1965 were used to develop regression equations used to make the projections. In a subsequent study, Wall (1973), identi- fied employment implications of projected timber output in the Douglas-fir region from 1970 to 2000. It was found that, ”With continuation of present levels of forest management, the timber output in the Douglas-fir region will decline by the year 2000. Based on this projection and estimates of changing industry mix and labor productivity, employment in the timber based industries is projected to drop 45% between 1970 and the year 2000 ---.” 87 Wall and Oswald (1975) examine techniques for employ- ment projections and past trends for principal forest products sectors in the United States Pacific Coastal area. Regression equations are developed by sector (logging, sawmills, and planing mills) for major wood producing regions over varying base periods. Beuter et al. (1976) use these estimates for projecting employment effects from changes in timber harvest for Oregon. Greig (1979) estimates employment coefficients for logging, wood products manufacturing, and outdoor recreation in Australia. He states that while confidence intervals for estimates are --- occasionally wide---", the estimates '--- are not expected to be biased and should therefore prove useful in predicting the approximate direct employment consequences of land-use changes that affect the flow of wood from and recreation visits to forest land." A model for estimating socio-economic impacts from timber management decisions has been developed by Fowler (1978). The decision variable is harvest rate which is determined by a linear programming submodel with the objec- tive being to maximize volume harvested during the analysis period plus the residual volume. Impacts of changes in harvest rates (20-year analysis period) are determined by four submodels which include employment, input-output, forest measurement simulator, and tax base. Forest related employment is projected by regression equations. The input-output submodel identifies impacts on non-forestry related employment and gross regional product. 88 Timber inventory is tracked by the forest measurement simulator. Finally, the tax base submodel provides impacts of harvest on property values and assessed valua- tion. Since harvest is affected by various management policies, a variety of alternatives can be evaluated. In addition to the above studies, there have been others concerned with changes in employment trends and productivity. These include Sandoe and Wayman (1977) for the Canadian forest products industry; Granskog and Guttenberg (1973) for Southern United States forest industries; Kaiser (1971, 1974); and Irland (1972). CHAPTER 5 PROBLEM COMPONENTS AND GENERAL RESEARCH APPROACH This chapter includes identification of the major problem components - the decisions, variables, and rela- tionships required to solve the problem, availability of information with respect to these components, and general research approach to requirements. The components are identified within the following major categories: identi- fying silvicultural opportunities; biological effects of alternatives; effects on allowable annual cut; costs and direct employment of alternatives; roundwood consumption, and impact on employment. Identifying Silvicultural Opportunities It is not possible to identify a best way of meeting objectives without first delineating the alternatives. Silvicultural opportunities or alternatives depend on dif- ferent types of treatments, where and under what condition those treatments are applied, and area available for treat- ment. Criteria for differentiating alternatives should be based on the problem - in this study, those that affect costs and employment impacts. In addition, consideration 89 90 must be given to data availability and planning capability within the Province. Types of Treatment Five principal types of treatments have been identi- fied on the basis of the above criteria, what is currently practiced or contemplated in the Province, and what is thought to be feasible given past experience. These are planting nonsatisfactory restocked (NSR) lands; planting recent cutover, burned, or insect-damaged areas; cleaning; commercial thinning; and the combination of cleaning and commercial thinning. Site preparation, control of vegeta- tive competition following planting, and roadbuilding are not considered as alternatives but rather as operations which may be a necessary part of principal treatments. Precommercial thinning and fertilization are not included because data available show that the feasibility is questionable. It is also assumed that all planted trees are from genetically improved seed sources. Planting NSR lands and planting recent cutover, burned, or insect damaged areas could be combined but site and stand conditions and therefore costs of treatments will differ. The main reason is that planting NSR land requires that a salvage be done first. It is estimated that there are about 400 000 ha in the Province, which are over 60 years old and less than 40% stocked (average only 33-45 m3/ha). These stands cannot be economically harvested. 91 Recent cutover area does not generally require a salvage but may need remnant removal. The same applies to burned or insect damaged stands. The conditions that distinguish salvage from remnant removal are not clear. As stated earlier, questions of efficiency are not considered. It is assumed that the optimal number of trees to plant per hectare, rotation age, spacing, and thinning regime are known. Application of Treatments There are four major criteria or considerations for differentiating the application of treatments. These are spatial, institutional, biological, and technical. Timber inventory, allowable cut, and roundwood consumption vary by geographic area within the Province. In general, the largest proportion of softwood merchantable volume and volume in the larger size classes is located in the western half of the Province. Comparison of AAC with harvest and other removals shows that the major problem is in the eastern region. Transportation costs prohibit move- ment of any significant volumes from areas with a surplus to areas with deficits. In addition, biological conditions and land ownership vary by geographical area. Much of the harvested area in the east regenerates well, naturally, while much of that in the west does not. However, in the east, a large proportion of the regeneration is balsam fir, which is particularly susceptible to spruce budworm. The 92 present policy for a portion for this area is to plant species that are less susceptible to budworm. The most critical institutional factor is land owner- ship. Fifty-two percent of the productive forest land is in small private ownership (less than 400 ha); 24% is Provincial Crown; 21% is large freehold; and 3% is Federal Crown. A small woodlot owner survey (MacQuarrie, 1981) showed that there are about 30,000 small woodlot owners in the Province. Past experience indicates that both costs and merchantable timber yields of silvicultural treatments vary by landowner-group. In general, costs for treatments are higher for small properties because of size. Equipment and crews have to be moved more often. At present, it is estimated that only about 60% of the potential merchantable volume is available from this group. The percentage, however, varies by geographical region. Land tenure influences not only the total amount of merchantable volume available but also its distribution. Much of the large freehold land is owned by pulp and paper companies and a few large lumber manufacturers. As might be expected timber from these lands is largely reserved for company-processing facilities. There is a limited amount of roundwood sold from these lands but volumes and condi- tions of sale vary considerably from year-to-year. About three-quarters of Provincial Crown land is under long-term 93 lease to pulp and paper companies. This land is managed by the companies and is essentially reserved for their use. Costs and impacts of the treatments identified vary by site and stand condition such as age, stocking, and cover type or species. For example, 1981-82 rates for govern- mental assistance for cleaning range from $335/ha for areas with 4,000-12,000 stems to $420/ha on areas with over 50,000 stems. Additional compensation is provided for rocky conditions, steep slopes, and windfalls. Yield is obviously influenced by site quality. The final consideration in identifying application of treatment is technical limitations. Each specific geo- graphic area, stand condition, or site cannot be differ- entiated. Data are not available and would be costly to obtain. In addition, computational limitations must be considered. Current practice in the Province is to differentiate treatment types as above. Specific criteria for applica- tion of treatments are not clear. In the AAC model, seven geographic regions (subdivisions) and the three principal landowner classes are distinguished. In addition, five inventory type classification categories (cover type, site, age, density, stand nature) can be used to identify 2,340 individual inventory types of stands. Theoretically then, it is possible to have almost 50,000 alternative sites (assuming all inventory types were available on each land- owner class in each geographic area) to apply the five 94 treatment types. In practice, the seven geographic regions and three landowner classes are distinguished. Planting NSR land (salvage) is prescribed for inventory types over 60 years of age with less than 40% stocking. Commercial thinning is recommended for stands that are over 40 years old, greater than 60% stocked, and on productive sites only. Cleaning is proposed for dense young stands (10-15 years of age), which regenerate naturally. But with regard to density, there does not appear to be a specified number of trees per hectare which is optimal. Governmental assis- tance is available for cleaning in stands from 4,000 to more than 50,000 stems per hectare. Planting is required on about one-third of the annual softwood area cutover. This is a guide determined from regeneration surveys. In brief, criteria for applying treatment types are specific for some treatments but not for others. It is therefore not possible at this point to identify the number of distinct alternatives. What is required is a more definitive breakdown of treatment application based on biological factors. This does not necessarily imply that, say, cleaning should be prescribed only on better sites at 15 years of age, and in stands with 30,000 stems per hec- tare. Two or three alternatives might be included. It is difficult to say how many alternatives should be delineat- ed. This has to be determined in respect to the criteria stated earlier such as treatment cost, expected impact, data availability, and computational capability. 95 Area Available Once criteria have been determined for treatment alternatives, the area that meets these criteria must be identified. Inventory data show the potential area for planting NSR lands by geographical area and biological factors. The problem is to estimate area by ownership class (particularly the area in small ownerships) that would, in fact, be available. This is a problem with all treatments. The small woodlot owner survey cited earlier (MacQuarrie, 1981) showed that 42% of the owners were "producers", that is, they intended to perform logging, or sell products in the near future, or produce Christmas trees. This group owns 61% of the small woodlot area (based on sample survey) and have an average of 125 ha. Currently, a varying percentage estimate is made on the basis of management plans for these owners (20% of the total area in small private ownership that meets the plant NSR criteria is assumed available from 1980 to 1985; 30% from 1985 to 1990). There are two problems in determining area of recent cutover, fire or insect damaged land that does not regen- erate adequately and which should therefore be planted. One is to identify how much harvest or damaged area there is, its geographic location, owner, site, and so on. The other problem is to determine how much of that area will require treatment. A rough approximation of area harvested can be made by simply dividing total volume harvested by 96 average volumes per hectare. For example in 1980, total softwood harvest was about 3.7 million m3- Average volume per hectare is assumed to be 120 m3. Therefore, about 30 000 ha are harvested annually. Some indication of geographic location and ownership can be determined since roundwood production figures are collected by county and landowner class. Cover type and site of cutover areas are not known. Area losses from fire are generally estimated and recorded by Provincial fire personnel. Information on in- sect-killed stands is more difficult to obtain. As noted earlier, spruce budworm is the most destructive insect. Stand susceptibility is due, in large part, to age and species composition. With the exception of the large contiguous area of overmature balsam fir stands in the eastern Highlands of Cape Breton Island, stand mortality is spotty. Broad areas with significant tree mortality can be identified through aerial defoliation surveys, but it is difficult to pinpoint specific stands. Area requiring treatment, that is, planting, is largely a function of cover type, site, and harvesting method. Mixedwood stands of shade-tolerant softwood stands that are clearcut will often not regenerate adequately to desirable species. Present practice on Crown lands is to survey, each year, areas that require planting. This is based on cutovers, which after 5 years have not adequately 97 restocked. On company lands, many areas are planted within 1 year of cutover. In summary, more information is required to identify area available for treatment. It is difficult to estimate the response of small woodlot owners to various programs. Biological data are necessary to show the relationship between regeneration and harvest method, stand cover type, and site. Another problem is data handling. Information on harvest area by method, location, and site is available on Crown and large freehold properties but it is time consuming and costly to input, store, and retrieve. An information management system is being developed by the Province but it is difficult to say whether this system will meet the needs outlined. The problem of obtaining reliable data or harvest from small woodlots is being alleviated, in part, through Provincial assistance pro- grams. Information is available from management plans for these properties. Biological Effects of Treatments Biological response to the alternatives identified above is a function of the initial stand condition, site quality, treatment, and competing factors such as fire, insects, and disease. Not only is information on gross net volume required, but it must also be determined how this 98 volume is distributed by species, size, and quality. These relationships are shown in Figure 12. Some data are available on biological effects for each of the five major treatments identified. The reliability of these data and extent or availability for specific treatment alternatives are not known. For example, it is stated, "plantations can yield 240 m3 per hectare (3O cords/ acres) in 40 years assuming an average seedling sur- vival of 70% on typical sites. If sawlogs are desired and one commercial thinning is carried out at age 30 (yielding 120 m3/ha), plantations can be expected to yield 16 bem/ha at 60 years. Natural softwood stands are not usually harvestable until 60 to 80 years after clearcutting and yield an average 112m3/ha" (N.S. Dept. Lands and Forests, 1981a). In addition to increased growth rates, yields are expected to be larger due to species change and reduced losses to insects and disease. Cleaned stands are expected to yield almost the same merchantable volume (215 m3/ha on typical sites) as plantations. Some information is available on product quality by cover type, site, and so on from a 3-P sampling survey. Fire, insect, and disease losses are difficult to estimate, particularly for the latter two. Spruce budworm populations and damage are a function of weather, stand conditions, and protection effort. In current AAC calcula- tions variable percentage losses by geographic area are estimated. These are based on past experience and the proportion of gross merchantable volume in susceptible species, as well as estimates of harvest or salvage. 99 Stand condition and site quality Silvicultural treatment Growth rates Time of harvest Gross volume yield Fire, insect and disease loss 1 net yield by species, size and quality Figure 12. Factors and relationships affecting biological yield of silvicultural treatments. 100 In summary, some data on biological response to various treatments are available but it is difficult to determine extent and reliability. A considerable amount of research on growth and yield is underway by the Province and the Canadian Forestry Service but it is too early to obtain definitive results from some of this work. Effects on Allowable Annual Cut Allowable annual cut is a function of inventory com- position and volume, growth and yield, harvest practices, land tenure, and governmental policy with respect to inven- tory control. The biological impacts discussed in the previous section are not directly translated into allowable cut impacts. Increased growth in immature stands may have no effect on allowable cut unless the inventory composition is such that an increased volume of mature timber can be removed while still maintaining a sustained capacity. Nova Scotia currently determines AAC by subdivision using the simulation model described earlier. In this model there are 21 different major land categories (seven subdivisions and three landowner groups). Each land cate- gory is further differentiated by a number of inventory descriptors (cover type, age, density, and so on). In total there is potential for distinguishing 2,340 inventory types. For each inventory type, area, volume per hectare, and basal area are identified from inventory data. 101 Inventory is updated at 10-year intervals by two growth functions - age and basal area. Stands that were 50 years old are advanced to 60 years. Basal area for that stand is then advanced by adding basal area growth per hectare for the period. The stand is checked to determine if basal area changes the density class. Harvested stands move back to a predesignated age and basal area class. In the updating procedure, land areas rather than volumes per hectare are changed. Allowable annual cut is identified for different harvesting and management scenarios. Harvesting level and pattern are specified; level by total area or volume, and pattern by age, site, density, cover, and so on. Total area available for harvest is reduced for inaccessibility, greenbelts, and stocking level. Management scenarios consist of specified silviculture treatments by geographic area, ownership, and inventory type. Volume reductions are made to reflect availability from small private ownership, cull and waste in logging, and abnormal insect and disease losses. Finally AAC is determined by finding the harvest level/pattern for each management scenario that yields a relatively even flow of timber (within 10-year periods) over the selected conversion or rotation period. Silvicultural alternatives affect total volumes by cover (softwood, hardwood), size, site, and density class. In addition, alternatives affect area available for harvest 102 by changing stocking level, losses from insects and dis- ease, and cull and waste. Since there is a high proportion of mature and overmature timber in the Province, any expected increases in yield affects current allowable cut. In other words, there is an allowable annual cut effect. The only apparent improvement on AAC determination1 is through better growth and yield information, insect and disease damage, and availability of timber from the small woodlot owners resulting from silvicultural inputs. Growth and yield and insect damage were discussed in the last section. With regard to timber availability from small private owners, MacQuarrie (1981) has developed a "...time depen- dent, deterministic population model." Timber availability is a function of attitude and tenure. Availability is in effect the timber volume on lands controlled by an owner willing to undertake logging, reforestation, spacing, or various other activities. Probabilities for change in attitude and land tenure (length of time owner held the property) were estimated. Initial results show that the percentage of area available (and presumable timber avail- able) approached 60%. An earlier attempt to determine small private supply curves for softwood and hardwood by geographic area was made as part of the demand/supply simulation model (Nova 1The Province is in the process of making changes in the current AAC model. 103 Scotia Research Formulation Corporation, 1976). In this survey, owners were asked the proportion of merchantable volume (stumpage) they would be willing to sell at various unit prices. The results indicate that the supply curve for hardwoods was above that for softwoods in all subdivi- sions. Also, hardwood supply was very inelastic (i.e., almost vertical) for prices up to $20 per m3 and levelled off sharply for higher prices.2 The softwood supply curve even for sawlogs was below that for hardwoods and much more elastic. It is questionable how meaningful the results of this survey are, and indeed how reliable the approach is. The sample size was only about 400, or slightly over 1% of the population. With regard to approach, it is questionable whether owners would actually react the way they said they would. Costs and Direct Employment of Alternatives Silvicultural Costs Silvicultural costs are a function of the variables or criteria used to differentiate alternatives. These are treatment type, geographical location of treatment, land- owner class, site, and stand condition. Costs by type of treatment depend on the operations required. For example, szenty dollars per cubic metre is probably about five to six times the average stumpage price at the time. 104 planting NSR lands may include roadbuilding for access; existing cover has to be removed; the area must be site prepared and planted; and finally there might have to be a follow up treatment to control competing vegetation in the plantation. Planting recent cutover, burned, or insect damaged stands probably will include remnant removal or cutover clearing, site preparation, and a herbicide appli- cation. Roadbuilding may not be required for cutovers since there was access during the harvesting operation. Geographic location of treatment is not likely to have much effect on costs with the possible exception of planting. Planting stock has to be shipped from one of the three nursery locations. Costs of treatments vary by landowner class. Most silvicultural work presently being done is funded to some extent through a Federal/Provincial forestry development agreement. This agreement is cost-shared with the Federal Govern- ment providing about 80% of funds. Two projects basically include all silviculture activities. These are private land management and Crown land management. Under the private land project, landowners are separated into three categories: those with less than 2000 ha; over 2000 ha; and pulp and paper companies. Various rates of assistance are provided for each group with the highest rates avail- able to those with smaller area. Each year a rate schedule is determined. In effect then, the Provincial cost is affected by landowner group. 105 Although area treated does not necessarily reflect effectiveness of silvicultural expenditures, there is little question that more work is done for less cost on large properties. Over the first four years of the current five-year agreement almost two-thirds of the total area treated was on pulp and paper company land for less than one-third of the total program expenditures (average cost per hectare treated was $59). For the same period, the treated area for small ownerships was less than one-third of the total at a cost of almost two-thirds of the program total (average cost per hectare treated was over $275) (MacLaren Plansearch, 1981). Silvicultural costs vary by stand condition. Under the development agreement, maximum allowable rates for each treatment are determined each year. These are expected average costs based on past experience and general price level increases. For example, maximum planting assistance is up to $90/ha for container seedlings and $120 for bareroot seedlings. Site preparation rates range from $135 per ha for chemicals to $275/ha for double pass with a rome disc. Cleaning rates vary from $245/ha (4,000-7,000 pine and/or larch stems per ha) to $450/ha (mixedwood or hardwood cover with more than 12,000 stems/ha). It is not known, how actual rates compare with the allowable set. In the case of some Crown land silvicultural treatments however, there are indications that the actual is larger than the maximum allowable (MacLaren Plansearch, 1981). 106 There are considerable data available on average costs of principal treatments but little on specific costs by geographic area, landowner group, and site. At present, it is difficult to determine the extent to which the Province uses cost differentials in setting treatment priorities. The practice appears to be to identify treatments, which will meet the biological Objective - that is, AAC require- ments, and then identify the cost of that package. Alter- natives within the package or scenario are not really examined to determine if there is a more cost-effective approach. Cost data are required for the alternatives identified earlier. Average costs by treatment type are not suffi- cient if there is to be effective choice among all the alternatives. Some of these data are available now. For example, costs of cleaning by stand condition or number of stems per hectare are known. Some data are available on costs of alternatives by landowner class. Additional information is required, for example, on access, site preparation, planting, and tending costs by site and stand condition. Direct Employment From Alternatives Employment resulting from silvicultural alternatives is largely a function of the operations required and the nature of Operation. Planting NSR land requires that the area be salvaged first. Cleaning stands with 50 000 stems 107 per hectare will require more labor input than cleaning those with 4,000 stems. There is little documented information on silvicul- tural employment. However, as in the case of costs, there is a general idea of requirements by principal treatment. The need is for more detailed information for the alterna- tives identified. This is relatively straightforward to obtain by simply observing or sampling a number of opera- tions. Roundwood Consumption The major impact of silviculture is increased quantity and quality of timber output. But it may also reduce costs of a given level of output. The extent to which these changes influence employment depends on biological effects, changes in the AAC, and finally roundwood consumption. Biological effects and changes in AAC have been discussed. Roundwood consumption is a function not only of supply but also demand. To some extent, AAC can be and often is used as a proxy for roundwood supply. In the AAC calculation, assumptions are made to eliminate stands that are uneconom- ical to harvest because of stocking levels or accessibil- ity. Also, constraints are included to account for parti- cipation of small owners or sustained yield on Crown lands. In effect, what is being assumed with regard to economic 108 logging chances and participation rates is that the price for additional quantities is above the equilibrium. Buyers are not willing to pay the price required by sellers. However, using this approach (AAC as a proxy), the assump- tions are generally short-run, that is, based on current conditions. They generally do not allow for changes in economic demand or supply. It is important therefore to think of consumption as a function of demand and suppy rather than as simply requirements and AAC. But it is not helpful to talk about supply and demand without considering their specific meaning. Roundwood supply is the quantity of a specific product someone or group is willing and able to provide at a given price at a particular place and time. Demand is the price someone or group is willing and able to pay for a specific product at a particular place and time. Products may be sawlogs, pulpwood, or boltwood either softwood or hardwood by species. Location can be roadside, collection yard, or mill yard. Group may mean small woodlot owners, large freehold owner, or Crown. The time may be now or at some future date. Consideration has been given to location, products, and landowner group but little has been said about the time frame of concern. 109 Planning Horizon Various silvicultural inputs will have different effects not only in terms of volume, species, and quality of timber output but also in terms of when in time those outputs occur. A commercial thinning results in immediate output from trees removed but it also affects the remaining crop and output at the time of harvest. Cleaning a stand at 10 to 15 years of age technically does not result in increased physical yield until about 20 years later when the stand is harvested. Planting might not affect physical yield of timber for 40 years or more. The selection of cost effective silvicultural alternatives depends not only on their initial unit costs and employment effects but also on when those effects are likely to occur. If the planning agency, in this case government, is concerned with only the next 10-year period, the alternatives are considerably more limited than if the interest were over 50 or 60 years. Planning is basically looking ahead - determining where to go (objectives), alternative ways of getting there, and evaluating consequences of alternatives. This provides the basis for management decision-making. In the problem discussion, there was no explicit mention of plan- ning horizon, that is, time period over which the objec- tives and alternatives would be considered. Implicitly, the planning horizon is specified by the types of alterna- tives contemplated. Although planting might have immediate effects on output and consumption (allowable cut effect) it 110 will also dictate the species and volumes that will be harvested 30 or 40 years in the future. The Province currently uses a planning horizon of 50 to 70 years depending on which assumption about the level of management inputs is used. The shorter period is used with the assumption of a high level of inputs and the longer period with no inputs. This basically reflects rotation lengths for softwood sawlog-quality trees. Because of the difficulty of accurately predicting future events that will affect not only timber output but also consumption levels, it is questionable what the appropriate planning period should be. The long-term (50- 70 years) cannot be ignored particularly with regard to the resource, but how much effort should be put into forecast- ing consumption and resulting employment that far into the future? One approach is to separate the planning horizon into several discrete periods based on expectations as to the type and magnitude of changes that will affect consump- tion. Special emphasis is then given to the current period, say the next 10 years with less effort on subse- quent periods. The principal variables and interrelation- ships in roundwood supply and demand are examined below. Roundwood Supply Roundwood supply at the mill is a function of the production functions and factor costs in growing, harves- ting, and tranSporting timber. In the short-term, supply 111 is relatively inelastic because the production function is limited by several fixed factors. There may be a sustained yield constraint limiting the physical volumes available from Crown lands. For private lands, yield per hectare is fixed and although logging utilization may be increased or harvesting may be extended into more inaccessible areas, marginal costs increase quite sharply. Silvicultural inputs can affect the shape and location of the supply curve. For example, commercial thinning will make additional timber available now, which might result in a shifting of the supply curve outward and to the right (Figure 13). Other operations such as planting or clean- ing, might result in a shifting of the curve depending on whether there is an allowable cut effect. If as a result of inputs, AAC can be increased immediately, the roundwood supply curve could be shifted to the right. In the medium-term (20-25 years), roundwood supply will likely be more elastic due to a reduction in the fixed factors. Additional land can be brought into production. The effects of cleaning or final harvest of commercially thinned areas will be reflected in a change in the AAC. In addition, technological advances in harvesting and trans- portation might reduce unit costs through productivity gains. There are virtually no fixed factors in the long-term. Productivity gains from planting will be reflected in the AAC. Accessibility of stands can be increased. Harvesting 112 and transportation costs can be reduced through technolog- ical improvements. The long-run supply curve therefore will be more elastic. Figure 13 shows what these supply curves might look like. Silvicultural inputs might shift each of these down- ward and to the right depending on the nature of productiv- ity increases and cost function. A shift will result in more roundwood being made available at a given price or the same quantity at a lower price. Effects on consumption, however, cannot be determined without considering the demand. P Q Figure 13. Hypothetical short-, medium-, and long-term roundwood supply curves. Roundwood Demand Roundwood demand is the quantity someone or group is willing and able to buy at a given price at a specific place and time. Because buyers differ in their ability or willingness to pay for the product (roundwood) there is a range of prices Offered for different quantities - this is the demand schedule or curve. Demand for roundwood is generally thought of as deriv- ed demand, that is, dependent on the demand for products 113 such as lumber, veneer, pulp, and so on. Demand for these latter products is derived from demand for houses, furni- ture, and paper products which is turn depends on prices, population, incomes, and technology. Basically all pulp and paper production in Nova Scotia is exported. Principal markets are the United States, United Kingdom, and the European Economic Community. Nova Scotia producers therefore are competing in these markets with other producers and therefore relative costs of production and transportation influence demand. About 70% of the lumber produced in the Province is sold locally. Lumber demand is therefore influenced strongly by construction activity in the Province. There is competition from outside and this can have a significant impact on demand for locally produced lumber. In addition to end-product demand, the demand for roundwood is influenced by factors such as quality Of inputs (sawlogs), processing technology, and industry capacity. For example, if quality of sawlogs increased while costs and other factors remained the same, the demand curve for logs would shift outward to the right. The same effect might be obtained by the introduction of new saw- milling technology, which increases the value of output. In the short-term, roundwood demand as with supply is relatively inelastic. This is due mainly to fixed proces- sing capability. There is some flexibility in that paper machines can be speeded up. Extra shifts can be added in 114 sawmills or within a year or less new sawmills can be built. But it may require 3 to 5 years to add a new line or build a new pulp mill. Hence, sawlog demand may be more elastic in the short-run than pulpwood demand. In the medium-term, roundwood demand is influenced by processing capability (new mills can be constructed), end- product markets, technological change, and the relative competitive position of Nova Scotia producers. Within this period new pulping technology can reduce or change fiber requirements. Demand for a specific size, species, or quantity of fiber can change substantially. As in the case of long-term supply there are essentially no fixed factors in demand. New mills can be constructed and there is the possibility of all types and sizes of roundwood being used. Figure 14 shows roundwood demand curves for alterna- tive time periods. Increases in end-product demand for housing, paper products, and so on might result in an upward shift of the curves. Q Figure 14. Hypothetical short-, medium-, and long-term demand curves for roundwood. 115 Roundwood Consumption and Price Consumption and price of roundwood are a function of both supply and demand. The basic purpose of silviculture is to shift the roundwood supply curve downward and to the right. The extent of changes in consumption and price depends on the shapes and positions of both supply of and demand for roundwood. Figure 15 shows some alternative effects of changing supply and demand. In (a) demand is perfectly inelastic; a downward shift in supply causes no change in consumption but a reduction in price. In (b) demand is perfectly elas- tic. A shift in supply from S to 8' causes no change in price but a significant change in consumption. Graph (c) shows that when the demand curve is downward sloping, a shift in roundwood supply has both an effect on price and consumption. // // / (a) (b) (C) Figure 15. Effects of changes in roundwood supply and demand on consumption and price. 116 The impact of a change in roundwood supply and there- fore on consumption and employment depends on the nature of demand. If demand is very inelastic as it probably is in the short-run, consumption effects are likely to be mini- mal. In the medium- to long-term, however, new processing capability can be constructed (assuming the competitive position of producers in end-product markets is favorable) and consumption will be increased. Approaches to Forecasting Roundwood Consumption A prerequisite to any type of forecasting is identi- fication of the product or product groups and geographic area of concern. Roundwood products can be softwood, hardwood, sawlogs, pulpwood, boltwood, fuelwood, and so on. The selection of products or groups should be dictated by the purpose of the forecasting exercise, that is, to determine probable future levels of consumption and employ- ment resulting from silvicultural inputs. Selection must also take into consideration data and analytical limita- tions. A reasonable approach in Nova Scotia might be to differentiate roundwood into softwood sawlogs, softwood, pulpwood, hardwood pulpwood, softwood other products, and hardwood other products. Roundwood is separated out by processing (sawlogs, pulpwood) because there are major differences in employment per unit of wood input by process. Major cover type (softwood, hardwood) is 117 separated because silvicultural inputs will have a major effect. Hardwood sawlogs are not delineated because rela- tively little effort is going into silviculture to produce them and it is not likely that any significant quantity of quality hardwood material will be available in the future. Softwood roundwood for other products would include mater- ial for wood-based panel or other reconstituted wood products. The other prerequisite to forecasting is the choice of geographic area for which consumption is to be estimated. This could be at the Provincial level or some smaller-sized unit such as subdivision or some unit based on concentra- tion of processing facilities. The criteria for choice should reflect the nature of the problem -- objectives, alternatives, and constraints. Consumption and price of roundwood depend on supply and demand, which are affected by transportation costs, location of facilities, and end- product markets. Timber inventory and markets are not evenly distributed geographically. Forecasting consumption at the Provincial level, therefore, will not provide an indication of sub-provincial activity. Richardson (1969) suggests three different approaches to defining regions. These are each based on different types of homogeneity. They may be homogeneous with respect to a variable such as timber type or site with respect to activity to administration - a planning region. A reason- able approach with regard to consumption would appear to be 118 to identify areas based on concentration of manufacturing or processing facilities. For example, delineation of a region for forecasting pulpwood production might be exis- ting pulp and paper mill locations and timbersheds; for sawlog consumption, it might be an area enclosing a "clump" of sawmills. Methods of forecasting can be categorized into four major groups (Gregory et al., 1971). These are naive methods, barometric techniques, Opinion polling, and econometrics. Naive Methods These include "...forecasting based on mere guess or simple extrapolation of historical data." (Gregory et al., 1971). The simplest approach here is to assume that future roundwood consumption will remain static, that is, remain at the current level. This is basically the approach now used in Nova Scotia. A slight variation of this is to assume that consumption will equal timber or roundwood output (AAC). The attitude here is that it is not possible to project future demand with any degree of reliability, but, if "quality" timber is produced it will likely be consumed. This is a popular attitude among many foresters particularly those involved with a strong interest in biological aspects of production. There are some varia- tions within this approach which range from a purely quantitative concern (the future is fiber, therefore 119 maximize quantity) to qualitative concern (there will always be a demand for large-sized timber, veneer and saw- logs therefore, maximize quality). A third variant Of the naive methods is to extrapolate the historic trend. Gregory et al., 1971, however, state that some trend projections may not be "naive". They "... are as objective as many other methods and are often suc- cessfully used for forecasting." The major problem with this approach is that if there are significant changes in important variables (silviculture inputs), which affect supply or demand, the historic trend may be misleading. Also, while extrapolation might yield reasonably accurate projections for short periods (year-to-year), it could be very inaccurate for long periods, mainly because of changes in key variables. Barometric Technique This approach is based on the assumption that future activities or events can be forecast on the basis of what is happening now with selected indicators. The indicators serve as a barometer and may include housing starts, busi- ness activity, and so on. They may also include "pressure indexes" such as "...the difference between the rate of family formation and growth rate in inventory of houses and apartments...” which might be used in estimating "...long term demand for new housing" (Gregory et al., 1971). 120 Opinion Polling This method involves a canvassing or survey of attitudes among manufacturers or marketing people about future intentions (capital investment) or opinions about future trends. Econometrics Econometric techniques are based on the assumption "... that changes in economic activity can be explained by a set of mathematical relationships between economic variables." (Gregory et al., 1971). A very simple model might look as follows: 1 1 Y = + + t a bxt “t where: Y: = lumber consumption in year t in region 1 a = constant b = coefficient x: = housing starts in year t in region 1 u = term accounting for unexplained t variation In this example, a, b, and u might be determined from his- torical data. To forecast Y, assumptions are made about future changes in housing starts (based on past trends or forecasts) and Y is then determined. Estimation of sawlog consumption is then based on forecasted lumber consumption. 121 This is a very simple approach. Intuitively, we know that lumber consumption is influenced by other types of activity (other uses for lumber) and other variables (prices). Also past relationships between lumber and hous- ing starts (expressed by coefficient b) may not be realis- tic because Of changes in lumber requirements per house. There is also the problem of forecasting housing starts. In most econometric studies multi-equation models are used, which determine demand, supply, consumption, and price simultaneously. An econometric approach therefore, involves defining the variables to be explained by the analysis (endogenous) and those that are external (exogenous); specifying the form of the equation, how the variables are related - which are dependent and which independent; and estimating the parameters or coefficients showing the nature of the rela- tionship between dependent and independent variables. There are a number of problems associated with this ap- proach (Chapelle, 1980; Gregory et al., 1971). Generally, a considerable amount of data is required. Computation can be time-consuming and costly. As stated earlier, the approach now used to estimate roundwood consumption in Nova Scotia is to assume that consumption will equal current requirements. This approach is intuitively reasonable and easy especially for the short-term. At present the annual softwood requirement without management inputs is higher than the AAC by about 1 122 3 and million 1113 (annual requirement is 3.3 million m AAC is 2.4 million m3)- There is a gap between SUPply and demand. The current level of management inputs is just expected to bring the AAC of softwoods up to current requirements. In effect then, consumption is constrained by the AAC. The problem with this approach, however, is that calculation of the AAC is based in part on assumptions of economic conditions (some stands are uneconomic to harvest because of stocking levels, tree size, accessibility), which are likely to change in the medium-term as harvesting technology or transportation facilities change. In effect, AAC and supply are shifted to the right. Even in the short-term, end-product demand is changing and causing a change in AAC. For example lumber demand (ignoring short- term cyclical changes) has increased substantially in the past 5 to 10 years. Lumber and sawlog prices have gone up dramaticlly. This has resulted in sawmill modernization (faster cutting, thinner kerf saws, more automated equip- ment) which has permitted the utilization of smaller-sized logs which in turn has increased the AAC of sawlogs. A reasonable approach to forecasting consumption at this point in time for purposes of evaluating silvicultural alternatives would appear to be to take a step beyond an estimate based on current requirements. The current requirements approach might be satisfactory for the short- term, that is, next 5 years, but for the medium-term 123 consideration should be given to end-product market chang- es, industry structure, and technology (for example increased use of hardwoods in pulp and paper), and factors affecting roundwood supply other than silvicultural inputs. This could be based, in part, on historic trends in con- sumption by product (volume in pulpwood versus sawlogs) but also on published data with respect to market for end- products. This might be combined with a polling of indus- try opinion. For example, end-product consumption projec- tions for the United States, the United Kingdom, and Europe could be used in conjunction with an assumption about Nova Scotia's share of that market. An econometric approach for the Province would require considerably more data and com- putational effort than presently available. Impact on Employment Changes in timber volume or quality resulting from silviculture can affect employment in harvesting, transpor- tation, and manufacturing. The extent of this impact depends on the volume of roundwood production and consump- tion and the nature of harvesting, transportation, and manufacturing operations. Because most logging firms also transport roundwood, it is difficult to separate these two activities. Therefore, for purposes of this study, they are considered together. 124 Harvesting Employment Despite a relatively large increase in roundwood production in Nova Scotia over the past 15 years, total employment in harvesting has declined. In 1965, roundwood 3 and the production was a little over 2.9 million m number of production and related workers was 1,221. In 1979, roundwood production increased to almost 4.2 million m3 while employment declined to 985. This represents an average annual increase in output per worker of over 5%. Production per man-hour was 0.86 m3 (Figure 16). Most of this increase in productivity has occurred since 1975. This appears to result, in large part, from severe labor shortage in the early 1970's and the subsequent introduc- tion of more mechanized equipment such as forwarders, feller-bunchers, and tree-length harvesters. The problem in forecasting harvesting employment resulting from silviculture inputs is to determine which systems are in use now (different landowner, location, product groups) and how these systems (and associated labor requirements) are likely to change over time. Systems currently used by landowner group are not too difficult to identify. Most of the operations on small woodlots are very labor-intensive, usually involving manual felling and skidding to roadside with farm tractors or conventional skidders. Harvesting methods on large freehold and pulp and paper company controlled lands usually involve some combination of manual felling and use of conventional Production/man-hour (m3) Man-hours paid (millions) 3) Roundwood production (millions m 125 0' I l"'/ ./ .5- 0_ /-/ \.\-/-/I~--" \- :I: n In 1 4 n n j n 1 n I 1 n r n .5 . ‘ ‘ Year Figure 16. Roundwood production, man-hours paid, and production per man-hour in Nova Scotia, 1965-1979. 126 skidders to mechanical felling and tree length processing. Productivity functions for various types of equipment have been developed (N.S. Research Foundation Corporation, 1976b; Aird et al., 1971; McGraw and Silversides, 1970). Estimating change in systems or equipment used and employment effects is more difficult. This will depend on labor availability, wage rates, capital and operating costs of equipment, and so on. It is unlikely however, that harvesting methods on small freehold property will change significantly due to the high capital and operating costs of alternatives. Increased mechanization on large freehold properties particularly pulp and paper company lands in the past five years has been dramatic. It is doubtful whether this practice will continue at a similar rate because of the high capital and operating costs of equipment. For many small, fragmented parcels of land highly mechanized equipment cannot be economically justified. Analysis of harvesting employment trends in the south- ern United States (Granskog and Guttenberg, 1973; Kaiser, 1971) show that labor productivity in harvesting "...rose at an average annual rate of 2.7% between 1954 and 1967 (Kaiser, 1971). Gains were rapid from 1958 to 1963 - 4.3% annually - but only 0.1% annually from 1963 to 1967" (Granskog and Guttenberg, 1973). Although increased mech- anization in logging has decreased the number of cutters, there has been an increase in equipment Operators, and maintenance and service personnel (Campbell and Power, 1966). 127 There are several approaches to forecasting employment effects of silviculture inputs. The simplest method is to determine average man-years per unit of output, based on current practice at the Provincial level for all land- owners, and apply this ratio to changes in output. How- ever, this approach would not permit differentiation of impacts by landowner group nor would it take into account technological change. Another approach would be to develop a regression equation based on historical data (Greig, 1979). The problem with this method is that it is implic- itly assumed that future developments (introduction of mechanical equipment) will mimic the past. Also this approach does not permit separation of impacts by product group or landowner. A third method is to separate land- owners (small and large), determine productivity based on most common harvesting system (for sawlogs and pulpwood), and assume some pattern of productivity change over time. Manufacturing Employment For purposes of this study, forest products manufac- turing is separated into three sectors or groups. These are sawmills and planing mills, pulp and paper mills, and other wood-using mills. These are examined below. Sawmills and Planing Mills Over the past 15 years, employment in sawmills has declined from about 1,500 employees to 1,000. During this same period, the number of 128 sawmills and lumber production has also declined.3 In 1965, there were 409 sawmills in the Province and lumber production was 577 000 m3. By 1979, the number of mills had dropped to 347 and lumber production was 443 000 m3- Figure 17 shows changes in lumber production, man-hours paid, and labor productivity over this 15-year period. Despite the fact that increase in the number of mills in the Province and lumber production appears to have levelled off since 1971, employment (both in terms of production and related workers and man-hours paid) has continued to decline. This is reflected by the increased labor productivity. The reason for the declining trend in both number of mills and lumber production appears to be due to increasing scarcity of quality sawlogs. Increasing productivity and declining employment results mainly from a change in structure of the industry. The number of smaller mills have been increasing while intermediate sized ones have been decreasing. Also, the larger mills are producing a higher proportion of total output. In 1965, mills with a 3 of lumber produced 35% of the capacity of over 7500 m total. In 1979, this group produced almost 60% of the total. The larger mills are more automated e.g., automatic carriages, faster cutting headsaws, resaws, double-end trimmers, and so on. 3This is production and related workers. Production/man-hour (m3) Man-hours paid (millions) 3) Lumber production (000 m 129 .25 /'\ I / .20 /l\. I I /. \_/ .15 '\ I l\ I—I/ .—." .10 T l 1 J I I l 4 J 1 4 j I I 1 l 4.0 A 3.5 Kl, \\A--A”‘\\\ A 3.0 2.5 A—-‘ “\ a-‘ A \‘N‘ ‘/A 2.0 \‘/ ::_- L 1 l 1 l 1 j J I 1 l L 1 4 l j 600 .\ O O . soo /\ . O O o. / O 0”, ’// . O \\\ "-.a" 1.00 / .1 O l l I 1 l _1 l 1 l I 1 A 1 1 l 1 1965 1970 1975 1979 Year Figure 17. Lumber production, man—hours paid, and production per man-hour in Nova Scotia, 1965-1979. 130 Future employment in sawmilling depends on availabil- ity of quality logs, consumption, and technology. Approaches to forecasting employment in sawmills and planing mills are much the same as those discussed for harvesting. Greig (1979) develops a regression equation based on time series data where the dependent variable is average employment per mill, and the independent variable is average sawn output per mill. Pulp and Paper Mills In 1965 there were four pulp and paper mills in the Province and two more were added in 1967.4 In 1972, the roundwood mill at Sheet Harbour was closed. At the present time there are five mills in operation. Over the 15-year period (1965-1979), employment (production and related workers) in pulp and paper manufac- ture has increased from 1,327 in 1965 to 1,865 in 1979. Man-hours paid have gone from 2.9 million to 4.2 million. It is not possible to determine the trend in labor productivity for this sector because output figures are not published. Output capacity is available but it is ques- tionable how meaningful these figures are. The alternative is to compare employment with wood input. In 1965, wood 4Because of confidentiality restrictions, Statistics Canada includes the Masonite Canada Ltd. hardboard mill with the pulp and paper sector. Therefore it is included here as a pulp and paper mill. 131 5 input used in pulp and paper production was about 1.7 million m3. By 1979, inputs increased to 3.1 million m3. Labor and wood inputs in pulp and paper production over this period are shown in Figure 18. As shown in the lower two graphs labor input has increased as wood inputs have increased. However, as shown in the top figure, labor input per unit of wood input is declining. Assuming future roundwood consumption levels were known, future employment in pulp and paper manufacturing could be estimated by extrapolating the trend shown in the top graph of Figure 18. These relationships, however, need to be analysed more closely. During the period under review two new mills have been built and one has been closed. There has been modernization and expansion of others. Other Forest Products Manufacturing Other types of manufacturing in the Province include firms that produce boxwood, cooperage stock, lath, spoolwood, and such products. Total employment for this group was about 500 in 1965 and increased to 760 by 1979. Wood input is difficult to estimate, but because many of these operations are very labor intensive, the labor input per unit of wood input 5This is softwood and hardwood roundwood equivalent classified as pulpwood by the Provincial Dept. of Lands and Forests (computer printout, unpublished). This includes sawmill and whole tree chips. Man-hours paid/unit wood input (mH/m3) Man-hours paid (millions) Wood input (000 m3) 132 ‘,,"“-A - a” ‘-‘ A/ /"‘-A/ . A A ‘/ \‘/ ‘1’ " A—A/ ..L— b ./. . /’.k\. I’O’/' I .\ /.\o/ \. .. / I—o ON/O Year Figure 18. Wood inputs in pulp and paper mills, man-hours paid, and man-hours paid/unit of wood input for Nova Scotia 1965—1979. 133 would probably be higher than that for sawmills or pulp and paper mills. Comparison of employment per unit of wood input in sawmills and pulp and paper mills in 1979 shows that man- hours paid in sawmills is about twice that for pulp and paper mills (2.4 man-hours per m3 in sawmills and 1.3 in pulp and paper mills). The trend, however, indicates that this gap may be closing (Figure 19). In addition to studies cited above there are others which are useful in analysing future employment trends in forest-products manufacturing. Wall (1969) projects the development of wood consumption, forest-based employment, and incomes for the northwestern United States from 1965 to 2020, based on regressions for the period 1950 to 1965. Expected changes include a declining wood consumption in the sawmill sector and increasing expansion of veneer, plywood, and pulp and paper. Industry employment is expected to decline further as productivity increases. In a later study of the Douglas fir region, Wall (1973) projects employment in timber-based industries to drop 45% between 1970 and the year 2000. This is based mainly on expected changes in industry mix and labor productivity. Kaiser (1974) examined labor productivity trends of forest industries in the United States from 1954-1972. During this period labor productivity in paper, paperboard, and pulp mills rose at an average annual rate of 4.5% while average increases in productivity for sawmills was 3%. 4.0 3.0 2.0 1.0 134 F I\ / .\ Sawmills / H“ I- . \I /l\ \ I I I’ ' \ / \- I b Pulp and paper mills /I.... I ..~ [’0 / .\. \ ’4. . .\ /. \I 0—0 N. J I l l l l l l 1 I l n L n 1 1965 1970 1975 1979 Year Figure 19. Man-hours paid per m3 of wood input in sawmills compared to pulp and paper mills in Nova Scotia, 1965-1979. 135 Kaiser (1974) states that "gains in labor productivity of the lumber industry did not follow the traditional pattern." In most industries, productivity increases result from increased output; employment remains about the same or changes only slightly. For the sawmill industry, however, output has stayed about the same while labor input has decreased. This is the trend in Nova Scotia. CHAPTER 6 RESEARCH STUDIES AND PRIORITIES Although a considerable amount of research is underway in the Province, the foregoing analysis indicates several areas where additional studies are required. Selection of these studies and priority are based on specific needs identified earlier, current availability of information, and work in progress. Although the study numbers indicate an order of priority, they are interrelated and the process should be iterative. For example, the identification and selection of silvicultural alternatives that are evaluated are partially dependent on expected biological results, effects on AAC, and cost and employment effects. These impacts cannot be determined without first specifying the treatment and application. In effect then, a tentative set of opportunities must be identified, but this might change as more information is made available on impacts. Study 1: Silvicultural Opportunities Problem: What are the silvicultural Opportunities in the Province? 136 137 Five principal treatments have been identified. These are planting nonsatisfactory restocked lands; planting recent cutover, burned, or insect damaged areas; cleaning; commercial thinning, and the combination of cleaning and commercial thinning. These treatments can be evaluated on the basis of average costs and effects on allowable annual cut but costs and effects differ by geographic area, land- owner class, site, and stand condition. Therefore, there is a need to further delineate treatment alternatives. The problem is - how should they be differentiated? And, how do we determine the area available or suitable for treat- ment? Objective: Identify silvicultural opportunities that will permit evaluation and choice given the Province's objec- tives, criteria, and constraints. Approach: Differentiation of treatment alternatives must be based on expected differences in impact, that is, on allowable annual cut and costs, on opportunities for those treatments, and on constraints. The recommended approach is to identify a number of silvicultural prescriptions based on the five treatments identified earlier, costs of alternatives, and expected responses. For example, one prescription is to plant non- satisfactory restocked areas with stands over 60 years of age and but which are less than 40% stocked. Another is to 138 plant areas with stands over 60 years old but less than 60% stocked. Cleaning alternatives might be differentiated on the basis of stems per hectare and site. The objective of this study is to differentiate no more than 50 or 60 alternatives. These might be based on the five treatments, three geographic regions, two land- owner classes, and two site/stand condition classes. The second step is to estimate area available for treatment given the above prescriptions. Inventory data can be used for estimating area for planting NSR land, cleaning, thinning, and the combination of cleaning and thinning. Recent cutover, burned, and insect damaged area can be estimated from harvest and regeneration surveys and fire and insect damage reports. Requirements: Prescriptions have to be based on the prac- tical experience and limited data available to the planning and operations staff of the Department of Lands and For- ests. The Department's staff have already identified what is thought to be the best alternative based on biological response. What is required is further differentiation based not only on biological response but also on estimated costs. More definitive response and cost data will be obtained in studies identified below. Additional effort is required in determining area harvested annually (burned or insect damaged) and not like- ly to regenerate adequately. This is required by 139 geographic area, landowner, and site. Some data are avail- able now (regeneration surveys) on the types of sites and proportion of total area not regenerating but this is limited mainly to Crown and large freehold property. The problem is determining area harvested and proportion of that area by site type. Little information is available on small freehold ownerships. A survey is required, which will show area harvested by geographic area, landowner, site, and probability of adequate natural regneration. Manpower and funding requirements for identifying prescriptions is expected to be minimal - no more than 2 to 3 man-months. A survey to determine harvest area (fire and insect-damaged) by landowner, geographic region, and site is estimated to require 10 to 12 man-months. This could be done by Crown and private land foresters in the respective regions. Study 2: Biological Effects of Treatments Problem: Growth and yield response for each alternative identified in Study 1 is needed to determine effects on AAC. Response is primarily a function of stand condition, site, treatment, growth rates, time of harvest, and losses to fire, insects, and disease. Data on gross merchantable volume yield are available for principal treatments (plant- ing, cleaning) by cover type (softwood, mixedwood, and hardwood stands), site, and stand condition class. 140 Reduction of gross merchantable yield due to noncatastroph- ic events (small fires, insects, and disease excluding major outbreaks of spruce budworm) are compensated for in the yield data. Losses due to large-scale fires and insect attack are not accounted for. These are handled separately in the AAC calculation. Objective: For each alternative identified in Study 1, identify growth and yield by major species groups and size class. Approach: Yield tables have been developed for the Prov- ince for softwood plantations. These show mean annual increment, basal area, and merchantable volume for three diameter classes and 13 land capability classes. Response to planting and cleaning alternatives by site can be obtained from these tables. However, there is no distinc- tion by species group; all softwoods are combined. Growth and yield for thinned stands can be Obtained from variable density yield tables that have just been deveIOped. Requirements: The need for this study is to identify response for treatments differentiated in Study 1. Some attempt should be made to distinguish differences by spec- ies group (softwoods) and increase in yield due to a cul- tural treatment such as thinning (reduction due to noncat- astrophic events). The requirement is basically assigning 141 response as shown in the yield tables to specific treat- ments 0 Study 3: Cost and Direct Employment of Silvicultural Alternatives Problem: Costs of silvicultural alternatives vary by prin- cipal treatment and the application of treatment. Planting nonsatisfactory restocked land may require roadbuilding to enable access, areas must be cleared, site prepared, plant- ed, and possibly followed up with a herbicide treatment to control competition. Costs for this treatment may vary by site due to preparation required and follow-up. Cleaning costs will vary by site and stand condition class. Areas with 5000 stems/ha are more costly to clean than those with 50 000 stems/ha, other factors being equal. These cost data are required to evaluate and compare alternatives. In addition, direct employment generated by alternatives will differ. The question is, how much? Objective: Determine costs and direct employment for silvicultural alternatives identified in Study 1. Approach: Average cost data and employment by principal treatment (planting, cleaning) are available but specific data by site, stand condition class, and landowner are not. Reliable data can only be obtained by identifying treatment 142 alternatives, cost components (for planting, access, site preparation, cost of seedlings, planting, and follow-up), cost by component, and labor input. This can probably best be approached by selecting three or four cases for each treatment alternative (if there are 50 alternatives, select 150 - 200 locations), develop a form for components, and determine the cost and employment effects for each compo- nent 0 Requirements: A survey form is required for each principal treatment (planting NSR land, cleaning) identifying geo- graphic location Of treatment, landowner class, size of area treated, site, and stand condition class. In addi- tion, cost component should be listed (access, site prepa- ration, and so on) as well as labor input. PrOSpective treatment applications need to be identi- fied and followed up by on-site monitoring, or completed by the work supervisor immediately following each job. The manpower requirement for developing forms is minimal. Collecting data could be accomplished during the spring, summer, and fall seasons by job supervisors. This could be done with only limited additional manpower effort. Collat- ing data is expected to require an additional one man- month. 143 Study 4: Effects on Allowable Annual Cut Problem: AAC is currently determined by specifying a management scenario, harvesting pattern, and conversion period. A management scenario consists of a schedule of treatments (clearing, planting, cleaning and thinning) designated by inventory type (also landowner and geograph- ic area) by simulation period (10 year periods now used). The level of programs can be specified by acreage or proportion of area eligible for treatment. The harvesting pattern is predetermined in the model. Harvest area or volume is specified in each simulation period based on a priority matrix. For example, in the initial 10-year simulation period, the harvest area might be set at 1000 ha for subdivision 1 (assumed that landowner group can be specified). The harvest rule might be to cut stands with the following characteristics: first - oldest age class, site 1, 61-80% density, softwood cover. If there is not sufficient area (or volume) in this category, the second priority stands are cut. By changing criteria for setting priorities, product groups (diameter size class) and cover types (softwood, hardwood) can be favored in the harvesting pattern. Output is the area and volume harvested (size, softwood, or hardwood), by lO-year period by inventory type description (age, site, cover, density). Allowable cut is then determined by selecting the harvesting level and 144 pattern, which provides an approximate even flow (over each 10-year period) of volume. The first problem is, what is a realistic approximation of the harvest pattern? This determines the volume and area available by location, site, and so on. Secondly, area and volume constraints must be determined to account for logging cull and waste, participation by small freehold owners, and catastrOphic losses due to fire, insects, and disease. Thirdly, it would not be efficient use of computer time to run each of, say, 60 alternatives separately. No single treamtent is likely to result in a realistic AAC target being met. Packages of treatments have to be specified. This is a trial and error approach. How are treatment combinations identified that meet AAC goals? Objective: Identify silvicultural combinations (from Study l) which will meet the AAC target for the Province by region, species group, and product class. Approach: The harvest pattern to be simulated can be approximated by results obtained in Study 1. Data on area and volume constraints for logging cull and waste, partici- pation by small freehold owners, and losses from fire, insects, and disease are available by geographic location. They are expressed as percentage reduction factors. Selection of alternative treatment packages that meet the 145 AAC target is difficult because it is not possible a priori to determine what contribution a single alternative will have on AAC. Combinations have to be specified and run through the model. One set has already been identified (this is the basis for the current silvicultural program). Other sets (say 3 or 4) have to be identified and run on the model. Since the objective is to meet an AAC target by region, at least cost, those alternatives which are thought to have the highest biological response, at least cost, should be grouped. As indicated earlier this is a trial and error approach. Requirements: This study requires the services of staff who have experience with the AAC model and sensitivity of AAC to various treatments. It is estimated that 6 man- months will be required in identifying silvicultural combinations and interpreting results. Computer time and costs are not possible to estimate with confidence because of the trial and error nature of the approach. Study 5: Roundwood Consumption Problem: In the previous studies, treatment alternatives, costs, and AAC are determined. There is a further need to relate AAC to roundwood consumption and employment. Given a target AAC specifying volumes by species group, size 146 class, and geographic area, what is the projected consumption level? Projection of roundwood demand, supply, and consumption will assist in setting AAC targets but it will also provide an indication of the level and type of manufacturing activity resulting from various AAC targets. Is it reasonable to assume that an AAC of 3.3 million m3 of softwood will be consumed? If so, will that consumption by industrial sector (pulp and paper, sawmilling) be in proportion to the relative amounts of product groups making up the AAC (pulpwood, sawlogs)? Objective: Identify probable future roundwood consumption levels associated with AAC by geographic region and major product groups. Approach: Delineate major consuming centers for principal products (sawlogs, pulpwood, other) based on current distribution of mills. Each of the five pulp and paper mills (including the hardboard mill) might be a separate center. Sawlog consumption centers would be based on "clumps" of sawmills presently in operation. Identify short-run or current "demand" for roundwood based on end-product markets, mill capacities, species and size constraints, and approximate timbershed (area within which the mill(s) could economically transport wood given current costs). Given AAC levels by geographic area, landowner, cover, and product, assume an allocation pattern based on 147 current distribution and demand at various centers. Repeat this process for the intermediate term (25 years) and long term (50 years) taking into account national (for Canada and the United States) projections of end-product demand, prices, and technological change. Requirements: This study is expected to require about 6 man-months. Study 6: Impact of Silvicultural Alternatives on Employment Problem: The ultimate concern of silvicultural investment is maintaining or expanding employment. Because of changes in labor per unit of wood imput in harvesting, transportation, and processing, it cannot be assumed that maintenance of AAC at the level of current requirements will result in maintenance of employment. Although employment in silvicultural Operations will increase as inputs increase, there is likely to be a reduction in labor input per unit of wood input in processing. The question is, what will these relationships be in the future? What is the expected employment in harvesting, transportation, and processing for principal product groups and consumption levels? 148 Objective: Determine expected employment resulting from different consumption levels by product group. Approach: Employment in logging and manufacturing is a function of the level of operation (roundwood production and consumption by product group) and methods. The level of operations is derived from Study 5. Methods of operation (logging, transporting, and manufacturing) vary by landowner group, raw material inputs, costs, and a host of other factors. There are reasonably good historic data on employment trends by activity (logging, manufacturing by principal product). A realistic approach is to extrapolate these past trends using regression analysis. Requirements: One to two man-months. Discussion of Research Studies and Priorities The study requirements identified above are relatively straight forward with the exception of Study 4 - effects of silvicultural alternatives on AAC. Identifying silvicultural input packages that meet AAC targets is a trial and error approach, which may be cumbersome. A particular package may not be optimal given all possible combinations. But it will be optimal given those compared. Another approach is to develop an optimizing model - 149 linear, dynamic, goal programming. However, given the Province's constraints, (planning staff, data, and so on), this type of model does not appear to be realistic. Another approach, which should be considered, is to rank silvicultural alternatives on the basis of internal rate of return and then use these results to identify packages. In effect then, the ranking provides a guideline for combining alternatives. The only additional requirement of this approach would be to identify per unit values for output (by product group and time), and using treatment cost data obtained in Study 3, calculate rate of return. Consideration should be given to uncertainty regarding treatment impacts. Biological effects are influenced by growth rates, losses to insects, diseases, and fire, and participation by woodlot owners. This can probably best be accounted for by assuming different response scenarios and running through the AAC model to determine sensitivity of results. CHAPTER 7 SUMMARY AND CONCLUSIONS The Province of Nova Scotia is currently faced with a softwood timber shortage. Past timber management practices and expansion of manufacturing facilities have resulted in a resource base that cannot continue to sustain current and projected requirements. There are a number of alterna- tives. Some are not politically acceptable. Others are limited because of environmental, institutional, or finan- cial reasons. The problem is one of choice from among a variety of alternatives given various constraints. There have been studies addressing this general issue but they have failed to provide useful guidelines in large part because the problem was not specifically identified. The provincial and federal governments are currently spending $7 - 8 million/year on forest management. This appears to be the level of expenditure required to maintain AAC based on the present planning approach. However, is it possible to obtain a higher level of AAC with the same expenditure, or to achieve the same AAC at lower cost? A forest management package has been selected which will meet the AAC target, but is it the cheapest package? 150 151 The purpose of this study is to identify the problem, that is, the decision-maker, Objectives, alternatives, and context; to determine information needs to solve the problem; and finally to indicate how that information might be obtained. The decision-maker of concern is the Provincial government, which owns 24% of the 4.1 million ha of forest land. The remainder is divided among the Federal Crown (3%), small freehold owners - those with holdings less than 400 ha (52%), and large freehold owners (21%). The Province's objective for forest management is to "increase the supply of wood fiber to maintain or increase aggregate employment directly dependent on the forest resource." In this study, it is assumed that the objective is to maintain current employment in the most cost-effective way. There are several alternatives for increasing wood supply and employment. Increased effort can be put into general administration and support, for example, research and extension services. Additional funds can be put into forest management-planting faster growing species, cleaning dense young stands to improve growth rates, and thinning older stands. Harvesting methods can be improved to permit logging of stands that are now uneconomic because of the small size of trees or low volumes per acre. Processing plants can be updated or modified to permit use of species, particularly hardwoods, which are presently underutilized. 152 Emphasis of this study is on forest management alternatives, that is, planting, cleaning, and thinning. The Province is currently spending the majority of its funds in this area and these alternatives appear to offer the most potential, given the objectives and constraints. Five types of forest management activities or alternatives are identified. These include planting nonsatisfactory restocked lands; planting recent cutover, burned, or insect-damaged areas; cleaning; commercial thinning; and the combination of cleaning and commericial thinning. The Opportunities for and impact of these alternatives on the Province's objective are affected by biological, economic, technological, and institutional factors. Area available for treatment depends in part on harvesting Operations and objectives of private landowners but there are about 30,000 small woodlot owners in the Province. Impact of alternatives depends on the location of wood- using industry now and in the future. This is affected by market conditions. Costs of alternatives vary by where they are undertaken, on whose land, and by method. A review of past and current work shows that two studies have recently been done in Nova Scotia to answer some of the questions. One looks at silvicultural alternatives as a group to determine costs and benefits. No attempt is made to distinguish among alternatives by type nor is there any attempt to identify the geographic area where they would be most cost-effective. 153 In the other study, a timber demand/supply simulation model was deveIOped for the Province. This was intended to address these questions. The model required three years to deveIOp. Since its completion in 1976, there has been no apparent use to date. The major reasons appear to be, failure to address a specific problem, inadequate user participation, inappropriate level of detail, and question- able validity of results. Timber management planning in Quebec is based on comprehensive analyses of timber demand and supply by sustained yield unit. An attempt is made to arrive at an optimal resource allocation plan given various management, harvesting, and utilization alternatives. This planning process and methods appear to be suitable for Quebec but Nova Scotia does not have the planning capability (staff and funding) at present to undertake such an exercise. Also, land tenure in Nova Scotia is different. A number of studies done in the United States are also reviewed. Some of the methods are useful in developing an approach for Nova Scotia. In the present study, major problem components, avail- ability of information with respect to components, and research approach to obtaining information are identified. Problem components are the decisions, variables, and inter- relationships required to solve the problem. Six compo- nents are discussed. These include identifying silvicul- tural opportunities; biological effects of alternatives; 154 effects on allowable annual cut; costs and direct employ- ment of alternatives; roundwood consumption; and impact on employment. It is not possible to identify the best way of meeting objectives without first identifying the opportunities. Types of silvicultural treatments must be determined as well as area available for treatment. Types of treatments (planting nonsatisfactory restocked areas, cleaning) were specified earlier. Area available depends on inventory status, harvesting, and insect, disease, and fire damage. Some of this information is available but it is not of sufficient detail. Biological effects of treatments must be determined. There appear to be reasonably good data on volume yields but more information is required on how this volume is distributed by species, quality, and size. These biolog- ical impacts must then be translated into effects on allow- able annual cut. Increased growth rates and yields on small woodlots will not affect AAC if owners choose not to harvest. Also AAC is affected by prOportion of timber in the mature age classes. The Province has a model for determining AAC. What is required is better data on growth response from alternative treatments including insect and disease losses, and availability of timber from small wood- lot owners. To determine the most cost effective alternatives, treatment costs have to be identified. Average costs are 155 available by treatment but more specific data are required on treatment cost by landowner, site and stand condition. In addition, more detailed information on employment for each of the alternatives must be obtained. The major impact of silvicultural inputs is increased quantity and quality of timber output. The extent to which changes in timber output affects employment depends on supply/demand relationships and consumption. In this study, the determinants of roundwood supply, demand, and consumption are discussed and methods of forecasting are examined. Manufacturing employment associated with various levels of consumption must finally be identified. Histori- cal data show that employment per unit of wood input has been decreasing in both sawmilling and pulp and paper manu- facture. The relationship between employment and roundwood consumption in the future can probably be projected reason- ably well from past trends. The final section of this study identifies specific research requirements, including the general approach, and their priority. These studies are 1) identify silvicul- tural alternatives; 2) determine biological effects of treatments; 3) determine effects on allowable annual cut; 4) identify costs and direct employment for each alterna- tive in 1); 5) estimate roundwood consumption; and 6) determine employment effects of alternatives. LIST OF REFERENCES REFERENCES Adams, D.M. and R. Haynes. 1980. The 1980 softwood timber assessment market model: structure, projections, and policy simulations. For. Sci. Monogr. 22. 64 p. Aird, K.L., and J. Ottens. 1979. The outlook for timber utilization in Canada to the year 2000. Can. Dep. Environ., Can. For. Serv., Unpubl. Rep. 305 p. Annino, J.S., and E.C. Russell. 1979. The ten most frequent causes of simulation analysis failure - and how to avoid them. Simulation Today, 60: 137-140. Beuter, J., and J. Handy. 1974. Economic guidelines to reforestation for different ownerships. Oreg. State Univ. For. Res. Lab., Res. Pap. 23. 69 p. Beuter, J., K.N. Johnson, and H.L. Scheurman. 1976. Timber for Oregon's Tomorrow. An analysis of reasonably possible occurrences. Oreg. State Univ. For. Res. Bull. 19. 61 p. Campbell, D.R., and E.B. Power. 1966. Manpower implications of prospective technological changes in the eastern Canadian pulpwood logging industry. Can. Dep. Man. Immigr., Res. Monogr. 1. 154 p. Canada Department Regional Expansion. 1981. The Atlantic Region silviculture program, an evaluation. Final Report. 142 p. Canada Department Regional Economic Expansion and Nova Scotia Department of Lands and Forests. 1977. Canada-Nova Scotia Subsidiary Agreement, Forestry. 47 p. Chappelle, D. 1979. RPA models, methodology and mythology. Paper presented at the annual meeting of the Midwest Forest Economists, Merrimac, Wis. Sept. 6. 10 p. Chappelle, D. 1980. Modeling problems in forecasting timber demand. Paper presented at the Conference, Timber Demand: the future is now. Sept. 9, 1980, For. Prod. Res. Soc., New Orleans, LA. 20 p. 156 157 Dobie, J., and D.M. Wright. 1978. Economics of thinning and pruning - a case study. For. Chron. 54: 34-38. Forest Products Research Society, 1979. Timber supply: issues and options. San Francisco, CA., Oct. 2-4, 1979. 218 p. Forest Products Research Society, 1980. Timber demand: the future is now. Proceedings presented in New Orleans, LAO, septo 9-11, 1980. 206 p0 Fowler, K.S. 1978. Toward a more integrated regional timber mOdelI For. 8C1. 24: 434-443. Gansner, D.A., J.E. Barnard, and S.F. Gringrich. 1973. Iden- tifying regional opportunities for accelerated timber management. USDA For. Serv., Res. Pap. NE-251. 6 p. Gregory, G.R., D. Hair, H.R. Josephson, V. Holopainen, K. Mantel, S.L. Pringle, P. Riihinen, R. Salther, and H. Vaux. 1971. Forecasting in forestry and timber economy. Folia For. 101. 49 p. Greig, p. 1974. Employment coefficients for forestry planning and management. Aust. For. 42: 181-189. Irland, L.C. 1972. Labor trends in southern forest industries, 1950 to 1969. USDA For. Serv., Res. Pap. SO-81. 21 p. Kaiser, H.F. 1971. Productivity gains in forest products industries. For. Prod. J. 21: 14-16. Kaiser, H.F. 1974. Forest products industries show advances in labour productivity. For. Ind. 101: 32-34. Lussier, L.J. 1976. The forest management plan operation in Quebec. Pulp Pap. Can. 77: 42-51. MacLaren Plansearch. 1981. Subsidiary agreement for forestry: interim evaluation. Final report. Prepared for the Province of Nova Scotia and Can. Dep. Reg. Econ. Exp. 175 p. MacQuarrie, P. 1974. Laws relating to forest resources in Nova Scotia. N.S. Dep. Lands For., Ext. Note 90. 28 p0 MacQuarrie, P. 1981. Nova Scotia private woodland owner survey. N.S. Dep. Lands For. 51 p. Marty, R. 1969. Some special problems of forecasting timber supplies. J. For. 67: 88-91. 158 Marty, R. 1973. Economic effectiveness of silvicultural investments for softwood timber production. Ip_Report of the President's Advisory panel on Timber and the Environment, Washington. D.C. pp. 141-147. Marty, R. 1973. Softwood sawtimber supply and demand projec- tions. In Report of President's Advisory Panel on Tim- ber and The Environment. Washington D.C. pp. 129-140. Marty, R., and W. Newman. 1969. Opportunities for timber management intensification on the national forests. J For. 67: 482-485. Marty, R., C. Rindt, and J. Fedkiw. 1966. A guide for evaluating reforestation and stand improvement projects. USDA Handb. 304. 24 p. Nadeau, J-P. 1970. Economics of reforestation. For. Chron. 46: 487-490. Nova Scotia Department of Lands and Forests. 19813. Forestry discussion paper. Unpublished. Halifax. Nova Scotia Department of Lands and Forests. 1981b. Proposed second generation agreement. Draft No. 2. June 5. Halifax. Nova Scotia Research Foundation Corporation. 1976. Nova Scotia forest resource model: final report. Report prepared for Nova Scotia Dep. Lands and Forest. 72 p. Runyon, K.L., J.V. Stewart, M.R.C. Massie, and R.M. Nacker. 1972. Analysis of the economic impact of sawmills and pulp and paper mills in Nova Scotia. Can. For. Serv., Marit. For. Res. Cent., Inf. Rep. M-X-33. 94 p. Runyon, K.L., J.V. Stewart, M.R.C. Massie, R.M. Nacker, and J.H. Smyth. 1973. Economic impact of the Nova Scotia forest industry. Can. For. Serv., Marit. For. Res. Cent., Inf. Rep. M-X-37. 78 p. Sandoe, M., and M. Wayman. 1977. Productivity of capital and labour in the Canadian forest products industry, 1965 to 1972. Can. J. For. Res. 7: 85-93. Wall, B.R. 1969. Projected developments of the timber economy of the Columbia-North Pacific Region. USDA For. Serv., Wall, B.R. 1973. Employment implications of projected timber output in the Douglas-Fir Region, 1970-2000. USDA For. Serv., Res. Note PNW-le. 11 p. 159 Wall, B.R., and D. Oswald. 1975. A technique and relationships for projections of employment in the Pacific Coast forest products industries. USDA For. serVI, Res. Pap. PNW-189, 49 p0 Wiksten, A. 1970. An economic guide to silviculture. For. Woods Gordon & Co. 1974. Forest resource model: the demand module. Report prepared for the Department of Lands and Forests, Govt. Nova Scotia. 153 p. MICHIGAN STATE UNIV. LIBRARIES 1|”111111111111111II""l11|HIIWHIHIWIHHI 31293105439271