UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Simulation models : a useful tool in the evaluation of planted coastal Douglas-fir silvicultural management Onate, Nilda Elvira 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1997-0361.pdf [ 4.18MB ]
JSON: 831-1.0087758.json
JSON-LD: 831-1.0087758-ld.json
RDF/XML (Pretty): 831-1.0087758-rdf.xml
RDF/JSON: 831-1.0087758-rdf.json
Turtle: 831-1.0087758-turtle.txt
N-Triples: 831-1.0087758-rdf-ntriples.txt
Original Record: 831-1.0087758-source.json
Full Text

Full Text

SIMULATION MODELS: A USEFUL TOOL IN THE EVALUATION OF PLANTED COASTAL DOUGLAS-FIR SILVICULTURAL MANAGEMENT by NILDA ELVIRA ONATE Agricultural Engineer, Universidad de La Plata, 1977 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Forest Resources Management We accept this thesis as conforming ^the/eguirecLstandard n THE UNIVERSITY OF BRITISH COLUMBIA June, 1997 © Nilda Elvira Onate, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. 1 further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT This study showed how selected growth and yield simulations models compare in predicting the results of a set of prescriptions for planted stands of coastal Douglas-fir (Pseudotsuga menziesii var. menziesii [Mirb.] Franco). Estimates of growth rate, mean diameter, and yield were compared over a rotation for various sites. An economic analysis of the projected estimates was conducted with particular focus on a site index of 30 m. The selected models showed a wide variation predicting the effects of treatments (individually or in combination) and in predicting yields. This variation was attributed mostly to the databases used to develop the mathematical expressions which predict growth. The response to commercial thinning, in particular, was based on diverse assumptions among the models. The growth resulting from the simulated use of a fertilizer differed in assumptions regarding the duration of the effects of the fertilizer, the increase and the acceleration of the volume growth rate, and the amount of volume produced. Model predictions were mostly in agreement for the simulation of stands established at low initial densities (400-1100 sph). Model predictions of total standing volume were in agreement for almost all the prescriptions for site index 30 m. The economic evaluation determined the best combination of a sequence of scheduled activities. Precommercial thinning in combination with commercial thinning proved to be a financially attractive management alternative for initial planting densities of 2500 sph. A stand planted at 1100 sph and commercially thinned at ages 40, 50 and 60 showed the best return, although entry at age 60 produced the highest net benefits. Stands planted at 2500 and 1600 sph ii and precommercially thinned did not offset the costs incurred in the prescription, planted (400-1100 sph) did not generate an acceptable return on the investment. i i i TABLE OF CONTENTS Page ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES vi ACKNOWLEDGMENT vii DEDICATION viii CHAPTER ONE. INTRODUCTION 1 CHAPTER TWO. LITERATURE REVIEW 4 2.1 Growth and yield projection systems 5 2.1.1 Introduction 5 2.1.2 Growth and yield modelling philosophies 6 2.1.3 Growth and yield models in the Pacific Northwest 8 2.1.4 Summary of model review 17 2.2 Experimental findings on cultural practices to grow planted stands 18 2.2.1 Introduction 18 2.2.2 Site preparation 19 2.2.3 Competing vegetation 20 2.2.4 Espacement 21 2.2.5 Precommercial thinning 23 2.2.6 Commercial thinning 25 2.2.7 Fertilization 28 2.2.8 Pruning 29 2.3 Computer simulated studies of silvicultural management alternatives 31 2.3.1 Background and recommendations for juvenile spacing standards in BC 31 2.3.2 Incremental silviculture: pruning financial analysis 33 2.3.3 Financial analysis for selected and combined silviculture treatments for the BC coast 34 2.3.4 An economic evaluation of commercial thinning Douglas-fir in the coastal region of British Columbia 35 2.3.5 PROGNOSIS, SPS and TASS yield comparisons for even-aged interior Douglas-fir 37 iv 2.3.6 Three-dimensional representation of Douglas-fir volume growth: comparison of growth and yield models with stand data 38 2.3.7 Summary 39 CHAPTER THREE . DATA AND METHODS 42 3.1 Prescriptions used 43 3.2 Start-up conditions for the four models chosen 45 3.3 Economic evaluation 47 3.3.1 Data requirements for the economic evaluation of the prescriptions 47 Log volume calculations 47 Gross log values 49 3.3.2 Assumptions used in the economic evaluation of a prescription 51 3.3.3 The economic evaluation criteria 52 3.3.4 Net present worth 54 3.3.5 Quicksilver 54 3.3.6 Interest rate 55 3.3.7 Pruning: volume calculation and economic evaluation 55 CHAPTER FOUR. GROWTH AND YIELD ESTIMATES 58 4.1 Results 58 4.1.1 Site index 20 m 58 4.1.2 Site index 30 m 71 4.1.3 Site index 40 m 72 4.2 Fertilization 73 4.3 Commercial thinning 75 4.4 Discussion 76 CHAPTER FIVE. ECONOMIC EVALUATION OF THE SIMULATED PRESCRIPTIONS 79 5.1 Results and discussion 79 5.2 Pruning: results and discussion 84 CHAPTER SIX. CONCLUSIONS AND RECOMMENDATIONS 87 6.1 Simulation models 87 6.2 Economic evaluation 88 LITERATURE CITED 90 LIST OF TABLES Table Page 1. Growth and Yield models in the Pacific Northwest region 8 2. Outline of prescriptions 44 3. Model simulations: summary of assumptions 46 4. Summary of coastal Douglas-fir and Pine grades as in Scaling Manual for British Columbia 50 5. Economic evaluation of a prescription: summary of assumptions 53 6. Economic evaluation of pruning: summary of assumptions 57 7. Mean dbh at site index 20 m 59 8. Mean dbh at site index 30 m 60 9. Mean dbh at site index 40 61 10. Maximum M.A.I, at site index 20 m 62 11. Maximum M.A.I, at site index 30 m 63 12. Maximum M.A.I, at site index 40 m 64 13. TSV at site index 20 m 65 14. TSV at site index 30 m 67 15. TSV at site index 40 m 69 16. Net present worth and optimum economic rotation age 81 17. Net present worth of pruning 85 vi ACKNOWLEDGMENTS I greatly appreciate all the support I have had from a great man, my father, as well as from my mother. I am pleased to acknowledge the Government of Canada for funding the first two and a half years of the program through the Canadian International Development Agency (CLDA). I am grateful to Dr. Andrew F. Howard whose funding from a Ministry of Forests project supported me for one year. I am also grateful to Dr. John McLean who has showed his best skills as a Graduate Student Advisor. I would also like to acknowledge Dr. Peter Marshall, whose kindness led my first steps as an international student. I have to acknowledge and thank Heather, my English tutor. She strongly encouraged and supported my work in order to achieve the last step of this project. Finally, my special thanks to my beloved friends without whom my world couldn't be this great. Thanks to you all, Nilda / want to dedicate this thesis to my grandfather. His high principles, his generosity, his appreciation of life, and his education advocacy have been inspiring. viii CHAPTER 1 INTRODUCTION Experimental evidence on managed second-growth and planted stands of Douglas-fir (Pseudotsuga Menziesii var. menziesii [Mirb.] Franco) and other species in coastal British Columbia shows that silvicultural treatments can help to mitigate the imminent harvest level decline in the forests of British Columbia (BC) (Baskerville,1986; Ministry of Forests, 1990; Smith, 1983, 1986; Weetman, 1991). In the short term, such treatments can shorten the time it takes to ensure operability of the stands to offset possible timber supply deficiencies. In the long term, specific prescriptions become strategically important in the planning of the desired outcome of a stand, as well as in the assessment of the economic benefits of the silvicultural management plan. The coastal forests of BC, which have high rainfall and long growing seasons, are associated with a low level of hazard in growing timber. Weetman (1991) remarked that the biological reasons that could explain these phenomena are not clearly understood. These coastal forests are also virtually pest and disease free. The above-mentioned biological and climatic reasons, the higher yields, and the higher market value for large trees of nearly all species in these coastal stands, all contribute greatly to the success of plantation silviculture. Douglas-fir is highly responsive to silvicultural management. This species is also flexible in the sense that it can express its responsiveness across a wide range of densities (Ministry of Forests, 1990; Mitchell and Cameron, 1985). The commercial importance of Douglas-fir is l recognized worldwide (Briggs and Fight, 1992; Cahill, 1988). These attributes make Douglas-fir an ideal candidate for plantation forestry. However, a problem arises with identifying the silvicultural practices and prescriptions most suitable to the species and the region under study. Therefore, forest managers need to rely on tools that allow them to estimate future growth and yield of the stands. Over the last thirty years, the development of computerized growth and yield simulation models has enabled forest managers to more easily predict levels of future yield. These computer simulation models may also be applied to assess the effects of silvicultural practices on diameter and volume growth rate along with other tree and stand variables. TASS (Mitchell and Cameron, 1985), XENO (Northway, 1990), PROGNOSIS (Stage, 1973), SPS (Arney, 1980) and DFSLM (Curtis etal, 1981) are widely used growth and yield simulation models in the Pacific Northwest region. However, little is known about how they compare to each other when predicting growth. In order to be able to use these models and to capitalize on their built-in capabilities, one needs to compare their outputs. Most simulation-based studies that attempt to address silvicultural management do not indicate the reasons why a particular model was selected. Some of the studies aim at comparing the structure of the models; others aim at the basic assumption that drives them; still others aim at the fact that certain models underestimate the growth for a specific regime. As a result, there is an array of projection systems that have been extensively used, but arbitrarily selected. Growth and yield predictions are a necessary component of the economic evaluation of a prescription or management plan. The assessment of the economic benefits of a prescription depends on the levels of yield projected at any point in time during the timber-growing cycle. The economic evaluation of a prescription may indicate the best combination of a sequence of scheduled treatments and the viability of the forest management plan over the span of the rotation. 2 Therefore, an economic evaluation is undertaken in order to know if the outcome of a prescribed stand results in an acceptable return on the investment. The objectives of this study are to assess how selected growth and yield simulation models compare in predicting the results of a set of prescriptions for planted stands of coastal Douglas-fir, and to identify the most economical range of prescriptions among the prescriptions examined. Firstly, estimates of growth rate, mean diameter, and yield are compared during the span of a rotation for various sites. Secondly, economic analyses of projected estimates are conducted with a particular focus on a site index 30 m. Chapter 2 reviews the literature on growth and yield simulation models, on findings from field trials of individual treatments, and on computer simulated studies of individual or combined treatments and prescriptions. Chapter 3 describes the data and methods used in the study, including the specifications of silvicultural prescriptions. Chapter 4 presents the results and the discussion of projection system estimates. Chapter 5 presents and discusses the results of the economic evaluation of the prescriptions on site index 30 m. Finally, Chapter 6 offers conclusions and recommendations based on the findings from this study. 3 CHAPTER 2 LITERATURE REVIEW This review concentrates on the specific aspects of growth and yield which are required for the silvicultural management of a stand. Its focus is on growth and yield simulation models, on cultural practices recommended in order to grow trees and, lastly, on particular studies which use simulation models to project silvicultural management alternatives and analyze the projected estimates from an economic viewpoint. This review is limited to the Pacific Northwest (PNW) of the United States and coastal BC. Firstly, growth and yield projection systems are defined and modelling philosophies are introduced to show the basic categories into which projection systems fit. Some of the simulation models existing in the PNW region are reviewed and grouped into basic categories in order to identify the models for use in this study. The literature review describes the models and their fundamental assumptions, the database used to build the models, the intent and objectives of the developer, the geographic regions for which they are developed, as well as their applications, data input, and output. Secondly, experimental silvicultural treatment results are examined for Douglas-fir in order to investigate timing, intensity, frequency, and type of treatment, as well as any other decision variable necessary to establish the host of prescriptions that provide the data to carry out this study. Finally, other studies which employ these simulation models to examine silvicultural management alternatives are reviewed with the objective of identifying their purposes 4 and results. These purposes and results will be analyzed in order to determine which aspects need to be investigated in order to maximize the effectiveness of silvicultural management, and to select the appropriate economic approach for assessing the stand management investment. 2 . 1 GROWTH AND YIELD PROJECTION SYSTEMS 2 . 1 . 1 Introduction A model, in the most general sense, is a physical or mathematical depiction of some item or process. "Growth and yield models are mathematical expressions that quantify a tree or forest change" (Buchman and Schieffly, 1983). "These mathematical expressions are relationships between the amount of yield or growth and the various factors that explain or predict this growth. The relationships used in the models are most conveniently stated as equations. These equations can be implicit or explicit in form, and linear or non-linear in their relationships" (Davis and Johnson, 1986). "The biological precision varies considerably from system to system due to the following factors: intent and objectives of a model, availability of good data, philosophy of the modeller, as well as many other practical considerations. Models are thus abstracts of reality which should be used only as one component of a decision-making process" (Goudie, 1987). Brand and Holdway (1983) stressed the user's need for performance information in order to evaluate models. According to these researchers, projection systems should provide extensive information about model performance presented in an easily understood and useful format. Furthermore, models must be evaluated and verified before they can effectively be compared and used. Evaluations typically require the synthesis of numerous quantitative and qualitative factors. The literature on evaluation criteria for projection systems often contains philosophical discussions providing a broad perspective to the evaluation process. Although authors differ in their attitudes and opinions, they recognize that no projection system can perfectly represent the 5 real system being modelled, and that little is gained from proving that a projection system is an inexact copy of nature. Rather, the concern should be how well a projection system performs in relation to alternative systems. As Forrester (1968, cited in Brand and Holdaway, 1983) states, "Model validity is a relative matter". The usefulness of a mathematical simulation model needs to be judged in comparison with the mental image or other abstract model which could be used instead. Most authors also recognize that projection systems cannot be evaluated in the absence of some stated objective (Goulding, 1979). "A system's utility depends upon the user's objectives. Only after objectives have been specified can the benefits of a system be judged" (Brand and Holdaway, 1983). Finally, Buchman and Shifley (1983) contend that evaluation is partially subjective. "System capabilities often fall short of desired performance, and for a given application each projection system has a unique set of strengths and weaknesses. Selecting the "best" system requires synthesizing a whole array of quantitative and qualitative factors". 2.1.2 Growth and yield modelling philosophies Munro (1974) described three modelling philosophies. These modelling philosophies can easily be distinguished on the basis of two features: primary unit and inter-tree dependency status (parameter requirements). The first philosophy of Munro assumes that the primary unit of stand modelling is the single tree and that inter-tree distance is a necessary parameter. All models which are single tree/distance dependent are capable of producing very detailed information about the structure of the stand and of providing tree summaries. Potentially powerful uses of those models which incorporate crown dimensional increments include studies of tree to tree competition, pruning, insect defoliation, mistletoe infections, top dieback, and bole form change. The effects of various cultural programs such as thinning, spacing, and fertilization also may be scrutinized. These crown models may not be able to project the growth of inventory plots unless 6 unconventional data or procedures are available (Mitchell, 19756). Crown models are costly to develop and to operate because of their attention to detail and because of their dependence on highly trained and specialized users. The second philosophy of Munro assumes that the primary unit of stand modelling is the single tree, and that inter-tree distance is not a necessary parameter. These individual tree models do not attempt to forecast the development of particular trees measured in the field. However, they do represent the stand as an aggregation of competing stems and simulate the growth of trees as a means of incorporating the dynamics of stand development into yield projections. The depth of detail or realism is limited only by the modeller's ability to understand and to simulate the biological process that drives the growth of trees and forests. The models are primarily designed to provide stand information. Advances in single tree distance-independent modelling show that much of the information thought to be available only from single tree/distance dependent models can now be obtained without inter-tree distance measures. The effects of spacing, thinning, and fertilization can be evaluated efficiently for operational purposes with distance independent models. The third philosophy of Munro assumes that the primary unit of stand modelling is the stand, and that information concerning individual trees is not necessary. Whole stand models are usually inexpensive to operate, and accurate within the range of the basic data. Species, age, site quality, and density or management regime are related directly to volume, basal area, diameter, height, number of trees, and other stand parameters. Whole stand distance independent models utilize conventional inventory information which produces readily available stand yield tables. 7 2.1.3 Growth and yield models in the Pacific Northwest Some of the growth and yield simulation models used in the PNW are outlined below according to Munro's philosophies (Table 1). The modelling level, spatial dependency and input and output of the models are given. The models are described in more detailed below. Table 1. Growth and yield models in the Pacific Northwest Model Modelling level (Primary unit) Spatial dependency (Parameter) Input Output TASS Single tree Distance/dependent Stem map + Indices Managed Stand Yield Tables XENO Single tree Distance/dependent Stem map + Indices Managed Stand Yield Tables SPS Single tree Di stance/independent Stand Table Stand Avg. PROGNOSIS Single tree Distance/independent Stand Table Stand Avg. DFSfM Stand Stand Table Stand Avg. TASS The development of the Tree and Stand Simulator (TASS) was initiated by Mitchell (1969) at the Canadian Forestry Service, with subsequent work completed mainly by Mitchell (1975) at Yale University, by Goudie (1980) at the University of Idaho, and by Mitchell and Cameron (1985) at the BC Ministry of Forests. TASS is one of the major components of a system of models called SYLVER. 8 TASS is a biologically-oriented model (crown model) designed to assess the effects of cultural practices and environmental factors on the growth and yield of Douglas-fir and other species. It consists of a series of mathematical equations which are used to grow trees in a simulated three dimensional space. The height-age curve for a particular site drives height increment and branch growth. The simulation assumes that some trees will grow more slowly than others and will succumb to competition earlier because of genetic variation in height and branch growth. Crowns expand and contract asymmetrically in response to internal growth processes, physical restrictions imposed by competing crowns, environmental factors (site quality, disease activity, insect defoliation, animal damage), and cultural practices (espacement, thinning, fertilization, and pruning). The volume of a tree is naturally increased by its foliage. This volume increment is annually distributed and accumulated on the bole. The model simulates this process and, consequently, produces the corresponding tree summaries and related stand information. TASS is also extended to simulate the production of juvenile wood. The model is calibrated to obtain the yield of remeasured plots on coastal BC. TASS overpredicts yields compared to those plots observed in the field. Field plots, which are not incorporated into TASS's calibration, usually have irregular stocking, pests, and other factors affecting growth and yield. The TASS model mimics the biological processes with considerable fidelity. High resolution, sensitivity, and a variety of treatment options are among the attractive features of the model. Small changes or perturbations can be traced through the system over the life of the stand. However, TASS is an expensive model to operate. It requires specialists on a mainframe or on mini-computers capable of completing computations in reasonable time and, as a consequence, this model is not convenient for general use by forest managers and decision makers. 9 In order to provide foresters with more readily available yield information, the BC Ministry of Forests Research Branch developed the TIPSY program (Mitchell, 1992). TASS was used to generate the yield tables in TIPSY's database. The TIPSY program is described in the following section. TIPSY TIPSY (Mitchell, 1992) is a table interpolation program for electronically retrieving stand yield information generated by TASS, available in reports such as Mitchell and Cameron, 1985. TIPSY is a simple program and requires a personal computer for operation. TIPSY retrieves the appropriate yield tables from its database, customizes the information, and displays yield summaries in seconds. TIPSY can be used at the stand level for evaluating silvicultural options in managed stands, or at the forest level for timber supply analysis in concert with VDYP (Variable Density Yield Prediction system; Ministry of Forest, 1995). TIPSY addresses managed even-aged stands of natural origin including the untreated controls which are needed in order to evaluate the response of the stand to silvicultural regimes. TIPSY can generate yield information for coniferous species of commercial importance growing on the coast and in the interior regions of BC. The management variables include species, regeneration method, establishment density, and precommercial thinning (PCT). Users may select a wide range of initial densities and/or residual densities by species in accordance with the input options developed. Precommercial thinning, when selected, occurs when stands reach a height of 6 m on the coast and 4 m in the interior. TIPSY-generated tables are based on the growth trends observed in fully-stocked research plots. Consequently, TIPSY reports the potential yield of a specific site, species and management regime. Operational Adjustment Factors (OAFs) are introduced in order to 10 compensate for nonproductive sites in stands, and for pest damage and other factors that lower the productivity of the stands in the operation. TIPSY yield tables include total and merchantable volumes, basal area, diameter, number of trees, mean annual increment and statistics for the largest 250 "prime" trees. A disadvantage with TIPSY is that it does not allow the user to change the age at which simulated spacing occurs. TIPSY can only incorporate precommercial thinning and initial planting density in order to show the effects of silvicultural treatments on a stand; this limits its application to the assesssment of limited treatments during the rotation. However, it is a very practical and user-friendly program which can be installed on a personal computer which makes it appropriate for general use. XENO XENO is an individual tree distance dependent model developed by Northway (1989) using McMillan Bloedel (MB) stand growth data. MB has maintained an aggressive research program on growth and yield since 1954, providing large amounts of data on natural stands and on the early development of spaced and planted stands of pure Douglas-fir and hemlock (Tsuga heterophylla (Raf.) Sarg.) in BC. XENO is based on the assumption that all the effects of site index may be accomodated by changing the age at which a stand attains a given site height. XENO is capable of growing single or mixed species under both natural and managed regimes. This model responds to establishment densities, to ingrowth, to stocking control, and to fertilization. The input procedure requires either one entry for all species, or entries for each particular species. In order to run the model, the input data requirements consist of site index values, number of trees/ha, percentage of survival in the plantation, age of trees and species. 11 Another choice for the user is the 'default' option for input procedures and input data that XENO provides. XENO's output consists of stand summaries at any specified utilization level and a stand table. The stand table output gives a breakdown by diameter class (0 to 60 cm, in 5 cm increments), and by species, including number of trees/ha, volume/ha, and volume per tree. Stand summary information refers to trees larger than the specified minimum diameter (pMIN) and to volumes excluding stump and top. The gross volume figure, which is also provided in the table, has no DMLN and it includes stump and top volumes. Volume estimates include: cumulative gross production (standing + mortality + thinning + spacing), cumulative merchantable production (standing + mortality + thinning + spacing), cumulative captured production (standing + thinning + spacing), and the mean annual increment of the cumulative captured production. Growth projections can be linked to a sub-model which offers an economic appraisal package. XENO is a flexible and user-friendly model, though it has not been used as extensively in this region as have other models. In order to know more about its performance, XENO must be utilized more intensively. This model was built to aid the investigation and add to the improvement of forest management as part of MB developmental programs. XENO may simulate the silvicultural management alternatives currently under discussion by foresters in BC. These alternatives appear to be the most effective type of silvicultural management for use in the near future. SPS The Stand Projection System (SPS) was originally developed for even-aged stands of coastal Douglas-fir. The data base consisted of individual tree and plot histories, originally 12 compiled and then edited to develop DFSIM (Douglas-fir simulator) from extensive PSP (Permanent Sample Plots) information in coastal Oregon and Washington. Arney (1972; 1980) selected the single tree distance-independent model structure using height growth of dominant trees to drive the projection system. Thus, even-aged stands of one or two species are the stand structure best projected by SPS. The SPS program may be used with existing stand tables from a forest inventory or from stand averages by species. The user can input an initial stand table to project existing stands or allow the system to generate start-up conditions for the projection of regenerated stands. Any site index, stand density, thinning, fertilization, or number of trees may be specified for projection of pure and of mixed stands. SPS produces estimates based on fully-stocked stands. The predictions of average yields can be estimated in SPS by using a clumping factor that increases the competitive index (CCF) in order to account for rock outcrops and for other nonproductive areas found in the general forest system. SPS is a simple and accurate computer simulation program in which a forest manager may produce yield tables or inventory updates. It provides a summary yield table and a mean annual increment chart. The stand table report provides an updated stand table by species and by diameter class for use in short term inventory growth updates. Growth projections can be linked to economic appraisal packages and to harvest scheduling systems. SPS is a flexible and user-friendly model. It can project the wide silvicultural management objectives that are currently discussed as possible future management strategies for BC forests. However, the database used to develop the model is the same one used to build DFSIM. This database reflects the concentrated effort to render much needed predictions in the Pacific Northwest region, though the plots used to make up the database represent stands located 13 predominantly in the U.S. Northwest (Washington and Oregon). Thus, the SPS user needs to be cautious when evaluating Douglas-fir silvicultural management for BC stands. PROGNOSIS PROGNOSIS is a single tree distance-independent model (Stage, 1973; Crookston, 1990, 1991; Ferguson, 1991, 1993; Moeur, 1985). Stage (1973) originally conceived the system in order to meet projections for the Inland Empire (western Montana, Idaho and eastern Washington). The stand structures that existed in 1973 ranged from even-aged stands of lodgepole pine (Pinus contorta var. latifolia Engelm.) originating from past wildfires, to uneven-aged stands containing up to eleven species. Uncertain man-made disturbances and periodic outbreaks of numerous pests added to the environmental complexity of the forest in the Inland Empire. The original intent was to develop a model that could project the development of managed and unmanaged existing stands and estimate yields of stands that had been regenerated by natural or artificial means. The conditions of the stands, the paucity of adequate PSP data, and the need for flexibility led to the diameter-based model structure which was primarily developed with existing inventory data. These result in projections of average conditions, although self-calibration features in the model allow for adjustments for stands that depart from average conditions. Another adjustment occurs during the course of the simulation. The large PROGNOSIS database describing long term future growth assimilates skewed predictions. Therefore, the data produced by the projection of trees growing fast at the initial stage of the simulation attenuates towards the average as the simulation proceeds. PROGNOSIS also allows the user to 'adjust' internal functions. The user may alter crown, height, basal area and other variables by using the appropriate values to modify estimates; however, it is recommended that adjustments be used with caution or left to experts. Model control is accessed by a keyword 14 system, which is represented by words associated with numeric data. Every keyword defines an activity and allows for data entry. The required data are species, diameter, inventory design, region and forest. PROGNOSIS was designed to be a component in a forest management planning system. Therefore, there are two levels of management application: planning for individual stands and planning for large ownerships that are comprised of many stands. PROGNOSIS can be used to update inventories and to generate yield tables. Summary tables describe stand characteristics such as basal area per acre, trees per acre, quadratic mean diameter, species composition, stand structure and volume (board foot, total cubic foot, merchantable cubic foot). The output also provides tree characteristics on diameter, total height, basal area percentile, per acre expansion factor, crown ratio, board foot volume, diameter growth, and height growth. The U.S. Forest Service has provided continuous support allowing the PROGNOSIS effort to be integrated for actual use. There now exists staff support, upgraded versions and reports of the model, guidelines to set up the program, and charts briefly describing intended use and instructions for the operation. The model continues to be developed and updated in order to improve predictive capability and to provide more efficient operation. Over the years, since the Inland Empire variant was first released, other variants have been developed, tested, and released by the U.S.D.A. Forest Service's Mensuration and Systems Development Group, located in Fort Collins, CO. PROGNOSIS, with its seventeen variants, is the most widely used forest growth model in the United States. The Inventory Branch of the Ministry of Forests in British Columbia is currently investigating its suitability for projecting stands in BC. At the time of this study, PROGNOSIS was being refined by the BC Ministry of Forests in order to calibrate it for the coastal and interior regions of BC and to convert its equations to metric units. 15 PROGNOSIS is a complex and powerful tool with great potential to explore silvicultural prescriptions and various harvesting systems. Its strengths as a single tree distance-independent model include links to other associated extensions (sub-models) which have been introduced to simulate pest activities and damage, the development of tree crowns and the understory vegetation, as well as other management alternatives. However, PROGNOSIS is not particularly user-friendly, and requires familiarization with a whole body of information and skills. DFSIM The Douglas-fir Simulator (DFSLM) is a 'stand average' model derived from small research plots. Private and governmental organizations in the Pacific Northwest combined their data to develop more broadly based yield estimates for managed stands of coastal Douglas-fir. This resulted in a whole stand model (DFSLM) developed by Curtis et al. (1981) which describes the development of relatively homogeneous, even-aged stands of Douglas-fir for a range of initial conditions and treatment regimes which could be modified to incorporate updated information. The developers combined regression relationships into a computational sequence determined, in part, by their concept of the growth of forest stands, by their predictions of the relative reliability of various estimated relationships, and by information available in the basic data. DFSIM represents the development of Douglas-fir and is not applicable to other types of stands. This model was not designed as a means of projecting inventory data. It can, however, be used to estimate probable future development of existing stands, providing that these are within the range of the basic data and that discretion is used in its application. DFSIM generates yield tables for managed stands, which include estimates of the effects of initial spacing, precommercial and commercial thinning, and nitrogen fertilization. Summary tables show statistics for a live stand, a cut stand and a residual stand, as well as cumulative values 16 for cut and mortality for stems over 4.1 cm diameter breast height (dbh). These statistics are produced for each thinning date and for final harvest. Intermediate summaries are optional and can be requested at any specified age. Optional summary tables are provided for merchantable volumes in cubic meters of stems over 14.2 cm dbh, and in cubic meters and board feet for stems over 19.3 cm dbh. DFSIM is constrained by limitations arising mainly from its basic data. One limitation pertinent to the variables of interest in this study is that the database includes some thinned stands representing thinnings which many foresters would not consider to be reasonable. Among the other limitations relevant to this study are that the basic data contain few plantations with less than 741 initially established stems per hectare (sph), there are few stands with early precommercial thinning to less than 741 sph, and there are few stands known to have had low initial stocking. 2.1.4 Summary of model review All of the models mentioned above are stand-level designed systems. They represent the most widely used growth and yield simulation models in coastal BC. The databases of the models are made up of research plots from fully-stocked stands with the exception of PROGNOSIS. Therefore, their projection estimates are expected to be higher than the estimates that could be obtained under operational conditions. The various models differ greatly in their assumptions. Except for TIPSY, the models can project the alternative silvicultural management options that are of potential interest to the region. The input information for each model varies with respect to the degree of detail used to design its database which, in turn, is dependent on the modelling philosophy employed by the modeller. The displayed output information (stand averages and/or tree summaries) of the models also differ from each other due to the fact that 17 each model reflects a different philosophy. More detailed information is usually obtained from single tree distance-dependent models; however, at the present time detailed information can be obtained from single tree distance-independent models as well, as is the case for PROGNOSIS. The models selected to carry out this study are XENO, SPS and DFSLM, as well as the TIPSY program. The following requirements were established for the selection of the models: inexpensiveness to operate, user-friendliness, ready availability, capacity to run numerous simulations in a day, and easy installation in a personal computer. The TASS model didn't fulfill any of the above-mentioned requirements. The PROGNOSIS model partially sastisfied these requirements, although it is not user-friendly and its operation is complex and requires familiarization with the applications. Another inconvenience is that PROGNOSIS output is not displayed in metric units. In sum, the selected models can project the alternative silvicultural management options that are of potential interest to the BC coastal region. The selected models have the ability to run different treatments with input data readily available to the user or forest manager. XENO, SPS, DFSIM and TIPSY will be used to compare their predictions on mean diameter, growth rate and total yield for a range of prescriptions for planted coastal Douglas-fir. The comparison of these models may allow the forest manager and other users to choose the model that best suits to his/her management plan. 2.2 EXPERIMENTAL FINDINGS ON CULTURAL PRACTICES TO GROW PLANTED STANDS 2.2.1 Introduction In order to grow a planted crop of trees over a rotation, foresters need to take into account many cultural practices including: site preparation, vegetation control, espacement, precommercial thinning, commercial thinning, fertilization and pruning. Experiments on these cultural practices 18 have been useful for identifying the decision variables involved in each activity. The decision variables represent parameters to be established in the application of a treatment or an activity. Once these parameters are defined, the forest manager or planner can specify the prescription or management plan. A review of the decision variables that comprise an activity was needed to determine the range of values for average conditions in order to specify the prescriptions for this study. The experimental findings also were examined in order to understand the interaction and complexity of the decision variables. These variables are dependent on various factors: species, site quality, genetic makeup, density and climate. Thus, the values of the decision variables are difficult to determine. Vast amounts of research on the application of decision variable values have been done; however, few conclusive results are available. Despite the wide range of data and the controversy among foresters on the application of these data, the aforementioned findings from field trials are the most valuable and reliable tool at hand for determining the characteristics and traits of Douglas-fir in the PNW region. In the following sections, literature on the cultural practices or management activities applied to Douglas-fir are reviewed to characterize the corresponding decision variables. 2.2.2 Site preparation Site preparation is a requisite management activity if a crop of planted trees is to be established. The forest land area to be used must be cleared either manually or mechanically, according to variable soil characteristics such as texture, fertility, and slope. These activities will differ if a tree crop existed before the establishment of a new crop. In this case, the time lapse between crops is also a possible factor to be taken into account. In the coastal region of BC, it is 19 common practice to remove the materials left after logging. These materials are moved to the edges of the forest land area and are then burned.1 Lately, in the coastal region, mechanical methods have been used as extensively as have manual methods in which broadcast burning is prescribed. Once the forest site has been prepared, tree planting follows. The decision variables for site preparation include type and timing. Depending on the type, decision variables may vary in terms of labour (manual) or machine hours (mechanical) employed. 2.2.3 Competing vegetation Competing vegetation refers to grass, brush and any other species besides the desired trees growing on the planted site. Trees may be affected by competing vegetation at early stages of growth, depending on site and planting density. Competing vegetation can decrease available resources on a site and the ability of the tree to utilize them (Daniels et al., 1979). Cole and Newton (1987) studied growth responses of Douglas-fir to crowding and to nonconiferous competition in 5-year-old plantations on three site types in the Oregon Cascades. Grass and red alder competition, as well as crowding, adversely affected the growth and stand biomass on a per tree basis. "When the primary objective of the stand is the production of maximum tree size, then low stocking and the control of competing vegetation will produce larger trees faster than with other management alternatives. Because of competition, early losses in individual tree growth may not be recovered after thinning" (Cole and Newton, 1987). Competing vegetation can be controlled by manual, mechanical, chemical and biological methods (Smith, 1986; Ministry of Forests, 1995). The control of vegetation takes place when the stand is 3-5 years old.1 The Ministry of Forests Guidebooks (1995) indicate control of vegetation at any time between 1-8 years, depending on site and growth of the stand. The decision variables for competing 1 Personal communication: J.W. Rodney, R.P.F.; Staff Silviculturalist, Weldwood of Canada Ltd. vegetation are type, timing and frequency. Depending on the type, the decision variables may vary in terms of labour (manual), machine hours (mechanical), dosage (chemical) and animal units (biological) employed. 2.2.4 Espacement Espacement refers to the distance between the trees planted on a unit area of land. Studies concerning espacement are carried out on trees grown at relatively constant spacing intervals until harvest. Provided undesirable vegetation is controlled, trees grow without intraspecific competition until the roots and crowns of adjacent trees meet; it is then that competition among the trees increases and comes into play before harvesting (Smith, 1959). Espacement decisions should be made based on management objectives; a particular regime would be financially attractive if it resulted in the highest monetary return while fulfilling specific product needs (stem diameter, rotation age, stand volume, wood quality). On a unit area of land, the distance between the trees (espacement) determines the number of trees to be planted (planting density). Thus, espacement and planting density define two aspects of the same activity: the planting operation. The decision variable for espacement is the number of trees to be planted per unit area (initial planting density). Although the results of many experiments have been published, there are few conclusive recommendations as to optimal planting density. According to Cole and Newton (1987), early losses in individual tree growth due to crowding may not be recovered after thinning. These authors also conclude that allowing Douglas-fir to reach a degree of crowding that maximizes current height growth probably jeopardizes both future height growth and the resistance to snow breakage. Among the normal range of espacements, the effects of differing intensity of competition on height growth are primarily attributed to site quality and is more common on poor sites (Reukema, 1979). 21 The benefits derived from fairly wide, uniformly planted plots on relatively poor site land have been recorded over the years from Douglas-fir stands planted in 1925 at Wind River, near Carson, Washington. Those plots, planted as wide as 3 m by 3 m (1100 sph) and 3.5 m by 3.5 m (816 sph), show earlier merchantability of trees on the more widely spaced plots, coupled with greater height growth than plots planted as close as 1.2 m by 1.2 m (7000 sph), 1.5 m by 1.5 m (4444 sph), and 2.4 m by 2.4 m (1700 sph) (Reukema, 1959; 1970). Reukema and Smith (1987) reported similar conclusions (except for the greater height growth) over 25 years for Douglas-fir planted at various spacings on a good site in BC. The results of five trials (planted range from 1 m to 5 m wide) showed that the choice of initial espacement is among the most important factors influencing bole and crown development, and stand growth and yield. Initial wide espacement, such as 6 m by 3 m (500 sph), produced large trees of high value and satisfactory quality. Reukema (1970) reports that diameter growth of even the largest trees is clearly affected by spacing. At age 43, the 100 largest (dbh) trees per acre were currently 67 percent larger in diameter and 39 percent taller on the 3.5 m by 3.5 m (816 sph) planted plot than on the 1.2 m by 1.2 m (7000 sph) planted plot. Thus, it appears that these differential increments, which favour the wider espacement, also increase with time. In order to prevent the decline of diameter growth rates, the trees in stands widely planted should be thinned. The 3 m by 3 m (1100 sph) and 3.5 by 3.5 m (816 sph) planted plots could support commercial thinning at approximately age 30, and the 2.4 m by 2.4 m (1700 sph) planted plots could support thinning at age 43. These responses are probably typical of what can be expected on most low to medium quality sites (Reukema, 1970). Espacement experiments indicate that a range between 1100 (3 m by 3 m) and 816 sph (3.5 m by 3.5 m) is the optimum number of trees to be planted per hectare for coastal Douglas-fir. 22 Financial analysis of these densities may determine if the desirable silvicultural aspects of the stand are reached when the silvicultural investment is the most financially attractive. 2.2.5 Precommercial thinning Precommercial thinning (PCT) is also known as juvenile spacing, spacing, and thinning to waste. The removal of selected trees early in the timber-growing cycle concentrates the growing capacity of the land on the optimum number of vigorous, healthy trees. PCT increases the quality and value of products, improves the health and vigour of the stand, and reduces the costs of final harvest (Olympic Natural Resource Centre, 1990). "This [PCT] is a practical means of substantially increasing the production of usable wood" (Reukema, 1975). Through PCT, the operability of forest stands is ensured at an earlier age, and this time reduction may partially offset timber supply deficits in some areas. "The larger the trees must be in order to qualify as merchantable, the greater are the gains from precommercial thinning" (Omule, 1985; Reukema, 1975). The decision variables for precommercial thinning include the tree height, the timing (age of the stand), the frequency (number of entries), and the intensity (number of cut trees). Ideally, Douglas-fir stands should be thinned when leave trees are about 3-4.5 m tall and 10-15 years old (Olympic Natural Resource Centre, 1990; Reukema, 1975; Reukema and Bruce, 1977). PCT should be delayed only long enough for trees to express their growth and quality characteristics, and to be above such deterrents as brush competition and animal browsing. Further delay will usually result in unnecessary loss in usable production and an increase in thinning costs (Reukema, 1975). The practice of precommercial thinning depends primarily on the tree maximum age or size (Omule, 1985; Reukema, 1975). The most critical of these factors is probably the size a tree can attain before commercial thinning will be applied (Reukema, 23 1972; 1975). If the stand is precommercially thinned, time is saved and the trees will have a chance to reach commercial size earlier; on the other hand, if the stand is not precommercially thinned, the time required to reach commercial size is longer (Reukema, 1975; Reukema and Bruce, 1977). Gains from precommercial thinning in stands where leave trees are up to 20 years old or 9 m tall, whichever comes first, can still be quite substantial (Reukema, 1975). Research on intensity of PCT indicates that stands which are heavily thinned past the recommended time for PCT show an immediate physiological shock and physical damage (Harrington and Reukema, 1983; Reukema, 1975). The occurrence, duration, and severity of this damage are apparently related to thinning intensity, site quality, tree species and vigour, and age (Reukema and Hamilton, 1983). Reports by the BC Ministry of Forests (1992, 1994), contend that stands simulated with as few as 1200 sph can be considered for spacing; they recommend post-spacing densities of400 or 500 sph to yield the greatest financial returns. Omule (1985) conducted a study of precommercial thinning and commercial thinning which incorporated frequency, as well as intensity, in dense (3000 sph), young, coastal Douglas-fir stands under 30 years on medium sites in British Columbia. In order to verify the effects of various frequencies and intensities, he established two series of plots: a precommercial and a commercial thinning series. In the case of the spacing series, he recommended low thinnings (1500-1000 sph) favouring larger and healthier trees. Omule indicated "the need for further studies in order to evaluate the practicality of repeated PCT regimes relative to their biological regimes". The studies on PCT discussed above are in agreement concerning the time (10-15 years) at which the activity takes place; one-time application seems to be reasonable on most sites. These studies are also conclusive regarding the flexibility of PCT intensity on the number of individual trees to be selected for one entry. However, due to the limited research on the interaction of 24 frequency (number and timing of entries) and intensity (number of cut trees), the data are inconclusive. 2.2.6 Commercial thinning Commercial thinning (CT) is an intermediate harvest of timber before the final harvest. Thinning has two major objectives: the redistribution of total stand growth to fewer trees, and the utilization of additional merchantable material as mortality is captured during a rotation. Thinnings will affect stand yields in two ways which, in turn, will influence the choice of rotation length: it will increase the diameter growth of the average tree, and it will delay the culmination of mean annual increment (Olympic Natural Resource Centre, 1990; Smith, 1964; Worthington and Staebler, 1961). The decision variables involved in commercial thinning are timing, frequency, intensity and type. There are five classical thinning types: thinning from below, thinning from above or crown thinning, selection thinning, mechanical thinning and free thinning (Smith, 1986). Thinning methods are not examined in this study. Most of the experimental findings on commercial thinning apply to thinning from below, which is the method used most extensively in the Pacific Northwest. However, the application of crown thinning has recently been introduced to a few natural stands in the forests of BC.1 Studies on commercial thinning explore the interrelationship between timing, intensity, frequency and type. It is exceedingly difficult to judge the interaction between these disparate variables; they are reviewed and grouped in the following studies in accordance with the options the authors chose to exercise. 1 Weetman, G.F. Professor of Silviculture. Columbia. From lecture in Forestry 306. Faculty of Forestry, The University of British 25 Worthington (1961) studied the intensity and timing of thinning in a Douglas-fir plantation in the Olympic National Forest of Northwest Washington. The study compared commercially thinned natural and planted stands. The results showed a higher growth rate for the plantation trees (1360 sph at age 31) and demonstrated that thinning is a feasible commercial practice at the early ages of 26 and 31 years. In a study in the Douglas-fir region of Oregon, Staebler and Worthington (1961) concluded that commercial thinning should be started in the 30-50 age class, or even earlier, in order to pursue growth and quality improvement. They stressed that the objective of thinning older stands (over 70 years) focuses on the salvage of mortality and on the realization of income before final harvest. Heideberg and Haddock (1955) confirmed the difficulties in judging frequency and intensity in thinning practices. They assumed, in accordance with the existing evidence at the time, that 25 or 30 percent of the basal area of a stand can be cut when the stand basal area is not below one-half of the maximum basal area possible at a given age. This results in a very small actual loss in volume production, if any. They stressed that volume may be lost through mortality if the cutting intervals are too long. They also maintained that timber quality is compromised when stands are heavily thinned. Fleischer (1949) supported the premise, that long thinning intervals mean heavy cuts which produce larger limbs on the remaining trees. In addition, larger limbs cause larger knots as well as wider and/or uneven rings; these characterisitics may give rise to the disqualification of high-grade products. Worthington (1966) contributed to commercial thinning research with a study on intensity. He studied the thirty years of growth, after the first thinning, of a 60-year-old Douglas-fir stand at the base of Mount Walker, Olympic National Forest near Quilcene, Washington on site class IV. Worthington concluded that "heavy thinning substantially depressed the gross increment in ensuing years. However, a moderate thinning reduced the gross increment only slightly. Growth 26 was well above normal on both moderately thinned and unthinned stands, but was a little less than normal in heavily thinned areas. Presumably, the stand was too old for heavy thinning to sufficiently speed up residual tree growth in order to compensate for the loss of increment due to cut trees". The Worthington (1962) and Omule (1988) studies on response to thinning removed 40-50% of the basal area or volume at one time on heavily thinned plots. Stone (Ministry of Forests, 1993) substantiated these studies on the intensity of commercial thinning. He reported that a commercially thinned planted Douglas-fir stand in the coastal region of BC, in the range of 200 to 400 leave trees per hectare, proved to be economically superior to the corresponding unthinned regime for all simulation runs. For each growing regime (1600 sph, 1975 sph and 2500 sph), five thinning intensities were simulated (200, 300, 400, 500, 600 and 800 leave trees per hectare). Stone concluded that the economic gain from commercial thinning is modest, but still worth pursuing. Reukema and Piennar (1973) studied volume gains, when thinnings are delayed beyond the optimum initial entry. The authors reported a slight gain in usable volume over standard thinnings after making delayed, light, frequent thinnings in a high-site-quality stand of nearly pure Douglas-fir in Grays Harbor County, Washington. The studies cited on commercial thinning substantiate the benefits of one-time entry in the age range of 30-60 years. In older stands, thinning mainly contributes to the salvage of mortality and the realization of income before the final harvest. The degree and the interaction of commercial thinning intensity and frequency is difficult to establish. Studies on intensity (number of cut trees), frequency (number of entries) and timing (years between intervals) suggest that these variables should be considered carefully whenever the volume gains and the timber quality of commercially thinned stands are to be maintained. Economic gains from commercial thinnings indicate heavy thinnings are commercially viable. 27 2.2.7 Fertilization Fertilizer in forests is usually employed to enhance tree growth. The most important incentive for forest fertilization is to produce more wood on a smaller land base. In most cases, forest fertilization results in a relatively short term increase in stand growth. "Only in rare cases where the amount of added nutrients is large in relation to the site nutrient capital does fertilization improve long-term site productivity" (Miller and Tarrant, 1983). The response to fertilizers usually results in a reduction in the rotation length, as fertilization increases the rate of stand development. In most cases, a fertilized crop will not differ significantly from one which has been grown over a longer rotation, but has not been fertilized. Therefore, fertilization can be used strategically to shorten rotation in order to bring stands to merchantability at a given point in time. These stands will then fill age class distribution gaps or quickly produce trees of a specified dimension. After the response to fertilization has ceased, trees will rejoin the growth curve at a point that is commensurate with their developmental stage, or size, rather than with their chronological age. "Consequently, although the response has ceased, fertilized trees have seemingly jumped several years ahead and thus will be growing either faster or slower than unfertilized trees" (Bfockley, personal communication) 1 . Fertilization decision variables include timing of application, dosage, type, form and method of application. The results of fertilizer applications are extremely variable over time, and possibly include interactions with other stand treatments (Brix, 1981; 1981; 1982; Brix and Mitchell, 1983; Stegemoeller and Chappell, 1990; Thomson and Barclay, 1984; Velazquez-Martinez etal, 1992). These results also depend on the conditions at the time of fertilization (Carter, 1989; Carter and Scagel, 1989; Darling and Omule, 1989), as well as on the possible differences associated with 1 Brockley, R. Research Scientist. BC Ministry of Forests. From notes provided in Forestry 306. 28 the form of nitrogen used, the method employed, and the time of the application (Barclay and Brix, 1984; Grier etal, 1984; Miller, 1979; Miller and Harrington, 1979). Fertilization is important early in the rotation. The greatest demands on the available nutrient capital of the site occurs before canopy closure, when the stand is about 15 years old (Klinka, 1981; Weetman1). Nitrogen, in the form of urea pellets (225-445 N kg/ha), is the most commonly applied fertilizer. The response to fertilization is evaluated relative to site index: the absolute highest response is in medium sites; an intermediate response is found on low sites; the lowest response is on high sites. However, the highest reponse in relative terms is on low sites (Daniels et al, 1979). "In Douglas-fir stands which are prone to thinning shock and which are nitrogen deficient, it appears that shock can be reduced or eliminated by fertilization" (Harrington and Reukema, 1975). 2.2.8 Pruning Pruning removes the branches from the bole of a tree at an early age in order to have knot-free wood accumulate on the bole. The current literature on pruning expresses a common theme: pruning is the only way to ensure the growth of significant volumes of clear wood in intensively managed stands. Given the short harvest cycles (less than 100 years) used in current management plans, and the limb retention characteristics of Douglas-fir, the growth of clear wood in unpruned second-growth stands will be minimal (Smith, 1960; Smith and Kennedy, 1983; Warrack, 1948). Pruning has little effect on tree growth if less than one-third of the live crown is removed (Fujimori, 1975; Staebler, 1963; Stein, 1955). "One of the benefits from pruning in the live crown is that the amount of juvenile wood in the bole may be reduced. Indirectly, pruning may also reduce tree taper" (Briggs and Smith, 1986). 1 Weetman, G.F. Professor of Silviculture. Faculty of Forestry, The University of British Columbia. From lecture in Forestry 306. 29 The most important decision variables for pruning are the age of the trees to be pruned and the height of pruning. Douglas-fir attains the required height and diameter for pruning between 12 and 25 years of age, depending on site quality. Other variables may be considered in a pruning investment. Pruning is sensitive to the market value of the clear wood produced and recovered (Briggs and Fight, 1992; Cahill etal, 1986, 1988; Fight etal, 1988), to the diameter of the trees (8-25 cm) being treated (Shaw and Staebler, 1950), to the number of trees being treated per hectare (Eversole, 1953), and to the growth rate per tree (Fight et al, 1987, 1988). "Investments in pruning would be most favorable on sites where the rate of growth of clear wood is the highest, as long as the rate of growth is not so rapid that it causes a decrease in quality" (Fight, 1987). The relatively high pruning cost per tree is presumably a reason why pruning has not been widely used in BC. The cost of pruning varies with the height of pruning (3-12 m), with the number of trees to be pruned (75-350 sph), with the size and hardness of the branches, and with the skill of the labour (Eversole, 1953; Finnis, 1953; Hedin, 1982; Nawitka Resource Consultants Ltd., 1992; Omule, 1994; Staebler, 1964). The results of studies on methods, time, and on costs of pruning young Douglas-fir are presented in Hedin (1982), Omule et al. (1994), and Warrack (1948). "The careful selection of trees to be pruned can make the difference between profit and loss on the pruning investment, especially in stands where no thinning is contemplated" (Eversole, 1953). "When combined with thinning, it will keep the pruned trees growing at a rapid rate. It will bring a maximum number of pruned trees to harvest age and size. This is a distinct advantage because the cost of pruning trees that drop out of the stand must be charged against the pruned trees that are harvested. If thinning is not possible, pruning should be concentrated in trees destined to retain a dominant position in the stand" (Warrack, 1948). 30 Reukema and Smith (1987) strongly recommend the pruning of widely spaced trees (500 sph) to enhance lower bole quality. "Since even in close espacement pruning will be necessary in order to produce clear lumber in 80-100 year rotations, it may be best to keep planting costs low by spacing trees as far apart as 3 m, with the expectation of pruning at an early age" (Eversole, 1955). The reviewed studies pointed out that pruning is very sensitive to various factors such as cost and methods, market value of clear wood and growth rate. Despite the above-mentioned factors, the practice of pruning appears to be necessary to produce clear wood in managed second-growth and planted stands of Douglas-fir. 2.3 COMPUTER SIMULATED STUDIES OF SILVICULTURAL MANAGEMENT ALTERNATIVES The following review illustrates the different approaches that researchers have used in order to address the three components of this study: Douglas-fir silvicultural stand management, growth and yield simulation models, and economic criteria. 2.3.1 Background and recommendations for juvenile spacing standards in BC (Ministry of Forests, 1994) This Ministry of Forests report is an updated version of an earlier report on the same project (Ministry of Forests, 1992). The 1994 report, based on a stand level analysis, summarized the effects of spacing decisions on yield and financial return. The report also discussed other factors affecting spacing decisions such as forest health, annual allowable cut shortfalls, jobs and community stability as well as biodiversity. The species directly addressed included coastal Douglas-fir, coastal western hemlock, interior Douglas-fir (Pseudotsuga menziesii var. glaucd), interior western hemlock, interior lodgepole pine and interior spruce (Picea glauca x engelmannii). All other species were equated to one of the principal species listed above. TASS 31 was used to produce yield tables. TIPSY was used in yield analysis whenever possible. Three different site indices were used for each species, and an operational adjustment factor was applied in order to reduce yields by 15% to account for stocking voids. Discounted net revenues (DNR) was the approach used to evaluate the economic potential of juvenile spacing. The discount rate was 4%. Discounting assumed that the time of spacing was the base year. The DNR of the spaced stand must exceed the DNR of the untreated stand by an amount that exceeds the anticipated cost of spacing. If the additional value created by treatment exceeded the cost of treatment, the stand was considered to be treatable and was assigned a benefit/cost ratio (B/C) in order to establish its priority for treatment. This report described how stand density affects yield, stem diameter, rotation, and wood quality. The conclusions were that spacing opportunities appear more attractive on the coast than in the interior due to the fact that larger diameter logs were produced in a shorter period of time. Spacing was financially feasible for most sites and for most establishment densities on the coast, while in the interior spacing might be cost-constrained on poor sites. The analysis shows clearly and exhaustively the benefits and feasibility of spacing. The report also reveals the range of densities that optimize financial return both on the coast and in the interior. However, research suggests that juvenile spacing could substantially increase timber quality and value when combined with commercial thinning. Although juvenile spacing proves to be efficient in itself, it may prove to be a more effective management tool and a more financially attractive investment opportunity when used in conjuction with other silvicultural treatments. This report evaluates juvenile spacing as a short term tool only. This limited approach fails to encompass the economic evaluation of the application of juvenile spacing as one component of a whole stand plan. 32 2.3.2 Incremental silviculture, pruning financial analysis (Nawitka Resource Consultants, 1992) This project analyzed pruning from a financial point of view. For the purposes of this report, pruning was considered for the major BC coastal species (Douglas-fir, hemlock, western redcedar (Thuja plicatd) and cypress (Chamaecyparus nootkatensis)) in both spaced and unspaced stands. The spaced stands were based on the best spacing options indicated in the report cited in Section 2.3.1. The unspaced stands were based on data provided by the simulation model TASS, with particular reference given to the statistics for the 250 crop trees. The cost of the first lift was assumed to be $2.00 per stem, and the cost of the second lift was $4.00 per stem. In the spaced stands, two pruning intensities were applied: pruned stems were either 500 or 400 in the first lift, and 450 or 350 in the second lift. In the natural stands, the best selected trees were 125 or 250; these trees had to be easily recognizable and treatable. The financial analysis used the DNR criteria at a 4% discount rate. The clear wood value was established by using a Lotus spreadsheet model. If the value attributable to one lift log exceeds the cost of one lift, the treatment was deemed feasible. The same procedure was used to evaluate a two-lift option. The investigation indicated that pruning appears to be feasible on the coast for various spaced and natural stands. The estimates of the costs of pruning seem to be too optimistic for the large number of trees pruned in the spaced stands. Selection and marking of trees for pruning requires skilled labour and consumes time which, given the number of trees indicated above, would escalate the costs. There is no evidence that stems over 30 cm up to 45 cm dbh (on medium to rich medium sites) at harvest age 70-80 are likely to produce the amount of clear wood that makes the treatment feasible. On average, approximately 100-175 trees would have attained diameter greater than 45 cm at age 80 on site index 30 m and given an initial density of 1200 sph (XENO's stand table 33 output produced for this study); this number of trees represents less than half of the pruned trees (350-500). The literature on pruning (Eversole, 1953; Fight et al, Hedin, 1982; Reeb, 1984) suggests selecting 75-350 stems for pruning. This report contributes to the financial analysis of pruning. The analysis shows some weaknesses because of the factors selected for the evaluation. The critical factors are the number of trees selected for pruning, the costs of pruning, and the amount of clear wood available at harvest age. 2.3.3 Financial analysis for selected and combined silvicultural treatments for the BC coast (Nawitka Resource Consultants, 1993) Three coastal regimes (natural Douglas-fir at 4444 sph, planted Douglas-fir at 1200 sph, and natural hemlock at 5000 sph) were used as the basis for this financial analysis. The silvicultural treatments were spacing, fertilizing and pruning, and some combinations of these treatments. Only timing and a small number of treatment intensities were considered in order to limit the number of the combinations. However, in the case of fertilization, more than one application was considered, and the timing was varied from early in the rotation to late in the rotation. Fertilizer effect estimates were compiled for the regimes of interest using the DFSIM model. These estimates were then reduced to reflect operational merchantable volumes in line with the TASS model estimates. The TASS model was primarily used to simulate the selected treatments. The economic analysis measured the feasibility of the management alternatives by using DNR at a 4% discount rate. A B/C ratio was used to indicate priority for treatment. The report showed that fertilization of Douglas-fir planted at 1200 sph was financially feasible. An unspaced stand initially planted at 1200 sph gives positive results for fertilizing or pruning 250 stems. The pruning of 450 stems up to a second lift in a spaced stand on a medium site was not feasible; however, pruning became marginally feasible if the stand had been fertilized. 34 Fertilization was particularly attractive on medium sites in many situations, and generally complemented previous treatments. The study indicated that the use of regional costs was a shortcoming in evaluating the B/C ratio. The report was not designed to consider what combination of treatments was the best for the stand over time. The basic purpose of the report was to explore the effect of an additional treatment (fertilization) on spacing and/or pruning; thus, the net revenues attributable to any treatment are discounted to a base year which coincides with year 10 of the stand. This report is unsatisfactory due to the fact that one of the limitations of DFSIM is that its database contains few stands treated with several applications of fertilizer. The study did not assess the reasons for preferring the TASS or DFSIM models to project either fertilization or other treatments. Fertilization projections were not compared or analyzed. Another limitation is that the report did not examine a complete prescription. Thus, the economic approach could not discover the best combination of treatments that provide the stand with the most efficient and financially attractive silvicultural regime. 2.3.4 An economic evaluation of commercial thinning Douglas-fir in the coastal region of British Columbia (Ministry of Forests, 1993) The purpose of this report was to conduct an economic evaluation of the commercial thinning of Douglas-fir in the coastal region of BC. Three regimes of planted stands (2500 sph, 1975 sph, and 1600 sph) were assumed for site indices 24 m, 30 m and 36 m, respectively. This report included the estimation of the effects of timing and of various intensities of thinning on the final harvest; the time of thinning was considered to be the base year for discounting of costs and revenues. TASS was used to simulate the growth and yield as a result of the commercial thinnings. Net present worth (NPW) was the economic criterion used to determine the full impact of one commercial thinning per rotation of an even-aged stand. The discount rate used was 4%. 35 This simulation led to the conclusion that commercial thinning would likely provide only a marginal increase in the cumulative merchantable volume available from a stand over a rotation. The economic evaluation showed that thinning in the range of 200 to 400 leave trees per hectare proved to be economically superior to the corresponding unthinned regime for all simulation runs. The optimum thinning regimes might potentially increase the net present value of the simulated stands by 13-17% at site index 36 m, by 10-18% at site index 30 m, and by 4-6% at site index 24 m. The report concluded that the economic gain from commercial thinning is modest, but potentially still worth pursuing. This report showed the viability of commercial thinning for unspaced and juvenile spaced stands, when the time of commercial thinning is considered as the base year. In order to measure the viability of commercial thinning over the span of a rotation, base year 0 has to be considered as the time of the establishment of the planted stand. Placing the base year at a different point in time would affect the evaluation over the rotation. The use of commercial thinning as the base year assumed that all the events that took place before commercial thinning were not taken into account. This assumption fails to evaluate commercial thinning as a component of an entire silvicultural regime. This report shows the economic efficiency of commercial thinning for three density regimes, each of which was assigned to a particular site index. An oversight in this research was that the same range of planted densities was not explored on each site index chosen. The viability of thinning as a commercial practice may depend on the number of trees at the time of establishment. The literature on thinning recommends the practice on stands planted at densities in the range of 816-1100 sph; these densities should be examined in order to demonstrate opportunity for investment. Another oversight, which the researcher acknowledged, was the need 36 to explore the assessment of commercial thinning when used in conjunction with other silvicultural treatments. 2.3.5 PROGNOSIS, SPS and TASS yield comparisons for even-aged interior Douglas-fir (Goudie, 1987) The objectives of this paper were: (1) to review the basic categories of individual tree forest growth models; and (2) to compare the mathematical structure and to graphically contrast the output of three representative systems, using sample simulations of managed interior Douglas-fir plantations. The experiment was designed to evaluate the performance of the PROGNOSIS, SPS, and TASS models on managed plantations of pure interior Douglas-fir. The base regime established a stand with 1236 trees/ha at age 0, unmanaged for at least 140 years. Option 2 investigated PCT of Option 1 (base regime) to 618 trees/ha at age 20. Options 3 through 5 analyzed commercial thinning of Option 2 (PCT+CT) to 247 trees/ha at age 50. Trees were thinned from below, systematically (d/D=l), and from above, for Options 3 through 5, respectively. The five options were repeated on site indices 15.2 m, 21.3m and 27.4 m. The simulations in this study compared the thinning response between the models by focusing on site index 27.4 m. Options 4 and 5 were not included in the report because they were not practised as often as thinning from below (Option 3). This paper examined the basic individual tree growth models designed for use in the region, as well as their underlying assumptions and structure. The study showed that there are considerable differences between projection systems. Most of the differences were related to the assumptions and structure built into the model and to the database used to develop the relationships. This report is included in this literature review even though it involved interior Douglas-fir, because it is one of the few studies other than Marshall's (1988, 1989, 1991) which compares growth and yield models used in the PNW region. 37 This report contributes to the understanding of growth and yield simulation models; the objectives are successful in categorizing and comparing the structure of the projection systems. Despite this reasonable representation and use of the growth and yield projection systems, the report limited the comparison of the output to the analysis of one regime and two treatments in the interior region of BC. This leaves some uncertainty about the use and the projection estimates of the models when simulating a range of densities for a range of silvicultural prescriptions. The small number of silvicultural options in the report also constrains the availability of growth and yield estimates for use in the analysis of the most profitable management plan. 2.3.6 Three-dimensional representation of Douglas-fir volume growth: comparison of growth and yield models with stand data (O'Hara and Oliver, 1988) Growth and yield estimates for unthinned stands from DFSIM and TASS were used to construct graphic three-dimensional representations of Douglas-fir stand growth for site index 44 m. The three-dimensional models used three variables: trees per hectare, breast height age, and either mean tree volume or stand volume. The objective was to assess the potential of Douglas-fir growth and yield at very wide spacings which resulted from initial establishment, or from thinning of narrowly spaced stands. The paper proceeded in two steps. Firstly, estimates from two growth and yield models were compared at wide spacings. Secondly, results from a carefully planned thinning trial, where trees were thinned to wide spacings, were compared to the growth and yield model estimates for unthinned stands at wide spacings. The study showed that TASS and DFSIM were in agreement over most of their common range of age and number of trees. However, the volumes predicted by the DFSIM model exceeded those predicted by the TASS model by as much as 25% at wider spacings and older ages. Comparisons of these three-dimensional models to unthinned and thinned stand data for a similar site quality found the models to be reasonably accurate representations of unthinned stand 38 growth. However, the thinned stands had greater mean tree and stand volumes than those indicated by the TASS model for unthinned stands at similar spacings. Complete comparisons were not possible with DFSIM because it has limitations on the number of trees that could be considered for treatments. The authors suggested that TASS and, to a lesser extent, DFSIM, may be underestimating the growth of widely spaced stands, or that thinning may actually increase the growth of thinned trees over that of trees which have always been grown at a post-thinning spacing. This report is a useful contribution to model prediction capabilities. The study investigated how simulation models compared in the prediction of a particular treatment. One limitation of this report is the small number of silvicultural options used in the comparison. The results suggest that the projections of silvicultural treatments, either individually or in combination, is necessary in order to know how models compare. The study also suggests that the great effort exercised in pursuing, collecting, and analyzing basic data has to continue uninterruptedly. Model development and prediction capabilities have to be intensified. 2.3.7 Summary On both natural and planted stands, forest management can be assisted by the use of various silvicultural treatments or prescriptions. Silvicultural treatments result in either qualitative or quantitative productivity gains. These short term tools can also produce effective results in the long run. On the other hand, a prescription consists of a sequence of scheduled activities. These activities (cultural treatments, harvests, or other events) are implemented over the rotation in order to achieve a desired outcome. The reviewed simulation-based studies overlook the use of prescriptions as a silvicultural stand management alternative. All of the reviewed studies 39 addressed alternative silvicultural treatments either individually or in combination, across various sites and for various species. The experimental findings suggested that coastal Douglas-fir has a high potential for silvicultural management in southwest BC. However, the cited studies did not exclusively represent Douglas-fir. Due to the commercial importance of coastal Douglas-fir, foresters and researchers need to thoroughly investigate the silvicultural management alternatives in the BC coastal region. The enormous amount of available data from field trials and reports provides the foundation to substantiate the need for Douglas-fir resource management. These data ensure attainable plantation management success on the coast. Although the research data contribute and confirm the practice of silviculture on Douglas-fir, foresters and planners have to be able to combine the cultural practices that will produce the most desired outcome for a stand. Researchers and foresters utilize growth and yield projection systems in order to predict the effects of treatment. Despite this, none of the studies discussed the reasons for preferring a particular simulation model; all utilized arbitrarily selected models. Consequently, none of the studies attempted to assess which model better suits the estimation of growth and yield. Most of the studies utilized TASS to predict silvicultural response. A few of the studies also used DFSIM, TIPSY, SPS, and XENO. According to Goudie (1987), there are differences in the prediction of growth and yield estimates between the simulation models used extensively in BC. O'Hara and Oliver (1988) also draw attention to the differences in model estimation. In essence, the cited simulation-based studies do not assess the output differences between models. As a result, users do not know if models are in agreement when simulating a management alternative; they do not know if the theoretical assumptions used to design the model are in agreement between models; they do not know how model estimates compare to one another. The user is not able to select the most suitable model for his/her management plan. This 40 arbitrary selection of a model may affect subsequent management decisions. Predicted estimates are a necessary component in the economic evaluation of the management plan and, thus, predicted estimates become as relevant in physical terms as in economic terms. These simulation-based studies also showed the results of silvicultural investments. The simulation-based estimates were used to determine the feasibility of the treatment and the opportunity for investment. DNR, PNW and B/C were the economic criteria used to analyze feasibility of treatment, net benefits and priority in the undertaking of a project, respectively. These criteria were used primarily to determine the wisest use of a limited budget in the short run. The cited studies do not assess the economic evaluation of a silvicultural prescription; therefore, they overlook the economic analysis of a sequence of scheduled activities over a rotation. The economic analysis of prescriptions can determine the most efficient and the most financially attractive combination of activities in obtaining the desired outcome of a stand. In summary, these reviewed simulation-based studies each omit some important factors: they do not incorporate the study of simulated prescriptions as another option in silvicultural management; they do not particularly address the silvicultural management of planted coastal Douglas-fir stands; they do not deliberately utilize selected models; they do not compare projections estimates for a range of entire silvicultural prescriptions; they do not assess the economic benefits of a simulated silvicultural prescription. The above-mentioned factors were identified as the rationale for this comprehensive study. 41 C H A P T E R 3 D A T A A N D M E T H O D S The reviewed literature revealed significant deficiencies in three areas: comparison of the output of growth and yield simulation models, simulation of entire silvicultural prescriptions and the economic analysis of the silvicultural prescriptions. The objective of this study was to address these deficiencies, at least partially, by simulating a range of silvicultural prescriptions for planted coastal Douglas-fir using four selected models, and by comparing projection estimates for various sites. In addition, a subset of the most promising prescriptions was included in the economic analysis in order to determine the most advantageous silvicultural regime. First, the silvicultural prescriptions used in the study are specified from findings in the literature review on individual treatments. Next, the start-up conditions used to run the models are described. Lastly, the criterion and assumptions used in the economic evaluation of the projected estimates are stated. The simulation of the prescriptions provided the estimates needed for comparing the output of the models and for the economic analysis. The analysis of the output of the models may allow foresters to choose the model best suited for particular regimes. The economic analysis of the projected growth and yield estimates may allow foresters to choose the most financially attractive prescriptions, as well as the most profitable range of densities among the prescriptions. 42 3 . 1 P R E S C R I P T I O N S U S E D Table 2 outlines the range of prescriptions used to establish and manage Douglas-fir plantations for this study. These prescriptions include all the cultural practices examined in the literature review and are based on the decision variables investigated in the literature review. The decision variables selected for PCT and CT were age (timing) and residual trees/number of cut trees (intensity). The decision variables selected for fertilization were age (timing), type and dose (dosage). The type of fertilizer (nitrogen in the form of urea pellets) is the same across the prescriptions; therefore, the type of fertilizer is not indicated in Table 2. The values for each decision variable were determined assuming average conditions. The total set of 2 1 prescriptions, covering a wide range of initial densities, was examined in order to identify the range of densities that best responds to treatment either individually or in combination. The decision variable values for each prescription were the same for all sites. Although pruning is listed as another treatment in Table 2, this practice is not included in a simulation run. Pruning scheduling shows a 1st and a 2 n d lift. The (age) timing, number of trees to be pruned (intensity), and the number of lifts (frequency) were the decision variables considered. The intensity is the same across prescriptions and, thus, the number of trees to be pruned is not indicated in Table 2. Details on pruning assumptions for the volume calculations and the economic evaluation are given in Section 3 . 3 . 7 . 43 Table 2. Outline of prescriptions Prescription Initial PCT CT Pruning Fertilization # Density Age Residual Age Residual 1st lift 2nd lift Age Dose trees trees Age Age (sph) (yr) (sph) (yr) (sph) (yr) (yr) (yr) (kg/ha) 1 2500 15 1100 2 1600 15 816 3 2500 15 1100 50 4 1600 15 816 50 5 1100 6 816 7 625 8 500 9 400 10 2500 15 1100 11 1600 15 816 12 816 15 400 13 1100 40 14 1100 50 15 1100 60 16 816 40 17 816 50 18 816 60 19 625 40 20 625 50 21 625 60 15 20 15 20 300 15 20 300 15 20 15 20 15 20 15 20 15 20 15 20 15 20 15 220 15 20 15 220 15 20 15 220 300 15 20 300 15 20 300 15 20 300 15 20 300 15 20 300. 15 20 300 15 20 300 15 20 300 15 20 44 3.2 START-UP CONDITIONS F O R T H E F O U R M O D E L S C H O S E N The models selected to carry out the study are DFSIM, SPS, XENO and the TIPSY program. The selection was based on their requirements for operation. The four models chosen can project the alternative silvicultural management options that are of potential interest in the region. Each of the models allows the user to enter site index, initial planting density, timing, intensity and type of treatment, except for the TIPSY program. TIPSY does not allow the user to change the timing decision variable for PCT. TIPSY assumes that, on the coast, PCT occurs when the stand is 6 m high. Except for TIPSY, all of the models can simulate precommercial thinning, as scheduled in Table 2. Not every model permits the user to get started from bare land. Therefore, the simulation of every prescription considered that all the planted trees were successfully established. Gains from genetically improved seedlings were not incorporated in the models, and could not be projected. Ingrowth also was not incorporated in the simulation. Fertilizer was applied only once, even if the models can project more than one application. Simulations where more than one application is applied are not substantially supported by the models because of the lack of treated plots in the database. Furthermore, the literature on fertilizer application results in inconclusive findings when more than one application takes place. CT was simulated as "thinning from below" which is the type of thinning that has been rigourously tested. The prescriptions were simulated up to age 100. There were two reasons to assume 100 years as the upper age for the comparison of the models and for the analysis of their projected estimates. One of these reasons was that DFSIM has extrapolation trouble past that age; the second reason was that plantations are expected to reach rotation age earlier than age 100. Growth rate, mean diameter and total standing volume were the variables chosen to compare model output. These variables can express the change in tree or stand growth due to the 45 application of silvicultural practices. For each prescription, the response to the sequence of scheduled activities was measured by growth rate, mean diameter, and total standing volume over the rotation in 10-year steps, for various sites. Every prescription was simulated for site index 20 m (poor), 30 m (medium) and 40 m (high). The comparison among the models was based on the grounds that all the models are designed for the same geographic area, and for sites with similar characteristics. A summary of the assumptions needed to run a simulation is given in Table 3. Table 3. Model simulations: summary of assumptions Simulation run Input label Site index Species Seedlings Trees Ingrowth Data entry for activities Pruning Simulation period Input data 20 m; 30 m; 40 m coastal Douglas-fir Not genetically improved Successfully established No Values from Table 2 No 100 years 46 3.3 ECONOMIC EVALUATION The economic evaluation of a prescription represents the evaluation of an entire regime. A prescription is composed of a sequence of scheduled activities. The sequence of activities is prescribed in order to provide the stand with the proper management to be able to attain a desired outcome. Some of these activities will generate a revenue; others will benefit the development of the stand in the long term, contributing indirectly to financial returns. The economic evaluation is then undertaken in order to ascertain if the outcome of the stand produced an acceptable return on the investment. The revenues gained by a prescription are dependent on the prediction of the volume yielded at any point in time during the timber-growing cycle. 3.3.1 Data requirements for the economic evaluation of the prescriptions In order to evaluate the prescriptions from an economic viewpoint, the unit area of a forest stand needs to be determined. Then, the volume of the unit area needs to be predicted. The benefits or disadvantages of a prescription are measured more precisely if the volume of the unit area can be determined and valued in terms of individual logs, rather than in terms of averages. Therefore, the economic evaluation of the prescriptions in Table 2 required the calculation of log volumes, and the determination of log values. Log grades, costs, and prices were the elements used to determine the value of a log. The following sections describe all the above-mentioned requirements for the economic evaluation of a prescription, as well as the economic criterion and the assumptions. Log volume calculations Log volumes are a relevant factor in the information that allows foresters to accurately value timber harvests. XENO's stand table output was chosen to determine log volumes. This output gives a breakdown by diameter class and by species, including number of trees/ha, 47 volume/ha, and volume per tree. The other models' stand table output do not provide the same amount of detail by diameter class and by species. XENO's diameter inside bark and volume formulas follow: dib = (dbh*0.9)*(H/(H-1.3)) h=0 dib = [((H-X/H)A ((1 + 0.8 (%CRN))/2)] dib h=x 100 h=o vol. = [(dib )A2] (H-X) id 40.000/(2 + 0.8 %CRN) h=x h=x 100 where: dib = diameter inside bark (cm) dib = diameter inside bark at a height of x meters (cm) h=x dbh = diameter breast height (cm) vol. = tree volume above x meters (m3) h=x CRN= % live crown H = total height (m) h = height above x meters (m) x = 5 meters height increments (log length) Log volume calculations were determined for each 5 m log length. The merchantable volume was based on a minimum dbh of 12.5 cm, a stump height of 0.3 m and a minimum top diameter outside bark of 10 cm. The Silviculture Manual (Ministry of Forests, 1995) recommends that silvicultural treatments be applied on good and medium sites. Good sites 48 represent 4% and medium sites represent 38% of the coastal region (Ministry of Forests, 1980a). Given the representative percentage of medium site for the coastal region, and the number of calculations needed in order to determine log volumes for the range of prescriptions described in Table 2, log volumes were determined on site index 30 m for diameter classes 0 to 60 cm, in 5 cm increments. Gross log values Log grades Log size, and other characteristics associated with form, health, and technological requirements make a log grade. Log grades are needed in order to determine the value of a log. In order to obtain grades, coastal Douglas-fir log requirements were followed as indicated in Code 6.5.9 Fir and Pine Grades (Ministry of Forests, 1995). A brief outline of all the requirements taken into account for Douglas-fir log grades is given in Table 4. The grade of a log is essentially determined by the product that can be made out of a log. Therefore, characteristics such as the number of rings (rings/cm), maximum size of knots (cm), % of merchantable lumber, % of clear wood, and log diameter are obviously relevant since they define the features to be observed when the scaling procedure is in place. Descriptive terms such as "merchantable" and "clear" are worth defining in order to clarify the meaning used in the Scaling Manual for BC (Ministry of Forests, 1995). "Clear" means free of knot or stain. "Merchantable" describes the percentage of lumber out of the gross scale that will be merchantable or better. "Gross scale" refers to the volume of a log inside bark, and includes unsound wood and holes in the log. 49 u PQ u a "« s e cs T3 eS e» •a es u W) S • mm a. e ce i_ C o cs E E s Tf 3 H J | X* CS cs .a " J M l en c C B 44 0 s 1_ e . S • s -J 3 •S'EI u u WD ' © -J B J _ o -J s •o o CS u o oo •4 o I 00 I TT O in O CI in r<"s in o m o in m r- m o m m o o m >n 00 c<1 o ON m 2 £ 2 J-00 ^ 00 I 00 ^ ON m o in 1 3 O in e 3 h-1 o in 60 o oo 60 O in © in 60 O 1 C/3 o in oo O Q. U X Log prices and costs Log values were determined, depending on the grade, using prices per volume unit (net of logging costs) from the Vancouver Log Market (VLM) (1994). It was assumed that the VLM accurately reflects timber values for the coastal region. Because of the restriction in use imposed by the local market for peeler logs, Grades B and C (peeler logs) were not used in the economic evaluation of a prescription. Logs with top diameters over 38 cm were valued according to Grade H price. Top diameters from 14-38 cm were valued at Grade J price, and top diameters from 10-14 cm were valued at Grade U price. The remainder of the logs were valued at Grade Y price (chips). Log costs were assumed to be included in the costs of thinning and final harvest. The costs used in this study are taken from various sources such as private firms (Fletcher Challenge Ltd.) and government agencies (Campbell River District, Ministry of Forests; Silviculture Branch, Victoria, Ministry of Forests). 3.3.2 Assumptions used in the economic evaluation of a prescription A prescription is comprised of the costs and returns stemming from the following sequence of scheduled treatments." site preparation, vegetation control, espacement (initial planting density), fertilization, precommercial thinning (spacing), commercial thinning and pruning. Site preparation, vegetation control, and planting contribute to the costs of the prescription. The seedlings were assumed to be planted without delay after site preparation. The gains obtained by utilizing high-quality seedlings were not expressed in this evaluation because of the lack of estimates provided by the simulation runs. The trees were assumed to be successfully established. The control of competing vegetation took place three years after planting. The application of fertilizer was not included in the economic evaluation in this study as a result of uncertainty in the predicted estimates. Spacing was not a profitable treatment in the sense that it 51 did not itself generate an income. Spacing was prescribed in order to render healthier, faster and more appropriate development of the stand. The economics of the forest may be altered with time; the harvest of second-growth stands will change the availability of the range of diameter classes that old-growth stands traditionally supplied. The spacing treatment may then constitute another opportunity to generate profits if logs of smaller size have a market. Thus, if spacing generates a revenue, the number or range of prescriptions representing an opportunity for investment may be modified. Commercial thinning requires the construction of a road system prior to final cut. This differs from what a crop requires when there are no intermediate harvests, and the road construction is concurrent with the final harvest. It was assumed that commercial thinning costs per hectare included road construction. Therefore, final harvest costs vary if the prescription does or does not include CT. Furthermore, it was assumed that the only benefit of the stand was commercial timber and the entire crop was clearcut when it reached harvesting age. After harvesting this one crop, the land was considered to be valueless, no taxes were owed, and all costs and prices remained constant during the growing period. All revenues and costs were measured in constant dollars in February 1994. The unit area of the stand was assumed to be an hectare. A summary of the assumptions described in this section is shown in Table 5. 3.3.3 The economic evaluation criteria Forest management involves planning a sequence of costs and benefits spread over time. Evaluation involves weighing the benefits against the costs and taking into account the timing of their occurrence in order to establish their relative efficiency. The most efficient management plan will generate the greatest possible benefits for the costs incurred. One of the most critical economic questions in forestry is the age at which trees should be harvested (i.e., the rotation 52 Table 5. Economic evaluation of a prescription: summary of assumptions Economic evaluation Economic variable Unit Land area ha Volume m3 Harvest system Period of evaluation Interest rate 4% Taxes Costs $/ha Prices $/m3 Grades Revenues $/ha Assumptions . Description Source valueless clearcut Feb. 1994 No constant various constant V L M H,J,U and Y BC Scaling Manual timber age). The concern is to find the age that will yield maximum economic returns, taking into account how the volume and technical characteristics of the forest change with age, and how these aspects are reflected in its economic value. Foresters have developed a variety of criteria for selecting the age of harvest for forest stands. These criteria can be classified into two broad groups: physical and economic. In this study, the rotation age is determined using NPW, an economic criterion. 53 3.3.4 Net present worth NPW was the criterion used for assessing the economic efficiency of the prescriptions outlined in Table 2. NPW is the net present value of the revenues and costs associated with a crop over a rotation. This method requires the calculation of revenues and costs that occur at different points in time. In order to compare these values, they have to be converted to their equivalent values at the same point in time. In order to find the present value of future revenues and costs, a discount rate is used. The discounted costs subtracted from the discounted revenues generate the net present worth, or net benefit, of the project. Net benefit is the appropriate basis for the selection of mutually exclusive uses of some fixed resource. Alternative feasible investment opportunities can be compared and ranked according to the magnitude of the net benefit they are expected to generate. Quick Silver (QS) was the program used to run the financial analysis (Vasievich etal, 1984). 3.3.5 Quick Silver QS is an interactive financial analysis program used to evaluate forest management investments. This program was developed by Dr. J.M. Vasievich and Ronald Frebis (1984) of the USDA Forest Service, Southeastern Forest Experiment Station. QS performs analysis on investment 'cases'. Each case describes a series of management activities carried out over a period of time 'the investment period'. The costs or revenues associated with each activity and their occurrence within the investment period are defined as 'transactions'. Based on these transactions, the program computes the series of annual and/or periodic incomes and expenses that will occur between the first and last year of the investment period. The 'cash flow' is used to calculate measures of financial return and generate reports. NPW, generated by QS, was calculated over one rotation for a range of ages in ten-year periods. 54 3.3.6 Interest rate The interest rate indicates a preference for present over future consumption. The interest rate also measures the opportunity cost of capital over time. In order to measure the cost of capital of public investments, the interest rate or discount rate can take two approaches: the discount rate, or the social opportunity cost of capital (SOC), and the social rate of time preference (SRTP). For silvicultural investments, SOC is generally preferred over SRTP. SOC can be measured objectively; it has an explicit recognition of trade-offs between private and public sector opportunities. Silvicultural investments are designed to produce mainly timber, which is a market good. SOC is calculated on the money transferred from the private sector to the public sector in taxation and bonus. SRTP reflects the preferences of society for present over future consumption. Individuals considered solely may have different behaviour from individuals considered collectively. SRTP bears no relationship at all with the rates that individual savers and investors, acting independently, generate in the capital market. To the extent that SRTP differs from the market rate of interest, it must be decided collectively through political processes. SRTP is difficult to measure and, despite extensive economic and philosophical enquiry, it remains elusive. A real discount rate of 4% for evaluating silvicultural projects in the public sector is recommended in BC. Cheap and Pratts (1989) recommend a real rate of return of 3-5% p.a. NPW, generated by QS, was performed using a discount rate of 4% over one rotation for a range of ages in ten-year periods. In addition, NPW was tested at 3 and 5% interest rates in order to examine responses to changes in the interest rate. 3.3.7 Pruning: volume calculation and economic evaluation DFSIM was the only selected model that has an extension model, named PRUNE, to simulate pruning. The PRUNE model was not used in this study because it would not be 55 possible to fulfill the objective of comparing models. However, pruning is another treatment allowed for within the prescriptions in Table 2. Pruning can be economically justified when the accumulated costs of the treatment are predicted to be less than the differential in value between pruned and nonpruned logs at the time of harvest. The financial return from pruning comes from the increase in value of clear wood that will be produced from pruned trees. The crucial economic factor is whether the increase in value from additional clear wood justifies the investment in pruning. Only the bottom log is commonly pruned, since this is the highest-value log in the tree, and because the costs of pruning rise dramatically with height. The butt log volume calculation follows the criteria used to determine log volumes as indicated in Section "Log volume calculations" in this chapter. Diameter classes ranging from 45 cm to 60 cm produced the butt logs used in the calculations. The butt log volumes calculated for every pruned stem were added to obtain a per hectare volume. Pruning was assumed to have no impact on growth, so the volume with and without pruning was the same. The number of trees to be pruned was 150 sph. It was considered that the 150 sph were correctly selected and that they continued to be the healthiest and most vigourous trees until the time of harvest. In order to complete the evaluation of pruning in physical terms, two lifts were considered. The first lift was 3 m high and was carried out at the time of spacing (15 years); the second lift up to 6 m when the trees were 20 years old. For the economic evaluation, it was assumed that no differences in costs for stand management and logging occurred after pruning. The pruned trees were assumed not to be damaged after logging when commercial thinning was applied. The harvest age for pruning was represented by the optimum rotation age that was calculated and determined in the economic evaluation of the prescriptions in Table 2. The costs of the two lifts, measured in constant dollars in February 1994, were $3/tree and $5/tree, respectively (Fletcher Challenge Ltd.). 56 NPW was used as the evaluation criterion; the value of the volume per hectare obtained from pruned logs at the time of harvest was discounted and its present value was compared to the present value of the discounted costs of pruning two lift logs. If the present value of two lift logs was equal to or greater than the present value of the costs of pruning them, then the treatment was deemed feasible. The discount rate used was 4%. A summary of the assumptions for the economic evaluation of pruning is given in Table 6. Table 6. Economic evaluation of pruning: summary of assumptions Economic evaluation of pruning Economic variable Assumptions Unit Description Source Land area ha Volume m3 Period of evaluation Feb. 1994 Interest rate 4% # of pruned trees 150 sph Costs $/tree 2 lift logs constant Fletcher Challenge L1 Prices $/m3 constant VLM Grades H BC Scaling Manual Revenues $/ha clear wood 57 CHAPTER 4 GROWTH AND YIELD ESTIMATES DFSIM, SPS, TIPSY, and XENO simulations provided the estimates for comparing the output of the models. The simulation runs followed the start-up conditions described in Chapter 3; the values in Table 2 provided the data necessary to complete a run. Every prescription was simulated for site index 20 m (poor), 30 m (medium) and 40 m (high). Growth rate, mean diameter and total standing volume (TSV) were the variables used to compare model output. These variables represent three important aspects of stand growth assisting in the assessment of the desired outcome of a stand. Analysis of the estimates of growth rate, TSV and mean diameter was carried out for each site in 10-year steps for ages 40-100. These estimates are summarized in Tables 7-9 for dbh, Tables 10-12 for maximum mean annual increment (MAI), and Tables 13-15 for TSV, 4.1 RESULTS 4.1.1 Site index 20 m Behaviour differed greatly among the models. No results were listed for several of the 21 prescriptions because of limitations within the models in simulating these prescriptions. In comparison to the other models, DFSIM projects dramatically bigger trees over the rotation. DFSLM and SPS provide the earliest estimates of age of the stand for a given dbh 58 Table 7. Mean dbh at site index 20 m Prescription # Mean dbh rem) Stand age (vr) DFSIM SPS TIPSY XENO 1 24 66 - 80 80 2 24 58 - 70 80 3 29 - - - 85 4 29 - - - 100 5 29 90 100 * * 6 29 80 80 100 * 7 29 70 70 80 100 8 29 62 66 70 100 9 29 58 63 66 90 10 24 56 - - 100 11 24 48 - - 100 12 24 @ - 70 13 29 - 58 - 90 14 29 - 58 - * 15 29 - 60 - * 16 29 - 56 - 85 17 29 - 58 - 95 18 29 - 60 - 100 19 29 - 58 - 85 20 29 - 58 - 96 21 29 - 59 - 90 - The model did not simulate the prescription. * The model predicts the corresponding estimates at some time later than the rotation age. @ The given diameter was attained some time earlier than age 40. 59 Table 8 . Mean dbh at site index 30 m Prescription # Mean dbh (cm) Stand age (yr) DFSIM SPS TIPSY XENO 1 45 * 100 * 100 2 45 93 80 * 95 3 45 64 75 - 80 4 45 66 64 - 86 5 40 90 85 * 100 6 40 85 75 100 95 7 40 76 63 73 86 8 45 80 70 80 95 9 45 70 63 70 86 10 45 93 70 - 96 11 45 82 65 - 86 12 45 54 48 - 74 13 45 60 68 - 90 14 45 68 78 - 93 15 45 75 80 - 97 16 45 58 65 - 87 17 45 68 70 - 90 18 45 78 75 - 95 19 45 - 65 - 80 20 45 65 65 - 87 21 45 77 63 - 90 - The model did not simulate the prescription. * The model predicts the corresponding estimates at some time later than the rotation age. 60 Table 9. Mean dbh at site index 40 m Prescription Mean Stand aee (yr) # dbh (cm) DFSIM SPS TIPSY XENO 1 45 76 63 66 60 2 45 68 52 62 58 3 45 59 56 - 56 4 45 59 49 - 54 5 45 76 63 70 66 6 45 70 56 68 61 7 45 66 50 58 58 8 45 60 46 50 57 9 45 56 41 46 54 10 45 70 50 - 60 11 45 69 44 - 59 12 45 54 @ - 50 13 45 54 50 - 56 14 45 60 52 - 57 15 45 68 56 - 59 16 45 56 46 - 56 17 45 60 49 - 58 18 45 68 53 - 57 19 45 53 44 - 56 20 45 62 46 - 56 21 45 68 50 56 - The model did not simulate the prescription. @ The given diameter was attained some time earlier than age 40. 61 Table 10. Maximum MAI at site index 20 m Prescription DFSIM SPS TIPSY XENO # Max. Stand Max. Stand Max. Stand Max. Stand M.A.I. age M.A.I. age M.A.I. age M.A.I, age (m3/yr) (yr) (m3/yr) (yr) (m3/yr» (yr) (m3/yr) (yr) 1 5.6 90 * * 2 5.8 90 - * * 3 - - - * 4 . . * 5 4.3 100 5.1 80 5.21 100 * 6 * 4.32 100 * * j * * * * g * * * * 9 * * * * 10 6.66 80 - - * 11 6.97 70 - - * 12 7.45 85 - - * 13 - 3.03 90 - * 14 - 2.92 90 - * 15 - * - * 16 - 3.02 90 - * 17 - 2.66 100 - * 18 - 2.9 80 - * 19 - 2.91 90 - * 20 - 2.93 90 - * 21 - 2.92 80 - * - The model did not simulate the prescription. * The model predicts the corresponding estimates at some time later than the rotation age 62 Table 11. Maximum MAI at site index 30 m Prescription DFSIM # Max. SPS TIPSY X E N O Stand M.A.I. age (m3/yr) (yr) Max. Stand M.A.I. age (m3/yr) (yr) Max. Stand Max. Stand M.A.I. age M.A.I. age (m3/vr) (yr) Cm3/vr) (yr) 1 10.25 70 11.3 60 10.0 90 11.4 90 2 10.5 70 11.6 60 9.7 90 * 3 * 7.28 80 - * 4 * 7.92 90 - * 5 9.82 80 11.2 60 10.1 90 * 6 9.95 80 10.52 70 9.8 90 * 7 9.97 90 9.28 80 9.8 90 * 8 10.01 90 8.12 80 9.3 90 * 9 .* 7.01 90 * * 10 11.28 60 12.74 50 - 13.0 80 11 11.65 60 13.14 50 - 13.6 90 12 15.83 80 10.21 50 - 9.7 90 13 10.66 90 6.97 80 - * 14 * 6.87 80 - * 15 * 6.77 80 - * 16 10.87 90 7.23 80 - * 17 9.46 90 7.22 80 - * 18 * 7.3 70 - * 19 - 7.47 80 -20 * 7.63 80 - * 21 8.61 90 7.8 80 * - The model did not simulate the prescription. * The model predicts the corresponding estimates at some time later than the rotation age. 63 Table 12. Maximum MAI at site index 40 m DFSIM SPS TIPSY XENO # Max. M.A.I. (m3/vr) Stand age fvr) Max. M.A.I. (m3/vr) Stand age Max. M.A.I. (m3/vr) Stand age (vr) Max. M.A.I. (m3/vr) Stand age (yr) 1 15.6 60 17.9 50 17.3 70 20.0 90 2 16.1 60 17.8 50 17.0 70 19.7 80 3 * 11.67 70 - 18.9 80 4 * 13.35 70 - * 5 15.4 70 17.2 50 16.9 60 21.7 80 6 15.5 70 17.4 60 16.7 70 22.1 80 7 15.5 80 16.25 70 17.1 70 20.9 90 8 15.4 80 14.7 70 17.0 80 19.7 90 9 15.3 90 12.97 70 17.2 90 * 10 16.45 60 18.26 50 - 20.7 60 11 16.07 70 18.76 50 - 20.9 60 12 15.83 80 17.0 60 - * 13 14.95 90 11.88 70 - 17.6 80 14 * 11.9 70 - 19.6 90 15 * 12.31 70 - 20.4 70 16 14.67 90 12.85 70 - * 17 * 13.11 70 - 17.3 80 18 13.68 70 - 19.5 80 19 15.05 90 13.55 70 - * 20 * 14.15 70 - * 21 14.88 70 _ 18.5 80 - The model did not simulate the prescription. * The model predicts the corresponding estimates at some time later than the rotation age. 64 Table 13. TSV at site index 20 m Prescription Stand age TSV (m3/ha) # DFSIM SPS TIPSY XEN 1 50 201 - 206 129 100 550 - 406 428 2 50 196 - 120 83 100 574 - 375 285 3 50 - - - 127 100 - - - 282 4 50 - - - 85 100 - - - 215 5 50 173 190 154 124 100 521 503 431 379 6 50 165 156 125 94 100 532 432 390 349 7 50 157 129 119 63 100 541 372 376 288 8 50 148 104 99 66 100 541 316 335 301 9 50 140 82 81 51 100 540 264 296 252 10 50 293 - - 212 100 627 - - 379 11 50 287 - - 191 100 655 - - 358 12 50 273 - - 98 100 713 _ _ 210 65 Table 13. (Continued) Prescription Stand age TSV (m3/ha) # (yr) DFSIM SPS TIPSY XENO 13 50 158 83 100 329 232 14 50 192 124 100 378 243 15 50 190 126 100 413 277 16 50 135 76 100 327 240 17 50 158 94 100 346 239 18 50 156 95 100 369 274 19 50 114 61 100 301 225 20 50 130 77 100 315 232 21 50 129 77 100 332 244 - The model did not simulate the prescription. 66 Table 14. TSV at site index 30 m Prescription Stand age TSV (m3/ha) # (vrl DFSIM SPS TIPSY XENO 1 50 477 562 406 466 100 955 965 991 1406 2 50 466 545 375 326 100 988 1005 970 1106 3 50 477 593 - 484 100 1184 940 - 1094 4 50 466 560 - 372 100 1167 961 - 836 5 50 565 534 431 393 100 940 949 986 1311 6 50 415 465 389 341 100 962 976 963 1202 7 50 399 400 375 318 100 980 889 968 1160 8 50 383 341 335 283 100 993 785 929 1038 9 50 363 286 295 251 100 1002 682 878 962 10 50 557 637 - 673 100 1018 1053 - 1625 11 50 561 657 - 640 100 1055 1071 - 1603 12 50 549 465 - 367 100 1151 956 _ 1065 67 Table 14. (Continued) Prescription Stand age TSV (m3/ha) # Or) DFSIM SPS TIPSY XENO 13 50 359 468 305 100 1152 844 861 14 50 580 473 406 100 1116 844 976 15 50 430 534 384 100 1069 969 986 16 50 433 418 292 100 1140 814 885 17 50 415 473 354 100 1103 850 926 18 50 415 465 348 100 1051 898 962 19 50 - 364 282 100 - 775 929 20 50 399 407 327 100 1100 811 943 21 50 399 400 322 100 1056 839 942 - The model did not simulate the prescription. 68 Table 15. TSV at site index 40 m Prescription Stand age TSV (m3/ha) # (yr) DFSIM SPS TIPSY XENO 1 50 768 844 790 917 100 1436 1388 1581 2662 2 50 779 892 770 852 100 1482 1465 1572 2506 3 50 768 894 - 995 100 1697 1422 - 2278 4 50 779 944 - 855 100 1458 1488 - 2169 5 50 731 861 807 1003 100 1433 1391 1536 2925 6 50 705 867 776 879 100 1463 1441 1522 2827 7 50 678 767 771 809 100 1484 1463 1576 2658 8 50 657 676 724 707 100 1500 1335 1625 2369 9 50 627 588 672 633 100 1506 1181 1714 2158 10 50 818 913 - 1128 100 1486 1495 - 2884 11 50 740 938 - 991 100 1496 1560 - 3071 12 50 663 826 - 716 100 1548 1410 2329 69 Table 15. (Continued) Prescription Stand age TSV (m3/ha) # (yr) DFSIM SPS TIPSY XENO 13 50 734 895 816 100 1662 1427 2087 14 50 731 921 961 100 1635 1455 2039 15 50 731 802 962 100 1593 1443 2364 16 50 709 820 690 100 1634 1384 1971 17 50 705 895 843 100 1594 1466 2055 18 50 705 867 892 100 1578 1501 2346 19 50 690 744 641 100 1614 1323 1936 20 50 678 790 769 100 1569 1384 2131 21 50 678 767 745 100 1544 1418 2194 - The model did not simulate the prescription 70 (Table 7) for all the prescriptions that these models were able to simulate. An adequate comparison of maximum growth rate (Table 10) was not possible because some of the simulations attain a maximum growth rate after age 100. TSV (Table 13) shows that for prescriptions 7-9 (densities 625, 500 and 400 sph), SPS and TIPSY indicated similar total standing volume estimates during the rotation. XENO predicted the lowest total standing estimates for the 21 prescriptions over the rotation, and DFSIM predicted the greatest total standing volume estimates for all the prescriptions it was able to simulate. DFSIM virtually doubled the other model estimates. 4.1.2 Site index 30 m Estimates of diameter growth and maximum growth rate varied extremely among the four models. Estimates of total standing volume were similar among models over the rotation. Table 8 shows how models differed when predicting a given mean dbh. A range of response to spacing was generated by the different models: the earliest was by SPS, the latest was by TIPSY. SPS predicted the greatest effect of spacing for the lowest density (1600 sph) in absolute terms. In general terms, SPS estimates ranked first, it was followed by DFSIM and then by TIPSY. XENO usually predicted the oldest age for a given dbh. The models generated a higher response on diameter growth for the combination of spacing and thinning as shown in prescriptions 3 and 4; XENO produced the most delayed response to a combined spacing and thinning treatment (Table 8). For prescriptions 7-9 (densities 625, 500 and 400 sph), DFSIM and TIPSY reached the given mean diameter at approximately the same age. DFSIM, SPS and XENO showed the same pattern of change in diameter growth when stands planted at low initial densities were thinned (prescriptions 13-21). These patterns showed that the same given mean dbh was attained at an earlier stand age when planting densities diminished. The thinning entry 71 had the same effects across densities; the later the entry, the later the given mean dbh was attained. In general terms, as shown in Table 11, the maximum growth rate estimates are roughly the same for all the prescriptions and models. However the age at which the maximum MAI was predicted differed greatly among the models. SPS projected the earliest culmination age (CA) followed by DFSLM, then by TIPSY, and finally by XENO. TIPSY predicted the same CA (90 years) for prescriptions 5-8 (densities 1100, 816, 625 and 500 sph); the growth rate projected by TIPSY was maintained at lower densities. For the same prescriptions (5-8), SPS simulated an earlier CA. SPS projected a delay in the CA when the initial planting density was reduced; in addition, in contrast with TIPSY, SPS simulated a slightly decreasing maximum growth rate for lower initial densities. However, DFSIM simulated an increasing growth rate for lower initial densities, which is the reverse of the TIPSY and SPS projection patterns. Model predictions for total standing volume were similar during the rotation and for all the prescriptions (Table 14). SPS total standing volume estimates were greater at earlier ages for a range of prescriptions (1-4) compared to the other models; these differences diminished as the stand matured. 4.1.3 Site index 40 m Estimates of mean dbh, maximum MAI and TSV were similar among the models, with the exception of XENO. XENO usually predicted the oldest age for the variables measured, except for site index 40 m. Table 9 shows that SPS estimates of age were slightly lower than TIPSY and XENO estimates during the rotation for a given mean dbh. The pattern of responses to spacing was similar among models; however, in relative terms, the response to spacing simulated by XENO 72 was larger than the other models (Table 9). SPS projections indicated the fastest diameter growth for the lowest initial densities (625, 500 and 400 sph). XENO estimates of age showed earlier diameter growth for these prescriptions on this site. When spacing and thinning were combined (prescriptions 3-4), model behaviour and estimates were similar for all the models. The response to spacing and thinning simulated by XENO (Table 9) was larger than the other models. The predicted maximum MAI (Table 12) was approximately the same for all the models except for XENO. XENO predicted the highest maximum growth rate in absolute terms across prescriptions for the site. However, the models differed greatly as to the time that the maximum growth rate was reached. SPS always ranked first, because it predicted the earliest CA. It was followed by DFSIM or TIPSY depending on the simulated prescription; XENO usually predicted the oldest CA. TSV (Table 15) estimates did not differ greatly among the models except for XENO. At age 50, XENO estimates are in agreement with the other model estimates for most of the simulated prescriptions; after age 50, XENO simulated a dramatic increase in volume yields, predicting the highest estimates in absolute terms across prescriptions. The TSV from XENO predictions were up to 100% larger than the other model estimates at age 100. For prescriptions 5-8 (densities 1100, 816, 625 and 500 sph) SPS, DFSIM and TIPSY showed similar values on TSV and MAI predictions during the rotation. 4.2 Fertilization The relevant literature found that fertilization produces more volume, and increases the rate of growth and the diameter size. The growth resulting from the simulated application of fertilization differed greatly among models. Projection estimates are discussed below according to site index. 73 Site index 20 m: DFSIM showed that the given dbh was attained 10 years earlier when the stand had been fertilized (Table 7). In contrast, XENO showed that the given diameter was attained 20 years later when a fertilizer had been added. XENO showed an increase in diameter at the time of the application; however, once the effects of the fertilizer ceased, the simulation seemed to project a slower pace of diameter and volume growth than the nonfertilized (NF) option. The volume yielded at age 100 for prescription 10 was 13% less than for prescription 1 (NF) (Table 13). Nonetheless, XENO projected the greatest percentage (age 50) in volume response for the site and for the fertilized (F) prescriptions 10-11 over prescriptions 1-2 (NF). Site index 30 m: The models are in agreement when projecting diameter growth at an earlier age in the developmental stage of the stand for F options. They differed in the projection of the time at which the effects of fertilization started. SPS projected a given dbh 30 years earlier for the F (prescription 10) than the NF (prescription 1) as shown in Table 8. The growth rate was accelerated approximately one decade for F prescriptions (10-12) across models (Table 11). The volume increase due to fertilization represented a moderate increment (15-20%) in the DFSIM and SPS simulations (Table 14). XENO continued to project a high increase in volume during the rotation which represented the greatest percentage (40-100%) at age 50. Site index 40 m: DFSIM and SPS projected that a given mean diameter was attained a few years earlier in the F prescriptions (10-11) than the NF prescriptions (1-2; Table 9). XENO estimates of age for a given mean diameter suggested that the effect of fertilization had a short term effect; F (10-11) 74 and NF (1-2) prescriptions showed the same age for a given mean diameter (Table 9). The models were not in agreement in the simulation of the CA (Table 12). SPS showed the same CA for both F and NF; DFSIM showed different CA for both F and NF; and XENO predicted CA 30 years earlier for F than for NF. DFSIM and SPS predicted a slight increase in volume (-6%) at age 50-100 for the F over the NF (Table 15). XENO predictions had increased volume by 22% at age 50 (Table 15). 4.3 Commercial thinning The response to CT differed greatly from the findings in the literature (Omule, 1988; Reukema, 1972; Reukema and Bruce, 1977; Warrak, 1979) for most of the models. XENO was in agreement with findings supported by the literature. Every model replicated its behavioural pattern except for XENO, which showed some variation when predicting the thinning response for site index 40 m. The thinning response of DFSIM was the most optimistic, in absolute terms, across models and prescriptions for site index 30 m. DFSLM surpassed the yield produced by the nonthinning prescription. Over the rotation, DFSIM predictions of TSV diminished slightly as the thinning entry was delayed. SPS showed a tendency to converge with the nonthinning prescription when thinning took place at ages 40 and 50. SPS surpassed the nonthinned timber production when the entry was at age 60. DFSIM and SPS predicted maximum M.A.I, and large mean dbh early in the life of a stand. These predictions, according to the literature, reflect an early development of the stand that is not found in BC coastal stands. However, the early predictions of DFSIM on thinning projections were reasonable for the stands investigated in Oregon and Washington. XENO projections produced less TSV for the thinning prescription than the nonthinning prescription; these projections represented 18 to 30% less volume produced, depending on the 75 time the thinning entry takes place (the later the entry the greater the volume produced). Across models, thinning delayed CA with the exception of XENO; XENO accelerated the CA across prescriptions when the thinning entry was performed at age 60 on site index 40 m. 4.4 DISCUSSION The selected models, with the exception of TIPSY, could simulate all the treatments considered for this study. Yet the comparison among the models was partially limited because DFSIM and SPS were not able to simulate, for each site, the complete range of prescriptions established for this study. DFSIM runs are questionable extrapolations when prescriptions call for less than 741 sph either in initial planting densities or spacing. DFSIM runs are also questionable extrapolations when commercial thinning is simulated by the model at an age later than that scheduled in a prescription. If this was the case, these simulation runs implied that the ratio d/D=0.80, required by the model when applying thinning from below, was attained at a later age than the scheduled age. The XENO model could simulate all the prescriptions used for this study and for the various sites. The large number of prescriptions, as well as the variation in output model estimates, ruled out the use of graphical representation. The analysis simply examined the differences observed in the projection estimates for each variable chosen and each site across models. The analysis of prediction estimates revealed that the models were in agreement for a small number of prescriptions on each site. These prescriptions (5-8) included a limited range of densities (625-1100 sph) and no treatments. The models produced a wide range of prediction estimates of age for a given diameter and of growth rate and levels of yield, as a consequence of the effects of the individual or combined treatments prescribed. The wide variation in model output represented a shortcoming; the 76 variation in the results was attributed mostly to the database used to develop the models. These differences in the results were more intense on certain sites. Goudie (1988) concluded that the selection of the driving equation that predicts growth also had a significant impact on model performance. He contended that many other differences between systems seem to be related to the assumptions, structure and the database of the models. The XENO model appears to be the most reliable model for the sequence of simulation runs performed in this study; its simulations of the prescriptions on site index 30 m are in agreement with the experimental evidence and the facts investigated in the literature review. In comparison with other models, XENO projected the growth of a tree at a slower rate during the rotation and for all the prescriptions and sites. XENO growth rate estimates on site index 20 m and 30 m were predicted to culminate past age 100 for the various prescriptions; these estimates, therefore, could not be used in the comparison. The models, although extremely different in their predictions, can prove to be useful tools for researchers, foresters and planners. Simulation models save time when predicting levels of yield and effects of treatments. Although these models have limitations when exploring a prescription, they enable researchers and foresters to assess the response of a range of management options and to compare these options in order to elaborate on his/her management plan. Even when the models were designed for the same geographic area and species there are some local variations that need to be addressed. The DFSIM and SPS databases are made up of plots representing stands located predominantly in Oregon and Washington. The SPS model projected dramatically big trees at an early stand age for widely planted stands (prescriptions 7-9), and for most of the prescriptions and sites. DFSLM projected a dramatic increase in diameter and volume during the rotation for those prescriptions which included commercial thinning 77 (prescriptions 3-4; 13-21) but not for those having low initial planting densities (prescriptions 5-9). XENO projected higher volume yields for stands grown at a wide espacement (prescriptions 5-9) than for stands thinned to low post-thinning densities (prescriptions 13-21). Therefore, users need to be aware of the underlying assumptions embodied in every model. The flexibility of the TIPSY, SPS, DFSIM and XENO models enabled them to project most of the prescriptions for this study, produce readily usable output and allow simple operation. However, in contrast to PROGNOSIS, these models lacked the flexibility to explore the prescriptions in a variety of stand management alternatives including thinning types, planting, site preparation, occurrence of natural events, harvest systems and development of understory vegetation. PROGNOSIS has the capacity to simulate these management alternatives which are extremely useful in the yield prediction of planted stands, and are particularly significant in the yield prediction of natural stands. The stand table output displayed by the single tree distance-independent models provided growth and yield information on the stand. This information proved to be useful in predicting the effects of treatment; however, more detailed information on tree attributes, such as the statistics that single tree distance-dependent models produce, would allow a more precise economic evaluation of the stand. Tree height, diameter and volume are needed not only to calculate log volume, but also to value logs in order to obtain an accurate appraisal of the stand volume. 78 CHAPTER 5 ECONOMIC EVALUATION OF THE SIMULATED PRESCRIPTIONS 5.1 RESULTS AND DISCUSSION The economic evaluation was undertaken using the criteria and assumptions described in Chapter 3. Thus, XENO's stand table output provided the growth and yield estimates used in the financial analysis. The prescriptions chosen for economic evaluation were comprised of prescription Nos. 1-9 and 13-15. These prescriptions were selected as a result of the growth and yield analysis in Chapter 4. This analysis showed the weaknesses of the fertilization performance and, as a consequence, those prescriptions which scheduled a fertilizer application were not included in the economic evaluation. When applied to stands established at low densities, commercial thinning could result in an investment opportunity. Prescriptions No. 13, 14 and 15 (initial density 1100 sph) were selected to provide information about the economic returns of commercial thinning when prescribed for stands planted at low densities. Because of the reduction in plantation costs at the time of establishment, a stand grown at wide espacement and commercially thinned may result in an economically efficient and financially attractive silvicultural investment alternative. The analysis did not consider the other prescriptions (16-21) which also apply thinning to widely spaced stands, because log volume calculations are time consuming. After showing the feasibility of prescriptions 13-15, the economic evaluation of commercial thinning in prescriptions 16-21 might 79 be investigated through further studies. The estimated NPW for each of the prescriptions chosen for the economic evaluation are presented in Table 16. Table 16 shows the net benefits, or present values, of each prescription at 3%, 4%, and 5% interest rates for a range of ages in 10-year steps. The highest NPW for a given prescription represents the optimum economic rotation or harvest age. At a 4 % interest rate, the highest NPW was shown at age 70 by prescription 15. Prescription 13 represented the second highest value at age 70, followed by prescriptions 14 and 3 at age 80. Prescription 4 showed its highest NPW at age 70; it is far below the returns obtained by prescriptions 13-14 and by prescription 3. Prescriptions 1 and 2 showed negative returns. Although returns increased moderately as initial densities decreased, prescriptions 5-9 yielded very low returns. An interest rate of 3% showed the highest NPW (age 90) for prescription 15. Prescription 3 ranked second with its highest NPW also at age 90 as well. Prescription 13 at age 90, and prescription 14 at age 80, ranked third and fourth, respectively. Prescriptions 4, 7, 8 and 9 generated lower returns than prescriptions 15, 3, 13 and 14. Prescriptions 6, 5, 2 and 1 had the lowest returns in the order given, and also reached optimum harvest age within 80-90 rotation age range. Prescription 9 was the only regime that reached the highest PNW at age 100 or later. A decrease by 1% in the interest rate across prescriptions delayed rotation age. At a 5% interest rate prescription 15 yielded the highest NPW at age 70. Prescriptions 13 and 14 at age 60 yielded the second and third highest NPW, respectively, at this interest rate. Prescriptions 1-9 yielded negative returns; therefore, an increase of 1% in the interest rate above 80 Table 16. Net present worth and optimum economic rotation age Prescription # Age fvr) NPW 3% 4% 5% 1 60 519 -739 -1402 70 934 -689 -1466 80 1158 -751 -1579 90 1405* -807 -1676 100 734 -1217 -1899 2 60 -23 -855 -1282 70 493 -702 -1266 80 898* -640 -1300 90 837 -795 -1426 100 869 -899 -1514 3 60 4021 1435 -47 70 4555 1559 -67 80 4956 1584* -136 90 4959* 1423 -278 100 4742 1193 -428 4 60 2700 814 -255 70 3081 887* -284 80 3291* 857 -363 90 3152 674 -498 100 2935 484 -616 5 60 825 -60 -547 70 1182 17 -567 80 1485 44* -611 90 1497* -65 -710 100 1393 -209 -809 6 60 954 72 -415 70 1305 146 -435 80 1555* 150* -490 90 1498 14 -599 100 1514 -79 -676 81 Table 16. (Continued) Prescription Age PNW . # (yr) 3% 4%_ 5% 7 60 1093 190 -309 70 1657 369* -277 80 1814 318 -363 90 2092* 316 -419 100 1843 102 -551 8 60 1103 221 -261 70 1527 332 -267 80 1854 369* -308 90 2001* 312 -386 100 1887 156 -494 9 60 1007 188 -264 70. 1461 322 -249 80 1914 422* -258 90 2038 355 -341 100 2050** 248 -429 13 60 3660' 1836 729* 70 4062 1922* 707 80 4250 1883 625 90 4261* 1765 519 100 4156 1613 414 14 60 3328 1466 389* 70 3818 1594 386 80 4120* 1600* 321 90 4060 1441 192 100 3920 1268 76 15 70 6276 2866* 1036* 80 6379 2798 946 90 6383* 2690 849 100 6295 2557 757 * Economic rotation age. * * Rotation age might occur at this age or later. 82 4% would not be viable. For prescriptions 14 and 15, an increase in the interest rate indicated that the stand could be liquidated at a profit one decade after thinning takes place. Prescription 13 also showed the optimum rotation age decreasing to shortly after thinning. Despite an increase of 1% in the interest rate, prescriptions 6 and 14 indicated no change in rotation age. The net benefits of the prescriptions for planted Douglas-fir expressed in Table 16 revealed that each of all the most financially attractive prescriptions included commercial thinning (prescriptions 3, 13, 14 and 15) at a 4% discount rate. For a stand planted at 1100 sph, commercial thinning is an appropriate cultural practice (prescriptions 13, 14 and 15). This practice made the stand the most financially attractive at ages 40, 50, and 60, although the entry at age 60 produced the highest net benefits (prescription 15). Precommercial thinning and commercial thinning represented the best combination of treatments when prescribed for initial densities of 2500 sph (prescription 3). The volume gains obtained from an intermediate harvest at age 50, and a final harvest at age 80, exceeded the costs of planting and precommercial thinning, turning this management option into one of the most financially attractive. The same regime did not prove to be so attractive for an initial density of 1600 sph (prescription 4). Precommercial thinning did not yield returns for stands planted at 2500 and 1600 sph. It appears that harvest gains did not offset either the costs of planting or precommercial thinning carried on during the rotation up to harvest time. Planting densities of 400, 500 and 625 sph generated very low revenues (prescriptions 9, 8 and 7). Despite the reduction in planting costs, the timber harvested at the optimum economic age did not turn widely spaced stands into an economically advantageous regime. Prescription 15 (a stand planted at 1100 sph and thinned to 300 sph at age 60) produced the highest NPW of all the discount rates tested in this study. The economic analysis also revealed 83 that the optimum rotation age is attained at the time that silviculturalists expect planted managed stands to reach the technical rotation age (70-90). 5.2 PRUNING: RESULTS AND DISCUSSION For pruning, the results of the economic evaluation were obtained using the butt log volume calculations based on XENO's stand table output for prescriptions (Table 2), as indicated in Section 3.3.7 in Chapter 3. The returns from the butt logs of pruned trees were analyzed across prescriptions 1-9 and 13-15; these prescriptions corresponded to the same set of prescriptions used in the determination of the optimum economic rotation age. Table 17 shows some of the results that are obtained when evaluating pruning. According to this economic evaluation, only prescriptions 3, 8 and 9 paid for pruning costs at the time of harvest. Prescriptions 4, 7 and 14 showed NPW results that were the closest to the amount required to equal the investment. Prescriptions 1, 2, 5, 6, 13 and 15 produced returns lower than the amount required to equal the investment. Prescription 3 may prove that precommercial thinning, pruning, and commercial thinning are the best combination of scheduled treatments for an initial density of 2500 sph. The benefits from prescriptions 8 and 9 greatly exceeded the costs of the treatments; however, there is no chance to show pruning may affect the economic evaluation of the silvicultural regime for these prescriptions. The economic evaluation of pruning measured feasibility for treatment and, consequently, pruning shows the viability of the investment independently of the prescribed regime. 84 Table 17. Net present worth of pruning Prescription Rotation Pruning Pruning # Age Returns Costs (yr) $/ha $/ha 3 80 723.53* 592.00 4 70 503.23 592.00 7 70 529.27 592.00 8 80 658.26* 592.00 9 80 724.68* . 592.00 14 80 542.19 592.00 * Pruning is feasible The financial analysis showed that the revenues obtained from pruning did not pay for the costs of the treatment in most of the selected cases. These results were conservative because premium prices for pruned logs were not available in the VLM. V L M prices represent the market prices for the coastal region. Using this market price, pruned logs were valued at a Grade H price. Grade H represented the largest dbh attained by a tree in the simulations for this study. According to the Scaling Manual for BC (Ministry of Forests, 1995), there is no percentage of clear wood in Grade F£ logs; in addition, there are some other specifications required for grade H, such as knot size (4-8 cm) and number of rings (6/cm) which, in the absence of pruning, are not applicable. In unmanaged second-growth stands, 60 years are necessary for complete occlusion of the wound once a branch is detached. In the natural process, this wound becomes an enclosed black or loose knot 15 cm or more in length (Kachin, 1940, as cited in Reeb, 1984). Log requirements, described in the Scaling Manual for BC (Ministry of Forests, 1995), are specific for 85 old-growth timber; these requirements may not apply to second-growth. It is likely that adjustments to these requirements should be made when upgrading pruned logs from second-growth or planted stands. Clear wood would certainly appear in lower diameter classes than those recognized by the currently used Scaling Manual. Therefore, a different set of requirements for second-growth and plantation log grades will be required. Pruning will then play a key role in satisfying grading measurements. Prescriptions 4, 7 and 14 indicated the potential feasibility of pruning. These prescriptions may show feasibility if diameter class 40 cm is taken into account for the volume calculation of the butt log, other things being equal. In the presence of pruning, the 40 cm diameter class will certainly show a recognizable percentage of clear wood. The economic evaluation of pruning might then have been justified for a broader range of prescriptions. Neither high costs nor the lack of premium markets justify not pruning. As cited in the literature review, second-growth and planted stands cannot provide high quality timber in the absence of pruning. Pruning may be the most effective action to be taken in order to maintain an international quality timber market share. 86 CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS 6.1 Simulation models The models were weak when projecting response to fertilization. Across the models, the growth resulting from the simulated use of a fertilizer differed in the duration of the effects of the fertilizer, the increase and the acceleration of the growth rate, and the amount of volume produced. There were also differences among the models with respect to response from commercial thinning. More field response data from trials would be helpful in reconciling these disparities. Model predictions were mostly in agreement for stands established at low initial planting densities (prescriptions 5-9). For these simulations, the estimates of diameter growth were roughly the same for DFSLM, TIPSY and SPS on site index 20 m and 30 m; and for DFSLM, TIPSY, SPS and XENO on site index 40 m. The estimates of growth rate were roughly the same for DFSIM, TIPSY and SPS on site index 30 m; TIPSY and DFSIM also agreed on CA. Estimates of TSV, with a few exceptions, were roughly the same for TIPSY, SPS and XENO on site index 20 m; for DFSLM, TIPSY, SPS and XENO on site index 30 m; and for SPS, DFSIM and TIPSY on site index 40 m. It is desirable that models produce more detailed information on tree attributes, as do single tree distance-dependent models, in order to enable more precise economic evaluation. The more 87 detailed the qualitative and quantitative physical information on the forest product, the more appropriate the economic evaluation becomes. The work already done by modellers, biometricians, silviculturalists and other researchers shows dedication, diligence and continuous effort. Despite the significant amount of research, growth and yield simulation needs more development. The research of biometricians constitutes a key element in modelling development and upgrading. The research of field foresters, mensurationists and analysts is also much needed in database development and enlargement. Model upgrading and development need ongoing research in order to render the models more efficient and to provide the model user with a more suitable variety of management options. 6.2 Economic evaluation The economic evaluation of the prescriptions tested determined the best combination of a sequence of scheduled activities, and the viability of the silvicultural regime over the span of the rotation. Precommercial thinning in combination with commercial thinning proved to be an effective management alternative for initial planting densities of 2500 sph (prescription 3) Commercial thinning manifested economic efficiency and flexibility for prescriptions with 1100 sph (prescriptions 13, 14 and 15) at the time of establishment; commercial thinning may demonstrate viability for prescriptions in stands planted at densities as low as 816 and 625 sph. Pruning proved to be a feasible treatment within the limits of an extremely conservative evaluation. The economic evaluation of pruning suggested that pruning might prove to be an efficient prescribed treatment for a range of 200-300 pruned trees. The financial analysis indicated that a lower discount rate than the recommended rate of 4% may delay optimum rotation age. However, a lower interest rate did increase the number of prescriptions that were financially attractive. A higher interest rate than the recommended rate 88 reduced the number of economically viable prescriptions. A higher interest rate may lower the optimum rotation age. In essence, this study showed that the economic evaluation of silvicultural prescriptions can be used to limit the number or range of prescriptions considered as alternative investments, to determine which treatments should be considered as components of a prescription, to indicate the planting densities that are financially attractive, and to generate decisions concerning timing of thinning and the optimum harvest age. The research effort required to provide the forest economy with more accurate basic data is worth continuing at a more intense pace. Accurate and refined information from growth and yield simulation modelling will undoubtedly benefit the economics of the forest industry, resulting in more precise, reasonable, and lucid stand and forest management decisions. 89 LITERATURE CITED Arney, J. D. 1972. Computer simulation of Douglas-fir tree and stand growth. Pacific Forest Research Centre, Canadian Forestry Service, Internal Report BC-27. 1980. Stand projection system (SPS). Mason, Bruce and Girard. Inc. Portland, Or. pp. 79. Barclay, H. J. and Brix, H. 1984. Effects of urea and ammonium nitrate fertilizer on growth of a young thinned and unthinned Douglas-fir stand. Canadian Journal Forestry Research Vol. 14:952-955. Baskerville, G. 1986. Understanding forest management. The Forestry Chronicle Vol. 62(4) 339-347. Brand, G. J. and Holdaway, M. R. 1983. Users need performance information to evaluate models. Journal of Forestry Vol. 91:235-237. Briggs, D. G. and Fight, R. D. 1992. Assessing the effects of silvicultural practices on product quality and value of coast Douglas-fir trees. Forest Products Journal 42(1): 40-46. Briggs, D. G. and Smith, W. R. 1986. Effects of silvicultural practices on wood properties of coniferous: a review. In Douglas-fir stand management for the future. College of Forest Resources. C. D. Oliver, D. P. Hanley and J. A. Johnson (editors). Institute of Forest Resources, University of Washington, Seattle, Contribution No. 55. pp. 108- 117. Brix, H. 1981. Effects of thinning and nitrogen fertilization on branch and foliage production in Douglas-fir. Canadian Journal Forestry Research Vol. 11: 502-511. 1981. Effects of nitrogen fertilizer source and application rates on foliar nitrogen concentration, photosynthesis and growth of Douglas-fir. Canadian Journal Forestry Research Vol. 11:775-780. 1982. Effects of thinning and nitrogen fertilization on growth of Douglas-fir: relative contribution of foliage quantity and efficiency. Canadian Journal Forestry Research Vol. 13: 167-175. Brix, H. and Mitchell, K. A. 1983. Thinning and nitrogen fertilization effects on sapwood development and relationships of foliage quantity to sapwood area 90 and basal area in Douglas-fir. Canadian Journal Forestry Research Vol. 13:384-389. Buchman, R. G. and Shifley, S. R. 1983. Guide to evaluating forest growth projection systems. Journal of Forestry Vol. 91:232-235. Cahill, J. M.; Snellgrove, T. A. and Fahey, T. D. 1986. The case for pruning young-growth stands of Douglas-fir. In: Douglas-fir stand management for the future. College of Forest Resources. C. D. Oliver, D. P. Hartley and J. A. Johnson (editors). Institute of Forest Resources, University of Washington, Contribution No. 55. pp. 121-131. Cahill, J. M.; Snellgrove, T. A. and Fahey, T. D. 1988. Lumber and veneer recovery from pruned Douglas-fir. Forest Products Journal 38(9) : 27-32. Carter, R. 1989. Variability of crop tree response to fertilization in unspaced Douglas-fir stand. FRDA. 085. Victoria, BC Carter, R. and Scagel, R. 1989. The effects of stand density and fertilization of stand (Table 7) for all the development in immature coastal Douglas-fir. FRDA. 094. Victoria, BC. Cheaps and Pratt 1989. The social discount rate for silvicultural investments. Ministry of Forestry. FRDA Report 071. Victoria, BC. Cole, E. C. and Newton, M. 1987. Fifth-year responses of Douglas-fir to crowding and nonconiferous competition.. Canadian Journal Forestry Research Vol. 17:181-186. Curtis, R. O.; Clendenem, G. W. and DeMars, D. J. 1981. A new stand simulator for coast Douglas-fir: DFSIM user's guide. USDA Forest Service, General Technical Report, PNW-128. Daniels, et al. 1979. Principles of silviculture. 2nd ed. McGraw-Hill Book Company, Series in Forest Resources, New York. pp. 500. Darling, L. M. and Omule, S. A. Y. 1989. Extensive studies of fertilizing and thinning coast Douglas-fir and western hemlock: an establishment report. FRDA. No 54. Ministry of Forests. Victoria, BC. Davis, L.S. and Johnson, K. N. 1986. Forest management. 3rd ed. McGraw-Hill Book Company, Series in Forest Resources, New York. pp. 790. Eversole, K. R. 1953. Better marking means cheaper pruning. USDA Forest Service, Research Note, PNW- 87, Portland, Or. 1955. Spacing tests in a Douglas-fir plantation. Forest Science Vol. 1(1): 14-18. Fight, R. D.; Cahill, J. M.; Fahey, T. D. and Snellgrove, T. A. 1987. Financial analysis of pruning coast Douglas-fir. USDA Forest Service, Research Paper, PNW-390 91 Fight, R. D.; Cahill, J. M. and FHA, T. D. 1988. A new look at pruning coast Douglas-fir. Western Journal Applied Forest 3(2): 46-48. Finnis, J. M. 1953. Experimental pruning of Douglas-fir in British Columbia. British Columbia Forest Service, Research Division, Research Notes No. 24. Victoria, BC. Fleischer, H. 1949. The suitability of second-growth Douglas-fir logs for veneer. Journal of Forestry 47: 533-537. Fujimori, T. 1975. Crown and canopy structure in relation to productivity. In Proceedings of an International workshop, 14 - 20 October 1985, Japan. Takao Fujimori and David Whitehead (editors). Goudie, J. W. 1980. Yield tables for managed stands of lodgepole pine in northern Idaho and southeastern British Columbia. M. Sc. thesis, University of Idaho, Moscow, ID. pp.111. 1987. PROGNOSIS, SPS and TASS: yield comparisons for even-aged interior Douglas-fir. Ministry of Forests and Lands, Research Branch, FPDS. Victoria, BC. pp. 69. Goulding, C. J. 1979. Validation of growth models used in forest management. New Zealand Journal of Forestry 24: 108-124. Grah, R. F. 1961. Relationship between tree spacing, knot size and log quality in young Douglas-fir stands. Journal of Forestry Vol. 59: 270-272. Grier, C. C; Lee, K. M and Archibald, R. M. 1984. Effect of urea fertilization on allometric relations in young Douglas-fir trees. Canadian Journal Forestry Research Vol. 14:914-904. Harrington, C. and Reukema, D. L. 1983. Initial shock and long term stand development following thinning in a Douglas-fir plantation. Forest Science Vol. 29(1) : 33-46. Hedin, I. B. 1982. Pruning Douglas-fir on coastal British Columbia. FERIC, Interim Report IR- 383-1. Heiberg, S. O. and Haddock, P. G. 1955. A method of thinning and forecast of yield in Douglas-fir. Journal of Forestry Vol. 53: 10-18. Klinka, K. 1980. Characterization of the most productive ecosystems for the growth of Pseudotsuga menziesii var. menziesii in Southwestern British Columbia. Land management, Report No. 6. Ministry of Forests. Victoria, BC. Krueger, K. W. 1959. Diameter growth of plantation grown Douglas-fir trees under varying degrees of release. USDA Forest Service, Research Note, PNW-168, Portland, Or. 92 Marshall, P. L. 1988. A decision analytic approach to silvicultural investment decisions. FEPA, Working paper No. 110 1989. The economic value of additional information about treatment-response information for coastal Douglas-fir. FEPA, Working paper No. 121. 1991. Using decision analysis for stand-level silvicultural decisions. The Forestry Chronicle Vol. 67(4): 384-388. Miller, R. E. and Harrington, C. A. 1979. Response to urea and ammonium nitrate fertilization in an 80-year-old Douglas-fir stand. USDA Forest Service, PNW- 330. pp. 1-7. Miller, R. E. and Tarrant, R. F. 1983. Long term growth response of Douglas-fir to ammonium nitrate fertilizer. Forest Science Vol. 29 (1) : 127-137. Miller, R. E. and Wert, S. 1979. Effects of soil and foliar applications of nitrogen fertilizers on a 20-year-old Douglas-fir stand. USDA Forest Service, PNW-329. pp. 1-72. Ministry of Forests. 1980a. Forest and Range Resource Analysis. Technical Report, Victoria, BC. 1982. Manual of silviculture, Silviculture Branch. Victoria, BC. 1990. Benefits of incremental silviculture. Ministry of Forests, Silviculture Branch. Victoria, BC. 1992. Background and recommendations for juvenile spacing standards in BC. Ministry of Forests, Silviculture Branch and Economic and Trade Branch, Zaing and McCulloch, Forest Management Services and Nawitka Research Consultants Ltd. Victoria, BC. 1993. An economic evaluation of commercial thinning Douglas-fir in the coastal region of British Columbia. Forest Economic Report by Michael Stone. Victoria, BC. 1994. Background and recommendations for juvenile spacing standards in BC. Ministry of Forests, Silviculture Branch and Economic and Trade Branch. Victoria, BC. 1995. Scaling manual. Ministry of Forests, Revenue Branch. Victoria, BC. 1995. Site preparation guidebook. Forest practices code of British Columbia. Ministry of Forests Act. Victoria, BC. 1995. VDYP interactive application. Ministry of Forests, Resources Inventory Branch. Victoria, BC. 93 Mitchell, K. J. 1969. Simulation of the growth of even-aged stands of white spruce. Bulletin No. 75. Yale University, School of Forestry, New Haven, CT. pp. 48. 1975. Dynamics and simulated yield of Douglas-fir. Forest Science. Monograph. 17. pp.39. 1975i. Stand description and growth simulation from low-level stereo photos of tree crowns. Journal of Forestry 73: 12-16,45. Mitchell, K. J. and Cameron, I. R. 1985. Managed stand yield tables for coastal Douglas-fir: initial density and precommercial thinning. BC. Ministry of Forests, Research Branch, Land Management Report No. 31. Victoria, BC. Mitchell, K. J. and Grout, S. E. 1992. User's guide for TIPSY. Ministry of Forests, Research Branch, Forest Productivity and Decision Group. Victoria, BC. Munro, D. D. 1974. Forest growth models, a prognosis. In J. Fries (ed.) Growth models for tree and stand simulation. Working party S4.01-4. Royal College of Forestry, Stockholm, Department of Forest Yield Research, Res. Notes No. 30. pp. 7-21. Nawitka Resource Consultants Ltd. 1992. Incremental silviculture: pruning financial analysis. Silvicultural Branch Report. Ministry of Forests. Victoria, BC. pp. 13. 1993. Financial analysis for selected and combined silviculture treatments for the BC coast. Ministry of Forests, Silviculture Branch. Victoria, BC. Northway, S. 1989. XENO Technical report. 1990. XENO user's guide. O'Hara, K. L. 1988. Stand structure and growing space efficiency following thinning in an even-aged Douglas-fir. Canadian Journal Forestry Research Vol. 18:859-866. O'Hara, K. L. and Oliver, C. D. 1988. Three-dimensional representation of Douglas-fir volume growth: comparison of growth and yield models with stand data. Forest Science 34: 724-743. Olympic Natural Resources Centre 1990. Non-timber strategies for precommercial and commercial thinning. Draft plan, University of Washington, Seattle, Wash. Omule, S. A. 1985. Response of coastal Douglas-fir to precommercial thinning on medium site in BC. Ministry of Forests. Victoria, BC. 1988. Growth and yield 35 years after commercially thinning 50-year-old Douglas-fir. Ministry of Forests and Lands, Research Branch. Victoria, BC. pp.15. 94 Omule, S. A.; Paul, D. E and Darling, L. M. 1994. Coast of pruning Douglas-fir in coastal British Columbia. The Forestry Chronicle Vol. 70(1): 80-83. Reeb, D. 1984. Influence of spacing and artificial pruning on the production of clearwood of Douglas-fir. M. Sc. thesis, Faculty of Forestry, University of British Columbia, Vancouver, BC. Reukema, D. L. 1959. Some recent development in the Wind River Douglas-fir plantations spacing tests. USDA Forest Service, Research Notes No. 167, Portland, Or. 1970. Forty-year development of Douglas-fir stands planted at various spacings. USDA Forest Service, Research Paper, PNW-100. 1972. Twenty-orte-year development of Douglas-fir stand repeatedly thinned at varying intervals. USDA Forest service, Research Paper, PNW-141. 1975. Guidelines for precommercial thinning of Douglas-fir. USDA Forest Service, Technical Report, PNW-30, Portland, Or. 1979. Fifty-year development of Douglas-fir stands planted at various spacings. USDA Forest Service, Research Paper, PNW-253. pp.121, Portland, Or. Reukema, D. L. and Bruce, D. 1977. Effects of thinning on yield of Douglas-fir: concepts and some estimates obtained by simulation. USDA Forest Service, General Technical Report, PNW-58, Portland, Or. Reukema, D. L. and Piennar, L. V. 1973. Yields with and without repeated commercial thinnings in a high-site-quality Douglas-fir stand. USDA Forest Service, Research Paper, PNW -155. Reukema, D. L. and Smith, J. Ff. G. 1987. Development over 25 years of Douglas-fir, western hemlock and western redcedar planted at various spacings on a very good site in British Columbia. USDA Forest Service, Research Paper, PNW-391. Shaw, E. W. and Staebler, G. R. 1950. Financial aspects of pruning. USDA Forest Service, PNW-and Range Experiment Station, Portland, Or. pp. 45. Smith, J. H. G. 1959. Comprehensive and economical designs for studies of spacing and thinning. Forest Science Vol. 5(3): 237-245. 1960. Comments on relationships between tree spacing, knot size, and log quality in young Douglas-fir stands. Journal of Forestry 59(9): 682-683. 95 1980. Influences of spacing on radial growth and percentage latewood of Douglas-fir, western hemlock and western redcedar. Canadian Journal Forestry Research Vol. 10: 169-175. 1986. The practice of silviculture. John Wiley and Sons, Inc. (editor), New York. pp. 527. Smith, H. G. and Kennedy, R. W. 1983. Effects of stand management on tree, log and wood quality of Douglas-fir. Faculty of Forestry, University of British Columbia, Vancouver, BC. Staebler, G. R. 1956. Evidence of shock following thinning of young Douglas-fir. Journal of Forestry Vol. 53:339-340. 1963. Growth along the stems of full crowned Douglas-fir trees after pruning to specified heights. Journal of Forestry 61(2) : 124-127. 1964. Height and diameter growth for four years following pruning of Douglas-fir. Journal of Forestry Vol. 62: 406. Stage, A. R. 1973. Prognosis model for stand development. USD A Forest Service and Range Experimental Station, Research Paper INT-137. Stein, W. I. 1955. Pruning to different heights in young Douglas-fir. Journal of Forestry, 53(5): 352-355. Steele, R. 1955. Growth after precommercial thinning in two stands of Douglas-fir. USDA Forest Service, Research Notes No 117, Portland, Or. Stegemoller, K. A. and Chapell, H. N. 1990. Effects of fertilization and thinning on 8-year growth responses of second-growth Douglas-fir stands. Canadian Journal Forestry Research Vol. 21:516-521. Tappeiner, J. C; Bell, J. F. and Brodie, J. D. 1982. Response of young Douglas-fir to 16 years of intensive thinning. School of Forestry, Oregon State University, Research Bulletin 38, Or. Thomson, A. J. and Barclay, H. J. 1984. Effects of thinning and urea fertilization on the distribution of area increment along the boles of Douglas-fir at Shawnigan Lake, British Columbia. Canadian Journal Forestry Research Vol. 14:879-884. Vasievich, J.; Frebis, R. and Wiethe, R 1984. Quick Silver, The forestry investment analysis program. USDA Forest Service, Southeastern Center for Forest Economics Research, NC. Velazquez-Martinez, A.; Perry, D. A. and Bell T. E. 1982. Response of above ground biomass increment, growth efficiency, and foliar nutrients to thinning, fertilization and pruning in 96 young Douglas-fir plantations in the central Oregon cascades. Canadian Journal Forestry Research Vol. 22: 1278-1289. Warrack, G. C. 1948. Thinning and pruning second-growth Douglas-fir in the coastal region of BC. Forest Service, Research Note No. 13. pp. 1-8. 1979. Successive thinning in a natural stand of Douglas-fir over a fifty year period. Ministry of Forests, Research Note No. 87. Victoria, BC. Weetman, G. C. 1991. Biological uncertainty in growth and yield benefits of second-growth silviculture actions and old-growth protection and liquidation strategies in BC FEPA., Working paper No. 148. Worthington, N. P. 1957. Some economic considerations in thinning Douglas-fir. USDA Forest Service, Research Note, PNW-137, Portland, Or. 1961. Some observations on yield and early thinning in a Douglas-fir plantation. Journal of Forestry Vol. 59:331-334. 1966. Response to thinning 60-year-old Douglas-fir. USDA Forest Service, Research Note, PNW-35. Worthington, N. P. and Staebler, G. R. 1961. Theory and practice of commercial thinning in Douglas-fir. Journal of Forestry Vol. 59:591-593. Worthington, N. P.; Reukema, D. L. and Staebler, G. R. 1962. Some effects of thinning on increment in Douglas-fir in western Washington. Journal of Forestry Vol. 60:115-119. 97 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items