Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relationship of cone production to wood traits of lodgepole pine Robertson, Donna L. 1989

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

Item Metadata

Download

Media
831-UBC_1989_A6_7 R06.pdf [ 7.14MB ]
Metadata
JSON: 831-1.0097545.json
JSON-LD: 831-1.0097545-ld.json
RDF/XML (Pretty): 831-1.0097545-rdf.xml
RDF/JSON: 831-1.0097545-rdf.json
Turtle: 831-1.0097545-turtle.txt
N-Triples: 831-1.0097545-rdf-ntriples.txt
Original Record: 831-1.0097545-source.json
Full Text
831-1.0097545-fulltext.txt
Citation
831-1.0097545.ris

Full Text

R E L A T I O N S H I P O F C O N E P R O D U C T I O N T O W O O D T R A I T S O F L O D G E P O L E P I N E By Donna L. Robertson B.S. (Genetics) University of California, Berkeley A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S D E P A R T M E N T O F F O R E S T R Y We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A August 1989 © Donna L. Robertson, 1989 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. I 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 Forestry The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: Abstract In order to examine the relationship between cone production and wood traits, wood samples were taken from a clonal seed orchard of 15-year-old lodgepole pine (Pinus contorta Dougl.) and analyzed using X-ray densitometry. The relative density of each tree at breast height (1.38 m above ground), per-year volume and per-year bolewood biomass were calculated. In addition, cone counts were recorded from the seed orchard over a period of several years. The earlywood, latewood, and total-year components of volume, biomass, and relative density were regressed against cone counts for each tree over several years. In addition, the densitometer data were combined with May-July weather data for multiple regressions to discover which other variables were influencing wood traits. Analyses of variance showed a strong annual effect, which reflected the growth trends of the orchard trees. In addition, there was a strong clonal effect in the analysis of cone count and relative density. Regression results showed that ramet age was the most important factor in predicting the volume and biomass increment each year. There was a small effect on wood traits from cone bearing, temperature, and precipitation. Heavy conebearing had a negative relationship with wood density, but a positive relationship with volume and biomass increment. The positive relationship with volume and biomass was unexpected and may be the result of the strong effect of ramet age. Progeny data showed a uniformly negative relationship between wood traits and family fecundity. This may imply that unless there is selection against high fecundity, future plantations may experience decreases in wood density. 11 Table of Contents Abstract ii List of Tables iv List of Figures v Acknowledgement vi 1 Introduction 1 2 Background 5 2.1 Phenotypic evidence 6 2.2 Genetic evidence 8 2.3 Wood traits 10 3 Materials and Methods 12 3.1 Background of Omineca-Pinchi seed orchard 12 3.2 Sample collection 15 3.3 X-ray densitometry 18 3.4 Transformation of densitometer output 22 3.5 Data analysis 27 3.5.1 Data sources . . . . . . 27 3.5.2 Preliminary analysis 30 3.5.3 Weather analysis 33 in 3.5.4 Individual-tree analysis 34 3.5.5 Clonal means 35 3.5.6 Progeny data . 35 4 Results 37 4.1 Data parameters 37 4.2 Analyses of variance 39 4.3 Weather data 40 4.4 Individual-tree data 41 4.5 Clonal means 49 4.5.1 Wood trait by cone count regressions 49 4.5.2 Trends over time 62 4.6 Means based on cone-production level 68 4.7 Progeny comparisons 70 5 D i s c u s s i o n 71 5.1 Data sources 71 5.2 Preliminary analyses 71 5.2.1 Data parameters 72 5.2.2 Analyses of variance . 72 5.2.3 Weather analysis 74 5.3 Individual-tree data : 74 5.3.1 Simple regressions . 75 5.3.2 Multiple regressions 76 5.4 Clonal means for analysis . 78 5.4.1 Regressions 78 5.5 Trends over time 79 iv 5.6 Progeny results 82 6 Conclusions 83 7 References Cited 85 Appendices 90 A Database for volume and biomass analysis 90 B Database for relative density analysis 91 C Weather data 92 D Swedish progeny trial data 94 E By-clone and by-year means 95 v List of Tables 3.1 Format for analysis of variance 33 4.2 Sample size (N), average of means, range and standard error of the mean (Sx) for cone production and wood traits, by clone and by year. Ring density is at breast height 38 4.3 Results from analysis of variance for four traits 39 4.4 Coefficient of determination (r2), standard error of the estimate (Se), and slope for the single regressions of wood traits, using individual-tree data. 45 4.5 Coefficients of determination (R2), standard error of the estimate (5e), and slope of cone count for the multiple regression of wood traits, using individual-tree data 47 4.6 Single regressions of wood traits by cone production, using clonal means as data 53 4.7 Coefficient of determination (R2), standard error of the estimate (i>e), a n d slope of cone count for the multiple regression of wood traits, using clonal means 58 4.8 Progeny comparisons 70 D.9 Swedish progeny trial data 94 vi List of Figures 3.1 Twin circular saws for preparing a 1.57 mm strip from increment cores. . 17 3.2 Diagram of X-ray densitometer for scanning wood samples 19 3.3 Diagram of X-ray scanning of a wood sample 20 3.4 Wood density profile, showing early and latewood components of ring density. 21 3.5 Diagram of tree growth by year, with the sheath of wood illustrated. . . 24 3.6 Flowchart showing data analysis steps 28 4.7 Volume regressions with individual-tree data 42 4.8 Biomass regressions with individual-tree data . 43 4.9 Density regressions with individual-tree data 44 4.10 Multiple regression surface for volume 50 4.11 Multiple regression surface for biomass 51 4.12 Multiple regression surface for relative density 52 4.13 Volume regressions with clonal means 55 4.14 Biomass regressions with clonal means 56 4.15 Density regressions with clonal means 57 4.16 Multiple regression surface for volume, with clonal means 59 4.17 Multiple regression surface for biomass, with clonal means 60 4.18 Multiple regression surface of relative density, using clonal means . . . . 61 4.19 Cone production over time 63 4.20 Volume sheath production over time 64 4.21 Biomass sheath production over time 65 vn 4.22 Relative density over time 66 4.23 May-July weather patterns, 1979-1987 67 4.24 Density profiles of heavy and light conebearers 69 viii A c k n o w l e d g e m e n t This study was funded in part by the Poldi Bentley Chair in Forest Genetics/ Tree Improvement at the University of British Columbia. Funding was also obtained from the Ministry of Forests of the Province of British Columbia. The co-operation of Forintek Canada Corporation is gratefully acknowledged. Diagrams on pages 17, 19, 20, 21, 24, and 69 were provided by Forintek. In addition, a great number of individuals have provided assistance in some capacity on this project. My supervisor, Dr. Donald T. Lester, was instrumental in the initi-ation and development of this thesis. Dr. Valerie LeMay and Dr. Judy Loo-Dinkins provided valuable statistical assistance throughout the project. Drs. John Worrall, Gor-don Weetman, Anton Kozak and Bob Kellogg examined the thesis and offered valued opinions. Carole Lee at the Prince George Tree Improvement Station provided assistance in the collection of the wood samples. Dr. Michael Carlson of the Ministry of Forests assisted in obtaining the funding for the project from the Ministry. Dr. Les Josza and Arlene Sawadsky at Forintek were instrumental in the X-ray densitometry. Swedish progeny data were provided by Dr. Anders Fries at the Swedish Agricultural University in Umea, Sweden. The two- dimensional graphs were created using a graphics program written by Ed Casas and Joel Bisson of the Deparment of Electrical Engineering at UBC, and constructed by Joel Bisson. Lastly, Kevin O'Donnell provided a great deal of assistance throughout the prepa-ration of this thesis, from wood sample collection to computer programming, as well as offering many useful insights and unfailing support. i x C h a p t e r 1 Int roduct ion Forest tree improvement can be considered as the union of silvicultural techniques with the application of genetic principles to produce plantations of forest species with desirable characteristics. As any tree is both a product of its specific genotype and the environment in which it has grown, both factors may be influenced by selecting trees with proven parentage and utilizing intensive silvicultural treatments to ensure the optimal microsite. Within the overall category of forest tree improvement, tree breeding is the practice of selecting, breeding and reselecting individual trees to produce a desirable crop tree. Tree breeding objectives differ from species to species. For five-needle soft pines in North America, tree breeding efforts center on resistance to white pine blister rust. For many other species, increased height or volume growth is the main objective. Typically, a tree breeding program consists of a specified path of endeavor. Wild stands are cruised and trees with phenotypically superior characteristics are recorded. Samples from these candidate "plus" trees are then collected by removing either scion material or cones or both. These samples are then established in orchards either by grafting the scions or growing the seedlings. In general, by the time a tree is recognizably superior in a wild stand, it is physiologically too mature to yield rootable cuttings. At present, seed orchards are the most commonly used method of accessing genetic improvement for most species. Grafts are usually preferred to seedlings for orchard es-tablishment because they capture the exact genotype of the plus tree. Grafting, however, is not a feasible way to produce millions of genetically improved trees for reforestation. 1 Chapter 1. Introduction 2 A graft is costly to produce operationally and may grow plagiotropically as a result of its level of maturity. The fact that it is genetically identical to the original plus tree is of benefit, however, as it in effect brings the plus tree to the seed orchard manager, permit-ting controlled crosses to be made. Both seeds and grafts are useful in tree improvement programs; seeds are more easily handled operationally. Optimally, the seed from the parent trees is tested in common-garden progeny tests, which leads to establishment of seed orchards made up of grafts from the "best" trees. However, the time lag between establishment of the progeny test and the seed orchard makes this solution generally unappealing. More commonly, grafts are established in seed orchards at the same time as the seedlings are established in progeny trials, and information from the progeny trials leads to roguing the seed orchard of grafts from undesirable parent trees. Controlled crosses are made within the improved orchard and the progeny of the crosses are tested. A second-generation orchard is established from grafts of the second progeny test, while the first-generation orchard is phased out. This cycle continues with selection for the most desirable traits balanced by awareness of possibly dangerous levels of inbreeding. Criteria for ranking progeny tests are based on the desirable characteristics, insofar as they can be shown at a young age. Height, diameter and crown form are common traits. As well, some programs have begun to include wood characteristics, especially relative density which has a major impact on the quality of both lumber and pulp. Seed orchards in British Columbia are funded by the provincial government through the Ministry of Forests. Although many of the orchards, are located on private lands, the Ministry has paid for many of the initial costs for irrigation systems and continues to lease the orchard lands and to supervise management of the orchards. There are legislated requirements for the use of improved seed in reforestation; the Chapter 1. Introduction 3 Ministry- has mandated that licensees shall acquire this seed at their own expense. Re-cently, the government of British Columbia has proposed two options to allow forest licensees to acquire improved seed. In the first option, the status quo will remain in place, in which the Ministry manages the seed orchards and sells the seed to the li-censees. The second option allows the seed orchards to be managed by the licensees on whose land they are located, and for the licensees to sell the seed. If the second option is chosen, certain conditions apply. Seed orchard seed must be used in preference to wild stand seed if it is available. Seed orchards must be certified before the seed is registered and used for reforesting Crown land. Older seed orchards may be decertified as advanced-generation seed becomes available. The Ministry will vegetatively propagate orchard stock of major licensees or cooperatives for a fee. The burden of reforestation was recently shifted from the Ministry of Forests to the companies doing the logging. There is an incentive, therefore, for licensees to choose the second option and maximize seed production in their seed orchards. As genetically improved seed must be used when it is available, a licensee with a seed orchard may have a virtual monopoly on seed for certain areas, and it may be sold in large quantities. Seed orchard seed, however, is not without disadvantages. Parental imbalance within seed orchards may create future problems. Due to the high degree of .variability among parents in seed production, certain families can be greatly over- or under-represented in an orchard's seed lot. This means that much of the genetic variability in the seed orchard is not being expressed in the outplanted seedlings, and that relatively few farnilies may be covering thousands of hectares in the future. Since the trees that make up seed orchards were originally chosen for their outstanding characteristics, there could be a great loss in the potential of future plantations if much of the genetic variation is unutilized. Another facet of the same problem is the quality of the trees that will grow from this seed. The quantity of cones produced is highly heritable. It follows, therefore, that Chapter 1. Introduction 4 the seedlings grown from trees that were covered with cones will themselves be heavy cone bearers. In the future, reforestation of genetically improved seedlings will be done not by natural regeneration but by outplanting nursery-grown stock from genetically improved orchards. Therefore, there is no need for the outplanted stock itself to bear cones, and there could be a very real disadvantage. If a tree produces a limited amount of photosynthate in a year, and reproduction is a high-priority sink for photosynthate, a loss in wood productivity is possible for extremely fecund trees. The present study seeks to examine if the presence or absence of consistent heavy conebearing has an effect on the wood traits of lodgepole pine (Pinus contorta Dough), notably the relative density, yearly volume and yearly stem biomass components. If there is a negative effect on these qualities that is related to high fecundity, this should be considered when establishing roguing criteria for seed orchards. Chapter 2 Background A plant can be viewed as a network of competing sinks; those sinks that are the largest and most active physiologically at a particular time attract the most photosynthate and grow most rapidly (Dickmann, 1985). During the course of a year, photosynthate may be allocated to foliage, branches, stems, reproductive structures, large roots, fine roots with associated root structures, respiration, and storage reserves. The proportion allocated to these sinks can vary with species, site, and genotype, as well as year-to-year climatic variation (Cannell, 1985). For the reproductive sink, photosynthate utilization begins with flower bud initiation. In many temperate tree species, flower primordia are formed early in the growing season preceding the spring in which the flowers appear. The critical period appears to lie between early May and the end of July. Flower bud initiation seems to be dependent on external factors such as light, temperature, water supply and nutrition (Matthews 1963). Certain conditions seem to promote abundant flowering. Ross and Pharis (1985) reported that abundant flowering is associated with factors that either promote a high rate of photosynthesis, encourage an accumulation of assimilates within the shoot, or check vegetative growth. Maguire (1956) noted that higher than average mean monthly temperatures in April and May was followed by profuse flowering the following spring. The reverse was also true. A reduction in water supply can also be associated with flower-bud formation (Matthews 1963). The same warm, dry spells which favor the initiation of flower buds and the accumulation of stored food also favor the maturing of the fruit 5 Chapter 2. Background 6 and seed (Matthews 1963). During the period of cone enlargement, a large amount of the current photosynthate produced may be diverted into growing cones (Kramer and Kozlowski, 1979.) Dickmann and Kozlowski (1968, 1970) reported a preferential mobilization of currently produced carbohydrates by reproductive tissues of red pine (Pinus resinosa Ait.). Reserve carbo-hydrate stores may also be mobilized. Cone enlargement occurred mainly in the second year of development, beginning with the onset of growth in the spring and.reaching max-imum size about one month before maximum dry weight was recorded. The period of greatest enlargement was in the month of June. By contrast, at the end of the previous growing season, the cones were only about one-fortieth the weight, one- thirtieth the volume, and one-third the length of mature cones (Dickmann and Kozlowski, 1969). A truism of crop physiology states that during crop evolution and domestication, yield has been increased mainly by increasing the proportion of assimilates partitioned to the harvested parts of the plants, and much less (or not at all) by increasing total biomass production (Evans, 1976). Forest tree breeders are generally interested in increasing the harvestable portion of a tree, which in most cases is the stem alone. In this goal there may be a conflict with the competing sink of reproductive structures. 2.1 P h e n o t y p i c evidence In 1951, Morris found that heavy flower production in balsam fir (Abies balsamea Mill.) was consistently attended by a conspicuous reduction in the size and quantity of the current season's foliage. When it was unaffected by abnormal weather conditions, radial growth was reduced in years with plentiful seed production. Morris concluded that in the great majority of cases, the growth of flowering trees shows depressions corresponding to the heavy seed years. Chapter 2. Background 7 The radial growth loss seen by Morris was borne out by Holmsgaard in 1956. He found that the ring width of beech (Fagus sylvatica L.) over 100 years of age was greatly reduced by the bearing of seed. On average, trees of greater than 130 years of age had ring widths only half the size in seed-bearing years as they were in non-bearing years. Younger stands of 60 to 100 years of age showed a 25 to 30% decrease in ring width. Eis et al. (1965) found that good cone producers were consistently good, while poor cone producers were consistently poor in stands of Douglas-fir, (Pseudotsuga menziesii (Mirb.) Franco), grand fir (Abies grandis Lindl.) and western white pine (Pinus monti-cola D. Don). In 1969, Tappeiner reported that there was a conspicuous reduction of shoot and needle growth in young-growth Douglas-fir during years of cone production. This confirms Morris' (1951) claim that foliage of the current year is greatly reduced by conebearing. Tappeiner went on to show that in two heavy cone years, the ring width in the cone-bearing Douglas-firs decreased while remaining the same or increasing in the non-bearing Douglas-firs. In a poor seed year shortly thereafter, the ring width in both cone-bearing and non-bearing Douglas-firs appeared to increase slightly. Other workers in the field (Messer 1956; Eklund 1957; Gross 1972) have supported the claim that increases in seed and cone production can cause reductions in diameter increment and foliage in forest tree species. Eis et al. (1965) and El-Kassaby et al. (1988) both found that a good cone year was followed by one or two "off" years to replenish carbohydrate reserves. This was supported by Matthews (1963), who observed that the seed crop of one year also affects the seed crops of the following year. This provides further evidence of the depletion of stored nutrients and loss of foliage that accompany seed bearing. Chapter 2. Background 8 2.2 Genetic evidence In 1960, Fielding found that in Monterey pine (Pinus radiata D. Don), there was an inverse genetic correlation between heaviness of flowering and inherent growth and vigor. He estimated that, on a medium-quality site over a rotation of 40 years, the total dry weight of materials expended annually on pollen, cones and seed was about 1,000 pounds (454 kg) per acre, or about 33 cubic feet of wood per acre per year (3 cubic metres per hectare). This is approximately 16% of the mean annual increment per acre. Cannell (1985) estimated that the figure may be less than 10% of the above-ground dry matter increment, although in heavy cone years it may be higher. Cannell did not, however, estimate pollen production. Fielding observed that, as seed and pollen are richer than wood in nutrients, the amount of growth energy annually expended on reproduction might be substantially higher than 16%. In order to divert a greater proportion of photosynthate into wood production, Fielding suggested that workers should select against rather than for profuse flowering, a proposal that Cannell supported. Ying et al. (1985) confirmed Eis' 1965 observation that good cone producers were consistently good and poor cone producers consistently poor. He found that clones that were good cone producers in lodgepole pine (Pinus contorta Dougl.) consistently tended to produce more cones year after year. These genetic differences in cone production among genotypes have been reported by others (Varnell et al. 1967; Bramlett and Be-langer 1976; Schmidtling 1981; Griffin 1982; Schmidthng 1983; Ying et al. 1985; Byram et al. 1986; Schoen et al. 1986; El-Kassaby et al. 1988). Schmidtling found in 1981 that flowering traits were negatively correlated with growth traits in loblolly pine (Pinus taeda L.). The flowering traits studied were precocity and fruitfulness, while the growth traits were height, crown width and diameter at breast height. Schmidtling suggested that the uniformly negative correlations indicate that Chapter 2. Background 9 selection based on fruitfulness or precocity alone would result in some loss in growth. In addition, Schmidtling reported that precocity and fruitfulness were very heritable, and that there were significant differences between clones. The average number of flowers, a quantitative trait, was highly heritable both on an individual tree basis and on a family basis, with heritabilities of 0.61 and 0.63, respectively. This was in good agreement with earlier work (Bramlett and Belanger 1976, Schmidtling 1980). Precocity was less heritable. When Schmidtling examined the progeny of the most fruitful parents, he found that growth traits in the progeny were negatively correlated. This work supported findings by Polk (1966) and Holmsgaard (1972). The d.b.h. and height at different years were exam-ined for the progeny, and the correlation between these traits and parental fruitfulness was almost uniformly, albeit weakly, negative. Schmidtling suggests that, although the magnitude of these correlations is not very important, they reinforce the caution against selection or roguing against clones with low fruitfulness without considering other as-pects, since there may be a tendency for the most fruitful parents to produce slower growing progeny. • > In a later paper, Schmidtling (1983) attributed 50% of the variation in female flow-ering and 40% of the variation in male flowering to genetic effects. The trees used in this study were grafts of loblolly pine. Broad-sense heritability for female strobilus pro-duction averaged around 0.50, but there was variation by year and age of the ramets. There was not a strong correlation between male and female flowering patterns on the same clones. The rootstocks used contributed substantially to the variation in flowering patterns, however. This rootstock variation was measured in the experimental design and contributed to the within-clone variation. In Schmidtling's study, the relative ranking of clones by cone production changed from year to year, which he suggested may serve to dispel the parental imbalance problem Chapter 2. Background 10 reported by North Carolina State University (1976). This paper showed that due to highly variable fecundity in seed orchards, 80% of seed collected from an orchard may be produced by only 20% of the clones represented. This situation encourages artificial selection for highly fecund clones, however, since they represent a much greater proportion of the seed lot from an orchard. Danbury (1971) has argued that selection based.on fecundity is a good idea, as it would aid seed managers in cutting costs by allowing them to rogue orchards of clones that were poor cone crops. However, in light of Schmidtling's results, a loss of potential growth increment could be expected. El-Kassaby (in press) confirmed that large family differences in cone crops were caus-ing over- and under-representation of specific families in the seed crop. He suggested that the "80/20" ratio of North Carolina State (Anon. 1976) was a very conservative estimate for the parental imbalance observed. In a poor cone year, the grafted Douglas-fir orchard El-Kassaby studied had a 80/6 ratio, while in a good year, the ratio improved to 88/50. In light of these findings, Romberger's 1967 characterization of the ideal tree as one that has few or no cones in the forest, but that can be induced to provide many cones in a seed orchard, has promise. Wild-stand selection for trees with sparse cone crops can be coupled with advanced methods for cone induction (see Ross and Pharis 1985) to yield satisfactory cone crops. 2.3 W o o d traits To a seed orchard manager, highly fruitful trees are desirable. However, the purpose of genetic improvement practices is to improve the end products, not the intermediates. Forest trees are useful because of their value as producers of wood and wood products. Tree breeders, therefore, select trees that show phenotypic properties associated with increases in volume, rate of growth, and other wood traits. Chapter 2. Background 11 Among the many traits that characterize wood, relative density has the greatest influence on both lumber and pulp production. Within existing structural grades of lumber, strength and stiffness are positively correlated with relative density (Barrett &: Kellogg 1984). Digester yields in chemical pulp manufacturing are also positively correlated to increases in relative density (Kellogg 1982; Namkoong et al. 1969). Relative density is highly heritable (Jefferson 1984). Kellogg (1978) suggests that relative density should be incorporated into tree improvement programs. The Coastal Tree Improvement Council of British Columbia has a stated goal of maintaining relative density levels while increasing the volume and growth rate of coniferous species (Anon. 1989). There have been no studies published which examine the effect of cone production on relative density. The current study examines this possible relationship by using X-ray densitometry to analyze the earlywood and latewood components of wood samples from a seed orchard. Chapter 3 Materials and Methods 3.1 Background of Omineca-Pinchi seed orchard In 1970, the Swedish Cellulose Company (SCA) undertook a project to collect selected lodgepole pines in British Columbia and the Yukon to form the basis of a breeding program in Sweden. Lodgepole pines were chosen for several reasons. The species is highly adaptive, able to grow well in many different microsites. This was seen as desirable as it would be widely planted in Sweden. Another positive attribute is the species' ability to withstand harsh winters with long periods of freezing temperatures. Since increased volume per hectare was considered a major objective in the project, the fact that lodgepole pine grows well in pure, high-density stands was attractive (Engleson, 1971). The SCA chose to use individual or "plus" tree selections, and made an agreement with the Faculty of Forestry at the University of British Columbia (UBC) to carry out the selections. Provenance trials of British Columbian lodgepole pines had previously been carried out in Sweden. The provenances that showed the best performance in Sweden came from the between 56° and 63°30' in central British Columbia. Therefore, selection efforts were concentrated on the area between the 58th and 61st latitudes, and between 120 and 130 degrees in longitude. The longitudinal, boundary ensured that all the trees chosen were from the interior rather than coastal areas, as the coastal provenances were less hardy to Swedish weather conditions and showed poor survival. 12 Chapter 3. Materials and Methods 13 In general, selected stands were at least 10 miles apart to minimize inbreeding. Plus trees were selected from stands that met five basic criteria outlined by the SCA (Hagner, 1970): 1. The trees on average are of good phenotype. 2. The stand is clean, of good density and uniform age. 3. The stand should preferably grow in mineral soil, and the site quality should be above the mean for the area in question. The stand should preferably be from 50 to 120 years old. 4. Stands with trees attacked by dwarf mistletoe (Arceuthobium americanum Nutt.), by western gall rust (Endocronartiurn harknessii Hirat.), or other parasites, must be avoided. 5. The stand must be situated close to a road that is kept open in winter. The three primary objectives for the Swedish program were increased volume, good seed production, and resistance to disease. Accordingly, the Plus Tree Rating Cards numerically rated the trees on height, diameter at breast height, straightness, crown form, branch thickness, cone production, and rust occurrence. Height and diameter at breast height were measured, while the other traits were assigned a subjective value from one to five. The trees chosen were compared to three surrounding dominant trees that were within 66 feet. These dominants were to be close in age to the chosen tree, with the plus tree clearly superior. As good seed production was a priority, trees with heavy cone crops received high scores. However, relative density was not one of the criteria used in selecting these trees, as the importance of the trait was not fully recognized at that time. Chapter 3. Materials and Methods 14 The trees that were ultimately chosen were marked with spray paint and flagging and left until late fall, when they had presumably entered dormancy. Each tree was felled to permit easy collection of scions, as tree climbing or shooting the scions was considered to be too time-consuming. A total of 180 scions were collected from the upper third of each tree, frozen, and shipped to UBC. Later, 70 scions from each tree were sent frozen to Sweden, where they were grafted onto potted grafting stock of lodgepole pine in 1971-1972. A short time later, they were outplanted into seed orchards at Norrberge, Logdo, Sor-Nedansjo and Galtstrom, all in the vicinity of Sundsvall, Sweden. Where possible, 200 cones were collected at the same time that the scions were col-lected. Not all trees had that many cones, however. The seed was extracted and samples were sent to Sweden for progeny testing. Wood samples in the form of two two-inch disks were collected as well, to be studied at UBC. The Stora Kopparberg (SK) Company of Sweden did a similar selection process during the summer of 1971. The British Columbian trees that they chose were from a more southern area than the SCA selections, between latitudes 54° and 56° and longitudes 124° and 124°30'. The elevation range was betwen 755 and 1035 metres (2475 and 3400 feet). Because of the narrow geographic band in which to make selections, stands were ~ chosen that were less than ten miles apart. Scion materials from the SK selections were grafted onto rootstocks in the spring of 1972! These grafts were planted in a clonebank at Red Rock nursery in August of that same year. The ramets that were left over from the clonebank were held until 1974, when they were planted in two seed orchards at the Prince George Tree Improvement Station, across the Fraser River from Red Rock. The eastern selections were established as the Dawson-Peace seed orchard, while the western selections became the Omineca-Pinchi seed orchard (OPSO). The Omineca-Pinchi orchard's stated purpose is to supply regular quantities of quality Chapter 3. Materials and Methods 15 seed for reforestation. The seed from the orchard is suitable for use in the Central Plateau seed orchard planning zone, at elevational ranges between 650 and 1100 metres. There are thirty-six clones in the orchard, with two ramets of each clone represented in each of nine blocks. The layout is a Modified Random Block, with none of the ramets of the same clone adjacent in the same block. The trees are planted at a spacing of 4 x 8 metres. The total size of the orchard is 4.4 hectares. 3.2 Sample collection In three of the eight blocks of the OPSO, cone counts were done every fall. Cones were collected and counted by tree on all of the living original grafts, beginning in 1980 when the orchard was six years old. Cone counts continued throughout the eighties with the exception of 1986. In May of 1988, wood samples were taken from each of the trees in the three blocks with cone count data. The method of increment core extraction was as described by Josza (1988). A hand-held 10-inch increment borer was used to obtain 5 millimetre samples. In order better to estimate volume and biomass, two opposite increment cores were taken from pith to bark at three locations on" the bole. Gonzalez (1989) has shown that two samples taken opposite each other at breast height will give a good approximation of mean density at that height. The lowest core was taken as close to the bottom of the tree as possible, above the graft union. This height was about 25 centimetres above ground. The central core was taken at breast height, defined as 138 centimetres above ground. The top core was taken at half the distance between the central core and the top of the tree. Each core was numbered with indelible pencil, and broken cores were taped together. At the same time, the tree height was recorded and the general vigor of the tree was noted. Chapter 3. Materials and Methods 16 At Forintek Canada Corporation in Vancouver, samples were prepared for X-ray densitometry. The cores were sawn into strips exactly 1.57 millimetres thick by a custom-built hydraulic saw (Figure 3.1). The cores were then renumbered directly on the sawn surface with indelible pencil. Broken cores were glued with polyvinyl acetate glue, using cross-hatches and numbering both sides to aid in rematching. Extraneous wood chemicals had to be removed from the wood because they absorb X-radiation at a different rate from the wood itself. If permitted to remain, incorrect density readings would occur when X-ray densitometry was done. The cores were extracted in a soxholet apparatus for 24 hours in a 2:1 ethanol/cyclohexane solution. This was followed by a 24-hour extraction in hot water. The extractive-free cores were covered with heavy objects to keep them from warping and permitted to dry. After drying, the cores were dated by assigning a calendar year to each ring, starting with the last complete ring beneath the bark surface and working inward to the pith. On cores without the pith present, the center was estimated. This was necessary for providing accurate basal area measurements to estimate radius and volume. The center is estimated by matching the rings present on the core to a target-like drawing, which shows where the pith should be. Once the rings are matched, the distance to the probable center location can be measured. To reduce the effort in densitometry, 12 of the 36 clones were chosen for analysis. The 12 were chosen by ranking all 36 clones on cone production over all the collection years, and choosing the four top cone producers, four bottom producers, and four intermediate producers. The remainder of the wood samples were stored for future reference. Chapter 3. Materials and Methods 18 3.3 X - r a y dens i tometry Initially, the densitometer is calibrated by reading a five-stepped Delrin plastic wedge. which is attached to the sample holder. The wedge steps were machined to the following nominal thicknesses: 0.25, 0.50, 0.75, 1.00, and 1.25 mm. These steps represent wood density equivalents from 0.10 to 1.00, respectively. This phase of the program automat-ically calculated the relationship between X-ray transmission (read by the detector and expressed in volts) and density equivalents for each step of the wedge (using previously determined densities contained in a data file for the plastic wedge). Once the relationship between current voltage readings and density valued is established, it allows immediate conversion of voltage values to density values during the sample scanning process. X-ray densitometry determines the ring density of the wood by comparing the ab-sorbtion of the X-radiation into the core sample to the calibrated Delrin plastic wedge. At the same time, it measures the width of the ring. The core is "read" from the pith outwards. As each core was taken completely through the tree, two sample readings could be taken from each core, going in opposite directions. Each core was clamped in the X-ray chamber and moved slowly in front of the X-ray emitter (Figures 3.2, 3.3). The core movement could be watched on a small T V monitor. As the X-ray emitter starts across a ring, the machine considers the wood to be "earlywood." The width and the average density of the earlywood section is recorded. When the ring density of the wood increases to a preset value (in this case .49 g/cc), the densitometer notes this as "latewood" and measures its width and average density separately. After the density falls below another preset value (in this case .40 g/cc), the densitometer is triggered to begin the next year's earlywood. The densitometer produces a density profile of the core sample (Figure 3.4). It is important to note that the X-ray densitometer's definition of latewood is not tree-ring sample-density calibration wedgex sample holder-precision linear motion table X-ray detector assembly ^stepping motor P Co 3. EL B a. o Cn Figure 3.2: Diagram of X-ray densitometer for scanning wood samples. Chapter 3. Materials and Methods 21 Figure 3.4: Wood density profile, showing early and latewood components of ring density. Density is on the Y-axiB in g/cc and calendar year is on the X-axis. Division between early and latewood is triggered at 0.49 g/cc. Chapter 3. Materials and Methods 22 the same as the definition of Mork (1928). Mork's definition is almost universally used because of its simplicity; latewood starts in an annual ring at a point where the double width of two adjacent cell walls equals or surpasses the radial width of the lumen. Evert-seh (1982) assigned earlywood/latewood boundaries to 84 annual rings, then measured the relative density at these points. There was significant variation, which indicates that a fixed boundary such as Mork's may not be valid. As X-ray densitometry is an auto-mated process, having it triggered at a preset latewood value lends consistency and speed to the data collection process. The densitometer output for each ring contains the distance from the pith, the total ring width, the width of the earlywood, the width of the latewood, the average density of the whole ring, the average density of the earlywood and the latewood, and the min-imum and maximum densities within the ring. This information is stored directly on a microcomputer, eliminating the tedious step of data entry. 3.4 Transformation of densitometer output After the densitometry was completed and the data stored, a computer program was developed to calculate the yearly volume increment, weight of bolewood, and average density of each individual tree. This program was later modified to give these values for their early and latewood components. To compute the volume, the program visualizes the tree as a series of stacked conical frustrums, sitting on a cylinder and topped with a cone (Figure 3.5). The sampling points where increment cores were taken demarcate the sections of the trunk. It is useful to think of a year's growth as a "sheath" of wood around the core of wood built up in previous years. To compute the volume of the sheath, the total volume up until the end of that year's growing season is calculated. Then the total volume up until the end of Chapter 3. Materials and Methods 23 the previous year's growing season is subtracted, leaving the volume of the sheath. The bottom increment core is used to calculate the volume of the bottom cylinder. There is no measurement of the radius of the tree at the root collar, so the program assumes it to be the same as that measured at the first sampling point. It is acknowledged that this will lead to minor underestimation of the volume of the bottom, as most trees will have a certain amount of flare near the root collar. The program uses the function for the volume of a cylinder, which is: volume = 7r * radius2 * height (31) Above the bottom cylinder, the tree trunk is a series of stacked conical frustrums. The computer program uses the radii of the tree at the top and the bottom of the frustrum to calculate the volume. Using the formula: 7T * height * (radius2 + radius2, + radiusi * radius2) . . volume = * - (3.2) Above the top core, the tree's growth is a conical shape. The computer program uses the radius of the tree at the top core and the estimated height of the tree for that year to estimate the volume of this cone, using the formula: 7T * height * radius2 volume = (3-3) 3 To calculate the volume grown in each year prior to 1987, it is necessary to estimate how tall the tree would have been at that time. The estimation is done by assigning the top of the tree a pith year of 1987. To find the tree height for a given year, each core is checked from the bottom up until a core is found with a pith year greater or equal to the year in question. The pith year and height of this core is then subtracted from the pith year and height of the core below to give a height increase and the number of years Chapter 3. Materials and Methods 24 Number of Year of annual rings internode A, formation^ Figure 3.5: Diagram of tree growth by year, with the sheath of wood illustrated. Chapter 3. Materials and Methods 25 over which it occured. The height increase is divided by the number of years to give the growth per year. The pith year of the lower core is also subtracted from the year in question and the result multiplied by the growth per year to give the estimated height above the lower core. Adding the height of the lower core gives the total estimated tree height. For example, for the year 1982, the program might find the pith years in the bottom core and the breast height core to be less than 1982. If the pith year of the upper core is 1983, the tree height in 1982 fell somewhere between the the breast height core and the upper core. If the pith year of the breast height core was 1979, subtracting the pith year of the breast height core from.the pith year of the upper core gives a difference of 4 years. If the height of the upper core was 278 cm, subtracting the height of the breast height core (138 cm) yields a frustrum height of 140 cm. Dividing this number by 4 (1983 -1979) yields a yearly growth estimate of 35 cm. The growth in the three years between 1979 and 1982 is then estimated to be 105 cm (3 years x 35 cm) above the breast height core. The total estimated height in 1982 is then 138 + 105 = 243 centimetres. After the volumes for the cylinder, the frustrums, and the cone have been calculated, the process is repeated to find the volume of these shapes at the end of the previous year. Subtracting the volume of the previous year gives the volume of the sheath grown in a single year for each shape. Then the volumes of the cylinder sheath, the frustrum sheaths, and the cone sheath are added to give the volume growth for a single year. The computer program also has a function to give an average density for a year's growth over the whole tree. This is done by taking each of the core values for density and simply averaging them. Although this average density value was used in the early part of the analysis, it was decided to substitute the average density at breast height only. The reason for this decision had to do with the perceived error of averaging the densities over such young trees. Over the first 15 to 20 years, the pattern of ring density Chapter 3. Materials and Methods 26 from pith to bark in lodgepole pine changes rapidly, first declining and then increasing. It was believed that, by taking cores from the bottom, middle and top of the tree, each core would fall in a different place in the juvenile density curve, according to its level of maturity. Thus, averaging these values would not in any way approximate what was really occurring in the stem of the tree. Although the increment core taken at the base of each tree had the greatest number of years represented, it was decided to use the breast height core to avoid any reaction wood that might surround the graft union, or compression wood associated with basal sweep. To estimate the weight of the yearly bolewood growth, an extension to the volume calculation procedure was done. For any given year, the volume of each geometric shape was calculated as above. The program then retrieves the density data for that year from the densitometer output. Looking at each shape individually, the program multiplies the volume of the sheath by the density of the cores surrounding it. For instance, to find the bolewood weight growth for 1984, the program would go through several steps. First, the volume of the bottom cylinder sheath would be calcu-lated. Then the density estimate for the 1984 annual ring in the bottom increment core would be multiplied by the volume of the cylinder sheath to yield the cylinder sheath weight. Next, the volume for the frustrum sheath between the bottom and breast height sampling points would be calculated. The densitometer data for 1984 from both the bottom and breast height cores would be retrieved and averaged, and that figure multi-plied by the frustrum sheath volume to yield the frustrum sheath weight. The process would be repeated for the next higher frustrum sheath, using the breast height and up-per core data. Lastly, the cone sheath volume would be calculated and multiplied by the 1984 density of the upper core, to yield the cone sheath weight. All these weights would then be added to give the total weight of the 1984 sheath. After the volume and bolewood weight were computed for each individual tree, the files were transferred onto Chapter 3. Materials and Methods 27 the mainframe computer at UBC. 3.5 Data analysis The main goal of the project was to quantify the relationship between wood traits and cone counts. Wood and cone data from the orchard trees, weather data, and data from a progeny trial were used. These datasets were initially analyzed individually to find the variance. Then, linear regressions were calculated between the wood traits and cone counts, first using individual- tree data and then clonal means. Lastly, linear regressions were performed between the wood traits of the progeny and the cone counts of the same OPSO clones (Figure 3.6). 3.5.1 Data sources The data sources included the OPSO cone counts from 1980-1987, the densitometer output of the 12 selected OPSO clones, weather data from Prince George, and data from a Swedish progeny trial of the original British Columbian trees. The densitometer data were transformed to give the yearly volume and biomass increment. Cone counts. The number of female cones were recorded for each tree in three of the OPSO's eight blocks. These counts were made each year between 1980 and 1987, with the exception of 1986. Cones were collected in the late fall, before snowfall. This data set included the clone number, block, and position within the block as well as the number of female cones collected. Wood traits. Once the densitometer data were brought from Forintek, two separate databases were constructed. One contained the transformed densitometer data, which included the calculated volume and biomass values. A different data set was used for the ring density analysis. As explained eariier (see section 3.4), the decision was made to Weather Data Cone Counts Wood Data Progeny Data Analysis of Variance Dataset #1 Individual Tree Data Dataset #2 Clonal Means Simple Regressions Between Cone Counts and Weather Simple Regressions Between Wood Traits and Cone Counts Dataset # 3 Cone Production Level Means ./V Simple Regressions Between Progeny Wood Traits and Family Cone Production Patterns Multiple Regressions Between Wood Traits and Cone Counts, Age and Weather Trends Over Time Relationship Between Wood Traits and Cone Production Chapter 3. Materials and Methods 29 not use the average densities calculated by the computer program, except in estimating bolewood weight. Therefore, the original densitometer data for breast height were used for the ring density calculations. Later, the program to compute the volume, density and bolewood biomass was improved to give the earlywood and latewood components of each of these qualities, then analysis was done using the new data set. The database used for volume and bolewood weight analysis held all the wood trait and cone production values for each ramet of the 12 study clones. The variables in the data file were clone number, ramet number, block number, year, number of cones pro-duced, earlywood volume, latewood volume, total volume, earlywood weight, latewood weight, total weight, earlywood density, latewood density, and total density. A sample of these data is included in Appendix A. The ring density database variables were the clone number, ramet number, block number, year, number of cones, RSX (a measure of the distance from the pith to the beginning of the ring), earlywood ring width, latewood ring width, total ring width, ear-lywood density, latewood density, overall ring density, and the minimum and maximum densities in the ring. A sample of these data is included in Appendix B. W e a t h e r . Precipitation amounts and temperature means were available from two weather stations near the Prince . George Tree Improvement Station. These locations were at the Prince George airport (Lat. 53°53'N Long. 122°40'W, Alt. 691 m) and at a location on the south end of the city (Lat. 53°53'N Long. 122°46'W, Alt. 579 m). The two weather stations were approximately equidistant from the tree improvement station to the north. Data were used for the months of May through July, the period of earlywood growth, cone enlargement and strobilus initiation. Data from the two stations were averaged, and means were calculated by month and by year. The weather data are available in Appendix C. Chapter 3. Materials and Methods 30 Swedish progeny. After the Stora Kopparberg (SK) scions and seeds arrived in Swe-den, progeny trials were established at Dala Jama (latitude 60°32'N, longitude 14°35'E, altitude 235 m). In addition, a seedling seed orchard, Sor Amsberg, was established at the SK nursery in Borlange (latitude 60°32'N, longitude 15°32'E, altitude 155 m). The seedlings for both the seed orchard and the progeny trial were sown in 1973 and raised together until outplanting in 1975. Heights were measured in the progeny trial and the seedling seed orchard in late 1981; these heights, as well as estimates of survival, damage, and crown form, were used to establish roguing criteria for the OPSO. In early 1983, when the trees were ten years old, between 9 and 15 increment cores were taken at breast height (1.3 m) as wood samples from each family in the seed orchard at Sor Amsberg. These cores were analyzed to find family means for volume, ring width, relative density, and biomass (Fries 1986). Data from these trees are available in Appendix D. 3.5.2 Preliminary analysis Statistical analysis was done throughout the project using the Statistical Analysis System (SAS Institute, 1985). This package was used because of its versatility, availability, and reputation for widespread use. Individual-tree data was available from 167 trees in the OPSO. These data were ana-lyzed initially without grouping by clone, to assess the total variation and to investigate the overall relationships between wood traits and cone counts. As the criteria for roguing seed orchards are based on clonal means rather than individual-tree data, it was neces-sary to do analysis by clone. Clonal means by year were calculated for the earlywood, latewood, and total ring components of volume increment, bolewood weight increment, and ring density. The clonal mean by year was also calculated for cone crop. Clonal means were derived from a minimum of three ramets to a maximum of six ramets per Chapter 3. Materials and Methods 31 clone. In only one case was a clone represented by three ramets; in most cases, five or six ramets were present in the orchard. A fundamental problem with the OPSO data is that the trees used are the same from year to year. For instance, cone crop data for clone 102 in position B2 in block 3 were taken for each year from 1980 through 1987. This situation creates serial correlation, and thus violates one of the assumptions of regression analysis. However, linear regressions of the individual-tree data were performed, followed by regressions on the clonal mean data. The regressions on the clonal mean data were calculated solely as a guide to breeders and others interested in overall family performance. The serial correlation has an effect on several regression statistics generated by the data sets. The coefficient of determination (r2) can be used as a descriptive statistic, but the square root (r) is not an unbiased estimate of the population correlation coefficient (p). In addition, the regression coefficients (b0,bi, etc) are unbiased estimates of the re-gression parameters (L3Q,L3\, etc). The standard error of the estimate, like the coefficient of determination, is acceptable as a descriptive statistic. However, as a measure of the variances of the observations (cr2), it is biased low. F-tests are also biased, and overes-timate the significance of the models by giving a smaller value than would be accurate for the significance level. For example, a significance level of .0001, or 99.9% probability, would actually be somewhat higher (which would make the probability lower). Never-theless, the significance of the models can still be discussed. A model with an estimate of 99.9% probability would probably remain significant if the serial correlation could be removed, although the probability would be lower. Multiple regression models were fitted using stepwise regression, which included variables in the model if they were significant at the 0.1500 level (85% probability). To compensate partially for the serial correlation, only variables that were significant at the .0001 level were included in the final model. Analysis of variance .was performed on the OPSO cone counts and the volume, biomass Chapter 3. Materials and Methods 32 increment and ring density from the densitometer output. This was done to establish where the variance was located in the data. A certain amount of variance between clones would permit selection by clones if this was justified in light of the overall results. In ad-dition, discovering the strength of the other variables, as well as what interactions might exist between clones, blocks, and years would determine the degree to which conclusions about clones might be drawn. The linear model for the ANOVA used the formula below, where p represents the overall mean, C{ represents the ith clone, Bj the jth block, and Yfc the A;th year. The interactions between the effects are also part of this model. As the clones were chosen from the OPSO population because of their cone production patterns, they were treated as a fixed effect. The autocorrelation from year to year of the trees renders the effect of years likewise fixed. In the expected mean squares, clones and years are therefore fixed, while blocks and ramets are random. There are minor variations in the degrees of freedom associated with years from trait to trait. This is because the data collection period varies; for instance, cone collection data is available for only seven years, while volume is available for 12 years. The effects, degrees of freedom, and expected mean squares of these analyses of variance are listed in Table 3.1. The fixed effects have a <f> rather than a cr to reflect this fact. From the expected mean squares, .F-tests of the effects were developed and then calculated for each wood trait and for cone count. The tests of the effects were: Xijkl = fJ, + Ci + Bj + (C*B)ij+Yk + (C*Y)ik +(B * Y)jk +(C*B* Y)ijk + (3.4) Clone Clone j C * B (3.5) Chapter 3. Materials and Methods 33 Effect D.F. Expected mean squares Clone 11 <r\ + nycr2CB + nyhcbc Block 2 <T\ + nyccrl C*B 20 <r| + nyo-2CB Year 6-11 oE + ncaBY + nbc(f)Y C * Y 62-104 <J2E + nbficY B*Y 12-20 <T2E + nc<TBY Exp. error 110-162 o\ + ncr2CBY Sampling error 163-208 Table 3.1: Format for analysis of variance. Block Block/Sampling error (3-6) Clone by Block C * BjSampling error (3.7) Year Year/B * Y (3.8) Clone by Year C * Yj Sampling error (3-9) Block by Year B * Y/Sampling error (3.10) Experimental error Exp. error/Sampling error (3-H) 3.5.3 W e a t h e r analysis The precipitation and temperature patterns for the years 1979-1987 were examined to' find the relationship between these variables and either cone production or wood qual-ity. In particular, the relationship between weather and cone production patterns was examined. Temperature and precipitation were regressed against cone production of the same year. In addition, both a one- year and a two-year lag of the weather data against cone production measured the relationship of weather with strobilus initiation (two years prior to harvest) and cone pollination (one year prior to harvest). Chapter 3. Materials and Methods 34 3.5.4 Individual-tree analysis As the main goal was to quantify the relationship between wood traits and cone counts, regression or correlation analysis was warranted. Regression analysis was used as it was suspected that the wood traits were related to cone count. A simple relationship between wood traits and cone counts was initially sought. Simple linear regressions between cone counts and wood traits were calculated. Vol-ume, biomass, and ring density were broken down into their earlywood, latewood and year-total components. Each of these components was regressed against cone count sep-arately. The linear model used was: Yi = fa + f3x(Cone county + e{ (3.12) where i represented the sample number from 1 to N; Y{ was either volume, biomass, or ring density; /3 0 and f3i were regression parameters; and e; was the error term. Parameters were estimated using least squares regression. Multiple regressions were performed in an attempt to improve the fit of the model and to discover to what extent other variables were influencing the wood traits. Besides cone count, variables investigated included the age of the tree, the amount of precipitation in May-July, the temperature in May-July, and the inverse of the" age. The inverse of age was included because a curvilinear relationship with age was indicated in some of the graphs of wood traits. Stepwise regression was used to find the best model for each component of volume, biomass and ring density. The full model with all variables was the following: Yi = Po + Pi{Cone county + f32(Age)i + /33(1/Age)i +(3^(Precipitation ); + (35(Temperature ); + €{ (3.13) where f30toB5 are regression parameters, and the other values are as described above Chapter 3. Materials and Methods 35 (Equation 3.12). 3.5.5 Clonal means Simple linear regressions were calculated using clonal means for volume, biomass, and ring density against cone count. The linear equation used for these simple regressions was the same as Equation 3.12, above, with the exception that the observations for the dependent and some of the independent variables were mean values by clone. Similarly, multiple regressions were performed for clonal averages using the same linear model as Equation 3.13, above. These regressions included the May-July weather data and ramet age to discover which other variables were having an impact on wood traits. Once again, the observations were clonal mean values. Since all ramets of a clone had the same age, the average age of the clone was equal to the age of any tree. Similarly, all ramets of a clone experienced the same precipitation and temperature, so these values are the same for each tree. After calculating the regressions, the relationships over time were plotted. Clonal means for each wood trait in a given year were plotted, as well as clonal means for cone counts. This permitted visual comparison of the clonal trends in each trait, and the variation by year that could be associated with conflicting demands on the trees' resources. Then, means by cone production level were shown, where the 12 clones were divided into high, intermediate and low conebearing levels. 3.5.6 Progeny data The Swedish data from the Sor Amsberg seedling seed orchard were used to establish the possible impact that highly fecund parent trees may have on their progeny. Simple linear regressions used progeny height, volume, ring width, relative density, and dry-stem biomass as dependent variables and cone production means of the same OPSO clones Chapter 3. Materials and Methods 36 as the independent variable. The same linear model as above (Equation 3.12) was used, except that the data used were family means rather than observations from individual trees. Chapter 4 Results 4.1 Data parameters The Omineca-Pinchi seed orchard experienced 23% mortality. Of the original 216 trees in the three cone-counted blocks in the orchard, 167 were alive at the time of the study. The number of ramets alive per clone varied from a low of three to a maximum of six. The trees ranged between 2.7 and 6.7 metres in height. There was very little obvious graft incompatibility in the orchard. Most of the trees had good form and appeared to be growing quickly, although there were a few that had sparse crowns and poor form. Pollen cone production appeared to be substantial. The cone counts and wood trait data for the OPSO are summarized in Table 4.2. Mean values were calculated by clone and by year, and these means were averaged to yield the values in the table. Yearly means and clonal means from which these averages were derived may be found in Appendix E. Data for the volume and bolewood biomass were measured from 1976 to 1987 (12 years), and ring density was measured from 1979 to 1987 (9 years). The cone counts were taken from 1980 through 1987 except 1986 (7 years). Both volume and bolewood biomass have most of their yearly increment in the earlywood. The standard errors of the mean show that volume varies from the regression line by as much as 115 cubic centimeters between clones, but between years, the variation is as much as 385 cubic centimeters. A similar pattern is shown for bolewood biomass. Variation about the regression line between clones can be 35 grams, but the variation around the regression 37 Chapter 4. Results 38 Trait N Average of minimum maximum mean values By clone Cone production- (#) 12 53.37 7.33 117.02 11.43 Volume increment (cc) Earlywood 12 1558.13 864.76 2193.01 105.04 Latewood 12 261.12 156.11 403.00 19.23 Total 12 1819.26 1072.21 2471.12 114.52 Bolewood biomass (gms) Earlywood 12 454.80 269.22 629.36 28.98 Latewood 12 134.07 77.31 203.98 10.47 Total 12 597.52 384.97 782.53 34.98 Density (gms/cc) Earlywood 12 0.2926 0.2717 0.3190 0.0040 Latewood 12 0.5093 0.4570 0.5422 0.0076 Total 12 0.3291 0.2991 0.3658 0.0054 By year Cone production (#) 7 61.50 0.92 126.62 18.29 Volume increment (cc) Earlywood 12 1278.74 38.43 3895.50 384.70 Latewood 12 210.67 9.99 644.53 62.62 Total 12 1489.41 48.87 4436.88 443.87 Bolewood biomass (gms) Earlywood 12 375.52 13.00 1113.00 109.75 Latewood 12 109.02 4.76 328.86 32.40 Total 12 491.51 18.33 1396.26 142.22 Density (gms/cc) • Earlywood 9 0.3007 0.2621 0.3394 0.0096 Latewood 9 0.4938 0.4014 0.5311 0.0143 Total 9 0.3349 0.3046 0.3712 0.0089 Table 4.2: Sample size (N), average of means, range and standard error of the mean (SW) for cone production and wood traits, by clone and by year. Ring density is at breast height. Chapter 4. Results 39 Variable Cone crop Volume Biomass Density r2 .91 .89 .89 .84 24.34 550.55 229.06 0.0195 Clone N.S. ' N.S. Year Clone*Year ** N.S. N.S. Block N.S. ** Block*Year N.S. N.S. N.S. Clone*Block ** Clone*Block*Year N.S. . N.S. N.S. N.S. Table 4.3: Results from analysis of variance for four traits. The fit of the model is measured by r 2 , and Se is the standard error of the estimate. ** = significant at the 99% level, N.S. = not significant at the 5% level. line between years is as much as 143 grams. On the other hand, the variation of ring density around the regression line is only 0.0054 grams per cubic centimeter between clones, and 0.0089 grams per cubic centimeter between years. 4.2 A n a l y s e s of variance Analyses of variance were performed on the cone counts and the wood trait data to establish where the majority of variation was located. The serial correlation of the data set underestimates the probability of a greater F. For this reason, only model components that have an extremely low significance level are included in Table 4.3. It was assumed that, although the significance level is underestimated, the variables that are significant with 99% probability in this analysis would probably remain significant if the serial correlation could be removed. The r2 values show how much of the variance is explained by the model for the analysis. The standard error of the estimate (5e) is the square root of the sum of squares of the sampling error divided by the error degrees of freedom. These analyses of variance showed that clones were an important effect in the variance Chapter 4. Results 40 of cone crop and relative density, but not in the variance of volume or biomass increment. Years were an important effect to all four traits. There was a substantial clone by year interaction in cone crop and density, but not in volume or biomass increment. Blocks were important to every trait except cone crop, while block by year interaction was only important for density. There was a substantial clone by block interaction, which was shown by all four traits. Therefore, analyses of variance were done for individual years in order to discover if the effect occurred every year, or only occasionally. None of these individual-year analyses, however, showed the clone by block interaction to be important. 4.3 Weather data Stepwise multiple regressions were calculated using cone crop as the dependent variable and May-June temperature and precipitation as independent variables. The cone counts were first regressed against weather data of the same year, then regressions were calcu-lated with cone production against both a one-year and a two-year lag of the weather data. In all cases, the regression model had extremely low r 2 values, from r2 = .16 for same-year data to r2 = .02 and .08 for the one-year and two-year lags, respectively. The linear model used to describe the relationship between same-year cone count and weather data was: Cone count = 145.93 — 1.58(Precipitation ) (4-14) The term Cone count refers to predicted cone count. Temperature was not important in this model. The fit of this model was r2 = .16. A similar model was used to describe the relationship of cone count and weather data from the previous year: Chapter 4. Results 41 Cone count = 118.50 — 0.95(Precipitation _i) (4-15) The fit of this model was r 2 = .02. Once again, temperature was not important in the model. However, in the model describing the relationship between cone count and weather two years previously, temperature did have an effect: Cone count = —384.56 + 35.66(Temperature _2) (4-16) — 0.^(Precipitation _2) This model had an R2 = .08, a small improvement over the one-year offset. 4.4 Individual-tree data Volume increment, bolewood weight increment, and ring density for the year were each regressed against the cone crop on an individual-tree basis. Each of these wood traits was broken down into earlywood, latewood, and total ring components (Figures 4.7, 4.8, 4.9). The results are summarized in Table 4.4. The linear models for these regressions are: Earlywood volume = 805.51 + 12.18(Ccme count) (417) Latewood volume — 155.41 + 1.89(Coree count) (418) TotaCvolume = 960.92 + 14.08(C7one count) (4.19) Earlywood biomass = 248.46 + 3.41(Cone count) (4.20) Latewood biomass = 79.06 + l.0(Cone count) (4-21) Totallnomass = 333.03 + 4.48(Cone count) (4.22) Earlywood density = .3150 — 0.0003(CWe count) (4-23) Chapter 4. Results 42 6000 0 * 1 *—I—i—i—i—I—i i i I < < . I . 0 40 80 120 ifin J—i—i—i i i i i i i _i_JL  160 200 240 280 320 360 Cone Production (#) 1500 1200 | 900 a> 600 300 I 1 1 1 I ' 1 1 I JET- . \ s > r i i - J — i — I — i — i i I i i 4 0 8 0 120 160 200 240 280 320 360 Cone Production (#) 7500 6000 I ' 1 1 I 1 1 1 I 1 1 1 I 0 B*'. •• i • • . | 40 80 — — • — • — 1 — • — • — • — ^ — • — ' — i — i — i — i — i — i — i i i i i i i_ 120 160 200 240 280 320 360 Cone Production (#) F i g u r e 4.7: D i a g r a m s of the re la t ionship of e a r l y w o o d , l a t e w o o d a n d t o t a l sheath v o l u m e to cone count . Chapter 4. Results 2000 43 «• 1600 E ro 3 j= 1200 O) 'cu "8 800 o | CO UJ 400 T 1 1 j 1 1 1— "| i " > | i—(—i—|—i—i—r~ wye' •" m 40 80 120 160 200 240 Cone Production (#) _L 280 320 360 1000 ~i | 1 ' i | i i — i — | — i — i — i — | — i — i — i — | — i — i — i — p 40 80 120 160 200 240 280 320 Cone Production (#) 360 500 -i i i | i — i — i — | — i — r — is—i——i i i i _ - J — i — i — i — i — i — i i i i i_ 40 80 120 160 200 240 280 320 360 Cone Production (#) Figure 4.8: Diagrams of the relationship of earlywood, latewood and total sheath biomass to cone count. Chapter 4. Results i E "8 0.5 0.4 0.3 0.2 -i—i—i—|—i—i—i—|—i—i—i—|—i—i—i—|—i—i—i—|—i—i—i—|—i—i—i i i ' i r n 0.1 ID I . . . I . . . I • . I . . . I 40 80 120 160 200 240 280 320 360 Cone Production (#) 0.75 £ 0.6 V) E <o T" 0.45 o "8 <5 0.3 0.15 1—I—1—I—I—I—1—|—I—I—r~ ~ i — i — I — i — i — i — l — i — i — I — l — I — I — l — l — i — I — I — [ — i — r -' • ••As*. 0 '—*- j i I i u , i I i i_ J_ ' ' I 1 L i I i i i I J I 1 I I L I I • 40 80 120 160 200 240 280 320 360 Cone Production (#) 0.5 8 0.4 -E ro S 0.3 >. In c CD Q 0.2 e > CD rr 0.1 i I i i i - i — i — i — | — i — i — i — | — i — i — r • ' I I • t I I I L . I i I . I i ' ' I ' 44 " 0 40 80 120 160 200 240 280 320 360 Cone Production (#) Figure 4.9: Diagrams of the relationship of earlywood, latewood and overall density at d.b.h. to cone count. Chapter 4. Results 45 Trait r2 slope Volume increment Earlywood .42 973.68 12.18 Latewood .30 195.64 1.89 Total .43 1106.41 14.08 Bolewood biomass Earlywood .39 287.35 3.41 Latewood .30 103.23 0.10 Total .40 371.22 4.48 Ring density Earlywood .27 0.031 -.0003 Latewood .01 0.073 .0001 Total .18 0.037 -.0002 Table 4.4: Coefficient of determination (r2), standard error of the estimate (Se), and slope for the single regressions of wood traits, using individual-tree data. Latewood density = .4995 — 0.0001(C7one count) (4.24) TotaCdensity = .3495 - 0.0002(C7orae count) (4.25) The r2 values of these regressions were moderate for growth traits, showing that there was a relationship between cone crop and these wood traits. Volume had a positive rela-tionship with cone crop. The majority of the total volume Is contributed by earlywood, as can be seen in their nearly identical regression lines. The relationship between early-wood and total year growth can be seen in bolewood biomass increment as well. Biomass also had a positive relationship with cone crop. Though bolewood biomass increment is a product of ring density and volume, the volume component, having a much greater range of variation (see Table 4.2), obscures the effect of density variation on bolewood weight. Unlike volume and bolewood weight increment, the relationship between ring density Chapter 4. Results 46 and cone crop was slightly negative. However, like volume and biomass, the majority of the yearly density is contributed by earlywood. There was a moderate relationship between cone crop and earlywood density, with an r2 value of .27. For total density, the r2 value was .18. The outliers on the bottom of the latewood density graph (Figure 4.9) are the result of an anomaly associated with the densitometry process. No wood with a density less than .40 gms/cc should be recorded as latewood, as that value automatically triggers the densitometer to begin a new ring. However, two things can occur to disrupt this. An included low-density section in the latewood ring can drop the. density briefly. The operator can also override the automatic trigger in favor of a visually determined manual triggering point; this is occasionally done when the densitometer reads a ring and does not record any latewood at all. In order to keep the years straight, the operator may trigger the densitometer to record a section as latewood if it appears to be latewood. The standard errors of the estimate shows that, in these simple regressions, the vari-ation around the regression line for volume of individual trees is as much as 1107 cubic centimeters. Likewise, the variation around the regression line for bolewood biomass is as much as 372 grams. The variation for ring density is much less, about 0.037 grams/cc. It was hypothesized that each clone might have a slightly different slope to its re-gression of wood traits against cone crops. An additional hypothesis was that clones with a higher level of cone production would show a tighter fit in a regression with ring density against cones. Accordingly, for each clone, earlywood density, latewood density and overall ring density were plotted as a function of cone crop. For earlywood density, r2 values ranged from .02 to .64, with 3 of the 12 clones having an value over .40. Late-wood density had r2 values from .02 to .18, with four clones having a value over .10. Overall ring density had r 2 values that ranged from .00 to .40, with three clones over .35. However, this was attributed to random effects rather than actual differences in the Chapter 4. Results 47 Trait Relationship R2 Se slope of cone count Volume increment Earlywood f(age, cone crop, precipitation) .64 770.58 6.49 Latewood f(age, temperature, cone crop) .61 146.33 0.58 Total f(age, cone crop, precipitation) .66 855.73 7.13 Bolewood biomass Earlywood f(age, cone crop, precipitation) .62 229.27 1.69 Latewood f(age, temperature, cone crop) .59 79.27 0.31 Total f(age, cone crop, precipitation) .64 290.22 2.11 Ring density Earlywood f(cone crop, temp., precip.) .46 .0260 - 0.0002 Latewood f(inverse of age, age) .10 .0677 N / A Total f(cone crop, temp., precip.) .42 0.312 - 0.0002 Table 4.5: Coefficients of determination (R2), standard error of the estimate (S^), and slope of cone count for the multiple regression of wood traits, using individual-tree data. relationship between different clones' ring density and cone crops, because there was no relationship between the r2 value and the level of cone production. As the relationship of wood traits to cone crop alone was fairly weak, model-fitting was done to improve the fit. The plots of the wood traits against the cone crop (Figures 4.7, 4.8, 4.9) indicated that a curvilinear model might be appropriate in some cases. Yearly cone crop, age of the ramets, precipitation and temperature were used. Stepwise regression was used to generate the best models, summarized in Table 4.5. Variables are listed in order of partial- R2 contribution to the model. Partial- F values are not shown because of the acknowleged bias of the F values due to the serial correlation in the data sets. The model will consistently underestimate the variances, rendering any inferences inaccurate. The linear models used for predicting each of these relationships were: Chapter 4. Results 48 Latewood volume Tot'il volume Earlywood biomass Earlywood volume - 320.13 + 187.77'(Age) + 6.49(Cone count) -21.07 (Precip.) (4.26) -1685.80 + Q5.15(Age) + 88.02(Temp.) +0.58(Cone count) (4.27) -206.24 + 254.54(Age) + 7.13(Cone count). -20.81(Precip.) (4.28) 29.43 + 59.35(Age) + 1.69(Cone count). -5.62(Precip.) (4.29) LatewooTbiomass = -845.03 + 33.41(ylsre) + 43.76(Temp.) +0.31(Cone count) (4.30) TotalbTomass = -162.83 + 90 Al(Age) + 2.11(C one count) -6.09(Precip.) (4.31) Earlywood density = -0.0064 - 0.0002(C7one count) + 0.02(Temp.) +0.0008(Predp.) (4.32) 1.21 - 4.39(1/A^e) - 0.03(A#e) (4.33) -.078 - .0002(Cone count) + .028(Temp.) -.0007(Predp.) (4.34) Latewood density Overall density On an in dividual-tree basis, the best model in terms of R2 values for predicting the total volume increment was volume as a function of ramet age, the cone crop, and the amount of precipitation received in May-July (Figure 4.10). The same model was also best for predicting earlywood volume increment. Latewood volume was best predicted by the age of the ramet, the temperature in May-July, and the cone crop. In all these cases, Chapter 4. Results 49 as before, there was a positive relationship between cone crop and volume increment. It is very slight in the latewood volume regression, however. Total bolewood weight increment was best predicted using the same model as total volume increment, where weight was a function of ramet age, cone crop, and precipitation (Figure 4.11). The R2 value was slightly less, reflecting the effect of the ring density. Earlywood weight increment was also best predicted using.this model. However, latewood weight increment was best predicted as a function of ramet age, May-July temperature, and the cone crop. The relationship between cone crop and bolewood weight increment was positive, although only weakly so in the latewood regression. The best model for predicting ring density was density as a function of cones, May-July temperature, and May-July precipitation (Figure 4.12). Age was not an important factor. However, when it is broken down into predicting the earlywood and latewood density, two other models were used, both of which contained age as a variable. Early-wood density was best predicted by the ramet age, the May-July temperature, and the cone crop. Latewood density was best predicted as a function of age and the inverse of age; cone crop was not a factor. However, this model was extremely weak, indicating that age is only a minor factor in determining the density of latewood. 4.5 Clonal means 4.5.1 Wood trait by cone count regressions Wood trait by cone count relationships were also estimated using clonal means. These regressions were calculated to aid tree breeders who make roguing decisions by family rather than based on individual-tree performance. The simple model of wood traits as a function of cone crop was used initially. Volume increment, bolewood biomass, and ring density are presented as a function of cone crop Chapter 4. Results 50 F i g u r e 4.10: D i a g r a m of the re la t ionship between t o t a l sheath v o l u m e , age, a n d cone count . A g e is g iven i n m o n t h s , cone p r o d u c t i o n i n numbers of female cones, a n d v o l u m e i n c u b i c c m . C**Pter4. R e s u I t s k 0 \ figure Hi- p. Chapter 4. Results 52 F i g u r e 4.12: D i a g r a m of the re la t ionship between re lat ive dens i ty at breast he ight , cone p r o d u c t i o n , a n d M a y - J u l y temperature . C o n e p r o d u c t i o n is g iven i n n u m b e r s of female cones a n d tempera ture i n degrees C e l s i u s . R e l a t i v e densi ty is uni t less . Chapter 4. Results 53 Trait r2 slope . Volume increment Earlywood .53 758.10 12.8924 Latewood .34 170.25 1.9525 Total .53 883.99 14.8449 Bolewood biomass Earlywood .50 224.62 .3.5835 Latewood .34 89.25 1.0172 Total .49 299.36 4.6794 Ring density Earlywood .45 .0243 -0.0004 Latewood .04 .0444 0.0001 Total .31 .0287 -0.0003 Table 4.6: Single regressions of wood traits by cone production, using clonal means as data. in Table 4.6. The linear models that describe the relationship are: Ave. earlywood volume = 773.15 + 12.89(Ave.cone count) (4.35) Ave. latewood volume — 155.61 + 1.95(Aue.cone count) (4.36) Ave. total volume = 928.77 + U.84(Ave.cone count) (4.37) Ave. earlywood biomass = 243.19 + 3.58(Ave.cone count) (4.38) Ave. latewood biomass = 80.49 + 1.0(Ave.cone count) (4.39) Ave. total biomass = 329.33 + 4.68(Ave.cone count) (4.40) Ave. earlywood density = .3219 - 0.0004(Ave.cone count) (4.41) Ave. latewood density = .4964 + 0.0001(Aue.ccme count) (4.42) Ave. total density = .3555 — 0.0003(At>e.cone count) (4.43) Chapter 4. Results 54 Using clonal means rather than individual-tree data increased the r2 value and reduced the Se value, but only because using averages flattens the overall variation. Volume increment again showed a positive relationship with cone crop (Figure 4.13), as did bolewood biomass (Figure 4.14). The regression between volume and cone count showed a strong similarity to the regression between biomass and cone count, once again reflecting the major contribution that volume makes to biomass. The r2 value of the regression between earlywood and cone crop was much higher than latewood in all cases. It is especially pronounced in the relationship between latewood density and cone crop, where the model only explains 4% of the variation (see Table 4.6). Ring density was again negatively correlated with cone crop (Figure 4.15). Model fitting was done as before, using the variables cone crop, age of ramets, the inverse of the age, the May-July precipitation, and the May-July temperature. Each model was generated using stepwise regression. The best models are summarized in Table 4.7. Variables used in the regression are listed in order of partial- R2 contribution to the model. The linear models used to describe the regressions are: Ave.earlywood volume = 263.52 + 195.97(Age) A- 6.10(Ave.cone count) -21A5(Precip.) (4.44) Ave.latewood volume = -1259.99 + 107.28(Age) + 3666.39(1/Age) (4.45) Ave.total volume = —272.98 + 266.36(A<7e) + 6.4\8(Ave.cone count) - 21.44( Precip.) (4.46) Ave.early wood biomass = 13.61 + 62.OA(Age) + 1.55(Ave. cone count) ~5.7Q(Precip.) (4.47) Ave.latewood biomass = -646.76 + 55.30(Age) + 1871.81(1/^) (4.48) Chapter 4. Results 55 5000 4000 CD | 3000 —i 1 1 1 1 1 1 1 1 r 1 1 1 1 1 1 1 1 1 r-120 160 200 Cone Production (ft) 1500 40 80 120 160 Cone Production (ft) 200 7500 6000 T I "'I I" ' F "I 200 Cone Production (#) 240 280 240 280 240 280 Figure 4.13: Diagrams of the relationship of earlywood, latewood and total sheath volume to cone count, using clonal means. Chapter 4. Results 56 1500 1200 ~l 1 1 1 I '—1—'—|—<~ 120 160 200 240 280 Cone Production (#) 500 ~i 1 1 1 1 r-J L _ J , L . 40 80 120 160 200 Cone Production (#) — i i i i_ 240 280 2000 1600 1 i i i | — i — i — r - ~1 1 1 1 r-_ l l I • > 40 J—i—i—i—I i i i I 80 120 —J I L 160 200 J i —1 I I I L . 240 280 Cone Production (#) F i g u r e 4.14: D i a g r a m s of the re la t ionsh ip of e a r l y w o o d , l a t e w o o d a n d t o t a l sheath b iomass to cone c o u n t , u s i n g c l o n a l means . Chapter 4. Results 57 0.5 ^ 0.4 E ra CD O "8 te 0.1 1X1 r - i — | i i i r-—r —t I I I 1 I L . _J I ' ' ' I L _ 40 80 120 160 Cone Production (#) 200 240 280 0.75 $ 0.6 E re c CD Q o i S re « 0.45 0.3 0.15 -T 1 1 1 1 r- T 1 1 1 1 1 1 1 1 1 1 1 1 r--i 1 i 1 i i i_ _J 1 I 1 I I L_ 40 80 120 160 Cone Production (#) •200 240 280 0.5 E re V) c CD O CD .> SSI CD rr 0.2 -0.1 I I I I I I I 40 80 120 160 Cone Production (#) 200 240 280 Figure 4.15: Diagrams of the relationship of earlywood, latewood and overall density at breast height to cone count, using clonal means. Chapter 4. Results 58 Trait relationship R2 Se slope Volume increment Earlywood f(age, cone crop, precipitation) .84 452.18 6.095 Latewood f(age, inverse of age) .79 103.71 N / A Total f(age, cone crop, precipitation) .85 509.25 6.477 Bolewood biomass Earlywood f(age, cone crop, precipitation) .82 136.54 1.548 Latewood f(age, inverse of age) .77 56.82 N / A Total f(age, cone crop, precipitation) .83 177.11 1.834 Ring density Earlywood f(cone crop, temperature, precip.) .68 .0188 -0.0003 Latewood f(inverse of age, age) .22 .0591 N / A Total f(cone crop, temperature, precip.) .62 .0215 - 0.0003 Table 4.7: Coefficient of determination (R2), standard error of the estimate (Se), and slope of cone count for the multiple regression of wood traits, using clonal means. Ave.total biomass Ave.earlywood density Ave.latewood density Ave.overall density -183.50 + 95.13(A#e) + 1.83(Ave.cone count) -6.39(Precip.) (4.49) 0.0061 - 0.0003(Ave.cone count) + 0.02(Temp.) -0.001{Precip.) (4.50) 1.23 - 0.03(Age) - 4.48(l/^e) (4.51) -0.067 - 0.Q00S( Ave. cone count) + 0.028(Temp.) -0.0007(Predp.) (4.52) Using the two most important independent variables, each of these relationships is shown. The regression surface for volume using clonal means is shown in Figure 4.16. The regression surface for biomass is shown in Figure 4.17, and the regression surface for density is shown in Figure 4.18. aPter 4 Results Chapter 4. Results 60 F i g u r e 4.17: D i a g r a m of the re la t ionship between age, cone p r o d u c t i o n , a n d t o t a l sheath b i o m a s s , u s i n g c l o n a l means . A g e is i n m o n t h s , cone p r o d u c t i o n i n n u m b e r of female cones, a n d b iomass i n grams. Chapter 4. Results 61 F i g u r e 4.18: D i a g r a m of the re la t ionship between cone p r o d u c t i o n , M a y - J u l y t e m p e r a -t u r e , a n d relat ive densi ty , us ing c l o n a l means . C o n e p r o d u c t i o n is g iven i n n u m b e r of female cones a n d t e m p e r a t u r e i n degrees C e l s i u s . R e l a t i v e densi ty is uni t less . Chapter 4. Results 62 As before, earlywood volume increment arid bolewood biomass are positively corre-lated with cone crop, and earlywood ring density is negatively correlated. As earlywood comprises the majority of the total year's growth, total volume increment and bolewood biomass are mostly a function of earlywood. The age of the ramet is a far more significant factor in the volume and biomass regressions; cone crop contributes significantly to the model, however. 4.5.2 Trends over time An effort was made to show the relationships between wood traits and cone crops more clearly, particularly the negative correlation between ring density and cone crop. There-fore, clonal means were plotted by year for the variables cone crop, volume increment, bolewood biomass, and ring density (Figures 4.19a, 4.20a, 4.21a, 4.22a). In addition, the weather data showing May-July temperature and precipitation over the period 1979-1987 are shown (Figure 4.23). Cone crop, volume increment and bolewood biomass all showed increases over time, which is unsurprising as this is a young, rapidly-growing seed orchard. Ring density is declining over time, which follows the typical pattern for juvenile lodgepole pine. If it continues to behave according to this pattern, ring density has already begun, or will soon begin, to increase. It can be noted that, in Figure 4.19a, certain clones are always near the top, and others are always near the bottom. The 12 study clones chosen from the original 36 were selected for their flowering patterns, with four heavy cone producers, four light, and four intermediate. A similar pattern may be observed for the other three graphs, where each clone maintains an only slightly varying position in relation to the other clones., Chapter 4. Results 63 Chapter 4. Results Figure 4.20: Sheath volume over time by clone (a) and by cone production level (b). Chapter 4. Results 65 Figure 4.21: Sheath biomass over time by clone (a) and by cone production level (b). Chapter 4. Results 66 Figure 4.22: Relative density at d.b.h. over time, by clone (a) and by cone production level (b). Chapter 4. Results 67 Figure 4.23: May-July weather patterns at Prince George over the period 1979-1987. Temperature (a) and precipitation (b). Chapter 4. Results 68 4.6 M e a n s b a s e d o n c o n e - p r o d u c t i o n l e v e l As explained above, the 12 study clones were chosen for their cone-production level. When the dataset is further refined by considering means of the heavy, intermediate, and fight cone-producing levels, much of the noise is removed. The four heavy cone bearers were clones 139, 102, 113, and 122. The four light cone bearers were clones 127, 134, 198 and 200. The four intermediate cone bearers were clones 116, 128, 129, and 133. The clonal trends observed above are much more pronounced when plotted by cone-production level (Figures 4.19b, 4.20b, 4.21b, 4.22b). It is clear from Figure 4.19b that there is very little crossover between the levels, with all three starting out at the same point but the heavy cone bearing clones quickly surpassing the others. As would be expected from the regression results, the volume increment and bolewood biomass graphs (Figures 4.20b and 4.21b) show heavy cone producers to be at the top of the graph, although not very different from the intermediate cone producers. Low cone producers are found at the bottom of the volume and bolewood biomass graphs. However, the low cone producers top the ring density trends (Figure 4.22b). In fact, the ring density graphs by cone-production level show the inverse of the cone-crop graph, with the heaviest cone producers having the lowest ring density and vice versa Figure 4.24 shows two density profiles, one from a heavy conebearer (sample 287, from a ramet of clone 139), and one from a light conebearer (sample 77, from a ramet of clone 127). Examination shows that the earlywood density of the light conebearer is rarely surpassed by the earlywood density of the heavy conebearer. In addition, the latewood density of the light conebearer is often well above the latewood density of the heavy conebearer. Figure 4.24: Density profile of a heavy conebearer (lower line) and light conebearer (upper line). Density on the Y-axis is given in g/cc; the X-axis is calendar year. C D Chapter 4. Results 70 Progeny trait r2 slope F PR> F Se Height .054 -.0111 0.52 .4909 1.85 Volume .016 -.0037 0.14 .7135 1.16 Ring width .001 -.0005 0.01 .9290 0.61 Relative density .228 -.0995 2.66 .1375 7.28 Biomass .076 -.0020 0.74 14112 0.27 Table 4.8: Coefficient of determination (r2), slope of the regression for growth and wood traits of progeny by the cone production patterns of same-family clones at OPSO, sig-nificance level (PR > F), and standard error of the estimate (Se)-4.7 P r o g e n y comparisons As cone production has been shown to be heritable (Schrmdtling, 1981), comparing the OPSO trees with progeny data from the SK selections planted in Sweden gave an op-portunity to establish whether the progeny of fecund trees had poorer qualities. Simple linear regressions were performed, using progeny height, volume, ring width, relative density, and dry-stem biomass as dependent variables and cone production means of the same OPSO clones as the independent variable. The results are summarized in Table 4.8. Since no year-to-year trends are being examined, serial correlation is not a problem, so the .F-values and the probability of a greater F are included. Although the relationships are very weak, they are uniformly negative. Relative density has a better fit than the other traits, and has the greatest slope, indicating a greater impact of cone production. There is also a slight impact of cone production on height. C h a p t e r 5 D i s c u s s i o n 5.1 D a t a s o u r c e s The main goal of the project was to determine whether there was a relationship between cone count and wood traits. The cone count data were appropriate for that goal because they included a wide range of values, although some of the variation in counts was associated with rapid increases in cone production as a function of age. Earlier studies of fecundity and vegetative growth were on older trees where trends associated with age were less apparent. Serial correlation limited the inferences that could be drawn from statistical analyses, but the r2 values of models testing various hypotheses provided an indication of which variables were influencing growth and wood traits. 5.2 P r e l i m i n a r y a n a l y s e s The problem of serial correlation invalidates the probabilities associated with the F-statistic. Thiel (1971) has shown that autocorrelation results when data are collected from the same individuals over time. Therefore, the statistical tests performed in this study, with the exception of the progeny test regression and the analyses of variance, do not include this statistic. Despite its overestimation of probability, the significance level was used to assist in variable selection for the multiple regressions. 71 Chapter 5. Discussion 72 5.2.1 D a t a p a r a m e t e r s All the wood traits were more variable between years than between clones. This was due to the large effect that age of these ramets has upon the wood qualities. Volume variation between years was much greater than between clones, as was biomass variation. Density was also more variable between years than between clones, although the difference was less substantial. Figures 18-21 show that all the levels of the wood traits are changing over time, with volume and biomass increasing and density decreasing. The increase in growth is not unexpected, as these are young trees showing the typical fast growth rate. The decrease in density is also normal for lodgepole pines at this level of maturity. However, there was variation of 115 cubic centimetres around the mean of volume by clone, and 35 grams variation around the mean for biomass by clone. Density had a smaller degree of variation, with 0.0054 gm/cc variation around the mean by clone. Cone count was also slightly more variable between years than between clones. As these trees are grafts, and therefore should be reproductively mature right away, the increasing upwards trend in cone count data (see Figure 4.19) should be a function of the increasing size of the crown and availability of locations for cones. However, there was some between-clone variation (12 cones/ramet) around the mean for cone count. Although overall cone count is increasing, there is a difference in number of cones produced by different clones. 5.2.2 A n a l y s e s o f v a r i a n c e An analysis of variance model including clones, blocks, years and their interactions ac-counted for much of the variation in cone counts, growth and wood traits. Year effects were pronounced for all variables and reflected trends with age (see Figures 4.19, 4.20, 4.21 and 4.22). Interpretation of clone effects for cone crop and density was clouded Chapter 5. Discussion 73 by interactions with year and block. The interaction between clone and year was not unexpected, as cone crops tend to cycle from year to year. This trend is less pronounced in lodgepole pine than in many other coniferous species. The block by clone interaction had no obvious explanation. The analyses of variance for cone crop, volume, bolewood biomass, and relative den-sity, had high r2 values and therefore explained much of the variance. There had been concern that the amount of variance within clones would be large, due to the effects of the rootstocks on different ramets. While it was impossible to test this hypothesis rigorously, as the rootstocks came from a single bulked seedlot and were unpedigreed, this appeared not to be the case. There was also a large clone by block interaction, which indicated that there were microsite differences affecting the performance of the same clone in different blocks. The clone by block interaction was studied at length by examining the analyses of variance for individual years. This procedure eliminates the serial correlation associated with examining the same samples over time. However, the clone by block interaction remained unexplained. It was not important when considering individual years, but only in the overall model. The OPSO site is level and fairly uniform with widely-spaced (4 x 8 m) trees; one possible explanation for the interaction is that two of the three blocks are on the edge of the orchard and may be receiving more wind, while the third is in the center. The substantial amount of variation between clones for density and cone crop is of note, as it indicates that selection for density increases or against cone crop is possible, should this be seen to be desirable. Seed orchard management programs have begun to incorporate methods to reduce the year-by- year variation, particularly to encour-age abundant flowering every year. This may make it easier to find the between-clqne variation. The clone by year interaction which is important in cone count and density is also of Chapter 5. Discussion. 74 note. This could be reflecting cyclical cone counts, which are typical of conifers. However, it may also indicate that density is cyclical as well, perhaps cycling inversely with cone crop. This is not well illustrated in Figures 4.19 and 4.22, which probably is a result of the strong growth trends associated with the age of these trees. 5.2.3 Weather analysis The lack of relationship between weather and cone crop was surprising. The probable explanation for this lack of relationship is again the overriding effect of age, where the fast growth of the trees permit them to carry a larger crop of cones each year. It may also be that lodgepole pine is less affected by weather in cone initiation and development than many other conifers. There had been concern that weather might be controlling both cone count and wood traits to some extent, so that the supposed effect of cone count on wood traits would actually, be a masked expression of the weather. However, the poor relationship between May-June temperature and cone counts, and between May-June precipitation and cone counts, would imply that this.is not the case. There was a relationship between weather and wood traits, which was shown in the multiple regressions. 5.3 Individual-tree data Initially, it was hoped that the relationship between wood traits and cone counts would be shown clearly by examining individual-tree data. Accordingly, simple regressions were calculated to show how strong this relationship might be, by regressing wood traits individually against cone count. Chapter 5. Discussion 75 5.3.1 Simple regressions In almost all cases, the simple regressions of wood traits on cone crop showed moderate r2 values. They indicated cone count did influence wood traits to a certain degree. The fit was particularly good for the volume and biomass increments (Figures 4.7 and 4.8). As biomass increment is a product of volume and density, it was expected that it would be similar to one or both of these traits. Because of the large amount of variation in volume, and the small amount of variation in density, the fit of regression models for biomass follow volume much more closely. Both volume and biomass increment were positively correlated with cone crop, an unexpected result. In light of the work by Tappeiner (1969), El-Kassaby (1988) and others, it was expected that these traits would be negatively correlated. The strong growth trend that these trees are showing at a young age may have obscured or changed this relationship, which may show a negative correlation at an older age. The fit of models using earlywood as the dependent variable was almost identical to overall models, which reflects its overwhelming contribution to the overall bolewood growth. Latewood models also showed a positive relationship with cone crop, although the r2 value was lower, nor was the impact of cone count as pronounced. Relative density regressed against cone crop showed a negative relationship, which could be a cause of some concern to tree breeders. The slopes of the regression lines (Figure 4.9), however, were not as steep as for volume or biomass. This implies a weaker impact on density from cone crop. The relationship between earlywood density and cone count was nearly identical to the relationship between total density and cone count. This was probably because the annual rings of young trees are overwhelmingly dominated by earlywood rather than latewood. Latewood density had virtually no relationship with cone crop, which is not surprising as latewood production begins after cone enlargement Chapter 5. Discussion 76 is essentially over, in late June or early July. 5.3.2 Multiple regressions Multiple regressions allowed the other variables influencing wood traits to be shown. In multiple regressions, cone production was still an important factor in the models. However, the regression of wood traits resulted in low partial R2 values for all variables except for age. The partialri?2 for age was approximately .50, while cone crop was .06 to .10. Age is an important variable because of the relative youth of these grafts and their rapid growth rate. The crown size is expanding each year, yielding more photosynthate for growth. In simple linear regressions, the partial-r2 contribution of cone crop was much greater than in multiple regressions. In multiple regressions, a much lower partial- R2 value is seen in volume and biomass, which probably better approximates its impact on these traits (Figures 4.10, 4.11). The small partial-r2 values for everything except age indicated that cone crop, precipitation, and temperature were only explaining a small amount of the variation around the regression for wood traits, and that the effect of age was large, perhaps large enough to obscure some of the effects of the other variables. The multiple regressions showed that earlywood volume and biomass are subject to impact from the amount of precipitation received in May-July. As the earlywood component is much larger than the latewood component, total volume and biomass were likewise subject to impact from the amount of precipitation. Interestingly, precipitation was negatively correlated with these traits. This was unexpected except in the light of the lower temperature and solar insolation that could be expected in a rainy month. If the amount of available water is not limiting, but the amount of sunlight is, a negative relationship such as the one predicted by these regressions is reasonable. Additionally,. the multiple regressions showed that earlywood and total density are Chapter 5. Discussion 77 positively correlated with temperature. It may be that higher temperatures imply more sunshine and therefore a higher photosynthetic rate, and that this increased amount of photosynthate is used by. the cambium to lay down thicker-walled cells. However, one would expect therefore to find a positive correlation with volume and biomass as well, which is not the case. Latewood volume, biomass and density were each influenced by both age and the inverse of age. The effect of age has been shown clearly as the overwhelming indicator of tree volume and biomass. Using the inverse of age introduces a curvilinearity into the model. The curvilinear rather than linear model shows that the relationship changes over the range of the independent variable, which in this case is age. The effect of age, therefore, varies from quite important early in the tree's lifetime to less important later. The slope of the regression of wood traits on cone counts varied between traits, indicating that cone count had a differing impact on each trait. It was relatively steep for the positive relationship between cone count and earlywood volume, as well as total volume. It was much less steep for the relationship between cone count and biomass. This was due to the effect of relative density in the calculation of biomass; it would indicate that biomass is much less affected by cone count than is volume. The slope of the relationship between earlywood density and cone count, and between overall density and cone count, was very small. This implied that the impact of cone count on density was weak, although it was negative. The cone count could be influencing the density of the tree by using much of the available photosynthate, which would leave less for cell wall construction. Chapter 5. Discussion 78 5 . 4 C l o n a l m e a n s f o r a n a l y s i s To develop guidelines that would be relevant for use as roguing criteria, clonal means were used as data for regression analyses. Using clonal means removed the within-clone variance, and conclusions could be drawn much more easily with clonal means as to whether to keep or remove a clone from an orchard. 5 . 4 . 1 R e g r e s s i o n s The simple regressions between wood traits and cone count were calculated using clonal means for both the trait in question and the cone count. Using the slope of the regression line to predict density, it would appear that a twofold increase in cone crop can cause a loss of 4.6% of the wood density, that the two are directly related and not being influence by a third factor. The linear models for the clonal multiple regressions between wood traits, cone count, age, and weather data were slightly different from the models using individual-tree data. However, this is due to the effect of flattening the variation by using means, and should be ignored. As seen in the individual-tree multiple regressions, the age of the ramet, the cone count, and the amount of precipitation in May-July had the greatest influence on yearly volume and biomass growth (Figures 4.16, 4.17), while the cone count, the temperature in May-July, and the precipitation in the same period most affected density (Figure 4.18). Again, the relationship between cone count and volume was positive, while it was weakly negative between cone count and density. Chapter 5. Discussion 79 5.5 Trends over time Cone production over time (Figure 4.19) steadily increased as the OPSO trees grew larger. It would appear, from the direction of the graph to date, that more cones will be produced in the future as these trees are not yet full-sized. Certain years appear to be high cone-producing years, such as 1983 and 1985. Lodgepole pines do not experience the dramatic swings in cone production observed in other conifers such as Douglas-fir, although there is a short cycle between heavy cone years. The by-clone graph (Figure 4.19a) shows that, although there are some localized crossovers within the graph, the clones near the top seem to stay there, while those near the bottom are similarly consistent. This is more striking in Figure 4.19b, which shows the clones grouped together as heavy, intermediate and light cone-bearers. There is no crossover at all, as the heavy conebearers remain substantially above the other groups. The intermediate group is also well above the light-conebearing group. This supports Eis (1965) and Ying et al. (1985), who reported that good conebearers were consistently good, while poor conebearers were consistently poor. Schmidtling (1983) reported that the top cone-producing trees shifted rankings each year. This was not born out by the study of the OPSO trees, although the clones studied were selected on the basis of their cone production patterns. Within the cone production level, there was minor rank shifting each year, although overall the ranks did not change very much. Volume and biomass increment over time (Figures 4.20 and 4.21) also show steady increases, which seem to be accelerating from 1984 on. There is a substantial amount of crossover in all of these graphs, indicating that these traits are not as closely linked with cone production. In Figures 4.20b and 4.21b, the clones are again grouped by cone- production level. There is a large amount of crossover between the heavy and the intermediate conebearers, although the light conebearers remain at the bottom of Chapter 5. Discussion 80 the graph. This indicates that the light conebearers are also not gaining in volume or biomass as fast as the intermediate and heavy conebearers. However, the difference may not be significant. These young trees are growing more volume each year. This trend is strong, and may obscure variation in impact from cone crop that might be occurring in older trees with a slower growth rate. Holmsgaard's 1956 study of European beeches, which noted a drop in ring width in heavy seed years, showed that the trees most affected were over 100 years of age. Younger trees, from 60 to 100, were much less affected. Density over time (Figure 4.22) shows a gradual decline with one sharp peak at 1982. In the same manner as the by-clone cone count graph (Figure 4.19a), there is localized crossover within the by-clone density graph (Figure 4.22a) but, as a general rule, clones near the top remain there and clones near the bottom stay there. Again, this is better, illustrated by grouping the clones into heavy, intermediate and light conebearers (Figure 4.22b). This graph shows that there is no crossover between the three levels. In addition, the order of the levels is reversed from the cone production graph (Figure 4.19b). The heaviest cone bearers have the lowest density, and the lightest cone bearers have the highest density. This may be interpreted in several ways. First, noting that the lightest cone bearers do not gain volume at the same rate as the heaviest cone bearers, one could assume that some of them are smaller than the trees from the heavy and intermediate cone count levels. They would, therefore, have fewer sites for cone production and be light cone producers. Likewise, the ratio of earlywood to latewood in these smaller trees might be less than in the bigger trees, as the smaller trees grow less earlywood which is the main component of the annual ring in young trees. If the latewood ring is comparable in size and density Chapter 5. Discussion 81 to the other trees in the orchard, this would increase the latewood proportion of the ring density. The density overall would therefore be raised. It may be that these trees also switch from early to latewood production earlier in the year, leading to a greater proportion of latewood. This effect may also be the result of carbohydrate partitioning away from reproduction and into wood cell wall thickness. The light conebearers may have a genetic predisposition towards low fecundity, although this does not suggest itself in view of natural selection. While there is a definite benefit to high fecundity, the selective advantage to higher density is less obvious. It may be that there has not been a selective pressure against high relative density, and the genes for this trait have therefore remained in the population, neither selected for or against. Or it may be that higher- density trees are stronger and thus able to withstand storm damage better than their lower-density compatriots. The two density profiles in Figure 4.24 show the comparison between a light and a heavy conebearer. Examination shows that the earlywood density of the light conebearer is rarely surpassed by the earlywood density of the heavy conebearer. In addition, the latewood density of the light conebearer is often well above the latewood density of the heavy conebearer. Comparison with heavy cone years (see Figure 4.19) shows that, during 1983 and 1985, the density profile of the light conebearers is well above that of the heavy conebearers. However, in light cone years such as 1984 and 1987, the two density profiles show very similar shapes and heights. This indicates that the heavier cone years are having a definite impact on the heavy conebearers, but not on the light conebearers. This would support and expand on Tappeiner's 1969 findings that conebearing depressed ring width. Although latewood density did not appear to be affected by conebearing in the regression analyses, the drop in latewood density in heavy cone years does seem to be noticeable in Figure 4.24. This may be due to a depletion of photosynthate reserves in the early part Chapter 5. Discussion 82 of the growing season. 5.6 Progeny results The simple regressions of the Swedish progeny trial were uniformly negative, which sup-ports Schmidtling's 1983 report that flowering traits of parents were negatively correlated with growth traits of progeny. At the time that the wood samples were taken, probably 50-60% of the trees in the Sor Amsberg seed orchard were flowering. The flowering traits of the clones in the OPSO were positively correlated with their own growth traits, which may be a result of the age of the ramets. These negative correlations with the progeny present a caution to seed orchard man-agers and others. Although the trees in the orchard themselves may not show the impact of heavy conebearing, their progeny may be at a disadvantage, showing poorer than average growth traits. Efforts should be made to follow this in the future. A final note on this subject comes from the results of the Dala Jarna progeny trial, which are guiding the roguing of the OPSO. The roguing criteria is a weighted index that considers their survival rate, percentage undamaged by weather, percentage without spike knots and fork tops, and height. Height was considered the most important criterion and was weighted three times heavier than the other components. The first two clones chosen for roguing from the OPSO were 102 and 107. Clone 102 was one of the 12 clones selected for the present study as the second-highest cone producer in the orchard. Conversely, the clone ranked third best in the Swedish trial, number 198, was one of the clones selected for this study for its light conebearing. C h a p t e r 6 C o n c l u s i o n s Lodgepole pine as a species is naturally highly fecund, as it is a pioneer species after fire. It does not show the wide annual variation in cone production that are common in other conifers. Its high fecundity and low year-to-year variation in cone crops may obscure the relationship between cone counts and wood traits. A species with more noticeable variation in cone counts from year to year might show a more pronounced effect on wood traits in heavy seed years. In addition, the grafts in the OPSO seed orchard are young and fast-growing, with crowns that are quickly expanding in size. The expanding crowns provide more photo-synthate each year, as well as more sites for cone growth. It is probable that the increase in available photosynthate is the major contributor to the yearly increase in volume and biomass increment. It was expected that a negative relationship would be discovered between cone count, volume and biomass, but a positive relationship was found. This may be due in part or in total to the expanding crowns providing more photosynthate for the rapid growth of the OPSO trees, while at the same time creating more sites for cones to be formed. A stand where the leaf area index and crown size had stabilized might show a more pronounced effect on volume and biomass increment in heavy cone years. The negative relationship between wood density arid cone count should be cause for action, or at least concern. This is confirmed by the negative relationship between progeny wood traits and family cone production patterns. Even a minor decrease in 83 Chapter 6. Conclusions 84 relative density could have serious consequences. If trees with heavy cone crops have more progeny, and these progeny inherit their parents' lower relative density, future plantations may be full of cone- covered trees with poor quality wood. Therefore, caution should be exercised in roguing either high- or low-fecundity clones from orchards, to evaluate what gains or losses might be expected in the wood traits of future plantations. C h a p t e r 7 References C i t e d Anonymous. 1976. Twentieth annual report on cooperative tree improvement and hard-wood research program. North Carolina State University, Raleigh, North Carolina. Barclay, H. J. and Y. A. El-Kassaby. 1988. Selection for cone production in Douglas-fir adversely affects growth. In Proc. 10th N. American For. Bio. Workshop. Worrall, J., J. Loo-Dinkins and D. Lester (eds.). Faculty of Forestry, Univ. British Columbia, Vancouver, Canada, pp. 149-151. Barrett, J. D. and R. M. Kellogg. 1984. Strength and stiffness of second growth Douglas-fir dimension lumber. Report to the Science Council of British Columbia. 57 PP-Bramlett, D. L. and R. P. Belanger. 1976. Fertilizer and phenotypic selection increase growth and flowering in young Virginia pine. For. Sci. 22, 461-467. Byram, T . D., W. J. Lowe, and J. A. McGriff. 1986. Clonal and annual variation in cone production in loblolly pine seed orchards. For. Sci. 32, 1067-1073. Cannell, M . G. R., and J. E. Jackson. 1985. Attributes of Trees as Crop Plants. Institute of Terrestrial Ecology. Huntingdon. 592 pp. Cannell, M . G. R. 1985. Dry matter partitioning in tree crops, pp. 160-193. In Cannell, M . G. R. and J. E. Jackson. 1985. Attributes of Trees as Crop Plants. Institute of Terrestrial Ecology. Huntingdon. 592 pp. Danbury, D. J. 1971. Economic implications of selection for seed production in radiata pine seed orchards. Aust. For. Res. 5, 37-44. Dickmann, D. I. and T. T. Kozlowski. 1968. Mobilization by Pinus resinosa cones and shoots of C14-photosynthate from needles of different ages. Am. J. Bot. 55, 85 Chapter 7. References Cited 86 900-906. Dickmann, D. I. and T. T. Kozlowski. 1969. Seasonal growth patterns of ovulate strobili of Pinus resinosa in central Wisconsin. Can. J. Bot. 47, 839-848. Dickmann, D. I. and T. T. Kozlowski. 1970. Mobilization and incorporation of photoas-similated 1 4 C by growing vegetative and reproductive tissues of adult Pinus resinosa Ait. trees. Plant Physiol. 45, 284- 288. Eis, S., E. H. Garman, and L. F. Ebell. 1965. Relation between cone production and diameter increment of Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco], grand fir [Abies grandis (Dougl.) Lindl.], and western white pine (Pinus mon-ticola Dougl.). Can. J. Bot. 4 3 , 1553- 1559. Eklund, B. 1957. Annual ring variations of spruce in central Norreland and its relation to climate. Medd. Statens Skogsforskningsinst. 47, 1-63. El-Kassaby, Y. A. , A. M . K. Fashler, and M. Crown. 1988. Genetic variation in fruit-fulness in a Douglas-fir clonal/seedling seed orchard and its effect on crop-management decisions. Silvae Genetica (in press). Engleson, J. B. 1971. Selection of Pinus contorta var. latifolia in northeastern British Columbia and southern Yukon. B.Sc. thesis, Forestry, Univ. British Columbia, Vancouver. 135 pp. Evans, L. T . 1976. Physiological adaptation to performance as crop plants. Phil. Trans. R. Soc, 275B , 71-83. Evertsen, J.A. 1982. Inter-laboratory standardisation survey. Wood Microdensitometry Bulletin 2(2), 3-24. Fielding, J. M . 1960. Branching and flowering characteristics of Monterey pine. Bull. Commonw. For. Timb. Bur. Austr. 37, 59 pp. Fries, Anders. 1986. Volume growth and wood density of plus tree progenies oi Pinus contorta in two Swedish field trials. Scand. J. For. Res. 1, 403-419. Gonzalez, J. S. 1989. Circumferential variation in the relative density of lodgepole pine. Can. J. For. Res. 19, 276-279. Chapter 7. References Cited 87 Griffin, A. R. 1982. Clonal variation in radiata pine seed orchards. I. Some flowering, cone and seed production traits. Aust. J. For. Res. 12, 295-302. Gross, H. L. 1972. Crown deterioration and reduced growth associated with excessive seed production by birch. Can. J. Bot. 50, 2431-2437. Hagner, S. 1970. Plans for establishing seed plantations of Pinus contorta. Svenska Cellulosa Aktiebolaget Forestry Department. 13 pp. Holmsgaard, E . 1956. Effect of seed-bearing on the increment of European beech (Fagus sylvatica L.) and Norway spruce [Picea abies (L.) Karst]. 12th Congr. Intern. Union Forest Res. Organ. Oxford, p. 3. Holmsgaard, E. 1972. Relations between climate and flowering, seed production and growth. In: For. Tree Improv.: Symposium on seed orchards in honor of C. Syrach-Larsen. Keiding, H., H. Roulund and H. Wellendorf (eds.). Arboretet Horsholm, Akademisk Forlag, Copenhagen, pp. 53-66. Jefferson, P. A. 1984. Heritability of relative density. Master's thesis, Univ. British Columbia, Vancouver. 92 pp. Josza, L. 1988. Increment core sampling techniques for high quality cores. Forintek Canada Corporation, Special Publication No. SP-30. 26 pp. Kellogg, R. M. 1978. Wood quality control in several tree improvement programs in the Southeastern United States. A report to the British Columbia Tree Improve-ment Council. 6 pp. Kellogg, R. M. 1982. Coming to grips with wood quality. For. Chron. 58(6), 254-257. Kramer, P. J. and T. T. Kozlowski. 1979. Physiology of Woody Plants, pp. 137-145. Academic Press, Inc. 811 pp. Maguire, W. P. 1956. Are Ponderosa pine cone crops predictable? J. For. 54(11), 778-779. Matthews, J. D. 1963. Factors affecting the production of seed by forest trees. Forestry Abstracts 24: i-xiii. Chapter 7. References Cited 88 Messer, H. 1956. Untersuchungen uber das Fruchten der Europ. Larche (Larix decidua Mill). Allgem. Forst Jagdzeitung 127, 8-16. Mork, E. 1928. Quality of pine wood, with special reference to its use as pulp. Papier J. 16, 40-44. Morris, R. F. 1951. The effect of flowering on the foliage production and growth of balsam fir. For. Chron. 27, 40-57. Namkoong, G. , A. C. Barefoot and R. G. Hitchings. 1969. Evaluating control of wood quality through breeding. Tappi 52(10), 1935-1938. Polk, R. B. 1966. Reproductive phenology and precocity as factors in seed orchard development. In: Proc. 5th Central States For. Tree Imrov. Conf, H. B. Kriebel (ed.). pp. 13-21. Romberger, J. A. 1967. Flowering as a problem in developmental physiology. Proc. XIV Congr. IUFRO 3, Sec. 22:2-14. Ross, S. D. and R. P. Pharis. 1985. Promotion of flowering in tree crops: different mechanisms and techniques, with special reference to conifers, pp. 383-397. In Cannell, M. G. R. and J. E. Jackson. 1985. Attributes of Trees as Crop Plants. Institute of Terrestrial Ecology. Huntingdon. 592 pp. SAS Institute, Inc. 1985. SAS User's Guide: Basics, Version 5 Edition. Cary, N.C.: SAS Institute Inc. 1290 pp. Schmidtling, R. C. 1980. The genetic and environmental basis of fruitfulness and growth in loblolly pines. Ph.D. diss. Univ. Fla., Gainesville. 104 pp. Schmidtling, R. C. 1981. The inheritance of precocity and its relationship with growth in loblolly pines. Silvae Genet. 30, 188-192. Schmidtling, R. C. 1983. Genetic variation in fruitfulness in a loblolly pine (Pinus taeda L.) seed orchard. Silvae Genet. 32, 76- 80. Schoen, D. J., D. Denti, and S. C. Stewart. 1986. Strobilus production in a clonal white spruce seed orchard: evidence for unbalanced mating. Silvae Genet. 35, 201-205. • Chapter 7. References Cited 89 Tappeiner, J. C. 1969. Effect of cone production on branch, needle and xylem ring growth of Sierra Nevada Douglas-fir. For. Sci. 15, 171-174. Thiel, H. 1971. Principles of Econometrics. John Wiley & Sons, Inc., New York. 736 pp. Varnell, R. J., A. E. Squillace, and G. W. Bengtson. 1967. Variation and heritability of fruitfulness in slash pine. Silvae Genet. 16, 125-128. Ying, C. C., J. C. Murphy, and S. Anderson. 1985. Cone production and seed yield of lodgepole pine grafts. For. Chron. 61, 223-228. Appendix A Database for volume and biomass analysis This is a sample of the data used for the volume and biomass analyses. E W T , LWT, and T W T refer, respectively, to earlywood biomass, latewood biomass, and total-year biomass. Likewise, E V O L , LVOL and T V O L refer to earlywood volume, latewood vol-ume, and total volume. E D E N , L D E N and R D E N refer to the computer-calculated averages for earlywood, latewood and overall relative density. These were not used in the analyses (see section 3.4). SAS CLONE RAMET BLOCK YEAR CONES EWT LWT TWT EVOL LVOL TVOL EDEN LDEN TDEN 133 1 1 1978 28 1 30 89 7 4 3 94 0 0 312 0 232 0 320 133 1 1 1979 76 22 98 233 3 43 4 276 7 0 328 0 498 0 354 133 1 1 1980 6 112 45 161 351 9 86 4 438 4 0 319 0 516 0 367 133 1 1 1981 11 408 60 484 1337 1 121 4 1458 5 0 305 0 493 0 332 133 1 1 1982 5 213 78 290 630 8 154 9 785 7 0 337 0 501 0 369 133 1 1 1983 39 441 203 645 1417 7 377 6 1795 2 0 311 0 538 0 359 133 1 1 1984 12 924 65 1002 3069 6 129 2 3198 8 0 301 0 501 0 313 133 1 1 1985 580 258 831 2025 2 495 1 2520 4 0 286 0 521 0 330 133 1 1 1986 1071 352 1412 3730 2 685 9 4416 2 0 287 0 513 0 320 90 A p p e n d i x B D a t a b a s e f o r r e l a t i v e d e n s i t y a n a l y s i s This is a sample of the output from the densitometer. Rings is a measure of how many rings out from the pith is a given years. RSX is a measure of how far out from the pith the ring begins, in millimetres. RW, E W and LW refer respectively to ring width, earlywood width and latewood width. Likewise, RD, E D and LD refer to overall relative density, earlywood density and latewood density. MIND and M A X D refer to the minimum and maximum values for density in the ring. This raw densitometer data was used for the relative density analyses. SAS CLONE RAMET BLOCK YEAR CONES RINGS RSX 133 1 1 1980 0 133 1 1 1981 11 2 7 500 133 1 1 1982 5 3 12 260 133 1 1 1983 39 4 15 585 133 1 1 1984 12 5 21 730 133 1 1 1985 6 26 635 133 1 1 1986 7 31 295 133 1 1 1987 95 8 37 935 RW EW LW RD ED LD MIND MAXD 4.760 4.180 0.585 0.331 0.308 0.490 0 3.325 2.655 0.675 0.364 0.329 0.503 0 6.145 5.020 1.130 0.345 0.301 0.539 0 4.905 4.660 0.240 0.288 0.276 0.516 0 4.660 3.885 0.775 0.315 0.272 0.528 0 6.640 5.610 1.030 0.310 0.270 0.526 0 4.320 3.760 0.555 0.323 0.293 0.506 0 91 Appendix C Weather data Precipitation and temperature for the months of May through July for 1979-1987 were used for analysis. These months were chosen as cone enlargement, strobilus initiation, and leader extension are all occurring at once, making great demands on the trees' photosynthate resources. Year Month S t a t i o n P r e c i p . '/, of normal Temp. Deg. from normal * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 1979 5 PGA 71.6 170 8.7 -0 . 7 1979 5 PGSTP 77.4 . 9.9 1979 6 PGA 72.6 124 12.5 -0 . ,5 1979 6 PGSTP 59.7 13.4 1979 7 PGA 24.6 42 16.0 1, .1 1979 7 PGSTP 37.0 17.1 1980 5 PGA 65.9 156 10.7 1, .3 1980 5 PGSTP 46.8 12.3 1980 6 PGA 84.4 145 13.5 0, .5 1980 6 PGSTP 82.4 15.1 1980 7 PGA 83.4 144 14.3 -0 . ,6 1980 7 PGSTP 59.0 15.5 1981 5 PGA 63.6 151 11.2 1, .8 1981 5 PGSTP 49.2 12.4 . 1981 6 PGA 89.7 154 11.1 -1 , .9 1981 6 PGSTP 78.9 12.0 1981 7 PGA 37.9 65 16.2 1 .3 1981 7 PGSTP 48.8 17.4 . 1982 5 PGA 63.3 134 8.8 -0 .5 1982 5 PGSTP 43.9 9.9 1982 6 PGA 15.0 22 16.2 3 .3 1982 6 PGSTP 12.6 17.6 1982 7 PGA 131.2 220 16.3 1 .2 1982 7 PGSTP 123.4 17.5 1983 5 PGA 16.5 35 11.7 2 .4 1983 5 PGSTP 11.4 13.2 92 Appendix C. Weather data 1983 6 PGA 145.5 217 12.8 -0 . ,1 1983 6 PGSTP 115.1 13.9 1983 7 PGA 111.3 186 14.7 -0 . A 1983 7 PGSTP 101.7 15.9 1984 5 PGA 98.6 208* 7.8 - 1 . .5 1984 5 PGSTP 73.4 9.1 1984 6 PGA 66.5 99 12.4 -0 . ,5 1984 6 PGSTP 53.0 13.7 . 1984 . 7 PGA 41.6 70 14.9 -0 . ,2 1984 7 PGSTP 46.2 16.2 1985 5 PGA 30.5 64 10.5 '1. .2 1985 5 PGSTP 25.0 11.6 . 1985 6 PGA 34.1 50 12.7 - o . .2 1985 6 PGSTP 34.6 13.7 1985 7 PGA 20.3 34 16.7 1 .6 1985 7 PGSTP 12.4 17.8 1986 5 PGA 38.8 82 9.2 -0, .1 1986 5 PGSTP 40.2 10.2 1986 6 PGA 56.9 85 13.4 0 .5 1986 6 PGSTP 45.9 14.2 1986 7 PGA 57.4 96 14.5 -0 .6 1986 7 PGSTP 38.8 15.6 1987 5 PGA 60.4 128 9.8 0 .5 1987 5 PGSTP 39.6 10.9 1987 6 PGA 21.7 32 14.1 . 1 .2 1987 6 PGSTP 16.1 5.3 1987 7. PGA 39.4 66 16.5 1 .4 1987 7 PGSTP 48.6 17.6 Appendix D Swedish progeny trial data Clone _# Mean Mean Mean Mean Mean number ind. height density ring width volume biomass (N) (dm) (mg/cc) (mm) (dm3) (kg) 102 11 32.18 229.09 8.49 13.49 3.12 116 12 35.92 222.00 7.70 13.18 2.92 122 13 33.69 220.69 7.73 12.61 2.80 127 11 32.82 236.09 7.05 11.75 2.71 128 10 35.50 234.90 7.71 14.16 3.34 129 5 35.00 247.80 6.96 12.07 2.88 133 12 33.17 227.92 8.18 11.17 2.57 134 12 35.08 230.42 8.06 12.12 2.79 139 11 36.91 230.45 8.82 15.02 3.47 198 12 37.33 222.35 8.14 12.61 2.82 200 13 37.15 235.46 8.46 12.85 3.05 Table D.9: Swedish progeny trial data, with half-sib family means for height, density, ring width, stem volume, and dry-stem biomass 94 Appendix E By-clone and by-year means The following pages contain SAS output that includes all the by-clone and by-year means for the cone counts and wood quality traits. These data were used to generate the data description table in the early part of Chapter 4, Results. 95 3 i r m - < r m - < r m n o o o a x x < < < o - I -n - * o o o z A A A U I U K n u i l J I U l O U l ^ J> N I U ft N w u ft ui a N O O O N U I U U I U I N U I u i u i r o - ~ < ^ o o u i c o o u i r o c o - s i u i f t - t u i o o u i O f t 0 9 - s i u i f t - t - * a o u i > J N l D ^ J t 7 l f t - « - « 0 3 S U J U I - s l — J U I f t - * - » 03 - * u i u i o - s t u i f t - . — coft r o r o u i - s i u t f t - t . * G o r o C O C O c O C D C T l f t — * C O C O m - < u o > r ro ro to ro 4 O O O n i U i s f t i O O O O OD ro u> ui ro ro O OD to ai ru cn a> N ui (O o - * CO OI - * o -* — U i - O O f t - O m ro u> to o - J o ro — ui ui Ul cn —* ro O ui co ui ui ft r o (i O CD I U l - » i CD - s i C r o u i -o u i t ftNl Q O O N O O C O O O O r o J - r o OD on ft u i ro - t O O O o a o a o o o o o o o o o o o o o a o o o o o o o o o o O U i o to o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o a ) ft r o 03 ao u i - * - * u i - * o cn -» 03 o o o w m N U H O ( O -ftUiuOOOuiiDsO o - ^ - ^ o o o o o o o ft — o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o -s i _* m ft u i - * - * ) 0 N C D ( D f u c n o o S O C D ^ r o — ftroui 3 0 ^ 0 ) I U f t ( O f t N k f t c a m u i u i u i c n u i B (D ru a a> cs cn N a - O f \ ) o j L n - * c n o » c o n ui ui ui ui co -ii ft ft co oo ro -9 u i ro - s i u i - » - J O O o o o a s a < < < o —< —4 —< o o o z oi - * m - * ro ao ao ro ui — ft a> ft a o o oi - f t i u u c o f t u i f t r o o u i u i - * o n f t o n r o-Njcocoo - 4 - « u t - s i a > - O J U l l f l O - . - s i O - ' O ) u i r o a o r o O f t a a O f t O ) u i t o r o a o O r o u i O r o c o C D r o t O U I O C S N O C I I O I c n c n r u - j o c n — » O u i c n to co - s i — o ft o -si - s i u i _» ft oa ro * ao -* to to ro t o o o N f u w o w -ro - * - J -» cn ui ui to oo — oo ro CO CO ft ft ft co — — - s i t o a o u i o s f t o _ » ~ » « c o u i r o ^ c n c o o 5 - » r o a s o o o o o u i c o o o ftftroOOOOOOO m - * u i a o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o ro — ro — Oi to -si Ul u i Ul 00 ro ui ^ 03 to to u> ft 03 OD Ul tO — O CO ft ft - * U l CO - » O O O ^ S O I W J O I W o o o f t O u i r o o m r o O O O r U f t C O f t ' l O C O - ' O O l \ l U 0 S % [ O - U I O ^ a o r o f t t o o u t c o Q «CDft(_rt->4—•CDCD — r o r u t o f t f t f t t o o D O a i r o t o o o - ^ o c o r o u i u i c o O O O s t U C 9 l \ J f t U t t O u l f t O U I U l O r o o o - * O O u i r o a o ^ c n o p r o - s i u i u i u i ft-Njcofttooroui r o t o o o - ^ f t u i t o r o N O M I S O I U U I O U l t O Q f t U l C O C O C O O O O I B O l oa ro O Ul CO 00 ui ro ro o ro o o ft O ui O O o — o o o o o o o o o o o o o o o o o o o o o o o o o o o O 03 O U l o o a o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o ro ro <o to 00 ft UI ft 00 O -4 O U l U l ft to UI U l O O O t o t o r o f t - t r o U T u i m u i o o o o r o t o o - s i U I U l O O O O O O O - t m r o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o O O O U I f O C O U I - s l U I U I o - » O f t ( D c o t o r u a > c o u i u i f t O u j - s t m o u i r o O f t l O O f t f t O l U l U I N — « — • 0 9 ( 0 0 0 0 ) — • G O — U) I O — t - * U I - » f t 0 3 u i — t - s l a o o i f t - M M r o u t a i u i u i a a a o a s a o - x o r o f t c o ui ui — -» O OD co - * oo ui ft -» ui oo o ro o> oi oo - - - a o i n s f t u i t o r o u i f t o i c a f t f t a i -• U l - O O O O ) ft ft o o o r o r o o o o o o o o U l C D C O O O O O O O O o o o o o o o o o o o o o o o o o o o o ui ro ui ro ui co ft - * ft ft to ui ro ui u to -si - * ft U l CO O ft CO 0D ui 09 to ui ao - J ro O O O C O C O L T I f t U T t O U l o o o u i o u i c s f t c o f t O O O c o - U i r o O i c o c D -*tO — — O 3 O 3 U I C 0 0 D U I O - s i O t o r u f t u i r o c o c o t o - * a o c o f t - « t o o 9 c o u i N C B u o o t o i o i n j u t j i a r m - i r m - i r m n o o o s x s < < < o - l - i -4 O O O z ftft<OJIXNO O O O I D O N U O I M O u i u i r o ^ u i - * u i c o o o u i O O N ( O f t l ] i O O ( 0 - * O u t — u i u i t o r o c o u i f t O f t ^ 4 f t f t O r o o - * r o U I O f t U I U ) C O - s t C 0 0 9 0 9 - s i o r o f t f t O r o o - > u i —'Ococncrioo->jcocD-»j ft00)0"iui->uj-tro-t u i — ft *si ro t co m -si - * i ) O O -» ui ro ui as c » O O ro - * O O ro i U I - * O 0 D C 9 C 0 t l ( 0 ( U ( 0 ( 0 O U N J ) ft (O tO ft CO CO U l c i u i c o f j i r o a i - » O D C . - » t o -si m ft r o - » ( i to - * co r o to o> ft t * U l O tO — — CO 03 ( O O O f t r o t o c o t O f t O ro <i r o ft CJl w ro to u i O O O O O O o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o a - * 0B ui O O O O o o o o o o o o o o o o o o o o o o o o o o o a o a j r - m - i r - r r t - H r - m n o o o « s * < < < o - i - i -4 o o o z U l U l U l CO CO ao CO CO CO ft O O O ft ft ft ft ft ft r o CO ft to r o CO ft U l 03 U l 03 U l O O O U l r o -* CO ft -* CO U l Ut U l U l 03 •s, a> ^ ao U l ui ru O -» U l U l ft CO -s i O U l to o U l UJ ui to ui CO U l o U l ft ft ao r o r o r o o OP U l r o O O O U l U l o r o ao o o o o o o U l CJl U l U l o o o o o o o O o "** _ -CO ft CO r o CO ft o CO 03 o O O 00 *•* — CO U l 03 ft O O O U l U l r o U l U l "Si CO U l U l ft - * CD to U l U l »o to N N I U O CO CO U l r o U l -s i co r o ft to U l r o r o U l ft U l ft -st o CO (O o -si ui ft O ft 03 CO U l c n 03 r o ui ft - J ft U l r o -s i U l oo -» ID 03 U l o ft as r o O O O 00 — r o to O ro J ro * ui ui - s i 0D O) o o o o o o o o o o o o o o o o o o ft u i r o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o in m -t < > ro — a) ui -> ft s i ft N CB U o ro — ui ao ft O O O r o u i m u i - s j c n r o u i m u i o o o — ft ro o O s u i m O O O O O O O — O O O O O O O O O O O O O O O O O O O O o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o z m 1] 03 CO CO U l U l - » o o o ft oo -si ro to ui o b b CO o —» ro co co o o o o o Ul Ul tO O ft -s i Ul U l Ul -si U l CO m Ul CO CO ro Ul o 03 -si ft O ft Ul ft ao to 00 ro ft -si oo ro ft ro ro s j ft ro ui oo to U l CO o ft Ul — CO CO 03 as o to ro ft -s i CO 03 O CO U l ru O — co ro CO co ui ro ui CO ro ui — ft ui CO 03 -s i ui ro ro — o to —* ft Ul ft 00 -s i 03 ui CO Ul o o O — CD U l O o 03 o o a o o o o ro CO o o o o o o o O o o o o o o o o o o o o o o o o o o o ui ro U l ro — ui O ft ft u i ft ro to U l -4 CO - * ( 0 CO to -s i Ul ft U l Ul O o o -s i U l en ui o -si ao o O O CO (O CO —J -si -si o O O ft ro ui ui -si ro o o ro s j CO -si CO -si —* ft -* to — CD ft - * O -sj - » C D - - C P f t — J U I - ' C T - s l r o f t u i u i * s 4 0 u i t o c o u i u i u i r o O c o u i - s i u i c o u i —« cn cn CD O O u i co co co co oo u i r o O O O - s i - . O f t u i U I C 0 ftUiftOOOoo-*cno r o - s j - t O O O O O O O O C O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O f t t O U I - t U i O O O O C o r o - * c o o ) t o c n u i f t u i u i u i r o t o c Q O a o U I C 0 t O 0 3 U I - * U l f t r O ' , s l U ) - * - s l O - s l - s l U > O a D C 0 a o c o ^ i O O f t f t c o c o c o r o c o < 0 u t - ~ a 3 u i f t r o O t o f t u i t o c o u i r o f t - s i u i ft u i ro - * O CO O U l 03 U l ft U l -si 03 CO r o -si ao - » r o - * - » O 3 u i f t r o - * t 0 c o a a u i u i u i r o u i u i t o c o a o u i o o o o u i - s i t o o ruco — o o o o o o o t o c o c o o o o o o o o o o o o o o o o o o o o o o o o o o o o u i ro u i ro co - s i 03 ro ro ft -si oo O - * ro ro ft ft s i O - s i - * Ul -si CO - » 09 ro ro oo ro U J *si O O O c n r o — - s i u i f t u i O O O O c O r o O — — u i O O O N ^ U I U I Q U I U - t - t - t t o c o m t o o i u i f t ft_co_ft_roa>00 r o - * o u i f t 0 3 u i o o c o - O D N O M L J U l O r U U I - . f t c n - s i o o — o o u i m c SUV3UI ivaA-Aq puv auop-Aq gr xipuaddy SAS VARIABLE MEAN STANDARD DEVIATION MINIMUM VALUE MAXIMUM VALUE STD ERROR OF MEAN 1:03 WEDNESDAY, JULY 12, 19B9 2 SUM VARIANCE C . V . CLONE=127 CONES 35 28 .40000000 30 .45169753 0 .00000000 108 .0000000 5 .14727635 994 .000000 927 .3059 107 224 EVOL 45 819 .89555556 925 .48864832 21 .10000000 3871 .8000000 137 .96370200 36895 .300000 856529. .2382 1 12 .879 LVOL 45 197 14444444 179 .05807280 7 .30000000 707 .6000000 26 .69240151 8871 .500000 32061 .7934 90 .826 TVOL 45 1017 .04444444 1090 .84866515 31 .10000000 4579 .4000000 162 .61411790 45767 .000000 1189950 .8103 107 257 EWT 45 256. 04444444 274 .98264041 8 .00000000 1159 .0000000 40 .99199177 11522 .000000 75615. .4525 107 . 396 LWT 45 105. .82222222 95. .61316210 4 .00000000 378 .0000000 14 .25316867 4762 .000000 9141 . 8768 90. .353 TWT 45 385. .64444444 364 .37300790 13 .00000000 1518 .0000000 54 .31752099 16454 .000000 132787 .6889 99 652 ED 32 0. .31900000 0 .04290725 0 ,23900000 0 .4260000 0 .00758500 10 .208000 0 .0018 13 .451 LD 32 0. .53481250 0 .02160113 0 .50200000 0 .5790000 0 .00381858 17 .114000 0 .0005 4 .039 RD 32 0. .36581250 0 .04518881 0 .29400000 0 .4490000 0 .00798833 11 .706000 0 .0020 12 . 353 CONES 20 32 .40000000 29 .52857670 0 00000000 87 .0000000 6 .60279048 648 .000000 871 .9368 91 . 138 EVOL 30 1245. 14333333 1141 .89003878 46 .50000000 3720 .2000000 208 .44312930 37354 .300000 1303458. 1446 91 691 LVOL 30 232. 43666667 241. 12425111 0 00000000 778 .7000000 44 .02306383 6973 .100000 58140. .9045 103. .738 TVOL 30 1477. 59000000 1351 48418220 54 80000000 4116 .9000000 246 .74612423 44327 .700000 1826509. 4947 91 . 465 EWT 30 393. 26666667 365. .27414702 19 .00000000 1109 .0000000 64 .86388814 11798 .000000 126219. 7195 90 .339 LWT 30 119. .36666667 125 .36745531 0. .00000000 414 . 0000000 22 .88886108 3581 .000000 15718 .9989 105 .027 TWT 30 518. .83333333 470. .33253704 23. .00000000 1408 .0000000 85 .87058002 15565 .000000 221212. .6954 90 652 ED 18 0. 30338889 0 .02490958 0. .26300000 0 .3460000 0 .00587124 5 .461000 0 .0006 8 210 LD 18 0 .50994444 0. .02168619 0. .46200000 0 .5380000 0 .00511148 9 .179000 0. .0005 4. .253 RD IB 0. .33466667 0. .02392144 0. .29400000 0 .3770000 0 .00563834 8 .024000 0. .0006 7 . 148 •a •a 3 & x i G. o CD e> S3 a. i »1 CD Co CL0NE=129 CONES 28 46. . 14285714 39 .39704283 0 ,00000000 133 0000000 7 44534127 1292. .000000 1552. . 1270 85 .381 EVOL 35 1593. .82000000 1482 .36828753 43 .00000000 6419 .8000000 250. 56597306 55776. .700000 2197415 .7399 93 019 LVOL 35 344. .15428571 288 ,80377151 21. 80000000 1208 1000000 48 .81674725 12045 .400000 83407. .6184 83. 917 TVOL 35 1937. .77714288 1727. .49689164 64 .90000000 7059 1000000 292. .00026960 67822 .200000 2984245 .5108 89. . 148 EWT 35 443 .85714286 390 .73271614 14 .00000000 1637 0000000 66. .04588350 15528. .000000 152672. .0555 88 07 1 LWT 35 188 .57142857 158 .35033921 11 .00000000 653, 0000000 26 .42803088 6530 ,000000 24445 .4286 83 .802 TWT 35 643. .48571429 542 .98519071 28 .00000000 2198. 0000000 91. . 77786822 22522 .000000 294811 . 1983 84 379 ED 28 0 .27667857 0 .03670609 0. .22400000 0, 3620000 0 .00693680 7 ,74 7000 0 .0013 13 .267 LD 28 0 .54221429 0 .03382237 0. ,44400000 0. 5960000 0 00639183 15. .182000 0 .0011 6 238 RD 28 0 .33135714 0 ,04357557 0. 26700000 0. 4340000 0. 00823501 9. 278000 0. 0019 13 151 CLONE=133 CONES 38 38 .50000000 42 .83156380 0 .00000000 151 .0000000 7 .13859397 1386 .00000 1834 .5429 111 .251 EVOL 80 1972 .83333333 1919 .94101690 60 .90000000 8276 .1000000 247 .86331947 118370 00000 3686173 , 5084 97 .319 LVOL 60 307 .24500000 308 .87092001 0 . 00000000 1329 .4000000 39 .87506431 18434 .70000 95401 .2452 100 .529 TVOL 60 2280 .08000000 2156 .64627737 81 .00000000 8934 . 9000000 278 .42183720 136804 .80000 4651123 . 1657 94 .586 EWT 60 585 .46666667 521 .77163869 19 00000000 2185 .0000000 67 .36042891 33928 .00000 272245 .6429 92 .273 LWT 80 158 .41666887 161 .22920748 0 .00000000 708 .0000000 20 .81460118 9505 .00000 25994 .8573 101 .775 TWT 60 731 .93333333 660 .90156365 29 . 00000000 2639 .0000000 85 32202498 43918 00000 4367gO .8768 90 .295 EO 51 0 .29274510 0 .03349259 0 .21700000 0 .3740000 0. .00468990 14. 93000 0 .00 11 11 ,441 LD 51 0 .50472549 0 .06061422 0 .22900000 0 5900000 0. 00848769 25. 74 100 0 0037 12. 009 RD 51 0. .32849020 0 .03959968 0 ,24400000 0 .4050000 0. 00554506 16. 75300 0. 0016 12. 055 CD SAS VARIABLE MEAN STANDARD DEVIATION MINIMUM VALUE MAXIMUM VALUE STD ERROR OF MEAN 1:03 WEDNESDAY, JULY 12, SUM VARIANCE CLONE=134 CLONE=139 C . V . CONES 30 16 23333333 17 . 19399120 0. OOOOOOOO 55 . 0000000 3 .13917894 487 .000000 295 .6333 105 918 EVOL 44 1627 .62500000 1413 .72505075 69. 20000000 5738 .2000000 213 . 12707046 71615 .500000 1998618 .5191 86 858 LVOL 44 402. .99545455 307 .04901734 8 20000000 1175 .3000000 46 .28938103 17731 .800000 94279 ,0990 78, 192 TVOL 44 2030 62500000 1851 .84631741 77 40000000 6403 .5000000 249 .02520210 89347 .500000 2728596 .2563 81 347 EWT 46 477. 82608696 424 .76918199 0. oooooooo 1686 .0000000 82 62879911 21980 .000000 180428 .8580 88 896 LWT 46 203 97826087 163 .53490815 0 oooooooo 607 .0000000 24 11190676 9383 000000 26743 6662 80 173 TWT 46 695 .97826087 571 .26921029 0. .oooooooo 2138 .0000000 84 .22904986 32015 000000 328348 5106 82. 081 ED 35 0. 30317143 0 02921068 0. 25100000 0 .3810000 0. .00493751 10 .611000 0 .0009 9 835 LD 35 0. .52551429 0 .05711514 0 25200000 0 .5950000 0 00965422 18 393000 0 .0033 10 868 RD 35 0. 35417143 0 .04005840 0. 29100000 0. .4730000 0 00677111 12. 396000 0 .0018,. 11. 310 CONES 42 117. .02380952 85 82440613 0 OOOOOOOO 355 .0000000 13 24299337 4915 .00000 7365 .8287 73 339 EVOL 57 1670, .74912281 1468 ,02246104 38 .70000000 5100 .6000000 194 44432505 95232 .70000 Z155089 .9461 87, .868 LVOL 57 211 .01929825 195 00553146 0 .oooooooo 746 .0000000 25 .82911362 12028 10000 38027 . 1573 92. .411 TVOL 57 1681, , 77017544 1630 14371397 44 .oooooooo 5668 .3000000 215 .91780958 107260 90000 2657368 .5282 86. .628 EWT 57 477, . 19298246 415 36758413 13 .oooooooo 1457 .ooooooo 55, .01678053 27200 .00000 172530 .2299 87. 044 LWT 57 104, ,80701754 100. 04810247 0 .oooooooo 372 .ooooooo 13 25169490 5974 00000 10009 .6228 95. .459 TWT 57 586 33333333 503. 76607858 18 .oooooooo 1755 .ooooooo 66 72544714 33421. . ooooo 253780 . 2619 85. 918 ED 47 0. .27744681 0. 03342468 0 .21800000 0 ,3390000 0.00487549 13. 04000 0 .0011 12. 047 LD 47 0. .43755319 0. 15644624 0 .oooooooo 0. 5560000 0, 02282003 20, 56500 0 0245 35. 755 RO 47 0, .30038298 0. 03227501 0 .22900000 0. ,3830000 0. 00470779 14 . 11800 0 0010 10. 745 3 a. S' i § p> a a. & > • I a o> fa S3 Cn CL0NE=198 CONES 30 22 .26668667 30 .88271839 0 OOOOOOOO 123 .000000 5 .59822084 668 .000000 940 .2023 137 .707 EVOL 42 1377, .12857143 1974 .48781976 40 .40000000 11046 .600000 304 .67008505 57839 .400000 3898602 . 1504 143 .377 LVOL 42 156, .11428571 169 .46987291 0 oooooooo 821 .600000 26 . 14973825 6556 .800000 28719 .9700 108 .555 TVOL 42 1533 .24047619 2107 .79841287 40 .40000000 11538 .400000 325 .24035615 64398 .100000 4442814 . 1493 137 .473 EWT 42 414. 92857143 573 .26209869 15 .oooooooo 3154 .000000 88 .45626223 17427 .000000 328629 .4338 138 . 159 LWT 42 77. .30952381 85 .39820301 0 .oooooooo 415 .000000 13 ,17722881 3247 .000000 7292 .8531 1 10 .463 TWT 42 497. ,59523810 647 .10515071 15 .oooooooo 3436 .000000 99 .85049253 20899 .000000 418745 .0761 130 046 ED 29 0 30306897 0 .03593042 0 .24700000 0 376000 0 .00667211 8 .789000 0 .0013 1 1 . .856 LD 29 0 48065517 0 .08704813 0. .23100000 0 .591000 0 .01616443 13 .939000 0 .0076 18 . 110 RD 29 0. 32920690 0 .03198814 0 .27400000 0 .394000 0. 00594005 9. 647000 0. .0010 9 .717 CLONE=200 CONES 24 7 .33333333 9 .11122894 0 .oooooooo 28 . OOOOOOO 1 .85982182 176 .000000 83 .0145 124 .244 EVOL 29 1163. .06551724 1156 .84759839 58 .60000000 4592 .2000000 214 .82120600 3372B .900000 1338296. .3659 99 .465 LVOL 29 217. .70344828 170 .32230053 13 .20000000 727. 8000000 31 .62805720 6313 .400000 29009. .6861 78 .238 TVOL 29 1380 .76896552 1301 .02454108 71 .90000000 5071. 5000000 241 .59419213 40042 .300000 1692664. .8565 94 .225 EWT 32 323 18750000 344 .90778922 0 .oooooooo 1313 ooooooo 80 .97185916 10342 .000000 118961. .3831 106 . 721 LWT 32 101 81250000 89 .23272852 0 .oooooooo 375. ooooooo 15, 77426686 3258 .000000 7962. 4798 87 .644 TWT 32 430 .18750000 425 .74825497 0 .oooooooo 1601. ooooooo 75, 26236954 13766 000000 181261. 5766 98 .968 ED 20 0 .29460000 0 .02958558 0 .24100000 0. 3610000 0 00661553 5 .892000 0. 0009 10 .043 LD 20 0. 51830000 0 .02389252 0 .48300000 0 5760000 0 .00534253 10 .366000 0. 0008 4 610 RD 20 0. 33005000 D .02859805 0 .29300000 0. 3820000 0. 00639427 8. .601000 0 0008 8 664 C O OO SAS VARIABLE MEAN STANDARD DEVIATION MINIMUM VALUE MAXIMUM VALUE STD ERROR OF MEAN 1:03 WEDNESDAY. JULY 12, 1989 4 SUM VARIANCE C . V . YEAR=1976 CONES 0 EVOL 3 38 43333333 17 .80149806 20 .50000000 58 .10000000 10 .27769970 115 .30000000 316 .89333333 48 .318 LVOL 3 10 43333333 7 .22864672 5 .30000000 18 .70000000 4 .17346113 31 .30000000 52. .25333333 69 .284 TVOL 3 48 .86666667 23 .87495201 27 .80000000 74 .80000000 13 .78420997 146 .60000000 570 .01333333 48 857 EWT 3 13. .00000000 6 .00000000 7 .00000000 19 00000000 3 46410162 39 .00000000 36. .00000000 46 154 LWT 3 5. .66666667 3 .78593890 3 .00000000 10 .00000000 2 .18581284 17 00000000 14 . 33333333 66 .811 TWT EO LD RO 3 0 0 0 18 33333333 8. ,73689495 11 .00000000 28 .00000000 5 04424865 55 .00000000 76. 33333333 47, 656 YEAR=1977 CONES 0 EVOL 13 90 .39230769 54 .04190444 44 .50000000 240 .00000000 14 .98852750 1175 .1000000 2920 .5274359 59 . 786 LVOL 13 21 .62307692 17 .01167020 7 .80000000 74 .80000000 4. .71818840 281 .1000000 289 .3969231 78 .674 TVOL 13 112 .00000000 89 .22837809 54 .80000000 314 .80000000 19. 20049746 1456 .0000000 4792 .5683333 61 .811 EWT 13 31 .69230769 19 .65275475 17 .00000000 88. .00000000 5. 45069346 412 .0000000 386. .2307692 62 .011 LWT 13 10 .84615385 8 .65877353 4 .00000000 38. .00000000 2 40151169 141 .0000000 74 .9743590 79 .833 TWT 13 43 .00000000 29 .64231210 23 .00000000 134. .00000000 8. 22129817 559 .0000000 878. .6666667 68 .936 ED LD RD 0 0 0 « -a 8 X' 3 o b a> s» >1 3 a fu 13 cn YEAR=1978 CONES 0 EVOL 29 117 .04482759 74, . 14458240 19 .20000000 311 .30000000 13 .78829959 3394 .3000000 5497 .4161330 63 .347 LVOL 29 9 .98620690 11 .82801529 0 .00000000 59 .40000000 2 , 19640730 289 . 6000000 139 9019458 1 18. 444 TVOL 29 127 .03103448 78. 15471882 21 .40000000 320 .90000000 14 .51296694 3683 .9000000 6108. . 1600739 61 . 524 EWT 29 38 .55172414 25.00083742 7 .00000000 98 .00000000 4 .64253896 1118 ,0000000 625 .0418719 64 .850 LWT 29 4 .75862069 8. .33953115 0 .00000000 32 .00000000 1 .17722138 138. .0000000 40 1896552 133 . 222 TWT ED LD RO 29 0 0 0 43 .96551724 27 .79964383 8 .00000000 121 .00000000 5 .16226426 1275 .0000000 772 8201970 63 .231 YEAR=1979 CONES 0 EVOL 39 251. ,12051282 193 .94881254 24 .70000000 LVOL 39 46. .74615385 28 .20045044 B .20000000 TVOL 39 297. 85897436 212 .93914465 39 30000000 EWT 39 82. 48717949 62 .77271943 9 .00000000 LWT 39 24 .23076923 14 .20796948 4 .00000000 TWT 39 108. 56410258 75 .80590163 17 .00000000 EO 18 0. 34311111 0.01997024 0 .31700000 LD 18 0. 41816667 0 . 19313246 0 00000000 RD 18 0. 36961111 0 .02084999 0 .32900000 968 .0000000 31 .05863326 9793 .700000 37616 .084305 77 .233 1 16 .6000000 4 . 19542976 1823 .100000 686 .463603 56 .048 1084 .5000000 34 .09755211 11616 .500000 45343 .079325 71 . 490 305 .0000000 10 .05167967 3217 .000000 3940 .414305 78 . 100 61 .0000000 2 .27509592 945 .000000 201 .866397 58 638 392 .0000000 12 . 10663345 4234 .000000 5716 .252362 69. 642 0 .3670000 0 00470703 6 .176000 0 .000399 5. .820 0 .5240000 0 .04552176 7 .527000 0 .037300 46 . 186 0 .3910000 0 00491439 6 .653000 0 000435 5. 641 CO CD SAS VARIABLE N MEAN STANOARD MINIMUM MAXIMUM DEVIATION VALUE VALUE YEAfl=19BO CONES 59 1 .00000000 1 .92084751 0 .00000000 10 . 0000000 EVOL 55 408 .68545455 302 .93519990 48 .80000000 1247 .1000000 LVOL 55 82. , 30000000 49 .54961599 2 .30000000 192 .5000000 TVOL 55 490 .98909091 340 .79214892 51 .10000000 1328 .6000000 EWT 59 122 40677966 101 .45495325 0 .00000000 399 .0000000 LWT 59 40 .05084746 27 .50512412 0 .00000000 94 .0000000 TWT 59 167 52542373 127 .49744500 0 .00000000 487 .0000000 ED 14 0 32735714 0 .03007070 0 .28500000 0 .4100000 LD 14 0 .49600000 0 .02031294 0 .45900000 0 .5260000 RO 14 0 35635714 0 .02895250 0 .29700000 0 .4200000 CONES 5B 17. .25882069 19 .15455179 0 . 00000000 118 ,0000000 EVOL 58 923 .00344828 615 .29708385 82 .90000000 2597 .7000000 LVOL 58 93. .03793103 72 .76191607 3 .40000000 297 .5000000 TVOL 58 1016. 04655172 653 .29306801 85 .60000000 2745 .0000000 EWT 59 277. . 13559322 189 .42093771 0 .00000000 783 .0000000 LWT 59 44 86440678 38 .31739392 0 .00000000 155 .0000000 TWT 59 330. ,55932203 219 .68302174 0 oooooooo 868 .0000000 ED 44 0, 313BSB36 0 .03386239 0 .26700000 0 4260000 LD 44 0 45702273 0 . 10376998 0. oooooooo 0 .5460000 RO 44 0. 33447727 0. .04071197 0. ,27400000 0 ,4490000 CONES 59 30 ,83050847 31 .84507353 1. oooooooo 152. .0000000 EVOL 59 671, .64237288 421 .20116588 40 .40000000 2620. ,8000000 LVOL 59 172. .23898305 90 .50232806 0. ,oooooooo 382. ,2000000 TVOL 59 843. 87966102 459 .35201729 40. .40000000 2882 .3000000 EWT 59 219. 54237288 130 .14982375 15. .oooooooo 764. , 0000000 LWT 59 88. 37288138 49 .58322966 0. .oooooooo 201. . 0000000 TWT 59 313 88135593 158 .45874964 15. oooooooo 972. .0000000 EO 52 0 33026923 0. .02287229 0. 2B700000 0. 3850000 LD 52 0. 47773077 0 . 10331906 0. oooooooo 0. .5640000 RO 52 0. .37117308 0 .03276280 0 .29900000 0. .4730000 CONES 59 79. .35593220 58, .47268571 0 oooooooo 210. ,0000000 EVOL 59 1231 .75593220 582 .23543445 126. oooooooo 2902. , 1000000 LVOL 59 319. . 10508475 142 .29997265 39. 90000000 671. 5000000 TVOL 59 1550. 88271186 880. .03015955 165. .90000000 3374, ,5000000 EWT 59 379. .64408780 172. .46836419 41. .oooooooo 861. ,0000000 LWT 59 172. 15254237 80 31464279 20. oooooooo 373. 0000000 TWT 59 552. 33898305 226 .02668994 62. oooooooo 1104, 0000000 EO 56 0. 30576788 0 .02661100 0. 25500000 0. 3700000 LD 58 0. 52933929 0. .04744654 0. 26300000 0. 5960000 RD 58 0. 34907143 0. .03215240 0. 27800000 0. 4270000 1:03 WEDNESDAY, JULY 12, 1989 5 STD ERROR SUM VARIANCE C.V. OF MEAN 0 .25007305 59 .000000 3 .88966 192 .085 40 .84777402 22477 .700000 91769 .73534 74 . 124 6 .68126886 4526 .500000 2455 .16444 60 .206 45 .95240399 27004 .400000 116139 .28877 69 .409 13 .20B3O986 7222 .000000 10293 .10754 82 .883 3 .58086216 2363 .000000 756 .53185 68 .676 16 .59875352 9884 .000000 16255 .59848 76 . 106 0 .00803673 4 .583000 0 .00090 9 . 186 0 005428B6 6 .944000 0 .00041 4 .095 0 .00773788 4 .989000 0 .00084 6 . 125 2 .51511587 1001 .000000 366 .89685 1 10 .985 80 .79246523 53534 .200000 378590 .50139 66 .662 9 .55410765 5396 .200000 5294 .29643 78 . 207 85 78158237 58930 .700000 426791 .83271 64 . 298 24 .86050580 16351 .000000 35880 .29164 68 . 350 4 .98849979 2647 .000000 1468 .22268 85 . 407 28 .60029336 19503 .000000 48260 .63004 68 .458 0 005104g5 13 .811000 0 .00115 10 788 0 .01564391 20 . 109000 0 .01077 22 706 0 .00813756 14 .717000 0 .00166 12, 172 4 . 14587544 1819 .000000 1014 , 1087 1 103. .291 54 .83572109 39626 .900000 177410 .42214 82. 712 11 .78239953 10162 .100000 8190 .67139 52. .545 59 80253889 49788 .900000 211004 .27579 54 .433 1B. .94406382 12953 .000000 16938 .97662 59 282 8 45258289 5214 .000000 2456 .51373 56 084 20 62957205 18519 .000000 25109 . 17534 50 484 0 00317182 17 .174000 0. 00052 6. 925 0 .01432778 24. .842000 0. 01067 21 627 0. 00454338 19. .301000 0. .00107 8 827 7. 35211745 4682. .000000 31B9. .16423 71 164 75 80059715 72673 .600000 338998. .10113 47 .269 IB. 52587847 18827. .200000 20249. 2B222 44 .593 B8. 53238591 91500, .900000 482441. .01790 43 849 22. 45346851 22399. .000000 29745 , 33665 45 429 10. 45607588 10157. 000000 6450. 44185 48. 653 29. 42616601 32588. 000000 5108B 05552 40. 922 0. 00355604 17. 123000 0. 0007 1 8. 703 0. 00634031 29. 643000 0. 00225 8. 963 0. 00429654 19. 548000 0. 00103 9. 211 SAS VARIABLE MEAN STANDARD DEVIATION MINIMUM VALUE MAXIMUM VALUE STD ERROR OF MEAN 1 : 0 3 WEDNESDAY, JULY 1 2 , SUM VARIANCE 1989 -a CD 13 ex X YEAR=1984 CONES 59 81 .71186441 53 .498 71303 2 .OOOOOOOO 202 .OOOOOOO 8 .96467881 3641 . 00000 2861 .89831 88 .688 EVOl 59 1613 .29361017 841 93364141 302 .70000000 3924 . 3000000 109 .61042389 95184 .50000 708852 .25654 52 . 187 LVOL 59 181 . 16440678 89 . 12223640 34 .60000000 389 6000000 11 .60272690 10688 .70000 7942 .77302 49 194 TVOL 59 1794 45762712 901 .08680840 385 .40000000 4313 .9000000 117 .31151029 105873 .00000 811957 .43628 50 215 EWT 59 431 38983051 245 .79126 391 104 oooooooo 1261 OOOOOOO 31 .99929698 27222 .00000 60413 .34541 53 272 LWT 59 96 .54237288 48 .54728304 17 .oooooooo 210 OOOOOOO 8 .32031791 5696 .ooooo 2356 .83869 50 286 TWT 59 566 .22033896 291 .31367546 150 .oooooooo 1608 OOOOOOO 37 .92581016 334C7 ooooo 84863 .65751 51 449 ED 59 0 .27577966 0 .03029343 0 .21300000 0 .3660000 0. 00394387 16 .33000 0 00092 10 945 LD 59 0 53108475 0 .04900010 D 23900000 0 .5950000 0 .00637927 31 33400 0 .00240 9. 226 RO 59 0 30462712 0 .03534212 0 .22900000 0 4080000 0 .00430115 17 .97 300 0 .00125 11 602 YEAR=1985 CONES 39 125 .48717949 60 .19480297 13 .OOOOOOOO 246 OOOOOOO 9 .63888267 4894 .00000 3623 .4143 47 969 EVOL 59 2908 .91186441 1810 .93039443 462 .OOOOOOOO 8276 .1000000 235 . 76305591 171625 .80000 3279468 ,8935 62 . 255 LVOL 59 388 .42203390 152 .15159077 114 .70000000 738 .3000000 19 .80844991 22916 .90000 23150 . 1066 39 . 172 TVOL 59 3297 .33220339 1923 .29198273 621 .70000000 8934. 9000000 250. .39128873 194542 .60000 3699052 ,0508 58 . 329 EWT 59 827. .76271188 513. .31362125 148 . OOOOOOOO 2185 .OOOOOOO 66 .82774134 48838 OOOOO 263490 .8738 62 .012 LWT 59 203 .55932203 82. .43477488 56 .OOOOOOOO 389 ooooooo 10 73209357 12010 .00000 6795 .4921 40 .497 TWT 59 1078. .93220339 818 .12884675 215 .OOOOOOOO 2639 ooooooo 80. .47352140 63657 ooooo 382083 ,2712 57 .291 EO 59 0 .26213559 0. .02626093 0 .21700000 0 .3300000 a .00341888 15 .46600 0 .0007 10 018 LD 59 0 .52894915 0 .02986272 D 42400000 0 5840000 0 00386176 31 .20800 0 .0009 5 .608 RD 59 0. .31052542 0 .03200989 0 24400000 0 3920000 0 00416733 18 .32100 0 0010 10. .308 to i a. o b CD t» !3 CL i fa CD fa S3 to YEAR=1986 CONES 0 EVOL 59 3833 .87457827 1686. .69363770 956 .8000000 11048 .600000 219 .58880785 226198 .60000 2844935 .4274 43 994 LVOL 59 536 . 10508475 223 .20416232 105 .3000000 1208 . 100000 29 .05870682 31630 .20000 49820 .0981 41 .634 TVOL 59 4369 .98644068 1732 . 14969922 1062. .1000000 11538 .400000 225 .50668300 257829 .20000 3000342 .5805 39 .837 EWT 59 1095 .35593220 463. .74687517 322. . OOOOOOO 3154 000000 60 .37470065 64626 OOOOO 215061 . 1642 42 .338 LWT 59 276 .89830508 121 .99034655 50 OOOOOOO 653.000000 15 .88179037 16337 ,OOOOO 148B1 .6447 44 .056 TWT 59 1375 .62711864 499. .15079258 370, ooooooo 3436. .000000 64 .98389810 81162 .OOOOO 249151 .5137 36 285 EO 59 0 .27276271 0. .02253882 0, .2270000 0 .342000 0 .00293404 16 .09300 0 0005 8 262 LD 59 0. .51676271 0. .02219222 0. 4750000 0 569000 0 .00288918 30 48900 0 .0005 4. .294 RD 59 0. ,305084 75 0. ,02811136 0 2530000 0. .381000 0 .00339941 18 OOOOO 0 .0007 8 559 YEAR=1987 CONES 59 116, .28813559 92 .08898449 5 .OOOOOOO 355.0000000 11 .98896460 6861 .OOOOO 8480 .3811 79 . 190 EVOL 59 3048. 45254237 1033 .95727457 1066 . 2000000 5813 .8000000 134 .60977158 179740 .70000 1069067 .6456 33 .940 LVOL 59 641. 80847458 271 . 12863869 124 . OOOOOOO 1329 .4000000 35 .29794221 37866 .70000 73510 . 7387 42 .244 TVOL 59 3688. .26271188 1231 .49410206 1252 .2000000 6634 .1000000 160. .32687603 217607 .50000 1516577 .7234 33 . 390 EWT 59 897. 93220339 293 .34028755 361 .OOOOOOO 1723 OOOOOOO 38. . 18964998 52978 ooooo 86048 .5126 32 .668 LWT 59 327 .76271188 147 .07098204 42 .ooooooo 708 OOOOOOO 19. . 14701099 19338 ooooo 21629 .8738 44 .871 TWT 59 1228. .38983051 403. ,34117572 445 ooooooo 2135 OOOOOOO 52. .51054842 72357 ,ooooo 162684 . 1040 32 .888 ED 59 0. .27801695 0 .02129553 0 .2400000 0 34 70000 0 00277244 16 40300 0. ,0005 7 660 LO 59 0. 50571186 0 .05629205 0 .2370000 0. 5750000 0. 00732880 29 .83700 0 .0032 11 131 RD 59 0. 31681017 0. ,02777311 0 .2450000 0 3800000 0. 00361575 18. 68000 0. 0008 8. 772 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0097545/manifest

Comment

Related Items