Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Weight-length relationships of coniferous wood tracheid skeletons Sastry, Cherla Bhaskararama 1971-04-21

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-UBC_1971_A1 S38.pdf [ 7.39MB ]
Metadata
JSON: 831-1.0075375.json
JSON-LD: 831-1.0075375-ld.json
RDF/XML (Pretty): 831-1.0075375-rdf.xml
RDF/JSON: 831-1.0075375-rdf.json
Turtle: 831-1.0075375-turtle.txt
N-Triples: 831-1.0075375-rdf-ntriples.txt
Original Record: 831-1.0075375-source.json
Full Text
831-1.0075375-fulltext.txt
Citation
831-1.0075375.ris

Full Text

WEIGHT-LENGTH RELATIONSHIPS OF CONIFEROUS WOOD TRACHEID SKELETONS by CHERLA BHASKARA RAMA SASTRY M. Sc. (Botany)' Andhra University, 1958 M. Sc. (Forestry) University of British Columbia, 1967 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of Forestry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH April, 1971 COLUMBIA 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 representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver 8, Canada ABSTRACT The hypothesis is examined that, among individual tracheids of coniferous woods, the weights of holocellulose and alpha-cellulose skeletons are a direct quantitative function of their length, independent of species. A total of 28 annual increments, representing wood from nine coniferous genera and seven families, separated into earlywood and latewood fractions, was delignified with peracetic acid and subsequent reduction with NaBH^. Further reduction to alpha-cellulose followed for a portion of the holocullulose skeletons. About 750 individual holo- and alpha-cellulose trac heid skeletons were measured for length and weight. A specially developed quartz ultra micro-balance, having a weighing range of 0.06 to 14 yg and a precision ± 0.03 yg, was constructed and used to weigh; individual tracheids. Statistical analyses indicated a significant positive curvilinear relationship between legnth of tracheid and weight of its carbohydrate fraction. Estimated variations accounted for in holocellulose and alpha-cellulose skeleton weights, by the length factor.; alone were, respectively, 91.9 and 95.7 per cent for pooled data of all the species. No significant dif ferences in holocellulose skeleton weights were evident within species for the same tracheid length, whereas weights of alpha-cellulose skeletons within species, and both the holo- and alpha-cellulose between species, differed significantly. Radial variation for single tracheid weights followed trends similar to those established by others for specific gravity, and percentage of cellulose based on gross wood analy ses. Individual tracheids of juvenile wood had significantly lower (1% level) alpha-cellulose skeleton weights than those from mature and overmature wood, while differences were non significant for holocellulose. Overmature wood tracheids were significantly lighter (carbohydrate skeleton weight) than those from mature wood, for the same tracheid length. Differences between earlywood and latewood were explored. For the same tracheid length, both earlywood and latewood trac heids contained similar amounts of alpha-cellulose,whereas the amount of holocellulose per tracheid was higher in latewood. Examination of compression wood also provided positive evidence for the length-weight relationship in tracheids. When weights of compression wood tracheids were compared with those of regular (normal) wood, no significant differences were apparent for holocellulose tracheid skeletons, whereas signifi cant differences were found for alpha-cellulose. It was con cluded that, for the same tracheid length, compression wood tracheids may.have a lower amount of alpha-cellulose than those from mature (normal) wood, but a higher amount than those from juvenile wood. Changes in Douglas-fir tracheid weights were studied in wood formed before and following tree fertilization. Variations, for the most part, were found to be associated with changes in tracheid length. Qualitative differences attri butable to treatment composition (Urea vs. NPK vs. (NH^^SO^) were also noted, in that some treatments resulted in less weight of alpha-cellulose per unit length of tracheid, when compared with normal wood tracheids. This reduction in cellulose frac tion was suggested as a possible factor for differences observed in gross wood specific gravity in wood of some fertilized trees. Results were compatible with the proposed hypothesis. iv TABLE OF CONTENTS Page ABSTRACT i TABLE OF CONTENTS iv LIST OF TABLES vi LIST,OF FIGURES viii LIST OF APPENDICES x ACKNOWLEDGEMENT xCHAPTER I. INTRODUCTION 1 II. LITERATURE REVIEW 4 Tracheid MorphologyTracheid Length Variation 11 Dimensional Relationships in Tracheids 15 Chemical Composition and Recovery of Carbohydrate Fraction 19 Estimates of Carbohydrate Fraction and Weight in Relation to Cellular Elements 26 Effects of Fertilization on Wood Properties and Chemical Composition 29 III. MATERIALS AND METHODS 3 5 Selection of Material 5 Wood Sample and Tracheid Skeleton Preparation 37 Analysis of Tracheid Skeletons 40 1. Sampling criterion 42. Tracheid length 1 3. Tracheid weight 42 V CHAPTER Page Statistical Analyses 45 IV. RESULTS AND DISCUSSION 7 Relationship Between Tracheid Length and Weight 48 1. Variation within species 42. Variation between species 52 3. Variation with increasing age 55 4. Variation between regular (normal) and compression wood 60 5. . Variation between earlywood and-latewood 64 Evidence for a Common Relationship in Conifers 68 Applicability of the Relationship 72 Some Practical Considerations 8V. RECOMMENDATIONS FOR FURTHER RESEARCH 8 6 VI. CONCLUSION 8 8 LITERATURE CITED 91 TABLES 105 FOREWORD TO FIGURES 120 FIGURES 121 APPENDICES 139 vi LIST OF TABLES NUMBER Page I. Single tracheid (holocellulose skeleton) weights and lengths for nine coniferous woods 105 II. Single tracheid (alpha-cellulose skeleton) weights and lengths for nine coniferous woods 107 III. Single tracheid (holocellulose skeleton) weights and-lengths for two Douglas-fir trees 109 IV. Single tracheid (alpha-cellulose skeleton) weights and lengths for two Douglas-fir trees 110 V. Analysis model (A) and (B) results of multiple curvilinear covariance analyses; differences among species (holocellulose skeletons). (Adjusted for differences in tracheid length) 111 VI. Analysis model (A) and (B) results of multiple curvilinear covariance analyses; differences among species (alpha-cellulose skeletons). (Adjusted for differences in tracheid length) 112 VII. Multiple curvilinear covariance analysis for tracheid weights (holocellulose skeletons) in juvenile, mature and overmature wood of a 500^year-old Douglas-fir tree.. (Adjusted for differences in tracheid length) 113 VIII. Multiple curvilinear covariance analysis for tracheid weights (alpha-cellulose skeletons) in juvenile, mature and overmature wood of a 500-year-old Douglas-fir tree. (Adjusted for differences in tracheid length) 114 IX. Analysis of covariance and adjusted mean values of earlywood and latewood tracheid weights in six Douglas-fir trees. (Adjusted for dif ferences in tracheid length) 115 X. Analysis of•covariance and adjusted mean values of earlywood and latewood tracheid weights in nine coniferous species. (Adjusted for dif ferences in tracheid length) 116 XI. Selected wood and tracheid properties of three fertilized Douglas-fir trees 117 vii NUMBER XII. Page Results of multiple curvilinear covariance analyses; differences within species in treated and untreated Douglas-fir. (Adjusted for differences in tracheid length) 118 viii LIST OF FIGURES NUMBER Page 1. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples, from Araucaria cunninghamii 121 2. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples~from Sequoia sempervirens 122 3. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples^" from Pinus lambertiana . 123 4. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples"'.from Podocarpus dacrydioides 124 5. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples.: from Picea sitchensis 125 6. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B)" skeletons) and length for samplesffrom Taxus brevifolia , 126 7. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples-" from Juniperus virginiana 127 8. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples." from Cephalotaxus wilsoniana 128 9. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Pseudotsuga menziesii 129 ix NUMBER Page 10. Relationship between tracheid weight (holo cellulose . (A) and alpha-cellulose (B) skeletons and - length for samples from nine coniferous species. (Least-squares fitted line) 130 11. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons and length for samples from nine coniferous species. (The line of best fit conditioned to pass through the origin) 131 12. Tracheid weight, mean tracheid length and specific gravity patterns, across the butt log of a 500-year-old Douglas-fir tree. (Each point repre sents the average tracheid weight or length for the increment. Specific gravity - values from Kennedy and Warren (70)) 132 13. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length for compression wood and normal wood of Douglas-fir. 133 14. Tracheid weight, increment width and mean tracheid length patterns across six consecutive incre ments of a fertilized (NH4NO3) Douglas-fir which produced compression wood 4 years after first fertilization. (Each point represents the average tracheid weight or length for the increment) 134 15. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length in Urea treated and normal Douglas-fir 135 16. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length in NPK treated and normal Douglas-fir 136 17. Relationship between tracheid weight (holo cellulose (A) and alpha-cellulose (B) skeletons) and length in (NB^^SO^ treated and normal Douglas-fir 137 18. Relationship between tracheid weight (holo cellulose ,-(A) and alpha-cellulose (B) skeletons) and length in NH4NO3 treated and normal Douglas-fir 13 8 X LIST OF APPENDICES NUMBER Page I. Description of growth increments included in the study 13 9 II. Summary of holocellulose and alpha-cellulose yields and estimates of lignin in some of the samples studied (based on ovendry, extractive-free wood) 141 III. Moisture content of the holocellulose and alpha-cellulose pulps after conditioning in the CTH Room for over one month 142 xi ACKNOWLEDGEMENT The writer acknowledges the continuous guidance, inspiration and constructive criticism given to him during his entire stay at the University of British Columbia, by his supervisor, Dr. R. W. Wellwood. Without the interest shown by him the thesis might not have been undertaken or completed. The help and kindness of Dr. J.W. Wilson is acknow ledged, not only for his many useful suggestions during the planning and experimental phases of the thesis, but also for his truly enlightening inspiration, encouragement and personal concern. Dr. A. Kozak helped the writer in clarifying many questions concerning statistical analyses. His advice is gratefully acknowledged. The thesis was reviewed by members of the advisory committee consisting of Drs. R.W. .Kennedy.,. A. .-Kozak, RvW.-. Wellwood, J.W. Wilson,.J. Worrall and D.J. Wort. For their valuable suggestions,comments and constructive criticism the writer is indebted. Many helpful suggestions from Drs. O.H. Lowry, Washington University School of Medicine, St. Louis, D.S. Skene, Cabot Foundation for Botanical Research, Harvard Forest, Peter sham, B.S. Wenger, Department of Anatomy, School of Medicine, University of Saskatchewan, and the time and talents of many xii persons on the Faculty of Forestry technical staff, Messrs. U. Rumma, K. Apt, and G. Bohnenkamp, made possible develop ment of the quartz ultra-microbalance, an integral part of the s tudy. Discussions with the following contributed greatly, in understanding some of the problems involved herein and their advice is appreciated. Dr. R.W. Kennedy Department of Fisheries and Forestry, Canadian Forestry Service Forest Products Laboratory Vancouver. Dr. L. Paszner Faculty of Forestry, University of British Columbia. Dr. J. Worrall Faculty of Forestry, University of British Columbia. Dr. D.J. Wort Department of Botany, University of British Columbia.-The writer sincerely thanks Miss L. Cowdell and Mrs. K. Hejjas for writing the computer programs; Mrs. M. Lambden for drafting work; Messrs H.J. Cho and C.P. Chen, Graduate Students, for technical assistance in the preparation of pulp samples; the National Research Council of Canada and the University.of British Columbia for financial assistance. Finally, he must express his special appreciation to his wife, Ratna, for her encouragement and monumental patience. CHAPTER I INTRODUCTION Much current research in wood and pulp science emphasizes variability at the cellular level. The principal cellular elements of coniferous woods are longitudinal tracheids which make up as much as 90 per cent by volume or 98 per cent by weight of the ovendry material. Since the tracheid is the basic unit of such woods, its isolation and study should assist in understanding and interpreting the biological activity and product behaviour of wood. Variability in wood within and between species has been studied extensively at the gross level. Similarly, varia tion within a single stem has been studied with respect to radial and height patterns, as well as wood zones (corewood, heartwood and sapwood). Less frequently, separate earlywood and latewood segments have been analyzed for differences in physical behaviour. At still another level of investigation, the variability in physical and chemical properties within individual growth increments has been examined by intra-incre-mental analyses. Further intensive study occurred at the individual cell level, in particular cell wall organization and chemical composition including spatial distribution of the constituents within cell walls. 2. Even though chemical constituents and physical proper ties of wood have been intensively studied, relatively little is known about the amount of carbohydrate fraction per indi vidual tracheid, and possible relation to an easily measured anatomical characteristic,for example, length. Though such information has been deduced by indirect methods such as fibre coarseness, or tissue fractions, these methods have an inherent weakness in that the estimate of individual cellular variation is not possible. Scientists in various segments of the wood industries are emphasizing methods of defining wood quality. These efforts have lead to a search for an anatomical basis for expression of wood quality. Dimensions of tracheids and their carbohydrate content are important because each is generally known to be closely related to gross wood characteristics that are used as quality indicators, such as density, and various strength properties. Also, the influence of tracheid length and strength of individual fibre1 on pulp and paper properties has been established. The significance of cellulose content on tensile strength of wood has been confirmed by various workers. Certainly, a thorough knowledge of the relationship between length of fibre and its carbohydrate amount, within and between species, is necessary in order to more fully evaluate and understand gross wood "*"The word 'fibre' and 'tracheid' are used inter changeably in this thesis. properties and to more fully utilize wood. In addition, know ledge regarding the type of relationship (i.e., linear or exponential), would be of importance to geneticists who are endeavouring to manipulate wood characteristics through tree breeding work. Even more important, any gains in this funda mental knowledge should be of general value to wood-based technologies. The objectives of this thesis are: 1. To test the hypothesis that, among individual tracheids of coniferous woods, the weights of holo-cullulose and alpha-cellulose skeletons are a direct quantitative function of their length, independent of species. 2. a. To apply the relationship within a single species, Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) by comparing regular (normal) wood with juvenile wood, overmature wood, compression wood, and wood formed within trees treated with chemical fer tilizers . b. To examine the relationship within increments by comparing individual carbohydrate skeletons from earlywood and latewood. CHAPTER II LITERATURE REVIEW During the past 40 to 50 years data regarding the anatomical structure, physical properties, chemical composition, and physical functioning of wood have accumulated in ever-increasing volume. Several books have been written on the subject within recent years (45, 62, 72, 101, 137) and these have been supplemented by review papers (19, 20, 25, 31, 34, 70, 127, 136, 153, 168). It would seem superfluous to re-cover the same ground. An attempt will be made to cover only those points of literature which are related directly to the thesis topic. Tracheid Morphology As stated by Wardrop (145), the term morphology was introduced originally by Goethe early in the last century to define that field of study concerned with plant and animal form, as distinct from the study of function. With time, the scope of the term has been widened to include the study of both structure and form, and more recently to include the study of structure even at the molecular level. Therefore, discussion is concerned not only with wood fibre external geometry but also with internal structure (where appropriate) and, briefly with the physiology of wood formation. 5. The production of wood results from activities of the vascular cambium which becomes circumferentially continuous a short distance below the growing tip of the main stem and the branches of a tree (45). There are two cell types in the vascular cambium: 1) fusiform initials; and 2) ray initials. Fusiform initials are elongated cells with tapering ends,whereas the ray initials are much smaller and almost isodiametric in shape (19). The former initials give rise to all the vertical elongate elements of the xylem and phloem, whereas the ray initials produce all elements of the transverse or ray system of the xylem and phloem (45). Two types of cell division occur in the vascular cambium of conifers; periclinal or additive and anticlinal or multiplicative. Periclinal divisions add to the xylem and phloem hence bringir. about radial growth, while the latter type result in multiplication of fusiform cambial cells with conse quent accommodation to circumferential expansion (8). Because the fusiform initials give rise to all the vertical elongate ele ments of the xylem and phloem, their initial and final dimensions, and the many physiological factors affecting the dimensions, are of great importance to the kind and quality of wood produced. This is particularly important to conifers since the length of derived vascular elements is traceable back to cambial initials 6. from which tracheids were derived, i.e., post-divisional elon gation of daughter cells is relatively slight in conifers (2, .8) . The pattern of secondary wall formation in fibres has been extensively studied by Australian workers, and Wardrop, in two recent articles, has summarized the work in this area (143, 144). The discussion that follows is a brief synthesis of the excellent and extensive work of Wardrop and his associates (143, 144, 145, 146, 147, 148). Development of tracheids from the cambium of conifer stems involves four phases: 1) cell division in the cambium; 2) enlargement of the daughter cells to the dimen sions of the mature cell; 3) the development of the.secondary wall; and 4) lignification (147). Daughter cells, formed as a result of cell division in the cam bium, immediately undergo considerable dimensional change, increasing radially by 200 to 500 per cent and in length by 10 to 20 per cent or more. Development of secondary wall begins, in part, before dimensional changes of differentiation are complete, first near the middle of the differentiating tracheids then extends toward their tips. During the progres sive development of the secondary wall, contact between the 7. primary wall and the protoplast is maintained by a minute terminal canal between the lumen and the primary wall (147). Wardrop and Dadswell (146) reported that the thickening of the cell wall in mature tracheids was often less at the tips than in the middle. This observation, besides supporting the pro gressive development of secondary wall from the middle towards the cell tip, also indicates the possibility that the tips per se should weigh : less compared to a similar portion of unit length from the tracheid body. The various factors which influence the extent of cell enlargement and wall development following cell division are still not fully understood. Cell enlargement apparently is a function of cell vacuolation and it has been suggested that this is related to the water regime of the trees at the time that cell enlargement occurs (14). As noted by Wardrop (148) : In the living condition the cells of the cambium exist in a state of turgor and the capacity of the cells to take up water is de termined by the difference between the actual pressure mutually exerted on one another by the wall and protoplast and the full (poten tial) pressure available through osmotic activity of substances in the cell sap. Cell elongation of fusiform initials is achieved by intrusive growth, i.e., tips of the elongating cell extend at one or both ends, between the other developing cells (2, 9, 45, 101). According to current concepts, radial expansion of the tracheid is regulated by auxin emanating from growth centers 8. in the crown (77, 80). In addition, Roelofsen (115) formulated the multinet theory for the growth of the primary wall. This theory implies that the innermost part of the wall consists of bundles of more or less transversely oriented microfibrils. During cell expansion, the fibrils are passively displaced and slip along each other as this membrane sheath is moved outward by the apposition of new wall material, and the strands become subsequently oriented in the direction of growth. However, the multinet theory does not offer an explanation for the transverse orientation of the fibrils prior to cell wall expansion. From the standpoint of its development, the mature conifer tracheid consists of two structures: the primary wall and the secondary wall. The presently accepted concept and terminology of cell wall structure.was developed by Kerr and Bailey (67). They divided the cell wall into the primary wall (cambial wall) and the secondary wall (secondary thickening laid down after the cell has reached its final size and shape). The secondary wall consists of three layers—an outer layer, a middle S^r and an inner S^--of differing optical properties (145). This optical heterogeneity has been attributed to dif-. ferences in helical orientation of the cellulose micellar strands in the three layers (143, 144>r145). In conifers, the innermost layer may be sculptured in the form of small excrescences (wart structures) (85, 145), possibly formed from protoplasmic debris. These aspects have been extensively 9. reviewed by Wardrop and his associates who developed a schematic model of the cell wall (142, 143, 144). In addition, Dunning (36) recently developed a.new method for microdissection and electron microscopy of single wood pulp fibres, and provided additional information on cell wall organization, in particular the layer. From studies of Jayme and Fengel (63) and several other workers referred to by them, it is understood that, for both earlywood and latewood, the S2 layer.is by.far the thickest and probably accounts for 80 to 95 per cent of the volume or weight of the fibre. Variations do occur in the wall organization, as in the case of reaction wood anatomy, and it is reported that even in normal wood some Picea spp. lack an layer (53) . Compres sion wood tracheids differ in cross section from normal ones in their typical rounded appearance and intercellular spaces. They usually have ,a wider layer and numerous regular radial discontinuities.in the S^r as well as a much greater micro fibrillar angle than normal wood tracheids. In addition, com pression wood tracheids contain an extra layer of lignin located between and (136). According to Timell (13 6), tension wood often lacks one or several of the three secondary wall layers and frequently possesses an additional wall layer, deposited next to the lumen, the so-called gelatinous layer. This contains little if any lignin and consists largely of highly crystalline microfibrils of cellulose, oriented parallel to the fibre axis "(136). Similarly, the morphology of tracheids, 10. i.e., dimensions and thickness of wall layers, differs between earlywood and latewood of the same growth increment (137). There is an indication, however, that the cross-sectional cell wall area of earlywood and of latewood tracheids may remain constant in conifers (14, 68). According to Wardrop (143) and as summarized by Stewart (127), the cell wall chemical constituents may be classified into three groups: 1) the matrix substances; 2) . the framework substances; and 3) the encrusting substances. The framework (cellulosic) and matrix (non-cellulosic) polysaccharides are deposited before the primary (lignin) and secondary (extractives) encrusting substances. The distribution of the chemical constituents is uneven throughout the cell wall. Bailey (1) reported the iso lation of the middle lamella of Douglas-fir by microdissection and indicated that it contained 71 per cent lignin, and 14 per cent pentosans. Lange and his associate.'. (75) applied a micro-spectrophotometric method on cross sections of wood samples of Swedish spruce (Picea abies (L.) Karst), and confirmed the earlier findings of Bailey. In addition, it was noted that the lignin concentration external to the layer is 73 per cent, compared with 16 per cent near the lumen. Berlyn and Mark (11) reviewed this aspect and concluded that although most of the compound middle lamella, is ligniri;-.this does not necessarily -mean that most-of the-lignin ' (amount).:.is in this region. 11. Information on the spatial distribution of poly saccharides in the cell wall is limited. The proportion of individual polysaccharides in each cell wall layer had been deduced very roughly by following microdissection techniques (90, 91) or by microspectrophotometric methods (74). In their recent book, Panshin and-de Zeeuw (101) summarized the work in this area as follows: ...less than 10 percent of the primary wall is cellulosic; in the S? layer the cellulose increases to more than 50 per cent of the material in that cell wall layer; a reduction in amount of cellulose-is "evident in the S^. It is clear from these studies that cellulose is the major component in the secondary wall, whereas lignin and non-cellulosic polysaccharides are dominant ' in the primary wall (145). Furthermore, a comparison of these.values with those of the lignin indicate that percentage of lignin is re lated inversely to that of the total polysaccharides (101). Thus .the work done so far has centred mostly on the spatial distribution.of components in the tracheid, i.e., qualitative aspects, but very little attempt has been made to quantify the various components per tracheid. Tracheid Length Variation An extensive literature has accumulated on tracheid length variations both within and between trees and on probable effects of the external environment. The pioneering work in this area is traceable to Sanio's first investigations (4) of tracheid length variation in Scotch pine (Pinus syIvestris L.). Within the present century the matter of fibre length has be come important to the pulp and paper industry, with a conse quent expansion.of information on both conifers and hardwoods (30, 31, 122). As a result, sufficient information is avail able on the patterns of tracheid length variation within a tree and the process of cell division in the cambial zone so that some relationships are now quite clear. Definite patterns of cell length changes are evident radially within a growth ring. It has been shown by macerating serial sections that cell length decreases following the in ception of yearly growth until a minimum is reached either in the first portion of the earlywood or close to the boundary between the earlywood and latewood (15, 31). Thereafter, the cell length increases to a maximum in the latewood, either rapidly or slowly, depending upon the nature of transition between the two intra-incremental zones (15). Other studies,-using isolated earlywood and latewood material from the same increment, have confirmed the results already cited (31, 122). Thus, latewood tracheids are on the average about 10 per cent longer than earlywood tracheids. Tracheid length variation vertically within an increment follows a general pattern of gradual increase from the base upward to a maximum below the crown, or between 3 0 to 4 0 per cent of the tree height (9,,30 13. 31, 82, .98) and a decrease in length above this height in the trunk. Radial variation of tracheid length, in successive growth increments, has been investigated in a great number of studies, the results of which are summarized below. At any given height, the tracheids of-the first growth increment are very short; coniferous tracheids range in length from 0.5 to 1.5 mm (31). Length increases rapidly in the second growth increment and in a few subsequent ones; then the rate of increase declines, but continues until a maximum length is reached. In conifers this may be three to five times greater than the initial length (26, 31). There is no general rule as to the number of years necessary for reaching maximum length which varies with the rate of growth (137) . Maximum length may be reached in as little as six to eight years, and may not be reached until a tree is 200 to 300 years of age; a continu ous increase has been observed throughout the growth increments of a 455-ryear-old Douglas-^fir tree (46) . After maximum is reached, there follows a period during which length may vary considerably about a mean maximum length; finally, in very old trees, the cell length may decrease from the maximum (5). In both parts of the increment, tracheids have a general increase in length from pith to bark, but earlywood exhibits a more gradual rise in the.early years than is shown by the latewood (155). Furthermore, the variability in length for the early- . wood cells is much less than that for latewood cells (30, 155). 14. There is no general agreement with respect to the relationship of increment width to cell length, but the balance of opinion appears to be in favor of a negative one; i.e., within a tree, wider increments have shorter cells other factors such as age being equal (31, 40, 122, 137). As regards height growth, a positive relation was. found to exist between this and cell length in the first increment from pith; thus .in comparing two trees of different growth rates, the tree of more rapid height growth was shown to have the longer tracheids (7, 23, 32, 82, 86, 106). The patterns of variation are often found to be modi fied if the tree has undergone changes in growth direction from vertical, due to environmental changes which involve the forma tion of structurally modified tracheid, such as takes place during the formation of compression wood (145), and cell length decreases. Tracheid length between trees in a particular species is also influenced by geographical location (8). How ever, it is to be emphasized that although slight modifications of tracheid length can be induced by environmental agents (108, 112, 113), considerable evidence has been presented for genetic control over tracheid length (8, 37, 47, 61, 100, 154, 169). Within the tree,tracheid length is controlled by physiological factors. Thus Dinwoodie (32) concluded "that tracheid length variation within the tree is a function of cambium which is strongly influenced by the activity of the terminal meristem, probably by means of auxin." Studies on the mechanisms that result in variations in length within the tree are difficult, but the extensive investigationsof Bannan (7, 8, 9) have shown that the rate of anticlinal divisions taking place in the cambium is a principal factor. It would be appropriate to quote a recent statement of Bannan, in this regard (9): In general, rate of anticlinal division and cell length are negatively related. This applies to both within-tree variations and to species-to-species differences. However, some" notable exceptions occur. Thus in Sequoia  sempervirens, where the cells are longer than in other conifers, the rate of'anticlinal division is substantially above average. . Con versely, Pinus contorta has both relatively short cells and a low rate of anticlinal division, an unusual combination. It therefore appears that the rate of anticlinal division in the cambium is probably a part of the operative mechanism by which cell dimensions, determined by the interplay of genetic, environmental and ontogenetic factors are regulated (9). Bannan's work is significant in that it bridges earlier studies of cell shape and dimensions with plant physiological investi gations of xylem differentiation. Dimensional Relationships4Jn Tracheids Investigations of variation in cell size within the tree were primarily concerned with cell length. Possible changes of cell diameter and wall thickness were given little attention. These and related aspects on conifers were discussed recently by Larson (77). Goggans (48), with special reference to southern pine species, reviewed the importance of diameter and wall thickness in relation to wood quality. Tangential diameters are primarily determined by the widths of cambial initials from which cells are derived, and ordinarily very little post-division expansion takes place (2). Nevertheless, tangential tracheid diameters increase consider ably as the widths of cambial initials increase with age on a cross section (3). As stated by Larson (77), "the,principal variation in tracheid diameter occurs in the radial direction, and it is this variation that contributes to latewood formation and the structural differences in the wood within and between trees." A general conclusion of Mohr and Roth (95) was that the radial width of tracheids increases from the pith outward and that the increase was rapid during the early years. Larson (77) quoted several authors and concluded that, with increasing age, radial diameter generally tends to increase to a maximum width and fluctuates thereafter. This is primarily applicable to early-wood tracheids since no definite age pattern can be established for latewood tracheids. Within a single growth increment, earlywood tracheid diameter usually increases from the stem base to a maximum at some point in mid-stem, then decreases towards the apex. 17. Investigations of Larson (77), Richardson (112)and Wodzicki(164) indicate that the physiological processes controlling secondary wall thickness are independent of those regulating tracheid diameter, although these two characteristics do vary more or less simultaneously - in nature. Whereas tracheid diameter was primarily determined by the amount of auxin reaching a devel oping tracheid, wall thickness appears to be determined by the amount of sucrose, or photosynthate, reaching each tracheid (80). Tracheid wall thickness increases with age on a cross section to a maximum that varies with growth conditions (65, 77, 120). Within a growth increment, wall thickness decreases with height to some point in mid-stem and may exhibit a slight increase within the crown. These patterns refer mostly to latewood tracheids and, although earlywood tracheid wall thickness also varies, the variations are minimal (77). Hiller (57) studied variation in radial wall thickness of latewood cells in several entire annual growth sheaths of loblolly pine (Pinus taeda L.). She found that wall thickness increased rapidly from the second to about the eighth internode from the sheath apex, then more slowly to the tree base. Larson (78) showed that in red pine (Pinus resinosa Ait.) earlywood wall thickness increased gradually with age, whereas in latewood the increase was more rapid but exhibited greater fluctuations. Finally, the cells that comprise the outer growth increment of very old trees (320 to 400 years) are often thin walled in com- . parison with wood formed during earlier stages of growth (52). 18. Within a tree there is some evidence that the micro fibril orientation, and to lesser degree the fibre composition, is related to the fibre dimension (145). A relationship between microfibril orientation in the cell wall and the morphology of wood fibres was first reported as long ago as 1934 by Preston (109, 110). He observed that for the layer in gymnosperms the angle of orientation of the cellulose molecules (micro fibrils) was greater in the short fibres of early growth incre ments than in longer fibres of later ones. The relationship was shown to apply to the tracheids from successive growth increments as well as to random samples from a number of growth increments, so that the relationship did not reflect a process of aging of meristem from which the tracheids were derived (see also 142, 143, 144, 145). This was subsequently shown to apply for the (145) and for the layers (89), as well as in the abnormal tracheids of compression wood (145, 146), fibres of angiosperms (145), phloem fibres (110) and cotton fibres (145). It is of interest to note that Hiller (57) similarly found a high degree of correlation between radial wall thickness and fibril angle in that, with increase in wall thickness, there is decrease in fibril angle. Likewise, Necesany (97) showed that microfibril thickness, cell wall thickness, and cell length are features of mutual direct proportionality. He further noted that an inverse proportional relationship exists between microfibril thickness and lignin content in the fibre. There are indications that cell length and width are also related. Thus a number of investigators, among them Bannan (6) , Graff and Miller (50) , Harlow (54) , Heinig r and Simmonds (56), and Wheeler et al. (158), have demonstrated that there is a direct positive relationship between fibre length and diameter within each species. In addition, a recent study by Vorreiter (140) brought out a rectilinear increase in fibre length with increasing thickness of fibre wall, for coniferous species.. Results of Preston (109, 110) and-Hiller, (57), if considered together, also point out that wall thickness and cell length may be related since both these features are in versely correlated with microfibril angle. The review presented above shows that definite inter relationships exist between cell dimensions, in particular length, and wall organization; hence with tracheid composition. Chemical Composition and Recovery of Carbohydrate  Fraction Wood consists of three main constituents, namely cellulose, hemicelluloses, and lignin (135). Holocellulose is the major constituent, and is, by definition,the lignin-free fibrous material comprising all of the hemicelluloses and cellu lose in wood (128). The total cell wall polysaccharides are usually obtained from wood substance in the form of holocellu-lose/,of which angiosperms contain from 70 to 80 per cent, and 20. conifers 60 to 75 per cent, based on extractive-free wood. The chief value of holocellulose preparation is that it offers a beginning material for further research. Glucan, as approximated by alpha-cellulose, constitutes 45 to 50 per cent of holocellulose in angiosperms and 40 to 45 per cent of that in conifers (20); this is customarily regarded as the true cellulosic fraction of wood carbohydrates. Alpha-cellulose is considered to be that fraction of holocellulose or pulp which is insoluble in a solu tion of 17.5 per cent NaOH at 20°C (131). The term, however, is strictly arbitrary, and does not imply exclusively a homogeneous glucan for, invariably, it contains small amounts of carbon dioxide- and furfural-yielding materials, as well as significant amounts of mannans from conifers (126, 162, 163). Standard methods for determining holocellulose or similar carbohydrate fractions, requiring 2 g of wood meal per sample, are available (128, 129). In addition, four micro-methods, requiring about 0.5 g samples, have been advanced for determination of holocellulose or similar carbohydrate fraction (see Ref. 161). Zobel and McElwee (171) and Erickson (43) applied chlorite, Watson (149) used chlorine-sulphate, and Leopold (83) employed peracetic acid followed by sodium boro hydride reduction. A fifth semi-micro-method has been used by Meier (90 , 91) and Larson (78) in which secondary xylem fibres are hydrolyzed to their monosaccharides, from which the poly saccharide composition is calculated from some basic assumptions 21. and chromatographic evidence. A comprehensive review of literature on these aspects is contained in a thesis by Squire (123). Squire also developed a new micro-method for cellulose determination, and quantitatively estimated intra-incremental alpha-cellulose yield as the corrected yield of nitrated wood meal. A major limitation of his technique was that it-could not be applied to all woods. Differences in analytical procedures will show varia tion of chemical composition in the same sample of wood (137). For example, the Cross and Bevan cellulose (128) does not con tain the entire hemicellulose fraction as some is extracted with the lignin. Thus the published data on holocellulose for a particular species of wood are inconsistent. As such, the most rigorous evaluation of analytical data is represented by summative analyses, in which the analyst aims to account for all constituents present by a summation that ideally should total 100 per cent, if no constituents have been overlooked and there is no overlapping (20). A summation that includes extrac tives, lignin, holocellulose, and ash ideally accounts for all chemical- constituents of wood. Many workers prefer, however, to use pre-extracted wood and to report summative analyses on the extractive-free and moisture-free basis, because of possible errors that may arise from variability in extractive content in different woods (20). In conifers, the proportions of cellulose, hemicellu-loses, and lignin, based on standardized methods of preparation, are approximately as follows (per cent of extractive-free, dry wood): cellulose 40 to 45 per cent, lignin 25 to 35 per cent, and hemicelluloses 20 fo.25 per - cent (20', 135). v The proportion of pectic substances is small (about 1 per cent) (135). Inter species variation in all chemical components, for several coni fers, is conveniently listed by Browning (20), Timell (135, 136) and Squire (123). Browning (20) also has shown wood elemental analyses for some of these species. Generally, chemical varia tion between species is less than that between genera; however, many exceptions have been reported. The chemical composition of reaction wood differs from that of normal wood in several respects (136) . Tension wood contains less lignin and xylan but more cellulose and more galactose residues than normal wood. Compression wood has a high lignin and a low cellulose content compared to normal wood. The relative amount of mannose residue is much reduced. The predominant hemicellulose is instead a polysaccharide based on galactose residues. Pillow and Bray- (107) reported that loblolly pine alpha-cellulose, from Cross and Bevan analysis, varied from 46 per cent in normal wood to 3 5 per cent in compression wood. Lignin content increased from 28 to 3 5 per cent, respec tively, with no significant change in total pentosans. Appar ently , the increase in lignin occurs at the expense of cellulose and/or glucomannan for conifers, and the reverse seems to operate in the.reaction wood chemistry of hardwoods. There is some evidence that compression wood also differs from normal wood, qualitatively. Thus, Dinwoodie's work (33) indicates low cellulose DP (degree of polymerization) for compression wood fibres. An extensive review of literature on anatomical and chemical properties of compression wood is given by Westing (151, 152). Another aspect related to the thesis topic includes variations in carbohydrate content across a tree diameter, which are significant because of juvenile wood. Zobel and McElwee (171) and Byrd et al. (22)^ reported that, based on dry weight, loblolly pine outer wood contained one to three per cent more holocellulose, and four to seven per cent more alpha-cellulose than that of juvenile wood from the.same trees. In general, it appears that, for coniferous wood, there is a rapid initial increase in cellulose content from the increments close to the pith, and somewhat less change beyond the juvenile core. It was shown by Kennedy and Jaworsky (69), for Douglas-fir, that Cross and Bevan cellulose yields increased from 57.4 per cent in increments 1 to 5 to 61.6 per cent in increments 16 to 25. Only a modest further increase of 1.7 per cent was reported in the remainder of the 8 0-year-old tree. Likewise, Wardrop (141) observed an increase in Cross and Bevan cellulose from 58 to 65 per cent within the first 20 increments from pith, for the same species. Chlorite holocellulose and alpha-cellulose patterns also follow the above trend. Holocellulose percentages appear to increase less rapidly than alpha-cellulose (52, 69), leading to the conclusion that there may be a greater hemicellulose fraction in increments close to the pith (70). This may be. attributed partly to the autohydrolysis of the older cell wall material closer to the pith (70). A point of interest is the similarity in behaviour of gross wood cellulose content in successive increments from the pith outward, and that of tracheid length across the stem. This similarity was shown by Hale and Clermont (52) working with Douglas-fir, and by Wardrop (141), with the same species and with radiata pine (Pinus radiata (D. Don.). Hale and Clermont also reported a low alpha-cellulose percentage in overmature wood (322 to 400 years) that generally contained thin-walled prosenchyma cells. They attributed this to the extremely slow growth of wood at that age. Other factors that could contribute to the differences in carbohydrate amount involve variation within the growth increment. Quantitative differences in growth zone chemistry were first reported in 1926 by Ritter and Fleck (114). They observed that cellulose forms a larger percentage (up to 3.4 per cent) of the total wood substance in latewood than in earlywood, of three coniferous woods, and an opposite situation prevails for lignin. These results were confirmed for conifers by subsequent workers for both holo- and alpha-cellulose fractions (52, 123, 124, 150, 161). Some of the later workers, however, showed greater difference in the magnitude of yields than reported by Ritter and Fleck. An anomaly occurs here, in/that the first-formed earlywood appears to retain some similarity at the chemical level of organization to last-formed tissues of the preceding season. Later-formed earlywood (i.e., from the present year), according to Squire (123), does not seem to retain such simi larity. Thus, minimum estimated cellulose yield occurred at considerable radial- cellular • depth1. Further analyses at - the intra-incremental. lewel; reported by-Wellwood. and .Wilson (156) , indicate that peak chlorite holocellulose (77 to 78 per cent, corrected for lignin) occurs at the point of latewood initiation, while the first- and last-formed cells of increments had lower values (72.5 to 73.5 per cent), in three adjacent increments of Douglas-fir. Besides quantitative differences> qualitative differences between earlywood and latewood were noted. It was shown that, for Douglas-fir, the average cellulose chain length of latewood is six per cent higher than that of earlywood (161). Carbohydrate content in the wood could be influenced by site index and the associated tree properties such as growth rate, diameter at breast height, total height, and specific gravity of the wood. Thus, it was shown by Worster and Sugiyama (165) that site index, which is a composite.of several basic factors, has a pronounced influence on carbohydrate content of conifers and angiosperms. Contrary to this, McMillan (87) found that the width of annual increment does not influence poly saccharide content when specific gravity effects are elimina ted. Meier's (90) analysis confirms this concept; he notes that much of the variation in polysaccharide concentration in the different cell wall layers is related to density. There fore, the observed differences in earlywood and latewood car bohydrate contents could be related to specific gravity or amount of cell wall substance per unit volume in the wood (124) . Estimates of Carbohydrate Fraction and Weight  in Relation to Cellular Elements Interest in the intensive examination of individual cells in woody tissue has centred mostly on determining the wall physical organization, and distribution of chemical con stituents in the wall layers. The weight factor (amount of carbohydrate fraction) of individual cells and its relation to any particular cell element has been studied very little. Some studies (10, 104, 105) reported the percentages (weight basis) of vertical tracheids (98 to 99 per cent) and ray parenchyma (one to two per cent) by quantitative fractionation of the total holocellulose derived from wood. In only one instance have compositional data been related to the weight of a biological unit. Thus, Thornber and Northcote (13 4) have estimated the average weight and length of single cells in four species consisting of three angiosperms and one conifer. They found that the mean weight of the cell increased with age by as much as 3 50 to 930 per cent in the four species examined by them. It was concluded that, at all stages, the cells from the coni fer were.considerably heavier than those from the angiosperms. No attempt was made to relate the observed weight differences in cells to length.. Estimates of tracheid weights were made by Yiannos (166) in connection with cell wall density of commercial pulp fibres. Weights were based on measurement of a group of fibres using a Mettler UM7 balance, and expressed as weight per unit length (coarseness). An ingenious method of weighing single tracheids was first attempted by Skene (121) using a quartz filament balance. Weights reported ranged from 0.04 to 0.1 yg. Neither the species employed, nor the tracheid composition are available. Weighing cells and relating results to a morphological property has several advantages over measuring lateral dimen sions. It is not only much easier and quicker, but includes also the effects of cell wall thickness, the size of the lumen,and the" amount of cellulosic material composing the cells. This property, i.e., fibre weight, is usually described as 'coarse ness' and is defined as the weight per unit length; it is expressed in milligrams per 100 m and called a decigrex, abbreviated to 'dg' (132). In view of its important effect on 28. many properties of paper, the.concept of fibre 'coarseness' (24) and a method for its measurement for pulp samples, have been described in TAPPI Suggested Method T234 su-67 (132). A new method for determination of fibre coasreness in wood, based upon the number of fibres per square mm of cross section and the specific gravity, was described by Britt (16, 17). Fibre coarseness measurements on various western coni ferous species, at various ages, were made following this tech nique (16). It was shown from such studies that there was a progressive but somewhat irregular increase in fibre coarse ness with increasing age of the tree, indicating that, per unit length, a fibre formed as the tree grows older weighs more. It is also known that fibre length increases from centre of a tree outward; thus •Clark\(24), based on studies on pulp samples has shown that per unit length, longer fibres weigh more than shorter ones. Such a conclusion can be justified on the basis of known inter-relationships within the tracheid morphological properties discussed earlier. Not only the weights, but chemical composition of specific wood cells differ (1, 55, 103, 133, 142). The higher pentosan content of ray cells compared to the entire wood has been known for some time (1, 55, 133, 136). Recent studies of Perila and his co-workers (103, 104, 105) confirm this for both angiosperms and conifers. Their data indicate that the charac teristic chemical feature of the parenchyma cells, as compared to prosenchyma, is their unusually high xylan content. Effects of Fertilization on Wood Properties  and Chemical Composition Klem (71) , in a recent paper, summarized the main effects of fertilization of forest stands on wood characteristics of spruce and pine species. Problems in assessing wood quality of fertilized coniferous trees were discussed by Larson (79). Detailed literature surveys have also been provided by Lee (81), Posey (108)and Sastry (117). Consequently, such information is not presented in detail in this thesis. In most instances, the application of fertilizers to forest stands results in an increased growth response and a slight change in the earlywood-latewood ratio. Changes in size and shape of cells are minimal and less consistent. Stud ies reported so far indicate that specific gravity is the characteristic most sensitive to fertilizing and that a maximum of 10 per cent reduction can occur, but that it is usually more than compensated for by the increase in volume production (44, 71, 117). Any decrease in specific gravity of the raw material produced by fertilization will be of greatest importance if that material is to be utilized by the pulp and paper industry since the yield per unit volume will be reduced. Additionally, and perhaps less appreciated, is the effect a change in density would have on the properties of the product. This could be brought about both by resulting changes in either the proportion 30. of latewood fibres in the raw material, or through changes in cell morphology, and possibly cell weight. Depending upon the type of product being produced the changes could either be advantageous of deleterious. The influence of such changes on the quality of this material when utilized as lumber is of less importance, not because the strength of the product could be altered but, with current grading practices, these factors would not be likely to affect lumber grade. If and when machine grading of structural lumber becomes important, changes in wood quality brought about by fertilization could possibly have a significant effect on lumber grade (66). Not all fertilizer experiments, however, have been successful in obtaining a substantial and consistent growth response (92, 93). Consequently, it is imperative to learn more about the basic factors affecting the quantity and quality of wood grown in fertilized stands. For example, .^Zobel et. al. (173) found that certain levels of NPK fertilizer treatment had drastic effects on specific gravity of loblolly pine wood. The same authors, as well as Posey (108), showed the existence of differential effects of fertilizer treatment depending on initial properties of wood. Posey also found that certain formulations of fertilizer brought greater difference in wood properties and growth rate compared to others. Earlier explora tory work by Sastry (117) indicated the possibility of differen tial effects of fertilizer composition on physical and mechani cal properties' of Douglas-fir wood. The possibility, exists that, in studies of fertilizer influence on older trees, i.e. , in field trials, several fac tors relating to environment and micro-site may confound the real effects of fertilizer application; this is part of the reason why some of the trials may have failed to elicit an increased growth response. Such a confounding influence could possibly be controlled in laboratory experiments employing seedlings. Thus Murphey et al. (96) conducted experiments in controlled nutrient cultures, for two growing seasons, using western larch (Larix occidentalis (Nutt.)) seedlings. Their results indicate differences in selected anatomical features such as cell size and dimensions, number of tracheids per 2 10,000 ym , resin canal diameter and number of epithelial cells per resin canal, as well as physical and mechanical properties. Changes in chemical composition following fertiliza tion have been studied to a very limited degree. Little signi ficant change was reported in Douglas-fir by Erickson and Lambert (44), but a slightly lower holocellulose and alpha-cellulose content was found in the wood formed after fertili zation. Likewise, in young loblolly pine, cellulose yield showed a decreasing trend, but the change was not significant (173). These experimental results would be expected. Fertili zation has been shown to increase the production of earlywood more than that of latewood, so that the greater number of earlywood cells with their higher lignin content would decrease cellulose percentage and increase that of lignin. These con clusions are in general agreement with Erickson's (42) findings on increase in the lignin content of fertilized spruce (Picea  abies (L.) Karst), and Malm's (88) data on the increase in middle lamella thickness of latewood cells in fertilized trees. Jensen et al. (64) have studied the effect of inc reased growth rate of pine and spruce in Finland on wood quality, particularly for pulp and paper. It was reported that paper properties were little affected by the increased growth rate due to fertilization. Studies in Sweden (51) have reported that paper produced from wood after fertilization was higher in burst and tensile strength and lower in tear than the corresponding material from unfertilized trees. Wood obtained after fertilization also produced a denser and less bulky paper. . Changes in paper characteristics can be explained readily on the basis of the known changes in wood characteris^„-tics due to fertilization. Thus, thin-walled fibres of large diameter, such as found in fertilized wood in high proportion, collapse to flattened ribbons in the conversion process, to form a compact, dense sheet, low in opacity, uniform in for mation, and smooth of surface. Furthermore, they present highly effective fibre to fibre contact areas, resulting in strong bonds, as revealed in high tensile and bursting strengths (13). Similarly, fibre length is a dominant factor controlling the tearing resistance of paper, long fibres producing paper with the highest tearing resistance by distributing over a large area the shearing forces involved in tearing paper (13). Both bulk and porosity increase to some extent with increasing tracheid length (29). In summary, is can be stated that the response of coniferous trees to fertilization appears to be related to the vigor of the tree, the initial level of the particular wood property, the genetically controlled ability of.the indi vidual to respond and, finally, the interplay of environment. Extremely slow growing trees which are producing low density 'starvation.wood1 may increase in density (51), while trees growing at a moderate rate are generally found to decrease in density. The most vigorous trees growing close to their poten tial often show no response to fertilization. The density changes observed so far are for the most part (up to 50 per cent) accounted for by changes in latewood percentage; but some studies have indicated that changing radial wall thickness (mostly latewood) may play some part in this density response. In no study, however, were changes in specific gravity related to changes in chemical properties of fertilized wood. Likewise, a decrease in cell length is expected since this is associated with the very rapid increase in growth, which could cause an increase in the number of cambial initials due to change in their pattern of division. Bannan (8) showed that anticlinal divisions are the means by which multiplication of fusiform cambial cells is achieved. Frequency of multi plicative divisions would influence the length of derived initials in the cambium, hence, length of derived vascular elements; a factor such as fertilization, which would cause an increase in radial growth, should cause a drop in the mean cell length. This concept is supported by the work of several authors (71, 108, 117, 141, 173). Finally, as stated by Zobel et al_. (173) , there is indication of a differential effect of fertilizer depending on the initial wood property of the tree. Also, there appears to be a large amount of individual tree difference in response to fertilizer treatments, suggesting genetic interaction. This 'individual response' may become very important in the develop ment of improved strains of trees. CHAPTER III MATERIALS AND METHODS Selection of Material The primary objective was to include as wide a range of tracheid lengths as possible, so that the criterion for the selection of wood material was tracheid length based on measure ment of length from macerated material. Within this objective, a broad taxonomical sorting of wood tissue was made, so that conclusions from the work might be interpreted oVer a reasonable range and more closely approach a general statement. In order to include the within-tree variation in tracheid length from pith to bark, several points were sampled in one disk, at stump height, of a 500-year-old Douglas-fir (Pseudotsuga men-ziesii (Mirb.) Franco) which was available from another study. The radial density pattern in this disk has been described by Kennedy and Warren (70). From Pinaceae, three species from separate genera were sampled: sugar pine (Pinus lambertiana Dougl.), Sitka spruce (Picea sitchensis (Bong.) Carr.) and Douglas-fir. One species of a representative genuss from each of the six other coniferous families were included; these were redwood (Sequoia  sempervirens (D. Don) Endl.) from Taxodiaceae, New Zealand white pine (Podocarpus dacrydiodes A. Rich) from Podocarpaceae, eastern redcedar (Juniperus virginiana L.) from Cupressaceae, Pacific yew (Taxus brevifolia Nutt.) from Taxaceae, Taiwan cow's-tail pine (Cephalotaxus wilsoniana Hay.) from Cephalo- taxaceae, and hoop pine (Araucaria cunninghamii Sweet) from Araucariaceae. In addition, two increments showing pronounced compression wood were also included from a tree of Douglas-fir. Material in the form of increment cores from fer-. tilized Douglas-fir trees was available from another study (119). Two increments, from one tree each of three fertilizer treatments, were selected from this study material. The treat ments consisted of formulations of Urea, NPK, and (NH^)^ which generated, respectively, an increase (Urea) and decrease (NPK, (NH^^SO^) in wood specific gravity following application. Finally, tracheid weight variation before and after fertili zation was traced in six consecutive increments of a Douglas-fir fertilized with NH^NO^. In this tree, sampling included one growth increment before fertilization and five increments formed coincident with and after the two applications. The entire sampling plan included nine genera, and nine species, and all seven coniferous families. A total of 28 growth increments were sampled. Wood material of sugar pine, redwood, hoop pine, New Zealand white pine, and eastern redcedar was available from the xylarium of the Faculty of Forestry, University of British Columbia. A small wood sample of Cephalo taxus , in the form of a disk, was obtained from National Taiwan University, Experimental Research Forest, in Taiwan. The . -remaining material was collected within British Columbia by the author, from the University of British Columbia Research Forest, near Haney and Robertson River Valley, South Central Vancouver Island. Origin, description of gross wood charac teristics, and measured tracheid length of these wood samples are given in Appendix 1. Wood Sample and Tracheid Skeleton Preparation General procedure: Selected wood growth increments were excised from wood blocks, disks or increment cores, then carefully separated into whole earlywood and latewood portions; each of these was then dissected into smaller pieces, with portions bundled separately in stainless steel wire screen (150-mesh) and made into small packets; each packet was labelled with a metal tag to identify the species, increment number, and earlywood and latewood fractions. Separate packets of some of the increments were also prepared to determine the percentage yield of holocellulose and alpha-cellulose. Holocellulose preparation: Packets containing wood pieces were extracted with 2:1 (V/V) ethanol-benzene; holo cellulose tracheid skeletons were obtained by treating wood pieces (in packets) in peracetic acid/ with subsequent sodium borohydride reduction, following the method described by Leopold (83). This method is reported to give virtually intact superior holocellulose fibres with 93 to 100 per cent recovery of major 38. polysaccharides and minimum carboxyl groups. The procedure included three steps: 1. The treatment was divided into a number of 3 0 minute cycles, consisting of a peracetic acid treatment (i.e., 5 g of peracetic acid and 2.5 g of sodium acetate per gram of wood; total volume 32 ml per gram of wood; temperature 70°C), followed by soaking in hot water (30 minutes, temperature 70°C). 2. At the end of each cycle (3 or 5; see below) the packets were washed in distilled water and the water removed with acetone. After air-drying, the packets were then placed in a suction flask, the flask was evacuated, and a new batch of re agent was drawn into it. 3. At the conclusion of the sodium borohydride reduc tion (3 hr, pH 9.5), the delignified, soft, wood pieces were defiberized by gently rubbing with a glass rod against the wall of a glass beaker in the presence of hot water. A total of five cycles was needed for latewood and three cycles for earlywood delignification as determined by physical checks. The holocellulose tracheid skeletons thus obtained were stored in test tubes containing distilled water. • Alpha-cellulose preparation: Approximately half the quantity of the holocellulose fraction, from each sample, was further reduced to alpha-cellulose skeletons essentially following TAPPI standard method T203 os-61 (131) scaled to accommodate micro-amounts of material. Converted yields obtained for holo- and alpha-cellulose fractions are given separately for earlywood and latewood in Appendix II. Tracheid skeletons of alpha-cellulose, after washing, were also kept stored in test tubes containing distilled water. At all stages holo- and alpha-cellulose pulps were kept in wet condi tion and never allowed to dry. This was done so that tracheid length measurements could be based on green condition and represent more closely their natural state in the tree. Thus, any possible effects of shrinkage due to drying were avoided as far as possible. The only exceptions were those samples which were intended for yield determination. These pulps (holocellulose) were processed into thin sheets and air-dried, before alkali treatment. Test for lignin: Lignin content in some of the holo cellulose and alpha-cellulose samples was determined by TAPPI standard method T236 m-60 (130) as modified for reduced quan tity of pulp by Berzins (12). The test showed that residual lignin contents were equal to or less than 1.53 and 0.29 per cent, respectively, for holo- and alpha-cellulose, in the samples examined (see Appendix II). 40. Analysis of Tracheid Skeletons 1. Sampling criterion The primary objective of this investigation was to relate tracheid weight to its length. Initially, all tracheids were treated as belonging to one population, irrespective of source. In order to examine such a relationship the sample replication number needed to be determined. Calculation of sample number for a statistically defensible determination could be obtained if the objective were to establish 'norms' for increments or tissue fraction within increments. However, in a study such as this where sampling included a wide range of tracheid lengths, from different origins, the primary function is not to estimate increment mean; rather it is to test the overall relationship between the two characteristics, namely tracheid length and weight. Moreover, within-tree tracheid length patterns, and those between and within species of coni fers, have already been established by several workers. There fore, standard procedures such as calculation of sample size by Stein's two-stage sample procedure (125), are of little help. Still, in order to maximize research effort within the time available, it is essential to determine sample size so as to get reliable estimates of the desired relationship. Previous experience from a similar study (118) indicated that at least two groups of 10 fibres each, taken at random from the early-wood and latewood regions of the increment, would provide a satisfactory number with reasonable estimates of the mean and standard deviations. For the above reasons, the number of tracheids analyzed from one particular sample (increment) was standar dized at approximately 20. In some cases where the variation in tracheid length was great, this number was increased even up to 40 for that sample. 2. Tracheid length A portion of the material from each of the labelled test tubes was transferred to a petri dish; randomly selected tracheids were picked, with the help of fine tweezers, using a binocular microscope. These were then transferred to a glass slide in a drop of distilled water and a cover glass applied. Care was taken that no damage was done to the fibre tips. Tracheid length measurements were made, in the wet condition, by the use of a fibre length recorder. Images of fibres are projected on an opaque glass screen and their lengths measured with the aid of a calibrated probe wheel. This instrument has-va precision of + 0.1042 mm at 56.X and + 0.0712 mm at 82 X magnification. Tracheids up to 4.9 8 mm length were measured using the higher magnification. As noted, a minimum of 10 tracheids wass. measured separately for earlywood and latewood for each of the selected increments from each of the holo- andv alpha-cellulose pulps. 42. Alpha-cellulose tracheid skeletons, showed a shortening in length of up to 15 per cent in relation to holocellulose tracheid skeletons. This was further verified by treating individual tracheid skeletons of holocellulose with a 17.5 per cent NaOH solution in separate petri dishes, for 75 minutes at 20°C. Length measurements before and after the treatment confirmed the reduction in tracheid length. The alkaline treatment of holocellulose fibrous skeletons could be expected to cause swelling (21) due to changes in lattice and thereby shortening of the tracheid, explainable by simple mechanical factors, i.e., reverse Poisson's ratio.. (160). 3. Tracheid weight After determining the length, each tracheid was transferred to a separate labelled plastic weigh boat. Several of these boats were then placed in a freeze dryer (Virtis Model 10-100) and dried in vacuum (-60°C and 5 u vacuum) for four hours. Later experience showed that one hour was sufficient to dry individual fibres. Subsequently the boats, each with its fibre, were transferred to a constant temperature and humidity (CTH) room (73 +3.5°F, 50 + 2% RH) and conditioned for 24 hours. Tracheids for which lengths had been determined were then weighed individually, using a specially designed quartz ultra-microbalance. Briefly, the balance consists of a thin quartz filament (20 y diameter and 6.35 cm (2 1/2 in.) long) firmly fixed at one end. The tracheid to be weighed is placed on the free end and its weight determined by measuring the deflection of the filament with a microscope. The principle is similar to the loading of a cantilever beam at the free end. The balance so developed has a weighting range from 0.06 to 14 yg,1 and provides a reproducibility (precision) within 0.03 yg, which is about 10 times higher than that of present commercial electrobalances and ultra-microbalances. The balance case consists of a 20 ml hypodermic syringe,to protect the filament, and a mount, i.e., the plunger, for holding the filament in position. The closed end of the syringe is cut off and left open to make loading of the filament possible, with arrangements made to cover the opening with a cover glass during weighings to protect it from air currents. A diagram of the ultra-microbalance is given below. A crucial step in the use of the balance is its calibration. This was achieved by a colorimetric method, in which small crystals or flakes of fluorescein were loaded on the filament and the deflection measured. These crystals were then collected and dissolved in a known volume of solvent (1 per cent Na^O^) and the optical density was measured against appropriate standards, using a Unicam SP 800 Recording Spectro photometer at absorption maxima of 493 and 460 m.U Results were then compared with the optical density of a series of 1 — 6 1 micro gram (yg) =10 g 44. 1 2 6 Cross section of the Ultra-Wicrobalance. 1. glass plunger 2. open-ended glass tube 3. plastic lock tip 4. plastic needle mount. 5. quartz filament 6. cover glass lid 7. radium foil solutions the strengths of which were known within close tolerance limits. The final .result was a graph showing de flection of the filament against weight of material placed on it. The relationship between sample weight and vertical dis placement of the filament was linear over the entire useful range of the balance. Once calibrated, the balance . (using the same quartz filament) was used to weigh tracheids. per minute; the results were reproducible when reweighing the same individual (+ 0.03 ug). Further details of the balance construction, design and calibration are discussed elsewhere (118, 121). Weights of some of the tracheids (1 to 2 ug) are Tracheids may be weighed at the rate of one or two comparable (+ 0.15 yg) to those obtained by using a Mettler UM7 balance, as determined by checks. Correction for moisture content: The weight of tracheids thus obtained represents that at the EMC of the CTH room. In order to express the weight in ovendry condition, adjustments were'made to the original weight. This was done as follows: the extra material available from the holo- and alpha-cellulose pulps (after selection of the required number of tracheids for weighing) was made into thin sheets of fibre, and water removed by applying suction. These "pulp sheets" were numbered for identification, and kept stored (in the open) in the CTH room until a constant weight was reached, following which they were ovendried. Individual tracheid weights were then adjusted to moisture-free basis, to the nearest 0.01 per cent. The moisture contents of the pulp samples examined are given in Appendix III. Statistical Analyses Several analyses were conducted in order to test the proposed hypothesis and its applicability to woods of different origin. Multiple regression analyses were used to determine interrelationships within and between species, and within increments (earlywood and latewood); to assess the measure of association between tracheid length and weight; finally, to develop regression equations of best fit. In addition to the least-squares models, conditioned regression models (with zero intercept) were also used. The regression technique required a stepwise elimination procedure, from an equation of multiple variables, as developed by. Kozak and Smith (73). Analyses of -covariance were"employed to test dif ferences within and between species; between wood formed at juvenile, mature and overmature stages of tree growth; between earlywood and latewood; between normal wood and compression wood; and to characterize the behaviour of wood from fertilized trees. Covariance is a technique used to adjust means of the dependent variable 'Y' (Tracheid Weight) for differences in sets of values of the corresponding independent variable 'X' (Tracheid Length) and to bring out differences, if any, within 2 or between the groups noted above. The analysis employed compares regression equations for parallelism and coincidence. Its purpose is to determine whether separate regression equations should be used for each population under study and/or if some or all of the populations could be described by one regression-equation . A simple method developed by Dr..A. Kozak, Faculty of Forestry, The University of British Columbia, and presented at the 3rd Conference of the Advisory Group of Forest Statis ticians, Section 25, I.U.F.R.O., France, September 7-11, 1970. CHAPTER IV RESULTS AND DISCUSSION The present study is not concerned with description of tracheids in species of different taxonomic origin nor, -for that matter, with development of procedures to test differences in tracheid weights within trees. Rather, its purpose and attention are focussed on: a) establishing a quantitative relationship between tracheid length and.the amount of carbohydrate fraction (holo- and alpha-cellulose) that may exist within and between coniferous species; b) evolving a basic relationship for coniferous woods, from biological and cell morphological concepts, in aid of practical considerations,for example carbohydrate yield based on tracheid length; and c) to test the usefulness of such a basic relationship with respect to observed effects of fertilization on wood characteristics in Douglas-fir. Such a fundamental relationship for coniferous species could also provide means to, or re-emphasize the importance of, trac heid length for selection of individuals within species in tree improvement work. 48. Relationship Between Tracheid Length, and Weight . 1. Variation within species A total of 1,534 individual tracheids (787 holo cellulose skeletons and 747 alpha-cellulose skeletons) was measured for length and weight in the nine species selected. Mean values for each of the increments examined are given separately for earlywood and latewood, in Tables I and II. Tracheid weight and length values, taken radially across the disks, for two Douglas-fir trees, are contained in Tables III and IV. Only the data from 'normal wood' (excluding compres sion wood and wood from fertilized trees) were considered when establishing a relationship between weight of fibre skeleton and its length. In addition, the relationshipslf:or both holo- and alph cellulose skeletons were examined separately. An advantage of knowing the amount of holocellulose is that it provides an estimate (within reasonable limits) of the total tracheid car bohydrate fraction, whereas alpha-cellulose relates to the basi skeletal component. Individual species relationships obtained between tracheid weight and length are shown in Figures 1 to 9, along with the regression relationships of best fit. It is seen that the correlations determined using either the curvilinear or the linear relationship, by least-squares method, are highly significant, accounting for 66.4 to 94.8 and 84.8 to 98.6 per 49. cent of variability, in holocellulose and alpha-cellulose skeleton weights, respectively, by the length factor alone. The variability accounted for by length is higher for alpha-cellulose than it is for holocellulose. Possibly this is related to the fact that alpha-cellulose, being more basic in the skeletal structure of carbohydrate framework, should be less variable. There could be two other possibilities, ~ these being: 1. The swelling and shrinkage changes that occurred due to strong alkali treatment, may have removed some of the minute variations in length-weight re lationships for these fibres; and 2. The technique of alpha-cellulose isolation (131), may be a better controlled one than that of holo cellulose (83) . The strong relationships.obetween tracheid length and weight, as observed from Figures 1 to 9 and the statistical analyses given therein, are highly significant, for each of the nine species examined. This is of importance since both tracheid length and gross wood carbohydrate content increase with age, i.e., from pith towards bark, for a number of years as des-" cribed previously. Results of gross wood analyses done by several workers suggested a relationship between carbohydrate composition of wood and cell length. But, due to confounding influence of the age effect, the relationship obtained was attributed to growth factors. Dadswell et a_l. (28), who obtained a highly significant correlation (1 per cent level) for radiata pine, suggested that: this is due no doubt to the increase in both with age, but it is a fact that is important because fiber length can possibly be con trolled genetically, and it would be most advantageous if high average fiber length also means high cellulose content. Ifju and Kennedy (60) also obtained a highly signi ficant 'r' value between tracheid length and gross wood cellu lose content for Douglas-fir. These authors also explained the correlations obtained on the basis of increase with age of both tracheid length and cellulose content in a single stem. In addition, the authors offered an explanation "that more compound middle lamellae will occur per unit of length in a longitudinal row of short tracheids," which should explain the observed tracheid length and gross cellulose content relation ship. This concept was supported by studies of Einspahr et al. (3 8, 39) who noted negative relationships between pulp fibre length and gross wood lignin content, for slash pine and loblolly pine. They suggested that "longer fibres could -result in fewer end walls per given weight of wood and would consequently result in less lignin (and less pectin compounds) and a higher pulp yield"; i.e., increased fibre length results in fewer fibres per gram and this in turn means less intercellular mat erial and less lignin. Further, they found that fibre length and cell diameter are highly correlated, from which they 51.. concluded that the relative proportion of the cell wall layers may also change due to general increase in cell size. In view of the age-related relationship between gross wood cellulose content and cell length within a particu lar stem, the question arises as to whether a similar relation ship holds for different stems of one species, and between species, and whether it is characteristic of a species. One approach can be found in the work of Zobel et al. (17 2).\ These authors noted a between-tree correlation for tracheid length at 30 years of age and per cent alpha-cellulose, among 308 loblolly pine trees. The correlation coefficient, although significant, was only 0.153, so that tracheid length explained only slightly more than two per cent of the variation in cellu lose content. Even though it is apparent that tracheid length and gross wood carbohydrate content are related within and between stems of a species, the relationship between species and its cause, is an area of relatively little information. Moreover, it may be appreciated that comparisons between gross cellulose content and single cell characteristics, such as length, are weak since the former is based on the estimate of numerous cells in a given volume or weight of wood, and the latter is measured for individual cells. Therefore, such comparisons become less sensitive and this may explain why some of the authors failed to show more than two per cent variability in cellulose content based on cell length. Nevertheless, these findings do provide earlier clues for relationship between length and carbohydrate content in wood. The results obtained in the present study (Figures',:l to 9) do show that, irrespective of species, the possibility exists for a strong relationship between length of an indivi dual cell and weight of its carbohydrate skeleton. However, as noted by some earlier workers, it is not apparent at this stage whether the relationship obtained is due to chance, i.e., that both increase with age, or whether there is a common fundamental cause for such a relationship between species. These and related aspects will be considered. 2. Variation between species As noted, the primary objective of this investigation wa's to relate tracheid weight to its length in conifers. In Figure 10, a definite relationship is shown when combined data from all the nine coniferous woods (excluding those from com pression wood and wood formed by fertilized trees) are plotted. The correlations obtained from the curvilinear relationships are,highly significant and accounted for 91.9 and 95.7 per cent of the variability in holocellulose and alpha-cellulose skeleton weights, respectively, by the length factor alone. When a conditioned regression model (where the intercept is zero) was fitted (see Figure 11), the variability accounted for was the same (91.9 per cent) for holocellulose skeleton weights, whereas"it was slightly less (95.2) for alpha-cellulose. In spite of the strong relationships obtained, indi vidual species differences were significant. Such differences could be expected since, for each of the species examined, tracheid characteristics are also different with respect to diameter and wall thickness (101, 102, 116). Thus, despite a strong relationship between length and weight in each species, some of these differed statistically from one another, in res pective slopes and levels.of the regression lines. A summary of these separate analyses is given in Tables V and VI, for holocellulose and alpha-cellulose tracheid skeletons,respec tively . Analyses for holocellulose tracheid skeletons (Table V) ..indicate that, in spite of the differences, there is simil arity in regression line slopes between species belonging to one family. Thus Pinus lambertiana, Picea sitchensis and Pseudotsuga  menziesii, belonging to the family Pinaceae, did not differ from one another in slopes; in addition, Pinus lambertiana and Picea  sitchensis showed no difference in the levels of the same regres sion lines. Similarly, Taxus and Cephalotaxus, which are closely related (159),did not differ in either slopes or levels. It is of interest to note the similarity in slopes between Podocarpus  dacrydiodes and Pseudotsuga menziesii, as well as similarity in slopes for the latter species together with Taxus, Juniperus and Cephalotaxus.. The behaviour of alpha-cellulose tracheid skeletons, with some exceptions,is like that of holocellulose, in showing similarities between species belonging to the same family or related families (Table VI). Sometimes, however, the similari ties noted were in species belonging to different families, but of the same tracheid length class. Again, Pinus lambertiana and Picea sitchensis did not differ from one another in slopes. Similarly, Taxus and Cephalotaxus showed the same behaviour with regard to slopes, and these two, together with Juniperus, did not differ significantly.in either slopes or levels. Interes tingly, although Pseudotsuga did not differ from Pinus and Picea of the same family, in slopes of regression lines for holo cellulose skeletons, it differed significantly (5% level) from these two in alpha-cellulose characteristics. Another point of interest is the similarity in slopes between Araucaria, Sequoia, Pinus and Picea, as well as that between Araucaria and and Podocarpus. In addition, Araucaria and Sequoia did riot-differ in levels. The slightly different behaviour pattern of alpha-cellulose skeletons, from those of holocellulose, could possibly be related to the shrinkage of the tracheids in the removal of hemicelluloses. Possibly the ratio of shrinkage may be dif ferent in different species. Therefore, it would be of interest to pursue this matter further in order to determine an exact relationship between shrinkage and alkali treatment in these fibre skeletons, for different species, including- earlywood and latewood. 3. Variation with increasing age The strong individual species relationships between tracheid length and weight, as well as that for all the species considered together, has already been pointed out. It is known, however, that for any particular anatomical or physical property, great differences can occur within a single stem. In view of this, variation of tracheid weight and length across the radius was studied. Average tracheid weight and length data are listed in Tables III and IV for two Douglas-fir trees. Radial tracheid weight patterns in the butt log of the 500-year-old tree are shown in Figure 12. For comparison, mean values for tracheid length obtained from macerated wood of certain increments, and specific gravity values taken from Kennedy and Warren (70) for this wedge sample, were also included in the same figure. It is seen that for both earlywood and latewood there is a marked increase in tracheid weight up to 80 years and a decrease after about 150 years. Tracheid length for this tree (macerated tracheids), as well as the average tracheid length of.the samples weighed (Tables III and IV), followed the same general trend. It is of interest to note that the radial (wood) den sity pattern described by Kennedy and Warren (70) is similar to the tracheid weight-length pattern obtained in the present study, indicating some inter-relationship between individual tracheid characteristics and gross wood specific gravity. The 56. literature does not show this clearly (47). Indirect evidence is available, however, from cell morphological studies. It is possible that with increase in length there may be an increase in percentage of cell wall as a portion of total cell volume, which is comparatively greater for longer fibres; i.e., cell wall volume per unit length would be higher for longer tracheids. It is reasonable, therefore, to assume that tracheid length is closely correlated with wood density within this tree. Thus, van Buijtenen (138), from his study on slash pine, concluded: There appear to be two complexes of proper ties which are rather strongly correlated. They are small fiber diameter, thin walls, a low sum merwood percentage, short fibers, and a low speci fic gravity on the one hand, and large fiber diameter, thick walls, a high summerwood percen tage, long fibers and a high specific gravity on the other hand. These two complexes are typically the juvenile properties versus the mature proper ties , although apparently even in mature wood these properties seem to be strongly associated with each other. This association, however, is not rigid and it is possible to find a variety of combinations. One can, for instance, find wood samples with small fiber diameter, thin walls, a high summerwood percentage, and an average specific gravity. The reasons for a relationship between tracheid length and specific gravity within this species could also be attributed to the increase in amount of carbohydrate fraction in the tracheid which, as shown above, is highly correlated with trac heid length. Not much is known on this aspect. Zobel and McElwee (171) found a poor relationship between cellulose yield and specific gravity on a weight basis, but a very strong 57. relationship on a volume basis i.e., "a given volume of high specific gravity wood giving higher yields of cellulose than the same volume of low specific gravity -wood." Several studies indicate highly significant relationship;', between gross wood cellulose content and fibre length (28, 60), as well as fibre length and density (99). Also, it was noted that the pattern of holocellulose variation across a growth increment is similar to that of tensile strength which is highly significantly cor related (r =,0.92) with specific gravity. (156, 157). It is possible that specific gravity is thus related indirectly to cellulose content in the tracheids. Wood formed in the region near the pith is called juvenile or corewood; when compared to mature wood it has different properties such as-shorter tracheids and lower cellu lose percentage, specific gravity, percentage of latewood, and tensile strength (41). Data from this study also show that weights of the individual tracheid skeletons from juvenile wood1 (Douglas-fir Increment 10), apparently, are well below the level found in mature and overmature wood increments (Figure 12). Statistical analyses, however, did not confirm this entirely (see Tables VII and VIII). Analyses for holo cellulose skeletons indicate that no significant difference in "'"Only Increment 10 is considered as juvenile wood, since McKimmy (94) showed for Douglas-fir that the juvenile period extends to about 10 annual increments from the pith. Also, tracheid length and specific gravity data for this tree do not warrant classifying Increment 20 as juvenile wood. 58. weight could be found between juvenile wood tracheids as compared with those from mature and overmature wood tracheids. Contrary to this, analyses for alpha-cellulose skeletons (Table VIII) indicate that, for the same tracheid length, the juvenile wood tracheids are significantly lower in weight com pared to mature (A) and overmature wood (B) tracheids. The obvious reason for these apparently conflicting conclusions is that qualitative differences may exist; i.e., for the same tracheid length, juvenile wood tracheids will have more hemi-cellulose fraction than those from mature and overmature wood. Larson (78) has noted some very definite qualitative changes in wall chemical composition with increase in age for red pine. From this and other considerations, he concluded that these changes appear to be inherent and may by typified by the.trend expressed in the earlywood. Some indirect support could also be found from Zobel and McElwee's (171) work on loblolly pine. They found that, based on dry weight, yields from mature wood ran 3.5 per cent higher for water-resistant carbohydrate (WRC), whereas those for alpha-cellulose were 7.5 per cent higher when compared to the yield from juvenile wood from-the same trees. As pointed out by Larson (76), formation of juvenile wood is associated with the prolonged influence of the apical meristem, during the growing season, in the region of the active crown. As the crown moves upward in the older tree, the cambium at a given height becomes less subject to the direct influences of the elongating crown region and adult wood is formed (111). Therefore, the reason for the significantly lower weight of alpha-cellulose skeletons in juvenile wood is evident, since wood produced during the juvenile period is of different quality as stated above, which quality will be reflected in the individual cell. Several arguments have been offered to explain the decrease in cellulose percentage and specific gravity in overmature wood (see Tables VII(C) and VIII(C); and Figure 12). For example, Hale and Clermont (52) explained that over mature wood (that formed in the outer portions) of relatively old trees (300 to 400 years) has characteristically thin-walled prosenchyma.with higher lignin and lower alpha-cellulose per centage than wood formed during earlier stages of growth. Kennedy and Warren (70) supported Doerner1s concept (3 5) in which he explains the increase and subsequent decrease in specific gravity on mechanistic principles. Doerner suggests that when the tree is growing appreciably in height, bending stresses are greater, thereby necessitating the,formation of dense peripheral wood to augment a relatively small section modulus. At advanced ages, when the rate of vertical to radial increment has decreased markedly, the tree continues to react efficiently to the reduced stress stimulus by producing less-dense wood to counteract the development of an excessively large 60. section modulus. For both cases he suggests that the density changes are produced by the reactions of developing tissues to the stimuli. The former authors (52) thus support a direct age effect independent of tree diameter, whereas Doerner infers an indirect one since diameter increases with age. Results of the present study on single tracheid weights are in agreement with the above conclusions based on gross wood analyses, and provide a further point of proof in that changes in growth behaviour of trees are reflected more directly in the development of the individual cell whether in normal or reaction wood formation (see below). Additional confirmation was obtained by the analysis of covariance test (Tables VII and VIII) in which tracheid weights of mature wood (Increments No. 20, 80, 150) were compared with those of over mature wood (Increments No. 300, 400), after adjusting for differences in tracheid length. The significant 'F' values for 'slopes' in both cases indicate that these populations differ. This means that the significantly lower tracheid weights in overmature wood could be attributed to some factor such as thin walled prosenchyma cells as suggested by Hale and Clermont (52). 4. Variation between regular (normal) and compression wood Reaction wood of conifers, usually referred to as com pression wood, is developed as part of an orientation mechanism, i.e., a geotropic reaction to an inertial force (151). As stated by Larson (80): The interaction between auxin and sucrose in producing compression wood results in trac heids with larger diameters and extremely thick cell walls. Physically, compression wood trac heids are somewhat similar to those of juvenile wood produced in the high auxin environment in close proximity to the crown. Chemically, the normal metabolic pathways appear to be altered during compression wood formation so that pro duction of-constituents normally confined to the outer wall layers is perpetuated across the wall. Wardrop and Dadswell (146) pointed out that the reduction in tracheid length (Chapter II) results from an increase in"the number of anticlinal divisions in the cambium associated with the rapid eccentric radial growth that is usual with the development of compression wood. Thus there appears to be a similarity between such cells and the tracheids formed during the youthful period of the life of the tree. In many investigations, comparisons have been frequen tly made with material from the quadrant adjacent to the com pression wood to typify normal wood. In this study, however, comparisons are made on the basis of same tracheid length, since they may be more meaningful and sensitive. Moreover, in spite of the obvious morphological differences, the main difference between compression wood and normal wood is one of degree only (mildest to extreme type), other developmental features being similar to those occurring in normal wood (see 146). Average values of compression wood tracheid weights and their mean lengths are contained in Tables I and II, for holo- and alpha-cellulose skeletons, respectively. Relation ships between tracheid length and weight are shown in Figure 13. It is apparent that, even in the formation of compression wood, there is again a highly significant correlation between tracheid weight and length, explaining 71.7 and 90.0 per cent variation in holocellulose and alpha-cellulose skeleton weights, respectively, by the length factor alone. This behaviour of compression wood reaffirms the previously established relation ship for normal wood (see Figure ,9) . A general comparison between compression wood and normal wood indicates that, for the same tracheid length, no significant difference can be found with respect to holocellu lose tracheid skeletons (Figure 13). In contrast, differences seem to exist for alpha-cellulose skeleton weights, suggesting the existence of qualitative and/or quantitative differences. Examination of the relationships given in Figure 13 as well as the original data, however, show that, for the same tracheid length, compression wood tracheids may have slightly higher alpha-cellulose than juvenile wood (where tracheid lengths were mainly under 2.3 0 mm) but a lower amount compared to mature wood tracheids (where tracheid lengths were mostly,over 2.30 mm). In addition, since compression wood tracheids did not differ from those of normal wood in holocellulose content, they must contain more hemicellulose fraction than comparable normal wood tracheids,i.e., the difference between holo- and alpha-cellulose . The above behaviour could be related to age-dependent influence. For example, in the present instance, the compres-. sion wood selected was from Increments 73 and 83 from the pith (Appendix I). With this background, it can be proposed from the data obtained that, despite a decrease in tracheid length associated with compression wood formation, the age effect is still superimposed in the cambial behaviour. Hence tracheids are produced with slightly lower amount of alpha-cellulose than comparable normal wood tracheids. Similarly, the slightly higher alpha-cellulose values as compared to those of juvenile wood tracheids could be explained on the same principle. The results here are consistent with the findings of Australian workers (27). These authors emphasized that all degrees of compression wood may be encountered from the mildest form to the very extreme types, and that there is no basic difference between these and normal wood, in the mechanism of cell division, in the readjustment of daughter cells, or in the progressive development of cell wall, and the difference is in degree only. Since, in the present study, care was taken to include the extreme case of compression wood it was possible to bring out these differences of degree. In summary, it can be stated that the basic relation ship, developed with normal wood tracheids, seems to hold for compression wood as well, indicating a common regulatory mech anism in conifer cambial morphology. However, compression wood tracheids differ from normal wood tracheids in showing some qualitiative differences with regard to the basic carbohy drate skeletal fraction (alpha-cellulose). The remarkable ability of the tree for quick readjustments at cellular level, to changes brought about by extrinsic factors, is clearly emphasized in the physiology of compression wood formation. 5. Variation between earlywood and latewood It has long been recognized that earlywood and late wood of coniferous woods, within a growth increment, have different morphological, physical and chemical properties and behaviours. In almost every respect the differences between these two tissue types may be essentially as great as, if not greater than, those between wood zones within a single stem,. wood from two members of a single tree species, or between two species (59). However, anatomical evidence seems to indicate no significant difference in the actual amounts of cell wall material per tracheid in earlywood and latewood tracheids in conifers (14, 68, 121a). The apparent variations observed in specific gravity across an increment have been attributed largely to differences in tracheid volume rather than wall substance (70). It is of interest, therefore, to examine the variation of individual tracheids from the two arbitrary-fractions of the increment. As a preliminary approach to this, the tracheid length-weight relationships obtained are re-examined separately for earlywood and latewood. The curvilinear multiple regression relationships of best fit, for the combined data of the nine coniferous woods, are given below. HOLOCELLULOSE ALPHA-CELLULOSE Earlywood Earlywood Y = 0.0291 + 0.0506X2 Y = -0.2418 + 0.1845X + 0.0377X2 2 r = 0.951; SEE = 0.198; R2 = 0.972; SEE = 0.130; DF = 345 DF = 356 Latewood Latewood Y = -0.3770 + 0.2365X Y = -0.3705 + 0.2602X + 0.0310X2 + 0.0301X2 0.939; SEE = 0.263; R2 = 0.959; SEE = 0.169; DF = 447 DF = 383 Earlywood + Latewood Earlywood + Latewood Y = -0.1708 + 0.1201X Y = -0.3019 + 0.2214X + 0.0412X2 + 0.0341X2 R2 = 0.936; SEE = 0.255; R2 = 0.965; SEE = 0.152; DF = 794 DF = 742 Where Y = ovendry weight of tracheid in g x 10 X = tracheid length in mm - green condition. These relationships confirm that, irrespective of origin of the wood within the increment, there is a high . degree of correlation between length and weight among tracheids, The 2 individual r values obtained,, for earlywood and latewood data, and their standard errors of estimate,do not differ much from those of the combined values. More detailed analyses were done of both holo-and alpha-cellulose data to explore further the behaviour of earlywood and latewood, as well as to determine differences between and within species. Results based on 't' tests indicate that, for holocullulose, weights of latewood tracheids frequently (six of the nine species) are significantly higher than those of earlywood tracheids (Table I). This tendency is less pro nounced for alpha-cellulose skeletons, where five of the nine species examined did not show any significant difference in tracheid weight between the two tissue fractions (Table II). Interestingly, in both cases (holo- and alpha-cellulose), 't' tests for the combined data showed that latewood tracheids are always significantly higher in weight than earlywood trac heids, thus confirming the findings of others based on gross analyses (1, 52, 114). It is to be noted, however, that where significant differences exist, latewood tracheids are longer than their counterparts in earlywood (Tables I and II). Because of the above reason, analyses of covariance tests were done to compare tracheid weights after adjusting for length differences. Results of these analyses are given in Tables IX and X. It can be seen that, for holocellulose skeletons, no significant differences can be found between the six trees of Douglas-fir examined (Table IX). As expected, differences exist between species (Table X) which, as already mentioned, are attributable to differences in the,'slopes' and 'levels' (Tables V and VI). The point of interest, however, is the behaviour of earlywood and latewood, in that latewood holocellulose tracheid skeletons are significantly heavier than earlywood, both within and between species. In contrast to the above, no such differences are evident in alpha-cellulose skeleton weights. This is not only true for the six Douglas-fir trees examined, but also for all the nine coniferous woods. In regard to overall behaviour, highly significant differences are found between species (Table X) while within species (i.e., six Douglas-fir trees), the dif ferences are significant only at the 5 per cent level (Table IX). The last mentioned differences could well result from the al ready described characteristics of juvenile wood and compression wood. In summarizing the variations in earlywood and latewood tracheid skeleton weights, for all the nine coniferous woods, it is evident that, for the same tracheid length, earlywood and latewood would have similar amounts of alpha-cellulose, yet differing amounts of hemicelluloses. Qualitative differences 68. in the hemicellulose fraction have also been reported by other workers from gross wood and tissue analyses (1, 78, 90, 114) but, in addition, these reports show differences in alpha- . cellulose percentage as well. In one study, however, based on tissue analysis on Scotch pine, Meier (90) obtained about the same percentage of alpha-cellulose for both earlywood and late wood. In this respect the present results are in agreement with Meier's findings. As noted,.gross analyses give only an estimate of the total fraction (percentage) of carbohydrate for the tissue as a whole, which adds to the difficulty of interpreting data based on individual cell weights. That is, gross analyses are not sensitive enough to reflect the true values for weights of individual tracheids from different origins. Evidence for a Common Relationship in Conifers From foregoing observations on the relationships between length of tracheid and its carbohydrate fraction, it is evident that a common relationship could be derived for all conifers irrespective of wood origin. In other words, the relationship is not necessarily an age dependent one in that both tracheid length and cellulose content increase with age in a stem within a species. As expected, there are statistical differences in the slopes of regression lines depending on species. If, from practical rather than strictly academic statistical considerations, a common regression line could be fitted for all these data, it would explain a significant part of the variability in tracheid weight, within tolerable limits, of standard error of the estimate. Such a randomi zation of tracheids would provide a unique measure for exami ning fundamental morphological and chemical behaviour of.the conifer wood basic unit, since variability introduced by species, environmental and ontogenetic factors is minimized. Thus, a general behavioural pattern could be deduced. The combined relationships of Figure 10 confirm that, despite statistical differences in slopes between species, a significant relationship exists among tracheidsof different conifers, indicating that the physiological mechanisms governing tracheid length and weight relationships are the same within and between species, including wood formed over a wide range of tree age, normal wood and reaction wood. Thus, if 'tracheid' is considered as one population, the relationships shown in Figures 10 and 11 are justified, since the correlations obtained are highly significant. In addition, the standard errors of estimates are no higher than those noted for some of the indi vidual species (e.g., Figure 9--Holocellulose and Figure 1--Alpha-cellulose), indicating the paramount importance of length factor with regard to the cell length-weight relationships. It is entirely possible that adding the effects of other morpholo gical characteristics, such as cell diameter.and wall thickness, not done herein, would improve the relationship. As has been 70. shown, these three morphological features are inter-related in conifer cambial morphology with length factor appearing to be the more dominant one. One advantage of this approach is that, based on the length factor alone, the amount of carbohydrate in any tracheid could be estimated irrespective of the origin of wood. In other words ,• for .practical- consideration, the car bohydrate. fraction would be similar for .tracheids of the same •length.;, in the coniferous woods examined.. "• - . - . - - • In view of such a strong relationship between cell length and weight, and a priori considerations from cell morphology, it is reasonable to hypothesize that the amount of carbohydrate fraction in an individual tracheid is a function of its length. It is not known whether the relationship is a direct one, in that the amount of carbohydrate fraction depends on the length of the cell; or whether it is an indirect one, in that both variables respond to some finite . factor. This is difficult to assess, since factors which influence the extent of cell enlargement and wall development following cell division are still not fully understood. As pointed out by Bannan (8), eventual size of derived coniferous elements is determined largely by size of the origi nating fusiform cambial cells, post-divisional elongation being relatively slight in conifers. In an earlier work Bannan and Bayly (10) showed that, in conifers, cell size and extent of ray.contacts are important factors for survival of cambial daughter cells. The longest cells with the best ray contacts have best opportunity to survive and develop, whereas the shortest fail or form new ray initials. This creates an effec tive feed-back regulative system which provides for an equidis tant initiation of new rays (167). The importance of rays to fusiform initials lies in their function of transfer and stor age of nutrients, required for the sustenance, growth, and division of the initials; where the longest initials are well provided with ray contacts they possibly have access to abun dant nutrients. A proper balance of ray contacts commensurate with tracheid length, would provide the derived cells with sufficient nutrients for the build-up of a rigid framework as the cell matures. It appears from these consideration that a direct relationship may exist between the amount of carbohydrate fraction and tracheid length. The results obtained herein are thus consistent with findings reported by others, in that the physiological mechanisms governing the length-weight relation ships are histories of growth relationships possibly predeter mined in the originating initials, according to their size. The relationship described above has another important implication in that longer tracheids have a greater weight per unit length than shorter ones (see Figures 10 and 11), a conclu sion also reached earlier by Clark (24), who used fibre 'coarse ness' to determine . fibre w£'i';ght per unit length of commercial sulphite pulp samples of coniferous wood. As noted, for longer tracheids the average percentage of cell wall volume per unit length will be higher since length is correlated positively with diameter (50, 54, 56, 158) and wall thickness (108, 140).. Also, the possibility exists that, for shorter tracheids, the relative contribution of tips, which undoubtedly, weigh less, (146, p.391), is more than for longer tracheids. If this is so, such a relationship may have important impli cations from the practical point of view. Applicability of the Relationship A somewhat different approach was taken to apply the relationship as developed to problems such as effects of fertilizers on wood properties. The relationship has already been applied to test differences between compression wood tracheids and those of normal wood. It is to be noted, however, that compression wood, on the one hand, is a morphogenetic phenomenon in trees for stem regulatory or reorientation func tion and, as such, it is a natural thing that trees respond automatically to produce such wood. On the other hand, a fac tor such as chemical fertilization may be artificial, created to raise the fertility level of a soil in order to increase the yield or bring non-productive land into production. Also, fer tilization appears to have little direct effect on cambial activity or wood development since there is no firm evidence to indicate that the cambium can mobilize nutrients directly 73. from the transpiration stream (70). Fertilization exerts its primary or direct effect on crown, and also on root develop- • ment; and its effect on wood growth is both secondary to and the result of crown responses (79). Thus it appears that, following the addition of a fertilizer, all the physiological processes in the tree are activated, so to speak. From the above viewpoint, it would be of interest to see what happens to the tracheid length-weight relationship that is operative in normal wood in response to the action of fertilization. Three possibilities exist: " (a) There might be a gain in carbohydrate amount per unit length of tracheid owing to the increasing availability of photosynthate, since there will be an increase in photosynthetic surface and pos sibly the photosynthetic efficiency of foliage (18, 139); (b) decrease in cell weight per unit length due to overproduction of cells in a given period; and (c) no change in spite of overproduction since the tree reverts to a temporary period of juvenility by producing shorter cells consequent to increase in the rate of anticlinal cambial cell divisions.. The best way to examine these possibilities would be to compare the tracheid length-weight relationships obtained from the fertilized trees, .with that obtained for the combined data for conifers. For obvious reasons, this cannot be done in the present case, i.e., because of 'slope' and 'level' differences between some of the species studied. In view of this, compari sons can only be made with the relationship obtained for normal Douglas-fir wood. Before investigating these aspects, it would be approp riate to trace the changing pattern of tracheid weight and length, starting before the fertilizati6n period. Such a pattern is shown in Figure 14, for a 29-year^old Douglas-fir tree which was fertilized at ages 21 and 22. For comparison, data for increment width, and mean tracheid length based on 50 macerated tracheids per increment, were also plotted in the same figure. It can be seen that tracheid weight of both holo- and alpha-cellulose skeletons decreased or remained static immediately following fertilization. In three of the four cases this dec rease in weight continued for two years following the year of application (or one year after refertilization); thereafter there was a sudden increase in weight followed again by a de crease due to the formation of compression wood. The only exceptions were the latewood holocellulose skeletons, which showed increase in weight after a temporary decrease, for only one year following the year of fertilization. Subsequently, their behaviour was similar to the general pattern described above. This increase in latewood tracheid weight (year 23) may have been due to the increased availability of current photosynthate to the cambial region, after already meeting the new needle requirements (49, 80). Further, this was primarily an increase in the hemicellulose fraction since no increase is evident for alpha-cellulose skeleton weights in that year. Both mean tracheid length for the increment (Figure 14) and average tracheid length of all the tracheids weighed (Tables III and IV) followed the same general patterns noted for tracheid weight. As expected, there was a steep increase in increment width for the years 21 and 22, following first fertilization (year 21); subsequent to year 23 there was no further increase, but growth rate was still quite high as compared to that before fertilization. The above pattern provides evidence of changes in tracheid weight following fertilization. However, it is not clear whether the apparent changes are independent of variation in tracheid length or whether they are directly related to tracheid length. In order to examine the variability in this tree and others, more fertilized wood material of Douglas-fir, available in the form of increment cores from another study (119), was employed. As noted, three fertilizer compositions had been selected, namely Urea, NPK„ and (NH^^SC^. Two incre ments, i.e., 15 and 16 (including and following refertilization), from one tree each of these treatments were studied. The reasons for selection of these treatments were that, while all generated response in increment width, they showed also definite increase (Urea) or decrease (NPK, (NH.)„S0.) in specific gravity following fertilization treatments. Some selected wood and tracheid properties for the three trees are given in Table XI. The main purpose here was to determine the carbohydrate skele- -ton weights per unit length, and to relate these to changes observed in gross wood specific gravity. Ifju and Kennedy (60) showed a high correlation between specific gravity.and percentage of cell wall, which would suggest a possible con version from cell wall area to density values. Since cell length is correlated with cell diameter and wall thickness as well as weight, a possibility of relating cell weights to specific gravity may be explored. Such a relationship has already been indicated for the 500-year-old Douglas-fir tree (Figure 12). Relationships between tracheid length and weight, in 2 the fertilized Douglas-fir trees, are indicated in Figures 15 to 18, along with that obtained for Douglas-fir normal wood. For effective comparison, the analyses done in each case are also given in each figure. _A summary of the overall comparisons among the treated and untreated Douglas-fir trees is shown in Table XII. It can be inferred from the combined analysis (Table XII A) that no significant difference exists, for holo cellulose skeletons, between the normal wood tracheid weight-2 It appears from Figure 15 that the regression line of holocellulose tracheid skeletons of the Urea treated tree is different from that of normal Douglas-fir. Statistical analyses, however, indicate that the apparent differences are not signifi cant. This is true because the standard error of estimate (SEE) of normal Douglas-fir (Figure 9—Holocellulose) is about three times that of the Urea treated tree and thus it would include all values of the latter. length relationship and that of fertilized wood. When each treatment was compared individually with normal Douglas-fir (see Figures 15 to 18; Holocellulose), with one exception (NH4N03 treatment—Figure 18), no significant difference could be observed. Contrary to the above, the combined analysis for alpha-cellulose skeletons points out significant differences (Table XII A) among and within the five Douglas-fir trees. The same behaviour was noted when treatments were compared indivi dually with normal Douglas-fir (Figures 16 to 18). Interestingly, alpha-cellulose skeletons' from treatment Urea did not differ either in slope or level from normal Douglas-fir (Figure 15). Comparisons between the treatments (Table XII B), for holocellulose skeletons, indicate that there is no significant difference in slopes, that is, they are parallel. This implies that the behaviour (response) of the different treatments, with respect to tracheid length-weight relationship, is similar. But the magnitude of response in these treatments is different since they differed in levels. The only exception was between Urea and NPK which did not differ in either levels or slopes. In view of the level differences, it can be concluded that these treatments may differ in their effects on wood properties. Thus, if the regression equations of best fit (Figures 15 to 18) are examined for predicted values, Urea treatment shows higher values in that holocellulose tracheids are heavier per unit length, than comparable ones fromthe other treatments. It is 78. therefore possible that changes associated with cell weight are responsible, along with other factors (see Table XI), for the observed changes in gross wood specific gravity in these fertilized trees. The behaviour of alpha-cellulose skeletons from dif ferent treatments is somewhat different than that of holocellu lose skeletons. Significant differences exist between the regression lines of treatments Urea, NPK and (NH^) in their respective slopes, while Urea and NH^NO^ did not show such difference (Table XII B). With regard to levels, Urea differed significantly from (NH^) 2S(-)4 kut not from NPK and NH^NO-j. Here again, when comparisons are made on the basis of regression equations of best fit (Figures 15 to 18), the Urea treated tree showed higher values (predicted tracheid weight), per unit length, than those of the other treatments. In this respect, results for alpha-cellulose skeletons are similar to the findings reported for holocellulose, thus confirming the argument advanced for the relationship between individual tracheid weight and gross, wood specific gravity. Although the above comments refer to tracheid weight-length relationships in fertilized wood, they also provide addi tional proof for the thesis hypothesis in that the physiological mechanism governing the tracheid weight-length relationships in conifers is once again the same in the fertilized trees. From the biological viewpoint, this means that while readjustments for increased growth rate are taking place, the suggested relationship is maintained. Obviously this is achieved by a temporary reversal to juvenility, so that changes in tracheid weight are basically a reflection of changes in tracheid length. While this seems to be the general case, within the limits of the suggested relationship there were some statistical differ ences among the treated trees in their magnitude of response (weight per unit length of tracheid) indicating the possibility for differential effects of fertilizers on wood properties. Many studies on the effects of fertilization on wood properties have shown a decrease in tracheid length associated with increased growth rate (71, 108, 117). It is evident that a tree has this remarkable type of mechanism of reverting to temporary juvenility as a result of an increased rate of anti clinal divisions in the cambial cells. These aspects and re lated processes are well documented in:the extensive works of Bannan (8). In view of such effects, comparisons were made between behaviour of fertilized wood and that of juvenile wood (Table XII B). From these analyses, it is clear that no signi ficant differences can be found either in slopes or levels of holocellulose regression lines, between "juvenile wood tracheids and those from the treatments Urea, NPK and (NH^^SO^. But a significant difference exists (in slopes only) when comparing juvenile wood tracheids with those from treatment NH^NO^ . This difference in behaviour could be attributed to the age factor of this tree. For example, the NH^NO^ fertilized tree was 21 years old at the time of first fertilization (Figure 14), 80. whereas the Urea, NPK, (NH^) trees were age 13 (very close to juvenile period—see Table XI) when fertilized. Consequently holocellulose tracheid skeletons from treatment NH^NO^ differed from those of juvenile wood. Analyses for alpha-cellulose indicate no significant differences in regression line slopes between the juvenile wood tracheids and those for the treatments Urea, NPK, and (NH^)2S04 (Table XII B). Once again results for treatment NH^NO^ differed significantly in slope from those of juvenile wood, suggesting that these two tracheid populations differ. The interesting point, however, is the significant difference in levels of regression lines between juvenile wood and treatments Urea and NPK. In contrast to this, alpha-cellulose skeletons from treatment (NH^) 2^4 not show significant difference from those of juvenile wood either in slopes or levels of regression lines. The implications of these findings are important in that treatment Urea and NPK differed from (NH^)2S04 with respect to resulting carbohydrate fraction per unit length of tracheid. Furthermore, when compared with normal Douglas-fir, only Urea (Figure 15) did not show any significant difference, whereas both NPK and (NH^)2S04 (Figures 16 and 17) differed significantly in the amount of alpha-cellulose per unit length of tracheid. The behaviour of tracheid skeletons from treatment NH^NO^ is somewhat different from the others. When compared to normal Douglas-fir, the NH.NO-. treated tree (Figure 18) showed 81. significant differences in slopes for both holo- and alpha-cellulose skeletons, suggesting that the tracheid populations from these two trees differ. Interestingly, the upper part of the curve in Figure 18 (Holocellulose—where X = 3.92) indi- . cates that these tracheids may have more holocellulose per unit length than comparable normal wood tracheids. Such behaviour may be attributed to the increase in weight noted in the latewood holocellulose skeletons (years 23 and 24-—Figure 14), following refertilization. In conclusion, it can be said that the basic relation ship suggested for normal wood, between length and weight, is maintained in the wood from fertilized trees; such trees adjust quickly to the increased growth rate by a temporary reversal to juvenility and produce shorter tracheids. In addition, the possibility exists for a decrease in weight (alpha-cellulose) per unit length as exemplified by treatments NPK and (NH^^SO^. The period of temporary juvenility and the degree of reduction in tracheid length, may well be related to rate of increase in growth as well as to other factors such as initial tree vigor and age. Where the trees are growing vigorously and still in the period of rapid growth (46a), they might recover speedily from this temporary reversal to juvenility. If the trees are old and past this period, it may take longer before they reach the expected norm in tracheid length-weight. In the present instance the fertilized trees used were young, at the time of fertilization, and it is not known whether the reported trends 82. in tracheid weight will be the same for older stands. Never theless, the results obtained do not indicate any drastic change in fibre properties even in older trees, when fertilizer is applied in moderation and with due consideration to other silvicultural and management practices (79). Some Practical Considerations Two practical aspects will be discussed in this sec tion, namely, experimental yield of carbohydrate fraction and the implication of the tracheid length-weight relationship obtained, to areas of wood and pulp science and industry. Yield of holocellulose fraction obtained by the present method, using peracetic acid (and NaBH4 reduction), has been reported by the originator (83) to be somewhat intermediate between chlorite holocellulose (163) and that of TAPPI (chlorine-ethanolamine (128)) methods. However, the peracetic acid method adopted has been suggested to give a superior holocellulose with a D.P. of 3050 to 3600, 93 to 100 per cent recovery of major polysaccharides and minimum of carboxyl groups (83). The yields obtained in the present study for both holo- and alpha-cellulose (Appendix II), are comparable to those reported by others (20, 123, 135). There appears to be some peculiarity for this method (peracetic acid) in that the holocellulose yields for earlywood are slightly higher than those from latewood. Similar differ ences in yield have also been"recorded by Leopold for loblolly 83. pine wood (83) . The reason for this difference is obvious since the cooking cycles needed for earlywood were only three as compared to five for latewood. But this behaviour does not seem to influence the alpha-cellulose yields, which showed the expected difference in that latewood values were higher (Appen dix II) . Corrections to the weight of individual tracheids, based on yield, have not been made since it is difficult to assess such minor differences in single tracheid weights. In addition, it is incorrect to assume that different woods will have the same yield, due to inherent variabilities. As such, comparisons based on the actual values obtained will be more meaningful and realistic than those based on adjusted values. However, in this work, care was taken to see that the same delignification conditions were employed for all the woods Spot checks of some of the pulp samples indicated comparable amounts of lignin (Appendix II). In the light of the present study,.it appears that the length-weight relationship of the carbohydrate skeletons obtained can be used profitably, by re-emphasizing the impor- . tance of tracheid length for selection of individuals in breedi and vegetative propagation work. Its importance to the pulp and paper industry lies in the judicious selection of indi viduals within species having longer fibres, not only because of the relationship between fibre morphology and paper properties 84. (13, 29, 34, 150), but also because of possible differences in carbohydrate yield. In addition, the relationship between length and weight is a curvilinear one, meaning that for every increase in unit length of fibre, there will be a dispropor tionate increase in the amount of cellulose (Figures 10, 11). It therefore appears that longer fibres could result in a higher pulp yield due to fewer end wall contact areas per given weight or volume of wood, which consequently would result in less lignin (39) . This brings up an important point related to pulp evaluation based on the number of fibres per gram, and its significance for analyzing the (pulp) characteristics as affected by species, yield or pulping process. The implica tions of this and related aspects have been discussed recently by Horn and Coens (58). At considerable risk of oversimplification, it can be stated that if longer fibres result in more cellulose content, per given weight or volume of wood, it would seem advantageous to grow trees with longer fibres rather than propagating species or trees with shorter fibres, if the wood is to be used in the pulp and paper industry; briefly, this is a high yield forestry concept. Similarly, Leslie (84) suggested some years ago the concept of 'cellulose forestry,' whereby forests would be scientifically managed for the continuous production of wood possessing optimum properties for paper and/or pulp making. Breeding for fibre quality is justified on these grounds. 85. Moreover, considerable evidence exists as to the heritable nature of fibre length (8, 9, 37, 47, 61, 100, 154, 169, 170). Although estimates for gross wood cellulose content do not show any significantly heritable nature, higher yields of cellulose can still be obtained by breeding for fibre length since the individual fibre composition is highly significantly correlated with length, as brought out in this study. Another point that may merit further consideration, is shortening rotation age of the stand because of increasing pressure for land to produce maximum fibre returns. The length-weight relationship of tracheids can be employed pro fitably, to suit individual company requirements, in deter mining the rotation age. This point may be worth investiga tion in particular for those countries setting up plantations or countries striving to produce enough wood to meet the increasing demands of the pulp and paper industry. CHAPTER V RECOMMENDATIONS FOR FURTHER RESEARCH Much .of the information in this paper on tracheid weight (carbohydrate skeleton) and length relationships, represents a reconnaissance of an overall relationship that exists between and within coniferous (stem) woods. Its primary objective was to determine the more striking phenomenon encoun tered, and provide some basis for subsequent and more detailed investigations. Although the project has been designed to include a wide range of material so that conclusions will be applicable to conifers in general, many questions needing further study have been raised. For example, a study of the relationship between individual fibre weight and intra-incre-mental specific gravity.merits further consideration. Since anatomical evidence indicates no significant difference in the actual amounts of cell wall material per tracheid between earlywood and latewood tracheids, it would be of interest to pursue this aspect further with regard to tracheid weights. A preliminary approach has already been made herein, but results are inconclusive. Other studies that need further research include the relationship that may exist between individual fibre weights and length, for angiosperms, so as to evaluate results for both practical and biological significance. 87. It has been observed in this study that a reduction in tracheid length occurs in alpha-cellulose tracheid skeletons subsequent to removal of the hemicellulose fraction in the holo cellulose tracheid skeletons. A study could be conducted to test whether the rate'or degree of reduction is related to tissue type (earlywood/latewood) and species. This would provide a more exact length-weight relationship for alpha-cellulose, so that more realistic comparisons can be made'with those of holocellulose skeletons.• This information could be useful in developmental studies involving the chemical changes in cells, accompanying differentiation into xylem. Finally,-weigh ing«o.f:other cell elements such as vessels and parenchyma could be done. Sastry and Wellwood (118) sugges ted several areas of research where tracheid weights could be employed profitably. For example, single fibre weights can be used in determining strength-weight or strength/density ratios for individual fibres, density of cell walls in cell wall porosity studies, in quantitative studies involving wood forma tion, and assessing the effects of silvical and environmental variables on wood properties. The uniqueness of this approach thus facilitates direct comparison of characteristics such as strength properties, dimensions and carbohydrate content on the same fibre. CHAPTER VI CONCLUSION 1. A quartz ultra-microbalance has been devised for accurate weighing of single cells. The balance has a weighing range of 0.06 to 14 ug and a precision ±0.03 yg. This balance was used for quantitative determination of both holo- and alpha-cellulose in individual tracheid skeletons of nine coni ferous woods, belonging to nine different genera and seven families. 2. Length of a tracheid and the amount.of carbohy drate in it are highly significantly correlated. The estimated variations accounted for in holocellulose fraction and that of alpha-cellulose, by the length factor alone, were respectively 91.9 and 95.7 per cent for all species studied. A direct, physiological relationship between length of tracheid and its weight (carbohydrate skeleton) is suggested. 3. Between species differences were observed with respect to slopes and levels of regression lines. Differences in cell diameter and wall thickness may have contributed to this statistical behaviour. For practical considerations, the coniferous tracheids could be grouped as one population. Small 2 standard deviation, - and highly significant r values, between tracheid weight and length, support this argument. 89. 4. The radial variation of single tracheid weight followed an increasing trend up to about 150 years and decreased subsequently in the overmature wood. 5. For the same tracheid length, individual tracheids of juvenile wood had significantly lower amounts of alpha-cellulose than those from mature and overmature wood, whereas differences were nonsignificant for holocellulose. 6. Overmature wood tracheids had significantly lower carbohydrate amount than those from mature wood, for the same tracheid length. Results are comparable to those reported by others:- based on gross analyses. 7. Compression wood tracheids did not differ from normal wood tracheids in.the amount of holocellulose, for the same tracheid length, but differed significantly in,respect to alpha-cellulose. It was concluded that compression wood tracheids may contain lower alpha-cellulose than mature (and normal) wood, but higher amount when compared to juvenile wood, for the same tracheid length. 8. Significant differences in morphological, phy sical, -and chemical properties between earlywood and latewood of coniferous species have long been recognized. Evidence ob tained in this study shows that, for the same tracheid length, both earlywood and latewood had similar amounts of alpha-cellu lose, but differing amounts of holocellulose, latewood tracheids being heavier. 90. 9. Variations in specific gravity of wood in fer tilized trees have been attributed formerly to differences in growth rate and percentage of latewood. Evidence was obtained in the present study that, in addition to the above variables, weight'per unit length of carbohydrate tracheid skeletons may also contribute to an alteration in specific gravity. Changes' in tracheid weight, associated with fertilization, for the most part, were related to tracheid length." Thus the basic relation between length and weight still exists in fertilized wood; where there is a reversion to a phase of temporary juvenility, trees produce shorter tracheids compared to normal wood of a compar able age.' The data indicate that some treatments may, however, generate less alpha-cellulose per unit length when compared to others. This difference in behaviour was suggested as a possible cause for changes observed in gross wood specific gravity of fertilized trees. LITERATURE CITED Bailey, A.J. 1936. Lignin in Douglas fir. Composition of the middle lamella. Ind. Eng. Chem. (Anal.) 8: 52-55. Bailey, I.W. 19 20. The cambium and its derivative tissues. II. Size variations of cambial initials in gymno-sperms and angiosperms. Am. J. Bot. 7: 355-367. 1923. The cambium and its derivative tissues. IV. The increase in girth of the cambium. Am. J. Bot. 10: 499-509., and H.B. Shephard. 1915. Sanio's laws for the variation of size in coniferous tracheids. Botan. Gaz. 60: 66-71. and A.F. Faull. 1934. The cambium and its derivitive tissues.. IX. Structural variability in redwood (Sequoia sempervirens) and its significance in the identification of fossil woods. J. Arnold Arboretum 15: 233-254. Bannan, M.W. 1965. The length, tangential width, and length/width ratio of conifer tracheids. Can. J. Bot. 43: 967-984. 1966. Cell length and rate of anticlinal divisions in the cambium of the sequoias. Can. J, Bot. 44: 209-218. . 1967. Anticlinal divisions and cell length in conifer cambium. For. Prod. J. 17(6): 63-69. 1970. A survey of cell length and frequency of multiplicative divisions in the cambium of conifers Can. J. Bot. 48: 1585-1589. 10. and LL. Bayly. 1956. Cell size and survival of conifer cambium. Can. J. Bot. 34: 769-776. 11. Berlyn, G.P. and R.E. Mark. 1965. Lignin distribution in wood cell walls. For. Prod. J. 15: 140-141. 12. Berzins, V. 1966. Micro kappa numbers. Pulp Paper Mag. Can. 67: T206-T208. 13. Besley, L. 1959. Relationship between wood fiber proper ties and paper quality. PPRIC, Montreal, Canada. Tech. Rept. 156. (Woodl. Res. Index No. 114). 92. 14. Bethel, J.S. 1964. The developing architecture of secondary xylem in conifers. For. Sci. 10: 89-91. 15. Bisset, I.J.W. and H.E. Dadswell. 1950. The variation in cell length within one growth ring of certain angiosperms and gymnosperms. Australian Forestry 14: 17-29. 16. Britt, K.W. 1965. Determination of fiber coarseness in wood samples. Tappi 48:.7-11. 17. . 1966. Fiber coarseness in wood. Tappi 49: 202-206. 18. Brix, H. and L.F. Ebell. 1969. Effects of nitrogen fer tilization on growth, leaf area, and photosynthesis rate in Douglas-fir. For. Sci. 15: 189-196. 19. Brown, CL. 1970. Physiology of wood formation in coni fers . Wood Sci . 3: 8-22. 20. Browning, B.L. 19 63. The composition and chemical reactions of wood. In.The Chemistry of Wood (ed. Browning). Inter Sci. Pub., New York. pp.57-102. 21. Bucher, H. and L.P. Widerkehr-Scherb. 1948. Morphology and Structure of Wood Fibres. Cellulosefabrik Attisholz AG. Switzerland. 22. Byrd, V.L., Ellwood, E.L., Hutchings, R.G. and A.C. Barefoot, Jr. 1965. Wood characteristics and kraft paper properties of four selected loblolly pines. Part II. Wood chemical constituents and their rela tionship to fiber morphology. For. Prod. J. 15: 313-320. 23. Chalk, L. and G. Ladell. 1962. Tracheid length in Corsi-can pine. Extr. from Rep. For. -Res. For. Comm. London. 1960/1961: 113-114. 24. Clark, J.d'A. 1962. Weight average fiber lengths - A quick visual method. Tappi 45: 38-45. 25. C6t£; W.A., Jr. (ed.) 1965. Cellular Ultrastructure of Woody Plants. Syracuse Univ. Press, Syracuse. 26. Dadswell, H.E. 1958. Wood structure variations occurring during tree growth and their influence on properties. J. Inst. Wood Sci. 1: 1-23. 93. 27. Dadswell, H.E.-, .Wardrop, A.B. and A.J. Watson. 1958. The morphology, chemistry and pulp characteristics of reaction wood. In Fundamentals of Paper Making Fibres (ed. Bolam). B.P. & B.M.A., London., pp.187-219. 28. , Watson, A.J. and J.W.P. Nicholls. 1959. What are the wood properties required by the paper industry in the trees of the future? Tappi 42: 521-526. 29. an& A.J. Watson. 1961. Influence of the morphology of wood pulp fibres on paper properties. C.S.I.R.O., DiVi of Forest Prod.,- Melbourne, Australia. Reprint 485. 30. de Zeeuw, C. 1965. Variability in wood. In Cellular Ultrastructure of Woody Plants (ed. cBt-) . Syracuse Univ. Press, Syracuse. pp.457-471. 31. Dinwoodie, J.M. 1961. Tracheid and fibre length in timber: A review of the literature. Forestry 34: 125-144. 32. . 1963 . Variation in tracheid length in Picea sitchensis Carr. Dept. Sci. and Ind. Res., Forest Prod. Research, London. Special Rept. No. 16. 33. - : . 1965. Tensile strength of individual compression wood fibres and its influence on proper ties of paper. Nature 205: 763-764. 34. . 1965. The relationship between fiber morphology and paper properties: A review of litera ture. Tappi 48: 440^-446. 35. Doerner, K. 1964. Some causes and effects of horizontal density variation in tree stems. For. Sci. 10: 24-27. 36. Dunning, C.E. 1968. Cell-wall morphology of longleaf pine latewood. Wood Sci. 1: 6 5-7 6. 37. Echols, R.M. 1955. Linear relationship of fibrillar angle to tracheid length, and the genetic control of tracheid length in slash.pine. Trop. Woods 102: 11-22. 38. Einspahr, D.W., Goddard, R.E. and H.S. Gardner. 1964. Slash pine, wood and fiber property hetitability study. Silvae Genet. 13: 103-109. 94. 39. • Einspahr, D.W., van Buijtenen, J.P. and J.R. Peckham. 1969.- Pulping characteristics'of ten-year lob lolly pine selected for extreme wood specific gravity. Silvae Genet. 18: 57-61. 40. Elliott, G.K. 1960. The distribution of tracheid length in a single stem of Sitka spruce. J. Inst. Wood Sci. 5: 38-47. 41. El-Osta, M.L. 1970. A literature survey on core and mature wood. Univ. British Columbia, Fac. For. Rept. (Unpubl.). 42. Erickson, B. 1962. Mojligheterna att oka skogevaxten genom markforbattrande atgarder. Sv. Skogsvn. for .,' Tidsskr. 2: 195-202. (Original not seen. From Klem, G.S., Ref. No. 71). 43. Erickson, H.D. 1962. Some aspects of method in deter mining cellulose in wood. Tappi 45: 710-719. 44. - and G.M. Lambert. 1958. Effect of fertili zation and thinning on chemical composition, growth and specific gravity of Douglas fir. For. Sci. 4: 307-315. 45. Esau, K. 1965. Plant Anatomy. 2nd ed. Wiley, New York. 46. Gerry, E. 1916. Fiber measurement studies: A comparison of tracheid dimensions in longleaf pine and Douglas-fir, with data on the strength and length, mean diameter and thickness of wall of the tracheids. Science 43: 360. 46a. Gessel, S.P., Turnbull, K.J. and F.T. Tremblay. 1960. How to fertilize trees and measure response. Published by National Plant Food Institute, Washington, D.C. 47. Goggans, J.F. 1962. The correlation, variation, and inheritance of wood properties in loblolly pine (Pinus taeda L.). N.C. State College, School of For., Tech. Rept. No. 14. 48. 1965. Variations of tracheid width and wall thickness within and between trees of southern pine species. Section 41, IUFRO Meeting, Melbourne, Australia. 49. Gordon, J.C. and P.R. Larson. 1968. Seasonal course of photosynthesis, respiration, and distribution of 14C in young Binus resinosa trees as related to wood formation. Plant Physiol. 43: 1617-1624. 95. 50. Graff, J.H. and R.W. Miller. 1939. Fiber dimensions. Paper Trade J. 109(6): 31-37. 51. Hagner, S.O. 1967. Fertilization as a production factor in industrial forestry. Univ. British Columbia, H. R. MacMillan Lecture series. No. 37. 52. Hale, J.D. and L.P. Clermont. 1963. Influence of prosen-chyma cell-wall morphology on basic physical and chemical characteristics of wood. J. Polymer Sci. C(2) : 253-261.*'' 53. Harada, H. 1965. Ultrastructure ,and organization of gymnosperm cell walls. In Cellular Ultrastructure of Woody Plants (ed. CoteT. Syracuse Univ. Press, Syracuse. pp.215-233. 54. Harlow, W.M. 1927. The effect of site on structure and growth of white cedar, (Thuja occidentalis L.). Ecology 8: 453-470. 55. and L.E. Wise. 1928. The chemistry of wood. I. Analysis of wood rays in hardwoods. Ind. Eng. Chem. (Anal.) 20: 720-722. 56. Heinig, M. and F.A. Simmonds. 1948. Length and width of unbleached sulphate pulp fibers from certain western woods. The Paper Ind. and Paper World 30: 738-741. 57. Hiller, CH. 1964. Correlation of fibril angle with wall thickness of tracheids in summerwood of slash and loblolly pine. Tappi 47: 125-128. 58. Horn,,R.A. and C.L. Coens. 1970. Rapid determination of the number of fibers per gram of pulp. Tappi 53: 2120-2122. 59. Ifju, G. 1963. Intra-increment variations in some mechanical, physical and chemical properties of Douglas-fir. Univ. British Columbia, Fac. For. Rept. (Unpubl.). 60. and R.Wi Kennedy.. 1962. Some variables affecting microtensile strength of Douglas-fir. For. Prod. J. 12: 213-217. 61. Jackson, L.W.R. and J.T. Greene. 1958. Tracheid length variations'and inheritance in slash and loblolly pine. For. Sci. 4: 316-318. 62. Jane, F.W. 1956. The Structure of Wood. A. & C. Black, London. 96. 63. Jayme, G. and D. Fengel. 1961. Beitrag zur Kenntnis des Feinbaus der Frlihholz-tracheiden. Beobachtungen and Ultradiinnschnitten von Fichtenholz. Holz Roh-Werkstoff 19: 50-55. 64. Jensen, W., Virkola, N.E., Huikari, 0. and K. Paorlahti. 1964. Wood grown in fertilized forests as a raw material for pulp. Paper and Timber 46: 443-452. 65. Jones, G.B. 1958. Variations in cell size and cell wall thickness in Norway spruce (Picea abies L. Karst). J. Oxf. Univ. For. Soc. (Ser. 4) No"^ 67 35-43. 66. Kellogg, ,R.M. 1970. Personal communication. 67. Kerr, T. and I.W. Bailey. 1934. The cambium and its derivative tissues. X. Structure, optical proper ties and chemical composition of the so-called middle lamella. J. Arnold Arboretum 15: 327-349. 68. Kennedy, R.W. 1961. Variation and periodicity of summer-wood in some second-growth Douglas-fir. Tappi 44: 161-166. 69. . and J.M. Jaworsky. 1960. Variation in cellulose content of Douglas-fir (Pseudotsuga  menziesii). Tappi 43: 25-27. 70. and W.G. Warren. 1969. Within tree varia tion in physical and chemical properties of Douglas-fir. Paper presented at Second World Consul, on Forest Tree Breeding, Washington, 7-16 August, 1969. 71. Klem, G.S. 1968. Quality of wood from fertilized trees. Tappi 51: 99A-103A. 72. Kollman, F. and W.A. Cote, Jr. 1968. Principles of Wood Science and Technology. 1. Solid Wood. Springer-Verlag, Berlin, New York. 73. Kozak, A. and J.H.G. Smith. 1965. A comprehensive and flexible multiple regression program for electronic computing. For. Chron. 41: 438-443. 74. Lange, P.W. 1958. The distribution of chemical consti tuents throughout the cell wall. In Fundamentals of Paper Making Fibres (ed. Bolam). B.P. & B.M.A., London, pp. 147-185. 75. and A. Kjaer. 1957. Quantitative chemical analysis of the.different parts of the woody cell wall in wood and cellulose fiber with the inter ference microscope. Norsk.,Skogind. 11: 425-432. 97. 76. Larson, P.R. 1962. A biological approach to wood quality. Tappi 45: 443-448. 77. ._. 1963. Microscopic wood characteristics and their variations with tree growth. Section 41, IUFRO Meeting, Madison, Wisconsin. 78. . 1966. Changes in chemical composition of wood cell walls associated with age...in Pinus  resinosa. For. Prod. J. 16(4): 37-45. 79. •. 1968. Assessing wood quality of fertilized coniferous trees. In Forest Fertilization Theory and Practice. Published by the Tennessee Valley Authority National Fertilizer Development Centre, Alabama, pp.275-280. 80. . 1969. Wood formation and the concept of wood.quality. Yale Univ., School of For. Bull. No. 74. 81. Lee, Y. 1968. A review of research literature on forest fertilization. Dept. of Fisheries & Forestry, For. Res. Lab., Can. For. Svce., Victoria. Information Rept. BC-X-18. 82. Lee, H.N. and E.M. Smith. 1916. Douglas fir fiber, with special reference to length. For.Quart. 14: 671-695. 83. Leopold, B. 1961. Chemical composition and physical properties of wood fibers. 1. Preparation of holo cellulose fibers from loblolly pinewood. Tappi 44: 230-232. 84. Leslie, A.J.- 1956. Cellulose forestry. Proc. APPITA. 10: 12-29. 85. Liese, W. 1965. The warty layer. In Cellular Ultra-structure of Woody Plants (ed. Cc^te). Syracuse Univ. Press, Syracuse. pp.251-269. 86. MacMillan, W.B. 1925. A study of comparative lengths of tracheids of red spruce grown under free and suppres sed conditions. J. For. 23: 34-42. 87. McMillan, C.W. 1968. Chemical composition of loblolly pinewood as related to specific gravity, growth rate and distance from pith. Wood Sci. Tech. 2: 233-240. 98. 88. Malm, E. 1962. "Studies av barrved Godslad ved." SCA Forksningslaboratoriet Rept. No. 68. (Original not seen. From Klem, G.S., Ref. No. 71). 89. Mark, R. 1967. Cell Wall Mechanics of Tracheids. Yale Univ. Press, New Haven and London. 90. Meier, H. 1964. General chemistry of cell walls and distribution of the chemical constituents across the walls. In The Formation of Wood in Forest Trees (ed. Zimmermann). Academic Press, New York, pp. 137-151. 91. and K.C.B. Wilkie. 1957. The distribution of polysaccharides in the cell wall of tracheids of pine (Pinus syIvestris L.). Holzforsch. 13: 177-182. 92. Merrifield, R.G. and R.R. Foil. 1967. The effects of fertilization on growth and nutrient concentration in young loblolly pine. Louisiana State Univ., Agri. Exp. Stn. Bull. No. 622. 93. Mitchell, K.J. and R.M. Kellogg. . 1970. Response of a 50-year-old Douglas-fir stand to Urea fertilization. Dept. of Fisheries & Forestry, For. Res. Lab., Can. For. Svce., Victoria. Information Rept. BC-X-45. 94. McKimmy, M.D. 1965. A variation and heritability study of wood specific gravity in 46-year-old Douglas-fir from known seed sources. Proc. Third For. Biol. Conf. of TAPPI, Madison, Wisconsin. Sec. 3(1): 1-7. 95. Mohr, C. and F. Roth. 1897. The timber pines of the southern United States - a discussion of the structure of their wood. USDA For. Bull. 13 (Revised). 96. Murphey, W.K., Brady, S.L. and M.J. Behan. 1969. Anatomy and physical properties of two-year western larch seedlings grown in mineral deficient solutions. For. Prod. J. 19(12): 28-31. 97. Necesany, V. 1969. The effect of some growth substances on the cell wall structure. Bull. Int. Ass. Wood Anatomists 1: 1-8. 98. Nicholls, J.W.P. and H.E.: Dadswell. 1962. Tracheid length in Pinus radiata D. Don. C.S.I.R.O., Div. of For. Prod., Melbourne, Australia. Tech. Paper No. 24. 99. 99. Nicholls, J.W.P. and H.E. Dadswell. 1965. Assessment of wood qualities for tree breeding. C.S.I.R.O., Div. of For. Prod., Melbourne, Australia. Tech. Paper No. 37. 100. ^ , Dadswell, H.E. and J.M. Fielding. 1964. The heritability of wood characteristics of Pinus  radiata. Silvae Genet. 13: 68-71. 101. Panshin, A.J. and C. de Zeeuw. 1970. Textbook of Wood Technology, Vol. 1. (Third ed.) McGraw-Hill, New York. 102. Penhallow, D.P. 1907. A Manual of the North American Gymnosperms. Ginn & Co., Boston. 103. Perila, 0. 1961. The chemical composition of carbohydrates of wood cells. J. Polymer Sci. 51: 19^26. 104. and P. Heitto. 1959. The chemical composition of wood cells. 1. Carbohydrates of pine cells. Swomen Kemistilehti, Helsinki. 32B: 76-80. 105. and T. Seppa. 1960. The chemical composition of wood cells. III. Carbohydrates of spruce cells. Swomen Kemistilehti, Helsinki. 32B: 114-116. 106. Phillips, E.W.J. 1962. The anatomy of soft woods and its influence on timber quality. Suppl. to Forestry 35: 15-21. 107. Pillow, M.Y. and M.W. Bray. 1935. Properties and sulphate pulping characteristics of compression wood. Paper Trade J. 101: 31-34. 108. Posey. CE. 1964. The effects of fertilization on wood properties of loblolly pine (Pinus taeda L.) . N.C State College, School of For., Tech. Rept. No. 22. 109. Preston, R.D. 1947. The fine structure of the wall of conifer tracheid. II. Optical properties of dissected cells in Pinus insignis. Proc. Roy. Soc. London. Series B 134: 202-218. 110. •/ • t 1965. Interdisciplinary approaches to wood structure. Irt Cellular Ultrastructure of Woody Plants (ed. C6te). Syracuse Univ. Press, Syracuse. pp.1-31. 111. Rendle, B.J. 1960. Juvenile and adult wood. J. Inst. Wood Sci. 5: 58-61. 100. 112. Richardson, S.D. 1964. The external environment and tracheid size in conifers. In The Formation of Wood in Forest Trees (ed. Zimmermann). Academic Press, New York. pp.367-388. 113. , • , . . _ an<3 J.M. Dinwoodie. 1960. Studies on the physiology of xylem development. I. The effect of night temperature on tracheid size and wood density in conifers. J. Inst. Wood Sci. 6: 3-13. 114. Ritter, G.J. and L.C. Fleck. 1926. Chemistry of wood. IX. Springwood and summerwood. Ind. Eng. Chem. 18: 608-609. 115. Roelofsen, P.A. 1958. Cell-wall structure as related to surface growth: Some supplementary remarks on multinet growth. Acta Botanika. Neerlandica 7: 77-89. 116. Rydholm, S.A. 1965. Pulping Processes. Inter. Sci. Pub., New York. 117. Sastry, C.B.R. 1967. Some effects of fertilizer appli cations on wood properties of Douglas-fir (Pseudot-suga menziesii (Mirb.) Franco). Univ. British Columbia, Fac. For. M.Sc. Thesis (Unpubl.). 118. and R.W. Wellwood. 1971. A sensitive quartz ultra-microbalance for weighing single trac heids. Wood Sci. 3: 179-182. 119. . 1970. Unpublished data. 120. Schultz-Dewitz, G. 1958. Einfluss der soziologischen Stellung auf die Jahrringbau sowie aud die Holzele-mente bei Kiefern, Spirken und Larchen eines Natur-waldes. (Influence of crown class on the annual ring sturcture and on the wood elements in Pinus  sylvestris, P. muflgo, and Larix decidua in a natural forest). Holz-zbl. 84: 849-851. 121. Skene, D.S. 1966. A sensitive microbalance for weighing single fibres. C.S.I.R.O., Div. of For. Prod., Melbourne, Australia. Newsletter No. 336, December, pp.6-7. 121a. ,.- • --. . 1969. The period of time taken by cambial derivatives to grow and differentiate into tracheids in Pinus radiata. Ann. Bot. 33: 253-262. 101. 122. Spurr, S.H. and M.J. Hyvarinen. 1954. Wood fiber length as related to position in tree and growth. Bot. Rev. 29: 561-575. 123. Squire, G.B. 1967. Examination of cellulose-lignin relationships within coniferous growth zones. Univ. British Columbia, Fac. For. Ph.D. Thesis. (Unpubl.). 124. Stamm, A.J. and H.T. Sanders. 1966. Specific gravity of the wood substance of loblolly pine as affected by chemical composition. Tappi 49: 397-400. 125. Stein, C. 1945. A two-sample test for a linear hypothesis whose power is independent of the variance. Ann. Math. Stat. 16: 243-258. 126. Stewart, CM. 1954. An outline of the composition and nomenclature of wood polysaccharides. Holzforsch. 8: 46-48. 127. . 1966. The chemistry of secondary growth in trees. C.S.I.R.O., Div. of For. Prod.,-Melbourne, Australia. Tech. Paper No. 43. 128. TAPPI Standards and Suggested Methods. 1954. Holocellu lose in wood. T9 m-54. Tech. Assoc. Pulp and Paper Ind., New York. 129. . . 1955. Cellulose in wood. T17 m-55. Tech. Assoc. Pulp and Paper Ind., New York. 130. . . 1960. Kappa number of pulp. T23 6 m-60. Tech. Assoc. Pulp and Paper Ind., New York. 131. - 1961. Alpha-, beta-, and gamma-cellulose in pulp. T203 os-61. Tech. Assoc. Pulp and Paper Ind., New York. 132. ' . 1967. Coarseness of pulp fibers. T234 su-67. Tech. Assoc. Pulp and Paper Ind., New York.-133. Thompson, N.S. and H.H. Heller. 1966. Isolation and the carbohydrate composition of epithelial cells of-longleaf -pine. Tappi 49: 401-405. 134. Thornber, J.P. and D.H. Northcote. 1961. Changes in the chemical composition of a cambial cell during its differentiation into xylem and phloem tissue in trees. I. Main components. Biochem. J. 81: 449-455. 102. 135. Timell, T.E. 1965. Wood and bark polysaccharides. In Cellular Ultrastructure of Woody Plants (ed. C6te). Syracuse Univ. Press, Syracuse, pp.127-154. 136. . 1967. Recent progress in the chemistry of wood hemicelluloses. Wood Sci. Tech. 1: 45-70. 137. Tsoumis, G. 1968. Wood as Raw Material. Pergamon Press, New York. 138. van Buijtenen, J.P. 1964. Anatomical factors influencing wood specific gravity of slash pines and the impli cations for the development of a high-quality pulpwood. Tappi 47:, 401-404. . 139. van den Driessche, -R. and P.F. Wareing. 1966. Dry-matter production and photosynthesis in pine seedlings. Ann. Bot. 30: 673-682. 140. Vorreiter, L. 1965. Regular correlations between the dimensions of the structural elements of wood fibres. Das Papier 19: 557-570. (CS .1 .R.O. , Australia Transln. No. 7928). 141. Wardrop, A.B. 1951. Cell wall organization and the properties of xylem. C.S.I.R.O., Div. of For. Prod. Melbourne, Australia. Reprint 146, Reprinted from Australian J. Sci. Res. Series B. Bio. Sci. 4: 391-414. 142. . 1962. Fundamental studies in wood and fibre structure relating to pulping processes. Appita 16: XV-XXX. 143. . . 1964. The structure and formation of the cell wall in xylem. In^The Formation of Wood in Forest Trees (ed. Zimmermann). Academic Press, New York. pp.87-134. • 144. . 1965. Cellular differentiation in xylem. In Cellular Ultrastructure of Woody Plants (ed. Co~t£) . Syracuse Univ. Press, Syracuse, pp.61-97. 145. ' . 1969. Fiber morphology and paper making. Tappi 52: 396-408. 146. and H.E. Dadswell. 1952. The nature of reaction wood. III. Cell division and cell-wall formation in conifer stems. Australian J. Sci. Res. Series B. Bio. Sci. 5: 385-398. 103. 147. Wardrop, A.B. and H.E. Dadswell. 1953. The development of the conifer tracheid. Holzforsch. 7: 33-39. 148. . 1954. The development and structure of wood fibres. Proc. APPITA 8: 6-26. 149. Watson, A.J. 1962. A semimicro procedure for estimating the resistant carbohydrate material in wood. Tappi 45: 722-724. 150. and I.G. Hodder. 1954. Relationship between fibre structure and handsheet properties in Pinus  taeda. Proc. APPITA 8: 290-310. 151. Westing, A.H. • 1965. Formation and function of compres sion wood in gymnosperms. Bot. Rev. 31: 381-480. 152. 1968. Formation and function of compres sion wood in gymnosperms II. Bot. Rev. 34: 51-78. 153. Wellons, J.D. 1970. Cell wall polysaccharides in southern pinewood - A review. Wood Sci. 2: 247-254. 154. Wellwood, R.W. 1960. Specific gravity and tracheid length variations in second-growth western hemlock. J. For. 58: 361-368. 155. . 1962. Tensile testing of small wood samples. Pulp Paper Mag. Can. 63: T61-T67. 156. and J.W. Wilson. - 1965. The growth incre ment as a guide to properties in coniferous woods. Section 41, IUFRO Meeting, Melbourne, Australia. 157. , Ifju, G. and J.W. Wilson. 1965. Intra increment physical properties in conifers. In Cellular Ultrastructure of Woody Plants (ed. Cdt&). Syracuse Univ. Press, Syracuse, pp.539-549. 158. Wheeler, E. Y. ,• Zobel, B.J. and D.L. Weeks. 1966. Tracheid length and diameter variation iiuthe bole of loblolly pine. Tappi 49: 484-490. 159. Willis, J.G. 1966. A Dictionary of Flowering Plants and Ferns. Cambridge Univ. Press, Cambridge. 160. Wilson, J.W. 1970. Personal communication. 161. and R.W. Wellwood. 1965. Intra-increment chemical properties of certain western Canadian coniferous species. In Cellular Ultrastructure of Woody Plants (ed. C6t£).. Syracuse Univ. Press, Syracuse, pp.551-559. 104. 162. Wise, L.E. 1958. What is wood cellulose - a semantic dilemma. Tappi 41: 14A-22A. 163. _____________ Murphy, M. and A.A. D'Addieco. 1946. Chlorite holocellulose, its fractionation and bearing on summative wood analysis and on studies on the hemi-celluloses. Paper Trade J. 122(2): 35-43. 164. Wodzicki, T. 1960. Investigation on the kind of Larix polonica Rac. Wood formed under various photoperiodic conditions. -'I. Plants growing in natural conditions. Acta Soc. Bot. Polon. 29: 713-730 (Forestry Abstracts No. 4998. Vol. 22). 165. Worster, H. and B.K. Sugiyama. 1962. The carbohydrate content and composition of some western woods related to growth factors. Pulp Paper Mag. Can. 63: T395-T401. 166. Yiannos, P.N. 1964.• The apparent cell-wall density of wood and pulp fibers. Tappi 47: 468-471. 167. Ziegler, H. 1964. Storage, mobilization and distribution of reserve material in trees. In The Formation of Wood in Forest Trees (ed. Zimmermann). Academic Press, New York. pp.303-320. 168. Zimmermann, M.H. (ed.). 1964. The Formation of Wood in Forest Trees. Academic Press, New York. 169. Zobel, B.J. 1961. Inheritance of wood properties in conifers. Silvae Genet. 10: 65-70. 170. and R.R. Rhodes. 1957. Specific gravity indices for use in breeding loblolly pine. For. Sci. 3: 281-285. 171. .and R.L. McElwee. 1958. Variation of cellu lose in loblolly pine. Tappi 41: 167-170. 172. , Thorb.j.ornsen, E. and F. Henson. 1960. Geographic, site and individual tree variation in wood properties of loblolly pine. Silvae Genet. 9: 149-158. 173. , Goggans, J.F., Maki, T.E. and F. Henson. 1961. Some effects of fertilizers on wood properties of loblolly pine. Tappi 44: 186-192. , TABLE I SINGLE TRACHEID (HOLOCELLULOSE SKELETON) WEIGHTS9, AND LENGTHS FOR NINE CONIFEROUS WOODS NO. OF TRACHEID TRACHEID LENGTH TRACHEID WEIGHT mm, ic ± 1 SD g SPECIES OBSERVATIONS DESIGNATION mm f  g x IO-6 x ± 1 SD Araucaria cunninghamii 46 EW 8.12 + 0.73 3.43 + 0.75 N.S. LW 8.05 + 0.79 3.54 + 0.82 Sequoia sempervirens 40 EW 6.15 + 0.50 1.66 + 0.42 * LW 6.73 + 0.64 2. 23 + 0.47 Pinus lambertiana 35 EW 6.48 + 0.71 2.53 + 0.41 N.S. LW 6.15 + 1. 29 2.46 + 0.86 Podocarpus dacrydioides 28 EW 4.31 + 0.64 1. 40 + 0.33 * LW 4.97 + 0.57 1.72 + 0.29 Picea sitchensis 25 EW 4.57 + 0.91 1.08 + 0.42 N.S. LW 4.54 + 0.90 1.08 + 0.40 Taxus brevifolia 27 EW 2.10 + 0.32 0. 26 + 0.08 * LW 2.51 + 0.30 0.32 + 0.06 Juniperus virginiana 33 EW 2.20 + 0.47 0.19 + 0.08 * LW 2.62 + 0.36 0.26 + 0.06 Cephalotaxus wilsoniana 21 EW 1.95 + 0.17 0. 21 + 0.04 * LW 2.18 + 0.20 0. 28 + 0.05 Pseudotsuga menziesii 204 EW 4.35 + 1.65 1.07 + 0.67 ** LW 4.89 + 1.52 1.70 + 0.90 2 Pseudotsuga menziesii 48 EW 2.65 + 0.38 0.41 + 0.10 ** LW 3.03 + 0.36 0.66 + 0.18 Pseudotsuga menziesii 116 EW 2.92 + 0.56 0 . 50 + 0.22 ** 4 ... Pseudotsuga menziesii LW 3.26 + 0.56 0.79 + 0.30 53 EW 2.64 + 0.22 0.42 + 0.14 ** LW 2.71 + 0.27 0.53 + 0.17 Pseudotsuga menziesii 52 EW 2.24 + 0.32 0.29 + 0.08 ** LW 2.45 + 0.27 0.40 + 0.11 Pseudotsuga menziesii 69 EW 2.57 + 0.24 0.32 + 0.11 ** LW 2.71 + 0.37 0. 47 + 0.13 Combined Average 797 EW 3.72 + 1.82 0.90 + 0.89 ** LW 4.18 + 1.88 1.27 + 1.06 (Continued) 106. TABLE I (Continued) a All weights expressed - in ovendry condition EW Earlywood LW Latewood x Average length or weight SD Standard deviation N.S. Earlywood and latewood not significantly different at the 5%level ('t' test) * Earlywirod and latewood significantly different at the 5%. level ( 11' test) ** Earlywood and latewood significantly different at the 1% level ('t' test) Groups within Douglas-fir 1. Butt log of a 500-year-old tree (Increments 10, 20, 80, 150 300, 400)and Increment 20 from group 3 below. 2. Compression wood - Breast height disk of a 120-year-old tree, 3. NH4NO3 fertilized - Breast height disk of a 29-year-old tree (Increments 21, 22, 23, 24, 25) fertilized at ages 21 and 22. 4. Urea fertilized - 12 mm increment core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 5. NPK fertilized -.12 mm increment core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 6. (NH4)2SO4 fertilized - 12 mm increment core from breast height, 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. TABLE II SINGLE TRACHEID (ALPHA-CELLULOSE SKELETON) WEIGHTS3 AND LENGTHS FOR NINE CONIFEROUS WOODS NO. OF TRACHEID TRACHEID LENGTH TRACHEID_WEIGHT OBSERVATIONS DESIGNATION mm, x ± 1 SD q x 10~6 x ± 1 SD Araucaria cunninghamii 33 EW 7.38 + 0.78 3.20 + 0. 53 N.S. LW 7.35 + 0.80 3.07 + 0.73 Sequoia sempervirens 32 EW 5.36 + 0.81 1.58 + 0.48 ** LW 5.81 + 0.60 1.81 + 0. 42 Pinus lambertiana 35 EW 5.37 + 0 . 62 1.92 + 0.37 ** LW 5.81 + 0.39 2.30 + 0.19 Podocarpus dacrydioides 43 EW 4.17 + 0.48 1.19 + 0. 24 N.S . LW 4.07 + 0.45 1.14 + 0. 29 Picea sitchensis 46 EW 3.98 + 0.51 0.89 + 0. 24 N.S . LW 3.95 0.46 0.88 + 0.19 Taxus brevifolia 27 EW 1.93 + 0.39 0.18 + 0.06 N.S. LW 2.13 + 0.29 0. 23 + 0 .06 Juniperus virginiana 26 EW 1.87 + 0.24 0.19 + 0.05 * LW 2.12 + 0.27 0. 23 + 0.04 Cephalotaxus wilsoniana 21 EW 1.41 + 0.32 0.14 + 0.03 N.S . LW 1.64 + 0.27 0.16 + 0.04 Pseudotsuga menziesii 199 EW 3.33 + 1.21 0.95 + 0. 63 ** LW 3.88 + 1.27 1.26 + 0.74 Pseudotsuga menziesii 43 EW 2.31 + 0. 54 0.40 + 0.17 * LW 2.62 + 0.46 0.52 + 0.17 Pseudotsuga menziesii ; 112 EW 2.16 + 0.37 0.33 + 0.13 ** LW 2.70 + 0.54 0.62 + 0. 24 4 Pseudotsuga menziesii 42 EW 2.08 + 0.17 0.31 + 0.06 ** LW 2.39 + 0.23 0.43 + 0.09 Pseudotsuga menziesii 42 EW 1.99 + 0.30 0.29 . + 0 .07 * LW 2.18 + 0 .25 0.35 + 0.08 Pseudotsuga menziesii 46 EW 1.97 + 0.31 0.24 + 0.09 ** • LW 2.33 + 0.46 0.37 + 0.12 Combined Average 745 EW 3.12 + 1.55 0.79 + 0.77 ** LW 3.56 + 1.57 1.01 + 0. 84 (Continued) 108. TABLE II (Continued) a All weights expressed in ovendry condition EW Earlywood LW Latewood x Average length or weight SD Standard deviation N.S. Earlywood and latewood not significantly different at the 5% level ("f test) * Earlywood and latewood significantly different at the 5%level ('t' test) ** Earlywood and latewood significantly different at the 1% level ('t' test) Groups within Douglas-fir 1. Butt log of a 500-year-old tree (Increments 10, 20, 80, 150 300, 400)and Increment 20 from group 3 below. 2. Compression wood - Breast height disk of a 120-year-old tree, 3. NH4NO3 fertilized - Breast height disk of a 29-year-old tree (Increments 21, 22, 23, 24, 25) fertilized at ages 21 and 22. 4. Urea fertilized - 12 mm increments core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 5. NPK fertilized - 12 mm increment core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 6. (NH4)2SC>4 fertilized - 12 mm increment core from breast height, 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. TABLE III SINGLE TRACHEID (HOLOCELLULOSE SKELETON) WEIGHTSa AND LENGTHS FOR TWO DOUGLAS-FIR TREES INCREMENT NO. NO. OF TRACHEID TRACHEID LENGTH TRACHEID_WEIGHT TREE NO. FROM PITH OBSERVATIONS DESIGNATION mm, x ± 1 SD g x 10~6 x ± 1 SD 10 27 EW 2.17 + 0.30 0.24 + 0.07 LW 2.41 + 0.38 0.39 + 0.13 20 26 EW 3.58 + 0.70 0.85 + 0.35 LW 3.98 + 0.66 1.26 + 0.44 80 23 EW 5.61 + 0.70 1.65 + 0.52 LW 5.85 + 0.52 2.25 + 0.51 150 39 EW 5.97 + 0.60 1.77 + 0.47 LW 6.26 + 0.48 2.62 + 0.68 300 26 EW 5.79 + 0.69 1.63 + 0.45 LW 5.46 + 0.95 1.90 + 0.56 400 27 EW 5.85 + 0.61 1.42 + 0.28 LW 6.00 + 0.71 1.87 + 0.49 20 36 EW 2.92 + 0.59 0.56 + 0.21 LW 3.20 + 0.48 0.84 + 0.33 21b 23 EW 2.92 + 0.57 0. 53 + 0.16 22b LW 3.40 + 0.33 0.70 + 0.32 24 EW 2.93 + 0.45 0. 54 + 0.18 LW 3.22 + 0.54 0.74 + 0. 21 23 28 EW ' 2.98 + 0.48 0.44 + 0.12 LW 3.34 + 0.72 0. 85 + 0.36 24 15 EW 3.34 + 0.23 0.71 + 0.28 LW 3.58 + 0.27 1.02 + 0.32 25° 26 EW 2.67 + 0.73 0.37 + 0. 21 LW 2.86 + 0.52 0.62 + 0. 20 a b c Tree Tree All weights expressed in ovendry condition Years fertilized, Compression wood x Average length or weight SD Standard deviation EW Earlywood LW Latewood• 1 - Butt log of a 500-year-old tree; 2 - NH4N03 fertilized - Breast height disk of a 29-year-old tree. TABLE IV SINGLE TRACHEID (ALPHA-CELLULOSE SKELETON) WEIGHTS3 AND LENGTHS FOR TWO DOUGLAS-FIR TREES TREE -NO. INCREMENT NO. FROM PITH NO. OF OBSERVATIONS TRACHEID DESIGNATION TRACHEID LENGTH mm, x ± .1 SD TRACHEID_WEIGHT g x IO-6 x ± 1 SD 10 28 EW 1.85 + 0.37 0.22 0.09 LW 2.18 + 0.33 0.31 + 0.12 20 34 EW 2.94 + 0.54 0.73 + 0.27 LW 3.46 + 0.62 0.96 + 0.32 80 24 EW 4.22 + 1.04 1.3 8 + 0.56 LW 4. 87 + 0.74 1.86 + 0.47 150 27 EW 4.26 + 1.22 1.50 + 0.70 LW 5.07 + 1.01 1.97 + 0.60 300 25 EW 4.-01 + 0.73 1. 24 + 0.37 LW 4.76 + 0.86 1.79 + 0.51 400 26 EW 4.48 + 0.70 1.33 + 0.44 LW 4 .54 + 0.94 1. 63 + 0.58 20 35 EW 2.39 + 0. 42 0.45 + 0.13 LW 2.80 + 0.42 0. 66 + 0. 21 21b 22 EW 2.38 + 0.44 0.36 + 0.15 LW 2.79 + 0.47 0.64 + 0.22 22b 20 EW 2.35 + 0.26 0.32 + 0.08 LW 2.66 + 0.28 0. 60 + 0.17 23 26 EW 2.03 + 0.19 0.31 + 0.07 LW 2.60 + 0.71 0.56 + 0.28 24 22 EW 2.12 + 0.33 0.38 . + 0.15 LW 2.98 + 0. 57 0.80 + 0. 27 25C 22 EW 1.90 + 0.40 0. 27 + 0.16 LW •2.54 + 0.50 0.51 + 0.18 a All weights expressed in ovendry condition?: b Years fertilized c Compression wood Tree 1 - Butt log of a 500-year-old tree. Tree 2 - NH4N03 fertilized - Breast height disk of a 29-year-old tree. x SD EW LW Average length or weight Standard deviation Earlywood Latewood TABLE V ANALYSIS MODEL (A) AND (B) RESULTS OF MULTIPLE CURVILINEAR COVARIANCE ANALYSES; DIFFERENCES AMONG SPECIES (HOLOCELLULOSE SKELETONS) (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH) A) Analysis Model . Group DF F Total Differences for testing slopes Sums Differences for testing levels Combined Regression X B) 1 vs 2 vs 3 1 vs 2 vs 3 6 vs 7 vs 8 1 vs 2 vs 4 3 vs 5 vs 9 1, 2, 4, vs 6, 7, 8 vs ! 1 vs 2 3 4 5 7 8 vs 4 vs vs 5 Group 5 vs 9 1 1 1 6 6 6, 8 vs 4 vs 9 3 vs 1 vs vs vs vs vs vs Significance in Slopes Levels 5 9 ** ** ** ** N.S , * N.S , ** ** * * ** * N.S, ** N.S , N.S , ** ** ** ** * ** ** ** * ** N.S. * ** N.S . ** ** N.S . ** Groups 1 Araucaria cunninghamii 2 Sequoia sempervirens 3 Pinus lambertiana . 4 Podocarpus dacrydioides 5 Picea sitchensis 6 Taxus brevifolia 7 Juniperus virginiana 8 Cephalotaxus wilsoniana 9 Pseudotsuga menziesii X -N.S. * ** Indicates 'Significance' or 'Non-significance' in slopes, Indicates 'Significance' or -'Non-significance1 in levels Not significant at the 5% lev Significant at the 5% level Significant at the 1% level 112. TABLE VI ANALYSIS MODEL (A) AND (B) RESULTS OF MULTIPLE CURVILINEAR COVARIANCE ANALYSES; DIFFERENCES AMONG SPECIES (ALPHA-CELLULOSE SKELETONS) (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH) A) Analysis Model  Group DF F Total Differences for testing slopes - X Sums Differences for testing levels - Y Combined RegressionB) Group Significance in Slopes Levels 1 vs 2 vs 3 vs 4 vs 5 vs 9 ** ** 1 vs 2 vs 3 vs 5 N.S . ** 6 vs 7 vs 8 vs 9 ** ** 1 vs 2 vs 4 N.S. * 3 vs 5 vs 9 * ** 1, 2, 4 vs 9 ** ** 6, 7, 8 vs 9 ** * * 1 vs 2 N.S . N.S . 1 vs 3 N.S . ** 1 vs 4 N.S . N.S. 1 vs 5 N.S. ** 6 vs 7 * * 6 vs 8 N.S . ** 6, 8 vs 7 N.S . N.S. • 4 vs 9 ** ** 3 vs 5 N.S . ** 1 vs 9 ** ** Groups 1 Araucaria cunninghamii 2 Sequoia sempervirens 3 Pinus lambertiana 4 Podocarpus dacrydioides 5 Picea sitchensis 6 Taxus brevifolia 7 Juniperus virginiana 8 Cephalotaxus wilsoniana 9 Pseudotsuga menziesii X - Indicates 'Significance' or 'Non-significance' in slopes Y - Indicates 'Significance' or 'Non-significance' in levels N.S. Not significant at the 5% level * Significant at the 5% level ** Significant at the 1% level 113. TABLE VII MULTIPLE CURVILINEAR COVARIANCE ANALYSIS FOR TRACHEID WEIGHTS (HOLOCELLULOSE SKELETONS) IN JUVENILE, MATURE AND OVERMATURE WOOD OF A 500-YEAR-OLD DOUGLAS-FIR TREE (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH) A Group DF F Juvenile wood (Increment"10) 24 Mature wood (Increments 20, 80, 150) 85 Total: 109 Differences for testing slopes 2 1 .190 N. S. Sums: 111 Difference for testing levels 1 2 .010 N. S. Combined Regression 112 B Group DF F Juvenile wood (Increment 10) 24 Overmature wood (Increments 300, 400) 50 Total: 74 Differences for testing slopes 2 0 .722 N. s. Sums: 76 Differences for testing levels 1 0 .355 N. s. Combined Regression 77 C Group DF F Mature wood (Increments 20, 80, 150) 85 Overmature wood (Increments 300, 400) 50 Total: 135 Differences for testing slopes 2 7 .401** Sums: 137 Differences for testing levels 1 29 .752** Combined Regression 138 ——— —————————— — ——— — —————————— — ———— — — N.S. Not significant at the 5% level ** Significant at the 1% level 114. TABLE VIII MULTIPLE CURVILINEAR COVARIANCE ANALYSIS FOR TRACHEID WEIGHTS (ALPHA-CELLULOSE SKELETONS) IN JUVENILE, MATURE AND OVERMATURE WOOD OF A 500-YEAR-OLD DOUGLAS-FIR TREE (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH) A Group Juvenile wood (Increment 10) Mature wood (Increments 20, 80, 150) Total: DF 25 82 .107 F Difference for testing slopes Sums: 2 109 2 .575 N. S. Difference for testing levels 1 7 .787 ** Combined Regression 110 B Group Juvenile wood (Increment 10) Overmature wood (Increments 3 00, 400) Total: DF 25 48 73 F Difference for testing slopes Sums : 2 75 - 1 .413 N. S . Difference for testing levels 1 15 .069 ** Combined Regression 76 C Group Mature wood (Increments 20, 80, Overmature wood (Increments 3 00, 150) 400) Total: DF 82 48 130 F Difference for testing slopes Sums : 2 132 3 .898 * Difference for testing levels 1 0 .034 N. S . Combined Regression 133 N.S. Not significant at the 5% level * Significant at the 5% level ** Significant at the 1% level 115. TABLE IX ANALYSIS OF COVARIANCE AND ADJUSTED MEAN VALUES OF EARLYWOOD AND LATEWOOD TRACHEID WEIGHTSa IN SIX DOUGLAS-FIR TREES (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH). HOLOCELLULOSE ALPHA-CELLULOSE SKELETONS SKELETONS Source DF F DF F Trees 5 1.23 N.S. 5 2.45* Earlywood/Latewood 1 97.92 ** 1 0.00 N.S, Error 534 475 Adjusted Mean Tracheid Skeleton Weights Earlywood 0.7 5 0.70 Latewood 0.94 0.7 0 — 6 a All weights expressed as g x 10 in ovendry condition N.S. Not significant at the 5% level * Significant at the 5% level ** Significant at the 1% level 116. TABLE X ANALYSIS OF COVARIANCE AND ADJUSTED MEAN VALUES OF EARLYWOOD AND LATEWOOD TRACHEID WEIGHTSa IN NINEb CONIFEROUS SPECIES (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH) HOLOCELLULOSE SKELETONS ALPHA-CELLULOSE SKELETONS Source DF Groups 13 20.91 ** Earlywood/Latewood 1 65.43 ** Error 781 DF 13 1 729 64.66 ** 1.88 N.S Adjusted Mean Tracheid Skeleton Weights Earlywood Latewood 1.02 1.17 0.90 0.91 a All weights expressed as g x 10 in ovendry condition b,c Includes all species and the six trees within Douglas-fir; N.S. Not significant at the 5% level ** Significant at the 1% level TABLE XI SELECTED WOOD AND TRACHEID PROPERTIES OF THREE FERTILIZED DOUGLAS-FIR TREES Characteristics Tree 84 Tree 90 Tree 74 Treatment Composition Urea NPK (NH4)2SO 4 Increment from pith 12 15* 16 12 15* 16 12 15* 16 Specific gravityl 0.414 0.430 0.436 0.403 0.374 0.352 0.381 0.371 0. 337 (3-year age segment) Percentage of latewood 28.6 23.1 20.0 24.7 18.6 18.6 24. 6 14.3 10. 7 Increment width, mm 4.2 6.5 5.5 5.1 10.5 10.7 5.7 7.7 7. 5 Tracheid length, mm 2.26 3.30 2.80 2.10 2.36 2.55 2.20 2. 26 2. 40 Tangential tracheid 27.0 28.0 27.5 29 . 8 29.5 29 .4 29.2 27. 8 26. 2 diameter, micron Radial tracheid 34.5 43.1 40.0 33.0 38.0 40.5 38.0 40.5 46. 0 diameter, micron Tangential double 7.6 7.5 7.0 7.5 7.5 7.5 7.0 7.5 8. 0 ,-;.wall thickness , micron Radial double wall 9.0 10.0 10.5 11.0 12.0 12.5 11.5 10. 2 10. 5 thickness,micron * First fertilized at the age of 13 and refertilized at age 15 at the rate of 200 lb Cpounds per acre of N, P2^§' or K2^ e<3uiva^-ents) 1. Maximum moisture content method. 118. TABLE XII RESULTS OF MULTIPLE CURVILINEAR COVARIANCE ANALYSES; DIFFERENCES WITHIN SPECIES IN TREATED AND UNTREATED DOUGLAS-FIR (ADJUSTED FOR DIFFERENCES IN TRACHEID LENGTH) A. (HOLOCELLULOSE (ALPHA-CELLULOSE SKELETONS) SKELETONS) Groupa DF F DF F 1 201 196 3 113 109 4 50 39 5 49 39 6 66 42 Total 479 425 Differences for testing slopes 8 1.915 N .S. 8 5 .650 ** Sums 487 433 Differences for testing levels 4 1.115 N .S . 4 7 .405 ** Combined Regression 491 437 B. Significance in Significance in Group Slopes Levels Slopes Levels 4 vs 5 vs 6 N.S. ** * * ** 4 vs 6 N.S. * * ** ** 4 vs 5 N.S . N.S . * * N.S . 3 vs 4 N.S . * N.S . N.S . 4 vs 7 N.S . N.S . N.S . ** 5 vs 7 N.S . N.S . N.S. ** 6 vs 7 N.S . N.S . N.S. N.S . 3 vs 7 ** N.S . ** N.S . N.S. Not significant . at the 5% level * Significant at the 5% level ** Significant at the 1% level a Group 2 - Compression wood of Douglas-fir is under separate heading (Continued) 119. TABLE XII (Continued) Groups within Douglas-fir 1. Butt log of a 500-year-old tree (Increments 10, 20, 80, 150, 300, 400) and Increment 20 from group 3 below. ? 3. NH4NO3 fertilized -Breast height disk of a 29-year-old tree (Increments,.'21, 22, 23, 24, 25) fertilized at ages 21 and 22. 4. Urea fertilized - 12 mm increment core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 5. NPK fertilized - 12 mm increment core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 6. (NH4)2SC>4 fertilized - 12 mm increment core from breast height; 24-year-old tree (Increments 15, 16) fertilized at ages 13 and 15. 7. Juvenile wood - Increment 10 from Group 1 above. 120. FOREWORD TO FIGURES Figures 1 to 10, 13, 15 to 18--Plotting and curve fitting by computer. Mathematical model for curve fitting: Y = bQ + b± X + b2 X2 _ g Where Y = Ovendry weight of tracheid in g x 10 X = Tracheid length in mm - green condition Figure 11—Plotting and curve fitting by computer. The line of best fit conditioned to pass through the origin, In all figures, the mathematical models given are the equations of best fit. Figure 1: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Araucaria cunninghamii. H I I I 1 1 1 1 1 1 1 3.0 0.98 1.96 2.94 3.92 4.90 5.98 6.86 7.B4 8.92 9 BC Tracheid Length, mm' Figure 2: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Sequoia sempervirens. H0LO-SP2 122, A. -2.2633 + 0.6580K r2 = 0,664, SE = 0.3OS; DF = 38 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x IO"6 X = TRACHEID LENGTH IN ffl - GREEN CONDITION l'.96 —I 2.94 -1 3.92 . 4.9© 5.88 6.86 ALPHA SP2 B, -0.1590 + O.OSStX2 •-2 = 0.933; SE = 0.119; DF = 30 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN HI - GREEN CONDITION —1 3.98 -r 2.94 3.92 4.90 J.flfl Tracheid Length,mm-Figure 3: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Pinus lambertiana. 123. A. 88 H0L0-SP3 Y = -1.1278 + 0.6222X r.2 = 0.933; SE = 0.183; DF = 33 WHERE Y = OVENDRY WEIGHT CF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN W - GREEN CONDITION —I 0.90 —I 2.54 —1 3.92 —1 4.90 —1 E.B6 -1 T.B4 ~1 8.82 RLPHfi SP3 B. -1.0880 + 0.5722X r2 = 0.860; SE  = 0.130; DF = 33 WHERE Y = OVENDRY WEIGHT CF TRACHEID IN G. x IO"6 X = TRACHEID LENGTH IN Wi - GREEN CONDITION -r-— 2.94 3.92 4.9<? Tracheid Length, mm-~T ?.8e Figure 4: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Podocarpus dacrydioides. H0L0-SP4 Y = 0 .0119 + 0.0t97X2 r'2 = 0.929.- SE = 0.109; DF = 26 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. X Iff* X = TRACHEID LENGTH IN HI - GREEN CONDITION 124 -1— 0.96 I 2.94 I 3.92 -1 5.88 -1 6.86 —1 7.64 —1 8.82 (£> 0-.0 RLPHfi SP4 B. Y = 0.0101 + 0.0527X2 r2 = 0.900; SE = 0.068; DF = 11 WHERE: Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN MM - GREEN CONDITION I 1.96 T —I 2.94 9.92 4.90 Tracheid Length, mm--r -1— 7.84 Figure 5: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Picea. sitchensis. H0L0-SP5 Y = -0.7389 + 0.H939X r2 = 0.948 SEE = 0.080; DF = 23 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN IT! - GREEN CONDITION —i— 0.9fl -1 1.96 —1 2.94 -1 3.92 —1 r 4.90 5.6 -1 6.86 ? 6.82 ALPHA SP5 Y = -0.0056 + 0, 1-2 = 0.920; SEE = 0.076; DF = W WHERE Y = OVENDRY WEIGHT CF TRACHEID IH G, x 10-6 X = TRACHEID LENGTH IN m - GREEN CONDITION —I— 2.94 -i r 4.9C 5.B 3.32 Tracheid Length, mm Figure 6: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Taxus brevifolia. 40L0-SP6 Y = -0.15)1 + 0.1917X r2 = 0.833; S__ = 0.033; DF = 25 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN ffl - GREEN CONDITION RLPHq SP6 Y = 0.0155 + O.COW i2 = 0.927; S_E = 0,017; DF = 25 WHEE Y = OVENDRY WEIGHT OF TRACHEID IN G, x 10-6 X = TRACHEID LENGTH IN Ml - GREEN CONDITION —I— 2.94 -1 1 3.92 4.9« Tracheid Length,mm-—I 6.96 Figure 7: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Juniperus virginiana. 127. HGL0-SP7 Y = 0,0110 + 0.0347X2 r2 = 0.894; SE  = 0.025; DF = 31 WHERE Y = OVENDRY WEIGHT OF TRACHEID ] G. x IO"6 X = TRACHEID LENGTH IN PM - GREEN CONDITION -1 1.98 —I 3.92 -1 5.BB —I G.B6 -1 T.B4 ? a.82 9.Bp flLPHfi SP7 = -0.1022 + 0.1549X r2 = 0.918; SE  = 0.013; DF = 24 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN W - GREEK CONDITION —I 2.94 3.92 4.90 Tracheid Length, mm Figure 8: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Cephalotaxus wilsoniana. 128 . H0L0-SP8 A. Y = -0.2790 + 0.2526X r2 = 0.921; SEE = 0.016; DF = 19 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN m - GREEN CONDITION -1 0.98 -1 1.96 -1 2.94 -1 3.92 RLPHfi SP8 B. Y = 0,0618 + 0.035GX2 r 2 = 0.818; SEE = 0.013; DF = 19 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G, x 10-6 X = TRACHEID LENGTH IN FN - GREEN CONDITION 8 ^ I I I I I 1 1 1 1 3.98 1.66 2.94 3.92 4.90 5.88 6.96 7.84 8.82 Q.80 Tracheid Length,mm-Figure 9: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from Pseudotsuga menziesii. 129. 384 A. 320 o Condil >. 2-56 ,endi o i o 1-92 X cn a> 1-28 13 ochei H 0-64 0-00"— 000 y = 00769+00554x2 r2 =0-843 3EE =0-342 DF=202 Where y = oven dry weight of tracheid in g xlO' x = tracheid length in mm -green condition least-squares fitted line 0-98 1-96 2-94 3-92 490 5-88 Tracheid Length, mm 6-86 7-84 8-82 B. 4-48 B 3-84 o o fT 3 20 T3 O 2 56 •92 1-28 -_ a> u £ 0-64 000 y = -0-3189+0-2184 x +00437 x2 R2 =0-986 SEE =0-083 DF = I96 Where y = oven dry weight of tracheid in g xlO""6 x= tracheid length in mm -green condition least-squares fitted line 000 0-98 2-94 3-92 4-90 5-88 6-86 7-84 Tracheid Length,mm-Figure 10: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from nine coniferous species. (Least-squares fitted line). HOLO SP1-9 130 Y = -0.1463 + 0.099OX + 0.0436X2 fi2 = 0.919; SEE = 0.319; DF =456 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G, X 10"6 X = TRACHEID LENGTH IN MM - GREEN CONDITION Tracheid Length, mm Figure 11: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for samples from nine coniferous species. (The line of best fit conditioned to pass through the origin). 131 HOLD SP1-9 - 0.03E8X + O.OWX2 R2 = 0,919, SE = 0,320; DF = 457 WHERE Y = OVENDRY WEIGHT CF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN ffl - GREEN CONDITION 1 1 r 1 1 1 r a.a 0 98 1.96 i.94 39i 4.9o 5.88 B.I —I— T.B4 ALPHA SP1-9 = 0.0758X + O.OWiX2 R2 = 0.952; SEE> 0.196; DF=459 WERE Y = OVENDRY '-.'EIGHT OF TRACHEID IN G, x 10-E X = TRA.CHEID LENGTH IN m - GREEN CONDITION -r —r —i 8.82 2.94 3.92 4.9C Tracheid Length,mm-Figure 12: Tracheid weight, mean tracheid length and specific gravity patterns across the butt log of a 500-year^old Douglas-fir tree. (Each point represents the average tracheid weight or length for the increment. Specific gravity values from Kennedy and Warren (70)) . •_£T A Figure 13: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length for compres sion wood and normal wood of Douglas-fir. HOLO COMP WOOD 133 A. MULTIPLE CURVILINEAR COVARIANCE ANALYSES CROUP DE 201 15 216 2 218 1 219 DOUGLAS-FIR (NORMAL WOOD) COMPRESSION WOOD TOTAL DIFFERENCE FOR TESTING SLOPES SUMS DIFFERENCE FOR TESTING LEVELS COMBINED REGRESSION N.S, NOT SIGNIFICANT AT THE 5? LEVEL 0.287N.S. 0.C60N.S, Y = -0.5959 + 0.1012X r 2= 0.717; SE = 0.103 DF = 16 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x 10-6 X = TRACHEID LENGTH IN MM - GREEN CONDITION ° s , least-squares fitted line ^for Douglas-fir least-squares fitted line for Douglas-fir Compression Wood —I 2.94 RLPHP, COMP WOOD MULTIPLE CURVILINEAR COVARIANCE ANALYSES B, GROUP DOUGLAS-FIR (NORMAL WOOD) COMPRESSION WOOD TOTAL DIFFERENCE FOR TESTING SLOPES SUMS DIFFERENCE FOR TESTING LEVELS COMBINED REGRESSION ** SIGNIFICANT AT THE 11 LEVEL DE 196 10 236 2 8.760" 1 7.613*' 239 Y = 0.0211 + 0.C v2 = 0.900; SE  = 0.057 DF = 11 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G, X 10-6 X = TRACHEID LENGTH IN MM - GREEN CONDITION least-squares fitted line for Douglas-fir least-squares fitted line for Douglas-fir Compression Wood —I— 1.96 -| 5.88 3.92 4.90 Tracheid Length, mm Figure 14: Tracheid weight, increment width and mean tracheid length patterns across six con secutive increments of a fertilized (NH4NO3) Douglas-fir which produced com pression wood 4.years after first fertili zation. (Each point represents the average tracheid weight or length for the increment). Tracheid Weight, g- xlO-6 (Ovendry Condition) Increment Width, mm- Mean Tracheid Length, mm-Figure 15: " Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length in Urea treated and normal Douglas-fir. 135. A. ion •o s c o-0 o >. •o > 0 O 0 g •v X Weic to •5 <u f 0 r— B, HOLQ UREA MULTIPLE CURVILINEAR COVARIANCE ANALYSES GROUP DF F DOUGLAS-FIR 201 UREA FERTILIZED DOUGLAS-FIR 50 TOTAL 251 DIFFERENCES FOR TESTING SLOPES 2 0.948 N.S. SIMS 253 DIFFERENCES FOR TESTING LEVELS 1 0.008 N.S. COMBINED REGRESSION 254 N.S. NOT SIGNIFICANT AT THE 5% LEVEL Y = -0.2423 + 0.0997X2 r2 = 0.635; SE - 0.103 DF = 51 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN e. x 10-6 X = TRACHEID LENGTH IN MM - GREEN CONDITION <r least-squares fitted line for Douglos-fir ^least-squares fitted line for Douglas-fir, Urea fertilized —I 1 1 1.96 2.94 3.92 —I 4.90 I 1 5.88 6.86 1.82 9.8c ALPHA UREA MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP DF F DOUGLAS-FIR 196 UREA FERTILIZED DOUGLAS-FIR 39 TOTAL 235 DIFFERENCES FOR TESTING SLOPES 2 0.721 N.S, SUMS 237 DIFFERENCES FOR TESTING LEVELS 1 2,517 N.S, COMBINED REGRESSION 238 N.S. NOT SIGNIFICANT AT THE 5Z LEVEL Y = 0.6502 - 0.6410X+ 0.2264X2 R.2 = 0.942; SE = 0.025 DF = 39 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G, X 10"6 X - TRACHEID LEN3TH IN Ml - GREEN CONDITION least-squares fitted line for Douglas-fir ~leost-squares fitted line for Douglas-fir.Urea fertilized T 2.94 3.92 4.90 Tracheid Length,mm-, A Figure 16: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length in NPK treated and normal Douglas-fir. HOLO NPK MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP DOUGLAS-FIR NPK FERTILIZED DOUGLAS-FIR TOTAL DIFFERENCES FOR TESTING SLOPES SUMS DIFFERENCES FOR TESTING LEVELS COMBINED REGRESSION DF F 201 19 250 2 0.106.N.S, 252 1 0.581 N.S. 253 Y = -0.0177 + 0.069W2 r2 = 0.710.- SE = 0.063 DF = 50 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. X 10-6 X = TRACHEID LENGTH IN MM - GREEN CONDITION N,S. NOT SIGNIFICANT AT THE 5? LEVEL ALPHA NPK MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP DF DOUGLAS-FIR 196 NPK FERTILIZED DOUGLAS-FIR 39 TOTAL 235 DIFFERENCES FOR TESTING SLOPES 2 SUMS 237 DIFFERENCES FOR TESTING LEVELS 1 COMBINED REGRESSION 238 " SIGNIFICANT AT THE V. LEVEL N.S. NOT SIGNIFICANT AT THE 5% LEVEL 1.753 N.S. = 0.030 Y = Q0335 + 0.C <2 = 0,861; SEE DF = 10 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. X IO"6 X = TRACHEID LENGTH IN MM - GREEN CONDITION Tracheid Length, mm-»A Figure 17: Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length in (NH4)2S04 treated and normal Douglas-fir. HOLO NS04 Y = -0.0758 + 0.0676X2 '2 = 0,687 SE = 0.079 DF = 67 WHERE Y = OVENDRY WEIGHT OF. TRACHEID IN G, x 10-6 X = TRACHEID LENGTH IN MM -GREEN CONDITION ALPHA NS04 MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP DF F DOUGLAS-FIR 201 (NHi|)2S0i( FERTILIZED DOUGLAS FIR 66 TOTAL 267 DIFFERENCES FOR TESTING SLOPES 2 0.309 N.S SUMS 269 DIFFERENCES FOR TESTING LEVELS 1 1.61)5 N.S COMBINED REGRESSION 270 N.S. NOT SIGNIFICANT AT THE 57. LEVEL MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP DF F DOUGLAS-FIR 196 (NhtyjSQi) FERTILIZED DOUSLAS-FIR 12 TOTAL 238 DIFFERENCES FOR TESTING SLOPES 2 13,631 SUMS 210 DIFFERENCES FOR TESTING LEVELS 1 21,103* COMBINED REGRESSION 211 Y = -0,2553 + 0.2599X ' <2 = 0.815,- SEE - 0.019 DF = 13 WHERE Y = OVENDRY WEIGHT OF TRACHEII IN G. X 10-6 X " TRACHEID LENGTH IN MM -GREEN CONDITION ' SIGNIFICANT AT THE 1% LEVEL Tracheid Length,mm-. A Figure 18: , Relationship between tracheid weight (holocellulose (A) and alpha-cellulose (B) skeletons) and length in NH4NO3 treated and normal Douglas-fir. HOLD HRNET 2-6 MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP jf p DOUGLAS-FIR 201 NH4NO3 FERTILIZED DOUGLAS FIR 113 TOTAL 311 DIFFERENCES FOR TESTING SLOPES 2 56,192*' SUMS 316 DIFFERENCES FOR TESTING LEVELS 1 11,581" COMBINED REGRESSION 317 ** SIGNIFICANT AT THE 1% LEVEL Y = 0.6561 - 0.5187X + 0.1625 X2 R2 = 0.781; SEE = 0.113; DF = 149 WHERE Y = OVENDRY WEIGHT OF TRACHEID IN G. x IO-6 X = TRACHEID LENGTH IN Ml - GREEN CONDITION least-squares fitted line for Dauglas-fir •least-squares fitted line for Douglas-fir, NH4 NO3 fertilized 1.96 2-94 -r 6.86 —1 8.82 flLPHR HRNEY 2-6 MULTIPLE CURVILINEAR COVARIANCE ANALYSIS GROUP DF F DOUGLAS FIR 196 NH1IO3 FERTILIZED DOUGLAS-FIR 109 TOTAL 305 DIFFERENCES FOR TESTING SLOPES 2 51,519" SUMS 307 DIFFERENCES FOR TESTING LEVELS 1 2.395 N.S. COMBINED REGRESSION 308 N.S. NOT SIGNIFICANT AT THE 57. LEVEL ** SIGNIFICANT AT THE 12 LEVEL Y = -0.0507 + 0.085D<2 T2 = 0.909; SEE = 0,072; DF = 115 WHERE Y = OVENDRY WEIGHT CF TRACHEID IN G, x 10-6 X = TRACHEID LENGTH IN Ml - GREEN CONDITION least-squares fitted line for Douglas-fir leost-squores filled line for Douglas-fir, NH4NO3 fertilized 3.92 4.90 Tracheid Length, mm-—I 6.86 APPENDICES APPENDIX I DESCRIPTION OF GROWTH INCREMENTS INCLUDED IN THE STUDY Species Age From Pith Yr. Wood Zone Increment Width mm Latewood Width mm Measured Tracheid Length ct Average nun Range -Araucaria cunninghamii — H 2.4 0.4 7.42 5.10-9.79 Sequoia serapervirens - H 2.5 0.5 5.97 • 4.20-8.50 Pinus lambertiana - H 1.4 0.3 6.29 3.75-8.12 Podocarpus dacrydioides - H 3.6 0.6 4.59 2.92-6.04 Picea sitchensis 36 H 4.4 1.4 4.70 2.92-5.84 Taxus brevifolia 70 H 2.3 0.6 2.14 1.44-2.96 Juniperus virginiana - H 3.4 0.8 2.37 1.42-2.99 Cephalotaxus wilsoniana 8 S 6.5 0.8 2.13 1.70-2.56 ^Pseudotsuga menziesii 10 H 2.0 0.9 2.37 1.88-2.71 20 H 5.8 2.4 3.97 2.14-4.55 80 H 3.4 1.4 5.46 3.96-6.67 150 H 1.7 0.7 5.92 3.33-7.08 300 H 1.0 0.3 4.96 3.13-6.88 2 400 H 0.7 0.2 5.89 3.45-7.08 Pseudotsuga menziesii 73 H 4.9 3.5 2.84 0.85-3.27 2 83 S 5.0 3.2 2.89 1.99-3.27 Pseudotsuga menziesii 20 H 5.9 2.5 2.92 1.20-3.80 21 H 7.7 3.7 2.50 1.45-4.20 22 H 10.0 3.3 2.51 1.65-3.95 23 H 8.5 2.4 2.40 1.90-4.20 24 S 8.5 2.6 2.81 1.65-3.77 25 S 8.6 3.7 2.30 1.05-3.85 Pseudotsuga menziesii 15 S 6.5 1.5 3.31 2.70-3.98 16 S 5.5 1.1 2.80 2.13-3.56 Pseudotsuga menziesii 15 H 10. 5 2.0 2.39 1.85-3.13 16 H 10.7 2.0 2.54 1.99-3.13 Pseudotsuga menziesii 15 H 7.7 1.1 2.23 1.42-2.99 16 H 7.5 0.9 2.39 1.99-3.56 (Continued) 140. APPENDIX I (Continued) Average based on a minimum of 50 tracheids Heartwood Sapwood Butt, log of a 500-year-old tree Compression wood - Breast height disk of a 120-year-old tree NH4NCU fertilized - Breast height disk of a 29-year-old treej fertilized at ages 21 and 22 (500 lb/acre) Urea fertilized - 12 mm increment core from breast height, 24-year-old tree; fertilized at ages 13 and 15 (200 lb/acre of N,P20 prK20 equi valents ) NPK fertilized - 12 mm increment core from breast height, 24-year-old tree; fertilized at ages 13 and 15 (200 lb/acre of N^-O^orK.O equi-valents) (NH^)2SO^ fertilized - 12 mm increment core from breast height, 24-year-old tree; fertilized at ages 13 and 15 (200 lb/acre of N,P205orK20 equivalents) APPENDIX II SUMMARY OF HOLOCELLULOSE AND ALPHA-CELLULOSE. YIELDS AND ESTIMATES OF LIGNIN IN SOME OF THE SAMPLES STUDIED (BASED ON OVENDRY, EXTRACTIVE-FREE WOOD) Description Increment Portion No. of Holocellu- Alpha-Cellu-Species of Species No. From in the Cycles3 lose Yield lose Yield Pith Increment % % Pseudotsuga menziesii 500-year-old 80 EW 3 7-4 .8(1. 30)b 45. 2(0. 27) LW 5 74 .1 46. 3 Pseudotsuga manziesii NH .NO, fertilized 23 EW 3 74 .3 44. 9 LW 5 73 .9 45. 8 Pseudotsuga menziesii Compression wood 73 LW 5 61 .2 38. 4 Sequoia sempervirens Mature wood - EW 3 70 .1 43. 3 LW 5 68 .9 44. 2 Picea sitchensis Mature wood 36 EW ' 3 74 .0 46. 0 . LW 5 73 .4 47. 5(0. 29) Pinus lambertiana Mature wood — LW 5 72 .2(1. 53) 46. 4 a Each cycle consists of a 30 minute treatment with peracetic acid, followed by a soaking in hot water (30 minutes) b No. in parenthesis indicates percentage of residual lignin EW Earlywood LW Latewood 142. APPENDIX III MOISTURE CONTENT OF THE HOLOCELLULOSE AND ALPHA-CELLULOSE PULPS AFTER CONDITIONING IN THE CTH ROOM FOR OVER ONE MONTH Earlywood/ Holocellulose Alpha-Cellulose Species Latewood Moisture Moisture Content Q. "5 Content O. *6 Araucaria cunninghamii EW 10.66 10.15 LW 10.87 10.38 Sequoia sempervirens EW 9.86 9.52 LW 9.79 9.62 Pinus lambertiana EW 9.80 9.83 LW 10.04 10.12 Podocarpus dacrydiodes EW 9.74 9.47 LW 9.81 9.52 Picea sitchensis EW 10.58 9.8 0 LW 10.73 9 . 54 Taxus brevifolia EW 10.00 9.65 LW 10.03 9.84 Juniperus virginiana EW 9.80 9.60 LW 10.07 9.96 Cephalotaxus wilsoniana EW 10.20 9.31 LW 10.20 9.31 Pseudotsuga menziesii EW 10.73 9.75 Increment 10 LW 10.01 10.10 increment 20 EW 10.08 10.11 LW 9.98 9.71 Increment 80 EW 9.99 10.07 LW 10.05 9. 94 Increment 150 EW 10.88 10.16 LW 10.98 10.24 Increment 3 00 EW 9.99 9.90 LW 10.01 9.73 Increment 400 EW 10.59 9.92 2 LW 10.11 9.96 Pseudotsuga menziesii EW 10.26 9.52 3 LW 10.28 9.54 Pseudotsuga menziesii EW 10.34 9.43 Increment 20 LW 10.35 9.46 Increment 21 EW 10.31 9.61 LW 10.34 9.64 Increment 22 EW 10.19 9.40 LW 10.18 9.46 (Continued) 143. APPENDIX III (Continued) Earlywood/ Holocellulose Alpha-Cellulose Species Latewood Moisture Moisture Content Q, *5 Content a *o Increment 23 EW 10.18 10.05 LW 10.28 9.86 Increment 24 EW 10.20 9.79 LW 10.18 9.78 Increment 25 EW 10.19 9.88 LW 10.88 9.90 Pseudotsuga menziesii EW 9.85 9.49 Increment 15 LW 9. 86 9.52 Increment 16 EW 9.88 9.53 LW 9.85 9.52 Pseudotsuga menziesii EW 10.47 9.56 Increment 15 LW 10.50 9.58 Increment 16 EW 10.42 9.50 LW 10.48 9.52 Pseudotsuga menziesii EW 10.62 9.64 Increment 15 LW 10.59 9.62 Increment 16 EW 10.58 9.59 LW 10.57 9.61 1. Butt log of a 500-year-old tree 2. Compression wood - Breast '. height disk of a 120-year-old tree 3. NH.NO_ fertilized - Breast height disk of a 29-year-old tree; 4 j fertilized at ages 21 and 22 (500 lb/acre 4. Urea fertilized - 12 mm increment core from breast height, 24-year-old tree; fertilized at ages 13 and 15 (200 lb/acre of N,P20,.orK 0 equivalents) 5. NPK fertilized - 12 mm increment core from breast height, 24-year-old tree; fertilized at ages 13 and 15 (200 lb/acre of-N,P205or K20 equivalents) 6. (NH^)2SO^ fertilized -.12 mm increment core from breast height. 24-year-old tree; fertilized at .ages 13 and 15 (200 lb/acre of N,P20 orK20 equivalents) 

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0075375/manifest

Comment

Related Items